Allele-Specific Histone Acetylation Accompanies Genomic Imprinting ...

0013-7227/00/$03.00/0 Endocrinology Copyright © 2000 by The Endocrine Society

Vol. 141, No. 12 Printed in U.S.A.

Allele-Specific Histone Acetylation Accompanies Genomic Imprinting of the Insulin-Like Growth Factor II Receptor Gene* JI-FAN HU, JUNG PHAM, INDIRAL DEY, TAO LI, THANH H. VU, ANDREW R. HOFFMAN

AND

Medical Service, VA Palo Alto Health Care System, and Division of Endocrinology, Department of Medicine, Stanford University, Palo Alto, California 94304 ABSTRACT The mouse insulin-like growth factor II receptor (Igf2r) gene encodes two reciprocally imprinted RNA transcripts: paternally imprinted Igf2r sense and maternally imprinted Igf2r antisense. Although DNA methylation has been implicated in the initiation and maintenance of genomic imprinting, acetylation of core histones has recently been appreciated as another important factor that regulates gene expression. To determine whether histone acetylation participates in the regulation of Igf2r imprinting, we examined the relative abundance of acetylated histones in interspecific mice (M. spretus ⫻ C57BL/6). Oligonucleosomes derived from liver were immunoprecipitated with acetyl-histone antiserum and were analyzed for the allelic distribution of DNA from the region of the sense and antisense Igf2r

A

LLELE-SPECIFIC DNA methylation is currently the favored mechanism to explain the initiation or maintenance of genomic imprinting. The expression of genes whose promoter regions are heavily methylated is usually decreased. DNA methyltransferase 1 (Dnmt1) can methylate CpG dinucleotides as well as provide maintenance methylation of the symmetrical cytosine in a hemi-methylated doublet (1). In mice that were made deficient in Dnmt 1 by gene targeting, there was no expression from the normally expressed maternal allele of Igf2r sense or from the normally expressed paternal allele of Igf2, whereas the normally imprinted H19 paternal allele was expressed, demonstrating the importance of DNA methylation in the transcription of imprinted genes (2). The degree of core histone acetylation has also been shown to modulate the expression of numerous genes. In general, histone deacetylation leads to transcriptional repression, whereas histone acetylation increases gene transcription. Histone acetylation is maintained during mitosis, so the acetylation pattern represents a heritable epigenetic imprint that can influence gene transcription (3). Thus, the degree of histone acetylation may represent another potential mechanism that could initiate or maintain genomic imprinting. DNA that is rich in methylated CpG is associated with Received July 13, 2000. Address all correspondence and requests for reprints to: Andrew R. Hoffman, M.D., Medical Service, Building 101, Room B2–125, VA Palo Alto Health Care System, 3801 Miranda Avenue, Palo Alto, California 94304. E-mail: [email protected]. * Supported by NIH Grant DK-36054 and by the Research Service of the Department of Veterans Affairs.

promoters. In nucleosomes associated with the Igf2r sense promoter, histone acetylation was demonstrated on the maternal allele, which is transcriptionally active. There was much less histone acetylation on the suppressed paternal allele. In nucleosomes associated with the Igf2r antisense promoter, the active paternal allele was heavily acetylated, whereas the suppressed maternal allele was underacetylated. Treatment of cultured fibroblasts with the histone deacetylase inhibitor Trichostatin A induces partial relaxation of genomic imprinting as well as decreased DNA methylation of both Igf2r sense and antisense promoters. These results demonstrate that increases in histone acetylation can lead to decreased DNA methylation, thereby modulating the regulation of the imprinted expression of Igf2r sense and antisense transcripts. (Endocrinology 141: 4428 – 4435, 2000)

hypoacetylated histone cores and increased histone H1, whereas DNA containing unmethylated CpG islands is associated with chromatin enriched in hyperacetylated histone cores and less histone H1 (4). Although the methylation of DNA may repress transcription by preventing the binding of transcription factors or by enhancing the binding of specific inhibitory proteins (5), it has also been shown that DNA methylation generates an inactive DNase-resistant local chromatin structure with hypoacetylated core histones. Moreover, the inactive chromatin structure generated by the methylated DNA can spread from the methylated area to adjacent nonmethylated DNA, thereby inhibiting gene transcription over a larger segment of the chromosome (6), and thus providing a potential mechanism for creating a region containing clusters of imprinted genes. Recent studies have demonstrated that the methyl-CpG-binding protein MeCP2 is in fact found in a complex with histone deacetylase and other proteins that might regulate transcription. Moreover, TSA, an inhibitor of histone deacetylase, overcomes DNA methylation-induced transcriptional repression, indicating a linkage between these two known mechanisms of transcriptional repression (7, 8). We previously demonstrated that inhibition of histone deacetylation by trichostatin A (TSA) induces the expression of the normally imprinted maternal IGF2 allele, leading to biallelic expression in human and murine cells (9). TSA-treated mouse conceptuses also demonstrated loss of H19 imprinting (10). In conjunction with DNA methylation, histone acetylation may represent a crucial molecular mechanism for initiating, maintaining and/or transmitting the genomic imprint. The reciprocally imprinted Igf2r sense and antisense tran-

4428

Igf2r IMPRINTING AND HISTONE ACETYLATION

scripts provide a convenient model for examining the molecular mechanisms underlying the imprinting process as the tissue-specific methylation imprint and transcript expression have been well delineated. The expression of the paternally imprinted Igf2r sense gene correlates with the methylation of the differentially methylated region (DMR) in its promoter region (region 1) (11–13). In peripheral tissues, where Igf2r sense is expressed only from the maternal allele, region 1 methylation is found only on the paternal allele. In the CNS, where biallelic expression is seen, neither allele demonstrates CpG methylation in this region (13). Igf2r antisense RNA is expressed only from the paternal allele in both peripheral tissues and CNS. The paternally derived antisense DNA sequence is unmethylated in its promoter region, DMR region 2, a CpG island that is methylated on the suppressed maternal allele (11, 13). We hypothesized that nonmethylated, nonimprinted alleles would preferentially be associated with acetylated histones. Furthermore, we proposed that the methylation status of an imprinted gene would be altered by modulating the acetylation of core histones. Therefore, we have studied the status of histone H4 acetylation and DNA methylation in the promoter regions of Igf2r sense and Igf2r antisense, where the genomic DNA has been shown to be differentially methylated (11, 13). Materials and Methods Interspecific mice F1 mice were generated from a cross between M. spretus males and C57BL/6 females, as previously described (14, 15). Female F1 mice were also crossed with C57BL/6 males to produce backcross mice. Backcross mice that were informative for Igf2r were killed at 2 months of age, and livers from three mice were collected for histone acetylation analysis.

Chromosome immunoprecipitation All of the following steps were carried out at 4 C in the presence of 5 mm sodium butyrate. Fresh livers were collected in PBS and minced into small pieces with scissors. Tissues were gently homogenized in 10 ml of PBS into single cells in a glass tissue grinder with a loose pestle. Cells were centrifuged at 600 ⫻ g for 10 min, and washed twice with PBS. Nuclei from these cells were prepared by the method described by O’Neill et al. (16). Briefly, liver cells were suspended in buffer A (10 mm Tris-HCI, pH 7.5, 3 mm CaCl2, 2 mm MgCl2, 1.5% Triton 100, and 0.1 m PMSF) and were homogenized with a tight glass pestle in a glass tissue grinder to exclude the nuclei. Nuclei were pelleted by centrifugation at 600 ⫻ g for 20 min, and were applied to 25% and 50% sucrose gradients in TBS (10 mm Tris-HCI, pH 7.5, 3 mm CaCl2, 2 mm MgCl2). Nuclei were collected by centrifugation (1,500 ⫻ g, 4 C, 20 min) and washed once in 25% sucrose. Oligonucleosomes were released from nuclei by mild digestion using micrococcal nuclease (16). Nuclear extracts (equivalent to 20 ␮g of DNA nuclei) were digested with 0.5–5 U of micrococcal nuclease (Amersham Pharmacia Biotech, Piscataway, NJ) for 15 min in buffer B (0.32 m sucrose, 50 mm Tris-HCl, pH 7.5, 4 mm MgCl2, 1 mm CaCl2, 0.1 mm PMSF, 5 mm sodium butyrate), and dialyzed overnight in lysis buffer (1 mm Tris-HCI, pH 7.4, 0.2 mm Na2 EDTA, 0.2 mm PMSF, 5 mm sodium butyrate). After centrifugation at 11,600 ⫻ g at 4 C for 20 min, an aliquot of the supernatant nucleosome DNA was extracted with phenol/chloroform and run on 2% agarose gel. An appropriate micrococcal nuclease digestion of chromatin produced a range of 1–5 oligonucleosomes. Oligonucleosomes (10 ␮g) were then incubated with antiserum to acetylated histone H4 antiserum (0 – 4 ␮l) (Serotec Inc., Raleigh, NC). Acetylated-histone nucleosomes were separated from unacetylatedhistone nucleosomes by precipitation with protein A-Sepharose (Amersham Pharmacia Biotech, Piscataway, NJ). DNA associated with acety-

4429

lated histones was extracted from nucleosomes with phenol/chloroform and was precipitated with ethanol for PCR analysis.

Allele-specific histone acetylation Allele-specific histone acetylation in Igf2r promoter regions was analyzed by PCR using specific Igf2r polymorphisms (13). Specifically, DNA associated with acetyl-H4 antiserum-immunoprecipitated oligonucleosomes was used as template for PCR analysis as previously described (14, 17). Briefly, immunoprecipitated DNA from Igf2r promoter regions was amplified in a 2.5-␮l reaction mixture in the presence of 50 ␮m dNTP, 0.2 ␮m Igf2r primers, 0.25 ␮Ci ␣-dCTP (Amersham Pharmacia Biotech), 0.125 U Tfl DNA polymerase (Epicentre Technologies, Madison, WI). DNA was amplified for 35 cycles at 94 C for 15 sec, 65 C for 40 sec, followed by a 30-sec extension at 72 C. The PCR products were diluted and digested with polymorphic restriction enzymes, and then separated on 5% polyacrylamide-urea gel to examine allelic distribution. For the Igf2r antisense promoter region, two restriction enzymes (Fok1 and XhoI) (13) were used to distinguish the parental alleles. Fok1 specifically digests the M. spretus allele, whereas XhoI cuts only the C57BL allele. For the Igf2r sense RNA promoter, a 16-bp deletion polymorphism (13) was used to separate the two parental alleles (Fig. 1). The oligonucleotide primers used for examining allelic histone acetylation were: 1) Fok1: no. 6282 (5⬘-primer): TCCCTTTCCCTCACCGCAACTCAG; no. 6188 (3⬘-primer): AGCACAACTCCAATTGTGCTGCGAT; 2) XhoI: no. 6647 (5⬘-primer): CGTGTAGTTCAGAACACTGGTGAGC; no. 6460 (3⬘-primer): TACGCGAGGTGAGGGTTCCACTGAT; and 3) 16-bp polymorphism: no. 6458 (5⬘-primer): GGTGCTGGACGGGGAAACTGAGGT; no. 8 – 41 (3⬘-primer): CCAGTCCCGGGTCACATGAGCATCG. To assess the abundance of the two parental alleles, primers (nos. 6188, 6647, and 8 – 41) used in the PCR amplification were end-labeled with [32P-␥] ATP (Amersham Pharmacia Biotech, Arlington Heights, IL) using T4 polynucleotide kinase (New England Biolabs, Inc., Beverly, MA).

Treatment of fibroblasts (C6W and C8W) with histone deacetylase inhibitor Fibroblasts were cultured from the skin of F1 mice (M. spretus males ⫻ C57BL/6 females) as previously described (9). Tissues were minced with scissors into small pieces and suspended in DMEM (Life Technologies, Inc., Gaithersburg, MD), supplemented with 15% FBS and 100 U/ml of penicillin and 100 ␮g/ml of streptomycin, and grown at 37 C with 5% CO2. The medium was replaced with fresh media 24 h after plating. When confluent, the fibroblasts were trypsinized and used for the treatment. C6W and C8W fibroblasts at passages 3–5 were seeded in 6-well plates at a density of 2 ⫻ 105 cells/ml and were treated with the histone deacetylase inhibitor, trichostatin A (TSA, Wako BioProducts, Richmond, VA) as previously described (9). The cells were randomly assigned into treatment and control groups. In treatment wells, the culture medium was replaced with media containing TSA (0.5 ␮m, 2 ␮m, and 6 ␮m) 24 h after seeding. Higher concentrations of TSA are potentially toxic to cells. For control wells, the cells were supplied with DMEM supplemented with the same volume of PBS as used in the treatment wells. After overnight treatment, the medium containing drugs was removed, and the cells were washed twice with PBS. The cells were allowed to grow in fresh media with FBS until they became confluent.

Measurement of Igf2r allelic expression Confluent fibroblasts from treatment and control wells were directly lysed with 0.6 ml Tri-reagent solution provided by the Tri-reagent kit (Sigma, St. Louis, MO). To avoid DNA contamination, RNA was digested with DNase I before cDNA was synthesized with RNA reverse transcriptase (9, 17). Igf2r allelic expression was analyzed by the same PCR condition as described above. The PCR products were diluted and digested with 1.0 U HaeIII for Igf2r sense RNA. For Igf2r antisense RNA, a 16-bp polymorphism was used to separate the two parental alleles (13). After restriction enzyme digestion, PCR products were separated on 5% polyacrylamide-urea gel to assess the allelic expression of Igf2r. The oligonucleotide primers used for examining sense Igf2r RNA

4430


Endo • 2000 Vol. 141 • No. 12

FIG. 1. Igf2r sense (pS) and antisense (pAS) promoters, PCR primers, and allelic polymorphisms. Asterisk-labeled primers are end-labeled with [32p-␥]ATP by T4 polynucleotide kinase. Except for PCR products derived from primer pairs no. 6287/no. 8 – 04 and no. 6194/no. 8 – 04, which have a 16 bp polymorphism, the two parental alleles are separated on 5% urea-polyacrylamide gel after polymorphic restriction enzyme digestion.

imprinting were: no. 6105 (5⬘-primer): CAGAAGAAGCTCGGGCGTGTCCTAC and no. 6294 (3⬘-primer): CTCCGCTCCTCGGCCTGAGTGAACT; and antisense Igf2r RNA imprinting were: no. 8 – 04 (5⬘-primer): GGCACGAGCGCCAGGTACCTACTCGA and no. 6458 (3⬘-primer). Primers (nos. 6105, and 8 – 04) were end-labeled with [32P-␥] ATP by T4 polynucleotide kinase (New England Biolabs, Inc.) for PCR amplification. Allelic expression was quantitated by PhosphorImager Analyzer (Molecular Dynamics, Inc., Sunnyvale, CA).

DNA methylation analysis DNA methylation in the regions of Igf2r sense and antisense promoters was examined by the DNA-sensitive restriction enzyme, HpaII, as previously described (11). Briefly, 2 ␮g of genomic DNA was digested by 6 U HpaII in a 10 ␮l reaction covered with liquid wax at 37 C overnight. The digested DNA was diluted and directly amplified by PCR. When genomic DNA is methylated, HpaII will not digest the DNA and PCR will amplify the DNA at full-length when separated on 5% polyacrylamide-urea gel. However, unmethylated DNA will be completely digested by HpaII and will not be amplified by PCR. Thus, PCR will provide a reliable quantitation of DNA methylation in a designated specific region (18). PCR conditions for DNA methylation quantitation were the same as for the quantitation of Igf2r allelic histone acetylation. After PCR, the amplified DNAs were electrophoresed on 5% polyacrylamide-urea gel and were exposed to the screen of a PhosphorImager Scanner (Molecular Dynamics, Inc., Sunnyvale, CA) for quantitation. As a PCR control, we also used the MspI restriction enzyme to digest genomic DNA. MspI is a DNA methylation-insensitive restriction enzyme, and it will digest DNA whether it is methylated or unmethylated. Furthermore, we also used a set of PCR primers that amplify a DNA region that does not contain any HpaII sites (CCpGG). This pair of PCR

primers will give an amplified PCR band whether genomic DNA is digested by HpaII or not. The oligonucleotide primers used for Igf2r sense RNA promoter were: no. 6194 (5⬘-primer): GTCCACCAGTCACCTTACATGCTGTA and no. 8 – 41 (3⬘-primer); 2) Igf2r antisense RNA promoter: no. 6647 (5⬘-primer) and no. 6460 (3⬘-primer); 3) Non-CCpGG PCR primers: no. 3302 (5⬘primer): GGCCAAACGTCATCGTCCCCTGAT, no. 3303 (3⬘-primer): CTGTCCCGCTCAAGAGGAGGTCA. Primer set for Igf2r antisense RNA promoter (nos. 6647 and 6460) covers one HapII site (CCpGG) located between the allele-discrimination signal and the de novo methylation signal in the imprinting box, which is critical for the establishment of Igf2r genomic imprinting (19). Primers (nos. 6647, 6194, and 3302) in the PCR amplification were end-labeled with [32P-␥]ATP by T4 polynucleotide kinase (New England Biolabs, Inc.). The conditions for PCR amplification were the same as those used for examining genomic imprinting as described above.

Results Differential histone acetylation in two parental alleles

To examine the role of histone acetylation on Igf2r imprinting, we used antiserum raised against acetylated histone H4 to immunoprecipitate oligonucleosomes derived from the mild digestion of liver cell nuclei using streptococcal nuclease. DNA was extracted from immunoprecipitated nucleosomes (i.e. acetylated histone-linked) and was analyzed for the allelic distribution of Igf2r sense and antisense using PCR primers derived from their respective promoter regions. Acetyl-H4 antiserum-immunoprecipitated DNA was am-


plified with PCR primers that are specific to Igf2r sense and antisense promoter regions (Fig. 1). The DNA was then digested with restriction enzymes which cut in sites that are polymorphic in C57BL/6 ⫻ M. spretus interspecific mice. As seen in Fig. 2A, in normal genomic DNA, PCR amplification yields equal amounts of the two parental Igf2r sense alleles (lane 4). However, when oligonucleosomes were immunoprecipitated with varying amounts of acetyl-histone H4 antiserum, Igf2r sense PCR products were shown to be derived primarily from the expressed maternal (M. spretus) allele (lanes 5– 8). The paternal allele, which is transcriptionally FIG. 2. Differential histone acetylation of Igf2r alleles. Oligonucleosomes prepared from mild micrococcal nuclease digestion of chromatin from livers of interspecific mice were immunoprecipitated with antiacetylated histone serum (0.0 – 4.0 ␮l). DNA was extracted from immunoprecipitated nucleosomes and Igf2r sense and antisense promoter regions were amplified with Igf2r promoter-specific primers. PCR products were digested with polymorphic restriction enzymes and analyzed by PhosphorImager Analyzer. A, Igf2r DMR region 1 (sense promoter). Use of a 16 bp deletion polymorphism. Lane 1, DNA ladder; lane 2, 162-bp fragment representing the C57BL (paternal, imprinted) allele. Lane 3, 146-bp fragment representing the M. spretus (maternal, expressed) allele. Lane 4, Equal amounts of maternal and paternal DNA from interspecific F1 mice. Lanes 5– 8, PCR products derived from DNA incubated with increasing amounts of antiacetylated histone antiserum. Most but not all of the acetylated histones are associated with the expressed maternal allele. B, Igf2r DMR region 2 (antisense promoter). Use of the Xho polymorphism. Lane 1, DNA ladder; lane 2, 169-bp fragment representing the M. spretus (maternal, imprinted) allele. Lane 3, 133-bp fragment representing the C57BL/6 (paternal, expressed) allele. Lane 4, Equal amounts of maternal and paternal DNA from interspecific F1 mice. Lanes 5– 8, PCR products derived from DNA incubated with increasing amounts of antiacetylated histone antiserum. Most but not all of the acetylated histones are associated with the expressed paternal allele. C, Igf2r DMR region 2 (antisense promoter). Use of the Fok 1 polymorphism. Lane 1, DNA ladder; lane 2; 196-bp fragment representing the C57BL (paternal, expressed) allele. Lane 3, 165-bp fragment representing the M. spretus (maternal, imprinted) allele. Lane 4, Equal amounts of maternal and paternal DNA from interspecific F1 mice. Lanes 5– 8, PCR products derived from DNA incubated with increasing amounts of antiacetylated histone antiserum. Most but not all of the acetylated histones are associated with the expressed paternal allele.

4431

suppressed, only accounts for a small portion of the genomic DNA that is associated with the acetylated histones. Thus, histone H4 is relatively hyperacetylated in the region of the expressed maternal allele, and is hypoacetylated in the region of the suppressed paternal allele. We also used this method to examine allele-specific histone acetylation of the Igf2r antisense promoter region. By DNA sequencing, we have identified two polymorphisms in this region (XhoI and Fok 1) between C57BL/6 and M. spretus mice (13, 20). XhoI is specific to the C57BL/6 allele (Fig. 2B) and Fok1 digests only the M. spretus allele

4432


Endo • 2000 Vol. 141 • No. 12

(Fig. 2C). After PCR amplification and restriction enzyme digestion, allelic histone acetylation in the Igf2r antisense promoter region can be easily visualized in polyacrylamide-urea gel. Histones are highly acetylated in the C57BL/6 allele (lanes 5– 8, Fig. 2, B and C), which represents the expressed paternal allele (13). Similar results were obtained using either Fok1 or XhoI polymorphic restriction enzymes. Histones associated with the imprinted maternal (M. spretus) allele are hypoacetylated. Thus, histone hyperacetylation is highly associated with the expressed alleles of Igf2r sense and antisense, whereas histone under-acetylation is associated with the imprinted alleles, suggesting that histone acetylation is intimately involved in the regulation of Igf2r imprinting.

lase inhibitor trichostatin A (TSA). Igf2r sense mRNA is derived from the maternal allele in all tissues except brain (13). In cultured mouse fibroblasts, the maternal imprinting of Igf2r sense mRNA is fully maintained (Fig. 3A, lanes 4 and 5). TSA treatment releases the imprinting of Igf2r sense mRNA transcripts to a small degree, but with much less efficacy than the from antisense RNA promoter (Fig. 3B and 3C, lanes 6 –11). Both C8W and C6W fibroblasts express the paternal (C57BL) but not the maternal (spretus) allele from the antisense promoter (region 2) (Fig. 3, B and C, lanes 2 and 3). After TSA treatment, however, the imprinted Igf2r antisense allele becomes expressed, although the loss of imprinting is not complete.

Relaxation of Igf2r imprinting following TSA treatment

Alteration of DNA methylation after treatment with histone deacetylase inhibitor

To further assess the importance of histone acetylation in Igf2r imprinting, we treated two interspecific F1 mouse fibroblast cell lines (C8W and C6W) with the histone deacety-

FIG. 3. Allelic expression of Igf2r sense and antisense RNAs in fibroblasts treated with the histone deacetylase inhibitor trichostatin A (TSA). A, Igf2r sense RNA in C6W fibroblasts. Lane 1, DNA ladder; lanes 2–3, liver cDNAs from C57BL and M. spretus; lanes 4 –5, control cells (0% relaxation of imprinting); lanes 6 –7, incubation with 0.5 ␮M TSA (7.6% relaxation of imprinting); lanes 8 –9, incubation with 2.0 ␮M TSA (8.5% relaxation of imprinting); lanes 10 –11, incubation with 6.0 ␮M TSA (19% relaxation of imprinting). B, Igf2r antisense RNA in C8W fibroblasts. Lane 1, DNA ladder. Lanes 2–3, control cells (0% relaxation of imprinting); lanes 4 –5, incubation with 0.5 ␮M TSA (9.0% relaxation of imprinting); lanes 6 –7: incubation with 2.0 ␮M TSA (21.9% relaxation of imprinting); lanes 8 –9, incubation with 6.0 ␮M TSA (30.4% relaxation of imprinting). C, Igf2r antisense RNA in C6W fibroblasts. Lane 1, DNA ladder; lanes 2–3: control cells (0% relaxation of imprinting); lanes 4 –5, incubation with 0.5 ␮M TSA (13.0% relaxation of imprinting); lanes 6 –7, incubation with 2.0 ␮M TSA (39.7% relaxation of imprinting); lanes 8 –9, incubation with 6.0 ␮M TSA (14.9% relaxation of imprinting).

To examine the interactions of histone acetylation and DNA methylation, we also examined the extent of DNA


methylation in fibroblasts that were treated with the histone deacetylase inhibitor TSA. DNA extracted from fibroblasts treated with varying concentrations of TSA were digested with the DNA methylation-sensitive restriction enzyme HpaII and the DNA methylation-insensitive restriction enzyme MspI. Methylated DNA cannot be digested by HpaII and thus can be amplified by PCR primers that encompass the regions of Igf2r sense and antisense RNA promoters. Unmethylated DNA will be digested, and will not be amplified in the PCR. In control samples, PCR primers amplified a 402-bp band from the Igf2r sense promoter region (Fig. 4A, lane 2) and a 169-bp band from Igf2r antisense RNA promoter region (Fig. 4B, lane 2). However, after fibroblasts were treated with TSA, the density of the PCR bands decreased in a dose-dependent manner (lanes 3–5), indicating a decrease in DNA methylation in that region. These results indicate that inhibition of histone deacetylase partially induces DNA demethylation in both Igf2r sense and antisense RNA promoter regions. As experimental controls, the same DNA samples were also digested with MspI, a DNA methylation insensitive restriction enzyme. As predicted, there was no PCR amplification from those MspI-digested control samples. We also amplified the same DNA samples with a set of control PCR primers that do not cover any CpG sites (Fig. 4C). There were no significant changes in the density of PCR bands between control and TSA-treated samples.

FIG. 4. DNA methylation in Igf2r sense and antisense promoter regions after treatment with the histone deacetylase inhibitor TSA. Cultured fibroblasts were incubated 0.0 – 6.0 ␮M TSA for 12 h and the DNA was then extracted and subjected to PCR and restriction enzyme analysis. For comparison, the density of PCR products of HapIIdigested DNA was also scanned by PhosphorImager and plotted in bar graph.

4433

Discussion

The mouse insulin-like growth factor-II receptor (Igf2r) gene was among the first genes that were shown to be imprinted. Multiple partially overlapping mechanisms have been adduced to explain the regulation of monoallelic imprinted expression, including differentially methylated regions of DNA, antisense competition, and the existence of an imprinting box. Because the gene expresses two transcripts that are reciprocally imprinted in a tissue-specific manner (11, 13), the Igf2r gene represents an excellent model to examine the mechanisms underlying the silencing of imprinted alleles. Previously, we and others (11) (13) have shown that Igf2r sense expression correlates with methylation in DMR1 in the promoter region and that antisense expression is always from the paternal allele, which is hypomethylated in DMR2 in its promoter region. In this study, we show that the expressed alleles of both the sense and the antisense transcripts are associated with histones that are more acetylated than is the chromatin, which is associated with the imprinted alleles. This is in accordance with earlier reports that acetylated histones are associated with gene expression. The positive charges on the lysine moieties of histones are able to bind tightly to DNA, preventing trans-activating proteins like transcription factors from gaining access to regulatory sites on the DNA. Acetylation of these lysine residues can neutralize the charges, releasing the histone from the DNA and thereby permitting gene regulators free access. Recently,

4434


Pedone et al. (21) studied another pair of reciprocally imprinted genes, and demonstrated that the expressed Igf2 and H19 alleles were associated with more abundantly acetylated histones than were the imprinted alleles. When cells are cultured in the presence of the histone deacetylase inhibitor trichostatin A, the pattern of Igf2r gene expression changes, as the relative concentration of acetylated histones increases. TSA treatment causes an increase in Igf2r antisense expression from the normally imprinted paternal allele. Although the expression of the imprinted Igf2r sense paternal allele is also increased by TSA, the relaxation of imprinting from the sense allele is far less extensive than is the loss of imprinting of Igf2r antisense. However, TSA treatment did not lead to total relaxation of genomic imprinting. It is likely that TSA exposure does not lead to complete histone acetylation, and it is probable that mechanisms other than histone acetylation and DNA methylation play an important role in maintaining suppression of the imprinted allele. We have previously shown that, whereas TSA treatment of cultured human cells leads to the loss of IGF2 imprinting (9), H19 imprinting remains intact (unpublished data), without any relaxation and expression of the paternal allele. These data indicate that altering the acetylation status of core histones has differential, gene-specific and transcript-specific effects, suggesting that the expression of some transcripts are more tightly controlled by chromatin structure than are others. Pedone et al. (21) showed that in the mouse, TSA had no effect on H19 imprinting, whereas it did lead to biallelic expression of Igf2. Moreover, they demonstrated that treatment with both TSA and the DNA methylation inhibitor 5-aza-2-deoxycytidine did lead to loss of H19 imprinting. Although Trichostatin A can reactivate some of the suppressed methylated genes in Neurospora, other methylated regions remain suppressed. In examining the synergy of DNA demethylation and histone deacetylase inhibition in the reexpression of genes silenced in cancer, Cameron and colleagues (22) showed that some hypermethylated genes cannot be transcriptionally reactivated with TSA alone. Following incubation with a low dose 5-aza-2-deoxycytidine that did not significantly result in demethylation, TSA treatment resulted in the expression of each repressed gene. However, TSA induced no further change in the methylation status of these newly expressed genes. The interrelationship between DNA methylation and histone acetylation has proven to be extremely intimate. The ability of methylated DNA to bind to methyl-CpG-binding proteins (MeCP), which then recruit histone deacetylases has been demonstrated in several systems. MeCP2 possesses a transcriptional repressor domain that binds the corepressor mSin3A, which is associated with histone deacetylases (HDAC) (7). HDAC1 binds to Dnmt1, the enzyme responsible for CpG methylation, suggesting that the modifications of DNA and histones are very tightly linked (23). In our study, we incubated cells with TSA and then estimated the amount of DNA methylation using methylation-sensitive restriction enzymes. We show that TSA led to the demethylation of Igf2r DNA, providing further proof of the interdependence of DNA methylation and histone acetylation.

Endo • 2000 Vol. 141 • No. 12

Selker (24) also demonstrated in Neurospora crassa that TSA treatment selectively demethylated and thus reactivated several genes that were originally suppressed by DNA hypermethylation. Despite the fact that TSA led to equivalent demethylation of the Igf2r sense and antisense alleles, the histone deacetylase inhibitor had significantly different effects on the expression of the sense and antisense transcripts in the same cell, stimulating the relaxation of antisense imprinting to a far greater extent than the relaxation of sense transcript imprinting. This phenomenon suggests that factors other than DNA methylation and histone acetylation regulate the transcription of the imprinted allele. It is likely in this case that the TSA-induced expression of the Igf2r maternal antisense transcript itself inhibited the relaxation of maternal Igf2r sense imprinting, in accordance with the antisense regulation of imprinting reported by Barlow and colleagues (12, 25). Acknowledgments We thank the Central Laboratory for Human Embryology Tissue, University of Washington for human fetal tissues. We also thank Dr. Rosemary Broom and Mrs. Marta Raygoza for their technical help with animal breeding.

References 1. Gruenbaum Y, Cedar H, Razin A 1982 Substrate and sequence specificity of a eukaryotic DNA methylase. Nature 295:620 – 622 2. Li E, Beard C, Jaenisch R 1993 Role for DNA methylation in genomic imprinting. Nature 366:362–365 3. Jeppesen P 1997 Histone acetylation: a possible mechanism for the inheritance of cell memory at mitosis. Bioessays 19:67–74 4. Tazi J, Bird A 1990 Alternative chromatin structure at CpG islands. Cell 60:909 –920 5. Boyes J, Bird A 1991 DNA methylation inhibits transcription indirectly via a methyl-CpG binding protein. Cell 64:1123–1134 6. Kass SU, Goddard JP, Adams RL 1993 Inactive chromatin spreads from a focus of methylation. Mol Cell Biol 13:7372–7379 7. Nan X, Ng HH, Johnson CA, Laherty CD, Turner BM, Eisenman RN, Bird A 1998 Transcriptional repression by the methyl-CpG-binding protein MeCP2 involves a histone deacetylase complex. Nature 393:386 –389 8. Jones PL, Veenstra GJ, Wade PA, Vermaak D, Kass SU, Landsberger N, Strouboulis J, Wolffe AP 1998 Methylated DNA and MeCP2 recruit histone deacetylase to repress transcription. Nat Genet 19:187–191 9. Hu JF, Oruganti H, Vu TH, Hoffman AR 1998 The role of histone acetylation in the allelic expression of the imprinted human insulin-like growth factor II gene. Biochem Biophys Res Commun 251:403– 408 10. Svensson K, Mattsson R, James TC, Wentzel P, Pilartz M, MacLaughlin J, Miller SJ, Olsson T, Eriksson UJ, Ohlsson R 1998 The paternal allele of the H19 gene is progressively silenced during early mouse development: the acetylation status of histones may be involved in the generation of variegated expression patterns. Development 125:61– 69 11. Stoger R, Kubicka P, Liu CG, Kafri T, Razin A, Cedar H, Barlow DP 1993 Maternal-specific methylation of the imprinted mouse Igf2r locus identifies the expressed locus as carrying the imprinting signal. Cell 73:61–71 12. Wutz A, Smrzka OW, Schweifer N, Schellander K, Wagner EF, Barlow DP 1997 Imprinted expression of the Igf2r gene depends on an intronic CpG island (see comments). Nature 389:745–749 13. Hu JF, Oruganti H, Vu TH, Hoffman AR 1998 Tissue-specific imprinting of the mouse insulin-like growth factor II receptor gene correlates with differential allele-specific DNA methylation. Mol Endocrinol 12:220 –232 14. Hu J, Vu T, Hoffman A 1995 Differential biallelic activation of three insulinlike growth factor II promoters in the mouse central nervous system. Mol Endocrinol 9:628 – 636 15. Hu JF, Nguyen PH, Pham NV, Vu TH, Hoffman AR 1997 Modulation of Igf2 genomic imprinting in mice induced by 5-azacytidine, an inhibitor of DNA methylation. Mol Endocrinol 11:1891–1898 16. O’Neill LP, Turner BM 1995 Histone H4 acetylation distinguishes coding regions of the human genome from heterochromatin in a differentiationdependent but transcription-independent manner. EMBO J 14:3946 –3957 17. Hu JF, Vu TH, Hoffman AR 1996 Promoter-specific modulation of insulin-like


18. 19. 20.

21.

growth factor II genomic imprinting by inhibitors of DNA methylation. J Biol Chem 271:18253–18262 Herman JG, Graff JR, Myohanen S, Nelkin BD, Baylin SB 1996 Methylationspecific PCR: a novel PCR assay for methylation status of CpG islands. Proc Natl Acad Sci USA 93:9821–9826 Birger Y, Shemer R, Perk J, Razin A 1999 The imprinting box of the mouse Igf2r gene. Nature 397:84 – 88 Hu JF, Balaguru KA, Ivaturi RD, Oruganti H, Li T, Nguyen BT, Vu TH, Hoffman AR 1999 Lack of reciprocal genomic imprinting of sense and antisense RNA of mouse insulin-like growth factor II receptor in the central nervous system. Biochem Biophys Res Commun 257:604 – 608 Pedone PV, Pikaart MJ, Cerrato F, Vernucci M, Ungaro P, Bruni CB, Riccio

22. 23. 24. 25.

4435

A 1999 Role of histone acetylation and DNA methylation in the maintenance of the imprinted expression of the H19 and Igf2 genes. FEBS Lett 458:45–50 Cameron EE, Bachman KE, Myohanen S, Herman JG, Baylin SB 1999 Synergy of demethylation and histone deacetylase inhibition in the re-expression of genes silenced in cancer. Nat Genet 21:103–107 Fuks F, Burgers WA, Brehm A, Hughes-Davies L, Kouzarides T 2000 DNA methyltransferase dnmt1 associates with histone deacetylase activity. Nat Genet 24:88 –91 Selker EU 1998 Trichostatin A causes selective loss of DNA methylation in Neurospora. Proc Natl Acad Sci USA 95:9430 –9435 Barlow DP 1997 Competition—a common motif for the imprinting mechanism? EMBO J 16:6899 – 6905