MCB
Home Help [Feedback] [For Subscribers] [Archive] [Search] [Contents]
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Thorvaldsen, J. L.
Right arrow Articles by Bartolomei, M. S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Thorvaldsen, J. L.
Right arrow Articles by Bartolomei, M. S.

 Previous Article  |  Next Article 

Molecular and Cellular Biology, February 2006, p. 1245-1258, Vol. 26, No. 4
0270-7306/06/$08.00+0     doi:10.1128/MCB.26.4.1245-1258.2006
Copyright © 2006, American Society for Microbiology. All Rights Reserved.

Developmental Profile of H19 Differentially Methylated Domain (DMD) Deletion Alleles Reveals Multiple Roles of the DMD in Regulating Allelic Expression and DNA Methylation at the Imprinted H19/Igf2 Locus

Joanne L. Thorvaldsen, Andrew M. Fedoriw, Son Nguyen, and Marisa S. Bartolomei*

Department of Cell and Developmental Biology, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19104

Received 28 September 2005/ Returned for modification 3 November 2005/ Accepted 25 November 2005


   ABSTRACT
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
The differentially methylated domain (DMD) of the mouse H19 gene is a methylation-sensitive insulator that blocks access of the Igf2 gene to shared enhancers on the maternal allele and inactivates H19 expression on the methylated paternal allele. By analyzing H19 DMD deletion alleles H19{Delta}DMD and H19{Delta}3.8kb-5'H19 in pre- and postimplantation embryos, we show that the DMD exhibits positive transcriptional activity and is required for H19 expression in blastocysts and full activation of H19 during subsequent development. We also show that the DMD is required to establish Igf2 imprinting by blocking access to shared enhancers when Igf2 monoallelic expression is initiated in postimplantation embryos and that the single remaining CTCF site of the H19{Delta}DMD allele is unable to provide this function. Furthermore, our data demonstrate that sequence outside of the DMD can attract some paternal-allele-specific CpG methylation 5' of H19 in preimplantation embryos, although this methylation is not maintained during postimplantation in the absence of the DMD. Finally, we report a conditional allele floxing the 1.6-kb sequence deleted from the H19{Delta}DMD allele and demonstrate that the DMD is required to maintain repression of the maternal Igf2 allele and the full activity of the paternal Igf2 allele in neonatal liver.


   INTRODUCTION
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Genomic imprinting is a mammalian epigenetic phenomenon whereby the parental origin of a gene determines whether or not it will be expressed. Over 75 imprinted genes have been identified, many of which are noncoding RNAs that are hypothesized to control the expression of linked protein coding genes that are also imprinted (59). Imprinted genes tend to be clustered with other imprinted genes with which they sometimes share common tissue-specific expression patterns. While an understanding of the mechanism by which imprinting occurs is dependent upon the identification of cis-acting control elements, few elements have been thus far characterized. As such, tissue-specific enhancers have only been elucidated for the imprinted locus encompassing the oppositely imprinted H19 and Igf2 genes. In addition to shared enhancers, most imprinting clusters harbor an imprinting control region (ICR) that governs the imprinting of the entire cluster.

ICRs are associated with sequences that carry imprinting marks (IMs), which are epigenetic modifications that distinguish the parental alleles. By definition, IMs are established in the germ line and maintained throughout development. To date, allele-specific methylation of CpG dinucleotides qualifies as an IM. At the H19/Igf2 locus, the ICR is associated with the paternally methylated 2-kb differentially methylated domain (DMD) upstream of H19 (54, 55). The DMD, which is located 2 kb 5' of H19 and approximately 90 kb 3' of Igf2, is a CTCF-dependent, methylation-sensitive insulator that blocks enhancer-driven Igf2 expression on the hypomethylated maternal allele and functions as a silencing element to repress H19 expression on the hypermethylated paternal allele (3, 15, 20, 59). Consistent with its function as an ICR, targeted deletions of sequence spanning the DMD result in loss of both Igf2 and H19 imprinted expression (19, 43, 51, 52). To date, the insulator model of imprinting control remains unique to the H19/Igf2 locus, although other ICRs are expected to bind CTCF and/or exhibit insulator activity in vitro, including the KvDMR1 at the Kcnq1 locus (8, 11, 14).

While targeted deletions of the DMD have suggested that at least three of the four CTCF binding sites are necessary for Igf2 and H19 imprinting (10, 51), DMD replacement alleles have been essential in validating the role of CTCF in regulating DMD function. The DMD replacement alleles with mutations that block CTCF binding perturb imprinting exclusively when maternally inherited: Igf2 becomes biallelically expressed due to loss of CTCF-dependent DMD insulator activity (31, 41, 47). In contrast, a DMD replacement allele that allowed CTCF binding on the paternal allele blocked enhancer-dependent Igf2 expression (12). While loss of parental-allele-specific methylation of the DMD in postimplantation embryos was reported for all of these alleles, paternal-allele-specific methylation of the mutant DMDs was detected in spermatozoa. These results suggest that the sequence responsible for establishing paternal-allele-specific methylation has yet to be defined. Furthermore, H19 and Igf2 expression were not analyzed in blastocysts and early postimplantation embryos in these studies to determine when loss of imprinting first occurs.

Our previous study of the H19{Delta}DMD and H19{Delta}3.8kb-5'H19 deletion alleles in neonatal tissues indicated that the mutations disrupted H19 and Igf2 imprinting to a similar extent, even though the H19{Delta}3.8kb-5'H19 allele also removes the 5' CTCF site (52). We and others (19) additionally showed that deleting the DMD from the maternal allele perturbs H19 expression in a tissue-specific manner. Similar defects were observed in one of the DMD replacement alleles that cannot bind CTCF, further suggesting that the DMD has a role in the transcriptional activation of H19, potentially through the activity of CTCF (41). This is consistent with CTCF's proposed role as a transcriptional activator (30). Our previous analysis did not, however, address the role of the DMD in establishment and maintenance of H19 and Igf2 imprinting.

Here we perform a developmental profile of allele-specific H19 and Igf2 expression and DNA methylation of the H19{Delta}DMD and H19{Delta}3.8kb-5'H19 alleles in pre- and postimplantation embryos. Our goals were to determine (i) when loss of imprinted expression is first detected, (ii) whether the extra sequence deleted from the H19{Delta}3.8kb-5'H19 allele, which includes the 5' CTCF site, affects H19 and Igf2 expression in the early embryo, and (iii) when parental-allele-specific methylation of the remaining 5' H19 sequence is lost at these deletion alleles. We found that the H19 DMD provides a crucial role for H19 and Igf2 imprinted expression throughout early development. We observed differences in H19 expression level from the H19{Delta}DMD and H19{Delta}3.8kb-5'H19 alleles, suggesting that additional sequence contributing to H19 expression was removed in the larger deletion. We detected parental-allele-specific methylation of sequence adjacent to both the H19{Delta}DMD and H19{Delta}3.8kb-5'H19 deletions in the germ line and blastocyst that was not maintained in the postimplantation embryos, indicating that the presence of a germ line-inherited DMD is not necessary to attract but is necessary to maintain allele-specific methylation at this locus. We also generated a conditional floxed DMD allele with loxP sites that flank the 1.6-kb DMD sequence that is deleted from the H19{Delta}DMD allele to determine if imprinting is maintained when the DMD is deleted in neonatal liver. While imprinting of H19 was not significantly affected when the DMD was deleted in neonatal liver, we detected both loss of Igf2 imprinting and a reduction in Igf2 expression when the DMD was deleted from the maternal and paternal alleles, respectively, suggesting that the methylated DMD is acting as an activator of paternal Igf2 expression in this tissue.


   MATERIALS AND METHODS
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Mice. Mice with the H19{Delta}DMD (51), H19{Delta}3.8kb-5'H19 (52), and H19lxDMD alleles were maintained on a C57BL/6J (B6) background. Reciprocal crosses between B6(CAST7) (27) and H19{Delta}DMD/+ or H19{Delta}3.8kb-5'H19/+ mice were used to generate postimplantation embryos for allelic expression and methylation analyses. Reciprocal crosses between B6(CAST7) and H19{Delta}DMD/{Delta}DMD or H19{Delta}3.8kb-5'H19/{Delta}3.8kb-5'H19 mice were used to generate blastocysts.

An albumin-Cre transgenic mouse line [C57BL6-TgN(AlbCre)21Mgn (AlbCreTg)] was maintained in the B6(CAST7) background to generate AlbCreTg mice that were homozygous for the Mus castaneus allele (Cast) at the H19/Igf2 locus (AlbCreCAST7) (35). Reciprocal crosses between AlbCreCAST7 and H19lxDMD/+ mice were used to generate the neonates to study the effect of deleting the DMD in neonatal liver. All studies adhered to procedures consistent with the Institutional Animal Care and Use Committee at the University of Pennsylvania.

Generation of a conditional allele in embryonic stem (ES) cells. H19 129/Sv genomic DNA fragments (51) were used for construction of the targeting vector lxDMD. The vector included a 6.6-kb BamHI-KpnI H19 5' arm, a loxP-containing fragment (39), a 1.6-kb KpnI-HindIII DMD fragment, a loxP-flanked PGK-neo cassette, and a 4.8-kb HindIII-BamHI H19 3' arm cloned into pBluescript II KS (pBIIKS; Stratagene). To synthesize the loxP site, oligonucleotides loxP1 (5'-TCGACGATATCATAACTTCGTATAGCATACATTATACGAAGTTATGGTAC-3') and loxP2 (5'-CATAACTTCGTATAATGTATGCTATACGAAGTTATGATATGC-3') were annealed and cloned into a modified pBIIKS. In the modified plasmid, pMut-BIIKS, the multiple cloning site restriction sites KpnI and SalI of pBIIKS are inverted. The DMD, PGK-neo, H19 5' arm, and H19 3' arm were sequentially subcloned into the starting loxP plasmid (P1) to generate lxDMD.

Vector lxDMD, linearized with NotI, was used to generate the conditional targeted allele H19lxDMDneo. Cell culture, electroporation methods, DNA analysis, generation of chimeric and germ line mice, and genotyping of tail biopsies were carried out as previously described (51), with the following exceptions. For excision of the PGK-neo cassette, 1 x 107 cells were electroporated with 2 to 5 µg of Cre-encoding plasmid, pIC-Cre (21). Two days later, cells were replated and colonies were picked as described elsewhere (58).

Embryo dissections. Dissections and staging of 3.5-, 6.5-, and 9.5-day-postcoitum (dpc) embryos were in accordance with the methods of Hogan et al. (16). The first day after mating was designated 0.5 dpc. Single and pooled blastocysts were collected in phosphate-buffered saline (PBS)-0.3% polyvinylpyrrolidone and stored at –80°C in a minimum amount of PBS-0.3% polyvinylpyrrolidone (≤3 µl) or in 20 to 100 µl Dynal lysis buffer (Dynal, Lake Success, NY) (26). The 6.5-dpc embryo (excluding the ectoplacental cone [EPC]) and the 9.5-dpc embryos, yolk sacs, and placentas were dissected in PBS-0.1% bovine serum albumin and stored at –80°C.

DNA isolation and analysis. DNA was extracted from tail and embryonic tissues (33), neonatal liver (2), and ES cells (51). To PCR genotype the DMD mutant alleles, 1 µl of supernatant was amplified using primer pairs illustrated below in Fig. 1A and 6A (G1/G5 for H19{Delta}DMD [380 bp], G4/G6 for H19{Delta}3.8kb-5'H19 [312 bp], G1/G7 for wild-type H19 [296 bp], and H19lxDMD [356 bp]). Primers G1 and G4 to G6 are described elsewhere (52), and the G7 sequence is 5'-CACACAAAGGATTCTTTGCAGAGAG-3'. To identify mice that were homozygous for Cast at the H19/Igf2 locus, the AlbCreCAST7 mice were genotyped by PCR using D7Mit207 and D7Mit362 markers (Research Genetics, Carlsbad, CA) (33). AlbCreTg was identified by PCR amplification using primers CreF (5'-CATCGTCGGTCCGGGCTGCC-3') and CreR (5'-CCCCCAGGCTAAGTGCCTTC-3') (354 bp). Genotyping by PCR was performed with Ready-To-Go PCR beads (Amersham) and primers at a final concentration of 0.4 µM, as follows: 2 min of denaturation at 95°C; 32 to 35 cycles of 10 s at 94°C, 15 s at 58°C, and 20 s at 72°C; and 5 min of extension at 72°C.


Figure 1
View larger version (28K):
[in this window]
[in a new window]
  		FIG. 1. H19 expression from the mutant paternal alleles in pre- and postimplantation embryos. (A) Schematic of the H19{Delta}DMD and H19{Delta}3.8kb-5'H19 deletion alleles. The 2-kb DMD (unfilled box), the conserved 21-bp CTCF binding repeats (R1 to R4) (triangles), and H19 transcription unit (shaded box) are shown. Cross-hatched boxes represent sequence that was deleted from the alleles. The 1.6-kb KpnI-HindIII fragment, including repeats R2 to R4, was deleted from the H19{Delta}DMD allele, and the 3.8-kb SacI-XbaI fragment, including the entire DMD, was deleted from the H19{Delta}3.8kb-5'H19 allele. Arrows designate the genotyping primers G1 and G4 to G7. (B) H19 expression from the mutant paternal allele in the 3.5-dpc blastocysts, 6.5-dpc embryos, and 9.5-dpc embryos, yolk sacs, and placentas. Paternal-allele-specific expression (percent) was measured relative to expression from the wild-type maternal allele. Dark gray and light gray bars depict the results in the +/{Delta}DMD and +/{Delta}3.8kb-5'H19 samples, respectively. The average value for the samples assayed (N) is shown. As low levels of H19 expression were occasionally detected from the wild-type paternal allele (data not shown), only expression levels of >5% of that normally observed on the wild-type maternal allele were considered to be significant.

 

Figure 6
View larger version (58K):
[in this window]
[in a new window]
  		FIG. 6. Conditional deletion of the 1.6-kb DMD sequence in neonatal liver. (A) Generation of the conditional H19lxDMDallele. Depicted from top to bottom are the wild-type H19 locus, targeting vector lxDMD, H19lxDMDneo targeted allele, H19lxDMD conditional allele after removal of the PGK-neo cassette, and the H19lx{Delta}DMD deletion allele after AlbCre-mediated excision of the DMD in liver. Restriction site locations (in kb) are relative to the start of H19 transcription. Probes A (1.7 kb EcoRV-EcoRI fragment), B (0.75 kb BamHI-StuI fragment), and C (0.9 kb SacI-KpnI fragment) were used to identify clones with the correctly targeted and excised alleles. The DMD and the H19 exons are boxed. Vector lxDMD includes the DMD and a PGK-neo cassette (neo) (boxed), which are both flanked by loxP sites (wide arrowheads), flanking H19 DNA (thin line), and pBKSII (thick line). Primers (G1, G2, G3, G5, and G7) used to identify targeted and excised alleles are below lxDMD. (B) Southern analysis of the H19lxDMDneo, H19lxDMD, and H19lx{Delta}DMD alleles. Probes A and B were hybridized to EcoRV- and StuI-digested DNA, respectively, to detect the parental E14 (lane 1) and targeted H19lxDMDneo ES cell clones (lane 2). Probe C was hybridized to SacI-digested liver DNA (with or without HhaI for methylation analysis) from 5-day and 10-day progeny of reciprocal crosses between H19lxDMD/+ and C7 AlbCre/+ mice (in lanes 1 to 5 H19lxDMD/+ was the mother; in lanes 6 to 10, H19lxDMD/+ was the father). The Cast (C, 1.6 kb), B6 (B, 3.8kb), H19lxDMD (lx, 3.9 kb) and H19lx{Delta}DMD (lx{Delta}, 2.3 kb) SacI fragments are indicated. In the presence of AlbCreTg (lanes 3, 4, and 7 to 11), the excision allele H19lx{Delta}DMD was detected in 40% and 60 to 70% of liver DNA from 5-day and 10-day progeny, respectively. The 5' DMD remained hypomethylated on the maternal DMD deletion allele (lanes 8 and 9) and hypermethlyated on the paternal DMD deletion allele (lanes 10 and 11). (C) Igf2 and H19 expression in neonatal livers from mice inheriting the maternal H19lx{Delta}DMD allele. Allele-specific Igf2 expression (top panel) and H19 expression relative to rpL32 (bottom panel) was measured in liver RNA from 5-day-old progeny of a mating between H19lxDMD/+ and AlbCreTgCAST7 mice. The presence AlbCreTg (Alb-Cre) is noted below each panel in sections C and D. The ratio of maternally to paternally derived Igf2 RNA is 0.01, 0.01, 0.02, 0.02, 0.02, 0.01, 0.24, 0.28 and 0.23 in lanes 1 to 9, respectively. The ratio of H19 to rpL32 RNA is 3.18, 3.96, 3.86 and 2.19 in lanes 1 to 4, respectively. (D) H19 and Igf2 expression in mice with the paternal H19lx{Delta}DMD allele in neonatal liver. Allele-specific H19 expression (top panel) and Igf2 expression relative to rpL32 (bottom panel) was measured in Li RNA from 5-day-old progeny of a mating between AlbCreTgCAST7 and H19lxDMD/+ mice. The ratio of paternally to maternally derived H19 RNA is 0.01, 0.02, 0.06, 0.07, 0.12, and 0.13 in lanes 1 to 6, respectively. The ratio of Igf2 to rpL32 RNA is 3.97, 5.17, 2.26, and 1.94 in lanes 1 to 4, respectively. The average Igf2/rpl32 RNA levels of four Cast/B6 and four Cast/lx{Delta}DMD neonatal livers were 3.93 and 1.69, respectively. The difference was statistically significant (P = 0.04).

 
For genotyping and DNA methylation detection by Southern analysis, DNA was digested and analyzed with probes illustrated below in Fig. 6A (see reference 52 for conditions).

RNA isolation and analysis. RNA from single blastocysts was purified on Dynabeads oligo(dT)25, and a Dynabead oligo(dT)25 covalently linked cDNA library was synthesized. Allele-specific H19 expression was measured on the Roche Light Cycler using second-strand reverse transcription-PCR (RT-PCR) (26).

Poly(A)+ RNA was isolated from 15 to 20 pooled blastocysts using the Dynabead RNA isolation kit (Dynal), and total RNA was isolated from 6.5-dpc embryos (excluding the EPC) and 9.5-dpc embryos, yolk sacs, and placentas using the High Pure RNA tissue kit (Roche). The RNA was reversed transcribed as previously described (13) except that 50 U SuperScript II (Invitrogen) was used with the postimplantation tissue RNA. Typically one to two embryo equivalents of blastocyst cDNA and 1 µl of postimplantation tissue cDNA were used in the following expression assays. H19 allele-specific expression was measured on cDNA using the previously described Light Cycler assay (26, 52). H19 RNA levels relative to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and Igf2 allelic-specific expression were measured on cDNA using Ready-to-Go PCR beads in a final volume of 25 µl, which included 0.1 µCi of {alpha}-32P-labeled dCTP (NEN). The reaction conditions were such that the product was in the linear region of semilog plots of the amount of product versus cycle number (49). Primers HE2 (5'-TGATGGAGAGGACAGAAGGG-3') and HE4 (5'-TTGATTCAGAACGAGACGGAC-3') were used to amplify H19 cDNA (235 bp). H19 PCR was carried out for 1 cycle of 94°C for 2 min followed by 28 cycles of 15 s at 94°C, 10 s at 58°C, and 15 s at 72°C. GAPDH cDNA PCR conditions and primers were described by Fedoriw (13). The H19 and GAPDH cDNA PCR products were resolved by 7% polyacrylamide gel electrophoresis and quantified using a PhosphorImager (Molecular Dynamics). The Igf2 primers Igf2-18 (5'-ATCTGTGACCTCTTGAGCAGG-3') and Igf2-20 (5'-GGGTTGTTTAGAGCCAATCAA-3') amplified a 200-bp region within exon 6 of Igf2. The amplified fragment includes a nucleotide polymorphism between the B6 and Cast allele (45) (see Fig. 3A, below) that is within a Tsp509I restriction site and is unique to the Cast allele. The amplified product was digested with Tsp509I (NEB) at 65°C for 2 hours and resolved on a 15% polyacrylamide gel. The Tsp509I B6 products were 180 bp and 20 bp, and the Cast products were 165, 20, and 15 bp. Cast- and B6-specific products were quantified on a PhosphorImager. Igf2 expression from the placenta-specific promoter was assayed by RT-PCR as described above but without {alpha}-32P-labeled dCTP. Primers Igf2-P1 (5'-GTGGAGAGCAGAAGCCACTT-3') and Igf2-P9 (5'-TCTGGCTGGACGAGAAGTTT-3') amplified a 431-bp product that was digested with BanI (NEB) at 37°C for 2 hours. The amplified fragment includes a nucleotide polymorphism between the B6 ("A" at position 880 of GenBank accession number AY849920) and Cast ("T" at position 847 of GenBank accession number AY849921) alleles that is within a BanI restriction site of the B6 allele. The products were resolved on a 12% polyacrylamide gel. The BanI B6 products were 200, 135, and 96 bp, and the Cast products were 231 and 200 bp (see Fig. 3C, below). Cast- and B6-specific products were quantified with Quantity One (Bio-Rad). Control Cast and B6 cDNAs were assayed to ensure complete digestion of PCR products. In addition, non-RT samples were assayed to ensure that RNA was free from contaminating DNA.


Figure 3
View larger version (35K):
[in this window]
[in a new window]
  		FIG. 3. Igf2 expression from the mutant maternal alleles in postimplantation embryos. (A) Igf2 exon 6 Tsp509I polymorphism. Schematic of the Tsp509I-digested Igf2 RT-PCR products derived from the wild-type B6 (+), H19{Delta}DMD and H19{Delta}3.8kb-5'H19 ({Delta}), and Cast (C) alleles. (B) Allelic analysis of all Igf2 transcripts. The genotype of the sample assayed is indicated above each lane (maternal/paternal). To the right of each panel, the undigested (U) and digested (+, {Delta}, and C) Igf2 RT-PCR products are noted. The presence (+) and absence (–) of Tsp509I and the tissue assayed are presented below each lane of the panels. In 6.5-dpc embryos (EM, top panel), Igf2 expression was measured in embryos inheriting the wild-type (+) or deletion allele from their mother or father. The marker lane (M) contains MspI-digested pBR322 DNA. The bottom panel shows Igf2 expression in 9.5-dpc placenta, embryo, and yolk sac (PL, EM, and YS) and in Cast (C) and B6 (+) neonatal liver. In {Delta}DMD/C and {Delta}3.8kb-5'H19/C 9.5-dpc placenta, embryo, and yolk sac, the ratio of maternally to paternally derived Igf2 ranged from 1.13 to 1.41 (PL), 0.96 to 1.15 (EM), and 1.06 to 1.18 (YS), respectively. There was no significant difference in Igf2 expression between {Delta}DMD/C and {Delta}3.8kb-5'H19/C samples. For all 6.5- and 9.5-dpc samples assayed, no PCR product was detected in the absence of the RT enzyme (data not shown). (C) Igf2 exon µ1 BanI polymorphism. The schematic shows the BanI-digested Igf2 RT-PCR products expressed from the wild-type B6 (+), H19{Delta}DMD and H19{Delta}3.8kb-5'H19 ({Delta}), and Cast (C) alleles. (D) Allelic Igf2 expression initiated from the Igf2 placenta-specific promoter P0. See the legend for panel B for panel description.

 
Liver and tongue were collected from 0- to 16-day neonatal mice, and total RNA was isolated by the lithium chloride method (1). RNase protection assays were performed, and products were resolved on 7% polyacrylamide-7 M urea gels to measure H19, Igf2, and rpL32 RNA (50).

DNA bisulfite modification and sequencing analysis. Blastocyst (pools of 20 to 30), 6.5-dpc embryo (excluding the EPC) and 9.5-dpc embryonic DNAs isolated from heterozygous embryos that inherited the mutant alleles maternally (H19{Delta}DMD/+ and H19{Delta}3.8kb-5'H19/+) and paternally (H19+/{Delta}DMD and H19+/{Delta}3.8kb-5'H19) were subjected to bisulfite modification, PCR amplification, subcloning, and sequencing as previously described (13, 27). Mutagenized DNAs were subjected to multiple independent PCR amplifications. DMD sequence from the wild-type allele (bp –3953 to –3532 relative to the H19 transcription start site) and the H19{Delta}DMD allele (sequence flanking the deletion between bp –3953 and bp –3622 and bp –2093 to –1894 relative to the H19 transcription start site) was analyzed. Promoter proximal (PP) sequence assayed was from the wild-type and H19{Delta}DMD alleles (bp –872 to –492 relative to the H19 transcription start site) and the H19{Delta}3.8kb-5'H19 allele (sequence flanking the deletion between bp –4777 and –4668 and –810 to –492 relative to the H19 transcription start site). Primer pairs used for the first and second rounds of PCR amplification of the DMD sequence were B1/B3 and B2/B4 for the wild-type allele and B1/B10 and B2/B11 for the H19{Delta}DMD allele (see Fig. 4A, below). Similarly, primer pairs used to amplify the PP region were B12/B14 and B13/B14 for the wild-type and H19{Delta}DMD alleles and B15/B14 and B16/B14 for the H19{Delta}3.8kb-5'H19 alleles (see Fig. 5A, below). Primers B1 to B4, B10, and B11 are described in reference 52. The sequences of primers B12 to B16 are as follows: B12 (5'-GTTGAGGATTTGTTAAGGTGTTATTGT-3'), B13 (5'-GAGTGGTTATGATTGGTTAGTTTTTGAG-3'), B14 (5'-AATAATAACTAATTTAAACACTCCTCACC-3'), B15 (5'-GGGTTATAGTGTGAGTTTAGG-3'), and B16 (5'-GGTTATGAATTTAGAAGAGAT-3').


Figure 4
View larger version (36K):
[in this window]
[in a new window]
  		FIG. 4. Methylation pattern of the 5' DMD sequence on the {Delta}DMD alleles in pre- and postimplantation embryos. (A) Schematic of the wild-type and H19{Delta}DMD regions analyzed by the bisulfite mutagenesis and sequencing assay. Sixteen and seven CpG dinucleotides were assayed on the wild-type and H19{Delta}DMD alleles, respectively. The seven CpG dinucleotides at the 5' DMD region (gray bar over open circles) are common to both the wild-type and deletion alleles. The H19{Delta}DMD CpG dinucleotides within the 5' DMD sequence (1 to 7) are depicted. Above each allele are the primers (B1 to B4, B10, and B11) used to amplify the assayed regions. See Materials and Methods for location of primers relative to the H19 transcription start site. (B) Methylation status of the individual maternal and paternal DNA strands of the wild-type and H19{Delta}DMD alleles in 3.5-dpc, 6.5-dpc, and 9.5-dpc embryos. Methylated and unmethylated cytosines are represented as filled and open circles, respectively. An absent circle indicates the residue was not assayed. Profiles observed more than once are indicated to the left of each strand. On the H19{Delta}DMD alleles, the CpG dinucleotides within the loxP/vector sequence and sequence downstream of the deleted DMD sequence were also methylated exclusively on the paternal allele in the blastocysts and hypermethylated on both parental alleles in postimplantation embryos (data not shown). (C) Summary of the methylation profiles of the seven 5' DMD CpG dinucleotides common to the wild-type and H19{Delta}DMD alleles in 3.5-dpc, 6.5-dpc, and 9.5-dpc embryos. White and black bars represent the fractions of methylated cytosines on the maternal and paternal wild-type alleles, respectively. Gray cross-hatched and gray solid bars represent the fractions of methylated cytosines on the maternal and paternal H19{Delta}DMD alleles, respectively.

 

Figure 5
View larger version (55K):
[in this window]
[in a new window]
  		FIG. 5. Methylation pattern of the promoter proximal (PP) region on the wild-type and H19{Delta}DMD and H19{Delta}3.8kb-5'H19 alleles in pre- and postimplantation embryos. (A) Schematic of the wild-type and H19{Delta}DMD and H19{Delta}3.8-5'H19 regions analyzed by the bisulfite mutagenesis and sequencing assay. The nine CpG dinucleotides spanning the PP region were assayed on the wild-type and H19{Delta}DMD alleles. On the H19{Delta}3.8-5'H19 allele, as specified beneath the schematic (gray bars), the eight remaining PP CpGs (2-9) were assayed. The primers (B12-B16) used to amplify the mutagenized DNA are shown above each allele. See Materials and Methods for location of the primers relative to the H19 transcription start site. (B) The methylation status of the individual maternal and paternal DNA strands of the wild-type, H19{Delta}DMD and H19{Delta}3.8-5'H19 alleles in 3.5-dpc, 6.5-dpc and 9.5-dpc embryos. See Fig. 4 legend for details. On the H19{Delta}3.8-5'H19 alleles, the CpG dinucleotides within the loxP/vector sequence and sequence upstream of the deleted DMD sequence were also methylated exclusively on the paternal allele in sperm and blastocysts and hypermethylated on both parental alleles in postimplantation embryos (data not shown). (C) Summary of the methylation profiles of the eight PP CpG dinucleotides common to the wild-type, H19{Delta}DMD and H19{Delta}3.8-5'H19 alleles in 3.5-dpc, 6.5-dpc and 9.5-dpc embryos. White and black bars represent the fraction of methylated cytosines on the maternal and paternal wild-type alleles, respectively. Gray crosshatched and gray bars represent the fraction of methylated cytosines on the maternal and paternal H19{Delta}DMD alleles, respectively. Black crosshatched and light gray bars represent the fraction of methylated cytosines on the maternal and paternal H19{Delta}3.8-5'H19 alleles, respectively.

 

   RESULTS
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
H19 remains inactive on the paternal H19{Delta}DMD and H19{Delta}3.8kb-5'H19 alleles in blastocysts. We previously generated two H19 alleles harboring mutations of the DMD, H19{Delta}DMD, which deletes 1.6 kb of the DMD, leaving the 5' CTCF site intact, and H19{Delta}3.8kb-5'H19, which removes the entire DMD and additional sequence between the DMD and promoter (Fig. 1A). We had determined that H19 is activated at similar levels from the paternal H19{Delta}DMD and H19{Delta}3.8kb-5'H19 deletion alleles in neonatal tissues (52), but this expression varied according to the tissue. It was of interest to determine if the DMD is required to prevent activation of the paternal H19 allele when maternal H19 is first expressed, in the trophectoderm of blastocysts (9, 34). We, therefore, examined allele-specific H19 expression in blastocysts, at 3.5 dpc. In single blastocysts and in pools of blastocysts, no significant H19 expression was detected from either of the paternally inherited mutant alleles (Fig. 1B and data not shown). These results suggest that either the deletion of the DMD does not perturb imprinted expression at this stage or that H19 cannot be expressed in blastocysts without the DMD sequence. To test if the DMD sequence was required for H19 expression in blastocysts, maternal transmission of the deletions was examined.

The DMD is required on the maternal allele for H19 expression in blastocysts. H19 expression from the maternal H19{Delta}DMD and H19{Delta}3.8kb-5'H19 alleles was similarly reduced, in a tissue-specific manner in neonatal tissues, suggesting that the DMD is required for H19 expression (52). It was not evident from these findings, however, whether the DMD is normally required for initiation of maternal-allele-specific expression or only for its maintenance throughout development. To address these possibilities, we analyzed H19 RNA relative to GAPDH RNA in pools of wild-type blastocysts versus pools of blastocysts with maternal inheritance of the DMD deletions. In contrast to the wild-type allele, H19 was not expressed in H19{Delta}3.8kb-5'H19 blastocysts and was barely detectable (<10% relative to expression from the wild-type allele) in H19{Delta}DMD blastocysts (Fig. 2A). These experiments indicate that the DMD is required to initiate maternal-allele-specific expression of H19 in blastocysts. In addition, the low but detectable level of H19 from the H19{Delta}DMD maternal allele, but not from the H19{Delta}3.8kb-5'H19 maternal allele, suggested that the extra sequence removed from the 3.8-kb deletion allele harbors elements required for full H19 expression.


Figure 2
View larger version (45K):
[in this window]
[in a new window]
  		FIG. 2. H19 expression from the mutant maternal alleles in pre- and postimplantation embryos. Results are shown for the 3.5-dpc blastocysts (A), 6.5-dpc embryos (B), and 9.5-dpc embryos, yolk sacs, and placentas (C). In each panel, H19 RNA relative to GAPDH (Gapd) RNA for wild-type and mutant littermates is presented graphically. Error bars reflect results from three experiments. A phosphorimage of the H19 and GAPDH RT-PCR products from one representative experiment is presented below each graph. Black, dark gray, and light gray bars depict the results in the +/+, {Delta}DMD/+, and {Delta}3.8kb-5'H19/+ samples, respectively.

 
Biallelic H19 expression is detected earlier and at higher levels from the H19{Delta}DMD alleles than from the H19{Delta}3.8kb-5'H19 alleles in postimplantation embryos. In neonatal tissue, H19 activation from the paternal deletion alleles varies according to tissue and ranges from <10% to 50% relative to the wild-type maternal allele (19, 52). To determine when expression was first activated from the mutant paternal alleles and if expression from the paternal H19{Delta}DMD and H19{Delta}3.8kb-5'H19 deletion alleles was similar throughout development, we analyzed RNA from 6.5-dpc embryos and 9.5-dpc embryo, yolk sac, and placenta of wild-type and mutant littermates. In 6.5-dpc embryos, where H19 is normally detected in extraembryonic tissues (34), H19 was expressed from the paternal H19{Delta}DMD allele at nearly 20% relative to the levels expressed from the wild-type maternal allele (Fig. 1B), but the H19{Delta}3.8kb-5'H19 allele remained inactive. As we previously reported for neonatal tissue, the levels of H19 from the mutant paternal H19{Delta}DMD allele relative to those from the wild-type maternal allele varied in a tissue-specific manner in the 9.5-dpc tissues (Fig. 1B): ~30% in yolk sac, ~20% in embryo, and ~15% in placenta. Comparatively higher levels of H19 expression were detected from the paternal H19{Delta}DMD than from the H19{Delta}3.8kb-5'H19 deletion allele in the assayed tissues (Fig. 1B). Additionally, H19 RNA levels from the paternal H19{Delta}3.8kb-5'H19 allele in placenta were negligible. These results further suggested that the extra sequence removed from the H19{Delta}3.8kb-5'H19 allele contributes positively to H19 expression.

To determine if H19 expression was similarly affected on the maternal deletion alleles in postimplantation embryos, we assayed H19 RNA in 6.5-dpc and 9.5-dpc embryos. H19 was expressed earlier and at higher levels from the maternal H19{Delta}DMD allele than from the maternal H19{Delta}3.8kb-5'H19 allele (Fig. 2B and C), similar to the behavior of the paternal deletion alleles. This was most evident in the 6.5-dpc embryo and 9.5-dpc placenta, with barely detectable H19 expression from the H19{Delta}3.8kb-5'H19 allele. These analyses confirmed that the additional sequence removed from the H19{Delta}3.8kb-5'H19 allele contributes to H19 expression either directly by promoting transcription or indirectly by providing partial insulator function that would allow H19 to access the shared enhancers. In all tissues assayed, H19 expression from the mutant maternal allele was equivalent to that seen from the mutant paternal allele (compare Fig. 1B to 2B and C), further demonstrating that the DMD is required for allele-specific expression throughout development.

Igf2 activation is similar from the maternal H19{Delta}DMD and H19{Delta}3.8kb-5'H19 alleles in postimplantation embryos. The DMD blocks access of Igf2 to shared enhancers on the maternal allele. To determine the Igf2 expression profiles on the mutant maternal alleles in early development, allele-specific Igf2 expression was measured using an RT-PCR assay that detects a Cast/B6 polymorphism within a Tsp509I restriction site in exon 6 of Igf2 (Fig. 3A) (45). Exon 6 is common to all Igf2 transcripts, which are initiated from various promoters (28). Igf2 was biallelically expressed at low levels in pools of blastocysts inheriting the wild-type or mutant maternal alleles (data not shown), indicating that enhancer-driven, imprinted Igf2 had not been activated. This result supports earlier studies suggesting that Igf2 imprinted expression is established in postimplantation embryos (22, 44). We also detected low levels of expression of Igf2 from the wild-type maternal allele in 6.5-dpc embryos (10 to 15% relative to the wild-type paternal allele), but not in 9.5-dpc tissues (compare +/C and C/+ lanes in Fig. 3B). This suggested that some of the expression in 6.5-dpc embryos is independent of enhancers downstream of the DMD or that the DMD insulator is not yet fully functional in blocking enhancer-dependent expression of Igf2. In 6.5-dpc embryos and 9.5-dpc embryos, yolk sacs, and placentas, Igf2 was fully expressed from the maternal mutant alleles (Fig. 3B). There was no detectable difference in Igf2 expression from the two deletion alleles, eliminating the possibility that the remaining CTCF site on the H19{Delta}DMD allele was imparting insulator activity.

Since H19 imprinted expression is first initiated in the extraembryonic lineages, it was of interest to determine if Igf2 was expressed from the placental specific promoter P0 in early postimplantation embryos (28). We developed an allele-specific assay to measure expression from this promoter utilizing a polymorphism between the Cast and B6 alleles (Fig. 3C). P0-specific monoallelic Igf2 expression was observed in wild-type 6.5-dpc embryos and 9.5-dpc placentas, yolk sacs, and embryos (Fig. 3D and data not shown). Analysis of expression in embryos inheriting the deletions maternally revealed a similar biallelic expression pattern as described above for all Igf2 transcripts (Fig. 3D). Although we did not observe size differences in the 9.5-dpc placentas and embryos of wild-type and mutant littermates inheriting the maternal deletion alleles, 13.5-dpc H19{Delta}3.8kb-5'H19/+ placentas and embryos were 40% (n = 9; P = 0.018) and 18% (n = 3; P = 0.028) heavier, respectively, than their wild-type counterparts.

Methylation analysis of the DMD deletion alleles. In our previous study, we reported that sequence at the first CTCF site on the H19{Delta}DMD allele exhibited a partial loss of differential methylation in neonatal tissues, with the paternal allele losing methylation and the maternal allele gaining methylation (52). To determine when differential methylation was compromised, we analyzed DNA methylation in pre- and postimplantation embryos (Fig. 4A). Paternal-allele-specific methylation of the remaining 5' part of the DMD sequence was detected in blastocysts, suggesting that the deleted DMD sequence is not required for establishment of paternal-allele-specific methylation at this locus. In postimplantation embryos, the paternal alleles maintain their hypermethylated state while the remaining DMD sequence on the maternal deletion allele begins to acquire methylation (Fig. 4B and C), suggesting that a single CTCF site on the maternal allele is not sufficient to antagonize DMD methylation, consistent with reports describing H19 alleles with mutated CTCF sites (31, 41, 47).

The PP region, which lies between the DMD and the H19 promoter, has been shown to have paternal-allele-specific methylation in sperm and in postimplantation embryos (54, 55). Here we assessed the parental-allele-specific methylation of the PP region on the wild-type alleles as well as on the H19{Delta}DMD and H19{Delta}3.8kb-5'H19 deletion alleles (Fig. 5A). A low level of paternal-allele-specific methylation of the PP region was detected on the wild-type and mutant deletion alleles in the preimplantation embryos (Fig. 5B and C). In contrast, in postimplantation embryos, the paternal PP region was preferentially methylated in wild-type embryos, while this region acquired a similar low level of methylation on both of the parental alleles in DMD mutant embryos. These results indicate that the germ line-inherited DMD is required to maintain paternal-allele-specific methylation of 5' H19 sequence in the early postimplantation embryo.

A conditional allele reveals that the 1.6-kb DMD is required for both maternal-allele-specific silencing and paternal-allele-specific activation of Igf2 in neonatal liver. To test if the DMD sequence is required for maintenance of imprinting in somatic cells, we generated mice with a conditional allele of the DMD, H19lxDMD, by introducing loxP sites that flank the 1.6-kb DMD sequence at the endogenous locus (Fig. 6A and B). We initially crossed the H19+/lxDMD mice to germ line-Cre mice (4) to generate mice with a germ line deletion of the DMD. These mice, which served as a control for the conditional allele, exhibited the same loss of imprinting that was previously described for the H19{Delta}DMD allele (data not shown). To delete the DMD in neonatal liver, we generated progeny from reciprocal matings of H19+/lxDMD and AlbCreCAST7 mice. The floxed 1.6-kb DMD was deleted in neonatal liver in 40 to 70% of the cells analyzed, dependent upon the age of the neonate but irrespective of the parental allele (Fig. 6B, bottom panel, lanes 3, 4, and 7 to 11), as previously demonstrated with AlbCreTg (35). Paternal-allele-specific methylation was maintained at the H19 locus upon deletion of the DMD in neonatal liver (Fig. 6B, bottom panel, lanes 8 to 11, and data not shown), as was also observed by Srivastava et al. upon deleting the DMD in another somatic tissue (42, 43).

To determine the effect of maternal deletion of the DMD in neonatal liver, allele-specific Igf2 expression was assessed. Maternal Igf2 expression was detected in neonatal liver RNA (Fig. 6C, lanes 7 to 9) but not in neonatal muscle RNA (data not shown). Control mice that inherited the H19lxDMD allele alone had no changes in Igf2 expression (Fig. 6C, top panel, compare lanes 1 to 3 with lanes 4 to 6). These data showed that the 1.6-kb DMD sequence is required to maintain paternal-allele-specific Igf2 expression in neonatal liver. In contrast to what is observed with the germ line DMD deletion, there were no changes in H19 RNA from the H19lx{Delta}DMD deletion allele (Fig. 6C, bottom panel).

Paternal inheritance of the H19lx{Delta}DMD allele was next evaluated in the livers of neonatal progeny. Low levels of H19 RNA were detected in neonatal liver RNA from the H19lx{Delta}DMD deletion allele (~12%) (Fig. 6D, lanes 5 and 6). These levels, however, were less than observed with germ line deletion of the DMD (Fig. 1B) (52). Inheriting the H19lxDMD allele also had a minor, yet reproducible, effect on parental-allele-specific H19 expression (Fig. 6D, top panel, compare lanes 1 and 2 with lanes 3 and 4), suggesting that the presence of the loxP site between R1 and R2 within the DMD slightly affects silencing of H19 in neonatal liver. However, this inserted sequence did not affect Igf2 expression. We also detected reduced Igf2 expression from the H19lx{Delta}DMD allele compared to the wild-type allele (reduction of >50%) (Fig. 6D, bottom panel, compare lanes 3 and 4 with lanes 1 and 2). Such a reduction is predicted, as the Igf2 RNA was ~70% lower in neonatal liver of mice inheriting the paternal H19{Delta}DMD allele (52). These experiments, which are concordant with similar experiments in neonatal muscle (43), demonstrate that the DMD is required for maternal-allele-specific silencing and paternal-allele-specific activation of Igf2 in neonatal liver.


   DISCUSSION
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
Molecular analyses of the endogenous wild-type and mutant H19/Igf2 loci have demonstrated that the H19 DMD is critical for silencing H19 on the paternal allele through DNA methylation and for silencing Igf2 on the maternal allele through enhancer blocking activity. In this study we addressed the role of the DMD in establishment versus maintenance of these functions by analyzing the DMD deletion alleles H19{Delta}DMD and H19{Delta}3.8kb-5'H19 in pre- and postimplantation development and the DMD conditional deletion allele, H19lxDMD, in neonatal liver. Our results demonstrate that the DMD is required to (i) initiate H19 expression in blastocysts and placenta, thereby revealing its role in promoting/enhancing transcription, (ii) establish imprinted expression of H19 and Igf2 when it is first detected, (iii) maintain Igf2 repression on the maternal allele and Igf2 activity on the paternal allele in neonatal liver, and (iv) maintain allele-specific CpG methylation at the H19 locus in postimplantation embryos if the DMD is deleted in the germ line.

Role of the DMD in promoting/enhancing H19 transcription. Here we demonstrate that the DMD is required for H19 expression in the preimplantation embryo. One possible explanation for this result is that in the absence of the DMD, Igf2 is assuming control of the downstream shared enhancers, thereby preventing H19 expression. Since Igf2 imprinted expression is not detected until postimplantation, such an explanation seems unlikely. Another possibility is that the DMD exerts a role in promoting/enhancing transcription in the preimplantation embryo. The function of the DMD on the maternal allele may evolve as development proceeds. In blastocysts and placenta the DMD may have a strong role in regulating H19 transcription, while in other tissues, such as the 9.5-dpc yolk sac and embryo, the DMD may have a less relevant transcriptional role. Rather, in neonatal tissues, where tissue-specific enhancers have been identified, the DMD functions mainly as a methylation-sensitive insulator (18, 19, 23). These studies as well as another deletion study of conserved 3' H19 sequence have not, however, identified enhancers responsible for H19 expression in preimplantation and early postimplantation embryos as well as late-stage extraembryonic tissues (58). Together, these data provide evidence for a novel function for the H19 DMD: as a promoter/enhancer of transcription. It remains to be determined whether this activity is dependent upon CTCF.

A CTCF-dependent function that is independent of its enhancer blocking activity is conceivable, since CTCF functions as a transcriptional activator at other loci (30). Reagents are already available to address this question. For example, H19 alleles harboring mutations in the CTCF binding sites could be analyzed early in development to determine if H19 is expressed (31, 41, 47). If H19 expression is CTCF dependent, these mutant alleles should exhibit reduced or negligible H19 expression in blastocysts and placenta.

Finally, because deletion of sequence immediately 3' of the 1.6-kb DMD deletion (H19{Delta}DMD) does not perturb H19 maternal-specific expression (37, 52), it is likely that the sequence 5' to the original 1.6-kb DMD deletion allele (H19{Delta}DMD) that was removed from the H19{Delta}3.8kb-5'H19 allele (including the 5' CTCF binding site) is contributing to H19 expression differences from the H19{Delta}DMD and H19{Delta}3.8kb-5'H19 alleles. Analysis of new and existing DMD deletion and mutant alleles will help resolve which part of the DMD facilitates transcription.

Role of the DMD in establishment/maintenance of H19/Igf2 imprinted expression. Our analysis of the deletion alleles in early development demonstrates that the DMD is necessary for the establishment of H19 and Igf2 expression. To determine if the DMD was required for maintenance of H19 and Igf2 imprinting, it was necessary to delete the DMD after the imprinting was fully established at the locus. We, therefore, introduced loxP sites at kb –3.7 and –2.1 relative to the start of H19 transcription and removed the sequence conditionally by the introduction of tissue-specific Cre recombinase. The deleted sequence is the same as that removed from the H19{Delta}DMD allele. Maternal-allele-specific deletion of this 1.6-kb region in neonatal liver caused derepression of Igf2, demonstrating that the DMD was continuously required to silence maternal Igf2. Previously, Srivastava et al. floxed sequence from bp –7.0 to –0.8 relative to the start of H19 transcription and deleted the DMD and flanking sequence in neonatal muscle (43). They too found that the deleted sequence was required to maintain the repression of the maternal Igf2 allele. Our study further narrows down the region responsible for Igf2 repression to 1.6 kb of the DMD. Neither of the conditional deletions disrupted H19 imprinting of the paternal allele. H19 silencing and the paternal-allele-specific methylation of the H19 promoter, which is hypomethylated in sperm and is hypermethylated in differentiated tissues, were retained despite the absence of the upstream sequence (data not shown) (42, 43). Thus, the DMD is required to maintain maternal Igf2 repression but not necessary to maintain the silencing of paternal H19 once the imprinted state is fully established.

Analyses of DMD deletion and mutant alleles have suggested that the DMD is required to establish full expression of paternal Igf2 in neonatal liver and maternal H19 in multiple tissues (19, 41, 52). Here we find that the DMD is also required to maintain maximal Igf2 expression in neonatal liver, suggesting that the methylated DMD is enhancing transcription of Igf2 in this tissue. However, we did not detect a significant reduction of H19 RNA upon deleting the DMD from the maternal allele in neonatal liver. Other studies suggest the DMD is necessary for maximal Igf2 as well as H19 expression. In a murine liver tumor model, maternal H19 and paternal Igf2 were reactivated in adult liver tumors, and this activation was dependent upon the endodermal enhancers (56). In addition, in mice inheriting a DMD deletion, Igf2 and H19 were both significantly less activated in these tumors, suggesting that the DMD is required for full expression of Igf2 and H19 (57). We have also observed that reactivation of H19 in adult regenerating liver is dependent upon the DMD (unpublished observation). These latter studies indicate that the DMD is required to reestablish transcription of Igf2 and H19 in adult liver under stressed conditions, while our deletion of the paternal DMD in neonatal liver indicates that the DMD is required to maintain maximal Igf2 expression. It remains to be determined if these results are a consequence of the DMD regulating transcription of H19 and Igf2 or controlling competition for shared enhancers.

Establishment versus maintenance of allele-specific DNA methylation. Allele-specific methylation of sequence upstream from the start of H19 transcription was observed in spermatozoa and preimplantation embryos in absence of all or part of the DMD. Since paternal-allele-specific methylation of upstream sequence on the H19{Delta}DMD allele was established prior to fertilization and persisted in the blastocyst, we initially considered that the methylation imprint was attracted (or established) by the remaining DMD sequence. However, paternal-allele-specific methylation was also detected in the blastocyst at the sequence immediately 3' of the 1.6-kb DMD deletion and, at reduced levels, at the PP region. Furthermore, the presence of paternal-allele-specific methylation at the PP region on the H19{Delta}3.8kb-5'H19 allele, which lacks all DMD sequence, suggested that acquisition of paternal-allele-specific methylation on the H19{Delta}DMD allele was independent of the deleted DMD sequence. While paternal-allele-specific methylation of the 5' DMD sequence was only partially lost in the postimplantation embryos, parental-allele-specific methylation of the sequence 3' of the 1.6-kb DMD deletion and the PP region was completely abolished in the postimplantation embryo. Consequently, the germ line-inherited DMD sequence is essential for maintenance of paternal-allele-specific methylation at the H19 locus in the early postimplantation embryo. Deletion of the DNA methyltransferase 1 gene demonstrated that DNA methylation is necessary for silencing of the paternal H19 allele in the early embryo (25). The time in early development when the DMD is no longer required for imprinted H19 expression and methylation remains to be determined.

A single-copy 15.7-kb H19 transgene that included 5.7 kb of sequence upstream of the H19 transcription start site contained the necessary sequence to acquire paternal-allele-specific methylation in sperm and exhibited maternal-allele-specific expression in neonatal liver. A low-copy-number 7.7-kb H19 transgene that included the same 5' H19 sequence, but not the 3' enhancer elements, also harbored the necessary sequence to acquire paternal-allele-specific methylation in sperm (5). Thus, while transgenic studies have determined the H19 sequence sufficient to attract paternal-allele-specific methylation in the germ line, the numerous deletion and mutant alleles generated at the endogenous H19 locus have not revealed the cis elements necessary for the DMD to acquire methylation in the male germ line. Deletion of conserved sequence 5' or 3' of the DMD as well as the H19 transcription unit did not significantly perturb Igf2 imprinted expression or acquisition of paternal-allele-specific DMD methylation in the male germ line (17, 24, 36-38, 40, 52). Additionally, none of the various DMD mutant alleles that have been generated thus far result in loss of paternal-allele-specific DMD methylation in spermatozoa (12, 31, 41, 47). In contrast, replacement of the DMD with the CpG-rich, CTCF-binding chicken globin insulator sequence resulted in insulator activity on both alleles, but the globin insulator sequence failed to acquire methylation in the male germ line (46). Thus, sequence surrounding the DMD was insufficient to attract DNA methylation to the locus in the presence of foreign sequence. Together, these results highlight the complexity of imprinting regulation at this locus.

How is parental identity conveyed to the H19 locus in the male germ line? One problem with addressing this question is determining (i) whether there is parental identity and (ii) the mark that designates parental identity. In the absence of imprinted expression, we and others have assayed allelic DNA methylation. Could there be another epigenetic modification that designates parental identity? In support of this idea, Davis et al. reported that after methylation is erased in migrating and early colonizing primordial germ cells, the DMD and PP regions are methylated asynchronously in the male germ line, with the paternal allele acquiring methylation prior to the maternal allele (6, 7). Additionally, in a study in which the DMD was inserted at the AFP locus (32) and in a second experiment inserting the DMD within a human ß-globin transgene (48), the DMD was unmethylated in spermatozoa but acquired paternal-allele-specific methylation after implantation. Thus, while these studies indicate that the DMD functioned as a methylation-sensitive and maternal-allele-specific insulator at an exogenous locus, it is not clear if the exogenous DMD recapitulated the paternal-allele-specific silencing function. Nonetheless, these studies support the thesis that an epigenetic mark in addition to allele-specific DNA methylation may distinguish the DMD alleles in preimplantation embryos.

Questions regarding H19/Igf2 imprinting control. In spite of being the best-studied imprinting control region, many questions remain regarding the function of the H19 DMD. Is there another epigenetic mark that precedes paternally specific methylation of the H19 DMD? When is the imprinted state of the H19 gene fully established/locked in, that is, when is the DMD no longer required to maintain H19 imprinted expression? It will also be important to define regions of the H19/Igf2 locus that physically interact in a temporal and tissue- and allele-specific manner and are critical for H19/Igf2 imprinting control. Tissue-specific physical interactions between distal enhancers and promoters at the ß-globin locus have been reported using chromosome conformation capture (3C) analysis, suggesting in vivo looping of ß-globin DNA (53). A recent study using 3C to study physical interactions at the H19/Igf2 locus suggested the presence of allele-specific interactions between the H19 DMD and Igf2 differentially methylated regions (29). Another interesting study characterized matrix attachment regions at the Igf2 locus in the vicinity of the Igf2 differentially methylated regions and identified paternal-allele-specific nuclear matrix associations of matrix attachment regions that may affect interactions of distal control elements and the looping of DNA (60). These approaches to studying the three-dimensional organization at the H19/Igf2 locus complement traditional strategies that identify essential regulatory cis elements. Ultimately, analysis of the DMD deletion and mutant alleles that we and others have generated will be critical for defining significant local and long-range temporal and tissue-specific chromatin interactions important for H19/Igf2 imprinting.


   ACKNOWLEDGMENTS
 
We thank J. Richa and the University of Pennsylvania Transgenic Core Facility for the production of chimeric mice and Grace Yang for technical assistance. We thank Mark Magnuson for the C57BL6-TgN(AlbCre)21Mgn mice. We thank Nora Engel and Raluca Verona for their comments on the manuscript.

This work was supported by U.S. Public Service grant GM51279 and the Howard Hughes Medical Institute. A.M.F. was supported by an NIH predoctoral training grant (T32 HD-07516).


   FOOTNOTES
 
* Corresponding author. Mailing address: Department of Cell and Developmental Biology, University of Pennsylvania School of Medicine, Philadelphia, PA 19104. Phone: (215) 898-9063. Fax: (215) 573-6434. E-mail: bartolom{at}mail.med.upenn.edu. Back


   REFERENCES
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 
1. Auffray, C., and F. Rougeon. 1980. Purification of mouse immunoglobulin heavy-chain messenger RNAs from total myeloma tumor RNA. Eur. J. Biochem. 107:303-314.[Medline]

2. Bartolomei, M. S., A. L. Webber, M. E. Brunkow, and S. M. Tilghman. 1993. Epigenetic mechanisms underlying the imprinting of the mouse H19 gene. Genes Dev. 7:1663-1673.[Abstract/Free Full Text]

3. Bell, A. C., and G. Felsenfeld. 2000. Methylation of a CTCF-dependent boundary controls imprinted expression of the Igf2 gene. Nature 405:482-485.[CrossRef][Medline]

4. Cho, H., J. Mu, J. K. Kim, J. L. Thorvaldsen, Q. Chu, E. B. Crenshaw III, K. H. Kaestner, M. S. Bartolomei, G. I. Shulman, and M. J. Birnbaum. 2001. Insulin resistance and a diabetes mellitus-like syndrome in mice lacking the protein kinase Akt2 (PKB beta). Science 292:1728-1731.[Abstract/Free Full Text]

5. Cranston, M. J., T. L. Spinka, D. A. Elson, and M. S. Bartolomei. 2001. Elucidation of the minimal sequence required to imprint H19 transgenes. Genomics 73:98-107.[CrossRef][Medline]

6. Davis, T. L., J. M. Trasler, S. B. Moss, G. J. Yang, and M. S. Bartolomei. 1999. Acquisition of the H19 methylation imprint occurs differentially on the parental alleles during spermatogenesis. Genomics 58:18-28.[CrossRef][Medline]

7. Davis, T. L., G. J. Yang, J. R. McCarrey, and M. S. Bartolomei. 2000. The H19 methylation imprint is erased and reestablished differentially on the parental alleles during male germ cell development. Hum. Mol. Genet. 9:2885-2894.[Abstract/Free Full Text]

8. Diaz-Meyer, N., C. D. Day, K. Khatod, E. R. Maher, W. Cooper, W. Reik, C. Junien, G. Graham, E. Algar, V. M. Der Kaloustian, and M. J. Higgins. 2003. Silencing of CDKN1C (p57KIP2) is associated with hypomethylation at KvDMR1 in Beckwith-Wiedemann syndrome. J. Med. Genet. 40:797-801.[Abstract/Free Full Text]

9. Doherty, A. S., M. R. W. Mann, K. D. Tremblay, M. S. Bartolomei, and R. M. Schultz. 2000. Differential effects of culture on imprinted H19 expression in the preimplantation mouse embryo. Biol. Reprod. 62:1526-1535.[Abstract/Free Full Text]

10. Drewell, R. A., J. D. Brenton, J. F. Ainscough, S. C. Barton, K. J. Hilton, K. L. Arney, L. Dandolo, and M. A. Surani. 2000. Deletion of a silencer element disrupts H19 imprinting independently of a DNA methylation epigenetic switch. Development 127:3419-3428.[Abstract]

11. Du, M., L. G. Beatty, W. Zhou, J. Lew, C. Schoenherr, R. Weksberg, and P. D. Sadowski. 2003. Insulator and silencer sequences in the imprinted region of human chromosome 11p15.5. Hum. Mol. Genet. 12:1927-1939.[Abstract/Free Full Text]

12. Engel, N. I., A. E. West, G. Felsenfeld, and M. S. Bartolomei. 2004. Antagonism between DNA hypermethylation and enhancer-blocking activity at the H19 DMD is revealed by CpG mutations. Nat. Genet. 36:883-888.[CrossRef][Medline]

13. Fedoriw, A. M., P. Stein, P. Svoboda, R. M. Schultz, and M. S. Bartolomei. 2004. Transgenic RNAi reveals essential function for CTCF in H19 gene imprinting. Science 303:238-240.[Abstract/Free Full Text]

14. Fitzpatrick, G. V., P. D. Soloway, and M. J. Higgins. 2002. Regional loss of imprinting and growth deficiency in mice with a targeted deletion of KvDMR1. Nat. Genet. 32:426-431.[CrossRef][Medline]

15. Hark, A. T., C. J. Schoenherr, D. J. Katz, R. S. Ingram, J. M. Levorse, and S. M. Tilghman. 2000. CTCF mediates methylation-sensitive enhancer-blocking activity at the H19/Igf2 locus. Nature 405:486-489.[CrossRef][Medline]

16. Hogan, B., F. Costantini, and E. Lacy. 1994. Manipulating the mouse embryo, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.

17. Jones, B. K., J. M. Levorse, and S. M. Tilghman. 1998. Igf2 imprinting does not require its own DNA methylation or H19 RNA. Genes Dev. 12:2200-2207.[Abstract/Free Full Text]

18. Kaffer, C. R., A. Grinberg, and K. Pfeifer. 2001. Regulatory mechanisms at the mouse Igf2/H19 locus. Mol. Cell. Biol. 21:8189-8196.[Abstract/Free Full Text]

19. Kaffer, C. R., M. Srivastava, K.-Y. Park, E. Ives, S. Hsieh, J. Batle, A. Grinberg, S.-P. Huang, and K. Pfeifer. 2000. A transcriptional insulator at the imprinted H19/Igf2 locus. Genes Dev. 14:1908-1919.[Abstract/Free Full Text]

20. Kanduri, C., V. Pant, D. Loukinov, E. Pugacheva, C. F. Qi, A. Wolffe, R. Ohlsson, and V. V. Lobanenkov. 2000. Functional association of CTCF with the insulator upstream of the H19 gene is parent of origin-specific and methylation-sensitive. Curr. Biol. 10:853-856.[CrossRef][Medline]

21. Kuhn, R., J. Lohler, D. Rennick, K. Rajewsky, and W. Muller. 1993. Interleukin-10-deficient mice develop chronic enterocolitis. Cell 75:263-274.[CrossRef][Medline]

22. Latham, K. E., A. S. Doherty, C. D. Scott, and R. M. Schultz. 1994. Igf2r and Igf2 gene expression in androgenetic, gynogenetic, and parthenogenetic preimplantation mouse embryos: absence of regulation by genomic imprinting. Genes Dev. 8:290-299.[Abstract/Free Full Text]

23. Leighton, P. A., J. R. Saam, R. S. Ingram, C. L. Stewart, and S. M. Tilghman. 1995. An enhancer deletion affects both H19 and Igf2 expression. Genes Dev. 9:2079-2089.[Abstract/Free Full Text]

24. Lewis, A., K. Mitsuya, M. Constancia, and W. Reik. 2004. Tandem repeat hypothesis in imprinting: deletion of a conserved direct repeat element upstream of H19 has no effect on imprinting in the Igf2-H19 region. Mol. Cell. Biol. 24:5650-5656.[Abstract/Free Full Text]

25. Li, E., C. Beard, and R. Jaenisch. 1993. Role for DNA methylation in genomic imprinting. Nature 366:362-365.[CrossRef][Medline]

26. Mann, M. R., Y. G. Chung, L. D. Nolen, R. I. Verona, K. E. Latham, and M. S. Bartolomei. 2003. Disruption of imprinted gene methylation and expression in cloned preimplantation stage mouse embryos. Biol. Reprod. 69:902-914.[Abstract/Free Full Text]

27. Mann, M. R., S. S. Lee, A. S. Doherty, R. I. Verona, L. D. Nolen, R. M. Schultz, and M. S. Bartolomei. 2004. Selective loss of imprinting in the placenta following preimplantation development in culture. Development 131:3727-3735.[Abstract/Free Full Text]

28. Moore, T., M. Constancia, M. Zubair, B. Bailleul, R. Feil, H. Sasaki, and W. Reik. 1997. Multiple imprinted sense and antisense transcripts, differential methylation and tandem repeats in a putative imprinting control region upstream of mouse Igf2. Proc. Natl. Acad. Sci. USA 94:12509-12514.[Abstract/Free Full Text]

29. Murrell, A., S. Heeson, and W. Reik. 2004. Interaction between differentially methylated regions partitions the imprinted genes Igf2 and H19 into parent-specific chromatin loops. Nat. Genet. 36:889-893.[CrossRef][Medline]

30. Ohlsson, R., R. Renkawitz, and V. Lobanenkov. 2001. CTCF is a uniquely versatile transcription regulator linked to epigenetics and disease. Trends Genet. 17:520-527.[CrossRef][Medline]

31. Pant, V., P. Mariano, C. Kanduri, A. Mattsson, V. Lobanenkov, R. Heuchel, and R. Ohlsson. 2003. The nucleotides responsible for the direct physical contact between the chromatin insulator protein CTCF and the H19 imprinting control region manifest parent of origin-specific long-distance insulation and methylation-free domains. Genes Dev. 17:586-590.[Abstract/Free Full Text]

32. Park, K. Y., E. A. Sellars, A. Grinberg, S. P. Huang, and K. Pfeifer. 2004. The H19 differentially methylated region marks the parental origin of a heterologous locus without gametic DNA methylation. Mol. Cell. Biol. 24:3588-3595.[Abstract/Free Full Text]

33. Percec, I., R. M. Plenge, J. H. Nadeau, M. S. Bartolomei, and H. F. Willard. 2002. Autosomal dominant mutations affecting X inactivation choice in the mouse. Science 296:1136-1139.[Abstract/Free Full Text]

34. Poirier, F., C.-T. J. Chan, P. M. Timmons, E. J. Robertson, M. J. Evans, and P. W. J. Rigby. 1991. The murine H19 gene is activated during embryonic stem cell differentiation in vitro and at the time of implantation in the developing embryo. Development 113:1105-1114.[Abstract]

35. Postic, C., and M. A. Magnuson. 2000. DNA excision in liver by an albumin-Cre transgene occurs progressively with age. Genesis 26:149-150.[CrossRef][Medline]

36. Reed, M. R., C. F. Huang, A. D. Riggs, and J. R. Mann. 2001. A complex duplication created by gene targeting at the imprinted H19 locus results in two classes of methylation and correlated Igf2 expression phenotypes. Genomics 74:186-196.[CrossRef][Medline]

37. Reed, M. R., A. D. Riggs, and J. R. Mann. 2001. Deletion of a direct repeat element has no effect on Igf2 and H19 imprinting. Mamm. Gen. 12:873-876.[CrossRef][Medline]

38. Ripoche, M.-A., K. Chantal, F. Poirier, and L. Dandolo. 1997. Deletion of the H19 transcription unit reveals the existence of a putative imprinting control element. Genes Dev. 11:1596-1604.[Abstract/Free Full Text]

39. Sauer, B.