The imprinted expression of the mouse Igf2/H19 locus is governed by the differential methylation of the imprinting control region (ICR), which is established initially in germ cells and subsequently maintained in somatic cells, depending on its parental origin. By grafting a 2.9-kbp H19 ICR fragment into a human β-globin yeast artificial chromosome in transgenic mice, we previously showed that the ICR could recapitulate imprinted methylation and expression at a heterologous locus, suggesting that the H19 ICR in the β-globin locus contained sufficient information to maintain the methylation mark (K. Tanimoto, M. Shimotsuma, H. Matsuzaki, A. Omori, J. Bungert, J. D. Engel, and A. Fukamizu, Proc. Natl. Acad. Sci. USA 102:10250-10255, 2005). Curiously, however, the transgenic H19 ICR was not methylated in sperm, which was distinct from that seen in the endogenous locus. Here, we reevaluated the ability of the H19 ICR to mark the parental origin using more rigid criteria. In the testis, the methylation levels of the solitary 2.9-kbp transgenic ICR fragment varied significantly between six transgenic mouse lines. However, in somatic cells, the paternally inherited ICR fragment exhibited consistently higher methylation levels at five out of six randomly integrated sites in the mouse genome. These results clearly demonstrated that the H19 ICR could acquire parent-of-origin-dependent methylation after fertilization independently of the chromosomal integration site or the prerequisite methylation acquisition in male germ cells.

Abnormal methylation at the maternally inherited H19 imprinted control region (H19 ICR) is one of the causative alterations leading to pathogenesis of Beckwith–Wiedemann syndrome (BWS). Recently, it was shown in human BWS patients, as well as mouse cell culture experiments, that Sox-Oct motifs (SOM) in the H19 ICR might play a role in protecting the maternal ICR from de novo DNA methylation. By grafting a mouse H19 ICR fragment into a human β-globin yeast artificial chromosome (YAC) followed by analysis in transgenic mice (TgM), we showed previously that the fragment carried sufficient information to establish and maintain differential methylation after fertilization. To examine possible functions of the SOM in the establishment and/or maintenance of differential methylation, two kinds of YAC-TgM were generated in this study. In the ΔSOM TgM, carrying the mouse H19 ICR bearing an SOM deletion, a maternally inherited transgenic ICR exhibited increased levels of methylation around the deletion site, in comparison to the wild-type control, after implantation. In the λ + CTCF + b (LCb) TgM, carrying a 2.3 kb λ DNA fragment supplemented with the fragment b including the SOM and four CTCF binding sites, maternally and some of the paternally inherited LCb fragments were significantly less methylated when compared with a control λ + CTCF fragment that was supplemented only with additional CTCF sites; the λ + CTCF was substantially methylated regardless of the parent of origin after implantation. These results demonstrated that the SOM in the maternal H19 ICR was required for maintaining surrounding sequences in the unmethylated state in vivo.

Genomic imprinting is a major monoallelic gene expression regulatory mechanism in mammals, and depends on gamete-specific DNA methylation of specialized cis-regulatory elements called imprinting control regions (ICRs). Allele-specific DNA methylation of the ICRs is faithfully maintained at the imprinted loci throughout development, even in early embryos where genomes undergo extensive epigenetic reprogramming, including DNA demethylation, to acquire totipotency. We previously found that an ectopically introduced H19 ICR fragment in transgenic mice acquired paternal allele-specific methylation in the somatic cells of offspring, whereas it was not methylated in sperm, suggesting that its gametic and postfertilization modifications were separable events. We hypothesized that this latter activity might contribute to maintenance of the methylation imprint in early embryos. Here, we demonstrate that methylation of the paternally inherited transgenic H19 ICR commences soon after fertilization in a maternal DNMT3A- and DNMT3L-dependent manner. When its germline methylation was partially obstructed by insertion of insulator sequences, the endogenous paternal H19 ICR also exhibited postfertilization methylation. Finally, we refined the responsible sequences for this activity in transgenic mice and found that deletion of the 5′ segment of the endogenous paternal H19 ICR decreased its methylation after fertilization and attenuated Igf2 gene expression. These results demonstrate that this segment of the H19 ICR is essential for its de novo postfertilization DNA methylation, and that this activity contributes to the maintenance of imprinted methylation at the endogenous H19 ICR during early embryogenesis.

Genomic imprinting is governed by allele-specific DNA methylation at imprinting control regions (ICRs), and the mechanism controlling its differential methylation establishment during gametogenesis has been a subject of intensive research interest. However, recent studies have reported that gamete methylation is not restricted at the ICRs, thus highlighting the significance of ICR methylation maintenance during the preimplantation period where genome-wide epigenetic reprogramming takes place. Using transgenic mice (TgM), we previously demonstrated that the H19 ICR possesses autonomous activity to acquire paternal-allele-specific DNA methylation after fertilization. Furthermore, this activity is indispensable for the maintenance of imprinted methylation at the endogenous H19 ICR during the preimplantation period. In addition, we showed that a specific 5′ fragment of the H19 ICR is required for its paternal methylation after fertilization, while CTCF and Sox-Oct motifs are essential for its maternal protection from undesirable methylation after implantation. To ask whether specific cis elements are sufficient to reconstitute imprinted methylation status, we employed a TgM co-placement strategy for facilitating detection of postfertilization methylation activity and precise comparison of test sequences. Bacteriophage lambda DNA becomes highly methylated regardless of its parental origin and thus can be used as a neutral sequence bearing no inclination for differential DNA methylation. We previously showed that insertion of only CTCF and Sox-Oct binding motifs from the H19 ICR into a lambda DNA (LCb) decreased its methylation level after both paternal and maternal transmission. We therefore appended a 478-bp 5′ sequence from the H19 ICR into the LCb fragment and found that it acquired paternal-allele-specific methylation, the dynamics of which was identical to that of the H19 ICR, in TgM. Crucially, transgene expression also became imprinted. Although there are potential binding sites for ZFP57 (a candidate protein thought to control the methylation imprint) in the larger H19 ICR, they are not found in the 478-bp fragment, rendering the role of ZFP57 in postfertilization H19 ICR methylation a still open question. Our results demonstrate that a differentially methylated region can be reconstituted by combining the activities of specific imprinting elements and that these elements together determine the activity of a genomically imprinted region in vivo.

Expression of the five β-like globin genes (ϵ, Gγ, Aγ, δ, β) in the human β-globin locus depends on enhancement by the locus control region, which consists of five DNase I hypersensitive sites (5′HS1 through 5′HS5). We report here a novel enhancer activity in 5′HS1 that appears to be potent in transfected K562 cells. Deletion analyses identified a core activating element that bound to GATA-1, and a two-nucleotide mutation that disrupted GATA-1 binding in vitro abrogated 5′HS1 enhancer activity in transfection experiments. To determine the in vivo role of this GATA site, we generated multiple lines of human β-globin YAC transgenic mice bearing the same two-nucleotide mutation. In the mutant mice, ϵ-, but not γ-globin, gene expression in primitive erythroid cells was severely attenuated, while adult β-globin gene expression in definitive erythroid cells was unaffected. Interestingly, DNaseI hypersensitivity near the 5′HS1 mutant sequence was eliminated in definitive erythroid cells, whereas it was only mildly affected in primitive erythroid cells. We therefore conclude that, although the GATA site in 5′HS1 is critical for efficient ϵ-globin gene expression, hypersensitive site formation per se is independent of 5′HS1 function, if any, in definitive erythroid cells.