Differential Differences in Methylation Status of Putative Imprinted Genes among Cloned Swine Genomes

DNA methylation is a major epigenetic modification in the mammalian genome that regulates crucial aspects of gene function. Mammalian cloning by somatic cell nuclear transfer (SCNT) often results in gestational or neonatal failure with only a small proportion of manipulated embryos producing live births. Many of the embryos that survive to term later succumb to a variety of abnormalities that are likely due to inappropriate epigenetic reprogramming. Aberrant methylation patterns of imprinted genes in cloned cattle and mice have been elucidated, but few reports have analyzed the cloned pig genome. Four surviving cloned sows that were created by ear fibroblast nuclear transfer, each with a different life span and multiple organ defects, such as heart defects and bone growth delay, were used as epigenetic study materials. First, we identified four putative differential methylation regions (DMR) of imprinted genes in the wild-type pig genome, including two maternally imprinted loci (INS and IGF2) and two paternally imprinted loci (H19 and IGF2R). Aberrant DNA methylation, either hypermethylation or hypomethylation, commonly appeared in H19 (45% of imprinted loci hypermethylated vs. 30% hypomethylated), IGF2 (40% vs. 0%), INS (50% vs. 5%), and IGF2R (15% vs. 45%) in multiple tissues from these four cloned sows compared with wild-type pigs. Our data suggest that aberrant epigenetic modifications occur frequently in the genome of cloned swine. Even with successful production of cloned swine that avoid prenatal or postnatal death, the perturbation of methylation in imprinted genes still exists, which may be one of reason for their adult pathologies and short life. Understanding the aberrant pattern of gene imprinting would permit improvements in future cloning techniques.


Introduction
Somatic cell nuclear transfer (SCNT) is the transmission of a differentiated somatic cell nucleus to an enucleated oocyte. SCNT is used to generate individuals with identical genetic backgrounds, increase the economic efficiency of animal preservation, produce transgenic animals, and cure genetic disorders or cancer [1,2]. However, SCNT-cloned mammals usually have a low survival rate due to abortion, neonatal death and postnatal defects. Animals that have been successfully cloned during the past decade include sheep, cattle, goats, pigs, rabbits, mice, cats, and dogs [3,4,5]. Various types of somatic cell are used for transfer, including mammary gland epithelial, ovary epithelial, cumulus, granulosa, and ear fibroblast cells. Although a variety of nuclear cell types and stages of oocytes have been tried, the success rate still remains low [6]. Most of surviving clones have physiological problems; for example, large offspring syndrome (LOS) and placental abnormalities have been found in cloned cattle, sheep, and mice [7,8,9]. Notably, offspring that are produced by the natural mating of clones that have an abnormal phenomenon do not inherit the abnormality, providing evidence that precise, dynamic epigenetic control is a major requirement during the period of fertilization to blastocyst [10]. Therefore, understanding aberrant methylation patterns and correcting perturbed epigenetic modification will help improve the health of cloned animals.
Epigenetic reprogramming is an essential process in mammals to regulate DNA methylation and gene expression during gametogenesis and embryogenesis. This reprogramming could be performed by demethylases and DNA methyltransferases (Dnmts) to produce demethylated and methylated DNA, respectively. However, no demethylase has been identified in mammals [11,12]. The loss of maternal nuclear Dnmt1 is the cause of aberrant methylation in imprinted genes during nuclear transfer [13]. The reprogramming process is divided into four parts: formation of primordial germ cells (PGCs), maturation of gametes, fertilization to produce a zygote, and embryonic stages. The methylation markers of imprinting are erased during the formation of PGCs and reestablished in the gamete genome. After fertilization, the non-imprinted methylation markers are demethylated, and then the methylation markers are reestablished during embryonic development. However, the methylated or nonmethylated imprinting markers are maintained from the one-cell to the blastocyst stage [14][15][16]. This reprogramming process affects DNA methylation, chromatin histone acetylation, and other embryo growth mechanisms. During embryonic development, the methylation status of some imprinted genes is dynamic and has spatial and temporal requirements [17]. Accumulating evidence indicates that incomplete or inappropriate epigenetic modification of donor nuclei used for nuclear transfer is likely to be the primary cause of failure when cloning animals [18].
DNA methylation at cytosine residues within CpG dinucleotides is one common regulatory modification of gene expression. Differentially methylated regions (DMRs) are DNA regions with methylation differences between parental alleles. Some DMRs are also imprinting control regions (ICRs), which control several imprinted genes in a cluster [15]. Most imprinted genes contain DMRs, which are crucial in maintaining imprinting in mammalian genomes. Genomic imprinting is a uniparentally expressed pattern that includes many reading mechanisms: promoter methylation, antisense transcripts, boundaries, and silencers. It is also involved in the regulation of normal embryonic development, placental growth, parental-specific expression, X-chromosome inactivation, clustering effects of ICRs, and tissue-specific expression [19,20]. A group of imprinted genes may encode signal transduction molecules, cell cycle regulators, transcription factors, enzymes, and non-coding RNAs [21]. So far, approximately 60 imprinted genes have been found in the human genome, but this number is expected to increase to at least 100 [22].
In general, the perturbation of mono-allelic expression of imprinted genes could cause the abortion of embryonic and fetal development during pregnancy. For example, H19 and insulin-like growth factor 2 (IGF2) are well-studied imprinted loci with specific expression patterns controlled by the DMR of H19. CTCFbinding protein is also involved in H19 downstream enhancer regulation [23]. The alteration of the imprinting status of an IGF2 allele (loss of imprinting, LOI) results in biallelic expression during embryonic growth, whereas IGF2 overexpression in mice causes prenatal or postnatal overgrowth that is similar to the symptoms of Beckwith-Wiedemann syndrome [24,25].
In this study, four surviving cloned sows created by ear fibroblast nuclear transfer with whole-cell microinjection, each with a different life span and multiple organ defects, were used as samples in this epigenetic study analyzing the aberrant methylation of maternally and paternally imprinted loci. Four imprinted genes, H19, IGF2, receptor of insulin-like growth factor 2 (IGF2R), and insulin-1 (INS), were selected as targets to verify changes in the DMR methylation patterns of the cloned swine genomes compared with the wild-type genomes. Both IGF2R and H19 are paternally imprinted, characteristically maternally expressed genes that encodes a growth-inhibitory factor and non-protein-coding RNA transcript of unknown function, respectively, whereas the other two genes, IGF2 and INS, are maternally imprinted, characteristically paternally expressed genes that encode growth-promoting factors [26]. To quantify the CpG island methylation status in DMRs of each selected imprinted gene, we used Southern blot hybridization, methylation-specific PCR (MS-PCR), bisulfite sequencing, and a combined bisulfite and restriction assay (COBRA). Significant changes (either hypermethylation or hypomethylation) in the levels of epigenetic methylation were observed in the analyzed imprinted loci in different tissues of cloned sows. Here, we describe the characterization of these epigenetic changes in the examined tissues of cloned swine genomes.

Materials and Methods
Tissue sample collection of cloned and wild-type sows Whole-cell intracytoplasmic microinjection of ear fibroblast cells was used to produce four surviving cloned piglets, named cloned pig Nos. 1 to 4 (CP1 to CP4), as described previously [27]. The sex of the cloned pigs was female, and the species of the enucleated oocytes, somatic cell donors, and recipient pigs were all of the Landrace breed. Three wild-type female pigs with the same genetic background and housed under the same husbandry conditions were used as normal controls. The control body weight was calculated as the average weight of normal newborn piglets from ten litters (98 piglets) of the Landrace breed of the same age as the cloned pigs. Animals were housed and handled according to the guidelines of the Animal Care Committee of the Animal Technology Institute Taiwan (ATIT approval ID: 93021). Physiological characteristics, including birth weight, life span, death weight, and major defects, were collected. Compared with wild-type piglets of the same age, the birth and death weights of all cloned piglets were significantly reduced. Tissues from all three germ layers of cloned pigs, including ectoderm-derived tissues (ear, brain), mesoderm-derived tissues (heart, muscle, kidney), and endoderm-derived tissues (liver, lung), and also extra-embryonic placenta were sampled. Tissues were separated into two parts for DNA and RNA extraction, snap-frozen in liquid nitrogen, and stored at 280uC until use [28].

Isolation of genomic DNA
High-molecular-weight genomic DNAs were extracted by the proteinase K/SDS method as described in our previous report [29]. Briefly, 300 mg of each tissue was homogenized and resuspended in 230 ml lysis buffer (50 mM Tris-HCl, 100 mM NaCl, 100 mM EDTA, pH 8.0), followed by adding 400 ml lysis buffer, 70 ml 10% SDS, 10 ml RNase A (10 mg/ml) and 35 ml proteinase K (10 mg/ ml). The tissue-digested reaction was mixed well and incubated at 55uC for 16 h. The reaction mixture was cleaned twice by adding equal volumes of phenol/chloroform (1:1). The supernatant was transferred, 1.56 volume of pure ethanol was added to precipitate DNA, and the genomic DNA pellet was washed twice with 70% ethanol. The dried DNA pellet was then dissolved in 40 ml of distilled deionized water and stored at 220uC.

Southern blot combined methylation-sensitive enzyme digestion
The pig H19 and IGF2 probes used for Southern blot hybridization were prepared by PCR amplification, with cloning and sequencing using the pGEM-T easy TA Cloning Kit (Promega, Madison, WI). The PCR primer sets were designed as follows: H19 probe: 59-GTGATCGGACTTCTGACCCT-39 and 59-TCTCCACACCCACAAGCCG-39; IGF2 DMR1 probe: 59-AGGGACCTGCCGCTCTGCT-39 and 59-AGGACTGG-GAAAG GAGAGGA-39. To analyze the DNA methylation status in a specific locus, 10 mg of genomic DNA was completely digested with a CpG island cutter (PvuII or NlaIII) alone or together with a methylation-sensitive restriction enzyme (EagI or BstUI; New England Biolabs, Ipswich, MA) and then electrophoresed on 1.2% agarose gels. DNA was transferred onto Hybond N + membranes (GE Healthcare, Piscataway, NJ) in 206 SSC [30]. Singlestranded sense and antisense probes were radio-labeled with a[ 32 P]-dCTP using the Rediprime II random prime labeling system (GE Healthcare) as previously described [31]. Hybridizations were carried out overnight in a 42uC incubator, and membranes were washed according to our previous report [32]. After autoradiograph exposure, the image plate was scanned under a phosphoimager (Typhoon 9200; GE Healthcare).

Methylation-specific PCR (MS-PCR)
Genomic DNA (0.5 mg) was treated with sodium bisulfite according to the manufacturer's recommendation (EZ DNA Methylation Kit TM ; Zymo Research Corp., Orange, CA) and amplified with specific primers for methylated or unmethylated DNA. All PCR reactions were performed on an ABI 2720 thermocycler (Applied Biosystems, Foster, CA) and in 25 ml volumes using the Platinum Taq DNA polymerase system (Invitrogen, Carlsbad, CA). PCR products were separated in 1.5% agarose gels. The M-set primers (H19 M-set: 59-TTTATTGTATTTTTGAACGGCG-39 and 59-CTAAAAACC-GAAACG AACCCG-39) contained at least three CpG sites to distinguish the methylation status of the investigated region. U-set primers (H19 U-set: 59-TTTTGAATGGTGTTGATGGTTTG-39 and 59-TAACCCATACTAAAAACCAAAACA-39) overlapping the M-set primers were used to amplify the unmethylated region. CO-BRA-PCR products were purified with phenol/chloroform, followed by ethanol precipitation. The DNA was resuspended in 8.5 ml of distilled deionized water. Purified PCR products were then digested with 10 U of restriction enzymes (New England Biolabs) as follows: putative DMR products of INS and IGF2 exon IX were digested with BstUI at 60uC; putative DMR products of IGF2R intron II were digested with HpyCH4IV at 37uC. The products of these digestions were electrophoresed in a 6% native acrylamide gel, stained with 200 mg/ml ethidium bromide (EtBr; Sigma, St Louis, MO), and visualized and quantified using a Kodak 1D Image Analysis Software (Eastman Kodak, Rochester, NY).

Bisulfite sequencing
To determine the methylation status of CpG sites within the INS of putative DMRs, primers were designed according to bisulfite standards (no CpG sites within primers) as described in the COBRA above. PCR reactions were performed in a total of 25 ml for each imprinted locus. Individual PCR products were purified with phenol/chloroform followed by ethanol precipitation. Purified PCR products were cloned into a pGEM T-easy Vector (Promega). Plasmid DNA was isolated using a Mini-M TM Plasmid Purification Kit (Viogene, Taipei, Taiwan) and sequenced using a BigDye Terminator cycle sequencing kit with an ABI PRISM 3100 DNA sequencer (Applied Biosystems).

RNA isolation and semi-quantitative RT-PCR
Total RNA was isolated from homogenized tissues using TriReagent TM (Invitrogen) according to the manufacturer's recommendations. One microgram of total RNA was treated twice with 10 U RNase-free DNase I (Invitrogen) to degrade any contaminating DNA, and the reaction was halted by heatinactivation. First-strand cDNAs were synthesized from 2 mg RNA with an oligo (dT) primer and MMLV reverse transcriptase (Promega) in a total volume of 25 ml [33]. RT-PCR was conducted using specific sets of primers for each imprinted gene analyzed:

The analysis and quantification of methylation changes
The definition of hypermethylation or hypomethylation (610% compared to the wild-type tissues) was as previously published in an analysis of the methylation changes of human cancer [34]. The methylation percentage of the H19 putative DMR was calculated according to the Southern blot and MS-PCR data. Southern blotting bands were quantified as previously described [35]. The following formula was used to calculate the methylation percentage from MS-PCR results used: (intensity of M-set band)/(intensity of M-set band + intensity of U-set band)6100 (%). The methylation percentages of INS, IGF2, and IGF2R were based on the COBRA data. The quantification method of COBRA was as previously published [36].

Identification of putative DMRs of imprinted genes in the swine genome
The DMRs of imprinted genes have regions that are highly conserved across various mammal species. First, we compared the well-known DMRs of IGF2, H19, INS, and IGF2R in the human, bovine, and mouse genomes to identify the counterparts of these DMRs in pig genome. The putative DMR1 of the pig IGF2 gene was predicted to be located in intron 3, and the putative DMR2 of pig IGF2 was in exon 9 ( Figure 1A). The putative DMR of the pig H19 gene was located upstream of the promoter, between 22856 and 21489 nucleotides (nt). We used the online software MethyPrimer with restrictive conditions (GC percent .50.0%, CpG observe/expect .0.6) to identify the distribution of CpG islands. The putative DMR of the pig INS gene was located before exon 1. Previously, there was no pig IGF2R intron sequence available in GenBank. The entire second intron of the pig IGF2R gene was newly cloned and sequenced (GenBank accession no. GQ888762) in this study. It contained a putative DMR based on comparisons with other species ( Figure 1A). We further demonstrated that these selected regions of the four imprinted genes exhibited potentially differential methylation patterns in various tissues from wild-type pigs via Southern blot, MS-PCR, and COBRA ( Figure 1B). The four putative DMRs in these imprinted genes were used to estimate the methylation perturbations in cloned pig genomes.

Aberrant methylation of the H19 gene in various tissues from cloned pigs
The H19 gene is a classical maternally expressed imprinted gene, and the aberrant methylation of the H19 DMR often occurs in genetic diseases, growth retardation, prenatal lethality, and many kinds of cancer [37,38]. We examined pig genomic DNA to determine whether an aberrant methylation pattern of the H19 putative DMR occurred in adult cloned pigs compared with wildtype pigs ( Figure 2B). Southern blot analysis indicated that the H19 DMR in wild-type pigs showed three methylation patterns: a full methylation pattern (2.6 kb band), a partial methylation pattern (967 bp band), and an unmethylated pattern (689 bp band) ( Figure 2B). Our data suggest that the CpG methylation of the first EagI restriction site (near the CTCF1-binding site of H19) exists as a unique methylation pattern of the cloned pig genome. However, extremely aberrant methylation of H19 DMR observed in the cloned pig genomes, including the lack of a 967 bp band in the placenta (Pl) and a 689 bp band in the ear (Ea) of CP2 (arrows in Figure 2B). We further categorized the extent of methylation into two levels. Level-1 was determined by three probed bands (2643 bp, 967 bp, and 689 bp). Level-2 was determined by two probed bands (967 bp and 689 bp). In the level-1 methylation, CP1 muscle, CP2 muscle, CP2 placenta, and CP3 heart showed hypomethylation patterns. In the level-2 methylation, CP2 muscle, CP2 ear, and CP4 placenta showed hypermethylation patterns, but CP2 placenta and CP3 brain showed hypomethylation patterns. Furthermore, we designed a primer set for MS-PCR upstream of the first EagI (black arrowheads in Figure 2A) to confirm the methylation status in the H19 promoter region. The wild-type (WT) control panel showed a normal methylation pattern in this region; however, various tissues from cloned pigs exhibited aberrant methylation statuses compared to the wild-type pig. CP1 kidney, CP2 ear, CP3 liver, CP3 kidney, and CP4 liver showed extremely aberrant methylation in the putative DMR of H19 ( Figure 2C; also see Table S1).

Aberrant methylation of the IGF2 gene in various tissues from cloned pigs
The IGF2 gene is paternally expressed and located upstream of the H19 gene. The methylation status of the H19 DMR can concurrently affect the IGF2 DMRs. The putative DMR1 of the pig IGF2 gene was chosen as a target. It was estimated to be located in intron 3 ( Figure 3A) based on the human, mouse, and cattle genomes. Southern blot combined with a methylation-sensitive enzyme assay showed that the putative IGF2 DMR1 from wild-type pigs exhibited an unmethylated status, whereas several cloned pig tissues, such as CP1 liver, CP3 ear, CP4 placenta, and CP4 ear, displayed a slight methylation pattern ( Figure 3B). The hybridized signal at 1.18 kb indicates full methylation and the band at 0.68 kb indicates the absence of methylation. The second putative DMR of IGF2 (DMR2), which was located between exon 8 and exon 9, was further analyzed ( Figure 3C). COBRA showed that this putative IGF2 DMR was differentially methylated in wildtype pig tissues ( Figure 3D). The methylation score (%) of the COBRA data were calculated as described previously [34], and any IGF2 DMR2 in cloned pig tissues containing more than 610% methylation changes compared with the wild-type pig was considered aberrantly methylated ( Figure 3D). For example, liver tissues of CP1, CP2, and CP3 showed a hypermethylated pattern in the IGF2 putative DMR2 region, but CP4 liver did not. Taking together all of the examined tissues from the cloned pigs except the placenta and umbilical cord, 40% showed a hypermethylated IGF2 pattern, 0% showed a hypomethylated pattern, and 60% showed a normal methylation pattern in the testing sample size (see Tables S2 and S5).

Aberrant methylation of the INS gene in various tissues from cloned pigs
The INS gene is a paternally expressed gene. The putative pig INS DMR is located in the promoter region ( Figure 4A). The methylation status of the INS putative DMR in cloned pigs was analyzed by COBRA-PCR and BstUI digestion. In this study, the definition of normal methylation was based on the methylation status of the INS putative DMR in wild-type pig tissues ( Figures 1B  and 4B). Several cloned pig tissues, including CP1 placenta, CP3 ear, CP4 ear, and CP4 umbilical cord, showed almost complete methylation of this putative INS DMR ( Figure 4B; also see Table  S3). Subsequently, bisulfite sequencing was performed to confirm the methylation pattern of these aberrant regions in the above tissues. The overall methylation percentage of CpG sites 1-20 in wild-type ear, CP4 liver, and CP3 ear were 75.7%, 71.8%, and 84.6%, respectively. Methylation percentage of the CpG sites 13-20 in wild-type ear, CP4 liver, and CP3 ear was 43.8%, 48.2%, and 75.9%, respectively. Moreover, the methylation percentage of CpG sites 15-20 in wild-type ear, CP4 liver, and CP3 ear was 48.8%, 36.9%, and 71.4%, respectively ( Figure 4C). The INS putative DMR had a significantly hypermethylated pattern in CP3 ear tissue when compared with wild-type. Taking together all of the examined tissues from the cloned pigs except the placenta and umbilical cord, 50% showed a hypermethylated INS pattern, 5% showed a hypomethylated pattern, and 45% showed a normal methylation pattern in the testing sample size (see Table S5).

Aberrant methylation of the IGF2R gene in various tissues from cloned pigs
Various tissues of cloned porcine were investigated for their methylation status at the putative DMR within IGF2R intron 2 ( Figure 5A). The control panel showed a differentially methylated status, ranging from 25.2% to 76.8% in wild-type tissues analyzed by COBRA ( Figure 5B). Aberrant methylation patterns were observed in the IGF2R putative DMR of cloned pigs; for example, the livers of CP1 and CP2 showed a hypermethylated pattern, while a hypomethylated pattern existed in CP3 and CP4 ( Figure 5B, also see Table S4). Taking together all of the examined tissues from cloned pigs except the placenta and umbilical cord, 15% showed a hypermethylated IGF2R pattern, 45% showed a hypomethylated pattern, and 40% showed a normal methylation pattern in the testing sample size (see Table  S5).

Effects of DNA methylation on the expression of the analyzed genes
Semi-quantitative RT-PCR was used to verify the aberrant expression status of the four analysed genes. For the putative DMR of the H19 gene, hypermethylation was shown in CP2 ear tissue ( Figure 6A). Thus, the mRNA expression of H19 was absent in the ear tissue of CP2 ( Figure 6B and 6C). Similar results were found for the putative INS DMR in CP4 ear tissue as well as in the putative IGF2 DMR2 in CP3 ear tissue ( Figure 6B and 6C). To validate the cause-effect relationship of putative DMR hypermethylation in the downregulation of imprinted genes, a demethylation study was performed. Treatment of pig ear fibroblasts with different concentrations (0.5-2.0 mM) of a demethylating agent, 5-aza-29-deoxycytidine (5-aza-dc), resulted in a reduction of the methylation levels of the IGF2, H19, and INS putative DMRs, thereby restoring the mRNA expression of these imprinted genes (see Figure S1).

Discussion
Our data reveal that cloned pigs exhibit widespread defects in the methylation of imprinted genes, in addition to morphological abnormalities. All analyzed samples of CP2 and CP4 showed aberrant methylation patterns in the H19 and IGF2R putative DMRs, respectively (Table 1 and Table S6). The aberrant methylation may occur during early embryonic development. We also found that aberrant methylation existed even though the cloned pigs grew into adult animals. These results indicate that the SCNT cloning process may disrupt the normal epigenetic reprogramming during embryogenesis but that some types of mechanisms are still maintained during normal fetal development. The mRNA expression levels of imprinted or non-imprinted genes can also be affected. The cloned pigs died naturally and had a shorter life-span than wild-type pigs. In these specific cloned pigs, physiological abnormalities occurred frequently in heart and bone (see Figure S2).
The cloned pigs used in this experiment exhibited a lower birth weight than wild-type pigs (Table 1). CP1 exhibited the oval foramen phenomenon and delay in the development of the limb skeleton. A significant reduction in birth weight was observed in CP2, which also suffered from gastric ulcer, peritonitis, and pericarditis, whereas CP3 showed a valvular disease and fibrosis of the heart. CP4 showed pneumonia and scoliosis. The SCNT processes in cloned pigs may cause the loss of regulation in gene expression or aberrant organ development. As the results of our methylation study suggest, the putative DMR of H19 was dramatically modified in CP2 ear and placenta (Figure 2A). The birth weight of all four cloned pigs was reduced. CP2 showed the lowest birth weight (42% body weight lower than wild-type), and the degree of aberrant methylation in CP2 was higher in the paternally imprinted genes (80%) than in the maternally imprinted genes (50%) ( Table 1). The opposite aberrant methylation pattern was presented in the other three cloned pigs. These results may be associated with the fact that paternally imprinted genes tend to inhibit embryonic growth, while the maternally imprinted genes tend to promote embryonic growth [39].
The cytoplasm of enucleated oocytes may possess the critical factors that control the normal reprogramming process [40]. Whereas incomplete demethylation changes are observed in cloned cattle [41], demethylation occurs more frequently in cloned pigs than in other species [42]. Previous studies have examined the methylation statuses of repetitive sequences in the cloned pig genome. The typical demethylation pattern is observed in the 2-cell through the blastocyst stage of porcine embryo development [42]. Wei et al. [13] also confirmed that demethylation status occurs in the first two cell cycles of the cloned porcine due to the lack of Dnmt1 in the oocyte, which affects the methylation of H19 and IGF2. Kang et al. [41] observed that the satellite genes of NT bovines were hypermethylated. During DNA methylation reprogramming, the methylation statuses of NT sheep are higher than in vivo embryos from the 2-cell to the 16-cell stage [43]. However, to date, there are few studies on the aberrant methylation of imprinted genes in cloned pigs. We propose that the typical demethylation pattern of the cloned pig genome would affect the methylation pattern in the imprinted genes. Repetitive sequences maintain the donor-type methylation status in cloned rabbit and bovine embryos. Thus, the cause of demethylation in cloned embryo genomes seems to be determined by the recipient oocyte but not by the donor cell [44]. This finding suggests that the mechanisms involved in epigenetic reprogramming are species-specific. As expected, the reduction of body weight in this study agreed with the results of previous reports [6,18]. Different strains of mouse oocytes with different epigenetic inheritance show differential cloning efficiency [40], which suggests that the factors that exist in oocytes may be critical to maintaining proper reprogramming processes during the embryonic development of cloned animals. Recent reports also suggest that the factors in the normal fibroblast cytosol can restore the aberrant imprinting status of tumor cells to the normal pattern [45]. Therefore, the procedure of nuclear transfer may prevent the accumulation or activity of factors that regulate the normal reprogramming process.
The methylation status of mouse and human H19 DMRs shows a semi-methylation pattern. Interestingly, a unique methylation pattern of the initial CpG sites of the H19 putative DMR existed in cloned pig genomes near the CTCF1 P1 site ( Figure 2B). The CTCF binding sites (P1, P2, and P3) in the H19 DMR are sensitive to methylation changes in cloned porcine genomes or human cancers [46,47]. The aberrant pattern observed in cloned porcine showed either hyper-or hypo-methylation in the H19 CTCF3 binding site, but there was no significant change in methylation status in the CTCF1-or CTCF2-binding sites [44]. Aberrant methylation of DMRs results in the loss of regulation of imprinted genes [48][49][50]. Therefore, the reprogramming process may be disrupted by the nuclear transfer technique, which may lead to LOI in the H19 gene. Moreover, mRNA expression of other genes that are controlled by events downstream of H19 may become dysregulated in a normal expression pattern [31]. Here, the aberrant methylation pattern in cloned pigs frequently appeared near the CTCF1 P1 site of the H19 DMR.
IGF2 and H19 are reciprocally imprinted genes in a boundary regulation phenomenon. The imprinting pattern (uniparental RNA expression and DMR characteristics) of IGF2 and H19 has also been confirmed in the porcine genome [51,52]. Based on Southern blotting, the putative DMR1 region of pig IGF2 showed no methylation in wild-type pig tissues ( Figure 3B). However, the CpG sites of exon 9 in the IGF2 putative DMR2 showed a differential methylation phenomenon ( Figure 3D). In the IGF2 putative DMR1, a few tissues of cloned pigs (CP4 placenta and ear, CP1 liver, and CP3 ear) showed some methylation. In the IGF2 putative DMR2, various tissues exhibited abnormal methylation in cloned pigs when compared with wild-type. These data indicate that DMR2, but not DMR1, exhibits a differential methylation phenomenon and easily acquires aberrant methylation in cloned pigs. These data agreed with previous findings that the putative DMRs of pig H19 and IGF2 exhibit a specific parental methylation at the 2-cell and blastocyst stages [53,46].
The presence of a variable-number tandem repeat (VNTR) in the 59 region of the INS promoter suggests that INS is potentially an imprinted gene, and its imprinting status could be associated with insulin-dependent diabetes mellitus [54]. The paternal   expression pattern of INS has also been found in the yolk sac of humans and mice [55,56]. The chromosomal location of INS is the same in the porcine, human, and mouse genome, i.e., upstream of IGF2 [57,58]. The promoter of the porcine INS gene had highdensity CpG sites located just upstream of exon 1 ( Figure 4A). By COBRA, this INS putative DMR was identified to contain a differential methylation pattern in the wild-type pig genome ( Figure 1B). Either hypermethylation or hypomethylation of this INS putative DMR occurred in the cloned pig tissues (Table S7). The precise location of the DMR was further confirmed by bisulfite sequencing, which showed that it existed in CpG sites 15 through 20 ( Figure 4C).
IGF2R and IGF2 are imprinted genes with opposing functions [59]. These genes are also reciprocally regulated during fetal growth. Loss of IGF2R imprinting correlates with LOS in sheep [50,60]. We used comparative sequencing analysis to define the organization of the pig, mouse, bovine, and human IGF2R putative DMRs and found that it was located in intron 2 of pig IGF2R gene ( Figure 5A). A normal differentially methylated pattern was shown in the IGF2R putative DMR of wild-type pig tissues. However, a largely hypomethylated phenomenon was detected in the tissues of cloned pigs (Tables S7 and S8).
In this study, the four cloned pigs we studied had many defects, especially in the heart, lung, and gastric tissues (Table 1). These defects may have been associated with the loss of regulation of DNA methylation or mRNA expression during embryonic development [45,61]. The delay in limb bone growth in CP1 may also have resulted from the NT process. A previous study of cloned pigs also showed abnormal development in the elbow joint bones [62]. Three hypermethylated tissue samples, CP2 ear (in the H19 putative DMR), CP4 ear (in the INS putative DMR), and CP3 ear (in the IGF2 putative DMR), were tested for the mRNA expression of the hypermethylated genes. Low expression levels of these imprinted genes were consistently observed in cloned pigs compared with the wild-type pigs ( Figure 6).
Aberrant DNA methylation induced obvious abnormalities in our SCNT-derived embryos and their offspring. The abnormal phenomenon in CP2 and CP3 included an enlarged tongue (macroglossia), which is similar to Beckwith-Wiedemann syndrome, and is caused by aberrant methylation and expression of IGF2 in humans [63]. Furthermore, the enlargement of the right ventricle in cloned pigs has also been reported [6]. CP3 and CP4 also suffered from this heart defect.
In conclusion, we confirmed that the putative DMRs of H19, IGF2, INS, and IGF2R in the wild-type porcine genome show differential methylation patterns. Specific proportions of epigenetic aberrations, either hypermethylation or hypomethylation, were observed in the adult tissues of the examined cloned pigs. These data will help further our understanding of the importance of imprinted genes during the development of normal or cloned swine and contribute to improvements in cloning techniques. The vertical pathological dissection of the right side of the heart showed defects in CP2. An anatomically normal wild-type heart is shown in the left panel (WT heart). The CP2 heart exhibited aberrant valve development and pericarditis. The heart also showed right ventricular hypertrophy and heart hypoplasia. (B) The femur of CP1 was shorter and obviously mineralized in the epiphysis compared with a femur from a WT pig of the same age. The growth of the radius of CP1 was retarded compared with WT. (TIF)