Disruption of imprinted expression of U2afbp-rs/U2af1-rs1 gene in mouse parthenogenetic fetuses.

The present study shows that the U2afbp-rs gene, a paternally expressed imprinted gene, is activated and expressed in a biallelic manner from maternal alleles in parthenogenetic mouse fetuses on day 9.5 of gestation. The mean expression was detected to be 88% (31-134%) of that in control biparental fetuses, using real-time quantitative reverse transcription and polymerase chain reaction. This disrupted expression of the gene was associated with changes in the chromatin structure but not with the methylation pattern in the regulation region. The present results show that parental specific expression of imprinted genes is not always maintained in uniparental embryos. This suggests that both parental genomes are necessary to establish parental specific expression of the U2afbp-rs gene.

tor (Igf2r) (15) genes were first identified as imprinted genes in 1991, around 40 of these types of genes have been identified. However, the genes identified thus far constitute only a small portion of the number of imprinted genes, which is estimated to be 1-2% of all genes. Recently, an inclusive attempt was made to identify imprinted genes by the subtraction method using parthenogenetic and androgenetic mouse embryos (16,17). Some imprinted genes, such as Peg1/Mest (16), Peg3 (18), and Meg/Grb10 (17) were successfully identified, but the number is smaller than expected. This limited success in identifying genes may suggest that the expression of some imprinted genes is unregulated in uniparental embryos.
The U2afbp-rs gene is a paternally expressed imprinted gene located on mouse chromosome 11 (19 -21). This gene was identified by a systematic screening performed using a restriction landmark genome scanning procedure for loci subject to imprinted methylation (19 -21). The gene is expressed in all tissues throughout development, with expression restricted to the paternal allele and predominantly in the brain of the adult mouse (21), but its function is unknown. In this study, we show that the paternally expressed U2afbp-rs gene is expressed abundantly from maternal alleles in parthenogenetic mouse fetuses. This suggests that the parental genome is necessary for establishing and/or maintaining the paternal expression of the gene.

MATERIALS AND METHODS
Production of Uniparental Embryos-B6CBF1 (C57BL/6N ϫ CBA) mice were used as oocyte and sperm donors to produce embryos. Oocytes at metaphase II were collected from superovulated mice and artificially activated with SrCl 2 (Sr 2ϩ ) (22,23). To produce diploid parthenogenetic embryos, during the culture in Sr 2ϩ medium and following the first 4 h of culture in M16 medium, 5 g/ml cytochalasin B was added to these media to inhibit a second polar body extrusion. Androgenetic embryos were produced by in vitro fertilization of enucleated oocytes according to the procedure reported previously (24). When required, pronuclear transfer (25) was performed to produce diploid androgenones in monospermic embryos. Control biparental embryos were produced by in vitro fertilization. The obtained blastocysts were transferred to the uterine horns of CD-1 females on day 3 of pseudopregnancy (2.5 days postcoitum), and fetuses were recovered at day 9.5 of gestation.
Expression Analysis-Total RNA was isolated from the fetuses at day 9.5 of gestation using an SV Total RNA Isolation System (Promega). The first strand of cDNA was synthesized from 1 g of total RNA from each fetus by SuperScript II reverse transcriptase (Life Technologies, Inc.) according to the manufacturer's instructions. The cDNA was subjected to PCR, which was carried out in a 50-l reaction buffer containing 1.25 units of Taq DNA polymerase (Takara Shuzo Co., Ltd., Japan), 1 pmol of each primer, 1.5 mM MgCl 2 , and 250 M dNTPs. The amplification consisted of a total of 35 cycles at 95, 63, and 72°C, respec-* This work was supported by grants from the Ministry of Education, Science, Culture and Sports of Japan, and the Japanese Society for the Promotion of Science. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. § Present address: CSK Research Park Inc., 1-135 Komakado, Gotemba-shi, Shizuoka 412-8513, Japan.
tively, for 60 s in a GeneAmp PCR System 2400 (PerkinElmer Life Sciences). Primers used were 5Ј-TACCACAGATAACCACGGATAC-CTG-3Ј and 5Ј-ATAGGCCTGCCCATGCAGTTA-3Ј. Quantitative analysis of gene expression was performed by means of real-time quantitative reverse transcription and polymerase chain reaction (RT-PCR). By adding a fluorescently labeled probe (5Ј-GGATAACTGGATAGACG-GATAACTGCGCA -3Ј) to the PCR reaction mixture, the quantity of PCR products was detected by monitoring the luminous intensity with an ABI PRISM 7700 Sequence Detection System (PerkinElmer Life Sciences).
To identify the activated allele of the U2afbp-rs gene, biparental and parthenogenetic embryos were constructed in which these embryos contained diploid genomes from B6CBF1 and JF1 mice, respectively. Polymorphisms of the U2afbp-rs gene between B6CBF1 and JF1 mice were detected by restriction enzyme fragment length polymorphism analysis using HpaII (Fig. 1).
Analysis of Methylation in the Promoter and Open Reading Frame (ORF) Regions-In an upstream CpG island (regions I, II, and IV) and the ORF (region III) of the U2afbp-rs gene (26) in parthenogenetic and normal biparental fetuses, the methylation pattern was analyzed by Southern hybridization using genome DNAs that had been digested by methyl-sensitive restriction enzymes (Fig. 1). Extracted genome DNAs were digested by DraI and HindIII to obtain DNA fragments containing the CpG islands or by EcoT14I and PshBI for ORF. Then the DNA fragments were digested by HpaII or HhaI and subjected to electrophoresis (1.5% NuSieve:agarose (3:1)). The DNAs were transferred to a nylon membrane (Hybond-Nϩ) and introduced into Southern hybridization. Hybridization was performed for 16 h at 60°C using AlkPhos Direct Labeling Reagent (RPN3680) (Amersham Pharmacia Biotech). The probes used were derived from PCR products using primer sets of 5Ј-GGTTTACCTGGTAAAGCCCTT-3Ј and 5Ј-ATAGGCCTGCCCATG-CAGTTA-3Ј for the CpG island and 5Ј-AAGATGGTACGCAGGAC-GAC-3Ј and 5Ј-TTTGTGGACCTGGACTGCTG-3Ј for the ORF and alkaline phosphatase labeled by the reagent. The hybridized membrane was treated with CDP-Star (RPN3682) (Amersham Pharmacia Biotech), and the intensity of the band was measured by an LAS1000 image analyzer (Fuji Photo Film Co., Ltd., Japan).
Analysis of Chromatin Structure-Chromatin conformation of the U2afbp-rs gene in parthenogenetic and control biparental fetuses was determined by DNase-I hypersensitivity assay, using DraI-EcoRI fragments of the genome DNAs (Fig. 1). Extracted genome DNAs were treated with a solution containing 5, 10, 20, 40, 80, and 160 units/ml DNase-I at 4°C for 5 min, as described previously (6). The DNAs were then digested with DraI and EcoRI and introduced into Southern hybridization, using the same probe for methylation analysis of the promoter region.

RESULTS
To confirm that the paternal expression of U2afbp-rs is maintained in uniparental fetuses, we first examined the expression of the U2afbp-rs gene in each parthenogenetic and androgenetic fetus at 9.5 days by semi-quantitative RT-PCR. The expression of the U2afbp-rs gene was unexpectedly detected in 38 of 39 parthenogenetic fetuses. The expression level was determined to be roughly 10 -50% of that in the control biparental fetuses. Analysis by real-time quantitative RT-PCR, which is able to detect the precise expression level in an individual fetus, showed that expression of the U2afbp-rs gene by parthenogenetic fetuses ranged from 31 to 134% (mean 88%) of that in controls (56 -124%, Fig. 2). In the controls (without RT), no signal of PCR product was detected. The expression allele was detected using parthenogenones constructed from JF1 and B6CBF1 mouse genomes that contain polymorphic sequences in the exon region between the strains (Fig. 3). After digestion of the RT-PCR products (243 bp) with HpaII, two bands (194 bp for B6CBF1 and 177 bp for JF1 alleles) were detected, showing that the gene was expressed from both alleles. It was also confirmed that, in the control biparental fetuses, the gene was not expressed from the maternal allele. These results suggest that the regulation mechanism for paternal specific expression of the U2afbp-rs gene is not established in parthenogenetic fetuses. The quantitative analysis in androgenetic fetuses, which have two paternal alleles, showed that expression of the U2afbp-rs gene was almost equal (mean 83%, 45-109%) to that in the controls (Fig. 2), suggesting the presence of an underlying dose-compensation mechanism.
Generally, parental expression of imprinting genes is closely related to the methylation level of the CpG islands in the promoter regions (6 -9). To investigate whether or not this phenomenon is seen in the expression of the U2afbp-rs in parthenogenetic fetuses, the methylation levels in the promoter and ORF regions of the U2afbp-rs gene were analyzed by Southern blot analysis (Fig. 4, a and b). When the regions are undermethylated, 1317-bp DraI-HindIII fragments for the promoter and 1207-bp EcoT14I-PshBI fragments for the ORF are digested into fragments of 252 bp by HpaII, 290 bp by HhaI, and 635 bp by HpaII, respectively. MspI digestion of the regions confirmed a positive effect of the HpaII digestion. These results show that in both regions in the control fetuses, two bands of digested and nondigested fragments, using HpaII and/or HhaI were formed, suggesting that only the maternal alleles were possibly hemimethylated. However, in the parthenogenetic fetuses, no sequence was digested in the HpaII nor in the HhaI site, indicating that both alleles were entirely hypermethylated despite considerable expression of the gene. This suggests that other regulation factors may be involved in the disruption of the regulated expression of the U2afbp-rs gene in the parthenogenetic mouse fetuses.
Because chromatin structure is also related to gene expression (27,28), chromatin conformation of the U2afbp-rs gene was determined by the DNase-I hypersensitivity assay (Fig. 5). The DNase-I hypersensitive sites in the DraI-EcoRI fragment were detected in both control and parthenogenetic fetuses (arrowheads in Fig. 5) when the fragments were digested by more than 20 units/ml DNase-I. The status of DNase-I hypersensitive sites of parthenogenetic fetuses was similar to that of the controls, suggesting that the chromosomes of the parthenogenetic fetuses have an open conformation. This perhaps caused the unregulated expression of the U2afbp-rs gene in the parthenogenetic fetuses. DISCUSSION We reported previously that expressions of mouse paternally expressed imprinted genes Peg1/Mest (16), Igf2 (14), Peg3 (18), and Snrpn (29) are not properly detected by RT-PCR in mouse parthenogenetic fetuses at day 9.5 of gestation (30). Further analysis by real-time quantitative RT-PCR confirmed parental specific expression of the Igf2 gene in both the parthenogenetic and androgenetic fetuses (31). These previously published observations show that parental expression of imprinted genes is generally maintained in uniparental fetuses. The current study demonstrated, however, that anti-theoretical gene expression of the paternally expressed imprinted gene U2afbp-rs is caused in the parthenogenetic fetuses. The expression reached an av-

FIG. 2. Quantitative analysis of U2afbp-rs gene expression in control biparental (C), parthenogenetic (P), and androgenetic fetuses (A) by real-time quantitative RT-PCR.
Ten samples per group were examined, and the expression level of each fetus is indicated with a dot. The average of the expression levels in the control fetuses was prescribed as a standard (100%). Shaded boxes indicate the averages of the expression levels (parthenogenetic fetuses, 88% and androgenetic fetuses, 83%, respectively).

FIG. 3. The expression allele of the U2afbp-rs gene was analyzed by restriction enzyme fragment length polymorphism analysis in biparental and parthenogenetic fetuses at day 9.5 of gestation. The marker used was a 25-bp DNA
Step Ladder (Promega, G4511). RT-PCR products (237 bp) from B6CBF1 and JF1 alleles were, respectively, digested into 194-and 177-bp fragments by HpaII. In B6CBF1 ϫ JF1 fetuses, expression was theoretically detected only from paternal JF1-derived alleles. In parthenogenetic fetuses, which contained diploid genomes from both B6CBF1 and JF1 mice, gene expression was detected from both alleles.

FIG. 4. Southern analysis of the DNA methylation pattern of the U2afbp-rs gene in control biparental (c) and parthenogenetic fetuses (p) at day 9.5 of gestation.
In the control fetuses, the sequences in the promoter (a, DraI-and HindIII-digested fragments, 1317 bp) and the ORF (b, EcoT14I-and PshBI-digested fragments, 1207 bp) regions were digested by HpaII (252-bp fragments for the promoter and 635-bp fragments for the ORF) or HhaI (290-bp fragments for the promoter) (see also Fig. 1). These findings indicate that the sequences of both regions were hypomethylated and that hypomethylation was related to the gene expression pattern. In contrast to the control fetuses, the sequences in both regions in the parthenogenetic fetuses were not digested by HpaII or HhaI and showed the status of hypermethylation regardless of the gene expression. erage level of 88%, ranging from 31 to 134% of that in biparental fetuses. Notably, the expression occurred at a level very different from that of leakage, suggesting that both maternal and paternal alleles are necessary for the regulated parentally specific expression of imprinted genes. U2afbp-rs was identified by the restriction landmark genome scanning procedure based on loci subject, but the subtraction method (16,17) failed to trap the U2afbp-rs gene, perhaps because the expression of the gene originates from the maternal alleles of parthenogenones.
During oocyte growth, the maternal genome is epigenetically modified, and maternally specific expression patterns of imprinted genes are established as a result of the primary imprinting. To demonstrate this, we obtained parthenogenetic mouse fetuses developed from constructed oocytes that contained nongrowing and fully grown oocyte-derived genomes that were derived from immature oocytes at diplotene stage and mature oocytes at metaphase II, respectively. The experiments clearly demonstrated that these fetuses develop to day 13.5 of gestation (32), and paternally expressed imprinted genes are expressed from the nongrowing oocyte alleles (30). It has also been found that U2afbp-rs is activated in the nongrowing oocyte allele in such parthenogenetic fetuses 2 . Before fertilization, the maternal allele of the U2afbp-rs gene has been imprinted to be repressed by the methylation of the CpG island in specific sites of its promoter regions (regions II and IV). A wide range of methylation occurs in the promoter region (region I) at the two-cell stage. This methylation may play a crucial role in determining allele-specific expression, although maternal expression was maintained at later stages (26). The question raised by the present results is whether the expression of U2afbp-rs from maternal alleles is associated with the demethylation of the regulatory region. Analysis of methylation status by Southern hybridization and HpaII-PCR (data not shown) showed that the region was hypermethylated and that expression of the U2afbp-rs gene was from the maternal allele in the control fetuses. A previous report (33) suggested that methylation levels in imprinted genes Igf2 (14), Igf2r (15), Snrpn (29), and H19 (34,35) are compensated for gene dosage in parthenogenetic and androgenetic mouse blastocysts. In this study of the U2afbp-rs gene in parthenogenetic and androgenetic fetuses, the dose compensation was also observed in the methylation level but not in the expression level.
To obtain further insight, we analyzed the chromatin conformation of the gene by DNase-I sensitivity assay (36 -38) and showed that the conformation was opened in the parthenogenones. This suggests that the expression of the U2afbp-rs gene in maternal alleles of the parthenogenones occurs because of the opened chromatin structure and that paternal alleles could be involved in the formation of a compacted chromatin structure but not in the methylation status. This discrepancy between methylation status and chromatin structure is not observed in the biparental embryos. In the mouse U2afbp-rs gene, the methylation pattern and the chromatin structure have also been correlated with the paternal specific expression of this gene (27,28).
A recent report on mechanisms underlying the chromatin conformation showed that DNA methylation causes histone deacetylation to modify chromatin and silence genes (39). Thus a further experiment is required to clarify the unregulated expression of the U2afbp-rs gene seen in parthenogenones. The explanation for the unregulated expression of the U2afbp-rs gene in the parthenogenetic fetus has not been established. One possible reason is that de novo gene modification by pa-ternally derived factors, which is necessary for the establishment of maternal imprinting, does not occur in the parthenogenones. The trans-acting factors may affect the ability to bind with the transcription factors to regulate the domain of the gene. A more likely possibility, however, is that a dose-compensation mechanism underlies this phenomenon. The finding that expression of the U2afbp-rs gene in androgenones was similar to that in the controls may support this conjecture, suggesting that the U2afbp-rs gene transcript may itself regulate the transcription activity.
Many imprinted genes, such as Igf2, Igf2r, Peg1/Mest, Peg3, and H19, maintain their parental specific expression in uniparental embryos (13,16,18,31,35), that is, in parthenogenetic/ gynogenetic and androgenetic embryos. However, the present results show that this is not the case for the U2afbp-rs. This suggests that both parental alleles are necessary for the establishment and maintenance of the parentally specific expression pattern. The molecular mechanisms underlying the disruption of imprinted status in parthenogenetic embryos remain to be clarified. Further investigation into this kind of disruption could provide greater insight into the characteristics of imprinting genes.