Fig 1.
PEG3 binding to the mouse Oxtr locus.
(A) The 40-kb genomic region surrounding the Oxtr locus is shown using the UCSC genome browser. The black boxes indicate the 4 exons of Oxtr, while the arrows indicate the transcriptional direction of Oxtr. The identified potential target region of PEG3 is also indicated with a vertical thick line (Oxtr-ChIP-BS4). Processed ChIP-seq results from WT and KO MEF cells (WT_Peg3.bw and KO_Peg3.bw) were uploaded onto the genome browser to show the enrichment levels by PEG3-ChIP. (B) Individual ChIP experiments. The chromatin from MEF and adult brains were immunoprecipitated with anti-PEG3 antibody. The subsequent DNA set, including Input, Negative without any antibody (Neg), the immunoprecipitated DNA with anti-PEG3 antibody (PEG3 IP), were used as templates for a fixed number of PCR cycle to test the in vivo binding of PEG3 to the target region (ChIP-BS4). The relative enrichment levels against Input were measured with qPCR experiments and further compared between Neg and PEG3 IP. Asterisks represent statistical significance of the observed differences between Neg and PEG3 IP (*, p value < 0.05).
Fig 2.
Potential DNA-binding sites of PEG3 within the Oxtr locus.
(A) The 20-kb genomic region of Oxtr was scanned with the consensus DNA-binding motif of PEG3, 5'-N-N-G-G-[C/G]-N-[C/G]-T with N being any nucleotide bases, identifying 5 potential binding sites, BS1-BS5. These binding sites are localized within three genomic regions with DNA hypomethylation, which are indicated with blue bars. (B) Competitive EMSA (Electrophoretic Mobility Shift Assay). The P32-radiolabeled oligonucleotide duplex probe derived from the Pgm2l1 locus was mixed with mouse brain nuclear extracts (lane 2), and further competed against a set of 5 potential binding sites (lane 3–7). The relative ratio of the labeled Pgm2l1 probe to BS1—BS5 oligonucleotide duplexes was 1 to 200. The sequences of the oligonucleotides used for this assay are shown on right. The potential binding sites are indicated with either underlines or red fonts. The sites with red font indicate the confirmed DNA-binding sites through EMSA, whereas the sites with underlines indicate the predicted but unconfirmed sites. The bold-typed nucleotide bases within each site represent the critical bases for PEG3 binding.
Fig 3.
Schematic representations of mutant alleles and breeding schemes.
(A) Targeted alleles of Oxtrvenus and Peg3CoKO. The 3rd exon of Oxtr has been replaced by an exogenous construct expressing the reporter Venus, thus the endogenous locus of Oxtr expresses the reporter Venus instead of the endogenous gene product, oxytocin receptor. In the case of Peg3CoKO, the 5th intron of Peg3 has been inserted with an expression cassette containing β-Gal and NeoR along with Poly(A) signals, thus causing transcriptional truncation and subsequent loss of the PEG protein. (B) Breeding scheme. Female heterozygotes for Oxtrvenus/+ were crossed with male heterozygotes for Peg3CoKO/+. The subsequent breeding results were presented with individual litter sizes, and also with a table summarizing the frequency of the four genotypes on right.
Fig 4.
Up-regulation of Oxtr in the mammary gland of nursing females with Peg3+/CoKO.
(A) At one day postpartum, a set of mammary glands were harvested from the following two females: OxtrVenus/+; Peg3+/+ (WT) and OxtrVenus/+; Peg3+/CoKO (KO). The spatial expression patterns of Oxtr were monitored through the signal of the reporter Venus, and the intensity of signals were also compared between WT and KO. The top panels represent the excretory ducts, while the bottom panels show the secretory acini of the mammary gland of the WT and KO mice. Lining epithelial cells show higher levels of Oxtr immunofluorescence in KO than in WT. The scale bars for both panels are 20 μm. (B) qRT-PCR analyses. Total RNA was isolated from the mammary epithelial cells of female littermates with the following genotypes: Oxtr+/+; Peg3+/+ (WT) and Oxtr+/+; Peg3+/CoKO (KO). These RNA were used for cDNA synthesis, which were subsequently used for qRT-PCR analyses. The expression levels of each gene were compared between WT and KO, and the observed differences were presented with statistical significance (*, p <0.01 and **, p < 0.001).
Fig 5.
Up-regulation of Oxtr in the hypothalamus of nursing females with Peg3+/CoKO.
(A) At one day postpartum, a set of brains were harvested from the following two females: OxtrVenus/+; Peg3+/+ (WT) and OxtrVenus/+; Peg3+/CoKO (KO). The harvested brains were fixed, sectioned and immunostained with anti-OXT antibody (blue) and anti-PEG antibody (red). The expression of Oxtr was monitored through the signal of the reporter Venus (green). The top panels represent a low power view of the hypothalamic region, while the bottom panels show a high power view of the boxed regions on top panels. Oxtr-Venus+ cells were localized along the 3rd ventricle and almost all co-localized with Peg3 in WT mice. Oxtr-Venus expression in KO mice was much broader and extended into the medial preoptic area. The scale bars for top and bottom panels are 0.5 mm and 200 μm, respectively. AC: Anterior Commisure, 3V: 3rd Ventricle, MPOA: Medial PreOptic Area. (B) qRT-PCR analyses. Total RNA was isolated from the hypothalamus of female littermates with the following genotypes: Oxtr+/+; Peg3+/+ (WT) and Oxtr+/+; Peg3+/CoKO (KO). These RNA were used for cDNA synthesis, which were then used for qRT-PCR analyses. The expression levels of each gene were compared between WT and KO, and the observed differences were presented with statistical significance (*, p <0.01 and **, p < 0.001).