Figure 1.
Removal of the PEG3 protein results in the up-regulation of Zim1.
(A) Schematic representation of the genomic structure of paternally expressed Peg3 and maternally expressed Zim1. In the mutant allele, a 7.1-kb cassette containing a promoterless β-galactosidase (β-Gal) and human β-actin promoter-driven neomycin resistant gene (NeoR) has been inserted between exon 5 and 6 of Peg3. The inserted cassette is flanked by two FRT sites (open ovals), thus can be removed through FLP-mediated recombination. (B) Expression analyses of Zim1 using a set of female mouse embryonic fibroblast (MEF) cells that had been prepared through breeding female and male heterozygotes with their wild-type littermates. The MEF cells with the wild-type and the heterozygote with the paternal transmission of the mutant allele were used for qRT-PCR. (C) Imprinting tests of Zim1 using the neonatal brains of the F1 hybrid derived from the crossing of a male heterozygote C57BL/6J (B6) and a female PWD/PhJ (PWD) breeder. The RT-PCR products from Zim1 were digested with DraI, showing two parental patterns (lane 1 and 2) as well as the maternal-specific expression pattern from the two neonates with WT and KO (lane 3 and 4), respectively. Schematic representation of this imprinting test was shown above the gel picture.
Figure 2.
PEG3 binds to the zinc finger exon of Zim1.
(A) Genomic structure of Zim1 and the relative positions of the primer sets used for ChIP experiments. The 8 exons of Zim1 are indicated with vertical lines and boxes. (B) PEG3-ChIP experiments using the two sets of chromatins prepared from WT and KO (+/−p) MEF cells. The DNA from Inputs, Negative controls (Neg), and Immunoprecipitates with anti-PEG3 antibody (PEG3 IP) was used for PCR amplification. This series of ChIP experiments also included another locus, Pgm2l1, as a positive control since this locus is known to be bound by PEG3.
Figure 3.
PEG3 binds to both alleles of Zim1.
(A) Genomic structure of Zim1 and the relative positions of the primer sets used for ChIP experiments. (B) Peg3-ChIP experiments using the two sets of chromatins prepared from WT and KO (+/−p) neonatal brains. The DNA from Inputs, Negative controls (Neg), and Immunoprecipitates with anti-PEG3 antibody (PEG3 IP) was used for PCR amplification. This series of ChIP experiments also included another locus, Pgm2l1, as a positive control. (C) Allele test of PEG3-ChIP DNA. One set of chromatin derived from the F1 neonatal brain of the interspecific crossing between a male C57BL/6J (B6) and female PWD/PhJ (PWD) was used for ChIP experiments. The DNA representing Region 1 and 3 were digested with TaqI and DraI, respectively, to differentiate two alleles. The restriction enzyme sites on both DNA fragments are shown along with the schematic representation of these allele tests. The results indicated that the immunoprecipitated DNA at these regions were derived equally from both alleles.
Figure 4.
Reduced levels of H3K9me3 in the mutant cells lacking PEG3.
(A) Histone modification profiles on the Zim1 locus. The histone modification profiles of H3K4me3 and H3K9me3 derived from ES (upper) and whole brains (lower) are presented along with the exon structure of Zim1. (B) H3K9me3-ChIP using the two sets of chromatins prepared from WT and KO (+/−p) MEF cells. The DNA from Inputs, Negative controls (Neg), and Immunoprecipitates with anti-H3K9me3 antibody (H3K9me3 IP) was used for PCR amplification. This series of ChIP experiments included another locus, the imprinting control region of H19, as a positive control since the paternal allele of this ICR is known to be modified with H3K9me3. (C) The reduced levels of H3K9me3 on Zim1 were further analyzed with qPCR. Shown are the relative values of the immunoprecipitated DNA to the negative controls derived from MEF cells. Region 1 does not show any difference whereas Region 3 shows reduced levels of H3K9me3 in KO compared to those from WT.
Figure 5.
Restoring the protein levels of PEG3 down-regulates Zim1.
(A) Genomic structure of the mutant allele of Peg3 and FLP-mediated recombination scheme to restore the expression of Peg3. The inserted cassette is flanked by two FRT sites, thus can be removed by Flippase (FLP). (B) Three pools of KO MEF cells were transfected with the following constructs: no vector as a mock control (lane 1), a Green Fluorescent Protein (GFP) expression vector as a negative control (lane 2), and a FLP expression vector (lane 3). The total RNA isolated from these cells were analyzed with RT-PCR to measure the expression levels of β-actin, Zim1, and Peg3. The bottom panel shows genotyping results confirming FLP-mediated recombination of the mutant allele (Rev KO) and endogenous allele (WT) of Peg3. (C) The observed down-regulation of Zim1 was further analyzed using qRT-PCR.
Figure 6.
Paternally expressed Peg3 controls maternally expressed Zim1 as a trans factor involving H3K9me3.
Schematic representation for Peg3's functional role in transcriptional control of Zim1. The gene product of paternally expressed Peg3 binds to the zinc finger exon of maternally expressed Zim1 on both alleles, resulting in transcriptional repression through H3K9me3. The protein PEG3 may interact with some unknown proteins, such as KAP-1, to recruit SETDB1 for the H3K9me3 modification on the Zim1 locus. This role of Peg3 is independent of the maternal-specific expression of Zim1, thus the observed up-regulation of Zim1 is still derived from the maternal allele of the Peg3 mutant animals.