Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2009 Feb;136(3):437-48.
doi: 10.1242/dev.032060.

An in vitro ES cell imprinting model shows that imprinted expression of the Igf2r gene arises from an allele-specific expression bias

Paulina A Latos  1 , Stefan H Stricker, Laura Steenpass, Florian M Pauler, Ru Huang, Basak H Senergin, Kakkad Regha, Martha V Koerner, Katarzyna E Warczok, Christine Unger, Denise P Barlow
Affiliations

An in vitro ES cell imprinting model shows that imprinted expression of the Igf2r gene arises from an allele-specific expression bias

Paulina A Latos et al. Development. 2009 Feb.

Abstract

Genomic imprinting is an epigenetic process that results in parental-specific gene expression. Advances in understanding the mechanism that regulates imprinted gene expression in mammals have largely depended on generating targeted manipulations in embryonic stem (ES) cells that are analysed in vivo in mice. However, genomic imprinting consists of distinct developmental steps, some of which occur in post-implantation embryos, indicating that they could be studied in vitro in ES cells. The mouse Igf2r gene shows imprinted expression only in post-implantation stages, when repression of the paternal allele has been shown to require cis-expression of the Airn non-coding (nc) RNA and to correlate with gain of DNA methylation and repressive histone modifications. Here we follow the gain of imprinted expression of Igf2r during in vitro ES cell differentiation and show that it coincides with the onset of paternal-specific expression of the Airn ncRNA. Notably, although Airn ncRNA expression leads, as predicted, to gain of repressive epigenetic marks on the paternal Igf2r promoter, we unexpectedly find that the paternal Igf2r promoter is expressed at similar low levels throughout ES cell differentiation. Our results further show that the maternal and paternal Igf2r promoters are expressed equally in undifferentiated ES cells, but during differentiation expression of the maternal Igf2r promoter increases up to 10-fold, while expression from the paternal Igf2r promoter remains constant. This indicates, contrary to expectation, that the Airn ncRNA induces imprinted Igf2r expression not by silencing the paternal Igf2r promoter, but by generating an expression bias between the two parental alleles.

PubMed Disclaimer

Figures

Fig. 1
Fig. 1. Imprinted gene expression in differentiating ES cells
(A)The imprinted Igf2r cluster on mouse chromosome 17 spans 490 kb from Plg to Mas1 and contains four imprinted (light grey boxes) and three non-imprinted (dark grey boxes) genes. Arrows indicate transcription orientation. Note the antisense Airn promoter in Igf2r intron 2 (black oval) and the 108 kb Airn ncRNA that overlaps the Igf2r promoter (grey oval) and the 3′ part of Mas1. The asterisk marks the exon 12 single nucleotide polymorphism (SNP), as described in Fig. 2. (B)Expression of Oct4 (pluripotent cell marker; diamond), Gata4 (endoderm marker; triangle) and Fgf5 (embryonic ectoderm marker; square) as assessed by QPCR during retinoic acic (RA)-induced differentiation of CCE (left) and D3 (right) ES cells. The mean and s.d. of three independent replicates are shown. The day-5 value was set to 100, except for Oct4 where day 0 was set to 100. (C)RNase protection assay (RPA) showing parallel upregulation of Igf2r (probe Igf2rex34 protects 133 bp) and Airn (probe AirF3b protects 261 bp) expression during RA-induced differentiation of CCE (left) and D3 (right) ES cells. Cycl., cyclophilin A loading control (protects 105 bp); #, undigested probe; *, non-specific band. 0, undifferentiated (day 0) ES cells; (0), longer exposure of the day-0 track; 1-5, days RA treatment; Y−, minus ribonuclease; Y+, plus ribonuclease. (D)Expression kinetics, as determined by QPCR, of Igf2r (square) and Airn (diamond) during RA-induced differentiation of CCE ES cells. (E)Expression kinetics, as determined by QPCR, of Slc22a2 (diamond) and Slc22a3 (square) during RA-induced differentiation of CCE ES cells. (F)Intronic RNA FISH for Igf2r detects nascent transcription in 17% of day-5 differentiated CCE cells, of which 88% show a single-spot signal indicating imprinted expression.
Fig. 2
Fig. 2. Generation of the S12 allele carrying a SNP in exon 12 of Igf2r
(A)Mouse wild-type (wt) Igf2r locus, showing exons 1-27 (black boxes) and (above) an enlargement of intron 2 that contains the genetically defined 3.65 kb ICE and the CpG island lying immediately downstream of the Airn transcription start (arrow). Below is shown the targeting vector and targeted locus. The targeting vector contains the introduced SNP in exon 12 (that mutates a PstI site) and a 3.8 kb Tk-Neo selection cassette (stippled box) flanked by loxP511 sites (triangles) inserted into the BsrGI site in Igf2r intron 11. Arrowheads indicate PCR primers used for cDNA analysis. MEi/X12, Southern blot probes; B, BglII; Bm, BamHI; Bs, BsrGI; E, EcoRV; M, MluI; Pc, PacI; P, PstI; AirF3b, RPA probe as used in Fig. 1C. (B)Southern blot showing homologous recombination and removal of the selection cassette after Cre recombination (BglII+BsrGI digest plus probe X12). (C)Parental origin of targeted allele analysed by pulsed-field gel electrophoresis (PFGE). ES cell DNA containing the selection cassette was digested with EcoRV or EcoRV+MluI and hybridised with probe MEi. In this assay, EcoRV generates a 50 kb wild-type and a 34.4 kb targeted allele (the selection cassette carries additional EcoRV sites), and a diagnostic 21.9 kb band is generated from a paternally targeted allele but not from a maternally targeted allele after digestion with the methyl-sensitive MluI enzyme. The blot shows that a EcoRV+MluI (E/M) digest of maternally targeted (S12/+) cells generates a 34.4 kb maternal targeted band (MluI does not cut the maternally methylated ICE) and a 37 kb wild-type paternal band (MluI cuts the unmethylated paternal ICE); however, these bands are not separated on the gel shown. An EcoRV+MluI digest of paternally targeted (+/S12) cells generates a 50 kb wild-type maternal band (MluI does not cut) and a 21.9 kb targeted paternal band (MluI cuts); the 37 kb band results from feeder contamination. Note that the MluI site outside the ICE is methylated on both parental alleles and does not participate in the assay.
Fig. 3
Fig. 3. Imprinted expression of Igf2r arises from a parental-specific expression bias
(A)Allele-specific QPCR distinguishes the mouse Igf2r wild-type (WT) and exon 12 SNP (S12) alleles. In the S12 allele, a C included in a PstI restriction site in Igf2r exon 12 is replaced by a T. Primer MUTSEF (ending on T, S12-assay) or WTSEF (ending on C, WT-assay) in combination with a common reverse primer (GESER2) in exon 11 distinguish the alleles. QPCR with the common reverse primer and MUTSEF or WTSEF on equal amounts of plasmid containing Igf2r cDNA with (S12) or without (WT) the SNP shows that the compatible QPCR assay reaches the detection threshold at least seven Ct cycles earlier than the incompatible assay (compare black bar with grey bar for each plasmid). This indicates a high specificity of the assay (~50:1). (B)Allele-specific QPCR showing the ratio of maternal to paternal Igf2r expression during RA-induced differentiation of maternally targeted (S12/+, dark grey bars) and paternally targeted (+/S12, light grey bars) ES cells. Mean values and s.d. of three replicates are shown. Since Igf2r is biallelically expressed in undifferentiated ES cells, the middle value of the three replicates on day (d) 1 is set to 1 (*). The maternal to paternal Igf2r expression ratio increases during differentiation in both S12/+ and +/S12 ES cells, showing that expression of the maternal allele is greater than that of the paternal allele. (C)Igf2r expression in feeder-depleted differentiating D3 ES cells with disruption of full-length Igf2r by lacZ-polyA insertion into exon 1 on the maternal (lacZ/+) or paternal (+/lacZ) or on both (lacZ/lacZ) alleles, analysed by northern blot using a downstream Igf2r probe (HX). Igf2r expression is upregulated from the wild-type maternal allele (+/lacZ) and is expressed at similar levels throughout differentiation from the wild-type paternal allele (lacZ/+). Igf2r expression in lacZ/lacZ ES cells indicates expression from irradiated feeder MEFs that contaminate earlier time points, despite feeder depletion. Oct4, control for ES cell differentiation; 18S rRNA, loading control. (D)Northern blot confirming biallelic Igf2r expression in undifferentiated +/lacZ and lacZ/+ D3 ES cells and the absence of Igf2r expression in lacZ/lacZ D3 ES cells (details as C). Wild-type D3 and CCE undifferentiated ES cells are shown for comparison. All ES cells were grown on mutant MEF feeders that completely lack Igf2r as they have a targeted deletion of the maternal Igf2r promoter and a Thp (hairpin-tail) deletion on the paternal chromosome that includes the whole Igf2r imprinted cluster. (E)QPCR analysis (assay Igf2rex48) of maternal and paternal wild-type Igf2r alleles in +/lacZ (light grey bars) and lacZ/+ (dark grey bars) differentiating ES cells grown as in D on mutant feeders. Bars indicate mean values with s.d. of three replicates. The middle value of the three replicates on day 1 was set to 1 (*). Expression of the wild-type maternal Igf2r allele increases during ES cell differentiation from day 2 of RA treatment onward, whereas expression of the wild-type paternal allele is constant during early time points (days 1-4) and increases slightly by day 5.
Fig. 4
Fig. 4. Gain of de novo methylation of the paternal Igf2r promoter
(A)DNA methylation of mouse Igf2r and Airn promoters during RA or embryoid body (EB) differentiation of CCE and D3 ES cells. Upper and middle panels show gain of methylation on the Igf2r promoter on a NotI site close to the transcription start. The lower panel shows stable methylation of the Airn promoter on a MluI site in the same DNA samples. Thp/+ and +/Thp uniparental deletion cells show that the Igf2r promoter is paternally methylated (Thp/+), whereas the Airn promoter is maternally methylated (+/Thp) in tail DNA. NIH3T3 diploid cells show a 50:50 ratio of methylated:unmethylated signal. Igf2r promoter: EcoRI+NotI digest with probes EEi (upper panel) and BEi (middle panel), both of which are contained in the EcoRI fragment. Airn promoter: EcoRI+MluI digest with probe MSi. M, methylated allele; UM, unmethylated allele; E, EcoRI; *, feeder cell contamination; nd, not done. (B)ImageJ quantification of Igf2r promoter methylation during RA or EB differentiation of CCE and D3 cells (probe EEi). Bars indicate percentage methylation on the paternal Igf2r promoter as compared with the fully methylated paternal Igf2r promoter in NIH3T3 cells. Methylation on the paternal Igf2r promoter in NIH3T3 cells was set to 100 (*). (C)DNA methylation kinetics at the Oct4 promoter (EcoRV + methyl-sensitive Eco47III) and the Igf2r promoter (EcoRI + methyl-sensitive NotI) during RA-induced differentiation of CCE cells assayed in the same DNA sample. Both promoters gain DNA methylation between days 2 and 3 of RA treatment (dashed line), but the Oct4 promoter gains substantially more methylation than the Igf2r CpG island promoter. (D)Details of the Southern blot methylation analyses in A and C (to scale).
Fig. 5
Fig. 5. Histone modification dynamics at Igf2r and Airn promoter regions upon ES cell differentiation
(A)Mouse Igf2r and Airn promoter regions. Arrows, direction of transcription; black boxes above line, CpG islands; black bars below line, regions analysed using 24 primer pairs spanning 12 kb around the Airn and Igf2r transcription start sites and spaced at ~1 kb intervals [from Regha et al., with permission (Regha et al., 2007)]; short grey bar, the 3.65 kb genetically defined ICE; grey and black asterisks, the MluI and NotI sites assayed in Fig. 4. (B)In undifferentiated ES cells, Igf2r is expressed biallelically and Airn is silent, the maternal Airn promoter carries a DNA methylation imprint and the Igf2r promoter is unmethylated. In differentiated ES cells, Igf2r is upregulated from the maternal allele and Airn is upregulated from the paternal allele, the paternal Igf2r promoter gains DNA methylation while the maternal Airn promoter methylation imprint is unchanged. Dotted arrow, low-level expression; thick arrow, high-level expression; black oval, methylated CpG island; white oval, unmethylated CpG island; grey oval, partially methylated CpG island. (C-N)Scanning ChIP-PCR analysis of Igf2r (C-H) and Airn (I-N) promoter regions for two repressive (H3K27me3, H3K9me3) and two active (H3K9Ac, H3K4me2) marks in undifferentiated (left panels) and differentiated (right panels) ES cells. Two to four replicates of each ChIP-PCR were performed and representative images are shown. M, size marker. (C)Input PCR for Igf2r promoter region for H3K27me3 and H3K9Ac ChIPs. (D,E)Scanning ChIP-PCR using H3K27me3 and H3K9Ac antibodies and PCR assays 13-24 (see A) to assay 12.7 kb around the Igf2r promoter. (F)Input PCR for Igf2r promoter region for H3K4me2 and H3K9me3 ChIPs. (G,H)Scanning ChIP-PCR using H3K4me2 and H3K9me3 antibodies. Details as D,E. (I)Input PCR for Airn promoter region for H3K27me3 and H3K9Ac ChIPs. (J,K)Scanning ChIP-PCR using H3K27me3 and H3K9Ac antibodies and primer pairs 1-12 (see A) to assay 11.6 kb around the Airn promoter. (L)Input PCR for Airn promoter region for H3K4me2 and H3K9me3 ChIPs. (M,N)Scanning ChIP-PCR using H3K4me2 and H3K9me3 antibodies. Details as J,K. (O)QPCR assays of H3K27me3 ChIP DNA on Igf2r (left) and Airn (middle) using primer pairs 18-21 for Igf2r and 6-7 for Airn (stars in C and I). Bars indicate the percentage ChIP/input (with s.d.). Arrows, mock antibody negative control showing background signals only. The Slc22a3 CpG island (right) was used as a positive control for the day-5 H3K27me3 ChIP.
Fig. 6
Fig. 6. Differentiating ES cells mimic the developmental gain of Igf2r/Airn imprinted expression
Changes in Igf2r/Airn expression (black lollipop, no expression; dotted arrow, low-level expression; solid arrow, medium- or high-level expression), DNA methylation (Me), repressive histone modifications (black; K9, H3K9me3; K27, H3K27me3), active histone modifications (white; K4, H3K4me2/me3; Ac, H3K9Ac), shown on the left-hand side for the mouse embryo and on the right-hand side for ES cells. In pre-implantation embryos, Igf2r is biallelically expressed and Airn is silent. The histone modification status is unknown, but DNA methylation is present on the maternal Airn CpG island and is absent from the Igf2r CpG island. These patterns are mimicked in undifferentiated ES cells that also show repressive H3K9me3 modifications on the maternal Airn CpG island and bivalent histone modifications comprising H3K27me3 and H3K4me3 on the Igf2r and Airn promoters, which is typical of all CpG islands in ES cells irrespective of expression status. In 11.5-13.5 dpc embryos, Igf2r shows imprinted maternal-specific expression, whereas Airn shows imprinted paternal-specific expression in all tissues except post-mitotic neurons (which lack Airn and express Igf2r biallelically). Persistent expression of the paternal Igf2r promoter is detected in some tissues of the post-implantation embryo. DNA methylation is maintained on the maternal Airn promoter and is also now present on the paternal Igf2r promoter (the latter is not fully methylated until after birth). Active histone marks (H3K4me2/3, H3K9Ac) are only found on the expressed paternal Airn and expressed maternal Igf2r CpG island promoters. Repressive histone marks (H3K9me3, H4K20me3) plus HP1 are only found on the silent maternal Airn and the silent paternal Igf2r promoters. Notably, both the expressed and silent Airn and Igf2r promoters and their gene bodies are free of H3K27me3. All these features, including the persistent low-level expression of the paternal Igf2r promoter and the loss of H3K27me3, are fully mimicked in differentiated ES cells. We propose two models, as discussed in the text, to explain the persistence of low-level Igf2r expression from the paternal allele: (1) maternal-specific upregulation and (2) paternal-specific silencing with stochastic ‘escapers’. References: 1(Szabo and Mann, 1995), 2(Lerchner and Barlow, 1997), 3(Wang et al., 1994), 4(Terranova et al., 2008), 5(Stoger et al., 1993), 6(Braidotti et al., 2004), 7(Mikkelsen et al., 2007), 8(Sleutels et al., 2002), 9(Regha et al., 2007). M, maternal; P, paternal; ?, unknown status; d RA, days of RA-induced differentiation.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Azuara V, Perry P, Sauer S, Spivakov M, Jorgensen HF, John RM, Gouti M, Casanova M, Warnes G, Merkenschlager M, et al. Chromatin signatures of pluripotent cell lines. Nat. Cell Biol. 2006;8:532–538. - PubMed
    1. Barlow DP, Bartolomei MS. Genomic imprinting in mammals. In: Allis C. David, Jenuwein T, Reinberg D., editors. Epigenetics. Cold Spring Harbor Laboratory Press; Cold Spring Harbor, New York: 2007.
    1. Bell AC, Felsenfeld G. Methylation of a CTCF-dependent boundary controls imprinted expression of the Igf2 gene. Nature. 2000;405:482–485. - PubMed
    1. Braidotti G, Baubec T, Pauler F, Seidl C, Smrzka O, Stricker S, Yotova I, Barlow DP. The Air noncoding RNA: an imprinted cis-silencing transcript. Cold Spring Harb. Symp. Quant. Biol. 2004;69:55–66. - PMC - PubMed
    1. da Rocha ST, Tevendale M, Knowles E, Takada S, Watkins M, Ferguson-Smith AC. Restricted co-expression of Dlk1 and the reciprocally imprinted non-coding RNA, Gtl2: implications for cis-acting control. Dev. Biol. 2007;306:810–823. - PubMed

Publication types

-