Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. 2003 Mar 4; 100(5): 2468–2473.
Published online 2003 Feb 14. doi:  10.1073/pnas.0434312100
PMCID: PMC151364
Developmental Biology

Rnf2 (Ring1b) deficiency causes gastrulation arrest and cell cycle inhibition


The highly homologous Rnf2 (Ring1b) and Ring1 (Ring1a) proteins were identified as in vivo interactors of the Polycomb Group (PcG) protein Bmi1. Functional ablation of Rnf2 results in gastrulation arrest, in contrast to relatively mild phenotypes in most other PcG gene null mutants belonging to the same functional group, among which is Ring1. Developmental defects occur in both embryonic and extraembryonic tissues during gastrulation. The early lethal phenotype is reminiscent of that of the PcG-gene knockouts Eed and Ezh2, which belong to a separate functional PcG group and PcG protein complex. This finding indicates that these biochemically distinct PcG complexes are both required during early mouse development. In contrast to the strong skeletal transformation in Ring1 hemizygous mice, hemizygocity for Rnf2 does not affect vertebral identity. However, it does aggravate the cerebellar phenotype in a Bmi1 null-mutant background. Together, these results suggest that Rnf2 or Ring1-containing PcG complexes have minimal functional redundancy in specific tissues, despite overlap in expression patterns. We show that the early developmental arrest in Rnf2-null embryos is partially bypassed by genetic inactivation of the Cdkn2a (Ink4a/ARF) locus. Importantly, this finding implicates Polycomb-mediated repression of the Cdkn2a locus in early murine development.

Polycomb Group (PcG) proteins and their genetic counterparts, the trithorax Group proteins (trxG), maintain Hox gene expression boundaries (14), which are critical for regional patterning along the antero-posterior (AP) axis (57). Based on biochemical characteristics, mammalian PcG proteins are currently grouped into at least two distinct functional groups: the first comprises Eed, Ezh1, and Ezh2 in the mouse (8, 9); the second consists of the highly related protein pairs Cbx2 (M33)/Cbx4 (MPc2), Bmi/Zfp144 (Mel18), and Edr1 and 2 (Rae28/Mph1 and 2), respectively (10, 11). For ease of this discussion we refer to them as groups I and II, respectively. Group I and II homologs are evolutionarily conserved from Drosophila to humans, only group I homologs are found in plants and Caenorhabditis elegans as well, supporting the concept of separate function (12, 13). In addition, complex composition differs throughout development in a temporal and cell-type-specific manner (14, 15). Interaction of Eed with histone deacetylases and the intrinsic histone methyltransferase activity of Ezh proteins suggest mechanisms for repression by group I complexes (1619). Although a mammalian hPRC-H (group II) complex harbors an intrinsic capacity to stabilize a repressive chromatin structure and counteract SWI/SNF chromatin remodeling complexes in vitro, its in vivo mode of action is not well understood (20). Association with histone methyl transferase activity may in part help explain the repressive action of some group II complexes (21).

Rnf2 and Ring1 have been identified as in vivo interactors of the group II PcG protein Bmi1 by us and others (22). These Ring finger proteins interact with a number of known class II PcG proteins, suggesting a central position in complex formation (refs. 2224 and this study). To address the role of Rnf2 in mammalian development, an Rnf2 null-mutant mouse was generated. We find that in contrast to the reported relatively mild phenotypes observed in other group II PcG-gene null mutants, the Rnf2 mutants display early developmental arrest, which also is observed for Eed and Ezh2 knockout (KO) mice. We further show a partial bypass of the early embryonic arrest by genetic inactivation of the Cdkn2a tumor suppressor locus. Together, our results highlight an essential function for both group I and II Polycomb repressors during early mammalian development.

Materials and Methods

Yeast Two-Hybrid Assay.

Two independent yeast two-hybrid screens were carried out as described elsewhere (9). Yeast strains Y190 and MAV103, which carry Gal4-inducible HIS3 and lacZ reporter genes, were transformed with Bmi1-Ring finger encoding sequences (1–780) carrying a C-terminal Gal4 DNA binding domain (DBD). Bait-transformed yeast strains were transformed with a human Jurkat T cell cDNA library or a B cell cDNA library fused to Gal4-transactivation (TA) domain sequences.

Genomic Library Screens and Targeting Construct.

The incomplete human Rnf2 cDNA was used to screen a 129/Ola mouse genomic phage library. Two ≈18-kb overlapping fragments were used to design a targeting construct. The position of exons encoding the Ring finger domain was determined by standard single stranded 32P-labeled oligonucleotide hybridization. A ±3.0-kb EcoRI–XhoI fragment containing sequences encoding the major part of the Ring finger domain was replaced by a neomycin selection cassette. Linearized gel-purified targeting vector DNA was electroporated into 129/Ola embryonic stem (ES) cells, which were selected on geneticin (G418) medium. DNA from individual G418-resistant ES cells colonies was BamHI- or HindIII-digested and analyzed by Southern hybridization with a 5′-prime external probe and internal Rnf2-derived and neomycin-based probes. Two independent, correctly targeted ES cells were injected into FVB blastocysts.

In Situ Hybridization.

The hybridization of embryo sections with 35S-labeled probes was carried out as originally described (25) by using modifications reported previously (26). The Brachyury T, Hoxb1, and Evx-1 probes were transcribed from a 1.9-kb EcoRI genomic and an 800-bp and a 588-bp cDNA fragment, respectively. Whole-mount in situ hybridization with a 1-kb mCer1 probe was carried out as described (27).

Cell Culture and Immunoprecipitation.

PcG cDNAs were cloned into pBABE (28) or LZRS-based (29) retroviral expression vectors. Cbx2 (M33) and Cbx4 (MPc2) were cloned into pMT-hemagglutinin (HA) plasmids; Myc-Rnf2 (Ring1b) and HA-Ring1 (Ring1a) and HA-Eed were cloned into pBABE-puromycin vector; Bmi1-PY, myc-Ezh and HA-Ezh were cloned into a LZRS-IRES-GFP plasmid. Rnf2 (Ring1b)-tandem affinity purification (TAP) fusions (30) were generated in a LZRS-IRES-GFP vector. Cells were harvested and PcG complexes were extracted by using a method described elsewhere (10). Immunoprecipitations were carried out with IgG beads or, as indicated, with antisera against cellular PcG proteins. Protein detection was done by standard Western analysis.

Reverse Transcription, Amplification, and Sequencing.

Individual embryonic day (E)7.5, pairs of E6.5, or 30 E3.5 embryos were transferred to clean tubes and snap-frozen on dry ice. For comparison, total RNA was extracted from ≈5 × 103 ES cells. RNA extraction was done as described (31), adapted to minute amounts (32). Reverse transcription (RT) was performed according to the manufacturer's specifications (SuperScript, GIBCO/BRL). Typically 0.15 (E7.5), 0.3 (E6.5), or 4.5 (E3.5) embryo equivalent was used in a subsequent PCR in PcG-cDNA amplification; RT-PCR of β-Actin served as an mRNA quality control. Semiquantitative RT-PCR was carried out as described (33); amplified products were detected by radiolabeled oligonucleotides. Reverse transcription and PCR conditions and amplimer sequences are available on request. Reverse transcribed, PCR-amplified total RNA from an Rnf2-heterozygous (HE) embryo was used to sequence the mRNA transcribed from the KO allele by using a Big-Dye Terminator method (Applied Biosystems) on an automated sequencing apparatus (ABI Prism 3700 DNA Analyzer, Applied Biosystems).

Histological and Skeletal Analysis.

Whole mouse brains were fixed in 4% paraformaldehyde. Four-micrometer sagittal, paraffin-embedded sections were stained with hematoxylin and eosin (H&E). Immunohistochemistry for glial fibrillary acidic protein (GFAP; polyclonal, 1:500; Dako), NeuN (monoclonal, 1:4,000; Chemicon), and calbindin (polyclonal, 1:200; Chemicon) was done. Biotinylated secondary antibodies (goat anti-rabbit and rabbit anti-mouse; Dako) were used at a dilution of 1:300. Chromogens used for visualization were biotin/avidin-peroxidase (Dako) and diaminobenzidine. Skeletal phenotypes of newborn Rnf2/Bmi1 compound mutant mice were determined as described (34).

Animal Procedures and Genotyping.

Chimeric Rnf2-HE mice were backcrossed to a 129/Ola or FVB background. 129/Ola Rnf2-HE or second-generation FVB-backcrossed Rnf2-HE animals were intercrossed to obtain full Rnf2-KO mice. Rnf2-HE/Bmi1-HE (35) mice were backcrossed to a FVB background for at least four generations, as were Rnf2-HE/Cdkn2a-KO (36) animals. Embryos until E7.5 from timed pregnancies (E0.5 at noon) were used entirely for DNA or RNA isolation. Surrounding membranes were used for genotyping older embryos. Rnf2-HE intercrosses yielded 19 litters comprising in total 272 analyzed embryos. In total six litters, comprising 65 embryos, were analyzed in the Rnf2/Cdkn2a-double KO (dKO) background. The Rnf2 genotype was determined by Southern analysis or PCR. Bmi1 and Cdkn2a-WT and KO alleles were analyzed by PCR as described (33).

Results and Discussion

Null Mutation of Rnf2 (Ring1b) Causes Early Developmental Arrest.

By using yeast two-hybrid assays with the murine Bmi1 Ring finger domain and directly flanking sequences as bait, several positive, overlapping human cDNAs were identified as two previously characterized Ring finger proteins, Ring1 (Ring1a) and Rnf2 (22, 23). To study the role of Rnf2 in mammalian development, a replacement vector was used to interrupt the murine Rnf2 locus in ES cells (Fig. (Fig.11a). Absence of functional Rnf2 mRNA was confirmed by RT-PCR (see below; Figs. Figs.11c and and33d). No null-mutant mice were detected among 101 10-day-old offspring and newborns tested, suggesting embryonic death. Indeed, the Rnf2-KO genotype was not detected beyond E10.5; Southern analysis did reveal the presence of Rnf2-KO embryos at E9.0–E9.5 at the expected 1 in 4 Mendelian ration (Fig. (Fig.11b). These observations held for two independently derived mouse lines in both the FVB and 129/Ola genetic background. E6.5, E7.5, and E8.5 null-mutant embryos all displayed an abnormal morphology, compared with that of WT and HE embryos. All Rnf2-KO embryos were delayed in development. Phenotypic variation was considerable: some KO concepti (≈10%) displayed an “empty” parietal yolk sac (data not shown), most had progressed further, but lagged in development behind the controls, never reaching the headfold stage (Fig. (Fig.22a).

Figure 1
Null mutation of the Rnf2 locus. (a) Rnf2 targeting construct. A ±3.0-kb EcoRI–XhoI 129/Ola genomic fragment containing exons (black boxes) encoding the major part of the Ring finger (RF) domain was replaced by a neomycin selection ...
Figure 2
Morphological abnormalities of Rnf2-KO embryos. (a) A normal day-7.5 embryo (Left) at the late headfold stage; neural folds are visible. Three smaller Rnf2-KO embryos (Right) manifest a characteristic phenotype. (b) Gene expression analysis in pre/early ...
Figure 3
Rnf2 interactions in whole cells and early embryonic PcG expression profiles. (a) Cotransfection experiments demonstrate interaction of Rnf2-TAP-tag with HA-tagged Cbx2 (M33) and Cbx4 (MPc2) in COS-7 cells (Right) and with endogenous Bmi1 and Edr (Mph) ...

Rnf2 Is Essential for Normal Gastrulation.

Rnf2-KO embryos do not progress through gastrulation normally (E6.5–E7; Fig. Fig.22b). The lethal effect of Rnf2-inactivation is surprising, because null mutation of genes encoding other PcG complex members and/or binding partners revealed redundancy between highly related homologs (35, 3741). Interestingly, genetic inactivation of Eed or Ezh2, both group I PcG proteins, results in gastrulation arrest as well (42, 43). Eed mutants arrest early, at least in part because Eed is necessary for stable maintenance of imprinted X inactivation in trophectoderm (44). Proper epiblast expansion also fails in these null-mutant embryos, possibly as a result of excessive ingression through the posterior primitive streak (45, 46); an excess of mesoderm localizes to the extraembryonic compartment, which may be caused by an anterior migration defect. Ezh2 null mutant embryos either do not initiate or fail to complete gastrulation. Growth of the inner cell mass and epiblast cells is impaired and accumulation of mesoderm in the extraembryonic compartment was reported (43). We analyzed E6.5 through E8.5 Rnf2-KO embryos histologically and regarding expression of developmental markers. The data show that some Rnf2-KO embryos exhibit an abnormal distal restriction of the anterior visceral endoderm (AVE) marker Cer1 (47) (Fig. (Fig.22b, rightmost embryo), whereas others show the normal anterior expression (Fig. (Fig.22b). These variations in Cer1 expression possibly reflect the considerable phenotypic variation observed in Rnf2-KO embryos.

In comparison to WT E7.5 embryos, a relatively far progressed Rnf2-KO reveals mesoderm accumulation in the primitive streak region posteriorly, but very little anteriorly expanding embryonic mesoderm (Fig. (Fig.2c2c A and B). Development of both the embryonic and extraembryonic part of the Rnf2-KO embryo is delayed. In particular, the posterior amniotic fold has hardly formed in KO embryos (Fig. (Fig.22cB), whereas in WT and HE embryos, it has developed posteriorly and fused with anterior extraembryonic ectoderm, ensuring amnion and chorion formation (Fig. (Fig.22cA). In most cases, the chorion appeared abnormal (data not shown). Analysis of expression patterns in situ on transverse sections of Rnf2-KO embryos reveals that Hoxb1, one of the earliest Hox genes activated, is not expressed in E7.5 KO embryos (Fig. (Fig.22cE), whereas it is expressed at the proper location in the E8.5 KO, in the accumulated mesoderm and epiblast at the level of the primitive streak (Fig. (Fig.22cG). However, expression of Hoxb1 in these tissues is significantly lower in E8.5 KO than in E7.5 controls (Fig. (Fig.2c2c G and C), in accordance with the severe developmental delay of the null mutants. The expression of the nascent mesoderm marker Brachyury T (Fig. (Fig.2c2c D, F, H, and J) and the posterior marker Evx-1 (data not shown) is correctly restricted, but T expression is weaker in the E7.5 KO than in controls (Fig. (Fig.2c2c D and F). The developmental stage of the E8.5 KO shown in Fig. Fig.2c2c G and H is closer to that of an E7.5 (Fig. (Fig.2c2c C and D) than to an E8.5 WT embryo (the latter not shown in these sections). Taken together, the above data illustrate a considerable delay in embryonic and extraembryonic development in Rnf2-KO embryos (cf. Fig. Fig.22a).

PcG Gene Expression Profiles Are Not Affected by Rnf2 Null Mutation.

To substantiate that the null phenotype is not caused by altered expression of Eed or Ezh2, we studied the expression profiles of different PcG genes in early WT and KO embryos. Rnf2 expression is detectable in blastocysts and ES cells, whereas Eed expression first appears shortly after implantation (Fig. (Fig.33d). This finding suggests that functional inactivation of Eed is unlikely to affect Rnf2 expression at preimplantation stages. Vice versa, Rnf2-KO embryos still express the Eed (Fig. (Fig.33d). The expression profiles of Ezh1 and Ezh2 are unchanged in the absence of Rnf2, as is the case for Ring1 and Bmi1 (Fig. (Fig.33d). Also, under the indicated experimental conditions (i.e., approximately equal material input; see Materials and Methods), Ring1 expression, in contrast to Rnf2, was not detectable in ES cells or blastocysts (Fig. (Fig.33d). Although more concentrated ES cell lysates reveal a low level of Ring1 expression (E. Boutsma and M.v.L., personal communication), the low levels of Ring1 detected clearly cannot compensate for the loss of Rnf2, both in ES cells and at later developmental stages. This observation further suggests a remarkable degree of functional divergence and specialization among highly related PcG proteins. The finding that compound null mutants for both Zfp144 (Mel18) and Bmi1 arrest later (i.e., at the 18th somite stage) than Rnf2 null mutants (41) further corroborates the notion of lineage-specific and temporal divergence among PcG complexes in murine development.

Because of temporally distinct expression profiles in early development, Ezh2-mediated silencing is thought to occur even before the onset of Eed expression (43). Likewise, Rnf2 is expressed relatively early and at a higher level than some of its binding partners, such as Bmi1. The observation that robust embryonic expression of Rnf2 appears to precede Eed also may suggest that early PcG-mediated repression functions in part by means of a mechanism independent of Eed-mediated histone deacetylase recruitment.

Differences in PcG Complex Composition in Early Mammalian Development.

We next studied PcG protein interactions in whole cells. Rnf2 is readily detectable in complex with mammalian group II PcG proteins (e.g., Bmi1, Cbx2, Cbx4, and Edr; Fig. Fig.33a). The phenotypic resemblance of group I and Rnf2 null mutants prompted us also to study Rnf2–group I PcG interactions. We found that Rnf2 does not interact with Eed or Ezh2 in several differentiated, established cell lines, as examined by coexpression and immunoprecipitation (Fig. (Fig.33 b and c). Although these findings are in full agreement with recent reports by other laboratories (8, 9, 2224), we cannot formally exclude that experimental conditions used so far do not permit detection of a more transient interaction between group I and group II complexes (see discussion below). Alternatively, group I/II interactions may be cell-type-specific or mediated by as-yet-unknown cellular factors in early murine development.

Compound Group-II PcG Mutations Reveal Genetic Interaction but also Suggest Divergences in Target Genes.

Although a hypomorphic Rnf2 mutant causes subtle homeotic transformations (48), the axial skeleton of Rnf2-HE newborn mice in this study reveals no obvious vertebral transformations or malformations (see Table 1, which is published as supporting information on the PNAS web site, www.pnas.org). In contrast, heterozygocity for Ring1 results in a very strong homeotic phenotype (40). When one Rnf2 allele is inactivated in addition to both Bmi1 alleles, only subtle additional changes to the axial skeleton are observed (Table 1). Taken together, the data on axial skeleton development further suggest that Rnf2 cannot compensate for Ring1-dosage variation during somite and vertebral development. Gross morphological appearance and body weight measurements indicate that the Rnf2-HE/Bmi1-KO mice are delayed in postnatal development at 10 days postpartum, from which point onward growth ceases almost entirely (Fig. (Fig.44a). In addition, the neurological disorder of Bmi1-KO mice (35) is aggravated in Rnf2-HE/Bmi1-KO mice: animals display a more seriously hampered gait and severe lack of coordination. The molecular and granular layer of the Rnf2-HE/Bmi1-KO cerebellum are severely depleted at the cellular level, significantly more so than in the Bmi1-KO cerebellum (Fig. (Fig.4b4b CE), which may explain the further deterioration of the motoric performance of Rnf2-HE/Bmi1-KO mice. Poor postnatal development combined with the motor neuron defects results in death at an earlier age than occurs in Bmi1-KO animals. The substantially reduced cellularity of the Rnf2-HE/Bmi1-KO cerebellum (Fig. (Fig.4b4b CE) further indicates that Ring1 cannot compensate for loss of Rnf2 on brain development, despite overlap in expression pattern (40). Despite their high homology, mammalian PcG protein homologs are known to have different binding properties to other (PcG) proteins (14). The phenotypic comparison between Ring1 and Rnf2 mutant mice further supports the notion of limited functional overlap and target gene specificity between Ring1 or Rnf2-containing PcG complexes in different tissues. The above and related issues will be addressed in an Rnf2 conditional-KO setting.

Figure 4
Genetic interaction of Rnf2 and Bmi1 mutations. (a) Poor postnatal growth of Rnf2-HE/Bmi1-KO mice. Gross appearance of Rnf2-HE/Bmi1-KO (Upper) and Bmi1-KO (Lower) mice at 4 weeks of age as compared with WT and double HE littermates. ( ...

Genetic Inactivation of Cdkn2a (Ink4a/ARF) Partially Bypasses the Early Developmental Arrest.

The epiblast in most Rnf2-KO embryos shows limited expansion, which could point to a proliferative defect. TrxG as well as counteracting PcG gene products are implicated in cell cycle regulation and malignant transformation (49, 50). Furthermore, Rnf2 and Bmi1 are found in complex with E2F transcription factors, which play key roles in regulation of cellular proliferation and differentiation (5153). Removal or overexpression of Bmi1 is associated with premature replicative senescence or cellular immortalization and tumorigenesis, respectively (reviewed in ref. 50), an effect that is mediated in part by means of regulation of the Cdkn2a locus (33, 54). The Ring finger domain of Bmi1, through which Bmi1 interacts with Rnf2, is essential for Bmi1's oncogenic potential and its ability to mediate cell cycle control by means of the Cdkn2a locus (33, 54). This finding prompted us to investigate whether inactivation of Rnf2 results in derepression of the Cdkn2a locus during early development. Indeed, we found expression of p16INK4a up-regulated in Rnf2-KO embryos at E7.0 (Fig. (Fig.55a). To unambiguously demonstrate that Rnf2 and Cdkn2a act in the same genetic axis, we crossed the Rnf2 null allele onto the Cdkn2a-KO background (36). Genetic removal of the INK4a/ARF cell cycle inhibitors results in a partial rescue of the Rnf2 null phenotype, and embryonic growth and gastrulation proceed normally until early somite stages (Fig. (Fig.55 b and c). Notably, a number of somites is laid down in paraxial mesoderm of dKO embryos, which is never observed in Rnf2-KO embryos (Fig. (Fig.5c).5c). Nevertheless dKO embryos still display abnormal morphology and their development does not progress beyond E11–E12. This observation suggests that an INK4a/ARF-activated proliferative block, caused by loss of PcG class II function, clearly contributes to the early developmental arrest in the Rnf2-KO.

Figure 5
Partial bypass of the Rnf2 null mutant developmental arrest by genetic inactivation of the Cdkn2a locus. (a) Up-regulation of P16INK4a expression in E7.5 embryos as detected by semiquantitative RT-PCR. (b Upper) Compared with Rnf2-KO embryos, E8 Rnf2 ...

In summary, we report here that Rnf2 inactivation is incompatible with early mouse development. In Rnf2 null mutants, the epiblast fails to expand normally, and mesoderm is not properly laid down or does not migrate anteriorly. Moreover, a clear delay in development of extraembryonic structures accompanies the delay in development of the embryo proper. This condition is reminiscent of the group I PcG Eed- and Ezh2-mutant phenotype. Although not completely identical to the Eed and Ezh2 null phenotype, the findings indicate that both groups of PcG protein complexes are required from early development onward to maintain correct target gene repression. The gastrulation arrest in the respective null mutant embryos (i.e., Eed, Ezh2, and Rnf2) suggests a mechanistic link between group I and II PcG function in early mammalian development. In mammals and in the fruit fly, Ezh2 and E(z), respectively, were recently uncovered as histone methyl transferases. They are thought to establish silencing through local chromatin structure modification (17): histone methylation by E(z) at lysine-27 in Drosophila may subsequently recruit PC, a PRC1 interaction partner of Dring, which recognizes and binds K27Me via its chromodomain (1719). It is therefore plausible that, early in development, Rnf2-containing protein complexes recognize a preexistent epigenetic repressional state generated by Ezh2 complexes. Alternatively, or possibly additionally, Rnf2 may be involved in mediating a transient group I/II interaction. Recent experiments suggest such a scenario exists in Drosophila (55). Elegant experiments in the fruit fly revealed that Esc is required at a transition stage, when early gene silencing is converted into stable repression. This is then maintained by (group II) PRC-1 complexes (5658). Subsequently it was shown that establishment of heritable Polycomb silencing requires a transient interaction between class I and II proteins during the first 3 h of embryonic development (59). It is conceivable that a similar interaction takes place during early mammalian development. Finally, in analogy with the central position of Dring in the Drosophila PRC-1 core-complex formation (60) Rnf2 may function as a crucial assembly factor in mammalian PRC complexes (20). Failure to assemble PcG complexes will result in derepression of critical target genes and subsequent developmental arrest. We show here that Rnf2-mediated repression of the Cdkn2a (Ink4a/ARF) locus is crucial for early development to proceed normally. This observation may be part of a mechanism ensuring heritable transcriptional maintenance during the rapid cell divisions that occur before and during early gastrulation. Besides a role in differentiation and determination, these data establish an important function for PcG proteins in cell cycle regulation in early mammalian development.

Supplementary Material

Supporting Table:


We are indebted to M. Dyer and H. Koseki for Ring1b antiserum; A. Ullrich for the Ezh2 antiserum; M. Djabali for Cbx2 cDNA; T. Jenuwein for Ezh2 cDNAs; M. Vidal for Ring1 and Rnf2 expression vectors; A. Schumacher for an Evx-1 probe and Eed cDNA; B. Herrmann, R. Krumlauf, and E. De Robertis for Brachyury T, Hoxb1, and mCer1 probes, respectively; and M. Serrano and R. DePinho for Cdkn2a (Ink4a/ARF) null mutant mice. We thank J. Korving, E. Boutsma, K. van Veen, E. Wientjes, and K. Kieboom for excellent technical assistance; A. Beverdam, F. Meijlink, R. Carvalho, and S. Chuva de Sousa Lopes for help with in situ hybridizations; L. Oomen for help with graphical work; and K. Lawson, A. Lund, R. Zwart, and R. Kingston for critical discussion. J.W.V. was supported by a grant from the Dutch Organization for Scientific Research.


PcGPolycomb Group
ES cellembryonic stem cell
Enembryonic day n
dKOdouble KO
TAPtandem affinity purification


This paper was submitted directly (Track II) to the PNAS office.


1. Paro R. Trends Genet. 1990;6:416–421. [PubMed]
2. Kennison J A. Annu Rev Genet. 1995;29:289–303. [PubMed]
3. Simon J. Curr Opin Cell Biol. 1995;7:376–385. [PubMed]
4. Pirrotta V. Cell. 1998;93:333–336. [PubMed]
5. Lewis E B. Nature. 1978;276:565–570. [PubMed]
6. Krumlauf R. Cell. 1994;78:191–201. [PubMed]
7. Deschamps J, van den Akker E, Forlani S, De Graaff W, Oosterveen T, Roelen B, Roelfsema J. Int J Dev Biol. 1999;43:635–650. [PubMed]
8. Sewalt R G, van der Vlag J, Gunster M J, Hamer K M, den Blaauwen J L, Satijn D P, Hendrix T, van Driel R, Otte A P. Mol Cell Biol. 1998;18:3586–3595. [PMC free article] [PubMed]
9. van Lohuizen M, Tijms M, Voncken J W, Schumacher A, Magnuson T, Wientjens E. Mol Cell Biol. 1998;18:3572–3579. [PMC free article] [PubMed]
10. Alkema M J, Bronk M, Verhoeven E, Otte A, van't Veer L J, Berns A, van Lohuizen M. Genes Dev. 1997;11:226–240. [PubMed]
11. Gunster M J, Satijn D P, Hamer K M, den Blaauwen J L, de Bruijn D, Alkema M J, van Lohuizen M, van Driel R, Otte A P. Mol Cell Biol. 1997;17:2326–2335. [PMC free article] [PubMed]
12. Korf I, Fan Y, Strome S. Development (Cambridge, UK) 1998;125:2469–2478. [PubMed]
13. Grossniklaus U, Vielle-Calzada J P, Hoeppner M A, Gagliano W B. Science. 1998;280:446–450. [PubMed]
14. Satijn D P, Otte A P. Biochim Biophys Acta. 1999;1447:1–16. [PubMed]
15. Brock H W, van Lohuizen M. Curr Opin Genet Dev. 2001;11:175–181. [PubMed]
16. van der Vlag J, Otte A P. Nat Genet. 1999;23:474–478. [PubMed]
17. Cao R, Wang L, Wang H, Xia L, Erdjument-Bromage H, Tempst P, Jones R S, Zhang Y. Science. 2002;298:1039–1043. [PubMed]
18. Czermin B, Melfi R, McCabe D, Seitz V, Imhof A, Pirrotta V. Cell. 2002;111:185–196. [PubMed]
19. Muller J, Hart C M, Francis N J, Vargas M L, Sengupta A, Wild B, Miller E L, O'Connor M B, Kingston R E, Simon J A. Cell. 2002;111:197–208. [PubMed]
20. Levine S S, Weiss A, Erdjument-Bromage H, Shao Z, Tempst P, Kingston R E. Mol Cell Biol. 2002;22:6070–6078. [PMC free article] [PubMed]
21. Sewalt R G, Lachner M, Vargas M, Hamer K M, den Blaauwen J L, Hendrix T, Melcher M, Schweizer D, Jenuwein T, Otte A P. Mol Cell Biol. 2002;22:5539–5553. [PMC free article] [PubMed]
22. Hemenway C S, Halligan B W, Levy L S. Oncogene. 1998;16:2541–2547. [PubMed]
23. Schoorlemmer J, Marcos-Gutierrez C, Were F, Martinez R, Garcia E, Satijn D P, Otte A P, Vidal M. EMBO J. 1997;16:5930–5942. [PMC free article] [PubMed]
24. Satijn D P, Gunster M J, van der Vlag J, Hamer K M, Schul W, Alkema M J, Saurin A J, Freemont P S, van Driel R, Otte A P. Mol Cell Biol. 1997;17:4105–4113. [PMC free article] [PubMed]
25. Wilkinson D G, Bailes J A, Champion J E, McMahon A P. Development (Cambridge, UK) 1987;99:493–500. [PubMed]
26. Vogels R, de Graaff W, Deschamps J. Development (Cambridge, UK) 1990;110:1159–1168. [PubMed]
27. Wilkinson D G, editor. In Situ Hybridization: A Practical Approach. Oxford: IRL; 1992.
28. Morgenstern J P, Land H. Nucleic Acids Res. 1990;18:3587–3596. [PMC free article] [PubMed]
29. Kinsella T M, Nolan G P. Hum Gene Ther. 1996;7:1405–1413. [PubMed]
30. Rigaut G, Shevchenko A, Rutz B, Wilm M, Mann M, Seraphin B. Nat Biotechnol. 1999;17:1030–1032. [PubMed]
31. Chomczynski P, Sacchi N. Anal Biochem. 1987;162:156–159. [PubMed]
32. Voncken J W, Griffiths S, Greaves M F, Pattengale P K, Heisterkamp N, Groffen J. Cancer Res. 1992;52:4534–4539. [PubMed]
33. Jacobs J J, Kieboom K, Marino S, DePinho R A, van Lohuizen M. Nature. 1999;397:164–168. [PubMed]
34. Bel S, Core N, Djabali M, Kieboom K, van der Lugt N, Alkema M J, van Lohuizen M. Development (Cambridge, UK) 1998;125:3543–3551. [PubMed]
35. van der Lugt N M, Domen J, Linders K, van Roon M, Robanus-Maandag E, te Riele H, van der Valk M, Deschamps J, Sofroniew M, van Lohuizen M, et al. Genes Dev. 1994;8:757–769. [PubMed]
36. Serrano M, Lee H, Chin L, Cordon-Cardo C, Beach D, DePinho R A. Cell. 1996;85:27–37. [PubMed]
37. Akasaka T, Kanno M, Balling R, Mieza M A, Taniguchi M, Koseki H. Development (Cambridge, UK) 1996;122:1513–1522. [PubMed]
38. Core N, Bel S, Gaunt S J, Aurrand-Lions M, Pearce J, Fisher A, Djabali M. Development (Cambridge, UK) 1997;124:721–729. [PubMed]
39. Takihara Y, Tomotsune D, Shirai M, Katoh-Fukui Y, Nishii K, Motaleb M A, Nomura M, Tsuchiya R, Fujita Y, Shibata Y, et al. Development (Cambridge, UK) 1997;124:3673–3682. [PubMed]
40. del Mar Lorente M, Marcos-Gutierrez C, Perez C, Schoorlemmer J, Ramirez A, Magin T, Vidal M. Development (Cambridge, UK) 2000;127:5093–5100. [PubMed]
41. Akasaka T, van Lohuizen M, van der Lugt N, Mizutani-Koseki Y, Kanno M, Taniguchi M, Vidal M, Alkema M, Berns A, Koseki H. Development (Cambridge, UK) 2001;128:1587–1597. [PubMed]
42. Schumacher A, Faust C, Magnuson T. Nature. 1996;383:250–253. [PubMed]
43. O'Carroll D, Erhardt S, Pagani M, Barton S C, Surani M A, Jenuwein T. Mol Cell Biol. 2001;21:4330–4336. [PMC free article] [PubMed]
44. Wang J, Mager J, Chen Y, Schneider E, Cross J C, Nagy A, Magnuson T. Nat Genet. 2001;28:371–375. [PubMed]
45. Faust C, Lawson K A, Schork N J, Thiel B, Magnuson T. Development (Cambridge, UK) 1998;125:4495–4506. [PubMed]
46. Faust C, Schumacher A, Holdener B, Magnuson T. Development (Cambridge, UK) 1995;121:273–285. [PubMed]
47. Belo J A, Bouwmeester T, Leyns L, Kertesz N, Gallo M, Follettie M, De Robertis E M. Mech Dev. 1997;68:45–57. [PubMed]
48. Suzuki M, Mizutani-Koseki Y, Fujimura Y, Miyagishima H, Kaneko T, Takada Y, Akasaka T, Tanzawa H, Takihara Y, Nakano M, et al. Development (Cambridge, UK) 2002;129:4171–4183. [PubMed]
49. Caldas C, Aparicio S. Cancer Metastasis Rev. 1999;18:313–329. [PubMed]
50. Jacobs J J, van Lohuizen M. Biochim Biophys Acta. 2002;1602:151–161. [PubMed]
51. Trimarchi J M, Fairchild B, Wen J, Lees J A. Proc Natl Acad Sci USA. 2001;98:1519–1524. [PMC free article] [PubMed]
52. Ogawa H, Ishiguro K, Gaubatz S, Livingston D M, Nakatani Y. Science. 2002;296:1132–1136. [PubMed]
53. Dahiya A, Wong S, Gonzalo S, Gavin M, Dean D C. Mol Cell. 2001;8:557–569. [PubMed]
54. Alkema M J, Jacobs H, van Lohuizen M, Berns A. Oncogene. 1997;15:899–910. [PubMed]
55. Beuchle D, Struhl G, Muller J. Development (Cambridge, UK) 2001;128:993–1004. [PubMed]
56. Struhl G. Nature. 1981;293:36–41. [PubMed]
57. Struhl G, Brower D. Cell. 1982;31:285–292. [PubMed]
58. Duncan I M. Genetics. 1982;102:49–70. [PMC free article] [PubMed]
59. Poux S, Melfi R, Pirrotta V. Genes Dev. 2001;15:2509–2514. [PMC free article] [PubMed]
60. Francis N J, Saurin A J, Shao Z, Kingston R E. Mol Cell. 2001;8:545–556. [PubMed]
61. Frohman M A, Boyle M, Martin G R. Development (Cambridge, UK) 1990;110:589–607. [PubMed]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
PubReader format: click here to try


Save items

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • Cited in Books
    Cited in Books
    NCBI Bookshelf books that cite the current articles.
  • Gene
    Gene records that cite the current articles. Citations in Gene are added manually by NCBI or imported from outside public resources.
  • GEO Profiles
    GEO Profiles
    Gene Expression Omnibus (GEO) Profiles of molecular abundance data. The current articles are references on the Gene record associated with the GEO profile.
  • HomoloGene
    HomoloGene clusters of homologous genes and sequences that cite the current articles. These are references on the Gene and sequence records in the HomoloGene entry.
  • MedGen
    Related information in MedGen
  • OMIM
    Genome Survey Sequence (GSS) nucleotide records reported in the current articles.
  • Pathways + GO
    Pathways + GO
    Pathways and biological systems (BioSystems) that cite the current articles. Citations are from the BioSystems source databases (KEGG and BioCyc).
  • PubMed
    PubMed citations for these articles

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...