- Email: [email protected]
Nanog is required for primitive endoderm formation through a non-cell autonomous mechanism
Developmental Biology 344 (2010) 129–137
Contents lists available at ScienceDirect
Developmental Biology j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / d e v e l o p m e n t a l b i o l o g y
Nanog is required for primitive endoderm formation through a non-cell autonomous mechanism Daniel M. Messerschmidt a,b,⁎, Rolf Kemler a a b
Department of Molecular Embryology, Max Planck Institute for Immunobiology, Stuebeweg 51, 79108 Freiburg, Germany Department of Mammalian Development, Institute of Medical Biology (IMB), 8A Biomedical Grove, #06-06 Immunos Singapore 138648, Singapore
a r t i c l e
i n f o
Article history: Received for publication 1 December 2009 Revised 9 April 2010 Accepted 22 April 2010 Available online 12 May 2010 Keywords: Nanog Epiblast Primitive endoderm Trophectoderm Blastocyst
a b s t r a c t Early lineage segregation in mouse development results in two, either CDX2- or OCT4/NANOG-positive, cell populations. CDX2-positive cells form the trophectoderm (TE), OCT4/NANOG-positive cells the inner cell mass (ICM). In a second lineage decision ICM cells segregate into Epiblast (EPI) and primitive endoderm (PE). EPI and PE formation depend on the activity of the transcription factors Nanog and Gata4/6. A role for Nanog, a crucial pluripotency factor, in preventing PE differentiation has been proposed, as outgrowths of mutant ICMs result in PE, but not EPI derivatives. We established Nanog-mutant mouse lines and analyzed EPI and PE formation in vivo. Surprisingly, Gata4 expression in mutant ICM cells is absent or strongly decreased, thus loss of Nanog does not result in precocious endoderm differentiation. However, Nanogdeficient embryos retain the capacity to form PE in chimeric embryos and, in contrast to recent reports, in blastocyst outgrowths. Based on our findings we propose a non-cell autonomous requirement of Nanog for proper PE formation in addition to its essential role in EPI determination. © 2010 Elsevier Inc. All rights reserved.
Introduction The first lineage segregations in the pre-implantation embryo establish the basis for mouse embryonic development. At the compacting morula stage, trophectoderm (TE) and inner cell mass (ICM) emerge; later the ICM gives rise to primitive endoderm (PE) and epiblast (EPI) (Gardner and Beddington, 1988). Recent evidence suggests that unlike TE/ICM segregation in the morula (Ralston and Rossant, 2008), lineage-specific gene expression precedes EPI- and PE-cell sorting in the ICM. Before the PE becomes evident, the ICM is already composed of two, committed cell types arranged in a ‘salt-and-pepper-like’ distribution (Chazaud et al., 2006). Transcription factors determine the adopted fate: PE cells express Gata6 and Gata4 (Koutsourakis et al., 1999; Morrisey et al., 1998), and EPI cells the homeobox transcription factor Nanog (Chambers et al., 2003; Mitsui et al., 2003). Nanog expression commences at the morula stage and becomes gradually restricted to the ICM during blastocyst formation and expansion (Hart et al., 2004; Mitsui et al., 2003; Zernicka-Goetz et al., 2009). Disruption of Nanog results in peri-implantation lethality and although mutant blastocysts are morphologically normal and induce decidualization they don't form a proper EPI. Embryonic stem (ES) cells cannot be isolated from Nanogdeficient blastocysts; the cells recovered from blastocyst outgrowths display a parietal endoderm-like morphology (Mitsui et al., 2003). Gata4/6-deficient embryos show deficits in establishing PE and fail to
⁎ Corresponding author. Fax: + 65 6464 2048. E-mail address: [email protected] (D.M. Messerschmidt). 0012-1606/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.ydbio.2010.04.020
maintain PE derivatives (Koutsourakis et al., 1999; Morrisey et al., 1998; Soudais et al., 1995), while over-expression of Gata4/6 in ES cells is sufficient to induce PE differentiation (Fujikura et al., 2002). A direct link between Nanog and Gata6 was proposed and NANOG was shown to bind the Gata6 promoter in ES cells (Mitsui et al., 2003; Singh et al., 2007). While early knock-out studies reported endodermal differentiation of Nanog-mutant ICMs, a recent study addressing the role of Nanog in the generation of pluripotency suggested that Nanog-mutant ICM cells are blocked in a transitional stage and are not capable of acquiring the pluripotent ground state and lose their differentiation potential (Silva et al., 2009). These authors propose that Nanog-null ICM cells are thus only capable of TE differentiation or else undergo apoptosis, as opposed to the original description of PE outgrowth (Mitsui et al., 2003). We established and examined Nanog gene-trap mouse lines and found that surprisingly PE formation in mutant blastocysts is profoundly repressed rather than precociously activated. However, in this study we took the analysis from Nanog mutants further, showing that surviving ICM cells from Nanog-null embryos retain the capacity for PE formation in vivo and at least transiently in vitro. Thus, our findings help resolve the recently emerged controversy of Nanog action in the early mouse embryo (Mitsui et al., 2003; Silva et al., 2009), and moreover, we demonstrate that PE formation depends on a functional EPI, hence directly or indirectly, on Nanog expression. In the model we propose, Nanog expression is required for the formation of the EPI lineage, which in turn through a non-cell autonomous mechanism, possibly mediated by FGF/Erk signaling (Chazaud et al., 2006; Nichols et al., 2009), is vital for proper PE formation.
130
D.M. Messerschmidt, R. Kemler / Developmental Biology 344 (2010) 129–137
Materials and methods
Results and discussion
Generation of gene-trap mice and genotyping
Nanog gene-trap embryos
Two ES cell lines (D042B07 [GT1] and D047A12 [GT2]) with disrupted Nanog alleles were obtained from the GSF (Munich) (Schnutgen et al., 2005) (Supplementary Fig. S1) and used for blastocyst injection. Chimeras were mated with C57BL/6 females to generate the Nanoggt1/+ and Nanoggt2/+ mouse lines. Genotyping was performed by PCR on genomic DNA isolated from embryos, yolk sacs or mouse-tail biopsies using primers specific for the GT1- or GT2-allele. For blastocyst complementation the Gt(Rosa)26Sor/J (ROSA26) line (Zambrowicz et al., 1997) was bred onto the GT1-Nanog background and matings performed as detailed in supplementary Fig. S2.
Embryos were flushed (E3.5) or dissected (E4.5) from the uterus following natural matings, counting noon of the day the plug was detected as E0.5. Antibody detection assays were performed as described in (Kan et al., 2007). Self-made affinity-purified rabbit antibody for NANOG was used at a dilution of 1:50, OCT4 (monoclonal mouse, Santa Cruz Biotechnology) at 1:200, CDX2 (monoclonal mouse, BioGenex) at 1:200 and GATA4 (polyclonal rabbit, Santa Cruz Biotechnology) at 1:100. Secondary antibodies used were Alexa Fluor 488 goat anti-rabbit and Alexa Fluor 594 goat anti-mouse (Molecular Probes; 1:500). Nuclei were stained with 10 µM DAPI (Molecular Probes) added to the last wash. For double staining of blastocyst outgrowths with GATA4 an additional αNANOG antibody (R&D, polyclonal goat, 1:10) was used.
Despite the relatively early expression of Nanog in pre-implantation embryos, Nanog mutants form normal morulae, which compact and give rise to morphologically normal blastocysts (Hart et al., 2004; Mitsui et al., 2003). We established two independent gene-trap mouse lines carrying a β-Geo insertion in the first intron of the Nanog locus (Supplementary Fig. S1A, B). We first investigated homozygous offspring of both lines for NANOG expression and found that embryos homozygous for either allele (Nanoggt1/gt1 or Nanoggt2/gt2) failed to express NANOG at pre-implantation stages (supplementary Fig. S1C). We failed to derive live homozygous offspring from either gene-trap line, and did not succeed in isolating homozygous mutants at E7.5, whereas phenotypically undistinguishable heterozygous and wild-type embryos were present. The number of empty deciduomas found, matched the expected Mendelian ratio of 25%. Thus, we concluded that, although genetically different in terms of the gene-trap integration site within the first intron, both, Nanoggt1/gt1 and Nanoggt2/gt2 embryos phenocopied the previously described Nanog-null mutant (Mitsui et al., 2003) and have identical phenotypes. Further, blastocysts carrying both gene-trap alleles (Nanoggt1/gt2) were indistinguishable from homozygous embryos for either allele (Nanoggt1/gt1 or Nanoggt2/gt2) (Fig. 1 A, G and supplementary Fig. S1C). All breeding for experiments shown was carried out using intercrosses of both strains allowing accurate identification of null-mutants by PCR. Thus, embryos carrying both disrupted alleles are referred to as ‘mutants’, and heterozygous and wild-type embryos are grouped as ‘controls’.
ES cell culture, blastocyst outgrowths and embryo culture
TE/ICM lineage segregation is independent of Nanog
ES cells were cultured on feeder cells in ES cell medium (DMEM with 15% FCS, supplemented with 500 U/ml LIF [Chemicon, ESGRO #ES1107]). For blastocyst outgrowths embryos were recovered at E3.5 and cultured in ES cell medium on gelatinized tissue culture chambers (Lab-tek #177402). After 5 days, attached outgrowths were fixed in 4% PFA/PBS, permeabilized in 0.5% Triton X-100/PBS and blocked in 10% FCS in 0.1% Triton X-100/PBS (PBSTx). First and second antibody incubation was performed in blocking solution as described above. Embryos were cultured in KSOM microdrops covered with mineral oil at 37 °C and 5% CO2 atmosphere. Inhibitor treatment on wild-type embryos was performed in KSOM microdrops containing 20 µM SU5402 in DMSO or DSMO only for controls. FGF treatment was performed on embryos from heterozygous intercrosses isolated at E2.5 and cultured for 48 in the presence of human recombinant FGF4 (10− 8 M; Sigma) in KSOM microdrops. Blastocysts were PCR genotyped after GATA4 immunostaining.
We analyzed TE and ICM formation by examining expression of the TE-marker CDX2 and the ICM-marker OCT4 in mutant and control E3.5 blastocysts. CDX2 was restricted to the nuclei of the TE cells in both mutants (Figs. 1A–C) and controls (Figs. 1D–F). Though NANOG was absent in the mutants, we did not find CDX2 in ICM cells as has been described recently (Chen et al., 2009). OCT4 was detectable in all cells of the early blastocyst stage (Figs. 1H, K), thus CDX2 expression does not inevitably drive Oct4 repression in vivo. Only later at implantation does OCT4 become restricted to the ICM (Dietrich and Hiiragi, 2007). Oct4 expression, however, was unaffected by the loss of Nanog (Figs. 1G–I). To ascertain whether loss of Nanog affects cell proliferation, we counted TE and ICM cells in mutant and control blastocysts on stacks of confocal zseries images of CDX2/DAPI stained blastocysts (Fig. 1M). CDX2-positive cells were subtracted from the total cell number to determine the number of ICM cells. Mutant embryos were compared to their control littermates to assure stage/age-matched blastocysts. In one litter ICM cell numbers were 20.3 ± 2.1 (standard deviation, sd) in mutant (n =3) and 22.3 ± 3.9 (sd) in control embryos (n = 9). TE cell numbers were 32.3 ± 2.3 (sd) in mutants and 30.1 ± 2.4 (sd) in controls. Therefore, we did not find significant differences in the number of ICM (Student's t-test, p = 0.3) or TE (p = 0.23) cells in mutants and controls. In a second litter ICM cell numbers were 24 (mutant, n =1) and 24.5 ± 1.9 (sd) (controls, n = 8) and TE cell numbers were 32 (mutant) and 33 ± 5.5 (sd) (control). We concluded that, whilst Nanog is expressed at the morula stage, it is neither required for segregation of TE and ICM, nor for cell proliferation until the E3.5 blastocyst stage.
Whole-mount immunodetection
Generation and analysis of chimeric embryos Wild-type W4 or stable pCAGGS-eGFP transfected W4 ES cells were injected into E3.5 blastocysts obtained from matings according to supplementary Fig. S2. All embryos resulting from these matings carry a ubiquitously expressed β-galactosidase reporter, allowing tracing of embryo-derived cells. After ES cell-injection the blastocysts were retransferred into pseudo-pregnant females and embryos isolated at indicated time points. LacZ staining was performed as described previously (Stemmler et al., 2005). For IF, embryos and yolk sacs were frozen in OCT (Tissue-tek) and sectioned at 14 µm using a cryostat (Leica). Sections were blocked in 10% FCS/PBSTx and incubated with αGFP (polyclonal sheep, 1:400; AbD) and α-β-galactosidase (polyclonal rabbit, 1:1000; MPBiomedicals) antibodies for 1 h at RT. GFP was visualized with biotin-conjugated donkey-anti-sheep Ig (Jackson Labs; 1:300) and FITC-conjugated streptavidin (Jackson Labs; 1:400); β-GAL with Cy3-conjugated donkey-anti-rabbit Ig (Jackson Labs; 1:300).
Nanog mutant ICMs show defects at implantation The same marker-gene analysis was conducted on implanting embryos (E4.5), which were directly dissected from the uterus. All embryos had hatched and attached to the uterine wall at the time of isolation. In control embryos NANOG-positive and NANOG-negative ICM cells were distinct populations (Figs. 2E, H), whereas at E3.5 they
D.M. Messerschmidt, R. Kemler / Developmental Biology 344 (2010) 129–137
131
Fig. 1. IF detection of NANOG (A, D, G, J), CDX2 (B, E) and OCT4 (H, K) in E3.5 Nanog-mutant (A–C, G–I) and control (D–F, J–L) blastocysts. NANOG-positive cells (yellow arrowheads) are intermingled with future PE cells (white arrowheads). Delineated areas (A–F) mark the ICM devoid of CDX2. (C, F, I, L) Merged images. (M) Quantification of ICM cells in mutant and control E3.5 blastocysts. Numbers of embryos are indicated in each bar. ICM cell numbers are not significantly reduced in mutants (Student's t-test, #p = 0.3). (Scale bars: 50 μm).
132
D.M. Messerschmidt, R. Kemler / Developmental Biology 344 (2010) 129–137
were arranged in a ‘salt-and-pepper-like’ manner (Figs. 1D, F, J, L, yellow and white arrowheads). OCT4 was absent in most TE cells in E4.5 embryos and the ICM expression domain was divided into two populations in controls, either overlapping with NANOG (Figs. 2E, H, outlined area) or GATA4 (Fig. 3L). The Nanog-deficient E4.5 embryos were comparable in size to their littermates and had implanted, indicating normal development. In mutants, all ICM cells expressed OCT4, however, the number of ICM cells appeared to be reduced (Figs. 2B, D). We observed frequent pyknotic and fragmented nuclei in mutant embryos, indicating apoptosis (Fig. 2C, red arrowheads). This was confirmed by the high incidence of active Caspase3 positive cells in mutant embryos isolated at E4.5 (Fig. 2I, red arrowhead). In contrast we observed few active Caspase3 positive cells in controls at the examined implantation stage (Fig. 2K). We confirmed our observation by examining additional E4.5 embryos isolated from C57BL/6 wild-type
mice. In two litters, containing 12 and 10 implanted embryos we found only two embryos each containing a single active Caspase3 positive cell localized to the ICM in one, and to the TE in the other embryo. At this stage the PE endoderm layer is formed and the sorting/migration of EPI and PE cells is completed (Figs. 2H and 3L). This result is in agreement with reports documenting decreasing cell death in embryos upon implantation (Copp, 1978), although apoptotic cells can be readily detected in wild-type pre-implantation embryos during the sorting process of EPI and PE cells within the ICM (Meilhac et al., 2009; Plusa et al., 2008). Additionally, our detection of apoptotic cells by staining for active Caspase3 reflects a ‘snap-shot’ of apoptotic events in the embryos, rather than observations made in an extended period of time during live cell imaging (Meilhac et al., 2009; Plusa et al., 2008). The reduction of ICM cell numbers in Nanog mutants was quantified by counting cells in embryos stained for CDX2 (TE) or
Fig. 2. IF detection of NANOG (A, E) and OCT4 (B, F) in Nanog-mutant and control E4.5 embryos. NANOG is undetectable in the mutant (A), confined to the EPI in the control (E, H, delineated area) but absent in the PE. OCT4 expression is absent in the TE, but present in mutant and control ICMs (B, F). (C, G) Red arrowheads indicate DAPI stained, fragmented and pyknotic nuclei detectable in the mutant (C), but not in control ICMs (G). (D, H) Overlay of NANOG, OCT4 and DAPI staining. (I–L) Active Caspase3 and DAPI staining; apoptotic cells are detectable in the mutant (I, J; red arrowheads) but not in the control (K, L). (M) Quantification of ICM cells in mutants and controls. Numbers of embryos are indicated in each bar. Cell numbers are highly significantly reduced in mutant embryos (Student's t-test, *p = 0.0059). (Scale bars: 50 μm).
D.M. Messerschmidt, R. Kemler / Developmental Biology 344 (2010) 129–137
133
Fig. 3. IF localization of GATA4 (A, D, G, J) and OCT4 (B, E, H, K) in Nanog-mutant (A–C, G–I) and control (D–F, J–L) E3.5 (A–F) and E4.5 (G–L) embryos. GATA4-positive cells in the mutant embryo are indicated (G, white arrowheads). (C, F, I, L) Overlay of GATA4, OCT4 and DAPI staining. (M–F) Blastocyst outgrowths of a Nanog-mutant (MP) and control (Q–T) blastocyst stained for GATA4 (M, Q), OCT4 (N, R) and Nuclei (O S). (P, S) Overlay of GATA4, OCT4 and DAPI staining. Lineage derivatives are indicated by white (PE), yellow (EPI) and red arrowheads (TE). (Scale bars: 100 μm).
OCT4 (ICM), and DAPI (total cell number). This revealed a highly significant (Student's t-test; p = 0.0059) loss of ICM cells in the Nanog mutants (26.25 ± 0.81 (sd) n = 4) compared to controls (35.3 ± 1.52
(sd) n = 3) (Fig. 2M). Thus, the first indications of a Nanog mutant phenotype become apparent in the ICM at E4.5 when PE and EPI segregate.
134
D.M. Messerschmidt, R. Kemler / Developmental Biology 344 (2010) 129–137
Nanog mutants don't form primitive endoderm To determine whether Nanog prevents endoderm differentiation in vivo, we analyzed mutant embryos for expression of the PE-marker GATA4. Surprisingly, Nanog-null-mutants did not display precocious PE marker-gene expression as expected (Mitsui et al., 2003), but were entirely devoid of Gata4-expressing cells (Fig. 3A), whereas single GATA4-positive cells could be observed in control ICMs at E3.5 (Fig. 3D). This lack of Gata4 expression raised the possibility of a Nanog requirement for PE formation. In E4.5 control embryos the PE was visible as a discrete structure lining the ICM towards the blastocoel (Fig. 3J). Nine of 41 embryos stained for Gata4 were identified as mutants by PCR. Their ICM was disarranged; a PE layer was not recognizable. Only three of these embryos contained few (four or less) GATA4-positive cells (Fig. 3G, white arrowheads). The other six mutant embryos did not express GATA4. This observation supports our hypothesis of Nanog requirement for proper PE formation, but also shows that this requirement is likely not cell autonomous, because Gata4expression can occur in mutant cells. It has previously been reported that, when cultured in vitro, Nanogdeficient ICMs differentiate into cells morphologically resembling parietal endoderm-like cells, but do not give rise to ES cells (Mitsui et al., 2003). A recent replication of these experiments disputes these observations claiming the loss of PE differentiation potential in the Nanog-mutant outgrowths as well (Silva et al., 2009). We obtained outgrowths of mutant and control E3.5 embryos and assessed Oct4 and Gata4 expression and genotypes five days after plating (Figs. 3M–T). Three litters, obtained from natural mating by intercrossing both genetrap strains yielded 25 blastocysts, which were plated in single wells. 22 blastocysts attached and formed outgrowths. Most control embryos (15/17) formed large outgrowths consisting of at least three distinct cell types: (1) OCT4-positive/GATA4-negative EPI-derived cells; (2) OCT4positive/GATA4-positive PE-derived cells; and (3) large OCT4-negative/GATA4-negative cells, presumptive TE-derived cells (Figs. 3Q–T). Two control outgrowths lacked the EPI lineage, however still contained PE cells. In an independent experiment we could show that NANOG expression never overlapped with GATA4 in controls (Supplementary Fig. S1D). All outgrowths of the mutant blastocysts (n = 5), however, consisted only of PE-derived cells and presumptive TE derivatives, lacking the OCT4-positive/GATA4-negative EPI-derived population (Figs.3 M–P). In summary, Nanog-deficient E3.5 blastocysts, although not expressing Gata4 in vivo, can give rise to GATA4-positive outgrowths in vitro. Hence, we corroborate that Nanog is indispensable for the formation of a functional EPI; however, we also show that it is not required for Gata4 repression and inhibition of PE differentiation in the embryo.
blastocysts exhibited the whole range of none to almost exclusive ES cell contribution to the EPI (not shown). Remarkably, though usually resorbed at this time point, embryos derived from Nanog-null blastocysts could also be isolated (Fig. 4B). LacZ staining and sectioning of these embryos revealed that EPI-derived tissues, such as the embryonic ectoderm and the embryonic and extra-embryonic mesoderm, were exclusively ES cell derived, i.e. LacZ-negative (Fig. 4B). This supports the idea of a cell-autonomous requirement of Nanog for formation of the EPI lineage, even in a wild-type niche. Conversely, TEand PE-derived tissues in the rescued null embryos were LacZ positive, i.e. host blastocyst-derived. The presence of PE-derived tissues such as the visceral yolk sac endoderm (Fig. 4B), suggests a Nanog/EPI requirement for proper PE formation and shows the preserved PE differentiation potential of Nanog-mutant cells. Nanog-deficient extra-embryonic tissues can support development post gastrulation. E10.5 embryos obtained by injecting GFP-expressing ES cells into Nanog-null blastocysts look morphologically normal and display strong fluorescence in the entire embryo (Figs. 4C, D). Sections from throughout the embryo were examined for GFP and ΒGAL by IF (Fig. 4E-E"). With the exception of a few β-GAL-positive/ GFP-negative (embryo-derived) cells in the gut endoderm, all cells of the embryo proper were GFP-positive, thus ES cell-derived. A possible minor contribution of the extra-embryonic visceral endoderm to the definitive gut endoderm has been shown previously (Kwon et al., 2008). Conversely, controls showed weaker live GFP fluorescence in whole embryos due to the chimerism of GFP- or Β-GAL-positive cells, as shown on the analyzed sections (Figs. 4G–I). Notably, the strong GFP signal in the heart appears to result from higher activity of the chicken Β-actin promoter in the heart mesoderm (unpublished observations). Stained sections of the yolk sac derived from mutant embryos demonstrated the dual origin of this structure: the PEderived layer was exclusively Β-GAL-positive, whereas the EPIderived mesodermal compartment was entirely GFP-positive (Fig. 4F). As expected, yolk sacs of controls were chimeric in the mesodermal compartment but not in the endoderm derived layer (Fig 4J). Remarkably, all extra-embryonic tissues of the mutant blastocyst-derived embryos were normal in size and shape and had to be identified by PCR genotyping. This suggests that Nanog is dispensable in extra-embryonic tissues, at least until E12.5, the latest stage we analyzed. Thus, the lethality of the Nanog-null mutation can be rescued by providing a wt EPI, but Nanog-null cells themselves are still not competent to contribute to the EPI. However, in the presence of a wt EPI they can form all the extra-embryonic lineages derived from the TE and PE, functional at least until mid-gestation. Nanog functions in the blastocyst
Nanog cell- and non-cell autonomous requirement in the blastocyst To determine the potential of the mutant ICM cells in PE/EPI formation and the role of Nanog in this process blastocyst complementation experiments were performed. Although we had observed Gata4 expression in blastocyst outgrowths previously, the functionality of such putative PE cells is unclear and can only be properly addressed in vivo. In the paradigm of blastocyst complementation ES cells do not contribute to TE or PE when injected into blastocysts, yet they have the competence to contribute to all EPI-derived tissues (Beddington and Robertson, 1989). Moreover, ES cells have the capacity to form viable, fully ES-derived embryos when aggregated with extra-embryonic tissues (Nagy et al., 1990). We generated chimeric embryos by injecting wild-type W4 ES cells or GFP-labeled W4 ES cells into E3.5 blastocysts derived from heterozygous GT1-/GT2-Nanog intercrosses and analyzed lineage contributions of host and injected cells. β-Galactosidase (β-Gal) expression from the ROSA26 locus in the host embryos allowed tracing of blastocyst-derived cells (see supplementary Fig. S2 for mating scheme). At E7.5 embryos derived from injections into control
Early lineage segregations in mouse embryonic development were originally described as a succession of discrete, well ordered fate decisions, driven by the antagonistic functions of lineage-specific transcription factors (Fig. 5A) (Zernicka-Goetz et al., 2009). Only recently has careful analysis of gene expression patterns in vivo disclosed the dynamics and flexibility of the decision-making processes in the embryo and brought into question the mutual repression proposed for the lineage-driving transcription factors (Dietrich and Hiiragi, 2007; Plusa et al., 2008; Zernicka-Goetz et al., 2009). It appears that the mechanisms directing the formation of PE, TE and EPI are more sophisticated and enmeshed than anticipated. For instance, new evidence involves Nanog in a cross-regulation with Cdx2 (Chen et al., 2009). However, we did not observe Cdx2 deregulation in our mutants, thus this interaction may be subordinate in vivo. The role of Nanog in PE formation Since its first description Nanog has been thought to prevent PE differentiation, perhaps through interaction with a binding site in the
D.M. Messerschmidt, R. Kemler / Developmental Biology 344 (2010) 129–137
135
Fig. 4. (A) Scheme of blastocyst complementation experiments. (B) Section of an E7.5 embryo derived from ES cell-injection in a Nanog-deficient blastocyst. ES/EPI-derived tissues are LacZ-negative: extra-embryonic mesoderm (ExM), embryonic mesoderm (EmM) and embryonic ectoderm (EmE). Blastocyst-derived tissues are LacZ-positive: extra-embryonic ectoderm (ExE) and endoderm of visceral yolk sac (VE). (C–J) Embryos derived from GFP-labeled ES cell-injection in mutant (C) and control (G) blastocysts. (D, H) GFP-detection in embryos shown in (C) and (G), respectively. (E) IF detection of GFP (green) and β-GAL (red) on transversal sections through the hindbrain (HB) region from the embryo shown in (C, section plane is indicated) (DAPI, blue; composite image of E' (GFP) and E" (β-GAL)). (I–I") Similar analysis for the control embryo shown in (G, section plane is indicated). Tissues indicated are neural ectoderm (E) and mesoderm (M). (F, J) Sections of the visceral yolk sac of the embryos shown in (C) and (D) stained for GFP (green), β-GAL (red) and nuclei (DAPI, blue). (Scale bars: 100 μm (B, E, I), 1 mm (C, D, G, H), 25 μm (F, J).
Gata6 promoter as described in ES cells (Fig. 5A) (Singh et al., 2007). Conversely, our data suggest that Nanog does not impede PE differentiation in vivo, a fact also supported by the latest knock-out studies in ES cells and embryos (Chambers et al., 2007; Silva et al., 2009). But have Nanog-mutant ICM cells indeed lost the ability to differentiate into PE (Silva et al. 2009)? We found they do give rise to PE in blastocyst outgrowth experiments. However, in contrast to the in vivo situation where PE formation is achieved within hours, in vitro outgrowths require several days to attach and form the typical
mixture of TE, PE and EPI derivatives. This extended culture period might be beneficial to PE formation. We also found occasional Gata4 expression in E4.5 Nanog-null embryos, suggesting that some signal required for PE induction must still be present in the mutant embryo. It is therefore possible that TE-derived signals are beneficial to PE outgrowth when plating whole embryos, while in experiments using ICMs only others merely observed TE-like derivatives (Silva et al., 2009). Culture conditions, media composition and the time-point of analysis could further influence Gata4 expression/detection.
136
D.M. Messerschmidt, R. Kemler / Developmental Biology 344 (2010) 129–137
conclusion, our careful analysis and manipulation of Nanog-deficient embryos reveals a new level of complexity in the processes that determine early cell fates in the mouse embryo and contributes a novel feature to the complex picture of PE formation. Acknowledgments We thank Verdon Taylor for helpful discussions, Heike Wollmann for critical reading of the manuscript, Riana Vogt for technical assistance. We also thank Caroline Johner and Benoit Kanzler from the MPI-IB animal and transgenic facilities. This work was funded by the MaxPlanck Society. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.ydbio.2010.04.020. References
Fig. 5. (A) Classical model of lineage decisions in the mouse pre-implantation embryo. Mutually antagonistic acting lineage-specific transcription factors drive successive differentiation events (TP, totipotent precursor). (B) Nanog is required cell autonomously for EPI induction but not for PE repression. (C) EPI cells are required non-cell autonomously for PE formation, putatively by a combination of paracrine signaling and cell–cell interaction.
Compound interactions in the blastocyst The upstream mechanisms leading to the mosaic composition of the ICM are as unknown as the processes involved in lineage restriction and positional sorting of PE and EPI cells (reviewed in (Zernicka-Goetz et al., 2009). Gata4 expression is dependent on Grb2–Ras–Map kinase signaling and one potential upstream component in this pathway is FGF4 (Chazaud et al., 2006). FGF4 as well as FGF4-receptor deletion affects PE formation (Arman et al., 1998; Chai et al., 1998; Feldman et al., 1995). However, the role, source, and timing of FGF and other signaling cascades are only beginning to be understood. Remarkably the inhibition of FGF/Erk prevents Gata4 expression in the ICM of embryos in vitro, resulting in the uniformly high expression levels of OCT4 and NANOG in all ICM cells (Supplementary Fig. S3) (Nichols et al., 2009). Once PE fate is induced however, FGF/Erk inhibition does not affect Gata4 expression (Nichols et al., 2009). Our results are in complete agreement with these findings, furthermore, they disclose the Nanog expressing EPI as the primary source of the PE inducing signal. We hypothesize that the loss of PE is the consequence of a nonfunctional EPI, unable to provide the cues, e.g. a paracrine signal such as FGF/Erk, required for promoting PE differentiation in neighboring ICM cells. Thus, we suggest that Nanog promotes the EPI fate, but does not prevent PE differentiation, in a cell autonomous manner (Fig. 5B). During fate commitment, these future EPI cells instruct their surroundings to adopt a PE fate, if they are perceptive (Fig. 5C). The ability to receive PE differentiation signals may again depend of Nanog/Oct4 expression levels (Chambers et al., 2007). The blastocyst complementation experiments clearly support this theory, as ES cell-derived EPIs can rescue lethality and PE formation in a non-cell autonomous manner in Nanog-null blastocysts. Whereas the inhibition of FGF/Erk blocks PE formation we were not able to induce Gata4 expression in Nanog mutants in a reverse approach, by FGF-4 (maximal 150 ng/ml) treatment. However, recent work by Yamanaka et al. published during the preparation of this manuscript, showed that treatment of embryos with very high doses of FGF-4 (500 ng/ml) plus heparin is indeed able to override the EPI-induction program resulting in 100% PE cells in the ICM (Yamanaka et al. 2010). This new evidence further supports our hypothesis of an EPI-derived PE inductive or promoting signal. In
Arman, E., Haffner-Krausz, R., Chen, Y., Heath, J.K., Lonai, P., 1998. Targeted disruption of fibroblast growth factor (FGF) receptor 2 suggests a role for FGF signaling in pregastrulation mammalian development. Proc. Natl. Acad. Sci. U S A 95, 5082–5087. Beddington, R.S., Robertson, E.J., 1989. An assessment of the developmental potential of embryonic stem cells in the midgestation mouse embryo. Development 105, 733–737. Chai, N., Patel, Y., Jacobson, K., McMahon, J., McMahon, A., Rappolee, D.A., 1998. FGF is an essential regulator of the fifth cell division in preimplantation mouse embryos. Dev. Biol. 198, 105–115. Chambers, I., Colby, D., Robertson, M., Nichols, J., Lee, S., Tweedie, S., Smith, A., 2003. Functional expression cloning of Nanog, a pluripotency sustaining factor in embryonic stem cells. Cell 113, 643–655. Chambers, I., Silva, J., Colby, D., Nichols, J., Nijmeijer, B., Robertson, M., Vrana, J., Jones, K., Grotewold, L., Smith, A., 2007. Nanog safeguards pluripotency and mediates germline development. Nature 450, 1230–1234. Chazaud, C., Yamanaka, Y., Pawson, T., Rossant, J., 2006. Early lineage segregation between epiblast and primitive endoderm in mouse blastocysts through the Grb2– MAPK pathway. Dev. Cell 10, 615–624. Chen, L., Yabuuchi, A., Eminli, S., Takeuchi, A., Lu, C.W., Hochedlinger, K., Daley, G.Q., 2009. Cross-regulation of the Nanog and Cdx2 promoters. Cell Res. 19, 1052–1061. Copp, A.J., 1978. Interaction between inner cell mass and trophectoderm of the mouse blastocyst. I. A study of cellular proliferation. J. Embryol. Exp. Morphol. 48, 109–125. Dietrich, J.E., Hiiragi, T., 2007. Stochastic patterning in the mouse pre-implantation embryo. Development 134, 4219–4231. Feldman, B., Poueymirou, W., Papaioannou, V.E., DeChiara, T.M., Goldfarb, M., 1995. Requirement of FGF-4 for postimplantation mouse development. Science 267, 246–249. Fujikura, J., Yamato, E., Yonemura, S., Hosoda, K., Masui, S., Nakao, K., Miyazaki Ji, J., Niwa, H., 2002. Differentiation of embryonic stem cells is induced by GATA factors. Genes Dev. 16, 784–789. Gardner, R.L., Beddington, R.S., 1988. Multi-lineage ‘stem’ cells in the mammalian embryo. J. Cell Sci. Suppl. 10, 11–27. Hart, A.H., Hartley, L., Ibrahim, M., Robb, L., 2004. Identification, cloning and expression analysis of the pluripotency promoting Nanog genes in mouse and human. Dev. Dyn. 230, 187–198. Kan, N.G., Stemmler, M.P., Junghans, D., Kanzler, B., de Vries, W.N., Dominis, M., Kemler, R., 2007. Gene replacement reveals a specific role for E-cadherin in the formation of a functional trophectoderm. Development 134, 31–41. Koutsourakis, M., Langeveld, A., Patient, R., Beddington, R., Grosveld, F., 1999. The transcription factor GATA6 is essential for early extraembryonic development. Development 126, 723–732. Kwon, G.S., Viotti, M., Hadjantonakis, A.K., 2008. The endoderm of the mouse embryo arises by dynamic widespread intercalation of embryonic and extraembryonic lineages. Dev. Cell 15, 509–520. Meilhac, S.M., Adams, R.J., Morris, S.A., Danckaert, A., Le Garrec, J.F., Zernicka-Goetz, M., 2009. Active cell movements coupled to positional induction are involved in lineage segregation in the mouse blastocyst. Dev. Biol. 331, 210–221. Mitsui, K., Tokuzawa, Y., Itoh, H., Segawa, K., Murakami, M., Takahashi, K., Maruyama, M., Maeda, M., Yamanaka, S., 2003. The homeoprotein Nanog is required for maintenance of pluripotency in mouse epiblast and ES cells. Cell 113, 631–642. Morrisey, E.E., Tang, Z., Sigrist, K., Lu, M.M., Jiang, F., Ip, H.S., Parmacek, M.S., 1998. GATA6 regulates HNF4 and is required for differentiation of visceral endoderm in the mouse embryo. Genes Dev. 12, 3579–3590. Nagy, A., Gocza, E., Diaz, E.M., Prideaux, V.R., Ivanyi, E., Markkula, M., Rossant, J., 1990. Embryonic stem cells alone are able to support fetal development in the mouse. Development 110, 815–821. Nichols, J., Silva, J., Roode, M., Smith, A., 2009. Suppression of Erk signalling promotes ground state pluripotency in the mouse embryo. Development 136, 3215–3222.
D.M. Messerschmidt, R. Kemler / Developmental Biology 344 (2010) 129–137 Plusa, B., Piliszek, A., Frankenberg, S., Artus, J., Hadjantonakis, A.K., 2008. Distinct sequential cell behaviours direct primitive endoderm formation in the mouse blastocyst. Development 135, 3081–3091. Ralston, A., Rossant, J., 2008. Cdx2 acts downstream of cell polarization to cellautonomously promote trophectoderm fate in the early mouse embryo. Dev. Biol. 313, 614–629. Schnutgen, F., De-Zolt, S., Van Sloun, P., Hollatz, M., Floss, T., Hansen, J., Altschmied, J., Seisenberger, C., Ghyselinck, N.B., Ruiz, P., Chambon, P., Wurst, W., von Melchner, H., 2005. Genomewide production of multipurpose alleles for the functional analysis of the mouse genome. Proc. Natl. Acad. Sci. U S A 102, 7221–7226. Silva, J., Nichols, J., Theunissen, T.W., Guo, G., van Oosten, A.L., Barrandon, O., Wray, J., Yamanaka, S., Chambers, I., Smith, A., 2009. Nanog is the gateway to the pluripotent ground state. Cell 138, 722–737. Singh, A.M., Hamazaki, T., Hankowski, K.E., Terada, N., 2007. A heterogeneous expression pattern for Nanog in embryonic stem cells. Stem Cells 25, 2534–2542.
137
Soudais, C., Bielinska, M., Heikinheimo, M., MacArthur, C.A., Narita, N., Saffitz, J.E., Simon, M.C., Leiden, J.M., Wilson, D.B., 1995. Targeted mutagenesis of the transcription factor GATA-4 gene in mouse embryonic stem cells disrupts visceral endoderm differentiation in vitro. Development 121, 3877–3888. Stemmler, M.P., Hecht, A., Kemler, R., 2005. E-cadherin intron 2 contains cis-regulatory elements essential for gene expression. Development 132, 965–976. Yamanaka, Y., Lanner, F., Rossant, J., 2010. FGF signal-dependent segregation of primitive endoderm and epiblast in the mouse blastocyst. Development 137, 715–724. Zambrowicz, B.P., Imamoto, A., Fiering, S., Herzenberg, L.A., Kerr, W.G., Soriano, P., 1997. Disruption of overlapping transcripts in the ROSA beta geo 26 gene trap strain leads to widespread expression of beta-galactosidase in mouse embryos and hematopoietic cells. Proc. Natl. Acad. Sci. U S A 94, 3789–3794. Zernicka-Goetz, M., Morris, S.A., Bruce, A.W., 2009. Making a firm decision: multifaceted regulation of cell fate in the early mouse embryo. Nat. Rev. Genet. 10, 467–477.
Contents lists available at ScienceDirect
Developmental Biology j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / d e v e l o p m e n t a l b i o l o g y
Nanog is required for primitive endoderm formation through a non-cell autonomous mechanism Daniel M. Messerschmidt a,b,⁎, Rolf Kemler a a b
Department of Molecular Embryology, Max Planck Institute for Immunobiology, Stuebeweg 51, 79108 Freiburg, Germany Department of Mammalian Development, Institute of Medical Biology (IMB), 8A Biomedical Grove, #06-06 Immunos Singapore 138648, Singapore
a r t i c l e
i n f o
Article history: Received for publication 1 December 2009 Revised 9 April 2010 Accepted 22 April 2010 Available online 12 May 2010 Keywords: Nanog Epiblast Primitive endoderm Trophectoderm Blastocyst
a b s t r a c t Early lineage segregation in mouse development results in two, either CDX2- or OCT4/NANOG-positive, cell populations. CDX2-positive cells form the trophectoderm (TE), OCT4/NANOG-positive cells the inner cell mass (ICM). In a second lineage decision ICM cells segregate into Epiblast (EPI) and primitive endoderm (PE). EPI and PE formation depend on the activity of the transcription factors Nanog and Gata4/6. A role for Nanog, a crucial pluripotency factor, in preventing PE differentiation has been proposed, as outgrowths of mutant ICMs result in PE, but not EPI derivatives. We established Nanog-mutant mouse lines and analyzed EPI and PE formation in vivo. Surprisingly, Gata4 expression in mutant ICM cells is absent or strongly decreased, thus loss of Nanog does not result in precocious endoderm differentiation. However, Nanogdeficient embryos retain the capacity to form PE in chimeric embryos and, in contrast to recent reports, in blastocyst outgrowths. Based on our findings we propose a non-cell autonomous requirement of Nanog for proper PE formation in addition to its essential role in EPI determination. © 2010 Elsevier Inc. All rights reserved.
Introduction The first lineage segregations in the pre-implantation embryo establish the basis for mouse embryonic development. At the compacting morula stage, trophectoderm (TE) and inner cell mass (ICM) emerge; later the ICM gives rise to primitive endoderm (PE) and epiblast (EPI) (Gardner and Beddington, 1988). Recent evidence suggests that unlike TE/ICM segregation in the morula (Ralston and Rossant, 2008), lineage-specific gene expression precedes EPI- and PE-cell sorting in the ICM. Before the PE becomes evident, the ICM is already composed of two, committed cell types arranged in a ‘salt-and-pepper-like’ distribution (Chazaud et al., 2006). Transcription factors determine the adopted fate: PE cells express Gata6 and Gata4 (Koutsourakis et al., 1999; Morrisey et al., 1998), and EPI cells the homeobox transcription factor Nanog (Chambers et al., 2003; Mitsui et al., 2003). Nanog expression commences at the morula stage and becomes gradually restricted to the ICM during blastocyst formation and expansion (Hart et al., 2004; Mitsui et al., 2003; Zernicka-Goetz et al., 2009). Disruption of Nanog results in peri-implantation lethality and although mutant blastocysts are morphologically normal and induce decidualization they don't form a proper EPI. Embryonic stem (ES) cells cannot be isolated from Nanogdeficient blastocysts; the cells recovered from blastocyst outgrowths display a parietal endoderm-like morphology (Mitsui et al., 2003). Gata4/6-deficient embryos show deficits in establishing PE and fail to
⁎ Corresponding author. Fax: + 65 6464 2048. E-mail address: [email protected] (D.M. Messerschmidt). 0012-1606/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.ydbio.2010.04.020
maintain PE derivatives (Koutsourakis et al., 1999; Morrisey et al., 1998; Soudais et al., 1995), while over-expression of Gata4/6 in ES cells is sufficient to induce PE differentiation (Fujikura et al., 2002). A direct link between Nanog and Gata6 was proposed and NANOG was shown to bind the Gata6 promoter in ES cells (Mitsui et al., 2003; Singh et al., 2007). While early knock-out studies reported endodermal differentiation of Nanog-mutant ICMs, a recent study addressing the role of Nanog in the generation of pluripotency suggested that Nanog-mutant ICM cells are blocked in a transitional stage and are not capable of acquiring the pluripotent ground state and lose their differentiation potential (Silva et al., 2009). These authors propose that Nanog-null ICM cells are thus only capable of TE differentiation or else undergo apoptosis, as opposed to the original description of PE outgrowth (Mitsui et al., 2003). We established and examined Nanog gene-trap mouse lines and found that surprisingly PE formation in mutant blastocysts is profoundly repressed rather than precociously activated. However, in this study we took the analysis from Nanog mutants further, showing that surviving ICM cells from Nanog-null embryos retain the capacity for PE formation in vivo and at least transiently in vitro. Thus, our findings help resolve the recently emerged controversy of Nanog action in the early mouse embryo (Mitsui et al., 2003; Silva et al., 2009), and moreover, we demonstrate that PE formation depends on a functional EPI, hence directly or indirectly, on Nanog expression. In the model we propose, Nanog expression is required for the formation of the EPI lineage, which in turn through a non-cell autonomous mechanism, possibly mediated by FGF/Erk signaling (Chazaud et al., 2006; Nichols et al., 2009), is vital for proper PE formation.
130
D.M. Messerschmidt, R. Kemler / Developmental Biology 344 (2010) 129–137
Materials and methods
Results and discussion
Generation of gene-trap mice and genotyping
Nanog gene-trap embryos
Two ES cell lines (D042B07 [GT1] and D047A12 [GT2]) with disrupted Nanog alleles were obtained from the GSF (Munich) (Schnutgen et al., 2005) (Supplementary Fig. S1) and used for blastocyst injection. Chimeras were mated with C57BL/6 females to generate the Nanoggt1/+ and Nanoggt2/+ mouse lines. Genotyping was performed by PCR on genomic DNA isolated from embryos, yolk sacs or mouse-tail biopsies using primers specific for the GT1- or GT2-allele. For blastocyst complementation the Gt(Rosa)26Sor/J (ROSA26) line (Zambrowicz et al., 1997) was bred onto the GT1-Nanog background and matings performed as detailed in supplementary Fig. S2.
Embryos were flushed (E3.5) or dissected (E4.5) from the uterus following natural matings, counting noon of the day the plug was detected as E0.5. Antibody detection assays were performed as described in (Kan et al., 2007). Self-made affinity-purified rabbit antibody for NANOG was used at a dilution of 1:50, OCT4 (monoclonal mouse, Santa Cruz Biotechnology) at 1:200, CDX2 (monoclonal mouse, BioGenex) at 1:200 and GATA4 (polyclonal rabbit, Santa Cruz Biotechnology) at 1:100. Secondary antibodies used were Alexa Fluor 488 goat anti-rabbit and Alexa Fluor 594 goat anti-mouse (Molecular Probes; 1:500). Nuclei were stained with 10 µM DAPI (Molecular Probes) added to the last wash. For double staining of blastocyst outgrowths with GATA4 an additional αNANOG antibody (R&D, polyclonal goat, 1:10) was used.
Despite the relatively early expression of Nanog in pre-implantation embryos, Nanog mutants form normal morulae, which compact and give rise to morphologically normal blastocysts (Hart et al., 2004; Mitsui et al., 2003). We established two independent gene-trap mouse lines carrying a β-Geo insertion in the first intron of the Nanog locus (Supplementary Fig. S1A, B). We first investigated homozygous offspring of both lines for NANOG expression and found that embryos homozygous for either allele (Nanoggt1/gt1 or Nanoggt2/gt2) failed to express NANOG at pre-implantation stages (supplementary Fig. S1C). We failed to derive live homozygous offspring from either gene-trap line, and did not succeed in isolating homozygous mutants at E7.5, whereas phenotypically undistinguishable heterozygous and wild-type embryos were present. The number of empty deciduomas found, matched the expected Mendelian ratio of 25%. Thus, we concluded that, although genetically different in terms of the gene-trap integration site within the first intron, both, Nanoggt1/gt1 and Nanoggt2/gt2 embryos phenocopied the previously described Nanog-null mutant (Mitsui et al., 2003) and have identical phenotypes. Further, blastocysts carrying both gene-trap alleles (Nanoggt1/gt2) were indistinguishable from homozygous embryos for either allele (Nanoggt1/gt1 or Nanoggt2/gt2) (Fig. 1 A, G and supplementary Fig. S1C). All breeding for experiments shown was carried out using intercrosses of both strains allowing accurate identification of null-mutants by PCR. Thus, embryos carrying both disrupted alleles are referred to as ‘mutants’, and heterozygous and wild-type embryos are grouped as ‘controls’.
ES cell culture, blastocyst outgrowths and embryo culture
TE/ICM lineage segregation is independent of Nanog
ES cells were cultured on feeder cells in ES cell medium (DMEM with 15% FCS, supplemented with 500 U/ml LIF [Chemicon, ESGRO #ES1107]). For blastocyst outgrowths embryos were recovered at E3.5 and cultured in ES cell medium on gelatinized tissue culture chambers (Lab-tek #177402). After 5 days, attached outgrowths were fixed in 4% PFA/PBS, permeabilized in 0.5% Triton X-100/PBS and blocked in 10% FCS in 0.1% Triton X-100/PBS (PBSTx). First and second antibody incubation was performed in blocking solution as described above. Embryos were cultured in KSOM microdrops covered with mineral oil at 37 °C and 5% CO2 atmosphere. Inhibitor treatment on wild-type embryos was performed in KSOM microdrops containing 20 µM SU5402 in DMSO or DSMO only for controls. FGF treatment was performed on embryos from heterozygous intercrosses isolated at E2.5 and cultured for 48 in the presence of human recombinant FGF4 (10− 8 M; Sigma) in KSOM microdrops. Blastocysts were PCR genotyped after GATA4 immunostaining.
We analyzed TE and ICM formation by examining expression of the TE-marker CDX2 and the ICM-marker OCT4 in mutant and control E3.5 blastocysts. CDX2 was restricted to the nuclei of the TE cells in both mutants (Figs. 1A–C) and controls (Figs. 1D–F). Though NANOG was absent in the mutants, we did not find CDX2 in ICM cells as has been described recently (Chen et al., 2009). OCT4 was detectable in all cells of the early blastocyst stage (Figs. 1H, K), thus CDX2 expression does not inevitably drive Oct4 repression in vivo. Only later at implantation does OCT4 become restricted to the ICM (Dietrich and Hiiragi, 2007). Oct4 expression, however, was unaffected by the loss of Nanog (Figs. 1G–I). To ascertain whether loss of Nanog affects cell proliferation, we counted TE and ICM cells in mutant and control blastocysts on stacks of confocal zseries images of CDX2/DAPI stained blastocysts (Fig. 1M). CDX2-positive cells were subtracted from the total cell number to determine the number of ICM cells. Mutant embryos were compared to their control littermates to assure stage/age-matched blastocysts. In one litter ICM cell numbers were 20.3 ± 2.1 (standard deviation, sd) in mutant (n =3) and 22.3 ± 3.9 (sd) in control embryos (n = 9). TE cell numbers were 32.3 ± 2.3 (sd) in mutants and 30.1 ± 2.4 (sd) in controls. Therefore, we did not find significant differences in the number of ICM (Student's t-test, p = 0.3) or TE (p = 0.23) cells in mutants and controls. In a second litter ICM cell numbers were 24 (mutant, n =1) and 24.5 ± 1.9 (sd) (controls, n = 8) and TE cell numbers were 32 (mutant) and 33 ± 5.5 (sd) (control). We concluded that, whilst Nanog is expressed at the morula stage, it is neither required for segregation of TE and ICM, nor for cell proliferation until the E3.5 blastocyst stage.
Whole-mount immunodetection
Generation and analysis of chimeric embryos Wild-type W4 or stable pCAGGS-eGFP transfected W4 ES cells were injected into E3.5 blastocysts obtained from matings according to supplementary Fig. S2. All embryos resulting from these matings carry a ubiquitously expressed β-galactosidase reporter, allowing tracing of embryo-derived cells. After ES cell-injection the blastocysts were retransferred into pseudo-pregnant females and embryos isolated at indicated time points. LacZ staining was performed as described previously (Stemmler et al., 2005). For IF, embryos and yolk sacs were frozen in OCT (Tissue-tek) and sectioned at 14 µm using a cryostat (Leica). Sections were blocked in 10% FCS/PBSTx and incubated with αGFP (polyclonal sheep, 1:400; AbD) and α-β-galactosidase (polyclonal rabbit, 1:1000; MPBiomedicals) antibodies for 1 h at RT. GFP was visualized with biotin-conjugated donkey-anti-sheep Ig (Jackson Labs; 1:300) and FITC-conjugated streptavidin (Jackson Labs; 1:400); β-GAL with Cy3-conjugated donkey-anti-rabbit Ig (Jackson Labs; 1:300).
Nanog mutant ICMs show defects at implantation The same marker-gene analysis was conducted on implanting embryos (E4.5), which were directly dissected from the uterus. All embryos had hatched and attached to the uterine wall at the time of isolation. In control embryos NANOG-positive and NANOG-negative ICM cells were distinct populations (Figs. 2E, H), whereas at E3.5 they
D.M. Messerschmidt, R. Kemler / Developmental Biology 344 (2010) 129–137
131
Fig. 1. IF detection of NANOG (A, D, G, J), CDX2 (B, E) and OCT4 (H, K) in E3.5 Nanog-mutant (A–C, G–I) and control (D–F, J–L) blastocysts. NANOG-positive cells (yellow arrowheads) are intermingled with future PE cells (white arrowheads). Delineated areas (A–F) mark the ICM devoid of CDX2. (C, F, I, L) Merged images. (M) Quantification of ICM cells in mutant and control E3.5 blastocysts. Numbers of embryos are indicated in each bar. ICM cell numbers are not significantly reduced in mutants (Student's t-test, #p = 0.3). (Scale bars: 50 μm).
132
D.M. Messerschmidt, R. Kemler / Developmental Biology 344 (2010) 129–137
were arranged in a ‘salt-and-pepper-like’ manner (Figs. 1D, F, J, L, yellow and white arrowheads). OCT4 was absent in most TE cells in E4.5 embryos and the ICM expression domain was divided into two populations in controls, either overlapping with NANOG (Figs. 2E, H, outlined area) or GATA4 (Fig. 3L). The Nanog-deficient E4.5 embryos were comparable in size to their littermates and had implanted, indicating normal development. In mutants, all ICM cells expressed OCT4, however, the number of ICM cells appeared to be reduced (Figs. 2B, D). We observed frequent pyknotic and fragmented nuclei in mutant embryos, indicating apoptosis (Fig. 2C, red arrowheads). This was confirmed by the high incidence of active Caspase3 positive cells in mutant embryos isolated at E4.5 (Fig. 2I, red arrowhead). In contrast we observed few active Caspase3 positive cells in controls at the examined implantation stage (Fig. 2K). We confirmed our observation by examining additional E4.5 embryos isolated from C57BL/6 wild-type
mice. In two litters, containing 12 and 10 implanted embryos we found only two embryos each containing a single active Caspase3 positive cell localized to the ICM in one, and to the TE in the other embryo. At this stage the PE endoderm layer is formed and the sorting/migration of EPI and PE cells is completed (Figs. 2H and 3L). This result is in agreement with reports documenting decreasing cell death in embryos upon implantation (Copp, 1978), although apoptotic cells can be readily detected in wild-type pre-implantation embryos during the sorting process of EPI and PE cells within the ICM (Meilhac et al., 2009; Plusa et al., 2008). Additionally, our detection of apoptotic cells by staining for active Caspase3 reflects a ‘snap-shot’ of apoptotic events in the embryos, rather than observations made in an extended period of time during live cell imaging (Meilhac et al., 2009; Plusa et al., 2008). The reduction of ICM cell numbers in Nanog mutants was quantified by counting cells in embryos stained for CDX2 (TE) or
Fig. 2. IF detection of NANOG (A, E) and OCT4 (B, F) in Nanog-mutant and control E4.5 embryos. NANOG is undetectable in the mutant (A), confined to the EPI in the control (E, H, delineated area) but absent in the PE. OCT4 expression is absent in the TE, but present in mutant and control ICMs (B, F). (C, G) Red arrowheads indicate DAPI stained, fragmented and pyknotic nuclei detectable in the mutant (C), but not in control ICMs (G). (D, H) Overlay of NANOG, OCT4 and DAPI staining. (I–L) Active Caspase3 and DAPI staining; apoptotic cells are detectable in the mutant (I, J; red arrowheads) but not in the control (K, L). (M) Quantification of ICM cells in mutants and controls. Numbers of embryos are indicated in each bar. Cell numbers are highly significantly reduced in mutant embryos (Student's t-test, *p = 0.0059). (Scale bars: 50 μm).
D.M. Messerschmidt, R. Kemler / Developmental Biology 344 (2010) 129–137
133
Fig. 3. IF localization of GATA4 (A, D, G, J) and OCT4 (B, E, H, K) in Nanog-mutant (A–C, G–I) and control (D–F, J–L) E3.5 (A–F) and E4.5 (G–L) embryos. GATA4-positive cells in the mutant embryo are indicated (G, white arrowheads). (C, F, I, L) Overlay of GATA4, OCT4 and DAPI staining. (M–F) Blastocyst outgrowths of a Nanog-mutant (MP) and control (Q–T) blastocyst stained for GATA4 (M, Q), OCT4 (N, R) and Nuclei (O S). (P, S) Overlay of GATA4, OCT4 and DAPI staining. Lineage derivatives are indicated by white (PE), yellow (EPI) and red arrowheads (TE). (Scale bars: 100 μm).
OCT4 (ICM), and DAPI (total cell number). This revealed a highly significant (Student's t-test; p = 0.0059) loss of ICM cells in the Nanog mutants (26.25 ± 0.81 (sd) n = 4) compared to controls (35.3 ± 1.52
(sd) n = 3) (Fig. 2M). Thus, the first indications of a Nanog mutant phenotype become apparent in the ICM at E4.5 when PE and EPI segregate.
134
D.M. Messerschmidt, R. Kemler / Developmental Biology 344 (2010) 129–137
Nanog mutants don't form primitive endoderm To determine whether Nanog prevents endoderm differentiation in vivo, we analyzed mutant embryos for expression of the PE-marker GATA4. Surprisingly, Nanog-null-mutants did not display precocious PE marker-gene expression as expected (Mitsui et al., 2003), but were entirely devoid of Gata4-expressing cells (Fig. 3A), whereas single GATA4-positive cells could be observed in control ICMs at E3.5 (Fig. 3D). This lack of Gata4 expression raised the possibility of a Nanog requirement for PE formation. In E4.5 control embryos the PE was visible as a discrete structure lining the ICM towards the blastocoel (Fig. 3J). Nine of 41 embryos stained for Gata4 were identified as mutants by PCR. Their ICM was disarranged; a PE layer was not recognizable. Only three of these embryos contained few (four or less) GATA4-positive cells (Fig. 3G, white arrowheads). The other six mutant embryos did not express GATA4. This observation supports our hypothesis of Nanog requirement for proper PE formation, but also shows that this requirement is likely not cell autonomous, because Gata4expression can occur in mutant cells. It has previously been reported that, when cultured in vitro, Nanogdeficient ICMs differentiate into cells morphologically resembling parietal endoderm-like cells, but do not give rise to ES cells (Mitsui et al., 2003). A recent replication of these experiments disputes these observations claiming the loss of PE differentiation potential in the Nanog-mutant outgrowths as well (Silva et al., 2009). We obtained outgrowths of mutant and control E3.5 embryos and assessed Oct4 and Gata4 expression and genotypes five days after plating (Figs. 3M–T). Three litters, obtained from natural mating by intercrossing both genetrap strains yielded 25 blastocysts, which were plated in single wells. 22 blastocysts attached and formed outgrowths. Most control embryos (15/17) formed large outgrowths consisting of at least three distinct cell types: (1) OCT4-positive/GATA4-negative EPI-derived cells; (2) OCT4positive/GATA4-positive PE-derived cells; and (3) large OCT4-negative/GATA4-negative cells, presumptive TE-derived cells (Figs. 3Q–T). Two control outgrowths lacked the EPI lineage, however still contained PE cells. In an independent experiment we could show that NANOG expression never overlapped with GATA4 in controls (Supplementary Fig. S1D). All outgrowths of the mutant blastocysts (n = 5), however, consisted only of PE-derived cells and presumptive TE derivatives, lacking the OCT4-positive/GATA4-negative EPI-derived population (Figs.3 M–P). In summary, Nanog-deficient E3.5 blastocysts, although not expressing Gata4 in vivo, can give rise to GATA4-positive outgrowths in vitro. Hence, we corroborate that Nanog is indispensable for the formation of a functional EPI; however, we also show that it is not required for Gata4 repression and inhibition of PE differentiation in the embryo.
blastocysts exhibited the whole range of none to almost exclusive ES cell contribution to the EPI (not shown). Remarkably, though usually resorbed at this time point, embryos derived from Nanog-null blastocysts could also be isolated (Fig. 4B). LacZ staining and sectioning of these embryos revealed that EPI-derived tissues, such as the embryonic ectoderm and the embryonic and extra-embryonic mesoderm, were exclusively ES cell derived, i.e. LacZ-negative (Fig. 4B). This supports the idea of a cell-autonomous requirement of Nanog for formation of the EPI lineage, even in a wild-type niche. Conversely, TEand PE-derived tissues in the rescued null embryos were LacZ positive, i.e. host blastocyst-derived. The presence of PE-derived tissues such as the visceral yolk sac endoderm (Fig. 4B), suggests a Nanog/EPI requirement for proper PE formation and shows the preserved PE differentiation potential of Nanog-mutant cells. Nanog-deficient extra-embryonic tissues can support development post gastrulation. E10.5 embryos obtained by injecting GFP-expressing ES cells into Nanog-null blastocysts look morphologically normal and display strong fluorescence in the entire embryo (Figs. 4C, D). Sections from throughout the embryo were examined for GFP and ΒGAL by IF (Fig. 4E-E"). With the exception of a few β-GAL-positive/ GFP-negative (embryo-derived) cells in the gut endoderm, all cells of the embryo proper were GFP-positive, thus ES cell-derived. A possible minor contribution of the extra-embryonic visceral endoderm to the definitive gut endoderm has been shown previously (Kwon et al., 2008). Conversely, controls showed weaker live GFP fluorescence in whole embryos due to the chimerism of GFP- or Β-GAL-positive cells, as shown on the analyzed sections (Figs. 4G–I). Notably, the strong GFP signal in the heart appears to result from higher activity of the chicken Β-actin promoter in the heart mesoderm (unpublished observations). Stained sections of the yolk sac derived from mutant embryos demonstrated the dual origin of this structure: the PEderived layer was exclusively Β-GAL-positive, whereas the EPIderived mesodermal compartment was entirely GFP-positive (Fig. 4F). As expected, yolk sacs of controls were chimeric in the mesodermal compartment but not in the endoderm derived layer (Fig 4J). Remarkably, all extra-embryonic tissues of the mutant blastocyst-derived embryos were normal in size and shape and had to be identified by PCR genotyping. This suggests that Nanog is dispensable in extra-embryonic tissues, at least until E12.5, the latest stage we analyzed. Thus, the lethality of the Nanog-null mutation can be rescued by providing a wt EPI, but Nanog-null cells themselves are still not competent to contribute to the EPI. However, in the presence of a wt EPI they can form all the extra-embryonic lineages derived from the TE and PE, functional at least until mid-gestation. Nanog functions in the blastocyst
Nanog cell- and non-cell autonomous requirement in the blastocyst To determine the potential of the mutant ICM cells in PE/EPI formation and the role of Nanog in this process blastocyst complementation experiments were performed. Although we had observed Gata4 expression in blastocyst outgrowths previously, the functionality of such putative PE cells is unclear and can only be properly addressed in vivo. In the paradigm of blastocyst complementation ES cells do not contribute to TE or PE when injected into blastocysts, yet they have the competence to contribute to all EPI-derived tissues (Beddington and Robertson, 1989). Moreover, ES cells have the capacity to form viable, fully ES-derived embryos when aggregated with extra-embryonic tissues (Nagy et al., 1990). We generated chimeric embryos by injecting wild-type W4 ES cells or GFP-labeled W4 ES cells into E3.5 blastocysts derived from heterozygous GT1-/GT2-Nanog intercrosses and analyzed lineage contributions of host and injected cells. β-Galactosidase (β-Gal) expression from the ROSA26 locus in the host embryos allowed tracing of blastocyst-derived cells (see supplementary Fig. S2 for mating scheme). At E7.5 embryos derived from injections into control
Early lineage segregations in mouse embryonic development were originally described as a succession of discrete, well ordered fate decisions, driven by the antagonistic functions of lineage-specific transcription factors (Fig. 5A) (Zernicka-Goetz et al., 2009). Only recently has careful analysis of gene expression patterns in vivo disclosed the dynamics and flexibility of the decision-making processes in the embryo and brought into question the mutual repression proposed for the lineage-driving transcription factors (Dietrich and Hiiragi, 2007; Plusa et al., 2008; Zernicka-Goetz et al., 2009). It appears that the mechanisms directing the formation of PE, TE and EPI are more sophisticated and enmeshed than anticipated. For instance, new evidence involves Nanog in a cross-regulation with Cdx2 (Chen et al., 2009). However, we did not observe Cdx2 deregulation in our mutants, thus this interaction may be subordinate in vivo. The role of Nanog in PE formation Since its first description Nanog has been thought to prevent PE differentiation, perhaps through interaction with a binding site in the
D.M. Messerschmidt, R. Kemler / Developmental Biology 344 (2010) 129–137
135
Fig. 4. (A) Scheme of blastocyst complementation experiments. (B) Section of an E7.5 embryo derived from ES cell-injection in a Nanog-deficient blastocyst. ES/EPI-derived tissues are LacZ-negative: extra-embryonic mesoderm (ExM), embryonic mesoderm (EmM) and embryonic ectoderm (EmE). Blastocyst-derived tissues are LacZ-positive: extra-embryonic ectoderm (ExE) and endoderm of visceral yolk sac (VE). (C–J) Embryos derived from GFP-labeled ES cell-injection in mutant (C) and control (G) blastocysts. (D, H) GFP-detection in embryos shown in (C) and (G), respectively. (E) IF detection of GFP (green) and β-GAL (red) on transversal sections through the hindbrain (HB) region from the embryo shown in (C, section plane is indicated) (DAPI, blue; composite image of E' (GFP) and E" (β-GAL)). (I–I") Similar analysis for the control embryo shown in (G, section plane is indicated). Tissues indicated are neural ectoderm (E) and mesoderm (M). (F, J) Sections of the visceral yolk sac of the embryos shown in (C) and (D) stained for GFP (green), β-GAL (red) and nuclei (DAPI, blue). (Scale bars: 100 μm (B, E, I), 1 mm (C, D, G, H), 25 μm (F, J).
Gata6 promoter as described in ES cells (Fig. 5A) (Singh et al., 2007). Conversely, our data suggest that Nanog does not impede PE differentiation in vivo, a fact also supported by the latest knock-out studies in ES cells and embryos (Chambers et al., 2007; Silva et al., 2009). But have Nanog-mutant ICM cells indeed lost the ability to differentiate into PE (Silva et al. 2009)? We found they do give rise to PE in blastocyst outgrowth experiments. However, in contrast to the in vivo situation where PE formation is achieved within hours, in vitro outgrowths require several days to attach and form the typical
mixture of TE, PE and EPI derivatives. This extended culture period might be beneficial to PE formation. We also found occasional Gata4 expression in E4.5 Nanog-null embryos, suggesting that some signal required for PE induction must still be present in the mutant embryo. It is therefore possible that TE-derived signals are beneficial to PE outgrowth when plating whole embryos, while in experiments using ICMs only others merely observed TE-like derivatives (Silva et al., 2009). Culture conditions, media composition and the time-point of analysis could further influence Gata4 expression/detection.
136
D.M. Messerschmidt, R. Kemler / Developmental Biology 344 (2010) 129–137
conclusion, our careful analysis and manipulation of Nanog-deficient embryos reveals a new level of complexity in the processes that determine early cell fates in the mouse embryo and contributes a novel feature to the complex picture of PE formation. Acknowledgments We thank Verdon Taylor for helpful discussions, Heike Wollmann for critical reading of the manuscript, Riana Vogt for technical assistance. We also thank Caroline Johner and Benoit Kanzler from the MPI-IB animal and transgenic facilities. This work was funded by the MaxPlanck Society. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.ydbio.2010.04.020. References
Fig. 5. (A) Classical model of lineage decisions in the mouse pre-implantation embryo. Mutually antagonistic acting lineage-specific transcription factors drive successive differentiation events (TP, totipotent precursor). (B) Nanog is required cell autonomously for EPI induction but not for PE repression. (C) EPI cells are required non-cell autonomously for PE formation, putatively by a combination of paracrine signaling and cell–cell interaction.
Compound interactions in the blastocyst The upstream mechanisms leading to the mosaic composition of the ICM are as unknown as the processes involved in lineage restriction and positional sorting of PE and EPI cells (reviewed in (Zernicka-Goetz et al., 2009). Gata4 expression is dependent on Grb2–Ras–Map kinase signaling and one potential upstream component in this pathway is FGF4 (Chazaud et al., 2006). FGF4 as well as FGF4-receptor deletion affects PE formation (Arman et al., 1998; Chai et al., 1998; Feldman et al., 1995). However, the role, source, and timing of FGF and other signaling cascades are only beginning to be understood. Remarkably the inhibition of FGF/Erk prevents Gata4 expression in the ICM of embryos in vitro, resulting in the uniformly high expression levels of OCT4 and NANOG in all ICM cells (Supplementary Fig. S3) (Nichols et al., 2009). Once PE fate is induced however, FGF/Erk inhibition does not affect Gata4 expression (Nichols et al., 2009). Our results are in complete agreement with these findings, furthermore, they disclose the Nanog expressing EPI as the primary source of the PE inducing signal. We hypothesize that the loss of PE is the consequence of a nonfunctional EPI, unable to provide the cues, e.g. a paracrine signal such as FGF/Erk, required for promoting PE differentiation in neighboring ICM cells. Thus, we suggest that Nanog promotes the EPI fate, but does not prevent PE differentiation, in a cell autonomous manner (Fig. 5B). During fate commitment, these future EPI cells instruct their surroundings to adopt a PE fate, if they are perceptive (Fig. 5C). The ability to receive PE differentiation signals may again depend of Nanog/Oct4 expression levels (Chambers et al., 2007). The blastocyst complementation experiments clearly support this theory, as ES cell-derived EPIs can rescue lethality and PE formation in a non-cell autonomous manner in Nanog-null blastocysts. Whereas the inhibition of FGF/Erk blocks PE formation we were not able to induce Gata4 expression in Nanog mutants in a reverse approach, by FGF-4 (maximal 150 ng/ml) treatment. However, recent work by Yamanaka et al. published during the preparation of this manuscript, showed that treatment of embryos with very high doses of FGF-4 (500 ng/ml) plus heparin is indeed able to override the EPI-induction program resulting in 100% PE cells in the ICM (Yamanaka et al. 2010). This new evidence further supports our hypothesis of an EPI-derived PE inductive or promoting signal. In
Arman, E., Haffner-Krausz, R., Chen, Y., Heath, J.K., Lonai, P., 1998. Targeted disruption of fibroblast growth factor (FGF) receptor 2 suggests a role for FGF signaling in pregastrulation mammalian development. Proc. Natl. Acad. Sci. U S A 95, 5082–5087. Beddington, R.S., Robertson, E.J., 1989. An assessment of the developmental potential of embryonic stem cells in the midgestation mouse embryo. Development 105, 733–737. Chai, N., Patel, Y., Jacobson, K., McMahon, J., McMahon, A., Rappolee, D.A., 1998. FGF is an essential regulator of the fifth cell division in preimplantation mouse embryos. Dev. Biol. 198, 105–115. Chambers, I., Colby, D., Robertson, M., Nichols, J., Lee, S., Tweedie, S., Smith, A., 2003. Functional expression cloning of Nanog, a pluripotency sustaining factor in embryonic stem cells. Cell 113, 643–655. Chambers, I., Silva, J., Colby, D., Nichols, J., Nijmeijer, B., Robertson, M., Vrana, J., Jones, K., Grotewold, L., Smith, A., 2007. Nanog safeguards pluripotency and mediates germline development. Nature 450, 1230–1234. Chazaud, C., Yamanaka, Y., Pawson, T., Rossant, J., 2006. Early lineage segregation between epiblast and primitive endoderm in mouse blastocysts through the Grb2– MAPK pathway. Dev. Cell 10, 615–624. Chen, L., Yabuuchi, A., Eminli, S., Takeuchi, A., Lu, C.W., Hochedlinger, K., Daley, G.Q., 2009. Cross-regulation of the Nanog and Cdx2 promoters. Cell Res. 19, 1052–1061. Copp, A.J., 1978. Interaction between inner cell mass and trophectoderm of the mouse blastocyst. I. A study of cellular proliferation. J. Embryol. Exp. Morphol. 48, 109–125. Dietrich, J.E., Hiiragi, T., 2007. Stochastic patterning in the mouse pre-implantation embryo. Development 134, 4219–4231. Feldman, B., Poueymirou, W., Papaioannou, V.E., DeChiara, T.M., Goldfarb, M., 1995. Requirement of FGF-4 for postimplantation mouse development. Science 267, 246–249. Fujikura, J., Yamato, E., Yonemura, S., Hosoda, K., Masui, S., Nakao, K., Miyazaki Ji, J., Niwa, H., 2002. Differentiation of embryonic stem cells is induced by GATA factors. Genes Dev. 16, 784–789. Gardner, R.L., Beddington, R.S., 1988. Multi-lineage ‘stem’ cells in the mammalian embryo. J. Cell Sci. Suppl. 10, 11–27. Hart, A.H., Hartley, L., Ibrahim, M., Robb, L., 2004. Identification, cloning and expression analysis of the pluripotency promoting Nanog genes in mouse and human. Dev. Dyn. 230, 187–198. Kan, N.G., Stemmler, M.P., Junghans, D., Kanzler, B., de Vries, W.N., Dominis, M., Kemler, R., 2007. Gene replacement reveals a specific role for E-cadherin in the formation of a functional trophectoderm. Development 134, 31–41. Koutsourakis, M., Langeveld, A., Patient, R., Beddington, R., Grosveld, F., 1999. The transcription factor GATA6 is essential for early extraembryonic development. Development 126, 723–732. Kwon, G.S., Viotti, M., Hadjantonakis, A.K., 2008. The endoderm of the mouse embryo arises by dynamic widespread intercalation of embryonic and extraembryonic lineages. Dev. Cell 15, 509–520. Meilhac, S.M., Adams, R.J., Morris, S.A., Danckaert, A., Le Garrec, J.F., Zernicka-Goetz, M., 2009. Active cell movements coupled to positional induction are involved in lineage segregation in the mouse blastocyst. Dev. Biol. 331, 210–221. Mitsui, K., Tokuzawa, Y., Itoh, H., Segawa, K., Murakami, M., Takahashi, K., Maruyama, M., Maeda, M., Yamanaka, S., 2003. The homeoprotein Nanog is required for maintenance of pluripotency in mouse epiblast and ES cells. Cell 113, 631–642. Morrisey, E.E., Tang, Z., Sigrist, K., Lu, M.M., Jiang, F., Ip, H.S., Parmacek, M.S., 1998. GATA6 regulates HNF4 and is required for differentiation of visceral endoderm in the mouse embryo. Genes Dev. 12, 3579–3590. Nagy, A., Gocza, E., Diaz, E.M., Prideaux, V.R., Ivanyi, E., Markkula, M., Rossant, J., 1990. Embryonic stem cells alone are able to support fetal development in the mouse. Development 110, 815–821. Nichols, J., Silva, J., Roode, M., Smith, A., 2009. Suppression of Erk signalling promotes ground state pluripotency in the mouse embryo. Development 136, 3215–3222.
D.M. Messerschmidt, R. Kemler / Developmental Biology 344 (2010) 129–137 Plusa, B., Piliszek, A., Frankenberg, S., Artus, J., Hadjantonakis, A.K., 2008. Distinct sequential cell behaviours direct primitive endoderm formation in the mouse blastocyst. Development 135, 3081–3091. Ralston, A., Rossant, J., 2008. Cdx2 acts downstream of cell polarization to cellautonomously promote trophectoderm fate in the early mouse embryo. Dev. Biol. 313, 614–629. Schnutgen, F., De-Zolt, S., Van Sloun, P., Hollatz, M., Floss, T., Hansen, J., Altschmied, J., Seisenberger, C., Ghyselinck, N.B., Ruiz, P., Chambon, P., Wurst, W., von Melchner, H., 2005. Genomewide production of multipurpose alleles for the functional analysis of the mouse genome. Proc. Natl. Acad. Sci. U S A 102, 7221–7226. Silva, J., Nichols, J., Theunissen, T.W., Guo, G., van Oosten, A.L., Barrandon, O., Wray, J., Yamanaka, S., Chambers, I., Smith, A., 2009. Nanog is the gateway to the pluripotent ground state. Cell 138, 722–737. Singh, A.M., Hamazaki, T., Hankowski, K.E., Terada, N., 2007. A heterogeneous expression pattern for Nanog in embryonic stem cells. Stem Cells 25, 2534–2542.
137
Soudais, C., Bielinska, M., Heikinheimo, M., MacArthur, C.A., Narita, N., Saffitz, J.E., Simon, M.C., Leiden, J.M., Wilson, D.B., 1995. Targeted mutagenesis of the transcription factor GATA-4 gene in mouse embryonic stem cells disrupts visceral endoderm differentiation in vitro. Development 121, 3877–3888. Stemmler, M.P., Hecht, A., Kemler, R., 2005. E-cadherin intron 2 contains cis-regulatory elements essential for gene expression. Development 132, 965–976. Yamanaka, Y., Lanner, F., Rossant, J., 2010. FGF signal-dependent segregation of primitive endoderm and epiblast in the mouse blastocyst. Development 137, 715–724. Zambrowicz, B.P., Imamoto, A., Fiering, S., Herzenberg, L.A., Kerr, W.G., Soriano, P., 1997. Disruption of overlapping transcripts in the ROSA beta geo 26 gene trap strain leads to widespread expression of beta-galactosidase in mouse embryos and hematopoietic cells. Proc. Natl. Acad. Sci. U S A 94, 3789–3794. Zernicka-Goetz, M., Morris, S.A., Bruce, A.W., 2009. Making a firm decision: multifaceted regulation of cell fate in the early mouse embryo. Nat. Rev. Genet. 10, 467–477.