The formation of proximal and distal definitive endoderm populations ...

2014. Published by The Company of Biologists Ltd | Journal of Cell Science (2014) 127, 2204–2216 doi:10.1242/jcs.134502

RESEARCH ARTICLE

The formation of proximal and distal definitive endoderm populations in culture requires p38 MAPK activity

ABSTRACT Endoderm formation in the mammal is a complex process with two lineages forming during the first weeks of development, the primitive (or extraembryonic) endoderm, which is specified in the blastocyst, and the definitive endoderm that forms later, at gastrulation, as one of the germ layers of the embryo proper. Fate mapping evidence suggests that the definitive endoderm arises as two waves, which potentially reflect two distinct cell populations. Early primitive ectoderm-like (EPL) cell differentiation has been used successfully to identify and characterise mechanisms regulating molecular gastrulation and lineage choice during differentiation. The roles of the p38 MAPK family in the formation of definitive endoderm were investigated using EPL cells and chemical inhibitors of p38 MAPK activity. These approaches define a role for p38 MAPK activity in the formation of the primitive streak and a second role in the formation of the definitive endoderm. Characterisation of the definitive endoderm populations formed from EPL cells demonstrates the formation of two distinct populations, defined by gene expression and ontogeny, that were analogous to the proximal and distal definitive endoderm populations of the embryo. Formation of the proximal definitive endoderm was found to require p38 MAPK activity and is correlated with molecular gastrulation, defined by the expression of brachyury (T). Distal definitive endoderm formation also requires p38 MAPK activity but can form when T expression is inhibited. Understanding lineage complexity will be a prerequisite for the generation of endoderm derivatives for commercial and clinical use. KEY WORDS: Embryonic stem cells, Endoderm, p38 MAP kinase, Gastrulation, BMP4

INTRODUCTION

Two distinct endoderm lineages arise during mammalian embryogenesis – the primitive endoderm, a derivative of the inner cell mass (ICM) of the blastocyst, which acts as a progenitor for the extraembryonic visceral and parietal endoderm, and the definitive endoderm, the progenitor of the embryonic endoderm populations. In mouse, a proportion of the definitive endoderm has been proposed to develop from a bipotent progenitor, the mesendoderm, which arises in the

1 Department of Zoology, University of Melbourne, Victoria, 3010, Australia. 2The Menzies Research Institute Tasmania, University of Tasmania, Tasmania, 7000, Australia. *These authors contributed equally to this work `

Author for correspondence ([email protected])

Received 2 May 2013; Accepted 4 January 2014

2204

primitive streak during gastrulation (Kinder et al., 2001; Lawson et al., 1991). The ability of embryonic stem (ES) cells to self-renew indefinitely and to give rise to all embryonic and adult tissues in response to the appropriate signals in vitro and in vivo make them attractive tools for modeling the developmental processes of gastrulation and formation of the definitive endoderm (Bradley et al., 1984; Doetschman et al., 1985; Evans and Kaufman, 1981; Martin, 1981). ES-cell-based approaches have been used to identify and characterise endoderm formation (Izumi et al., 2007; Kubo et al., 2004; Tada et al., 2005), and the addition of activin A has been shown to increase the formation of endoderm during ES cell differentiation (Kubo et al., 2004; Nostro et al., 2011). Furthermore, bipotent progenitors that differentiate into mesoderm and definitive endoderm have been identified in culture – a brachyury (T)-positive cell population (Kubo et al., 2004) and a cell population that is positive for goosecoid (GSC), E-cadherin (ECD, also known as CDH1) and PDGFRa that diverges to Gsc+ECD+PDGFRa2 and Gsc+ECD2PDGFRa+ populations (Tada et al., 2005). It is becoming clear, however, that definitive endoderm formation and subsequent differentiation are complex processes, and outcomes are influenced by induction strategies and positional specification (Gadue et al., 2009; Gadue et al., 2006; Jackson et al., 2010; Nostro et al., 2011). Analysing the mechanisms that regulate gastrulation and endoderm formation in systems that initiate differentiation from mouse ES cells is impacted on by two confounding factors. Initially, ES cells differentiate to form a later pluripotent population [a population that is equivalent to the embryonic primitive ectoderm (Rathjen et al., 2003a)] and primitive endoderm (Soudais et al., 1995). The primitive endoderm lineage is a potent source of signals that regulate pluripotent cell differentiation (Beddington and Robertson, 1998; Beddington and Robertson, 1999). Signals emanating from the primitive endoderm have the potential to synergise or compete with exogenous signals to influence the differentiation outcome. The maturation of ES cells to later pluripotent cell populations requires the generation of endogenous signals within the differentiation system (Li et al., 2004; Coucouvanis and Martin, 1995). Primitive ectoderm formation can be controlled in culture by the differentiation of ES cells to a second pluripotent cell population, early primitive ectoderm-like (EPL) cells, in response to MEDII, a medium conditioned by the human hepatocarcinoma cell line HepG2 (Lake et al., 2000; Rathjen et al., 1999; Tan et al., 2011; Washington et al., 2010). EPL cells are similar to cells of the primitive ectoderm. This has been shown using morphology, gene expression, cytokine response and differentiation potential (Lake et al., 2000; Rathjen et al., 2002; Tan et al., 2011; Washington et al., 2010). ES cell differentiation in response to MEDII is not accompanied by the concomitant formation of primitive endoderm (Rathjen et al., 2002; Vassilieva et al., 2012).

Journal of Cell Science

Charlotte Yap1,*, Hwee Ngee Goh1,*, Mary Familari1, Peter David Rathjen1,2 and Joy Rathjen1,2,`

Using EPL cells as the starting material for differentiation has resulted in a reliable model of gastrulation, which has been used to determine the role of growth factors, environmental manipulations and intracellular signalling in cell differentiation and to allow the identification of transient developmental intermediates (Harvey et al., 2010; Hughes et al., 2009a; Hughes et al., 2009b; Zheng et al., 2010). Here, the differentiation of EPL cells is used to investigate the control of molecular gastrulation, with a focus on the formation of definitive endoderm. Cells formed as a result of molecular gastrulation are referred to here as primitive streak intermediates; this term covers all cells expressing T, a marker of molecular gastrulation, and includes, but is not limited to, bipotent mesendoderm. p38 MAPK is a member of the mitogen-activated protein kinase (MAPK) family of kinases (Martı´n-Blanco, 2000; Nebreda and Porras, 2000), and was originally identified through its involvement in stress and inflammatory responses (Han et al., 1994; Raingeaud et al., 1995; Rouse et al., 1994). Developmental roles for p38 MAPK activity have been shown in the formation of cardiocytes (Aouadi et al., 2006), myocytes (Perdiguero et al., 2007), adipocytes (Engelman et al., 1998), chondrocytes (Nakamura et al., 1999), erythroid cells (Nagata et al., 1998) and neurons (Nebreda and Porras, 2000). Inhibiting p38 MAPK during ES cell differentiation promotes neurogenesis at the expense of cardiogenesis and suggests a role for p38 MAPK in germ layer specification (Aouadi et al., 2006; Barruet et al., 2011; Wu et al., 2010) and cardiogenesis (Wang et al., 2012). In this study, roles for p38 MAPK activity in molecular gastrulation and in definitive endoderm formation are identified. The inhibition of p38 MAPK during EPL cell differentiation in response to serum reduced the expression of T and promoted the formation of neural lineages. By contrast, when cells were differentiated in response to BMP4 the inhibition of p38 MAPK did not alter differentiation outcomes or the expression of differentiation markers. The analysis of the differentiation outcomes from cells differentiated in the presence of activin A, BMP4 or serum and the p38 MAPK inhibitor showed that the formation of definitive endoderm from EPL cells was dependent on p38 MAPK activity. Further characterisation suggested that EPL-cell-derived definitive endoderm comprised two distinct populations, representative of the proximal and distal definitive endoderm of the embryo, which formed in response to alternative signalling environments and potentially from distinct progenitor populations. RESULTS BMP4 and serum induce the formation of primitive streak intermediates through distinct signalling pathways

Primitive streak intermediates can be induced from EPL cells by BMP4 (Harvey et al., 2010; Zheng et al., 2010) or serum (Hughes et al., 2009b). The requirement for p38 MAPK activity in the formation of primitive streak intermediates in response to these inducers was investigated. Phosphorylated p38 MAPK (p-p38 MAPK) and phosphorylated heat shock protein 27 (pHsp27), a downstream target of p38 MAPK signalling, were detected by western blot in EPL cells incubated in serum-free medium (SFM) with or without BMP4 or serum (p-p38 MAPK only) (Fig. 1Ai,Aii). No consistent increase in p-p38 MAPK was seen in EPL cells after the addition of serum. p38 MAPK activity was inhibited pharmacologically with SB203580 [4-(49-fluorophenyl)2-(49-methylsulfinylphenyl)-5-(49-pyridyl)-imidazole; SB]. This chemical inhibits p38a (MAPK14) and p38b (MAPK11) homologues by competing for ATP-binding pockets (Cuenda

Journal of Cell Science (2014) 127, 2204–2216 doi:10.1242/jcs.134502

et al., 1995). The levels of p-p38 MAPK and pHsp27 were reduced in cells exposed to serum and SB as compared with cells exposed to serum alone (Fig. 1Aii; supplementary material Fig. S1A,B). In the absence of serum or BMP4, aggregates of EPL cells almost exclusively formed neurons (Fig. 1B; data not shown) (Zheng et al., 2010). The addition of BMP4, in BMP4-containing medium (BCM; supplementary material Table S1), or serum, in serum-containing medium (SCM; supplementary material Table S1), during differentiation resulted in the formation of cardiocytes and erthryocytes within the aggregates, as expected from previous analyses of differentiation (Fig. 1B) (Harvey et al., 2010; Zheng et al., 2010). The addition of SB to serum reduced the percentage of aggregates that formed erthryocytes and increased the percentage of aggregates that formed neurons, whereas cardiocyte formation was unaffected (Fig. 1B). In the presence of BMP4, however, there was no significant effect of the inhibitor on the production of erythrocytes, neurons or cardiocytes. SB did not affect the ability of aggregates to adhere to plasticware for scoring or the survival of mesoderm or ectoderm lineages scored in the assay (supplementary material Table S3). These data suggest a role for p38 MAPK in differentiation. Previous analysis of EPL cell differentiation has suggested that the frequency of blood, cardiocyte and neuron formation reflects the efficiency of the formation of primitive streak intermediates (Fig. 1C; supplementary material Fig. S1B) (Zheng et al., 2010). Differentiating EPL cells were analysed for the expression of established markers of the primitive streak (T, Bmp4, Tgfb1, Wnt3 and Fgf8) (Crossley and Martin, 1995; Dickson et al., 1995; Liu et al., 1999; Wilkinson et al., 1990; Winnier et al., 1995), ectoderm (Sox1 and Ascl1) (Guillemot et al., 1993; Pevny et al., 1998) and mesoderm (Mesp1, Hbb-b1, Nkx2-5 and Osr1) (Farace et al., 1984; Lints et al., 1993; Saga et al., 1996; So and Danielian, 1999). Analysis was performed when gene expression was most reproducibly detected (at day two for T, Wnt3 and Fgf8 and at day four for Bmp4 and Tgfb1). The expression of marker genes of the primitive streak and mesoderm was reduced in EPL cells that were differentiated in SCM+SB compared to the expression in controls, consistent with the reduced formation of mesoderm from these cells (Fig. 1C,D; supplementary material Fig. S2D). By contrast, only Bmp4 expression was decreased in EPL cells that were differentiated in BCM+SB (Fig. 1C). Lower expression of T, Wnt3 and Bmp4 was detected in cells that were differentiated in response to SCM compared with those differentiated in BCM (supplementary material Fig. S2E), suggesting that differentiation in response to these inducers was not equivalent. The significantly reduced expression of markers of the primitive streak intermediates and reduced formation of erythrocytes from EPL cells differentiated in serum when SB was added suggested that there is a role for p38 MAPK in the formation of the primitive streak intermediate. Cardiocyte formation and the residual levels of erythrocyte formation in a proportion of the aggregates that were differentiated in SCM+SB demonstrated the formation of a population of primitive streak intermediates, albeit a reduced one, and a discord between the gene expression and differentiation data. Most notably, the percentage of aggregates containing cardiocytes was unaffected by the inhibitor SB but the expression of Nkx2-5, a cardiocyte marker, was undetectable, suggesting that SB did affect the establishment of cardiocytes. Differentiation assays score the presence, but not the abundance, of a cell type in an aggregate. Potentially, differentiation within 2205


RESEARCH ARTICLE

RESEARCH ARTICLE


aggregates cultured in serum and SB was reduced, delayed and asynchronous, leading to a reduction in Nkx2-5 transcript levels at any given timepoint and affecting the ability of the analysis to detect expression. Collectively, these data suggest that the induction of primitivestreak intermediates by serum, but not by BMP4, requires p38 MAPK activity. The suppression of differentiation that resulted from the addition of the p38 MAPK inhibitor SB to serum-containing medium was specific to the formation of the primitive-streak intermediate. Neurectoderm was formed and the prevalence of the lineage was increased in these conditions. The formation of primitive streak intermediates in response to serum is reduced, but not abolished, by BMP4 inhibition

BMP4 and serum potentially induce primitive streak intermediates by independent pathways. Alternatively, serum might contain BMP activity at levels that are sufficient to induce primitive-streak intermediates in cells with p38 MAPK activity or it might induce BMP expression and signalling during cell differentiation. Some sera have been shown to contain BMP activity (Herrera and Inman, 2009; Kodaira et al., 2006). Western blot analysis showed that the levels of phosphorylated Smad1, 5 and 8 2206

(pSmad1/5/8) were not increased in EPL cells that were exposed to the serum used in this analysis (Fig. 2A). By contrast, pSmad1/ 5/8 was markedly induced in cells that were exposed to BMP4. These data indicate that our serum contained little or no BMP activity. The possibility that BMP4 expression was induced in EPL cells as they differentiated in response to serum, and that endogenously expressed BMP4 subsequently acted to induce the primitive streak intermediate, was investigated by differentiating cells in BCM or SCM in the presence of the BMP inhibitor noggin. Noggin binds to BMP4, BMP2 and, to a lesser extent, BMP7 proteins, and inhibits their interaction with receptors (Zimmerman et al., 1996). The inhibition of BMP4 signalling in BCM will abrogate signalling from endogenous BMP activity and exogenous BMP4. In BCM+noggin a higher percentage of aggregates contained neurons when compared with controls, and effectively no aggregates contained cardiocytes and erythrocytes (Fig. 2B). Consistent with this, the expression of primitive-streak markers was decreased in cells that were differentiated in BCM+noggin compared with their expression in BCM controls (Fig. 2C,D). A significant decrease in the percentage of aggregates containing cardiocytes and erythrocytes was also seen in aggregates that were cultured in SCM+noggin,


Fig. 1. Inhibition of p38 MAPK signalling affects lineage choice in EPL cells. (Ai) Western blot of p-p38 MAPK and p38 MAPK in EPL cells incubated in SFM, SFM+serum (Serum), or SFM+BMP4 (BMP4) for 10, 30 or 60 min. b-tubulin was used as a loading control. The appearance of increased levels of p-p38 MAPK with the addition of serum was variable. n53. (Aii) Western blot of pHsp27 and Hsp27 in serum-starved EPL cells that were transferred to SFM, SFM+serum and DMSO (Serum) or SFM+serum and SB for 15, 30 or 60 min. b-tubulin was used as a loading control. n53. (B) EPL cells differentiated in SFM+SB, or BCM or SCM with either SB or DMSO were scored for the formation of cardiocytes, erythrocytes and neurons. Raw data for this experiment can be found in supplementary material Table S3. Results are means6s.e.m. n53. (C,D). Primitive streak (T, Bmp4, Tgfb1, Wnt3 and Fgf8), ectoderm (Sox1 and Ascl1) and mesoderm (Mesp1, Hbb-b1, Nkx2-5 and Osr1) markers were detected by RT-PCR in EPL cells differentiating in BCM or SCM with DMSO or SB. (C) Primitive-streak markers were quantified on day three and have been normalised to the expression of Gapdh. Results are means6s.e.m., n53. (D) Mesoderm and ectoderm markers were analysed on day four and a representative result is shown. n53. 2R, no reverse transcriptase control; 2D, no cDNA control. *P,0.05, #P,0.01.

RESEARCH ARTICLE


SCM controls (Fig. 2C,D). Mesoderm formation in cells that were differentiated in SCM+noggin was reduced but not abolished (Fig. 2B), suggesting that two pathways mediate the induction of the primitive streak intermediate in these aggregates, a pathway that is dependent on endogenously generated BMP activity and a second pathway that is independent of BMP signalling.

Fig. 2. Inhibition of BMP signalling affects lineage choice in differentiating EPL cells. (A) Western blot of pSmad1/5/8 in EPL cells treated with SFM or SFM containing BMP4 or serum for 5, 10 or 30 min. btubulin was used as a loading control. A representative result is shown. n52. (B) EPL cell aggregates differentiated in BCM or SCM with or without noggin were scored for the formation of cardiocytes, erythrocytes and neurons. The data represent the mean6s.e.m. n53. Raw data for this experiment can be found in supplementary material Table S4. (C,D) The expression of primitive-streak markers in EPL cells differentiated for 2 (C) and 4 (D) days as in B. The data have been normalised to b-actin transcript levels. Results are means6s.e.m. n53. *P,0.05, #P,0.01.

although ,20% of aggregates still formed mesoderm-derived lineages (Fig. 2B). Similarly, this reduction in mesoderm formation was reflected in a reduction in the expression of markers of primitive streak intermediates in cells that were differentiated in SCM+noggin when compared with

Mixl1, which has been implicated in the formation of the definitive endoderm (Robb et al., 2000; Hart et al., 2002), was found to be expressed in cells that were differentiated in BCM and SCM, and was significantly decreased in both conditions on the addition of the p38 MAPK inhibitor SB (Fig. 3A), raising the possibility that p38 MAPK was involved in the formation of the definitive endoderm. The expression of additional endoderm markers Sox17, Ttr, Gata4, Trh and Eya2 (Gu et al., 2004; KanaiAzuma et al., 2002; Kwon et al., 2008; Lickert et al., 2002; McKnight et al., 2007) in cells that were differentiated in BCM and SCM was examined. A novel endoderm marker, serine protease inhibitor Kazal type 3 (Spink3), was included in the analysis. Previously, Spink3 has been shown to be expressed in endoderm-derived populations, including cells of the gut and pancreas in embryonic day (E)9.5 mouse embryos (Wang et al., 2008). As shown here, Spink3 expression was detected in a band of definitive endoderm that was located immediately below the embryonic and extraembryonic boundary of gastrulating E7.5 embryos (Fig. 3B). The expression of endoderm-marker genes was detected in cells that were differentiated in BCM or SCM but higher levels of expression were generally detected in cells differentiated in SCM (Fig. 3C). Whole-mount in situ hybridisation (WISH) detected endoderm markers on the surface of aggregates that were differentiated in BCM or SCM, but more Ttr+ and Trh+ cells were seen on aggregates differentiated in SCM when compared with those in BCM (Fig. 3D). Inhibition of p38 MAPK led to a reduction in the expression of most endoderm markers in EPL cells that were differentiated in SCM, but only a subset of markers (Spink3 and Ttr) in EPL cells differentiated in BCM (Fig. 3C). The addition of SB resulted in similar expression levels of all endoderm markers, with the exception of Gata4, regardless of the differentiation induction strategy used. The sustained expression of Gata4 in the BCM+SB condition might reflect the expression of the gene in another lineage. Some of the endoderm markers used here (Sox17 and Ttr) also mark visceral endoderm (Kanai-Azuma et al., 2002; Kwon et al., 2008). Visceral endoderm can be distinguished from definitive endoderm and parietal endoderm by the ability to endocytose horseradish peroxidase (HRP) from the surrounding medium – internalised HRP can be detected colorimetrically (Kanai-Azuma et al., 2002; Vassilieva et al., 2012). Embryoid bodies (EBs), which contain visceral endoderm (Kubo et al., 2004; Vassilieva et al., 2012), developed areas of brown staining on their surface (Fig. 3E), indicating the presence of visceral endoderm. By contrast, EPL cells that were differentiated in SFM or SCM formed few cells that were capable of taking up HRP from the medium, demonstrating that little or no visceral endoderm was formed in these conditions. The role of p38 MAPK in endoderm formation was investigated further using a second inhibitor, SB202190 (Lee et al., 1994), which acts by binding the ATP-binding site of p38 2207


The formation of definitive endoderm in response to BMP4 or serum is impaired by the inhibition of p38 MAPK

RESEARCH ARTICLE


Fig. 3. Inhibition of p38 MAPK signalling reduces the expression of endoderm marker genes in differentiating EPL cells. (A) Mixl1 expression in EPL cells differentiated in BCM or SCM, with or without SB or DMSO for 2 days. Expression has been normalised to bactin and is shown relative to the expression in SCM. The data represent the mean6s.e.m. n53. (B) WISH analysis of Spink3 expression in E7.5 mouse embryos. Panels i and iii show lateral views, panel ii is an anterior view. The asterisk (*) marks the anterior embryonic-extraembryonic boundary. Panel iv shows a cross section of an E7.5 Spink3-stained mouse embryo. The embryo is ,500 mm from anterior to posterior. (C) EPL cells differentiated in BCM or SCM with or without SB for 4 days were analysed for the expression of endoderm markers by qPCR. Expression was normalised to the expression of the b-actin gene. Results are means6s.e.m. n53–7. (D) WISH analysis of Ttr or Trh expression in EPL cells differentiated in BCM or SCM. Aggregates are ,250 mm in diameter. (E) EPL cells differentiated in SCM and SFM were incubated with HRP. Day seven EBs were used as a positive control. Cells that took up HRP stained brown (arrowheads). *P,0.05, **P,0.01.

2208

Inhibition of BMP signalling during differentiation promotes the formation of a definitive endoderm population that expresses a subset of markers

The formation of endoderm in the embryo has been suggested to proceed through the formation of a bipotent progenitor referred to as mesendoderm (Kinder et al., 2001; Lawson et al., 1991); mesendoderm arises in the primitive streak and is encompassed within the primitive-streak-intermediate population. Mesendoderm formation could underpin endoderm formation in SCM and BCM, with BMP4, or other factors, inducing a primitive-streak intermediate with the properties of mesendoderm and a p38-MAPKdependent mechanism inducing an endoderm fate from this progenitor on further differentiation. The inhibition of BMP4 signalling in SCM by noggin did not affect the expression of the endoderm markers (Fig. 4A), suggesting that cells cultured in SCM do not rely on endogenously generated BMP4 to form mesendoderm. In EPL cells differentiated in BCM+noggin, analysis of the expression of marker genes and differentiation outcomes showed that formation of the primitive-streak intermediate was


MAPKs (Young et al., 1997). EPL cells that were differentiated in BCM or SCM supplemented with SB202190 showed reduced expression of endoderm markers (supplementary material Fig. S3A). Treatment with SB202190, however, also reduced the expression of markers of the primitive streak intermediate in cells that were differentiated in SCM or BCM (supplementary material Fig. S3B), suggesting that, in comparison to SB203580, this compound inhibited p38 MAPK and additional pathways that were required for molecular gastrulation. These data are consistent with a role for p38 MAPK activity in the formation of definitive endoderm but suggest heterogeneity in the endoderm outcomes from EPL cells that are differentiated in serum and BMP4. The documentation of endoderm formation in response to BMP4 without the addition of activin A is unprecedented but perhaps not unexpected, given that mesoderm and endoderm can form from a common progenitor in culture (Kubo et al., 2004; Tada et al., 2005) and that BMP4 can function in conjunction with activin A or other growth factors to induce definitive endoderm (Mathew et al., 2012; Phillips et al., 2007).

RESEARCH ARTICLE


S4A), suggesting the formation of endoderm by cells cultured in BCM+noggin. The presence of endoderm on the surface of aggregates differentiated in BCM or BCM+noggin was confirmed morphologically and by localisation of Trh expression (Fig. 4B,C). Cell aggregates that were cultured for an extended period expressed markers of endoderm derivatives including Afp and Ttr (gut and liver), Fabp2 (intestine) and Pdx1 (gut and pancreas) (supplementary material Fig. S4B). In EPL cells differentiated in BCM+noggin and SB, the expression of Trh and Eya2 was decreased by 10% and 40%, respectively, when compared with cells differentiated in BCM+noggin (supplementary material Fig. S4C), suggesting a requirement for endogenous p38 MAPK in Trh and Eya2 expression. The co-regulation of Spink3, Ttr and Gata4 is distinct from the co-regulation of Trh and Eya2, and this could arise from the formation of two endoderm populations during EPL-cell differentiation. Spink3, Ttr and Gata4 potentially mark endoderm that is produced in response to both BMP4 and serum but is reduced during differentiation in BCM when noggin is present. A second population that expresses Trh and Eya2 can be hypothesised, the formation of which is maintained in the absence of BMP4. In situ hybridisation was used to confirm the generation of two genetically distinct populations of endoderm during differentiation (Fig. 4D). Double staining with probes against Spink3 and Trh detected distinct populations of cells on the surface of aggregates that expressed either Spink3 (blue) or Trh (magenta).

Fig. 4. Inhibition of BMP signalling affects endoderm formation in differentiating EPL cells. (A) EPL cells that were differentiated in BCM or SCM with or without noggin or DMSO for 4 days were analysed by qPCR for the expression of endoderm-marker genes. The data have been normalised to b-actin transcript levels. Results are means6s.e.m. n53 or 4. (B,C) EPL cells differentiated in BCM (B) and BCM+noggin (C) were sectioned and stained with haematoxylin and eosin for morphology (i) or for the expression of Trh (ii). Arrowheads indicate squamous endoderm-like cells on the surface of the aggregates. Scale bars: 50 mm. (D) EPL cells differentiated in SCM were analysed by double WISH for the expression of Spink3 (blue, arrowhead) and Trh (magenta, arrow). *P,0.05, **P,0.01. Scale bar: 100 mm.

significantly reduced (Fig. 2C,D), the formation of mesoderm was effectively ablated (Fig. 2B) and the expression of the endoderm markers Spink3, Ttr and Gata4 decreased (Fig. 4A). This is consistent with a two-step process that is reliant on the initial formation of a primitive-streak intermediate in response to BMP4. Trh and Eya2 expression, however, was increased to levels equivalent to those in cells differentiated in SCM when compared with control (Fig. 4A; supplementary material Fig.

The induction of Spink3+, Ttr+ and Gata4+ endoderm in response to BMP4 is a two-step process proceeding through a bipotent primitive streak intermediate, or mesendoderm. The likely role for BMP4 in this process is indirect, by way of the induction of mesendoderm from EPL cells (Harvey et al., 2010). The possibility exists, however, that BMP4 has a role in the subsequent induction of endoderm from mesendoderm. Activin A can induce primitive streak intermediates independently of BMP4 signalling from ES cells and has been identified as a potent inducer of endoderm lineages in culture (Izumi et al., 2007; Jackson et al., 2010; Kubo et al., 2004; Tada et al., 2005). As expected, EPL cells that were differentiated in activin-A-containing medium (ACM) or ACM+noggin expressed markers of the primitive streak (Fig. 5A). In cells differentiated in ACM+noggin+SB, the expression of primitive streak markers (except Bmp4) was reduced and the expression of the neural marker Sox1 was increased (Fig. 5A), suggesting that the ability of activin A to induce primitive streak intermediates required p38 MAPK. Western blot showed p-p38 MAPK in cells that had been treated with ACM (Fig. 5B). The regulation of the formation of primitive streak intermediates in response to activin A, therefore, is distinct from that observed in response to BMP4 or serum. In EPL cells differentiated in ACM, SFM+noggin or ACM+noggin, Sox17, Spink3 and Ttr were expressed equivalently (Fig. 5C). The expression of these genes in the absence of BMP4 signalling suggests that BMP4 does not play additional roles in the specification of endoderm from mesendoderm. Eya2 expression was higher in EPL cells differentiated in ACM+noggin when compared with cells differentiated in ACM. The expression of all endoderm markers 2209


BMP4 regulates the formation of endoderm expressing Spink3, Ttr and Gata4 through induction of the primitive streak intermediate

RESEARCH ARTICLE


in cells that were cultured in ACM+noggin suggests that activin A, like serum, is able to induce both of the proposed endoderm populations from EPL cells (Fig. 4A). The increased expression of Eya2 suggests that the formation of endoderm that expresses these markers can be enhanced by BMP4 inhibition. The expression of Sox17, Spink3, Ttr and Eya2 (Fig. 5C) was decreased in cells differentiated in ACM+noggon+SB, consistent with a role for p38 MAPK activity in the activin-Ainduced formation of endoderm.

increased, when compared with cells formed in BCM (Fig. 5D). These data are consistent with formation of Trh+ and Eya2+ endoderm in the absence of primitive-streak intermediate formation and with a requirement for the initial formation of primitive-streak intermediates in the formation of Spink3+ and Ttr+ endoderm. Further differentiation of aggregates cultured in BCM+DAPT showed the expression of markers of later endoderm populations, consistent with the formation of an endoderm progenitor (supplementary material Fig. S4B).

The primitive streak intermediate is the progenitor for endoderm that expresses Spink3, Ttr and Gata4 but not necessarily for endoderm that expresses Trh and Eya2

DISCUSSION p38 MAPK and the formation of primitive-streak intermediates

The formation of endoderm that expresses Spink3, Ttr and Gata4 appears to be dependent on the prior formation of primitive streak intermediates. By contrast, the expression of Trh and Eya2 in EPL cells differentiated in BCM+noggin suggests the formation of an endoderm that does not rely on the prior formation of this population. The formation of primitive-streak intermediates from differentiating EPL cells can be inhibited by DAPT {N-[N-(3,5-difluorophenacetyl-L-alanyl)]-S-phenylglycine t-butyl ester}, an antagonist of c-secretase (Hughes et al., 2009a). In cells formed from EPL cells differentiated in BCM+DAPT, the expression of Spink3 and Ttr was decreased, and Eya2 expression

The inhibition of p38 MAPK during EPL-cell differentiation disrupts the formation of primitive-streak intermediates from EPL cells in response to activin A or serum. Others have shown that p38 MAPK inhibition during ES cell differentiation in serum promotes neurogenesis at the expense of cardiogenesis (Aouadi et al., 2006; Barruet et al., 2011; Wu et al., 2010), and that there is a requirement for p38 MAPK activity early in cell differentiation (Barruet et al., 2011; Davidson and Morange, 2000; Duval et al., 2004; Wu et al., 2010). The inhibition of p38 MAPK activity did not affect the ability of BMP4 to induce primitive streak intermediates or mesoderm derivatives, suggesting that this

2210


Fig. 5. The formation of primitive-streak intermediates and endoderm in response to activin A signalling requires p38 MAPK activity. (A) EPL cells that were differentiated in ACM, ACM+noggin with or without SB or DMSO or in SFM+noggin for 2 or 4 days were analysed by RT-PCR for the expression of markers of primitive streak intermediates (day two and four) or ectoderm (day four). 2R, no reverse transcriptase control; 2D, no cDNA control. A representative result is shown. n53. (B) Serum-starved aggregates were transferred to SFM or SFM containing 25 ng/ml activin A. Aggregates were collected 15, 30 and 60 min after transfer and were analysed by western blot for the phosphorylation of p38 MAPK. (C) EPL cells differentiated in ACM, ACM+noggin with or without SB or DMSO or in SFM+noggin for 4 days were analysed by qPCR for the expression of endoderm markers. The data have been normalised to b-actin transcript levels. Results are means6s.e.m. n53. (D) The expression of endoderm markers in EPL cells that were differentiated in BCM+DMSO or DAPT on day four of treatment. The data have been normalised to b-actin transcript levels. Results are means6s.e.m. n53. *P,0.05, **P,0.01.

pathway is not essential for the formation of primitive streak intermediates per se. The data presented here are consistent with a role for p38 MAPK in lineage allocation or specification during differentiation, specifically in the formation of primitive streak intermediates, but only when molecular gastrulation is induced by serum or activin A. Moreover, the inability of cells cultured in SFM to form primitive streak intermediates, despite intracellular p-p38 MAPK, suggests that p38 MAPK activity is not sufficient for differentiation but works in conjunction with other pathways. In cells differentiated in SCM, but not ACM, one role for p38 MAPK appears to be the upregulation of Bmp4, which in turn initiates differentiation. The increased expression of Bmp4 in cells differentiated in SCM, and the significant reduction in its expression in cells differentiated in BCM supplemented with SB, is consistent with p38 MAPK acting upstream of Bmp4. Endogenously produced BMP4 acts in turn to induce the primitive streak intermediate, an activity that is blocked by noggin. Noggin did not completely suppress molecular gastrulation in response to serum, suggesting the presence of additional, potentially p38 MAPK dependent, pathways. By contrast, markers of molecular gastrulation were expressed robustly in cells differentiated in activin-A-containing medium supplemented with noggin, suggesting that activin A activity was independent of BMP. This is consistent with previous reports that activin A does not induce the expression of Bmp4 during ES cell differentiation (Jackson et al., 2010). The involvement of p38 MAPK in lineage specification from ES cells has proven difficult to demonstrate, with conflicting reports on the need for p38 MAPK activity during mesoderm specification. The difficulty in resolving a role for p38 MAPK is most likely a consequence of the complexity of the molecular mechanisms that regulate gastrulation coupled with the use of illdefined and poorly understood culture reagents. The variability of outcomes elicited in cells cultured in the presence of p38 MAPK inhibitors (Aouadi et al., 2006; Barruet et al., 2011; Wu et al., 2010) or from Mapk142/2 cells (also known as p38a2/2 cells) (Allen et al., 2000; Chakraborty et al., 2009; Guo et al., 2007; Aouadi et al., 2006) could be attributed to the confounding use of serum in these experiments. Some sera have been reported to contain exogenous BMP activity (Herrera and Inman, 2009; Kodaira et al., 2006). BMP activity within sera would be able to specify primitive streak intermediates and mesoderm lineages in the absence of p38 MAPK activity. Our interpretation of the respective roles of serum and growth factors suggests that serum variability is a contributing factor to variability in the analysis of molecular gastrulation, and in the analysis of the role of p38 MAPK specifically. The ability of p38 MAPK inhibition to promote cardiocyte formation from human ES cells (Graichen et al., 2008) contradicts the findings from mouse pluripotent cells reported here and by others (Barruet et al., 2011; Aouadi et al., 2006; Wu et al., 2010). The differentiation of human ES cells was induced in cell aggregates by a conditioned medium in which the signalling activity was largely uncharacterised. Potentially, increased cardiocyte formation resulted from an increase in the number of primitive streak intermediates (formed in response to BMP or to similar signalling activity within the conditioned medium) adopting a mesoderm fate. This would occur when p38 MAPK was inhibited in these cells, preventing their differentiation to the endoderm. p38 MAPK comprises a, b, c and d isoforms that are encoded separately. Of these, p38a and p38b are expressed in the germ layers and primitive ectoderm, respectively, at gastrulation (Zohn


et al., 2006). The disruption of the gene encoding p38a resulted in placental defects and embryonic death mid-gestation (Adams et al., 2000; Mudgett et al., 2000). Double-mutant embryos, lacking p38a and p38b in embryonic tissues, gastrulate but fail mid-gestation with diverse developmental defects (del Barco Barrantes et al., 2011). Embryos with mutations in the MKK3, MKK4 and MKK6 kinases that have been shown to activate p38 MAPK also survive beyond gastrulation (Lu et al., 1999; Tanaka et al., 2002; Yang et al., 1997). Mice that are deficient in p38interacting protein (p38IP, also known as SUPT20H) show a loss of p38 MAPK phosphorylation in the primitive streak, and the cells that lack p-p38 MAPK fail to migrate (Zohn et al., 2006). This mutation did not, however, prevent the formation of primitive-streak intermediates or mesoderm specification, but raises the possibility that p38 MAPK has a role in the regulation of an epithelial-to-mesenchymal transition during differentiation. It is unlikely, therefore, that p38 MAPK is essential for the formation of primitive streak intermediates in vivo, but its role is revealed during in vitro differentiation in response to serum or activin A. A novel role for p38 MAPK in the formation of definitiveendoderm populations

Endoderm formation in the mammal is complex, with the formation of two endoderm lineages from the pluripotent lineage during early development – primitive or visceral endoderm and definitive endoderm. Analysis of EPL cell differentiation suggests that the definitive endoderm might form as two distinct cell populations that can be distinguished by gene expression and ontogeny. The formation of both lineages required p38 MAPK activity. p38 MAPK inhibition during the differentiation of EPL cells in response to BMP4 resulted in the reduced expression of Mixl1, Spink3 and Ttr. When p38 MAPK was inhibited in cells that differentiated in response to serum there was reduced expression of Mixl1, Sox17, Spink3, Ttr, Gata4 and Eya2. These results suggest a previously unidentified role for p38 MAPK in the formation of definitive endoderm. In the mouse, definitive endoderm formation is dependent on signalling through nodal (Tremblay et al., 2000; Vincent et al., 2003), a member of the TGFb family with similar signalling properties to activin A (Conlon et al., 1994; Vincent et al., 2003). Similarly, ES cell differentiation induced by activin A results in the enrichment of definitive endoderm (Gadue et al., 2006; Kubo et al., 2004; Nostro et al., 2011). Canonically, nodal signalling is mediated by Smad2/3 and the downstream effectors of Smads (Heldin et al., 1997; Massaous and Hata, 1997). p38 MAPK has been shown to be activated by TGFbs and to mediate signalling in response to these factors (Hanafusa et al., 1999; Hu et al., 2004; Yue and Mulder, 2000). The inhibition of p38 MAPK impaired the induction of definitive endoderm markers by activin A, suggesting that activin A signalling was mediated, in part, by p38 MAPK. A role for p38 MAPK has been shown in the positional specification of the visceral endoderm in response to nodal (Clements et al., 2011) and a similar requirement may exist for p38 MAPK in the induction of definitive endoderm formation in response to Nodal. The differential effects of the inhibitor SB on the expression of endoderm markers in aggregates that were differentiated in SCM or BCM infers that the specification of endoderm from pluripotent cells can occur through multiple pathways and results in two populations. The expression of Spink3, Ttr and Gata4 2211


RESEARCH ARTICLE

marked one of these populations. Ttr has been used previously to mark visceral endoderm (Kwon et al, 2008), but the widespread expression of Ttr in EBs derived from EPL cells, in which visceral endoderm is rarely formed (Vassilieva et al., 2012), and the reliance of Ttr expression on molecular gastrulation by EPL cells suggests that Ttr can also be expressed in an endoderm formed during molecular gastrulation. The formation of a Spink3+, Ttr+ and Gata4+ endoderm population from differentiating EPL cells correlated with the prior expression of primitive streak markers. When the formation of primitive streak intermediates was inhibited, as happened when cells were differentiated in BCM supplemented with noggin or DAPT, the expression of Spink3, Ttr and Gata4 was reduced. The second endoderm population was marked by the expression of Trh and Eya2. The expression of these markers is maintained in cells differentiated in BCM supplemented with noggin or DAPT, suggesting that Trh+ and Eya2+ endoderm can form in the absence of a T-expressing primitive-streak intermediate. The differential expression of Spink3, Ttr and Gata4 between conditions that enriched or suppressed the formation of the primitive-streak intermediate, coupled with the persistence of Trh+ and Eya2+ endoderm when the formation of the primitive streak intermediate was suppressed, is consistent with the formation of two endoderm populations during EPL cell differentiation. A model for the regulation of molecular gastrulation

Based on this analysis of the formation of the primitive-streak intermediate and endoderm from EPL cells, we propose a revised paradigm for molecular gastrulation (Fig. 6). This model proposes that primitive streak intermediates, which express T and other primitive streak markers, can be induced from EPL cells through multiple pathways, including pathways dependent on BMP4 signalling, growth factors/cytokines, including activin A and the active components of serum, which require p38 MAPK signalling and, as has been reported by others, WNT signalling (Tanaka et al., 2009). These pathways most likely generate distinct primitive streak intermediates that are distinguished by


divergent differentiation potential, as has been shown for BMP4 and Wnt (Tanaka et al., 2009). We propose that the primitive streak intermediate induced by BMP4 can differentiate to form mesoderm and a population of endoderm expressing Spink3, Ttr and Gata4. In gene expression, this population is similar to the ring of endoderm marked by Spink3 in the proximal region of the embryo. Formation of this endoderm population requires p38 MAPK activity in the primitive streak intermediate; activation of this pathway is achieved through endogenous signalling in aggregates that are differentiated in the presence of BMP4. Alternatively, EPL cells can differentiate into a Trh+ and Eya2+ endoderm that can be formed independently of the BMP4-induced primitive streak intermediate; this population also requires p38 MAPK activity for its formation. The gene expression profile of this population mirrors the endoderm of the distal region of the embryo (Gu et al., 2004; McKnight et al., 2007). The endoderm populations defined here are both are products of EPL cell differentiation and arise during molecular gastrulation. We propose, therefore, that these populations are subpopulations of the definitive endoderm and suggest that the terms ‘proximal definitive endoderm’ and ‘distal definitive endoderm’ are used to describe them. Two waves of definitive endoderm, which populate the more proximal (lateral endoderm) and more distal (medial endoderm) regions of the endoderm, have been proposed to occur during embryogenesis (Tam, 2007). These populations have been distinguished by their time of exit from the primitive streak, their direction of migration across the egg cylinder and their allocation to different regions of the gut tube in later development. The populations defined from in vitro differentiation here potentially represent the populations identified by fate mapping in vivo. Our proposed model (Fig. 6) addresses the role of the mesendoderm in mammalian gastrulation, suggesting a population that (1) satisfies the criteria of mesendoderm within the population of primitive streak intermediates induced by BMP4 and (2) acts as a progenitor of the proximal definitive endoderm and mesoderm. The model also describes distal definitive endoderm that can form independently of the BMP Fig. 6. A model of endoderm formation from EPL cells. Formation of the proximal definitive endoderm is dependent on p38 MAPK activity and correlates with the prior expression of markers of the primitive-streak intermediate. The initial formation of the T-expressing primitivestreak intermediate can occur in response to a number of pathways including those regulated by BMP4, activin A, serum and, we hypothesize, Wnt (WNT3), and likely results in a mixed population of progenitors (indicated by the use of several ovals). The formation of the distal definitive endoderm is dependent on p38 MAPK signalling but is not correlated with the initial formation of the T-expressing primitive-streak intermediate. PSI, primitive streak intermediate; DE, definitive endoderm.

2212


RESEARCH ARTICLE

signalling, suggesting that not all definitive endoderm formation proceeds through the mesendoderm and raising the possibility of an endoderm-specific progenitor. Definitive endoderm formation independent of a bipotent progenitor has been suggested previously by fate mapping of the embryo (Lawson et al., 1991; Rouse et al., 1994). A goal of stem cell research is to generate, in sufficient quantity, functional cell types with commercial and clinical applications. Many approaches have been reported for forming definitive endoderm and derivatives from ES cells. These rely almost exclusively on the prior formation of primitive streak intermediates and, with few exceptions (Mathew et al., 2012; Morrison et al., 2008), the positional identity of the endoderm population that is formed has not been considered. What is clear from the literature is that the formation of later endoderm populations is generally inefficient; this is potentially a consequence of the inefficient generation of the appropriate progenitor at the onset of differentiation. Positional specification could restrict the developmental potential of ES-cell-derived definitive endoderm and success might depend on enrichment for a specific endoderm population. Characterisation of the derivatives of proximal and distal definitive endoderm populations in the embryo and in culture will allow differentiation protocols to be tailored to ensure the enrichment of the appropriate definitive endoderm for subsequent differentiation. MATERIALS AND METHODS Cell culture

Mouse ES D3 ES cells (Doetschman et al., 1985) were used throughout. ES cell maintenance, EPL cell formation (as aggregates), EB formation and the production of the conditioned medium MEDII were performed as described previously (Rathjen and Rathjen, 2003). All treatments were administered to EPL cells that had been maintained in 50% MEDII for 72 h. Differentiation assays EPL cells were transferred to SCM (supplementary material Table S1). The fetal calf serum (FCS, Life Technologies) used in these experiments was chosen for the maintenance of pluripotency. Alternatively, EPL cells were transferred to SFM (supplementary material Table S1). SFM was supplemented with BMP4 (10 ng/ml, R&D Systems) (BCM; supplementary material Table S1). Noggin (90 ng/ml, R&D Systems), SB (10 mM, Sigma) and/or 0.1% DMSO (Sigma) were added as described in the text and EPL cells were cultured for 3 days with daily medium change. The formation of recognisable cell types [erythrocytes (scored as the presence of red patches of cells), pulsing cardiocytes (scored as cell movement) and neurons (scored as long cell extensions emanating from the aggregates)] from aggregates was determined as described previously (Hughes et al., 2009a; Hughes et al., 2009b). For each experimental condition §24 aggregates were analysed. Ideally, a single aggregate was analysed per well but in reality some wells contained more than one analysed aggregate. Alternatively, aggregates were treated for 4 days in suspension culture before they were mass-seeded in a 9.6 cm2 dish and maintained in SCM with regular medium change for a further 7 days. Gene expression assays EPL cells were transferred to SCM, BCM or ACM [SFM+activin A (25 ng/ml); supplementary material Table S1] and were supplemented with noggin (90 ng/ml), SB (10 mM), DAPT (50 mM) and/or 0.1 or 0.2% DMSO, as described in the text, and cultured for 4 days with daily medium change. Cells were collected after 2, 3 and 4 days of treatment. RT-PCR and quantitative PCR Total cytoplasmic RNA was isolated using TRIzolH (Invitrogen). cDNA was synthesised as per the manufacturer’s protocol (Promega). Primers


(supplementary material Table S2) were validated on differentiated ES (mouse) or genomic DNA (human), and PCR products were sequenced. For RT-PCR, 25 ml reactions contained 1 ng/ml of forward and reverse primers, 16GoTaqH Green Master Mix (Promega) and cDNA. Reactions were heated to 94 ˚C for 2 min before cycles of 94 ˚C for 30 s, 60 ˚C for 30 s and 72 ˚C for 30 s, ending with 5 min at 72 ˚C, in an MJ Research thermocycler. PCR products were visualised with a Molecular ImagerH ChemiDocTM XRS Imaging System (BioRad) with SYBRH Gold (Invitrogen). Gene expression was quantified using Quantity One 1D band analysis software (BioRad). For quantitative PCR (qPCR), reactions containing 16 Absolute blue QPCR SYBR Green Mix (Thermo Scientific), cDNA and 200 nM of forward and reverse primers were performed on an MJ research thermocycler with a Chromo4 Continuous Fluorescence Detection system (MJ Research). Reactions were heated to 95 ˚C for 15 min before cycling at 95 ˚C for 15 s, 60 ˚C for 15 s and 72 ˚C for 30 s. The raw data was analysed using the Q-Gene software package (Muller et al., 2002; Simon, 2003). Whole-mount in situ hybridisation (WISH) Embryos from time-mated Swiss mice and cell aggregates were fixed in 4% paraformaldehyde (PFA) and dehydrated in methanol. WISH was performed as described previously (Lake et al., 2000; Rosen and Beddington, 1993) with modifications. Rehydrated embryos were treated with 6% H2O2. Probes were labelled using digoxigenin-11-dUTP or fluorescein-12-UTP (Roche). The hybridisation and post-hybridisation washes were performed at 65 ˚C. Embryos and aggregates were incubated overnight with antidigoxigenin–AP Fab fragments (1:2000) (Roche) or anti-fluorescein– AP Fab fragments (1:2000) (Roche) and were developed with NBT/ BCIP or INT/BCIP (Roche), respectively, as per the manufacturer’s instructions. The samples were photographed using an Olympus UC30 camera mounted on a Motic SMX-143 stereomicroscope. Riboprobes were synthesized from pGEMT-easy vectors (Promega) containing 300 bp of Spink3, 460 bp of Ttr or 408 bp of Trh cDNA fragments, linearized with NcoI or SalI, and transcribed with SP6 (antisense) or T7 (sense) RNA polymerases. Aggregates were embedded in paraffin and sectioned as required. Western blot

EPL cell aggregates were serum starved for 2 h in SFM before BMP4 (10 ng/ml), 10% FCS or activin A (25 ng/ml) were added. Aggregates were pretreated with noggin (90 ng/ml), 0.1 or 0.035% DMSO, SB (10 mM) or LDN (350 nM) for 1 h before the addition of BMP4 or FCS. Total proteins was analysed by western blotting. Membranes were developed with ECL substrate (Amersham Pharmacia Biotech), scanned with a Molecular ImagerH ChemiDocTM XRS Imaging System (BioRad) or Fujifilm LAS-3000 (Berthold Australia Pty Ltd) and analysed by Quantity OneTM. Primary antibodies were against p38, p-p38, pSmad1/5/ 8 (Cell Signaling Technologies) and b-tubulin I (Sigma). The secondary antibodies were HRP-conjugated (Cell Signaling Technologies and DakoCytomation). HRP-uptake assay

The HRP-uptake assay was performed as described previously (KanaiAzuma et al., 2002; Vassilieva et al., 2012). Statistical analysis

Experiments were analysed using unpaired one- or two-tailed Student’s ttests in MicrosoftH Excel software. The statistical significance is denoted as follows: *P,0.05, **P,0.01. Comparisons are made between outcomes in SB, noggin and DAPT compared to the equivalent base medium, with or without DMSO. Acknowledgements The authors would like to thank members of the Rathjen laboratory for insightful discussions of the project.

2213


RESEARCH ARTICLE


Competing interests

Engelman, J. A., Lisanti, M. P. and Scherer, P. E. (1998). Specific inhibitors of p38 mitogen-activated protein kinase block 3T3-L1 adipogenesis. J. Biol. Chem. 273, 32111-32120. Evans, M. J. and Kaufman, M. H. (1981). Establishment in culture of pluripotential cells from mouse embryos. Nature 292, 154-156. Farace, M. G., Brown, B. A., Raschella`, G., Alexander, J., Gambari, R., Fantoni, A., Hardies, S. C., Hutchison, C. A., 3rd and Edgell, M. H. (1984). The mouse beta h1 gene codes for the z chain of embryonic hemoglobin. J. Biol. Chem. 259, 7123-7128. Gadue, P., Huber, T. L., Paddison, P. J. and Keller, G. M. (2006). Wnt and TGFbeta signaling are required for the induction of an in vitro model of primitive streak formation using embryonic stem cells. Proc. Natl. Acad. Sci. USA 103, 16806-16811. Gadue, P., Gouon-Evans, V., Cheng, X., Wandzioch, E., Zaret, K. S., Grompe, M., Streeter, P. R. and Keller, G. M. (2009). Generation of monoclonal antibodies specific for cell surface molecules expressed on early mouse endoderm. Stem Cells 27, 2103-2113. Graichen, R., Xu, X., Braam, S. R., Balakrishnan, T., Norfiza, S., Sieh, S., Soo, S. Y., Tham, S. C., Mummery, C., Colman, A. et al. (2008). Enhanced cardiomyogenesis of human embryonic stem cells by a small molecular inhibitor of p38 MAPK. Differentiation 76, 357-370. Gu, G., Wells, J. M., Dombkowski, D., Preffer, F., Aronow, B. and Melton, D. A. (2004). Global expression analysis of gene regulatory pathways during endocrine pancreatic development. Development 131, 165-179. Guillemot, F., Lo, L.-C., Johnson, J. E., Auerbach, A., Anderson, D. J. and Joyner, A. L. (1993). Mammalian achaete-scute homolog 1 is required for the early development of olfactory and autonomic neurons. Cell 75, 463476. Guo, Y. L., Ye, J. and Huang, F. (2007). p38alpha MAP kinase-deficient mouse embryonic stem cells can differentiate to endothelial cells, smooth muscle cells, and neurons. Dev. Dyn. 236, 3383-3392. Han, J., Lee, J. D., Bibbs, L. and Ulevitch, R. J. (1994). A MAP kinase targeted by endotoxin and hyperosmolarity in mammalian cells. Science 265, 808-811. Hanafusa, H., Ninomiya-Tsuji, J., Masuyama, N., Nishita, M., Fujisawa, J., Shibuya, H., Matsumoto, K. and Nishida, E. (1999). Involvement of the p38 mitogen-activated protein kinase pathway in transforming growth factor-betainduced gene expression. J. Biol. Chem. 274, 27161-27167. Hart, A. H., Hartley, L., Sourris, K., Stadler, E. S., Li, R., Stanley, E. G., Tam, P. P., Elefanty, A. G. and Robb, L. (2002). Mixl1 is required for axial mesendoderm morphogenesis and patterning in the murine embryo. Development 129, 3597-3608. Harvey, N. T., Hughes, J. N., Lonic, A., Yap, C., Long, C., Rathjen, P. D. and Rathjen, J. (2010). Response to BMP4 signalling during ES cell differentiation defines intermediates of the ectoderm lineage. J. Cell Sci. 123, 1796-1804. Heldin, C. H., Miyazono, K. and ten Dijke, P. (1997). TGF-beta signalling from cell membrane to nucleus through SMAD proteins. Nature 390, 465-471. Herrera, B. and Inman, G. J. (2009). A rapid and sensitive bioassay for the simultaneous measurement of multiple bone morphogenetic proteins. Identification and quantification of BMP4, BMP6 and BMP9 in bovine and human serum. BMC Cell Biol. 10, 20. Hu, M. C., Wasserman, D., Hartwig, S. and Rosenblum, N. D. (2004). p38MAPK acts in the BMP7-dependent stimulatory pathway during epithelial cell morphogenesis and is regulated by Smad1. J. Biol. Chem. 279, 1205112059. Hughes, J. N., Dodge, N., Rathjen, P. D. and Rathjen, J. (2009a). A novel role for gamma-secretase in the formation of primitive streak-like intermediates from ES cells in culture. Stem Cells 27, 2941-2951. Hughes, J. N., Washington, J. M., Zheng, Z., Lau, X. K., Yap, C., Rathjen, P. D. and Rathjen, J. (2009b). Manipulation of cell:cell contacts and mesoderm suppressing activity direct lineage choice from pluripotent primitive ectodermlike cells in culture. PLoS ONE 4, e5579. Izumi, N., Era, T., Akimaru, H., Yasunaga, M. and Nishikawa, S. (2007). Dissecting the molecular hierarchy for mesendoderm differentiation through a combination of embryonic stem cell culture and RNA interference. Stem Cells 25, 1664-1674. Jackson, S. A., Schiesser, J., Stanley, E. G. and Elefanty, A. G. (2010). Differentiating embryonic stem cells pass through ‘temporal windows’ that mark responsiveness to exogenous and paracrine mesendoderm inducing signals. PLoS ONE 5, e10706. Kanai-Azuma, M., Kanai, Y., Gad, J. M., Tajima, Y., Taya, C., Kurohmaru, M., Sanai, Y., Yonekawa, H., Yazaki, K., Tam, P. P. et al. (2002). Depletion of definitive gut endoderm in Sox17-null mutant mice. Development 129, 2367-2379. Kinder, S. J., Tsang, T. E., Wakamiya, M., Sasaki, H., Behringer, R. R., Nagy, A. and Tam, P. P. (2001). The organizer of the mouse gastrula is composed of a dynamic population of progenitor cells for the axial mesoderm. Development 128, 3623-3634. Kodaira, K., Imada, M., Goto, M., Tomoyasu, A., Fukuda, T., Kamijo, R., Suda, T., Higashio, K. and Katagiri, T. (2006). Purification and identification of a BMP-like factor from bovine serum. Biochem. Biophys. Res. Commun. 345, 1224-1231. Kubo, A., Shinozaki, K., Shannon, J. M., Kouskoff, V., Kennedy, M., Woo, S., Fehling, H. J. and Keller, G. (2004). Development of definitive endoderm from embryonic stem cells in culture. Development 131, 1651-1662. Kwon, G. S., Viotti, M. and 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.

The authors declare no competing interests.

Author contributions C.Y.: conception and design, collection and/or assembly of data, data analysis and interpretation, manuscript writing. H.N.G.: conception and design, collection and/or assembly of data, data analysis and interpretation. M.F. conception and design, data analysis and interpretation. P.D.R.: conception and design. J.R.: conception and design, data analysis and interpretation, manuscript writing, final approval of manuscript.

Funding This work was supported by the University of Melbourne and the Albert Shimmins Postgraduate Writing up Award. C.Y. and H.N.G. were supported by Australian Postgraduate Awards, C.Y. received additional support from the Australian Stem Cell Centre.

Supplementary material Supplementary material available online at http://jcs.biologists.org/lookup/suppl/doi:10.1242/jcs.134502/-/DC1

References Adams, R. H., Porras, A., Alonso, G., Jones, M., Vintersten, K., Panelli, S., Valladares, A., Perez, L., Klein, R. and Nebreda, A. R. (2000). Essential role of p38alpha MAP kinase in placental but not embryonic cardiovascular development. Mol. Cell 6, 109-116. Allen, M., Svensson, L., Roach, M., Hambor, J., McNeish, J. and Gabel, C. A. (2000). Deficiency of the stress kinase p38alpha results in embryonic lethality: characterization of the kinase dependence of stress responses of enzymedeficient embryonic stem cells. J. Exp. Med. 191, 859-870. Aouadi, M., Bost, F., Caron, L., Laurent, K., Le Marchand Brustel, Y. and Bine´truy, B. (2006). p38 mitogen-activated protein kinase activity commits embryonic stem cells to either neurogenesis or cardiomyogenesis. Stem Cells 24, 1399-1406. Barruet, E., Hadadeh, O., Peiretti, F., Renault, V. M., Hadjal, Y., Bernot, D., Tournaire, R., Negre, D., Juhan-Vague, I., Alessi, M. C. et al. (2011). p38 mitogen activated protein kinase controls two successive-steps during the early mesodermal commitment of embryonic stem cells. Stem Cells Dev. 20, 12331246. Beddington, R. S. and Robertson, E. J. (1998). Anterior patterning in mouse. Trends Genet. 14, 277-284. Beddington, R. S. and Robertson, E. J. (1999). Axis development and early asymmetry in mammals. Cell 96, 195-209. Bradley, A., Evans, M., Kaufman, M. H. and Robertson, E. (1984). Formation of germ-line chimaeras from embryo-derived teratocarcinoma cell lines. Nature 309, 255-256. Chakraborty, S., Kang, B., Huang, F. and Guo, Y. L. (2009). Mouse embryonic stem cells lacking p38alpha and p38delta can differentiate to endothelial cells, smooth muscle cells, and epithelial cells. Differentiation 78, 143-150. Clements, M., Pernaute, B., Vella, F. and Rodriguez, T. A. (2011). Crosstalk between Nodal/activin and MAPK p38 signalling is essential for anteriorposterior axis specification. Curr. Biol. 21, 1289-1295. Conlon, F. L., Lyons, K. M., Takaesu, N., Barth, K. S., Kispert, A., Herrmann, B. and Robertson, E. J. (1994). A primary requirement for nodal in the formation and maintenance of the primitive streak in the mouse. Development 120, 1919-1928. Coucouvanis, E. and Martin, G. R. (1995). Signals for death and survival: a twostep mechanism for cavitation in the vertebrate embryo. Cell 83, 279-287. Crossley, P. H. and Martin, G. R. (1995). The mouse Fgf8 gene encodes a family of polypeptides and is expressed in regions that direct outgrowth and patterning in the developing embryo. Development 121, 439-451. Cuenda, A., Rouse, J., Doza, Y. N., Meier, R., Cohen, P., Gallagher, T. F., Young, P. R. and Lee, J. C. (1995). SB 203580 is a specific inhibitor of a MAP kinase homologue which is stimulated by cellular stresses and interleukin-1. FEBS Lett. 364, 229-233. Davidson, S. M. and Morange, M. (2000). Hsp25 and the p38 MAPK pathway are involved in differentiation of cardiomyocytes. Dev. Biol. 218, 146-160. del Barco Barrantes, I., Coya, J. M., Maina, F., Arthur, J. S. and Nebreda, A. R. (2011). Genetic analysis of specific and redundant roles for p38alpha and p38beta MAPKs during mouse development. Proc. Natl. Acad. Sci. USA 108, 12764-12769. Dickson, M. C., Martin, J. S., Cousins, F. M., Kulkarni, A. B., Karlsson, S. and Akhurst, R. J. (1995). Defective haematopoiesis and vasculogenesis in transforming growth factor-beta 1 knock out mice. Development 121, 18451854. Doetschman, T. C., Eistetter, H., Katz, M., Schmidt, W. and Kemler, R. (1985). The in vitro development of blastocyst-derived embryonic stem cell lines: formation of visceral yolk sac, blood islands and myocardium. J. Embryol. Exp. Morphol. 87, 27-45. Duval, D., Malaise´, M., Reinhardt, B., Kedinger, C. and Boeuf, H. (2004). A p38 inhibitor allows to dissociate differentiation and apoptotic processes triggered upon LIF withdrawal in mouse embryonic stem cells. Cell Death Differ. 11, 331341.

2214


RESEARCH ARTICLE

Lake, J., Rathjen, J., Remiszewski, J. and Rathjen, P. D. (2000). Reversible programming of pluripotent cell differentiation. J. Cell Sci. 113, 555-566. Lawson, K. A., Meneses, J. J. and Pedersen, R. A. (1991). Clonal analysis of epiblast fate during germ layer formation in the mouse embryo. Development 113, 891-911. Lee, J. C., Laydon, J. T., McDonnell, P. C., Gallagher, T. F., Kumar, S., Green, D., McNulty, D., Blumenthal, M. J., Heys, J. R., Landvatter, S. W. et al. (1994). A protein kinase involved in the regulation of inflammatory cytokine biosynthesis. Nature 372, 739-746. Li, L., Arman, E., Ekblom, P., Edgar, D., Murray, P. and Lonai, P. (2004). Distinct GATA6- and laminin-dependent mechanisms regulate endodermal and ectodermal embryonic stem cell fates. Development 131, 5277-5286. Lickert, H., Kutsch, S., Kanzler, B., Tamai, Y., Taketo, M. M. and Kemler, R. (2002). Formation of multiple hearts in mice following deletion of beta-catenin in the embryonic endoderm. Dev. Cell 3, 171-181. Lints, T. J., Parsons, L. M., Hartley, L., Lyons, I. and Harvey, R. P. (1993). Nkx2.5: a novel murine homeobox gene expressed in early heart progenitor cells and their myogenic descendants. Development 119, 419-431. Liu, P., Wakamiya, M., Shea, M. J., Albrecht, U., Behringer, R. R. and Bradley, A. (1999). Requirement for Wnt3 in vertebrate axis formation. Nat. Genet. 22, 361-365. Lu, H. T., Yang, D. D., Wysk, M., Gatti, E., Mellman, I., Davis, R. J. and Flavell, R. A. (1999). Defective IL-12 production in mitogen-activated protein (MAP) kinase kinase 3 (Mkk3)-deficient mice. EMBO J. 18, 1845-1857. Martin, G. R. (1981). Isolation of a pluripotent cell line from early mouse embryos cultured in medium conditioned by teratocarcinoma stem cells. Proc. Natl. Acad. Sci. USA 78, 7634-7638. Martı´n-Blanco, E. (2000). p38 MAPK signalling cascades: ancient roles and new functions. Bioessays 22, 637-645. Massaous, J. and Hata, A. (1997). TGF-beta signalling through the Smad pathway. Trends Cell Biol. 7, 187-192. Mathew, S., Jaramillo, M., Zhang, X., Zhang, L. A., Soto-Gutie´rrez, A. and Banerjee, I. (2012). Analysis of alternative signaling pathways of endoderm induction of human embryonic stem cells identifies context specific differences. BMC Syst. Biol. 6, 154. McKnight, K. D., Hou, J. and Hoodless, P. A. (2007). Dynamic expression of thyrotropin-releasing hormone in the mouse definitive endoderm. Dev. Dyn. 236, 2909-2917. Morrison, G. M., Oikonomopoulou, I., Migueles, R. P., Soneji, S., Livigni, A., Enver, T. and Brickman, J. M. (2008). Anterior definitive endoderm from ESCs reveals a role for FGF signaling. Cell Stem Cell 3, 402-415. Mudgett, J. S., Ding, J., Guh-Siesel, L., Chartrain, N. A., Yang, L., Gopal, S. and Shen, M. M. (2000). Essential role for p38alpha mitogen-activated protein kinase in placental angiogenesis. Proc. Natl. Acad. Sci. USA 97, 1045410459. Muller, P. Y., Janovjak, H., Miserez, A. R. and Dobbie, Z. (2002). Short technical report processing of gene expression data generated by quantitative real-time RT-PCR. Biotechniques 32, 1372-1374. Nagata, Y., Takahashi, N., Davis, R. J. and Todokoro, K. (1998). Activation of p38 MAP kinase and JNK but not ERK is required for erythropoietin-induced erythroid differentiation. Blood 92, 1859-1869. Nakamura, K., Shirai, T., Morishita, S., Uchida, S., Saeki-Miura, K. and Makishima, F. (1999). p38 mitogen-activated protein kinase functionally contributes to chondrogenesis induced by growth/differentiation factor-5 in ATDC5 cells. Exp. Cell Res. 250, 351-363. Nebreda, A. R. and Porras, A. (2000). p38 MAP kinases: beyond the stress response. Trends Biochem. Sci. 25, 257-260. Nostro, M. C., Sarangi, F., Ogawa, S., Holtzinger, A., Corneo, B., Li, X., Micallef, S. J., Park, I. H., Basford, C., Wheeler, M. B. et al. (2011). Stagespecific signaling through TGFb family members and WNT regulates patterning and pancreatic specification of human pluripotent stem cells. Development 138, 861-871. Pelton, T. A., Sharma, S., Schulz, T. C., Rathjen, J. and Rathjen, P. D. (2002). Transient pluripotent cell populations during primitive ectoderm formation: correlation of in vivo and in vitro pluripotent cell development. J. Cell Sci. 115, 329-339. Perdiguero, E., Ruiz-Bonilla, V., Gresh, L., Hui, L., Ballestar, E., Sousa-Victor, P., Baeza-Raja, B., Jardı´, M., Bosch-Comas, A., Esteller, M. et al. (2007). Genetic analysis of p38 MAP kinases in myogenesis: fundamental role of p38alpha in abrogating myoblast proliferation. EMBO J. 26, 1245-1256. Pevny, L. H., Sockanathan, S., Placzek, M. and Lovell-Badge, R. (1998). A role for SOX1 in neural determination. Development 125, 1967-1978. Phillips, B. W., Hentze, H., Rust, W. L., Chen, Q. P., Chipperfield, H., Tan, E. K., Abraham, S., Sadasivam, A., Soong, P. L., Wang, S. T. et al. (2007). Directed differentiation of human embryonic stem cells into the pancreatic endocrine lineage. Stem Cells Dev. 16, 561-578. Raingeaud, J., Gupta, S., Rogers, J. S., Dickens, M., Han, J., Ulevitch, R. J. and Davis, R. J. (1995). Pro-inflammatory cytokines and environmental stress cause p38 mitogen-activated protein kinase activation by dual phosphorylation on tyrosine and threonine. J. Biol. Chem. 270, 7420-7426. Rathjen, J. and Rathjen, P. D. (2003). Lineage specific differentiation of mouse ES cells: formation and differentiation of early primitive ectoderm-like (EPL) cells. Methods Enzymol. 365, 1-25. Rathjen, J., Lake, J. A., Bettess, M. D., Washington, J. M., Chapman, G. and Rathjen, P. D. (1999). Formation of a primitive ectoderm like cell population,


EPL cells, from ES cells in response to biologically derived factors. J. Cell Sci. 112, 601-612. Rathjen, J., Haines, B. P., Hudson, K. M., Nesci, A., Dunn, S. and Rathjen, P. D. (2002). Directed differentiation of pluripotent cells to neural lineages: homogeneous formation and differentiation of a neurectoderm population. Development 129, 2649-2661. Rathjen, J., Washington, J. M., Bettess, M. D. and Rathjen, P. D. (2003). Identification of a biological activity that supports maintenance and proliferation of pluripotent cells from the primitive ectoderm of the mouse. Biol. Reprod. 69, 1863-1871. Robb, L., Hartley, L., Begley, C. G., Brodnicki, T. C., Copeland, N. G., Gilbert, D. J., Jenkins, N. A. and Elefanty, A. G. (2000). Cloning, expression analysis, and chromosomal localization of murine and human homologues of a Xenopus mix gene. Dev. Dyn. 219, 497-504. Rosen, B. and Beddington, R. S. (1993). Whole-mount in situ hybridization in the mouse embryo: gene expression in three dimensions. Trends Genet. 9, 162167. Rouse, J., Cohen, P., Trigon, S., Morange, M., Alonso-Llamazares, A., Zamanillo, D., Hunt, T. and Nebreda, A. R. (1994). A novel kinase cascade triggered by stress and heat shock that stimulates MAPKAP kinase-2 and phosphorylation of the small heat shock proteins. Cell 78, 10271037. Saga, Y., Hata, N., Kobayashi, S., Magnuson, T., Seldin, M. F. and Taketo, M. M. (1996). MesP1: a novel basic helix-loop-helix protein expressed in the nascent mesodermal cells during mouse gastrulation. Development 122, 27692778. Simon, P. (2003). Q-Gene: processing quantitative real-time RT-PCR data. Bioinformatics 19, 1439-1440. So, P. L. and Danielian, P. S. (1999). Cloning and expression analysis of a mouse gene related to Drosophila odd-skipped. Mech. Dev. 84, 157-160. Soudais, C., Bielinska, M., Heikinheimo, M., MacArthur, C. A., Narita, N., Saffitz, J. E., Simon, M. C., Leiden, J. M. and 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, 38773888. Tada, S., Era, T., Furusawa, C., Sakurai, H., Nishikawa, S., Kinoshita, M., Nakao, K., Chiba, T. and Nishikawa, S. (2005). Characterization of mesendoderm: a diverging point of the definitive endoderm and mesoderm in embryonic stem cell differentiation culture. Development 132, 43634374. Tam, P. P. L., Khoo, P. L., Lewis, S. L., Bildsoe, H., Wong, N., Tsang, T. E., Gad, J. M. and Robb, L. (2007). Sequential allocation and global pattern of movement of the definitive endoderm in the mouse embryo during gastrulation. Development 134, 251-260. Tan, B. S., Lonic, A., Morris, M. B., Rathjen, P. D. and Rathjen, J. (2011). The amino acid transporter SNAT2 mediates L-proline-induced differentiation of ES cells. Am. J. Physiol. Cell Physiol. 300, C1270-1279. Tanaka, N., Kamanaka, M., Enslen, H., Dong, C., Wysk, M., Davis, R. J. and Flavell, R. A. (2002). Differential involvement of p38 mitogen-activated protein kinase kinases MKK3 and MKK6 in T-cell apoptosis. EMBO Rep. 3, 785791. Tanaka, M., Jokubaitis, V., Wood, C., Wang, Y., Brouard, N., Pera, M., Hearn, M., Simmons, P. and Nakayama, N. (2009). BMP inhibition stimulates WNTdependent generation of chondrogenic mesoderm from embryonic stem cells. Stem Cell Res. 3, 126-141. Tremblay, K. D., Hoodless, P. A., Bikoff, E. K. and Robertson, E. J. (2000). Formation of the definitive endoderm in mouse is a Smad2-dependent process. Development 127, 3079-3090. Vassilieva, S., Goh, H. N., Lau, K. X., Hughes, J. N., Familari, M., Rathjen, P. D. and Rathjen, J. (2012). A system to enrich for primitive streak-derivatives, definitive endoderm and mesoderm, from pluripotent cells in culture. PLoS ONE 7, e38645. Vincent, S. D., Dunn, N. R., Hayashi, S., Norris, D. P. and Robertson, E. J. (2003). Cell fate decisions within the mouse organizer are governed by graded Nodal signals. Genes Dev. 17, 1646-1662. Wang, J., Ohmuraya, M., Hirota, M., Baba, H., Zhao, G., Takeya, M., Araki, K. and Yamamura, K. (2008). Expression pattern of serine protease inhibitor kazal type 3 (Spink3) during mouse embryonic development. Histochem. Cell Biol. 130, 387-397. Wang, J., Chen, L., Ko, C. I., Zhang, L., Puga, A. and Xia, Y. (2012). Distinct signaling properties of mitogen-activated protein kinase kinases 4 (MKK4) and 7 (MKK7) in embryonic stem cell (ESC) differentiation. J. Biol. Chem. 287, 27872797. Washington, J. M., Rathjen, J., Felquer, F., Lonic, A., Bettess, M. D., Hamra, N., Semendric, L., Tan, B. S., Lake, J. A., Keough, R. A. et al. (2010). LProline induces differentiation of ES cells: a novel role for an amino acid in the regulation of pluripotent cells in culture. Am J Physiol Cell Physiol 298, C982992. Wilkinson, D. G., Bhatt, S. and Herrmann, B. G. (1990). Expression pattern of the mouse T gene and its role in mesoderm formation. Nature 343, 657-659. Winnier, G., Blessing, M., Labosky, P. A. and Hogan, B. L. (1995). Bone morphogenetic protein-4 is required for mesoderm formation and patterning in the mouse. Genes Dev. 9, 2105-2116. Wu, J., Kubota, J., Hirayama, J., Nagai, Y., Nishina, S., Yokoi, T., Asaoka, Y., Seo, J., Shimizu, N., Kajiho, H. et al. (2010). p38 Mitogen-activated protein

2215


RESEARCH ARTICLE

RESEARCH ARTICLE

Yue, J. and Mulder, K. M. (2000). Activation of the mitogen-activated protein kinase pathway by transforming growth factor-b. Methods Mol. Biol. 142, 125131. Zheng, Z., de Iongh, R. U., Rathjen, P. D. and Rathjen, J. (2010). A requirement for FGF signalling in the formation of primitive streak-like intermediates from primitive ectoderm in culture. PLoS ONE 5, e12555. Zimmerman, L. B., De Jesu´s-Escobar, J. M. and Harland, R. M. (1996). The Spemann organizer signal noggin binds and inactivates bone morphogenetic protein 4. Cell 86, 599-606. Zohn, I. E., Li, Y., Skolnik, E. Y., Anderson, K. V., Han, J. and Niswander, L. (2006). p38 and a p38-interacting protein are critical for downregulation of Ecadherin during mouse gastrulation. Cell 125, 957-969.


kinase controls a switch between cardiomyocyte and neuronal commitment of murine embryonic stem cells by activating myocyte enhancer factor 2Cdependent bone morphogenetic protein 2 transcription. Stem Cells Dev. 19, 1723-1734. Yang, D., Tournier, C., Wysk, M., Lu, H. T., Xu, J., Davis, R. J. and Flavell, R. A. (1997). Targeted disruption of the MKK4 gene causes embryonic death, inhibition of c-Jun NH2-terminal kinase activation, and defects in AP-1 transcriptional activity. Proc. Natl. Acad. Sci. USA 94, 3004-3009. Young, P. R., McLaughlin, M. M., Kumar, S., Kassis, S., Doyle, M. L., McNulty, D., Gallagher, T. F., Fisher, S., McDonnell, P. C., Carr, S. A. et al. (1997). Pyridinyl imidazole inhibitors of p38 mitogen-activated protein kinase bind in the ATP site. J. Biol. Chem. 272, 12116-12121.


2216