Crosstalk between Fgf and Wnt signaling in the zebrafish tailbud

Developmental Biology 369 (2012) 298–307

Contents lists available at SciVerse ScienceDirect

Developmental Biology journal homepage: www.elsevier.com/locate/developmentalbiology

Crosstalk between Fgf and Wnt signaling in the zebrafish tailbud Michael J. Stulberg a, Aiping Lin b, Hongyu Zhao b,c, Scott A. Holley a,n a b c

Department of Molecular, Cellular, and Developmental Biology, New Haven, CT, USA Keck Biostatistics Resource, New Haven, CT, USA Department of Epidemiology and Public Health, Yale University, New Haven, CT 06511, USA

a r t i c l e i n f o

a b s t r a c t

Article history: Received 5 March 2012 Received in revised form 26 June 2012 Accepted 2 July 2012 Available online 14 July 2012

Fibroblast growth factor (Fgf) and Wnt signaling are necessary for the intertwined processes of tail elongation, mesodermal development and somitogenesis. Here, we use pharmacological modifiers and time-resolved quantitative analysis of both nascent transcription and protein phosphorylation in the tailbud, to distinguish early effects of signal perturbation from later consequences related to cell fate changes. We demonstrate that Fgf activity elevates Wnt signaling by inhibiting transcription of the Wnt antagonists dkk1 and notum1a. PI3 kinase signaling also increases Wnt signaling via phosphorylation of Gsk3b. Conversely, Wnt can increase signaling within the Mapk branch of the Fgf pathway as Gsk3b phosphorylation elevates phosphorylation levels of Erk. Despite the reciprocal positive regulation between Fgf and Wnt, the two pathways generally have opposing effects on the transcription of co-regulated genes. This opposing regulation of target genes may represent a rudimentary relationship that manifests as out-of-phase oscillation of Fgf and Wnt target genes in the mouse and chick tailbud. In summary, these data suggest that Fgf and Wnt signaling are tightly integrated to maintain proportional levels of activity in the zebrafish tailbud, and this balance is important for axis elongation, cell fate specification and somitogenesis. & 2012 Elsevier Inc. All rights reserved.

Keywords: Fgf signaling Wnt signaling Paraxial mesoderm Tailbud Axis elongation

Introduction Vertebrate axis elongation and segmentation are closely linked processes that occur in the tailbud, which is the posterior leading edge of the embryo (Griffith et al., 1992; Holmdahl, 1925). Recent studies have resolved some long-standing questions regarding the potency of tailbud cells and indicate that the tailbud contains bipotential neural/mesodermal stem cells (Martin and Kimelman, 2012; Takemoto et al., 2011; Tzouanacou et al., 2009; Wilson et al., 2009). The stem zone of the tailbud, which contains these bipotential cells, is one of many discrete domains of gene expression within the tailbud. However, there is also extensive cell movement before mesodermal cells enter the presomitic mesoderm (PSM) and undergo segmental patterning (Benazeraf et al., 2010; Kanki and Ho, 1997; Mara et al., 2007; Zamir et al., 2006). Morphological segmentation manifests as the formation of somites, which are anlagen of the vertebral column and skeletal muscle. Somites form sequentially from anterior to posterior in conjunction with elongation of the embryo. Canonical (b-catenin dependent) Wnt and Fibroblast growth factor (Fgf) signaling link axis elongation and somitogenesis, and perturbation of either

n

Corresponding author. E-mail address: [email protected] (S.A. Holley).

0012-1606/$ - see front matter & 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ydbio.2012.07.003

pathway results in tail truncation and/or abnormal somitogenesis (Kimelman, 2006; Pourquie, 2011). Fgf and Wnt signaling cooperate with a network of T-box family transcription factors to regulate both cell fate within the tailbud and axis elongation. This interaction appears conserved across mouse, chick, zebrafish and Xenopus (Kimelman, 2006; Wilson et al., 2009). In zebrafish, four T-box genes, ntl, spt, bra and tbx6, function in combination to specify mesoderm (Amacher and Kimmel, 1998; Goering et al., 2003; Griffin et al., 1998; Griffin and Kimelman, 2002; Martin and Kimelman, 2008). Wnt signaling has been shown to regulate tbx6 expression and promote mesodermal fates (Szeto and Kimelman, 2004). Meanwhile, ntl and spt target Fgf genes, which in-turn regulate T-box gene expression, forming a positive autoregulatory loop also required for mesoderm identity (Draper et al., 2003; Griffin et al., 1998; Griffin and Kimelman, 2003). Recent data point to another positive feedback loop between ntl, bra and canonical Wnt signaling in paraxial mesoderm precursors. This study also suggests that the T-box/Fgf loop is restricted to the axial mesoderm (Martin and Kimelman, 2008). The relationship between Wnt and Fgf signaling has been explored in the context of somitogenesis. Conditional knock-outs (cKOs) of Fgfr1 or Fgf4 and Fgf8 in mesoderm progenitors exhibit tail extension defects, alteration in cell fates and aberrant somitogenesis (Naiche et al., 2011; Niwa et al., 2007; Wahl et al., 2007). In Fgfr1 cKO mice, Wnt target genes are no longer expressed in the anterior PSM, but a high level of posterior expression of the Wnt

M.J. Stulberg et al. / Developmental Biology 369 (2012) 298–307

targets axin2, dkk1, and snail1 was revealed by in situ hybridization. These studies were confirmed with a chemical inhibitor of Fgfr1 and suggest that Wnt signaling oscillations are downstream of Fgf signaling (Wahl et al., 2007). Conversely, other evidence suggests Fgf signaling is downstream of Wnt activity. Expression of constitutively active (ca) b-catenin in the paraxial mesoderm expands the unsegmented mesoderm tissue, while the vestigial tail (vt) mutant (a Wnt3a hypomorph) has a shortened axis and malformed somites (Aulehla et al., 2008; Dunty et al., 2008; Greco et al., 1996; Takada et al., 1994). In vt/vt mouse embryos Fgf8 levels are reduced (Aulehla et al., 2003), and in ca b-catenin embryos Fgf activity is increased (Aulehla et al., 2008; Dunty et al., 2008), suggesting Wnt signaling regulates Fgf. When ca b-catenin embryos are combined with either of the Fgf cKOs, a partial rescue of somitogenesis, but not tail elongation, is observed (Aulehla et al., 2008; Naiche et al., 2011). Together, these data suggest that the Fgf and Wnt signaling pathways regulate each other and do not have a simple epistatic relationship. Fgf signaling is transduced through a number of downstream pathways, including the mitogen-activated protein kinase (Mapk) pathway and phosphoinositide 3-kinase (PI3k) pathway. Fgf activity is in a gradient in the tailbud and PSM, with activity being highest in the posterior and progressively diminishing toward the anterior (Dubrulle et al., 2001; Sawada et al., 2001). Both Mapk and PI3k exhibit graded activity in the PSM and multiple studies have examined the role of Mapk in somitogenesis and PSM cell motility (Benazeraf et al., 2010; Delfini et al., 2005; Dubrulle and Pourquie´, 2004; Niwa et al., 2007; Wahl et al., 2007). While it has no reported role in segmentation, PI3k signaling has been linked to cell motility in migrating primordial germ cells and neutrophil migration in zebrafish (Dumstrei et al., 2004; Yoo et al., 2010). Additionally, Fgf ligands have been shown to be both an attractant (Fgf4) and repellent (Fgf8b) to chick primitive streak cells (Yang et al., 2002), implicating the Fgf pathway in chemotaxis. The chemotactic property of Fgf ligands, combined with the presence of multiple downstream effectors involved in cell movement, suggest that inhibition of the Fgfr could impact cell migration within the tailbud. Here, we use pharmacological modifiers and time-resolved quantitative analysis of both nascent transcription and protein phosphorylation in the tailbud to distinguish early effects of signal perturbation from later consequences related to cell fate changes. We demonstrate that Fgf activity elevates Wnt signaling by inhibiting transcription of the Wnt antagonists dkk1 and notum1a. PI3 kinase signaling also increases Wnt signaling via phosphorylation of Gsk3b. In addition, Wnt can increase signaling within the Mapk branch of the Fgf pathway. Thus, while the Wnt and Fgf pathways each function in positive feedback loops with T-box genes, Wnt and Fgf signaling appear to be more directly integrated. More broadly, Fgf and Wnt activity have opposing effects on the transcription of co-regulated genes. This opposing regulation of target genes by the two pathways may represent a rudimentary relationship that manifests as out-of-phase oscillation of Fgf and Wnt target genes in the mouse and chick tailbud.

299

60–90 embryos in a 2 mg/mL solution of Pronase in E2. The embryos were washed several times in E2 before being incubated in 500 mL of E2 (15 mM NaCl, 0.5 mM KCl, 1 mM MgSO4, 0.15 mM KH2PO4, 0.05 mM Na2HPO4, 1 mM CaCl2, 0.7 mM NaHCO3) plus chemical modifier. Concentrations used were 50 mM SU5402 (Tocris), 100 mM U0126 (LC Laboratories), 60 mM LY294002 (LC Laboratories), and 0.3 M LiCl (Sigma). A solution of 1% DMSO in E2 was used as a vehicle control. Embryos were incubated for different lengths of time at 28 1C in glass depression slides. LiCl treatment lasted 30 min, after which embryos were transferred to E2 plus DMSO until removed for dissection. Tailbud dissections and RNA extraction Embryos were dissected in Marc’s Modified Ringers solution (MMR) modified to include a final concentration of 10 mM Hepes (MMR 10 mM Hepes). Tailbuds were collected at room temperature into ice-cold RA1 buffer for RNA extraction following kit protocols (Macherey–Nagel RNA XS) or MMR 10 mM Hepes plus a protease inhibitor cocktail (Roche) and stored at  80 1C for Western blotting. Western blots Samples were prepared by adding lysis buffer and dithiothreitol (DTT) (12 mg/mL final concentration) to tailbuds in MMR 10 mM Hepes plus protease inhibitors, and incubated at 95 1C for 5 min. Lysed samples were loaded into 4% stacking, 10% resolving bis-acrylamide gels. Proteins were transferred to nitrocellulose paper and incubated overnight in primary antibodies. Primary antibodies were used at a 1:1000 dilution and included anti-bTubulin (T4026, Sigma–Aldrich), anti-phospho-Gsk3b (Ser9) (D85E12, Cell Signaling), anti-phospho-Akt (Ser 473) (D9E XP, Cell Signaling), and anti-Map Kinase, Activated (diphosphorylated Erk 1&2) (M8159 Sigma). Primary antibodies were detected using 1:50,000 dilutions of secondary antibodies, goat anti-rabbit IgG Horseradish Peroxidase conjugated (G-21234 Molecular Probes) and rabbit anti-mouse IGG peroxidase conjugated (A9044 Sigma), followed by ECL plus Detection Reagent (GE Healthcare). Film was digitally imaged and pixel intensities calculated using ImageJ. Statistics were calculated using the Student’s unpaired t test. In situ hybridization

Methods

Probe synthesis and standard in situ hybridization protocols were performed as previously described for digoxygenin-labeled probes (Julich et al., 2005). Embryos were raised and treated as described for RNA or protein analysis, but fixed in 4% paraformaldehyde (PFA) after treatment. The experiment was repeated 4 times, staining in parallel and stopped after the same duration. Images were processed with the same magnification and image crop size. The dkk1 probe was synthesized using a forward primer 50 -tgggctgcatcaaagtggccgg-30 and reverse primer 50 -ccattccaagtcctgcttccctc-30 with a T7 sequence added to the 50 end of the reverse primer.

Zebrafish

Quantitative real-time PCR

All fish used in this study were of TLF or TLAB strains. Zebrafish were handled in accordance to protocols approved by Yale University Institutional Animal Care and Use Committee.

Extracted RNA was converted to cDNA following High-Capacity cDNA Reverse Transcription kit protocol (Applied Biosystems). cDNA was then mixed with primers, buffer, and power SYBR green (Applied Biosystems) and loaded in a 7900 HT Applied Biosystems machine. Cycle parameters were 95 1C for 10 min, followed by 40 cycles of 95 1C for 10 s, 58 1C for 1 min. Primer concentrations were adjusted to attain primer efficiencies between 90–120%. Fold change was determined by 2^( DDCt), where expression was normalized

Chemical treatment of zebrafish embryos Embryos were raised at 24 1C and moved to 20 1C overnight. At the 4 somite stage, embryos were dechorionated in batches of

300

axin2 dkk1 notum1a spred3 her1 her7 hes6 sef sprouty4 dusp4 pea3 tbx6 tbx24 fgf8a cdx4

M.J. Stulberg et al. / Developmental Biology 369 (2012) 298–307

Forward 50 -30

Reverse 50 -30

Exon/Intron

CCAGCAGCAAAGCCTTCAGT GCTTGGCATGGAAGAGTTCG CTGCCCAAACAATCGGACGC GGACAGAGGATCTGACTCGCCA TGCATGCCTTTCCACTCTCCCTAAC CGGGCTGCTTTTTGAAGACA GCAACACTCACGACGAGGAT TGAGCTCACAGCCCTTCTCA ATGAGGACGAGGAAGGCTCC ACAACGCGCTGTGTAGGGATAC TATGACCGAGCGGCCCACAT TCCATCCAGACTCACCCGCC TCACTAAGAAGGGCAGGCGAATGT CATATGTTGCGAGTCACAGTGTGGA TCGGCGGTTCTTCCTCCACCA

GCGCGCACAAAGTAGACGTA AGTGACGAGCGCAGCAAAGT CTCAGAGGGGTGTCATCTTCGC AGGGGCGCACAAAACAGGTG ATGGCATCTGGGGTCTCCTT TGTCAACTCTTATGTTTTGTAGCAACC AGTTGTGGGAAACGTCGACAA GCAGAAAAGATGGCGGAAAG GCATTTCTGCGAAAGCTTGG GGACCACCCTGGTCGTGTTG GCAGCCAGGCACATCAGGTT AGTGAAGAACCACCAGGCCGT ACCACCTGCCAGCTGTTGTCTTG TGGTTACCTGAGCATAGTAGCAAAACG CGTTTTGCCGGTTGATGACGACT

Intron Exon Exon Exon Intron Intron Intron Intron Exon Exon Exon Exon Exon Exon Exon

to b-actin and compared to a wt control, and converted to log10. Statistical comparisons were made using the Student’s unpaired t test. b-actin primers were previously described (Keegan et al., 2002), but all other primers were designed for this study as indicated above. Microarray analysis Extracted RNA was cleaned via an additional step using an RNA Clean and Concentrator kit (Zymo Research) for increased purity. A fraction of extracted RNA was converted to cDNA and tested by qPCR to verify drug effect. The RNA then was shipped to Ambry Genetics for hybridization and reading using 4  44k zebrafish arrays (v3) from Agilent Technologies. Microarrays were performed in triplicate. For the SU5402 3 h versus 7 h treatment comparison, a minimum 2-fold change and p-value of less than 0.05 was the cutoff. For comparison of microarray data for 3 h SU5402 and LiCl treatments, a 1.5-fold cutoff was used with po0.05. GEO accession number: GSE38969. Gene ontology We used DAVID Bioinformatics Resources 6.7 to cluster genes regulated from the microarray (Huang da et al., 2009a, b). Genes were identified by Official_Gene_Symbol using the zebrafish background. Genes were clustered according to KEGG annotation or Biological process FAT. Classifications selected had a p-value of less than 0.05.

Results Time course of Fgf inhibition and Wnt activation following pharmacological modification Chemical treatments were initiated at the beginning of segmentation to allow for long chemical incubations, and we dissected tailbuds from embryos undergoing somitogenesis to gain tissue specificity (Fig. 1A and B). After 3 h of chemical treatment, gene expression levels changed significantly. Known direct targets of Fgf (sef and sprouty4) and Wnt (dkk1 and axin2) signaling were used to verify drug efficacy (Chamorro et al., 2005; Furthauer et al., 2002; Furthauer et al., 2001; Jho et al., 2002; Tsang et al., 2002). After 3 h of treatment with the Gsk3b inhibitor, LiCl (Klein and Melton, 1996; Stambolic et al., 1996), a significant increase in nascent axin2 RNA and dkk1 mRNA was observed, indicating induction of Wnt signaling (Fig. 1C).

Nascent qPCR primers include one primer in an intron region and the other in an exon, while mRNA qPCR primers are exon primers that span a spliced intron site. Embryos treated with the Fgfr1 inhibitor, SU5402 (Mohammadi et al., 1997), displayed a decrease in both sef and sprouty4 nascent RNA and mRNA levels, respectively, after 3 h of treatment, indicating Fgf inhibition (Fig. 1D). After 7 h of treatment with SU5402, nascent her1 and her7 transcription and tbx6 mRNA levels were significantly down-regulated, marking a second timepoint in SU5402 treatment at which the number of presomitic mesoderm cells is reduced (Fig. 1E). Agilent zebrafish (v3) microarrays were performed at both SU5402 time-points and the number of regulated genes nearly doubled at 7 h (Fig. 1F). This latter dataset includes a much larger Gene Ontology profile, suggesting many more secondary effects at the later time point (Fig. 1G). Therefore, the 3 h time-point was used for most of our analyses as we conclude that at 7 h much of the change in gene expression is due to alteration in cell fate. Fgf and Wnt signaling oppositely affect 95% of co-regulated genes We performed microarray analysis of dissected tailbuds from embryos treated with LiCl or SU5402 for 3 h. The LiCl and SU5402 microarrays had an overlap of 274 targets (Fig. 2A). Interestingly, 263 of the 274 targets (95%) were oppositely regulated by the Wnt and Fgf pathways meaning that expression either increased after both LiCl and SU5402 treatment or decreased after both LiCl and SU5402 treatment. About half of the co-regulated genes were up-regulated by both Wnt and Fgf signaling and roughly half were down-regulated by the two pathways. Two of the genes up-regulated were dkk1 and notum1a, known Wnt inhibitors (Flowers et al., 2012). Additionally, one of the down-regulated genes was spred3, a known Fgf inhibitor. Most of the genes that were classifiable by Gene Ontology were down-regulated in both treatments including cell adhesion genes that may be involved in cell migration in the tailbud (Fig. 2B). These data suggest that Fgf promotes cell–cell adhesion while Wnt may antagonize it. Gene Ontology analysis using the KEGG classification of genes from the LiCl array represents many groups. Spliceosome and neural genes were up-regulated, and many metabolism genes were down-regulated (Fig. 2C). Tailbud precursors begin to downregulate amino-acid metabolism genes and increase oxidative metabolism genes as they differentiate (Ozbudak et al., 2010). Our LiCl treatment exhibits an increase in aminoacyl-tRNA biosynthesis and a decrease in many metabolism pathway genes, which suggest an increase in Wnt activity leads to reduction in cell differentiation, consistent with models of Wnt maintaining an

M.J. Stulberg et al. / Developmental Biology 369 (2012) 298–307

301

Fig. 1. Time course of the transcriptional response in the zebrafish tailbud following activation of Wnt signaling or inhibition of Fgf signaling. (A) Graphical representation of the experimental protocol. (B) DIC images of embryos after 3 h treatment of DMSO, LiCl, and SU5402. (C) Time course of chemical treatment. LiCl activation of Wnt signaling was measured by monitoring transcription of known direct Wnt targets dkk1 and axin2. (D) Time course of SU5402 treatment, an inhibitor of Fgfr1. Fgf signal inhibition was assayed by quantifying transcription of direct Fgf targets sef and spry4. (E) A reduction of presomitic mesoderm cells upon 7 h SU5402 treatment is revealed by reduction in her1, her7, and tbx6 expression. (F) Microarray analysis at the early time point (3 h) and later time point (7 h) suggests that many secondary changes in gene expression are observed at the later time point due to alterations in cell fate and/or cell survival. The data were processed via Agilent’s Feature Extraction software version 10.7.3.1. Samples were normalized using the percentile shift algorithm. No baseline transformation was performed. A 2-fold change cutoff was used to define regulated transcripts. (G) Gene Ontology classification based on Biological Processes Fat indicates that genes representing many more processes are differentially expressed at the later time point than the early, suggesting a loss of cell fates. Fat clustering is a version of Gene Ontologies (GO) that filters out the broadest GO terms so that they do not overshadow more specific GO terms. Fat contrasts to Slim, which provides a broader overview of the ontology content. Error bars are SEM, n¼ 3.

undifferentiated state in the tailbud (Pourquie, 2011). Two downregulated classification groups, VEGF and Insulin, include many members of the Mapk and PI3k family, which are shared effectors of Fgf signaling. KEGG pathway classification indicated genes from the Wnt pathway were enriched among those down-regulated by SU5402 (Fig. 2D). Also represented among down-regulated genes were the TGFb signaling pathway and Jak/Stat signaling pathway, the latter of which utilizes many of the same downstream effectors as Fgf signaling. Among genes up-regulated by SU5402 treatment, retinol metabolism genes were enriched, which was expected due to the opposing gradients of Fgf and retinoic acid signaling in the PSM (Diez del Corral et al., 2003). We note the Wnt pathway genes were found to be either up- or down-regulated after inhibition of Fgf signaling

(Fig. 2D), with the Wnt inhibitors dkk1 and notum1a being up-regulated. Fgf appears to influence Wnt signaling by both increasing expression of Wnt ligands and repressing expression of Wnt antagonists. To visualize the spatial change in dkk1 mRNA, we performed in situ hybridization following treatment with LiCl, SU5402, and a combination of LiCl and SU5402 (Fig. 3A). dkk1 transcript levels appear increased in the posterior tailbud following all treatment conditions and displayed little evidence of an expansion in the spatial domain of expression. Indeed, there was little evidence of a change in spatial pattern of other tailbud expressed genes that we examined (SFig. 1). The exception was the occasional extra her1 stripe due to an anterior shifting of the segmentation wavefront. A similar observation had been previously made in mice expressing

302

M.J. Stulberg et al. / Developmental Biology 369 (2012) 298–307

Biological Process GO: Overlap genes

Wnt

ventricular system development

Fgf

up regulated

fourth ventrical development

down regulated

cell proliferation in hindbrain transcription cell-cell adhesion LiCl 2305 genes

274 genes

regulation of transcription

SU5402 979 genes

biological adhesion cell adhesion regulation of RNA metabolic process regulation of transcription, DNA-dependent 0

KEGG GO analysis: LiCl

2

4

6 8 10 12 14 16 Number of genes

KEGG GO analysis: SU5402

Aminoacyl-tRNA biosynthesis Terpenoid backbone biosynthesis Spliceosome Insulin signaling pathway beta-Alanine metabolism Propanoate metabolism Lysine degradation Butanoate metabolism VEGF signaling pathway p53 signaling pathway Apoptosis Pantothenate and CoA biosynthesis Glycolysis / Gluconeogenesis

Retinol metabolism

up regulated down regulated

Hedgehog signaling pathway Wnt signaling pathway up regulated

Adherens junction

down regulated

Jak-STAT signaling pathway TGF-beta signaling pathway 0

0

LiCl regulated mapkapk3

2

4 6 8 10 Number of genes p value

fold change

3.47E-05

1.73585

12

14

2

4 6 8 10 Number of genes

p value

fold change

dkk1

0.002422

2.3264

notum1a wnt11r wnt3

0.000186

3.44682

0.017037

1.87192

0.01302

1.7602

wnt8a wnt5b

1.52E-05 0.02757

1.65415

1.38E-05

-1.50424 -3.26844

8.43E-06

-16.812 -2.09263

SU5402 regulated

mapk7

0.00698

-1.59705

map2k1

0.00741

-1.60936

LOC555662

0.02055

-1.82608

wif1 wnt5a

LOC100005183

0.00435

-1.94306

sfrp1b

prkab1a

0.00073

-1.50544

sfrp1a

0.031695 0.001185

fzd8a

0.002038

lef1

0.002221

LOC100149794

0.00128

-1.70677

12

-1.70135 -1.75045 -1.87208

Fig. 2. Microarray analysis comparing Fgf and Wnt regulated genes in the tailbud. (A) Venn diagram of genes regulated in the two arrays, including overlap genes. (B–D) Gene Ontology classification of the overlap genes is based on Biological Processes Fat (B), while the classification for LiCl (C) and SU5402 (D) experiments uses KEGG pathway classification. (C–D) A list of genes both up- and down- regulated in LiCl and SU5402 experiments that represent Mapk and PI3k genes in the LiCl array (C), and Wnt pathway genes in the SU5402 array (D).

a constitutively active beta-catenin in the tailbud (Aulehla et al., 2008; Dunty et al., 2008). To validate our microarrays, we used qPCR to measure changes in gene expression of select Wnt and Fgf targets, as well as other genes involved in segmentation. Treatment with LiCl produced an increase in dkk1, axin2, and notum1a and a reduction of spred3 mRNA in our microarray data; these changes were validated by qPCR. None of the segmentation genes (her1, her7, and hes6) showed differential expression by either microarray or qPCR analysis following LiCl treatment. Treatment with SU5402 resulted in an expression change of a number of Wnt, Fgf, and segmentation genes that were validated by qPCR (Fig. 3B). Fgf and Wnt signaling are integrated via phosphorylation of Gsk3b, Erk and Akt To analyze the phosphorylation states of Fgf and Wnt signaling pathways, we selected canonical components of each pathway. For the PI3k branch of the Fgf pathway, phosphorylated Akt (pAkt) levels

were measured. Double phosphorylated Erk (dpErk) levels were used to quantify signaling through the Mapk branch. Gsk3b phosphorylation (pGsk3b) was used as read-out of Wnt signaling. Western blots were performed on pooled dissected tailbuds, and pools were run on two gels that were subsequently probed for different phosphoproteins along with a b-Tubulin control for normalizing total protein levels. To determine whether Wnt signaling regulates Fgf signaling in the zebrafish tailbud, we measured the levels of dpErk and pAkt following LiCl treatment, with pGsk3b levels acting as a positive control. Western blot analysis indicated that treatment with LiCl increased pGsk3b and dpErk levels, but had no effect on pAkt levels (Fig. 4A–C), suggesting that Wnt signaling regulates Mapk signaling but not PI3k signaling in the zebrafish tailbud. Previous studies indicate that PI3k signaling increases pGsk3b levels in vitro (Hashimoto et al., 2002), but whether this regulation occurs during vertebrate axis elongation is unknown. We inhibited PI3k using the small molecule LY294002 (Vlahos et al., 1994) and examined phospho-protein levels and gene expression.

M.J. Stulberg et al. / Developmental Biology 369 (2012) 298–307

303

dkk1

DMSO

LiCl

SU5402

LiCl / SU5402

6 4 3 2 1.5 0 -1.5 -2 -3 -4 -6 -8 -12

SU5402

** ** Fold Change

Fold Change

LiCl

*

*

6 4 3 2 1.5 0 -1.5 -2 -3 -4 -6 -8 -12

**

** **

**

*

* ***

***

***

** **

*** tbx24

tbx6

spred3

pea3

dusp4

sprouty4

sef

hes6

her7

her1

fgf8

notum1a

dkk1

hes6

her7

her1

spred3

notum1a

dkk1

axin2

***

Fig. 3. In situ hybridization of dkk1 following LiCl and SU5402 treatment and qPCR validation of selected microarray data. (A) In situ hybridization of dkk1 following LiCl, SU5402, and LiCl/SU5402 treatment. Embryos were stained in parallel for the same duration and images were processed identically. The spatial pattern of dkk1 does not appear to expand, but levels increase following treatment with LiCl, SU5402, and a combination of LiCl and SU5402. Posterior is down. n¼ 4 experiments, 15 embryos per treatment per experiment. (B) We used direct transcriptional targets of the Wnt and Fgf pathways, along with genes involved in somitogenesis, to validate the microarray data. No change in transcription was observed via microarray or qPCR for genes involved in somitogenesis (her1, her7, hes6) following LiCl treatment. Changes identified via microarray in axin2, dkk1, notum1a, and spred3 expression following LiCl treatment were confirmed by qPCR. Inhibition of Fgf signaling using SU5402 resulted in a change in dkk1, notum1a, fgf8, her1, her7, hes6, sef, sprouty4, dusp4, pea3, spred3, tbx6, and tbx24 expression detected via microarray, and these changes were confirmed by qPCR (n ¼ 4). Error bars are SEM. Statistical comparisons were made using the Student’s unpaired t test. *p r 0.06, **p r 0.01, ***p r 0.001.

Treatment for 3 h led to the expected decrease in pAkt levels, and also a reduction in pGsk3b levels (Fig. 4A and B). This decrease in pGsk3b levels correlates with the lower Wnt target gene levels measured by qPCR (Fig. 5A), suggesting that PI3k interfaces with Wnt signaling in the tailbud. Regulation of Wnt signaling by Fgfr and Mapk is more complex, and the data are more difficult to interpret. cdx4 expression is regulated by both Wnt and Fgf signaling (Shimizu et al., 2005), and we find that cdx4 expression is down-regulated after Mapk inhibition (Fig. 5C). Inhibition of Mapk signaling also increases dkk1 and notum1a expression in a manner similar to that observed after inhibition of Fgfr (Fig. 5B and C). The elevation in expression of Wnt antagonists following inhibition of Fgfr and Mapk signaling would be predicted to lead to a decrease in Wnt signaling, but neither inhibition of Fgfr nor Mapk signaling reduced the level of pGsk3b (Fig. 4A). The expression of Wnt antagonists such as notum1a and dkk1 would normally decline when Wnt signaling is abrogated but the regulation of these genes by Fgf signaling makes their expression unsuitable reporters of Wnt activity in the SU5402 and U0126 experiments. Fgf signaling in the tailbud utilizes the Mapk pathway more than the PI3k pathway To test whether Fgfr signaling in the tailbud is mediated by Mapk or PI3k signaling, we quantified phosphorylation levels and gene expression following SU5402 treatment as well as after inhibition of Mapk or PI3k. Treatment with SU5402 did not alter pAkt or pGsk3b levels, but did reduce the amount of dpErk

(Fig. 4A–C). Inhibiting Mapk using the small molecule U0126 (Favata et al., 1998) resulted in a decrease of dpErk levels, as expected (Fig. 4C). Analysis of gene expression by qPCR indicates that many of the same genes regulated by SU5402 are also affected by U0126. Two notable exceptions are notum1a (Fig. 5C) and fgf8, which are both up-regulated in response to SU5402 in both our microarray and our qPCR data. We conclude that in zebrafish, Mapk mediates much of the transcription regulation downstream of Fgfr in the PSM, but Mapk inhibition does not account for all of the changes observed after SU5402 treatment. Inhibition of Fgf signaling via SU5402 and inhibition of PI3k signaling via LY294002 treatment produces few shared effects on levels of transcription and phosphorylation. PI3k signaling regulates dusp4, spry4 and hes6, consistent with SU5402 treatment, but no major change is observed by qPCR on other Fgf targets, and the effect on the Wnt inhibitors dkk1 and notum1a is opposite that seen with SU5402 treatment (Fig. 5D). However, as previously reported, inhibition of PI3k signaling does reduce tail elongation, whereas Mapk inhibition does not (data not shown)(Finkielsztein and Kelly, 2009; Hawkins et al., 2008). Thus, PI3k may function independently of Fgfr and play a more important role in regulating tail morphogenesis than in governing tailbud transcription.

Discussion We report cross-regulation between the Fgf and Wnt signaling pathways in the zebrafish tailbud. Previous studies attempted to

304

M.J. Stulberg et al. / Developmental Biology 369 (2012) 298–307

Fig. 4. Quantification of Gsk3b, Akt and Erk phosphorylation in the tailbud following perturbation of Fgfr1, Mapk, PI3k and Gsk3b signaling. (A) Western blot analysis of pGsk3b after treatment with LiCl, SU5402, U0126 (Mapk inhibitor), and LY294002 (PI3k inhibitor), reveals levels are significantly down-regulated in LY294002 (n¼ 11) and up-regulated in LiCl (n¼ 9) treatments. (B) pAkt is significantly down-regulated by LY294002 (n¼10) and (C) dpErk is down-regulated by U0126 (n¼4), SU5402 (n¼ 5), and is up-regulated by LiCl (n¼5). Phospho-protein levels normalized to b-Tubulin levels. Statistical comparisons were made using the Student’s unpaired t test. *pr0.05, **pr0.01, ***pr0.001. Error bars are SEM.

establish a clear epistatic hierarchy between the Fgf and Wnt pathways, and we provide an explanation for their inability to unambiguously do so (Aulehla and Pourquie, 2008; Dunty et al.,

2008; Wahl et al., 2007). We find that these pathways are integrated via protein phosphorylation and transcriptional regulation (Fig. 5E and F). For example, inhibition of Fgfr and Mapk leads to an increase in expression of the Wnt inhibitor dkk1, consistent with observations by ISH in mouse embryos (Wahl et al., 2007). These data suggest that Fgf signaling promotes Wnt signaling via inhibition of Wnt antagonists. Inhibition of Fgfr results in a change of expression of some Wnt target genes. This result is consistent with experiments with chick extracts (Hardy et al., 2011). The Wnt pathway genes regulated by Fgfr include both canonical pathway genes and some non-canonical Wnt ligands. Notably, non-canonical Wnt signaling is also required for zebrafish tail elongation (Marlow et al., 2004). We also find that activating Wnt signaling increases Fgf signaling. This interaction has been observed in mouse via ISH analysis of fgf8 and pea3 following constitutively active b-catenin induction (Aulehla et al., 2008; Dunty et al., 2008). In our study, inhibition of Gsk3b leads to an increase in dpErk, suggesting a direct response of Fgf activity to changes in Wnt activity. pGsk3b can promote the phosphorylation of Erk by de-regulating its upstream kinases (Kim et al., 2012). Our microarray data also showed LiCl treatment down-regulates spred3 which acts as an inhibitor of Mapk signaling. The communication between the Fgf and Wnt pathways suggests that it is important for Fgf and Wnt signaling to remain proportional during tail elongation and somitogenesis. Previous experiments with b-catenin and Erk, along with mathematical modeling, suggest that it is not absolute levels of signal that are important for cell differentiation, but the relative change in signal levels (CohenSaidon et al., 2009; Goentoro and Kirschner, 2009; Goentoro et al., 2009). Additionally, if the pathways cause opposing effects on cell adhesion, a proper balance may be necessary to ensure normal cell flow in the tailbud. The crosstalk between Wnt and Fgf likely begets more robust signaling, which buffers environmental and genetic variation so that the ratio of signaling remains consistent from embryo to embryo. Prior studies of somitogenesis suggest that Mapk is the main effector of Fgf signaling, and we verify that Mapk is downstream of Fgf signaling in the zebrafish tailbud (Niwa et al., 2007; Sawada et al., 2001). Our data indicate that PI3k signaling does regulate some Fgf targets, but inhibition of PI3k does not significantly affect sef expression and generally has distinct effects on gene regulation relative to inhibition of Mapk. PI3k signaling was not affected by inhibition of Fgfr, but we did observe an axis elongation defect after PI3k inhibition suggesting that PI3k signaling functions independently of Fgfr signaling in regulating axis elongation. While PI3k signaling is canonically downstream of Fgf, PI3k/Akt can function within other signaling pathways. For example, BMP can signal through PI3k/Akt in embryonic stem cells (Lee et al., 2009), and BMP signaling is also active in the zebrafish tailbud (Row and Kimelman, 2009; Szeto and Kimelman, 2004). Additionally, a recent study demonstrated that in skeletal muscle, non-canonical Wnt signaling via Wnt7a/Fzd7 can trigger the phosphorylation of Akt through PI3k (von Maltzahn et al., 2012). While our data indicate that PI3k has an important role in regulating trunk and tail morphogenesis, it is unclear which pathway(s) signal through PI3k in the tailbud. Comparison of gene expression after inhibition of PI3k to activation of Wnt reveals correlations that suggest PI3k signaling normally represses Wnt signaling in the tailbud. Indeed, we find that PI3k signaling promotes phosphorylation, i.e., inhibition, of Gsk3b (Fig. 6). This interaction with the Wnt/beta-catenin pathway is hypothesized to be integral to PI3k/Akt signaling in cancer metastasis (Qiao et al., 2008). We observed that transcription of the Wnt signaling antagonists dkk1 and notum1a is governed by both Wnt and Fgf. These

M.J. Stulberg et al. / Developmental Biology 369 (2012) 298–307

8

8 6

**

*

1.5

*

0 -1.5

**

2 ***

*

1.5

*

0 -1.5

* *

**

-2

*

-2

*

**

4

**

**

3 2

spry4

sef

**

**

3

SU5402 U0126

***

dusp4

-4

dusp4

her1

spry4

hes6

notum1a

notum1a

dkk1

dkk1

axin2

axin2

-4

SU5402 LY294002

2

***

*

***

-3

-3

1.5

*

*

**

spred3

**

2

LiCl U0126

3 Fold Change

Fold Change

3

**

4

LiCl LY294002

pea3

4

tbx6

6

4

305

1.5 * 0

-1.5

*

-2

*

*

**

** ***

-3

**

***

-4

***

***

***

* *

-2

*

**

*

-3

**

-6

**

-8

***

***

***

-8

**

-12

-12 cdx4

pea3

tbx6

spry4

*** sef

dusp4

her1

notum1a

hes6

*** dkk1

-16

-1.5

-4

**

-6

**

Fold Change

Fold Change

0

-16

dkk1 notum1a hes6

*** her1 dusp4

Fgfr

sef

*** spry4

tbx6

pea3

pGsk3β Mapk

Wnt

Fgf

PI3k spry4, sef, dusp4 Fgf inhibitors

p

Fgfr

Akt

p Gsk3β

Fgfr PI3k

PI3k Mapk

pGsk3β

Mapk

p Erk

p

hes6, her1, tbx6 Mesoderm genes

dkk1, notum1a Wnt inhibitors

Fig. 5. Comparison of the transcriptional responses within the tailbud following perturbation of Fgfr1, Mapk, PI3k and Gsk3b signaling. (A) qPCR analysis comparing LiCl to LY294002. LiCl activates Wnt (via Gsk3b), while LY294002 inhibits PI3k. Genes up-regulated by LiCl are down-regulated after treatment with LY294002. (B) LiCl treatment response compared to U0126 treatment. (C) The Mapk inhibitor U0126 and SU5402 appear to regulate the same genes in a similar manner. The Wnt inhibitor dkk1 is upregulated after treatment with both SU5402 and U0126, and genes in the Fgf synexpression group are down-regulated by both SU5402 and U0126, consistent with Mapk mediating Fgfr signaling in the tailbud. (D) The PI3k inhibitor LY294002 and Fgfr inhibitor SU5402 exhibit more limited similarities in gene regulation. n¼ 3 or 4. Statistical analyses utilized Student’s unpaired t test. *p r 0.05, **pr 0.01, ***p r0.001. Error bars are SEM. E and F are graphical depictions of Western blot and qPCR data. (E) Fgf signals more strongly through the Mapk pathway. PI3k promotes Wnt signaling by increasing pGsk3b levels. Both Fgfr and pGsk3b increase dpErk levels. Arrows indicate phosphorylation events. (F) qPCR analysis indicates that Fgfr and Mapk promote both Fgf inhibitor and mesoderm-specific transcription, while repressing transcription of Wnt antagonists. PI3k promotes Wnt antagonist transcription but has weaker effects on Fgf inhibitor and mesoderm-specific transcription. Activating Wnt signaling by inhibiting Gsk3b (pGsk3b) induces Wnt inhibitor gene expression and causes a moderate induction of some Fgf inhibitors. Green arrows signify positive regulation while red lines with stops represent inhibition. The arrow thickness represents relative strength of regulation.

306

M.J. Stulberg et al. / Developmental Biology 369 (2012) 298–307

Acknowledgements

Fgf

Wnt PI3k p Gsk3β p

p

Erk

We would like to thank Patrick McMillen and Jamie Schwendinger-Schreck for comments on the manuscript and Jamie for providing the hes6 qPCR primers. Research support provided by an NIH predoctoral Genetics training grant (T32 GM007499 to MJS) and by the National Institute of Child Health and Human Development (RO1 HD045738 to SAH).

Appendix A. Supporting information ntl, spt, tbx6 T-box genes

spred3 dkk1, notum1a Microarray overlap genes Fig. 6. A simple model of the Fgf and Wnt gene network in the zebrafish tailbud. In addition to the previously described autoregulatory loops between Fgf, Wnt and T-box genes (grey arrows), we find evidence of more direct intertwining of the Fgf and Wnt pathways (colored arrows). Fgf signaling inhibits transcription of the Wnt inhibitors dkk1 and notum1a. PI3k signaling activates the Wnt pathway by phosphorylating (inactivating) Gsk3b. Phosphorylation of Gsk3b also leads to an increase in active Erk levels, suggesting Wnt signaling promotes Mapk activity. Fgf and Wnt have opposing effects on transcription of co-regulated genes, which include Fgf (spred3) and Wnt (dkk1, notum1a) inhibitors.

two genes are oppositely regulated by the two pathways, in addition to 261 other genes identified by microarray analysis (Fig. 6). In mouse embryos, dkk1 expression normally oscillates out-of-phase with Fgf target gene oscillations in the PSM, suggesting an antagonistic relationship between the two pathways (Dequeant et al., 2006). In zebrafish, dkk1 is expressed in a posterior to anterior gradient in the tailbud and exhibits no evidence of oscillation (Hashimoto et al., 2000). We find that increases in dkk1 expression after activation of Wnt signaling or inhibition of Fgf signaling manifest as an increase in the slope of the dkk1 gradient rather than an expansion of the expression domain. The antagonistic regulation of dkk1 by Wnt and Fgf in zebrafish may be a rudimentary relationship related to the antiphase oscillations of dkk1 and sprouty in the mouse. Future experiments exploring BMP signal integration with Wnt and Fgf in the tailbud should expand the gene network governing tailbud development. A number of studies in zebrafish indicate the importance of BMP in trunk and tail elongation (Agathon et al., 2003; Row and Kimelman, 2009; Szeto and Kimelman, 2004). Indeed, a study in zebrafish utilized ectopic expression of BMP during axis elongation and found that BMP regulates Wnt signaling (Row and Kimelman, 2009). Our microarray results showed many members of the BMP pathway being regulated by both Wnt and Fgf signaling, thus BMP signaling may also constitute part of the broader regulatory network governing mesoderm and neural development in the vertebrate trunk and tail.

Conclusion Fgf signaling inhibits Wnt inhibitors and promotes the expression of some Wnt ligands. Wnt, in turn, promotes the phosphorylation of Erk, a downstream effector of Fgf. These interactions contrast to the indirect activation provided by T-box gene expression that forms an autoregulatory loop with Wnt and Fgf signaling (Fig. 6). Our data indicate that Fgf and Wnt signaling are intertwined to maintain proportional signaling levels, facilitating normal axis elongation, cell differentiation and segmentation.

Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.ydbio.2012.07.003.

References Agathon, A., Thisse, C., Thisse, B., 2003. The molecular nature of the zebrafish tail organizer. Nature 424, 448–452. Amacher, S.L., Kimmel, C.B., 1998. Promoting notochord fate and repressing muscle development in zebrafish axial mesoderm. Development 125, 1397–1406. Aulehla, A., Pourquie, O., 2008. Oscillating signaling pathways during embryonic development. Curr. Opin. Cell. Biol. 20, 632–637. Aulehla, A., Wehrle, C., Brand-Saberi, B., Kemler, R., Gossler, A., Kanzler, B., Herrmann, B.G., 2003. Wnt3a plays a major role in the segmentation clock controlling somitogenesis. Dev. Cell. 4, 395–406. Aulehla, A., Wiegraebe, W., Baubet, V., Wahl, M.B., Deng, C., Taketo, M., Lewandoski, M., Pourquie, O., 2008. A beta-catenin gradient links the clock and wavefront systems in mouse embryo segmentation. Nat. Cell. Biol. 10, 186–193. Benazeraf, B., Francois, P., Baker, R.E., Denans, N., Little, C.D., Pourquie, O., 2010. A random cell motility gradient downstream of FGF controls elongation of an amniote embryo. Nature 466, 248–252. Chamorro, M.N., Schwartz, D.R., Vonica, A., Brivanlou, A.H., Cho, K.R., Varmus, H.E., 2005. FGF-20 and DKK1 are transcriptional targets of beta-catenin and FGF-20 is implicated in cancer and development. EMBO J. 24, 73–84. Cohen-Saidon, C., Cohen, A.A., Sigal, A., Liron, Y., Alon, U., 2009. Dynamics and variability of ERK2 response to EGF in individual living cells. Mol. Cell. 36, 885–893. Delfini, M.C., Dubrulle, J., Malapert, P., Chal, J., Pourquie, O., 2005. Control of the segmentation process by graded MAPK/ERK activation in the chick embryo. Proc. Nat. Acad. Sci. U.S.A. 102, 11343–11348. Dequeant, M.L., Glynn, E., Gaudenz, K., Wahl, M., Chen, J., Mushegian, A., Pourquie, O., 2006. A complex oscillating network of signaling genes underlies the mouse segmentation clock. Science 314, 1595–1598. Diez del Corral, R., Olivera-Martinez, I., Goriely, A., Gale, E., Maden, M., Storey, K., 2003. Opposing FGF and retinoid pathways control ventral neural pattern, neuronal differentiation, and segmentation during body axis extension. Neuron 40, 65–79. Draper, B.W., Stock, D.W., Kimmel, C.B., 2003. Zebrafish fgf24 functions with fgf8 to promote posterior mesodermal development. Development 130, 4639–4654. Dubrulle, J., McGrew, M.J., Pourquie, O., 2001. FGF signaling controls somite boundary position and regulates segmentation clock control of spatiotemporal Hox gene activation. Cell 106, 219–232. Dubrulle, J., Pourquie´, O., 2004. fgf8 mRNA decay establishes a gradient that couples axial elongation to patterning in the vertebrate embryo. Nature 427, 419–422. Dumstrei, K., Mennecke, R., Raz, E., 2004. Signaling pathways controlling primordial germ cell migration in zebrafish. J. Cell. Sci. 117, 4787–4795. Dunty Jr., W.C., Biris, K.K., Chalamalasetty, R.B., Taketo, M.M., Lewandoski, M., Yamaguchi, T.P., 2008. Wnt3a/beta-catenin signaling controls posterior body development by coordinating mesoderm formation and segmentation. Development 135, 85–94. Favata, M.F., Horiuchi, K.Y., Manos, E.J., Daulerio, A.J., Stradley, D.A., Feeser, W.S., Van Dyk, D.E., Pitts, W.J., Earl, R.A., Hobbs, F., Copeland, R.A., Magolda, R.L., Scherle, P.A., Trzaskos, J.M., 1998. Identification of a novel inhibitor of mitogen-activated protein kinase kinase. J. Biol. Chem. 273, 18623–18632. Finkielsztein, A., Kelly, G.M., 2009. Altering PI3K-Akt Signalling in Zebrafish Embryos Affects PTEN Phosphorylation and Gastrulation. Biology of the Cell/ Under the Auspices of the European Cell Biology Organization 101, pp. 661– 678664 p following 678. Flowers, G.P., Topczewska, J.M., Topczewski, J., 2012. A zebrafish Notum homolog specifically blocks the Wnt/beta-catenin signaling pathway. Development 139, 2416–2425. Furthauer, M., Lin, W., Ang, S.L., Thisse, B., Thisse, C., 2002. Sef is a feedbackinduced antagonist of Ras/MAPK-mediated FGF signalling. Nat. Cell. Biol. 4, 170–174.

M.J. Stulberg et al. / Developmental Biology 369 (2012) 298–307

Furthauer, M., Reifers, F., Brand, M., Thisse, B., Thisse, C., 2001. sprouty4 acts in vivo as a feedback-induced antagonist of FGF signaling in zebrafish. Development 128, 2175–2186. Goentoro, L., Kirschner, M.W., 2009. Evidence that fold-change, and not absolute level, of beta-catenin dictates Wnt signaling. Mol. Cell. 36, 872–884. Goentoro, L., Shoval, O., Kirschner, M.W., Alon, U., 2009. The incoherent feedforward loop can provide fold-change detection in gene regulation. Mol. Cell. 36, 894–899. Goering, L.M., Hoshijima, K., Hug, B., Bisgrove, B., Kispert, A., Grunwald, D.J., 2003. An interacting network of T-box genes directs gene expression and fate in the zebrafish mesoderm. Proc. Nat. Acad. Sci. U.S.A. 100, 9410–9415. Greco, T.L., Takada, S., Newhouse, M.M., McMahon, J.A., McMahon, A.P., Camper, S.A., 1996. Analysis of the vestigial tail mutation demonstrates that Wnt-3a gene dosage regulates mouse axial development. Genes Dev. 10, 313–324. Griffin, K.J., Amacher, S.L., Kimmel, C.B., Kimelman, D., 1998. Molecular identification of spadetail: regulation of zebrafish trunk and tail mesoderm formation by T-box genes. Development 125, 3379–3388. Griffin, K.J., Kimelman, D., 2002. One-Eyed Pinhead and Spadetail are essential for heart and somite formation. Nat. Cell. Biol. 4, 821–825. Griffin, K.J., Kimelman, D., 2003. Interplay between FGF, one-eyed pinhead, and T-box transcription factors during zebrafish posterior development. Dev. Biol. 264, 456–466. Griffith, C.M., Wiley, M.J., Sanders, E.J., 1992. The vertebrate tail bud: three germ layers from one tissue. Anat. embryol. 185, 101–113. Hardy, K.M., Yatskievych, T.A., Konieczka, J., Bobbs, A.S., Antin, P.B., 2011. FGF signalling through RAS/MAPK and PI3K pathways regulates cell movement and gene expression in the chicken primitive streak without affecting E-cadherin expression. BMC Dev. Biol. 11, 20. Hashimoto, H., Itoh, M., Yamanaka, Y., Yamashita, S., Shimizu, T., Solnica-Krezel, L., Hibi, M., Hirano, T., 2000. Zebrafish Dkk1 functions in forebrain specification and axial mesendoderm formation. Dev. Biol. 217, 138–152. Hashimoto, M., Sagara, Y., Langford, D., Everall, I.P., Mallory, M., Everson, A., Digicaylioglu, M., Masliah, E., 2002. Fibroblast growth factor 1 regulates signaling via the glycogen synthase kinase-3beta pathway. Implications for neuroprotection. J. Biol. Chem. 277, 32985–32991. Hawkins, T.A., Cavodeassi, F., Erdelyi, F., Szabo, G., Lele, Z., 2008. The small molecule Mek1/2 inhibitor U0126 disrupts the chordamesoderm to notochord transition in zebrafish. BMC Dev. Biol. 8, 42. Holmdahl, D.E., 1925. Experimentelle Untersuchungen uber die Lage der Grenze primarer und sekundarer Korperentwicklung beim Huhn. Anat. Anz. 59, 393–396. Huang da, W., Sherman, B.T., Lempicki, R.A., 2009a. Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists. Nucleic Acids Res. 37, 1–13. Huang da, W., Sherman, B.T., Lempicki, R.A., 2009b. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat. Protoc. 4, 44–57. Jho, E.H., Zhang, T., Domon, C., Joo, C.K., Freund, J.N., Costantini, F., 2002. Wnt/betacatenin/Tcf signaling induces the transcription of Axin2, a negative regulator of the signaling pathway. Mol. Cell. Biol. 22, 1172–1183. Julich, D., Hwee Lim, C., Round, J., Nicolaije, C., Schroeder, J., Davies, A., Geisler, R., Lewis, J., Jiang, Y.J., Holley, S.A., 2005. Beamter/deltaC and the role of Notch ligands in the zebrafish somite segmentation, hindbrain neurogenesis and hypochord differentiation. Dev. Biol. 286, 391–404. Kanki, J.P., Ho, R.K., 1997. The development of the posterior body in zebrafish. Development 124, 881–893. Keegan, B.R., Feldman, J.L., Lee, D.H., Koos, D.S., Ho, R.K., Stainier, D.Y., Yelon, D., 2002. The elongation factors Pandora/Spt6 and Foggy/Spt5 promote transcription in the zebrafish embryo. Development 129, 1623–1632. Kim, T.W., Michniewicz, M., Bergmann, D.C., Wang, Z.Y., 2012. Brassinosteroid regulates stomatal development by GSK3-mediated inhibition of a MAPK pathway. Nature 482, 419–422. Kimelman, D., 2006. Mesoderm induction: from caps to chips. Nat. Rev. Genet 7, 360–372. Klein, P.S., Melton, D.A., 1996. A molecular mechanism for the effect of lithium on development. Proc. Nat. Acad. Sci. U.S.A. 93, 8455–8459. Lee, M.Y., Lim, H.W., Lee, S.H., Han, H.J., 2009. Smad, PI3K/Akt, and Wnt-dependent signaling pathways are involved in BMP-4-induced ESC self-renewal. Stem Cells (Dayton, Ohio) 27, 1858–1868. Mara, A., Schroeder, J., Chalouni, C., Holley, S.A., 2007. Priming, initiation and synchronization of the segmentation clock by deltaD and deltaC. Nat. Cell. Biol. 9, 523–530.

307

Marlow, F., Gonzalez, E.M., Yin, C., Rojo, C., Solnica-Krezel, L., 2004. No tail cooperates with non-canonical Wnt signaling to regulate posterior body morphogenesis in zebrafish. Development 131, 203–216. Martin, B.L., Kimelman, D., 2008. Regulation of canonical Wnt signaling by Brachyury is essential for posterior mesoderm formation. Dev. Cell. 15, 121–133. Martin, B.L., Kimelman, D., 2012. Canonical Wnt signaling dynamically controls multiple stem cell fate decisions during vertebrate body formation. Dev. Cell. 22, 223–232. Mohammadi, M., McMahon, G., Sun, L., Tang, C., Hirth, P., Yeh, B.K., Hubbard, S.R., Schlessinger, J., 1997. Structures of the tyrosine kinase domain of fibroblast growth factor receptor in complex with inhibitors. Science 276, 955–960. Naiche, L.A., Holder, N., Lewandoski, M., 2011. FGF4 and FGF8 comprise the wavefront activity that controls somitogenesis. Proc. Nat. Acad. Sci. U.S.A. 108, 4018–4023. Niwa, Y., Masamizu, Y., Liu, T., Nakayama, R., Deng, C.X., Kageyama, R., 2007. The initiation and propagation of Hes7 oscillation are cooperatively regulated by Fgf and notch signaling in the somite segmentation clock. Dev. Cell. 13, 298–304. Ozbudak, E.M., Tassy, O., Pourquie, O., 2010. Spatiotemporal compartmentalization of key physiological processes during muscle precursor differentiation. Proc. Nat. Acad. Sci. U.S.A. 107, 4224–4229. Pourquie, O., 2011. Vertebrate segmentation: from cyclic gene networks to scoliosis. Cell 145, 650–663. Qiao, M., Sheng, S., Pardee, A.B., 2008. Metastasis and AKT activation. Cell Cycle (Georgetown, Tex.) 7, 2991–2996. Row, R.H., Kimelman, D., 2009. Bmp inhibition is necessary for post-gastrulation patterning and morphogenesis of the zebrafish tailbud. Dev. Biol. 329, 55–63. Sawada, A., Shinya, M., Jiang, Y.J., Kawakami, A., Kuroiwa, A., Takeda, H., 2001. Fgf/ MAPK signalling is a crucial positional cue in somite boundary formation. Development 128, 4873–4880. Shimizu, T., Bae, Y.K., Muraoka, O., Hibi, M., 2005. Interaction of Wnt and caudalrelated genes in zebrafish posterior body formation. Dev. Biol. 279, 125–141. Stambolic, V., Ruel, L., Woodgett, J.R., 1996. Lithium inhibits glycogen synthase kinase-3 activity and mimics wingless signalling in intact cells. Curr. Biol. 6, 1664–1668. Szeto, D.P., Kimelman, D., 2004. Combinatorial gene regulation by Bmp and Wnt in zebrafish posterior mesoderm formation. Development 131, 3751–3760. Takada, S., Stark, K.L., Shea, M.J., Vassileva, G., McMahon, J.A., McMahon, A.P., 1994. Wnt-3a regulates somite and tailbud formation in the mouse embryo. Genes Dev. 8, 174–189. Takemoto, T., Uchikawa, M., Yoshida, M., Bell, D.M., Lovell-Badge, R., Papaioannou, V.E., Kondoh, H., 2011. Tbx6-dependent Sox2 regulation determines neural or mesodermal fate in axial stem cells. Nature 470, 394–398. Tsang, M., Friesel, R., Kudoh, T., Dawid, I.B., 2002. Identification of Sef, a novel modulator of FGF signalling. Nat. Cell. Biol. 4, 165–169. Tzouanacou, E., Wegener, A., Wymeersch, F.J., Wilson, V., Nicolas, J.F., 2009. Redefining the progression of lineage segregations during mammalian embryogenesis by clonal analysis. Dev. Cell. 17, 365–376. Vlahos, C.J., Matter, W.F., Hui, K.Y., Brown, R.F., 1994. A specific inhibitor of phosphatidylinositol 3-kinase, 2-(4-morpholinyl)-8-phenyl-4H-1-benzopyran-4-one (LY294002). J. Biol. Chem. 269, 5241–5248. von Maltzahn, J., Bentzinger, C.F., Rudnicki, M.A., 2012. Wnt7a-Fzd7 signalling directly activates the Akt/mTOR anabolic growth pathway in skeletal muscle. Nat. Cell. Biol. 14, 186–191. Wahl, M.B., Deng, C., Lewandoski, M., Pourquie, O., 2007. FGF signaling acts upstream of the NOTCH and WNT signaling pathways to control segmentation clock oscillations in mouse somitogenesis. Development 134, 4033–4041. Wilson, V., Olivera-Martinez, I., Storey, K.G., 2009. Stem cells, signals and vertebrate body axis extension. Development 136, 1591–1604. Yang, X., Dormann, D., Munsterberg, A.E., Weijer, C.J., 2002. Cell movement patterns during gastrulation in the chick are controlled by positive and negative chemotaxis mediated by FGF4 and FGF8. Dev. Cell. 3, 425–437. Yoo, S.K., Deng, Q., Cavnar, P.J., Wu, Y.I., Hahn, K.M., Huttenlocher, A., 2010. Differential regulation of protrusion and polarity by PI3K during neutrophil motility in live zebrafish. Dev. Cell. 18, 226–236. Zamir, E.A., Czirok, A., Cui, C., Little, C.D., Rongish, B.J., 2006. Mesodermal cell displacements during avian gastrulation are due to both individual cell-autonomous and convective tissue movements. Proc. Nat. Acad. Sci. U.S.A. 103, 19806–19811.