• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Dev Biol. Author manuscript; available in PMC Apr 15, 2012.
Published in final edited form as:
PMCID: PMC3065516
NIHMSID: NIHMS271126

Siamois and Twin are redundant and essential in formation of the Spemann organizer

Abstract

The Spemann organizer is an essential signaling center in Xenopus germ layer patterning and axis formation. Organizer formation occurs in dorsal blastomeres receiving both maternal Wnt and zygotic Nodal signals. In response to stabilized βcatenin, dorsal blastomeres express the closely related transcriptional activators, Siamois (Sia) and Twin (Twn), members of the paired homeobox family. Sia and Twn induce organizer formation and expression of organizer-specific genes, including Goosecoid (Gsc). In spite of the similarity of Sia and Twn sequence and expression pattern, it is unclear whether these factors function equivalently in promoter binding and subsequent transcriptional activation, or if Sia and Twn are required for all aspects of organizer function. Here we report that Sia and Twn activate Gsc transcription by directly binding to a conserved P3 site within the Wnt-responsive proximal element of the Gsc promoter. Sia and Twn form homodimers and heterodimers by direct homeodomain interaction and dimer forms are indistinguishable in both DNA-binding and activation function. Sequential chromatin immunoprecipitation reveals that the endogenous Gsc promoter can be occupied by either Sia or Twn homodimers or Sia-Twn heterodimers. Knockdown of Sia and Twn together, but not individually, results in a failure of organizer gene expression and a disruption of axis formation, consistent with a redundant role for Sia and Twn in organizer formation. Furthermore, simultaneous knockdown of Sia and Twn blocks axis induction in response to ectopic Wnt signaling, demonstrating an essential role for Sia and Twn in mediating the transcriptional response to the maternal Wnt pathway. The results demonstrate the functional redundancy of Sia and Twn and their essential role in direct transcriptional responses necessary for Spemann organizer formation.

Keywords: Siamois, Twin, Goosecoid, Organizer, Transcription, Wnt, Homeodomain

Introduction

Vertebrate axial development is dependent on the correct formation and function of the dorsal signaling center known as the Spemann organizer (reviewed in Harland and Gerhart, 1997). Spemann organizer function is essential for the dorsoventral and anteroposterior patterning of the embryonic germ layers that serves as a foundation for subsequent axial development (reviewed in Harland and Gerhart, 1997). The organizer is a source of multiple negative regulatory factors, including the secreted antagonists Cerberus, Chordin, and Noggin, and transcriptional repressors such as Goosecoid (Gsc), which act to silence or moderate the activity of TGFβ and Wnt signals within the organizer and adjacent domains (reviewed in De Robertis, 2006). The combined action of these antagonists and repressors establishes signaling gradients and boundaries that confer spatial pattern in the gastrula and organize the embryonic axes during gastrulation (reviewed in De Robertis, 2006).

The organizer forms in response to the combined action of two distinct signaling inputs, the Wnt and Nodal signaling pathways (Harland and Gerhart, 1997). Shortly after fertilization, dorsal determinants localized to the vegetal hemisphere of the embryo are translocated, in a microtubule dependent manner, to the future dorsal side of the embryo (Heasman, 2006). These dorsal determinants likely include components of the Wnt signaling pathway, such as Wnt11 and LRP6, leading to localized stabilization of βcatenin in a dorsal domain (Kofron et al., 2007; Tao et al., 2005). The maternal Wnt pathway directly activates transcription of Siamois (Sia) and Twin (Twn), closely related paired-type homeodomain proteins, which function as transcriptional activators and zygotic effectors of maternal Wnt signaling (Brannon et al., 1997; Brannon and Kimelman, 1996; Carnac et al., 1996; Crease et al., 1998; Fan et al., 1998; Kodjabachian and Lemaire, 2004; Laurent et al., 1997; Nelson and Gumbiner, 1998; Nishita et al., 2000).

Sia and Twn were identified in functional screens for factors capable of mimicking the developmental activity of the Spemann organizer (Kodjabachian and Lemaire, 2004; Laurent et al., 1997; Lemaire et al., 1995). Targeted ventral expression of Sia or Twn induces ectopic organizer gene expression, as well as the formation of a complete secondary axis consisting of head, trunk and tail tissues (Laurent et al., 1997; Lemaire et al., 1995). The expression profiles of Sia and Twn are identical, both temporally and spatially, and the onset of expression in dorsal blastomeres at the mid-blastula transition, just prior to the initiation of organizer gene expression, is consistent with a significant role for Sia and Twn in activating organizer gene transcription (Laurent et al., 1997; Lemaire et al., 1995). With near identity within the paired-type homeodomains, mediating DNA-binding and target selection, Sia and Twn likely share the same targets for transcriptional activation (Laurent et al., 1997). Given these similarities in expression and DNA-binding domains, it was suggested that Sia and Twn may function as redundant or cooperative regulatory factors in activation of organizer gene expression (Laurent et al., 1997).

Expression of a dominant repressive form of Sia, a fusion of the Engrailed repressor domain with the Sia homeodomain (Eng-Sia), in the dorsal domain of the gastrula results in a complete suppression of organizer gene expression and axis formation, demonstrating that Sia and/or Sia-related proteins are essential for organizer formation (Fan and Sokol, 1997; Kessler, 1997). However, recent knockdown analysis suggests that Sia and Twn are necessary only for anterior axial development (Ishibashi et al., 2008). Injection of a mixture of morpholino antisense oligonucleotides specific for Sia and Twn resulted in a loss of head structures, but trunk and tail development was normal (Ishibashi et al., 2008), suggesting that Sia and Twn are required for head organizer function, but not for the full activity of the Spemann organizer. So while the gain-of-function and dominant repressor studies suggest that Sia and Twn confer full organizer activity (head and trunk organizer) (Kodjabachian and Lemaire, 2001; Laurent et al., 1997; Lemaire et al., 1995), the knockdown studies suggest a role limited to anterior development (head organizer) (Ishibashi et al., 2008). These apparent differences could reflect off-target effects resulting from overexpression of Sia, Twn and Eng-Sia. Alternatively, the knockdown phenotype could represent a partial loss-of-function for endogenous Sia and Twn. Given these contrasting results, further analysis is necessary to define the developmental requirement for Sia and Twn in organizer formation and function.

Sia and Twn are likely direct transcriptional regulators of multiple organizer genes. Sia has been shown to cooperate with Xlim1, Xotx2 and Mix.1 in the direct regulation of Cerberus, and both Sia and Twn directly activate Gsc (Fan and Sokol, 1997; Kessler, 1997; Laurent et al., 1997; Yamamoto et al., 2003). Gsc is expressed specifically within the organizer domain (Blumberg et al., 1991; Cho et al., 1991; De Robertis, 2004) where it functions as a transcriptional repressor to suppress Wnt and BMP signaling and maintain organizer identity (Sander et al., 2007; Yao and Kessler, 2001). The Gsc promoter contains a distal element (DE) responsive to TGFβ signals and a proximal element (PE) responsive to Wnt signals (Watabe et al., 1995). These two response elements are found in close proximity within the Gsc promoter and are conserved in all vertebrate Gsc genes (Fig. 1A). Previous studies have shown that the Wnt-responsive PE is necessary for Sia and Twn-mediated activation of a Gsc reporter construct (Fan and Sokol, 1997; Kessler, 1997; Laurent et al., 1997; Yao and Kessler, 2001), and in vitro experiments have revealed that Twn binds to a conserved region within the PE (Laurent et al., 1997). The similarities in the structure, expression and function of Sia and Twn suggest that these proteins likely bind the same sequence in the Gsc promoter to activate transcription. However, it is unknown whether Sia and Twn contribute equivalently to the activation of Gsc expression. Furthermore, it remains unclear whether Sia and Twn function in an entirely redundant manner in organizer formation, and whether these factors are required for the complete function of the Spemann organizer. Further analysis of the regulation of Gsc and other organizer genes by Sia and Twn would provide insight to the developmental and transcriptional mechanisms of organizer formation.

Fig. 1
Siamois and Twin bind an identical conserved region within the Gsc Proximal Element. (A) Schematic of the Gsc promoter indicating sequence conservation within the Proximal Element (PE) across species. The P3 element and upstream half site are indicated ...

We sought to address these questions, first by defining the conserved sequences within the Gsc PE that are required for stable binding of Sia and Twn and consequent transcriptional activation of Gsc. In protein interaction assays Sia and Twn form both homo- and heterodimers through direct protein-protein interactions, and we found that the different dimer forms are indistinguishable in both DNA-binding and transcriptional activation function. In vivo, Sia and Twn can together occupy the endogenous Gsc promoter, consistent with both homo- and heterodimer formation at the Gsc promoter. Knockdown of both Sia and Twn together, but not individually, results in a loss of organizer gene expression and a complete disruption of axis formation. Furthermore, we confirm the prediction that Sia and Twn together are required downstream of the Wnt signaling pathway in axis formation. The results demonstrate the functional redundancy of Sia and Twn and their essential role in direct activation of organizer gene expression and regulation of Spemann organizer formation.

Materials and methods

Embryo manipulation and microinjection

Xenopus embryos were collected, fertilized, injected and cultured as previously described (Yao and Kessler, 2001). Embryonic stage was determined according to Nieuwkoop and Faber (Nieuwkoop and Faber, 1967). Ectopic axis induction was scored at the neurula stage as partial axis induction (containing trunk but no head structures) or complete (containing trunk and head structures). Results represent at least five independent experiments. Explants were prepared using a Gastromaster microsurgery instrument (Xenotek Engineering). Capped, in vitro transcribed mRNA for microinjection was synthesized from linearized DNA templates using the SP6 mMessage Machine kit (Ambion); 10nl of RNA solution was injected per embryo. Templates for in vitro transcription were pCS2+Siamois (Kessler, 1997), pSP64-Twin (Laurent et al., 1997), pCS2+myc-Twin (this study), pCS2+myc-Siamois (this study), pCS2+myc-SiaQ191E (Kessler, 1997) this study), pCS2+GST-Sia (this study), pCS2+GST-Twn (this study), and pCS2+XWnt8 (Kessler, 1997).

Plasmid constructs

pCS2+myc-Sia and pCS2+myc-SiaQ191E were generated by PCR amplification of the coding region of Sia or SiaQ191E (Kessler, 1997). The amplified products were subcloned into the BamHI site of pCS2+myc. For pCS2+myc-Twn, the coding region of Twn (Laurent et al., 1997) was amplified by PCR and cloned into the EcoRI site of either pCS2+ or pCS2+myc. pCS2+GST-Sia and pCS2+GST-Twn were generated by subcloning the coding regions of Sia or Twn into the XbaI site of pCS2+GST (Yaklichkin et al., 2007). All constructs were verified by sequencing and in vitro translation assays. For DNAse footprinting, a plasmid containing the – 226Gsc promoter (Watabe et al., 1995) was digested with BamHI and HindIII and subcloned into pBSII-KS+ to make pBS-226Gsc. pBS-226Gsc was digested with BamHI and HincII for bottom strand labeling, and HindIII and SacII for top strand labeling. For preparation of tagged recombinant proteins, 6xHis- or GST-tagged Sia and Twn were amplified by PCR and subcloned into the pet28b or pGEX vectors, respectively. Reporter constructs with mutations in the Gsc promoter sequence were generated by PCR-mediated mutagenesis. Specific mutations introduced into the Gsc promoter are indicated in Fig. 2A.

Fig. 2
The Gsc P3 element is required for stable binding and transactivation by Sia and Twn. (A) Sequence of oligonucleotide probes used in EMSA experiments with the P3 element and upstream half site indicated with gray shading. Mutated nucleotides are indicated ...

Protein purification, pulldown and crosslinking

Histidine-tagged and GST-tagged proteins were purified using standard methods (Novagen and Pharmacia Biotech). The in vitro GST pulldown assay was performed as previously described (Yaklichkin et al., 2007). GST or GST-Siamois (2μg) were incubated with full length His-Sia or His-Twn (2μg), protein complexes were recovered using Glutathione Sepharose 4B (GE Healthcare, 17-0756-01) and subjected to western analysis using an anti-6X His tag antibody (AbCam). For the protein crosslinking studies, EGS (Ethylene Glycol-bis (succinic acid N-hydroxysuccinimide ester) (Sigma, E3257) dissolved in DMSO was added to each protein sample and incubated for 30 minutes at room temperature. DMSO alone was used for control reactions. The crosslinking reaction was stopped by addition of glycine to a final concentration of 75mM. Crosslinking of proteins in the presence of the DNA-binding site was performed in a similar manner by incubating oligonucleotides with proteins for 20 minutes on ice prior to addition of EGS. Crosslinked protein complexes were detected by western analysis using an anti-His tag antibody.

EMSA and DNase footprinting

Electrophoretic mobility shift assay (EMSA) was performed according to manufacturer's instructions (Promega Gel Shift Assay System). Full length Sia protein-DNA complexes were resolved on a 5% native polyacrylamide gel in 0.25X Tris-Borate-EDTA buffer for one hour at 240V. Sia and Twn homeodomain (HD) fragments, complexes were resolved on an 8% native polyacrylamide gel. Stability of protein-DNA complexes for wild-type and mutated probes was determined by addition of a 100-fold molar excess of cold unlabeled wild-type oligonucleotide as a competitor after the initial binding reaction. The bound complex was collected at specific time points, resolved by EMSA, and protein-DNA complex formation was quantified using the ImageQuant program (Molecular Dynamics). For heterodimerization of Sia and Twn when bound to DNA, EMSA was performed with increasing concentrations of His-Sia112-215 and constant concentration of Twn HD. DNase footprinting was performed according to standard procedures (Brenowitz et al., 2001). End labeled DNA was incubated with 0.5 – 2.0μg recombinant Sia or Twn protein. Upon completion of DNase cleavage, DNA was extracted with phenol/chloroform, ethanol precipitated and radiolabelled DNA fragments were resolved on a 6% denaturing polyacrylamide gel.

Luciferase reporter assay

One-cell stage Xenopus embryos were injected in the animal pole with in vitro transcribed mRNA encoding Sia or Twin. At the two-cell stage, one blastomere was injected with 100pg of pGL3-Gsc-Luciferase containing the wild-type or mutated -226Gsc promoter in combination with 10pg of pGL3-CMV-Renilla as an internal control (Renilla luciferase under the control of the constitutive CMV promoter) (Kessler, 1997). Animal pole explants prepared at the blastula stage were collected at midgastrula stage and luciferase activity was determined using the Dual Luciferase Assay Kit (Promega) on a TD-20/20 luminometer (Turner Designs). The data presented are the results of at least five independent experiments, with error bars representing standard error.

Chromatin immunoprecipitation

Chromatin immunoprecipitation (ChIP) was performed as described (Blythe et al., 2009). One-cell embryos were injected with 50pg of either myc-Sia or myc-Twn mRNA. An average of 50 embryos were collected at stage 10.25 and processed for ChIP. Sequential chromatin immunoprecipitation was performed as described (Geisberg and Struhl, 2004) with two immunoprecipitations using polyclonal anti-myc antibody (Millipore, 06-549) and anti-GST antibody (GE Lifesciences, 27-4577-01). Briefly, 150pg of mRNA encoding differentially tagged (either GST or myc) Sia or Twn was injected into one-cell embryos. An average of 75 embryos was collected at stage 10.25 and processed for ChIP. The eluate from the first immunoprecipitation was subdivided, with half processed for ChIP and half used for the second immunoprecipitation. The second immunoprecipitation was performed by adding 1.4ml of RIPA buffer to 100μl of eluate, and addition of the second antibody according to the ChIP protocol. Quantitative PCR was performed using primers specific for Gsc, Ef1α or Xmlc2 as previously described (Blythe et al., 2009).

In situ hybridization and histology

For whole mount in situ hybridization, embryos were fixed and hybridized with antisense digoxygenin-labeled RNA probes as described (Sive et al., 2000). Hybridized probe was detected using alkaline phosphatase-conjugated anti-digoxygenin Fab fragments (Roche) and BMpurple (Roche) as a substrate for color development. Antisense probes were synthesized from linearized plasmid DNA using the Megascript kit (Ambion) supplemented with 2mM digoxygenin-11-UTP (Roche). Templates for in situ probes were pCS2+Gsc (Yao and Kessler, 2001), pCS2+Chd (Sasai et al., 1994), pGEM-Xbra (Wilson and Melton, 1994), pBS-Opl (Kuo et al., 1998), and pGEM-XWnt8 (Sokol et al., 1991). For histology, 10μm sections were prepared from paraplast-embedded embryos and dewaxed sections were stained with Hematoxylin/Eosin before coverslipping with Permount as previously described (Sive et al., 2000).

Morpholino oligonucleotides

The Sia and Twn morpholino antisense oligonucleotides (Sia MO and Twn MO) are complementary to nucleotides of 1-25 of Xenopus Sia (5’-GCTCCATTTCAGCCTCATAGGTCAT -3’) and nucleotides 1-25 of Xenopus Twin (5’-GCTCAAGTTCAGAGTCACAAGTCAT-3’) (Gene Tools). Individual or mixed oligonucleotides were injected at a total dose of 50ng per embryo. As a control, embryos were injected with equal doses of the standard control morpholino (5'-CCTCTTACCTCAGTTACAATTTATA-3') (Gene Tools).

Results

Siamois and Twin bind identical sequences in the Goosecoid proximal element

Sia and Twn have each been identified as direct regulators of Gsc expression in previous studies (Fan and Sokol, 1997; Kessler, 1997; Laurent et al., 1997). Sia and Twn share high homology, especially within the third helix of the homeodomain, which is predicted to be the region of the DNA binding domain that imparts specific recognition of target DNA sequences (Laurent et al., 1997; Wilson et al., 1995). Previous biochemical studies indicated that Twn binds to the PE of the Gsc promoter at a sequence that contains two consensus homeodomain binding half sites (Laurent et al., 1997). As paired-type homeodomain proteins, Sia and Twn are predicted to bind preferentially to P3 sites, consisting of two inverted TAAT motifs separated by 3 base pairs (Wilson et al., 1995). Examination of the Xenopus Gsc PE reveals a near perfect consensus P3 site (-129 to -119) with an additional upstream half site (-136 to -133) (Fig. 1A). Alignment of Gsc promoter sequences of Xenopus laevis, human, mouse and zebrafish reveals a striking conservation of the P3 site and the upstream half site (Fig. 1A). The presence of this conserved P3 site within the Gsc PE suggests a role for paired-type homeodomain proteins in Gsc transcriptional regulation across species. We sought to investigate whether this site plays a role in mediating the transcriptional response to Sia and Twn in Xenopus.

To precisely map the region bound by Sia and Twn within the Gsc promoter, DNase footprinting was performed. A fragment of the Gsc promoter (-226 to +1) (Watabe et al., 1995) was labeled either on the top or bottom strand, incubated with full length Sia protein, Sia homeodomain (HD), Twn HD or a mixture of Sia HD and Twn HD, and subjected to DNase1 digestion to identify the regions bound and protected. A nearly identical region, containing the conserved P3 site and upstream half site, was protected on both the top (-146 to -115) and bottom strands (-145 to -115) (Fig. 1B-D). Sia HD, Twn HD or a mixture of Sia HD and Twn HD protected the same area as full-length Sia, suggesting that the homeodomain alone is sufficient to confer specific binding to the Gsc promoter (Fig. 1B,C). These results are consistent with previous footprinting analysis with the Twn homeodomain, which showed a protection of -114 to -127 within the Gsc PE (Laurent et al., 1997). Two minor protected regions (-103 to -93 and -15bp to +1bp) were detected as well (Fig. 1 B,C), but these did not contain apparent homeodomain binding sites and may be either non-specific or cryptic homeodomain binding sites. These results demonstrate that Sia and Twn bind to and protect an identical region of the Gsc promoter, which includes a conserved P3 site and upstream half site. The near identity of the Sia and Twn homeodomains predicts that Sia and Twn likely share transcriptional targets (Laurent et al., 1997); our results suggest that Sia and Twn regulate Gsc transcription by binding to the highly conserved P3 site within the Gsc promoter.

Siamois and Twin binding to the Goosecoid promoter is dependent on conserved homeodomain binding sites

To determine whether the conserved P3 site and upstream half site are required for Sia binding to the Gsc promoter, electrophoretic mobility shift assays (EMSA) were performed using a double-stranded oligonucleotide probe containing the region protected by Sia and Twn (-146 to -115, referred to as wild-type or WT probe) (Fig. 2A). When bound by full-length recombinant Sia protein, the WT probe formed two distinct protein-DNA complexes, a higher mobility complex and a lower mobility complex (Fig. 2B). Formation of the higher mobility complex was seen at lower protein concentrations, whereas the lower mobility complex was observed only at higher protein concentrations. Paired-type homeodomain proteins are known to dimerize at higher protein concentration (Wilson et al., 1993), suggesting that the higher mobility complex represents the binding of a Sia monomer to one half site, while the lower mobility complex results from formation of a Sia dimer at the P3 site. Consistent with this idea, palindromic P3 sites have been shown to be occupied by two paired-type homeodomain proteins in a high affinity complex (Wilson et al., 1993), which would suggest that Sia and Twn might both occupy the Gsc promoter to regulate transcription.

To assess the contribution of the upstream half site and P3 site to Sia binding, complex formation was assessed for probes with mutations introduced into the upstream half site (136 MT), the P3 site (127 MT) or both the upstream half site and P3 site (2X MT) (Fig. 2A). Sia binding was unaffected by mutation of the upstream half site (136 MT), and both monomer and dimer complexes formed at near identical protein concentrations as observed for the WT probe (Fig. 2B). To disrupt the P3 site, one of the half sites comprising the P3 site was mutated (127 MT), and this resulted in a dramatic reduction of complex formation (Fig. 2B). Only at the highest concentrations of Sia protein were monomeric and dimeric complexes detected, but at greatly reduced levels compared to the WT probe (Fig. 2B). The continued presence of both the monomer and dimer complexes may reflect low affinity binding of Sia to the two half sites still present in the probe. When both the P3 site and the upstream half site were mutated (2X MT), only a monomeric complex was weakly observed at the highest concentrations of Sia protein (Fig. 2B), likely due to Sia binding to the single remaining half site. When all three half sites were mutated (3X MT), no binding of Sia was observed, even at the highest protein concentration (data not shown). Taken together, these results demonstrate that high affinity binding of Sia to the Gsc PE is dependent on the conserved P3 site.

To further assess the sequence requirements for stable binding of Sia to the Gsc promoter, EMSA competition assays were performed. Binding of Sia protein to radiolabeled probe (WT, 136 MT, 127 MT or 2X MT) was allowed to reach equilibrium (20 min), a 100-fold molar excess of unlabeled WT probe was then added, and the resulting levels of Sia-DNA complex were assessed at 5, 10, 20 or 30 min after competitor addition (Fig. 2C). As expected, Sia binding to the WT probe formed a stable complex with ~50% of the complex still intact 30 min after competitor addition (Fig. 2C). Sia binding to a probe mutated for the upstream half site (136 MT) was nearly as stable as WT, while mutation of the P3 site (127 MT) or both sites (2X MT) resulted in an unstable complex that was fully dissociated within 5 min of competitor addition (Fig. 2C). The extent of complex dissociation following competitor addition suggests that the P3 site, but not the upstream half site, is essential for stable binding of Sia to the Gsc promoter. To assess whether Sia HD or Twn HD is sufficient for complex formation at the Gsc promoter, as suggested by the DNase footprinting results (Fig. 1), we tested the ability of recombinant Sia HD or Twn HD to bind to the WT and mutant probes (see Fig. S1 in supplementary materials). When compared to the results with full-length Sia, no differences in complex formation or sequence requirements were observed for the Sia HD (Fig. S1A) or Twn HD (Fig. S1B) alone, suggesting that the homeodomain confers the complete binding activity of the full-length protein. In addition, the formation of apparent dimeric protein-DNA complexes by the homeodomains alone suggests that dimer formation for Sia and Twn may be mediated by direct homeodomain interactions. Taken together, these results indicate that Sia and Twn have identical sequence requirements for binding to the Gsc PE, and that the conserved P3 site is required for stable dimeric complex formation.

Siamois and Twin activation of Goosecoid transcription is dependent on conserved homeodomain binding sites

To determine if the conserved homeodomain binding sites required for Sia and Twn complex formation at the Gsc promoter are also required for transcriptional activation of the Gsc promoter, transcriptional reporters containing either the wild-type Gsc promoter (-226 to +1) or the mutated forms described above (Fig. 2A) were tested in vivo. Xenopus embryos were injected at the one-cell stage with either Sia or Twn mRNA, a Gsc-luciferase reporter and an internal control renilla reporter (Fig. 2D). As expected, Sia or Twn strongly activates the wild-type Gsc promoter (~10-fold activation) and no significant difference between Sia and Twn transcriptional activity was observed (Fig. 2D). However, mutation of the upstream half site (136 MT) caused an ~60% reduction in reporter activity in response to both Sia and Twn, suggesting that this half site, which has a marginal effect on Sia and Twn complex formation on the PE, is required for maximal activity of the Gsc promoter in this assay (Fig. 2D). While this site might not contribute to complex formation in vitro, it does seem to contribute to transcriptional activity, perhaps by providing additional contacts for Sia and Twn, or by providing the proper DNA conformation for complex maintenance or cofactor recruitment. Mutation of the P3 site resulted in a near complete loss of transcriptional response (~2-fold activation), while mutation of two (2X MT) or all three half sites (3X MT) fully blocked the response to Sia and Twn (Fig. 2D). These results confirm the functional importance of the P3 site in mediating the transcriptional response of Gsc to Sia or Twn, but also reveal a role for the upstream half site in promoting maximal transcriptional response. Given the striking conservation of these homeodomain binding sites in other vertebrate Gsc genes, it is likely that this region of the Gsc promoter is essential for Gsc regulation in other species.

Siamois and Twin form homodimers and heterodimers that occupy the Goosecoid promoter

The ability of paired-type homeodomain proteins to dimerize (reviewed in (White, 1994), the similar expression and structure of Sia and Twn (Laurent et al., 1997), and the apparent formation of Sia and Twn dimer complexes in DNA-binding assays (Fig. 2 and Supplemental Fig. S1), suggested that Sia and Twn may form homodimer or heterodimer complexes in regulating Gsc transcription. As an initial assessment of the ability of Sia and Twn to form heterodimers, DNA-protein complexes were examined by EMSA using a mixture of recombinant Sia homeodomain and Twn homeodomain (Fig. 3A). Since the Sia and Twn homeodomains are nearly identical in length, a fragment of Sia encompassing the homeodomain and flanking sequence (Sia112-215) was used to distinguish it from the Twn HD (136-195) based on mobility (diagrammed in Fig. 3H). Sia112-215 alone or Twn HD alone each formed two distinct complexes when bound to the WT probe (Fig. 3A, lanes 2 and 7), and these correspond to predicted monomer and dimer complexes observed in the studies above (Fig. 2 and Supplementary Fig. S1). When Sia112-215 and Twn HD were combined in the DNA-binding assay, an additional complex formed that was intermediate in size to the Sia112-215 homodimer and the Twn HD homodimer, consistent with the formation of a Sia-Twn heterodimer (Fig. 3A, lanes 3-6). The results suggest that Sia and Twn can form both homodimers and heterodimers on the Gsc PE. These dimer forms are likely a result of direct protein-protein interactions, as purified His-Sia and His-Twn binds to a purified GST-Sia fusion protein (Fig. 3B, lanes 5-6), but not to GST alone (Fig. 3B, lanes 3-4). Therefore, direct and stable protein interaction, in the absence of a DNA-binding site, mediates the formation of Sia homodimers and Sia-Twn heterodimers, and the homodimers and heterodimers appear to form at equal efficiency.

Fig. 3
Siamois and Twin form homodimers and heterodimers through direct protein-protein interactions. (A) EMSA analysis of complex formation for Sia112-215 (lane 7), Twn HD (lane 2) or a combination of both proteins (lanes 3-6) bound to the WT Gsc probe. Twn ...

The DNA-binding analyses described above (Fig. 2B,C, Supplementary Fig. S1, and Fig. 3A) suggest that the homeodomain alone is sufficient for homo-and heterodimerization of Sia and Twn. To determine if the homeodomain alone is sufficient for dimerization in the absence of DNA, recombinant His-Sia and His-Twn proteins (Fig. 3H) were combined and crosslinked to stabilize protein complexes (Fig. 3C-G). The Sia HD (142-201), a larger fragment of Sia (112-215), and the Twn HD (136-195) each formed homodimers, as well as higher molecular weight complexes (Fig. 3C-E). When Sia112-215 was combined with either the Sia HD or Twn HD, intermediate sized complexes were formed that demonstrate the formation of a Sia homodimer and a Sia-Twn heterodimer (Fig. 3F,G). Taken together, these observations indicate that Sia and Twn homodimers and heterodimers can form by direct protein interactions of the homeodomain in the absence of a DNA-binding site. These results are consistent with previous structural predictions of paired type homeodomain proteins suggesting that the homeodomain can mediate both protein-protein interactions as well as DNA-protein interactions (Wilson et al., 1995). Furthermore, the results strongly predict that Sia and Twn homodimers, as well as Sia-Twn heterodimers occupy the Gsc PE to activate transcription.

To assess the occupancy of the endogenous Gsc promoter by Sia and Twn, chromatin immunoprecipitation (ChIP) in whole embryos was performed (Blythe et al., 2009). Myc-tagged Sia or myc-tagged Twn were immunoprecipitated using an anti-myc antibody and quantitative PCR was performed for either the endogenous Gsc promoter or for the Ef1α genomic locus as control (Fig. 4A). Both Sia and Twn bound robustly and specifically to the Gsc promoter (~18-fold enrichment over background) (Fig. 4A). This occupancy is dependent on the DNA-binding function of the homeodomain, as an inactivating point mutation (SiaQ191E) in the critical third helix of the homeodomain (Kessler, 1997) impairs occupancy of the Gsc promoter (Fig. 4A). As predicted, these data indicate that Sia and Twn occupy the endogenous Gsc promoter, and that this occupancy is dependent on a functional homeodomain.

Fig. 4
Siamois and Twin homodimers and heterodimers occupy the endogenous Gsc promoter. (A) Genomic regions recovered by chromatin immunoprecipitation for myc-Sia, myc-Twn or myc-SiaQ191E were evaluated by quantitative PCR (QPCR) for either the Gsc promoter ...

While the standard ChIP analysis demonstrates that Sia and Twn occupy the endogenous Gsc promoter, it cannot determine whether Sia and Twn occupy the Gsc promoter at the same time, which is predicted for Sia-Twn heterodimer formation in vivo. To assess the occupancy of the Gsc promoter by Sia and Twn homodimers and heterodimers, sequential ChIP was performed in gastrula stage embryos. Differentially tagged forms of Sia or Twn were coexpressed and sequential immunoprecipitations were carried out for each eptiope-tagged form to define the composition of the protein complex bound at the Gsc promoter. Western blot analysis confirmed equivalent levels of Sia and Twn expression in these studies (data not shown). Genomic DNA recovered in each round of immunoprecipitation was analyzed by QPCR for the Gsc promoter and the Xmlc2 genomic region as control (Fig. 4B). The sequential ChIP results are consistent with formation of both Sia-Sia homodimers and Twn-Twn homodimers. The Gsc promoter was highly enriched in sequential ChIP for either myc-Sia and GST-Sia or myc-Twn and GST-Twn, while Xmlc2 genomic sequences were not recovered (Fig. 4B). As additional controls, if GST alone was coexpressed with myc-Sia, the Gsc promoter was not recovered in GST-containing complexes (Fig. 4B). These sequential ChIP studies demonstrate that Sia and Twn homodimers can occupy the endogenous Gsc promoter.

Finally, to assess Gsc occupancy by Sia-Twn heterodimers, myc-Twn and GST-Sia were coexpressed and subjected to sequential ChIP. The Gsc promoter was robustly recovered in both rounds of immunoprecipitation (~60-fold and ~30-fold for myc-Twn and GST-Sia, respectively), consistent with occupancy of the Gsc promoter by Sia-Twn heterodimers (Fig. 4B). A similar result was obtained when coexpressing myc-Sia and GST-Twn (~50-fold and ~40-fold, respectively), further supporting the conclusion that Sia-Twn heterodimers occupy the endogenous Gsc promoter (Fig. 4B). Consistent with direct protein-protein interactions in dimer formation at the Gsc promoter, sequential ChIP of myc-Sia or myc-Twn with GST-SiaQ191E, a DNA-binding inactive mutant, also results in recovery of the Gsc promoter (data not shown). This suggests that SiaQ191E interacts directly with wild-type Sia or Twn at the Gsc promoter. Taken together, the ChIP data confirm that Gsc is a direct target of Sia and Twn, and that these factors are capable of occupying the endogenous promoter as homodimers or heterodimers.

Siamois and Twin homodimers and heterodimers have similar transcriptional and developmental function

The expression, DNA-binding, protein interaction, transcriptional and developmental analyses of Sia and Twn, presented both here and in previous studies (Kodjabachian and Lemaire, 2001; Laurent et al., 1997; Lemaire et al., 1995), suggest that Sia and Twn function equivalently within the context of all available studies. However, our demonstration that Sia-Twn heterodimers form and can occupy the endogenous Gsc promoter raises the possibility that the heterodimer complex has distinct function, and may differ from the homodimer forms in either transcriptional or developmental function. To assess the transcriptional and developmental function of Sia-Twn heterodimers, dose response analysis was performed for Sia alone, Twn alone, or the combination of Sia and Twn in a luciferase reporter assay and in an ectopic axis induction assay. The transcriptional response of the WT Gsc-luciferase reporter to increasing doses of Sia alone or Twn alone (3, 10 or 30pg mRNA) were similar, with maximal responses of ~5-fold for Sia and ~4-fold for Twn (Fig. 5A). When Sia and Twn mRNAs were combined and injected at a total dosage equal to that used for the individual factors (1.5+1.5, 5+5, or 15+15pg), a similar transcriptional dose response was observed (~3.5-fold maximal response) (Fig. 5A). These results suggest that Sia and Twn homodimers and heterodimers have similar transactivation function.

Fig. 5
Siamois and Twin homodimer and heterodimers have indistinguishable transactivation and axis induction function in vivo. (A) At the one-cell stage the animal pole was injected with Sia, Twn or a mixture of both mRNAs at the indicated doses and at the two-cell ...

Sia and Twn were originally identified based on their ability to mimic the axis-inducing activity of the Spemann organizer when ectopically expressed in ventral blastomeres of the Xenopus embryo (Laurent et al., 1997; Lemaire et al., 1995). To assess the developmental function of Sia and Twn homodimers and heterodimers, a single ventral blastomere was injected at the four-cell stage with Sia alone, Twn alone or a combination of Sia and Twn. At low dosage (1pg) Sia or Twn induced partial ectopic axes consisting of tail and trunk structures, but lacking head structures (~20% and ~30% for Sia and Twn, respectively) (Fig. 5B). At higher dosage (3 and 10pg) complete ectopic axes, including head structures, were observed at increasing frequency (~25% and ~45% for Sia and Twn at 10pg, respectively) (Fig. 5B). When Sia and Twn were injected at a combined dosage equal to the individual mRNAs (0.5+0.5, 1.5+1.5, or 5+5pg), a similar response profile for axis induction was observed. At low dose, Sia+Twn induced partial ectopic axes (~25% at 0.5+0.5pg), and with higher dosage an increasing frequency of complete ectopic axes was observed (~15% and ~45% for Sia+Twn at 1.5+1.5 and 5+5pg, respectively) (Fig. 5B). Therefore, under conditions where Sia-Twn heterodimers would likely form, no cooperative or synergistic transcriptional activity or induction of axis formation is observed, but rather the response observed is similar to that obtained with equivalent doses of Sia or Twn alone. Taken together, these results indicate that Sia and Twin homodimers and Sia-Twn heterodimers have indistinguishable function in vivo, both in their ability to activate transcription and induce axis formation.

Siamois and Twin are redundant and essential for axial development and organizer formation

In previous studies the function of Sia and Twn was disrupted either with a dominant repressive Eng-Sia fusion protein (Fan and Sokol, 1997; Kessler, 1997) or by simultaneous knockdown of Sia and Twn (Ishibashi et al., 2008). In both cases Sia and Twn were found to be essential for organizer formation and axial development, although the disruption of organizer function differs in severity for these two approaches. While Eng-Sia completely suppressed organizer and axis formation (Fan and Sokol, 1997; Kessler, 1997), the double knockdown resulted in a less severe phenotype, with loss of head, but not trunk or tail structures (Ishibashi et al., 2008). These differences could reflect off-target effects of Eng-Sia or incomplete knockdown of Sia and Twn. Despite this discrepancy in the functional analysis of Sia and Twn, our results strongly predict that Sia and Twn function equivalently and redundantly in organizer formation. To more clearly establish the requirement for Sia and Twn in organizer formation, and to assess their predicted functional redundancy, Sia and Twn were knocked down individually and in combination.

Translation-blocking morpholino oligonucleotides were designed to specifically target Sia or Twn. The specificity and efficacy of these oligonucleotides was assessed in protein translation and axis induction assays (see Fig. S2 in supplementary materials). In an in vitro translation assay, the Sia-specific morpholino oligonucleotide (MO) blocked translation of Sia, but not Twn. Conversely, the Twn MO blocked Twn translation, but not Sia (Fig. S2A). Myc-tagged forms of Sia and Twn, in which a distinct translational start site is used, were insensitive to either Sia MO or Twn MO (Fig. S2A). To assess the function blocking activity of the MOs, their ability to inhibit ectopic axis induction by Sia or Twn mRNA was examined. Injection of Sia or Twn mRNA into a single ventral blastomere at the four-cell stage resulted in induction of complete ectopic axes in most embryos (94% and 72% for Sia and Twn, respectively) (Fig. S2C,D), and axis induction was greatly reduced in the presence of the corresponding MO (28% and 6% for Sia and Twn, respectively) (Fig. S2I,M), but not with the unmatched MO (95% and 73% for Sia and Twn, respectively) (Fig. S2J,L). The axis-inducing activity of myc-Sia and myc-Twn was unaffected by either MO (insets Fig. S2I,M). Therefore, the Sia MO and Twn MO are effective and specific in blocking the translation and developmental function of Sia and Twn.

To determine the requirement for Sia and Twn in axial development and organizer formation, Sia and Twn were knocked down in the dorsal domain of the embryo, the region of their endogenous expression. At the four-cell stage, both dorsal blastomeres were injected with the Sia MO or Twn MO individually, or with a combination of both MOs, and axial development was assessed at the tailbud stage (Fig. 6A-B,E-F,I-J). Injection of each individual MO, or a control non-specific MO unrelated to Sia or Twn, had little or no effect on axial development (90-100% normal axis formation) (Fig. 6A-B,E-F,M). Embryos injected with both Sia MO and Twn MO displayed severe axial defects with loss of head structures, and reduction or loss of trunk and tail structures (Fig. 6I-J). Phenotypic severity for the double knockdown embryos ranged from complete ventralization with loss of all axial structures (DAI 0) (Fig. 6J) to loss of head with reduction of trunk and tail (DAI 1-2) (Fig. 6I), and the majority of injected embryos displayed severe axial defects (90% DAI 0-1 at highest MO dosage) (Fig. 6M) (Kao and Elinson, 1989). Histological analysis was performed to examine axial development in the single and double knockdown embryos (Fig. 6C-D,G-H,K-L). Axis formation was normal in embryos injected with the Sia MO, Twn MO or the control MO, with notochord, somitic muscle and neural tube formation indistinguishable from uninjected controls (Fig. 6C-D,G-H). Double knockdown embryos displayed axial defects ranging from partial ventralization (loss of notochord and fusion of somitic muscle across the midline) (Fig. 6I,K) to complete ventralization (loss of notochord, muscle and neural tube) (Fig. 6J,L). The severity of axial defects was dependent on the dosage of Sia and Twn MOs. At lower doses (25 ng of each MO), less severe axial defects were observed (loss of head and reduction of trunk and tail; 72% with DAI 2-4), while at higher doses of MOs (50 ng of each MO), 90% of embryos displayed a near compete loss of axial structures (DAI 0-1) (Fig. 6M). To confirm the specificity of the developmental defects observed, rescue experiments were performed (see Fig. S3 in supplementary materials). The severe axial defects observed for the double knockdown (79% axial defects) (Fig. S3B) were fully rescued by expression of either myc-Sia or myc-Twn (71% and 76% normal for myc-Sia and myc-Twn, respectively) (Fig. S3D,F). These studies demonstrate that Sia and Twn are functionally redundant and together are essential for development of head, trunk and tail structures of the body axis. We note that the severity of the axial defects observed are consistent with the Eng-Sia studies (Fan and Sokol, 1997; Kessler, 1997), but not with the previous knockdown studies (Ishibashi et al., 2008), suggesting that a more complete knockdown reveals a requirement for Sia and Twn in tail, trunk and head development. Consistent with this idea, injection of a mixture Sia MO and Twn MO at lower dosage resulted in reduction of head development with little effect on trunk and tail formation (Fig. 6M) similar to the previously reported phenotype (Ishibashi et al., 2008).

Fig. 6
Siamois and Twin function redundantly in axial development and organizer formation. (A-L) At the 4-cell stage both dorsal blastomeres were injected with (B, D) a non-specific control morpholino oligonucleotide (NSMO, 50ng), (E,G) a Sia-specific oligonucleotide ...

To establish the developmental origins of the axial defects resulting from Sia and Twn knockdown, gene expression was examined at the early gastrula stage. Knockdown of Sia or Twn individually had no effect on organizer (Gsc, Chordin), ventral mesodermal (Xwnt8), panmesodermal (Brachyury) or neural plate (Opal) gene expression (Fig. 6X-G’), as was the case for the non-specific control MO (Fig. 6S-W). In contrast, simultaneous knockdown of both Sia and Twn resulted in a near complete loss of Gsc (77% reduced or absent expression) (Fig. 6H’) and Chordin (100% reduced or absent expression) (Fig. 6I’), an expansion of Xwnt8 into the organizer domain (Fig. 6J’), a loss of Opal in the neural plate (Fig. 6K’), but no change in Brachyury expression (Fig. 6L’). These gene expression defects indicate a dramatic loss of organizer formation at the early gastrula stage, and are consistent with the severity of the axial defects observed later in development. In contrast, the direct Wnt target, Xnr3, is unaffected by Sia/Twn knockdown (data not shown). The results indicate that Sia and Twn together are essential regulators of organizer formation and subsequent axial development. Furthermore, we find that in response to loss-of-function for either Sia or Twn, the individual proteins can functionally compensate and support normal development.

Siamois and Twin are required to mediate Xwnt8-induced axis induction

Sia and Twn expression is activated in response to maternal Wnt signals at the midblastula transition, and multiple Tcf binding sites within the Sia and Twn promoters mediate direct activation by ßcatenin (Brannon et al., 1997; Brannon and Kimelman, 1996; Carnac et al., 1996; Fan et al., 1998; Laurent et al., 1997; Nelson and Gumbiner, 1998). Previous reports suggest that Sia is required downstream of both maternal Wnt signals and ßcatenin in axis induction (Fan and Sokol, 1997; Kessler, 1997), and Sia and Twn function are required for LiCl-mediated dorsalization of the embryo (Ishibashi et al., 2008). Furthermore, the axial defects we report for Sia-Twn knockdown are those predicted for inhibition of maternal Wnt signaling (reviewed in Heasman, 2006). The requirement for Sia and Twn in mediating the response to maternal Wnt signaling was determined by examining the influence of Sia-Twn knockdown on Xwnt8-induced axis induction (Fig. 7). At the four-cell stage, both ventral blastomeres were injected with Sia MO or Twn MO individually, or with the combination of both MOs, and at the eight-cell stage a single ventral blastomere was injected with Xwnt8 mRNA. Complete axis formation was induced at high frequency in response to Xwnt8 (90%) (Fig. 7B), and this response was unaffected by Sia MO, Twn MO or the non-specific control MO (77%, 85%, and 97%, respectively) (Fig. 7D,F,H). Simultaneous knockdown of Sia and Twn abrogated Xwnt8-mediated axis induction in most embryos (66% normal development) (Fig. 7J), with 34% displaying a weaker partial axis induction, while none of the embryos displayed a complete ectopic axis (data not shown). These data may reflect a partial inhibition of Xwnt8 activity, which could suggest an incomplete knockdown of Sia and Twn, as well as the presence of other effectors of the Wnt pathway in organizer formation. We note that ventral injection of the MOs alone had no effect on axial development (Fig. 7C,E,G,I). The results indicate that Sia and Twn together are required for Wnt induction of axis formation, demonstrating an essential and redundant role for Sia and Twn in mediating the transcriptional response to maternal Wnt signaling in axis formation.

Fig. 7
Siamois and Twin are required for Xwnt8 induction of ectopic axis formation. At the 4-cell stage both ventral blastomeres were injected with (C,D) a non-specific control morpholino oligonucleotide (NSMO, 50ng), (E,F) a Sia-specific oligonucleotide (SiaMO, ...

Discussion

Our results demonstrate an essential role for Sia and Twn in the transcriptional activation of Gsc and the formation of the Spemann organizer in Xenopus. Sia and Twn form functionally equivalent homodimers and heterodimers that occupy a conserved Wnt-responsive element of the Gsc promoter. Knockdown of Sia and Twn together, but not individually, results in severe axial defects, characterized by a loss of organizer gene expression and a failure of organizer formation. The results demonstrate that Sia and Twn are functionally redundant, as predicted from their structural, expression profile and functional similarities, and together are essential for formation of the Spemann organizer. Furthermore, Sia and Twn are required transcriptional mediators of the response to maternal Wnt signals in organizer formation and axial development. These studies establish Sia and Twn as essential and redundant activators of the Spemann organizer transcriptional program in the Xenopus gastrula.

Siamois and Twin are essential for Spemann organizer formation

Sia and Twn show striking similarity in structure, expression pattern, transcriptional activity and developmental function. With nearly identical homeodomains (88% identity) (Laurent et al., 1997; Lemaire et al., 1995), Sia and Twn likely bind to and activate a common set of target genes within the organizer domain of the early gastrula. Our results demonstrate that Sia and Twn transactivate target genes as homodimers or heterodimers with equivalent function. In gain-of-function studies, ventral expression of Sia or Twn induced a complete axial duplication containing head and trunk structures (Laurent et al., 1997; Lemaire et al., 1995). Taken together, these observations predict that Sia and Twn function redundantly in the regulation of organizer formation.

Previous developmental studies of Sia and Twn are consistent with redundant and essential roles in organizer formation, but did not provide a definitive analysis. Overexpression of a dominant repressive form of Sia (Eng-Sia), fully inhibits organizer gene expression, resulting in disruption of head, trunk and tail structures, consistent with complete ventralization of the body axis (Fan and Sokol, 1997; Kessler, 1997). At the time of these studies, Twn had not yet been identified, but it is predicted that Eng-Sia strongly represses the common targets of both Sia and Twn, and therefore, the phenotypic response to Eng-Sia likely reflects the consequence of interfering with both Sia and Twn function. Given the overexpression of a dominant repressive fusion protein, it is possible that the severity of the development defects reflects off-target effects. However, the complete rescue of axis formation by coexpression of native Sia argues for specificity in the phenotypic defects obtained (Fan and Sokol, 1997; Kessler, 1997). While not true loss-of-function analyses, these studies support an essential role for Sia and Twn in organizer formation and axial development.

In contrast to the Eng-Sia studies, a recent knockdown analysis of Sia and Twn demonstrated redundancy, but a requirement only for anterior axial development (Ishibashi et al., 2008). The conclusion that Sia and Twn are required for head, but not trunk formation, suggested that Sia and Twn are not required for the full activity of the Spemann organizer. The knockdown results we obtained are consistent with the Eng-Sia studies (Fan and Sokol, 1997; Kessler, 1997), but not with the prior knockdown analyses (Ishibashi et al., 2008). We find that knockdown of both Sia and Twn results in a complete loss of organizer formation (Fig. 6H’,I’), and consequently neither head nor trunk structures form in the most severe phenotypic class (Fig. 6J). The discrepancy in the severity of axial defects likely reflects a difference in knockdown efficiency, with the prior results representing a partial loss-of-function for Sia and Twn, while in our studies a more complete knockdown was achieved. In support of this interpretation, we find that lower dosage of the mixture of Sia MO and Twn MO phenocopies the anterior defects previously reported (Fig. 6M). Therefore, our results confirm that Sia and Twn are redundant and essential for formation of the Spemann organizer, including both head and trunk organizer activity.

Sia and Twn are redundant factors, and together are essential for the formation of the Spemann organizer. Sia and Twn appear to play equivalent roles in organizer formation, as knockdown of either Sia or Twn alone has no effect on axis formation (Fig. 6E,F). This suggests that Sia or Twn homodimers can compensate for the loss of the Sia-Twn heterodimer. The overall structure of Sia and Twn are highly similar, with high sequence conservation with their homeodomains, as well as within small regions N-terminal to the homeodomain (Laurent et al., 1997). It is likely that Sia and Twn were formed as a result of the duplication of an ancestral Sia-Twn-like gene, whether a local or genome-wide duplication, but despite significant sequence divergence outside of the homeodomain, it appears that the transcriptional and developmental functions of these genes have not diverged (Van de Peer et al., 2009). Further studies may reveal whether Sia and Twn have discrete functions, perhaps in a target-specific or context-specific manner.

An intriguing observation is the apparent absence of Sia and Twn orthologs in non-amphibian vertebrates. While Sia and Twn orthologs have been identified in the closely related amphibian Xenopus tropicalis, other vertebrate orthologs have yet to be identified despite extensive efforts. Given the presence of the conserved Sia-Twn response element in all vertebrate Gsc promoters, the apparent absence of Sia and Twn orthologs raises questions about the conservation of Sia and Twn and the role of functional homologs in organizer formation of other vertebrates. Interestingly, a similar conundrum is found in zebrafish bozozok, a homeodomain protein that functions as a transcriptional repressor (Yamanaka et al., 1998; Fekany et al., 1999; Koos and Ho, 1999). bozozok is essential for organizer formation and expression of organizer genes such as gsc and the Nodal-related gene, squint (Shimizu et al., 2000; Solnica-Krezel and Driever, 2001), yet no vertebrate orthologs have been identified. While true orthologs of Sia, Twn or Bozozok may be identified in other vertebrates, it seems likely that the developmental functions of these Xenopus- and zebrafish-specific factors may reside in functional homologs that are employed in other species to regulate organizer formation and organizer gene expression. The presence of species-specific transcriptional regulators of organizer formation in Xenopus and zebrafish suggests an unexpected regulatory diversity, perhaps reflecting either distinct developmental demands in these species or an evolutionary flexibility at this discrete step of organizer formation.

Transcriptional regulation of Goosecoid and other organizer genes by Siamois and Twin

Our results suggest that Sia and Twn regulate Gsc transcription by binding to a conserved HD binding site within the Wnt responsive proximal element of the Gsc promoter. As direct targets of maternal Wnt signals (Brannon and Kimelman, 1996; Carnac et al., 1996; Crease et al., 1998; Fan et al., 1998; Nelson and Gumbiner, 1998; Nishita et al., 2000), Sia and Twn are expressed at the onset of zygotic gene expression in the blastula (Blythe et al., 2010; Laurent et al., 1997; Lemaire et al., 1995), and likely play a role in the initiation of the expression of organizer genes at the onset of gastrulation. Consistent with this mechanism, Gsc and Chd expression is reduced or absent at the start of gastrulation in Sia/Twn knockdown embryos (Fig. 6H’,I’). The BMP antagonists Chordin and Noggin, which are required for proper organizer function (Khokha et al., 2005), can partially rescue axis formation in Sia/Twn knockdown embryos (data not shown), placing Chordin and Noggin downstream of Sia and Twn in Spemann organizer function.

The Gsc promoter also contains a highly conserved Nodal-responsive element (DE) in addition to the Wnt-responsive element (PE) (Watabe et al., 1995). Our results provide strong evidence that Sia and Twn mediate the zygotic response to maternal Wnt signals through direct binding to a conserved P3 site within the PE element of the Gsc promoter. However, which Nodal effectors are involved in the initiation of Gsc expression and how those may interact with the Wnt effectors Sia/Twn remains to be determined. The Nodal signaling pathway has been shown to signal through several pathway effectors, including Fast1 (FoxH1), a Fox family transcription factor that is maternally expressed throughout the embryo (Chen et al., 1996), and Mix family members such as Mixer or Milk, which are paired-type homeodomain transcriptional activators that are zygotically expressed throughout the endoderm (Germain et al., 2000). Fast1 is present prior to and during gastrula stages (Watanabe and Whitman, 1999), suggesting that it likely plays a role in initiation of Gsc expression, perhaps in cooperation with Sia and Twn. Consistent with this idea, maternal knockdown of Fast1 results in decreased expression of Gsc (Kofron et al., 2004), and Fast1 has been shown to directly occupy the endogenous Gsc promoter (Blythe et al., 2009). Mixer and Milk interact with the signaling mediator Smad2 in a Nodal-dependent manner, and can form a complex on the DE of the Gsc promoter (Germain et al., 2000). The zygotic expression of Mix family members suggests a later role in the maintenance of Gsc expression.

The Nodal-responsive DE and the Wnt-responsive PE are nearly adjacent (~50 bp separation) in all Gsc promoters (Watabe et al., 1995), raising the possibility that transcriptional effectors of the two pathways may interact or cooperate to activate Gsc transcription. Our preliminary results indicate that Nodal and Wnt pathway effectors synergistically enhance transcription of Gsc (Reid and Kessler, unpublished results), consistent with an interaction of pathway effectors at the Gsc promoter. The strong conservation of both the DE and the PE in vertebrate Gsc promoters suggests a conserved mechanism of Gsc regulation involving transcriptional integration of Nodal and Wnt signaling inputs.

Given the conserved structure of the Gsc promoter, it is interesting to consider whether the function of Gsc is conserved across species. Disruption of Gsc function in Xenopus, either by knockdown or expression of a dominant activating form of Gsc, leads to severe anterior defects, including a reduction or loss of head structures anterior to the hindbrain (Sander et al., 2007; Yao and Kessler, 2001). In contrast, a mouse knockout of Gsc results in no developmental defects associated with organizer function (Rivera-Perez et al., 1995; Wakamiya et al., 1998; Yamada et al., 1995; Zhu et al., 1998). Gsc mutant mice gastrulate normally and show normal development of the primary body axes. However, the mutants do die shortly after birth due to severe craniofacial defects, as well as improperly formed sternum and ribs (Rivera-Perez et al., 1995). If the function of Gsc is not conserved in higher vertebrates, it remains to be seen whether the regulatory control of Gsc expression is conserved. The P3 site within the Gsc promoter is conserved in vertebrates (Fig. 1A), indicating that a paired-type homeodomain-containing protein likely regulates the expression of Gsc in all vertebrates. However, the identity of such proteins, their role in the initiation and/or maintenance of Gsc transcription, and their ability to mediate the transcription response to Wnt signals remain unknown. The mouse PE is Wnt-responsive in Xenopus explants (Watabe et al., 1995), suggesting that Wnt pathway inputs may influence the control of Gsc expression in mammals, but whether the PE confers Wnt-responsiveness in mammals remains to be determined. The availability of complete genome sequences and the introduction of powerful computational approaches should aid in the identification of Gsc regulators that may serve as the functional homologs of Sia and Twn in higher vertebrates.

Sia and Twn have been identified as direct regulators of Gsc, and likely mediate the Wnt-dependent transcriptional activation of multiple organizer genes (Fan and Sokol, 1997; Kessler, 1997; Laurent et al., 1997; Yamamoto et al., 2003). Sia, in cooperation with other paired-type homeodomain proteins, has been implicated in the transcription of several organizer genes, including Cerberus (Yamamoto et al., 2003) and Crescent (Shibata et al., 2000). However, it is unclear how Sia may be interacting with other homeodomain proteins to affect gene transcription for other organizer-specific genes. Xlim-1 and Lim Domain Binding Protein-1 were shown to influence Gsc transcription, although through a site upstream of the PE (Mochizuki et al., 2000). Whether Sia and Twn initiate expression of Gsc and other organizer genes in cooperation with Nodal signals (Engleka and Kessler, 2001) remains to be determined, as is the role of other promoter elements and regulatory proteins that maintain organizer gene expression through the gastrula and neurula stages.

Formation of the organizer domain within the gastrula embryo is essential for germ layer patterning and axial development. Sia and Twn act redundantly downstream of the Wnt pathway to regulate formation of the organizer. Sia and Twn, and likely other factors, play an essential role in specifying the proper spatial and temporal expression of the organizer-specific gene Gsc. As mediators of the transcriptional response to maternal Wnt signals, and through cooperative interactions with other pathways, Sia and Twn control the expression of multiple organizer genes, thus contributing to the establishment of the organizer transcriptional program.

Supplementary Material

01

02

03

4

Supplementary Materials

Supplementary Figure 1. The Siamois or Twin homeodomain is sufficient for DNA-binding and complex formation at the Gsc proximal element. Increasing amounts of purified Sia homeodomain (A) or Twn homeodomain (B) was incubated with the indicated radiolabeled EMSA probes. Probe sequences are for wild-type and mutated forms of the Gsc proximal element are shown in Fig. 2A. Monomer (M) and dimer (D) complexes were observed for the WT, 136 MT and 127 MT probes. The monomer complex only was observed for the 2X MT probe and no complex formation was observed for 3X MT (data not shown). F, free probe.

Supplementary Figure 2. Morpholino antisense oligonucleotides specifically block the translation and biological activity of Siamois and Twin. (A) In vitro translation reactions programmed with DNA constructs (1μg) encoding native Sia or Twn, or myc-tagged forms of Sia or Twn, in the presence of oligonucleotides (100ng) specific for Sia or Twn, or a non-specific control oligonucleotide (NS). Translation products were labeled with 35S-methionine, resolved by 12% SDS-PAGE, and visualized by autoradiography. Protein size markers are on the left. The Sia MO blocked translation of Sia, but not Twn. The Twn MO blocked translation of Twn, but not Sia. Neither oligonucleotide blocked translation of myc-Sia or myc-Twn, which have distinct upstream translation start sites. The NSMO oligonucleotide had no translation blocking activity for any of the proteins. (B-M) Inhibition of axis induction by Sia- or Twn-specific oligonucleotides. At the 4-cell stage both ventral blastomeres were injected with (E-G) a non-specific control morpholino oligonucleotide (NSMO, 25ng), (H-J) a Sia-specific oligonucleotide (SiaMO, 25ng), or (K-M) a Twn-specific oligonucleotide (TwnMO, 25ng). At the 8-cell stage a single ventral blastomere was injected with 20pg of (C,F,I,L) Sia, (D,G,J,M) Twn, (I, inset) myc-Sia, or (M, inset) myc-Twn mRNA. The Sia MO blocked axis induction by Sia, but not Twn. The Twn MO blocked axis induction by Twn, but not Sia. myc-Sia and myc-Twn were insensitive to the corresponding oligonucleotides and the NSMO oligonucleotide did not block axis induction for either Sia or Twn. Whole embryo morphology (dorsal up, anterior right) is shown at the tailbud stage, with percentage of embryos displaying the representative phenotype and total embryos analyzed indicated in the lower right for each panel. (B) Uninjected control embryo.

Supplementary Figure 3. Rescue of axial development in the Siamois-Twin double knockdown embryo. (B,D,F) At the 4-cell stage both dorsal blastomeres were injected with a combination of the Sia and Twn oligonucleotides (SiaMO+TwnMO, 25ng+25ng). At the 8-cell stage a single dorsal blastomere was injected with 50pg of (C,D) myc-Sia or (E,F) myc-Twn. myc-Sia and myc-Twn fully rescued axial development in double knockdown embryos (D,F), and resulted in mild dorsalization in control embryos (C,E). Whole embryo morphology (dorsal up, anterior right) is shown at the tailbud stage, with percentage of embryos displaying the representative phenotype and total embryos analyzed indicated in the lower right for each panel. (A) Uninjected control embryo.

Acknowledgments

We are grateful to Shelby Blythe, Doug Epstein, Peter Klein and Shawn Little for criticalreading of the manuscript. We thank Ken Cho, John Gurdon, and Patrick Lemaire for providing plasmids. This work was supported by grants from the NIH (T32-HD007516) to C.D.R. and by grants from the NIH (R01-GM64768) and NSF (IOS-0718961) to D.S.K.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  • Blumberg B, Wright CV, De Robertis EM, Cho KW. Organizer-specific homeobox genes in Xenopus laevis embryos. Science. 1991;253:194–6. [PubMed]
  • Blythe SA, Cha SW, Tadjuidje E, Heasman J, Klein PS. beta-Catenin primes organizer gene expression by recruiting a histone H3 arginine 8 methyltransferase, Prmt2. Dev Cell. 2010;19:220–31. [PMC free article] [PubMed]
  • Blythe SA, Reid CD, Kessler DS, Klein PS. Chromatin immunoprecipitation in early Xenopus laevis embryos. Dev Dyn. 2009;238:1422–32. [PMC free article] [PubMed]
  • Brannon M, Gomperts M, Sumoy L, Moon RT, Kimelman D. A beta-catenin/XTcf-3 complex binds to the siamois promoter to regulate dorsal axis specification in Xenopus. Genes Dev. 1997;11:2359–70. [PMC free article] [PubMed]
  • Brannon M, Kimelman D. Activation of Siamois by the Wnt pathway. Dev Biol. 1996;180:344–7. [PubMed]
  • Brenowitz M, Senear DF, Kingston RE. DNase I footprint analysis of protein-DNA binding. Curr Protoc Mol Biol. 2001:4. Chapter 12, Unit 12. [PubMed]
  • Carnac G, Kodjabachian L, Gurdon JB, Lemaire P. The homeobox gene Siamois is a target of the Wnt dorsalisation pathway and triggers organiser activity in the absence of mesoderm. Development. 1996;122:3055–65. [PubMed]
  • Chen X, Rubock MJ, Whitman M. A transcriptional partner for MAD proteins in TGF-beta signalling. Nature. 1996;383:691–6. [PubMed]
  • Cho KW, Blumberg B, Steinbeisser H, De Robertis EM. Molecular nature of Spemann's organizer: the role of the Xenopus homeobox gene goosecoid. Cell. 1991;67:1111–20. [PMC free article] [PubMed]
  • Crease DJ, Dyson S, Gurdon JB. Cooperation between the activin and Wnt pathways in the spatial control of organizer gene expression. Proc Natl Acad Sci U S A. 1998;95:4398–403. [PMC free article] [PubMed]
  • De Robertis EM. Goosecoid and Gastrulation. In: Stern C, editor. Gastrulation: From Cells to Embryo. Cold Spring Harbor Laboratory Press; Cold Spring Harbor, NY: 2004. pp. 581–590.
  • De Robertis EM. Spemann's organizer and self-regulation in amphibian embryos. Nat Rev Mol Cell Biol. 2006;7:296. [PMC free article] [PubMed]
  • Engleka MJ, Kessler DS. Siamois cooperates with TGFbeta signals to induce the complete function of the Spemann-Mangold organizer. Int J Dev Biol. 2001;45:241–50. [PubMed]
  • Fan MJ, Gruning W, Walz G, Sokol SY. Wnt signaling and transcriptional control of Siamois in Xenopus embryos. Proc Natl Acad Sci U S A. 1998;95:5626–31. [PMC free article] [PubMed]
  • Fan MJ, Sokol SY. A role for Siamois in Spemann organizer formation. Development. 1997;124:2581–9. [PubMed]
  • Geisberg JV, Struhl K. Quantitative sequential chromatin immunoprecipitation, a method for analyzing co-occupancy of proteins at genomic regions in vivo. Nucleic Acids Res. 2004;32:e151. [PMC free article] [PubMed]
  • Germain S, Howell M, Esslemont GM, Hill CS. Homeodomain and winged-helix transcription factors recruit activated Smads to distinct promoter elements via a common Smad interaction motif. Genes Dev. 2000;14:435–51. [PMC free article] [PubMed]
  • Harland R, Gerhart J. Formation and function of Spemann's organizer. Annu Rev Cell Dev Biol. 1997;13:611–67. [PubMed]
  • Heasman J. Patterning the early Xenopus embryo. Development. 2006;133:1205–17. [PubMed]
  • Ishibashi H, Matsumura N, Hanafusa H, Matsumoto K, De Robertis EM, Kuroda H. Expression of Siamois and Twin in the blastula Chordin/Noggin signaling center is required for brain formation in Xenopus laevis embryos. Mech Dev. 2008;125:58–66. [PMC free article] [PubMed]
  • Kao KR, Elinson RP. Dorsalization of mesoderm induction by lithium. Dev Biol. 1989;132:81–90. [PubMed]
  • Kessler DS. Siamois is required for formation of Spemann's organizer. Proc Natl Acad Sci U S A. 1997;94:13017–22. [PMC free article] [PubMed]
  • Khokha MK, Yeh J, Grammer TC, Harland RM. Depletion of three BMP antagonists from Spemann's organizer leads to a catastrophic loss of dorsal structures. Dev Cell. 2005;8:401–11. [PubMed]
  • Kodjabachian L, Lemaire P. Siamois functions in the early blastula to induce Spemann's organiser. Mech Dev. 2001;108:71–9. [PubMed]
  • Kodjabachian L, Lemaire P. Role of Siamois before and during Gastrulation. In: Stern C, editor. Gastrulation: From Cells to Embryo. Cold Spring Harbor Laboratory Press; Cold Spring Harbor, NY: 2004. pp. 609–17.
  • Kofron M, Birsoy B, Houston D, Tao Q, Wylie C, Heasman J. Wnt11/beta-catenin signaling in both oocytes and early embryos acts through LRP6-mediated regulation of axin. Development. 2007;134:503–13. [PubMed]
  • Kofron M, Puck H, Standley H, Wylie C, Old R, Whitman M, Heasman J. New roles for FoxH1 in patterning the early embryo. Development. 2004;131:5065–78. [PubMed]
  • Kuo JS, Patel M, Gamse J, Merzdorf C, Liu X, Apekin V, Sive H. Opl: a zinc finger protein that regulates neural determination and patterning in Xenopus. Development. 1998;125:2867–82. [PubMed]
  • Laurent MN, Blitz IL, Hashimoto C, Rothbacher U, Cho KW. The Xenopus homeobox gene twin mediates Wnt induction of goosecoid in establishment of Spemann's organizer. Development. 1997;124:4905–16. [PubMed]
  • Lemaire P, Garrett N, Gurdon JB. Expression cloning of Siamois, a Xenopus homeobox gene expressed in dorsal-vegetal cells of blastulae and able to induce a complete secondary axis. Cell. 1995;81:85–94. [PubMed]
  • Mochizuki T, Karavanov AA, Curtiss PE, Ault KT, Sugimoto N, Watabe T, Shiokawa K, Jamrich M, Cho KW, Dawid IB, Taira M. Xlim-1 and LIM domain binding protein 1 cooperate with various transcription factors in the regulation of the goosecoid promoter. Dev Biol. 2000;224:470–85. [PubMed]
  • Nelson RW, Gumbiner BM. Beta-catenin directly induces expression of the Siamois gene, and can initiate signaling indirectly via a membrane-tethered form. Ann N Y Acad Sci. 1998;857:86–98. [PubMed]
  • Nieuwkoop PD, Faber J. Normal Table of Xenopus laevis (Daudin). North Holland Publishing Company; Amsterdam: 1967.
  • Nishita M, Hashimoto MK, Ogata S, Laurent MN, Ueno N, Shibuya H, Cho KW. Interaction between Wnt and TGF-beta signalling pathways during formation of Spemann's organizer. Nature. 2000;403:781–5. [PubMed]
  • Rivera-Perez JA, Mallo M, Gendron-Maguire M, Gridley T, Behringer RR. Goosecoid is not an essential component of the mouse gastrula organizer but is required for craniofacial and rib development. Development. 1995;121:3005–12. [PubMed]
  • Sander V, Reversade B, De Robertis EM. The opposing homeobox genes Goosecoid and Vent1/2 self-regulate Xenopus patterning. EMBO J. 2007;26:2955–65. [PMC free article] [PubMed]
  • Sasai Y, Lu B, Steinbeisser H, Geissert D, Gont LK, De Robertis EM. Xenopus chordin: a novel dorsalizing factor activated by organizer-specific homeobox genes. Cell. 1994;79:779–90. [PMC free article] [PubMed]
  • Shibata M, Ono H, Hikasa H, Shinga J, Taira M. Xenopus crescent encoding a Frizzled-like domain is expressed in the Spemann organizer and pronephros. Mech Dev. 2000;96:243–6. [PubMed]
  • Shimizu T, Yamanaka Y, Ryu SL, Hashimoto H, Yabe T, Hirata T, Bae YK, Hibi M, Hirano T. Cooperative roles of Bozozok/Dharma and Nodal-related proteins in the formation of the dorsal organizer in zebrafish. Mech Dev. 2000;91:293–303. [PubMed]
  • Sive HL, Grainger RM, Harland RM. Early Development of Xenopus laevis: A Laboratory Manual. Cold Spring Harbor Laboratory Press; Cold Spring Harbor, NY: 2000.
  • Sokol S, Christian JL, Moon RT, Melton DA. Injected Wnt RNA induces a complete body axis in Xenopus embryos. Cell. 1991;67:741–52. [PubMed]
  • Solnica-Krezel L, Driever W. The role of the homeodomain protein Bozozok in zebrafish axis formation. Int J Dev Biol. 2001;45:299–310. [PubMed]
  • Tao Q, Yokota C, Puck H, Kofron M, Birsoy B, Yan D, Asashima M, Wylie CC, Lin X, Heasman J. Maternal wnt11 activates the canonical wnt signaling pathway required for axis formation in Xenopus embryos. Cell. 2005;120:857–71. [PubMed]
  • Van de Peer Y, Maere S, Meyer A. The evolutionary significance of ancient genome duplications. Nat Rev Genet. 2009;10:725–32. [PubMed]
  • Wakamiya M, Lindsay EA, Rivera-Perez JA, Baldini A, Behringer RR. Functional analysis of Gscl in the pathogenesis of the DiGeorge and velocardiofacial syndromes. Hum Mol Genet. 1998;7:1835–40. [PubMed]
  • Watabe T, Kim S, Candia A, Rothbacher U, Hashimoto C, Inoue K, Cho KW. Molecular mechanisms of Spemann's organizer formation: conserved growth factor synergy between Xenopus and mouse. Genes Dev. 1995;9:3038–50. [PubMed]
  • Watanabe M, Whitman M. FAST-1 is a key maternal effector of mesoderm inducers in the early Xenopus embryo. Development. 1999;126:5621–34. [PubMed]
  • White R. Homeodomain proteins. Homeotic genes seek partners. Curr Biol. 1994;4:48–50. [PubMed]
  • Wilson D, Sheng G, Lecuit T, Dostatni N, Desplan C. Cooperative dimerization of paired class homeo domains on DNA. Genes Dev. 1993;7:2120–34. [PubMed]
  • Wilson DS, Guenther B, Desplan C, Kuriyan J. High resolution crystal structure of a paired (Pax) class cooperative homeodomain dimer on DNA. Cell. 1995;82:709–19. [PubMed]
  • Wilson PA, Melton DA. Mesodermal patterning by an inducer gradient depends on secondary cell-cell communication. Curr Biol. 1994;4:676–86. [PubMed]
  • Yaklichkin S, Steiner AB, Lu Q, Kessler DS. FoxD3 and Grg4 physically interact to repress transcription and induce mesoderm in Xenopus. J Biol Chem. 2007;282:2548–57. [PMC free article] [PubMed]
  • Yamada G, Mansouri A, Torres M, Stuart ET, Blum M, Schultz M, De Robertis EM, Gruss P. Targeted mutation of the murine goosecoid gene results in craniofacial defects and neonatal death. Development. 1995;121:2917–22. [PubMed]
  • Yamamoto S, Hikasa H, Ono H, Taira M. Molecular link in the sequential induction of the Spemann organizer: direct activation of the cerberus gene by Xlim-1, Xotx2, Mix.1, and Siamois, immediately downstream from Nodal and Wnt signaling. Dev Biol. 2003;257:190–204. [PubMed]
  • Yao J, Kessler DS. Goosecoid promotes head organizer activity by direct repression of Xwnt8 in Spemann's organizer. Development. 2001;128:2975–87. [PubMed]
  • Zhu CC, Yamada G, Nakamura S, Terashi T, Schweickert A, Blum M. Malformation of trachea and pelvic region in goosecoid mutant mice. Dev Dyn. 1998;211:374–81. [PubMed]
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...