Drs. Patil and Alexander contributed equally to this work.
Novel gene ashwin functions in Xenopus cell survival and anteroposterior patterning
Article first published online: 5 MAY 2006
Copyright © 2006 Wiley-Liss, Inc.
Volume 235, Issue 7, pages 1895–1907, July 2006
How to Cite
Patil, S. S., Alexander, T. B., Uzman, J. A., Lou, C.-H., Gohil, H. and Sater, A. K. (2006), Novel gene ashwin functions in Xenopus cell survival and anteroposterior patterning. Dev. Dyn., 235: 1895–1907. doi: 10.1002/dvdy.20834
- Issue published online: 22 MAY 2006
- Article first published online: 5 MAY 2006
- Manuscript Accepted: 25 MAR 2006
- Texas Advanced Research Program. Grant Number: 003652-0176-2001
- NSF. Grant Number: IBN-9723183
- neural induction;
- head formation
The novel gene ashwin was isolated in a differential display screen for genes activated or up-regulated early in neural specification. ashwin is expressed maternally and zygotically, and it is up-regulated in the neural ectoderm after the midgastrula stage. It is expressed in the neural plate and later in the embryonic brain, eyes, and spinal cord. Overexpression of ashwin in whole embryos leads to anterior truncations and other defects. However, a second Organizer does not form, and the secondary axial structures may result from splitting of the Organizer, rather than axis duplication. Morpholino oligonucleotide-mediated reduction in ashwin expression leads to lethality or abnormalities in gastrulation, as well as significant apoptosis in midgastrula embryos. Apoptosis is also observed in midgastrula embryos overexpressing ashwin. Coexpression of ashwin with the bone morphogenetic protein-4 antagonist noggin has a synergistic effect on neural-specific gene expression in isolated animal cap ectoderm. Ashwin has no previously characterized domains, although two nuclear localization signals can be identified. Orthologues have been identified in the human, mouse, chicken, and pufferfish genomes. Our results suggest that ashwin regulates cell survival and anteroposterior patterning. Developmental Dynamics 235:1895–1907, 2006. © 2006 Wiley-Liss, Inc.
Vertebrate neural specification requires an inhibition of bone morphogenetic protein-4 (BMP4) signals (reviewed in Weinstein and Hemmati-Brivanlou,1999; Harland,2000; Stern,2004). Extracellular BMP4 antagonists such as noggin and chordin play a critical role, but antagonism of BMP4 also may involve the inhibitory Smads Smad6 (Nakayama et al.,1998) and Smad7 (Casellas and Brivanlou,1998) or inhibitory phosphorylation of Smad1 by mitogen-activated protein kinase (MAPK) in response to fibroblast growth factor (FGF; Kretzschmar et al.,1997; Gohil and Sater, unpublished results). In addition, β-catenin has been shown to mediate transcriptional repression of BMP4 in gastrula ectoderm (Baker et al.,1999). Although inhibition of BMP4 signals is a primary step in the establishment of anterior neural fate, insulin-like growth factor (IGF) signaling also contributes to formation of the head and anterior neural structures (Pera et al.,2001; Richard-Parpaillon et al.,2002). Specification of posterior neural identity requires additional signals involving FGFs (Lamb and Harland,1995; Hongo et al.,1999; Holowacz and Sokol,1999) and Wnt3A (McGrew et al.,1995).
The dorsal animal ectoderm shows a neural bias by the onset of gastrulation, and several genes involved in neural specification are expressed in the dorsal ectoderm of early gastrulae. These genes are down-regulated in presumptive epidermal ectoderm by the midgastrula stage. In the midgastrula neural plate, however, they are up-regulated, and their expression is required for commitment to neural fate. A key group of these genes are transcriptional regulators that stabilize the new state of specification and promote progression of cells along a neural developmental pathway. Such genes include opl (Kuo et al.,1998), otx2 (Blitz and Cho,1995; Pannese et al.,1995), and foxD5a (Sullivan et al.,2001), among others. In a screen for genes expressed at the earliest stages of neural specification, we have identified the novel gene ashwin, a maternally expressed gene that is up-regulated during neural specification. Because overexpression leads to formation of a secondary, or split, axis, the gene has been named for one of the twin children of the sun in Hindu mythology, who together became the morning star and the evening star. Initial functional analyses indicate that ashwin can participate in ectodermal specification and may also contribute to axial and anteroposterior neural pattern formation.
Isolation and Characterization of ashwin
We carried out a differential display screen for genes up-regulated early in neural specification by comparing amplified polymerase chain reaction (PCR) fragments from cDNA prepared from five different embryonic tissues. Three tissues served as “positive samples” that were expected to express genes characteristic of early neural specification. These tissues included (1) midgastrula (stage [st.] 11) neural plates, (2) animal caps isolated from embryos microinjected with 0.25 ng of noggin and cultured until st. 11, and (3) animal caps isolated at late blastula and treated at st. 10.5 with 10 mM ammonium chloride for 1 hr. Two negative control samples included (4) animal caps isolated at midblastula (st. 8) and cultured until st. 11 and (5) ventral ectoderm isolated from a midneurula (st. 15) embryo. For each sample, preliminary reverse transcriptase-polymerase chain reaction (RT-PCR) assays were performed to confirm the identity of the tissues (not shown).
Polyacrylamide gels (Genomyx Differential Display Gel System) were manually scanned for amplified PCR products that were consistently detected in all three positive samples and absent from both negative control samples. After a second round of screening by means of colony hybridization, 13 different clones were obtained and sequenced. Among these were clones for previously identified neural specific genes (e.g., XANF-1, Zaraisky et al.,1992), as well as novel cDNA sequences. The expression of some of these novel cDNA sequences was examined by in situ hybridization. One of these novel cDNAs was expressed within embryonic neural tissues and was chosen for a more detailed analysis.
ashwin (GenBank accession no. AF281253) encodes a protein of 226 amino acids that contains no previously identified domains. Ashwin contains two nuclear localization signals and, like many nuclear proteins, is highly basic, containing 15.9% basic residues. Figure 1A shows the amino acid sequence of ashwin in comparison with hypothetical proteins deduced from the human, mouse, chicken, and Takifugu genomes. Overall, ashwin is 47% identical and 58% similar to the human orthologue MGC5509. There are three highly conserved regions of between 29 and 79 amino acids, interspersed with two variable regions. The N-terminal conserved domain is 67% identical to the human orthologue, whereas the middle conserved domain shows 88% identity, and the C-terminal conserved domain is 70% identical to the other genes (Fig. 1B). At the N-terminus, ashwin contains a leucine-rich region adjacent to a basic region (Kneller et al.,1990) that falls within one of four conserved helical regions predicted by the algorithm NNPredict. Thus, this region shows some of the hallmarks of a leucine-zipper motif (reviewed in Hurst,1994).
Scans of expressed sequence tag (EST) databases for human, mouse, and Xenopus indicate that ashwin and its orthologues are expressed in a wide range of embryonic and adult tissues, including derivatives of all three germ layers. Among adult mammalian tissues, the ashwin orthologue has been identified in ESTs from brain, pituitary, retina, kidney, blood, liver, lung, thymus, germ cell, pancreas, and placenta. The ashwin orthologue is detected among ESTs from mouse embryonic tissues, including eye, spinal cord, branchial arches, urinary bladder, and mammary gland. In adult Xenopus, ashwin has been found among ESTs from the eye, heart, liver, spleen, lung, and ovary.
Temporal and Spatial Expression of ashwin
Northern analysis of ashwin expression reveals two transcripts, which are present in the unfertilized egg and are detected through the tailbud stage (Fig. 2N). In situ hybridization studies show that ashwin transcripts become detectable in the dorsal marginal zone just before the onset of gastrulation (st. 9+, not shown). Before gastrulation, ashwin is expressed at low levels in the animal cap as detected by RT-PCR, and we found no evidence for localization of maternal transcripts within the oocyte (not shown). After the onset of gastrulation, ashwin becomes expressed in a broad dorsolateral domain (st. 10, Fig. 2A) and is strongly expressed throughout the dorsal hemisphere by midgastrulation (st. 12, Fig. 2B,C). There is also a strong circumblastoporal ring of expression that persists throughout early development and culminates in the tail bud (Fig. 2C,L). Ashwin expression persists throughout the forming brain and spinal cord at all stages analyzed (Fig. 2E–J). However, visualization of the deep expression pattern in dissected or cleared embryos reveals strong anterior and posterior domains overlying its pan-neural expression pattern, as well as staining in the notochord and prechordal plate (Fig. 2G). By midneurula, ashwin expression is restricted to the ventral brain and spinal cord, as well as throughout the eye anlage (Fig. 2J,K). Overall, expression of ashwin coincides both temporally and spatially in regions involved in neural induction and anteroposterior patterning.
Overexpression of ashwin in Whole Embryos Leads to Phenotypic Abnormalities
As a preliminary assessment of ashwin function, ashwin was overexpressed in embryos by microinjection of mRNA encoding ashwin in either myc-tagged or native forms. Microinjection of more than 250 pg/embryo of mRNA encoding ashwin was generally lethal, whereas injection of less than 200 pg had no visible phenotypic effect. Overexpression of ashwin in the animal region leads to a range of abnormalities, including anterior truncations (Fig. 3A), poorly developed secondary axes (Fig. 3B), open neural folds (Fig. 3C), and epidermal shedding at pre–tail bud stages (Fig. 3F). These abnormalities affected approximately 80% of injected embryos (Table 1). Embryonic mortality was high because of epidermal shedding beginning after neural tube closure.
|Treatment||Ashwin RNA||Ashwin RNA||Ashwin DNA|
|Site of injection||AP||Ventral||AP|
|2nd Dorsal lip (%)||7 (5)||22 (10)||NS|
|Split axis (%)||34 (24)||4 (2)||16 (19)|
|Anterior truncation (%)||60 (43)||0||47 (55)|
|Neurulation defects (%)||37 (26)||0||NS|
|Epidermal shedding (%)||75 (54)||0||29 (34)|
|Number of embryos (experiments)||140 (9)||227 (3)||85 (4)|
Lineage tracing studies were carried out to determine whether the phenotypic variability could be attributed to differences in the distribution of the ashwin protein. Animally directed injection of 200 pg/embryo of RNA encoding myc-tagged ashwin, followed by immunolocalization of the protein using anti-myc antibody, showed that embryos with well-developed heads had little or no myc-ashwin expression in the head (Fig. 3D). In contrast, embryos with head defects had significant myc-ashwin expression in the head region (Fig. 3E).
Phenotypic abnormalities were also generated by overexpression of ashwin in the Organizer region. Microinjection of 0.2 ng/embryo RNA encoding myc-ashwin in the equatorial region of the two dorsal blastomeres led to two major phenotypes, in roughly equal proportions: the first group had distended anterior regions, open anterior neural folds, and a loss of eye development (Fig. 3G), whereas the remainder, in the second group, had anterior truncations as described earlier (Fig. 3H). This apparent contradiction was resolved by lineage labeling studies using the myc-tagged ashwin; in embryos with anterior truncations, the myc-ashwin was present in the dorsal mesoderm, including the notochord, whereas the embryos with distended heads and anterior neurulation defects had very high levels of myc-ashwin expression in the head ectoderm, especially the neural plate.
The effects of ventrally directed misexpression of ashwin were evaluated to determine whether ashwin is sufficient to elicit ectopic Organizer formation. Four-cell embryos were microinjected with 0.2 ng/embryo ashwin mRNA and monitored for second dorsal lip formation, secondary axis formation, and ectopic expression of goosecoid (gsc), expressed in the anterior Organizer (head mesoderm) during gastrulation and throughout the notochord at subsequent stages (Cho et al.,1991). Ventrally directed overexpression of ashwin produced phenotypically normal embryos in 90% of cases, with no evidence of a second dorsal lip or secondary axis (Table 1). Only 10% of the embryos showed a secondary dorsal lip, and only 2% formed a secondary axis visible by midneurula stage (Table 1; Fig. 3I). Ectopic expression of gsc was not observed (data not shown). Ventral misexpression of ashwin led to relatively normal dorsoventral patterning, in contrast to ventral misexpression of either noggin or β-catenin (data not shown).
The major class of phenotypic abnormalities resulting from ashwin overexpression were anterior truncations. Abnormalities in anterior pattern, observed in 40% of animally injected embryos, included a complete loss of head structures, eye defects, or asymmetrical development of bilateral structures. Embryos overexpressing ashwin were stained using the notochord-specific Tor70 antibody, before sectioning. In some cases, anterior tissues did not form, and the anterior end of the notochord coincided with the anterior end of the embryo, indicating a complete absence of the prechordal mesoderm (Fig. 4B). In others, tissue anterior to the notochord appeared undifferentiated (Fig. 4D). In milder cases, heads formed, but they were small, and the eyes were reduced or absent (Fig. 4E). Even in embryos in which formation of the forebrain and eye were highly abnormal, more posterior structures such as the otic vesicle appeared largely undisturbed (Fig. 4F).
To evaluate the effects of ashwin on anteroposterior neural pattern, embryos overexpressing ashwin were hybridized in situ with probes for otx2, expressed in anterior neural ectoderm, krox-20, expressed in rhombomeres 3 and 5 in the hindbrain, and HoxB9, expressed throughout the spinal cord (Fig. 5). At midneurula stages, otx2 expression in embryos overexpressing ashwin was indistinguishable from that seen in controls (Fig. 5A,C). At progressively later stages, however, otx2 expression decreased, and it was restricted to a small area centered at the midline in early tail bud embryos overexpressing ashwin (Fig. 5B,D). Expression patterns for krox-20 and HoxB9 were examined in embryos injected unilaterally with 200 pg/embryo ashwin mRNA (Fig. 5E). Krox-20 was expressed in a single stripe on the injected side; in contrast, HoxB9 expression was identical on the injected and uninjected sides. By early tail bud (st. 28), however, Krox-20 expression in embryos overexpressing ashwin was similar to that seen in control embryos, suggesting that ashwin overexpression leads to a delay in Krox-20 expression, rather than inhibition of Krox-20 expression or a loss of Krox-20–expressing tissue.
Formation of Axial Structures
Rudimentary secondary axial structures were seen in 28% of cases (Fig. 6A). These structures generally appeared at the level of the head, and they were most readily apparent at late neurula stages (e.g., st. 18–20). In many cases, secondary axial structures visible in neurulae did not expand further and were detectable as a fork ending in a small protrusion at the level of the hindbrain in early tail bud embryos. In some cases, however, sections of embryos that showed no overt secondary axial structure revealed a second notochord (Figs. 4F, 6H).
The formation of secondary axial structures in ashwin-injected embryos led us to ask whether overexpression of ashwin elicits formation of a second Organizer. We first analyzed dorsal lip formation at early gastrula stages when ashwin overexpression was directed to the animal pole. Whereas only 5% of these embryos had two dorsal lips, often located close to the primary lip (data not shown), approximately 22% of embryos showed disrupted primary dorsal lips. This phenotypic frequency is closely correlated with that of secondary axis formation, suggesting that ashwin may act to split the Organizer into two domains by means of a localized inhibitory mechanism. A splitting, rather than duplication, mechanism would also account for the presence of two small, closely apposed notochords in many ashwin-injected embryos. We also observed variations in the shape of the archenteron in midgastrula embryos overexpressing ashwin. In over 75% of embryos, the archenteron has a single leading edge, visible through the dorsal ectoderm after clearing in benzyl benzoate: benzyl alcohol (BB:BA; Fig. 6I). In other embryos, however, the tip of the archenteron appears to bifurcate (Fig. 6J). Such a bifurcation could be consistent with morphogenetic disruptions that might arise from splitting of the anterior Organizer.
We next examined the expression of two Organizer genes in gastrula embryos overexpressing ashwin. Embryos were injected in the animal pole with 0.25 ng/embryo mRNA encoding either ashwin or β-galactosidase. Embryos were collected at early gastrula stages, fixed, and hybridized in situ with probes for Xnot1, expressed in the posterior Organizer and notochord (Von Dassow et al.,1993) and the anterior Organizer marker gsc. Expression patterns for Xnot1 were essentially indistinguishable in embryos overexpressing ashwin and control embryos (Fig. 6C,D). In contrast, gsc expression was strongly reduced in ashwin-injected embryos (Fig. 6E,F). Because only 20–25% of embryos overexpressing ashwin form a secondary axis, we examined at least 50 embryos in four independent experiments for each probe to ensure detection of any consistent variations in expression pattern. No evidence of a second region of gsc or Xnot1 expression was observed.
Loss of gsc expression in embryos overexpressing ashwin could result from either a failure to establish anterior pattern or from a loss of cells committed to anterior fate, by means of either cell death or a lack of proliferation. To determine whether ashwin overexpression leads to a large-scale loss of cells, embryos were injected with 200 pg of mRNA encoding myc-ashwin, hybridized in situ with the gsc probe, and then stained to visualize expression of the myc-ashwin protein. Embryos were photographed before and after anti-myc immunohistochemistry, and individual embryos were tracked to evaluate the correspondence between gsc and myc-ashwin expression patterns. In all embryos, myc-ashwin was expressed across a large region, and there was no evidence of necrotic cells or other abnormalities. Gsc expression was barely detected in embryos in which the region of myc-ashwin expression spanned the entire dorsal marginal zone (Fig. 7A,B). In cases in which myc-ashwin was expressed across only part of the dorsal marginal zone, gsc expression was considerably stronger where myc-ashwin was absent. These findings indicate that cells overexpressing ashwin persist after the onset of gastrulation.
Reduction in ashwin Expression Causes Gastrulation Defects
We initiated a loss-of-function analysis of ashwin using a morpholino oligonucleotide (MO) directed against the exon–intron boundary at the 3′end of exon I. The MO was based on genomic sequence obtained by PCR amplification of genomic DNA across a region corresponding to mammalian exons 1 and 2. Injection of this MO (Ex1 MO) led to a partial inhibition of splicing of ashwin mRNA, as revealed by RT-PCR using primers located to either side of the boundary between exons 1 and 2 (data not shown). In preliminary experiments, a five-base mispair MO was used to establish the appropriate concentration for Ex1 MO, and the most consistent results were obtained at a concentration of 10 ng/embryo. Although the mispair sequence does not correspond to any known Xenopus sequence, we were unable to identify a concentration of the mispair MO that did not produce some abnormalities at a low frequency. A second MO corresponding to the ashwin translation start site (5′ MO) was also used in some experiments.
A partial reduction in Ashwin protein levels, using either the Ex1 or 5′ MOs, was either lethal or led to gastrulation defects in many cases (Fig. 8A,B). Gastrulation would be initiated normally, but convergence and extension movements would fail during the second half of gastrulation, leading either to arrest or incomplete blastopore closure. In most cases, embryos surviving past the end of gastrulation were often outwardly normal, but would then die from a rupture of yolk cells within the archenteron, visible as a release of cells or yolk from the interior. Additional defects, including reductions in head formation or embryo size, were also observed, but these varied greatly across different batches of embryos.
Open-face Keller explants of the dorsal marginal zone (DMZ) were isolated from embryos injected with 10 ng of Ex1 MO or the Ex1 mispair MO and observed to determine whether ashwin is required for convergence and extension within the dorsal marginal zone. Explants from embryos injected with either MO underwent normal morphogenetic movements and were indistinguishable from explants prepared from uninjected embryos (data not shown).
Alterations in the Level of ashwin Expression Lead to Apoptosis
Gastrula embryos injected with 10 ng of either the Ex1 MO or the standard control MO, as well as embryos injected with 200 pg/embryo of RNA encoding either ashwin or β-galactosidase, were subjected to terminal uridine nucleotide end labeling (TUNEL) assays. Reduction of ashwin levels led to a striking appearance of apoptotic nuclei throughout the animal ectoderm and marginal zone (Fig. 8B), as well as in cells scattered within the dorsal mesendoderm (data not shown); a lower level of apoptosis was observed in embryos injected with the 5′ MO (data not shown). Interestingly, overexpression of ashwin also generated apoptotic nuclei (Fig. 8C). Apoptotic nuclei were not observed in embryos injected with either the control MO (Fig. 8D) or β-galactosidase RNA (Fig. 8E).
Ectodermal Overexpression of ashwin Leads to Neural Gene Expression
Because ashwin was first isolated in a screen for genes showing differential expression in neural vs. non-neural ectoderm, we sought to determine whether ashwin could alter ectodermal gene expression. mRNA encoding either ashwin or β-galactosidase was microinjected into the animal pole region of two-cell embryos. Animal caps were isolated at the late blastula stage and cultured until controls reached late neurula (st. 18), then collected for RNA isolation and RT-PCR assays. Overexpression of ashwin in isolated ectoderm leads to expression of NCAM (Fig. 9A). Coexpression of ashwin and noggin leads to synergistic up-regulation of noggin, zic3, and NCAM. Animal caps did not undergo visible phenotypic changes, such as elongation or extensive vesiculation; they did, however, begin to disintegrate after controls reached the end of neurulation, consistent with the onset of epidermal shedding in whole embryos overexpressing ashwin.
The effects of ashwin on the expression of several BMP4-inducible genes were also examined, because initiation of neural gene expression in vivo involves a down-regulation of BMP4 signals. Embryos were microinjected with 0.25 ng of ashwin mRNA, and ectoderm was isolated at the late blastula stage and cultured until intact controls reached stage 12. Ectodermal isolates were then collected for RNA isolation and RT-PCR assays. Overexpression of ashwin has no significant effect on the expression of the BMP4-inducible genes Xvent1 or msx1, nor does it alter expression of BMP4 itself (Fig. 9B), which is up-regulated by means of a Smad1-dependent positive feedback loop (Schuler-Metz et al.,2000). These results indicate that ashwin promotes the establishment of neural fate without disrupting BMP4-inducible gene expression.
Our results identify ashwin, a novel gene so far detected only in vertebrates, as a regulator of cell survival and axial pattern. We have been unable to identify genes that include any of the conserved regions of ashwin from the Drosophila, Caenorhabditis elegans, or Ciona genomes, suggesting that this gene may have arisen in a lineage leading to vertebrates. The range of expression in vivo coupled with the specificity of the overexpression and knockdown phenotypes suggests that ashwin may play distinct roles in different tissues and invites the speculation that the specificity of ashwin action is dependent upon as yet-unidentified binding partners.
Ashwin as a Regulator of Cell Survival
The appearance of apoptotic cells in gastrula embryos under conditions in which ashwin protein levels are altered indicates that ashwin regulates cell survival during gastrulation and suggests that the level of ashwin is critical for the balance of pro- and anti-apoptotic processes. The appearance of apoptosis under both gain- and loss-of-function conditions is consistent with a biochemical model in which ashwin is required to link two proteins into a functional complex: when ashwin levels are too low, both components are largely unbound. At high ashwin levels, each component may be bound by different pools of ashwin molecules; the abundance of ashwin protein titrates out the other components, and the ternary complex does not form. While evaluation of this model awaits the identification of ashwin binding partners, now in progress, Bix3 has been identified as another apoptotic regulator that triggers apoptosis under both gain- and loss-of-function conditions in Xenopus embryos (Trinidade et al.,2003).
The widespread onset of apoptosis could account for the failure of gastrulation in embryos injected with ashwin MOs. Whereas convergence and extension in dorsal mesoderm are unaffected by a reduction in ashwin function, these movements do not spread properly to the lateral and ventral marginal zones, leading to a failure of blastopore closure. Apoptosis within the marginal zone could disrupt the integration of tensional mechanical forces within the marginal zone; this “hoop stress” is required for the coordination of cell intercalation behavior in a sequential manner throughout the marginal zone, which in turn generates convergence and extension (reviewed in Keller et al.,2003).
Although apoptosis may account for some of the phenotypic abnormalities observed after overexpression or reduction of ashwin, several findings suggest that some of the effects of ashwin alteration are independent of apoptosis. First, overexpression of ashwin alone or in combination with noggin produces changes in gene expression in isolated ectoderm. Second, overexpression also leads to incomplete neural tube closure, which cannot directly be attributed to increased apoptosis. Finally, some embryos with anterior defects had considerable tissue in the anterior region that failed to show evidence of anterior identity or morphological differentiation. These findings, as well as studies now in progress, indicate that ashwin may act by means of other processes in addition to apoptosis.
ashwin May Act as a Negative Regulator of Head Formation
Our results suggest that ashwin may play a role in the regulation of anteroposterior pattern. Overexpression of ashwin in dorsal mesoderm or ectoderm leads to reduction, loss, or abnormal development of eye and forebrain structures, suggesting that it may have a negative regulatory effect on anterior neural development or cell survival within the anterior neural ectoderm or the anterior Organizer. Ashwin overexpression could alter the balance of cell proliferation and apoptosis in the anterior region. Alternatively, ashwin could disrupt head formation by acting within the dorsal mesoderm, either to promote expression of genes that block or override the establishment of the head Organizer, or to disrupt the expression or activity of genes involved in head formation. Head formation requires an inhibition of BMP4 and Wnt signals (Glinka et al.,1997). Key genes include the multifunctional BMP4/Wnt/Nodal antagonist cerberus (Piccolo et al.,1999), the Wnt antagonists dickkopf (Glinka et al.,1998) and frzb (Wang et al.,1997; Leyns et al.,1997), as well as transcriptional regulators such as Hex (Jones et al.,1999; Brickman et al.,2000), Lim1 (Taira et al.,1997), and gsc (Yao and Kessler,2001).
ashwin and Axis Formation
Most instances of experimentally induced secondary axis induction occur by means of ventrally directed misexpression. Secondary axes can be induced by canonical Wnt pathway components (reviewed in Moon and Kimelman,1998), some transforming growth factor-beta (TGF-β) family members such as activin (Thomsen et al.,1990), or targets of Wnt and/or TGF-β family signaling such as gsc (Cho et al.,1991) or siamois (Lemaire et al.,1995). The TGF-β family member derriere induces a secondary axis lacking anterior structures (Sun et al.,1999) as does an activated form of FAST-1, a transcription factor that participates in the induction of mesoderm by Smad2-dependent TGF-β family signals, such as activin, nodals, or Vg1 (Watanabe and Whitman,1999). Ventral misexpression of BMP antagonists can also lead to secondary axis formation (e.g., Schmidt et al.,1995), as does expression of Xiro-1, an inhibitor of BMP4 transcription (Glavic et al.,2001).
In contrast, the secondary axis phenotype generated by ashwin overexpression differs from these examples in several ways: (1) ashwin induces secondary axial structures more frequently when overexpressed animally, rather than ventrally; (2) frequency of second dorsal lip formation in ashwin-injected embryos does not correlate with the frequency of secondary axial structures; (3) the frequency of secondary axial structures is similar between RNA-injected embryos, which express ashwin early, or DNA-injected embryos, which express ashwin after MBT; (4) neither ventral nor animal ashwin overexpression is able to induce gsc expression in the ventral region; and (5) the morphology of the primary dorsal lip is often perturbed by animally directed ashwin overexpression. Furthermore, overexpression of ashwin in isolated ectoderm does not mimic the effects of Smad2-dependent TGF-β family signaling, nor does it antagonize BMP4 signals. Together, these results suggest that ashwin acts downstream of early dorsal axis specification and cannot produce a second Organizer region.
Instead, the formation of secondary axial structures in ashwin-injected embryos may reflect a splitting of the original Organizer region. We speculate that a local loss of anterior mesoderm cells or inhibition of anterior identity within one region of the Organizer could lead to a loss of integration with flanking regions of the Organizer. For example, ashwin disrupts gsc expression, and gsc has been shown to promote anteriorward migratory movement in cells within the anterior Organizer (Niehrs et al.,1993). Therefore, the reduction in gsc expression in ashwin-injected embryos could disrupt position-specific behavior or identity within the anterior Organizer, resulting in some of the diverse phenotypes arising from ashwin overexpression. (1) Medial disruptions in the Organizer could lead to splitting of the notochord, as normal Organizer cells migrate and extend around ashwin-expressing cells. (2) Asymmetric accumulation of ashwin-expressing cells in the Organizer could lead to curving of the notochord, generating asymmetric signaling to the overlying neurectoderm at later stages (3). Disruption of the entire anterior Organizer could lead to anterior truncations. The identification of binding partners and downstream targets of ashwin and the analysis of ashwin function in the mesoderm vs. ectoderm will delineate the role of ashwin in early development.
Microinjection and Microsurgery
Preparation of embryos, microsurgical manipulations, and treatment of isolated ectoderm with NH4Cl or purified recombinant Xenopus basic FGF were performed as described (Uzman et al.,1998, Uzgare et al.,1998). For overexpression studies, synthetic mRNA was prepared by in vitro transcription of linearized plasmid DNA, using the mMessage Machine in vitro transcription system (Ambion, Austin, TX). Unless otherwise indicated, approximately 10 nl of mRNA stock at the appropriate concentration was pressure-injected near the animal pole into each blastomere at the two-cell stage. For dorsally or ventrally directed injections, embryos were injected in the equatorial region of dorsal or ventral blastomeres at the four-cell stage, unless otherwise indicated. Open-face Keller explants of the dorsal marginal zone were prepared and cultured as described previously (Sater et al.,1993).
Differential Display RT-PCR and Colony Screens
Differential Display RT-PCR (DD-RT-PCR) was carried out using the RNAimage kit (GenHunter Corp., Nashville, TN) and the Genomyx Differential Display Gel System (Genomyx, Palo Alto, CA), both according to manufacturer's instructions. For each of the five samples tested, RT-PCR was carried out using specific combinations of anchored and arbitrary primers. DNA fragments that appeared exclusively in presumptive neural tissue were excised, re-amplified, and collectively cloned into the pGEM-T Easy vector system (Promega, Madison, WI). Recombinant clones were screened by colony hybridization (Sambrook et al.,1989) using Hybond N nylon filters (Amersham). Colonies that showed differential hybridization to neural ectoderm cDNA were picked and replated for another round of screening by means of colony hybridization. Confirmed positive clones were retained for use as probes for Northern analysis and in situ hybridization.
Isolation and Subcloning of Full-Length ashwin
A gastrula-stage cDNA library in pBluescript SK- (gift of D. DeSimone) was screened to obtain a full-length cDNA clone for one confirmed positive clone (clone 7). This clone represented an mRNA of novel sequence and specific expression in the early neural plate, based on DNA sequencing, Northern analysis and in situ hybridization. Positive plaques were isolated and sequenced. The clone containing the longest open reading frame was designated clone 11a.
Sequence data from clone 11a were compared with entries in the GenBank database using the Basic Local Alignment Search Tool (BLAST). Retrieved sequences were aligned using CLUSTAL W. Strong sequence similarity to putative human and mouse proteins indicated that additional 5′ coding sequence existed. To obtain the full-length 5′ end of ashwin, nested PCR was performed on a neurula-stage cDNA library (T. Mohun) using the high-fidelity polymerase Pwo (Roche, Indianapolis, IN). PCR products were gel purified (Qiagen, Valencia, CA) and cloned into the TA cloning vector PCR4-TOPO (Invitrogen, Carlsbad, CA). Clones were sequenced using T3 and SP6 primers (Lone Star Labs, Inc., Houston, TX). Sequence analysis revealed a 101-bp 5′ fragment that has 85% and 82% identity to mouse and human cDNAs, respectively. Full-length ashwin was amplified from st. 10 cDNA using primers with added restriction sites, and subsequently cloned into the pCS2+ vector with and without myc-epitope tags.
Northern blots were prepared using total RNA from two oocytes or embryos at each stage. The RNA samples were separated on formaldehyde gels, transferred to nylon membrane (HyBond-N, Amersham, Piscataway, NJ), and fixed to the membrane with ultraviolet crosslinking. The Northern blots were hybridized with a 32P-labeled antisense ashwin probe generated using the Strip-EZ PCR kit (Ambion, Austin, TX). Hybridization and washing were carried out as described in Seshadri et al. (1993).
RNA isolation, preparation of cDNA, and PCR assays were carried out as described (Uzman et al.,1998). PCR primers for Zic3 and N-tubulin were taken from the Xenopus Molecular Marker Resource (http://vize222.zo.utexas.edu). Additional primers were obtained from the following sources: Msx1 (Suzuki et al.,1997), Xvent1 (Gawantka et al.,1995). PCR primers for BMP4 were Forward-GGCTGGAATGATTGGATTGTGG; Reverse-ATCTCAGACTCAACGGCACC; product size of 276. The RT-PCR products were separated on 6% polyacrylamide gels and visualized by autoradiography.
The genomic sequence for the X. laevis ashwin gene was compared with that of mammalian orthologues to identify the boundary between exon I and the first intron within the coding region; a sequence across this boundary was identified and obtained from Gene Tools (Philomath, OR). This sequence of the exon1 morpholino oligonucleotide was 5′-AATGGCGAGGGCGCCTCACCCGCTC-3′. In many experiments, a five-base mispair morpholino oligonucleotide was used as a control; the sequence of the five-base mispair was 5′-AATCGCGAGGGCGCCTCACGCGGTC-3′ (mispaired bases underlined). A morpholino oligonucleotide directed against the translation start site was also used in some experiments; the sequence for this 5′ MO was 5′-CAGCACTGCCAACAGCTTTAGCAT-3′.
In Situ Hybridization, Immunohistochemistry, Histology, and TUNEL Assay
Probes for in situ hybridization were prepared using the Maxiscript in vitro transcription kit (Ambion). In situ hybridization was carried out as described in Harland (1991), using the modifications suggested by Cox and Hemmati-Brivanlou (1995). Immunohistochemistry was performed as described in Sater et al. (1993). In some cases, embryos were cleared by immersion in 2:1 benzyl benzoate: benzyl alcohol (BB:BA), as described in Sive et al. (2000). Immunostained embryos were fixed in MEMFA, dehydrated, and embedded in paraffin, then sectioned at 10 μm and mounted in Permount. TUNEL assays were carried out as described in Van Stry et al. (2004), using the Cell Death Detection Kit – AP (Roche Diagnostics).
We thank Doug DeSimone and Tim Mohun for cDNA libraries, Heithem El-Hodiri for advice and comments on the manuscript, Paul Hardin and Arnold Eskin for advice regarding differential display, Richard Harland for the Tor70 monoclonal antibody, Karen Symes for the TUNEL protocol, and Richard Harland, Ken Cho, Eddy De Robertis, David Wilkinson, David Kimelman, Gerry Thomsen, and Pierre McCrea for cDNA clones. The Xen1 monoclonal antibody developed by Ariel Ruiz i Altaba was obtained from the Developmental Studies Hybridoma Bank developed under the auspices of the NICHD and maintained by The University of Iowa, Department of Biological Sciences (Iowa City, IA). The Joint Genome Institute supported by the Department of Energy provided fugu sequence data. A.K.S. and J.A.U. were funded by Texas Advanced Research Program and A.K.S. was funded by the NSF.
- 1999. Wnt signaling in Xenopus embryos inhibits bmp4 expression and activates neural development. Genes Dev 13: 3149–3159. , , .
- 1995. Anterior neurectoderm is progressively induced during gastrulation: the role of the Xenopus homeobox gene orthodenticle. Development 121: 993–1004. , .
- 2000. Hex is a transcriptional repressor that contributes to anterior identity and suppresses Spemann organiser function. Development 127: 2303–2315. , , , , .
- 1998. Xenopus Smad7 inhibits both the activin and BMP pathways and acts as a neural inducer. Dev Biol 198: 1–12. , .
- 1991. Molecular nature of Spemann's Organizer: the role of the Xenopus homeobox gene goosecoid. Cell 67: 1111–1120. , , , .
- 1995. Caudalization of neural fate by tissue recombination and bFGF. Development 121: 4349–4358. , .
- 1995. Antagonizing the Spemann Organizer: role of the homeobox gene Xvent-1. EMBO J 14: 6268–6279. , , , , .
- 2001. Xiro-1 controls mesoderm patterning by repressing BMP4 expression in the Spemann Organizer. Dev Dyn 222: 368–376. , , .
- 1997. Head induction by simultaneous repression of Bmp and Wnt signalling in Xenopus. Nature 389: 517–519. , , , , .
- 1998. Dickkopf-1 is a member of a new family of secreted proteins and functions in head induction. Nature 391: 357–362. , , , , , .
- 1991. In situ hybridization: an improved whole-mount method for Xenopus embryos. Methods Cell Biol 36: 685–695. .
- 2000. Neural induction. Curr Opin Genet Dev 10: 357–362. .
- 1999. FGF is required for posterior neural patterning but not for neural induction. Dev Biol 205: 296–308. , .
- 1999. FGF signaling and the anterior neural induction in Xenopus. Dev Biol 216: 561–581. , , .
- 1994. Transcription factors. 1: bZIP proteins. Protein Profile 1: 123–168. .
- 1999. An anterior signalling centre in Xenopus revealed by the homeobox gene XHex. Curr Biol 9: 946–954. , , , , .
- 2003. How we are shaped: the biomechanics of gastrulation. Differentiation 71: 171–205. , , .
- 1990. Improvements in protein secondary structure prediction by an enhanced neural network. J Mol Biol 214: 171–182. , , .
- 1997. Opposing BMP and EGF signaling pathways converge on the TGF-β family mediator Smad1. Nature 389: 618–622. , , .
- 1998. Opl: a zinc finger protein that regulates neural determination and patterning in Xenopus. Development 125: 2867–2882. , , , , , , .
- 1995. Fibroblast growth factor is a direct neural inducer, which combined with noggin generates anterior-posterior neural pattern. Development 121: 3627–3636. , .
- 1995. Expression cloning of Siamois, a Xenopus homeobox gene expressed in dorsal-vegetal cells of blastulae and able to induce a complete secondary axis. Cell 81: 85–94. , , .
- 1997. Frzb-1 is a secreted antagonist of Wnt signaling expressed in the Spemann Organizer. Cell 88: 747–756. , , , , .
- 1995. Specification of the anteroposterior neural axis through synergistic interaction of the Wnt signaling cascade with noggin and follistatin. Dev Biol 172: 337–342. , , .
- 1998. From cortical rotation to Organizer gene expression: toward a molecular explanation of axis specification in Xenopus. Bioessays 20: 536–545. , .
- 1998. Smad6 functions as an intracellular antagonist of some TGF-beta family members during Xenopus embryogenesis. Genes Cells 3: 387–394. , , , .
- 1993. The homeobox gene goosecoid controls cell migration in Xenopus embryos. Cell 72: 491–503. , , , .
- 1995. The Xenopus homologue of Otx2 is a maternal homeobox gene that demarcates and specifies anterior body regions. Development 121: 707–720. , , , , , , .
- 2001. Neural and head induction by insulin-like growth factor signals. Dev Cell 1: 655–665. , , , .
- 1999. The head inducer Cerberus is a multifunctional antagonist of Nodal, BMP, and Wnt signals. Nature 397: 707–710. , , , , , , .
- 2002. The IGF pathway regulates head formation by inhibiting Wnt signaling in Xenopus. Dev Biol 244: 407–417. , , , , .
- 1989. Molecular cloning: a laboratory manual. 2nd ed. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press. , , .
- 1993. Induction of neuronal differentiation by planar signals in Xenopus embryos. Dev Dyn 197: 268–280. , , .
- 1995. Drosophila short gastrulation induces an ectopic axis in Xenopus: evidence for conserved mechanisms of dorsal-ventral patterning. Development 121: 4319–4328. , , , .
- 2000. The homeodomain transcription factor Xvent-2 mediates autocatalytic regulation of BMP4 expression in Xenopus embryos. J Biol Chem 275: 34365–34374. , , , .
- 1993. Identification of a transcript that is down-regulated in senescent human fibroblasts. Cloning, sequence analysis, and regulation of the human L7 ribosomal protein gene. J Biol Chem 268: 18474–18480. , , , .
- 2000. Early development of Xenopus laevis: a laboratory manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press. , , .
- 2004. Neural induction. In: SternCD, editor. Gastrulation: from cells to embryo. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press. p 419–432. .
- 2001. FoxD5a, a Xenopus winged helix gene, maintains an immature neural ectoderm via transcriptional repression that is dependent on the C-terminal domain. Dev Biol 232: 439–457. , , .
- 1999. derriere: a TGF-beta family member required for posterior development in Xenopus. Development 126: 1467–1482. , , , , , , , .
- 1997. Xenopus msx1 mediates epidermal induction and neural inhibition by BMP4. Dev Biol 124: 3037–3044. , , .
- 1997. Role of the Xim-1 and Xbra genes in anteroposterior patterning of neural tissue by the head and trunk Organizer. Proc Natl Acad Sci U S A 94: 895–900. , , .
- 1990. Activins are expressed early in Xenopus embryogenesis and can induce axial mesoderm and anterior structures. Cell 63: 485–493. , , , , , .
- 2003. Regulation of apoptosis in the Xenopus embryo by Bix3. Development 130: 4611–4622. , , , , , , .
- 1998. Mitogen-activated protein kinase and neural specification in Xenopus. Proc Natl Acad Sci U S A 95: 14833–14838. , , , .
- 1998. The role of intracellular alkalinization in the establishment of anterior neural fate in Xenopus. Dev Biol 193: 10–20. , , , .
- 2004. The mitochondrial-apoptotic pathway is triggered in Xenopus mesoderm cells deprived of PDGF receptor signaling during gastrulation. Dev Biol 268: 232–242. , , , .
- 1993. Induction of the Xenopus organizer: expression and regulation of Xnot, a novel FGF and activin-inducible homeobox gene. Genes Dev 7: 355–366. , , .
- 1997. Frzb, a secreted protein expressed in the Spemann Organizer, binds and inhibits Wnt-8. Cell 88: 757–766. , , , ,
- 1999. FAST-1 is a key maternal effector of mesoderm inducers in the early Xenopus embryo. Development 126: 5621–5634. , .
- 1999. Neural induction. Annu Rev Cell Dev Biol 15: 411–433. , .
- 2001. Goosecoid promotes head Organizer activity by direct repression of Xwnt8 in Spemann's Organizer. Development 128: 2975–2987. , .
- 1992. A novel homeobox gene expressed in the anterior neural plate of the Xenopus embryo. Dev Biol 152: 373–382. , , , , , .