Nodal regulates neural tube formation in the Ciona intestinalis embryo
- First Online:
- Cite this article as:
- Mita, K. & Fujiwara, S. Dev Genes Evol (2007) 217: 593. doi:10.1007/s00427-007-0168-x
- 272 Views
Overexpression of a lefty orthologue, Ci-lefty, caused a failure of neural tube closure in the protochordate ascidian Ciona intestinalis. The body bent dorsally, and anterior–posterior elongation was inhibited. A similar phenotype was observed in embryos treated with SB431542, an inhibitor of Nodal receptors, suggesting that Ci-Lefty antagonized Nodal signaling as reported in other deuterostome species. Overexpression of Ci-nodal also resulted in a similar phenotype, suggesting that a correct quantity and/or a spatial restriction of Nodal signaling are important for the neural tube to form. In addition to known Ci-Nodal target genes, orthologues of Zic (Ci-ZicL) and cdx (Ci-cdx) were activated by Ci-Nodal. Expression of a dominant negative Ci-cdx caused defects in neural tube formation similar to those obtained on treatment with SB431542 or overexpression of Ci-lefty. A regulatory cascade composed of Ci-Nodal, Ci-ZicL, and Ci-Cdx may play an important role in neural tube formation in the Ciona embryo.
KeywordsAscidianNeural tube formationLeftyNodalCdxZicL
Inductive cellular interactions mediated by growth factors are important for embryogenesis. nodal, a member of the transforming growth factor-β (TGF-β) superfamily, has been identified exclusively in deuterostomes, including vertebrates, amphioxus, ascidians, and sea urchins (Chea et al. 2005). No counterpart is present in genomes of protostomes, such as Caenorhabditis elegans and Drosophila melanogaster (Boorman and Shimeld 2002). In vertebrates, Nodal signaling is required for multiple developmental processes, such as specification of the endoderm and mesoderm, neural patterning, and the formation of the anterior–posterior and left–right body axes (Whitman 2001; Schier and Shen 2000).
lefty, also named antivin, belongs to the TGF-β superfamily and is known as an antagonist of Nodal in vertebrates (Juan and Hamada 2001). Orthologues of lefty have also been reported only in deuterostomes (Duboc et al. 2004; Fujiwara et al. 2002; Imai 2003). Overexpression of sea urchin Antivin/Lefty leads to an abnormal oral–aboral axis, indistinguishable from the phenotypes obtained by the injection of nodal-specific antisense morpholino oligonucleotides (MO; Duboc et al. 2004). Therefore, Antivin/Lefty seems to inhibit Nodal signals in sea urchin embryos (Duboc et al. 2004). Expression of Ci-lefty, a C. intestinalis orthologue of lefty, starts from the 16-cell stage (Imai et al. 2004). Ci-lefty and Ci-nodal show a similar temporal expression pattern, suggesting an interaction. However, the function of Ci-lefty is still unclear.
In the present study, we examined the role of the Nodal signal in the formation of the neural tube in C. intestinalis embryos. In Xenopus laevis, a disturbance of nodal-related genes (Xnr2, Xnr3, and Xnr5) causes an abnormal neural tube to form (Osada and Wright 1999; Onuma et al. 2002; Yokota et al. 2003). However, there is no report concerning a possible role of Nodal function in neural tube formation in other deuterostome species. We show here that Ci-Lefty antagonizes Nodal signaling. When Ci-lefty was over-expressed in embryos, closure of the neural tube was incomplete and the body curved dorsally. Overexpression of Ci-nodal caused similar abnormal phenotypes. The inhibition or promotion of Nodal signaling affected the expression pattern of Ci-cdx and Ci-ZicL in the A-line neural cells at the late gastrula stage. Overexpression of a dominant negative form of Ci-cdx disturbed neural tube formation. These results suggest that Nodal signaling is involved in the formation of the neural tube in the C. intestinalis embryo.
Materials and methods
Adult C. intestinalis was collected around the Uranouchi Inlet near the Usa Marine Biological Institute. Juvenile adults were kindly provided by Kazuko Hirayama and Nori Satoh at Kyoto University. They were cultivated for a few months in corves. Gametes were surgically obtained from the gonoducts of mature adults. Eggs were inseminated with non-self sperm. Fertilized eggs were dechorionated with 0.05% actinase E (Kaken Pharmaceutical) and 1% sodium thioglycolate.
A genomic DNA fragment encompassing the 5′ flanking region of Ci-FoxD was amplified by polymerase chain reaction (PCR) using the primers 5′-AAAGTCTCGAGCACAAATATAGCGGTTTTGAAGTC-3′ and 5′-AAAAACCCGGGCCATCATCACACAACGGATTCGAT-3′. The PCR product was digested with XhoI and XmaI, and inserted into a lacZ-containing plasmid vector, 72-1.27 (Corbo et al. 1997b). This construct was named Ci-FoxD/lacZ. A Ci-nodal complementary DNA (cDNA) fragment was amplified by reverse transcription-PCR using the C. intestinalis embryonic RNA as a template. The primers used were 5′-ATATGATTTCTATGTTTAATATCGCTGC-3′ and 5′-AAGAATTCTTATCTGCAACCGCATTCG-3′. The PCR product was inserted into pGEM-T (Promega). The cDNA was then excised from pGEM-T using EcoRI and ApaI, and inserted into pBluescript II SK+ (Stratagene). With this construct as a template, the translated region was amplified by PCR using 5′-AATGTCCCGGGCGGCTGCTTTCGTCTTCACT-3′ and 5′-AAGAATTCTTATCTGCAACCGCATTCG-3′. A Ci-lefty cDNA fragment was amplified by PCR using a cDNA clone (ID cicl007p08; Fujiwara et al. 2002) as a template. The primers used were 5′-TATGTCCCGGGCGAAGACTTTCTTACTTATA-3′ and 5′-AAGAATTCTTACACACCAAATACACTGTCCAGGGA-3′. The Ci-lefty and Ci-nodal cDNA fragments were digested with XmaI and EcoRI. Ci-FoxD/lacZ was also excised with XmaI and EcoRI, and the lacZ translated region was substituted with the cDNA fragments. These constructs were named Ci-FoxD/Ci-lefty and Ci-FoxD/Ci-nodal, respectively.
For the construction of a dominant negative Ci-cdx, a Ci-cdx cDNA fragment was amplified by PCR using a cDNA clone (ID citb004h19; Satou et al. 2002b) as a template. The primers used were 5′-ATGTTACGTAACCCCACCCCGGACCCGAA-3′ and 5′-AAACTCGAGGTCATTTCCGTTGCCGTG-3′. The PCR product was digested with SnaBI and XhoI. The cohesive end produced by XhoI was ligated with a cDNA fragment encoding the repression domain of D.melanogaster Engrailed (EnR; Jaynes and O’Farrell 1991). The other end of the EnR fragment was cut with EcoRI. Ci-FoxD/lacZ was digested with SmaI and EcoRI, and the lacZ translated region was substituted with the Ci-cdx-EnR fusion cDNA.
Nucleotide sequences were determined using a BigDye Terminator v3.1 cycle sequencing kit and PRISM 3100-Avant Genetic Analyzer (Applied Biosystems). Plasmid DNA was purified using QIAGEN tip-100 (Qiagen). Transgenes were introduced into dechorionated C. intestinalis embryos by electroporation as described by Corbo et al. (1997b).
A stock solution of SB431542 (Sigma) was prepared at a concentration of 5 mM in dimethylsulfoxide (DMSO). The stock solution was diluted with filtered seawater. Embryos were continuously treated with 1–5 μM SB431542 from the 16-cell stage until they were fixed. Control embryos were reared in filtered seawater containing 0.1% (v/v) DMSO.
Whole-mount in situ hybridization
RNA probes were labeled with digoxigenin (DIG) as described by Nagatomo et al. (2003).
Templates for the synthesis of the probes were cDNA clones obtained from the C. intestinalis Gene Collection Release 1 (Satou et al. 2002b) and from cDNA clones analyzed by Fujiwara et al. (2002). The cDNA clones used were cicl002e04 (Ci-ZicL), cicl016e09 (Ci-chordin), cieg005o22 (Ci-Delta-like), cign044b23 (Ci-Mnx), cilv005f18 (Ci-cdx), cilv050a24 (Ci-FoxC), and citb028e11 (Ci-ETR; Satou et al. 2002a; the C. intestinalis cDNA project web site: http://ghost.zool.kyoto-u.ac.jp/indexr1.html). A Ci-sna cDNA (Corbo et al. 1997a) and the above-mentioned pBluescript II SK+ containing the Ci-nodal cDNA were also used to prepare the probe. T3 RNA polymerase (Promega) was used for the synthesis of an antisense Ci-nodal probe. For the other clones, T7 RNA polymerase (Takara) was used. The entire translated region of lacZ was obtained by cutting pSV-β-galactosidase vector (Promega) using BamHI and HindIII. This fragment was inserted in pBluescript II SK+. T3 RNA polymerase was used for the synthesis of an antisense lacZ probe.
Embryos were fixed with 4% paraformaldehyde in 0.1 M MOPS (pH 7.5) and 0.5 M NaCl at 4°C for more than 8 h and stored in 80% ethanol at −30°C. After equilibration with phosphate-buffered saline containing 0.1% Tween 20 (PBST), the embryos were treated with 2 μg/ml proteinase K in PBST for 30 min at 37°C. They were re-fixed with 4% paraformaldehyde in PBS (PBST without Tween 20) at room temperature for 1 h. The specimens were washed with PBST and incubated in a hybridization solution [50% formamide (FA), 5× SSC (750 mM NaCl and 75 mM sodium citrate, pH 7.0), 100 μg/ml yeast RNA (Roche), 5× Denhardt’s solution (Eppendorf), 0.1% Tween 20] at 56°C for 2 h. Hybridization with DIG-labeled probes was performed at 56°C for 18 h. The specimens were washed twice with 4× Wash (4× SSC, 50% FA, and 0.1% Tween 20) at 60°C for 15 min, and twice with 2× Wash (2× SSC, 50% FA, and 0.1% Tween 20) at 60°C for 15 min. They were immersed twice in Solution A [0.5 M NaCl, 10 mM Tris–HCl (pH 8.0), 5 mM ethylenediamine tetraacetic acid, and 0.1% Tween 20] at 60°C for 10 min. The samples were incubated at 37°C in Solution A containing 25 μg/ml RNaseA for 30 min. They were washed at 60°C sequentially with Solution A, 2× Wash, 1× Wash (1× SSC, 50% FA, and 0.1% Tween 20), and a 1:1 mixture of 1× Wash and PBST for 10 min each. They were then washed four times with PBST at 60°C for 10 min. The hybridization signal was immunologically detected essentially according to Fujiwara et al. (2002), except that blocking was carried out overnight at 4°C and incubation with anti-DIG antibody (Roche) was performed for 1 h at room temperature.
Morphological abnormality caused by overexpression of Ci-lefty and Ci-nodal
To analyze the roles of Nodal signaling in the C. intestinalis embryo, we caused Ci-lefty and Ci-nodal to be overexpressed. A single lefty/antivin orthologue was identified in the C. intestinalis genome (Figure S1; Hino et al. 2003). The gene’s ID at the C. intestinalis genome project web site (http://genome.jgi-psf.org/Cioin2/Cioin2.home.html) is gw1.03q.154.1 (Dehal et al. 2002). We refer to this gene as Ci-lefty in this paper, as “lefty” is used in many species. Normal expression of Ci-lefty starts at the 16-cell stage in A5.1, A5.2, and B5.1 blastomeres (Imai et al. 2004; Fig. 1d). Normal expression of Ci-nodal starts at the 32-cell stage in A6.1, A6.3, B6.1, and b6.5 blastomeres (Imai et al. 2004; Fig. 1f,g). We decided to use the Ci-FoxD promoter/enhancer for driving transgene expression. Ci-FoxD is expressed at the 16-cell stage in A5.1, A5.2, and B5.1 blastomeres (Imai et al. 2004). The C. intestinalis genome contains two closely linked Ci-FoxD genes (gw1.08q.1850.1 and gw1.08q.1853.1) on chromosome 8 (Fig. 1a). We isolated a 3-kb genomic DNA fragment that contained a promoter region of gw1.08q.1850.1, including putative transcription and translation initiation sites (Fig. 1a). Imai et al. (2002) showed that binding sites for a transcription factor Tcf were necessary for activation of a FoxD gene in another ascidian species Ciona savignyi. Two of the Tcf-binding sites were also conserved in the promoter region of gw1.08q.1850.1 (Fig. 1b). This DNA fragment was fused in-frame to lacZ and introduced into the one-cell embryo by electroporation. The transgene, named Ci-FoxD/lacZ, was expressed in A5.1, A5.2, and B5.1 blastomeres of the 16-cell embryo, as revealed by in situ hybridization (Fig. 1c,d). The result indicates that the 3-kb upstream region recapitulates initial transcriptional activation of endogenous Ci-FoxD gene. The lacZ messenger RNA (mRNA) was detected in all the daughter cells at the 32-cell stage (Fig. 1e–g). This expression pattern was similar to that of endogenous Ci-lefty and Ci-nodal (Fig. 1c–g). The lacZ mRNA was expressed broadly in the vegetal hemisphere at later stages (Fig. 1h,i). The lacZ translated region of Ci-FoxD/lacZ was substituted with Ci-lefty or Ci-nodal cDNA fragments. The resultant genes were named Ci-FoxD/Ci-lefty and Ci-FoxD/Ci-nodal, respectively.
Embryos treated with SB431542, an inhibitor of the Nodal receptors, ALK4, ALK5, and ALK7 (Inman et al. 2002), from the 16-cell stage showed dose-dependent defects in morphology (Fig. 2n–q). A high dose (5 μM) of SB431542 strongly inhibited gastrulation movement (Hudson and Yasuo 2005; Fig. 2q). Embryos treated with 1–2 μM SB431542 exhibited a morphology similar to that of embryos electroporated with the Ci-FoxD/Ci-lefty transgene (Fig. 2o,p). The neural tube did not close, anterior–posterior elongation was prevented, and the body curved dorsally (Fig. 2o–q). These results suggest the overexpression of Ci-lefty specifically blocked Nodal signaling.
Alteration of gene expression pattern by overexpression of Ci-lefty and Ci-nodal
Ci-Nodal activates Ci-ZicL and Ci-cdx expression in A-line nerve cord lineage cells
A cdx homologue (Hrcad) is required for the neural tube to form in H. roretzi (Katsuyama et al. 1999). Embryos injected with a Hrcad-specific phosphorothioate antisense oligo or dominant negative form show defects in neural tube formation and tail elongation, and their body curves dorsally (Katsuyama et al. 1999). This phenotype seems similar to that obtained by inhibiting Nodal signaling (see Fig. 2). We examined the expression of a C. intestinalis homologue of cdx (Ci-cdx) in transgenic embryos. Ci-cdx is expressed normally in A9.15, A9.29, and A9.31 blastomeres in late gastrulae (Imai et al. 2006). The Ci-FoxD/Ci-lefty transgene mediated repression of Ci-cdx expression (Fig. 3e). SB431542 had a similar effect on Ci-cdx expression (Fig. 3e). In contrast, the Ci-FoxD/Ci-nodal transgene induced ectopic Ci-cdx expression in A9.13 cells (Fig. 3e).
Expression of Ci-cdx in the late gastrula was repressed when the one-cell embryo was injected with a Ci-ZicL-specific MO (Imai et al. 2006). Ci-ZicL is expressed normally in A9.15 and A9.29 nerve cord-lineage cells, as well as in some a-line and b-line neural-lineage cells (Imai et al. 2006). In the present study, we found that Ci-ZicL was weakly expressed in A9.13, A9.14, A9.16, A9.30, and A9.32 nerve cord-lineage cells in normal embryos (Fig. 3f). In embryos carrying the Ci-FoxD/Ci-lefty transgene, expression of Ci-ZicL mRNA in the A-line cells was diminished while that in the a-line and b-line cells was not affected (Fig. 3f). Suppression was particularly evident in A9.15 and A9.29 cells where the amount of Ci-ZicL mRNA became equivalent to that in the other A-line nerve cord-lineage cells (Fig. 3f). SB431542 similarly affected Ci-ZicL expression in the A-line cells (Fig. 3f). In contrast, the Ci-FoxD/Ci-nodal transgene mediated enhanced expression of Ci-ZicL in all the A-line nerve cord-lineage cells (Fig. 3f). Ci-ZicL is also expressed in A7.3, A7.4, A7.7, and A7.8 blastomeres at the 64-cell stage (Yagi et al. 2004; Fig. 3g). The A-line nerve cord cells derive from A7.4 and A7.8 blastomeres. Neither promotion nor inhibition of the Nodal signaling affected the expression pattern of Ci-ZicL at the 64-cell stage (Fig. 3g). The results indicate that Ci-Nodal is required for later expression of Ci-ZicL at the gastrula stage but not for early expression at the 64-cell stage.
A dominant negative Ci-Cdx inhibits neural tube formation
Ci-Lefty antagonizes Nodal signaling
Lefty functions as a competitive inhibitor of Nodal in vertebrates (Juan and Hamada 2001). A similar function of a Lefty orthologue was suggested in the sea urchin (Duboc et al. 2004). In the C. intestinalis embryo, the Ci-FoxD/Ci-lefty transgene and SB431542 had almost the same effect on the spatial expression of Nodal target genes. These altered expression patterns are consistent with those obtained by injection of a Ci-nodal-specific MO (Hudson and Yasuo 2005; Imai et al. 2006). These results suggest that Ci-Lefty antagonizes Nodal signaling in the C. intestinalis embryo. This is also supported by the observation that the Ci-FoxD/Ci-lefty transgene and SB431542 caused a similar morphological abnormality. Pasini et al. (2006) observed a similar malformation by overexpression of Ci-lefty in the ectodermal region. The function of Lefty as an antagonist of Nodal seems conserved in deuterostomes.
Ci-Nodal activates Ci-ZicL and Ci-cdx in A-line lateral nerve cord lineage cells
The Drosophila caudal gene and its orthologues, designated cdx, in vertebrates play an important role in the establishment of the posterior structures in the embryo (Epstein et al. 1997; Isaacs et al. 1998; Macdonald and Struhl 1986; Marom et al. 1997; van den Akker et al. 2002). Expression of cdx in the posterior region of developing embryos is regulated by extracellular signals, such as retinoic acid, Wnt, and FGF (Lohnes 2003). There is no report to suggest that Nodal regulates the expression of cdx. In the C. intestinalis embryo, injection of a Ci-SoxC-specific MO induced ectopic expression of Ci-nodal and Ci-cdx in A-line nerve cord-lineage cells (Imai et al. 2006). Based on this observation, Ci-Nodal was speculated to be a candidate activator of Ci-cdx expression (Imai et al. 2006). In the present study, Ci-FoxD/Ci-nodal mediated ectopic activation of Ci-cdx, while Ci-FoxD/Ci-lefty and SB431542 repressed Ci-cdx. From these results, we conclude that Ci-Nodal is an activator of Ci-cdx expression.
Transcriptional activation of Ci-ZicL in A-line cells is intricately regulated (Anno et al. 2006). Ci-ZicL is first activated in A6.2 and A6.4 cells at the 32-cell stage (Yagi et al. 2004). The expression becomes weak at the early gastrula stage and is resumed at the middle/late gastrula stage (Yagi et al. 2004; Anno et al. 2006). The present study demonstrated that only this re-activation requires Nodal signaling. A Ci-ZicL-specific MO suppressed the expression of Ci-cdx at the late gastrula stage, suggesting that Ci-ZicL is an activator of Ci-cdx expression (Imai et al. 2006). This raised the possibility that Ci-Nodal activates Ci-ZicL, which in turn activates Ci-cdx. Overexpression of Ci-lefty, as well as treatment with SB431542, diminished Ci-ZicL expression in A9.15 and A9.29 blastomeres where Ci-cdx is normally expressed. This is consistent with our hypothesis, although this does not exclude the possibility of direct activation of Ci-cdx by Ci-Nodal.
Role for Ci-Nodal in neural tube formation
In the present study, both the inhibition and accentuation of Nodal signaling caused defects in neural tube formation. This suggests that the appropriate dose and/or spatial restriction of the Ci-Nodal influence are important for normal morphogenesis. We propose two possible mechanisms that mediate Ci-Nodal-dependent neural tube formation. (1) Ci-Nodal induces the lateral/dorsal fate and suppresses the ventral fate in the nerve cord (Hudson and Yasuo 2005). Consequently, both the inhibition and enhancement of Nodal signaling lead to interference in the pattern formation of the neural plate (Hudson and Yasuo 2005). This may affect the morphogenetic movement required for the neural groove to form. (2) Ci-Cdx may be responsible for the morphogenesis of the nerve cord. Inhibition of the function of Ci-Cdx caused defects in gastrulation, anterior–posterior elongation, and neural tube formation. This phenotype was similar to that observed in embryos caused by SB431542 treatment or overexpression of Ci-lefty. These two possibilities are not mutually exclusive, although the involvement of Ci-Cdx was demonstrated. Early Ci-Nodal target genes, such as Ci-Delta-like and Ci-sna, may activate Ci-cdx in late gastrulae. Functional analysis of these early target genes is important for understanding of the entire molecular network that regulates neural tube morphogenesis.
Ci-cdx is also expressed in the tail epidermis (Imai et al. 2004). However, in the present study, the Ci-cdx-EnR transgene was expressed under the control of the Ci-FoxD enhancer that is not activated in the epidermis. This suggests that the Ci-Cdx function is required in the nerve cord lineage. Ci-Nodal is also necessary for specification of the notochord and muscle (Hudson and Yasuo 2006). However, possible involvement of these mesodermal tissues in neural tube formation is now unclear. Katsuyama et al. (1999) proposed a hypothesis that Hrcad regulates cell motility and intercalation in the neuroectoderm and notochord, which is essential to body elongation. Convergence and intercalation in the neuroectoderm and notochord are important for closure of the blastopore and neural tube in vertebrates (Wallingford and Harland 2001). It is therefore likely that the neural tube’s formation depends, at least in part, on a regulatory cascade composed of Ci-Nodal, Ci-ZicL, and Ci-Cdx, in addition to genes involved in the medial–lateral patterning of the neural plate.
We thank Nori Satoh and Kazuko Hirayama at Kyoto University and Zenji Imoto at the Usa Marine Biological Institute of Kochi University for providing animals. We thank Hidetoshi Saiga at Tokyo Metropolitan University for kindly providing the plasmid containing EnR. We are also grateful to You Katsuyama for valuable advice. This work was supported by MEXT Japan. K. M. was supported by the Sasakawa Scientific Research Grant from The Japan Science Society.