Wawersik S et al. (2005), Conditional BMP inhibition in Xenopus reveals s...

XB-ART-2540

Dev Biol 2005 Jan 15;2772:425-42. doi: 10.1016/j.ydbio.2004.10.002.

Show Gene links Show Anatomy links

Conditional BMP inhibition in Xenopus reveals stage-specific roles for BMPs in neural and neural crest induction.

Wawersik S , Evola C , Whitman M .

???displayArticle.abstract???
Bone morphogenetic protein (BMP) inhibition has been proposed as the primary determinant of neural cell fate in the developing Xenopus ectoderm. The evidence supporting this hypothesis comes from experiments in explanted "animal cap" ectoderm and in intact embryos using BMP antagonists that are unregulated and active well before gastrulation. While informative, these experiments cannot answer questions regarding the timing of signals and the behavior of cells in the more complex environment of the embryo. To examine the effects of BMP antagonism at defined times in intact embryos, we have generated a novel, two-component system for conditional BMP inhibition. We find that while blocking BMP signals induces ectopic neural tissue both in animal caps and in vivo, in intact embryos, it can only do so prior to late blastula stage (stage 9), well before the onset of gastrulation. Later inhibition does not induce neural identity, but does induce ectopic neural crest, suggesting that BMP antagonists play temporally distinct roles in establishing neural and neural crest identity. By combining BMP inhibition with fibroblast growth factor (FGF) activation, the neural inductive response in whole embryos is greatly enhanced and is no longer limited to pre-gastrula ectoderm. Thus, BMP inhibition during gastrulation is insufficient for neural induction in intact embryos, arguing against a BMP gradient as the sole determinant of ectodermal cell fate in the frog.

???displayArticle.pubmedLink??? 15617685
???displayArticle.link??? Dev Biol

Species referenced: Xenopus
Genes referenced: actl6a acvr1 foxd3 hes4 mrc1 msx1 myc ncam1 nodal1 nog smad1 smad2 smad7 sox2 tbxt

???attribute.lit??? ???displayArticles.show???

	Fig. 1. δSmad73TEVGR is an inducible inhibitor of Smad1 phosphorylation. (A) Diagram of inducible inhibitor constructs. δSmad7 is a deletion of the MH1 domain of Xenopus Smad7, using the methionine at amino acid 172 as a start codon. A glucocorticoid receptor domain is fused to the C terminus of δSmad7 in the δSmad7GR construct. In the MTδSmad73TEVGR construct, three cleavage sites for the tobacco etch virus (TEV) protease are inserted between the C terminus of δSmad7 and the GR domain, and a myc tag is added to the N-terminus. The dexamethasone (DEX) inducible TEV protease (TEV2GR) consists of two GR domains flanking the TEV protease itself. (B) δSmad7 inhibits phosphorylation of Smad1, but not of Smad2. Lysates from uninjected St. 10.5 embryos or from embryos injected at the animal pole at the two-cell stage with mRNA encoding activated ALK2 (ALK2*; 100 pg/embryo), noggin (100 pg/embryo), or increasing amounts of δSmad7 were analyzed by Western blot with anti-phosphoSmad1 and anti-phosphoSmad2 antibodies. Cytoskeletal actin served as a loading control. (C) δSmad7 inhibits Smad1 phosphorylation but causes no reduction in Xnr1 stimulated Smad2 phosphorylation. Embryos were analyzed as in (B), after injection at the 2-cell stage with the indicated doses of RNA encoding Xnr1 or δSmad7 (D) TEV2GR cleaves the GR domain from δSmad73TEVGR in a regulated manner. N-terminal myc tags were added to δSmad7 and δSmad73TEVGR, and mRNA encoding these proteins were injected into two-cell stage embryos (δSmad7: 100 pg/embryo; δSmad73TEVGR: 300 pg/embryo, along with increasing amounts of TEV2GR. 10 μM DEX was added to embryos at the 8-cell stage. After lysis at St. 10.5, accumulation of the cleaved form of δSmad73TEVGR was analyzed by Western blotting with an anti-myc antibody, while concomitant inhibition of Smad1 phosphorylation was assessed using the anti-phosphoSmad1 antibody.
	Fig. 2. Phenotypic consequences of δSmad7TEVGR/TEV2GR expression in the absence and presence of DEX. Phenotypes of uninjected embryos (A) or embryos injected with 200 pg (B), 300 pg (C) or 400 pg (D) δSmad7TEVGR mRNA together with 10 pg TEV2GR mRNA. Note the head defects and partial duplication of axes at the highest concentration of δSmad7TEVGR (arrows, D). In the presence of 10 μM DEX, uninjected embryos are phenotypically normal (E), but injected embryos show defects reflecting dorsalization of the embryo (F–H).
	Fig. 3. Inhibition of pSmad1 occurs rapidly after DEX addition. (A) RNA was into both cells of two-cell stage embryos, animal cap explants were isolated at stage 8 and DEX was added when sibling embryos were at either stage 9 or stage 11. The caps were then allowed to develop overnight at 14°C, until sibling embryos had reached stage 16, when they were harvested and phosphoSmad1 levels were analyzed by Western blot. (B) mRNA encoding 300 pg δSmad73TEVGR and 10 pg TEV2GR was injected into both cells at the two-cell stage. Animal caps were isolated and assayed by Western blot with anti-phosphoSmad1 antibody. Cytoskeletal actin was used as a loading control. Culture of embryos and caps was carried out at 14°C. DEX was added either before caps were cut (St. 5) or 6, 3, 2, and 1 h before harvesting at stage 10. One hour after DEX addition, pSmad1 inhibition is substantially inhibited. By 3 h after addition, pSmad1 levels are equivalent to that observed in caps from embryos to which DEX was added at stage 5.
	Fig. 4. Stage-dependent neural induction in isolated embryonic ectoderm. Neural induction in embryos injected with 300 pg δSmad7TEVGR and 10 pg TEV2GR was examined by RT-PCR for N-CAM or Sox-2 (A) or in situ hybridization for Sox2 (B–F). (A) Addition of DEX induces strong neural expression in caps from embryos to which DEX was added at stage 4. Neural markers are also up-regulated upon addition of DEX at stage 9, though the neural response is decreased. Neural marker expression is absent in caps to which DEX is added at stage 11. (B) Little or no Sox2 expression is observed in β-gal-injected caps or in δSmad7TEVGR/TEV2GR-injected caps to which no DEX was added (C). Caps from δSmad7TEVGR/TEV2GR-injected embryos to which DEX was added at stage 4 show strong Sox2 upregulation (D). Up-regulation of Sox2 is diminished in δSmad7TEVGR/TEV2GR-injected caps to which Dex was added at stage 9 (E), and absent upon addition of DEX at stage 11 (F). (A) This effect was also observed by RT-PCR. Little induction of the neural markers Sox2 or NCAM is observed in the absence of DEX.
	Fig. 5. DEX-dependent inhibition of pSmad1 and of the BMP transcriptional target Msx1. mRNA encoding δSmad7-3TEVGR (300 pg) and TEV2GR (10 pg) were injected into 1 cell of 2-cell stage embryos. To mark the injected side, mRNA encoding either GFP (A–H) or β-galactosidase (I–L) was also included. 10 μM DEX was added at either St. 5, St. 10, or St. 14, with control embryos cultured in the absence of DEX. When DEX is added, inhibition of Smad1 phosphorylation on the injected side is observed by anti-phosphoSmad1 immunostaining (B–D) and by in situ hybridization for Msx1, a transcriptional target of BMP signaling (J–K). Little or no effect is observed on either of these markers in controls where no DEX is added (A, I).
	Fig. 6. Stage-dependent ectopic neural induction by δSmad7TEVGR-mediated BMP inhibition. Embryos were injected with 300 pg δSmad7TEVGR and 10 pg TEV2GR. After injection, embryos were cultured in the absence of DEX (A, F) or DEX was added at stage 6 (B, G), stage 8 (C, H), stage 9 (D, I), or stage 11 (E, J) and harvested at stage 16. Neural identity was assayed by in situ hybridization for Sox2 (A–E) or NCAM (F–J).
	Fig. 7. Effects of δSmad7-mediated stage-dependent BMP inhibition on neural crest and neural plate border. Embryos were injected with 300 pg δSmad7TEVGR RNA and 10 pg TEV2GR RNA and cultured in the absence of DEX (A, E, I, M), or with DEX added at stage 5 (B, F, J, N), stage 9 (C, G, K, O), or stage 11 (D, H, L, P). The effects of stage-specific BMP inhibition on the neural plate border were examined by in situ hybridization for the neural border markers N-tubulin (A–D), and Hairy2a (E–H). Neural crest development was examined by in situ hybridization for Hairy2a (red arrows, E–H), FoxD3 (I–L), and PDGFRα (M–P).(For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
	Fig. 8. FGF activation must be combined with BMP inhibition to induce ectopic neural expression during gastrulation. Embryos were injected with 20 pg iFGFR1 RNA (A–C), 300 pg δSmad7TEVGR and 10 pg TEV2GR RNA (D–F), or 20 pg iFGFR1 + 300 pg δSmad7TEVGR/10 pg TEV2GR RNA (G–I). Inducer (10 μM DEX and/or 1.25 μM AP20187) was added at stage 5 (B, E, H) or stage 9 (C, F, I). In the absence of inducer, ectopic neural induction is not observed in injected embryos (A, D, G). No ectopic neural induction is observed on the injected side (arrows) in iFGFR1-injected embryos when AP20187 is added at stage 5 (B) or stage 9 (C). Ectopic neural induction is observed in δSmad7TEVGR/TEV2GR-injected embryos upon addition of DEX at stage 5 (arrow, E), but induction is minimal when DEX is added at stage 9 (arrow, F). FGF activation and BMP inhibition appear to act synergistically, as strong ectopic neural induction is observed when iFGFR and δSmad7TEVGR/TEV2GR are combined, both when inducer is added at stage 5 (H) or at stage 9 (I).
	Fig. 9. Inhibition of FGF signaling reduces the size of the neural plate. Embryos were cultured in 0.1× MMR with either 0.005% DMSO (A) or 20 μM SU5402 + 0.005% DMSO added at stage 9 (B) or stage 11 (C). Neural plate formation was examined by in situ hybridization for Sox2. (A) Control embryos treated with DMSO exhibit normal neural plates (66/79 normal). (B) Addition of SU5402 at stage 9 causes spina bifida (red arrow) and reduction of the anterior neural plate (white arrow). Only 27/94 embryos examined exhibited normal-sized anterior neural plates. (C) 61/76 embryos to which SU5402 was added at stage 11 exhibit normal-sized neural plates. Spina bifida is absent in most embryos.(For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
	Fig. 10. FGF activation lengthens the period of neural competence in animal caps and induces expression of Xbra. Embryos were injected with 20 pg iFGFR1 and 300 pg δSmad7tevGR/10 pg TEV2GR RNA. Inducers (10 μM DEX and/or 1.25 μM AP20187) were added at stage 5 (C) or stage 11 (D). Animal caps were cut at stage 9 and collected at stage 14, when they were subjected to either to in situ hybridization for Sox2 or to RT-PCR. No Sox2 expression is detected in caps from uninjected embryos (A), and minimal Sox2 is present in caps from injected embryos to which no inducer was added (B). Strong up-regulation of Sox2 is observed in caps to which inducers were added at stage 5 (C). Addition of inducers at stage 11 also strongly activates Sox2 expression (D). This is in contrast to caps from embryos injected with δSmad7tevGR and TEV2GR and activated at stage 11 (compare with Fig. 4F). Note the elongation in caps induced at both stage 5 and stage 11 (arrows, C, D), suggesting mesoderm induction due to the activation of FGF. (F) RT-PCR for NCAM confirms the ability of iFGFR and δSmad7tevGR, but not δSmad7tevGR alone, to induce neural development when activated at stage 11. RT-PCR for Xbra further supports the induction of mesoderm in iFGFR- and δSmad7tevGR/TEV2GR-injected caps, but not in caps injected with δSmad7tevGR and TEV2GR alone.