Azbazdar Y and De Robertis EM (2024), Molecular analysis of a self-organizing signali...

XB-ART-60778

Proc Natl Acad Sci U S A 2024 Jul 09;12128:e2408346121. doi: 10.1073/pnas.2408346121.

Show Gene links Show Anatomy links

Molecular analysis of a self-organizing signaling pathway for Xenopus axial patterning from egg to tailbud.

Azbazdar Y , De Robertis EM .

???displayArticle.abstract???
Xenopus embryos provide a favorable material to dissect the sequential steps that lead to dorsal-ventral (D-V) and anterior-posterior (A-P) cell differentiation. Here, we analyze the signaling pathways involved in this process using loss-of-function and gain-of-function approaches. The initial step was provided by Hwa, a transmembrane protein that robustly activates early β-catenin signaling when microinjected into the ventral side of the embryo leading to complete twinned axes. The following step was the activation of Xenopus Nodal-related growth factors, which could rescue the depletion of β-catenin and were themselves blocked by the extracellular Nodal antagonists Cerberus-Short and Lefty. During gastrulation, the Spemann-Mangold organizer secretes a cocktail of growth factor antagonists, of which the BMP antagonists Chordin and Noggin could rescue simultaneously D-V and A-P tissues in β-catenin-depleted embryos. Surprisingly, this rescue occurred in the absence of any β-catenin transcriptional activity as measured by β-catenin activated Luciferase reporters. The Wnt antagonist Dickkopf (Dkk1) strongly synergized with the early Hwa signal by inhibiting late Wnt signals. Depletion of Sizzled (Szl), an antagonist of the Tolloid chordinase, was epistatic over the Hwa and Dkk1 synergy. BMP4 mRNA injection blocked Hwa-induced ectopic axes, and Dkk1 inhibited BMP signaling late, but not early, during gastrulation. Several unexpected findings were made, e.g., well-patterned complete embryonic axes are induced by Chordin or Nodal in β-catenin knockdown embryos, dorsalization by Lithium chloride (LiCl) is mediated by Nodals, Dkk1 exerts its anteriorizing and dorsalizing effects by regulating late BMP signaling, and the Dkk1 phenotype requires Szl.

???displayArticle.pubmedLink??? 38968117
???displayArticle.pmcLink??? PMC11252917
???displayArticle.link??? Proc Natl Acad Sci U S A
???displayArticle.grants??? [+]

Species referenced: Xenopus tropicalis Xenopus laevis
Genes referenced: admp axin1 bmp4 cer1 chrd dkk1 dvl2 frzb hwa lefty1 ncam1 nodal nodal1 nodal6 nog otx2 rax sia1 sox2 szl ventx1.2 wnt1 wnt8a
GO keywords: embryo development [+]
???displayArticle.morpholinos??? ctnnb1 MO1 szl MO1

Phenotypes: Xla wt + bmp4 (Fig. S4 C) [+]

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

	Fig. 1. Hwa mRNA did not rescue -catenin knockdown and -catenin depletion inhibited Wnt signals through tadpole stages in Xenopus. (A and B) A single ventral injection of Hwa mRNA (10 pg) induced complete secondary axes. (C) -catenin depletion with -cat MO injected four times at 2- to 4-cell caused complete ventralization. (D) Hwa mRNA 1x ventral failed to rescue the body axis in radially -cat MO-injected embryos. (EH) Group views of the same experiment one day later. (IK) Luciferase assays in -catenin-depleted embryos showing that xWnt8 DNA was not sufficient to induce BAR transcriptional activity at stage 11, 13, and 20 in -catenin-depleted embryos. Five embryos were used for each group using biological triplicates. (L and M) Embryos injected with -cat MO twice ventral at late four-cell stage inhibited late Wnt signaling with expanded head and cement gland structures. The number of embryos were as follows: (A) n = 53, all normal; (B) n = 38, all with twinned axes; (C) n = 22, all ventralized with a complete absence of dorsal axes; (D) n = 18, all ventralized; (L) n = 16, all normal; (M) n = 20, all with Dkk1-like phenotype with expanded cement gland and enlarged bellies; (N) Diagram for Hwa and -catenin pathway. Error bars indicate SD (*P < 0.001, P < 0.01, and *P < 0.05). (Scale bars: 500 m for individual embryo photos and 2 mm for group photos.)
	Fig. 2. Microinjection of Xnr mRNA into a ventral blastomere rescued a complete dorsal axis in β-catenin-depleted embryos, and the Xenopus Nodal-related antagonists CerS or Lefty mRNA inhibited Hwa mRNA ectopic secondary axes in wt embryos. (A) Uninjected control embryo. (B) Xnr1 mRNA microinjection induced partial incomplete secondary axes. (C) β-cat MO radially injected embryos with complete ventralization. (D–F) Xnr1 (25 pg 1 × ventral) induced complete primary axes in β-catenin-depleted embryos. (G and G’) Control embryos injected with LacZ. (H–I’) High doses of CerS and Lefty caused smaller heads but the primary axis remained. (J and J’) Hwa mRNA injected embryos with twinned axes. (K–L’) CerS and Lefty inhibited secondary axis formation. Results from three independent experiments. (M) Diagram of the Hwa/β-catenin/Nodal/Chordin pathway. The number of embryos were as follows: (A) n = 48, all normal; (B) n = 36, 28 embryos with incomplete axis; (C) n = 18, all completely ventralized; (D–F) n = 28, 26 embryos with complete axes with large cement gland and 2 ventralized embryos; (G) n = 60, all normal; (H) n = 20, 14 embryos with no cement gland and small heads and six embryos with smaller heads with cement glands; (I) n = 44, all with head ventralization; (J) n = 26, all with complete second axes; (K) n = 108, 87 with no secondary axes, 17 embryos normal, four embryos with small ectopic heads; (L) n = 55, 43 embryos with no axes and 12 embryos with small ectopic bumps. (Scale bars: 500 µm.)
	Fig. 3. The trunk-tail organizer was rescued by a single injection of the Nodal antagonists CerS or Lefty in LiCl dorsalized embryos consisting of head structures. (A) Control embryos with 1× ventral lacZ mRNA staining in Inset. (B) Embryos treated with LiCl at the 32-cell stage have radial heads lacking trunk-tails. (C) CerS single ventral injection caused ventralization with small heads. (D) CerS restored trunk and tail structures in LiCl-treated embryos. (E) Ventralized embryo injected with Lefty mRNA. (F) Lefty mRNA single ventral injection at the four-cell stage rescued the LiCl effect increasing trunk-tail structures. Results from three independent experiments. (G) Diagram for the Hwa/β-catenin/Nodal/Chordin pathway. The number of embryos were as follows: (A) n = 163, all normal; (B) n = 151, all with strong dorsalization; (C) n = 137, all ventralized; (D) n = 123, 71 embryos with trunks and small heads, 52 embryos were partially dorsalized but less than LiCl; (E) n = 68, all with small heads; (F) n = 67, all with trunk-tails and small heads. (Scale bar, 500 µm.)
	Fig. 4. Chordin mRNA rescued both D-V and A-P axial structures in a dose-dependent manner. (A) Control embryo. (B) Embryos with 4 x injection of β-cat MO (34 pg total) were completely ventralized. (C–F) Chordin mRNA injection rescued A-P and D-V axial structures in β-catenin-depleted embryos in a dose-dependent way; note that phenotypes in both axes are intertwined. Results from three independent experiments. (G) Diagram of the Hwa/β-catenin/Nodal/Chordin/BMP4 pathway. The number of embryos were as follows: (A) n = 176, all normal; (B) n = 131, all ventralized; (C) n = 18, 1 embryo with weak axis and others ventralized; (D) n = 38, all with complete axes including eyes and cement glands; (E) n = 21, with expanded cement gland; (F) n = 118, all with very large CNS, eyes, and cement glands. In addition, experiments were performed for β-cat MO 4 x + 2 pg of Chordin, n = 16, 7 embryos with incomplete axes with no cement gland, 4 with cement gland and 5 with no axis; and β-cat MO 4 x + 50 pg of Chordin 1 x ventral, n = 14, all with strong axial rescue. (Scale bars: 500 µm.)
	Fig. 5. The Wnt antagonist Dkk1 synergized with Hwa mRNA in dorso-anterior CNS development, and Szl depletion blocked this effect. (A–D’ ) Hwa mRNA in combination with Dkk1 mRNA had striking anteriorized CNS structures with almost radial cement gland and lacking trunk-tail structures. (E and E’ ) Uninjected control embryo. (F and F’ ) 4 x Szl MO injection caused strong ventralization with small heads and expanded ventro-posterior tissues. (G and G’ ) Embryo coinjected with Hwa and Dkk1 with radial cement bland and anterior CNS. (H and H’ ) Szl depletion strongly antagonized the synergy between Hwa and Dkk1. Results from three independent experiments. Numbers of embryos analyzed were as follows: (A and A’) n = 59, all normal; (B and B’) n = 46, 45 with twinned axes; (C and C’) n = 46, 45 with strong dorsalization; (D and D’) n = 77, all strongly dorsalized; (E and E’) n = 131, all normal; (F and F ‘) n = 43, all ventralized with small heads and expanded ventral tissues; (G and G’) n = 65, 33 embryos with radial cement glands and no trunk, 32 embryos with shortened double axes with cement glands covering over 50% of the radius of the embryo; (H and H’) n = 41, 38 embryos were ventralized with the Szl MO phenotype of small heads and cement glands, two embryos with residual large cement glands, one embryo with two weak axes. (Scale bars: 500 µm.)
	Fig. 6. Proposed molecular pathway for self-organizing axial development in Xenopus embryos. Black activator or inhibitory arrows indicate interactions at the protein level and blue arrows show that transcriptional activation steps. Hwa is an egg maternal mRNA that stabilizes the β-catenin transcriptional activator by increasing Axin1 degradation (14). Dorsal β-catenin promotes high Nodal signaling levels from the Nieuwkoop center in endoderm at early blastula, as well as the BCNE (Blastula Chordin and Noggin Expression center) in dorsal ectoderm (43). The BCNE expresses Siamois, a homeobox protein that plays an important role in the induction of Spemann organizer genes but is not indicated in this diagram (44). Expression of Xenopus Nodal-related genes is required for the induction of secondary axes by Hwa. At the gastrula stage, the Spemann–Mangold organizer center secretes a cocktail of proteins such as the BMP antagonists Noggin and Chordin, the Wnt antagonists Dkk1 and Frzb-1, ADMP, and the multivalent inhibitors Lefty (antagonist of Nodals and activing) and Cerberus (antagonist of Nodal, BMP, and Wnt). At the opposite pole of the embryo, the ventral center has high BMP4/7 signals that oppose the influence of the organizer. xWnt8 is induced by BMP and, in a positive feedback loop, increases BMP signaling through GSK3 inhibition. Sizzled is an inhibitor of the Tolloid proteinase that degrades Chordin, which is transcribed at high levels of BMP, forming a negative feedback loop that increases Chordin levels when BMP is high. Dkk1 and Frzb-1 inhibit the activity of xWnt8 at late gastrula, reducing BMP signaling.
	Fig. S1. Hwa mRNA had a high penetrance of secondary axis induction, and Hwa mRNA signals earlier than xWnt8 mRNA. (A-D) Hwa mRNA (10 pg) injected 1 x ventrally induced secondary axes in 100% of the cases. Number of embryos were as follows: A and B n=20, all normal; C and D n=21, all with complete twinned axesincluding eyes. (E-H) At blastula and early gastrula stages, Hwa mRNA was more active than Wnt8 mRNA in BAR β-catenin transcriptional activity assays, but not in late gastrula stage. Five embryos were used for each group as biological triplicates. Error bars indicates standard deviation (*p < 0.001, p < 0.01 and *p < 0.05). (I-L) DNA injections showing Hwa did not have an effect when expressed after midblastula transition, while xWnt8 and Dkk1 had late Wnt/β-catenin signaling activity. Number of embryos were as follows: I n=66 for 1 x ventral and n=16 for 4 x injections, all normal; J n=46 for 1 x ventral (inset) and n=40 for 4 x injection, all phenotypically normal; K n=56, 43 ventralized, rest wt for 1 x ventral and n=35, all ventralized with small heads; L n=62, 51 embryos dorsalized with enlarged cement gland for 1 x ventral injections and n=26, 1 normal embryo and rest with enlarged cement gland and bellies for 4 x Dkk1 DNA injection. Scale bars: 500 µm.
	Fig. S2. Xnr6 mRNA induced the Spemann organizer gene Chordin. (A-C) Xnr6 was able to induce Chordin expression in β-catenin depleted embryos in in situ hybridizations, and (D) by qRT-PCR. (E) Xnr6 was not able to activate the Wnt target gene Siamois in 4 x β-cat MO injected embryos in qRT-PCR assays. (F and G) The ventral genes Szl and Vent1 were inhibited by a single injection of Xnr6 in β-catenin depleted embryos. Five embryos were used for each group in biological triplicates. Error bars indicates standard deviation (*p < 0.001 and p < 0.01). Number of embryos were as follows: A n=53, all normal; B n=50, all ventralized; C n=35, with some Chordin expression. Scale bar, 500 µm.
	Fig. S3. Molecular analysis of axial induction by Chordin, and rescue of β-catenin knockdown by Noggin. (A-C) In situ hybridizations of the pan-neural marker Sox2 at tailbud stage showing that 1 x ventral Chordin mRNA injection was able to rescue the CNS in β-catenin depleted embryos. (D-F) In situ hybridizations at gastrula stage showing induction of the anterior marker Otx2 by Chordin. (G-I) qRT-PCR showing Chordin was able to increase transcripts of the anterior markers Otx2 and Rx2a and the neural marker NCAM at early neurula stage 13. Five embryos were used for each group as biological triplicates. (*p < 0.001 and p < 0.01). Error bars indicate standard deviation (J-L) The BMP antagonist Noggin was also able to induce D-V and A-P axial structures in β-catenin depleted embryos. Numbers of embryos analyzed were as follows: J n=96, all normal; K n=40, all ventralized; L n=63, 61 embryos with large heads and strong dorsalization. Scale bars: 500 µm. (M) Diagram for Hwa/β-catenin/Nodal/Chordin/BMP4 pathway.
	Fig. S4. Hwa secondary axes were antagonized by coinjection of BMP4 mRNA. (A) Uninjected embryo. (B) Hwa mRNA injected embryo with secondary axis. (C) mBMP4 injected partially ventralized embryo. (D) mBMP4 mRNA coinjection blocked the Hwa axis. (E-H) In situ hybridizations of Sox2 for control, Hwa, xBMP4 and coinjection of Hwa and xBMP4. (I) Diagram for Hwa/β-catenin/Nodal/Chordin/BMP4 pathway. Number of embryos analyzed were as follows: A n=50, all normal. B n=60, 48 embryos with second axis and 12 embryos strong dorsalization; C n=41, 14 embryos with ventralization with smaller heads, 3 embryos with no cement gland, 24 embryos normal; D n=73, 60 embryos with no secondary axis, 4 embryos with incomplete axis, 9 embryos with axes with cement gland. Results from three independent experiments. Scale bars: 500 µm.
	Fig. S5. Hwa mRNA dorsalized cell fate, expanded the late organizer, and cooperated with Dkk1 or Frzb-1. (A-D) Dkk1 mRNA injected in a single ventral cell remained in the trunk ventral region but coinjection of Hwa and Dkk1 redirected the fate of cells to the superficial layers of the expanded head structures. (E and F) In situ hybridizations of Chordin at the mid gastrula stage 11 showing embryos injected with Hwa mRNA 1 x ventral had two organizer centers. (G) Dkk1 mRNA injected embryo with strong Chordin transcription in the Spemann-Mangold organizer center but no ectopic expression. (H) The Chordin transcriptional domains were expanded in embryos coinjected with Hwa and Dkk1. (I-L) Frzb-1 coinjection with Hwa mRNA increased dorsalization but not as strongly as Dkk1. Numbers of embryos analyzed were as follows: I n=115, all normal; J n=63, 56 embryos with second axis and 7 embryos with dorsalization; K n=101, all with enlarged cement gland; L n=113, all stronger than Frzb-1 mRNA alone but without reaching radial structures. Scale bars: 500 µm.
	Fig. S6. Sizzled depletion was epistatic over the Dkk1 mRNA phenotype, and the Wnt inhibitor Dkk1 inhibited BMP signaling at early neurula but not early gastrula. (A and B) 4 x Szl MO injection caused strong ventralization of embryos (a typical high-BMP phenotype). (C and D) Szl depletion eliminated the dorsalizing effect of Dkk1 overexpression. (E and F) Dkk1 was unable to affect BMP reporter gene expression or overexpressed xBMP4 signals at early gastrula stage but was able to inhibit both at early neurula stage. Three embryos were used for each group as biological triplicates. Numbers of embryos analyzed were as follows: A n=57, all normal; B n=40, 38 embryos strongly ventralized with small heads and cement glands absent or very small; C n=102, 97 embryos with typical Dkk1 phenotype with expanded cement gland, head, and anterior belly; D n=41, all ventralized; Error bars indicate standard deviation (*p < 0.001 and p < 0.01). Scale bars: 500 µm.