Leibovich A et al. (2018), ADMP controls the size of Spemann's organizer t...

XB-ART-54527

BMC Biol 2018 Jan 22;161:13. doi: 10.1186/s12915-018-0483-x.

Show Gene links Show Anatomy links

ADMP controls the size of Spemann's organizer through a network of self-regulating expansion-restriction signals.

Leibovich A , Kot-Leibovich H , Ben-Zvi D , Fainsod A .

???displayArticle.abstract???
BACKGROUND: The bone morphogenetic protein (BMP) signaling gradient is central for dorsoventral patterning in amphibian embryos. This gradient is established through the interaction of several BMPs and BMP antagonists and modulators, some secreted by Spemann's organizer, a cluster of cells coordinating embryonic development. Anti-dorsalizing morphogenetic protein (ADMP), a BMP-like transforming growth factor beta ligand, negatively affects the formation of the organizer, although it is robustly expressed within the organizer itself. Previously, we proposed that this apparent discrepancy may be important for the ability of ADMP to scale the BMP gradient with embryo size, but how this is achieved is unclear. RESULTS: Here we report that ADMP acts in the establishment of the organizer via temporally and mechanistically distinct signals. At the onset of gastrulation, ADMP is required to establish normal organizer-specific gene expression domains, thus displaying a dorsal, organizer-promoting function. The organizer-restricting, BMP-like function of ADMP becomes apparent slightly later, from mid-gastrula. The organizer-promoting signal of ADMP is mediated by the activin A type I receptor, ACVR1 (also known as activin receptor-like kinase-2, ALK2). ALK2 is expressed in the organizer and is required for organizer establishment. The anti-organizer function of ADMP is mediated by ACVRL1 (ALK1), a putative ADMP receptor expressed in the lateral regions flanking the organizer that blocks expansion of the organizer. Truncated ALK1 prevents the organizer-restricting effects of ADMP overexpression, suggesting a ligand-receptor interaction. We also present a mathematical model of the regulatory network controlling the size of the organizer. CONCLUSIONS: We show that the opposed, organizer-promoting and organizer-restricting roles of ADMP are mediated by different receptors. A self-regulating network is proposed in which ADMP functions early through ALK2 to expand its own expression domain, the organizer, and later functions through ALK1 to restrict this domain. These effects are dependent on ADMP concentration, timing, and the spatial localization of the two receptors. This self-regulating temporal switch may control the size of the organizer and the genes expressed within in response to genetic and external stimuli during gastrulation.

???displayArticle.pubmedLink??? 29357852
???displayArticle.pmcLink??? PMC5778663
???displayArticle.link??? BMC Biol
???displayArticle.grants??? [+]

Species referenced: Xenopus laevis
Genes referenced: acvr1 acvrl1 admp bmp4 bmpr1a cer1 chrd fst gsc lhx1 myc nog sia1 szl ventx1 ventx2 ventx3 wnt8a
GO keywords: BMP signaling pathway [+]
???displayArticle.morpholinos??? acvr1 MO1 acvrl1 MO1 admp MO1

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

	Fig. 1. ADMP restricts the size of the organizer domain. ADMP overexpression was induced by injection of increasing amounts of capped RNA (32–800 pg/embryo). Embryos were collected during early gastrula and analyzed for changes in gene expression patterns by qPCR and in situ hybridization. a, b qPCR analysis of the expression levels of dorsal (a) and ventral (b) genes. c–j Embryos were injected in dorsal or ventral blastomeres with ADMP mRNA (100 pg/embryo) together with a lineage tracer (turquoise) followed by in situ hybridization analysis with chordin (c–f) and gsc (g–j) probes to determine the effect on their expression domain. The relative arc of the expression domain was determined. f, j Graphs summarizing all the measurements of the relative sizes (%) of the expression domains. Chordin analysis: control, n = 75; ADMP dorsal, n = 46; ADMP ventral, n = 33, and gsc analysis: control, n = 16; ADMP dorsal, n = 19; ADMP ventral, n = 21. Statistical test; Dunnett’s analysis of variance (ANOVA) multiple comparisons test. p < 0.05, p < 0.001, **p < 0.0001, ns not significant
	Fig. 2. ADMP provides dorsal signal required in Spemann's organizer at the onset of gastrulation. ADMP knockdown was induced by ADMPMO injection (7 ng/embryo) and subjected to analysis of changes in gene expression. a, b qPCR analysis of the effects on dorsal and ventral gene expression during early gastrulation (stage 10–10.25) (a) and later during early/mid-gastrula (stage 10.25–10.5) (b). c–n Effect of ADMPMO injection on the domains of chordin (c–f), gsc (g–j), and ADMP (k–n) expression. Embryos injected with ADMPMO (4 ng) either dorsally or ventrally were fixed at the onset of gastrulation. f, j, n The arc of the domain of expression was measured, and the size of the domain of expression relative to control embryos is shown (%). o–r The specificity of the ADMPMO was further tested in rescue experiments. Embryos were injected dorsally (turquoise) with ADMPMO (3.4 ng/embryo) alone or with increasing amounts of RNA encoding the zebrafish (ZF) ADMP protein (10 or 20 pg RNA/embryo). o–q In situ hybridization analysis of the changes in the chordin expression domain in control (o), ADMPMO injected (p), and embryos co-injected with ADMPMO and zebrafish ADMP mRNA (20 pg/embryo) (q). r Summary of the quantitation of the changes in the size of the chordin expression domain. Chordin analysis: control, n = 81; ADMPMO dorsal, n = 48; ADMPMO ventral, n = 33, gsc analysis: control, n = 19; ADMPMO dorsal, n = 20; ADMPMO ventral, n = 15, ADMP analysis: control, n = 20; ADMPMO dorsal, n = 18; ADMPMO ventral, n = 14; and chordin rescue analysis: control, n = 25; ADMPMO, n = 25; ADMPMO + ZF ADMP 10, n = 19; ADMPMO + ZF ADMP 20, n = 25. p values were calculated compared to controls (a, b, f, j, n) except in the rescue experiments that were compared to the ADMPMO sample (r). Statistical test; Dunnett’s (ANOVA) multiple comparisons test. p < 0.05, p < 0.01, p < 0.001, **p < 0.0001, ns not significant
	Fig. 3. ALK2 is required for the expansion of the organizer domain. a The efficacy of the ALK2MO was determined by western blot analysis. RNA encoding a myc-tagged version of the ALK2 protein with the ALK2MO recognition sequence was co-injected into embryos with increasing amounts of ALK2MO. Early gastrula protein extracts were subjected to electrophoresis and immunodetection with the 9E10 anti-myc monoclonal antibody. To control the loading and transfer steps, the blot was stripped and re-probed with an anti-actin antibody. b–m Embryos were injected with an antisense morpholino oligonucleotide specific for ALK2 (ALK2MO; 400 pg/embryo) or control MO (coMO). At the onset of gastrulation, embryos were analyzed by in situ hybridization with chordin (b–e), gsc (f–i), and ADMP (j–m) specific probes to determine the effect on their expression domain. The domain of expression was determined by measuring the arc in degrees and calculating the relative size compared to control embryos. The relative (%) arc of the expression domain was determined. Chordin analysis: control, n = 13; ALK2MO, n = 12; coMO, n = 13, gsc analysis: control, n = 11; ALK2MO, n = 12; coMO, n = 10, and ADMP analysis: control, n = 9; ALK2MO, n = 10; coMO, n = 8. p values were calculated compared to controls. Statistical test; Dunnett’s (ANOVA) multiple comparisons test. p < 0.05, *p < 0.01, ns not significant
	Fig. 4. The ALK2 receptor positively regulates the expression of organizer genes. The truncated, dominant negative form of ALK2 (tALK2) was employed to corroborate the function of this receptor in the expansion of the organizer. Embryos were injected with mRNA (1.3 ng/embryo) encoding tALK2 or the truncated ALK3 (tALK3) in order to block their activity. At the onset of gastrulation, the effect of blocking these receptors on the expression domains of chordin (a–d), gsc (e–h), and ADMP (i–l) was determined by in situ hybridization with specific probes. The relative size of the expression domain (%) was determined. Chordin analysis: control, n = 28; tALK2, n = 20; tALK3, n = 15, gsc analysis: control, n = 20; tALK2, n = 26; tALK3, n = 23, and ADMP analysis: control, n = 10; tALK2, n = 18; tALK3, n = 19. p values were calculated compared to controls. Statistical test; Dunnett’s (ANOVA) multiple comparisons test. p < 0.01, p < 0.001, ***p < 0.0001
	Fig. 5. ALK1 is required to restrict the size of the organizer domain. a–d To determine the function of the ALK1 receptor during early gastrulation, embryos were injected with increasing amounts of ALK1MO (3.4 and 1.7 ng/embryo). The effect of blocking this receptor on the chordin expression domain was determined by in situ hybridization. The relative (%) arc of the expression domain was determined. e–h The specificity of the ALK1MO (1.7 ng/embryo) was analyzed by co-injection with RNA encoding the mouse ALK1 receptor (2.2 ng/embryo). Changes in the domain of chordin expression were quantitated. i–p RNA encoding the truncated, dominant negative form of the mouse ALK1 receptor (tmALK1) was injected into Xenopus embryos either in the dorsal or ventrolateral regions of the embryo. The effect of the ALK1 activity was determined by in situ hybridization. Analysis of the changes in the chordin (i–l) and gsc (m–p) expression domains was performed at the onset of gastrulation. The relative (%) arc of the expression domain was determined in each case. ALK1MO amounts analysis: control, n = 18; ALK1MO 1.7, n = 18; ALK1MO 3.4, n = 15, ALK1MO rescue analysis: control, n = 83; ALK1MO, n = 62; ALK1MO + mALK1, n = 40, tmALK1 chordin analysis: control, n = 17; tmALK1 dorsal, n = 19; tmALK1 ventrolateral, n = 10, and tmALK1 gsc analysis: control, n = 34; tmALK1 dorsal, n = 15; tmALK1 ventrolateral, n = 20. p values were calculated compared to controls except in the rescue experiments that were compared to the ALK1MO sample (h). Statistical test; Dunnett’s (ANOVA) multiple comparisons test. p < 0.01, p < 0.001, ***p < 0.0001, ns not significant
	Fig. 6. Blocking the ALK1 receptor preferentially blocks the anti-organizer activity of ADMP. a–d, i Embryos were injected with RNA encoding the tALK1, tALK2, and tALK3 dominant negative receptors. During early gastrula stages the changes in the chordin expression domain were determined by in situ hybridization. e–h, j To determine the receptor mediating the organizer-repressive activity of ADMP, four-cell-stage embryos were injected in a single dorsal blastomere with ADMP mRNA (50 pg/embryo) alone or together with one of the dominant negative type I receptors, tALK1, tALK2, or tALK3 (320 pg/embryo). Also, fluorescein isothiocyanate (FITC)-dextran was included as a lineage tracer (turquoise). At the onset of gastrulation, the changes in the size of the chordin expression domain were studied. The embryos were analyzed for the expression domain changes induced by ADMP gain of function and the extent of rescue by the different dominant negative receptors. The relative (%) arc of the expression domain was determined. Comparative truncated receptor analysis: control, n = 24; tALK1, n = 24; tALK2, n = 15; tALK3, n = 21, and block ADMP ventral activity analysis: control, n = 18; ADMP, n = 21; +tALK1, n = 16; +tALK2, n = 21; tALK3, n = 14. p values were calculated compared to controls (i) except in the rescue experiments that were compared to the ADMP sample (j). Statistical test; Dunnett’s (ANOVA) multiple comparisons test (j) and two-tailed t test (i). p < 0.01, **p < 0.0001, ns not significant
	Fig. 7. Mathematical model for the interaction between ALK1, ALK2, and ADMP. a Model design: ADMP can diffuse, degrade, or bind ALK1 or ALK2 receptor. The occupied receptors then induce (ADMP/ALK2) or repress (ADMP/ALK1) the expression of ADMP and other organizer genes. ADMP is degraded after binding the receptor, and the receptor is recycled back to the surface. b Size of the organizer induction domain at the end of simulation, t = 1 h, as a function of ADMP flux relative to its size for the reference parameter set where ADMP flux is 1 μmnM s–1. The organizer size is robust to ADMP flux in the dual receptor model where ALK1 represses organizer induction (black circles), while size is sensitive and continues to increase in a model that does not include ADMP/ALK1-mediated organizer repression (red circles). c, d Heat map of the organizer induction domain (θ), where 1 denotes full induction, and 0 no induction in the two-receptor model (c) and in a classic morphogen model, with no repression by ADMP/ALK1 (d). X axis denotes position along the dorsoventral axis, dashed vertical gray line denotes the steady state organizer induction domain. Y axis denotes time. tsteady is the time when steady state organizer induction is achieved in the two-receptor model (c). The organizer induction domain does not reach steady state in the simulation time in the absence of ADMP/ALK1 repression (d)
	Fig. 8. ADMP establishes a self-regulatory network that controls the size of the organizer domain. Model for the ADMP-mediated self-regulatory control of organizer size. a During the onset of gastrulation ADMP and Alk2 are already robustly expressed within the organizer. This ligand-receptor pair promotes the enlargement of the organizer and expression of the organizer-specific genes. At this stage, the expression of chordin is low. b With the progression of gastrulation, the expression of ADMP continues to increase, thus allowing the diffusion or shuttling of ADMP to lateral regions flanking the organizer. c In the lateral marginal zone (LMZ), the ALK1 receptor is expressed and the ADMP/ALK1 ligand-receptor pair prevents further expansion of the organizer

References [+] :

Ben-Zvi, Scaling of the BMP activation gradient in Xenopus embryos. 2008, Pubmed, Xenbase

	Fig. 1. ADMP restricts the size of the organizer domain. ADMP overexpression was induced by injection of increasing amounts of capped RNA (32–800 pg/embryo). Embryos were collected during early gastrula and analyzed for changes in gene expression patterns by qPCR and in situ hybridization. a, b qPCR analysis of the expression levels of dorsal (a) and ventral (b) genes. c–j Embryos were injected in dorsal or ventral blastomeres with ADMP mRNA (100 pg/embryo) together with a lineage tracer (turquoise) followed by in situ hybridization analysis with chordin (c–f) and gsc (g–j) probes to determine the effect on their expression domain. The relative arc of the expression domain was determined. f, j Graphs summarizing all the measurements of the relative sizes (%) of the expression domains. Chordin analysis: control, n = 75; ADMP dorsal, n = 46; ADMP ventral, n = 33, and gsc analysis: control, n = 16; ADMP dorsal, n = 19; ADMP ventral, n = 21. Statistical test; Dunnett’s analysis of variance (ANOVA) multiple comparisons test. p < 0.05, p < 0.001, **p < 0.0001, ns not significant
	Fig. 2. ADMP provides dorsal signal required in Spemann's organizer at the onset of gastrulation. ADMP knockdown was induced by ADMPMO injection (7 ng/embryo) and subjected to analysis of changes in gene expression. a, b qPCR analysis of the effects on dorsal and ventral gene expression during early gastrulation (stage 10–10.25) (a) and later during early/mid-gastrula (stage 10.25–10.5) (b). c–n Effect of ADMPMO injection on the domains of chordin (c–f), gsc (g–j), and ADMP (k–n) expression. Embryos injected with ADMPMO (4 ng) either dorsally or ventrally were fixed at the onset of gastrulation. f, j, n The arc of the domain of expression was measured, and the size of the domain of expression relative to control embryos is shown (%). o–r The specificity of the ADMPMO was further tested in rescue experiments. Embryos were injected dorsally (turquoise) with ADMPMO (3.4 ng/embryo) alone or with increasing amounts of RNA encoding the zebrafish (ZF) ADMP protein (10 or 20 pg RNA/embryo). o–q In situ hybridization analysis of the changes in the chordin expression domain in control (o), ADMPMO injected (p), and embryos co-injected with ADMPMO and zebrafish ADMP mRNA (20 pg/embryo) (q). r Summary of the quantitation of the changes in the size of the chordin expression domain. Chordin analysis: control, n = 81; ADMPMO dorsal, n = 48; ADMPMO ventral, n = 33, gsc analysis: control, n = 19; ADMPMO dorsal, n = 20; ADMPMO ventral, n = 15, ADMP analysis: control, n = 20; ADMPMO dorsal, n = 18; ADMPMO ventral, n = 14; and chordin rescue analysis: control, n = 25; ADMPMO, n = 25; ADMPMO + ZF ADMP 10, n = 19; ADMPMO + ZF ADMP 20, n = 25. p values were calculated compared to controls (a, b, f, j, n) except in the rescue experiments that were compared to the ADMPMO sample (r). Statistical test; Dunnett’s (ANOVA) multiple comparisons test. p < 0.05, p < 0.01, p < 0.001, **p < 0.0001, ns not significant
	Fig. 3. ALK2 is required for the expansion of the organizer domain. a The efficacy of the ALK2MO was determined by western blot analysis. RNA encoding a myc-tagged version of the ALK2 protein with the ALK2MO recognition sequence was co-injected into embryos with increasing amounts of ALK2MO. Early gastrula protein extracts were subjected to electrophoresis and immunodetection with the 9E10 anti-myc monoclonal antibody. To control the loading and transfer steps, the blot was stripped and re-probed with an anti-actin antibody. b–m Embryos were injected with an antisense morpholino oligonucleotide specific for ALK2 (ALK2MO; 400 pg/embryo) or control MO (coMO). At the onset of gastrulation, embryos were analyzed by in situ hybridization with chordin (b–e), gsc (f–i), and ADMP (j–m) specific probes to determine the effect on their expression domain. The domain of expression was determined by measuring the arc in degrees and calculating the relative size compared to control embryos. The relative (%) arc of the expression domain was determined. Chordin analysis: control, n = 13; ALK2MO, n = 12; coMO, n = 13, gsc analysis: control, n = 11; ALK2MO, n = 12; coMO, n = 10, and ADMP analysis: control, n = 9; ALK2MO, n = 10; coMO, n = 8. p values were calculated compared to controls. Statistical test; Dunnett’s (ANOVA) multiple comparisons test. p < 0.05, *p < 0.01, ns not significant
	Fig. 4. The ALK2 receptor positively regulates the expression of organizer genes. The truncated, dominant negative form of ALK2 (tALK2) was employed to corroborate the function of this receptor in the expansion of the organizer. Embryos were injected with mRNA (1.3 ng/embryo) encoding tALK2 or the truncated ALK3 (tALK3) in order to block their activity. At the onset of gastrulation, the effect of blocking these receptors on the expression domains of chordin (a–d), gsc (e–h), and ADMP (i–l) was determined by in situ hybridization with specific probes. The relative size of the expression domain (%) was determined. Chordin analysis: control, n = 28; tALK2, n = 20; tALK3, n = 15, gsc analysis: control, n = 20; tALK2, n = 26; tALK3, n = 23, and ADMP analysis: control, n = 10; tALK2, n = 18; tALK3, n = 19. p values were calculated compared to controls. Statistical test; Dunnett’s (ANOVA) multiple comparisons test. p < 0.01, p < 0.001, ***p < 0.0001
	Fig. 5. ALK1 is required to restrict the size of the organizer domain. a–d To determine the function of the ALK1 receptor during early gastrulation, embryos were injected with increasing amounts of ALK1MO (3.4 and 1.7 ng/embryo). The effect of blocking this receptor on the chordin expression domain was determined by in situ hybridization. The relative (%) arc of the expression domain was determined. e–h The specificity of the ALK1MO (1.7 ng/embryo) was analyzed by co-injection with RNA encoding the mouse ALK1 receptor (2.2 ng/embryo). Changes in the domain of chordin expression were quantitated. i–p RNA encoding the truncated, dominant negative form of the mouse ALK1 receptor (tmALK1) was injected into Xenopus embryos either in the dorsal or ventrolateral regions of the embryo. The effect of the ALK1 activity was determined by in situ hybridization. Analysis of the changes in the chordin (i–l) and gsc (m–p) expression domains was performed at the onset of gastrulation. The relative (%) arc of the expression domain was determined in each case. ALK1MO amounts analysis: control, n = 18; ALK1MO 1.7, n = 18; ALK1MO 3.4, n = 15, ALK1MO rescue analysis: control, n = 83; ALK1MO, n = 62; ALK1MO + mALK1, n = 40, tmALK1 chordin analysis: control, n = 17; tmALK1 dorsal, n = 19; tmALK1 ventrolateral, n = 10, and tmALK1 gsc analysis: control, n = 34; tmALK1 dorsal, n = 15; tmALK1 ventrolateral, n = 20. p values were calculated compared to controls except in the rescue experiments that were compared to the ALK1MO sample (h). Statistical test; Dunnett’s (ANOVA) multiple comparisons test. p < 0.01, p < 0.001, ***p < 0.0001, ns not significant
	Fig. 6. Blocking the ALK1 receptor preferentially blocks the anti-organizer activity of ADMP. a–d, i Embryos were injected with RNA encoding the tALK1, tALK2, and tALK3 dominant negative receptors. During early gastrula stages the changes in the chordin expression domain were determined by in situ hybridization. e–h, j To determine the receptor mediating the organizer-repressive activity of ADMP, four-cell-stage embryos were injected in a single dorsal blastomere with ADMP mRNA (50 pg/embryo) alone or together with one of the dominant negative type I receptors, tALK1, tALK2, or tALK3 (320 pg/embryo). Also, fluorescein isothiocyanate (FITC)-dextran was included as a lineage tracer (turquoise). At the onset of gastrulation, the changes in the size of the chordin expression domain were studied. The embryos were analyzed for the expression domain changes induced by ADMP gain of function and the extent of rescue by the different dominant negative receptors. The relative (%) arc of the expression domain was determined. Comparative truncated receptor analysis: control, n = 24; tALK1, n = 24; tALK2, n = 15; tALK3, n = 21, and block ADMP ventral activity analysis: control, n = 18; ADMP, n = 21; +tALK1, n = 16; +tALK2, n = 21; tALK3, n = 14. p values were calculated compared to controls (i) except in the rescue experiments that were compared to the ADMP sample (j). Statistical test; Dunnett’s (ANOVA) multiple comparisons test (j) and two-tailed t test (i). p < 0.01, **p < 0.0001, ns not significant
	Fig. 7. Mathematical model for the interaction between ALK1, ALK2, and ADMP. a Model design: ADMP can diffuse, degrade, or bind ALK1 or ALK2 receptor. The occupied receptors then induce (ADMP/ALK2) or repress (ADMP/ALK1) the expression of ADMP and other organizer genes. ADMP is degraded after binding the receptor, and the receptor is recycled back to the surface. b Size of the organizer induction domain at the end of simulation, t = 1 h, as a function of ADMP flux relative to its size for the reference parameter set where ADMP flux is 1 μmnM s–1. The organizer size is robust to ADMP flux in the dual receptor model where ALK1 represses organizer induction (black circles), while size is sensitive and continues to increase in a model that does not include ADMP/ALK1-mediated organizer repression (red circles). c, d Heat map of the organizer induction domain (θ), where 1 denotes full induction, and 0 no induction in the two-receptor model (c) and in a classic morphogen model, with no repression by ADMP/ALK1 (d). X axis denotes position along the dorsoventral axis, dashed vertical gray line denotes the steady state organizer induction domain. Y axis denotes time. tsteady is the time when steady state organizer induction is achieved in the two-receptor model (c). The organizer induction domain does not reach steady state in the simulation time in the absence of ADMP/ALK1 repression (d)
	Fig. 8. ADMP establishes a self-regulatory network that controls the size of the organizer domain. Model for the ADMP-mediated self-regulatory control of organizer size. a During the onset of gastrulation ADMP and Alk2 are already robustly expressed within the organizer. This ligand-receptor pair promotes the enlargement of the organizer and expression of the organizer-specific genes. At this stage, the expression of chordin is low. b With the progression of gastrulation, the expression of ADMP continues to increase, thus allowing the diffusion or shuttling of ADMP to lateral regions flanking the organizer. c In the lateral marginal zone (LMZ), the ALK1 receptor is expressed and the ADMP/ALK1 ligand-receptor pair prevents further expansion of the organizer