|
Fig 1. The BMP type I receptor Alk1/2 is essential for D/V patterning.(A) Morphology of embryos at 72 hours after fertilization (hpf) injected with morpholinos targeting either the alk3/6, alk1/2, or bmp2/4 transcripts. Note the striking similarity of the phenotypes of alk1/2 and bmp2/4 morphants that both develop with a ciliary band on the dorsal side (black arrowheads) compared to the less severe phenotype of alk3/6 morphants that is evidenced by the presence of pigment cells (black arrows) and of a less well-developed ciliary band on the dorsal side. (B) Expansion of the ventral and ciliary band fates at the expense of the dorsal ectoderm in alk1/2 morphants was revealed by the analysis of marker genes. Controls and alk1/2 morphants embryos were stained by in situ hybridization with the indicated probes. In alk1/2 morphants at mesenchyme blastula, the territory expressing the ventral marker genes, nodal, chordin, and foxA is largely normal, but consistently, a slight broadening of nodal expression is observed (white arrowheads), while expression of the dorsal gene hox7 is suppressed. At the gastrula stage, however, this ventralization is patent with chordin and foxA expression extending towards the dorsal side in alk1/2 morphants (black arrowheads). Also note the dramatic dorsal expansion of the ciliary band genes foxG and onecut in the alk1/2 morphants. (C) Injection of high doses (2 mM) of alk1/2 morpholino caused a massive ectopic expression of nodal in about 50% of the embryos at the mesenchyme blastula stage. (D) Phospho-Smad1/5/8 immunostaining in control or alk1/2 morphants. p.Smad1/5/8 in the ectoderm and in the dorsal chain of primary mesenchyme cells (PMCs) (white arrowheads) of alk1/2 morphants is largely abolished. LV, lateral view; VV, vegetal pole view; AV, animal pole view; D, dorsal; V, ventral.
|
|
Fig 2. The double inactivation of alk1/2 and alk3/6 causes massive ectopic expression of nodal, resulting in extreme ventralization.(A) Morphology of embryos at 72 hpf injected with morpholinos targeting either the alk3/6, alk1/2, or bmp2/4 transcripts or injected with a mixture of the alk1/2 and alk3/6 morpholinos. Simultaneous down-regulation of Alk1/2 and Alk3/6 caused a strong ventralization similar to that resulting from overexpression of nodal or from treatment with nickel chloride (a treatment that ventralizes sea urchin embryos by causing massive ectopic expression of nodal). Note the presence of a ciliary band in the vegetal pole region (black arrowheads) and the prominent proboscis (white arrowheads) in the animal pole region in the double alk1/2 + alk3/6 morphants and in nodal-overexpressing or nickel-treated embryos. (B) In situ hybridization on controls and alk1/2 + alk3/6 morphants at the blastula and gastrula stages with ventral, ciliary band, and dorsal marker genes. The strong ventralization of alk1/2 + alk3/6 morphants is presaged by the massive ectopic expression of nodal at blastula stages. Note the radial expression of the ciliary band markers foxG and onecut in the vegetal pole region of alk1/2 + alk3/6 morphants at the gastrula stage. (C) Scheme describing the changes in fate maps caused by the single or double inactivation of type I BMP receptors. In the simple alk3/6 knockdown, the ventral ectoderm remains unaffected and the dorsal ectoderm is converted into ciliary band, while in the alk1/2 morphants, the ventral ectoderm is expanded, giving rise to a partial ventralization. In contrast, in the double alk1/2 + alk3/6 morphants, the whole ectoderm is converted into ventral ectoderm. LV, lateral view; VV, vegetal pole view; AV, animal pole view; FV, frontal view; V, ventral; D, dorsal.
|
|
Fig 3. Panda is the TGF-β ligand that cooperates with BMP2/4 to restrict nodal during D/V patterning.(A) Simple inactivation of alk1/2, alk3/6, bmp5/8, bmp2/4, panda, double inactivation of bmp5/8 +bmp2/4 or of bmp2/4 +admp, or triple inactivation of bmp2/4+ bmp5/8 + admp affects D/V polarity to various extents but does not cause full ventralization of the embryo. In contrast, double inactivation of panda and bmp2/4 causes an extreme ventralization of the embryo, mimicking the phenotype caused by nodal overexpression or by the double inactivation of alk1/2 and alk3/6. The ventralized phenotype of the double panda + bmp2/4 morphants is so strong that the proboscis in the animal pole frequently detaches from the rest of the embryo as a consequence of formation of a circular stomodeum. (B) In situ hybridization on controls and double panda + bmp2/4 morphants at the blastula and gastrula stages with ventral, ciliary band, and dorsal marker genes. Simultaneous inactivation of panda and bmp2/4 causes massive ectopic expression of nodal, suppresses dorsal marker gene expression, and restricts ciliary band markers to the vegetal pole, mimicking the effects of the double knockdown of Alk1/2 + Alk3/6. (C) Scheme describing the changes in fate maps caused by the single inactivation of bmp2/4, by the double inactivation of bmp2/4 and panda, or by the double inactivation of alk1/2 and alk3/6. LV, lateral view; VV, vegetal pole view; AV, animal pole view; V, ventral; D, dorsal.
|
|
Fig 4. Panda belongs to a subfamily of TGF-β that includes Drosophila Maverick and vertebrate GDF15.Phylogenetic analysis of TGF-β ligands using the maximum likelihood method. The analysis was performed using the full-length proteins. Representative taxa from deuterostomes, protostomes, and cnidarians were used (see S1 Text for a list of these taxa). These 162 sequences were collected from diverse databases using the National Center for Biotechnology Information (NCBI) research tool (http://www.ncbi.nlm.nih.gov/). Full-length sequences were aligned using ClustalOmega with default parameters (http://www.ebi.ac.uk/clustalw/), and gap optimisation and obvious alignment error correction were made using Bioedit 7.0.5.3 (http://www.mbio.ncsu.edu/BioEdit/bioedit.html). The tree was calculated using the maximum likelihood method with PhyML with substitution model WAG (http://atgc.lirmm.fr/phyml/). A consensus tree with a 45% cutoff value was derived from 500 bootstrap analysis using Mega 3.1 (http://www.megasoftware.net/). Numbers above nodes represent a percentage of bootstrap values supporting this node. The original tree is presented in S5 Fig.
|
|
Fig 5. The TGF-β ligand Panda is expressed maternally in a D/V gradient.(A) RT-PCR analysis of Panda mRNA. An aliquot of the PCR reaction was run on a 1% agarose gel, and the gel was stained with Syber safe. Top panel, panda expression. Bottom panel, expression of mkk3 used as a control. Panda is expressed in ovaries and unfertilized eggs as well as during the cleavage, blastula, gastrula, and pluteus stages. (B) Expression of panda mRNA analyzed by in situ hybridization. Whole mount in situ hybridizations with panda alone or with panda (red) and nodal (blue) probes. A gradient of maternal panda mRNA is detected in immature ovocytes and to a lesser extent in the unfertilized mature eggs, whereas during the cleavage and blastula stages, panda mRNA is detected in a shallow D/V gradient. VEB, very early blastula (about 120 cells); EB, early blastula (about 220 cells); PHB, prehatching blastula (about 300 cells); HB, hatching blastula (about 400 cells); MB, mesenchyme blastula; LV, lateral view; VV, vegetal view; V, ventral; D, dorsal; FV, frontal view.
|
|
Fig 6. Panda plays a pivotal role during D/V axis formation.(A) Morphological phenotypes resulting from injection into the egg of antisense morpholino oligonucleotide targeting the translation start site of the panda transcript. Down-regulation of panda completely radializes the embryos during the first 48 h, but a partial recovery of D/V polarity occurs afterwards, as evidenced by the formation of a short dorsal apex. This phenotype is similar to that caused by treatments with recombinant Nodal protein at 1 μg/ml. (B) In situ hybridization on control embryos and panda morphants at the blastula and gastrula stages with ventral, ciliary band, and dorsal marker genes. Note the dramatic ectopic expression of nodal, chordin, and foxA at the blastula stage in panda morphants. At the late gastrula stage, despite their radialized morphology, panda morphants are patterned along the D/V axis, as evidenced by the restricted expression of nodal, foxA, foxG, onecut, and hox7. Note, however, the extended expression of these ventral markers compared to control embryos. (C) A second morpholino oligonucleotide targeting the 5' UTR of the panda transcript produces similar phenotypes and radializes the expression of nodal and chordin. (D) Scheme describing the changes in fate maps caused by the single inactivation of panda or bmp2/4 or by the double inactivation of panda and bmp2/4. LV, lateral view; VV, vegetal pole view; AV, animal pole view; V, ventral; D, dorsal.
|
|
Fig 7. Maternal but not zygotic Panda function is required for the spatial restriction of nodal expression.(A) Injection of a morpholino oligonucleotide targeting the translation start site of panda mRNA, but not of a morpholino targeting a splice junction, disrupts the establishment of D/V polarity. The lineage tracer Fluoresceinated Lysine-Fixable Dextran (FLDX) was coinjected with the morpholino. (B) In situ hybridizations against the nodal transcript at early stages. In the absence of maternal, but not of zygotic, Panda, a massive ectopic expression of nodal is observed starting at the 60-cell stage. Note that nodal expression remains radially expressed up to the mesenchyme blastula stage. Massive and early ectopic expression of nodal is also observed in the double alk1/2+alk3/6 morphants. VEB, very early blastula (about 120 cells); EB, early blastula (about 220 cells); PHB, prehatching blastula (about 300 cells); HB, hatching blastula (about 400 cells); SB, swimming blastula; LB, late blastula; MB, mesenchyme blastula. (C) Scheme summarizing the perturbations of nodal expression caused by blocking Panda or Alk1/2+Alk3/6. VV, vegetal pole view; V, ventral; D, dorsal.
|
|
Fig 8. Panda activity is required locally to orient the D/V axis.(A) Global overexpression of panda at the one-cell stage does not perturb establishment of the D/V axis. Normal morphology of pluteus larvae developing after injection with panda mRNA at 1,000 μg/ml. (B) Effects of local overexpression or down-regulation of various components of the Nodal and BMP pathway on the orientation of the D/V axis. Injection of panda (1,000 μg/ml) or of the activated form of alk3/6 (alk3/6Q230D) mRNA (200 μg/ml) into one blastomere at the two-cell stage imposes a dorsal identity to the progeny of the injected cell in nearly 100% of the injected embryos. Local down-regulation of nodal or acvrII also imposes a dorsal identity. Conversely, down-regulation of panda or alk3/6 mRNA, and to a lesser extent of alk1/2, strongly biases the orientation of the D/V axis and promotes ventral fates. V, ventral; D, dorsal. The same result was also obtained by local inhibition of Nodal signaling after injection into one blastomere at the two-cell stage of a morpholino oligonucleotide targeting either the nodal transcript or the type II Nodal receptor acvrII. Consistent with this strong effect on the orientation of the D/V axis, in all embryos injected with panda or alk3/6QD mRNAs or with the acvrII or nodal morpholinos, at the blastula stage, nodal was expressed in a sector of the embryo located on the opposite side of the clone of injected cells (Fig 8B). (C) Expression of tbx2/3 at the early blastula or mesenchyme blastula stages in embryos injected with panda mRNA. Overexpression of panda precociously and ubiquitously activates tbx2/3 at the early blastula stage. Note, however, that expression of tbx2/3 becomes polarized along the D/V axis at the mesenchyme blastula stage.
|
|
Fig 9. Panda activity has to be provided locally to efficiently rescue panda morphants.Differential interference contrast (DIC) and fluorescence images of embryos injected with a panda morpholino into the egg and then with panda mRNA either into the egg or into one blastomere at the two-cell stage. While providing Panda activity into the egg does not rescue D/V polarity, injection of panda mRNA or of low doses (50 μg/ml) of mRNA encoding the activated form of Alk3/6 (Alk3/6QD) into one blastomere at the two-cell stage fully rescues D/V polarity of panda morphants. LV, lateral view; V, ventral; D, dorsal.
|
|
Fig 10. Panda, like BMP2/4, requires Alk1/2 and Alk3/6 to pattern the D/V axis, but Panda, unlike BMP2/4, does not appear to activate phosphorylation of Smad1/5/8.(A) Coinjection of the panda and alk1/2 morpholinos or of the panda and alk3/6 morpholinos causes a strong ventralization, as does the double inactivation of bmp2/4 + alk1/2 or of bmp2/4 and alk3/6. (B) While the simple knockdown of bmp2/4 is not sufficient to cause ectopic expression of nodal, the double knockdown of bmp2/4 + alk1/2 or bmp2/4 + alk3/6 causes massive ectopic expression of nodal. (C) Panda requires Alk3/6 to orient the D/V axis when misexpressed. Local overexpression of panda orients the D/V axis in nearly 100% of the injected embryos. However, if the eggs are first injected with the alk3/6 morpholino, panda is no longer able to orient the D/V axis. (D) Western blot of phospho-Smad1/5/8 in control embryos at the indicated stages or in embryos overexpressing panda or injected with a panda morpholino. Note that phosphoSmad1/5/8 is undetectable before the late blastula stage and that overexpression of panda does not appear to cause phosphorylation of Smad1/5/8. B1, very early blastula; B2, early blastula. (E) Phospho-Smad1/5/8 immunostaining at very early and early blastula stages or mesenchyme blastula stages in control embryos and in embryos overexpressing panda, bmp2/4, alk3/6QD mRNA, or alk1/2QD. The highly sensitive alkaline phosphatase-based detection of pSmad1/5/8 does not allow detection of Smad1/5/8 signaling at early stages in control embryos. In contrast, following overexpression of bmp2/4, alk3/6QD, or alk1/2QD into the egg or into one blastomere at the two-cell stage, strong nuclear phosphoSmad1/5/8 immunostaining is easily detected at early blastula. This pSmad1/5/8 immunoreactivity is not detected following injection of panda mRNA. Nevertheless, overexpression of panda induces weak ectopic Smad1/5/8 signaling at mesenchyme blastula. (F) Overexpression of panda expands Smad1/5/8 signaling at the mesenchyme blastula stage. pSmad1/5/8 is normally restricted to the dorsal ectoderm and dorsal PMCs at mesenchyme blastula. Overexpression of panda expands the territory in which pSmad1/5/8 is detected toward the ventral side. LV, lateral view; VV, vegetal view; SVV, surface and vegetal view; V, ventral; D, dorsal.
|
|
Fig 11. Model of D/V axis formation: The sequential activities of Panda, Lefty, and BMP2/4 establish the D/V axis.(A) Expression and functional requirements of the main ligands and receptors involved in D/V patterning. (B,C) The sequential activities of Panda, Lefty, and BMP2/4 progressively define the D/V axis. (1) During cleavage, a gradient of maternal Panda activity, acting through Alk3/6 and Alk1/2, first antagonizes nodal expression and breaks the radial symmetry of the embryo. (2) At the early blastula stage (7 h 30 hpf), following this initial symmetry breaking, spatially restricted Nodal induces Lefty, and Lefty, which diffuses more than Nodal, prevents Nodal autoregulation outside the presumptive ventral ectoderm. This second phase of Nodal antagonism is required to maintain the spatial restriction of nodal expression. (3) Starting at the hatching blastula stage (9 hpf), Nodal induces BMP2/4, and BMP2/4 signaling synergizes with Lefty to antagonize Nodal on the dorsal side. V, ventral; D, dorsal.
|
