Fig. 5. Nodal-independent signaling is sensitized but is sufficient for mesoderm induction.

a Model of Nodal signaling gradient formation in WT and MZsmad4a mutant embryos. In contrast to WT (upper panel), in MZsmad4a embryos (lower panel) a homomeric trimer of activated pSmad2 can still be recruited to chromatin via Foxh1, but with reduced efficiency. Consequently, some Nodal target genes, in particular, lft1/2 are initially inefficiently transcribed. As Nodal induction of Ndr1/2 is not inhibited by Lft1/2, this leads to higher premature induction of the Nodal–pSmad2 signaling gradient. By 60% epiboly, once Lft1/2 levels have reached those equivalent to WT, the signaling gradient in MZsmad4a mutant embryos has adjusted to WT-like levels. b Dose responses of SB-505124 for lft1/2 and ndr2 in WT and MZsmad4a embryos, treated with inhibitor at sphere stage. The dose marked 0 corresponds to DMSO. Black dots represent the mean for three biological replicates for each dose and each genotype. Orange shading indicates the SEM. lft1 0.4 µM: p(adj) = 0.012, lft1 0.8 µM: p(adj) = 0.036, ndr2 0.4 µM: p(adj) = 0.005, ndr2 0.8 µM: p(adj) = 0.003, ndr2 2 µM: p(adj) = 0.003. Unpaired multiple comparison t-test with Holm-Sidack correction. *p < 0.05; **p < 0.01. c Upper panel: MIP of ISH for tbxta (cyan) in CTRL and MZsmad4a embryos at 90% epiboly (epi). Bottom panel: MIP of ISH for tbx16 (cyan) at 90% epiboly. Nuclei are stained with DAPI (white). Scale bars correspond to 150 µm. Dorsal views are shown. d MIP of ISH for sox32 (green) in CTRL and MZsmad4a embryos at 75% epiboly. Nuclei are stained with DAPI (white). Scale bars correspond to 150 µm. Dorsal views are shown. Images in c, d are representative of 25 embryos each and three independent experiments, except for the tbx16 ISH, which is the result of two independent experiments.