Rel/NFκB negatively regulates dorsal organizer formation.

a Schematic diagrams of NFκB-tkP:dGFP reporter. NFκB BS: consensus sequence of the NFκB-binding element. PolyA: SV40 polyadenylation sequence. b, c NFκB reporter is activated in the dorsal region of early zebrafish embryos. b Whole-mount in situ hybridization (WISH) for dGFP at the indicated stage in NFκB-tkP:dGFP-transgenic embryos. The dGFP-expressing dorsal regions are indicated with red arrowheads. Scale bar = 200 μm. c Double fluorescent in situ hybridization (FISH) for dGFP (green) and chordin (dorsal marker, magenta) in NFκB-tkP:dGFP-transgenic embryos. Animal views with dorsal to the right. Scale bar = 200 μm. d–f Inhibition of NFκB activity by iκbαb leads to expansion of organizer size and dorsal tissue. d WISH for dGFP in NFκB-tkP:dGFP-transgenic embryos, (e) dharma (organizer marker) and chordin in embryos injected with control (mKO2) or iκbαb mRNA at indicated stages. Animal views. Scale bar = 200 μm. Box plots of the angle of marker genes show first and third quartile, median is represented by a line, whiskers indicate the minimum and maximum. Each dot represents one embryo. P-values from unpaired two-tailed t-tests are indicated. f Phenotypes of larvae injected with control (mKO2) or iκbαb mRNA at 27 hpf. The strength of dorsalization was scored using the C1–5 classification scheme. Lateral views with anterior to the left. The loss of ventral tail fin (a typical dorsalization phenotype) is indicated with red arrowheads. Scale bar = 200 μm. g–irel over-expression leads to reduction of organizer size and dorsal tissue. g WISH for dGFP in NFκB-tkP:dGFP-transgenic embryos, (h) dharma and chordin in embryos injected with control (mKO2) or rel mRNA at indicated stages. Scale bar =  200 μm. Box plots of the angle of marker genes show first and third quartile, median is represented by a line, whiskers indicate the minimum and maximum. Each dot represents one embryo. P-values from unpaired two-tailed t-tests are indicated. i Phenotypes of 27 hpf larvae injected with control (mKO2) or rel mRNA. The strength of ventralization was scored. Lateral views with anterior to the left. Scale bar = 200 μm. Source data are provided as a Source Data file.

Specific rel MO reveals that Rel restricts dorsal organizer formation.

a Rel is a main NFκB acting in early zebrafish embryos. WISH for dGFP in NFκB-tkP:dGFP-transgenic embryos injected with control MO (ctrl MO) or rel MO at dome stage, animal view. Scale bar = 200 μm. b–drel MO leads to expansion of organizer size and dorsal tissue in wild-type (WT) embryos whereas rela compensates the loss of rel in MZrel embryos. WISH for (b) dharma, chordin, and (c) vent (ventral marker) in WT and maternal-zygotic rel mutants (MZrel) embryos injected with ctrl MO, rel MO or rela MO at 30% epiboly. Animal views with dorsal to the right. Scale bar = 200 μm. Box plots of the angle of marker genes show first and third quartile, median is represented by a line, whiskers indicate the minimum and maximum. Each dot represents one embryo. P-values from two-tailed one-way ANOVAs with Sidak correction are indicated. ns: not significant (p > 0.05). d Representative pictures of 27 hpf WT and MZrel larvae injected with ctrl MO, rel MO or rela MO. The strength of dorsalization was scored. The loss of ventral tail fin is indicated with red arrowheads. Lateral views with anterior to the left. Scale bar = 200 μm. erela is upregulated in MZrel but not rel morphants. qPCR analysis for expression of rela in MZrel or rel morphants at sphere stage. Normalized values are shown as means ± SEM. n = 3, biologically independent samples. P-values from unpaired two-tailed t-tests are indicated. f Model of the mechanism of rela-mediated genetic compensation and the specific inhibition by rel MO. Source data are provided as a Source Data file.

Rel negatively regulates Wnt/β-catenin signaling.

Wnt/β-catenin signaling is enhanced by rel knockdown (a, b) and decreased by rel overexpression (c, d) at the dome stage. WISH (a, c) and qPCR analysis (b, d) of axin2 and sp5l in dome-stage WT embryos injected with ctrl MO, rel MO, control (mKO2) or rel mRNA. In a and c, animal views are shown. Scale bar = 200 μm. In b and d, normalized values are shown as mean ± SEM. n = 3 biologically independent samples. P-values from unpaired two-tailed t-tests are indicated. Source data are provided as a Source Data file.

Rel directly activates a secreted Wnt antagonist Frzb to restrict Wnt/β-catenin-mediated dorsal organizer formation.

a, b Rel positively regulates frzb expression. a qPCR analysis for expression of Wnt antagonists (frzb, sfrp1a, dkk1b and notum1a) in dome stage embryos injected with ctrl MO or rel MO. Normalized values are shown as means ± SEM. n = 3, biologically independent samples. P-values from unpaired two-tailed t-tests are indicated. b WISH for frzb in dome stage embryos injected with ctrl MO or rel MO, animal view. c, d Rel inhibits Wnt/β-catenin signaling through Frzb. WISH for (c) Wnt targets axin2 and sp5l; (d) dharma and chordin in embryos injected with ctrl MO, rel MO or rel MO with frzb mRNA at indicated stage, animal view. Box plots of the angle of marker genes show first and third quartile, median is represented by a line, whiskers indicate the minimum and maximum. Each dot represents one embryo. P-values from two-tailed one-way ANOVAs with Sidak correction are indicated. e, f Rel activates promoter of frzb. e Top panel shows the schematic diagram of the upstream region of zebrafish frzb gene. The NFκB-binding element possessing the strongest potential to bind to Rel homodimer is marked with red “pin” while others are marked by black “pins”. Gray and blue boxes indicate Exons and UTRs, respectively. Bottom panel shows the schematic diagrams of the reporter constructs, frzb:luc and frzb:luc MT. f FISH for luciferase (magenta) and gfp (green) in dome-stage embryos injected with pDha-GFP and frzb:luc or frzb:luc MT, with MOs as indicated, animal view. Scale bar = 200 μm. Source data are provided as a Source Data file.

Tlr4 activates NFκB to stimulate frzb-mediated restriction of dorsal organizer formation.

a–c Inhibition of Tlr4 reduces frzb expression and enhances Wnt/β-catenin signaling and dorsal organizer formation. Embryos were injected with mKO2 (control) or Tlr4 DN mRNA. WISH for (a) dGFP in NFκB-tkP:dGFP-transgenic; frzb and axin2 in WT, (b) dharma and chordin in WT embryos at the indicated stage. Animal views. Box plots of the angle of marker genes show first and third quartile, median is represented by a line, whiskers indicate the minimum and maximum. Each dot represents one embryo. P-values from unpaired two-tailed t-tests are indicated. c Representative pictures of 27 hpf larvae, lateral views with anterior to the left. The strength of dorsalization was scored. d–f Forced activation of Tlr4 by injection of lipopolysaccharide (LPS) activates NFκB signaling and inhibits dorsal organizer formation. d Phenotypes of 27 hpf larvae injected with LPS, uninjected as control. The strength of ventralization was scored. Lateral views with anterior to the left. e Embryos were injected with LPS and treated with DMSO or TAK-242 or co-injected with Tlr4 DN, uninjected as control. DMSO and TAK-242 were treated from 3 hpf to dome stage. WISH for dGFP in NFκB-tkP:dGFP-transgenic; frzb in WT embryos at dome stage. f WISH for axin2, dharma and chordin in embryos injected with LPS, uninjected as control. Animal views. Scale bar = 200 μm. Source data are provided as a Source Data file.

β-catenin stimulates Wnt5b-mediated Tlr4/NFκB activation.

a, b β-catenin activates NFκB via Tlr4. a Embryos were injected with control (mKO2) or constitutively active β-catenin mutant (β-cat CA) mRNA, with or without Tlr4 DN mRNA. b Embryos were injected with control or ctnnb2 (zebrafish β-catenin) MO. WISH for dGFP in NFκB-tkP:dGFP-transgenic; frzb in WT embryos at the dome stage, animal view. Scale bar = 200 μm. c NFκB is activated following Wnt/β-catenin activation in the developing dorsal organizer. WISH for dGFP in OTM:d2EGFP and NFκB-tkP:dGFP-transgenic embryos and frzb in WT embryos at the indicated stage. The dGFP-expressing and frzb-expressing dorsal regions are indicated by red arrowheads. Scale bar = 200 μm. d β-catenin activates wnt5b expression in early embryos. Embryos were injected with control (mKO2) or constitutively active β-catenin mutant (β-catCA) mRNA. WISH for wnt5a and wnt5b at the sphere stage, animal view. Scale bar = 200 μm. e, f Wnt/β-catenin signaling activates NFκB via Wnt5b. e Embryos were injected with ctrl or wnt5b MO. f Embryos were injected with control (mKO2) or constitutively active β-catenin mutant (β-cat CA) mRNA, with or without wnt5b MO. WISH for dGFP in NFκB-tkP:dGFP-transgenic embryos at the dome stage, animal view. Scale bar = 200 μm. g, h Wnt5b activates frzb expression through Tlr4/Rel. g Embryos were injected with control (mKO2) or wnt5b mRNA, and co-injected with Tlr4 DN mRNA or rel MO. WISH for dGFP in NFκB-tkP:dGFP-transgenic, frzb in WT embryos at the dome stage, animal view. Scale bar = 200 μm. h WISH for dharma in WT embryos injected with control (mKO2) or wnt5b mRNA at 30% epiboly stage. Scale bar = 200 μm. Box plots of the angle of marker gene show first and third quartile, median is represented by a line, whiskers indicate the minimum and maximum. Each dot represents one embryo. P-values for unpaired two-tailed t-tests are indicated. Source data are provided as a Source Data file.

Negative feedback loop between canonical/non-canonical Wnts and Tlr4/NFκB determines the precise size of zebrafish dorsal organizer.

Model of the role of Tlr/NFκB signaling in the initiation of embryonic DV axis formation. In Drosophila, Wnt-mediated negative feedback regulation of Toll/NFκB signaling determines the embryonic DV pattern. In zebrafish, negative feedback regulation between canonical Wnt (caWnt), non-canonical Wnt (Wnt5b), and Tlr4/NFκB generates the precise size of the dorsal organizer and the consequent DV pattern.