Phenotypes of different mutants at various stages. (A) Generation of eomesa^tsu007 mutant allele. Left: genomic structures and putative coding products of eomesa WT allele and tsu007 allele. The exons were colored and positions of nucleotides and amino acids were indicated. The mutant allele carries a 353-bp deletion. Right: morphology of eomesa^+/tsu007 (heterozygote) and eomesa^{tsu007/tsu007} (zygotic mutant) adults. Note that the posterior dorsal fin (indicated by an hollow arrowhead) is absent in mutant adult. (B) Scheme of generation of Mhwa;Meomesa double mutants. Zygotic genotypes were indicated. Zhwa^−/−;Zeomesa^−/− homozygous female adults could not ovulate naturally, and eggs squeezed from these fish were used for in vitro fertilization (IVF) using sperms squeezed from wildtype males (Zhwa^+/+;Zeomesa^+/+). (C) Wildtype (WT) embryos at indicated stages. (D,D’) Meomesa mutants. Mutant embryos exhibited variable phenotypes at 8 hpf and 24 hpf, and the ratios of 24-hpf mutants categorized into different classes (I-III), which were derived from several homozygous females, were shown in (D’). (E) Mhwa mutants at indicated stages. (F,F’) Meomesa;Mhwa double mutants. The ratios of embryos with different phenotypes at 24 hpf were shown in (F’). Embryos were laterally positioned when the dorsal or tail was recognizable. The scale bar in (C) was also applied to (D–F).

Expression patterns of mesendodermal markers and nodal genes in WT and mutant embryos. The expression of the endodermal marker sox32, the mesodermal marker tbxta and the epidermal marker gata2a as well as ndr1 and ndr2 was examined by WISH (A,C) or qRT-PCR (B,D) at indicated stages. Embryos in (A,C) were shown in animal-pole view with dorsal to the right if the dorsal was recognizable. The ratio of embryos with the representative pattern was indicated at the right bottom. Note that the majority of Meomesa;Mhwa embryos completely lacked tbxta expression while the other had some tbxta expression. Scale bars: 100 μm. For RT-PCR analysis, 15 embryos were pooled for each assay, and the expression level was normalized to that of eif4g2a in WT embryos at the same stage. Error bars indicated S.D. based on three biological replicates (indicated by small circles). Color keys for embryo types were shown in (A,C). Statistically significant levels: ns, nonsignificant; **, p < 0.01; ***, p < 0.001.

Induction of nodal genes and mesendodermal markers in Meomesa;Mhwa mutants by ectopic hwa or/and eomesa. One-cell stage mutant embryos were injected with corresponding mRNA or/and MO and observed for morphology at 6 hpf (A) or harvested at indicated stages for detection of selected genes by WISH (B,D) or by RT-PCR analysis (C,E). Embryos were positioned laterally (A) or in animal-pole view (B,D) with dorsal to the right if the dorsal side was perceptible. The ratio of embryos with the representative pattern was indicated in the right bottom. Scale bars: 100 μm. Injection doses of mRNA or MO: hwa and myc-eomesa, 50 pg/embryo; cMO (as control MO) and β-cat2-MO, 20 ng/embryo. qRT-PCR analysis was performed using 15 embryos per sample, and the expression level was normalized to that of eif4g2a in WT embryos at the same stage. Error bars indicated S.D. based on three biological replicates (indicated by small circles). Color keys for embryo types and treatments were shown in (A,B,D). Statistically significant level: ns, nonsignificant; *, p < 0.05; **, p < 0.01; ***, p < 0.001.

Overexpression effect of eomesa and hwa in WT and different types of mutant embryos on morphological changes at 24 hpf. (A–D) Morphology of WT or mutants at 24 hpf under different mRNAs injection. One-cell stage WT or mutant embryos were injected with corresponding mRNA and observed for morphology at 24 hpf. Embryos were positioned laterally with the head to the left. The ratio of embryos with the representative pattern was indicated at bottom. Injection doses: hwa and myc-eomesa, 50 pg/embryo. Scale bars: 100 μm.

Responses of Meomesa and Mhwa mutants to Nodal signaling inhibition. One-cell stage embryos (10 min postfertilization) were incubated in Holfreter’s water with 1% DMSO (control) or 50 μM SB431542 (SB, the Nodal signaling inhibitor), and harvested at 24 hpf for morphological observation (A) or for detection of marker gene expression by WISH (B,C) or qRT-PCR analysis (D) at indicated stages. Note that inhibition of Nodal signaling aggravated mesendodermal defects in both Meomesa and Mhwa mutants (A). Embryos were positioned laterally (A) or in animal-pole view with dorsal to the right (B,C) if the dorsal or tail was perceptible. The embryonic shield in WT embryo at the shield stage was indicated by an arrowhead. Scale bars, 100 μm. The ratio of embryos with the representative pattern was indicated in the right bottom (B,C). qRT-PCR analysis was performed using 15 embryos per sample, and the expression level was normalized to that of eif4g2a in WT embryos at the same stage. Error bars indicated S.D. based on three biological replicates (indicated by small circles). Statistically significant level: ***, p < 0.001.

Effect of Nodal signaling inhibition on ndr1 expression in WT and mutant embryos. WT, Meomesa or Mhwa embryos at the one-cell stage were incubated in Holfreter’s water with 1% DMSO (control) or 50 μM SB431542 (SB) and harvested at indicated stages for detection of ndr1 expression by WISH (A,C,E) or qRT-PCR analysis (B,D,F). Embryos (A,C,E) were positioned in animal-pole view with dorsal to the right if the dorsal side was distinguishable. The ratio of embryos with the representative pattern was indicated in the right bottom. Scale bars, 100 μm. qRT-PCR analysis was performed using 15 embryos per sample. The expression level at 3.7, 4, 4.3, 5 and 6 hpf in WT or Mhwa embryos was normalized to that at 4.7 hpf in WT embryo, while the expression level at different stages in WT or Meomesa embryos was normalized to that at 6 hpf in WT embryos. Error bars indicated S.D. based on three biological replicates (indicated by small circles). Statistically significant levels: ns, nonsignificant; *, p < 0.05; **, p < 0.01; ***, p < 0.001.

Effect of Nodal signaling inhibition on ndr2 expression in WT and mutant embryos. Embryo treatment and data presentation were similar to those described in Figure 6 legend.

Illustration of contributions of maternal eomesa, hwa/β-catenin signaling and Nodal signaling to ndr1 and ndr2 expression. The expression pattern of ndr1 and ndr2 in the late blastula (in lateral view with dorsal to the right and animal pole to the top) is depicted. The overall expression level of ndr1 or ndr2 is the sum of contributions from maternal eomesa (as seen in Mhwa without Nodal signaling), maternal hwa/β-catenin signaling (as seen in Meomesa without Nodal signaling) in and Nodal signaling (autoregulation). In wildtype (WT) embryos, all of the three forces make a significant contribution to ndr1 expression; however, maternal eomesa makes a predominant contribution to ndr2 expression while maternal hwa/β-catenin signaling may contribute a little to ndr2 expression by activating ndr1 in the dorsal blastodermal margin and thereof Nodal signaling. In the last column, ndr1 and ndr2 levels contributed by eomesa and hwa were shown as empty to highlight the contribution of Nodal signaling.