Analysis of dvl2 and dvl3a mutant phenotypes.

(A-C) WT embryos, the insets show the eyes in ventral view at 30 hpf and in dorsal view at 5 dpf. (D-F) Zdvl2 mutants, obtained by crosses between heterozygous dvl2^+/- carriers, display weak axis extension defect at 11.5 hpf. They are normal at 30 hpf, and show reduced swim bladder (sb) at 5 hpf. (G-I) MZdvl2 mutants, obtained from crosses between female dvl2^-/- fish and male dvl2^+/- fish, display a reduced AP axis at 11.5 hpf and 30 hpf. They present fused eyes at 30 hpf (inset, ventral view; arrowhead indicates fused lenses), and develop craniofacial defects and cyclopia (inset, dorsal view), with pharyngeal cartilages protruding outward (arrow) at 5 dpf. (J-L) MZdvl3a mutants from crosses between dvl3a^-/- carriers display weak axis extension defect, but are indistinguishable from WT embryos at 30 hpf and 5 dpf. (M) Statistical analysis of the extent of axis extension delay. The embryos were imaged at 11.5 hpf followed by genotyping. Those embryos with expected genotypes were used to measure the angle between the anterior end and posterior end, with vertex at the geometric center of the embryo (inset). Bars represent the mean ± s.d. from indicated numbers of embryos (****, P<0.0001). (N) Alcian blue staining of head cartilages at 5 dpf. Cartilage structures of the basicranium are outlined in grey, showing the fusion of trabeculae and the absence of ethmoid plate (ep) in MZdvl2 mutants. (O) Quantitative analysis of MZdvl2 mutant phenotypes at 5 dpf in offspring from three independent female dvl2^-/- and male dvl2^+/- fish pairs. Control embryos were obtained from crosses between female WT fish and male dvl2^+/- fish. Numbers on the top of each column indicate total embryos analyzed. (P) Genotyping of dvl2 mutants with normal and defective phenotypes. All embryos with a normal phenotype are dvl2^+/- mutants, whereas all defective embryos are MZdvl2 mutants. Numbers on the top of each column indicate total embryos genotyped from three independent fish pairs. Scale bars: (A, D, G, J) 400 μm; (B, E, H, K) 400 μm; (C, F, I, L) 400 μm; (N) 100 μm.

Zygotic Dvl2 and Dvl3a in axis extension.

Mutant phenotypes were compared with WT embryos at indicated stages. Insets show eye phenotypes in ventral view, with bidirectional arrows indicating the distance of the two eyes. (A-C) WT embryos. (D-F) Double heterozygous dvl2^+/-;dvl3a^+/- mutants, from crosses between female dvl2^-/- fish and male dvl3a^-/- fish, show weakly delayed axis extension at 11.5 hpf, but are phenotypically normal at later stages. (G-I) Triallelic dvl2^+/-;Zdvl3a mutants,from crosses between dvl2^+/-;dvl3a^+/- carriers, display weak axis extension defect at 11.5 hpf and 30 hpf, and recover at 5 dpf. (J-L) Triallelic Zdvl2;dvl3a^+/- mutants, from crosses between dvl2^+/-;dvl3a^+/- carriers, exhibit more obvious axis extension defect at 11.5 hpf and 30 hpf, and develop shortened AP axis, compressed head, and reduced swim bladder at 5 dpf, which are referred as type I phenotype. (M-O) Zdvl2;Zdvl3a mutants, from crosses between female dvl2^+/-;dvl3a^+/- fish and male dvl2^+/-;Zdvl3a fish, show more strong axis extension defect at 11.5 hpf and 30 hpf, and present a severe type II phenotype, with shortened and wavy axis, craniofacial defects, and complete disappearance of swim bladder (sb) at 5 dpf. (P) Statistical analysis of the extent of axis extension delay, after genotyping of imaged embryos at 11.5 hpf. Capital letters of the abscissa correspond to the images at 11.5 hpf. Bars represent the mean ± s.d. from indicated numbers of embryos, and asterisks above the bars refer to comparison with WT embryos (***, P<0.001; ****, P<0.0001; NS, not significant). (Q) Quantitative analysis of type I and type II phenotypes at 5 dpf from three independent female dvl2^+/-;dvl3a^+/- and male dvl2^+/-;dvl3a^-/- fish pairs. Control embryos were from crosses between female WT fish and male dvl2^+/-;dvl3a^-/- fish. (R) Genotypes of type I and type II embryos. Numbers on the top of each column indicate total embryos analyzed. Scale bars: (A, D, G, J, M) 400 μm; (B, E, H, K, N) 400 μm; (C, F, I, L, O) 400 μm.

Dvl2 and Dvl3a dosages in axis extension and AP patterning.

Both dvl2^+/-;MZdvl3a and Zdvl2;MZdvl3a mutants were from crosses between dvl2^+/-;dvl3a^-/- carriers. Representative mutant embryos were imaged at indicated stages. Eye phenotypes are shown in the insets as ventral view. (A-C) WT embryos. (D-F) dvl2^+/-;MZdvl3a mutants display moderate axis extension defect at 11.5 hpf and 30 hpf, and recover to a normal phenotype at 5 dpf. (G-I) Zdvl2;MZdvl3a mutants show strong axis extension defect at different stages, and display caudal truncation, craniofacial defects, cardiac edema (arrow), and fused eyes or cyclopia (inset) at 30 hpf and at 5 dpf. (J) Quantitative analysis of the posterior truncation phenotype at 5 dpf in offspring derived from three independent dvl2^+/-;dvl3a^-/- carriers. Control embryos were from crosses between female WT fish and male dvl2^+/-;dvl3a^-/- fish. Posterior deficiency is present in the offspring from all three fish pairs with a proportion that follows the Mandel inheritance (about 25%). Numbers on the top of each column indicate total embryos analyzed. (K) Genotyping of normal and posteriorly truncated embryos. All defective embryos are dvl2^-/-;dvl3a^-/- mutants. Scale bars: (A, D, G) 400 μm; (B, E, H) 400 μm; (C, F, I) 400 μm.

Analysis of MZdvl2;MZdvl3a mutants.

(A) Schema illustrating the strategy to generate female mosaic mdvl2^+(-)/-;dvl3a^-/- adult fish with a new mutant allele (red) in dvl2. MZdvl2;MZdvl3a mutant embryos are present at varied proportions in the offspring from crosses between female mdvl2^+(-)/-;dvl3a^-/- and male dvl2^+/-;dvl3a^-/- fish, depending on the efficiency of germline mutations in the remaining dvl2 WT allele. (B) An example of the sequencing chromatogram with both the original mutant allele and a new indel in dvl2 locus. (C-F) Lateral (C, D) and dorsal (E, F) views of a representative WT embryo (C, E), and an MZdvl2;MZdvl3a mutant (D, F) at 11.5 hpf. Notice that the mutant embryo displays most severely impaired AP axis and convergence of paraxial mesoderm, with strongly widened somites (arrowheads). (G) A WT embryo at 30 hpf. (H) Lateral view of a representative MZdvl2;MZdvl3a mutant at 30 hpf shows deficiency of trunk and posterior regions, and cyclopia (arrowhead; see also S12 Fig). (I) The phenotype of a Zdvl2;MZdvl3a mutant with characteristic caudal truncation at 30 hpf. (J) Genotyping of dvl2 alleles in MZdvl2;MZdvl3a mutant embryos from 7 independent fish pairs. A novel indel (red) along with the original mutation (blue) are present in the mutants. The left and right TALEN targeting sites are indicated in green. Dots are introduced to optimize sequence alignment. (K) Quantitative analyses of the occurrence of MZdvl2;MZdvl3a and Zdvl2;MZdvl3a mutants among offspring from 7 independent fish pairs. Each fish pair was crossed three times, and numbers on the top of each column indicate total embryos scored. Scale bar: (C-F) 400 μm; (G-I) 400 μm.

The expression of organizer genes is not affected in MZdvl2;MZdvl3a mutants.

In situ hybridization analysis of the expression patterns of goosecoid, chordin, and tbxta at dome stage. Animal pole viewed embryos with dorsal region on the right. (A) Schematic representation for the analysis of MZdvl2;MZdvl3a embryos. (B-F) Representative images show similar expression patterns of goosecoid in WT embryos and in different mutants. (G) Knockdown of ß-catenin2 strongly inhibits goosecoid expression. (H-L) Similar expression patterns of chordin in WT embryos and all indicated mutants. (M) Knockdown of ß-catenin2 blocks chordin expression. (N-R) The expression patterns of tbxta are similar between WT embryos and different mutants. (S) Knockdown of ß-catenin2 has no effect on tbxta expression. Scale bar: (B-J) 400 μm.

Knockdown of other dvl genes has no effect in MZdvl2;MZdvl3a mutants.

The embryos were injected with coMO (2.5 ng) or a mixture of equal amount of dvl1a/1b/3bMOs (6 ng in total) at 1-cell stage, except for immunostaining. All mutant embryos were imaged after different analyses, followed by genotyping. (A-D) In situ hybridization analysis of goosecoid (gsc) expression in indicated embryos at dome stage. Animal pole view with dorsal region on the right. (E) Endogenous ß-catenin nuclear accumulation (arrows) in dorsal marginal cells of a WT embryo at high stage. (F-I) In situ hybridization analysis of chordin expression in indicated embryos at dome stage. Animal pole view with dorsal region on the right. (J) Endogenous ß-catenin nuclear accumulation (arrows) in dorsal marginal cells of an MZdvl2;MZdvl3a mutant at high stage. (K-O) Phenotypes of indicated embryos at 11.5 hpf. Lateral view, with a dvl1a/1b/3bMOs-injected MZdvl2;MZdvl3a embryo also shown in dorsal view (O). (P-S) Phenotypes of indicated embryos at 30 hpf. Lateral view, note that injection of dvl1a/1b/3bMOs does not change the phenotype of WT and MZdvl2;MZdvl3a embryos. (T) Statistical analysis of the extent of axis extension delay in indicated embryos at 11.5 hpf. Bars represent the mean ± s.d. from indicated numbers of embryos (NS, not significant). Scale bars: (A-D, F-I) 400 μm; (E, J) 25 μm; (K-S) 400 μm.

The inducing-activity of Wnt8a is blocked in MZdvl2;MZdvl3a mutants.

In situ hybridization analysis of ectopic organizer gene expression at dome stage following Wnt8a overexpression in indicated embryos. Animal pole viewed embryos with dorsal region on the right. (A-F) goosecoid expression pattern in indicated embryos. (G-M) chordin expression pattern in indicated embryos. Notice that reduction of Dvl dosage progressively inhibits the inducing activity of Wnt8a, and MZdvl2;MZdvl3a embryos display a complete absence of ectopic goosecoid and chordin expression. All mutant embryos were derived from crosses between a female mdvl2^+(-)/-;dvl3a^-/- and a male dvl2^+/-;dvl3a^-/- fish, and were genotyped after in situ hybridization. Scale bar: 400 μm.

Cooperation of maternal and zygotic Dvl2 and Dvl3a in AP patterning.

Mutant embryos were selected by the extent of CE defects at 12 hpf, and genotyped after in situ hybridization. Immunostaining was performed at shield stage, followed by genotyping. (A-B’) Expression patterns of sp5l and tbx16l in WT embryos. Arrowheads in lateral viewed embryos indicate the anterior end and posterior end. Other embryos are dorsal-posterior view. (C, C’) The expression of sp5l is not affected in Zdvl2;MZdvl3a mutants. (D, D’) Strongly reduced tbx16l expression in Zdvl2;MZdvl3a mutants. (E, E’) Strongly reduced sp5l expression in MZdvl2;MZdvl3a mutants. (F, F’) Residual expression of tbx6l in MZdvl2;MZdvl3a mutants. (G) TOPFlash luciferase reporter activity in Zdvl2;MZdvl3a and MZdvl2;MZdvl3a mutants at 12 hpf. Bars represent the mean ± s.d. from three independent experiments (*, P<0.05; ***, P<0.001). (H) Endogenous ß-catenin nuclear accumulation (arrows) in ventral marginal cells of a WT embryo at shield stage. (I) Absence of ß-catenin nuclear accumulation in ventral marginal cells of an MZdvl2;MZdvl3a mutant at shield stage. (J-P) In situ hybridization analysis of indicated genes in WT and MZdvl2;MZdvl3a embryos at 12 hpf. Dorsal view with anterior to the top (J, K), dorsal view (L-O), and dorsal-posterior view (P). (R-U) Analysis of hoxb1b and zygotic Wnt/ß-catenin target genes by RNA-sequencing at 12 hpf. Note the significant decrease in RPKM (reads per kilobase million) for these genes in MZdvl2;MZdvl3a mutants. Bars represent the mean ± s.d. from three independent samples (*, P<0.05; **, P<0.01). (V, W) Lateral (V) and dorsal-posterior (W) views show rescue of axin2 and tbx16l expression in MZdvl2;MZdvl3a mutants at 12 hpf, following LiCl (0.3 M) treatment for 8 minutes at 5 hpf. op, otic placode; p, pronephric mesoderm; tb, tailbud. Scale bar: (A-F’) 400 μm; (H, I) 20 μm; (J-P) 400 μm; (V, W) 400 μm.

Dvl2 and Dvl3a dosages in CE movements and Wnt/PCP signaling activity.

(A-H) Representative live images of WT and mutant embryos at 11.5 hpf. Lateral view, with anterior region on the top. (I-P) Dorsal view of indicated embryos simultaneously hybridized with ctslb, dlx3, and tbxta probes to reflect the position of prechordal plate mesoderm, neural plate borders, and notochord, respectively. MZdvl3a mutants were from intercrosses between dvl3a^-/- carriers; MZdvl2 mutants were from crosses between female dvl2^-/- fish and male dvl2^+/- fish; Zdvl2;Zdvl3a mutants were from crosses between female dvl2^+/-;dvl3a^-/- fish and male dvl2^+/-;dvl3a^+/- fish; Zdvl2;MZdvl3a mutants were from intercrosses between dvl2^+/-;dvl3a^-/- carriers; MZdvl2;MZdvl3a mutants were from crosses between female mdvl2^+(-)/-;dvl3a^-/- fish and male dvl2^+/-;dvl3a^-/- fish. (Q) Statistical analysis shows that progressive reduction of Dvl2 and Dvl3a dosages increasingly aggravates axis extension defect. The embryos were imaged at 11.5 hpf, and genotyped before measuring the angle (inset). Bars represent the mean ± s.d. from indicated numbers of embryos collected from three independent experiments, and asterisks above the bars indicate significance with respect to WT embryos (***, P<0.001; ****, P<0.0001). (R) Reduced AP1 reporter activity in Zdvl2;MZdvl3a and MZdvl2;MZdvl3a mutants at 12 hpf. Bars represent the mean ± s.d. from three independent experiments (*, P<0.05, ***, P<0.001, ****, P<0.0001). (S) Rescue of cell polarity of MZdvl2;MZdvl3a mutant cells by caJNK. Vertical bidirectional arrows indicate AP orientation. (T) Statistical analysis of the length (l) to width (w) ratio in indicated cells. Bars represent the mean ± s.d. from at least 10 cells in two representative images (*, P<0.05, **, P<0.01, ****, P<0.0001). (U-X) Still frames from live time-lapse images show the dorsal convergence and movement behaviors of lateral cells in indicated embryos (see also S1–S4 Movies). The trajectories of 10 randomly selected cells are traced. (Y) Statistical analysis of the ratio between the trajectory and the net mediolateral distance (as indicated by a horizontal arrow). Bars represent the mean ± s.d. from at least 15 cells in two or three representative images (***, P<0.001; ****, P<0.0001). Scale bars: (A-H) 400 μm; (I-P) 400 μm; (S) 20 μm; (U-X) 50 μm.

Phenotypes of MZdvl1a, MZdvl1b and MZdvl3b mutants at different stages.

(A-C) WT embryos. (D-F) MZdvl1a mutants. (G-I) MZdvl1b mutants. (J-L) MZdvl3b mutants. All embryos are lateral view. The anterior region of 11.5 hpf embryos is positioned on the top. Scale bars: (A, D, G, J) 400 μm; (B, E, H, K) 400 μm; (C, F, I, L) 400 μm.

Analysis of mutant dvl2 and dvl3a transcripts.

(A) Semi-quantitative RT-PCR analysis to detect mutant dvl2 and dvl3a transcripts at 1-cell stage. ß-actin served as a loading control. NMD can be observed for dvl2 and dvl3a transcripts, respectively. (B, C) Quantification of mutant dvl2 and dvl3a mRNA levels in Mdvl2 and MZdvl3a mutants. The expression level in WT embryo is set as 1 after normalization with ß-actin. Bars represent the mean ± s.d. from three experiments (*, P<0.05; ****, P<0.0001). (D, E) In situ hybridization analysis of dvl2 transcripts in WT and Mdvl2 embryos. RT, reverse transcriptase.

Development of cyclopia in dvl2 and dvl3a mutants.

(A-E) Ventral view of representative images of normal and different degrees of eye phenotypes at 3 dpf. (F) Quantitative analysis of different degrees of eye phenotypes in indicated mutants. Except for WT embryos, all mutants were analyzed from three independent crosses using the same fish pair (indicated below the horizontal line). Numbers on the top of each column indicate total embryos carrying the indicated genotypes (above the horizontal line). Scale bar: (A-E) 400 μm.

Absence of dorsalization in MZdvl2;MZdvl3a mutants.

In situ hybridization analysis of dorsoventral markers. (A, B) The expression of axin2 at 50% epiboly is inhibited in MZdvl2;MZdvl3a mutants. (C-F’) The expression domains of goosecoid (gsc) and chordin in MZdvl2;MZdvl3a mutants at 60% epiboly do not show ventral expansion, but reflect impaired AP extension. Scale bar: 400 μm.

Overexpression of ß-catenin partially rescues AP patterning in MZdvl2;MZdvl3a mutants.

Embryos at 1-cell stage derived from crosses between female mdvl2^+(-)/-;dvl3a^-/- fish and male dvl2^+/-;dvl3a^-/- fish were injected with ΔN-ß-catenin mRNA (50 pg). Following in situ hybridization or phenotype analysis, the embryos were subjected to genotyping. (A-F) ΔN-ß-catenin partially rescues axin2 and tbx16l expression in MZdvl2;MZdvl3a mutants. Arrowheads indicate the axin2 anterior expression domain. (G-I) ΔN-ß-catenin partially rescues tail development in Zdvl2;MZdvl3a mutants (arrow). Scale bar: (A-F) 400 μm; (G-I) 400 μm.

Maternal contribution of Dvl2 and Dvl3a in axis extension.

The embryos were imaged at 11.5 hpf and 30 hpf, followed by genotyping, and the extent of axis extension defect was reflected by the angle between the anterior end and the posterior end at 11.5 hpf. (A, B) WT embryos. (C, D) Mdvl2;Mdvl3a mutants from crosses between female mdvl2^+(-)/-;dvl3a^-/- fish and male WT fish were genotyped for the presence of a novel indel (red) along with a WT allele in the dvl2 locus. They have dvl2 and dvl3a heterozygous mutations. (E, H) The offspring with the two possible genotypes (dvl2^+/+;dvl3a^+/- and dvl2^+/-;dvl3a^+/-; the effects of these mutations are indicated in parenthesis), derived from a cross between female dvl2^+/-;dvl3a^-/- fish and male WT fish, are maternal mutants for dvl3a, with a reduced dosage of maternal dvl2 in both cases, despite of the genotype. (I) Statistical analysis of the extent of axis extension delay in three types of maternal mutants. Bars represent the mean ± s.d. from indicated numbers of embryos, and asterisks above the bars show significance with respect to WT embryos (***, P<0.001; ****, P<0.0001). (J) Quantitative analysis of defective axis extension at 11.5 hpf. Each type of cross was done using three independent fish pairs, and total numbers of embryos analyzed are indicated on the top of each column. Subjective measures of axis extension defect are shown on the embryos at 11.5 hpf. Scale bar: (A, C, E, G) 400 μm; (B, D, F, H) 400 μm.