The wnt9a/b paralogous genes are essential for myocardial morphogenesis. (A) Model depicting the zebrafish embryonic wild-type heart during stages of cardiac cone formation and leftward jogging. At 15 hpf, bilateral populations of cardiomyocyte progenitor cells start converging towards the embryonic midline (black arrows). After 20 hpf, endocardial and myocardial progenitor cells initiate unilateral migrations towards the left, which causes a tilting of the heart cone and an elongation of the heart tube. (B-E) Maximum projections of confocal z-sections of 48 hpf zebrafish hearts. Unlike the looped wild-type heart (B), the wnt9^DKO heart is collapsed and the atrium (highlighted by Myh6 staining) surrounds the ventricle (C). The ventricle fails to fuse anteriorly (white arrowhead; n=10/10 embryos analyzed). wnt9b^sa20083 homozygous; wnt9a^sd49 heterozygous mutants are phenotypically similar to wnt9^DKO mutants (D; n=24/24 embryos analyzed), whereas wnt9a^sd49 homozygous;wnt9b^sa20083 heterozygous mutants have a wild-type phenotype (E; n=23/23 embryos analyzed). (F-I) Maximum projections of confocal z-scan sections with ventral views of the zebrafish heart field. At 20 hpf, wild-type cardiomyocyte progenitor cells have reached the midline and generated the heart cone (F). In wnt9a/b double morphants, anterior cardiomyocyte progenitor cells have not reached the embryonic midline and failed to fuse into the heart cone (G; n=10/10 embryos analyzed). At 24 hpf, cardiomyocyte progenitor cells have migrated towards the left side and generated an elongated heart tube (H). In wnt9a/b double morphants, cardiomyocyte progenitor cells have failed to fuse anteriorly and remain at the embryonic midline (I; n=10/10 morphants analyzed). The embryonic midline is marked with anti-ZO-1 staining. (J-M) Maximum projections of confocal z-scan sections with ventral views of endocardial progenitor cells. At 20 hpf, wild-type (J) and wnt9a/b double morphant (K) endocardial progenitor cells have reached the embryonic midline (n=6/6 morphants analyzed). At 24 hpf, endocardial progenitor cells have migrated towards the left side of the embryo and contribute to the elongating heart tube (L). In wnt9a/b double morphants, endocardial progenitor cells initiate a leftward movement (white arrowhead) but fail to complete leftward jogging (M; n=7/7 morphants analyzed). The embryonic midline is marked with anti-ZO-1 staining. (N-Q) Whole-mount in situ hybridization of vmhc cardiac expression during the stages of cardiac cone formation and leftward jogging. At 20 hpf, vmhc is expressed within the cardiac cone (N). In wnt9^DKO mutants, expression of vmhc reveals the failure of the cardiac cone to close anteriorly (O; n=7/7 embryos analyzed). At 24 hpf, vmhc marks the wild-type heart tube during cardiac leftward jogging (P), whereas in wnt9^DKO mutants, the heart remains at the embryonic midline (Q; n=7/7 embryos analyzed). White dashed lines indicate the embryonic midline. (R) Model depicting the zebrafish embryonic wnt9^DKO heart during the stages of cardiac cone formation and leftward jogging. The bilateral population of cardiomyocyte progenitor cells fails to complete anterior migrations and instead forms a bilateral wing-like structure (black arrows), which fails to undergo leftward jogging. Scale bars: 30 µm.

Wnt9b has a permissive role during the anterior convergence movements and fusion of cardiomyocyte progenitor cells. (A,B) Single images from light-sheet microscopy-derived time-lapse movies of cardiomyocyte progenitor cell migrations. Tracks of cardiac progenitor cell migrations during cardiac cone formation are highlighted. In the wildtype (A), cardiomyocyte progenitor cells migrate towards the embryonic midline and fuse first posteriorly and then anteriorly. In wnt9a/b double morphants (B), cardiomyocyte progenitor cells complete migrations and fuse posteriorly, but fail to converge anteriorly. Cardiomyocyte progenitor cells included in the tracking analysis (shown in C,D) have been marked. Dashed yellow lines and white arrowheads indicate the converging anterior part of the heart cone. All the other cell tracks are marked with white dots. (C,D) Tracking analyses of cardiomyocyte progenitor cells during cardiac cone formation. Quantification of the straightness index (C) and velocity (D) of wild-type versus wnt9a/b double morphant cardiomyocyte progenitor cells in anterior positions (wildtype, n=5 embryos; wnt9a/b double morphants, n=5 embryos). Box plots show mean values (middle bars) and first to third interquartile ranges (boxes). Lower and upper whiskers indicate minimum and maximum values, respectively (*P<0.05, **P<0.01; two-tailed, unpaired Student's t-test). (E-F′) Fluorescent whole-mount in situ hybridization of wnt9b expression at stage of cardiomyocyte progenitor cell convergence towards the embryonic midline. (E,E′) The lateral plate mesoderm (white arrows, dashed line) is devoid of wnt9b expression. Instead, wnt9b expression is in the neural tube (white asterisks) and neural retina (n=10/10 embryos analyzed). (F,F′) Neural progenitor cells expressing Tg(elavl3:EGFP)^knu3 do not co-express wnt9b (n=10/10 embryos analyzed). (G-J) Maximum projections of confocal z-scan sections with ventral views of cardiomyocyte progenitor cells. In wildtype, cardiomyocyte progenitor cells have fused into the heart cone (G). Upon heat shock-induced overexpression of wnt9b, the size of the heart cone increases (n=22/22 embryos analyzed) (H). At 24 hpf, wild-type cardiomyocyte progenitor cells undergo leftward cardiac jogging (I). Upon heat shock-induced overexpression of wnt9b, cardiomyocyte progenitor cells undergo leftward jogging but are more dispersed (J; n=13/13 embryos analyzed). The embryonic midline is marked by ZO-1 staining (G,H) and dashed lines (I,J). Scale bars: 40 µm (A,B); 30 µm (E-J).

Wnt9a/b signal via canonical and non-canonical pathways during cardiac cone formation. (A-C) Quantification of cardiomyocyte numbers in wildtype versus wnt9^DKO mutants during cardiac cone formation. Cardiomyocyte numbers increase in wnt9^DKO mutants (wildtype, n=9 embryos; wnt9^DKO, n=11 embryos). Data are mean±s.e.m. Dots represent single values (**P<0.01; two-tailed, unpaired Student's t-test). (D-G) Maximum projections of confocal z-scans with ventral views of cardiomyocyte progenitor cells during cardiac cone formation. In DMSO-treated control embryos, the heart cone has formed by 20 hpf (D) whereas in IWR-1- and TNP-470-treated embryos, cardiomyocyte progenitor cells have not reached the embryonic midline and failed to fuse into a heart cone (E; n=21/21 embryos analyzed). Heart cones of embryos treated only with IWR-1 (F; n=6/6 embryos analyzed) or TNP-470 (G; n=11/11 embryos analyzed) have no anterior fusion defects. The treatment with IWR-1, TNP-470, their combination or DMSO (control) was carried out from 14 to 20 hpf. The embryonic midline is labeled for ZO-1. Scale bars: 30 µm.

Wnt9a/b control cardiac cone formation and leftward jogging independently of epithelial cell polarity or endoderm. (A-B″) Maximum projections of confocal z-scans with ventral views (A,A′ and B,B′) or x-z section views (A″,B″) of the zebrafish heart field. Maximum projection of a 7 µm deep z-stack showing subapical tight junctional ZO-1 staining within the myocardium at 20 hpf (A′). x-z views (A″) show the correct subapical localization of tight junctions between cardiomyocytes (white arrowheads; n=9/9 embryos analyzed). In wnt9^DKO mutants, apico-basal cell polarity is intact, as indicated by a maximum projection of confocal z-scans (B) or of a 7 µm deep z-stack (B′), which reveals intact tight junctions. x-z view (B″) of the wnt9^DKO heart field which shows intact tight junctions within the myocardium (white arrowheads; n=12/12 embryos analyzed). (C-D″) Maximum projections of confocal z-scans with ventral views of cardiomyocyte progenitor cells during cardiac cone formation and of the underlying endoderm. In the wildtype (C-C″), the endoderm is present at the midline while the cardiac cone is completing its fusion. In wnt9a/b double morphants (D-D″), the endoderm is present at the embryonic midline, whereas the anterior portion of the cardiac cone does not migrate towards the embryonic midline and fails to fuse into the heart cone (n=15/15 embryos analyzed). Scale bars: 30 µm (A,B,C-D″); 10 µm (A′,B′).

The activity of Wnt9a/b does not depend on the endocardium and prevents premature cardiac differentiation in the heart cone. (A-F) Maximum projections of confocal z-scans with ventral views of cardiomyocyte progenitor cells. In the wildtype (A), cardiomyocyte progenitor cells complete fusion into the heart cone by 20 hpf and undergo leftward jogging by 24 hpf (B). In klf2a/b double morphants (C), cardiomyocyte progenitor cells are not defective during the fusion of the cardiac cone (n=10/10 morphants analyzed) or during leftward jogging (D; n=9/9 morphants analyzed). In npas4l morphants (E), which lack all endocardial cells, cardiomyocyte progenitor cells do not fuse correctly at the embryonic midline (n=8/8 embryos analyzed). Yet, the heart cone undergoes leftward jogging (F; n=10/10 embryos analyzed). (G-J) Maximum projections of confocal z-scans of 48 hpf zebrafish hearts. Different to the looped wild-type heart (G), the npas4l morphant heart (H) is strongly ballooned. In comparison, the wnt9a/b double morphant mutant heart (I) is collapsed. The knockdown of wnt9a/b in npas4l^m378 mutants produces the wnt9a/b double morphant cardiac phenotype with a collapsed heart (J; n=30/30 embryos analyzed). This suggests that the wnt9a/b morphant cardiac phenotype is not due to endocardial defects. (K) Quantifications of changes of mRNA expression levels of myocardial differentiation markers by quantitative real-time PCR. Shown is the comparison in Tg(hsp70l:wnt9b_IRES_EGFP)^pbb48 transgenic embryos with their wild-type heat-shocked siblings (n=4 experiments; *P<0.05; two-tailed, paired Student's t-test). (L) Model depicting how canonical and non-canonical signaling by Wnt9a/b may affect the midline-directed migration of cardiomyocyte progenitor cells, formation of the cardiac cone and leftward jogging. Scale bars: 30 µm.