Zebrafish stat3 mutants generated using TALEN method are strong/null alleles.

(A) Design of TALEN pair targeting zebrafish stat3 gene (yellow filled text). (B) Sequence alignments and schematics of Stat3 proteins encoded by stl27, stl28 and sa15744 alleles. (C) Expression patterns of stat3 transcripts detected by WISH in WT and MZstat3 embryos (lateral views). Arrowhead, nm, neuromast. (D) RT-PCR analysis of stat3 RNA levels in WT embryos at 1.3, 6, 8.3 and 24 hpf. gapdh was used as an internal control. (E) qRT-PCR analysis of stat3 transcript levels in WT and MZstat3 embryos normalized to gapdh. (F) Western blot detecting total Stat3 in WT, MZstat3 and MZstat3 embryos overexpressing Flag-tagged Stat3 at 6 hpf. β-actin was used as a loading control. See also S1 Fig.

Zebrafish stat3 mutants develop late-onset scoliosis and cannot survive to adulthood.

(A) Growth curve of stat3^stl28/stl28 fish and siblings. (B) Body length of stat3^stl27/stl27 and stat3^stl27/+ and WT siblings at 18 dpf. Different genotypes were kept in the same tank and fed exclusively with rotifers to diminish food competition. (C) Survival rate of stat3^stl27/stl27, stat3^stl27/+ fish and WT siblings. (D) Images of live animals and Alizarin stained skeletons showing WT and zygotic stat3 mutant fish at 15 dpf (anterior to the left). Vertebral structures in the boxed region were revealed via Alizarin Red live staining and confocal imaging, shown in insets. (E) Images of live animals and Alizarin stained skeletons showing a WT and a zygotic stat3^stl28/stl28 mutant fish with two curves near the caudal fin (inset) at 22 dpf. Anterior to the left. (F) Images of live animals and Alizarin stained skeletons showing a WT and a scoliotic stat3^stl27/stl27 mutant fish at 46 dpf (anterior to the left). (G) μCT imaging showing body curvatures of WT sibling and stat3^stl27/stl27 mutant at 41 dpf (anterior to the left). (H) μCT imaging showing body curvatures of WT sibling and stat3^{sa15744/sa15744} mutant at 43 dpf (anterior to the left, arrow denote bend in spine). (I-J) Quantification of bone mineral density (H) and bone/tissue volume ratio (Bv/Tv, I) of stat3^stl27/stl27 (G). (K-L) Transcript levels of tnfα and il6 in WT and stat3^stl27/stl27 mutant animals detected by qRT-PCR at indicated developmental stages. (M-N) Survival rate and percentage of mutant fish with various numbers of body curves of MZ stat3^stl27/stl27 mutants. *p<0.05, ***p<0.001, ****p<0.0001, error bars = SEM. See also S2 Fig.

MZstat3 mutants exhibit transient and mild extension defects in axial mesoderm during gastrulation.

(A) Live images of WT and MZstat3 embryos shown in lateral (a, c) and dorsal view (b). The insets in c show the part of notochord above yolk extension in 30 hpf embryos. Arrowhead, nt, notochord. (B) Morphometric analysis of AP axis extension of 30 hpf embryos shown in A(c). (C) papc in presomitic mesoderm and dlx3b marking neuroectoderm boundary in 1-somite stage WT and MZstat3 embryos (dorsal view). (D) Measurement of ML width of papc expression domain (pink in C). (E) Expression of ntl in notochord and tail in 1-somite stage WT, Zstat3, Mstat3, MZstat3, and MZstat3 embryos overexpressing Stat3-F, as well as WT and MZstat3 embryos injected with 5 ng MO1-stat3 (lateral view). Phenol red was used as injection control. (F) Measurement of notochord length in embryos in E (blue lines in E). ****p<0.0001, n.s. = non-significant, error bars = SEM. See also S3 Fig.

MZstat3 mutants show neither obvious cell polarity defects during C&E nor interaction with zebrafish PCP mutants.

(A) Confocal image of dorsal mesoderm in WT, Mstat3, and MZstat3 1-somite stage embryos labeled with mGFP (anterior to the top). Cell shape and orientation of notochord cells outlined in yellow were analyzed as illustrated in D. (B) Measurement of notochord width at 1-somite stage. (C) Quantification of the number of cells across the ML notochord axis at 1-somite stage. (E) Cumulative distribution of notochord cell orientation in WT, Mstat3, and MZstat3 embryos. (F) Cell shape analysis represented by length-to-width ratio (LWR). (G-I) Long axis (length, G), short axis (width, H) and average area (I) of in WT, Mstat3 and MZstat3 notochord cells. (J) A spectrum of eye separation phenotypes at 3 dpf with C1 representing WT eye spacing and C5 representing the most severe phenotype, cyclopia. Ventral view, anterior to the top. (K) Penetrance and expressivity of eye separation phenotypes of Ztri, Ztri;Zstat3^stl27/+, and Ztri;Zstat3 embryos. Eye separation phenotypes were also quantified by cyclopia index (CI) as previously described [36].(L) Penetrance and expressivity of eye separation phenotypes of MZslb, MZslb;Zstat3^stl27/+, and MZslb;Zstat3 embryos. ****p<0.0001, n.s. = non-significant, error bars = SEM. See also S4 Fig.

Stat3 promotes cell cycle progression during zebrafish embryogenesis.

(A) Immunofluorescent anti-pH3 labeling of proliferating cells (red) and DAPI labeling all nuclei (blue) in WT, Mstat3, and MZstat3 embryos at 6 hpf (animal view) and 10 hpf (dorsal view). (B, D) Quantification of mitotic cell number at 6 and 10 hpf. (C, E) Quantification of total cell number at 6 and 10 hpf. (F and G) Average length of each cell cycle (F) and timing of mitosis (G) from Cycle 5 to Cycle 9 in embryos from WT, stat3^stl27/+, and stat3^stl27/+ females. (H-K) Analyses of cell divisions from Cycle 5 to Cycle 9 in 10 pg (H and I) and 25 pg (J and K) stat3-F injected MZstat3 embryos. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001, n.s. = non-significant, error bars = SEM. See also S5 Fig.

Stat3 promotes post-MBT cell divisions but does not regulate cell division orientation during zebrafish embryogenesis.

(A) Experimental design for post-MBT cell cycle analyses. Embryos labeled ubiquitously with histone2B-RFP (H2B-RFP) were mosaically injected with memCherry or mGFP + stat3-F mRNA at 8-cell stage for lineage tracing. Labeled clones (B and C) within the same embryo were monitored using confocal time-lapse microscopy (See also Materials and methods). (D) Analyses of cell cycle lengths for Cycle 11–13 in WT, WT overexpressing Stat3-F, MZstat3, and MZstat3 overexpressing Stat3-F embryos. (E-G) Cell division orientation in dorsal neuroectoderm in WT (E) and MZstat3 (F) embryos during 1–3 somite stages, with 0 degrees representing mediolaterally aligned cell divisions and 90 degrees representing anteroposteriorly aligned cell divisions, quantified in G. *p<0.05, **p<0.01, ****p<0.0001, n.s. = non-significant, error bars = SEM. See also S7 Fig.

Cell number reduction correlates with axis extension defects in stat3 mutant embryos.

(A-D) Confocal image of dorsal mesoderm expressing mGFP in WT (A), Mstat3 (B), MZstat3 (C) embryos, and MZstat3 embryos overexpressing Stat3-F (D) at 11 hpf (anterior to the top). Somite boundaries are outlined in green. Adaxial cells are outlined in orange. Green arrows show AP dimension of each somite. Adaxial cells and notochord cells aligning along the AP axis are numbered in yellow. (E-G) Quantification of the average somite AP dimension (E), number of notochord cells (F) and number of adaxial cells (G) between adjacent furrows of the first two somites. (H-K) Schema of AP extension of the notochord and presomitic mesoderm. st, somite; nt, notochord. **p<0.01, ****p<0.0001, error bars = SEM.

Inhibition of cell proliferation using hydroxyurea and aphidicolin leads to axis extension defects in zebrafish gastrulae.

(A) Immunofluorescent anti-pH3 labeling of proliferating cells (red) and total nuclei labeled with DAPI labeling (blue) in DMSO-treated control embryos and hydroxyurea+aphidicolin (H+A)-treated embryos at 6 hpf (animal view) and 10 hpf (dorsal view, anterior to the top). (B-E) Quantification of pH3+ cells (B and D) and DAPI+ cells (C and E) in A. (F and G) Expression of ntl at 1-somite stage (lateral view, anterior to the top). (H) Quantification of notochord length (blue lines in F and G). (I and J) Confocal image of dorsal mesoderm in 3-somite stage embryos expressing mGFP with somite AP dimension illustrated with green arrow, somitic boundaries outlined in green, adaxial cells outlined in orange, adaxial cells and notochord cells between adjacent somitic boundaries numbered in yellow (dorsal view, anterior to the top). (K-M) Quantification of somite AP dimension (K), numbers of adaxial cells (L) and notochord cells (M) in I and J. (N and O) Schema of AP extension of the notochord and presomitic mesoderm. st, somite; nt, notochord. (P-V) Dorsal view showing cells labeled with mGFP in DMSO-treated (P) and drug-treated (Q) embryos at 1-somite stage (anterior to the top). Analyses of notochord cells’ orientation (R), shape (S), long axis (length, T), short axis (width, U) and size (V). ****p<0.0001, n.s. = non-significant, error bars = SEM.

Stat3 promotes cell cycle progression during zebrafish embryogenesis via upregulation of cdc25a expression.

(A) qRT-PCT analysis of stat3 and cdc25a transcript levels at 1.5 hpf. All results shown were normalized to gapdh. (B) qRT-PCR analysis of cdc25a transcript levels at 8.3 hpf. (C and D) Pre-MBT (C) and post-MBT (D) cell cycle length analyses in 1-somite stage MZstat3 embryos injected with 25 or 50 pg cdc25a RNA. Phenol Red was used as injection control. (E and F) ntl expression in 1-somite stage MZstat3 and WT gastrulae misexpressing Cdc25a (lateral view, anterior to the top). Notochord length (blue line) was quantified in F. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001, error bars = SEM. See also S9 Fig.

Various vertebral abnormalities contribute to the scoliosis phenotype in stat3 animals.

(A) Vertebrae of a WT larva at 29 dpf; anterior to the left. (B-H) Images of Alizarin stained skeletons showing various vertebral abnormalities. B and C, normal vertebral body and end plates, tilted intervertebral discs; D and E, bent vertebral body and non-perpendicular end plates; F-H, fractures and extra bony matrix. (I) Variations in larvae body length of stat3 mutant and siblings at 22 dpf.

Injection of stat3 morpholino results in dose-dependent axis extension defect.

(A) ntl WISH in the notochord tissue in stage-matched control and stat3 morphant embryos at 2-somite stage injected with various doses of MO1-stat3 at one cell stage (lateral view, dorsal to the right, anterior to the top).

Convergence of the notochord into a single-cell wide column is not affected in M and MZstat3 embryos.

(A-C) Confocal microscope image of 5-somite stage embryos in dorsal view, in which cell membranes are labeled with mGFP: WT (A), Mstat3 (B), and MZstat3 (C) (anterior to the top).

Pre-MBT cell divisions are lengthened in MZstat3 mutant embryos compared to WT.

Depicted are confocal microscope snap-shots from a full pre-MBT cell cycle (Cycle 6, 32-cell stage to 64-cell stage) in WT and MZstat3 embryos from S phase to S phase.

Stat3 limits apoptosis during embryogenesis.

(A-E) Apoptosis in WT (A, D) and MZstat3 (B, E) embryos at 10 hpf (A, B) and 11 hpf (D, E) detected by TUNEL labeling (dorsal view, anterior to the top). Number of TUNEL-positive cells are quantified in C and F. (G, H) birc5a/survivin (G) and bcl2a (H) transcript levels in WT and MZstat3 embryos at 1.5 hpf and 8.3 hpf determined by qRT-PCR. *p<0.05, **p<0.01, n.s. = non-significant, error bars = SEM.

Effect of cdc25a overexpression in WT on notochord cell size and morphogenesis.

(A and B) Dorsal view of 1-somite stage embryos showing cells labeled with mGFP: control embryos (A) and embryos injected with 25 pg to 50 pg cdc25a mRNA (B) (anterior to the top). (C-H) Analyses of notochord cells’ orientation (D), shape (E), long axis (length, F), short axis (width, G) and size (H) in A and B. (I and J) Confocal image of dorsal mesoderm in 3-somite stage control and cdc25a-overexpressing embryos labeled with mGFP with somite AP dimension illustrated with green arrow and somitic boundaries outlined in green (dorsal view, anterior to the top). (K-M) Quantification of somite AP dimension (K), numbers of adaxial cells (L) and notochord cells (M) in I and J. ****p<0.0001, error bars = SEM.

Onset of zygotic gene expression in MZstat3 and WT embryos.

(A) Expression of bozozok/dharma (boz), a zygotic target of β-catenin, in WT (top, animal and lateral view, AP, animal pole) and MZstat3 embryos (bottom) at 4 hpf. Yellow arrows denote boz expression domain. (B) Expression of floatinghead (flh), an early zygotic gene whose expression domain rapidly changes shape with convergence and extension of the axial mesoderm, in WT and MZstat3 gastrulae in dorsal view and animal pole towards the top at 6 hpf. Yellow line indicates width of expression domain. (C) The mediolateral dimension of the flh expression domain at 6 hpf is not significantly different between WT and MZstat3 embryos. n.s. = non-significant.