Hipk2 Is Required for Dvl-Mediated Early Embryonic Events in Zebrafish

(A) Hipk2 is required for the stabilization of β-catenin in zebrafish. Zebrafish embryos were injected with MOs. Extracts were harvested from embryos at 8 hr postfertilization (hpf) and immunoblotted with anti-zebrafish β-catenin. α-Tubulin was used as a loading control.

(B and C) Hipk2 is essential for the β-catenin pathway-mediated posterior mesoderm formation and convergent extension (CE). Embryos were injected with MOs. Panels in (B) show whole-mount in situ hybridization of d2EGFP, tbx6, or ntla in OTM:d2EGFP-transgenic zebrafish embryos. Panels in (C) show whole-mount in situ hybridization of ntla and dlx3b or myod1 in zebrafish embryos. Blue, red, and orange two-way arrows indicate the width of neural plate, the length of notochord, and the width of myod1-expression domain, respectively. The percentages of embryos showing similar expression patterns and total number of MO-injected embryos (n) are shown under each image. In all images, A, P, V, and D indicate anterior, posterior, ventral, and dorsal sides. Scale bar, 200 μm.

(D and E) Hipk2 is required for the protein stability of endogenous Dvl (D) and exogenous mouse Dvl1 (E) Dvl in zebrafish. Zebrafish embryos were injected with MOs without (D) or with mouse Myc-Dvl1 and hemagglutinin (HA)-GFP mRNA (E). Extracts were harvested from the embryos at 8 hpf. In (D), embryo extracts were immunoprecipitated and then immunoblotted with anti-Dvl (see te Experimental Procedures for details). In (E), extracts were immunoblotted with indicated antibodies. GFP was used as a loading control.

See also Figures S1 and S2.

Hipk2 Regulates Dvl Stability in a Kinase Activity-Independent Manner in Mammalian Cells and Zebrafish

(A) Hipk2 is required for endogenous Dvl stability in HeLa cells. Cells were treated with either control siRNA or Hipk2 siRNA #1 or #2, and cell extracts were then were immunoblotted with anti-Dvl1/2/3 (which recognizes all types of Dvl), anti-Hipk2, and anti-β-tubulin (7beta;-tub). β-tub was used as a loading control. Relative Dvl protein levels were calculated by determining the ratio of Dvl to β-tub. Values are presented below the top panel as the relative percentage.

(B) Hipk2 WT and C, but not Hipk2 N, reversed the Hipk2-knockdown-induced reduction in Dvl expression. HeLa cells were treated with either control siRNA or Hipk2 siRNA#1 and then transfected with Myc-Dvl1 and Myc-tagged GFP (Myc-GFP) with or without Flag-tagged human Hipk2 (Flag-Hipk2) WT, N, and C.

(C) Hipk2 increased the Dvl protein levels via its C-terminal domain. HeLa cells were transfected with Myc-Dvl1 and Myc-GFP with empty vector () or Flag-Hipk2 WT, N, HC, H, or C.

(D) Schematic diagram of the human Hipk2 deletion mutants. Kinase, kinase domain; HD, homeodomain-interacting domain; PEST, PEST sequence.

(E) The C-terminal domain of Hipk2 is essential and sufficient for binding to Dvl. HeLa cells were transfected with empty vector () or Flag-tagged human Hipk2 (Flag-Hipk2) WT, N, HC, H, or C. The cell extracts were mixed with extracts prepared from cells transfected with Myc-Dvl1 and then immunoprecipitated with anti-Flag. Immunoprecipitated complexes were then immunoblotted with indicated antibodies. The expression of Myc-Dvl1 proteins in cell extracts was confirmed by immunoblotting with anti-Myc (“Input” lane).

(F) Hipk2 increases the protein level of Dvl1 in a kinase activity-independent manner. HeLa cells were transfected with Dvl1, Myc-tagged mouse Hipk2 (Myc-Hipk2) WT and KN, and Myc-GFP, as indicated.

(G and H) Hipk2 regulates the β-catenin pathway-mediated posterior mesoderm formation and CE via its C-terminal domain. Embryos were injected with MOs with or without human Hipk2 WT, N, or C mRNA. Panels show whole-mount in situ hybridization of d2EGFP in OTM:d2EGFP-transgenic zebrafish embryos (G) or ntla and dlx3b in nontransgenic zebrafish embryos (H) fixed at the indicated stages. The percentages of embryos showing similar expression patterns and total number of MO-injected embryos (n) are shown under each image.

See also Figures S2 and S3.

Hipk2-PP1c Stabilizes Dvl through Dephosphorylating the Conserved CK1 Sites

(A) Amino acid sequence alignment of the C-terminal CK1 phosphorylation regions within vertebrate Dvl proteins. The potential CK1 phosphorylation sites are indicated by red letters.

(B) Anti-pDvl1 recognizes the phosphorylation of the conserved CK1 sites on Dvl1 C-terminal region. HeLa cells were transfected with Myc-Dvl1 WT and mutants as indicated. Cell lysates were immunoblotted with indicated antibodies. To clearly show the migration-shift of Dvl proteins, equal quantities of Dvl proteins were resolved.

(C) The conserved CK1 sites on Dvl are phosphorylated in zebrafish. Extracts were harvested from the embryos injected with Myc-tagged mouse Dvl1 WT and 3A mRNA at 8 hpf and immunoblotted with indicated antibodies. In (B) and (C), High and low level phosphorylated forms of Dvl are indicated with red and blue arrowheads, respectively.

(D) Endogenous Dvl is phosphorylated at the conserved CK1 sites. HeLa cells were treated with DMSO () or 2 nM CalA and 10 μM MG132 for 3 hr. Cell extracts were subjected to immunoprecipitation with anti-Dvl1/2/3 (anti-Dvl). Immunoprecipitates were immunoblotted with indicated antibodies.

(E) Hipk2-PP1c dephosphorylates Dvl at the conserved CK1 sites in vivo. HeLa cells were transfected with Myc-Dvl1 WT and 3A and Flag-Hipk2. The cells were then treated with DMSO () or 2 nM CalA and 25 μM MG132 for 3 hr. Cell lysates were immunoblotted with indicated antibodies.

(F) Hipk2-PP1c directly dephosphorylates Dvl1 at the conserved CK1 sites. Aliquots of Dvl1 proteins were incubated with or without PP1c and Hipk2 and then immunoblotted with indicated antibodies. In (E) and (F), relative Dvl phosphorylation levels were calculated by determining the ratio of phospho-Dvl1 to total Dvl and these values are presented below the panels as the relative percentages.

(G–J) Dvl1 3A is less sensitive to Hipk2 and PP1c than Dvl WT. In (G)–(I), HeLa cells were treated with or without control siRNA, Hipk2 siRNA#2, or PP1c siRNA#1 and then transfected with Myc-Dvl1 WT and 3A, Myc-GFP, and Flag-Hipk2. In (J), zebrafish embryos were injected with MOs with mouse Myc-Dvl1 and HA-GFP mRNA. Extracts were immunoblotted with indicated antibodies. Relative Dvl1 protein levels were calculated by determining the ratio of Dvl1 to GFP. These values are presented below the top panel as the relative percentages.

(K–N) Hipk2 regulates the β-catenin pathway through Dvl dephosphorylation in zebrafish. In (K) and (M), classes of phenotypes induced by hipk2 spl MO injection are shown. In (K), class I, weak reduction of OTM:d2EGFP (d2EGFP); class II, strong reduction of d2EGFP. In (M), class I, weak reduction of tbx6; class II, strong reduction of tbx6. In (L) and (N), the distribution of phenotypes in 8 hpf OTM:d2EGFP-transgenic (L) or nontransgenic (N) zebrafish embryos injected with MOs with or without mouse Dvl1 WT or 3A mRNA (40 pg) is shown. n = the total number of MO-injected embryos.

See also Figures S5 and S6.

In Zebrafish, PP1c Cooperates with Hipk2 to Sustain Wnt Signal Transduction while Itch Counteracts Hipk2 Activity

(A–C) PP1c is required for the protein stability of endogenous Dvl (A) and β-catenin (B) and exogenous mouse Dvl1 (C) in zebrafish. Zebrafish embryos were injected with MOs without (A and B) or with (C) mouse Myc-Dvl1 and HA-GFP mRNA. Extracts were harvested from the embryos at 8 hpf. In (A), embryo extracts were immunoprecipitated and then immunoblotted with anti-Dvl. In (B) and (C), extracts were immunoblotted with indicated antibodies.

(D and E) PP1c is required for tbx6 expression and CE. Embryos were injected with control MO or ppp1caa MO (2 ng in D, 4 ng in E) with or without MO-insensitive ppp1caa mRNA. Panels show whole-mount in situ hybridization of tbx6, ntla, or dlx3b in zebrafish embryos.

(F) Itch counteracts Hipk2 activity in zebrafish. Embryos were injected with MOs as indicated. Panels show whole-mount in situ hybridization of OTM:d2EGFP (d2EGFP) or tbx6 in OTM:d2EGFP-transgenic (top) or nontransgenic (bottom) zebrafish embryos. In (D)–(F), the percentages of embryos showing similar expression patterns and the total number of MO-injected embryos (n) are shown under each image.

The expression patterns of zebrafish hipk2, ppp1caa, and itch and the effects of hipk2 MOs, ppp1caa MO, and itcha MO. Related to Figure 1.
(A) hipk2 mRNA, ppp1caa mRNA, and itcha mRNA are ubiquitously expressed throughout early embryogenesis. Panels show lateral views of whole mount in situ hybridization of hipk2, ppp1caa, and itcha in zebrafish embryos fixed at the indicated stages. Scale bar = 200 μm.
(B, C) hipk2 MO blocks translation of the hipk2 gene. To examine the effect of hipk2 MO, mRNA including the 5′UTR and 5′coding region (1–640 bp) of the hipk2 gene fused in-frame with a Myc-tag (B, 5′UTR-hipk2-Myc) was injected into zebrafish embryos with HA-GFP and hipk2 MO. The hipk2 MO annealing site is indicated by the red line in (B). In (C), embryo extracts were prepared at the indicated stages and then immunoblotted with anti-Myc and anti-HA antibodies.
(D, E) hipk2 spl MO inhibits the proper splicing of hipk2 mRNA. The genomic structure of the hipk2 gene is shown in (D). The hipk2 spl MO annealing site (red line), and hipk2 primers (P1: 5′- gaggtgctggagttcctgggtcga-3′, P2: 5′- ctctttatcccggagcatca-3′, P3: 5′- gcattcctgctcaataaggg-3′, and P4: 5′- ctctgctggcaagccctgcgtttg-3′) were used to monitor the effects on splicing by RT-PCR. Total RNA was extracted from 20 embryos at the indicated stages, reverse transcribed, and amplified by PCR. PCR of mRNA isolated from control MO-injected embryos using primers P1 and P4 produced a 669 bp band that corresponds to the correctly spliced transcript (E). Injection of hipk2 spl MO reduced the expression of this transcript. PCR of mRNA from embryos injected with hipk2 spl MO using primers P2 and P3 produced a 330 bp band (E). This product was mis-spliced (intron inserted) as shown in (D) and corresponds to a protein with a nonsense mutation immediately after exon 3. Elongation factor 1 alpha (ef1&alpjha;) mRNA was used as an internal control.
(F, G) ppp1caa MO blocks the translation of the ppp1caa gene in zebrafish embryos. To examine the effect of ppp1caa MO, mRNA including the 5′UTR and 5′coding region (1–267 bp) of the ppp1caa gene fused in-frame with a Myc-tag
(F, 5′UTR-ppp1caa-Myc) was injected into zebrafish embryos with HA-GFP and ppp1caa MO. The ppp1caa MO annealing site is indicated by the red line in (F). In (G), Embryo extracts were prepared at the indicated stages and then immunoblotted with anti-Myc and anti-HA antibodies.
(H, I) itcha MO blocks the translation of the itcha gene in zebrafish embryos. To examine the effect of itcha MO, mRNA including the 5′UTR and 5′coding region (1–372 bp) of the itcha gene fused in-frame with a Myc-tag (H, 5′UTR-itcha-Myc) was injected into zebrafish embryos with HA-GFP and itcha MO. The itcha MO annealing site is indicated by the red line in (H). In (I), embryo extracts were prepared at the indicated stages and then immunoblotted with anti-Myc and anti-HA antibodies.

Hipk2 and PP1c are involved in Dvl–mediated Wnt signaling during zebrafish early embryogenesis. Related to Figures 1 and 2.
(A) Hipk2 is required for the expression of tbx6 mRNA, but not the expression of endogenous β-catenin mRNA in zebrafish. Control MO or hipk2 MO were injected into one cell-stage zebrafish embryos as indicated. At 8 hpf, 30 embryos were collected and total RNA was purified. The relative mRNA expression levels of zebrafish β-catenins, ctnnb1 (β-cat1) and ctnnb2 (β-cat2), and tbx6 were analyzed by quantitative RT-PCR. The mean and standard deviation of triplicate measurements from one of two independent experiments are presented.
(B, C) Hipk2 is required for the activation of Wnt/β-catenin pathway in 4 and 10 hpf zebrafish embryos.
(D) Hipk2 regulates brain anterior-posterior patterning.
(E) hipk2 knockdown embryos failed to extend properly around the yolk at 14 hpf. In (B–E), embryos were injected with control MO, hipk2 MO, or hipk2 spl MO. In (B-D), panels show whole mount in situ hybridization of OTM:d2EGFP (d2EGFP), pax2a, eng2a, or otx2 in OTM:d2EGFP-transgenic (B, C) or non-transgenic (D) zebrafish embryos fixed at the indicated stages. In (E), panels show the lateral views of embryos. A, P, V, and D indicate anterior, posterior, ventral, and dorsal sides. Scale bar = 200 μm. The percentages of embryos showing similar gene expression patterns and total number of MO-injected embryos (n) are shown in each image.
(F, G) Hipk2 appears to function upstream of GSK-3β and downstream of Wnt8. Non-transgenic (F) or OTM:d2EGFP-transgenic (G) zebrafish embryos were injected with control MO or hipk2 spl MO without (F) or with zebrafish Wnt8 mRNA (G), and then untreated (G) or treated with DMSO or 10 μM of the GSK-3β inhibitor BIO from 5 to 8 hpf (F) and fixed at 8 hpf. Panels show whole mount in situ hybridization of otx2 or OTM:d2EGFP (d2EGFP). The percentages of embryos showing similar gene expression patterns and total number of MO-injected embryos (n) are shown in each image. Scale bar = 200 μm. (H) Anti-Dvl antibody recognizes zebrafish Dvl2 and Dvl3a proteins. HEK293 cells were transfected with empty vector (mock) or expression plasmids encoding zebrafish Dvl2 or Dvl3a and cell lysates were immunoblotted with anti-Dvl antibody (Abcam ab106844).
(I) Hipk2 regulates the β-catenin pathway–mediated posterior tissue formation via its C-terminal domain. Embryos were injected with control MO or hipk2 spl MO with or without 40 pg of MO=insensitive human Hipk2 WT, N, or C mRNA, as indicated. Panels show whole mount in situ hybridization of OTM:d2EGFP (d2EGFP) in OTM:d2EGFP-transgenic zebrafish embryos fixed at 10 hpf.
(J) Hipk2 functions in a kinase activity-independent manner in vivo. Embryos were injected with control MO or hipk2 spl MO with or without 100 pg (top and second panels) or 200 pg (bottom panels) of MO-insensitive mouse Hipk2 KN mRNA, as indicated. Panels show whole mount in situ hybridization of OTM:d2EGFP (d2EGFP) in OTM:d2EGFP-transgenic zebrafish embryos (top panels) or tbx6 or ntla and dlx3b in non-transgenic zebrafish embryos (second and bottom panels) fixed at the indicated stages.
(K) PP1c functions upstream of GSK-3β. Embryos were injected with control MO or ppp1caa MO as indicated, and then treated with DMSO or 10 μM of the GSK-3β inhibitor BIO from 5 to 8 hpf. Panels show whole mount in situ hybridization of tbx6 in embryos.
(L) PP1c and Hipk2 cooperate to regulate CE. Embryos were injected with control MO or a low dose of hipk2 MO (1 ng) or ppp1caa MO (2 ng) as indicated. Panels show whole mount in situ hybridization of ntla and dlx3b in embryos fixed at 10 hpf. In (I–L), the percentages of embryos showing similar gene expression patterns and total number of MO-injected embryos (n) are shown in each image. Scale bar = 200 μm.