DEV000026 Supplementary Material

Files in this Data Supplement:

• Supplemental Tables S1-S12 - Adobe PDF

• Supplemental Figure S1 -

  Fig. S1. Translational inhibition of Ttrap protein by Ttrap MO. Western blot analysis of HA-tagged zebrafish Ttrap, either alone (left lane), after co-injection with scrambled control MO (SCMO) (middle lane), or after co-injection with TtrapMO (1:1 mix of Ttrap MO1 and Ttrap MO2 (16 ng total), as used throughout this study) (right lane). Since none of the available antibodies against mammalian Ttrap were capable of detecting zebrafish Ttrap in vivo, we resorted to mRNA overexpresssion of full-length, HA-tagged zebrafish Ttrap (zfTtrap-HA) in fish embryos. Western blot analysis using an anti-HA antibody resulted in a single band of 55 kDa which is clearly downregulated in Ttrap^MO embryos. An anti-tubulin antibody was used to control for input and loading.

• Supplemental Figure S2 -

  Fig. S2. Multiple sequence alignment of vertebrate TTRAP proteins. Comparison of N-terminal residues shows relatively low conservation in this region, whereas C-terminal portion shows high degree of conservation beginning at the phosphohydrolase domain (underlined in blue) as described (Rodrigues-Lima et al., 2001). Black highlighting denotes 100% conserved residues, dark gray highlighting denotes 80% conserved residues, and light gray highlighting denotes 60% conserved residues. Red and black arrows denote threonine residues T88 and T92 respectively, which are phosphorylated by ALK4 kinase.

• Supplemental Figure S3 -

  Fig. S3. Test for functionality and expression of mutant and wild-type human TTRAP in zebrafish embryos. (A) Co-IP of Smad3 with TTRAP or TTRAP^T88A,^T92A. TTRAP-pCS3 or TTRAP^T88A,^T92A-pCS3 were co-expressed with HA-tagged Smad3 or Flag-tagged TTRAP-frame-shift as control in HEK293T cells and co-immunoprecipitated using anti-HA antibody. The precipitates were immunoblotted and proteins detected with monoclonal anti-HA (middle panel) or monoclonal anti-TTRAP antibody (top and bottom panel). (B) Western blot analysis of TTRAP and TTRAP^T88A,^T92A expression in zebrafish embryos. Total protein of zebrafish embryos injected with TTRAP-pCS2 or TTRAP^T88A,^T92A-pCS2 mRNA was isolated 2 hours, 4 hours and 6 hours after injection and incubated with TTRAP antibody to check expression levels. An anti-tubulin antibody was used to control for input and loading.

• Supplemental Figure S4 -

  Fig. S4. Targeting of eGFP mRNA to DFCs and expression in Kupffer’s Vesicle. (A-C, A′-C′) Live embryos at 7-somite stage. (A-C) Fluorescence images of eGFP^DFCOE embryos showing eGFP expression only in the yolk and KV and not in the rest of the embryo. (A′-C′) Corresponding transmitted light images of embryos in A-C. (B) Occasionally, eGFP could be found filling the entire KV (red arrowhead), whereas in the majority of embryos, such as in (C), eGFP only outlines part of the KV (red arrowhead), implying that only a subset of DFCs take up and/or express eGFP RNA/protein. Embryos in (B′ and C′) showing fully formed KV (black arrowheads). All eGFP^DFCOE embryos (n=160) developed normally. (A and A′) Lateral view. (B, B′, C, and C′) Posterior views.

• Supplemental Figure S5 -

  Fig. S5. Smad2 morpholino knockdown phenotype. Live observation of control^MO or Smad2^MO embryos at 30 hpf. (A,A′) Embryo injected with standard fluorescent control morpholino possesses normal morphology and clearly visible floorplate (fp, arrow). (B,B′) Smad2 morpholino-treated embryo displays anterior truncations, a curved shortened body axis, and absence of floorplate. This phenotype was observed at 4 ng and 8 ng MO doses. Lateral views, anterior to the left, dorsal at top.

• Supplemental Figure S6 -

  Fig. S6. Smad3b morpholino knockdown phenotype. Live observation of control^MO or Smad3b^MO embryos at 30 hpf. (A,A′) Wild-type embryo possesses normal morphology and clearly visible floorplate (fp, arrow). (B,B′) Smad3b morpholino-treated embryo displays anterior truncations, a curved shortened body axis, absence of floorplate, and a slightly enlarged intermediate cell mass (ICM, blue arrowhead), the latter phenotype indicative of partial ventralization. This phenotype was observed at 4 ng and 8 ng MO doses. Lateral views, anterior to the left, dorsal at top. Note: embryo in (B) is a composite photo to show all structures of embryo in focus.

• Supplemental Figure S7 -

  Fig. S7. Behavior of Ttrap^MO cells in maternal-zygotic one-eyed pinhead (MZoep) embryos. A total of 23 MZoep embryos were transplanted with a mix of wild-type and Ttrap^MO cells. Wild-type embryos were pre-labeled at the one-cell stage with either Alexa Fluor 568 dye (red fluorescence; Invitrogen), or with fluorescein-labeled Ttrap MO (green fluorescence; Genetools). In all 23 cases, Ttrap^MO cells failed to internalize. Left panels (bright field), middle panels (rhodamine filter), right panels (GFP filter). Time-course begins approximately at germ ring to shield stage (5.5 to 6 hpf), and ends at around late tailbud to 1- to 3-somite stage (10.5 to 11 hpf). (A,B) Representative time-lapse images of wild-type vs Ttrap^MO cell migration in an MZoep background. Green arrows indicate wild-type cells in region of endodermal progenitors and in a deeper cell layer and are therefore in a different focal plane, hence becoming more and more out of focus over the course of the time-lapse study. Red arrows indicate Ttrap^MO cells, which, like MZoep (or oep) cells fail to involute and remain at the surface, accumulating at the vegetal-dorsal thickening (black arrow), but not in the neighboring endodermal progenitor regions. All embryos oriented with dorsal side to the right. MZoep embryos were obtained from a homozygous cross of parental fish generated from an oep ^t₃⁵⁷/ ^t₃⁵⁷ incross followed by cripto rescue. Digital images were captured using the Lumar V12 and AxioCam MRc5 and the images processed using Axiovision 4.5 Software (Zeiss). Cell transplantations were carried out using a Cell Tram Vario (Eppendorf).

• Supplemental Figure S8 -

  Fig. S8. Comparative quantification of DFC number and domain width along lateral margin between Ttrap^MO and control^MO embryos. Embryos were microinjected with either Ttrap MO or scrambled control MO and fixed with 4% PFA at shield, 70-80% epiboly and bud stages. DFCs were visualized via whole-mount in situ hybridization using sox32/casanova as a probe. Cell counts were performed manually using a 200× magnification of the DFC stained region. Widths of DFC sox32/cas expression domains were obtained and recorded by measuring the points between the outermost cells. Photographs and measurements of each embryo were taken using a Deltapix digital camera and the annotations function of the DpxView Pro Image Management Software (Deltapix AS, Denmark; see example of screen capture image below) respectively. Note: embryos are somewhat wider in diameter (approx. 1-1.2 mm) than live equivalent, most likely due to processing during in situ hybridization and the final rehydration process after clearing. Zoom factor in program also results in final width measurement that is 1.75× greater than actual sample. However, this zoom factor was taken into account when calculating final distances.

• Supplemental Figure S9 -

  Fig. S9. Graphical depiction of DFC count in Ttrap^MO vs control^MO embryos at shield and 70-80% epiboly stages. (A) DFC numbers in Ttrap^MO (blue bar) vs control^MO (beige bar) embryos at shield stage. (B) DFC numbers in Ttrap^MO (blue bar) vs Cont^MO (beige bar) embryos at 70-80% epiboly stage. At either gastrulation stage tested, there is no significant difference in DFC count between Ttrap^MO and control^MO embryos as visualized by whole-mount in situ hybridization analysis using sox32/cas as a probe. Statistical analysis was performed using Student’s unpaired t-test. ttmo, Ttrap MO; scmo, scrambled control MO.

• Supplemental Figure S10 -

  Fig. S10. Graphical depiction of DFC domain width in Ttrap^MO vs control^MO embryos throughout gastrulation. (A) DFC numbers in Ttrap^MO (blue bar) vs control^MO (beige bar) embryos at shield stage (B) at 70-80% epiboly stage, and (C) at bud stage. For all gastrulation stages tested, DFC domain widths for Ttrap^MO embryos are significantly wider than for control^MO embryos, as visualized by whole-mount in situ hybridization analysis using sox32/cas as a probe. Statistical analysis was performed using Student’s unpaired t-test. ttmo, Ttrap MO; scmo, scrambled control MO.

• Supplemental Figure S11 -

  Fig. S11. ttrap:snail1a MO double knockdown rescue of heart looping defects. Results of rescue experiment in graphical format. Y-axis represents percentage of embryos displaying phenotype and X-axis represents type of MO treatment (wild-type untreated (left); ttrap MO only (4 ng) (middle); ttrap:snail1a MO (4 ng + 2 ng) (right). Blue denotes percentage of embryos displaying left looping, maroon, right looping, and beige, no looping.