DEV020842 Supplementary Material

Files in this Data Supplement:

• Supplemental Figure S1 -

  Fig. S1. In situ hybridisation of endogenous Xenopus Bα and Bδ mRNA. Embryos were collected at stage 8, stage 10 and tailbud stage, and hybridised with an antisense probe specific for Xenopus Bα mRNA or Xenopus Bδ mRNA. Whole embryos and sagittal sections are shown. At Stage 10, the sections were through the blastopore lip. The position of the dorsal blastopore lip is indicated by asterisks. At Stage 8 and 10, both B subunits are predominantly expressed in the animal cap and more weakly in the marginal zone. Bδ mRNA is particularly well expressed in the ectodermal inner layer of stage 10 embryos (arrow). At tailbud stages, Bα is strongly expressed in the branchial arches, whereas Bδ is much more weakly expressed.

• Supplemental Figure S2 -

  Fig. S2. Bα and Bδ mRNA expression specifically rescues the phenotype induced by the cognate morpholinos. One-cell embryos were injected with control morpholinos (MoC), morpholinos against Bα (MoBα) or morpholinos against Bδ (MoBδ), either alone (-) or in combination with Bα mRNA (Bα) or Bδ mRNA (Bδ). Embryos were fixed when control embryos had reached stage 10.25. The blastopore lip is absent in MoBα-injected embryos (d). This phenotype can be rescued by the co-expression of Bα (e), but not Bδ mRNA (f). MoBδ-injected embryos exhibit a slightly more advanced blastopore lip (compare g with a), which is restored to its expected size by co-expression of Bδ mRNA (i), but not Bα mRNA (h).

• Supplemental Figure S3 -

  Fig. S3. Suppression of tws overexpression by smox or baboA. Males with A9-Gal4 driven overexpression survived poorly, so experiments were performed at several temperatures. At higher temperature (25°C), expression levels from the Gal4-UAS system are higher, so that phenotypes are stronger than at lower temperature (22°C) (Brand and Perrimon, 1993). Phenotypes of wings from experimental genotypes were compared with the phenotype of wings from A9-Gal4/+; UAS-tws23/+ and grouped into classes. At both 25°C (n=14) and 22°C (n=16), the majority of wings from A9-Gal4/+; UAS-tws23/+; UAS-smox8D3/+ were wild type in phenotype. Fewer A9-Gal4/+; UAS-tws23/+; UAS-baboA/+ males survived (n=8), and most wings had an intermediate phenotype. Phenotype classes: complete suppression, wild-type shape and venation; strong suppression, slight loss of tissue at posterior margin (Bajpai et al., 2004) or small changes in vein thickness; moderate suppression, normal wing size with longitudinal fold down middle or decreased wing size with thick or split veins; weak or no suppression, small wings with no apparent veins. For graphical depiction, control genotypes were set arbitrarily at 16 wings.

• Supplemental Figure S4 -

  Fig. S4. Overexpression of Bδ inhibits nuclear accumulation of Smad2 in animal cap explants. One-cell embryos were injected with EGFP-Smad2 mRNA and mRNA expressing HA-tagged constitutively active ALK4 (HA-ALK4^ca) with or without Bδ mRNA as indicated. At stage 8, animal caps were dissected and visualised by confocal microscopy. Two separate caps are shown for each condition.

• Supplemental Figure S5 -

  Fig. S5. Overexpression of Bδ in animal caps blocks Smad2 phosphorylation. Embryos were injected at the one-cell stage with EGFP-Smad2 and/or Flag-tagged mouse Bδ mRNA as indicated and cultivated until control embryos reached stage 8. Animal caps were dissected, treated with Activin as indicated and processed for western blot analysis.

• Supplemental Figure S6 -

  Fig. S6. (A) Specificity of siRNA-induced knockdown of Bα and Bδ in HeLa EGFPSmad2 cells as assayed by RT-PCR. HeLa EGFPSmad2 cells were transfected with different siRNAs against Bα or Bδ, RNA was extracted and subjected to RT-PCR analysis. Levels of Bα and Bδ mRNA are given relative to a GAPDH control. (B) Specificity of siRNA-induced knockdown of Bα and Bδ in HeLa EGFPSmad2 cells assayed by western blotting. HeLa EGFPSmad2 cells were transfected with siRNA SMARTpools and individual siRNAs against Bα or Bδ, induced with TGFβ for the times indicated and analysed by western blotting with antibodies against phosphorylated Smad2 (pSmad2), pan B subunits and Smad2/3 as a loading control. (C) Specificity of siRNA-induced knockdown of Bα and Bδ in the HeLa EGFPSmad2 cell line as assayed by EGFPSmad2 localisation. HeLa EGFPSmad2 cells were transfected with siRNA SMARTpools and individual siRNAs against Bα or Bδ as indicated, induced with TGFβ for the times indicated and fixed for confocal microscopy. (D) Specificity of the pan B antibody. Flag-tagged mouse Bδ, Xenopus Bδ (XL-Bδ) and mouse Bα were transcribed and translated in vitro and analysed by immunoblotting with anti-pan B and anti-Flag antibodies. The anti-pan B antibody recognises mouse Bα and Bδ equally well, but fails to detect Xenopus Bδ. As endogenous Bα and Bδ co-migrate on SDS-PAGE and Bα is much more abundant than Bδ (Strack et al., 1999), we could easily monitor knockdown of Bα using this antibody, but could not detect knockdown of Bδ directly on the protein level. Bδ knockdown was thus monitored at the RNA level (A).

• Supplemental Figure S7 -

  Fig. S7. TGF-β-dependent growth arrest is inhibited by knockdown of Bα. HaCaT cells were treated with siRNAs as indicated. After 48 hours, cells were serum starved and after 24 hours of starvation shifted into 10% serum in the absence or the presence of different amounts of TGFβ (100 pg/ml and 400 pg/ml). After a further 24 hours, the percentage of cells in G1 was determined by FACS analysis and normalised to the percentage of cells in G1 under control conditions (no TGFβ).

• Supplemental Figure S8 -

  Fig. S8. Clustering of ALK4 receptor is dependent on ALK4 kinase activity. Embryos were injected at the one-cell stage with HA-ALK4 and EGFP-Smad2 mRNA, and cultured until control embryos reached Stage 8. Animal caps were dissected and either left untreated or induced with Activin for 1 hour in the absence or presence of 10 µM SB-431542 to inhibit ALK4 kinase activity. EGFP-Smad2 was visualised directly, and HA-ALK4 was detected by immunostaining using anti-HA antibodies. Clustering of HA-ALK4 upon Activin induction (arrows) is inhibited by treatment with SB-431542.

• Supplemental Figure S9 -

  Fig. S9. (A) ALK5 mRNA levels are not affected by knockdown of either Bα or Bδ. HeLa EGFP-Smad2 cells were transfected with siRNAs as indicated. C, control siRNA. Total RNA was prepared 72 hours later and reverse transcription with (RT) or without (no RT) reverse transcriptase was performed. cDNA fragments were then amplified. Grb-2 was used as a control. (B) Protein levels of exogenous HA-tagged ALK5 are strongly reduced in Bα knockdown cells. HeLa EGFP-Smad2 cells were transfected with Bα- or control (C) siRNA. After 48 hours, cells were transfected with a constitutively active version of ALK5 (HA-ALK5^ca), and at the same time treated with the ALK5 receptor inhibitor SB-431542. After a further 24 hours, SB-431542 was washed out to reactivate the receptor and to induce Smad2 phosphorylation. Protein extracts were immunoblotted using antibodies against HA, phospho-Smad2, pan-B subunits and Smad2 (as a loading control). Protein levels of HA-ALK5^ca are strongly reduced in Bα knockdown conditions. Correspondingly, TGF-β-independent Smad2 phosphorylation by HA-ALK5^ca is attenuated under Bα knockdown conditions. A similar downregulation of protein levels was seen for wild-type HA-ALK5 (data not shown). (C) Downregulation of HA-ALK5^ca in Bα-knockdown cells is not due to impaired transfection efficiency or lower transcription of transgenes. β-Galactosidase activity from a co-transfected lacZ-plasmid is unaffected by Bα knockdown.

• Supplemental Figure S10 -

  Fig. S10. Bα interacts with ALK4 in Xenopus embryos. HA-ALK4 was immunoprecipitated from stage 9 embryos injected with HA-ALK4 and Bα or Bδ as indicated. Immunoprecipitates (IP) and inputs were immunoblotted with the indicated antibodies. An immunoprecipitation control using beads alone is also shown.

• Supplemental Figure S11 -

  Fig. S11. ALK5 degradation triggered by knockdown of Bα is partially rescued by Bafilomycin A1 treatment, but does not occur via a Dapper2-dependent mechanism. (A) HeLa TK- cells were transfected with control siRNA or an siRNA against Bα. After 48 hours, they were transfected with a plasmid expressing HA-ALK5. After a further 24 hours, cells were treated (or not) with 10 nM bafilomycin A1 or 25 µM MG132 for 6 hours. Whole cell extracts were analysed by immunoblotting using antibodies against HA, pan B subunits and Smad2. (B) HaCaT cells were transfected with control siRNA or an siRNA against Bα. After 72 hours, cells were treated or not with 10 nM bafilomycin A1, 25 µM MG132 or 25 µM lactacystin for 6 hours. Whole-cell extracts were analysed by immunoblotting using antibodies against ALK5, pan B subunits and Smad2/3. (C) Knockdown of Dapper2 in HaCaT cells increases protein levels of endogenous ALK5 (compare lanes 4-6 with lanes 1-3 in the top panel), however, fails to prevent the ALK5 degradation triggered by Bα knockdown (lane 5). Knockdown and loading controls are shown (lower two panels).