|
Figure 1. The endocytic adaptor protein Eps15R interacts with Smad1. (a) Xenopus Eps15R and deletion constructs. The N-terminal portion contains three Eps15 homology (EH) domains (green, orange and purple), followed by a coiled coil domain (blue) and carboxy-terminal DPF tripeptide repeat domain (black). (b) Eps15R binds to Smad1 in the yeast two-hybrid assay, as detected by β-galactosidase activity. Interaction requires the DPF domain of Eps15R. (c) Eps15R and Smad1 interact in Xenopus embryos. Embryos were injected with Myc-tagged full-length Eps15R mRNA alone or with Flag-tagged Smad1 mRNA. Lysates were immunoprecipitated (IP) with anti-FLAG antibodies and then western blotted with anti-Myc antibodies. Whole embryo lysates were immunoblotted with anti-Myc and anti-FLAG antibodies. (d) GFP-Eps15R is enriched in bright punctate foci located both juxta-membraneously and deeper within the cytoplasm of Xenopus animal cap cells. Lower fluorescence levels are found in the nucleus. (e) Histone B4-RFP and mCherry-GPI expression in cells shown in (d). (f) Limited co-localization of GFP-Eps15R and mCherry-Hrs 1 h after stimulation with 100 ng ml–1 BMP4/7.
|
|
Figure 2. Eps15R enhances Smad1 signalling and is required for transcription in response to BMP signalling. (a,b) Over-expression of a control RNA encoding lactate dehydrogenase (LDH) does not impair development (a), whereas over-expression of Eps15R RNA causes defects in head and anterior mesoderm (b). (c–e) Phenotypes of embryos injected with an antisense MO oligonucleotide targeting Eps15R. (c) Controls. (d) Embryos injected with Eps15R MO exhibit shortened axes and ventrolateral defects. (e) The phenotype in (d) is significantly rescued by co-injection of MT-Eps15R RNA. (f) Morphometric analysis of PAT : AP length ratio in these embryos. Measurements were subjected to ANOVA and Tukey's test for least significant difference. (g) Diagrammatic depiction of the PAT and AP measurements. (h) Expression of BMP targets Xhox3 and Xbra is elevated when Eps15R is co-expressed with Flag-Smad1 in the animal cap assay; caps were harvested at NF11.5. (i) Eps15R MO inhibits transcription of BMP-responsive genes in animal caps treated with 100 ng ml–1 BMP4/7 heterodimers and harvested at NF11.5, yet there is no decrease in activation of these genes in response to 100 ng ml–1 FGF4 (FGF) or 10 ng ml–1 activin (act). Fold induction was calculated relative to control cap levels, and samples were normalized to the expression of ornithine decarboxylase.
|
|
Figure 3. The DPF domain transactivates gene expression and differentially modulates Smad signalling. (a–c) Effects of over-expression of full-length Eps15R, or the DPF domain alone, on Xenopus development. (a) Uninjected embryos. (b) Embryos injected with 3 ng RNA encoding Eps15R. (c) Embryos injected with 3 ng RNA encoding the Eps15R DPF domain. (d) RT-PCR showing globin induction by the DPF domain in NF21 animal caps. (e) The DPF domain mimics the activity of Smad1 in the induction of Xbra and Xhox3 in NF11 gastrula caps but the full-length Eps15R lacks this ability. (f) Both full length Eps15R and the DPF domain can transactivate transcription in the yeast assay. (g) The cytoplasmic localization of GFP-Eps15R-δDPF is disrupted, displaying a diffuse, reticulated cytoplasmic distribution, in contrast to the punctate cytoplasmic localization of the full-length version shown in figure 1d. Nuclear enrichment is retained in the absence of the DPF domain. (h) The DPF domain antagonizes the ability of Smad2 to induce expression of genes such as chordin, goosecoid and Frzb in NF11 gastrula animal caps, which instead activates the ventral marker Xhox3.
|
|
Figure 4. Differential compartmentalization of Eps15R/Smad complexes. Live imaging of Eps15R/Smad complexes monitored by bimolecular fluorescence complementation (BiFC). (a,c,e,g) Greyscale images of BiFC fluorescence. (b,d,f,h) Merged images of BiFC fluorescence in green and CFP-histone H2B (to label nucleus) and CFP-GPI (to label membranes) in red. Total numbers of cells scored for each BiFC complex were: n = 78 (Eps15R/Smad1); n = 27 (Eps15R-δDPF /Smad1); n = 45 (Eps15R/Smad2); n = 24 (Eps15R-δDPF/Smad2). (a,b) Cells injected with VN-Eps15R and VC-Smad1 have nuclear BiFC fluorescence, with enrichment in localized regions. (c,d) VN-Eps15R-δDPF does not interact with VC-Smad1. (e,f) Complexes of VN-Eps15R and VC-Smad2 are distributed in punctate dots throughout the cell. Some complexes are closely associated with the membrane (arrows), consistent with the known association of Eps15R with coated pits. (g,h) VN-Eps15R-δDPF does not interact with VC-Smad2.
|
|
Supplementary Figure 1: Expression of Xenopus Eps15R during embryonic development.
(A) A Northern blot analysis shows that Eps15R is expressed at all stages of early development.
Embryonic stages correspond to blastula (NF6), late gastrula (NF12), early neurula (NF16),
early tadpole (NF23) and tailbud tadpole (NF32). Smad1 expression is unchanged over early
embryogenesis and serves as a loading control. (B, C and D) Whole-mount in situ hybridization
of Eps15R in developing embryos. (B) At blastula stages Eps15R transcripts are present in the
animal pole. The top right embryo was stained with sense probe for background staining. (C) In
the neurula, Eps15R expression is enriched in the neural folds, cement gland and proctodeum.
(D) In the early swimming tadpole (NF36) Eps15R is expressed in head (h), spinal cord (sc) and
kidney (k). The upper embryo in panels C-D is a negative control (sense probe) for background
staining. (E) Close up of the head of the lower tadpole in panel D showing expression in the eye
(e) brain (b) and branchial arches (ba).
|
|
Supplementary Figure 2: Synergism of GFP-Eps15R with Smad1
GFP-Eps15R mRNA replicates the activity of untagged Eps15R mRNA shown in Fig.2(h), synergising with Flag-Smad1 to induce the BMP targets Xbra and Xhox3 in animal caps. 2ng of each RNA were injected and the caps were harvested at NF11.5.
|
|
eps15l1 (epidermal growth factor receptor pathway substrate 15-like 1) gene expression in Xenopus laevis embryo, assayed via in situ hybridization, NF stage 36, lateral view, anterior left, dorsal up.
|
|
Figure 1. The endocytic adaptor protein Eps15R interacts with Smad1. (a) Xenopus Eps15R and deletion constructs. The N-terminal portion contains three Eps15 homology (EH) domains (green, orange and purple), followed by a coiled coil domain (blue) and carboxy-terminal DPF tripeptide repeat domain (black). (b) Eps15R binds to Smad1 in the yeast two-hybrid assay, as detected by β-galactosidase activity. Interaction requires the DPF domain of Eps15R. (c) Eps15R and Smad1 interact in Xenopus embryos. Embryos were injected with Myc-tagged full-length Eps15R mRNA alone or with Flag-tagged Smad1 mRNA. Lysates were immunoprecipitated (IP) with anti-FLAG antibodies and then western blotted with anti-Myc antibodies. Whole embryo lysates were immunoblotted with anti-Myc and anti-FLAG antibodies. (d) GFP-Eps15R is enriched in bright punctate foci located both juxta-membraneously and deeper within the cytoplasm of Xenopus animal cap cells. Lower fluorescence levels are found in the nucleus. (e) Histone B4-RFP and mCherry-GPI expression in cells shown in (d). (f) Limited co-localization of GFP-Eps15R and mCherry-Hrs 1 h after stimulation with 100 ng ml–1 BMP4/7.
|
|
Figure 2. Eps15R enhances Smad1 signalling and is required for transcription in response to BMP signalling. (a,b) Over-expression of a control RNA encoding lactate dehydrogenase (LDH) does not impair development (a), whereas over-expression of Eps15R RNA causes defects in head and anterior mesoderm (b). (c–e) Phenotypes of embryos injected with an antisense MO oligonucleotide targeting Eps15R. (c) Controls. (d) Embryos injected with Eps15R MO exhibit shortened axes and ventrolateral defects. (e) The phenotype in (d) is significantly rescued by co-injection of MT-Eps15R RNA. (f) Morphometric analysis of PAT : AP length ratio in these embryos. Measurements were subjected to ANOVA and Tukey's test for least significant difference. (g) Diagrammatic depiction of the PAT and AP measurements. (h) Expression of BMP targets Xhox3 and Xbra is elevated when Eps15R is co-expressed with Flag-Smad1 in the animal cap assay; caps were harvested at NF11.5. (i) Eps15R MO inhibits transcription of BMP-responsive genes in animal caps treated with 100 ng ml–1 BMP4/7 heterodimers and harvested at NF11.5, yet there is no decrease in activation of these genes in response to 100 ng ml–1 FGF4 (FGF) or 10 ng ml–1 activin (act). Fold induction was calculated relative to control cap levels, and samples were normalized to the expression of ornithine decarboxylase.
|
|
Figure 3. The DPF domain transactivates gene expression and differentially modulates Smad signalling. (a–c) Effects of over-expression of full-length Eps15R, or the DPF domain alone, on Xenopus development. (a) Uninjected embryos. (b) Embryos injected with 3 ng RNA encoding Eps15R. (c) Embryos injected with 3 ng RNA encoding the Eps15R DPF domain. (d) RT-PCR showing globin induction by the DPF domain in NF21 animal caps. (e) The DPF domain mimics the activity of Smad1 in the induction of Xbra and Xhox3 in NF11 gastrula caps but the full-length Eps15R lacks this ability. (f) Both full length Eps15R and the DPF domain can transactivate transcription in the yeast assay. (g) The cytoplasmic localization of GFP-Eps15R-ΔDPF is disrupted, displaying a diffuse, reticulated cytoplasmic distribution, in contrast to the punctate cytoplasmic localization of the full-length version shown in figure 1d. Nuclear enrichment is retained in the absence of the DPF domain. (h) The DPF domain antagonizes the ability of Smad2 to induce expression of genes such as chordin, goosecoid and Frzb in NF11 gastrula animal caps, which instead activates the ventral marker Xhox3.
|
|
Figure 4. Differential compartmentalization of Eps15R/Smad complexes. Live imaging of Eps15R/Smad complexes monitored by bimolecular fluorescence complementation (BiFC). (a,c,e,g) Greyscale images of BiFC fluorescence. (b,d,f,h) Merged images of BiFC fluorescence in green and CFP-histone H2B (to label nucleus) and CFP-GPI (to label membranes) in red. Total numbers of cells scored for each BiFC complex were: n = 78 (Eps15R/Smad1); n = 27 (Eps15R-ΔDPF /Smad1); n = 45 (Eps15R/Smad2); n = 24 (Eps15R-ΔDPF/Smad2). (a,b) Cells injected with VN-Eps15R and VC-Smad1 have nuclear BiFC fluorescence, with enrichment in localized regions. (c,d) VN-Eps15R-ΔDPF does not interact with VC-Smad1. (e,f) Complexes of VN-Eps15R and VC-Smad2 are distributed in punctate dots throughout the cell. Some complexes are closely associated with the membrane (arrows), consistent with the known association of Eps15R with coated pits. (g,h) VN-Eps15R-ΔDPF does not interact with VC-Smad2.
|
