Syntenin is recruited by Fz 7 and stimulates c-jun phosphorylation. Confocal micrographs of MCF-7 cells overexpressing eGFP-syntenin alone (A), overexpressing eGFP-syntenin together with Fz 7 (B), or overexpressing eGFP-syntenin together with an Fz 7 carrying a mutated C-terminal PDZBM (T/A) (C). (D) Results are expressed as the mean percentage of cells where the syntenin fluorescence was concentrated at the plasma membrane; bars represent standard deviations. Note that the recruitment of eGFP-syntenin to the plasma membrane relies on Fz 7 and on the integrity of its C-terminal PDZBM. (E) Cell lysates originating from HEK293T cells transfected as indicated on top were tested for c-jun expression and c-jun phosphorylation (ser-63). Actin was used as a loading control. Note that syntenin addition stimulates c-jun phosphorylation in a concentration dependent manner.

Xsyntenin expressions during X. laevis early development and Xsyntenin-a corecruitment with XDsh, by XFz 7, to the plasma membrane of animal caps. (A) RNA extracts from embryos at different stages were used to analyze Xsyntenin-a expression by RT-PCR. Histone H4 amplification was used as an internal control. (B–I) Xsyntenin-a, -a', and b and XFz 3, 7, and 8 mRNA distributions (purple) at different stages (st) of development. Hd, head; nt, neural tube; pn, pronephros; ba; branchial arches; cg, cement gland. (J) Syntenin-Fz 7 interaction is conserved in Xenopus. Hu Fz7 and XFz7 cytoplasmic domain sequences are shown on the right. The sequences vary for two amino acids (underlined). Fz 7 cytoplasmic domain from Xenopus or human, mutated (T/A) or not mutated in the PDZBM, was overlayed with Xsyntenin-a (top). The quality and concentration of the fusion proteins were controlled in Coomassie (bottom). (K–U) Confocal micrographs of Xenopus animal caps at stage 9 showing the subcellular distribution of XDsh-myc (K–M and Q), Xkermit-myc (S and T), Xsyntenin-a-myc (N–P) or Xsyntenin-a-HA (R and U) as indicated at the bottom. The caps originate from embryos injected at two-cell stage with different combinations of mRNAs encoding proteins indicated on each micrograph. Note the translocation of XDsh to the plasma membrane upon XFz 7 and XFz 7 T/A expression (compare M and L with K) and the translocation of Xsyntenin-a upon XFz 7 but not XFz7 T/A expression (compare O and P with N). Note also that although Xsyntenin-a impairs Xkermit translocation (T), XDsh translocation is maintained (Q).

Xsyntenin supports the XFz 7/ XPKCα/XCDC42 branch of the noncanonical Wnt signaling pathway. (A, C, and E) Pictures of stage 19 animal cap explants originating from embryos injected at two-cell stage with different components as indicated on top, dissected at stage 9, and cultured in the presence of 50 ng/ml activin. Note that the block of elongation observed with MOs for XFz 7 (A, right image) is rescued by the coinjection of a low dose of Xsyntenin-a RNA (C, middle picture). Note that the block of elongation induced by a high dose of Xsyntenin-a RNA is rescued by the coinjection of RNA encoding a DN form of the small GTPase XCDC42 (compare E, middle, with Figure 5F). (B, D, and F) Results are expressed as the mean percentage of elongated caps per condition; bars represent standard deviations. (G and H) Confocal micrographs of Xenopus animal caps at stage 9 showing the subcellular distribution of XPKCα-GFP as indicated at the bottom. (G) Comparison of the XPKCα-GFP distribution upon the overexpression of XFz 7 alone (left), together with Xsyntenin-a-HA (middle), or together with X-syntenin MOs (right). (H) Comparison of the XPKCα-GFP distribution upon the overexpression of Xsyntenin-a-HA (top) or the down-regulation of X-syntenins (bottom). Note that Xsyntenin enhances the plasma membrane recruitment of XPKCα-GFP.

Syntenin interacts with Fz in a PDZ-dependent mode. (A) Sequences of the last 25 cytosolic amino acids of all human Fz and Fz 7 mutants; overview of syntenin binding to these peptides as detected in overlay; nd, not determined. C-terminal PDZBMs are in bold. For Fz 1, 2, and 7, the membrane proximal PDZBM for Dsh is also present in the 25 last amino acids, and it is indicated in gray. (B) Overlays illustrating syntenin interaction with Fz. GST-Syndecan-2 cytoplasmic domain (WT) was used as a positive control, and GST or GST-Syndecan-2 ΔPDZBM (cytoplasmic domain deleted for the last 2 amino acids) were used as negative controls. Note the interaction of syntenin with Fz 7, 3, and 8 last 25 amino acids, and the lack of interaction with Fz 7 T/A and Fz 7 T/A V/A mutants. The quality and concentration of the fusion proteins were controlled in Coomassie as shown at the bottom. (C) Respective roles of the PDZ domains of syntenin in Fz interaction. Structure of syntenin and coordinates of the amino acids that define the different domains are shown on the right. The relevant GST fusions were overlayed with recombinant proteins containing different combinations of the PDZ domains of syntenin as indicated. Note that Fz 7 interacts preferentially with the PDZ1 domain, whereas syndecan 2, Fz 3, and 8 interact preferentially with the PDZ2 domain of syntenin. The quality and concentration of the fusion proteins were controlled in Coomassie as shown at the bottom. (D) Interaction of syntenin and syntenin PDZ domains with Fz 7 cytoplasmic domain in surface plasmon resonance. RU, response units. Note that the binding relies primarily on the PDZ1 domain of syntenin. (E) Coimmunoprecipitation of endogenous syntenin with eYFP-tagged Fz 7 cytoplasmic domain. MCF-7 cells overexpressing the Fz 7 cytoplasmic domain fused N-terminally to eYFP were extracted with detergent. The cell lysate was immunoprecipitated with anti-eYFP antibodies and protein G beads before immunoblotting (right lanes, IP Fz7) with anti-eYFP to detect the Fz 7 fusion (top) or anti-syntenin antibodies (bottom). The cell lysate was used as positive control (left), in the negative control the anti-eYFP antibodies were omitted (middle). Note that endogenous syntenin coimmunoprecipitates with the eYFP-Fz 7 fusion (right bottom lane).

(A) Sequence alignment of the human syntenin (Husyntenin) with the three orthologues found in X. laevis (Xsyntenin-a, -a', and -b). (B) Percentage of identity between the different domains of the various syntenins. Note the extensive conservation of the PDZ domains.

Xsyntenin overexpression and down-regulation block the elongation of activin-treated animal caps and of whole embryos. (A–I) Pictures of stage 19 animal cap explants, originating from embryos injected at two cell stage with different mRNAs as indicated on top, dissected at stage 9, and cultured in the absence or in the presence of 50 ng/ml activin (as indicated at the bottom). In situ hybridization with the mesodermal marker cardiac actin (B, D, G, and I) shows that the mesodermal differentiation is sustained in all activin-treated caps. Note the specific block of elongation induced by high dose of X-syntenin-a mRNA (F–G). (J) Results are expressed as the mean percentage of elongated caps per condition, bars represent standard deviations. (K–Q) Illustration of selected anomalies in X. laevis embryos depleted for Xsyntenin-a, a', and b. Embryos injected with Xsyntenin MOs show a delayed gastrulation (compare L with K), and they have a shorter body-axis (compare N–O with M). Injections were performed at the four-cell stage with 7, 5, or 10 ng of each Xsyntenin MO or with 22 or 30 ng mismatch MO in each blastomere. Embryos with body-axis length <60% (O) and embryos with body-axis between 60 and 80% (N) of the average length of noninjected controls were pooled for quantitative analysis (Q). The percentage of embryos showing delayed gastrulation (P) and shortened body-axis (Q) was scored in three independent experiments, with at least 80 embryos. Results are expressed as the mean percentage of embryos showing the defects at the stages indicated; bars represent standard deviations.

Model for the role of syntenin scaffolding in noncanonical Wnt signaling. Syntenin interacts with Fz 7 via its PDZ1 domain and with Syndecan-4 (A) or Syndecan-2 (B) via its PDZ2 domain. Syndecan and/or Fz 7 can recruit syntenin to the plasma membrane. (A) Syndecan-4 interacts directly with PKCα. Syndecan-4 triggers the activation of PKCα. PKCα functions upstream of CDC42 in noncanonical Wnt signaling. (B) Syndecan-2 might also activate CDC42. See Discussion for details and references.