Axin and dishevelled are required for reorientation in a DIX-dependent mechanism. (A) The different Dvl2 constructs used in microinjection experiments. (B) Golgi and centrosome reorientation were measured in cells microinjected with RFP-F control, or various Dvl2 domain-constructs. *, P < 0.01; **, P > 0.05, in t test when compared with control. (C) Axin is efficiently depleted with siRNA, also when combined with Dvl2 and 3 (right panel). aPKC serves as loading control. (D) Golgi and centrosome reorientation were measured in axin-depleted cells and in cells depleted of axin together with Dvl2 and 3. (E) Golgi reorientation was measured in cells microinjected with full-length axin or axinΔDIX. Red bar marks basal level expected from random Golgi/centrosome reorientation (33%). *, P < 0.005 in t test when compared with control.

Wnt signaling is required for the accumulation of APC on microtubules at the leading edge. (A–D) Confocal images of cells treated with control siRNA (A); all three Dvl isoform siRNAs (B); Axin siRNA (C); Wnt5a siRNA (D) and stained for APC (green) and microtubules (red). Higher magnification of areas marked with dashed-line squares are shown in bottom panel (a–d). Arrowheads point to regions of APC accumulation, at the front of control-siRNA cells, at the back of All-Dvl-siRNA cells, or at the perinuclear region of Axin-siRNA or Wnt5a-siRNA cells. (E) APC polarization was quantified as described in Materials and methods.

GSK-3 inhibition is mediated by dishevelled and not by phosphorylation. (A) Golgi reorientation toward the scratch was quantified in GSK-3^WT and GSK-3^SA MEFs. Cells were pretreated for 1 h with indicated inhibitors, or 0.1% DMSO as control: 20 μM SB216763 or 5 μM CHIR99021 (both are GSK-3 inhibitors). *, P < 0.001; **, P < 0.01 in t test when compared with DMSO control. (B) Dvl1, 2, and 3 are efficiently depleted using specific siRNAs, also in combination (depletion >95%), compared with nonspecific control siRNA oligo. aPKC serves as loading control. Arrows indicate position of Dvl protein. (C) Golgi and centrosome reorientation were measured in cells depleted of single Dvl isoforms or their combinations. Red bar marks basal level expected from random Golgi/centrosome reorientation (33%). *, P < 0.01; **, P > 0.05, in t test when compared with control.

Wnt5a is required for reorientation and Dvl phosphorylation. (A) Cells were pretreated for 1 h with the indicated concentrations of recombinant Dkk1 or sFRP1 proteins (these concentrations were based on previous reports [Endo et al., 2005] and were also determined by titration) or with PBS as control. Golgi and centrosome reorientation were measured 4.5 h after wounding. (B) Dvl2 and Dvl3 phosphorylation can be detected as a band-shift in Western blots preformed on 7.5% SDS/polyacrylamide gels, indicated by arrowheads. 5 μg/ml sFRP1 or 1 μg/ml Dkk1 was used. aPKC serves as loading control. (C) Wnt5a protein can be detected in fibroblast cell extracts by Western blot. Protein levels do not change over the time course of migration. Tubulin serves as loading control. (D) Wnt5a protein can be efficiently depleted using two different siRNA oligos. Tubulin serves as loading control. (E) Centrosome and Golgi reorientation were measured in Wnt5a-depleted cells. Red bar marks basal level expected from random Golgi/centrosome reorientation (33%). (F) Dvl2 and Dvl3 phosphorylation was visualized in cells treated with control or two different Wnt5a siRNA oligos. Arrows indicate differently phosphorylated forms of Dvl and aPKC serves as loading control.

Dvl2 and aPKC form a complex after scratching in a Wnt- and Cdc42-dependent manner. (A) Endogenous Dvl was immunoprecipitated using monoclonal antibodies against Dvl2 or Dvl3 from cell extracts 30 min after scratching. Top panel shows coimmunoprecipitated aPKC and immunoprecipitated Dvl2 (Western blotting was done using a rabbit antibody against Dvl2), and middle panel shows the coimmunoprecipitated aPKC and immunoprecipitated Dvl3. aPKC levels in 10% of total input extract were used as loading control (bottom panel). Control immunoprecipitation (right panel) using a nonspecific antibody (mouse anti-Myc) demonstrates the specificity of aPKC immunoprecipitations. (B) Endogenous Dvl2 was immunoprecipitated from extracts of cells that had been pretreated for 1 h with 8 μM Gö6983 (aPKC inhibitor), 5 μg/ml sFRP1 (Wnt inhibitor), or 5 ng/ml Toxin B10463 (Rho GTPase inhibitor) and then scratched and left for 30 min. Top two panels show coimmunoprecipitated aPKC and immunoprecipitated Dvl2. aPKC levels in 10% of total input extract were used as loading control. (C) Quantification of fold increase in Dvl2-aPKC interaction after scratching cell monolayer is shown. aPKC levels were normalized to Dvl2 levels (i.e., the sum of the two migrating species). Average of three independent experiments is shown. Red bar marks basal (un-induced) levels.