Upon stimulation with Wnt3A for 6 days, hESC colonies exhibit overt phenotypic changes that include loss of compact colony structure (top panels, Fig. 2B), decreased co-expression of cell surface markers of undifferentiated hESCs (GCTM2 and CD9, Fig. 2C) and decreased steady-state RNA abundance of pluripotency genes (NANOG and POU5F1, Fig. 2D). Additionally, treatment of hESCs with WNT3A leads to increased expression of genes associated with endoderm (SOX17 and GATA6, Fig. 2E) and mesoderm (T and KDR, Fig. 2F) differentiation. We found that in hESCs treated with both Wnt3A and WIKI4, the WNT3A-dependent effects that we typically observe on colony morphology (bottom panels, Fig. 2B), expression of cell surface markers (Fig. 2C) and expression of markers of pluripotency and differentiation (Fig. 2D, 2E, 2F) were eliminated. We conclude that WIKI4 inhibits Wnt/?catenin-mediated processes in hESCs, as well as in DLD1 cells, suggesting that WIKI4 acts on a conserved component of the Wnt/?catenin signaling pathway.levels similar to those seen with treatment with the TNKS inhibitor XAV-939. To further investigate how WIKI4 regulates AXIN protein abundance, we queried whether WIKI4 treatment promotes expression of AXIN mRNA or whether it prevents the degradation of AXIN by the proteasome. Using quantitative PCR (qPCR) analyses of DLD1 colorectal carcinoma cells, we found that steady state levels of AXIN1 (Fig. 3C) and AXIN2 (Fig. 1D) transcripts were not increased upon treatment with WIKI4. To test whether WIKI4 inhibits AXIN protein turnover, we treated DLD1 cells overnight with WIKI4, and then released them from treatment the next day for two hours (wash-off).
We found that cells continuously treated with WIKI4 during the wash-off period exhibited increased abundance of AXIN1 and AXIN2 relative to cells treated with DMSO (Fig. 3D, compare lanes two and three), suggesting that WIKI4 prevents turnover of the AXIN proteins. When DLD1 cells were treated with the proteasome inhibitor MG132 during the wash-off period, AXIN1 and AXIN2 protein abundance remained elevated (Fig. 3D, compare lanes 3 and 4). Taken together, the qPCR and wash-off experiments suggest that WIKI4 increases the steady-state abundance of AXIN proteins by preventing their degradation by the proteasome.
WIKI4 Blocks the Activity of TNKS2 and Prevents AXIN Ubiquitylation
AXIN1 is modified sequentially by two enzymes in order for it to be recognized by the proteasome for degradation. First, AXIN1 is ADP-ribosylated by the TNKS1 and TNKS2 enzymes [33]. Subsequently, ADP-ribosylated AXIN1 is bound by the E3 ubiquitin ligase RNF146, which specifically catalyzes its ubiquitylation (Fig. 4A, [48,49]). To test whether WIKI4 prevents ubiquitylation of AXIN protein, we treated SW480 (Fig. 4B) and DLD1 (Figure S3A, S3B) colorectal carcinoma cells overnight with WIKI4, and subsequently incubated for two hours with either MG132 alone or MG132 and WIKI4 (wash-off). We found that inhibition of the proteasome during the wash-off period with MG132 led to an increase in the abundance proteins bound to ubiquitin (Fig. 4B, Fig. S3A, S3B left panels). We further observed that WIKI4 treatment during the wash-off period reduced the detection of ubiquitin in immunoprecipitated AXIN2 (Fig. 4B, Figure S3A) and AXIN1 (Figure S3B), suggesting that WIKI4 indeed inhibits AXIN ubiquitylation. One possible explanation for WIKI4-dependent inhibition of AXIN ubiquitylation is that WIKI4 directly inhibits TNKSmediated ADP-ribosylation of AXIN. The ADP-ribosylation activity of TNKS proteins can be assayed in vitro by quantifying their ability to catalyze auto ADP-ribosylation. To investigate the hypothesis that WIKI4 inhibits the catalytic activity of TNKS proteins, we performed in vitro auto-ADP-ribosylation assays using recombinant TNKS2. Similar to what is observed for the known TNKS inhibitor XAV-939, we found that WIKI4 prevents autoADP-ribosylation of TNKS2 at an IC50 of ,15 nM (Fig. 4C). In contrast to the effects of XAV-939 and WIKI4, a second ATP analog, U0126, failed to inhibit auto-ADP-ribosylation of TNKS2, demonstrating that our assay is specific (Figure 4C). Taken WIKI4 Increases Steady-state Abundance of AXIN1
After stimulation of A375 melanoma cells with Wnt3A, we observed that the steady-state abundance of the scaffold protein AXIN1 is reduced (left time course, Fig. 3A) and conversely, abundance of cytosolic CTNNB1 increases (left time course, Fig. 3A). Additionally, we observed that the abundance of CTNNB1 that is phosphorylated at sites that are regulated by the destruction complex components CSNK1A1 (S45, left time course Fig. 3A) and GSK3B (S33, left time course, Fig. 3A) is decreased following Wnt3A stimulation. We next investigated whether WIKI4 regulates the biochemical changes associated with Wnt/?catenin signaling. We found that WIKI4 inhibits WNT3Adependent increases in the steady-state abundance of cytosolic CTNNB1, inhibits Wnt3A-dependent decreases in steady-state abundance of AXIN1, and inhibits Wnt3A-dependent decreases in abundance of phosphorylated of ?catenin (S33 and S45) (right time course, Fig. 3A). Taken together, our findings indicate that WIKI4 modulates Wnt-dependent changes in the abundance and phosphorylation of known core components of the Wnt/?catenin signaling pathway. We next examined whether WIKI4 alters steady-state abundance of AXIN1 and the related AXIN2 in another cell type. Increases in the steady-state abundance of the AXIN scaffolding proteins have been shown to correlate with decreases in the steadystate abundance of cytosolic CTNNB1, even in APC-mutant colon cancer cells [33,34]. To test the effects of WIKI4 on AXIN levels in APC-mutant cells, DLD1 colorectal cancer cells were treated with WIKI4 for two, four, six or 24 hours and processed for western blotting. Figure 3. WIKI4 increases the steady-state abundance of the Wnt/?catenin inhibitory protein, AXIN1. (A) WIKI4 prevents degradation of AXIN1 following stimulation with Wnt3A. A375 melanoma cells were stimulated with 10% (vol/vol) Wnt3A CM for the indicated time periods with or without WIKI4 treatment, lysed and analyzed by western blot using the indicated antibodies. (B) WIKI4 increases the steady-state abundance of AXIN1 and AXIN2 protein. DLD1 colorectal carcinoma cells were incubated with DMSO, WIKI4 or XAV-939 for the indicated times, lysed and analyzed by western blot. (C) WIKI4 does not significantly affect the steady-state RNA abundance of AXIN1. DLD1 colorectal carcinoma cells were incubated with WIK4 for the indicated times, and processed for qPCR to assess changes in the steady-state abundance of AXIN1 transcript. This data is representative of two independent experiments and the error bars represent standard deviation. (D) WIKI4-dependent increases in AXIN1 protein abundance can be maintained by treatment with a proteasome inhibitor. DLD1 colorectal carcinoma cells were treated overnight with WIKI4, and after washing were then incubated for two hours with DMSO (D), WIKI4 (W), or the proteasome inhibitor MG132 (M). The cells were lysed and analyzed by western blotting for the indicated antibodies. together, our data suggest that WIKI4 inhibits Wnt/?catenin signaling by inhibiting tankyrase activity, and thus preventing the ubiquitylation and degradation of AXIN proteins.
