Supplementary Materialscancers-11-01392-s001. importance, POSTN expression and functional phenotypes in the osteogenic front of developing cranial sutures are differentially regulated by specific TW dimer motifs [17]. Together these observations support functionally relevant mechanistic links between TW phosphorylation, dimerization and regulation of expression. However, similar mechanisms have not been established in cancer studies. Hong et al. found no evidence for an association between AN3365 TW S68 phosphorylation and TW:E12 heteromdimerization in a yeast two-hybrid assay [14]. In a prostate carcinoma model malignant phenotypes generated by phospho-mimetic TW were highly correlated with those of a TW:E12 tethered dimer but no direct connection between phosphorylation and dimer formation was shown [21]. In pancreatic cancer TW phosphorylations at S123, T148 and S184 were associated with preferential TW homodimerization and EMT phenotypes, but functional activity of the TW homodimer was not directly studied [23]. Hhex Collectively these observations support the importance of TW phosphorylation dependent TW dimerization but direct validation and functional comparisons of TW dimers with regard to invasion and regulation of expression are lacking. Therefore, here we sought to test the hypothesis that TW mediates mesenchymal changes and expression through site-specific TW phosphorylation dependent regulation of TW dimerization motifs. To test this hypothesis we studied the role of TW S68 phosphorylation in regulating TW dimerization motifs and POSTN expression using hypo-phosphorylation TW mutants and forced TW:TW homodimer or TW:E12 heterodimer constructs in GBM cells. Our results demonstrated a novel mechanism whereby coordinated TW and E12 phosphorylation are required for preferential formation of pro-invasive TW:E12 heterodimers that drive maximal transcriptional activation of expression. This new understanding may provide new targets for intervention that could be leveraged to inhibit the TW-POSTN signaling axis in GBM and other cancers. 2. Results 2.1. TW S68 Phosphorylation Detected in Human GBM and GBM Cells Promotes Invasion To establish the potential relevance of TW S68 phosphorylation for GBM functional phenotypes we first confirmed its presence in patient-derived GBM samples using a S68 phospho-specific TW antibody and normal brain samples (Figure 1A). Higher levels of pTWS68 and total TW are detected in tumors compared to normal brain. However, levels of pTWS68 in tumors do not always correlate with expression levels of the total TW. We then AN3365 confirmed the presence of TW S68 phosphorylation at the endogenous level in glioma cells and patient-derived GBM cell lines to establish its potential relevance for GBM tumor cell specific phenotypes. We performed immunoprecipitation using phospho-TW S68 antibody and detected phosphorylated protein form AN3365 with total TW antibody in T98G cells (Figure 1B). This experiment demonstrated TW phosphorylation at the endogenous levels in glioma cells and verified the specificity of the pTWS68 antibody AN3365 by detecting immunoprecipitated protein with an unrelated TW antibody. Subsequently we detected pTWS68 expression in glioma primary cells (GBM4 and G131) using Western blot (Figure 1C). Open in a separate window Figure 1 Identification and functional characterization of phosphorylation at S68 residue in TW. (A) Detection of TW S68 phosphorylation and total TW in patient GBM samples compared to normal brains. AN3365 (B) TW S68 phosphorylation at the endogenous level in T98G cells detected by immunoprecipitation with pTW S68 antibody followed by detection using a total TW antibody. As a negative control non-specific same isotype IgG was used. Inputs were 2% of total proteins used for immunoprecipitation. (C) Expression of pTW S68 and total TW in primary glioma stem cells GBM4 and G131. (D) Relative fraction of pTW S68 in total amount of TW immunoprecipitated from U87MG (dTW-A) and GBM8 cells with TW overexpression. The percent of pTWS68 is defined as area under the curve of phospho-peptide divided by sum of pTWS68 + Non-phospho-peptide and averaged from three biological replicates. (E) Non-modified and phosphopeptides detected in U87MG (dTW-A) and GBM8 cells with TW overexpression and used for calculation of S68 phosphorylation fraction in panel.