Figure 6. TGF-β1 is required for NRP2-mediated EMT.

A, Western-blotting experiments were performed on nuclear extracts of HT29^NRP2 and Colo320 for the detection of nuclear smad proteins after treatment with increased doses of TGFRI kinase-inhibitor SB-431542 (10 µM, 50 µM). DMSO is the control medium. Histone-H1 was used as a control of protein loading. Whereas only 10 µM of SB-431542 decreases the nuclear level of phosphorylated Smad2 in Colo320 cancer cells, doses of 50 µM should be used to decrease the nuclear level of phosphorylated Smad2 in HT29^NRP2 cancer cells. Smad proteins have been quantified with the BIO-1D advanced software and reported to the Histone H1 level. B, Overnight treatment with increased doses of TGFRI kinase-inhibitor SB431542 (10 µM, 50 µM, 100 µM) inhibited dose-dependently vimentin expression in Colo320 and in HT29^NRP2 cells. TGF-β1 signaling inhibition restored in part E-cadherin in HT29^NRP2 cells. Proteins have been quantified with the BIO-1D advanced software and reported to the β-actin level. Data represent results of a representative experiment out of 3. C, Western-blotting experiment realized with Colo320 and HT29^NRP2 whole cell lysates overnight treated with increased doses of TGFRI kinase-inhibitor SB431542 (10 µM, 50 µM, 100 µM). SB431542 treatment inhibited dose-dependently snail1 and twist1 expression in Colo320 and in HT29^NRP2 cells. Proteins have been quantified with the BIO-1D advanced software and reported to the β-actin level. D, Confocal microscopy analysis of localization of β-catenin in HT29^NRP2 cells after overnight treatment with increased doses of TGFRI kinase-inhibitor SB431542 (10 µM, 50 µM). Cells (HT29^NRP2) were stained with DAPI (blue) for nuclear staining, with mitotracker (purple) for cytotoxicity dectection and also with an anti-β-catenin antibody (green). Mitotracker staining (purple) was realized in order to assess potential cytotoxicity of the SB-431542 treatment. Loss of staining of this marker correlates with increased cytotoxicity. 16 hours treatment with 50 µM of SB-431542 induces relocalization of beta-catenin at the membrane of the cells, indicating that TGFRI activity neutralization by SB-431542 can reverse EMT in HT29^NRP2 colorectal cancer cells. No cytotoxicity was observed at any dose of SB-431542 used. For fluorescence quantification, ratio of fluorescence intensity was calculated in each condition. A decreased ratio of nuclear fluorescence/cytoplasm-membrane fluorescence intensity after treatment with 50 µM of SB-431542 in HT29^NRP2 cells was observed indicating that beta-catenin relocalizes at the membrane of the cells.