Figure 1

TIAF1 physically binds Smad4 and blocks Smad4-regulated promoter activation. (a) By yeast two-hybrid library screening,8, 12, 13 two Smad4 clones (No. 7 and 17), with an identical DNA sequence, were isolated as a TIAF1-binding protein. Both Smad4/TIAF1 interactions and positive control bindings for WOX1/p53 and MafB self-interaction are shown, as evidenced by the growth of yeast at 37 °C using a selective galactose-containing agarose plate (six representative colonies). Empty pSos/pMyr vectors and p53/TIAF1 plasmid constructs8, 12, 13 were regarded as negative controls. (b) COS7 cells were cultured in 10% FBS and treated with or without TGF-β1 (5 ng/ml) for 30 min, followed by isolating cytosolic and nuclear fractions for co-immunoprecipitation with anti-TIAF1 IgG. TGF-β1 increased the binding of TIAF1 with Smad4. Non-immune IgG was used as a negative control in co-immunoprecipitation. (c) COS7 cells were infected with retroviral TIAF1si or empty retrovirus and cultured for 48 h, followed by determining protein nuclear localization for Smad3 and 4 by immunofluorescence microscopy. TIAF1si knocked down endogenous TIAF1 protein by ∼70%. (d) COS7 cells were transfected with EGFP-Smad4 plasmid (using liposome) and infected with retroviral TIAF1si or empty retrovirus. The cells were grown for 48 h, fixed and immunostained with TIAF1. Nuclei were stained with DAPI. Knockdown of TIAF1 significantly induced nuclear accumulation of EGFP-Smad4. (e and f) Human monocytic U937 cells and mouse L929 fibroblasts were infected with TIAF1si or empty retrovirus and cultured for 48 h, followed by isolating nuclear and cytosolic fractions. Spontaneous accumulation of Smad3 and 4 in the nuclei was shown in cells infected with TIAF1si. Reduction of endogenous TIAF1 protein by siRNA is shown (60–75% reduction; a representative set of data from two experiments). (g) COS7 cells were transfected with ECFP or ECFP-TIAF1 expression constructs, grown for 24 h, and treated with TGF-β1 (5 ng/ml) for 1–8 h. The extent of accumulation of Smad2, 3 or 4 in the nuclei post TGF-β1 treatment for 4 h is shown (n=3, ∼100 cells counted per experiment), as determined by immunofluorescence microscopy. (h) COS7 cells were transfected with a SMAD-responsive element DNA construct (using GFP as a reporter), in the presence or absence of Smad4 and/or TIAF1 expression constructs. After 48 h, positive promoter activation in cells with green fluorescence was counted. Compared with Smad4 alone, TIAF1 significantly blocked the ectopic Smad4-induced activation of the promoter (*P<0.001 for the last two bars at the right, ∼200 cells counted in each experiment with three repeats; mean±S.D., Student's t test). SP, promoter driven by SMAD. (i) COS7 cells were transfected with the SMAD promoter reporter construct and simultaneously infected with the TIAF1si or empty retrovirus. Knockdown of TIAF1 resulted in spontaneous activation of the SMAD promoter (∼1.5-fold increase). Ectopic Smad4-regulated promoter activation was increased by ∼2-fold in the TIAF1-knockdown cells