Smad1 knockdown correlates with activation of SMAD2, a TGF-β downstream effector. (A) miSmad1 EBs were left untreated or induced with 1 μg/mL doxycycline to deplete Smad1 on day 2 or day 4, and EBs were harvested and fractionated to generate nuclear extracts. Nuclear fractions were subjected to Western blotting to quantify SMAD2 and SMAD5 levels, and HDAC-1 was used as a loading control for total nuclear protein. (B) Densitometric analysis of relative nuclear SMAD2 and SMAD5 protein levels, with induced samples normalized to uninduced controls. (C) Whole-cell lysates were generated from day 2– or day 4–induced miSmad1 EBs and subjected to immunoprecipitation using a SMAD2-specific Ab. Immunoprecipitates were analyzed by Western blotting for SMAD2 and SMAD4. (D) Densitometric quantification of SMAD2/SMAD4 interaction, with day 2– and day 4–induced samples normalized to their respective uninduced controls. (E) An ES-cell line was engineered to allow doxycycline-induced expression of Smad2 (see supplemental Figure 1). Smad2 expression was induced on day 4 and on day 5.75 the EBs were analyzed for EryP colony-forming potential, as in Figure 2. For each sample, n = 3, and the graph is a representative example taken from a pool of at least 4 experiments. Error bars indicate SEM; *P < .01 compared with uninduced controls.