Curcumin retards activation of β-catenin/Slug pathway in breast cancer stem cells, thereby restoring E-cadherin. (A) β-catenin-associated E-cadherin was assayed by co-immunoprecipitation from cell lysates of MCF-7-derived 2° mammospheres with or without curcumin treatment using specific antibodies (left panel) or with normal human immunoglobulin G (IgG) as a negative control (right panel). To ensure comparable protein loading, 20% of supernatant from immunoprecipitation (IP) sample was subjected to determination of α-actin by Western blotting. (B) Western blotting was conducted to study the levels of total β-catenin and nuclear β-catenin in 2° mammospheres in presence or absence of curcumin exposure. (C) The relative nuclear expression of β-catenin in 2° mammospheres with or without curcumin treatment was visualized by immunofluorescence. (D, E) Under similar conditions, Western blotting was performed to study the expression levels of β-catenin target genes Cyclin-D1, c-Myc, Slug, Snail, Vimentin, MMP-2 and MMP-9 in curcumin treated/untreated 2° mammospheres. (F) Protein expression of E-cadherin in 2° mammospheres with or without curcumin or Slug-cDNA (or both) were determined by Western blotting (left panel). The efficiency of transfection was determined by Western blot analysis (right panel). (G) Graphical quantifications of cell adhesion, spreading, three-dimensional (3D) invasion, and migration assays of MCF-7-derived 2° mammospheres with or without curcumin and Slug-cDNA treatment/transfection. (H) Transwell migration assay was performed under similar experimental conditions in T47D-derived 2° mammospheres. α-Actin/histone H1/glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as an internal loading control. Data are presented as mean ± standard error of mean or representative of three independent experiments. Cont, control; Cur, curcumin.