A. Cells were treated with verapamil (50 µM), nicotine (100 µM), nicotine plus siRNA targeted galectin-3 or nicotine plus negative control siRNA, and stained with DyeCycle Violet and SP was measured using Dye cycle violet dye (DCV) on a flow cytometer. The SP gated populations are displayed as scatter plots for red (650 nm LP) vs. blue (450/40 nm BP) emissions of DCV (excited with 405 nm laser source). The ratio of the SP cells to total population of MCF-7 cells is indicated as percentage on each scatter plot (as represented by pink square). B. Bar diagram showing results of side population cells in untreated and treated MCF-7 cells as described in Fig. 5A. Data were expressed as mean ± S.D, where *** and ** represent p≤0.001 and p≤0.01, respectively. C. Gene expression microarray showing (i) Heat map and (ii) The Scatter Plot. (i). The heat map is depicting the overall impression of the change in gene regulation of PCR array specific genes. The color code specifies the magnitude of change of gene expression (Green to Red for upregulation). In the heat map, left panel is showing the basal gene expression of the untreated MCF-7 cells and right panel represents nicotine treated group. (ii). MCF-7 cells were treated with nicotine (100 µM, 48 h) and expression of all the genes (genes included in PCR array 96 well plate) was measured in comparison to the untreated control. The scatter plot shows the expression level (2 ^ (−ΔΔCt)) of each gene in the nicotine treated cells compared to the untreated control. The black line indicates fold changes (2 ^ (−ΔΔCt)) of 1. Twist-1 was indicated as most up-regulated gene. The expression fold of each gene was shown in Table 2. D. Validation of TWIST1 expression in nicotine treated cells by Western blot. The number on top of blot represents arbitrary unit showing quantification of the Twist expression as measured by Image J software. E. Expression of TWIST1. Representative diagrams showing expression of TWIST1 in tissues of stage III breast cancer patients who never smoked (3 patients as described in Table 1) and smoked (5 patients) as determined by immunostaining with anti-TWIST1 antibody. Adjacent bar graph shows quantitative expression of TWIST1 in these samples. Each sample was performed in duplicate or in triplicate. The number of cells showing positive staining of TWIST1 was scored independently and blindly in 3 different area of each slide and then data were expressed as mean ± S.D, where *** represents p≤0.001. F. Effect of TWIST1 silencing on SP. Cells were treated with nicotine or nicotine plus si-TWIST1 and SP was measured as described in A.