Figure 7

CD133 expression in PyMT mammary tumor epithelial cells is regulated by HIF-1α. (A) Subconfluent wild-type (WT) or knockout (KO) cultured cells were exposed to acute hypoxia (Hyp) (6 hours), then immunostained for CD133 (green) and counterstained with DAPI. Images were captured at ×200 original magnification. Scale bar = 50 μm. (B) Representative CD133-phycoerythrin (CD133-PE) staining profiles of WT (red histogram) and KO (blue histogram) cells cultured at normoxia (Nor) and subjected to fluorescence-activated cell sorting analysis. The isotype-only antibody control is also plotted (green histogram). The percentage of cells that were CD133-positive was determined on the basis of the live, singlet, hematopoietic lineage panel-negative (Lin^neg) parent population using the FlowJo data analysis software package. (C) Secondary WT or KO tumorspheres cultured at normoxia were harvested at the study end point from ultralow adhesion wells, dried onto glass slides, stained with CD133-PE, counterstained with DAPI and imaged by confocal microscopy. The highlighted area (white boxes) shows a higher-magnification image, which demonstrates that CD133 is localized to the cell surface. (D) Histogram of CD133-PE (red) in the live, singlet, Lin^neg parent population of cells isolated from tumors that arose in the PyMT transgenic mouse as compared to the isotype antibody control (green). Approximately 6% of cells were defined in this experiment as CD133^hi. (E) A representative experiment showing the mean sphere formation efficiency (SFE)/well ± SEM of CD133^hi versus CD133^neg subpopulations that were isolated by flow sorting and cultured at normoxia at a density of 10,000 cells/well in six-well format (n ≥ 8 wells/genotype, unpaired Student's t-test). The SFE was determined as the percentage of cells capable of forming spheres per the total number of single cells plated.