Figure 4

Colony-forming efficiency assay and skin organotypic assay. (A), (F) Colony-forming efficiency (CFE) of ABCG2-positive, ABCG2-negative and unsorted keratinocytes. (A) ABCG2-positive and ABCG2-negative keratinocytes were sorted as in Figure 3, 100 cells of each population were seeded in a 100 mm Petri dish preseeded with a lethally irradiated 3T3 feeder layer, and CFE was evaluated at day 12. (F) The ABCG2-positive cells showed a greater number of colonies than the ABCG2-negative cells (CFE, 85% ± 2.1% vs. 27% ± 1.8%, P <0.01, n = 5), whereas there was no difference between ABCG2-negative cells and unsorted cells (CFE, 27% ± 1.8% vs. 31 ± 2.3%, P >0.05, n = 5). (B) to (E) Epidermal reconstruction (organotypic) assay of ABCG2-positive and ABCG2-negative keratinocytes. Human epidermal ABCG2-positive and ABCG2-negative keratinocytes were sorted as in Figure 3 and plated onto a de-epidermized dermis (DED) substrate. (B) ABCG2-positive cells were seeded on top of DED; 1 week after seeding, stratified epidermis was formed, with a typical stratum granulosum and stratum corneum. (C) ABCG2-negative cells were seeded on top of DED; 1 week later, a disordered multilayered tissue had formed. Scale bars: 100 μm. (D) Immunostaining of involucrin in ABCG2-positive culture showed strong positive staining (green) in the suprabasal layer. (E) Weak and abnormal staining for involucrin was detected in ABCG2-negative culture. (G) Clonal analysis of ABCG2-positive epidermis keratinocytes. ABCG2-positive cells were sorted as in Figure 3, and a total of 300 single cells from five donors were picked and inoculated into a 35 mm Petri dish preseeded with lethally irradiated 3 T3 cells. Seven days later, one-quarter of each resulting colony was trypsinized, and transferred to an indicator dish. These dishes were fixed, stained and analyzed 12 days later. The holoclones, meroclones and paraclones shown here are typical figures. Scale bars: D, E, 50 μm.