Fig. 1

A putative human HpSC population was identified in the CDCP1⁺CD90⁺CD66^– subpopulation. a Schematic of the HpSC derivation protocol. b Human primary FLCs costained with human antibodies against CD66, CDCP1, and CD90 for flow cytometry analysis. CD66^– human primary FLCs were reanalyzed into four fractions based on the expression of CDCP1 and CD90: CDCP1^–CD90⁺CD66^–, CDCP1⁺CD90⁺CD66^–, CDCP1^–CD90^–CD66^–, and CDCP1⁺CD90^–CD66^–. Numbers represent the percentage of each fraction in CD66^– human primary FLCs (n = 7 independent experiments). c Four fractionated cell populations were sorted onto collagen IV-coated plates at a density of 100 cells/cm². After 18 days of culture, colony-forming capabilities of the cell populations were characterized by Giemsa staining. d Efficiency of large-colony formation (containing more than 100 cells per colony) derived from clonal cell sorting as one cell per well of a 96-well plate determined from each fractionated subpopulation on day 18. Results presented as mean colony count ± SD (n = 4 independent experiments). See also Additional file 1: Figure S1, Additional file 2: Figure S2, and Additional file 3: Figure S3. FACS fluorescence-activated cell sorting, CDCP1 CUB domain-containing protein 1, SSC side scatter, FSC forward scatter. ***P < 0.0001