Fig. 4

Further validation of CDC50A as a biomarker of cancer-initiating cells in primary epithelial ovarian cancer tissues. A CDC50A⁺ cells could be detected through FACS in human ovarian tumours (PID045) and ascites (PID046). B shows the much higher sphere-forming activity of CDC50A⁺Lin⁻ cells than CDC50A⁻Lin⁻ cells sorted from individual epithelial ovarian cancer tumours (PID003, PID006 and PID008, input: 10⁴ cells/well). Scale bar, 50 μm. N = 3, P < 0.001. C Higher fractions of CDC50A⁺Lin⁻ cells in sphere culture with cells isolated from epithelial ovarian cancer tumours (4.6% vs. 22.7% on average, n = 4). D A representative result of semiquantitative PCR analysis of markers for stem cells and epithelial-mesenchymal transition in CDC50A⁺ cells from a primary ovarian cancer (PID003). The grouping of gel cropped from different gels. The raw data was shown in Supplementary Fig. 9. E A representative FACS analysis of primary and secondary xenografts in NSG mice generated with CDC50A⁺ cells sorted from human epithelial ovarian cancer (PID003). The percentage of CDC50 A⁺ cells was similar to that of the original cancer. F. When the percentage of CDC50A⁺ cells in EOC tissue was over 4.145%, these patients were separated into a high CDC50A group, and the rest were sorted into a low CDC50A group. Between the two groups, after adjusting for the optimal debulking surgery, the number of CDC50A-positive cells was correlated with PFI (p = 0.031, HR 0.260, 95% CI 0.770 ~ 0.885)