An emerging question about putative cancer stem cells in established cell lines-are they true stem cells or a fluctuating cell phenotype?

It has been proposed that established cell lines contain populations of cancer stem cells (CSCs), which are responsible for expansion of these cell lines and their metastatic potential. To address this issue better, we employed a human ovarian cancer cell line, A2780, and sorted cells according to the postulated highly mestatatic cancer stem cell phenotype, CD24⁺CD44^-, and the less-metastatic CD24^-CD44⁺ and CD24^-CD44^- phenotypes. These cells were employed in chemotaxis assays in vitro to migrate in response to conditioned media harvested from bone marrow or liver cells damaged by irradiation and in in vivo assays to grow tumors after injection into immunodeficient mice. We also sorted single cells expressing all three phenotypes by FACS and expanded them to grow clones. We found that the CD24⁺CD44^- cells are a highly migratory population compared with CD24^-CD44⁺ and CD24^-CD44^- cells and were seeded in higher numbers in murine bone marrow and liver after intravenous injection. Most importantly, we observed that singly sorted cells efficiently expanded ex vivo into cell populations that represented all phenotypes of the parental cell line. Thus, our data indicate that cells expressing a certain set of markers, e.g., CD24, have at any given moment a higher potential to migrate and metastasize. However, cells that are CD24-negative, if expanded from a singly sorted cell, may give rise to cells containing all of the markers, including CD24. Based on this finding, we propose that the CSC phenotype in cell lines fluctuates with cell expansion.