Abstract In our studies of hematopoietic malignancies, acute (Nature 441:475) and chronic myeloid leukemia (Cancer Cell 16:137) appear to follow a cancer stem cell model. In both cases, leukemogenic cells are rare, phenotypically distinct from the vast majority of other leukemia cells, and hierarchically organized. However, tumorigenic capacity is a common attribute of cells from a variety of other human and mouse cancers we have studied, including melanoma. Many cancers are not driven by rare cancer stem cells. We recently showed that the optimization of xenotransplantation assay conditions, including the use of more highly immunocompromised NOD/SCID IL2Rγnull mice, can increase the detection of tumorigenic cells in some cancers by orders-of-magnitude (Nature 456:593). This indicated that tumorigenic cells are common in some human cancers, and that to accurately estimate the frequency of cancer cells with tumorigenic potential it is critical to optimize assay conditions. In this regard, the characterization of human cancer stem cells is no different from the characterization of normal human hematopoietic stem cells, which also depended on the use of the most permissive possible xenotransplantation assay conditions, including the use of highly immunocompromised NOD/SCID IL2Rγnull mice, rather than NOD/SCID mice (which retain natural killer and other innate immune cells that reject some transplanted human cells). Optimization of xenotransplantation assay conditions revealed that cells with tumorigenic potential are common in human melanoma (at least 25% of cells from stage 3 and 4 melanomas have the potential to form tumors after transplantation), suggesting that melanoma growth and progression are unlikely to be driven by rare cancer stem cells. We have also been unable to find any markers that can distinguish tumorigenic from non-tumorigenic melanoma cells, suggesting that melanoma lacks the obvious hierarchical organization observed in other cancers that follow a stem cell model, such as leukemia. Despite observing extensive phenotypic heterogeneity among melanoma cells, we have been unable to detect any differences in tumorigenic capacity among melanoma cells from multiple patients that differed in the expression of ABCB5, CD133, CD166, L1-CAM, CD49f, A2B5, p75, CD44, CD54, CD29, MCAM, HNK1, CD49b, CD49d, E-Cadherin, N-Cadherin, or c-kit. When transplanted into NOD/SCID IL2Rγnull (NOG) mice, only 10 melanoma cells from either the positive or negative fractions of each of these markers readily formed tumors. We therefore remain unable to find any evidence that human melanoma cells are segregated into a hierarchy of intrinsically distinct tumorigenic and non-tumorigenic fractions. Moreover, all fractions of melanoma cells appear to be able to recapitulate the phenotypic diversity of the tumors from which they are drawn. For example, irrespective of whether melanomas arose from CD133+ or CD133- melanoma cells, they contained similar proportions of CD133+ and CD133- cells (Cell 138:822). This suggests that the capacity to recapitulate the phenotypic diversity of primary tumors is not necessarily unique to cancer stem cells but rather that many cells can form phenotypically diverse progeny in certain cancers. Finally, we observed significant differences in the growth rates of tumors from different patients, but these differences were not associated with differences in the frequency of cells with tumorigenic potential. This suggests there are likely to be biologically and clinically important differences among melanoma cells that are not explained by the cancer stem cell model. The available data suggest that many melanoma cells are capable of contributing to disease progression and that it will not be possible to cure melanoma by targeting rare populations of cancer stem cells. Preliminary studies of additional mouse and human cancers suggest that melanoma is not unusual in having common tumorigenic cells. Overall, some cancers appear to be driven by minority populations of cancer stem cells, while other cancers appear to have common tumorigenic cells with little evidence of hierarchical organization. It will be critical to determine which cancers follow a stem cell model and which do not so that efforts to identify and target stem cells can be appropriately focused. Citation Format: Elsa Quintana, Mark Shackleton, Sean J. Morrison. Some cancers follow a stem cell model, while other cancers have common tumorigenic cells with little hierarchical organization [abstract]. In: Proceedings of the AACR 101st Annual Meeting 2010; 2010 Apr 17-21; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2010;70(8 Suppl):Abstract nr PL03-02
Read full abstract