Tumors arise as clones of a single cell, but classic transplant experiments suggested that few malignant cells within an established tumor have the potential to drive its growth. When transferred into immunocompromised mice, only a few human leukemic cells initiate tumors (an estimated 1 in 106 cells). This observation led investigators to propose that tumors require rare cancer stem cells for growth, but it did not account for any effects that differences between donor and host might have on tumor growth, such as incompatibility of growth factors and their receptors.
Priscilla Kelley et al. tested these microenvironmental influences with a simple and elegant experiment. Instead of transferring human leukemic cells into mice, the authors transferred 105 mouse leukemic cells into a histocompatible mouse, then serially diluted this amount to determine the lowest number of cells that had tumor-initiating potential. They found that mice that had received only ten cells developed leukemia. The only difference between the small and large transfers was a correlated time of leukemia onset, which is consistent with multiple cells in the large sample growing independently and simultaneously. This astonishing result indicates that most leukemic cells have tumor-growth sustaining potential and strikes a blow to the stem-cell hypothesis for leukemia.
This is a preview of subscription content, access via your institution
