Leukemia and Cancer Stem Cells
Cancers and normal tissue stem cells have much in common: Both have self-renewal capacity, and both develop into differentiated progeny. But do true cancer stem cells exist? We believe that they do and that this realization will have a major impact on the understanding and treatment of cancers. Putative cancer stem cells can be recognized by three attributes: They constitute a homogenous cell population; they, on their own, can initiate cancer; and they both self-renew and undergo differentiation into nontumorigenic progeny.
Many normal tissues start with stem cells. In a tightly regulated sequence, daughter cells undergo successive quantal steps in differentiation and have limited self-renewal capacity. This is an idea that has been around for a long time. For example, chromosome-marking experiments supporting the existence of stem cells for the hematopoietic system were published in 1967. Yet it took 20 years for the definitive demonstration of hematopoietic stem cells (HSC), namely long-term regeneration of multiple lineages of donor-derived blood cells in lethally irradiated mice.1 And it took another 12 years to complete the isolation of the downstream blood cell progenitors, all of them non-self-renewing.
This schema set the stage to examine hematopoietic neoplasms (leukemias and lymphomas). Somewhere in the hierarchy of stem and progenitor cells lie cancer stem cells. These progress, through multiple genetic and gene-expression events that are proto-oncogenic, to full-blown cancer.
Identifying cancer stem cells has been difficult. Looking at fixed tissues and surmising schemes of cell transitions proved to be a recipe for confusion, as was the adoption of molecular pathway morphologies. The first serious attempt to isolate a cancer stem cell was the system that John Dick and his colleagues developed. They transferred cells from patients with acute myelogenous leukemia (AML) successfully into immunodeficient mice.3 These leukemia-initiating cells shared part of the phenotype of normal HSC; that is, they were CD34+38lo/-, but true human HSC are additionally Thy1+ and lacked blood lineage markers. We sought to determine if the leukemia stem cells (LSC) were derived from HSC or progenitors.
We used cells from patients with AML that bore the aml1/eto chromosomal translocation, which plays a proto-oncogenic role in these leukemias. To our surprise, the true HSC had the aml1/eto chromosome, but lacked the potential to produce leukemia blast cells in culture, yielding only normal-looking myeloerythroid colonies. The Thy1- progenitors were the LSC.4 This proved to be consistent with the clinical data: Many of these treated patients with leukemia were healthy for as long as 150 months, yet their marrows contained detectable normal HSC with the aml1/eto chromosomal translocation. We interpreted these findings to mean that the aml1/eto translocation was probably necessary but not sufficient for the full AML disease.
Several independent events are required for the progression of chronic to acute leukemias in mice, and in mouse and human myelopoiesis, only HSC self-renew. Our interpretation (see figure, right) is that most or all proto-oncogenic events short of acute leukemia occur in a succession of HSC clonal progeny; had such early events initially occurred in progenitors, they would be lost as the progenitor lifespan was completed. However, the emergence of the acute leukemic clone could occur at the HSC or progenitor level when the self-renewal pathway genes are activated. To give an example, in a particular group of patients (bearing bcr/abl translocations) the chronic phase leukemia is at the level of HSC, producing myeloid, erythroid, and B lymphoid cells. But when myeloid blast crisis emerges, it is progenitors that are responsible, mainly from the granulocyte-macrophage stage of hematopoiesis.5
This pattern is repeated in solid tumors in other tissues that are arranged in a cellular hierarchy, such as the brain (see Stem cells for brain cancer) and the breast. We have found that breast cancers have a similar cellular hierarchy to the normal gland from which they arose. In malignant breast tumors, a minority population of CD24-/loCD44+ cancer cells has the ability to self-renew while the majority of the cancer cells are destined to stop proliferating.6 The finding that cancers arising in such diverse tissues as the blood, epithelium, and brain all contain cancer stem cells suggests that many, if not all, tumors fit this paradigm.
Identification of cancer stem cells has far reaching implications. Just as normal stem cells have enabled the identification of molecular pathways that regulate self-renewal,7 so cancer stem cells will reveal the differences between these key malignant cells and their normal counterparts. It is possible that such differences could be exploited to eliminate the cancer stem cells, resulting in improved outcome for patients with these devastating diseases.
Irving Weissman is director of Stanford's Institute of Stem Cell Biology and Regenerative Medicine. Michael Clarke is a professor in Medicine and Cancer Biology at Stanford.
1. G.S. Spangrude et al., "Purification and characterization of mouse hematopoietic stem cells," Science, 241:58-62, 1988.
2. K. Akashi et al., "A clonogenic common myeloid progenitor that gives rise to all myeloid lineages," Nature, 404:193-7, 2000.
3. D. Bonnet, J.E. Dick, "Human acute myeloid leukemia is organized as a hierarchy that originates from a primitive hematopoietic cell," Nat Med, 3:730-7, 1997.
4. T. Miyamoto et al., "AML1/ETO-expressing nonleukemic stem cells in acute myelogenous leukemia with 8;21 chromosomal translocation," Proc Natl Acad Sci, 97:7521-6, 2000.
5. C.H. Jamieson et al., "Granulocyte-macrophage progenitors as candidate leukemic stem cells in blast-crisis CML," N Engl J Med, 351:657-67, 2004.
6. T. Reya et al., "Stem cells, cancer, and cancer stem cells," Nature, 414:105-11, 2004.
7. I.K. Park et al., "Bmi-1 is required for maintenance of adult self-renewing haematopoietic stem cells," Nature, 423:302-5, 2003.
A new analysis finds is no link between a researcher’s citations and Twitter mentions of her scientific research.