Most human acute myeloid leukaemia (AML) cells have limited proliferative capacity, suggesting that the leukaemic clone may be maintained by a rare population of stem cells. This putative leukaemic stem cell has not been characterized because the available in vitro assays can only detect progenitors with limited proliferative and replating potential. We have now identified an AML-initiating cell by transplantation into severe combined immune-deficient (SCID) mice. These cells homed to the bone marrow and proliferated extensively in response to in vivo cytokine treatment, resulting in a pattern of dissemination and leukaemic cell morphology similar to that seen in the original patients. Limiting dilution analysis showed that the frequency of these leukaemia-initiating cells in the peripheral blood of AML patients was one engraftment unit in 250,000 cells. We fractionated AML cells on the basis of cell-surface-marker expression and found that the leukaemia-initiating cells that could engraft SCID mice to produce large numbers of colony-forming progenitors were CD34+ CD38-; however, the CD34+ CD38+ and CD34- fractions contained no cells with these properties. This in vivo model replicates many aspects of human AML and defines a new leukaemia-initiating cell which is less mature than colony-forming cells.

Mutations of the BRCA1 gone in humans are associated with predisposition to breast and ovarian cancers. We show here that Brca1+/- mice are normal and fertile and lack tumors by age eleven months. Homozygous Brca1(5-6) mutant mice die before day 7.5 of embryogenesis. Mutant embryos are poorly developed, with no evidence of mesoderm formation. The extraembryonic region is abnormal, but aggregation with wild-type tetraploid embryos does not rescue the lethality. In vivo, mutant embryos do not exhibit increased apoptosis but show reduced cell proliferation accompanied by decreased expression of cyclin E and mdm-2, a regulator of p53 activity. The expression of cyclin-dependent kinase inhibitor p21 is dramatically increased in the mutant embryos. Buttressing these in vivo observations is the fact that mutant blastocyst growth is grossly impaired in vitro. Thus, the death of Brca1(5-6) mutant embryos prior to gastrulation may be due to a failure of the proliferative burst required for the development of the different germ layers.

Mutations in the SMAD4/DPC4 tumor suppressor gene, a key signal transducer in most TGF␤-related pathways, are involved in 50% of pancreatic cancers. Homozygous Smad4 mutant mice die before day 7.5 of embryogenesis. Mutant embryos have reduced size, fail to gastrulate or express a mesodermal marker, and show abnormal visceral endoderm development. Growth retardation of the Smad4-deficient embryos results from reduced cell proliferation rather than increased apoptosis. Aggregation of mutant Smad4 ES cells with wild-type tetraploid morulae rescues the gastrulation defect. These results indicate that Smad4 is initially required for the differentiation of the visceral endoderm and that the gastrulation defect in the epiblast is secondary and non-cell autonomous. Rescued embryos show severe anterior truncations, indicating a second important role for Smad4 in anterior patterning during embryogenesis.