“Hierarchical” Induction and “Stochastic” Maintenance of Leukemia in An In Vivo Model of t(6;9) Positive Acute Myeloid Leukemia.

Abstract 2953

Poster Board II-929

Stem cells have been shown to play an important role in the pathogenesis and maintenance of a significant number of malignancies, including leukemias. Similar to normal hematopoiesis the AML cell population is thought to be hierarchically organized. According to this model, only a few stem cells (LSC) are able to initiate and maintain the disease. The inefficient targeting of the leukemic stem cells (LSC) is considered responsible for relapse after the induction of complete hematologic remission (CR) in AML. t(6;9)-positive AML is classified as a separate entity, because of its young age of onset and poor prognosis. The t(6;9) associated fusion protein is DEK/CAN. Assuming that in AML the genetic aberration, here the t(6;9) and the expression of DEK/CAN, represents the initiation event of the leukemogenic process we wanted i.) to disclose its effects on the biology of primitive hematopoietic stem cells (HSC) and its leukemogenic potential and ii.) to characterize the leukemia-initiating cell and the cell population able to maintain the disease in vivo. The model was based on a classical transduction/transplantation system of murine Sca1+/lin- HSC combined with a novel approach for the enrichment of transformed cells with long-term stem cell properties. We found that i.) DEK/CAN induced leukemia from the Sca1⁺/lin⁻ HSC with a frequency of 20% and a long latency of 8-12 months. ii.) DEK/CAN did not efficiently block the differentiation of committed progenitors; iii.) DEK/CAN increased number of colony forming cells in Sca1⁺/lin⁻ HSC which did not exhibit increased replating efficiency as compared to controls; iv.) DEK/CAN augmented ST-HSC potential but not LT-HSC of murine Sca1⁺/lin⁻ HSCs, most likely due to its incapacity to up-regulate p21Cip1/Waf1 expression. Based on the hypothesis that DEK/CAN exerts its leukemogenic effects on only a small proportion of the Sca1+1/lin- population, we proceeded to select and to amplify rare DEK/CAN-positive cells with the leukemia-initiating potential, by a negative selection of cell populations with proliferation potential without long term stem cell-capacity (LT). Therefore we expressed DEK/CAN in Sca1⁺/lin⁻ cells and enriched this population for LT- (lin⁻/Sca1⁺/c-Kit⁺/Flk2⁻) and ST-HSC (lin⁻/Sca1⁺/c-Kit⁺/Flk2⁺). After a passage first in semi-solid medium for 7 days and subsequent transplantation into lethally irradiated mice, cells from the ensuing CFU-S day12 were again transplanted into sublethally recipient mice. We here report that i.) after 4 to 42 weeks, 6/6 mice developed AML without signs of differentiation in the group transplanted with the lin⁻/Sca1⁺/c-Kit⁺/Flk2⁻ population but not from that transplanted with lin⁻/Sca1⁺/c-Kit⁺/Flk2⁺ cells; ii.) the DEK/CAN-induced AML was efficiently transplanted into secondary recipients exhibiting a very aggressive clinical picture; iii.) the leukemic cell population gave origin to four different clearly distinct subpopulations defined by surface marker pattern as an expression of populations with distinct differentiation status, all able - after sorting - to give leukemia in sublethally irradiated recipients: lin⁻/Sca1⁺/c-Kit⁺/CD34⁻ (LT) lin⁻/Sca1⁺/c-Kit⁺/CD34⁺ (ST), Sca1⁻/c-Kit⁺/Mac1⁺/Gr1⁺, Sca1⁻/c-Kit⁺/Mac1⁻/Gr1⁻. These findings strongly suggest that there is a difference between a leukemia-initiating (L-IC) and leukemia-maintaining (L-MC) cell population in the murine DEK/CAN leukemia model. In contrast to the L-IC, represented by a very rare subpopulation of primitive HSC, recalling a hierarchical stem cell model, the L-MC is represented by a larger cell population with a certain grade of phenotypical heterogeneity, but a high grade of functional homogeneity recalling a stochastic cancer induction model.