Acute Promyelocytic Leukemia Stem Cell Derived from Leukemic Progenitor.

Efforts to identify the leukemic stem cell (or leukemia-repopulating cell, LRC) of acute myeloid leukemia (AML) have led to considerable insights, but also to a number of unresolved questions. While the LRC is believed to arise primarily from the hematopoietic stem cell, data from chronic myelogenous leukemia blast crisis and acute promyelocytic leukemia (APL) suggest that they may arise from more committed cells.

In order to clarify the potential of committed progenitors to become LRCs in some human AMLs, we utilized the MRP8-PML/RARA transgenic murine model of APL to perform initial studies directed at the identification of the LRC in APL. Unsorted spleen cells of MRP8-PML/RARA leukemic mice injected into cohorts of sublethally irradiated mice using limiting dilutions showed that the LRC of APL is present in 1:100 to 1:300 cells. FACS analysis of six primary and serially transplanted leukemic specimens revealed that normal hematopoetic stem cells and myeloid progenitors were rare, comprising .007% and .078% of total splenic cells, respectively. These data raised the possibility that the LRC is present at higher frequency than these normal progenitors and does not belong to any of these populations.

Given the scarcity of normal progenitors, we surmised that the LRC may be part of the leukemic population. We performed FACS analysis of seven primary leukemias to elucidate potential heterogeneity within the leukemic cell population. Myeloblasts had the immunophenotype lymph⁻Gr1⁺ckit⁺CD16/32⁺CD34^mod/+. A smaller population that made up 1.4–12.7% of all cells was lymph⁻Gr1⁻ckit^lo/mod and had variable expression of CD16/32 and CD34. Such cells have been identified as myelomonocytic cells and represent a stage in myelopoiesis beyond the GMP. Our findings suggest a hierarchy in which Gr1⁻ckit^lo/mod leukemia repopulating cells proliferate and differentiate to maintain the population of abnormal promyelocytes.

We infer from the dilution analysis data that the LRC in this MRP8-PML/RARA mouse model of acute promyelocytic leukemia cannot be a hematopoietic stem cell or a normal myeloid progenitor cell. Also, it is apparent that the entire blast population is not capable of transferring disease; otherwise, all the mice injected would have developed leukemia. While our data does indicate that while leukemogenesis in APL is a hierarchy, it originates from a newly identified leukemia progenitor cell that is Gr1⁻ckit^lo/mod that may be more mature than a GMP. It is quite likely that the LRC lies within this population of cells. Future work will further delineate the immunophenotype of the LRC in this model and in human leukemia, as well as molecular abnormalities that are important in differentiating from the LRC into the myeloblast phenotype.