| The hematopoietic system produces appropriate levels of blood cells over an individual's lifetime through a careful balance of differentiation, proliferation and self-renewal. The deregulation of these processes leads to the development of malignancies including myeloproliferative disorders (MPDs) and acute myeloid and lymphoid leukemias (AML and ALL, respectively). A prerequisite for targeted therapies is an understanding of the processes which subvert the developmental program of the hematopoietic system in these diseases. This involves identifying the cell of origin and the sequence of molecular lesions required to generate cells capable of initiating disease, as well as phenotypically characterizing the cells responsible for sustaining the malignant clone in vivo. Significant progress has been made in these areas, particularly through the identification of recurrent genetic alterations, such as activating mutations of the Janus kinase 2 ( JAK2) gene in MPDs and translocations involving the mixed-lineage leukemia (MLL) gene in acute leukemias. Characterization of these oncogenes has been performed using cell lines and mouse models; however, to precisely model the development of hematopoietic malignancies, it is necessary to perform similar studies in primary human cells. To this end, we have developed an experimental system where genetic hits are introduced into lineage-depleted human umbilical cord blood (Lin-CB) by retroviral-mediated oncogene overexpression. Using this approach, we have shown that TEL-JAK2, a constitutively active variant of JAK2, drives erythropoietin-independent erythropoiesis in vitro and induces the rapid development of myelofibrosis in vivo. This work indicated that activated JAK2 signaling is sufficient to drive MPD-associated disease processes and also established the feasibility of utilizing xenotransplantation systems to develop models of human hematopoietic malignancies. Using a similar approach, we subsequently demonstrated that Lin-CB cells expressing MLL fusion genes reproducibly generate AML and B-ALL in vivo. Serial transplantation showed that these diseases were sustained by leukemia-initiating cells (L-ICs) that undergo phenotypic evolution, establishing that L-ICs in established disease need not resemble the cell of origin. Moreover, some L-ICs could initiate both myeloid and B-lymphoid disease, and in these instances, microenvironmental cues could select for the resulting leukemia lineage. Collectively, the models described here provide numerous insights into the pathobiology of human hematopoietic malignancies. |
