The steps to leukemia following an in utero fusion of MLL (HRX, ALL-1) to a partner gene in humans are not. known. introduction of the MII-AF9 fusion gene into embryonic stem cells results in leukemia in mice with cell-type specificity similar to. humans. In this study we used myeloid colony assays, immunophenotyping, and transplantation to evaluate myelopoiesis in MII-AF9 mice. Colony assays demonstrated that both prenatal and postnatal MII-AF9 tissues have significantly increased numbers of CP11b(+)/CD117(+)/Gr-1(+/-) myeloid cells, often in compact clusters. The self-renewal capacity of prenatal myeloid progenitors was found to decrease following serial replating of colony-forming cells. In contrast, early postnatal myeloid progenitors increased following replating; however, the enhanced self-renewal of early postnatal myeloid progenitor cells was limited and did not result in long-term cell, lines or leukemia in vivo. Unlimited replating, longterm CD11b/Gr-1(+) myeloid call lines, and the ability to produce early leukemia in Vivo in transplantation experiments, were found only in mice with overt leukemia. Prenatal MII-AF9 tissues had reduced total I (mature and progenitor) CD11b/Gr-1+ cells compared with wild-type tissues. Colony replating, immunophenotyping, and cytochemistry suggest that any perturbation of cellular differentiation from the prenatal stage onward is partial and largely reversible. We describe a no vel informative in vitro and in vivo model system that permits study of the stages in the pathogenesis of MII fusion gene leukemia, beginning in prenatal myeloid cells, I progressing to a second stage, in the postnatal period and, finally, resulting in overt leukemia in adult animals. (C) 2003 by The American Society of Hematology.