In an attempt to explore novel treatment modalities in acute myeloid leukemia (AML), we studied the role of costimulatory and cytokine gene immunotherapy in murine AML. We have previously shown that leukemic mice can be cured with CD80 transfected leukemic cells (B7. 1-AML vaccine) administered early in the course of the disease and that the failure B7.1-AML vaccines administered late cannot be attributed to immunosuppression induced by tumor growth. CD8+ T cells, which are necessary for tumor rejection, are activated rather than suppressed during the first half of the leukemic course in nonvaccinated mice. In this report, we question whether CD86 (B7.2) or the cytokines granulocyte-macrophage colony-stimulating factor (GM-CSF), interleukin-4 (IL-4), or tumor necrosis factor-alpha (TNF-alpha) can improve the vaccination potential of AML cells. The choice of cytokines was based on their combined and alone as well ability to direct the differentiation of CD34+ cells into potent antigen-presenting dendritic cells in vitro. Our studies show that (1) mice vaccinated with a leukemogenic number of AML cells engineered to express B7.2 (B7.2-AML) or to secrete GM-CSF, IL-4, or TNF-alpha (GM-, IL-4-, TNF-alpha-AML) do not develop leukemia; (2) GM-AML cells are tumorigenic in sublethally irradiated SJL/J mice but not in Swiss nu/nu mice, indicating that killing of tumor cells is not T-cell-dependent; (3) vaccines with irradiated GM-AML, but not B7.2-, IL-4-, or TNF-alpha-AML cells, can elicit leukemia-specific protective and therapeutic immunity; and (4) in head-to-head comparison experiments, vaccination with irradiated GM-AML is more potent than B7.1-AML, curing 80% and providing 20% prolonged survival of the leukemic mice at week 2, as opposed to cures only up to 1 week with B7.1-AML vaccines. These preclinical data emphasize that GM-CSF gene immunotherapy deserves clinical evaluation in AML.