DOT1L and Histone H3 Lysine79 Methylation as a Therapeutic Target in Mixed Lineage Leukemia.

Abstract 212 Epigenetic gene regulation is emerging as a major mechanism of regulating genetic programs and pathways involved in various forms of cancer. Specifically, recently published data and preliminary evidence suggests that histone modifications directly and indirectly affected by the Mixed Lineage Leukemia (MLL) protein may play a fundamental role in the pathogenesis of certain leukemias. Rearrangements of the MLL gene are found in a subset of pediatric and adult acute lymphoid and myeloid leukemia (ALL and AML). Leukemias with MLL-rearrangements tend to have a poor prognosis, particularly infant-ALL and treatment associated AML. MLL is a histone modifying enzyme, methylating histone 3 at lysine 4 (H3K4). In most MLL fusions, the domain harboring the methyl transferase activity (Set-domain) is lost. However, several fusion partners of MLL, such as AF4, AF9, AF10 and ENL, have been shown to bind and potentially recruit another histone methyl transferase, DOT1L, that methylates histone 3 at lysine 79 (H3K79). Chromatin immunoprecipitation studies in MLL-rearranged leukemia cells indeed revealed elevated H3K79 methylation at MLL-fusion target loci. These results were consistent in leukemic cells from Mll-AF4 conditional knock in mice, MLL-AF4 expressing cell lines and primary human t(4;11) (MLL-AF4) leukemia cells. H3K79 is a chromatin modification associated with actively transcribed genes, and H3K79 methylation profiles correspond well to expression profiles in MLL-rearranged cells. This prompted the hypothesis that certain MLL-fusions transform cells in part by mis-targeting DOT1L, and promoting inappropriate histone methylation. We decided to test this hypothesis using an RNAi approach. Transduction of human leukemia cell lines carrying a t(4;11) translocation (MLL-AF4) with 2 different lentiviral shRNA constructs directed against DOT1L show a 60-80% reduction in global H3K79 methylation. This reduction was also observed on known MLL target loci such as the 5' HoxA cluster genes, which are central to MLL-mediated leukemogenesis. HoxA5 and HoxA9 expression levels from hypomethylated loci were greatly reduced in cell expressing the DOT1L shRNAs. Phenotypically, DOT1L knockdown adversely affected in vitro viability and proliferation of 2 ALL cell lines expressing the MLL-AF4 fusion gene, SEM-K2 and RS4;11. Nalm-6 and Jurkat ALL control cells were unaffected by DOT1L suppression. We developed a xenograft mouse model using SEM-K2 and Jurkat cells which stably express luciferase, thus allowing monitoring of leukemia development in live mice using bioluminescence imaging. In this in vivo model, DOT1L suppression led to a significant reduction in the time to onset of leukemia in t(4;11) SEM-K2 cells, but not Jurkat control cells. We are currently extending our studies to include a larger panel of human leukemia cells with different cytogenetic abnormalities, including AML cell lines. The results presented here, particularly if confirmed in a larger panel of cell lines and primary patient cells, should establish DOT1L as a highly promising therapeutic target for MLL-rearranged leukemias. Disclosures: Armstrong: Epizyme Pharmaceuticals: Consultancy.