Ezh2 Plays a Critical Role in the Progression of MLL-AF9-Induced Acute Myeloid Leukemia

Abstract 57

The polycomb group proteins function in gene silencing through histone modifications. They have been characterized as a general regulator of stem cells, but also play a critical role in cancer. EZH2 is a catalytic component of the polycomb repressive complex 2 (PRC2) and tri-methylates histone H3 at lysine 27 to transcriptionally repress the target genes. Although EZH2 is over-expressed in various cancers including hematological malignancies, it remains unknown how EZH2 contributes to the initiation and/or progression of acute myeloid leukemia (AML).

To understand the role of EZH2 in AML, we transformed granulocyte macrophage progenitors (GMPs) from Cre-ERT;Ezh2^+/+ and Cre-ERT;Ezh2^flox/flox mice with the MLL-AF9 fusion gene. Then, Ezh2 was deleted by inducing nuclear translocation of Cre by adding tamoxifen to culture. We found that proliferation of Ezh2^δ/δ transformed cells was severely compromised upon deletion of Ezh2 (Ezh2^δ/δ) in liquid culture. They gave rise to a significantly reduced number of colonies in replating assays. Of note, while Ezh2^+/+ cells formed compact colonies composed of immature myeloblasts, Ezh2^δ/δ cells formed dispersed colonies composed of differentiated myeloid cells. We next transplanted Cre-ERT;Ezh2^+/+ and Cre-ERT;Ezh2^flox/flox GMPs transformed by MLL-AF9 into recipient mice. All the recipient mice developed AML by 3 weeks after transplantation. At 3 weeks after transplantation, we depleted Ezh2 by intraperitoneal injection of tamoxifen. Deletion of Ezh2 significantly prolonged the survival of the recipient mice (60 days vs. 76 days, p<0.0001), although all the mice eventually died of leukemia. Nonetheless, as was detected in vitro, Ezh2^δ/δ AML cells in BM were apparently differentiated in morphology compared with the control. Ezh2^δ/δ AML cells in BM gave rise to 10-fold fewer colonies in methylcellulose medium compared with Ezh2^+/+ AML cells, and again showed an obvious tendency of differentiation. These observations imply that Ezh2 is critical for the progression of MLL-AF9 AML and maintains the immature state of AML cells. To elucidate the mechanism how Ezh2 promotes the progression of MLL-AF9-induced AML, we examined the genome-wide distribution of tri-methylation of histone H3 at lysine 27 (H3K27me3) by ChIP-sequencing and microarray-based expression analysis. ChIP-sequencing using Ezh2^+/+ and Ezh2^δ/δ BM AML cells identified 3525 and 89 genes exhibiting a ≧ 10-fold enrichment in H3K27me3 levels in Ezh2^+/+ and Ezh2^δ/δ AML cells, respectively, confirming a drastic reduction in the levels of global H3K27me3 in the absence of Ezh2. Microarray analysis using lineage marker (except for Mac1)⁻Sca-1⁻c-Kit⁺FcγRII/III^hi BM AML cells revealed 252 upregulated and 154 downregulated genes (≧ 2-fold) in Ezh2^δ/δ AML cells compared with Ezh2^+/+ AML cells. Of interest, the absence of Ezh2 did not affect the transcriptional activation of the major target genes of MLL-AF9, including HoxA9 and Meis1. Because Ezh2 functions as transcriptional repressor, de-repressed genes could be direct targets of Ezh2. Based on these data, we are now engaged in further comprehensive analysis to narrow down the direct target genes of Ezh2 responsible for the progression of AML. Collectively, our findings suggest that Ezh2 is the major enzyme for H3K27me3 in AML and contributes to the progression of AML by regulating transcription a cohort of genes that are supposedly relevant to the self-renewal capacity and perturbed differentiation of AML stem cells.