C/ebpα and Pu.1 Are Members of a Critical Regulatory Network Required for Hoxa9-Mediated Immortalization

Abstract 650

HOXA9 is a homeodomain-containing transcription factor that plays important roles in both development and hematopoiesis. Deregulation of HOXA9 occurs in a variety of acute lymphoid and myeloid leukemias and plays a key role in their pathogenesis. More than 50% of acute myeloid leukemia (AML) cases show up-regulation of HOXA9, which correlates strongly with poor prognosis. Nearly all cases of AML with mixed lineage leukemia (MLL) translocations have increased HOXA9 expression, as well as cases with mutation of the nucleophosmin gene NPM1, overexpression of CDX2, and fusions of NUP98. Despite the crucial role that HOXA9 plays in development, hematopoiesis and leukemia, its transcriptional targets and mechanisms of action are poorly understood. Previously we identified Hoxa9 and Meis1 binding sites in myeloblastic cells, profiled their epigenetic modifications, and identified the target genes regulated by Hoxa9. Hoxa9 and Meis1 co-bind at hundreds of promoter distal, highly evolutionarily conserved sites showing high levels of histone H3K4 monomethylation and CBP/p300 binding characteristic of enhancers. Hoxa9 association at these sites correlates strongly with increases in histone H3K27 acetylation and activation of downstream target genes, including many proleukemic gene loci. De novo motif analysis of Hoxa9 binding sites shows a marked enrichment of motifs for the transcription factors in the C/EBP and ETS families, and C/ebpα and the ETS transcription factor Pu.1 were found to cobind at Hoxa9-regulated enhancers. Both C/ebpα and Pu.1 are known to play critical roles in the establishment of functional enhancers during normal myeloid development and are mutated or otherwise deregulated in various myeloid leukemias. To determine the importance of co-association of Hoxa9, C/ebpα and Pu.1 at myeloid enhancers, we generated cell lines from C/ebpα and Pu.1 conditional knockout mice (kindly provided by Dr. Daniel Tenen, Harvard University) by immortalization with Hoxa9 and Meis1. In addition we transformed bone marrow with a tamoxifen-regulated form of Hoxa9. Strikingly, loss of C/ebpα or Pu.1, or inactivation of Hoxa9, blocks proliferation and leads to myeloid differentiation. ChIP experiments show that both C/ebpα and Pu.1 remain bound to Hoxa9 binding sites in the absence of Hoxa9. After the loss of Pu.1, both Hoxa9 and C/ebpα dissociate from Hoxa9 binding sites with a corresponding decrease in target gene expression. In contrast, loss of C/ebpα does not lead to an immediate decrease in either Hoxa9 or Pu.1 binding, suggesting that C/ebpα may be playing a regulatory as opposed to a scaffolding role at enhancers. Current work focuses on performing ChIP-seq analysis to assess how C/ebpα and Pu.1 affect Hoxa9 and Meis1 binding and epigenetic modifications genome-wide, and in vivo leukemogenesis assays to confirm the requirement of both Pu.1 and C/ebpα in the establishment and maintenance of leukemias with high levels of Hoxa9. Collectively, our findings implicate C/ebpα and Pu.1 as members of a critical transcription factor network required for Hoxa9-mediated transcriptional regulation in leukemia.