Coordinated regulation of Hox gene expression during hematopoiesis is epigenetically controlled via chromatin modification by Polycomb group (PcG) and Trithorax (MLL) protein complexes. Whereas the oncogenic potential of certain HOX genes in leukemia has already been defined, little is known about their role in Diffuse Large B-cell Lymphomas (DLBCL). The primary focus of our studies is to determine the contribution of PcG-mediated repression of HOX and other genes to DLBCL pathogenesis. The PcG protein, Ezh2, is vital for maintaining both pluripotency of stem cells and identity of differentiated cells. Ezh2 tri-methylates lysine K27 of histone 3 (H3K27me3), a histone modification associated with gene silencing. Importantly, Ezh2 is frequently overexpressed in DLBCLs suggesting a role for EZH2 in lymphomagenesis. In support to this notion we discovered that Ezh2 is essential for DLBCL cell survival. By depleting Ezh2 level using RNAi, we found that loss of Ezh2 triggers cell cycle arrest and death of DLBCL cells. This finding prompted us to initiate functional studies aimed at uncovering Ezh2 target genes that mediate the observed cellular response in DLBCL cells. We first focused on a potential role of Ezh2 in regulation of HOX genes. We compared and contrasted Ezh2 targets in both normal Germinal Center (GC) B-cells and GC-derived DLBCLs to determine the normal and pathologic function of EZH2. We employed a tiling ChIP-chip approach covering the four human HOX clusters and mapped Ezh2 and H3K27m3 within HOX gene clusters. We further verified gene expression status of a subset of Hox genes by QPCR. These data indicated that

  • Ezh2 and its cognate H3K27m3 mark are present at promoters of HoxC genes in both mature GC B-cells and GC-derived lymphoma cells, thereby driving the HoxC locus silent, suggesting that both rapidly dividing GC cells and GC-derived lymphoma cells require epigenetic silencing of this locus in order to maintain their phenotype.

  • Both Ezh2 and the corresponding H3K27m3 transcription repression mark are absent within the promoter region of HoxA9 gene. HoxA9 promotes stem cell self-renewal and it is aberrantly activated in AML cells. This observation is especially striking as the HoxA9 is embedded into the Ezh2-sealed region in DLBCL cells, suggesting an Ezh2-independent mode of regulation. We are in the process of testing functional significance of this finding for lymphoma pathogenesis.

  • we found that HoxB genes that are differentially expressed in progenitor vs. lineage committed cells are silent in DLBCL cells according to H3K27m3/Ezh2 pattern and gene expression analysis.

Intriguingly, the early progenitor specific gene, HoxB3, is uniquely not bound by EZH2 nor H3K27 methylated and was highly expressed in lymphoma cells. This finding underscores a potential functional significance of re-expression of genes that control cell self-renewal in malignances that derive from mature B cells. We also examined transcriptional programming by EZH2 at the genomic level by ChIP-on-chip using NimbleGen 24,000 promoter arrays. EZH2 was bound to ∼1700 promoters in DLBCL cells and a similar number of genes displayed H3K27 methylation. Gain and loss of function studies are underway to identify the contribution of the most likely EZH2 direct targets genes to the DLBCL survival including both HOX genes and other genomic direct target genes. Taken together, our data suggest a critical role for EZH2 mediated epigenetic silencing of HOX and other genes in DLBCL - and implicate aberrant HOX gene expression in DLBCL pathogenesis.

Author notes

Disclosure: No relevant conflicts of interest to declare.