Abstract
Abstract 3465
Poster Board III-353
EZH2 is the catalytic subunit of the PRC2 Polycomb complex and mediates transcriptional repression through its histone methyltransferase activity. It is over-expressed in many types of aggressive tumors, e.g., prostate and breast cancer and this over-expression is generally associated with poor patient prognosis. It is also over-expressed in certain lymphomas, e.g., follicular lymphoma; however its exact role and importance in normal and malignant B-cells remains unclear. Most B-cell lymphomas arise from germinal center (GC) B-cells. We thus first set out to investigate the function and activity of EZH2 in normal B-cells. We confirmed a previously published observation that expression of EZH2 protein is greatly elevated during developmental transition from resting Naive B-cells to rapidly proliferating GC B-lymphocytes. Using ChIP-on-chip, we then determined the genome-wide pattern of EZH2 binding in GC B cells and found that EZH2 targets ∼1,800 promoters in these cells (at FDR<0.1). We also mapped the genome-wide distribution of H3K27me3 histone marks in GC B cells; consistent with the known role of EZH2 in catalyzing H3K27me3, a majority (62%) of EZH2 target promoters also displayed a very strong H3K27me3 peak (p=0, hypergeometric test). Also consistent with the repressive nature of H3K27me3, expression arrays rezvealed that EZH2 targets are usually expressed at a lower level in GC B cells than other genes (p<1e-80). However, contrary to a previously postulated role of EZH2 in promoting DNA methylation, our results indicate that EZH2-bound promoters are largely hypomethylated in GC B cells (p=0, Wilcoxon test). From the biological standpoint EZH2 target genes are enriched with transcription factors (p<1e-5), kinases (p<1e-6), and other components of signal transduction pathways such as TGF-beta, WNT, EGFR, PDGFR, and VEGF. EZH2 also targets and represses many tumor suppressor genes, e.g., CDKN1A/p21 and CDKN1B/p27, CDKN2A/p16 and CDKN2A/p14. Using an unbiased motif discovery procedure, we associate EZH2 binding with sequences highly similar to those bound by orthologous PRC2 in Drosophila; we find that EZH2 binding is also associated with the highly statistically significant depletion of regulatory sequences typically bound by transcriptional activators. We then compared the genome-wide binding patterns of EZH2 in GC B cells and embryonic stem cells and observed a strong overlap of EZH2 targets between these cell types (>30% of GC B cells targets are also bound in hESCs, p<1e-378). However, we also observed a large GC B cell-specific EZH2 regulatory program with >1,000 genes. Seeking to extrapolate our binding data to GC-derived Diffuse Large B Cell Lymphoma (DLBCL), we found that the expression profile of many EZH2 target genes is anti-correlated with EZH2 mRNA levels in expression profiles of primary DLBCL tumors. Surprisingly, we found that this anti-correlation was most pronounced among GC B cell-specific EZH2 targets (p<1e-26). In turn, the EZH2 mRNA level was itself positively correlated with cellular proliferation in primary DLBCL tumors, as measured by Ki67 staining (Pearson correlation = 0.3, p<0.001). Finally siRNA-mediated down-regulation of EZH2 in SUDHL4 DLBCL cells resulted in acute cell cycle arrest at the G1/S transition in SUDHL4 cells and upregulation of EZH2 target genes with cell cycle inhibitory functions such as those mentioned above. Altogether, these data suggest a scenario whereby EZH2 upregulation in GC B-cells leads to its recruitment to genes containing Polycomb Response Elements with consequent H3K27 trimethylation and silencing of a GC B cell context specific cohort of genes including those involved in restraining cellular proliferation, thus contributing to the ability of these cells to undergo massive clonal expansion. This function of EZH2 may also contribute to the malignant transformation of GC B-cells into DLBCLs and facilitate their proliferative phenotype. Thus, our results indicate that therapeutic targeting of EZH2 might have significant anti-lymphoma effects and support the rationale for development of inhibitors of the EZH2 SET domain.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.