A Role for BCL6 in the Higher-Order Organization of Genes in DLBCL

Abstract 558

Diffuse large B cell lymphomas (DLBCL), which account for 40% of adult non-Hodgkin's lymphomas, arise from malignant transformation of germinal center (GC) B cells. DLBCL cells display prominent and aberrant transcriptional programming; however, the molecular basis for these alterations is not fully understood. There is now emerging evidence that in addition to aberrations in the DNA sequence and dysregulation of the epigenetic code, the three-dimensional organization of the genome plays a fundamental role in the control of transcriptional programs that determine cell fate. Recent studies in lymphocytes have shown that coordinate gene expression is facilitated by the physical co-localization of genes in the nucleus and by long-range DNA interactions between gene control regions mediated by recruitment of cell-specific transcription factors. The BCL6 oncogenic transcriptional repressor is a key mediator of GC gene expression programs and is essential for maintaining lymphoma cell survival. BCL6 is a member of the BTB/POZ class of proteins, which have been shown to mediate chromatin looping and large-scale chromosome organization through protein oligomerization. Moreover, BCL6 localizes with corepressor proteins to discrete nuclear speckles of unknown function. We hypothesize that BCL6 coordinates gene expression in lymphoma cells through a novel mechanism that involves long-range DNA interactions and the formation of gene regulatory hubs in the nucleus. Using ChIP-sequencing, we surveyed BCL6 binding in DLBCL cell lines and observed that BCL6 binds to multiple sites at each of its target loci including promoters, enhancers, intronic elements, and a novel class of conserved, intergenic (non-enhancer) binding sites, which we postulate may be chromosomal anchoring points. Using chromosome conformation capture (3C) in DLBCL cells, we discovered that chromatin looping occurs across the GC B cell differentiation gene, PRDM1, which is repressed by BCL6, and is bound by BCL6 at its promoter region, a cryptic promoter in intron 3, and a non-enhancer site 100 kb upstream. We have detected long-range DNA interactions occurring between these sites of BCL6 binding, but not between regions of DNA to which BCL6 is not bound. Higher-order folding of the PRDM1 gene appears to be dependent on BCL6 binding, since looping is abolished in cells in which BCL6 is depleted by siRNA and PRDM1 is re-expressed. Thus, repression of the PRDM1 gene in DLBCL cells requires BCL6-mediated chromatin folding. We have extended this analysis to a larger cohort of genes required for DLBCL survival to discover novel mechanisms of BCL6-mediated gene organization in lymphoma. The data suggest that long-range DNA interactions occur at sites of BCL6 binding on BCL6 target genes and could represent a more global mechanism by which these genes are regulated (and dysregulated) in lymphoma. Determining the mechanism by which BCL6 helps to direct regional chromatin organization in lymphoma cells will enable us to understand transcriptional programming and refine strategies for therapeutic targeting of BCL6 and other factors in DLBCL.