Discovery and Characterization of Promoter and Super-Enhancer-Associated Dependencies through E2F and BET Bromodomains in Multiple Myeloma

Recently we reported the concomitant inhibition of MYC and E2F activity in multiple myeloma (MM) upon treatment with the BET bromodomain inhibitor JQ1. BET bromodomains (BETs) are transcriptional co-activators that occupy active promoters and enhancers, but are asymmetrically localized to a small number of “super-enhancer” domains. JQ1 treatment results in disproportionate displacement of BETs from super-enhancers leading to potent and selective downregulation of super-enhancer target genes, with MYC levels > 90% depleted after 6 hours. In contrast, E2F protein levels are relatively unperturbed by JQ1 treatment. Prompted by these observations, we here explored the global genomic occupancy and interaction of E2F and BET bromodomains in regulating MM cell state.

We have previously reported that E2F1 and its heterodimerization partner DP1 promote MM tumor proliferation both in vitro and in vivo with a statistically significant inverse correlation between DP1 expression and patient outcome suggesting a role of E2F1/DP1 in the pathogenesis of MM. To better understand how E2F1 and DP1 drive proliferation, we mapped the global occupancy of E2F1/DP1 in MM. Integration of E2F1 and DP1 genomic localization in MM reference epigenome revealed specific co-occupancy of the factors at promoters of active genes marked by H3K4me3 promoters, with a strong positive correlation between E2F and RNA Polymerase II (RNA Pol II) levels at transcription start sites. In contrast, active enhancers, as defined by promoter distal Mediator (MED1) peaks and marked by H3K27ac, showed virtually no E2F binding.

Using unbiased hierarchical clustering, we organized different factors/epigenetic modifications in MM1.S by spatial similarity of binding patterns. From this analysis, we identify two distinct active regulatory axes in MM1.S. The first comprised Mediator, P-TEFb (CDK9), RNA Pol II, BRD4, IRF4, and H3K27ac — factors/epigenetic modifications that are found at both promoters and enhancers. The second comprised MYC/MAX, E2F1/DP1, and the transcription start site specific histone modification H3K4me3. Although MYC is also found at enhancers in MM1.S, these data are consistent with MYC and E2F collaborative regulation of E2F target genes at promoters and suggest a role for E2F in regulation of MM transcription separate from BETs. Functional analysis of genes governed by superenhancers (SE) or E2F revealed divergent functionality, with SE-associated genes involved in cell signaling, apoptosis, and hypoxia, and E2F associated genes involved in cell cycle regulation and canonical E2F/MYC regulation. Therefore, our global chromatin analysis reveals two distinct regulatory axes for E2F and BETs in the MM epigenome, with E2F predominantly localized to active gene promoters and BETs disproportionately found at super-enhancers.

We hypothesized that the presence of BETs and E2F in distinct regulatory axes divides active genes in MM into those that can be selectively influenced by BET inhibition or E2F perturbation, but not both. In line with this we have observed that dual E2F and BET inhibition is synergistic for MM cell growth. Moreover, we observed that MM cell lines absent of IgH/MYC translocations fail to place MYC directly under SE control, and consequently BET inhibition fails to rapidly downregulate MYC or E2F transcriptional activity. In these non MYC translocated lines, direct E2F inhibition either by genetic approach or by a dimerization inhibiting stapled peptide is synergistic with BET inhibition.

In conclusions, our results implicate E2F and superenhancer-driven transcriptional programs as dependencies in MM and suggest an unexplored collaboration between Myc, E2F and BET in maintenance of MM that can be synergistically targeted.