Genome-Wide Mapping Reveals BRD4 in Regulation of Tumor-Driver Genes in Cutaneous T-Cell Lymphoma

Cutaneous T-cell lymphoma (CTCL) is a non-HodgkinÕs lymphoma of skin homing malignant CD4+ T cells.  Early treatment of patients with skin-directed therapies often yields successful short-term outcomes; however, survival of patients with late-stage CTCL is extremely poor, highlighting the need to identify novel therapies to inhibit key oncogenic processes. Although the underlying factors driving CTCL pathogenesis are poorly understood, recent findings from our lab highlight the critical role of epigenetic dysregulation in its development (Mishra et al., Blood, 122:1826, 2013). While many studies have been carried out on the role of epigenetic writers (histone acetyltransferases) and erasers (histone deacetylases) in CTCL gene regulation, little is known about the epigenetic readers (bromodomains; tandem PHD fingers; Pleckstrin homology domains) that coordinate transcriptional reprogramming. Previous studies have identified bromodomain-containing protein-4 (BRD4) as a super-enhancer transcriptional regulator that interacts with pairs of acetylated lysine residues on histone H3 to upregulate oncogenic transcription in multiple myeloma and diffuse large B-cell lymphoma  (Loven, J. et al. Cell 153, 2013; Chapuy, B. et al. Cancer Cell 24, 2013).  We analyzed BRD4 occupancy patterns in the CTCL genome and characterized downstream oncogenic pathways. Additionally, we investigated the effect of treatment with a BRD4 inhibitor, JQ1, on CTCL in vitro and in vivo. In order to gain insights into the role of BRD4 in CTCL, we generated a genome-wide map of BRD4 occupancy in CD4+ T-cells from CTCL patients and normal donors using ÒChIP-seqÓ, which combines chromatin immunoprecipitation (ChIP) with sequencing of BRD4-associated DNA. The analysis revealed an increased occupancy of BRD4 in patientÕs CD4+ T-cells compared to normal donor CD4+ T-cells at putative active regions, promoter active regions, distal active regions, and super-enhancers (Figure 1).  The abnormal binding patterns in patients were reversed within 24 hours of treatment with JQ1 (Figure 1). PatientÕs cells treated with 1.0 μM JQ1 showed reduced cell viability (mean ± SEM of relative % viability in JQ1 vs control treated cells: 39.67 ± 1.838 vs 100.0 ± 2.335, n=4 each).  Furthermore, we saw a similar effect of JQ1 in all the patient-derived CTCL cell lines tested (HH, Hut78, Hut102, SeAx, and MyLa) with EC50 concentrations ranging from 0.167 to 21.0 μM. Since both primary CD4+ T-cells from CTCL patients and CTCL cell lines displayed sensitivity to the cytotoxic effect of JQ1, we performed subsequent functional assays on Hut78 cell line. Propidium iodide/Annexin staining of JQ1-treated Hut78 cells (1.0 μM for 48 hours) revealed an increased frequency of cells in the G1 phase, indicating significant disruption of cell cycle progression upon BRD4 inhibition. Furthermore, considering that a large portion of BRD4 binding was detected in promoter and super-enhancer regions, we speculated that BRD4 regulates tumor driver genes. When examining BRD4 occupancy, we found that BRD4 binding at NOTCH1, and MYC was higher in patientsÕ vs normal donorsÕ CD4+ T-cells. Similar differential increase in BRD4 occupancy was also noted at the DNA-binding factor gene RBPJ, a known NOTCH1 activator and T-cell oncogenic co-factor. In order to determine whether BRD4 occupancy at these genes exerts transcriptional regulation, we examined expression levels of RBPJ, NOTCH1, and MYC in JQ1 treated cells. Semi-quantitative RT-PCR analysis on HuT78 cells treated with JQ1 showed significant reduction in RBPJ (mean ± SEM of relative expression in Control vs JQ1 treatment= 100.5 ± 07.086 vs. 35.39 ± 0.7832, n=3, P=0.0008), NOTCH1 (mean ± SEM of relative expression in Control vs JQ1 treatment= 100.0 ± 0.95 vs 0.61.77 ± 02.844, n=3, P=0.0002), and MYC (mean ± SEM of relative expression in Control vs JQ1 treatment= 100.1 ± 03.019 vs. 57.97 ± 01.997, n=3, P=0.0003). Additionally, in vivo treatment of CTCL mice with JQ1 (50mg/kg, 5 days/week for 4-weeks) halted as well as regressed progression of CTCL compared to placebo treated controls. Together, these findings demonstrate that BRD4 regulates the expression of the oncogenic drivers in CTCL, and that bromodomain inhibitors offer the opportunity to interrogate the mechanisms of BRD4-mediated T-cell oncogenesis in the CTCL mouse model and design new anti-cancer therapies in CTCL patients.