Inhibition of BET bromodomain targets genetically diverse glioblastoma

Abstract

Purpose: Glioblastoma is refractory to conventional therapies. The bromodomain and extraterminal domain (BET) proteins are epigenetic readers that selectively bind to acetylated lysine residues on histone tails. These proteins recently emerged as important therapeutic targets in NUT midline carcinoma and several types of hematopoietic cancers. In this study, the therapeutic potential of a novel BET bromodomain inhibitor, JQ1, was assessed in a panel of genetically heterogeneous glioblastoma samples.

Experimental design: The antineoplastic effects of JQ1 were shown using ex vivo cultures derived from primary glioblastoma xenograft lines and surgical specimens of different genetic background. The in vivo efficacy was assessed in orthotopic glioblastoma tumors.

Results: We showed that JQ1 induced marked G1 cell-cycle arrest and apoptosis, which was phenocopied by knockdown of individual BET family members. JQ1 treatment resulted in significant changes in expression of genes that play important roles in glioblastoma such as c-Myc, p21(CIP1/WAF1), hTERT, Bcl-2, and Bcl-xL. Unlike the observations in some hematopoietic cancer cell lines, exogenous c-Myc did not significantly protect glioblastoma cells against JQ1. In contrast, ectopically expressed Bcl-xL partially rescued cells from JQ1-induced apoptosis, and knockdown of p21(CIP1/WAF1) attenuated JQ1-induced cell-cycle arrest. Cells genetically engineered for Akt hyperactivation or p53/Rb inactivation did not compromise JQ1 efficacy, suggesting that these frequently mutated signaling pathways may not confer resistance to JQ1. Furthermore, JQ1 significantly repressed growth of orthotopic glioblastoma tumors.

Conclusion: Our results suggest potentially broad therapeutic use of BET bromodomain inhibitors for treating genetically diverse glioblastoma tumors.

Figure 1

JQ1 reduces proliferation and survival of glioblastoma stem cells and matched non-stem cancer cells. A, CD133+ and CD133− cells derived from T4302 primary glioblastoma xenograft tumor were exposed to JQ1 with a 2-fold serial dilution starting at 10 μmol/L for 17 dilutions in total. Cell viability was determined 5 days posttreatment and normalized to the vehicle-treated group (dimethyl sulfoxide). Data points are presented as mean ± SD (n = 3) unless otherwise indicated. B, neurosphere formation of T4302 CD133+ cells were measured in the presence of JQ1 at indicated concentrations. *, P < 0.05 by Student t test. C, T4302 CD133+ cells were exposed to vehicle or 1 μmol/L JQ1 for 48 hours. Cell-cycle distribution was determined by the ModFit LT software. Data are representative of 3 independent experiments. D, presence of the Annexin V apoptotic marker was determined 72 hours after treatment. The percentage of cells stained positively for Annexin V but negatively for propidium iodine (PI) is indicated. Data are representative of 3 independent experiments.

Figure 2

Knockdown of BET proteins leads to antiproliferative and antisurvival effects resembling JQ1 treatment. A, T4302 CD133+ cells were infected with lentivirus directing expression of nontargeting shRNA (NT) or shRNA sequences specific to BRD2, BRD3 or BRD4 (clone 2). After selection with 1 μg/mL puromycin for 2 days, relative cell growth was determined for the following 5 days. B, three days after puromycin selection, relative caspase-3/7 activity was determined by normalizing the activity of caspase-3/7 to the corresponding cell titers. *, P < 0.001 by Student t test.

Figure 3

c-Myc plays a relatively minor role in mediating the antineoplastic effects of JQ1. A, qRT-PCR of c-Myc, p21, hTERT, Bcl-2, and Bcl-xL in T4302 CD133+ cells treated with 2.5 μmol/L JQ1 for 24 hours. β-Actin was used as loading control for qRT-PCR or immunoblotting throughout this study. B, T4302 CD133+ cells were infected with control lentivirus or virus directing expression of c-Myc under control of a CMV promoter. The protein levels of c-Myc were determined 24 hours after JQ1 treatment. C, qRT-PCR of selected JQ1-responsive genes in control or exogenous c-Myc-expressing T4302 CD133+ cells 24 hours after treatment with JQ1 at 2.5 μmol/L. *, P < 0.05; #, P < 0.001 by Student t test. D, dose–response curves of JQ1 in control or exogenous c-Myc–expressing T4302 CD133+ cells determined as described in Fig. 1A.

Figure 4

Hyperactivation of Akt or knockdown of p53 or Rb does not compromise the efficacy of JQ1. T4302 CD133+ cells were infected with control lentivirus or virus directing expression of myristoylated Akt1 (A and B), p53-specific shRNA (C), or Rb-specific shRNA (D). A, immunoblotting was conducted 24 hours after JQ1 treatment. B–D, dose–response curves of JQ1 in control or genetically engineered T4302 CD133+ cells determined as described in Fig. 1A.

Figure 5

JQ1 induces broad antineoplastic effects in genetically heterogeneous glioblastoma samples partially via p21 upregulation and Bcl-xL downregulation. A, immunoblotting of indicated proteins in selected glioblastoma spheroid cultures treated with 2.5 μmol/L JQ1 for 24 hours. B, T4302 CD133+ cells were infected with lentivirus directing expression of nontargeting shRNA (NT) or shRNA sequences specific to p21. Following puromycin selection, cells were exposed to JQ1 at indicated concentrations for 48 hours. Cell-cycle distribution was determined by the ModFit LT software. C, three days after treatment with JQ1 at indicated concentrations, relative caspase activity was determined in control or exogenous Bcl-xL–expressing T4302 CD133+ cells as described in Fig. 2B. *, P < 0.01 by Student t test. D, dose–response curves of JQ1 in control or exogenous Bcl-xL–expressing T4302 CD133+ cells.

Figure 6

JQ1 represses orthotopic glioblastoma progression. A, expression of indicated target genes was determined by qRT-PCR in samples extracted from intracranial T4302 tumors following a single dose of JQ1 at 100 mg/kg at varying time points as indicated. *, P < 0.05 by Student t test. Mice bearing intracranial T4302 (B) or T4597 (C) tumors were treated intraperitoneally with 50 mg/kg JQ1, twice daily, from day 11 after tumor implantation to day 30 (T4302) or day 45 (T4597). The P values were calculated by the log-rank test. D, representative bioluminescence images of T4597 intracranial tumors 3 weeks after JQ1 treatment.