BET Inhibitors Potentiate Chemotherapy and Killing of SPOP-Mutant Colon Cancer Cells via Induction of DR5

Colorectal cancer is the second leading cause of cancer-related deaths in the United States (1). Patients with colorectal cancer are typically treated with conventional chemotherapeutic drugs including 5-fluorouracil (5-FU), oxaliplatin, and irinotecan (2), which have limited specificity and efficacy in advanced colorectal cancers (3, 4). Patients with colorectal cancer are also treated with targeted drugs such as the anti-EGFR antibodies, cetuximab and panitumumab, and the multikinase inhibitor, regorafenib (2). Nonetheless, the tumors treated with targeted drugs almost invariably develop resistance shortly after initial therapy (5). The benefit of anti-PD-1 immunotherapy is limited in 10%–15% of colorectal cancers deficient in DNA mismatch repair (MMR; ref. 6). There is a critical need for developing novel therapies against colorectal cancers.

Epigenetic alterations such as DNA hypermethylation and histone modifications play an important role in colorectal cancer progression (7, 8). The bromodomain and extraterminal domain (BET) family proteins, including BRD2, BRD3, BRD4, and BRDT, are epigenetic readers of acetylated histone marks (9, 10). They function as transcriptional coactivators by recognizing acetylated lysine residues through their bromodomains to regulate transcriptional elongation by RNA polymerase II (10). BET family proteins are frequently assembled into super-enhancers and control the expression of oncogenes such as c-Myc (11). Recent studies show that BRD4, the most extensively characterized BET protein, is critical for colorectal cancer cell proliferation (12, 13).

Targeting the BET family proteins has recently emerged as a promising anticancer approach (14). BET inhibitors (BETi), alone or in combination with other anticancer agents, have exhibited efficacy in a variety of tumors (14). Recent studies revealed that mutations in speckle-type POZ protein (SPOP), encoding a component of the E3 ubiquitin ligase of BET proteins, dictate BETi sensitivity in some cancer cells (15–17). Although SPOP-inactivating mutations in prostate cancer cells stabilize BET proteins and confer resistance to BETi, SPOP-activating mutations sensitize endometrial cancer cells to BETi (15–17). The molecular basis of resistance and differential BETi sensitivity caused by SPOP mutations is unclear.

Cell killing is a key mechanism of anticancer therapies (18). BETi sensitivity was shown to be mediated by the induction of apoptosis (19, 20), which is regulated by the extrinsic (death receptor) and intrinsic (mitochondrial) pathways. The extrinsic pathway is engaged upon activation of the TNF family receptors such as death receptor 5 (DR5; TRAILR2; TNFRSF10B) and DR4, which further recruit other proteins to activate caspase-8 and downstream caspases (21). DR5 can also be induced by p53 upon DNA damage (22), or by C/EBP homologous protein (CHOP) in response to endoplasmic reticulum (ER) stress (23). The mitochondrial pathway is activated by the Bcl-2 family members via mitochondrial dysfunction (24, 25). Relative to the mitochondrial pathway, the role of the extrinsic pathway in anticancer therapies is less well-characterized.

In this study, we investigated the molecular basis of differential response to BETi in colorectal cancer cells. Our results suggest that DR5-mediated apoptosis plays a critical role in chemosensitization by BETi in colorectal cancer cells, and is responsible for increased BETi sensitivity in colorectal cancer cells with SPOP-activating mutations.

Human colorectal cancer cell lines, including HCT116, HT29, RKO, SW48, Lim1215, and NCI-H508, were purchased from the ATCC and cultured in McCoy 5A modified media (Invitrogen). SNU-407 was purchased from AddexBio and cultured in RPMI1640 (Invitrogen). DR5-knockout (KO) HCT116 cells were described previously (26). DR5-KO RKO cells were generated by CRISPR/Cas9 using a single guide RNA sequence 5′-CGCGGCGACAACGAGCACAA-3′ as described previously (27). Cells were authenticated in 2018 by genotyping and analysis of protein expression by Western blotting, and routinely checked for Mycoplasma contamination by PCR. All cell lines were maintained at 37°C and 5% CO₂ atmosphere. Cell culture media were supplemented with 10% defined FBS (HyClone), 100 U/mL penicillin, and 100 μg/mL streptomycin (Invitrogen). For drug treatment, cells were plated in 12-well plates at 20%–30% density 24 hours before treatment. DMSO (Sigma) stocks of JQ1, I-BET151 (Apexbio), OTX015, I-BET762, 5-FU, and oxaliplatin (Sigma) were prepared and diluted in cell culture media before adding to cells.

Cells were transfected using Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions. pDONR223 vectors expressing wild-type (WT; plasmid #81856), E47A (plasmid #81645), and E50K (plasmid #81631) mutant SPOP were described previously (28), and obtained from Addgene. Small interfering RNA (siRNA) transfection was performed 24 hours before drug treatment using 200 pmol of control scrambled, CHOP (5′-GCACAGCUAGCUGAAGAGAdTdT-3′), DR5 (5′-AAGACCCUUGUGCUCGUUGUCdTdT-3′) (Dharmacon), BRD4 (sc-43639), or SPOP (sc-63056) siRNA (Santa Cruz Biotechnology).

Total RNA was prepared from HCT116 cells transfected with either control scrambled or BRD4 siRNA for 24 hours using the Quick-RNA Kit (Zymo Research) according to manufacturer's instructions. Library construction, RNA sequencing (RNA-seq), and data analysis were performed by Novogene using the Illumina HiSeq platform. Sample quality was assessed by HTSeq v0.6.1 to count the read numbers mapped of each gene. Data quality was ensured by the percentage of bases with a sequencing quality score above Q30. FPKM (fragments per kilobase of transcript per million mapped reads) of each gene was calculated on the basis of the length of a gene and read counts mapped to this gene. Differential expression analysis was performed using the DESeq R package (2_1.6.3).

Western blotting was performed as described previously (29) using antibodies listed in Supplementary Table S1.

Total RNA was isolated from cells using the Mini RNA Isolation II Kit (Zymo Research) according to the manufacturer's protocol. One microgram of total RNA was used to generate cDNA using the SuperScript II reverse transcriptase (Invitrogen). PCR was performed with the previously described cycle conditions (30) and primers (23), except for CHOP (5′-GGTCCTGTCTTCAGATGAAAATG-3′/5′-CAGCCAAGCCAGAGAAGCA-3′).

Cells seeded in 96-well plates at a density of 1 × 10⁴ cells/well were treated with different agents for 72 hours. Cell viability was evaluated by 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS) assay (Promega) according to the manufacturer's instructions. Chemiluminescence was measured using a Wallac Victor 1420 Multilabel Counter (Perkin Elmer). Each assay was conducted in triplicate and repeated three times.

pGL3-based DR5 luciferase reporter constructs containing WT or mutant CHOP-binding site were described previously (31, 32). To measure reporter activities, cells were transfected with WT or mutant DR5 reporter along with the transfection control β-galactosidase reporter, pCMVβ (Promega). Cell lysates were collected and luciferase activities were measured and normalized as described previously (33). All reporter experiments were performed in triplicate and repeated three times.

Chromatin immunoprecipitation (ChIP) was performed using the Chromatin Immunoprecipitation Assay Kit (EMD Millipore) according to the manufacturer's instructions. The precipitates were analyzed by PCR for DR5 promoter using the primer pair 5′-AGGTTAGTTCCGGTCCCTTC-3′/5′-CAACTGCAAATTCCACCACA-3′.

Apoptosis was measured by counting cells with condensed and fragmented nuclei after nuclear staining with Hoechst 33258 (Invitrogen) as described previously (33). At least 300 cells were analyzed for each group. Apoptosis was also analyzed by flow cytometry of cells stained with Annexin V/propidium iodide (PI; ref. 33). Viable cells in 12-well plates after drug treatment for 36 hours were visualized by staining with crystal violet (Sigma). Colony formation was assayed by plating equal numbers of treated cells in 12-well plates at appropriate dilutions, followed by visualization of colonies by crystal violet staining after 14 days, and counting of colony numbers.

All animal experiments were approved by the University of Pittsburgh Institutional Animal Care and Use Committee. Female 5- to 6-week-old Nu/Nu mice (Charles River) and 4- to 6-week-old NOD.Cg-Prkdc^scid Il2rg^tm1Wjl/SzJ (NSG) mice (Jackson Laboratory) were housed in a sterile environment with micro isolator cages and allowed access to water and chow ad libitum.

Cell line xenografts were established by subcutaneously injecting 4 × 10⁶ WT or DR5-KO HCT116 cells into both flanks of Nu/Nu mice. Tumors were allowed to grow for 7 days before treatment. Tumor-bearing mice were randomized into 4 groups and subjected to the following treatments for 3 weeks: (i) untreated; (ii) JQ1 (i.p.; 30 mg/kg daily); (iii) 5-FU (i.p.; 25 mg/kg every other day); and (iv) JQ1+5-FU. JQ1 (AstaTech) was dissolved in DMSO to 50 mg/mL and then diluted 1:10 in 10% hydroxypropyl-β-cyclodextrin (Sigma). Injectable 5-FU solution (APP Pharmaceuticals) was diluted with Hank's Balanced Salt Solution (Invitrogen). Six mice were used for each group.

Two patient-derived xenograft (PDX) models were established using treatment-naïve primary tumors resected from patients with newly diagnosed colorectal cancers under written informed consent and approved Institutional Review Board protocol. PDX1 was established from a KRAS-mutant (G13D), NRAS-mutant (G12D), and MMR-proficient tumor (T4N0M1) in the sigmoid colon of a 77-year-old male. PDX2 was established from an MMR-proficient tumor (T2N0) in the right colon of a 69-year-old female. Tumors were implanted subcutaneously and passaged as described previously (34). Briefly, deidentified patient colorectal cancer samples were delivered to the laboratory in 1 × Antibiotic/Antimycotic Solution (Invitrogen) within 4 hours after tumor resection. Tissues were cut into 25-mg pieces and directly implanted subcutaneously into both flanks of NSG mice. Tumors passaged and expanded for two generations (P2) in NSG mice were used for the described experiments. Treatment was performed for 20 days in the same way as described for the cell line xenografts, with each group including 4 mice.

Tumor growth was monitored by calipers, and tumor volumes were calculated according to the formula ½ × length × width². Ethical endpoint was defined as a time point when a tumor reached 1.5 cm or more in any dimension. Tumors were dissected and fixed in 10% formalin and embedded in paraffin. Terminal deoxynucleotidyl transferase mediated dUTP Nick End Labeling (TUNEL; EMD Millipore), active caspase-3 (Cell Signaling Technology), active caspase-8 (Cell Signaling Technology), and Ki67 (Dako) immunostaining was performed on 5 μm paraffin-embedded tumor sections as described previously (33), by using an AlexaFluor 488–conjugated secondary antibody (Invitrogen) for detection and 4′6-Diamidino-2-phenylindole (DAPI) for nuclear counter staining.

Statistical analysis was carried out using the GraphPad Prism VI software. For cell culture experiments, P values were calculated using the Student t tests. Means ± SD are reported in the figures. For animal experiments, P values were calculated by repeated measures (RM) ANOVA with Fisher LSD post hoc tests. Means ± SEM are reported in the figures. Differences were considered significant if P < 0.05.

To determine the effects of BET inhibition in colorectal cancer cells, RNA-seq was performed on HCT116 cells transfected with BRD4 or control scrambled siRNA. Analysis of differentially expressed genes revealed a marked induction of DR5 mRNA (FKPM 34.78 vs. 15.63; P < 0.01) and two isoforms of DR5 protein upon BRD4 knockdown (Fig. 1A). Treating HCT116 cells with different BETi, including JQ1, OTX015, I-BET151, and I-BET762, induced DR5 in a dose- and time-dependent manner (Fig. 1B and C). DR5 induction by BRD4 knockdown and BETi treatment was observed in other colorectal cancer cell lines with different colorectal cancer driver mutations (Supplementary Table S2), including those of KRAS, BRAF, PIK3CA, p53, SMAD4, SPOP, and FBW7, and associated with reduced c-Myc expression (Fig. 1D and E). The fold changes of both isoforms of DR5 protein were found to be higher than those of DR5 mRNA (Fig. 1A, C, and E), suggesting a more prominent role of DR5 protein induction in determining response to BETi. Furthermore, other members of the death receptor family, including DCR1, DCR2, TNFR1, and TNFR2, were upregulated after BRD4 knockdown or BETi treatment (Fig. 1A; Supplementary Fig. S1A). In contrast, the mRNA and protein expression of DR4 did not change significantly (Fig. 1A; Supplementary Fig. S1A–S1C). The Bcl-2 family members including PUMA, Bim, Bad, Bid, and Bax were either downregulated or unchanged, except for Noxa (Fig. 1A). These results indicate that BETi treatment activates DR5 and the death receptor pathway in colorectal cancer cells irrespective of common drivers, and prompted us to further investigate the functional role of DR5 and the death receptor pathway.

DR5 can be induced by p53 and other transcription factors in response to various stresses (22, 23). As expected, the induction of DR5 by BETi is p53-independent and unaffected in p53-deficient and -KO cells (Fig. 1D and E; Supplementary Fig. S1D). RNA-seq data revealed upregulation of several ER stress response genes including CHOP following BRD4 depletion (Supplementary Fig. S1E), which was confirmed by protein expression (Fig. 2A). Treating colorectal cancer cells with different BETi also induced CHOP mRNA and protein (Fig. 2B; Supplementary Fig. S1F), as well as other ER stress markers including BiP expression and eIF2α (Ser51) and PERK (Thr980) phosphorylation (Fig. 2B; ref. 35). Knockdown of CHOP by siRNA abrogated the induction of DR5 mRNA and protein by JQ1 (Fig. 2C). ChIP analysis showed enhanced binding of CHOP to the DR5 promoter upon JQ1 treatment (Fig. 2D). Furthermore, JQ1 exposure activated a DR5 promoter reporter containing a CHOP-binding site (31), but did not affect a similar reporter with mutations that alter the CHOP-binding site (Fig. 2E; ref. 32). Together, these results indicate that BETi treatment induces DR5 through ER stress/CHOP-mediated transcription.

BRD4 knockdown or inhibition, while ineffective alone (Supplementary Fig. S2A and S2B), markedly suppressed colorectal cancer cell growth when combined with 5-FU or oxaliplatin (Fig. 3A and B; Supplementary Fig. S2B and S2C). The combinations induced marked apoptosis in HCT116 cells, as shown by reduced cell viability, increased nuclear fragmentation, caspases 3 and 8 activation, and Annexin V staining, which could be blocked by the pan-caspase inhibitor, z-VAD-fmk (Supplementary Fig. S2C–S2G). Enhanced apoptosis by the 5-FU/BETi combination was also observed in RKO cells (Supplementary Fig. S2H). Using isogenic DR5-KO HCT116 cells generated by homologous recombination (26), we found that DR5-KO markedly suppressed all signs of apoptosis, including cell viability loss, nuclear fragmentation, and caspases 3 and 8 cleavage, in response to BRD4 knockdown or JQ1 combined with 5-FU or oxaliplatin (Fig. 3C–E; Supplementary Fig. S3A–S3C), and significantly improved the long-term clonogenic survival (Fig. 3F). Furthermore, KO of DR5 by CRISPR/Cas9 in RKO cells also abolished apoptosis induced by BRD4 knockdown or JQ1 in combination with 5-FU or oxaliplatin (Supplementary Fig. S3D and S3E). Together, these results demonstrate that the chemosensitization effects of BETi are dependent on DR5-mediated apoptosis in colorectal cancer cells.

Recent studies showed that the cancer cells harboring activating mutations in SPOP, a substrate-binding subunit of E3 ubiquitin ligase of BRD4, are exquisitely sensitive to BETi due to reduced BRD4 expression (16). Upon analysis of the Catalogue Of Somatic Mutations In Cancer (COSMIC) database (Supplementary Table S2), we identified another colorectal cancer line, SNU-407, which contains a SPOP mutation (G75E), in addition to NCI-H508 (SPOP-E47K; ref. 16). SNU-407 and NCI-H508 cells were more sensitive to BETi than SPOP-WT colorectal cancer cells lines, with IC₅₀ values of JQ1 and OXT015 ranging from 0.69 to 1.05 μmol/L, compared with 4.26 to 12.02 μmol/L in HCT116, HT29, RKO, SW48, and Lim1205 cells (Fig. 4A). JQ1 at 0.5 μmol/L potently suppressed colony formation of SNU-407 and NCI-H508 cells but barely affected that of HCT116 cells (Fig. 4B). Compared with SPOP-WT colorectal cancer cells, SNU-407 and NCI-H508 cells expressed lower levels of BRD4 (Fig. 4C). Interestingly, we detected strong induction of both DR5 isoforms in SNU-407 and NCI-H508 cells by JQ1 at only 0.5 μmol/L, in contrast to an obvious DR5 induction with JQ1 at 4 μmol/L in the SPOP-WT cells (Fig. 4D). Consistent with the strong DR5 induction, robust activation of caspases 3 and 8 was found in JQ1-treated SNU-407 and NCI-H508 cells (Fig. 4E). Knockdown of SPOP in SNU-407 and NCI-H508 cells resulted in BRD4 accumulation, abrogated JQ1-induced DR5 expression, caspase-3 and 8 activation (Fig. 4E), nuclear fragmentation (Fig. 4F), and reduced JQ1 sensitivity (Fig. 4G), suggesting that increased BETi sensitivity in these cells is attributed to SPOP mutations and enhanced apoptosis induction. Consistent with the on-target CHOP and DR5 induction in SPOP-WT cells, BRD4 knockdown also induced DR5 and CHOP in SNU-407 and NCI-H508 cells (Supplementary Fig. S4A and S4B).

Knockdown of DR5 by siRNA suppressed cell viability loss, nuclear fragmentation, and caspase activation induced by JQ1 and OTX015 in both SNU-407 and NCI-H508 cells (Fig. 5A–C). Transfecting WT HCT116 cells with the activating mutants, including E47A or E50K (16), but not WT SPOP, increased JQ1 and OTX015 sensitivity (Fig. 5D–F; Supplementary Fig. S5A), which was accompanied by enhanced BRD4 depletion (Fig. 5D), apoptosis induction (Fig. 5G; Supplementary Fig. S5B), and caspase activation (Fig. 5D; Supplementary Fig. S5C). In contrast, these effects of SPOP mutants were almost completely absent in DR5-KO HCT116 cells (Fig. 5E–H; Supplementary Fig. S5A–S5C). The DR5-dependent effects of SPOP mutants on BETi sensitivity were also observed in DR5-KO RKO cells transfected and treated in a similar fashion (Supplementary Fig. S5D–S5F). Together, these results indicate that increased BETi sensitivity in SPOP-mutant colorectal cancer cells is mediated by enhanced DR5 induction.

To determine the role of DR5 in BETi chemosensitization in vivo, athymic nude mice with xenograft tumors established from WT and DR5-KO HCT116 cells were treated with 5-FU at a low, ineffective dose (i.p.; 25 mg/kg every other day), with or without JQ1 (i.p.; 30 mg/kg daily; Fig. 6A). The combination treatment, but not JQ1 or 5-FU alone, effectively suppressed the growth of WT HCT116 tumors (Fig. 6A and B). Compared with 5-FU or JQ1 alone, the combination treatment markedly enhanced DR5 induction (Fig. 6C) and activation of caspases 3 and 8 (Fig. 6D–F). In contrast, the antitumor and apoptotic effects of the JQ1/5-FU combination treatment were largely suppressed in the DR5-KO tumors (Fig. 6A–F).

PDX models better recapitulate heterogeneity, histology, and molecular alterations of patient tumors (36). We further explored the efficacy of the JQ1/5-FU combination in two colorectal cancer PDX models. The JQ1/5-FU combination, but not JQ1 (30 mg/kg daily) or 5-FU at a low dose (25 mg/kg every other day) alone, markedly suppressed PDX tumor growth (Fig. 7A–D; Supplementary Fig. S6A), without a significant effect on body weight (Supplementary Fig. S6B). Compared with JQ1 or 5-FU alone, the combination treatment also induced higher levels of DR5 protein (Fig. 7E), tumor cell loss (Fig. 7F), suppression of cell proliferation detected by Ki67 staining (Fig. 7G), and apoptosis as analyzed by active caspase-3 and TUNEL staining (Fig. 7H; Supplementary Fig. S6C). Collectively, our in vitro and in vivo data demonstrate a critical role of DR5 induction in mediating the chemosensitization and antitumor effects of BETi in colorectal cancer cells with different genetic backgrounds.

We found that BET depletion or inhibition triggers ER stress and CHOP-mediated DR5 induction. Perturbations in ER function and ER stress can occur in normal cells in response to misfolded proteins or other stimuli (35). If unresolved, ER stress leads to eIF2α phosphorylation, CHOP and DR5 induction, and subsequent apoptotic death (37). This pathway can be engaged in cancer cells upon inhibition of addicted oncogenic signaling (38). ER stress and DR5 induction by BETi may result from downregulation of c-Myc, a key oncogenic driver activated by the defective APC/β-catenin pathway and aberrant Wnt signaling in colorectal cancer cells (7). Sustained c-Myc expression in tumor cells often relies on BET proteins, and is subject to inhibition by BETi (11). JQ1 has been shown to induce apoptosis in Eμ-Myc lymphoma cells (39). The extent of c-Myc depletion was found to correlate with BETi sensitivity in colorectal cancer cells (40). Although directly targeting c-Myc is challenging, inhibiting its expression by BETi or other agents is an attractive approach for targeting Myc-driven cancers including colorectal cancers. In line with our findings, other studies have also shown apoptosis induction as a key effect of BETi (19, 20).

A combination of BETi with 5-FU or oxaliplatin markedly improved therapeutic efficacy in colorectal cancer cells, and may offer hope to patients with chemotherapy-refractory and metastatic colorectal cancers (3). The improved therapeutic activity is dependent on enhanced DR5 induction, likely involving both p53-dependent and -independent mechanisms (22), leading to stronger apoptotic signaling via both the intrinsic and extrinsic pathways. The acute and potent induction of apoptosis by the combination treatment may provide a therapeutic window necessary for BETi dosing to minimize normal tissue toxicity caused by sustained BET inhibition, such as intestinal stem cell depletion and lymphoid and hematopoietic abnormalities (41, 42). Our results showed that JQ1 at 30 mg/kg daily combined with 5-FU was well-tolerated and efficacious (Fig. 6 and 7). BETi were previously shown to synergize with other classes of anticancer agents including targeted drugs and immune checkpoint blockers (14). In those cases, BETi inhibited the protective feedback mechanisms mediated by BRD-dependent upregulation of kinase signaling (14). Enhanced DR5 induction may contribute to improved therapeutic response via cell killing and subsequent effects on tumor microenvironment.

Several recent studies showed that mutations in SPOP, which encodes a substrate-binding subunit of a cullin-3–based E3 ubiquitin ligase, dictate BETi sensitivity in some cancer cells (15–17). Depending on the tumor types and specific residues affected, SPOP mutations could be activating or inactivating, generating opposite effects on BETi sensitivity (15–17). Our results suggest that enhanced DR5 induction underlies the increased BETi sensitivity in SPOP-mutant colorectal cancer cells. The mutations we analyzed, including E47K in NCI-H508 and G75E in SNU-407, are located in the vicinity of other activating mutations, such as E47A, E50K, E78K, and S80R, which were mostly found in endometrial cancer cells (15–17). Our mechanistic data show that colorectal cancer cells expressing a SPOP-activating mutation have reduced BRD4 expression (Fig. 4C), and are more prone to DR5 induction in response to BETi (Fig. 4D). Analyzing additional SPOP mutations in different cancer types will be necessary to determine whether DR5 induction is broadly involved in modulating BETi sensitivity.

BETi have been evaluated in a number of clinical trials and shown single-agent efficacy against some hematologic malignancies and NUT carcinomas (14). Biomarkers for predicting BETi sensitivity are essential for developing personalized and precision therapies. In addition to SPOP mutations, colorectal cancer cells with a CpG island methylator phenotype (CIMP) are known to be sensitive to JQ1 and c-Myc repression (12). BETi resistance in triple-negative breast cancer cells involves hyperphosphorylation and bromodomain-independent activities of BRD4 (43). Our findings suggest that ER stress and DR5 induction might be useful molecular markers for predicting BETi sensitivity. Several genes in the death receptor pathway, such as caspase-8, DCR1, and DCR3, are inactivated or downregulated in some colorectal cancers (5). Our next step is to determine whether DR5 or other genes in the death receptor pathway are aberrantly regulated in cells with acquired resistance to BETi due to changes in BRD4 and/or c-Myc.

In conclusion, we demonstrated a critical role of DR5-mediated apoptosis in determining the therapeutic response to BETi alone and BETi-mediated chemosensitization in colorectal cancer cells in vitro and in vivo. Our results provide a better understanding of the proapoptotic action of BETi, particularly the heightened sensitivity of SPOP-mutant colorectal cancers.

No potential conflicts of interest were disclosed.

Conception and design: X. Tan, L. Zhang

Development of methodology: R. Fletcher, L. Zhang

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): X. Tan, J. Tong, Y.-J. Wang, R. Fletcher, J. Yu, L. Zhang, R.E. Schoen

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): X. Tan, J. Tong, Y.-J. Wang, L. Zhang

Writing, review, and/or revision of the manuscript: X. Tan, Y.-J. Wang, R. Fletcher, J. Yu, L. Shen, L. Zhang, R.E. Schoen

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): J. Yu, L. Shen, L. Zhang

Study supervision: L. Shen, L. Zhang

We thank Dr. Darleny Y. Lizardo and other lab members for critical reading, Dr. Shi-Yong Sun at Emory University (Atlanta, GA) for providing the DR5 reporter constructs, and Dr. Zhanzhan Li at Xiangya Hospital of Central South University (Changsha, Hunan, China) for help on statistical analyses of animal experiments. This work was supported by the NIH grants (R01CA172136, R01CA213028, and R01CA217141 to L. Zhang; U19AI068021 and R01CA215481 to J. Yu; and U01CA152753 to R.E. Schoen), and UPMC Hillman Cancer Center institutional funds (to L. Zhang). This project used the UPMC Hillman Cancer Center Animal Facility, Cytometry Facility, and Tissue and Research Pathology Services, which are supported, in part, by P30CA047904.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.