Bromodomain-containing protein 4 silencing by microRNA-765 produces anti-ovarian cancer cell activity

Bromodomain-containing protein 4 (BRD4) overexpression promotes ovarian cancer progression, and represents an important therapeutic oncotarget. This current study identified microRNA-765 (miR-765) as a novel BRD4-targeting miRNA. We showed that miR-765 directly associated with and silenced BRD4. In primary ovarian cancer cells and established cell lines (SKOV3 and CaOV3), ectopic overexpression of miR-765 inhibited cancer cell proliferation, migration and invasion, and induced apoptosis activation. In contrast, miR-765 inhibition by its anti-sense induced BRD4 upregulation to promote ovarian cancer cell proliferation, migration and invasion. Significantly, miR-765 overexpression-induced anti-ovarian cancer cell activity was largely attenuated by restoring BRD4 expression through an UTR-null BRD4 construct. Moreover, CRISPR/Cas9-induced BRD4 knockout (KO)inhibited proliferation and activated apoptosis in ovarian cancer cells. BRD4 KO in ovarian cancer cells abolished the functional impact of miR-765. miR-765 expression levels were downregulated in human ovarian cancer tissues and cells, correlating with the upregulation of BRD4 mRNA. Collectively, BRD4 silencing by miR-765produces significant anti-ovarian cancer cell activity. miR-765 could be further tested for its anti-ovarian cancer potential.


INTRODUCTION
Ovarian cancer is the seventh most common cancer and the eighth most common cause of cancer-related mortalities among women [1,2]. It is estimated that each year there are over 1.2 million new cases of ovarian cancers and 160,000 deaths around the world [1,2]. The majority of ovarian cancer patients are diagnosed at the late stages, possibly due to a lack of typical clinical symptoms or early screening methods [3][4][5]. The current therapies for this devastating disease are surgery and/or platinum-based chemotherapy [3][4][5]. As a result of profound resistance to current available therapies, studies testing agents targeting novel cell signaling pathways and oncogenic genes required for ovarian cancer tumorigenesis and progression are necessary [3][4][5].
To further support a direct binding between miR-765 and BRD4 mRNA, two mutant miR-765 mimics, namely "Mu1" and "Mu2", were generated. The two mutants contain mutations at proposed bindings sites to BRD4 3'-UTR ( Figure 1I). The wild-type (WT) or the mutant mimics were separately transfected to pOC-1 cells. Only WT miR-765 mimic resulted in significant reductions in BRD4 3'-UTR luciferase reporter activity ( Figure 1J) as well as BRD4 mRNA expression ( Figure 1K), while the two mutants were completely ineffective ( Figure 1J, 1K). In other ovarian cancer cells, including pOC-2 primary cells and established cell lines (SKOV3 and CaOV3), stable transfection of lv-pre-miR-765 similarly resulted in robust miR-765 upregulation ( Figure 1L) and BRD4 mRNA downregulation ( Figure 1M). These results together indicated that miR-765 directly associates with and silences BRD4 in ovarian cancer cells.

Ectopic overexpression of miR-765 induces significant anti-ovarian cancer cell activity
Results in Figure 1 suggested that miR-765 is a potential BRD4-targeting miRNA. Since BRD4 is an important therapeutic target of ovarian cancer, we next tested whether miR-765 could alter ovarian cancer cell functions. Cell growth curve results in Figure 2A demonstrated that the growth of miR-765-overexpressed pOC-1 cells, lv-pre-miR-765-Stb-L1/L2 (see Figure 1), was significantly slower than the control pOC-1 cells. When testing cell viability using CCK-8 assays, we demonstrated that overexpression of miR-765 inhibited the viability of pOC-1 cells ( Figure 2B). Moreover, colony formation assay results in Figure 2C showed that the number of viable pOC-1 cell colonies was decreased in lv-pre-miR-765-expressing pOC-1 cells. These results supported the anti-survival activity by miR-765 overexpression.

AGING
EdU incorporation was tested to quantitatively measure cell proliferation. As shown, in pOC-1 cells, the EdUpositive nuclei ratio was largely decreased following miR-765 overexpression ( Figure 2D). To test pOC-1 cell migration in vitro, "Transwell" assays were performed. Results showed that ectopic miR-765 overexpression inhibited pOC-1 cell migration. As the number of migrated cells was significantly decreased in miR-765overexpressed pOC-1 cells ( Figure 2E). Furthermore, pOC-1 cell invasion, tested by "Matrigel Transwell" assays, was largely inhibited as well ( Figure 2F). These results showed that ectopic overexpression of miR-765 inhibited pOC-1 cell growth, proliferation, migration and invasion.
Next we tested the potential effect of miR-765 overexpression on oncogenic behaviors in other ovarian cancer cells, including pOC-2 primary cells and established cell lines (SKOV3 and CaOV3). EdU staining assay results in Figure 2G indicated that miR-765 overexpression by lv-pre-miR-765 (see Figure 1) potently inhibited proliferation of the primary and established ovarian cancer cells. Furthermore, CCK-8 viability was decreased in lv-pre-miR-765-expressing ovarian cancer cells ( Figure 2H). In pOC-2 cells and established cell lines, ectopic overexpression of miR-765 largely inhibited cell migration and invasion, tested by "Transwell" ( Figure 2I) by "Matrigel Transwell" ( Figure 2J) assays, respectively. These results further were transduced with the lentiviral construct encoding pre-miR-765 sequence (lv-pre-miR-765) or scramble non-sense miRNA (lv-miRC). After selection by puromycin stable cells were established, expression of miR-765 and listed genes was tested by qPCR and Western blotting assays (D, F-H, L, M), and results quantified; the BRD4 3'-UTR luciferase reporter activity was tested as well (E). The pOC-1 cells were transfected with the wild-type (WT) or the mutant miR-765 mimic (sequences listed in I), at 500 nM each for 36h, the relative BRD4 UTR luciferase reporter activity (J) and BRD4 mRNA expression (K) were shown. "Pare" stands for the parental control cells. "miRC" stands for microRNA control mimic. For each assay, n=5 (five replicate well/dishes). Data were presented as mean ± standard deviation (SD). * p < 0.05 vs. "lv-miRC"/"miRC" cells. Experiments in this figure were repeated five times with similar results obtained. , as well as cell migration ("Transwell" assays, E, I) and invasion ("Matrigel Transwell" assays, F, J) were tested, and results quantified. "Pare" stands for the parental control cells. For each assay, n=5 (five replicate well/dishes). Data were presented as mean ± standard deviation (SD). * p < 0.05 vs. "lv-miRC" cells. Experiments in this figure were repeated five times with similar results obtained. Scale bar=100 μm (D-F).

AGING
AGING supported that miR-765 overexpression produced significant anti-ovarian cancer cell activity. As expected, the non-sense miRNA control (lv-miRC) did not alter ovarian cancer cell behaviors (Figure 2A-2J).

Ectopic overexpression of miR-765 induces apoptosis activation in ovarian cancer cells
BRD4 inhibition or silencing should be capable of inducing apoptosis in ovarian cancer cells [13,14]. Results in Figure 2 showed that ectopic overexpression of miR-765 inhibited ovarian cancer cell viability and proliferation. We therefore tested whether miR-765 could induce apoptosis activation. In miR-765-overexpressed pOC-1 cells (lv-pre-miR-765-Stb-L1/L2) the caspase-3 activity was significantly higher than that in control pOC-1 cells ( Figure 3A). In addition Histone-bound DNA contents were significantly increased in miR-765overexpressed pOC-1 cells, indicating accumulation of DNA breaks ( Figure 3B).
In apoptotic cells with mitochondrial depolarization, JC-1 fluorescence dye can aggregate to decompose into a monomer form, and the emitted fluorescence shall change from red to green. As demonstrated, there was accumulation of JC-1 green monomers in lv-pre-miR-765-expressing pOC-1 cells, indicating mitochondrial depolarization ( Figure 3C). To confirm cell apoptosis activation, we showed that the ratio of TUNEL-positive nuclei was significantly increased in miR-765overexpressed pOC-1 cells ( Figure 3D).

AGING
in ovarian cancer cells. The non-sense miRNA control, lv-miRC, failed to induce significant apoptosis activation ( Figure 3A-3G).

miR-765 is downregulated in ovarian cancer tissues and cells
At last, we tested expression of miR-765 in human ovarian cancer tissues. A set of 10 (n=10) different ovarian cancer tissues ("T") and surrounding paracancerous normal tissues ("N") were obtained and tested. As shown, miR-765 expression levels in the cancer tissues were significantly lower than those in the normal tissues ( Figure 6A). In the contrast, BRD4 mRNA expression was significantly elevated in cancer tissues ( Figure 6B). When compared to miR-765-BRD4 mRNA expression in primary human ovarian epithelial cells (pOE-1 and pOE-2), miR-765 downregulation ( Figure 6C) and BRD4 mRNA upregulation ( Figure 6D) were observed in primary ovarian cancer cells (pOC-1 and pOC-2) and established cell lines (CaOV3 and SKOV3). These results showed that miR-765 is downregulated in ovarian cancer tissues and cells, and it is correlated with BRD4 mRNA upregulation.
We discovered that miR-765 could be a novel BRD4targeting miRNA. In ovarian cancer cells, miR-765 mainly located at the cytosol fraction where it directly associated with BRD4 mRNA. Ectopic overexpression of miR-765 by lv-pre-miR-765 robustly inhibited BRD4 3'-UTR luciferase reporter activity and significantly downregulated BRD4 expression in primary and AGING established ovarian cancer cells. Conversely, lv-antagomiR-765-induced miR-765 silencing had opposite effects and it upregulated BRD4 expression in ovarian cancer cells. Interestingly, the two mutant miR-765 mimics with mutations at the binding sites to BRD4 3'-UTR failed to alter BRD4 expression in ovarian cancer cells. Importantly, in human ovarian cancer tissues, miR-765 was downregulated and negatively correlated with BRD4 mRNA upregulation. Thus, miR-765 is a novel BRD4-targeting miRNA in ovarian cancer.
Ectopic expression of BRD4-targeting miRNA has proven to be a good strategy to inhibit BRD4-driven cancer progression. Zheng et al., reported that expression of miR-4651 silenced BRD4 to inhibit nonsmall cell lung cancer (NSCLC) cell growth and proliferation [28]. Li et al., found that miR-608 induced pancreatic ductal adenocarcinoma cell apoptosis by silencing BRD4 [29]. Similarly, BRD4-targeting miR-608 inhibited HCC cell proliferation [30]. Kang and colleagues found that targeting BRD4 by miR-612 inhibited malignant development of NSCLC cells [31]. In cutaneous T-cell lymphoma, miR-29b downregulation is associated with BRD4-mediated activation of several oncogenes [32].
Here in primary ovarian cancer cells and established cell lines (SKOV3 and CaOV3), ectopic overexpression of miR-765 by lv-pre-miR-765 inhibited cell proliferation, migration and invasion, and simultaneously induced cell apoptosis activation. In contrast, lv-antagomiR-765-induced miR-765 inhibition augmented ovarian cancer cell proliferation, migration and invasion. Significantly, miR-765 overexpression-induced anti-ovarian cancer cell activity was largely inhibited by restoring BRD4 expression using the UTR-null BRD4. In addition, mimicking lv-pre-miR-765's actions, CRISPR/Cas9induced BRD4 KO induced significant proliferation inhibition and apoptosis in ovarian cancer cells. Importantly, BRD4 KO in ovarian cancer cells abolished the functional impact of miR-765. Therefore, miR-765 targeted and silenced BRD4 to potently inhibit ovarian cancer cell progression.

Chemicals and reagents
Puromycin, polybrene, the anti-BRDT antibody, fetal bovine serum (FBS) and other cell culture reagents were provided by Sigma-Aldrich Chemicals (St Louis, MO, USA). Other antibodies utilized in this study were purchased from Cell Signaling Tech China (Shanghai, China). From Genechem Co (Shanghai, China), primers, nucleotide sequences, constructs and viruses were obtained. RNA assay reagents were provided by Thermo-Fisher Invitrogen (Beijing, China).

Human tissues
From ten (10) primary ovarian cancer patients (all with written-informed consent, Stage II-III), ovarian cancer tissues ("T") and para-cancer normal ovarian epithelial tissues ("N") were obtained. Fresh human tissue specimens were washed, minced, homogenized in tissue lysis buffer (Beyotime Biotechnology, Wuxi, China). Experiments using human tissues and cells were approved by the Ethics Board of Soochow University, according to the Declaration of Helsinki.

Cells
Primary human ovarian epithelial cells, pOE-1 and pOE-2 were provided by Dr. Bi [33]. Cells were cultured in MCDB109/M199 medium plus 15% FBS. To obtain primary human ovarian cancer cells, fresh ovarian cancer tissues were washed, minced into small pieces, and digested with Collagenase I (Sigma-Aldrich) and DNase (Sigma-Aldrich). The resulting single-cell suspensions were pelleted and washed. Fibroblasts, blood vessel cells, and immune cells were abandoned. Purified cancer cells were cultured in the described medium [34]. Two different primary ovarian cancer cells, pOC-1 and pOC-2 were established. Established ovarian cancer cell lines, SKOV3 and CaOV3, were purchased from Cell Bank of Chinese Academy of Sciences (Shanghai, China). Cells were cultured in RPMI-1640 medium supplemented with 10% FBS.

Quantitative reverse transcription-polymerase chain reaction (qPCR)
Total cellular RNA was isolated from human tissues and cells using a RNA Kit (TIANGEN, Beijing, China). For RNA quantification, a NanoDrop™ 2000 Spectrophotometer (Invitrogen Thermo Fisher Scientific, Shanghai, China) was utilized. The PrimeScript RT Master Mix (Perfect Real Time; Takara Bio, Tokyo, Japan) was utilized for reverse transcription to generate complementary DNA (cDNA). TB Green Premix Ex Taq™ II (Takara Bio, Inc, Beijing, China) was utilized to test expression of targeted genes under the ABI 7900 system (Applied Biosystems, Foster City, CA). GAPDH was tested as the internal reference gene. A miRcute Plus miRNA qPCR Kit (SYBR Green; TIANGEN) was used for testing miR-765 expression, the result was normalized to U6 small nuclear RNA. All data were quantified through the 2 −ΔΔCt method. mRNA primers for BRD4, U6, GAPDH, c-Myc, Bcl-2 and FoxM1 were provided by Dr. Zheng [28]. MiR-765 primers, F: TGGAGGAGAAGGAAGGTG and R: GAACATGTCT AGING GCGTATCTC, were provided by Genechem (Shanghai, China).

Subcellular fractionation location
A Cytoplasmic and Nuclear RNA Purification Kit (Norgen, Belmont, CA, USA) was utilized for the isolation of RNA from the cytoplasmic and nuclear fractions based on the attached protocols.

"Transwell" migration and invasion assays
For migration assays, OS cells with the applied genetic modifications were trypsinized, rinsed, centrifuged, and resuspended in serum-free medium. For each well, 200 μL cell suspension containing 3 × 10 4 OS cells were added to the upper chamber of the "Transwell" inserts (BD Biosciences, Shanghai, China). The complete medium with 10% FBS was added to the lower chamber. After incubation, the non-migrated cells were removed. The migrated cells were fixed, stained and counted. Five visual fields of each condition were randomly selected and the average number of migrated cells was calculated. Invasion assays were conducted using the same protocol, only the "Transwell" inserts were pre-coated with Matrigel (BD Biosciences).

Western blotting
OS cells were incubated with RIPA protein lysis buffer (TIANGEN) to isolate total cellular protein. A BCA Protein Assay Kit (TIANGEN) was utilized to quantify protein concentration. Equal amounts of protein extracts (40 μg proteins per lane) were separated by 10-12% sodium dodecyl sulfate-polyacrylamide (SDS-PAGE) gels and were transferred to PVDF blots (Millipore, Billerica, MA, USA). After blocking, blots were incubated with the applied primary antibodies overnight at 4° C and subsequently incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies at room temperature for 1h. An Immobilon ECL Ultra Western HRP Substrate (Millipore) kit was utilized to detect the blotting signals.

Transfection of miRNA mimics
OS cells were seeded into six-well plates at approximately 60% confluence. Lipofectamine 3000 protocol was utilized for the transfection of 500 nM of the wild-type ("WT") or the mutant ("Mut") miR-765 mimics (for 48h). Expression of miR-765 was determined by qPCR after transfection.

BRD4 knockout (KO)
A lenti-CRISPR/Cas9-BRD4-KO-GFP construct was provided by Dr. Zhao [35]. OS cells were initially seeded into six-well plates and transfected with the construct by Lipofectamine 3000. GFP-positive cells were sorted by FACS. Cells were then cultured in 96well plates to form monoclonal cells. BRD4 knockout (KO) was screened by Western blotting and qPCR assays, and single stable cells were established.

UTR-null BRD4
The lentiviral construct containing3'-UTR-null BRD4 was provided by Dr. Zheng [28]. The construct was transfected to cultured OS cells by Lipofectamine 3000. Expression of UTR-null BRD4 was verified by Western blotting and qPCR assays.

RNA pull-down
A Pierce Magnetic RNA pull-down Kit was utilized for RNA pull-down assay through the described AGING protocols [36,37]. Briefly, OS cells were transfected with biotinylated miR-765 mimic or control mimic (Genechem, each at 300 nM) for 24h, and cells were harvested [37]. Cell lysates were incubated with streptavidin-coated magnetic beads to pull-down biotincaptured RNA complex [36]. The miR-765-bound BRD4 mRNA was tested by qPCR, and its level was normalized to "Input" controls.

Colony formation
Ovarian cancer cells were seeded into 10-cm dishes (at 1×10 4 cells per dish). Complete medium was renewed every two days. After 10 days, the number of large colonies was counted.

EdU incorporation
Cells were seeded into 96-well plates at 3, 000 cells per well and cultured for applied time periods. Afterwards, cells were cultured for 2h with EdU medium diluents. Cells were then incubated with DAPI solution for15 min, followed by observation under the fluorescence microscope (Olympus, Tokyo, Japan).

TUNEL assay
OS cells with applied genetic treatments were treated with 2% formaldehyde and 0.1% Triton X-100 for 5 min. Cells were then incubated with TUNEL reaction buffer for 1h and co-stained with DAPI. Apoptotic cells were visualized under Olympus microscope.

Histone DNA ELISA assay
A cell suspension (200 μL) containing 3 × 10 3 viable OS cells with applied genetic modifications was seeded into each well of 96-well plates. Cells were cultured for 72h and a Histone DNA ELISA kit was utilized to test Histone-bound DNA contents [39].

Mitochondrial depolarization
Ovarian cancer cells with the indicate genetic treatments were seeded into six-well plates (1 ×10 5 cells per well). JC-1 staining of mitochondrial membrane potential was described previously [40,41]. JC-1 green monomer intensity (at 488 nm) was measured.

Statistical analysis
Data were presented as the mean ± standard deviation (SD). The Student's t-test (Excel 2007) was utilized for the comparison between two specific treatment groups. Alternatively, one-way analysis of variance (ANOVA) with Tukey's test was utilized to test difference between multiple groups (SPSS23.0, SPSS Co. Chicago, CA). p values <0.05 were considered statistically significant.

CONFLICTS OF INTEREST
The listed authors have no conflicts of interest.