BRDT promotes ovarian cancer cell growth

Bromodomain testis-specific factor (BRDT) is a member of the bromodomain and extra-terminal (BET) family proteins. Its expression and potential functions in ovarian cancer were examined. We show that BRDT is overexpressed in human ovarian cancer tissues and in established (CaOV3)/primary ovarian cancer cells. However, its expression is low in ovarian epithelial tissues and cells. Significantly, shRNA-induced silencing or CRISPR/Cas9-mediated knockout of BRDT inhibited ovarian cancer cell growth, viability, proliferation and migration, and induced significant apoptosis activation. Conversely, exogenous overexpression of BRDT, by a lentiviral construct, augmented CaOV3 cell proliferation and migration. In CaOV3 cells expression of two key BRDT target genes, polo-like kinase 1 (PLK1) and aurora kinase C (AURKC), was downregulated by BRDT shRNA or knockout, but upregulated with BRDT overexpression. In vivo, xenograft tumors-derived from BRDT-knockout CaOV3 cells grew significantly slower than control tumors in severe combined immunodeficient (SCID) mice. Furthermore, intratumoral injection of BRDT shRNA lentivirus potently inhibited the growth of primary ovarian cancer xenografts in SCID mice. Downregulation of PLK1 and AURKC was detected in BRDT-knockout and BRDT-silenced tumor tissues. Collectively, BRDT overexpression promotes ovarian cancer cell progression. Targeting BRDT could be a novel strategy to treat ovarian cancer.


Introduction
Despite significant progresses have been achieved to improve the overall survival of ovarian cancer in past decades, it is still one primary cause of female mortality 1 . In the United States alone, the American Cancer Society estimates 22,440 new cases of ovarian cancer and 14,080 related deaths each year 2 . Ovarian cancer is typically diagnosed at late stages, due to the lack of effective screening strategy [3][4][5] . Standard treatments for newly diagnosed ovarian cancers include surgery and platinumbased chemotherapy [3][4][5] . Yet, because of the significant resistance and recurrence, researchers are focusing on novel oncogenes and cell signaling pathways essential for cancer progression [3][4][5] .
Bromodomain testis-specific factor (BRDT) is a member of the bromodomain and extra-terminal (BET) family proteins 6 . BRDT, like other BET family proteins (BRD1-4), epigenetically regulates targeted genes expression through interacting with acetylated lysines 7,8 , critical for normal development and disease (cancer) progression 6 . Recent studies have implied that BET family proteins are key oncogenic proteins, being overexpressed in a number of different cancers 7,8 . Inhibitors of BET proteins have displayed promising efficiency against human cancer cells in vitro and in vivo 9,10 .
BRDT is expressed in testes, regulating the meiotic and post-meiotic genes expression to promote spermatogenesis 11 . Cancer cells have the ability to reactivate the normally silent testis-restricted genes, which are important for cancer progression 6 . Recent studies have shown that BRDT is ectopically activated and expressed in human cancers 6 . It has been proposed that BRDT could be a novel biomarker and a possible therapeutic target for human cancer 6 . Several BRDT-regulated genes, including polo-like kinase 1 (PLK1) 12 and aurora kinase C (AURKC) 13 , are key oncogenic proteins 6,11 . BRDT expression and potential functions in ovarian cancer have not been studied thus far.

Human tissues
From a total of six (6) primary ovarian cancer patients, the fresh human ovarian cancer tissues ("Ca") and paracancer ovarian epithelial tissues ("S") were acquired. Tissues were washed, minced, and homogenized in the tissue lysis buffer (Beyotime Biotechnology, Shanghai, China), stored in liquid nitrogen. Expression of BRDT was tested by quantitative reverse transcriptase PCR ("qPCR") and western blotting assays. The written-informed consent was obtained from each participant. The enrolled patients received no prior chemotherapy before surgeries. The surgically excised normal testis tissue was provided by a written-informed testicular cancer patient administrated at Affiliated Kunshan Hospital of Jiangsu University. Experiments and protocols requiring human tissues and cells were approved by the Ethics Board of Affiliated Kunshan Hospital of Jiangsu University, according to Declaration of Helsinki.

Cell culture
Ovarian cancer cell line CaOV3 was purchased from the Cell Bank of Shanghai Institute of Biological Science, CAS (Shanghai, China). Cells were cultured in the medium previously described 14 . The normal ovarian epithelial ("OE") cells were provided by Dr. Bi 14 , which were cultured in MCDB109/M199 medium with 20% fetal bovine serum (FBS). The achieved ovarian cancer tissues were washed in phosphate-buffered saline for five times and minced into small pieces (1 mm 3 ), which were digested by Collagenase I (Sigma-Aldrich) and DNase (Sigma-Aldrich). Single-cell suspensions were pelleted and washed in Dulbecco's modified Eagle medium (DMEM). Fibroblasts, blood vessel cells, and immune cells were abandoned via mechanical means and centrifugation. Purified primary human ovarian cancer cells were cultured in the medium previously described 15 . Four different ovarian cancer cells were established, named "pOC-1/4". All cells were regularly checked to exclude possible mycoplasma and microbial contamination. Authentication by STR profiling, population doubling time, and cell morphology were also routinely examined to verify the genotypes every 3-4 months.
qPCR Trizol reagents (Promega, Madison, WI) were added for RNA extraction. A SYBR Green PCR kit (Applied Biosystems, Foster City, CA) was utilized for reverse transcription under the ABI Prism7500 Fast Real-Time PCR system. A melt curve analysis was performed to calculate product melting temperatures. To quantify targeted gene expression a 2 −ΔΔCt method was applied, with GAPDH tested as the internal control. The primers were listed in Table 1.

Western blotting
Cells or tissues were incubated with the described lysis buffer 14 . Aliquots of 40 µg of proteins of each condition were separated by 10% sodium dodecyl sulphate-polyacrylamide gel electrophoresis gels and then transferred to PVDF blots (Millipore, Bedford, MA). After blocking, blots were incubated with specific primary antibodies (overnight at 4°C) and corresponding secondary antibodies (2 h at room temperature) 14 . Antibody-antigen bindings were tested by an enhanced chemiluminescence (ECL) system (Amersham Biosciences, Piscataway, NJ). An Image J software (NIH) was utilized for data quantification.

MTT assay
Cells, with different genetic manipulations, were initially seeded into 96-well plates at a density of 2.5 × 10 3 cells per well. After incubation for 72 h, the cell viability was tested by MTT assay. MTT optical density (OD) was measured at 550 nm.

BrdU assay
Ovarian cancer cells were seeded into 96-well plates at a density of 2.5 × 10 3 cells per well. Following incubation with BrdU (10 μM, Cell Signaling Tech, Shanghai, China) for 48 h, cells were washed and BrdU incorporation was tested by an ELISA kit (Cell Signaling Tech), with ELISA absorbance tested at 405 nm.
"Transwell" assay Genetically modified CaOV3 cells (1.5 × 10 5 cells in 250 μL serum-free medium) were seeded into "Transwell" upper chambers (16-μm pore size, BD Biosciences, Shanghai, China). The lower compartments were filled with 10% FBS medium. After 24 h, the non-migrated cells on the upper surface were removed. The migrated cells, on the lower surface, were fixed, stained and counted.

Caspase-3 activity assay
Ovarian cancer cell lysates (20 μg of each treatment) were mixed with the described assay buffer 16 , together with the caspase-3 substrate (Ac-DEVD-AFC). Following extensive washes, the caspase-3 enzymatic AFC activity was measured at the excitation wavelength of 380 nm and the emission wavelength of 440 nm.

Mitochondrial depolarization
JC-1 fluorescence dye can form green monomers by aggregating in the mitochondria in apoptotic cells with mitochondrial depolarization 19 . Ovarian cancer cells with applied genetic modifications were seeded into 24-well plates (3 × 10 4 cells per well) and cultured for 48 h. Cells were then stained with JC-1 (10 μg/mL, Sigma). JC-1 fluorescence images, integrating green (at 488 nm) and red (at 625 nm) wavelengths were presented. JC-1 green fluorescence intensity (at 488 nm) was recorded as well.

Short-hairpin RNA (shRNA)-mediated gene silencing
There different shRNAs, targeting non-overlapping sequences ("−1/−2/−3", listed in Table 1) of BRDT, were individually sub-cloned into a GV248 (hU6-MCS-Ubiquitin-EGFP-IRES-puromycin) construct. The shRNA construct, along with the lentivirus package plasmids, were co-transfected to HEK-293 cells to generate BRDT shRNA lentivirus. Following filtration and enrichment, the virus was added to ovarian cancer cells (cultured in the polybrene medium, same for all virus procedures). The infected cells were then subjected to selection by puromycin (2.0 μg/mL) for another 4-5 passages. Knockdown of BRDT in the stable cells was verified by qPCR and Western blotting assays. Silencing of PLK1 and AURKC was through the same procedure, with their shRNA sequences listed in Table 1.

BRDT knockout
The single-guide RNA (sgRNA) targeting BRDT (two different sequences, "sgRNA-1/-2", listed in Table 1, provided by Shanghai Genechem Co.) was sub-cloned into a lenti-CRISPR-GFP-puro construct, which was transfected to HEK-293 cells along with the lentivirus package plasmids to generate BRDT-KO virus. The virus was filtered, enriched and added to ovarian cancer cells for 24 h. Cells were subjected to selection with puromycin (2.0 μg/mL) for 4-5 passages. In the resulting Table 1 Sequences utilized in this study. stable cells BRDT KO was verified by western blotting and qPCR assays.

Overexpression of BRDT and others
The full-length BRDT cDNA was synthesized and sequence-verified by Shanghai Genechem Co, sub-cloned to a GV248 vector. The construct was then transfected to HEK-293 cells with the lentiviral packaging plasmids 20 , generating BRDT-expressing lentivirus ("LV-BRDT"). Following filtration and enrichment, LV-BRDT was added to ovarian cancer cells. Afterwards, puromycin (2.0 μg/ mL) was included to select stable cells, where BRDT overexpression was verified by Western blotting and qPCR assays. Control cells were infected with lentivirus with empty vector ("LV-C"). Ectopic overexpression of PLK1 and AURKC was through the same protocol.

Xenograft assay
The severe combined immunodeficient (SCID) mice (17.5-18.5 g, 4-5-week-old) were obtained from the Animal Center of Chinese Academy of Science (Shanghai, China). CaOV3 or pOC-1 primary cells (for each mouse, 5 × 10 6 cells in 100 μL DMEM plus 100 μL Matrigel, no serum) were subcutaneously (s.c.) injected to the right flanks of SCID mice. After 3 weeks the subcutaneous xenografts were established (around 100 mm 3 ), and recordings were initiated (Day-0, or "D0"). Mice body weights and bi-dimensional tumor measurements were recorded every seven days for total 35 days. The animal protocols were approved by the Ethics Board and IACUC of Affiliated Kunshan Hospital of Jiangsu University.

Statistical analyses
In vitro experiments were repeated at least three times and similar results were obtained. Values were normalized when necessary and expressed as mean ± standard deviation (SD, normal distribution). For statistical analyses the SPSS software (version 21.0, using one-way ANOVA) was employed. To test significance between two treatment groups, a two-tailed unpaired t-test (Excel 2007) was utilized. All differences were considered significant at P < 0.05.

BRDT overexpression in ovarian cancer
First, the database proteomicsdb (www.proteomicsdb. org) was consulted to verify BRDT expression in human tissues. As shown, BRDT protein is primarily expressed in testis, lung and ovarian tissues (Fig. 1A), very few was detected in other human tissues. To verify the data, a total of six ovarian cancer tissues ("Ca") and paired surrounding normal ovarian epithelial tissues ("S") were tested. These tissues were derived from six different primary ovarian cancer patients: Pat-1 to Pat-6. The qPCR results, Fig. 1B, demonstrated that BRDT mRNA expression was relatively low in normal ovarian epithelial tissues (Fig. 1B), but was significantly upregulated in five out of six cancer tissues (Pat-1 to Pat-5, Fig. 1B). BRDT protein upregulation was detected as well in the five ovarian cancer tissues (Fig. 1C). Again, low BRDT protein expression was detected in ovarian epithelial tissues (Fig. 1C). BRDT expression in human testis tissue was shown as the positive control (Fig. 1B, C).
In established (CaOV3 cell line) and primary human ovarian cancer cells, BRDT mRNA (Fig. 1D) and protein (Fig. 1E) expression was significantly higher than that in ovarian epithelial ("OE") cells. The primary cancer cells were derived from the four ovarian cancer tissues with significant BRDT upregulation ("Pat-1/-2/-3/-4", see Fig.  1B, C). These results together show that BRDT is overexpressed in human ovarian cancer tissues and cells.

BRDT overexpression promotes ovarian cancer proliferation and migration
Based on these results, we hypothesized that forced BRDT overexpression might promote ovarian cancer cell growth. To test this hypothesis, a lentiviral BRDT expression construct ("LV-BRDT") was transfected to CaOV3 cells. Puromycin was applied again to select two stable cell lines ("sL-1/sL-2"). Testing BRDT mRNA expression, by qPCR, confirmed that in the LV-BRDT stable cells BRDT mRNA levels increased over 10 folds (vs. control cells with empty vector/"LV-C") (Fig. 5A). BRDT protein levels were increased as well (Fig. 5A). As shown, exogenous BRDT overexpression augmented CaOV3 cell viability (Fig. 5B) and proliferation (by recording EdU-positive nuclei ratio, Fig. 5C). BRDT overexpression also promoted CaOV3 cell migration, tested by "Transwell" assays ( Fig. 5D) These results further confirmed that BRDT plays an essential role in ovarian cancer cell progression.
Conversely, forced BRDT overexpression in CaOV3 cells (see Fig. 5) significantly increased PLK1 and AURKC expression (both mRNA and protein) (Fig. 6C, D). These results suggest that BRDT is indeed important for PLK1 and AURKC expression in ovarian cancer cells.

BRDT silencing inhibits ovarian cancer xenograft growth in SCID mice
In order to study the potential effect of BRDT in ovarian cancer cell growth in vivo, control CaOV3 cells ("Cas9-C") and BRDT-KO (with "sgRNA-1") CaOV3 cells were inoculated to SCID mice to form subcutaneous xenografts. Tumor growth curve results, Fig. 7A, demonstrated that xenografts-derived from BRDT-KO CaOV3 cells grew significantly slower than the control tumors (formed by Cas9-C cells). By calculating the estimated daily tumor growth, using the formula [tumor volume at Day-35 (D35) subtracting tumor volume at Day-0 (D0)]/35, we show that BRDT-KO CaOV3 tumor growth was significantly inhibited (Fig. 7B). At D35 tumors of the two groups were isolated and weighted. As demonstrated, BRDT-KO CaOV3 xenografts were significantly lighter than the control tumors (Fig. 7C). The mice body weights, on the other hand, were not significantly different between the two groups (Fig. 7D). There were no noticeable signs of apparent toxicity. These results suggest that BRDT-KO significantly inhibited CaOV3 cell growth in vivo.
To testing signaling changes, at D7, three tumors of each group were isolated and tissues lysates were analyzed for signaling proteins. As shown, BRDT was depleted in BRDT-KO CaOV3 xenografts (Fig. 7E), where PLK1 and AURKC expression was significantly decreased (Fig. 7E). These signaling results in vivo are in line with the in vitro findings.
To further support the role of BRDT in ovarian cancer cell growth in vivo, primary ovarian cancer cells, pOC-1, were s.c. injected to the flanks of SCID mice, forming pOC-1 xenografts within 3 weeks. The mice were then randomly assigned into two groups, receiving intratumoral injection of either BRDT shRNA lentivirus ("shBRDT-3") or control shRNA lentivirus (shSCR). When recording tumor growth, we found that shBRDT-3 injection potently inhibited pOC-1 xenograft growth in mice (Fig. 7F), but without affecting mice body weights (Fig. 7G). At Day-7 (D7), two tumors of each group were isolated and tissues lysates were subjected to Western blotting assays. Results confirmed BRDT silencing in shBRDT-3-injected tumors (Fig. 7H), with PLK1 and AURKC downregulation observed (Fig. 7H). These results further indicated that BRDT is important for ovarian cancer cell growth in vivo.

Discussion
BRDT is a driver of meiotic and post-meiotic gene expression 6,28 . BRDT binds to acetylated lysines to regulate epigenetic processes, essential for chromatin structure formation in mitosis progression 6 . Furthermore, BRDT recruits positive transcription elongation factor b (p-TEFb), regulating transcription elongation and expression of several key proteins 6 , including PLK1 and AURKC 11 . A threefold higher BRDT expression is detected in adult testis than that in the embryo testis 29,30 . BRDT expression is correlated with histone H4 hyperacetylation during spermiogenesis 29,30 . In patient with abnormal spermatogenesis, few or no BRDT expression is detected in testis 29,30 . Recent studies have shown that BRDT is reactivated and expressed in various human cancers, including non-small cell lung cancer (NSCLC), head and neck squamous cell carcinomas (HNSCC) and esophagus squamous cell carcinomas, but not in melanoma or in cancers of the colon, breast, kidney and bladder 6,28,30 . The proteomicsdb database and our results in human tissues confirmed that BRDT is expressed in ovarian tissues.
Our results here suggest that BRDT is possibly one important oncogenic gene and therapeutic target of ovarian cancer. Its expression is significantly upregulated in ovarian cancer tissues and in established (CaOV3)/ primary human ovarian cancer cells. Low BRDT expression, however, is detected in ovarian epithelial tissues and cells. In ovarian cancer cells, BRDT shRNA or CRISPR-Cas9-indueced BRDT knockout potently inhibited cell growth, survival, proliferation and migration, whereas inducing apoptosis activation. Conversely, forced BRDT overexpression augmented CaOV3 cell proliferation and migration. Importantly, BRDT-KO CaOV3 xenograft tumors grew significantly slower than the control tumors. Furthermore, intratumoral injection of BRDT shRNA lentivirus potently inhibited pOS-1 xenograft growth in SCID mice. These results suggest that targeting BRDT could be a novel and efficient strategy to inhibit ovarian cancer cells, in vitro and in vivo.
PLK1 plays an essential role in regulating cell cycle progression 12 . It is required for a number of cell cycle Fig. 7 BRDT silencing inhibits ovarian cancer xenograft growth in SCID mice. The SCID mice were injected s.c. with control and BRDT-KO CaOV3 cells (5 × 10 6 cells in 100 μL DMEM plus 100 μL Matrigel, no serum) at the right flanks, within 3 weeks the xenografts were established (Day-0, "D0"). The tumor volumes (A) and mice body weights (B) were recorded every seven days for total 35 days; The estimated daily tumor growth was calculated (C); At D35 all tumors were isolated and weighted (D). At D7 three tumors of each group were isolated and tissues lysates were subjected to Western blotting assays of listed proteins (E). The SCID mice were injected s.c. with pOC-1 primary ovarian cancer cells (5 × 10 6 cells in 100 μL DMEM plus 100 μL Matrigel, no serum) at the right flanks. Within 3 weeks the xenografts were established (Day-0, "D0"). The mice were then subjected to intratumoral injection of either BRDT shRNA lentivirus ("shBRDT-3") or control shRNA lentivirus (shSCR). Thereafter, the tumor volumes (F) and mice body weights (G) were recorded every 7 days; At Day-7/D7, two tumors of each group were isolated and tissues lysates were subjected to Western blotting assays (H). Listed protein expression was quantified and normalized to Tubulin (E and H). n = 8 mice per group *P < 0.05 vs. "Cas9-C"/"shSCR" tumors.
processes, including cell mitotic entry and G2/M checkpoint, centrosome coordination, spindle assembly and chromosome segregation 12 . PLK1 also exerts key functions at the spindle midzone during abscission, facilitating DNA replication and cytokinesis 12 . It is, therefore, vital for cell cycle progression and cell proliferation 12 . Studies have indicated an oncogenic activity of overexpressed PLK1 in ovarian cancer 21,27 , that is important for cancer cell progression and chemoresistance 21,27 . Additionally, AURKC, another BRDT-regulated gene, is reactivated and overexpressed in multiple different human cancer cells 13 , essential for cancerous behaviors, including cell proliferation and migration 13 . AURKC could promote xenograft tumor growth 13 . Kinase-dead AURKC inhibited HeLa cell proliferation, whereas constitutively active AURKC promoted cancer cell progression [31][32][33] .
The results of the current study show that PLK1 and AURKC are expressed in ovarian cancer cells. Significantly, the two were downregulated by BRDT shRNA or KO, but being upregulated with ectopic BRDT overexpression. PLK1 and AURKC downregulation was also detected in BRDT-KO CaOV3 tumor tissues and in pOS-1 xenografts with BRDT shRNA injection. Importantly, BRDT-KO-induced anti-ovarian cancer cell activity was largely attenuated with PLK1 and AURKC re-expression. PLK1 plus AURKC double silencing-induced proliferation inhibition and apoptosis activation in CaOV3 cells, mimicking BRDT silencing-induced actions. Therefore, BRDT-mediated ovarian cancer cell progression is associated with regulation of PLK1 and AURKC expression. The detailed mechanisms may warrant further characterizations.

Conclusion
The current treatments for ovarian cancer, with combination of platinum-based chemotherapy and surgery, can only result in a 5-year overall survival close to 45%. It goes down to 25% for the advanced cancers 34 . The poor survival highlights the limitations in the biological understanding of this devastating disease [35][36][37][38] . It is, therefore, urgent to indentify novel therapeutic strategies/ molecular targets [35][36][37][38] . Recent genomic studies have demonstrated the extensive alternations of epigenetic regulators in ovarian cancers, which might be utilized as basis for developing new epigenetic drugs [35][36][37][38] . Our results indicate that BRDT overexpression promotes ovarian cancer cell progression in vitro and in vivo. Targeting BRDT could be a novel strategy to treat ovarian cancer.