Target prediction tools TargetScan (version 7.1; http://www.targetscan.org/vert_71/) and miRanda (http://www.mirbase.org/) were used to search potential target miRNAs for PIK3CA, ARID1A, ETS1 and DNA-dependent protein kinase catalytic subunit (PRKDC).

The ovarian cancer cell lines TOV21G and TOV112D were obtained from the American Type Culture Collection (Manassas, VA, USA). All cell lines were grown in a 1:1 mixture of complete Medium 199 (Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA) and MCDB 105 (Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) supplemented with 15% fetal bovine serum (FBS; Gibco; Thermo Fisher Scientific, Inc.) and 1% penicillin/streptomycin (Gibco; Thermo Fisher Scientific, Inc.). All cells were cultured in a humidified incubator at 37°C with 5% CO₂. The procedure used to isolate stromal cells from an ectopic endometriotic implant has been described previously (14,15). Ovarian endometrial tissue from 3 patients of reproductive age (aged 21 to 42 years) who underwent laparoscopic surgery was obtained from the Department of Obstetrics and Gynecology, Kaohsiung Medical University Hospital (Kaohsiung, Taiwan). For each case, a final diagnosis of endometriosis was made according to the revised American Society of Reproductive Medicine classification (1997) in female patients that had been surgically resected between June 2014 and April 2016 (16). All samples were histologically confirmed by pathologists. All patients had regular menstrual cycles and had not received any hormone treatment within the 6 months prior to gynecological surgery.

The tissue samples were enzymatically dissociated using collagenase II, then stromal cells were separated from epithelial glands by wire sieves. The procedure used to isolate stromal cells from an ectopic endometriotic implant has been described previously (14,15). Filtered stromal cells were then cultured in Dulbecco's Modified Eagle's Medium Nutrient Mixture F12 (Gibco; Thermo Fisher Scientific, Inc.) supplemented with 10% FBS. All cells were cultured in a humidified incubator at 37°C with 5% CO₂. The study protocol was approved by the Ethics Committee of the Institutional Review Board of the Kaohsiung Medical University Hospital (approval no. KMUH-IRB-20140031), and informed written consent was obtained from each patient.

Serum samples were obtained from patients who underwent surgery with indications of gynecological disease at the Department of Obstetrics and Gynecology of Kaohsiung Medical University Hospital. A total of 9 subjects aged 21 to 55 years participated in the present study, including 3 healthy controls, 3 patients with endometriosis, and 3 patients with ovarian cancer (1 endometrioid and 2 clear cell carcinoma). In each case of endometriosis-associated ovarian cancer (EAOC), a final diagnosis of endometriosis-associated clear cell cancer, endometrioid cancer or ovarian cancer was made according to FIGO and WHO criteria in patients who were surgically resected between June 2014 and April 2016 (17). Each diagnosis of endometriosis, either by laparoscopy or laparotomy, followed the classification of the revised American Society of Reproductive Medicine, 1997 (16). The patients with endometriosis and EAOC were histologically confirmed by examination of a laparoscopy. All patients had regular menstrual cycles and had not received any hormone treatment in the 6 months prior to gynecological surgery. Healthy controls were diagnosed as having other benign diseases, including urinary incontinence, pelvic organ prolapse or ovarian hemorrhagic cyst. Exclusion criteria included post-menopausal status, endometrial cancer, hyperplasia, endometrial polyps, infectious diseases, adenomyosis, inflammatory diseases, malignancy, autoimmune disease or cardiovascular disease.

The study protocol was approved by the Institutional Review Board of Kaohsiung Medical University Hospital (approval no. KMUHIRB-G(I)-20150046), and written informed consent was obtained from each participant. Whole blood was drawn in EDTA-coated tubes (BD Biosciences, Franklin Lakes, NJ, USA) and centrifuged at 2,200 × g for 10 min at room temperature. The resultant sera were aliquoted into Eppendorf tubes and stored at −80°C for use in reverse transcription (RT) and quantitative polymerase chain reaction (qPCR) studies.

RNA was extracted from patient sera using bulk reagents from the MasterPure Complete DNA and RNA Purification kit (Epicentre; Illumina, Inc., San Diego, CA, USA) or purified from cultured cells using TRIzol reagent (Thermo Fisher Scientific, Inc.). RT was performed using the Deoxy + HiSpec RT kit (cat. FYT501-100R; Yeastern Biotech Co., Ltd., Taipei, Taiwan) and the TaqMan MicroRNA Reverse Transcription kit (cat. 4366596; Applied Biosystems; Thermo Fisher Scientific, Inc.). Then, qPCR was performed using the ABI 7900 Real-Time PCR system (Applied Biosystems; Thermo Fisher Scientific, Inc.). mRNA was amplified using the Eztime Fast Real-Time PCR Premix (2X, SYBR-Green, ROX; Yeastern Biotech Co., Ltd.), and 18S mRNA served as the internal control. The levels of miRNAs were measured using specific primers to miR-381 (000571), miR-203 (000507) and RNU6 (001093; Applied Biosystems; Thermo Fisher Scientific, Inc.). RNU6 served as the internal control. The reaction mixture included 2 µl cDNA, 5 µl TaqMan Universal PCR Master Mix II (2X; cat. no. 4440042; Applied Biosystems; Thermo Fisher Scientific, Inc.), 1 µl specific primers and 2.5 µl RNase-free water for a final reaction volume of 10 µl. The thermocycling conditions were initiated by uracil-N glycosylase activation at 50°C for 2 min and initial denaturation at 95°C for 10 min, followed by 40 cycles at 95°C for 15 sec, and annealing at 60°C for 60 sec. Expression levels were normalized to that of an endogenous control. Relative miRNA and mRNA levels were determined using the 2^−∆∆Cq method (18). The qPCR primers for HOXD10, 18S, PIK3CA, PRKDC, ETS1 and ARID1A were synthesized by Genemessenger Scientific Co., Ltd. (Kaohsiung, Taiwan), and their sequences are listed in Table I.

TOV21G and TOV112D cells were transiently transfected with 45 nM of miRNA mimics, inhibitors, and 1 µg of short hairpin RNAs (shRNAs) or other plasmids was performed using Turbofect Transfection Reagent (Thermo Fisher Scientific, Inc.). The miRIDIAN miRNA mimic (miR-381 mimics, cat. no. C-300690-03-0010) and mimic negative control (cat. no. CN-001000-01-05) were purchased from GE Healthcare Dharmacon, Inc. (Lafayette, CO, USA). The mirVana miRNA inhibitor (miR-381 inhibitor, cat. no. MH10242) and inhibitor negative control (cat. no. 4464078) were obtained from Ambion (Thermo Fisher Scientific, Inc.). The shRNAs included shRNA-homeobox D10 (HOXD10)#1 (cat. no. TRCN0000274091), shRNA-HOXD10#2 (cat. no. TRCN0000274093), shRNA-HOXD10#3 (cat. no. TRCN0000274028) and control shRNA (cat. no. TRCN0000274091; National RNAi Core Facility at the Institute of Molecular Biology, Academia Sinica, Taipei, Taiwan). pDNA-HOXD10 was purchased from Addgene, Inc. (Cambridge, MA, USA). Following 24 h of transfection, the cells were subjected to assays measuring mRNA and protein expression, cell viability, migration and colony formation.

Cell viability was measured using a Cell Counting Kit-8 (CCK-8) assay (Honeywell Fluka; Thermo Fisher Scientific, Inc.). Briefly, ovarian cancer cell lines were seeded into 96-well plates at a density of 5×10³ cells/well in a total volume of 100 µl medium with 15% (v/v) FBS, transfected with miRNA mimics, inhibitors or scrambled-sequence controls, and incubated at 37°C in an atmosphere of 5% CO₂. The following day, 90 µl culture medium and 10 µl CCK-8 reagent were added to each well. Cell viability was assayed after 2 h using an enzyme-linked immunosorbent assay reader (Multiskan EX; Thermo Fisher Scientific, Inc.) at 450 nm (reference, 650 nm). In each case, three independent experiments were performed.

Ovarian cancer cells transduced with the miR-381 mimic, miR-381 inhibitor, or scrambled-sequence controls were plated in triplicate into 6-well plates (500 cells/well), cultured in a humidified incubator at 37°C with 5% CO₂ for 2 weeks, then fixed with 4% paraformaldehyde and stained with 0.1% crystal violet at room temperature for 1 h. Colonies were counted by visual inspection. Digital images of the colonies were obtained using a camera. Colonies were counted using ImageJ software (version 1.47; National Institutes of Health, Bethesda, MD, USA). Triplicate wells were measured for each treatment group.

A total of 5×10⁴ cells were seeded into the upper chamber of each Transwell insert (8 µm pore size) in a 24-well plate in serum-free medium. In the lower chambers, medium containing 15% FBS was added as a chemoattractant. Following incubation for 18 h at 37°C, non-migrated cells in the upper chamber were removed with cotton swabs. The migrated cells on the lower chamber surface were fixed with 4% paraformaldehyde for 15 min and stained with 0.1% crystal violet at room temperature for 15–20 min. The number of migrated cells was calculated by counting three different fields of view under a phase-contrast microscope. In each case, three independent experiments were performed.

Proteins were extracted using radioimmunoprecipitation lysis buffer (EMD Millipore, Billerica, MA, USA) containing several protease and phosphatase inhibitors (Roche Applied Science, Madison, WI, USA). Protein concentration was determined with the Bio-Rad DC Protein Assay kit from Bio-Rad Laboratories, Inc. (Hercules, CA, USA). Cell lysates were analyzed by western blotting. Proteins (30 µg) were resolved on a 10% BIS-TRIS SDS gel) at 100 V and transferred onto nitrocellulose membranes at 90 V for 1.5 h. Non-specific binding was blocked by incubating the membranes in 5% nonfat dry milk in PBS for 1 h. Then, the membranes were incubated with primary antibodies overnight at 4°C. Primary antibodies were as follows: Mouse monoclonal anti-HOXD10 (cat. no. TA800777; 1:500; OriGene Technologies, Inc., Rockville, MD, USA) or rabbit monoclonal anti-PIK3CA (cat. no. TA302178; 1:1,000; OriGene Technologies, Inc.). Expression of β-actin was used as a loading control and was detected with monoclonal anti-actin (cat. no. A5316; 1:5,000, Sigma-Aldrich; Merck KGaA). The secondary antibodies used were horseradish peroxidase-conjugated goat anti-mouse IgG or anti-rabbit IgG (Santa Cruz Biotechnology, Inc., Dallas, TX, USA). Secondary antibodies were incubated at room temperature for 1 h. Enhanced chemiluminescence reagents (EMD Millipore) were used for immunodetection.

ChIP assays were performed using the Chromatin Immunoprecipitation Assay kit (Upstate Biotechnology, Inc., Lake Placid, NY, USA). The ALGGEN-PROMO prediction software (http://alggen.lsi.upc.es/cgi-bin/promo_v3/promo/promoinit.cgi?dirDB=TF_8.3) was used to identify potential transcription factor binding sites within the miR-381 promoter (19). Chromatin was incubated with anti-HOXD10 (cat. no. TA800777; 1:500; OriGene Technologies, Inc.) or goat anti-mouse IgG (control, cat. no. TA130004; 1:500: OriGene Technologies, Inc.). PCR was conducted with primers complementary to the miR-381 promoter region. The primer sequences are presented in Table I. RT semi-quantitative PCR was performed to amplify the transcription factor binding sequence of miR-381. RT-PCR analysis of relative expression was performed with the one-step Emerald Amp GT PCR Master mix (Takara Bio, Inc., Otsu, Japan). Thermocycling was performed under the following conditions: 94°C for 5 min; 35 cycles of 94°C for 30 sec, 58°C for 30 sec and 72°C for 30 sec. PCR products were analyzed by electrophoresis with 2% agarose gels and visualized by UV light following staining with EtB‘Out’ Nucleic Acid Staining Solution (cat. no. FYD007-200P; Yeastern Biotech Co., Ltd).

Data are presented as the mean ± standard deviation from at least three independent experiments by GraphPad Prism 7 (GraphPad Software, Inc., La Jolla, CA, USA). For the in vitro cell migration assays, groups were compared using the Mann-Whitney U test. For all other results, one-way analysis of variance followed by Tukey's honest significant difference test was used to analyze differences among multiple groups. Student's t-test was used to analyze differences between two groups. P<0.05 was considered to indicate a statistically significant difference.

In our previous work, the genetic alterations that are commonly observed in endometriosis-associated ovarian cancers were characterized; the genes PIK3CA, ARID1A, ETS1 and PRKDC were identified to be the most frequently mutated genes in patient samples (13). Further examination revealed potential interactions between miRNAs and the mutated mRNAs in ovarian tumors. To characterize the mechanisms by which genetic alterations promote ovarian cancer, TargetScan (version 7.1) and miRanda were used to predict potential upstream regulators of the mutated gene (20). The two most commonly mutated genes encoded mRNAs are regulated by two miRNAs, namely miR-203 and miR-381 (Fig. 1A). Therefore, in the present study, the expression level of miR-203 and miR-381 was determined in tumor samples from patients with ovarian cancer and in ovarian cancer cell lines. The results revealed that miR-381 and miR-203 were expressed at a significantly lower level in the sera of patients with ovarian cancer compared with healthy patients and endometriosis patients, and in ovarian cancer cell lines compared with controls (Fig. 1B and C). Subsequent studies focused on miR-381 because it was the most consistently downregulated miRNA in patient sera (Fig. 1B) and ovarian cancer cell lines (Fig. 1C).

To address whether miR-381 affects the cellular levels of mRNAs encoding PIK3CA, ARID1A, EST1 and PRKDC, ovarian cancer cell lines were transfected with an miRNA mimic or miRNA inhibitor specific for these individual mRNAs and then mRNA expression was assessed. The PIK3CA mRNA level was significantly reduced in ovarian cancer cell lines transfected with miR-381 mimics compared with a non-specific control mimic; by contrast, PIK3CA mRNA level was significantly increased following transfection with an miR-381 inhibitor compared with a control non-specific inhibitor (Fig. 2A and B). However, the cellular levels of ARID1A, ETS1 and PRKDC mRNAs did not respond consistently to transfection with various miRNA mimics and inhibitors (Fig. 2A and B). Next, the effect of various concentrations of miR-381 mimics or inhibitors on the cellular level of PIK3CA mRNA was examined in the various transiently transfected ovarian cancer cell lines (Fig. 2C). The results indicated that 45 nM of miR-381 mimic or inhibitor affected the cellular level of PIK3CA mRNA and protein (Fig. 2C and D), suggesting that miR-381 effectively suppresses the cellular level of both PIK3CA mRNA and protein.

To investigate how miR-381 suppression promotes ovarian cancer progression, cell proliferation and migration assessed following targeted knockdown or overexpression of miR-381. Cell proliferation was evaluated by CCK-8 and colony formation assays. The downregulation of miR-381 significantly increased cell proliferation, and inhibition of miR-381 activity significantly decreased proliferation (Fig. 3A). Similarly, in a colony formation assay, which is an in vitro cell survival assay based on the ability of a single cell to grow into a colony (21), clonogenic survival decreased following transfection of cells with miR-381 mimics, whereas survival was enhanced in miR-381 inhibitor-transfected cells (Fig. 3B). Furthermore, cell migration was evaluated by Transwell assays, in which overexpression of miR-381 significantly decreased migration, whereas an miR-381 inhibitor significantly increased migration (Fig. 3C). These results indicated that miR-381 suppresses the growth and migration of ovarian cancer cells.

The current results indicated that miR-381 is an upstream regulator of PIK3CA in ovarian cancer cells. Previous studies have identified that dysregulation of the phosphatidylinositol 3-kinase (PI3K) pathway, which is commonly observed in human tumors, drives tumorigenesis; therefore, the PI3K pathway is an appealing therapeutic target (22). Hence, the upstream regulators of miR-381 with respect to the progression of ovarian cancer were investigated. The ALGGEN-PROMO prediction program (http://alggen.lsi.upc.es/cgi-bin/promo_v3/promo/promoinit.cgi?dirDB=TF_8.3) was used to identify potential transcription factor binding sites within the miR-381 promoter (19,21,23). HOXD10 is a tumor-suppressor gene that suppresses tumor invasion and hence could potentially modulate the cellular level of miR-381 (24). To determine whether miR-381 level is regulated by HOXD10, ovarian cancer cells were transfected with an overexpression plasmid encoding HOXD10. Subsequently, HOXD10 was efficiently overexpressed (Fig. 4A), resulting in a significantly increased cellular level of miR-381 expression (Fig. 4B). Conversely, HOXD10 knockdown resulted in a decreased level of miR-381 in cells (Fig. 4B). Furthermore, it was investigated whether HOXD10 modulates the cellular level of miR-381. The ALGGEN-PROMO program was used to predict the promoter region for miR-381. ChIP assays revealed that HOXD10 positively regulated transcription of the miR-381 gene by binding to the region −565 to −574 bp upstream of the miR-381 coding sequence (Fig. 4C).