There were 112 paraffin-embedded tissue samples, including 83 epithelial ovarian cancer, 18 benign epithelial tumor and 11 normal ovarian tissues (resected for non-ovarian diseases) that were collected from the Department of Pathology, the Second Affiliated Hospital of Chongqing Medical University between 2010 and 2013. All the ovarian cancer patients underwent surgical exploration and cytoreduction as the initial treatment. The clinicopathological characteristics are presented in Table I. The patients were staged according to the International Federation of Gynecology and obstetrics (FIGO). The tissue blocks were re-evaluated by two senior pathologists for histological type and histopathologic grading. Tissue samples were obtained following informed consent by the patients. The study protocol was approved by the Ethics Committee of the Second Affiliated Hospital of Chongqing Medical University.

The tissue slides were deparaffinized in xylene and rehydrated through graded ethanol. Antigen retrieval was performed in 10 mmol/l boiling sodium citrate buffer at pH 6.0 for 15 min by microwave irradiation. The slides were then incubated with 3% hydrogen peroxide (H₂O₂) for 10 min at room temperature. After rinsing, the non-specific binding site was blocked with 10% normal goat serum for 15 min at room temperature. A mouse monoclonal anti-human CFTR antibody (diluted 1:200, ab2784; Abcam, Cambridge, Uk) was applied to slides in a moist chamber at 4°C overnight. After washing with phosphate-buffered saline (PBS), the slides were incubated with biotinylated secondary antibody for 15 min at room temperature (diluted 1:1,000, cat. no. SP-9002; Zhongshan Golden Bridge Inc., China). The slides were treated with ABC reagent for 15 min at room temperature, and stained with 3,3′-diaminobenzidine (DAB), followed by counterstaining with hematoxylin.

Cells with brown staining for the membrane and/or cytoplasm were considered positive. A semi-quantitative scoring system was applied to assess protein level based on intensity (0–3: 0, absence of staining; 1, weakly stained; 2, moderately stained; and 3, strongly stained) and percentage of positive tumor cells (0–3: absence of positive cells; 1, 0.1–33% of cells positive; 2, 33.1–66% of cells positive; 3, >66.1% of cells positive) and generated a score ranging from 0 to 9 when multiplied (16). Low expression was defined as a staining score of 0–4, and high expression was defined as a staining score of 5–9.

The human A2780 and SKOV3 epithelial ovarian cancer cell lines were obtained from the Laboratory of Obsterics and Gynecology, the Second Affiliated Hospital of Chongqing Medical University. The cells were cultured in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum and antibiotics (100 U/ml penicillin and 100 μg/ml streptomycin) (all from Invitrogen, Carlsbad, CA, USA). The cells were cultured in an incubator with 5% CO₂ at 37°C.

Total proteins were extracted using lysis buffer (Beyotime, Jiangsu, China) consisting of 50 mM Tris (pH 7.4), 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS and a protease inhibitor mixture supplemented with 1 mM phenylmethanesulfonyl fluoride (PMSF) according to the manufacturer’s instructions. The protein concentration was measured using a BCA protein assay kit (Beyotime). Total proteins (100 μg) of each sample were separated on 8% SDS-PAGE gel by electrophoresis and transferred to polyvinylidene fluoride (PVDF) membranes (Millipore Corporation, Billerica, MA, USA). The membranes were then blocked in 5% skim milk at 37°C for 1 h. Subsequently, the membranes were incubated, respectively, in mouse monoclonal anti-human CFTR antibody (diluted 1:1,000, ab2784) and rabbit monoclonal anti-β-actin antibody (diluted 1:1,000, ab133626) (both from Abcam) at 4°C overnight. After washing with TBST, the membranes were incubated with HRP-conjugated secondary antibodies for 1 h at room temperature. Proteins were visualized with the ECL system (GE Healthcare, Pittsburgh, PA, USA) using the ChemiDoc XRS system (Bio-Rad, Philadelphia, PA, USA).

To knockdown CFTR in the ovarian cells, four human CFTR-specific small hairpin RNA (shRNA) expression vectors in pGFP-V-RS plasmid were purchased from OriGene (TG313958; OriGene Technologies, Inc., Rockville, MD, USA). Cells transfected with non-effective shRNA cassette in pGFP-V-RS plasmid were used as a negative control. Vector DNA (1 μg) was transfected into A2780 and SKOV3 cells by PolyJet™ transfection reagent (SignaGen Laboratories, Ijamsville, MD, USA). At two days after transfection, 2.5×10⁶ cells were digested and replanted into five 10-cm dishes in full medium containing puromycin at concentration of 0.9 μg/ml for A2780 and 0.5 μg/ml for SKOV3 cells, and selected for 2–3 weeks. Puromycin-resistant and GFP-positive cell colonies were separated and verified by western blotting. The stably transfected cell lines were then cultured in medium containing 0.5 μg/ml of puromycin for A2780 and 0.25 μg/ml for SKOV3 cells for subsequent studies.

For cell invasion and migration assays, a cell invasion assay kit (8-μm pore size; Cell Biolabs Inc., San Diego, CA, USA) was used according to the manufacturer’s instructions. Approximately 1×10⁵ cells were placed into the Transwell insert with serum-free media. After 24–48 h, the cells on the inner layer of the Transwell insert were wiped away, then inserts were fixed and stained, and photographed with a light microscope under high magnification objective. Each insert was then transferred to an empty well, and extracted with extraction solution for 10 min on an orbital shaker. Extraction solution (100 μl) from each sample was transferred to a 96-well microtiter plate and OD 560 nm was measured using a microplate reader (Thermo Scientific, waltham, MA, USA). The procedure was repeated independently three times with triplicate inserts for each group.

We seeded 200 cells into 6-well culture plates in triplicate. The cells were collected after 14 days and then stained with Giemsa stain. The colony containing >50 cells was counted under the microscope.

We pre-coated 96-well plates with 100 μl Matrigel at concentration of 0.04 μg/ml (BD Matrigel, Franklin Lakes, NJ, USA). Prior to cell seeding, the plates were incubated in serum-free medium at 37°C for 30 min for rehydration. Cells (3×10⁴) were seeded in 100 μl growth medium in triplicate and incubated for 1 h. After rinsing, the adherent cells were fixed using 4% formaldehyde for 10 min, and stained by crystal violet. Adherent cell staining was extracted by acetic acid and quantified using a microplate reader (Thermo Scientific) at OD 490 nm.

For the proliferation assay, 500 cells were seeded in 96-well plates in triplicate and cultured for days 1, 2, 3, 4 and 5. The ST-8 Cell Counting Kit (KeyGen Biotech, Nanjing, China) and a modified 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay method, were applied to estimate the proliferation of viable cells. The cells were collected at specific time-points (6, 24, 48, 72, 96 and 120 h) and incubated for an additional 2 h in 10% WST-8 solution at 37°C. Absorbance was measured at 450 nm using a microplate reader (Thermo Scientific). The procedure was repeated independently three times.

To investigate tumorigenicity, 5-week-old athymic female nude mice were provided by the Institution Animal Care Committee at Chongqing Medical University. Approximately 4×10⁶ cells in 100 μl suspension of SKOV3-CFTRi cells (7 implants) with PBS and SKOV3-NC cells (cells transfected with non-effective shRNA cassette) as a control (7 implants) were injected subcutaneously. The tumor size was measured on a weekly basis and monitored over 5 weeks. At the end of the experiment, the mice were euthanized and the xenografts collected for subsequent analysis. The tumor size was calculated as: 0.5234 × [long diameter (short diameter)²]. The experimental procedure was approved by the Institution Animal Care Committee at Chongqing Medical University.

Independent Student’s t-test was used to make a comparison of the data between two groups. Relevance analysis between CFTR immunostaining and clinicopathological characteristics was performed by Pearson’s χ² or Fisher’s exact tests. Statistical analysis was carried out by SPSS 17.0 software (SPSS, Inc., Chicago, IL, USA). P<0.05 (two-sided) was considered to indicate a statistically significant result.

CFTR staining was predominantly detected in the cell membrane and cytoplasm (Fig. 1A). The expression of CFTR in ovarian cancer specimens was significantly higher than that in benign (P<0.05) and normal ovaries (P<0.05, Fig. 1B). Moreover, in a majority of benign tumors (77.8%) and normal ovaries (90.1%), it was difficult to detect CFTR, which was in contrast to the cancer group (Fig. 1B). Regarding the clinicopathological characteristics, the statistical analysis revealed that the CFTR protein level was well-related to advanced clinical stages (stage III/IV vs. I/II, P<0.05), poor histological grade (grade 2–3 vs. 1, p<0.05), and a higher serum Ca-125 level (P<0.05, Table I). Furthermore, we observed that CFTR staining was stronger in the serous type and clear cell type as compared to the remaining types (P<0.05, Table I).

To examine the role of CFTR in serous ovarian cancer cells, we knocked down CFTR in SKOV3 and A2780 cells. The cells were transfected with CFTR-specific shRNA expression pGFP-V-RS vectors and the stably transfected cells were selected (referred to as CFTRi in Fig. 2A and B). Non-effective shRNA-infected cells were designated as a negative control, referred to as SKOV3-NC and A2780-NC. Western blot results revealed that the CFTR protein level of CFTRi cells was significantly reduced compared with that of its corresponding control cells (Fig. 2C). Thus, SKOV3-CFTRi3 and A2780-CFTRi2 were used in the subsequent experiments.

SKOV3-CFTRi3 cells exhibited an impaired migration capacity of 60.89% compared to SKOV3-NC cells (P<0.05) (Fig. 3A and B). Similarly, A2780-CFTRi2 cells had a reduced migration capacity of 63.93% compared to A2780-NC cells (P<0.05). For SKOV3-CFTRi3 and A2780-CFTRi2, an ~41.72 and 63.11% reduction in the number of migratory cells was observed compared with SKOV3-NC and A2780-NC cells (P<0.05 for both, Fig. 3C and D).

As shown in Fig. 3E, there was a 42.31 and 38.52% decreased percentage of optical density (OD) value in SKOV3-CFTRi3 and A2780-CFTRi2 cells compared to the control cells (P<0.05 for both).

The results of the WST-8 assay (Fig. 4C and D) showed that the growth ability of SKOV3-CFTRi3 and A2780-CFTRi2 cells decreased compared with the corresponding control cells. Moreover, CFTR knockdown cells also had a low colony formation ability as the number of colonies of CFTRi cells were decreased when compared with the corresponding control cells (Fig. 4A and B).

Cells were subcutaneously injected into the flanks to form xenograft tumors in nude mice. We measured the size of xenograft tumors once a week for a total of five weeks. We observed that the tumor size of the SKOV3-CFTRi3 group was statistically significantly smaller than that of the SKOV3-NC group from the third week until the end of the experiment (Fig. 5A). The size of xenograft tumors of SKOV3-CFTRi3 (183±61 cm³) was significantly smaller than that of the control (713±270 cm³) at the end of experiment (Fig. 5B and C). At the same time, the average tumor weight of SKOV3-CFTRi3 group was lower than that of the control group (Fig. 5D). Notably, these results showed that CFTR knockdown affected the tumor size significantly rather than the incidence of tumor formation, suggesting that CFTR knockdown suppresses tumor progression rather than tumor initiation in vivo.