Utilization of all EOC and normal tissue specimens was approved by our hospital Internal Review Board (IRB) in Chinese PLA General Hospital. Before treatment by radiotherapy or chemotherapy, 48 human EOC tissue specimens were obtained by surgical resection. Twenty-two normal human ovarian tissue specimens were collected also by surgical resection. All tissue specimens were stored at −70°C.

The OVCAR-3 and SK-OV-3 EOC cell lines were provided by the cell resource center of Chinese Academy of Medical Sciences and were grown respectively in RPMI-1640 medium (Invitrogen, Carlsbad, CA, USA) and L-15 medium (Gibco, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (FBS; Gibco, Rockville, MD, USA), 50 μg/ml penicillin and 50 μg/ml streptomycin. The cells were incubated at 37°C, with 5% CO₂. For hypoxia treatment, cells were placed in a hypoxia incubator infused with a gas mixture of 5% CO₂ and nitrogen to obtain 3% oxygen concentration. Oxygen concentration was monitored continuously (Forma 3130; Thermo Scientific, Rockford, IL, USA). siHIF-1α, siHIF-2α and siRNA control oligos were synthesized by GenePharma Technology (Shanghai, China) and were transfected into OVCAR-3 cells with at a concentraction of 40 nM of lipofectamine 2000 to suppress the HIF-1α or HIF-2α. miR-210 mimics (miR con as control) (Qiagen, Valencia, CA, USA) and miR-210 inhibitor (Sigma-Aldrich, St. Louis, MO, USA) were utilized to manipulate the miR-210 level. A total of 20 or 40 nM miR-210 mimics/miR con, or 10 nM miR-210 inhibitor was transfected into EOC cells with lipofectamine 2000. 5-FU (Sigma-Aldrich) was utilized to induce EOC apoptosis with a concentration of 5 or 40 μg/ml.

Total mRNA was extracted from tissue or cell samples by using the RNeasy mini kit (Qiagen). RT-qPCR analysis of the HIF-1α, HIF-2α, CASP3 or CASP9 expression in mRNA level was performed using SYBR-Green with the LightCycle 2.0 (Roche, Mannheim, Germany). All mRNA expression levels were normalized to GAPDH (glyceraldehyde-3-phosphate dehydrogenase). miRNA extraction was performed using the mirVana miRNA isolation kit (Ambion, Austin, TX, USA). Quantification of miR-210 expression was conducted using the mirVana qRT-PCR miRNA detection kit (Ambion), and the U6 small nuclear RNA was used as internal control. ΔΔCt method was used for relative quantification (43). A non-radioactive northern blot method, LED, for small RNA (about 15–40 bases) detection using digoxigenin (DIG)-labeled oligonucleotide probes containing locked nucleic acids (LNA) and 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide was utilized to confirm the miR-210 and U6 expression, according to the protocol (44).

Whole cell extracts were prepared with a cell lysis reagent (Sigma-Aldrich) according to the manual; CASP3 or CASP9 were detected by western blot analysis using anti-CASP3, anti-CASP9 or anti-GAPDH rabbit polyclonal antibody (1:500; Sigma-Aldrich). Goat anti-rabbit IgG conjugated to horseradish peroxidase (Pierce, Rockford, IL, USA) and ECL detection systems (Super Signal West Femto; Pierce) were used for detection.

For cell colony formation assay, 5×10² cells were incubated in 12-well plates at 37°C containing 5% CO₂, and were transfected with 0 or 40 nM miR-210 mimics or miR-210 control, 5–8 days post incubation; the cells were stained with crystal violet (0.005%) for 20 min and recorded the colony numbers by imaging J software. For proliferation assay, post transfecting with miR-210 mimics or miR-210 control, cells were incubated in CCK-8 (Dojindo, Kumamoto, Japan). The 450 nm absorbance of each well was detected after visual color occurrence at 24, 48 or 72 h. OVCAR-3 or SK-OV-3 cell apoptosis was examined with an Annexin V-FITC apoptosis detection kit (Sigma-Aldrich). Briefly, 1–5×10⁵ cells were stained with Annexin V-FITC and propidium iodide and detected by a FACScan flow cytometer (BD Biosciences, Franklin Lakes, NJ, USA) to analyze cellular apoptosis. The results were calculated using CellQuest™ Pro software (BD Biosciences) and expressed as the percentage of apoptotic cells from the total cells.

For luciferase reporter experiments, according to the prediction to interact with miR-210, the 3′-UTR or the mutated 3′-UTR of PTPN1 was amplified by PCR from human genomic DNA and inserted into the MluI and HindIII sites of pGL3 vector immediately downstream from the stop codon of luciferase. OVCAR-3 cells were cotransfected in 12-well plates with 0.4 μg of the firefly luciferase report vector and 0.1 μg of the control vector containing Renilla luciferase, pRLTK (Promega, Madison, WI, USA) to eliminate differences in cell number and transfection efficiency, as well as with or without 40 nM of miR-210 mimics or miR-con. At 24 h post-transfection, firefly and Renilla luciferase activities were measured consecutively using dualluciferase assays (Promega). Experiments were carried out in triplicate and means were determined.

Statistical analyses were performed using SPSS16.0 software (IBM SPSS, Armonk, NY, USA). Correlations between miR-210 expression and the various clinical and laboratory data of patients with HIF-1α were analyzed using the Spearman rank correlation. The HIF-1α, HIF-2α, CASP3 or CASP9 expression in mRNA level, the CASP3 or CASP9 expression in protein level, or miR-210 expression between two groups were analyzed by Student’s t-test. A p-value ≤0.05 was considered statistically significant.

We detected the expression of miR-210 in epithelial ovarian cancer tissues, compared to the normal ovarian tissues. Forty-eight primary EOC patients with a mean (± SD) age of 57.2±7.1 years at diagnosis were included in the study. Of these, five were FIGO stage I, 12 were stage II, 22 were stage III and seven were stage IV. Ovarian tissue samples from 22 healthy age-matched volunteers were used as controls. The mean (± SD) ΔΔCT value was 2.77±1.3 in the 48 samples from patients with EOC and 1.00±0.42 in healthy controls (p<0.01; Fig. 1A). Thus, the miR-210 in ovarian tissue was significantly upregulated in patients with EOC. The miR-210 overexpression was correlated with the tumor stage or post-operative residual tumor size (Table I; p<0.01 and p<0.05, respectively). To reconfirm the significant miR-210 upregulation in EOC, the miRNA samples were examined by northern blot analysis, and as shown in Fig. 1B the OC group was significantly higher than the normal group (p<0.05).

miR-210 is induced by hypoxia via HIF-1α (36,37). To assess the hypoxic regulation of hsa-miR-210 in EOC specimens, the HIF-1α mRNA expression in the EOC and normal specimens was also determined by RT-qPCR. As shown in Fig. 1C, a significant high level of HIF-1α mRNA was also confirmed (p<0.01). The correlation between miR-210 and of HIF-1α mRNA expression was explored. In EOC patients, miR-210 expression showed a positive correlation with HIF-1α mRNA (R²=0.3054, p<0.01) (Fig. 1D). To reconfirm the correlation of miR-210 expression with hypoxia in EOC, the miRNA samples from OVCAR-3 and SK-OV-3 under normoxia or hypoxia, were analyzed using quantitative-PCR (Q-PCR) too. In agreement with results of clinical specimens, we found a substantial and significant induction of miR-210 expression in two kinds of epithelial ovarian cancer cell lines under hypoxia (Fig. 2A and B). The in vitro results were also be reconfirmed by a HIF knockdown experiment. Effective siRNAs targeting HIF-1α or HIF-2α (Fig. 2C) abrogated the miR-210 induction by hypoxia (p<0.01 and p<0.05, respectively).

To determine the possible contribution of miR-210 to EOC cell proliferation, we manipulated the miR-210 expression level in OVCAR-3 and SK-OV-3 cell lines by transfecting with miR-210 mimics or miRNA control. The significant increase of miR-210 in OVCAR-3 and SK-OV-3 cells post transfection with miR-210 mimics is shown in Fig. 3A (both p<0.01). Then, the proliferation of OVCAR-3 and SK-OV-3 cells post-miR-210 mimics or miRNA control was tested by CCK-8 assay. Both in OVCAR-3 and in SK-OV-3 cells, the miR-210 mimics promoted cell proliferation rather than miRNA control time-dependently (Fig. 3B and C). We also detected the differences in colony formation of OVCAR-3 and SK-OV-3 cells transfected with miR-210 mimics or miRNA control. Fig. 4 shows the higher capability of colony formation for both cell lines post-transfection with miR-210 mimics than post-transfection with miRNA control. The above findings demonstrated that upregulated miR-210 enhanced the proliferative capability and colony formation of EOC cells in vitro.

Hypoxia induces apoptosis via HIF-1α (45–47), in various tumor cells, including ovarian cancer (48). The most direct induction of hypoxia-induced apoptosis is the inhibition of the electron transport chain at the inner membrane of the mitochondria. We tested the sensitivity of OVCAR-3 or in SK-OV-3 ovarian cancer cells to 5-FU. At normoxia culture conditions, the 5-FU induced low level of apoptosis with a slight time-dependent increase, while hypoxia costimulates higher level of apoptosis with 5-FU in both cell types (Fig. 5A and B). We also tested in OVCAR-3 cells, the expression of caspase 3 (CASP3) and caspase 9 (CASP9), both of which are executional molecules in apoptosis. The western blot analysis results demonstrated that significantly high levels of CASP3 and CASP9 were induced by a low dose of 5-FU under hypoxic culture condition rather than in normoxia (Fig. 5C and D). However, the hypoxia-induced apoptosis could be reversed by miR-210 inhibitor, the increase in apoptotic rate and the expression level of CASP3 or CASP9 were blocked by the miR-210 inhibitor transfecion (Fig. 5E–H).

To unveil the anti-apoptosis effect of miR-210, we predicted the possible targets of miR-210 by miRanda. PTPN1 is one of the screened targets, and has not been determined to play roles in ovarian cancer, though it is well known to induce apoptosis (49–52). To evaluate the possible targeting of PTPN1 by miR-210 in ovarian cancer cells, the PTPN1 expression in mRNA and protein levels was determined with the miR-210 mimics or miRNA control transfection in OVCAR-3 or SK-OV-3 cells. It was shown that the proapoptotic protein was downregulated after transfection with miR-210 mimics while there was no effect for the control transfection (Fig. 6A and B). Fig. 6C shows that PTPN1 had a consequential pairing with the miR-210. To reconfirm the PTPN1 downregulation by miR-210, we construct a luciferase reporter constructs containing the miR-210 recognition sequence of PTPN1 mRNA (Fig. 6C). As shown in Fig. 6D, transfection with the miR-210 mimics suppressed, while the miRNA control did not regulate the activity of pGL3-PTPN1 reporter in OVCAR-3 cells; and the miR-210 mimics had no suppression effect on the luciferase activity of pGL3-PTPN1_mut reporter. These results agree with the fact that miR-210 regulates PTPN1 by targeting the 3′-UTR and inducing translation repression of the gene. Therefore, suppression of the particular gene contributes to the improvement of ovarian cancer by inhibiting apoptosis of ovarian cancer cells under hypoxia.