A total of 41 patients with OC were recruited to this study between March 2015 and September 2018. A total of 41 paired tissues of OC and adjacent normal tissues (>3 cm from tumor sites) were acquired from these patients with OC (age range 18–60 years old) subjected to surgical excision in the Shengli Oilfield Central Hospital, after obtaining informed consents from the patients. These tissues were divided into two groups (I+II: n=23; III+IV: n=18) according to the tumor stage. All specimens were promptly stored in liquid nitrogen for subsequent RNA or protein extraction. The present study was approved and supported by the Shengli Oilfield Central Hospital.

Human ovarian surface epithelial cells (IOSE-80) and OC cell line OVCAR-3 were bought from American Type Culture Collection and another OC cell line, SKOV3, was obtained from the cell bank of the Chinese Academy of Sciences (Shanghai, China). These cells were cultured in basic medium RPMI 1640 (Invitrogen; Thermo Fisher Scientific, Inc.) complemented with 10% fetal bovine serum (FBS; Gibco; Thermo Fisher Scientific, Inc.) in a humidified incubator with a constant temperature of 37°C and CO₂ concentration at 5%.

Small interfering (si) RNA against OIP5-AS1 (si-OIP5-AS1, 5′-GGCTTTGTGTTCCTTATCACAGG-3′), siRNA against CCNG1 (si-CCNG1, 5′-TTGAAGTAAAAGATCTTCTTAGT-3′), siRNA negative control (si-control, 5′-TTCTCCGAACGTGTCACGTTT-3′), short hairpin (sh) RNA vector against OIP5-AS1 (sh-OIP5-AS1, 5′-CCGGGCTCCTAGGATTCCAGTTATCCTCGAGGCAGAAGGCTGAGTTTCATTTTTTTTG-3′), shRNA vector negative control (sh-control, 5′-CACCGTTCTCCGAACGTGTCACGTCAAGAGATTACGTGACACGTTCGGAGAATTTTTTG-3′), miR-128-3p mimic (miR-128-3p, 5′-UCACAGUGAACCGGUCUCUUU-3′), mimic negative control (miR-control, 5′-UUGUACUACACAAAAGUACUG-3′), miR-128-3p inhibitor (anti-miR-128-3p, 5′-AAAGAGACCGGUUCACUGUGA-3′), inhibitor negative control (anti-control, 5′-CAGUACUUUUGUGUAGUACAA-3′) and the overexpression vector pCE-RB-Mam-OIP5-AS1 (OIP5-AS1) vector or the empty vector pCE-RB-Mam were all purchased from Guangzhou RiboBio Co., Ltd. Cell transfection was performed using Lipofectamine^® 3000 (Invitrogen; Thermo Fisher Scientific, Inc.). Cells were seeded into the 24-well plates and cultured overnight to 50% coverage, then transfected with different concentrations of RNAs or vectors (40 nM siRNA, 40 nM shRNA vector, 40 nM mimic, 20 nM inhibitor or 2 µg pCE-RB-Mam/OIP5-AS1 vector). Further experimentation was performed after cells were incubated at 37°C for 48 h.

The extraction of total RNA was performed using TRIzol^® (Thermo Fisher Scientific, Inc.) from OC tissues and cells (2×10⁵) in accordance with the manufacturer's protocol, followed by the quantification of RNA concentration. Then 1 µg RNA was reversely transcribed to synthesis the complementary DNA (cDNA) using Prime Script RT regent kit (Takara Bio, Inc.) according to the manufacturer's protocols. PCR reaction system was prepared in 384 well-plates using a TB Green Premix EX Taq II kit (Takara Bio, Inc.) following the manufacturer's protocol and amplified via ABI Step One Real-time PCR System (Thermo Fisher Scientific, Inc.). The reaction system was prepared on ice with a 20-µl volume: 10 µl TB Green Premix EX Taq II (2×), 0.8 µl forward primer, 0.8 µl reverse primer, 0.4 µl ROX Reference Dye, 2 µl cDNA, 6 µl sterile water. Three technical replications were set for each sample. The reaction protocols were as follows: Predenaturation at 95°C for 30 sec, 40 cycles of denaturation at 95°C for 5 sec and annealing at 60°C for 30 sec. GAPDH and U6 were used as respective internal control for lncRNA or mRNA and miRNA. Primers were synthesized by Sangon Biotech Co., Ltd. and the sequences were: OIP5-AS1 (forward: 5′-TGCGAAGATGGCGGAGTAAG-3′, reverse: 5′-TAGTTCCTCTCCTCTGGCCG-3′); miR-128-3p (forward: 5′-GCCGAGTCACAGTGAACCGGT-3′, reverse: 5′-CAGTGCAGGGTCCGAGGTAT′); CCNG1 (forward: 5′-GTTACCGCTGAGGAGCTGCAGTC-3′, reverse: 5′-ATAGCCATCATGGATAGACTCAG-3′); GAPDH (forward: 5′-GTCTCCTCTGACTTCAACAGCG-3′, reverse: 5′- ACCACCCTGTTGCTGTAGCCAA-3′); U6 (forward: 5′-ATTGGAACGATACAGAGAAGATT-3′, reverse: 5′-GGAACGCTTCACGAATTTG-3′). The collected data from three independent experiments were analyzed using the 2^−∆∆Cq method (30).

Cell viability was examined by MTT assay. At 0, 24, 48 and 72 h post-transfection, 15 µl MTT solution (5 mg/ml, pH=7.4; Thermo Fisher Scientific, Inc.) was added to cells in the 96-well plates and incubated for another 4 h. Then the supernatants were slowly removed and 150 µl dimethylsulfoxide (DMSO; Thermo Fisher Scientific, Inc.) was added to each well for 10 min to resolve the generated formazan. The optical density (OD) value of each well at the wavelength of 490 nm was determined using a microplate reader (Thermo Fisher Scientific, Inc.).

An FITC Annexin V Apoptosis Detection kit I (BD Biosciences) was used to distinguish the apoptotic cells. After 48 h post-transfection, OVCAR-3 and SKOV3 cells were collected following digestion by 0.25% trypsin (GIBCO), then cell pellets were washed by pro-cooled phosphate buffered saline (PBS; GIBCO) and resuspended in 1X binding buffer. Afterwards, cells were stained by 5 µl FITC Annexin V and 5 µl PI for 15 min in the dark at room temperature according to the manufacturer's protocol. The apoptotic cells were examined using a BD Accuri C6 flow cytometer (BD Biosciences) and analyzed using the BD Accuri C6 system (32-bit) software (BD Biosciences). The apoptosis rate was calculated as the percentage of early apoptotic cells (Annexin⁺/PI⁻) and late apoptotic cells (Annexin⁺/PI⁺).

A Transwell 24-well chamber (Corning Life Sciences) was used to detect the abilities of cell migration and invasion. Cell suspension in serum-free medium was pipetted into the upper chamber at room temperature. Simultaneously, 600 µl RPMI 1640 medium with 10% FBS was added to the lower chamber at room temperature. The chamber was incubated at 37°C for 24 h, and cells that passed through the lower membranes were fixed with methyl alcohol (Sangon Biotech Co., Ltd.) for 10 min and stained with crystal violet (Sangon Biotech Co., Ltd.) for 10 min at room temperature. Finally, the cells were counted under an inverted light microscope (Olympus Corporation) and cell images were acquired at ×100 magnification. For invasion assay, the lower surface of the upper chamber was coated with Matrigel (Corning Life Sciences) at 37°C overnight before seeding the cells.

According to the manufacturer's protocol, the consumption of glucose and the production of lactate were assessed using glucose detection kit (cat. no. K188-200) and lactate detection kit (cat. no. K209-100) purchased from Beijing Zhonghao Biotechnology Co., Ltd.).

Total proteins from tissues and cells were extracted by RIPA Lysis Buffer (Sangon Biotech Co., Ltd.). The obtained proteins were quantified using BCA Protein Assay kit (Sangon Biotech Co., Ltd.) and 30 µg proteins were separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) on 10% gels for 2 h. Next, the separated proteins were transferred onto the polyvinylidene fluoride (PVDF) membranes (Thermo Fisher Scientific, Inc.). These membranes were then immersed in 5% skimmed milk (Sangon Biotech Co., Ltd.) diluted with phosphate buffered saline and 0.1% Tween 20 (PBST) at room temperature for 3 h to prevent the non-specific protein binding, followed by incubation with primary antibodies overnight at 4°C. The primary antibodies were: anti-hexokinase 2 (anti-HK2; Abcam; cat. no. ab209847; 1:1,000), anti-CCNG1 (Abcam; cat. no. ab170389; 1:1,000) and anti-β-actin (Abcam; cat. no. ab8227; 1:3,000). Finally, the chromogenic reaction was performed by an HRP/DAB (ABC) detection kit (Abcam) after the incubation of anti-rabbit secondary antibody (Abcam; cat. no. ab205718; 1:5,000) for 1 h at room temperature. The gray level of each band was obtained by the statistical analysis of ImageJ software (National Institutes of Health) using β-actin as the endogenous reference.

The online software applications StarBase (http://starbase.sysu.edu.cn) and TargetScan (http://www.targetscan.org) were used to seek the miRNA target of OIP5-AS1 and target gene for miR-128-3p, respectively. To construct the recombinant luciferase plasmids, the wild-type (WT) OIP5-AS1 and 3′UTR of CCNG1 sequences were respectively cloned into the pGL-3 luciferase basic vector (Promega Corporation) to acquire WT-OIP5-AS1 and WT-CCNG1, then their mutant-types (MUT) with the mutant sites for miR-128-3p were also constructed as MUT-OIP5-AS1 and MUT-CCNG1. Each constructed luciferase reporter plasmid was transfected into OVCAR-3 and SKOV3 cells with miR-128-3p or miR-control by Lipofectamine 3000. After transfection for 48 h, cells were lysed and the luciferase intensity was determined by the dual-luciferase reporter system (Promega Corporation) following the manufacturer's protocol. The luciferase activity ratio of firefly and Renilla was considered to be the relative luciferase activity.

RIP assay was performed using Magna RNA immunoprecipitation kit (EMD Millipore). OC cells were centrifuged at 4°C and 16,000 × g for 10 min and 2×10⁷ cells were lysed using RIPA lysis buffer (EMD Millipore). The cells were then incubated with magnetic beads pro-covered with antibodies against Argonaute2 (Anti-Ago2; Abcam; cat. no. ab32381) using anti-Immunoglobulin G (Anti-IgG; Abcam; cat. no. ab205718) as the negative reference. RNA was extracted using TRIzol and the RNA enrichment was detected via RT-qPCR. Finally, the levels of OIP5-AS1 and miR-128-3p in Anti-IgG and Anti-Ago2 groups were compared.

Pierce™ Biotinylated Protein Interaction Pull-Down kit (Thermo Fisher Scientific, Inc.) was used to perform the pull-down assay. OVCAR-3 and SKOV3 cells were transfected with biotin-labeled miR-128-3p (Bio-miR-128-3p) or Bio-NC (Guangzhou RiboBio Co., Ltd) using Lipofectamine 3000. Following transfection at 37°C for 48 h, cells were harvested by centrifugation at 4°C and 16,000 g for 10 min. The cell lysates were then incubated with streptavidin-coupled agarose beads at 4°C overnight. Following the isolation of RNA from the washed beads by TRIzol, the enrichment of OIP5 was measured by RT-qPCR analysis.

A total of 12 BALB/c nude mice (six-week-old, 20–25 g) were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. and maintained in a specific-pathogen-free environment under temperature of 25°C, humidity of 70% and a 12 h light/dark cycle. These mice received food and water ad libitum. Mice were arbitrarily divided into two groups with 6 mice/group, comprising sh-control and sh-OIP5-AS1 groups. The xenograft tumor model was established through the subcutaneous injection of 2×10⁶ OVCAR-3 cells stably expressed sh-OIP5-AS1 or sh-control into the left flank of mouse's back. The health and behavior of mice were monitored every 2 days and tumor volume (length × width² × 0.5) was measured using a digital caliper every week. After cell injection for 4 weeks, tumor volume reached 900 mm³ (<1,000 mm³, which was set as the humane endpoint) and the mice were sacrificed. The 30% air of environment (per min) was displaced using the flow rate of CO₂ according to the current guidelines of the American Veterinary Medical Association (31), then mice were verified to have succumbed by monitoring the breathing. No mice died during the 4-week experimental period. The tumor tissues were excised from mice and weighed on an electronic scale. The examination of OIP5-AS1, miR-128-3p and CCNG1 was performed via RT-qPCR or western blotting. This experiment was ratified by the Animal Ethics Committee of the Shengli Oilfield Central Hospital and all procedures followed the guidelines provided by the National Institutes of Health for the Care and Use of Laboratory Animals, 8th edition (32).

All data were given as the mean ± standard deviation from three independent experiments. SPSS 19.0 (IBM Corp.) was used to conduct statistical analysis and figure plotting was performed using GraphPad Prism 7 (GraphPad Software, Inc.). The differences of two groups and more than two groups were analyzed through Student's t-test and a one-way analysis of variance followed by Tukey's post hoc test. P<0.05 was considered to indicate a statistically significant difference.

RT-qPCR analysis was firstly conducted to examine the expression of OIP5-AS1. OIP5-AS1 expression was markedly increased in OC tissues compared with the adjacent normal tissues (Fig. 1A). Among 41 OC tissues, OIP5-AS1 was upregulated in tissues at III+IV stage compared with those tissues at I+II stage (Fig. 1B). OIP5-AS1 expression was higher in both OVCAR-3 and SKOV3 cells compared with normal human ovarian IOSE-80 cells (Fig. 1C). This dysregulation of OIP5-AS1 implied its vital role in OC.

OVCAR-3 and SKOV3 cells were transfected with si-OIP5-AS1 or si-control to investigate the function of OIP5-AS1 in OC. As shown in Fig. 2A, the decreased OIP5-AS1 expression in si-OIP5-AS1 group compared with the si-control group demonstrated that OIP5-AS1 was successfully knocked down in the two cell lines. MTT assay demonstrated that cell viabilities of OVCAR-3 and SKOV3 cells were distinctly reduced following the introduction of si-OIP5-AS1 compared with si-control group (Fig. 2B), while the apoptosis rate was shown to be increased (Fig. 2C). Transwell assay suggested that the migrated and invaded cells were decreased after downregulation of OIP5-AS1 (Fig. 2D and E). Analysis of glycolysis was performed by the detection of glucose consumption and lactate production, as well as the protein level of HK2 (an important enzyme of glycolysis). The data revealed that the transfection of si-OIP5-AS1 suppressed the consumption of glucose (Fig. 2F), lactate production (Fig. 2G) and HK2 protein expression (Fig. 2H) in both OVCAR-3 and SKOV3 cells, indicating that OIP5-AS1 knockdown caused the inhibition of glycolysis. These results suggested that knockdown of OIP5-AS1 repressed the progression of OC.

LncRNAs exert their regulatory effects through sponging miRNAs generally (33). As demonstrated in Fig. 3A, StarBase predicted that OIP5-AS1 contained the binding sequences (UCACUGUG) for miR-128-3p and these sites were mutated into AGUGACAC in mutant OIP5-AS1. After performing the dual-luciferase reporter assay, it was found that the luciferase activity of WT-OIP5-AS1 plasmid was definitely inhibited by miR-128-3p mimic (compared with the miR-control group) while there was no obvious change in MUT-OIP5-AS1 group of OVCAR-3 and SKOV3 cells (Fig. 3B). RIP assay demonstrated that OIP5-AS1 and miR-128-3p were enriched in Ago-2 pellet compared with the IgG group (Fig. 3C). Moreover, the enrichment of OIP5-AS1 was much higher in Bio-miR-128-3p group compared with the Bio-NC group, suggesting that OIP5-AS1 was pulled down by miR-128-3p (Fig. 3D). Subsequently, the miR-128-3p expression in OC cells was examined and the results indicated that miR-128-3p level was downregulated in OVCAR-3 and SKOV3 cells compared with normal IOSE-80 cells (Fig. 3E). After verifying the successful overexpression effects on miR-128-3p and OIP5-AS1 by miR-128-3p mimic and OIP5-AS1 vector (Fig. 3F and G), it was noticed that miR-128-3p overexpression could recover the OIP5-AS1-induced miR-128-3p suppression (Fig. 3H). The results demonstrated that OIP5-AS1 could directly sponge miR-128-3p.

To explore the regulatory mechanism of OIP5-AS1 and miR-128-3p, OVCAR-3 and SKOV3 cells were transfected with miR-128-3p, miR-128-3p+OIP5-AS1 or their matched controls. MTT and flow cytometry demonstrated that miR-128-3p had a suppressive effect on cell viability (Fig. 4A and B) and a promoting effect on apoptosis (Fig. 4C and D) in OVCAR-3 and SKOV3 cells, whereas the overexpression of OIP5-AS1 reversed these effects. Cell migration (Fig. 4E and F) and invasion (Fig. 4G and H) were inhibited by miR-128-3p transfection, which was abrogated following the upregulation of OIP5-AS1. OIP5-AS1 transfection counteracted the inhibition of glucose consumption (Fig. 4I and J) and lactate production (Fig. 4K and L) caused by miR-128-3p. The miR-128-3p-induced downregulation of HK2 protein level was clearly alleviated by the promotion of OIP5-AS1 expression (Fig. 4M and N). Hence, the inhibitory effect of miR-128-3p on OC progression was abated by OIP5-AS1 overexpression.

TargetScan software predicted the binding domain of miR-128-3p in the 3′-UTR sequence of CCNG1 (Fig. 5A), indicating that CCNG1 might be a potential target of miR-128-3p. Dual-luciferase reporter assay proved the combination of miR-128-3p and CCNG1 because the overexpression of miR-128-3p markedly repressed the luciferase activity of WT-CCNG1 group but failed to decrease that of MUT-CCNG1 group (Fig. 5B). Then RT-qPCR and western blotting revealed that the mRNA and protein expression levels of CCNG1 were higher in OVCAR-3 and SKOV3 cells than those in normal IOSE-80 cells (Fig. 5C and D). The RT-qPCR analysis showed that the repressive impact of anti-miR-128-3p on the expression of miR-128-3p was significant (Fig. 5E). Subsequently, the introduction of miR-128-3p was presented to suppress the CCNG1 mRNA and protein levels, while the opposite effects were observed following anti-miR-128-3p transfection (Fig. 5F and G). Collectively, miR-128-3p directly targeted CCNG1.

To explore whether CCNG1 was associated with the influence of miR-128-3p on OC development, transfection of si-CCNG1, si-CCNG1 + anti-miR-128-3p or the relative controls was conducted in OVCAR-3 and SKOV3 cells. Western blotting indicated that si-CCNG1 transfection markedly decreased the protein level of CCNG1, while miR-128 inhibitor relieved this expression downregulation in OVCAR-3 and SKOV3 cells (Fig. S1). Subsequent experiments demonstrated that anti-miR-128-3p partly abolished the si-CCNG1-induced cell viability repression (Fig. 6A and B), apoptosis enhancement (Fig. 6C and D) and migration (Fig. 6E and F) or invasion (Fig. 6G and H) inhibition. Additionally, the decline of glucose consumption (Fig. 6I and J), lactate production (Fig. 6K and L) and HK2 protein level (Fig. 6M and N) caused by CCNG1 knockdown was also weakened following the downregulation of miR-128-3p. Above data clarified that miR-128-3p inhibition lightened the effects of CCNG1 knockdown on OC cells, hinting that CCNG1 inhibition was responsible for the anti-cancer effect of miR-128-3p on OC.

To ascertain whether CCNG1 could be regulated by OIP5-AS1, OVCAR-3 and SKOV3 cells were respectively transfected with miR-control, miR-128-3p, miR-128-3p+vector or miR-128-3p+OIP5-AS1, followed by the analysis of RT-qPCR and western blotting. As shown in Fig. 7A and B, miR-128-3p transfection resulted in reducing the CCNG1 mRNA and protein levels, which was ameliorated by ectopic high expression of OIP5-AS1. In sum, OIP5-AS1 could upregulate the CCNG1 level by targeting miR-128-3p.

The xenograft tumor model was constructed to explore the effect of OIP5-AS1 on OC in vivo. As Fig. 8A and B demonstrate, tumor volume and weight were significantly lower in sh-OIP5-AS1 group compared with the sh-control group, implying that OIP5-AS1 knockdown restrained the OC carcinogenesis in vivo. In contrast to the sh-control group, the expression of OIP5-AS1 was downregulated (Fig. 8C) but miR-128-3p was upregulated (Fig. 8D) in sh-OIP5-AS1 group according to the results of RT-qPCR. CCNG1 protein level was decreased with the downregulation of OIP5-AS1 in western blot assay (Fig. 8E). Therefore, OIP5-AS1 knockdown impeded the oncogenesis of OC by the miR-128-3p/CCNG1 axis in vivo.