GSE119056 (25) and GSE23392 (26) downloaded from Gene Expression Omnibus (GEO) datasets (27) included the mRNA expression profiles. With log fold change (logFC) ≥1.5 and P-value <0.01, the upregulated differentially expressed genes (DEGs) were screened out. Then, the overlapping DEGs from GSE119056 and GSE23392 were uploaded to STRING to perform the networking analysis. Gene Expression Profiling Interactive Analysis (GEPIA, http://gepia.cancer-pku.cn/index.html) was used to show the expression of RHPN1-AS1, our gene of interest, in ovarian cancer. Finally, the key miRNA binding to RHPN1-AS1, our gene of interest, was found by ENCORI Starbase (http://starbase.sysu.edu.cn/panCancer.php) and TargetScan (http://www.targetscan.org/vert_71/), respectively.

A total of 37 tumor tissues and corresponding adjacent normal tissues from ovarian cancer patients were obtained from Yantai Affiliated Hospital of Binzhou Medical University between June 2019 and December 2020. The age range of the 37 patients was 45–70 years with the 60 years mean age. Our study methodologies conformed to the standards set by the Ethics Committee of Yantai Affiliated Hospital of Binzhou Medical University (approval no. 20190601001, Yantai, Shandong, China). All the patients provided their written informed consent. The clinical characteristics of the 37 patients are listed in Table I.

Ovarian cancer cell lines (SKOV3, OVCAR3, CaoV3 and OV90) and normal ovarian epithelial cells (HOSE) collected from the American Type Culture Collection (ATCC) were cultured in RPMI-1640 medium (Sigma; Merck KGaA) with 10% (v/v) fetal calf serum (Invitrogen; Thermo Fisher Scientific, Inc.) under 5% CO₂ at 37°C.

The lncRNA RHPN1-AS1 probe was designed and synthesized by Guangzhou RiboBio Co., Ltd. (China). The cells were cultured in a 24-well plate for 24 h incubation. Then, the cells were immobilized with 4% paraformaldehyde for 10 min, and treated with 0.5% Triton X-100 permeate solution for 5 min. After preheating the pre-hybridized solution at 37°C, the pre-hybridized solution was added to the cells to incubate cells for 30 min. The lncRNA RHPN1-AS1 probe diluted by the hybrid solution at the ratio of 1:50 was added into the culture plate at 200 µl per well for incubation overnight at 37°C. The next day, removing the hybridized mixture, the saline sodium citrate (SSC) was preheated at 45°C, and the cells were washed in the order of 4X SSC, 2X SSC, and 1X SSC for three times each. Next, 20 µl DAPI was added to each well for staining at room temperature for 10 min. After washing with PBS, the slides were fixed with anti-fluorescence quenching sealing tablet, and the images were captured using fluorescence microscope (Olympus IX70; Olympus) under ×400 magnification.

The RHPN1-AS1 silencing plasmids, miR-6884-5p inhibitor and TOP2A silencing plasmids and their corresponding negative control (NC) were constructed and obtained from Guangzhou RiboBio Co., Ltd. Before transfections, the cells were cultured in a 96-well-plate (1×10⁵/ml), a 12-well-plate (1×10⁶/ml) or a 6-well-plate (2.5×10⁶/ml) to a density of 50%. Lipofectamine 2000 (Invitrogen; Thermo Fisher Scientific, Inc.) was applied to conduct cell transfections. As described in the protocol, the transfection concentrations of RHPN1-AS1 silencing plasmids, TOP2A silencing plasmids and their NC were 100 ng for a 96-well plate, 1,000 ng for a 12-well-plate and 2,500 ng for a 6-well-plate. The primer sequence of RHPN1-AS1 silencing plasmids, miR-6884-5p inhibitor and TOP2A silencing plasmids are showed in Table SI. Transfection concentrations of the miR-6884-5p inhibitor were 3 pmol for a 96-well-plate, 30 pmol for a 12-well-plate and 75 pmol for a 6-well-plate. After transfection for 48 h, the transfection efficiency was detected by RT-qPCR. The sequences of all the constructed vectors are shown in Table SI.

Total RNA Extraction Kit (Solarbio, China) was used to extract total RNAs from the tissues and cells. The cDNAs were synthesized by PrimeScript RT reagent kit (Takara Bio, Inc.). RT-qPCR was performed by TB Green^® Premix Ex Taq II kit (Takara Bio, Inc.). β-actin was applied as the reference gene for RHPN1-AS1 and TOP2A, and U6 was the reference gene for miR-6884-5p. 2^−ΔΔCq method (28) was applied for the calculation of the relative expression level. The sequence information of primers is listed in Table II.

Nuclear and cytoplasmic fractionations of SKOV3 and OVCAR3 cells were separated with NE-PER Nuclear and Cytoplasmic Extraction Kit (Thermo Fisher Scientific, Inc.). RHPN1-AS1 expression was detected using RT-qPCR. GAPDH/U6 was the cytoplasmic/nuclear control.

BrdU assay was conducted by the 5-bromo-2′-deoxyuridine (BrdU) kit (YEASEN, China). SKOV3 and OVCAR3 cells were seeded in 96-well plates. After a 48-h transfection, 100 µl BrdU liquid was used to treat the cells in each well for a 2-h incubation. The absorbance at 450 nm was determined in a microplate reader (BioTek).

MTT cell proliferation and cytotoxicity assay kit (Sangon Biotech Co., Ltd.) was applied to determine the cell viability of SKOV3 and OVCAR3 cells. The cells were incubated for 0, 24, 48 and 72 h transfection. Before determination, 10 µl MTT reagent was used to treat the cells in each well. After a 4-h incubation, the absorbance was monitored at 570 nm using a microplate reader (BioTek).

The transfected cells were seeded into a 6-well cell culture plate at 1×10³ cells/well to culture for 2 weeks. After culturing, the medium was removed and washed with PBS for 3 times. Then, the cells were fixed with 1 ml methanol for 20 min and stained with 0.1% crystal violet solution for 30 min. After that, the remaining crystal violet solution was slowly washed with running water. Finally, the cells were photographed with an optical microscope with ×200 magnification, and the number of clones was counted.

CultreCoat^® BME 96 Well Cell Adhesion Assay Kit (Trevigen, Inc.) was used to carry out this detection. A 96-well-plate was coated with Laminin-1 overnight. The plate was washed by washing buffer for twice. Then the plate was blocked with blocking buffer at 37°C for 1 h. An amount of 50 µl cells after a 48-h transfection was added into each well for an incubation for 30 min. The cells were fixed by 4% paraformaldehyde. Crystal violet (0.5%) was added to stain the cells. After washing with washing buffer, the plate was turned upside down and dried up completely, and 2% SDS was added for a 30-min incubation. Finally, the absorbance was determined at 570 nm by a microplate reader (BioTek).

Caspase-3 activity assay kit (Cell Signaling Technology, Inc.) was used for the analysis of caspase-3 activity of SKOV3 and OVCAR3 cells after transfection for 72 h. Based on the protocol, cells were digested by trypsin and collected by centrifuging at 630 × g for 5 min. Supernatant was discarded and cells were resuspended gently and counted. At the proportion of 50 µl lysate per 2×10⁶ cells, the cells were treated by precooled lysis buffer. The cells were treated with a centrifugation at 4°C at 3,780 × g for 10 min. Supernatant was carefully absorbed and transferred to a new EP tube. Ac-DEVD-pNA was added for a 4-h incubation. The absorbance was determined at 405 nm by a microplate reader (BioTek, USA). The sample OD value/blank group OD value ratio was used to determine caspase-3 activation.

The apoptosis rate was detected by Annexin V-FITC/PI apoptosis kit (eBioscience; Thermo Fisher Scientific, Inc.). SKOV3 and OVCAR3 cells after transfection for 48 h were seeded into 6 well plates at 2×10⁵ cells/well. After the cell confluence reached 80%, the cells were collected. Then, 400 µl premixed 1X buffer binding was added into the cells, and the cell suspension was transferred into a flow tube. Next, into the cell suspension in the flow tube was added 5 µl of Annexin V-FITC and 5 µl of propidium iodide (PI) to incubate for 15 min in the dark. After incubation, the cells were analyzed using the BD FACScalibur Flow cytometry (BD Biosciences), which is equipped with Cell Quest software version 5.1 (BD Biosciences).

Wound healing assay was applied to demonstrate the migration ability of the SKOV3 and OVCAR3 cells. The cells after a 48-h transfection were seeded in 6-well plates at the density of 2×10⁵ cells/well. When the cell confluence was more than 90%, 200-µl pipette tips were used to make the wounds in the middle of the 6-well plates plated, and the wounds at 0 h were photographed using an optical microscope. After incubation with serum-free medium for 24 h, the wounds at 24 h were also photographed using an optical microscope. The cell migration rate was calculated by the change of wounds width at 0 and 24 h.

SKOV3 and OVCAR3 cells were transfected with RHPN1-AS1 wild-type (psiCHECK2-RHPN1-AS1-WT, WT-lnc), RHPN1-AS1 mutant-type (psiCHECK2-RHPN1-AS1-MUT, MUT-lnc), TOP2A 3′UTR wild-type (psiCHECK2-TOP2A-WT, WT) or TOP2A 3′UTR mutant-type (psiCHECK2-TOP2A-MUT, MUT) from Guangzhou RiboBio Co., Ltd., which was transfected with miR-6884-5p mimic or mimic negative control (NC). Dual-Luciferase Reporter Assay System (Promega Corp.) was applied to detect the firefly and Renilla luciferase activities. Relative luciferase activity was normalized to the firefly luciferase internal control. Fluorescence/Multi-Detection Microplate Reader (BioTek) was used to detect luciferase activity.

Magna RIP RNA-binding protein immunoprecipitation kit (Millipore, USA) was used to perform RNA immunoprecipitation (RIP) assay. Cells transfected with miR-6884-5p mimic or negative control (NC) were lysed by RIP lysis buffer and incubated in RIP buffer containing magnetic beads supplemented with Anti-Ago2 or negative control (Anti-IgG). After incubation with Proteinase K and the immunoprecipitated RNA, the extracted RNAs isolated by TRIzol reagent (Invitrogen; Thermo Fisher Scientific, Inc.) were detected by RT-qPCR.

Pierce™ RNA 3′ End Desthiobiotinylation Kit (Thermo Fisher Scientific, Inc.) was used to conduct RNA pull-down assay. In brief, first, biotin-labeled miR-6884-5p mimic (Bio-miR-6884-5p) or biotin-labeled miR-6884-5p NC (Bio-NC) was obtained from Guangzhou RiboBio Co., Ltd. After transfection, harvested cells were lysed and treated with an incubation of streptavidin magnetic beads, forming RNA-protein complexes, after which protein K and DNase A were added. RNA was separated from the complex and TOP2A mRNA expression was detected by RT-qPCR.

After transfection for 48 h, total proteins were extracted from cells by RIPA lysis buffer (Sigma; Merck KGaA) and qualified by BCA detection kit (Thermo Fisher Scientific, Inc.). After separating by 10% SDS-PAGE, proteins were transferred to PVDF membranes (Millipore, USA). After blocking by 5% skim milk powder for 1 h, the membranes were treated with the TOP2A antibody (anti-TOP2A, dilution 1:1,000, ab52934, Abcam) and β-actin antibody (rabbit anti-β-actin, dilution 1:1,000, ab8227, Abcam) at 4°C for 12 h. Next day, the secondary antibody (goat anti-rabbit, dilution 1:10,000, ab6721, Abcam) was added into the membranes for 1 h. Protein bands were visualized via Immobilon Western Chemiluminescent HRP Substrate (Millipore), and the densitometry was analyzed using AlphaEase FC 6.0.2 (Alpha Innotech Corp.).

Our statistical analysis was performed by GraphPad Prism 8.0 (GraphPad Software, Inc.). The Student's t-test (two-tailed) was applied to assess the statistical significance between two groups, while the statistical significance among three or more groups was determined by one-way ANOVAs with Dunnett's multiple comparisons test. Correlation analysis in GraphPad Prism 8.0 was used to analyze the correlation among RHPN1-AS1, miR-6884-5p and TOP2A expression level. Data are presented as the mean ± standard deviation (SD). P<0.05 was regarded as statistically significant.

We firstly screened out the significantly upregulated genes from GSE119056 and GSE23392 data series downloaded from GEO database. There were 50 overlapped genes between the two datasets (Fig. 1A). The 50 genes then went through a STRING networking analysis at the highest confidence level 0.9. Among the 10 genes in the network, we noted that TOP2A (Fig. 1B) had been studied in ovarian cancer and was considered to be a tumor promoter; however, it had not been studied in a ceRNA network in ovarian cancer before. Thus, TOP2A was selected as the gene of interest in our study. According to the GEPIA database, TOP2A was found to be significantly upregulated in ovarian cancer as well (Fig. 1C). On the other hand, we have long paid attention to lncRNA RHPN1-AS1, and we noted that it was significantly upregulated in ovarian cancer (from GEPIA, Fig. 1D) and is considered to be a tumor promoter in ovarian cancer by working a ceRNA mechanism (29,30). There were two overlapped miRNAs between the target miRNAs of RHPN1-AS1 and the target miRNAs of TOP2A: miR-485-5p and miR-6884-5p (Fig. 1E). miR-485-5p has been extensively studied in ovarian cancer, whereas miR-6884-5p has been limitedly studied. Thus, we chose miR-6884-5p as our miRNA of interest. The RHPN1-AS1/miR-6884-5p/TOP2A interactome was then studied for the first time in our study.

It was necessary to firstly understand the characteristic of RHPN1-AS1 in ovarian cancer. After eliminating the tissue samples with borderline type (n=12) and unknown type (n=3), we analyzed RHPN1-AS1 expression in 22 ovarian cancer tissues and 22 corresponding normal tissues by RT-qPCR. The expression of RHPN1-AS1 in the tumor tissues was nearly 2.7 times of that in the corresponding normal tissues (Fig. 2A). RT-qPCR was also applied to compare how RHPN1-AS1 was expressed in ovarian cancer cell lines (SKOV3, OVCAR3, CaoV3 and OV90) and a normal ovarian epithelial cell line (HOSE). The amount of RHPN1-AS1 expression in every ovarian cancer cell line was more than twice of that in normal ovarian epithelial cell line (Fig. 2B). In addition, SKOV3 and OVCAR3 cells displayed the highest RHPN1-AS1 expression, triple higher than that in the HOSE cells. Thus, SKOV3 and OVCAR3 cells were chosen for further cellular behavioral experiments. Subsequently, we researched the intracellular distribution of RHPN1-AS1. RHPN1-AS1 was mainly distributed in the cell cytoplasm (Fig. 2C). At the same time, the FISH assay further verified that the distribution of RHPN1-AS was in the cell cytoplasm (Fig. 2D).

Before the ovarian cancer cellular behavioral experiments, the transfection efficiency of the RHPN1-AS1 silencing plasmids (si-lnc) was detected by RT-qPCR. si-lnc was able to downregulate RHPN1-AS1 expression compared to the blank control group (Fig. S1A). After transfection with si-lnc for 72 h, the cell viability was significantly decreased especially in SKOV3 and OVCAR3 cells (Fig. S1B). Then, the cellular behavioral experiments were performed in SKOV3 and OVCAR3 cells. BrdU assay revealed that knockdown of RHPN1-AS1 obviously repressed cell proliferation by approximately 40% compared to the blank control group in the OVCAR3 and SKOV3 cells (Fig. 3A). MTT assay showed silencing of RHPN1-AS1 significantly impaired the cell viability of the OVCAR3 and SKOV3 cells (Fig. 3B). Colony formation assay revealed that knockdown of RHPN1-AS1 reduced the number of colonies by 35% in the SKOV3 cells and 60% in OVCAR3 cells (Fig. 3C). Cell adhesion assay demonstrated si-lnc suppressed cell adhesion by approximately 30% compared to the blank control group (Fig. 3D). Caspase-3 activity assay demonstrated that the silencing of RHPN1-AS1 increased cell apoptosis to over twice the level as the blank control group (Fig. 3E). Moreover, flow cytometry was similar to the results of the caspase-3 activity assay, which showed that the apoptosis rate of the si-lnc group was increased by more than 2 times compared with the blank group (Fig. 3F). Additionally, wound healing assay demonstrated that the silencing of RHPN1-AS1 induced inhibition of cell migration (Fig. 3G). These data suggest that silencing of RHPN1-AS1 functionally impedes ovarian cancer progression in vitro.

ENCORI starbase algorithm (http://starbase.sysu.edu.cn/) was employed for the exploration of potential targets for RHPN1-AS1 in ovarian cancer. As a result, miR-6884-5p is a target of RHPN1-AS1, and the potential binding site between RHPN1-AS1 and miR-6884-5p is shown in Fig. 4A. Compared to the mimic NC group, the luciferase activity was significantly reduced by about 50% only in cells that were co-transfected with the miR-6884-5p mimic and RHPN1-AS1-WT plasmids (Fig. 4B). RIP assay confirmed that RHPN1-AS1 binding to miR-6884-5p was dramatically pulled down by anti-Ago2 antibody (Fig. 4C). After eliminating the tissue samples with borderline type (n=12) and unknown type (n=3), miR-6884-5p expression in ovarian cancer tissues was 50% lower than that observed in the corresponding normal tissues (Fig. 4D). The correlation analysis displayed an inverse correlation between miR-6884-5p and RHPN1-AS1 expression (Fig. 4E). Moreover, in Fig. 4F, miR-6884-5p exhibited a lower expression in ovarian cancer SKOV3 and OVCAR3 cells than that noted in the normal ovarian HOSE cells, showing that miR-6884-5p expression in ovarian cancer cells were only approximately 55% of that in HOSE cells. As expected, silencing of RHPN1-AS1 positively regulated miR-6884-5p expression, increasing it by more than 3-fold compared to the blank control group (Fig. 4G). These findings strongly confirmed RHPN1-AS1 sponges miR-6884-5p in ovarian cancer.

Further, rescue experiments were performed to ascertain whether RHPN1-AS1 facilitates ovarian cancer progression via miR-6884-5p. As a result, compared to the blank control group, si-RHPN1-AS1 increased miR-6884-5p expression by twice while decreasing RHPN1-AS1 expression by 75%, and the miR-6884-5p inhibitor decreased miR-6884-5p expression by 70% while it had no effect on RHPN1-AS1 expression (Fig. 5A). In addition, miR-6884-5p inhibitor + si-RHPN1-AS1 did not change miR-6884-5p expression while this co-transfection suppressed RHPN1-AS1 expression by 70% compared to the blank control group (Fig. 5A). In BrdU assay, compared to the blank control group, silencing of RHPN1-AS1 repressed cell proliferation by 30 to 40% while the miR-6884-5p inhibitor promoted cell proliferation by 40%, which was restored in the co-transfection group (Fig. 5B). MTT assay revealed that the miR-6884-5p inhibitor exerted an oncogenic effect on cell viability, and the oncogenic effect induced by the miR-6884-5p inhibitor was abrogated by the silencing of RHPN1-AS1 (Fig. 5C). Colony formation assay showed that the downregulation of miR-6884-5p increased the number of cell colonies compared to the blank group, but this effect was reversed by silencing of RHPN1-AS1 (Fig. 5D). Cell adhesion assay demonstrated that the silencing of RHPN1-AS1 decreased cell adhesion by 30%, while the miR-6884-5p inhibitor promoted it by about 40% compared to the blank control group, and these effects were restored by the miR-6884-5p inhibitor + si-RHPN1-AS1 (Fig. 6A). The result of caspase-3 activity assay demonstrated that the silencing of RHPN1-AS1 increased cell apoptosis to a level twice that of the blank control, while the miR-6884-5p inhibitor suppressed it by about 40% compared to the blank control group, and these effects were restored by the miR-6884-5p inhibitor + si-RHPN1-AS1 (Fig. 6B). The result of the flow cytometric assay showed that downregulation of miR-6884-5p inhibited cell apoptosis, while miR-6884-5p inhibitor eliminated the pro-apoptotic effect of the silencing of RHPN1-AS1 (Fig. 6C). In addition, wound healing assay confirmed that the silencing of RHPN1-AS1 attenuated the promotion of migration in response to miR-6884-5p inhibition (Fig. 6D). Thereby, our results suggest that RHPN1-AS1 promotes ovarian cancer progression via a miR-6884-5p-dependent mechanism.

TargetScan was applied to predict targets for miR-6884-5p. TOP2A might be the most suitable target. The sequence of TOP2A mRNA that contains miR-6884-5p's binding site is shown in Fig. 7A. Dual-luciferase reporter assay demonstrated that the luciferase activity of SKOV3 and OVCAR3 cells transfected with TOP2A-WT was obviously decreased by the introduction of the miR-6884-5p mimic by about 50% compared to the mimic NC (Fig. 7B). RNA pull-down assay revealed that TOP2A was significantly pulled down by bio-miR-6884-5p (Fig. 7C). After eliminating the tissue samples with borderline type (n=12) and unknown type (n=3), TOP2A expression in invasive ovarian cancer tissues was more than twice that noted in the corresponding normal tissues (Fig. 7D). Moreover, TOP2A expression exhibited a negative correlation with miR-6884-5p expression (Fig. 7E). Following analysis of RT-qPCR, TOP2A was obviously highly expressed in ovarian cancer SKOV3 and OVCAR3 cells than that noted in the HOSE cells (Fig. 7F). The miR-6884-5p inhibitor was identified to promote TOP2A mRNA expression by 80% compared to the blank control group (Fig. 7G). The western blot assay further confirmed that the miR-6884-5p inhibitor enhanced the TOP2A protein expression by 1.3-fold, as well as the silencing of RHPN1-AS1 led to a decrease in TOP2A protein expression (Fig. 7H). Therefore, miR-6884-5p targets TOP2A in ovarian cancer, which is regulated by RHPN1-AS1.

To explore the effect of the miR-6884-5p/TOP2A axis in ovarian cancer, TOP2A siRNA was used to transfect ovarian cancer cells. As expected, silencing of TOP2A dramatically inhibited TOP2A protein expression compared to the blank control group (Figs. 8A and S2); at the same time, miR-6884-5p inhibitor restored the effect of the silencing of RHPN1-AS1 (Fig. 8A). As determined by the BrdU assay, silencing of TOP2A in SKOV3 and OVCAR3 cells caused about 50% inhibition in cell viability compared to the blank control group, which was restored by miR-6884-5p inhibitor + TOP2A siRNA (Fig. 8B). According to the MTT assay, silencing of TOP2A led to inhibition of cell viability, which was restored by miR-6884-5p inhibitor + TOP2A siRNA (Fig. 8C). Colony formation assay showed that interference with TOP2A reduced the colony number to approximately 50% of the blank group, while miR-6884-5p inhibitor reversed the influence of TOP2A siRNA on the number of cell colonies (Fig. 8D). Cell adhesion assay demonstrated that silencing of TOP2A decreased cell adhesion by 30% compared to the blank control group, and the decrease was restored by the miR-6884-5p inhibitor + TOP2A siRNA (Fig. 9A). Caspase-3 activity assay revealed that the miR-6884-5p inhibitor abolished the increase (more than 1.5 times compared to the blank control group) of cell apoptosis induced by silencing if TOP2A in the SKOV3 and OVCAR3 cells (Fig. 9B). The result of the flow cytometry assay showed that knockdown of TOP2A promoted cell apoptosis, while the miR-6884-5p inhibitor eliminated the apoptosis-promoting effect caused by the silencing of TOP2A (Fig. 9C). As for wound healing assay, silencing of TOP2A obviously suppressed the ability of cell migration, while the miR-6884-5p inhibitor abrogated the suppressive effect on cell migration modulated by the silencing of TOP2A (Fig. 9D). Herein, miR-6884-5p inhibition functionally decreased cell apoptosis and increased cell proliferation, viability and migration in ovarian tumor by releasing TOP2A. These lines of evidence confirm our hypothesis that RHPN1-AS1 promotes ovarian carcinogenesis through sponging miR-6884-5p and releasing TOP2A.