LncRNA MIR4435-2HG triggers ovarian cancer progression by regulating miR-128-3p/CKD14 axis

Background Accumulating studies showed that long noncoding RNAs (lncRNAs) played vital roles in cancer progression. LncRNA MIR4435-2HG was proved to act as an oncogene in various tumors. However, the underlying function of MIR4435-2HG in ovarian cancer (OC) remains unclear. Methods The expression levels of MIR4435-2HG, miR-128-3p and cyclin-dependent kinase 14 (CDK14) were analyzed by quantitative real-time polymerase chain reaction (qRT-PCR). Cell proliferation and apoptosis in OC cells were detected by 3-(4, 5-dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide (MTT) assay and flow cytometric analysis, respectively. Transwell assay was applied to evaluate cell migration and invasion. Wound healing assay was performed to monitor the migration rate. Western blot assay was performed to detect the protein levels of Bcl-2, Cleaved PARP, E-cadherin, Vimentin and CDK14 in OC cells. The binding sites between miR-128-3p and MIR4435-2HG or CDK14 were predicted by online tool starBase and their relationship was confirmed by dual-luciferase reporter assay, RIP assay and pull-down experiment. Results MIR4435-2HG and CDK14 were over-expressed in OC tissues and cells. Patients with high MIR4435-2HG expression had poorer overall survival (OS) than patients with low MIR4435-2HG expression. MIR4435-2HG knockdown inhibited proliferation, invasion and migration but induced apoptosis of OC cells via miR-128-3p/CDK14 axis. In conclusion, MIR4435-2HG knockdown suppressed the progression of OC cells through downregulating CDK14 expression by the promotion of miR-128-3p.

clinical application to explore the mechanisms of OC progression.
Long non-coding RNAs (LncRNAs) are endogenous non-coding RNAs with the length of over 200 nucleotides, and ncRNA targeting protein-coding genes was reported to be an important regulatory molecule that can regulate the gene expression [5,6]. A recent report suggested that lncRNA has a significant effect on the progression of numerous tumors and exerts its function as tumor suppressor or oncogene during tumorigenesis [7].
MIR4435-2HG is considered as a carcinogenic lncRNA in various cancers. For instance, up-regulated MIR4435-2HG interacted with poorer progression-free survival (PFS) and overall survival in colorectal cancer patients, and MIR4435-2HG was involved in the occurrence and evolution of colorectal cancer through regulating P38/ MAPK and VEGF pathways [8]. MIR4435-2HG was overexpressed in hepatocellular carcinoma and promoted tumor cell proliferation by activating miR-487a [9]. A recent study presented that MIR4435-2HG was elevated in OC and the high expression of MIR4435-2HG triggered the migration and invasion of OC cells via activating TGF-β1 [10]. Moreover, miR-128-3p could participate in the progression of diver cancers through targeting different genes, and miR-128-3p regulated cell growth in breast cancer by targeting LIMK1 and overexpression of miR-128-3p in breast cancer patients presaged a good prognosis [11]. In glioma, the level of miR-128-3p expression was dramatically attenuated in glioma clinical tissues, and miR-128-3p regulated the progression of glioma via targeting NPTX1 and activating the IRS-1/PI3K/ AKT signaling pathway [12]. Cyclin-dependent kinase 14 (CDK14) is a member of cyclin-dependent kinases, and a previous study demonstrated that CDK14 was augmented in the tissues and cells of OC [13].
In our study, we aimed to explore the role of MIR4435-2HG and its underlying mechanism in OC cells, and the results showed a novel mechanism of MIR4435-2HG/ miR-128-3p/CKD14 axis and provided new potential therapeutic targets for OC.

Materials and methods
Patients' samples 42 pairs of OC tissues and matched adjacent normal tissues were obtained from The First People's Hospital of Shangqiu. The project was supported by the Ethics Committee of the First People's Hospital of Shangqiu and informed consents were signed by all participants. All patients received not any preoperative treatment, the characteristics of clinical samples are enumerated in Table 1. Tumor stage was defined based on the Federation of Gynecology and Obstetrics (FIGO) guidelines [14]. Tumor-node-metastasis (TNM) stage was defined by the American Joint Committee on Cancer [15]. The isolated tissues were quickly placed in liquid nitrogen and preserved at − 80 °C ultra low-temperature refrigerator for subsequent experiments.

Flow cytometry assay
A PI/Annexin V Kit (Sigma, Saint Louis, MO, USA) was used to examine cell apoptosis rate. SKOV3 and OVCAR3 cells in exponential phase were collected and washed twice with PBS, then 5 × 10 5 cells were resuspended in 100 μL binding buffer, and incubated with propidium iodide (PI) and Annexin V for 10 min in the dark at 37 °C, the apoptotic rate was observed by FAC-Scan flow cytometer (Beckman Coulter, Inc. San Jose, CA, USA).

Transwell assay
Migration assay was completed in 24-well transwell chambers (Millipore, Bedford, MA, USA). Transfected cells were cultured in the upper chambers without matrigel (Millipore) and 600 μL medium containing 10% FBS was put into the lower chambers. After 24 h, cells transferred to the surface of the lower chamber were fixed with 4% paraformaldehyde and stained with 0.1% crystal violet. Finally, the migratory cells were observed by a microscope (Olympus, Tokyo, Japan).
Transwell invasion experiments are carried out by the same method with the upper chambers pre-coated with matrigel.

Wound healing assay
SKOV3 and OVCAR3 cells were seeded in 6-well plates (5 × 10 5 cells/well) and exposed to mitomycin C (Sigma) for 2 h. The cell monolayer was wounded with a pipette tip to create a scratch. Then, the cells were washed with PBS and the serum-free medium was added. Images were captured every 6 h following the initial scratch to evaluate cell migration rate.

In vivo experiments
BALB/c nude mice (n = 6, 4-week-old, female) were purchased from HFK bioscience Co., LTD (Beijing, China). SKOV3 cells were stably transfected with sh-MIR4435-2HG or sh-NC. Then, stably transfected SKOV3 cells were subcutaneously inoculated into the right flank of mice back. Five days after inoculation, the tumor volume was calculated every five days with a vernier caliper according to the formula: length × width 2 × 0.5. After 30 days, all tumor tissues were excised and used for weighting and expression analysis. All animal procedures were approved by the Animal Care and Use Committee of The First People's Hospital of Shangqiu.

RNA immunoprecipitation (RIP) assay
RIP assay was performed using the Magna RIP RNA-Binding Protein Immunoprecipitation Kit (Millipore) in agreement with the manufacturer's instruction. SKOV3 and OVCAR3 cell lysates were obtained and incubated with RIP buffer containing magnetic beads conjugated with human anti-Argonaute2 (anti-Ago2) antibody (Millipore) or normal mouse IgG (control; Millipore). RNA was extracted from immunoprecipitate and analyzed by qRT-PCR.

Pull-down assay
Biotinylated miR-128-3p or NC mimics were transfected into SKOV3 and OVCAR3 cells, respectively. The active cells were collected after transfection and treated with lysis buffer (Ambion, Austin, Texas, USA). Following the protocol of manufacturer, the cell lysate was incubated with Dynabeads M-280 Streptavidin (Invitrogen). Beads were rinsed and incubated with biotinylated miR-128-3p at 4 °C overnight. Then, the RNase-free lysis buffer was added into the beads with the immobilized miR-128-3p fragment and incubated for 1 h at 37 °C, then interacted RNAs were purified and evaluated by qRT-PCR.
Statistical analysis SPSS 22.0 (IBM Corp, Armonk, NY, USA) was used to complete the statistical analyses. The experimental data were displayed as mean ± standard deviation. The t test and one-way analysis of variance were adopted to analyze experimental data. The survival rate was evaluated with Kaplan-Meier. The difference was statistical significance when P < 0.05.

MIR4435-2HG was highly expressed in OC tissues and cell lines
The expression of MIR4435-2HG was examined in OC tissues and adjacent normal tissues. The results showed that the expression of MIR4435-2HG in OC tissues (n = 42) was significantly increased by 1.97 folds on average compared with adjacent normal tissues (n = 42) (P < 0.05) (Fig. 1a). Besides, CISH assay revealed that there was strong staining in tumor tissues but not in adjacent normal tissues, suggesting the high abundance of MIR4435-2HG in OC tissues (Fig. 1a). Compared with normal ovarian cell line ISOE80, the expression of MIR4435-2HG in OC cell lines (SKOV3, Caov-3, A2780, and OVCAR3) was notably increased, while SKOV3 and OVCAR3 cell lines represented with the higher MIR4435-2HG expression (P < 0.05, Fig. 1b). Subsequently, the prognostic values of MIR4435-2HG expression were analyzed by Kaplan-Meier, and it was shown that the survival time of patients with low MIR4435-2HG expression was significantly higher than those with high MIR4435-2HG expression (P < 0.05, Fig. 1c). Meanwhile, to explore the clinical significance of MIR4435-2HG in OC, the relationship between its expression pattern and clinicopathological characteristics was analyzed, and the data implied that the expression level of MIR4435-2HG was closely correlated with tumor size, FIGO stage and the lymph distant metastasis (P < 0.05, Table 1). These results demonstrated that high MIR4435-2HG expression was associated with poor prognosis.

Knockdown of MIR4435-2HG inhibited malignant behaviors of OC cells
To examine the biological functions of MIR4435-2HG in OC cells, the expression of MIR4435-2HG was prevented by si-MIR4435-2HG in SKOV3 and OVCAR3 cells. The lowest MIR4435-2HG expression was caused by si-MIR4435-2HG #1 (0.42 folds on average), therefore si-MIR4435-2HG #1 was chosen for subsequent experimentations (P < 0.05, Fig. 2a). By performing MTT assay, MIR4435-2HG knockdown was shown to significantly retard the proliferative capacity of SKOV3 and OVCAR3 cells compared with the NC group (P < 0.05, Fig. 2b, c). The flow cytometry results showed that SKOV3 and OVCAR3 cells transfected with si-MIR4435-2HG induced apoptosis augment (P < 0.05, Fig. 2d). In transwell assay, the migration and invasion of SKOV3 and OVCAR3 cells were inhibited in the si-MIR4435-2HG group compared with the si-NC group (P < 0.05, Fig. 2e and f ). Besides, the wound healing assay presented that MIR4435-2HG knockdown suppressed the migration rate of SKOV3 and OVCAR3 cells compared with NC group (Fig. 2g  and h). Moreover, the protein expression of Cleaved PARP and E-cadherin was activated by MIR4435-2HG knockdown, while Bcl-2 and Vimentin were restricted (P < 0.05, Fig. 2i). All the data indicated that depletion of MIR4435-2HG promoted apoptosis pathway but inhibited Epithelial-to-mesenchymal transition (EMT) progression, and MIR4435-2HG knockdown can act as a tumor inhibitor in the development of OC.

MIR4435-2HG knockdown inhibited tumor growth in vivo
To ascertain the role of MIR4435-2HG in vivo, the sh-MIR4435-2HG transfected SKOV3 cells were subcutaneously injected into the mice. Compared with sh-NC injection, the tumor volume and tumor weight were strikingly reduced in the sh-MIR4435-2HG groups (Fig. 3a-c). Then, the expression of MIR4435-2HG in excised tumor tissues was checked, and the result showed that the level of MIR4435-2HG was significantly decreased by 0.428 folds on average in the sh-MIR4435-2HG group relative to the sh-NC group (Fig. 3d). These data demonstrated that MIR4435-2HG knockdown suppressed tumor growth in vivo.

MiR-128-3p overexpression abrogated the role of CDK14 overexpression in OC cells
To investigate whether miR-128-3p interacted with CDK14 to inhibit the role of CDK14, SKOV3 and OVCAR3 cells were introduced with CDK14 or CDK14 + miR-128-3p mimics, Vector or CDK14 + NC mimics as the control. We noticed that the expression of CDK14 at both mRNA and protein levels elevated by CDK14 transfection was decreased in the CDK14 + miR-128-3p mimics group (P < 0.05, Fig. 8a-c). CDK14 overexpression could significantly enhance the proliferative capacity of SKOV3 and OVCAR3 cells, while the enhanced proliferation could be abolished by miR-128-3p mimics (P < 0.05, Fig. 8d and e). On the contrary, cell apoptosis was decreased by CDK14 overexpression but partially reversed by miR-128-3p mimics (P < 0.05, Fig. 8f ). Next, we evaluated migration and invasion of SKOV3 and OVCAR3 cells using transwell assays. CDK14 transfection could significantly elevate the migration and invasion abilities of SKOV3 and OVCAR3 cells, while these enhanced abilities could be partially repressed by miR-128-3p mimics (P < 0.05, Fig. 9a and b). The wound healing assay revealed that CDK14 overexpression promoted the migration ratio, while miR-128-3p mimics partially restrained the role of CDK14 (P < 0.05, Fig. 9c and d). Additionally, the data from western blot assay depicted that CDK14 upregulation significantly increased the levels of Bcl-2 and Vimentin, whereas miR-128-3p overexpression reversed the positive effects of CDK14 on these proteins. Similarly, CDK14 generated a significant reduction in the levels of Cleaved PRAP and E-cadherin both in SKOV3 and OVCAR3 cells, which were augmented after the transfection of miR-128-3p mimics (P < 0.05, Fig. 9e and f ). These analyses manifested that miR-128-3p bound to CDK14 and blocked the role of CDK14.

Knockdown of MIR4435-2HG downregulated the expression of CDK4 via regulating the expression of miR-128-3p
Further analysis revealed that CDK14 expression was positively correlated with the MIR4435-2HG The expression level of CDK14 in ISOE80, SKOV3, and OVCAR3 cells was examined by qRT-PCR and western blot, respectively. g The specific binding regions between CDK14 and miR-128-3p were obtained from the online software starbase. h-k The relationship between CDK14 and miR-128-3p was confirmed by dual-luciferase reporter assay, RIP assay and RNA pull-down assay. l The expression of miR-128-3p in SKOV3 and OVCAR3 cells transfected with miR-128-3p inhibitor was analyzed by qRT-PCR. m-p The expression of CDK14 at mRNA and protein levels in SKOV3 and OVCAR3 cells transfected with miR-128-3p inhibitor or mimics was measured by qRT-PCR and western blot. *P < 0.05 expression in OC tissues (R 2 = 0.606, P < 0.0001, Fig. 10a). The qRT-PCR analysis showed that CKD14 was reduced after the transfection of si-MIR4435-2HG#1 in SKOV3 and OVCAR3 cells, which was accelerated by miR-128-3p inhibitor (P < 0.05, Fig. 10b). We found that knockdown of MIR4435-2HG decreased the protein level of CDK14, but the effect was significantly reversed after the inhibition of miR-128-3p (P < 0.05, Fig. 10c and d). These data suggested that MIR4435-2HG knockdown depleted the expression of CDK14 by competitively binding to miR-128-3p in SKOV3 and OVCAR3 cells.

Discussion
Accumulated research suggested that MIR4435-2HG played a critical role in multiple tumors, which can function as an oncogene to induce cancer cell reproduction and metastasis [9,16]. Qian et al. found that over-expression of MIR4435-2HG contributed to lung cancer progression and was related to histological grade and lymph node metastasis [17]. Previous data showed that MIR4435-2HG could be regarded as a prognostic marker of OC [10]. However, the relationship between MIR4435-2HG and OC progression remains elusive. Thereby, we aimed to explore the specific role of MIR4435-2HG in OC.
In the study, qRT-PCR results suggested that MIR4435-2HG was greatly increased in both OC tissues and cell lines, and MIR4435-2HG overexpression was closely related to large tumor size, FIGO stage (III + IV), distant lymph metastasis, and poor survival in OC tissues. We also found that knockdown of MIR4435-2HG repressed proliferation, invasion, metastasis, and triggered apoptosis in SKOV3 and OVCAR3 cells. These results indicated that MIR4435-2HG might act as a carcinogen in OC, which was consistent with the previous study [10].
LncRNAs affected the occurrence and development of tumors via post-transcriptional regulation [18]. Multiple lncRNAs may regulate gene expression by isolating miRNAs, thus reducing the number of available miR-NAs in cells. Therefore, lncRNA acts as a depressor of miRNA function and further serves as an activator of gene expression, which served as a competing endogenous RNA (ceRNA) to modulate the expression of the target gene via sponging miRNAs [19,20]. In the current literatures, MIR4435-2HG was reported to exert ceRNA function in osteoarthritis [21], glioma [22], hepatocellular carcinoma [9] and lung cancer [17]. Similarly, we assumed that MIR4435-2HG functioned as an oncogene in OC via a ceRNA mode. To test this hypothesis, bioinformatic analysis of MIR4435-2HG-miRNA prediction was performed through online software, and the target binding was verified by a dual-luciferase reporter and pull-down assay. All the data indicated that miR-128-3p was a direct target of MIR4435-2HG. It was well known that miR-128-3p was an inhibitor in various tumors [23,24]. Our data showed that the level of miR-128-3p was dramatically reduced in OC tissues and cell lines, which was inversely correlated with MIR4435-2HG expression Fig. 9 MiR-128-3p hindered the effects of CDK14 on cell migration and invasion in OC cells. a, b Cell migration and invasion in SKOV3 and OVCAR3 cells were determined by transwell assay. c, d The migration ratio was monitored according to the wound healing assay. e, f The expression of E-cadherin, Vimentin, Cleaved PARP and Bcl-2 in SKOV3 and OVCAR3 cells was measured by western blot. *P < 0.05 in OC tissues. These results indicated that miR-128-3p was a tumor inhibitor in OC.
MiRNA acted as a tumor suppressor through the restraint of its target genes to participate in OC progression. We inferred CDK14 as a target of miR-128-3p containing the putative miRNA response sequences within its 3′-UTR by starBase v2.0, and the result indicated that CDK14 was a promising candidate gene of miR-128-3p. CDK14 was reported to be involved in the progression of various cancers, including OC. Jin et al. showed that high expression of SNHG15 facilitated NSCLC development through the acceleration of CDK14 and sponging and OVCAR3 cells transfected with si-MIR4435-2HG#1, si-NC, si-MIR4435-2HG#1 + miR-128-3p inhibitor or si-MIR4435-2HG#1 + NC inhibitor were examined using qRT-PCR and western blot. *P < 0.05 Fig. 11 The MIR4435-2HG/miR-128-3p/CDK14 axis in OC progression miR-486 [25]. Li et al. pointed that overexpression of CDK14 reversed the depressant effects of miR-542-3p on OC cells [26], while CDK14 knockdown restrained cell proliferation, invasion and migration in OC [27]. Here, we found a remarkable negative correlation between the expression levels of CDK14 and miR-128-3p in OC tissues. CDK14 was a direct target of miR-128-3p in SKOV3 and OVCAR3 cells, which indicated that the ceRNA system existed among MIR4435-2HG, miR-128-3p and CDK14 in OC. Next, we wondered that whether MIR4435-2HG/miR-128-3p/CDK14 axis conduced to the progression of OC in vitro. The data presented that the effect of MIR4435-2HG knockdown reduced CDK14 expression level by the promotion of miR-128-3p.

Conclusions
In summary, lncRNA MIR4435-2HG was regarded as a vital mediator of cell growth in OC. MIR4435-2HG was obviously upregulated in OC tissues and cell lines, and MIR4435-2HG knockdown repressed OC progression by acting as a ceRNA to downregulate CDK14 through competitively binding to miR-128-3p (Fig. 11). These results indicated that MIR4435-2HG/miR-128-3p/CDK14 axis may be a potential therapeutic basis for the treatment of OC.