Long non-coding RNA PVT1-mediated miR-543/SERPINI1 axis plays a key role in the regulatory mechanism of ovarian cancer

Abstract Purpose: To investigate the regulation mechanism of long non-coding RNA (lncRNA) plasmocytoma variant translocation 1 (PVT1) in ovarian cancer (OC). Methods: The levels of PVT1, microRNA (miR)-543, serpin peptidase inhibitor-clade I (neuroserpin)-member 1 (SERPINI1) in OC tissues and OVCAR-3, A2780, TOV-112D of OC cell lines were detected by quantitative real-time PCR (qRT-PCR) and Western Blot (WB). Cell proliferation, migration, invasion, apoptosis and the regulatory relationship between genes and target genes were analyzed by 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), Transwell, flow cytometry and dual luciferase reporter (DLR). The OC patients were followed up for 5 years to analyze the relationship between PVT1 and 5-year overall survival (OS). Results: In contrast with miR-543, PVT1 and SERPINI1 were highly expressed in OC tissues and cell lines, and high levels of PVT1 were significantly associated with lower 5-year OS of patients. Down-regulating PVT1 not only inhibited the malignant proliferation, migration and invasion of OC cells, but promoted cell apoptosis. PVT1 regulated miR-543 in a targeted manner, and its overexpression could attenuate the anticancer effect of miR-543 on OC cells. In addition, miR-543 also directly targeted SERPINI1, and miR-543 knockdown weakened the inhibitory effect of down-regulated SERPINI1 on OC progression. Furthermore, we found that PVT1 acted as a competitive endogenous RNA to sponge miR-543, thereby regulating the expression of SERPINI1. Conclusion: PVT1 can mediate the molecular mechanism of OC by miR-543/SERPINI1 axis regulatory network, which is a new therapeutic direction for OC.


Introduction
Ovarian cancer (OC) is the eighth leading cause of cancer-related deaths worldwide, and its risk increases with the number of ovulation cycles in women [1,2]. According to epidemiological statistics of OC, in 2018, this disease plagued approximately 22240 new cases and accounted for 14070 deaths in the U.S.A., and the 5-year relative survival rate of advanced OC was only 29% [3,4]. Currently, the mainstream treatment approaches for OC are constantly optimized, which however, will inevitably trigger recurrence and chemotherapy resistance that cannot be ignored, resulting in unsatisfactory improvement of OC patients' condition [5,6]. Therefore, in-depth understanding of the molecular mechanism of OC and further elucidating its pathological process is essential for the treatment and prevention of this disease.
As molecular regulators that have attracted increasing attention in recent years, long non-coding RNAs (lncRNAs) can control basic biological processes and tumor progression, and research on them is conducive to better understanding the gene regulatory network of cancer [7,8]. The lncRNA family not only plays a role in the development of gynecological tumors such as breast cancer, cervical cancer and endometrial cancer, but participates in the pathological mechanism of OC [9][10][11][12]. For example, CACS15 can promote the malignant development of OC by recruiting EZH2 to suppress the expression of APC, and DANCR can mediate tumor growth and angiogenesis of OC by targeting regulation of microRNA (miR)-145 [13,14]. Plasmocytoma variant translocation 1 (PVT1), which is the focus of this study, is often used as a carcinogen to induce the aggravation of the disease in cancer, and can regulate the tumor microenvironment through the negative regulation of miRNA. Currently, it has been found to be tumorigenic in breast cancer, OC and other gynecological cancers [15]. Through Starbase, we predicted the potential target sites between PVT1 and miR-543, of which the latter serves as a tumor inhibitor in cervical cancer while playing a malignant role in promoting tumor growth and metastasis in prostate cancer [16,17]. In addition, miR-543 shows a low level in OC tissues and cells, which can play an anticancer role in OC by targeting TWIST1, and can also be regulated by placental growth factors and silenced to promote MMP7-mediated tumor invasion in OC [18,19]. Besides, we found via TargetScan that miR-543 had potential binding sites with serpin peptidase inhibitor-clade I (neuroserpin)-member 1 (SERPINI1). As a double-sided secreted protein, SERPINI1 has been found to be abnormally highly expressed in hepatocellular carcinoma and colorectal cancer, while abnormally down-regulated in gastric cancer, which may be carcinogenic or anticancer in different cancers [20,21]. In addition, SERPINI1 also mediates the mechanism of brain metastasis, in which pathological changes promote the survival of tumor cells and vascular combination, and is also closely associated with the brain recurrence of primary tumors, suggesting that SERPINI1 may be a threat to human health in most cases [22].
We hypothesized that the PVT1/miR-543/SERPINI1 axis regulatory network played a key part in the pathological development of OC, and hereby we are here to explore it.

Tissue sample collection
Forty-two patients with OC admitted to China-Japan Union Hospital of Jilin University from January 2013 to January 2015 were selected, with an average age of (49.66 + − 5.41) years. After obtaining the consent of the patients, cancer tissues and adjacent tissues were collected as the research samples during tumor resection, and then stored in liquid nitrogen tanks for future use. Inclusion criteria: patients diagnosed with OC [23] by pathological diagnosis, who received treatment for the first time, without any medications such as hormones within the past 6 months, and with normal cognition and unobstructed communication. Exclusion criteria: patients with other ovarian diseases, those with other malignant tumors, severe complicated diseases of heart, lung, kidney and other organs or severe infectious diseases, or those who refused to provide laboratory specimens. The present study was carried out under the guidance of the Helsinki Declaration of the World Medical Association and the approval of the Ethics Committee of China-Japan Union Hospital of Jilin University. All the enrolled patients were informed and who agreed to participate in the present study had given written informed consent.

Follow-up
The patients were followed up once every 3 months by telephone, visit, medical records or other means. The overall survival (OS) started from the day of diagnosis and ended with the death of the patient or the last follow-up over the 5-year follow-up period.

Quantitative real-time PCR
First, the total RNA was isolated from tissues and cells with TRIzol reagent (EK-Bioscience Co., Ltd., Shanghai, China, 5007050). Then 5 μg of the total RNA were processed for reverse transcription of cDNA as instructed by the kit instructions (Qiming Biotechnology Co., Ltd., Shanghai, China, OX02700). After transcription, 1 μl of synthetic cDNA was amplified. GADPH was used as the internal reference for PVT1 and SERPINI1, while U6 for miR-543, and the data were analyzed by 2 − C t .

Western Blot detection
The total protein was extracted from OC tissues and cultured OC cells by RIPA lysis, and the protein concentration was determined by BCA kit (Rongbai Biotechnology Co., Ltd., Shanghai, China, LCB004). After adjusting the concentration to 4 μg /μL, the protein was separated by 12% SDS-PAGE electrophoresis and transferred to a membrane for sealing in 5% blocking solution. The dilution ratio of SERPINI1, β-Actin and other primary antibodies was 1:1000. The antibodies were obtained from Shanghai Xiyuan Biotechnology Co., Ltd., among which all the primary antibodies were detected with rabbit polyclonal antibodies. Each membrane was rinsed three times with PBS (Zhenyu Biotechnology Co., Ltd., Shanghai, China, 130-070-525) for 15 min each, and then incubated with horseradish peroxidase-labeled goat anti-rabbit secondary antibody. After blotting the excess liquid on the membrane, it was developed by ECL, and the gray value was analyzed.

Transwell for cell migration and invasion
Cells were collected 24 h after transfection, adjusted to 3 × 10 4 cells/well, and inoculated on 24-well plates. Then trypsin digestion was performed before the cells were transferred to the upper chamber, where 200 μl RPMI1640 (Shuangru Biotechnology Co., Ltd., Shanghai, China, LS115-001) was added, while the lower chamber was added with 600 ml RPMI1640 (containing 10% FBS). After 48 h of culture at 37 • C, the upper chamber was swabbed and washed three times with PBS, fixed with paraformaldehyde for 10 min, followed by three times of rinsing with double steam water. After being dried, the cells were stained with 0.5% Crystal Violet, and the number of cell migration was observed and recorded under a microscope. For the determination of cell invasion, the bottom membrane was diluted with 50 mg/l Matrigel (1:8; Hengfei Biotechnology Co., Ltd., Shanghai, China, M8370) on the surface of the upper chamber, and the rest of the steps were the same as above.

Flow cytometry for apoptosis
After digestion with 0.25% trypsin, the transfected cells were washed twice with PBS, and mixed with 100 μl binding buffer to prepare a 1 × 10 6 cells/ml suspension. AnnexinV-FITC and propidium iodide (PI) were then added in turn with 10 μl each, incubated at room temperature in darkness for 5 min, and tested by NovoCyte flow cytometer (FC) (Ranger Apparatus Co., Ltd., Shanghai, China). The experiment was repeated three times to average the value.

RNA immunoprecipitation
RNA immunoprecipitation (RIP) was utilized to verify the interaction between PVT1 and miR-543, and the experiment was conducted strictly following the instructions of the EZMagna RIP kit (Goy Biotechnology Co., Ltd., Shanghai, China, GOY-E5944). After lysis, the cells were incubated with protein A magnetic beads and conjugated with antibodies at 4 • C. Six hours later, the beads were washed with buffer and then incubated with 0.1% SDS/0.5 mg/ml proteinase K at 55 • C for 30 min to discard the protein. Finally, quantitative real-time PCR (qRT-PCR) analysis was performed on the immunoprecipitated RNA to verify the presence of PVT1 and miR-543 using specific primers.

RNA pull-down experiment
The template PVT1 DNA was prepared by PCR, and the product was ligated to pGEM-T Easy Vector and transformed into competent cells. The plasmid prepared by 5 μg was purified with QIAquick PCR purification kit (Huishen Biotechnology Co., Ltd., Shanghai, China, 28104). The transcription reaction of PVT1 was labeled with biotin-UTP (Yage Biotechnology Co., Ltd., Shanghai, China, YS-10258RJ) in vitro, and OC cells were transfected with biotinylated miR-543-Wt, miR-543-Mut and negative control Bio-NC, respectively. Forty-eight hours later, the cell lysate were incubated with M-280 Streptomyces magnetic beads according to the production instructions. The washed beads were then placed in an RNase-free solution, followed by a 10-min incubation with equal volume of biotinylated miR-543 in binding and washing buffer on the rotator at room temperature, and then the beads with fixed miR-543 fragments were cultivated with 10 mM EDTA pH 8.2 and 95% formamide at 65 • C for 5 min. TRIzol was used to purify the bound RNA and qRT-PCR was utilized to detect the PVT1 level in the bead-bound RNA complex.

Statistical methods
In the present study, GraphPad 6 was used for data analysis and picture drawing. Intergroup comparisons were performed by the independent sample t test, while multigroup comparisons were conduced by one-way ANOVA, and post-hoc pairwise comparisons were done by LSD t test. Multi-time profiles were analyzed by repeated measures analysis of variance, and the post-hoc test was conducted by Bonferroni. Pearson test was utilized for correlation analysis, Kaplan-Meier method was applied to draw the survival curve, and Log-rank test was used to evaluate the difference in survival time between groups. A statistically significant difference was assumed at P<0.05.

Up-regulated PVT1 in OC samples and cell lines
PVT1 was dramatically up-regulated in OC patients' cancer tissue samples. The follow-up was successfully completed, and the 5-year OS was 38.10% (16/42). We used the median (0.90) as the cut-off point of high and low expression, and found that high PVT1 expression was significantly correlated with the lower 5-year OS in OC patients. In addition, PVT1 up-regulation was also observed in OC cell lines, and the up-regulation was more pronounced in OVCAR-3 and TOV-112D cell lines, so the two were selected for further analysis. All the above results were statistically significant (P<0.05) (Figure 1).

Knocking down PVT1 was not conducive to OC cell proliferation, migration and invasion, but induced apoptosis
After knocking down PVT1 and transfecting it into OC cells, we found that PVT1 was evidently down-regulated, cell proliferation, migration and invasion behaviors were remarkably suppressed, and apoptosis was noticeably induced, with statistically significant differences (P<0.05) (Figure 2).

Targeting relationship between PVT1 and miR-543
We found through Star Base (http://starbase.sysu.edu.cn/index.php) that PVT1 and miR-543 had potential target sites. After up-regulation of miR-543, the PVT1-Wt luciferase activity was noticeably decreased (P<0.05), while the PVT1-Mut luciferase activity changed little (P>0.05). In the RIP experiment, PVT1 and miR-543 levels precipitated by Argonaute2 (Ago2) antibodies were evidently elevated than those of IgG, and in RNA pull-down experiments, PVT1 was only pulled down by biotin-labeled miR-543-WT. Subsequently, clinical findings revealed a significantly lower level of miR-543 in OC patients' cancer tissues, and a marked negative correlation between PVT1 and miR-543 (r = −0.638, P<0.001). In OC cell lines, we noticed that miR-543 was also notably reduced, especially in OVCAR-3 and TOV-112D, so further analysis was performed on these two cell lines. Besides, PVT1 was identified to be enhanced and miR-543 was blocked in OC cells transfected with PVT1, while miR-543 was enhanced after transfection of si-PVT1. All the above results were statistically significant (P<0.05) (Figure 3).

PVT1 reversed the anticancer effect of miR-543 on OC cells
MiR-543 expression was enhanced by transfection of miR-543 mimics in OC cells, while its level was attenuated by transfection of PVT1 on this basis. In addition, it was found that the proliferation, migration and invasion of OC cells were inhibited and apoptosis was induced after miR-543 mimics transfection, but after further transfection of PVT1, the effect of miR-543 on OC cells was reduced. All the above results were statistically significant (P<0.05) (Figure 4).

Targeted regulation of SERPINI1 by miR-543
Through TargetScan (http://www.targetscan.org/vert 72/), it was found that there were potential binding sites between miR-543 and SERPINI1, and the up-regulation of miR-543 only significantly reduced the SERPINI1-Wt luciferase activity. The expression and protein levels of SERPINI1 in OC cancer tissues were dramatically elevated than those in adjacent tissues, and were markedly negatively correlated with miR-543 (r = −0.707, P<0.001). SERPINI1 expression and protein levels were also obviously increased in OC cells, and they were the most notably in OVCAR-3 and TOV-112D, which were hereby used for further analysis. After anti-miR-543 transfection, miR-543 was inhibited, but the expression and protein levels of SERPINI1 were increased, while the opposite result was obtained by transfecting miR-543 mimics. All the above results were statistically significant (P<0.05) ( Figure 5).

Down-regulation of miR-543 attenuated si-SERPINI1's inhibitory effect on OC progression
Si-SERPINI1 mediated the reduction in SERPINI1 messenger RNA (mRNA) and protein levels, inhibited the proliferation, migration and invasion of OC cells, and promoted apoptosis. When miR-543 was further down-regulated, the above results were evidently reversed. All the above results were statistically significant (P<0.05) ( Figure 6). Compared between the two groups, *P<0.05 and **P<0.01.

SERPINI1 expression regulated by PVT1 and miR-543
The mRNA and protein levels of SERPINI1 were significantly inhibited after transfection of miR-543 mimics, while those were markedly increased after further transfection of PVT1, both with statistically significant differences (P<0.05) (Figure 7).

Discussion
OC is a fatal gynecological tumor, and patients with metastatic OC often present high mortality and short survival [24]. lncRNAs show abnormal dysregulation in the body and are crucial in regulating the occurrence and development of tumors [25]. In this study, we focused on the role of PVT1 in OC and tapped its underlying regulatory mechanisms.
Located on human chromosome 8q24.21, PVT1 is carcinogenic in various diseases such as liver cancer and colorectal cancer [26]. In studies such as Gou et al. [27], PVT1 regulated the progression of liver cancer by targeting miR-214. And as reported by Wang et al. [28], PVT1 could participate in the metastatic mechanism of colorectal cancer through sponging miR-145. In the present study, we first evaluated the expression of PVT1 in OC tissues and cells. It was found that PVT1 showed high levels in both OC tissues and cell lines, suggesting that it may be the key in the disease progression of OC. The prognostic results demonstrated that high PVT1 level was remarkably associated with lower 5-year OS of patients, indicating that high PVT1 level may predict poor prognosis of patients. While cell analysis of PVT1 exhibited that PVT1 down-regulation inhibited OC cells' proliferation, migration and invasion, and induced apoptosis, which suggested that the down-regulating PVT1 is helpful to the improvement of OC. LncRNA can act as a molecular sponge of miRNA to negatively regulate its expression and cause miRNA targets to be repressed, thus regulating biological functions [29]. Therefore, we explored whether PVT1 functioned as a molecular sponge for miRNA. DLR assay, RIP and RNA pull-down experiments all confirmed the target-controlled relationship between PVT1 and miR-543. Additionally, we detected the high miR-543 expression in OC tissues and cells, and the marked negative correlation between PVT1 and miR-54 through correlation analysis. Further investigation into the role of miR-543 in OC cell behaviors revealed that overexpressed miR-543 greatly prevented the proliferation, migration and invasion of OC cells, and enhanced the apoptosis ability, indicating that miR-543 had anti-tumor activity in OC. Furthermore, up-regulating PVT1 was found to be able to eliminate the positive effect of miR-543 on OC cells, suggesting that the anti-tumor activity of miR-543 might be affected by PVT1.
We further explored the ways in which miR-543 played a part in OC progression. The potential binding sites between SERPINI1 and miR-543 were identified by TargetScan, and later their binding relationship was confirmed by DLR. SERPINI1 belongs to the family of serine protease inhibitors and has been reported to present differential expression in the probe group of epithelial OC cell lines [30,31]. Our research found that inhibiting SERPINI1 was not conducive to the proliferation, migration or invasion of OC cells, but was beneficial to apoptosis. However, when  we further down-regulated miR-543, the above results were obviously reversed, suggesting that although SERPINI1 exerted a regulatory effect on OC cell behaviors, this effect was also affected by miR-543. Finally, we found that SERPINI1 was positively regulated by PVT1 and negatively regulated by miR-543.
Although the present study confirmed that PVT1, as an endogenous RNA, positively regulates SERPINI1 through sponging miR-543, there is still some room for improvement. First of all, whether PVT1 is an independent prognostic factor of OC, and whether PVT1 has the potential to predict the prognosis of OC remain to be explored. Secondly, the influence of PVT1 on OC chemical resistance still need to be discussed to clarify its potential regulatory mechanism on OC chemical sensitivity. Furthermore, the potential value of PVT1 in predicting the recurrence of OC can be supplemented in the future studies. Last but not the least, we can supplement the correlation diagram of star database for evidence.

Conclusion
To sum up, we propose for the first time the molecular mechanism by which PVT1-mediated miR-543/SERPINI1 axis regulates the progression of OC, which may be beneficial to the treatment of OC patients.