CircLNPEP promotes the progression of ovarian cancer through regulating miR-876-3p/ WNT5A axis.

CircRNA LNPEP has been shown to promote the development of hepatocellular carcinoma, but its function in ovarian cancer (OC) remains unclear. The Kaplan-Meier method was used to analyze the clinical significance of circLNPEP expression in OC patients. The stability of circLNPEP was detected by actinomycin D and RNase R treatment. The correlations between miR-876-3p and two genes (circLNPEP and WNT5A) were predicted by bioinformatics analysis and confirmed by dual-luciferase reporter assay. Expressions of circLNPEP, miR-876-3p, and WNT5A were determined by qRT-PCR and western blot. The effect of circLNPEP/miR-876-3p/WNT5A axis on viability, proliferation, migration, and invasion, and angiogenesis of cells was determined by cell function experiment and rescue experiment. Xenograft tumor mice were constructed for in vivo verification. Expressions of apoptosis, epithelial mesenchymal transition (EMT)-related genes, and CD34 were determined by qRT-PCR, western blot and immunohistochemistry. High level of circLNPEP was related to poor prognosis in OC. CircLNPEP was highly expressed in OC tissues and cell lines, mainly distributed in the cytoplasm, while miR-876-3p was the opposite. MiR-876-3p targeted and negatively correlated with circLNPEP and WNT5A. Sh-circLNPEP repressed cell viability, proliferation, migration, invasion, angiogenesis, and EMT but promoted apoptosis, which were related to its regulation of expression of EMT- and apoptosis-related genes, WNT5A, and CD34. The regulatory effect of sh-circLNPEP can be reversed by miR-876-3p inhibitor, and that of miR-876-3p inhibitor can be reversed by sh-WNT5A. CircLNPEP promoted cancer cell proliferation, EMT and angiogenesis, and inhibited apoptosis by regulating miR-876-3p/WNT5A axis.

Introduction

Ovarian cancer (OC) is a malignant tumor with the highest mortality rate of female reproductive system [1]. The long eclipse period of OC makes early diagnosis difficult. The study found that only 25% of the patients were in stage I or II at the time of treatment, whilst over 75% of the patients were already in stage III–IV [2]. The main treatment for OC is surgery plus continuous postoperative chemotherapy. Although the cure rate of OC has improved in recent years, the 5-year survival rate remains as low as 30%[3], but that of the early-detected OC can reach 90% after surgery and chemotherapy [4]. Therefore, the early diagnosis of OC is crucial to the prognosis of patients with OC.

Non-coding RNA is a type of RNA transcript without protein coding function, including microRNA (miRNA), long non-coding RNA (lncRNA) and circular RNA (circRNA). They can form a competitive endogenous RNA (ceRNA) regulatory network, which plays an important role in the occurrence and development of malignant tumors [5]. MiRNA exerts a gene regulation function by combining with the target gene. Evidence has proved that miRNA is abnormally expressed in OC, which is closely correlated with the diagnosis, prognosis, and treatment of OC [6]. Metastasis of OC is orchestrated by several interconnected biological processes, such as epithelial mesenchymal transition (EMT), increment in cell migration, destruction of the ECM, avoidance of apoptosis, angiogenesis, and immune suppression [7]. Some miRNAs have been reported to facilitate metastasis, primarily by targeting negative regulators of these processes in OC [8].

The current research on the function of circRNA mainly focuses on the circRNA-miRNA-mRNA axis. The 3’end of circRNA is connected to the 5’end to form a closed loop structure, which is extremely stable and can resist degradation by RNase [9]. The exon-derived circRNA, mainly distributed in the cytoplasm, can function as a sponge to adsorb miRNA, thereby up-regulating the expressions of miRNA target genes. The intron- and exon-intron-derived circRNAs are mainly located in the nucleus and participate in transcriptional regulation and selective splicing [10]. In the field of tumor research, circRNA participates in tumor progression by up-regulating proto-oncogenes or tumor suppressor genes. CircMYLK promotes cell proliferation, migration, and EMT by adsorbing miR-29a [11]. The circAGFG1/miR-195-5p/cyclin E1 axis is involved in regulating the proliferation and metastasis of breast cancer cells [12]. CircLNPEP (has_circ_0130087) is considered as a risk factor for OC with hazard ratio (HR) values >1 [13]. However, its role in OC remains unclear. Therefore, this article is dedicated to exploring the role and specific mechanisms of circLNPEP in OC.

Materials and methods

Sample collection

Forty pairs of OC tissues and adjacent tissues were collected surgically from inpatients of gynecology department of the First Affiliated Hospital of Zhengzhou University from May 2015 to June 2019. All tissue samples were taken immediately after operation and stored in liquid nitrogen, which were confirmed by highly professional pathologists. None of the patients have received radiotherapy, chemotherapy, or other immunobiological treatments before surgery. The written informed consent was signed by all patients. The study was approved by the Ethics Committee of the First Affiliated Hospital of Zhengzhou University with the following reference number: T201504027.

Cell culture

Human OC cell lines A2780 (CL-0013), SKOV-3 (CL-0215), OVCAR3 (CL-0178), SK-BR-3 (CL-0211), OV-56 (96,020,759) and TOV-21 G (CRL-11,730) were purchased from Procell life science and technology Co., Ltd. (Wuhan, China) (https://www.procell.com.cn/) or Biofeng company (http://www.biofeng.com/). Human normal ovarian cell-line IOSE80 (CC-Y1679) were purchased from EK-Bioscience (Shanghai, China) (http://www.elisakits.cn/Index/index.html). Cells were grown in Roswell Park Memorial Institute (RPMI)-1640 medium (72,400,120, Gibco, USA) containing 10% fetal bovine serum (FBS) (10,091, Gibco, USA) and 1% penicillin-streptomycin (15,140–122, Gibco, USA) at 37°C with 5% CO2 (Forma Steri-Cycle, Thermo Scientific, USA).

Quantitative reverse transcription polymerase chain reaction (qRT-PCR)

The RNA from cytoplasmic or nuclear fractions of cells was extracted by Cytoplasmic & Nuclear RNA Purification Kit (Norgen Biotek, 21,000, Canada). According to the manufacturer’s instructions, total RNA of OC tissues and cells was extracted by TRIzol (B511311, Sangon, China). Concerning circRNA and mRNA, reverse transcription and fluorescent quantitative PCR were performed by One-Step qRT-PCR kit (D7268M, Beyotime, China). Regarding miRNA, qRT-PCR was conducted by miRNA first Strand cDNA Synthesis Kit (MR101, vazyme, China) and miRNA Universal SYBR® qPCR Master Mix (MQ101, vazyme, China). Real-time PCR Detection system (CFX96, Bio-rad, USA) with 2−ΔΔCt method was applied to detect the relative expression levels of genes [14]. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as a reference gene for circRNA and mRNA. U6 was used as an internal control for miRNA. The details of primers are listed in (Table 1).

Table 1.

Primers and RNA sequences used in this study

List of oligonucleotide sequences	5’–> 3’
Primers for PCR
circLNPEP Forward	CAGTCAGAGGCAACCTTCCG
circLNPEP Reverse	TGTCACGACTGCCCACGAC
linear-LNPEP Forward	GTCGTGGGCAGTCGTGACA
linear-LNPEP Reverse	CGTGACGGACCCTTGACATTA
GAPDH Forward	CAATGACCCCTTCATTGACC
GAPDH Reverse	TTGATTTTGGAGGGATCTCG
U6 Forward	CTCGCTTCGGCAGCACA
U6 Reverse	ACGCTTCACGAATTTGCGT
hsa- miR-140-3p Forward	GCGCGTACCACAGGGTAGAA
hsa- miR-140-3p Reverse	AGTGCAGGGTCCGAGGTATT
hsa- miR-361-3p Forward	ACACTCCAGCTGGGTCCCCCAGGTGTGATTC
hsa- miR-361-3p Reverse	CTCAACTGGTGTCGTGGAGTCGGCAATTCAGTTGAGAAATCAGA
hsa- miR-330-5p Forward	TCTCTGGGCCTGTGTC
hsa- miR-330-5p Reverse	CCAGTTTTTTTTTTTTTTTGCCTAAG
hsa-miR-876-3p Forward	CTGTGGTGGTTTACAAAGTAATT
hsa-miR-876-3p Reverse	GTGCAGGGTCCGAGGT
Bcl-2 Forward	GGTGGGGTCATGTGTGTGG
Bcl-2 Reverse	CGGTTCAGGTACTCAGTCATCC
Bax Forward	TTTTGCTTCAGGGTTTCAT
Bax Reverse	ACACTCGCTCAGCTTCTTG
E-cadherin Forward	CTCGGCCTGAAGTGACTCGTAAC
E-cadherin Reverse	CAGCAACGTGATTTCTGCATTTC
N-cadherin Forward	TGAAGTCCCCAATGTCTCCA
N-cadherin Reverse	GCATCATCATCCTGCTTATCC
Vimentin Forward	GGACCAGCTAACCAACGACA
Vimentin Reverse	AAGGTCAAGACGTGCCAGAG
c-Met Forward	CTTTGTGAGCAGATGCGGAG
c-Met Reverse	GGTTTATCTTTCGGTGCCCAG
FOXA2 Forward	AAGACCTACAGGCGCAGCTA
FOXA2 Reverse	CCTTCAGGAAACAGTCGTTGA
WNT5A Forward	TCGTTAGCAGCATCAGTCCACA
WNT5A Reverse	GACCTGTGCCTTCGTGCCTA
ZEB1 Forward	GCAGTCTGGGTGTAATCGT
ZEB1 Reverse	TTGCAGTTTGGGCATTCAT

Actinomycin D and RNase R treatment

To test the stability of circ-leucyl and cystinyl aminopeptidase (LNPEP), the cells were treated with actinomycin D (A1410, Merck, German) and RNase R (RNR07250, epicenter, USA). Total RNA (10 μg) was incubated with RNase R (40 U) for 60 minat 37°C. Actinomycin D (2 mg/mL) or the same concentration of dimethyl sulfoxide (DMSO, D2650, Merck, German) was added to the medium for cell culture for 4 h, 8 h, 12 h and 24 h. After the cells were treated with RNase R or actinomycin D, the cells were collected for qRT-PCR to detect the expression levels of circLNPEP and LNPEP.

Cell transfection

MiR-876-3p mimic (M, B01001), mimic control (MC, B04001), miR-876-3p inhibitor (I, B03001), inhibitor control (IC, B04006), full-length circLNPEP and sh-circLNPEP were synthesized by Genepharma (Shanghai, China). Full-length circLNPEP was inserted into pcDNA3.1+ vector (V87020, ThermoFisher, USA), while sh-circLNPEP or sh-WNT5A was inserted into pGPU6 vector (Genepharma, C02001, China). The empty vector was used as control. Cells (4 × 104 per well) were cultured to 80% confluence in a 6-well plate. Lipofectamine 3000 (L3000-015, Invitrogen, USA) and vectors/miRNAs were diluted by Opti-MEM medium (31,985,062, Gibco, USA). After standing for 20 min, the vectors/miRNAs were transfected into cells using diluted Lipofectamine 3000 and then incubated for 24 h. The transfection efficiency was tested by qRT-PCR.

The experiments were divided into four groups: vector + IC group (SKOV3 and SK-BR-3 cells were transfected with pGPU6 vector and miR-876-3p IC), sh-circLNPEP/sh-WNT5A + IC group (SKOV3 and SK-BR-3 cells were transfected with pGPU6 vector containing sh-circLNPEP/sh-WNT5A sequences and miR-876-3p IC), vector + I group (SKOV3 and SK-BR-3 cells were transfected with pGPU6 vector and miR-876-3p I) and sh-circLNPEP/sh-WNT5A + I group (SKOV3 and SK-BR-3 cells were transfected with pGPU6 vector containing sh-circLNPEP/sh-WNT5A sequences and miR-876-3p I).

Cell counting kit-8 (CCK-8)

CCK8 kit (C0037, Beyotime, China) was applied to examine the effect of sh-circLNPEP on viability of OC cell lines. Cells (3 × 103 per well) transfected with sh-circLNPEP or sh-NC were seeded into a 96 well plate and incubated for 24 h or 48 h. After incubation, the cells reacted with 10 μl CCK8 and then incubated for 2 h. The absorbance at 450 nm was counted by a multiple detection reader (SpectraMax5, Molecular Devices, USA).

Colony formation

Cells (200 per well) transfected with sh-NC/sh-circLNPEP or cotransfected with vector/sh-circLNPEP + miR-876-3p IC/I were seeded into a 6-well plate and cultured with RPMI-1640 medium containing 10% FBS for 10 days. After incubation, cells were fixed by 4% paraformaldehyde (E672002, Sangon, China) for 20 min and stained with crystal violet (E607309, Sangon, China) for 30 min. The relative colony formations were counted by ImageJ2x (Rawak Software, Germany).

Wound healing assay

Transfected cells (2 × 104 per well) were cultured in a 6-well plate. After the cells reached 80% confluence, the cultured medium was replaced with serum-free medium. The plates were scratched by a pipette tip and floating cells were removed by phosphate buffer saline (PBS) (C0221A, Beyotime, China). The migration rates were observed under an optical microscope (BX53M, Olympus, Japan, 100×) and calculated by Image J2x (Rawak Software, Germany) after 48 h.

Transwell assay

Transfected cells (2 × 104 per well) were resuspended with serum-free medium and injected into upper chamber (354,480, Coring, USA). The lower chamber contained medium with 10% FBS. Following invasion for 48 h, cells on the upper membrane were wiped out, while cells on the bottom of the membrane were washed by PBS and fixed by 4% paraformaldehyde for 15 min. After fixing, cells were stained with 0.1% crystal violet (C0121, Beyotime, China) for 20 min at room temperature. The invasion rates were observed under an optical microscope (200×) and counted by ImageJ2x.

Tube formation assay

The matrigel (354,230, BD Biosciences, USA) was melted in advance and placed in the icebox during the experiment. Matrigel was added into a 96-well plate at 50 μL/well. After the matrigel solidifies, Human Umbilical Vein Endothelial Cells (HUVECs) (2 × 104) were seeded into a 96-well plate. Then, the supernant of OC cells transfected with sh-NC/sh-circLNPEP or cotransfected with vector/sh-circLNPEP + miR-876-3p IC/I was added onto 96-well plates and incubated for 4 h. The tube lengths was observed under an optical microscope (200×) and measured by ImageJ2x.

Biotinylated RNA pull-down assay

The biotin-labeled circLNPEP probe, oligo probe, biotin-miR-876-3p-wild type and biotin-miR-876-3p-mutuant (MUT) were synthesized by GenePharma (Shanghai, China). The biotinylated RNA pull-down assay was conducted as previously described [15]. Briefly, for circLNPEP pulled down miRNAs, the circLNPEP probe and its control probe were incubated with streptavidin magnetic beads (65,001, ThermoFisher, USA) for 2 h at room temperature. Cells (1 × 107) were lysed and incubated with probe-coated beads at 4°C overnight. The abundance of miRNAs was analyzed by qRT-PCR. For miR-876-3p pulled down circLNPEP, SKOV3 and SK-BR-3 cells were transfected with biotin-miR-876-3p-WT or biotin-miR-876-3p-MUT by lipofectamine 3000. Subsequently, the cells were processed in the same manner as above, including lysis, incubation with magnetic beads and qRT-PCR.

Luciferase activity assay

The binding site of miR-876-3p and circLNPEP was predicted using CircInteractome (https://circinteractome.nia.nih.gov/), whilst that of miR-876-3p and WNT5A was predicted by Targetscan (https://www.targetscan.org). Then the strength of combination was validated by Dual-Luciferase Reporter Assay System (E1910, Promega, USA). We constructed pmirGLO-circLNPEP-WT, pmirGLO-circLNPEP-MUT, pmirGLO-WNT5A-WT and pmirGLO-WNT5A-MUT using pmir-GLO System (E1330, Promega, USA). Then pmirGLO-circLNPEP-WT/pmirGLO-circLNPEP-MUT, pmirGLO-WNT5A-WT/pmirGLO-WNT5A-MUT and miR-876-3p M/MC were co-transfected into cells by lipofectamine 3000 for 24 h. Multifunctional microplate reader (Fluoroskan Ascent FL, Thermo, USA) was applied to examine luciferase activity.

Western blot

Western blot was performed as previously described [16]. The protein sample was collected from OC tissues and cell lines by RIPA buffer (P0013B, Beyotime, China) supplemented with protease inhibitors (p1005, Beyotime, China) and Phenylmethylsulfonyl Fluoride (PMSF) (ST505, Beyotime, China). The protein concentration was quantified by BCA protein assay kit (P0009, Beyotime, China). Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was conducted to separate proteins. After electrophoresis, the proteins were transferred into polyvinylidene fluoride membrane (FFP24, Beyotime, China). After blocking by 5% skim milk for 1 h, the membrane was incubated with following primary antibodies overnight at 4°C: anti-cleaved caspase-3 (1:200, 17 kDa, ab2302, Abcam, UK), anti-bcl-2 (1:1000, 26 kDa, ab26, Abcam, UK), anti-bax (1:1000, 21 kDa, ab32503, Abcam UK), anti-E-cadherin (1:10,000, 97 kDa, ab40772, Abcam, UK), anti-N-cadherin (1:1000, 130 kDa, ab18203, Abcam, UK), anti-vimentin (1:1000, 54 kDa, ab92547, Abcam, UK), anti-WNT5A (1:1000, 42 kDa, ab179824, Abcam, UK) and GAPDH (1:1000, 36 kDa, ab8245, Abcam, UK). Next, the membrane was washed with 1% Tris-buffered saline with Tween 20 (TBST) and incubated with secondary antibody goat anti rabbit (1:10,000, ab205718, Abcam, UK) for 2 h at room temperature. The protein bands were visualized by gel imaging system (FluorChem FC3, Alpha, USA) with an ECL luminescence kit (PE0010, Solarbio, China), the intensity of which was counted by ImageJ2x (Rawak Software, Germany). GAPDH was used as internal control.

Animals

Male BALB/c nude mice (4 weeks old) were purchased from Guangdong Medical Experimental Animal Center (http://www.gdmlac.com.cn/) and maintained under specific pathogen-free conditions at The First Affiliated Hospital of Zhengzhou University Central Animal Laboratory. The room temperature and humidity were constant at 22°C and 50%, respectively. The rats had free access to food and water. Animal experiments were approved by The First Affiliated Hospital of Zhengzhou University Animal Ethics Committee with the reference number T201908035.

Nude mice were randomly divided into four groups, each with eight mice. SKOV3 cells (5 × 106 cells each) transfected with vector/sh-circLNPEP and miR-876-3p IC/I were subcutaneously injected into the right flank of nude mice. The volume of the tumor was measured every fifth day (volume = width2 × length × 1/2). Nude mice were euthanized by an anesthetic overdose (intraperitoneal injection of pentobarbital (P0500000, Merck, German) at a dose of 200 mg/kg) after 4 weeks. The tumor was removed and the tumor volume and weight were measured. A part of the tumor sample was used to extract RNA and protein, and the rest was fixed with 4% paraformaldehyde and subjected to immunohistochemistry experiments.

Immunohistochemistry (IHC)

IHC was conducted as aforementioned [17]. In brief, the fixed tissues were embedded in paraffin. The paraffin sections were dewaxed, hydrated, incubated in 3% hydrogen peroxide (H1009, Merck, German) for 8 min at room temperature, blocked with 5% BSA (E661003, Sangon, China) at room temperature for 10 min and then reacted with primary antibody CD34 (ab81289, Abcam, UK) at 4°C overnight. The sections were then visualized with Metal Enhanced DAB Substrate Kit (DA1015, Solarbio, China). Afterward, the sections were counterstained by hematoxylin, dehydrated and mounted by neutral balsam (G8590, Solarbio, China). The image was captured using an optical microscope (magnification, 200 ×).

Statistical analysis

Graph Prism v8.0 (Graphpad software, California, USA) and SPSS 20.0 (SPSS, Chicago, USA) were employed to analyze the data. The results were represented as mean ± standard deviation. Prognostic analysis of circLNPEP expression was performed in both progression-free and overall survival using the Kaplan–Meier method. The expressions of circLNPEP and miR-876-3p in tumor tissues and adjacent tissues were compared by paired-sample t test. Spearman analysis was used to analyze the correlations between circLNPEP and miR-876-3p, and between WNT5A and miR-876-3p. Differences among multiple groups were analyzed by one-way analysis of variance (ANOVA), whereas pairwise comparisons between groups was measured by Tukey. A value of P < 0.05 was considered as statistically significant.

Results

High level of circLNPEP was related to ovarian cancer progression and poor prognosis

The relationship between clinical characteristics and circLNPEP expression in OC was shown in (Table 2). The level of circLNPEP (high or low) in tissues was distinguished according to the mean expression of circLNPEP in patients. High level of circLNPEP was associated with advanced tumor stage, advanced International Federation of Gynecology and Obstetrics (FIGO) stage, and larger residual tumor. There was no difference in circLNPEP expression in the histological subtypes of OC. Prognostic analysis of circLNPEP expression was also performed in both overall survival and progression-free survival using the Kaplan–Meier method, indicating that high level of circLNPEP was related to lower overall survival (Figure 1(a), p = 0.021) and progression-free survival (Figure 1(b), p = 0.014).

Table 2.

The relationship between the clinical characteristics and circLNPEP expression in ovarian cancer

Features	circLNPEP expression		P value
	Low (n = 20)	High (n = 20)
Age			0.525
<55	8	10
≥55	12	10
Tumor grade			0.027
G1	14	7
G2+ G3	6	13
FIGO stage			0.025
I-II	12	5
III-IV	8	15
Residual tumor			0.004
NO	15	6
YES	5	14
Histological subtypes			0.836
Serous	10	11
Muncinous	2	3
Endometrioid	4	2
clear-cell	1	2
mixed	3	2

Low/high expression was obtained by the sample mean. FIGO stage: International Federation of Gynecology and Obstetrics stage.

Figure 1.

The expression of circLNPEP in ovarian cancer and the stability of circLNPEP

Upregulation of circLNPEP in ovarian cancer tissues and cell lines and the stability of circLNPEP

The level of circLNPEP in OC tissues and adjacent tissues was determined, confirming the notably higher level in OC tissues compared to that in adjacent tissues (Figure 1(c), p < 0.001). Similarly, compared with normal ovarian epithelial cell-line IOSE80, circLNPEP was highly expressed in OC cell lines A2780, SKOV3, OVCAR3, SK-BR-3, OV-56 and TOV-21 G (Figure 1(d), p < 0.001). On the grounds that circLNPEP was highly expressed in SKOV-3 and SK-BR-3, these two cell lines were selected for subsequent experiments. Then we determined the stability of circLNPEP. After Rnase R treatment, the level of LNPEP was decreased markedly, while that of circLNPEP remained unchanged (Figure 1(e), p < 0.001). In addition, after treatment with the transcription inhibitor Actinomycin D, the half-life of circLNPEP surpassed 24 h, while the level of linear LNPEP was decreased obviously after 4 h, suggesting that circLNPEP was a stable ring structure (Figure 1(f,g), P < 0.001).

Silencing of circLNPEP inhibited the viability, proliferation, migration, and invasion of ovarian cancer cells, as well as angiogenesis of HUVECs

CircLNPEP was silenced to observe its effect on the biological behavior of OC cells. We first determined the efficiency of silencing circLNPEP, and sh-circLNPEP effectively suppressed the expression of circLNPEP, but had no significant effect on the expression of linear LNPEP (Figure 2(a), p < 0.001). CCK-8 experiment proposed that sh-circLNPEP transfection for 24 h and 48 h repressed the viability of OC cells (Figure 2(b), p < 0.05). The relative colony formations were attenuated in sh-circLNPEP group compared with sh-NC group (Figure 2(c,d), P < 0.01). Additionally, circLNPEP silencing diminished the migration rate and invasion rate of OC cells and OC cell-induced angiogenesis of HUVECs (Figure 2(e-j), P < 0.001).

Figure 2.

Sh-circLNPEP inhibited the viability, proliferation, migration, and invasion of ovarian cancer cells, as well as angiogenesis of HUVECs

MiR-876-3p targeted and negatively correlated with circLNPEP which was mainly expressed in the cytoplasm

We identified circLNPEP in OC cells, with the results demonstrating that circLNPEP is mainly located in the cytoplasm (Figure 3(a)). In order to determine the interaction between circLNPEP and these miRNAs, we designed a biotinylated circLNPEP probe at the 3’end, verifying that the circLNPEP probe could pull down circLNPEP in SKOV3 and SK-BR-3, while circLNPEP overexpression improved pull-down efficiency (Figure 3(b), p < 0.001). We, therefore, predicted that circLNPEP might target through CircInteractome (miR-140-3p, miR-361-3p, miR-330-5p, miR-876-3p). Among the 4 candidate miRNAs, only miR-876-3p was significantly pulled down by circLNPEP in SKOV3 and SK-BR-3 (Figure 3(c), p < 0.001). In order to further confirm the combination of miR-876-3p and circLNPEP, we used wild type and mutuant type of biotin-labeled miR-876-3p to pull down overexpressed circLNPEP. The results showed that the wild-type miR-876-3p exerted more pull-down effects on circLNPEP compared with the mutant type (Figure 3(d), p < 0.001). The binding site of circLNPEP and miR-876-3p was predicted by CircInteractome (Figure 3(e)). Dual-luciferase reporter gene experiment pointed out that miR-876-3p mimic suppressed the luciferase activity in circLNPEP-WT group instead of circLNPEP-MUT group (Figure 3(f), p < 0.001). Moreover, miR-876-3p was found to be alleviated in OC tissues (Figure 3(i), p < 0.001), the expression of which was upregulated by sh-circLNPEP but downregulated by pc-circLNPEP (Figure 3(g,h), P < 0.001). Spearman analysis showed that miR-876-3p was negatively correlated with circLNPEP level (Figure 3(j-r)r = −0.446, P = 0.004).

Figure 3.

CircLNPEP serves as a sponge for miR-876-3p in ovarian cancer cell lines

Co-transfection of miR-876-3p I + sh-circLNPEP reversed the regulation of miR-876-3p I or sh-circLNPEP on biological behavior of ovarian cancer cells

We co-transfected vector + miR-876-3p IC, sh-circLNPEP + miR-876-3p IC, vector + miR-876-3p I and sh-circLNPEP + miR-876-3p I into SKOV3 and SK-BR-3 cells, respectively. Compared with vector + IC group, the expression of miR-876-3p was upregulated in sh-circLNPEP + IC group but downregulated in vector + I group. However, the regulatory effects of sh-circLNPEP + IC group and vector + I group were neutralized by the co-transfection of sh-circLNPEP + I (Figure 4(a), p < 0.01). Then we observed the changes in biological functions of OC cells after co-transfection.Cell clone formation (Figure 4(b,c), P < 0.01), migration rates (Figure 5(a,b), P < 0.01), invasion rates (Figure 5(c,d), P < 0.01) and OC cells-induced angiogenesis of HUVECs (Figure 5(e,f), P < 0.05) were inhibited in sh-circLNPEP + IC group but promoted in vector + I group. However, the effects on biological functions of the OC cells regulated by miR-876-3p I or sh-circLNPEP were overturned by co-transfection of miR-876-3p I + sh-circLNPEP (P < 0.05). The above results indicate that circLNPEP regulates the proliferation, migration, and invasion of ovarian cancer cells and the angiogenesis of HUVECs induced by OC cells through targeting miR-876-3p.

Figure 4.

Co-transfection of miR-876-3p I and sh-circLNPEP reversed regulation of miR-876-3p I or sh-circLNPEP on proliferation of ovarian cancer cell lines

Figure 5.

Co-transfection of miR-876-3p I and sh-circLNPEPreversed regulation of miR-876-3p I or sh-circLNPEPon migration and invasion of ovarian cancer cells, as well as angiogenesis of HUVECs

Co-transfections of miR-876-3p I + sh-circLNPEP reversed the regulation of miR-876-3p I or sh-circLNPEP on expressions of apoptosis-related gene, EMT-related gene and WNT5A

We detected the expressions of apoptosis and EMT-related genes by qRT-PCR and western blot. The protein levels of cleaved caspase-3, Bax and E-cadherin were elevated in sh-circLNPEP + IC group but decreased in vector + I group, whereas those of Bcl-2, N-cadherin and vimentin showed opposite changes in the two groups (Figure 6(a,b,d,e), P < 0.05). Nevertheless, co-transfection of miR-876-3p I + sh-circLNPEP neutralized the sh-circLNPEP +IC- or vector + I-regulated expressions of cleaved caspase-3, Bax, Bcl-2, E-cadherin, N-cadherin and vimentin (P < 0.05). The mRNA levels of Bcl-2, Bax, E-cadherin, N-cadherin and vimentin presented similar trends to the western blot results (Figure 6(c-f), P < 0.05). Additionally, co-transfection of vector + miR-876-3p IC, sh-circLNPEP + miR-876-3p IC, vector + miR-876-3p I and sh-circLNPEP + miR-876-3p I had no effects on the expressions of mesenchymal-epithelial transition factor (c-MET), Forkhead Box A2 (FOXA2) and Zinc-finger E-box binding protein 1 (ZEB1) (Figure 7(a,b)). However, the mRNA and protein levels of WNT5A were suppressed by sh-circLNPEP + IC but promoted by vector + I (Figure 7(a-d), P < 0.05). The suppression or promotion induced by sh-circLNPEP or miR-876-3p I was also reversed by co-transfection of miR-876-3p I + sh-circLNPEP (P < 0.01). The above results proved that circLNPRP targets miR-876-3p to regulate the expression of apoptosis- and EMT-related molecules and WNT5A in OC cells.

Figure 6.

Co-transfection of miR-876-3p I + sh-circLNPEP reversed the regulation of miR-876-3p I or sh-circLNPEP on expressions of apoptosis-related gene and EMT-related gene

Figure 7.

Co-transfection of miR-876-3p I + sh-circLNPEP reversed the regulation of miR-876-3p I or sh-circLNPEP on expression of WNT5A

WNT5A bound to miR-876-3p and WNT5A silencing rescued the miR-876-3p I-induced migration and invasion of ovarian cancer cells

Through Targetscan and luciferase activity assay, we found that miR-876-3p directly targeted WNT5A (Figure 8(a-b), P < 0.001). Moreover, WNT5A expression was up-regulated (Figure 8(c), p < 0.001) and negatively associated with miR-876-3p expression in OC tissues (Figure 8(d), p = 0.012). Then, we co-transfected vector + miR-876-3p IC, sh-WNT5A + miR-876-3p IC, vector + miR-876-3p I and sh-WNT5A + miR-876-3p I into SKOV3 and SK-BR-3 cells, respectively. In accordance with wound-healing and transwell assays, cell migration and invasion were restrained by sh-WNT5A+ IC, which were promoted by miR-876-3p I; moreover, the promotion effect of miR-876-3p I was reversed by co-transfection of sh-WNT5A and I (Figure 8(e-h), P < 0.01). The above results confirmed that miR-876-3p regulates the migration and invasion of OC cells by targeting WNT5A.

Figure 8.

WNT5A bound to miR-876-3p and WNT5A silencing rescued the miR-876-3p suppression-induced migration and of ovarian cancer cells

We observed the effects of circLNPEP and miR-876-3p on tumor volume and weight through nude mouse tumor formation experiments. As shown in (Figure 9(a)), the tumor volume and weight were the smallest in sh-circLNPEP + IC group and the largest in vector + I group (Figure 9(b,c), P < 0.001), and moreover, those in sh-circLNPEP + I group were higher than in sh-circLNPEP + IC group but were lower than in vector + I group (P < 0.001). The expressions of apoptosis-related genes in tumor tissues were the same as those in OC cell lines (Figure 9(d-f), P < 0.05). Similarly, the expressions of circLNPEP and miR-876-3p in xenograft tumor tissue showed same trends in tumor tissues and OC cell lines (Figure 9(g,h), P < 0.01). Immunohistochemistry and western blot were performed to determine the effect of circLNPEP/miR-876-3p axis on the expressions of CD34 and WNT5A. The expression of CD34 plays a vital role in mediating cell adhesion and migration. Moreover, inhibiting the expression of CD34 is conducive to suppressing the metastatic ability of cancer cells. The results revealed that co-transfection of miR-876-3p I + sh-circLNPEP overturned the promoting effect of miR-876-3p I and the inhibitory effect of sh-circLNPEP on CD34 and WNT5A expressions (Figure 9(i-k), P < 0.05). The above experiment verified the regulatory effect of circLNPRP/miRA-876-3p axis on apoptosis-related molecules, CD34 and WNT5A in vivo.

Figure 9.

Co-transfection of miR-876-3p I + sh-circLNPEP reversed the regulation of miR-876-3p I or sh-circLNPEP on tumor progression in vivo.

Disscussion

The existing research on circRNA in OC mainly focuses on the influence of circRNA/miRNA/mRNA axis on the biological function and drug resistance of cancer cells [18,19,20]. However, few circRNAs have been reported to exert regulatory effects on OC. Hence, it is urgent to discover more circRNAs to shed new light on the research of circRNA in tumors.

In our study, we found that high level of circLNPEP was associated with advanced tumor stage, advanced FIGO stage, larger residual tumor as well as a poor prognosis. Moreover, circLNPEP was highly expressed in OC tissues and cell lines, indicating that the upregulation of circLNPEP was related to OC. The present results revealed that the viability, proliferation, migration, and invasion of OC cells, as well as angiogenesis of HUVECs were significantly reduced after silencing circLNPEP. Thus, circLNPEP may be the potential target for OC treatment.

Different distributions of circRNA in cells perform diversified functions [10]. We detected that circLNPEP was mainly distributed in the cytoplasm and speculated that its function was mainly to adsorb miRNA to regulate the expression of downstream genes. By means of bioinformatics and molecular biology, we confirmed the targeting relationship between miR-876-3p and circLNPEP. Studies have reported that miR-876-3p is inhibited in gastric cancer, pancreatic adenocarcinoma, and hepatocellular carcinoma, which is inseparable from the metastasis, proliferation, apoptosis, and drug resistance of cancer cells [21,22,23]. In our study, we found that miR-876-3p was also inhibited in OC and miR-876-3p level was negatively correlated with circLNPEP level, suggesting that circLNPEP may function as a sponge to absorb miR-876-3p, thus affecting the biological function of OC cells.

To further verify our conclusion, sh-circLNPEP and miR-876-3p I were co-transfected into cells, which demonstrated that sh-circLNPEP promoted cell proliferation, migration, invasion, and EMT process, and inhibited apoptosis of ovarian cancer cells, as well as promoted angiogenesis of HUVECs through regulating miR-876-3p. Bax and Bcl-2, as members of Bcl-2 protein family, are key regulators of cell apoptosis [24]. The difference lies in that Bax is a pro-apoptotic molecule, while Bcl-2 is an inhibitor of apoptosis [25]. Caspase-3 is the executor of apoptosis, whose activation indicates the occurrence of apoptosis [26]. E-cadherin is involved in cell adhesion, so its absence can promote cell migration, while N-cadherin and vimentin can promote cell angiogenesis and metastasis [27,28]. In line with the effects of sh-circLNPEP and miR-876-3p I on the expressions of these molecules, we determined that circLNPEP suppressed apoptosis and promoted EMT in OC by sponging miR-876-3p. The data suggested that circLNPEP as a sponge to absorb miR-876-3p affects the biological function of OC cells.

The occurrence of EMT is an extremely complicated process, in which multiple transcription factors and signaling pathways play synergistic roles. The c-MET is a receptor of hepatocyte growth factor (HGF), which binds to regulate tissue and angiogenesis [29]. FOXA2 makes profound impacts upon maintaining the morphology and structure of epithelial cells and regulating EMT related to tumor metastasis [30]. WNT5A can initiate the occurrence of EMT in OC [31]. ZEB1, as a member of ZEB family, is a transcriptional regulator of EMT [32]. We confirmed that sh-circLNPEP and miR-876-3p I had no effect on the expressions of c-Met, FOXA2, and ZEB1. However, circLNPEP silencing inhibited WNT5A expression and abrogated the promotive effect of downregulated miR-876-3p on WNT5A expression. Furthermore, we uncovered that WNT5A level was up-regulated in OC, and negatively assosicated with miR-876-3p level. Besides, our data also revealed that downregulated miR-876-3p promoted migration and invasion of OC cells through directly regulating WNT5A expression. These results suggested that circLNPEP may exert regulatory effects on the development of OC through miR-876-3p/WNT5A axis.

We also conducted in vivo experiments as verification. The results pointed out that when circLNPEP was silenced, the size of tumor cell became smaller, apoptosis was increased, and the expressions of CD34 (a sign of blood vessel growth in tumor growth) and WNT5A were decreased, whilst the inhibition of miR-876-3p showed the opposite effects. Nevertheless, co-transfection of sh-circLNPEP and miR-876-3p I neutralized the effects induced by silent circLNPEP and inhibited miR-876-3p. All of these results demonstrated that circLNPEP regulates growth, apoptosis, and vasculature in OC by regulating miR-876-3p/WNT5A axis. Collectively, our research provided the first evidence with regard to the role of circLNPEP in OC, clarifying that its mechanism was to regulate the expression of downstream genes by adsorbing miR-876-3p, thereby promoting cancer cell proliferation, EMT and inhibiting apoptosis. In addition, the inherent weakness of the design in this study is the fact that only one shRNA was used for circLNPEP and WNT5A.

Conclusion

Overall, our research highlights the potential of circLNPEP in the targeted therapy of OC. Also, our findings provide robust evidence that circLNPEP serves as a novel oncogenic circRNA by regulating miR-876-3p/WNT5A axis, as well as a promising prognostic biomarker in OC.