Upregulation of circ_0000069 stimulates cervical cancer development by enhancing proliferation, migration, invasion while inhibiting apoptosis through the miR-4429/ZIC2 axis

Background Cervical cancer (CC) is a common female cancer according to global cancer statistics. The current study was used to investigate the regulatory mechanism of circ_0000069 in CC. Methods The real-time quantitative polymerase chain reaction (RT-qPCR) was used to assess circ_0000069, miR-4429, and zinc finger protein of the cerebellum 2 (ZIC2) expression in CC tissues and cells. Kaplan-Meier analysis was performed in CC patients to analyze the relationship between survival time and circ_0000069 expression. The proliferation, apoptosis, cell cycle of CC cells were detected by 3-(4, 5-dimethylthiazol-2-yl)-2, 5-diphenyl-2H-tetrazol-3-ium bromide (MTT), colony formation, and flow cytometry analyses, respectively. In addition, western blot assay was employed to show expression levels of apoptosis/cell cycle-related proteins, as well as ZIC2. The migration and invasion of CC cells were assessed by transwell analysis. Possible target miRNAs of circ_0000069, along with the interaction between ZIC2 and miR-4429 were confirmed by pull-down and dual-luciferase reporter assays. Eventually, the functional role of circ_0000069 in vivo was clarified with the xenograft experiment in nude mice. Results Circ_0000069 was overexpressed in CC tissues and cells than controls. Furthermore, the silencing of circ_0000069 inhibited proliferation, migration, and invasion while included apoptosis and cell cycle arrest of CC cells, which was overturned by downregulation of miR-4429. Importantly, ZIC2 was a direct target of miR-4429 in CC cells, and we further confirmed that overexpression of miR-4429 suppressed CC progress by decreasing ZIC2 expression in CC cells. Surely, silencing of circ_0000069 inhibited tumorigenesis in vivo. Conclusion Our current results suggested that circ_0000069 exerted its tumorigenic roles by regulation of proliferation, apoptosis, cell cycle, migration, and invasion of si-NC: siRNA FITC: isothiocyanate; short hairpin RNAs.


Background
Cervical cancer (CC) ranks fourth for both incidence and cancer-related death in women according to global cancer statistics 2018 [1]. The metastasis and easy recurrence were major obstacles for improving the survival time of CC patients [2,3]. Therefore, further molecular research on the fundamental mechanism of CC development was meaningful, which may provide a new treatment strategy for CC patients.
Circular RNAs (circRNAs) are new members of the non-coding RNAs with covalently linked ends, generating from back-splicing of exons in various cell lines and across different species [4,5]. Emerging evidence has indicated that circRNAs could function as crucial mediators to participate in initiation and development of multiple diseases [6]. For example, circ_0000069 was derived from the SCL/TAL1 interrupting locus (STIL) gene and located on chr1 (47745912-47748131). Circ_0000069 was reported to be associate with worse clinical outcome of colorectal cancer patients [7]. As far as we know, few studies focused on the function and regulator mechanism of circ_0000069 in CC.
MicroRNAs(miRNAs), non-coding RNAs with 18-22 nucleotides in length, was implicated in cervical carcinogenesis [8]. It had been reported that miRNAs caused mRNA degradation by targeting the 3' untranslated region (3'UTR) [9]. Recently, the anti-oncogenic role of miR-4429 has been confirmed in clear cell renal cell carcinoma [10]. Furthermore, miR-4429 was differentially downregulated in CC cells when compared to normal cells [11]. The potential role of miR-4429 in CC was further investigated in the current study.
Zinc finger protein of the cerebellum 2 (ZIC2), transcription factors, was a member of ZIC family, involving in vertebrate development [12], oncogenesis [13], and transcriptional regulation [14]. Recent studies reported that ZIC2 acted as an oncogene in breast cancer [15]. Interestingly, Chan et al. revealed that overexpression of ZIC2 in CC cells obviously increased cell proliferation and metastasis abilities [16]. Consequently, a previous study 4 suggested that the downregulation of ZIC2 inhibited the invasion potential of nasopharyngeal carcinoma cells [17].
Based on those findings, our investigation focused on the implications and roles of circ_0000069 in CC. We aimed to demonstrate the association relationship among circ_0000069, miR-4429, and ZIC2 network in the development of CC, which may provide potential novel targets for CC therapeutics.

Patient specimens
In the total of 62 paired samples of CC and neighboring non-tumorous tissues were harvested from CC patients with prior written informed consent from the patients at The Affiliated Hospital of Qingdao University. All tissues were confirmed as CC samples by two pathologists following surgery and then frozen in liquid nitrogen. Sample tissues were transferred to a -80℃ refrigerator until total RNA or protein extraction. Patients with CC were assigned to two groups (I+II group and III+ IV groups) according to the patient's clinical stage. In addition, CC patients were divided into high (n=31) and low (n=31) circ_0000069 expression groups with median circ_0000069 expression value as the cutoff.
The correlations between hsa_circ_0000069 expression and clinical characteristics of cervical cancer patients were listed in Table 1. All the procedures were permitted by the Ethics Committee of The Affiliated Hospital of Qingdao University.

Cell culture
The normal cervical epithelial cells (HcerEpic) were bought from the Chinese Academy of Sciences Cell Bank (Shanghai, China). In addition, CC cell lines (HeLa and SW756) were acquired from the American Type Culture Collection (Rockville, MD, USA). The above cells were maintained in RPMI1640 media (Invitrogen, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum (FBS; Invitrogen) and antibodies (penicillin/streptomycin; 5 Invitrogen) in a humidified atmosphere containing 5% CO 2 at 37℃.

RNA isolation and real-time quantitative polymerase chain reaction (RT-qPCR)
RT-qPCR assay was conducted to determine RNA level in cells and tumor tissues. In brief,

Apoptosis and cell cycle
For cell apoptosis analysis was conducted by using an Annexin V labeled with fluorescein isothiocyanate (FITC) Apoptosis Detection Kit I (Thermo Fisher Scientific) following the manufacturer's recommended instructions. HeLa and SW756 cells were collected by 7 trypsin and then re-suspended in binding buffer supplemented with 5 μL of Annexin V labeled with FITC and 5 μL of propidium iodide (PI). After incubation for 20 min, apoptotic cells were monitored by flow cytometry (Applied Biosystems). Apoptosis rate = (Annexin V + PI + cell numbers + Annexin V + PIcell numbers)/total cells×100%). For cell cycle detection, 100 uL of cell suspension (1×10 6 /mL) was treated with PI staining solution contained RNase R and TritonX-100 for 30 min at 4℃, then flow cytometry was conducted to detect cell cycle distribution.

Western blot assay
Briefly, proteins were segregated on 10% sodium dodecyl sulfate polyacrylamide gels. The

Statistical analysis
All data were exhibited as mean ± standard deviation from at least 3 times independent experiments, and P <0.05 was considered statistically significant. All analyses were carried out using the SPSS 21.0 software (IBM, Somers, NY, USA) based on Student's t-test and one-way analysis of variance. In addition, the survival curves of CC patients were plotted using the Kaplan-Meier method and the log-rank test. The correlation between clinicopathological characteristics of CC patients and circ_0000069 expression was assessed by Chi-square test. Pearson's correlation analysis was used to analyze the 10 relationship between miR-4429 and circ_0000069, or ZIC2 mRNA in CC tissues.

Circ_0000069 was increased in CC tissues and cells
As presented in Figure 1A, circ_0000069 was upregulated in cervical cancer tissues than normal tissues with microarray analysis (GSE102686; https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE102686). Furthermore, the schematic model showed that circ_0000069 derived from the STIL gene exons 6 and 7 ( Figure 1B). Our results also confirmed that circ_0000069 was increased in CC tissues when compared with paired normal tissues ( Figure 1C). Interestingly, circ_0000069 showed higher expression in III+ IV grade than I+II grade ( Figure 1D). Consistently, CC cells (HeLa and SW756) showed higher expression of circ_0000069 than HcerEpic cells.
Besides, circ_0000069 was more resistant to RNase R than linear STIL mRNA ( Figure 1F-1G). According to the Kaplan-Meier, CC patients with high expression of circ_0000069 had a poor overall survival when contrasted to low expression group ( Figure 1H). Moreover, circ_0000069 level was associated with tumor size, the tumor, node, and metastasis (TNM) stage, and lymph node metastasis in CC patients (Table 1). Collectively, circ_0000069 was upregulated in CC tissues and cells, and its expression was associated with poor prognosis of CC patients.

Circ_0000069 silencing inhibited proliferation, migration, and invasion while induced cell cycle arrest and apoptosis in CC cells
In subsequent experiments, HeLa and SW756 cells were transfected with si-circ_0000069 or si-NC to explore the functional role of circ_0000069 in CC. As shown in Figure 2A-2B, circ_0000069 was downregulated in HeLa and SW756 cells infected with si-circ_0000069 than si-NC group. After transfection, si-circ_0000069 obviously repressed cell proliferation in HeLa and SW756 cells with MTT assay (Figure 2C-2D). Consistently, colony formation assay results indicated that the silencing of circ_0000069 inhibited colony formation in HeLa and SW756 cells ( Figure 2E). Conversely, apoptotic cells were increased in si-circ_0000069 group than that in si-NC group by performing flow cytometry assay ( Figure   2F). In addition, the proportion of cells in G0/G1 phase was increased while S phase was decreased in HeLa and SW756 cells transfected with si-circ_0000069 than control group ( Figure 2G-2H). To measure cell cycle and apoptosis-related proteins expression, western blot assay was conducted in transfected HeLa and SW756 cells. As displayed in Figure 2I

MiR-4429 was a target of circ_0000069 in CC cells
As shown in Figure 3A-3B, overexpression of circ_0000069 enhanced the pull-down efficiency of biotinylated-circ_0000069 probe. 18 potential target miRNAs were clustered in Figure 3C by starbase and circBank databases analysis. After the biotinylated-circ_0000069 probe interacting with cell lysates from HeLa and SW756 cells, we found that miR-4429 showed the highest abundance among 18 potential target miRNAs in both of HeLa and SW756 cells ( Figure 3D-3E). Binding regions between circ_0000069 and miR-4429 were displayed in Figure 3F. After RNA pull-down assay, circ_0000069 was enriched in the product isolated from Biotin-miR-4429-WT group than Biotin-miR-4429-MUT group ( Figure 3G-3H). The dual-luciferase assay results suggested that miR-4429 mimic repressed the luciferase activity of circ_0000069 WT than control, while luciferase activity of circ_0000069 MUT was not affected by miR-4429 mimic ( Figure 3I-3J). Importantly, we also noticed that miR-4429 was decreased in CC tissues and cells than that in controls ( Figure 3K-3L). MiR-4429 was negatively correlated with circ_0000069 expression in CC tissues ( Figure 3M). Additionally, the silencing of circ_0000069 increased miR-4429 expression in HeLa and SW756 cells ( Figure 3N). Therefore, miR-4429 was negatively regulated by circ_0000069 in HeLa and SW756 cells.

Inhibition of miR-4429 reversed the effects of circ_0000069 knockdown on proliferation, apoptosis, cell cycle, migration, and invasion in CC cells
As shown in Figure 4A

MiR-4429 regulated ZIC2 expression in CC cells
Bioinformatics database analysis predicted that miR-4429 might directly interact 3'UTR of ZIC2 mRNA ( Figure 5A). After that, the results of dual-luciferase reporter assay suggested that luciferase activity of ZIC2-3'UTR-WT was declined by transfection with miR-4429 13 mimic in HeLa and SW756 cells, while luciferase activity of ZIC2-3'UTR-MUT was not affected by miR-4429 mimic when compared with control group (Figure 5B-5C).
Additionally, we noticed that ZIC2 was increased in cervical squamous cell carcinoma and endocervical adenocarcinoma tissues than normal tissues (http://gepia.cancerpku.cn/detail.php) ( Figure 5D). Currently, we confirmed that mRNA and protein expression levels of ZIC2 were overexpressed in CC tissues and cells in comparison with controls ( Figure 5E-5H). In addition, ZIC2 was negatively correlated with miR-4429 expression in CC tissues ( Figure 5I). What's more, overexpression of miR-4429 suppressed ZIC2 expression in CC cells ( Figure 5J-5K). All data indicated that ZIC2 was a functional target of miR-4429 in CC cells.

Upregulation of ZIC2 counteracted the effect of miR-4429 overexpression on proliferation, apoptosis, cell cycle, migration, and invasion in CC cells
The underlying regulatory mechanism between ZIC2 and miR-4429 was further investigated. As shown in Figure 6A-6D, transfection with ZIC2 into HeLa and SW756 cells abolished the inhibitory effects on ZIC2 expression induced by miR-4429 mimic.
Furthermore, transfection with ZIC2 could abolish the inhibiting role of miR-4429 in CC cell proliferation by performing MTT and colony formation assays ( Figure 6E-6G). In addition, overexpression of ZIC2 attenuated miR-4429 mimic-induced apoptosis in HeLa and SW756 cells ( Figure 6H). For cell cycle analysis, G0/G1 phase was increased, and S phase was decreased in miR-4429-overexpressing cells, which was overturned by upregulation of ZIC2 in HeLa and SW756 cells ( Figure 6I-6J). Furthermore, the expression of cell apoptosisrelated proteins and cell cycle protein proteins were measured by western blot assay. Bax was increased, and Bcl-2, CDK2, and CDK4 were decreased in HeLa and SW756 cells after overexpression of miR-4429, which was overturned by overexpression of ZIC2 ( Figure 6K-6L). Results of transwell assay suggested that the upregulation of ZIC2 restored the loss 14 of migration and invasion abilities of HeLa and SW756 cells induced by overexpression of miR-4429 ( Figure 6M-6N). In summary, miR-4429 regulated proliferation, apoptosis, cell cycle, migration, and invasion of CC cells by targeting ZIC2.

Circ_0000069 regulated ZIC2 expression by targeting miR-4429 in CC cells
Initially, our results confirmed that circ_0000069 was positively correlated with ZIC2 mRNA expression in CC tissues ( Figure 7A). Importantly, mRNA and protein expression levels of ZIC2 were declined in HeLa and SW756 cells after silencing of circ_0000069, which was weakened by co-transfection with si-circ_0000069 and anti-miR-4429 in HeLa and SW756 ( Figure 7B-7E). We could conclude that ZIC2 was regulated by circ_0000069/miR-4429 axis in CC cells.

Knockdown of circ_0000069 inhibited CC tumor growth in vivo
The animal experiment was performed to explore the role of circ_0000069 in tumorigenesis in vivo. As presented in Figure 8A-8B, the silencing of circ_0000069 significantly suppressed the tumor growth rate and tumor weight when compared with sh-NC group. Besides, circ_0000069 and ZIC2 mRNA were decreased while miR-4429 was increased in sh-circ_0000069 group than sh-NC group ( Figure 8C). Consistently, western blot assay revealed that the protein expression of ZIC2 was downregulated after silencing of circ_0000069 ( Figure 8D). Therefore, the downregulation of circ_0000069 impeded CC tumor growth in vivo by regulating miR-4429 and ZIC2.

Discussion
Collectively, current results indicated that circ_0000069 was significantly upregulated in CC tissues and cells than the negative groups. Furthermore, the CC-associated circRNA expression profile (GSE102686) also revealed that circ_0000069 was upregulated in CC.
Currently, loss-of-function assays revealed that silencing of circ_0000069 could impede proliferation, invasion, and migration, as well as induce cell apoptosis and cell cycle arrest 15 in CC cell, at least in part, by regulating miR-4429/ZIC2 network.
Previous studies indicated that circRNAs served as biomarkers in cancer by functioning oncogenic or tumor-suppressive circRNAs [18]. However, the roles and regulatory mechanism of circ_0000069 in CC remained largely unknown. In this study, patients with high circ_0000069 expression had shorter survival time than CC patients with the low expression of circ_0000069, implying that circ_0000069 might be a promising prognostic biomarker in CC. Additionally, we found that knockdown of circ_0000069 inhibited proliferation, invasion, and migration, as well as induced cell apoptosis and cell cycle arrest at G0/G1 phase in CC. Similar effects of circ_0000069 silencing were confirmed in colorectal cancer patients [7].
Recently, numerous studies have shown that circRNAs acted biological functions by serving as sponges or competing endogenous RNAs (ceRNA) of miRNAs to regulate mRNA expression [19,20]. In this study, we found that circ_0000069 could interact with miR-4429 in CC cells by biotinylated RNA pull-down and dual-luciferase reporter assays, which suggested that miR-4429 was a target gene of circ_0000069. Besides, the silencing of circ_0000069-induced anti-tumor effects were abolished by combination of circ_0000069 knockdown and miR-4429 inhibitor in CC cell.
MiR-4429 was demonstrated to act as an anti-tumor miRNA in many malignancies, including clear cell renal cell carcinoma [10], glioblastoma [21], and gastric cancer [22]. reported that the overexpression of miR-4429 overturned the oncogenic role of LINC00313 in papillary thyroid cancer, revealing the anti-tumor role of miR-4429 [23]. In addition, miR-4429 could increase the radio-sensitivity of CC cell, revealing the potential therapeutic potential of miR-4429 for CC patients [11].
Moreover, overexpression of ZIC2 may be an adverse prognostic factor for oral squamous cell carcinoma patients [24]. Lu et al. also reported that ZIC2 expression was significantly in hepatocellular carcinoma tissues than control, which may result to poor outcomes in patients with hepatocellular carcinoma [25]. Consistently, ZIC2 was upregulated in CC tissues and cells than controls. More importantly, miRNAs were known to play a pivotal role in human diseases by targeting 3'UTR of target mRNA [9]. ZIC2 was reported as a target gene of multiple miRNA in tumor, such as miR-1271-5p [26]. Surely, a miRNA may target many genes in diseases, which may ascribe to the complexity genome and multiplex regulation network of gene expression [27,28]. In this study, we identified ZIC2

Ethics approval and consent to participate
The study was approved by the Ethics Committee of The Affiliated Hospital of Qingdao University and written informed consents were collected from all patients and hospitals.
The animal experiment was permitted by the Animal Research Committee of The Affiliated

Hospital of Qingdao
University and performed in accordance with the guidelines of the National Animal Care and Ethics

Consent for publication
Not applicable

Availability of data and materials
The data sets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no conflicts of interest.