Circ_0072088 Promotes Proliferation, Migration, and Invasion of Esophageal Squamous Cell Cancer by Absorbing miR-377

Circular RNA (circRNA) is an endogenous noncoding RNA. Accumulative investigations have confirmed that circRNAs play a vital role in carcinogenesis and tumor progression. Herein, we examined the expression and mechanism of circ_0072088 in esophageal squamous cell carcinoma (ESCC). As a result, circ_0072088 was significantly overexpressed in ESCC tissues and cells, which was closely associated with tumor size, invasion depth, clinical stage, and lymph node metastasis of esophageal cancer. Nuclear and cytoplasmic separation as well as FISH assays showed that circ_0072088 was mainly localized in the cytoplasm of ESCC cells. RNase R treatment assay revealed that circ_0072088 was steadier than linear ZFR mRNA. circ_0072088 promoted ESCC cell proliferation, migration and invasion in vitro, and cell proliferation in vivo. Mechanistically, circ_0072088 upregulated VEGF gene expression by acting as the sponge of miRNA-377. In conclusion, circ_0072088 might be used as a diagnostic biomarker and therapeutic target for ESCC.


Introduction
Esophageal cancer (EC) is one of the most common malignancies worldwide, ranking the seventh in terms of incidence and the sixth in terms of cancer-related death among the 10 most common malignancies globally [1]. EC is mainly divided into two major pathological categories: esophageal squamous cell carcinoma (ESCC) and esophageal adenocarcinoma (EAC), with ESCC accounting for as high as 90% [2]. Due to a lack of typical symptoms in the early stage in most EC patients, the diagnosis rate in the early stage is low, while the majority of EC patients are already in an advanced stage at diagnosis [3]. e 5-year survival rate of EC is only 29.7% in China [4]. us, it is an urgent need to investigate the pathogenesis of EC at the molecular level and to explore molecular markers as well as therapeutic targets for early diagnosis of EC.
Circular RNAs (circRNAs) are a novel type of noncoding RNAs existing in eukaryotes with abundant expression [5]. According to sources, circRNAs consist of exon circularization, intron circularization, and cocirculation of exons and introns, with exon circRNAs as the most common [6]. Unlike linear RNA, circRNAs have a covalently closed loop structure without 5′ to 3′ polarity, making it more resistant to RNase R, which also makes circRNA more stable than its linear counterparts in cells [7]. circRNAs are highly conserved among different species and specifically expressed in tissues [8]. A recent study has disclosed that the differential expressions of circRNAs affect the occurrence and progression of carcinomas [9][10][11][12], including the proliferation, migration, and invasion of EC [13,14]. However, there are relatively limited studies concerning the effects of circRNA on ESCC overall. Based on the differential expression of circRNA chip from 10 pairs of ESCC and matched adjacent tissues in our previous study [14], in this study, the top ten circRNAs with elevated expression were selected and further validated in cancer tissues and matched adjacent tissues of another 25 ESCC patients, revealing circ_0072088 with the most significant differential expression. circ_0072088 expression was elevated in ESCC tissues and cells and was positively correlated with malignant phenotypes both in vitro and in vivo.
erefore, circ_0072088 may importantly involve in the development of ESCC.

Collection of Specimens from ESCC Patients.
Surgical samples were collected from 83 patients with pathologically diagnosed ESCC undergoing surgical resection in the Affiliated People's Hospital of Jiangsu University from January 2014 to December 2018. None of them underwent chemotherapy, radiotherapy, or other treatments before the surgical operation. Samples were collected within approximately 5 min after tumor resection and stored in liquid nitrogen. e study was gained approval from the Medical Ethics Committee of the Affiliated People's Hospital of Jiangsu University, and written informed consent was signed by all subjects.

Quantitative Real-Time PCR (qRT-PCR).
Total RNA was extracted from cells and tissues by Trizol reagent (Invitrogen, Shanghai, China) and GeneJET RNA Purification Kit (Invitrogen), respectively. RT SuperMix Reverse Transcription Kit (Vazyme Biotech, Nanjing, China) was employed for reverse transcription from RNA into cDNA, followed by qRT-PCR using SYBR Premix Ex Taq II Kit (Vazyme Biotech). e relative expression was calculated by the 2 −ΔΔCT method, GAPDH was utilized to normalize the relative expression levels of circRNA and linear mRNA, and U6 was utilized to normalize the expression level of microRNA.
e primer sequences were listed in Supplementary Table S1.
e sequences of sh-circ_0072088, si-ZFR, miR-377 mimic/NC, and miR-377 inhibitor/NC were shown in Supplementary Table S2. 2.5. Cell Proliferation Assay. Cell proliferation ability was examined by Cell Counting Kit-8 (CCK-8) as well as colony formation assay. In the CCK-8 assay, ESCC cells were inoculated into the 96-well plate (2000 cells/per well), and CCK-8 solution (Dojindo, Kumamoto, Japan) was added to each well at 24, 48, 72, and 96 h separately. Afterward, cells at each time point were incubated at 37°C for an additional 2 h, followed by measurement of absorbance at a wavelength at 450 nm. In the colony formation assay, 500 ESCC cells were inoculated in each well of the 6-well plates. After culture for 14 days, cells were then fixed, stained, photographed, and counted.
2.6. Transwell Assays. 2 × 10 4 /100 μl of ESCC cells in serumfree medium were added to the top Transwell chambers (8 µm, pore size, Corning, NY, USA) with or without Matrigel (BD Biosciences, USA). In addition, the lower chambers were added with a medium containing 10% FBS. After 24 h, cells passing through the membrane were fixed, stained, photographed, and counted.

RNase R Treatment
Assay. 2 µg RNA and 6-unit RNase R (Geneseed Biotech, Guangzhou, China) were mixed and incubated at 37°C for 20 min. qRT-PCR for determining the mRNA level of circ_0072088 and linear ZFR before and after RNase R treatment.

Nuclear and Cytoplasmic Separation
Assay. PARIS kit (Invitrogen, Shanghai, China) was purchased to extract nuclear and cytoplasmic RNA in line with the standard protocol. Subsequently, the relative expression of circ_0072088 in the nucleus and cytoplasm was measured by qRT-PCR.

RNA Fluorescence In Situ Hybridization (FISH).
e cy3labeled circ_0072088 probe was purchased from Gene-Pharma. A FISH Kit was purchased (Ribobio, Guangzhou, China) and performed accordingly. Afterward, Zeiss AIM software along with a Zeiss LSM 700 confocal microscope system (Carl Zeiss Jena, Oberkochen, Germany) was utilized to capture confocal images of cells.

RIP Assay.
EZMagna RIP kit (Merck, Darmstadt, Germany) was utilized. In brief, ESCC cells were lysed with RIP lysis buffer. e lysate was incubated with magnetic beads that had bounded with anti-Argonaute 2 (Ago2) or anti-IgG antibody at 4°C for 6 h. e magnetic beads-bound RNA was extracted, purified, and detected by qRT-PCR. 2 Journal of Oncology

Western Blot
Analysis. RIPA lysis buffer ( ermo Scientific, USA) was utilized for protein extraction from ESCC cells after transfection for 48 h, followed by protein quantification using a BCA kit (Beyotime Biotechnology, Nantong, China). Afterward, 20 µg of protein sample was separated by electrophoresis, transferred to the PVDF membrane, and blocked in the blocking solution at 37°C on a shaker for 1 h. Afterward, membranes were reacted with primary antibody at 4°C overnight, reacted with proper secondary antibody at 37°C for 1 h, visualized by chemiluminescence (ECL) (YEASEN, Shanghai, China), and processed and analyzed by a gel imager. Tubulin was used as the internal control in the experiment. e antibodies used in Western blot were as follows: VEGF (1 : 1000, YEASEN), Tubulin (1 : 1000, Beyotime Biotechnology), and HRP-conjugated secondary antibody (1 : 1000, Beyotime Biotechnology). e assay was performed in triplicate.

Xenograft Mouse Model.
e animal experiment procedure gained approval from the Animal Care and Use Committee of Jiangsu University. Six-week-old BALB/c nude mice were purchased from GemPharmatech (Nanjing, China) and housed in the SPF room in the Animal Experimental Center of Jiangsu University. Twelve mice were randomly divided into the sh-circ_0072088 group and the control group (N � 6 each). TE13 cells with stable expression of sh-circ_0072088 or sh-control were injected subcutaneously into the right axilla of mice (4 × 10 6 /each mouse). After seven days, the tumor volume of nude mice was measured once a week to plot the tumor growth curve. After 28 days, mice were anesthetized and sacrificed by cervical dislocation. e tumor mass was isolated and subjected to assays of immunohistochemistry (IHC).
2.14. Statistical Analysis. SPSS 21.0 software (IBM, Chicago, USA) was employed for statistical analysis. Experimental data were shown as mean ± standard deviation. e t-test was used for pairwise comparison, and Pearson's Chi-square test was utilized to analyze the correlation between circ_0072088 expression and clinical characteristics. A P < 0.05 indicated statistical significance.

circ_0072088 Expression Was Increased in ESCC Tissues
and Cells. Based on the previous circRNA microarray analysis of ESCC and matched adjacent tissues (N � 10), the top ten circRNAs with elevated expression were validated by qRT-PCR in another 25 cases of ESCC cancer tissues and matched adjacent tissues, which revealed circ_0072088 as the most significantly differentially expressed circRNAs (Figure 1(a)). Afterward, qRT-PCR was adopted to determine the expression of circ_0072088 in cancer tissues and paired adjacent tissues of another 58 ESCC patients, which confirmed the significantly increased circ_0072088 expression in ESCC tissue than in adjacent tissue (Figure 1(b)). We further analyzed the correlation of circ_0072088 expression with clinicopathological characteristics, demonstrating that the expression of circ_0072088 was significantly correlated with tumor size, invasion depth, TNM stage, and LNM (Table 1). qRT-PCR also revealed the significantly elevated circ_0072088 expression in different ESCC cell lines compared to the normal esophageal epithelial cells HET-1A (Figure 1(c)). Meanwhile, linear ZFR expression was significantly elevated in ESCC tissues and cells compared to relative controls (Figures 1(d) and 1(e)). Afterward, three si-ZFRs were designed and transfected into TE-13 cells. As a result, si-ZFR not only successfully knocked down the expression of linear ZFR but also reduced circ_0072088 expression in ESCC cells (Figures 1(f ) and 1(g)). ese data suggested that the overexpression of circ_0072088 may associate with the malignant phenotypes of ESCC.

3.2.
Biological Characteristics of circ_0072088. circ_0072088 is located on chromosome 5 and is circularized by exons 13-17 on the linear ZFR mRNA, with 693 bases in length (Figure 2(a)). e circularization sites in circ_0072088 were confirmed by gene amplification and base sequencing using back-to-back primers (Figure 2(b)). RNase R treatment in the total RNA extracted from TE-13 cells and subsequent qRT-PCR demonstrated that the mRNA expression of linear ZFR was reduced after treatment, while the expression of circ_0072088 was not significantly declined, confirming that circ_0072088 was steadier than linear ZFR mRNA ( Figure 2(c)). Nuclear and cytoplasmic separation and FISH assays showed that circ_0072088 was mostly located in the cytoplasm of ESCC cells (Figures 2(d) and 2(e)).

circ_0072088 Promotes the Proliferation, Migration, and Invasion of ESCC Cells In Vitro.
To further examine the roles of circ_0072088 on ESCC cells, lentiviral sh-circ_0072088 was designed and transfected into TE-13 cells. qRT-PCR further showed that sh-circ_0072088 transfection could significantly decrease circ_0072088 expression in TE-13 cells (Figure 3(a)), which was more obvious in the cytoplasm than in the nucleus (Figures 3(b) and 3(c)). However, knocking down circ_0072088 did not affect the expression of linear ZFR (Figure 3(d)). CCK-8 and colony formation assays showed that the proliferation capacity of TE-13 cells was significantly decreased after knocking down circ_0072088 (Figures 3(e) and 3(f )). Transwell assay revealed that the migration and invasion abilities of TE-13 cells were significantly decreased after the downregulation of circ_0072088 (Figure 3(g)).
e biological effects of circ_0072088 on malignant phenotypes of ESCC were also verified in ECA109 cells. Transfection of circ_0072088 overexpression plasmid into ECA109 cells could significantly increase the expression level of circ_0072088, which did not affect the expression of linear ZFR (Figures 3(h) and 3(i)). In addition, circ_0072088 overexpression could enhance the proliferation, migration, and invasion of ESCC cells (Figures 3(j)-3(l)).

circ_0072088 Acts as a Sponge for miRNA-377.
Because circ_0072088 was an exon circRNA mainly located in the cytoplasm, it may act as the sponge to absorb miRNAs to regulate its downstream target genes, thereby influencing phenotypes of ESCC cells. By searching and analyzing the circular RNA interactome and starBase databases, we found that nine miRNAs had the binding sites for circ_0072088, including miRNA-223, miRNA-330-3p, miRNA-377, miRNA-532-3p, miRNA-545, miRNA-616, miRNA-1270, miRNA-620, and miRNA-624 (Figure 4(a)). e qRT-PCR assay showed that miRNA-377 expression was significantly upregulated after knocking down circ_0072088 in TE-13 cells, while the expression of other miRNAs was not significantly changed (Figure 4(b)). Similar data were received from ECA109 cells. After the upregulation of circ_0072088 in ECA109 cells, the expression of miRNA-377 was significantly downregulated, while the expression of other miRNAs was not significantly changed (Figure 4(c)). ese results suggested that circ_0072088 could regulate the expression of miRNA-377. e binding site of miRNA-377 to circ_0072088 was then predicted by TargetScan (Figure 4(d)). RIP assay was performed to examine whether miR-377 could bind to circ_0072088 by means of Argonaute protein.
circ_0072088-wt/circ_0072088-mut plasmid and miRNA-377 mimic or miRNA-NC mimic were cotransfected into TE-13 cells, revealing that miRNA-377 could only decrease the luciferase activity of the circ_0072088-wt plasmid, but not circ_0072088-mut plasmid (Figure 4(h)). ese results confirmed that circ_0072088 could act as a sponge to absorb miRNA-377 in ESCC cells.

circ_0072088 Acts as a Sponge for miRNA-377 to Target
VEGF. e prediction data showed that VEGF is one of the potential binding targets of miRNA-377. Western blot and qRT-PCR demonstrated that the downregulation of miRNA-377 in ECA109 cells could significantly increase both mRNA and protein expression levels of VEGF (Figures 5(a) and 5(b)), while the upregulation of miRNA-377 in ECA109 cells could significantly inhibit both mRNA and protein expression levels of VEGF (Figures 5(c) and 5(d)), indicating that VEGF was a direct target gene of miRNA-377. Western blot showed that circ_0072088 downregulation could significantly decrease the protein level of VEGF ( Figure 5(e)), while circ_0072088 upregulation could significantly increase the protein level of VEGF ( Figure 5(f )). Meanwhile, Western blot revealed that the overexpression of miRNA-377 could reverse the upregulatory effects of circ_0072088 on VEGF protein ( Figure 5(f )). Transwell assay showed that the overexpression of miRNA-377 could reverse the promoting effects of circ_0072088 on migration in ECA109 cells ( Figure 5(g)). e above findings indicated that circ_0072088 could promote the migration of ESCC cells through upregulating VEGF by acting as a sponge for miRNA-377.

circ_0072088 Promotes ESCC Cell Proliferation In Vivo.
To confirm the functions of circ_0072088 on ESCC cell proliferation in vivo, sh-circ_0072088 TE-13 cells and shcontrol TE-13 cells were subcutaneously injected into the right axilla of 6-week-old nude mice, respectively (4 × 10 6 / each mouse). Tumor volume was determined every week, and mice were sacrificed 28 days after subcutaneous injection. As a result, tumor growth was significantly lower in the sh-circ_0072088 group than in the control group. After 28 days, tumor volume was significantly smaller in the sh-circ_0072088 group than in the control group (Figures 6(a)-6(c)). IHC showed significantly lower VEGF expression of tumor tissue in the sh-circ_0072088 group than in the shcontrol group (Figure 6(d)). qRT-PCR indicated that VEGF expression in tumor tissue was significantly lower in the sh-circ_0072088 group than in the sh-control group (Figure 6(e)).

Discussion
circRNA, as a novel endogenous RNA, has recently become a research focus. Accumulative studies have shown that circRNAs not only play roles in carcinogenesis and tumor progression but also might be novel biomarkers for early diagnosis as well as therapeutic targets of tumors [15]. e expression profiles of our previous circRNA chip from 10 pairs of ESCC and matched adjacent tissues were validated, which further identified circ_0072088 as the most significantly differentially expressed circRNA. In this study, we further confirmed that the expressions of circ_0072088 and linear ZFR were increased in ESCC tissues and cells. si-ZFR transfection in ESCC cells could decrease linear ZFR expression and simultaneously reduce circ_0072088 expression, indicating that the elevated expression  Journal of Oncology circ_0072088 in ESCC tissues and cells was partially due to the increased expression of linear ZFR. In addition, the expression of circ_0072088 in ESCC was positively related to tumor size, invasion depth, clinical stage, and LNM. Large tumor size, deep invasion, advanced clinical stage, and LNM in ESCC patients suggest poor prognosis [16][17][18][19]; therefore, circ_0072088 might be adopted as a diagnostic and prognostic predictor of ESCC.
circRNAs are enriched in circulating exosomes, plasma, and cells, which are more stable than linear RNAs [15,20]. In this study, RNase R treatment was performed, which confirmed that circ_0072088 was steadier than linear ZFR mRNA. However, it requires more in-depth investigations on whether the plasma level of circ_0072088 could be used as a diagnostic and prognostic biomarker for ESCC.
Because high expression of circ_0072088 was related to the aggressive phenotype of ESCC, we hypothesized that circ_0072088 was likely to be involved in tumor biology. To begin with, in vitro functional assays validated that high expression of circ_0072088 promoted the proliferation, migration, and invasion of ESCC cells, which was inhibited by circ_0072088 downregulation. Consistently, circ_0072088 downregulation could suppress ESCC cell proliferation in vivo.
ZFR is located at 5p13.3 and with 20 exons and 97416 in length, which encodes a zinc finger protein. circ_0072088 is circularized from linear ZFR and mainly exists in the cytoplasm as an exon circRNA [21]. In addition, nuclear and cytoplasmic separation assay and FISH assay also confirmed that circ_0072088 was mainly located in the cytoplasm, indicating that it was able to posttranscriptionally regulate miRNAs. e functions of circRNAs include miRNA sponges, regulation of alternative splicing and gene transcription, and protein decoys [5,22]. Among them, the sponge absorption function of circRNA has been most studied. circRNAs can function as the miRNA sponges to competitively absorb miRNAs to attenuate the inhibitory effects of miRNAs on their target genes, thereby affecting tumorigenesis and tumor progression [23][24][25]. By analyzing and predicting circ_0072088-related miRNAs, miRNA-377 was screened by interfering with the expression level of circ_0072088 in ESCC cells and miR-377 contains seven paired nucleotides with circ_0072088. MiRNAs can form the RNA-induced silencing complex (RISC) with Argonaute protein to silence or degrade the target mRNA through base pairing [26]. In this study, the RIP experiment indicated that the relative enrichment of circ_0072088 and miR-377 in the Ago2 coprecipitation group was remarkably increased relative to that in the IgG immunoprecipitation group, followed by dual-luciferase reporter assay to confirm the direct interaction between circ_0072088 and miRNA-377.
miRNA-377 has been revealed as a tumor suppressor gene to regulate various phenotypes in tumors [27,28]. e expression of miR-377 is significantly lower in the ESCC tissue than in the adjacent tissue, and the miR-377 expression level is negatively related to the clinical stage as well as distant metastasis but is positively related to the survival rate of patients [29]. e growth and metastasis of invasive tumors are closely associated with angiogenesis [30]. VEGF, the most important regulatory factor in angiogenesis, is a mitogenic factor to promote the proliferation of endothelial cells and angiogenesis, which is continuously overexpressed throughout the tumor life * (k) Cell colony formation assay was utilized for examining ECA109 cell proliferation after circ_0072088 overexpression. (l) Transwell assay was adopted for assessing the roles of circ_0072088 overexpression on migration as well as invasion of ECA109 cells. * P < 0.05, * * P < 0.01, and * * * P < 0.001.     Figure 5: circ_0072088 acts as a sponge for miRNA-377 to target VEGF to further promote the migration of ESCC cells. (a-d) qRT-PCR and Western blot were employed for examining the effects of downregulation and upregulation of miRNA-377 on the VEGF gene and protein expression. (e) e regulatory role of circ_0072088 downexpression on the VEGF protein level was examined by Western blot. (f ) Western blot was adopted for determining the regulatory effect of circ_0072088 overexpression on the VEGF protein level and the regulatory role of miRNA-377 mimic on the upregulation of VEGF induced by circ_0072088 overexpression. (g) Transwell assay was conducted for examining the functions of miRNA-377 overexpression on the migration ability of ECA109 cells with circ_0072088 overexpression. * P < 0.05, * * P < 0.01, and * * * P < 0.001. cycle [31,32]. VEGF is widely expressed in multiple types of malignancies, including lung cancer, EC, head and neck squamous cell carcinoma, colorectal cancer, breast cancer, and ovarian cancer, and the high expression of VEGF is positively related to LNM and tumor angiogenesis [29,[31][32][33][34]. VEGF can be regulated by a variety of miRNAs. For example, miR-655 inhibits the proliferation and invasion of ovarian cancer cells by directly targeting VEGF [35]. Low expression of miR-125a inhibits the proliferation and metastasis of hepatocellular carcinoma cells by targeting VEGF [36]. Li et al. have reported that miR-377 can bind to the 3′-UTR of VEGF and play a negative regulatory role in EC [31].
In this study, we have also confirmed that miR-377 negatively regulates VEGF in ESCC by Western blot and simultaneously validated the mechanism of the regulatory role of circ_0072088 on VEGF expression in ESCC. In other words, circ_0072088 can act as a sponge for miR-377 to attenuate the inhibitory effect of miR-377 on VEGF expression, which could elevate the expression of VEGF, thereby promoting proliferation, migration, and invasion of ESCC. ese findings sufficiently illustrate that the circ_0072088-miR-377-VEGF signaling axis is vitally involved in regulating ESCC ( Figure 6(f)).

Conclusion
We demonstrated the elevated expression of circ_0072088 in ESCC tissues and cells, which was positively related to tumor size, invasion depth, clinical stage, and LNM in ESCC patients. Functionally, circ_0072088 could enhance proliferation, migration, and invasion of ESCC cells in vitro, and circ_0072088 downregulation was capable of suppressing ESCC cell proliferation in vivo. circ_0072088 could regulate the expression of VEGF to promote proliferation, migration, and invasion of ESCC by acting as a sponge for miR-377. Collectively, circ_0072088 might be used as a new molecular marker for early diagnosis and therapeutic target for ESCC. * P < 0.05, * * P < 0.01.

Data Availability
All data and materials are available upon request from the corresponding author.

Conflicts of Interest
e authors declare no conflicts of interest.

Authors' Contributions
NF, JZ, and YQS designed the research and drafted the paper with the input of all authors. NF and YJS conducted the experiments, analyzed the results, and edited the paper. YF provided circRNA chip data. TL was responsible for collecting specimens. All authors read and approved the final version.