circ_0002346 Suppresses Non-Small-Cell Lung Cancer Progression Depending on the Regulation of the miR-582-3p/STXBP6 Axis

Background Accumulating articles have reported the pivotal regulatory roles of circular RNAs (circRNAs) in non-small-cell lung cancer (NSCLC) tumorigenesis. Here, our purpose was to explore the role of circ_0002346 in NSCLC progression and its associated mechanism. Methods Cell proliferation ability was assessed by a 5-ethynyl-2′-deoxyuridine (EDU) assay and a colony formation assay. Transwell assays were conducted to analyze cell migration and invasion abilities. Cell apoptosis was analyzed by flow cytometry and by using a caspase3 activity assay kit. The glycolysis of NSCLC cells was analyzed using a fluorescence-based glucose/lactate assay kit. A dual-luciferase reporter assay and an RNA pull-down assay were performed to verify the binding relationship between microRNA-582-3p (miR-582-3p) and circ_0002346 or syntaxin-binding protein 6 (STXBP6). Results circ_0002346 level was prominently downregulated in NSCLC tissues and cell lines. circ_0002346 overexpression significantly suppressed the proliferation, migration, invasion, and glycolysis and triggered the apoptosis of NSCLC cells. circ_0002346 directly interacted with miR-582-3p, and circ_0002346 overexpression-mediated antitumor effects in NSCLC cells were partly reversed by miR-582-3p overexpression. miR-582-3p directly interacted with the 3′ untranslated region (3′UTR) of STXBP6, and STXBP6 silencing partly counteracted circ_0002346 overexpression-mediated antitumor influences in NSCLC cells. circ_0002346 can upregulate the expression of STXBP6 by acting as a miR-582-3p sponge in NSCLC cells. circ_0002346 overexpression suppressed xenograft tumor growth in vivo. Conclusion circ_0002346 overexpression suppressed the malignant properties of NSCLC cells by binding to miR-582-3p to induce the expression of STXBP6.


Introduction
Non-small-cell lung cancer (NSCLC) is a predominant type of lung cancer, including adenocarcinoma, squamous cell carcinoma, and large cell carcinoma [1,2]. Due to the lack of effective markers for early diagnosis, most patients are diagnosed with NSCLC at an advanced stage, leading to poor prognosis [3,4]. Therefore, the development of novel and effective diagnostic, therapeutic, and prognostic markers is important in the treatment of NSCLC.
Circular RNAs (circRNAs) are natural noncoding RNAs characterized by a closed circular structure [5,6]. circRNAs are conserved during evolution, with high stability and a tissue-specific expression pattern. These characteristics make circRNAs ideal markers for human diseases [7,8]. cir-cRNAs are widely dysregulated in NSCLC, and accumulating evidence has identified a series of circRNA biomarkers for NSCLC, including circ_0001649 [9], circ_0043278 [10], and circ_0102533 [11]. We investigated the function of a novel circRNA, circ_0002346, in NSCLC progression.
MicroRNAs (miRNAs), as another class of noncoding RNAs, play important regulatory roles in NSCLC progression as oncogenes or tumor suppressor genes [12,13]. cir-cRNAs can indirectly regulate gene expression by sequestering miRNAs, thus releasing the downstream genes from the inhibition of miRNAs [14]. For instance, circ_ 101237 is reported to contribute to NSCLC development by upregulating MAPK1 via sponging miR-490-3p [15]. circ_0018818 is reported to aggravate NSCLC progression by increasing NID1 level via sequestering miR-767-3p [16]. We predicted the downstream miRNA/messenger RNA (mRNA) axis of circ_0002346 using bioinformatics databases, and its working mechanism was tested by rescue experiments.
In the current study, we first analyzed the characteristics and functions of circ_0002346 in NSCLC cells. The downstream miRNA/mRNA axis of circ_0002346 was predicted by bioinformatics analysis, and its working mechanism was tested by rescue experiments. Lastly, the in vivo role of circ_0002346 in the growth of xenograft tumors was evaluated using a xenograft tumor formation assay.

Materials and Methods
2.1. Clinical Specimens. NSCLC tissue specimens (n = 45) and paracancer normal tissue specimens (n = 45) were harvested from NSCLC patients at the Affiliated Xiangshan Hospital of Wenzhou Medical University. All patients were diagnosed by histopathological examination. The clinical tissue specimens were immediately frozen in liquid nitrogen and then stored at -80°C. A written informed consent form had been signed by all the participants, and this clinical study was authorized by the ethics committee of the Affiliated Xiangshan Hospital of Wenzhou Medical University.

Cell
Lines. The human normal lung epithelial cell line (HBE) and two NSCLC cell lines (A549 and H1299) were purchased from the Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China). All cell lines were cultured in Roswell Park Memorial Institute-1640 medium (RPMI-1640, Gibco, Carlsbad, CA, USA) with the addition of 10% fetal bovine serum (FBS, Gibco) and 1% penicillin/streptomycin (Sangon Biotech, Shanghai, China) at 37°C with 5% CO 2 .

Reverse Transcription-Quantitative Polymerase Chain
Reaction (RT-qPCR). RNA samples were isolated with the TRIzol Reagent (Ambion, Austin, TX, USA). For circRNA and mRNA, reverse transcription (RT) was conducted using a complementary DNA (cDNA) kit (Thermo Fisher Scientific, Waltham, MA, USA), and thermal cycling reaction was conducted using SYBR Green reagents (Invitrogen, Carlsbad, CA, USA). The fold changes were analyzed by the 2 −ΔΔCt method and normalized to β-actin. RT-qPCR of miR-582-3p was conducted using its specific Bulge-Loop™ miRNA RT-qPCR primer sets (RiboBio, Guangzhou, China), and the relative expression of miR-582-3p was analyzed by the 2 −ΔΔCt method and normalized to U6. Primer sequences are shown in Table 1. 2.4. RNase R Treatment. RNase R was purchased from Epicentre Technologies (Madison, WI, USA). RNA samples (4 μg) were incubated with 12 U RNase R for 40 min at 37°C. The levels of circ_0002346 and its host gene cysteine rich transmembrane BMP regulator 1 (CRIM1) were examined by RT-qPCR.

Isolating RNAs in Nuclear and Cytoplasmic Fractions.
The subcellular localization of circ_0002346 in NSCLC cells was analyzed by this assay. The nuclear and cytoplasmic fractions were isolated using the PARIS™ Kit (Invitrogen). U6 and 18S rRNA were regarded as positive references for the nucleus and cytoplasm, respectively.
2.8. Colony Formation Assay. Cell proliferation ability was evaluated by a colony formation assay. NSCLC cells were  3 International Journal of Genomics dispersed in culture medium at low density, and cells were seeded onto 12-well plates at 200 cells/well. Cells were incubated for 2 weeks to form visible colonies. The colonies were immobilized with 4% paraformaldehyde (Sigma-Aldrich) and stained with 0.1% crystal violet (Sigma-Aldrich). After washing twice and being air-dried, the number of colonies was analyzed.
2.9. Transwell Assays. A transwell migration or invasion assay was implemented to assess cell migration or invasion capacity using uncoated or Matrigel-(BD Biosciences, San Jose, CA, USA) covered transwell plates, respectively. NSCLC cells were dispersed in serum-free medium, and cells were then seeded onto the above chambers. The below chambers were added with 500 μL medium plus 20% FBS (chemokines). After 24 h incubation, unmigrated or uninvaded NSCLC cells were scraped using a cotton swab, and cells that passed through the membrane were immobilized with 4% paraformaldehyde (Sigma-Aldrich) and stained with 0.5% crystal violet (Sigma-Aldrich). Cell number in five random fields was analyzed under a light microscope (Olympus, Osaka, Japan) at 100x.

Flow
Cytometry. The apoptosis of NSCLC cells was analyzed by flow cytometry. The percentage of apoptosis cells (FITC + /PI +/-) was evaluated using an Annexin Vfluorescein isothiocyanate (FITC)/propidium iodide (PI) kit (Vazyme, Nanjing, China). After transfection for 72 h, NSCLC cells in 400 μL binding buffer were simultaneously stained with 5 μL Annexin V-FITC and 5 μL PI in the dark. Apoptosis cells with FITC + /PI +/were captured by using a FACSCanto II flow cytometer (BD Biosciences), and the apoptosis rate was analyzed.
2.11. Analysis of Caspase3 Activity. The activity of caspase3 was analyzed using a caspase3 activity assay kit (Beyotime, Beijing, China). Briefly, cell extracts were incubated with the assay reagent and its specific substrate Ac-DEVD-pNA for 2 h. The absorbance was determined at the spectrum of 405 nm. The activity of caspase3 was assessed through drawing the standard curve.

Analysis of Glucose Uptake and Lactate Production.
A fluorescence-based glucose/lactate assay kit (BioVision, Milpitas, California, USA) was used to analyze glucose uptake and lactate production according to the manufacturer's instructions.
The partial fragment of circ_0002346 or STXBP6 was subcloned to the luciferase vector pmirGLO (Promega, Madison, WI, USA), and the constructed plasmids were termed as circ_0002346 WT/MUT and STXBP6-3 ′  (e, f) The RNase R resistance of circ_0002346 and its host gene CRIM1 mRNA was analyzed by RT-qPCR. (g, h) The subcellular localization of circ_0002346 was analyzed, and 18S rRNA and U6 served as cytoplasmic and nuclear markers, respectively. * * * P < 0:001.    International Journal of Genomics untranslated region (3 ′ UTR) WT/MUT. After cotransfecting these luciferase plasmids and oligonucleotides into NSCLC cells for 24 h, the luciferase intensities were examined via a commercial dual-luciferase reporter assay system kit (Promega).

RNA Pull-Down
Assay. miR-582-3p was biotinylated to construct Bio-miR-582-3p, and Bio-miR-NC was regarded as the control. Cell extracts (3 μg) were incubated with 100 pmol Bio-miR-582-3p or Bio-miR-NC. The streptavidin agarose beads were then added to the tubes. RNA levels were examined by RT-qPCR. 2.17. Statistical Analysis. Statistical analysis was conducted using GraphPad Prism 7.0 software (GraphPad, La Jolla, CA, USA), and the data were represented as mean ± standard deviation (SD). The differences were evaluated by Student's t-test and one-way analysis of variance (ANOVA) followed by Tukey's test. Differences were considered to be statistically significant at P < 0:05.

circ_0002346 Expression Is Downregulated in NSCLC
Tissues and Cell Lines. We first analyzed the expression pattern of circ_0002346 in NSCLC tissues and cell lines. circ_ 0002346 expression was markedly reduced in NSCLC tissues and cell lines compared with adjacent healthy tissues and the HBE cell line (Figures 1(a) and 1(b)). circ_0002346 was generated from the back-splicing of exon 2 (174 nt), 3 (243 nt), and 4 (121 nt) of its host gene CRIM1 (Figure 1(c)). To confirm the circular structure of circ_0002346, divergent and convergent primers were used to amplify circ_0002346 and its linear counterpart CRIM1, respectively. The results revealed that circ_0002346 can be amplified in the cDNA group when using divergent primers, not in the gDNA group, suggesting that circ_0002346 was indeed a circular transcript (Figure 1(d)). circRNAs are resistant to RNase R due to their covalently closed continuous loop structure. circ_0002346 was resistant to RNase R degradation (Figures 1(e) and 1(f)), which further demonstrated that circ_0002346 was a circular transcript. The functions of circRNAs are closely associated with their subcellular localization. We found that circ_0002346 was majorly localized in the cytoplasmic fraction of NSCLC cells (Figures 1(g) and 1(h)), suggesting that circ_0002346 might regulate gene expression at the posttranscriptional level. Taken together, circ_0002346 was notably downregulated in NSCLC. 7 International Journal of Genomics into NSCLC cells. circ_0002346 plasmid was effective in upregulating circ_0002346 level in NSCLC cells (Figure 2(a)). Cell proliferation ability was assessed by an EDU assay and a colony formation assay. circ_0002346 overexpression reduced the incorporation of EDU in NSCLC cells (Figure 2(b)), suggesting that circ_0002346 overexpression suppressed the proliferation of NSCLC cells. The number of colonies was decreased following the overexpression of circ_0002346 (Figure 2(c)), which further demonstrated that circ_0002346 overexpression restrained the proliferation ability of NSCLC cells. The numbers of migrated and invaded cells were both reduced after the  (Figure 2(f)). Also, the activity of caspase3 was markedly increased following the overexpression of circ_0002346 (Figure 2(g)). circ_0002346 overexpression elevated the levels of proapoptotic proteins (Bax and C-casp3) (Figure 2(h)). These results suggested that circ_ 0002346 overexpression induced the apoptosis of NSCLC cells. circ_0002346 overexpression reduced the uptake of glucose and the production of lactate (Figures 2(i) and 2(j)). Moreover, circ_0002346 accumulation reduced the levels of two glycolysis-associated rate-limiting enzymes (HK2 and PKM2) in NSCLC cells (Figure 2(k)). These results suggested that circ_0002346 overexpression restrained cell glycolytic metabolism. Taken together, circ_ 0002346 played a tumor suppressor role by inhibiting the malignant behaviors of NSCLC cells.

circ_0002346 Directly Binds to miR-582-3p.
Accumulating evidence has demonstrated that circRNAs can function as oncogenes or tumor suppressors by sponging miRNAs [6,17]. Through bioinformatics analysis using the Circinteractome database, miR-582-3p was predicted as a target of circ_0002346. The putative binding sites between circ_ 0002346 and miR-582-3p are shown in Figure 3(a). miR-582-3p expression was notably upregulated in NSCLC tissues and cell lines relative to that in adjacent normal tissues and the HBE cell line (Figures 3(b) and 3(c)). Subsequently, a dual-luciferase reporter assay and an RNA pull-down assay were conducted to confirm the interaction between circ_ 0002346 and miR-582-3p. The dual-luciferase reporter assay showed that miR-582-3p overexpression can markedly reduce the luciferase activity driven by the wild-type plasmid (circ_0002346 WT) instead of the mutant plasmid (circ_ 0002346 MUT) (Figures 3(d) and 3(e)), suggesting that circ_0002346 directly interacted with miR-582-3p via the predicted sites. circ_0002346 was pulled down when using a Bio-miR-582-3p probe compared with that in the Bio-miR-NC group (Figures 3(f) and 3(g)), demonstrating the binding relationship between circ_0002346 and miR-582-3p. Taken together, miR-582-3p was a direct target of circ_ 0002346 in NSCLC cells.

circ_0002346 Overexpression-Induced Antitumor Impacts Are Partly Reversed by the Addition of miR-582-3p
Mimics in NSCLC Cells. An RT-qPCR assay confirmed the high transfection efficiency of miR-582-3p mimics in NSCLC cells (Figure 4(a)). We wondered whether circ_ 0002346 overexpression-mediated antitumor effects were partly based on its sponge activity for miR-582-3p, and rescue experiments were performed by transfecting NSCLC cells with circ_0002346 alone or together with miR-582-3p mimics. An EDU assay and a colony formation assay together demonstrated that the addition of miR-582-3p mimics partly rescued the proliferation ability of NSCLC cells (Figures 4(b) and 4(c)). The introduction of miR-582-3p also largely recovered the migration and invasion abilities of NSCLC cells (Figures 4(d) and 4(e)). miR-582-3p overexpression attenuated circ_0002346 overexpression-induced apoptosis in NSCLC cells (Figures 4(f)-4(h)), evidenced by the reduced apoptosis rate, the activity of caspase3, and the protein levels of Bax and C-casp3. circ_0002346 overexpression restrained cell glycolytic metabolism, and the introduction of miR-582-3p mimics largely rescued the glycolytic rate of NSCLC cells (Figures 4(i)-4(k)), verified by the increased rates of glucose consumption and lactate  13 International Journal of Genomics production and the protein levels of HK2 and PKM2. Taken together, circ_0002346 overexpression suppressed NSCLC progression partly by downregulating miR-582-3p.
3.5. STXBP6 Is a Target of miR-582-3p. miRNAs can bind to the 3′UTR of mRNAs to induce the degradation or translational repression of target mRNAs [18]. Bioinformatics analysis using the TargetScan database showed that the 3′UTR of STXBP6 harbored the complementary sites with miR-582-3p ( Figure 5(a)). The mRNA and protein expression of STXBP6 was significantly reduced in NSCLC tissues compared with normal tissues (Figures 5(b) and 5(c)). Compared with the HBE cell line, the mRNA and protein levels of STXBP6 were reduced in NSCLC cell lines (Figures 5(d) and 5(e)). The luciferase activity of the wild-type plasmid (STXBP6-3 ′ UTR WT) was markedly reduced by the overexpression of miR-582-3p (Figures 5(f) and 5(g)). However, the luciferase activity of the mutant plasmid (STXBP6-3 ′ UTR MUT) remained unchanged upon the transfection of miR-NC or miR-582-3p (Figures 5(f) and 5(g)), suggesting that miR-5582-3p interacted with the 3 ′ UTR of STXBP6 via the predicted sites. The overexpression of miR-582-3p significantly reduced STXBP6 protein level in NSCLC cells (Figures 5(h) and 5(i)). circ_0002346 overexpression notably increased the protein expression of STXBP6, and the introduction of miR-582-3p mimics led to a prominent reduction in the protein expression of STXBP6 in NSCLC cells (Figures 5(j) and 5(k)). Taken together, these results indicated that STXBP6 was a direct target of miR-582-3p, and it was regulated by the circ_0002346/miR-582-3p axis.

STXBP6 Silencing Partly Offsets circ_0002346
Overexpression-Induced Effects in NSCLC Cells. RT-qPCR showed that the transfection efficiency of si-STXBP6 was high in NSCLC cells (Figure 6(a)). To explore the functional association between circ_0002346 and STXBP6, we conducted rescue experiments by transfecting NSCLC cells with the circ_0002346 plasmid alone or together with si-STXBP6. The addition of si-STXBP6 largely rescued the proliferation ability of NSCLC cells (Figures 6(b) and 6(c)), evidenced by the increased ratio of EDU incorporation and colony number. The introduction of si-STXBP6 also largely rescued the migration and invasion abilities of NSCLC cells (Figures 6(d) and 6(e)). STXBP6 silencing attenuated circ_ 0002346 overexpression-induced apoptosis in NSCLC cells (Figure 6(f)). Furthermore, the introduction of si-STXBP6 reduced the activity of caspase3 and decreased the protein levels of Bax and C-casp3 in circ_0002346-overexpressed NSCLC cells (Figures 6(g) and 6(h)). circ_0002346 overexpression-induced suppressive effects on the glucose consumption and lactate production were largely reversed by the addition of si-STXBP6 (Figures 6(i) and 6(j)). The addition of si-STXBP6 largely rescued the protein levels of HK2 and PKM2 in circ_0002346-overexpressed NSCLC cells (Figure 6(k)). Taken together, circ_0002346 overexpression suppressed the malignant behaviors of NSCLC cells partly by upregulating STXBP6.

16
International Journal of Genomics tumor formation assay in vivo. The results indicated that the average volume of xenograft tumors in the lenti-circ_ 0002346 group was notably smaller than that in the lenti-NC group (Figure 7(a)). Meanwhile, tumor weight was also reduced in the lenti-circ_0002346 group relative to the lenti-NC group (Figure 7(b)). The results of an IHC assay suggested that the staining intensity of Ki-67 was notably decreased in circ_0002346-overexpressed tumor tissue compared with the lenti-NC group (Figure 7(c)). In addition, we found that the staining intensity of STXBP6 was markedly increased in circ_0002346-overexpressed tumor tissue in comparison with the lenti-NC group (Figure 7(c)). Moreover, we measured the expression of the circ_ 0002346/miR-582-3p/STXBP6 axis in tumor tissues. The expression of circ_0002346 and STXBP6 was significantly upregulated in circ_0002346-overexpressed tissues, and miR-582-3p level was reduced in the circ_0002346-overexpressed group (Figures 7(d)-7(f)). Taken together, circ_ 0002346 overexpression restrained tumor growth in vivo.

Discussion
Owing to the rapid progression of sequencing technology and genomic research, more and more circRNAs have been identified and studied [19,20]. Unlike linear noncoding RNAs, such as miRNAs, circRNAs are stable circular noncoding RNAs that are resistant to exonuclease digestion [21]. Previous evidence has demonstrated that circRNAs are implicated in NSCLC progression. For example, circ_ 0000376 is reported to contribute to the proliferation, motility, and chemoresistance of NSCLC cells by sequestering miR-384 [22]. circ_0002483 is reported to increase the Taxol sensitivity of NSCLC cells by absorbing miR-182-5p [23]. Wang et al. found that circ_0002346 expression is decreased in lung adenocarcinoma, and it restrains the invasion and motility of lung adenocarcinoma cells by targeting miR-182 and miR-93 [24]. Consistent with a former study, we found that circ_0002346 level was markedly reduced in NSCLC tissue specimens and cell lines. circ_0002346 overexpression restrained the proliferation, migration, invasion, and glycolysis and facilitated the apoptosis of NSCLC cells.
Accumulating articles pointed out that circRNAs can regulate gene expression and multiple cellular biological processes by absorbing miRNAs [17,25]. Moreover, one circRNA can absorb many diverse miRNAs. For instance, circ_0008532 is reported to contribute to bladder cancer development by upregulating MTGR1 via sponging miR-155-5p and miR-330-5p [26]. circ-PRMT5 is reported to aggravate gastric cancer development by absorbing miR-145 and miR-1304 [27]. We found that miR-582-3p was a novel target of circ_0002346 in addition to miR-182 and miR-93 [24]. miR-582-3p level was elevated in NSCLC. Fang et al. reported that miR-582-3p maintains the malignant properties of lung cancer stem cells by activating the Wnt/β-