Circ_0017956 promotes the proliferation and metastasis of non-small cell lung cancer through regulating miR-515-5p/ITGB8 axis

ABSTRACT

Circular RNAs (circRNAs) have been reported to exert vital roles in the tumorigenesis of non-small cell lung cancer (NSCLC). The study aimed to probe the function of circ_0017956 in NSCLC development. The expression of circ_0017956, microRNA (miR)-515-5p and integrin subunit beta 8 (ITGB8) was gauged by quantitative real-time polymerase chain reaction (qRT-PCR) or western blot. The proliferation detection was conducted employing Cell Counting Kit-8 (CCK-8) and colony formation assays. Transwell assay was performed to determine cell migratory and invasive abilities. Western blot was implemented for the measurement of related proteins. The targeted interactions among circ_0017956, miR-515-5p and ITGB8 were affirmed by dual-luciferase reporter assay, RNA immunoprecipitation (RIP) assay and RNA pull-down assay. The role of circ_0017956 in NSCLC tumor growth in vivo was studied by xenograft mice model. Circ_0017956 and ITGB8 abundances were overtly raised whereas miR-515-5p was low expressed in NSCLC tissues and cells. Circ_0017956 knockdown caused inhibitory effects on the proliferative and metastasizing capacities of NSCLC cells. Mechanistically, circ_0017956 could sponge miR-515-5p, and circ_0017956 depletion blocked NSCLC cell malignant behaviors by increasing miR-515-5p expression. Furthermore, miR-515-5p targeted ITGB8 and ITGB8 overexpression also neutralized the miR-515-5p-triggered inhibition effects on NSCLC cell progression. Moreover, circ_0017956 could regulate ITGB8 expression through sponging miR-515-5p. In addition, circ_0017956 knockdown repressed NSCLC tumorigenesis by targeting miR-515-5p/ITGB8 axis in vivo. Circ_0017956 could promote NSCLC carcinogenesis at least partly through sponging miR-515-5p and upregulating ITGB8 level, providing a novel theoretical basis for NSCLC treatment.

Introduction

Lung cancer is the most frequent pernicious disease that gives rise to cancer-related death globally, and approximately 85% of the lung cancer cases are non-small cell lung cancer (NSCLC) [1,2]. Despite remarkable improvements in NSCLC treatment, these patients still have gloomy survival rates, and the pathogenesis of NSCLC is still not entirely clear [3]. Therefore, there is an imperative need to search molecular markers for NSCLC diagnosis and develop effective therapies against the metastatic and advanced NSCLC.

Circular RNAs (circRNAs) are endogenous noncoding RNAs (ncRNAs) which have covalently closed-loop structures [4–6]. Numerous circRNAs have been identified as crucial molecular regulators in NSCLC pathogenesis by serving as natural microRNA (miRNA) sponges [7]. For instance, Yang et al. underscored that circEPSTI1 contributed to the malignant progression of NSCLC via miR-1248/TRIM24 axis [8]. Wang et al. disclosed that circRACGAP1 functioned as a miR-144-5p sponge to accelerate NSCLC development [9]. Meanwhile, Zhu et al. revealed that circMLLT10 was upregulated in gastric cancer tissues, and it could accelerate cancer cell advancement via sponging miR-509-3-5p [10]. Besides, circ_0017956, derived from MLLT10, has been reported to be highly expressed in the NSCLC tissues [11]. However, the effects of circ_0017956 on regulating NSCLC progression and the mechanism under it have not been illuminated yet.

MiRNAs are small ncRNAs which exert their regulatory functions by interacting with the 3ʹ untranslated regions (3ʹUTRs) of target mRNAs [12,13]. MiRNAs have been attested to be concerned with the tumorigenesis and development of NSCLC [14]. Previous studies have confirmed that miR-515-5p could serve as a tumor inhibitor in multifarious human cancers, including prostate cancer [15], gastric cancer [16] and bladder cancer [17]. For example, miR-515-5p blocked cell propagation and transferability in bladder cancer via sponging GINS2 [17]. Furthermore, miR-515-5p could impede the progression of NSCLC cells by serving as a sponge of CXCL6 [18]. Nevertheless, the potential targeted relationship between circ_0017956 and miR-515-5p in NSCLC still remains unexplored.

Here, this research mainly inquired the expression and involvement of circ_0017956 in NSCLC malignant progression. Circ_0017956 was identified to be a sponge for miR-515-5p to influence integrin subunit beta 8 (ITGB8) expression, thereby regulating NSCLC cell malignant behaviors in vitro and in vivo. The novel circRNA/miRNA/mRNA signal mechanism based on circ_0017956 was also elucidated in NSCLC during this current study.

Materials and methods

Clinical samples

A sum of 55 paired NSCLC tissue specimens and adjacent noncancerous tissue specimens were obtained from NSCLC patients after the surgical resection at the Second Affiliated Hospital of Dalian Medical University and then were stored at −80°C for standby application. None of the patients enrolled in the research received any preoperative treatment and provided written informed consents before the operation. All procedures of the work were permitted by the Ethical Committee of the Second Affiliated Hospital of Dalian Medical University.

Cell culture

Two NSCLC cell lines (A549 and H1299) and human bronchial epithelial cell line (HBE) were acquired from Procell (Wuhan, China). The cells were cultured in RPMI-1640 medium (Procell) plus 10% fetal bovine serum (FBS; Solarbio, Beijing, China) and 1% antibiotics (Solarbio) in a moist CO₂ (5%) incubator controlled temperature at 37°C.

Cell transfection

Small interference RNA (siRNA) against circ_0017956 (si-circ_0017956#1, si-circ_0017956#2 and si-circ_0017956#3) for circ_0017956 knockdown and negative control (si-NC) were established from Ribobio (Guangzhou, China). For circ_0017956 stable knockdown, circ_0017956 lentiviral short hairpin RNA (sh-circ_0017956) from Ribobio was established, with sh-NC served as negative control. Meanwhile, miR-515-5p mimic or inhibitor (miR-515-5p or anti-miR-515-5p) and matched controls (miR-NC or anti-miR-NC) were also obtained from Ribobio. The pcDNA3.1 vector (RiboBio) was used to construct ITGB8 overexpression vector, with the empty vector (pcDNA) served as the contrast. The transfection of NSCLC cells with above plasmids or oligonucleotides was implemented via Lipofectamine 3000 Reagent (Invitrogen, Carlsbad, CA, USA).

Quantitative real-time polymerase chain reaction (qRT-PCR)

RNA isolation was finished via TRIzol (Beyotime, Shanghai, China). For removing DNA contamination, RNA samples were treated with RNase-free DNase I (Promega, Madison, WI, USA). Subsequently, High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Carlsbad, CA, USA) for circ_0017956 and ITGB8 or TaqMan Advanced miRNA cDNA Synthesis Kit (Applied Biosystems) for miR-515-5p was applied to reversely transcribe the extracted RNA into cDNA. Next, qRT-PCR reaction was executed using BeyoFast™ SYBR Green qPCR Mix (Beyotime) with the obtained cDNA as the amplified template on a PCR system (Applied Biosystems). The expression was assessed with the 2^−∆∆Ct method with β-actin or U6 served as control. The primers were listed as follows: circ_0017956-F: 5ʹ-CGCTATGATCAACCAGGCAAC-3ʹ and circ_0017956-R: 5ʹ-TCAATAACTGAAGCCGTTCCT-3ʹ; miR-515-5p-F: 5ʹ-CGGGTTCTCCAAAAGAAAGCA-3ʹ and 5ʹ-CAGCCACAAAAGAGCACAAT-3ʹ; ITGB8-F: 5ʹ-CTGTTTGCAGTGGTCGAGGAGT-3ʹ and ITGB8-R: 5ʹ-TGCCTGCTTCACACTCTCCATG-3ʹ; β-actin-F: 5ʹ-CACCATTGGCAATGAGCGGTTC-3ʹ and β-actin-R: 5ʹ-AGGTCTTTGCGGATGTCCACGT-3ʹ; U6-F: 5ʹ-GCTCGCTTCGGCAGCACA-3ʹ and U6-R: 5ʹ-GAGGTATTCGCACCAGAGGA-3ʹ.

Cell counting kit-8 (CCK-8) assay

CCK-8 assay was manipulated for evaluating cell proliferative ability. In brief, A549 and H1299 cells were, respectively, plated into the 96-well plates. At specific times after transfection, cells were added with CCK-8 solution (Solarbio) for 4 h at 37°C. Then a microplate reader was utilized to inspect the optical density at 450 nm.

Colony formation assay

Cell proliferation was also assessed using colony formation assay. Following relevant transfection, A549 and H1299 cells (300 cells/well) were sowed into 6-well plates and cultivated for 14 days. Afterward, the macroscopic colonies were immobilized and dyed by methanol (Beyotime) and crystal violet (Beyotime), respectively. The number of colonies was counted with an inverted microscope.

Western blot

Protein extraction was implemented via RIPA buffer (Beyotime). The protein was split by SDS-PAGE (Solarbio) and subsequently blotted on polyvinylidene difluoride membranes (Millipore, Billerica, MA, USA). The samples were obstructed in slim milk for 2 h and mixed with primary antibodies at 4°C for 24 h. Thereafter, the membranes were interacted with secondary antibody (1:3000; ab205718; Abcam, Cambridge, UK) for 2 h. At last, the protein blots were visualized with an enhanced chemiluminescence reagent (Beyotime). The antibodies acquired from Abcam were listed as below: anti-Ki67 antibody (ab243878; 1/1000), anti-proliferating cell nuclear antigen (PCNA) antibody (1:3000; ab92552), anti-ITGB8 antibody (1:2000; ab80673), anti-E-cadherin antibody (ab133597; 1: 2000), anti-N-cadherin antibody (ab207608; 1: 1000), anti-Vimentin antibody (ab137321; 1: 2000) and anti-β-actin antibody (1:200; ab115777). β-actin acted as the internal control in this assay.

Transwell migration and invasion assays

Transwell chambers (BD Bioscience, San Jose, CA, USA) coated with or without Matrigel (BD Bioscience) on the top surface of the membranes were used for the detection of invasive and migratory abilities, respectively. Briefly, transfected cells in the serum-free medium were sowed into the upper compartment, whereas complete medium plus 10% FBS was placed into the bottom compartment. 24 h later, the metastatic cells on the low surface of the membranes were fixed in methanol (Sangon Biotech, Shanghai, China) followed by dyeing with crystal violet (Sangon Biotech). Ultimately, the cells were counted under an inverted microscope at 100× magnification.

Subcellular fractionation assay

The RNA was obtained from the cytoplasm and nucleus of A549 and H1299 cells using Cytoplasmic & Nuclear RNA Purification Kit (Norgen Biotek, Thorold, ON, Canada). Then, the subcellular distribution of circ_0017956 in cytoplasm and nucleus was determined by qRT-PCR, with U6 and 18S rRNA used as the internal nuclear and cytoplasmic controls, respectively.

Dual-luciferase reporter assay

The sequences of circ_0017956 or ITGB8 3ʹUTR harboring the binding sequences of wild-type (WT) or mutant (MUT) miR-515-5p were synthesized and inserted into pmirGLO plasmid (Promega, Madison, WI, USA) to produce circ_0017956-WT, circ_0017956-MUT, ITGB8 3ʹUTR-WT and ITGB8 3ʹUTR-MUT, respectively. Next, an indicated luciferase reporter plasmid was introduced into NSCLC cells together with miR-NC or miR-515-5p. After 48 h, the luciferase activities were inspected by Dual-Luciferase Reporter Assay Kit (Promega).

RNA immunoprecipitation (RIP) assay

RIP assay was enforced by Imprint RNA Immunoprecipitation Kit (Sigma, St. Louis, MO, USA). Firstly, A549 and H1299 cells were dissolved by RIP lysis buffer, and then equal cell lysates were reacted with magnetic beads conjugated with antibody against either Argonaute-2 (Anti-Ago2) (Abcam) or nonspecific Immunoglobulin G (Anti-IgG) (Abcam) at 4°C. Finally, total RNA was extracted from precipitated RNA after purification, and the analysis of circ_0017956 and miR-515-5p enrichments were completed by qRT-PCR.

RNA pull-down assay

Biotin-labeled miR-515-5p probe harboring circ_0017956-matched region (Bio-miR-515-5p) were obtained from Sangon Biotech, with non-target sequence labeled by biotin as the negative control (Bio-miR-NC). A549 and H1299 cells were homogenized in lysis buffer and then incubated with Bio-miR-515-5p or Bio-miR-NC probe conjugated with streptavidin-coated magnetic beads overnight at 4°C. Subsequently, the level of circ_0017956 in total RNA after elution was tested by qRT-PCR.

Xenograft models in mice

An amount of 12 Male BALB/c nude mice (about 6-weeks old, 25 g) were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China). A549 cells stably transduced with sh-circ_0017956 or sh-NC vector were hypodermically inoculated into the right flank of mice (N = 6 each group) to establish the xenograft models in vivo. The tumor volume was computed with the equation of (length × width² × 0.5) via monitoring tumor length and width every week. The tumors were excised at 35 d and the xenografts were collected for weigh and succedent assays. The animal experiment was implemented after the empowerment was obtained from the Animal Ethical Committee of the Second Affiliated Hospital of Dalian Medical University.

Statistical analysis

The data of the experiments with 3 repetitions were analyzed by SPSS version 24.0 software (IBM Corp., Armonk, NY, USA) and exhibited as mean ± standard deviation. The differences were compared through Student’s t-test or one-way analysis of variance. The linear correlations among circ_0017956, miR-515-5p and ITGB8 in tumor samples were evaluated by the Spearman’s correlation coefficient test. P < 0.05 was defined as statistically significant.

Results

The expression of circ_0017956 in NSCLC tissues and cells

To inquiry the function of circ_0017956 in NSCLC, the level of circ_0017956 in 55 paired NSCLC tissues and surrounding normal tissues was investigated. qRT-PCR results manifested the elevated level of circ_0017956 in NSCLC tissues relative to that in normal tissues (Figure 1a). Then, whether circ_0017956 expression level was associated with lymph node metastasis (LNM) and tumor node metastasis (TNM) was explored. A higher level of circ_0017956 was observed in patients with LNM (N = 29) than that in patients without LNM (N = 26) (Figure 1b). In comparison with the TNM stage I+ II group (N = 23), patients with TNM stage III (N = 32) had higher circ_0017956 expression (Figure 1c). Besides, the expression level of circ_0017956 was overtly enhanced in NSCLC cells (A549 and H1299) versus that in normal HBE cells (Figure 1d). These data indicated that circ_0017956 might be involved in NSCLC development.

Figure 1.

The expression of circ_0017956 in NSCLC tissues and cells. (a) qRT-PCR was used to test circ_0017956 level in NSCLC tissues (N = 55) and adjacent normal tissues (N = 55). (b) The enrichment of circ_0017956 in patients with lymph node metastasis (LNM) or not. (c) The level of circ_0017956 in patients with TNM stage I+ II or TNM stage III. (d) qRT-PCR was implemented to analyze the abundance of circ_0017956 in NSCLC cell lines (A549 and H1299) and normal cell line (HBE). *P < 0.05

Circ_0017956 silence inhibited proliferation, migration and invasion of NSCLC cells

In consideration of the upregulation of circ_0017956 in NSCLC, the function of circ_0017956 in NSCLC progression was explored through loss-of-function experiments. The expression of circ_0017956 was apparently decreased in A549 and H1299 cells after si-circ_0017956 (si-circ_0017956#1, si-circ_0017956#2 and si-circ_0017956#3) transfection compared to the control si-NC, indicating that the successful knockdown of circ_0017956 (Figure 2a and Supplementary Fig. 1AS). Therein, si-circ_0017956#2 was selected for subsequent experiments for its highest knockdown efficiency. CCK-8 assay presented that circ_0017956 knockdown impeded cell proliferative ability in A549 cells (Figure 2b). And the colony number of circ_0017956-slienced A549 cells was reduced (Figure 2c). Moreover, the levels of proliferation-related marker proteins (PCNA and Ki67) were decreased in A549 cells with si-circ_0017956#2 transfection (Figure 2d), which further validated that circ_0017956 depletion could suppress NSCLC cell proliferation. Transwell assay exhibited that circ_0017956 silence curbed cell migratory and invasive capacities in A549 cells (Figures 2e and 2f). Besides, the expression of E-cadherin was raised and the enrichments of Vimentin and N-cadherin were declined (Figure 2g), which attested that circ_0017956 knockdown lessened the metastasis of NSCLC cells. In like manner, the suppressive effects of circ_0017956 silence on the proliferation, colony formation and metastasis in H1299 cells were demonstrated (Supplementary Figs. 1BS-1GS). Collectively, above-mentioned results certified the anti-cancer role of circ_0017956 silence in the development of NSCLC cells.

Figure 2.

Circ_0017956 knockdown hindered NSCLC cell proliferation, migration and invasion. (a) Circ_0017956 abundance was judged by qRT-PCR in A549 cells transfected with si-NC, si-circ_0017956#1, si-circ_0017956#2 or si-circ_0017956#3. (b-g) A549 cells were introduced with si-NC or si-circ_0017956#2. (b and c) Cell proliferation was assessed via CCK-8 (b) and colony formation assay (c). (d) The expression of PCNA and Ki67 was detected by western blot. (e and f) Cell migratory and invasive abilities were evaluated using transwell assay. (g) The metastasis-related proteins of E-cadherin, Vimentin and N-cadherin were measured by western blot. *P < 0.05

Circ_0017956 could sponge miR-515-5p in NSCLC cells

To determine the subcellular localization of circ_0017956 in NSCLC cells, we performed subcellular fractionation assays. These results showed that circ_0017956 was mainly localized in the cytoplasm of A549 and H1299 cells (Figure 3a). Then, the precise mechanism of circ_0017956 in NSCLC was explored. The interacting miRNAs that possessed the potential targeted sequence of circ_0017956 were predicted through circinteractome, circBank and StarBase 3.0 databases. The analysis of Venn diagram displayed that miR-515-5p was the shared miRNA that had the potential-binding sequence for circ_0017956 (Figure 3b). The complementary sites between circ_0017956 and miR-515-5p were exhibited in Figure 3c. The miR-515-5p expression was overtly increased in A549 and H1299 cells after miR-515-5p overexpression (Figure 3d), implying the high transfection efficiency of miR-515-5p. Meanwhile, dual-luciferase reporter assay testified that miR-515-5p impeded the luciferase activity of circ_0017956-WT luciferase reporter vectors, whereas it had no noticeable influence in circ_0017956-MUT reporters (Figures 3e and 3f). The RNA immunoprecipitation (RIP) assay exhibited enhanced expression of circ_0017956 and miR-515-5p in the Ago2 pellets (Figures 3g and 3h). Also, circ_0017956 was dramatically pulled down by Bio-miR-515-5p probe using RNA pull-down assay (Figure 3i). Furthermore, miR-515-5p abundance was elevated in circ_0017956-inhibited A549 and H1299 cells (Figure 3j). Moreover, the level of miR-515-5p was declined in NSCLC tissues relative to that in normal tissues (Figure 3k). Contrary to circ_0017956, miR-515-5p expression was declined in patients with LNM and TNM stage III relative to that in patients without LNM and with the TNM stage I+ II group, respectively (Figures 3l and 3m). Additionally, miR-515-5p was low expressed in NSCLC cells, and the abundances of circ_0017956 and miR-515-5p in NSCLC tissues were negatively associated (Figures 3n and 3o). From these data, it was speculated that miR-515-5p was sponged by circ_0017956 in NSCLC cells.

Figure 3.

Circ_0017956 directly interacted with miR-515-5p in NSCLC cells. (a) The subcellular localization of circ_0017956 in NSCLC cells using subcellular fractionation assays, with U6 and 18S rRNA as internal controls. (b) Venn diagram represented the potential miRNAs with potential complementary sites of circ_0017956 predicted by circinteractome, circBank and StarBase 3.0 databases. (c) The binding sequence between circ_0017956 and miR-515-5p predicted by circinteractome was presented. (d) qRT-PCR examined the overexpression efficiency of miR-515-5p. (e-i) Dual-luciferase reporter assay (e and f), RIP assay (g and h) and pulldown assay (i) were implemented for testifying the binding relation of circ_0017956 and miR-515-5p. (j) The level of miR-515-5p in A549 and H1299 cells introduced with si-circ_0017956#2 or si-NC was examined by qRT-PCR. (k) The expression of miR-515-5p by qRT-PCR in 55 pairs of tumor tissues and matched non-tumor tissues. (l) qRT-PCR tested the expression of miR-515-5p in patients with lymph node metastasis (LNM) or not. (m) The level of miR-515-5p in patients with TNM stage I+ II or TNM stage III was examined by qRT-PCR. (n) The miR-515-5p level by qRT-PCR in HBE, A549 and H1299 cells. (o) The correlation between circ_0017956 and miR-515-5p in NSCLC tissues was assessed using Spearman’s correlation coefficient. *P < 0.05

MiR-515-5p inhibitor reversed circ_0017956 knockdown-mediated effects on proliferation and metastasis of NSCLC cells

To explore whether miR-515-5p was concerned with the NSCLC progression regulated by circ_0017956, the rescue experiments were performed by reducing the expression of miR-515-5p in circ_0017956 knocked-down cells. The knockdown efficiency of anti-miR-515-5p was first verified in A549 and H1299 cells (Figure 4a and Supplementary Fig. 2AS). It was presented that miR-515-5p inhibition abolished the inhibitory influences of circ_0017956 downregulation on cell proliferation and colony formation of A549 cells (Figures 4b and 4c). Consistently, western blot assay revealed that the repressive effects of circ_0017956 reduction on the levels of PCNA and Ki67 were also reversed by miR-515-5p inhibitor in A549 cells (Figure 4d). Furthermore, the suppressive impacts of circ_0017956 silence on cell metastasis of A549 cells were abrogated by miR-515-5p knockdown (Figures 4e and 4f). Moreover, the circ_0017956 silence-mediated effects on metastasis-related proteins were also overturned by anti-miR-515-5p in A549 cells (Figure 4g). The above results of functional experiments were also verified in H1299 cells (Supplementary Fig. 2BS-Fig. 2GS). Together, circ_0017956 regulated NSCLC development through sponging miR-515-5p.

Figure 4.

Silence of circ_0017956 decelerated NSCLC advancement by promoting miR-515-5p expression. (a) The level of miR-515-5p was measured in A549 cells transduced with anti-miR-NC or anti-miR-515-5p. (b-g) A549 cells were transduced with si-NC, si-circ_0017956#2, si-circ_0017956#2+ anti-miR-NC, or si-circ_0017956#2+ anti-miR-515-5p. (b and c) CCK-8 and cell colony formation assays were conducted to assess the proliferative or colony formation abilities of transfected cells. (d) Western blot assay detected the proteins associated with proliferation. (e and f) Transwell assay was employed to assess cell migration or invasion. (g) Western blot was executed to examine the expression of metastasis-related proteins. *P < 0.05

ITGB8 was a target of miR-515-5p

To understand the downstream theory of the circ_0017956/miR-515-5p pathway, four bioinformatics tools database (miRDB, OncomiR, Starbase 3.0 and TargetScanHuman) were combined to predict the target genes that had the possible binding sequence for miR-515-5p (Figure 5a). The common data predicted that miR-515-5p harbored regions that match the combinative sites of 18 genes (Figure 5a). Among the 18 candidates predicted by above four prediction algorithms, ITGB8 was the most significant upregulated gene in NSCLC tissues (Figure 5b), so it was assumed as a target gene of miR-515-5p. The latent complementary sequence between miR-515-5p and ITGB8 was shown in Figure 5c. And transient introduction of ITGB8 3ʹUTR-WT in the presence of miR-515-5p mimic caused a significant reduction in the luciferase activity, and this effect was dramatically abrogated by ITGB8 3ʹUTR-MUT (Figures 5d and 5e). Meanwhile, miR-515-5p mimic repressed the mRNA and protein enrichments of ITGB8 in NSCLC cells (figures 5f and 5g). Moreover, ITGB8 expression was visibly raised in NSCLC tissues and cells (Figure 5h-Figure 5k). Further analysis showed that ITGB8 expression was positively correlated with circ_0017956 expression but inversely correlated with miR-515-5p expression level (Figures 5l and 5m). The aforementioned findings corporately revealed that miR-515-5p regulated ITGB8 expression via targeting ITGB8 in NSCLC cells.

Figure 5.

ITGB8 was a target of miR-515-5p in NSCLC cells. (a) Venn diagram represented the potential genes with possible binding sequence of miR-515-5p predicted by miRDB, OncomiR, Starbase 3.0, TargetScanHuman databases. (b) The mRNA expression of the potential targeted genes of miR-515-5p was tested by qRT-PCR. (c) Starbase 3.0 predicted the complementary sites between miR-515-5p and ITGB8. (d and e) The interaction between miR-515-5p and ITGB8 was verified by dual-luciferase reporter assay. (f and g) The mRNA and protein levels of ITGB8 in A549 and H1299 cells transduced with miR-NC or miR-515-5p were measured by qRT-PCR and western blot. (h-k) The mRNA and protein levels of ITGB8 in NSCLC tissues and cells were detected by qRT-PCR and western blot. (l and m) The linear relation between ITGB8 and circ_0017956 or miR-515-5p in NSCLC tissues was assessed using Spearman’s correlation coefficient. *P < 0.05

ITGB8 introduction abolished the influence of miR-515-5p overexpression on the malignant evolvement of NSCLC cells

Considering the relation of miR-515-5p and ITGB8, whether miR-515-5p exerted its function through sponging ITGB8 in NSCLC cells was further probed. The transfection efficiency of ITGB8 overexpression was verified via qRT-PCR and western blot in A549 cells (Figures 6a and 6b) and H1299 cells (Supplementary Fig. 3AS and 3BS). Moreover, ITGB8 increase relieved the depressant impacts of miR-515-5p overexpression on proliferation and colony formation of A549 cells (Figure 6c-Figure 6e). Additionally, the suppressive impacts of miR-515-5p upregulation on cell metastasis of A549 cells were reversed by ITGB8 overexpression (figure 6f-Figure 6h). Besides, we also probed the influences of miR-515-5p, miR-515-5p+ITGB8 and or corresponding controls on the proliferation and metastasis in H1299 cells (Supplementary Fig. 3CS-Fig. 3IS). Consequently, it was evidenced that miR-515-5p functioned in NSCLC cells vy targeting ITGB8.

Figure 6.

ITGB8 upregulation mitigated the impact of miR-515-5p overexpression on NSCLC cell progression. (a and b) After A549 cells were introduced with pcDNA or ITGB8, the expression of ITGB8 was determined by qRT-PCR and western blot. (c-h) A549 cells were transfected with miR-NC, miR-515-5p, miR-515-5p+pcDNA or miR-515-5p+ITGB8. (c and d) CCK-8 and colony formation assays were administrated for analyzing the proliferative ability of the above transfected cells. (e) The proliferation-related proteins PCNA and Ki67 were inspected by western blot. (f and g) The assessment of cell metastasis was achieved by transwell assay. (h) The analyses of proteins regarding metastasis were executed by western blot. *P < 0.05

Circ_0017956 regulated ITGB8 expression through sponging miR-515-5p

The interactions among circ_0017956, miR-515-5p and ITGB8 were also investigated by detecting ITGB8 expression in transfected NSCLC cells. The results depicted that ITGB8 expression was signally lessened in the si-circ_0017956#2-transfected cells, whereas introduction of si-circ_0017956#2 and anti-miR-515-5p counteracted the influence (Figures 7a and 7b). Thus, the above experiments indicated that circ_0017956 sponged miR-515-5p to elevate ITGB8 expression in NSCLC cells.

Figure 7.

Circ_0017956 regulated ITGB8 by working as a miR-515-5p sponge. (a and b) The mRNA and protein levels of ITGB8 were measured in A549 and H1299 cells introduced with si-NC, si-circ_0017956#2, si-circ_0017956#2+ anti-miR-NC or si-circ_0017956#2+ anti-miR-515-5p. *P < 0.05

Circ_0017956 silence inhibited NSCLC growth in vivo

To probe into the effects of circ_0017956 on NSCLC tumor growth in vivo, the sh-circ_0017956-infected or sh-NC-infected A549 cells were implanted into the nude mice to generate the xenograft model. Circ_0017956 silence resulted in reduced tumor volume and weight (Figures 8a and 8b). Also, circ_0017956 was low expressed and miR-515-5p was highly expressed in circ_0017956-silenced mice tumor tissues (Figure 8c). In addition, ITGB8 mRNA and protein levels were decreased in mice tumor tissues of the sh-circ_0017956 group (Figures 8d and 8e). These data collectively pointed out that circ_0017956 knockdown could curb NSCLC growth via miR-515-5p/ITGB8 pathway in vivo.

Figure 8.

Circ_0017956 knockdown impeded NSCLC growth in vivo. A549 cells stably transduced with sh-circ_0017956 or sh-NC vector were subcutaneously inoculated into nude mice. (a) The tumor volume was measured weekly. (b) The representative images were presented and the xenograft tumors were weighed after 35 d. (c-e) The abundances of circ_0017956, miR-515-5p and ITGB8 in mice tumor tissues were inspected by qRT-PCR or western blot. *P < 0.05

Discussion

Accumulating evidence has pointed that circRNAs are key players and could be used as diagnostic and prognostic biomarkers in multifarious cancers [19], including NSCLC [7]. In NSCLC cells, the circRNA/miRNA/mRNA regulatory network has been frequently reported [11]. In the present exploration, we selected circ_0017956 that was claimed to be upregulated in NSCLC as a research subject and identified the precise actions of circ_0017956 in NSCLC pathogenesis.

CircRNAs have previously been considered as miRNA sponges and participate in the carcinogenesis in multifold cancers as tumor promoters or suppressors [20,21]. For example, circMAGI3 functioned an oncogenic role in the tumorigenesis through miR-515-5p/HDGF axis in NSCLC [22]. Also, circMYLK facilitated NSCLC cell proliferation by sponging miR-195-5p [23]. While circ-SLC7A6 acted as a tumor suppressor in NSCLC through abundantly sponging miR-21 [24]. However, the character of circ_0017956 in NSCLC is indistinct. In this study, similar to foregoing report [11], circ_0017956 was verified to be upregulated in NSCLC tissues and cells. Functionally, circ_0017956 silence impeded cell proliferation, colony formation, migration or invasion in NSCLC cells. And the proliferation and metastasis were again verified by measuring the protein expression of proliferation-related and metastasis-related markers. These discoveries suggested that circ_0017956 was involved in the development of NSCLC cells. Besides, the subcellular fractionation assay attested that circ_0017956 was mainly localized in the cytoplasm of NSCLC cells, which provided the possibility for its combination with miRNAs.

Considering the importance of circ_0017956 in regulating NSCLC progression, we then explored its molecular basis through a series of rescue experiments. For the first time, we validated that circ_0017956 sponged miR-515-5p via base pairing. Numerous researches have indicated that miR-515-5p can function as a tumor depressor in many cancers. For instance, miR-515-5p suppressed the proliferation and transferability of bladder cancer cells via SNHG3/miR-515-5p/GINS2 pathway [17]. Meanwhile, miR-515-5p could restrain the evolution of gastric cancer via miR-515-5p/HIF-1α axis [25] and impede cell malignant behaviors by enhancing CBX4 expression in breast cancer [26]. Furthermore, miR-515-5p has a negative effect on NSCLC evolution, and the miR-515-5p/mRNA axis in NSCLC has also been demonstrated in previous studies. Li et al. indicated that miR-515-5p could suppress NSCLC survival and metastasis via serving as a target of CXCL6 in NSCLC cells [18]. Guo et al. point out that miR-515-5p could block the glycolysis of NSCLC through miR-515-5p/HDGF axis [22]. Here, the enhancive miR-515-5p level exerted a repressive effect on NSCLC progression, and miR-515-5p inhibitor reversed circ_0017956 silence-induced impacts on proliferative and metastatic capacities of NSCLC cells. Taken conjointly, these data certified that circ_0017956 could expedite the malignant biological properties of NSCLC cells via targeting miR-515-5p. However, the downstream target gene of circ_0017956/miR-515-5p/mRNA axis has not been reported. Here, dual-luciferase reporter analysis evidenced that ITGB8 was targeted by miR-515-5p, which was a newfound target gene of miR-515-5p/mRNA pathway.

ITGB8 is a member of the integrin beta chain family, and has been verified to be involved in some types of cancers, including prostate cancer [27], colorectal cancer [28] and glioma [29]. Meanwhile, ITGB8 also played a crucial part in lung cancer. For instance, ITGB8 level was obviously raised in Squamous Cell Lung Carcinoma (LUSC) and NSCLC tissues [30]. And ITGB8 knockdown suppressed cell propagation, mobility and facilitated apoptotic ability in NSCLC through miR-145-5p/ITGB8 axis [31]. Consistent with above reports, ITGB8 enrichment was enhanced in NSCLC tissues and cells in the current exploration. Moreover, the repressive influence of miR-515-5p overexpression on NSCLC cell development was neutralized by ITGB8 addition. Therefore, the miR-515-5p/ITGB8 axis was a paralleled pathway with above miR-145-5p/ITGB8 axis. Besides, circ_0017956 could positively regulate ITGB8 expression via sponging miR-515-5p in NSCLC cells. Animal studies also showed that circ_0017956 depletion blocked tumor growth and brought about enhancive miR-515-5p level and reductive ITGB8 level in the tumor tissues. Thus, all the data confirmed that circ_0017956 participated in NSCLC tumorigenesis through sponging miR-515-5p and increasing ITGB8 level.

In conclusion, circ_0017956 was upregulated and contributed to NSCLC cell growth and metastasis by elevating ITGB8 expression via targeting miR-515-5p. This study proposed a circ_0017956/miR-515-5p/ITGB8 regulatory axis in NSCLC carcinogenesis and a novel insight for NSCLC therapy.

Funding Statement

The authors have no funding to report.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplementary material

Supplemental data for this article can be accessed here.