Hsa_circ_0018818 knockdown suppresses tumorigenesis in non-small cell lung cancer by sponging miR-767-3p

Differentially expressed circRNAs in NSCLC

To detect differentially expressed circRNAs in NSCLC, we performed a bioinformatics analysis. In Figure 1A–1C, circRNAs differentially expressed as compared to normal tissues in the GSE101586, GSE101586 and GSE112214 datasets are presented as volcano plots. Overlap among the three datasets is illustrated by the Venn diagram in Figure 1D. Among the differentially expressed circRNAs, 890 in GSE101684 were related to early stage NSCLC.

Figure 1. Differentially expressed circRNAs in NSCLC. Volcano plots illustrating the circRNAs differentially expressed in NSCLC detected in the (A) GSE101684, (B) GSE112214 and (C) GSE101586 datasets. Red indicates a higher expression level, while blue indicates a lower expression level. (D) Venn diagram showing the overlap among the differentially expressed circRNAs in the three datasets. In GSE101684, 890 circRNAs were related to NSCLC.

Gene ontology (Go) and pathway analyses performed to determine the host genes of the circRNAs showed that the most common biological process was nucleocytoplasmic transport. The most enriched cellular component was cell-cell junction, and the most enriched molecular function was GTPase binding (Figure 2A). Pathway analysis revealed that the phosphatidylinositol 3 kinase (PI3K) signaling pathway was related to the development of NSCLC (Figure 2B).

Figure 2. Profiles of circRNAs in NSCLC analyzed with GO and pathway analyses. (A) Go analysis exploring the potential functions of differentially expressed circRNAs. (B) Pathway analysis exploring the signaling pathways related to NSCLC.

Silencing hsa_circ_0018818 significantly inhibits NSCLC cell proliferation

The top differentially expressed circRNAs in GSE101684 were related to early stage NSCLC (Figure 3A). Among them, hsa_circ_0018818 (host gene: USP54) closely correlated with NSCLC tumorigenesis. Therefore, hsa_circ_0018818 was selected for further analysis. In addition, hsa_circ_0018818 was significantly upregulated in NSCLC tissues compared with adjacent normal tissues (Figure 3B). Besides, hsa_circ_0018818 was closely associated with the metastasis of NSLC (Table 1). RT-qPCR analysis showed that hsa_circ_0018818 expression was upregulated in several NSCLC cell lines as compared to BEAS-2B normal lung epithelial cells (Figure 3C). Further analysis of A549 and NCI-H1650 cells showed that transfecting the cells with shRNA1 or shRNA2 targeting hsa_circ_0018818 significantly downregulated its expression (Figure 3D, 3E). Although both shRNAs were stably transfected, hsa_circ_0018818 shRNA1 exhibited better transfection efficiency and was therefore used in subsequent experiments. CCK-8 assays demonstrated that silencing of hsa_circ_000018818 significantly inhibited NSCLC cell proliferation (Figure 3F, 3G).

Figure 3. Downregulation of hsa_circ_0018818 inhibits NSCLC cells proliferation. (A) The overlap of these differentially expressed circRNAs was analyzed. (B) RT-qPCR analysis of hsa_circ_0018818 gene expression in NSCLC and adjacent normal tissues. (C) RT-qPCR analysis of hsa_circ_0018818 gene expression in BEAS-2B, A549, PC-9, NCI-H1441 and NCI-H1650 cells. (D, E) Suppression of hsa_circ_0018818 expression after transfection of A549 and NCI-H1650 cells with shRNA for 24 h. (F, G) CCK-8 assays of A549 and NCI-H1650 cell proliferation. Shown are the OD₄₅₀ values. **P< 0.01 vs. control.

Hsa_circ_0018818 shRNA1 induces apoptosis and reduces the invasiveness of NSCLC cells

Flow cytometry exemplified by the results presented in Figure 4A, 4B showed that downregulating hsa_circ_0018818 obviously induced apoptosis among both A549 and NCI-H1650 cells. Moreover, transwell assays revealed that transfection with hsa_circ_0018818 shRNA1 substantially reduced the invasiveness of these cells (Figure 4C, 4D). Because, NCI-H1650 cells were more sensitive to hsa_circ_0018818 shRNA1 than A549 cells, NCI-H1650 cells were used in the following experiments.

Figure 4. Hsa_circ_0018818 shRNA1 induces apoptosis and inhibits invasion by NSCLC cells. (A, B) The incidence of apoptosis was detected using FACS after double staining cells with Annexin V and PI. X axis: the level of Annexin-V FITC fluorescence; Y axis: the PI fluorescence. (C, D) Transwell assays testing the invasiveness of A549 and NCI-H1650 cells. Magnification: 400×. **P< 0.01 compared to control.

MiR-767-3p is a downstream target of hsa_circ_0018818

To investigate the mechanism by which hsa_circ_0018818 regulates the progression of NSCLC, its interactome was examined using the web tool CircInteractome (https://circinteractome.nia.nih.gov/). We found that miR-767-3p was the most likely downstream target of hsa_circ_0018818 (Figure 5A, 5B). In addition, RT-qPCR analysis demonstrated that miR-767-3p expression was notably upregulated by miR-767-3p agonist and but downregulated by miR-767-3p antagonist (Figure 5C). Dual luciferase reporter assays confirmed that miR-767-3p is a downstream target of hsa_circ_0018818 (Figure 5D). This was further verified by fluorescence in situ hybridization (FISH), which showed their colocalization with cells (Figure 5E). Taken together, these findings indicate that miR-767-3p is a downstream target of hsa_circ_0018818.

Figure 5. MiR-767-3p is the downstream target of hsa_circ_0018818. (A, B) Gene structure of hsa_circ_0018818 indicating the predicted miR-767-3p binding site in its 3'UTR. (C) RT-qPCR analysis miR-767-3p expression in NCI-H1650 cells. (D) The luciferase activity in NCI-H1650 cells after co-transfecting a plasmid encoding the wild-type (WT) or mutant (MT) hsa_circ_0018818 3′-UTR and miR-767-3p. (E) Co-localization of hsa_circ_0018818 and miR-767-3p detected using FISH. **P< 0.01 vs. control. (F) Gene structure of NID1 at the position of bp 161-167 showing the predicted miR-767-3p binding site in its 3'UTR. (G) Luciferase activity in NCI-H1650 cells after co-transfecting a plasmid encoding the WT or MT NID1 3′-UTR and miR-767-3p. **P< 0.01 vs. control.

Nidogen 1 (NID1) is a direct target of miR-767-3p

To determine the target of miR-767-3p, Targetscan (http://www.targetscan.org/vert_71/), miRDB (http://www.mirdb.org/), and dual luciferase assays were used. As illustrated in Figure 5F, 5G, NID1 is a direct target of miR-767-3p.

Hsa_circ_0018818 knockdown inhibits NSCLC progression by inactivating PI3K signaling

Subsequent western blot analysis demonstrated that hsa_circ_0018818 knockdown significantly decreased expression of NID1 (Figure 6A, 6B). This inhibitory effect of hsa_circ_0018818 shRNA1 on NID1 was partially reversed by miR-767-3p antagonist (Figure 6B). Moreover, expression of Twist-2 and E-cadherin in NSCLC cells was notably increased by knockdown of hsa_circ_0018818. In contrast, hsa_circ_0018818 shRNA1 greatly decreased the expression of Vimentin. Meanwhile, downregulation of miR-136 partially suppressed the inhibitory effect of hsa_circ_0018818 shRNA on EMT process of NSCLC (Figure 6A, 6C–6E). Besides, expression of p-Akt and p-ERK in NSCLC cells was significantly downregulated by hsa_circ_0018818 knockdown, but was partially rescued in the presence of miR-767-3p antagonist (Figure 6A, 6F, 6G). This suggests that hsa_circ_0018818 silencing inhibits the progression of NSCLC by inactivating EMT process and PI3K/Akt signaling.

Figure 6. Silencing Hsa_circ_0018818 inhibits NSCLC progression by inactivating EMT process and PI3K/Akt signaling. (A) Western blot analysis of NID1, E-cadherin, Vimentin, Twist-2, Akt, ERK, p-Akt and p-ERK expression in NCI-H1650 cells. (B–G) Relative levels of NID1, Vimentin, E-cadherin, Twist-2,p-Akt and p-ERK expression in NCI-H1650 cells normalized to β-actin expression. **P< 0.01 vs. control. ^##P< 0.01 vs. shRNA1.

Akt inhibitor further enhanced the inhibitory effect of hsa_circ_0018818 shRNA on the progression of NSCLC

To further verify the mechanism by which hsa_circ_0018818 mediated the progression of NSCLC, CCK-8 assay was performed. The data confirmed that anti-proliferative effect of hsa_circ_0018818 shRNA on NSCLC was further increased in the presence of AZD5363 (Figure 7A). Consistently, AZD5363 enhanced the apoptotic effect of hsa_circ_0018818 shRNA (Figure 7B). Moreover, the inhibitory effect of hsa_circ_0018818 shRNA on cell invasion was enhanced by AZD5363 as well (Figure 7C). To sum up, Akt inhibitor further enhanced the inhibitory effect of hsa_circ_0018818 shRNA on progression of NSCLC.

Figure 7. Akt inhibitor further enhanced the inhibitory effect of hsa_circ_0018818 shRNA on progression of NSCLC. (A) The OD value of NCI-H1650 cells was detected by CCK-8 assay. (B) The apoptotic NSCLC cells were examined by flow cytometry. (C) The invasion of NSCLC cells was tested by transwell assay. **P< 0.01 vs. control. ^##P< 0.01 vs. shRNA1.

Hsa_circ_0018818 knockdown significantly inhibits NSCLC tumor growth in vivo

Finally, a xenograft mouse model was established to detect the function of hsa_circ_0018818 in NSCLC in vivo. Four weeks after subcutaneous injection of NC1-H1650 cells, tumor size and weight were significantly lower when cells were transfected hsa_circ_0018818 shRNA1 prior to injection (Figure 8A, 8B). RT-qPCR confirmed that hsa_circ_0018818 levels were stably suppressed within the tumor tissue (Figure 8C). In addition, western blotting revealed that expression levels of NID1, p-Akt and p-ERK were all significantly reduced in tumor tissues expressing hsa_circ_0018818 shRNA1 (Figure 8D–8G). These results demonstrate that downregulating hsa_circ_0018818 significantly attenuates NSCLC tumorigenesis in vivo.

Figure 8. Hsa_circ_0018818 shRNA1 suppresses NSCLC tumor growth in vivo. Mice were subcutaneously injected NC1-H1650 transfected with vector-control or hsa_circ_0018818 shRNA1 or left untreated (Blank), after which tumors were allowed to grow for 4 weeks. (A) Volumes of tumors collected at the indicted times after transplantation. (B) Images of tumors (left) and tumor weights (right) after 4 weeks. (C) RT-qPCR analysis of hsa_circ_0018818 gene expression in tumor tissues. (D) Western blot analysis of NID1, Akt, ERK, p-Akt and p-ERK levels in tumor tissues. (E–G) Relative levels of NID1, p-Akt and p-ERK expression normalized to β-actin expression. **P< 0.01 vs. control.

According to Union for International Cancer Control, version 8.

Cell culture

The BEAS-2B, A549, PC-9, NCI-H441, NCI-H1650 and 293T cell lines were obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA) and cultured in Dulbecco’s Modified Eagle’s Medium (DMEM, Thermo Fisher Scientific, Waltham, MA, USA) supplemented with 10% FBS (Thermo Fischer Scientific), 1% penicillin and streptomycin (Thermo Fisher Scientific) at 37°C under a 5% CO₂ atmosphere. AZD5363 (Akt inhibitor) was obtained from MedChemExpress (MCE, Monmouth Junction, NJ, USA).

Bioinformatics analysis

Three datasets (GSE101586, GSE101684 and GSE112214) containing the gene expression data for NSCLC and adjacent normal tissue (controls) were obtained from the GEO database (https://www.ncbi.nlm.nih.gov/geo/). Among them, GSE101684 was a dataset from a patient with early stage of NSCLC. Gene Ontology (GO) analysis was performed to explore the functional roles of circRNA-targeted genes in terms of biological processes, cellular components, and molecular functions. Biological pathways were defined using the Kyoto Encyclopedia of Genes and Genomes (KEGG).

Tissue collection

In total, 30 pairs of NSCLC samples and adjacent normal tissues were collected from Peking Union Medical College Hospital between June 2018 and June 2019. The clinical and pathological data of these patients were collected with their written informed consent. Each tissue sample was stored at -80°C until RNA extraction. The present study was approved by the Ethics Committee of Peking Union Medical College Hospital. The correlation of hsa_circ_0018818 and clinic-pathological parameters of patients with NSCLC was presented in Table 1.

Quantitative real time polymerase chain reaction (RT-qPCR)

Total RNA was extracted from NSCLC cell lines or tissues using TRIzol reagent (TaKaRa, Tokyo, Japan) according to the manufacturer's protocol. cDNA was synthesized using a reverse transcription kit (TaKaRa, Ver.3.0) according to the manufacturer's protocol. Real-Time qPCRs were performed in triplicate using the following protocol: 2 min at 94°C, followed by 35 cycles or 30 s at 94°C and 45 s at 55°C. The primers for hsa_circ_0018818, miR-767-3p, β-actin and U6 were obtained from GenePharma (Shanghai, China): for Hsa_circ_0018818, 5’-CCCACAGTTTTGCTCTGCAG-3’ (forward) and 5’-AGGTGTAGCCTGAGAAGTACGC-3’ (reverse); for MiR-767-3p, 5’-TCCATTTGTTTTGATGATGGACT-3’ (forward) and 5’-CTCAACTGGTGTCGTGGAGTC-3’ (reverse); for β-actin, 5’-GTCCACCGCAAATGCTTCTA-3’ (forward) and 5’-TGCTGTCACCTTCACCGTTC-3’ (reverse); and for U6, 5’-CTCGCTTCGGCAGCACAT-3’ (forward) and 5’-AACGCTTCACGAATTTGCGT-3’ (reverse). The relative fold changes were calculated using the 2^-ΔΔCt method with the formula 2-(sample ΔCt – control ΔCt), where ΔCt is the difference between the amplification fluorescent thresholds of the gene of interest and the internal reference gene (U6 or β-actin) used for normalization.

Cell transfection

PcDNA3.1 expression vector encoding short-hairpin RNA (shRNA1 or shRNA2) targeting hsa_circ_0018818 or a non-targeted sequence (negative control) were obtained from GenScript Co., Ltd (Nanjing, China). NSCLC cells were then transfected with the vector using Liposome 2000, after which the transfectants were selected by incubation in the presence of G418 (Sigma Aldrich, St. Louis, MO, USA). RT-qPCR was used to verify the transfection efficiency. The shRNA sequences were as follows: hsa_circ_0018818 shRNA1, 5’-CTTCCCACAGTTTTGCTCTG-3’ (forward) and AAAACTTCCCACAGTTTTG (reverse); hsa_circ_0018818 shRNA2, 5’-CAGTTTTGCTCTGCAGACGG-3’ (forward) and AAAAACAGTTTTGCTCTGCA (reverse).

In addition, NCI-H1650 cells were transfected with miR-767-3p agonist, miR-767-3p antagonist or negative control (NC) RNAs (GenePharma, Shanghai, China) using Lipofectamine 2000 as previously described [35]. The transfection efficiency was determined using q-PCR.

CCK-8 assays

A549 or NCI-H1650 cells were seeded into 96-well plates (5×10³ per well) and incubated overnight at 37°C. The cells were then treated with NC, hsa_circ_0018818 shRNA1 or hsa_circ_0018818 shRNA1 + AZD5363 (Akt inhibitor) for 0, 24, 48 or 72 h, after which 10 μl of CCK-8 reagent were added to each well, and the cells were incubated for an additional 2 h. Finally, the absorbance at 450 nm was measured using a microplate reader (Thermo Fisher Scientific).

Cell apoptosis analysis

A549 or NCI-H1650 cells were trypsinized, washed with phosphate-buffered saline, resuspended in Annexin V Binding Buffer, and stained with 5 μl of FITC and 5 μl of propidium (PI) for 15 min. The cells were the analyzed using a flow cytometer (BD, Franklin Lake, NJ, USA) to assess the incidence of cell apoptosis.

Transwell assays

The invasiveness of the cells was assessed using transwell assays. The upper chamber is pre-treated with 100 μl of Matrigel, after which A549 or NCI-H1650 cells were seeded into the upper chamber in RPMI1640 medium with 1% FBS. The density was adjusted to about 1.0×10⁶ cells per chamber. RPMI1640 medium with 10% FBS was added in the lower chamber. After incubation for 48 h at 37°C, the non-invading cells in the upper chamber were removed with a cotton swab. Cells in the lower chamber were stained with 0.1% crystal violet and counted under a microscope (LEICADMLB2, Frankfurt, Germany).

Dual luciferase reporter assay

The partial sequences of hsa_circ_0018818 and the NID1 3’-UTR containing putative miR-767-3p binding sites were obtained from Sangon Biotech (Shanghai, China) and cloned into pmirGLO Dual-Luciferase miRNA Target Expression Vectors (Promega, Madison, WI, USA) to construct wild-type reporter vectors hsa_circ_0018818 (WT) and NID1 (WT), respectively. Mutant hsa_circ_0018818 and NID1 3’-UTR sequences containing the putative miR-767-3p binding site were generated using a Q5 Site-Directed Mutagenesis Kit (New England Biolabs, Ipswich, MA, USA) and then cloned into pmirGLO vectors to generate the mutant-type reporter vectors hsa_circ_0018818 (MUT) and NID1 (MUT). hsa_circ_0018818 (WT) or hsa_circ_0018818 (MUT) were transfected into 293T cells together with NC or miR-767-3p agonist using Lipofectamine 2000 (Thermo Fisher Scientific). Similarly, NID1 (WT) or NID1 (MUT) was transfected into 293T cells together with NC or miR-767-3p agonist. Relative luciferase activities were then analyzed using a Dual-Glo Luciferase Assay System (Promega).

Fluorescence in situ hybridization (FISH)

To explore the relation between hsa_circ_0018818 and miR-767-3p, their colocalization in the cytoplasm was investigated using FISH as previously described [36].

Western blotting

Total protein was isolated from cell lysates or tumor tissues using RIPA buffer, and quantified using a BCA protein assay kit (Beyotime, Shanghai, China). Proteins were resolved on 10% SDS-PAGE and then transferred to PVDF (Bio-Rad) membranes. After blocking, the membranes were incubated first with primary antibodies at 4°C overnight and then with secondary anti-rabbit antibody (Abcam; 1:5000) at room temperature for 1 h. Membranes were scanned using an Odyssey Imaging System and analyzed using Odyssey v2.0 software (LICOR Biosciences, Lincoln, NE, USA). The primary antibodies were anti-Akt (Abcam, Cambridge, MA, USA; 1:1000), anti-ERK (Abcam; 1:1000), anti-NID1 (Abcam; 1:1000), anti-E-cadherin (Abcam; 1:1000), anti-Twist2 (Abcam; 1:1000), anti-Vimentin (Abcam; 1:1000) and anti-β-actin (Abcam; 1:1000). β-actin served as an internal control.

In vivo study

Eighteen BALB/c nude mice (6-8 weeks old) were purchased from Vital River (Beijing, China) and housed within a dedicated SPF facility. NCI-H1650 cells stably expressing hsa_circ_0018818 shRNA1 were transplanted subcutaneously into each mouse as described previously [37]. Tumor volume was then measured weekly as described previously [38]. At the end of the experiment, the mice were sacrificed and the tumors were collected and weighted. All in vivo experiments were performed in accordance with National Institutes of Health Guide for the Care and Use of Laboratory Animals. The protocol was approved by the Ethics Committees of Peking Union Medical College Hospital.

Statistical analysis

All data were expressed as the mean ± standard deviation (SD) of at least three independent experiments. Differences were analyzed using one-way analysis of variance (ANOVA) followed by Tukey’s test (more than 2 groups, Graphpad Prism7). Values of P<0.05 was considered statistically significant.