Circular RNA circFBXO7 attenuates non-small cell lung cancer tumorigenesis by sponging miR-296-3p to facilitate KLF15-mediated transcriptional activation of CDKN1A

Abstract

Background: Accumulating evidence indicates that circular RNAs (circRNAs) play important roles in various cancers. Hsa_circ_0008832 (circFBXO7) is a circRNA generated from the second exon of the human F-box only protein 7 (FBXO7). Mouse circFbxo7 is a circRNA generated from the second exon of mouse F-box only protein 7 (Fbxo7). The role of human circFBXO7 and mouse circFbxo7 in non-small cell lung cancer (NSCLC) has not been reported.

Methods: The expression of circFBXO7 was measured by quantitative real-time PCR. Survival analysis was performed to explore the association between the expression of circFBXO7 and the prognosis of patients with NSCLC. Lung cancer cell lines were transfected with plasmids. Cell proliferation, cell cycle, and tumorigenesis were evaluated to assess the effects of circFBXO7. Fluorescence in situ hybridization assay was used to identify the location of circFBXO7 and circFbxo7 in human and mouse lung cancer cells. Luciferase reporter assay was conducted to confirm the relationship between circFBXO7 and microRNA.

Results: In this study, we found that circFBXO7 was downregulated in NSCLC tissues and cell lines. NSCLC patients with high circFBXO7 expression had prolonged overall survival. Overexpression of circFBXO7 inhibited cell proliferation both in vitro and in vivo. Mechanistically, we demonstrated that circFBXO7 upregulated the expression of miR-296-3p target gene Krüppel-like factor 15 (KLF15) and KLF15 transactivated the expression of CDKN1A.

Conclusions: CircFBXO7 acts as a tumor suppressor by a novel circFBXO7/miR-296-3p/KLF15/CDKN1A axis, which may serve as a potential biomarker and therapeutic target for NSCLC.

Keywords: CDKN1A; Circular RNAs; KLF15; Non-small cell lung cancer; miR-296-3p.

Graphical abstract

Fig. 1

Identification and expression of circFBXO7 in NSCLC cells and tissues. (A) CircFBXO7 is constructed by the cyclization of exon 2 of FBXO7 from chromosome 22. Sanger sequencing confirmed the back-spliced site of circFBXO7. (B) qRT-PCR detected the differential relative expression of circFBXO7 in 52 paired NSCLC tissues and adjacent normal tissues. (C) The Kaplan-Meier method demonstrated the overall survival probability of 52 NSCLC patients with low and high expression of circFBXO7. (D) The existence of cricFBXO7 was confirmed in A549 and H226 cells by agarose gel electrophoresis. The divergent primers only amplified circFBXO7 in cDNA. (E) qRT-PCR for the abundance of circFBXO7 in A549 cells treated with RNase R. (F) qRT-PCR for the abundance of circFBXO7 and linear FBXO7 mRNA in H226 cells treated with Actinomycin D at different time points. (G) The abundance of circFBXO7 in the cytoplasmic and nuclear fractions of A549 and H226 cells. (H) Fluorescence in situ hybridization (FISH) showed the location of circFBXO7 A549 and H226 cancer cells. *P < 0.05, **P < 0.01, ***P < 0.001. NSCLC, non-small cell lung cancer.

Fig. 2

Overexpression of circFBXO7 inhibited NSCLC cell proliferation in vitro and in vivo. (A) The efficiency of transfected circFBXO7 overexpression vector (circFBXO7-OE) in A549 and H226 cells was detected by qRT-PCR. (B,C,D,E,F) The effect of circFBXO7-OE on A549 and H226 cell proliferation was detected by CCK-8 assays (B), colony formation assays (C, D), and EdU assays (E, F). Blue, Hoechst-stained nuclei; red, EdU-positive nuclei; scale bars =100 μm. (G, H) The percentages of cells in the G1, S, or G2 phases in A549 and H226 cells were detected by flow cytometry. CircFBXO7 arrested the A549 and H226 cell cycle at the G1/S phase. (I, J) Tumor weights and volumes of resected tumors were measured 4 weeks after inoculation of circFBXO7-transfected H226 cells (n = 3). ns P > 0.05, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. NSCLC, non-small cell lung cancer. Vector-NC, negative control vector.

Fig. 3

CircFBXO7 upregulated the expression of KLF15 by sponging miR-296-3p. (A) Cluster heatmap of differentially expressed mRNAs in paired A549 and H226 cells transfected with circFBXO7 overexpression vector (circFBXO7-OE) and negative control vector (vector-NC). (B) The qRT-PCR analysis detected the differential abundance of KLF15 in 52 paired NSCLC and adjacent normal tissues. (C) The Pearson correlation analysis of KLF15 and circFBXO7 in 52 paired NSCLC and adjacent normal tissues. (D) The qRT-PCR analysis detected the differential abundance of CDKN1A in 52 paired NSCLC and adjacent normal tissues. (E) The Pearson correlation analysis of CDKN1A and circFBXO7 in 52 paired NSCLC and adjacent normal tissues. (F) A schematic model demonstrated the assumed binding sites of miR-296-3p on circFBXO7 and KLF15. (G, H) Dual-luciferase reporter activity of circFBXO7 and KLF15–3′UTR in H226 cells co-transfected with miR-296-3p mimics or mimics negative control (NC). *P < 0.05, **P < 0.01. NSCLC, non-small cell lung cancer.

Fig. 4

KLF15 inhibited NSCLC cell proliferation by inducing a cell cycle arrest at the G1/S phase. (A) The efficiency of transfected KLF15 overexpression vector (KLF15-OE) in A549 and H226 cells was detected by western blotting assay. (B,C,D,E,F) The effect of KLF15-OE on A549 and H226 cell proliferation was detected by CCK-8 assays (B), colony formation assays (C, D), and EdU assays (E, F). Blue, Hoechst-stained nuclei; red, EdU-positive nuclei; scale bars =100 μm. (G, H) The percentages of cells in the G1, S, or G2 phases in A549 and H226 cells were detected by flow cytometry. KLF15 arrested the A549 and H226 cell cycle at the G1/S phase. (I, J) Tumor weights and volumes of resected tumors were measured 4 weeks after inoculation of H226 cells transfected with negative control vector (vector-NC), miR-296-3p mimics, KLF15-OE, and miR-296-3p mimics plus KLF15-OE (n = 7). ns P > 0.05, *P < 0.05, **P < 0.01, ***P < 0.001. NSCLC, non-small cell lung cancer.

Fig. 5

KLF15 transactivated CDKN1A by binding the promoter region of it. (A) The western blotting assay and qRT-PCR analysis detected the increasing expression of KLF15 and CDKN1A in A549 cells after transfecting circFBXO7-OE. (B) The association between CDKN1A and KLF15 in 52 paired NSCLC tissues was detected by qRT-PCR. (C) The expression of CDKN1A protein after KLF15 overexpression was detected by immunofluorescence assay. (D) The putative transcription factor binding sites of KFL15 in the JASPAR database. (E) The CDKN1A promoter structure was constructed and relative luciferase activities were measured. (F) ChIP-qPCR analysis revealed the presence of KLF15-binding sites in the promoter region (−1000 to −500) of the CDKN1A gene in H226 cells. (G) Schematic model of the human CDKN1A promoter region in 2000 bp upstream or 500 bp downstream of transcription start site (TSS, designated as + 1). ChIP PCR products for putative KLF15 binding sites and an upstream region not expected to associate with KLF15 are depicted with bold lines. ns P > 0.05, *P < 0.05, ** P < 0.01.

Fig. 6

Mouse homologous circular RNA circFbxo7 suppresses tumor growth. (A) CircFbxo7 is constructed by the cyclization of exon 2 of mouse Fbxo7 from chromosome 10. Sanger sequencing confirmed the back-spliced site of circFbxo7. (B) The existence of cricFbxo7 was confirmed in LLC cells by agarose gel electrophoresis. The divergent primers only amplified circFbxo7 in LLC cDNA. (C) Pairwise alignment of the human circFBXO7 and mouse circFbxo7 sequences. (D) The qRT-PCR analysis for the abundance of circFbxo7 in the cytoplasmic and nuclear fractions of LLC cells. (E) Fluorescence in situ hybridization showed the location of circFbxo7 in LLC cells. (F, G) Tumor weights and volumes of resected tumors from wild-type mice were measured after 4 weeks postinoculation of LLC cells transfected with circFbxo7-OE and vector-NC (n = 5). *P < 0.05, **P < 0.01, ***P < 0.001. Vector-NC, negative control vector.

Fig. 7

Hypothesis diagram illustrates the role of circFBXO7 in human NSCLC cells.