A two-circular RNA signature as a noninvasive diagnostic biomarker for lung adenocarcinoma

Abstract

Background: Recently, circular RNAs (circRNAs) have been reported to be microRNA sponges and play essential roles in cancer development. This study aimed to evaluate whether circulating circRNAs could be used as diagnostic biomarkers for lung adenocarcinoma (LUAD).

Methods: The Gene Expression Omnibus (GEO) dataset was used to investigate differentially expressed circRNAs (DEcircRNAs) in paired LUAD tissues and adjacent nontumor tissues. The expression levels of the host genes were analyzed in The Cancer Genome Atlas (TCGA)-LUAD dataset, and the prognostic value was assessed using the Kaplan-Meier plotter. Quantitative real-time PCR (qRT-PCR) was performed to validate the expression of candidate circRNAs in the LUAD plasma and cells. The CCK8 assay was used to measure the function of circRNAs in cell proliferation. Competing endogenous RNA (ceRNA) network, gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses were performed to predict the possible mechanisms and functions of circRNAs in LUAD.

Results: Two upregulated and two downregulated circRNAs were identified as candidate circRNAs using bioinformatics analysis. qRT-PCR demonstrated that hsa_circ_0005962 was upregulated in LUAD plasma and cells, whereas hsa_circ_0086414 was downregulated. Receiver operating characteristic (ROC) curve analysis confirmed that a signature comprising the two circRNAs had good diagnostic potential, with an area under the ROC curve (AUC) of 0.81 (P < 0.0001). In addition, we observed that overexpression of plasma hsa_circ_0086414 was related to EGFR mutations (P = 0.001). Plasma hsa_circ_0005962 displayed significantly different expression before and after surgery in patients with LUAD (P < 0.0001). In vitro experiments suggested that hsa_circ_0005962 promoted LUAD cell proliferation. For future studies, we predicted the circRNA-miRNA-mRNA network for hsa_circ_0005962. Bioinformatics analysis revealed that hsa_circ_0005962 might be involved in LUAD development.

Conclusion: A circRNA signature was identified as a potential noninvasive biomarker for LUAD diagnosis.

Keywords: Biomarker; CircRNA; Diagnosis; Hsa_circ_0005962; Hsa_circ_0086414; LUAD; Plasma.

Fig. 1

Circular RNA (circRNA) expression profiles of lung adenocarcinoma (LUAD) and matched nontumor tissues. a Volcano plot showing the differential expression of circRNAs between the tumor and normal groups. b Heatmap of the differential expression of the 23 selected LUAD-specific circRNAs in the GSE101586 dataset. c The expression of host genes of hsa_circ_0005962, hsa_circ_0003958, hsa_circ_0086414 and hsa_circ_0001936 in The Cancer Genome Atlas (TCGA)-LUAD dataset. d Overall survival analysis for the host genes of the four circRNAs in LUAD patients was performed using the Kaplan–Meier plotter. Log-rank tests were used to determine statistical significance. ****P < 0.0001

Fig. 2

Quantitative real-time PCR (qRT-PCR) analysis of the candidate circRNAs in the LUAD patients and normal controls. a Hsa_circ_0005962 expression was upregulated in the plasma of the LUAD patients compared with that of the normal controls. b Hsa_circ_0086414 was downregulated in the LUAD patients. Hsa_circ_0003958 (c) and hsa_circ_0001936 (d) were detected without significantly differential expression between the two groups. e Hsa_circ_0005962 was upregulated in LUAD patients with TNM stage I. f Hsa_circ_0086414 was downregulated in LUAD patients with TNM stage I. ****P < 0.0001

Fig. 3

ROC analysis of hsa_circ_0005962, hsa_circ_0086414 and the combination of the two markers for the diagnosis of LUAD. a TNM stages I-IV; Combination: The area under the ROC curve (AUC) 0.81, P < 0.0001; Hsa_circ_0005962: AUC 0.73, P < 0.0001; Hsa_circ_0086414: AUC 0.78, P < 0.0001. b TNM stage I; Combination: AUC 0.83, P < 0.0001; Hsa_circ_0005962: AUC 0.79, P < 0.0001; Hsa_circ_0086414: AUC 0.77, P < 0.0001

Fig. 4

Comparison of plasma hsa_circ_0005962 and hsa_circ_0086414 expression in preoperative and postoperative LUAD patients. a Upregulated hsa_circ_0005962 was detected at a lower level in the plasma samples after surgery. b No significant differential expression of hsa_circ_0086414 was detected in the patients before and after surgery. ****P < 0.0001

Fig. 5

Hsa_circ_0005962 promotes LUAD cell proliferation in vitro. a Hsa_circ_0005962 expression was upregulated in the LUAD cells (A549, H1299 and HCC827) compared to that of the normal human bronchial epithelial cells (16HBE). b Hsa_circ_0086414 was downregulated in the LUAD cells compared to that of the 16HBE cells and was differentially expressed between the LUAD cells with (HCC827) and without (A549 and H1299) EGFR mutations. c qRT-PCR analysis of hsa_circ_0005962 expression in A549 and H1299 cells after transfection with a specifically synthesized siRNA. d The CCK-8 assay showed that hsa_circ_0005962 promoted A549 and H1299 cell proliferation. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001

Fig. 6

Bioinformatics analysis of the hsa_circ_0005962–miRNA–mRNA interactions. The competing endogenous RNA (ceRNA) network was constructed and visualized with the Cytoscape software

Fig. 7

Functional enrichment analysis for hsa_circ_0005962. a Gene ontology (GO) analysis of hsa_circ_0005962 based on the ceRNA network. The top 10 significantly enriched biological processes (BPs), cellular components (CCs) and molecular functions (MFs) are listed. b Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis of hsa_circ_0005962 based on the ceRNA network