Identification of a potentially functional circRNA-miRNA-mRNA regulatory network for investigating pathogenesis and providing possible biomarkers of bladder cancer

Abstract

Background: Circular RNAs (circRNAs) have received considerable attention in human cancer research. However, many circRNAs remain to be detected. In our study, we determined novel circRNAs and investigated their effects on bladder cancer (BCa).

Methods: Microarray dataset GSE92675 was downloaded from Gene Expression Omnibus (GEO). Then, we combined computational biology with quantitative real-time polymerase chain reaction (qRT-PCR) to select related circRNAs in BCa. The selected circRNA-microRNA (miRNA)-messenger RNA (mRNA) regulatory subnetwork was determined by Gene Oncology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses.

Results: The regulatory network constructed from the microarray dataset (GSE92675) contained 49 differentially expressed circRNAs (DECs). GO and KEGG analyses showed that the MAPK and PI3K-AKT signaling pathways were statistically significant. On the basis of qRT-PCR and the degree value calculated by the cytoHubba plugin of Cytoscape, hsa_circ_0011385 was finally confirmed. The subnetwork around hsa_circ_0011385 was constructed. In addition, we created a protein-protein interaction (PPI) network composed of 67 nodes and 274 edges after removing independent nodes. GO and KEGG analyses showed that hubgenes were involved in cell cycle activities. Moreover, they could be regulated by miRNAs and play an eventful role in BCa pathogenesis.

Conclusions: We proposed a novel circRNA-miRNA-mRNA network related to BCa pathogenesis. This network might be a new molecular biomarker and could be used to develop potential treatment strategies for BCa.

Keywords: Biomarker; Bladder cancer; ceRNA; circRNA.

Fig. 1

Flowchart of the study. First, we downloaded microarray datasets from GEO to obtain DECs. Next, we intersected miRNAs that they sponge and differentially expressed miRNAs obtained from TCGA. Using similar methods, we obtained overlapping mRNAs and analyzed them for GO and KEGG analyses. Then, we selected nine candidate circRNAs and detected their expression by qRT-PCR. A regulatory network was successfully constructed around hsa_circ_0011385. Subsequently, we detected the expression of target miRNAs and identified hubgenes

Fig. 2

a The PCA plot showed the PC1 and PC2 coefficients of GSE92675. b Volcano plot for DECs in BCa on the basis of GSE92675. A total of 428 DECs were identified, which consisted of 261 upregulated and 167 downregulated circRNAs. We ranked all DECs using a robust method and an adjusted P-value of < 0.05. Twenty-one upregulated and 28 downregulated circRNAs were in the top rankings. c Heatmap for DECs in BCa on the basis of GSE92675

Fig. 3

a, b Differentially expressed miRNAs were obtained from TCGA. c, d We confirmed 28 target miRNAs by intersecting target miRNAs from CSCD and differentially expressed miRNAs obtained from TCGA

Fig. 4

a, b Differentially expressed mRNAs were obtained from TCGA. c, d By intersecting target mRNAs and differentially expressed mRNAs, 476 target mRNAs were selected, of which 79 were upregulated and 397 were downregulated

Fig. 5

CircRNA–miRNA–mRNA regulatory network for DECs

Fig. 6

a, b GO and KEGG analyses were conducted to detect the potential biological functions of target mRNAs. The results showed that the MAPK and PI3K–AKT signaling pathways were statistically significant

Fig. 7

Heatmap for nine circRNAs (hsa_circ_0001955, hsa_circ_0032821, hsa_circ_0060219, hsa_circ_0011385, hsa_circ_0084171, hsa_circ_0040039, hsa_circ_0082582, hsa_circ_0009172, hsa_circ_0077526) in GSE92675

Fig. 8

a Violin plot for the expression of hsa_circ_0011385. b Structural patterns of hsa_circ_0011385 from CSCD

Fig. 9

Heatmap for the significant signaling pathways mediated by the four miRNAs (mir-211-5p, mir-204-5p, mir-182-5p, mir-96-5p) according to DIANA-miRPath

Fig. 10

CircRNA–miRNA–mRNA regulatory network for hsa_circ_0011385

Fig. 11

a, d, g, i Violin plots for the expression of mir-96-5p, mir-182-5p, mir-211-5p, and mir-204-5p in 16 BCa patients. b, e, h, k Differentially expressed miRNAs between high-grade and low-grade BCa except mir-96-5p. c, f, i, l Mir-96-5p, mir-182-5p, mir-211-5p, and mir-204-5p showed significant difference in different stages of BCa. m High expression of mir-204-5p was correlated with poor prognosis in BCa. n Correlation analysis revealed a moderate negative correlation between the expression of hsa_circ_0011385 and miR-204 (r = − 0.43, P = 0.01268)

Fig. 12

a PPI network of target genes. b PPI network of the 18 hubgenes extracted from a

Fig. 13

Violin plots for the expression of the top 18 hubgenes on the basis of TCGA. a RRM2, b MKI67, c CENPA, d AURKB, e FOXM1, f DLGAP5, g DTL, h RACGAP1, i CHEK1, j CDCA8

Fig. 14

a, b GO and KEGG analyses were conducted to detect the potential biological functions of hubgenes