Identification of Diagnostic Exosomal LncRNA-miRNA-mRNA Biomarkers in Colorectal Cancer Based on the ceRNA Network

Abstract

Background: Colorectal cancer (CRC) is currently the fourth most common cancer worldwide. The roles of exosomal competing endogenous RNAs (ceRNAs) in CRC remain unclear. In this study, we constructed an exosomal ceRNA network to identify the core ceRNAs and investigate the diagnostic biomarkers in CRC. Methods and Patients: Serum exosomes were isolated from four CRC patients and two healthy donors by ultracentrifugation, and then subjected to RNA isolation, sequencing and microarray. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway and Gene Ontology (GO) analyses were performed to identify functional enrichment implications of differentially expressed exosomal mRNAs. TargetScan and miRanda were used for identifying the miRNA-mRNA and miRNA-LncRNA interactions. The predicted lncRNAs and mRNAs were intersected with the differentially expressed genes, for which the screening criterion was fold change >1.5 in the microarray. Differentially expressed exosomal miRNAs were identified in the GSE71008 dataset, and differentially expressed mRNAs (DEmRNAs) were further summarized from The Cancer Genome Atlas (TCGA) database. Results: A total of 1186 exosomal DEmRNAs, 2088 exosomal DElncRNAs and 29 exosomal miRNAs were detected in CRC patients compared to the healthy donors. Functional enrichment analysis suggested that exosomal DEmRNAs might participate in pathways related to carcinogenesis and development of cancer. An exosomal ceRNA regulatory network of CRC was constructed based on 40 lncRNAs, two miRNAs, and five mRNAs. Exosomal miR-150-5p and miR-10b-5p expression levels were increased in healthy donors compared with CRC patients in the GSE71008 dataset, and five DEmRNAs (TOMM70A, RBM48, BEND3, RHOBTB1, and ADAMTS2) were significantly upregulated in TCGA database. Two potential exosomal regulatory axes of lncRNA G016261-miR-150-5p-RBM48 and lncRNA XLOC_011677-miR-10b-5p-BEND3 were identified from the network. Conclusion: The current study revealed potential molecular biological regulation pathways and diagnostic biomarkers through the exosomal ceRNA regulatory network.

Keywords: colorectal cancer; diagnostic biomarker; exosomes; lncRNA; mRNA; miRNA.

FIGURE 1

Identification of isolated exosomes. (A) TEM image of exosomes from CRC patients with 50–150 nm diameter (scale bar: 100 nm; high voltage (HV) = 100 kV). (B) Size distribution of exosomes with 50–150 nm diameter were measured by qNano system. (C) Western blot analysis of CD63, TSG101, and GM130 as exosomal markers.

FIGURE 2

Identification of exosomal DERNAs and Gene ontology and pathway analysis of DEmRNAs (A) Exosomal lncRNA and mRNA were compared by the volcano graph the expression fold changes in healthy donors and CRC patients. The green dots indicated the downregulated lncRNAs and mRNAs, the red dots indicated the upregulated lncRNAs and mRNAs and the black dots represent no differential expressed. (B) DE mRNAs were clustered in top 10 terms from a GO analysis of molecular function, biological process and cellular component. (C) DE mRNAs were clustered in Kyoto Encyclopedia ofGenes and Genomes (KEGG) pathways.

FIGURE 3

Expression analysis of exosomal DElncRNA-DEmiRNAs pairs in CRC (A) Exosomal DElncRNA-DEmiRNAs pairs, the green dots represents up-regulated DElncRNAs, the red dots represents DEmiRNAs (B) Mann‐Whitney U test shows that there is a significant differences in miR-10b-5p expression levels between CRC patients and healthy donors in GSE71008 dataset. Data were expressed as median with interquartile range (*p = 0.0125). (C) Mann‐Whitney U test showed that there is a significant difference in miR-150-5p expression levels between CRC patients vs. healthy donors in GSE71008 dataset. Data were expressed as median with interquartile range (p = **0.0014). (D,E) The AUC of exosomal miR-10b-5p and exosomal miR-105-5p was 0.625 and 0.658 in 100 CRC patients vs. 50 healthy donors.

FIGURE 4

Expression analysis of exosomal miRNA-mRNA internections in CRC (A) Flowchart of the DEmRNAs in CRC. TCGA-COAD: The Cancer Genome Atlas-colon adenocarcinoma, TCGA-READ: The Cancer Genome Atlas-rectal adenocarcinoma. (B) Analysis of exosomal miR-10b-5p and miR-150-5p targeted up-regulated DEmRNAs, red dots represent DEmiRNAs and the blue dots reprsent DEmRNAs. (C–G) The relative expression of BEND3, ADAMTS2, TOMM70A, RBM48, RHOBTB1 from TCGA database (CRC patients n = 623, Healthy donors n = 51). The AUC of BEND3, ADAMTS2, TOMM70A, RBM48, RHOBTB1 were 0.887, 0.695, 0.73, 0.66, 0.692 for CRC diagnostics, respectively. (H) BEND3, ADAMTS2, TOMM70A, RBM48 and RHOBTB1 were significantly up-regulated in CRC tissues compared to the precancerous in the TCGA database. (I) The combination of five DEmRNAs had an AUC of 0.925 with a specificity of 90.2% and sensitivity of 82.8%.

FIGURE 5

Construction and Correlation analyses of exosomal ceRNA network in CRC. (A) Pearson correlation analysis for exosomal DElncRNAs and associated DEmRNAs. Horizontal axis indicates normalized exosomal DElncRNA expression levels; vertical axis indicates normalized exosomal DEmRNA expression levels (lncRNA xloc_011677 correlates to BEND3, lncRNA G016261 correlates to RBM48).(B,C)The exosomal lncRNA G016261-miR-150-5p-RBM48 and exosomal lncRNA XLOC_01167-miR-10b-5p-BEND3 regulatory axis, miRNAs, lncRNAs and mRNAs are represented by ellipse, round rectangle and diamonds, respectively.