Identification of seven circRNAs that are repeatedly down-regulated in gastric cancer as a biomarker and their potential miRNA targets

Abstract Objective: This study aimed to screen and identify the differentially-expressed circRNAs (DE-circRNAs) in multiple microarray data sets of gastric cancer (GC), predict their corresponding miRNAs, and explore their regulatory relationship, thereby identifying potential molecular markers for early diagnosis of GC. Results : Seven DE-circRNAs were screened and identified: hsa_circ_0007991, hsa_circ_0013048, hsa_circ_0048607, hsa_circ_0050745, hsa_circ_0054971, hsa_circ_0059802 and hsa_circ_0067934. The host genes of these DE-circRNAs are enriched in cancer-related signaling pathways, including ERb signaling, mTOR signaling, chemokine signaling, and B cell receptor signaling. Moreover, these seven DE-circRNAs were all down-regulated in GC tissues. The 7 miRNAs predicted by DE-circRNAs were expressed in GC tissues, namely hsa-miR-1182, hsa-miR-1225-3p, hsa-miR-1275, hsa-miR-193a-5p, hsa-miR-543, hsa-miR-622 and hsa-miR-630. Among them, miR-193a targets were abundant and highly enriched in cancer-related pathways (ERb signaling and mTOR signaling), which was highly consistent with the enrichment of the host gene where the DEcircRNA was located. These results suggested that circRNA may regulate the occurrence and development of GC through the regulation of the expression of the host genes and miRNA. In clinical samples, DE-circRNAs were less expressed in GC tissues when compared to normal tissues. Conclusion : Seven down-regulated DE-circRNAs and 7 targeted miRNAs were identified in GC, which contributed to the understand of the underlying molecular mechanism of GC and provided a basis for the early diagnosis of GC and the development of targeted drugs. Methods: Four sets of circRNA microarray data and one set of miRNA microarray data in GC and adjacent tissues were downloaded from the Gene Expression Omnibus (GEO) and analyzed using GEO2R tool to identify DE-circRNAs and differentially-expressed miRNAs (DE-miRNAs). Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses were used to analyze the enrichment pathway of host mRNAs of DE-circRNAs. Miranda and TargetScan were used to predict miRNA-binding DE-circRNAs. The predicted DE-circRNA interacting miRNAs were intersected with DE-miRNAs, and KEGG and Reactome analyzed the targeted mRNAs corresponding to the obtained miRNAs. DE-circRNAs in GC tissues and adjacent tissues was verified in clinical samples.