Upregulated miR-18 a-5 p and its regulatory roles in hepatocellular carcinoma : a study based on bioinformatics analysis with miRNA-seq and miRNA-microarray data

Xiaojv Chi,Kaiteng Cai,Wenliang Guo,Aihua Lan,Gang Chen,Jie Ma,Weijia Mo
2018-01-01
Abstract:Background: Expression levels of miR-18a-5p and its molecular mechanisms in hepatocellular carcinoma (HCC) remain poorly understood. This present study aimed to examine expression profiling of miR-18a-5p in HCC samples from available high throughput data and to assess possible related pathways of miR-18a-5p. Material and methods: Expression levels of precursor miR-18a and miR-18a-5p were extracted and analyzed from Cancer Genome Atlas (TCGA) data as well as different miRNA microarrays from Gene Expression Omnibus (GEO) and ArrayExpress databases. This study further conducted meta-analyses to determine clinical expression levels of miR-18a-5p in HCC, combining all data from miRNA-seq and miRNA microarrays with standard mean difference (SMD) and summarizing receiver operating characteristic (sROC) methods. Predicted genes from 12 online platforms were then collected and differentially expressed genes (DEGs) in both TCGA and GEO datasets were interacted. Furthermore, bioinformatics analyses of the intersecting genes were performed to identify possible pathways related to miR-18a5p in HCC using Gene ontology (GO) classification, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways, and protein-protein interaction (PPI) analyses. Results: Levels of precursor miR-18a were more upregulated in 371 HCC tissues (6.2620±1.46955) than in 50 adjacent liver tissues (5880±0.6416, P=0.0015). Results from metaanalyses with 902 HCC cases also indicated that levels of miR-18a-5p were obviously elevated in HCC tissues. AUC of sROC reached 0.88 (0.85-0.91), implying a certain distinguishing capacity of miR-18a-5p in HCC. In total, 78 genes were gathered to gain great possibilities of becoming prospective targets of miR-18a-5p in HCC. Additionally, GO analysis indicated the following centralized pathways for miR-18a-5p in HCC: striated muscle cell differentiation, cell fraction, and electron carrier activity. KEGG pathway analyses indicated that pathways of ‘retinol metabolism’ were the most centralized pathways for miR-18a-5p in HCC. CXCL12 and IGF1 could be core target genes of miR18a-5p in HCC, as provided by the PPI network, whose mRNA levels were downregulated in HCC tissues from TCGA data. Conclusion: miR-18a-5p might contribute to tumorigenesis and progression of HCC via influencing various related pathways. CXCL12 and IGF1 could be novel targets of miR-18a-5p in HCC.
What problem does this paper attempt to address?