INTEGRATED ANALYSIS OF LNCRNA-MIRNA-MRNA CERNA NETWORK IN THE SCREENING OF POTENTIAL BIOMARKERS IN THE PROGNOSIS OF ORAL SQUAMOUS CELL CARCINOMA

Background: A large number of studies have shown that the long noncoding RNAs (lncRNAs) can be used as competitive endogenous ribonucleic acids (ceRNAs) to bind with microRNA (miRNA), thereby affecting and regulating the expression of mRNA and target genes. These lncRNA-related ceRNAs are believed to play an important role in the occurrence and development of cancer. However, its role and function in the oral squamous cell carcinoma (OSCC) remains unclear. Methods: GSE84805 data set was selected and divided into control group and tumor group. The differential expression probes were screened at P < 0.05 and logFC greater than 2 or less than -2 through R language limma package. The target genes of lncRNA and miRNAs were predicted and analyzed using the Starbase online prediction tool. Go enrichment analysis and KEGG pathway analysis was used to analyze the differentially expressed target genes, and a PPI interaction network was constructed to screen the hub genes. The survival analysis of the screened hub genes was performed using the clinic-pathological data of the TCGA-HSCC data set and follow-up data. Oral squamous cell carcinoma cell lines CAL-27, TCA-8113, SCC-9, and the normal oral epithelial cell line HIOEC were cultured in vitro as controls and the mRNA expression levels of lncRNA hcg22 in each cell line were determined by qRT-PCR. Results: We found five genes and mRNAs that were differentially expressed in the normal and tumor tissues. Through GO and KEGG analysis, we further analyzed the behavior and functions of the participation of differentially expressed genes (DEGs). Results of the GO enrichment analysis showed that specific genes were concentrated in several process fields, such as extracellular structure, muscle contraction, and nuclear division. Through survival analysis, we identified the hub genes, such as FN1, COL1A1, co14a2, COL7A1 and ITGA3 exhibited the highest degree score. A ceRNA network with OSCC specific miRNA and lncRNA expression was constructed. Conclusion: Our study showed that age (P < 0.05, or = 1.022 [1.009-1.034]), and a high expression of FN1 (P < 0.05, or = 1.179 [0.74-1.877]) and ITGA3 (P < 0.05, or = 1.376 [1.012-1.871]) in tumor tissues were the important risk factors for the poor prognosis of OSCC.