Abstract
OBJECTIVE: This study aimed to construct a ceRNA network and identify potential diagnostic biomarkers for HSCC. METHODS: The R software was used to identify Differentially Expressed circular RNAs (DE-circRNAs) and mRNAs (DE-mRNAs) between HSCC and Normal Control (NC) specimens. Univariate Cox analysis and survival curves were used to identify biomarkers associated with the prognosis of HSCC. In addition, the diagnostic value of the biomarkers and their relevance to the tumor microenvironment and immunotherapy were evaluated. A ceRNA network based on mRNAs, circRNAs, and miRNAs was established. Finally, qRT-PCR was used to evaluate the expression of biomarkers in HSCC and para-cancerous tissue samples. RESULTS: A network containing 90 miRNAs, 47 circRNAs, and 111 mRNAs was established via bioinformatic analysis. Three genes (NRG1, CCNG2, and CHSY1) were identified as prognostic and diagnostic biomarkers for HSCC. The low-CCNG2-expression and low-CHSY1-expression groups had higher survival rates, whereas the low-NRG1-expression group had a lower survival rate. The AUC values of the three biomarkers > 0.9, which indicated that the diagnostic value of all biomarkers was excellent. Moreover, CCNG2 and CHSY1 were associated with the severity of malignancy and the effectiveness of immunotherapy. Subsequently, a ceRNA network containing 3 mRNAs, 7 miRNAs, and 7 circRNAs was established. The results of qRT-PCR validated that the expression patterns of NRG1 and CHSY1 in clinical HSCC and para-cancerous samples were consistent with those observed in the GSE2379 dataset. CONCLUSION: In this study, we identified three diagnostic biomarkers for HSCC (NRG1, CCNG2, and CHSY1) and established their corresponding ceRNA network via bioinformatic analysis, providing novel insights into the screening and treatment of HSCC.