AUTHOR=Pan Qi , Yi Caiyu , Zhang Yijie TITLE=Overall Survival Signature of 5-Methylcytosine Regulators Related Long Non-Coding RNA in Hepatocellular Carcinoma JOURNAL=Frontiers in Oncology VOLUME=12 YEAR=2022 URL=https://www.frontiersin.org/journals/oncology/articles/10.3389/fonc.2022.884377 DOI=10.3389/fonc.2022.884377 ISSN=2234-943X ABSTRACT=Purpose

Studies reported that 5-methylcytosine (m5C) RNA transferase alters tumor progression; however, studies of m5C-related lncRNA remain lacking. This article intends to study the lncRNA modified by m5C RNA transferase in hepatocellular carcinoma using a combination of computational biology and basic experiments.

Method

We identified 13 m5C RNA transferase-related genes and selected long non-coding RNAs with a Pearson correlation coefficient greater than 0.4. Univariate Cox regression analysis was used to screen m5C RNA transferase lncRNA related to survival phenotype. We divided TCGA-LIHC into two types of m5C RNA using non-negative matrix decomposition. According to WGCNA, the co-expression models of two lncRNA regulation modes were constructed to analyze the characteristic biological processes of the two m5C RNA transferase-related lncRNA gene models. Then, a predictive model of m5C RNA transferase lncRNA was using LASSO regression. Finally, we used cell experiments, transwell experiments, and clone formation experiments to test the relationship between SNHG4 and tumor cell proliferation in Hep-G2 and Hep-3b cells line.

Results

We identified 436 m5C RNA transferase-related lncRNAs. Using univariate Cox regression analysis, 43 prognostic-related lncRNAs were determined according to P < 0.001. We divided TCGA-LIHC into two regulation modes of m5C RNA transferase using non-negative matrix factorization. The two regulation modes showed significant differences in overall and disease-free survival. We used LASSO to construct m5c-related lncRNA prognostic signature. Thus, a predictive m5C-lncRNA model was established using four lncRNAs: AC026412.3, AC010969.2, SNHG4, and AP003392.5. The score calculated by the m5C-lncRNA model significantly correlated with the overall survival of hepatocellular carcinoma. The receiver operating characteristic curve and decision curve analysis verified the accuracy of the predictive model. We observed a more robust immune response in the high-risk score group. The transwell experiments and clone formation experiments suggested that m5C RNA transferase-related lncRNA SNHG4 promotes the proliferation and migration of Hep-G2 and Hep-3b cells line.

Conclusion

Two lncRNA expression patterns regulated by m5C RNA transferase were identified. The difference between the two expression patterns and the survival phenotype in the biological process was pointed out. A 5-methylcytosine RNA methyltransferases-related lncRNA overall survival signature was constructed. These results provide some understanding of the influence of m5C transferase on hepatocellular carcinoma. The prediction model of m5C transferase lncRNA has potential clinical value in managing hepatocellular carcinoma.