Comprehensive Analysis of METTLs (METTL1/13/18/21A/23/25/2A/2B/5/6/9) and Associated mRNA Risk Signature in Hepatocellular Carcinoma.

ABSTRACT

Currently, 80%-90% of liver cancers are hepatocellular carcinomas (HCC). HCC patients develop insidiously and have an inferior prognosis. The methyltransferase-like (METTL) family principal members are strongly associated with epigenetic and tumor progression. The present study mainly analyzed the value of METTLs (METTL1/13/18/21A/23/25/2A/2B/5/6/9) and associated mRNA risk signature for HCC. METTLs expression is upregulated in HCC and is a poor prognostic factor in HCC. METTLs were upregulated in patients older than 60 and associated with grade. Except for METTL25, the remaining 10 genes were associated with the HCC stage, invasion depth (T). In addition, METTLs showed an overall alteration rate of 50%. Except for METTL13/2A/25/9, the expression of the other seven genes was significantly associated with overall survival, disease-specific survival, and progression-free survival. Multivariate studies have shown that METTL21A/6 can be an independent prognostic marker in HCC. A total of 664 mRNAs were selected based on Pearson correlation coefficient (R > 0.5), unsupervised consensus clustering, weighted coexpression network analysis, and univariate Cox analysis. These mRNAs were significantly associated with METTLs and were poor prognostic factors in HCC patients. The least absolute shrinkage and selection operator (lasso) was used to construct the best METTLs associated with mRNA risk signature. The mRNA risk signature was significantly associated with age, stage, and t grade. The mRNA high-risk group had higher TP53 and RB1 mutations. This study constructed a nomogram with the mRNA risk profile and clinicopathological features, which could better predict the OS of individuals with HCC. We also analyzed associations between METTLs and mRNA risk signatures in epithelial-mesenchymal transition, immune checkpoints, immune cell infiltration, tumor mutational burden, microsatellite instability, cancer stem cells, tumor pathways, and drug sensitivity. In addition, this study constructed a protein interaction network network including METTLs and mRNA risk signature genes related to tumor microenvironment remodeling based on single-cell sequencing. In conclusion, this study provides a theoretical basis for the mechanism, biomarker screening, and treatment of HCC.