Integrated investigation of the clinical implications and targeted landscape for RNA methylation modifications in hepatocellular carcinoma

RNA methylation (RM) is a crucial post-translational modification (PTM) that directs epigenetic regulation. It mostly consists of N¹-methyladenosine (m¹A), 5-methylcytosine (m⁵C), N³-methylcytidine (m³C), N⁶-methyladenosine (m⁶A), and 2′-O-methylation (Nm). The “writers” mainly act as intermediaries between these modifications and associated biological processes. However, little is known about the interactions and potential functions of these RM writers in hepatocellular carcinoma (HCC).

The expression properties and genetic alterations of 38 RM writers were assessed in HCC samples from five bioinformatic datasets. Two patterns associated with RM writers were identified using consensus clustering. Then, utilizing differentially expressed genes (DEGs) from different RM subtypes, we built a risk model called RM_Score. Additionally, we investigated the correlation of RM_Score with clinical characteristics, tumor microenvironment (TME) infiltration, molecular subtypes, therapeutic response, immunotherapy effectiveness, and competing endogenous RNA (ceRNA) network.

RM writers were correlated with TME cell infiltration and prognosis. Cluster_1/2 and gene.cluster_A/B were shown to be capable of distinguishing the HCC patients with poor prognosis after consensus and unsupervised clustering of RNA methylation writers. Additionally, we constructed RNA modification pattern-specific risk model and subdivided the cases into RM_Score high and RM_Score low subgroups. In individual cohorts or merged datasets, the high RM_Score was related to a worse overall survival of HCC patients. RM_Score also exhibited correlations with immune and proliferation related pathways. In response to anti-cancer treatments, the RM_Score had a negative correlation (drug sensitive) with drugs that focused on the MAPK/ERK and metabolism signaling, and a positive correlation (drug resistant) with compounds targeting RKT and PI3K/mTOR signaling pathway. Notably, the RM_Score was connected to the therapeutic effectiveness of PD-L1 blockage, implying that RM writers may be the target of immunotherapy to optimize clinical outcomes. Additionally, a ceRNA network was generated including 2 lncRNAs, 4 miRNAs, and 7 mRNAs that was connected to RM writers.

We thoroughly investigated the potential functions of RNA methylation writers and established an RM_patterns-based risk model for HCC patients. This study emphasized the critical functions of RM modification in TME infiltration, targeted therapy, and immunotherapy, providing potential targets for HCC.