Integrating transcriptomes and somatic mutations to identify RNA methylation regulators as a prognostic marker in hepatocellular carcinoma

Abstract

Background

RNA methylation modifying plays an important role in the occurrence and progression of a range of human cancers including hepatocellular carcinoma (HCC), which is characterized by a mass of genetic and epigenetic alterations. However, the treatment targeting these alterations is limited.

Methods

We used comprehensive bioinformatics analysis to analyze the correlation between cancer-associated RNA methylation regulators and HCC malignant features in network datasets.

Results

We identified two HCC subgroups (cluster 1 and 2), which had clearly distinct clinicopathological, biofunctional and prognostic characteristics, by consensus clustering. The cluster 2 subgroup correlated with malignancy of the primary tumor, higher tumor stage, higher histopathological grade and higher frequency of TP53 mutation, as well as with shorter survival when compared with cluster 1. Gene enrichment indicated that the cluster 2 correlated to the tumor malignancy signaling and biological processes. Based on these findings, an 11-gene risk signature was built, which not only was an independent prognostic marker but also had an excellent power to predict the tumor features.

Conclusions

Our study indicated that RNA methylation regulators are vital for HCC malignant progression and provide an important evidence for RNA methylation, methylation regulators are actionable targets for anticancer drug discovery.

Introduction

Hepatocellular carcinoma (HCC), with approximately 750,000 deaths each year, is one of the most common lethal cancers worldwide [1]. At present, HCC occurrence and progression are considered an intricate molecular pathological process involving multi-gene alterations and multi-step processes [2]. Persistent inflammatory microenvironment, such as viral infection, obesity and diabetes, causes hepatic parenchymal necrosis and regeneration. The long term inflammation results in the chromosomal instability and gene mutations in the hepatocytes [3], [4], [5]. However, the specific mechanics are still unclear and therefore, the potent drugs for molecular targeted therapy are sparse. Sorafenib is so far the only FDA-approved drug for advanced HCCs [6].

Emerging data indicated that RNA methylation regulates RNA splicing, nuclear export, translation and degradation. RNA methylation impacts almost all RNAs, like messenger RNAs (mRNAs) [7], tRNAs [8], microRNAs (miRNAs) [9], circular RNAs (cirRNAs) [10], ribosomal RNAs (rRNAs) [11], small nuclear RNAs (snRNAs) [12] and small nucleolar RNAs (snoRNAs) [13]. The majority of those RNAs are in the forms of N1-methyladenosine (m1A), 5-methylcytosine (m5C) and N6,2′-O-dimethyladenosine (m6A). However, these cancer-associated RNA methylations are on the RNA molecules, and the downstream biological functions are not consistent. A recent study found that m6A demethylase fat mass and obesity associated protein (FTO), acting as an oncogene, is up-regulated in acute myelocytic leukemia (AML), resulting in the decreased m6A and differentiation inhibition; FTO depletion enhances differentiation and decreases cell proliferation, cell colonies formation and promotes apoptosis [14]. Agent inhibiting demethylase FTO catalytic activity has anti-tumor effect via targeting FTO/m6A/MYC/CEBPA signaling pathway [15]. Methyltransferase like 3 (METTL3) has an inhibitory reaction to haematopoietic stem cell differentiation and is up-regulated in AMLs, while METTL3 depletion induces myeloid differentiation and cellular apoptosis [16,17]. This irreconcilable opposition is also seen in the gliomas. ALKBH5 as a demethylase is up-regulated in glioblastomas and correlated with poor prognosis, and its down-regulation blocks self-renewal of glioblastoma stem cells and prolongs mice survival [18]. Similarly, depletion of the METTL3-METTL14 complex, which is vital for initiation of m6A methylation process, decreases m6A burdens and facilitates glioblastoma stem cells growth leading to a poor survival [19]. But the recent research has suggested that reducing levels of m6A by silencing METTL3 could inhibit glioblastomas progression and improve the survival [20].

The data above indicate that different tumors or the same tumor at different phases have the differential sensitivity to methylations. HCC is characterized by a mass of genetic and epigenetic alterations, including aberrant activation of WNT signaling pathway and loss-of-function mutation in the TP53 suppressor gene [21,22]. This study aimed to elucidate the role of methylation regulators in HCC at different phase and different differentiation grade, the correlation between each key regulator and HCC clinicopathological feature. We tried to uncover the specific mechanisms of HCC evolution and to facilitate further anticancer drugs discovery.