RNA epigenetics has become a hot topic in recent years. Among different modifications, the m6A is the most abundant RNA modification modulated by methyltransferases, demethylases and binding proteins, which regulates almost each step of mRNA metabolism and multiple biological processes. Emerging data from recent studies have indicated that m6A RNA methylation regulators play a crucial role in most of cancers, contributing to the self-renewal of cancer stem cell, promotion of cancer cell proliferation, and so on. For instance, METTL3, a methyltransferase of m6A, has been reported to be significantly up-regulated in human hepatocellular carcinoma (HCC) and the overexpression of METTL3 is associated with poor prognosis in patients with HCC. WTAP, a methyltransferase of m6A, has been found to regulate migration and invasion of glioblastoma cells. The m6A demethylase ALKBH5 has been reported to demonstrate a high expression in glioblastoma stem-like cells (GSCs) and to maintain tumorigenicity of GSCs by sustaining FOXM1 expression and cell proliferation program. Esophageal cancer is one of the most aggressive malignant tumors with two main histological subtypes, EAC and ESCC. Several studies have explored the association between m6A modifications and ESCC[28, 34, 35]; however, relevant data on EAC remains rare. Given the rising role of m6A modifications in multiple tumors, its exact role in esophageal cancer, especially EAC, needs to be further investigated.
In this present study, we proved that 10 out of 16 m6A RNA methylation regulators were highly expressed in EAC. This is consistent with the results from previous studies focusing on gastric carcinoma, lung cancer, HCC and other malignant tumors[36-38]. Compared with the expression level in normal tissues, the expression of three genes, HNRNPA2B1, HNRNPC, and YTHDF1, were significantly up-regulated in EAC tissues with p<0.001. There was a similar phenomenon in a study on melanoma, where YTHDF1 and HNRNPA2B1 affected modification by genes related to p53-signaling, further up-regulating their expression and facilitated their roles in inhibiting p53 to suppress tumorigenesis. The m6A demethylase FTO represented a relatively higher expression in normal tissues than in EAC tissues in this study (Fig. 1b). Interestingly, FTO has been reported to be highly expressed in ESCC and other malignant tumors such as melanoma and gastric cancer, implying its distinct role in EAC[28, 40, 41]. In the correlation analysis, KIAA1429 was found to present the strongest correlation with YTHDF3. A previous study found that YTHDF3 could enhance the mRNA stability of Zeb1, the downstream target of KIAA1429, thus leading to cell metastasis in HCC. This may contribute to the strong correlation between the two genes. Furthermore, through consensus clustering, three clusters of EAC subgroups were identified based on the expression pattern of m6A RNA methylation regulators. A significant difference in OS was observed among the three subgroups, which indicated that the expression pattern of m6A RNA methylation regulators had a close relationship to the prognosis of patients with EAC. Taken together, these results unveiled the vital role of m6A RNA methylation in EAC.
In our study, we performed univariate, LASSO Cox regression analyses to develop a prognostic related risk signature with 5 genes, HNRNPA2B1, KIAA1429, WTAP, METTL16 and ALKBH5, and then divided the EAC patients into high- and low-risk group. Among the 5 genes, ALKBH5 was the only protective gene for the prognosis of EAC. Nonetheless, a previous study investigated the association between ALKBH5 and prognosis in 177 patients with ESCC, and identified ALKBH5 as the first demethylase that accelerated cell cycle progression and promoted cell proliferation of ESCC cells, suggesting that the up-regulation of ALKBH5 was associated with poor prognosis in ESCC. Conversely, ALKBH5 has been reported to inhibit pancreatic cancer motility by demethylating lncRNA KCNK15-AS1, which agrees with the result in our study to some extent. Considering the opposite effect of ALKBH5 in different types of malignant tumors, more preclinical and clinical evidences are required to find out the exact role of ALKBH5 in EAC. HNRNPA2B1, KIAA1429, WTAP, METTL16 were identified as protective genes, which was consistent with the results in many studies[44-47]. At present, the prognosis assessment mainly relies on TNM stage which includes limited risk factors. Establishing an effective prognosis model with molecular biology risk factors becomes an issue of increasing importance. In the present study, a five-gene risk score model was built (Risk score = (0.145) × WTAP + (0.257) × KIAA1429 + (0.108) × METTL16 + (0.019) × HNRNPA2B1+ (-0.043) × ALKBH5) and used to stratify the survial of EAC patients into high and low risk categories, which exhibited significant difference in OS (p=0.0016), showing a potential value for EAC treatment.
To our best knowledge, this is the first study to compare the expression pattern of m6A RNA methylation regulators in EAC and normal tissues. Our analyses revealed the predictive value of five identified m6A RNA methylation regulators in prognosis of EAC patients and built a risky signature. However, there are some limitations in our study. First, the sample size of normal tissues is limited, which may lead to some biases during the statistical analysis. Second, further molecular biology experiments are required to investigate the mechanisms by which the five identified genes influence the progression of EAC. In addition, the prognostic efficacy of the five-gene signature constructed in the present study still needs further verification in a large-scale population of EAC patients.