PeerJ (Sep 2020)

The role of m6A-related genes in the prognosis and immune microenvironment of pancreatic adenocarcinoma

  • Rong Tang,
  • Yiyin Zhang,
  • Chen Liang,
  • Jin Xu,
  • Qingcai Meng,
  • Jie Hua,
  • Jiang Liu,
  • Bo Zhang,
  • Xianjun Yu,
  • Si Shi

DOI
https://doi.org/10.7717/peerj.9602
Journal volume & issue
Vol. 8
p. e9602

Abstract

Read online Read online

Background Pancreatic adenocarcinoma (PAAD) is among the most lethal diseases and has a dismal prognosis; however, efficient treatment is currently limited. Several studies have observed epigenetic variation during tumorigenesis, suggesting the potential role of RNA methylation, especially N6-methyladenosine (m6A) modification, as a novel epigenetic modification mediating PAAD prognosis. Methods The expression levels of m6A-related genes were downloaded from The Cancer Genome Atlas-Pancreatic Adenocarcinoma (TCGA) and Genotype-Tissue Expression (GTEx) projects, and the findings were validated in four Expression Omnibus (GEO) datasets. A predictive model was constructed using a lasso regression and evaluated by a survival analysis and receiver operating characteristic curve. Consensus clustering identified two distinct subgroups with different immune activity signatures based on the expression pattern of m6A-related genes. The relationship between the mutation state of m6A-related genes and infiltration of immune cells was established and visualized using Tumor Immune Estimation Resource (https://cistrome.shinyapps.io/timer/). Results Fourteen of twenty-one m6A-related genes were differentially expressed between PAAD and normal tissues in TCGA-GTEx cohort. Among these genes, HNRNPC, IGF2BP2 and YTHDF1 were further validated in four GEO datasets. Moreover, an m6A-based model exhibited moderate accuracy in predicting overall survival in PAAD samples. Additionally, potential m6A modification targets were screened by selecting genes from a set of 23,391 genes that not only harbored the most m6A-modified sites but also showed a robust correlation with PAAD survival. Moreover, we correlated the expression level of m6A-related genes with the immune microenvironment of pancreatic cancer for the first time. Specifically, both arm-level gain and deletion of ALKBH5 decreased the infiltration of CD8+T cells (P < 0.05 and P < 0.01, respectively). Conclusion Collectively, our findings suggest a novel anticancer strategy for restoring balanced RNA methylation in tumor cells and guide clinical physicians in developing a new practical approach for considering the impact of related genes on prognosis.

Keywords