Abstract N6-Methyladenosine (m6A), the most prevalent epitranscriptomic modification in eukaryotic mRNA, determines the fate of target RNAs. Key m6A regulators have been elucidated to be essential for lung cancer cell proliferation. However, the relationship between individual m6A markers and lung cancer risk remains inadequately explored. We conducted an epitranscriptome-wide association study (EpiTWAS) to investigate genetically predicted m6A in association with lung cancer risk. Genetic and m6A data of 30 lung, 12 heart, 28 muscle, and 50 brain tissues from 87 donors predominantly of European ancestry were obtained from Genotype-Tissue Expression (GTEx). For each m6A marker, we built a single-tissue model via elastic net, focusing on lung tissue, and a cross-tissue model through the unified test for molecular signatures (UTMOST) approach, capturing genetic influences on m6A shared by lung and other tissues. SPrediXcan was used to apply these models to data of genome-wide association studies (GWAS) among 38,422 lung cancer cases and 677,930 controls of European descent from International Lung Cancer Consortium (ILCCO), UK Biobank, and FinnGen. For m6A markers significantly associated with lung cancer risk at false discovery rate (FDR)<0.05, we assessed the differential expression of proteins encoded by their target mRNAs in lung cancer versus adjacent normal tissues (200 pairs), using data from Clinical Proteomic Tumor Analysis Consortium (CPTAC). Of the 38,841 high-quality m6A markers in 11,715 genes analyzed, prediction models were built for 6,809 m6A markers in 4,548 genes with performance of R>0.1 (P<0.05), ~85.2% (5,801) of which were single-tissue models. Among these, 28 m6A markers in 26 genes were significantly associated with overall lung cancer risk (FDR<0.05). Notably, 12 (~43%) of these markers in 12 loci are >2 megabases away from any of GWAS-identified lung cancer risk variants. When conditioning on nearby GWAS signals, these 12 associations did not change materially, whereas 15 of the remaining 16 associations lost statistical significance. Analyses by lung cancer histology and smoking status revealed no significant heterogeneity among the 28 associations but identified eight additional m6A markers significantly (FDR<0.05) associated with the risk of adenocarcinoma (n=2), squamous cell carcinoma (n=3), and overall lung cancer among ever- (n=2) or never-smokers (n=1). Of the 33 target RNAs of those 36 m6A markers, 30 encode proteins and 22 of these proteins displayed a significant tumor-normal differential expression at fold change >1.5 and FDR<0.05 in CPTAC lung tissue data. Our study demonstrates the ability of EpiTWAS to decipher GWAS loci and uncover signals potentially missed by GWAS. Our results imply that certain m6A markers may impact lung cancer risk by modulating mRNA translation, underscoring the need for further research to fully unravel the role of m6A in lung cancer etiology. Citation Format: Yaohua Yang, Yaxin Chen, Yawen Qi, Hongmo Liu, Guochong Jia, Jie Ping, Xiao-Ou Shu, Wei Zheng, Weimin Li, Jirong Long, Qiuyin Cai. RNA m6A modification and lung cancer risk: An epitranscriptome-wide association study [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 7002.
Read full abstract