Cuixiang Xu,1,2 Xiangrong Zhao,1,2 Huiting Li,1,2 Yaping Li,1,2 Yangmeng Feng,1,2 Guoan Zhang,3 Xiaoyan Huang1,2 

1Shaanxi Provincial Key Laboratory of Infection and Immune Diseases, Shaanxi Provincial People’s Hospital, Xi’an, People’s Republic of China; 2Shaanxi Engineering Research Center of Cell Immunology, Shaanxi Provincial People’s Hospital, Xi’an, People’s Republic of China; 3Department of Cardiovascular Surgery, Shaanxi Provincial People’s Hospital, Xi’an, People’s Republic of China

Correspondence: Xiaoyan Huang, Email huangxy08@163.com

Background: RNA modifications are associated to various human diseases. However, the functions of RNA modification-related genes have yet to be thoroughly investigated in dilated cardiomyopathy (DCM). This study sought to conduct a comprehensive analysis of RNA modification-associated genes for the diagnosis and subtype classification of DCM.
Methods: We collected DCM and control sample RNA modification-related genes from Gene Expression Omnibus (GEO) microarray datasets. Differential expression analysis was performed on these using the “Limma” package in R. Univariate logistic regression, and the LASSO algorithm were used to identify optimal genes for diagnostic model establishment. Furthermore, ConsensusClusterPlus was used to identify RNA modification-molecular subtypes. Lastly, the expression of the hub RNA modification-related genes and their connection to DCM were confirmed using the clinical samples and mouse models.
Results: Twenty-six RNA modification-related genes were identified as dysregulated in DCM, with strong connections noted among these genes. A diagnostic model based on 13 genes (TRMT61B, MBD2, YTHDC2, NOP2, TRMT10C, WDR4, CPSF2, CSTF3, ZBTB4, UNG, NSUN6, TET1, and DNMT3B) with an AUC of 0.980 predicted DCM well. Infiltrating plasma B cells, eosinophils, CD8 T cells, and regulatory T cells correlated strongly with TRMT61B, MBD2, YTHDC2, and CPSF2. Two RNA modification-molecular subtypes (clusters 1 and 2) were identified. Cluster 1 had greater RNA modification scores, lower immune ratings, and lower HLA-DRB1 and HLA-DPB1 expression than Cluster 2. Cluster 2 engaged metabolism-related pathways, while Cluster 1 activated renin-angiotensin system pathways.We further found a substantial link between lower cardiac function and up-regulation of TET1, DNMT3B, and down-regulation of MBD2, TRMT61B in the 13 hub RNA modification-related genes.
Conclusion: In conclusion, our RNA modification-related diagnostic model predicts DCM well. The discovery of two RNA modification-molecular subgroups and four key pivotal genes may assist stratify DCM patients by risk.

Keywords: dilated cardiomyopathy, RNA modification, N6-methyladenosine, 5-methylcytosine, immune infiltration