Zhenghao Li,* Changying Li,* Yue Shao, Haoyu Ran, Haoming Shi, Ruiqin Zhou, Xuanyu Liu, Qingchen Wu, Cheng Zhang

Department of Cardiothoracic Surgery, The First Affiliated Hospital of Chongqing Medical University, Chongqing, People’s Republic of China

*These authors contributed equally to this work

Correspondence: Cheng Zhang, Department of Cardiothoracic Surgery, The First Affiliated Hospital of Chongqing Medical University, Chongqing, People’s Republic of China, Email zhangchengcqmu@126.com

Purpose: The aim of this study is to investigate the underlying molecular mechanism of oxidative stress (OS) involved in aortic dissection (AD).
Methods: Datasets of AD and OS-related genes were obtained from the Gene Expression Omnibus (GEO) and the GeneCards database, respectively. Differential expression analysis and weighted gene correlation network analysis (WGCNA) were employed to screen genes. After enrichment analysis, a protein–protein interaction (PPI) network was constructed, and machine learning algorithms were used to determine signature genes. Comprehensive bioinformatics analyses on the signature genes were executed, and a clinical prediction model was established and evaluated. External datasets, in vitro experiment, and Mendelian randomization (MR) analysis were applied to validation.
Results: We identified CCL2, ITGB4, MYC, SOCS3, SPP1 and TEK as OS-related signature genes in AD. The area under the ROC curve of all the signature genes was greater than 0.75. The clinical prediction model based on the signature genes showed satisfactory diagnostic efficacy in both training and validation cohorts. In validation cohort and in vitro experiment, CCL2, MYC, SPP1 and TEK were further validated. However, the MR results showed no causal association between the expression of the signature genes and AD.
Conclusion: This study demonstrated that OS participates in and affects the progression of AD. Six biomarkers associated with OS could be perceived as crucial targets for the diagnosis and treatment of AD.

Keywords: bioinformatics analysis, machine learning, Mendelian randomization, expression quantitative trait locus, aortic dissection