Weighted Co-Expression Network Analysis Identifies RNF181 as a Causal Gene of Coronary Artery Disease

Ruoyu Dang; Bojian Qu; Kaimin Guo; Shuiping Zhou; He Sun; Wenjia Wang; Jihong Han; Ke Feng; Jianping Lin; Yunhui Hu

doi:10.3389/fgene.2021.818813

Weighted Co-Expression Network Analysis Identifies RNF181 as a Causal Gene of Coronary Artery Disease

Front Genet. 2022 Feb 10:12:818813. doi: 10.3389/fgene.2021.818813. eCollection 2021.

Authors

Ruoyu Dang¹, Bojian Qu^{1

2}, Kaimin Guo³, Shuiping Zhou⁴, He Sun³, Wenjia Wang³, Jihong Han⁵, Ke Feng⁵, Jianping Lin^{1

2}, Yunhui Hu³

Affiliations

¹ State Key Laboratory of Medicinal Chemical Biology, College of Pharmacy and Tianjin Key Laboratory of Molecular Drug Research, Nankai University, Tianjin, China.
² Pharmaceutical Intelligence Platform, Tianjin International Joint Academy of Biomedicine, Tianjin, China.
³ GeneNet Pharmaceuticals Co. Ltd., Tianjin, China.
⁴ The State Key Laboratory of Core Technology in Innovative Chinese Medicine, Tasly Academy, Tasly Holding Group Co., Ltd, Tianjin, China.
⁵ College of Life Sciences, State Key Laboratory of Medicinal Chemical Biology, Key Laboratory of Bioactive Materials of Ministry of Education, Nankai University, Tianjin, China.

Abstract

Background: Coronary artery disease (CAD) exerts a global challenge to public health. Genetic heritability is one of the most vital contributing factors in the pathophysiology of CAD. Co-expression network analysis is an applicable and robust method for the interpretation of biological interaction from microarray data. Previous CAD studies have focused on peripheral blood samples since the processes of CAD may vary from tissue to blood. It is therefore necessary to find biomarkers for CAD in heart tissues; their association also requires further illustration. Materials and Methods: To filter for causal genes, an analysis of microarray expression profiles, GSE12504 and GSE22253, was performed with weighted gene co-expression network analysis (WGCNA). Co-expression modules were constructed after batch effect removal and data normalization. The results showed that 7 co-expression modules with 8,525 genes and 1,210 differentially expressed genes (DEGs) were identified. Furthermore, Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses were conducted. Four major pathways in CAD tissue and hub genes were addressed in the Hybrid Mouse Diversity Panel (HMDP) and Human Protein Atlas (HPA), and isoproterenol (ISO)/doxycycline (DOX)-induced heart toxicity models were used to validate the hub genes. Lastly, the hub genes and risk variants were verified in the CAD cohort and in genome-wide association studies (GWAS). Results: The results showed that RNF181 and eight other hub genes are perturbed during CAD in heart tissues. Additionally, the expression of RNF181 was validated using RT-PCR and immunohistochemistry (IHC) staining in two cardiotoxicity mouse models. The association was further verified in the CAD patient cohort and in GWAS. Conclusion: Our findings illustrated for the first time that the E3 ubiquitination ligase protein RNF181 may serve as a potential biomarker in CAD, but further in vivo validation is warranted.

Keywords: GWAS meta-analysis; WGCNA (weighted gene co-expression network analyses); coronary artery disease; diagnostic biomarkers; heart tissue.