Lingling Zhu,1,* Ya Dong,2,* Hang Guo,1 Jie Qiu,1 Jun Guo,1 Yonghui Hu,3 Congqing Pan1 

1NHC Key Laboratory of Hormones and Development, Tianjin Key Laboratory of Metabolic Diseases, Chu Hsien-I Memorial Hospital & Tianjin Institute of Endocrinology, Tianjin Medical University, Tianjin, 300134, People’s Republic of China; 2Department of Endocrinology, Shenzhen Second People’s Hospital, the First Affiliated Hospital of Shenzhen University, Health Science Center of Shenzhen University, Shenzhen Clinical Research Center for Metabolic Diseases, Shenzhen Center for Diabetes Control and Prevention, Shenzhen, Guangdong Province, People’s Republic of China; 3Department of Endocrinology, Nanjing First Hospital, Nanjing Medical University, Nanjing, People’s Republic of China

*These authors contributed equally to this work

Correspondence: Yonghui Hu, Email 13132139005@163.com; Congqing Pan, Email profpancq@163.com

Background: Diabetic Cardiomyopathy (DCM) is a heart condition that arises specifically from diabetes mellitus, characterized by cardiac dysfunction in the absence of coronary artery disease or hypertension. The prevalence of DCM is rising in tandem with the global increase in diabetes, necessitating the development of early diagnostic markers and therapeutic targets. This study integrates bioinformatics analysis with experimental validation to identify potential biomarkers for DCM.
Methods: We performed gene expression data mining from the Gene Expression Omnibus (GEO) database. We employed Weighted Gene Co-expression Network Analysis (WGCNA) coupled with machine learning techniques to sift through hub differentially expressed genes (DEGs). Functional enrichment and protein-protein interaction (PPI) network analysis were also conducted to pinpoint key genes functions. Subsequent in vitro and in vivo experiments were performed to validate the findings.
Results: Our analysis revealed six core genes significantly associated with DCM. The expression of Dusp15 was notably downregulated and validated in both high-glucose cultured cardiomyocytes and DCM animal models, suggesting its potential role in DCM pathogenesis.
Conclusion: The integration of bioinformatics with experimental approaches has identified Dusp15 as a promising candidate for a DCM biomarker, offering valuable insights for early diagnosis and potential therapeutic development.

Keywords: diabetic cardiomyopathy, biomarker, WGCNA, machine learning, Dusp15