Identification and Validation of a Four-Gene Signature as a Novel Potential Diagnostic Biomarker Panel for Endometriosis: Causal and Mechanistic Insights from Mendelian Randomization and Bioinformatics

Qiu Meng,1 Xingrun Xie,2 Haiyan Wang,1 Shouguo Huang,1 Jie Wang,1 Jiantao Fu,1 Jing Zhang11Department of Obstetrics and Gynecology, Haikou People’s Hospital, The First Affiliated Hospital of Hainan University, Haikou, 570208, People’s Republic of China; 2Department of Radiology, Haikou People’s Hospital, The First Affiliated Hospital of Hainan University, Haikou, 570208, People’s Republic of ChinaCorrespondence: Jing Zhang, Department of Obstetrics and Gynecology, Haikou People’s Hospital, The First Affiliated Hospital of Hainan University, No. 43 Renmin Avenue, Meilan District, Haikou, 570208, People’s Republic of China, Tel +86 0898 66151304, Fax +86 0898 66189688, Email [email protected]; [email protected]: Endometriosis (EMS) is an estrogen-dependent disease that can lead to chronic pain and infertility. Small ubiquitin-related modifier (SUMO)ylation, a key post-translational modification, has been demonstrated to be associated with endometrial dysfunction, but its role in EMS remains unclear. This study aims to identify SUMOylation-related genes as potential biomarkers for EMS utilizing bioinformatics approaches.Methods: The GSE51981, GSE7305, and GSE214411 datasets and 194 SUMOylation-related genes were assessed. The differential expression analysis and weighted gene co-expression network analysis were performed to identify candidate genes associated with EMS and SUMOylation. Subsequently, the univariable and multivariable Mendelian randomization analyses were used to investigate causal relationships. The receiver operating characteristic curve and nomogram were constructed for assessing the performance of candidate genes. Furthermore, the biological roles of candidate biomarkers were evaluated through gene set enrichment analysis. Additionally, immune infiltration levels were analyzed and pseudo-time trajectory analysis was conducted to assess cell differentiation dynamics.Results: CRYZ, FRMD4B, FZD6, and TPST1 were identified as biomarkers associated with SUMOylation in EMS, all with area under the curve values above 0.75, suggesting good diagnostic performance. The constructed nomogram, which incorporated these biomarkers, demonstrated exceptional predictive capabilities for EMS risk. Furthermore, these biomarkers functioned through common enriched pathways, including the cell cycle, neuroactive ligand-receptor interaction, and ubiquitin-mediated proteolysis. Notably, immune cell analysis revealed significant infiltration of regulatory T cells and memory B cells in EMS. Pseudo-time trajectory analysis indicated that epithelial cells displayed distinct expression patterns for the biomarkers across five differentiation stages.Conclusion: This study identifies CRYZ, FRMD4B, FZD6, and TPST1 as potential diagnostic and therapeutic targets associated with SUMOylation in EMS. These findings provide potential clinical insights for the non-invasive detection and risk stratification of EMS, which might help improve the efficiency of early diagnosis and the development of personalized treatment strategies.Keywords: endometriosis, small ubiquitin-like modifier ylation, biomarkers, Mendelian randomization, bioinformatics analysis