Identification and validation of an endoplasmic-reticulum-stress-related gene signature as an effective diagnostic marker of endometriosis

Endometriosis is one of the most common benign gynecological diseases and is characterized by chronic pain and infertility. Endoplasmic reticulum (ER) stress is a cellular adaptive response that plays a pivotal role in many cellular processes, including malignant transformation. However, whether ER stress is involved in endometriosis remains largely unknown. Here, we aimed to explore the potential role of ER stress in endometriosis, as well as its diagnostic value.

Our work not only provides new insights into the impact of ER stress in endometriosis but also provides a novel biomarker with high clinical value.

We retrieved data from the Gene Expression Omnibus (GEO) database. Data from the GSE7305 and GSE23339 datasets were integrated into a merged dataset as the training cohort. Differentially expressed ER stress-related genes (DEG-ERs) were identified by integrating ER stress-related gene profiles downloaded from the GeneCards database with differentially expressed genes (DEGs) in the training cohort. Next, an ER stress-related gene signature was identified using LASSO regression analysis. The receiver operating characteristic curve was used to evaluate the discriminatory ability of the constructed model, which was further validated in the GSE51981 and GSE105764 datasets. Online databases were used to explore the possible regulatory mechanisms of the genes in the signature. Meanwhile, the CIBERSORT algorithm and Pearson correlation test were applied to analyze the association between the gene signature and immune infiltration. Finally, expression levels of the signature genes were further detected in clinical specimens using qRT-PCR and validated in the Turku endometriosis database.

In total, 48 DEG-ERs were identified in the training cohort. Based on LASSO regression analysis, an eight-gene-based ER stress-related gene signature was constructed. This signature exhibited excellent diagnostic value in predicting endometriosis. Further analysis indicated that this signature was associated with a compromised ER stress state. In total, 12 miRNAs and 23 lncRNAs were identified that potentially regulate the expression of ESR1, PTGIS, HMOX1, and RSAD2. In addition, the ER stress-related gene signature indicated an immunosuppressive state in endometriosis. Finally, all eight genes showed consistent expression trends in both clinical samples and the Turku database compared with the training dataset. Conclusions: Our work not only provides new insights into the impact of ER stress in endometriosis but also provides a novel biomarker with high clinical value.