Abstract

Introduction: Endometriosis (EM) is an inflammatory condition that affects approximately 10% of the female-born population. It is characterized by the presence of endometrial tissue outside the uterine cavity, leading to chronic pelvic pain, dysmenorrhea, infertility, and a significant reduction in quality of life. The molecular feature of endometriosis remains poorly understood. Material and methods: Endometriosis-related transcriptomic datasets from the Gene Expression Omnibus (GEO) were normalized and analyzed, and candidate hub genes were identified using two machine learning algorithms. These protein expressions were further validated using proteomic data derived from clinical ectopic and eutopic endometrial tissue samples. Functional assays, including gene silencing and cell migration experiments, were conducted to investigate their biological roles. Results: Glutathione Peroxidase 3 (GPX3) emerged as a candidate diagnostic protein. Silencing GPX3 significantly reduced the migratory capacity of endometriosis cells, likely through modulation of antioxidant activity. In clinical samples, ectopic lesions with high GPX3 expression exhibited increased immune cell infiltration, particularly showing elevated CD68⁺ macrophages and PD-1-positive T cells. Conclusion: Collectively, our findings suggest that GPX3 serves as a potential biomarker for the diagnosis and prognosis of endometriosis, with its expression positively correlating with immune cell infiltration in ectopic endometrial tissue.