Abstract
Endometriosis (EMS) is a common gynecological condition with apparent heterogeneity, lack of diagnostic markers, and unclear pathogenesis. A series of bioinformatics methods were employed to explore EMS's pathological mechanisms and potential biomarkers by analyzing the combined datasets of EMS (GSE7305, GSE7307, GSE58198, E-MTAB-694), which included 34 normal, 127 eutopic, and 46 ectopic endometrium samples. Then, wet-laboratory experiments (including Western blot, qRT-PCR, and Immunohistochemistry, Immunofluorescence, CCK-8, EdU, Wound healing, Transwell, and Adhesion assays) were applied to examine the biomarkers' expression and function in primary endometrial stromal cells. Bioinformatic analysis indicated that the core pathogenesis of EMS was dysregulated immune-inflammation and tissue remolding processes. Among the upregulated DEGs, BST2 was screened as a potential diagnostic biomarker in EMS, which associated with the revised American Fertility Society (r-AFS) stage and immune-inflammation processes of EMS. Moreover, BST2's overexpression was affirmed in the RNA and protein levels in EMS tissues. In vitro experiments demonstrated that TNF-α promoted the expression of BST2 in ESCs. And BST2 knockdown inhibited migration, invasion, adhesion, and inflammation except for the proliferation of ESCs, probably via the TNF-α/NF-κB pathway. Through a combination of wet and dry studies, we concluded that the core pathogenesis of endometriosis was dysregulated immune-inflammation and tissue remolding, and BST2 might be a potential diagnostic and therapeutic target in endometriosis.