Single-Cell Transcriptomics Reveals Biomarkers for NK Cell Dysfunction in Endometriosis-Associated Immune Dysregulation.

Background: Endometriosis (EM) is associated with immune dysregulation, while dysfunction of natural killer (NK) cells is regarded as a key mechanism underlying immune escape and the persistent growth of ectopic lesions. Method: This study used single-cell RNA sequencing (scRNA-seq) on lesions from three patients with EM and on three normal endometrium samples and integrated these data with three bulk RNA-seq datasets from GEO (GSE105765, GSE7305, and GSE6364). Seurat, Monocle, limma, least absolute shrinkage and selection operator (LASSO), and support vector machine recursive feature elimination (SVM-RFE) were used for cell clustering, trajectory inference, differential expression analysis, and feature selection. Immune-cell composition and pathway activity were evaluated with CIBERSORT and GSVA. Gene expression was validated by qPCR, and cell migration and invasiveness were assessed using wound healing and Transwell assays. Result: scRNA-seq resolved 11 clusters assigned to eight major cell types. By integrating pseudotime features with bulk data, 20 differentially expressed genes (DEGs) were prioritized, and machine-learning analyses identified three key genes: granulysin (GNLY), perforin 1 (PRF1), and ENTPD1. The three-gene model showed good discrimination in the training set and two external validation cohorts (AUCs 0.84, 0.67, and 0.77, respectively). GNLY and PRF1 were predominantly expressed in NK cells and CD8 Conclusion: This study highlights the central role of NK-cell dysfunction in EM pathogenesis and proposes GNLY, PRF1, and ENTPD1 as potential molecular diagnostic biomarkers. Notably, ENTPD1 appears to have dual functions, including immunomodulation and promotion of stromal cell migration, which promotes lesion formation. These findings provide a mechanistic rationale and actionable targets for earlier screening and targeted therapy in EM.