Integrative Transcriptomic Analysis Identifies a Key Candidate Gene Bridging Inflammation and DNA Repair in Endometriosis.

BACKGROUND: Endometriosis is a chronic, estrogen-dependent inflammatory disorder characterized by the ectopic implantation of endometrial-like tissue. Although retrograde menstruation is highly prevalent, only a subset of women develops the disease. This epidemiologic paradox suggests that intrinsic molecular alterations in the eutopic endometrium may precondition refluxed cells to survive under inflammatory and oxidative stress. METHODS: Eutopic endometrial transcriptomes from the GSE6364 dataset (21 endometriosis patients and 16 controls) were analyzed across menstrual phases. Differential expression analysis, Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment, and hallmark inflammatory gene set profiling were performed. Protein-protein interaction (PPI) networks and three machine-learning algorithms maximal clique centrality (MCC), random forest (RF), and least absolute shrinkage and selection operator (LASSO) were applied to identify robust feature genes. Immune cell infiltration was estimated using CIBERSORT . Excision repair cross-complementation group 1 (ERCC1) was further evaluated via cross-species evolutionary conservation and in silico structural modeling of high-risk variants. Additionally, an exploratory stemness-related single-sample gene set enrichment analysis (ssGSEA) was conducted, and correlations between the stemness score and the feature genes were examined. RESULTS: A total of 443 differentially expressed genes (DEGs) were identified, which were significantly enriched in inflammatory cascades, including the NF-κB, Toll-like receptor, and cytokine signaling pathways. Four convergent feature genes ERCC1, SOX3, P75NTR (encoded by NGFR), and FPR1 were prioritized. Immune deconvolution revealed selective immune remodeling in the eutopic endometrium, characterized by elevated activated natural killer (NK) cells and reduced CD8+ T cells, which correlated significantly with the expression of the feature genes. ERCC1 exhibited high evolutionary conservation, and structural modeling of missense variants predicted the disruption of protein function within conserved DNA repair domains. Exploratory stemness analysis revealed no significant overall case-control difference after adjusting for menstrual phase, though the early secretory subset exhibited a marginally higher score in endometriosis. CONCLUSIONS: These findings highlight a stress-adaptive transcriptomic state in the eutopic endometrium driven by inflammatory signaling, selective immune remodeling, and altered DNA repair capacity. Specifically, ERCC1 may serve as a critical mechanistic link between inflammatory pressure and impaired genomic maintenance, thereby facilitating cellular persistence and lesion establishment. Furthermore, our data indicate that this four-gene signature primarily reflects a DNA-repair-adaptive program rather than a global bulk-tissue stemness shift.