[Transcriptomic analysis of NT5E/CD73-related macrophage state remodeling in the maternal-fetal immune...

Objective To investigate changes in the immune microenvironment of placental villous tissue in recurrent spontaneous abortion (RSA) based on transcriptomic analysis, and to preliminarily explore the potential relationship between the adenosine-NT5E/CD73 axis and macrophage state remodeling. Methods The placental villous tissue transcriptome dataset GSE237732, including RSA and normal pregnancy samples, was obtained from the gene expression omnibus (GEO) database. Candidate key genes were screened by differential expression analysis, weighted gene co-expression network analysis (WGCNA), and protein-protein interaction (PPI) network analysis. Immune infiltration analysis was performed to assess the association between key genes and relative immune cell abundance, and gene set enrichment analysis (GSEA) was used to explore related signaling pathways. The placental microarray dataset GSE22490 was used for independent validation of NT5E expression and macrophage/inflammation-related signatures. The decidual CD45⁺ immune cell single-cell dataset GSE164449 was also analyzed as an exploratory bypass dataset. Finally, a THP-1-derived macrophage model was used to evaluate the effects of CD73 inhibition and adenosine supplementation on inflammatory markers of M1 macrophages. Results Nine key genes were identified from GSE237732, including PDGFRB, DCN, COL6A1, COL5A2, PDGFRA, LTBP2, TIMP1, ENG, and NT5E. Immune infiltration analysis showed that all nine key genes were significantly correlated with M1 macrophages; ENG and NT5E were negatively correlated, whereas the other genes were positively correlated. GSEA indicated that NT5E was involved in multiple pathways related to immune dysregulation and inflammatory responses. In GSE22490, NT5E expression showed a decreasing trend in the recurrent miscarriage group, whereas CD86 and M1/inflammatory signatures showed increasing trends, although these changes were not statistically significant. The M2 macrophage signature was significantly increased, suggesting overall macrophage state remodeling rather than isolated M1 elevation. Exploratory single-cell analysis of GSE164449 showed increased macrophage-like cell proportions and increased M1- and M2-related scores in RPL-associated decidual CD45⁺ immune cells. Quantitative real-time PCR and flow cytometry showed that APCP-mediated inhibition of NT5E/CD73 activity increased inflammatory factor expression and the CD86^-positive rate in M1 macrophages, whereas adenosine supplementation partially reversed these effects. Conclusion NT5E-encoded CD73 may participate in the regulation of macrophage inflammatory phenotypes through adenosine signaling and may be associated with immune microenvironment remodeling at the maternal-fetal interface in RSA. Together with independent dataset validation, public single-cell exploratory analysis, and cellular experiments, these findings suggest that the adenosine-NT5E/CD73 axis may be a candidate pathway related to immune imbalance in RSA. However, the current evidence cannot replace validation of NT5E/CD73 protein expression and spatial colocalization of CD68/CD86/CD206 in patient villous tissues. Further studies using larger clinical cohorts, tissue localization, and maternal-fetal interface-related experimental models are required.