Molecular subtype of recurrent implantation failure reveals distinct endometrial etiology of female infertility
Abstract Background Recurrent implantation failure (RIF) remains a significant barrier in assisted reproductive technology (ART), where multiple transfers of high-quality embryos fail to achieve pregnancy. While embryo-related factors have been extensively investigated, the contribution of endometri...
Saved in:
| Main Authors: | , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
BMC
2025-07-01
|
| Series: | Journal of Translational Medicine |
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s12967-025-06771-1 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Abstract Background Recurrent implantation failure (RIF) remains a significant barrier in assisted reproductive technology (ART), where multiple transfers of high-quality embryos fail to achieve pregnancy. While embryo-related factors have been extensively investigated, the contribution of endometrial dysfunction to RIF remains poorly characterized. This study aimed to determine whether biologically distinct molecular subtypes of endometrial dysfunction exist in RIF and whether such subtypes could guide more personalized and effective treatment strategies. Methods We conducted a comprehensive computational analysis integrating publicly available endometrial transcriptomic datasets with prospectively collected samples. Multi-platform data were harmonized using a random-effects model. Differentially expressed genes (DEGs) between RIF and normal samples were identified using MetaDE. Clinical and hormonal correlations were used to assess heterogeneity among RIF samples. Unsupervised clustering (ConsensusClusterPlus) identified RIF subtypes, and their biological characteristics were analyzed using Gene Set Enrichment Analysis (GSEA). Immunohistochemistry (IHC) was used to evaluate the protein-level expression of selected subtype-associated genes. A molecular classifier (MetaRIF) was developed using the optimal F-score from 64 combinations of machine learning algorithms. Candidate therapeutic compounds were predicted using the Connectivity Map (CMap) database. Results A total of 1,776 robust DEGs were identified between RIF and normal samples. Clustering analysis revealed two reproducible RIF subtypes: an immune-driven subtype (RIF-I) and a metabolic-driven subtype (RIF-M). RIF-I was enriched for immune and inflammatory pathways (e.g., IL-17 and TNF signaling, p < 0.01) and showed increased infiltration of effector immune cells. RIF-M was characterized by dysregulation of oxidative phosphorylation, fatty acid metabolism, steroid hormone biosynthesis, and altered expression of the circadian clock gene PER1. Immunohistochemical analysis showed that the T-bet/GATA3 expression ratio mirrored the expected subtype distribution, with higher values in RIF-I and lower values in RIF-M. The MetaRIF classifier accurately distinguished subtypes in independent validation cohorts (AUC: 0.94 and 0.85) and outperformed previously published models (AUC: MetaRIF = 0.88; koot_sig = 0.48; Wang_sig = 0.54; OSR_score = 0.72). CMap-based drug predictions identified sirolimus as a candidate for RIF-I and prostaglandins for RIF-M. Conclusions Our findings reveal two biologically distinct endometrial subtypes of RIF, highlighting the heterogeneous nature of its pathogenesis. By addressing immune and metabolic dysregulation through subtype-specific approaches, our findings provide a foundation for improving diagnosis and tailoring treatment in RIF, potentially enhancing implantation outcomes in ART. |
|---|---|
| ISSN: | 1479-5876 |