EIF4A1 and YBX3 as stress granule-related screenability biomarkers in osteoarthritis synovial tissue: an integrated transcriptomic and scRNA-seq study
- 1Department of Orthopedics, Affiliated Changsha Hospital of Hunan Normal University (The Fourth Hospital of Changsha)
- 2Department of Emergency, The Eighth Hospital of Changsha
Abstract
Osteoarthritis (OA) is the most prevalent degenerative joint disease, yet its molecular pathogenesis remains incompletely understood. Stress granules (SGs) are membraneless organelles that regulate mRNA translation and cell fate under cellular stress. However, the role of stress granule-related genes (SGRGs) in OA has not been systematically investigated. OA synovial-tissue transcriptomic datasets (GSE55235, GSE55457, GSE82107) from the Gene Expression Omnibus (GEO) were integrated as the training cohort. GSE12021, an independent OA synovial-tissue dataset, served as the external validation cohort. Differentially expressed genes (DEGs) were identified using limma, and while weighted gene co-expression network analysis (WGCNA) was used to detect modules correlated with OA-related traits. Candidate genes were defined by overlapping DEGs, WGCNA hub genes, and 464 SGRGs, followed by functional enrichment and protein–protein interaction (PPI) analyses. Four machine learning approaches—random forest (RF), eXtreme Gradient Boosting (XGBoost), least absolute shrinkage and selection operator (LASSO), and support vector machine–recursive feature elimination (SVM-RFE)—together with maximal clique centrality (MCC) analysis was employed to construct composite models. Single-cell RNA sequencing data (GSE216651) were further analyzed to determine cell-type-specific expression patterns. In addition, immune infiltration assessment, competitive endogenous RNA (ceRNA) network construction, transcription factor (TF) prediction, and molecular docking analyses were carried out. Finally, quantitative real-time PCR (RT-qPCR) validation was performed using synovial tissue samples from 10 patients with OA and 10 healthy controls. We identified 1165 DEGs and 2857 WGCNA hub genes, yielding 15 candidate SGRGs enriched in Salmonella infection, pentose phosphate, and metabolic pathways. Multi-machine learning composite model identified Eukaryotic Translation Initiation Factor 4A1 (EIF4A1) and Y-box binding protein 3 (YBX3) as key genes with acceptable screening performance (Training set: EIF4A1 = AUC: 0.848, YBX3 = AUC: 0.731; independent validation set: EIF4A1 = AUC: 0.822, YBX3 = AUC: 0.778). Single-cell analysis revealed EIF4A1 downregulation in fibroblasts and YBX3 upregulation in endothelial cells (ECs) during OA. A ceRNA network was constructed, with Forkhead box C1 (FOXC1) and cAMP response element-binding protein 1 (CREB1) identified as shared TFs. Molecular docking confirmed favorable predicted binding interactions for vincristine-EIF4A1 and danazol-YBX3. RT-qPCR validated significant differential expression of both genes in OA tissues. This integrated transcriptomic and single-cell analysis nominates EIF4A1 and YBX3 as SG-related screening biomarker candidates and structurally engageable targets in OA synovial tissue, providing hypotheses for future mechanistic and translational investigation.
Related
- Synovial oxidative stress biomarkers are not associated with radiological severity in knee osteoarthritis: a prospective cross-sectional study.
- Integrated serum and synovial biomarker profiling reveals inflammatory and cartilage remodeling signatures in age-restricted knee osteoarthritis.
- An epithelial–mesenchymal transition-like synovial stromal-remodeling programme associated with knee osteoarthritis pain: Exploratory transcriptomic discovery and spatial analysis
- Identification of candidate selenium-associated transcriptomic biomarkers in chronic obstructive pulmonary disease: an integrated bioinformatics and expression validation study
- Integrated single-cell and bulk transcriptomic analyses reveal UTF1-associated spermatogenic dysfunction and identify ACTRT2 and AKAP4 as biomarkers in non-obstructive azoospermia
BibTeX
@article{he2026eif4a1,
title = {{EIF4A1 and YBX3 as stress granule-related screenability biomarkers in osteoarthritis synovial tissue: an integrated transcriptomic and scRNA-seq study}},
author = {Ning He and Zhipeng Han and Fan Hu and Meijuan Hu},
journal = {Scientific Reports},
year = {2026},
month = aug,
doi = {10.1038/s41598-026-67301-7},
url = {https://doi.org/10.1038/s41598-026-67301-7},
note = {CC BY},
}
RIS
TY - JOUR
TI - EIF4A1 and YBX3 as stress granule-related screenability biomarkers in osteoarthritis synovial tissue: an integrated transcriptomic and scRNA-seq study
AU - Ning He
AU - Zhipeng Han
AU - Fan Hu
AU - Meijuan Hu
PY - 2026
DA - 2026/08/01
JO - Scientific Reports
DO - 10.1038/s41598-026-67301-7
UR - https://doi.org/10.1038/s41598-026-67301-7
AB - Osteoarthritis (OA) is the most prevalent degenerative joint disease, yet its molecular pathogenesis remains incompletely understood. Stress granules (SGs) are membraneless organelles that regulate mRNA translation and cell fate under cellular stress. However, the role of stress granule-related genes (SGRGs) in OA has not been systematically investigated. OA synovial-tissue transcriptomic datasets (GSE55235, GSE55457, GSE82107) from the Gene Expression Omnibus (GEO) were integrated as the training cohort. GSE12021, an independent OA synovial-tissue dataset, served as the external validation cohort. Differentially expressed genes (DEGs) were identified using limma, and while weighted gene co-expression network analysis (WGCNA) was used to detect modules correlated with OA-related traits. Candidate genes were defined by overlapping DEGs, WGCNA hub genes, and 464 SGRGs, followed by functional enrichment and protein–protein interaction (PPI) analyses. Four machine learning approaches—random forest (RF), eXtreme Gradient Boosting (XGBoost), least absolute shrinkage and selection operator (LASSO), and support vector machine–recursive feature elimination (SVM-RFE)—together with maximal clique centrality (MCC) analysis was employed to construct composite models. Single-cell RNA sequencing data (GSE216651) were further analyzed to determine cell-type-specific expression patterns. In addition, immune infiltration assessment, competitive endogenous RNA (ceRNA) network construction, transcription factor (TF) prediction, and molecular docking analyses were carried out. Finally, quantitative real-time PCR (RT-qPCR) validation was performed using synovial tissue samples from 10 patients with OA and 10 healthy controls. We identified 1165 DEGs and 2857 WGCNA hub genes, yielding 15 candidate SGRGs enriched in Salmonella infection, pentose phosphate, and metabolic pathways. Multi-machine learning composite model identified Eukaryotic Translation Initiation Factor 4A1 (EIF4A1) and Y-box binding protein 3 (YBX3) as key genes with acceptable screening performance (Training set: EIF4A1 = AUC: 0.848, YBX3 = AUC: 0.731; independent validation set: EIF4A1 = AUC: 0.822, YBX3 = AUC: 0.778). Single-cell analysis revealed EIF4A1 downregulation in fibroblasts and YBX3 upregulation in endothelial cells (ECs) during OA. A ceRNA network was constructed, with Forkhead box C1 (FOXC1) and cAMP response element-binding protein 1 (CREB1) identified as shared TFs. Molecular docking confirmed favorable predicted binding interactions for vincristine-EIF4A1 and danazol-YBX3. RT-qPCR validated significant differential expression of both genes in OA tissues. This integrated transcriptomic and single-cell analysis nominates EIF4A1 and YBX3 as SG-related screening biomarker candidates and structurally engageable targets in OA synovial tissue, providing hypotheses for future mechanistic and translational investigation.
SN - 2045-2322
ER -