Integrated transcriptomics identifies immune-metabolic dysregulation and candidate diagnostic biomarkers in oligoasthenozoospermia with experimental validation.

Objective: Oligoasthenozoospermia is a major cause of male infertility and is characterized by reduced sperm concentration and motility. However, candidate molecular biomarkers and integrated mechanistic frameworks for disease characterization remain limited. This study aimed to identify candidate diagnostic biomarkers for oligoasthenozoospermia and to characterize the immune-metabolic dysregulation associated with impaired spermatogenesis. Methods: This study integrated gene expression profiles from two Gene Expression Omnibus (GEO) datasets, GSE45887 and GSE45885, to analyze transcriptomic alterations in oligoasthenozoospermia. After data normalization and batch correction, differential expression analysis and weighted gene co-expression network analysis (WGCNA) were performed to identify disease-related genes. Functional enrichment was further evaluated using Gene Set Variation Analysis (GSVA) and Gene Set Enrichment Analysis (GSEA). Immune-related transcriptomic signature variation was estimated by single-sample gene set enrichment analysis (ssGSEA). To identify candidate biomarkers, three machine learning algorithms-least absolute shrinkage and selection operator (LASSO), random forest, and support vector machine-recursive feature elimination (SVM-RFE)-were applied to screen core genes, followed by construction and internal evaluation of logistic regression, LASSO, random forest, and support vector machine (SVM) diagnostic models. Finally, a smoking- and ethanol-induced mouse model of oligoasthenozoospermia was established, and sperm quality, histopathology, transcriptomic alterations, immunofluorescence, and Western blotting were used for experimental validation. Results: Differential expression analysis identified reproducible transcriptomic alterations in oligoasthenozoospermia, and intersection with WGCNA-derived key module genes yielded 86 candidate genes. Functional enrichment analysis showed that these genes were mainly associated with immune- and metabolism-related pathways. GSVA and GSEA demonstrated coordinated activation of complement, IL6-JAK-STAT3, interferon-γ, and PI3K-AKT-mTOR/mTORC1 signaling, together with marked suppression of spermatogenesis-related programs. Immune signature analysis based on ssGSEA suggested potential immune microenvironment alterations at the transcriptomic level and correlations between key genes and multiple immune-related signatures. Further analysis using three machine learning algorithms identified 12 core genes. Among the tested classifiers, the random forest model showed the best overall performance in internal validation; however, the near-perfect performance observed in several models should be interpreted cautiously given the limited sample size. In the mouse model, sperm concentration, viability, and motility were significantly decreased, whereas the sperm abnormality rate was significantly increased, accompanied by abnormal testicular histology and reduced PCNA expression. Transcriptomic and protein-level validation further supported dysregulation of representative candidate genes and pathways, supporting the biological relevance of the human transcriptomic findings rather than providing exhaustive mechanistic validation. Conclusion: This study identified a 12-gene candidate biomarker panel for oligoasthenozoospermia and revealed a coordinated immune-metabolic-spermatogenic dysregulation pattern. These findings provide a transcriptomic framework for molecular characterization of oligoasthenozoospermia and preliminary evidence for future biomarker-based diagnostic development, which requires external validation in larger independent and more clinically homogeneous human cohorts.