Identification of Progesterone-Associated Gene Signatures in Men with Alzheimer’s Disease Using Public Transcriptomic Data

Introduction Alzheimer’s disease (AD) exhibits marked molecular heterogeneity, with increasing evidence that sex-specific biological mechanisms influence disease onset and progression. Progesterone is a neuroactive steroid involved in mitochondrial regulation, neuroinflammation, and synaptic function; however, its transcriptional regulation in men with AD remains insufficiently characterized. Objective To identify and quantify progesterone-associated gene signatures in men with AD using public transcriptomic data. Methods Bulk RNA-sequencing data from 412 male subjects were analyzed across three AMP-AD cohorts (ROSMAP, MSBB, and Mayo Clinic). Differential expression analyses were conducted using DESeq2 with adjustment for demographic, technical, and neuropathological covariates. Progesterone-related gene sets were curated from established databases. Co-expression network analysis, pathway enrichment, sex-by-diagnosis interaction testing, and CIBERSORTx-based cell-type deconvolution were performed. Robustness was evaluated through replication, sensitivity analyses, and cross-method validation with edgeR. Results PGRMC1 was consistently downregulated in male AD brains (meta-analytic log2FC = −0.64; 95% CI: −0.89 to −0.39; p = 4.2×10⁻⁶). PGRMC1 expression inversely correlated with Braak stage (ρ = −0.38, p = 1.3×10⁻⁵) and amyloid-β burden (ρ = −0.29, p = 0.0012). PGRMC1-centered co-expression networks contracted by 77.1% in AD, with marked loss of mitochondrial and oxidative phosphorylation gene associations (FDR <10⁻⁷). Steroidogenic enzymes, including HSD3B1, were significantly reduced (log2FC = −0.52, p = 0.0087). Cell-type deconvolution revealed decreased neuronal proportions (−8.3%, p = 0.00034) and increased microglia (+4.7%, p = 0.0012), while cell-adjusted models confirmed persistent PGRMC1 suppression (log2FC = −0.59, p = 0.0051). Sex-stratified analyses identified 18 genes with significant sex-by-diagnosis interactions (FDR <0.05). Conclusion Male AD is characterized by a distinct progesterone-associated transcriptional profile marked by PGRMC1 downregulation and mitochondrial network disruption, supporting progesterone signaling as a biologically relevant, sex-informed therapeutic target.