Integrative Mendelian Randomization and Whole-Blood Transcriptomic Analysis Implicate a Myeloid–Inflammation Axis in Polycystic Ovary Syndrome

Yiman Fu,* Weihao Li,* Lin Ma Laboratory for Assisted Reproduction and Reproductive Genetics, The Reproductive Medical Center, The Seventh Affiliated Hospital of Sun Yat-Sen University, Shenzhen, Guangdong, People’s Republic of China*These authors contributed equally to this workCorrespondence: Lin Ma, Email [email protected]: Immune dysregulation and low-grade inflammation are central to the pathophysiology of polycystic ovary syndrome (PCOS). However, putative causal relationships inferred from genetic instruments and their consistency with transcriptomic readouts remain underexplored.Methods: We conducted two-sample Mendelian randomization (MR) to estimate potential causal relationships between 731 immune immunophenotypes and PCOS. In an independent, small whole-blood transcriptomic dataset (Gene Expression Omnibus [GEO]: GSE54248; PCOS n=4, controls n=4), we applied gene-signature–based deconvolution (xCell and Microenvironment Cell Populations-counter [MCP-counter]) and single-sample pathway scoring to profile myeloid cell composition and IL-6/JAK/STAT3 and tumor necrosis factor-α/nuclear factor κB (TNF-α/NF-κB) pathway activity. Robustness of the MR findings was examined by standard sensitivity analyses.Results: MR indicated opposite-direction associations for key myeloid traits: higher absolute monocyte count was inversely associated with PCOS risk (odds ratio [OR] per standard deviation [SD] increase 0.52, 95% confidence interval [CI] 0.39– 0.69, P=6.21x10− 6, false discovery rate [FDR] q=0.002), whereas higher CD33+HLA-DR+ absolute count was positively associated with PCOS (OR per SD 1.61, 95% CI 1.27– 2.03, P=6.39x10− 5, FDR q=0.024). In the exploratory GSE54248 dataset, transcriptomic deconvolution suggested myeloid enrichment in PCOS, with tendencies toward higher signatures of monocytes, macrophages (M1/M2), and myeloid dendritic cells (DCs) and higher composite myeloid scores versus controls. Single-sample pathway analyses were consistent with increased IL-6/JAK/STAT3 and TNF-α/NF-κB activities in PCOS, with large effect sizes (Cohen’s d=2.27 and 2.16, respectively) but modest P-values (Wilcoxon P=0.057; Benjamini–Hochberg FDR q=0.143 for both pathways).Conclusion: Genetic and transcriptomic evidence jointly support the concept of a myeloid–inflammation axis in PCOS. The inverse MR association for absolute monocyte count alongside positive risk for CD33+HLA-DR+ compartments, together with cross-cohort evidence of myeloid enrichment and inflammatory pathway activation, is compatible with a potential mechanistic link between immune imbalance and PCOS and provides a hypothesis-generating framework for future therapeutic investigation.Keywords: polycystic ovary syndrome, immune cells, immunophenotypes, Mendelian randomization, inflammation, transcriptomics, myeloid cells