Endocrine Abstracts

JOINT1489

Polycystic ovary syndrome (PCOS) is a common endocrine disorder among females characterized by hyperandrogenism, polycystic ovarian morphology, menstrual irregularities, and metabolic dysfunctions such as hyperinsulinemia and an increased risk of type 2 diabetes. Immune involvement in PCOS is indicated by a higher incidence of asthma, higher prevalence of autoimmune thyroid disease and respiratory infections. Although PCOS presents with chronic inflammation in peripheral blood, it remains poorly understood how metabolic and reproductive abnormalities impacts the inflammatory and immune profile in PCOS. To fill this knowledge gap, we profiled 92 immune and inflammatory serum proteins in 106 PCOS patients and 58 controls using the Olink Target Inflammation panel. In addition, general patient information (e.g., age) and levels of markers related to metabolic (e.g., BMI, insulin resistance) and reproductive function (e.g., sex hormone-binding globulin, testosterone) were measured and accounted for in our analysis. After adjusting for age and batch effects, our statistical model shows that several proteins are differentially expressed between PCOS and controls. We observe higher concentrations of pro-inflammatory (IL18R1, IL18, and IL1A) and immune exhaustion (PDL1) proteins, which confirm with the chronic inflammation state in PCOS. Women with PCOS have a clear different metabolic and reproductive profiles compared to controls, including higher level of HOMA-IR, triglycerides, and circulating androgens. Principal component (PC) analysis reveals metabolic-driven PC1 and reproductive-driven PC2, indicating that the metabolic profile and reproductive profile contribute differently to inter-individual differences. We computed metabolic and reproductive scores based the sum of z-score transformed clinical variable values for each individual. Interestingly, metabolic and reproductive scores are positively correlated in controls while the correlation disappeared in PCOS, which suggests that reproductive and metabolic relation is disrupted in PCOS. Overall, the levels of differentially expressed immune proteins in PCOS positively correlate with metabolic measurements, while only a few proteins were found to correlate with reproductive measurements. Interestingly, although PCOS patients have a skewed inflammatory profile and higher testosterone concentrations, we identified that IL6, S100A12, and IL20 negatively correlate with testosterone levels in PCOS. Finally, k-means clustering of the PCOS immune proteome revealed two immune subgroups of PCOS, with subgroup A showing a more pro-inflammatory immune profile and a more severe metabolic profile compared to subgroup B. Our study uncovers novel immune proteome alterations in PCOS and the metabolic and reproductive regulation of such proteins. This has important implications for understanding PCOS pathogenesis and identifying pathways for immunomodulatory intervention.