Identification and validation of biomarkers associated with glycolysis in polycystic ovarian syndrome

Polycystic ovary syndrome (PCOS), an endocrine disorder emerging in adolescence and reproductive years, has been linked to glycolysis in prior studies, though the precise mechanistic role of glycolysis in its pathogenesis remains unclear. Therefore, this study sought to identify glycolysis-related biomarkers in PCOS and elucidate their regulatory mechanisms to provide novel therapeutic strategies. Utilizing publicly available datasets, biomarkers were identified via differential analysis, various PPI algorithms, and validation of expression patterns. Subsequent analyses included functional enrichment, tissue and cell-specific expression profiling, m6A modification site prediction, compound screening, molecular network construction, and molecular docking. RT-qPCR was performed on clinical samples for experimental validation. Two biomarkers, TXNIP and TGFBI, were identified and jointly enriched in "complement and coagulation cascades". TXNIP showed elevated expression in tongue and endocrine cells, whereas TGFBI was highly expressed in placental and adipocyte tissues. TGFBI had 14 high-confidence m6A modification sites and TXNIP had 1 high-confidence m6A modification site. The identified regulatory networks included hsa-miR-6761-5p-TXNIP-PPARG and hsa-miR-6761-5p-TGFBI-RB1. Four key compounds-acetaminophen, bisphenol A, tetrachlorodibenzodioxin, and valproic acid-were prioritized, with molecular docking revealing strongest binding affinities between bisphenol A and both biomarkers (TXNIP: -5.9 kcal/mol; TGFBI: -13.1 kcal/mol). RT-qPCR validation in granulosa cells from PCOS patients confirmed significant upregulation of TGFBI and TXNIP, aligning with bioinformatics predictions. These findings suggest that TXNIP and TGFBI may serve as potential biomarkers associated with glycolytic dysregulation in PCOS, offering insights into the interplay between metabolic dysfunction and disease mechanisms.