Methylation reader MBD2-mediated GPX4 transcriptional repression drives ovarian granulosa cell ferroptosis in PCOS

Arrested follicular development and anovulation are hallmarks of polycystic ovary syndrome (PCOS), in which granulosa cell (GC) ferroptosis is emerging as a potential contributor. However, its precise role and regulation remain largely unknown. Here, we identify a methyl-CpG-binding domain protein 2 (MBD2)-driven ferroptotic program as a central pathogenic mechanism in PCOS. In a dehydroepiandrosterone (DHEA)-induced PCOS mouse model, GCs exhibited marked ferroptotic alterations and transcriptional suppression of glutathione peroxidase 4 (GPX4), a key anti-ferroptotic enzyme. GC-specific Gpx4 knockout exacerbated ferroptosis, impaired follicular maturation, reduced corpora lutea formation, and aggravated PCOS pathology. GPX4 repression was associated with increased DNA methyltransferases (DNMTs), elevated DNA Methyl-reading protein MBD2 and hypermethylation of the Gpx4 promoter. Pharmacological inhibition of MBD2 with KCC-07, or DNMT blockade with 5-Azacytidine, restored GPX4 expression, reduced lipid peroxidation and GC ferroptosis, and alleviated ovarian dysfunction. Integrative ATAC-seq and RNA-seq analyses revealed enhanced Gpx4 promoter accessibility in PCOS ovaries, where MBD2, MAZ, HDAC3 and NCoR assembled into a repressive complex that was interrupted by KCC-07 treatment. Importantly, pharmacologic GPX4 inhibition with RSL3 or GC-specific Gpx4 deletion abrogated the protective effects of MBD2 inhibition, establishing GPX4 repression as the critical downstream effector. Collectively, these findings uncover an MBD2-driven epigenetic program that silences GPX4, triggers GC ferroptosis, and promotes PCOS pathogenesis. Targeting MBD2 to restore epigenetic control of ferroptosis offers a promising therapeutic strategy for PCOS.