Decreased miR-128-3p in serum exosomes from polycystic ovary syndrome induces ferroptosis in granulosa cells via the p38/JNK/SLC7A11 axis through targeting CSF1

Abstract

Increasing evidence suggests that non-coding small RNAs (miRNAs) carried by exosomes (EXOs) play important roles in the development and treatment of polycystic ovary syndrome (PCOS). In this study, we demonstrate that PCOS mouse serum-derived EXOs promote granulosa cells (GCs) ferroptosis, and induce the occurrence of a PCOS-like phenotype in vivo. Notably, EXO miRNA sequencing combined with in vitro gain- and loss-of-function assays revealed that miR-128-3p, which is absent in the serum-derived EXOs of mice with PCOS, regulates lipid peroxidation and GC sensitivity to ferroptosis inducers. Mechanistically, overexpression of CSF1, a direct target of miR-128-3p, reversed the anti-ferroptotic effect of miR-128-3p. Conversely, ferroptosis induction was mitigated in CSF1-downregulated GCs. Furthermore, we demonstrated that miR-128-3p inhibition activates the p38/JNK pathway via CSF1, leading to NRF2-mediated down-regulation of SLC7A11 transcription, which triggers GC iron overload. Moreover, intrathecal miR-128-3p AgomiR injection into mouse ovaries ameliorated PCOS-like characteristics and restored fertility in letrozole-induced mice. The study reveals the pathological mechanisms of PCOS based on circulating EXOs and provides the first evidence of the roles of miR-128-3p and CSF1 in ovarian GCs. This discovery is expected to provide promising therapeutic targets for the treatment of PCOS.

Fig. 1. PCOS mouse serum and serum-derived exosomes (EXOs) promote granulosa cell (GCs) ferroptosis.

A Cell viability of GCs incubated with control and PCOS mouse serum for 0, 24, 48, or 72 h. *P < 0.05 vs Control serum, # P < 0.05 vs Control serum. B mRNA expression levels of ferroptosis-related genes in GCs as determined using qRT-PCR. C, D Determination of mitochondrial membrane potential (MMP) levels in GCs. Scale bar: 30 μm. E Transmission electron microscopic (TEM) analysis of serum-derived EXOs. Scale bar: 500 nm. F Western blot analysis of the EXO-specific marker proteins, CD9, CD63, and CD81. G EXO particle size as determined through the ZetaView analysis. H Cell viability of GCs incubated with serum EXOs for 0, 24, 48, or 72 h. ***P < 0.001 vs Control serum exo, ****P < 0.0001 vs Control serum exo, # P < 0.05 vs Control serum exo. I Relative mRNA levels of SLC7A11 and GPX4 in serum EXO-treated GCs. J MMP levels in GCs treated with EXOs as determined using JC-1 staining. K Co-culture assay showing that labelled serum EXOs can enter GCs. Scale bar: 200 μm. Data are presented as Mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 2. PCOS mouse serum-derived EXOs induce oestrous disorders and increase ferroptosis in mice.

A Immunofluorescence imaging of ovarian tissues 24 h after serum-derived EXO injection, Scale bar: 100 μm. B, C Measurement of body and ovarian weights in mice in each group. D Steroid sex hormone concentrations in the sera of mice from the different groups as determined via ELISA. E Evaluation of representative oestrous cycles after injection of serum EXOs, “P:” proestrous, “E:” oestrous, “M:” metoestrous, “D:” dioestrus. F, G Corpus luteum and cystic follicle counts following H&E staining. Scale bar: 300 μm. H Ferroptosis-related gene mRNA expression levels. Data are presented as Mean ± SD. Ns P > 0.05, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 3. miRNA expression profiles of serum-derived EXOs from PCOS and control mice.

A Venn diagram showing common and specific miRNA expression in control and PCOS mouse serum-derived EXOs. B Hierarchical clustering analysis showing 18 downregulated and 17 upregulated miRNAs, as well as significantly decreased miR-128-3p expression, in PCOS mouse serum-derived EXOs. C GO and (D) KEGG enrichment analyses of differentially expressed miRNAs. E miRNAs that are co-downregulated and co-upregulated in humans and mice in the database. Serum samples from 250 mice in each group were pooled for EXO isolation. The extracted EXO were subjected to miRNA-seq with three technical replicates performed per group to ensure reliability. Data are presented as Mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001.

Fig. 4. Overexpression of miR-128-3p inhibits GC ferroptosis.

A Analysis of signalling pathways involving miR-128-3p using Diana Tools, the red numbers indicate the number of target genes enriched in this pathway. B Sequence of mature miR-128-3p for different species. C miR-128-3p mimic treatment increased miR-128-3p levels in GCs. D GC cell viability as determined through the CCK-8 assay. E ROS, (F) GSH, (G) Fe²⁺ and (H) MDA levels in GCs treated with erastin and the miR-128-3p mimic. I Measurement of MMP levels in GCs treated with erastin and the miR-128-3p mimic. Scale bar: 30 μm. J Lipid Peroxidation Levels Assessed by Liperfluo Staining. Scale bar: 30 μm. K Low- and high-magnification images obtained by TEM, The red arrows indicate the outer mitochondrial membrane was ruptured and the mitochondrial cristae decreased or disappeared. Scale bar: 1 μm and 500 nm. L Relative mRNA levels for oxidative stress-related genes in GCs. M, N mRNA and protein levels for ferroptosis-related genes. Data are presented as Mean ± SD. Ns P > 0.05, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 5. CSF1 is a direct target gene for miR-128-3p.

A Alignment results for the 3′ untranslated region (3′ UTR) sequence of CSF1 from different species. B, C Dual luciferase reporter assay for miR-128-3p on the CSF1 3′ UTR in mGCs and KGN cells 48 h post transfection with miRNA and a luciferase reporter sequence containing mutant (MUT) CSF1 3′ UTR (pmiRGLO -3′ UTR of CSF1-MUT) or wild type (WT) CSF1 3′ UTR (pmiRGLO-3′ UTR of CSF1-WT). D, E The mRNA and protein levels of CSF1 in GCs were detected after transfection with miR-128-3p mimics. F, G The mRNA and protein levels of CSF1 in GCs were detected after transfection with miR-128-3p inhibitor. H The mRNA expression of CSF1 in ovary was detected by RT-qPCR. I The mRNA expression levels of CSF1 in human and rat databases. Data are presented as Mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 6. CSF1 mediates the effects of miR-128-3p on GC ferroptosis.

A The CCK-8 assay was performed to determine the cell viability of GC. B ROS levels were measured using flow cytometry. C Fe²⁺ content. D MDA levels. E GSH levels. F Determination of MMP levels in GCs through JC-1 staining. Scale bar: 30 μm. G Lipid Peroxidation Levels Assessed by Liperfluo Staining. Scale bar: 30 μm. H Low- and high-magnification images obtained by TEM, The red arrows indicate the outer mitochondrial membrane was ruptured and the mitochondrial cristae decreased or disappeared. Scale bar: 1 μm and 500 nm. I Relative SOD, CAT, and NRF2 mRNA levels. J, K Ferroptosis-related gene mRNA and protein levels in GCs as determined through RT-qPCR and western blotting. L Volcano plot analysis showing that there are 2 downregulated and 70 upregulated lncRNAs in EXO. M The lncRNA-miRNA-mRNA (miR-128-3p) regulatory network. Data are presented as Mean ± SD. Ns P > 0.05, *P<0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 vs (NC mimic + pcNC). # P < 0.05, ## P < 0.01, ### P < 0.001, #### P < 0.0001 vs (miR-128-3p mimic + pcNC).

Fig. 7. Inhibition of miR-128-3p suppresses SLC7A11 transcription through the activation of the P38/JNK pathway.

A, B p38 MAPK/JNK/ERK phosphorylation levels in mouse ovaries. C–F Western blot analysis of p38 MAPK/JNK/ERK phosphorylation levels in GCs transfected with the miR-128-3p inhibitor or mimic. G, H Phosphorylated-p38, p38, SLC7A11, GPX4, and NRF2 protein levels in GCs treated with the miR-128-3p inhibitor and/or SB203580 (p38 inhibitor, 20 μM). I, J Phosphorylated-JNK, JNK, SLC7A11, GPX4, and NRF2 protein levels in GCs treated with the miR-128-3p inhibitor and/or SP600125 (JNK inhibitor, 20 μM). K ChIP-qPCR analysis showing NRF2 binding to the SLC7A11 promoter. Data are presented as Mean ± SD. Ns P > 0.05, *P < 0.05, **P < 0.01, ****P < 0.0001.

Fig. 8. Overexpression of miR-128-3p can restore PCOS-like features in mice.

A Schematic representation of the animal experimentation process in mice. B miR-128-3p expression as determined through qPCR following intrabursal injection of miR-128-3p AgomiR into mouse ovaries. C, D Body and ovary weights of mice in each group. E Representative estrous cycles after ovary-subcutaneous injection of miR-128-3p agomir. F Steroid sex hormone concentrations in mouse serum. G Corpus luteum and antral follicle counts following H&E staining. Scale bar: 300 μm. H, I oxidative stress- and ferroptosis-related gene mRNA expression. J Mouse litter size following intrabursal injection of miR-128-3p AgomiR into mouse ovaries. Data are presented as Mean ± SD. Ns P > 0.05, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.