Upregulation of PLAU in granulosa cells disrupts steroid hormone synthesis and promotes apoptosis by activating NF-κB signaling pathway in PCOS.

Table  1 demonstrates the characteristics of the patients included in this study, which shows that 20 patients were enrolled in the PCOS group based on Rotterdam criteria and 20 patients were included in the control group. There were no statistically significant differences in age, estradiol, Prolactin (PRL), and Progesterone (P). In PCOS group, Follicle Stimulating Hormone (FSH) is lower than control group. However, Body Mass Index (BMI), Luteinizing Hormone (LH), LH/FSH, Testosterone (T), Anti-Mullerian Hormone (AMH), and Antral Follicle Counting (AFC) at basal state were significantly higher in the PCOS group compared to the control group. Table 1 Clinical characteristics in women with and without PCOS Control ( n  = 20) PCOS ( n  = 20) P -value Age (year) 30.3 ± 1.0 28.15 ± 0.81 0.1023 Body Mass Index, BMI 20.79 ± 0.45 23.08 ± 0.60 0.0041** Follicle-Stimulating Hormone, FSH (mIU/ml) 5.79 ± 0.31 4.75 ± 0.24 0.0117* Luteinizing Hormone, LH (mIU/ml) 4.76 ± 0.46 8.18 ± 1.23 0.0127* LH/FSH 0.85 ± 0.08 1.75 ± 0.27 0.0026** Estradiol, E 2 (ng/L) 31.71 ± 3.08 35.96 ± 4.1 0.4124 Prolactin, PRL (ng/ml) 19.21 ± 2.78 19.61 ± 2.67 0.9189 Progesterone, P (ng/ml) 0.24 ± 0.04 0.29 ± 0.03 0.3709 Testosterone, T (ng/ml) 0.32 ± 0.02 0.77 ± 0.05 <0.0001**** Anti-Mullerian Hormone, AMH (ng/ml) 3.78 ± 0.74 8.14 ± 1.0 0.0012** Antral Follicle Counting, AFC 13.65 ± 1.04 23.6 ± 1.53 <0.0001****  All data are expressed as the mean ± S.E.M. Data were analyzed by two-tailed Student’s t-test BMI Body Mass Index, FSH Follicle-Stimulating Hormone, LH Luteinizing Hormone, E 2 Estradiol, PRL Prolactin, P Progesterone, AMH Anti-Mullerian Hormone, AFC Antral Follicle Counting * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001 Clinical characteristics in women with and without PCOS All data are expressed as the mean ± S.E.M. Data were analyzed by two-tailed Student’s t-test BMI Body Mass Index, FSH Follicle-Stimulating Hormone, LH Luteinizing Hormone, E 2 Estradiol, PRL Prolactin, P Progesterone, AMH Anti-Mullerian Hormone, AFC Antral Follicle Counting * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001 To quantify the principal components of mRNA for all samples, Principal component analysis (PCA) was performed, and the aggregation between samples indicates the degree of variability and different mRNA expression levels among samples (Fig.  1 a). The DEGs are presented by a volcano plot as shown in Fig.  1 b, a total of 1734 DEGs were found between the two groups. Of these, 1545 genes were upregulated, and 189 genes were downregulated in the PCOS group. To further understanding the DEGs, we conducted functional analyses. In terms of GO analysis, notably enriched biological processes included positive regulation of cytokine production, leukocyte migration and regulation of immune effector process. In the analysis of cellular components, DEGs were observed in secretory granule membrane, specific granule and endocytic vesicle, and DEGs were particularly enriched in GTPase regulator activity, nucleoside-triphosphatase regulator activity and phospholipid blinding based on the molecular function analysis (Fig.  1 c). As for the enrichment analysis of KEGG, DEGs were mainly concentrated in the pathways such as chemokine signaling pathway, osteoclast differentiation, B cell receptor signaling pathway, leukocyte transendothelial migration, tuberculosis and NF-kappa B signaling pathway (Fig.  1 d). Fig. 1 Analyses and identification of DEGs in human ovarian granulosa cells in PCOS and normal control groups. a Presentation of principal component analysis results for human samples. b Volcano plot of differential expression genes. The thresholds for differentially expressed genes (DEGs) analysis were set at P  < 0.05 and log2 FC ≥ 1. c DEGs were represented by dot plots displaying GO enrichment. d DEGs were represented by bar plots displaying KEGG enrichment Analyses and identification of DEGs in human ovarian granulosa cells in PCOS and normal control groups. a Presentation of principal component analysis results for human samples. b Volcano plot of differential expression genes. The thresholds for differentially expressed genes (DEGs) analysis were set at P  < 0.05 and log2 FC ≥ 1. c DEGs were represented by dot plots displaying GO enrichment. d DEGs were represented by bar plots displaying KEGG enrichment To reveal the difference gene expression level in control and PCOS group, we constructed a mice model of PCOS that induced by dehydroepiandrosterone (DHEA) according to previous study [ 20 ]. Figure 2 a illustrates the process of animal experimentation in this study, and there is no significant difference in initial body weight. After DHEA administration for 28 days, a statistically significant difference in body weight is observed (Fig. 2 b). Glucose tolerance test (GTT) and insulin tolerance test (ITT) experiments show a significant difference between the control and PCOS groups, and the blood glucose in PCOS group is always higher than the control group (Fig. 2 c, d). In Fig. 2 e, Hematoxylin and Eosin (H&E) stained ovarian tissue showed more corpus luteum and growing follicles in the control group. In contrast, there were almost no corpus luteum in the ovary of mice with polycystic ovary syndrome. Additionally, we observed the estrous cycle of mice for 16 consecutive days, and the results show that comparing to the control group, the estrous cycle of mice with PCOS is abnormal, which was predominantly in oestrus, and less proestrus/metestrus was observed. These results suggested that the PCOS mice well mimic the PCOS phenotype of obesity, glucose abnormalities, insulin tolerance, and ovulation abnormalities. Fig. 2 Phenotypic evaluation of normal control and dehydroepiandrosterone (DHEA)-induced mice. a Mice were divided into two cohorts (8 mice/cohort): sham control mice received corn oil; DHEA cohorts were injected with DHEA subcutaneously for 28 days. b Comparison of body weight changes before and after DHEA administration in mice between two groups. c GTT. d ITT. e Hematoxylin and eosin (H&E) stain of representative ovaries. #cystic follicle; *corpora lutea. f , g , h , i Quantitative analysis of estrous cycles. P, proestrus; E, estrus; M/D, metestrus/diestrus. Data are represented as mean ± Standard Error of the Mean (SEM). The SEM reflects the variability of the sampling distribution of the sample mean. A smaller value of SEM indicates that the sample mean is more likely to be close to the population mean, thereby enhancing the reliability and precision of the estimate. p-values were determined by unpaired two-tailed Student’s t-test. * p  < 0.05; ** p  < 0.01; *** p  < 0.001; **** p  < 0.0001 Phenotypic evaluation of normal control and dehydroepiandrosterone (DHEA)-induced mice. a Mice were divided into two cohorts (8 mice/cohort): sham control mice received corn oil; DHEA cohorts were injected with DHEA subcutaneously for 28 days. b Comparison of body weight changes before and after DHEA administration in mice between two groups. c GTT. d ITT. e Hematoxylin and eosin (H&E) stain of representative ovaries. #cystic follicle; *corpora lutea. f , g , h , i Quantitative analysis of estrous cycles. P, proestrus; E, estrus; M/D, metestrus/diestrus. Data are represented as mean ± Standard Error of the Mean (SEM). The SEM reflects the variability of the sampling distribution of the sample mean. A smaller value of SEM indicates that the sample mean is more likely to be close to the population mean, thereby enhancing the reliability and precision of the estimate. p-values were determined by unpaired two-tailed Student’s t-test. * p  < 0.05; ** p  < 0.01; *** p  < 0.001; **** p  < 0.0001 Principal component analysis (PCA) based on mouse ovary was shown in Fig.  3 a, the samples in the control group and the samples in the PCOS group are far apart, which indicating greater variability between the two groups. However, clusterization of samples within the control group and the PCOS group showed less variability of samples within the group. Then, we performed a volcano plot to present the DEGs as shown in Figs.  3 b and 987 genes were identified between the two groups, including 522 upregulated and 465 downregulated genes in the PCOS group. GO and KEGG enrichment analyses were employed to uncover the trends in functional enrichment of DEGs. DEGs were enriched in regulation of membrane potential, muscle system process and calcium ion transport in the analysis of biological process. In terms of cellular components, the GO analysis revealed an abundance of DEGs in apical plasma membrane, transmembrane transporter complex and transporter complex. Metal ion transmembrane transporter activity, icon channel activity and cation channel activity were enriched in the analysis of molecular functions (Fig.  3 c). The KEGG functional enrichment analysis was shown in Fig.  3 d, which highlighted associations of DEGs with neuroactive ligand-receptor interaction, calcium signaling pathway, steroid hormone biosynthesis, cAMP signaling pathway, metabolism of xenobiotics by cytochrome P450 and primary bile acid biosynthesis. To gain a deeper understanding of the DEGs enriched in steroid hormone biosynthesis, we performed protein–protein interaction (PPI) network. 15 DEGs were labeled as nodes in the figure, with red and blue indicating upregulated and downregulated expression, respectively (Fig.  3 e). Fig. 3 Analyses and identification of DEGs in mouse ovary in PCOS and normal control groups. a Presentation of principal component analysis results for all samples. b Volcano plot of differential gene expression. P  < 0.05 and log2 FC ≥ 1 were set for differentially expressed genes (DEGs) analysis. c DEGs were represented by dot plots displaying GO enrichment. d DEGs were represented by bar plots displaying KEGG enrichment. e PPI network of DEGs in steroid hormone synthesis pathway Analyses and identification of DEGs in mouse ovary in PCOS and normal control groups. a Presentation of principal component analysis results for all samples. b Volcano plot of differential gene expression. P  < 0.05 and log2 FC ≥ 1 were set for differentially expressed genes (DEGs) analysis. c DEGs were represented by dot plots displaying GO enrichment. d DEGs were represented by bar plots displaying KEGG enrichment. e PPI network of DEGs in steroid hormone synthesis pathway Figure  4 a revealed the PCA of RNA-sequencing analysis based on mouse granulosa cells, indicating a greater variability of samples in the control and PCOS groups. As shown in Fig.  4 b, a total of 389 DEGs were identified between the PCOS and control groups, with 210 upregulated genes and 179 downregulated genes in the PCOS group. Based on these DEGs, we performed GO analysis, which revealed DEGs were enriched in biological processes including ossification, cell junction assembly and renal system development. The analysis of cellular components found that the DEGs were mainly enriched in collagen-containing extracellular matrix, receptor complex and collagen trimer. As for molecular function, actin binding, glycosaminoglycan binding and extracellular matrix structural constituent were enriched. KEGG pathway enrichment analysis demonstrated that DEGs were principally concentrated in processes such as cytoskeleton in muscle cells, AGE-RAGE signaling pathway in diabetic complications, focal adhesion, amoebiasis, ECM-receptor interaction, cAMP signaling pathway, TNF signaling pathway, proteoglycans in cancer, regulation of lipolysis in adipocytes, TGF-beta signaling pathway, Ras signaling pathway, Type II diabetes mellitus and platelet activation (Fig.  4 d). Fig. 4 Analyses and identification of DEGs in mouse ovarian granulosa cells in PCOS and normal control groups. a Presentation of principal component analysis results for all samples. b Volcano plot of differential gene expression. P  < 0.05 was defined as significant and the threshold was set as log2 FC ≥ 0.602. c DEGs were represented by dot plots displaying GO enrichment. d DEGs were represented by bar plots displaying KEGG enrichment Analyses and identification of DEGs in mouse ovarian granulosa cells in PCOS and normal control groups. a Presentation of principal component analysis results for all samples. b Volcano plot of differential gene expression. P  < 0.05 was defined as significant and the threshold was set as log2 FC ≥ 0.602. c DEGs were represented by dot plots displaying GO enrichment. d DEGs were represented by bar plots displaying KEGG enrichment Venn diagram was used to identify critical genes in the control and PCOS groups among DEGs from human granulosa cells, DEGs from mouse ovary, DEGs from mouse granulosa cells. Seven genes were found out as shown in Fig.  5 a, and they are Cadherin Related 23 (CDH23), Calmin (CLMN), RAS And EF-Hand Domain Containing (RASEF), ATP Binding Cassette Subfamily B Member 4 (ABCB4), Plasminogen Activator, Urokinase (PLAU), STEAP4 Metalloreductase (STEAP4) and Cytochrome P450 Family 2 Subfamily S Member 1 (CYP2S1). Table  1 demonstrates the 7 differentially expressed genes in granulosa cells of women with PCOS. Table 2 The 7 differentially expressed genes (DEGs) in human Gene Name Offical Full Name Log2FoldChange p value significance CDH23 Cadherin Related 23 2.49085479 0.0000228 up CLMN Calmin 2.074105286 0.000107203 up RASEF RAS And EF-Hand Domain Containing 2.533883065 0.000121495 up ABCB4 ATP Binding Cassette Subfamily B Member 4 1.860608158 0.000767763 up PLAU Plasminogen Activator, Urokinase 1.722773989 0.001153793 up STEAP4 STEAP4 Metalloreductase 2.832150361 0.001672879 up CYP2S1 Cytochrome P450 Family 2 Subfamily S Member 1 1.97145957 0.003173471 up The 7 differentially expressed genes (DEGs) in human Fig. 5 GSEA for the single critical gene. a Critical genes were identified by venn diagram. b GSEA analysis for CDH23. c GSEA analysis for CLMN. d GSEA analysis for RASEF. e GSEA analysis for ABCB4. f GSEA analysis for PLAU. g GSEA analysis for STEAP4. h GSEA analysis for CYP2S1. Normalized enrichment score (NES) is calculated by normalizing the enrichment score (ES) for gene set size and is used to assess the significance of gene set enrichment. A positive NES indicates that the gene set is enriched at the top of the ranked list, indicating the upregulation of enriched gene set. A negative NES indicates enrichment at the bottom, indicating the downregulation of enriched gene set. Larger absolute values of NES indicate higher enrichment of the gene set and greater biological relevance GSEA for the single critical gene. a Critical genes were identified by venn diagram. b GSEA analysis for CDH23. c GSEA analysis for CLMN. d GSEA analysis for RASEF. e GSEA analysis for ABCB4. f GSEA analysis for PLAU. g GSEA analysis for STEAP4. h GSEA analysis for CYP2S1. Normalized enrichment score (NES) is calculated by normalizing the enrichment score (ES) for gene set size and is used to assess the significance of gene set enrichment. A positive NES indicates that the gene set is enriched at the top of the ranked list, indicating the upregulation of enriched gene set. A negative NES indicates enrichment at the bottom, indicating the downregulation of enriched gene set. Larger absolute values of NES indicate higher enrichment of the gene set and greater biological relevance To further understand the functions of these seven genes, we performed GSEA pathway enrichment analysis. CDH23 demonstrated positive correlations with hematopoietic cell lineage, intestinal immune network for IgA production, primary immunodeficiency, Th1 and Th2 cell differentiation, viral protein interaction with cytokine and cytokine receptor (Fig.  5 b). CLMN is positively correlated with chemokine signaling pathway, cytokine-cytokine receptor interaction, hematopoietic cell lineage and tuberculosis, but negative with ribosome (Fig.  5 c). RASEF demonstrated positive correlations with herpes simplex virus 1 infection and olfactory transduction, but negative with oxidative phosphorylation, ribosome and systemic lupus erythematosus (Fig.  5 d). The GSEA analysis of ABCB4 shows that ABCB4 has positive correlations with asthma, herpes simplex virus 1 infection, influenza A, phagosome and rheumatoid arthritis (Fig.  5 e). As for PLAU, as a plasminogen activator, which is positively associate with hematopoietic cell lineage, lysosome, phagosome, rheumatoid arthritis and tuberculosis (Fig.  5 f). Besides, STEAP4 demonstrated positive correlations with cytokine-cytokine receptor interaction, hematopoietic cell lineage, natural killer cell mediated cytotoxicity, NOD-like receptor signaling pathway and tuberculosis (Fig.  5 g). In addition, CYP2S1 demonstrated positive correlations with leishmaniasis, lysosome, phagosome, rheumatoid arthritis and tuberculosis (Fig.  5 h). Considering the possible inaccuracies of RNA-sequencing, we performed RT-qPCR to validate the mRNA expression of critical genes in mice. GAPDH was used as the reference gene for normalization because GAPDH is a housekeeping gene and stably expressed across different experimental conditions and tissue types [ 21 ]. Figure 6 a shows the expression of critical genes in mouse ovarian tissue, and Fig. 6 b shows the expression of critical genes in mouse granulosa cells. Fig. 6 PLAU is upregulated in granulosa cells of PCOS mice. a Validation of critical genes by RT-qPCR in mouse ovary. b Validation of critical genes by RT-qPCR in mouse ovarian granulosa cells. c Western bolt presented PLAU is upregulated in mouse ovarian granulosa cells of PCOS. d Immunofluorescence with FSHR antibody and PLAU antibody evaluating PLAU expression in mouse granulosa cells. e Immunohistochemistry with PLAU in mouse ovary f Immunofluorescence with PLAU in mouse ovary. Data are represented as mean ± Standard Error of the Mean (SEM). The SEM reflects the variability of the sampling distribution of the sample mean. A smaller value of SEM indicates that the sample mean is more likely to be close to the population mean, thereby enhancing the reliability and precision of the estimate. p-values were determined by unpaired two-tailed Student’s t-test. * p  < 0.05; ** p  < 0.01; *** p  < 0.001; **** p  < 0.0001 PLAU is upregulated in granulosa cells of PCOS mice. a Validation of critical genes by RT-qPCR in mouse ovary. b Validation of critical genes by RT-qPCR in mouse ovarian granulosa cells. c Western bolt presented PLAU is upregulated in mouse ovarian granulosa cells of PCOS. d Immunofluorescence with FSHR antibody and PLAU antibody evaluating PLAU expression in mouse granulosa cells. e Immunohistochemistry with PLAU in mouse ovary f Immunofluorescence with PLAU in mouse ovary. Data are represented as mean ± Standard Error of the Mean (SEM). The SEM reflects the variability of the sampling distribution of the sample mean. A smaller value of SEM indicates that the sample mean is more likely to be close to the population mean, thereby enhancing the reliability and precision of the estimate. p-values were determined by unpaired two-tailed Student’s t-test. * p  < 0.05; ** p  < 0.01; *** p  < 0.001; **** p  < 0.0001 RT-qPCR shows that the mRNA expression of PLAU is upregulated in PCOS mice, which is consistent with the results of RNA-sequencing. Subsequently, we used WB to further identify the protein expression of PLAU in mouse granulosa cells, which found PLAU is upregulated in PCOS group (Fig.  6 c, d). Then, immunofluorescence (IF) analysis was engaged to investigate the expression pattern of PLAU at the protein level using mouse ovary. As shown in Fig.  6 e, PLAU was strongly expressed in GCs, whereas it was weakly expressed in oocytes and pericytes within mouse ovaries and showed a trend toward increased expression in the PCOS group than the control group. Besides, we performed immunohistochemistry (IHC) to further confirm the localization of PLAU in mouse ovary (Fig.  6 f), which is in line with the results of IF that the PLAU is expressed in granulosa cells of mouse ovary. FSHR is a marker for granulosa cells, so we performed IF assay to confirm the cell types and to explore the localization of PLAU, PLAUR and CYP19A1 in cultured mouse granulosa cells (Fig.  7 ). Summarizing all these results, we concluded that PLAU is upregulated in granulosa cells of PCOS mice. Fig. 7 IF demonstrated the protein expression location of FSHR, PLAU, PLAUR and CYP19A1 in mouse ovarian granulosa cells. Scale bar = 50 μm IF demonstrated the protein expression location of FSHR, PLAU, PLAUR and CYP19A1 in mouse ovarian granulosa cells. Scale bar = 50 μm To investigate the potential role of PLAU in the pathogenesis of PCOS, we designed three siRNAs targeting PLAU and selected the siRNA with the highest interference efficiency for subsequent experiments (Fig. 8 a). As shown in Fig. 8 b, knockdown of PLAU in KGN cells resulted in increased mRNA expression of PLAR, while the mRNA expression of CYP11A1, CYP19A1 and STAR decreased (Fig. 8 b). To assess protein expression, we conducted Western blot analysis, which revealed a decrease in CYP19A1 levels following PLAU knockdown (Fig. 9 a). Conversely, the protein expression of PLAUR increased, consistent with the qPCR results (Fig. 9 b). Previous studies demonstrated that the activation of NF-κB signaling pathway promotes apoptosis in granulosa cells [ 22 ]. Combining literature search and biological analysis of KEGG enrichment, we hypothesized that PLAU may promote apoptosis through NF-κB signaling pathway. Western blot analysis further explored the relationship between PLAU and the NF-κB signaling pathway, showing decreased levels of p65, p-p65, and p-IKB in KGN cells after PLAU knockdown. Additionally, we observed a reduction in apoptosis of granulosa cells following PLAU silencing, with decreased mRNA and protein expression levels of Bax, Bax/Bcl2 and cleaved caspase 3 (Fig. 9 c, d, e). Fig. 8 PLAU interferes with steroid hormone synthesis and promotes apoptosis. a Validation of PLAU transfection efficiency. b The effects of PLAU knockdown on steroid hormone synthesis and apoptosis. Data are represented as mean ± Standard Error of the Mean (SEM). The SEM reflects the variability of the sampling distribution of the sample mean. A smaller value of SEM indicates that the sample mean is more likely to be close to the population mean, thereby enhancing the reliability and precision of the estimate. p-values were determined by unpaired two-tailed Student’s t-test. * p  < 0.05; ** p  < 0.01; *** p  < 0.001; **** p  < 0.0001 PLAU interferes with steroid hormone synthesis and promotes apoptosis. a Validation of PLAU transfection efficiency. b The effects of PLAU knockdown on steroid hormone synthesis and apoptosis. Data are represented as mean ± Standard Error of the Mean (SEM). The SEM reflects the variability of the sampling distribution of the sample mean. A smaller value of SEM indicates that the sample mean is more likely to be close to the population mean, thereby enhancing the reliability and precision of the estimate. p-values were determined by unpaired two-tailed Student’s t-test. * p  < 0.05; ** p  < 0.01; *** p  < 0.001; **** p  < 0.0001 Fig. 9 Western blot showed that CYP19A1 and NF-κB signaling pathway related gene and apoptosis related genes (Bax, Bcl2 and cleaved caspase 3) levels in KGN cells after the knockdown of PLAU. a The expression of CYP19A1 was decreased in KGN cells after the knockdown of PLAU. b The expression of PLAUR was increased in KGN cells after the knockdown of PLAU. c Western bolt presented the expression of p65, p-p65, Bcl2 and Bax in KGN cells after the knockdown of PLAU. d The expression of p-IKB was decreased in KGN cells after the knockdown of PLAU. e The expression of cleaved caspase 3 was decreased in KGN cells after the knockdown of PLAU. Data are represented as mean ± Standard Error of the Mean (SEM). The SEM reflects the variability of the sampling distribution of the sample mean. A smaller value of SEM indicates that the sample mean is more likely to be close to the population mean, thereby enhancing the reliability and precision of the estimate. p-values were determined by unpaired two-tailed Student’s t-test. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001 Western blot showed that CYP19A1 and NF-κB signaling pathway related gene and apoptosis related genes (Bax, Bcl2 and cleaved caspase 3) levels in KGN cells after the knockdown of PLAU. a The expression of CYP19A1 was decreased in KGN cells after the knockdown of PLAU. b The expression of PLAUR was increased in KGN cells after the knockdown of PLAU. c Western bolt presented the expression of p65, p-p65, Bcl2 and Bax in KGN cells after the knockdown of PLAU. d The expression of p-IKB was decreased in KGN cells after the knockdown of PLAU. e The expression of cleaved caspase 3 was decreased in KGN cells after the knockdown of PLAU. Data are represented as mean ± Standard Error of the Mean (SEM). The SEM reflects the variability of the sampling distribution of the sample mean. A smaller value of SEM indicates that the sample mean is more likely to be close to the population mean, thereby enhancing the reliability and precision of the estimate. p-values were determined by unpaired two-tailed Student’s t-test. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001

The study protocol received approval from the Ethics Committee of the Second Xiangya Hospital of Central South University, following the guidelines of the Council for International Organizations of Medical Sciences. All participants gave informed consent prior to recruitment. Written informed consent was obtained from all participants prior to recruitment. All experimental protocols using animals were approved by the Experimental Animal Ethics Committee of the Second Xiangya Hospital of Central South University of China. Our study was involved 3 normal control individuals and 10 women diagnosed with PCOS. Diagnosis of PCOS in women was based on the 2003 Rotterdam criteria, requiring the presence of at least two of the following clinical manifestations: (1) clinical and/or biochemical hyperandrogenemia; (2) oligo-ovulation and/or anovulation; (3) Polycystic ovary morphology. Individuals with thyroid disease, diabetes, cardiovascular disease, hypertension, neoplasia, endometriosis, renal disease, or recent use of hormonal drugs within the last three months were excluded. Only male azoospermia or infertile patients with tubal occlusion were included in the control group. All individuals were on the first in the vitro fertilization cycle and treated with gonadotropin-releasing hormone antagonist regimen. Transvaginal ultrasound-guided follicular aspiration. Follicular fluid samples from each subject were centrifuged, and granulosa cells with the supernatant discarded were collected and washed in phosphate-buffered saline (PBS) as described previously [ 16 ]. Briefly, washed cell precipitates were resuspended in PBS, layered on Ficoll (LTS1077; TBD Science) solution and separated from erythrocytes by centrifugation. The cell layer at the Ficoll/PBS interface was aspirated and rinsed with PBS to remove residual Ficoll. The final cell sediment was incubated in DMEM-F12 medium containing 10% fetal bovine serum and 1% penicillin-streptomycin in a humidified atmosphere of 5% CO2 at 37 °C for 12 h. The granulosa cells were then collected for subsequent RNA extraction. Total RNA was extracted from collected human granulosa cells according to TRIzol reagent (CoWin Biotech, China) under the manufacturer’s protocol. RNA sequencing (RNA-seq) analysis was performed as previous [ 17 ]. All downstream analyses are based on high quality clean data. Reference genome and gene model annotation files were downloaded directly from the Genome website. Bioinformatics analysis was performed using the R studio tool. To screening for differentially expressed genes, we conducted differentially expressed genes analysis using the R package “LIMMA”, calculating the differences between PCOS and control groups. The thresholds for differentially expressed genes (DEGs) analysis were set at P < 0.05 and log2 FC ≥ 1. Dynamic log2 FC was set in mouse granulosa cells due to the fewer DEGs. Formulas was used to calculated the threshold according to previous study [ 17 ] and finally the threshold was set as log2 FC ≥ 0.602. To identify the functions of differential genes in PCOS, the “clusterProfiler” R package was employed to conduct enrichment analysis on Gene Ontology (GO) terms and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways. Three-week-old female C57BL/6J mice (SPF grade) were obtained from Hunan SJA Laboratory Animal Co., Ltd. (Changsha, China). All animals were housed in ventilated micro-isolated cages illuminated for 12 h per day, with 50 ± 15% humidity and 22 ± 2 °C temperature. The mice were divided into two groups ( n  = 16 per group) randomly at postnatal day 25. Mice in the control group were subcutaneously injected in the neck with 0.2 ml of corn oil daily and fed with a normal chow, and the research group mice was injected daily with DHEA (6.5 mg per 100 g body weight) dissolved in 0.2 ml of corn oil for 28 consecutive days and fed with HFD (60% calories from fat; medicience). The mice were weighed every 4 days. After 28 consecutive days of injection, blood was collected from mice that had been fasted for 12 h after cervical dislocation, and the supernatant was centrifuged and stored in a freezer at − 80 °C for subsequent biochemical experiments. The left ovary of mouse was taken for immunohistochemical analysis and hematoxylin and eosin (H&E), the right ovary was used to extract RNA for qPCR and protein for western blot. Blood glucose levels were measured through the tail vein using the Gold AQ Blood Glucose Monitoring System (Sinocare, China). Mice were fasted for 12 h prior to GTT and 4 h prior to ITT and were then given an intraperitoneal injection of glucose (2 g/kg body weight) for GTT or insulin (1 IU/kg body weight) for ITT. Blood was collected from 0, 15, 30, 60, 90, and 120 min after the injections. To observe changes over at least 2 estrous cycles, we collected vaginal smears from mice at regular intervals starting on days 13–28 of treatment (10 am daily). The smears were stained with Pap smears, and the nuclei of the cells appeared bluish-purple or black, and the cytoplasm of the non-cornified cells appeared light blue nuclei or light green. The cytoplasm of cornified cells appeared pink or orange, and the main cell types were observed under the microscope to determine the stage of the estrus cycle: proestrus = round, nucleated epithelial cells; estrus = cornified squamous epithelial cells; metestrus = epithelial cells and leukocytes; and diestrus = nucleated metestrus = epithelial cells and leukocytes; and diestrus = nucleated epithelial cells and leukocytes. Hematoxylin and eosin staining were used to detect the morphology of the ovaries. After mice were executed by cervical dislocation, the ovaries were dissected and the surrounding tissues and fat were carefully removed, then dehydrated and fixed in 4% paraformaldehyde and 70% ethanol. Paraffin-embedded organs were cut into five consecutive µm sections longitudinally along the largest surface of the ovary using a slicer. Tissue sections were deparaffinised with xylene and then dehydrated in a graded ethanol series for staining. Luteal and follicular cysts were identified and quantified by a trained and licensed pathologist. Mouse granulosa cells (mGCs) were isolated from PCOS-like mouse model group and control group as previous study described with minor modifications [ 18 ]. Mice were cervically executed after 44–48 h of PMSG 10IU injection [ 19 ], and ovaries were removed carefully. After removal of peripheral organs and adipose tissue, ovaries were placed in cell culture dishes containing DMEM/F12 medium (Gibco, China). The ovaries were washed 2–3 times. Using a 25-gauge needle to repeatedly puncture ovarian sinus follicles under the microscope for obtaining granulosa cells. Suspensions of mGCs were collected by centrifugation at 1000 rpm for 5 min and after resuspension three times in DMEM/F12 medium containing 10% fetal bovine serum, 100 IU/mL penicillin and 100 µg/mL streptomycin, and the mGCs were subsequently transferred to DMEM/F12 medium. After overnight incubation, the cells were washed with PBS to remove unattached cells. Cell morphology was observed under a microscope. Venn diagram was performed to screening for critical genes among DEGs from human granulosa cells, mouse ovary and mouse granulosa cells. To illustrate the expression patterns of genes highly enriched functional pathways, we employed the “clusterProfiler” package in R for Gene Set Enrichment Analysis (GSEA). Statistical significance was defined by adjusted P  < 0.05. Then, we performed Real-time fluorescence quantitative PCR (RT-qPCR) to quantify the expression levels of critical genes (CDH23, CLMN, RASEF, ABCB4, PLAU, STEAP4, CYP2S1) between the normal control and PCOS groups of mice. The primer sequences of genes were listed in supplemental materials (Table S1). Mean and standard error (SEM) was used to express the data. Student’s t-test was used to statistically analyze the comparison between groups. P  < 0.05 was considered as significant difference. Paraffin-embedded mouse ovary tissue sections were baked in a thermostatic oven at 60 °C for 1 h to remove paraffin. Samples were hydrated by xylene and gradient ethanol. After high-pressure antigen repair with citrate repair solution and removal of endogenous peroxidase, the samples were blocked with 5% goat serum for 30 min. They were incubated overnight at 4 °C with PLAU antibody (Rabbit, 1:100). The next day, the samples were incubated with goat anti-rabbit secondary antibody (Goat, 1:100) for 1 h. After DAB staining for 5–10 min, nucleus was stained with hematoxylin, dehydrated and blocked, and the sections were visualized under a microscope and recorded. Cells were fixed and cultured in 4% paraformaldehyde and then permeabilized with 0.2% trixon-100 at 25 °C for 10 min, followed by blocking with 3% BSA. Incubate overnight at 4 °C with primary antibody, including PLAU (Rabbit, 1:500), FSHR (Rabbit, 1:500), PLAUR (Rabbit, 1:500), CYP19A1 (Rabbit, 1:500), and they were followed by incubation with fluorescently labeled secondary antibody (Goat, 1:100) protected from light for 1 h. Incubate with DAPI (C1005, Beyotime, China) for 3–5 min, wash and add anti-fluorescence quencher to it for observe under a fluorescence microscope. PLAU-siRNA was designed and synthesized by Hanbio (Shanghai, China). PLAU-siRNA sequence is as follows: Forward: 5'-GCUCAAGGCUUAACUCCAATT-3'. Reverse 5'-UUGGAGUUAAGCCUUGAGCTT-3'. Forward: 5'-GCUCAAGGCUUAACUCCAATT-3'. Reverse 5'-UUGGAGUUAAGCCUUGAGCTT-3'. KGN cells were cultured in 6-well plates until the density reached 50%-70%. siRNA was diluted in serum-free medium for 5 min. According to the manufacturer’s instruction for Lipofectamine ™ 2000, diluted siRNA was added to diluted Lipofectamine ™ 2000 transfection reagent (Thermo Fisher, China) at 50nM (1:1 ratio). Mix gently and incubate for 20 min at room temperature. Finally, added the oligomer-Lipofectamine ™ 2000 complexes to the 6-well plates containing cells and medium. Medium was changed to the medium mixed with 10% FBS after 6 h. 24–36 h after cell transfection, RNA was collected for RT-qPCR experiments. Protein was extracted 48–96 h after transfection. Proteins were extracted with RIPA lysate and phosphatase inhibitor (CoWin Biotech, China). Protein concentration was quantified by BCA protein assay kit (Bioss, China). Protein samples were fractionated by SDS-PAGE fractionation, and then transferred to PVDF membrane (Millipore, USA), which was closed by 5% BSA and then conjugated with primary antibodies, including PLAU (huabio, China), GAPDH (Proteintech, China), P-65, p-p65, p-IKB (Abmart, China), Bax, Bcl2 (Zenbio, China), cleaved caspase 3 (Cell signaling, America) and then probed with secondary antibodies. Protein chemiluminescence was detected by ECL chemiluminescence reagent (Bioss, China). Finally, exposure was performed with a visualizer.

According to Rotterdam Criteria, polycystic ovary syndrome (PCOS) is defined by a combination of clinical features including hyperandrogenemia, oligo-/non-ovulation and polycystic ovarian morphologic pattern, in the absence of related endocrine disorders [ 1 ]. The global prevalence of PCOS is estimated to be almost 10% in 2016, which is the most common endocrine disorders among women of childbearing age [ 2 ]. The results showed that PCOS is associated with an increased prevalence and extent of insulin resistance (IR), type II diabetes [ 3 ] and coronary artery disease [ 4 ], which led to an increase in the burden of disease in almost all countries over the past 30 years, and may exacerbate the global burden of infertility. Subsequently, these diseases may result in significant health and economic costs, making it a major public health problem [ 5 ]. Based on 2023 international evidence-based guidelines, emphasis on continuing a healthy lifestyle in the assessment and management of polycystic ovary syndrome has been suggested, while emphasizing the lack of awareness. Despite recommendations and evidence having improved over the past five years, the overall quality is still low to moderate, and more studies are now required [ 6 ]. Further elucidation of the underlying etiology of PCOS is necessary if we are eager to further improve the quality of managing PCOS. Granulosa cells (GCs) are essential for oocyte development, which not only synthesizing and secreting steroid hormones to regulate the reproductive cycle [ 7 ], but also regulating cellular communication between oocytes and granulosa cells to facilitate ovulation [ 8 ]. Given the critical role of GCs in oocyte development, finding biomarkers of targeting granulosa cells maybe a potential treatment strategy for PCOS [ 9 ]. Therefore, more and more research focusing on granulosa cells to explore genes related to PCOS based on RNA sequencing (RNA-seq) with the development of high-throughput sequencing technology, which is an effective method to identification of possible disease-associated alterations in genes and molecular modifications by analyzing the epigenome and causal inference [ 10 ]. Previous study performed RNA-seq within 6 control and 6 PCOS patients with hyperandrogenemia and found that androgen excess-induced abnormal granulosa cell metabolism plays a role in ovarian dysfunction in patients with PCOS [ 11 ]. However, the pathogenesis of PCOS is ambiguous. Current research seems to recognizes the pathogenesis of PCOS as a combination of polygenic, epigenetic, and developmental contributions that are highly correlated with lifestyle [ 12 ]. PCOS is strongly familial, and hyperandrogenism is the most inheritable phenotypic trait [ 13 ]. In addition, hyperandrogenism is recognized as a major driver of the highly heterogeneous phenotype of polycystic ovary syndrome [ 14 ]. Mendelian randomization analyses and genome-wide association study (GWAS) also identified that PCOS risk genes are related to hyperandrogenemia, including abnormalities in testosterone levels and gonadotropin regulation [ 15 ]. Therefore, to deeper understand the potential risk genes involved in the development of PCOS, we performed RNA-seq in patients with hyperandrogenemia. However, studies in human are limited, so DHEA-induced animal models are necessary to deepen the understanding of the pathogenesis and underlying pathophysiologic mechanisms of PCOS.

Polycystic ovary syndrome is a substantially understudied disease that affecting more than 10% women globally [ 6 ], was marked by hyperandrogenism and insulin resistance, which significantly contributes to the incidence of early-onset type 2 diabetes and cardiovascular disease [ 23 , 24 ]. Up to now, the treatment remains provisional due to the limited understanding of the underlying mechanisms. Therefore, it is vital to explore novel biomarkers as well as the etiology and pathophysiology of the syndrome for research that helps to enhancing the efficacy and precision of interventions [ 25 ]. As technology develops, more and more studies are applying RNA-sequencing technology to medical research in order to probe the alteration of genes at the transcriptional level [ 26 ]. To find potential targets and gain insight into how they are involved in the pathogenesis of polycystic ovary syndrome, we also performed RNA-sequencing on human granulosa cells in patients with PCOS. In our study, we have identified 1734 DEGs including 1545 upregulated genes and 189 downregulated genes in the patients with PCOS comparing to control. Based on these DEGs, GO analysis was applied and shows that notably enriched in positive regulation of cytokine production, leukocyte migration, regulation of immune effector process, secretory granule membrane, specific granule and endocytic vesicle, GTPase regulator activity, nucleoside-triphosphatase regulator activity and phospholipid blinding. As for the enrichment analysis of KEGG, DEGs were mainly concentrated in the pathways such as chemokine signaling pathway, osteoclast differentiation, B cell receptor signaling pathway, leukocyte transendothelial migration, tuberculosis and NF-kappa B signaling pathway. To further explore and identify the potential targets involved in the pathogenesis of PCOS, we constructed PCOS mice models. RNA-sequencing was also performed in mouse ovary and mouse ovarian granulosa cells, respectively. In mouse ovary, 987 genes were identified between PCOS and control groups, including 522 upregulated and 465 downregulated genes in the PCOS group. GO and KEGG analyses showed that DEGs were enriched in signaling conduction, neuroactive ligand-receptor interaction, calcium signaling pathway, steroid hormone biosynthesis, cAMP signaling pathway, metabolism of xenobiotics by cytochrome P450 and primary bile acid biosynthesis. Granulosa cells is essential in the development of follicles [ 27 ]. Previous study reveals that increased granulosa cells apoptosis is one of the key factors for the abnormal follicular development [ 28 ]. Consequently, we conducted RNA-seq in mouse ovarian granulosa cells, which found 389 DEGs between two groups, with 210 upregulated genes and 179 downregulated genes in the PCOS group. GO and KEGG analyses presented that DEGs were enriched in collagen-containing extracellular matrix, cytoskeleton in muscle cells, AGE-RAGE signaling pathway in diabetic complications, focal adhesion, amoebiasis, ECM-receptor interaction, cAMP signaling pathway, TNF signaling pathway, proteoglycans in cancer, regulation of lipolysis in adipocytes, TGF-beta signaling pathway, Ras signaling pathway, Type II diabetes mellitus and platelet activation. Finally, 7 potential risk genes (CDH23, CLMN, RASEF, ABCB4, PLAU, STEAP4, CYP2S1) were identified by combining the DEGs among three sets of sequencing data. In order to have a better understanding of their functions, GSEA analysis was applied and we found that they were mainly enriched in cytokine-cytokine receptor interaction, immune function and hematopoietic cell lineage. The findings may provide a new sight into the alterations of function of granulosa cells in PCOS. Subsequently, we performed RT-qPCR in mouse ovary, the mRNA expression levels of 7 potential genes (CDH23, CLMN, RASEF, ABCB4, PLAU, STEAP4, CYP2S1) were upregulated in PCOS, which is consistent with the RNA-seq analysis findings. However, in mouse ovarian granulosa cell, only PLAU is identified statistically significantly upregulated in PCOS. Then, we employed WB to further identify the protein expression level of PLAU in mouse granulosa cells, which is the same as the results of RT-qPCR that PLAU is increased in PCOS. Plasminogen activator urokinase (PLAU), was known as Urokinase (uPA), encodes a secreted serine protease that converts plasminogen to plasmin [ 29 ]. Plasminogen activates epithelial-mesenchymal transition (EMT), which is involved in the degradation of basement membrane and extracellular mesenchymal components and is crucial in the proliferation and migration of tumor cells [ 30 ]. In gynecological tumors diseases, PLAU was categorized as an unfavorable role [ 31 ]. In recent years, PLAU was reported as a potential role involved in the progress of ovulatory. Zhao et al., found that PLAU/PLAUR induced bovine GC proliferation by cAMP-ERK1/2 signaling pathway, which identified the important role of PLAU in follicular development [ 32 ]. Besides, Berenji et al., targeting lncRNA X-inactive specific transcript and its associated competitive endogenous RNA network towards transcriptome data analysis, first time revealed PLAU mRNA levels is overexpression in human granulosa cells in PCOS compared with non-PCOS women [ 21 ], which is in line with our results. Combining the results of RT-qPCR, WB, IF in mouse ovary and mouse granulosa cells and IHC in mouse ovary, we further identified PLAU is important and overexpressed in granulosa cells in PCOS, but possible mechanisms have not been studied so far. Therefore, we knocked PLAU down in KGN to explore the potential mechanisms involved in the pathogenesis of PCOS. Our study is the first research that reveals hormone secretion-related genes (CYP11A1, CYP19A1 and STAR) were decreased after the downregulated of PLAU, especially CYP19A1, which is a crucial enzyme that localizes to the endoplasmic reticulum and converts androgens into estrogens [ 33 ]. Thus, we believe that PLAU may participate in the pathogenesis of PCOS by interfering with steroid biosynthesis. Besides, in our results of KEGG in three sets of transcriptome data, we found PLAU is involved in nuclear factor kB (NF-κB) signaling pathway. Previous study confirmed the knockdown of PLAU attenuates inflammation by inhibiting NF-κB signaling pathway in human dental pulp tissues [ 34 ]. NF-κB signaling pathway encompasses a wide range of transcription factors which participate in the regulation of several biological responses [ 35 ]. Recently, some study found steroid hormone secretion is associated with NF-κB signaling pathway in the reproductive system [ 36 – 38 ]. Guan et al., ’s study found that TLR4 activation inhibits CYP19A1 expression in human granulosa cells via NF-κB signaling pathway [ 37 ]. Substantial evidence linking PLAU to the NF-κB signaling pathway. Thus, we performed RT-qPCR and western blot to verify the changes of p-p65, p65 and p-IKB, and found they were decreased after silencing PLAU in KGN cells. The results indicated that overexpressed PLAU in granulosa cells maybe disrupts steroid hormone synthesis by activating NF-κB signaling pathway in PCOS. In addition, some studies found the activation of NF-κB signaling pathway induces apoptosis [ 39 ]. Apoptosis is a widely studied form of non-inflammatory programmed death, which is characterized by the changes in pro-apoptotic, anti-apoptotic members and caspase family [ 40 ]. Recent studies have gradually confirmed that apoptosis of ovarian granulosa cells plays a role in the pathogenesis of PCOS [ 41 ]. Tan et al., used KGN cells, a cell lines that is functionally similar to ovarian granulosa cells, and uncovered that miR-93-5p promotes granulosa cell apoptosis in PCOS through the NF-κB signaling pathway [ 22 ]. Accordingly, we set up the hypothesis that high expression of PLAU in PCOS may promote granulosa cell apoptosis through activating the NF-κB signaling pathway, and the results have verified this point finally. Bax, Bax/Bcl2 and cleaved caspase 3 are decreased after the silencing of PLAU in KGN cells. However, our study has some limitations. First of all, due to the difficulty of obtaining samples, the expanded sample validation was not performed in the population. Secondly, in our study, we only investigated pro-apoptotic (BAX), anti-apoptotic proteins (BCL2) and cleaved caspase 3 without further exploring the changes in other caspase family members, which is one of the drawbacks of this study. Secondly, the functional experiments of PLAU were conducted in KGN cells in vitro, which may not fully recapitulate the complex physiological environment of primary granulosa cells. Additionally, further exploration of the role of PLAU in the pathogenesis of PCOS is needed. In this study, we not only provided transcriptome data analysis of human granulosa cells between PCOS and control groups, but also revealed the transcriptome data analysis of mouse ovary and mouse ovarian granulosa cells between PCOS and control groups. Based on these DEGs, we screened seven potential risk genes. GSEA presented the functions of seven genes for us. According to the results of RT-qPCR, we believe PLAU maybe plays an essential role in the development of PCOS. Thus, we further identified PLAU is upregulated in granulosa cells of PCOS. Finally, we constructed siRNA of PLAU to explore the regime, and we found that CYP19A1, p65, p-p65, p-IKB, Bax, Bax/Bcl2 and cleaved caspase 3 were decreased after the knockdown of PLAU. Hence, we concluded that upregulated PLAU in granulosa cells interferes with steroid hormone synthesis and promotes apoptosis by activating NF-κB signaling pathway in polycystic ovary syndrome (Fig.  10 ). Fig. 10 Overview of the effect of PLAU in granulosa cells. PLAU activates the NF-κB signaling pathway by binding to PLAUR and increases the expression of Bax, ultimately leading to apoptosis. Additionally, upregulated PLAU existed in nucleus increased the expression of CYP19A1 Overview of the effect of PLAU in granulosa cells. PLAU activates the NF-κB signaling pathway by binding to PLAUR and increases the expression of Bax, ultimately leading to apoptosis. Additionally, upregulated PLAU existed in nucleus increased the expression of CYP19A1 In summary, we have offered a more comprehensive picture of the transcriptomic changes that occur in PCOS, which may provide new perspectives for future research. Our study is the first study identifying the expression of PLAU in granulosa cells and exploring its important role participated in the pathogenesis of PCOS. These findings will facilitate the development of personalized management and treatment strategies for polycystic ovary syndrome.

In this study, we describe the mRNA expression profiles of ovarian granulosa cells from patients with PCOS, mouse ovary and mouse granulosa cells in PCOS-like models, both suggesting differences at the transcriptional level. Besides, our study first time revealed the location of PLAU in granulosa cell, and verified upregulated expression of PLAU in granulosa cell interferes with steroid hormone synthesis and promotes apoptosis by activating NF-κB signaling pathway in PCOS. Our study provides a new perspective on the research of PCOS and indicates that PLAU may be a new molecular target for improving the function of granulosa cells. However, much more exploration of PLAU is necessary to reveal the role involved in PCOS.

Supplementary Material 1. Supplementary Material 1. Supplementary Material 2. Supplementary Material 2. Supplementary Material 3. Supplementary Material 3.