Big data analysis reveals miR-874 as a suppressor and therapeutic target in the progression from endometriosis to ovarian cancer

We conducted a bidirectional two-sample MR analysis using the EMS dataset (ebi-a-GCST90018839) from the EBI GWAS database as the exposure and the CCOC dataset (ieu-a-1124, n  = 42,307) from the IEU database as the outcome. The causal effect of EMS on CCOC development was assessed using the Inverse Variance Weighting (IVW) method. Leave-one-out sensitivity analysis indicated that the removal of any single SNP did not substantially alter the causal effect estimates, suggesting that no single SNP disproportionately influenced the overall effect (Fig.  1 A). The combined effect was significant, and the effect direction of individual SNPs was generally consistent, supporting the credibility of the causal association (Fig.  1 B). Funnel plot analysis suggests that our MR results are not biased, suggesting the robustness of our findings (Fig.  1 C). Our results suggest that EMS is a potential causal factor for CCOC development (Fig.  1 D), whereas the reverse relationship was not supported. The occurrence of EMS may be involved in the development of CCOC (Fig.  1 E). However, no significant causal association was observed between EMS and EOC. Fig. 1 Two-sample Mendelian randomization (MR) analysis reveals a potential causal relationship between endometriosis (EMS) and clear cell ovarian cancer (CCOC) development by the TwoSampleMR package. A By sequentially removing each SNP and rerunning the MR analysis, we observed the overall effect estimates remained consistent, indicating that the results are robust. B Forest plot for the effect of individual SNPs on EMS and CCOC development. C The distribution of individual SNP effect estimates was nearly symmetric, located on both sides of the vertical line, indicating minimal horizontal pleiotropy for the SNPs. D Scatter Plot for SNP Effects on EMS and CCOC. E Five analysis algorithms identified 13 SNPs associated with a positive impact of EMS on CCOC Development: The figure shows the effect direction (b), standard error (se), and statistical significance (p-value) of these SNPs Two-sample Mendelian randomization (MR) analysis reveals a potential causal relationship between endometriosis (EMS) and clear cell ovarian cancer (CCOC) development by the TwoSampleMR package. A By sequentially removing each SNP and rerunning the MR analysis, we observed the overall effect estimates remained consistent, indicating that the results are robust. B Forest plot for the effect of individual SNPs on EMS and CCOC development. C The distribution of individual SNP effect estimates was nearly symmetric, located on both sides of the vertical line, indicating minimal horizontal pleiotropy for the SNPs. D Scatter Plot for SNP Effects on EMS and CCOC. E Five analysis algorithms identified 13 SNPs associated with a positive impact of EMS on CCOC Development: The figure shows the effect direction (b), standard error (se), and statistical significance (p-value) of these SNPs We analyzed miRNA expression changes and their potential clinical significance using multiple independent datasets. In the plasma miRNA dataset for EMS patients ( GSE46735 ), several differential miRNAs were identified. By intersecting data from different stages, we found 10 miRNAs that showed differential expression across various stages (Fig.  2 A). We found miR-874 was consistently down-regulated in endometrioma (Fig.  2 B) and endometrioma associated CCOC (Fig.  2 C). To assess the potential clinical relevance of these miRNAs, we performed survival analysis using KMplot. Four miRNAs (miR-148a, miR-155, miR-502, and miR-874) were found to have consistent prognostic roles across three common gynecological cancers including cervical squamous cell carcinoma, ovarian cancer, and endometrial cancer. Notably, higher expression levels of these miRNAs were associated with longer patient survival, suggesting a protective effect (Fig.  3 A–C). We further validated these findings using dataset GSE94533 , which contains serum miRNA transcriptome data from 180 ovarian cancer cases spanning different histopathological subtypes. Among the four miRNAs, only miR-874 was consistently differentially expressed, showing significant down-regulation in EMS patients (Fig.  2 B). In dataset GSE239685 ( n  = 20), we confirmed the down-regulation of serum miR-874 in CCOC compared with healthy donor. In dataset GSE214239 ( n  = 8), we found that miR-874 in tissue-exudative extracellular vesicles was lower in the CCOC compared to normal ovaries. Moreover, transcriptome analysis of CCOC patients with EMS ( GSE230956 ) also revealed miR-874 down-regulation (Fig.  2 C). Therefore, the consistent down-regulation of miR-874 across multiple datasets suggests that its decreased expression in EMS patients may be associated with the development of CCOC and potentially other gynecological malignancies. Fig. 2 Down-regulation of miR-874 expression in endometriosis and clear cell ovarian cancer (CCOC) with concurrent endometriosis. A Comparison of plasma miRNA differential expression between the follicular phase (early proliferative and late proliferative phases) and the luteal phase ( p  < 0.05) in GSE46735 . B miR-874 is down-regulated in the plasma of EMS patients (* p  < 0.05) in GSE94533 . C miR-874 is down-regulated in CCOC patients with EMS (* p  < 0.05) in GSE230956 Down-regulation of miR-874 expression in endometriosis and clear cell ovarian cancer (CCOC) with concurrent endometriosis. A Comparison of plasma miRNA differential expression between the follicular phase (early proliferative and late proliferative phases) and the luteal phase ( p  < 0.05) in GSE46735 . B miR-874 is down-regulated in the plasma of EMS patients (* p  < 0.05) in GSE94533 . C miR-874 is down-regulated in CCOC patients with EMS (* p  < 0.05) in GSE230956 Fig. 3 Differentially expressed miR-148a, miR-155, miR-502, and miR-874 in endometriosis are associated with prognosis in three common gynecological cancers according to KMplot web tool. A Cervical squamous cell carcinoma. B Ovarian cancer. C Endometrial cancer Differentially expressed miR-148a, miR-155, miR-502, and miR-874 in endometriosis are associated with prognosis in three common gynecological cancers according to KMplot web tool. A Cervical squamous cell carcinoma. B Ovarian cancer. C Endometrial cancer To investigate the potential mechanisms of miR-874, we performed differential gene expression analysis using the TCGA ovarian cancer dataset. Samples were stratified into high- and low-expression groups based on miR-874 levels, with the low-expression group serving as the control. A total of 194 differentially expressed genes (DEGs) were identified (Fig.  4 A, B), among which ASB9, TMSB10P1, SMIM24, and PDLIM4 were the most significantly up-regulated. A literature search in the PubMed database revealed that ASB9 functions as a negative regulator of cell growth, which aligns with previous findings suggesting that miR-874 may act as a tumor suppressor [ 39 , 40 ]. GO functional analysis showed that these DEGs are associated with mitochondrial matrix, cell adhesion, proteasomal protein degradation, and GTPase activity regulation (Fig.  4 C–E). Pathway analysis revealed that down-regulated genes were enriched in the p53 signaling pathway, including MTBP, RAD18, ZNF217, and NDRG1. Given that miRNAs typically function by negatively regulating target gene expression, we conducted a survival analysis to assess the prognostic relevance of these DEGs. The analysis demonstrated that most down-regulated genes were significantly associated with survival outcomes in ovarian cancer patients. Specifically, ZNF623, NDRG1, ZNF7, MTBP, and PHF20L1 were down-regulated in tissues with high miR-874 expression, and lower expression of these genes was correlated with improved survival (Fig.  5 ). Conversely, among the up-regulated genes, ASB9 was associated with better survival outcomes, while high expression of PDLIM4 and ECRG4 was linked to poorer prognosis. Further validation using the GSE230956 dataset confirmed that some of the down-regulated genes identified in the TCGA tissues with high miR-874 expression were up-regulated in CCOC patients with EMS (Figure S1), suggesting that these genes may contribute to CCOC pathogenesis. In summary, these findings provide mechanistic insights into the potential tumor-suppressive role of miR-874 and its involvement in CCOC progression. Fig. 4 Transcriptomic differential expression analysis between high and low miR-874 expression in TCGA ovarian cancer tissues. A Volcano plot showing 194 differentially expressed genes identified under the conditions of P  < 0.01 and log2FC = 0.3, with 37 down-regulated and 157 up-regulated genes. B Boxplot of the top 10 most significantly differentially expressed genes. C GO cellular component, D GO biological process, and E GO molecular function enrichment analysis of the differentially expressed genes Transcriptomic differential expression analysis between high and low miR-874 expression in TCGA ovarian cancer tissues. A Volcano plot showing 194 differentially expressed genes identified under the conditions of P  < 0.01 and log2FC = 0.3, with 37 down-regulated and 157 up-regulated genes. B Boxplot of the top 10 most significantly differentially expressed genes. C GO cellular component, D GO biological process, and E GO molecular function enrichment analysis of the differentially expressed genes Fig. 5 Survival analysis of the top 10 differentially expressed genes based on miR-874 high and low expression in TCGA ovarian cancer tissues by KMplot web tool. Eight genes, ZNF623, DRG1, ZNF7, MTBP, PHF20L1, ASB9, PDLIM4, and ECRG4, are associated with survival prognosis. Among them, ZNF623, DRG1, ZNF7, MTBP, and PHF20L1 are down-regulated in tissues with high miR-874 expression, while ASB9, PDLIM4, and ECRG4 are up-regulated Survival analysis of the top 10 differentially expressed genes based on miR-874 high and low expression in TCGA ovarian cancer tissues by KMplot web tool. Eight genes, ZNF623, DRG1, ZNF7, MTBP, PHF20L1, ASB9, PDLIM4, and ECRG4, are associated with survival prognosis. Among them, ZNF623, DRG1, ZNF7, MTBP, and PHF20L1 are down-regulated in tissues with high miR-874 expression, while ASB9, PDLIM4, and ECRG4 are up-regulated EcoTyper is an advanced tool used for analyzing transcriptomic data, particularly for deconvoluting the cellular composition and cellular states of complex tissues, including tumors. Using machine learning-based approaches, EcoTyper identifies distinct cell subtypes and their abundance from gene expression data. Given that tumors comprise diverse cell populations, we applied EcoTyper to quantitatively analyze cell type abundance in ovarian cancer tissues (Fig.  6 ), enabling a deeper understanding of the tumor microenvironment concerning miR-874 expression. This analysis revealed three cell subtypes that are associated with high miR-874 expression. Specifically, Endothelial.cells.5 (myoepithelial-like cells) and Fibroblasts.1 (myofibroblast-like cells) were up-regulated in tissues with high miR-874 expression, whereas Fibroblasts.8 (migratory-like cells) was down-regulated (Fig.  7 A). Among these cell types, a higher abundance of fibroblasts.1 is associated with longer survival (Fig.  7 B). These findings suggest that high miR-874 expression is linked to a more favorable tumor phenotype, consistent with the improved prognosis observed in patients with elevated miR-874 levels. Overall, these results provide insight into the potential cellular mechanisms underlying miR-874 function in the tumor microenvironment. Furthermore, these cell subtypes could serve as potential prognostic cellular markers in ovarian cancer. Fig. 6 Tumor microenvironment cell abundance analysis of TCGA ovarian cancer tissue samples by Ecotyper tool. A total of 9 cell types were quantified Tumor microenvironment cell abundance analysis of TCGA ovarian cancer tissue samples by Ecotyper tool. A total of 9 cell types were quantified Fig. 7 Tumor microenvironment cell abundance difference analysis between high and low expression of miR-874 in TCGA ovarian cancer. A Endothelial.cells.5, annotated as Myoendothelium-like cells, is up-regulated in samples with high miR-874 expression. Fibroblasts.1, annotated as Myofibroblast-like cells, is up-regulated in high miR-874 expression samples. Fibroblasts.8, annotated as Pro-migratory-like cells, is down-regulated in high miR-874 expression samples. B A higher abundance of fibroblasts.1 is associated with longer survival. Green and red lines indicate high and low miR-874 expression Tumor microenvironment cell abundance difference analysis between high and low expression of miR-874 in TCGA ovarian cancer. A Endothelial.cells.5, annotated as Myoendothelium-like cells, is up-regulated in samples with high miR-874 expression. Fibroblasts.1, annotated as Myofibroblast-like cells, is up-regulated in high miR-874 expression samples. Fibroblasts.8, annotated as Pro-migratory-like cells, is down-regulated in high miR-874 expression samples. B A higher abundance of fibroblasts.1 is associated with longer survival. Green and red lines indicate high and low miR-874 expression To further investigate the molecular mechanisms underlying miR-874 function, we performed WGCNA on the DEGs identified in previous analyses. This analysis suggested a potential regulatory relationship between miR-874 and NDRG1 as well as ZNF217 (Fig.  8 A–H). To validate these findings, we conducted target gene prediction using miWalk, which identified NDRG1 and ZNF217 as potential direct targets of miR-874 among the down-regulated genes. Mechanistically, miR-874 may exert its effects by binding complementarily to the 3’ UTR of NDRG1 and the CDS region of ZNF217 (Fig.  9 A, B). These findings suggest that miR-874 may play a tumor-suppressive role by regulating key oncogenes in ovarian cancer. Fig. 8 miR-874 may influence the occurrence and development of CCOC by regulating NDRG1 and ZNF217. A Network topology of the WGCNA analysis at different values of the power parameter, where a network is considered scale-free if R 2 is greater than 0.8. The red horizontal line represents R² = 0.8. B When the power parameter is greater than 12, the connectivity of nodes in the network reaches a stable state. C With the power parameter set to 12, the constructed co-expression network meets the scale-free requirement, with the fitted R 2 of the linear regression exceeding 0.8. D WGCNA identified two co-expression modules, with 26 and 34 genes, respectively. E The expression of co-expression module ME2 is negatively correlated with miR-874 expression. F Co-expression network construction for module ME2. G The regulatory network between downregulated genes and miR-874 constructed using FFLtool. H Patients with high expression of ZNF217 have a shorter survival time according to KMplot web tool miR-874 may influence the occurrence and development of CCOC by regulating NDRG1 and ZNF217. A Network topology of the WGCNA analysis at different values of the power parameter, where a network is considered scale-free if R 2 is greater than 0.8. The red horizontal line represents R² = 0.8. B When the power parameter is greater than 12, the connectivity of nodes in the network reaches a stable state. C With the power parameter set to 12, the constructed co-expression network meets the scale-free requirement, with the fitted R 2 of the linear regression exceeding 0.8. D WGCNA identified two co-expression modules, with 26 and 34 genes, respectively. E The expression of co-expression module ME2 is negatively correlated with miR-874 expression. F Co-expression network construction for module ME2. G The regulatory network between downregulated genes and miR-874 constructed using FFLtool. H Patients with high expression of ZNF217 have a shorter survival time according to KMplot web tool Fig. 9 miR-874 may influence gene expression by binding to the mRNA of NDRG1 and ZNF217 according to the miRWalk database. A Detailed information on the miR-874:NDRG1 complex and its complementary pairing with the 3’UTR region of NDRG1. B Detailed information on the miR-874:ZNF217 complex and its complementary pairing with the CDS region of ZNF217 miR-874 may influence gene expression by binding to the mRNA of NDRG1 and ZNF217 according to the miRWalk database. A Detailed information on the miR-874:NDRG1 complex and its complementary pairing with the 3’UTR region of NDRG1. B Detailed information on the miR-874:ZNF217 complex and its complementary pairing with the CDS region of ZNF217 Single-cell analysis revealed that NDRG1 is primarily expressed in ovarian stromal cells (Fig.  10 ), while ZNF217 is predominantly expressed in ovarian stromal cells, smooth muscle cells, and immune cells (Fig.  10 ). To further investigate the functional relevance of these genes, we analyzed the ME2 co-expression module, which was significantly associated with the p53 signaling pathway, as determined using the Reactome database. Literature evidence highlights several genes within this module that are directly linked to p53 regulation, including NDRG1, MTBP, PVT1, RAD18, and ZNF217 [ 41 , 42 ]. Protein expression analysis further confirmed that NDRG1 and ZNF217 are up-regulated in EMS-associated ovarian cancer (Fig.  11 ). Notably, ZNF217, a transcription factor with a well-established role in oncogenesis [ 43 ], is significantly associated with patient survival. Given its direct regulatory role in the p53 signaling pathway, we hypothesize that ZNF217 may be regulated by miR-874. To explore this regulatory network, ChIP-Seq (Chromatin Immunoprecipitation Sequencing) data analysis was performed, revealing that ZNF217 binds to the upstream promoter region of MTBP. As miR-874 is down-regulated in the blood of EMS. We propose a hypothesis that the decrease in ovarian tissue miR-874 levels results in the up-regulation of the transcription factor ZNF217. This further activates the transcription of MTBP, which stabilizes the p53-negative regulator MDM2, leading to decreased p53 signaling pathway activity, inducing epithelial-mesenchymal transition (EMT) in ovarian epithelial cells, and promoting the development of CCOC (Fig.  12 ). Based on these findings, we propose a miR-874–ZNF217–MTBP–MDM2–p53 regulatory axis that may play a crucial role in the development of EMS to CCOC (Fig.  12 ). Future experiments will be designed to validate this hypothesis. Fig. 10 Single-cell analysis reveals that the NDRG1 gene is primarily expressed in ovarian stromal cells, while the ZNF217 gene is mainly expressed in ovarian stromal cells, smooth muscle cells, and immune cells according to the Human Protein Atlas database Single-cell analysis reveals that the NDRG1 gene is primarily expressed in ovarian stromal cells, while the ZNF217 gene is mainly expressed in ovarian stromal cells, smooth muscle cells, and immune cells according to the Human Protein Atlas database Fig. 11 According to the Human Protein Atlas database, the expression of NDRG1 and ZNF217 is increased in endometriosis-associated ovarian cancer tissues. Magnification: 100x According to the Human Protein Atlas database, the expression of NDRG1 and ZNF217 is increased in endometriosis-associated ovarian cancer tissues. Magnification: 100x Fig. 12 Potential mechanism of miRNA-induced transformation of endometriosis (EMS) into clear cell ovarian cancer (CCOC) Potential mechanism of miRNA-induced transformation of endometriosis (EMS) into clear cell ovarian cancer (CCOC)

We utilized the publicly available dataset GSE46735 from the Gene Expression Omnibus (GEO) database. This dataset comprises plasma samples collected from 16 women, divided equally between those diagnosed with endometriosis and asymptomatic controls. Each participant provided blood samples during three distinct phases of the menstrual cycle: early proliferative, late proliferative, and mid-luteal, resulting in a total of 47 plasma samples. The menstrual cycle phases were confirmed through hormonal profiling [ 24 ]. GSE46735 was used to identify consistently dysregulated miRNAs in EMS. GSE94533 comprises pre-treatment serum samples from 179 women presenting with an adnexal mass. Total serum RNA was extracted, converted into microRNA next-generation sequencing libraries, and sequenced using the Illumina NextSeq 500 platform [ 25 ]. GSE230956 includes RNA sequencing (RNA-seq) and small RNA sequencing (small RNA-seq) data from eight primary tissue samples. Among them, four samples were CCOC with EMS, while the remaining four were benign endometriotic cysts [ 26 ]. GSE239685 includes circulating miRNA profiles from blood samples from 16 CCOC patients and 4 healthy subjects. GSE214239 focuses on miRNAs derived from tissue-exudative extracellular vesicles (Te-EVs) isolated from cancer and adjacent normal ovarian tissues of 8 CCOC patients. GSE94533 , GSE230956 , GSE239685 , and GSE214239 were used to validate the dysregulated miRNAs identified in EMS and CCOC. We conducted a comprehensive analysis using data from The Cancer Genome Atlas (TCGA) ovarian cancer cohort. Specifically, we selected 426 patients for whom both gene expression and miRNA profiling data were available [ 27 ]. To investigate the potential causal relationship between EMS and CCOC occurrence, a bidirectional two-sample Mendelian randomization (MR) analysis was conducted. Summary-level genetic association data for EMS were obtained from the ebi-a-GCST90018839 dataset ( n  = 231,771), while data for CCOC were sourced from the ieu-a-1124 dataset ( n  = 42,307) in the MRC IEU OpenGWAS database [ 12 ]. Single nucleotide polymorphisms (SNPs) strongly associated with EMS ( p  < 1 × 10 − 6 ) were selected as instrumental variables (IVs), ensuring they were independent (r 2  < 0.01) and not in linkage disequilibrium. The inverse-variance weighted (IVW) method was employed as the primary analytical approach to estimate causal effects, while MR-Egger regression and the weighted median method were used for sensitivity analyses to address potential pleiotropy and heterogeneity. The F-statistic was computed to evaluate IV strength. Additionally, Cochran’s Q test was performed to assess heterogeneity, and MR-Egger regression intercepts were examined for evidence of horizontal pleiotropy. All statistical analyses were conducted using the R package TwoSampleMR [ 28 ]. The results were expressed as odds ratios (OR) or beta coefficients with 95% confidence intervals (CI) to quantify the effect of EMS on CCOC, with a reverse analysis conducted to assess the potential impact of CCOC on EMS. Differential plasma miRNA expression analysis was performed using GEO2R, an online tool provided by NCBI GEO, to preprocess and normalize the data. GEO2R, which is based on the limma package in R, was used to identify differentially expressed miRNAs between EMS and control samples. Significant miRNAs were defined by an adjusted P-value  1. For differential gene expression analysis in tissue samples, gene expression data were preprocessed by filtering based on mean expression, standard deviation (SD), and coefficient of variation (CV). Genes with a mean expression and CV above the 10th percentile were retained for further analysis. Differential gene expression was assessed using the eBayes method implemented in the R package limma, with significance thresholds set at P   0.3. Our study focuses on the comparison of molecular subtypes within tumor samples, rather than comparing tumor tissues with adjacent normal tissues. In such intra-tumoral comparisons, the transcriptional differences are typically more subtle. Therefore, we adopted a relaxed log2 fold change threshold of > 0.3 to capture biologically meaningful but moderate transcriptional shifts that are characteristic of subtype-specific signatures. This threshold yielded a manageable number of genes for downstream analysis while retaining potential biological signal. Functional annotation of differentially expressed genes (DEGs) was conducted using the clusterProfiler package [ 29 ]. Survival analysis was performed using the KMplot web tool ( https://kmplot.com/ ) [ 30 ]. Kaplan-Meier survival curves were generated to compare ovarian cancer patients with high and low expression levels of the identified miRNAs. The log-rank test was applied to assess statistical significance [ 31 ]. WGCNA was performed using the R package WGCNA [ 32 ]. First, the DEG expression matrix was transformed into a similarity matrix based on Pearson correlation coefficients between gene pairs [ 33 ]. A soft-thresholding power of 12 was selected using the pickSoftThreshold function to achieve scale-free topology, and an adjacency matrix was constructed. This matrix was subsequently converted into a topological overlap matrix (TOM) to measure network connectivity. Genes were clustered using hierarchical clustering with dynamic tree cut, and modules were identified based on similar expression patterns with parameter minModuleSize = 20. Module-miR-874 relationships were assessed by correlating module eigengenes with miR-874 expression. The module of interest was selected based on the highest correlation coefficient. Intramodular connectivity and gene significance were used to identify hub genes within the key module. Gene co-expression network was visualized by the visNetwork package. Functional enrichment analysis of module genes was conducted using the clusterProfiler package. The target genes of miR-874 were predicted using the miWalk web tool [ 34 ]. The minimum binding site score threshold for miRNA-target gene duplex was set to 0.99. To explore potential feed-forward loops involving transcription factors (TF), miRNAs, and target genes, the FFLtool web tool was utilized [ 35 ]. We integrated comprehensive TF–target and miRNA–target regulatory networks into FFLtool and constructed functional modules based on genes identified in the prior WGCNA. The resulting networks were visualized to highlight key interactions between miRNAs and TF. Cellular composition profiles of TCGA samples were obtained using the Ecotyper deconvolution algorithm [ 36 , 37 ]. Heat maps were created to display the differences in cellular subtype abundance between samples with high and low miR-874 expression. The most significantly differentially expressed cell subtypes were identified through t-tests, and their association with patient survival was assessed using the survminer and survival packages in R. Kaplan-Meier survival curves were generated for ovarian cancer patients, stratified based on high and low levels of cellular subtype abundance. Single-cell transcriptomic and histopathological data were retrieved from the Human Protein Atlas database [ 38 ]. All statistical analyses were performed using R software. P-values < 0.05 were considered statistically significant. Multiple testing corrections were applied where appropriate, using methods such as the Benjamini-Hochberg false discovery rate.

With the rapid development of technologies such as bioinformatics, genomics, and transcriptomics, research on ovarian cancer and endometriosis is gradually evolving toward multidimensional and systematic approaches. This study integrates big data analysis to delineate the underlying mechanisms of ESM transformation to ovarian cancer. We used MR analysis to confirm that EMS is a key predisposing factor for the development of CCOC. Bioinformatics analysis of multiple datasets further revealed that down-regulation of miR-874 is closely associated with EMS occurrence, ovarian cancer progression, and patient survival. In tumor tissues with low miR-874 expression, significant alterations were observed in ZNF217, MTBP, NDRG1, RAD18, and PVT1, all of which are related to the p53 signaling pathway. Furthermore, we explored how miR-874 down-regulation induces ZNF217 up-regulation, which is correlated with MTBP overexpression and increased MDM2 stability, ultimately suppressing p53 signaling activity and promoting tumor progression. Clinical samples confirmed abnormalities in miR-874 associated molecules ZNF217 and NDRG1. These findings provide a scientific basis and novel insights for improving EMS and CCOC prognosis and developing personalized therapeutic strategies. GWAS has identified SNPs associated with ovarian cancer and EMS, providing potential genetic markers for early diagnosis and personalized treatment [ 44 , 45 ]. MR analysis, by using genetic variants to infer causal relationships, minimizes confounding biases and has been increasingly applied to gynecological diseases, offering new insights into their prevention and intervention. In our study, MR analysis confirmed EMS as a key causal factor for the development of CCOC, further supporting the hypothesis that genetic predispositions associated with EMS can influence ovarian cancer susceptibility. The use of large-scale genetic data from the EBI and IEU databases allowed for a robust assessment of the causal relationship between EMS and CCOC, providing strong evidence for the role of EMS as a primary risk factor for CCOC. Our subsequent bioinformatics analyses of miRNA expression in EMS and ovarian cancer tissues identified miR-874 as a critical modulator in the development of these conditions. Notably, miR-874 was found to be significantly down-regulated in both EMS and CCOC tissues. The expression pattern of miR-874 suggests that its loss may disrupt critical molecular pathways involved in tumor suppression, thereby promoting cancer progression. Our study further demonstrated that miR-874 down-regulation was associated with the up-regulation of key genes such as ZNF217, MTBP, and NDRG1, which are involved in the p53 signaling pathway, a central regulator of cell cycle and apoptosis [ 46 – 50 ]. Although the GSE230956 dataset provides useful information for exploratory validation, its relatively small sample size limits the statistical power of our conclusions. Therefore, the observed gene expression trends should be interpreted cautiously and require further validation in larger independent cohorts. The interplay between miR-874 and ZNF217 is particularly noteworthy. As a transcription factor, ZNF217 has been implicated in the regulation of genes related to cell growth and survival [ 43 ]. ChIP-seq data showed that ZNF217 could regulate MTBP, a known stabilizer of MDM2 [ 51 ]. MDM2 is a negative regulator of p53 [ 52 ]. We hypothesize that miR-874 may exert its effects by directly targeting ZNF217, leading to the activation of MTBP and stabilization of MDM2, which in turn inhibits p53 signaling. This molecular axis may contribute to the unchecked cell proliferation and tumor progression observed in CCOC. Single-cell RNA sequencing and co-expression analyses further elucidated the role of these molecular interactions within the tumor microenvironment. We identified distinct cell subtypes associated with high miR-874 expression, including endothelial cells and fibroblasts, which suggest a potential role for miR-874 in maintaining a more benign tumor phenotype. This finding is consistent with the observation that high miR-874 expression correlates with better patient prognosis. The absence of tissue-level validation is a limitation, and future studies should perform tumor microenvironment profiling using IHC or multiplex immunofluorescence to confirm key cell subsets such as Fibroblasts.1 and immune components in EMS-associated CCOC. The identification of these cell types as potential prognostic markers offers potential avenues for future research focused on cellular compositions within the tumor microenvironment and their role in CCOC progression. Additionally, the down-regulation of miR-874 and subsequent up-regulation of ZNF217 in CCOC suggests that miR-874 acts as a tumor suppressor in EMS-associated ovarian cancer. This study provides the first comprehensive analysis of how miR-874 may inhibit tumorigenesis by targeting critical genes involved in COCC development. Our findings not only shed light on the complex molecular mechanisms of EMS-associated CCOC but also highlight miR-874 as a potential biomarker and therapeutic target for improving the prognosis of CCOC patients. In conclusion, this study proposes a potential molecular link between EMS and CCOC, highlighting the miR-874-ZNF217-p53 axis as a candidate pathway involved in tumor suppression. While our findings offer preliminary insights that may inform future therapeutic strategies, they are based on bioinformatic analyses of general ovarian cancer datasets and should be considered hypothesis-generating. Further experimental validation in CCOC-specific models is warranted to confirm the relevance and therapeutic potential of the identified mechanisms.

Despite the comprehensive multi-omics analysis and integration of large public datasets, our study has several limitations. First, some of the datasets included relatively small sample sizes, which may introduce sampling bias and limit the generalizability of the findings. The reduced statistical power may also increase the risk of false positives or false negatives in differential gene expression analyses. Second, while we used multiple independent datasets to cross-validate the expression patterns of miR-874 in EMS and CCOC, the regulatory mechanisms and downstream effects were primarily inferred through bioinformatics approaches and lack in vivo or in vitro experimental validation. Third, the proposed miR-874–ZNF217–MTBP–MDM2–p53 axis requires further experimental confirmation to establish a causal relationship. It is important to note that the differentially expressed genes (DEGs) used in our analysis were derived from the TCGA ovarian cancer cohort, and therefore, the findings are more broadly representative of ovarian cancer rather than specific to CCOC, which remains to be experimentally validated in future studies. Future work should include experimental approaches such as miRNA manipulation in CCOC cell lines, expression analysis in clinical samples, target verification assays, and immunohistochemistry or multiplex immunofluorescence to validate key cell subtypes. These steps will help to substantiate the molecular mechanisms predicted by our computational analyses.

Endometriosis (EMS) is a common chronic gynecological disorder primarily affecting women of reproductive age. It is characterized by the presence of endometrial-like tissue outside the uterine cavity, which remains responsive to cyclical ovarian hormonal changes [ 1 ]. This leads to periodic bleeding, chronic inflammation, and fibrosis, causing a range of symptoms including chronic pelvic pain and infertility. The overall prevalence of endometriosis in women of reproductive age is approximately 5%-10% [ 2 ]. In women with chronic pelvic pain, the prevalence of endometriosis can range from 30 to 50%, and among women with infertility, approximately 25–50% have endometriosis. While EMS is typically considered a benign condition, chronic inflammation, tissue hyperplasia, and abnormal behavior of ectopic endometrial tissue may increase the risk of certain types of tumors. Notably, women with EMS are at a significantly heightened risk of developing ovarian cancer. Studies suggest that the risk of ovarian cancer in women with EMS is approximately three times greater than in the general population [ 3 ]. Clear cell ovarian carcinoma (CCOC) is a subtype of ovarian cancer that is clinically challenging to treat and associated with a poor prognosis. Its pathogenesis is complex and not yet fully understood [ 4 ]. CCOC accounts for 5–25% of all ovarian malignancies, second only to high-grade serous carcinoma (70%), with a mortality rate ranking second among all ovarian cancers and a 5-year survival rate of approximately 50% [ 5 ]. Studies have shown that endometriosis (EMS) is one of the significant risk factors for CCOC. Additionally, endometrioid ovarian cancer (EOC), another subtype closely linked to EMS, also exhibits histological features similar to those of endometrial cancer. EOC accounts for 10–15% of ovarian epithelial cancers and typically has a 5-year survival rate of about 70% [ 6 ]. Despite evidence suggesting that certain ovarian cancers, particularly CCOC and EOC, may evolve from EMS, the exact molecular mechanisms underlying this progression remain unclear [ 7 , 8 ]. Although EMS is typically regarded as benign, growing evidence points to its potential role in the malignant transformation to ovarian cancer, especially CCOC, which is associated with poorer prognosis [ 9 ]. However, the lack of systematic clinical and molecular mechanistic studies has hindered a clear understanding of the connection between EMS and CCOC [ 10 ]. In recent years, with the development of genomics, transcriptomics, and big data analysis technologies, the genetic background and molecular mechanisms of many diseases have been studied in unprecedented depth. In particular, genome-wide association studies (GWAS) have made significant progress in revealing the genetic factors underlying diseases [ 11 ]. By utilizing large-scale genetic data from GWAS databases and combining them with Mendelian randomization (MR) analysis, strong evidence for inferring causal relationships between diseases has been provided [ 12 – 14 ]. This study systematically analyzes large-scale genetic data and employs bidirectional two-sample MR analysis to verify the potential causal relationship between EMS and CCOC, laying a solid theoretical foundation for further experimental research. Blood is considered a fluid environment that reflects the status of various organs in the body, and it plays a crucial role in the occurrence, diagnosis, prognosis, and treatment of diseases [ 15 ]. One of the emerging areas of interest is the role of microRNAs (miRNAs), small non-coding RNAs that regulate gene expression by binding to messenger RNAs (mRNAs) and inhibiting their translation or promoting their degradation [ 16 ]. MiRNAs have been implicated in various biological processes, including cell growth, differentiation, apoptosis, and cancer progression [ 17 – 19 ]. In the context of EMS and ovarian cancer, miRNAs may serve as key regulators of tumorigenesis, influencing the cellular processes that drive the transformation of benign endometriotic lesions into malignant ovarian tumors [ 20 , 21 ]. We have identified miR-874 as a potential molecular link in the progression from EMS to CCOC. miR-874, a microRNA known to be dysregulated in various cancers including ovarian cancer, has emerged as a potential contributor to this transformation. Recent studies suggest that miR-874 is down-regulated in ovarian cancer tissues, and its reduced expression may be associated with tumor progression and poor prognosis [ 22 , 23 ]. However, the specific molecular mechanisms by which miR-874 mediates the transition from EMS to CCOC remain poorly understood. In this study, we integrated multiple datasets to explore the molecular mechanisms underlying the transition from EMS to CCOC, with a focus on the role of miR-874. Our findings indicate that miR-874 is consistently and significantly down-regulated in both EMS and CCOC, and its reduced expression is associated with poor prognosis. Bioinformatics analysis and preliminary experimental evidence suggest that miR-874 down-regulation is associated with the up-regulation of ZNF217, which in turn may promote MTBP overexpression and enhances MDM2 stability, ultimately suppressing p53 signaling and facilitating tumor progression. These findings offer new insights into the potential mechanistic link between EMS and CCOC and support the hypothesis that the miR-874–ZNF217–MTBP–MDM2–p53 axis may contribute to CCOC pathogenesis. Further in vivo and clinical studies are warranted to validate these mechanisms and to assess the therapeutic potential of miR-874 in improving outcomes for patients with CCOC.

Below is the link to the electronic supplementary material. Supplementary Material 1. Supplementary Material 2. Supplementary Material 1. Supplementary Material 2.