Identification of potential drug targets for pelvic organ prolapse using a proteome-wide Mendelian randomization approach.

The FinnGen study was a large-scale genomic project in which over 500,000 samples from Finnish biobanks were analysed, linking genetic variants with health data to better understand disease mechanisms and susceptibilities. This project involved collaboration among Finnish research institutions, biobanks, and international industry partners 17 . We extracted summary data related to FGP from the FinnGen database, which included 23,074 cases of POP and 130,160 controls. For cross-trait genetic correlation analysis with POP, we utilized the interactive Cross-Phenotype GWAS database (iCPAGdb) to explore the genetic correlation between POP and other traits. The iCPAGdb contains linkage disequilibrium (LD)-based specific association data for cross-trait enrichment analysis 18 . We utilized data from the GWAS Catalogue of the National Human Genome Research Institute and the European Bioinformatics Institute to explore genetic correlations with POP. The iCPAGdb contains information on shared signals and LD proxy single-nucleotide polymorphisms (SNPs) between the traits. We then calculated the genetic correlation between POP and these traits using LD score regression, limiting the analysis to SNPs with minor allele frequencies greater than 1%. The default 1000 Genomes European reference dataset was used for LD calculations. Finally, we performed MR using the inverse variance-weighted (IVW) approach, selecting significant and independent SNPs ( P  ≤ 5 × 10–8, R 2  < 0.001, kb = 10,000) to explore causal links between traits and POP. We utilized the FUMA platform, an online platform that integrates various resources, such as functional annotation and gene prioritization 19 . We employed default parameters in the SNP2GENE module to identify lead and candidate SNPs. The threshold for lead SNPs was set at a maximum P value of less than 5e−8, whereas candidate SNPs had a P value threshold of < 0.05. Independent significant SNPs were defined as those with an r 2  ≤ 0.6, and for lead SNPs, the r 2 threshold was set at ≤ 0.1. The reference panel was the European population from Phase 3 of the 1000 Genomes Project (1000G Phase 3 EUR). We analysed POP via multimarker analysis of genomic annotation (MAGMA), which is based on GTEx v8 and comprises 30 tissue categories. We sourced proteomics data from two extensive GWASs focused on plasma protein levels, specifically from the UKB-PPP and the deCODE Health Study. In both studies, cis-pQTLs were used as instrumental variables. The deCODE dataset contains detailed information on 4,907 plasma proteins from 35,559 Icelandic participants predominantly of European descent 20 . The UKB-PPP dataset contains GWAS data from a cohort of 54,219 individuals, specifically data on 2,940 proteins measured using the Olink Explore 3072 antibody platform 21 . We used the R package ‘TwoSampleMR’ (version 0.6.8) 22 to perform two-sample MR analyses for both the UKB-PPP and deCODE cohorts in relation to POP. To satisfy the criteria of importance, we selected SNPs with P values below 5 × 10 − ⁸. To mitigate bias from LD, precise specifications were established (r 2  = 0.01, kb = 10,000) to ensure the independence of the selected SNPs 23 . We utilized the IVW method 24 and MR-Egger regression 25 to confirm the causal link between the cis-pQTLs and POP. The Wald ratio method was used for a single genetic instrument; for multiple SNPs, IVW meta-analysis was applied to consolidate Wald estimates 26 . False discovery rate (FDR) correction was used, and an FDR P value less than 0.05 was considered indicative of a causal relationship. Cochran’s Q test was used to assess heterogeneity, with a P value < 0.05 indicating significant heterogeneity. Horizontal pleiotropy was evaluated using the intercept from MR-Egger regression, where a P value below 0.05 was interpreted as evidence of pleiotropy 27 . Finally, to ensure robustness, the Mendelian Randomization Pleiotropy RESidual Sum and Outlier (MR-PRESSO) method was used to detect and remove potential pleiotropic biases. We applied Bayesian statistical techniques to assess the likelihood of two or more traits sharing the same causal SNP 28 . During this analysis, we established the prior probabilities for SNP associations as follows: for trait 1 (p1), 1 × 10 −4 ; for trait 2 (p2), 1 × 10 −4 ; and for SNPs linked to both traits (p12), 1 × 10 −5 . A posterior probability (PPH4) of ≥ 0.75 was interpreted as strong evidence of colocalization 29 . To investigate whether genes encoding plasma proteins are enriched in POP tissues, we analysed single-cell RNA sequencing (scRNA-seq) data from the Gene Expression Omnibus (GEO) database ( https://www.ncbi.nlm.nih.gov/geo/ ). Specifically, we used the GSE151192 dataset, which contains 21 samples: 5 normal tissue samples and 16 POP samples 30 . The R package ‘Seurat’ (version 5.1.0) was used to process the scRNA-seq data 31 . Cells expressing fewer than 500 or more than 3,500 genes were excluded, as were cells with more than 10% mitochondrial gene expression and more than 3% haemoglobin gene expression. Cells with a total RNA count greater than 100,000 were excluded to ensure data quality. To reduce batch effects in the scRNA-seq data, we utilized the R package ‘Harmony’ (version 1.2.1) 32 . Clustering of cells was conducted to identify distinct cell types, followed by annotation using known marker genes to ensure accuracy. To capture horizontal pleiotropy and potential side effects not detected by MR-Egger or MR-PRESSO, we conducted a phenome-wide association study (PheWAS) using the AstraZeneca PheWAS portal ( https://azphewas.com/ ) 33 . This analysis was based on exome sequencing data from 394,695 UKB participants of European ancestry covering 11,669 phenotypes. The details of the study design can be found in the original article 33 . Evaluating protein-drug interactions is critical for determining whether a gene can serve as a viable drug target. We used the Drug Signatures Database (DSigDB) 34 to predict potential drug candidates for the identified target genes. The chemical structures of the drugs were obtained from PubChem in the structure data file format. The structures were transformed into PDB format using OpenBabel software (version 3.1.1) ( http://openbabel.org ) 35 . Protein crystal structures were sourced from the AlphaFold Protein Structure Database, which provides highly accurate protein structures comparable to those determined by advanced techniques, such as cryo-electron microscopy 36 . The protein receptors and small-molecule ligands were loaded into AutoDock Vina (version 4.2) for molecular docking 37 and visualized using PyMOL software( https://pymol.org/2/ ). The ligand-receptor pairs with binding energies lower than − 1.2 kcal/mol were considered to have strong binding affinities, with lower values indicating better docking capability.

Using data from the iCPAGdb, we performed a cross-phenotype genetic correlation analysis of the significant SNPs associated with POP ( P  ≤ 5 × 10 −8 ). We identified 75 traits that were significantly associated with POP (FDR threshold < 0.05) (Supplementary Table 1). These traits mainly include anthropometric measurements (e.g., BMI-adjusted WHR, WHR, BMI-adjusted waist circumference, vital capacity, forced expiratory volume (FEV)/forced expiratory capacity (FEC) ratio, body height, BMI-adjusted hip circumference, and appendicular lean mass), skeletal health (e.g., bone density, grip strength, spine bone size, and bone quantitative ultrasound measurement), reproductive system disorders (e.g., uterine fibroid, female reproductive system disorders, ovarian malignant epithelial tumours, endometrial cancer, endometriosis, and bone fracture), metabolic factors (e.g., triglyceride level, red blood cell distribution width, obese BMI status, and body fat percentage), and cardiovascular factors (e.g., diastolic blood pressure, pulse pressure measurement, and cardiovascular diseases). Additionally, POP was potentially associated with other conditions, including inguinal hernia, carpal tunnel syndrome, and chronic lymphocytic leukaemia, as well as traits such as apolipoprotein A1 level; asthma (adult-onset asthma, and childhood-onset asthma); eosinophil count; high-density lipoprotein cholesterol level; calcium channel blocker use; and other conditions, such as anxiety, melanoma, and varicose veins. These findings suggest that POP may be influenced by a combination of metabolic factors, skeletal health, cardiovascular factors, and reproductive system disorders, highlighting the importance of considering these factors when investigating the pathogenesis and risk management of POP. Next, we selected 16 traits from the IEU GWAS and UKB datasets for further analysis and used LD score regression to assess SNP heritability and genetic correlations with POP. Among these, seven traits surpassed the significance threshold ( P  < 0.05). Specifically, POP was significantly negatively correlated with inguinal hernia (rg = − 0.1497; P  = 4.27 × 10 −3 ), whereas positive correlations were observed for endometriosis (rg = 0.2279; P  = 6.10 × 10 −3 ), uterine fibroids (rg = 0.1872; P  = 1.12 × 10 −2 ), heel bone mineral density (rg = 0.063; P  = 1.36 × 10 −2 ), ovarian cysts (rg = 0.521; P  = 1.50 × 10 −2 ), WHR (rg = 0.105; P  = 1.98 × 10 −2 ), and the reduction or fixation of bone fractures (rg = 0.1466; P  = 3.44 × 10 −2 ) (Supplementary Table 2). To investigate the potential causal relationships between these 16 traits and POP, we performed MR analysis. Two significant risk factors were identified: FVC and body fat percentage, whereas inguinal hernia emerged as a protective factor (Supplementary Table 3). We used the FUMA platform to map and annotate POP, identifying 4,220 candidate SNPs, 136 independent significant SNPs, 40 lead SNPs, and 29 genomic risk loci (Supplementary Tables 4–7). Candidate SNPs represent potential variants related to POP, whereas independent significant SNPs are statistically identified as markers that are independently associated with the phenotype. Through MAGMA analysis, we further identified 310 genes significantly associated with POP (FDR threshold of < 0.05). These genes are likely to play important roles in POP pathogenesis and display distinct expression patterns in different tissues. Notably, these genes are highly expressed in reproductive system tissues (e.g., uterus and ovary), and significant associations have also been observed in blood and vascular tissues. These findings suggest that the genetic risk of POP is primarily concentrated in tissues related to the reproductive system, and these genes may also be involved in biological processes such as vascular function, providing valuable insights for further research into the biological basis of POP (Fig.  2 A, B , Supplementary Table 8). Fig. 2 Functional mapping and annotation of genome-wide association studies: ( A ) The gene Manhattan plot illustrates the distribution and significance of genes associated with the trait of interest across chromosomes. ( B ) Results from MAGMA analysis display gene expression across different tissues, based on data from the GTEx v8 database. Functional mapping and annotation of genome-wide association studies: ( A ) The gene Manhattan plot illustrates the distribution and significance of genes associated with the trait of interest across chromosomes. ( B ) Results from MAGMA analysis display gene expression across different tissues, based on data from the GTEx v8 database. Using our genetic instrument selection strategy, we successfully analysed 1,777 proteins from the deCODE dataset. After applying strict multiple testing correction (FDR < 0.05), we identified 7 plasma proteins with significant causal relationships with POP (Fig.  3 A, Supplementary Table 9). To validate the robustness of these findings, we independently analysed 1971 proteins in the UKB-PPP dataset and identified 13 plasma proteins that were significantly associated with POP (Fig.  3 B, Supplementary Table 10). The deCODE dataset is considered the discovery phase, whereas the UKB-PPP dataset is regarded as the validation phase. By combining the MAGMA analysis results, EFEMP1 and MFAP4 were ultimately confirmed as potential drug targets. Various sensitivity analyses, such as MR-Egger and MR-PRESSO, showed minimal evidence of pleiotropy, suggesting that the results were not significantly affected by potential bias due to horizontal pleiotropy. Furthermore, the F statistics for the genetic instruments corresponding to each protein were high, indicating that these instrumental variables had strong and reliable effects, further supporting the robustness of the causal inferences. Fig. 3 PW-MR and colocalization analysis ( A ) PW-MR results for POP using deCODE p-QTLs ( B ) PW-MR results for POP using UKB-PPP p-QTLs ( C ) Colocalization plots for EFEMP1 and MFAP4, showing their association with POP in both deCODE and UKB-PPP datasets. PW-MR and colocalization analysis ( A ) PW-MR results for POP using deCODE p-QTLs ( B ) PW-MR results for POP using UKB-PPP p-QTLs ( C ) Colocalization plots for EFEMP1 and MFAP4, showing their association with POP in both deCODE and UKB-PPP datasets. We evaluated the genetic colocalization of proteins associated with POP. EFEMP1 and MFAP4 exhibited moderate or strong colocalization (Fig.  3 C, Supplementary Table 11). EFEMP1 showed strong colocalization in both the deCODE dataset (PPH4 = 99.18%) and the UKB-PPP dataset (PPH4 = 77.02%). A causal variant (rs11899888) was identified in both datasets. MFAP4 showed moderate colocalization in the deCODE dataset (PPH4 = 99.08%) and the UKB-PPP dataset (PPH4 = 99.30%), with the same causal variant (rs139356332) identified in both datasets. These findings provide support for the use of EFEMP1 and MFAP4 as potential drug targets for POP. We analysed 21 single-cell RNA samples from normal and POP patients (Fig.  4 A). After filtering, 84,336 cells were retained, including 68,287 from POP cases and 16,049 from controls. Through uniform manifold approximation and projection (UMAP) clustering, 16 cell subpopulations were identified, representing seven major cell types (Fig.  4 B–C). Compared with the control group, the POP group presented fewer B cells and endothelial cells but more fibroblasts (Fig.  4 D). Biological process enrichment analysis of the top 20 genes with high expression revealed that fibroblasts were involved mainly in extracellular matrix (ECM) organization, collagen fibril formation, and Wnt signalling—processes linked to tissue repair and POP pathology—suggesting their key role in POP development (Fig.  4 E). Fig. 4 Single-cell RNA sequencing analysis ( A ) UMAP plot showing cell distribution colored based on their origin (normal and POP samples). ( B ) DotPlot showing the expression levels of known cell-type marker genes across different cell clusters, with color indicating average expression and dot size representing the percentage of cells expressing the gene. ( C ) Unbiased clustering of 84,336 cells reveals 16 cell subpopulations and 7 major cell types. ( D ) Proportional changes in the 7 cell types between normal and POP samples. ( E ) The left panel shows dynamic patterns of representative differentially expressed genes (DEGs); the middle panel displays a heatmap of DEGs across clusters; the right panel presents Gene Ontology (GO) enrichment results for each cluster. ( F ) DotPlot illustrating the expression of EFEMP1 and MFAP4 across different cell clusters, with color indicating average expression and dot size representing the percentage of cells expressing the gene. ( G ) UMAP plot showing the expression patterns of EFEMP1 and MFAP4 across different cell subpopulations. Single-cell RNA sequencing analysis ( A ) UMAP plot showing cell distribution colored based on their origin (normal and POP samples). ( B ) DotPlot showing the expression levels of known cell-type marker genes across different cell clusters, with color indicating average expression and dot size representing the percentage of cells expressing the gene. ( C ) Unbiased clustering of 84,336 cells reveals 16 cell subpopulations and 7 major cell types. ( D ) Proportional changes in the 7 cell types between normal and POP samples. ( E ) The left panel shows dynamic patterns of representative differentially expressed genes (DEGs); the middle panel displays a heatmap of DEGs across clusters; the right panel presents Gene Ontology (GO) enrichment results for each cluster. ( F ) DotPlot illustrating the expression of EFEMP1 and MFAP4 across different cell clusters, with color indicating average expression and dot size representing the percentage of cells expressing the gene. ( G ) UMAP plot showing the expression patterns of EFEMP1 and MFAP4 across different cell subpopulations. We analysed the distribution of EFEMP1 and MFAP4 across cell subpopulations and found significant enrichment in fibroblasts, with MFAP4 also showing greater enrichment in smooth muscle cells (Fig.  4 F, G ). EFEMP1 had a log2FC of 2.67 in fibroblasts, whereas MFAP4 had a log2FC of 1.73 in fibroblasts, both of which were statistically significant. These findings highlight the critical roles of EFEMP1 and MFAP4 in the pathophysiology of local tissues, further reinforcing their potential as therapeutic targets for POP. PheWAS analysis revealed a correlation between genetically determined protein expression and particular diseases or features, and no significant associations (genome-wide association P value < 5E−8) were detected between EFEMP1 and MFAP4 and other traits at the genomic level (Fig.  5 A, B , Supplementary Table 12). These findings suggest that drugs targeting these genes may have minimal potential side effects and horizontal pleiotropy, further supporting their use as drug targets. Fig. 5 PheWAS analysis ( A ) The Manhattan plot shows the PheWAS association results of EFEMP1 with various binary and continuous traits.( B ) The Manhattan plot shows the PheWAS association results of MFAP4 with various binary and continuous traits. PheWAS analysis ( A ) The Manhattan plot shows the PheWAS association results of EFEMP1 with various binary and continuous traits.( B ) The Manhattan plot shows the PheWAS association results of MFAP4 with various binary and continuous traits. We used the DSigDB to predict potential therapeutic drug candidates. Using a significance cut-off of P  < 0.05, we identified drugs that target both EFEMP1 and MFAP4 and considered their toxicity profiles. Three promising drug candidates were selected (Supplementary Table 13): dasatinib (CTD 00004330), progesterone (CTD 00006624), and retinoic acid (CTD 00006918). We performed molecular docking studies to assess the interactions between potential drugs and target proteins, focusing on binding affinities and energy calculations (Figs.  6 A–F). Lower binding energy values indicate more stable binding. Our studies revealed high binding affinities between the drug candidates and the two protein targets, with docking energies below − 5 kJ/mol, suggesting strong and stable binding and indicating that these drugs can effectively target genes (Supplementary Table 14). Fig. 6 Molecular docking analysis visualization using PyMOL software. ( A ) Molecular docking of Dasatinib with EFEMP1 ( B ) Molecular docking of Dasatinib with MFAP4 ( C ) Molecular docking of Progesterone with EFEMP1 ( D ) Molecular docking of Progesterone with MFAP4 ( E ) Molecular docking of Retinoic acid with EFEMP1 ( F ) Molecular docking of Retinoic acid with MFAP4. Molecular docking analysis visualization using PyMOL software. ( A ) Molecular docking of Dasatinib with EFEMP1 ( B ) Molecular docking of Dasatinib with MFAP4 ( C ) Molecular docking of Progesterone with EFEMP1 ( D ) Molecular docking of Progesterone with MFAP4 ( E ) Molecular docking of Retinoic acid with EFEMP1 ( F ) Molecular docking of Retinoic acid with MFAP4.

Initially, we employed the MAGMA tool on the FUMA platform to identify 310 genes associated with POP progression. We subsequently used various MR methods, including the Wald ratio/IVW, MR-Egger, heterogeneity tests, MR-PRESSO, and Cochran’s Q test, to deCODE and UKB-PPP data. We identified two plasma proteins, EFEMP1 and MFAP4, with significant causal relationships with POP. These methods effectively addressed confounding factors, and colocalization analysis provided strong supporting evidence. Single-cell transcriptomic analysis revealed that EFEMP1 and MFAP4 are significantly enriched in fibroblasts. Finally, potential drugs targeting these genes were predicted, and their drug development potential was confirmed through molecular docking analysis. EFEMP1 (fibulin-3), a member of the fibulin glycoprotein family, is an ECM protein involved in the formation of ECM scaffolds and promotes adhesion between cells and the ECM in connective tissue 38 , 39 . Through our MR analysis, we identified EFEMP1 as a factor that has a protective effect on POP at the plasma level. The role of fibulin-3 in pelvic organ support has been evaluated in previous studies. Female Efemp1 knockout mice presented significantly impaired pelvic organ support, with 26.9% of the animals developing obvious vaginal, perineal, and rectal prolapse 40 . Another study revealed a significant association between EFEMP1 gene variants (e.g., rs3791675) and POP as well as other connective tissue pathologies 41 . These findings suggest that the loss or dysfunction of EFEMP1 may be an important factor in the development of POP. Previous studies have also revealed the critical role of EFEMP1 in the development of POP, which primarily involves an elastin fibre homeostasis disruption, abnormal matrix metalloproteinase activity, and limitations in compensatory mechanisms. In gene knockout mice, the absence of Efemp1 led to core rupture of elastin fibres and microfibril disorganization in the vaginal wall, with the problem worsening with age. Moreover, Efemp1 deletion was accompanied by significantly increased activity of MMP-2 and MMP-9, accelerating the degradation of the ECM. Although compensatory increases in Fibulin-5 and tropoelastin occurred in the absence of Efemp1, these increases could not lead to the repair of elastin fibre function, and the ultimate result was the loss of pelvic floor support 42 . The findings of these studies suggest that the loss of EFEMP1 may promote the progression of POP through multiple mechanisms. In our study, we found specific enrichment of EFEMP1 in fibroblasts, a discovery that may help further elucidate the role of EFEMP1 in POP. Connective tissue is one of the main mechanisms of support for maintaining the position of pelvic organs near the vagina. Connective tissue consists mainly of cellular components and the ECM, with fibroblasts being the most important cellular component 30 . Fibroblasts are responsible for synthesizing ECM proteins, including type I/III collagen and fibronectin, which are essential for the connective tissue network forming the pelvic base 43 . However, fibroblasts are susceptible to damage from various factors, such as oxidative stress-induced damage, an ageing-related decline in cellular metabolic capacity 42 , and long-term mechanical stress (e.g., stress resulting from childbirth, chronic coughing, or heavy lifting), which can damage the cytoskeleton and metabolic function 44 . Such damage may affect the ability of fibroblasts to synthesize ECM proteins, leading to changes in ECM components, including altered EFEMP1 expression, which may trigger the loss or dysfunction of EFEMP1, which is potentially closely related to the occurrence of POP. Therefore, intervention strategies targeting the protection of fibroblast function and the metabolic regulation of EFEMP1 may have therapeutic value. Microfibril-associated protein 4 (MFAP4) is an ECM protein belonging to the fibrinogen-related domain superfamily 45 . Our MR analysis revealed for the first time a significant positive correlation between elevated plasma MFAP4 levels and the risk of POP. Notably, although MFAP4 has been established as a predictive biomarker for various fibrotic diseases 46 , 47 , no studies have directly investigated its association with POP. This gap in research highlights the potential unknown role of MFAP4 in pelvic floor connective tissue remodelling. The upregulation of MFAP4 expression has been confirmed to be associated with various fibrotic diseases 48 . In atrial fibrotic rats, increased levels of MFAP4 in the serum and atria have been reported 49 . Similarly, in a heart-kidney syndrome rat model, MFAP4 protein expression was significantly upregulated in the kidneys of rats with an aortic fistula over a period of more than 21 weeks. In vascular fibrosis, the expression of MFAP4 is widely upregulated in vascular tissues, which activates the ECM and promotes fibrosis in vascular tissue 50 . A chronic obstructive pulmonary disorder (COPD) cohort study revealed that MFAP4 levels in serum increase with disease severity, suggesting that MFAP4 levels can be used to estimate the severity of COPD and serve as a biomarker for disease stability 51 . In liver fibrosis, elevated MFAP4 levels in the blood and increased MFAP4 protein expression in tissues indicate a significant association between MFAP4 and the occurrence and development of liver fibrosis. Our single-cell transcriptomic analysis further revealed that MFAP4 is expressed in both fibroblasts and smooth muscle cells. POP is associated with fibrosis. Vasin et al. noted that fibrosis of the vaginal wall connective tissue predominates in POP. The authors reported increased amounts of type III collagen, decreased amounts of type I collagen, and obvious fragmentation of elastic fibres in scar areas 52 . Fibrosis is a common pathological change in various diseases and is characterized by excessive fibroblast activity leading to myofibroblast differentiation and abnormal ECM deposition, resulting in tissue fibrosis 53 . The high enrichment of MFAP4 in fibroblasts suggests that MFAP4 may play an important role in the fibrotic process associated with POP. Future research can be focused on the specific molecular mechanisms of MFAP4 in POP and explorations of its potential as a therapeutic target, which may lead to new insights for the prevention and treatment of POP. In recent years, with the in-depth study of the pathogenesis of POP, multiple potential drug treatment targets have been revealed. Among them, the matrix metalloproteinase (MMP) and tissue inhibitor of metalloproteinase (TIMP) system has gained significant attention for its key role in the ECM remodelling process. Research indicates that an imbalance in the MMP/TIMP system may lead to connective tissue defects, thereby triggering POP 54 . These findings suggest that modulating the MMP/TIMP system to block ECM degradation could be an effective strategy to intervene in the development and progression of POP. Among the numerous potential therapeutic targets, TGF-β1 has garnered considerable attention because of its unique mechanism of action. Studies have confirmed that TGF-β1 not only has significant value in predicting the severity of POP but also represents a highly promising therapeutic target 55 . Through multiple mechanisms, such as inducing fibroblast differentiation, promoting ECM deposition, and regulating cell proliferation and ECM metabolism, TGF-β1 plays an essential role in the development and progression of POP. Additionally, the cell cycle regulatory proteins p53 and p21 have been confirmed to be involved in the pathological process of POP. By regulating p21, p53 inhibits abnormal proliferation and maintains normal ECM synthesis, thereby playing a critical role in the development of POP 54 . Like the above targets, EFEMP1 and MFAP4 may also participate in the development and progression of POP by modulating ECM-related factors. Moreover, the significant causal relationship between EFEMP1 and MFAP4 at the genetic level and POP has been supported by the findings of MR studies, which, to some extent, reduce confounding factors and reverse causality, providing strong evidence for the association between the two proteins and POP. Therefore, gene-level intervention strategies based on EFEMP1 and MFAP4 may serve as new treatment directions for POP, although further exploration and validation of these findings to determine the clinical application value of these strategies are still required. This study has several notable strengths. First, we identified drug targets using MAGMA, MR, and Bayesian colocalization analyses. Second, through single-cell transcriptomics analysis, we explored the pathological mechanisms of drug targets and POP at the local tissue level. Finally, we evaluated the potential therapeutic value and side effects of these drug targets using PheWAS and molecular docking analyses. However, this study still has certain limitations. First, although the potential value of EFEMP1 and MFAP4 as drug treatment targets in POP research has been demonstrated, numerous challenges are faced in their clinical application. Currently, clinical studies on EFEMP1 and MFAP4 are relatively limited, especially the lack of large-scale clinical trials that systematically evaluate the efficacy, safety, and long-term effects of drugs that target these proteins for POP treatment. Additionally, EFEMP1 and MFAP4 are widely expressed in various tissues, which may lead to insufficient specificity of drugs targeting these proteins and an increased risk of side effects. Moreover, POP, as a multifactorial disease, involves complex interactions among genetic, hormonal, and environmental factors, and this heterogeneity may lead to significant differences in the efficacy of EFEMP1- and MFAP4-targeted therapies across different patient populations. Second, the GWAS data analysed in this study were primarily from European populations, which somewhat limits the generalizability of the findings to other ethnic groups or populations. Third, since the pQTL data used in this study were derived mainly from plasma samples, they may not fully reflect tissue-specific regulatory mechanisms; thus, further tissue-specific validation experiments are needed in the future. Finally, although we predicted potential drug-target interactions through molecular docking analysis, the feasibility and reliability of these predictions still need to be verified through subsequent in vitro and in vivo experiments. In conclusion, we identified two plasma proteins, EFEMP1 and MFAP4, with significant causal relationships with POP. These proteins may serve as therapeutic targets for POP, highlighting the need for further research to explore their roles in the pathogenesis of POP and as potential therapeutic targets.

The plasma proteins EFEMP1 and MFAP4 have significant causal relationships with POP and may represent therapeutic targets for this condition.

Pelvic organ prolapse (POP), which primarily involves uterine and vaginal prolapse (i.e., female genital prolapse [FGP]), is a common gynaecological disorder characterized by pelvic floor dysfunction. With the increasing ageing population, the global incidence of POP has been steadily increasing 1 – 3 . Epidemiological studies revealed that the global prevalence of POP in women ranges from 3 to 6% 4 , with a prevalence of 6% in women aged 20–29 years, 31% in those aged 50–59 years, and nearly 50% in women over 80 years 5 . POP is associated with a range of symptoms, including pelvic discomfort, a sensation of heaviness, vaginal bleeding, urinary dysfunction, sexual dysfunction, and vulvar infections 6 . These symptoms significantly impair quality of life and impose considerable health and economic burdens on individuals and society. Current treatment options for POP include pelvic floor muscle exercises, vaginal pessaries, and surgical repair 7 , but these treatments are often limited in efficacy and are associated with high recurrence rates. These findings indicate that physical repair alone cannot comprehensively address the issue of POP. Therefore, there is an urgent need for research into more effective treatment options for POP. Mendelian randomization (MR) is an innovative statistical method in which genetic variants are used as instrumental variables to assess the causal relationships between exposures and outcomes 8 . MR reduces confounding and reverse causality because genetic variants are randomly assigned at conception and are not influenced by environmental factors throughout life 9 . This intrinsic capacity to reduce bias highlights the dependability and essential functions of MR in genetic epidemiological research 10 . MR has already been applied in POP research. For example, previous studies have demonstrated significant causal relationships between metabolic factors such as body mass index (BMI) and waist-hip ratio (WHR) and educational attainment and POP 11 . These findings highlight the potential of MR to inform more precise prevention and treatment strategies for POP. The plasma proteome, which comprises proteins circulating in the bloodstream, regulates essential bodily functions and facilitates intertissue communication. The aberrant expression of circulating proteins is linked to pathological changes in numerous diseases, making these proteins promising therapeutic targets 12 . Research has revealed that disease-associated proteins with genetic correlations are more likely to obtain market approval as therapeutic targets 13 . Thus, MR analysis of protein quantitative trait loci (pQTLs) associated with POP could effectively elucidate the causal relationships between proteins and diseases and thereby identify new drug targets. Recent research findings suggest that systemic factors, such as lipid metabolism, metabolic disorders, hormonal changes, and inflammatory responses, play significant roles in the pathogenesis of POP 14 – 16 . Plasma proteins, key components of these systemic factors, may play a role in the initiation and progression of POP. Consequently, conducting an MR analysis to explore the causal associations between plasma proteins and POP could lead to new insights for diagnosis and treatment. The aim of this study was to identify therapeutic targets for POP using plasma proteomics. Figure  1 illustrates the study framework. Initially, we utilized the Functional Mapping and Annotation (FUMA) platform to perform functional annotation and gene prioritization for the FinnGen cohort. We subsequently applied MR analysis using Genome-Wide Association Study (GWAS) summary data from the deCODE and UKB Pharmaceutical Proteomics Project (UKB-PPP) cohorts to investigate the causal relationships between circulating proteins and POP. Fig. 1 Flowchart of the research. EUR, European; LDSC, Linkage disequilibrium score regression; MR, Mendelian randomization; PW-MR, Proteome-wide mendelian randomization. Flowchart of the research. EUR, European; LDSC, Linkage disequilibrium score regression; MR, Mendelian randomization; PW-MR, Proteome-wide mendelian randomization.

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