Shared Risk Genes and Common Molecular Pathways Between PCOS and RPL by Integrated Transcriptomic Analysis and Machine Learning

Shared Risk Genes and Common Molecular Pathways Between PCOS and RPL by Integrated Transcriptomic Analysis and Machine Learning | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Shared Risk Genes and Common Molecular Pathways Between PCOS and RPL by Integrated Transcriptomic Analysis and Machine Learning Wanzhen LI, Yunlong Li, Xu hongxia, Yamin Kong, Jiahong Tan, Aiqi Cai, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6337423/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background Polycystic ovary syndrome (PCOS) is an endocrine-related factor contributing to recurrent pregnancy loss (RPL). PCOS and RPL may share common risk genes and potential pathological mechanisms. Methods Three PCOS and three RPL datasets were obtained from the GEO database. Weighted gene co-expression network analysis (WGCNA), differential expression analysis, and three external immune gene datasets were used to identify shared immunological genes. Enrichr analysis, Gene-TF-miRNA, and Gene-pro networks suggested potential pathogenic mechanisms. Machine learning algorithms were then applied to identify the key risk gene. ROC curves and RT-qPCR tested the performance of the key gene in validation datasets for both PCOS and RPL. Gene Set Enrichment Analysis (GSEA) validated pathway changes, and immune infiltration analysis identified immune cells involved in both diseases. Conclusions This study highlighted the association of the NF-κB pathway by involvement of 19 shared immunological genes and one key risk gene, IL1RN in RPL with PCOS . It might provide a novel understanding of the molecular pathology for RPL with PCOS. Polycystic ovary syndrome Recurrent Pregnancy Loss IL1RN NF-κB pathway GEO WGCNA Machine learning Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 1. Introduction Polycystic ovary syndrome (PCOS) is commonly observed in women of reproductive age, with a prevalence ranging from 6–20%[ 1 ]. It can adversely affect multiple systems, including reproductive, endocrine, metabolic, and psychological features, as PCOS patients often exhibit a state of low-grade systemic and local inflammation. Recurrent pregnancy loss (RPL) has a complex etiology and its relationship with PCOS is still not fully understood. Although various guidelines differ in their definitions, RPL is generally defined as the loss of two or more consecutive pregnancies before 28 weeks of gestation with the same partner, including biochemical pregnancies[ 2 ]. Currently, the incidence of RPL is approximately 2–4%, with some RPL patients exhibiting immune factor abnormalities[ 3 ]. Both the RCOG and ESHRE guidelines recognize that PCOS increases the risk of RPL[ 4 , 5 ]. More researches are required on the relationship between PCOS with RPL, as studies examining their correlation are scarce and inconsistent. The pathogenic mechanisms remain unclear, and treatment approaches are still ambiguous. Liu et al. indicated that the levels of interleukin-18 (IL-18) in follicular fluid are higher in PCOS patients than in controls, especially in overweight PCOS patients, where IL-18 levels in follicular fluid are significantly higher compared to those with normal weight. This increase in IL-18 can activate the NF-κB inflammatory pathway[ 6 ]. Hu et al. found that the TLR-mediated NF-κB signaling pathway was abnormally expressed in the endometrial tissue of women with PCOS[ 7 ]. Multiple studies have demonstrated that HMGB1 activated pyroptosis through the TLR2/TLR4-NF-κB pathway to cause aseptic inflammation, leading to the occurrence and development of unexplained RPL[ 8 , 9 ]. These pieces of evidence suggest that the two diseases might share common pathogenic pathways in local and systemic inflammation. In terms of immune cells, an imbalance in the Th1/Th2 cell ratio and a decrease in the percentage of CD56 + /CD56 − NK cells and CD56 bright /CD56 dim NK cells in the late secretory endometrium of PCOS patients may play a crucial role in implantation and maintenance of pregnancy[ 10 , 11 ]. These findings have also been confirmed in the RPL patients[ 12 ]. CD4 + CD25 + Foxp3 + regulatory T cells regulate immune balance in unexplained RPL through the TLR4/nuclear factor-κB pathway[ 13 ].This suggests that NK cells and T cells may play important immunoregulatory roles in the pathogenesis of both diseases. To explore the common pathogenic mechanisms of PCOS and RPL, we integrated transcriptome data from GEO and used the "limma" package, WGCNA, and external immune gene datasets to identify shared immune genes. Enrichment analysis was performed using Enrichr, and Gene-TF-miRNA and Gene-Pro networks were constructed. Using machine learning, we identified IL1RN as a key risk gene and validated it with external datasets. Gene set enrichment analysis (GSEA) revealed common pathways, while immune infiltration analysis suggested the NF-kB pathway activation as a key factor in the co-occurrence of RPL in PCOS patients. 2. Materials and methods 2.1 Data collection and preparation To study these two diseases, we screened PCOS and RPL datasets in the GEO database ( http://www.ncbi.nlm.nih.gov/geo/ ) using the keywords "PCOS," "RPL," and "RSA." The inclusion criteria required datasets with both patients and normal controls. For PCOS, we selected GSE34526 and GSE98461 (GPL570 platform). For RPL, we chose GSE22490 and GSE26787 (GPL570 platform). GSE137684 (GPL17077 platform) was used as an external validation dataset for PCOS, and GSE139180 (GPL21282 platform) for RPL. The analysis workflow is shown in Fig. 1A, and Fig. 1B lists the dataset details. When preparing the datasets, PCA showed significant batch effects between the two disease groups. The "sva" R package was used to remove these effects, and the cleaned data was visualized with the "FactoMineR" and "factoextra" R packages. 2.2 Differential gene expression analysis After the datasets for each disease were prepared, differential expression analysis was performed using the "limma" package in R(version 4.3.0). For both PCOS and RPL, the Differentially expressed genes (DEGs) thresholds were set at P-value 1. Subsequently, heatmaps and volcano plots were used to display the DEGs for each group. In these plots, blue indicates low expression, and red indicates high expression. 2.3 Weighted gene co-expression network analysis (WGCNA) WGCNA is widely used for identifying gene modules, linking them to clinical traits, and exploring biological networks, making it a powerful tool in genomics studies[ 14 ]. WGCNA was performed on merged PCOS and RPL datasets to identify gene modules linked to clinical traits. The top 5000 genes, selected by median absolute deviation, were analyzed after removing missing values and low-variance genes. The "pickSoftThreshold" function from the "WGCNA" package was used to select soft threshold values. An adjacency matrix was created based on the scale-free topology criterion and transformed into a topological overlap matrix (TOM). Hierarchical clustering grouped genes into modules, and Pearson correlation analysis identified modules most relevant to PCOS and RPL. Key associated genes were further analyzed. 2.4 Identification of Immunological shared genes and Enrichr analysis To explore common pathogenic mechanisms of PCOS and RPL, we separately intersected the DEGs from both diseases and gene modules screened by WGCNA from both diseases. Subsequently, we downloaded immune gene sets from InnateDB database ( https://www.innatedb.com/index.jsp )[ 15 ], ImmPort database ( https://immport.org/shared/home )[ 16 ], and Immunome Database (linked on the ImmPort website) [ 17 ], and intersected them again. Eventually, we obtained a collection of immune-related shared genes. Through Enrichr website( https://maayanlab.cloud/Enrichr ), we analyzed the functional annotations and enriched pathways of these genes[ 18 ]. We aim to investigate the biological changes related to PCOS and RPL. The Enrichr visualization displays bar charts where bar length represents the P-value and color brightness indicates term significance. We identified candidate factors from terms that performed well across multiple databases, aiming to uncover common mechanisms for both diseases. 2.5 Construction of Genes-TFs-miRNAs and Genes-Protein Interaction Networks We utilized Network Analyst 3.0 ( https://www.networkanalyst.ca/ ) to analyze the regulatory interactions among shared genes, transcription factors (TFs), and microRNAs (miRNAs), as well as the interactions between genes and proteins [ 19 ]. A After inputting shared genes, we constructed a Genes-TFs-miRNA network using TF-miRNA coregulatory interactions from RegNetwork [ 20 ]. The Genes-Protein network was constructed using the STRING Interactome[ 21 ], selecting the network with the most edges. Our goal was to identify key transcription factors, miRNAs, and proteins related to both diseases and explore potential pathway connections among them. 2.6 Identification of Potential Key Risk Genes Based on Machine Learning Algorithms We used three algorithms—LASSO, SVM-RFE, and Random Forest—to further select candidate risk genes for PCOS and RPL based on previously selected shared genes. In the PCOS group, LASSO regression fitting was performed using "glmnet" package[ 22 ]. and then Random Forest was used to classify the data and select the top 5 most important genes. Additionally, the "caret" package was used to configure the parameters for the SVM-RFE algorithm, with cross-validation to evaluate the model and select the optimal feature subset [ 23 ]. The same three algorithms were applied to the RPL group. Finally, the intersection of the genes from PCOS and RPL was determined to identify the key risk genes for both diseases. 2.7 Prediction performance in validation cohorts To further assess the accuracy of the key risk gene, we selected the GSE137684 dataset for PCOS and the GSE139180 dataset for RPL as external validation datasets. We downloaded these datasets from GEO and extracted the expression matrices. For GSE137684, we used 4 healthy controls and 8 PCOS samples; for GSE139180, we used 3 healthy controls and 3 RPL samples. Using the "pROC" R package, we plotted the expression patterns of diagnostic genes in both sets and calculated the AUC (area under the ROC curve). 2.8 RT-qPCR Total RNA was extracted from endometrial tissues of PCOS, RPL, and healthy control groups using the FastPure RNA Extraction Kit (Qinkeo, China). RNA purity and concentration were assessed with a NanoDrop 2000 (Thermo Fisher Scientific, USA). Genomic DNA was removed using the Prime Script RT Reagent Kit (Takara, Japan), and RNA was reverse transcribed into cDNA using SynScript®Ⅲ RT SuperMix (Qinkeo, China). PCR was performed with ArtiCanCEO SYBR qPCR Mix (Qinkeo, China) and GAPDH as the reference gene. Primer specificity was confirmed by melting curve analysis, and gene expression was calculated using the 2^-△△CT method. 2.9 Implementation of GSEA for key risk gene After obtaining the key risk gene, we conducted single-gene set enrichment analysis (GSEA) for key risk gene in both groups using the "clusterProfiler" package[ 24 ]. The gseKEGG function was employed to perform KEGG pathway enrichment analysis on the gene lists, allowing us to compare the biological signaling pathways between the disease and control groups. Enrichment plots were utilized to display the top 5 activated and inhibited pathways for each gene in the two disease groups. 2.10 Immune cell abundance CIBERSORT analysis was performed on each group to determine the relative levels of immune cells based on gene expression data [ 25 ]. According to the CIBERSORT website ( http://cibersort.stanford.edu/ ), the LM22 signature, containing 22 annotated gene sets, was used to estimate immune cell composition for each sample. The results were visualized using boxplots and stacked bar charts to compare immune cell differences between disease and control groups. Spearman correlation analysis was then conducted to assess the correlation between immune cells and key risk genes, and the results were visualized. 3. Results 3.1 GEO information Based on our inclusion criteria, we selected four GEO datasets for analysis: GSE34526, GSE98461, GSE22490, and GSE26787. Among them, GSE34526 and GSE98461 were used as the discovery cohorts for PCOS, while GSE22490 and GSE26787 were used as the discovery cohorts for RPL. Additionally, GSE137684 and GSE139180 served as the validation cohorts for PCOS and RPL, respectively. 3.2 Identification of DEGs Before bioinformatics analysis, we addressed batch effects in the datasets using the "sva" package to ensure reliable results. The "Limma" package was then used to identify DEGs between groups (Fig. 2 A, B, E, F), with thresholds set at P-value 1. In PCOS, 191 DEGs were found, including 149 upregulated and 42 downregulated. In RPL, 303 DEGs were identified, with 221 upregulated and 82 downregulated. Volcano plots (Fig. 2 C, G) and heatmaps (Fig. 2 D, H) visualized DEGs for both diseases. 3.3 Screening for key modules by WGCNA We performed WGCNA to identify co-expression gene modules and their association with phenotypic traits. Using the "pickSoftThreshold" function from the "WGCNA" package, the optimal soft threshold was determined as 16 for both PCOS and RPL (Fig. 3 A, D). The adjacency matrix was constructed based on the scale-free topology criterion and converted to a TOM. Hierarchical clustering was done using TOM dissimilarity (Fig. 3 B, E). The ME Turquoise module, with the highest correlation, was selected for both PCOS (Fig. 3 C) and RPL (Fig. 3 F). A total of 2,696 genes were identified for PCOS, and 843 for RPL. 3.4 Analysis of the shared genes and functional enrichment To explore the common pathogenesis of PCOS and RPL, we intersected the DEGs with gene modules identified by WGCNA and immune gene sets from the InnateDB, ImmPort, and Immunome databases. This analysis identified 19 shared immune-related genes: C1QA , HLA-DRB4 , PECAM1 , PTPRCAP , TLR2 , MBP , IL1RN , TAP2 , MAP3K12 , HNRNPL , ARRB2 , CCNT1 , OLFM4 , TNFRSF11B , SEMA4A , NR3C2 , SYTL1 , SORT1 , and SEMA5B . These genes were analyzed using Enrichr for functional annotations. GO analysis showed involvement in axon guidance (Fig. 4 A), vesicle membrane formation (Fig. 4 B), and semaphorin receptor binding (Fig. 4 C). KEGG analysis linked them to axon guidance and cell adhesion molecules (Fig. 4 D), with additional information from BioCarta, NCI-Nature, TRRUST, Panther, Human Phenotype Ontology, and WikiPathway, further connecting them to NF-κB activation (Fig. 4 E), the ACEI pathway (Fig. 4 F), and microvascular complications of diabetes. These results suggest that the pathogenesis of both diseases is linked to the NF-κB and ACEI pathways, IL1 receptor binding, and MHC protein binding. Additionally, based on the NF-κB signaling pathway map (map04064) from the KEGG database ( https://www.genome.jp/kegg/ ) [ 26 ], we found that these related terms are all linked to the canonical NF-κB pathway. 3.5 Genes-TFs-miRNAs and Genes-Protein interaction networks We constructed DEG-TF-miRNA and DEG-Protein networks using NetworkAnalyst. Seventeen genes successfully became seeds, generating 316 nodes and 418 edges. The top TFs with the highest betweenness were SP1 , CTCF , and ESR1 , suggesting their potential as therapeutic targets for RPL and PCOS. Among miRNAs, hsa-miR-19a and hsa-miR-211 showed the strongest interactions (Fig. 5 A). A Protein-DEGs network was built with STRING, revealing 7 shared genes and 126 nodes. Top proteins with high betweenness included FYN , NFKB1 , RELA , CXCR4 , and TRAF6 , linking them to the NF-κB pathway and the shared pathogenesis of both diseases (Fig. 5 B). 3.6 Identify potential shared diagnostic genes based on machine learning algorithms We used LASSO, SVM-RFE, and Random Forest algorithms to identify candidate risk genes from 19 shared immunological genes. For PCOS, LASSO identified 3 genes (TNFRSF11B, NR3C2, IL1RN ) (Fig. 6 A), Random Forest selected 5 genes (Fig. 6 B), and SVM-RFE identified 2 genes (Fig. 6 E), with 3 common biomarkers found ( TNFRSF11B , NR3C2 , and IL1RN ) (Fig. 6 G). For RPL, LASSO identified 6 genes (Fig. 6 C), Random Forest selected 5 (Fig. 6 D), and SVM-RFE identified 13 (Fig. 6 F), with 8 shared biomarkers ( TLR2 , ARRB2 , OLFM4 , HNRNPL , PECAM1 , IL1RN , CCNT1 , and HLA-DRB4 ) (Fig. 6 H). IL1RN was the common risk gene for both diseases. 3.7 Prediction performance in validation cohorts Next, to evaluate the specificity and sensitivity of the risk gene for diagnosing the two diseases, we performed ROC analysis. IL1RN exhibited good diagnostic accuracy in the discovery cohorts for PCOS (AUC = 0.961) and RPL (AUC = 0.879) (Fig. 6 I, K). Additionally, we confirmed the reliability of IL1RN as a risk gene for PCOS and RPL through external validation. Similarly, IL1RN showed good diagnostic accuracy in the external validation cohorts for PCOS (AUC = 0.750) and RPL (AUC = 0.778) (Fig. 6 J, L). The results indicate that IL1RN can serve as a shared risk gene for both PCOS and RPL. 3.8 Verification of hub biomarkers To verify the mRNA of the hub gene, Endometrial samples were collected from 3 control,5 PCOS patients and 5 RPL patients. RT-qPCR results showed that the relative expression of IL1RN mRNA in the RPL and PCOS groups was significantly higher than that in the control group (Fig. 7 A, B). The primer information can be found in Supplementary Material 1. 3.9 Single-Gene GSEA for Risk Genes Single-gene GSEA analysis of IL1RN in both PCOS and RPL datasets identified upregulated pathways in PCOS related to inflammation and autoimmune diseases (Fig. 8A), and in RPL related to lipoic acid and thiamine metabolism (Fig,8B). Downregulated pathways in PCOS included linoleic acid metabolism and peroxisomes, while in RPL, they involved natural killer cell-mediated cytotoxicity. These pathways are associated with the NF-κB pathway 3.10 Immune cell abundance We utilized CIBERSORT to analyze immune cell abundance in PCOS and RPL groups. In PCOS, there was a decrease in CD8 + T cells, CD4 memory resting T cells, and resting NK cells, while Macrophages M0 increased (Fig. 8C, D). In RPL, resting Dendritic cells and Neutrophils increased (see Fig. 8E, F). IL1RN in PCOS was positively correlated with Monocytes and negatively with memory B cells; in RPL, IL1RN was positively correlated with naive B cells and negatively with activated NK cells (see Fig. 8G, H). These findings suggest that T cells, B cells, and NK cells play significant roles in immune regulation in both diseases. 4. Discussions This study explores the shared immune mechanisms between PCOS and RPL by identifying 19 common immunological genes using WGCNA and limma, constructing gene networks, and conducting enrichment, machine learning, GSEA, and immune infiltration analyses to uncover key risk genes and pathways involved in both diseases. As shown in Fig. 2 , we identified 19 shared immunological genes through DEGs and WGCNA combined with external immune datasets. As shown in Fig. 3 , We used the Enrichr database to analyze the shared risk genes, revealing that these genes are associated with the canonical NF-κB pathway. NF-κB is a family of transcription factors, including NFKB1 (p50), NFKB2 (p52), RelA (p65), RelB, and c-Rel. NFKBIA (also known as IκBα) is a regulatory protein that inhibits NF-κB activity by masking its nuclear localization signal and interfering with DNA binding.[ 27 ]. Multiple studies have shown that miRNAs such as hsa-miR-19a and hsa-miR-204 are involved in regulating the NF-κB signaling pathway[ 28 , 29 ]. As shown in Fig. 4 , those nodes associated with NF-κB pathway mentioned above are highly correlated with the shared immune genes. As shown in Fig. 5 , We identified IL1RN as a key risk gene among 19 shared immune genes and validated its differential expression in validation cohorts. IL1RN , located on chromosome 2q14, encodes interleukin-1 receptor antagonist (IL-1RA), a cytokine that inhibits IL-1α and IL-1β activities, modulating immune and inflammatory responses.[ 30 ]. It binds to the IL-1 receptor but does not transmit an activation signal, acting as a physiological inhibitor of pre-formed IL-1. IL-1RA is expressed in the endometrium and blastocysts[ 31 ]. Studies have shown that IL1RA, TNFα, and US-CRP are significantly elevated in the serum of PCOS patients, suggesting that chronic inflammation may be related to the pathogenesis of PCOS[ 32 , 33 ]. Montazeri et al. proposed that introducing IL1RA into the endometrium could serve as a diagnostic marker for infertility. IL1RA regulates the NF-κB pathway and inhibits trophoblast adhesion to endometrial cells [ 34 ]. Sabah Linjawi et al. found that the frequencies of the 2,2 and 4,2 genotypes were higher in PCOS women, while the 4,4 genotype was less frequent, and the frequency of allele 2 was increased[ 35 ]. However, this group (n = 10) was very small, so further research with a larger group is needed. IL1RN may be a key immunological risk gene for both PCOS and recurrent miscarriage. Based on above results, multiple databases indicated the association of both diseases with the NF-κB pathway. Additionally, the key immune risk gene IL1RN , encoding IL1RA, regulates the NK-κB pathway. The NF-κB pathway causes immune dysregulation in both PCOS and RPL through these mechanisms by participating in the maturation and differentiation of T cells, B cells, and NK cells[ 36 – 39 ]. Therefore, we decided to conduct single-gene enrichment analysis and immune infiltration analysis to explore the changes in pathway and immune cell regulation in these two diseases. As shown in Fig. 6 , we found that some pathways highly associated with these two diseases have been shown to be related to the NF-κB pathway. For example, glucosinolates and thiamine can both regulate the NF-κB pathway[ 40 , 41 ]. The non-steroidal anti-inflammatory drug aspirin influences the metabolism of arachidonic acid by reducing the expression of NF-κB downstream genes, particularly COX-2, thereby improving pregnancy retention rates in PCOS patients[ 42 ][ 43 ]. M A Morsy et.al found found that peroxisome proliferator-activated receptor (PPAR) α activation, associated with the inactivation of the NF-κB pathway, improves polycystic ovary syndrome symptoms in rats [ 44 ][ 45 ]. These evidences suggest that the NF-κB pathway plays a role in the pathogenesis of these two diseases. As shown in Fig. 6 , immune infiltration analysis of both diseases suggested their main association with the immune regulation of NK cells, B cells, and T cells. This indicated their potential key roles in immune regulation in these two diseases. This immune regulation may operate through the NF-κB signaling pathway to regulate the occurrence and development of immune cells. However, the present study remains certain limitations. First, the sample size included in this study is insufficient. Second, the experiment lacks more comprehensive pathway validation. Third, the study is relatively lacking in population representativeness, as it does not include independent experiments across different regions and ethnicities. Nonetheless, this study has provided a comprehensive and integrated bioinformatics analysis of the common pathogenesis of PCOS with RPL, profiles the involvement of NF-κB in these two diseases, identifies key immune risk gene, and lays a foundation for future research on these two diseases. Conclusion This study identified the common pathogenic NF-κB pathway between PCOS and RPL through comprehensive bioinformatics analysis. Additionally, it identified 19 shared immune genes and one key risk gene, IL1RN , for both diseases. These findings provide a foundation for future research and treatment of PCOS with RPL. Abbreviations PCOS Polycystic Ovary Syndrome RPL Recurrent Pregnancy Loss RSA Recurrent Spontaneous Abortion GEO Gene Expression Omnibus PCA Principal Component Analysis DEGs Differentially Expressed Genes log2FC Log2 Fold Change WGCNA Weighted Gene Co-expression Network Analysis TOM Topological Overlap Matrix IL-1β Interleukin-1 Beta NF-κB Nuclear Factor Kappa B TLR4 Toll-like Receptor 4 IRF-7 Interferon Regulatory Factor 7 TcR T-cell Receptor HMGB1 High Mobility Group Box 1 Foxp3 Forkhead Box P3 IL1RN Interleukin 1 Receptor Antagonist uNK Uterine Natural Killer cells KEGG Kyoto Encyclopedia of Genes and Genomes GSEA Gene Set Enrichment Analysis SVM-RFE Support Vector Machine-Recursive Feature Elimination STRING Search Tool for the Retrieval of Interacting Genes/Proteins miR MicroRNA GPL Gene Expression Omnibus Platform Declarations Authors’ contributions Wanzhen Li, Jinman Zhang and Baosheng Zhu designed the study. Wanzhen Li,Yunlong Li, Aiqi Cai, Youmou Fu analyzed and visualized the data. Wanzhen Li, Jiahong Tan, Yunlong Li, Baosheng Zhu contributed to writing and revising the paper. Wanzhen Li, Hongxia Xu Jiahong Tan All authors have read and agreed to the published version of the manuscript. Funding This work is supported by the Yunnan Ten Thousand Talents Plan Yunling Scholar Project (NO. YNWR-YLXZ-2019-005). Data Availability The datasets employed in our study can be acquired in the GEO repository (https://www.ncbi.nlm.nih.gov/geo/). The accession numbers are GSE34526, GSE98461, GSE137684, GSE22490, GSE26787, GSE139180. Disclosure statement The authors declare no competing interests. Ethics approval and consent to participate GEO belongs to public databases. The patients involved in the database have obtained ethical approval.The endometrial samples required for RT-qPCR were approved by the Ethics Committee of the First People's Hospital of Yunnan Province, and were collected without causing any additional effects on the patients. Consent for publication Not applicable. Author details Kunming University of Science and Technology, 650500 Kunming, China The First People's Hospital of Yunnan, 650032 Kunming, China References Legro RS, Arslanian SA, Ehrmann DA, Hoeger KM, Murad MH, Pasquali R, et al. Diagnosis and treatment of polycystic ovary syndrome: an Endocrine Society clinical practice guideline. J Clin Endocrinol Metab. 2013;98(12):4565–92. [Chinese expert consensus on diagnosis and management of recurrent spontaneous abortion (2022)]. Zhonghua Fu Chan Ke Za Zhi. 2022;57(9):653 – 67. El Hachem H, Crepaux V, May-Panloup P, Descamps P, Legendre G, Bouet PE. Recurrent pregnancy loss: current perspectives. Int J Womens Health. 2017;9:331–45. Regan L, Rai R, Saravelos S, Li T-C, Obstetricians tRCo, Gynaecologists. Recurrent MiscarriageGreen-top Guideline No. 17. 2023;130(12):e9-e39. Bender Atik R, Christiansen OB, Elson J, Kolte AM, Lewis S, Middeldorp S, et al. ESHRE guideline: recurrent pregnancy loss: an update in 2022. Hum Reprod Open. 2023;2023(1):hoad002. Liu Y, Liu H, Li Z, Fan H, Yan X, Liu X, et al. The Release of Peripheral Immune Inflammatory Cytokines Promote an Inflammatory Cascade in PCOS Patients via Altering the Follicular Microenvironment. Front Immunol. 2021;12:685724. Hu M, Zhang Y, Li X, Cui P, Sferruzzi-Perri AN, Brännström M, et al. TLR4-Associated IRF-7 and NFκB Signaling Act as a Molecular Link Between Androgen and Metformin Activities and Cytokine Synthesis in the PCOS Endometrium. J Clin Endocrinol Metab. 2021;106(4):1022–40. Taylor DD, Bohler HC, Gercel-Taylor C. Pregnancy-linked suppression of TcR signaling pathways by a circulating factor absent in recurrent spontaneous pregnancy loss (RPL). Mol Immunol. 2006;43(11):1872–80. Zou H, Yin J, Zhang Z, Xiang H, Wang J, Zhu D, et al. Destruction in maternal-fetal interface of URSA patients via the increase of the HMGB1-RAGE/TLR2/TLR4-NF-κB signaling pathway. Life Sci. 2020;250:117543. Qin L, Xu W, Li X, Meng W, Hu L, Luo Z, et al. Differential Expression Profile of Immunological Cytokines in Local Ovary in Patients with Polycystic Ovarian Syndrome: analysis by Flow Cytometry. Eur J Obstet Gynecol Reprod Biol. 2016;197:136–41. Matteo M, Serviddio G, Massenzio F, Scillitani G, Castellana L, Picca G, et al. Reduced percentage of natural killer cells associated with impaired cytokine network in the secretory endometrium of infertile women with polycystic ovary syndrome. Fertil Steril. 2010;94(6):2222-7, 7.e1-3. Acar N, Ustunel I, Demir R. Uterine natural killer (uNK) cells and their missions during pregnancy: a review. Acta Histochem. 2011;113(2):82–91. Qin S, Li L, Liu J, Zhang J, Xiao Q, Fan Y, et al. CD4(+)CD25(+)Foxp3(+) regulatory T cells regulate immune balance in unexplained recurrent spontaneous abortion via the Toll-like receptor 4/nuclear factor-κB pathway. J Int Med Res. 2020;48(12):300060520980940. Langfelder P, Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics. 2008;9:559. Breuer K, Foroushani AK, Laird MR, Chen C, Sribnaia A, Lo R, et al. InnateDB: systems biology of innate immunity and beyond–recent updates and continuing curation. Nucleic Acids Res. 2013;41(Database issue):D1228-33. Bhattacharya S, Andorf S, Gomes L, Dunn P, Schaefer H, Pontius J, et al. ImmPort: disseminating data to the public for the future of immunology. Immunol Res. 2014;58(2–3):234–9. Ortutay C, Vihinen M. Immunome: a reference set of genes and proteins for systems biology of the human immune system. Cell Immunol. 2006;244(2):87–9. Kuleshov MV, Jones MR, Rouillard AD, Fernandez NF, Duan Q, Wang Z, et al. Enrichr: a comprehensive gene set enrichment analysis web server 2016 update. Nucleic Acids Res. 2016;44(W1):W90-7. Zhou G, Soufan O, Ewald J, Hancock REW, Basu N, Xia J. NetworkAnalyst 3.0: a visual analytics platform for comprehensive gene expression profiling and meta-analysis. Nucleic Acids Res. 2019;47(W1):W234-w41. Liu ZP, Wu C, Miao H, Wu H. RegNetwork: an integrated database of transcriptional and post-transcriptional regulatory networks in human and mouse. Database (Oxford). 2015;2015. Szklarczyk D, Kirsch R, Koutrouli M, Nastou K, Mehryary F, Hachilif R, et al. The STRING database in 2023: protein-protein association networks and functional enrichment analyses for any sequenced genome of interest. Nucleic Acids Res. 2023;51(D1):D638-d46. Engebretsen S, Bohlin J. Statistical predictions with glmnet. Clin Epigenetics. 2019;11(1):123. Sanz H, Valim C, Vegas E, Oller JM, Reverter F. SVM-RFE: selection and visualization of the most relevant features through non-linear kernels. BMC Bioinformatics. 2018;19(1):432. Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005;102(43):15545–50. Newman AM, Liu CL, Green MR, Gentles AJ, Feng W, Xu Y, et al. Robust enumeration of cell subsets from tissue expression profiles. Nat Methods. 2015;12(5):453–7. Ogata H, Goto S, Sato K, Fujibuchi W, Bono H, Kanehisa M. KEGG: Kyoto Encyclopedia of Genes and Genomes. Nucleic Acids Res. 1999;27(1):29–34. Gilmore TD. Introduction to NF-kappaB: players, pathways, perspectives. Oncogene. 2006;25(51):6680–4. Gantier MP, Stunden HJ, McCoy CE, Behlke MA, Wang D, Kaparakis-Liaskos M, et al. A miR-19 regulon that controls NF-κB signaling. Nucleic Acids Res. 2012;40(16):8048–58. Song X, Zhu M, Sun Y, Liu B, Yan Z, Yin Y. MiR-204 enhances the progression of osteoarthritis by suppressing the production of IL-1β. Pharmazie. 2017;72(10):587–92. Patterson D, Jones C, Hart I, Bleskan J, Berger R, Geyer D, et al. The human interleukin-1 receptor antagonist (IL1RN) gene is located in the chromosome 2q14 region. Genomics. 1993;15(1):173–6. Simón C, Piquette GN, Frances A, Polan ML. Localization of interleukin-1 type I receptor and interleukin-1 beta in human endometrium throughout the menstrual cycle. J Clin Endocrinol Metab. 1993;77(2):549–55. Yang Yan HW, Qiao Jie, Li Meizhi. Study on the correlation between interleukin-1 receptor antagonist and tumor necrosis factor alpha, high sensitivity C-reactive protein, and pathogenesis of polycystic ovary syndrome. Chinese Journal of Birth Health and Heredity. 2007;15(5):35–7. (In Chinese). Ma Qinglian YJ, Wang Weixiang. Correlation studies of tumor necrosis factor, caseepin, interleukin 1 receptor antagonists and polycystic ovary syndrome. Maternal and Child Health Care of China. 2017;32(11):2315–7.(In Chinese). Montazeri M, Sanchez-Lopez JA, Caballero I, Maslehat Lay N, Elliott S, Fazeli A. Interleukin-1 receptor antagonist mediates toll-like receptor 3-induced inhibition of trophoblast adhesion to endometrial cells in vitro. Hum Reprod. 2016;31(9):2098–107. Linjawi S, Li TC, Laird S, Blakemore A. Interleukin-1 receptor antagonist and interleukin-1 beta polymorphisms in women with recurrent miscarriage. Fertil Steril. 2005;83(5):1549–52. Hayden MS, Ghosh S. NF-κB in immunobiology. Cell Res. 2011;21(2):223–44. Vallabhapurapu S, Karin M. Regulation and function of NF-kappaB transcription factors in the immune system. Annu Rev Immunol. 2009;27:693–733. Wang J, Yin T, Liu S. Dysregulation of immune response in PCOS organ system. Front Immunol. 2023;14:1169232. Yang X, Tian Y, Zheng L, Luu T, Kwak-Kim J. The Update Immune-Regulatory Role of Pro- and Anti-Inflammatory Cytokines in Recurrent Pregnancy Losses. Int J Mol Sci. 2022;24(1). Tibullo D, Li Volti G, Giallongo C, Grasso S, Tomassoni D, Anfuso CD, et al. Biochemical and clinical relevance of alpha lipoic acid: antioxidant and anti-inflammatory activity, molecular pathways and therapeutic potential. Inflamm Res. 2017;66(11):947–59. Ghaiad HR, Ali SO, Al-Mokaddem AK, Abdelmonem M. Regulation of PKC/TLR-4/NF-kB signaling by sulbutiamine improves diabetic nephropathy in rats. Chem Biol Interact. 2023;381:110544. Sclabas GM, Uwagawa T, Schmidt C, Hess KR, Evans DB, Abbruzzese JL, et al. Nuclear factor kappa B activation is a potential target for preventing pancreatic carcinoma by aspirin. Cancer. 2005;103(12):2485–90. Chakraborty P, Banerjee S, Saha P, Nandi SS, Sharma S, Goswami SK, et al. Aspirin and low-molecular weight heparin combination therapy effectively prevents recurrent miscarriage in hyperhomocysteinemic women. PLoS One. 2013;8(9):e74155. Morsy MA, El-Hussieny M, Zenhom NM, Nair AB, Venugopala KN, Refaie MMM. Fenofibrate ameliorates letrozole-induced polycystic ovary in rats via modulation of PPARα and TNFα/CD95 pathway. Eur Rev Med Pharmacol Sci. 2022;26(20):7359–70. Korbecki J, Bobiński R, Dutka M. Self-regulation of the inflammatory response by peroxisome proliferator-activated receptors. Inflamm Res. 2019;68(6):443–58. Additional Declarations No competing interests reported. Supplementary Files Supplementarymaterial1.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6337423","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":451871995,"identity":"f5972512-771a-4438-a0d4-8ea997a5f556","order_by":0,"name":"Wanzhen LI","email":"","orcid":"","institution":"Kunming University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Wanzhen","middleName":"","lastName":"LI","suffix":""},{"id":451871996,"identity":"5717713a-12cb-4027-9719-93934ed892ee","order_by":1,"name":"Yunlong Li","email":"","orcid":"","institution":"the First People's Hospital of Yunnan Province","correspondingAuthor":false,"prefix":"","firstName":"Yunlong","middleName":"","lastName":"Li","suffix":""},{"id":451871999,"identity":"477aef3e-65c5-4006-8679-0583435f6de6","order_by":2,"name":"Xu hongxia","email":"","orcid":"","institution":"the First People's Hospital of Yunnan Province","correspondingAuthor":false,"prefix":"","firstName":"Xu","middleName":"","lastName":"hongxia","suffix":""},{"id":451872007,"identity":"b5038e2e-950e-4018-8b2c-6637df6335b6","order_by":3,"name":"Yamin Kong","email":"","orcid":"","institution":"the First People's Hospital of Yunnan Province","correspondingAuthor":false,"prefix":"","firstName":"Yamin","middleName":"","lastName":"Kong","suffix":""},{"id":451872011,"identity":"b1fbcc13-639e-4648-b798-682df8c56362","order_by":4,"name":"Jiahong Tan","email":"","orcid":"","institution":"the First People's Hospital of Yunnan Province","correspondingAuthor":false,"prefix":"","firstName":"Jiahong","middleName":"","lastName":"Tan","suffix":""},{"id":451872017,"identity":"1da1ddc1-acab-456c-9110-5ec4e2f14b10","order_by":5,"name":"Aiqi Cai","email":"","orcid":"","institution":"the First People's Hospital of Yunnan Province","correspondingAuthor":false,"prefix":"","firstName":"Aiqi","middleName":"","lastName":"Cai","suffix":""},{"id":451872019,"identity":"bd243912-b223-4e1a-aa6f-ebb2a3628245","order_by":6,"name":"Youmou LI","email":"","orcid":"","institution":"Kunming University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Youmou","middleName":"","lastName":"LI","suffix":""},{"id":451872026,"identity":"8c964005-7917-40f9-b3ff-ef7b2a461422","order_by":7,"name":"Jinman Zhang","email":"","orcid":"","institution":"the First People's Hospital of Yunnan Province","correspondingAuthor":false,"prefix":"","firstName":"Jinman","middleName":"","lastName":"Zhang","suffix":""},{"id":451872028,"identity":"a2c7ab8b-d7a1-46a6-8e86-ff8d65ba6910","order_by":8,"name":"Baosheng Zhu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABCUlEQVRIiWNgGAWjYBACAyBmBrMkGBgfJFSwybGxtx8gWguzwYczfMZ8PGcSiNbCJjmzTS5xnoSDAV4t5uyHD38uqLhjN392j4E0D5tZepsEQwLDj4ptOLVY9qSlSc848yy5cc4ZA2MenrTcNunGA4w9Z27jdtiBHDNm3rbDycwSOQbJPBLHcttkDiQwM7bh0XL+jfFn3n+Hk9mAWg7zGPxPZ5NIMMCv5UaOgTRvw2E7Hokcw8YZCWwJBLVYzniWJs1z7HCChERaMcOHA2yGbcBAPojPL+b8yYc/89Qctpefkbz9R+I/Nnn59vaDD35U4NYCA4kNyLwDBNUDgT0xikbBKBgFo2CEAgDjHVcHGvXO9AAAAABJRU5ErkJggg==","orcid":"","institution":"the First People's Hospital of Yunnan Province","correspondingAuthor":true,"prefix":"","firstName":"Baosheng","middleName":"","lastName":"Zhu","suffix":""}],"badges":[],"createdAt":"2025-03-30 08:08:20","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6337423/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6337423/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":82175116,"identity":"e554337d-7f38-4bb1-a505-da2fdadaf47f","added_by":"auto","created_at":"2025-05-07 11:08:10","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":235554,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe flow chart for the whole design and details of GEO datasets used in the study.\u003c/strong\u003e (A). The flow chart for the whole design. (B). details of GEO datasets used in the study. PCOS, Polycystic Ovarian Syndrome; RPL, Recurrent Pregnancy Loss; WGCNA, Weighted Gene Co-expression Network Analysis; DEGs, Differentially Expressed Genes; LASSO, Least Absolute Shrinkage and Selection Operator; SVM-RFE, Support Vector Machine- Recursive Feature Elimination; GSEA, Gene Set Enrichment Analysis; ROC, Receiver Operating Characteristic; GSE, Gene Expression Omnibus Series; GEO, Gene Expression Omnibus; GPL, GEO Platform.\u003c/p\u003e","description":"","filename":"figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-6337423/v1/6e1a5c92b5a6f4a8fa066e43.png"},{"id":82178493,"identity":"0544ed03-72c2-4d30-b6bc-6d3378fc19b5","added_by":"auto","created_at":"2025-05-07 11:24:10","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":721726,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eRemoval of batch effects and identification of DEGs in PCOS.\u003c/strong\u003e (A, B). PCA plots showed the expression pattern in two datasets of PCOS before and after eliminating the batch effects. (C, D). DEG heatmap and volcano plot in PCOS group. (E, F). PCA plots showed the expression pattern in two datasets of RPL before and after removing the batch effects. (G, H). Heatmap and the volcano plot of DEGs in RPL group. DEGs, Differentially Expressed Genes; PCOS, Polycystic Ovary Syndrome; RPL, Recurrent Pregnancy Loss; PCA, Principle-component Analysis.\u003c/p\u003e","description":"","filename":"Figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-6337423/v1/6497497b80c00bc96bc4e98b.png"},{"id":82175118,"identity":"da592dc5-11df-4d44-b7ae-a60d7af9a682","added_by":"auto","created_at":"2025-05-07 11:08:10","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":944283,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eWeighted gene co-expression network analysis (WGCNA).\u003c/strong\u003e(A). Determination of soft-threshold power for PCOS. (B). Cluster dendrogram of PCOS highly connected genes in key modules. (C). Relationships between modules and traits in PCOS. Correlations and P values are included in each cell. (D). Calculation of soft-threshold power for RPL. (E). Cluster dendrogram of RPL modules with highly connected genes. (F). Relationships between modules and traits in RPL\u003c/p\u003e","description":"","filename":"Figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-6337423/v1/02f92ae54c3c89b0967da507.png"},{"id":82177027,"identity":"a8e3e0b3-f2db-4568-8dd1-746b749459bb","added_by":"auto","created_at":"2025-05-07 11:16:10","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":1469681,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eEnrichr analysis of immune-related shared genes between PCOS and RPL.\u003c/strong\u003e (A). GO Biological Process. (B). GO Cellular Component. (C). GO Molecular Function. (D). KEGG Pathway Analysis. (E). NCI-Nature. (F). WikiPathway Human. PCOS, Polycystic Ovary Syndrome; RPL, Recurrent Pregnancy Loss; GO, Gene Ontology; KEGG, Kyoto Encyclopedia of Genes and Genomes; NCI, National Cancer Institute.\u003c/p\u003e","description":"","filename":"figure4.png","url":"https://assets-eu.researchsquare.com/files/rs-6337423/v1/0dec05a98f4e97a483ea9acd.png"},{"id":82175117,"identity":"8378a747-4f3c-46d2-9abf-94cfe7f76a63","added_by":"auto","created_at":"2025-05-07 11:08:10","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":860866,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eConstruction of Genes-TFs-miRNAs and Genes-Protein interaction networks.\u003c/strong\u003e (A). The Genes-TFs-miRNAs interaction network: gradient circles from red to orange represent genes, green diamonds represent transcription factors, and blue squares represent miRNAs. (B). The Genes-Protein interaction network: the darker the color, the more important the node. TF, transcription factor; miRNA, microRNA.\u003c/p\u003e","description":"","filename":"figure5.png","url":"https://assets-eu.researchsquare.com/files/rs-6337423/v1/e84cf1d0133b2362838bf09e.png"},{"id":82175121,"identity":"8edef760-cd94-4d4c-8af9-3509684765b1","added_by":"auto","created_at":"2025-05-07 11:08:10","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":732142,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe screening of candidate PCOS and RPL risk gene using three machine learning algorithms and validation of the shared risk gene.\u003c/strong\u003e(A). Coefficient profile plot of the LASSO model for PCOS showing the final parameter selection λ (lambda). (B). Top-5 genes for PCOS based on their discriminant ability in the RF algorithm. (C). Coefficient profile plot of the LASSO model for RPL showing the final parameter selection λ (lambda). (D). Top-5 genes for RPL based on their discriminant ability in the RF algorithm. (E). Genes selected using the SVM-RFE algorithm for PCOS. (F). Genes selected using the SVM-RFE algorithm for RPL. (G). The Venn diagram showed three candidate risk genes in PCOS by intersecting the results of three algorithms. (H). The Venn diagram showed eight candidate risk genes in RPL by intersecting the results of three algorithms. (I). ROC curve of IL1RN in the training group for PCOS. (J). ROC curve of IL1RN in the validation group for PCOS. (K). ROC curve of IL1RN in the training group for RPL. (L). ROC curve of IL1RN in the validation group for RPL. PCOS, Polycystic Ovary Syndrome; RPL, Recurrent Pregnancy Loss; LASSO, Least Absolute Shrinkage and Selection Operator; SVM-RFE, Support Vector Machine-Recursive Feature Elimination; RF, Random Forest; ROC, Receiver Operating Characteristic; AUC, Area Under the Curve.\u003c/p\u003e","description":"","filename":"Figure6.png","url":"https://assets-eu.researchsquare.com/files/rs-6337423/v1/dde047c530fba4bbcb4c68a5.png"},{"id":82178492,"identity":"51f1393f-2aa3-4c01-a788-ebd30ee19b91","added_by":"auto","created_at":"2025-05-07 11:24:10","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":58276,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe mRNA expression validation between normal and PCOS, RPL tissues by RT-qPCR.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A). PCOS; (B) RPL. PCOS, Polycystic Ovary Syndrome; RPL, Recurrent Pregnancy Loss.\u003c/p\u003e","description":"","filename":"figure7.png","url":"https://assets-eu.researchsquare.com/files/rs-6337423/v1/8378a8b88711dc47ef2c9332.png"},{"id":82175125,"identity":"e227d5b3-11d1-42d3-a1af-6f12c3db4dcd","added_by":"auto","created_at":"2025-05-07 11:08:10","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":1848331,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eGSEA analysis for the single diagnostic risk gene and the immune cell composition of PCOS and RPL.\u003c/strong\u003e (A). GSEA analysis for IL1RN in the PCOS group. (B). GSEA analysis for IL1RN in the RPL group. (C). Stacked bar chart showed the characteristics of infiltrating immune cells in the PCOS group. (D). Violin plot indicated that the PCOS group exhibited significantly different types of immune cells. (E). Stacked bar chart showed the characteristics of infiltrating immune cells in the RPL group. (F). Violin plot indicated that the RPL group exhibited significantly different types of immune cells. (G). Correlation between IL1RN expression and immune cells in the PCOS group. (H). Correlation between IL1RN expression and immune cells in the RPL group. PCOS, Polycystic Ovary Syndrome; RPL, Recurrent Pregnancy Loss; GSEA, Gene Set Enrichment Analysis.\u003c/p\u003e","description":"","filename":"figure8.png","url":"https://assets-eu.researchsquare.com/files/rs-6337423/v1/9e5ecce263624c4fcbb7676c.png"},{"id":85625940,"identity":"bc6ebb31-46fe-4f8b-b912-0593eb836dbd","added_by":"auto","created_at":"2025-06-29 22:01:36","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":7446200,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6337423/v1/3b9d006b-6726-4c1e-9fc6-78b0a8db5e8c.pdf"},{"id":82175115,"identity":"784c38bc-b6ae-4f3c-9c29-be7b95dc8ca9","added_by":"auto","created_at":"2025-05-07 11:08:10","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":12799,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementarymaterial1.docx","url":"https://assets-eu.researchsquare.com/files/rs-6337423/v1/29a95f2a921b5325d34b5ea2.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Shared Risk Genes and Common Molecular Pathways Between PCOS and RPL by Integrated Transcriptomic Analysis and Machine Learning","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003ePolycystic ovary syndrome (PCOS) is commonly observed in women of reproductive age, with a prevalence ranging from 6\u0026ndash;20%[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. It can adversely affect multiple systems, including reproductive, endocrine, metabolic, and psychological features, as PCOS patients often exhibit a state of low-grade systemic and local inflammation. Recurrent pregnancy loss (RPL) has a complex etiology and its relationship with PCOS is still not fully understood. Although various guidelines differ in their definitions, RPL is generally defined as the loss of two or more consecutive pregnancies before 28 weeks of gestation with the same partner, including biochemical pregnancies[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. Currently, the incidence of RPL is approximately 2\u0026ndash;4%, with some RPL patients exhibiting immune factor abnormalities[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Both the RCOG and ESHRE guidelines recognize that PCOS increases the risk of RPL[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. More researches are required on the relationship between PCOS with RPL, as studies examining their correlation are scarce and inconsistent. The pathogenic mechanisms remain unclear, and treatment approaches are still ambiguous. \u003cem\u003eLiu et al.\u003c/em\u003e indicated that the levels of interleukin-18 (IL-18) in follicular fluid are higher in PCOS patients than in controls, especially in overweight PCOS patients, where IL-18 levels in follicular fluid are significantly higher compared to those with normal weight. This increase in IL-18 can activate the NF-κB inflammatory pathway[\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. \u003cem\u003eHu et al.\u003c/em\u003e found that the TLR-mediated NF-κB signaling pathway was abnormally expressed in the endometrial tissue of women with PCOS[\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. Multiple studies have demonstrated that \u003cem\u003eHMGB1\u003c/em\u003e activated pyroptosis through the TLR2/TLR4-NF-κB pathway to cause aseptic inflammation, leading to the occurrence and development of unexplained RPL[\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. These pieces of evidence suggest that the two diseases might share common pathogenic pathways in local and systemic inflammation. In terms of immune cells, an imbalance in the Th1/Th2 cell ratio and a decrease in the percentage of CD56\u003csup\u003e+\u003c/sup\u003e/CD56\u003csup\u003e\u0026minus;\u003c/sup\u003eNK cells and CD56\u003csup\u003ebright\u003c/sup\u003e/CD56\u003csup\u003edim\u003c/sup\u003e NK cells in the late secretory endometrium of PCOS patients may play a crucial role in implantation and maintenance of pregnancy[\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. These findings have also been confirmed in the RPL patients[\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. CD4\u003csup\u003e+\u003c/sup\u003eCD25\u003csup\u003e+\u003c/sup\u003eFoxp3\u003csup\u003e+\u003c/sup\u003e regulatory T cells regulate immune balance in unexplained RPL through the TLR4/nuclear factor-κB pathway[\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e].This suggests that NK cells and T cells may play important immunoregulatory roles in the pathogenesis of both diseases.\u003c/p\u003e \u003cp\u003eTo explore the common pathogenic mechanisms of PCOS and RPL, we integrated transcriptome data from GEO and used the \"limma\" package, WGCNA, and external immune gene datasets to identify shared immune genes. Enrichment analysis was performed using Enrichr, and Gene-TF-miRNA and Gene-Pro networks were constructed. Using machine learning, we identified \u003cem\u003eIL1RN\u003c/em\u003e as a key risk gene and validated it with external datasets. Gene set enrichment analysis (GSEA) revealed common pathways, while immune infiltration analysis suggested the NF-kB pathway activation as a key factor in the co-occurrence of RPL in PCOS patients.\u003c/p\u003e"},{"header":"2. Materials and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\n \u003ch2\u003e2.1 Data collection and preparation\u003c/h2\u003e\n \u003cp\u003eTo study these two diseases, we screened PCOS and RPL datasets in the GEO database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.ncbi.nlm.nih.gov/geo/\u003c/span\u003e\u003c/span\u003e) using the keywords \u0026quot;PCOS,\u0026quot; \u0026quot;RPL,\u0026quot; and \u0026quot;RSA.\u0026quot; The inclusion criteria required datasets with both patients and normal controls. For PCOS, we selected GSE34526 and GSE98461 (GPL570 platform). For RPL, we chose GSE22490 and GSE26787 (GPL570 platform). GSE137684 (GPL17077 platform) was used as an external validation dataset for PCOS, and GSE139180 (GPL21282 platform) for RPL. The analysis workflow is shown in Fig.\u0026nbsp;1A, and Fig.\u0026nbsp;1B lists the dataset details.\u003c/p\u003e\n \u003cp\u003eWhen preparing the datasets, PCA showed significant batch effects between the two disease groups. The \u0026quot;sva\u0026quot; R package was used to remove these effects, and the cleaned data was visualized with the \u0026quot;FactoMineR\u0026quot; and \u0026quot;factoextra\u0026quot; R packages.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\n \u003ch2\u003e2.2 Differential gene expression analysis\u003c/h2\u003e\n \u003cp\u003eAfter the datasets for each disease were prepared, differential expression analysis was performed using the \u0026quot;limma\u0026quot; package in R(version 4.3.0). For both PCOS and RPL, the Differentially expressed genes (DEGs) thresholds were set at P-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05 and |log\u003csub\u003e2\u003c/sub\u003eFC (fold change)| \u0026gt; 1. Subsequently, heatmaps and volcano plots were used to display the DEGs for each group. In these plots, blue indicates low expression, and red indicates high expression.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\n \u003ch2\u003e2.3 Weighted gene co-expression network analysis (WGCNA)\u003c/h2\u003e\n \u003cp\u003eWGCNA is widely used for identifying gene modules, linking them to clinical traits, and exploring biological networks, making it a powerful tool in genomics studies[\u003cspan class=\"CitationRef\"\u003e14\u003c/span\u003e]. WGCNA was performed on merged PCOS and RPL datasets to identify gene modules linked to clinical traits. The top 5000 genes, selected by median absolute deviation, were analyzed after removing missing values and low-variance genes. The \u0026quot;pickSoftThreshold\u0026quot; function from the \u0026quot;WGCNA\u0026quot; package was used to select soft threshold values. An adjacency matrix was created based on the scale-free topology criterion and transformed into a topological overlap matrix (TOM). Hierarchical clustering grouped genes into modules, and Pearson correlation analysis identified modules most relevant to PCOS and RPL. Key associated genes were further analyzed.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\n \u003ch2\u003e2.4 Identification of Immunological shared genes and Enrichr analysis\u003c/h2\u003e\n \u003cp\u003eTo explore common pathogenic mechanisms of PCOS and RPL, we separately intersected the DEGs from both diseases and gene modules screened by WGCNA from both diseases. Subsequently, we downloaded immune gene sets from InnateDB database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.innatedb.com/index.jsp\u003c/span\u003e\u003c/span\u003e)[\u003cspan class=\"CitationRef\"\u003e15\u003c/span\u003e], ImmPort database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://immport.org/shared/home\u003c/span\u003e\u003c/span\u003e)[\u003cspan class=\"CitationRef\"\u003e16\u003c/span\u003e], and Immunome Database (linked on the ImmPort website) [\u003cspan class=\"CitationRef\"\u003e17\u003c/span\u003e], and intersected them again. Eventually, we obtained a collection of immune-related shared genes. Through Enrichr website(\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://maayanlab.cloud/Enrichr\u003c/span\u003e\u003c/span\u003e), we analyzed the functional annotations and enriched pathways of these genes[\u003cspan class=\"CitationRef\"\u003e18\u003c/span\u003e]. We aim to investigate the biological changes related to PCOS and RPL. The Enrichr visualization displays bar charts where bar length represents the P-value and color brightness indicates term significance. We identified candidate factors from terms that performed well across multiple databases, aiming to uncover common mechanisms for both diseases.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e\n \u003ch2\u003e2.5 Construction of Genes-TFs-miRNAs and Genes-Protein Interaction Networks\u003c/h2\u003e\n \u003cp\u003eWe utilized Network Analyst 3.0 (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.networkanalyst.ca/\u003c/span\u003e\u003c/span\u003e) to analyze the regulatory interactions among shared genes, transcription factors (TFs), and microRNAs (miRNAs), as well as the interactions between genes and proteins [\u003cspan class=\"CitationRef\"\u003e19\u003c/span\u003e]. A After inputting shared genes, we constructed a Genes-TFs-miRNA network using TF-miRNA coregulatory interactions from RegNetwork [\u003cspan class=\"CitationRef\"\u003e20\u003c/span\u003e]. The Genes-Protein network was constructed using the STRING Interactome[\u003cspan class=\"CitationRef\"\u003e21\u003c/span\u003e], selecting the network with the most edges. Our goal was to identify key transcription factors, miRNAs, and proteins related to both diseases and explore potential pathway connections among them.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\n \u003ch2\u003e2.6 Identification of Potential Key Risk Genes Based on Machine Learning Algorithms\u003c/h2\u003e\n \u003cp\u003eWe used three algorithms\u0026mdash;LASSO, SVM-RFE, and Random Forest\u0026mdash;to further select candidate risk genes for PCOS and RPL based on previously selected shared genes. In the PCOS group, LASSO regression fitting was performed using \u0026quot;glmnet\u0026quot; package[\u003cspan class=\"CitationRef\"\u003e22\u003c/span\u003e]. and then Random Forest was used to classify the data and select the top 5 most important genes. Additionally, the \u0026quot;caret\u0026quot; package was used to configure the parameters for the SVM-RFE algorithm, with cross-validation to evaluate the model and select the optimal feature subset [\u003cspan class=\"CitationRef\"\u003e23\u003c/span\u003e]. The same three algorithms were applied to the RPL group. Finally, the intersection of the genes from PCOS and RPL was determined to identify the key risk genes for both diseases.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\n \u003ch2\u003e2.7 Prediction performance in validation cohorts\u003c/h2\u003e\n \u003cp\u003eTo further assess the accuracy of the key risk gene, we selected the GSE137684 dataset for PCOS and the GSE139180 dataset for RPL as external validation datasets. We downloaded these datasets from GEO and extracted the expression matrices. For GSE137684, we used 4 healthy controls and 8 PCOS samples; for GSE139180, we used 3 healthy controls and 3 RPL samples. Using the \u0026quot;pROC\u0026quot; R package, we plotted the expression patterns of diagnostic genes in both sets and calculated the AUC (area under the ROC curve).\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\n \u003ch2\u003e2.8 RT-qPCR\u003c/h2\u003e\n \u003cp\u003eTotal RNA was extracted from endometrial tissues of PCOS, RPL, and healthy control groups using the FastPure RNA Extraction Kit (Qinkeo, China). RNA purity and concentration were assessed with a NanoDrop 2000 (Thermo Fisher Scientific, USA). Genomic DNA was removed using the Prime Script RT Reagent Kit (Takara, Japan), and RNA was reverse transcribed into cDNA using SynScript\u0026reg;Ⅲ RT SuperMix (Qinkeo, China). PCR was performed with ArtiCanCEO SYBR qPCR Mix (Qinkeo, China) and GAPDH as the reference gene. Primer specificity was confirmed by melting curve analysis, and gene expression was calculated using the 2^-△△CT method.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\n \u003ch2\u003e2.9 Implementation of GSEA for key risk gene\u003c/h2\u003e\n \u003cp\u003eAfter obtaining the key risk gene, we conducted single-gene set enrichment analysis (GSEA) for key risk gene in both groups using the \u0026quot;clusterProfiler\u0026quot; package[\u003cspan class=\"CitationRef\"\u003e24\u003c/span\u003e]. The gseKEGG function was employed to perform KEGG pathway enrichment analysis on the gene lists, allowing us to compare the biological signaling pathways between the disease and control groups. Enrichment plots were utilized to display the top 5 activated and inhibited pathways for each gene in the two disease groups.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\n \u003ch2\u003e2.10 Immune cell abundance\u003c/h2\u003e\n \u003cp\u003eCIBERSORT analysis was performed on each group to determine the relative levels of immune cells based on gene expression data [\u003cspan class=\"CitationRef\"\u003e25\u003c/span\u003e]. According to the CIBERSORT website (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://cibersort.stanford.edu/\u003c/span\u003e\u003c/span\u003e), the LM22 signature, containing 22 annotated gene sets, was used to estimate immune cell composition for each sample. The results were visualized using boxplots and stacked bar charts to compare immune cell differences between disease and control groups. Spearman correlation analysis was then conducted to assess the correlation between immune cells and key risk genes, and the results were visualized.\u003c/p\u003e\n\u003c/div\u003e"},{"header":"3. Results","content":"\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\n \u003ch2\u003e3.1 GEO information\u003c/h2\u003e\n \u003cp\u003eBased on our inclusion criteria, we selected four GEO datasets for analysis: GSE34526, GSE98461, GSE22490, and GSE26787. Among them, GSE34526 and GSE98461 were used as the discovery cohorts for PCOS, while GSE22490 and GSE26787 were used as the discovery cohorts for RPL. Additionally, GSE137684 and GSE139180 served as the validation cohorts for PCOS and RPL, respectively.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\n \u003ch2\u003e3.2 Identification of DEGs\u003c/h2\u003e\n \u003cp\u003eBefore bioinformatics analysis, we addressed batch effects in the datasets using the \u0026quot;sva\u0026quot; package to ensure reliable results. The \u0026quot;Limma\u0026quot; package was then used to identify DEGs between groups (Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eA, B, E, F), with thresholds set at P-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05 and |log2FC| \u0026gt; 1. In PCOS, 191 DEGs were found, including 149 upregulated and 42 downregulated. In RPL, 303 DEGs were identified, with 221 upregulated and 82 downregulated. Volcano plots (Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eC, G) and heatmaps (Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eD, H) visualized DEGs for both diseases.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec16\" class=\"Section2\"\u003e\n \u003ch2\u003e3.3 Screening for key modules by WGCNA\u003c/h2\u003e\n \u003cp\u003eWe performed WGCNA to identify co-expression gene modules and their association with phenotypic traits. Using the \u0026quot;pickSoftThreshold\u0026quot; function from the \u0026quot;WGCNA\u0026quot; package, the optimal soft threshold was determined as 16 for both PCOS and RPL (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eA, D). The adjacency matrix was constructed based on the scale-free topology criterion and converted to a TOM. Hierarchical clustering was done using TOM dissimilarity (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eB, E). The ME Turquoise module, with the highest correlation, was selected for both PCOS (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eC) and RPL (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eF). A total of 2,696 genes were identified for PCOS, and 843 for RPL.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec17\" class=\"Section2\"\u003e\n \u003ch2\u003e3.4 Analysis of the shared genes and functional enrichment\u003c/h2\u003e\n \u003cp\u003eTo explore the common pathogenesis of PCOS and RPL, we intersected the DEGs with gene modules identified by WGCNA and immune gene sets from the InnateDB, ImmPort, and Immunome databases. This analysis identified 19 shared immune-related genes: \u003cem\u003eC1QA\u003c/em\u003e, \u003cem\u003eHLA-DRB4\u003c/em\u003e, \u003cem\u003ePECAM1\u003c/em\u003e, \u003cem\u003ePTPRCAP\u003c/em\u003e, \u003cem\u003eTLR2\u003c/em\u003e, \u003cem\u003eMBP\u003c/em\u003e, \u003cem\u003eIL1RN\u003c/em\u003e, \u003cem\u003eTAP2\u003c/em\u003e, \u003cem\u003eMAP3K12\u003c/em\u003e, \u003cem\u003eHNRNPL\u003c/em\u003e, \u003cem\u003eARRB2\u003c/em\u003e, \u003cem\u003eCCNT1\u003c/em\u003e, \u003cem\u003eOLFM4\u003c/em\u003e, \u003cem\u003eTNFRSF11B\u003c/em\u003e, \u003cem\u003eSEMA4A\u003c/em\u003e, \u003cem\u003eNR3C2\u003c/em\u003e, \u003cem\u003eSYTL1\u003c/em\u003e, \u003cem\u003eSORT1\u003c/em\u003e, and \u003cem\u003eSEMA5B\u003c/em\u003e. These genes were analyzed using Enrichr for functional annotations. GO analysis showed involvement in axon guidance (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003eA), vesicle membrane formation (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003eB), and semaphorin receptor binding (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003eC). KEGG analysis linked them to axon guidance and cell adhesion molecules (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003eD), with additional information from BioCarta, NCI-Nature, TRRUST, Panther, Human Phenotype Ontology, and WikiPathway, further connecting them to NF-\u0026kappa;B activation (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003eE), the ACEI pathway (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003eF), and microvascular complications of diabetes. These results suggest that the pathogenesis of both diseases is linked to the NF-\u0026kappa;B and ACEI pathways, IL1 receptor binding, and MHC protein binding. Additionally, based on the NF-\u0026kappa;B signaling pathway map (map04064) from the KEGG database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.genome.jp/kegg/\u003c/span\u003e\u003c/span\u003e) [\u003cspan class=\"CitationRef\"\u003e26\u003c/span\u003e], we found that these related terms are all linked to the canonical NF-\u0026kappa;B pathway.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec18\" class=\"Section2\"\u003e\n \u003ch2\u003e3.5 Genes-TFs-miRNAs and Genes-Protein interaction networks\u003c/h2\u003e\n \u003cp\u003eWe constructed DEG-TF-miRNA and DEG-Protein networks using NetworkAnalyst. Seventeen genes successfully became seeds, generating 316 nodes and 418 edges. The top TFs with the highest betweenness were \u003cem\u003eSP1\u003c/em\u003e, \u003cem\u003eCTCF\u003c/em\u003e, and \u003cem\u003eESR1\u003c/em\u003e, suggesting their potential as therapeutic targets for RPL and PCOS. Among miRNAs, hsa-miR-19a and hsa-miR-211 showed the strongest interactions (Fig. \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003eA). A Protein-DEGs network was built with STRING, revealing 7 shared genes and 126 nodes. Top proteins with high betweenness included \u003cem\u003eFYN\u003c/em\u003e, \u003cem\u003eNFKB1\u003c/em\u003e, \u003cem\u003eRELA\u003c/em\u003e, \u003cem\u003eCXCR4\u003c/em\u003e, and \u003cem\u003eTRAF6\u003c/em\u003e, linking them to the NF-\u0026kappa;B pathway and the shared pathogenesis of both diseases (Fig. \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003eB).\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec19\" class=\"Section2\"\u003e\n \u003ch2\u003e3.6 Identify potential shared diagnostic genes based on machine learning algorithms\u003c/h2\u003e\n \u003cp\u003eWe used LASSO, SVM-RFE, and Random Forest algorithms to identify candidate risk genes from 19 shared immunological genes. For PCOS, LASSO identified 3 genes (TNFRSF11B, NR3C2, \u003cem\u003eIL1RN\u003c/em\u003e) (Fig. \u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eA), Random Forest selected 5 genes (Fig. \u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eB), and SVM-RFE identified 2 genes (Fig. \u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eE), with 3 common biomarkers found (\u003cem\u003eTNFRSF11B\u003c/em\u003e, \u003cem\u003eNR3C2\u003c/em\u003e, and \u003cem\u003eIL1RN\u003c/em\u003e) (Fig. \u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eG). For RPL, LASSO identified 6 genes (Fig. \u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eC), Random Forest selected 5 (Fig. \u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eD), and SVM-RFE identified 13 (Fig. \u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eF), with 8 shared biomarkers (\u003cem\u003eTLR2\u003c/em\u003e, \u003cem\u003eARRB2\u003c/em\u003e, \u003cem\u003eOLFM4\u003c/em\u003e, \u003cem\u003eHNRNPL\u003c/em\u003e, \u003cem\u003ePECAM1\u003c/em\u003e, \u003cem\u003eIL1RN\u003c/em\u003e, \u003cem\u003eCCNT1\u003c/em\u003e, and \u003cem\u003eHLA-DRB4\u003c/em\u003e) (Fig. \u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eH). \u003cem\u003eIL1RN\u003c/em\u003e was the common risk gene for both diseases.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec20\" class=\"Section2\"\u003e\n \u003ch2\u003e3.7 Prediction performance in validation cohorts\u003c/h2\u003e\n \u003cp\u003eNext, to evaluate the specificity and sensitivity of the risk gene for diagnosing the two diseases, we performed ROC analysis. \u003cem\u003eIL1RN\u003c/em\u003e exhibited good diagnostic accuracy in the discovery cohorts for PCOS (AUC\u0026thinsp;=\u0026thinsp;0.961) and RPL (AUC\u0026thinsp;=\u0026thinsp;0.879) (Fig. \u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eI, K). Additionally, we confirmed the reliability of \u003cem\u003eIL1RN\u003c/em\u003e as a risk gene for PCOS and RPL through external validation. Similarly, \u003cem\u003eIL1RN\u003c/em\u003e showed good diagnostic accuracy in the external validation cohorts for PCOS (AUC\u0026thinsp;=\u0026thinsp;0.750) and RPL (AUC\u0026thinsp;=\u0026thinsp;0.778) (Fig. \u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eJ, L). The results indicate that \u003cem\u003eIL1RN\u003c/em\u003e can serve as a shared risk gene for both PCOS and RPL.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec21\" class=\"Section2\"\u003e\n \u003ch2\u003e3.8 Verification of hub biomarkers\u003c/h2\u003e\n \u003cp\u003eTo verify the mRNA of the hub gene, Endometrial samples were collected from 3 control,5 PCOS patients and 5 RPL patients. RT-qPCR results showed that the relative expression of IL1RN mRNA in the RPL and PCOS groups was significantly higher than that in the control group (Fig. \u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003eA, B). The primer information can be found in Supplementary Material 1.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec22\" class=\"Section2\"\u003e\n \u003ch2\u003e3.9 Single-Gene GSEA for Risk Genes\u003c/h2\u003e\n \u003cp\u003eSingle-gene GSEA analysis of \u003cem\u003eIL1RN\u003c/em\u003e in both PCOS and RPL datasets identified upregulated pathways in PCOS related to inflammation and autoimmune diseases (Fig. 8A), and in RPL related to lipoic acid and thiamine metabolism (Fig,8B). Downregulated pathways in PCOS included linoleic acid metabolism and peroxisomes, while in RPL, they involved natural killer cell-mediated cytotoxicity. These pathways are associated with the NF-\u0026kappa;B pathway\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec23\" class=\"Section2\"\u003e\n \u003ch2\u003e3.10 Immune cell abundance\u003c/h2\u003e\n \u003cp\u003eWe utilized CIBERSORT to analyze immune cell abundance in PCOS and RPL groups. In PCOS, there was a decrease in CD8\u0026thinsp;+\u0026thinsp;T cells, CD4 memory resting T cells, and resting NK cells, while Macrophages M0 increased (Fig. 8C, D). In RPL, resting Dendritic cells and Neutrophils increased (see Fig. 8E, F). \u003cem\u003eIL1RN\u003c/em\u003e in PCOS was positively correlated with Monocytes and negatively with memory B cells; in RPL, \u003cem\u003eIL1RN\u003c/em\u003e was positively correlated with naive B cells and negatively with activated NK cells (see Fig. 8G, H). These findings suggest that T cells, B cells, and NK cells play significant roles in immune regulation in both diseases.\u003c/p\u003e\n\u003c/div\u003e"},{"header":"4. Discussions","content":"\u003cp\u003eThis study explores the shared immune mechanisms between PCOS and RPL by identifying 19 common immunological genes using WGCNA and limma, constructing gene networks, and conducting enrichment, machine learning, GSEA, and immune infiltration analyses to uncover key risk genes and pathways involved in both diseases.\u003c/p\u003e \u003cp\u003eAs shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e2\u003c/span\u003e, we identified 19 shared immunological genes through DEGs and WGCNA combined with external immune datasets. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e3\u003c/span\u003e, We used the Enrichr database to analyze the shared risk genes, revealing that these genes are associated with the canonical NF-κB pathway. NF-κB is a family of transcription factors, including NFKB1 (p50), NFKB2 (p52), RelA (p65), RelB, and c-Rel. NFKBIA (also known as IκBα) is a regulatory protein that inhibits NF-κB activity by masking its nuclear localization signal and interfering with DNA binding.[\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. Multiple studies have shown that miRNAs such as hsa-miR-19a and hsa-miR-204 are involved in regulating the NF-κB signaling pathway[\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e]. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e4\u003c/span\u003e, those nodes associated with NF-κB pathway mentioned above are highly correlated with the shared immune genes.\u003c/p\u003e \u003cp\u003eAs shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e5\u003c/span\u003e, We identified \u003cem\u003eIL1RN\u003c/em\u003e as a key risk gene among 19 shared immune genes and validated its differential expression in validation cohorts. \u003cem\u003eIL1RN\u003c/em\u003e, located on chromosome 2q14, encodes interleukin-1 receptor antagonist (IL-1RA), a cytokine that inhibits IL-1α and IL-1β activities, modulating immune and inflammatory responses.[\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. It binds to the IL-1 receptor but does not transmit an activation signal, acting as a physiological inhibitor of pre-formed IL-1. IL-1RA is expressed in the endometrium and blastocysts[\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]. Studies have shown that IL1RA, TNFα, and US-CRP are significantly elevated in the serum of PCOS patients, suggesting that chronic inflammation may be related to the pathogenesis of PCOS[\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. \u003cem\u003eMontazeri et al.\u003c/em\u003e proposed that introducing IL1RA into the endometrium could serve as a diagnostic marker for infertility. IL1RA regulates the NF-κB pathway and inhibits trophoblast adhesion to endometrial cells [\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]. \u003cem\u003eSabah Linjawi et al.\u003c/em\u003e found that the frequencies of the 2,2 and 4,2 genotypes were higher in PCOS women, while the 4,4 genotype was less frequent, and the frequency of allele 2 was increased[\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e]. However, this group (n = 10) was very small, so further research with a larger group is needed. \u003cem\u003eIL1RN\u003c/em\u003e may be a key immunological risk gene for both PCOS and recurrent miscarriage.\u003c/p\u003e \u003cp\u003eBased on above results, multiple databases indicated the association of both diseases with the NF-κB pathway. Additionally, the key immune risk gene \u003cem\u003eIL1RN\u003c/em\u003e, encoding IL1RA, regulates the NK-κB pathway. The NF-κB pathway causes immune dysregulation in both PCOS and RPL through these mechanisms by participating in the maturation and differentiation of T cells, B cells, and NK cells[\u003cspan additionalcitationids=\"CR37 CR38\" citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e–\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e]. Therefore, we decided to conduct single-gene enrichment analysis and immune infiltration analysis to explore the changes in pathway and immune cell regulation in these two diseases. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e6\u003c/span\u003e, we found that some pathways highly associated with these two diseases have been shown to be related to the NF-κB pathway. For example, glucosinolates and thiamine can both regulate the NF-κB pathway[\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e, \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e]. The non-steroidal anti-inflammatory drug aspirin influences the metabolism of arachidonic acid by reducing the expression of NF-κB downstream genes, particularly COX-2, thereby improving pregnancy retention rates in PCOS patients[\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e][\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e]. \u003cem\u003eM A Morsy et.al\u003c/em\u003e found found that peroxisome proliferator-activated receptor (PPAR) α activation, associated with the inactivation of the NF-κB pathway, improves polycystic ovary syndrome symptoms in rats [\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e][\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e]. These evidences suggest that the NF-κB pathway plays a role in the pathogenesis of these two diseases. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e6\u003c/span\u003e, immune infiltration analysis of both diseases suggested their main association with the immune regulation of NK cells, B cells, and T cells. This indicated their potential key roles in immune regulation in these two diseases. This immune regulation may operate through the NF-κB signaling pathway to regulate the occurrence and development of immune cells.\u003c/p\u003e \u003cp\u003eHowever, the present study remains certain limitations. First, the sample size included in this study is insufficient. Second, the experiment lacks more comprehensive pathway validation. Third, the study is relatively lacking in population representativeness, as it does not include independent experiments across different regions and ethnicities. Nonetheless, this study has provided a comprehensive and integrated bioinformatics analysis of the common pathogenesis of PCOS with RPL, profiles the involvement of NF-κB in these two diseases, identifies key immune risk gene, and lays a foundation for future research on these two diseases.\u003c/p\u003e "},{"header":"Conclusion","content":"\u003cp\u003eThis study identified the common pathogenic NF-κB pathway between PCOS and RPL through comprehensive bioinformatics analysis. Additionally, it identified 19 shared immune genes and one key risk gene, \u003cem\u003eIL1RN\u003c/em\u003e, for both diseases. These findings provide a foundation for future research and treatment of PCOS with RPL.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003ePCOS Polycystic Ovary Syndrome\u003c/p\u003e\n\u003cp\u003eRPL Recurrent Pregnancy Loss\u003c/p\u003e\n\u003cp\u003eRSA Recurrent Spontaneous Abortion\u003c/p\u003e\n\u003cp\u003eGEO Gene Expression Omnibus\u003c/p\u003e\n\u003cp\u003ePCA Principal Component Analysis\u003c/p\u003e\n\u003cp\u003eDEGs Differentially Expressed Genes\u003c/p\u003e\n\u003cp\u003elog2FC Log2 Fold Change\u003c/p\u003e\n\u003cp\u003eWGCNA Weighted Gene Co-expression Network Analysis\u003c/p\u003e\n\u003cp\u003eTOM Topological Overlap Matrix\u003c/p\u003e\n\u003cp\u003eIL-1β Interleukin-1 Beta\u003c/p\u003e\n\u003cp\u003eNF-κB Nuclear Factor Kappa B\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTLR4 Toll-like Receptor 4\u003c/p\u003e\n\u003cp\u003eIRF-7 Interferon Regulatory Factor 7\u003c/p\u003e\n\u003cp\u003eTcR T-cell Receptor\u003c/p\u003e\n\u003cp\u003eHMGB1 High Mobility Group Box 1\u003c/p\u003e\n\u003cp\u003eFoxp3 Forkhead Box P3\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eIL1RN\u003c/em\u003e Interleukin 1 Receptor Antagonist\u003c/p\u003e\n\u003cp\u003euNK Uterine Natural Killer cells\u003c/p\u003e\n\u003cp\u003eKEGG Kyoto Encyclopedia of Genes and Genomes\u003c/p\u003e\n\u003cp\u003eGSEA Gene Set Enrichment Analysis\u003c/p\u003e\n\u003cp\u003eSVM-RFE Support Vector Machine-Recursive Feature Elimination\u003c/p\u003e\n\u003cp\u003eSTRING Search Tool for the Retrieval of Interacting Genes/Proteins\u003c/p\u003e\n\u003cp\u003emiR MicroRNA\u003c/p\u003e\n\u003cp\u003eGPL Gene Expression Omnibus Platform\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAuthors’ contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWanzhen Li, Jinman Zhang and Baosheng Zhu designed the study. Wanzhen Li,Yunlong Li, Aiqi Cai, Youmou Fu analyzed and visualized the data. Wanzhen Li, Jiahong Tan, Yunlong Li, Baosheng Zhu contributed to writing and revising the paper. Wanzhen Li, Hongxia Xu Jiahong Tan All authors have read and agreed to the published version of the manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work is supported by the Yunnan Ten Thousand Talents Plan Yunling Scholar Project (NO. YNWR-YLXZ-2019-005).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets employed in our study can be acquired in the GEO repository (https://www.ncbi.nlm.nih.gov/geo/). The accession numbers are GSE34526, GSE98461, GSE137684, GSE22490, GSE26787, GSE139180.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDisclosure statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eGEO belongs to public databases. The patients involved in the database have obtained ethical approval.The endometrial samples required for RT-qPCR were approved by the Ethics Committee of the First People's Hospital of Yunnan Province, and were collected without causing any additional effects on the patients.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor details\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eKunming University of Science and Technology, 650500 Kunming, China\u003c/p\u003e\n\u003cp\u003eThe First People's Hospital of Yunnan, 650032 Kunming, China\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eLegro RS, Arslanian SA, Ehrmann DA, Hoeger KM, Murad MH, Pasquali R, et al. Diagnosis and treatment of polycystic ovary syndrome: an Endocrine Society clinical practice guideline. J Clin Endocrinol Metab. 2013;98(12):4565\u0026ndash;92.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003e[Chinese expert consensus on diagnosis and management of recurrent spontaneous abortion (2022)]. Zhonghua Fu Chan Ke Za Zhi. 2022;57(9):653\u0026thinsp;\u0026ndash;\u0026thinsp;67.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEl Hachem H, Crepaux V, May-Panloup P, Descamps P, Legendre G, Bouet PE. Recurrent pregnancy loss: current perspectives. Int J Womens Health. 2017;9:331\u0026ndash;45.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRegan L, Rai R, Saravelos S, Li T-C, Obstetricians tRCo, Gynaecologists. Recurrent MiscarriageGreen-top Guideline No. 17. 2023;130(12):e9-e39.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBender Atik R, Christiansen OB, Elson J, Kolte AM, Lewis S, Middeldorp S, et al. ESHRE guideline: recurrent pregnancy loss: an update in 2022. Hum Reprod Open. 2023;2023(1):hoad002.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu Y, Liu H, Li Z, Fan H, Yan X, Liu X, et al. The Release of Peripheral Immune Inflammatory Cytokines Promote an Inflammatory Cascade in PCOS Patients via Altering the Follicular Microenvironment. Front Immunol. 2021;12:685724.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHu M, Zhang Y, Li X, Cui P, Sferruzzi-Perri AN, Br\u0026auml;nnstr\u0026ouml;m M, et al. TLR4-Associated IRF-7 and NFκB Signaling Act as a Molecular Link Between Androgen and Metformin Activities and Cytokine Synthesis in the PCOS Endometrium. J Clin Endocrinol Metab. 2021;106(4):1022\u0026ndash;40.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTaylor DD, Bohler HC, Gercel-Taylor C. Pregnancy-linked suppression of TcR signaling pathways by a circulating factor absent in recurrent spontaneous pregnancy loss (RPL). Mol Immunol. 2006;43(11):1872\u0026ndash;80.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZou H, Yin J, Zhang Z, Xiang H, Wang J, Zhu D, et al. Destruction in maternal-fetal interface of URSA patients via the increase of the HMGB1-RAGE/TLR2/TLR4-NF-κB signaling pathway. Life Sci. 2020;250:117543.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQin L, Xu W, Li X, Meng W, Hu L, Luo Z, et al. Differential Expression Profile of Immunological Cytokines in Local Ovary in Patients with Polycystic Ovarian Syndrome: analysis by Flow Cytometry. Eur J Obstet Gynecol Reprod Biol. 2016;197:136\u0026ndash;41.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMatteo M, Serviddio G, Massenzio F, Scillitani G, Castellana L, Picca G, et al. Reduced percentage of natural killer cells associated with impaired cytokine network in the secretory endometrium of infertile women with polycystic ovary syndrome. Fertil Steril. 2010;94(6):2222-7, 7.e1-3.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAcar N, Ustunel I, Demir R. Uterine natural killer (uNK) cells and their missions during pregnancy: a review. Acta Histochem. 2011;113(2):82\u0026ndash;91.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQin S, Li L, Liu J, Zhang J, Xiao Q, Fan Y, et al. CD4(+)CD25(+)Foxp3(+) regulatory T cells regulate immune balance in unexplained recurrent spontaneous abortion via the Toll-like receptor 4/nuclear factor-κB pathway. J Int Med Res. 2020;48(12):300060520980940.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLangfelder P, Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics. 2008;9:559.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBreuer K, Foroushani AK, Laird MR, Chen C, Sribnaia A, Lo R, et al. InnateDB: systems biology of innate immunity and beyond\u0026ndash;recent updates and continuing curation. Nucleic Acids Res. 2013;41(Database issue):D1228-33.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBhattacharya S, Andorf S, Gomes L, Dunn P, Schaefer H, Pontius J, et al. ImmPort: disseminating data to the public for the future of immunology. Immunol Res. 2014;58(2\u0026ndash;3):234\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOrtutay C, Vihinen M. Immunome: a reference set of genes and proteins for systems biology of the human immune system. Cell Immunol. 2006;244(2):87\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKuleshov MV, Jones MR, Rouillard AD, Fernandez NF, Duan Q, Wang Z, et al. Enrichr: a comprehensive gene set enrichment analysis web server 2016 update. Nucleic Acids Res. 2016;44(W1):W90-7.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhou G, Soufan O, Ewald J, Hancock REW, Basu N, Xia J. NetworkAnalyst 3.0: a visual analytics platform for comprehensive gene expression profiling and meta-analysis. Nucleic Acids Res. 2019;47(W1):W234-w41.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu ZP, Wu C, Miao H, Wu H. RegNetwork: an integrated database of transcriptional and post-transcriptional regulatory networks in human and mouse. Database (Oxford). 2015;2015.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSzklarczyk D, Kirsch R, Koutrouli M, Nastou K, Mehryary F, Hachilif R, et al. The STRING database in 2023: protein-protein association networks and functional enrichment analyses for any sequenced genome of interest. Nucleic Acids Res. 2023;51(D1):D638-d46.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEngebretsen S, Bohlin J. Statistical predictions with glmnet. Clin Epigenetics. 2019;11(1):123.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSanz H, Valim C, Vegas E, Oller JM, Reverter F. SVM-RFE: selection and visualization of the most relevant features through non-linear kernels. BMC Bioinformatics. 2018;19(1):432.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSubramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005;102(43):15545\u0026ndash;50.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNewman AM, Liu CL, Green MR, Gentles AJ, Feng W, Xu Y, et al. Robust enumeration of cell subsets from tissue expression profiles. Nat Methods. 2015;12(5):453\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOgata H, Goto S, Sato K, Fujibuchi W, Bono H, Kanehisa M. KEGG: Kyoto Encyclopedia of Genes and Genomes. Nucleic Acids Res. 1999;27(1):29\u0026ndash;34.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGilmore TD. Introduction to NF-kappaB: players, pathways, perspectives. Oncogene. 2006;25(51):6680\u0026ndash;4.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGantier MP, Stunden HJ, McCoy CE, Behlke MA, Wang D, Kaparakis-Liaskos M, et al. A miR-19 regulon that controls NF-κB signaling. Nucleic Acids Res. 2012;40(16):8048\u0026ndash;58.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSong X, Zhu M, Sun Y, Liu B, Yan Z, Yin Y. MiR-204 enhances the progression of osteoarthritis by suppressing the production of IL-1β. Pharmazie. 2017;72(10):587\u0026ndash;92.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePatterson D, Jones C, Hart I, Bleskan J, Berger R, Geyer D, et al. The human interleukin-1 receptor antagonist (IL1RN) gene is located in the chromosome 2q14 region. Genomics. 1993;15(1):173\u0026ndash;6.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSim\u0026oacute;n C, Piquette GN, Frances A, Polan ML. Localization of interleukin-1 type I receptor and interleukin-1 beta in human endometrium throughout the menstrual cycle. J Clin Endocrinol Metab. 1993;77(2):549\u0026ndash;55.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYang Yan HW, Qiao Jie, Li Meizhi. Study on the correlation between interleukin-1 receptor antagonist and tumor necrosis factor alpha, high sensitivity C-reactive protein, and pathogenesis of polycystic ovary syndrome. Chinese Journal of Birth Health and Heredity. 2007;15(5):35\u0026ndash;7. (In Chinese).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMa Qinglian YJ, Wang Weixiang. Correlation studies of tumor necrosis factor, caseepin, interleukin 1 receptor antagonists and polycystic ovary syndrome. Maternal and Child Health Care of China. 2017;32(11):2315\u0026ndash;7.(In Chinese).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMontazeri M, Sanchez-Lopez JA, Caballero I, Maslehat Lay N, Elliott S, Fazeli A. Interleukin-1 receptor antagonist mediates toll-like receptor 3-induced inhibition of trophoblast adhesion to endometrial cells in vitro. Hum Reprod. 2016;31(9):2098\u0026ndash;107.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLinjawi S, Li TC, Laird S, Blakemore A. Interleukin-1 receptor antagonist and interleukin-1 beta polymorphisms in women with recurrent miscarriage. Fertil Steril. 2005;83(5):1549\u0026ndash;52.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHayden MS, Ghosh S. NF-κB in immunobiology. Cell Res. 2011;21(2):223\u0026ndash;44.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVallabhapurapu S, Karin M. Regulation and function of NF-kappaB transcription factors in the immune system. Annu Rev Immunol. 2009;27:693\u0026ndash;733.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang J, Yin T, Liu S. Dysregulation of immune response in PCOS organ system. Front Immunol. 2023;14:1169232.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYang X, Tian Y, Zheng L, Luu T, Kwak-Kim J. The Update Immune-Regulatory Role of Pro- and Anti-Inflammatory Cytokines in Recurrent Pregnancy Losses. Int J Mol Sci. 2022;24(1).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTibullo D, Li Volti G, Giallongo C, Grasso S, Tomassoni D, Anfuso CD, et al. Biochemical and clinical relevance of alpha lipoic acid: antioxidant and anti-inflammatory activity, molecular pathways and therapeutic potential. Inflamm Res. 2017;66(11):947\u0026ndash;59.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGhaiad HR, Ali SO, Al-Mokaddem AK, Abdelmonem M. Regulation of PKC/TLR-4/NF-kB signaling by sulbutiamine improves diabetic nephropathy in rats. Chem Biol Interact. 2023;381:110544.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSclabas GM, Uwagawa T, Schmidt C, Hess KR, Evans DB, Abbruzzese JL, et al. Nuclear factor kappa B activation is a potential target for preventing pancreatic carcinoma by aspirin. Cancer. 2005;103(12):2485\u0026ndash;90.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChakraborty P, Banerjee S, Saha P, Nandi SS, Sharma S, Goswami SK, et al. Aspirin and low-molecular weight heparin combination therapy effectively prevents recurrent miscarriage in hyperhomocysteinemic women. PLoS One. 2013;8(9):e74155.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMorsy MA, El-Hussieny M, Zenhom NM, Nair AB, Venugopala KN, Refaie MMM. Fenofibrate ameliorates letrozole-induced polycystic ovary in rats via modulation of PPARα and TNFα/CD95 pathway. Eur Rev Med Pharmacol Sci. 2022;26(20):7359\u0026ndash;70.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKorbecki J, Bobiński R, Dutka M. Self-regulation of the inflammatory response by peroxisome proliferator-activated receptors. Inflamm Res. 2019;68(6):443\u0026ndash;58.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Polycystic ovary syndrome, Recurrent Pregnancy Loss, IL1RN, NF-κB pathway, GEO, WGCNA, Machine learning","lastPublishedDoi":"10.21203/rs.3.rs-6337423/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6337423/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground \u003c/strong\u003ePolycystic ovary syndrome (PCOS) is an endocrine-related factor contributing to recurrent pregnancy loss (RPL). PCOS and RPL may share common risk genes and potential pathological mechanisms.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods \u003c/strong\u003eThree PCOS and three RPL datasets were obtained from the GEO database. Weighted gene co-expression network analysis (WGCNA), differential expression analysis, and three external immune gene datasets were used to identify shared immunological genes. Enrichr analysis, Gene-TF-miRNA, and Gene-pro networks suggested potential pathogenic mechanisms. Machine learning algorithms were then applied to identify the key risk gene. ROC curves and RT-qPCR tested the performance of the key gene in validation datasets for both PCOS and RPL. Gene Set Enrichment Analysis (GSEA) validated pathway changes, and immune infiltration analysis identified immune cells involved in both diseases.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusions \u003c/strong\u003eThis study highlighted the association of the NF-κB pathway by involvement of 19 shared immunological genes and one key risk gene, \u003cem\u003eIL1RN \u003c/em\u003ein RPL with PCOS\u003cem\u003e.\u003c/em\u003eIt might provide a novel understanding of the molecular pathology for RPL with PCOS.\u003c/p\u003e","manuscriptTitle":"Shared Risk Genes and Common Molecular Pathways Between PCOS and RPL by Integrated Transcriptomic Analysis and Machine Learning","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-05-07 11:08:05","doi":"10.21203/rs.3.rs-6337423/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"c86b5188-42be-4f96-986c-19959f2a37af","owner":[],"postedDate":"May 7th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-06-29T21:53:25+00:00","versionOfRecord":[],"versionCreatedAt":"2025-05-07 11:08:05","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6337423","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6337423","identity":"rs-6337423","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}