Bioinformatic analysis reveals endoplasmic reticulum stress-related molecular cluster and immune characterization in endometriosis：implications for disease subtyping and therapeutic strategies

Bioinformatic analysis reveals endoplasmic reticulum stress-related molecular cluster and immune characterization in endometriosis：implications for disease subtyping and therapeutic strategies | 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 Bioinformatic analysis reveals endoplasmic reticulum stress-related molecular cluster and immune characterization in endometriosis：implications for disease subtyping and therapeutic strategies Erqing Huang, Ling Zhang, Jie Lou, Xiaoli Wang, Lijuan Chen This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4212798/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 Numerous investigations have demonstrated the implication of endoplasmic reticulum stress (ERS) in the etiology of endometriosis. Employing bioinformatics methodologies, we conducted an analysis to ascertain the participation of genes associated with endoplasmic reticulum stress in endometriosis disease subtyping and immune infiltration, with the aim of constructing a diagnostic model for the disease. Materials and Methods Differential expression analysis, weighted gene co-expression network analysis (WGCNA), protein-protein interaction (PPI) network construction, and three machine learning algorithms were employed to identify hub genes associated with endoplasmic reticulum stress in endometriosis. Unsupervised cluster analysis was conducted to identify the ERS cluster. The ERS score and immune infiltration score were computed for distinct clusters using the CIBERSORT algorithm. Functional and pathway enrichment analysis was conducted based on the differential expression profiles of genes within the clusters to elucidate their potential biological functions. The differential expression profiles of genes within the clusters were submitted to the Connectivity Map database to identify candidate therapeutic compounds. A diagnostic model was developed utilizing hub genes, and its predictive performance for endometriosis was assessed. Endometrial tissue specimens obtained from patients were subjected to RT-qPCR and immunohistochemistry (IHC) analyses to evaluate the mRNA and protein expression levels of the hub genes. Results Von Willebrand factor (VWF), vascular cell adhesion molecule 1 (VCAM1), endothelial PAS domain protein 1 (EPAS1), and coagulation factor VIII (F8) were identified as the ERS-related hub genes in endometriosis. Unsupervised consensus clustering analysis revealed the presence of two stable clusters. Cluster B exhibited significantly higher immune scores compared to cluster A, thereby characterizing cluster B as an immune-enriched cluster and cluster A as a less immune-enriched cluster. Functional enrichment analysis revealed that the differentially expressed genes across the clusters predominantly participated in processes related to cell adhesion and regulation of immune cell activation. Decision curves, clinical impact curves, and calibration curves collectively underscored the robust diagnostic utility of the endometriosis diagnostic model derived from four hub genes. In cluster A, certain adrenergic receptor antagonists, progesterone or progesterone receptor agonists, androgen receptor modulators, and NF-κB pathway inhibitors exhibit promising therapeutic prospects. In contrast, cluster B presents potential therapeutic benefits with certain PKC activators, PPAR receptor agonists, insulin sensitizers, adenylate cyclase activators, and caspase activators. Moreover, the findings obtained from RT-qPCR and IHC assays corroborated the outcomes of the bioinformatic analysis, demonstrating elevated expression levels of both mRNA and protein of endoplasmic reticulum stress (ERS) hub genes in endometriosis tissues. Endometriosis Endoplasmic Reticulum Stress Bioinformatic analysis immune infiltration stromal cells Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Background Endometriosis (EMS) is delineated as the ectopic presence and infiltrative expansion of endometrial-like tissue beyond the confines of the uterine cavity[ 1 ]. Approximately 10% of women of reproductive age globally are estimated to be afflicted by this ailment[ 2 ]. The main manifestations of endometriosis encompass dysmenorrhea, chronic pelvic pain, infertility, painful intercourse, and if the diseased tissue accumulates in the bladder and rectum, it may also lead to corresponding dysfunctions in the urinary and digestive systems, such as constipation and hematuria[ 3 – 5 ]. Currently, laparoscopic surgery to remove the lesions is still the gold standard in the treatment of endometriosis therapy. Nonetheless, the disorder is predisposed to postoperative recurrence, thereby rendering long-term management a formidable challenge. Estrogen-mediated inflammation and immune dysregulation are hypothesized to play pivotal roles in the pathogenesis of endometriosis [ 6 – 9 ]. Recent years have seen significant strides in uncovering the mechanisms driving inflammatory immune responses in endometriosis. Multiple investigations have postulated a tight association between aberrant immune cell recruitment, heightened activation of inflammatory mediators, and biological processes including oxidative stress, autophagy, and apoptosis in endometriosis.[ 10 ]. Peritoneal macrophages and neutrophils are overactive in patients with endometriosis. The number of CD8 + T lymphocytes is higher in ectopic endometrial tissues than in ectopic endometrial tissues[ 11 ]. Studies have shown that TGF-β levels are higher in the peritoneal fluid of patients with endometriosis compared to healthy women[ 12 ].TGF-β is one of the inflammatory mediators released by mast cells, which promotes the expression of fibrotic factors, mediates epithelial-mesenchymal transition, and facilitates the transformation of mesothelial cells into fibroblasts[ 13 , 14 ].TGF-β is also involved in the differentiation of T-cells, stimulating the release of IL-17 and IL -10 release[ 15 , 16 ]. In addition, immune cells produce pro-angiogenic and pro-neurogenic factors in the peritoneal immune microenvironment of endometriosis patients[ 17 ]. For example, macrophages release nerve growth factor (NGF), which promotes the expression of pain-associated receptors, leading to clinical symptoms of pain in patients[ 18 ]. The intricate immune-inflammatory response orchestrated by immune cells constitutes a pivotal element contributing to the pathogenesis of endometriosis. The endoplasmic reticulum serves as a crucial organelle within organisms, facilitating the synthesis and modification of proteins. Proper protein folding processes are pivotal in dictating cellular outcomes [ 19 ]. The unfolded protein response (UPR) is initiated by transmembrane proteoceptors including IRE1α, PERK, and ATF6, whereby the chaperone GRP78/BiP typically binds to these receptors. When exogenous or endogenous endoplasmic reticulum stressors result in misfolded endoplasmic reticulum proteins, BiP dissociates from the receptor and binds to unfolded proteins, which activates the UPR[ 20 ]. The UPR initially responds to adaptive endoplasmic reticulum stress, however, under irremediable endoplasmic reticulum stress, the UPR can trigger pro-inflammatory and pro-death signals. Emerging evidence indicates that multiple endoplasmic reticulum stress sensors are activated within the tumor immune microenvironment, and the protective function of innate immune cells is disrupted by endoplasmic reticulum stress, and promote tumor cell proliferation and progression[ 21 – 23 ]. Notably, endometriosis lesions share certain characteristics with malignant tumors, including the capacity for growth, implantation, metastasis, and dissemination. Moreover, the immune microenvironment of ectopic endometrium exhibits tumor-like features [ 24 ]. This prompts the question: Is endoplasmic reticulum stress implicated in the pathogenesis of endometriosis? Multiple studies have demonstrated that induced endoplasmic reticulum stress can impede the proliferation and invasion of endometriosis lesions by modulating signaling pathways such as Akt/mTOR, MAPK/ERK, and NF-κB [ 25 – 29 ]. For instance, a study on the first-line therapeutic agent for endometriosis, dienogest, confirmed that this fourth-generation synthetic progestin upregulates endoplasmic reticulum stress in endometriotic stromal cells. This augmentation results in the upregulation of CHOP, thereby promoting apoptosis, restraining cell proliferation and invasion, and mitigating lesion progression in endometriosis [ 30 ]. Another study on ovarian endometriosis granulosa cells found that endoplasmic reticulum stress mediated apoptosis and subsequent ovarian dysfunction in patients' granulosa cells through caspase-3 and caspase-8[ 31 ]. These findings underscore the significant implication of endoplasmic reticulum stress in the pathogenesis of endometriosis. Nevertheless, the specific molecular subtypes of the disease linked to endoplasmic reticulum stress, as well as their immunological attributes, remain to be elucidated. Since limited research has elucidated the specific roles of ERS in endometriosis through bioinformatic approaches, we performed a bioinformatic analysis to establish an ERS- and immune-associated subtyping model, aiming to facilitate prospective clinical applications in the diagnosis and management of endometriosis. Therefore, exploring the relationship between endoplasmic reticulum stress-related genes (ERSRGs) and endometriosis may help clarify the endometriosis heterogeneity at a molecular level. Materials and Methods Patients and sample collection This research was approved by the Ethics Committee of Union Hospital, Tongji Medical College, Huazhong University of Science and Technology. Written informed consent was obtained from patients before the collection of human tissues, in accordance with the guidelines of the Declaration of Helsinki. Normal endometrial tissue samples were obtained from patients without endometriosis who underwent hysteroscopy and endometrial biopsy prior to pathological diagnosis, which confirmed as normal endometrium post-procedure (n = 10). Furthermore, ectopic and eutopic endometrial samples were sourced from individuals (n = 17) diagnosed with stage III or IV endometriosis[ 32 ]. All the collected endometrial tissues were diagnosed as proliferative endometrium after pathological histological diagnosis. All menstrual cycles were normal, non-pregnant or non-lactation, and no hormonal medication was taken 6 months before the operation, and no obvious medical and surgical diseases and complications were found. Data collection GSE7305 ( https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE7305 ) and GSE11691( https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE11691 ) datasets were downloaded from the Gene Expression Omnibus (GEO) database as training set by using the R package “GEOquery”. Details of datasets are listed in Table 1 . GSE11691 was in GPL96 platform, which consisted 9 endometriosis and 9 normal endometrial samples (Control samples). GSE7305 was in GPL570 platform, which consisted 10 endometriosis and 10 normal endometrial samples (Control samples). Then, we combined two datasets by using the “limma” package, the batch effect of the original gene expression landscapes in two datasets was eliminated using the comBat function based on the “sva” package. A total of 19 endometriosis samples and 19 normal endometrium samples with complete mRNA expression data and corresponding clinical materials were selected for subsequent analysis. Table 1 Information on the GEO dataset used in this study Sample Platforms Use in this article Normal Endometriosis All GSE7305 GPL570 Training set 10 10 20 GSE11691 GPL96 Training set 9 9 18 GSE51981 GPL570 Training set 71 77 148 GSE23339 GPL6102 Test set 9 10 19 GSE25628 GPL571 Test set 6 16 22 Identification of differentially expressed endoplasmic reticulum stress-related genes in endometriosis We extracted 1350 endoplasmic reticulum stress-related genes (ERSRGs) from previous articles. Genes with p 0.5 were considered differentially expressed genes (DEGs) by “limma” R package. The results were visualized using “ggplot2” and “ComplexHeatmap” packages. Analyzing and comparing immune infiltrating cells between endometriosis and normal endometrium CIBERSORT algorithm ( https:/cibersort.stanford.edu/ ) was applied to estimate the abundance of 21 immunocyte subtypes in endometriosis and normal samples, which was performed on R software. ESTIMATE algorithm was adopted to measure immunocyte infiltration degree (ImmuneScore) in these samples. Student’s t-test was applied to verify the differences between the two groups, and results were presented using the “ggboxplot”, “pheatmap” and “corrplot” packages. Single sample gene set enrichment analysis (ssGSEA) was performed to determine enrichment scores for each coupling of a sample and immune reaction gene sets in the ImmPort database. The package “GSVA” was used for the ssGSEA analysis. Construction of weighted gene coexpression networks (WGCNA) The package “WGCNA” was used for constructing the co-expression network of all genes between the 19 endometriosis and 19 normal samples. We obtained the difference between gene modules and endometriosis related modules hub genes. Hierarchical clustering was conducted on these samples to detect the outliers and remove the abnormal samples. The optimal power value was selected to transform the gene expression matrix into a weighted adjacency matrix, which was further transformed into a topological overlap matrix (TOM). Correlations between modules and traits were then calculated using the WGCNA package. Modules with high correlation coefficients were considered as candidates related to endometriosis and selected for subsequent analyses. With the candidate module selected, we defined |MM| (|Module membership|) > 0.8 and |GS| (|gene significance|) > 0.5 as the screening criteria for filtering key genes in the candidate module. The intersection of differentially expressed endometriosis genes related to ERS and hub genes in key modules were performed using the “jvenn” online website ( https://jvenn.toulouse.inrae.fr/app/example.html ). Identification of ERS-related key genes in endometriosis and chromosome location Based on above genes obtained intersecting genes, we constructed of protein-protein interaction (PPI) networks in STRING database ( http://string-db . org/). Subsequently, Cytoscape software was used to visualize the PPI network. And then, we used the Maximal Clique Centrality (MCC) algorithm in the cytoHubba (a Cytoscape plugin) to screen the key genes with high connectivity in PPI networks. Genomic locations of these key genes in chromosomes were demonstrated using “RCircos” package. Then, these key genes were processed with three machine learning algorithms. Least Absolute Shrinkage and Selection Operator (LASSO) to identify the genes by using the “glmnet” package. Random Forest Algorithm Analysis (RF) was conducted by the “randomForest” package. Meanwhile, a support vector machine-recursive feature elimination (SVM-RFE) model was established with a “SVM” package. Genes converged by three machine learning methods were confirmed as the hub ERS-related genes in endometriosis. The diagnostic and clustering discriminative ability of these selected hub genes was assessed through receiver operating characteristic (ROC) curves. ROC curve analysis We chose GSE23339 and GSE25628 dataset as testing sets, performed receiver operating characteristic (ROC) curve analysis on each screened hub ERS-related genes to verify its accuracy. The “pROC” package was used for ROC curve analysis. The hub genes with average AUC > 0.75 in both training and testing set were considered useful for EMS diagnosis. Unsupervised clustering of hub ERSRGs in endometriosis According to the expression level of hub ERSRGs, an unsupervised clustering analysis was carried out to classify 77 endometriosis samples of GSE51981 into distinct clusters by using the R package of “ConsensusClusterPlus”. The optimal number of classifications was determined by the cumulative distribution function (CDF) curves, consistency clustering score of each cluster and consensus clustering plot. We used Principal Component Analysis (PCA) to verify the ERS gene expression patterns. Besides, we calculate the “ERSscore” and “ImmuneScore” in distinct ERS cluster. Those with ERS scores greater than zero were categorized as “high ERSscore group” and those less than zero were categorized as “low ERSscore group”. The DEGs between the two clusters were identified by the “limma” package (adjusted P 1). Functional enrichment analysis and identification of potential drugs Based on the cluster DEGs, the GO and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses between the two subgroups were performed using the “clusterProfiler” package. Gene set enrichment analysis (GSEA) is used to evaluate the correlation of cluster DEGs in pre-defined gene set with each ERS cluster. The “c7.all.v2023.1.Hs. symbols” and “c2. kegg. immune. v2023.1.Hs. symbols” gene sets in the MSigDB database were subjected to GSEA using the “clusterProfiler” package. P-value < 0.05 was considered statistically significant. For each ERS cluster, we uploaded the cluster DEGs to the connective map (CMap) database and then applied enrichment analysis to pull out significantly enriched drugs. We used filter (with p < 0.05) to identify top 60 compounds per ERS cluster that are predicted to treat EMS. Clinical trait analysis and diagnostic model construction Besides, the R package “ggalluvial” constructed a Sankey diagram for demonstrating the distribution trend of ERS cluster, endometriosis severity and ERS score type. Each row represents a feature variable, different color represents different typing or clinical trait, lines represent the distribution of the same sample in different feature variables. Based on above hub ERSRGs obtained from intersection of machine learning methods, we built a nomograph using the “rms” and “rmda” package and evaluated its predictive power using the calibration curve to predict the probability of EMS. Immunohistochemistry (IHC) and image analysis The immunohistochemical staining of paraffin-embedded sections from all samples underwent morphological screening examination and was conducted following the established protocol previously described in the literature[ 33 ]. Briefly, the sections were stained with the primary antibody against F8 (1:500, affinity, USA), VCAM1(1:200, Proteintech, Wuhan, China), VWF (1:400, Proteintech, Wuhan, China), and EPAS1(1:50, Proteintech, Wuhan, China) and then scanned with a digital scanner. The intensity of staining was quantified using the ImageJ software. Quantitative real-time PCR (qRT-PCR) Total RNA was isolated from ovarian endometriosis, eutopic endometrium and normal control endometrium. We used Trizol (Yeasen, Shanghai, China) to isolate RNA from tissues according to the manufacturer’s instructions. Then, one microgram of RNA was reversely transcribed into cDNA using BeyoRT™ III cDNA First Strand cDNA Synthesis Kit (Beyotime, Shanghai, China). Amplification was performed using gene-specific primers (Qingke, Beijing, China) and SYBR Green qPCR Mix 2× (Beyotime, Shanghai, China) on a qRT-PCR device (ABI StepOnePlus, America). GAPDH was used as an internal control. The relative expression of the genes was calculated using the 2 –ΔΔCT method. The primers used were as follows Table 2 qRT-PCR primers Genes Forward primers Reverse primers EPAS1 GGCTGTGTCTGAGAAGAGTAACT TCCCGAAATCCAGAGAGATGATG F8 TGGAGTTGATGGGCTGTGATTTA CCAGGTGGCAAACATATTGGTAAA VCAM1 CGGAGACAGGAGACACAGTACTA GCACGAGAAGCTCAGGAGAAA VWF GGGAAGACTGTGATGATCGATGT GCAAACATCTCCCACAACATTCA GAPDH GGAGTCCACTGGCGTCTTCA GTCATGAGTCCTTCCACGATACC Statistical analysis All statistical analyses were performed using the R programming software (Version R4.3.1). Comparisons between two groups were performed using the t-test. Spearman test was utilized for correlation analysis. A difference of p < 0.05 indicated statistical significance unless specified otherwise. (∗∗∗ represents p < 0.001, ∗∗ represents p < 0.01, and ∗ represents p < 0.05) Results The landscape of different expressed ERSRGs and immune cell characteristics between EMS and normal samples After moving batch effect, differential gene expression analysis was performed using combined dataset of GSE7305 and GSE11691. We totally gained 1350 ERSRGs from published articles, among which 1180 ERSRGs are included in the combined dataset. Heatmaps of top 50 ERS-related DEGs are shown (Fig. 1a), the DEGs exhibit significantly different expression patterns between the EMS and normal samples. A total of 209 ERS-related DEGs (|log 2 FC|>0.5 and p < 0.01) were obtained, including 92 up-regulated and 117 down-regulated genes (Fig. 1b). The proportion of different infiltrating immune cell types between the EMS and normal groups was evaluated using the CIBERSORT algorithm (Fig. 1c and Fig. 1d). After removing populations with a sum of immune abundance value zero, Wilcox test was used to assess the enrichment scores representing immunocyte abundance and immune response activity between EMS and normal samples. As can be concluded from the heatmap, the level of immune cell infiltration in EMS was significantly higher than that in normal tissues, suggesting that EMS is a disease closely related to immune dysregulation, which is consistent with previous findings (Fig. 1e). In comparison to normal endometrium, activated.B.cell, CD56bright.natural.killer.cell, immature.B.cell, myeloid-derived suppressor cells(MDSCs), macrophage, mast.cell, natural.killer.T.cell, natural.killer.cell, regulatory.T.cell, T.follicular.helper.cell, type.1.T.helper.cell, effector.memory.CD4.T.cell, central.memory.CD4.T.cell, central.memory.CD8.T.cell were found to have higher level of infiltration in EMS (Fig. 1f). Besides, we conducted the correlation analysis between 21 immunocyte subtypes. We found that dendritic cells activated and plasma cells were the most positive relevant immune cells in EMS samples (r = 0.64), monocytes and eosinophils were the most positive relevant immune cells in normal samples (r = 0.62), which may indicate that these immunocytes work together (Fig. 1g and 1h). Identification of key modules associated with the ERS through WGCNA The soft -threshold power was set to 3 based on the scale-free network construction (Fig. 2a-2d). Five co-expression modules were identified through WGCNA analysis (Fig. 2e). The turquoise module showed highly positively correlation with EMS mRNA (r = 0.86, p = 5e-12) (Fig. 2f). Next, the turquoise module was selected as key modules relevant to EMS for further analysis. In Fig. 3F, the significant correlations between gene significance (GS) and module membership (MM) were presented in the turquoise module, 239 module hub genes were found in the two modules by GS > 0.5 and MM > 0.8 (Fig. 2g). Hub genes of module turquoise were intersected with 209 different expressed ERSRGs. There were 29 intersecting genes in the venn diagram (Fig. 2h). Identification of hub ERSRGs in endometriosis PPI network analysis was performed on 29 genes. The cytoHubba of Cytoscape was used to determine the hub genes in the PPI network (Fig. 3a). Top 10 genes in network ranked by MCC method were obtained, namely ESR1, KPNA2, CAV1, TXN, CDK1, VWF, EZH2, VCAM1, EPAS1, F8. Figure 3b displays the chromosomal locations of top 10 ERSRGs that are differentially expressed. LASSO regression was performed on top 10 ERSRGs for feature selection and dimensionality reduction to exclude unimportant regulators, which ultimately identified 4 hub ERSRGs (Fig. 3c-3d). For the random forest algorithm, six characteristic genes with relative importance > 2 were determined, including VCAM1, EPAS1, F8, VWF, ESR1, CAV1. (Fig. 3e-3f). For the SVM-RFE algorithm, when the feature number was six, the classifier had the minimum error, containing VCAM1, EPAS1, TXN, F8, VWF, ESR1 (Fig. 3g). Following intersection, four characteristic genes shared by LASSO, RF, and SVM-RFE algorithms were finally identified (VWF, VCAM1, EPAS1, F8). (Fig. 3h) We calculate the average AUC of training set (Fig. 4a) and test sets (Fig. 4b-4c), the average AUC values for all four genes exceeded 0.75 (EPAS1 = 0.786, F8 = 0.774, VCAM1 = 0.890, VWF = 0.866), demonstrating the good discriminatory efficacy of screened hub ERSRGs. Identification of ERS clusters Four hub ERSRGs in EMS, including five up-regulated and five down-regulated genes, were used to cluster the EMS datasets GSE51981, which includes 77 endometriosis samples. We performed unsupervised consensus cluster analysis based on the expression of four hub ERSRGs using the “ConsensusClusterPlus” R package. We observed stable isoform numbers when k = 2, and significant differences in the relative changes in the area under the CDF curve from k = 2 to k = 9 (Fig. 5a-d). The consistency scores of the subtypes were all over 0.9 when k = 2 (Fig. 5e). We found two distinct ERS modification subclusters, with 51 samples in cluster A and 26 samples in cluster B.PCA verified the remarkable difference between the clusters (Fig. 5f). To better understand the molecular characteristics between subtypes, we evaluated the differences in the expression of 4 ERSRGs (Fig. 6a). The results showed that EPAS1, F8, VCAM1 presented a higher expression in B than A cluster. We used ten violin plots revealed significant differences in gene expression patterns between the two clusters (Fig. 6b-k), indicating that the various ERS clusters may have various transcriptome or other characteristics. We also calculated the ERS scores based on the four hub ERSRGs obtained above and compared them using the PCA method. The ERS score of EMS were higher than control samples (Fig. 6l) and ERS score of cluster B were higher than those of cluster A (Fig. 6m). Characteristics of the immune microenvironment in distinct ERS cluster To identify differences in immune microenvironmental characteristics, we evaluated immune cells between these different ERS cluster. Significantly, Cluster B is clearly more immunologically active, while cluster A shows a relative defective immune cell infiltration. In Fig. 7a, we noticed that there was remarkable heterogeneity in in the abundance of immune cell infiltration between distinct cluster. Cluster B presented higher infiltration levels of activated.B.cell, activated.CD8.T.cell, activated.dendritic.cell, CD56bright.natural.killer.cell, CD56dim.natural.killer.cell, gamma.delta.T.cell, immature.B.cell, MDSC, macrophage, mast.cell, natural.killer.T.cell, natural.killer.cell, neutrophil, regulatory.T.cell, T.follicular.helper.cell, type.1.T.helper.cell, type.17.T.helper.cell, effector.memory.CD4.T.cell, memory.B.cell, central.memory.CD4.T.cell, effector.memory.CD8.T.cell (Fig. 7b). After that, we calculated the immune scores of the samples using the ESTIMATE algorithm, the immune scores of cluster B were significantly higher than that of cluster A (Fig. 7c). Collectively, we identified cluster B as an immune subtype and cluster A as a less-immune subtype. Furthermore, we explored the correlation between 10 ERSRGs and immune cells (Fig. 7d). Correlation analysis revealed that 10 ERSRGs were closely associated with most immune cells. For example, KPNA2 had the strongest positive correlation with immature.dendritic.cell abundance (r = 0.72) (Fig. 7e), and CAV1 had the strongest negative correlation with activated.dendritic.cell abundance (r=-0.71) (Fig. 7f). This indicates that KPNA2 and CAV1 play important roles in immunoinflammatory response in EMS. Explore the difference biological behaviors between ERS clusters We identified ERS cluster DEGs in order to explain the gene profiles related to cluster-mediated biological function regulation. A total of 451 cluster DEGs were screened out (Fig. 8a), including 127 DEGs with upregulation in cluster A and 324 DEGs with upregulation in cluster B. In the BP analysis of GO, cluster DEGs mainly participated in cell adhesion, regulation of T cell activation, ERK1 and ERK2 cascade, natural killer cell mediated immunity. In CC analysis, cluster DEGs mainly focused on collagen − containing extracellular matrix, external side of plasma membrane, vesicle lumen, external side of plasma membrane, cytoplasmic vesicle lumen and secretory granule lumen. MF analysis showed that cluster DEGs mainly related to extracellular matrix structural constituent, immune receptor activity, cytokine binding (Fig. 8b-d). In addition, in the KEGG analysis, the selected biological process of DEG enrichment significantly participated in biological pathways such as natural killer cell mediated cytotoxicity, PI3K − Akt signaling pathway, progesterone-mediated oocyte maturation, focal adhesion, chemokine signaling pathway, cytokine–cytokine receptor interaction pathway, VEGF signaling pathway, TNF signaling pathway and growth hormone regulation (Fig. 8e-f). We also conducted the GSEA analysis, the results showed that role of mammalian E proteins E2A and HEB in the development of T cells and N-ras in T cell development and function were significantly enriched in ERS cluster B (Fig. 8g), while functions TRAF6 regulated CD8 T cell memory development following infection by modulating fatty acid metabolism, extrathymic Treg development and STAT6 down-regulated in bone marrow-derived macrophages were significantly enriched in cluster A (Fig. 8h). Find potential drugs and clinicopathological features of ERS clusters We employed CMap database to identify potential therapeutic drugs for EMS. Differences between cluster A and cluster B were also apparent in compound prediction. Ten relevant compounds associated with distinct ERS cluster were identified. Next, we further evaluated the mechanism of actions (MOA) and drug target of these drugs to explore their potential mechanism for treating EMS. Notably, in cluster A, some adrenergic receptor antagonists, progesterone or progesterone receptor agonists, androgen receptor modulators, NF-κB pathway inhibitors, dipeptidyl peptidase inhibitors, and 5-hydroxytryptamine receptor agonists may have potential therapeutic roles (Fig. 9a), whereas in cluster B, some histone deacetylase inhibitors, protein kinase C (PKC) activators, PPAR receptor agonists and insulin sensitizers, adenylate cyclase activators, and caspase activators show a possible therapeutic role (Fig. 9b). Sankey diagram to visualize the relationships between the ERS phenotypes and EMS severity. The Sankey diagram verified the cluster A was associated with low ERS score, moderate or severe EMS. However, in cluster B with high ERS scores, mild EMS and moderately severe EMS had distributions with little difference. It is evident that the degree of clinical manifestation of EMS has greater heterogeneity among different ERS clusters and even within ERS cluster (Fig. 9c). We constructed a nomogram (Fig. 9d) to evaluate its predictive power using the calibration curve to predict the risk of EMS more clearly (Fig. 9e). Decision curve analysis (DCA) (Fig. 9f) and clinical impact curve (Fig. 9g) indicated that the “nomogram” curve was higher than the gray line. The calibration curve indicated a minimal difference between the real and predicted EMS risks, suggesting that the nomograph model of EMS is precise. Validation mRNA and protein expression in collected human endometrial tissue The results of bioinformatics analysis suggested that EPAS1, F8, VCAM1, and VWF might be highly expressed in endometriosis. After RT-qPCR, the mRNAs of the four ERS hub genes in ectopic endometrium were significantly higher than those in the eutopic endometrium and normal control endometrium (Fig. 10a). We performed immunohistochemical staining on normal control endometrial specimens, ectopic endometrial specimens and eutopic endometrial specimens in order to detect the protein expression. The results showed that the protein expression levels of EPAS1, F8, VCAM1, and VWF were consistent with those of mRNA (Fig. 10b-c). Discussion Currently, there is a lack of biomarkers with both accuracy and sensitivity for the diagnosis of endometriosis. Endometriosis is an estrogen-dependent inflammatory immune disorder. In addition to surgical removal of endometriosis lesions and loosening of pelvic adhesions, hormonal therapy remains the first line of pharmacologic treatment for endometriosis. Therefore, it is necessary to find the hub genes associated with the diagnosis of endometriosis and to explore the role of these hub genes in the pathogenesis of endometriosis, and drug therapy. In this study, we explored the role of endoplasmic reticulum stress-related genes in endometriosis immune infiltration, disease typing, potential therapeutic agents, biological functions, and pathways. Our study confirmed that the levels of infiltration of a wide range of immune cells were significantly elevated in ectopic lesion samples, which is consistent with previous research, suggesting the presence of a heavy immune-inflammatory response in endometriosis. Several studies in recent years have mechanistically confirmed the relationship between endoplasmic reticulum stress and inflammatory immune responses. For example, endoplasmic reticulum stress-induced inflammation and production of the pro-inflammatory cytokine IL-6 have been reported to be dependent on two members of the NOD-like receptor family, NOD1 and NOD2[ 34 ]. Another study on inflammatory bowel disease confirmed that endoplasmic reticulum stress induced by deletion of the transcription factor X-box binding protein-1 (XBP1) can lead to aberrant responses of intestinal epithelial cells to inflammatory signals[ 35 ]. In a study of endometriosis, treated endometrial stromal cells with peritoneal fluid from patients with endometriosis and normal control patients, the result suggested that endoplasmic reticulum stress-associated UPR pathways were activated in endometriosis[ 36 ]. Our study also showed that a relatively high degree of endoplasmic reticulum stress was present in ectopic lesion samples of endometriosis and that different endoplasmic reticulum stress subtypes had different degrees of UPR activation. However, initial activation of the UPR contributes to the endoplasmic reticulum's emergency response to adverse external signals, and over-activation leads to cell death, which is the entry point for many current studies in related fields to confirm the therapeutic effects of drugs on endometriosis. Our study of endoplasmic reticulum subtypes used the GSE51981 database, and Sankey plots showed that the degree of endoplasmic reticulum stress did not significantly correlate with the degree of clinical manifestations of endometriosis in this dataset. Is the specific level of endoplasmic reticulum stress in endometriosis associated with case typing and disease severity? This question remains to be further elucidated using large-scale transcriptomic data and clinical samples. Estrogen and progesterone strictly regulate the physiologic cycle of the endometrium. These steroids also have receptors in the endoplasmic reticulum that are involved in the regulation of protein folding, calcium homeostasis in the endoplasmic reticulum, and degradation of misfolded proteins. Abnormal endoplasmic reticulum stress response to progesterone enhances the invasiveness of endometrial stromal cells in endometriosis through the AKT/mTOR pathway, which has attracted our interest in the role of endoplasmic reticulum stress in endometriosis[ 2 ]. In this article, for the first time, we found the hub endoplasmic reticulum stress-related genes in endometriosis by bioinformatics methods and performed endoplasmic reticulum stress typing in endometriosis. First, we identified 10 characterized endoplasmic reticulum stress genes by differential analysis, WGCNA, PPI and Cytoscape. Second, machine learning was used to further screen these 10 genes for endoplasmic reticulum stress-centered genes. Subsequently, unsupervised clustering was performed using the screened VWF, VCAM1, EPAS1 and F8. We categorized endometriosis into two subtypes, A and B, which showed significant differences in the degree of immune cell infiltration and the type of immune cell infiltration. type B had a higher endoplasmic reticulum stress score and tended to be an immune cell-rich type, whereas type A had a lower endoplasmic reticulum stress score and a relatively low immune cell infiltration. In the comparative analysis of fractional immune infiltration, we found that several immune cell types with P < 0.001, such as activated B cells, CD56 strongly positively expressing NK cells, immature B cells, myeloid-derived suppressor cells (MDSC), NKT cells, NK cells, regulatory T cells, follicular helper T cells, and helper T cells 1 (Th1), which have also been shown to be associated with immunomodulation in endometriosis. For example, in a 1995 study, researchers found a strong correlation between NKT cell-mediated lysis of ectopic endometrial cells and downregulation of HLA1-like receptors[ 37 ]. Patients with severe endometriosis had significantly higher levels of CTLA-4 + T cells than those with mild endometriosis. In addition, CTLA-4 + T lymphocytes were negatively correlated with the percentage of NK and NKT-like cells in women with both endometriosis and infertility[ 38 ], which indicate that immunologic mechanisms of associated infertility may differ between endoplasmic reticulum stress subtypes and endometriosis clinicopathologic types and degrees of clinical manifestations of endometriosis. In the functional enrichment analysis of genes differing in endoplasmic reticulum stress subtypes, we noted significant functional enrichment of genes for extracellular matrix and cell adhesion. Endoplasmic reticulum stress enhances leukocyte recruitment through extracellular signals, remodels the immune microenvironment, and alters the behavior of immune cells through the secretion of polarizing cytokines, and in doing so restores tissue protein homeostasis. Muscle cells and airway epithelial cells have been reported to secrete a functional leukocyte-adherent hyaluronic acid matrix after various forms of endoplasmic reticulum stress[ 39 ]. Endometriosis lesions are highly resistant to apoptosis and cell adhesion, and attenuating endoplasmic reticulum stress-induced ectopic cell adhesion may be an important treatment for endometriosis. A related study showed that Frankincense could alleviate endometriosis by reducing the adhesion and proliferation of ectopic endometrial cells through endoplasmic reticulum stress/p53-apoptosis and chemokine-migration/adhesion pathways[ 40 ]. Notably, activated T cells showed significant enrichment in functional enrichment analysis and gene-immunocyte correlation analysis. We found that CAV1 was negatively correlated with multiple T cell subtypes. In a previous study, knockdown of Sirt1 in endothelial cells induced endoplasmic reticulum stress and miR-204 expression in endothelial cells, decreased CAV1, and impaired endothelium-dependent vasodilation, but there are no studies on CAV1 in relation to endometriosis[ 41 ]. In addition, we found a positive correlation between immature dendritic cells (iDC) and several endoplasmic reticulum-related genes. The proportion of iDCs was increased in the peritoneal cavity of patients with endometriosis compared to mature dendritic cells (mDCs), and maturation of dendritic cells in the peritoneal cavity plays an important role in the development of endometriosis[ 42 , 43 ]. However, in our immune cell differential analysis, there were no significant differences in iDCs between endometriosis and control samples, or in endoplasmic reticulum stress cluster A versus cluster B. Our study identified four endoplasmic reticulum stress center genes, VWF, VCAM1, EPAS1, and F8. VWF is a macromolecular plasma protein that plays a key role in maintaining normal coagulation and contributes to thrombotic disorders following endothelial and platelet dysfunction. The VWF gene is expressed predominantly in vascular endothelium and megakaryocytes, and its expression level is commonly used as a measure of angiogenesis capacity[ 44 , 45 ]. A Mendelian randomization study based on GWAS data from a large population showed a causal relationship between elevated VWF and an increased risk of endometriosis[ 46 ]. In lesions treated with GnRH agonists (GnRH-a), microvessel density was significantly reduced in patients with positive VWF expression[ 47 ]. However, protein expression of VWF did not show r-ASRM stage-dependent changes in endometriosis[ 48 ]. VCAM1 is a cell adhesion molecule, which is often used as a relevant measure of inflammation and malignancy cell adhesion capacity in studies related to a number of diseases, and has also been associated with transvascular endothelial migration of immune cells[ 49 ]. And estrogen-mediated upregulation of VCAM1 contributes to mast cell recruitment and differentiation[ 50 ]. As a classical inflammation-associated gene, knockdown of VCAM1 impedes TGF-β1-mediated endometrial cell proliferation, migration and invasion as well as attenuates inflammatory responses in endometriotic lesions[ 51 , 52 ]. Mechanistic studies of endometriosis associated with EPAS1 are lacking, but it is certain that EPAS1, as a key transcription factor in cellular response to hypoxia, is closely associated with inflammatory response and angiogenesis under hypoxic conditions, and more importantly, EPAS1 expressed in the endometrial stroma has been associated with invasion of trophoblast cells during embryo implantation, and mice knocked out of EPAS1 expression were infertile due to infertility due to failure of implantation[ 53 , 54 ]. Whether EPAS1 is involved in endometriosis-associated infertility may be a new research direction for the future. A study in 2018 confirmed that EPAS1 expression was significantly upregulated in CD73 + CD90 + CD105 + pluripotent stem cells isolated from ectopic endometrium compared to paired in situ endometrium, which may be associated with the progression of ovarian endometriosis to associated ovarian cancer[ 55 ]. The F8 gene is required for the production of coagulation factor VIII, which is essential for clot formation. In a mouse endometriosis model study, researchers found that the F8 antibody had a strong targeting effect on endometriotic tissues, based on which the study significantly reduced the size of endometriosis lesions using the F8 antibody-IL-10 fusion protein[ 56 ], and a similar study found that the F8 antibody-IL4 can similarly reduce the size of endometriosis lesions[ 57 ]. We look forward to the future discovery of upstream and downstream pathways and other molecules interacting with these four key genes, and to the thorough elucidation of the involvement of endoplasmic reticulum stress-related mechanisms in the pathogenesis and potential therapeutic drug molecular biology of endometriosis. There exists significant heterogeneity in clinical symptomatology and disease severity among individuals with endometriosis; individuals exhibiting high ASMR and EFI scores may manifest mild or asymptomatic pain, while those presenting with minimal lesions or mild pelvic adhesions observed laparoscopically may experience more severe pain. Within the scope of our investigation, the classification and grading of endoplasmic reticulum stress did not demonstrate a significant correlation with the severity of endometriosis. There is a notable absence of diagnostic markers demonstrating high sensitivity and specificity for endometriosis. The diagnostic model constructed in this investigation underwent validation through diverse algorithms, demonstrating favorable performance in discriminating endometriosis samples from control endometrial specimens. Furthermore, the endoplasmic reticulum stress-associated core genes identified herein hold promise for clinical application in the diagnosis of endometriosis. Presently, the primary modalities for addressing endometriosis encompass laparoscopic surgery and hormonal therapy. Moreover, our investigation has delineated compounds that target distinct subtypes of endoplasmic reticulum stress, potentially offering novel avenues for both scientific inquiry and clinical intervention in the management of endometriosis. It is reasonable to hypothesize that these compounds may promote the development and progression of endometriosis by inducing or inhibiting endoplasmic reticulum stress. Interestingly, our study found that subtypes with high endoplasmic reticulum stress scores also had higher immune cell infiltration scores, and this feature was not only reflected among endoplasmic reticulum stress subtypes, but also showed the same trend between endometriosis samples and control samples. Our study also presents certain limitations. Specifically, we did not investigate the potential association between endoplasmic reticulum stress subtypes and clinical phenotypes of endometriosis. It is pertinent to inquire whether disparities exist in endoplasmic reticulum stress-related gene expression profiles and the associated immune cell infiltration patterns between ovarian and non-ovarian endometriosis. Additionally, despite the utilization of multiple algorithms to validate the accuracy of our diagnostic models, the diagnostic utility of endoplasmic reticulum stress-related genes in endometriosis diagnosis warrants further substantiation through comprehensive investigations. Moreover, the absence of publicly available transcriptome sequencing data encompassing samples representing diverse pathologic types of endometriosis poses a challenge, and the outcomes of bioinformatics analyses may be subject to certain biases attributable to variations in sequencing platforms, statistical methodologies, and databases employed. Finally, this study did not use cell lines ore animal models to validate the screened genes, immune cells, and potential therapeutic compounds. In order to draw more convincing conclusions, further studies on the involvement of endoplasmic reticulum stress-related genes in the pathogenesis of endometriosis are necessary next. Conclusions The expression profiles of key endoplasmic reticulum stress-related genes, including EPAS1, F8, VCAM1, and VWF, exhibited significant disparities between endometriosis specimens and both adjacent normal endometrial tissues and control endometrial tissues. These pivotal genes were utilized for stratifying endometriosis cases into distinct immune and non-immune subgroups. Furthermore, distinct patterns were observed in the expression levels of endoplasmic reticulum stress-associated genes, endoplasmic reticulum stress scores, immune cell infiltration levels, biological functionalities, signaling pathways, and potential therapeutic targets across different endoplasmic reticulum stress subtypes of endometriosis. Notably, a diagnostic model based on the expression profiles of endoplasmic reticulum stress-related genes demonstrated robust discriminatory capacity for identifying endometriosis cases. List of abbreviations AUC Area Under Curve BP Biological Processes CC Cellular Component EMS Endometriosis ERS Endoplasmic reticulum stress GEO Gene Expression Omnibus GO Gene Ontology GS Gene Significance GSEA Gene set enrichment analysis KEGG Kyoto Encyclopedia of Genes and Genomes LASSO Least Absolute Shrinkage and Selection Operator MF Molecular Function MM Module Membership PCA Principal Component Analysis PPI Protein-Protein Interaction networks RF Random Forest RMSE Root Mean Squared Error ROC Receiver Operating Characteristic SVM-RFE Support Vector Machine-Recursive Feature Elimination WGCNA Weighted Correlation Network Analysis Declarations Ethics approval and consent to participate The collection of endometriosis samples from human subjects was approved by the Ethics Committee of Tongji Medical College, Huazhong University of Science and Technology and followed the tenets of the Declaration of Helsinki. The patients provided their written informed consent to participate in this study. Data availability statement The public datasets GSE7305, GSE11691, GSE23339, GSE25628 and GSE51981 used in this paper are available on the NCBI website (https://www.ncbi.nlm.nih.gov/geo/). Competing interests The authors declare that they have no competing interests. Funding This work was supported by grant from National Natural Science Foundation of China (No. 82371657). Acknowledgements Not applicable. Author contributions L Chen and X Wang proposed the idea and reviewed the manuscript, E Huang and L Zhang drafted and revised the initial manuscript. E Huang and J Lou performed the experiments and analyzed the data. All authors read and approved the final manuscript. References Vercellini P, Viganò P, Somigliana E, Fedele L. Endometriosis: pathogenesis and treatment. Nat Rev Endocrinol. 2014;10(5):261-75.http://doi.org/10.1038/nrendo.2013.255 Chapron C, Marcellin L, Borghese B, Santulli P. Rethinking mechanisms, diagnosis and management of endometriosis. Nat Rev Endocrinol. 2019;15(11):666-82.http://doi.org/10.1038/s41574-019-0245-z Olive DL, Schwartz LB. Endometriosis. N Engl J Med. 1993;328(24):1759-69.http://doi.org/10.1056/nejm199306173282407 Greene AD, Lang SA, Kendziorski JA, Sroga-Rios JM, Herzog TJ, Burns KA. Endometriosis: where are we and where are we going? Reproduction. 2016;152(3):R63-78.http://doi.org/10.1530/rep-16-0052 de Ziegler D, Borghese B, Chapron C. Endometriosis and infertility: pathophysiology and management. Lancet. 2010;376(9742):730-8.http://doi.org/10.1016/s0140-6736(10)60490-4 Shen HH, Zhang T, Yang HL, Lai ZZ, Zhou WJ, Mei J, et al. Ovarian hormones-autophagy-immunity axis in menstruation and endometriosis. Theranostics. 2021;11(7):3512-26.http://doi.org/10.7150/thno.55241 Liang Y, Yao S. Potential role of estrogen in maintaining the imbalanced sympathetic and sensory innervation in endometriosis. Mol Cell Endocrinol. 2016;424:42-9.http://doi.org/10.1016/j.mce.2016.01.012 Dutta S, Banu SK, Arosh JA. Endocrine disruptors and endometriosis. Reprod Toxicol. 2023;115:56-73.http://doi.org/10.1016/j.reprotox.2022.11.007 Bulun SE, Zeitoun K, Takayama K, Noble L, Michael D, Simpson E, et al. Estrogen production in endometriosis and use of aromatase inhibitors to treat endometriosis. Endocr Relat Cancer. 1999;6(2):293-301.http://doi.org/10.1677/erc.0.0060293 Berkkanoglu M, Arici A. Immunology and endometriosis. Am J Reprod Immunol. 2003;50(1):48-59.http://doi.org/10.1034/j.1600-0897.2003.00042.x Kisovar A, Becker CM, Granne I, Southcombe JH. The role of CD8+ T cells in endometriosis: a systematic review. Front Immunol. 2023;14:1225639.http://doi.org/10.3389/fimmu.2023.1225639 Young VJ, Ahmad SF, Duncan WC, Horne AW. The role of TGF-β in the pathophysiology of peritoneal endometriosis. Hum Reprod Update. 2017;23(5):548-59.http://doi.org/10.1093/humupd/dmx016 Mwaura AN, Riaz MA, Maoga JB, Mecha E, Omwandho COA, Scheiner-Bobis G, et al. Activin A Modulates Betaglycan Shedding via the ALK4-SMAD3-Dependent Pathway in Endometriotic Cells. Biomolecules. 2022;12(12).http://doi.org/10.3390/biom12121749 Arablou T, Aryaeian N, Khodaverdi S, Kolahdouz-Mohammadi R, Moradi Z, Rashidi N, et al. The effects of resveratrol on the expression of VEGF, TGF-β, and MMP-9 in endometrial stromal cells of women with endometriosis. Sci Rep. 2021;11(1):6054.http://doi.org/10.1038/s41598-021-85512-y Lurje I, Gaisa NT, Weiskirchen R, Tacke F. Mechanisms of organ fibrosis: Emerging concepts and implications for novel treatment strategies. Mol Aspects Med. 2023;92:101191.http://doi.org/10.1016/j.mam.2023.101191 Ghoreschi K, Laurence A, Yang XP, Tato CM, McGeachy MJ, Konkel JE, et al. Generation of pathogenic T(H)17 cells in the absence of TGF-β signalling. Nature. 2010;467(7318):967-71.http://doi.org/10.1038/nature09447 McKinnon BD, Evers J, Bersinger NA, Mueller MD. Induction of the neurokinin 1 receptor by TNFα in endometriotic tissue provides the potential for neurogenic control over endometriotic lesion growth. J Clin Endocrinol Metab. 2013;98(6):2469-77.http://doi.org/10.1210/jc.2013-1019 Velho RV, Taube E, Sehouli J, Mechsner S. Neurogenic Inflammation in the Context of Endometriosis-What Do We Know? Int J Mol Sci. 2021;22(23).http://doi.org/10.3390/ijms222313102 Rashid HO, Yadav RK, Kim HR, Chae HJ. ER stress: Autophagy induction, inhibition and selection. Autophagy. 2015;11(11):1956-77.http://doi.org/10.1080/15548627.2015.1091141 Bhardwaj M, Leli NM, Koumenis C, Amaravadi RK. Regulation of autophagy by canonical and non-canonical ER stress responses. Semin Cancer Biol. 2020;66:116-28.http://doi.org/10.1016/j.semcancer.2019.11.007 Sen Santara S, Lee DJ, Crespo Â, Hu JJ, Walker C, Ma X, et al. The NK cell receptor NKp46 recognizes ecto-calreticulin on ER-stressed cells. Nature. 2023;616(7956):348-56.http://doi.org/10.1038/s41586-023-05912-0 Di Conza G, Ho PC. ER Stress Responses: An Emerging Modulator for Innate Immunity. Cells. 2020;9(3).http://doi.org/10.3390/cells9030695 Cubillos-Ruiz JR, Mohamed E, Rodriguez PC. Unfolding anti-tumor immunity: ER stress responses sculpt tolerogenic myeloid cells in cancer. J Immunother Cancer. 2017;5:5.http://doi.org/10.1186/s40425-016-0203-4 Spagnolo E, Martinez A, Mascarós-Martínez A, Marí-Alexandre J, Carbonell M, González-Cantó E, et al. Evaluation of Immune Infiltrates in Ovarian Endometriosis and Endometriosis-Associated Ovarian Cancer: Relationship with Histological and Clinical Features. Int J Mol Sci. 2023;24(15).http://doi.org/10.3390/ijms241512083 Choi J, Jo M, Lee E, Lee DY, Choi D. Nuclear factor-kappa B signaling in endometriotic stromal cells is not inhibited by progesterone owing to an aberrant endoplasmic reticulum stress response: a possible role for an altered inflammatory process in endometriosis. Mol Hum Reprod. 2021;27(2).http://doi.org/10.1093/molehr/gaab002 Park S, Lim W, Bazer FW, Song G. Naringenin induces mitochondria-mediated apoptosis and endoplasmic reticulum stress by regulating MAPK and AKT signal transduction pathways in endometriosis cells. Mol Hum Reprod. 2017;23(12):842-54.http://doi.org/10.1093/molehr/gax057 Ham J, Park W, Song J, Kim HS, Song G, Lim W, et al. Fraxetin reduces endometriotic lesions through activation of ER stress, induction of mitochondria-mediated apoptosis, and generation of ROS. Phytomedicine. 2024;123:155187.http://doi.org/10.1016/j.phymed.2023.155187 Ham J, Kim J, Bazer FW, Lim W, Song G. Silibinin-induced endoplasmic reticulum stress and mitochondrial dysfunction suppress growth of endometriotic lesions. J Cell Physiol. 2019;234(4):4327-41.http://doi.org/10.1002/jcp.27212 Choi J, Jo M, Lee E, Lee DY, Choi D. Involvement of endoplasmic reticulum stress in regulation of endometrial stromal cell invasiveness: possible role in pathogenesis of endometriosis. Mol Hum Reprod. 2019;25(3):101-10.http://doi.org/10.1093/molehr/gaz002 Choi J, Jo M, Lee E, Lee DY, Choi D. Dienogest regulates apoptosis, proliferation, and invasiveness of endometriotic cyst stromal cells via endoplasmic reticulum stress induction. Mol Hum Reprod. 2020;26(1):30-9.http://doi.org/10.1093/molehr/gaz064 Kunitomi C, Harada M, Takahashi N, Azhary JMK, Kusamoto A, Nose E, et al. Activation of endoplasmic reticulum stress mediates oxidative stress-induced apoptosis of granulosa cells in ovaries affected by endometrioma. Mol Hum Reprod. 2020;26(1):40-52.http://doi.org/10.1093/molehr/gaz066 Revised American Society for Reproductive Medicine classification of endometriosis: 1996. Fertil Steril. 1997;67(5):817-21.http://doi.org/10.1016/s0015-0282(97)81391-x Li X, Jin J, Long X, Weng R, Xiong W, Liang J, et al. METTL3-regulated m6A modification impairs the decidualization of endometrial stromal cells by regulating YTHDF2-mediated degradation of FOXO1 mRNA in endometriosis-related infertility. Reprod Biol Endocrinol. 2023;21(1):99.http://doi.org/10.1186/s12958-023-01151-0 Sweet LA, Kuss-Duerkop SK, Keestra-Gounder AM. IRE1α-Driven Inflammation Promotes Clearance of Citrobacter rodentium Infection. Infect Immun. 2022;90(1):e0048121.http://doi.org/10.1128/iai.00481-21 Kaser A, Lee AH, Franke A, Glickman JN, Zeissig S, Tilg H, et al. XBP1 links ER stress to intestinal inflammation and confers genetic risk for human inflammatory bowel disease. Cell. 2008;134(5):743-56.http://doi.org/10.1016/j.cell.2008.07.021 Ekiz-Yilmaz T, Isildar B, Gezer A, Kankaya D, Cansiz-Ersoz C, Kayisli UA, et al. The role of unfolded protein response in the pathogenesis of endometriosis: contribution of peritoneal fluid. Reprod Biomed Online. 2021;42(1):1-15.http://doi.org/10.1016/j.rbmo.2020.09.012 Semino C, Semino A, Pietra G, Mingari MC, Barocci S, Venturini PL, et al. Role of major histocompatibility complex class I expression and natural killer-like T cells in the genetic control of endometriosis. Fertil Steril. 1995;64(5):909-16.http://doi.org/10.1016/s0015-0282(16)57901-1 Abramiuk M, Bębnowska D, Hrynkiewicz R, Polak PNG, Kotarski J, Roliński J, et al. CLTA-4 Expression is Associated with the Maintenance of Chronic Inflammation in Endometriosis and Infertility. Cells. 2021;10(3).http://doi.org/10.3390/cells10030487 Lauer ME, Erzurum SC, Mukhopadhyay D, Vasanji A, Drazba J, Wang A, et al. Differentiated murine airway epithelial cells synthesize a leukocyte-adhesive hyaluronan matrix in response to endoplasmic reticulum stress. J Biol Chem. 2008;283(38):26283-96.http://doi.org/10.1074/jbc.M803350200 Cho MK, Jin JS, Jo Y, Han JH, Shin S, Bae SJ, et al. Frankincense ameliorates endometriosis via inducing apoptosis and reducing adhesion. Integr Med Res. 2023;12(2):100947.http://doi.org/10.1016/j.imr.2023.100947 Kassan M, Vikram A, Kim YR, Li Q, Kassan A, Patel HH, et al. Sirtuin1 protects endothelial Caveolin-1 expression and preserves endothelial function via suppressing miR-204 and endoplasmic reticulum stress. Sci Rep. 2017;7:42265.http://doi.org/10.1038/srep42265 Qiaomei Z, Ping W, Yanjing Z, Jinhua W, Shaozhan C, Lihong C. Features of peritoneal dendritic cells in the development of endometriosis. Reprod Biol Endocrinol. 2023;21(1):4.http://doi.org/10.1186/s12958-023-01058-w Schulke L, Berbic M, Manconi F, Tokushige N, Markham R, Fraser IS. Dendritic cell populations in the eutopic and ectopic endometrium of women with endometriosis. Hum Reprod. 2009;24(7):1695-703.http://doi.org/10.1093/humrep/dep071 Randi AM, Smith KE, Castaman G. von Willebrand factor regulation of blood vessel formation. Blood. 2018;132(2):132-40.http://doi.org/10.1182/blood-2018-01-769018 Randi AM, Laffan MA. Von Willebrand factor and angiogenesis: basic and applied issues. J Thromb Haemost. 2017;15(1):13-20.http://doi.org/10.1111/jth.13551 Li Y, Liu H, Ye S, Zhang B, Li X, Yuan J, et al. The effects of coagulation factors on the risk of endometriosis: a Mendelian randomization study. BMC Med. 2023;21(1):195.http://doi.org/10.1186/s12916-023-02881-z Khan KN, Kitajima M, Hiraki K, Fujishita A, Sekine I, Ishimaru T, et al. Changes in tissue inflammation, angiogenesis and apoptosis in endometriosis, adenomyosis and uterine myoma after GnRH agonist therapy. Hum Reprod. 2010;25(3):642-53.http://doi.org/10.1093/humrep/dep437 Khan KN, Masuzaki H, Fujishita A, Kitajima M, Sekine I, Ishimaru T. Immunoexpression of hepatocyte growth factor and c-Met receptor in the eutopic endometrium predicts the activity of ectopic endometrium. Fertil Steril. 2003;79(1):173-81.http://doi.org/10.1016/s0015-0282(02)04535-1 Kong DH, Kim YK, Kim MR, Jang JH, Lee S. Emerging Roles of Vascular Cell Adhesion Molecule-1 (VCAM-1) in Immunological Disorders and Cancer. Int J Mol Sci. 2018;19(4).http://doi.org/10.3390/ijms19041057 McCallion A, Nasirzadeh Y, Lingegowda H, Miller JE, Khalaj K, Ahn S, et al. Estrogen mediates inflammatory role of mast cells in endometriosis pathophysiology. Front Immunol. 2022;13:961599.http://doi.org/10.3389/fimmu.2022.961599 Zhang J, Li H, Yi D, Lai C, Wang H, Zou W, et al. Knockdown of vascular cell adhesion molecule 1 impedes transforming growth factor beta 1-mediated proliferation, migration, and invasion of endometriotic cyst stromal cells. Reprod Biol Endocrinol. 2019;17(1):69.http://doi.org/10.1186/s12958-019-0512-9 Lyu D, Tang N, Wang J, Zhang Y, Chang J, Liu Z, et al. TGR5 agonist INT-777 mitigates inflammatory response in human endometriotic stromal cells: A therapeutic implication for endometriosis. Int Immunopharmacol. 2019;71:93-9.http://doi.org/10.1016/j.intimp.2019.02.044 Bhurke A, Kannan A, Neff A, Ma Q, Laws MJ, Taylor RN, et al. A hypoxia-induced Rab pathway regulates embryo implantation by controlled trafficking of secretory granules. Proc Natl Acad Sci U S A. 2020;117(25):14532-42.http://doi.org/10.1073/pnas.2000810117 Matsumoto L, Hirota Y, Saito-Fujita T, Takeda N, Tanaka T, Hiraoka T, et al. HIF2α in the uterine stroma permits embryo invasion and luminal epithelium detachment. J Clin Invest. 2018;128(7):3186-97.http://doi.org/10.1172/jci98931 Ponandai-Srinivasan S, Andersson KL, Nister M, Saare M, Hassan HA, Varghese SJ, et al. Aberrant expression of genes associated with stemness and cancer in endometria and endometrioma in a subset of women with endometriosis. Hum Reprod. 2018;33(10):1924-38.http://doi.org/10.1093/humrep/dey241 Schwager K, Bootz F, Imesch P, Kaspar M, Trachsel E, Neri D. The antibody-mediated targeted delivery of interleukin-10 inhibits endometriosis in a syngeneic mouse model. Hum Reprod. 2011;26(9):2344-52.http://doi.org/10.1093/humrep/der195 Quattrone F, Sanchez AM, Pannese M, Hemmerle T, Viganò P, Candiani M, et al. The Targeted Delivery of Interleukin 4 Inhibits Development of Endometriotic Lesions in a Mouse Model. Reprod Sci. 2015;22(9):1143-52.http://doi.org/10.1177/1933719115578930 Additional Declarations No competing interests reported. Supplementary Files SupplementMaterial.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-4212798","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":289185775,"identity":"a60501a0-9480-4b45-b27f-fd6743784477","order_by":0,"name":"Erqing Huang","email":"","orcid":"","institution":"Huazhong University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Erqing","middleName":"","lastName":"Huang","suffix":""},{"id":289185776,"identity":"f6e9d440-3731-4b72-bc2f-f00f45e45d82","order_by":1,"name":"Ling Zhang","email":"","orcid":"","institution":"Huazhong University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Ling","middleName":"","lastName":"Zhang","suffix":""},{"id":289185777,"identity":"4d0feaf7-8ebb-491c-8f05-0980eaf74fc3","order_by":2,"name":"Jie Lou","email":"","orcid":"","institution":"Huazhong University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Jie","middleName":"","lastName":"Lou","suffix":""},{"id":289185778,"identity":"9065a9f4-2ac7-4da7-99c4-45570e956aa3","order_by":3,"name":"Xiaoli Wang","email":"","orcid":"","institution":"Huazhong University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Xiaoli","middleName":"","lastName":"Wang","suffix":""},{"id":289185779,"identity":"3145032c-8cfd-4c88-b481-e8a7bc9f35bf","order_by":4,"name":"Lijuan Chen","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAzklEQVRIiWNgGAWjYDACCSB+wGOTAOGxEaslgScNqIWZJC0Mh0nQwj+7+diDBJnzeQbnzx9g+FB2GCjSQMCSO8fSDRJ4bhcb3EhmYJxx7jBQ5AB+LQYSOWYSQC2JG24wMzDzth0GiiQQ0pL/DajlXOKG84cZmP8SpyWHDajlQOKGA8kMzIzEaJG4kQZyWHLizBvJBgd7zqXzSNwgoIV/RvIziY89dol95w8+fPCjzFqOfwYBLWDA2AOhDwAxDxHqQeAHkepGwSgYBaNgZAIAVIFDJyVplyYAAAAASUVORK5CYII=","orcid":"","institution":"Huazhong University of Science and Technology","correspondingAuthor":true,"prefix":"","firstName":"Lijuan","middleName":"","lastName":"Chen","suffix":""}],"badges":[],"createdAt":"2024-04-03 12:38:46","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4212798/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4212798/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":54597198,"identity":"1a4f30a9-4d7b-44f7-bd41-a529639e1254","added_by":"auto","created_at":"2024-04-12 19:13:36","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":7827604,"visible":true,"origin":"","legend":"\u003cp\u003eThe expression pattern and immune characteristics of the ERSRGs in EMS. (a) Heatmap of the top fifty differentially expressed ERSRGs. (b) The volcano plot of the 209 different expressed ERSRGs in EMS and normal samples. (c-d) Bar plot of the relative proportion of 21 infiltrated immune cells in EMS(c) and normal sample(d). (e) Distribution of 28 immune cells and their subtypes in EMS and normal samples. (f) Violin plot showing differential infiltration of the 28 immune cell populations. (g-h) Correlation heatmap of 21 immune cell types in EMS(g) and normal(h) samples. *p \u0026lt; 0.05, **p \u0026lt; 0.01, ∗∗∗ p\u0026lt; 0.001.\u003c/p\u003e","description":"","filename":"figure1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-4212798/v1/b5aaf22025c9c4d95483b621.jpg"},{"id":54597851,"identity":"19a27b7c-6f89-4477-ab12-7e87139985e1","added_by":"auto","created_at":"2024-04-12 19:29:36","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":5910949,"visible":true,"origin":"","legend":"\u003cp\u003eWGCNA finds characteristic EMS genes. (a) Analysis of the scale-free ft index and analysis of the mean connectivity for various soft-thresholding powers. (b) Topology in scale free network. (c-d) gene dendrogram and module colors map. (e) Bar plot of the importance of gene modules. (f) Correlation between the different modules in EMS and normal sample. (g) Gene significance (GS) scatter plot of EMS and module member (MM) in the turquoise module. (h) Venn diagram showing the intersection of WGCNA turquoise module key genes and ERS related DEGs in EMS.\u003c/p\u003e","description":"","filename":"figure2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-4212798/v1/2d0dabae9c7898c21e8b1f0d.jpg"},{"id":54597852,"identity":"86b1e8f5-7846-497b-b4ea-e330e8cc3d75","added_by":"auto","created_at":"2024-04-12 19:29:37","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":4293339,"visible":true,"origin":"","legend":"\u003cp\u003ePPI Machine learning arthrograms find hub ERSRGs in EMS. (a) PPI network of 29 genes. Top 10 differentially expressed ERSRGs were in red, yellow or orange color. (b) The location of top 10 differentially expressed ERSRGs on chromosomes. (c) The optimal lambda value was selected in the LASSO regression. (d) The LASSO coefficient profiles of the top 10 genes. (e) Relationship between the number of random forest trees and error rates. (f) Ranking of the relative importance of genes. (g) SVM-RFE algorithm for feature selection. (h) Venn diagram showing the feature genes shared by LASSO, random forest, and SVM-RFE algorithms.\u003c/p\u003e","description":"","filename":"figure3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-4212798/v1/1a95ca58d4c17d66d310b8b8.jpg"},{"id":54597557,"identity":"a5b57e8c-9ff4-42f7-8403-b394085a4931","added_by":"auto","created_at":"2024-04-12 19:21:36","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":352229,"visible":true,"origin":"","legend":"\u003cp\u003eROC curves were used to analyze the ability of four hub genes to distinguish between EMS and normal samples, and the AUC value was used to evaluate the distinction ability. (a) ROC curves in training set. (b-c) ROC curves in test sets GSE23339(b) and GSE25628(c).\u003c/p\u003e","description":"","filename":"figure4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-4212798/v1/b8d454ef3e694bda2fb0ea18.jpg"},{"id":54597203,"identity":"ff56913d-64a0-45dd-a642-88499660bceb","added_by":"auto","created_at":"2024-04-12 19:13:37","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":5318249,"visible":true,"origin":"","legend":"\u003cp\u003eIdentification two different ERS clusters in EMS. (a) Consensus clustering matrix when k=2. (b) Cumulative distribution function (CDF) curves of clustering. (c) CDF delta area curves. (d) The tracking plot shows the cluster assignment of items (columns) by color for each k (rows). (e) Consensus clustering score of each cluster. (f) PCA visualizes the distribution of two clusters.\u003c/p\u003e","description":"","filename":"figure5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-4212798/v1/bf2e126564cfd12fbc80572a.jpg"},{"id":54597560,"identity":"3e8f23bf-495a-4676-ab03-772477ac24fe","added_by":"auto","created_at":"2024-04-12 19:21:38","extension":"jpg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":8200107,"visible":true,"origin":"","legend":"\u003cp\u003eHub gene expression and immune microenvironmental features between two ERS clusters. (a) Visualization of the expression of the four hub ERSRGs in cluster A and B. (b-k) Ten violin plots revealed significant differences in gene expression patterns between the two clusters. (l-m) Comparison of ERS score between EMS and normal samples(l), ERS cluster A and B(m).\u003c/p\u003e","description":"","filename":"figure6.jpg","url":"https://assets-eu.researchsquare.com/files/rs-4212798/v1/b8272b2d4de923e66626752c.jpg"},{"id":54597204,"identity":"ea31e555-930e-4506-97e3-11feb03f5a19","added_by":"auto","created_at":"2024-04-12 19:13:38","extension":"jpg","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":11646918,"visible":true,"origin":"","legend":"\u003cp\u003eThe abundance difference of infiltrating immunocytes between each cluster and gene-immune cell correlation analysis. (a-b) The heatmap(a) and boxplot(b) shows the immune cell abundance in two ERS cluster. (c)Comparison the immune score between the two cluster. (d)Gene-immune cell correlation heatmap. (e-f)The dot plot shows the most positively(e) and negatively(f) associated immune cells and genes.\u003c/p\u003e","description":"","filename":"figure7.jpg","url":"https://assets-eu.researchsquare.com/files/rs-4212798/v1/1c30e0ea2d5bfed73cef8cb8.jpg"},{"id":54597207,"identity":"1ce15f31-4905-4a3f-aa25-479b7f7693f1","added_by":"auto","created_at":"2024-04-12 19:13:39","extension":"jpg","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":14995343,"visible":true,"origin":"","legend":"\u003cp\u003eFunctional analysis of genes involved in the ERS clusters in EMS. (a) 451 genes were related to the ERS clusters. (b-d) GO analysis showed the biological features of cluster DEGs. (e-f) KEGG analysis showed pathway enrichment of cluster DEGs. (g-h) GSEA enrichment analysis showing significantly activated immune-related functions in two clusters.\u003c/p\u003e","description":"","filename":"figure8.jpg","url":"https://assets-eu.researchsquare.com/files/rs-4212798/v1/fc05f0daec1ef77993ce8c5f.jpg"},{"id":54597202,"identity":"8118d465-a372-42ec-bd96-812faf480717","added_by":"auto","created_at":"2024-04-12 19:13:37","extension":"jpg","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":4767441,"visible":true,"origin":"","legend":"\u003cp\u003eTherapeutic drug prediction, clinical characterization, and diagnostic modeling. (a-b) Heatmap displaying the mechanisms of action shared by prospective therapeutic drugs for cluster A (a) and cluster B (b), respectively. (c) Sankey diagram for two ERS clusters and clinical traits, ERS high and low score groups. (d) Representative nomogram for predicting the risk of EMS based on the four hub ERSRGs. (e) Calibration curves for evaluating the predictive ability of the nomogram model. (f) DCA curves for assessing the clinical value of the nomogram model. (g) Clinical impact curve of the nomogram model.\u003c/p\u003e","description":"","filename":"figure9.jpg","url":"https://assets-eu.researchsquare.com/files/rs-4212798/v1/f80324dd6b699f799c00b37b.jpg"},{"id":54597197,"identity":"9b8d3b66-c2ee-4396-9a80-740069158aec","added_by":"auto","created_at":"2024-04-12 19:13:36","extension":"jpg","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":18473934,"visible":true,"origin":"","legend":"\u003cp\u003eqRT-PCR and IHC validation the hub gene expression level. (a)Hub ERS genes mRNA expression of normal, eutopic and ectopic endometrium. (b) Immunostaining of EPAS1, F8, VCAM1, VWF protein expression. (c)Semi-quantitative analysis of IHC results.\u003c/p\u003e","description":"","filename":"figure10.jpg","url":"https://assets-eu.researchsquare.com/files/rs-4212798/v1/9459c19739f7aa7ffdf36cc6.jpg"},{"id":54597923,"identity":"ce2da779-b313-46ee-964e-6a3d54005306","added_by":"auto","created_at":"2024-04-12 19:29:50","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1991255,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4212798/v1/54576863-bec1-4da9-89b4-03d21de5e32c.pdf"},{"id":54597206,"identity":"9ee0222f-95ec-4e4c-b19c-019bf4ef7fed","added_by":"auto","created_at":"2024-04-12 19:13:38","extension":"docx","order_by":13,"title":"","display":"","copyAsset":false,"role":"supplement","size":131318,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementMaterial.docx","url":"https://assets-eu.researchsquare.com/files/rs-4212798/v1/3f8c174ebe7473748bf3d549.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Bioinformatic analysis reveals endoplasmic reticulum stress-related molecular cluster and immune characterization in endometriosis：implications for disease subtyping and therapeutic strategies","fulltext":[{"header":"Background","content":"\u003cp\u003eEndometriosis (EMS) is delineated as the ectopic presence and infiltrative expansion of endometrial-like tissue beyond the confines of the uterine cavity[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. Approximately 10% of women of reproductive age globally are estimated to be afflicted by this ailment[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. The main manifestations of endometriosis encompass dysmenorrhea, chronic pelvic pain, infertility, painful intercourse, and if the diseased tissue accumulates in the bladder and rectum, it may also lead to corresponding dysfunctions in the urinary and digestive systems, such as constipation and hematuria[\u003cspan additionalcitationids=\"CR4\" citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. Currently, laparoscopic surgery to remove the lesions is still the gold standard in the treatment of endometriosis therapy. Nonetheless, the disorder is predisposed to postoperative recurrence, thereby rendering long-term management a formidable challenge.\u003c/p\u003e \u003cp\u003eEstrogen-mediated inflammation and immune dysregulation are hypothesized to play pivotal roles in the pathogenesis of endometriosis [\u003cspan additionalcitationids=\"CR7 CR8\" citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. Recent years have seen significant strides in uncovering the mechanisms driving inflammatory immune responses in endometriosis. Multiple investigations have postulated a tight association between aberrant immune cell recruitment, heightened activation of inflammatory mediators, and biological processes including oxidative stress, autophagy, and apoptosis in endometriosis.[\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. Peritoneal macrophages and neutrophils are overactive in patients with endometriosis. The number of CD8\u003csup\u003e+\u003c/sup\u003e T lymphocytes is higher in ectopic endometrial tissues than in ectopic endometrial tissues[\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. Studies have shown that TGF-β levels are higher in the peritoneal fluid of patients with endometriosis compared to healthy women[\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e].TGF-β is one of the inflammatory mediators released by mast cells, which promotes the expression of fibrotic factors, mediates epithelial-mesenchymal transition, and facilitates the transformation of mesothelial cells into fibroblasts[\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e].TGF-β is also involved in the differentiation of T-cells, stimulating the release of IL-17 and IL -10 release[\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. In addition, immune cells produce pro-angiogenic and pro-neurogenic factors in the peritoneal immune microenvironment of endometriosis patients[\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. For example, macrophages release nerve growth factor (NGF), which promotes the expression of pain-associated receptors, leading to clinical symptoms of pain in patients[\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. The intricate immune-inflammatory response orchestrated by immune cells constitutes a pivotal element contributing to the pathogenesis of endometriosis.\u003c/p\u003e \u003cp\u003eThe endoplasmic reticulum serves as a crucial organelle within organisms, facilitating the synthesis and modification of proteins. Proper protein folding processes are pivotal in dictating cellular outcomes [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. The unfolded protein response (UPR) is initiated by transmembrane proteoceptors including IRE1α, PERK, and ATF6, whereby the chaperone GRP78/BiP typically binds to these receptors. When exogenous or endogenous endoplasmic reticulum stressors result in misfolded endoplasmic reticulum proteins, BiP dissociates from the receptor and binds to unfolded proteins, which activates the UPR[\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. The UPR initially responds to adaptive endoplasmic reticulum stress, however, under irremediable endoplasmic reticulum stress, the UPR can trigger pro-inflammatory and pro-death signals. Emerging evidence indicates that multiple endoplasmic reticulum stress sensors are activated within the tumor immune microenvironment, and the protective function of innate immune cells is disrupted by endoplasmic reticulum stress, and promote tumor cell proliferation and progression[\u003cspan additionalcitationids=\"CR22\" citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]. Notably, endometriosis lesions share certain characteristics with malignant tumors, including the capacity for growth, implantation, metastasis, and dissemination. Moreover, the immune microenvironment of ectopic endometrium exhibits tumor-like features [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. This prompts the question: Is endoplasmic reticulum stress implicated in the pathogenesis of endometriosis?\u003c/p\u003e \u003cp\u003eMultiple studies have demonstrated that induced endoplasmic reticulum stress can impede the proliferation and invasion of endometriosis lesions by modulating signaling pathways such as Akt/mTOR, MAPK/ERK, and NF-κB [\u003cspan additionalcitationids=\"CR26 CR27 CR28\" citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e]. For instance, a study on the first-line therapeutic agent for endometriosis, dienogest, confirmed that this fourth-generation synthetic progestin upregulates endoplasmic reticulum stress in endometriotic stromal cells. This augmentation results in the upregulation of CHOP, thereby promoting apoptosis, restraining cell proliferation and invasion, and mitigating lesion progression in endometriosis [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. Another study on ovarian endometriosis granulosa cells found that endoplasmic reticulum stress mediated apoptosis and subsequent ovarian dysfunction in patients' granulosa cells through caspase-3 and caspase-8[\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]. These findings underscore the significant implication of endoplasmic reticulum stress in the pathogenesis of endometriosis. Nevertheless, the specific molecular subtypes of the disease linked to endoplasmic reticulum stress, as well as their immunological attributes, remain to be elucidated.\u003c/p\u003e \u003cp\u003eSince limited research has elucidated the specific roles of ERS in endometriosis through bioinformatic approaches, we performed a bioinformatic analysis to establish an ERS- and immune-associated subtyping model, aiming to facilitate prospective clinical applications in the diagnosis and management of endometriosis. Therefore, exploring the relationship between endoplasmic reticulum stress-related genes (ERSRGs) and endometriosis may help clarify the endometriosis heterogeneity at a molecular level.\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003ePatients and sample collection\u003c/h2\u003e \u003cp\u003eThis research was approved by the Ethics Committee of Union Hospital, Tongji Medical College, Huazhong University of Science and Technology. Written informed consent was obtained from patients before the collection of human tissues, in accordance with the guidelines of the Declaration of Helsinki.\u003c/p\u003e \u003cp\u003eNormal endometrial tissue samples were obtained from patients without endometriosis who underwent hysteroscopy and endometrial biopsy prior to pathological diagnosis, which confirmed as normal endometrium post-procedure (n\u0026thinsp;=\u0026thinsp;10). Furthermore, ectopic and eutopic endometrial samples were sourced from individuals (n\u0026thinsp;=\u0026thinsp;17) diagnosed with stage III or IV endometriosis[\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e]. All the collected endometrial tissues were diagnosed as proliferative endometrium after pathological histological diagnosis. All menstrual cycles were normal, non-pregnant or non-lactation, and no hormonal medication was taken 6 months before the operation, and no obvious medical and surgical diseases and complications were found.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003eData collection\u003c/h2\u003e \u003cp\u003eGSE7305 (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE7305\u003c/span\u003e\u003cspan address=\"https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE7305\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) and GSE11691(\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE11691\u003c/span\u003e\u003cspan address=\"https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE11691\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) datasets were downloaded from the Gene Expression Omnibus (GEO) database as training set by using the R package \u0026ldquo;GEOquery\u0026rdquo;. Details of datasets are listed in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. GSE11691 was in GPL96 platform, which consisted 9 endometriosis and 9 normal endometrial samples (Control samples). GSE7305 was in GPL570 platform, which consisted 10 endometriosis and 10 normal endometrial samples (Control samples). Then, we combined two datasets by using the \u0026ldquo;limma\u0026rdquo; package, the batch effect of the original gene expression landscapes in two datasets was eliminated using the comBat function based on the \u0026ldquo;sva\u0026rdquo; package. A total of 19 endometriosis samples and 19 normal endometrium samples with complete mRNA expression data and corresponding clinical materials were selected for subsequent analysis.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eInformation on the GEO dataset used in this study\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSample\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePlatforms\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eUse in this article\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eNormal\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eEndometriosis\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eAll\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGSE7305\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eGPL570\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eTraining set\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e20\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGSE11691\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eGPL96\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eTraining set\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e18\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGSE51981\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eGPL570\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eTraining set\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e71\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e77\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e148\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGSE23339\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eGPL6102\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eTest set\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e19\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGSE25628\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eGPL571\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eTest set\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e22\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cdiv id=\"Sec5\" class=\"Section3\"\u003e \u003ch2\u003eIdentification of differentially expressed endoplasmic reticulum stress-related genes in endometriosis\u003c/h2\u003e \u003cp\u003eWe extracted 1350 endoplasmic reticulum stress-related genes (ERSRGs) from previous articles. Genes with p\u0026thinsp;\u0026lt;\u0026thinsp;0.01 and |log\u003csub\u003e2\u003c/sub\u003eFC|\u0026gt;0.5 were considered differentially expressed genes (DEGs) by \u0026ldquo;limma\u0026rdquo; R package. The results were visualized using \u0026ldquo;ggplot2\u0026rdquo; and \u0026ldquo;ComplexHeatmap\u0026rdquo; packages.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003eAnalyzing and comparing immune infiltrating cells between endometriosis and normal endometrium\u003c/h2\u003e \u003cp\u003eCIBERSORT algorithm (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps:/cibersort.stanford.edu/\u003c/span\u003e\u003cspan address=\"https://cibersort.stanford.edu/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) was applied to estimate the abundance of 21 immunocyte subtypes in endometriosis and normal samples, which was performed on R software. ESTIMATE algorithm was adopted to measure immunocyte infiltration degree (ImmuneScore) in these samples. Student\u0026rsquo;s t-test was applied to verify the differences between the two groups, and results were presented using the \u0026ldquo;ggboxplot\u0026rdquo;, \u0026ldquo;pheatmap\u0026rdquo; and \u0026ldquo;corrplot\u0026rdquo; packages. Single sample gene set enrichment analysis (ssGSEA) was performed to determine enrichment scores for each coupling of a sample and immune reaction gene sets in the ImmPort database. The package \u0026ldquo;GSVA\u0026rdquo; was used for the ssGSEA analysis.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003eConstruction of weighted gene coexpression networks (WGCNA)\u003c/h2\u003e \u003cp\u003eThe package \u0026ldquo;WGCNA\u0026rdquo; was used for constructing the co-expression network of all genes between the 19 endometriosis and 19 normal samples. We obtained the difference between gene modules and endometriosis related modules hub genes. Hierarchical clustering was conducted on these samples to detect the outliers and remove the abnormal samples. The optimal power value was selected to transform the gene expression matrix into a weighted adjacency matrix, which was further transformed into a topological overlap matrix (TOM). Correlations between modules and traits were then calculated using the WGCNA package. Modules with high correlation coefficients were considered as candidates related to endometriosis and selected for subsequent analyses. With the candidate module selected, we defined |MM| (|Module membership|)\u0026thinsp;\u0026gt;\u0026thinsp;0.8 and |GS| (|gene significance|)\u0026thinsp;\u0026gt;\u0026thinsp;0.5 as the screening criteria for filtering key genes in the candidate module. The intersection of differentially expressed endometriosis genes related to ERS and hub genes in key modules were performed using the \u0026ldquo;jvenn\u0026rdquo; online website (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://jvenn.toulouse.inrae.fr/app/example.html\u003c/span\u003e\u003cspan address=\"https://jvenn.toulouse.inrae.fr/app/example.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eIdentification of ERS-related key genes in endometriosis and chromosome location\u003c/h2\u003e \u003cp\u003eBased on above genes obtained intersecting genes, we constructed of protein-protein interaction (PPI) networks in STRING database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://string-db\u003c/span\u003e\u003cspan address=\"http://string-db\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. org/). Subsequently, Cytoscape software was used to visualize the PPI network. And then, we used the Maximal Clique Centrality (MCC) algorithm in the cytoHubba (a Cytoscape plugin) to screen the key genes with high connectivity in PPI networks. Genomic locations of these key genes in chromosomes were demonstrated using \u0026ldquo;RCircos\u0026rdquo; package. Then, these key genes were processed with three machine learning algorithms. Least Absolute Shrinkage and Selection Operator (LASSO) to identify the genes by using the \u0026ldquo;glmnet\u0026rdquo; package. Random Forest Algorithm Analysis (RF) was conducted by the \u0026ldquo;randomForest\u0026rdquo; package. Meanwhile, a support vector machine-recursive feature elimination (SVM-RFE) model was established with a \u0026ldquo;SVM\u0026rdquo; package. Genes converged by three machine learning methods were confirmed as the hub ERS-related genes in endometriosis. The diagnostic and clustering discriminative ability of these selected hub genes was assessed through receiver operating characteristic (ROC) curves.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003eROC curve analysis\u003c/h2\u003e \u003cp\u003eWe chose GSE23339 and GSE25628 dataset as testing sets, performed receiver operating characteristic (ROC) curve analysis on each screened hub ERS-related genes to verify its accuracy. The \u0026ldquo;pROC\u0026rdquo; package was used for ROC curve analysis. The hub genes with average AUC\u0026thinsp;\u0026gt;\u0026thinsp;0.75 in both training and testing set were considered useful for EMS diagnosis.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003eUnsupervised clustering of hub ERSRGs in endometriosis\u003c/h2\u003e \u003cp\u003eAccording to the expression level of hub ERSRGs, an unsupervised clustering analysis was carried out to classify 77 endometriosis samples of GSE51981 into distinct clusters by using the R package of \u0026ldquo;ConsensusClusterPlus\u0026rdquo;. The optimal number of classifications was determined by the cumulative distribution function (CDF) curves, consistency clustering score of each cluster and consensus clustering plot. We used Principal Component Analysis (PCA) to verify the ERS gene expression patterns. Besides, we calculate the \u0026ldquo;ERSscore\u0026rdquo; and \u0026ldquo;ImmuneScore\u0026rdquo; in distinct ERS cluster. Those with ERS scores greater than zero were categorized as \u0026ldquo;high ERSscore group\u0026rdquo; and those less than zero were categorized as \u0026ldquo;low ERSscore group\u0026rdquo;. The DEGs between the two clusters were identified by the \u0026ldquo;limma\u0026rdquo; package (adjusted P\u0026thinsp;\u0026lt;\u0026thinsp;0.05, |log\u003csub\u003e2\u003c/sub\u003eFC| \u0026gt; 1).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eFunctional enrichment analysis and identification of potential drugs\u003c/h2\u003e \u003cp\u003eBased on the cluster DEGs, the GO and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses between the two subgroups were performed using the \u0026ldquo;clusterProfiler\u0026rdquo; package. Gene set enrichment analysis (GSEA) is used to evaluate the correlation of cluster DEGs in pre-defined gene set with each ERS cluster. The \u0026ldquo;c7.all.v2023.1.Hs. symbols\u0026rdquo; and \u0026ldquo;c2. kegg. immune. v2023.1.Hs. symbols\u0026rdquo; gene sets in the MSigDB database were subjected to GSEA using the \u0026ldquo;clusterProfiler\u0026rdquo; package. P-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05 was considered statistically significant. For each ERS cluster, we uploaded the cluster DEGs to the connective map (CMap) database and then applied enrichment analysis to pull out significantly enriched drugs. We used filter (with p\u0026thinsp;\u0026lt;\u0026thinsp;0.05) to identify top 60 compounds per ERS cluster that are predicted to treat EMS.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eClinical trait analysis and diagnostic model construction\u003c/h2\u003e \u003cp\u003eBesides, the R package \u0026ldquo;ggalluvial\u0026rdquo; constructed a Sankey diagram for demonstrating the distribution trend of ERS cluster, endometriosis severity and ERS score type. Each row represents a feature variable, different color represents different typing or clinical trait, lines represent the distribution of the same sample in different feature variables. Based on above hub ERSRGs obtained from intersection of machine learning methods, we built a nomograph using the \u0026ldquo;rms\u0026rdquo; and \u0026ldquo;rmda\u0026rdquo; package and evaluated its predictive power using the calibration curve to predict the probability of EMS.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eImmunohistochemistry (IHC) and image analysis\u003c/h2\u003e \u003cp\u003eThe immunohistochemical staining of paraffin-embedded sections from all samples underwent morphological screening examination and was conducted following the established protocol previously described in the literature[\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. Briefly, the sections were stained with the primary antibody against F8 (1:500, affinity, USA), VCAM1(1:200, Proteintech, Wuhan, China), VWF (1:400, Proteintech, Wuhan, China), and EPAS1(1:50, Proteintech, Wuhan, China) and then scanned with a digital scanner. The intensity of staining was quantified using the ImageJ software.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eQuantitative real-time PCR (qRT-PCR)\u003c/h2\u003e \u003cp\u003eTotal RNA was isolated from ovarian endometriosis, eutopic endometrium and normal control endometrium. We used Trizol (Yeasen, Shanghai, China) to isolate RNA from tissues according to the manufacturer\u0026rsquo;s instructions. Then, one microgram of RNA was reversely transcribed into cDNA using BeyoRT\u0026trade; III cDNA First Strand cDNA Synthesis Kit (Beyotime, Shanghai, China). Amplification was performed using gene-specific primers (Qingke, Beijing, China) and SYBR Green qPCR Mix 2\u0026times; (Beyotime, Shanghai, China) on a qRT-PCR device (ABI StepOnePlus, America). GAPDH was used as an internal control. The relative expression of the genes was calculated using the 2\u003csup\u003e\u0026ndash;ΔΔCT\u003c/sup\u003e method. The primers used were as follows\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eqRT-PCR primers\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGenes\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eForward primers\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eReverse primers\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEPAS1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eGGCTGTGTCTGAGAAGAGTAACT\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eTCCCGAAATCCAGAGAGATGATG\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eF8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTGGAGTTGATGGGCTGTGATTTA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCCAGGTGGCAAACATATTGGTAAA\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVCAM1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCGGAGACAGGAGACACAGTACTA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eGCACGAGAAGCTCAGGAGAAA\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVWF\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eGGGAAGACTGTGATGATCGATGT\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eGCAAACATCTCCCACAACATTCA\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGAPDH\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eGGAGTCCACTGGCGTCTTCA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eGTCATGAGTCCTTCCACGATACC\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eStatistical analysis\u003c/h2\u003e \u003cp\u003eAll statistical analyses were performed using the R programming software (Version R4.3.1). Comparisons between two groups were performed using the t-test. Spearman test was utilized for correlation analysis. A difference of p\u0026thinsp;\u0026lt;\u0026thinsp;0.05 indicated statistical significance unless specified otherwise. (\u0026lowast;\u0026lowast;\u0026lowast; represents p\u0026thinsp;\u0026lt;\u0026thinsp;0.001, \u0026lowast;\u0026lowast; represents p\u0026thinsp;\u0026lt;\u0026thinsp;0.01, and \u0026lowast; represents p\u0026thinsp;\u0026lt;\u0026thinsp;0.05)\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eThe landscape of different expressed ERSRGs and immune cell characteristics between EMS and normal samples\u003c/h2\u003e \u003cp\u003eAfter moving batch effect, differential gene expression analysis was performed using combined dataset of GSE7305 and GSE11691. We totally gained 1350 ERSRGs from published articles, among which 1180 ERSRGs are included in the combined dataset. Heatmaps of top 50 ERS-related DEGs are shown (Fig.\u0026nbsp;1a), the DEGs exhibit significantly different expression patterns between the EMS and normal samples. A total of 209 ERS-related DEGs (|log\u003csub\u003e2\u003c/sub\u003eFC|\u0026gt;0.5 and p\u0026thinsp;\u0026lt;\u0026thinsp;0.01) were obtained, including 92 up-regulated and 117 down-regulated genes (Fig.\u0026nbsp;1b).\u003c/p\u003e \u003cp\u003eThe proportion of different infiltrating immune cell types between the EMS and normal groups was evaluated using the CIBERSORT algorithm (Fig.\u0026nbsp;1c and Fig.\u0026nbsp;1d). After removing populations with a sum of immune abundance value zero, Wilcox test was used to assess the enrichment scores representing immunocyte abundance and immune response activity between EMS and normal samples. As can be concluded from the heatmap, the level of immune cell infiltration in EMS was significantly higher than that in normal tissues, suggesting that EMS is a disease closely related to immune dysregulation, which is consistent with previous findings (Fig.\u0026nbsp;1e).\u003c/p\u003e \u003cp\u003eIn comparison to normal endometrium, activated.B.cell, CD56bright.natural.killer.cell, immature.B.cell, myeloid-derived suppressor cells(MDSCs), macrophage, mast.cell, natural.killer.T.cell, natural.killer.cell, regulatory.T.cell, T.follicular.helper.cell, type.1.T.helper.cell, effector.memory.CD4.T.cell, central.memory.CD4.T.cell, central.memory.CD8.T.cell were found to have higher level of infiltration in EMS (Fig.\u0026nbsp;1f).\u003c/p\u003e \u003cp\u003eBesides, we conducted the correlation analysis between 21 immunocyte subtypes. We found that dendritic cells activated and plasma cells were the most positive relevant immune cells in EMS samples (r\u0026thinsp;=\u0026thinsp;0.64), monocytes and eosinophils were the most positive relevant immune cells in normal samples (r\u0026thinsp;=\u0026thinsp;0.62), which may indicate that these immunocytes work together (Fig.\u0026nbsp;1g and 1h).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eIdentification of key modules associated with the ERS through WGCNA\u003c/h2\u003e \u003cp\u003eThe soft -threshold power was set to 3 based on the scale-free network construction (Fig.\u0026nbsp;2a-2d). Five co-expression modules were identified through WGCNA analysis (Fig.\u0026nbsp;2e). The turquoise module showed highly positively correlation with EMS mRNA (r\u0026thinsp;=\u0026thinsp;0.86, p\u0026thinsp;=\u0026thinsp;5e-12) (Fig.\u0026nbsp;2f). Next, the turquoise module was selected as key modules relevant to EMS for further analysis. In Fig.\u0026nbsp;3F, the significant correlations between gene significance (GS) and module membership (MM) were presented in the turquoise module, 239 module hub genes were found in the two modules by GS\u0026thinsp;\u0026gt;\u0026thinsp;0.5 and MM\u0026thinsp;\u0026gt;\u0026thinsp;0.8 (Fig.\u0026nbsp;2g). Hub genes of module turquoise were intersected with 209 different expressed ERSRGs. There were 29 intersecting genes in the venn diagram (Fig.\u0026nbsp;2h).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003eIdentification of hub ERSRGs in endometriosis\u003c/h2\u003e \u003cp\u003ePPI network analysis was performed on 29 genes. The cytoHubba of Cytoscape was used to determine the hub genes in the PPI network (Fig.\u0026nbsp;3a). Top 10 genes in network ranked by MCC method were obtained, namely ESR1, KPNA2, CAV1, TXN, CDK1, VWF, EZH2, VCAM1, EPAS1, F8. Figure\u0026nbsp;3b displays the chromosomal locations of top 10 ERSRGs that are differentially expressed. LASSO regression was performed on top 10 ERSRGs for feature selection and dimensionality reduction to exclude unimportant regulators, which ultimately identified 4 hub ERSRGs (Fig.\u0026nbsp;3c-3d). For the random forest algorithm, six characteristic genes with relative importance\u0026thinsp;\u0026gt;\u0026thinsp;2 were determined, including VCAM1, EPAS1, F8, VWF, ESR1, CAV1. (Fig.\u0026nbsp;3e-3f). For the SVM-RFE algorithm, when the feature number was six, the classifier had the minimum error, containing VCAM1, EPAS1, TXN, F8, VWF, ESR1 (Fig.\u0026nbsp;3g). Following intersection, four characteristic genes shared by LASSO, RF, and SVM-RFE algorithms were finally identified (VWF, VCAM1, EPAS1, F8). (Fig.\u0026nbsp;3h) We calculate the average AUC of training set (Fig.\u0026nbsp;4a) and test sets (Fig.\u0026nbsp;4b-4c), the average AUC values for all four genes exceeded 0.75 (EPAS1\u0026thinsp;=\u0026thinsp;0.786, F8\u0026thinsp;=\u0026thinsp;0.774, VCAM1\u0026thinsp;=\u0026thinsp;0.890, VWF\u0026thinsp;=\u0026thinsp;0.866), demonstrating the good discriminatory efficacy of screened hub ERSRGs.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003eIdentification of ERS clusters\u003c/h2\u003e \u003cp\u003eFour hub ERSRGs in EMS, including five up-regulated and five down-regulated genes, were used to cluster the EMS datasets GSE51981, which includes 77 endometriosis samples. We performed unsupervised consensus cluster analysis based on the expression of four hub ERSRGs using the \u0026ldquo;ConsensusClusterPlus\u0026rdquo; R package. We observed stable isoform numbers when k\u0026thinsp;=\u0026thinsp;2, and significant differences in the relative changes in the area under the CDF curve from k\u0026thinsp;=\u0026thinsp;2 to k\u0026thinsp;=\u0026thinsp;9 (Fig.\u0026nbsp;5a-d). The consistency scores of the subtypes were all over 0.9 when k\u0026thinsp;=\u0026thinsp;2 (Fig.\u0026nbsp;5e). We found two distinct ERS modification subclusters, with 51 samples in cluster A and 26 samples in cluster B.PCA verified the remarkable difference between the clusters (Fig.\u0026nbsp;5f).\u003c/p\u003e \u003cp\u003eTo better understand the molecular characteristics between subtypes, we evaluated the differences in the expression of 4 ERSRGs (Fig.\u0026nbsp;6a). The results showed that EPAS1, F8, VCAM1 presented a higher expression in B than A cluster. We used ten violin plots revealed significant differences in gene expression patterns between the two clusters (Fig.\u0026nbsp;6b-k), indicating that the various ERS clusters may have various transcriptome or other characteristics. We also calculated the ERS scores based on the four hub ERSRGs obtained above and compared them using the PCA method. The ERS score of EMS were higher than control samples (Fig.\u0026nbsp;6l) and ERS score of cluster B were higher than those of cluster A (Fig.\u0026nbsp;6m).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec21\" class=\"Section2\"\u003e \u003ch2\u003eCharacteristics of the immune microenvironment in distinct ERS cluster\u003c/h2\u003e \u003cp\u003eTo identify differences in immune microenvironmental characteristics, we evaluated immune cells between these different ERS cluster. Significantly, Cluster B is clearly more immunologically active, while cluster A shows a relative defective immune cell infiltration. In Fig.\u0026nbsp;7a, we noticed that there was remarkable heterogeneity in in the abundance of immune cell infiltration between distinct cluster. Cluster B presented higher infiltration levels of activated.B.cell, activated.CD8.T.cell, activated.dendritic.cell, CD56bright.natural.killer.cell, CD56dim.natural.killer.cell, gamma.delta.T.cell, immature.B.cell, MDSC, macrophage, mast.cell, natural.killer.T.cell, natural.killer.cell, neutrophil, regulatory.T.cell, T.follicular.helper.cell, type.1.T.helper.cell, type.17.T.helper.cell, effector.memory.CD4.T.cell, memory.B.cell, central.memory.CD4.T.cell, effector.memory.CD8.T.cell (Fig.\u0026nbsp;7b). After that, we calculated the immune scores of the samples using the ESTIMATE algorithm, the immune scores of cluster B were significantly higher than that of cluster A (Fig.\u0026nbsp;7c). Collectively, we identified cluster B as an immune subtype and cluster A as a less-immune subtype.\u003c/p\u003e \u003cp\u003eFurthermore, we explored the correlation between 10 ERSRGs and immune cells (Fig.\u0026nbsp;7d). Correlation analysis revealed that 10 ERSRGs were closely associated with most immune cells. For example, KPNA2 had the strongest positive correlation with immature.dendritic.cell abundance (r\u0026thinsp;=\u0026thinsp;0.72) (Fig.\u0026nbsp;7e), and CAV1 had the strongest negative correlation with activated.dendritic.cell abundance (r=-0.71) (Fig.\u0026nbsp;7f). This indicates that KPNA2 and CAV1 play important roles in immunoinflammatory response in EMS.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec22\" class=\"Section2\"\u003e \u003ch2\u003eExplore the difference biological behaviors between ERS clusters\u003c/h2\u003e \u003cp\u003eWe identified ERS cluster DEGs in order to explain the gene profiles related to cluster-mediated biological function regulation. A total of 451 cluster DEGs were screened out (Fig.\u0026nbsp;8a), including 127 DEGs with upregulation in cluster A and 324 DEGs with upregulation in cluster B. In the BP analysis of GO, cluster DEGs mainly participated in cell adhesion, regulation of T cell activation, ERK1 and ERK2 cascade, natural killer cell mediated immunity. In CC analysis, cluster DEGs mainly focused on collagen\u0026thinsp;\u0026minus;\u0026thinsp;containing extracellular matrix, external side of plasma membrane, vesicle lumen, external side of plasma membrane, cytoplasmic vesicle lumen and secretory granule lumen. MF analysis showed that cluster DEGs mainly related to extracellular matrix structural constituent, immune receptor activity, cytokine binding (Fig.\u0026nbsp;8b-d). In addition, in the KEGG analysis, the selected biological process of DEG enrichment significantly participated in biological pathways such as natural killer cell mediated cytotoxicity, PI3K\u0026thinsp;\u0026minus;\u0026thinsp;Akt signaling pathway, progesterone-mediated oocyte maturation, focal adhesion, chemokine signaling pathway, cytokine\u0026ndash;cytokine receptor interaction pathway, VEGF signaling pathway, TNF signaling pathway and growth hormone regulation (Fig.\u0026nbsp;8e-f). We also conducted the GSEA analysis, the results showed that role of mammalian E proteins E2A and HEB in the development of T cells and N-ras in T cell development and function were significantly enriched in ERS cluster B (Fig.\u0026nbsp;8g), while functions TRAF6 regulated CD8 T cell memory development following infection by modulating fatty acid metabolism, extrathymic Treg development and STAT6 down-regulated in bone marrow-derived macrophages were significantly enriched in cluster A (Fig.\u0026nbsp;8h).\u003c/p\u003e \u003cdiv id=\"Sec23\" class=\"Section3\"\u003e \u003ch2\u003eFind potential drugs and clinicopathological features of ERS clusters\u003c/h2\u003e \u003cp\u003eWe employed CMap database to identify potential therapeutic drugs for EMS. Differences between cluster A and cluster B were also apparent in compound prediction. Ten relevant compounds associated with distinct ERS cluster were identified. Next, we further evaluated the mechanism of actions (MOA) and drug target of these drugs to explore their potential mechanism for treating EMS. Notably, in cluster A, some adrenergic receptor antagonists, progesterone or progesterone receptor agonists, androgen receptor modulators, NF-κB pathway inhibitors, dipeptidyl peptidase inhibitors, and 5-hydroxytryptamine receptor agonists may have potential therapeutic roles (Fig.\u0026nbsp;9a), whereas in cluster B, some histone deacetylase inhibitors, protein kinase C (PKC) activators, PPAR receptor agonists and insulin sensitizers, adenylate cyclase activators, and caspase activators show a possible therapeutic role (Fig.\u0026nbsp;9b).\u003c/p\u003e \u003cp\u003eSankey diagram to visualize the relationships between the ERS phenotypes and EMS severity. The Sankey diagram verified the cluster A was associated with low ERS score, moderate or severe EMS. However, in cluster B with high ERS scores, mild EMS and moderately severe EMS had distributions with little difference. It is evident that the degree of clinical manifestation of EMS has greater heterogeneity among different ERS clusters and even within ERS cluster (Fig.\u0026nbsp;9c).\u003c/p\u003e \u003cp\u003eWe constructed a nomogram (Fig.\u0026nbsp;9d) to evaluate its predictive power using the calibration curve to predict the risk of EMS more clearly (Fig.\u0026nbsp;9e). Decision curve analysis (DCA) (Fig.\u0026nbsp;9f) and clinical impact curve (Fig.\u0026nbsp;9g) indicated that the \u0026ldquo;nomogram\u0026rdquo; curve was higher than the gray line. The calibration curve indicated a minimal difference between the real and predicted EMS risks, suggesting that the nomograph model of EMS is precise.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec24\" class=\"Section2\"\u003e \u003ch2\u003eValidation mRNA and protein expression in collected human endometrial tissue\u003c/h2\u003e \u003cp\u003eThe results of bioinformatics analysis suggested that EPAS1, F8, VCAM1, and VWF might be highly expressed in endometriosis. After RT-qPCR, the mRNAs of the four ERS hub genes in ectopic endometrium were significantly higher than those in the eutopic endometrium and normal control endometrium (Fig.\u0026nbsp;10a). We performed immunohistochemical staining on normal control endometrial specimens, ectopic endometrial specimens and eutopic endometrial specimens in order to detect the protein expression. The results showed that the protein expression levels of EPAS1, F8, VCAM1, and VWF were consistent with those of mRNA (Fig.\u0026nbsp;10b-c).\u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eCurrently, there is a lack of biomarkers with both accuracy and sensitivity for the diagnosis of endometriosis. Endometriosis is an estrogen-dependent inflammatory immune disorder. In addition to surgical removal of endometriosis lesions and loosening of pelvic adhesions, hormonal therapy remains the first line of pharmacologic treatment for endometriosis. Therefore, it is necessary to find the hub genes associated with the diagnosis of endometriosis and to explore the role of these hub genes in the pathogenesis of endometriosis, and drug therapy. In this study, we explored the role of endoplasmic reticulum stress-related genes in endometriosis immune infiltration, disease typing, potential therapeutic agents, biological functions, and pathways.\u003c/p\u003e \u003cp\u003eOur study confirmed that the levels of infiltration of a wide range of immune cells were significantly elevated in ectopic lesion samples, which is consistent with previous research, suggesting the presence of a heavy immune-inflammatory response in endometriosis. Several studies in recent years have mechanistically confirmed the relationship between endoplasmic reticulum stress and inflammatory immune responses. For example, endoplasmic reticulum stress-induced inflammation and production of the pro-inflammatory cytokine IL-6 have been reported to be dependent on two members of the NOD-like receptor family, NOD1 and NOD2[\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]. Another study on inflammatory bowel disease confirmed that endoplasmic reticulum stress induced by deletion of the transcription factor X-box binding protein-1 (XBP1) can lead to aberrant responses of intestinal epithelial cells to inflammatory signals[\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e]. In a study of endometriosis, treated endometrial stromal cells with peritoneal fluid from patients with endometriosis and normal control patients, the result suggested that endoplasmic reticulum stress-associated UPR pathways were activated in endometriosis[\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e]. Our study also showed that a relatively high degree of endoplasmic reticulum stress was present in ectopic lesion samples of endometriosis and that different endoplasmic reticulum stress subtypes had different degrees of UPR activation. However, initial activation of the UPR contributes to the endoplasmic reticulum's emergency response to adverse external signals, and over-activation leads to cell death, which is the entry point for many current studies in related fields to confirm the therapeutic effects of drugs on endometriosis. Our study of endoplasmic reticulum subtypes used the GSE51981 database, and Sankey plots showed that the degree of endoplasmic reticulum stress did not significantly correlate with the degree of clinical manifestations of endometriosis in this dataset. Is the specific level of endoplasmic reticulum stress in endometriosis associated with case typing and disease severity? This question remains to be further elucidated using large-scale transcriptomic data and clinical samples.\u003c/p\u003e \u003cp\u003eEstrogen and progesterone strictly regulate the physiologic cycle of the endometrium. These steroids also have receptors in the endoplasmic reticulum that are involved in the regulation of protein folding, calcium homeostasis in the endoplasmic reticulum, and degradation of misfolded proteins. Abnormal endoplasmic reticulum stress response to progesterone enhances the invasiveness of endometrial stromal cells in endometriosis through the AKT/mTOR pathway, which has attracted our interest in the role of endoplasmic reticulum stress in endometriosis[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. In this article, for the first time, we found the hub endoplasmic reticulum stress-related genes in endometriosis by bioinformatics methods and performed endoplasmic reticulum stress typing in endometriosis. First, we identified 10 characterized endoplasmic reticulum stress genes by differential analysis, WGCNA, PPI and Cytoscape. Second, machine learning was used to further screen these 10 genes for endoplasmic reticulum stress-centered genes. Subsequently, unsupervised clustering was performed using the screened VWF, VCAM1, EPAS1 and F8. We categorized endometriosis into two subtypes, A and B, which showed significant differences in the degree of immune cell infiltration and the type of immune cell infiltration. type B had a higher endoplasmic reticulum stress score and tended to be an immune cell-rich type, whereas type A had a lower endoplasmic reticulum stress score and a relatively low immune cell infiltration. In the comparative analysis of fractional immune infiltration, we found that several immune cell types with P\u0026thinsp;\u0026lt;\u0026thinsp;0.001, such as activated B cells, CD56 strongly positively expressing NK cells, immature B cells, myeloid-derived suppressor cells (MDSC), NKT cells, NK cells, regulatory T cells, follicular helper T cells, and helper T cells 1 (Th1), which have also been shown to be associated with immunomodulation in endometriosis. For example, in a 1995 study, researchers found a strong correlation between NKT cell-mediated lysis of ectopic endometrial cells and downregulation of HLA1-like receptors[\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e]. Patients with severe endometriosis had significantly higher levels of CTLA-4\u003csup\u003e+\u003c/sup\u003e T cells than those with mild endometriosis. In addition, CTLA-4\u003csup\u003e+\u003c/sup\u003eT lymphocytes were negatively correlated with the percentage of NK and NKT-like cells in women with both endometriosis and infertility[\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e], which indicate that immunologic mechanisms of associated infertility may differ between endoplasmic reticulum stress subtypes and endometriosis clinicopathologic types and degrees of clinical manifestations of endometriosis.\u003c/p\u003e \u003cp\u003eIn the functional enrichment analysis of genes differing in endoplasmic reticulum stress subtypes, we noted significant functional enrichment of genes for extracellular matrix and cell adhesion. Endoplasmic reticulum stress enhances leukocyte recruitment through extracellular signals, remodels the immune microenvironment, and alters the behavior of immune cells through the secretion of polarizing cytokines, and in doing so restores tissue protein homeostasis. Muscle cells and airway epithelial cells have been reported to secrete a functional leukocyte-adherent hyaluronic acid matrix after various forms of endoplasmic reticulum stress[\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e]. Endometriosis lesions are highly resistant to apoptosis and cell adhesion, and attenuating endoplasmic reticulum stress-induced ectopic cell adhesion may be an important treatment for endometriosis. A related study showed that Frankincense could alleviate endometriosis by reducing the adhesion and proliferation of ectopic endometrial cells through endoplasmic reticulum stress/p53-apoptosis and chemokine-migration/adhesion pathways[\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e]. Notably, activated T cells showed significant enrichment in functional enrichment analysis and gene-immunocyte correlation analysis. We found that CAV1 was negatively correlated with multiple T cell subtypes. In a previous study, knockdown of Sirt1 in endothelial cells induced endoplasmic reticulum stress and miR-204 expression in endothelial cells, decreased CAV1, and impaired endothelium-dependent vasodilation, but there are no studies on CAV1 in relation to endometriosis[\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e]. In addition, we found a positive correlation between immature dendritic cells (iDC) and several endoplasmic reticulum-related genes. The proportion of iDCs was increased in the peritoneal cavity of patients with endometriosis compared to mature dendritic cells (mDCs), and maturation of dendritic cells in the peritoneal cavity plays an important role in the development of endometriosis[\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e, \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e]. However, in our immune cell differential analysis, there were no significant differences in iDCs between endometriosis and control samples, or in endoplasmic reticulum stress cluster A versus cluster B.\u003c/p\u003e \u003cp\u003eOur study identified four endoplasmic reticulum stress center genes, VWF, VCAM1, EPAS1, and F8. VWF is a macromolecular plasma protein that plays a key role in maintaining normal coagulation and contributes to thrombotic disorders following endothelial and platelet dysfunction. The VWF gene is expressed predominantly in vascular endothelium and megakaryocytes, and its expression level is commonly used as a measure of angiogenesis capacity[\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e, \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e]. A Mendelian randomization study based on GWAS data from a large population showed a causal relationship between elevated VWF and an increased risk of endometriosis[\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e]. In lesions treated with GnRH agonists (GnRH-a), microvessel density was significantly reduced in patients with positive VWF expression[\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e]. However, protein expression of VWF did not show r-ASRM stage-dependent changes in endometriosis[\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e]. VCAM1 is a cell adhesion molecule, which is often used as a relevant measure of inflammation and malignancy cell adhesion capacity in studies related to a number of diseases, and has also been associated with transvascular endothelial migration of immune cells[\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e]. And estrogen-mediated upregulation of VCAM1 contributes to mast cell recruitment and differentiation[\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e]. As a classical inflammation-associated gene, knockdown of VCAM1 impedes TGF-β1-mediated endometrial cell proliferation, migration and invasion as well as attenuates inflammatory responses in endometriotic lesions[\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e, \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e]. Mechanistic studies of endometriosis associated with EPAS1 are lacking, but it is certain that EPAS1, as a key transcription factor in cellular response to hypoxia, is closely associated with inflammatory response and angiogenesis under hypoxic conditions, and more importantly, EPAS1 expressed in the endometrial stroma has been associated with invasion of trophoblast cells during embryo implantation, and mice knocked out of EPAS1 expression were infertile due to infertility due to failure of implantation[\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e, \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e]. Whether EPAS1 is involved in endometriosis-associated infertility may be a new research direction for the future. A study in 2018 confirmed that EPAS1 expression was significantly upregulated in CD73\u003csup\u003e+\u003c/sup\u003eCD90\u003csup\u003e+\u003c/sup\u003eCD105\u003csup\u003e+\u003c/sup\u003e pluripotent stem cells isolated from ectopic endometrium compared to paired in situ endometrium, which may be associated with the progression of ovarian endometriosis to associated ovarian cancer[\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e]. The F8 gene is required for the production of coagulation factor VIII, which is essential for clot formation. In a mouse endometriosis model study, researchers found that the F8 antibody had a strong targeting effect on endometriotic tissues, based on which the study significantly reduced the size of endometriosis lesions using the F8 antibody-IL-10 fusion protein[\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e], and a similar study found that the F8 antibody-IL4 can similarly reduce the size of endometriosis lesions[\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e]. We look forward to the future discovery of upstream and downstream pathways and other molecules interacting with these four key genes, and to the thorough elucidation of the involvement of endoplasmic reticulum stress-related mechanisms in the pathogenesis and potential therapeutic drug molecular biology of endometriosis.\u003c/p\u003e \u003cp\u003eThere exists significant heterogeneity in clinical symptomatology and disease severity among individuals with endometriosis; individuals exhibiting high ASMR and EFI scores may manifest mild or asymptomatic pain, while those presenting with minimal lesions or mild pelvic adhesions observed laparoscopically may experience more severe pain. Within the scope of our investigation, the classification and grading of endoplasmic reticulum stress did not demonstrate a significant correlation with the severity of endometriosis. There is a notable absence of diagnostic markers demonstrating high sensitivity and specificity for endometriosis. The diagnostic model constructed in this investigation underwent validation through diverse algorithms, demonstrating favorable performance in discriminating endometriosis samples from control endometrial specimens. Furthermore, the endoplasmic reticulum stress-associated core genes identified herein hold promise for clinical application in the diagnosis of endometriosis.\u003c/p\u003e \u003cp\u003ePresently, the primary modalities for addressing endometriosis encompass laparoscopic surgery and hormonal therapy. Moreover, our investigation has delineated compounds that target distinct subtypes of endoplasmic reticulum stress, potentially offering novel avenues for both scientific inquiry and clinical intervention in the management of endometriosis. It is reasonable to hypothesize that these compounds may promote the development and progression of endometriosis by inducing or inhibiting endoplasmic reticulum stress. Interestingly, our study found that subtypes with high endoplasmic reticulum stress scores also had higher immune cell infiltration scores, and this feature was not only reflected among endoplasmic reticulum stress subtypes, but also showed the same trend between endometriosis samples and control samples.\u003c/p\u003e \u003cp\u003eOur study also presents certain limitations. Specifically, we did not investigate the potential association between endoplasmic reticulum stress subtypes and clinical phenotypes of endometriosis. It is pertinent to inquire whether disparities exist in endoplasmic reticulum stress-related gene expression profiles and the associated immune cell infiltration patterns between ovarian and non-ovarian endometriosis. Additionally, despite the utilization of multiple algorithms to validate the accuracy of our diagnostic models, the diagnostic utility of endoplasmic reticulum stress-related genes in endometriosis diagnosis warrants further substantiation through comprehensive investigations. Moreover, the absence of publicly available transcriptome sequencing data encompassing samples representing diverse pathologic types of endometriosis poses a challenge, and the outcomes of bioinformatics analyses may be subject to certain biases attributable to variations in sequencing platforms, statistical methodologies, and databases employed. Finally, this study did not use cell lines ore animal models to validate the screened genes, immune cells, and potential therapeutic compounds. In order to draw more convincing conclusions, further studies on the involvement of endoplasmic reticulum stress-related genes in the pathogenesis of endometriosis are necessary next.\u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003eThe expression profiles of key endoplasmic reticulum stress-related genes, including EPAS1, F8, VCAM1, and VWF, exhibited significant disparities between endometriosis specimens and both adjacent normal endometrial tissues and control endometrial tissues. These pivotal genes were utilized for stratifying endometriosis cases into distinct immune and non-immune subgroups. Furthermore, distinct patterns were observed in the expression levels of endoplasmic reticulum stress-associated genes, endoplasmic reticulum stress scores, immune cell infiltration levels, biological functionalities, signaling pathways, and potential therapeutic targets across different endoplasmic reticulum stress subtypes of endometriosis. Notably, a diagnostic model based on the expression profiles of endoplasmic reticulum stress-related genes demonstrated robust discriminatory capacity for identifying endometriosis cases.\u003c/p\u003e"},{"header":"List of abbreviations","content":"\u003cp\u003eAUC Area Under Curve\u003c/p\u003e\n\u003cp\u003eBP Biological Processes\u003c/p\u003e\n\u003cp\u003eCC Cellular Component\u003c/p\u003e\n\u003cp\u003eEMS Endometriosis\u003c/p\u003e\n\u003cp\u003eERS Endoplasmic reticulum stress\u003c/p\u003e\n\u003cp\u003eGEO Gene Expression Omnibus\u003c/p\u003e\n\u003cp\u003eGO Gene Ontology\u003c/p\u003e\n\u003cp\u003eGS Gene Significance\u003c/p\u003e\n\u003cp\u003eGSEA Gene set enrichment analysis\u003c/p\u003e\n\u003cp\u003eKEGG Kyoto Encyclopedia of Genes and Genomes\u003c/p\u003e\n\u003cp\u003eLASSO Least Absolute Shrinkage and Selection Operator\u003c/p\u003e\n\u003cp\u003eMF Molecular Function\u003c/p\u003e\n\u003cp\u003eMM Module Membership\u003c/p\u003e\n\u003cp\u003ePCA Principal Component Analysis\u003c/p\u003e\n\u003cp\u003ePPI Protein-Protein Interaction networks\u003c/p\u003e\n\u003cp\u003eRF Random Forest\u003c/p\u003e\n\u003cp\u003eRMSE Root Mean Squared Error\u003c/p\u003e\n\u003cp\u003eROC Receiver Operating Characteristic\u003c/p\u003e\n\u003cp\u003eSVM-RFE Support Vector Machine-Recursive Feature Elimination\u003c/p\u003e\n\u003cp\u003eWGCNA Weighted Correlation Network Analysis\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe collection of endometriosis samples from human subjects was approved by the Ethics Committee of Tongji Medical College, Huazhong University of Science and Technology and followed the tenets of the Declaration of Helsinki. The patients provided their written informed consent to participate in this study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe public datasets GSE7305, GSE11691, GSE23339, GSE25628 and GSE51981 used in this paper are available on the NCBI website (https://www.ncbi.nlm.nih.gov/geo/).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was supported by grant from National Natural Science Foundation of China (No. 82371657).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eL Chen and X Wang proposed the idea and reviewed the manuscript, E Huang and L Zhang drafted and revised the initial manuscript. E Huang and J Lou performed the experiments and analyzed the data. All authors read and approved the final manuscript.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eVercellini P, Vigan\u0026ograve; P, Somigliana E, Fedele L. Endometriosis: pathogenesis and treatment. Nat Rev Endocrinol. 2014;10(5):261-75.http://doi.org/10.1038/nrendo.2013.255\u003c/li\u003e\n\u003cli\u003eChapron C, Marcellin L, Borghese B, Santulli P. Rethinking mechanisms, diagnosis and management of endometriosis. Nat Rev Endocrinol. 2019;15(11):666-82.http://doi.org/10.1038/s41574-019-0245-z\u003c/li\u003e\n\u003cli\u003eOlive DL, Schwartz LB. Endometriosis. N Engl J Med. 1993;328(24):1759-69.http://doi.org/10.1056/nejm199306173282407\u003c/li\u003e\n\u003cli\u003eGreene AD, Lang SA, Kendziorski JA, Sroga-Rios JM, Herzog TJ, Burns KA. Endometriosis: where are we and where are we going? Reproduction. 2016;152(3):R63-78.http://doi.org/10.1530/rep-16-0052\u003c/li\u003e\n\u003cli\u003ede Ziegler D, Borghese B, Chapron C. Endometriosis and infertility: pathophysiology and management. Lancet. 2010;376(9742):730-8.http://doi.org/10.1016/s0140-6736(10)60490-4\u003c/li\u003e\n\u003cli\u003eShen HH, Zhang T, Yang HL, Lai ZZ, Zhou WJ, Mei J, et al. Ovarian hormones-autophagy-immunity axis in menstruation and endometriosis. Theranostics. 2021;11(7):3512-26.http://doi.org/10.7150/thno.55241\u003c/li\u003e\n\u003cli\u003eLiang Y, Yao S. Potential role of estrogen in maintaining the imbalanced sympathetic and sensory innervation in endometriosis. Mol Cell Endocrinol. 2016;424:42-9.http://doi.org/10.1016/j.mce.2016.01.012\u003c/li\u003e\n\u003cli\u003eDutta S, Banu SK, Arosh JA. Endocrine disruptors and endometriosis. Reprod Toxicol. 2023;115:56-73.http://doi.org/10.1016/j.reprotox.2022.11.007\u003c/li\u003e\n\u003cli\u003eBulun SE, Zeitoun K, Takayama K, Noble L, Michael D, Simpson E, et al. Estrogen production in endometriosis and use of aromatase inhibitors to treat endometriosis. Endocr Relat Cancer. 1999;6(2):293-301.http://doi.org/10.1677/erc.0.0060293\u003c/li\u003e\n\u003cli\u003eBerkkanoglu M, Arici A. Immunology and endometriosis. Am J Reprod Immunol. 2003;50(1):48-59.http://doi.org/10.1034/j.1600-0897.2003.00042.x\u003c/li\u003e\n\u003cli\u003eKisovar A, Becker CM, Granne I, Southcombe JH. The role of CD8+ T cells in endometriosis: a systematic review. Front Immunol. 2023;14:1225639.http://doi.org/10.3389/fimmu.2023.1225639\u003c/li\u003e\n\u003cli\u003eYoung VJ, Ahmad SF, Duncan WC, Horne AW. The role of TGF-\u0026beta; in the pathophysiology of peritoneal endometriosis. Hum Reprod Update. 2017;23(5):548-59.http://doi.org/10.1093/humupd/dmx016\u003c/li\u003e\n\u003cli\u003eMwaura AN, Riaz MA, Maoga JB, Mecha E, Omwandho COA, Scheiner-Bobis G, et al. Activin A Modulates Betaglycan Shedding via the ALK4-SMAD3-Dependent Pathway in Endometriotic Cells. Biomolecules. 2022;12(12).http://doi.org/10.3390/biom12121749\u003c/li\u003e\n\u003cli\u003eArablou T, Aryaeian N, Khodaverdi S, Kolahdouz-Mohammadi R, Moradi Z, Rashidi N, et al. The effects of resveratrol on the expression of VEGF, TGF-\u0026beta;, and MMP-9 in endometrial stromal cells of women with endometriosis. Sci Rep. 2021;11(1):6054.http://doi.org/10.1038/s41598-021-85512-y\u003c/li\u003e\n\u003cli\u003eLurje I, Gaisa NT, Weiskirchen R, Tacke F. Mechanisms of organ fibrosis: Emerging concepts and implications for novel treatment strategies. Mol Aspects Med. 2023;92:101191.http://doi.org/10.1016/j.mam.2023.101191\u003c/li\u003e\n\u003cli\u003eGhoreschi K, Laurence A, Yang XP, Tato CM, McGeachy MJ, Konkel JE, et al. Generation of pathogenic T(H)17 cells in the absence of TGF-\u0026beta; signalling. Nature. 2010;467(7318):967-71.http://doi.org/10.1038/nature09447\u003c/li\u003e\n\u003cli\u003eMcKinnon BD, Evers J, Bersinger NA, Mueller MD. Induction of the neurokinin 1 receptor by TNF\u0026alpha; in endometriotic tissue provides the potential for neurogenic control over endometriotic lesion growth. J Clin Endocrinol Metab. 2013;98(6):2469-77.http://doi.org/10.1210/jc.2013-1019\u003c/li\u003e\n\u003cli\u003eVelho RV, Taube E, Sehouli J, Mechsner S. Neurogenic Inflammation in the Context of Endometriosis-What Do We Know? Int J Mol Sci. 2021;22(23).http://doi.org/10.3390/ijms222313102\u003c/li\u003e\n\u003cli\u003eRashid HO, Yadav RK, Kim HR, Chae HJ. ER stress: Autophagy induction, inhibition and selection. Autophagy. 2015;11(11):1956-77.http://doi.org/10.1080/15548627.2015.1091141\u003c/li\u003e\n\u003cli\u003eBhardwaj M, Leli NM, Koumenis C, Amaravadi RK. Regulation of autophagy by canonical and non-canonical ER stress responses. Semin Cancer Biol. 2020;66:116-28.http://doi.org/10.1016/j.semcancer.2019.11.007\u003c/li\u003e\n\u003cli\u003eSen Santara S, Lee DJ, Crespo \u0026Acirc;, Hu JJ, Walker C, Ma X, et al. The NK cell receptor NKp46 recognizes ecto-calreticulin on ER-stressed cells. Nature. 2023;616(7956):348-56.http://doi.org/10.1038/s41586-023-05912-0\u003c/li\u003e\n\u003cli\u003eDi Conza G, Ho PC. ER Stress Responses: An Emerging Modulator for Innate Immunity. Cells. 2020;9(3).http://doi.org/10.3390/cells9030695\u003c/li\u003e\n\u003cli\u003eCubillos-Ruiz JR, Mohamed E, Rodriguez PC. Unfolding anti-tumor immunity: ER stress responses sculpt tolerogenic myeloid cells in cancer. J Immunother Cancer. 2017;5:5.http://doi.org/10.1186/s40425-016-0203-4\u003c/li\u003e\n\u003cli\u003eSpagnolo E, Martinez A, Mascar\u0026oacute;s-Mart\u0026iacute;nez A, Mar\u0026iacute;-Alexandre J, Carbonell M, Gonz\u0026aacute;lez-Cant\u0026oacute; E, et al. Evaluation of Immune Infiltrates in Ovarian Endometriosis and Endometriosis-Associated Ovarian Cancer: Relationship with Histological and Clinical Features. Int J Mol Sci. 2023;24(15).http://doi.org/10.3390/ijms241512083\u003c/li\u003e\n\u003cli\u003eChoi J, Jo M, Lee E, Lee DY, Choi D. Nuclear factor-kappa B signaling in endometriotic stromal cells is not inhibited by progesterone owing to an aberrant endoplasmic reticulum stress response: a possible role for an altered inflammatory process in endometriosis. Mol Hum Reprod. 2021;27(2).http://doi.org/10.1093/molehr/gaab002\u003c/li\u003e\n\u003cli\u003ePark S, Lim W, Bazer FW, Song G. Naringenin induces mitochondria-mediated apoptosis and endoplasmic reticulum stress by regulating MAPK and AKT signal transduction pathways in endometriosis cells. Mol Hum Reprod. 2017;23(12):842-54.http://doi.org/10.1093/molehr/gax057\u003c/li\u003e\n\u003cli\u003eHam J, Park W, Song J, Kim HS, Song G, Lim W, et al. Fraxetin reduces endometriotic lesions through activation of ER stress, induction of mitochondria-mediated apoptosis, and generation of ROS. Phytomedicine. 2024;123:155187.http://doi.org/10.1016/j.phymed.2023.155187\u003c/li\u003e\n\u003cli\u003eHam J, Kim J, Bazer FW, Lim W, Song G. Silibinin-induced endoplasmic reticulum stress and mitochondrial dysfunction suppress growth of endometriotic lesions. J Cell Physiol. 2019;234(4):4327-41.http://doi.org/10.1002/jcp.27212\u003c/li\u003e\n\u003cli\u003eChoi J, Jo M, Lee E, Lee DY, Choi D. Involvement of endoplasmic reticulum stress in regulation of endometrial stromal cell invasiveness: possible role in pathogenesis of endometriosis. Mol Hum Reprod. 2019;25(3):101-10.http://doi.org/10.1093/molehr/gaz002\u003c/li\u003e\n\u003cli\u003eChoi J, Jo M, Lee E, Lee DY, Choi D. Dienogest regulates apoptosis, proliferation, and invasiveness of endometriotic cyst stromal cells via endoplasmic reticulum stress induction. Mol Hum Reprod. 2020;26(1):30-9.http://doi.org/10.1093/molehr/gaz064\u003c/li\u003e\n\u003cli\u003eKunitomi C, Harada M, Takahashi N, Azhary JMK, Kusamoto A, Nose E, et al. Activation of endoplasmic reticulum stress mediates oxidative stress-induced apoptosis of granulosa cells in ovaries affected by endometrioma. Mol Hum Reprod. 2020;26(1):40-52.http://doi.org/10.1093/molehr/gaz066\u003c/li\u003e\n\u003cli\u003eRevised American Society for Reproductive Medicine classification of endometriosis: 1996. Fertil Steril. 1997;67(5):817-21.http://doi.org/10.1016/s0015-0282(97)81391-x\u003c/li\u003e\n\u003cli\u003eLi X, Jin J, Long X, Weng R, Xiong W, Liang J, et al. METTL3-regulated m6A modification impairs the decidualization of endometrial stromal cells by regulating YTHDF2-mediated degradation of FOXO1 mRNA in endometriosis-related infertility. Reprod Biol Endocrinol. 2023;21(1):99.http://doi.org/10.1186/s12958-023-01151-0\u003c/li\u003e\n\u003cli\u003eSweet LA, Kuss-Duerkop SK, Keestra-Gounder AM. IRE1\u0026alpha;-Driven Inflammation Promotes Clearance of Citrobacter rodentium Infection. Infect Immun. 2022;90(1):e0048121.http://doi.org/10.1128/iai.00481-21\u003c/li\u003e\n\u003cli\u003eKaser A, Lee AH, Franke A, Glickman JN, Zeissig S, Tilg H, et al. XBP1 links ER stress to intestinal inflammation and confers genetic risk for human inflammatory bowel disease. Cell. 2008;134(5):743-56.http://doi.org/10.1016/j.cell.2008.07.021\u003c/li\u003e\n\u003cli\u003eEkiz-Yilmaz T, Isildar B, Gezer A, Kankaya D, Cansiz-Ersoz C, Kayisli UA, et al. The role of unfolded protein response in the pathogenesis of endometriosis: contribution of peritoneal fluid. Reprod Biomed Online. 2021;42(1):1-15.http://doi.org/10.1016/j.rbmo.2020.09.012\u003c/li\u003e\n\u003cli\u003eSemino C, Semino A, Pietra G, Mingari MC, Barocci S, Venturini PL, et al. Role of major histocompatibility complex class I expression and natural killer-like T cells in the genetic control of endometriosis. Fertil Steril. 1995;64(5):909-16.http://doi.org/10.1016/s0015-0282(16)57901-1\u003c/li\u003e\n\u003cli\u003eAbramiuk M, Bębnowska D, Hrynkiewicz R, Polak PNG, Kotarski J, Roliński J, et al. CLTA-4 Expression is Associated with the Maintenance of Chronic Inflammation in Endometriosis and Infertility. Cells. 2021;10(3).http://doi.org/10.3390/cells10030487\u003c/li\u003e\n\u003cli\u003eLauer ME, Erzurum SC, Mukhopadhyay D, Vasanji A, Drazba J, Wang A, et al. Differentiated murine airway epithelial cells synthesize a leukocyte-adhesive hyaluronan matrix in response to endoplasmic reticulum stress. J Biol Chem. 2008;283(38):26283-96.http://doi.org/10.1074/jbc.M803350200\u003c/li\u003e\n\u003cli\u003eCho MK, Jin JS, Jo Y, Han JH, Shin S, Bae SJ, et al. Frankincense ameliorates endometriosis via inducing apoptosis and reducing adhesion. Integr Med Res. 2023;12(2):100947.http://doi.org/10.1016/j.imr.2023.100947\u003c/li\u003e\n\u003cli\u003eKassan M, Vikram A, Kim YR, Li Q, Kassan A, Patel HH, et al. Sirtuin1 protects endothelial Caveolin-1 expression and preserves endothelial function via suppressing miR-204 and endoplasmic reticulum stress. Sci Rep. 2017;7:42265.http://doi.org/10.1038/srep42265\u003c/li\u003e\n\u003cli\u003eQiaomei Z, Ping W, Yanjing Z, Jinhua W, Shaozhan C, Lihong C. Features of peritoneal dendritic cells in the development of endometriosis. Reprod Biol Endocrinol. 2023;21(1):4.http://doi.org/10.1186/s12958-023-01058-w\u003c/li\u003e\n\u003cli\u003eSchulke L, Berbic M, Manconi F, Tokushige N, Markham R, Fraser IS. Dendritic cell populations in the eutopic and ectopic endometrium of women with endometriosis. Hum Reprod. 2009;24(7):1695-703.http://doi.org/10.1093/humrep/dep071\u003c/li\u003e\n\u003cli\u003eRandi AM, Smith KE, Castaman G. von Willebrand factor regulation of blood vessel formation. Blood. 2018;132(2):132-40.http://doi.org/10.1182/blood-2018-01-769018\u003c/li\u003e\n\u003cli\u003eRandi AM, Laffan MA. Von Willebrand factor and angiogenesis: basic and applied issues. J Thromb Haemost. 2017;15(1):13-20.http://doi.org/10.1111/jth.13551\u003c/li\u003e\n\u003cli\u003eLi Y, Liu H, Ye S, Zhang B, Li X, Yuan J, et al. The effects of coagulation factors on the risk of endometriosis: a Mendelian randomization study. BMC Med. 2023;21(1):195.http://doi.org/10.1186/s12916-023-02881-z\u003c/li\u003e\n\u003cli\u003eKhan KN, Kitajima M, Hiraki K, Fujishita A, Sekine I, Ishimaru T, et al. Changes in tissue inflammation, angiogenesis and apoptosis in endometriosis, adenomyosis and uterine myoma after GnRH agonist therapy. Hum Reprod. 2010;25(3):642-53.http://doi.org/10.1093/humrep/dep437\u003c/li\u003e\n\u003cli\u003eKhan KN, Masuzaki H, Fujishita A, Kitajima M, Sekine I, Ishimaru T. Immunoexpression of hepatocyte growth factor and c-Met receptor in the eutopic endometrium predicts the activity of ectopic endometrium. Fertil Steril. 2003;79(1):173-81.http://doi.org/10.1016/s0015-0282(02)04535-1\u003c/li\u003e\n\u003cli\u003eKong DH, Kim YK, Kim MR, Jang JH, Lee S. Emerging Roles of Vascular Cell Adhesion Molecule-1 (VCAM-1) in Immunological Disorders and Cancer. Int J Mol Sci. 2018;19(4).http://doi.org/10.3390/ijms19041057\u003c/li\u003e\n\u003cli\u003eMcCallion A, Nasirzadeh Y, Lingegowda H, Miller JE, Khalaj K, Ahn S, et al. Estrogen mediates inflammatory role of mast cells in endometriosis pathophysiology. Front Immunol. 2022;13:961599.http://doi.org/10.3389/fimmu.2022.961599\u003c/li\u003e\n\u003cli\u003eZhang J, Li H, Yi D, Lai C, Wang H, Zou W, et al. Knockdown of vascular cell adhesion molecule 1 impedes transforming growth factor beta 1-mediated proliferation, migration, and invasion of endometriotic cyst stromal cells. Reprod Biol Endocrinol. 2019;17(1):69.http://doi.org/10.1186/s12958-019-0512-9\u003c/li\u003e\n\u003cli\u003eLyu D, Tang N, Wang J, Zhang Y, Chang J, Liu Z, et al. TGR5 agonist INT-777 mitigates inflammatory response in human endometriotic stromal cells: A therapeutic implication for endometriosis. Int Immunopharmacol. 2019;71:93-9.http://doi.org/10.1016/j.intimp.2019.02.044\u003c/li\u003e\n\u003cli\u003eBhurke A, Kannan A, Neff A, Ma Q, Laws MJ, Taylor RN, et al. A hypoxia-induced Rab pathway regulates embryo implantation by controlled trafficking of secretory granules. Proc Natl Acad Sci U S A. 2020;117(25):14532-42.http://doi.org/10.1073/pnas.2000810117\u003c/li\u003e\n\u003cli\u003eMatsumoto L, Hirota Y, Saito-Fujita T, Takeda N, Tanaka T, Hiraoka T, et al. HIF2\u0026alpha; in the uterine stroma permits embryo invasion and luminal epithelium detachment. J Clin Invest. 2018;128(7):3186-97.http://doi.org/10.1172/jci98931\u003c/li\u003e\n\u003cli\u003ePonandai-Srinivasan S, Andersson KL, Nister M, Saare M, Hassan HA, Varghese SJ, et al. Aberrant expression of genes associated with stemness and cancer in endometria and endometrioma in a subset of women with endometriosis. Hum Reprod. 2018;33(10):1924-38.http://doi.org/10.1093/humrep/dey241\u003c/li\u003e\n\u003cli\u003eSchwager K, Bootz F, Imesch P, Kaspar M, Trachsel E, Neri D. The antibody-mediated targeted delivery of interleukin-10 inhibits endometriosis in a syngeneic mouse model. Hum Reprod. 2011;26(9):2344-52.http://doi.org/10.1093/humrep/der195\u003c/li\u003e\n\u003cli\u003eQuattrone F, Sanchez AM, Pannese M, Hemmerle T, Vigan\u0026ograve; P, Candiani M, et al. The Targeted Delivery of Interleukin 4 Inhibits Development of Endometriotic Lesions in a Mouse Model. Reprod Sci. 2015;22(9):1143-52.http://doi.org/10.1177/1933719115578930\u003c/li\u003e\n\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":"Endometriosis, Endoplasmic Reticulum Stress, Bioinformatic analysis, immune infiltration, stromal cells","lastPublishedDoi":"10.21203/rs.3.rs-4212798/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4212798/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003eNumerous investigations have demonstrated the implication of endoplasmic reticulum stress (ERS) in the etiology of endometriosis. Employing bioinformatics methodologies, we conducted an analysis to ascertain the participation of genes associated with endoplasmic reticulum stress in endometriosis disease subtyping and immune infiltration, with the aim of constructing a diagnostic model for the disease.\u003c/p\u003e\u003ch2\u003eMaterials and Methods\u003c/h2\u003e \u003cp\u003eDifferential expression analysis, weighted gene co-expression network analysis (WGCNA), protein-protein interaction (PPI) network construction, and three machine learning algorithms were employed to identify hub genes associated with endoplasmic reticulum stress in endometriosis. Unsupervised cluster analysis was conducted to identify the ERS cluster. The ERS score and immune infiltration score were computed for distinct clusters using the CIBERSORT algorithm. Functional and pathway enrichment analysis was conducted based on the differential expression profiles of genes within the clusters to elucidate their potential biological functions. The differential expression profiles of genes within the clusters were submitted to the Connectivity Map database to identify candidate therapeutic compounds. A diagnostic model was developed utilizing hub genes, and its predictive performance for endometriosis was assessed. Endometrial tissue specimens obtained from patients were subjected to RT-qPCR and immunohistochemistry (IHC) analyses to evaluate the mRNA and protein expression levels of the hub genes.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eVon Willebrand factor (VWF), vascular cell adhesion molecule 1 (VCAM1), endothelial PAS domain protein 1 (EPAS1), and coagulation factor VIII (F8) were identified as the ERS-related hub genes in endometriosis. Unsupervised consensus clustering analysis revealed the presence of two stable clusters. Cluster B exhibited significantly higher immune scores compared to cluster A, thereby characterizing cluster B as an immune-enriched cluster and cluster A as a less immune-enriched cluster. Functional enrichment analysis revealed that the differentially expressed genes across the clusters predominantly participated in processes related to cell adhesion and regulation of immune cell activation. Decision curves, clinical impact curves, and calibration curves collectively underscored the robust diagnostic utility of the endometriosis diagnostic model derived from four hub genes. In cluster A, certain adrenergic receptor antagonists, progesterone or progesterone receptor agonists, androgen receptor modulators, and NF-κB pathway inhibitors exhibit promising therapeutic prospects. In contrast, cluster B presents potential therapeutic benefits with certain PKC activators, PPAR receptor agonists, insulin sensitizers, adenylate cyclase activators, and caspase activators. Moreover, the findings obtained from RT-qPCR and IHC assays corroborated the outcomes of the bioinformatic analysis, demonstrating elevated expression levels of both mRNA and protein of endoplasmic reticulum stress (ERS) hub genes in endometriosis tissues.\u003c/p\u003e","manuscriptTitle":"Bioinformatic analysis reveals endoplasmic reticulum stress-related molecular cluster and immune characterization in endometriosis：implications for disease subtyping and therapeutic strategies","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-04-12 19:13:28","doi":"10.21203/rs.3.rs-4212798/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":"7a3f1058-0e9c-46f7-8b13-3b8e6f66b247","owner":[],"postedDate":"April 12th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2024-04-12T19:13:32+00:00","versionOfRecord":[],"versionCreatedAt":"2024-04-12 19:13:28","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-4212798","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4212798","identity":"rs-4212798","version":["v1"]},"buildId":"2u56kwukJI3zHK-uzyFNs","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}