DLK1 as a Potential Biomarker and shows NOTCH signaling could be the potential target for Endometriosis: A Machine Learning Approach

DLK1 as a Potential Biomarker and shows NOTCH signaling could be the potential target for Endometriosis: A Machine Learning Approach | 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 Article DLK1 as a Potential Biomarker and shows NOTCH signaling could be the potential target for Endometriosis: A Machine Learning Approach Liting Liao, Zhijian Pan This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-3990509/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 Purpose: The objective of this research is to pinpoint potential diagnostic markers for endometriosis and explore the immune infiltration patterns linked with this condition through the utilization of machine learning techniques. Methods: A total of five gene expression datasets (GSE7305, GSE7307, GSE25628, GSE23339, and GSE120103) were examined in order to identify differentially expressed genes (DEGs) that distinguish normal patients from those with endometriosis. The algorithms Random Forest and Lasso regression were utilised to identify diagnostic biomarkers. GSEA and Go&KEGG database were utilised to determine the potential pathway in which the biomarker was implicated. With the ailment. Furthermore, an assessment of immune cell infiltration in endometriosis tissues relative to normal tissues was conducted using CIBERSORT analysis. In order to investigate the relationship between diagnostic markers and immune cell populations, a correlation analysis was performed. Results: DLK1 (Delta-like 1 homolog) has emerged as a potential diagnostic biomarker for endometriosis, with indications suggesting that Notch signalling could be pivotal in the development of endometriosis. Conclusion: DLK1 emerges as a promising diagnostic biomarker for endometriosis, as our study indicates a complex interplay between immune dysregulation and disease pathogenesis. Notably, our findings elucidate that DLK1 regulates endometriosis through Notch signaling, highlighting the potential of Notch signalling as a therapeutic target for future interventions. Biological sciences/Computational biology and bioinformatics Biological sciences/Computational biology and bioinformatics/Databases Biological sciences/Computational biology and bioinformatics/Machine learning Endometriosis Machine learning Notch signalling DLK1 biomarker Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 What does this study add to the clinical work: This research uses a machine learning approach to identify the biomarker of endometriosis, which reveals that NOTCH signalling might be a novel target for the endometriosis treatment. Take home Message: DLK1 is a potential biomarker for endometriosis and it causes the disease by the regulation of NOTCH signalling. Introduction Endometriosis, a chronic inflammatory disease, impacts an estimated 10% of women in their reproductive years. World Health Organisation data links endometriosis to pelvic pain and infertility.( 1 ). This condition is distinguished by the existence of endometrial-like tissue beyond the confines of the uterus, predominantly in the pelvic peritoneum, ovaries, and rectovaginal septum. In addition to inducing pelvic pain, infertility, and diminished quality of life, it imposes a substantial financial strain on healthcare systems( 2 ). Endometriosis has a complex and multifactorial pathogenesis, which is influenced by environmental, hormonal, genetic, and immunological factors. However, the precise mechanisms that govern the initiation and advancement of endometriosis are still not fully comprehended, and the existing diagnostic and therapeutic alternatives are constrained in scope and frequently prove to be ineffectual.( 3 ). One of the hallmarks of endometriosis is the dysregulation of the immune system, which contributes to the establishment, maintenance, and growth of ectopic endometrial lesions. Several studies have shown that endometriosis is associated with altered immune cell populations, cytokine profiles, and inflammatory responses in the peritoneal cavity and systemic circulation( 4 ). Moreover, endometriosis lesions exhibit features of immune evasion and resistance to apoptosis, similar to cancer cells. Therefore, immunotherapy, which aims to modulate the immune system to eliminate or control disease, has emerged as a promising novel strategy for the treatment of endometriosis. Hormone therapy is developed based on it, which aims to suppress the grow of endometrial tissue by reducing or blocking the production of estrogen. Hormone therapy may include birth control pills, progestogens, gonadotropin-releasing hormone (GnRH) agonists or antagonists, or aromatase inhibitors( 5 ). Hormone therapy may help to reduce pain and bleeding associated with endometriosis, but it may also cause side effects such as weight gain, mood changes, bone loss, or menopausal symptoms, which shows the importance of immunotherapy in endometriosis( 6 ). In our study, we utilized GSE7305, GSE7307 and GSE25628 from GEO to identify DEGs between endometriosis and normal patients. Lasso regression and Random Forest are used to screen diagnostic biomarker for endometriosis. Furthermore, GSE23339 and GSE120103 is used for the validation set. Following this, we analysed the difference in immune infiltration between endometriosis tissues and normal tissues in 22 immune cell subsets using CIBERSORT. To help comprehend the molecular immune mechanisms involved in the progression of endometriosis, we investigate the correlation between infiltrating immune cells and diagnostic markers. we present the results of our investigation into the differential gene expression patterns associated with endometriosis. By integrating and analyzing gene expression data from multiple datasets, we seek to identify potential biomarkers, unravel the biological processes at play, and shed light on the immunological and molecular intricacies of this enigmatic condition. Furthermore, we explore the clinical implications of our findings, particularly in the context of diagnosis and the development of novel therapeutic strategies. Method Data retrieved We used the GEOPARSE package of Python software to download the endometriosis expression profile datasets GSE7305, GSE7307 and GSE25628. GSE7305, GSE7307 and GSE25628 are merged into the single file. GSE23339 and GSE120103 as validating set. There are 74 samples in total, with 38 control group and 36 of disease group. Then we use Scanpy to load and analysis the gene expression file. After concatenation, combat package is used to remove the batch effect of different datasets. PCA plot is shown to visualise the effects of batch correction. DEGs were screened by the pydeseq2 packages, Degs logFC > 1.5 and p < 0.05 are considered as statistically significant. GO&KEGG analysis To elucidate gene functionalities, Gene Ontology (GO) ( http://geneontology.org/ ) is used, highlighting biological implications across three dimensions: Biological Processes (BP), Cellular Component (CC), and Molecular Function (MF). Analytical proceedings were facilitated using the “clusterProfiler” package( 7 ), adhering to a significance threshold defined by a P value of < 0.05. To delve deeper into the mechanistic intricacies underpinning DEGs, Gene Set Enrichment Analysis (GSEA) was harnessed. This aimed to unearth potential differential characteristics between endometriosis-afflicted samples and their healthy counterparts( 8 ). For this purpose, the reference gene set “c2.cp.kegg.v7.4.symbols.gmt” was procured from the Molecular Signatures Database (MSigDB). An adjusted P value of < 0.05 coupled with a False Discovery Rate. Model development Feature selection was conducted using Lasso regression and Random Forest on the DEGs that were previously identified in order to further eliminate the diagnostic markers for endometriosis. The model development was performed utilising the scikit-ilearn package module. Intersection genes between two machine learning models were eliminated and retained for subsequent analysis. It was determined that a two-sided p < 0.05 was statistically significant. Analysis of the receiver operating characteristic (ROC) curve To assess the predictive accuracy of the identified biomarkers, we utilised the "pROC" package, which generated ROC curves based on the candidate genes. We then utilised the area under the ROC curve (AUC) to compare the predictive value of the candidate genes. Relationship between identified genes and infiltrating immune cells in endometriosis. Utilising the CIBERSORT platform ( https://cibersortx.stanford.edu/ ), the degree of immune cell infiltration in samples of control and endometrioid tissue was determined. In order to analyse the interrelationships of infiltrated immune cells and visually interpret these results, the R packages "vioplot" and "corrplot" were utilised. Moreover, the correlation between potential gene markers and the levels of immune cell infiltration was determined through visualisation utilising the "ggpubr" package in R. Result In this study, we integrated three datasets into a single dataset as the input. Figure 1 illustrates the batch differences between each dataset through PCA plots. By applying the Combat function to mitigate batch effects, Figure 1B demonstrates the successful removal of the batch effect. After preprocessing the datasets, we utilized the Pydeseq2 package to identify differentially expressed genes (DEGs) based on the criteria of adjusted p 1.5. The volcano plot in Figure 2A reveals a total of 335 DEGs (161 downregulated and 174 upregulated). As our primary interest lies in discovering biomarkers for endometriosis, we focused on the upregulated DEGs. The heatmap in Figure 2B displays the top 20 upregulated DEGs in the endometriosis group compared to the normal group. After the DEGs had been identified, our subsequent aim was to clarify the fundamental pathways that are linked to the disease. GO and KEGG analyses were conducted on the DEGs. The results for three categories of GO terms are illustrated in Figure 3A. We determined that biological processes were predominantly enriched in the development of the urogenital system, reproductive system, and reproductive structure, using a significance level of p < 0.05. The spindle, endoplasmic reticulum lumen, and collagen-containing extracellular matrix were found to be associated with cellular components. Extracellular matrix structural components, peptidase inhibitor activity, and peptidase regulator activity exhibited an enrichment of molecular functions. Malaria, complement and coagulation cascades, and cell adhesion molecules were identified as being enriched by KEGG analysis. Pathways that were enriched, including cell cycle checkpoints and cell cycle mitotic, were identified using GSEA with a significance level of P < 0.05. We employed Lasso regression and Random Forest, two machine learning algorithms, to create a more concise model. Lasso regression was chosen due to the presence of more features than outcomes, while Random Forest was selected for its lower risk of overfitting compared to other models. To identify the most appropriate biomarkers for endometriosis, we applied both Lasso regression and Random Forest and identified overlapping genes. Figures 4A and 4B display the importance rankings generated by these models, while Figure 4C demonstrates the intersection between the two models. Subsequently, Random Forest identified 26 key DEGs, and Lasso regression identified 10 key DEGs as potential biomarkers, with C3 and DLK1 being common to both. We assessed the sensitivity and specificity of the intersection genes for identifying endometriosis using the Area Under the Curve (AUC). As shown in Figure 5A, both C3 and DLK1 exhibited AUC values greater than 0.7, emphasizing their diagnostic capability. Figure 5B demonstrates the validation of C3 and DLK1 in an external dataset, indicating significantly higher expression in the endometriosis group compared to the normal group in both datasets, consistent with the testing group. Since C3 has been reported as a biomarker for endometriosis, we decide to focus on DLK1. To gain further insights into the role of DLK1 in endometriosis, we examined DLK1-associated genes using GO and KEGG analysis as well as GSEA. The results in Figure 6A revealed that biological processes were primarily related to complement activation and the regulation of humoral immune response. Cellular components were linked to the collagen-containing extracellular matrix, blood microparticles, and endoplasmic reticulum lumen. Molecular functions were enriched in peptidase inhibitor activity, extracellular matrix structural constituents, and endopeptidase inhibitor activity. KEGG analysis showed enrichment in cell adhesion molecules, complement and coagulation cascades, Malaria, and Pertussis. GSEA results indicated enriched pathways such as cell cycle checkpoints and the PLK1 pathway, with a cutoff value of P < 0.05. In order to dive deeper into the potential roles of DLK1 and its influence on endometriosis, we categorised the samples into two groups according to DLK1 expression levels: the high DLK1 expression group (DLK1 expression is greater than the median) and the low DLK1 expression group (DLK1 expression is less than the median). As illustrated in Figure 7A, we estimated the extent of immune cell infiltration utilising Cibersort. Following this, noteworthy immune cell types were identified, and Spearman correlation analyses were performed to ascertain the relationship between DLK1 and these cells, as depicted in Figure 7B. The findings indicated that the strongest correlations were found between quiescent T cell CD3 memory and macrophage M2 and NK activated cells, whereas negative correlations were observed with follicular helper T cells and T cells. Discussion Endometriosis, characterized by the presence of endometrial-like tissue beyond the uterine cavity, represents a chronic and inflammatory gynaecological ailment. Research indicates that diagnosing endometriosis often faces delays, and the current diagnostic techniques, in addition to being invasive, may lack reliability. Therefore, there is a pressing need to identify biomarkers capable of enhancing prognostic accuracy. Furthermore, the pathogenesis of endometriosis remains enigmatic, although existing research has linked its occurrence to genetic, immune, and endocrine factors, comprehensive comprehension of the pathogenic and molecular mechanisms underlying eutopic endometrial damage remains limited. Investigating the mechanisms involved in biomarker identification and their interactions with immune cells promises to significantly advance our understanding of the disease's pathophysiology. Therefore, the identification of biomarkers is crucial in order to transform non-invasive and early diagnostic methods. In this study, we merged three datasets into one and employed PCA to visualize batch differences, followed by batch effect correction using the Combat function. Subsequently, we identified differentially expressed genes (DEGs), focusing on those upregulated in endometriosis. DEGs are associated with crucial biological processes, cellular components, and molecular functions, as determined by functional enrichment analysis. Machine learning technique were used to identify potential biomarkers, with C3 and DLK1 emerging as promising candidates. These biomarkers were validated using external datasets, and their diagnostic capabilities were assessed through AUC analysis. Since a recent publication has confirmed the prognostic value of C3, we only explored the functional roles of DLK1 through GO, KEGG analysis, and GSEA. Finally, we investigated immune cell infiltration and correlations with DLK1 expression, highlighting specific immune cell types associated with endometriosis. DLK1, a transmembrane protein related to Notch, has been implicated in several articles as a growth factor that sustains the proliferative state of undifferentiated cells. However, its expression tends to decrease during the differentiation of immune cells( 9 ). Most of the DLK1 related articles focus on the association with stem cells and terminal differentiation( 10 , 11 ). The majority of evidence indicates that DLK1 can participate in the development process by inhibiting Notch signalling; this is consistent with the stimulation of growth and inhibition of differentiation observed in various organs( 12 ). The Notch signalling pathway is essential for somatic stem cell maintenance and tissue homeostasis. The activation of ligand receptor interaction induces the transcription of Notch target genes, including HEY1, HEY2, and HES19. There is a scarcity of literature utilising DLK1 as a biomarker; only one publication suggests that serum DLK1 levels could potentially function as prognostic biomarkers for patients with HCC( 13 ), and another article evidences that DLK1 is linked to obesity and insulin resistance( 9 ). In endometriosis field, 1 publication has reported that it is upregulated in Ectopic endometrium than eutopic endometrium( 14 ), however they did not have systematically go thought of the role that DLK1 play in the endometriosis. Therefore, the relationship between DLK1 and endometriosis is still confounded, and the role that DLK1 plays in endometriosis is still yet to be determined. In our research, we observed that DLK1 has a reasonable AUC in predicting the endometriosis compared to normal patients. According to the findings, DLK1 concentration might function as a prognostic biomarker for patients with endometriosis. Previous reported have shown that the key role that DLK1 play in Notch signalling, which means that DLK1 could be a potential therapeutical target in order to treat endometriosis. A lot of articles have investigated the relationship between Notch signalling and endometriosis, with the conservative view that Notch signalling is associated with endometriosis( 15 , 16 ). Endometrial dysregulation of molecules associated with Notch signalling, including Notch1, DLL1, and JAG1, has been identified in infertile women( 17 ). Since DLK1 serves as a negative regulator of notch signalling, it might cause the decreased signalling, induce the endometriosis, impaired decidualization and eventually cause infertility. Investigating the molecular mechanism underlying the association between DLK1 and endometriosis will contribute to the advancement of targeted therapeutic strategies. Patients were categorised into high and low DLK1 subgroups according to the median value of DLK1 expression in order to construct a differential gene-based functional network. The differential genes associated with DLK1 expression are involved in a vast array of biological processes, including structural components of the regulation of humoral immune responses, which have been identified as a pathophysiological factor in endometriosis, according to GO and KEGG pathway analyses. Furthermore, its molecular function inhibitory activity may indicate that DLK1 inhibits the Notch signalling pathway. Two 'Cell Cycle Checkpoint' and 'PLK1 pathways' are associated with DLK1 overexpression, which is reported to be associated with Notch signalling, according to the GSEA result. PLK1 may be one of the kinases implicated in NOTCH1 regulation, which is comparable to the function of DLK1, according to one study. As PLK1 activity is both required for mitotic entry and inhibited in the presence of DNA damage, it controls the expression of NOTCH1 during the G2/M transition. When cells in the G2 phase are exposed to DNA damage, however, PLK1 is inhibited to prevent mitotic entry( 18 ). Furthermore, it has been established that Notch signalling initiates the cell cycle, as it facilitates cell-to-cell communication and can convert membrane-based receptor activation into changes in gene expression. This supports the hypothesis that DLK1 has been significantly involved in endometriosis via Notch signalling. The signalling pathways and processes are linked to the pathogenesis of endometriosis. However, additional experiments are required to validate the mechanisms by which DLK1 regulates these pathways in endometriosis. The development of endometriosis is associated with immune cell infiltration and immune dysfunction in the immune microenvironment; nevertheless, there is a lack of information regarding the correlation between immune infiltration and DLK1 expression. By assessing the level of immune cell infiltration in endometriosis using the Cibersort technique and a correlation threshold exceeding |0.3|, our findings revealed that there was an elevation in the infiltration of Macrophage M2 and T cell CD4 memory resting (CD4), while the infiltration of T cell follicular helper (Tfh) and NK cells was reduced. Accumulation of M2 macrophages during adenomyosis increases the capacity of adenomyotic and healthy endometrial cells to invade, suggesting that infiltration of macrophages may be adequate to promote the disease( 19 ). Furthermore, T cells contribute to the advancement of the disease through the secretion of numerous cytokines that modulate the operations of other immune cells, facilitate ectopic implantation and endometrial cell proliferation, and stimulate angiogenesis( 20 ). On the other hand, endometriosis development is associated with aberrant expression of NK cell receptors and diminished cytotoxicity of NK cells, the primary immune system defence mechanism. These findings are consistent with prior research( 21 ), the under expressed of Tfh also consistent with the previous publication( 22 ). Our research consists predominantly of an examination of publicly accessible datasets that have already been analysed; clinical data and experimental validation are absent. Hence, although our results indicate a possible correlation between DLK1 and endometriosis, extensive prospective clinical studies are necessary to validate the association and determine the diagnostic and prognostic significance of DLK1. In addition, our investigation into the pathogenesis of endometriosis is predominately based on bioinformatics analyses of DLK1. In order to attain a more comprehensive comprehension of the complex molecular mechanisms that govern the participation of DLK1 in the pathogenesis and advancement of endometriosis, it is imperative to undertake additional in vitro and in vivo investigations. In this study, we made a discovery revealing that DLK1 is significantly upregulated in endometriosis and is intricately associated with immune cell infiltration, particularly T cells, B cells, and NK cells. Furthermore, our findings indicate that DLK1 regulates endometriosis through the Notch signalling pathway, suggesting that Notch signalling may play a pivotal role in the development of this condition. Overall, our research contributes to the growing body of knowledge surrounding the treatment of endometriosis by identifying Notch signalling as a potential therapeutic target for future interventions. Declarations Author Contribution First author Liting Liao is responsible for the data analysis and manuscript writing. Corresponding author Zhijian Pan is responsible for the research planning and providing suggestions and project development. References Parasar P, Ozcan P, Terry KL. Endometriosis: Epidemiology, Diagnosis and Clinical Management. Curr Obstet Gynecol Rep. 2017;6(1):34–41. Burney RO, Giudice LC. Pathogenesis and pathophysiology of endometriosis. Fertil Steril. 2012;98(3):511–9. Lamceva J, Uljanovs R, Strumfa I. The Main Theories on the Pathogenesis of Endometriosis. Int J Mol Sci. 2023;24(5). Chen S, Liu Y, Zhong Z, Wei C, Liu Y, Zhu X. Peritoneal immune microenvironment of endometriosis: Role and therapeutic perspectives. Front Immunol. 2023;14:1134663. Gheorghisan-Galateanu AA, Gheorghiu ML. HORMONAL THERAPY IN WOMEN OF REPRODUCTIVE AGE WITH ENDOMETRIOSIS: AN UPDATE. Acta Endocrinol (Buchar). 2019;15(2):276–81. Vannuccini S, Clemenza S, Rossi M, Petraglia F. Hormonal treatments for endometriosis: The endocrine background. Rev Endocr Metab Disord. 2022;23(3):333–55. Yu G, Wang LG, Han Y, He QY. clusterProfiler: an R package for comparing biological themes among gene clusters. Omics. 2012;16(5):284–7. Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005;102(43):15545–50. Yevtodiyenko A, Schmidt JV. Dlk1 expression marks developing endothelium and sites of branching morphogenesis in the mouse embryo and placenta. Dev Dyn. 2006;235(4):1115–23. Traustadóttir G, Lagoni LV, Ankerstjerne LBS, Bisgaard HC, Jensen CH, Andersen DC. The imprinted gene Delta like non-canonical Notch ligand 1 (Dlk1) is conserved in mammals, and serves a growth modulatory role during tissue development and regeneration through Notch dependent and independent mechanisms. Cytokine Growth Factor Rev. 2019;46:17–27. Grassi ES, Pietras A. Emerging Roles of DLK1 in the Stem Cell Niche and Cancer Stemness. J Histochem Cytochem. 2022;70(1):17–28. Falix FA, Aronson DC, Lamers WH, Gaemers IC. Possible roles of DLK1 in the Notch pathway during development and disease. Biochimica et Biophysica Acta (BBA) - Molecular Basis of Disease. 2012;1822(6):988–95. Li H, Cui ML, Chen TY, Xie HY, Cui Y, Tu H, et al. Serum DLK1 is a potential prognostic biomarker in patients with hepatocellular carcinoma. Tumour Biol. 2015;36(11):8399–404. Cui D, Liu Y, Ma J, Lin K, Xu K, Lin J. Identification of key genes and pathways in endometriosis by integrated expression profiles analysis. PeerJ. 2020;8:e10171. Su RW, Strug MR, Joshi NR, Jeong JW, Miele L, Lessey BA, et al. Decreased Notch pathway signaling in the endometrium of women with endometriosis impairs decidualization. J Clin Endocrinol Metab. 2015;100(3):E433-42. Kopan R, Ilagan MX. The canonical Notch signaling pathway: unfolding the activation mechanism. Cell. 2009;137(2):216–33. Afshar Y, Jeong JW, Roqueiro D, DeMayo F, Lydon J, Radtke F, et al. Notch1 mediates uterine stromal differentiation and is critical for complete decidualization in the mouse. Faseb j. 2012;26(1):282–94. De Blasio C, Zonfrilli A, Franchitto M, Mariano G, Cialfi S, Verma N, et al. PLK1 targets NOTCH1 during DNA damage and mitotic progression. J Biol Chem. 2019;294(47):17941–50. Stratopoulou CA, Cussac S, d'Argent M, Donnez J, Dolmans MM. M2 macrophages enhance endometrial cell invasiveness by promoting collective cell migration in uterine adenomyosis. Reprod Biomed Online. 2023;46(4):729–38. Khan KN, Yamamoto K, Fujishita A, Muto H, Koshiba A, Kuroboshi H, et al. Differential Levels of Regulatory T Cells and T-Helper-17 Cells in Women With Early and Advanced Endometriosis. J Clin Endocrinol Metab. 2019;104(10):4715–29. Freitag N, Pour SJ, Fehm TN, Toth B, Markert UR, Weber M, et al. Are uterine natural killer and plasma cells in infertility patients associated with endometriosis, repeated implantation failure, or recurrent pregnancy loss? Arch Gynecol Obstet. 2020;302(6):1487–94. Geng R, Huang X, Li L, Guo X, Wang Q, Zheng Y, et al. Gene expression analysis in endometriosis: Immunopathology insights, transcription factors and therapeutic targets. Front Immunol. 2022;13:1037504. Additional Declarations No competing interests reported. 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-3990509","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":276686422,"identity":"5c1b1990-a2a1-411b-b495-e3abea7668cd","order_by":0,"name":"Liting Liao","email":"","orcid":"","institution":"Qinzhou First People's Hospital","correspondingAuthor":false,"prefix":"","firstName":"Liting","middleName":"","lastName":"Liao","suffix":""},{"id":276686423,"identity":"60d880b3-c741-473c-b498-3a9c27be285b","order_by":1,"name":"Zhijian Pan","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABFUlEQVRIie2PMUvEMBTHEwOdUrJeEfQr9AgcfheXHAWnxsWlw3kGhLul2LUFPb9CXTqnBOpS8QO4VApON9yY7UxFuOnSVTC/4cF7/H+P9wBwOP4gUAAEgN7fEZJ9diwxI4TEuAJTxIK8ReGuHRRoV4YIAB5ioYy9oFj9rrHm16qnGHvXgXhT1N/cXpK1UXRSHT8svaIRvsA3BD3Me7965bmCAqbth+WXGCmMJ7AQklK/argwCoIri5JtjeKFsJRsduo/Nvx5VMljFGGPzUsZz4JCLHg5rnzR6VMqaZA3UbhrJH8xSm37ZZpF/WSr5Rkh93XHFku+eVd1pxOLIgA4wYde/VR5NG84H47Th35pCzscDsc/5RufiWF2pBPDCwAAAABJRU5ErkJggg==","orcid":"","institution":"Qinzhou First People's Hospital","correspondingAuthor":true,"prefix":"","firstName":"Zhijian","middleName":"","lastName":"Pan","suffix":""}],"badges":[],"createdAt":"2024-02-26 09:07:00","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-3990509/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-3990509/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":52191734,"identity":"e52f21dc-5de1-4e9d-9925-fd0d3fbc95c7","added_by":"auto","created_at":"2024-03-07 19:30:49","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":99982,"visible":true,"origin":"","legend":"\u003cp\u003eElimination of batch effects using the Combat function. (A) Three datasets before normalization. (B) After normalization and batch correction of the three datasets.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-3990509/v1/a9a287a00f7e5473fd227ad2.png"},{"id":52191735,"identity":"9a79d311-5ca6-4eb6-9b54-d325ee2ce507","added_by":"auto","created_at":"2024-03-07 19:30:49","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":273787,"visible":true,"origin":"","legend":"\u003cp\u003eDifferentially expressed genes (DEGs) from the merged datasets. (A) Volcano plot of downregulated and upregulated DEGs. (B) Heatmap showing the top 20 upregulated DEGs in endometriosis tissue compared to control samples.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-3990509/v1/d1313e569215444be92a3dab.png"},{"id":52191731,"identity":"b629b8f4-4cd3-44f4-a673-3889d059aa6f","added_by":"auto","created_at":"2024-03-07 19:30:48","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":211763,"visible":true,"origin":"","legend":"\u003cp\u003eFunctional annotation of Differentially Expressed Genes (DEGs). Subfigure (A) illustrates the Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) functional enrichment of the DEGs. Subfigure (B) showcases Gene Set Enrichment Analysis (GSEA) for pathways that are significantly enriched.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-3990509/v1/b9fb8b65f2d23ca3409546fc.png"},{"id":52191733,"identity":"dc1f1602-c597-4fa1-876d-e702f6b09908","added_by":"auto","created_at":"2024-03-07 19:30:48","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":100480,"visible":true,"origin":"","legend":"\u003cp\u003eMachine learning approach for identifying key genes. (A) Random Forest identifies 26 genes ranked by importance. (B) Lasso regression identifies 10 genes ranked by importance. (C) Venn diagram shows overlapping genes identified by Random Forest and Lasso regression.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-3990509/v1/bc117d691969ce486e03bad4.png"},{"id":52191738,"identity":"89980e3c-5cda-4b65-9b76-602f9841b45d","added_by":"auto","created_at":"2024-03-07 19:30:50","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":86697,"visible":true,"origin":"","legend":"\u003cp\u003eDiagnostic marker validation through the utilisation of an external dataset. (A) Assessment of the diagnostic efficacy exhibited by two biomarkers comprising the dataset. (B) The expression of two biomarkers in the validation set is validated.\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-3990509/v1/01c5af0205f3f5bd201ff87f.png"},{"id":52191732,"identity":"534c4f5a-2a9b-4196-9c23-6e19ebde667c","added_by":"auto","created_at":"2024-03-07 19:30:48","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":239169,"visible":true,"origin":"","legend":"\u003cp\u003eFunctional annotation of DEGs associated with DLK1. (A) Functional enrichment of DEGs by KEGG and GO. (B) GSEA for pathways that are significantly enriched.\u003c/p\u003e","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-3990509/v1/8f802fe8673d3d291805a7f5.png"},{"id":52191739,"identity":"7cacfd69-b7b5-4e76-a3f3-500906df6533","added_by":"auto","created_at":"2024-03-07 19:30:50","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":139748,"visible":true,"origin":"","legend":"\u003cp\u003eThe correlation analysis and distribution of immune cell infiltration are depicted. (A) Immune cell composition in patients with endometriosis versus healthy individuals. (B) Examination of the correlation between DLK1 and endometriosis-infiltrating immune cells.\u003c/p\u003e","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-3990509/v1/4fb9fe0ae28f6d0bc4c7ccb9.png"},{"id":63601628,"identity":"94bc006c-1c6f-4759-8ad4-a00a95158e68","added_by":"auto","created_at":"2024-08-30 05:30:17","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1416384,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3990509/v1/97fc9f5b-d0d0-4680-b887-7e9280409a0b.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"DLK1 as a Potential Biomarker and shows NOTCH signaling could be the potential target for Endometriosis: A Machine Learning Approach","fulltext":[{"header":"What does this study add to the clinical work:","content":"\u003cp\u003eThis research uses a machine learning approach to identify the biomarker of endometriosis, which reveals that NOTCH signalling might be a novel target for the endometriosis treatment.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTake home Message:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDLK1 is a potential biomarker for endometriosis and it causes the disease by the regulation of NOTCH signalling.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e"},{"header":"Introduction","content":"\u003cp\u003eEndometriosis, a chronic inflammatory disease, impacts an estimated 10% of women in their reproductive years. World Health Organisation data links endometriosis to pelvic pain and infertility.(\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e). This condition is distinguished by the existence of endometrial-like tissue beyond the confines of the uterus, predominantly in the pelvic peritoneum, ovaries, and rectovaginal septum. In addition to inducing pelvic pain, infertility, and diminished quality of life, it imposes a substantial financial strain on healthcare systems(\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e). Endometriosis has a complex and multifactorial pathogenesis, which is influenced by environmental, hormonal, genetic, and immunological factors. However, the precise mechanisms that govern the initiation and advancement of endometriosis are still not fully comprehended, and the existing diagnostic and therapeutic alternatives are constrained in scope and frequently prove to be ineffectual.(\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eOne of the hallmarks of endometriosis is the dysregulation of the immune system, which contributes to the establishment, maintenance, and growth of ectopic endometrial lesions. Several studies have shown that endometriosis is associated with altered immune cell populations, cytokine profiles, and inflammatory responses in the peritoneal cavity and systemic circulation(\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e). Moreover, endometriosis lesions exhibit features of immune evasion and resistance to apoptosis, similar to cancer cells. Therefore, immunotherapy, which aims to modulate the immune system to eliminate or control disease, has emerged as a promising novel strategy for the treatment of endometriosis. Hormone therapy is developed based on it, which aims to suppress the grow of endometrial tissue by reducing or blocking the production of estrogen. Hormone therapy may include birth control pills, progestogens, gonadotropin-releasing hormone (GnRH) agonists or antagonists, or aromatase inhibitors(\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e). Hormone therapy may help to reduce pain and bleeding associated with endometriosis, but it may also cause side effects such as weight gain, mood changes, bone loss, or menopausal symptoms, which shows the importance of immunotherapy in endometriosis(\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn our study, we utilized GSE7305, GSE7307 and GSE25628 from GEO to identify DEGs between endometriosis and normal patients. Lasso regression and Random Forest are used to screen diagnostic biomarker for endometriosis. Furthermore, GSE23339 and GSE120103 is used for the validation set. Following this, we analysed the difference in immune infiltration between endometriosis tissues and normal tissues in 22 immune cell subsets using CIBERSORT. To help comprehend the molecular immune mechanisms involved in the progression of endometriosis, we investigate the correlation between infiltrating immune cells and diagnostic markers.\u003c/p\u003e \u003cp\u003ewe present the results of our investigation into the differential gene expression patterns associated with endometriosis. By integrating and analyzing gene expression data from multiple datasets, we seek to identify potential biomarkers, unravel the biological processes at play, and shed light on the immunological and molecular intricacies of this enigmatic condition. Furthermore, we explore the clinical implications of our findings, particularly in the context of diagnosis and the development of novel therapeutic strategies.\u003c/p\u003e"},{"header":"Method","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eData retrieved\u003c/h2\u003e \u003cp\u003eWe used the GEOPARSE package of Python software to download the endometriosis expression profile datasets GSE7305, GSE7307 and GSE25628. GSE7305, GSE7307 and GSE25628 are merged into the single file. GSE23339 and GSE120103 as validating set. There are 74 samples in total, with 38 control group and 36 of disease group. Then we use Scanpy to load and analysis the gene expression file. After concatenation, combat package is used to remove the batch effect of different datasets. PCA plot is shown to visualise the effects of batch correction. DEGs were screened by the pydeseq2 packages, Degs logFC\u0026thinsp;\u0026gt;\u0026thinsp;1.5 and p\u0026thinsp;\u0026lt;\u0026thinsp;0.05 are considered as statistically significant.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003eGO\u0026amp;KEGG analysis\u003c/h2\u003e \u003cp\u003eTo elucidate gene functionalities, Gene Ontology (GO) (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://geneontology.org/\u003c/span\u003e\u003cspan address=\"http://geneontology.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) is used, highlighting biological implications across three dimensions: Biological Processes (BP), Cellular Component (CC), and Molecular Function (MF). Analytical proceedings were facilitated using the \u0026ldquo;clusterProfiler\u0026rdquo; package(\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e), adhering to a significance threshold defined by a P value of \u0026lt;\u0026thinsp;0.05.\u003c/p\u003e \u003cp\u003eTo delve deeper into the mechanistic intricacies underpinning DEGs, Gene Set Enrichment Analysis (GSEA) was harnessed. This aimed to unearth potential differential characteristics between endometriosis-afflicted samples and their healthy counterparts(\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e). For this purpose, the reference gene set \u0026ldquo;c2.cp.kegg.v7.4.symbols.gmt\u0026rdquo; was procured from the Molecular Signatures Database (MSigDB). An adjusted P value of \u0026lt;\u0026thinsp;0.05 coupled with a False Discovery Rate.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003eModel development\u003c/h2\u003e \u003cp\u003eFeature selection was conducted using Lasso regression and Random Forest on the DEGs that were previously identified in order to further eliminate the diagnostic markers for endometriosis. The model development was performed utilising the scikit-ilearn package module. Intersection genes between two machine learning models were eliminated and retained for subsequent analysis. It was determined that a two-sided p\u0026thinsp;\u0026lt;\u0026thinsp;0.05 was statistically significant.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003eAnalysis of the receiver operating characteristic (ROC) curve\u003c/h2\u003e \u003cp\u003eTo assess the predictive accuracy of the identified biomarkers, we utilised the \"pROC\" package, which generated ROC curves based on the candidate genes. We then utilised the area under the ROC curve (AUC) to compare the predictive value of the candidate genes.\u003c/p\u003e \u003cp\u003e \u003cb\u003eRelationship between identified genes and infiltrating immune cells in endometriosis.\u003c/b\u003e \u003c/p\u003e \u003cp\u003eUtilising the CIBERSORT platform (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://cibersortx.stanford.edu/\u003c/span\u003e\u003cspan address=\"https://cibersortx.stanford.edu/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), the degree of immune cell infiltration in samples of control and endometrioid tissue was determined. In order to analyse the interrelationships of infiltrated immune cells and visually interpret these results, the R packages \"vioplot\" and \"corrplot\" were utilised. Moreover, the correlation between potential gene markers and the levels of immune cell infiltration was determined through visualisation utilising the \"ggpubr\" package in R.\u003c/p\u003e \u003c/div\u003e"},{"header":"Result","content":"\u003cp\u003eIn this study, we integrated three datasets into a single dataset as the input. Figure 1 illustrates the batch differences between each dataset through PCA plots. By applying the Combat function to mitigate batch effects, Figure 1B demonstrates the successful removal of the batch effect.\u003c/p\u003e\n\u003cp\u003eAfter preprocessing the datasets, we utilized the Pydeseq2 package to identify differentially expressed genes (DEGs) based on the criteria of adjusted p \u0026lt; 0.05 and |log fold change| \u0026gt; 1.5. The volcano plot in Figure 2A reveals a total of 335 DEGs (161 downregulated and 174 upregulated). As our primary interest lies in discovering biomarkers for endometriosis, we focused on the upregulated DEGs. The heatmap in Figure 2B displays the top 20 upregulated DEGs in the endometriosis group compared to the normal group.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAfter the DEGs had been identified, our subsequent aim was to clarify the fundamental pathways that are linked to the disease. GO and KEGG analyses were conducted on the DEGs. The results for three categories of GO terms are illustrated in Figure 3A. We determined that biological processes were predominantly enriched in the development of the urogenital system, reproductive system, and reproductive structure, using a significance level of p \u0026lt; 0.05. The spindle, endoplasmic reticulum lumen, and collagen-containing extracellular matrix were found to be associated with cellular components. Extracellular matrix structural components, peptidase inhibitor activity, and peptidase regulator activity exhibited an enrichment of molecular functions. Malaria, complement and coagulation cascades, and cell adhesion molecules were identified as being enriched by KEGG analysis. Pathways that were enriched, including cell cycle checkpoints and cell cycle mitotic, were identified using GSEA with a significance level of P \u0026lt; 0.05.\u003c/p\u003e\n\u003cp\u003eWe employed Lasso regression and Random Forest, two machine learning algorithms, to create a more concise model. Lasso regression was chosen due to the presence of more features than outcomes, while Random Forest was selected for its lower risk of overfitting compared to other models. To identify the most appropriate biomarkers for endometriosis, we applied both Lasso regression and Random Forest and identified overlapping genes. Figures 4A and 4B display the importance rankings generated by these models, while Figure 4C demonstrates the intersection between the two models. Subsequently, Random Forest identified 26 key DEGs, and Lasso regression identified 10 key DEGs as potential biomarkers, with C3 and DLK1 being common to both.\u003c/p\u003e\n\u003cp\u003eWe assessed the sensitivity and specificity of the intersection genes for identifying endometriosis using the Area Under the Curve (AUC). As shown in Figure 5A, both C3 and DLK1 exhibited AUC values greater than 0.7, emphasizing their diagnostic capability. Figure 5B demonstrates the validation of C3 and DLK1 in an external dataset, indicating significantly higher expression in the endometriosis group compared to the normal group in both datasets, consistent with the testing group.\u003c/p\u003e\n\u003cp\u003eSince C3 has been reported as a biomarker for endometriosis, we decide to focus on DLK1. To gain further insights into the role of DLK1 in endometriosis, we examined DLK1-associated genes using GO and KEGG analysis as well as GSEA. The results in Figure 6A revealed that biological processes were primarily related to complement activation and the regulation of humoral immune response. Cellular components were linked to the collagen-containing extracellular matrix, blood microparticles, and endoplasmic reticulum lumen. Molecular functions were enriched in peptidase inhibitor activity, extracellular matrix structural constituents, and endopeptidase inhibitor activity. KEGG analysis showed enrichment in cell adhesion molecules, complement and coagulation cascades, Malaria, and Pertussis. GSEA results indicated enriched pathways such as cell cycle checkpoints and the PLK1 pathway, with a cutoff value of P \u0026lt; 0.05.\u003c/p\u003e\n\u003cp\u003eIn order to dive deeper into the potential roles of DLK1 and its influence on endometriosis, we categorised the samples into two groups according to DLK1 expression levels: the high DLK1 expression group (DLK1 expression is greater than the median) and the low DLK1 expression group (DLK1 expression is less than the median). As illustrated in Figure 7A, we estimated the extent of immune cell infiltration utilising Cibersort. Following this, noteworthy immune cell types were identified, and Spearman correlation analyses were performed to ascertain the relationship between DLK1 and these cells, as depicted in Figure 7B. The findings indicated that the strongest correlations were found between quiescent T cell CD3 memory and macrophage M2 and NK activated cells, whereas negative correlations were observed with follicular helper T cells and T cells.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eEndometriosis, characterized by the presence of endometrial-like tissue beyond the uterine cavity, represents a chronic and inflammatory gynaecological ailment. Research indicates that diagnosing endometriosis often faces delays, and the current diagnostic techniques, in addition to being invasive, may lack reliability. Therefore, there is a pressing need to identify biomarkers capable of enhancing prognostic accuracy. Furthermore, the pathogenesis of endometriosis remains enigmatic, although existing research has linked its occurrence to genetic, immune, and endocrine factors, comprehensive comprehension of the pathogenic and molecular mechanisms underlying eutopic endometrial damage remains limited. Investigating the mechanisms involved in biomarker identification and their interactions with immune cells promises to significantly advance our understanding of the disease's pathophysiology. Therefore, the identification of biomarkers is crucial in order to transform non-invasive and early diagnostic methods.\u003c/p\u003e \u003cp\u003eIn this study, we merged three datasets into one and employed PCA to visualize batch differences, followed by batch effect correction using the Combat function. Subsequently, we identified differentially expressed genes (DEGs), focusing on those upregulated in endometriosis. DEGs are associated with crucial biological processes, cellular components, and molecular functions, as determined by functional enrichment analysis. Machine learning technique were used to identify potential biomarkers, with C3 and DLK1 emerging as promising candidates. These biomarkers were validated using external datasets, and their diagnostic capabilities were assessed through AUC analysis. Since a recent publication has confirmed the prognostic value of C3, we only explored the functional roles of DLK1 through GO, KEGG analysis, and GSEA. Finally, we investigated immune cell infiltration and correlations with DLK1 expression, highlighting specific immune cell types associated with endometriosis.\u003c/p\u003e \u003cp\u003eDLK1, a transmembrane protein related to Notch, has been implicated in several articles as a growth factor that sustains the proliferative state of undifferentiated cells. However, its expression tends to decrease during the differentiation of immune cells(\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e). Most of the DLK1 related articles focus on the association with stem cells and terminal differentiation(\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e). The majority of evidence indicates that DLK1 can participate in the development process by inhibiting Notch signalling; this is consistent with the stimulation of growth and inhibition of differentiation observed in various organs(\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e). The Notch signalling pathway is essential for somatic stem cell maintenance and tissue homeostasis. The activation of ligand receptor interaction induces the transcription of Notch target genes, including HEY1, HEY2, and HES19. There is a scarcity of literature utilising DLK1 as a biomarker; only one publication suggests that serum DLK1 levels could potentially function as prognostic biomarkers for patients with HCC(\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e), and another article evidences that DLK1 is linked to obesity and insulin resistance(\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e). In endometriosis field, 1 publication has reported that it is upregulated in Ectopic endometrium than eutopic endometrium(\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e), however they did not have systematically go thought of the role that DLK1 play in the endometriosis. Therefore, the relationship between DLK1 and endometriosis is still confounded, and the role that DLK1 plays in endometriosis is still yet to be determined.\u003c/p\u003e \u003cp\u003eIn our research, we observed that DLK1 has a reasonable AUC in predicting the endometriosis compared to normal patients. According to the findings, DLK1 concentration might function as a prognostic biomarker for patients with endometriosis. Previous reported have shown that the key role that DLK1 play in Notch signalling, which means that DLK1 could be a potential therapeutical target in order to treat endometriosis. A lot of articles have investigated the relationship between Notch signalling and endometriosis, with the conservative view that Notch signalling is associated with endometriosis(\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e). Endometrial dysregulation of molecules associated with Notch signalling, including Notch1, DLL1, and JAG1, has been identified in infertile women(\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e). Since DLK1 serves as a negative regulator of notch signalling, it might cause the decreased signalling, induce the endometriosis, impaired decidualization and eventually cause infertility.\u003c/p\u003e \u003cp\u003eInvestigating the molecular mechanism underlying the association between DLK1 and endometriosis will contribute to the advancement of targeted therapeutic strategies. Patients were categorised into high and low DLK1 subgroups according to the median value of DLK1 expression in order to construct a differential gene-based functional network. The differential genes associated with DLK1 expression are involved in a vast array of biological processes, including structural components of the regulation of humoral immune responses, which have been identified as a pathophysiological factor in endometriosis, according to GO and KEGG pathway analyses. Furthermore, its molecular function inhibitory activity may indicate that DLK1 inhibits the Notch signalling pathway.\u003c/p\u003e \u003cp\u003eTwo 'Cell Cycle Checkpoint' and 'PLK1 pathways' are associated with DLK1 overexpression, which is reported to be associated with Notch signalling, according to the GSEA result. PLK1 may be one of the kinases implicated in NOTCH1 regulation, which is comparable to the function of DLK1, according to one study. As PLK1 activity is both required for mitotic entry and inhibited in the presence of DNA damage, it controls the expression of NOTCH1 during the G2/M transition. When cells in the G2 phase are exposed to DNA damage, however, PLK1 is inhibited to prevent mitotic entry(\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e). Furthermore, it has been established that Notch signalling initiates the cell cycle, as it facilitates cell-to-cell communication and can convert membrane-based receptor activation into changes in gene expression. This supports the hypothesis that DLK1 has been significantly involved in endometriosis via Notch signalling. The signalling pathways and processes are linked to the pathogenesis of endometriosis. However, additional experiments are required to validate the mechanisms by which DLK1 regulates these pathways in endometriosis.\u003c/p\u003e \u003cp\u003eThe development of endometriosis is associated with immune cell infiltration and immune dysfunction in the immune microenvironment; nevertheless, there is a lack of information regarding the correlation between immune infiltration and DLK1 expression. By assessing the level of immune cell infiltration in endometriosis using the Cibersort technique and a correlation threshold exceeding |0.3|, our findings revealed that there was an elevation in the infiltration of Macrophage M2 and T cell CD4 memory resting (CD4), while the infiltration of T cell follicular helper (Tfh) and NK cells was reduced. Accumulation of M2 macrophages during adenomyosis increases the capacity of adenomyotic and healthy endometrial cells to invade, suggesting that infiltration of macrophages may be adequate to promote the disease(\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e). Furthermore, T cells contribute to the advancement of the disease through the secretion of numerous cytokines that modulate the operations of other immune cells, facilitate ectopic implantation and endometrial cell proliferation, and stimulate angiogenesis(\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e). On the other hand, endometriosis development is associated with aberrant expression of NK cell receptors and diminished cytotoxicity of NK cells, the primary immune system defence mechanism. These findings are consistent with prior research(\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e), the under expressed of Tfh also consistent with the previous publication(\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eOur research consists predominantly of an examination of publicly accessible datasets that have already been analysed; clinical data and experimental validation are absent. Hence, although our results indicate a possible correlation between DLK1 and endometriosis, extensive prospective clinical studies are necessary to validate the association and determine the diagnostic and prognostic significance of DLK1. In addition, our investigation into the pathogenesis of endometriosis is predominately based on bioinformatics analyses of DLK1. In order to attain a more comprehensive comprehension of the complex molecular mechanisms that govern the participation of DLK1 in the pathogenesis and advancement of endometriosis, it is imperative to undertake additional in vitro and in vivo investigations.\u003c/p\u003e \u003cp\u003eIn this study, we made a discovery revealing that DLK1 is significantly upregulated in endometriosis and is intricately associated with immune cell infiltration, particularly T cells, B cells, and NK cells. Furthermore, our findings indicate that DLK1 regulates endometriosis through the Notch signalling pathway, suggesting that Notch signalling may play a pivotal role in the development of this condition. Overall, our research contributes to the growing body of knowledge surrounding the treatment of endometriosis by identifying Notch signalling as a potential therapeutic target for future interventions.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eFirst author Liting Liao is responsible for the data analysis and manuscript writing. Corresponding author Zhijian Pan is responsible for the research planning and providing suggestions and project development.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eParasar P, Ozcan P, Terry KL. Endometriosis: Epidemiology, Diagnosis and Clinical Management. Curr Obstet Gynecol Rep. 2017;6(1):34\u0026ndash;41.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBurney RO, Giudice LC. Pathogenesis and pathophysiology of endometriosis. Fertil Steril. 2012;98(3):511\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLamceva J, Uljanovs R, Strumfa I. The Main Theories on the Pathogenesis of Endometriosis. Int J Mol Sci. 2023;24(5).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen S, Liu Y, Zhong Z, Wei C, Liu Y, Zhu X. Peritoneal immune microenvironment of endometriosis: Role and therapeutic perspectives. Front Immunol. 2023;14:1134663.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGheorghisan-Galateanu AA, Gheorghiu ML. HORMONAL THERAPY IN WOMEN OF REPRODUCTIVE AGE WITH ENDOMETRIOSIS: AN UPDATE. Acta Endocrinol (Buchar). 2019;15(2):276\u0026ndash;81.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVannuccini S, Clemenza S, Rossi M, Petraglia F. Hormonal treatments for endometriosis: The endocrine background. Rev Endocr Metab Disord. 2022;23(3):333\u0026ndash;55.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYu G, Wang LG, Han Y, He QY. clusterProfiler: an R package for comparing biological themes among gene clusters. Omics. 2012;16(5):284\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSubramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005;102(43):15545\u0026ndash;50.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYevtodiyenko A, Schmidt JV. Dlk1 expression marks developing endothelium and sites of branching morphogenesis in the mouse embryo and placenta. Dev Dyn. 2006;235(4):1115\u0026ndash;23.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTraustad\u0026oacute;ttir G, Lagoni LV, Ankerstjerne LBS, Bisgaard HC, Jensen CH, Andersen DC. The imprinted gene Delta like non-canonical Notch ligand 1 (Dlk1) is conserved in mammals, and serves a growth modulatory role during tissue development and regeneration through Notch dependent and independent mechanisms. Cytokine Growth Factor Rev. 2019;46:17\u0026ndash;27.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGrassi ES, Pietras A. Emerging Roles of DLK1 in the Stem Cell Niche and Cancer Stemness. J Histochem Cytochem. 2022;70(1):17\u0026ndash;28.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFalix FA, Aronson DC, Lamers WH, Gaemers IC. Possible roles of DLK1 in the Notch pathway during development and disease. Biochimica et Biophysica Acta (BBA) - Molecular Basis of Disease. 2012;1822(6):988\u0026ndash;95.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi H, Cui ML, Chen TY, Xie HY, Cui Y, Tu H, et al. Serum DLK1 is a potential prognostic biomarker in patients with hepatocellular carcinoma. Tumour Biol. 2015;36(11):8399\u0026ndash;404.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCui D, Liu Y, Ma J, Lin K, Xu K, Lin J. Identification of key genes and pathways in endometriosis by integrated expression profiles analysis. PeerJ. 2020;8:e10171.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSu RW, Strug MR, Joshi NR, Jeong JW, Miele L, Lessey BA, et al. Decreased Notch pathway signaling in the endometrium of women with endometriosis impairs decidualization. J Clin Endocrinol Metab. 2015;100(3):E433-42.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKopan R, Ilagan MX. The canonical Notch signaling pathway: unfolding the activation mechanism. Cell. 2009;137(2):216\u0026ndash;33.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAfshar Y, Jeong JW, Roqueiro D, DeMayo F, Lydon J, Radtke F, et al. Notch1 mediates uterine stromal differentiation and is critical for complete decidualization in the mouse. Faseb j. 2012;26(1):282\u0026ndash;94.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDe Blasio C, Zonfrilli A, Franchitto M, Mariano G, Cialfi S, Verma N, et al. PLK1 targets NOTCH1 during DNA damage and mitotic progression. J Biol Chem. 2019;294(47):17941\u0026ndash;50.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStratopoulou CA, Cussac S, d'Argent M, Donnez J, Dolmans MM. M2 macrophages enhance endometrial cell invasiveness by promoting collective cell migration in uterine adenomyosis. Reprod Biomed Online. 2023;46(4):729\u0026ndash;38.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKhan KN, Yamamoto K, Fujishita A, Muto H, Koshiba A, Kuroboshi H, et al. Differential Levels of Regulatory T Cells and T-Helper-17 Cells in Women With Early and Advanced Endometriosis. J Clin Endocrinol Metab. 2019;104(10):4715\u0026ndash;29.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFreitag N, Pour SJ, Fehm TN, Toth B, Markert UR, Weber M, et al. Are uterine natural killer and plasma cells in infertility patients associated with endometriosis, repeated implantation failure, or recurrent pregnancy loss? Arch Gynecol Obstet. 2020;302(6):1487\u0026ndash;94.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGeng R, Huang X, Li L, Guo X, Wang Q, Zheng Y, et al. Gene expression analysis in endometriosis: Immunopathology insights, transcription factors and therapeutic targets. Front Immunol. 2022;13:1037504.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Endometriosis, Machine learning, Notch signalling, DLK1, biomarker","lastPublishedDoi":"10.21203/rs.3.rs-3990509/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-3990509/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003ePurpose:\u003c/h2\u003e \u003cp\u003eThe objective of this research is to pinpoint potential diagnostic markers for endometriosis and explore the immune infiltration patterns linked with this condition through the utilization of machine learning techniques.\u003c/p\u003e\u003ch2\u003eMethods:\u003c/h2\u003e \u003cp\u003eA total of five gene expression datasets (GSE7305, GSE7307, GSE25628, GSE23339, and GSE120103) were examined in order to identify differentially expressed genes (DEGs) that distinguish normal patients from those with endometriosis. The algorithms Random Forest and Lasso regression were utilised to identify diagnostic biomarkers. GSEA and Go\u0026amp;KEGG database were utilised to determine the potential pathway in which the biomarker was implicated. With the ailment. Furthermore, an assessment of immune cell infiltration in endometriosis tissues relative to normal tissues was conducted using CIBERSORT analysis. In order to investigate the relationship between diagnostic markers and immune cell populations, a correlation analysis was performed.\u003c/p\u003e\u003ch2\u003eResults:\u003c/h2\u003e \u003cp\u003eDLK1 (Delta-like 1 homolog) has emerged as a potential diagnostic biomarker for endometriosis, with indications suggesting that Notch signalling could be pivotal in the development of endometriosis.\u003c/p\u003e\u003ch2\u003eConclusion:\u003c/h2\u003e \u003cp\u003eDLK1 emerges as a promising diagnostic biomarker for endometriosis, as our study indicates a complex interplay between immune dysregulation and disease pathogenesis. Notably, our findings elucidate that DLK1 regulates endometriosis through Notch signaling, highlighting the potential of Notch signalling as a therapeutic target for future interventions.\u003c/p\u003e","manuscriptTitle":"DLK1 as a Potential Biomarker and shows NOTCH signaling could be the potential target for Endometriosis: A Machine Learning Approach","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-03-07 19:30:39","doi":"10.21203/rs.3.rs-3990509/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":"a74b44b1-f316-4d06-beeb-6cb533eb1d0f","owner":[],"postedDate":"March 7th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":29161814,"name":"Biological sciences/Computational biology and bioinformatics"},{"id":29161815,"name":"Biological sciences/Computational biology and bioinformatics/Databases"},{"id":29161816,"name":"Biological sciences/Computational biology and bioinformatics/Machine learning"}],"tags":[],"updatedAt":"2024-08-30T05:22:09+00:00","versionOfRecord":[],"versionCreatedAt":"2024-03-07 19:30:39","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-3990509","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-3990509","identity":"rs-3990509","version":["v1"]},"buildId":"B-jG_2CBjPDmsCi4Wdhf-","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}