Machine learning-based identification of biomarkers associated with NEBL in diabetic nephropathy

Machine learning-based identification of biomarkers associated with NEBL in diabetic nephropathy | 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 Machine learning-based identification of biomarkers associated with NEBL in diabetic nephropathy Yunxi Tao, Shenglong Xu, Xuhua Ge This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4361592/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 Patients with diabetes had a significantly increased risk of cardiovascular disease by reporter, and the nebulette ( NEBL ) gene were closely related with cardiovascular disease. However, the impact of the NEBL gene on diabetic nephropathy (DN) and the underlying molecular mechanisms, have yet to be conclusively validated. Therefore, this study aims to mine NEBL related biomarkers in DN by bioinformatics analysis. A total of 157 differentially expressed genes (DEGs) associated with DN and NEBL gene were excavated, and they were associated with biological processes of mesonephric development and AGE-RAGE signaling pathway in diabetic complications. Besides, totally 19 candidate genes were screened by correlation and expression analyses, among which, MPP5, TGFBR3, PCMTD2 and C1orf21 related to NEBL were deemed as biomarkers via Support Vector Machine-Recursive Feature Elimination (SVM-RFE), Boruta and least absolute shrinkage and selection operator (LASSO) algorithms, the ability of NEBL and biomarkers to distinguish DN from controls was accurate. Immunoinfiltration certified that the contents of 12 differential immune cells were significantly increased in DN group. Then, the gene set enrichment analysis (GSEA) revealed that NEBL and biomarkers were observably enriched in ribosome, intestinal immune network for immunoglobulin A (IgA) production and so on. Finally, drug-gene network revealed bisphenol A and Valproic Acid might be pivotal drugs regulating the expression of NEBL and biomarkers.In this sutudy, totally four biomarkers ( MPP5, TGFBR3, PCMTD2 and C1orf21 ) related to NEBL were screened by bioinformatic analysis, providing a novel reference for effective clinical diagnosis and treatment of DN. Diabetic nephropathy NEBL Biomarkers Bioinformatics analysis Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Synopsis The impact of the NEBL gene on DN and its underlying molecular mechanisms have not been definitively validated. This study identified four NEBL -related biomarkers ( MPP5, TGFBR3, PCMTD2 and C1orf21 ) diagnosing DN through machine learning algorithms. NEBL and biomarkers had the ability to accurately distinguish DN from control samples. NEBL and biomarkers were significantly enriched in immune-related pathways. CD56dim NK cells, mast cells, myeloid-derived suppressor cells (MDSCs), regulatory T cells (Tregs) and follicular T helper cells were negatively correlated with NEBL and biomarkers. 1. Introduction Diabetic nephropathy (DN), characterized by proteinuria, hypertension, and progressive reductions in kidney function, which is a form of chronic kidney disease (CKD) that stands as the primary cause of kidney failure globally 1 . Based on prior research, the growing global prevalence of diabetes, particularly in developing nations, has led to an increase in the number of patients with DN 2 , and patients with DN have a much higher risk of cardiovascular disease (CVD) than those without DN 3 . The onset of DN is insidious, often lacking clear clinical signs in the early stages. The irreversible kidney lesions can lead to delayed treatment, underscoring the importance of early diagnosis and intervention 4, 5 . According to the previous study, the targeted treatment of individuals with DN can be categorized into four main areas: reducing cardiovascular risk, managing glycemic levels, controlling blood pressure, and utilizing RAS inhibition, along with emerging therapies like SGLT-2 inhibitors 6, 7 . Chronic kidney diseases, including DN, typically progress over several years and exhibit a prolonged clinical latency. Consequently, relevant biomarkers hold significant importance in the diagnosis, assessment, and treatment of these conditions 8 . Although the potential early biomarkers of DN were reported, their impact on DN have not been mined. Additionally, there is a pressing need for innovative targets in the diagnosis and treatment of DN. The progress in bioinformatics has been actively harnessed in recent years to investigate potential targets for various conditions, including DN. Nebulin proteins are a group of actin binding proteins that bind a single actin subunit and participate in the assembly of troponin 9 . Nebulette ( NEBL ) gene specifically expresses Nebulin family proteins in myocardium, which is involved in the early development of myocardial tissue and the formation of muscle fiber tissue 1, 9 . Prior investigations have elucidated those mutations in the NEBL gene result in diverse degrees of myocardial detriment 10 . Additionally, it has been substantiated that NEBL gene mutations serve as early indicators of dynamic myocardial changes 11 . Although these collective findings underscore the intimate association between the NEBL gene and cardiovascular disease, the impact of the NEBL gene on the complications of cardiovascular diseases has yet to be confirmed, as well as its influence on DN and the underlying molecular mechanisms. This study aims to investigate the NEBL gene-based biomarkers in DN through bioinformatics analysis. We further seek to and analyze the potential molecular mechanisms underlying the NEBL gene and its associated biomarkers for the sake of thereby offering novel insights as a reference for the clinical diagnosis and treatment of DN. 2. Materials and methods 2.1 Retrieval of data Microarray sequencing data of datasets GSE30528 and GSE47183 were stemmed from the Gene Expression Omnibus (GEO) database ( https://www.ncbi.nlm.nih.gov/geo/ ). In detail, the GSE30528 dataset contained nine DN and 13 control samples, while the GSE47183 dataset embraced seven DN and 14 control samples. 2.2 Differential expression analysis Firstly, the R package ‘limma’ (version 3.52.4) was carried out to screen differentially expressed genes 1 (DEGs1) between DN and control groups from the GSE30528 dataset via differential expression analysis (|Log 2 fold-change (FC) |>1 and p value < 0.05) 12 . At the same time, Wilcoxon rank-sum test was utilized to analyse the expression levels of nebulette ( NEBL ) in DN and control samples of GSE30528 and GSE47183. Subsequently, the DN samples in two dataset were grouped into high and low expression cohorts according to the median of NEBL expression, and differentially expressed genes 2 (DEGs 2) were screened between two cohorts by means of R package ‘limma’ (|Log2FC|>1 and p value < 0.05). Afterwards, DEGs in DN were screened by intersecting upregulated and downregulated differential genes in DEG1 and DEG2, separately. 2.3 Enrichment analysis of DEGs as well as screening of candidate genes For the sake of exploring the biological functions as well as latent pathways of DEGs, Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genome (KEGG) enrichment analyses were executed with the aid of ‘clusterProfiler’ package in R (version 4.7.1.003) (p value < 0.05) 13 . Moreover, in order to probe into the interactions of DEGs at the protein level, a protein-protein interaction (PPI) network of DEGs was synthesized and visualized by feat of the Search Tool for the Retrieval of Interacting Genes database (STRING, https://cn.string-db.org/ ) and Cytoscape software (confidence value = 0.4), respectively. Besides, the correlations between NEBLs and DEGs were analyzed by Spearman’s correlation analysis to screen differentially expressed NEBL (DE- NEBL ) for subsequent analysis (|r|>0.8, p value < 0.05). Ulteriorly, significantly expressed DE- NEBL both in DN versus normal samples and high versus low expression cohorts were identified as candidate genes by expression analysis in GSE30528 and GSE47183 (p value < 0.05). 2.4 Screening of biomarkers in DN Three kinds of machine learning (ML) methods were executed to seek the biomarkers that could accurately diagnose the onset of DN. Firstly, a Support Vector Machine-Recursive Feature Elimination (SVM-RFE) algorithm was executed with R package ‘e1071’ (version 1.7–13) 13 to acquire feature genes on the basis of candidate genes. Meanwhile, feature genes were also obtained from candidate genes by using Boruta algorithm with the ‘Boruta’ package in R (version 8.0.0) 14 . In addition, the least absolute shrinkage and selection operator (LASSO) regression analysis was proceeded with the aid of ‘glmnet’ package in R (version 4.1-8) 15 to identify feature genes. Afterwards, biomarkers for DN were identified by overlapping feature genes derived from the three ML algorithms. Finally, to further evaluate the ability of NEBL and biomarkers to distinguish DN samples from control samples, the R package ‘pROC’ (version 1.7–11) was used to draw receiver operating characteristic (ROC) curve of them in GSE30528 and GSE47183 respectively 16 , and the area under curve (AUC) of ROC was calculated. 2.5 Functional analysis of NEBL and biomarkers In order to further probe into the biological functions of NEBL and biomarkers in DN, the functionally similar genes with NEBL and biomarkers were primarily mined with the aid of GeneMANIA database ( http://genemania.org/ ). The correlations between NEBL and biomarkers were then analyzed by Spearman’s correlation analysis (|Cor|>0.3, p value < 0.05). Furthermore, gene set enrichment analysis (GSEA) was carried out via the R package ‘clusterProfiler’ (version 4.7.1.003) to explore biological functions and latent pathways related with NEBL and biomarkers (p value < 0.05) 13 . The correlation between NEBL , biomarkers as well as other genes in GSE30528 was calculated and ranked according to their correlation coefficients. Subsequently, the KEGG signalling pathway set acquired from the Molecular Signatures Database (MSigDB, https://www.gsea-msigdb.org/ ) was invoked as a background set to perform GSEA. 2.6 Characterization of immune infiltration in DN The immune cell infiltration landscap in DN was characterized to investigate the effect of NEBL and biomarkers on immune cell infiltration in DN. Firstly, based on dataset GSE30528, single-sample gene set enrichment analysis (ssGSEA) was utilized by ‘GSVA’ package in R (version 1.46.0) to severally estimate the immune cells enrichment score of 28 immune cells types in DN as well as control group 17 , and the distinctions of immune cell scores between two groups were compared. In addition, the relationships between NEBL and differential immune cells were computed by Spearman’ correlation analysis, as well as the correlations between biomarkers and differential immune cells (|Cor|>0.3, p value < 0.05) 2.7 Contruction of molecular regulatory networks A lncRNA-miRNA-mRNA network was synthesized with the aim of profoundly researching the latent mechanism of NEBL and biomarkers in the occurrence and process of DN. Firstly, the critical miRNAs targeting NEBL and biomarkers were predicted by intersecting miRNAs acquired from miRDB database ( https://mirdb.org/ ), miRWalk database ( http://129.206.7.150/ ), and Targetscan database ( https://www.targetscan.org/vert_80/ ). Subsequently, lncRNAs targetting critical miRNAs were screened by Starbase ( https://rnasysu.com/encori/ ) (clipExpNum > 15). Besides, for the sake of exploring whether NEBL and biomarkers had an effect on DN at the molecular level, the analysis of transcription factors (TFs) targetting NEBL and biomarkers were indispensable. The JASPAR database ( https://ngdc.cncb.ac.cn/databasecommons/database/id/176 ) was employed to search TFs targetting NEBL and biomarkers. The study of drugs targeting NEBL and biomarkers might provide new strategies for the treatment of DN, consequently, the CTD database ( http://ctdbase.org/ ) were utilized to predict drugs related with NEBL and biomarkers. Finally, the lncRNA-miRNA-mRNA, TF-gene and drug-gene networks of NEBL and biomarkers were structured via Cytoscape software (version 3.7.1) 18 . 2.8 Statistical Analysis In this study, the R software was engaged in statistical analysis. Differences between two groups were analyzed via the Wilcoxon rank-sum test, and p value of less than 0.05 was represented a significant difference. 3. Results 3.1 A total of 157 DEGs were mined The expression of NEBL was significantly downregulated in DN group of both GSE30528 and GSE47183 datasets (Fig. 1 ). A number of 424 DEGs1 were obtained between DN and control groups, including 113 up-regulated and 311 down-regulated genes (Fig. 2 A-B). A total of 176 DEGs2 were screened, the numbers of upregulation and downregulation were 56 and 120, respectively (Fig. 2 C-D). Totally 157 DEGs covering 41 up-regulated and 116 down-regulated genes were obtained by intersecting the DEGs1 with DEGs2 (Fig. 2 E). 3.2 A sum of 19 candidate genes were excavated GO and KEGG were used to analyse DEGs with the aim of delving into their potential biological functions and signaling pathways. GO analysis indicated that DEGs were enriched to 237 entries, encompassing four cellular components (CCs), 12 molecular functions (MFs) and 226 biological process (BPs). To be specific, DEGs were associated with mesonephric epithelium development, mesonephros development and renal system development (Fig. 3 A). With respect to seven KEGG pathway involved by DEGs, AGE − RAGE signaling pathway in diabetic complications, Rap1 signaling pathway and complement and coagulation cascades and other pathways were notable (Fig. 3 B). In order to investigate the interactions of DEGs at the protein level, a PPI network of DEGs was constructed, which contained 80 nodes and 202 edges. Thereinto, the interaction pairs with greater co-expression included COL1A2-LUM, SPP1-COL1A2 and COL1A2-COL6A3 (Fig. 3 C). A sum of 64 DEGs strongly associated with NEBL were noted as DE-NEBL by Spearman’s correlation analysis. The result manifested that 26 DE- NEBL expressed discrepantly between DN and control samples in GSE30528 and GSE47183 datasets (Fig. 3 D-E), and a number of 31 DE- NEBL kept different expression levels between high and low expression groups (Fig. 3 F-G). To sum up, a total of 19 DE- NEBL were were counted as candidate genes as a result of their discrepant expression levels in DN and control groups, as well as high and low expression groups. 3.3 Totally four biomarkers were identified Three ML methods were exploited to screen biomarkers among candidate genes obtained above. Firstly, a total of 15 feature genes were acquired via SVM-RFE algorithm (Fig. 4 A). Meanwhile, Boruta algorithm indicated that all the 19 candidate genes obtained above were retained and ranked according to their importance on the basis of their importance (Fig. 4 B). Additionally, LASSO regression analysis was performed to screen out four feature genes (Fig. 4 C-D). As a result, totally four biomarkers ( MPP5, TGFBR3, PCMTD2 and C1orf21 ) were screened via intersecting the feature genes acquired above (Fig. 4 E). Whereafter, ROC curves were drawn to evaluate the diagnostic performance of NEBL as well as four biomarkers for DN, and the results manifested that the AUC values of ROC curves were all greater than 0.7 in GSE30528, the validation results in dataset GSE47183 were consistent with the above results, implying the diagnostic performance of NEBL and four biomarkers was fine (Fig. 4 F-G). 3.4 NEBL and biomarkers were significantly enriched in immune-related pathways The functionally alike genes with NEBL and biomarkers were mined and a gene-gene interaction (GGI) network showing the interactions among them was synthesized. NEBL and four biomarkers were associated with similar genes through co-localization, genetic interactions and physical interactions, etc. Additionally, gene pairs of NEBL-PATJ , C1orf21-TGFB1 and MPP5-CRB3 were included in the network (Fig. 5 A). In addition, NEBL and biomarkers were mainly related to cellular response to transforming growth factor beta stimulus, transmembrane receptor protein serine/threonine kinase signaling pathway and tight junction organization, etc. Synchronously, correlation analysis revealed that there was a remarkable positive correlation between NEBL and MPP5 (r = 0.874, p value < 0.001), TGFBR3 (r = 0.844, p value < 0.001), PCMTD2 (r = 0.836, p value < 0.001) as well as C1orf21 (r = 0.802, p value < 0.001) (Fig. 5 B). Then, pathways involved in NEBL and biomarkers were analyzed, the result revealed that NEBL and biomarkers were observably enriched in ribosome, intestinal immune network for immunoglobulin A (IgA) production and systemic lupus erythematosus, etc. (Fig. 5 C-G) (p value < 0.05). 3.5 NEBL and biomarkers might affect DN by altering the expression of 12 immune cells Since NEBL and biomarker function were relevant to immunologically, immune infiltration analysis was proceed, and the abundance of 28 immune cells for each sample in GSE30528 was presented on heatmap (Fig. 5 H). Totally 12 immune cells appeared discrepant immune infiltration between DN and control groups, such as CD56dim natural killer (NK) cells, regulatory T cells (Tregs) and mast cells, etc. And the contents of them were significantly increased in DN group (Fig. 5 I). Moreover, CD56dim NK cells, mast cells, myeloid-derived suppressor cells (MDSCs), Tregs and follicular T helper cells were negatively correlated with NEBL and biomarkers among the 12 discrepant immune cells, there was no significantly correlated between T helper Type 1 cells and NEBL as well as biomarkers (p > 0.05). In addition, CD56dim NK cells maintained the strongest correlation with NEBL (r=-0.718, p value < 0.001), MPP5 (r=-0.752, p value < 0.001), TGFBR3 (r=-0.787, p value < 0.001), PCMTD2 (r=-0.674, p value < 0.001) and C1orf21 (r=-0.804, p value < 0.001) (Fig. 5 J). 3.6 The lncRNA-miRNA-mRNA, TF-gene and drug-gene networks of NEBL and biomarkers were constructed A total of 91 critical miRNAs targetting NEBL and biomarkers were predicted with the aid of three databases ( TGFBR3 : 24, PCMTD2 : 25, NEBL : 9, C1orf21 : 6 and M PP5: 27) (Fig. 6 A-E). A number of 103 lncRNAs were acquired based on critical miRNAs (clipExpNum > 15). The ceRNA network was composed of five miRNAs ( NEBL, MPP5, TGFBR3, PCMTD2 and C1orf21 ), 88 miRNAs (hsa-miR-130a-5p, hsa-miR-613 and hsa-miR-501-5p, etc.), and 10 lncRNAs (XIST, NEAT1 and MALAT1, etc.), among which the interaction pairs included TGFBR3 -HSA-Mir-130A-5P-AC010980.2, MPP5 -hsa-miR-199b-5p-LINC01783, NEB L-hsa-miR-497-5p-XIST, etc. (Fig. 6 F). Ulteriorly, TFs targeting NEBL and MPP5 , TGFBR3 , PCMTD2 w ere predicted in addition to C1orf21 . As presented in Fig. 6 G, TFs that co-target PCMTD2 and NEBL were POU2F2, STAT3 and TFAP2A, etc.. GFBR3 and NEBL were regulated jointly by SREBF1, CEBPB, FOS and FOXC1, etc.. Besides, TFs such as USF2, NFIC and TEAD1 jointly targeted to NEBL and MPP5 . With the respect of drug prediction, compounds interacted with NEBL and biomarkers added up to 210, among which 42 compounds (Reference Count > 1) were regarded as drugs to construct the gene-drug network (Fig. 6 H). The network was consisted of 31 nodes and 42 edges. It was worth noiticted that NEBL, PCMTD2 and TGFBR3 were jointly targeted by bisphenol A, and the common targets of Valproic Acid were PCMTD2, TGFBR3, NEBL and C1ORF21 , meaning that bisphenol A and Valproic Acid might provide reference for the treatment of DN. 4. Discussion The onset of DN is instigated by a confluence of various factors, nevertheless, the specific mechanisms of this phenomenon are yet to be elucidated 19, 20 . Currently, the early diagnosis and treatment of DN present challenges, necessitating the exploration of novel molecular pathways to solving this problem. Based on prior studies and experiments, the onset of DN, to some extent, is associated with immune infiltration, immune mediators, and oxidative stress 21–23 . The NEBL gene, located on chromosome 10q22.2, codes for a protein called Nebulette, which has traditionally been linked to heart function. However, recent studies have revealed its relationship with nephropathy, arousing curiosity about its potential significance in DN 24 . This study involved the retrieval of transcriptomic datasets from the GEO database, employing bioinformatics methodologies to pinpoint differentially expressed genes pivotal in DN. Through ML-based screening, four NEBL-related genes, MPP5, TGFBR3, PCMTD2 , and C1orf21 , were identified. ROC analysis showed that the NEBL gene and four NEBL -related genes possessed the btility of distinguishing DN samples and controls, suggesting that they might be developed as diagnostic biomarkers. In addition, the functions of NEBL genes and biomarkers were investigated with the aim of understanding their potential molecular mechanisms of them on occurrence and development of DN. Membrane Protein Palmitoylated 5 ( MPP5 ), also recognized as Protein Associated with Lin7 ( PALS1 ), functions as a fundamental constituent of the apical membrane determining CRB complex within the nephron, and the absence of Pals1 leads to tubular dilation and cyst formation 25 . However, there are a scarcity of studies investigating the role of MPP5 in the development of DN. In our study, we observed that down-regulated MPP5 also had better diagnostic efficacy as a result of its AUC value of 1. TGFBR3 , a membrane proteoglycan, serves as a co-receptor with other transforming growth factor receptors and is expressed in glomerular podocytes, mesangial cells, and endothelial cells 26, 27 . With a high specificity for the glomerulus, TGFBR3 exhibits a particular affinity for two subtypes of TGF-β, forming specific binding interactions A decrease in TGF-β secretion triggers cytokines to stimulate T helper1 (Th1) cells, leading to the production of macrophage M1 and fostering the inflammatory response of Th1 cells, which is an essential factor in the progression of DN. The study suggested that a potential therapeutic target for DN involving intervention with TGF-β could be TGFBR3 28 . Additionally, we observed elevated levels of TGFBR3 in DN patients, demonstrating diagnostic accuracy values with AUC exceeding 0.70. The outcomes from GeneMANIA indicated that TGFBR3, TGF-β1, TGF-β2 , and TGF-β3 might collaboratively influence the progression of DN through physical interactions. This reinforces the credibility of the findings in this study. Protein-L-Isoaspartate O-Methyltransferase Domain Containing 2 ( PCMTD2 ) is anticipated to possess protein-L-isoaspartate (D-aspartate) O-methyltransferase activity and play a role in protein methylation, and its predicted activity is situated in the cytoplasm. Chromosome 1 Open Reading Frame 21 ( C1orf21 ) is classified as a protein-coding gene. Diseases linked to C1orf21 encompass thymoma type B2 and renal artery atheroma. PCMTD2 and C1orf21 , particularly in the context of kidney disease, have limited existing knowledge. However, our study heightened the enhanced expression levels of PCMTD2 and C1orf21 in DN patients compared to controls. Considering the limited available literature regarding the influence of PCMTD2 and C1orf21 on the progression of DN, it is prudent to make inference based on the results of this study, generally, increased gene expression levels is believed to be linked with an elevated disease risk. The study's findings revealed increased expression of PCMTD2 and C1orf21 in DN patients, suggesting that these genes might serve as potential risk factors for DN. However, further analysis and experimental validation are requisite to substantiate this hypothesis. GO and KEGG enrichment analysis found that DEGs mainly enriched in BPs related to kidney development, indicating that these DEGs might be related to the development of DN, and this result further verified the reliability of DEGs. In our analysis, we found that the AGE-RAGE signaling pathway played an important role in DN. Previous studies have demonstrated that the presence and expression of AGE and RAGE have increased in human diabetic kidneys, especially in the glomerulus, glomerular epithelial cells (podocytes) and endothelial cells 29 . The absence of RAGE has been found to effectively mitigate glomerular sclerosis, glomerular basement membrane thickening, podocyte loss, as well as decline in glomerular filtration rate (GFR). In addition, RAGE is also present in the angiotensin II (angII) axis, which is also associated with the expression of RAGE and the occurrence of ND 30 . GSEA revealed that the NEBL gene demonstrated enrichment in immune-related pathways. Additionally, GSEA demonstrated significant enrichment of the NEBL gene and four biomarkers in pathways associated with ribosomes, the IgA-produced intestinal immune network, systemic lupus erythematosus, primary immunodeficiency, leishmania infection, cell adhesion molecules (CAM), glutathione metabolism, and others. Besides, a significant enrichment of biomarkers, such as the NEB L and TGFBR3 , in pathways associated with immunity. In conclusion, we infered that NEBL and TGFBR3 might play similar immune functions in the development of DN. The findings of immune infiltration analysis elucidated a close association between the development of DN and CD56 NK cells, mast cells, and regulatory T cells. According to previous studies, Natural Killer T (NKT) cells can be found to be associated with kidney damage as well as vascular damage in type 2 diabetes, but few studies have focused on NKT cells and DN, especially CD56 NK cells, which would be a new focus on DN 31–33 . Drug prediction revealed two compounds that were most closely targeted towards NEBL gene and biomarkers, bisphenol A (BPA) and Valproic Acid. BPA demonstrates an association with low-grade albuminuria, a manifestation characteristic of the initial phases of DN 34 . Moreover, findings certified that valproic acid mitigated the diabetes-induced upregulation of complement C5a receptors, concurrently diminishing indicators of cellular senescence and the senescence-associated secretory phenotype. The attenuation of cellular senescence in the diabetic context assumes significance, given its implication in the pathogenesis of diabetic kidney disease 35 . Consequently, therapeutic interventions targeting cellular senescence, such as complement inhibitors, emerge as a novel and promising avenue for the treatment of diabetic kidney disease. Based on current research, the reduction of NEBL in DN may exacerbate the condition. We hypothesize that BPA may reverse the expression of NEBL in DN, thereby potentially alleviating the progression of DN. Nevertheless, it is crucial to acknowledge the limitations of this study. The evidence relies on publicly available data, and while we validated the expression using another dataset, additional experiments are imperative to substantiate NEBL and the four associated genes as diagnostic markers before their clinical applicability can be established. At the same time, the clinical application research of drugs also needs experimental proof, and we will continue to pay attention to the relevant research progress of DN. Declarations Funding: This work was supported by the National Natural Science Foundation of China to Xuhua Ge (Nos. 82170733). Author Contribution YT performed the data analyses and wrote the manuscript; SX contributed to the analysis andmanuscript preparation; XG contributed to the conception of the study and helped perform theanalysis with constructive discussions. All authors participated in the conference workgroups, thedevelopment of the summary statement, and the review of the manuscript. All authors read andapproved the final manuscript. Acknowledgments: I really appreciate Dr.Xuhua Ge of Children’s Hospital of Nanjing Medical University (Nanjing, China) for her guidance on the paper and her assistance with the bioinformatics analysis. References Esham M, Bryan K, Milnes J, Holmes WB, Moncman CL. Expression of nebulette during early cardiac development. Cell Motil Cytoskeleton 2007, 64: 258-73. Yang C, Wang H, Zhao X, Matsushita K, Coresh J, Zhang L, Zhao MH. CKD in China: Evolving Spectrum and Public Health Implications. Am J Kidney Dis 2020, 76: 258-64. Palsson R, Patel UD. Cardiovascular complications of diabetic kidney disease. Adv Chronic Kidney Dis 2014, 21: 273-80. Xin C, Sun X, Li Z, Gao T. Relationship of Soluble Klotho and Early Stage of Diabetic Nephropathy: A Systematic Review and Meta-Analysis. Front Endocrinol (Lausanne) 2022, 13: 902765. Viigimaa M, Sachinidis A, Toumpourleka M, Koutsampasopoulos K, Alliksoo S, Titma T. Macrovascular Complications of Type 2 Diabetes Mellitus. Curr Vasc Pharmacol 2020, 18: 110-6. Alicic R, Nicholas SB. Diabetic Kidney Disease Back in Focus: Management Field Guide for Health Care Professionals in the 21st Century. Mayo Clin Proc 2022, 97: 1904-19. Wang L, Zhang Y, Liu S, Huang N, Zeng W, Xu W, Zhou T, et al. Comparison of Anti-Microbic and Anti-Biofilm Activity Among Tedizolid and Radezolid Against Linezolid-Resistant Enterococcus faecalis Isolates. Infect Drug Resist 2021, 14: 4619-27. Lopez-Giacoman S, Madero M. Biomarkers in chronic kidney disease, from kidney function to kidney damage. World J Nephrol 2015, 4: 57-73. Bonzo JR, Norris AA, Esham M, Moncman CL. The nebulette repeat domain is necessary for proper maintenance of tropomyosin with the cardiac sarcomere. Exp Cell Res 2008, 314: 3519-30. Purevjav E, Varela J, Morgado M, Kearney DL, Li H, Taylor MD, Arimura T, et al. Nebulette mutations are associated with dilated cardiomyopathy and endocardial fibroelastosis. J Am Coll Cardiol 2010, 56: 1493-502. Maiellaro-Rafferty K, Wansapura JP, Mendsaikhan U, Osinska H, James JF, Taylor MD, Robbins J, et al. Altered regional cardiac wall mechanics are associated with differential cardiomyocyte calcium handling due to nebulette mutations in preclinical inherited dilated cardiomyopathy. J Mol Cell Cardiol 2013, 60: 151-60. Liu S, Xie X, Lei H, Zou B, Xie L. Identification of Key circRNAs/lncRNAs/miRNAs/mRNAs and Pathways in Preeclampsia Using Bioinformatics Analysis. Med Sci Monit 2019, 25: 1679-93. Yu G, Wang LG, Han Y, He QY. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS 2012, 16: 284-7. Shubin NJ, Navalkar K, Sampson D, Yager TD, Cermelli S, Seldon T, Sullivan E, et al. Serum Protein Changes in Pediatric Sepsis Patients Identified With an Aptamer-Based Multiplexed Proteomic Approach. Crit Care Med 2020, 48: e48-e57. Wu X, Qin K, Iroegbu CD, Xiang K, Peng J, Guo J, Yang J, et al. Genetic analysis of potential biomarkers and therapeutic targets in ferroptosis from coronary artery disease. J Cell Mol Med 2022, 26: 2177-90. Robin X, Turck N, Hainard A, Tiberti N, Lisacek F, Sanchez JC, Muller M. pROC: an open-source package for R and S+ to analyze and compare ROC curves. BMC Bioinformatics 2011, 12: 77. Ren Z, He Y, Yang Q, Guo J, Huang H, Li B, Wang D, et al. A Comprehensive Analysis of the Glutathione Peroxidase 8 (GPX8) in Human Cancer. Front Oncol 2022, 12: 812811. Costa-Silva J, Domingues D, Lopes FM. RNA-Seq differential expression analysis: An extended review and a software tool. PLoS One 2017, 12: e0190152. Natesan V, Kim SJ. Diabetic Nephropathy - a Review of Risk Factors, Progression, Mechanism, and Dietary Management. Biomol Ther (Seoul) 2021, 29: 365-72. Hu F, Zhang T. Study on Risk Factors of Diabetic Nephropathy in Obese Patients with Type 2 Diabetes Mellitus. Int J Gen Med 2020, 13: 351-60. Yu K, Li D, Xu F, Guo H, Feng F, Ding Y, Wan X, et al. IDO1 as a new immune biomarker for diabetic nephropathy and its correlation with immune cell infiltration. Int Immunopharmacol 2021, 94: 107446. Kahles F, Meyer C, Mollmann J, Diebold S, Findeisen HM, Lebherz C, Trautwein C, et al. GLP-1 secretion is increased by inflammatory stimuli in an IL-6-dependent manner, leading to hyperinsulinemia and blood glucose lowering. Diabetes 2014, 63: 3221-9. Celejewski-Marciniak P, Wolinowska R, Wroblewska M. Molecular Characterization of Class 1, 2 and 3 Integrons in Serratia spp. Clinical Isolates in Poland - Isolation of a New Plasmid and Identification of a Gene for a Novel Fusion Protein. Infect Drug Resist 2021, 14: 4601-10. Ge X, Zhang T, Yu X, Muwonge AN, Anandakrishnan N, Wong NJ, Haydak JC, et al. LIM-Nebulette Reinforces Podocyte Structural Integrity by Linking Actin and Vimentin Filaments. J Am Soc Nephrol 2020, 31: 2372-91. Weide T, Vollenbroker B, Schulze U, Djuric I, Edeling M, Bonse J, Hochapfel F, et al. Pals1 Haploinsufficiency Results in Proteinuria and Cyst Formation. J Am Soc Nephrol 2017, 28: 2093-107. Chen S, Li B, Chen L, Jiang H. Identification and validation of immune-related biomarkers and potential regulators and therapeutic targets for diabetic kidney disease. BMC Med Genomics 2023, 16: 90. Yamamoto T, Watanabe T, Ikegaya N, Fujigaki Y, Matsui K, Masaoka H, Nagase M, et al. Expression of types I, II, and III TGF-beta receptors in human glomerulonephritis. J Am Soc Nephrol 1998, 9: 2253-61. Zhu Y, Usui HK, Sharma K. Regulation of transforming growth factor beta in diabetic nephropathy: implications for treatment. Semin Nephrol 2007, 27: 153-60. Ramasamy R, Yan SF, Schmidt AM. Receptor for AGE (RAGE): signaling mechanisms in the pathogenesis of diabetes and its complications. Ann N Y Acad Sci 2011, 1243: 88-102. Ruster C, Franke S, Wenzel U, Schmidthaupt R, Fraune C, Krebs C, Wolf G. Podocytes of AT2 receptor knockout mice are protected from angiotensin II-mediated RAGE induction. Am J Nephrol 2011, 34: 309-17. Wang H, Cao K, Liu S, Xu Y, Tang L. Tim-3 Expression Causes NK Cell Dysfunction in Type 2 Diabetes Patients. Front Immunol 2022, 13: 852436. Uchida T, Ito S, Kumagai H, Oda T, Nakashima H, Seki S. Roles of Natural Killer T Cells and Natural Killer Cells in Kidney Injury. Int J Mol Sci 2019, 20: Lv X, Gao Y, Dong T, Yang L. Role of Natural Killer T (NKT) Cells in Type II Diabetes-Induced Vascular Injuries. Med Sci Monit 2018, 24: 8322-32. Jiang W, Ding K, Huang W, Xu F, Lei M, Yue R. Potential effects of bisphenol A on diabetes mellitus and its chronic complications: A narrative review. Heliyon 2023, 9: e16340. Coughlan MT, Ziemann M, Laskowski A, Woodruff TM, Tan SM. Valproic acid attenuates cellular senescence in diabetic kidney disease through the inhibition of complement C5a receptors. Sci Rep 2022, 12: 20278. 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-4361592","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":300814244,"identity":"2f5644a9-f359-467c-a1c8-b7a3cb155011","order_by":0,"name":"Yunxi Tao","email":"","orcid":"","institution":"Nanjing Medical University","correspondingAuthor":false,"prefix":"","firstName":"Yunxi","middleName":"","lastName":"Tao","suffix":""},{"id":300814245,"identity":"8e0403b8-3f23-41f0-80e8-1adc951b4a92","order_by":1,"name":"Shenglong Xu","email":"","orcid":"","institution":"Nanjing Medical University","correspondingAuthor":false,"prefix":"","firstName":"Shenglong","middleName":"","lastName":"Xu","suffix":""},{"id":300814246,"identity":"38060e92-1f79-409f-b6b1-443298211a84","order_by":2,"name":"Xuhua Ge","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA0klEQVRIiWNgGAWjYJACZiCWY2A4AGMTqcWYdC2JDQg2ASDv3mMmXdhml76d8XSaBEOFdWID+9kDeLUYnjmWJj2zLTl3Z8PZbRIMZ9ITG3jyEvBrmZF8TJq3jTl3wwGgFsa2w4kNEjwG+LXMf9gG1FKfbgDW8o8ILfISzCBbDidAtDQQocWAJy3ZmufccUOgwzZbJBxLN27jySFgS/sZw9s8ZdXyBjfObrzxocZatp/9DAFbDsBYEkBWApBmw6seZEsDjMXfgFvVKBgFo2AUjGwAANLWRVDqM3OgAAAAAElFTkSuQmCC","orcid":"","institution":"Children’s Hospital of Nanjing Medical University","correspondingAuthor":true,"prefix":"","firstName":"Xuhua","middleName":"","lastName":"Ge","suffix":""}],"badges":[],"createdAt":"2024-05-03 02:40:09","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4361592/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4361592/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":56410367,"identity":"31830923-84fa-492a-bbbd-68411ed6659c","added_by":"auto","created_at":"2024-05-13 20:19:38","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":133130,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDifferentia expression of NEBL gene in DN and control samples of GSE30528 and GSE47183 datasets.\u003c/strong\u003e *P\u0026lt;0.05, **P\u0026lt;0.01, ***P\u0026lt;0.001.\u003c/p\u003e","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-4361592/v1/8878fd5ef3ea55b34e95e174.png"},{"id":56410368,"identity":"d011875c-1865-4b17-8d67-55758dbbe15a","added_by":"auto","created_at":"2024-05-13 20:19:38","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":594967,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAcquisition of differentially expressed genes (DEGs)\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eVolcano plot of DEGs1 in GSE30528 (A) and DEGs2 in high and low NEBL expression cohorts of GSE30528 (C). Heat map of DEGs1 in GSE30528 (B) and DEGs2 in high and low NEBL expression cohorts of GSE30528 (D). (E) Intersection of DEGs1 and DEGs2 in GSE30528. *adj.P\u0026lt;0.05\u003c/p\u003e","description":"","filename":"Figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-4361592/v1/dfee70c81a3deadd50e146ac.png"},{"id":56410372,"identity":"ee89bd39-7c79-4aab-b334-a997f8a8d790","added_by":"auto","created_at":"2024-05-13 20:19:39","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1175604,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eExploration of biological functions associated with DEGs and Identification of candidate genes related to NEBL gene\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIdentification of KEGG (A) and GO (B) entries involved by DEGs. (C) Construction of a protein-protein interaction (PPI) network of DEGs. Differential expression of DEGs in DN and control samples of GSE30528 (D) and GSE47183 (E) datasets. Differential expression of DE-NEBLin DN and control samples of GSE30528 (F) and GSE47183 (G) datasets. *P\u0026lt;0.05, **P\u0026lt;0.01, ***P\u0026lt;0.001.\u003c/p\u003e","description":"","filename":"Figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-4361592/v1/1ec6eb2b0f949ce45b133c20.png"},{"id":56410369,"identity":"7b0e55f2-78e4-437a-99d7-509449b60b93","added_by":"auto","created_at":"2024-05-13 20:19:39","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":420059,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eScreening of NEBL-related biomarkers in DN\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAcquirement of biomarkers (E) in DN by overlapping feature genes derived from SVM-RFE algorithm (A), Boruta algorithm (B) and LASSO regression analysis (C-D). Evaluation and validation of NEBL genes and its associated biomarkers in GSE30528 (F) and GSE47183 (G) datasets.\u003c/p\u003e","description":"","filename":"Figure4.png","url":"https://assets-eu.researchsquare.com/files/rs-4361592/v1/b1e244888526c2d91cc838b0.png"},{"id":56410373,"identity":"f9b16123-107f-43a9-b832-b55058e0184c","added_by":"auto","created_at":"2024-05-13 20:19:39","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":2045539,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFunctional exploration of NEBL genes and biomarkers and immunoinfiltration characterization in GSE30528 dataset\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eGene-gene interaction (GGI) network of NEBL genes and biomarkers (A). The correlation among NEBL genes and biomarkers (B). GSEA-KEGG pathway analysis in TGFB13\u003cem\u003e \u003c/em\u003e(C), MPP5\u003cem\u003e \u003c/em\u003e(D), NEBL (E), PCMTD2\u003cem\u003e \u003c/em\u003e(F) and C1orf21\u003cem\u003e \u003c/em\u003e(G). Heatmap (H), box plot (I) of immune cell infiltration abundance in DN and control groups. The correlation heatmap between NEBL genes, biomarkers and differential immune cells (J).*P\u0026lt;0.05, **P\u0026lt;0.01, ***P\u0026lt;0.001.\u003c/p\u003e","description":"","filename":"Figure5.png","url":"https://assets-eu.researchsquare.com/files/rs-4361592/v1/ffab5122dac65790d04ea10f.png"},{"id":56410371,"identity":"d964e2f6-0768-4332-883c-dbcad5ee301c","added_by":"auto","created_at":"2024-05-13 20:19:39","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":846564,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe construction of lncRNA-miRNA-mRNA, TF-gene and drug-gene networks\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe Venn diagram showing the critical miRNAs of C1orf21 (A), PCMTD2\u003cem\u003e \u003c/em\u003e(B), TGFB3 (C), MPP5\u003cem\u003e \u003c/em\u003e(D) and NEBL (E). The lncRNA-miRNA-mRNA (F), TF-gene (G) and drug-gene (H) networks.\u003c/p\u003e","description":"","filename":"Figure6.png","url":"https://assets-eu.researchsquare.com/files/rs-4361592/v1/8d912f36b56eeef2a858dd15.png"},{"id":56778156,"identity":"85e68d97-406f-4945-ac0e-6cdc2cad9a67","added_by":"auto","created_at":"2024-05-20 10:51:19","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":6063370,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4361592/v1/3b1780db-6aae-448b-a511-3a45a6adde79.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Machine learning-based identification of biomarkers associated with NEBL in diabetic nephropathy","fulltext":[{"header":"Synopsis","content":"\u003cp\u003eThe impact of the \u003cem\u003eNEBL\u003c/em\u003e gene on DN and its underlying molecular mechanisms have not been definitively validated. This study identified four \u003cem\u003eNEBL\u003c/em\u003e-related biomarkers (\u003cem\u003eMPP5, TGFBR3, PCMTD2\u0026nbsp;\u003c/em\u003eand \u003cem\u003eC1orf21\u003c/em\u003e) diagnosing DN through machine learning algorithms.\u003c/p\u003e\n\u003cul\u003e\n \u003cli\u003e\u003cem\u003eNEBL\u003c/em\u003e and biomarkers had the ability to accurately distinguish DN from control samples.\u003c/li\u003e\n \u003cli\u003e\u003cem\u003eNEBL\u003c/em\u003e and biomarkers were significantly enriched in immune-related pathways.\u003c/li\u003e\n \u003cli\u003eCD56dim NK cells, mast cells, myeloid-derived suppressor cells (MDSCs), regulatory T cells (Tregs) and follicular T helper cells were negatively correlated with \u003cem\u003eNEBL\u003c/em\u003e and biomarkers.\u003c/li\u003e\n\u003c/ul\u003e"},{"header":"1. Introduction","content":"\u003cp\u003eDiabetic nephropathy (DN), characterized by proteinuria, hypertension, and progressive reductions in kidney function, which is a form of chronic kidney disease (CKD) that stands as the primary cause of kidney failure globally\u003csup\u003e1\u003c/sup\u003e. Based on prior research, the growing global prevalence of diabetes, particularly in developing nations, has led to an increase in the number of patients with DN\u003csup\u003e2\u003c/sup\u003e, and patients with DN have a much higher risk of cardiovascular disease (CVD) than those without DN\u003csup\u003e3\u003c/sup\u003e. The onset of DN is insidious, often lacking clear clinical signs in the early stages. The irreversible kidney lesions can lead to delayed treatment, underscoring the importance of early diagnosis and intervention\u003csup\u003e4, 5\u003c/sup\u003e. According to the previous study, the targeted treatment of individuals with DN can be categorized into four main areas: reducing cardiovascular risk, managing glycemic levels, controlling blood pressure, and utilizing RAS inhibition, along with emerging therapies like SGLT-2 inhibitors\u003csup\u003e6, 7\u003c/sup\u003e. Chronic kidney diseases, including DN, typically progress over several years and exhibit a prolonged clinical latency. Consequently, relevant biomarkers hold significant importance in the diagnosis, assessment, and treatment of these conditions\u003csup\u003e8\u003c/sup\u003e. Although the potential early biomarkers of DN were reported, their impact on DN have not been mined.\u003c/p\u003e \u003cp\u003eAdditionally, there is a pressing need for innovative targets in the diagnosis and treatment of DN. The progress in bioinformatics has been actively harnessed in recent years to investigate potential targets for various conditions, including DN.\u003c/p\u003e \u003cp\u003eNebulin proteins are a group of actin binding proteins that bind a single actin subunit and participate in the assembly of troponin\u003csup\u003e9\u003c/sup\u003e. Nebulette (\u003cem\u003eNEBL\u003c/em\u003e) gene specifically expresses Nebulin family proteins in myocardium, which is involved in the early development of myocardial tissue and the formation of muscle fiber tissue\u003csup\u003e1, 9\u003c/sup\u003e. Prior investigations have elucidated those mutations in the \u003cem\u003eNEBL\u003c/em\u003e gene result in diverse degrees of myocardial detriment\u003csup\u003e10\u003c/sup\u003e. Additionally, it has been substantiated that \u003cem\u003eNEBL\u003c/em\u003e gene mutations serve as early indicators of dynamic myocardial changes\u003csup\u003e11\u003c/sup\u003e. Although these collective findings underscore the intimate association between the \u003cem\u003eNEBL\u003c/em\u003e gene and cardiovascular disease, the impact of the \u003cem\u003eNEBL\u003c/em\u003e gene on the complications of cardiovascular diseases has yet to be confirmed, as well as its influence on DN and the underlying molecular mechanisms.\u003c/p\u003e \u003cp\u003eThis study aims to investigate the \u003cem\u003eNEBL\u003c/em\u003e gene-based biomarkers in DN through bioinformatics analysis. We further seek to and analyze the potential molecular mechanisms underlying the \u003cem\u003eNEBL\u003c/em\u003e gene and its associated biomarkers for the sake of thereby offering novel insights as a reference for the clinical diagnosis and treatment of DN.\u003c/p\u003e"},{"header":"2. Materials and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Retrieval of data\u003c/h2\u003e \u003cp\u003eMicroarray sequencing data of datasets GSE30528 and GSE47183 were stemmed from the Gene Expression Omnibus (GEO) database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ncbi.nlm.nih.gov/geo/\u003c/span\u003e\u003cspan address=\"https://www.ncbi.nlm.nih.gov/geo/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). In detail, the GSE30528 dataset contained nine DN and 13 control samples, while the GSE47183 dataset embraced seven DN and 14 control samples.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Differential expression analysis\u003c/h2\u003e \u003cp\u003eFirstly, the R package \u0026lsquo;limma\u0026rsquo; (version 3.52.4) was carried out to screen differentially expressed genes 1 (DEGs1) between DN and control groups from the GSE30528 dataset via differential expression analysis (|Log\u003csub\u003e2\u003c/sub\u003e fold-change (FC) |\u0026gt;1 and p value\u0026thinsp;\u0026lt;\u0026thinsp;0.05)\u003csup\u003e12\u003c/sup\u003e. At the same time, Wilcoxon rank-sum test was utilized to analyse the expression levels of nebulette (\u003cem\u003eNEBL\u003c/em\u003e) in DN and control samples of GSE30528 and GSE47183. Subsequently, the DN samples in two dataset were grouped into high and low expression cohorts according to the median of \u003cem\u003eNEBL\u003c/em\u003e expression, and differentially expressed genes 2 (DEGs 2) were screened between two cohorts by means of R package \u0026lsquo;limma\u0026rsquo; (|Log2FC|\u0026gt;1 and p value\u0026thinsp;\u0026lt;\u0026thinsp;0.05). Afterwards, DEGs in DN were screened by intersecting upregulated and downregulated differential genes in DEG1 and DEG2, separately.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3 Enrichment analysis of DEGs as well as screening of candidate genes\u003c/h2\u003e \u003cp\u003eFor the sake of exploring the biological functions as well as latent pathways of DEGs, Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genome (KEGG) enrichment analyses were executed with the aid of \u0026lsquo;clusterProfiler\u0026rsquo; package in R (version 4.7.1.003) (p value\u0026thinsp;\u0026lt;\u0026thinsp;0.05)\u003csup\u003e13\u003c/sup\u003e. Moreover, in order to probe into the interactions of DEGs at the protein level, a protein-protein interaction (PPI) network of DEGs was synthesized and visualized by feat of the Search Tool for the Retrieval of Interacting Genes database (STRING, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://cn.string-db.org/\u003c/span\u003e\u003cspan address=\"https://cn.string-db.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) and Cytoscape software (confidence value\u0026thinsp;=\u0026thinsp;0.4), respectively. Besides, the correlations between \u003cem\u003eNEBLs\u003c/em\u003e and DEGs were analyzed by Spearman\u0026rsquo;s correlation analysis to screen differentially expressed \u003cem\u003eNEBL\u003c/em\u003e (DE-\u003cem\u003eNEBL\u003c/em\u003e) for subsequent analysis (|r|\u0026gt;0.8, p value\u0026thinsp;\u0026lt;\u0026thinsp;0.05). Ulteriorly, significantly expressed DE-\u003cem\u003eNEBL\u003c/em\u003e both in DN versus normal samples and high versus low expression cohorts were identified as candidate genes by expression analysis in GSE30528 and GSE47183 (p value\u0026thinsp;\u0026lt;\u0026thinsp;0.05).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e2.4 Screening of biomarkers in DN\u003c/h2\u003e \u003cp\u003eThree kinds of machine learning (ML) methods were executed to seek the biomarkers that could accurately diagnose the onset of DN. Firstly, a Support Vector Machine-Recursive Feature Elimination (SVM-RFE) algorithm was executed with R package \u0026lsquo;e1071\u0026rsquo; (version 1.7\u0026ndash;13)\u003csup\u003e13\u003c/sup\u003e to acquire feature genes on the basis of candidate genes. Meanwhile, feature genes were also obtained from candidate genes by using Boruta algorithm with the \u0026lsquo;Boruta\u0026rsquo; package in R (version 8.0.0) \u003csup\u003e14\u003c/sup\u003e. In addition, the least absolute shrinkage and selection operator (LASSO) regression analysis was proceeded with the aid of \u0026lsquo;glmnet\u0026rsquo; package in R (version 4.1-8) \u003csup\u003e15\u003c/sup\u003e to identify feature genes. Afterwards, biomarkers for DN were identified by overlapping feature genes derived from the three ML algorithms. Finally, to further evaluate the ability of \u003cem\u003eNEBL\u003c/em\u003e and biomarkers to distinguish DN samples from control samples, the R package \u0026lsquo;pROC\u0026rsquo; (version 1.7\u0026ndash;11) was used to draw receiver operating characteristic (ROC) curve of them in GSE30528 and GSE47183 respectively\u003csup\u003e16\u003c/sup\u003e, and the area under curve (AUC) of ROC was calculated.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e2.5 Functional analysis of \u003cem\u003eNEBL\u003c/em\u003e and biomarkers\u003c/h2\u003e \u003cp\u003eIn order to further probe into the biological functions of \u003cem\u003eNEBL\u003c/em\u003e and biomarkers in DN, the functionally similar genes with \u003cem\u003eNEBL\u003c/em\u003e and biomarkers were primarily mined with the aid of GeneMANIA database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://genemania.org/\u003c/span\u003e\u003cspan address=\"http://genemania.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). The correlations between \u003cem\u003eNEBL\u003c/em\u003e and biomarkers were then analyzed by Spearman\u0026rsquo;s correlation analysis (|Cor|\u0026gt;0.3, p value\u0026thinsp;\u0026lt;\u0026thinsp;0.05). Furthermore, gene set enrichment analysis (GSEA) was carried out via the R package \u0026lsquo;clusterProfiler\u0026rsquo; (version 4.7.1.003) to explore biological functions and latent pathways related with \u003cem\u003eNEBL\u003c/em\u003e and biomarkers (p value\u0026thinsp;\u0026lt;\u0026thinsp;0.05)\u003csup\u003e13\u003c/sup\u003e. The correlation between \u003cem\u003eNEBL\u003c/em\u003e, biomarkers as well as other genes in GSE30528 was calculated and ranked according to their correlation coefficients. Subsequently, the KEGG signalling pathway set acquired from the Molecular Signatures Database (MSigDB, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.gsea-msigdb.org/\u003c/span\u003e\u003cspan address=\"https://www.gsea-msigdb.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) was invoked as a background set to perform GSEA.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e2.6 Characterization of immune infiltration in DN\u003c/h2\u003e \u003cp\u003eThe immune cell infiltration landscap in DN was characterized to investigate the effect of \u003cem\u003eNEBL\u003c/em\u003e and biomarkers on immune cell infiltration in DN. Firstly, based on dataset GSE30528, single-sample gene set enrichment analysis (ssGSEA) was utilized by \u0026lsquo;GSVA\u0026rsquo; package in R (version 1.46.0) to severally estimate the immune cells enrichment score of 28 immune cells types in DN as well as control group\u003csup\u003e17\u003c/sup\u003e, and the distinctions of immune cell scores between two groups were compared. In addition, the relationships between \u003cem\u003eNEBL\u003c/em\u003e and differential immune cells were computed by Spearman\u0026rsquo; correlation analysis, as well as the correlations between biomarkers and differential immune cells (|Cor|\u0026gt;0.3, p value\u0026thinsp;\u0026lt;\u0026thinsp;0.05)\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e2.7 Contruction of molecular regulatory networks\u003c/h2\u003e \u003cp\u003eA lncRNA-miRNA-mRNA network was synthesized with the aim of profoundly researching the latent mechanism of \u003cem\u003eNEBL\u003c/em\u003e and biomarkers in the occurrence and process of DN. Firstly, the critical miRNAs targeting \u003cem\u003eNEBL\u003c/em\u003e and biomarkers were predicted by intersecting miRNAs acquired from miRDB database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://mirdb.org/\u003c/span\u003e\u003cspan address=\"https://mirdb.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), miRWalk database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://129.206.7.150/\u003c/span\u003e\u003cspan address=\"http://129.206.7.150/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), and Targetscan database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.targetscan.org/vert_80/\u003c/span\u003e\u003cspan address=\"https://www.targetscan.org/vert_80/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). Subsequently, lncRNAs targetting critical miRNAs were screened by Starbase (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://rnasysu.com/encori/\u003c/span\u003e\u003cspan address=\"https://rnasysu.com/encori/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) (clipExpNum\u0026thinsp;\u0026gt;\u0026thinsp;15). Besides, for the sake of exploring whether \u003cem\u003eNEBL\u003c/em\u003e and biomarkers had an effect on DN at the molecular level, the analysis of transcription factors (TFs) targetting \u003cem\u003eNEBL\u003c/em\u003e and biomarkers were indispensable. The JASPAR database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://ngdc.cncb.ac.cn/databasecommons/database/id/176\u003c/span\u003e\u003cspan address=\"https://ngdc.cncb.ac.cn/databasecommons/database/id/176\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) was employed to search TFs targetting \u003cem\u003eNEBL\u003c/em\u003e and biomarkers. The study of drugs targeting \u003cem\u003eNEBL\u003c/em\u003e and biomarkers might provide new strategies for the treatment of DN, consequently, the CTD database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://ctdbase.org/\u003c/span\u003e\u003cspan address=\"http://ctdbase.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) were utilized to predict drugs related with \u003cem\u003eNEBL\u003c/em\u003e and biomarkers. Finally, the lncRNA-miRNA-mRNA, TF-gene and drug-gene networks of \u003cem\u003eNEBL\u003c/em\u003e and biomarkers were structured via Cytoscape software (version 3.7.1) \u003csup\u003e18\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e2.8 Statistical Analysis\u003c/h2\u003e \u003cp\u003eIn this study, the R software was engaged in statistical analysis. Differences between two groups were analyzed via the Wilcoxon rank-sum test, and p value of less than 0.05 was represented a significant difference.\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Results","content":"\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e3.1 A total of 157 DEGs were mined\u003c/h2\u003e \u003cp\u003eThe expression of \u003cem\u003eNEBL\u003c/em\u003e was significantly downregulated in DN group of both GSE30528 and GSE47183 datasets (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). A number of 424 DEGs1 were obtained between DN and control groups, including 113 up-regulated and 311 down-regulated genes (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA-B). A total of 176 DEGs2 were screened, the numbers of upregulation and downregulation were 56 and 120, respectively (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eC-D). Totally 157 DEGs covering 41 up-regulated and 116 down-regulated genes were obtained by intersecting the DEGs1 with DEGs2 (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eE).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e3.2 A sum of 19 candidate genes were excavated\u003c/h2\u003e \u003cp\u003eGO and KEGG were used to analyse DEGs with the aim of delving into their potential biological functions and signaling pathways. GO analysis indicated that DEGs were enriched to 237 entries, encompassing four cellular components (CCs), 12 molecular functions (MFs) and 226 biological process (BPs). To be specific, DEGs were associated with mesonephric epithelium development, mesonephros development and renal system development (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA). With respect to seven KEGG pathway involved by DEGs, AGE\u0026thinsp;\u0026minus;\u0026thinsp;RAGE signaling pathway in diabetic complications, Rap1 signaling pathway and complement and coagulation cascades and other pathways were notable (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eB). In order to investigate the interactions of DEGs at the protein level, a PPI network of DEGs was constructed, which contained 80 nodes and 202 edges. Thereinto, the interaction pairs with greater co-expression included COL1A2-LUM, SPP1-COL1A2 and COL1A2-COL6A3 (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eC). A sum of 64 DEGs strongly associated with \u003cem\u003eNEBL\u003c/em\u003e were noted as DE-NEBL by Spearman\u0026rsquo;s correlation analysis. The result manifested that 26 DE-\u003cem\u003eNEBL\u003c/em\u003e expressed discrepantly between DN and control samples in GSE30528 and GSE47183 datasets (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eD-E), and a number of 31 DE-\u003cem\u003eNEBL\u003c/em\u003e kept different expression levels between high and low expression groups (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eF-G). To sum up, a total of 19 DE-\u003cem\u003eNEBL\u003c/em\u003e were were counted as candidate genes as a result of their discrepant expression levels in DN and control groups, as well as high and low expression groups.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003e3.3 Totally four biomarkers were identified\u003c/h2\u003e \u003cp\u003eThree ML methods were exploited to screen biomarkers among candidate genes obtained above. Firstly, a total of 15 feature genes were acquired via SVM-RFE algorithm (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA). Meanwhile, Boruta algorithm indicated that all the 19 candidate genes obtained above were retained and ranked according to their importance on the basis of their importance (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB). Additionally, LASSO regression analysis was performed to screen out four feature genes (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eC-D). As a result, totally four biomarkers (\u003cem\u003eMPP5, TGFBR3, PCMTD2\u003c/em\u003e and \u003cem\u003eC1orf21\u003c/em\u003e) were screened via intersecting the feature genes acquired above (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eE). Whereafter, ROC curves were drawn to evaluate the diagnostic performance of \u003cem\u003eNEBL\u003c/em\u003e as well as four biomarkers for DN, and the results manifested that the AUC values of ROC curves were all greater than 0.7 in GSE30528, the validation results in dataset GSE47183 were consistent with the above results, implying the diagnostic performance of \u003cem\u003eNEBL\u003c/em\u003e and four biomarkers was fine (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eF-G).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e3.4 \u003cem\u003eNEBL\u003c/em\u003e and biomarkers were significantly enriched in immune-related pathways\u003c/h2\u003e \u003cp\u003eThe functionally alike genes with \u003cem\u003eNEBL\u003c/em\u003e and biomarkers were mined and a gene-gene interaction (GGI) network showing the interactions among them was synthesized. \u003cem\u003eNEBL\u003c/em\u003e and four biomarkers were associated with similar genes through co-localization, genetic interactions and physical interactions, etc. Additionally, gene pairs of \u003cem\u003eNEBL-PATJ\u003c/em\u003e, \u003cem\u003eC1orf21-TGFB1\u003c/em\u003e and \u003cem\u003eMPP5-CRB3\u003c/em\u003e were included in the network (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA). In addition, NEBL and biomarkers were mainly related to cellular response to transforming growth factor beta stimulus, transmembrane receptor protein serine/threonine kinase signaling pathway and tight junction organization, etc. Synchronously, correlation analysis revealed that there was a remarkable positive correlation between \u003cem\u003eNEBL\u003c/em\u003e and \u003cem\u003eMPP5\u003c/em\u003e (r\u0026thinsp;=\u0026thinsp;0.874, p value\u0026thinsp;\u0026lt;\u0026thinsp;0.001), \u003cem\u003eTGFBR3\u003c/em\u003e (r\u0026thinsp;=\u0026thinsp;0.844, p value\u0026thinsp;\u0026lt;\u0026thinsp;0.001), \u003cem\u003ePCMTD2\u003c/em\u003e (r\u0026thinsp;=\u0026thinsp;0.836, p value\u0026thinsp;\u0026lt;\u0026thinsp;0.001) as well as \u003cem\u003eC1orf21\u003c/em\u003e (r\u0026thinsp;=\u0026thinsp;0.802, p value\u0026thinsp;\u0026lt;\u0026thinsp;0.001) (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eB). Then, pathways involved in \u003cem\u003eNEBL\u003c/em\u003e and biomarkers were analyzed, the result revealed that \u003cem\u003eNEBL\u003c/em\u003e and biomarkers were observably enriched in ribosome, intestinal immune network for immunoglobulin A (IgA) production and systemic lupus erythematosus, etc. (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eC-G) (p value\u0026thinsp;\u0026lt;\u0026thinsp;0.05).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003e3.5 \u003cem\u003eNEBL\u003c/em\u003e and biomarkers might affect DN by altering the expression of 12 immune cells\u003c/h2\u003e \u003cp\u003eSince \u003cem\u003eNEBL\u003c/em\u003e and biomarker function were relevant to immunologically, immune infiltration analysis was proceed, and the abundance of 28 immune cells for each sample in GSE30528 was presented on heatmap (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eH). Totally 12 immune cells appeared discrepant immune infiltration between DN and control groups, such as CD56dim natural killer (NK) cells, regulatory T cells (Tregs) and mast cells, etc. And the contents of them were significantly increased in DN group (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eI). Moreover, CD56dim NK cells, mast cells, myeloid-derived suppressor cells (MDSCs), Tregs and follicular T helper cells were negatively correlated with \u003cem\u003eNEBL\u003c/em\u003e and biomarkers among the 12 discrepant immune cells, there was no significantly correlated between T helper Type 1 cells and \u003cem\u003eNEBL\u003c/em\u003e as well as biomarkers (p\u0026thinsp;\u0026gt;\u0026thinsp;0.05). In addition, CD56dim NK cells maintained the strongest correlation with \u003cem\u003eNEBL\u003c/em\u003e (r=-0.718, p value\u0026thinsp;\u0026lt;\u0026thinsp;0.001), \u003cem\u003eMPP5\u003c/em\u003e (r=-0.752, p value\u0026thinsp;\u0026lt;\u0026thinsp;0.001), \u003cem\u003eTGFBR3\u003c/em\u003e (r=-0.787, p value\u0026thinsp;\u0026lt;\u0026thinsp;0.001), \u003cem\u003ePCMTD2\u003c/em\u003e (r=-0.674, p value\u0026thinsp;\u0026lt;\u0026thinsp;0.001) and \u003cem\u003eC1orf21\u003c/em\u003e (r=-0.804, p value\u0026thinsp;\u0026lt;\u0026thinsp;0.001) (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eJ).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003e3.6 The lncRNA-miRNA-mRNA, TF-gene and drug-gene networks of NEBL and biomarkers were constructed\u003c/h2\u003e \u003cp\u003eA total of 91 critical miRNAs targetting \u003cem\u003eNEBL\u003c/em\u003e and biomarkers were predicted with the aid of three databases (\u003cem\u003eTGFBR3\u003c/em\u003e: 24, \u003cem\u003ePCMTD2\u003c/em\u003e: 25, \u003cem\u003eNEBL\u003c/em\u003e: 9, \u003cem\u003eC1orf21\u003c/em\u003e: 6 \u003cem\u003eand M\u003c/em\u003ePP5: 27) (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eA-E). A number of 103 lncRNAs were acquired based on critical miRNAs (clipExpNum\u0026thinsp;\u0026gt;\u0026thinsp;15). The ceRNA network was composed of five miRNAs (\u003cem\u003eNEBL, MPP5, TGFBR3, PCMTD2\u003c/em\u003e and \u003cem\u003eC1orf21\u003c/em\u003e), 88 miRNAs (hsa-miR-130a-5p, hsa-miR-613 and hsa-miR-501-5p, etc.), and 10 lncRNAs (XIST, NEAT1 and MALAT1, etc.), among which the interaction pairs included \u003cem\u003eTGFBR3\u003c/em\u003e-HSA-Mir-130A-5P-AC010980.2, \u003cem\u003eMPP5\u003c/em\u003e-hsa-miR-199b-5p-LINC01783, \u003cem\u003eNEB\u003c/em\u003eL-hsa-miR-497-5p-XIST, etc. (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eF). Ulteriorly, TFs targeting \u003cem\u003eNEBL\u003c/em\u003e and \u003cem\u003eMPP5\u003c/em\u003e, \u003cem\u003eTGFBR3\u003c/em\u003e, \u003cem\u003ePCMTD2 w\u003c/em\u003eere predicted in addition to \u003cem\u003eC1orf21\u003c/em\u003e. As presented in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eG, TFs that co-target \u003cem\u003ePCMTD2\u003c/em\u003e and \u003cem\u003eNEBL\u003c/em\u003e were POU2F2, STAT3 and TFAP2A, etc.. \u003cem\u003eGFBR3\u003c/em\u003e and \u003cem\u003eNEBL\u003c/em\u003e were regulated jointly by SREBF1, CEBPB, FOS and FOXC1, etc.. Besides, TFs such as USF2, NFIC and TEAD1 jointly targeted to \u003cem\u003eNEBL\u003c/em\u003e and \u003cem\u003eMPP5\u003c/em\u003e. With the respect of drug prediction, compounds interacted with \u003cem\u003eNEBL\u003c/em\u003e and biomarkers added up to 210, among which 42 compounds (Reference Count\u0026thinsp;\u0026gt;\u0026thinsp;1) were regarded as drugs to construct the gene-drug network (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eH). The network was consisted of 31 nodes and 42 edges. It was worth noiticted that \u003cem\u003eNEBL, PCMTD2\u003c/em\u003e and \u003cem\u003eTGFBR3\u003c/em\u003e were jointly targeted by bisphenol A, and the common targets of Valproic Acid were \u003cem\u003ePCMTD2, TGFBR3, NEBL\u003c/em\u003e and \u003cem\u003eC1ORF21\u003c/em\u003e, meaning that bisphenol A and Valproic Acid might provide reference for the treatment of DN.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eThe onset of DN is instigated by a confluence of various factors, nevertheless, the specific mechanisms of this phenomenon are yet to be elucidated\u003csup\u003e19, 20\u003c/sup\u003e. Currently, the early diagnosis and treatment of DN present challenges, necessitating the exploration of novel molecular pathways to solving this problem. Based on prior studies and experiments, the onset of DN, to some extent, is associated with immune infiltration, immune mediators, and oxidative stress\u003csup\u003e21\u0026ndash;23\u003c/sup\u003e. The \u003cem\u003eNEBL\u003c/em\u003e gene, located on chromosome 10q22.2, codes for a protein called Nebulette, which has traditionally been linked to heart function. However, recent studies have revealed its relationship with nephropathy, arousing curiosity about its potential significance in DN\u003csup\u003e24\u003c/sup\u003e. This study involved the retrieval of transcriptomic datasets from the GEO database, employing bioinformatics methodologies to pinpoint differentially expressed genes pivotal in DN. Through ML-based screening, four NEBL-related genes, \u003cem\u003eMPP5, TGFBR3, PCMTD2\u003c/em\u003e, and \u003cem\u003eC1orf21\u003c/em\u003e, were identified. ROC analysis showed that the NEBL gene and four \u003cem\u003eNEBL\u003c/em\u003e-related genes possessed the btility of distinguishing DN samples and controls, suggesting that they might be developed as diagnostic biomarkers. In addition, the functions of \u003cem\u003eNEBL\u003c/em\u003e genes and biomarkers were investigated with the aim of understanding their potential molecular mechanisms of them on occurrence and development of DN.\u003c/p\u003e \u003cp\u003eMembrane Protein Palmitoylated 5 (\u003cem\u003eMPP5\u003c/em\u003e), also recognized as Protein Associated with Lin7 (\u003cem\u003ePALS1\u003c/em\u003e), functions as a fundamental constituent of the apical membrane determining CRB complex within the nephron, and the absence of Pals1 leads to tubular dilation and cyst formation\u003csup\u003e25\u003c/sup\u003e. However, there are a scarcity of studies investigating the role of \u003cem\u003eMPP5\u003c/em\u003e in the development of DN. In our study, we observed that down-regulated \u003cem\u003eMPP5\u003c/em\u003e also had better diagnostic efficacy as a result of its AUC value of 1.\u003c/p\u003e \u003cp\u003e \u003cem\u003eTGFBR3\u003c/em\u003e, a membrane proteoglycan, serves as a co-receptor with other transforming growth factor receptors and is expressed in glomerular podocytes, mesangial cells, and endothelial cells\u003csup\u003e26, 27\u003c/sup\u003e. With a high specificity for the glomerulus, \u003cem\u003eTGFBR3\u003c/em\u003e exhibits a particular affinity for two subtypes of TGF-β, forming specific binding interactions A decrease in TGF-β secretion triggers cytokines to stimulate T helper1 (Th1) cells, leading to the production of macrophage M1 and fostering the inflammatory response of Th1 cells, which is an essential factor in the progression of DN. The study suggested that a potential therapeutic target for DN involving intervention with TGF-β could be \u003cem\u003eTGFBR3\u003c/em\u003e\u003csup\u003e28\u003c/sup\u003e. Additionally, we observed elevated levels of TGFBR3 in DN patients, demonstrating diagnostic accuracy values with AUC exceeding 0.70. The outcomes from GeneMANIA indicated that \u003cem\u003eTGFBR3, TGF-β1, TGF-β2\u003c/em\u003e, and \u003cem\u003eTGF-β3\u003c/em\u003e might collaboratively influence the progression of DN through physical interactions. This reinforces the credibility of the findings in this study.\u003c/p\u003e \u003cp\u003eProtein-L-Isoaspartate O-Methyltransferase Domain Containing 2 (\u003cem\u003ePCMTD2\u003c/em\u003e) is anticipated to possess protein-L-isoaspartate (D-aspartate) O-methyltransferase activity and play a role in protein methylation, and its predicted activity is situated in the cytoplasm. Chromosome 1 Open Reading Frame 21 (\u003cem\u003eC1orf21\u003c/em\u003e) is classified as a protein-coding gene. Diseases linked to \u003cem\u003eC1orf21\u003c/em\u003e encompass thymoma type B2 and renal artery atheroma. \u003cem\u003ePCMTD2\u003c/em\u003e and \u003cem\u003eC1orf21\u003c/em\u003e, particularly in the context of kidney disease, have limited existing knowledge. However, our study heightened the enhanced expression levels of \u003cem\u003ePCMTD2\u003c/em\u003e and \u003cem\u003eC1orf21\u003c/em\u003e in DN patients compared to controls. Considering the limited available literature regarding the influence of \u003cem\u003ePCMTD2\u003c/em\u003e and \u003cem\u003eC1orf21\u003c/em\u003e on the progression of DN, it is prudent to make inference based on the results of this study, generally, increased gene expression levels is believed to be linked with an elevated disease risk. The study's findings revealed increased expression of \u003cem\u003ePCMTD2\u003c/em\u003e and \u003cem\u003eC1orf21\u003c/em\u003e in DN patients, suggesting that these genes might serve as potential risk factors for DN. However, further analysis and experimental validation are requisite to substantiate this hypothesis.\u003c/p\u003e \u003cp\u003eGO and KEGG enrichment analysis found that DEGs mainly enriched in BPs related to kidney development, indicating that these DEGs might be related to the development of DN, and this result further verified the reliability of DEGs. In our analysis, we found that the AGE-RAGE signaling pathway played an important role in DN. Previous studies have demonstrated that the presence and expression of AGE and RAGE have increased in human diabetic kidneys, especially in the glomerulus, glomerular epithelial cells (podocytes) and endothelial cells\u003csup\u003e29\u003c/sup\u003e. The absence of RAGE has been found to effectively mitigate glomerular sclerosis, glomerular basement membrane thickening, podocyte loss, as well as decline in glomerular filtration rate (GFR). In addition, RAGE is also present in the angiotensin II (angII) axis, which is also associated with the expression of RAGE and the occurrence of ND\u003csup\u003e30\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eGSEA revealed that the \u003cem\u003eNEBL\u003c/em\u003e gene demonstrated enrichment in immune-related pathways. Additionally, GSEA demonstrated significant enrichment of the \u003cem\u003eNEBL\u003c/em\u003e gene and four biomarkers in pathways associated with ribosomes, the IgA-produced intestinal immune network, systemic lupus erythematosus, primary immunodeficiency, leishmania infection, cell adhesion molecules (CAM), glutathione metabolism, and others. Besides, a significant enrichment of biomarkers, such as the \u003cem\u003eNEB\u003c/em\u003eL and \u003cem\u003eTGFBR3\u003c/em\u003e, in pathways associated with immunity. In conclusion, we infered that \u003cem\u003eNEBL\u003c/em\u003e and \u003cem\u003eTGFBR3\u003c/em\u003e might play similar immune functions in the development of DN.\u003c/p\u003e \u003cp\u003eThe findings of immune infiltration analysis elucidated a close association between the development of DN and CD56 NK cells, mast cells, and regulatory T cells. According to previous studies, Natural Killer T (NKT) cells can be found to be associated with kidney damage as well as vascular damage in type 2 diabetes, but few studies have focused on NKT cells and DN, especially CD56 NK cells, which would be a new focus on DN\u003csup\u003e31\u0026ndash;33\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eDrug prediction revealed two compounds that were most closely targeted towards NEBL gene and biomarkers, bisphenol A (BPA) and Valproic Acid. BPA demonstrates an association with low-grade albuminuria, a manifestation characteristic of the initial phases of DN\u003csup\u003e34\u003c/sup\u003e. Moreover, findings certified that valproic acid mitigated the diabetes-induced upregulation of complement C5a receptors, concurrently diminishing indicators of cellular senescence and the senescence-associated secretory phenotype. The attenuation of cellular senescence in the diabetic context assumes significance, given its implication in the pathogenesis of diabetic kidney disease\u003csup\u003e35\u003c/sup\u003e. Consequently, therapeutic interventions targeting cellular senescence, such as complement inhibitors, emerge as a novel and promising avenue for the treatment of diabetic kidney disease. Based on current research, the reduction of \u003cem\u003eNEBL\u003c/em\u003e in DN may exacerbate the condition. We hypothesize that BPA may reverse the expression of \u003cem\u003eNEBL\u003c/em\u003e in DN, thereby potentially alleviating the progression of DN.\u003c/p\u003e \u003cp\u003eNevertheless, it is crucial to acknowledge the limitations of this study. The evidence relies on publicly available data, and while we validated the expression using another dataset, additional experiments are imperative to substantiate \u003cem\u003eNEBL\u003c/em\u003e and the four associated genes as diagnostic markers before their clinical applicability can be established. At the same time, the clinical application research of drugs also needs experimental proof, and we will continue to pay attention to the relevant research progress of DN.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eFunding:\u003c/h2\u003e \u003cp\u003eThis work was supported by the National Natural Science Foundation of China to Xuhua Ge (Nos. 82170733).\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eYT performed the data analyses and wrote the manuscript; SX contributed to the analysis andmanuscript preparation; XG contributed to the conception of the study and helped perform theanalysis with constructive discussions. All authors participated in the conference workgroups, thedevelopment of the summary statement, and the review of the manuscript. All authors read andapproved the final manuscript.\u003c/p\u003e\u003ch2\u003eAcknowledgments:\u003c/h2\u003e \u003cp\u003eI really appreciate Dr.Xuhua Ge of Children\u0026rsquo;s Hospital of Nanjing Medical University (Nanjing, China) for her guidance on the paper and her assistance with the bioinformatics analysis.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eEsham M, Bryan K, Milnes J, Holmes WB, Moncman CL. Expression of nebulette during early cardiac development. Cell Motil Cytoskeleton 2007, 64: 258-73.\u003c/li\u003e\n\u003cli\u003eYang C, Wang H, Zhao X, Matsushita K, Coresh J, Zhang L, Zhao MH. CKD in China: Evolving Spectrum and Public Health Implications. Am J Kidney Dis 2020, 76: 258-64.\u003c/li\u003e\n\u003cli\u003ePalsson R, Patel UD. Cardiovascular complications of diabetic kidney disease. Adv Chronic Kidney Dis 2014, 21: 273-80.\u003c/li\u003e\n\u003cli\u003eXin C, Sun X, Li Z, Gao T. Relationship of Soluble Klotho and Early Stage of Diabetic Nephropathy: A Systematic Review and Meta-Analysis. Front Endocrinol (Lausanne) 2022, 13: 902765.\u003c/li\u003e\n\u003cli\u003eViigimaa M, Sachinidis A, Toumpourleka M, Koutsampasopoulos K, Alliksoo S, Titma T. Macrovascular Complications of Type 2 Diabetes Mellitus. Curr Vasc Pharmacol 2020, 18: 110-6.\u003c/li\u003e\n\u003cli\u003eAlicic R, Nicholas SB. Diabetic Kidney Disease Back in Focus: Management Field Guide for Health Care Professionals in the 21st Century. Mayo Clin Proc 2022, 97: 1904-19.\u003c/li\u003e\n\u003cli\u003eWang L, Zhang Y, Liu S, Huang N, Zeng W, Xu W, Zhou T, et al. Comparison of Anti-Microbic and Anti-Biofilm Activity Among Tedizolid and Radezolid Against Linezolid-Resistant Enterococcus faecalis Isolates. Infect Drug Resist 2021, 14: 4619-27.\u003c/li\u003e\n\u003cli\u003eLopez-Giacoman S, Madero M. Biomarkers in chronic kidney disease, from kidney function to kidney damage. World J Nephrol 2015, 4: 57-73.\u003c/li\u003e\n\u003cli\u003eBonzo JR, Norris AA, Esham M, Moncman CL. The nebulette repeat domain is necessary for proper maintenance of tropomyosin with the cardiac sarcomere. Exp Cell Res 2008, 314: 3519-30.\u003c/li\u003e\n\u003cli\u003ePurevjav E, Varela J, Morgado M, Kearney DL, Li H, Taylor MD, Arimura T, et al. Nebulette mutations are associated with dilated cardiomyopathy and endocardial fibroelastosis. J Am Coll Cardiol 2010, 56: 1493-502.\u003c/li\u003e\n\u003cli\u003eMaiellaro-Rafferty K, Wansapura JP, Mendsaikhan U, Osinska H, James JF, Taylor MD, Robbins J, et al. Altered regional cardiac wall mechanics are associated with differential cardiomyocyte calcium handling due to nebulette mutations in preclinical inherited dilated cardiomyopathy. J Mol Cell Cardiol 2013, 60: 151-60.\u003c/li\u003e\n\u003cli\u003eLiu S, Xie X, Lei H, Zou B, Xie L. Identification of Key circRNAs/lncRNAs/miRNAs/mRNAs and Pathways in Preeclampsia Using Bioinformatics Analysis. Med Sci Monit 2019, 25: 1679-93.\u003c/li\u003e\n\u003cli\u003eYu G, Wang LG, Han Y, He QY. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS 2012, 16: 284-7.\u003c/li\u003e\n\u003cli\u003eShubin NJ, Navalkar K, Sampson D, Yager TD, Cermelli S, Seldon T, Sullivan E, et al. Serum Protein Changes in Pediatric Sepsis Patients Identified With an Aptamer-Based Multiplexed Proteomic Approach. Crit Care Med 2020, 48: e48-e57.\u003c/li\u003e\n\u003cli\u003eWu X, Qin K, Iroegbu CD, Xiang K, Peng J, Guo J, Yang J, et al. Genetic analysis of potential biomarkers and therapeutic targets in ferroptosis from coronary artery disease. J Cell Mol Med 2022, 26: 2177-90.\u003c/li\u003e\n\u003cli\u003eRobin X, Turck N, Hainard A, Tiberti N, Lisacek F, Sanchez JC, Muller M. pROC: an open-source package for R and S+ to analyze and compare ROC curves. BMC Bioinformatics 2011, 12: 77.\u003c/li\u003e\n\u003cli\u003eRen Z, He Y, Yang Q, Guo J, Huang H, Li B, Wang D, et al. A Comprehensive Analysis of the Glutathione Peroxidase 8 (GPX8) in Human Cancer. Front Oncol 2022, 12: 812811.\u003c/li\u003e\n\u003cli\u003eCosta-Silva J, Domingues D, Lopes FM. RNA-Seq differential expression analysis: An extended review and a software tool. PLoS One 2017, 12: e0190152.\u003c/li\u003e\n\u003cli\u003eNatesan V, Kim SJ. Diabetic Nephropathy - a Review of Risk Factors, Progression, Mechanism, and Dietary Management. Biomol Ther (Seoul) 2021, 29: 365-72.\u003c/li\u003e\n\u003cli\u003eHu F, Zhang T. Study on Risk Factors of Diabetic Nephropathy in Obese Patients with Type 2 Diabetes Mellitus. Int J Gen Med 2020, 13: 351-60.\u003c/li\u003e\n\u003cli\u003eYu K, Li D, Xu F, Guo H, Feng F, Ding Y, Wan X, et al. IDO1 as a new immune biomarker for diabetic nephropathy and its correlation with immune cell infiltration. Int Immunopharmacol 2021, 94: 107446.\u003c/li\u003e\n\u003cli\u003eKahles F, Meyer C, Mollmann J, Diebold S, Findeisen HM, Lebherz C, Trautwein C, et al. GLP-1 secretion is increased by inflammatory stimuli in an IL-6-dependent manner, leading to hyperinsulinemia and blood glucose lowering. Diabetes 2014, 63: 3221-9.\u003c/li\u003e\n\u003cli\u003eCelejewski-Marciniak P, Wolinowska R, Wroblewska M. Molecular Characterization of Class 1, 2 and 3 Integrons in Serratia spp. Clinical Isolates in Poland - Isolation of a New Plasmid and Identification of a Gene for a Novel Fusion Protein. Infect Drug Resist 2021, 14: 4601-10.\u003c/li\u003e\n\u003cli\u003eGe X, Zhang T, Yu X, Muwonge AN, Anandakrishnan N, Wong NJ, Haydak JC, et al. LIM-Nebulette Reinforces Podocyte Structural Integrity by Linking Actin and Vimentin Filaments. J Am Soc Nephrol 2020, 31: 2372-91.\u003c/li\u003e\n\u003cli\u003eWeide T, Vollenbroker B, Schulze U, Djuric I, Edeling M, Bonse J, Hochapfel F, et al. Pals1 Haploinsufficiency Results in Proteinuria and Cyst Formation. J Am Soc Nephrol 2017, 28: 2093-107.\u003c/li\u003e\n\u003cli\u003eChen S, Li B, Chen L, Jiang H. Identification and validation of immune-related biomarkers and potential regulators and therapeutic targets for diabetic kidney disease. BMC Med Genomics 2023, 16: 90.\u003c/li\u003e\n\u003cli\u003eYamamoto T, Watanabe T, Ikegaya N, Fujigaki Y, Matsui K, Masaoka H, Nagase M, et al. Expression of types I, II, and III TGF-beta receptors in human glomerulonephritis. J Am Soc Nephrol 1998, 9: 2253-61.\u003c/li\u003e\n\u003cli\u003eZhu Y, Usui HK, Sharma K. Regulation of transforming growth factor beta in diabetic nephropathy: implications for treatment. Semin Nephrol 2007, 27: 153-60.\u003c/li\u003e\n\u003cli\u003eRamasamy R, Yan SF, Schmidt AM. Receptor for AGE (RAGE): signaling mechanisms in the pathogenesis of diabetes and its complications. Ann N Y Acad Sci 2011, 1243: 88-102.\u003c/li\u003e\n\u003cli\u003eRuster C, Franke S, Wenzel U, Schmidthaupt R, Fraune C, Krebs C, Wolf G. Podocytes of AT2 receptor knockout mice are protected from angiotensin II-mediated RAGE induction. Am J Nephrol 2011, 34: 309-17.\u003c/li\u003e\n\u003cli\u003eWang H, Cao K, Liu S, Xu Y, Tang L. Tim-3 Expression Causes NK Cell Dysfunction in Type 2 Diabetes Patients. Front Immunol 2022, 13: 852436.\u003c/li\u003e\n\u003cli\u003eUchida T, Ito S, Kumagai H, Oda T, Nakashima H, Seki S. Roles of Natural Killer T Cells and Natural Killer Cells in Kidney Injury. Int J Mol Sci 2019, 20:\u003c/li\u003e\n\u003cli\u003eLv X, Gao Y, Dong T, Yang L. Role of Natural Killer T (NKT) Cells in Type II Diabetes-Induced Vascular Injuries. Med Sci Monit 2018, 24: 8322-32.\u003c/li\u003e\n\u003cli\u003eJiang W, Ding K, Huang W, Xu F, Lei M, Yue R. Potential effects of bisphenol A on diabetes mellitus and its chronic complications: A narrative review. Heliyon 2023, 9: e16340.\u003c/li\u003e\n\u003cli\u003eCoughlan MT, Ziemann M, Laskowski A, Woodruff TM, Tan SM. Valproic acid attenuates cellular senescence in diabetic kidney disease through the inhibition of complement C5a receptors. Sci Rep 2022, 12: 20278.\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":"Diabetic nephropathy, NEBL, Biomarkers, Bioinformatics analysis","lastPublishedDoi":"10.21203/rs.3.rs-4361592/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4361592/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003ePatients with diabetes had a significantly increased risk of cardiovascular disease by reporter, and the nebulette (\u003cem\u003eNEBL\u003c/em\u003e) gene were closely related with cardiovascular disease. However, the impact of the \u003cem\u003eNEBL\u003c/em\u003e gene on diabetic nephropathy (DN) and the underlying molecular mechanisms, have yet to be conclusively validated. Therefore, this study aims to mine \u003cem\u003eNEBL\u003c/em\u003e related biomarkers in DN by bioinformatics analysis. A total of 157 differentially expressed genes (DEGs) associated with DN and \u003cem\u003eNEBL\u003c/em\u003e gene were excavated, and they were associated with biological processes of mesonephric development and AGE-RAGE signaling pathway in diabetic complications. Besides, totally 19 candidate genes were screened by correlation and expression analyses, among which, \u003cem\u003eMPP5, TGFBR3, PCMTD2\u003c/em\u003e and \u003cem\u003eC1orf21\u003c/em\u003e related to \u003cem\u003eNEBL\u003c/em\u003e were deemed as biomarkers via Support Vector Machine-Recursive Feature Elimination (SVM-RFE), Boruta and least absolute shrinkage and selection operator (LASSO) algorithms, the ability of \u003cem\u003eNEBL\u003c/em\u003e and biomarkers to distinguish DN from controls was accurate. Immunoinfiltration certified that the contents of 12 differential immune cells were significantly increased in DN group. Then, the gene set enrichment analysis (GSEA) revealed that \u003cem\u003eNEBL\u003c/em\u003e and biomarkers were observably enriched in ribosome, intestinal immune network for immunoglobulin A (IgA) production and so on. Finally, drug-gene network revealed bisphenol A and Valproic Acid might be pivotal drugs regulating the expression of \u003cem\u003eNEBL\u003c/em\u003e and biomarkers.In this sutudy, totally four biomarkers (\u003cem\u003eMPP5, TGFBR3, PCMTD2\u003c/em\u003e and \u003cem\u003eC1orf21\u003c/em\u003e) related to \u003cem\u003eNEBL\u003c/em\u003e were screened by bioinformatic analysis, providing a novel reference for effective clinical diagnosis and treatment of DN.\u003c/p\u003e","manuscriptTitle":"Machine learning-based identification of biomarkers associated with NEBL in diabetic nephropathy","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-05-13 20:19:34","doi":"10.21203/rs.3.rs-4361592/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":"18937558-23f9-42e5-b671-7f062be742c5","owner":[],"postedDate":"May 13th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2024-05-20T10:42:59+00:00","versionOfRecord":[],"versionCreatedAt":"2024-05-13 20:19:34","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-4361592","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4361592","identity":"rs-4361592","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}