Unveiling the role of EGR1 and hub senescencerelated genes in Type II alveolar epithelial cells senescence for Obstructive sleep Apnea

Unveiling the role of EGR1 and hub senescencerelated genes in Type II alveolar epithelial cells senescence for Obstructive sleep Apnea | 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 Unveiling the role of EGR1 and hub senescencerelated genes in Type II alveolar epithelial cells senescence for Obstructive sleep Apnea Caili Li, Yuxiang Zhang, Xia Yang, Yubao Wang, Haiyan Zhao, Jing Feng This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6563621/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 19 Dec, 2025 Read the published version in BMC Biotechnology → Version 1 posted 11 You are reading this latest preprint version Abstract Backgroud : A frequently encountered breathing condition, obstructive sleep apnea (OSA) primarily manifests while sleeping and is characterized by total or incomplete blockage of the upper respiratory tract. This disorder disrupts normal airflow, often leading to repeated pauses in breathing throughout the night. Aging significantly increases the risk of OSA, yet the underlying biomolecular connections between aging and OSA remain incompletely understood. Methods : This research integrates bioinformatics and machine learning methods. To discover and confirm possible biomarkers, a combination of WGCNA and machine learning techniques was utilized. Functional characterization of genes was achieved through GO and KEGG enrichment studies, which provided insights into biological pathways and molecular roles. A predictive nomogram was developed based on hub hub OSA-ARDEGs.Comprehensive immune infiltration analysis was conducted to elucidate the immunological microenvironment associated with key biomarkers.To experimentally validate computational predictions, RNA-seq and Western blotting analyses were performed to confirm EGR1 expression patterns in human type II alveolar epithelial cells. Results : Investigative studies revealed genes exhibiting differential expression patterns, along with interconnected gene networks that showed notable associations with OSA. The analysis further demonstrated that these molecular networks are intricately tied to mechanisms of biological aging and immune system activity. Enrichment studies revealed that these genes are involved in multiple biological mechanisms, including processes related to inflammation and signaling cascades mediated by immune cells. Furthermore, EGR1 was validated experimentally as a critical gene involved in cellular senescence, immune regulation, and DNA damage response. Conclusion : Our findings establish EGR1 as a crucial mediator in OSA pathogenesis, potentially driving disease progression through cellular senescence mechanisms. These results position EGR1 as a promising molecular target for developing therapeutic interventions against OSA-associated respiratory dysfunction. Obstructive sleep apnea aging senescencerelated genes EGR1 Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 Figure 13 Figure 14 Figure 15 Figure 16 Figure 17 Figure 18 Figure 19 Figure 20 1. Introduction A prevalent breathing condition associated with sleep, obstructive sleep apnea (OSA) is marked by recurring total or partial obstructions in the upper airway while sleeping. These events trigger pronounced drops in blood oxygen levels and disruptions to sleep continuity [ 1 ] . OSA affects approximately 9–38% of adults worldwide, with prevalence rates demonstrating a positive correlation with advancing age and elevated body mass index [ 2 – 4 ] . Patients with OSA often exhibit various symptoms such as loud snoring, observed apneas, nocturnal arousals, morning headaches, and excessive daytime somnolence, which significantly degrade the quality of life. Individuals with OSA suffer from multiple symptoms that significantly impact their quality of life, including loud snoring, observed pauses in breathing, frequent waking, morning headaches, and excessive daytime sleepiness. These symptoms typically come with various neuropsychiatric issues like concentration difficulties, memory impairments, and mood fluctuations, all linked to the fragmented sleep and hypoxemia associated with the disorder [ 5 ] . OSA independently correlates with cardiovascular conditions, including hypertension, heart failure, atrial fibrillation, and stroke. It has been associated with metabolic conditions such as type 2 diabetes and dyslipidemia, while also heightening the likelihood of vehicular collisions resulting from daytime drowsiness [ 6 – 8 ] . Chronic untreated OSA leads to reduced neurocognitive function and overall life expectancy, highlighting the necessity for effective diagnostic and treatment approaches to alleviate these risks [ 9 ] Advancing age represents a major risk factor for OSA [ 10 , 11 ] . The relationship between OSA and aging holds significant clinical relevance, especially given the worldwide shift toward an aging population. Age-related physiological decline involves complex biological mechanisms marked by progressive weakening of cellular repair and homeostasis, heightening susceptibility to numerous age-associated pathologies [ 12 , 13 ] . Biomarkers like telomere attrition, inflammation, alterations in synthesis pathways, and hormonal dysregulation are core aspects of aging.These factors not only advance typical aging phenotypes but also impact diverse pathophysiological pathways. Regarding sleep and respiratory functions, aging correlates with increased neck fat accumulation, decreased airway muscle tone, and changes in lung and chest wall mechanics, all of which contribute to a higher prevalence of OSA in older individuals [ 14 ] . Aging alters sleep architecture, diminishing deep sleep phases and increasing nocturnal awakenings—changes that can worsen the severity and outcomes of OSA [ 15 ] . The shared mechanisms between aging and OSA include systemic inflammation, oxidative stress, and endothelial dysfunction. Aging impacts OSA by increasing vulnerability to the hypoxic challenges of OSA due to a decline in physiological reserves associated with aging. Aging also affects the neural control of upper airway muscles and respiratory drive during sleep, which may exacerbate the severity of OSA or complicate the management of elderly patients [ 16 ] . The interaction between aging and OSA is recognized, but the exact mechanisms by which aging regulates OSA pathogenesis via molecular networks remain unclear. This significantly hampers the development of targeted interventions for elderly OSA patients. This underscores the critical requirement for additional studies to comprehensively clarify these pathways [ 17 , 18 ] . The zinc finger-containing regulatory protein EGR1 (early growth response 1) is swiftly activated by external signals and orchestrates the regulation of target genes. It has been associated with various diseases, including acute kidney injury [ 19 ] and multiple malignancies such as gastric cancer [ 20 ] , pancreatic cancer [ 21 ] , and endometrial cancer [ 22 ] . However, its role in the pathogenesis of OSA remains to be clarified. To establish a gene co-expression network, WGCNA method was applied, which facilitated the detection of central OSA-associated hub genes. Differences in immune cell infiltration levels across individuals with OSA were assessed using ssGSEA. Meanwhile, GSVA was applied to compare functional and pathway variations between distinct OSA subtypes, further uncovering the molecular heterogeneity of OSA.Within machine learning applications, LASSO and Random Forest approaches were applied to identify critical variables and construct predictive models. Ultimately, key discoveries from bioinformatics analyses were validated in the lab through experimental validation steps. Through the application of these methods, the research delivers an in-depth exploration of the biological pathways linked to aging-associated OSA. By uncovering prospective biomarkers and targets for therapy, this work establishes a foundation for advancing diagnostic strategies and treatment approaches for the disorder. 2. Materials and Methods 2.1. Data Collection and Preprocessing This study utilized the OSA gene expression profile database, including GSE135917 [ 25 ] GSE38792 [ 26 ] , and GSE75097 [ 27 ] . The GSE135917 and GSE38792 datasets utilize the GPL6244 platform (Affymetrix Human Gene 1.0 ST Array for genome-wide expression profiling). In contrast, the GSE75097 dataset originated from the GPL10904 platform (Illumina HumanHT-12 V4.0 beadchip). Earlier research has shown that continuous positive airway pressure (CPAP) therapy in OSA patients markedly alters transcriptional activity [ 28 , 29 ] , with evidence suggesting restoration of gene expression patterns toward baseline levels after intervention. To minimize possible biases arising from CPAP-induced transcriptional alterations affecting OSA-related molecular signatures, 24 individuals receiving CPAP therapy were excluded from the GSE135917 cohort. These retained samples were then combined with data from the GSE38792 dataset for subsequent analysis. Technical batch effects were adjusted by applying the "removeBatchEffect" function from the limma R package, aiming to achieve consistent data quality between integrated datasets. The combined dataset comprised 16 healthy controls and 44 OSA patients. During analysis of the GSE75097 dataset, background adjustments and data normalization were performed via the neqc method available in the limma R package. 14 CPAP-treated OSA patients were excluded from the analysis to maintain consistency with our exclusion criteria. The final validation cohort consisted of 6 healthy controls and 28 OSA patients. The comprehensive data processing workflow is presented in Fig. 1 . 2.2 Differential Gene Expression Analysis To identify differentially expressed genes (DEGs) between OSA patients and healthy controls, the "limma" R package [ 30 ] was utilized for statistical analysis. The merged gene expression dataset was analyzed using stringent criteria, with |log2 fold change (logFC)| > 0.58 and false discovery rate (FDR) < 0.05, to identify OSA-related DEGs (OSA-DEGs). The results were visualized using volcano plots and heatmaps, which were generated with the "ComplexHeatmap" package and the "ggplot2" package [ 31 ] . 2.3. Weighted Gene Co-Expression Network Analysis This study applied a weighted gene co-expression network analytical framework (WGCNA) to pinpoint hub genes implicated in OSA and connected to biological aging mechanisms. A scale-free co-expression network was developed for genes exhibiting high variability within the training cohort, implemented through the WGCNA package in R [ 24 ] .Following initial quality assessment using the "goodSamplesGenes" method, an optimal soft threshold (β = 6) was selected via the "pickSoftThreshold" algorithm.The gene co-expression network was developed by initially computing an adjacency matrix based on Pearson correlation measures, which was later converted into a topological overlap matrix (TOM) to assess network interconnectedness. Functional modules within the network were identified via hierarchical clustering integrated with dynamic branch-cutting methodology, establishing a lower limit of 50 genes per module and a module consolidation cutoff set at 0.15 similarity. Statistical relationships between module eigengenes (MEs) and phenotypic characteristics were systematically evaluated, while key molecular candidates were selected through comparative analysis of intra-modular connectivity (MM) indices and biological relevance (GS) parameters. 2.4. Functional Enrichment Analysis To investigate the molecular roles of genes linked to obstructive sleep apnea (OSA), Gene Ontology (GO) and KEGG pathway enrichment analyses were conducted in R using the "clusterProfiler" package [ 31 ] .A Benjamini-Hochberg-adjusted p-value cutoff of < 0.05 was applied to define statistically significant terms. Enrichment results for GO categories, such as biological processes, cellular components, and molecular functions—were plotted as dot diagrams. KEGG pathway enrichment findings were further mapped with the ​"pathview" package [ 32 ] .This integrative approach yielded a holistic perspective on the biological mechanisms and pathway interactions involving OSA-associated genes. 2.5 Download and Organization of Aging-Related Genes Following established protocols [ 33 ] , A comprehensive collection of human aging-related genes (ARGs) was sourced from the Human Aging Genomic Resources (HAGR) database [ 34 ] . This dataset included 307 genes from the GenAge database and 279 genes from the CellAge database. Upon integration and de-duplication, a consolidated list of 543 unique ARGs was established. 2.6 Evaluation of Immune Infiltration in OSA Patients Enrichment levels for 19 immune cell populations and 13 immune-related processes across healthy and OSA cohorts were assessed via the "GSVA" package in R, with results visualized as violin-style plots using the "vioplot" R package [ 35 ] . To investigate linkages between OSA-ARDEGs (aging-associated differentially expressed genes in OSA) and immune activity, Spearman's correlation coefficient was applied to evaluate connections with both immune cell subsets and functional immunological pathways. 2.7 Unsupervised Clustering Analysis and Gene Set Variation Analysis Samples from OSA cohorts were clustered in an unsupervised manner utilizing the "ConsensusClusterPlus" toolkit in R [ 36 ] , employing the k-means algorithm across 1,000 resampling iterations with an upper cluster limit (k = 6). Optimal cluster partitioning (k = 2) was selected through systematic evaluation of agreement matrices, cumulative density function (CDF) curves, and partition robustness indices. To investigate functional divergence across molecular subgroups, Gene Set Variation Analysis (GSVA) was implemented. Pathway enrichment patterns were examined via the MSigDB repository, integrating canonical KEGG pathways and Gene Ontology-derived biological processes. A significance threshold of |GSVA score| > 2 (t-test) was established to identify differentially enriched pathways and immune microenvironment patterns. 2.8 Feature Selection, Modeling, and Validation Critical variables were identified via LASSO regression implemented in the "glmnet" R package. A prediction model was then constructed using the random forest method. The data were divided randomly, with 80% allocated for training and 20% reserved for validation. Within the LASSO framework, variables exhibiting non-zero coefficients were selected to build the random forest-based predictive signature. 2.9 Using Machine Learning (ML) Models to Identify Characteristic Biomarkers Six machine learning approaches—AdaBoost, Decision Trees (DT), k-Nearest Neighbors (KNN), LightGBM, Naïve Bayes, and XGBoost—were applied to build diagnostic models. The data were partitioned randomly, allocating 70% for model training and 30% for testing. AUC values served as the primary metric to evaluate the ability of OSA-associated DEGs to distinguish patients from healthy individuals. Model execution relied on multiple R packages: AdaBoost, DT, and KNN utilized the "caret" package; LightGBM leveraged "LightGBM" and "tidymodels"; XGBoost employed the "XGBoost" package; and Naïve Bayes was operationalized via "e1071". The "DALEX" package supported explainability assessments to improve model transparency. 2.10 Construction and Evaluation of Binary Logistic Regression (LR) Model The investigation utilized a binary logistic regression (LR) model to establish association metrics for aging-associated OSA-related differentially expressed genes (OSA-ARDEGs). Model calibration plots were created using the ​RMS package to assess predictive performance. The prognostic value of risk stratification metrics was assessed via decision curve analysis (DCA), calibration plots, and area under the curve (AUC) measurements. Model generalizability and robustness were evaluated using both internal and external validation cohorts. External validation utilized the GSE75097 transcriptomic dataset acquired from the Gene Expression Omnibus (GEO) repository, while a separate internal validation cohort comprising independent samples was established for additional verification. 2.11 Gene Set Enrichment Analysis (GSEA) This study performed gene set enrichment analysis (GSEA) on key OSA-ORDEGs to further investigate the functional roles of EGR1 and GLB1 in OSA pathogenesis. Pathway enrichment analysis of key genes is conducted using GSEA, identifying signaling pathways implicated in the development of OSA through GSEA.Sort according to the standardized enrichment score (NES), with a P-value 1 as the screening criteria. 2.12 Gene Set Variation Analysis (GSVA) Gene Set Variation Analysis (GSVA) operates as a non-parametric, unsupervised analytical approach. Unlike GSEA, this method eliminates the need for predefined sample categorization and directly computes enrichment scores for defined gene sets at the individual sample level. The fundamental concept of GSVA is to transform gene expression data from an expression matrix of individual genes to one featured by gene sets. Annotation databases are loaded into the R environment. Gene groupings are established according to GO and KEGG annotations. Gene Set Variation Analysis (GSVA) is performed via the "GSVA" package in R, leveraging transcriptomic sequencing-derived gene expression profiles. 2.13.1 Cell culture The alveolar type II epithelial cell line (AT2), human bronchial epithelial cell line (BEAS-2B), and human fetal lung fibroblast cell line (MRC-5) were sourced from Zhongshan Hospital, Fudan University, Shanghai, China. Cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM, HyClone, USA) enriched with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin (HyClone, USA). Cell cultures were grown under controlled humidity at 37℃ and 5% CO2, with no mycoplasma contamination detected. Cells were subcultured upon reaching 90% confluency. The cell suspension was collected and spun at 1000 rpm for 5 min. The pellet was resuspended in 1 mL of complete medium, then transferred to a culture dish with 7 mL of fresh DMEM to facilitate continued growth. 2.13.2 Extraction of Total Protein Upon reaching 80–90% confluency, AT2, BEAS-2B, and MRC-5 cells were harvested for protein isolation. Cell lysis was performed using RIPA lysis buffer (Cell Signaling Technology, Jiangsu, China) on ice for 30 minutes. Following detachment via a cell scraper, cells were moved into a new centrifuge tube for total protein lysate isolation. Samples underwent brief sonication (10 seconds) and were subsequently spun at 12,000 rpm for 15 minutes under refrigeration (4℃). The supernatant obtained post-centrifugation was harvested for further downstream applications. 2.13.3 WB assay Total protein was isolated from cellular or tissue samples using RIPA lysis buffer (Cell Signaling Technology, Jiangsu, China). Protein separation was achieved using 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), followed by electrophoretic transfer to a polyvinylidene difluoride (PVDF) membrane employing a submerged transfer methodology. Following transfer, membranes were washed in Tris-buffered saline containing 0.1% Tween-20 (TBST) and blocked with 5% skimmed milk powder prepared in TBST to reduce non-specific interactions. Immunoblotting with rabbit-derived primary antibodies and β-actin (1:10,000 dilution, AC026; ABclonal), serving as the loading reference, was conducted overnight at 4°C under constant low-speed rotation. Membranes were subsequently incubated with horseradish peroxidase (HRP)-linked secondary antibodies (Goat Anti-Rabbit IgG-HRP) diluted in antibody buffer at room temperature with gentle agitation for signal development. Western blot assays were conducted following the standardized protocol provided by Abcam, and protein bands were visualized using a Bio-Rad gel imaging system (Bio-Rad, Hercules, CA, USA). Comprehensive details regarding the commercial antibodies employed in this investigation can be found in the Supplementary Materials (Table S1 ). 2.13.4 Reverse transcription-quantitative polymerase chain reaction (RT-qPCR) of mRNA Total RNA isolation was performed using TRIzol reagent (Thermo Fisher Scientific, USA), following the manufacturer’s guidelines (Abcam protocol: https://www.abcam.cn/protocols/rna-isolation-protocol-cells-in-culture ). Complementary DNA (cDNA) synthesis was achieved with the PrimeScript Reverse Transcription Kit (Takara Bio, Japan). Quantitative real-time PCR (RT-qPCR) assays were executed with Bimake SYBR Green qPCR master mix on a Bio-Rad CFX Connect Real-Time PCR platform. The thermal cycling protocol included an initial denaturation step (95℃for 30 seconds), succeeded by 45 cycles of amplification (95℃ for 5 seconds, 55℃ for 30 seconds, and 72℃ for 30 seconds). Cycle threshold (Ct) values for candidate genes were standardized relative to ACTB expression levels in matched samples, with differential expression calculated using the 2 − ΔΔCt formula. Oligonucleotide primer sequences utilized in this workflow are provided in the Supplementary Materials (Table S1 ). 2.13.5 siRNAs and cell transfection. The siRNAs used in this study were obtained from TSINKE (Beijing, China) and included a negative control, EGR1#1 siRNA (F-CCAUGGACAACUACCCUAATT, R-UUAGGGUAGUUGUCCAUGGTT), and EGR1#2 siRNA (F-GCCUAGUGAGCAUGACCAATT, R-UUGGUCAUGCUCACUAGGCTT). Transfection of siRNAs and plasmids was performed using lipo8000 reagent (Beyotime, Shanghai, China) or Lipo3000 transfection reagent (Thermo Fisher Scientific, Waltham, MA, USA). The NC group was transfected with a negative control plasmid, while the OE-EGR1 group was transfected with an EGR1 overexpression plasmid. Cells were used for subsequent experiments after transfection. 2.13.6 Measurement of Cell Apoptosis AT2 cells were collected by digestion with trypsin without EDTA. Subsequently, cells were washed twice with PBS (1,000 rpm, 5 minutes of centrifugation), and 1×10⁵ to 5×10⁵ cells were collected. 5 µL of Annexin V-FITC reagent was added, followed by 5 µL of Propidium Iodide (PI). After incubation, apoptosis was immediately analyzed using a flow cytometer (ACEA NovoCyte, USA) to determine the apoptosis rate. 2.13.7 β-Galactosidase Assay Cellular senescence was evaluated by rinsing cells with PBS. Fixation was performed with 1 mL of β-galactosidase (β-gal) staining fixative under ambient conditions (15 min). Post-fixation, cells underwent additional PBS rinsing and were incubated with the chromogenic substrate. Cells were then maintained overnight at 37℃ in a CO 2 -free environment. After staining, microscopic examination was performed using an Olympus CKX31 optical microscope (Japan). The percentage of senescent cells was determined by analyzing randomly sampled microscopic fields and calculating the ratio of β-gal-positive cells per 100 total cell. 2.13.8 γ-H2AX Assay After the HEI-OC1 cells were collected, immunofluorescence staining was carried out. The process started by eliminating the culture medium from 6-well plates, followed by a single PBS rinse. Following this, the samples underwent incubation with rabbit-derived monoclonal γ-H2AX antibodies for 60-minute room temperature exposure. For nuclear visualization, DAPI solution was applied for five minutes at ambient temperature. Post-staining, specimens were embedded in fluorescence-preserving mounting agent and secured under a coverslip. Fluorescence microscopy showed that γ-H2AX immunoreactivity appeared as bright green fluorescence, while DAPI counterstaining highlighted the nuclei with blue fluorescence. 2.13.9 Enzyme-Linked Immunosorbent Assay (ELISA) Cytokine concentrations within AT2 cell culture supernatants were assessed via enzyme-linked immunosorbent assay (ELISA). Following 48 hours of transfection, supernatants were harvested, subjected to centrifugation, and cleared of cell fragments. IL-6, IL-17A, IL-1β, and TNF-α protein levels were specifically quantified during the assay, employing commercially available ELISA kits. Prior to initiating the procedure, all components were equilibrated at ambient temperature for half an hour. The protocol involved pipetting 50 µL of calibrators and test samples into designated wells, after which 100 µL of horseradish peroxidase-linked detection antibodies were introduced. 2.13.10 Telomere Length Measurement Genomic DNA isolation from AT2 cells was performed, normalized, and analyzed with the Telomere Relative Length Detection Kit (Shanghai's Wing and Applied Biotechnology Co.) in compliance with the supplier's guidelines. Telomere length assessment utilized the qPCR-based protocol established by Cawthon [ 37 ] , employing graded concentrations of reference DNA and a calibration-based approach [ 37 ] . SYBR Green fluorescence detection facilitated quantification of telomere relative length. To derive final measurements, the sample's fluorescence signals were analyzed to determine the proportional value reflecting telomeric DNA quantity. 2.14.11 Data Analysis Data analysis was conducted with R statistical software. Group differences were evaluated through Wilcoxon rank-sum, chi-square, and Fisher's exact tests, chosen based on data suitability. Correlations were analyzed via Spearman's rank-order approach, utilizing the ggpubr and stats packages for computational tasks. qPCR results, expressed as mean ± SD from triplicate runs, underwent two-tailed unpaired Student's t-test comparisons. Significance markers included ns (non-significant), *p < 0.05, **p < 0.01, and ***p < 0.001. To validate reproducibility, triplicate experimental iterations were executed, with statistical relevance defined at p 0.58 and FDR < 0.05 as selection criteria, we identified 94 DEDEGs between the control group and OSA patients.Volcano plots and heatmaps are shown in Figs. 2 A and 2 B. 3.2 Identifying Key Modules in OSA The WGCNA analysis included 44 patients with OSA and 15 healthy controls. A scale-free co-expression network was built using a soft power parameter (β = 10), achieving a scale-free topology fit (R2 = 0.9), as illustrated in Fig. 3 A. Hierarchical tree-based clustering detected 14 distinct modules, each comprising over 50 genes. These modules were then assessed for their linkages to OSA (Fig. 3 C). Findings revealed that the green (r = 0.67, p = 6e − 9), magenta (r = 0.33, p = 0.01), and pink (r = 0.32, p = 0.01) modules displayed positive correlations with OSA. Conversely, the yellow (r = − 0.73, p = 3e − 11), red (r = − 0.44, p = 4e − 4), and tan (r = − 0.36, p = 0.005) modules showed inverse relationships. The green module, demonstrating the most significant linkage (r = 0.67) and harboring 334 genes, was prioritized for deeper investigation. A scatter plot (Fig. 3 E) graphically represented the association between module connectivity and OSA-related trait relevance. 3.3 Functional Enrichment Analysis of DEDEs Cross-referencing 94 DEGs against the green module's gene set yielded 21 OSA-associated DEDEGs. The biological roles of these overlapping genes were investigated via GO and KEGG pathway enrichment studies. GO analysis demonstrated marked enrichment in key biological pathways, including transcriptional regulation of microRNAs, control of neuronal apoptosis, and calcium signaling-mediated cellular responses. Cellular compartment evaluation underscored the relevance of distinct subcellular structures, such as lysosomal compartments and RNA polymerase II-containing transcriptional regulatory complexes. Molecular function analysis demonstrated significant interactions between RNA polymerase II transcription factors and SMAD protein binding. KEGG pathway enrichment revealed prominent involvement in immune-related signaling cascades, including IL-17 and TNF signaling pathways, along with rheumatoid arthritis pathogenesis. Additionally, significant enrichment was observed in osteogenic differentiation pathways and infection-related mechanisms, particularly those associated with Kaposi's sarcoma-associated herpesvirus and enteropathogenic Escherichia coli. 3.4 Correlation Between OSA-ARDEGs By cross-referencing 21 OSA-associated genes with genes linked to aging, we pinpointed 4 aging-associated OSA-linked genes (OSA-ARDEGs), as visualized in Fig. 5 A. The relationships among OSA-ARDEGs were evaluated using Pearson correlation coefficients.High correlations were found between EGR1 and FOS (cor = 0.95), and JUN (cor = 0.93), as well as between JUN and FOS (cor = 0.95) (Fig. 5 B). Figure 5 C illustrates the chromosomal positions of the OA-ARDEGs. 3.5 Immune Infiltration Analysis To examine variations in the immunological landscape of OSA patients compared to healthy individuals, we evaluated the prevalence and functional activity of immune cell subsets across both cohorts (Fig. 6 A). Differences in immune cell abundance between groups, along with interrelationships among 22 immune subtypes, are illustrated in Figs. 6 B and 6 C. Results revealed a marked rise in M1 macrophage levels among OSA subjects relative to controls. Conversely, OSA cases displayed decreased proportions of plasma cells, naïve CD4 + T lymphocytes and activated dendritic cells. We then examined the relationships between infiltrative immune cells and four types of OSA-ARDEGs (Fig. 7 ). Immune infiltration analysis revealed that EGR1 exhibited notable associations with dendritic cell activation (r = 0.59) and regulatory T cell abundance (Tregs; r = − 0.44) (Fig. 7 A), indicating a pivotal function in modulating these populations. FOS displayed strong linkages to dendritic cell activation (r = 0.66) and inversely correlated with M0 macrophage prevalence (r = − 0.53) (Fig. 7 B), highlighting its potential regulatory influence on these immune subsets. GLB1 demonstrated inverse relationships with monocyte infiltration (r = − 0.65) and dendritic cell activation (r = − 0.49) (Fig. 7 C), suggesting its role in suppressing these cellular responses. JUN exhibited connections with dendritic cell activation (r = 0.56) and a negative association with memory B cell frequency (r = − 0.4) (Fig. 7 D), implying involvement in immune activation dynamics and memory regulation. 3.6 Identification of Biological Functional Characteristics of Various Aging-Related Subtypes Unsupervised clustering of OSA samples using core regulatory factor expression profiles identified two distinct molecular subtypes (Figs. 8 A-E). Differential expression analysis of regulatory factors revealed subtype-specific molecular signatures, with subtype 2 demonstrating marked upregulation of GLB1 and concurrent downregulation of EGR1, FOS, and JUN (Fig. 8 A). Comparative evaluation further highlighted distinct immune infiltration profiles between the two subgroups. Specifically, subgroup 1 demonstrated elevated effector memory CD8 + T cell and eosinophil infiltration, whereas subgroup 2 exhibited higher central memory CD8 + T cells and both CD56dim/bright natural killer cell populations (Fig. 8 F). GSVA analysis indicated that subtype 1 exhibited upregulation in various biological processes and pathways, primarily involving lipoprotein particle clearance, intracellular lipid transport, and oxidoreductase activity, related to metabolism and molecular transport (Fig. 9 C).Moreover, this subtype exhibited enrichment in metabolic and biosynthetic pathways, including glutathione metabolism, porphyrin and chlorophyll metabolism, folate biosynthesis, and peroxisomes, possibly indicating subtype 1’s significant roles in oxidative stress response and metabolic regulation (Fig. 9 D).In contrast, subtype 2 exhibited significant activity in processes such as regulation of cell proliferation, stem cell development, inflammation, and immune regulation (Fig. 9 C). Key enriched pathways were primarily linked to immune system modulation and intracellular communication, encompassing the JAK-STAT signaling cascade, NOD receptor-mediated pathway, TCR signaling pathways, and TGF-β signaling cascades (as shown in Fig. 9 D). 3.7 Construction and Validation of OSA-ARDEGs Gene Markers To enhance the diagnostic accuracy of hub OA-ARDEGs in OSA, we employed three distinct ML algorithms—LASSO regression (Fig. 10 A-B), SVM-RFE analysis (Fig. 10 C-D), and Random Forest modeling (Fig. 10 E-F)—for feature selection. Through integrative analysis of these algorithmic outputs, GLB1 and EGR1 emerged as robust candidates, defining two key OSA-linked hub genes (OSA-ARDEGs). 3.8 Validating Characteristic Biomarkers Using ML Models Multiple machine learning algorithms, including AdaBoost, Decision Tree, KNN, LightGBM, Naïve Bayes, and XGBoost, were employed for OSA characteristic gene identification. Model performance evaluation demonstrated recall rates exceeding 50% across all algorithms, as detailed in Table 1 . The predictive performance of each model was further validated through ROC curve analysis, presented in Fig. 11 .Specifically, AdaBoost excelled in all metrics, achieving 100% accuracy, recall, F1 score, and an AUC of 1.0, demonstrating its superiority in identifying characteristic genes in OSA.Among them, AdaBoost had the highest AUC (1.0), while LightGBM performed best in terms of accuracy (0.9), Kappa value (0.7443), and F1 score (0.9318).The importance plots of the four OSA-ARDEGs for different models are shown in Fig. 12 . EGR1 and GLB1 emerged as predominant features across multiple models, with GLB1 specifically identified as a key feature in the Decision Tree algorithm. Table 1 Comparison of diagnostic effects of six different machine learning models. ML models TP TN FP FN Accuracy Kappa Precision Recall F1-Score AUC AdaBoost 44 16 0 0 1.0 1.0 1.0 1.0 1.0 1 Decision Tree; 41 11 5 3 0.8667 0645 0.9318 0.8913 0.9111 0.905 KNN 41 10 6 3 0.85 0.5921 0.9318 0.8723 0.9011 0.915 LightGBM 41 13 3 3 0.9 0.7443 0.9318 0.9318 0.9318 0.964 Naïve Bayes 33 11 5 11 0.7333 0.3909 0.7500 0.8684 0.8049 0.834 XGBoost 39 9 7 5 0.8 0.4675 0.8864 0.8478 0.8667 0.9048 3.9 Construction and Evaluation of the LR Model Feature relevance assessment revealed two top-ranking genes, mutually prioritized by both AdaBoost and LightGBM frameworks, which were designated as core OSA-DEDEGs. Using these key genes, EGR1 and GLB1, a logistic regression (LR) model was developed. Investigation demonstrated significantly lower EGR1 levels in OSA patients relative to control subjects, whereas GLB1 expression was elevated in OSA cases (Fig. 13 A). Furthermore, a nomogram developed through LR analysis is shown in Fig. 13 B. The model's robustness was assessed via 1000 bootstrap resampling cycles applied to the training dataset. Calibration assessment validated the high predictive performance of the LR algorithm, as illustrated in Fig. 13 C. Decision curve analysis revealed superior clinical utility of the model-based approach for OSA patients, as evidenced by the consistent superiority of the model's benefit curve (red line) over the default strategies (gray line) across threshold probabilities (Fig. 13 D). The training set evaluation demonstrated robust diagnostic performance, with Hub OSA-ARDEGs achieving AUC values of 0.801 and 0.865 (Fig. 14 A), while the logistic regression model attained an AUC of 0.865 in the validation cohort (Fig. 14 B). External validation using the GSE75097 dataset yielded AUC values of 0.607 and 0.577 for Hub OSA-ARDEGs (Fig. 14 C), with the LR model achieving an AUC of 0.667 (Fig. 14 D). These findings confirm the superior diagnostic predictive capacity of the logistic regression model across both internal and external validation datasets. 3.6 Pathway Enrichment Analysis (GSEA) of Hub OSA-ARDEGs Pathway enrichment analysis for GLB1 and EGR1 was performed using GSEA. The results reveal distinct biological processes associated with each gene. EGR1 exhibited marked enrichment in biological pathways associated with cardiomyocyte differentiation, interleukin-1-mediated signaling responses, and DNA-binding transcriptional activation mechanisms, as illustrated in Fig. 15 A. These processes appear more prominently among the most differentially expressed genes, indicated by earlier peaks in the ranked list for EGR1 compared to GLB1.For GLB1, enrichment analysis highlights its involvement in chromosome segregation, mitochondrion organization, and nuclear chromosome segregation, shown in Fig. 15 C. Extended pathway enrichment profiling of EGR1-related networks revealed associations with cytokine receptor cross-talk, TCR signaling cascades, and olfactory signal transduction mechanisms (as shown in Fig. 15 B). In contrast, GLB1 pathways are enriched in oxidative phosphorylation, ribosome function, and the response to Vibrio cholerae infection (Fig. 15 D). Each gene's unique enrichment profile underscores its potential regulatory mechanisms and interactions within cellular processes. 3.7 GSVA for Hub OSA-ARDEGs Gene Set Variation Analysis (GSVA) was applied to assess GLB1 and EGR1 differential activity, specifically probing their functional involvement across distinct biological pathways. EGR1 predominantly upregulates pathways associated with metabolic processes and immune response modulation. Key upregulated pathways include Hemoglobin Alpha Binding, Toll-like Receptor 4 Binding, and Interleukin-21 Production. Notably, EGR1 downregulates pathways related to granulocyte chemotaxis and extracellular matrix disassembly, indicating a lesser role in these functions (Fig. 16 A).Additionally, EGR1 upregulates pathways crucial to cellular metabolism, such as Porphyrin and Chlorophyll Metabolism, Sulfur Metabolism, Glutathione Metabolism, and Folate Biosynthesis. Conversely, it downregulates pathways involved in specialized or stress responses, including the Renin-Angiotensin System and Complement and Coagulation Cascades (Fig. 16 B). Similarly, GLB1 exhibits strong upregulation in broader immune-related pathways, including Regulation of T-helper 17 Cell Differentiation and Interleukin 21 Production. Pathways such as Chitin Metabolic Process and Chitinase Activity are notably downregulated (Fig. 16 C). GLB1 significantly enhances pathways critical to immune function and inflammation, including Glycosaminoglycan Biosynthesis, Glycosylphosphatidylinositol (GPI)-Anchor Biosynthesis, and Galactose Metabolism, while downregulating the Renin-Angiotensin System and pathways associated with Systemic Lupus Erythematosus and Primary Immunodeficiency (Fig. 16 D). 3.8 Selection of Basal Protein and Verification of Overexpression Plasmid To further explore the potential mechanisms of cellular senescence, this study focused on EGR1. During preliminary screening to identify a suitable cellular system, EGR1 baseline expression was assessed across three human cell lines: AT2 alveolar epithelial cells (type II), BEAS-2B bronchial epithelial cells, and MRC-5 fetal lung fibroblasts. Comparative analysis of EGR1 basal expression across three cell types revealed significantly higher expression levels in AT2 cells (Fig. 17 A), establishing them as the experimental model for subsequent investigations. To elucidate EGR1's role in cellular senescence mechanisms, an EGR1 overexpression vector was developed, with successful transfection confirmed through Western blot analysis (Fig. 17 B). 3.9 Selection of Senescence-Related Genes After transfection with the EGR1 overexpression plasmid, to improve the accuracy of predicting downstream signaling pathways, candidate genes for 12 Hub OSA-ARDEGs were predicted using the GEO and HCA databases. These genes included: CIITA, NAIP, NOD1, NOD2, NLRC5, NLRP1, NLRP3, NLRP4, NLRP5, NLRP13, NLRP14, and NLRX1. The accuracy of the candidate genes was subsequently verified by qPCR, as shown in Fig. 18 (A). Due to issues with the CIITA and NLRP13 primers, new primers were designed for a second qPCR experiment. Finally, the three most markedly dysregulated genes—CIITA, NAIP, and NLRX1—were prioritized for immunoblot validation (Fig. 18 B). Data revealed markedly elevated expression levels of these genes in EGR1-overexpressing AT2 cells. 3.10 Apoptosis Detection Apoptosis analysis through flow cytometry revealed distinct cellular states between experimental groups. The negative control (NC) group exhibited predominant cell viability, with the majority of cells localized in the healthy quadrant (Q3-3) and minimal apoptotic populations (Q3-1 and Q3-2), indicating low basal apoptosis rates and active cell proliferation. In contrast, EGR1 overexpression significantly enhanced apoptotic cell proportions, as demonstrated in Fig. 19 (A). 3.11 Cellular Senescence and DNA Damage To assess the impact of EGR1 overexpression on cellular senescence in AT2 cells, β-galactosidase activity levels were quantified in both non-modified control (NC) and EGR1-overexpressing cohorts. Relative to NC cells, EGR1-overexpressing AT2 cells exhibited markedly elevated senescence levels (Fig. 19 B). Furthermore, qPCR analysis of telomere length revealed no significant difference in telomere length between the NC and EGR1 overexpressing AT2 cells (Fig. 19 D), suggesting that EGR1 does not induce senescence through a telomere-dependent mechanism. To explore the involvement of EGR1 in modulating the DNA damage response (DDR), γ-H2AX immunofluorescence staining was utilized to identify DNA double-strand breaks (DSBs). In the negative control (NC) group, γ-H2AX foci were scarcely detectable after treatment. In contrast, cells overexpressing EGR1 exhibited sustained γ-H2AX signals (Fig. 19 C). These results suggest that EGR1 promotes cellular senescence and DDR activation in AT2 cells via a mechanism independent of telomere regulation. 3.12 ELISA Pro-inflammatory cytokines, particularly TNF-α and IL-6, are central drivers of OSA progression. Cytokine quantification assays identified robust upregulation of inflammatory mediators (IL-6, IL-17A, IL-1β, TNF-α) in EGR1-overexpressing AT2 cells (Fig. 20 ). These findings suggest that EGR1 potentially drives OSA progression by regulating inflammatory signaling pathways. 4. Discussion OSA is increasingly recognized as a critical health problem, particularly prevalent in aging populations. OSA is marked by frequent episodes of upper airway collapse during sleep, causing intermittent oxygen desaturation, fragmented sleep patterns, and consequent excessive daytime fatigue [ 38 , 39 ] . The formal study of aging mechanisms traces its origins to 1989, when Irwin Rosenberg defined "sarcopenia" to characterize the progressive loss of skeletal muscle mass and strength associated with aging [ 40 ] . Later investigations have deepened insights into aging biology, uncovering multifaceted relationships between aging and the gradual impairment of systemic organ functions, dysregulated metabolic processes, and the emergence of age-related chronic pathologies [ 41 ] . For instance, aging contributes to the onset of diabetes, cancer, cardiovascular diseases, and neurodegenerative diseases through mechanisms such as mitochondrial autophagy dysfunction [ 42 ] , inflammatory pathway activation [ 43 ] , and changes in hormone levels [ 44 ] . Recent research has highlighted significant interactions between the aging process and OSA. Muscle mass loss caused by aging, such as the decline in pharyngeal dilator muscle function, and metabolic dysregulation, like insulin resistance, may worsen the pathological progression of OSA [ 42 , 45 ] . Epidemiological studies indicate that OSA affects a substantial number of adults, with prevalence expected to rise as the global population ages [ 11 ] . Advanced age significantly elevates the risk of OSA onset. Progressive age-related physiological alterations—including diminished muscular tone, heightened adipose accumulation in cervical regions, and respiratory control dysregulation—collectively underlie OSA pathogenesis [ 46 , 47 ] . In older adults, OSA is often associated with accelerated cognitive deterioration and a heightened risk of Alzheimer’s disease. These findings indicate that OSA may not simply be a marker of normal aging but might actively contribute to its progression [ 48 , 49 ] . Understanding the interplay between OSA and aging is essential for developing targeted interventions that can mitigate OSA's impact on the health and quality of life of older adults [ 50 ] . However, although recent research has increasingly recognized the significant role of aging in OSA, the exact mechanisms and contributions remain elusive. An integrated analysis of transcriptomic datasets initially detected 94 genes (OSA-DEGs) exhibiting significant differential expression between OSA patients and controls. To investigate their disease relevance, co-expression network analysis (WGCNA) uncovered a module comprising 334 OSA-correlated genes. Cross-validation via two orthogonal algorithms refined the candidate list to 21 DEDEGs, enhancing confidence in the reproducibility and precision of these biomarkers.To delineate the functional attributes of DEGs, GO and KEGG pathway enrichment investigations were performed. GO analysis revealed significant enrichment in processes associated with transcriptional regulation mediated by miRNAs—non-coding RNA molecules involved in post-transcriptional silencing mechanisms. Emerging evidence emphasizes the diagnostic and therapeutic potential of miRNAs in OSA, as highlighted by recent research. For instance, Moriondo et al. identified specific miRNA signatures correlating with OSA severity, underscoring their utility as biomarkers and intervention targets [ 51 ] . Altered miRNA expression may affect the pathogenesis of OSA through pathways involving inflammation, oxidative stress, and metabolic dysfunction, which are prevalent in OSA patients [ 52 ] . The "IL-17 signaling pathway" was significantly enriched. This pathway encompasses the binding of IL-17 to its receptors, tinitiating a cascade of intracellular signaling events that culminate in the activation of NF-κB and other transcription factors, which promote the expression of inflammatory genes [ 53 ] . Emerging research underscores a robust pathophysiological link between IL-17 signaling dysregulation and OSA—a condition characterized by recurrent nocturnal upper airway collapse, intermittent hypoxia, and chronic systemic inflammation [ 54 , 55 ] , This association establishes a mechanistic rationale for downstream experimental validation. To investigate the contribution of aging-associated genes to OSA pathogenesis, we analyzed 543 ARDEGs, narrowing this to four hub OSA-ARDEGs with disease-specific relevance. SsGSEA revealed an elevated enrichment score for M1 macrophages in OSA patients compared to non-OSA controls. Concurrently, reduced infiltration levels were observed for plasma cells, naïve CD4 + T lymphocytes, monocytes, and activated dendritic cells in OSA cohorts. The prominence of M1 macrophages aligns mechanistically with the chronic inflammatory milieu and oxidative stress burden hallmarking OSA. We assessed the efficacy of six machine learning algorithms to select the optimal classifier for discerning OSA-associated signature genes. Specifically, AdaBoost outperformed all other models across all metrics, achieving perfect accuracy, recall, F1 score, and AUC (all at 100%), highlighting its distinct advantage in precisely identifying OSA characteristic genes. In the global and local interpretability analysis of four OSA-arDEGs, two key genes, EGR1 and GLB1, were specifically identified. EGR1 is a zinc-finger transcription factor that orchestrates critical biological functions, including cellular proliferation, differentiation, and apoptosis. Conversely, GLB1, a lysosomal enzyme, exhibits progressive upregulation in senescent tissues and serves as a molecular indicator of cellular aging and age-related functional deterioration. This increase exacerbates the decline in lung function, which is critical in cases of OSA. GLB1, encoded at the GLB1 gene locus, is essential for lysosomal function and is characteristically elevated in the tissues of aging organisms, underscoring its role as a biomarker of cellular senescence [ 56 ] . The increase in GLB1 is linked to a decline in age-related physiological functions, especially in the lung system, where systemic aging and GLB1 activity may relate to pathologies such as fibrosis and cellular senescence, worsening the decline in lung function [ 57 ] . To validate the results of bioinformatics analysis, we selected the EGR1 gene for further investigation.Studies across multiple fields show that EGR1 regulates key signaling pathways, influencing aging and immune homeostasis. In aging - related diseases, EGR1 affects cell function by regulating mitophagy and oxidative stress. For example, in an intervertebral disc degeneration model, EGR1 exacerbates oxidative - stress - induced nucleus pulposus cell senescence by inhibiting PINK1 - Parkin - dependent mitophagy [ 58 ] . In immune regulation, EGR1 directly modulates inflammatory enhancer activity. During macrophage differentiation, it suppresses pro - inflammatory genes (e.g., IL1β, TNFα) by recruiting the NuRD repressor complex, limiting excessive inflammatory responses [ 59 ] . Despite this, the role of EGR1 in OSA remains poorly understood. Apoptosis refers to the programmed cell death process triggered by internal or external signals under specific physiological or pathological conditions. Several studies have shown that OSA induces apoptosis in tissues such as the brain, heart, and kidneys through intermittent hypoxia, further exacerbating cognitive impairment, metabolic disorders, and organ dysfunction [ 60 – 62 ] . Based on this, we analyzed the apoptosis levels in AT2 cells. The results indicated that the occurrence of apoptosis provided further support for the hypothesis of the EGR1 pathway predicted by bioinformatics analysis. Cellular senescence manifests through two distinct molecular pathways [ 63 ] : (1) replicative senescence due to telomerase dysfunction and progressive telomere shortening, causing irreversible cell - cycle arrest linked to lost proliferative capacity [ 64 ] ; and (2) stress - induced premature senescence (SIPS), a potentially reversible process triggered by external stressors such as oxidative damage, radiation, or high glucose [ 65 ] . Persistent exposure to oxidative stress or environmental stressors activates DNA damage response (DDR) pathways, culminating in either permanent cell cycle arrest or transient proliferative arrest (SIPS), dictated by the magnitude and duration of stress. Mechanistically, our findings reveal two parallel senescence - inducing pathways: the telomere - dependent pathway, marked by significant telomere attrition tied to aging; and the DNA damage response pathway, characterized by notable γ - H2AX accumulation, a sign of genomic instability and stress - induced senescence. Importantly, our results suggest that EGR1 may mechanistically promote SIPS development by activating stress - response pathways, particularly those linked to the DNA damage response. Immune system dysregulation in OSA patients has long been a central focus of clinical research. Empirical studies demonstrate that EGR1 drives transcriptional activation of CIITA, NAIP, and NLRX1 within AT2 alveolar epithelial cells.CIITA (MHC Class II Transactivator) serves as the principal controller governing MHC-II gene transcriptional activation. Recognized as the archetypal member within the NLR family (nucleotide-binding oligomerization domain-like receptors), this regulator plays a critical role in mediating adaptive immune responses. Research has demonstrated that CIITA modulates immune responses within the tumor microenvironment through the regulation of MHC-II molecule expression [ 66 ] . NLRX1, a mitochondrially localized member of the nucleotide-binding oligomerization domain-like receptor (NLR) family, mediates pleiotropic immunomodulatory roles across innate and adaptive immunity, including pathogen sensing and inflammatory homeostasis. It negatively regulates type I interferons and pro-inflammatory responses by modulating the NF-κB signaling pathway, thereby influencing cell death, autophagy, and proliferation [ 67 ] . NAIP (NLR Family Apoptosis Inhibitory Protein) is a core component of the NAIP/NLRC4 inflammasome and primarily initiates the inflammatory response by recognizing pathogenic factors such as bacterial flagellin. The NAIP/NLRC4 inflammasome plays an important role in host defense, particularly in immune responses against bacterial infections by recruiting and activating Caspase-1 [ 68 , 69 ] . EGR1 plays a central role in OSA pathophysiology by orchestrating immune regulatory networks, particularly through modulation of CIITA, NAIP, and NLRX1 activity. Mechanistically, EGR1 activation triggers transcriptional induction of pro-inflammatory cytokines (e.g., IL-6, IL-17A, IL-1β, TNF-α), which are cornerstones of OSA-associated inflammation and pathogenic drivers of disease initiation and progression. These findings position EGR1 as a master regulator of immune dysregulation in OSA, offering a biologically plausible therapeutic target. This mechanistic insight aligns with and extends prior computational predictions of EGR1’s centrality in OSA pathogenesis. Advancing age profoundly impacts pulmonary cellular homeostasis, resulting in diminished lung capacity and elastic recoil, thereby amplifying the severity of obstructive respiratory pathologies such as OSA[70]. Emerging evidence implicates gerontogenes—particularly those regulating cellular senescence and mitochondrial dynamics—as potential diagnostic biomarkers for OSA. This supports a paradigm wherein genetic susceptibility to OSA exhibits age-dependent escalation, attributable to lifelong accrual of genomic instability and environmental exposures [ 50 ] . Furthermore, epigenetic modifications in these aging genes can lead to metabolic imbalances, further complicating the cellular environment and promoting the progression of OSA. This investigation is subject to certain constraints. Primarily, while we performed rigorous internal and external validation of the diagnostic framework, its generalizability necessitates additional validation across diverse, independent cohorts, particularly in prospective, multi-center studies. In this research, we utilized all available OSA datasets. The role of EGR1 in OSA remains incompletely understood: while experimental results suggest that EGR1 may influence immune factors and DNA damage response in OSA, its specific mechanisms need to be further explored. The complex interactions between EGR1 and aging have not been fully examined either. Furthermore, to further validate the robustness and reliability of the diagnostic method, it is recommended to include more external validation cohorts. 5. Conclusion In this study, we employed bioinformatics and machine learning approaches to investigate age - related gene expression patterns and molecular mechanisms in OSA. We identified two key senescence - related genes that are closely associated with OSA and developed a logistic regression (LR) diagnostic model based on their expression profiles. This model significantly enhanced the accuracy of OSA diagnosis. The expression levels of these genes were significantly associated with OSA severity, indicating their potential as novel diagnostic biomarkers. In our preliminary exploration of the mechanisms of senescence - related genes in OSA, we found that EGR1, a marker of cellular senescence and immune response, may influence OSA progression by regulating immune - related factors and SIPS. EGR1 may contribute to OSA through senescence and immune - related pathways, making it a potential therapeutic target. Declarations Funding: This study was funded by the National Natural Science Foundation of China (81270144, 30800507, 81170071 and 81400063). Conflict of Interest: Author Li Caili declares that she has no conflict of interest. Author Zhou Wei declares that she has no conflict of interest. Author Xu Chong declares that she has no conflict of interest. Author Cao Jie declares that she has no conflict of interest. Author Zhang Jing declares that she has no conflict of interest. Ethical approval The ethical and methodological aspects of the study was approved by Institutional Review Board of Tianjin Medical University General Hospital (TMU IRB Approving Number: EA-20-120002) and performed in accordance with the Guide for the Care and Use of Laboratory Animals. Acknowledgements ：None. Data Availability Statements The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request. References JOOSTEN S A, LANDRY S A, WONG A M, et al. Assessing the Physiologic Endotypes Responsible for REM- and NREM-Based OSA [J]. Chest. 2021;159(5):1998–2007. BAJAJ R, SINGH N. Properties of octenyl succinic anhydride (OSA) modified starches and their application in low fat mayonnaise [J]. Int J Biol Macromol. 2019;131:147–57. LAM JC, MAK J C, IP MS. Obesity, obstructive sleep apnoea and metabolic syndrome [J]. Respirol (Carlton Vic). 2012;17(2):223–36. MESSINEO L, BAKKER J P, CRONIN J, et al. Obstructive sleep apnea and obesity: A review of epidemiology, pathophysiology and the effect of weight-loss treatments [J]. Sleep Med Rev. 2024;78:101996. KHALYFA A, MARIN JM, QIAO Z et al. Plasma exosomes in OSA patients promote endothelial senescence: effect of long-term adherent continuous positive airway pressure [J]. Sleep, 2020, 43(2). LIN Y N, LI Q Y, ZHANG X J. Interaction between smoking and obstructive sleep apnea: not just participants [J]. Chin Med J. 2012;125(17):3150–6. NARANJO M, WILLES L, PRILLAMAN B A, et al. Undiagnosed OSA May Significantly Affect Outcomes in Adults Admitted for COPD in an Inner-City Hospital [J]. Chest. 2020;158(3):1198–207. NG N B H, LIM C Y S TANS, et al. Screening for obstructive sleep apnea (OSA) in children and adolescents with obesity: A scoping review of national and international pediatric obesity and pediatric OSA management guidelines [J]. Obes reviews: official J Int Association Study Obes. 2024;25(5):e13712. DHARMAKULASEELAN L, BOULOS M I. Sleep Apnea and Stroke. Narrative Rev [J] Chest. 2024;166(4):857–66. BHASIN S, BRITO J P, CUNNINGHAM G R, et al. Testosterone Therapy in Men With Hypogonadism: An Endocrine Society Clinical Practice Guideline [J]. J Clin Endocrinol Metab. 2018;103(5):1715–44. GASPAR LS, ÁLVARO A R, MOITA J, et al. Obstructive Sleep Apnea and Hallmarks of Aging [J]. Trends Mol Med. 2017;23(8):675–92. LI Y, WANG Y. Obstructive Sleep Apnea-hypopnea Syndrome as a Novel Potential Risk for Aging [J]. Aging disease. 2021;12(2):586–96. LIU P Y, REDDY RT. Sleep, testosterone and cortisol balance, and ageing men [J]. Reviews Endocr metabolic disorders. 2022;23(6):1323–39. CARROLL JE, IRWIN M R, SEEMAN T E et al. Obstructive sleep apnea, nighttime arousals, and leukocyte telomere length: the Multi-Ethnic Study of Atherosclerosis [J]. Sleep, 2019, 42(7). COONEY L G DOKRASA. Beyond fertility: polycystic ovary syndrome and long-term health [J]. Fertil Steril. 2018;110(5):794–809. LAL C, AYAPPA I, AYAS N, et al. The Link between Obstructive Sleep Apnea and Neurocognitive Impairment: An Official American Thoracic Society Workshop Report [J]. Annals Am Thorac Soc. 2022;19(8):1245–56. WEIHS A, FRENZEL S, WITTFELD K et al. Associations between sleep apnea and advanced brain aging in a large-scale population study [J]. Sleep, 2021, 44(3). YEREVANIAN A, SOUKAS AA, Metformin. Mechanisms in Human Obesity and Weight Loss [J]. Curr Obes Rep. 2019;8(2):156–64. CHEN J W, HUANG M J, CHEN X N, et al. Transient upregulation of EGR1 signaling enhances kidney repair by activating SOX9(+) renal tubular cells [J]. Theranostics. 2022;12(12):5434–50. JIN Y, WANG C, ZHANG B, et al. Blocking EGR1/TGF-β1 and CD44s/STAT3 Crosstalk Inhibits Peritoneal Metastasis of Gastric Cancer [J]. Int J Biol Sci. 2024;20(4):1314–31. WANG Y, QIN C, ZHAO B, et al. EGR1 induces EMT in pancreatic cancer via a P300/SNAI2 pathway [J]. J translational Med. 2023;21(1):201. WANG B, WANG Y, WANG W, et al. WTAP/IGF2BP3 mediated m6A modification of the EGR1/PTEN axis regulates the malignant phenotypes of endometrial cancer stem cells [J]. J experimental Clin cancer research: CR. 2024;43(1):204. NGIAM K Y, KHOR I W. Big data and machine learning algorithms for health-care delivery [J]. Lancet Oncol. 2019;20(5):e262–73. XU M, ZHOU H, HU P, et al. Identification and validation of immune and oxidative stress-related diagnostic markers for diabetic nephropathy by WGCNA and machine learning [J]. Front Immunol. 2023;14:1084531. GHARIB SA, HURLEY A L, ROSEN M J et al. Obstructive sleep apnea and CPAP therapy alter distinct transcriptional programs in subcutaneous fat tissue [J]. Sleep, 2020, 43(6). GHARIB SA, HAYES A L, ROSEN M J, et al. A pathway-based analysis on the effects of obstructive sleep apnea in modulating visceral fat transcriptome [J]. Sleep. 2013;36(1):23–30. CHEN YC, CHEN K D, SU M C, et al. Genome-wide gene expression array identifies novel genes related to disease severity and excessive daytime sleepiness in patients with obstructive sleep apnea [J]. PLoS ONE. 2017;12(5):e0176575. ADEGUNSOYE A, NEBORAK J M, ZHU D, et al. CPAP Adherence, Mortality, and Progression-Free Survival in Interstitial Lung Disease and OSA [J]. Chest. 2020;158(4):1701–12. LORENZI-FILHO G, ALMEIDA F R, STROLLO P J. Treating OSA. Current and emerging therapies beyond CPAP [J]. Respirology (Carlton, Vic), 2017, 22(8): 1500-7. RITCHIE M E, PHIPSON B. Nucleic Acids Res. 2015;43(7):e47. WU D, et al. limma powers differential expression analyses for RNA-sequencing and microarray studies [J]. WU T, HU E, XU S, Innovation et al. (Cambridge (Mass)), 2021, 2(3): 100141. LUO W. Pathview: an R/Bioconductor package for pathway-based data integration and visualization [J]. Bioinf (Oxford England). 2013;29(14):1830–1. ZHOU J, HUANG J, LI Z, et al. Identification of aging-related biomarkers and immune infiltration characteristics in osteoarthritis based on bioinformatics analysis and machine learning [J]. Front Immunol. 2023;14:1168780. DE MAGALHãES JP, ABIDI Z, DOS SANTOS G A, et al. Human Ageing Genomic Resources: updates on key databases in ageing research [J]. Nucleic Acids Res. 2024;52(D1):D900–8. ZUCCATO JA, PATIL V. Cerebrospinal fluid methylome-based liquid biopsies for accurate malignant brain neoplasm classification [J]. Neurooncology. 2023;25(8):1452–60. GU Z, EILS R. Complex heatmaps reveal patterns and correlations in multidimensional genomic data [J]. Bioinf (Oxford England). 2016;32(18):2847–9. CAWTHON R M. Telomere length measurement by a novel monochrome multiplex quantitative PCR method [J]. Nucleic Acids Res. 2009;37(3):e21. DESHMUKH GOMASEVG. Obstructive Sleep Apnea and Its Management: A Narrative Review [J]. Cureus. 2023;15(4):e37359. WHITE D P, YOUNES M K. Obstructive sleep apnea [J]. Compr Physiol. 2012;2(4):2541–94. NISHIKAWA H, FUKUNISHI S, ASAI A et al. Pathophysiology and mechanisms of primary sarcopenia (Review) [J]. Int J Mol Med, 2021, 48(2). OH HS, RUTLEDGE J. Organ aging signatures in the plasma proteome track health and disease [J]. Nature. 2023;624(7990):164–72. KITADA M. Autophagy in metabolic disease and ageing [J]. Nat reviews Endocrinol. 2021;17(11):647–61. ARMENTO A, UEFFING M, CLARK SJ. The complement system in age-related macular degeneration [J]. Cell Mol Life Sci. 2021;78(10):4487–505. GAUTHIER B R, SOLA-GARCíA A, CáLIZ-MOLINA M, et al. Thyroid hormones in diabetes, cancer, and aging [J]. Aging Cell. 2020;19(11):e13260. LIU D, WANG S, LIU S et al. Frontiers in sarcopenia: Advancements in diagnostics, molecular mechanisms, and therapeutic strategies [J]. Molecular aspects of medicine, 2024, 97: 101270. PERGER E, MATTALIANO P. LOMBARDI C Menopause Sleep Apnea [J] Maturitas. 2019;124:35–8. YEO E J. Hypoxia and aging [J]. Exp Mol Med. 2019;51(6):1–15. ANDRADE A G, BUBU O M, VARGA A W, et al. The Relationship between Obstructive Sleep Apnea and Alzheimer's Disease [J]. J Alzheimer's disease: JAD. 2018;64(s1):S255–70. KINUGAWA K. Obstructive sleep apnea and dementia: A role to play? [J]. Rev Neurol. 2023;179(7):793–803. MULLINS A E, KAM K, PAREKH A, et al. Obstructive Sleep Apnea and Its Treatment in Aging: Effects on Alzheimer's disease Biomarkers, Cognition, Brain Structure and Neurophysiology [J]. Neurobiol Dis. 2020;145:105054. MORIONDO G, SOCCIO P, TONDO P et al. Obstructive Sleep Apnea: A Look towards Micro-RNAs as Biomarkers of the Future [J]. Biology, 2022, 12(1). ZHANG K, WANG C, WU Y, et al. Identification of novel biomarkers in obstructive sleep apnea via integrated bioinformatics analysis and experimental validation [J]. PeerJ. 2023;11:e16608. YANG J. Role and mechanism of IL – 17 and its gene polymorphisms in dyslipidemia caused by obstructive sleep apnea syndrome in children [J]. Cellular and molecular biology (Noisy-le-Grand, France), 2022, 68(2): 208–12. BHATT SP, GULERIA R, KABRA SK. Metabolic alterations and systemic inflammation in overweight/obese children with obstructive sleep apnea [J]. PLoS ONE. 2021;16(6):e0252353. HUANG YS, CHIN W C, GUILLEMINAULT C et al. Inflammatory Factors: Nonobese Pediatric Obstructive Sleep Apnea and Adenotonsillectomy [J]. J Clin Med, 2020, 9(4). SUN J, WANG M, ZHONG Y, et al. A Glb1-2A-mCherry reporter monitors systemic aging and predicts lifespan in middle-aged mice [J]. Nat Commun. 2022;13(1):7028. PHADKE M, KRYNETSKAIA N, MISHRA A, et al. Accelerated cellular senescence phenotype of GAPDH-depleted human lung carcinoma cells [J]. Biochem Biophys Res Commun. 2011;411(2):409–15. WU ZL, WANG K P, CHEN Y J, et al. Knocking down EGR1 inhibits nucleus pulposus cell senescence and mitochondrial damage through activation of PINK1-Parkin dependent mitophagy, thereby delaying intervertebral disc degeneration [J]. Volume 224. Free radical biology & medicine; 2024. pp. 9–22. TRIZZINO M, ZUCCO A, DELIARD S et al. EGR1 is a gatekeeper of inflammatory enhancers in human macrophages [J]. Sci Adv, 2021, 7(3). GUO X, SHI Y, DU P, et al. HMGB1/TLR4 promotes apoptosis and reduces autophagy of hippocampal neurons in diabetes combined with OSA [J]. Life Sci. 2019;239:117020. LV R, ZHAO Y, WANG X et al. GLP-1 analogue liraglutide attenuates CIH-induced cognitive deficits by inhibiting oxidative stress, neuroinflammation, and apoptosis via the Nrf2/HO-1 and MAPK/NF-κB signaling pathways [J]. Int Immunopharmacol, 2024, 142(Pt B): 113222. SHI Y, GUO X, ZHANG J et al. DNA binding protein HMGB1 secreted by activated microglia promotes the apoptosis of hippocampal neurons in diabetes complicated with OSA [J]. Brain, behavior, and immunity, 2018, 73: 482–92. OPRESKO PL, SHAY JW. Telomere-associated aging disorders [J]. Ageing Res Rev. 2017;33:52–66. MOHAMAD KAMAL N S, SAFUAN S, SHAMSUDDIN S, et al. Aging of the cells: Insight into cellular senescence and detection Methods [J]. Eur J Cell Biol. 2020;99(6):151108. CHEN MS, LEE R T, GARBERN JC. Senescence mechanisms and targets in the heart [J]. Cardiovascular Res. 2022;118(5):1173–87. LIU H, HE J, BAGHERI-YARMAND R, et al. Osteocyte CIITA aggravates osteolytic bone lesions in myeloma [J]. Nat Commun. 2022;13(1):3684. NAGAI-SINGER M A, MORRISON H A. ALLEN I C. NLRX1 Is a Multifaceted and Enigmatic Regulator of Immune System Function [J]. Front Immunol. 2019;10:2419. BAUER R, RAUCH I. The NAIP/NLRC4 inflammasome in infection and pathology [J]. Mol Aspects Med. 2020;76:100863. VANCE R E. The NAIP/NLRC4 inflammasomes [J]. Curr Opin Immunol. 2015;32:84–9. ARAYA J, TSUBOUCHI K, SATO N, et al. PRKN-regulated mitophagy and cellular senescence during COPD pathogenesis [J]. Autophagy. 2019;15(3):510–26. Additional Declarations No competing interests reported. Supplementary Files CellApoptosisinoverexpressedEGR1ATIIs.tiff CellApoptosisinNC.tiff CellApoptosisinNC2.tiff CellApoptosisinoverexpressedEGR1ATIIs2.tiff CellApoptosisinNC3.tiff CellApoptosisinoverexpressedEGR1ATIIs3.tiff GADPHinthreecells2.tif NLRX12.tif NLRX11.tif EGR1inthreecells2.tif CIITA2.tif NAIP2.tif GADPHinthreecells.tif GADPH.tif EGR1inthreecells.tif EGR1overexpression2.tif CIITA.tif NAIP.tif EGR1overexpression1.tif Cite Share Download PDF Status: Published Journal Publication published 19 Dec, 2025 Read the published version in BMC Biotechnology → Version 1 posted Editorial decision: Revision requested 09 Jun, 2025 Reviews received at journal 06 Jun, 2025 Reviews received at journal 04 Jun, 2025 Reviewers agreed at journal 30 May, 2025 Reviewers agreed at journal 30 May, 2025 Reviewers agreed at journal 30 May, 2025 Reviewers invited by journal 20 May, 2025 Editor assigned by journal 20 May, 2025 Editor invited by journal 19 May, 2025 Submission checks completed at journal 19 May, 2025 First submitted to journal 19 May, 2025 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-6563621","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":459560677,"identity":"692888d7-edfb-43a5-ad32-e39b39199b79","order_by":0,"name":"Caili Li","email":"","orcid":"","institution":"Tianjin Medical University General Hospital","correspondingAuthor":false,"prefix":"","firstName":"Caili","middleName":"","lastName":"Li","suffix":""},{"id":459560678,"identity":"461610e5-0439-43e8-8347-20b9a4bd0a48","order_by":1,"name":"Yuxiang Zhang","email":"","orcid":"","institution":"Tianjin Medical University General Hospital","correspondingAuthor":false,"prefix":"","firstName":"Yuxiang","middleName":"","lastName":"Zhang","suffix":""},{"id":459560679,"identity":"4909927d-f4cf-4fb9-b451-408944b46923","order_by":2,"name":"Xia Yang","email":"","orcid":"","institution":"Tianjin Medical University General Hospital","correspondingAuthor":false,"prefix":"","firstName":"Xia","middleName":"","lastName":"Yang","suffix":""},{"id":459560680,"identity":"c50a74d7-2c2b-45d7-8131-9f0bfae8eb47","order_by":3,"name":"Yubao Wang","email":"","orcid":"","institution":"Tianjin Medical University General Hospital","correspondingAuthor":false,"prefix":"","firstName":"Yubao","middleName":"","lastName":"Wang","suffix":""},{"id":459560681,"identity":"544d3b7f-be44-4bf7-b799-5053d08744b6","order_by":4,"name":"Haiyan Zhao","email":"","orcid":"","institution":"Tianjin Medical University General Hospital","correspondingAuthor":false,"prefix":"","firstName":"Haiyan","middleName":"","lastName":"Zhao","suffix":""},{"id":459560682,"identity":"ec2d72e4-8c9f-4ea2-9d5e-63f904ed33a6","order_by":5,"name":"Jing Feng","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABBklEQVRIie3PMUsDMRTA8RcOzuWVc8xh6X2FlEKLUKwfJXIgDhmcXA0UdCnOKS5+hE6dUwJ2Obw10KUiOKe4iNxgr7Rj046C+U9veD8eDyAU+pvFALpVD1wD9DE5kUeRzo5cN9ORPp7UmT6zl/79TInuBxas1Ute3s3tT4lggbiV2E+IEr22sqxzrpbcjJ8WSJ5llI6n+0lERZc6V11NrOamMVpg1NRx1PCQuCbcsfstecOYcj/BzRXLOCslN/itEQ8Rip93qSpYe2JhfUXmSHE29P6SPebTM3xlGSuL/Auri8FgPpy5lYfA6Q3b3hMcyEM9EenZX5fMl7tBA1T+5VAoFPqf/QJbpVn1577NZAAAAABJRU5ErkJggg==","orcid":"","institution":"Tianjin Medical University General Hospital","correspondingAuthor":true,"prefix":"","firstName":"Jing","middleName":"","lastName":"Feng","suffix":""}],"badges":[],"createdAt":"2025-04-30 09:53:03","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6563621/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6563621/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1186/s12896-025-01067-0","type":"published","date":"2025-12-19T15:57:25+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":83328778,"identity":"c89bdaf0-cdc7-4a73-b317-79d4a0bc959c","added_by":"auto","created_at":"2025-05-23 07:08:52","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":118135,"visible":true,"origin":"","legend":"\u003cp\u003eFlowchart of the comprehensive analysis of aging-related genes in OSA.\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/e93717c8e6d53eaad43e2b03.png"},{"id":83329065,"identity":"890e8bf2-0d4b-4980-9062-4f4cc1d6dd6c","added_by":"auto","created_at":"2025-05-23 07:16:47","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1509727,"visible":true,"origin":"","legend":"\u003cp\u003eVolcano plots and heatmaps.(A) Volcano plot of differentially expressed genes (DEDEGs). Genes were filtered based on |logFC| \u0026gt; 0.58 and FDR \u0026lt; 0.05. Downregulated genes in OSA are represented by green dots, while upregulated DEDEGs are highlighted in red.(B) Heatmap displaying the expression patterns of DEDEGs in healthy controls and OSA patients.\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/4b68ac3ae7a60ba4e34592cc.png"},{"id":83328696,"identity":"37413b0f-8181-4831-b6ca-772a88d5adef","added_by":"auto","created_at":"2025-05-23 07:08:46","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":2042788,"visible":true,"origin":"","legend":"\u003cp\u003eWGCNA of differentially expressed genes.(A) Sample clustering dendrogram with clinical trait heatmap.(B) Determination of soft-thresholding power, with the red line indicating a scale-free topology fit (R² = 0.9).(C) Module-trait associations showing correlation coefficients (upper values) and statistical significance (lower values).(D) Gene module clustering dendrogram, with color-coded modules and grey representing unassigned genes.(E) Association of trait relevance with network connectivity in the green co-expression module.\u003c/p\u003e","description":"","filename":"image3.png","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/19ea4a62d542844ee4ca4ae9.png"},{"id":83328689,"identity":"f47d31ae-7a52-445a-b577-413ce3563a0d","added_by":"auto","created_at":"2025-05-23 07:08:45","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":2276161,"visible":true,"origin":"","legend":"\u003cp\u003eFunctional and molecular profiling of OSA-associated DEGs and pathway enrichment.(A) Venn diagram illustrating the overlap between co-expressed gene clusters and differentially expressed DEGs identified in OSA.(B)GO enrichment across BP, CC, and MF domains. A graphical bubble chart displays the five most enriched terms with statistical significance, where bubble diameter correlates with the quantity of overlapping genes (larger = greater enrichment), and a chromatic gradient indicates the degree of statistical confidence (darker shades = more stringent FDR-corrected p-values). (C) KEGG pathway analysis, with a bubble chart ranking the 20 most enriched pathways based on significance. (D) Network visualization of functional associations among enriched KEGG pathways.\u003c/p\u003e","description":"","filename":"image4.png","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/b9106fd8da3b20a878421b8e.png"},{"id":83329748,"identity":"2a1d9744-d807-401f-a441-49211e627941","added_by":"auto","created_at":"2025-05-23 07:24:47","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":1524772,"visible":true,"origin":"","legend":"\u003cp\u003eCorrelation analysis of OSA-ARDEGs. (A) Venn diagram depicting intersections of aging-associated genes and DEGs derived from the co-expression network. (B) Correlation matrix depicting the relationships among OSA-ARDEGs. (C) Chromosomal localization of the four identified OSA-ARDEGs.\u003c/p\u003e","description":"","filename":"image5.png","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/c2ba38bf25f289cdc79740d4.png"},{"id":83328782,"identity":"798abcb4-178e-4d8f-b4b5-501d27b3ca43","added_by":"auto","created_at":"2025-05-23 07:08:52","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":2981702,"visible":true,"origin":"","legend":"\u003cp\u003eImmune Cell Correlation Analysis.(A) Heatmap visualization of 22 immune cell subtype distributions across OSA patients and healthy controls.(B) Correlation matrix of immune cell interactions, with blue and red gradients representing positive and negative associations, respectively, where color intensity reflects correlation strength.(C) Comparative box plot analysis of immune cell population distributions between OSA patients and controls, with statistical significance denoted as *p \u0026lt; 0.05 and **p \u0026lt; 0.01.\u003c/p\u003e","description":"","filename":"image6.png","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/64d7df520115c9c0fab6525a.png"},{"id":83329069,"identity":"a9026a55-aa6f-4e08-b8ef-25cc50354440","added_by":"auto","created_at":"2025-05-23 07:16:47","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":2654894,"visible":true,"origin":"","legend":"\u003cp\u003eCorrelation of OSA-ARDEGs with Immune Cell Infiltration.Association patterns between OSA-associated differentially expressed genes (EGR1 (A), FOS (B), GLB1 (C), JUN (D)) and infiltrating immune cell populations are shown. Dot size corresponds to correlation coefficient magnitude, while color intensity represents statistical significance levels.\u003c/p\u003e","description":"","filename":"image7.png","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/6f97857000f4c807b9ed2064.png"},{"id":83328795,"identity":"2915d147-92ab-4a1f-b0d0-738621ded9e4","added_by":"auto","created_at":"2025-05-23 07:08:54","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":1405576,"visible":true,"origin":"","legend":"\u003cp\u003eCharacterization of OSA-ARDEGs and their immune infiltration patterns. (A) Consensus matrix for k=2 clustering. (B) Cumulative distribution curves (CDF) for clustering stability across k=2-6. (C) Delta area shifts under CDF curves for varying k. (D) Consensus scores for k=2-6 based subgroup assignments. (E) PCA plot illustrating subtype 1 (blue) versus subtype 2 (yellow) sample distribution. (F) Split-violin plots comparing expression profiles of EGR1/FOS/GLB1/JUN between subtypes. (G) Box plots comparing immune infiltration levels (subtype 1 vs. 2) (* p\u0026lt;0.05, ** p\u0026lt;0.01, *** p\u0026lt;0.001).\u003c/p\u003e","description":"","filename":"image8.png","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/3084b5a6132f97c371f396d9.png"},{"id":83328765,"identity":"93c092fb-9433-4f7c-bfbc-583e0600f31d","added_by":"auto","created_at":"2025-05-23 07:08:51","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":2834198,"visible":true,"origin":"","legend":"\u003cp\u003eMolecular signatures and pathway alterations in aging-associated subtypes. (A) Differential gene expression patterns visualized through volcano plots. (B) Heatmap representation of subtype-specific gene expression profiles. (C-D) Functional pathway analysis showing GO and KEGG pathway alterations, organized by GSVA-derived t-values. Subtype 2 exhibits distinct pathway regulation patterns, with blue bars representing upregulated processes and green/orange bars indicating downregulated biological functions and signaling cascades.\u003c/p\u003e","description":"","filename":"image9.png","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/5cd2a17997488218aa2d13b1.png"},{"id":83328724,"identity":"74a7ca01-c70e-4e5e-9dee-590fac6717eb","added_by":"auto","created_at":"2025-05-23 07:08:48","extension":"png","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":728393,"visible":true,"origin":"","legend":"\u003cp\u003eDiscovery of Core OSA-Linked Aging-Related Genes via ML Algorithms. (A) LASSO regression coefficient trajectories, with optimal λ parameter demarcated by dashed vertical lines. (B) Cross-validation error curves for LASSO model parameter optimization, with individual gene trajectories displayed. (C-D) SVM-RFE algorithm performance metrics showing maximum accuracy and minimum error thresholds for optimal gene selection. (E) Feature importance ranking of OSA-ARDEGs. (F) Random forest error rate as a function of tree number. (G) Composite Venn diagram depicting overlapping genes selected via LASSO, Random Forest, and SVM-RFE methodologies.\u003c/p\u003e","description":"","filename":"image10.png","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/be4d7285754bca590fc26866.png"},{"id":83328752,"identity":"22a4456b-2477-46e7-971f-801f5c69740b","added_by":"auto","created_at":"2025-05-23 07:08:51","extension":"png","order_by":11,"title":"Figure 11","display":"","copyAsset":false,"role":"figure","size":1194963,"visible":true,"origin":"","legend":"\u003cp\u003edisplays the performance assessment of six ML algorithms via ROC curve evaluation: (A) AdaBoost, (B) Decision Tree, (C) KNN, (D) LightGBM, (E) Naïve Bayes, and (F) XGBoost. Model efficacy was quantified by AUC metrics.\u003c/p\u003e","description":"","filename":"image11.png","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/1e5b7ace0ebafb272485bed7.png"},{"id":83328727,"identity":"bc73a294-8be7-40c3-b86e-f2a9fa1064a1","added_by":"auto","created_at":"2025-05-23 07:08:49","extension":"png","order_by":12,"title":"Figure 12","display":"","copyAsset":false,"role":"figure","size":2843514,"visible":true,"origin":"","legend":"\u003cp\u003ePredictive feature relevance visualizations in ML classifiers. (A) AdaBoost classifier; (B) Decision Tree model; (C) KNN algorithm; (D) LightGBM framework; (E) Naïve Bayes predictor; (F) XGBoost system.\u003c/p\u003e","description":"","filename":"image12.png","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/fbaf4842625849323255c744.png"},{"id":83328769,"identity":"ccd97646-f0d0-41e5-bac4-986af7db8d95","added_by":"auto","created_at":"2025-05-23 07:08:51","extension":"png","order_by":13,"title":"Figure 13","display":"","copyAsset":false,"role":"figure","size":344518,"visible":true,"origin":"","legend":"\u003cp\u003eAssessing the Risk of OSA. (A) Comparative distribution analysis of Hub OSA-ARDEGs in OSA patients versus healthy controls. (B) Nomogram construction for OSA risk prediction in the training cohort, incorporating characteristic gene expression profiles. (C) Calibration curve analysis demonstrating the nomogram's predictive accuracy in the training set. (D) Decision curve analysis evaluating the clinical utility of the nomogram in the training cohort.\u003c/p\u003e","description":"","filename":"image13.png","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/a946282b98177f0e7d3f4f5a.png"},{"id":83329076,"identity":"38ca495f-b477-4dca-88d9-50ab2877682b","added_by":"auto","created_at":"2025-05-23 07:16:49","extension":"png","order_by":14,"title":"Figure 14","display":"","copyAsset":false,"role":"figure","size":288455,"visible":true,"origin":"","legend":"\u003cp\u003eDiagnostic Model Validation. (A) ROC evaluation of core OSA-linked aging-related DEGs in the validation cohort. (B) ROC curve-based assessment of the model’s diagnostic capability within the testing dataset. (C) ROC curve assessment of Hub OSA-ARDEGs in the external validation cohort. (D) Model diagnostic performance verification in the validation set using ROC curve analysis.\u003c/p\u003e","description":"","filename":"image14.png","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/a72b5e5b179803e4661443a9.png"},{"id":83328695,"identity":"8e4cbf4c-0a63-4ba4-811c-e2ffb44bd194","added_by":"auto","created_at":"2025-05-23 07:08:46","extension":"png","order_by":15,"title":"Figure 15","display":"","copyAsset":false,"role":"figure","size":3371567,"visible":true,"origin":"","legend":"\u003cp\u003eGSEA Results for Key Genes.GLB1 GSEA enrichment analysis in GO (A) and KEGG (B).EGR1 GSEA enrichment analysis in GO (C) and KEGG (D).\u003c/p\u003e","description":"","filename":"image15.png","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/00d6ea977f94390f5cfe95a9.png"},{"id":83329094,"identity":"726290ea-dece-42ae-8163-7456567ab127","added_by":"auto","created_at":"2025-05-23 07:16:51","extension":"png","order_by":16,"title":"Figure 16","display":"","copyAsset":false,"role":"figure","size":2505341,"visible":true,"origin":"","legend":"\u003cp\u003eFunctional Enrichment Analysis of Key Genes. (A-B) GSVA results showing GO and KEGG pathway enrichment patterns associated with EGR1 expression. (C-D) GSVA assessment of GO/KEGG pathway shifts associated with GLB1 expression. Pathway enrichment rankings are rank-ordered based on GSVA-calculated t-statistics. Red and blue bars denote upregulated (biological processes) and downregulated (signaling cascades) pathways, signifying subtype 2-specific dysregulation.\u003c/p\u003e","description":"","filename":"image16.png","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/f2b3be1764fcbb0741139e66.png"},{"id":83328730,"identity":"f12a1484-f208-4e56-b0b2-8a108c05960c","added_by":"auto","created_at":"2025-05-23 07:08:49","extension":"png","order_by":17,"title":"Figure 17","display":"","copyAsset":false,"role":"figure","size":408317,"visible":true,"origin":"","legend":"\u003cp\u003eBasal Protein Expression and Overexpression Verification. (A) Baseline EGR1 protein levels across AT2 alveolar epithelial cells (type II), BEAS-2B bronchial epithelial cells, and MRC-5 fetal lung fibroblasts (triplicate experiments). (B)Immunoblot validation of lysates post-transfection with the EGR1 overexpression construct (triplicate experiments). *** p\u0026lt;0.001.\u003c/p\u003e","description":"","filename":"image17.png","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/df0c8240d7bb276b1cef9769.png"},{"id":83329064,"identity":"c79fd290-4caf-48ae-a725-5fcc957bd70f","added_by":"auto","created_at":"2025-05-23 07:16:46","extension":"png","order_by":18,"title":"Figure 18","display":"","copyAsset":false,"role":"figure","size":1075504,"visible":true,"origin":"","legend":"\u003cp\u003eScreening and Verification of Senescence-Related Genes. (A) Verification of the accuracy of 12 Hub OSA-ARDEGs candidate genes (CIITA, NAIP, NOD1, NOD2, NLRC5, NLRP1, NLRP3, NLRP4, NLRP5, NLRP13, NLRP14, NLRX1) by qPCR. (B) Selection of differentially expressed genes (CIITA, NAIP, NLRX1) for Western blot (WB) analysis (N=3). (ns indicates no significance; ** p \u0026lt; 0.01; *** p \u0026lt; 0.001).\u003c/p\u003e","description":"","filename":"image18.png","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/5a4c8fdfe1ff3740bab912e7.png"},{"id":83329072,"identity":"214c17cc-1080-43f4-82b7-5f129cf79afa","added_by":"auto","created_at":"2025-05-23 07:16:48","extension":"png","order_by":19,"title":"Figure 19","display":"","copyAsset":false,"role":"figure","size":2549687,"visible":true,"origin":"","legend":"\u003cp\u003eEGR1 Promotes Cell Apoptosis, Senescence, and DNA Damage.C (A) Annexin V-FITC flow cytometric analysis of apoptosis across experimental cohorts (triplicate experiments). (B) Senescence evaluation via SA-β-Gal staining. Blue chromogenic substrate denotes senescent cells (triplicate experiments). (C) Quantification of γ-H2AX foci in wild-type vs. EGR1-overexpressing AT2 cells. Immunofluorescence revealed elevated γ-H2AX in EGR1-overexpressing cells versus controls (nuclear counterstain: DAPI [blue]; γ-H2AX [green]) (triplicate experiments). (D) Telomere length assessment in non-modified controls (NC) and EGR1-overexpressing AT2 cells via qRT-PCR (triplicate experiments; ns: non-significant, *** p\u0026lt;0.001).\u003c/p\u003e","description":"","filename":"image19.png","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/4e518c24d24c9285e571084e.png"},{"id":83329075,"identity":"c0366bc9-cfb2-4522-a6dc-338ff93dfd36","added_by":"auto","created_at":"2025-05-23 07:16:49","extension":"png","order_by":20,"title":"Figure 20","display":"","copyAsset":false,"role":"figure","size":327758,"visible":true,"origin":"","legend":"\u003cp\u003eELISA Analysis. Secreted levels of IL-6, IL-17A, IL-1β, and TNF-αin cell culture supernatants from non-modified control (NC) and EGR1-overexpressing AT2 cells were assayed via ELISA. Data were analyzed via Student’s t-test (triplicates; *** p\u0026lt;0.001).\u003c/p\u003e","description":"","filename":"image20.png","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/f98e2caf113a8791b911dca5.png"},{"id":98813905,"identity":"9d887e76-2f38-4885-ba8e-fccb1583da10","added_by":"auto","created_at":"2025-12-22 16:07:23","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":29974177,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/fd4719da-3f27-4af1-a340-cfba72f02adb.pdf"},{"id":83329068,"identity":"dfa9c5cb-9d43-401c-9e66-81bcf197088f","added_by":"auto","created_at":"2025-05-23 07:16:47","extension":"tiff","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":89160,"visible":true,"origin":"","legend":"","description":"","filename":"CellApoptosisinoverexpressedEGR1ATIIs.tiff","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/962a4049173ae0e7b9da93fd.tiff"},{"id":83328717,"identity":"4f1e6dd5-3a3f-4cd4-ad5e-c82ec908f7f9","added_by":"auto","created_at":"2025-05-23 07:08:48","extension":"tiff","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":87492,"visible":true,"origin":"","legend":"","description":"","filename":"CellApoptosisinNC.tiff","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/a6bf6b89e2a13262ea4954d1.tiff"},{"id":83328780,"identity":"e5bff935-aacb-49dc-8986-8779bd0d886b","added_by":"auto","created_at":"2025-05-23 07:08:52","extension":"tiff","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":126984,"visible":true,"origin":"","legend":"","description":"","filename":"CellApoptosisinNC2.tiff","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/4634bf0fe030e9e4eb7215e3.tiff"},{"id":83329091,"identity":"1fe5cd0e-24eb-4762-b594-2efd3245e9b0","added_by":"auto","created_at":"2025-05-23 07:16:51","extension":"tiff","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":129376,"visible":true,"origin":"","legend":"","description":"","filename":"CellApoptosisinoverexpressedEGR1ATIIs2.tiff","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/9a542771658ee9604eebdb9a.tiff"},{"id":83328763,"identity":"4649af15-21f1-4ee4-a6a5-549f52f63f3e","added_by":"auto","created_at":"2025-05-23 07:08:51","extension":"tiff","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":155704,"visible":true,"origin":"","legend":"","description":"","filename":"CellApoptosisinNC3.tiff","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/24083062eddb1510b56c4df8.tiff"},{"id":83328768,"identity":"0be4c68c-7df0-41e6-8d2e-92680d0fa69a","added_by":"auto","created_at":"2025-05-23 07:08:51","extension":"tiff","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":169442,"visible":true,"origin":"","legend":"","description":"","filename":"CellApoptosisinoverexpressedEGR1ATIIs3.tiff","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/4c0a3604723b8fe5a5086ca6.tiff"},{"id":83329077,"identity":"11f84725-d6f9-4ff2-b118-050b3b45df73","added_by":"auto","created_at":"2025-05-23 07:16:49","extension":"tif","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":1473456,"visible":true,"origin":"","legend":"","description":"","filename":"GADPHinthreecells2.tif","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/28d8d61e2fecfb923c878dc8.tif"},{"id":83329062,"identity":"62f4e1f5-7cb7-4d17-8812-968802f93139","added_by":"auto","created_at":"2025-05-23 07:16:45","extension":"tif","order_by":7,"title":"","display":"","copyAsset":false,"role":"supplement","size":1473456,"visible":true,"origin":"","legend":"","description":"","filename":"NLRX12.tif","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/540b25df67b81b6732645e52.tif"},{"id":83328705,"identity":"2d67f863-eb8e-444d-b168-0620c89319ec","added_by":"auto","created_at":"2025-05-23 07:08:47","extension":"tif","order_by":8,"title":"","display":"","copyAsset":false,"role":"supplement","size":1473456,"visible":true,"origin":"","legend":"","description":"","filename":"NLRX11.tif","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/ead510f69c690deaf58abdc9.tif"},{"id":83329063,"identity":"eef1f2bc-c3a5-48b9-8153-aaabcdf61ce9","added_by":"auto","created_at":"2025-05-23 07:16:46","extension":"tif","order_by":9,"title":"","display":"","copyAsset":false,"role":"supplement","size":1473456,"visible":true,"origin":"","legend":"","description":"","filename":"EGR1inthreecells2.tif","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/0eb68960225f1bb37f9ffd19.tif"},{"id":83328750,"identity":"e4d2c065-d207-415a-ae63-36c482a2a149","added_by":"auto","created_at":"2025-05-23 07:08:50","extension":"tif","order_by":10,"title":"","display":"","copyAsset":false,"role":"supplement","size":1473456,"visible":true,"origin":"","legend":"","description":"","filename":"CIITA2.tif","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/43baec8ce504b0e175af3c9f.tif"},{"id":83329088,"identity":"ba169f46-ed14-457a-bfc7-5c930f5285a9","added_by":"auto","created_at":"2025-05-23 07:16:50","extension":"tif","order_by":11,"title":"","display":"","copyAsset":false,"role":"supplement","size":1473456,"visible":true,"origin":"","legend":"","description":"","filename":"NAIP2.tif","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/7ece4e56ec626e46be95c4cc.tif"},{"id":83328790,"identity":"7d623a7f-3f90-4fe6-ac89-6b14eb7db0f2","added_by":"auto","created_at":"2025-05-23 07:08:54","extension":"tif","order_by":12,"title":"","display":"","copyAsset":false,"role":"supplement","size":1473456,"visible":true,"origin":"","legend":"","description":"","filename":"GADPHinthreecells.tif","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/888e5531cb6f7399bbd38046.tif"},{"id":83329081,"identity":"ec08dd7b-9039-4e69-95ba-1f34b8ffdaa0","added_by":"auto","created_at":"2025-05-23 07:16:50","extension":"tif","order_by":13,"title":"","display":"","copyAsset":false,"role":"supplement","size":1473456,"visible":true,"origin":"","legend":"","description":"","filename":"GADPH.tif","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/4330a2c20eb2175b347ba8fc.tif"},{"id":83328732,"identity":"81070742-3857-4387-b264-7f35c7835cb6","added_by":"auto","created_at":"2025-05-23 07:08:49","extension":"tif","order_by":14,"title":"","display":"","copyAsset":false,"role":"supplement","size":1473456,"visible":true,"origin":"","legend":"","description":"","filename":"EGR1inthreecells.tif","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/8e5dfdbf82426c2219c50015.tif"},{"id":83328701,"identity":"75375dbe-d89e-4155-89f9-35ee882593b7","added_by":"auto","created_at":"2025-05-23 07:08:47","extension":"tif","order_by":15,"title":"","display":"","copyAsset":false,"role":"supplement","size":1473456,"visible":true,"origin":"","legend":"","description":"","filename":"EGR1overexpression2.tif","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/b4b75883a2b569030e27758b.tif"},{"id":83328797,"identity":"de3e079b-71b4-40eb-b529-16e91a278ae8","added_by":"auto","created_at":"2025-05-23 07:08:54","extension":"tif","order_by":16,"title":"","display":"","copyAsset":false,"role":"supplement","size":1473456,"visible":true,"origin":"","legend":"","description":"","filename":"CIITA.tif","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/f728c42b4334220971620315.tif"},{"id":83329098,"identity":"100be98d-cb8a-448d-a7eb-5dbaf36e6c82","added_by":"auto","created_at":"2025-05-23 07:16:52","extension":"tif","order_by":17,"title":"","display":"","copyAsset":false,"role":"supplement","size":1473456,"visible":true,"origin":"","legend":"","description":"","filename":"NAIP.tif","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/6a2cb2cefe5ed155d005dc9b.tif"},{"id":83328711,"identity":"ea85f89e-125f-4f98-8706-4aa96096a20e","added_by":"auto","created_at":"2025-05-23 07:08:47","extension":"tif","order_by":18,"title":"","display":"","copyAsset":false,"role":"supplement","size":1473456,"visible":true,"origin":"","legend":"","description":"","filename":"EGR1overexpression1.tif","url":"https://assets-eu.researchsquare.com/files/rs-6563621/v1/95bb7bd2a7259c5e096bc10a.tif"}],"financialInterests":"No competing interests reported.","formattedTitle":"Unveiling the role of EGR1 and hub senescencerelated genes in Type II alveolar epithelial cells senescence for Obstructive sleep Apnea","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eA prevalent breathing condition associated with sleep, obstructive sleep apnea (OSA) is marked by recurring total or partial obstructions in the upper airway while sleeping. These events trigger pronounced drops in blood oxygen levels and disruptions to sleep continuity\u003csup\u003e[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]\u003c/sup\u003e. OSA affects approximately 9\u0026ndash;38% of adults worldwide, with prevalence rates demonstrating a positive correlation with advancing age and elevated body mass index\u003csup\u003e[\u003cspan additionalcitationids=\"CR3\" citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]\u003c/sup\u003e. Patients with OSA often exhibit various symptoms such as loud snoring, observed apneas, nocturnal arousals, morning headaches, and excessive daytime somnolence, which significantly degrade the quality of life. Individuals with OSA suffer from multiple symptoms that significantly impact their quality of life, including loud snoring, observed pauses in breathing, frequent waking, morning headaches, and excessive daytime sleepiness. These symptoms typically come with various neuropsychiatric issues like concentration difficulties, memory impairments, and mood fluctuations, all linked to the fragmented sleep and hypoxemia associated with the disorder \u003csup\u003e[\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]\u003c/sup\u003e. OSA independently correlates with cardiovascular conditions, including hypertension, heart failure, atrial fibrillation, and stroke. It has been associated with metabolic conditions such as type 2 diabetes and dyslipidemia, while also heightening the likelihood of vehicular collisions resulting from daytime drowsiness\u003csup\u003e[\u003cspan additionalcitationids=\"CR7\" citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]\u003c/sup\u003e. Chronic untreated OSA leads to reduced neurocognitive function and overall life expectancy, highlighting the necessity for effective diagnostic and treatment approaches to alleviate these risks\u003csup\u003e[\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eAdvancing age represents a major risk factor for OSA\u003csup\u003e[\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]\u003c/sup\u003e. The relationship between OSA and aging holds significant clinical relevance, especially given the worldwide shift toward an aging population. Age-related physiological decline involves complex biological mechanisms marked by progressive weakening of cellular repair and homeostasis, heightening susceptibility to numerous age-associated pathologies\u003csup\u003e[\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]\u003c/sup\u003e. Biomarkers like telomere attrition, inflammation, alterations in synthesis pathways, and hormonal dysregulation are core aspects of aging.These factors not only advance typical aging phenotypes but also impact diverse pathophysiological pathways. Regarding sleep and respiratory functions, aging correlates with increased neck fat accumulation, decreased airway muscle tone, and changes in lung and chest wall mechanics, all of which contribute to a higher prevalence of OSA in older individuals\u003csup\u003e[\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]\u003c/sup\u003e. Aging alters sleep architecture, diminishing deep sleep phases and increasing nocturnal awakenings\u0026mdash;changes that can worsen the severity and outcomes of OSA\u003csup\u003e[\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe shared mechanisms between aging and OSA include systemic inflammation, oxidative stress, and endothelial dysfunction. Aging impacts OSA by increasing vulnerability to the hypoxic challenges of OSA due to a decline in physiological reserves associated with aging. Aging also affects the neural control of upper airway muscles and respiratory drive during sleep, which may exacerbate the severity of OSA or complicate the management of elderly patients \u003csup\u003e[\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]\u003c/sup\u003e. The interaction between aging and OSA is recognized, but the exact mechanisms by which aging regulates OSA pathogenesis via molecular networks remain unclear. This significantly hampers the development of targeted interventions for elderly OSA patients. This underscores the critical requirement for additional studies to comprehensively clarify these pathways\u003csup\u003e[\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]\u003c/sup\u003e. The zinc finger-containing regulatory protein EGR1 (early growth response 1) is swiftly activated by external signals and orchestrates the regulation of target genes. It has been associated with various diseases, including acute kidney injury\u003csup\u003e[\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]\u003c/sup\u003e and multiple malignancies such as gastric cancer \u003csup\u003e[\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]\u003c/sup\u003e, pancreatic cancer\u003csup\u003e[\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]\u003c/sup\u003e, and endometrial cancer \u003csup\u003e[\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]\u003c/sup\u003e. However, its role in the pathogenesis of OSA remains to be clarified.\u003c/p\u003e \u003cp\u003eTo establish a gene co-expression network, WGCNA method was applied, which facilitated the detection of central OSA-associated hub genes. Differences in immune cell infiltration levels across individuals with OSA were assessed using ssGSEA. Meanwhile, GSVA was applied to compare functional and pathway variations between distinct OSA subtypes, further uncovering the molecular heterogeneity of OSA.Within machine learning applications, LASSO and Random Forest approaches were applied to identify critical variables and construct predictive models. Ultimately, key discoveries from bioinformatics analyses were validated in the lab through experimental validation steps. Through the application of these methods, the research delivers an in-depth exploration of the biological pathways linked to aging-associated OSA. By uncovering prospective biomarkers and targets for therapy, this work establishes a foundation for advancing diagnostic strategies and treatment approaches for the disorder.\u003c/p\u003e"},{"header":"2. Materials and Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1. Data Collection and Preprocessing\u003c/h2\u003e \u003cp\u003eThis study utilized the OSA gene expression profile database, including GSE135917 \u003csup\u003e[\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]\u003c/sup\u003e GSE38792\u003csup\u003e[\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]\u003c/sup\u003e, and GSE75097\u003csup\u003e[\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]\u003c/sup\u003e. The GSE135917 and GSE38792 datasets utilize the GPL6244 platform (Affymetrix Human Gene 1.0 ST Array for genome-wide expression profiling). In contrast, the GSE75097 dataset originated from the GPL10904 platform (Illumina HumanHT-12 V4.0 beadchip). Earlier research has shown that continuous positive airway pressure (CPAP) therapy in OSA patients markedly alters transcriptional activity\u003csup\u003e[\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e]\u003c/sup\u003e, with evidence suggesting restoration of gene expression patterns toward baseline levels after intervention. To minimize possible biases arising from CPAP-induced transcriptional alterations affecting OSA-related molecular signatures, 24 individuals receiving CPAP therapy were excluded from the GSE135917 cohort. These retained samples were then combined with data from the GSE38792 dataset for subsequent analysis.\u003c/p\u003e \u003cp\u003eTechnical batch effects were adjusted by applying the \"removeBatchEffect\" function from the limma R package, aiming to achieve consistent data quality between integrated datasets. The combined dataset comprised 16 healthy controls and 44 OSA patients. During analysis of the GSE75097 dataset, background adjustments and data normalization were performed via the neqc method available in the limma R package. 14 CPAP-treated OSA patients were excluded from the analysis to maintain consistency with our exclusion criteria. The final validation cohort consisted of 6 healthy controls and 28 OSA patients. The comprehensive data processing workflow is presented in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Differential Gene Expression Analysis\u003c/h2\u003e \u003cp\u003eTo identify differentially expressed genes (DEGs) between OSA patients and healthy controls, the \"limma\" R package\u003csup\u003e[\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]\u003c/sup\u003e was utilized for statistical analysis. The merged gene expression dataset was analyzed using stringent criteria, with |log2 fold change (logFC)| \u0026gt; 0.58 and false discovery rate (FDR)\u0026thinsp;\u0026lt;\u0026thinsp;0.05, to identify OSA-related DEGs (OSA-DEGs). The results were visualized using volcano plots and heatmaps, which were generated with the \"ComplexHeatmap\" package and the \"ggplot2\" package \u003csup\u003e[\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3. Weighted Gene Co-Expression Network Analysis\u003c/h2\u003e \u003cp\u003eThis study applied a weighted gene co-expression network analytical framework (WGCNA) to pinpoint hub genes implicated in OSA and connected to biological aging mechanisms. A scale-free co-expression network was developed for genes exhibiting high variability within the training cohort, implemented through the WGCNA package in R\u003csup\u003e[\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]\u003c/sup\u003e.Following initial quality assessment using the \"goodSamplesGenes\" method, an optimal soft threshold (β\u0026thinsp;=\u0026thinsp;6) was selected via the \"pickSoftThreshold\" algorithm.The gene co-expression network was developed by initially computing an adjacency matrix based on Pearson correlation measures, which was later converted into a topological overlap matrix (TOM) to assess network interconnectedness. Functional modules within the network were identified via hierarchical clustering integrated with dynamic branch-cutting methodology, establishing a lower limit of 50 genes per module and a module consolidation cutoff set at 0.15 similarity. Statistical relationships between module eigengenes (MEs) and phenotypic characteristics were systematically evaluated, while key molecular candidates were selected through comparative analysis of intra-modular connectivity (MM) indices and biological relevance (GS) parameters.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e2.4. Functional Enrichment Analysis\u003c/h2\u003e \u003cp\u003eTo investigate the molecular roles of genes linked to obstructive sleep apnea (OSA), Gene Ontology (GO) and KEGG pathway enrichment analyses were conducted in R using the \"clusterProfiler\" package\u003csup\u003e[\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]\u003c/sup\u003e.A Benjamini-Hochberg-adjusted p-value cutoff of \u0026lt;\u0026thinsp;0.05 was applied to define statistically significant terms. Enrichment results for GO categories, such as biological processes, cellular components, and molecular functions\u0026mdash;were plotted as dot diagrams. KEGG pathway enrichment findings were further mapped with the ​\"pathview\" package\u003csup\u003e[\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e]\u003c/sup\u003e.This integrative approach yielded a holistic perspective on the biological mechanisms and pathway interactions involving OSA-associated genes.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e2.5 Download and Organization of Aging-Related Genes\u003c/h2\u003e \u003cp\u003eFollowing established protocols\u003csup\u003e[\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]\u003c/sup\u003e, A comprehensive collection of human aging-related genes (ARGs) was sourced from the Human Aging Genomic Resources (HAGR) database\u003csup\u003e[\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]\u003c/sup\u003e. This dataset included 307 genes from the GenAge database and 279 genes from the CellAge database. Upon integration and de-duplication, a consolidated list of 543 unique ARGs was established.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e2.6 Evaluation of Immune Infiltration in OSA Patients\u003c/h2\u003e \u003cp\u003eEnrichment levels for 19 immune cell populations and 13 immune-related processes across healthy and OSA cohorts were assessed via the \"GSVA\" package in R, with results visualized as violin-style plots using the \"vioplot\" R package\u003csup\u003e[\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e]\u003c/sup\u003e. To investigate linkages between OSA-ARDEGs (aging-associated differentially expressed genes in OSA) and immune activity, Spearman's correlation coefficient was applied to evaluate connections with both immune cell subsets and functional immunological pathways.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e2.7 Unsupervised Clustering Analysis and Gene Set Variation Analysis\u003c/h2\u003e \u003cp\u003eSamples from OSA cohorts were clustered in an unsupervised manner utilizing the \"ConsensusClusterPlus\" toolkit in R\u003csup\u003e[\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e]\u003c/sup\u003e, employing the k-means algorithm across 1,000 resampling iterations with an upper cluster limit (k\u0026thinsp;=\u0026thinsp;6). Optimal cluster partitioning (k\u0026thinsp;=\u0026thinsp;2) was selected through systematic evaluation of agreement matrices, cumulative density function (CDF) curves, and partition robustness indices. To investigate functional divergence across molecular subgroups, Gene Set Variation Analysis (GSVA) was implemented. Pathway enrichment patterns were examined via the MSigDB repository, integrating canonical KEGG pathways and Gene Ontology-derived biological processes. A significance threshold of |GSVA score| \u0026gt; 2 (t-test) was established to identify differentially enriched pathways and immune microenvironment patterns.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e2.8 Feature Selection, Modeling, and Validation\u003c/h2\u003e \u003cp\u003eCritical variables were identified via LASSO regression implemented in the \"glmnet\" R package. A prediction model was then constructed using the random forest method. The data were divided randomly, with 80% allocated for training and 20% reserved for validation. Within the LASSO framework, variables exhibiting non-zero coefficients were selected to build the random forest-based predictive signature.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e2.9 Using Machine Learning (ML) Models to Identify Characteristic Biomarkers\u003c/h2\u003e \u003cp\u003eSix machine learning approaches\u0026mdash;AdaBoost, Decision Trees (DT), k-Nearest Neighbors (KNN), LightGBM, Na\u0026iuml;ve Bayes, and XGBoost\u0026mdash;were applied to build diagnostic models. The data were partitioned randomly, allocating 70% for model training and 30% for testing. AUC values served as the primary metric to evaluate the ability of OSA-associated DEGs to distinguish patients from healthy individuals. Model execution relied on multiple R packages: AdaBoost, DT, and KNN utilized the \"caret\" package; LightGBM leveraged \"LightGBM\" and \"tidymodels\"; XGBoost employed the \"XGBoost\" package; and Na\u0026iuml;ve Bayes was operationalized via \"e1071\". The \"DALEX\" package supported explainability assessments to improve model transparency.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e2.10 Construction and Evaluation of Binary Logistic Regression (LR) Model\u003c/h2\u003e \u003cp\u003eThe investigation utilized a binary logistic regression (LR) model to establish association metrics for aging-associated OSA-related differentially expressed genes (OSA-ARDEGs). Model calibration plots were created using the ​RMS package to assess predictive performance. The prognostic value of risk stratification metrics was assessed via decision curve analysis (DCA), calibration plots, and area under the curve (AUC) measurements. Model generalizability and robustness were evaluated using both internal and external validation cohorts. External validation utilized the GSE75097 transcriptomic dataset acquired from the Gene Expression Omnibus (GEO) repository, while a separate internal validation cohort comprising independent samples was established for additional verification.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e2.11 Gene Set Enrichment Analysis (GSEA)\u003c/h2\u003e \u003cp\u003eThis study performed gene set enrichment analysis (GSEA) on key OSA-ORDEGs to further investigate the functional roles of EGR1 and GLB1 in OSA pathogenesis. Pathway enrichment analysis of key genes is conducted using GSEA, identifying signaling pathways implicated in the development of OSA through GSEA.Sort according to the standardized enrichment score (NES), with a P-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05 and a corrected absolute enrichment score (| NES |)\u0026thinsp;\u0026gt;\u0026thinsp;1 as the screening criteria.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003e2.12 Gene Set Variation Analysis (GSVA)\u003c/h2\u003e \u003cp\u003eGene Set Variation Analysis (GSVA) operates as a non-parametric, unsupervised analytical approach. Unlike GSEA, this method eliminates the need for predefined sample categorization and directly computes enrichment scores for defined gene sets at the individual sample level. The fundamental concept of GSVA is to transform gene expression data from an expression matrix of individual genes to one featured by gene sets. Annotation databases are loaded into the R environment. Gene groupings are established according to GO and KEGG annotations. Gene Set Variation Analysis (GSVA) is performed via the \"GSVA\" package in R, leveraging transcriptomic sequencing-derived gene expression profiles.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e2.13.1 Cell culture\u003c/h2\u003e \u003cp\u003eThe alveolar type II epithelial cell line (AT2), human bronchial epithelial cell line (BEAS-2B), and human fetal lung fibroblast cell line (MRC-5) were sourced from Zhongshan Hospital, Fudan University, Shanghai, China. Cells were cultured in Dulbecco\u0026rsquo;s Modified Eagle Medium (DMEM, HyClone, USA) enriched with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin (HyClone, USA). Cell cultures were grown under controlled humidity at 37℃ and 5% CO2, with no mycoplasma contamination detected. Cells were subcultured upon reaching 90% confluency. The cell suspension was collected and spun at 1000 rpm for 5 min. The pellet was resuspended in 1 mL of complete medium, then transferred to a culture dish with 7 mL of fresh DMEM to facilitate continued growth.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003e2.13.2 Extraction of Total Protein\u003c/h2\u003e \u003cp\u003eUpon reaching 80\u0026ndash;90% confluency, AT2, BEAS-2B, and MRC-5 cells were harvested for protein isolation. Cell lysis was performed using RIPA lysis buffer (Cell Signaling Technology, Jiangsu, China) on ice for 30 minutes. Following detachment via a cell scraper, cells were moved into a new centrifuge tube for total protein lysate isolation. Samples underwent brief sonication (10 seconds) and were subsequently spun at 12,000 rpm for 15 minutes under refrigeration (4℃). The supernatant obtained post-centrifugation was harvested for further downstream applications.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003e2.13.3 WB assay\u003c/h2\u003e \u003cp\u003eTotal protein was isolated from cellular or tissue samples using RIPA lysis buffer (Cell Signaling Technology, Jiangsu, China). Protein separation was achieved using 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), followed by electrophoretic transfer to a polyvinylidene difluoride (PVDF) membrane employing a submerged transfer methodology. Following transfer, membranes were washed in Tris-buffered saline containing 0.1% Tween-20 (TBST) and blocked with 5% skimmed milk powder prepared in TBST to reduce non-specific interactions. Immunoblotting with rabbit-derived primary antibodies and β-actin (1:10,000 dilution, AC026; ABclonal), serving as the loading reference, was conducted overnight at 4\u0026deg;C under constant low-speed rotation. Membranes were subsequently incubated with horseradish peroxidase (HRP)-linked secondary antibodies (Goat Anti-Rabbit IgG-HRP) diluted in antibody buffer at room temperature with gentle agitation for signal development. Western blot assays were conducted following the standardized protocol provided by Abcam, and protein bands were visualized using a Bio-Rad gel imaging system (Bio-Rad, Hercules, CA, USA). Comprehensive details regarding the commercial antibodies employed in this investigation can be found in the Supplementary Materials (Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003e2.13.4 Reverse transcription-quantitative polymerase chain reaction (RT-qPCR) of mRNA\u003c/h2\u003e \u003cp\u003eTotal RNA isolation was performed using TRIzol reagent (Thermo Fisher Scientific, USA), following the manufacturer\u0026rsquo;s guidelines (Abcam protocol: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.abcam.cn/protocols/rna-isolation-protocol-cells-in-culture\u003c/span\u003e\u003cspan address=\"https://www.abcam.cn/protocols/rna-isolation-protocol-cells-in-culture\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). Complementary DNA (cDNA) synthesis was achieved with the PrimeScript Reverse Transcription Kit (Takara Bio, Japan). Quantitative real-time PCR (RT-qPCR) assays were executed with Bimake SYBR Green qPCR master mix on a Bio-Rad CFX Connect Real-Time PCR platform. The thermal cycling protocol included an initial denaturation step (95℃for 30 seconds), succeeded by 45 cycles of amplification (95℃ for 5 seconds, 55℃ for 30 seconds, and 72℃ for 30 seconds). Cycle threshold (Ct) values for candidate genes were standardized relative to ACTB expression levels in matched samples, with differential expression calculated using the 2\u0026thinsp;\u0026minus;\u0026thinsp;ΔΔCt formula. Oligonucleotide primer sequences utilized in this workflow are provided in the Supplementary Materials (Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003e2.13.5 siRNAs and cell transfection.\u003c/h2\u003e \u003cp\u003eThe siRNAs used in this study were obtained from TSINKE (Beijing, China) and included a negative control, EGR1#1 siRNA (F-CCAUGGACAACUACCCUAATT, R-UUAGGGUAGUUGUCCAUGGTT), and EGR1#2 siRNA (F-GCCUAGUGAGCAUGACCAATT, R-UUGGUCAUGCUCACUAGGCTT). Transfection of siRNAs and plasmids was performed using lipo8000 reagent (Beyotime, Shanghai, China) or Lipo3000 transfection reagent (Thermo Fisher Scientific, Waltham, MA, USA). The NC group was transfected with a negative control plasmid, while the OE-EGR1 group was transfected with an EGR1 overexpression plasmid. Cells were used for subsequent experiments after transfection.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003e2.13.6 Measurement of Cell Apoptosis\u003c/h2\u003e \u003cp\u003eAT2 cells were collected by digestion with trypsin without EDTA. Subsequently, cells were washed twice with PBS (1,000 rpm, 5 minutes of centrifugation), and 1\u0026times;10⁵ to 5\u0026times;10⁵ cells were collected. 5 \u0026micro;L of Annexin V-FITC reagent was added, followed by 5 \u0026micro;L of Propidium Iodide (PI). After incubation, apoptosis was immediately analyzed using a flow cytometer (ACEA NovoCyte, USA) to determine the apoptosis rate.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec21\" class=\"Section2\"\u003e \u003ch2\u003e2.13.7 β-Galactosidase Assay\u003c/h2\u003e \u003cp\u003eCellular senescence was evaluated by rinsing cells with PBS. Fixation was performed with 1 mL of β-galactosidase (β-gal) staining fixative under ambient conditions (15 min). Post-fixation, cells underwent additional PBS rinsing and were incubated with the chromogenic substrate. Cells were then maintained overnight at 37℃ in a CO\u003csub\u003e2\u003c/sub\u003e -free environment. After staining, microscopic examination was performed using an Olympus CKX31 optical microscope (Japan). The percentage of senescent cells was determined by analyzing randomly sampled microscopic fields and calculating the ratio of β-gal-positive cells per 100 total cell.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec22\" class=\"Section2\"\u003e \u003ch2\u003e2.13.8 γ-H2AX Assay\u003c/h2\u003e \u003cp\u003eAfter the HEI-OC1 cells were collected, immunofluorescence staining was carried out. The process started by eliminating the culture medium from 6-well plates, followed by a single PBS rinse. Following this, the samples underwent incubation with rabbit-derived monoclonal γ-H2AX antibodies for 60-minute room temperature exposure. For nuclear visualization, DAPI solution was applied for five minutes at ambient temperature. Post-staining, specimens were embedded in fluorescence-preserving mounting agent and secured under a coverslip. Fluorescence microscopy showed that γ-H2AX immunoreactivity appeared as bright green fluorescence, while DAPI counterstaining highlighted the nuclei with blue fluorescence.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec23\" class=\"Section2\"\u003e \u003ch2\u003e2.13.9 Enzyme-Linked Immunosorbent Assay (ELISA)\u003c/h2\u003e \u003cp\u003eCytokine concentrations within AT2 cell culture supernatants were assessed via enzyme-linked immunosorbent assay (ELISA). Following 48 hours of transfection, supernatants were harvested, subjected to centrifugation, and cleared of cell fragments. IL-6, IL-17A, IL-1β, and TNF-α protein levels were specifically quantified during the assay, employing commercially available ELISA kits. Prior to initiating the procedure, all components were equilibrated at ambient temperature for half an hour. The protocol involved pipetting 50 \u0026micro;L of calibrators and test samples into designated wells, after which 100 \u0026micro;L of horseradish peroxidase-linked detection antibodies were introduced.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec24\" class=\"Section2\"\u003e \u003ch2\u003e2.13.10 Telomere Length Measurement\u003c/h2\u003e \u003cp\u003eGenomic DNA isolation from AT2 cells was performed, normalized, and analyzed with the Telomere Relative Length Detection Kit (Shanghai's Wing and Applied Biotechnology Co.) in compliance with the supplier's guidelines. Telomere length assessment utilized the qPCR-based protocol established by Cawthon\u003csup\u003e[\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e]\u003c/sup\u003e, employing graded concentrations of reference DNA and a calibration-based approach\u003csup\u003e[\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e]\u003c/sup\u003e. SYBR Green fluorescence detection facilitated quantification of telomere relative length. To derive final measurements, the sample's fluorescence signals were analyzed to determine the proportional value reflecting telomeric DNA quantity.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec25\" class=\"Section2\"\u003e \u003ch2\u003e2.14.11 Data Analysis\u003c/h2\u003e \u003cp\u003eData analysis was conducted with R statistical software. Group differences were evaluated through Wilcoxon rank-sum, chi-square, and Fisher's exact tests, chosen based on data suitability. Correlations were analyzed via Spearman's rank-order approach, utilizing the ggpubr and stats packages for computational tasks. qPCR results, expressed as mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD from triplicate runs, underwent two-tailed unpaired Student's t-test comparisons. Significance markers included ns (non-significant), *p\u0026thinsp;\u0026lt;\u0026thinsp;0.05, **p\u0026thinsp;\u0026lt;\u0026thinsp;0.01, and ***p\u0026thinsp;\u0026lt;\u0026thinsp;0.001. To validate reproducibility, triplicate experimental iterations were executed, with statistical relevance defined at p\u0026thinsp;\u0026lt;\u0026thinsp;0.05.\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Results","content":"\u003cdiv id=\"Sec27\" class=\"Section2\"\u003e \u003ch2\u003e3.1. Differential Expression Analysis between Controls and OSA Patients\u003c/h2\u003e \u003cp\u003eUsing |logFC| \u0026gt; 0.58 and FDR\u0026thinsp;\u0026lt;\u0026thinsp;0.05 as selection criteria, we identified 94 DEDEGs between the control group and OSA patients.Volcano plots and heatmaps are shown in Figs.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA and \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec28\" class=\"Section2\"\u003e \u003ch2\u003e3.2 Identifying Key Modules in OSA\u003c/h2\u003e \u003cp\u003eThe WGCNA analysis included 44 patients with OSA and 15 healthy controls. A scale-free co-expression network was built using a soft power parameter (β\u0026thinsp;=\u0026thinsp;10), achieving a scale-free topology fit (R2\u0026thinsp;=\u0026thinsp;0.9), as illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA. Hierarchical tree-based clustering detected 14 distinct modules, each comprising over 50 genes. These modules were then assessed for their linkages to OSA (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eC). Findings revealed that the green (r\u0026thinsp;=\u0026thinsp;0.67, p\u0026thinsp;=\u0026thinsp;6e\u0026thinsp;\u0026minus;\u0026thinsp;9), magenta (r\u0026thinsp;=\u0026thinsp;0.33, p\u0026thinsp;=\u0026thinsp;0.01), and pink (r\u0026thinsp;=\u0026thinsp;0.32, p\u0026thinsp;=\u0026thinsp;0.01) modules displayed positive correlations with OSA. Conversely, the yellow (r\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;0.73, p\u0026thinsp;=\u0026thinsp;3e\u0026thinsp;\u0026minus;\u0026thinsp;11), red (r\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;0.44, p\u0026thinsp;=\u0026thinsp;4e\u0026thinsp;\u0026minus;\u0026thinsp;4), and tan (r\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;0.36, p\u0026thinsp;=\u0026thinsp;0.005) modules showed inverse relationships. The green module, demonstrating the most significant linkage (r\u0026thinsp;=\u0026thinsp;0.67) and harboring 334 genes, was prioritized for deeper investigation. A scatter plot (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eE) graphically represented the association between module connectivity and OSA-related trait relevance.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec29\" class=\"Section2\"\u003e \u003ch2\u003e3.3 Functional Enrichment Analysis of DEDEs\u003c/h2\u003e \u003cp\u003eCross-referencing 94 DEGs against the green module's gene set yielded 21 OSA-associated DEDEGs. The biological roles of these overlapping genes were investigated via GO and KEGG pathway enrichment studies. GO analysis demonstrated marked enrichment in key biological pathways, including transcriptional regulation of microRNAs, control of neuronal apoptosis, and calcium signaling-mediated cellular responses. Cellular compartment evaluation underscored the relevance of distinct subcellular structures, such as lysosomal compartments and RNA polymerase II-containing transcriptional regulatory complexes. Molecular function analysis demonstrated significant interactions between RNA polymerase II transcription factors and SMAD protein binding. KEGG pathway enrichment revealed prominent involvement in immune-related signaling cascades, including IL-17 and TNF signaling pathways, along with rheumatoid arthritis pathogenesis. Additionally, significant enrichment was observed in osteogenic differentiation pathways and infection-related mechanisms, particularly those associated with Kaposi's sarcoma-associated herpesvirus and enteropathogenic Escherichia coli.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec30\" class=\"Section2\"\u003e \u003ch2\u003e3.4 Correlation Between OSA-ARDEGs\u003c/h2\u003e \u003cp\u003eBy cross-referencing 21 OSA-associated genes with genes linked to aging, we pinpointed 4 aging-associated OSA-linked genes (OSA-ARDEGs), as visualized in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA. The relationships among OSA-ARDEGs were evaluated using Pearson correlation coefficients.High correlations were found between EGR1 and FOS (cor\u0026thinsp;=\u0026thinsp;0.95), and JUN (cor\u0026thinsp;=\u0026thinsp;0.93), as well as between JUN and FOS (cor\u0026thinsp;=\u0026thinsp;0.95) (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eB). Figure\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eC illustrates the chromosomal positions of the OA-ARDEGs.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec31\" class=\"Section2\"\u003e \u003ch2\u003e3.5 Immune Infiltration Analysis\u003c/h2\u003e \u003cp\u003eTo examine variations in the immunological landscape of OSA patients compared to healthy individuals, we evaluated the prevalence and functional activity of immune cell subsets across both cohorts (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eA). Differences in immune cell abundance between groups, along with interrelationships among 22 immune subtypes, are illustrated in Figs.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eB and \u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eC. Results revealed a marked rise in M1 macrophage levels among OSA subjects relative to controls. Conversely, OSA cases displayed decreased proportions of plasma cells, na\u0026iuml;ve CD4\u0026thinsp;+\u0026thinsp;T lymphocytes and activated dendritic cells.\u003c/p\u003e \u003cp\u003eWe then examined the relationships between infiltrative immune cells and four types of OSA-ARDEGs (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e). Immune infiltration analysis revealed that EGR1 exhibited notable associations with dendritic cell activation (r\u0026thinsp;=\u0026thinsp;0.59) and regulatory T cell abundance (Tregs; r\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;0.44) (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eA), indicating a pivotal function in modulating these populations. FOS displayed strong linkages to dendritic cell activation (r\u0026thinsp;=\u0026thinsp;0.66) and inversely correlated with M0 macrophage prevalence (r\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;0.53) (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eB), highlighting its potential regulatory influence on these immune subsets. GLB1 demonstrated inverse relationships with monocyte infiltration (r\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;0.65) and dendritic cell activation (r\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;0.49) (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eC), suggesting its role in suppressing these cellular responses. JUN exhibited connections with dendritic cell activation (r\u0026thinsp;=\u0026thinsp;0.56) and a negative association with memory B cell frequency (r\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;0.4) (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eD), implying involvement in immune activation dynamics and memory regulation.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec32\" class=\"Section2\"\u003e \u003ch2\u003e3.6 Identification of Biological Functional Characteristics of Various Aging-Related Subtypes\u003c/h2\u003e \u003cp\u003eUnsupervised clustering of OSA samples using core regulatory factor expression profiles identified two distinct molecular subtypes (Figs.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eA-E). Differential expression analysis of regulatory factors revealed subtype-specific molecular signatures, with subtype 2 demonstrating marked upregulation of GLB1 and concurrent downregulation of EGR1, FOS, and JUN (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eA). Comparative evaluation further highlighted distinct immune infiltration profiles between the two subgroups. Specifically, subgroup 1 demonstrated elevated effector memory CD8\u0026thinsp;+\u0026thinsp;T cell and eosinophil infiltration, whereas subgroup 2 exhibited higher central memory CD8\u0026thinsp;+\u0026thinsp;T cells and both CD56dim/bright natural killer cell populations (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eF).\u003c/p\u003e \u003cp\u003eGSVA analysis indicated that subtype 1 exhibited upregulation in various biological processes and pathways, primarily involving lipoprotein particle clearance, intracellular lipid transport, and oxidoreductase activity, related to metabolism and molecular transport (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003eC).Moreover, this subtype exhibited enrichment in metabolic and biosynthetic pathways, including glutathione metabolism, porphyrin and chlorophyll metabolism, folate biosynthesis, and peroxisomes, possibly indicating subtype 1\u0026rsquo;s significant roles in oxidative stress response and metabolic regulation (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003eD).In contrast, subtype 2 exhibited significant activity in processes such as regulation of cell proliferation, stem cell development, inflammation, and immune regulation (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003eC). Key enriched pathways were primarily linked to immune system modulation and intracellular communication, encompassing the JAK-STAT signaling cascade, NOD receptor-mediated pathway, TCR signaling pathways, and TGF-β signaling cascades (as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003eD).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec33\" class=\"Section2\"\u003e \u003ch2\u003e3.7 Construction and Validation of OSA-ARDEGs Gene Markers\u003c/h2\u003e \u003cp\u003eTo enhance the diagnostic accuracy of hub OA-ARDEGs in OSA, we employed three distinct ML algorithms\u0026mdash;LASSO regression (Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e10\u003c/span\u003eA-B), SVM-RFE analysis (Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e10\u003c/span\u003eC-D), and Random Forest modeling (Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e10\u003c/span\u003eE-F)\u0026mdash;for feature selection. Through integrative analysis of these algorithmic outputs, GLB1 and EGR1 emerged as robust candidates, defining two key OSA-linked hub genes (OSA-ARDEGs).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec34\" class=\"Section2\"\u003e \u003ch2\u003e3.8 Validating Characteristic Biomarkers Using ML Models\u003c/h2\u003e \u003cp\u003eMultiple machine learning algorithms, including AdaBoost, Decision Tree, KNN, LightGBM, Na\u0026iuml;ve Bayes, and XGBoost, were employed for OSA characteristic gene identification. Model performance evaluation demonstrated recall rates exceeding 50% across all algorithms, as detailed in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. The predictive performance of each model was further validated through ROC curve analysis, presented in Fig.\u0026nbsp;\u003cspan refid=\"Fig11\" class=\"InternalRef\"\u003e11\u003c/span\u003e.Specifically, AdaBoost excelled in all metrics, achieving 100% accuracy, recall, F1 score, and an AUC of 1.0, demonstrating its superiority in identifying characteristic genes in OSA.Among them, AdaBoost had the highest AUC (1.0), while LightGBM performed best in terms of accuracy (0.9), Kappa value (0.7443), and F1 score (0.9318).The importance plots of the four OSA-ARDEGs for different models are shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig12\" class=\"InternalRef\"\u003e12\u003c/span\u003e. EGR1 and GLB1 emerged as predominant features across multiple models, with GLB1 specifically identified as a key feature in the Decision Tree algorithm.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eComparison of diagnostic effects of six different machine learning models.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"11\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c11\" colnum=\"11\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eML models\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTP\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eTN\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eFP\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eFN\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eAccuracy\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eKappa\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003ePrecision\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003eRecall\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c10\"\u003e \u003cp\u003eF1-Score\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c11\"\u003e \u003cp\u003eAUC\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAdaBoost\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e44\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e1.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e1.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e1.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e1.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e1.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDecision Tree;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e41\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.8667\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0645\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.9318\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.8913\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e0.9111\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e \u003cp\u003e0.905\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKNN\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e41\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.85\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.5921\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.9318\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.8723\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e0.9011\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e \u003cp\u003e0.915\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLightGBM\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e41\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e13\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.7443\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.9318\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.9318\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e0.9318\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e \u003cp\u003e0.964\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNa\u0026iuml;ve Bayes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e33\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.7333\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.3909\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.7500\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.8684\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e0.8049\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e \u003cp\u003e0.834\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eXGBoost\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e39\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.4675\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.8864\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.8478\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e0.8667\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e \u003cp\u003e0.9048\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec35\" class=\"Section2\"\u003e \u003ch2\u003e3.9 Construction and Evaluation of the LR Model\u003c/h2\u003e \u003cp\u003eFeature relevance assessment revealed two top-ranking genes, mutually prioritized by both AdaBoost and LightGBM frameworks, which were designated as core OSA-DEDEGs. Using these key genes, EGR1 and GLB1, a logistic regression (LR) model was developed. Investigation demonstrated significantly lower EGR1 levels in OSA patients relative to control subjects, whereas GLB1 expression was elevated in OSA cases (Fig.\u0026nbsp;\u003cspan refid=\"Fig13\" class=\"InternalRef\"\u003e13\u003c/span\u003eA). Furthermore, a nomogram developed through LR analysis is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig13\" class=\"InternalRef\"\u003e13\u003c/span\u003eB.\u003c/p\u003e \u003cp\u003eThe model's robustness was assessed via 1000 bootstrap resampling cycles applied to the training dataset. Calibration assessment validated the high predictive performance of the LR algorithm, as illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig13\" class=\"InternalRef\"\u003e13\u003c/span\u003eC. Decision curve analysis revealed superior clinical utility of the model-based approach for OSA patients, as evidenced by the consistent superiority of the model's benefit curve (red line) over the default strategies (gray line) across threshold probabilities (Fig.\u0026nbsp;\u003cspan refid=\"Fig13\" class=\"InternalRef\"\u003e13\u003c/span\u003eD).\u003c/p\u003e \u003cp\u003eThe training set evaluation demonstrated robust diagnostic performance, with Hub OSA-ARDEGs achieving AUC values of 0.801 and 0.865 (Fig.\u0026nbsp;\u003cspan refid=\"Fig14\" class=\"InternalRef\"\u003e14\u003c/span\u003eA), while the logistic regression model attained an AUC of 0.865 in the validation cohort (Fig.\u0026nbsp;\u003cspan refid=\"Fig14\" class=\"InternalRef\"\u003e14\u003c/span\u003eB). External validation using the GSE75097 dataset yielded AUC values of 0.607 and 0.577 for Hub OSA-ARDEGs (Fig.\u0026nbsp;\u003cspan refid=\"Fig14\" class=\"InternalRef\"\u003e14\u003c/span\u003eC), with the LR model achieving an AUC of 0.667 (Fig.\u0026nbsp;\u003cspan refid=\"Fig14\" class=\"InternalRef\"\u003e14\u003c/span\u003eD). These findings confirm the superior diagnostic predictive capacity of the logistic regression model across both internal and external validation datasets.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec36\" class=\"Section2\"\u003e \u003ch2\u003e3.6 Pathway Enrichment Analysis (GSEA) of Hub OSA-ARDEGs\u003c/h2\u003e \u003cp\u003ePathway enrichment analysis for GLB1 and EGR1 was performed using GSEA. The results reveal distinct biological processes associated with each gene. EGR1 exhibited marked enrichment in biological pathways associated with cardiomyocyte differentiation, interleukin-1-mediated signaling responses, and DNA-binding transcriptional activation mechanisms, as illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig15\" class=\"InternalRef\"\u003e15\u003c/span\u003eA. These processes appear more prominently among the most differentially expressed genes, indicated by earlier peaks in the ranked list for EGR1 compared to GLB1.For GLB1, enrichment analysis highlights its involvement in chromosome segregation, mitochondrion organization, and nuclear chromosome segregation, shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig15\" class=\"InternalRef\"\u003e15\u003c/span\u003eC.\u003c/p\u003e \u003cp\u003eExtended pathway enrichment profiling of EGR1-related networks revealed associations with cytokine receptor cross-talk, TCR signaling cascades, and olfactory signal transduction mechanisms (as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig15\" class=\"InternalRef\"\u003e15\u003c/span\u003eB). In contrast, GLB1 pathways are enriched in oxidative phosphorylation, ribosome function, and the response to Vibrio cholerae infection (Fig.\u0026nbsp;\u003cspan refid=\"Fig15\" class=\"InternalRef\"\u003e15\u003c/span\u003eD). Each gene's unique enrichment profile underscores its potential regulatory mechanisms and interactions within cellular processes.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec37\" class=\"Section2\"\u003e \u003ch2\u003e3.7 GSVA for Hub OSA-ARDEGs\u003c/h2\u003e \u003cp\u003eGene Set Variation Analysis (GSVA) was applied to assess GLB1 and EGR1 differential activity, specifically probing their functional involvement across distinct biological pathways. EGR1 predominantly upregulates pathways associated with metabolic processes and immune response modulation. Key upregulated pathways include Hemoglobin Alpha Binding, Toll-like Receptor 4 Binding, and Interleukin-21 Production. Notably, EGR1 downregulates pathways related to granulocyte chemotaxis and extracellular matrix disassembly, indicating a lesser role in these functions (Fig.\u0026nbsp;\u003cspan refid=\"Fig16\" class=\"InternalRef\"\u003e16\u003c/span\u003eA).Additionally, EGR1 upregulates pathways crucial to cellular metabolism, such as Porphyrin and Chlorophyll Metabolism, Sulfur Metabolism, Glutathione Metabolism, and Folate Biosynthesis. Conversely, it downregulates pathways involved in specialized or stress responses, including the Renin-Angiotensin System and Complement and Coagulation Cascades (Fig.\u0026nbsp;\u003cspan refid=\"Fig16\" class=\"InternalRef\"\u003e16\u003c/span\u003eB).\u003c/p\u003e \u003cp\u003eSimilarly, GLB1 exhibits strong upregulation in broader immune-related pathways, including Regulation of T-helper 17 Cell Differentiation and Interleukin 21 Production. Pathways such as Chitin Metabolic Process and Chitinase Activity are notably downregulated (Fig.\u0026nbsp;\u003cspan refid=\"Fig16\" class=\"InternalRef\"\u003e16\u003c/span\u003eC). GLB1 significantly enhances pathways critical to immune function and inflammation, including Glycosaminoglycan Biosynthesis, Glycosylphosphatidylinositol (GPI)-Anchor Biosynthesis, and Galactose Metabolism, while downregulating the Renin-Angiotensin System and pathways associated with Systemic Lupus Erythematosus and Primary Immunodeficiency (Fig.\u0026nbsp;\u003cspan refid=\"Fig16\" class=\"InternalRef\"\u003e16\u003c/span\u003eD).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec38\" class=\"Section2\"\u003e \u003ch2\u003e3.8 Selection of Basal Protein and Verification of Overexpression Plasmid\u003c/h2\u003e \u003cp\u003eTo further explore the potential mechanisms of cellular senescence, this study focused on EGR1. During preliminary screening to identify a suitable cellular system, EGR1 baseline expression was assessed across three human cell lines: AT2 alveolar epithelial cells (type II), BEAS-2B bronchial epithelial cells, and MRC-5 fetal lung fibroblasts. Comparative analysis of EGR1 basal expression across three cell types revealed significantly higher expression levels in AT2 cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig17\" class=\"InternalRef\"\u003e17\u003c/span\u003eA), establishing them as the experimental model for subsequent investigations. To elucidate EGR1's role in cellular senescence mechanisms, an EGR1 overexpression vector was developed, with successful transfection confirmed through Western blot analysis (Fig.\u0026nbsp;\u003cspan refid=\"Fig17\" class=\"InternalRef\"\u003e17\u003c/span\u003eB).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec39\" class=\"Section2\"\u003e \u003ch2\u003e3.9 Selection of Senescence-Related Genes\u003c/h2\u003e \u003cp\u003eAfter transfection with the EGR1 overexpression plasmid, to improve the accuracy of predicting downstream signaling pathways, candidate genes for 12 Hub OSA-ARDEGs were predicted using the GEO and HCA databases. These genes included: CIITA, NAIP, NOD1, NOD2, NLRC5, NLRP1, NLRP3, NLRP4, NLRP5, NLRP13, NLRP14, and NLRX1. The accuracy of the candidate genes was subsequently verified by qPCR, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig18\" class=\"InternalRef\"\u003e18\u003c/span\u003e(A). Due to issues with the CIITA and NLRP13 primers, new primers were designed for a second qPCR experiment. Finally, the three most markedly dysregulated genes\u0026mdash;CIITA, NAIP, and NLRX1\u0026mdash;were prioritized for immunoblot validation (Fig.\u0026nbsp;\u003cspan refid=\"Fig18\" class=\"InternalRef\"\u003e18\u003c/span\u003eB). Data revealed markedly elevated expression levels of these genes in EGR1-overexpressing AT2 cells.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec40\" class=\"Section2\"\u003e \u003ch2\u003e3.10 Apoptosis Detection\u003c/h2\u003e \u003cp\u003eApoptosis analysis through flow cytometry revealed distinct cellular states between experimental groups. The negative control (NC) group exhibited predominant cell viability, with the majority of cells localized in the healthy quadrant (Q3-3) and minimal apoptotic populations (Q3-1 and Q3-2), indicating low basal apoptosis rates and active cell proliferation. In contrast, EGR1 overexpression significantly enhanced apoptotic cell proportions, as demonstrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig19\" class=\"InternalRef\"\u003e19\u003c/span\u003e(A).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec41\" class=\"Section2\"\u003e \u003ch2\u003e3.11 Cellular Senescence and DNA Damage\u003c/h2\u003e \u003cp\u003eTo assess the impact of EGR1 overexpression on cellular senescence in AT2 cells, β-galactosidase activity levels were quantified in both non-modified control (NC) and EGR1-overexpressing cohorts. Relative to NC cells, EGR1-overexpressing AT2 cells exhibited markedly elevated senescence levels (Fig.\u0026nbsp;\u003cspan refid=\"Fig19\" class=\"InternalRef\"\u003e19\u003c/span\u003eB). Furthermore, qPCR analysis of telomere length revealed no significant difference in telomere length between the NC and EGR1 overexpressing AT2 cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig19\" class=\"InternalRef\"\u003e19\u003c/span\u003eD), suggesting that EGR1 does not induce senescence through a telomere-dependent mechanism. To explore the involvement of EGR1 in modulating the DNA damage response (DDR), γ-H2AX immunofluorescence staining was utilized to identify DNA double-strand breaks (DSBs). In the negative control (NC) group, γ-H2AX foci were scarcely detectable after treatment. In contrast, cells overexpressing EGR1 exhibited sustained γ-H2AX signals (Fig.\u0026nbsp;\u003cspan refid=\"Fig19\" class=\"InternalRef\"\u003e19\u003c/span\u003eC). These results suggest that EGR1 promotes cellular senescence and DDR activation in AT2 cells via a mechanism independent of telomere regulation.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec42\" class=\"Section2\"\u003e \u003ch2\u003e3.12 ELISA\u003c/h2\u003e \u003cp\u003ePro-inflammatory cytokines, particularly TNF-α and IL-6, are central drivers of OSA progression. Cytokine quantification assays identified robust upregulation of inflammatory mediators (IL-6, IL-17A, IL-1β, TNF-α) in EGR1-overexpressing AT2 cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig20\" class=\"InternalRef\"\u003e20\u003c/span\u003e). These findings suggest that EGR1 potentially drives OSA progression by regulating inflammatory signaling pathways.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eOSA is increasingly recognized as a critical health problem, particularly prevalent in aging populations. OSA is marked by frequent episodes of upper airway collapse during sleep, causing intermittent oxygen desaturation, fragmented sleep patterns, and consequent excessive daytime fatigue\u003csup\u003e[\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e]\u003c/sup\u003e. The formal study of aging mechanisms traces its origins to 1989, when Irwin Rosenberg defined \"sarcopenia\" to characterize the progressive loss of skeletal muscle mass and strength associated with aging\u003csup\u003e[\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e]\u003c/sup\u003e. Later investigations have deepened insights into aging biology, uncovering multifaceted relationships between aging and the gradual impairment of systemic organ functions, dysregulated metabolic processes, and the emergence of age-related chronic pathologies\u003csup\u003e[\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e]\u003c/sup\u003e. For instance, aging contributes to the onset of diabetes, cancer, cardiovascular diseases, and neurodegenerative diseases through mechanisms such as mitochondrial autophagy dysfunction\u003csup\u003e[\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e]\u003c/sup\u003e, inflammatory pathway activation\u003csup\u003e[\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e]\u003c/sup\u003e, and changes in hormone levels\u003csup\u003e[\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e]\u003c/sup\u003e. Recent research has highlighted significant interactions between the aging process and OSA. Muscle mass loss caused by aging, such as the decline in pharyngeal dilator muscle function, and metabolic dysregulation, like insulin resistance, may worsen the pathological progression of OSA\u003csup\u003e[\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e, \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eEpidemiological studies indicate that OSA affects a substantial number of adults, with prevalence expected to rise as the global population ages\u003csup\u003e[\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]\u003c/sup\u003e. Advanced age significantly elevates the risk of OSA onset. Progressive age-related physiological alterations\u0026mdash;including diminished muscular tone, heightened adipose accumulation in cervical regions, and respiratory control dysregulation\u0026mdash;collectively underlie OSA pathogenesis\u003csup\u003e[\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e, \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e]\u003c/sup\u003e. In older adults, OSA is often associated with accelerated cognitive deterioration and a heightened risk of Alzheimer\u0026rsquo;s disease. These findings indicate that OSA may not simply be a marker of normal aging but might actively contribute to its progression\u003csup\u003e[\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e, \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e]\u003c/sup\u003e. Understanding the interplay between OSA and aging is essential for developing targeted interventions that can mitigate OSA's impact on the health and quality of life of older adults\u003csup\u003e[\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e]\u003c/sup\u003e. However, although recent research has increasingly recognized the significant role of aging in OSA, the exact mechanisms and contributions remain elusive.\u003c/p\u003e \u003cp\u003eAn integrated analysis of transcriptomic datasets initially detected 94 genes (OSA-DEGs) exhibiting significant differential expression between OSA patients and controls. To investigate their disease relevance, co-expression network analysis (WGCNA) uncovered a module comprising 334 OSA-correlated genes. Cross-validation via two orthogonal algorithms refined the candidate list to 21 DEDEGs, enhancing confidence in the reproducibility and precision of these biomarkers.To delineate the functional attributes of DEGs, GO and KEGG pathway enrichment investigations were performed. GO analysis revealed significant enrichment in processes associated with transcriptional regulation mediated by miRNAs\u0026mdash;non-coding RNA molecules involved in post-transcriptional silencing mechanisms. Emerging evidence emphasizes the diagnostic and therapeutic potential of miRNAs in OSA, as highlighted by recent research. For instance, Moriondo et al. identified specific miRNA signatures correlating with OSA severity, underscoring their utility as biomarkers and intervention targets\u003csup\u003e[\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e]\u003c/sup\u003e. Altered miRNA expression may affect the pathogenesis of OSA through pathways involving inflammation, oxidative stress, and metabolic dysfunction, which are prevalent in OSA patients \u003csup\u003e[\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e]\u003c/sup\u003e. The \"IL-17 signaling pathway\" was significantly enriched. This pathway encompasses the binding of IL-17 to its receptors, tinitiating a cascade of intracellular signaling events that culminate in the activation of NF-κB and other transcription factors, which promote the expression of inflammatory genes\u003csup\u003e[\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e]\u003c/sup\u003e. Emerging research underscores a robust pathophysiological link between IL-17 signaling dysregulation and OSA\u0026mdash;a condition characterized by recurrent nocturnal upper airway collapse, intermittent hypoxia, and chronic systemic inflammation\u003csup\u003e[\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e, \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e]\u003c/sup\u003e, This association establishes a mechanistic rationale for downstream experimental validation.\u003c/p\u003e \u003cp\u003eTo investigate the contribution of aging-associated genes to OSA pathogenesis, we analyzed 543 ARDEGs, narrowing this to four hub OSA-ARDEGs with disease-specific relevance. SsGSEA revealed an elevated enrichment score for M1 macrophages in OSA patients compared to non-OSA controls. Concurrently, reduced infiltration levels were observed for plasma cells, na\u0026iuml;ve CD4\u0026thinsp;+\u0026thinsp;T lymphocytes, monocytes, and activated dendritic cells in OSA cohorts. The prominence of M1 macrophages aligns mechanistically with the chronic inflammatory milieu and oxidative stress burden hallmarking OSA.\u003c/p\u003e \u003cp\u003eWe assessed the efficacy of six machine learning algorithms to select the optimal classifier for discerning OSA-associated signature genes. Specifically, AdaBoost outperformed all other models across all metrics, achieving perfect accuracy, recall, F1 score, and AUC (all at 100%), highlighting its distinct advantage in precisely identifying OSA characteristic genes. In the global and local interpretability analysis of four OSA-arDEGs, two key genes, EGR1 and GLB1, were specifically identified. EGR1 is a zinc-finger transcription factor that orchestrates critical biological functions, including cellular proliferation, differentiation, and apoptosis. Conversely, GLB1, a lysosomal enzyme, exhibits progressive upregulation in senescent tissues and serves as a molecular indicator of cellular aging and age-related functional deterioration. This increase exacerbates the decline in lung function, which is critical in cases of OSA. GLB1, encoded at the GLB1 gene locus, is essential for lysosomal function and is characteristically elevated in the tissues of aging organisms, underscoring its role as a biomarker of cellular senescence \u003csup\u003e[\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e]\u003c/sup\u003e. The increase in GLB1 is linked to a decline in age-related physiological functions, especially in the lung system, where systemic aging and GLB1 activity may relate to pathologies such as fibrosis and cellular senescence, worsening the decline in lung function \u003csup\u003e[\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eTo validate the results of bioinformatics analysis, we selected the EGR1 gene for further investigation.Studies across multiple fields show that EGR1 regulates key signaling pathways, influencing aging and immune homeostasis. In aging - related diseases, EGR1 affects cell function by regulating mitophagy and oxidative stress. For example, in an intervertebral disc degeneration model, EGR1 exacerbates oxidative - stress - induced nucleus pulposus cell senescence by inhibiting PINK1 - Parkin - dependent mitophagy\u003csup\u003e[\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e]\u003c/sup\u003e. In immune regulation, EGR1 directly modulates inflammatory enhancer activity. During macrophage differentiation, it suppresses pro - inflammatory genes (e.g., IL1β, TNFα) by recruiting the NuRD repressor complex, limiting excessive inflammatory responses \u003csup\u003e[\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e]\u003c/sup\u003e. Despite this, the role of EGR1 in OSA remains poorly understood.\u003c/p\u003e \u003cp\u003eApoptosis refers to the programmed cell death process triggered by internal or external signals under specific physiological or pathological conditions. Several studies have shown that OSA induces apoptosis in tissues such as the brain, heart, and kidneys through intermittent hypoxia, further exacerbating cognitive impairment, metabolic disorders, and organ dysfunction \u003csup\u003e[\u003cspan additionalcitationids=\"CR61\" citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e]\u003c/sup\u003e. Based on this, we analyzed the apoptosis levels in AT2 cells. The results indicated that the occurrence of apoptosis provided further support for the hypothesis of the EGR1 pathway predicted by bioinformatics analysis.\u003c/p\u003e \u003cp\u003eCellular senescence manifests through two distinct molecular pathways\u003csup\u003e[\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e]\u003c/sup\u003e: (1) replicative senescence due to telomerase dysfunction and progressive telomere shortening, causing irreversible cell - cycle arrest linked to lost proliferative capacity\u003csup\u003e[\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e]\u003c/sup\u003e; and (2) stress - induced premature senescence (SIPS), a potentially reversible process triggered by external stressors such as oxidative damage, radiation, or high glucose\u003csup\u003e[\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e65\u003c/span\u003e]\u003c/sup\u003e. Persistent exposure to oxidative stress or environmental stressors activates DNA damage response (DDR) pathways, culminating in either permanent cell cycle arrest or transient proliferative arrest (SIPS), dictated by the magnitude and duration of stress.\u003c/p\u003e \u003cp\u003eMechanistically, our findings reveal two parallel senescence - inducing pathways: the telomere - dependent pathway, marked by significant telomere attrition tied to aging; and the DNA damage response pathway, characterized by notable γ - H2AX accumulation, a sign of genomic instability and stress - induced senescence. Importantly, our results suggest that EGR1 may mechanistically promote SIPS development by activating stress - response pathways, particularly those linked to the DNA damage response.\u003c/p\u003e \u003cp\u003eImmune system dysregulation in OSA patients has long been a central focus of clinical research. Empirical studies demonstrate that EGR1 drives transcriptional activation of CIITA, NAIP, and NLRX1 within AT2 alveolar epithelial cells.CIITA (MHC Class II Transactivator) serves as the principal controller governing MHC-II gene transcriptional activation. Recognized as the archetypal member within the NLR family (nucleotide-binding oligomerization domain-like receptors), this regulator plays a critical role in mediating adaptive immune responses. Research has demonstrated that CIITA modulates immune responses within the tumor microenvironment through the regulation of MHC-II molecule expression\u003csup\u003e[\u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e]\u003c/sup\u003e. NLRX1, a mitochondrially localized member of the nucleotide-binding oligomerization domain-like receptor (NLR) family, mediates pleiotropic immunomodulatory roles across innate and adaptive immunity, including pathogen sensing and inflammatory homeostasis. It negatively regulates type I interferons and pro-inflammatory responses by modulating the NF-κB signaling pathway, thereby influencing cell death, autophagy, and proliferation \u003csup\u003e[\u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e67\u003c/span\u003e]\u003c/sup\u003e. NAIP (NLR Family Apoptosis Inhibitory Protein) is a core component of the NAIP/NLRC4 inflammasome and primarily initiates the inflammatory response by recognizing pathogenic factors such as bacterial flagellin. The NAIP/NLRC4 inflammasome plays an important role in host defense, particularly in immune responses against bacterial infections by recruiting and activating Caspase-1 \u003csup\u003e[\u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e68\u003c/span\u003e, \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e69\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eEGR1 plays a central role in OSA pathophysiology by orchestrating immune regulatory networks, particularly through modulation of CIITA, NAIP, and NLRX1 activity. Mechanistically, EGR1 activation triggers transcriptional induction of pro-inflammatory cytokines (e.g., IL-6, IL-17A, IL-1β, TNF-α), which are cornerstones of OSA-associated inflammation and pathogenic drivers of disease initiation and progression. These findings position EGR1 as a master regulator of immune dysregulation in OSA, offering a biologically plausible therapeutic target. This mechanistic insight aligns with and extends prior computational predictions of EGR1\u0026rsquo;s centrality in OSA pathogenesis.\u003c/p\u003e \u003cp\u003eAdvancing age profoundly impacts pulmonary cellular homeostasis, resulting in diminished lung capacity and elastic recoil, thereby amplifying the severity of obstructive respiratory pathologies such as OSA[70]. Emerging evidence implicates gerontogenes\u0026mdash;particularly those regulating cellular senescence and mitochondrial dynamics\u0026mdash;as potential diagnostic biomarkers for OSA. This supports a paradigm wherein genetic susceptibility to OSA exhibits age-dependent escalation, attributable to lifelong accrual of genomic instability and environmental exposures\u003csup\u003e[\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e]\u003c/sup\u003e. Furthermore, epigenetic modifications in these aging genes can lead to metabolic imbalances, further complicating the cellular environment and promoting the progression of OSA.\u003c/p\u003e \u003cp\u003eThis investigation is subject to certain constraints. Primarily, while we performed rigorous internal and external validation of the diagnostic framework, its generalizability necessitates additional validation across diverse, independent cohorts, particularly in prospective, multi-center studies. In this research, we utilized all available OSA datasets. The role of EGR1 in OSA remains incompletely understood: while experimental results suggest that EGR1 may influence immune factors and DNA damage response in OSA, its specific mechanisms need to be further explored. The complex interactions between EGR1 and aging have not been fully examined either. Furthermore, to further validate the robustness and reliability of the diagnostic method, it is recommended to include more external validation cohorts.\u003c/p\u003e"},{"header":"5. Conclusion","content":"\u003cp\u003eIn this study, we employed bioinformatics and machine learning approaches to investigate age - related gene expression patterns and molecular mechanisms in OSA. We identified two key senescence - related genes that are closely associated with OSA and developed a logistic regression (LR) diagnostic model based on their expression profiles. This model significantly enhanced the accuracy of OSA diagnosis. The expression levels of these genes were significantly associated with OSA severity, indicating their potential as novel diagnostic biomarkers. In our preliminary exploration of the mechanisms of senescence - related genes in OSA, we found that EGR1, a marker of cellular senescence and immune response, may influence OSA progression by regulating immune - related factors and SIPS. EGR1 may contribute to OSA through senescence and immune - related pathways, making it a potential therapeutic target.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eFunding:\u003c/strong\u003e This study was funded by the National Natural Science Foundation of China (81270144, 30800507, 81170071 and 81400063).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflict of Interest:\u003c/strong\u003e Author Li Caili declares that she has no conflict of interest. Author Zhou Wei declares that she has no conflict of interest. Author Xu Chong \u0026nbsp; declares that she has no conflict of interest. Author Cao Jie declares that she has no conflict of interest. Author Zhang Jing declares that she has no conflict of interest.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthical approval\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe ethical and methodological aspects of the study was approved by Institutional Review Board of Tianjin Medical University General Hospital (TMU IRB Approving Number: EA-20-120002) and performed in accordance with the Guide for the Care and Use of Laboratory Animals.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e：None.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Availability Statements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eJOOSTEN S A, LANDRY S A, WONG A M, et al. Assessing the Physiologic Endotypes Responsible for REM- and NREM-Based OSA [J]. Chest. 2021;159(5):1998\u0026ndash;2007.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBAJAJ R, SINGH N. Properties of octenyl succinic anhydride (OSA) modified starches and their application in low fat mayonnaise [J]. Int J Biol Macromol. 2019;131:147\u0026ndash;57.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLAM JC, MAK J C, IP MS. Obesity, obstructive sleep apnoea and metabolic syndrome [J]. Respirol (Carlton Vic). 2012;17(2):223\u0026ndash;36.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMESSINEO L, BAKKER J P, CRONIN J, et al. Obstructive sleep apnea and obesity: A review of epidemiology, pathophysiology and the effect of weight-loss treatments [J]. Sleep Med Rev. 2024;78:101996.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKHALYFA A, MARIN JM, QIAO Z et al. Plasma exosomes in OSA patients promote endothelial senescence: effect of long-term adherent continuous positive airway pressure [J]. Sleep, 2020, 43(2).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLIN Y N, LI Q Y, ZHANG X J. Interaction between smoking and obstructive sleep apnea: not just participants [J]. Chin Med J. 2012;125(17):3150\u0026ndash;6.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNARANJO M, WILLES L, PRILLAMAN B A, et al. Undiagnosed OSA May Significantly Affect Outcomes in Adults Admitted for COPD in an Inner-City Hospital [J]. Chest. 2020;158(3):1198\u0026ndash;207.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNG N B H, LIM C Y S TANS, et al. Screening for obstructive sleep apnea (OSA) in children and adolescents with obesity: A scoping review of national and international pediatric obesity and pediatric OSA management guidelines [J]. Obes reviews: official J Int Association Study Obes. 2024;25(5):e13712.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDHARMAKULASEELAN L, BOULOS M I. Sleep Apnea and Stroke. Narrative Rev [J] Chest. 2024;166(4):857\u0026ndash;66.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBHASIN S, BRITO J P, CUNNINGHAM G R, et al. Testosterone Therapy in Men With Hypogonadism: An Endocrine Society Clinical Practice Guideline [J]. J Clin Endocrinol Metab. 2018;103(5):1715\u0026ndash;44.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGASPAR LS, \u0026Aacute;LVARO A R, MOITA J, et al. Obstructive Sleep Apnea and Hallmarks of Aging [J]. Trends Mol Med. 2017;23(8):675\u0026ndash;92.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLI Y, WANG Y. Obstructive Sleep Apnea-hypopnea Syndrome as a Novel Potential Risk for Aging [J]. Aging disease. 2021;12(2):586\u0026ndash;96.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLIU P Y, REDDY RT. Sleep, testosterone and cortisol balance, and ageing men [J]. Reviews Endocr metabolic disorders. 2022;23(6):1323\u0026ndash;39.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCARROLL JE, IRWIN M R, SEEMAN T E et al. Obstructive sleep apnea, nighttime arousals, and leukocyte telomere length: the Multi-Ethnic Study of Atherosclerosis [J]. Sleep, 2019, 42(7).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCOONEY L G DOKRASA. Beyond fertility: polycystic ovary syndrome and long-term health [J]. Fertil Steril. 2018;110(5):794\u0026ndash;809.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLAL C, AYAPPA I, AYAS N, et al. The Link between Obstructive Sleep Apnea and Neurocognitive Impairment: An Official American Thoracic Society Workshop Report [J]. Annals Am Thorac Soc. 2022;19(8):1245\u0026ndash;56.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWEIHS A, FRENZEL S, WITTFELD K et al. Associations between sleep apnea and advanced brain aging in a large-scale population study [J]. Sleep, 2021, 44(3).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYEREVANIAN A, SOUKAS AA, Metformin. Mechanisms in Human Obesity and Weight Loss [J]. Curr Obes Rep. 2019;8(2):156\u0026ndash;64.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCHEN J W, HUANG M J, CHEN X N, et al. Transient upregulation of EGR1 signaling enhances kidney repair by activating SOX9(+) renal tubular cells [J]. Theranostics. 2022;12(12):5434\u0026ndash;50.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJIN Y, WANG C, ZHANG B, et al. Blocking EGR1/TGF-β1 and CD44s/STAT3 Crosstalk Inhibits Peritoneal Metastasis of Gastric Cancer [J]. Int J Biol Sci. 2024;20(4):1314\u0026ndash;31.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWANG Y, QIN C, ZHAO B, et al. EGR1 induces EMT in pancreatic cancer via a P300/SNAI2 pathway [J]. J translational Med. 2023;21(1):201.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWANG B, WANG Y, WANG W, et al. WTAP/IGF2BP3 mediated m6A modification of the EGR1/PTEN axis regulates the malignant phenotypes of endometrial cancer stem cells [J]. J experimental Clin cancer research: CR. 2024;43(1):204.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNGIAM K Y, KHOR I W. Big data and machine learning algorithms for health-care delivery [J]. Lancet Oncol. 2019;20(5):e262\u0026ndash;73.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXU M, ZHOU H, HU P, et al. Identification and validation of immune and oxidative stress-related diagnostic markers for diabetic nephropathy by WGCNA and machine learning [J]. Front Immunol. 2023;14:1084531.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGHARIB SA, HURLEY A L, ROSEN M J et al. Obstructive sleep apnea and CPAP therapy alter distinct transcriptional programs in subcutaneous fat tissue [J]. Sleep, 2020, 43(6).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGHARIB SA, HAYES A L, ROSEN M J, et al. A pathway-based analysis on the effects of obstructive sleep apnea in modulating visceral fat transcriptome [J]. Sleep. 2013;36(1):23\u0026ndash;30.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCHEN YC, CHEN K D, SU M C, et al. Genome-wide gene expression array identifies novel genes related to disease severity and excessive daytime sleepiness in patients with obstructive sleep apnea [J]. PLoS ONE. 2017;12(5):e0176575.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eADEGUNSOYE A, NEBORAK J M, ZHU D, et al. CPAP Adherence, Mortality, and Progression-Free Survival in Interstitial Lung Disease and OSA [J]. Chest. 2020;158(4):1701\u0026ndash;12.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLORENZI-FILHO G, ALMEIDA F R, STROLLO P J. Treating OSA. Current and emerging therapies beyond CPAP [J]. Respirology (Carlton, Vic), 2017, 22(8): 1500-7.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRITCHIE M E, PHIPSON B. Nucleic Acids Res. 2015;43(7):e47. WU D, et al. limma powers differential expression analyses for RNA-sequencing and microarray studies [J].\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWU T, HU E, XU S, Innovation et al. (Cambridge (Mass)), 2021, 2(3): 100141.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLUO W. Pathview: an R/Bioconductor package for pathway-based data integration and visualization [J]. Bioinf (Oxford England). 2013;29(14):1830\u0026ndash;1.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZHOU J, HUANG J, LI Z, et al. Identification of aging-related biomarkers and immune infiltration characteristics in osteoarthritis based on bioinformatics analysis and machine learning [J]. Front Immunol. 2023;14:1168780.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDE MAGALH\u0026atilde;ES JP, ABIDI Z, DOS SANTOS G A, et al. Human Ageing Genomic Resources: updates on key databases in ageing research [J]. Nucleic Acids Res. 2024;52(D1):D900\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZUCCATO JA, PATIL V. Cerebrospinal fluid methylome-based liquid biopsies for accurate malignant brain neoplasm classification [J]. Neurooncology. 2023;25(8):1452\u0026ndash;60.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGU Z, EILS R. Complex heatmaps reveal patterns and correlations in multidimensional genomic data [J]. Bioinf (Oxford England). 2016;32(18):2847\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCAWTHON R M. Telomere length measurement by a novel monochrome multiplex quantitative PCR method [J]. Nucleic Acids Res. 2009;37(3):e21.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDESHMUKH GOMASEVG. Obstructive Sleep Apnea and Its Management: A Narrative Review [J]. Cureus. 2023;15(4):e37359.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWHITE D P, YOUNES M K. Obstructive sleep apnea [J]. Compr Physiol. 2012;2(4):2541\u0026ndash;94.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNISHIKAWA H, FUKUNISHI S, ASAI A et al. Pathophysiology and mechanisms of primary sarcopenia (Review) [J]. Int J Mol Med, 2021, 48(2).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOH HS, RUTLEDGE J. Organ aging signatures in the plasma proteome track health and disease [J]. Nature. 2023;624(7990):164\u0026ndash;72.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKITADA M. Autophagy in metabolic disease and ageing [J]. Nat reviews Endocrinol. 2021;17(11):647\u0026ndash;61.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eARMENTO A, UEFFING M, CLARK SJ. The complement system in age-related macular degeneration [J]. Cell Mol Life Sci. 2021;78(10):4487\u0026ndash;505.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGAUTHIER B R, SOLA-GARC\u0026iacute;A A, C\u0026aacute;LIZ-MOLINA M, et al. Thyroid hormones in diabetes, cancer, and aging [J]. Aging Cell. 2020;19(11):e13260.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLIU D, WANG S, LIU S et al. Frontiers in sarcopenia: Advancements in diagnostics, molecular mechanisms, and therapeutic strategies [J]. Molecular aspects of medicine, 2024, 97: 101270.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePERGER E, MATTALIANO P. LOMBARDI C Menopause Sleep Apnea [J] Maturitas. 2019;124:35\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYEO E J. Hypoxia and aging [J]. Exp Mol Med. 2019;51(6):1\u0026ndash;15.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eANDRADE A G, BUBU O M, VARGA A W, et al. The Relationship between Obstructive Sleep Apnea and Alzheimer's Disease [J]. J Alzheimer's disease: JAD. 2018;64(s1):S255\u0026ndash;70.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKINUGAWA K. Obstructive sleep apnea and dementia: A role to play? [J]. Rev Neurol. 2023;179(7):793\u0026ndash;803.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMULLINS A E, KAM K, PAREKH A, et al. Obstructive Sleep Apnea and Its Treatment in Aging: Effects on Alzheimer's disease Biomarkers, Cognition, Brain Structure and Neurophysiology [J]. Neurobiol Dis. 2020;145:105054.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMORIONDO G, SOCCIO P, TONDO P et al. Obstructive Sleep Apnea: A Look towards Micro-RNAs as Biomarkers of the Future [J]. Biology, 2022, 12(1).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZHANG K, WANG C, WU Y, et al. Identification of novel biomarkers in obstructive sleep apnea via integrated bioinformatics analysis and experimental validation [J]. PeerJ. 2023;11:e16608.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYANG J. Role and mechanism of IL \u0026ndash;\u0026thinsp;17 and its gene polymorphisms in dyslipidemia caused by obstructive sleep apnea syndrome in children [J]. Cellular and molecular biology (Noisy-le-Grand, France), 2022, 68(2): 208\u0026ndash;12.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBHATT SP, GULERIA R, KABRA SK. Metabolic alterations and systemic inflammation in overweight/obese children with obstructive sleep apnea [J]. PLoS ONE. 2021;16(6):e0252353.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHUANG YS, CHIN W C, GUILLEMINAULT C et al. Inflammatory Factors: Nonobese Pediatric Obstructive Sleep Apnea and Adenotonsillectomy [J]. J Clin Med, 2020, 9(4).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSUN J, WANG M, ZHONG Y, et al. A Glb1-2A-mCherry reporter monitors systemic aging and predicts lifespan in middle-aged mice [J]. Nat Commun. 2022;13(1):7028.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePHADKE M, KRYNETSKAIA N, MISHRA A, et al. Accelerated cellular senescence phenotype of GAPDH-depleted human lung carcinoma cells [J]. Biochem Biophys Res Commun. 2011;411(2):409\u0026ndash;15.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWU ZL, WANG K P, CHEN Y J, et al. Knocking down EGR1 inhibits nucleus pulposus cell senescence and mitochondrial damage through activation of PINK1-Parkin dependent mitophagy, thereby delaying intervertebral disc degeneration [J]. Volume 224. Free radical biology \u0026amp; medicine; 2024. pp. 9\u0026ndash;22.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTRIZZINO M, ZUCCO A, DELIARD S et al. EGR1 is a gatekeeper of inflammatory enhancers in human macrophages [J]. Sci Adv, 2021, 7(3).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGUO X, SHI Y, DU P, et al. HMGB1/TLR4 promotes apoptosis and reduces autophagy of hippocampal neurons in diabetes combined with OSA [J]. Life Sci. 2019;239:117020.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLV R, ZHAO Y, WANG X et al. GLP-1 analogue liraglutide attenuates CIH-induced cognitive deficits by inhibiting oxidative stress, neuroinflammation, and apoptosis via the Nrf2/HO-1 and MAPK/NF-κB signaling pathways [J]. Int Immunopharmacol, 2024, 142(Pt B): 113222.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSHI Y, GUO X, ZHANG J et al. DNA binding protein HMGB1 secreted by activated microglia promotes the apoptosis of hippocampal neurons in diabetes complicated with OSA [J]. Brain, behavior, and immunity, 2018, 73: 482\u0026ndash;92.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOPRESKO PL, SHAY JW. Telomere-associated aging disorders [J]. Ageing Res Rev. 2017;33:52\u0026ndash;66.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMOHAMAD KAMAL N S, SAFUAN S, SHAMSUDDIN S, et al. Aging of the cells: Insight into cellular senescence and detection Methods [J]. Eur J Cell Biol. 2020;99(6):151108.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCHEN MS, LEE R T, GARBERN JC. Senescence mechanisms and targets in the heart [J]. Cardiovascular Res. 2022;118(5):1173\u0026ndash;87.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLIU H, HE J, BAGHERI-YARMAND R, et al. Osteocyte CIITA aggravates osteolytic bone lesions in myeloma [J]. Nat Commun. 2022;13(1):3684.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNAGAI-SINGER M A, MORRISON H A. ALLEN I C. NLRX1 Is a Multifaceted and Enigmatic Regulator of Immune System Function [J]. Front Immunol. 2019;10:2419.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBAUER R, RAUCH I. The NAIP/NLRC4 inflammasome in infection and pathology [J]. Mol Aspects Med. 2020;76:100863.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVANCE R E. The NAIP/NLRC4 inflammasomes [J]. Curr Opin Immunol. 2015;32:84\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eARAYA J, TSUBOUCHI K, SATO N, et al. PRKN-regulated mitophagy and cellular senescence during COPD pathogenesis [J]. Autophagy. 2019;15(3):510\u0026ndash;26.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-biotechnology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bbit","sideBox":"Learn more about [BMC Biotechnology](http://bmcbiotechnol.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/bbit/default.aspx","title":"BMC Biotechnology","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Obstructive sleep apnea, aging, senescencerelated genes, EGR1","lastPublishedDoi":"10.21203/rs.3.rs-6563621/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6563621/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackgroud\u003c/strong\u003e: A frequently encountered breathing condition, obstructive sleep apnea (OSA) primarily manifests while sleeping and is characterized by total or incomplete blockage of the upper respiratory tract. This disorder disrupts normal airflow, often leading to repeated pauses in breathing throughout the night. Aging significantly increases the risk of OSA, yet the underlying biomolecular connections between aging and OSA remain incompletely understood.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods\u003c/strong\u003e: This research integrates bioinformatics and machine learning methods. To discover and confirm possible biomarkers, a combination of WGCNA and machine learning techniques was utilized. Functional characterization of genes was achieved through GO and KEGG enrichment studies, which provided insights into biological pathways and molecular roles. A predictive nomogram was developed based on hub hub OSA-ARDEGs.Comprehensive immune infiltration analysis was conducted to elucidate the immunological microenvironment associated with key biomarkers.To experimentally validate computational predictions, RNA-seq and Western blotting analyses were performed to confirm EGR1 expression patterns in human type II alveolar epithelial cells.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults\u003c/strong\u003e: Investigative studies revealed genes exhibiting differential expression patterns, along with interconnected gene networks that showed notable associations with OSA. The analysis further demonstrated that these molecular networks are intricately tied to mechanisms of biological aging and immune system activity. Enrichment studies revealed that these genes are involved in multiple biological mechanisms, including processes related to inflammation and signaling cascades mediated by immune cells. Furthermore, EGR1 was validated experimentally as a critical gene involved in cellular senescence, immune regulation, and DNA damage response.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusion\u003c/strong\u003e: Our findings establish EGR1 as a crucial mediator in OSA pathogenesis, potentially driving disease progression through cellular senescence mechanisms. These results position EGR1 as a promising molecular target for developing therapeutic interventions against OSA-associated respiratory dysfunction.\u003c/p\u003e","manuscriptTitle":"Unveiling the role of EGR1 and hub senescencerelated genes in Type II alveolar epithelial cells senescence for Obstructive sleep Apnea","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-05-23 07:08:27","doi":"10.21203/rs.3.rs-6563621/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-06-10T03:48:25+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-06-07T00:56:47+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-06-04T12:57:29+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"315005287149916953945468467920915481078","date":"2025-05-31T03:05:12+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"65498852290696050757165935874459440218","date":"2025-05-30T22:33:37+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"8803019066576061149431532460589851388","date":"2025-05-30T15:07:35+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-05-20T13:46:03+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-05-20T13:34:29+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-05-19T05:39:23+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-05-19T04:02:32+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Biotechnology","date":"2025-05-19T04:01:27+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-biotechnology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bbit","sideBox":"Learn more about [BMC Biotechnology](http://bmcbiotechnol.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/bbit/default.aspx","title":"BMC Biotechnology","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"29a4afc2-8829-4297-979e-e1079d0c7560","owner":[],"postedDate":"May 23rd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-12-22T16:00:25+00:00","versionOfRecord":{"articleIdentity":"rs-6563621","link":"https://doi.org/10.1186/s12896-025-01067-0","journal":{"identity":"bmc-biotechnology","isVorOnly":false,"title":"BMC Biotechnology"},"publishedOn":"2025-12-19 15:57:25","publishedOnDateReadable":"December 19th, 2025"},"versionCreatedAt":"2025-05-23 07:08:27","video":"","vorDoi":"10.1186/s12896-025-01067-0","vorDoiUrl":"https://doi.org/10.1186/s12896-025-01067-0","workflowStages":[]},"version":"v1","identity":"rs-6563621","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6563621","identity":"rs-6563621","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}