{"paper_id":"0790f52d-ff58-48ed-9c2d-c3664cd789aa","body_text":"CYP1B1-Mediated Ferroptosis Defines a Biomarker and Therapeutic Target in COPD Across Multi-omics and Single-Cell | 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 CYP1B1-Mediated Ferroptosis Defines a Biomarker and Therapeutic Target in COPD Across Multi-omics and Single-Cell Lingfeng Liu, Mingjun Jiang, Muzi Lei, Kun Du, Dafei Wei, Xiangyang Ye, and 5 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8049999/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background: Chronic obstructive pulmonary disease (COPD) causes progressive airflow limitation and remains without disease-modifying therapy. Ferroptosis—iron-dependent lipid peroxidation—has emerged as a potential mechanism driving epithelial dysfunction and chronic inflammation; however, its upstream regulators in COPD remain incompletely defined. We hypothesized that integrative multi-omics could identify robust biomarkers and pathways, with translational potential for diagnosis and targeted therapy. Objective Primary: discover and validate biomarkers/pathways underlying COPD via bulk transcriptomics, single-cell analyses, machine learning (ML), and experimental validation. Secondary: evaluate CYP1B1 as a diagnostic biomarker and explore druggability through molecular docking. Methods: Public cohorts (GSE47460, GSE76925, GSE37768; total discovery/validation samples reported) were integrated with ComBat batch correction. Single-cell RNA-seq datasets (GSE196341, GSE135893; 12 COPD vs 12 controls) profiled cell-type localization. Differential expression, WGCNA, and enrichment (GO/KEGG/GSEA) were performed. A 12-algorithm ML framework (113 combinations) plus ANN modeling assessed diagnostic performance (ROC/AUC, confusion matrices, calibration). CIBERSORT estimated immune infiltration. A cigarette-smoke mouse model and 16HBE cell assays provided experimental validation; small-n proteomics were deposited (PXD068247; 3 vs 3). Molecular docking screened candidate CYP1B1-binding compounds and recorded binding energies. Statistical reporting included n, effect sizes, 95% CIs, and multiple-testing control. Results: Twenty-four hub genes were initially identified; four (BHLHE22, DPP6, DHRS9, CYP1B1) were prioritized by ML across 113 model combinations. In the training set, the combined model achieved an AUC of 0.996 (95% CI, 0.991–0.999), with an external validation AUC of 0.834 (95% CI, 0.755–0.906). Single-gene ROC in an external cohort yielded AUCs of 0.764–0.795, with CYP1B1 being the most consistent. Single-cell analysis localized CYP1B1 upregulation to airway secretory cells (ASCs) and linked high CYP1B1 expression to the activation of the ferroptosis pathway. In mouse and 16HBE models, cigarette smoke increased lung inflammation/fibrosis and upregulated CYP1B1; proteomics corroborated expression changes. Docking identified α-/β-naphthoflavone, chrysin, and naringenin as CYP1B1 binders (best energies ≈ −7.11 to −5.45 kcal/mol). Immune deconvolution associated CYP1B1 with macrophage and plasma-cell signals. Conclusions: Integrative multi-omics implicates CYP1B1-mediated ferroptosis in ASCs as a central pathway in COPD, with diagnostic promise and a tractable chemistry space. Prospective validation in larger cohorts, causal perturbation of CYP1B1–ferroptosis in vitro/in vivo, and pharmacology against prioritized ligands are warranted to translate these findings. COPD CYP1B1 ferroptosis multi-omics single-cell RNA sequencing biomarker discovery 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 Highlights 1. Multi-omics integration identifies CYP1B1 as a potential COPD diagnostic biomarker, with external validation. 2. CYP1B1 is upregulated in airway secretory cells and is linked to the activation of ferroptosis. 3. ML and ANN demonstrate strong predictive performance (AUC up to 0.996 in training; 0.834 in external validation). 4. Mouse and cell models demonstrate increased CYP1B1 expression in the context of inflammation and fibrosis; proteomics corroborate these findings. 5. Docking highlights α/β-naphthoflavone, chrysin, and naringenin as CYP1B1 binders ( ≈ − 7.11 to − 5.45 kcal/mol). Introduction Chronic obstructive pulmonary disease (COPD) remains a leading global cause of morbidity and mortality [ 1 ]. Clinically, COPD is defined by persistent respiratory symptoms and progressive airflow limitation, quantified by spirometry; however, current diagnostic criteria incompletely capture the disease heterogeneity, which spans complex structural remodeling of the lung, diverse inflammatory endotypes, and shifts in the airway microbiome [ 2 , 3 ]. Against this backdrop, there is a sustained need for mechanism-anchored biomarkers and actionable targets. Increasing evidence implicates oxidative stress and lipid peroxidation as central features of chronic airway injury, with ferroptosis—an iron-dependent, lipid peroxidation-driven cell-death program—emerging as a plausible contributor to epithelial dysfunction and tissue remodeling in COPD. Airflow obstruction in COPD typically progresses over several years and reflects a fixed pathobiology, including fibrotic narrowing of the small airways and loss of elastic recoil, leading to peripheral airway collapse and air trapping [ 4 ]. The small airway epithelium is the principal locus of cellular and histologic change [ 5 ], comprising a pseudostratified layer of secretory (club) cells, ciliated cells, and basal cells [ 6 ]. Inflammatory remodeling features increased neutrophils and macrophages in the airway lumina, as well as the accumulation of macrophages, T lymphocytes, and B lymphocytes in the airway wall and parenchyma [ 7 , 8 ]. These observations place epithelial injury and immune–epithelial crosstalk at the center of COPD progression, motivating an interrogation of epithelial cell-intrinsic death pathways that may amplify chronic inflammation. Genomic and transcriptomic studies suggest that interindividual susceptibility to COPD is related to gene expression programs within the airway epithelium [ 9 ], yet the upstream regulators that couple xenobiotic/redox metabolism to ferroptosis remain poorly understood. CYP1B1, a cytochrome P450 implicated in xenobiotic biotransformation and oxidative metabolism, is a biologically plausible node at the interface of redox stress and ferroptosis in airway secretory cells, where lipid peroxidation may promote persistent epithelial dysfunction. However, causal links between CYP1B1 activity, ferroptosis signaling, and COPD pathogenesis are not established, underscoring the need for systems-level discovery integrated with cellular and in vivo validation. To address this gap, we employed an integrative multi-omics strategy combining bulk transcriptomics with weighted gene co-expression network analysis (WGCNA), differential expression, and pathway enrichment (GO/KEGG/GSEA), followed by a machine-learning framework (12 algorithms across 113 combinations) and an artificial neural network (ANN) to derive and test diagnostic models (ROC/AUC). Immune deconvolution (CIBERSORT) characterized the inflammatory context, while single-cell RNA sequencing mapped candidate gene expression to specific lung cell types. Experimental validation in cigarette–smoke–exposed models and orthogonal proteomics supported the biological plausibility, and molecular docking explored the therapeutic tractability. We hypothesized that CYP1B1 marks—and may contribute to—ferroptosis-linked epithelial pathology in COPD, would generalize across multi-omics layers, demonstrate diagnostic utility, and reveal druggable chemistry space. We aimed to identify robust COPD biomarkers/pathways, localize them to disease-relevant cell types, link them to ferroptosis signatures, validate them in models, and nominate candidate CYP1B1-targeting ligands for future experimental testing. Materials and Methods Data sources and preprocessing (bulk expression, single-cell, proteomics) We integrated bulk gene-expression datasets, single-cell RNA sequencing (scRNA-seq) data, and discovery proteomics. Bulk datasets were retrieved from GEO and included GSE47460 (platforms GPL6480 and GPL14550; healthy controls n = 17 and COPD n = 75 for GPL6480; healthy controls n = 91 and COPD n = 145 for GPL14550) [ 10 ], with GSE76925 (COPD n = 111, controls n = 40) [ 11 ] and GSE37768 (COPD n = 18, controls n = 9) serving as independent validation cohorts. Probe-level preprocessing removed empty probes and those mapping to multiple genes; for genes with multiple probes, the probe with the highest mean expression was retained, and when multiple probes per gene were required, median aggregation was used. Datasets were merged after re-annotation to gene symbols using the official platform annotation files and were batch-corrected and normalized using ComBat (R package sva) [ 12 ]. Subsequent analyses used these normalized matrices. scRNA-seq data were obtained from GSE196341 and GSE135893 (12 COPD and 12 control donors in total) [ 13 ]. Discovery proteomics were deposited to ProteomeXchange (iProX partner) under PXD068247 [ 14 , 15 ]. A summary of datasets is provided in Table 1 . Table 1 Datasets included in the study Datasets ID Year Date type Country N (samples) GSE47460 2013 Bulk RNA-Seq American Case = 220, Control = 108 GSE76925 2017 Bulk RNA-Seq American Case = 111, Control = 40 GSE37768 2016 Bulk RNA-Seq Spain Case = 18, Control = 9 GSE196341 2022 scRNA-Seq American Case = 12 GSE135893 2019 scRNA-Seq American Control = 12 PXD068247 2025 Proteomics China Case = 3, Control = 3 Differentially expressed genes (DEGs) Differential expression between COPD and control groups was assessed using limma, with model terms including group and batch/platform, where applicable. Unless otherwise indicated, significance is used |log2FC| > 1.5 and p < 0.01 after multiple-testing correction. Results were visualized with volcano plots and clustered heatmaps (ggplot2, pheatmap) [ 16 – 18 ]. Weighted gene co-expression network analysis (WGCNA) Co-expression networks were constructed using WGCNA. Signed networks were generated from pairwise Pearson correlations, selecting the soft-thresholding power β to approximate scale-free topology (reporting scale-free fit R 2 and mean connectivity). Adjacency matrices were transformed to topological overlap matrices (TOM), and hierarchical clustering with dynamic tree cutting identified gene modules (report minimum module size and merge thresholds). Module eigengenes were correlated with COPD status, and the midnight blue module, showing the strongest association with COPD, was selected for downstream analysis [ 19 ]. Machine-learning feature selection and classification Batch-corrected, z-score–standardized expression matrices were used to derive features and train classifiers. We implemented a multi-algorithm screen comprising 12 algorithms across 113 combinations (e.g., Logistic Regression, LASSO/Elastic Net, Ridge, SVM-Linear/RBF, Random Forest, ExtraTrees, GBM, XGBoost, k-NN, Naïve Bayes), with feature selection embedded within folds to prevent information leakage. Model selection employed stratified k-fold cross-validation on the training data, followed by external validation on independent cohorts (GSE76925, GSE37768). Performance metrics included ROC-AUC with 95% CIs, PR-AUC, accuracy, F1, and calibration (Brier score, calibration curves). Feature importance was summarized via permutation importance and SHAP, where applicable. The highest mean external AUC chose the optimal model. Gene set enrichment analysis (GSEA) and functional annotation To contextualize biological processes, we performed GSEA using MSigDB KEGG (c2.cp.kegg.Hs.symbols) and GO (c5.go.symbols) collections, and complementary over-representation analyses with clusterProfiler; visualizations employed enrichplot. Significance thresholds were based on field standards for GSEA and ORA, as specified in the Results [ 20 ]. Single-cell RNA-seq analysis scRNA-seq data (GSE196341, GSE135893) were processed in Seurat. Cells were filtered by mitochondrial RNA > 15%, ribosomal RNA > 5%, erythrocyte RNA > 1%, detected genes < 300 or > 7,000. Data were normalized (method as specified in the Results), and batch effects were mitigated via Harmony integration. Highly variable genes (n = 2,000) were used for PCA and UMAP (dimensions = 22). Clustering utilized a shared nearest neighbor graph (resolution 0.2) and was manually annotated using the top 10 marker genes per cluster. We visualized donor-wise UMAPs and cell-type distributions, and mapped CYP1B1 expression to airway secretory cells. Pseudotime analysis was performed using Monocle, and intercellular communication was inferred using CellChat [ 13 ]. Artificial neural network (ANN) A feed-forward ANN was trained to classify Control vs COPD using binarized signatures derived from characteristic genes (upregulated genes coded 1 if ≥ median; downregulated genes reverse-coded). The network consisted of an input layer (size equal to the number of features), a single hidden layer (5 neurons, using the sigmoid activation function), and a binary output. Training was performed using a neural network with Rprop + optimization (seed = 12345678). Performance was evaluated via confusion matrices, accuracy per class, and ROC-AUC; feature contributions were estimated using the Garson algorithm. Immune cell deconvolution From normalized bulk expression, immune cell proportions were estimated using CIBERSORT (LM22), with default permutations (state number) and quantile normalization settings appropriate for the tissue type. Samples with CIBERSORT p < 0.05 were retained. Group differences were tested using the Wilcoxon rank-sum test with the Benjamini–Hochberg correction across 22 cell types. Spearman correlations between characteristic genes and immune subsets were computed with multiple-testing correction. Cigarette smoke extract (CSE) preparation Mainstream smoke from Hongta brand filter-tipped cigarettes (Yunnan, China) was drawn by a peristaltic pump and bubbled through 10 mL DMEM per cigarette. The pH was adjusted to 7.2, and the solution was sterilized using a 0.22-µm filter to yield 10% CSE (v/v); working concentrations were prepared freshly for each experiment, for particulate matter (PM2.5), stock suspensions were prepared in DMSO at 400 mg/mL, vortexed and sonicated for 2 h, centrifuged to remove insoluble particulates, and supernatants stored at − 20°C [21] . Cell culture and treatments 16HBE cells were cultured in DMEM + 10% FBS at 37°C, 5% CO₂. Cells were exposed to 2.5% CSE for 24 h. Unless otherwise stated, vehicle-treated cells served as controls [21] . Mouse COPD model Specific pathogen-free C57BL/6J mice (15–20 g) were housed at 25°C, 12/12 h light–dark, with food and water ad libitum. The institutional ethics committee approved protocols. Mice were randomized into two groups: control (n = 8) and COPD model (n = 8). COPD was induced by whole-body exposure in a 60×57×100 cm chamber to 9 cigarettes twice daily, 6 days/week, for 6 months; each session lasted ~ 2 h with ≥ 4 h interval. Control mice were exposed to room air for the same duration. Before exposure, mice received sterile saline aerosol (30 minutes, twice daily). Tissues (lung), bronchoalveolar lavage fluid (BALF), and blood were collected at endpoint [22] . Lung function testing Lung mechanics were evaluated using a FinePointe system (Buxco) according to the manufacturer’s instructions, recording airway resistance/compliance indices [23] . Enzyme-linked immunosorbent assays (ELISAs) Concentrations of IL-1β, IL-6, IL-8, and TNF-α in BALF and serum were quantified per the manufacturers’ protocols. Histology and staining The lungs were fixed in 4% paraformaldehyde, processed, embedded in paraffin, sectioned, and stained with H&E, PAS, and Picrosirius Red using standard protocols. Immunohistochemistry (IHC) IHC was performed on paraffin-embedded lung sections using standard antigen retrieval and detection procedures. Molecular docking To explore therapeutic tractability, we performed molecular docking against CYP1B1. Three-dimensional structures for ligands were retrieved (PubChem) and prepared in AutoDockTools 1.5.7; receptor structure (CYP1B1) was obtained from RCSB PDB ( www.rcsb.org ) and prepared analogously. The grid center and box size were defined around the active site, and docking was performed using the Lamarckian Genetic Algorithm under default parameters. Rigid docking was performed (AutoDockTools 1.6.7), and the top-scoring pose was retained. Statistical analysis Analyses were performed in R 4.4.2. Unless otherwise stated, two-sided tests were used, with α = 0.05. For multiple comparisons, the Benjamini–Hochberg correction was applied. Classification performance was summarized by ROC-AUC (95% CI), PR-AUC, accuracy, F1 score, confusion matrices, and calibration curves. Data are reported as mean ± SD (or median [IQR]) with n indicating biological replicates. Results Multi-omics discovery and hub-gene prioritization A study overview is provided in Fig. 1 . Using WGCNA on the training cohort, we selected a soft-threshold power of β = 5 (Fig. 2 A) and identified 20 gene modules through dynamic tree cutting (Fig. 2 B). The midnightblue module showed the strongest association with COPD status (module–trait r = 0.893, P = 5.4×10 − 32 ; Fig. 2 C–D) and contained 89 genes. After normalization of the GSE47460 expression matrix, differential expression analysis identified 170 DEGs between COPD and control samples (volcano plot, Fig. 2 E; top 50 heat map, Fig. 2 F). Intersecting module genes with DEGs yielded 24 hub genes (Fig. 2 G), which were advanced for biomarker screening. Machine-learning screen and diagnostic performance We evaluated 12 algorithms across 113 model combinations to prioritize diagnostic features (Fig. 3 A). In the training set, the combined model achieved an AUC of 0.996 (95% CI, 0.991–0.999; Fig. 3 B). External validation in GSE76925 yielded an AUC of 0.834 (95% CI, 0.755–0.906; Fig. 3 C), indicating preserved discrimination in an independent cohort. Confusion matrices for the training and validation sets demonstrated high sensitivity and specificity (Figs. 3 D and 3 E). Across models, BHLHE22, DPP6, DHRS9, and CYP1B1 were most frequently selected and ranked highly in terms of importance. Correlation analyses revealed strong positive co-variation between BHLHE22 and CYP1B1 (r = 0.62) and an inverse correlation between DPP6 and CYP1B1 (r = -0.55; Fig. 3 G). Differential expression visualizations confirmed that three genes were upregulated in COPD, and DPP6 was downregulated (Fig. 3 F). External validation and pathway context for top biomarkers Per-gene ROC curves in the independent GSE37768 cohort showed that BHLHE22, DPP6, and CYP1B1 achieved AUCs of 0.764–0.795, whereas DHRS9 performed lower (Fig. 4 A). A nomogram integrating these predictors emphasized CYP1B1 as the strongest contributor to diagnostic probability (Fig. 4 B–C). To contextualize biology, GSEA revealed that BHLHE22-high samples were enriched for collagen fibril organization and extracellular matrix remodeling (GO) and for chemokine signaling, complement/coagulation, cytokine–receptor interaction, and NK cell–mediated cytotoxicity (KEGG) (Fig. 4 D–E). Similarly, CYP1B1-high samples showed GO enrichment for collagen processes, response to interleukin-1, and metallopeptidase activity, alongside KEGG enrichment for chemokine signaling, complement/coagulation, intestinal immune network for IgA, and NK cell–mediated cytotoxicity (Fig. 4 F–G). ANN construction, calibration, and independent checks We built an ANN using BHLHE22, DPP6, and CYP1B1 (architecture and weights, Fig. 5 A). The model demonstrated close agreement between the predicted and observed risks in both the training and validation datasets (calibration plots, Fig. 5 B). In GSE37768, all three biomarkers were significantly upregulated in COPD versus controls, consistent with discovery analyses (Fig. 5 C–D). Collectively, the ANN achieved robust discrimination with good calibration, supporting the translational feasibility of a parsimonious gene-based classifier. Single-cell resolution of CYP1B1 and ferroptosis signatures Integration of human lung scRNA-seq from COPD and control donors resolved 16 major cell types (marker panels in Fig. 6 A; UMAP in Fig. 6 B), with proportion differences summarized in Fig. 6 C. While BHLHE22, DPP6, and CYP1B1 were detected across multiple lineages, CYP1B1 showed significantly higher expression in airway secretory cells (ASCs) from COPD relative to controls (Fig. 6 D; report log₂FC, adj. P). Stratifying ASCs by CYP1B1 median defined high- and low-expressing subpopulations; differential analysis indicated ferroptosis pathway enrichment in CYP1B1-high ASCs (KEGG) with increased AUCell ferroptosis scores (Fig. 6 E–F). Pseudotime and cell–cell communication implicate ASC programs in the pathogenesis of COPD. Monocle-based trajectories suggested bifurcating epithelial lineages, with ASCs enriched at a terminal branch (Fig. 7 A–B), consistent with a remodeled secretory program in COPD. Global CellChat networks (Figs. 7 C–D) revealed increased outgoing and incoming signaling to/from ASCs, particularly with alveolar macrophages and other immune lineages. An ASC-centric view (Fig. 7 E) highlighted ASCs as communication hubs, supporting a model in which secretory epithelial programs shape the COPD inflammatory microenvironment. Immune infiltration correlates with hub genes. Bulk deconvolution revealed significant differences in multiple immune subsets between COPD and control subjects, including plasma cells, M0 macrophages, activated mast cells, and follicular helper T cells (Fig. 8 A), with inter-subset correlations summarized in Fig. 8 B. Expression of BHLHE22 and CYP1B1 correlated positively with plasma cells and M0 macrophages, whereas DPP6 correlated negatively with these subsets and positively with M2 macrophages (Fig. 8 C–E), suggesting coordinated epithelial–immune remodeling in COPD. In vivo and in vitro evidence of CYP1B1 upregulation with smoke exposure In a 6-month cigarette-smoke mouse model, systemic and airway inflammation increased over time, as indicated by ELISAs (Fig. 9 A–H), accompanied by higher pulmonary CYP1B1 expression (Fig. 9 I). In 16HBE cells exposed to CSE, proteomic profiling corroborated CYP1B1 upregulation (Fig. 9 J). Histopathology in smoke-exposed mice revealed enlarged alveolar septa, increased glycogen deposition, and fibrosis (H&E, PAS, Picrosirius Red; Fig. 10 A–C), accompanied by a significant impairment of lung function (Fig. 10 D–K). Ferroptosis-related indicators also increased significantly (Fig. 10 L-N). These findings support a link between cigarette smoke exposure, epithelial remodeling, and CYP1B1 induction. Inhibition of CYP1B1 attenuates ferroptosis signals In the CSE model constructed by cigarette smoke-induced 16-HBE cells, lentiviral transfection was used to knock down the expression of CYP1B1, and we found that ferroptosis signal-related indicators such as GPX4, MDA, and 4-HNE had a significant decline (Fig. 11 A-C). This suggests that inhibition of CYP1B1 expression can inhibit ferroptosis in airway epithelial cells, which may be helpful in the treatment of COPD. Docking nominates natural compounds with predicted CYP1B1 binding. DGIdb-guided filtering of traditional Chinese medicine (TCM) compounds prioritized α-naphthoflavone, β-naphthoflavone, chrysin, and naringenin as putative CYP1B1 binders. Rigid docking (PyMOL/AutoDock workflow) predicted binding energies < − 5.0 kcal/mol (− 7.11, − 6.86, − 5.56, − 5.45 kcal/mol; Fig. 12 A–D), consistent with favorable interactions in the CYP1B1 active site. These in silico data are hypothesis-generating and motivate biochemical validation of enzyme inhibition and selectivity. Discussion COPD remains a leading global contributor to morbidity and mortality [ 24 ]. While the disease is heterogeneous [ 25 ], its clinical definition centers on persistent airflow limitation and chronic respiratory symptoms. To enhance diagnostic precision and mechanistic understanding, we employed an integrative multi-omics pipeline to identify and validate biomarkers with potential clinical utility. Our analyses identify CYP1B1 as a promising biomarker and mechanistic node in COPD, with convergent evidence spanning bulk transcriptomics, machine learning, single-cell mapping, and experimental models. Using publicly available COPD cohorts, we defined 170 DEGs, mapped them onto co-expression modules, and intersected signals to yield 24 hub genes. A 12-algorithm/113-combination screen prioritized BHLHE22, CYP1B1, and DPP6, with external validation supporting diagnostic discrimination (combined model AUC up to 0.996 in training and 0.834 in GSE76925; single-gene AUCs 0.764–0.795 in GSE37768). Single-cell analyses localized CYP1B1 upregulation to airway secretory cells (ASCs) in COPD, where high CYP1B1 expression coincided with enrichment of ferroptosis signatures. In smoke-exposed mouse lungs and CSE-treated bronchial epithelial cells, CYP1B1 expression increased alongside histologic remodeling and impaired lung function. Collectively, these observations support an association between CYP1B1 upregulation, epithelial stress programs consistent with ferroptosis, and the pathobiology of COPD, while acknowledging that causality requires targeted perturbation. CYP1B1, a xenobiotic-metabolizing cytochrome P450, bioactivates environmental toxicants (including polycyclic aromatic hydrocarbons) and modulates steroid and eicosanoid pathways [ 26 , 27 ]. In the lung, CYP1B1 activity has been implicated in oxidative stress and inflammatory signaling pertinent to chronic airway disease [ 27 , 28 ]. Altered CYP1B1 expression has been linked to epithelial dysfunction—airway remodeling, impaired mucociliary clearance, and heightened susceptibility to inhaled insults—features that align with COPD pathogenesis [ 29 ]. Moreover, CYP1B1 polymorphisms have been associated with interindividual variation in lung function decline and COPD risk [ 30 ]. Integrating these observations, our data position CYP1B1 as a biologically plausible mediator of smoke-related redox imbalance and epithelial remodeling in COPD, and a tractable node for therapeutic exploration [ 31 , 32 ]. Primary immunodeficiencies (PIDs) encompass defects in cellular and humoral immunity that predispose to recurrent airway infections and chronic inflammation. More than half of PID subtypes impair antibody production and contribute to recurrent sinusitis and pulmonary infections; persistent upper/lower respiratory infections can drive airway inflammation, obstruction, and, in some cases, structural remodeling compatible with COPD phenotypes [ 33 ]. Respiratory complications of PIDs are associated with downstream risks—including severe asthma, bronchiectasis, and COPD—and increased mortality [ 34 ]. In the context of our findings, these observations highlight how host-defense perturbations can exacerbate epithelial injury and immune–epithelial crosstalk in susceptible individuals. Single-cell results indicated that differentially expressed genes were enriched in epithelial compartments, with ASCs showing the clearest CYP1B1 signal. Epithelial–macrophage crosstalk is central to airway remodeling, as alveolar and airway macrophages orchestrate pathogen clearance, surfactant turnover, tissue repair, and homeostasis. In COPD, they exhibit impaired phagocytosis/efferocytosis, dysregulated cytokine release, and heightened oxidative stress [ 35 ]. Beyond the lung, macrophage inflammatory phenotypes and lysosomal signaling influence disease processes across organs [ 36 , 37 ]; by analogy, macrophage dysfunction in COPD may intersect with epithelial redox pathways, including lipid peroxidation and iron handling. Our data suggest that CYP1B1-high ASCs reside within an inflamed network where macrophage interactions are increased, consistent with a feed-forward loop of epithelial stress and immune activation. Limitation This study has limitations. First, although we leveraged multiple GEO cohorts, sample availability limits power for some subgroup analyses. Second, the external validation cohorts differ in clinical covariates and platforms, which may affect generalizability. Third, while single-cell analyses map CYP1B1 to airway secretory cells and associate high expression with ferroptosis signatures, we did not perform genetic or pharmacologic perturbation of CYP1B1 with standardized ferroptosis rescue controls (e.g., ferrostatin-1, liproxstatin-1, deferoxamine). Finally, docking results are in silico and require biochemical confirmation (enzyme inhibition, selectivity, and cellular target engagement). Conclusion In aggregate, our integrative analyses identify CYP1B1 as a potential biomarker for COPD and a putative mediator of epithelial stress, consistent with ferroptosis in airway secretory cells. Clinically, CYP1B1 expression and derived gene-based models may aid diagnostic stratification; mechanistically, CYP1B1 presents a druggable entry point for modulating redox-lipid peroxidation pathways. Prioritized natural compounds from docking provide testable leads, but their translation will require validating target engagement, ferroptosis-specific rescue, and safety/PK. Considerations. Prospective, multi-center studies integrating single-cell profiling, circulating biomarkers, and interventional perturbations should establish clinical utility and clarify therapeutic potential. Abbreviations Abbreviation Definition COPD Chronic Obstructive Pulmonary Disease scRNA Single-cell RNA DEGs Differentially Expressed Genes WGCNA Weighted Gene Co-Expression Network Analysis ROC Receiver Operating Characteristic AUC Area Under the Curve GSEA Gene Set Enrichment Analysis GO Gene Ontology KEGG Kyoto Encyclopedia of Genes and Genomes ANN Artificial Neural Network UMAP Uniform Manifold Approximation and Projection CSEs Cigarette Smoke Extracts Ams Alveolar Macrophages AT2s Alveolar Type II epithelial cells CD8Ts CD8+ T cells ECs Endothelial cells Cils Ciliated cells Monos Monocytes ASCs Airway secretory cells BCs B cells AT1s Alveolar Type I cells PMVECs Pulmonary microvascular endothelial cells Fbs Fibroblasts Sers Airway serous cells MCs Mast cells LECs Lymphatic endothelial cells PCs Proliferating cells Mac_Monos Macrophages/Monocytes TCM Traditional Chinese medicine Declarations Ethics approval and consent to participate The study involving animals was reviewed and approved by Ethics Committee of Guangzhou Medical University (GY2025-036). Availability of data and materials All data generated or analysed during this study are included in this published article. Competing interest The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Funding The author(s) declare that financial support was received for the research and/or publication of this article. This work was sponsored by National Natural Science Foundation of China (81900044 and 82170500), Shanghai Municipal Education Commission (2021 Technology and Innovation-03-163), Natural Science Foundation of Hunan Province (2021JJ40484 and 2024JJ5350), Major Project of Guangzhou National Laboratory (GZNL2024A02005) and the grant of State Key Laboratory of Respiratory Disease (SKLRD-Z-202315). Authors’ contributions LL and MJ: Conceptualization, Data curation, Formal Analysis, Investigation, Methodology, Software, Visualization, Writing– original draft. ML and KD: Data curation, Investigation, Methodology, Software, Visualization. DW, XY and XT: Data curation, Methodology, Software, Visualization. ZM, YT and RS: Data curation, Software, Visualization. SL: Conceptualization, Funding acquisition, Investigation, Project administration, Resources, Supervision, Writing– review & editing. Acknowledgments We thank the participants and staff of the National Center for Biotechnology Information. We thank Wen Qianmei and Zuo yujie form the First Affiliated Hospital of Guangzhou Medical University for their assistance in the construction of the mouse model. At the same time, we would also like to acknowledge assistance from proofreaders and editors. Generative AI statement The author(s) declare that no Generative AI was used in the creation of this manuscript. References Christenson SA, Smith BM, Bafadhel M, Putcha N. Chronic obstructive pulmonary disease. Lancet. 2022;399(10342):2227–42. Labaki WW, Rosenberg SR. Chronic Obstructive Pulmonary Disease. Ann Intern Med. 2020;173(3):Itc17–32. Brightling C, Greening N. Airway inflammation in COPD: progress to precision medicine. Eur Respir J, 2019. 54(2). Barnes PJ. Cellular and molecular mechanisms of asthma and COPD. Clin Sci (Lond). 2017;131(13):1541–58. Wistuba II, Mao L, Gazdar AF. Smoking molecular damage in bronchial epithelium. Oncogene. 2002;21(48):7298–306. Mercer RR, Russell ML, Roggli VL, Crapo JD. Cell number and distribution in human and rat airways. Am J Respir Cell Mol Biol. 1994;10(6):613–24. Brusselle GG, Joos GF, Bracke KR. New insights into the immunology of chronic obstructive pulmonary disease. Lancet. 2011;378(9795):1015–26. Hogg JC, Chu F, Utokaparch S, et al. The nature of small-airway obstruction in chronic obstructive pulmonary disease. N Engl J Med. 2004;350(26):2645–53. Zhao J, Cheng W, He X, et al. Chronic Obstructive Pulmonary Disease Molecular Subtyping and Pathway Deviation-Based Candidate Gene Identification. Cell J. 2018;20(3):326–32. Tan J, Tedrow JR, Dutta JA, et al. Expression of RXFP1 Is Decreased in Idiopathic Pulmonary Fibrosis. Implications for Relaxin-based Therapies. Am J Respir Crit Care Med. 2016;194(11):1392–402. Morrow JD, Zhou X, Lao T, et al. Functional interactors of three genome-wide association study genes are differentially expressed in severe chronic obstructive pulmonary disease lung tissue. Sci Rep. 2017;7:44232. Leek JT, Johnson WE, Parker HS, et al. The sva package for removing batch effects and other unwanted variation in high-throughput experiments. Bioinformatics. 2012;28(6):882–3. Habermann AC, Gutierrez AJ, Bui LT et al. Single-cell RNA sequencing reveals profibrotic roles of distinct epithelial and mesenchymal lineages in pulmonary fibrosis. Sci Adv, 2020. 6(28): p. eaba1972. Ma J, Chen T, Wu S, et al. iProX: an integrated proteome resource. Nucleic Acids Res. 2019;47(D1):D1211–7. Chen T, Ma J, Liu Y, et al. iProX in 2021: connecting proteomics data sharing with big data. Nucleic Acids Res. 2022;50(D1):D1522–7. Ritchie ME, Phipson B, Wu D, et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 2015;43(7):e47. Kolde R. pheatmap: Pretty Heatmaps. 2010. Wickham H. Ggplot2: Elegant graphics for data analysis. Cham, Switzerland: Springer International Publishing; 2016. Langfelder P, Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics. 2008;9:559. Subramanian A, Tamayo P, Mootha VK, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005;102(43):15545–50. Fan X, Dong T, Yan K et al. PM2.5 increases susceptibility to acute exacerbation of COPD via NOX4/Nrf2 redox imbalance-mediated mitophagy. Redox Biol, 2023. 59. Guan R, Wang J, Cai Z et al. Hydrogen sulfide attenuates cigarette smoke-induced airway remodeling by upregulating SIRT1 signaling pathway. Redox Biol, 2020. 28. He F, Liao B, Pu J et al. Exposure to Ambient Particulate Matter Induced COPD in a Rat Model and a Description of the Underlying Mechanism. Sci Rep, 2017. 7(1). Raherison C, Girodet PO. Epidemiology of COPD. Eur Respir Rev. 2009;18(114):213–21. Mannino DM. COPD: epidemiology, prevalence, morbidity and mortality, and disease heterogeneity. Chest. 2002;121(5 Suppl):s121–6. Esteve-Codina A, Hofer TP, Burggraf D, et al. Gender specific airway gene expression in COPD sub-phenotypes supports a role of mitochondria and of different types of leukocytes. Sci Rep. 2021;11(1):12848. Zhu Y, Dutta S, Han Y, et al. Oxidative stress promotes lipid-laden macrophage formation via CYP1B1. Redox Biol. 2025;79:103481. Zhang S, Duan H, Yan J. Identifying biomarkers of endoplasmic reticulum stress and analyzing immune cell infiltration in chronic obstructive pulmonary disease using machine learning. Front Med (Lausanne). 2024;11:1462868. Reddy KD, Lan A, Boudewijn IM, et al. Current Smoking Alters Gene Expression and DNA Methylation in the Nasal Epithelium of Patients with Asthma. Am J Respir Cell Mol Biol. 2021;65(4):366–77. Obernolte H, Niehof M, Braubach P, et al. Cigarette smoke alters inflammatory genes and the extracellular matrix - investigations on viable sections of peripheral human lungs. Cell Tissue Res. 2022;387(2):249–60. Misiukiewicz-Stępien P, Mierzejewski M, Zajusz-Zubek E et al. RNA-Seq Analysis of UPM-Exposed Epithelium Co-Cultivated with Macrophages and Dendritic Cells in Obstructive Lung Diseases. Int J Mol Sci, 2022. 23(16). Fitzgerald LF, Lackey J, Moussa A, et al. Chronic aryl hydrocarbon receptor activity impairs muscle mitochondrial function with tobacco smoking. J Cachexia Sarcopenia Muscle. 2024;15(2):646–59. Berger M, Geng B, Cameron DW, et al. Primary immune deficiency diseases as unrecognized causes of chronic respiratory disease. Respir Med. 2017;132:181–8. Jang JH, Kim JH, Park HS. Current Issues in the Management of IgG Subclass Deficiencies in Adults With Chronic Respiratory Diseases. Allergy Asthma Immunol Res. 2023;15(5):562–79. Hu YS, Shao XJ, Xing L et al. Single-Cell Sequencing of Lung Macrophages and Monocytes Reveals Novel Therapeutic Targets in COPD. Cells, 2023. 12(24). Ogawa D, Stone JF, Takata Y, et al. Liver x receptor agonists inhibit cytokine-induced osteopontin expression in macrophages through interference with activator protein-1 signaling pathways. Circ Res. 2005;96(7):e59–67. Weber K, Schilling JD. Distinct lysosome phenotypes influence inflammatory function in peritoneal and bone marrow-derived macrophages. Int J Inflam, 2014. 2014: p. 154936. Additional Declarations No competing interests reported. Supplementary Files WB.zip Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {\"props\":{\"pageProps\":{\"initialData\":{\"identity\":\"rs-8049999\",\"acceptedTermsAndConditions\":true,\"allowDirectSubmit\":true,\"archivedVersions\":[],\"articleType\":\"Research Article\",\"associatedPublications\":[],\"authors\":[{\"id\":567312858,\"identity\":\"a9fe2e3b-a615-4f55-9f27-37d250721acd\",\"order_by\":0,\"name\":\"Lingfeng Liu\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"University of South China\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Lingfeng\",\"middleName\":\"\",\"lastName\":\"Liu\",\"suffix\":\"\"},{\"id\":567312859,\"identity\":\"92bb95b3-cc2f-444e-84be-0886be0cb370\",\"order_by\":1,\"name\":\"Mingjun Jiang\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Tongji University\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Mingjun\",\"middleName\":\"\",\"lastName\":\"Jiang\",\"suffix\":\"\"},{\"id\":567312860,\"identity\":\"f4818485-915a-442b-a94d-6c7b17894904\",\"order_by\":2,\"name\":\"Muzi Lei\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"University of South China\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Muzi\",\"middleName\":\"\",\"lastName\":\"Lei\",\"suffix\":\"\"},{\"id\":567312861,\"identity\":\"3ac2e107-22cb-47e3-addc-f6b54e953eee\",\"order_by\":3,\"name\":\"Kun Du\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"University of South China\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Kun\",\"middleName\":\"\",\"lastName\":\"Du\",\"suffix\":\"\"},{\"id\":567312862,\"identity\":\"747e1121-5731-4fc1-b26c-2d49054d8aaf\",\"order_by\":4,\"name\":\"Dafei Wei\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"University of South China\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Dafei\",\"middleName\":\"\",\"lastName\":\"Wei\",\"suffix\":\"\"},{\"id\":567312863,\"identity\":\"124e30e3-1bcb-4ebc-8c36-ad95c1449ed6\",\"order_by\":5,\"name\":\"Xiangyang Ye\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Xiangya Hospital Central South University\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Xiangyang\",\"middleName\":\"\",\"lastName\":\"Ye\",\"suffix\":\"\"},{\"id\":567312864,\"identity\":\"2f7220d8-937b-441b-bc57-8a1c13464a7b\",\"order_by\":6,\"name\":\"Xiaowu Tan\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"University of South China\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Xiaowu\",\"middleName\":\"\",\"lastName\":\"Tan\",\"suffix\":\"\"},{\"id\":567312865,\"identity\":\"cd8ce670-051d-4c51-b14e-f473d6def98e\",\"order_by\":7,\"name\":\"Zhuang Mi\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"University of South China\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Zhuang\",\"middleName\":\"\",\"lastName\":\"Mi\",\"suffix\":\"\"},{\"id\":567312866,\"identity\":\"cf13b7dd-8406-4c41-a1be-889ec514815e\",\"order_by\":8,\"name\":\"Yao Tang\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"University of South China\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Yao\",\"middleName\":\"\",\"lastName\":\"Tang\",\"suffix\":\"\"},{\"id\":567312867,\"identity\":\"499f1ee4-fcf8-4dc6-a498-1f768506a70c\",\"order_by\":9,\"name\":\"Ruiting Sun\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Guangzhou Medical University\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Ruiting\",\"middleName\":\"\",\"lastName\":\"Sun\",\"suffix\":\"\"},{\"id\":567312868,\"identity\":\"19917a39-ac33-420c-ac3b-63f59f6ed7da\",\"order_by\":10,\"name\":\"Sha Liu\",\"email\":\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA7klEQVRIie3RPQrCMBiA4ZRAuny2a0JBrxDpqOBVEgRnRwfBlkAdXestnJwrBV16gLpZBGdBEF3E+LMJqeDikBeyBB6SfEHIZvvDfOzEOzGiQDDOXltZDWFTpfix6DZ9l4jvCC82CZsng5DNgH9HUClVCCSXixzO7DJGTa8UzmloEE4q4z3AgzSWAaxRyEqBg9RAMH2cQt/EiZBclIJgMBBCZRIAf17swK4RmtQSgFXCUqGfr4DQRoQEryPUjfWQMz1kTMIOrGl7XlQqMJFe7lY7edNf6efV9jLutrxNf3UykY9T9dJDsNlsNttv3QFByEzMiZcLCwAAAABJRU5ErkJggg==\",\"orcid\":\"\",\"institution\":\"University of South China\",\"correspondingAuthor\":true,\"prefix\":\"\",\"firstName\":\"Sha\",\"middleName\":\"\",\"lastName\":\"Liu\",\"suffix\":\"\"}],\"badges\":[],\"createdAt\":\"2025-11-06 16:23:20\",\"currentVersionCode\":1,\"declarations\":\"\",\"doi\":\"10.21203/rs.3.rs-8049999/v1\",\"doiUrl\":\"https://doi.org/10.21203/rs.3.rs-8049999/v1\",\"draftVersion\":[],\"editorialEvents\":[],\"editorialNote\":\"\",\"failedWorkflow\":false,\"files\":[{\"id\":99788399,\"identity\":\"40af81ad-51ac-4309-b199-d4c9c34fcad5\",\"added_by\":\"auto\",\"created_at\":\"2026-01-08 12:46:35\",\"extension\":\"docx\",\"order_by\":0,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"acdc-reference\",\"size\":24774181,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"RevisedManuscript.docx\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/671331af9cf973b354efd05d.docx\"},{\"id\":99343757,\"identity\":\"9ac790a7-4375-4a5f-8b3e-c91318a2d253\",\"added_by\":\"auto\",\"created_at\":\"2026-01-01 08:01:13\",\"extension\":\"json\",\"order_by\":1,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"acdc-reference\",\"size\":12613,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"b7f0e3a8a0604c6a838626e24a30ec96.json\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/f691a2fbab0aa5c404834460.json\"},{\"id\":99788974,\"identity\":\"a6454c9e-1cb5-410b-ab34-0d7652d4d4cc\",\"added_by\":\"auto\",\"created_at\":\"2026-01-08 12:48:26\",\"extension\":\"zip\",\"order_by\":2,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"acdc-reference\",\"size\":93492254,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"WB.zip\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/cc9c5c476574f841995331e3.zip\"},{\"id\":99343758,\"identity\":\"986243b9-5705-48c5-965e-bb0a00db7adf\",\"added_by\":\"auto\",\"created_at\":\"2026-01-01 08:01:13\",\"extension\":\"xml\",\"order_by\":3,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"acdc-reference\",\"size\":113232,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"b7f0e3a8a0604c6a838626e24a30ec961enriched.xml\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/f735949b69fc3a67804946cd.xml\"},{\"id\":99788708,\"identity\":\"6d60b237-6e58-43f3-be85-067ca56aa463\",\"added_by\":\"auto\",\"created_at\":\"2026-01-08 12:47:35\",\"extension\":\"jpeg\",\"order_by\":4,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"acdc-reference\",\"size\":489552,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"floatimage1.jpeg\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/2b397af0b619acd1d17453bb.jpeg\"},{\"id\":99788654,\"identity\":\"77f25ac3-047a-485f-851b-1cca9f822df3\",\"added_by\":\"auto\",\"created_at\":\"2026-01-08 12:47:26\",\"extension\":\"jpeg\",\"order_by\":5,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"acdc-reference\",\"size\":1340194,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"floatimage10.jpeg\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/70237210305aba1ac068adc0.jpeg\"},{\"id\":99343763,\"identity\":\"a20e8b31-70e7-4710-b067-c1e074f2bbf3\",\"added_by\":\"auto\",\"created_at\":\"2026-01-01 08:01:13\",\"extension\":\"jpeg\",\"order_by\":6,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"acdc-reference\",\"size\":227966,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"floatimage11.jpeg\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/3787ff31a8f436727731909f.jpeg\"},{\"id\":99789115,\"identity\":\"36604fc8-262e-4909-aae4-07d29e48b840\",\"added_by\":\"auto\",\"created_at\":\"2026-01-08 12:48:48\",\"extension\":\"jpeg\",\"order_by\":7,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"acdc-reference\",\"size\":2565236,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"floatimage12.jpeg\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/e60e80140f06716a1ca77d91.jpeg\"},{\"id\":99788387,\"identity\":\"58b989b8-5c40-4c79-89f5-92e755e5536c\",\"added_by\":\"auto\",\"created_at\":\"2026-01-08 12:46:33\",\"extension\":\"jpeg\",\"order_by\":8,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"acdc-reference\",\"size\":4368076,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"floatimage2.jpeg\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/115618f7d28269a4f0817b4d.jpeg\"},{\"id\":99343768,\"identity\":\"91bbf453-711d-4288-89ab-f1a10e698a4f\",\"added_by\":\"auto\",\"created_at\":\"2026-01-01 08:01:14\",\"extension\":\"jpeg\",\"order_by\":9,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"acdc-reference\",\"size\":3065984,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"floatimage3.jpeg\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/045c682cb4bc8a7a9d2500b6.jpeg\"},{\"id\":99788368,\"identity\":\"f7d22912-81c7-4e36-bb51-fab41320ad7f\",\"added_by\":\"auto\",\"created_at\":\"2026-01-08 12:46:31\",\"extension\":\"jpeg\",\"order_by\":10,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"acdc-reference\",\"size\":2412472,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"floatimage4.jpeg\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/7eeab4d6ca385837e546b9b0.jpeg\"},{\"id\":99343770,\"identity\":\"886a77ae-95a0-4382-9590-b9e10c3f187c\",\"added_by\":\"auto\",\"created_at\":\"2026-01-01 08:01:14\",\"extension\":\"jpeg\",\"order_by\":11,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"acdc-reference\",\"size\":1039040,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"floatimage5.jpeg\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/fdf4984b4efaba39fd677034.jpeg\"},{\"id\":99788426,\"identity\":\"6c45ec31-0b41-4924-bd57-996d572f6fb0\",\"added_by\":\"auto\",\"created_at\":\"2026-01-08 12:46:40\",\"extension\":\"jpeg\",\"order_by\":12,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"acdc-reference\",\"size\":1960680,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"floatimage6.jpeg\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/0b67a2cf7f8dfdef3b2179b4.jpeg\"},{\"id\":99788810,\"identity\":\"16a68dab-ffb7-44e1-a1df-a4c66d1efb50\",\"added_by\":\"auto\",\"created_at\":\"2026-01-08 12:47:56\",\"extension\":\"jpeg\",\"order_by\":13,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"acdc-reference\",\"size\":2575820,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"floatimage7.jpeg\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/03a5bd2b1d008dc264054c47.jpeg\"},{\"id\":99343775,\"identity\":\"71dc4ef7-88d5-41fb-8319-e0962bdc60cb\",\"added_by\":\"auto\",\"created_at\":\"2026-01-01 08:01:14\",\"extension\":\"jpeg\",\"order_by\":14,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"acdc-reference\",\"size\":2325264,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"floatimage8.jpeg\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/8b6d08b41c8eae0b0a14d057.jpeg\"},{\"id\":99343795,\"identity\":\"ce899f11-d0fb-4659-9019-f31b4c819c81\",\"added_by\":\"auto\",\"created_at\":\"2026-01-01 08:01:14\",\"extension\":\"jpeg\",\"order_by\":15,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"acdc-reference\",\"size\":2345964,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"floatimage9.jpeg\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/89d46361fa12a739fcdbd732.jpeg\"},{\"id\":99788551,\"identity\":\"580571f3-f853-4656-8110-94a71fe09950\",\"added_by\":\"auto\",\"created_at\":\"2026-01-08 12:47:05\",\"extension\":\"png\",\"order_by\":16,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"acdc-reference\",\"size\":36534,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"Onlinefloatimage1.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/74ef2a5be3d3cff3d9d213c0.png\"},{\"id\":99788973,\"identity\":\"b43493bf-0ad6-435f-b332-4747654ab815\",\"added_by\":\"auto\",\"created_at\":\"2026-01-08 12:48:26\",\"extension\":\"png\",\"order_by\":17,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"acdc-reference\",\"size\":159299,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"Onlinefloatimage10.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/3aa028fddbfda6b353a6cd69.png\"},{\"id\":99788505,\"identity\":\"f815cbfc-3dcc-411f-8d8a-6fded856c340\",\"added_by\":\"auto\",\"created_at\":\"2026-01-08 12:46:56\",\"extension\":\"png\",\"order_by\":18,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"acdc-reference\",\"size\":33976,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"Onlinefloatimage11.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/26ea927260402f106cc8d789.png\"},{\"id\":99788950,\"identity\":\"d132cb81-da4b-4389-90e7-d8550b132b30\",\"added_by\":\"auto\",\"created_at\":\"2026-01-08 12:48:22\",\"extension\":\"png\",\"order_by\":19,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"acdc-reference\",\"size\":151708,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"Onlinefloatimage12.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/3092a00d4debc7157d393226.png\"},{\"id\":99343780,\"identity\":\"79d4f9b6-34c7-4c43-bdb0-50ef61915c8c\",\"added_by\":\"auto\",\"created_at\":\"2026-01-01 08:01:14\",\"extension\":\"png\",\"order_by\":20,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"acdc-reference\",\"size\":197555,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"Onlinefloatimage2.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/211ad4f5a03b82c8ae7fc82d.png\"},{\"id\":99788356,\"identity\":\"3d61b5d1-2120-4320-ac81-0459523121e9\",\"added_by\":\"auto\",\"created_at\":\"2026-01-08 12:46:29\",\"extension\":\"png\",\"order_by\":21,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"acdc-reference\",\"size\":247191,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"Onlinefloatimage3.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/dc022e260452603ec2921876.png\"},{\"id\":99343784,\"identity\":\"7ec2ec4f-416a-4ba2-bdbf-119474d46012\",\"added_by\":\"auto\",\"created_at\":\"2026-01-01 08:01:14\",\"extension\":\"png\",\"order_by\":22,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"acdc-reference\",\"size\":136393,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"Onlinefloatimage4.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/1a663bcdea427ee5b6b123f4.png\"},{\"id\":99343785,\"identity\":\"b0ff9eee-f657-4c59-89a9-c6cae9418a51\",\"added_by\":\"auto\",\"created_at\":\"2026-01-01 08:01:14\",\"extension\":\"png\",\"order_by\":23,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"acdc-reference\",\"size\":96210,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"Onlinefloatimage5.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/9a8da28e496c17322e2236a8.png\"},{\"id\":99343788,\"identity\":\"22ff9b1b-a5f7-48b8-9855-fcb3e3b013db\",\"added_by\":\"auto\",\"created_at\":\"2026-01-01 08:01:14\",\"extension\":\"png\",\"order_by\":24,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"acdc-reference\",\"size\":168607,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"Onlinefloatimage6.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/934eeac107b159788c1334ba.png\"},{\"id\":99343789,\"identity\":\"c1befc07-6613-4920-b4aa-b81b3971d1e7\",\"added_by\":\"auto\",\"created_at\":\"2026-01-01 08:01:14\",\"extension\":\"png\",\"order_by\":25,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"acdc-reference\",\"size\":225748,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"Onlinefloatimage7.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/ef31104ff2178a541f86cc03.png\"},{\"id\":99788679,\"identity\":\"83eaad82-b4b4-4dad-8228-0d77c052b168\",\"added_by\":\"auto\",\"created_at\":\"2026-01-08 12:47:32\",\"extension\":\"png\",\"order_by\":26,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"acdc-reference\",\"size\":166134,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"Onlinefloatimage8.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/bf16eed662309a3a05324ef6.png\"},{\"id\":99343793,\"identity\":\"86559f26-1eb2-4285-be81-9a564e860682\",\"added_by\":\"auto\",\"created_at\":\"2026-01-01 08:01:14\",\"extension\":\"png\",\"order_by\":27,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"acdc-reference\",\"size\":116710,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"Onlinefloatimage9.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/074f2c216815854ad0353be5.png\"},{\"id\":99343781,\"identity\":\"9939bdbc-60a7-4f37-8283-051b722924eb\",\"added_by\":\"auto\",\"created_at\":\"2026-01-01 08:01:14\",\"extension\":\"xml\",\"order_by\":28,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"acdc-reference\",\"size\":110521,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"b7f0e3a8a0604c6a838626e24a30ec961structuring.xml\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/07faaf2d1716f1943c0e03ba.xml\"},{\"id\":99343786,\"identity\":\"2fb0dc64-8941-40ce-8f5f-b053652928d8\",\"added_by\":\"auto\",\"created_at\":\"2026-01-01 08:01:14\",\"extension\":\"html\",\"order_by\":29,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"acdc-reference\",\"size\":127560,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"earlyproof.html\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/ec122bd70c54121fb67356da.html\"},{\"id\":99343756,\"identity\":\"fe88d446-5190-427a-a539-5cce9bc08411\",\"added_by\":\"auto\",\"created_at\":\"2026-01-01 08:01:13\",\"extension\":\"png\",\"order_by\":1,\"title\":\"Figure 1\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":365797,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003eStudy workflow for integrative discovery and validation.\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eOverview of the multi‑omics and validation pipeline. Steps include: dataset assembly (bulk transcriptomics, scRNA‑seq, proteomics), preprocessing and batch correction, DEG and WGCNA discovery, machine‑learning screening across 12 algorithms/113 combinations, single‑cell localization, immune deconvolution, experimental validation in CSE‑treated cells and cigarette smoke–exposed mice, proteomics confirmation, and molecular docking against CYP1B1. Arrows indicate analysis flow; boxes list key outputs and Figures referenced in the Results.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"floatimage1.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/8ac72e0e3f7539a019eedb58.png\"},{\"id\":99343755,\"identity\":\"50cfe455-9b7f-412b-b6c7-b903793cd6ec\",\"added_by\":\"auto\",\"created_at\":\"2026-01-01 08:01:13\",\"extension\":\"png\",\"order_by\":2,\"title\":\"Figure 2\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":1558211,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003eWGCNA module associated with COPD and differential‑expression filtering.\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003e(A) Scale‑free topology analysis used to select soft threshold β=5. (B) Dendrogram with dynamic tree cutting identifying 20 modules. (C) Module–trait heat map showing correlations with COPD. The midnightblue module shows the strongest association. (D) Gene significance versus module membership in the midnight blue module. (E) Volcano plot of DEGs (thresholds: |log2FC|\\u0026gt;0.585). (F) Heat map of top 50 DEGs (z‑score normalized). (G) Venn/UpSet plot intersecting DEGs with midnightblue module genes to yield 24 hub genes.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"floatimage2.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/532cd73c569d6864e06e5eec.png\"},{\"id\":99343765,\"identity\":\"f3d2bd31-1866-43b3-b53e-d03ad97312ef\",\"added_by\":\"auto\",\"created_at\":\"2026-01-01 08:01:13\",\"extension\":\"png\",\"order_by\":3,\"title\":\"Figure 3\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":1974695,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003eMachine‑learning feature prioritization and diagnostic performance.\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003e(A) Heat map of feature selection frequency/importance across 12 algorithms and 113 model combinations. (B) ROC curve in training (AUC with 95% CI). (C) ROC in external validation cohort GSE76925 (AUC with 95% CI). (D–E) Confusion matrices for training and validation sets. (F) Volcano plots highlighting BHLHE22, CYP1B1, DHRS9, and DPP6 expression differences. (G) Inter‑gene correlation matrix.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"floatimage3.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/0044349457978c4fe99a9b72.png\"},{\"id\":99789024,\"identity\":\"2a5b7308-bc13-4b0b-bc55-6c5ee2ed7420\",\"added_by\":\"auto\",\"created_at\":\"2026-01-08 12:48:32\",\"extension\":\"png\",\"order_by\":4,\"title\":\"Figure 4\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":1267151,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003eExternal Validation and Functional Enrichment for Prioritized Biomarkers.\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003e(A) Per‑gene ROC curves in GSE37768 with AUC (95% CI) for BHLHE22, DPP6, DHRS9, and CYP1B1; include DeLong P values vs 0.5. (B) Nomogram integrating top predictors; points, linear predictor, and predicted probability scales are displayed. (C) Calibration curve for nomogram. (D–E) GSEA results for BHLHE22‑high group of GO/KEGG. (F–G) GSEA results for the CYP1B1‑high group.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"floatimage4.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/f116dc33d73c38afcee66759.png\"},{\"id\":99343760,\"identity\":\"1197caee-5530-4239-a4a9-b1ddda163acc\",\"added_by\":\"auto\",\"created_at\":\"2026-01-01 08:01:13\",\"extension\":\"png\",\"order_by\":5,\"title\":\"Figure 5\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":751826,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003eANN architecture, calibration, and external discrimination.\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003e(A) ANN topology (input: 3 genes; hidden layer with 5 neurons; sigmoid activation); edge weights visualized. (B) Calibration curves (Brier score) for training and external validation. (C) ROC and PR curves with 95% CIs. (D) Decision‑curve analysis demonstrating net benefit across clinically relevant thresholds.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"floatimage5.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/30169d6038d93342995184c6.png\"},{\"id\":99788290,\"identity\":\"efb284f3-7ea1-4187-a2b8-ace881f1d505\",\"added_by\":\"auto\",\"created_at\":\"2026-01-08 12:46:01\",\"extension\":\"png\",\"order_by\":6,\"title\":\"Figure 6\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":1297410,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003eSingle‑cell mapping of biomarker expression and ferroptosis signatures.\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003e(A) Marker‑based annotation of 16 cell types; top markers per cluster listed. (B) UMAP of integrated donors showing major epithelial and immune populations. (C) Bar plots of cell‑type proportions in control vs COPD donors. (D) Feature plots/violin plots showing higher CYP1B1 in airway secretory cells (ASCs) in COPD. (E) KEGG enrichment for CYP1B1‑high vs low ASCs showing ferroptosis pathway. (F) AUCell/GSVA ferroptosis scores (adjusted P\\u0026lt;0.05).\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"floatimage6.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/44396f7a891bf8946a8a50ab.png\"},{\"id\":99343767,\"identity\":\"7610a0c8-277d-4fd2-a78c-e780c8175964\",\"added_by\":\"auto\",\"created_at\":\"2026-01-01 08:01:13\",\"extension\":\"png\",\"order_by\":7,\"title\":\"Figure 7\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":1029246,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003ePseudotime trajectories and cell–cell communication networks.\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003e(A) Monocle pseudotime ordering of epithelial lineages with bifurcation; ASCs are enriched in a terminal branch. (B) Same trajectory colored by annotated cell types. (C–D) CellChat global signaling networks comparing control vs COPD; edge width reflects communication strength. (E) ASC‑centric subnetworks highlighting increased outgoing/incoming interactions with alveolar macrophages and immune cells.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"floatimage7.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/ceb6ea38e04b070901bfd161.png\"},{\"id\":99343774,\"identity\":\"ab523d48-5fa9-41df-93b3-edd3078c79d4\",\"added_by\":\"auto\",\"created_at\":\"2026-01-01 08:01:14\",\"extension\":\"png\",\"order_by\":8,\"title\":\"Figure 8\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":1476231,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003eImmune infiltration landscape and correlations with hub genes.\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003e(A) CIBERSORT LM22 deconvolution across cohorts showing differences in immune subset proportions between COPD and controls. (B) Correlation heat map among immune subsets. (C–E) Associations of BHLHE22, CYP1B1, and DPP6 expression with selected immune subsets.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"floatimage8.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/866880fe12decf56db718610.png\"},{\"id\":99788688,\"identity\":\"64809408-57bf-4a84-8ae2-772f5c07d47d\",\"added_by\":\"auto\",\"created_at\":\"2026-01-08 12:47:32\",\"extension\":\"png\",\"order_by\":9,\"title\":\"Figure 9\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":998757,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003eSmoke exposure increases CYP1B1 expression.\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003e(A–H) BALF/serum cytokines (IL‑1β, IL‑6, IL‑8, TNF‑α) across exposure time points (mean±SD; n per group; statistical tests with exact P and multiple‑testing correction). (I) Lung CYP1B1 expression (IHC/Western/qPCR as applicable; effect size and CI). (J) Proteomics confirmation of CYP1B1 upregulation in CSE‑treated 16HBE cells (FDR\\u0026gt;0.5).\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"floatimage9.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/46a0759bdf5be2a0ef89b710.png\"},{\"id\":99343772,\"identity\":\"7c2356a4-4e02-43ac-8672-3fdc07c4b6c8\",\"added_by\":\"auto\",\"created_at\":\"2026-01-01 08:01:14\",\"extension\":\"png\",\"order_by\":10,\"title\":\"Figure 10\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":1035761,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003eSmoke exposure causes airway dysfunction in the lungs.\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003e(A-C) Representative histology (H\\u0026amp;E, PAS, Picrosirius Red) with quantification. (D-K) Lung function parameters. (L-N) Altered levels of ferroptosis-related indicators (GSH/GSSG, MDA, GPX4 and SOD1).\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"floatimage10.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/33f27ec7540c405d778e7079.png\"},{\"id\":99788237,\"identity\":\"082ae174-449a-4cb0-8be7-c65d5a3b447f\",\"added_by\":\"auto\",\"created_at\":\"2026-01-08 12:45:47\",\"extension\":\"png\",\"order_by\":11,\"title\":\"Figure 11\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":239645,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003eKnockdown of CYP1B1 expression inhibits ferroptosis signaling.\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003e(A-C) Expression of CYP1B1 and altered levels of ferroptosis-related indicators (GSH/GSSG, MDA, GPX4 and 4-HNE).\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"floatimage11.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/cd048e765d31e2254cf97a5a.png\"},{\"id\":99789073,\"identity\":\"d132db3b-aed4-4b34-86fc-cab18d951791\",\"added_by\":\"auto\",\"created_at\":\"2026-01-08 12:48:40\",\"extension\":\"png\",\"order_by\":12,\"title\":\"Figure 12\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":671155,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003eMolecular docking identifies natural compounds that target the CYP1B1 enzyme.\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003e(A–D) Predicted binding poses and scores (kcal/mol) for α‑naphthoflavone, β‑naphthoflavone, chrysin, and naringenin within the CYP1B1 active site. Annotate key interactions (H‑bonds, π–π, hydrophobic contacts).\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"floatimage12.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/bf03beb2e5a347b10dfc3879.png\"},{\"id\":104835119,\"identity\":\"87a836db-8e29-42fe-88ce-0d24e4e8cb9b\",\"added_by\":\"auto\",\"created_at\":\"2026-03-17 17:40:28\",\"extension\":\"pdf\",\"order_by\":0,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"manuscript-pdf\",\"size\":14171788,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"manuscript.pdf\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/17270815-bddf-4096-8b84-79ca9a464000.pdf\"},{\"id\":99343797,\"identity\":\"7e6d356c-80a2-468f-a199-62ec2888a68f\",\"added_by\":\"auto\",\"created_at\":\"2026-01-01 08:01:16\",\"extension\":\"zip\",\"order_by\":0,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"supplement\",\"size\":93492254,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"WB.zip\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-8049999/v1/e66fe1bd3f79487e74f9ea0d.zip\"}],\"financialInterests\":\"No competing interests reported.\",\"formattedTitle\":\"\\u003cp\\u003eCYP1B1-Mediated Ferroptosis Defines a Biomarker and Therapeutic Target in COPD Across Multi-omics and Single-Cell \\u003c/p\\u003e\",\"fulltext\":[{\"header\":\"Highlights\",\"content\":\"\\u003cp\\u003e1. Multi-omics integration identifies CYP1B1 as a potential COPD diagnostic biomarker, with external validation.\\u003c/p\\u003e\\u003cp\\u003e2. CYP1B1 is upregulated in airway secretory cells and is linked to the activation of ferroptosis.\\u003c/p\\u003e\\u003cp\\u003e3. ML and ANN demonstrate strong predictive performance (AUC up to 0.996 in training; 0.834 in external validation).\\u003c/p\\u003e\\u003cp\\u003e4. Mouse and cell models demonstrate increased CYP1B1 expression in the context of inflammation and fibrosis; proteomics corroborate these findings.\\u003c/p\\u003e\\u003cp\\u003e5. Docking highlights α/β-naphthoflavone, chrysin, and naringenin as CYP1B1 binders (\\u0026thinsp;\\u0026asymp;\\u0026thinsp;\\u0026minus;\\u0026thinsp;7.11 to \\u0026minus;\\u0026thinsp;5.45 kcal/mol).\\u003c/p\\u003e\"},{\"header\":\"Introduction\",\"content\":\"\\u003cp\\u003eChronic obstructive pulmonary disease (COPD) remains a leading global cause of morbidity and mortality [\\u003cspan citationid=\\\"CR1\\\" class=\\\"CitationRef\\\"\\u003e1\\u003c/span\\u003e]. Clinically, COPD is defined by persistent respiratory symptoms and progressive airflow limitation, quantified by spirometry; however, current diagnostic criteria incompletely capture the disease heterogeneity, which spans complex structural remodeling of the lung, diverse inflammatory endotypes, and shifts in the airway microbiome [\\u003cspan citationid=\\\"CR2\\\" class=\\\"CitationRef\\\"\\u003e2\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR3\\\" class=\\\"CitationRef\\\"\\u003e3\\u003c/span\\u003e]. Against this backdrop, there is a sustained need for mechanism-anchored biomarkers and actionable targets. Increasing evidence implicates oxidative stress and lipid peroxidation as central features of chronic airway injury, with ferroptosis\\u0026mdash;an iron-dependent, lipid peroxidation-driven cell-death program\\u0026mdash;emerging as a plausible contributor to epithelial dysfunction and tissue remodeling in COPD.\\u003c/p\\u003e \\u003cp\\u003eAirflow obstruction in COPD typically progresses over several years and reflects a fixed pathobiology, including fibrotic narrowing of the small airways and loss of elastic recoil, leading to peripheral airway collapse and air trapping [\\u003cspan citationid=\\\"CR4\\\" class=\\\"CitationRef\\\"\\u003e4\\u003c/span\\u003e]. The small airway epithelium is the principal locus of cellular and histologic change [\\u003cspan citationid=\\\"CR5\\\" class=\\\"CitationRef\\\"\\u003e5\\u003c/span\\u003e], comprising a pseudostratified layer of secretory (club) cells, ciliated cells, and basal cells [\\u003cspan citationid=\\\"CR6\\\" class=\\\"CitationRef\\\"\\u003e6\\u003c/span\\u003e]. Inflammatory remodeling features increased neutrophils and macrophages in the airway lumina, as well as the accumulation of macrophages, T lymphocytes, and B lymphocytes in the airway wall and parenchyma [\\u003cspan citationid=\\\"CR7\\\" class=\\\"CitationRef\\\"\\u003e7\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR8\\\" class=\\\"CitationRef\\\"\\u003e8\\u003c/span\\u003e]. These observations place epithelial injury and immune\\u0026ndash;epithelial crosstalk at the center of COPD progression, motivating an interrogation of epithelial cell-intrinsic death pathways that may amplify chronic inflammation.\\u003c/p\\u003e \\u003cp\\u003eGenomic and transcriptomic studies suggest that interindividual susceptibility to COPD is related to gene expression programs within the airway epithelium [\\u003cspan citationid=\\\"CR9\\\" class=\\\"CitationRef\\\"\\u003e9\\u003c/span\\u003e], yet the upstream regulators that couple xenobiotic/redox metabolism to ferroptosis remain poorly understood. CYP1B1, a cytochrome P450 implicated in xenobiotic biotransformation and oxidative metabolism, is a biologically plausible node at the interface of redox stress and ferroptosis in airway secretory cells, where lipid peroxidation may promote persistent epithelial dysfunction. However, causal links between CYP1B1 activity, ferroptosis signaling, and COPD pathogenesis are not established, underscoring the need for systems-level discovery integrated with cellular and in vivo validation.\\u003c/p\\u003e \\u003cp\\u003eTo address this gap, we employed an integrative multi-omics strategy combining bulk transcriptomics with weighted gene co-expression network analysis (WGCNA), differential expression, and pathway enrichment (GO/KEGG/GSEA), followed by a machine-learning framework (12 algorithms across 113 combinations) and an artificial neural network (ANN) to derive and test diagnostic models (ROC/AUC). Immune deconvolution (CIBERSORT) characterized the inflammatory context, while single-cell RNA sequencing mapped candidate gene expression to specific lung cell types. Experimental validation in cigarette\\u0026ndash;smoke\\u0026ndash;exposed models and orthogonal proteomics supported the biological plausibility, and molecular docking explored the therapeutic tractability. We hypothesized that CYP1B1 marks\\u0026mdash;and may contribute to\\u0026mdash;ferroptosis-linked epithelial pathology in COPD, would generalize across multi-omics layers, demonstrate diagnostic utility, and reveal druggable chemistry space. We aimed to identify robust COPD biomarkers/pathways, localize them to disease-relevant cell types, link them to ferroptosis signatures, validate them in models, and nominate candidate CYP1B1-targeting ligands for future experimental testing.\\u003c/p\\u003e\"},{\"header\":\"Materials and Methods\",\"content\":\"\\u003cdiv id=\\\"Sec3\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eData sources and preprocessing (bulk expression, single-cell, proteomics)\\u003c/h2\\u003e \\u003cp\\u003eWe integrated bulk gene-expression datasets, single-cell RNA sequencing (scRNA-seq) data, and discovery proteomics. Bulk datasets were retrieved from GEO and included GSE47460 (platforms GPL6480 and GPL14550; healthy controls \\u003cem\\u003en\\u003c/em\\u003e\\u0026thinsp;=\\u0026thinsp;17 and COPD \\u003cem\\u003en\\u003c/em\\u003e\\u0026thinsp;=\\u0026thinsp;75 for GPL6480; healthy controls \\u003cem\\u003en\\u003c/em\\u003e\\u0026thinsp;=\\u0026thinsp;91 and COPD \\u003cem\\u003en\\u003c/em\\u003e\\u0026thinsp;=\\u0026thinsp;145 for GPL14550) [\\u003cspan citationid=\\\"CR10\\\" class=\\\"CitationRef\\\"\\u003e10\\u003c/span\\u003e], with GSE76925 (COPD \\u003cem\\u003en\\u003c/em\\u003e\\u0026thinsp;=\\u0026thinsp;111, controls \\u003cem\\u003en\\u003c/em\\u003e\\u0026thinsp;=\\u0026thinsp;40) [\\u003cspan citationid=\\\"CR11\\\" class=\\\"CitationRef\\\"\\u003e11\\u003c/span\\u003e] and GSE37768 (COPD \\u003cem\\u003en\\u003c/em\\u003e\\u0026thinsp;=\\u0026thinsp;18, controls \\u003cem\\u003en\\u003c/em\\u003e\\u0026thinsp;=\\u0026thinsp;9) serving as independent validation cohorts. Probe-level preprocessing removed empty probes and those mapping to multiple genes; for genes with multiple probes, the probe with the highest mean expression was retained, and when multiple probes per gene were required, median aggregation was used. Datasets were merged after re-annotation to gene symbols using the official platform annotation files and were batch-corrected and normalized using ComBat (R package sva) [\\u003cspan citationid=\\\"CR12\\\" class=\\\"CitationRef\\\"\\u003e12\\u003c/span\\u003e]. Subsequent analyses used these normalized matrices. scRNA-seq data were obtained from GSE196341 and GSE135893 (12 COPD and 12 control donors in total) [\\u003cspan citationid=\\\"CR13\\\" class=\\\"CitationRef\\\"\\u003e13\\u003c/span\\u003e]. Discovery proteomics were deposited to ProteomeXchange (iProX partner) under PXD068247 [\\u003cspan citationid=\\\"CR14\\\" class=\\\"CitationRef\\\"\\u003e14\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR15\\\" class=\\\"CitationRef\\\"\\u003e15\\u003c/span\\u003e]. A summary of datasets is provided in Table\\u0026nbsp;\\u003cspan refid=\\\"Tab1\\\" class=\\\"InternalRef\\\"\\u003e1\\u003c/span\\u003e.\\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\\u003eDatasets included in the study\\u003c/p\\u003e \\u003c/div\\u003e \\u003c/caption\\u003e \\u003ccolgroup cols=\\\"5\\\"\\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=\\\"left\\\" class=\\\"colspec\\\" colname=\\\"c3\\\" colnum=\\\"3\\\"\\u003e\\u003c/div\\u003e \\u003cdiv align=\\\"left\\\" class=\\\"colspec\\\" colname=\\\"c4\\\" colnum=\\\"4\\\"\\u003e\\u003c/div\\u003e \\u003cdiv align=\\\"left\\\" class=\\\"colspec\\\" colname=\\\"c5\\\" colnum=\\\"5\\\"\\u003e\\u003c/div\\u003e \\u003cthead\\u003e \\u003ctr\\u003e \\u003cth align=\\\"left\\\" colname=\\\"c1\\\"\\u003e \\u003cp\\u003eDatasets ID\\u003c/p\\u003e \\u003c/th\\u003e \\u003cth align=\\\"left\\\" colname=\\\"c2\\\"\\u003e \\u003cp\\u003eYear\\u003c/p\\u003e \\u003c/th\\u003e \\u003cth align=\\\"left\\\" colname=\\\"c3\\\"\\u003e \\u003cp\\u003eDate type\\u003c/p\\u003e \\u003c/th\\u003e \\u003cth align=\\\"left\\\" colname=\\\"c4\\\"\\u003e \\u003cp\\u003eCountry\\u003c/p\\u003e \\u003c/th\\u003e \\u003cth align=\\\"left\\\" colname=\\\"c5\\\"\\u003e \\u003cp\\u003eN (samples)\\u003c/p\\u003e \\u003c/th\\u003e \\u003c/tr\\u003e \\u003c/thead\\u003e \\u003ctbody\\u003e \\u003ctr\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e \\u003cp\\u003eGSE47460\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c2\\\"\\u003e \\u003cp\\u003e2013\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e \\u003cp\\u003eBulk RNA-Seq\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e \\u003cp\\u003eAmerican\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e \\u003cp\\u003eCase\\u0026thinsp;=\\u0026thinsp;220, Control\\u0026thinsp;=\\u0026thinsp;108\\u003c/p\\u003e \\u003c/td\\u003e \\u003c/tr\\u003e \\u003ctr\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e \\u003cp\\u003eGSE76925\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c2\\\"\\u003e \\u003cp\\u003e2017\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e \\u003cp\\u003eBulk RNA-Seq\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e \\u003cp\\u003eAmerican\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e \\u003cp\\u003eCase\\u0026thinsp;=\\u0026thinsp;111, Control\\u0026thinsp;=\\u0026thinsp;40\\u003c/p\\u003e \\u003c/td\\u003e \\u003c/tr\\u003e \\u003ctr\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e \\u003cp\\u003eGSE37768\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c2\\\"\\u003e \\u003cp\\u003e2016\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e \\u003cp\\u003eBulk RNA-Seq\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e \\u003cp\\u003eSpain\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e \\u003cp\\u003eCase\\u0026thinsp;=\\u0026thinsp;18, Control\\u0026thinsp;=\\u0026thinsp;9\\u003c/p\\u003e \\u003c/td\\u003e \\u003c/tr\\u003e \\u003ctr\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e \\u003cp\\u003eGSE196341\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c2\\\"\\u003e \\u003cp\\u003e2022\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e \\u003cp\\u003escRNA-Seq\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e \\u003cp\\u003eAmerican\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e \\u003cp\\u003eCase\\u0026thinsp;=\\u0026thinsp;12\\u003c/p\\u003e \\u003c/td\\u003e \\u003c/tr\\u003e \\u003ctr\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e \\u003cp\\u003eGSE135893\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c2\\\"\\u003e \\u003cp\\u003e2019\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e \\u003cp\\u003escRNA-Seq\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e \\u003cp\\u003eAmerican\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e \\u003cp\\u003eControl\\u0026thinsp;=\\u0026thinsp;12\\u003c/p\\u003e \\u003c/td\\u003e \\u003c/tr\\u003e \\u003ctr\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e \\u003cp\\u003ePXD068247\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c2\\\"\\u003e \\u003cp\\u003e2025\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c3\\\"\\u003e \\u003cp\\u003eProteomics\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c4\\\"\\u003e \\u003cp\\u003eChina\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c5\\\"\\u003e \\u003cp\\u003eCase\\u0026thinsp;=\\u0026thinsp;3, Control\\u0026thinsp;=\\u0026thinsp;3\\u003c/p\\u003e \\u003c/td\\u003e \\u003c/tr\\u003e \\u003c/tbody\\u003e \\u003c/colgroup\\u003e \\u003c/table\\u003e\\u003c/div\\u003e \\u003c/p\\u003e \\u003c/div\\u003e\\n\\u003ch3\\u003eDifferentially expressed genes (DEGs)\\u003c/h3\\u003e\\n\\u003cp\\u003eDifferential expression between COPD and control groups was assessed using limma, with model terms including group and batch/platform, where applicable. Unless otherwise indicated, significance is used |log2FC| \\u0026gt; 1.5 and p\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.01 after multiple-testing correction. Results were visualized with volcano plots and clustered heatmaps (ggplot2, pheatmap) [\\u003cspan additionalcitationids=\\\"CR17\\\" citationid=\\\"CR16\\\" class=\\\"CitationRef\\\"\\u003e16\\u003c/span\\u003e\\u0026ndash;\\u003cspan citationid=\\\"CR18\\\" class=\\\"CitationRef\\\"\\u003e18\\u003c/span\\u003e].\\u003c/p\\u003e\\n\\u003ch3\\u003eWeighted gene co-expression network analysis (WGCNA)\\u003c/h3\\u003e\\n\\u003cp\\u003eCo-expression networks were constructed using WGCNA. Signed networks were generated from pairwise Pearson correlations, selecting the soft-thresholding power β to approximate scale-free topology (reporting scale-free fit R\\u003csup\\u003e2\\u003c/sup\\u003e and mean connectivity). Adjacency matrices were transformed to topological overlap matrices (TOM), and hierarchical clustering with dynamic tree cutting identified gene modules (report minimum module size and merge thresholds). Module eigengenes were correlated with COPD status, and the midnight blue module, showing the strongest association with COPD, was selected for downstream analysis [\\u003cspan citationid=\\\"CR19\\\" class=\\\"CitationRef\\\"\\u003e19\\u003c/span\\u003e].\\u003c/p\\u003e\\n\\u003ch3\\u003eMachine-learning feature selection and classification\\u003c/h3\\u003e\\n\\u003cp\\u003eBatch-corrected, z-score\\u0026ndash;standardized expression matrices were used to derive features and train classifiers. We implemented a multi-algorithm screen comprising 12 algorithms across 113 combinations (e.g., Logistic Regression, LASSO/Elastic Net, Ridge, SVM-Linear/RBF, Random Forest, ExtraTrees, GBM, XGBoost, k-NN, Na\\u0026iuml;ve Bayes), with feature selection embedded within folds to prevent information leakage. Model selection employed stratified k-fold cross-validation on the training data, followed by external validation on independent cohorts (GSE76925, GSE37768). Performance metrics included ROC-AUC with 95% CIs, PR-AUC, accuracy, F1, and calibration (Brier score, calibration curves). Feature importance was summarized via permutation importance and SHAP, where applicable. The highest mean external AUC chose the optimal model.\\u003c/p\\u003e\\n\\u003ch3\\u003eGene set enrichment analysis (GSEA) and functional annotation\\u003c/h3\\u003e\\n\\u003cp\\u003eTo contextualize biological processes, we performed GSEA using MSigDB KEGG (c2.cp.kegg.Hs.symbols) and GO (c5.go.symbols) collections, and complementary over-representation analyses with clusterProfiler; visualizations employed enrichplot. Significance thresholds were based on field standards for GSEA and ORA, as specified in the Results [\\u003cspan citationid=\\\"CR20\\\" class=\\\"CitationRef\\\"\\u003e20\\u003c/span\\u003e].\\u003c/p\\u003e \\u003cdiv id=\\\"Sec8\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eSingle-cell RNA-seq analysis\\u003c/h2\\u003e \\u003cp\\u003escRNA-seq data (GSE196341, GSE135893) were processed in Seurat. Cells were filtered by mitochondrial RNA\\u0026thinsp;\\u0026gt;\\u0026thinsp;15%, ribosomal RNA\\u0026thinsp;\\u0026gt;\\u0026thinsp;5%, erythrocyte RNA\\u0026thinsp;\\u0026gt;\\u0026thinsp;1%, detected genes\\u0026thinsp;\\u0026lt;\\u0026thinsp;300 or \\u0026gt;\\u0026thinsp;7,000. Data were normalized (method as specified in the Results), and batch effects were mitigated via Harmony integration. Highly variable genes (n\\u0026thinsp;=\\u0026thinsp;2,000) were used for PCA and UMAP (dimensions\\u0026thinsp;=\\u0026thinsp;22). Clustering utilized a shared nearest neighbor graph (resolution 0.2) and was manually annotated using the top 10 marker genes per cluster. We visualized donor-wise UMAPs and cell-type distributions, and mapped CYP1B1 expression to airway secretory cells. Pseudotime analysis was performed using Monocle, and intercellular communication was inferred using CellChat [\\u003cspan citationid=\\\"CR13\\\" class=\\\"CitationRef\\\"\\u003e13\\u003c/span\\u003e].\\u003c/p\\u003e \\u003c/div\\u003e\\n\\u003ch3\\u003eArtificial neural network (ANN)\\u003c/h3\\u003e\\n\\u003cp\\u003eA feed-forward ANN was trained to classify Control vs COPD using binarized signatures derived from characteristic genes (upregulated genes coded 1 if\\u0026thinsp;\\u0026ge;\\u0026thinsp;median; downregulated genes reverse-coded). The network consisted of an input layer (size equal to the number of features), a single hidden layer (5 neurons, using the sigmoid activation function), and a binary output. Training was performed using a neural network with Rprop\\u0026thinsp;+\\u0026thinsp;optimization (seed\\u0026thinsp;=\\u0026thinsp;12345678). Performance was evaluated via confusion matrices, accuracy per class, and ROC-AUC; feature contributions were estimated using the Garson algorithm.\\u003c/p\\u003e\\n\\u003ch3\\u003eImmune cell deconvolution\\u003c/h3\\u003e\\n\\u003cp\\u003eFrom normalized bulk expression, immune cell proportions were estimated using CIBERSORT (LM22), with default permutations (state number) and quantile normalization settings appropriate for the tissue type. Samples with CIBERSORT p\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.05 were retained. Group differences were tested using the Wilcoxon rank-sum test with the Benjamini\\u0026ndash;Hochberg correction across 22 cell types. Spearman correlations between characteristic genes and immune subsets were computed with multiple-testing correction.\\u003c/p\\u003e \\u003cdiv id=\\\"Sec11\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eCigarette smoke extract (CSE) preparation\\u003c/h2\\u003e \\u003cp\\u003eMainstream smoke from Hongta brand filter-tipped cigarettes (Yunnan, China) was drawn by a peristaltic pump and bubbled through 10 mL DMEM per cigarette. The pH was adjusted to 7.2, and the solution was sterilized using a 0.22-\\u0026micro;m filter to yield 10% CSE (v/v); working concentrations were prepared freshly for each experiment, for particulate matter (PM2.5), stock suspensions were prepared in DMSO at 400 mg/mL, vortexed and sonicated for 2 h, centrifuged to remove insoluble particulates, and supernatants stored at \\u0026minus;\\u0026thinsp;20\\u0026deg;C\\u003csup\\u003e[21]\\u003c/sup\\u003e.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec12\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eCell culture and treatments\\u003c/h2\\u003e \\u003cp\\u003e16HBE cells were cultured in DMEM\\u0026thinsp;+\\u0026thinsp;10% FBS at 37\\u0026deg;C, 5% CO₂. Cells were exposed to 2.5% CSE for 24 h. Unless otherwise stated, vehicle-treated cells served as controls\\u003csup\\u003e[21]\\u003c/sup\\u003e.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec13\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eMouse COPD model\\u003c/h2\\u003e \\u003cp\\u003eSpecific pathogen-free C57BL/6J mice (15\\u0026ndash;20 g) were housed at 25\\u0026deg;C, 12/12 h light\\u0026ndash;dark, with food and water ad libitum. The institutional ethics committee approved protocols. Mice were randomized into two groups: control (n\\u0026thinsp;=\\u0026thinsp;8) and COPD model (n\\u0026thinsp;=\\u0026thinsp;8). COPD was induced by whole-body exposure in a 60\\u0026times;57\\u0026times;100 cm chamber to 9 cigarettes twice daily, 6 days/week, for 6 months; each session lasted\\u0026thinsp;~\\u0026thinsp;2 h with \\u0026ge;\\u0026thinsp;4 h interval. Control mice were exposed to room air for the same duration. Before exposure, mice received sterile saline aerosol (30 minutes, twice daily). Tissues (lung), bronchoalveolar lavage fluid (BALF), and blood were collected at endpoint\\u003csup\\u003e[22]\\u003c/sup\\u003e.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec14\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eLung function testing\\u003c/h2\\u003e \\u003cp\\u003eLung mechanics were evaluated using a FinePointe system (Buxco) according to the manufacturer\\u0026rsquo;s instructions, recording airway resistance/compliance indices\\u003csup\\u003e[23]\\u003c/sup\\u003e.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec15\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eEnzyme-linked immunosorbent assays (ELISAs)\\u003c/h2\\u003e \\u003cp\\u003eConcentrations of IL-1β, IL-6, IL-8, and TNF-α in BALF and serum were quantified per the manufacturers\\u0026rsquo; protocols.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec16\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eHistology and staining\\u003c/h2\\u003e \\u003cp\\u003eThe lungs were fixed in 4% paraformaldehyde, processed, embedded in paraffin, sectioned, and stained with H\\u0026amp;E, PAS, and Picrosirius Red using standard protocols.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec17\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eImmunohistochemistry (IHC)\\u003c/h2\\u003e \\u003cp\\u003eIHC was performed on paraffin-embedded lung sections using standard antigen retrieval and detection procedures.\\u003c/p\\u003e \\u003c/div\\u003e\\u003cdiv id=\\\"Sec18\\\" class=\\\"Section2\\\"\\u003e\\n \\u003ch2\\u003eMolecular docking\\u003c/h2\\u003e\\n \\u003cp\\u003eTo explore therapeutic tractability, we performed molecular docking against CYP1B1. Three-dimensional structures for ligands were retrieved (PubChem) and prepared in AutoDockTools 1.5.7; receptor structure (CYP1B1) was obtained from RCSB PDB (\\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ewww.rcsb.org\\u003c/span\\u003e\\u003c/span\\u003e) and prepared analogously. The grid center and box size were defined around the active site, and docking was performed using the Lamarckian Genetic Algorithm under default parameters. Rigid docking was performed (AutoDockTools 1.6.7), and the top-scoring pose was retained.\\u003c/p\\u003e\\n\\u003c/div\\u003e\\n\\u003cdiv id=\\\"Sec19\\\" class=\\\"Section2\\\"\\u003e\\n \\u003ch2\\u003eStatistical analysis\\u003c/h2\\u003e\\n \\u003cp\\u003eAnalyses were performed in R 4.4.2. Unless otherwise stated, two-sided tests were used, with \\u0026alpha;\\u0026thinsp;=\\u0026thinsp;0.05. For multiple comparisons, the Benjamini\\u0026ndash;Hochberg correction was applied. Classification performance was summarized by ROC-AUC (95% CI), PR-AUC, accuracy, F1 score, confusion matrices, and calibration curves. Data are reported as mean\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;SD (or median [IQR]) with n indicating biological replicates.\\u003c/p\\u003e\\n\\u003c/div\\u003e\"},{\"header\":\"Results\",\"content\":\"\\u003cdiv id=\\\"Sec21\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eMulti-omics discovery and hub-gene prioritization\\u003c/h2\\u003e \\u003cp\\u003eA study overview is provided in Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig1\\\" class=\\\"InternalRef\\\"\\u003e1\\u003c/span\\u003e. Using WGCNA on the training cohort, we selected a soft-threshold power of β\\u0026thinsp;=\\u0026thinsp;5 (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig2\\\" class=\\\"InternalRef\\\"\\u003e2\\u003c/span\\u003eA) and identified 20 gene modules through dynamic tree cutting (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig2\\\" class=\\\"InternalRef\\\"\\u003e2\\u003c/span\\u003eB). The midnightblue module showed the strongest association with COPD status (module\\u0026ndash;trait r\\u0026thinsp;=\\u0026thinsp;0.893, P\\u0026thinsp;=\\u0026thinsp;5.4\\u0026times;10\\u003csup\\u003e\\u0026minus;\\u0026thinsp;32\\u003c/sup\\u003e; Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig2\\\" class=\\\"InternalRef\\\"\\u003e2\\u003c/span\\u003eC\\u0026ndash;D) and contained 89 genes. After normalization of the GSE47460 expression matrix, differential expression analysis identified 170 DEGs between COPD and control samples (volcano plot, Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig2\\\" class=\\\"InternalRef\\\"\\u003e2\\u003c/span\\u003eE; top 50 heat map, Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig2\\\" class=\\\"InternalRef\\\"\\u003e2\\u003c/span\\u003eF). Intersecting module genes with DEGs yielded 24 hub genes (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig2\\\" class=\\\"InternalRef\\\"\\u003e2\\u003c/span\\u003eG), which were advanced for biomarker screening.\\u003c/p\\u003e \\u003cp\\u003e \\u003c/p\\u003e \\u003cp\\u003e \\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec22\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eMachine-learning screen and diagnostic performance\\u003c/h2\\u003e \\u003cp\\u003eWe evaluated 12 algorithms across 113 model combinations to prioritize diagnostic features (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig3\\\" class=\\\"InternalRef\\\"\\u003e3\\u003c/span\\u003eA). In the training set, the combined model achieved an AUC of 0.996 (95% CI, 0.991\\u0026ndash;0.999; Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig3\\\" class=\\\"InternalRef\\\"\\u003e3\\u003c/span\\u003eB). External validation in GSE76925 yielded an AUC of 0.834 (95% CI, 0.755\\u0026ndash;0.906; Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig3\\\" class=\\\"InternalRef\\\"\\u003e3\\u003c/span\\u003eC), indicating preserved discrimination in an independent cohort. Confusion matrices for the training and validation sets demonstrated high sensitivity and specificity (Figs.\\u0026nbsp;\\u003cspan refid=\\\"Fig3\\\" class=\\\"InternalRef\\\"\\u003e3\\u003c/span\\u003eD and \\u003cspan refid=\\\"Fig3\\\" class=\\\"InternalRef\\\"\\u003e3\\u003c/span\\u003eE). Across models, BHLHE22, DPP6, DHRS9, and CYP1B1 were most frequently selected and ranked highly in terms of importance. Correlation analyses revealed strong positive co-variation between BHLHE22 and CYP1B1 (r\\u0026thinsp;=\\u0026thinsp;0.62) and an inverse correlation between DPP6 and CYP1B1 (r = -0.55; Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig3\\\" class=\\\"InternalRef\\\"\\u003e3\\u003c/span\\u003eG). Differential expression visualizations confirmed that three genes were upregulated in COPD, and DPP6 was downregulated (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig3\\\" class=\\\"InternalRef\\\"\\u003e3\\u003c/span\\u003eF).\\u003c/p\\u003e \\u003cp\\u003e \\u003c/p\\u003e \\u003cdiv id=\\\"Sec23\\\" class=\\\"Section3\\\"\\u003e \\u003ch2\\u003eExternal validation and pathway context for top biomarkers\\u003c/h2\\u003e \\u003cp\\u003ePer-gene ROC curves in the independent GSE37768 cohort showed that BHLHE22, DPP6, and CYP1B1 achieved AUCs of 0.764\\u0026ndash;0.795, whereas DHRS9 performed lower (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig4\\\" class=\\\"InternalRef\\\"\\u003e4\\u003c/span\\u003eA). A nomogram integrating these predictors emphasized CYP1B1 as the strongest contributor to diagnostic probability (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig4\\\" class=\\\"InternalRef\\\"\\u003e4\\u003c/span\\u003eB\\u0026ndash;C). To contextualize biology, GSEA revealed that BHLHE22-high samples were enriched for collagen fibril organization and extracellular matrix remodeling (GO) and for chemokine signaling, complement/coagulation, cytokine\\u0026ndash;receptor interaction, and NK cell\\u0026ndash;mediated cytotoxicity (KEGG) (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig4\\\" class=\\\"InternalRef\\\"\\u003e4\\u003c/span\\u003eD\\u0026ndash;E). Similarly, CYP1B1-high samples showed GO enrichment for collagen processes, response to interleukin-1, and metallopeptidase activity, alongside KEGG enrichment for chemokine signaling, complement/coagulation, intestinal immune network for IgA, and NK cell\\u0026ndash;mediated cytotoxicity (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig4\\\" class=\\\"InternalRef\\\"\\u003e4\\u003c/span\\u003eF\\u0026ndash;G).\\u003c/p\\u003e \\u003cp\\u003e \\u003c/p\\u003e \\u003c/div\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec24\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eANN construction, calibration, and independent checks\\u003c/h2\\u003e \\u003cp\\u003eWe built an ANN using BHLHE22, DPP6, and CYP1B1 (architecture and weights, Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig5\\\" class=\\\"InternalRef\\\"\\u003e5\\u003c/span\\u003eA). The model demonstrated close agreement between the predicted and observed risks in both the training and validation datasets (calibration plots, Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig5\\\" class=\\\"InternalRef\\\"\\u003e5\\u003c/span\\u003eB). In GSE37768, all three biomarkers were significantly upregulated in COPD versus controls, consistent with discovery analyses (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig5\\\" class=\\\"InternalRef\\\"\\u003e5\\u003c/span\\u003eC\\u0026ndash;D). Collectively, the ANN achieved robust discrimination with good calibration, supporting the translational feasibility of a parsimonious gene-based classifier.\\u003c/p\\u003e \\u003cp\\u003e \\u003c/p\\u003e \\u003cdiv id=\\\"Sec25\\\" class=\\\"Section3\\\"\\u003e \\u003ch2\\u003eSingle-cell resolution of CYP1B1 and ferroptosis signatures\\u003c/h2\\u003e \\u003cp\\u003eIntegration of human lung scRNA-seq from COPD and control donors resolved 16 major cell types (marker panels in Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig6\\\" class=\\\"InternalRef\\\"\\u003e6\\u003c/span\\u003eA; UMAP in Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig6\\\" class=\\\"InternalRef\\\"\\u003e6\\u003c/span\\u003eB), with proportion differences summarized in Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig6\\\" class=\\\"InternalRef\\\"\\u003e6\\u003c/span\\u003eC. While BHLHE22, DPP6, and CYP1B1 were detected across multiple lineages, CYP1B1 showed significantly higher expression in airway secretory cells (ASCs) from COPD relative to controls (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig6\\\" class=\\\"InternalRef\\\"\\u003e6\\u003c/span\\u003eD; report log₂FC, adj. P). Stratifying ASCs by CYP1B1 median defined high- and low-expressing subpopulations; differential analysis indicated ferroptosis pathway enrichment in CYP1B1-high ASCs (KEGG) with increased AUCell ferroptosis scores (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig6\\\" class=\\\"InternalRef\\\"\\u003e6\\u003c/span\\u003eE\\u0026ndash;F).\\u003c/p\\u003e \\u003cp\\u003e \\u003c/p\\u003e \\u003cp\\u003e \\u003cb\\u003ePseudotime and cell\\u0026ndash;cell communication implicate ASC programs in the pathogenesis of COPD.\\u003c/b\\u003e \\u003c/p\\u003e \\u003cp\\u003eMonocle-based trajectories suggested bifurcating epithelial lineages, with ASCs enriched at a terminal branch (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig7\\\" class=\\\"InternalRef\\\"\\u003e7\\u003c/span\\u003eA\\u0026ndash;B), consistent with a remodeled secretory program in COPD. Global CellChat networks (Figs.\\u0026nbsp;\\u003cspan refid=\\\"Fig7\\\" class=\\\"InternalRef\\\"\\u003e7\\u003c/span\\u003eC\\u0026ndash;D) revealed increased outgoing and incoming signaling to/from ASCs, particularly with alveolar macrophages and other immune lineages. An ASC-centric view (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig7\\\" class=\\\"InternalRef\\\"\\u003e7\\u003c/span\\u003eE) highlighted ASCs as communication hubs, supporting a model in which secretory epithelial programs shape the COPD inflammatory microenvironment.\\u003c/p\\u003e \\u003cp\\u003e \\u003c/p\\u003e \\u003cp\\u003e \\u003cb\\u003eImmune infiltration correlates with hub genes.\\u003c/b\\u003e \\u003c/p\\u003e \\u003cp\\u003eBulk deconvolution revealed significant differences in multiple immune subsets between COPD and control subjects, including plasma cells, M0 macrophages, activated mast cells, and follicular helper T cells (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig8\\\" class=\\\"InternalRef\\\"\\u003e8\\u003c/span\\u003eA), with inter-subset correlations summarized in Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig8\\\" class=\\\"InternalRef\\\"\\u003e8\\u003c/span\\u003eB. Expression of BHLHE22 and CYP1B1 correlated positively with plasma cells and M0 macrophages, whereas DPP6 correlated negatively with these subsets and positively with M2 macrophages (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig8\\\" class=\\\"InternalRef\\\"\\u003e8\\u003c/span\\u003eC\\u0026ndash;E), suggesting coordinated epithelial\\u0026ndash;immune remodeling in COPD.\\u003c/p\\u003e \\u003cp\\u003e \\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec26\\\" class=\\\"Section3\\\"\\u003e \\u003ch2\\u003eIn vivo and in vitro evidence of CYP1B1 upregulation with smoke exposure\\u003c/h2\\u003e \\u003cp\\u003eIn a 6-month cigarette-smoke mouse model, systemic and airway inflammation increased over time, as indicated by ELISAs (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig9\\\" class=\\\"InternalRef\\\"\\u003e9\\u003c/span\\u003eA\\u0026ndash;H), accompanied by higher pulmonary CYP1B1 expression (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig9\\\" class=\\\"InternalRef\\\"\\u003e9\\u003c/span\\u003eI). In 16HBE cells exposed to CSE, proteomic profiling corroborated CYP1B1 upregulation (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig9\\\" class=\\\"InternalRef\\\"\\u003e9\\u003c/span\\u003eJ). Histopathology in smoke-exposed mice revealed enlarged alveolar septa, increased glycogen deposition, and fibrosis (H\\u0026amp;E, PAS, Picrosirius Red; Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig10\\\" class=\\\"InternalRef\\\"\\u003e10\\u003c/span\\u003eA\\u0026ndash;C), accompanied by a significant impairment of lung function (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig10\\\" class=\\\"InternalRef\\\"\\u003e10\\u003c/span\\u003eD\\u0026ndash;K). Ferroptosis-related indicators also increased significantly (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig10\\\" class=\\\"InternalRef\\\"\\u003e10\\u003c/span\\u003eL-N). These findings support a link between cigarette smoke exposure, epithelial remodeling, and CYP1B1 induction.\\u003c/p\\u003e \\u003cp\\u003e \\u003c/p\\u003e \\u003cp\\u003e \\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec27\\\" class=\\\"Section3\\\"\\u003e \\u003ch2\\u003eInhibition of CYP1B1 attenuates ferroptosis signals\\u003c/h2\\u003e \\u003cp\\u003eIn the CSE model constructed by cigarette smoke-induced 16-HBE cells, lentiviral transfection was used to knock down the expression of CYP1B1, and we found that ferroptosis signal-related indicators such as GPX4, MDA, and 4-HNE had a significant decline (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig11\\\" class=\\\"InternalRef\\\"\\u003e11\\u003c/span\\u003eA-C). This suggests that inhibition of CYP1B1 expression can inhibit ferroptosis in airway epithelial cells, which may be helpful in the treatment of COPD.\\u003c/p\\u003e \\u003cp\\u003e \\u003c/p\\u003e \\u003cp\\u003e \\u003cb\\u003eDocking nominates natural compounds with predicted CYP1B1 binding.\\u003c/b\\u003e \\u003c/p\\u003e \\u003cp\\u003eDGIdb-guided filtering of traditional Chinese medicine (TCM) compounds prioritized α-naphthoflavone, β-naphthoflavone, chrysin, and naringenin as putative CYP1B1 binders. Rigid docking (PyMOL/AutoDock workflow) predicted binding energies\\u0026thinsp;\\u0026lt;\\u0026thinsp;\\u0026minus;\\u0026thinsp;5.0 kcal/mol (\\u0026minus;\\u0026thinsp;7.11, \\u0026minus;\\u0026thinsp;6.86, \\u0026minus;\\u0026thinsp;5.56, \\u0026minus;\\u0026thinsp;5.45 kcal/mol; Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig12\\\" class=\\\"InternalRef\\\"\\u003e12\\u003c/span\\u003eA\\u0026ndash;D), consistent with favorable interactions in the CYP1B1 active site. These in silico data are hypothesis-generating and motivate biochemical validation of enzyme inhibition and selectivity.\\u003c/p\\u003e \\u003cp\\u003e \\u003c/p\\u003e \\u003c/div\\u003e \\u003c/div\\u003e\"},{\"header\":\"Discussion\",\"content\":\"\\u003cp\\u003eCOPD remains a leading global contributor to morbidity and mortality [\\u003cspan citationid=\\\"CR24\\\" class=\\\"CitationRef\\\"\\u003e24\\u003c/span\\u003e]. While the disease is heterogeneous [\\u003cspan citationid=\\\"CR25\\\" class=\\\"CitationRef\\\"\\u003e25\\u003c/span\\u003e], its clinical definition centers on persistent airflow limitation and chronic respiratory symptoms. To enhance diagnostic precision and mechanistic understanding, we employed an integrative multi-omics pipeline to identify and validate biomarkers with potential clinical utility. Our analyses identify CYP1B1 as a promising biomarker and mechanistic node in COPD, with convergent evidence spanning bulk transcriptomics, machine learning, single-cell mapping, and experimental models.\\u003c/p\\u003e \\u003cp\\u003eUsing publicly available COPD cohorts, we defined 170 DEGs, mapped them onto co-expression modules, and intersected signals to yield 24 hub genes. A 12-algorithm/113-combination screen prioritized BHLHE22, CYP1B1, and DPP6, with external validation supporting diagnostic discrimination (combined model AUC up to 0.996 in training and 0.834 in GSE76925; single-gene AUCs 0.764\\u0026ndash;0.795 in GSE37768). Single-cell analyses localized CYP1B1 upregulation to airway secretory cells (ASCs) in COPD, where high CYP1B1 expression coincided with enrichment of ferroptosis signatures. In smoke-exposed mouse lungs and CSE-treated bronchial epithelial cells, CYP1B1 expression increased alongside histologic remodeling and impaired lung function. Collectively, these observations support an association between CYP1B1 upregulation, epithelial stress programs consistent with ferroptosis, and the pathobiology of COPD, while acknowledging that causality requires targeted perturbation.\\u003c/p\\u003e \\u003cp\\u003eCYP1B1, a xenobiotic-metabolizing cytochrome P450, bioactivates environmental toxicants (including polycyclic aromatic hydrocarbons) and modulates steroid and eicosanoid pathways [\\u003cspan citationid=\\\"CR26\\\" class=\\\"CitationRef\\\"\\u003e26\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR27\\\" class=\\\"CitationRef\\\"\\u003e27\\u003c/span\\u003e]. In the lung, CYP1B1 activity has been implicated in oxidative stress and inflammatory signaling pertinent to chronic airway disease [\\u003cspan citationid=\\\"CR27\\\" class=\\\"CitationRef\\\"\\u003e27\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR28\\\" class=\\\"CitationRef\\\"\\u003e28\\u003c/span\\u003e]. Altered CYP1B1 expression has been linked to epithelial dysfunction\\u0026mdash;airway remodeling, impaired mucociliary clearance, and heightened susceptibility to inhaled insults\\u0026mdash;features that align with COPD pathogenesis [\\u003cspan citationid=\\\"CR29\\\" class=\\\"CitationRef\\\"\\u003e29\\u003c/span\\u003e]. Moreover, CYP1B1 polymorphisms have been associated with interindividual variation in lung function decline and COPD risk [\\u003cspan citationid=\\\"CR30\\\" class=\\\"CitationRef\\\"\\u003e30\\u003c/span\\u003e]. Integrating these observations, our data position CYP1B1 as a biologically plausible mediator of smoke-related redox imbalance and epithelial remodeling in COPD, and a tractable node for therapeutic exploration [\\u003cspan citationid=\\\"CR31\\\" class=\\\"CitationRef\\\"\\u003e31\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR32\\\" class=\\\"CitationRef\\\"\\u003e32\\u003c/span\\u003e].\\u003c/p\\u003e \\u003cp\\u003ePrimary immunodeficiencies (PIDs) encompass defects in cellular and humoral immunity that predispose to recurrent airway infections and chronic inflammation. More than half of PID subtypes impair antibody production and contribute to recurrent sinusitis and pulmonary infections; persistent upper/lower respiratory infections can drive airway inflammation, obstruction, and, in some cases, structural remodeling compatible with COPD phenotypes [\\u003cspan citationid=\\\"CR33\\\" class=\\\"CitationRef\\\"\\u003e33\\u003c/span\\u003e]. Respiratory complications of PIDs are associated with downstream risks\\u0026mdash;including severe asthma, bronchiectasis, and COPD\\u0026mdash;and increased mortality [\\u003cspan citationid=\\\"CR34\\\" class=\\\"CitationRef\\\"\\u003e34\\u003c/span\\u003e]. In the context of our findings, these observations highlight how host-defense perturbations can exacerbate epithelial injury and immune\\u0026ndash;epithelial crosstalk in susceptible individuals.\\u003c/p\\u003e \\u003cp\\u003eSingle-cell results indicated that differentially expressed genes were enriched in epithelial compartments, with ASCs showing the clearest CYP1B1 signal. Epithelial\\u0026ndash;macrophage crosstalk is central to airway remodeling, as alveolar and airway macrophages orchestrate pathogen clearance, surfactant turnover, tissue repair, and homeostasis. In COPD, they exhibit impaired phagocytosis/efferocytosis, dysregulated cytokine release, and heightened oxidative stress [\\u003cspan citationid=\\\"CR35\\\" class=\\\"CitationRef\\\"\\u003e35\\u003c/span\\u003e]. Beyond the lung, macrophage inflammatory phenotypes and lysosomal signaling influence disease processes across organs [\\u003cspan citationid=\\\"CR36\\\" class=\\\"CitationRef\\\"\\u003e36\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR37\\\" class=\\\"CitationRef\\\"\\u003e37\\u003c/span\\u003e]; by analogy, macrophage dysfunction in COPD may intersect with epithelial redox pathways, including lipid peroxidation and iron handling. Our data suggest that CYP1B1-high ASCs reside within an inflamed network where macrophage interactions are increased, consistent with a feed-forward loop of epithelial stress and immune activation.\\u003c/p\\u003e \\u003cdiv id=\\\"Sec29\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eLimitation\\u003c/h2\\u003e \\u003cp\\u003eThis study has limitations. First, although we leveraged multiple GEO cohorts, sample availability limits power for some subgroup analyses. Second, the external validation cohorts differ in clinical covariates and platforms, which may affect generalizability. Third, while single-cell analyses map CYP1B1 to airway secretory cells and associate high expression with ferroptosis signatures, we did not perform genetic or pharmacologic perturbation of CYP1B1 with standardized ferroptosis rescue controls (e.g., ferrostatin-1, liproxstatin-1, deferoxamine). Finally, docking results are in silico and require biochemical confirmation (enzyme inhibition, selectivity, and cellular target engagement).\\u003c/p\\u003e \\u003c/div\\u003e\"},{\"header\":\"Conclusion\",\"content\":\"\\u003cp\\u003eIn aggregate, our integrative analyses identify CYP1B1 as a potential biomarker for COPD and a putative mediator of epithelial stress, consistent with ferroptosis in airway secretory cells. Clinically, CYP1B1 expression and derived gene-based models may aid diagnostic stratification; mechanistically, CYP1B1 presents a druggable entry point for modulating redox-lipid peroxidation pathways. Prioritized natural compounds from docking provide testable leads, but their translation will require validating target engagement, ferroptosis-specific rescue, and safety/PK. Considerations. Prospective, multi-center studies integrating single-cell profiling, circulating biomarkers, and interventional perturbations should establish clinical utility and clarify therapeutic potential.\\u003c/p\\u003e\"},{\"header\":\"Abbreviations\",\"content\":\"\\u003cdiv align=\\\"center\\\"\\u003e\\n \\u003ctable border=\\\"1\\\" cellspacing=\\\"0\\\" cellpadding=\\\"0\\\" width=\\\"100%\\\"\\u003e\\n \\u003cthead\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd style=\\\"width: 23px;\\\"\\u003e\\n \\u003cp\\u003eAbbreviation\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 76px;\\\"\\u003e\\n \\u003cp\\u003eDefinition\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003c/thead\\u003e\\n \\u003ctbody\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd style=\\\"width: 23px;\\\"\\u003e\\n \\u003cp\\u003eCOPD\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 76px;\\\"\\u003e\\n \\u003cp\\u003eChronic Obstructive Pulmonary Disease\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd style=\\\"width: 23px;\\\"\\u003e\\n \\u003cp\\u003escRNA\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 76px;\\\"\\u003e\\n \\u003cp\\u003eSingle-cell RNA\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd style=\\\"width: 23px;\\\"\\u003e\\n \\u003cp\\u003eDEGs\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 76px;\\\"\\u003e\\n \\u003cp\\u003eDifferentially Expressed Genes\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd style=\\\"width: 23px;\\\"\\u003e\\n \\u003cp\\u003eWGCNA\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 76px;\\\"\\u003e\\n \\u003cp\\u003eWeighted Gene Co-Expression Network Analysis\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd style=\\\"width: 23px;\\\"\\u003e\\n \\u003cp\\u003eROC\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 76px;\\\"\\u003e\\n \\u003cp\\u003eReceiver Operating Characteristic\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd style=\\\"width: 23px;\\\"\\u003e\\n \\u003cp\\u003eAUC\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 76px;\\\"\\u003e\\n \\u003cp\\u003eArea Under the Curve\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd style=\\\"width: 23px;\\\"\\u003e\\n \\u003cp\\u003eGSEA\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 76px;\\\"\\u003e\\n \\u003cp\\u003eGene Set Enrichment Analysis\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd style=\\\"width: 23px;\\\"\\u003e\\n \\u003cp\\u003eGO\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 76px;\\\"\\u003e\\n \\u003cp\\u003eGene Ontology\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd style=\\\"width: 23px;\\\"\\u003e\\n \\u003cp\\u003eKEGG\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 76px;\\\"\\u003e\\n \\u003cp\\u003eKyoto Encyclopedia of Genes and Genomes\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd style=\\\"width: 23px;\\\"\\u003e\\n \\u003cp\\u003eANN\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 76px;\\\"\\u003e\\n \\u003cp\\u003eArtificial Neural Network\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd style=\\\"width: 23px;\\\"\\u003e\\n \\u003cp\\u003eUMAP\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 76px;\\\"\\u003e\\n \\u003cp\\u003eUniform Manifold Approximation and Projection\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd style=\\\"width: 23px;\\\"\\u003e\\n \\u003cp\\u003eCSEs\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 76px;\\\"\\u003e\\n \\u003cp\\u003eCigarette Smoke Extracts\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd style=\\\"width: 23px;\\\"\\u003e\\n \\u003cp\\u003eAms\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 76px;\\\"\\u003e\\n \\u003cp\\u003eAlveolar Macrophages\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd style=\\\"width: 23px;\\\"\\u003e\\n \\u003cp\\u003eAT2s\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 76px;\\\"\\u003e\\n \\u003cp\\u003eAlveolar Type II epithelial cells\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd style=\\\"width: 23px;\\\"\\u003e\\n \\u003cp\\u003eCD8Ts\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 76px;\\\"\\u003e\\n \\u003cp\\u003eCD8+ T cells\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd style=\\\"width: 23px;\\\"\\u003e\\n \\u003cp\\u003eECs\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 76px;\\\"\\u003e\\n \\u003cp\\u003eEndothelial cells\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd style=\\\"width: 23px;\\\"\\u003e\\n \\u003cp\\u003eCils\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 76px;\\\"\\u003e\\n \\u003cp\\u003eCiliated cells\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd style=\\\"width: 23px;\\\"\\u003e\\n \\u003cp\\u003eMonos\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 76px;\\\"\\u003e\\n \\u003cp\\u003eMonocytes\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd style=\\\"width: 23px;\\\"\\u003e\\n \\u003cp\\u003eASCs\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 76px;\\\"\\u003e\\n \\u003cp\\u003eAirway secretory cells\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd style=\\\"width: 23px;\\\"\\u003e\\n \\u003cp\\u003eBCs\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 76px;\\\"\\u003e\\n \\u003cp\\u003eB cells\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd style=\\\"width: 23px;\\\"\\u003e\\n \\u003cp\\u003eAT1s\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 76px;\\\"\\u003e\\n \\u003cp\\u003eAlveolar Type I cells\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd style=\\\"width: 23px;\\\"\\u003e\\n \\u003cp\\u003ePMVECs\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 76px;\\\"\\u003e\\n \\u003cp\\u003ePulmonary microvascular endothelial cells\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd style=\\\"width: 23px;\\\"\\u003e\\n \\u003cp\\u003eFbs\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 76px;\\\"\\u003e\\n \\u003cp\\u003eFibroblasts\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd style=\\\"width: 23px;\\\"\\u003e\\n \\u003cp\\u003eSers\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 76px;\\\"\\u003e\\n \\u003cp\\u003eAirway serous cells\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd style=\\\"width: 23px;\\\"\\u003e\\n \\u003cp\\u003eMCs\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 76px;\\\"\\u003e\\n \\u003cp\\u003eMast cells\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd style=\\\"width: 23px;\\\"\\u003e\\n \\u003cp\\u003eLECs\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 76px;\\\"\\u003e\\n \\u003cp\\u003eLymphatic endothelial cells\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd style=\\\"width: 23px;\\\"\\u003e\\n \\u003cp\\u003ePCs\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 76px;\\\"\\u003e\\n \\u003cp\\u003eProliferating cells\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd style=\\\"width: 23px;\\\"\\u003e\\n \\u003cp\\u003eMac_Monos\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 76px;\\\"\\u003e\\n \\u003cp\\u003eMacrophages/Monocytes\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003ctr\\u003e\\n \\u003ctd style=\\\"width: 23px;\\\"\\u003e\\n \\u003cp\\u003eTCM\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003ctd style=\\\"width: 76px;\\\"\\u003e\\n \\u003cp\\u003eTraditional Chinese medicine\\u003c/p\\u003e\\n \\u003c/td\\u003e\\n \\u003c/tr\\u003e\\n \\u003c/tbody\\u003e\\n \\u003c/table\\u003e\\n\\u003c/div\\u003e\"},{\"header\":\"Declarations\",\"content\":\"\\u003cp\\u003e\\u003cstrong\\u003e\\u003cem\\u003eEthics approval and consent to participate\\u003c/em\\u003e\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eThe study involving animals was reviewed and approved by Ethics Committee of Guangzhou Medical University\\u0026nbsp;(GY2025-036).\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003e\\u003cem\\u003eAvailability of data and materials\\u003c/em\\u003e\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eAll data generated or analysed during this study are included in this published article.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003e\\u003cem\\u003eCompeting interest\\u003c/em\\u003e\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eThe authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003e\\u003cem\\u003eFunding\\u003c/em\\u003e\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eThe author(s) declare that financial support was received for the research and/or publication of this article. This work was sponsored by National Natural Science Foundation of China (81900044 and 82170500), Shanghai Municipal Education Commission (2021 Technology and Innovation-03-163), Natural Science Foundation of Hunan Province (2021JJ40484 and 2024JJ5350), Major Project of Guangzhou National Laboratory (GZNL2024A02005) and the grant of State Key Laboratory of Respiratory Disease (SKLRD-Z-202315).\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003e\\u003cem\\u003eAuthors\\u0026rsquo; contributions\\u003c/em\\u003e\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eLL and MJ: Conceptualization, Data curation, Formal Analysis, Investigation, Methodology, Software, Visualization, Writing\\u0026ndash; original draft. ML and KD: Data curation, Investigation, Methodology, Software, Visualization. DW, XY and XT: Data curation, Methodology, Software, Visualization. ZM, YT and RS: Data curation, Software, Visualization. SL: Conceptualization, Funding acquisition, Investigation, Project administration, Resources, Supervision, Writing\\u0026ndash; review \\u0026amp; editing.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003e\\u003cem\\u003eAcknowledgments\\u003c/em\\u003e\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eWe thank the participants and staff of the National Center for Biotechnology Information. We thank Wen Qianmei and Zuo yujie form the First Affiliated Hospital of Guangzhou Medical University for their assistance in the construction of the mouse model. At the same time, we would also like to acknowledge assistance from proofreaders and editors.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003e\\u003cem\\u003eGenerative AI statement\\u003c/em\\u003e\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eThe author(s) declare that no Generative AI was used in the creation of this manuscript.\\u003c/p\\u003e\"},{\"header\":\"References\",\"content\":\"\\u003col\\u003e\\u003cli\\u003e\\u003cspan\\u003eChristenson SA, Smith BM, Bafadhel M, Putcha N. Chronic obstructive pulmonary disease. Lancet. 2022;399(10342):2227\\u0026ndash;42.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eLabaki WW, Rosenberg SR. Chronic Obstructive Pulmonary Disease. Ann Intern Med. 2020;173(3):Itc17\\u0026ndash;32.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eBrightling C, Greening N. Airway inflammation in COPD: progress to precision medicine. Eur Respir J, 2019. 54(2).\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eBarnes PJ. Cellular and molecular mechanisms of asthma and COPD. Clin Sci (Lond). 2017;131(13):1541\\u0026ndash;58.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eWistuba II, Mao L, Gazdar AF. Smoking molecular damage in bronchial epithelium. Oncogene. 2002;21(48):7298\\u0026ndash;306.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eMercer RR, Russell ML, Roggli VL, Crapo JD. Cell number and distribution in human and rat airways. Am J Respir Cell Mol Biol. 1994;10(6):613\\u0026ndash;24.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eBrusselle GG, Joos GF, Bracke KR. New insights into the immunology of chronic obstructive pulmonary disease. Lancet. 2011;378(9795):1015\\u0026ndash;26.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eHogg JC, Chu F, Utokaparch S, et al. The nature of small-airway obstruction in chronic obstructive pulmonary disease. N Engl J Med. 2004;350(26):2645\\u0026ndash;53.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eZhao J, Cheng W, He X, et al. Chronic Obstructive Pulmonary Disease Molecular Subtyping and Pathway Deviation-Based Candidate Gene Identification. Cell J. 2018;20(3):326\\u0026ndash;32.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eTan J, Tedrow JR, Dutta JA, et al. Expression of RXFP1 Is Decreased in Idiopathic Pulmonary Fibrosis. Implications for Relaxin-based Therapies. Am J Respir Crit Care Med. 2016;194(11):1392\\u0026ndash;402.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eMorrow JD, Zhou X, Lao T, et al. Functional interactors of three genome-wide association study genes are differentially expressed in severe chronic obstructive pulmonary disease lung tissue. Sci Rep. 2017;7:44232.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eLeek JT, Johnson WE, Parker HS, et al. The sva package for removing batch effects and other unwanted variation in high-throughput experiments. Bioinformatics. 2012;28(6):882\\u0026ndash;3.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eHabermann AC, Gutierrez AJ, Bui LT et al. Single-cell RNA sequencing reveals profibrotic roles of distinct epithelial and mesenchymal lineages in pulmonary fibrosis. Sci Adv, 2020. 6(28): p. eaba1972.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eMa J, Chen T, Wu S, et al. iProX: an integrated proteome resource. Nucleic Acids Res. 2019;47(D1):D1211\\u0026ndash;7.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eChen T, Ma J, Liu Y, et al. iProX in 2021: connecting proteomics data sharing with big data. Nucleic Acids Res. 2022;50(D1):D1522\\u0026ndash;7.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eRitchie ME, Phipson B, Wu D, et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 2015;43(7):e47.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eKolde R. pheatmap: Pretty Heatmaps. 2010.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eWickham H. Ggplot2: Elegant graphics for data analysis. Cham, Switzerland: Springer International Publishing; 2016.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eLangfelder P, Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics. 2008;9:559.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eSubramanian A, Tamayo P, Mootha VK, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005;102(43):15545\\u0026ndash;50.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eFan X, Dong T, Yan K et al. PM2.5 increases susceptibility to acute exacerbation of COPD via NOX4/Nrf2 redox imbalance-mediated mitophagy. Redox Biol, 2023. 59.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eGuan R, Wang J, Cai Z et al. Hydrogen sulfide attenuates cigarette smoke-induced airway remodeling by upregulating SIRT1 signaling pathway. Redox Biol, 2020. 28.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eHe F, Liao B, Pu J et al. Exposure to Ambient Particulate Matter Induced COPD in a Rat Model and a Description of the Underlying Mechanism. Sci Rep, 2017. 7(1).\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eRaherison C, Girodet PO. Epidemiology of COPD. Eur Respir Rev. 2009;18(114):213\\u0026ndash;21.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eMannino DM. COPD: epidemiology, prevalence, morbidity and mortality, and disease heterogeneity. Chest. 2002;121(5 Suppl):s121\\u0026ndash;6.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eEsteve-Codina A, Hofer TP, Burggraf D, et al. Gender specific airway gene expression in COPD sub-phenotypes supports a role of mitochondria and of different types of leukocytes. Sci Rep. 2021;11(1):12848.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eZhu Y, Dutta S, Han Y, et al. Oxidative stress promotes lipid-laden macrophage formation via CYP1B1. Redox Biol. 2025;79:103481.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eZhang S, Duan H, Yan J. Identifying biomarkers of endoplasmic reticulum stress and analyzing immune cell infiltration in chronic obstructive pulmonary disease using machine learning. Front Med (Lausanne). 2024;11:1462868.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eReddy KD, Lan A, Boudewijn IM, et al. Current Smoking Alters Gene Expression and DNA Methylation in the Nasal Epithelium of Patients with Asthma. Am J Respir Cell Mol Biol. 2021;65(4):366\\u0026ndash;77.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eObernolte H, Niehof M, Braubach P, et al. Cigarette smoke alters inflammatory genes and the extracellular matrix - investigations on viable sections of peripheral human lungs. Cell Tissue Res. 2022;387(2):249\\u0026ndash;60.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eMisiukiewicz-Stępien P, Mierzejewski M, Zajusz-Zubek E et al. RNA-Seq Analysis of UPM-Exposed Epithelium Co-Cultivated with Macrophages and Dendritic Cells in Obstructive Lung Diseases. Int J Mol Sci, 2022. 23(16).\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eFitzgerald LF, Lackey J, Moussa A, et al. Chronic aryl hydrocarbon receptor activity impairs muscle mitochondrial function with tobacco smoking. J Cachexia Sarcopenia Muscle. 2024;15(2):646\\u0026ndash;59.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eBerger M, Geng B, Cameron DW, et al. Primary immune deficiency diseases as unrecognized causes of chronic respiratory disease. Respir Med. 2017;132:181\\u0026ndash;8.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eJang JH, Kim JH, Park HS. Current Issues in the Management of IgG Subclass Deficiencies in Adults With Chronic Respiratory Diseases. Allergy Asthma Immunol Res. 2023;15(5):562\\u0026ndash;79.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eHu YS, Shao XJ, Xing L et al. Single-Cell Sequencing of Lung Macrophages and Monocytes Reveals Novel Therapeutic Targets in COPD. Cells, 2023. 12(24).\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eOgawa D, Stone JF, Takata Y, et al. Liver x receptor agonists inhibit cytokine-induced osteopontin expression in macrophages through interference with activator protein-1 signaling pathways. Circ Res. 2005;96(7):e59\\u0026ndash;67.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eWeber K, Schilling JD. Distinct lysosome phenotypes influence inflammatory function in peritoneal and bone marrow-derived macrophages. Int J Inflam, 2014. 2014: p. 154936.\\u003c/span\\u003e\\u003c/li\\u003e\\u003c/ol\\u003e\"}],\"fulltextSource\":\"\",\"fullText\":\"\",\"funders\":[],\"hasAdminPriorityOnWorkflow\":false,\"hasManuscriptDocX\":true,\"hasOptedInToPreprint\":true,\"hasPassedJournalQc\":\"\",\"hasAnyPriority\":false,\"hideJournal\":true,\"highlight\":\"\",\"institution\":\"\",\"isAcceptedByJournal\":false,\"isAuthorSuppliedPdf\":false,\"isDeskRejected\":\"\",\"isHiddenFromSearch\":false,\"isInQc\":false,\"isInWorkflow\":false,\"isPdf\":false,\"isPdfUpToDate\":true,\"isWithdrawnOrRetracted\":false,\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"researchsquare\",\"isNatureJournal\":false,\"hasQc\":true,\"allowDirectSubmit\":true,\"externalIdentity\":\"\",\"sideBox\":\"\",\"snPcode\":\"\",\"submissionUrl\":\"/submission\",\"title\":\"Research Square\",\"twitterHandle\":\"researchsquare\",\"acdcEnabled\":true,\"dfaEnabled\":false,\"editorialSystem\":\"\",\"reportingPortfolio\":\"\",\"inReviewEnabled\":false,\"inReviewRevisionsEnabled\":true},\"keywords\":\"COPD, CYP1B1, ferroptosis, multi-omics, single-cell RNA sequencing, biomarker discovery\",\"lastPublishedDoi\":\"10.21203/rs.3.rs-8049999/v1\",\"lastPublishedDoiUrl\":\"https://doi.org/10.21203/rs.3.rs-8049999/v1\",\"license\":{\"name\":\"CC BY 4.0\",\"url\":\"https://creativecommons.org/licenses/by/4.0/\"},\"manuscriptAbstract\":\"\\u003cp\\u003e\\u003cstrong\\u003eBackground: \\u003c/strong\\u003eChronic obstructive pulmonary disease (COPD) causes progressive airflow limitation and remains without disease-modifying therapy. Ferroptosis—iron-dependent lipid peroxidation—has emerged as a potential mechanism driving epithelial dysfunction and chronic inflammation; however, its upstream regulators in COPD remain incompletely defined. We hypothesized that integrative multi-omics could identify robust biomarkers and pathways, with translational potential for diagnosis and targeted therapy.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eObjective \\u003c/strong\\u003ePrimary: discover and validate biomarkers/pathways underlying COPD via bulk transcriptomics, single-cell analyses, machine learning (ML), and experimental validation. Secondary: evaluate CYP1B1 as a diagnostic biomarker and explore druggability through molecular docking.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eMethods: \\u003c/strong\\u003ePublic cohorts (GSE47460, GSE76925, GSE37768; total discovery/validation samples reported) were integrated with ComBat batch correction. Single-cell RNA-seq datasets (GSE196341, GSE135893; 12 COPD vs 12 controls) profiled cell-type localization. Differential expression, WGCNA, and enrichment (GO/KEGG/GSEA) were performed. A 12-algorithm ML framework (113 combinations) plus ANN modeling assessed diagnostic performance (ROC/AUC, confusion matrices, calibration). CIBERSORT estimated immune infiltration. A cigarette-smoke mouse model and 16HBE cell assays provided experimental validation; small-n proteomics were deposited (PXD068247; 3 vs 3). Molecular docking screened candidate CYP1B1-binding compounds and recorded binding energies. Statistical reporting included n, effect sizes, 95% CIs, and multiple-testing control.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eResults: \\u003c/strong\\u003eTwenty-four hub genes were initially identified; four (BHLHE22, DPP6, DHRS9, CYP1B1) were prioritized by ML across 113 model combinations. In the training set, the combined model achieved an AUC of 0.996 (95% CI, 0.991–0.999), with an external validation AUC of 0.834 (95% CI, 0.755–0.906). Single-gene ROC in an external cohort yielded AUCs of 0.764–0.795, with CYP1B1 being the most consistent. Single-cell analysis localized CYP1B1 upregulation to airway secretory cells (ASCs) and linked high CYP1B1 expression to the activation of the ferroptosis pathway. In mouse and 16HBE models, cigarette smoke increased lung inflammation/fibrosis and upregulated CYP1B1; proteomics corroborated expression changes. Docking identified α-/β-naphthoflavone, chrysin, and naringenin as CYP1B1 binders (best energies ≈ −7.11 to −5.45 kcal/mol). Immune deconvolution associated CYP1B1 with macrophage and plasma-cell signals.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eConclusions: \\u003c/strong\\u003eIntegrative multi-omics implicates CYP1B1-mediated ferroptosis in ASCs as a central pathway in COPD, with diagnostic promise and a tractable chemistry space. Prospective validation in larger cohorts, causal perturbation of CYP1B1–ferroptosis in vitro/in vivo, and pharmacology against prioritized ligands are warranted to translate these findings.\\u003c/p\\u003e\",\"manuscriptTitle\":\"CYP1B1-Mediated Ferroptosis Defines a Biomarker and Therapeutic Target in COPD Across Multi-omics and Single-Cell\",\"msid\":\"\",\"msnumber\":\"\",\"nonDraftVersions\":[{\"code\":1,\"date\":\"2026-01-01 08:01:08\",\"doi\":\"10.21203/rs.3.rs-8049999/v1\",\"editorialEvents\":[{\"type\":\"communityComments\",\"content\":0}],\"status\":\"published\",\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"researchsquare\",\"isNatureJournal\":false,\"hasQc\":true,\"allowDirectSubmit\":true,\"externalIdentity\":\"\",\"sideBox\":\"\",\"snPcode\":\"\",\"submissionUrl\":\"/submission\",\"title\":\"Research Square\",\"twitterHandle\":\"researchsquare\",\"acdcEnabled\":true,\"dfaEnabled\":false,\"editorialSystem\":\"\",\"reportingPortfolio\":\"\",\"inReviewEnabled\":false,\"inReviewRevisionsEnabled\":true}}],\"origin\":\"\",\"ownerIdentity\":\"a3afcf31-9a23-4b46-890a-e3a4c45796e6\",\"owner\":[],\"postedDate\":\"January 1st, 2026\",\"published\":true,\"recentEditorialEvents\":[],\"rejectedJournal\":[],\"revision\":\"\",\"amendment\":\"\",\"status\":\"posted\",\"subjectAreas\":[],\"tags\":[],\"updatedAt\":\"2026-03-01T12:39:26+00:00\",\"versionOfRecord\":[],\"versionCreatedAt\":\"2026-01-01 08:01:08\",\"video\":\"\",\"vorDoi\":\"\",\"vorDoiUrl\":\"\",\"workflowStages\":[]},\"version\":\"v1\",\"identity\":\"rs-8049999\",\"journalConfig\":\"researchsquare\"},\"__N_SSP\":true},\"page\":\"/article/[identity]/[[...version]]\",\"query\":{\"redirect\":\"/article/rs-8049999\",\"identity\":\"rs-8049999\",\"version\":[\"v1\"]},\"buildId\":\"XKTyCvWXoU3ODBz1xrDgd\",\"isFallback\":false,\"isExperimentalCompile\":false,\"dynamicIds\":[84888],\"gssp\":true,\"scriptLoader\":[]}","source_license":"CC-BY-4.0","license_restricted":false}