Integrative GWAS and Mendelian Randomization Analysis Identifies IREB2 and CD27+ Memory B Cells as Core Drivers of COPD to Lung Cancer Progression

Abstract

Chronic obstructive pulmonary disease (COPD) associates with increased lung cancer incidence and shares genetic susceptibility, yet its independent causal role and driver mechanisms are poorly understood. We integrated data from the National Health and Nutrition Examination Survey (NHANES) cohort with genome-wide association studies (GWAS) summary statistics and Mendelian randomization analyses to map genetic correlations and infer causality between COPD phenotypes and lung cancer. Post-GWAS methods-including transcriptome-wide association study, colocalization, partitioned heritability via heritability estimation from summary statistics (ρ-HESS), and cross-phenotype association (CPASSOC)-identified shared susceptibility loci, highlighting IREB2 and CD27⁺ B cells as potential mediators. Elevated IREB2 expression correlated with accelerated lung-function decline in COPD but predicted improved prognosis in lung cancer B cells, whereas higher CD27⁺ B cell levels in COPD were associated with protumorigenic activity. Single-cell transcriptomic analysis and in vitro knockdown experiments confirmed IREB2's role in modulating B-cell activation and apoptosis pathways within tumors. These results support COPD as an independent lung cancer risk factor and implicate IREB2 and CD27⁺ B cells in COPD-to-cancer progression, laying groundwork for early detection and targeted intervention in high-risk individuals.

Keywords: GWAS; IREB2; chronic obstructive pulmonary disease; lung cancer; mendelian randomization.

FIGURE 1

LDSC indicates that COPD, emphysema, chronic bronchitis, and lung function are genetically correlated with lung cancer. (A) MR analyses of COPD, emphysema, chronic bronchitis, lung function, lung cancer, LUAD, LUSC, and SCLC, after conditioning on the effects of SNPs on other cell types. Assumption 1, IVs are associated the exposure; assumption 2, IVs are not affected by confounders; assumption 3, IVs influence outcome solely through exposure, excluding alternative pathways. (B and C) Whole‐genome genetic correlations of COPD, emphysema, chronic bronchitis, and lung function in IEU GWAS (B) and FinnGen GWAS (C) with lung cancer using LDSC. Colors represent the magnitude of the genetic correlation of COPD, emphysema, chronic bronchitis, and lung function with lung cancer (lung cancer, LUAD, LUSC, and SCLC), using LDSC, with red indicating positive genetic correlation and blue indicating negative genetic correlation. Numbers represent the genetic correlation. * (p < 0.05), ** (p < 0.005), and *** (p < 0.001) represent significance. All p values are two sided.

FIGURE 2

MR results suggest COPD is a driving factor for lung cancer. (A) Forest plot of forward and reverse two‐sample MR analyses between COPD, with lung cancer, LUAD, LUSC, and SCLC after using MR PRESSO after Steiger filtering outlier SNPs. Effect sizes (Beta, 95% CI) are shown as the standard deviation change in lung cancer, LUAD, LUSC, and SCLC per standard deviation increase in COPD. Points on the forest plot represent effect size estimates, while whiskers denote 95% confidence intervals (CIs). Nsnp represents the final number of SNPs included in the analysis. All p values are two sided. (B) The diagram illustrates Mendelian randomization assessing COPD‐mediated pathways in smoking‐associated lung carcinogenesis: lung cancer (left), LUSC (middle), and LUAD (right). Rectangular nodes denote exposure (smoke), mediators (COPD), and outcome (lung cancer). Directional arrows represent causal effects with annotated parameters: β coefficients for mediation effects (β1–β4) and total effect β, alongside corresponding p values.

FIGURE 3

Overlapping gene loci and functions of COPD and lung cancer in lung tissue and peripheral blood. (A and B) Enrichment (A) and PPI network (B) analysis of overlapping genes from TWAS of COPD and lung cancer in lung tissue and peripheral blood using Metascape. (C and D) Upset plot showing the overlapping genes from TWAS of COPD (C) and emphysema/chronic bronchitis (D) with lung cancer in lung tissue and peripheral blood. (E) This workflow begins with COPD (ieu‐b‐106, FinnGen_COPD) and lung cancer (ieu‐a‐966, ieu‐a‐987, GCST004784, FinnGen_Lung_cancer) GWAS datasets. Cross‐phenotype associations are established through p‐HESS analysis and verified by MTAG/CPASSOC integration. Subsequent trans‐ethnic TWAS and colocalization analyses identify pleiotropic SNPs. cis‐/trans‐eQTL mapping (highlighted in light green) and secondary colocalization/TWAS validation (light red) prioritize causal genes underlying disease comorbidity. Arrows indicate analytical progression from raw data to gene prioritization. (F) Genome‐wide local heritability and genetic correlation between ieu‐b‐106 (COPD) and ieu‐a‐966 (lung cancer). The top two panels display the local SNP heritability estimates for COPD and lung cancer, where statistically significant positive correlations are indicated in blue and statistically significant negative correlations are indicated in red. The third panel presents local genetic covariance, with positive covariance shown in blue and negative covariance shown in red.

FIGURE 4

Overlapping gene loci and functions of COPD and lung cancer in lung tissue and peripheral blood. SMR (A) and colocalization (B) results of IREB2, PSMA4, and CHRNA5 in lung tissue and peripheral blood for COPD, emphysema, chronic bronchitis, lung function, lung cancer, LUAD, LUSC, and SCLC. Colors represent the effects of IREB2, PSMA4, and CHRNA5 on these conditions, with red indicating positive genetic correlation and blue indicating negative genetic correlation. Numbers represent the beta value of SMR. * (p < 0.05), ** (p < 0.005), and *** (p < 0.001) represent significant. All p values are two sided. For colocalization results, deeper colors indicate higher SNP.PP.H4 values, and numbers represent the SNP.PP.H4 values that passed the colocalization analysis threshold (SNP.PP.H4 > 0.8). (B) Expression of IREB2 (left) and PSMA4 (right) in LUAD and LUSC within the TCGA database. Two‐tailed t‐test. (C) Prognostic significance of IREB2 and PSMA4 expression in lung cancer subtypes: a Kaplan–Meier (KM) survival analysis of total lung cancer, LUAD, and LUSC cohorts.​ (D and E) RNA expression of IREB2 and PSMA4 in lung tissues of COPD patients from GSE76925 (D) and in whole blood from ECOPD cohort (E). Data are presented as mean ± SD. Two‐tailed t‐test. (F) Smoking status‐dependent peripheral blood expression heterogeneity of IREB2 and PSMA4 in the ECOPD cohort. Data are presented as mean ± SD. One‐way ANOVA. (G–I) Bar charts quantifying adjusted mean differences in annual lung function decline between high/low IREB2 expression groups: pre‐/postbronchodilator FEV₁ (G) and FVC (H); and between PSMA4 expression groups: pre‐/postbronchodilator FVC (I) in the ECOPD cohort. ​​Stratification based on median mRNA levels (qRT‐PCR), covariate‐adjusted for age, sex, BMI, smoking status/pack‐years; SEM error bars shown.

FIGURE 5

IREB2 and memory B cells as drivers of COPD progression to lung cancer. (A) Two‐sample MR analysis of IREB2 in the lung (GTEx_v8 database) and peripheral blood (eQTLGen) for COPD, emphysema, chronic bronchitis, lung function, lung cancer, LUAD, LUSC, and SCLC. * (p < 0.05), ** (p < 0.01), and *** (p < 0.001) represent significance. (B) The diagram illustrates Mendelian randomization assessing IREB2‐eQTLs‐mediated pathways in COPD‐associated lung cancer. Rectangular nodes denote exposure (COPD), mediators (IREB2 eQTLs), and outcome (lung cancer). Directional arrows represent causal effects with annotated parameters: β coefficients for mediation effects (β1–β4) and total effect β, alongside corresponding p values. (C) UMAP plot illustrating the clustering of cell subpopulations in the single‐cell dataset of NSCLC tumors. (D) UMAP plot showing the expression distribution of IREB2 in adjacent normal and cancerous tissues. (E) This schematic illustrates the analytical pipeline: causality testing of 731 peripheral immune cell traits (exposures) against COPD/lung cancer outcomes (FinnGen and IEU‐GWAS datasets) using inverse‐variance weighted MR, then Venn diagram identifying replicable causal mediators through cross‐dataset intersection, revealing two concordant features. (F) The diagram illustrates Mendelian randomization assessing CD27 on IgD⁻ CD38 dim B cell (left) or CD27 on switched memory B cell (right) mediated pathways in COPD‐associated lung cancer. Rectangular nodes denote exposure (COPD), mediators (IREB2 eQTLs), and outcome (lung cancer). Directional arrows represent causal effects with annotated parameters: β coefficients for mediation effects (β1–β3), alongside corresponding p values.

FIGURE 6

IREB2‐ and CD27‐expressing B cells as pivotal regulatory immune cells driving COPD‐to‐lung cancer progression. (A) UMAP plot illustrating the clustering of cell subpopulations in the single‐cell dataset of COPD patients from GSE173896. (B) Numbers and proportions of CD27‐positive and CD27‐negative B cells in nonsmokers, smokers, and COPD, derived from single‐cell sequencing data from GSE173896. (C) Venn diagram showing the overlap between DEGs in B cells from COPD patients and DEGs between CD27‐positive and CD27‐negative B cells from GSE173896. (D) Prognostic analysis of the signature constructed from 45 overlapping genes between B cells in COPD and CD 27⁺ B cells in lung cancer, LUAD, and LUSC, using KMplot. (E) Venn diagram showing the overlap of upregulated DEGs between IREB2‐positive B cells and IREB2+CD27+ double‐positive B cells. (F) Hazard ratios (HR) with 95% confidence intervals for a gene set evaluated in lung cancer, LUAD, and LUSC using KM analysis. Data are presented as bar chart with error bars depicting 95% CIs. Prognostic markers colored blue indicate favorable outcomes. (G) UMAP illustrates the integration and reclustering of B cells and plasma cells extracted from COPD and lung cancer scRNA‐seq datasets into five distinct subsets, with IREB2 expression distribution across these subsets. (H) Violin plots depict differential IREB2 expression among B cell subsets. (I) UMAP presents the integrated pan‐cancer B cell scRNA‐seq dataset annotated with 20 refined B cell clusters and five canonical subsets, alongside IREB2 expression patterns (Cirrocumulus database). (J) Dot plot displaying expression of IREB2 and B cell function/prognosis‐associated genes across pan‐cancer B cell subsets. (K) UMAP illustrates B memory cell subsets within the integrated pan‐cancer B memory cell scRNA‐seq dataset, with specific visualization of the NSCLC compartment and IREB2 expression distribution. (L) Dot plot showing IREB2 and B cell function/prognosis‐related gene expression in pan‐cancer and lung cancer B memory subsets. (M) Hazard ratios (HR) with 95% confidence intervals for a gene set evaluated in lung cancer, LUAD, and LUSC using KM analysis. Data are presented as bar chart with error bars depicting 95% CIs. Prognostic markers colored blue indicate favorable outcomes. (N) UMAP illustrates B memory cell subsets within the integrated pan‐cancer cycling B cell scRNA‐seq dataset, with specific visualization of the NSCLC compartment and IREB2 expression distribution. (O) Dot plot illustrating IREB2 and B cell function/prognosis‐linked gene expression in pan‐cancer versus lung cancer B cycling subsets. Data visualization (I, J, K, L, N, O) partly generated by pan‐B (http://pan‐b.cancer‐pku.cn/).

FIGURE 7

IREB2 exhibits regulatory potential for both intrinsic B cell functions and antigen‐presenting capabilities. (A) Multiplex immunofluorescence demonstrates the expression patterns of CD19, IREB2, and CD27 on lung cancer tissue microarrays. Red: IREB2; green: CD19; yellow: CD27. (B) The proportions of CD19⁺, CD27⁺, and IREB2⁺ cells and the CD19/CD27, IREB2/CD27 ratios across histological grades of lung tumors. (C) Forest plots show prognostic correlations of CD19⁺/CD27⁺/IREB2⁺ cell proportions and CD19/CD27, IREB2/CD27 ratios with survival by Kaplan–Meier analysis. (D) GSEA results showcase KEGG, Reactome, and Hallmark pathway enrichment in si‐IREB2 versus si‐NC in SU‐DHL‐4 cells from RNA sequencing. (E) UMAP visualization of B cells (left), scPAS‐derived risk scores (middle), and scPAS‐selected cells (right) from integration of B cells and plasma cells from COPD and lung cancer scRNA‐seq datasets, comparing si‐NC group versus si‐IREB2 group. Colors reflect risk score gradients (red: high; blue: low). (F) Distribution of scPAS+ and scPAS− cells across B cell subpopulations. (G) Violin plots depict expression distributions of CD19, FCRL4, MS4A1, CD80, CD86, FAS, CD27, IGHA1, IGHG1, and HLA‐DQA1 across scPAS⁺ and scPAS⁻ cell populations.​ (H) Western blot analysis showing changes in protein expression levels of CD19, CD27, and HLA‐DPB1 in SU‐DHL‐4 cells following CSE stimulation and IREB2 knockdown (n = 5). Multiple t‐tests. Data are presented as mean ± SD.