Computational AOP modeling links 1-nitropyrene exposure to asthma exacerbation and COPD progression

Nitropyrene (1-NP), a PM2.5-bound nitro-PAH, is a high-priority air pollutant with genotoxic and inflammatory potential, but the cascade linking exposure to asthma exacerbation and COPD progression remains unclear. We developed a multiscale computational respiratory toxicology pipeline integrating network toxicology, clinical transcriptomics, machine-learning feature selection with SHAP (LASSO, SVM-RFE), immune deconvolution (CIBERSORT), single-cell validation, protein–ligand docking, 100-ns molecular dynamics with MM/PBSA, and single-cell virtual knockout (scTenifoldKnk). Outputs were synthesized into an OECD-aligned computational pathway for adverse outcome prediction (cpAOPs) using AOP-HelpFinder 3.0. Nine core targets were linked to 1-NP–associated asthma exacerbation (CTNNB1, CXCL2, CXCL3, EHD1, GPER1, IFRD1, MSR1, SLC7A5, TFRC) and five to COPD progression (TP53, EPHX1, THBS1, LTB4R, LTF). These signatures discriminated severity (asthma AUC 0.824–0.969; COPD AUC ≥ 0.70), with CTNNB1 and TP53 showing the highest SHAP contributions. Asthma severity was associated with increased CD8 + T cells, resting CD4 memory T cells and mast cells, and reduced macrophage subsets. Single-cell data showed broad expression of asthma targets across macrophage, neutrophil, endothelial and epithelial compartments, whereas COPD targets were more cell-type restricted. Docking/MD supported stable multi-target binding (MM/PBSA ΔG − 12.43 to − 47.03 kcal/mol). Virtual knockout implicated IL-17/Th differentiation and NF-κB/MAPK–cytokine signaling in asthma, and oxidative stress, p53 responses, senescence/autophagy and cytokine amplification in COPD. This OECD-consistent cpAOPs framework links 1-NP exposure to airway outcomes, providing mechanistically anchored biomarkers for severity stratification and regulatory-grade risk assessment.