Abstract
BACKGROUND: Airborne microplastic polyethylene terephthalate (PET) accumulates in human lungs and is linked to respiratory pathologies; however, its molecular role in lung adenocarcinoma (LUAD) remains unclear. This study aims to explore the potential carcinogenic mechanisms of PET exposure in LUAD. METHODS: We integrated single-cell RNA sequencing, machine learning algorithms [including Classification and Regression Trees (CART), Naïve Bayes (NB), random forest (RF), and support vector machine (SVM)], molecular docking, survival analysis, and multi-omics data. Through differential expression screening across datasets combined with Venn analysis, we identified seven PET-associated oncogenic targets. Seven PET-associated oncogenic targets were identified via differential gene screening and Venn analysis. RESULTS: MYL9 was validated as a downregulated, LUAD-protective biomarker associated with significant survival benefit [hazard ratio (HR) =0.59, 0.16, 0.23; all P<0.05]. These findings were consistent across the Human Protein Atlas (HPA) database, co-expression networks, and three independent LUAD datasets. SPI1 was identified as a key transcriptional regulator, showing strong co-expression with MYL9 (R=0.556, P<0.05) and concurrent downregulation in LUAD. Molecular docking revealed that PET bound to the DNA-binding pocket of SPI1 (ΔG =-5.30 kcal·mol-1), suggesting its transcriptional inhibition of MYL9. CONCLUSIONS: Our integrated bioinformatics approach supports a novel "PET-SPI1-MYL9" transcriptional axis, revealing a potential non-genotoxic carcinogenic pathway for PET. While the computational evidence is robust, further wet-lab experiments are needed to validate the binding and transcriptional inhibition mechanism. This model provides a framework for understanding airborne microplastic toxicity in LUAD. We propose that PET promotes LUAD by disrupting the SPI1-MYL9 transcriptional axis, highlighting a potential environmental trigger and candidate targets for diagnostic and therapeutic strategies.