Integrating Network Pharmacology, Machine Learning, and Experimental Validation to Elucidate the Mechanism of Cardamonin in Treating Idiopathic Pulmonary Fibrosis

Wenyue Zhang, Yi Guo, Qiushi Wang, Kai Wang, Huning Zhang, Sirong Chang, Anning Yang, Zhihong Liu, Yue Sun

Idiopathic pulmonary fibrosis (IPF) is a chronic and irreversible interstitial lung disease characterized by progressive scarring of the lungs. The available therapeutic strategies are limited and primarily focus on slowing disease progression rather than achieving fibrosis reversal. Cardamonin (CDN), a food-derived natural chalcone, has exhibited anti-fibrotic activity in liver and kidney fibrosis models; however, its role and underlying mechanism in IPF remain unelucidated. Herein, we integrated network pharmacology, machine learning, molecular simulations, and in vitro experiments. Network pharmacology identified 135 overlapping targets between CDN and IPF, which demonstrated a significant enrichment in the Phosphatidylinositol 3-Kinase/Protein Kinase B signaling pathway (PI3K/AKT). Machine learning further prioritized 6 core targets, with IGF1 emerging as a key candidate. Molecular docking revealed a favorable binding energy of −7.9 kcal/mol for the CDN-IGF1 complex. Subsequent 100 ns molecular dynamics simulations further confirmed its robust binding stability, yielding a mean binding free energy of −150.978 kcal/mol. In vitro, CDN significantly mitigated fibrosis in bleomycin (BLM)-challenged A549 cells, downregulating the expression of α-smooth muscle actin (α-SMA) and fibronectin. This effect was accompanied by a beneficial reversal of epithelial–mesenchymal transition (EMT), as indicated by increased E-cadherin levels and decreased vimentin expression. Mechanistically, CDN significantly suppressed the IGF1/PI3K/AKT axis; this inhibitory effect was partially reversed by exogenous IGF1 supplementation and further enhanced by the PI3K-specific inhibitor LY294002. This work provides the evidence that CDN alleviates BLM-induced pulmonary fibrosis by targeting the IGF1/PI3K/AKT-EMT axis. These findings lend support to a robust mechanistic basis for developing CDN as a potential therapeutic candidate for IPF. It should be noted that these conclusions are drawn from in vitro experiments using A549 cells, and further validation in primary alveolar epithelial cells and animal models is warranted to confirm their physiological relevance.