Exploring the molecular mechanisms underlying the impact of plasticizer exposure on keloid: insight from network toxicology and multi-omics analyses

Introduction The potential association between phthalate plasticizers and keloid (KD), characterized by abnormal fibrous hyperplasia after skin healing, still remains ambiguous. Methods Herein, this study incorporated network toxicology, multi-omics sequencing, and molecular docking to investigate the potential molecular associations between phthalate plasticizer-related targets and keloid. Results & conclusions First, screening the GEO, CTD, ToxCast, ChEMBL, and STITCH databases against 278 plasticizer targets (PTs) identified 4 compounds with skin toxicity (diethyl phthalate DEP, dimethyl phthalate DMP, dioctyl phthalate DOP, tributyl O-acetylcitrate ATBC). Subsequently, 21 candidate genes were determined by cross-referencing keloid-related GeneCards and the OMIM database with predicted targets, which are mainly enriched in the Ribosome, Olfactory Transduction, and so on. Afterwards, by employing the machine learning algorithm LASSO, which was used for dimensionality reduction, overfitting prevention and predictive factor selection, two hub genes potentially associated with KD-related molecular alterations were identified, namely ADAMTS4 and RGS4. In addition, immune infiltration analysis revealed a remarkably positive correlation between ADAMTS4 and monocytes (cor = 0.68, p < 0.01). Molecular docking analyses predicted potential in silico binding affinities between RGS4 and DEP/DMP, with binding energies of −5.8 and −5.2 kcal/mol, respectively. Single-cell RNA sequencing further suggested that smooth muscle cells may represent a potentially relevant cell population characterized by relatively high ADAMTS4 expression, whereas RGS4 showed minimal expression. To conclude, our study systematically explored a putative network linking plasticizer-related targets with keloid, providing preliminary computational clues for future mechanistic validation.