Abstract:Although nanotoxicology studies have shown that respiratory exposure of titanium dioxide nanoparticles (TiO2 NPs) could induce adverse health effects, limited biomarkers associated with occupational exposure of TiO2 NPs were reported. The purpose of this study is to screen serum biomarkers among workers occupationally exposed to TiO2 NPs using metabolomics. Compared with the control group, a total of 296 serum metabolites were differentially expressed in the TiO2 NPs-exposed group, of which the relative expression of 265 metabolites increased, and the remaining 31 decreased. Three machine learning methods including random forest (RF), support vector machines (SVM), and boruta screened eight potential biomarkers and simultaneously selected a metabolite, Liquoric acid. Through multiple linear regression analysis to adjust the influence of confounding factors such as gender, age, BMI, smoking and drinking, occupational exposure to TiO2 NPs was significantly related to the relative expression of the eight potential biomarkers. Meanwhile, the receiver operating characteristic curves (ROCs) of these potential biomarkers had good sensitivity and specificity. These potential biomarkers were related to lipid peroxidation, and had biological basis for occupational exposure to TiO2 NPs. Therefore, it was demonstrated that the serum metabolites represented by Liquoric acid were good biomarkers of occupational exposure to TiO2 NPs.

Metabolomics Screening Of Serum Biomarkers For Occupational Exposure Of Titanium Dioxide Nanoparticles