Reaction network model for the prediction of mammalian metabolism of benzo[a]pyrene.

摘要

Humans are exposed to mixtures of environmental pollutants on daily bases. Many of these chemicals undergo biotransformation in our body and often produce toxic metabolites. The biotransformation of mixtures involves complex reaction networks that are difficult to study using conventional experimental techniques. As a first step of developing a predictive tool for the biotransformation of chemical mixtures, a chemical engineering approach, Reaction Network (RN) modeling, was utilized to study the mammalian metabolism of benzo[ a]pyrene (BaP), a priority environmental carcinogen. A RN pathway model which predicts the theoretically possible reaction network for BaP was first developed based on the existing modeling technology for predicting the reaction networks in petroleum refinery processes, mechanistic organic chemistry, as well as the commonly observed biochemical reactions for mammalian metabolism of BaP. The resulting RN pathway model for BaP predicts that 246 reactions can occur, resulting in unique 150 products in the presence of mammalian cytochrome P450 and epoxide hydrolase. Some of these predicted products might not be experimentally detected due to the slow reactions for their formation or the production of reactive species. A RN kinetics model which reflects the experimentally measurable metabolic pathways was then established to determine the reaction rates of BaP metabolism. To obtain proper separation of eleven BaP metabolites with high detection sensitivities, high-performance liquid chromatography methods were developed and validated. The RN kinetics model was calibrated and validated using experimental data of BaP metabolism catalyzed by recombinant human enzymes. The biotransformation of BaP and the production of nine BaP metabolites were accurately described by the RN kinetics model. Finally, the RN kinetics model of BaP was linked to a physiologically based pharmacokinetic (PBPK) model to describe the distribution and disposition of BaP and its metabolites in rats. The major advantages of applying RN modeling to study toxicology are: (1) their capabilities of handling complex metabolic systems; (2) their potential for predicting reaction networks of chemicals with limited knowledge on their metabolic pathways; and (3) their abilities to predict the reactive intermediates that are not readily measurable in experiments.