You are here:
Application of Biologically-Based Lumping To Investigate the Toxicological Interactions of a Complex Gasoline Mixture
Citation:
Jasper, M., S. Martin, W. Oshiro, J. Ford, P. Bushnell, AND H. El-Masri. Application of Biologically-Based Lumping To Investigate the Toxicological Interactions of a Complex Gasoline Mixture. ENVIRONMENTAL SCIENCE & TECHNOLOGY. American Chemical Society, Washington, DC, 50(6):3231-3238, (2016).
Impact/Purpose:
Most experimental efforts in toxicology are focused on single chemicals or small discrete mixtures. Yet humans are often exposed to large complex mixtures, such as gasoline, diesel, tobacco smoke, combustion exhaust, or drinking water, which can often consist of hundreds of chemicals. These large complex mixture are difficult to investigate experimentally and computationally. The Biologically Based Lumping Method (BBLM) presented in this research is one of the first examples of applying clustering algorithms and biologically based chemical lumping to a complex mixture, specifically gasoline. The gasoline mixture sampled in this research consisted of 109 identified and quantified chemicals. Using our computational methods, it is shown that there was a minimal difference between simulating a simplified mixture with 10 target chemicals and 15 lumps and simulating all 109 interacting chemicals. Though gasoline was specifically considered in this research, the lumping method presented can be applied to any kind of complex mixture to lump the chemicals and simplify the mixture based on biologically relevant chemical properties.
Description:
People are often exposed to complex mixtures of environmental chemicals such as gasoline, tobacco smoke, water contaminants, or food additives. However, investigators have often considered complex mixtures as one lumped entity. Valuable information can be obtained from these experiments, though this simplification provides little insight into the impact of a mixture's chemical composition on toxicologically-relevant metabolic interactions that may occur among its constituents. We developed an approach that applies chemical lumping methods to complex mixtures, in this case gasoline, based on biologically relevant parameters used in physiologically-based pharmacokinetic (PBPK) modeling. Inhalation exposures were performed with rats to evaluate performance of our PBPK model. There were 109 chemicals identified and quantified in the vapor in the chamber. The time-course kinetic profiles of 10 target chemicals were also determined from blood samples collected during and following the in vivo experiments. A general PBPK model was used to compare the experimental data to the simulated values of blood concentration for the 10 target chemicals with various numbers of lumps, iteratively increasing from 0 to 99. Large reductions in simulation error were gained by incorporating enzymatic chemical interactions, in comparison to simulating the individual chemicals separately. The error was further reduced by lumping the 99 non-target chemicals. Application of this biologically-based lumping method resulted in a simplification of the gasoline mixture into 15 chemical lumps and 10 target chemicals that closely approximated the interaction effects of the entire complex mixture.
