Multiscale physiologically based toxicokinetic (PBTK) modeling: A case study for arsenic compounds
S.K. Stamatelos, C.J. Brinkerhoff, A.F. Sasso, S.S. Isukapalli, P.G. Georgopoulos
Environmental and Occupational Health Sciences Institute, Piscataway, NJ
Whole-body Physiologically Based Toxicokinetic (PBTK) models are used to estimate target tissue doses of toxicants and their metabolites following exposures to environmental toxicants. Development of human PBTK is usually hindered by availability to data from in vivo experiments (biomarkers); for the reason it is important to utilize additional information from in vitro to in vivo, and from cross-species extrapolations.
Information supporting the understanding of toxicokinetics of xenobiotics is available at multiple scales. These include in vivo biomarker data reflecting organism exposure to the chemical of interest, in vitro experiments involving tissue slices, and in vitro experiments involving cell cultures, as well as computational chemistry tools for estimating biochemical properties of xenobiotics using molecular-scale descriptors. However, usually there is significant inconsistency between models developed using data available at different scales. Hence, there is a need for a multiscale integrated approach for incorporating information from different scales, into a comprehensive mechanistic model of xenobiotic–organism interactions.
Arsenic is studied here as a model chemical since it is a potent human toxicant with a multiplicity of health effects associated with both acute and chronic exposures. Arsenic is present in environmental and microenvironmental media in many forms, with varying degrees of bioavailability. Inorganic arsenicals are the predominant forms found in water, while a wide range of organic species is detected in seafood and other dietary components. Intake from food and from water account for the most significant environmental arsenic exposures.
The present modeling effort focuses on refining PBTK modeling of arsenic using data from in vitro experiments. Since liver is the primary organ of biotransformation of arsenic and hepatocytes constitute the primary part of its cytoplasmic mass, this study focuses on incorporating data from hepatocytes exposed to arsenic into a whole-body PBTK model for arsenic. This provides a first step towards resolving inconsistencies between in vitro cellular-level data and whole-body PBTK models.
A cellular-level model has been developed here that describes the kinetics of various arsenic species, including arsenite (iAsIII), monomethylated As (MMA) and dimethylated As (DMA), inside hepatocytes and across the cellular membrane. This model considers uptake of iAsIII by hepatocytes via aquaporin isozymes 9 (AQP9s) and metabolism through a series of methylation and glutathione conjugation reactions. Efflux of arsenic glutathione adducts (ATG, MADG) is considered saturable via multidrug resistant proteins (MRPs). Methylation reactions of arsenic are assumed to follow Michaelis-Menten kinetics with uncompetitive inhibition. All the other biotransformation processes are modeled through linear kinetics. The parameterization of this model utilizes experimental data from human hepatocytes exposed to arsenite in vitro at various doses. They are estimated using Bayesian Markov Chain Monte Carlo (MCMC) simulations and reflect inter-individual variation as distributions of metabolic parameters across individuals. This variation reflects different arsenic methylation patterns among individuals due to genetic and epigenetic factors. Such a modeling description will enhance the current understanding of relative roles of transport and biotransformation reactions in the production of various arsenic metabolites with different toxicological potency. Major challenges and issues associated with the integration of the hepatocyte toxicokinetic model for arsenic compounds with a whole-body PBTK model are highlighted; while a framework for the systematic assessment of mechanistic consistency across multiple scales (molecular, cellular, tissue, organ, and organism) is described.