Humans are continuously exposed to arsenic, as it is very often present in drinking water (typically as an inorganic ion) as well as in various foods (mostly in organic forms) [1]. Arsenic is a known carcinogen and its ability to induce various types of cancer, such as skin, liver, bladder and lung cancer has been demonstrated in various epidemiological studies involving the inhalation and ingestion exposure routes. Arsenic causes oxidative stress by increasing production of Reactive Oxygen Species (ROS) [2]. Furthermore, low doses of arsenic in mice during gestation have been associated with gene expression changes related to steroid pathways as well as with liver cancer in the adult stage [3].

Inorganic arsenic in water is mainly present in the form of arsenate (iAsV), which is uptaken slowly by cells, as it is negatively charged at physiological pH. Arsenate in the blood is quickly converted to arsenite (iAsIII) [4].

The goal of the effort presented here is to examine in silico specific hepatic effects of arsenic exposures. The first step is the quantitative characterization of the time-course dynamics of different chemical forms of arsenic within human hepatocytes. This is pursued through the development of a “cellular level pharmacokinetic (PK) model” that simulates the kinetics of various “arsenic species,” i.e. arsenite (iAsIII), monomethylated As (MMA) and dimethylated As (DMA), inside hepatocytes. The PK model considers iAsIII uptake by hepatocytes via aquaporin isozymes 9 (AQP9's) [5] and metabolism through a series of methylation and glutathione conjugation reactions [6]. MMAIII is effluxed via AQP9's [7] while DMA simply diffuses across the cell membrane. Glutathione adducts of arsenicals are effluxed via multidrug resistant proteins 1/2 (MRP's) [8]. The parameterization of the model utilizes primarily experimental data from human hepatocytes exposed to arsenite in vitro [9]. However, the kinetic structure employed in this model is also able to replicate experimental results involving mouse hepatocytes exposed to arsenic [10]. Bayesian Markov Chain Monte Carlo (MCMC) simulations and systematic sensitivity runs have been conducted in order to develop estimates of the distributions of model parameters and to characterize essential features of the behavior of the model under different conditions. These parameterizations have then been used in evaluating the PK model with respect to dose-extrapolations, and for incorporation into whole-body Physiologically Based Toxicokinetic (PBTK) models of arsenic.

On-going and future components of this research effort involve the coupling of this PK model with various endpoint-specific pharmacodynamic (PD) models for the cellular-level effects of arsenic species. One specific initial objective within this effort involves the development of hypotheses and corresponding computational models, such as indirect response PD models [11], that will link the levels of arsenic species within hepatocytes, and especially of the methylated arsenic forms, with ROS profiles and induced DNA damage. This PK/PD modeling approach is going to provide linkages among exposure to a toxicant, target tissue dosimetry and effects in the subcellular level.

References

1. Georgopoulos, P.G., et al., Biologically based modeling of multimedia, multipathway, multiroute population exposures to arsenic. J Expo Sci Environ Epidemiol, 2007.

2. Qian, Y., et al., Cdc42 regulates arsenic-induced NADPH oxidase activation and cell migration through actin filament reorganizatlion. Journal of Biological Chemistry, 2005. 280(5): p. 3875-3884.

3. Liu, J., et al., Global gene expression associated with hepatocarcinogenesis in adult male mice induced by in utero arsenic exposure. Environmental Health Perspectives, 2006. 114(3): p. 404-411.

4. Vahter, M. and E. Marafente, Reduction and binding of arsenate in marmoset monkeys. Archives of Toxicology, 1985. 57(2): p. 119-124.

5. Liu, Z.J., et al., Arsenite transport by mammalian aquaglyceroporins AQP7 and AQP9. Proceedings of the National Academy of Sciences of the United States of America, 2002. 99(9): p. 6053-6058.

6. Hayakawa, T., et al., A new metabolic pathway of arsenite: arsenic-glutathione complexes are substrates for human arsenic methyltransferase Cyt19. Archives of Toxicology, 2005. 79(4): p. 183-191.

7. Liu, Z.J., M. Styblo, and B.P. Rosen, Methylarsonous acid transport by aquaglyceroporins. Environmental Health Perspectives, 2006. 114(4): p. 527-531.

8. Kala, S.V., et al., Formation and urinary excretion of arsenic triglutathione and methylarsenic diglutathione. Chemical Research in Toxicology, 2004. 17(2): p. 243-249.

9. Styblo, M., et al., Metabolism of arsenic in primary cultures of human and rat hepatocytes. Chemical Research in Toxicology, 1999. 12(7): p. 560-565.

10. Kedderis, G.L., et al., Kinetics of arsenic methylation by freshly isolated B6C3F1 mouse hepatocytes. Chemico-Biological Interactions, 2006. 161(2): p. 139-145.

11. Sharma, A. and W.J. Jusko, Characteristics of indirect pharmacodynamic models and applications to clinical drug responses. British Journal of Clinical Pharmacology, 1998. 45(3): p. 229-239.