![[logo] US EPA](http://www.epa.gov/epafiles/images/logo_epaseal.gif){kind=logo}
Computational Toxicology Research Program
Background: EPA’s Office of Research and Development (ORD) has initiated a research program on Computational Toxicology, http://epa.gov/comptox, to better understand the relationships between sources of environmental pollutant exposure and adverse outcomes. Abbreviated ‘CompTox’ – it is the integration of computing and information technology with the technologies of molecular biology and chemistry and is used to improve EPA’s prioritization of data requirements and risk assessments for toxic chemicals. Strategic objectives of CompTox are to: (1) improve understanding of the linkages in the continuum between the source of a chemical in the environment and adverse outcomes, (2) provide predictive models for screening and testing and (3) improve quantitative risk assessment. ORD has other closely-related activities and an FY04 Accomplishment Report will be posted soon at http://epa.gov/comptox.
Role of Physiologically-Based Pharmacokinetic and Pharmacodynamic (PBPK/PD) Research: The overall success of ORD’s CompTox initiative is dependent on the development and coupling of new computational methods, for example, Quantitative Structure-Activity Relationships (QSARs) with high-throughput “omic” technologies (e.g., genomic, proteomic and metabonomic). The goal is to contribute to ORD’s next-generation of PBPK/PD models by (1) developing methods to predict absorption, distribution, metabolism and elimination (ADME) and toxic effects on the basis of physicochemical properties via QSARs and (2) developing advanced representations of organs and tissues. In addition to improving PBPK/PD models, the Exposure and Dose Research Branch (EDRB) is working on a Daphnia Pilot Project (see details below) to determine the utility of using data from high-throughput “omic” technologies in support of risk assessment through their application in PBPK/PD modeling.
Integration of Mechanistic QSARs within a PBPK/PD Model Framework: The use of QSARs is one way to enhance the predictive capability of PBPK/PD modeling. In the QSAR approach, statistical correlations are found between chemical structure indices and chemical activity (or behavior) in biological systems. These correlations are then used to assign parameter values to new chemicals for which activity data are not available. Traditionally, QSARs have been used to estimate coarse physicochemical properties, but the potential exists to establish correlations with the important mechanisms related to ADMEs, e.g., mechanistic QSARs. The integration of mechanistic QSARs within a PBPK/PD model framework supports the overarching goal of EPA’s Computational Toxicology initiative: to enable health risk assessment of chemicals without the need for extensive in vivo animal testing of each new chemical.
Computational Virtual Organs: Computational virtual organs which include robust details of metabolic reactions, transport and biochemical transformations will improve the prediction of ADME and toxic effects of a PBPK/PD model. A virtual liver and virtual blood-brain barrier are being developed because of their importance in chemical modulation/metabolism and distribution, respectively. Ultimately, these virtual organs can be integrated into a whole-organism PBPK modeling framework such as the Exposure-Related Dose-Estimating Model (ERDEM).
Liver Model: A computational virtual liver contains a database of the known interactions between the various enzymes, such as cytochrome P450s, and chemical substrates, as well as the known inducers and inhibitors for each isozyme. The database would also have values corresponding to the estimated titer of each isoform within the liver. When a substrate is introduced into the virtual liver in the PBPK/PD model, the program would match the substrate to the isoform which metabolizes it, to derive an estimate of the kinetics of the metabolism. Other factors that affect the rate of isoform production by induction or inhibition would be included in the modeled liver and would influence the metabolic rate parameter values. In this manner, genomic information, such as functional effects of a gene polymorphism that affect the rate of enzyme production, can be directly incorporated within the model to modulate the production of a metabolite, according to the distribution of Vmax in a target population (Vmax is the maximum speed at which an enzymatic reaction can go when saturated with substrate).
Blood-Brain Barrier Model: In this project, development is proceeding on a predictive computer model of chemical transport to the brain that includes the blood-brain barrier, blood-cerebrospinal fluid barrier, and the brain parenchyma. The permeability of chemicals to the brain by passive diffusion has been generally observed to be a function of molecular size and lipophilicity, but the correlations lose predictive power when selective active-transport mechanisms are involved. The prediction of chemical penetration to the brain is difficult due to the presence of selective active-transport systems and heterogeneity of the tissue in contact with the circulation. The mechanistic computer model of brain transport will include passive diffusion as well as selective active-transport systems. So the contribution of each transport mechanism can be simulated for a chemical concentration in the blood to predict the penetration to the central nervous system.
Daphnia Pilot Project: The goal of the Daphnia pilot project is to improve the fundamental understanding of the relationships among currently disparate exposure, dose and toxicity data in animal systems. Non-invasive “omics” (metabonomics, proteomics, genomics) techniques can provide a holistic view of the biochemical status of an organism.
 |
Metabonomics provides the entire complement of all the small molecular weight metabolites inside a sample of interest (e.g., urine, buccal cell homogenate) determined using nuclear magnetic resonance (NMR) spectroscopy. | |
|
Proteomics characterizes the dynamics of protein function on a global scale. Proteomics relying on two-dimensional (2-D) gel electrophoresis of proteins, followed by spot identification with mass spectrometry, is an excellent experimental tool for chemical-effects studies since many environmental chemicals interact directly with cellular proteins to modify protein function and interactions. | |
|
Genomics data can be used to find correlations between responses to chemicals and changes in the genetic profiles of animals exposed to such chemicals. The expression of literally thousands of genes can be simultaneously evaluated using microarray technology. |
However, these above technologies are not currently being utilized in predictive toxicology because their relationships to standard measurable toxicological end-points are unclear or unknown altogether. It is imperative that we understand how the various endpoints of toxicological evaluation are related in order to know which endpoints are representative of known PBPK/PD pathways. Extracting information from this data will enable better understanding in the modeling of biological systems (from biochemical interactions to physiological responses) and to increased utilization of these models for diagnostic and predictive toxicology.
Partnerships: Research partnerships with universities, Federal agencies, and industry are imperative for the success of this research program. Through collaborative efforts, data- and knowledge-bases which are currently compiling requisite data will be approached for collaboration and partnerships. Existing and planned partnerships include:
| |
U.S. Department of Agriculture, Agricultural Research Service |
|
National Center for Toxicogenomics, National Institute of Environmental Health Sciences, Biomedical Sciences Division, Imperial College, London |
|
U.S. Department of Energy, Sandia and Pacific Northwest Laboratories |
|
Environmental Health Sciences Center, Oregon State University |
|
U.S. Food and Drug Administration, National Center for Toxicological Research |
|
Armed Forces Institute of Pathology, Department of Environmental and Toxicological Pathology |
|
Daphnia Genomics Consortium |
|
Iconix Pharmaceuticals |
|
Cellnomica, Inc. |
|
University of North Texas, Developmental Physiology and Genetics Research Group |
|
Indiana University School of Medicine, Biotechnology and Research Training Center |
For more information contact:
Dr. Curtis C. Dary
U.S. Environmental Protection Agency/ORD/NERL
Human Exposure & Atmospheric Sciences Division
Exposure and Dose Research Branch
944 East Harmon Avenue
Las Vegas, NV 89119-6748
Dary.Curtis@epa.gov
(702)798-2286