Citation:

JONES, W. J., P. K. SCHMIEDER, R. C. KOLANCZYK, AND O. MEKENYAN. SIMULATING METABOLISM OF XENOBIOTIC CHEMICALS AS A PREDICTOR OF TOXICITY. Presented at BOSC Review of the Computational Toxicology Program, Research Triangle Park, NC, June 19 - 20, 2006.

Impact/Purpose:

This task is divided into four major research areas: (1) Development of computational tools and databases for screening-level modeling of the environmental fate of organic chemicals; (2) Metabolism of xenobiotics: Enhancing the development of a metabolic simulator; (3) Metabonomics: The use of advanced analytical tools to identify toxicity pathways; and (4) Software infrastructure to support development and application of transformation/metabolic simulators.

For many chemicals, multiple transformation/metabolic pathways can exist. Consequently, transformation/metabolic simulators must utilize transformation rate data for prioritization of competing pathways. The prioritization process thus requires the integration of reliable rate data. When this data is absent, it is necessary to generate a database with metabolic and transformation rate constants based on: (1) experimentally measured values, including those requiring the use of advanced analytical techniques for measuring metabolic rate constants in vivo and in vitro; (2) rate constants derived from SPARC and mechanistic-based QSAR models; and (3) data mined from the literature and Program Office CBI. A long-term goal of this project is to build this database. This information will be used to enhance the predictive capabilities of the transformation/metabolic simulators. As indicated previously, exposure genomics, which provide early signs of chemical exposure based on changes in gene expression, will be used to guide chemical fate and metabolism studies. The incorporation of exposure genomics into fate studies will provide information concerning (1) the minimal concentrations at which biological events occur; and (2) the identification of biologically relevant chemicals(s) in mixtures.

The capability of categorizing chemicals and their metabolites based on toxicity pathway is imperative to the success of the CompTox Research Program. Metabonomics, which is the multi-parametric measurement of metabolites in living systems due to physiological stimuli and/or genetic modification, provides such a capability. The application of metabonomics to toxicity testing involves the elucidation of changes in metabolic patterns associated with chemical toxicity based on the measurement of component profiles in biofluids, and enables the generation of spectral profiles for a wide range of endogenous metabolites. Metabolic profiles can provide a measure of the real outcome of potential changes as the result of xenobiotic exposure.

Description:

EPA is faced with long lists of chemicals that need to be assessed for hazard. A major gap in evaluating chemical risk is accounting for metabolic activation resulting in increased toxicity. The goals of this project are to develop a capability to forecast the metabolism of xenobiotic chemicals of EPA interest, to predict the most likely formed metabolites, and to interface that information with toxic effect models. Results will identify metabolites of equal or greater toxicity than the parent chemical. An existing metabolism simulator is being refined by focusing on reactions leading to increased toxicity for the effects endpoint of concern and which are currently simulated with low reliability. The toxic effect endpoint considered is endocrine disruption mediated by direct chemical binding to the estrogen receptor (ER). The principal reactions under investigation, ring oxidation and O-dealkylation, are those that result in hydroxylated metabolites predicted to bind the ER with greater affinity than the parent chemical. Chemicals for study were selected from EPA concern lists (food use inerts) provided by the Office of Prevention, Pesticides, and Toxic Substances (OPPTS) and include chemicals not predicted to be estrogenic as parent but for which forecasted metabolites are predicted to bind ER. Concurrent research will expand and enhance a QSAR for prediction of chemical binding to the ER as a means of assessing potential estrogenicity of parent chemicals and forecasted metabolites. Model results will be used to prioritize chemicals for further study and provide transformation reliability estimates.

To enhance the simulator performance, chemical metabolism maps will be collected from the published literature and determined from in vitro rat hepatic microsome experiments. Newly acquired maps (and transformations) will be used to re-train the metabolism simulator and improve reliability estimates. In conjunction with metabolism experiments, analytical methods used to verify bioactivated metabolites formed in rat microsomes will be optimized to detect metabolites formed in fish liver tissue slices. Metabolically-competent liver slices from male fish that produce vitellogenin when exposed to xenobiotics will be used to study chemical bioactivation to ER-active forms. Data from these studies are used to improve the metabolic simulator and prioritize chemicals for testing that have the potential to be bioactivated to more toxic species. Metabolism data for training and improvement of the simulator will be stored and accessed using a database manager software under development. In addition, the database software is capable of search functions, depiction of metabolic maps, and provides access via structures to metabolism information and associated data collected from EPA's Office of Pesticide Programs (OPP). The database is specifically designed to store information submitted to EPA in support of pesticide registration such as pesticide metabolism in laboratory animals, plants, and livestock as well as environmental degradates. The database will be used by OPP risk assessors to increase efficiency of metabolism data access and performance of risk assessments. The system includes the ability to: depict hierarchical connection sequences of parent chemical and metabolites; track radiolabel within a pathway and combine/separate maps from associated studies; identify all maps (and parent structure) containing a specific metabolite of concern; search for sub-structures of toxicological concern and indicate its presence across chemicals and maps; and compare map similarity across chemicals and species. The database contains associated chemical identifiers as well as bioassay and analytical chemistry data. In its simplest mode, the database will furnish curated (verified) structures of chemicals/pesticides and their metabolites suitable for searches in other databases and provide metabolic maps plus tabulations of amounts of metabolites and other parameters. In a more advanced mode, the database will allow the risk assessor to perform searches for specific compounds and toxicophores and identify metabolism commonalities and differences across pesticides and species. At present, this process is performed manually and the results may vary among individual risk assessors. The database/software is expected to make this process more efficient, to furnish more reliable results, and to serve as a tool for hypothesis formulation by EPA researchers.

Finally, prioritized chemical lists (based upon predicted toxic effects of parent chemical and metabolites) with transformation reliability estimates will be provided to OPPTS for chemical evaluations (risk assessments) and ranking for toxicity testing. This research will expand the knowledge-base of metabolic pathways and transformation products for important groups of toxic chemicals and demonstrate an approach that integrates metabolism simulation with toxic effects modeling.

Record Details: