MetaPath: an electronic knowledge base for collating, exchanging and analyzing case studies of xenobiotic metabolism.

The MetaPath knowledge base was developed for the purpose of archiving, sharing and analyzing experimental data on metabolism, metabolic pathways and crucial supporting metadata. The MetaPath system grew out of the need to compile and organize the results of metabolism studies into a systematic database to facilitate data comparisons and evaluations. Specialized MetaPath data evaluation tools facilitate the review of pesticide metabolism data submitted for regulatory risk assessments as well as exchange of results of complex analyses used in regulation and research. Customized screen editors called Composers were developed to automate data entry into MetaPath while also streamlining the production of agency specific study summaries such as the Data Evaluation Records (DER) used by the US EPA Office of Pesticide Programs. Efforts are underway through an Organization for Economic Co-operation and Development (OECD) work group to extend the use of DER Composers as harmonized templates for rat metabolism, livestock residue, plant residue and environmental degradation studies.

2D-3D migration of large chemical inventories with conformational multiplication. Application of the genetic algorithm.

Mathematical chemistry has afforded a variety of research areas with important tools to understand and predict the behavior of chemicals without having to consider the complexities of three-dimensional conformations of molecules. Predictive toxicology, an area of increasing importance to toxicity assessments critical to molecular design and risk management, must be based on more explicit descriptions of structure, however. Minimum energy conformations are often used for convenience due, in part, to the difficulty of computing a representative population of conformers in all but rigid structures. Such simplifying assumptions fail to reveal the variance of the stereoelectronic nature of molecules as well as the misclassification of chemicals which initiate receptor-based toxicity pathways. Because these errors impact both the success in discovering new lead and the identification of possible hazards, it is important that mathematical chemistry develop additional tools for conformational analysis. This paper presents a new system for automated 2D-3D migration of chemicals in large databases with conformer multiplication. The main advantages of this system are its straightforward performance, reasonable execution time, simplicity and applicability to building large 3D chemical inventories. The module for conformer multiplication within the 2D-3D migration system is based on a new formulation of the genetic algorithm for computing populations of possible conformers. The performance of the automated 2D-3D migration system in building a centralized 3D database for all chemicals in commerce worldwide is discussed. The applicability of the 3D database in assessing the impact of molecular flexibility on identifying active conformers in QSAR analysis and assessing similarity between chemicals is illustrated.

Use of trout liver slices to enhance mechanistic interpretation of estrogen receptor binding for cost-effective prioritization of chemicals within large inventories.

The cost of testing chemicals as reproductive toxicants precludes the possibility of evaluating large chemical inventories without a robust strategyfor prioritizing chemicals to test. The use of quantitative structure-activity relationships in early hazard identification is a cost-effective prioritization tool, but in the absence of systematic collection of interpretable test data upon which models are formulated, these techniques fall short of their intended use. An approach is presented for narrowing the focus of candidate ED chemicals using two in vitro assays: one optimized to measure the potential of chemicals to bind rainbow trout estrogen receptors (rtER), and a second to enhance interpretation of receptor binding data in a relevant biological system (i.e., fish liver tissue). Results of rtER competitive binding assays for 16 chemicals yielded calculable relative binding affinities (RBA) from 179 to 0.0006% for 13 chemicals and partial or no binding for an additional 3 chemicals. Eleven lower to no affinity chemicals (RBA < 0.1%) were further tested in trout liver slices to measure induction of rtER-dependent vitellogenin (VTG) mRNA in the presence of chemical passive partitioning (from media to multiple hepatocyte layers in the slice) and liver xenobiotic metabolism. VTG induction in slices was observed in a concentration-dependent manner for eight chemicals tested that had produced complete displacement curves in binding assays, including the lowest affinity binder with an RBA of 0.0006%. Two chemicals with only partial binding curves up to their solubility limit did not induce VTG. The monohydroxy metabolite of methoxychlor was the only chemical tested that apparently bound rtER but did not induce VTG mRNA. Data are presented illustrating the utility of the two assays in combination for interpreting the role of metabolism in VTG induction, as well as the sensitivity of the assays for measuring enantiomer selective binding and ER-mediated induction. The combined approach appears particularly useful in interpreting the potential relevance of extremely low affinity chemical binding to fish receptors (RBA = 0.01-0.0001%) within a defined toxicity pathway as a basis for prioritizing within large chemical inventories of environmental concern.

A systematic approach to simulating metabolism in computational toxicology. I. The TIMES heuristic modelling framework.

Designing biologically active chemicals and managing their risks requires a holistic perspective on the chemical-biological interactions that form the basis of selective toxicity. The balance of therapeutic and adverse outcomes for new drugs and pesticides is managed by shaping the probabilities for transport, metabolism, and molecular initiating events. For chemicals activated as well as detoxified by metabolism, selective toxicity may be considered in terms of relative probabilities, which shift dramatically across species as well as within a population, depending on many factors. The complexity in toxicology that results from metabolism has been troublesome in QSAR research because the parent structure is less relevant to predicting ultimate effects and finding reference species/conditions for metabolic rates seems hopeless. Even the complexity of comparative xenobiotic metabolism itself seems paradoxical in light of the evidence of highly conserved catabolic processes across species. Clearly, predicting the role of metabolism in selective toxicity and adverse health outcomes requires a probabilistic framework for deterministic models as well as the many factors shaping the metabolic probability distributions under specific conditions. This paper presents a tissue metabolism simulator (TIMES), which uses a heuristic algorithm to generate plausible metabolic maps from a comprehensive library of biotransformations and abiotic reactions and estimates for system-specific transformation probabilities. The transformation probabilities can be calibrated to specific reference conditions using transformation rate information from systematic testing. In the absence of rate data, a combinatorial algorithm is used to translate known metabolic maps taken from reference systems into best-fit transformation probabilities. Finally, toxicity test data itself can be used to shape the transformation probabilities for toxicity pathways in which the metabolic activation is the rate-limiting process leading to a toxic effect. The conceptual approach for metabolic simulation will be presented along with practical uses in forecasting plausible activated metabolites.