Identifying Metabolically Active Chemicals Using a Consensus Quantitative Structure

Identifying Metabolically Active Chemicals Using a Consensus Quantitative Structure Activity Relationship Model for Estrogen Receptor Binding

PRESENTING AUTHOR: 

Jeremy A. Leonard

INSTITUTION / COMPANY : 

Oak Ridge Institute for Science and Education

AUTHOR(S): 

Jeremy A. Leonard, Caroline Stevens, Kamel Mansouri, Daniel Chang and Cecilia Tan

REFERENCES: 
National Exposure Research Laboratory, U.S. EPA, Athens, GA
ScitoVation, Durham, NC
Office of Pollution Prevention and Toxics, U.S. EPA, Washington, D.C.
National Exposure Research Laboratory, U.S. EPA, Durham, NC
ABSTRACT CONTENT / DETAILS: 

Traditional toxicity testing provides insight into the mechanisms underlying toxicological responses but requires a high investment in a large number of resources. The new paradigm of testing approaches involves rapid screening studies able to evaluate thousands of chemicals across hundreds of biological targets through use of in vitro assays. Endocrine disrupting chemicals (EDCs) are of concern due to their ability to alter neurodevelopment, behavior, and reproductive success of humans and other species. A recent integrated computational model examined results across 18 ER-related assays in the ToxCast in vitro screening program to eliminate chemicals that produce a false signal by possibly interfering with the technological attributes of an individual assay. However, in vitro assays can also lead to false negatives when the complex metabolic processes that render a chemical bioactive in a living system might be unable to be replicated in an in vitro environment. In the current study, the influence of metabolism was examined for over 1,400 chemicals considered inactive using the integrated computational model. Over 2,000 first-generation and over 4,000 second-generation metabolites were generated for the inactive chemicals using in silico techniques. Next, a consensus model comprised of individual structure activity relationship (SAR) models was used to predict ER-binding activity for each of the metabolites. Binding activity was predicted for 8-10% of the metabolites within each generation. Additionally, it was found that approximately 20% of the inactive parents have at least one potentially active metabolite. The approaches presented here can be used to identify potential parents that are inactive under in vitro conditions but that might become metabolically active in a living organism.   

Disclaimer: The views expressed in this presentation are those of the authors and do not necessarily represent the views or policies of the U.S. EPA.