The Toxicologist - Society of Toxicology
The Toxicologist - Society of Toxicology
The Toxicologist - Society of Toxicology
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
high-content imaging. Thirty-two compounds were confirmed in both fluorescence<br />
plate reader and imaging assay formats. To study the structure-activity relationship<br />
<strong>of</strong> these mitochondrial disruptors, we clustered these compounds by structural<br />
similarity. This analysis resulted in four structural clusters and 15 singletons.<br />
<strong>The</strong>se clusters may be useful for identifying structural features associated with mitochondrial<br />
toxicity. Our results confirm the robustness <strong>of</strong> this assay for identifying<br />
mitochondrial membrane-potential disruptors in qHTS format. Supported by<br />
NIEHS Interagency Agreement Y2-ES-7020-01.<br />
494 QUANTITATIVE STRUCTURE-ACTIVITY<br />
RELATIONSHIP (QSAR) MODELING OF ESTROGEN<br />
RECEPTOR (ER) BINDING AFFINITY AND VIRTUAL<br />
SCREENING FOR POTENTIAL ENDOCRINE<br />
DISRUPTING COMPOUNDS (EDCS).<br />
L. Zhang 1 , H. Zhu 1 , A. Afantitis 2 , G. Melagraki 2 , H. Sarimveis 2 , I. Rusyn 3 and<br />
A. Tropsha 1 . 1 Eshelman School <strong>of</strong> Pharmacy, UNC, Chapel Hill, NC,<br />
2 NovaMechanics Ltd., Nicosia, Cyprus and 3 Environmental Sciences and Engineering,<br />
UNC, Chapel Hill, NC.<br />
Regulatory agencies, both in Europe and in the US, require substantial, and costly,<br />
animal testing <strong>of</strong> EDCs. <strong>The</strong> use <strong>of</strong> computational predictors to prioritize potential<br />
candidates for in vivo endocrine disruption tests could significantly reduce the cost<br />
<strong>of</strong> experiments and the number <strong>of</strong> compounds subject to testing. In this study, we<br />
have employed combinatorial QSAR modeling workflow, incorporating rigorous<br />
model validation procedures (Tropsha, Mol. Inf., 2010, 29, 476-488), to the largest<br />
publicly available dataset <strong>of</strong> ER binding compounds. <strong>The</strong> dataset containing 436<br />
ERα and 113 ERβ ligands was compiled from public sources and carefully curated.<br />
Each compound was characterized by chemical descriptors calculated with Dragon,<br />
CDK, and Mold2 S<strong>of</strong>tware. To evaluate the true predictivity <strong>of</strong> models, 5-fold external<br />
cross validation procedure was applied. For each modeling set including 80%<br />
<strong>of</strong> all compounds, k-nearest neighbors (kNN) QSAR approach and each <strong>of</strong> the<br />
three descriptor sets were used for model development. <strong>The</strong> external predictive accuracy<br />
<strong>of</strong> models ranged between R2 values <strong>of</strong> 0.63-0.80 and 0.66-0.85 for ERα<br />
and ERβ models, respectively. Y-randomization test yielded no statistically significant<br />
models, indicating the robustness <strong>of</strong> our models. <strong>The</strong> consensus models generated<br />
by combining predictions from all independent models passing both training<br />
and test set accuracy thresholds had the highest external accuracy, characterized by<br />
R2 values <strong>of</strong> 0.72 and 0.76 for ERα and ERβ models, respectively. We have applied<br />
our models for virtual screening <strong>of</strong> several chemical libraries <strong>of</strong> interest including<br />
the 30K compounds registered under REACH as well as EPA-10K compound set.<br />
Approximately 3,000 compounds were identified and prioritized as potential EDCs<br />
for future in vivo endocrine disruption tests.<br />
495 A COLLABORATIVE PROJECT FOR DATA SHARING<br />
AND HARMONIZATION OF DEVELOPMENTAL<br />
TOXICITY DATABASES.<br />
E. Busta 1 , S. Espada 2 , M. Martin 3 , D. Hristozov 1 , K. Arvidson 1 , D. Mattison 4<br />
and C. Yang 1 . 1 U.S. FDA, College Park, MD, 2 Pennsylvania State University,<br />
College Station, PA, 3 U.S. EPA, Research Triangle Park, NC and 4 NICHD,<br />
Bethesda, MD.<br />
Developmental toxicity is one <strong>of</strong> the most important non-cancer endpoints for<br />
both environmental and human health. Despite the fact that numerous developmental<br />
studies are conducted as required for regulatory decisions, there are not yet<br />
sufficient data available to develop predictive computational methods. <strong>The</strong> Food<br />
and Drug Administration’s (FDA) Center for Food Safety and Applied Nutrition<br />
(CFSAN) has been involved in developing ToxML (a toxicity database model) and<br />
toxicity databases. <strong>The</strong> National Center for Computational <strong>Toxicology</strong> (NCCT)<br />
team at Environmental Protection Agency (EPA) implemented ToxRefDB with its<br />
data model inspired by ToxML. <strong>The</strong> new initiative at FDA CFSAN to build an institutional<br />
knowledge-base, the CERES (Chemical Evaluation and Risk Estimation<br />
System) project, has imposed the need to generate another database model for the<br />
Center’s regulatory data. <strong>The</strong> prospect <strong>of</strong> creating yet another toxicity data model<br />
was a real concern; hence, as part <strong>of</strong> the CERES project, an experiment was conducted<br />
at FDA CFSAN to evaluate EPA’s ToxRefDB data entry tool for capturing<br />
developmental effects from FDA’s approved drugs. In collaboration with National<br />
Institute <strong>of</strong> Child Health and Human Development (NICHD), we have added developmental<br />
data on 80 more drugs and merged with the existing database <strong>of</strong> 720<br />
chemicals from EPA and FDA. By incorporating pharmaceutical data into the current<br />
agrochemical domain, both the chemical space and biological pr<strong>of</strong>iles were ex-<br />
106 SOT 2011 ANNUAL MEETING<br />
panded. This experiment also assisted our database modeling while expanding and<br />
harmonizing the controlled vocabulary. It is our further goal to combine the data<br />
with existing developmental toxicity databases to support computational method<br />
development. This poster will present comparisons <strong>of</strong> developmental toxicity data<br />
models, pr<strong>of</strong>iles <strong>of</strong> distinct and common effects, and the chemical domains in both<br />
ToxRefDB and CERES. This abstract does not necessarily reflect policies <strong>of</strong> US<br />
EPA, US FDA, and NICHD.<br />
496 DEHP: COMPARISON OF IN SILICO PREDICTIONS<br />
WITH IN VITRO AND IN VIVO TOXICITY.<br />
M. M. Dingemans 1 , E. Rorije 2 , A. Efremenko 3 , H. Clewell 3 and B. Blaauboer 1 .<br />
1 Institute for Risk Assessment Sciences (IRAS), Utrecht University, Utrecht,<br />
Netherlands, 2 National Institute for Public Health and the Environment (RIVM),<br />
Bilthoven, Netherlands and 3 <strong>The</strong> Hamner Institutes, Research Triangle Park, NC.<br />
This research was conducted within a larger program to investigate possibilities and<br />
limitations <strong>of</strong> the use <strong>of</strong> non-animal data for predicting human risk. In the part described<br />
here, the in silico predicted effects for bis(2-ethylhexyl) phthalate (DEHP;<br />
using knowledge-based systems DEREK and OECD Toolbox, and QSAR-based<br />
TOPKAT®) are compared with those identified in vitro and in vivo, to investigate<br />
whether in vitro studies can confirm predicted effects.<br />
Toxic effects by DEHP are predicted to be teratogenicity and testicular toxicity<br />
(DEREK), carcinogenicity and developmental toxicity (TOPKAT, DEHP in training<br />
set) and DNA binding (metabolism required; OECD Toolbox). In particular<br />
endocrine disruption, reproductive toxicity (including testicular toxicity) and hepatotoxicity<br />
by peroxisome proliferation, and non-genotoxic carcinogenicity are identified<br />
in vitro. <strong>The</strong> predicted DNA binding is not confirmed in vitro, possibly due<br />
to a lack <strong>of</strong> metabolism. In vivo toxicity data for rodents after oral exposure were<br />
obtained from North American and European risk assessment agencies, revealing<br />
increased liver weight and testicular toxicity as the most critical effects (