SRA Logo (print)


Society For Risk Analysis Annual Meeting 2008

Risk Analysis: the Science and the Art

Session Schedule & Abstracts


* Disclaimer: All presentations represent the views of the authors, and not the organizations that support their research. Please apply the standard disclaimer that any opinions, findings, and conclusions or recommendations in abstracts, posters, and presentations at the meeting are those of the authors and do not necessarily reflect the views of any other organization or agency. Meeting attendees and authors should be aware that this disclaimer is intended to apply to all abstracts contained in this document. Authors who wish to emphasize this disclaimer should do so in their presentation or poster. In an effort to make the abstracts as concise as possible and easy for meeting participants to read, the abstracts have been formatted such that they exclude references to papers, affiliations, and/or funding sources. Authors who wish to provide attendees with this information should do so in their presentation or poster.

Common abbreviations

M3-C
Analysis of Genomic Dose-Response Data for Risk Assessment

Room: Grand Ballroom E   2:00-3:30 PM

Chair(s): Jeff Gift, Thomas Russell



M3-C.1    Automated Quantitative Dose Response Modeling. Burgoon LD*, Zacharewski TR; Michigan State University   burgoonl@msu.edu

Abstract: In 2007 the National Research Council released a report on toxicity testing in the 21st century that promoted the development of computational modeling and in vitro high throughput screening (HTS) capabilities, including toxicogenomics, in order to support mechanistically-based quantitative safety and risk assessments. This includes the creation of high performance computing methods for dose-response modeling of large toxicogenomic and HTS data sets. We have developed the ToxResponse Modeler which uses the particle swarm optimization algorithm and a weighted voting method to identify the best-fit model for each response. This approach has been used to calculate probabilistic points of departure and ED50 values to rank and prioritize putative biomarkers of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) that can be associated with specific functions and toxic responses. The ToxResponse Modeler has also been used to evaluate the relative effect potencies (REPs) of 2,3,7,8-tetrachlorodibenzofuran (TCDF)- and PCB126 –elicited responses compared to TCDD. In principle the automated ToxResponse Modeler can be used to analyze any large dose response dataset including outputs from high throughput screening assays to assist with the ranking and prioritization of compounds that warrant further investigation or development, and to support homeland security compound identification and screening.

M3-C.2    Analysis of Genomic Dose-Response Data in the EPA ToxCast Program. Dix D*; Environmental Protection Agency   dix.david@epa.gov

Abstract: The U.S. EPA must assess the potential adverse effects of thousands of chemicals, often with limited toxicity information. Accurate toxicity predictions will help prioritize chemicals for further testing, focusing resources on the greater potential hazards or risks. In vitro genomics is part of EPA’s ToxCast research program, which is developing the ability to forecast toxicity based on bioactivity profiles derived from high-throughput screening (HTS) assays. From several of the HTS assays, ToxCast is generating gene expression data from in vitro cell systems designed to model the rodent or human liver. Following exposure to each of the 320 ToxCast phase I chemicals, analysis of these data will be used to generate in vitro concentration-response curves for individual gene or protein targets, and for the biological pathways populated by these targets. The suitability of specific in vitro targets and pathways for predictive modeling of in vivo toxicity will be assessed based on several factors. First, whether the in vitro concentration-response for gene expression changes suggests plausible in vivo tissue concentrations could have similar effects on the same targets and pathways in vivo. A second suitability factor will be the plausibility of the mode of action inferred by affected genes, targets and pathways, relative to the demonstrated toxicity for specific ToxCast chemicals. An initial focus in study design and data analysis is examining the concentration-response of gene expression regulated by the nuclear receptors CAR, PXR and PPARalpha. These nuclear receptors were selected because they regulate key metabolic pathways modulated by xenobiotics such as the ToxCast chemicals, and critical to non-genotoxic carcinogenic modes of action. This work has been reviewed by EPA and approved for presentation but does not necessarily reflect official Agency policy.

M3-C.3    Useful Lessons for Toxicogenomics using Systems Based Approaches for Dose and Temporal Response Modeling. Faustman EM*, Xiao Y, Griffith WC, Robinson JF; University of Washington   faustman@u.washington.edu

Abstract: The explosion of "Omic" information available for risk assessment requires a reassessment of quantitative risk assessment tools. Although microarray technology has emerged as a powerful tool to explore expression levels of thousands of genes or even complete genomes after exposure to toxicants, the functional and quantitative interpretation of microarray datasets still is a time-consuming and challenging task. Gene ontology (GO) and cell signaling pathway mapping have both been shown to be powerful approaches to generate an overall global view of biological processes and cellular responses impacted by toxicants rather than focusing on single genes. However, current methods do not allow for adequate comparisons across dose and time points and have limited capabilities for translation across levels of biological complexity. Frequently results are presented in extensive gene lists with minimal or limited quantitative information, data that is crucial in risk assessment. To facilitate quantitative interpretation of dose or time dependent genomic data, we propose several systems based approaches. First, we have developed a program (GO-Quant) to extract quantitative gene expression values and to calculate average intensities or ratios based on functional gene categories using MAPPFinder results. An application of this approach will be given. Secondly, computational approaches will be discussed that integrate GO-Quant across cellular, organ, and organism level responses. To evaluate these responses across biological systems necessitates the use of toxicokinetic and dynamic models that can incorporate biological information. A discussion of benchmark doses (BMDs) and minimal effective levels (MELs) will be included and challenges for using such responses within the context of the risk assessment framework will be presented. Supported by NIEHS 5-P01-ES009601, NIEHS P30-ES007033, NIEHS P50-ES012762, NIEHS U10-ES011387, NSF OCE-0434087, EPA RD-83170901, and EPA RD-83273301.

M3-C.4    Integration of toxicogenomics data in mode of action analyses and cancer risk assessment. Keshava C*, Keshava N, Davis A, Gift J; US Environmental Protection Agency   keshava.channa@epa.gov

Abstract: Development of new experimental approaches capable of differentiating among a wide range of mechanisms of action is expected to significantly improve risk assessment (RA) methodologies. US EPA’s 2005 Guidelines for Carcinogen Risk Assessment encourages incorporation of data from new technologies to better inform RA. Toxicogenomic (TG) data are beginning to be used in RA to inform qualitative mode of action (MOA) analyses and support weight-of-evidence evaluations for the characterization of carcinogenicity of environmental chemicals. One of the hallmarks of TG is identification of molecular signatures of specific classes of toxic compounds that discriminates direct-acting genotoxicants from indirect-acting genotoxicants based on their gene expression patterns. TG approaches have demonstrated that changes in pathway-associated gene expression profiles could be used to differentiate genotoxic from non-genotoxic hepatocarcinogens in rodents. Additional quantitative methods for analysis of TG data include benchmark dose modeling of dose-responses from results of microarray studies. This methodology allows for the analysis of changes in transcriptional profiles due to chemical exposure in individual genes as well as families of genes associated with specific cellular processes, and calculates benchmark doses which identify the most susceptible single gene and cellular process. These quantitative methodologies can in turn be informative in the overall determination of carcinogenic MOAs. The goal of this presentation is to summarize current scientific progress in TG methodologies and to integrate such data for evaluating and characterizing genotoxic from non-genotoxic carcinogens. The information obtained from such analyses can be utilized to improve understanding of the MOA for environmental carcinogens and to facilitate better cancer risk characterization. (The views expressed are those of the authors and do not necessarily reflect the vies or policies of the US EPA)

M3-C.5    Potential Impacts of Genomics on EPA Regulatory and Risk Assessment Applications. Benson WH, Birchfield N, Gallagher K*; U.S. Environmental Protection Agency   benson.william@epa.gov

Abstract: Advances in molecular technology have led to the elucidation of full genomic sequences of several multicellular organisms. The related molecular fields of proteomics and metabolomics are now beginning to advance rapidly as well. In addition, advances in bioinformatics and mathematical modeling provide powerful approaches for analyzing patterns of biological response imbedded in the massive data sets produced through genomics research. Many defensive responses to external stimuli are common among many organisms, including wildlife species (fish, birds, invertebrates) and humans. In view of this, genomic technologies may provide great insight into how diverse organisms respond to environmental stressors and provide information for regulatory and risk assessment applications at the U.S. Environmental Protection Agency (EPA). In this regard, a cross-Agency Genomics Task Force developed a White Paper which outlined implications for the use of genomics technologies in EPA. Four areas were identified as those likely to be influenced by the generation of genomics information within EPA and the submission of such information to EPA: (1) prioritization of contaminants and contaminated sites, (2) monitoring, (3) reporting provisions; and (4) risk assessment. It is important to note that significant research by EPA and other agencies and researchers will be necessary to fully understand and apply genomics technologies to human health and ecological risk assessment. A critical need in the area of technical development was identified as the need to establish a framework for the analysis and acceptance of genomics information for scientific and regulatory purposes. This presentation will discuss the various activities of the EPA’s Genomics Task Force in the context of implications for regulatory and risk assessment applications with particular emphasis on environmental risk assessment.

M3-C.6    Comparative Benchmark Dose Estimates for Genomic Data from TCDD-, TCDF-, or PeCDF-Treated Human and Rat Normal Hepatocytes. Rowlands JC*, Budinsky R, Gollapudi B, Dombkowski A, Thomas R; The Dow Chemical Company   JCRowlands@dow.com

Abstract: Assessing health risks associated with mixtures of dioxins and dioxin-like compounds (DLCs) utilizes the toxic equivalency factor (TEF) approach that calculates a potency estimate for a DLC relative to that of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). The TEF concept assumes exposure additivity for the DLCs in a mixture and the single TEF value for a DLC is assumed to be the same in all species. Thus, TEFs are central to the risk assessment of DLCs. The application of genomic technologies has the potential for improving the accuracy of the TEF approach by allowing comparisons of potency across a wide-range of signaling pathways and in multiple species and cell types. This study used such an approach by measuring microarray-based gene expression changes in primary human and rat hepatocytes treated with seven log-order concentrations of TCDD, 2,3,7,8-tetrachlorodibenzofuran (TCDF), or 2,3,4,7,8-pentachlorodibenzofuran (PeCDF). The data were analyzed using the BMDExpress software application that combines traditional benchmark dose (BMD) methods with gene ontology classification in the analysis of dose-response data from microarray experiments. In this presentation, the BMD estimates for affected pathways in human and rat hepatocytes will be discussed and compared with the standard TEF approach.

M3-C.7    The Potential of Genomic Dose-Response Data to Define Mode-of-Action and Low-Dose Behavior of Chemical Toxicants. Thomas RS*, Allen BC, Longlong Y, Clewell HJ, Andersen ME; The Hamner Institutes for Health Sciences   rthomas@thehamner.org

Abstract: There is increasing acceptance within the toxicology community that high dose animal studies are not predictive of low dose risks in humans or even in the test animals themselves. New genomic technologies now provide a unique opportunity to evaluate the relevance of current low dose default assumptions used in chemical risk assessment and identify dose-dependent transitions in modes of action. In this presentation, we describe the application of benchmark dose analysis to gene expression microarray data collected following ninety day exposures to five different lung and liver carcinogens. The benchmark dose methods were used to estimate doses at which different cellular processes are altered and showed that benchmark dose values for certain processes mirrored the tumor response in a two-year rodent bioassay. The results show that dose-response changes in gene expression, when related to higher-order biological processes and pathways, reflect the dose dependent changes in key events in the carcinogenic process and can support nonlinear modeling of the events in the low dose region within the framework of a mode of action risk assessment.



[back to schedule]