An Official Journal of the Society of Toxicology

www.toxsci.oxfordjournals.org

The Toxicologist

47th Annual Meeting and ToxExpoTMSeattle, Washington

47th Annual Meeting and ToxExpoTMMarch 16–20, 2008

Society of Toxicology 1821 Michael Faraday Drive, Suite 300 • Reston, VA 20190

T: (703) 438-3115 • F: (703) 438-3113 • E-mail: sothq@toxicology.org

www.toxicology.org

The Toxicologist March 2008

48thAnnual Meeting and ToxExpo™Baltimore, Maryland

Baltimore Convention CenterMarch 15–19, 2009

Photos courtesy of Baltimore Area Convention and Visitors Association

ISSN 1096-6080Volume 102, Number 1, March 2008

2008

SEATTLE

2008

SEATTLE

Supplement to Toxicological Sciences

We are proud to print this publication entirely on Forest Stewardship Council certified paper. FSC certification ensures that the paper in this publication contains fiber from well-managed and responsibly harvested forests that meet strict environmental and socioeconomic standards.SOT saved the following resources: 33.6 trees preserved for the future • 23,800,000 BTUs energy not consumed • 3,109 lbs net greenhouse gases prevented • 14,272 gallons wastewater flow saved • 97.02 lbs waterborne waste not created • 1,579 lbs solid waste not generated

Cert no. XXX-XXX-XXXX

SeSSion TypeSContinuing Education—Emphasis on quality presentations of generally accepted, state-of-the-art knowledge in toxicology

Note: CE Courses will be held on Sunday.

Symposia—“Cutting-edge” science; new areas, concepts, or data

Workshops—State-of-the-art knowledge in toxicology

Roundtables—Controversial subjects

Historical Highlights—Review of a historical body of science that has impacted toxicology

Informational Sessions—Scientific planning or membership development

WHy SUBMiT A pRopoSAL?1. To present new developments in toxicology.

2. To provide attendees an opportunity to learn about state-of-the-art technology and how it applies to toxicological research.

3. To provide attendees an opportunity to learn about the emerging fields and how they apply to toxicology.

1821 Michael Faraday Drive, Suite 300, Reston, VA 20190 Phone: (703) 438-3115

Deadline for proposals for SoT 2009 Annual Meeting Sessions: April 30, 2008

2009 Thematic ApproachThe Program Committee will carry forward the thematic approach for the 2009 Annual Meeting. All proposal submissions will be reviewed for their relevance under the following themes:

Biomarkers•

Epigenetics•

InflammationandDisease•

NeurodegenerativeDisease•

Due to the timeliness of the Nanotechnology theme, it will again be highlighted for the 2009 meeting.

2 0 0 948thAnnual Meeting and

ToxExpo™

Baltimore, Maryland

March 15–19, 2009Baltimore Convention Center

Photos courtesy of Baltimore Area Convention and Visitors Association

Submit your proposal on-line at www.toxicology.org

SOT 2008 ANNUAL MEETING 1

1 MINI-PIGS AS AN ALTERNATIVE NON-RODENTSPECIES IN TOXICOLOGY AND SAFETY STUDIES.

G. Washer1 and A. Makin2. 1LAB Research Inc., Montréal, QC, Canada and2Research Denmark (Scantox), Copenhagen, Denmark.

This course will focus on the regulatory acceptability and use of minipigs as an al-ternative to traditional non-rodent species for non-clinical studies of pharmaceuti-cals. Attention will be given to their use in general toxicology and safety studies, in-cluding non-routine routes and embryofetal development studies. The choice ofnon-rodent species for safety assessment of new pharmaceuticals is becoming in-creasingly important. Researchers have adopted the paradigm that we use the dog,and if the dog is no good, then we use the primate. With the ever increasing ethicalconcerns relating to the use of primates, and other groups concerned with the use ofman’s best friend, attention has increasingly focused on the potential of the pig andminipig. An increasing appreciation of the similarities between man and the pig instructure of skin, gastrointestinal tract, urogenital system and metabolism is theprincipal reason for the increased popularity. With 13 species of pig, however, bio-medical research has been largely restricted to the domestic pig derived from theEurasian wild boar and the warthog. Smaller strains have increasingly come toprominence in non-clinical testing and are widely employed as non-rodent alterna-tives for regulatory-driven safety programs. For pharmaceutical compounds whoseroute is dermal in man, there is a growing realization that the minipig is a validmodel. Pig skin has been shown to be anatomically, physiologically, biochemicallyand immunologically similar to human skin. Although the use of the minipig in thedevelopment of dermal products is well accepted by regulators, its suitability for awide range of other study types has been somewhat overlooked although in manyways it is more representative of man than other species. The minipig is now recog-nized as a suitable non-rodent species for efficacy and safety studies for many typesof pharmaceutical products, and are suitable experimental models using many ofthe commonly applied techniques used in other species. They have also been usedas alternative non-rodent species for embryofetal studies.

2 INTRODUCTION TO PATHOLOGY FORTOXICOLOGISTS AND STUDY DIRECTORS.

L. M. Fomby1 and P. Haley2. 1Battelle, Columbus, OH and 2Incyte Corp,Wilmington, DE.

Today, even with the promise of ‘omic’ technologies, pathology plays a critical rolein the evaluation of chemicals and drugs. Toxicological pathology uses specializedterms and methods and can generates complex interpretative problems. Effectivecommunication between the toxicologist and the study pathologists will allow thetoxicologist to better integrate pathology data into their reports and optimize thescientific content of the study. In this basic course, experienced toxicologicalpathologists will introduce important concepts in pathology to toxicologists asstudy directors in order to help the study director appreciate the role of pathologyin toxicological studies. The course will begin with a discussion of basic conceptsand the role of the toxicological pathologist followed by a discussion of global reg-ulations as they relate to pathology endpoints. Later lectures will discuss the lesionsand interpretation of pathology data in non-oncogenicity studies and oncogenicitystudies. By the end of the session, the study director should have a better under-standing of pathology and a basic background that will allow for effective commu-nication with a study pathologist.

3 STEM CELLS AND THEIR MULTI-POTENTIAL USESAND POTENTIAL DANGERS.

D. Lawrence1 and M. J. Pallardy2. 1Wadsworth Center, Albany, NY and 2Universityof Paris, Chatenay Malabry, France.

Stem cells can theoretically divide without limit to replenish cells and each daugh-ter cell has the potential to remain a stem cell or develop into a fully functional dif-ferentiated cell. Therefore, stem cells have the potential to develop into many celltypes that could benefit the health status of many individuals but the types of cellsthat could differentiate from stem cells must be considered in that certain growthand differentiation factors could give rise to conditions that could be harmful in-cluding development of a cancer stem cell. Additionally, environmental factors canskew the developmental patterns of stem cells leading to harmful deficiencies or ex-pansions of certain types of progenitors. This course is designed as a basic levelcourse to provide an overview of the field of stem cell biology, touch upon the dif-ferences of adult versus embryonic stem cells and provide a foundation for under-standing the implications of the use of stem cells in toxicological research.Presentations will cover in vitro and in vivo proliferation and differentiation of stemcells and the influences of toxicants on these processes. Quantification of normaland aberrant immunophenotypic changes with development, mechanisms that in-fluence proliferation, differentiation, and functional changes, and in vivo control ofprogenitors in their developmental niche will be discussed. The effect of environ-

mental toxicants on a single molecular pathway controlling cell signaling and de-velopment of the CNS will be presented. Finally, the challenges in pharmacologi-cally regulated cell therapy including the ability to regulate the fate of engineeredcells will be deliberated.

4 DOSE-RESPONSE MODELING FOR OCCUPATIONALAND ENVIRONMENTAL RISK ASSESSMENT.

D. G. Dolan1 and A. Maier2. 1Amgen, Inc., Thousand Oaks, CA and 2ToxicologyExcellence for Risk Assessment, Cincinnati, OH.

The development non-cancer exposure guidance values - whether for environmen-tal or occupational sources of exposure - is rooted in the ‘critical adverse effect’ con-cept. Thus, derivation of such guidance values requires understanding the methodsand approaches for estimating thresholds for the onset of adverse effects. Thiscourse will describe the history of the development of methods to develop ‘safe’ ex-posure limits using adverse effect level estimation. The lecture will focus on currentmethods and issues for using dose-response modeling for developing potency andpoint of departure estimates for limit-setting. Each lecture will include presenta-tions followed by hands-on application of key elements from the lecture as appliedto model toxicological datasets. The participants are expected to obtain sufficientknowledge to estimate critical effect levels for risk assessment using a variety ofmodeling techniques and addressing key considerations of current interest.

5 THE USE OF TRANSGENIC ANIMAL TECHNOLOGY INTOXICOLOGICAL RESEARCH.

D. K. Ness1 and R. Forster2. 1Elan Pharmaceuticals, Inc., South San Francisco, CAand 2CIT, Evreux, France.

The ability to direct genetic changes at the molecular level in vivo has resulted in arevolution in biology. Nowhere has this been more apparent than in the productionof transgenic animals. A host of techniques has been used to effect change in geneexpression and develop new toxicological testing paradigms. Genetically modifiedanimals are commonly produced and often yield important information relevant tosafety/toxicological assessment. This session will help guide the toxicologist in theuse and interpretation of data derived from transgenic models. This course will in-troduce this topic and review the history of genetic engineering technologies lead-ing to the development of loss-of-function, gain-of-function modeling technologiesacross mammalian platforms.

6 PROCESS BASED APPROACHES TO MODULATINGGENE AND PROTEIN EXPRESSION IN VIVO AND INVITRO.

R. Pollenz1 and R. Tanguay2. 1University of South Florida, Tampa, FL and 2OregonState University, Corvallis, OR.

The mechanistic analysis of cellular responses to xenobiotics requires the ability tomodulate important genes involved in specific pathways. Such genes include thosethat encode receptors that associate with xenobiotics as well as the enzymes in-volved in xenobiotic metabolism. The ability to modulate these genes and proteinsin vitro and in vivo has become accessible to more laboratories with the refinementof techniques such RNA interference (RNAi), viral gene delivery, morpholino-me-diated gene knock down and targeted gene disruption. However, the ability to uti-lize these techniques and generate reproducible results requires a detailed under-standing of the advantages and applications of each procedure. Thus, the goal ofthis course is to provide the investigator with an overview of experimental designand the use of proper controls for four cutting edge techniques. The first talk willfocus on experimental design and analysis of RNAi to reduce endogenous targetproteins in culture cells with emphasis on controls and endpoint analysis. The sec-ond presentation will move to the zebrafish model system and discuss the use ofmorpholino-mediated gene knock down to reduce the expression of specific pro-teins in embryos. The third presentation will discuss gene delivery utilizing the ade-novirus system for reduction of gene expression in mice. The forth presentation willdetail the use or transgenic approaches in mouse models to modulate the expressionof specific target genes or knock-in genes from other species. This course should beof broad interest to laboratories considering a mechanistic approach to understand-ing signal transduction pathways, gene expression and protein-protein interactionsas well as those currently investigating these endpoints.

7 BASIC EMBRYOLOGY AND DEVELOPMENTALTOXICOLOGY.

L. Dostal1 and J. M. Rogers2. 1Private Consultant, Ann Arbor, MI and 2U.S. EPA,Research Triangle Park, NC.

Embryonic and fetal prenatal development in mammalian species is a complexprocess which is sensitive to the effects of maternal and environmental factors. Thetiming of development of major organ systems varies between humans and various

2 SOT 2008 ANNUAL MEETING

test species used in assessment of developmental toxicity of xenobiotics. This basiccourse will cover normal development from fertilization through early stages of im-plantation and embryogenesis, through development of the major organ systems(cardiovascular, central nervous system, craniofacial, skeletal, urogenital). The ef-fects of known human and animal teratogens on fetal development will be pre-sented as examples, including etiologies of abnormal development. Current re-quirements and study designs for developmental toxicology studies required fornew chemicals and new drugs will be covered in the context of the biological con-cepts of organ systems described in the initial lectures.

8 NANOTOXICOLOGY: THE SCIENCE OF DEVELOPINGA SAFE TECHNOLOGY.

C. M. Sayes1 and J. Teeguarden2. 1DuPont Haskell Labs, Newark, DE and 2PacificNorthwest National Laboratory, Richland, WA.

The objective of this course is to prepare toxicologists and risk assessors to meet therapidly growing need to understand and evaluate the risks that engineered nanoma-terials may pose to human health. Toxicological and risk assessment of nanomateri-als requires an understanding of the unique differences between these ‘new’ materi-als and their chemical and large-particle predecessors. Few currently have themultidisciplinary understanding (toxicology, nanoscience, applied physics, dosime-try) to effectively approach the problem of risk and safety assessment of these mate-rials. This basic course will provide an overview of the state of the science of thefield of nanotoxicology and develop the scientific basis for evaluating and charac-terizing exposure, toxicity and the risks associated with engineered nanomaterials.Nanoscience and nanomaterials will be described with a focus on how they may bedifferent than chemicals and previous studied particulates. A summary of currentevidence regarding important routes of exposure and materials humans are exposedto will be presented. Current methods for, and challenges to toxicological testing ofnanomaterials will be covered with specific examples relating the toxicity ofnanoscale TiO2 to more well characterized particulates (e.g. crystalline silica).Guidelines for testing have been proposed by several groups and will be summa-rized. Finally, the dosimetric basis for extrapolating the results of toxicology studiesto humans will then be developed and presented. This course promotes the applica-tion of best scientific practices in the toxicological evaluation and risk assessment ofthis rapidly growing family of new materials.

9 CLINICAL DOSE SETTING FOR BIOTHERAPEUTICS.

J. D. Green1 and S. M. Heidel2. 1Biogen Idec, Inc., Cambridge, MA and 2Eli Lillyand Company, Indianapolis, IN.

Preclinical development programs that are designed to support the safe clinical useof biopharmaceuticals have considerations that are very different from programsdesigned to support the development of small molecule drugs. The InternationalConference of Harmonization ICH S6 document provides guidance regarding sci-entific and regulatory consensus for key preclinical study design elements. Since theadoption of ICH S6 considerable experience has been developed across a wide vari-ety of biological drug and antibody constructs. This experience has indicated that ifICH S6 guidance is followed and expert assessment of data sets is provided, thesafety considerations for first in human and later stage clinical trials has been as-sured. However, the recent experience involving the humanized monoclonal anti-body TGN1412 has raised questions regarding the adequacy of current nonclinicalapproaches. This session will review the following current and proposed regulatoryguidance documents that determine safe starting doses, basic and advanced con-cepts in PK and PD that support preclinical and clinical dosimetry, basic and ad-vanced concepts of toxicology that are involved in the assessment of safety of bio-therapeutics and the current approaches will be illustrated by several case studiesacross a wide range of product classes. The course attendee will learn key conceptsand design considerations for successful preclinical programs that support the initi-ation of safe human clinical trials.

10 USE OF DATA FOR DEVELOPMENT OF UNCERTAINTYFACTORS IN NON-CANCER RISK ASSESSMENT.

J. Lipscomb1 and L. T. Haber2. 1U.S. EPA, Cincinnati, OH and 2ToxicologyExcellence for Risk Assessment, Cincinnati, OH.

Promoting clarity and objectivity is a primary concern when advancing the techni-cal basis for human health risk assessment. Default values for uncertainty factorsare intended to serve as place-holders, to be replaced when relevant data becomeavailable. The default values for the uncertainty factors for inter and intraspecies ex-trapolation were established before many toxicologists were born, and methods for

incorporation of chemical-specific or categorical data continue to evolve. To under-stand the rationale for these uncertainty factors and methods for their replacement,this course will present a continuum of approaches to develop non-default valuesfor uncertainty factors, culminating in the application of advanced toxicokineticmethods to quantify differences in internal dosimetry. The course will describe thehistory of the development of default values for uncertainty factors, including theirrecent subdivision into toxicokinetic and toxicodynamic components; methodsthat make use of generally-applicable species differences in anatomy and physiologyand allometric scaling; specific instruction in the Chemical Specific AdjustmentFactor (CSAF) methodology developed by the International Programme onChemical Safety (IPCS); and the application of physiologically based pharmacoki-netic (PBPK) modeling to develop quantitative data for use in replacing default as-sumptions for inter and intraspecies differences in tissue dosimetry. The concludinglecture will present a series of examples to illustrate how the various approaches in-troduced in the course can be applied to derive non-default uncertainty factors forenvironmental contaminants and pharmaceutical compounds.

11 ESSENTIAL INFORMATICS FOR TOXICOLOGISTS:KNOWLEDGE MANAGEMENT END TO END.

W. B. Mattes1 and L. D. Burgoon2. 1The Critical Path Institute, Rockville, MD and2Toxicogenomic Informatics and Solutions, LLC, Lansing, MI.

The combination of the Internet, automated data acquisition, and genomic infor-mation has transformed the role of the computer in the modern scientist’s life. A fa-miliarity with word processing and simple spreadsheets is simply not adequatepreparation for dealing with large datasets such as those generated by large toxicol-ogy studies, toxicogenomics or high-throughput screens. Increasingly, the softwaretools used to deal with such data require an understanding of basic concepts incomputer science, database design, bioinformatics and statistics. This ‘basic’ levelcourse hopes to provide the beginnings of such an understanding. Thus the first lec-ture will cover some of the essential concepts of operating systems, file and dataconcepts and programming concepts. This will be followed by a talk discussing theessentials of database design and use, contrasting flat-file and relational databases. Athird lecture will provide an overview of the information necessary so that data maybe integrated, e.g. pathology with genomics. The final lecture will cover concepts ofvisual analysis of large data sets, and contrast some of the various approaches used.Hopefully after this course the student will be conversant in informatics to the levelof effectively interacting with computer scientists, as well as collecting and manipu-lating datasets with reasonable skill.

12 EPIDEMIOLOGY FOR TOXICOLOGISTS:INTRODUCTION.

R. A. Parent1 and G. Olsen2. 1Consultox Ltd., Damariscotta, ME and 23MCorporation, St. Paul, MN.

Over the course of the last 50 years, epidemiology has gained respect as an estab-lished scientific discipline. Whereas epidemiologic results were once greeted withscorn and skepticism they are now accorded some degree of respect and in some in-stances perhaps too much respect. This course will begin by providing an overviewof the terminology used to describe the techniques used to measure and quantifydisease in populations. Toxicologist will become familiar the with the various studydesigns used in epidemiologic research to generate and evaluate hypotheses regard-ing disease occurrence. Next, the distinction between experiments and observa-tional studies will be made with an overview on epidemiologic study designs, andbias in epidemiologic data. We will cover the types of observational study designscharacterized according to whether the unit of observation is a group, the individ-ual, or some hybrid of the two. A summary will provide the audience with somegeneral guidelines regarding the tendency of the different types of epidemiologicstudies to provide unbiased results.

13 METALS, MICROGLIA, AND NEUROINFLAMMATION.

D. Lawrence. Wadsworth Center, Albany, NY.

The major immune cell that permanently resides within the brain parenchyma isthe microglial cell. Current consensus suggests that microglia are derived from thehematopoietic monocyte lineage and that blood monocytes differentiate intoperivascular macrophages which develop into microglia as they traffic into multiplebrain regions. Another hematopoietic cell residing in the brain is the mast cell.Most other immune cells only transiently migrate through the brain. A major ques-tion that continues to be debated is whether neurotoxic metals directly activatedimmune cells within the brain that subsequently induce neuronal damage orwhether metal toxicants induce neuronal damage, which subsequently activates the

SOT 2008 ANNUAL MEETING 3

immune cells (mainly microglia) leading to exacerbated neuronal damage. An alter-native hypothesis is that microglia are neuroprotective but metals cause loss of mi-croglial protective factors because of microglial degeneration and/or elevated pro-duction of potential cytotoxicants. Numerous cytokines have been implicated inthe neuroinflammation and neurodegeneration, but their cellular sources have notbeen fully delineated. The various neuroimmune mechanisms posited to regulateneuroinflammation, neuroprotection, and neuronal degeneration will be discussed.

14 ROLE OF MICÏROGLIA IN METAL NEUROTOXICITY.

M. Aschner. Pediatrics, Vanderbilt University Medical Center, Nashville, TN.

Microglia respond early to neurotoxic insults. Increased evidence suggests that acti-vation of these cells contributes to neuronal damage in multiple neurodegenerativediseases, including Parkinson’s disease and Alzheimer’s disease. Activation of mi-croglia may ensue as a result of (1) recognition of pro-inflammatory stimuli, such aslipopolysaccharide (LPS) or (2) in response to neuronal damage (reactive mi-crogliosis), which is then toxic to neighboring neurons, resulting in a perpetuatingcycle of neuronal demise. Microglial activation can also enhance ongoing neurode-generation. For example, microglia-derived LPS synergistically enhances both theactivation as well as the neurotoxic effects of rotenone and MPTP, suggesting am-plification of neuroinflammation processes and the associated neurotoxicity by ac-tivated microglia. A common mechanism invoked in mediating microglia-inducedneurotoxicity is via release of reactive oxygen species (ROS). The primary and mostcommon pathway by which diverse toxin signals are transduced into ROS produc-tion in microglia is via pattern recognition receptors (PRRs). Stimulation ofNADPH oxidase activity appears to be the predominant mechanism by which mi-croglia generate ROS. NADPH oxidase is also implicated as a mechanism of pro-inflammatory signaling in microglia. This talk will address the role of microglia inlinking the environment and disease, using several environmental agents as exam-ples. Major objectives of the talk are (1) to discuss the propensity of microglia to ac-cumulate toxins, such as methylmercury (MeHg); (2) to discuss mechanisms asso-ciated with microglial activation by meHg and other xenobiotics; and (3) toenumerate common downstream effects associated with microglial activation andneurotoxicity.

15 POTENTIAL ROLE OF MICROGLIA IN MANGANESENEUROTOXICITY.

N. M. Filipov, P. L. Crittenden and S. Lee. CEHS, Basic Sciences, Mississippi StateUniversity, Mississippi State, MS.

Proinflammatory cytokines and other inflammatory mediators, when produced insubstantial amounts, may be involved in neurodegenerative diseases such asParkinsons Disease (PD). Microglia, the main producers of proinflammatory cy-tokines in the brain, are not uniformly distributed and are concentrated in the basalganglia, an area of the brain that is also a target for excessive manganese (Mn) ex-posure. This presentation will (i) describe our findings that Mn enhances the pro-duction of lipopolysaccharide (LPS)-induced proinflammatory cytokines such asIL-1β, IL-6, and TNF-α by microglial cell lines and microglia-rich brain slices and(ii) present new data pertaining to the mechanism(s) Mn uses to enhance the pro-duction of proinflammatory cytokines by activated microglia. More specifically, inseries of studies, we have determined that the increased proinflammatory cytokineproduction by activated microglia exposed to Mn is associated with persistent acti-vation of a key intracellular kinase, p38. Moreover, the prolonged p38 activationcaused by Mn exposure appears to be caused by an inhibition of a key phosphatase.In addition, a transcription factor that is potentially involved in proinflammatorycytokine production is markedly increased by Mn in LPS-activated microglia.Increased production of proinflammatory cytokines by microglial cells may causedamage to neighboring neurons. Thus, determining the mechanism(s) of Mn-en-hanced proinflammatory cytokine production will contribute to our understandingof Mn neurotoxicity and similar neuropathologies.

16 LEAD AND MERCURY INDUCED NEUROIMMUNECHANGES IN THE BRAIN.

D. Lawrence1, J. Kasten-Jolly1, D. Gao1, T. Mondal1 and C. Hudson2.1Wadsworth Center, Albany, NY and 2SUNY Upstate Medical University, Syracuse, NY.

Glia activity is modified by developmental lead (Pb) or mercury (Hg) exposure, andthese exposures altered behaviors in adults. Exposure via the drinking water to thedam continued from gestational day 8 until weaning at postnatal day (PND) 21. AtPND 21, some pups were separated by gender and brains were assayed for mRNAand protein levels for cytokines and immunoregulatory factors. Some mice per lit-ter also were aged an additional 6 wk without metal in their drinking water and as-

sayed for behavioral changes. The two major immune cells residing in brainparenchyma (microglia and mast cells) were specifically assessed. Metal effects onastrocytes also were measured with regard to GFAP expression. Both the Hg and Pbexposure differentially affected the neuroimmune profiles as well as the behaviors ofmale and female mice. In general, both Pb and Hg appear to enhance processes as-sociated with neuroinflammation. Cytokines that currently appear to be intimatelyinvolved in the responses to the metal exposures are IL-6, IL-18, and TGF-beta.Potential mechanisms implicated in the elicited processes will be discussed.

17 MICROGLIAL ACTIVATION OR MICROGLIALDEGENERATION?

W. J. Streit. Neuroscience, University of Florida, Gainesville, FL. Sponsor: D.Lawrence.

Neurotoxic metals are thought of as causing injury primarily to neurons, followedby glial activation secondarily. Much interest has been focused on the activation ofmicroglia because activated microglia are often seen as potentially dangerous im-mune effector cells believed to exacerbate the initial toxic insult. Activated mi-croglia have also been linked causally to the development of neurodegenerative dis-eases, such as Alzheimer’s and Lou Gehrig’s disease. On the other hand, there isreason to believe that activated microglia are neuroprotective and important for res-cuing injured neurons. Based on this idea, we have reexamined microglial activa-tion in neurodegenerative disease states in both human and rodent brain, and ourfindings strongly suggest that many ostensibly activated microglia are, in fact, de-generating microglial cells. These observations have led us to propose a novel ‘mi-croglial dysfunction hypothesis’ regarding the role of activated microglia in thepathogenesis of neurodegeneration. The hypothesis states that it is the loss of mi-croglia and/or microglia neuroprotective functions that triggers neurodegeneration.This view represents a paradigm shift offering a revised perspective on the func-tional significance of activated microglia. It should prompt studies to investigate iftraditional neurotoxins might also have microgliotoxic properties.

18 MOLECULAR BASIS FOR SUSCEPTIBILITY TOCHEMICAL TOXICITY AND DISEASE.

R. Hines. Medical College of Wisconsin & Children’s Research Institute, Milwaukee, WI.

The completion of the Human Genome Project in 2003, and more recently, thefirst phase of the International HapMap project, have provided valuable toolswhich we are still learning how to use effectively in our goal to improve humanhealth and prevent disease. For example, the cost-effective and rigorous ability toperform whole genome association studies for complex diseases has only beenachieved within this past year. Although exciting and of clear utility in identifyingmultiple genetic loci whose variability will contribute to disease risk, such associa-tion studies still must be followed-up with more detailed analyses to understand thecausative variants that contribute to differential susceptibility. Further, there now iswidespread acceptance that most if not all complex disease arises because of an in-teraction between our genome and the environment. Thus, a careful assessmentand definition of phenotype, including environmental exposure, is critical. We alsonow have clear examples wherein epigenetic changes caused by environmental fac-tors, including transgenerational epigenetics changes, can also have a significant im-pact on disease risk. This symposium will provide state-of-the-art updates on ourunderstanding of the molecular basis for individual differences in susceptibility tochemical toxicity and disease. Topics to be covered include how regulatory poly-morphisms resulting in variation in gene expression can impact risk, chemical-in-duced epigenetic changes and their impact on susceptibility, the use of genetic vari-ation to predict risk and efficacy during drug development and the use of pathwayanalyses to understand complex disease risk using Parkinson’s disease as an example.Thus, this symposium will span from fundamental, mechanistic studies, to transla-tional investigations and clinical applications, demonstrating the strengths of an in-terdisciplinary approach to environmental health issues.

19 REGULATORY POLYMORPHISMS AND THEIR IMPACTON SUSCEPTIBILITY.

R. N. Hines. Pediatrics, Medical College of Wisconsin & Children’s Research Institute,Milwaukee, WI.

CYP2C9 accounts for approximately 20% of the total cytochrome P450 content inadult human liver and is responsible for the metabolism of approximately 16% ofclinically used drugs, as well as numerous environmental toxicants. Although varia-tion within the CYP2C9 structural gene has been extensively characterized, theidentified variants fail to fully explain interindividual CYP2C9 expression differ-ences. To address this problem, 31 SNPs were identified in 10 kbp of CYP2C9 up-stream information using the 24 DNA sample set from the Coriell PDR. Individual

4 SOT 2008 ANNUAL MEETING

SNP frequencies were determined in 193 DNA samples obtained from HispanicAmericans (Mexican descent) and used to infer common haplotypes using PHASE2.1. Nine common variant haplotypes were inferred, two of which included theCYP2C9*2 and two the CYP2C9*3 causative SNPs. The variant haplotypes wereintroduced into a CYP2C9/luciferase reporter construct and their impact on pro-moter activity evaluated by transient expression in HepG2 cells. Relative to the ref-erence, 3 haplotypes exhibited a 65% decrease in constitutive promoter activitywhile 2 exhibited a 50% decrease or a 40% increase in PXR-mediated, rifampicin-induced promoter activity.The flavin-containing monooxygenases (FMOs) also are important for xenobiotic

metabolism. FMO3 is expressed at levels comparable to CYP2C9 in the adult liverand is the predominant FMO enzyme in this tissue. However, significant interindi-vidual variation is observed. Using a similar approach as that used for CYP2C9,FMO3 population-specific SNP frequencies were determined and haplotypes wereinferred. Of the 7 common promoter region haplotypes identified, one resulted inan 8-fold increase in FMO3 promoter activity, while two exhibited a near completeloss of activity.

Although the in vivo significance of both the CYP2C9 and FMO3 regulatorypolymorphisms has yet to be determined, their ability to significantly alter their re-spective promoter activities would suggest such genetic variants will likely con-tribute to a risk phenotype for xenobiotics dependent on these enzymes for theirdisposition.

20 EPIGENETICS AND ITS IMPORTANCE INTOXICOLOGY.

R. L. Jirtle. Radiation Oncology, Duke University Medical Center, Durham, NC.

Human epidemiological and animal experimental data indicate that the risk of de-veloping adult-onset diseases, such as asthma, diabetes, obesity, and cancer, is influ-enced by persistent adaptations to prenatal and early postnatal exposure to environ-mental factors. Moreover, the link between what we are exposed to in utero anddisease formation in adulthood appears to involve epigenetic modifications likeDNA methylation at metastable epiallele and imprinted gene loci.

Genomic imprinting is an epigenetic form of gene regulation that results inmonoallelic, parent-of-origin dependent gene expression. The functional haploidstate of imprinted genes makes them susceptibility loci for diseases since a single ge-netic or epigenetic mutation can alter their function. We have recently developed acomputer-learning algorithm that identified 600 imprinted-candidate genes in themouse. Interestingly, humans are predicted to not only contain fewer imprintedgenes, but the repertoire is strongly species dependent (unpublished results). Bymapping the human imprinted-gene candidates onto the landscape of disease riskdefined by linkage analysis, we are now poised to determine the importance of im-printing in the etiology of complex human diseases.

Genes with metastable epialleles have highly variable expression because of sto-chastic allelic changes in the epigenome rather than mutations in the genome. Theviable yellow agouti (Avy) mouse harbors a metastable Agouti gene because of an in-sertion of a transposable element. We have used the Avy mouse to investigate the im-portance of nutrition in determining the susceptibility of offspring to adult dis-eases. We have shown that maternal dietary supplementation during pregnancy,with either methyl donors (i.e. folic acid, vitamin B12, choline and betaine) orgenistein, decreases adult disease incidence in the offspring by increasing DNAmethylation at the Avy locus. Moreover, these nutritional supplements can blockCpG hypomethylation caused by the endocrine disruptor, bisphenol A. (Supportedby NIH grants ES13053, ES08823, ES015165 and T32-ES07031, and DOE grantDE-FG02-05ER64101)

21 THE USE OF GENETIC VARIATION TO PREDICT RISKAND EFFICACY DURING DRUG DEVELOPMENT.

A. D. Roses. Joseph and Kathleen Bryan Alzheimer’s Disease Research Center,Durham, NC. Sponsor: R. Hines.

Safety and efficacy pharmacogenetics can be applied successfully to the drug dis-covery and development pipeline at multiple phases. Several examples will be dis-cussed. We have conducted screening using high throughput SNP associations withAlzheimer’s disease, testing more than 1,800 candidate targets with approximately7,000 SNPs in an attempt to identify novel drug-targets. A Phase IIB clinical trialfor Alzheimer’s disease will be described that exemplifies the major pipeline deci-sion between program attrition and further clinical development. In this case, therewas no significant improvement in 511 intention-to-treat patients but, applying aconfirmed prognostic biomarker (APOE4) to segment the clinical trial population,all three doses of rosiglitazone demonstrated improvement in patients who did notcarry the APOE4 allele. The data for the APOE4 carriers demonstrated no signifi-cant improvement but suggested that there may be a need for higher doses. Thus, adevelopment program that would have been terminated progressed to Phase III

registration trials based on the results of prospective efficacy pharmacogeneticanalyses. The implications of using APOE genotype as a biomarker to predict effi-cacy and possibly dose, as well as supporting the basic neurobiology and pharma-cology that provided the original target validation, will be discussed. Data will bepresented that support a slow neurotoxic effect of a specific fragment of apoE pro-tein on mitochondria and the use of rosiglitazone to increase mitochondrial bio-genesis and improve glucose utilization. In summary, pharmacogenetics is currentlybeing used across the pipeline to prevent attrition and to create safer and more ef-fective medicines.

22 A GENOMIC PATHWAY APPROACH TO COMPLEXDISEASES: PROMISE, PITFALLS, AND REMEDIES.

D. Maraganore. Mayo Clinic College of Medicine, Rochester, MN. Sponsor: R. Hines.

We recently described a genomic pathway approach to study complex diseases[Lesnick et al., 2007]. We illustrated this approach using the example of the axonguidance pathway and Parkinson’s disease (PD), and using available data from awhole-genome association study of PD [Maraganore et al., 2005]. We producedsingle nucleotide polymorphism (SNP) models that were highly predictive of PDsusceptibility (odds ratio = 90.8, p = 4.64 × 10−38), survival free of PD (hazardsratio = 19.0, p = 5.43 × 10−48), and PD age at onset (R2 = 0.68, p = 1.68 × 10−51). We were able to validate our findings for the axon guidance pathway and PDusing available SNP data from a second whole-genome association study of PD andalso using available probe set data from a whole-genome expression profiling studyof PD [Lesnick et al., 2007]. Our findings for the axon guidance pathway and PDhave biological plausibility. With respect to a possible brain wiring mechanism,they support the fetal basis for adult diseases (FeBAD) hypothesis (consistent alsowith animal models of neurodevelopmental toxicity and PD). With respect to apossible axonal maintenance and repair mechanism, they support the role of cumu-lative toxicity with aging. A genomic pathway approach might similarly providemechanistic insights into the causes of other diseases of relevance to toxicology.However, we acknowledge that findings for genomic pathways and complex dis-eases may prove difficult to replicate across diverse populations, particularly usingSNP markers. Furthermore, traditional experimental models may not embrace thecomplexity of the genomic pathway findings observed for human subjects. In lightof such pitfalls, we will discuss additional approaches that can be employed to vali-date findings for genomic pathway and complex diseases, again using the exampleof the axon guidance pathway and PD, with hopes of fulfilling the promise of thegenomic pathways approach.

23 NEW DEVELOPMENTS IN LIVER TUMOR BIOLOGY.

R. Roberts1 and Y. Dragan2. 1R&D Safety Assessment, AstraZeneca, Macclesfield,United Kingdom and 2R&D Safety Assessment, AstraZeneca, Wilmington, DE.

The liver is one of the most widely studied organs in cancer biology since it providesa flexible model for analysis of tumor biology that is amenable to in vitro and invivo correlative approaches. Because of this, many other fields of cancer biologyhave been informed by data emerging from studies of the liver. Over the last decade,the inter-related roles between the pathways of apoptosis, cell survival and cell pro-liferation and their relevance to toxicology have been more rigorously described.Key to this is the interplay between ligand-activated transcription factors such asthe peroxisome proliferators activated receptor α (PPARα) and the constitutivelyactive/androstane receptor (CAR) and the networks that initiate and execute apop-tosis. Against this background of knowledge, several new findings are particularlyresonant. Specifically, increasing evidence suggests a pivotal role for epigeneticmodification of DNA and its impact on survival and proliferation. Also of great rel-evance is the role played by the wnt-signalling pathway in cell signalling and tu-morigenesis. Another recent and informative approach involves the use of compar-ative genomics of human and model organism hepatocellular carcinoma to identifyaberrant phenotypes, reflecting molecular pathways that are evolutionarily con-served. This is further illustrated by use of a systems toxicology approach to analysisof liver tumors. This symposium brings together leading experts in human cancerbiology and liver toxicology to provoke cross-fertilization of ideas and will be of in-terest both to non-experts looking to gain a greater understanding of tumor biologyand those with a specific interest in the liver and mechanisms of carcinogenesis.

24 THE ROLE OF CELL SURVIVAL SIGNALLING INNORMAL AND TUMOR CELL BIOLOGY: LESSONSFROM LIVER.

R. Roberts. Safety Assessment, AstraZeneca, Macclesfield, United Kingdom.

There are multiple pathways implicated in regulating cell survival and apoptosis,and much of this has been elucidated in the liver as well as in other systems.Apoptosis can be induced via two main pathways, the stress/damage pathway and

SOT 2008 ANNUAL MEETING 5

the death receptor pathway. In general, stresses such as DNA damage, oxidativestress and chemical stress signal via the mitochondria to activate caspase 9 which inturn feeds into Caspase 3. In contrast, death ligands such as FasL and TNFa activatean extensive family of specific cell surface receptors and cascade through to caspase3 activation via the Death Inducing Signalling Complex (DISC). These pathwaysof induced cell death are highly relevant to our understanding of tumor cell biologyand will be described in order to set the scene for the subsequent presentations onsignalling, epigenetics, and carcinogenesis. Overall, key developments in elucidat-ing cross-talk between survival and proliferation networks is highly relevant to ourunderstanding of both normal and tumor cell biology.

25 THE ROLE OF THE WNT PATHWAY AND β-CATENININ NORMAL AND TUMOR CELL BIOLOGY.

S. A. Aaronson. Department of Oncological Sciences, Mount Sinai School ofMedicine, New York, NY. Sponsor: R. Roberts.

The evolution of multicellular organisms requires intercellular communicationfor processes including embryonic development, tissue differentiation, and sys-temic responses to wounds and infections. These complex signaling networks arein large part mediated by growth factors, cytokines and hormones. The Wnt fam-ily of highly conserved, secreted signaling molecules is essential in embryonic in-duction, cell polarity generation, and cell fate specification. Mutations in compo-nents, which constitutively activate Wnt canonical signaling, have been identifiedin several tumor types. Wnts bind to two co-receptors, the Frizzled-type seven-transmembrane-domain receptor and the low-density receptor related protein(LRP) 5/6. These interactions cause β-catenin stabilization through inhibition ofits phosphorylation by glycogen synthase kinase 3β (GSK3β). Phosphorylationof β-catenin normally targets it for proteosome degradation, and inhibition of itsphosphorylation leads to stabilization and translocation to the nucleus, where itforms a complex with a family of high mobility group (HMG)-like transcriptionfactors including leukocyte enhancer factor-1 (LEF-1) and T cell factors (TCF),and activates the transcription of target genes. Several naturally occurring, se-creted antagonists of Wnt signaling have been identified, and these act either tocompetitively inhibit Wnt/receptor interactions or by specifically interfering withsignaling by LRP5/6. It is well established that genetic aberrations leading eitherto loss of function of APC or activating mutations in β-catenin, which impair itsproteosome degradation, result in increased levels of uncomplexed β-catenin andconstitutive activation of Wnt canonical signaling in human malignancies includ-ing hepatocellular carcinoma. Current efforts are focused on how activation ofWnt canonical signaling plays such an important role in such a wide array ofhuman tumors

26 ORPHAN NUCLEAR RECEPTOR CONSTITUTIVEACTIVE/ANDROSTANE RECEPTOR (CAR)-MEDIATEDALTERATIONS IN DNA METHYLATION DURINGPHENOBARBITAL (PB) PROMOTION OF LIVERTUMORIGENESIS.

J. I. Goodman1 and J. M. Phillips2. 1Pharmacology and Toxicology, Michigan StateUniversity, East Lansing, MI and 2Biochemistry and Mollecular Biology, MichiganState University, East Lansing, MI.

Altered DNA methylation is an epigenetic mechanism that plays a key role in thecarcinogenesis process, and the nongenotoxic rodent hepatocarcinogen phenobar-bital (PB) alters the methylation status of DNA in mouse liver. The constitutive ac-tive/androstane nuclear receptor (CAR) mediates half of the PB-induced hepaticgene expression changes and it is essential for liver tumor promotion in PB-treatedmice. Methylation-sensitive restriction digestion, arbitrarily primed PCR, and cap-illary electrophoresis were utilized to detect PB-induced regions of altered DNAmethylation (RAMs) in CAR wildtype (WT) mice that are sensitive to promotionby PB and resistant CAR knockout (KO) mice, both on a C3H/He background.The CAR WT mice developed preneoplastic lesions after 23 wk of PB treatment(precancerous) and liver tumors after 32 wk, while the CAR KO mice did not de-velop tumors (Yamamoto et al., 2004). Our goal is to discern those RAMs that areplaying important roles in tumor formation by comparing the RAMs that form insensitive and resistant groups of mice. Using this novel approach, 42 unique RAMswere identified in the precancerous as compared to the CAR KO, 23 wk PB-treatedtissue. Of these 42 RAMs, 14 carried forward to the tumor tissue, and additionally,104 total unique RAMs were observed in the tumor tissue. These results indicatethat unique RAMs occur in the sensitive CAR WT mice and a subset of these isseen in both the precancerous and tumor tissue. Furthermore, 7 of the uniqueRAMs discerned in the precancerous liver and 13 of those in the tumor tissue werealso identified in the livers of B6C3F1 mice treated with a promoting dose of PB

for only 4 wk. We hypothesize that the unique RAMs may facilitate the tumorige-nesis process, and these data support the view that DNA methylation plays acausative role in PB-induced tumorigenesis. The implications for safety assessmentwill be discussed.

27 GENOMIC DECODING OF HUMAN LIVER CANCER.

S. S. Thorgeirsson. Laboratory of Experimental Carcinogenesis, NCI, Bethesda, MD.Sponsor: R. Roberts.

Much is known about both the sequential cellular changes that precede the forma-tion of hepatocellular carcinoma (HCC) and the etiological agents (i.e. HBV, HCVinfection, and alcohol) responsible for the majority of HCC. We have investigatedthe possibility that variations in gene expression of HCC at diagnosis would permitthe identification of distinct subclasses of HCC patients with different prog-noses.Unsupervised classification methods based solely on gene expression patternsrevealed two subclasses of HCC strongly associated with the length of patients’ sur-vival. Also, when the classifiers used in a training set to optimized classification ofthe tumors were applied to the validation set, all the classifiers successfully separatedpoorer survival patients (cluster A) from longer survival patients (cluster B). The variability in the prognosis of individuals with HCC may result from activa-tion of different oncogenic pathways during tumorigenesis and/or from a differentcell of origin. We have address whether the transcriptional characteristics of HCCcan provide insight into the cellular origin of the tumor. We integrated gene expres-sion data from rat fetal hepatoblasts and adult hepatocytes with HCC from humanand mouse models. Individuals with HCC who shared a gene expression patternwith fetal hepatoblasts had a poor prognosis. The gene expression program that dis-tinguished this subtype from other types of HCC included markers of hepatic ovalcells, suggesting that HCC of this subtype may arise from hepatic progenitor cells.Analyses of gene networks showed that activation of AP-1 transcription factors inthis newly identified HCC subtype might have key roles in tumor development.These findings support the notion that multiple molecular pathways dictate the de-velopment and different clinical outcomes of HCC. Therefore, use of gene expres-sion profiling promises to improve molecular classification and prediction of out-comes in HCC. Furthermore, molecular stratification of individuals with HCCinto homogeneous subgroups may provide opportunities for the development ofnew treatment modalities.

28 SYSTEMS TOXICOLOGY: APPLICATIONS TO RATMODELS OF CARCINOGENESIS.

Y. Dragan1, R. Beger2, R. Edmondson2, J. Fuscoe2, H. Fang2 and W. Tong2.1Safety Assessment-U.S., AstraZeneca, Wilmington, DE and 2Systems Toxicology,NCTR/FDA, Jefferson, AR.

A number of compounds can cause an increased incidence of liver tumors in ratsthrough mechanisms that may not be applicable in the human. In order to gain aclearer understanding of the mechanisms that underlie liver tumor promotion in therat, an integrated OMICs analysis was performed on the livers of rats administeredselected rodent liver carcinogens including mestranol, phenobarbital or tamoxifenfor six months. Microarray analysis was performed on RNA isolated from these liv-ers and 70 genes were changed by any one of the treatments. Proton NMR-basedmetabonomic analysis was performed and indicated that perturbation of lipids oc-curred with tamoxifen and phenobarbital administration. Proteomic analyses indi-cated that 125 proteins were altered in level between the tamoxifen-treated and con-trol liver. This integrated OMIC data analysis produced with ArrayTrack providesimportant insight into the mechanisms of liver carcinogenesis. Furthermore, thesystems toxicology approach is a mecahnism for generating safety biomarkers thatcan be used to bridge preclinical and clinical datasets. The perturbations of path-ways by agents can be compared using a systems toxicology approach.

29 INTRODUCTION.

D. S. Barber. University of Florida, Gainesville, FL.

Concern is increasing regarding the potential adverse health effects of materials inthe nanoscale range, including both ambient air pollutants and manufacturednanoparticles. The extremely high surface area of these materials relative to theirmass suggests the potential for interactions with biological molecules, resulting ineither inactivation or local concentration of the biological materials. Nanomaterialsurface properties, including charge, reactivity, hydrophobicity, and granularity,vary as a function of the source and chemical composition. There is, therefore, po-tential for some degree of selectivity in the interactions with biological materials,which may influence subsequent responses to the particles. This symposium focuseson recent investigations into these interactions and their biological implications.

6 SOT 2008 ANNUAL MEETING

30 THE ADSORPTION OF BIOLOGICALMACROMOLECULES BY COMBUSTION-DERIVEDNANOPARTICLES.

K. BéruBé1 and T. Jones2. 1School of Biosciences, Cardiff University, Cardiff, UnitedKingdom and 2School of Earth, Ocean and Planetary Sciences, Cardiff University,Cardiff, United Kingdom. Sponsor: J. Seagrave.

Protein separation by SDS-PAGE can be used to estimate relative molecular mass,to determine the relative abundance of major proteins in a sample, and to deter-mine the distribution of proteins among fractions. For biological purposes, SDS-PAGE can be used to evaluate the adsorptive capacity of biological macromoleculesby particulate dusts. The physicochemical properties of dust particles (i.e. size,shape, surface chemistry) plays an important role in their biological activity, in as-sociation with biological solutions (e.g. serum proteins and pulmonary surfactant),and hence, their impact of respiratory health. SDS-PAGE of proteins bound to par-ticles can be used to demonstrate whether particles selectively adsorb and retainthem. Differential absorption of lung lining fluids by combustion-derived nanopar-ticles (carbon black, fly ash, diesel particles) may play a role in their pathogenicmechanisms of action in the lung environment.

31 USING PROTEOMIC ANALYSIS TO DETERMINECOMPETITIVE BINDING OF PROTEINS TONANOPARTICLES.

S. C. Wasdo. University of Florida, Gainesville, FL. Sponsor: D. Barber.

Proteins readily adsorb to the surface of nanoparticles in a physiological environ-ment, and these proteins can strongly influence the nature and magnitude of ananoparticle-induced biological response. To more fully understand how varioussurface characteristics influence protein binding to nanoparticles, it is necessary toinvestigate the simultaneous and competitive binding of hundreds of proteins.Isotope labeling proteomics techniques, principally ICAT and ITRAQ, have beenused to identify and quantify the relative abundance of over 100 adsorbed proteinson various nanoparticles following their exposure to human plasma or serum.Characteristic patterns of protein adsorption were observed in nickel, aluminum,diamond and functionalized gold nanoparticles, including the preferential adsorp-tion of several complement components and other potentially inflammatory factorssuch as vitronectin. From these data, it is possible to estimate the relative bindingaffinities of the adsorbed proteins and model the process of competitive protein ad-sorption in complete physiological fluids.

32 SURFACE PROPERTIES CONTROL INTERACTIONSWITH CYTOSKELETAL ELEMENTS.

J. L. McGrath. Biomedical Engineering, University of Rochester, Rochester, NY.Sponsor: D. Barber.

We have modified the surface of polystyrene nanoparticles (NPs) with a variety offunctional groups and examined the internalization kinetics and processing of theseNPs by cultured endothelial cells and fibroblasts. Results reveal several principlesthat appear to simplify our understanding of NP fate during cellular interactions.First, we show that the capacity of NPs to bind serum proteins directly correlateswith the tendency of particles to be internalized by cells. Second, we show that par-ticles that are internalized through endocytosis steadily accumulate in perinuclearstores in a manner that is independent of NP surface chemistry but dependent oncytoskeletal activity. Finally, we show that NPs that emerge free in cytoplasm willinteract directly with elements of the cytoskeleton and that these interactions con-trol the diffusive movement of NPs within cells. The extent to which NPS in cy-tosol bind to either intermediate filaments or actin filaments varies with differentsurface chemistry.

33 MECHANISMS AND IMPLICATIONS OFPARTICLE/CHEMOKINE INTERACTIONS.

J. Seagrave. Lovelace Respiratory Research Institute, Albuquerque, NM.

Inflammation induced by inhalation of air pollutant particles has been implicatedas a mechanism for the adverse health effects associated with exposure to air pollu-tion. The inflammatory response is associated with upregulation of various pro-in-flammatory cytokines and chemokines. Both the neutrophil chemokines humanIL-8 and rodent MIP-2 become associated with diesel exhaust particles, and remainactive. This presentation will describe the mechanisms for adsorption of thesechemokines to DEP and other particles. The biological implications include localconcentration of the chemokines and prolonged retention.

34 ROLE OF IRON SEQUESTRATION IN HOST RESPONSETO PARTICLES AND FIBERS.

A. J. Ghio. Human Studies Division, National Health and Environmental EffectsResearch Laboratory, U.S. Environmental Protection Agency, Research TrianglePark, NC.

The host response to particles and fibers includes an increased local expression ofiron transport and storage proteins in respiratory epithelial cells which function tosequester iron away from the particle and fiber. Decreased iron complexation byfunctional groups at the surface of the particle or fiber decreases the oxidative stress.This response is recognized by the presence of ferruginous bodies in a tissue.

35 CURRENT PERSPECTIVES ON OCULAR AND SAFETYRISKS OF THERAPEUTICS.

M. D. Aleo and C. A. Somps. Drug Safety R&D, Pfizer Global R&D, Groton, CT.

Ophthalmic safety risks can occur from systemically administered therapeuticswhile systemic safety risks can occur from ophthalmic therapeutics. Such adversesafety effects identified during the research, development, and marketing phase of anew therapeutic agent may drastically limit further progression or marketing de-pending upon: the nature and severity of the finding, the stage of development, theproposed indication for the new drug, therapeutic index, and relevance of the lesionto humans based on mechanistic elucidation. The goal of this symposium is to pres-ent the Yin and Yang of managing ocular safety risks of systemic therapeutics andsystemic/ocular safety risks of ophthalmic therapeutics in the pharmaceutical in-dustry and potential methods for managing risks through mechanistic evaluation.Specific emphasis will be placed on the following concepts: 1) unexpected lenticu-lar toxicity associated with therapeutic agents intended for systemic administrationalong with ways to screen potential back-up candidates and risk assessment strate-gies, 2) unintended systemic toxicity associated with therapeutic agents intendedfor ophthalmic administration, 3) newest concepts and challenges associated withthe development of antisense, nanotechnology, and biologic-based therapeutics forophthalmic indications, implantable ophthalmic delivery devices and special for-mulations for ophthalmic delivery and 4) a perspective on how such cases may beviewed by medical reviewers for regulatory agencies. As a workshop this course willpromote interaction with the audience to share experiences in managing ocular/sys-temic safety risks in a pharmaceutical setting.

36 LENTICULAR TOXICITY OF SYSTEMICTHERAPEUTICS IN THE RAT: CASE STUDY ANDMECHANISM-BASED SCREENING APPROACHES.

M. D. Aleo. Drug Safety R&D, Pfizer Global R&D, Groton, CT.

In the pharmaceutical industry the discovery of ocular toxicity associated with ther-apeutic agents intended for prolonged systemic administration can lead to com-pound or program attrition before there is an opportunity to prove clinical efficacyor benefit at sufficient therapeutic margins of safety. Since safety evaluation of ther-apeutic agents in the rodent is required by regulatory agencies as a nonclinicalspecies for registration of first in human trials, the ability to extrapolate ocularsafety risks uncovered during safety testing in rodents is important. A long recog-nized mechanism of cataract liability in the rodent is related to inhibition of lentic-ular cholesterol biosynthesis. We have examined this mechanism across a samplingof rodent cataractogens using two distinct mechanistic assays (ex vivo lens choles-terol biosynthesis rate and cholesterol domain formation in model and rat nativemembranes). Information on the sensitivity and specificity of the assays will be pro-vided. A proposed risk management strategy will be presented for discussion basedon the mechanism of toxicity and its relevance to human exposure.

37 LENTICULAR TOXICITY OF SYSTEMICTHERAPEUTICS IN THE DOG: CASE STUDY ANDMECHANISM-BASED SCREENING APPLICATIONS.

C. J. Somps. Drug Safety R&D, Pfizer, Groton, CT.

During drug development, ocular lens toxicity may be observed in one or more pre-clinical animal models following systemic administration of candidate medicines.Understanding the mechanism(s) by which a systemically delivered compoundcauses lens toxicity can inform risk management strategies for that particular drugcandidate, and may be used to support higher throughput screening strategies forbackup compounds or alternative drug development programs. In this session wewill review the potential role of membrane ion transport mechanisms in lens toxic-ity by considering a case study in which systemically delivered drug candidates pro-

SOT 2008 ANNUAL MEETING 7

duced lens opacities in the dog which correlated with inhibition of potassium cur-rents in lens epithelium. This mechanism was consistent with differences in lensopacity observed across preclinical species, may inform human risk assessment, andprovided a path for developing an in vitro, higher throughput tool for screeningcompounds of similar structure for cataract risk.

38 SYSTEMIC CONSIDERATIONS WITH OCULARPHARMACOLOGIC THERAPIES.

R. B. Bhisitkul. Department of Ophthalmology, UCSF, Beckman Vision Center, SanFrancisco, CA. Sponsor: M. Aleo.

It is not generally well-recognized that severe unintended systemic toxicity can beassociated with administration of ophthalmic drugs. For example, ocular adminis-tration of chloramphenicol or timolol can be associated with clinical evidence ofaplastic anemia or alterations in cardiopulmonary function, respectively. This ses-sion will present the latest research regarding the possibility of an increased risk ofstroke with Lucentis and off-label use of Avastin. Clinical findings to date suggestthat even small amounts of an ophthalmic drug, even a biologic, administered intothe eye may have undesirable systemic side-effects in humans.

39 NONCLINICAL STUDIES FOR OPHTHALMIC DRUGDEVELOPMENT.

Z. Chen. U.S. FDA, Silver Spring, MD. Sponsor: M. Aleo.

This section will focus on nonclinical studies required for ophthalmic drugs. Nonclinical evaluation of ophthalmic drug products includes review of preIND, IND,and NDA submissions, and special study protocols. The evaluation is guided by re-lated FDA guidance, and the International Conference for Harmonization (ICH)guidelines. Ophthalmic drug products can be administered topically (as drops ap-plied to the eye), intravitreally, and by several other routes. Regardless of the modeof administration, ocular and systemic toxicity concerns need to be addressed fordeveloping these products. The pharmacokinetics of ophthalmic drugs, includingsystemic bioavailability and ocular distribution following topical exposure, shouldbe defined early in the development process. As a general rule, ophthalmic drugsshould be evaluated in general toxicology, genetic toxicology, reproductive toxicol-ogy, and, if applicable, carcinogenicity studies as defined by ICH guidelines. Tofully define the toxicity profile of ophthalmic drugs, general toxicology studiesshould be conducted using both ocular and systemic routes. Ocular toxicity shouldbe assessed in two species. To support human trials, the dosing frequency and drugconcentrations in the nonclinical studies should be at least equal and preferably ex-ceed the maximum frequency and concentration of clinical human studies. The as-sessment of ocular toxicity should include slit lamp biomicroscopy, funduscopy,tonometry, and histopathology of ocular tissues at appropriate times. If the drugbinds to melanin, animals with pigmented eyes should be used.

40 STRATEGIES FOR NONCLINICAL SAFETYASSESSMENT OF NOVEL OCULAR PRODUCTS.

D. J. Kornbrust. Preclinsight, Reno, NV. Sponsor: M. Aleo.

This session presentation will address general considerations, strategies and regula-tory expectations for achieving proper safety assessment of specific types of novelocular therapeutics, with a primary focus on oligonucleotide-based products (in-cluding siRNA programs), as well as other examples of ocular biotechnology-de-rived therapeutics and novel sustained release technology applications. Attendeesshould derive a sense for how a rational case-by-case approach to toxicity evaluationand effective collaboration with regulatory agencies can yield the most appropriatenonclinical support for enabling clinical trials.

41 THE LLNA, FALSE POSITIVES, FALSE NEGATIVES ANDALTERNATIVE ENDPOINTS.

D. Basketter. St John’s Institute of Dermatology, St Thomas’ Hospital, London, UnitedKingdom.

The local lymph node assay (LLNA) is being used increasingly in the identificationof skin sensitising chemicals for regulatory purposes. In the context of new chemi-cals legislation (REACH) in Europe, it is the preferred assay, as it is for an increas-ing number of regulations and regulatory agencies. An important part of the ra-tionale for this is that the LLNA offers a quantitative and objective approach to skinsensitisation testing allied with the important animal welfare benefits. However, aswith certain guinea pig sensitisation tests before it, this increasing use also brings

experience with an increasingly diverse range of industrial and non-industrialchemicals where the outcome of the assay does not always necessarily meet with theexpectations of those conducting it. Sometimes, the result appears to be a false neg-ative, but recently more common has been the suggestion that some chemicals rep-resent false positives. Typically, this situation can arise when the chemistry andguinea pig test history with similar materials suggests possible overestimation of thehazard. Against this background this session aims to review instances where falsepositive and false negative results have been described and will attempt to reconcilescience with expectation. The consequences of confusion surrounding this subjectwill be considered in relation to efforts to produce valid non-radioactive endpointsfor the LLNA as well as in the context of integrating potency measurement and riskassessments into classification and labelling schemes whose aim is to manage poten-tial risks to human health.

42 THE LOCAL LYMPH NODE ASSAY: INTRODUCTION,CONDUCT AND PERFORMANCE.

I. Kimber. Central Toxicology Laboratory, Syngenta, Cheshire, United Kingdom.

The mouse local lymph node assay (LLNA)is a method for the identification (andcharacterization) of chemicals with the potential to cause skin sensitization. Activityis measured as a function of the ability of test chemicals to cause an increase in theproliferative activity of of cells within lymph nodes draining the site of topical ex-posure. In the LLNA skin sensitizing chemicals are defined as those that elicit a 3-fold or greater increase in lymph node cell proliferation (measured by incorporationof radiolabelled thymidine) compared with concurrent vehicle-treated controls.Based upon extensive national and international inter-laboratory evaluations, anddetailed comparions with other available methods, the LLNA has been validated asa stand-alone method for the purposes of hazard identification. In the period sinceformal validation in both the US and Europe the LLNA has found increasing usefor assessment of skin sensitizing activity of a wide range of chemicals, and with thisaccumulating experience has emerged a more detailed understanding of the per-formance characteristics of the method. In tandem there has been interest in con-sidering the potential utility of alternative end points for the LLNA. The back-ground to these developments will be described and the current status of the assaydiscussed.

43 ALTERNATIVE METHODS AND ENDPOINTS IN“ENHANCED” LOCAL LYMPH NODE ASSAYS AND THEPHOTO-LLNA.

G. L. DeGeorge. MB Research Laboratories, Spinnerstown, PA.

The LLNA has evolved greatly from the original concepts of Basketter, Kimber, andGerberick, recognized as the co-inventors of the assay in the 1990s. Several alterna-tive endpoints and methods have been developed, e.g. Flow Cytometry-based (FC-LLNA), the IMDS test, and the Enhanced-LLNA (E-LLNA), as well as attempts atmeasuring induction of cytokine levels and release in ex-vivo methods using cul-tured LNCs. Earlier methods were of varied design and more crude endpoints, ie.“ear weight” and “node weight” for endpoints, and a 4 or 10-day study length incontrast to the current validated standard 6-day protocol. These earlier methodswere plagued by false-positives (due to irritancy properties of the test material) andconcordance values only in the mid-80% range, at best. Also, the ‘pooling’ of nodesfrom all animals in a treatment group did not allow for statistical analyses, includ-ing the variance and STDev to assist in identification of ‘outlier’ data points. Thenewer methods include quantitative measurements of ear thickness to better detectirritancy; immunophenotyping by flow cytometry to measure cell surface markersof lymphocyte activation; and the separate counting by flow cytometry to better ex-clude non-lymphocytes, RBCs, and other LN debris. Other factors, including se-lection of proper vehicle, and range-finding the ideal test doses will be discussed. Asummary or timeline of contributions or achievements from Europe, the USA andJapan will also be discussed, along with the latest discoveries, and promising futuremethods and protocols.

44 USE OF THE LOCAL LYMPH NODE ASSAY FORSENSITIZATION HAZARD IDENTIFICATION: A FIVE-YEARS EXPERIENCE FROM A REGULATOR’S POINTOF VIEW.

A. Oberemm, H. B. Richter-Reichhelm and U. Gundert-Remy. Safety ofSubstances and Preparations, Federal Institute for Risk Assessment, Berlin, Germany.Sponsor: D. Basketter.

Since clinical data are generally sparse, the evaluation of the potential of a substanceto cause skin sensitization in humans is mainly based on results prospectively ob-tained from animal studies. For decades, sensitization testing was performed in

8 SOT 2008 ANNUAL MEETING

guinea pigs (guinea pig maximization test/ GPMT; exceptionally: Buehler patchtest). In 2002, the local lymph node assay (LLNA) was accepted as an alternativestand-alone method. Here, we present data from German EU-notifications of newchemicals. Submission of significant numbers of LLNA reports started in 2003.Over time, the proportion of LLNA reports within totally submitted sensitizationstudy reports in Germany increased to about nearly 50% in 2005. However, no fur-ther increase was observed thereafter. This may depend on still existing reluctanceof notifiers, since the LLNA is presumed to produce an increased rate of false posi-tives compared to guinea pig testing. Accordingly, some companies still did not es-tablish the LLNA in-house and most of the studies are performed by contract labs.In one company, a non-radioactive LLNA was developed which considerably devi-ates from the standard LLNA, and makes it somewhat difficult to compare results.Regarding the technical performance and robustness of protocols, the LLNA is su-perior compared to GPMT, which demands more experience, resulting in morestudies showing methodological flaws compared to the LLNA. In addition, LLNAresults are often easier to evaluate compared to guinea pig data. In contrast to con-cerns that the LLNA would generally lead to on increase of positively evaluatedsubstances, the proportion of substances classified as sensitizers remained in thesame range. In conclusion, our experiences support the use of the LLNA for routineregulatory use. However, there exist substances with specific properties, which stillmay require testing in guinea pigs.

45 DISCORDANT RESULTS BETWEEN THE LLNA ANDGUINEA PIG TESTS: SPECIFICITY VS. SELECTIVITY.

A. Mehling2, 9, N. Ball1, 9, H. Certa8, 9, D. Eigler6, 9, H. Esch4, 9, C. Haux3, 9, R.Kreiling7, 9, K. Skirda9 and G. Veenstra5, 9. 1Dow Europe GmbH, Horgen,Switzerland, 2Cognis GmbH, Düsseldorf, Germany, 3Akzo Nobel SurfaceChemistry AB, Stenungsund, Sweden, 4BASF AG, Ludwigshafen, Germany, 5ShellInternational BV, The Hague, Netherlands, 6Goldschmidt GmbH, Essen, Germany,7Clariant Produkte (Deutschland) GmbH, Sulzbach, Germany, 8SASOL GermanyGmbH, Marl, Germany and 9CESIO, Brussels, Belgium. Sponsor: D. Basketter.

Although the local lymph node assay (LLNA), when using the [3H]-thymidine in-corporation method described in OECD 429, gives reliable results for many sub-stances, recent experience by Industry suggests it may overestimate or more rarelyunderestimate the sensitisation potential of some chemicals. Due to ethical aspectsand animal welfare laws, comparative or repeated testing for regulatory purposes isnormally not acceptable in the European Union. Therefore, defining unexpectedresults as false positives or negatives from an industry viewpoint can be difficult es-pecially with regards to regulatory acceptance. From a consumer perspective, riskassessments could be based on inaccurate data. Published data and hitherto unpub-lished data from industry presented here, point to some limitations of the LLNA.Unexpected positive results were observed in LLNA tests performed with classesgenerally regarded to lack a substantial sensitization potential, e.g. surfactants,siloxanes and polyols. Most of these chemical classes were poorly represented in theoriginal validation data set. Furthermore, discrepancies between other results ob-tained from guinea pig tests (GPTs), human experience and LLNAs were found.Various surfactants yielded unexpected positive results. High molecular weightpolyols induced positive reactions although structurally similar substances did notsensitise in GPTs. Conflicting results were also obtained with some siloxanes,aliphatic hydrocarbons, fatty alcohols and unsaturated fatty acids. The presence ofdouble-bonds in fatty acids as well as irritation may play a role but other mecha-nisms are feasible. These results indicate that the LLNA has limitations that requireimproved characterisation and understanding.

46 WHERE ARE WE NOW: TRUE AND FALSE POSITIVESAND NEGATIVES.

D. A. Basketter. St John’s Institute of Dermatology, St Thomas’ Hospital, London,United Kingdom.

The LLNA is now the de facto benchmark against which in vitro alternatives willlargely be judged. Accordingly, it is important to be clear about the quality of thisstandard, for which a proper appreciation of true false positives and true false nega-tives is vital. This presentation will seek to provide a digest of the preceding mate-rial and draw out the main issues and conclusions. In particular, the local lymphnode assay (LLNA) is being used increasingly in the identification of skin sensitis-ing chemicals for regulatory purposes. In the context of new chemicals legislation(REACH) in Europe, it is the preferred assay. The rationale for this is that theLLNA quantitative and objective approach to skin sensitisation testing, allied withthe important animal welfare benefits that the method offers. However, as with cer-tain guinea pig sensitisation tests before it, this increasing use also brings experiencewith an increasingly wide range of industrial and other chemicals where the out-come of the assay does not always necessarily meet with the expectations of thoseconducting it. Sometimes, the result appears to be a false negative, but rather morecommonly, the suggestion is that the chemical represents a false positive. We have

here reviewed a number of instances where false positive and false negative resultshave been described and have sought to reconcile science with expectation. Basedon these analyses, false positives and false negatives do occur in the LLNA, as theydo with any other skin sensitisation assay (and indeed with all tests used for hazardidentification), and that this occurs for a number of reasons. However, false positiveresults in the LLNA, as with the guinea pig maximisation test, can arise via failureto distinguish what is scientifically correct from that which is unpalatable, espe-cially where classification and labeling is entirely hazard based. The consequences ofthis will be discussed, particularly in relation to the need to integrate both potencymeasurement and risk assessments into classification and labelling schemes whichaim to manage potential risks to human health.

47 COMPUTATIONAL TOOLS AND TESTING METHODS:DELIBERATIVE AND JUDICIOUS USE IN RISKASSESSMENT.

M. Mumtaz1, P. Ruiz1, W. Stokes2, H. Hansen1 and C. De Rosa1. 1Division ofToxicology and Environmental Medicine, ATSDR/CDC, Atlanta, GA and2NICEATM, NIEHS, Research Triangle Park, NC.

Humans are exposed daily to chemicals in the environment, mostly as mixtures,through interaction with consumer products, through diet, and in occupational set-tings. The process of assessing the health risk associated with human exposure tosuch chemicals inevitably relies on a number of assumptions, estimations, and ra-tionalizations. Out of necessity, computational methods are increasingly being usedto meet the greatest challenges to extrapolate from conditions in experimental stud-ies providing evidence of toxicity of the chemical to those anticipated in the envi-ronment. However, integrating these methods into useful decision-making tools in-volves a thorough understanding of the inherent assumptions, the associateddatabases, and a critical interpretation of the results because they tend to give a falsesense of precision. This will be illustrated through the use of applications, atATSDR, of structure activity relationships (SAR) to evaluate mutagenicity and de-velopmental toxicity of Alkoxyethanols, and application of physiologically basedpharmacokinetic (PBPK) modeling for derivation of minimal risk levels (MRLs) ofmethylene chloride and to define threshold for interactions of binary chemical mix-tures.

Approaches and strategies that capture current biological knowledge, can help de-sign model directed research, promote development of new in vitro methods, andintegrate them with computational methods provide powerful alternatives and en-able the scientific community to use them to address real life issues. These can beused to fill research needs without the expense of extensive laboratory or field test-ing. Parallel with the developments in the scientific world and our understanding ofadvantages and limitations of the computational methods “virtual toxicity testing”without animal use can be set as a long term goal of toxicology.

48 MODULATION OF GENE EXPRESSION BY BTEX: RISKASSESSMENT IMPLICATIONS.

D. A. Sarigiannis, G. Cimino-Reale, E. Marafante, A. Collotta and A. Gotti.Institute for Health and Consumer Protection, European Commission - Joint ResearchCentre, Ispra, Italy. Sponsor: L. Birnbaum.

Benzene, toluene, ethylbenzene, and xylene (BTEX) are volatile monoaromatic hy-drocarbons commonly found in gasoline and in outdoor and indoor air mixtures.Environmental monitoring studies demonstrate significant differences in the meanexposure of the general population. Although exposure to VOCs has been associ-ated with health risks very little is known about the modulation of toxic responsesby the exposure levels of BTEX and the variation in its composition. In this studywe investigate the modulation of gene expression in vitro on human bron-choepitelial cells A549 and in vivo in the lungs of mice exposed i.tr. to two differentBTEX mixtures representative of European cities. Human and Mouse GenomeSurvey Microarray and TaqMan Low Density Arrays by Applied Biosystems wereused. Results from the in vitro studies demonstrated dose-response relationships forsub-toxic concentrations ranging from 10ng to 10 ug/l. The gene expression pat-tern correlates to different ratios of benzene/toluene in the BTEX mixtures. Similarresults were obtained from in vivo studies on mice at 24 hours following i.tr. expo-sure. Increasing concentration of toluene was a major factor in modulating the ex-pression of genes related to apoptosis, cell cycle control and proliferation, detoxifi-cation and inflammation. These results were confirmed by Quantitative PCR assaysand are consistent with the output of a physiologically-based pharmacokineticmodel of BTEX and its metabolites, which shows differences in both the transientand steady-state biologically effective doses of the BTEX components as a functionof mixture composition.Our findings suggest that toluene plays a critical role in thetoxicological response of the BTEX mixture. The composition of the mixture af-fects the bio-persistence of the primary compounds and their metabolites. This may

SOT 2008 ANNUAL MEETING 9

bear a strong impact on BTEX health risk assessment. For proper risk assessmentwe need to account not only for personal exposure levels, but also for differences inthe chemical composition of the mixture.

49 EXPLORING THE BOUNDARIES OF ADDITIVITY:MIXTURES OF NADH-QUINONE OXIDOREDUCTASEINHIBITORS.

J. W. Boyd, J. B. Patrone, N. T. Boggs, H. T. Le and M. L. Theodore.Eisenhower Research Center, Johns Hopkins University Applied Physics Lab., Laurel,MD. Sponsor: M. Trush.

Mitochondrial NADH:quinone oxidoreductase (complex I) is composed of up to46 subunits with a mass reaching 1000+ kDa, and is the largest and most compli-cated enzyme in the electron transport chain (ETC). Many complex I inhibitorstarget a single large binding domain with partially overalapping sites localized tothe PSST subunit which couples electron transfer from the iron-sulfur cluster N2to ubiquinone. In light of the shared binding site and the implied “common modeof action” of these inhibitors, our laboratory has tested the theory of dose-additionwith HepG2 cells exposed to mixtures of three classic complex I inhibitors:rotenone, deguelin and pyridaben. We chose two isoeffective mixture rays, basedupon the EC01 and EC50 values of the individual inhibitors, and examined all bi-nary and tertiary combinations. Additionally, the impact of upstream potentiation(via glycolysis inhibition) on downstream additivity was tested with co-administra-tion of 2-deoxyglucose in combination with mixtures of complex I inhibitors.

50 OPENREP: A COMMUNITY-BASED DELIBERATIONPLATFORM FOR DEBATING DIOXIN AND DIOXIN-LIKE CHEMICAL TOXIC EQUIVALENCY DATA ANDASSIGNED FACTORS.

L. D. Burgoon and T. R. Zacharewski. Toxicogenomic Informatics & Solutions,LLC, Lansing, MI.

Toxic equivalency factors (TEFs) are used to assess the relative risk of mixtures oftoxic polychlorinated dibenzo-p-dioxins, dibenzofurans and biphenyls relative to2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). The World Health Organization’sInternational Programme on Chemical Safety (IPCS) has selectively used relativeeffect potency (REP) data reported in the literature to establish TEFs with limitedinput from affected stakeholders, advocacy groups and data providers. OpenREPwas created to facilitate open and public deliberation, in order to capture commu-nity input using a wiki framework to support REP and TEF data presentation,stakeholder comments and facilitate the debate regarding the weighting of REP val-ues, as well as the (re)assessment of current TEF values. In OpenREP each toxiccongener has a Risk Assessment Dossier consisting of general chemical information,the current WHO TEFs, stakeholder-based recalculated TEFs, stakeholder discus-sion forums, and the in vivo and in vitro study summaries used to establish the cur-rent TEF values. General chemical information includes the chemical name, EntrezPubChem link, the IUPAC name, and structure information (canonical SMILES,and InChI structure codes). The in vivo and in vitro study summaries include REPvalues, assay system, route of administration, study quality weighting, and qualityweighting justification sections. Links to the PubMed record for each citation arealso listed. Stakeholders may also submit new literature for consideration. Curatorswill monitor the dossiers and discussion sections to ensure that posted data accu-rately reflect current stakeholder discussions regarding REP calculations and studyweights. All stakeholders will be able to participate in the deliberation process, andthe contents and deliberation materials will be freely available to government regu-lators and members of the WHO TEF committee. The underlying literature data-base will also be freely available to the scientific community.

51 A FRAMEWORK TO EVALUATE CHILD-ADULTDIFFERENCES IN INHALATION DOSIMETRY OFGASES: APPLICATION TO SELECTED SYSTEMICALLY-ACTING VOLATILE ORGANIC COMPOUNDS (VOCS).

L. T. Haber1, R. Gentry2, T. Adamou3 and K. Krishnan3. 1TERA, Cincinnati,OH, 2Environ, Monroe, LA and 3U Montréal, Montréal, QC, Canada.

We have developed a framework and analytical approach for evaluating the relativetissue dosimetry in adults and children for systemic effects of inhaled gases. Meaninternal dose in adults and children is compared, based on dose metrics for the ap-propriate chemical form. Such information can help to focus efforts for obtainingadditional data and for more refined analyses for classes of chemicals where there isthe greatest potential or likelihood of children being at greater risk. The results canbe combined with information on variability, to evaluate the adequacy of defaultuncertainty factors for protecting children, or to determine the need for chemical-

specific modifications to uncertainty factors. Case studies at steady state were con-ducted for systemic effects of VOCs to provide perspective on the potential range ofinternal dose in children at different ages (3 months, 1 year, 5 years, and 10 years)vs. adults for various combinations of physicochemical and metabolic characteris-tics. Calculations were conducted for the parent, reactive metabolite, and stablemetabolite for blood:air partition coefficients of 1-50, hepatic extraction ratio of 0-1, and for high or low intrinsic clearance (1000 or 0.1 L/hr). Illustrative analyseswere also developed on relative dosimetry for chemicals with flow-limited and en-zyme-limited clearance, and those that show significant age-related variability inenzyme capacity. These analyses indicate that in most cases, the ratio of internaldose for children:adults for these chemicals is likely to be within a factor of 2.3when the toxic moiety is the parent, and within a factor of 1.5 when it is a reactivemetabolite. Ratios were generally ≤1 for stable metabolites, but higher ratios wereobtained when the toxic moiety is a stable metabolite that is cleared efficiently inadults, but poorly in children. Overall, the framework provides a structure for eval-uating age-related kinetic differences and to focus future research in this area.

52 AGE-RELATED DIFFERENCES IN SUSCEPTIBILITY TOCARCINOGENESIS III. TENTATIVE CLASSIFICATIONFOR AGENTS WITH PUTATIVE NON-GENETICMODES OF ACTION, AND USE FOR ANALYSIS OFAGE-RELATED SUSCEPTIBILITY DIFFERENCES.

D. Hattis1, M. Chu2 and R. L. Goble1. 1George Perkins Marsh Institute, ClarkUniversity, Worcester, MA and 2NCEA, U.S. Environmental Protection Agency,Washington, DC.

This paper first proposes a system of categories/groupings for nonmutagenic modesof action for carcinogenesis, and then tentatively applies this system for analysis ofavailable age-related carcinogenesis bioassay data for six putative nonmutagenic car-cinogens (amitrole, DDT, dieldrin, diphenylhydantoin, ethylene thiourea, andpolybrominated biphenyls). The classification is of modes of action rather than in-dividual carcinogens, because the same compound can affe