An Offi cial Journal of the Society of Toxicology

www.toxsci.oxfordjournals.org

The Toxicologist Supplement to Toxicological Sciences

ISSN 1096-6080Volume 96, Number 1, March 2007

46th Annual Meeting and ToxExpoTMCharlotte, North Carolina

PrefaceThis issue of The Toxicologist is devoted to the abstracts of the presentations for the symposium, platform, poster discussion, workshop, and poster sessions of the 46th Annual Meeting of the Society of Toxicology, held at the Charlotte Convention Center, Charlotte, March 25–29, 2007.

An alphabetical Author Index, cross referencing the corresponding abstract number(s), begins on page 449.

The issue also contains a Keyword Index (by subject or chemical) of all the presentations, beginning on page 480.

The abstracts are reproduced as accepted by the Program Committee of the Society of Toxicology and appear in numerical sequence.

Copies of The Toxicologist are available at $45 each plus $5 postage and handling (U.S. funds) from:

Society of Toxicology 1821 Michael Faraday Drive, Suite 300

Reston, VA 20190

www.toxicology.org

© 2007 Society of Toxicology

All text and graphics © 2007 by the Society of Toxicology unless noted. The North Carolina photos are courtesy of Visit Charlotte and the photos of Seattle, Washington are courtesy of Washington State Tourism. All rights reserved. No text or graphics may be copied or used without written permission from the Society of Toxicology.

This abstract book has been produced electronically by ScholarOne, Inc. Every effort has been made to faithfully reproduce the abstracts as submitted. The author(s) of each article appearing in this publication is/are solely responsible for the content thereof; the publication of an article shall not constitute or be deemed to constitute any representation by the Society of Toxicology or its boards that the data presented therein are correct or are sufficient to support the conclu-sions reached or that the experiment design or methodology is adequate. Because of the rapid advances in the medical sciences, we recommend that independent verification of diagnoses and drug dosage be made.

SOT 2007 ANNUAL MEETING 1

1 REACH – A NEW FRAMEWORK FOR THEREGULATION OF CHEMICALS.

J. Galvin2, E. S. Williams1 and M. J. Dastin von Rijn3. 1Houston, ChemRisk,Houston, TX, 2Conoco-Phillips, Bartlesville, OK and 3U.S. Mission to the EU,Department of State, Brussels, Belgium.

The European Commission (EC) is increasing its regulation of chemicals through acomprehensive new program known as REACH (Registration, Evaluation, andAuthorization of Chemicals). The legislation, which is based on the precautionaryprinciple, shifts the burden from government to industry to prove that chemicalscan be used safely before they are marketed. As a part of this new system, manufac-turers and importers of chemicals will be required to submit hazard, use, and riskdata for many substances manufactured in or imported into the EU. Chemicalswhich exist in preparations or as part of articles sold in the marketplace will also besubject to these regulations. Thus, while REACH is a European regulation, it willimpact manufacturers worldwide. A key aspect of the program is the requirementto submit data concerning the physicochemical properties, toxicity, and environ-mental effects of those chemicals. Registration will be mandatory before a newchemical can be marketed, as well as to keep an existing chemical on the market,and chemicals of greatest concern will be subject to formal authorization proce-dures much like pharmaceuticals. The amount of information required for registra-tion will be proportional to the chemical’s health risks and production volumes.Chemicals that are classified as carcinogenic, mutagenic, toxic to reproduction, per-sistent, or bioaccumulative may be banned entirely if the manufacturer cannotdemonstrate that risks associated with the chemical can be controlled. REACH will impact the global chemicals industry and their downstream users. Theestimated costs to industry for complying with REACH have been reported to be ashigh as $64 billion ([| |]50 billion). Chemicals produced in quantities greater than1000 tons, as well as the most toxic chemicals, will have to be registered within 3years of the regulation entering into force, and all chemicals over 1 ton must be reg-istered within 11 years. Compliance with this regulation will require significanttoxicology and laboratory resources.

2 CONCEPTS, METHODS AND APPLICATIONS OFDISCOVERY AND INVESTIGATIVE TOXICOLOGY.

J. Davis1, D. Badger2, J. McKim3, J. Steidl-Nicholls4, B. Car5 and K. Kolaja6.1DSRD, PGRD, Chesterfield, MO, 2Drug Safety Evaluation, Allergan, Irvine, CA,3Cee-Toxicology, Kalamazoo, MI, 4DSRD, PGRD, La Jolla, CA, 5DiscoveryToxicology, Bristol-Myers Squibb, Princeton, NJ and 6Investigative Toxicology, Roche,Palo Alto, CA.

Drug development is a long, complex and expensive process. Typical developmenttimelines are between 10 and 15 years with attrition rates that are often too high forcompanies to sustain productive pipelines. Investigational and discovery toxicologyare extensions of the field of general toxicology, created to fulfill the growing needfor generating higher throughput, integrative, and predictive toxicological informa-tion, in an effort to reduce attrition at later stages of drug development. These novelideas have begun to be employed more frequently and its widely anticipated thatthis will pave the way for future drug testing paradigms. This course will providefirst an overview on the role of toxicology in discovery units including a descriptionof how the discovery toxicologist can support non-clinical drug development. Thenext lecture will cover in vitro methods and interpretation of cytotoxicity.Following this, an iterative approach to de-risking strategies of QT prolongationwill be discussed. The forth speaker will discuss strategies to partner with pharma-cology colleagues to probe and understand toxicology. The course will concludewith a lecture discussing emerging technologies that hold near-term potential forimproving drug safety evaluation. By the conclusion of this course it is anticipatedthat the attendees will be able to better assess which approaches would be most ap-propriate for their organization.

3 GENOMICS: FROM NOVICE TO EXPERT, FROMCHALLENGES TO PROMISES.

B. Sen1, J. Tucker2, S. Hester1, H. Auer3 and H. Hamadeh4. 1U.S.Environmental Protection Agency, Research Triangle Park, NC, 2National Institute ofEnvironmental Health Sciences, Research Triangle Park, NC, 3Columbus Children’sResearch Institute, Columbus, OH and 4Amgen, Inc., Thousand Oaks, CA.

Microarrays, when used properly, can be a powerful tool for genomics research. Thecomplexity of the technology provides numerous challenges and potential pitfalls tothe uninitiated. The FDA has developed guidelines for submission of genomicsdata and the EPA has guidelines for submission under development. It is becomingincreasingly necessary for toxicologists in many fields, including regulatory review,to have some level of understanding of the design and interpretation of microarraybased studies. The ability to utilize this technology and generate reproducible re-

sults requires a detailed understanding of the process. The first presentation willfocus on the design considerations of a successful microarray experiment. In a typi-cal microarray experiment, lists of significant differentially expressed genes are cre-ated. The common task faced by any researcher is to translate these gene lists into abetter understanding of the biological phenomenon involved. The second presenta-tion will discuss the use of various bioinformatic tools for data mining. From drugdiscovery, to a better understanding of the disease process, to unraveling the toxic-ity profile of unknown chemicals, genomics offers many promises. The second halfof the course will highlight the applications and limitations of this emerging tech-nology in disease and toxicity profiling. The third presentation will focus on the useof microarray in disease outcome prediction. The final presentation will examinethe potential of genomics in evaluating toxicity and identifying biomarkers. Thiscourse is aimed at providing investigators contemplating or pursuing studies in dif-ferential gene expression, and practicing toxicologists involved in data interpreta-tion for safety assessment and regulation, valuable knowledge of experimental de-sign, sample processing, data analysis, and an understanding of the scope andlimitations of this powerful tool.

4 ALLERGY AND ALLERGIC DISEASE: A PRIMER FORTOXICOLOGISTS.

I. Kimber1 and G. Gerberick2. 1Central Toxicology Laboratory, Syngenta, Cheshire,United Kingdom and 2Miami Valley Innovation Center, Procter & Gamble,Cincinnati, OH.

Allergy to chemicals and proteins is an important health issue, and some forms ofallergy are increasingly prevalent diseases. Moreover, there is no doubt that skin sen-sitization represents the most common manifestation of immunotoxicity in hu-mans. Allergy may take a variety of forms, those of greatest relevance to toxicologybeing allergic contact dermatitis resulting from skin sensitization, respiratory al-lergy and occupational asthma associated with both proteins and chemicals, gas-trointenstinal and systemic allergic responses caused by food proteins, and a varietyof local and systemic adverse reactions to drugs where allergic reactions are fre-quently implicated. The purpose of this Continuing Education Course is to ex-plore, in the context of toxicological sciences, and following a general introductionto the principles of allergy, each of the four main types of allergic disease mentionedabove. In each case clinical manifestations of disease, biological mechanisms, andapproaches available for toxicological evaluation will be described. This Basic Levelcourse will be of value to immunotoxicologists, and also to those with interests inrespiratory, skin and gastrointestinal toxicology, occupational medicine, the devel-opment of approaches to hazard characterization, and risk assessment.

5 FUNDAMENTALS OF HUMAN HEALTH RISKASSESSMENT WITH A CASE STUDY APPROACH.

T. Leavens1, J. Orme-Zavaleta2, G. Foureman2, E. Cohen-Hubal2, H. Clewell1

and M. Gargas3. 1CIIT Centers for Health Research, Research Triangle Park, NC,2U.S. EPA, Research Triangle Park, NC and 3The Sapphire Group, Inc.,Beavercreek, OH.

This course will provide a general overview of the process of human health risk as-sessment to educate students and post doctorates interested in careers in risk assess-ment and researchers interested in how their work can impact and strengthen thescience upon which the assessments are based. Risk assessment is used by toxicolo-gists to estimate the likelihood of an adverse event from exposure to a hazardousagent. In 1983 the National Academy of Sciences-National Research Council pub-lished “Risk Assessment in the Federal Government: Managing the Process” inwhich the process of risk assessment was clearly outlined into 4 components in-cluding hazard identification, dose-response assessment, exposure assessment, andrisk characterization. This course will focus on each of these 4 components and willuse vinyl chloride, especially the U.S. EPA IRIS assessment, as a case study to illus-trate the development and application of the 4 components. The presentations willcover historical studies incorporated into hazard identification including acute,subchronic, and chronic animal assays as well as epidemiology studies, in vitro as-says, predictive models, and genomics research. Information will also be presentedon determination of the mode-of-action from non tumor data, and how it mayguide and direct the assessment process to reduce uncertainty in extrapolating ef-fects. Various methods of dose-response assessment will be presented including sta-tistical and mechanistic models used to extrapolate effects among doses and species.Available exposure factor data, exposure modeling tools, biomonitoring data, andpotential sources of variability and uncertainty in exposure assessment will be high-lighted. The final component, risk characterization, will address statistical and bio-logical issues, such as appropriate models and target populations, as well as connectthe process of risk assessment with risk management.

SOT_2007.final 2/28/07 10:09 AM Page 1

2 SOT 2007 ANNUAL MEETING

6 TOXICOLOGY AND MOLECULAR BIOLOGY OF TISSUEREPAIR.

H. M. Mehendale2 and C. Corton1. 1NHEERL, U.S.-EPA, Research Triangle Park,NC and 2Department of Toxicology, University of Louisiana, Monroe, LA.

After chemical exposure, injury may progress unabated leading to tissue and organdysfunction and failure, or regress leading to recovery. When recovery occurs, tissueinjury resolves as a result of cell division, restored tissue architecture, structure andfunction. The molecular biology of these events, and their dynamic relationships toprogression or regression of tissue injury are of significant mechanistic interest, be-cause risk of adverse effects from chemical exposure should consider not only thelevel of tissue damage but the dynamics of recovery from that damage. In addition,interventional therapeutic opportunities may be developed through this under-standing. These dynamic events apply to acute injuries such as ischemia-reperfu-sion, drug or toxicant-induced necrosis, or chronic tissue injuries. The first lecturewill cover the sequential events that occur during tissue regeneration using recentgenomic analyses of the partial hepatectomy model and other models of tissue re-generation. The second lecture will describe various models of liver injury, dose-re-sponse relationships and modulations of injury and progression leading to hepaticfailure, and regression of injury leading to recovery from injury. The third lecturewill cover the cellular and biochemical changes associated with renal injuries, andhow injury initiated by drugs, toxicants, and ischemia-reperfusion may progressleading to renal failure. Both renal as well as hepatic injury and tissue repair modelswill be discussed in the context of modulation by disease, dose, duration of is-chemia before reperfusion, and nutritional factors. The last lecture will cover therole of the dynamic events of tissue destruction and restoration focusing on hydrol-ysis of extracellular matrix components by activated matrix metalloproteinases inpreparation for restoration of tissue architecture. The role of angiogenesis in restor-ing extracellular matrix, cell mobility and placement are dynamic and critical eventsneeded in re-establishing tissue homeostasis.

7 THE BIOLOGY AND TOXICOLOGY OF THE PERI- ANDPOST-NATAL DEVELOPMENT.

D. J. Auyeung-Kim1, R. Chapin2, R. Tyl3, L. W. Jackson4, G. Cappon2 and S.Makris5. 1Toxicology, Charles River Laboratories, Preclinical Services, Sparks, NV,2Pfizer Global Research & Development, Groton, CT, 3Center for Life Sciences &Toxicology, RTI International, Research Triangle Park, NC, 4Epidemiology &Biostatistics, Case Western Reserve University School of Medicine, Cleveland, OH and5ORD NCEA, U.S. Environmental Protection Agency, Washington, DC.

The period of post-natal development has always been of societal, scientific, andregulatory concern. This concern is heightened by the growing literature whichshows that exposures and health status in early (prenatal and post-natal) life canhave significant health consequences much later in life; in humans, this may mean50-60 years later. This basic course will begin with a review of the recent data show-ing the connection between early-life events and later-life health, and move on to adiscussion of the comparative developmental biology of rodents and humans. Thethird talk will describe regulatory and tailored study designs which address peri-natal exposures and adult outcomes and discuss endpoint selection. The last speakerwill address the issues and challenges of data interpretation and regulatory action.Attendees will take home an appreciation of the complex biology, the length of thehealth trajectory that is set up by very early exposures, and the challenges of dealingwith these issues from a regulatory perspective.

8 TOXICOLOGICAL EVALUATION OF CHEMICAL ANDDRUG MIXTURES.

C. Borgert1, J. Simmons2, A. Constan3, M. Mumtaz4 and K. Krishnan5.1Applied Pharmacology and Toxicology Inc., Gainsville, FL, 2U.S. EPA, ResearchTriangle Park, NC, 3Infinity Pharmaceuticals, Cambridge, MA, 4ATSDR, Atlanta,GA and 5SEST, University of Montréal, Montréal, QC, Canada.

Drugs, consumer and personal care products, pesticides, environmental chemicals,and foods are often assessed for safety and risk without extensive consideration ofnumerous potential pharmacological and toxicological interactions that mightoccur as these agents are encountered as mixtures by patients, consumers, andthrough environmental exposures (e.g. mixtures present in air, water, soil). Such as-sessments are complex due to the potential pharmacokinetic and pharmacody-namic mechanisms that underlie interactions in mixtures, necessitating study de-signs, methods of analysis, and limits on interpretation not required for singlechemicals. This course will cover the fundamentals of study design and data analy-sis for mixtures that apply to all uses of chemicals, regardless of market application.The first presentation will cover the basic principles underlying component-basedand whole mixture techniques as well as study designs available for evaluating thetoxicity of mixtures. The second presentation will review the concepts for evaluat-

ing data on mixtures, including data on binary and higher order interactions, andwill apply those concepts to various examples from the published scientific litera-ture. The third talk will focus on the pharmacokinetic and pharmacodynamicmechanisms of interactions among drugs and chemicals. The final talk will presentthe current approaches and guidance for conducting risk assessment of chemicalmixtures and safety assessment of drug combinations. The course will be of interestto experimentalists who wish to conduct studies on mixtures that are more proba-tive and relevant for safety and risk assessment, as well as for safety and risk assessorswho must evaluate and apply data on mixtures and interactions in assessments.

9 PHYSIOLOGICALLY BASED PHARMACOKINETICMODELING FOR RISK ASSESSMENT APPLICATIONS.

H. A. Barton1 and J. G. Teeguarden2. 1ORD National Center for ComputationalToxicology, U.S. Environmental Protection Agency, Research Triangle Park, NC and2Biological Monitoring and Modeling, Pacific Northwest National Laboratory,Richland, WV.

Pharmacokinetics is a key step in mechanistically based risk assessments, linkingchemicals exposures with the mode of action leading to toxicity. Physiologicallybased pharmacokinetic (PBPK) modeling is one approach for incorporating phar-macokinetics into risk assessment, particularly quantitative dose-response analyses.This course will guide the student through the PBPK modeling process, frommodel development through the use of the modeling results in risk assessment. Thebasic principals of PBPK model development will be described, focusing on therange of different model structures that have been developed to address compoundswith a variety of physicochemical characteristics, toxicological properties, and expo-sure pathways. Models for endocrine active compounds will be described to illus-trate issues of serum protein and receptor binding and modeling in vitro and invivo systems. Characterizing variability of the human population can be incorpo-rated in PBPK models using distributions of physiological and chemical-specificparameters (e.g., metabolism)and Monte Carlo methods. Bayesian statistical ap-proaches implemented with Markov Chain Monte Carlo analyses and maximumlikelihood methods will be described as these have been the major approaches tocharacterizing uncertainty in model estimated dose metrics. Qualitative and quan-titative evaluations of PBPK models are a critical step in their application in risk as-sessments. A variety of risk assessment applications (e.g., route-to-route extrapola-tion) and examples of using PBPK models to develop referenceconcentrations/doses and cancer unit risks will be presented.

10 PROTEIN THERAPEUTICS: ASSESSMENT AND IMPACTOF IMMUNOGENICITY ON STUDY DESIGN ANDINTERPRETATION.

Y. Ouyang1, L. LeSauteur2, D. Wierda3, D. Finco-Kent4, R. House5 and S.Kirshner6. 1CDER/OND/OODP/DMIHP, U.S. FDA, Silver Spring, MD,2Laboratory Sciences, Preclinical Services, Charles River Laboratories, Senneville, QC,Canada, 3Biopharmaceutical Immunotoxicology, Eli Lilly & Company, Greenfield, IN,4World Wide Safety Sciences, Pfizer Inc., Groton, CT, 5DynPort Vaccine CompanyLLC, Frederick, MD and 6CDER/OPS/OBP/DTP, U.S. FDA, Bethesda, MD.

Immunogenicity is a unique and significant safety and efficacy concern in develop-ing protein therapeutics and could have significant impact on safety, pharmacoki-netics, and efficacy. Anti-erythropoietin (EPO) antibodies-elicited pure red cellaplasia has highlighted the significant impact of immunogenicity on human safetyof protein therapeutics. Immunogenicity is also a critical issue in designing and in-terpreting animal toxicity studies. The purpose of this basic CE course is to high-light important aspects of immunogenicity including an overview of current FDAregulation with regards to nonclinical and clinical immunogenicity assessment.Specifically, the course will address what makes protein therapeutics potentially im-munogenic, and how an antibody response is produced. The development and val-idation of common methods used to detect an antibody response will be presentedby describing development/optimization parameters and validation criteria for im-munogenicity assays, followed by case studies describing antibody screening ap-proaches, and total and neutralizing antibodies determination. This course will alsoreview the current practice in immunogenicity evaluation in animal models and re-cent development of new animal models. The consequences of immunogenicitywill be discussed using specific case studies highlighting safety evaluation, risk as-sessment, potential clinical consequence and the impact of immunogenicity on de-signing and interpreting toxicity studies. With the current drug development ap-proach on targeted therapeutics, protein therapeutics are becoming increasinglyimportant. Immunogenicity is a challenge for anyone involved in immunotoxicityresearch and in the research and development of protein therapeutics.

SOT_2007.final 2/28/07 10:09 AM Page 2

SOT 2007 ANNUAL MEETING 3

11 RODENT IMAGING OVERVIEW FOR THETOXICOLOGIST.

K. Gabrielson1, R. Maronpot2, B. Tsui3 and P. Contag4. 1Molecular andComparative Pathobiology, Johns Hopkins University, Baltimore, MD, 2Laboratory ofExperimental Pathology, National Institute of Environmental Health Sciences,Research Triangle Park, NC, 3Radiology, Johns Hopkins University, Baltimore, MDand 4Imaging, Xenogen Corporation, Alameda, CA.

In the last decade, in vivo imaging methods have become established tools in basicscience research. This course is designed to provide a general overview of imagingmodalities utilized in basic science with applications to drug discovery and toxico-logical research. This review will focus on five imaging modalities including: MR,PET, SPECT, optical and ultrasound. Each speaker will give a basic overview of themodality and present specific examples relevant to drug discovery and toxicologicalresearch. In this review, several applications of MRI, Magnetic ResonanceSpectroscopy and Electron Paramagnetic Resonance in oncology, cardiovasculardisease and morphological phenotyping during toxicological studies will be de-scribed. PET and SPECT imaging of radioactively labeled compounds will be re-viewed in the setting of drug discovery and toxicological research. The applicationof the multi-modality imaging scanners will also be discussed. These multi-modali-ties studies provide comprehensive evaluation that enables the toxicologist to studystructure, function, in tandem with distribution and metabolism of radioactivelylabeled drugs. PET imaging of gene expression during pharmacology studies will bereviewed. Optical imaging will be reviewed in the context of imaging gene expres-sion in transgenic mice engineered with gene promoters from enzymes (CYP1A2 orCYP3A4) driving luciferase expression. Signal transduction pathway discovery intoxicological research will be discussed using optical imaging in transgenic micewith a Smad-responsive luciferase reporter. Ultrasound imaging will be overviewedusing applications of toxicity and drug discovery in cardiovascular and cancermouse models. Following completion of this workshop, attendees should have basicunderstanding of the various imaging modalities and applications to drug discoveryand toxicological research.

12 APPLICATION OF HUMANIZED MOUSE LINES TOTOXICOLOGICAL AND PHARMACOLOGICALSTUDIES.

R. S. Pollenz. Biology, University of South Florida, Tampa, FL.

A major challenge to human risk assessment is the translation of toxicological andpharmacological results produced in rodents to human populations. This is furthercomplicated by the fact that many toxicologically relevant enzymes and receptorsexhibit differences in response to activating compounds that are species specific. Inaddition, for any chemicals that have carcinogenic properties, it is often impossibleto obtain results in human populations for comparisons to rodents. In an effort toaddress some of these issues, “humanized” mouse lines are being generated in whichhuman genes have been used to replace orthologous mouse genes. Such models aretermed “knock-in” mouse lines. Alternatively, mouse lines can be generated inwhich human hepatocytes have been transplanted into mice. In these studies itoften possible to obtain mice that show 70-98% replacement ratios of the mousehepatocytes. The advantage to such humanized mouse lines is that they are likely toshow phenotypes closer to that of humans and be applicable to a wide variety of invivo studies. Such lines make it feasible to carry out risk assessment studies in ro-dents that may be directly applicable to human populations. This symposium ses-sion is aimed at exploring the generation and analysis of mouse models harboringhCYP1A1, hCYP1A2, pregnane X receptor (hPXR), and the peroxisome prolifera-tor-activated receptor (hPPAR). These studies will be contrasted to those in miceharboring human hepatocytes. These novel animal models have relevance to phar-macology, toxicology, drug development and human risk assessment.

13 MOUSE METABOLOMICS IN THE SEARCH FORENDOGENOUS P450 METABOLITES AND METABOLICPROFILING.

F. Gonzalez1, C. Chen1, K. Krausz1 and J. Idle2. 1NCI, Bethesda, MD and2Charles Univeristy, Prague, Czech Republic. Sponsor: D. Nebert.

Metabolomics is the study of changes in small molecules in cells, tissues and bio-logical fluids such as serum and urine. Metabolomics, using 1H-NMR, has gainedprominence and been embraced by the pharmaceutical industry as a means to de-termine the potential toxic effects of drugs. For example, a chemical is administeredto rodents and urine examined for the presence of metabolites that reflect organ-specific toxicity. In an attempt to find biomarkers for cytochromes P450 expres-sion, P450-null and humanized mice were compared. By use of metabolic cages,urine was collected and subjected to ultra performance liquid chromatography cou-pled time-of-flight mass spectrometry (UPLC-TOFMS). The UPLC-TOFMS data

were deconvoluted by principle component analysis software in order to find com-pounds that correlate with the presence or absence of a specific P450 form. Tostudy drug metabolism, drugs were administered to mice and serum and urine ex-amined for the production of metabolites that were not found in untreated mice. Inthis manner, metabolites derived from the drug and endogenous metabolites thatreflect the drugs biological activity or toxicity can be resolved. This high resolutionof MS-based metabonomics offers great promise in the fields of drug metabolismand toxicology.

14 PREGNANE X RECEPTOR (PXR) HUMANIZED MICE INTOXICOLOGY STUDIES.

W. Xie. University of Pittsburgh, Pittsburgh, PA. Sponsor: D. Nebert.

The regulation of hepatic CYP enzymes is important in drug metabolism and toxi-cological studies. The CYP genes are up-regulated by numerous foreign chemicals(xenobiotics), yet the inducibility shows distinct species specificity. The orphan nu-clear receptor PXR (NR1I2) in the mouse and its human ortholog, have been es-tablished as species-specific xeno-sensors that regulate CYP3A enzymes. By knock-ing-out the mouse Nr1i2 gene (encoding mPXR) and replacing it with the humanNR1I2 gene (encoding hPXR), a “humanized” mouse model has been established.In addition to the original liver-specific hPXR mouse line, we have recently alsocreated a mouse line with hPXR expressed in both the liver and intestine. We havefound that the humanized mice readily responded to human P450 inducer drug ri-fampicin in the equivalent range of the standard oral dosing regimen in humans(300-600 mg per 70-kg man). These humanized mouse lines, which display humandrug response profiles, represent unique tools for dissecting responses to drugs andforeign chemicals and should aid in the development of safer drugs. —-Supportedin part by NIH grants ES012479 and CA107011.

15 REPOPULATION OF MOUSE LIVER WITH HUMANHEPATOCYTES: CHARACTERIZATION OF THISMOUSE LINE AND USE IN DRUG METABOLISIMSTUDIES.

K. Yoshizato. Biological Science, Hiroshima University, Higashihiroshima,Hiroshima, Japan. Sponsor: D. Nebert.

Chimeric mice were produced by transplantation of human hepatocytes into im-muno-deficient and liver-diseased mice (urokinase-type plasminogen activator-transgenic severe combined immune deficiency mice). These animals were normaland survived for more 100 days post-transplantation. Some animals were given bro-modeoxy uridine (BrdU). Liver sections were stained with antibodies againsthuman cytokeratin 8/18 (hCK8/18) and BrdU. The hCK8/18-positive-hepatocytecolonies gradually increased their size. Some chimeric mice showed an almost-com-plete replacement with human hepatocytes after 9 weeks post-transplantation.Approximately a half of the chimeric mice had a replacement index (RI, the num-ber of hCK8/18-positive human hepatocytes divided by total number of hepato-cytes) of more than 70%. BrdU-labeling index was about 9% at 1 week and gradu-ally decreased to a background revel at 11 weeks post-transplantation. The chimericmice showed normal liver lobules and sinusoid structures: the sinusoids were linedup with mouse endothelial cells; the space of Disse was occupied with mouse stel-late cells; mouse Kupffer cells were distributed on the luminal aspect of the en-dothelium. The human hepatocytes in the chimeric liver expressed major humancytochrome P450 (hCYP) subtypes in a manner nearly equivalent to those of theoriginal donor hepatocytes. hCYP3A4 and hCYP1A1/1A2 were specifically in-duced in the liver by rifampicin and 3-methylcholanthrene, respectively. The pro-file of bile acid composition of taurocholic and glycocholic acids was closer tohuman than to mouse. Efficient infection and propagation of both hepatitis B andC viruses were observed in the chimeric mice: known anti-hepatitis virus agentsdramatically decreased the level of viremia in the infected mice. This animal modelwill have wide usefulness for human liver-related studies, especially for predictingthe metabolism of drugs in the human liver and to screen an anti-human hepatitisvirus drug. —-Supported in part by CLUSTER Project.

16 REPLACEMENT OF THE MOUSE CYP1A1 AND CYP1A2GENES WITH THE HUMAN HCYP1A1_CYP1A2 LOCUS.

D. W. Nebert, S. Derkenne, T. Doetschman and T. P. Dalton. Department ofEnvironmental Health, University Cincinnati Medical Center, Cincinnati, OH.

The human CYP1A1 and CYP1A2 genes are oriented head-to-head, sharing a com-mon 5’-flanking region of 23.3 kb. The mouse Cyp1a1 and Cyp1a2 genes are alsooriented head-to-head and share a common 5’-flanking region of only 13.3 kb; thisregion is so small that one would need to generate thousands of offspring from a

SOT_2007.final 2/28/07 10:09 AM Page 3

4 SOT 2007 ANNUAL MEETING

Cyp1a1(-/-) x Cyp1a2(-/-) cross to obtain a double-knockout. Using Cre-mediatedinterchromosomal recombination between loxP sites inserted at the 3’ ends of boththe Cyp1a1 and Cyp1a2 genes, we have succeeded in generating a Cyp1a1/1a2(-/-)double-knockout mouse line. This lab has also created a bacterial artificial chromo-some (BAC)-transgenic mouse line, carrying both human CYP1A1 and CYP1A2genes, the 23.3-kb spacer region, >55 kb of CYP1A1 3’ flank, and >85 kb ofCYP1A2 3’ flank. Both lines have been backcrossed into C57BL/6J (B6) so thattheir genomes comprise >99.8% B6 genes. The humanized hCYP1A1_1A2line––in the absence of the mouse Cyp1a1 and Cyp1a2 orthologous genes––ex-presses human CYP1A1 and CYP1A2 mRNA, protein and enzyme activities in alltissues in a pattern very similar to that of the mouse: basal CYP1A1 is nil, whereasTCDD-inducible CYP1A1 is robust and occurs in all tissues examined; basalCYP1A2 is prominent in liver, whereas TCDD-inducible CYP1A2 occurs in liver,duodenum, stomach and lung. Theophylline was chosen to prove human P450participation in metabolism; the drug is predominantly metabolized by CYP1A2 inhumans and rodents. Replacement of mouse Cyp1a2 with the functional humanCYP1A2 gene restored the capacity to metabolize theophylline, and the metabolismchanged from the mouse to the human profile (i.e. 3-methylxanthine formationwas now seen, which does not occur in wild-type mice). This humanizedhCYP1A1_1A2 BAC-transgenic line will be helpful in estimates of human risk as-sessment, following exposure to a variety of drugs, carcinogens, and other environ-mental pollutants that are CYP1A1 and/or CYP1A2 substrates or inducers.––Supported in part by NIH grants R01 ES08147 and P30 ES06096.

17 GENOMICS AND DOSE-RESPONSE MODELING.

M. Andersen. CIIT Centers for Health Research, Research Triangle Park, NC.

In recent years, with the completion of the Human Genome project and evolvingcapabilities and technologies in the area of genomic research, there has been a sig-nificant increase in information about consequences of chemical exposure on geneexpression. While these types of information are currently being used qualitativelyto make decisions, the challenge remains to be able to incorporate genomic infor-mation quantitatively into the risk/safety assessment process. The main challengefacing scientists in the regulatory arena is not only attempting to identify a genomicsignature or threshold associated with chemical exposures, but to also correlate thisinformation with a phenotypic threshold, whether that is at the level of a protein,cell, tissue or behavior. The purpose of this symposium is to present examples of thegathering of dose-dependent genomic information on toxic compounds and to dis-cuss how this information might be considered in dose-response modeling for riskassessment application. Several case studies will be presented, as well as reviews ofthe types of information or studies needed to facilitate the consideration of ge-nomic data in dose-response modeling.

18 THE CURRENT STATE OF THE USE OF GENOMICINFORMATION IN RISK ASSESSMENT.

R. J. Preston. U.S Environmental Protection Agency, Research Triangle Park, NC.Sponsor: R. Gentry.

The recently released U.S. EPA Guidelines for Carcinogen Risk Assessment providea framework for the consideration of mechanistic data in the cancer risk assessmentprocess. A tumor is induced by a set of key events that are required for the conver-sion of a normal cell into a transformed one and ultimately to produce a malignantphenotype. The key events for a particular chemical are described by the mode-of-action (MoA) for that chemical; for example, DNA-reactivity, cytotoxicity and re-generative cell proliferation, and receptor interactions. It is likely that multiplemodes of actions are operating for any particular chemical. Given the complexity oftumor formation and that multiple key events are involved, it is of considerablevalue, if not essential, to assess effects of potential or known carcinogens at thewhole-genome level, and at the tissue/organ level. It is this systems (cell or tissue)approach that is likely to be the most fruitful for assessing cancer risks at low, envi-ronmental exposure levels. In this way, it is feasible to enhance the ability to con-duct the various extrapolations that have to be a part of the risk assessment process.These extrapolations include high to low dose, animal to human and tumor site tosite. Genomics data at the DNA, mRNA and protein levels have already been in-formative in the identification of specific genetic and phenotypic alterations associ-ated with tumor formation. This approach can lead to the development of inform-ative biomarkers of tumor response and their use as surrogates for tumors for thetypes of extrapolation noted above. The use of genomics approaches is just enteringthe realm of risk assessments; there remains considerable groundwork before aclearly defined use can be established. (This abstract does not necessarily reflectU.S. EPA policy).

19 GENOMIC CONSIDERATIONS IN THE DOSE-RESPONSE MODELING FOR FORMALDEHYDE.

M. Andersen. CIIT Centers for Health Research, Research Triangle Park, NC.

This presentation will provide a discussion of the development of a mechanisticdose-response model with genomic data for formaldehyde. Inhalation exposures inrats were conducted at concentrations ranging from 0.7 to 15 ppm, 6 hours perday, 5 days per week, for three weeks. These concentrations were previously exam-ined in long term inhalation studies in rats. Instillation exposure to 400 mMformaldehyde served as a positive control. Gene expression was evaluated with awhole genome rat array. Dose-response analyses of genomic responses at severaltime-points showed dose-dependent transition for the mode of action of formalde-hyde nasal carcinogenicity, in which increasing exposure concentration activatesdifferent gene pathways. Qualitatively large scale changes in gene expression oc-curred with increasing concentration from 6 to 15 ppm. The 15 ppm more resem-bled the instillation results. The differential activation of these pathways correlateswith different cellular responses noted by histological observations and measure-ments of cell proliferation rates.

20 USING GENOMIC DOSE-RESPONSE INFORMATIONTO INFORM A POTENTIAL MODE OF ACTION FORARSENIC.

R. Gentry1, T. McDonald1, J. Yager3 and H. Clewell2. 1ENVIRON, Ruston, LA,2CIIT, Research Triangle Park, NC and 3EPRI, Palo Alto, CA.

A comprehensive literature search was conducted that identified more than 400studies containing information on gene expression changes following exposures toinorganic arsenic compounds. This information was organized by compound, ex-posure, dose/concentration, species, tissue, and cell type. Initial dose-responseanalysis focused on in vitro data obtained with primary human cells exposed to ar-senite. A concentration-related hierarchy of responses was observed, beginning withchanges gene expression associated with adaptive responses (e.g., pre-inflammatoryresponses, delay of apoptosis). Between 0.1 and 10 µM, additional gene expressionchanges related to oxidative stress, proteotoxicity, inflammation, and proliferativesignaling occur along with those related to DNA repair, cell cycle G2/M checkpointcontrol, and induction of apoptosis. At higher concentrations (10 - 100µM),changes in apoptotic genes dominate. Comparisons of primary cell results withthose obtained from immortalized or tumor-derived cell lines were also evaluated todetermine the extent to which cell lines predict the responses of primary cells.While immortalized cells appear to respond similarly to primary cells, caution mustbe exercised in using gene expression data from tumor-derived cell lines, where in-activation or overexpression of key genes (e.g., p53, Bcl-2) may lead to altered ge-nomic responses. Data from acute in vivo exposures is of limited value for evaluat-ing the dose-response for gene expression, due to the transient, variable, anduncertain nature of tissue exposure in these studies. The available in vitro gene ex-pression data, together with information on the metabolism and protein binding ofarsenic compounds, supports a mode of action for inorganic arsenic carcinogenicityinvolving the superposition of highly-specific direct interactions with critical pro-teins such as those involved in DNA repair, overlaid against a background of chem-ical stress, including proteotoxicity and depletion of non-protein sulfhydryls, re-sulting in increased oxidative damage to DNA.

21 GENE EXPRESSION DOSE-RESPONSE IN THE MOUSEBLADDER FOLLOWING EXPOSURE TO ARSENATE INDRINKING WATER.

H. Clewell1, R. Thomas1, E. Kenyon2, M. Hughes2 and J. Yager3. 1CIIT Centersfor Health Research, Research Triangle Park, NC, 2EPA/NHEERL, Research TrianglePark, NC and 3EPRI, Palo Alto, CA.

The association between drinking water exposures to inorganic arsenic and life-threatening tumors in the human is strongest for bladder cancer. Moreover, a work-ing model for the pathogenesis of human bladder cancer has been developed. To in-vestigate the mode of action for inorganic arsenic carcinogenicity in the bladder, astudy was conducted to explore the dose-response for genomic expression in blad-der epithelial cells resulting from exposure of C53Bl/6J mice to arsenate in drinkingwater for up to 12 weeks at concentrations ranging from 0.05 to 50 mg As/L.Bladder and urine concentrations of all arsenic species, including arsenite and themethylated metabolites, were also determined in order to support the design of invitro analyses on primary bladder epithelial cells at concentrations of arsenite equiv-alent to those achieved in the in vivo study. Animals exposed at 0.05 mg/L arsenateshowed no significant differences from controls in urinary arsenic species, tissueconcentrations, or bladder gene expression. However, tissue concentrations weresignificantly increased at the higher exposures, with MMA as the dominant speciesin kidney tissue and DMA in bladder and urine. In addition, a dose-dependent up-

SOT_2007.final 2/28/07 10:09 AM Page 4

SOT 2007 ANNUAL MEETING 5

regulation in a variety of genes related to DNA repair, cell cycle control, proteotox-icity, and apoptosis was observed. This presentation will discuss the implications ofthese genomic data for understanding the mode of action of arsenic bladder car-cinogenicity and will also describe how similar data obtained in primary humanbladder epithelial cells will be used to inform the quantitative dose-response assess-ment for the bladder carcinogenicity of inorganic arsenic.

22 THE FUTURE OF GENOMICS IN DOSE RESPONSEMODELING: USEFUL LESSONS FROM SYSTEMS BASEDAPPROACHES.

E. Faustman, X. Yu and B. Griffith. Environmental and Occupational HealthSciences, University of Washington, Seattle, WA.

The increasing availability of “Omic” information requires a reassessment of quan-titative risk assessment interpretative tools. Although microarray technology hasemerged as a powerful tool to explore expression levels of thousands of genes oreven complete genomes after exposure to toxicants, the functional and quantitativeinterpretation of microarray datasets still represents a time-consuming and chal-lenging task. Gene ontology (GO) and cell signaling pathway mapping have bothbeen shown to be powerful approaches to generate an overall global view of biolog-ical processes and cellular responses impacted by toxicants. However, current meth-ods do not allow for adequate comparisons across dose and time points and havelimited capabilities for translation across levels of biological complexity. Frequentlyresults are presented in extensive gene lists with minimal or limited quantitative in-formation, data that is crucial in the application of toxicogenomic data for risk as-sessment. To facilitate quantitative interpretation of dose or time dependent ge-nomic data, we propose several approaches. First, we have developed an extendedprogram (GO-Quant) to extract quantitative gene expression values and to calcu-late the average intensity or ratio for those significantly altered by functional genecategory based on MAPPFinder results. An application of this approach will begiven. Secondly, computational approaches will be discussed that allow for integra-tion of GO-Quant across cellular, organ, and organism level responses. To evaluatethese responses across biological systems necessitates the use of toxicokinetic anddynamic models that can incorporate biological information. These studies were supported by USEPA and NIEHS ES08601, ES07033, ES11387, ES 10613 and R 826882.

23 ENVIRONMENTAL RISKS FACTORS FOR AUTISM.

E. Silbergeld. Environmental Health Sciences, Johns Hopkins School of Public Health,Baltimore, MD.

Autism spectrum disorders (ASD) are a prevalent and severe health issue for chil-dren and their families, with estimated prevalence rates betwenn 1:200 and 1:600live births. It has been generally recognized that ASD involves gene:environmentinteractions, but very little information has been developed on environmental riskfactors that may contribute to these serious diseases. There are many new initiativesin research on ASD with both private and public support. The opportunity to dis-cuss current research on the ceullar and molecular biology of brain development,new information on the molecular events in ASD as well as on candidate suscepti-bility genes makes a workshop on environmental risks timely and appropriate. Thisworkshop will bring together experts on ASD, neuroscientists, and neurotoxicolo-gists to discuss the state of knowledge, promising experimentalmodels, and meth-ods for translating research into clinical practice and public health policy.

24 TRENDS IN THE INCIDENCE AND PREVALENCE OFAUTISM SPECTRUM DISORDERS: CLUES FORGENERATING HYPOTHESES ON ENVIRONMENTALRISK FACTORS.

C. J. Newschaffer. Epidemiology, Johns Hopkins University, Baltimore, MD.Sponsor: E. Silbergeld.

Data on secular trends in autism prevalence have fueled the current debate over thepotential role of environmental risk factors in autism etiology. This presentationwill summarize existing data on secular and cross-national trends in autism preva-lence. Challenges to the meaningful etiologic interpretation of these ecologic datawill be highlighted. Results from a simulation analysis will also be presented thatwill underscore the complexities of interpreting incidence, prevalence, and cumula-tive incidence trend data for a developmental condition, like autism, whose diagno-sis is behaviorally based.

25 CELL AND MOLECULAR MECHANISMS OFNEURODEVELOPMENT: CLUES FOR IDENTIFYINGENVIRONMENTAL RISK FACTORS FOR AUTISM?

P. Lein. Center for Research on Occupational and Environmental Toxicology, OregonHealth & Science University, Portland, OR.

Autism spectrum disorder (ASD) is a behaviorally defined syndrome with diverseclinical presentations. Although altered neurodevelopment is widely recognized asthe underlying neuropathological cause of ASD, the etiology and pathogenesis ofASD is poorly defined at the cellular and molecular levels. Genetic factors exert astrong influence in aggregate; however, it is clear that ASD susceptibility is deter-mined by not only complex interactions between genes but also gene-environmentinteractions. There is, therefore, significant interest in identifying and characteriz-ing epigenetic and environmental risk factors for ASD. One approach towards real-izing this goal is to identify and characterize the cell and molecular mechanismsthat control normal neurodevelopment in order to identify molecular targets whosedisruption contributes to ASD-related traits. This presentation will provide anoverview of recent advances in our understanding of the cellular and molecularmechanisms of neurodevelopment in the context of the neurobiology of autism.The major goals are to familiarize attendees with developmental mechanisms thatcontrol (1) cell migration; (2) the patterning of the balance between excitatory andinhibitory signaling pathways; and (3) synaptogenesis and synaptic plasticity. In ad-dition, recent research delineating the influence of immune mediators on neurode-velopment will be discussed in light of the emerging theme that immunologicalchallenge represents a point of convergence between genetic and environmentalcausal factors in ASD.

26 IDENTIFYING ENVIRONMENTAL CONTRIBUTORS TOAUTISM RISK; INSIGHTS FROM NEUROBIOLOGY ANDIMMUNOLOGY.

I. N. Pessah1, 2. 1VM: Molecular Biosciences, University of California, Davis, Davis,CA and 2Center for Children’s Environmental Health, Davis, CA.

Autism is among the most complex neurodevelopmental disorders with potentialsusceptibility genes (current estimates range from 3 up to 100 genes) spread acrossthe entire genome. Concordance between monozygotic twins for autism is signifi-cantly less than unity and phenotypic presentation can vary broadly between sibs.Autism therefore presents unique opportunities and challenges to study interactionsamong multiple susceptibility genes, and how epigenetic factors and exposure toenvironmental modifiers may contribute to variable expression of autism, andautism-related traits. The major goals of this presentation are to familiarize atten-dees with (1) genetic and epigenetic mechanisms implicated in altering the ratio ofexcitation/inhibition within central processing circuits of the autistic brain; includ-ing nicotinic, glutamatergic, gabaergic, and calcium signaling systems; (2) early im-munologic differences in humoral and cellular immunity identified in autistic chil-dren that may impact brain development and behavior; and (3) approaches thattoxicologist can develop in mouse models to better understanding how specific ge-netic or epigenetic defects alter sensitivity to chemical exposure(s) of current con-cern to children’s health and autism. The development of mouse models to studygene-environment interaction relevant to autism present special challenges becausethe impact of altered immunity on the development of complex behaviors, keybrain structures and neurotransmitter receptors relevant to autism must be studiedin concert. How these challenges can be met will be discussed by presenting specificapproaches used by the UC Davis Center for Children’s Environmental Health andDisease Prevention. A better understanding of the mechanisms by which environ-mental, immunologic, and molecular factors interact to influence the risk andseverity of autism is essential for developing relevant strategies for interveningand/or mitigating autism’s devastating impact.

27 TERATOGENS AND AUTISM.

P. M. Rodier. Departments of Obstetrics and Gynecology and Environmental HealthSciences, University of Rochester School of Medicine, Rochester, NY.

Five teratogens are known to increase the risk of autism: rubella, thalidomide, val-proic acid, ethanol and misoprostol. The critical period of exposure is early in thefirst trimester for each of these. The discovery of autism cases among the thalido-mide-exposed population in Sweden has led to greater understanding of autism’sorigins by pinpointing the time of neural tube closure as important in the develop-ment of the disorder. In these cases, dysfunctions of the cranial motor nerves indi-cated brain stem injury and hearing impairments occurred with ear malformations.A newly discovered genetic syndrome (BSA) includes similar anomalies of themotor innervation of the eye muscles and facial muscles, deafness, and autism. It is

SOT_2007.final 2/28/07 10:09 AM Page 5

6 SOT 2007 ANNUAL MEETING

caused by homozygous inheritance of a large mutation of the gene, HOXA1. Ananimal model based on valproic acid exposure during neural tube closure has beenstudied anatomically and behaviorally. It resembles the histology of human cases ofautism in having a reduction of the volume of the cerebellar cortex, reduced num-bers of Purkinje cells, reduced volume and neuron number in the inferior olive andreduced volume and neuron number in the deep nuclei, especially the nucleus in-terpositus. Because these injuries involve several parts of the neural circuits respon-sible for eyeblink conditioning, one might expect that behavior to be disturbed inboth human cases of autism and the animal model. The result is the opposite: eye-blink conditioning is enhanced in both. Further, the animal and human data ex-hibit identical abnormalities in the timing of the blink response. These results arenoteworthy because many neurological and psychiatric disorders have been testedfor eyeblink conditioning and autism is the only one in which enhancement hasbeen observed. Human cases exposed to valproate in utero are now being studiedfor eyeblink conditioning. We are comparing exposed children who developedautism to exposed children who did not, and comparing both to typically-develop-ing children.

28 AUTISM: WHAT DOES MERCURY HAVE TO DO WITH IT?

E. Silbergeld. Environmental Health Sciences, Johns Hopkins School of Public Health,Baltimore, MD.

Research on complex neurological disorders has utilized animal models exposed toagents that do not contribute to the burden of human disease, such as the use of 6-hydroxydopamine to emulate pathophysiological mechanisms in Parkinsonism orspecific gene knockouts in mice to study dementias. In this sense, research onmechanisms of mercury-induced developmental neurotoxicity may be valuable inunderstanding pathophysiological mechanisms involved in some aspects of autismspectrum disorders. Exposures to inorganic and organic mercury (Hg) compoundsinduces developmental neurotoxicity, involving glial activation and altered neu-ronal migration. The neuroimmunological effects of Hg are highly dependent upongenotype in rodent models. In addition to developmental neurotoxicity, early expo-sure to Hg also affects the development of the immune system, and increases thelikelihood and severity of autoimmunity consequent to infection. Because theseevents are also reported in many cases of ASD, these similarities suggest that furthermechanistic research on Hg, the role of genotype, and interactions with infectionmay be valuable in increasing understanding of early and potentially reversible orcompensible events in ASD. Research supported by Cure Autism Now Foundation.

29 THE FUTURE OF RISK ASSESSMENT INENVIRONMENTAL HEALTH – CAN CURRENTAPPROACHES FULFIL PUBLIC HEALTH NEEDS?INTRODUCTORY THOUGHTS.

A. Tritscher1 and C. Portier2. 1World Health Organization, Geneva, Switzerland and2National Institute of Environmental Health Sciences, Research Triangle Park, NC.

Risk assessment needs to use all available information to address the relevance ofenvironmental exposures to human health. In practice, current risk assessment ap-proaches tend to consider single hazards, most sensitive toxic endpoints, and oftensimple exposure scenarios. Most available data are from basic toxicological tests,which have been standardized and accepted for regulatory purposes. Also, most riskassessments are based on deterministic approaches and variability in the populationand during life-stages may not be taken sufficiently into account. Real life situa-tions are much more complex and human exposures occur to mixtures of agentsand also never in isolation from other factors such as life-style and environmentalissues. New scientific developments may provide tools to address more complexquestions in the evaluation of environmental hazards, quantification of exposure,calculation of risks and quantification of associated uncertainties. However, theyare often not sufficiently integrated into the regualtory process. Application of newscientific knowledge and tools to medicine and public health could change the na-ture of risk assessments in the future. Questions to be addressed are e.g.: can new‘omics’ technologies be used to quantify interindividual variation in susceptibility?Can these tools be used to identify and quantify consequences of exposures to mul-tiple environmental agents? Can dose-response assessment methods address expo-sures to mixtures of compounds? How can new scientific developments and newmechanistic knowledge be integrated in the regulatory risk assessment process? Thissession tries to address if currently applied risk assessment procedures, following thedefined steps of hazard identification, hazard characterization with dose-responseassessment, exposure assessment and ultimately risk characterization, are appropri-ate to address complex real life situations, and how new scientific knowledge andtools can be integrated to improve risk assessment.

30 THE FUTURE OF HAZARD IDENTIFICATION.

S. M. Roberts. Center for Environmental & Human Toxicology, University ofFlorida, Gainesville, FL.

Currently, hazard identification is based primarily on the results of toxicity studiesconducted in laboratory animals. Hazard identification that is well conducted usingthe current approach is both costly and time consuming. New tools made availableby advances in biotechnology, such as the “omics” and high-throughput screening,have the potential to provide a more comprehensive assessment of hazard, morerapidly, and perhaps at reduced cost. “Omics” tests can provide a much more com-plete picture of the biological responses to an agent than traditional studies, offer-ing alerts for possible adverse effects as well as aiding in the identification of modesof action. With high throughput screening, information on specific effects at thecellular and sub-cellular level can be obtained for tens of thousands of chemicals ina short period of time. There is general agreement regarding the value of new tech-nologies such as these as aids in hazard identification and as screening tools to helpprioritize chemicals for conventional toxicity studies. What it would take for thesetechnologies to supplant conventional in vivo studies in hazard identification ismore controversial, especially for use in risk assessments to support regulatory ac-tion. If hazard identification is to move from a foundation in observational studiesto one in predictive studies (i.e., from in vivo to in vitro and in silico), several ob-stacles must be overcome. Perhaps one of the more difficult challenges will be to re-sist the temptation to oversimplify toxic responses, trying to reduce them to one ora few key biological events. Lessons learned from decades of trying to predict car-cinogenesis, for example, indicate that the complexity of pathophysiology needs tobe acknowledged and embraced. Nevertheless, a future can be imagined in whichhazard identification is shaped primarily by predictive modelling of biological re-sponses, with in vivo studies reserved for validation and confirmation.

31 EXPOSURE ASSESSMENT.

L. W. Reiter and C. Trent. National Exposure Research Laboratory, U.S.Environmental Protection Agency, Research Triangle Park, NC.

Environmental exposures are a known etiology of preventable disease. The quanti-tative assessment of these exposures is a critical component of the risk assessmentparadigm. Yet, despite its obvious importance in the risk assessment continuum, ex-posure assessment often is the weak link, resulting in large uncertainties that canperpetuate throughout the risk assessment process. Major sources of these uncer-tainties arise from either a lack of adequate data to characterize exposure or a poorunderstanding of the relationships between environmental concentrations, expo-sure and health outcomes. This presentation will discuss how recent scientific/tech-nical advances will likely improve our ability to collect data needed to reduce theseuncertainties. Advances in sensor technology for example, will provide more sensi-tive, real time and affordable methods to measure both ambient concentrations andhuman exposures. The availability of biosensors when applied in tandem will allowus to better define and understand the relationships between external exposure andinternal dose. Novel applications of emerging electronics and information tech-nologies (e.g. GIS) provide new possibilities to accurately record human activitiesand describe their relationship to exposure. Improvements in computational meth-ods are providing new opportunities to link exposure models with pharmacokineticand dose-response models and hence our ability to conduct more integrated quan-titative risk assessments. In the specific case of air pollution, data fusion techniquesare also being developed to add air quality models to these linkages. Finally, therapid advances in genomics are providing new and better biomarkers of exposurewhich are mechanistically linked to health outcomes. Consequently, they serve toclosely link exposure assessments to health outcome including how gene environ-ment interactions affect sensitivity/susceptibility. Potential applications as well aschallenges for using these biomarkers as part of the exposure assessment process willbe discussed.

32 DOSE-RESPONSE ANALYSIS – THE QUANTITATIVECHALLENGE.

C. J. Portier. Office of Risk Assessment Research, NIEHS, Research Triangle Park, NC.

Dose-response analysis is that part of risk assessment tasked with developing aquantitative description of the relationship between a given level of exposure andthe risk of disease morbidity or mortality in human populations. The major chal-lenges in quantifying these risks are presenting an accurate picture of the variabilityof the estimated risks for known components that contribute to variation in re-sponse and providing a clear discussion of the uncertainty in the risk estimates dueto untestable assumptions that were needed for the analysis. This talk will focus onthe future of dose-response analysis in its broadest context (modelling at all levels,sensitivity to assumptions, statistical objectivity, uncertainty analysis, etc.) in light

SOT_2007.final 2/28/07 10:09 AM Page 6

SOT 2007 ANNUAL MEETING 7

of our emerging understanding of interactions betweenorganisms/organs/tissues/cells/biomolecules with exposures that alter these systems.The talk will be from the perspective of the analyst attempting to use the broadestarray of current and emergingtools and approaches to evaluate dose-response rela-tionships with a focus on the “the good, the bad and the ugly” regarding thestrengths of new tools and approaches for risk assessments, their weaknesses and thereally difficult aspects of using tools that are at the cutting edge of risk assessmentresearch. Additionally, there will be discussion of the linkage between the usage ofthese data for risk assessment and feedback into the laboratory setting to developappropriate designs to improve their utility. Finally, the discussion will switch toemerging data sources and some of the advantages and challenges seen for their usein the next 5 years. Examples of dose-response evaluations that have used emergingtools historically and currently will be included in the talk to illustrate the pointsbeing made.

33 RISK CHARACTERIZATION - THE BOTTOM LINE.

S. Olin. Research Foundation, ILSI, Washington, DC.

The true product of a risk assessment is not simply a number (e.g., a tolerable dailyintake or a cancer slope factor). Rather, it is an integrative analysis and discussion ofthe risk assessment findings, including their scope and applicability, assumptionsand uncertainties, and interpretation(s) for human health or the environment. Thisintegrative analysis and discussion has been called the risk characterization. It is thefoundation, the basis for communication between the risk assessor and those whoneed to understand and use the risk assessment findings. Because a risk characteri-zation reflects the identified hazard(s), whatever is known about the relationshipbetween dose and effect(s), the available information on exposure, and the ap-proach(es) chosen to describe and sometimes quantify the risks, it must adapt todevelopments in all of these areas as they are incorporated into the risk assessment.Some specific examples include the focus on problem formulation as a first step ina risk assessment and its impact on the risk characterization; the availability of newtools and approaches for characterizing the human relevance of modes of action oc-curring in test species; the increasing use of Margin of Exposure in risk characteri-zation; the incorporation of lifestage-related differences; and the call for improvedcharacterization of uncertainties. The complex problem of characterizing combinedrisks from multiple stressors or at multiple endpoints has received some attention,and sophisticated models for estimating cumulative risks (from exposure to multi-ple chemicals operating by the same mode of action) have been developed and ap-plied in a few cases; issues and prospects for further progress will be noted.Overlaying all of these is the challenge of finding better ways to reach consensus onthe adoption of this new science in risk assessments at national and internationallevels. Strengths and weaknesses of current approaches will be discussed, and op-portunities to improve “the bottom line” will be considered.

34 RISK ASSESSMENT AND REGULATION.

L. L. Smith. CTL, Syngenta, Cheshire, United Kingdom.

Risk assessment should be a scientific process with the outcome based on a holisticevaluation of good quality data on hazard characterisation and exposure.Regulation of risk requires an appreciation of the process of risk assessment and theincorporation of precautionary safety factors. Also, regulation must necessarily takeaccount of societal mores, often resulting from non-scientific drivers particular toindividual countries. However, it is vital for the good governance of regulation thatthe process is transparent, and the rationalé for the decision clearly stated. This isimportant so that the scientific community, NGO’s and other regulatory authori-ties can critique the outcome. Originally hazard characterisation was essentially an observational science that re-lied on an understanding of the physiology and biochemical function of individualorgans and of various cell types. Progressively, hazard characterisation has improvedwith the advent of mechanistic toxicology which increases the reliability of the ex-trapolation of animal data to humans. However, this extrapolation requires expertinterpretation and greater openness on the limitations of the utility of animal data.The availability of new technology platforms provides insights into the individualgenes and biochemical pathways that are altered as a consequence of exposure toparticular chemicals. While our ability to improve hazard characterisation contin-ues, the competence and speed of some regulatory authorities to incorporate thesenew concepts is mixed. However, perhaps the real challenge to better risk assess-ment, and hence regulation, is our ability to determine realistic exposure measure-ments. Hence, scientifically robust models of exposure need to be developed. The future of risk assessment should be driven by a more mechanistic approach tohazard characterisation combined with improved measures of exposure. Only bythe thoughtful application of these capabilities will the predictability of human

health hazards improve now. This approach, together with the exquisite balance ofjudgement, by regulators should lead to a more certain protection of humans with-out the damaging consequences of exaggerated concerns.

35 HEALTH RISKS OF OF CARBON NANOTUBES: WHATCAN WE LEARN FROM MINERAL FIBERS OR ULTRA-FINE PARTICULATES?

J. S. Tsuji1 and F. S. Mowat2. 1Exponent, Bellevue, WA and 2Exponent, MenloPark, CA.

As major building blocks in nanotechnology, carbon nanotubes (CNTs) are amongthe most studied nanomaterials, although understanding of their health risk is farfrom complete. Factors affecting toxicity include chemistry, shape and structure,persistence, and surface characteristics. Exposure characterization is also criticalgiven the complexity of particle size and shape distributions. Similar issues havebeen extensively investigated for mineral fibers and ultrafine particulates (UFPs).For example, increasing toxicity with aspect ratio has been established for asbestosfibers, and biopersistence of synthetic vitreous fibers (SVFs) influences disease po-tential. Although CNTs may have similar properties, they differ in their chemistryand in some other aspects (e.g., electrical conductivity, photochemistry, tensilestrength). In addition, differences among types of CNTs are apparent. Single-walled CNTs (SWCNTs) agglomerate and have different morphology from SVFs,as they are typically very flexible. By contrast, multi-walled CNTs (MWCNTs) donot agglomerate, are more rigid, and will therefore react differently than SWCNTs.Besides producing lung lesions, CNTs may potentially cause cardiovascular toxicityby either translocation of CNTs from the lung or by cytokine release (as seen withUFPs). Recent identification of MWCNTs in fine particulate matter (≤PM2.5)from natural gas combustion indicates that environmental exposures to MWCNTsmay be common and could play a role in air pollutant-induced diseases, although adirect link has yet to be established. Information from other well-studied aerosolsmay help guide research and assessments of the health effects of CNTs, althoughthe complexity of these structures may also inform understanding of exposure andhealth effects of different types of aerosols in general.

36 CHALLENGES OF MONITORING EXPOSURE TOAIRBORNE CARBON NANOTUBES.

A. Maynard. Project on Emerging Nanotechnologies, Woodrow Wilson InternationalCenter for Scholars, Washington, DC. Sponsor: J. Tsuji.

Single and multi-walled carbon nanotubes form complex aerosol particles that arenot easliy characterized using conventional exposure monitoring methods. For in-stance, recent research on single walled carbon nanotubes has demonstrated thatthe chemistry and morphology of aggregates released from as-produced materialmay differ over different size ranges. At the same time, research is indicating thatphysical and chemical structure is critical to determining the behavior of these par-ticles in the lungs. The characteristics of carbon nanotube aerosols will depend on the source mate-rial—whether it is as-produced or processed in some manner—and the mecha-nisms of aerosolization. Likewise, the characteristics of carbon nanotube aggregatesused in toxicity studies will depend on how they are prepared and introduced to thetest system. Sophisticated methods are needed to characterize and monitor the ma-terial appropriately if potential impact is to be evaluated. Advanced aerosol analysistechniques such as differential mobility analysis, aerosol particle mass analysis, andhigh resolution electron microscopy can be used to provide insight into the natureof airborne carbon nanotube aggregates. Simpler approaches are needed for routine exposure monitoring. However, repre-senting such a complex material with a single index of exposure presents many chal-lenges. The diverse morphologies of carbon nanotube aggregates makes analogywith other airborne materials difficult. Nevertheless, there may be sufficient simi-larities to inform future approaches. For instance, some aggregates of carbon nan-otubes will be typified by high surface-to-area ratios, as are many aerosols of aggre-gated nanometer-diameter particles. Others may share similar morphologies tomineral fibers—both natural and synthetic—that would suggest microscopy-basedmonitoring would be applicable. But while we will be able to learn from previousexperience when monitoring exposure to carbon nanotubes, future methods mustbe responsive to biologically relevant characteristics and unique carbon nanotubeattributes.

37 BIO-PERSISTENCE OF SYNTHETIC VITREOUS FIBERSAND TUMORIGENIC POTENTIAL.

T. W. Hesterberg. International Truck and Engine Corp., Warrenville, IL.

Because inhalation of asbestos, a naturally occurring, inorganic fibrous material, isassociated with lung fibrosis and thoracic cancers, concerns have been raised aboutthe possible health effects of synthetic vitreous fibers (SVFs). SVFs include a very

SOT_2007.final 2/28/07 10:09 AM Page 7

8 SOT 2007 ANNUAL MEETING

broad variety of inorganic fibrous materials: fiber glass, mineral wool, and refrac-tory ceramic fiber. Ongoing epidemiology studies of SVF manufacturing workershave provided very little evidence of harmful effects in humans. Research using ro-dents exposed by inhalation have confirmed that SVF pulmonary effects are deter-mined by fiber dose (to the lung), dimension, and durability. Fiber dose over timeis determined by fiber deposition and biopersistence in the lung. Deposition is in-versely related to fiber diameter. Biopersistence is directly related to fiber length andinversely related to fiber dissolution and fragmentation rates. Inhaled short fibers (<5µm) are cleared from the lung relatively quickly by mobile phagocytic cells, butlong fibers (> 20 µm) persist until they dissolve or fragment. In contrast to asbestos,most of the SVFs tested in rodent inhalation studies are cleared rapidly from thelung and were innocuous. However, several biopersistent SVFs induced chronic in-flammation, lung scarring (fibrosis), and thoracic neoplasms. Thus, biopersistenceof fibers is now generally recognized as a key determinant of the toxicologic poten-tial of SVFs. In vitro dissolution of fibers in simulated extracellular fluid correlatesfairly well with fiber biopersistence in the lung and pulmonary toxicity, but severalexceptions suggest that biopersistence involves more than dissolution rate. In vitrocell culture studies indicate that the biological effects of fibers are coordinated by acomplex system of inter-cellular messengers and cytotoxic cell products; phagocyticlung cells appear to be central players. Research demonstrating the relationship be-tween biopersistence and SVF toxicity has provided a scientific basis for hazard clas-sification and regulation of SVFs. It is unclear whether the mechanisms of SVFlung clearance and toxicity apply to carbon nanotubes.

38 HOW THE PROGRESSION OF PULMONARY TOXICITYRESEARCH IS INFORMING COMPARISONS OF CNTWITH MINERAL FIBERS.

D. B. Warheit. DuPont Haskell Laboratory, Newark, DE.

Previously we reported that intratracheal instillation exposures to laser ablation-type single wall carbon nanotubes (SWCNT) in rats produced a non-dose-depend-ent series of multifocal granulomas; evidence of a foreign tissue body reaction. Inaddition, instillation of high-dose SWCNT produced some mortality, resultingfrom mechanical blockage of the upper airways by the agglomerated instillate.Other investigators have demonstrated similar granulomatous lesions in mice ex-posed to a variety of SWCNT-types. In our study in rats, we have noted that the de-velopment of these granulomatous lesions is not consistent with a normal dust-re-lated paradigm, including adverse pulmonary reactions caused by cytotoxic dustssuch as quartz particles or asbestos fibers. These lesions generally develop followinglung inflammatory responses at the anatomical sites of bronchoalveolar junctions.In contrast, we observed the presence of SWCNT-nanorope-induced multifocalgranulomas located primarily in the distal airways. As a consequence, we have ques-tioned the physiological relevance of our findings following instillation exposures.Moreover, it is conceivable that due to the electrostatic nature and tendency ofSWCNT to agglomerate into nanorope structures, instillation of SWCNT mayproduce very different pathological results when compared to exposures via the in-halation route. In contrast to SWCNT, chrysotile asbestos fibers are highly res-pirable, tend to fibrillate, and produce similar lung pathological responses follow-ing either inhalation or instillation exposures. The main criteria for thepathogenicity of chrysotile asbestos fibers is biopersistence concomitant with thelength dimension, with chrysotile fibrils greater than 10 microns producing lunginflammation, fibrosis and tumors in rats. Indeed, the pathological sequelae ofthese pulmonary lesions are very different from those reported thusfar with expo-sures to SWCNT. This issue of similarities or differences in responses to SWCNTvs. asbestos can be reconciled following the implementation of inhalation toxicitystudies in rats with aerosolized SWCNT samples.

39 COMPARISON OF PULMONARY RESPONSES TOSINGLE-WALLED VS. MULTI-WALLED CARBONNANOTUBES.

V. Castranova. PPRB, NIOSH, Morgantown, WV.

Exposure of mice to purified single-walled carbon nanotubes (

SOT 2007 ANNUAL MEETING 9

42 METHODS FOR INVESTIGATING SKIN IRRITATION.

M. Sachdeva1 and J. McDougal2. 1Pharmacy, Florida A&M, Tallahassee, FL and2Pharmacology and Toxicology, Wright State University, Dayton, OH.

Occupational skin diseases are believed to be more prevalent than reported with an-nual costs (including lost workdays and loss of productivity) reaching over $1bil-lion. Skin irritation can be defined as a non-specific inflammatory response of theskin to direct chemical damage which is characterized by erythema, mild edemaand scaling (Irritant contact dermatitis). All the chemical agents (e.g. soaps,cleansers, solvents) damage the skin by causing structural change in lipids leading toa release of inflammatory mediators predominantly from cells in epidermis.Absorption of the chemical into skin reflects the potential for irritation. Differencesin blood flow, vascular permeability, skin thickness and leukocyte infiltration mayvary depending on the irritant. Understanding the signaling mechanisms involvedin chemical induced irritation has been recognized to be a very important topic andthis is being currently studied at several different levels. Most mechanistic studieshave used cultured cells and also human stratified epidermis. Also, various publica-tions do exist from whole animal studies but there are very few studies which havea) correlated in vitro and in vivo data and b) used a model capable of investigatingthe molecular effects of chemicals in the epidermis and dermis based on measuredskin concentrations in an effort to demonstrate cause and effect. This workshopwill discuss the basic issues with skin irritation (irritant and contact dermatitis), up-date the audience about the current status of in vivo methods to assess skin irrita-tion, discuss some novel in vitro culture methods being employed in skin irritation(3-D cultures), enumerate the use of skin microdialysis in assessing skin irritationby monitoring various neuropeptides (Substance P, alpha-MSH etc.) and finallydiscuss the regulatory issues important in this process. This symposium is timelybecause of the international interest in protecting individuals from skin irritationfrom chemicals.

43 OVERVIEW OF SKIN IRRITATION: INCIDENCE,MECHANISMS AND MANAGEMENT.

E. Corsini. Department of Pharmacological Sciences, University of Milan, Milan, Italy.

A major task of the epidermis is to defend the organism against outside harm,which can be of biological, physical or chemical nature. Today’s lifestyle involvesfrequent exposure to chemicals with irritating properties, making irritant contactdermatitis a common dermatologic problem, with a prevalence of up to 10% of thepopulation. In addition, occupational contact dermatitis resulting from exposure tochemicals in the workplace, depending on the country, comprise from 20 to 70 percent of all occupational diseases. Skin irritation is a local inflammatory reaction induced by primary contact withchemicals, which is characterized by erythema, edema and scaling. Irritant contactdermatitis is a multifactorial disease, the onset of which depends on both intrinsic(i.e. age, genetic background, race, gender) and extrinsic factors (i.e chemical prop-erties of the molecule itself, concentration applied, duration of exposure, etc.).Substances that are keratin solvents, dehydrating agents, oxidizing or reducingagents, and others maybe irritants. Skin inflammatory reactions involve resident epidermal cells and fibroblasts of der-mis, as well as invading leukocytes, each interacting under the control of a networkof cytokines and lipid mediators. Generally, the inflammatory reaction is initiatedby external damage to the stratum corneum, resulting in disruption of the epider-mal barrier and release of inflammatory mediators by keratinocytes, which in turnrapidly puts the innate immune system on alert in order to prevent invasion ofpathogens through the weakened barrier. Considering the increasing number of occupational skin diseases, it is essential thatthe mechanism of protection and efficacy be assessed together with a proper thera-peutic approach. The prevention of skin diseases is accepted as an important part ofoccupational dermatology. Correctly identifying the inciting irritants permits ap-propriate personal avoidance. Barrier creams, mild topical cortisones may representvaluable treatment options, while pramoxine, phenol, camphor, and menthol maybe valuable in allaying symptoms of pruritus.

44 UNDERSTANDING THE CASCADE OF EVENTS INIRRITATION AT A MOLECULAR LEVEL - IN VIVOMETHODS.

J. N. McDougal. Pharmacology and Toxicology, Wright State University School ofMedicine, Dayton, OH.

Tests for chemical-induced skin irritation have identified a variety of chemicals thathave the potential to initiate inflammatory processes in the skin. The Draize rabbitskin test has been used for over 50 years to identify chemicals that may irritate theskin. Patch testing for chemical irritation in humans has been used as an alternative

and perhaps more reliable test. Once a chemical irritant is identified, rational ther-apeutic or prophylactic interventions require an understanding of the irritant cas-cade at the molecular level. The irritant response could be blocked or lessened withappropriate pharmacological intervention. In vivo experiments in laboratory ani-mals have been used to aid in the understanding of the molecular events that occurduring skin irritation. Many of the mediators of inflammation involved in the acuteinflammatory response and vascular permeability have been identified, but the“triggers” or initiating events in chemical-induced irritation are not completely un-derstood. These first responses in the skin would be expected to be the most likelypoints for intervention if they can be recognized. Several types of molecular studieshave the potential to help in our understanding of the sequence of events in skin ir-ritation. The sequence of changes in signaling protein levels, including cytokinesand chemokines, can be used to understand the crosstalk between the epidermisand the dermis. Gene expression studies may be useful to identify activation of mo-lecular pathways involved. Immunohistochemistry studies can be used to localizethe molecular responses to specific skin layers or appendages. A systems biology ap-proach may be useful to integrate a variety of responses and provide a coherentframework for the inflammatory response. The potential for understanding chemi-cal-induced skin irritation is tremendously improving as molecular methods rapidlyadvance.

45 IN VITRO MODELS FOR ASSESSING SKIN IRRITATIONPOTENTIAL.

P. J. Hayden, J. Kubilus, S. Ayehunie, Y. Kaluzhny and M. Klausner. MatTekCorp., Ashland, MA.

In vitro models for assessing skin irritation include submerged monolayer cell cul-tures (e.g. mouse 3T3 cells or human keratinocytes), as well as differentiated 3-Dhuman epidermal (partial-thickness) or epidermal/dermal (full-thickness) skinequivalents cultured at the air-liquid interface. Submerged monolayer cultures maybe useful for irritation screening in some instances, and have been validated for ap-plications such as phototoxicity. However, submerged cultures do not adequatelymodel the physiologic and biochemical properties of differentiated epithelial tis-sues, and do not allow for topical application of water insoluble materials.Endpoints for submerged monolayer culture assays include cell viability (e.g. neu-tral red assay) or cytokine release (e.g. IL-1α. IL-8, TNFα). 3-D human skin equiv-alents cultured at the air-liquid interface offer differentiated properties includingbarrier function (i.e. stratum corneum) and Phase I/II drug metabolizing enzymeactivity, and allow for more realistic in vivo-like treatment options, including topi-cal application of creams, lotions and other water insoluble treatments. Endpointsmay be culture viability, cytokine release or histological evaluation as well as morerecent genomic and proteomic technologies. Ongoing validation programs for skinirritation with 3-D skin equivalents have shown promising correlation with in vivoresults.

46 IN SITU METHODS FOR IRRITATION- SKINMICRODIALYSIS.

M. Sachdeva and S. Fulzele. Pharmacy, Florida A&M University, Tallahassee, FL.

Microdialysis is an in vivo sampling technique for measuring endogenous and ex-ogenous solutes in the extracellular space of tissues. It is of great value in pharmaco-kinetic & pharmacodynamic investigations and in monitoring disease status andprogression. Applications of Dermal microdialysis can be classified as (a) measure-ment of inflammatory mediators in the dermis (b) studies of skin metabolism (c)determination of the absorption of drugs or other agents in the skin and (d) utiliza-tion as an alternate route of drug administration. Skin irritation is an observablephysiological reaction to a stimulus and is visualized as erythema and edema as a re-sult of local inflammatory processes. Neurogenic inflammation refers to the abilityof the neurological system to evoke local inflammatory responses through the re-lease of neuropeptides such as Substance P (SP), calcitonin gene-related peptide(CGRP) and alpha-melanocyte stimulating hormone (alpha-MSH). There is accu-mulating evidence to suggest that these mediators contribute to inflammation andskin irritation. Conventional methods to assess irritation include measurements ofskin condition [transepidermal water loss (TEWL), skin capacitance & erythema),release of inflammatory cytokines (IL-1a & TNFa) and chemokines. We hypothe-size that the relationship between skin irritation and exposure period can be quan-tified well enough by measurement of neuropeptides in the skin by dermal micro-dialysis in response to topical exposures with irritant chemicals. Jet fuel (JP-8) andits component chemicals were used as model skin irritants. In order to validate der-mal microdialysis, release of neurochemicals after topical exposure with irritantchemicals was correlated with TEWL and cytokine (IL-1a & TNFa) expression inskin & blood from experimental animals exposed under similar conditions. The re-sults from theses studies will be discussed.

SOT_2007.final 2/28/07 10:09 AM Page 9

10 SOT 2007 ANNUAL MEETING

47 REGULATORY ISSUES WITH SKIN IRRITATION: ANFDA PERSPECTIVE.

A. Jacobs. CDER, U.S. FDA, Silver Spring, MD.

Guidelines for evaluation of pharmaceuticals have been developed by theInternational Conference on Harmonization (ICH). However, these guidelines donot address skin irritation, other than under the umbrella of local effects. TheOrganization for Economic Cooperation and Development (OECD), which theUnited States is a party to, develops guidelines that describe protocols, includingthose for skin irritation and skin corrosion. Recently, laws in the EU have been en-acted that either ban the use of animals in testing for certain products or mandatethe development and validation of alternative methods to animal testing. Researchhas resulted in much activity in the development of in vitro methods intended foruse as screens, adjuncts, and replacements for current in vivo standards. In theUnited States, the Interagency Committee for the Validation of AlternativesMethods coordinates validation of alternative methods and the FDA is a very activeparticipant. Because it is desirable to use fewer animals for evaluation of toxicityand to cause less pain and suffering and because of the impetus of legislation in theEU regarding cosmetics evaluation, several in vitro skin corrosion assays have beendeveloped and validated and several in vitro ski