Skip Navigation
Johns Hopkins Bloomberg School of Public HealthCAAT

Technical Report No. 7

Molecular and Cellular Approaches to Extrapolation for Risk Assessment

A Report of the CAAT Technical Workshop of May 5-6, 1993


The idea of conducting a workshop on the topic of molecular and cellular approaches to extrapolation for risk assessment was conceived at a meeting to review the progress of a research program grant. The overall goal of that interdisciplinary program was to fill critical gaps in our knowledge of the molecular and cellular events that occur in cells following exposure to dioxin. Much information is available about the molecular events that occur subsequent to exposure to this chemical in rodents, but as is often the case, considerably less is known about human responses. As they reviewed the program, investigators explained the molecular and cellular approaches used to link early events in cellular responses to this chemical to differential changes in gene activity, and to evaluate qualitative and quantitative differences observed in rodent and human cells. After a detailed discussion of mechanism and biological response, one board member asked a simple question: "Would such information be used by the regulatory agencies?" No one present could provide an answer.

Further consideration of this key question indicated a general lack of information about the potential uses of data generated from in vitro systems in the risk assessment process. Scientific advancements in two areas suggested that an important opportunity to clarify and redefine the role played by in vitro information in this process currently exists. First, significant improvements have occurred in methods used to culture human cells, and to maintain the differentiated functions of these cells in culture. Second, the field of molecular biology has provided toxicologists with sensitive tools to study mechanisms of chemical action. With this information in hand, and a simple paradigm that focused on species and in vitro/in vivo extrapolations, I approached Alan Goldberg, director of CAAT, with a potential workshop topic. Like the original question, Alan's answer was also quite simple, "Let's do it."

By bringing together individuals from government, industry, and academia at the workshop summarized in this technical report, we initiated a discussion of the current potential to improve our knowledge of the biological responses that occur in one very relevant animal - Homo sapiens. Since the goal of safety evaluation is to protect human health, much would be gained by an improved understanding of the risks associated with the exposure of people to drugs and chemicals. One approach to obtaining such information is through the use of human in vitro systems. While the potential offered by such an approach is clear, the realization of this potential will not be easy to achieve. As stated by Goldberg in his summary of the workshop (CAAT newsletter, Vol. 11 [No. 1], Fall 1993, Director's Diary),

"Professional judgement is needed. We can no longer focus on looking for a single answer to a simple question -- is a substance toxic or nontoxic, carcinogenic or noncarcinogenic. In a sense, we are once again acknowledging the truth of the Paracelsian dictum - it is the dose that makes the poison. The difference is that we now know that dose cannot be defined as the exposure dose but must be redefined as the dose delivered to the target site."

The Molecular and Cellular Approaches to Extrapolation workshop held in Baltimore May 5-6, 1993 addressed our current ability to ask a key question: Does the same molecular target exist in people, as in other animals, and if so, does it respond to the same dose? The abstracts contained in this technical report represent our first attempt to construct a framework for improving interspecies extrapolations through the use of in vitro systems.

Thomas R. Sutter, PhD
February, 1994


The Johns Hopkins Center for Alternatives to Animal Testing depends on financial support from public and private institutions to conduct its activities. This workshop would not have been possible without the generous sponsorship of the American Forest and Paper Association, whose support is gratefully acknowledged.

Many people contributed to the organization and administration of this workshop. I thank Alan M. Goldberg, director of CAAT, co-chairman of the workshop, and Joanne Zurlo, associate director of CAAT, co-organizer, for their time and effort. I would also like to acknowledge and thank John Frazier for his helpful discussions concerning the scientific content of this workshop, and Richard N. Hill, Science Advisor to the Assistant Administrator, U.S. Environmental Protection Agency, for his invaluable insight about the originator of the "parallelogram" approach, Frederik Sobels.

Finally, I thank Marilyn Principe, special assistant to Goldberg; Deborah Rudacille, research writer; Ann Kerr, administrative secretary; and Richelle Lewis, budget analyst, for their assistance in organizing the workshop and in the preparation of this technical report. I acknowledge and thank two students in my laboratory, Carrie Hayes and Jonathan Gastel, for their helpful discussions and literature review.



The scientific methods of toxicology are most often employed as a means of identifying potential hazards, and/or evaluating the safety of specific substances under certain experimental conditions. Through the process of risk assessment, such information is used to establish a quantitative index of risk. This index is often expressed as the probability of incremental risk as a function of exposure. The process by which data estimating potential hazards are converted into decisions of risk management is complex (see Boorman). Equally complex are the specific applications of risk assessments. These include evaluations of therapeutic drugs and commercial compounds, assessments of individual risk from occupational or accidental exposure, and assessments of population-based exposures that occur through the environment. Implicit in this process is the concept of "acceptable risk," a realization that it is virtually impossible to eliminate risk completely without adverse social or economic consequences. Although current scientific risk assessments are quantitative in nature, they are based upon assumptions inherent in methods of safety evaluation and hazard identification. To date, the predominantly accepted method is to test using animals, and then extrapolate these results to humans. Extrapolations include route of exposure, dose, and species-to-species. Due to the lack of information about relevant human responses to chemical exposures, such extrapolations lead to uncertainties; these uncertainties result in decreased confidence in risk estimates.

Workshop Premise and Goal

An important opportunity to improve the use of

in vitro

data in risk assessments currently exists. For most toxicants, we find more information gaps between the available animal and human data sets. Despite the widespread use and advancements of

in vitro

methods throughout the biomedical research community, including significant achievements in the ability to culture human cells, the current use of

in vitro

data in risk assessments is limited. The goal of this workshop was to develop a working framework to promote the use of

in vitro

data in risk assessments.

A Framework for Discussion

Traditionally, ultimate decisions concerning safety or hazard have been based on extrapolation from animals to humans. Even in tier test systems that are significantly based on alternative tests, confirmation is obtained through selective testing in animal species (see Stitzel). As such, in vitro data are often viewed as an additional level of extrapolation (Fig. 1A). In such schemes (Fig. 1A), in vitro data may be used to bolster knowledge about specific issues of extrapolation, but the data themselves represent an additional source of uncertainty. For this workshop, we proposed an alternate scheme (Fig. 1B), in vitro data are used to support investigations of mechanism of action, specifically to evaluate the assumption of conserved mechanism of action among different species.

Figure 1. The Role of In Vitro Data in Extrapolations for Risk Assessment

A) The traditional approach. In vitro data provides specific knowledge about important issues of extrapolation: route, species and dose.
  B) A four-cornered experimental approach to knowledge of mechanism. In vitro data is used to test the hypothesis that specific mechanism of action is conserved among rodent and human species. Note that the alternate hypothesis will still provide information about the action of the test compound in humans.

Workshop Scope

Topics were chosen that emphasized each corner of the alternate scheme (Fig. 1), or highlighted related issues. The workshop began with a plenary lecture on the topic of in vitro/in vivo extrapolation. Session one focused on the in vivo corners, session two on specific extrapolation issues. A discussion of structure activity relationships highlighted approaches to improve predictions of toxic potential. On day two, morning discussions focused on a case study of dioxin. Beginning with a review of the current status of the U.S. Environmental Protection Agency's reassessment of the risk of exposure to dioxin, three presentations followed, each emphasizing distinct but equally elegant molecular and cellular approaches to comparative studies of the mechanism of action of dioxin in rodent and human species. The final session included two discussions of strategies for implementation and current and future uses of in vitro data in risk assessments, by scientists from the Procter & Gamble Company and the Food and Drug Administration. The workshop concluded with an open discussion of the two day event. As expected, discussions often focused on selected toxins, organ systems, and endpoint toxicity. Given the diversity of the participants it would have been easy to expand this workshop into a major conference. However, the small size was conducive for discussion, and although not every topic was covered by specific example, the diversity of representation was sufficient to explore the issues, opportunities, and limitations of the potential use of in vitro data in risk assessments.

Issues and Opportunities

Prior to the meeting, each speaker provided a statement of his or her beliefs concerning the three most relevant issues and opportunities associated with molecular and cellular approaches to extrapolation for risk assessment. Compilation of these submissions identified issues related to four major topics:

  1. prediction;
  2. humans
  3. mechanism; and
  4. regulatory agencies and risk assessment.

Prediction: Issues

In vivo responses are often the result of complex pathological processes, that is, the long term result of multiple factor, multicellular interactions. Even when utilizing sensitive molecular and cellular approaches, can in vitro data be used to predict likely outcomes of such processes? For example, can early markers predict chronic toxicity? A related issue of prediction concerns the equivalency of sensitive biological responses. If one determines different concentration-response curves for different markers, which one(s) predicts in vivo toxicity?

These issues related to in vitro/in vivo extrapolations are significant. When viewed within the context of a complex process, for example, carcinogenesis, it is difficult to conceptualize an approach for addressing these difficulties. However, if complex biological processes are broken down into a biologically-based dose-response paradigm, then the specific in vitro/in vivo comparisons become more focused (see Frazier and Greenlee). As shown in Figure 2, complex processes can be subdivided into discrete components that provide a context for investigations of specific mechanistic steps. While each individual component is truly a connected series of mechanistic steps, the broader stroke depicted in Figure 2 emphasizes the point that for the current status of risk assessments it may be more useful to obtain increased knowledge of the entire process, albeit at a level that is less than comprehensive, than it is to have complete knowledge about one or more of the steps in the process, with little knowledge of others. In general, the overall assessment will only be as good as the least understood component in the mechanism of the endpoint of interest.

Figure 2. The Principle of Dose-Response in Risk and Safety Assessments

In the expose-dose-response paradigm, complex biological processes are divided into discrete components that provide a context for investigations of specific mechanistic steps. A) Components of in vivo toxicity. B) Components of in vitro toxicity. Note that most of the components of in vov toxicity can be studies in vitro.

Several speakers raised an interesting question concerning prediction: are the responses observed in rodents valid predictors of human toxicity? As discussed above, the alternative scheme (Fig. 1B) provides a framework in which to test the hypothesis of conserved mechanism of action among different species.

Prediction: Opportunities

Molecular and cellular approaches, coupled with comparative in vitro systems, provide a means to explore early biological response(s) to chemical or physical agents, and the role of these early effects in altered cellular structure and function. Such studies may lead to an improved understanding of mechanism of action, and of biological determinants of specificity. Also, studies of the relationship between concentration and biologically effective dose may provide insights into the shape of the dose-response curve in humans, and at lower levels of exposure. The potential for this latter opportunity (high to low dose) comes from the sensitivity of biological endpoints based on specific molecular and cellular targets.

In terms of linking exposure to dose-response relationships, several significant advancements have been made in the area of physiologically-based pharmacokinetic (PBPK) and pharmacodynamic (see Andersen and Conolly). The concept of "surrogate dose," or dose at the site of molecular action, provides a bridge between in vivo exposure and specific biological responses measured either in vivo or in vitro. As such, these modeling techniques may provide a continuum in investigations of mechanism, as experimental systems move between animals and cells in culture.

Human Cells: Issues

At present, understanding of the mechanism of action of many chemicals in humans is limited. This includes knowledge of distribution, metabolism, specific cellular targets, sensitivity of specific cell populations, and repair capacity. This general lack of information concerning toxicity in humans is further complicated by the understanding that people are exposed to multiple chemicals over a lifetime that is considerably longer than that of rodents. A second issue relates to the limited availability of human specimens.

Human Cells: Opportunities

Human specimens, including tissues, slices, organ cultures, co-cultures, or primary cells in culture, provide tremendous opportunity to investigate human biological response(s) to a variety of chemical and physical agents. When coupled with modern methods of molecular biology and biochemistry to provide human recombinant DNA probes, and expressed and purified human proteins, such studies can be used to identify primary biological endpoints relevant to human exposures (see case study on dioxin), and to determine whether the same critical cellular target (see Lehman-McKeeman) and mechanism (see Smith and Sipes) that is responsible for toxicity in animals exist in people. Corollary to this approach is the understanding that the methods developed using human in vitro systems can be easily imported as biomarkers into human epidemiology studies. Thus, human in vitro studies support both human corners of the alternate scheme (Fig. 1B), and provide an opportunity for improved understanding of human in vivo responses (see Groopman). Without such improved sensitivity of the methods of human epidemiology of the incorporation of human in vitro data into the risk characterization process, biologically based risk assessments will simply represent improved models for the interpretation of data generated by animal experimentation.

Mechanism: Issues

Mechanism versus correlation. Should all tests be relevant mechanistically? Is correlation, especially as it relates to screens, sufficient, or is it necessary to demonstrate a mechanistic link to biological response? Considerable discussion centered around the importance of knowledge of mechanism in decision-making processes. Both lectures on strategies for implementation (see Stitzel and Green), made clear the sufficiency of correlation as a criteria for the application of screens (rapid tests to determine general or specific toxicity). In these cases, knowledge of the mechanism resulting in the endpoint of toxicity is not required. It was clear that in this area of screens, current and future uses of in vitro data offer great potential to reduce, or eventually eliminate, the use of animals. The importance of these advancements should not be understated. However, it should be noted that while correlative studies may provide useful "in-house" information for decision making, they advance neither the specific understanding of the endpoint of toxicity nor the methods to detect and quantitate such toxicity. For example, the Draize test, an in vivo screen for ocular and dermal irritancy, was widely used as a correlative screen for human-use product safety assessments. Had more emphasis been placed on obtaining a mechanistic understanding of this test, its replacement by cell or organ culture methods would have been greatly facilitated. Correlative studies do not provide a foundation for scientific advancement, and as such, should be used judiciously to reduce the use of animals, while mechanistically-based screen replacements, having inherent potential for continued improvement, are developed.

Mechanism: Opportunities

Mechanism-based approaches to risk assessment tend towards identification of true risk. Risk assessments that are based on such information will be based on the best available science. In turn, this should motivate good research and promote a self-advancing field that provides an improved understanding of human risk. Computer-based chemical databases facilitate the collection, storage, and retrieval of large amounts of information. Inherent in these chemical structures are features that determine biological activity. Studies of structure-activity relationships provide the opportunity to advance from chemical specific risk assessments to chemical class-based risk assessments (see Rosenkranz). Both the concepts of structure activity and surrogate dose imply the presence of a critical cellular target. Mechanism-based approaches demand the identification of such targets, and raise the question of their conservation among species.

Regulatory Agencies and RIsk Assessment: Issues

Several issues were identified that relate to certainty/uncertainty in risk estimates. Currently, both the regulatory and legal systems attempt to classify everything unambiguously as safe or hazardous. Is it possible to move away from this towards a weight-of-evidence approach? Is the most sensitive response observed in animals necessarily the most relevant for human risk assessment? To what extent must we define a toxic mechanism in vivo as a prerequisite to gaining regulatory acceptance?

Regulatory Agencies and Risk Assessment: Opportunities

Currently, use of in vitro data in the risk assessment process is limited. Advancements in our abilities to grow and maintain human specimens, coupled with improvements in the ability to detect and quantitate specific human molecular targets, suggest that an important opportunity exists to improve the understanding of the human component of information available for risk assessment. To understand how to use this in vitro data and to incorporate such data into the risk assessment process will be difficult but is achievable.


For most of the topics discussed during this workshop we are currently knowledge-limited, that is, we lack sufficient information to move to a totally in vitro based approach. This information gap encompasses both understanding of mechanism of action and the availability of reliable data sets of sufficient size to facilitate the recognition of underlying general principles. Because of these limitations, there continues to be a prerequisite for a preliminary understanding of mechanism that, in general, is currently obtainable only through in vivo studies. Such information includes knowledge of: i) distribution, including route of administration, dose, and duration; ii) metabolism and identification of the proximate toxicant; iii) target tissue/target cell, including critical cellular concentration and relative cell and tissue sensitivities; iv) injury progression, cell:cell interactions, including measurements of both binary and continuous endpoints; v) capacity for repair, compensatory responses, adaptation; and vi) the potential for chemical interactions, including exposures to mixtures, and interactions with endogenous chemicals. In general, for the complex biological responses depicted in Figure 2, we have little understanding of the events that link altered structure/function to disease.

The availability of human specimens is limited. The quality of such samples varies, and this further complicates the issues of inter-sample and inter-individual variability. In addition, little information is available about the influence of cell culture conditions and specific medium constituents on measurements of biological responses determined in vitro.

In general, mechanism-based approaches are expensive and time consuming to develop. In addition to being technically demanding, the results tend to be chemical specific and indicative of selective toxicity, as opposed to more general or "universal" mechanisms.

Current and Future Uses of In Vitro Data

As discussed by Dr. Green, several current and future uses of in vitro data exist.

Current Uses:

  • To select the most appropriate animal model of humans;
  • To provide mechanistic information about in vivo responses;
  • To screen series of toxicants rapidly;
  • To screen for ocular, dermal, neurological, and developmental toxicity;
  • To establish potential mutagenicity and carcinogenicity;
  • To document further the hazardous nature of a carcinogen.

Future Uses:

  • Expanded use as screens;
  • To reduce or eliminate the use of animals for assessments of dermal irritation;
  • To determine specific parameters for PBPK;
  • Expanded use in investigations of mechanism of action, specifically as such information relates to risk assessment.

The "Parallelogram"

The "parallelogram" approach to extrapolation to man was proposed in the late 1970s by Dr. Frits Sobels. While this approach was originally described for its application to chemical mutagenesis, its underlying principal, "to obtain information on damage that is hard to measure directly" (see F.H. Sorbels Archives of Toxicology 46: 21-30, 1980), is relevant to most, if not all, biological endpoints of toxicity. In Figure 3, we have modified the parallelogram to emphasize two important issues that were considered in this workshop: interspecies and in vitro/in vivo extrapolation.

Figure 3. The Parallelogram Approach to Extrapolation

Modified to emphasize the issues of inter-species and in vitro/in vivo extrapolation.

While the original four-cornered scheme (Fig. 1B) provided a framework for discussion of molecular and cellular approaches to extrapolation for risk assessment, the parallelogram goes further to provide a process for systematic, comparative biology. By superimposing the parallelogram (Fig. 3) onto the components of toxicity identified in Figure 2, a rationale is established for systematic step-wise comparisons of specific mechanism of action. Through such comparisons, it should be possible to establish whether specific mechanistic steps are conserved among species. Furthermore, once conservation of mechanism is established, subsequent studies can be used to determine, and to compare between species, quantitative aspects of dose-response relationships. Thus, as previously noted by Sorbels, the parallelogram approach can be used to provide both qualitative and quantitative information that is directly relevant to estimates of human risk.


This workshop explored many aspects of the complex issues related to interspecies extrapolation. It is suggested that the parallelogram approach provides a rationale for systematic stepwise comparisons, including in vitro/in vivo comparisons of rodent and human biology that provide knowledge of both response and sensitivity to chemical action. Applications of modeling provide important methods to link in vivo exposures to other endpoints of in vivo and in vitro biological response. In reviewing the available methods and experimental systems, a major information gap was identified concerning the events that mechanistically link altered structure/function to toxicity or disease. Future studies need to focus on this important area of limited knowledge, as it appears to be rate-limiting in the overall process to determine accurate risk estimates.

Given the understanding that chemical specific risk assessments are both time consuming and expensive, considerable concern remains about the issue of selective versus universal mechanisms of toxicity. For now, no simple solution is evident. Minimally, advancements in structure-activity relationships should permit us to move from chemical specific risk assessments to those based on chemical class. Moreover, from the history of mutagenesis, it is clear that complete knowledge of specific mechanism is not required for effective determinations of risk estimates. As in the case of chemical mutagenesis, unifying concepts of general mechanisms may make it possible to develop systems to detect and quantify specific chemical activity. It remains possible that such unifying concepts are inherent in other complex biological processes such as dermal irritancy, or even cancer, and that once realized, such concepts will supersede the need for complete and specific knowledge of mechanism of action, and permit the development of effective, general screens based on common mechanism.


General Toxicology & Alternatives

  • Stone, R. (1993). Toxicology goes molecular. Science 259: 1394-1399.
  • Brudnoy, S. (1993). Pushing for a paradigm shift in cancer risk assessment. The Scientist March 8: 14-16.
  • Zbinden, G. (1992). The three eras of research in experimental toxicology Trends in Pharmacological Science 13: 221-223.
  • Zurlo, J., Rudacille, D. and A. Goldberg (1993) Animals and Alternatives in Testing - History, Science and Ethics. Mary Ann Liebert, New York.

Risk Assessment & Prediction

  • Johannsen, F.R. (1990). Risk assessment of carcinogenic and noncarcinogenic chemicals. Critical Reviews in Toxicity 20(5): 341-367.
  • Huff, J. (1993). Issues and Controversies Surrounding Qualitative Strategies for Identifying and Forecasting Cancer Causing Agents in the Human Environment. Pharmacological Toxicology 72 Suppl. 1: 12-27.
  • Wassom, J.S. (1989). Origins of genetic toxicology and the environment mutagen society. Environmental and Molecular Mutagenesis 14 Supple. 16: 1-6.
  • Ashby, J. (1991). Determination of the genotoxic status of a chemical. Mutagen Research 248: 221-231.
  • Prival, M.J. and V.L. Dellarco (1989). Evolution of social concerns and environmental policies for chemical mutagens. Environmental and Molecular Mutagenesis 14: 46-50.
  • Tennant, R.W., Margolin, B.H., Shelby, M.D., Zeiger, E., Haseman, J.K., Spalding, J., Caspary, W., Resnick, M., Stasiewics, S., Anderson, B. and R. Minor (1987). Prediction of chemical carcinogenicity in rodents from in vitro genetic toxicity assays. Science 236: 933-941.
  • Cohen, S.M., Ellwein, L.B. (1990). Cell proliferation in carcinogenesis. Science 249: 1007-1011.
  • Moolgavkar, S.H. and E.G. Luebeck (1992). Risk assessment of non-genotoxic carcinogens. Toxicology Letters 64/65: 631-636.
  • Bohrman, J.S. (1980). Identification and assessment of tumor-promoting and cocarcinogenic agents: state-of-the-art in vitro methods. CRC Critical Reviews in Toxicology 11:2: 121-167.

Experimental Models & Transgenic Animals

  • Irving, G.W. (1991). A perspective on the selection of experimental models. Neuroscience & Biobehavioral Reviews 15: 15-20.
  • Goldsworthy, T.L., Recio, L., Brown, K., Donehower, L.A., Mirsalis, J.C., Tennant, R.W. and I.F.H. Purchase (1994). Transgenic animals in toxicology. Fundamental and Applied Toxicology 11: 8-19.


  • Davidson, I.W.F., Parker, J.D. and R.P. Beliles (1986). Biological basis for extrapolation across mammalian species. Regulatory Toxicology and Pharmacology 6: 211-237.
  • Cohen, S.M. and L.B. Ellwein (1992). Risk assessment on high-dose animal exposure experiments. Chemical Research Toxicology 5: 742-748.
  • Swenberg, J.A., Richardson, F.C., Boucheron, J.A., Deal, F.H., Belinsky, S.A., Charbonneau, M. and B.G. Short (1987). High-to low-dose extrapolation: critical determinants involved in the dose response of carcinogenic substances. Environmental Health Perspectives 76: 57-63.

Human Epidemiology, Biomarkers, Human Tissue & Cells

  • Shugart, L.R., McCarthy, J.F. and R.S. Halbrook (1992) Biological markers of environmental and ecological contamination: an overview. Risk Analysis 12: 353-360.
  • Wogan, G.N. (1992). Molecular epidemiology in cancer risk assessment and prevention: recent progress and avenues for future research. Environmental Health Perspectives 98: 167-178.
  • Harris, C.C. (1987). Human tissues and cells in carcinogenesis research. Cancer Research 47: 1-10.
  • Freshney, R.I. (ed.) (1992). Culture of Epithelial Cells. New York: John Wiley & Sons.

The Parallelogram

  • Sobels, F.H. (1977). Some problems associated with the testing for environmental mutagens and a perspective for studies in "comparative mutagenesis." Mutation Research 46: 245-260.
  • Sobels, F.H. (1980). Evaluating the mutagenic potential of chemicals: the minimal battery and extrapolation problems. Archives of Toxicology 46: 21-30.
  • Sobels, F.H. (1987). Environmental mutagenesis in retrospect. Mutagen Research 181: 299-310.
  • Aaron, S., Lee, B. and J. Wassom. In Memoriam: Frederik H. Sobels (1993). Environmental and Molecular Mutagenesis 22: 125-126.


New Perspectives on In Vitro/In Vivo Extrapolation for Risk Assessment

John M. Frazier, The Johns Hopkins Center for Alternatives to Animal Testing, 615 N. Wolfe Street, Baltimore, MD 21205

If in vitro testing methods are ever to become more than just adjunct methods in the risk assessment process, then it is necessary to develop new techniques to extrapolate data generated using in vitro models to the in vivo situation. The scientific basis for this extrapolation is the focus of this presentation.

In vitro test systems (consisting of a biological component, an endpoint measurement, and a test protocol) are useful models to investigate the central events in the toxicological process, which occurs at the molecular level in the environment of the cell. The concentration-response relationship (CRR) developed in an in vitro model provides quantitative toxicological data representing the probability that a particular cellular response will result given a known concentration of the chemical in the extracellular fluid space. Usually CRRs are determined at a fixed time of exposure; however, it should be kept in mind that the concentration-time exposure profile determines the type of toxicity being evaluated. High concentrations for short exposure times are more relevant to acute toxicity considerations while low concentrations over long exposure times may be more important for the evaluation of chronic toxicity. The determination of CRRs for several endpoints over time provides the optimum toxicological information.

Figure 4: A) The Toxicological Process In Vivo and B) The Toxicological Process In Vitro

A) Expression of in vivo toxicity in any biological system is the culmination of many interacting processes which can be divided into three categories--toxicokinetics, initiation, and toxicodynamics. Toxicokinetics controls the deliverty of the active form of the toxicant (T) to the site of the interaction with the molecular target (M). The physical-chemical interaction of the active form of the toxicant and the molecular target results in the alteration of the target (M1), initiating the cascade of biological events that constitute the toxicodynamic phase. The toxicodynamic phase begins at the molecular level and propogates to high levels of biological organization ultimately producing observatble effects which are described as in vivo toxicity.
  B) In vitro cellular systems are models for the central portion of the in vivo process, including the initiation process and early molecular and cellular events. The components of the in vitro processes are similar to the in vivo process except that the level of biological organization which is observed is only at the molecular and cellular level.

At the present time, in vitro toxicity test systems can provide concentration-time-response relationships for multiple endpoint measurements as input data to the risk assessment process. Without additional experimental input, what can this information tell us? Appropriately designed test batteries can tell us much about the basic mechanisms of toxicity of the chemical. If we want to extend the usefulness of in vitro toxicity data in a meaningful way, it will be necessary to develop new quantitative techniques in two additional areas:

  1. Prediction of chemical toxicokinetics which relates chemical exposure to the concentration of the chemical and its metabolites at the cellular targets within the organism, and
  2. Prediction of toxicodynamics which describe the relationship between molecular/cellular perturbations and the ultimate expression of toxicological alterations at higher levels of biological organization.

The first problem is being addressed by in vitro models of xenobiotic metabolism and toxicokinetics. Data produced by these models can be integrated using physiologically-based toxicokinetic modeling to simulate in vivo kinetics of the chemical and its metabolites. The second area, toxicodynamics, is more difficult to address for two reasons. First, this area is poorly understood in vivo and we have little mechanistic knowledge about the sequence of events connecting cellular events to organ pathology, particularly in the case of chronic toxicity. Secondly, these events most likely involve interactions between various cell types which are difficult to reconstruct in vitro. The solution to the toxicodynamic problem is not simple; however, complex models, such as the multistage cancer models, are one approach. A second approach lies in the identification of early in vivo biomarkers of chronic pathologies which can be detected in in vitro test systems allowing for the prediction of chronic pathologies.

The key to the successful application of in vitro methods in risk assessment is an understanding of the relationship between the mechanistic basis of in vitro test system and the mechanistic basis of in vivo toxicity. Much of the knowledge required to accomplish this goal is still to be discovered. In the interim, a certain degree of empiricism will be necessary. However, the research agenda must be defined now to reach the desired objective in an efficient manner.

Portions of this abstract appeared previously in an abstract published by the Japanese Society for Alternatives to Animal Experimentation.


  • Frazier, J.M. (1992). In Vitro Toxicity Testing - Applications to Safety Evaluation. New York: Marcel Dekker, Inc.
  • Frazier, J.M. (1993). In vitro models for toxicological research and testing. Toxicological Letters 67: (In press).
  • Frazier, J.M. (1992). Scientific perspectives on the role of in vitro toxicity testing in chemical safety evaluation. In Vitro Methods in Toxicology, G. Jolles and A. Cardier (eds.) Academic Press: 521-529.


Animal Toxicity/Carcinogenicity Studies

Gary A. Boorman and Scot L. Eustis, National Institute of Environmental Health Sciences, Research Triangle Park, NC 27709

Toxicological studies in rodents have been important in identifying chemicals that are potentially hazardous to humans. Nearly all of the known human carcinogens are also carcinogenic in one or more rodent species; however, there are numerous chemicals that are carcinogenic in one or more rodent species with little or no evidence that would suggest that they might also be carcinogenic in humans. While some of the differences in the carcinogenic results can be explained by species differences or routes of exposure, some of the discrepancies between the early studies may be related to the adequacy of the studies. During the past 25 years, the quality and consistency of the rodent studies has improved considerably and much has been learned about the mechanisms whereby chemicals initiate or promote the carcinogenic process in rats and mice. While the process of identifying chemicals that cause toxicity or carcinogenicity in rodents is quite well established, the procedures for extrapolating this data for risk management decisions in the protection of human health has lagged far behind. A major impediment to progress in meaningful risk assessment has been the tendency of the U.S. regulatory and legal systems to classify each chemical as unambiguously safe or hazardous. Thus, all subsequent mechanistic in vivo and in vitro studies were usually viewed as supporting one side or the other, either the chemical was safe or not. The adversarial atmosphere was not conducive to resolving complex scientific questions and in a binary world (safe or hazardous) less emphasis was placed on defining the boundaries of certainty and uncertainty. When a chemical is found to be unequivocally hazardous for an animal model, or structure-activity relationships suggest a hazard, two major questions need to be addressed.

These questions are as follows:

  1. Does the hazard extrapolate across species where the effect was found to humans; and if yes, then
  2. What is the dose extrapolation for humans.

While newer molecular biological techniques offer exciting possibilities for better risk assessment, it is the combination of well-designed rodent studies with appropriate mechanistic studies that offers the best hope for regulatory decisions based on sound scientific principles. Acceptance of policy decisions will depend on good communication between the scientific community and the public. This will require that policy makers and toxicologists actively become involved with scientists specializing in communication.

Figure 5: The Risk Assessment Process


  • Cohen, S.M. and L.B. Ellwein (1990). Cell proliferation in carcinogenesis. Science 249: 1007-1011.
  • Woodward, K.N., McDonald, A. and S. Joshi (1991). Ranking of chemicals for carcinogenic potency - comparative study of 13 carcinogenic chemicals and an examination of some of the issues involved. Carcinogenesis 12: 1061-1066.
  • McClellan, R.O., Cuddihy, R.G., Griffith, W.C. and J.L. Mauderly (1989). Integrating diverse data sets to assess the risks of airborne pollutants. Assessment of Inhalation Hazards. U. Mohr, D.V. Bates, D.L. Dungworth, P.N. Lee, R.O. McClellan and F.J.C. Roe, (eds.) Heidelberg Springer-Verlag: 3-22.
  • Boorman, G.A., Eustis, S.L., Elwell, M.R. and R.A. Griesemer (1989). Rodent carcinogenesis studies: their value and limitations. Assessment of Inhalation Hazards. U. Mohr, D.V. Bates, D.L. Dungworth, P.N. Lee, R.O. McClellan and F.J.C. Roe. (eds.): 61-68.

Human Epidemiology/Biomarkers: Aflatoxin and Liver Cancer as a Model

John D. Groopman, Johns Hopkins University, Department of Environmental Health Sciences, Baltimore, MD 21205

Progress in basic research over the past several years has resulted in dramatic advances in the technologies used to measure the biologic events occurring in people from the time of exposure to environmental carcinogens to the time of disease outcomes. The molecular biomarkers indicative of these biologic events have generated great interest in both the research and regulatory communities because of their potential for elucidating etiologic agents of disease, thus permitting better characterization of risk from exposure. While this research is still in its early stages, these molecular biomarkers hold promise to represent a fundamental advance in public health practice. This abstract reviews some of the recent data illustrating the current status of the field with respect to aflatoxins and human liver cancer and their possible application to preventative interventions.

In a study carried out in Guangxi Autonomous Region, People's Republic of China, the diets of 42 people were monitored for one week with aflatoxin intake levels determined for each day (Groopman et al, 1992). Aflatoxin-N7-guanine and other aflatoxin metabolites were measured in urines collected in consecutive 12 hour fractions during the last three to four days of the seven day monitoring period using immunoaffinity chromatography on each sample. The levels of aflatoxin-N7-guanine adduct excreted in the urine showed a strong linear correlation with dietary aflatoxin intake, supporting the usefulness of this biomarker as an indicator of exposure (r=0.80, P<0.00001); while other metabolites, such as aflatoxin P1, gave no correspondence between intake and excretion.

Most recently, data is being analyzed from a prospective nested case control study in Shanghai in 1986 to examine relationships between markers for AF and hepatitis B virus (HBV) and the development of liver cancer (Ross et al, 1992). Over a three and a half year period, 18,244 urine samples were collected from healthy males between the ages of 45 and 64. Twenty-two of these individuals subsequently developed liver cancer and their urine samples were each age-matched with five to ten controls and analyzed for AF biomarkers and HBV surface antigen status. The data revealed a highly significant increase in the relative risk (RR = 4.9) for those liver cancer cases where AF-N7-guanine was detected. There were also elevated risks for other AF urinary markers. The relative risk for people who tested positive for the HBV surface antigen was also about five, but individuals with both urinary AFs and positive HBV surface antigen status had a relative risk for developing liver cancer of about 50. These results show for the first time that a specific biomarker for a chemical carcinogen is related with cancer risk and also for the first time that there is a multiplicative interaction between two major risk factors for liver cancer, namely HBV and AFB1.


  • Groopman, J.D., Hasler, J.A., Trudel, L.J. et al (1992). Molecular dosimetry in rat urine of aflatoxin-N7-guanine and other aflatoxin metabolites by multiple monoclonal antibody affinity chromatography and immunoaffinity/high performance liquid chromatography. Cancer Research 52: 267-274.
  • Ross, R.K., Yuan, J.M., Yu, M.C. et al (1990). Urinary aflatoxin biomarkers and risk of hepatocellular carcinomas. The Lancet 339: 943-946.
  • Groopman, J.D., Zhu, J., Chen, J.S. and G.N. Wogan (1992). Molecular dosimetry of urinary aflatoxin DNA adducts in people living in Guangxi Autonomous Region, People's Republic of China. Cancer Research 52: 45-52.
  • Ross, R., Yuan, J.M., Yu, M. et at (1992). Urinary aflatoxin biomarkers and risk of hepatocellular carcinoma. The Lancet 339: 943-946.
  • Groopman, J.D., Wild, C.P., Hasler, J. et al (1992). Molecular epidemiology of aflatoxin exposures: validation of aflatoxin-N7-guanine levels in urine as a biomarker for experimental rat models and humans. Environmental Health Perspectives 99: (in press).


How Will We Know What In Vitro and Molecular Approaches Really Tell Us About What Goes on in the Living Animal?

Melvin E. Andersen, Health Effects Research Laboratory, USEPA, Research Triangle Park, NC 27711

In vitro and molecular approaches can be seductive. Intellectually, they provide intriguing mechanistic data about the effect of chemicals on simplified biological systems. On an institutional level, they are sometimes pursued simply because they require fewer resources than do animal experiments and involve a much lesser degree of interdisciplinary cooperation. Yet, when we have these data in hand, the question persists about the significance of these in vitro observations, either for responses in the intact animal or for informed assessment of the risk posed to people from chemical exposure at environmentally relevant concentrations. In other words, what do these data from simplified biological preparations tell us about what goes on in the living animal? Can we ever really expect these techniques to have a role in quantitative health risk assessment?

Quantitative simulation models of the delivery of chemicals to target tissues in the body (i.e. pharmacokinetic models) and of the responses of tissues to the presence of toxic chemicals (i.e. dose-response models) provide an opportunity for integrating diverse in vitro and molecular results quantitatively into health risk assessments. Physiologically-based pharmacokinetic (PBPK) models described processes at the molecular, biochemical, cellular, and organ system level that determine the delivery of chemicals to target tissues and the reactions by which chemicals interact with these tissues. Biologically-based dose response (BBDR) models links target tissue dose, interactions of chemicals with tissue, and progression to overt tissue responses via biologically realistic quantitative relationships. Both types of models encode biological relationships in mathematical form, and, in principle, permit extrapolation from high doses to lower doses, from one species to another, and between dose routes. These models also provide a method to extrapolate from relevant parameters derived from in vitro tests to predict dose-response behavior in vivo. In vitro to in vivo extrapolation is accomplished by obtaining quantitative in vitro results related to tissue solubilities, rates of metabolism, rates of interactions with biological macromolecules, alterations in growth parameters, altered cell viability, toxicity to cells in culture etc. These parameters can be used in PBPK and BBDR models to predict dose-response curves in other species, including predictions of expected outcome in people. Such a model-based in vitro/in vivo extrapolation strategy is not at all unreasonable based on the biological resources now available for in vitro/molecular studies and the computational resources available for the development and implementation of these simulation models.

To date, there has been little effort to assure the consistency between actual in vivo behavior and the quantitative predictions based on incorporation of in vitro data into physiologically realistic models of dosimetry and response. With appropriate validation of this extrapolation strategy, the goal of using results from in vitro and molecular approaches for quantitative risk assessment will be much closer to becoming a reality. After some experience with these validation strategies, we also will be in a much better position to determine the relevance of these in vitro data for the living animal.


  • Clewell, H.J. and M.E. Andersen (1985). Risk assessment extrapolations and physiological modeling. Toxicology and Industrial Health. 1: 111-131.
  • Conolly, R.B. and M.E. Andersen (1991). Biologically-based pharmacodynamic models: tools for toxicological research and risk assessment. Annual Review of Pharmacological Toxicology 31: 503-523.
  • Krishnan, K. and M.E. Andersen (1991). The role of physiological modeling in reducing animal use in toxicology research. Alternative Methods in Toxicology, Vol. 8, A.M. Goldberg (ed.) New York: Mary Ann Liebert, Inc.: 113-134.
  • Krishnan, K., Gargas, M.L., and M.E. Andersen (1992). In vitro toxicology and risk assessment. Alternative Methods in Toxicology, Vol. 9. A.M. Goldberg (ed.) New York: Mary Ann Liebert, Inc.: 185-206.

Ovarian Toxicity of 4-Vinylcyclohexene and Related Compounds I

Glenn Sipes, Bill J. Smith, and Stephen B. Hooser, Center for Toxicology, University of Arizona, College of Pharmacy, Tucson, AZ 85721

Present in off gases produced during the curing of tires is 4-vinylcyclohexene (VCH), a dimer of 1,3-butadiene. Because of this potential for human exposure, VCH was tested for possible carcinogenic effects by the National Toxicology Program. These studies found that VCH caused a loss of follicles and was an ovarian carcinogen in mice (B6C3F1) but not rats (Fischer 344). Since destruction of follicles is a critical step in the induction of ovarian tumors, we investigated the ability of VCH and epoxide metabolites of VCH to destroy small follicles in the ovaries of mice and rats. Daily i.p. treatment of mice with VCH for 30 days destroys small follicles with an ED50 of 295 mg/kg/day. Similar treatment of rats with VCH failed to destroy small follicles at doses from 100 to 800 mg/kg/day. Possible cytochrome P-450 metabolites of VCH; VCH-1,2-epoxide, VCH-7,8-epoxide, and VCH diepoxide were seven, four and 18-fold more potent than VCH, respectively, at destroying small follicles of mice when given i.p. for 30 days. These epoxides also caused loss of small follicles in F-344 rats. Since VCH epoxides were more ovotoxic and VCH, evidence of their production was sought in vivo and in vitro. When mice and rats were treated with VCH (800 mg/kg/i.p.), VCH-1,2-epoxide was present in the blood of mice from 0.5 to six hours with the highest concentration at two hours (40 nmol/ml), whereas, the concentration of VCH-1,2-epoxide was <2.5 nmol/ml in the blood of rats over this time period. Studies performed in vitro suggest that this may be due to a species difference in the rate of epoxide formation in the liver. Mouse hepatic microsomes incubated with VCH (1 nM) formed VCH-1,2-epoxide at a rate 6.8-fold greater than rat hepatic microsomes. In vitro and in vivo evidence has also been obtained that rats can detoxify the epoxides of VCH more readily than mice. Therefore, hepatic metabolism of VCH may explain differences in the species variation of ovarian toxicity and carcinogenicity of VCH. Since human liver microsomes produce the ovotoxic 1,2-epoxide which is critical to the induction of ovarian tumors by VCH; the rat may be the more appropriate species for predicting the ovarian toxicity of VCH in humans. Long-term treatment of females with VCH results in increased circulating levels of FSH, suggesting disruption of the pituitary-ovarian axis. Elevation of FSH may play an important role in the ovarian cancer produced by VCH.

Supported by NIEHS N01 ES 85230 and the March of Dimes Grant #15-165.


  • Smith, B.J., Sipes, I.G., Stevens, J.C. and J.R. Halpert (1990). The biochemical basis for the species difference in hepatic microsomal 4-vinylcyclohexene epoxidation between female mice and rats. Carcinogenesis 11: 1951-1957.
  • Smith, B.J., Mattison, D.R. and I.G. Sipes (1990). The role of epoxidation in 4-vinylcyclohexene-induced ovarian toxicity. Toxicology and Applied Pharmacology 105: 373-381.
  • Smith, B.J. and I.G. Sipes (1991). Epoxidation of 4-vinylcyclohexene by human hepatic microsomes. Toxicology and Applied Pharmacology 109: 376-371.

Male Rat Specific ?2u-Globulin Nephropathy: In Vivo and In Vitro Assessment

L.D. Lehman-McKeeman, The Procter & Gamble Company, Cinncinati, OH 45239-8707

A diverse group of hydrocarbons has been shown to produce a male rat-specific nephrotoxicity manifested acutely as the accumulation of protein droplets in phagolysosomes of renal proximal tubule cells. It is now well established that ?2u-globulin, the major urinary protein excreted by adult male rats, is the only protein involved in this syndrome, referred to as hyaline droplet or ?2u-globulin nephropathy. Mechanistic studies have shown that the rate-limiting step in the development of this male rat-specific nephrotoxicity is the binding of the xenobiotic (or a metabolite) to ?2u-globulin, which subsequently renders the protein more resistant to proteolytic degradation. With chronic exposure to hyaline droplet inducing agents, the syndrome progresses to renal cell injury, compensatory renal cell proliferation, and ultimately, nephrocarcinogenicity.

Given the carcinogenic outcome associated with hyaline droplet nephropathy, it is essential to establish whether this syndrome could occur in humans. Short-term studies in other animals, including female rats, male and female mice, guinea pigs, dogs, and monkeys have shown no hyaline droplet nephropathy or any other renal toxicity; and long-term studies in female rats and male and female mice have shown no renal cancer associated with these chemicals, thereby emphasizing the male rat-specificity of this syndrome. However, there has been a family of proteins identified that share significant sequence and structural homology with ?2u-globulin and which function primarily to bind and transport hydrophobic ligands. Given these characteristics, it is possible that these proteins could interact with hyaline droplets inducing agents and cause renal toxicity in humans.

To address the significance of the ?2u-globulin protein superfamily, in vitro assays have been established to determine whether these structurally similar proteins have the ability to bind to hyaline droplet inducing agents. A systematic evaluation of these proteins, including retinol-binding protein, ?1-acid glycoprotein, B-lactoglobulin, and human protein-1 has demonstrated that ?2u-globulin is the only protein capable of binding to these male rat renal toxicants. Therefore, because of the unique ability of hyaline droplet inducing agents to bind exclusively to ?2u-globulin, this syndrome of renal protein overload, and ultimately renal cancer, is limited only to male rats and is not predictive of a carcinogenic risk to humans. Regulatory agencies (EPA) have accepted this mechanistic data for human risk assessment purposes, concluding that any chemical that produces a male rat-specific renal syndrome involving the accumulation of ?2u-globulin does not present a human health hazard.

Finally, the development of an in vitro assay to evaluate binding to ?2u-globulin or the other members of the protein superfamily, coupled with a mechanistic understanding of the significance of binding to ?2u-globulin to the development of this syndrome, has created the opportunity for the development of an in vitro assay whereby the ability of a chemical to produce this syndrome can be evaluated. This assay has proven to be very specific and highly predictive of the potential for chemically-induced hyaline droplet nephropathy.

Figure 6. Differences in Specific Binding to Rodents and Human ?2u-globulin Proteins

Equilibrium saturation binding isotherms demonstrating the binding of two known hyaline droplet inducing agents, d-limonene-1,2-epoxide (left panels) and 2,4,4-trimethyl-2-pentanol (right panels) to ?2u-globulin, and no interaction with 2 proteins in the ?2u-globulin protein superfamily, human protein-I (top panel) and a1-acid glycoprotein (lower panels). The dissociation constant for the interaction of the d-limonene epoxide and pentanol metabolites with ?2u-globulin is approximately 10-7 M.

Figure 7. Exposure-Response Model for Chloroform Carcinogenicity

Schematic representation of a quantitative, mechanism-based description of chloroform carcinogenicity. A PBPK model translates from chloroform in the exposure medium to chloroform in the target tissue. The pharmaco-dynamic description consists of a large tissue dose surrogate, amount of chloroform metabolized in the previous 30 minutes, its linkage to a short-term tissue effect, cell death, and a two-stage clonal growth cancer model. The cancer model links cell death, via regenerative cellular replication and mutation to the development of preneoplastic lesions comprised of intermediate cells, and, ultimately, to malignant tumors which arise by clonal expansion of single malignant cells.


  • Lehman-McKeeman, L.D. and D. Caudill (1992). ?2u-globulin is the only member of the lipocalin protein superfamily that binds to hyaline droplet inducing agents. Toxicology and Applied Pharmacology 116:170-176.
  • Lehman-McKeeman, L.D. (1993). Male rat-specific light hydrocarbon nephropathy in toxicology of the kidney. Toxicology of the kidney. R.S. Goldstein and J.B. Hook (eds.) New York: Raven Press: 477-494.
  • EPA Risk Assessment Forum (1991). ?2u-globulin: association with chemically-induced renal toxicity and neoplasia in the male rat. EPA/625/3-91/019F.

Pharmacodynamic Modeling: Quantitative Descriptions of the Linkage Between Tissue Dose and Toxic Response

Rory B. Conolly, Chemical Industries Institute of Toxicology, 6 Davis Drive, Research Triangle Park, NC 27709

Pharmacodynamic (PD) models describe the relationship between tissue dose and effect, be it pharmacologic or toxic. Empirical PD models, i.e. models describing the behavior of the data but not the biological mechanism giving rise to the data, have been in use for a number of years (e.g. Sheiner et at., Clinical Pharmacology and Therapeutics 25: 258-71, 1979). However, just as recent developments in physiologically-based pharmacokinetic (PBPK) modeling emphasize the mechanisms determining pharmacokinetic behavior, so there is a growing interest mechanism-based PD models. Like PBPK models, mechanism-based PD models are tools for hypothesis formulation and evaluation and have potential for use as chemical-specific risk assessment models. By "mechanism" we mean "how" an event takes place. For both pharmacokinetic and pharmacodynamic models, mechanism may be defined at the cellular or the molecular level, depending on the information available. Across the exposure-to-response continuum encompassed by PBPK/PD models, this information typically includes: (1) descriptions of factors controlling disposition of the chemical throughout the body, especially the target tissue; (2) the biochemical interaction of the toxic chemical; and (3) ensuing tissue responses, including short-term alterations such as cytolethality, and longer-term changes such as cancer.

In developing mechanism-based PD models, particular attention must be given to identification of the target tissue dose surrogate. It is important to ensure that the chemical-specific definition of the surrogate is consistent with the mechanism of action of that chemical. For example, while area-under-the-curve of the parent compound may be an appropriate dose surrogate for ethylene oxide, which is directly reactive, it would not be so for dioxin, which acts by a different mechanism. The biological effects of dioxin are mediated by its binding to high affinity receptors. The state-of-the-art in mechanism-based PD models lags behind that for PBPK models. This reflects the general lack of detailed understanding of biochemical mechanisms of tissue response. Consequently, in constructing mechanism-based PD models, it is often necessary to use empirical descriptions of some components of the model within the framework of the overall mechanism-based model. New mechanistic information can be incorporated, replacing the empirical components, as it becomes available.

These principles for development of mechanism-based PD models are illustrated by a CIIT research project whose goal is the development of a mechanism-based PBPK/PD model for chloroform. Chloroform is a renal and hepatic cytotoxicant in rodents and people and a carcinogen in mouse liver and male rat kidney. Evidence to date indicates that chloroform-induced cancer is secondary to initiation and promotion events associated with induced cytolethality and regenerative cell proliferation. We are refining an existing PBPK model for chloroform and collecting extensive dose-response data on cytolethality and induced cell replication in the target organs of mice and rats exposed to chloroform by inhalation, drinking water, and corn oil gavage. Exposure durations of up to 13 weeks are being used to determine if prior exposure to chloroform affects its pharmacokinetic behavior and the sensitivity of target tissues to the development of toxic responses. The target tissue dose of chloroform, as defined by the PBPK model, is linked to cell killing by an empirical function. The form of this function is constrained at the "dose" end by the reactive nature of the major chloroform metabolites - phosgene and HHCL - and at the "response" end by the extensive dose-response data for cytolethality and cell replication. The reactivity of the chloroform metabolites suggests a dose surrogate that is closely related to rate of chloroform metabolism rather than to total amount metabolized.

In vitro approaches, under development at CIIT, will be used to estimate model parameter values from human tissue samples. Quantitative estimates will be obtained of: (a) the metabolism of chloroform, and (b) the relationship between chloroform metabolism and cytolethality. This project will provide a PBPK/PD model for chloroform capable of predicting renal and hepatic cytolethality and regenerative cellular replication in people exposed to chloroform by inhalation or drinking water. The model will also be used to examine the correlation between these measures of the acute toxicity of chloroform and carcinogenic risks associated with chloroform exposure.


  • Sheiner, L.B., Stansky, D.R., Vozeh, S. et al (1979). Simultaneous modeling of pharmacokinetics and pharmacodynamics: Application to d-turbocurarine. Clinical Pharmacology and Therapeutics 25: 258-71.
  • Conolly, R.B. and M.E. Andersen (1991). Biologically-based pharmacodynamic models: Tools for toxicological research and risk assessment. Annual Review of Pharmacology and Toxicology 31: 503-23.
  • Conolly, R.B., Monticello, T.M., Morgan, K.T., et. al (1992). A biologically-based risk assessment strategy for inhaled formaldehyde. Comments on Toxicology 4: 269-293.
  • Reitz, R.H., Mendrala, A.L., Corley, R.A., et al (1990). Estimating the risk of liver cancer associated with human exposures to chloroform using physiologically based pharmacokinetic modeling. Toxicology and Applied Pharmacology 105: 443-459.

Applications of SAR to Extrapolation from In Vitro to In Vivo Assays

Herbert S. Rosenkranz, Center for Environmental and Occupational Health and Toxicology, Graduate School of Public Health, University of Pittsburgh, 260 Kappa Drive, Pittsburgh, PA 15238

Newer techniques in SAR allow the analysis of large congeneric databases such that mechanistic information encoded therein can be retrieved. Initially, this new SAR approach is not hypothesis-driven but derives solely from the information content of learning sets (databases). This new SAR approach yields structural information which relates to the mechanism of action of a particular biological phenomenon under study and can also be used to predict the activity of chemicals as yet not tested.

The new SAR technique allows a qualitative comparison of structural determinants present in different databases (e.g., a comparison of structural determinants of cell toxicity with those associated with, for example, the MTD in rodents or a comparison between the structural determinants of carcinogenicity with those associated with cell toxicity or with mutagenicity in salmonella). Overlaps in structural determinants are taken to indicate commonalities in mechanisms of action. These can then be used to ascertain which in vitro assays may be predictive of which in vivo effect and whether this is restricted to specific classes of chemicals.

In addition, when test results are expressed quantitatively (e.g., mg/kg/day or mutants/nanomole), the QSAR portion of the program identifies the structural mollities contributing most to the potency. A comparison between in vitro and in vivo structural determinants then allows extrapolation of the extent of activity based upon the presence of structural commonalities between the databases.

This approach has been used on, among others, the following group of assays: (1) mutagenicity in vitro vs. mutagenicity in vivo; (2) cell toxicity vs. MTD in rodents; (3) cell toxicity, MTD, mutagenicity in salmonella vs. carcinogenicity in rodents; and (4) developmental toxicity in mice, rats, and humans.


  • Klopman, G., Ptchelinstsev, D., Frierson, M. et al (1993). Multiple computer automated structure evaluation methodology as an alternative to in vivo eye irritation testing. ATLA 21: 14-27.
  • Yang, W.L., Klopman, G., and H.S. Rosenkranz (1992). Structural basis of the in vivo induction of micronuclei. Mutation Research, 171:11-124.
  • Rosenkranz, H.S. and G. Klopman (1993) Structural relationship between mutagenicity, the maximum tolerated dose, and carcinogenicity in rodents. Environmental and Molecular Mutagenesis 21: 100-115.


Case Study: Dioxin - Current and Future Uses of In Vitro Data

William H. Farland, Office of Health and Environmental Assessment, United States Environmental Protection Agency, Washington, D.C. 20460

The practice of risk assessment is evolving within the United States Environmental Protection Agency (EPA). Current approaches focus on attempts to incorporate more of the available data into weight of evidence decisions on hazard and risk. Past practice of identifying hazards through toxicologic testing is being replaced by efforts to characterize hazards by evaluating mechanistic data and developing biologically-based models. The EPA's current biologically-based efforts on dioxin and related compounds provide a good example of this evolution.

Risk assessment plays a central role in the translation of scientific information for use in regulatory decision making. As the interface between laboratory investigations and implementation of laws, it has the ability not only to use scientific data but to inform the process which drives its collection. The process of risk assessment allows the identification of critical data needed to replace the necessary scientific inference or default methodology which occurs in the absence of such data. At the same time, it feeds the regulatory decision making process as just one source of information. Scientific knowledge with its attendant uncertainties must be factored into broad based considerations of environmental and public health protection in a legal, political, and scientific context. Emerging data must always meet the test of scientific scrutiny but application to risk decision making requires broader consensus on its implications.

In recent years, a greater sense of importance has emerged for data which is useful to explain empirical observations from traditional toxicological testing. These data are often derived from cellular and molecular studies in vitro. They become critical links between whole animal studies carried out at high dose and evaluations of potential human health impacts at lower doses. Greater emphasis is now placed on availability of such data, and efforts to reach consensus on their use is underway in many instances.

The agency's reassessment of the toxicity of dioxin and related compounds provides a prime example of this trend. Recently published information from cellular and molecular studies has advanced our thinking regarding the implications of earlier observations in whole animals. These data argue strongly for the similarity of dioxin responses in humans and animals at the cellular level. Most, if not all, responses seem to be modulated by Ah receptor binding. Differential gene activity in response to dioxin-receptor interactions is being investigated as the mechanistic link with potential adverse effects in animals and humans. Additional data will be required to close the gap between these mechanistic observations and manifestation of measurable health endpoints.

Approaches developed using rodent cells or tissues are now being applied to human systems in vitro to further strengthen mechanistic inferences. The relative value of these data in the risk assessment process should not be underestimated. The implications of this approach for future dioxin research and data collection will be discussed along with the potential for the more general use of this approach in the evaluation of other potential toxicants.

Molecular Modeling of Dioxin Action

Christopher A. Bradfield, Department of Pharmacology, Northwestern University Medical School, Chicago IL 61611

Our research is directed towards developing molecular models to describe the toxic actions of compounds like 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Our experimental approach is guided by two approximations. The first is that all of the toxic effects of TCDD are mediated through stereospecific interactions with a ligand-activated transcription factor known as the Ah-receptor. The second is that the biological responses elicited by TCDD are related to alterations in gene expression. Given these precepts, we predict that the differential sensitivities to the toxic effects of TCDD that are observed both within and across species, will be related to either: 1) structural variations in the Ah-receptor protein, 2) quantitative differences in receptor expression, or 3) differential patterns of receptor-mediated gene regulation.

In our modeling studies, we have initially focused on the development of molecular tools and in vitro systems to describe receptor signaling pathways. To this end, we have recently cloned and functionally expressed both the human and murine Ah-receptor cDNAs. Analysis of these cDNAs has revealed their sequence similarly to the basic-region/helix-loop-helix (bHLH) family of transcription factors that form heterodimers prior to DNA-binding. In addition to the bHLH domain, a second unique sequence motif is found near the N-terminus and appears to define a new family of transcriptional regulators. This distinctive motif, referred to as a PAAS domain, is found in the Drosophila proteins, Sim and Per, and the recently characterized ARNT protein, a gene product required for proper Ah-receptor function.

Using coupled transcription/translation of cDNA clones, we have recently developed a system to study receptor signalling in vitro and have determined that this system accurately reflects differences in receptor biochemistry known to exist within and across species. Using this system, we have been able to reproduce three important steps using receptor signalling: 1) ligand-binding to the receptor; 2) receptor interactions with ARNT; and 3) ligand-induced DNA binding. This model system has allowed us to make a number of important observations. First, we have found that the Ah-receptor does not require the ARNT protein for ligand-binding, but does require ARNT for DNA-binding. Second, structure-function studies in this system have allowed identification of domains required for ligand-binding, DNA binding, and dimerization with ARNT. Finally, we have identified two previously undescribed domains involved in receptor activation to a DNA-binding form. The overlapped nature of many of these domains leads us to suggest a model in which ligand induced conformational changes are transduced over short distances in the receptor to induce the dissociation of an inhibitor protein, possibly Hsp90. This dissociation reveals domains required for localization to the emergence of sequence specific DNA-binding properties and the activation of a subset of genes intimately involved in the toxic pathway.


  • Burbach, K.M., Poland, A. and C.A. Bradfield (1991). Cloning of the Ah-receptor cDNA reveals a novel ligand-activated transcription factor. Proceedings of the National Academy of Sciences (USA) 89: 8185-8189.
  • Hoffman, E.C., Reyes, H., Chu, F.F. et al (1991). Cloning of a factor required for activity of the Ah (dioxin) receptor. Science 252 (5008): 954-8.

Palatal Organ Culture in the Study of Dioxin-Induced Cleft Palate

Barbara D. Abbott, Developmental Toxicology Division, Environmental Protection Agency, Research Triangle Park, NC 27711

A widespread environmental contaminant is 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). TCDD is teratogenic in C57BL/6N mice inducing cleft palate and hydronephrosis at doses which are not overtly maternally toxic. The mechanism of cleft palate induction by TCDD has been studied in our laboratory using in vivo and in vitro methods. After exposure to TCDD, palatal shelves of normal size form and come into contact, but fail to fuse. The inability to fuse correlates with altered proliferation and differentiation of the medial palatal epithelial cells which are critically involved in fusion. In the supported organ culture mode, palatal shelves are dissected on Gestation Day 12 and placed on a supported filter in contact with culture medium. This model maintains the structural and morphological organization of the tissue and allows interactions of the cell types. The medial cells develop in a manner analogous to that observed in vivo and endpoints such as peridermal cell degeneration, epithelial proliferation, and ultrastructural morphology are observed following culture. Thus, the morphological and cellular responses to TCDD were determined at various concentrations and compared to responses occurring after an in vivo dose in the mouse. A similar study was performed for cultured F344 rat palatal shelves and compared to the mouse data. Also, human embryonic palatal shelves were obtained from the Human Embryology Laboratory at the University of Washington, Seattle. The human tissues were developmentally comparable to those of the rat and mouse, regarding stage of palatal development. From each sample, one palate was cultured with control medium and the other with TCDD-containing medium. The morphology, proliferation, and peridermal degeneration were examined after exposure to TCDD at various concentrations. The relative sensitivity of the mouse, rat, and human tissues to disruption of these parameters was compared.

At the present time, studies are in progress using a palatal culture model in which mid-craniofacial tissues (including the primary palate, the maxillary arch, and the early outgrowths of palatal shelves) are submerged in medium. In this model the shelves grow, elevate, and fuse during culture. This expands the parameters which can be studied and allows scoring of the fusion of the opposing shelves. Glucocorticoids, 5-fluoracil, methanol, TCDD, and retinoic acid were studied with this model.

Palatal organ culture provides dose-response data and allows interspecies comparisons. The in vivo response of the mouse embryo must be studied in order to select meaningful endpoints to monitor in culture. Inferences about the relative sensitivity between murine and human tissue can be drawn using a parallelogram approach which relates data from mouse in vivo exposure to mouse and human responses in culture.


  • Abbott, B.D., Diliberto, J.J., and L.S. Birnbaum (1989). TCDD alters embryonic palatal medial epithelial cell differentiation in vitro. Toxicology and Applied Pharmacology 100: 119-131.
  • Abbott, B.D. and L.S. Birnbaum (1990). Rat embryonic palatal shelves respond to TCDD in organ culture. Toxicology and Applied Pharmacology. 103: 441-451.
  • Abbott, B.D. and L.S. Birnbaum (1991). TCDD exposure of human embryonic palatal shelves in organ cultures alters the differentiation of medial epithelial cells. Teratology 43: 119-132.

Human Response to Dioxin: Identification of Interspecies Determinants of Specificity

William F. Greenlee, Department of Pharmacology and Toxicology, Purdue University, West Lafayette, IN 47907

Many of the toxic responses elicited by dioxin, including chloracne in humans and the promotion of liver tumors in rats, are thought to result from the regulation of specific growth regulatory genes by a ligand-activated intracellular receptor (designated the Ah receptor). In its unoccupied form, the Ah receptor complex contains at least one ligand-binding subunit (LBS), one or more 90-kD heat shock proteins, and possibly other proteins. In the presence of dioxin, this complex dissociates and a dioxin-occupied heterodimeric nuclear transcription factor consisting of the LBS and a second protein (designated ARNT for Ah-receptor nuclear translocator) form.

With the cloning of LBS and ARNT, detailed descriptions of the interactions of dioxin with the LBS and activation to the transcriptionally active LBS/ARNT heterodimer are emerging. These early events in the receptor pathway (recognition -» transduction -» response) represent sites for multiple regulatory feedback control that may be important determinants of tissue- and species-specific responses to dioxin. For example, the expression of different allelic forms of LBS and/or ARNT could result in the formation of heterodimers that bind to unique DNA recognition motifs on multiple genes. Several lines of evidence support involvement of phosphorylation events in the formation of the activated LBS/ARNT heterodimer. Thus, regulation of cellular kinases can be an important determinant for the ligand-dependent activation of the Ah-receptor. Studies on the regulation of dioxin-responsive genes reveal multiple mechanisms that include direct transcriptional activation, suppression of gene transcription, and stabilization of specific messenger RNAs. The latter mechanisms have been demonstrated for genes involved in growth regulation. Taken together, these and other findings suggest that multiple factors act in concert with the ligand-activated Ah-receptor to produce changes in the levels of specific gene products.

Within the contact of an exposure -» dose -» response paradigm, knowledge of the mechanisms of dioxin-dependent gene regulation, particularly for gene targets likely to play a role in growth and differentiation, can serve to integrate a large body of descriptive knowledge of the toxicities produced by dioxin. The greatest impact of research on the molecular mechanisms of action is likely to be the identification of interspecies determinants of specificity. The development of molecular dosimetry models, while providing a quantitative description of the actions of dioxin on a specific target gene, are by themselves of limited value. Recognizing that the individual molecular events are integrated to produce the biological response observed in the intact animal, it is difficult to imagine any mathematical model that would incorporate successfully all of the complexities of biological processes that occur within the context of overlapping, redundant, and tightly regulated control systems.

Figure 8. Scheme for the Relationship Between Dioxin-Mediated Changes in Gene Expression and Toxic Responses


  • Sutter, T.R., Guzman, K., Dold, D.M. and W.F. Greenlee (1991). Targets for dioxin: genes for plasminogen activator inhibitor-1 and interleukin-1?. Science 254: 415-418.
  • Andersen, M.E., Mills, J.J., Gargas, M.L. et al (1993). Modeling receptor-mediated processes with dioxin: implications for pharmacokinetics and risk assessment. Risk Analysis 13: 25-36.
  • Greenlee, W.F., Sutter, T.R. and C. Marcus (1993). Molecular basis of dioxin actions on rodent and human target tissues. Progress in Clinical and Biological Research (in press).


Current and Future Uses of In Vitro Data

Katherine A. Stitzel, Miami Valley Laboratories, The Procter & Gamble Company, PO Box 398707, Cincinnati, OH 45239-8707

The body's responses to possible toxic agents involve a wide variety of changes which can be observed and measured. Methods to assess the safety of new materials rely on endpoints as varied as whole organism responses (i.e. death) to measurements of changes in specific chemical reactions occurring within specific cell populations. Our ability to use in vitro data in the assessment of the safety of new compounds depends upon our ability to model these responses in a cell culture system.

For certain endpoints such as eye and skin irritation, scientists are making progress in modeling the in vivo endpoints using in vitro systems. Both correlative data and a basic understanding of mechanisms are important in evaluating in vitro methods. Currently in vitro methods are being used as part of tier-testing approach in the assessment of the eye irritancy potential of some types of consumer products.

The tier approach includes evaluation of historical data on similar materials and bench marking against well-characterized compounds of a similar class. Using this approach, in some cases, we are able to make a safety assessment based on in vitro data alone.

Figure 9. Representative Tier-Testing Process for Eye Irritancy Currently Used by Procter & Gamble


  • Bruner, L.H. (1992). Ocular irritation. In Vitro Toxicity Testing, J.M. Frazier (ed.) New York: Marcel Dekker: 149-190.
  • Bruner, L.H. (1992). Alternatives to the use of animals in household products and cosmetic testing. JAVMA 200(5): 669-673.
  • Goldberg, A.M., Frazier, J.M., Brusnick, D. et al (1993). Framework for the validation and implementation of in vitro toxicity tests: REport of the validation and technology transfer committee of the Johns Hopkins Center for Alternatives to Animal Testing. Xenobiotica 23 (5): 563-72.

Current and Future Uses of In Vitro Data

Sidney Green, United States Food and Drug Administration, Laurel MD 20708

Data from in vitro tests are used in a variety of ways by regulatory agencies. Illustrative of some of these uses are the following: (1) use of cell cultures of various species, including humans in an attempt to determine the most appropriate animal model in which to conduct toxicological studies; (2) in vitro studies are useful in augmenting results from animal studies, particularly in identifying mechanisms of toxicity at cellular, subcellular, and molecular levels of organization; (3) cells in culture can serve to quickly screen a series of toxicants to identify the most likely candidates for further study in animals. In this context, ocular, dermal, neurobehavioral, and developmental toxicology have benefitted from this approach. Genetic toxicology has used in vitro systems for years in a variety of ways. They have been used to establish potential mutagenicity carcinogenicity, and to further document the hazardous nature of an agent shown to be carcinogenic. They are also recommended as tests for antiviral, antipsoritic or immunosuppressant drugs. Based on results from such testing, a decision may be made to limit human testing until carcinogenicity testing is completed.

Most of the above tests need not be mechanistically-based. That is, the in vitro mechanism need not be similar to the in vivo except the first, that of metabolism. It is an added advantage if they are, but it is not necessary. Genetic toxicology presents a special case, in that target as far as site of action is concerned, is the same, whether we are speaking of in vitro or in vivo assays, that is DNA. This cannot be said for the other areas of toxicology, that is the target is the same among the assays.

Future uses of in vitro data will almost certainly include expanded utility as screens in toxicology. These will range from methods which screen for systemic effects, e.g. acute toxicity, to the more specialized toxicological effects such as reproductive/developmental. It seems very likely that in vitro tests will develop to the point of almost if not totally replacing animals for dermal irritation testing. Cell cultures will also be used more in the future to provide physiological data to be used in establishing physiologically-based pharmacokinetic models. In vitro tests will contribute immensely to our understanding of mechanisms and lead to more scientifically defensible risk assessments. It does not seem probable at this time, that in vitro tests will replace animals for risk assessment purposes, whether we are speaking of no-observed-adverse-effect levels or quantitative risk for carcinogenesis. Finally, what appears to be missing and is sorely needed, are in vitro assays of human origin. Results from such assays linked to clinical or other information from humans could aid in interpreting the in vitro and certainly the in vivo rodent data.


  • Green, S. and J. Bradlaw (1992). Regulatory law and the use of in vitro methods for the assessment of various toxicities. In Vitro Toxicity Testing. J.M. Frazier (ed.) New York: Marcel Dekker, Inc.
  • Balls, N. and J.H. Fentem (1992). The use of basal cytotoxicity and target organ toxicity tests in hazard identification and risk assessment. ATLA 20 (3): 368-388.
  • U.S. Food and Drug Administration (1982). Toxicological Principles for the Safety Assessment of Direct Food Additives and Color Additives Used in Food. National Technical Information Services Pub. No. PB-83-170696. Department of Health and Human Services, Washington, D.C.


Barbara D. Abbott
Developmental Toxicology Div, USEPA
Research Triangle Park, NC 27711
Carrie Hayes
Department of Environmental Health Sciences
Johns Hopkins University Bloomberg School of Public Health
615 N. Wolfe Street
Baltimore, MD 21205
Linda Arterburn
W.R. Grace & Co.
7379 Route 32
Columbia, MD 21044
Stewart E. Holm
Georgia-Pacific Corporation
1875 I St., NW, Ste. 775
Washington, DC 20006
Melvin E. Andersen
Research Triangle Park, NC 27711
Sherif A. Kafafi
Assistant Professor of Environmental Chemistry and Biology
Department of Environmental Health Sciences
Johns Hopkins University Bloomberg School of Public Health
615 Wolfe Street
Baltimore, MD 21205
Gary A. Boorman
Chief, Pathology Branch
Environmental Toxicology Program, NIEHS
PO Box 12233
Research Triangle Park, NC 27709
Lois D. Lehman-McKeeman
Miami Valley Laboratories
The Procter & Gamble Company
PO Box 398707
Cincinnati, OH 45239
Chris Bradfield
Assistant Professor of Pharmacology
Northwester University Medical School
303 E. Chicago Avenue
Chicago, IL 60611
John Lipscomb
United States Air Force
2856 G Street
Wright Patterson AFB, OH 45422-7400
David Brusick
Hazleton Washington
9200 Leesburg Pike
Vienna, VA 22182
Nan Ni
Department of Environmental Health Sciences
Johns Hopkins University Bloomberg School of Public Health
615 N. Wolfe Street
Baltimore, MD 21205
Craig Berchtold
Department of Environmental Health Sciences
Johns Hopkins University Bloomberg School of Public Health
615 N. Wolfe Street
Baltimore, MD 21205
Rowena L. Roberts
W.R. Grace & Co.
7379 Route 32
Columbia, MD 21044-4041
Chris Cody
Department of Environmental Health Sciences
Johns Hopkins University Bloomberg School of Public Health
615 N. Wolfe Street
Baltimore, MD 21205
Herbert S. Rosenkranz
Environmental & Occupational Health
Graduate School of Publice Health
University of Pittsburgh
260 Kappa Drive
Pittsburgh, PA 15261
Rory B. Conolly, Sc.D.
Chemical Industry Institute of Toxicology, (CIIT)
6 Davis Drive
Research Triangle Park, NC 27709
Deborah Rudacille
Science Writer
The Johns Hopkins Center for Alternatives to Animal Testing
615 N. Wolfe Street
Baltimore, MD 21205
Valeria Culotta
Assistant Professor of Toxicological Sciences
Department of Environmental Health Sciences
Johns Hopkins University Bloomberg School of Public Health
615 N. Wolfe Street
Baltimore, MD 21205
William Seacat
Department of Environmental Health Sciences
Johns Hopkins University Bloomberg School of Public Health
615 N. Wolfe Street
Baltimore, MD 21205
Bill Farland
Office of Health & Env. Assessment
401 M Street, SW
Washington DC 20460
Bill Smith
Center for Toxicology and
Department of Pharmacology and Toxicology
University of Arizona College of Pharmacy
1703 E. Mabel Street
Tucson, AZ 85721
John L. Festa
American Forest and Paper Association
1250 Connecticut Ave., NW
Washington, DC 20036
Katherine A. Stitzel
Associate Director
Miami Valley Laboratories
The Procter & Gamble Company
PO Box 398707
Cincinnati, OH 45239-8707
John M. Frazier
Associate Professor of Toxicological Sciences
Department of Environmental Health Sciences
Johns Hopkins University Bloomberg School of Public Health
615 N. Wolfe Street
Baltimore, MD 21205
Peter Styzcynski
Department of Environmental Health Sciences
Johns Hopkins University Bloomberg School of Public Health
615 N. Wolfe Street
Baltimore, MD 21205
Jonathan Gastel
Department of Environmental Health Sciences
Johns Hopkins University Bloomberg School of Public Health
615 N. Wolfe Street
Baltimore, MD 21205
Thomas R. Sutter
Assistant Professor of Toxicological Sciences
Department of Environmental Health Sciences
Johns Hopkins University Bloomberg School of Public Health
615 N. Wolfe Street
Baltimore, MD 21205
Mamata Gokhale
Department of Environmental Health Sciences
Johns Hopkins University Bloomberg School of Public Health
615 N. Wolfe Street
Baltimore, MD 21205
John P. Tedeschi
Cosmair, Inc.
285 Terminal Avenue
Clark, NY 07066
Alan M. Goldberg
Director, Johns Hopkins Center for Alternatives to Animal Testing
Professor of Toxicological Sciences, Department of Environmental Sciences
Associate Dean for Research
Johns Hopkins University Bloomberg School of Public Health
615 N. Wolfe Street
Baltimore, MD 21205
Michael A. Trush
Associate Professor of Toxicological Sciences
Johns Hopkins University Bloomberg School of Public Health
615 N. Wolfe Street
Baltimore, MD 21205
Sidney Green
Director, Division of Toxicological Research
8301 Muirkirk Road
Laurel, MD 20708
Nigel Walker
Department of Environmental Health Sciences
Johns Hopkins University Bloomberg School of Public Health
615 N. Wolfe Street
Baltimore, MD 21205
William Frank Greenlee
Department of Pharmacology and Toxicology
Purdue University
1334 Robert E. Heine Bldg
West Lafayette, IN 47907-1334
James Yager
Professor and Director, Division of Toxicological Sciences
Department of Environmental Sciences
Johns Hopkins University Bloomberg School of Public Health
615 N. Wolfe Street
Baltimore, MD 21205
John D. Groopman
Professor & Chair
Department of Environmental Health Sciences
Johns Hopkins University Bloomberg School of Public Health
615 N. Wolfe Street
Baltimore, MD 21205
Hong Yin
Department of Environmental Health Sciences
Johns Hopkins University Bloomberg School of Public Health
615 N. Wolfe Street
Baltimore, MD 21205
Joanne Zurlo
Assistant Professor of Toxicological Sciences
Department of Environmental Health Sciences
Johns Hopkins University Bloomberg School of Public Health
615 N. Wolfe Street
Baltimore, MD 21205