Skip Navigation
Johns Hopkins Bloomberg School of Public HealthCAAT

CAAT Grants Program


Research Grants 2010-2011

Summary of Research Grants

Human Culture Model as a Replacement to the Animal Assays for Assessing the Potential of Cosmetic Ingredients to Cause Non-Immunological Contact Urticaria (NICU)
Francesca Levi-Schaffer
The Hebrew University of Jerusalem

Proposal: This proposal is a joint project of The Hebrew University in Jerusalem, Israel and Lipo Chemicals Inc., USA to find an alternative method to assess cosmetic ingredients as potential irritants in humans. It is well known that a growing number of ingredients present in fragrance mix, nail polish, hair dyes and emulsions can provoke the skin contact urticaria syndrome. This reaction which is characterized by wheal and flare of healthy skin in within 30-60 min of the cosmetic application, can be either immunologic (IgE-dependent) or non immunologic (non-IgE-dependent). The last reaction can therefore appear in the skin of non atopic/ normal individuals even at the first application , and it is probably due to a direct dermal mast cell activating effect and consequent histamine release.

Our goal is to develop a totally human cell system which will allow to assess the potential of substances present in cosmetics to cause the non immunologic type of contact urticaria .This will avoid therefore using the known animal models such as mouse ear swelling assay and subjective assessment using guinea pigs.Specifically the study is designed to  evaluate the potential  of a human mast cell/fibroblast co-culture, that we have developed in the past, to be used as a simplified  in vitro model for the detection of substances inducing the contact urticaria syndrome. In this unique model, mast cells are in co-cultures with fibroblasts in a manner that resembles the in vivo condition in the dermis. Moreover, we have previously demonstrated that only when human mast cells are co-cultured with fibroblasts they are fully responsive-as in in vivo - to non IgE-dependent stimuli such as selected drugs, some neuropeptides, compound 48/80, etc. Our co-culture model will consist of human cord blood derived mast cells (CBMC) co-cultured on human dermal fibroblast monolayers. For selected experiments we will co-culture primary human skin mast cells with human dermal fibroblasts. Co-cultures will be incubated for 30 min with the cosmetic substance under investigation at different concentrations. At this time point supernatants will be evaluated for either histamine release by ELISA or tryptase release by an enzymatic-colorimetric method. Our co-culture system has several advantages: 1. It uses human mast cells that are the target of the skin applied cosmetics and not the blood-borne basophils that are often used as a not reliable counterpart of tissue mast cells. 2. It uses ßßuniquely human mast cells and fibroblasts, the last being the main cells found in the dermis in close contact with the mast cells. 3. It is a totally human system. 4. It gives quick responses-incubation is 30min long and the  histamine or tryptase assays take a few hours. 5. Unlike a few of the other suggested proposals to investigate potential skin allergic and allergic-like reactions (non-IgE-dependent), this model is focused on a cellular interactive reaction, and not on the detection of individual cytokines that may be too isolated and not representative of a cascade of events. 6. Unlike some of the animal/ human models, it is quantitative and can potentially lead to the evaluation of concentration- effect relationships. 7) Last but not least, the system is not very expensive.The proposed series of testing will include three classes of chemicals: 1. Known sensitizers (positive control), 2. Compounds that are known as non-sensitizers (negative control), 3. Unknowns (compounds that were introduced to the market in the last decade and were not extensively tested). We believe that our system can be highly predictive of the human contact urticaria syndrome even more than any in vivo animal test.

A Novel Anterior Stromal Construct for In Vitro Ocular Irritancy Testing
W. Matthew Petroll, Ph.D.
U.T. Southwestern Medical Center

Two key barriers to successful tissue engineering of an in vitro corneal equivalent have been: (1) the development of a 3-D extracellular matrix which has high mechanical stiffness, yet supports maintenance of the quiescent keratocyte phenotype, and (2) the development of a corneal epithelial cell line that differentiates and stratifies in vitro. The goal of this proposal is to evaluate a newly developed in vitro construct that overcomes these limita-tions, as an alternative to in vivo ocular toxicity testing.

The corneal stroma consists of a rigid, compacted collagen extracellular matrix (ECM). Existing techniques for producing a compacted collagen matrix rely on cellular contraction of a loose hydrated collagen matrix, cellular invasion of a synthetic matrix, or cellular synthesis of a dense ECM. Unfortunately, all of these approaches require the use of activated corneal fibroblasts. Since accurately assessing ocular toxicity requires the use of quiescent (i.e. non-contractile) corneal keratocytes, a model is needed that does not require the use of activated corneal fibroblasts for its generation. Recently, a novel approach to dermal tissue engineering has been described in which standard cell-seeded hydrated collagen matrices are compacted by external compression. This process results in the formation of a flat, cell/collagen sheet. Standard compressed matrices have a collagen concentration and mechanical stiffness similar to that of native corneal tissue. In addition, cells are not required to be contractile or activated at the start of the experiment. Importantly, our lab has demonstrated that corneal keratocytes survive the matrix compression procedure, and differentiate normally and respond to growth factors within these matrices. Furthermore, we have pilot data suggesting that corneal epithelial cells differentiate and stratify when plated on top of these matrices.

In Specific Aim 1, we will established definitively whether compressed collagen matrices support the normal differentiation and stratification of corneal epithelium under serum-free conditions in vitro. All studies will use a novel human telomerase-immortalized corneal epithelial cell line (hTCEpi), which expresses key differentiation markers under air-lifted culture conditions. hTCEpi cells will be plated on top of keratocyte-populated compressed matrices, and epithelial stratification and the expression and localization of key differentiation markers will be assessed over time.  In Specific Aim 2, the effects of slight, mild and severe ocular irritants on the epithelium and keratocytes will be assessed by performing a 3-D temporal analysis of viability, apoptosis, proliferation and cytoskeletal organization.  These results will be compared with historical data from in vivo studies.

The 3-Dimensional MAME Culture Model: An In Vitro Screen for Agents Affecting Mammary Gland Development   
Reiners, Jr., John Joseph
Wayne State University

The mammary gland is a complex organ whose development is a carefully orchestrated, chronological process. Development is very similar in humans and rodents. In the case of rodents, mammary epithelial buds form on gestation days (GD) 12-14 at the site of the nipple. By GD 16-17 bud cells begin to migrate into, and populate the stromal portion of the gland. By birth the epithelium has entered the fat pad and formed a rudimentary ductal tree, consisting of primary and secondary ducts. Exposure of pregnant rodents on GD 16-18 to a variety of chemicals (e.g., the environmental toxicant dioxin and the ubiquitous herbicide Atrazine) stunts mammary gland duct development and suppresses duct branching in female progeny. These effects are persistent and carry into puberty, resulting in the accumulation of carcinogen susceptible, undifferentiated terminal end buds. Conversely, exposure to bisphenol A (a ubiquitous chemical used in the manufacturing of polycarbonate products) on GD 15-19 programs the gland such that it exhibits enhanced ductal branching and alveolar development in female progeny when they reach puberty. The only assay available for the screening of agents capable of modulating mammary gland development entails in vivo animal treatment, followed by whole mount analyses of removed mammary tissue. The Sloane and Reiners laboratories have collaborated on the development of an in vitro 3D co-culture model that recapitulates several aspects of normal mammary gland development and architecture. In the ‘Mammary Architecture and Microenvironment Engineering’ (MAME) co-culture model, single cell suspensions of normal human mammary fibroblasts, epithelial, and myoepithelial cell lines are plated in a semi-solid Cultrex matrix. Within 9-15 days of plating the cells form primary and secondary ducts, with attached alveoli-like structures. The latter exist in clusters resembling lobular units. Both the ducts and alveoli-like structures have lumens, like their in vivo counterparts. Structures can be observed by differential interference contrast (DIC) phase microscopy, or confocal fluorescence microscopy. Culturing in the presence of low nM concentrations of dioxin suppresses duct development and branching in the MAME co-culture model, without affecting alveoli formation. We hypothesize that the MAME co-culture model maybe useful in the detection of agents that affect mammary gland duct development. The overall goal of this application is to determine if the MAME model can be used as a screen to identify chemicals whose in vivo administration during pregnancy alters fetal mammary gland development/programming. Acceptance of the MAME model as a surrogate for actual in vivo testing rests on a) how well the culture model recapitulates normal mammary gland development, and b) if known disrupters of in vivo fetal mammary duct development exhibit a similar activity in the MAME model. The specific aims of this application are to: 1) characterize the kinetics and ontogeny of duct and alveoli development in the MAME model; 2) characterize the spatial arrangements of the myoepithelial and epithelial cells in the ducts and alveolar-like structures; 3) characterize the role of fibroblasts in the ontogeny of mammary gland duct development; and 4) to survey the effects of a select group of xenobiotics for their abilities to modulate alveoli and duct formation in the MAME model. Agents to be tested include chemicals for which fetal exposure has been documented to either stunt/suppress (e.g., dioxin, perfluorooctanoic acid,
Atrazine), or enhance (e.g., bisphenol-A) mammary gland development in vivo. The first three aims assess how well the MAME model replicates in vivo mammary gland development and architecture; whereas, the fourth aim represents a first step in determining whether the MAME model can substitute for in vivo animal testing in the identification of agents capable of altering mammary gland duct development.

Biological Pathway Analysis in Human Dendritic Cells after Exposure to Sensitizing Chemicals   
Schoeters, Greet
The Flemish Institute for Technological Research

From a regulatory perspective there is an urgent need to develop non-animal models for the identification of skin allergens. Dendritic cells (DC) of the skin, also named Langerhans cells, take up the antigens and after processing, present them further to naïve T cells, thus evoking the allergic cascade. These are the earliest steps in the complex process of skin sensitization. Using transcriptomics, we tested whether CD34+ progenitor-derived DC (CD34-DC) from human cord blood, an in vitro model for skin DC, are good candidates for discrimination of chemicals with sensitizing capacity (Schoeters et al., 2007). By running further RT-PCR studies to confirm a panel of selected genes, we were able to come up with a classification model, VITOSENS®, based on combined gene expression patterns that successfully classifies (concordance of 89%, specificity of 97% and sensitivity of 82%) chemicals as sensitizers and non-sensitizers (21 chemicals tested) (Hooyberghs et al.,2008). The long-term goal of this project is to increase the comprehension of how skin sensitizer exposure impacts upon DC by elucidating how the previously identified gene markers potently discriminate between sensitizers and non-sensitizers (Schoeters et al., 2007; Hooyberghs et al.,2008). After all, besides the development of a classification model, we also have a powerful tool to learn more about potential key players in the complex skin sensitization process. Acceptance of in vitro test systems like VITOSENS® will be much more convincing if they can be linked to the essential mode of action. The latter can support further validation steps considering a prediction model and might improve risk assessment for human safety. In the first year of the project, knowledge is being gathered on the functional participation of 3 novel biomarkers, CREM, CCR2 and PTGS2, in the process of skin sensitization. Up to now, experiments show that both CCR2 and PTGS2 genes are differentially translated into proteins, and that the PTGS2 protein appears to be functionally involved in DC maturation. Experiments regarding CREM and CCR2 function are still ongoing. In the second year, we want to further elaborate the mechanism by which the VITOSENS® biomarkers potently discriminate sensitizers from non-sensitizers. Since cell signaling is at the core of most biological processes, we want to investigate whether the signaling cascades that have been reported to induce maturation and migration markers of sensitizer exposed DC also address our biomarkers and whether or not the cascades that are addressed by different sensitizers operate independent. To this end, CD34-DC will be exposed to 2 sensitizers (dinitrochlorobenzene and nickel sulfate) that have been reported in literature to operate via MAPK and NF-?? to address maturation and migration markers (Ade et al., 2007; Boisleve et al., 2005; Boisleve et al., 2004). This knowledge will be investigated in our test system by evaluating the effect of pharmacological inhibition of these signaling routes on the differential
biomarker gene expression. Also, previously obtained microarray data on exposure of CD34-DC to these sensitizers are available to the project. In a next step, it will be evaluated if both sensitizers operate via different and independent signaling routes in CD34-DC. Therefore, CD34-DC will be exposed to a mixture of both sensitizers, each added at a concentration that induces no biomarker effect when exposed to the cells separately. Evaluation of the differential expression of the biomarkers induced by the combined exposure will yield valuable information regarding effect addition in chemical mixtures. Elucidating how DC detect sensitizing exposure and transform this into a specific response of previously identified biomarkers may contribute significantly to the understanding of the mechanism of skin sensitization. The incorporation of this mechanistic understanding in an in vitro assay like VITOSENS® will enhance its acceptance and reduce the need for animals for skin sensitization testing. Furthermore, current skin sensitization testing has only dealt with single chemical substances, while in modern society we mostly deal with co-exposures. Exposing DC to a mixture of sensitizing chemicals might identify the key components of sensitization pathways and may therefore aid in more accurate risk assessment for human safety.

A Microarray Analysis Software for Comparisons and Extrapolations Relevant to Toxicology
Sutter, Thomas R.
The University of Memphis

Analysis and interpretation of structure activity and dose response relationships, and extrapolations between species and from the in vitro to in vivo situation continue to be major challenges in toxicological sciences. Establishing mechanism of action and identifying the extent of conservation of mechanism across comparisons is essential to the development of in vitro assays that are predictive of human health consequences. One promising tool for establishing such relationships is high throughput microarray analysis of mRNA expression. While proteomic and metabolomic approaches are in development, microarray analysis remains the current standard for high throughput implementation. Despite the high throughput capacity of microarray, difficulties in analysis including data
reduction and visualization (informative clustering) continue to limit the interpretation of such data. We have developed a multiple comparisons combinatorial analysis approach to the problem of structure activity relationships and have refined the statistical analysis of such data for structure activity relationships and mode of action studies. Here, we propose to develop prototype software to make these methods available and easily implemented for the toxicological comparisons described above. Such software is necessary for the important mission of CAAT in the reduction, refinement and replacement of animal use in toxicology.

Innovative microRNA-based differentiation strategy to develop metabolically stable long-term hepatocyte cultures, applicable in early pre-clinical drug research  
Vanhaecke, Tamara
Vrije Universiteit Brussel

Liver toxicity observed during (pre)clinical studies is often a major reason for stopping the development of promising drug candidates. Therefore, the pharmaceutical industry is strongly interested in establishing in vitro screening models to detect hepatotoxicity, and especially chronic liver toxicity, early in the drug development process. However, the currently available liver-based in vitro models suffer from phenotypic instability that cannot be fully overcome by classical approaches such as the introduction of cell-cell and cell-extracellular matrix contacts. Therefore, new strategies need to be explored. Our group has built up the know-how to favour the differentiated hepatocellular (in vivo-like) phenotype in vitro by interfering with epigenetic regulation of gene expression, including
DNA methylation and histone acetylation (PCT/EP2004/0012134 and PCT/EP2008/056706). Indeed, our research showed that inhibitors of these processes, like 5’-aza-2’-deoxycytidine and Trichostatin A, respectively, cause proliferative blocks, counteract spontaneous apoptotic cell death, and promote functional and morphological differentiation in primary cultures of hepatocytes. The present research project is aimed at the further extension of the already established epigenetically-based hepatocellular
differentiation-promoting strategy, i.e. by modulation of the microRNA (miRNA) expression profile in primary hepatocyte cultures. Indeed, based on the current scientific literature in the field of miRNA research, it is thought conceivable that dedifferentiation of primary hepatocytes is associated with, and probably governed by, alterations in miRNA expression patterns. Therefore, the main objective of the current research project is to develop a primary hepatocyte culture system in which functional differentiated hallmarks are kept stable for extended periods via in vitro modulation of the hepatic miRNA repertoire in favour of the in vivo-like hepatocellular phenotype. Furthermore, a combination with the epigenetic strategy already developed in the host laboratory, could increase the efficacy of this very innovative miRNA-based methodology. It is clear that this liver-based in vitro model will aid in the further implementation of the 3Rs concept, especially replacement, into research and testing. With respect to the latter, functional metabolising liver cells are not only necessary in alternatives for the pharmaceutical industry to increase preclinical safety and efficacy of new chemical entities (NCEs), but also for testing tiers for the new EU chemical legislation (REACH) as well as cosmetic testing to overcome the testing and marketing bans of March 2009 and 2013, respectively.

Use of High Throughput Methods Improves Computational Models For in vivo Low-Dose Toxicity
Zhu, Hao
The University of North Carolina at Chapel Hill

Low-dose toxicity testing is broadly utilized both in pharmaceutical industry and environmental organizations to determine the relative health hazards of drug candidates and environmental chemicals. The traditional approaches for low-dose toxicity testing in vivo are costly, time-consuming, and of a low throughput. Cell based High Throughput Screening (HTS) assays have been used recently as a possible alternative to in vivo toxicity testing. To this end, the U.S. Environmental Protection Agency (EPA) initiated a project, ToxCast, to study the relationship between in vitro screening or microarray data and in vivo low-dose toxicity for 320 compounds. However, previous studies have found little correlations between cytotoxicity data (e.g., HTS results in PubChem) and animal toxicity. Nevertheless, in our current funded grant, we have combined the biological screening data obtained from the HTS bioassays and chemical descriptors to develop statistically significant and externally predictive Quantitative Structure-Activity Relationship (QSAR) models of in vivo toxicity endpoints (including carcinogenicity and acute toxicity). We believe the same or similar approach could be used to study low-dose toxicity and develop the associated toxicity models. In this project, we will develop novel “hybrid” (i.e., using both chemical descriptors and HTS biological descriptors) QSAR models for multiple low-dose toxicity endpoints in vivo. We expect to validate and use our computational low-dose toxicity predictors to directly evaluate the toxicity potential and prioritize chemicals for future low-dose toxicity animal testing.

centrifuge tubes