Fourth World Congress Abstract Book

Session A3: In Vitro Reconstituted Human Tissue Models

Chairs: John Harbell (USA) and Roland Roguet (France)

A3: Application of Human Tissue Constructs to In Vitro Toxicology
John W. Harbell and Rodger D. Curren. Institute for In Vitro Sciences, Inc. jharbell@iivs.org.

The functional differential of eucaryotic cells is often dependent on the three-dimensional arrangement of the cells into an organized tissue. To properly model certain biological processes in vitro, including many manifestations of toxicity, a certain level of tissue organization and differentiation is required. Preformed tissue may be harvested and studied, ex vivo, in organ culture but this approach is limited to the species (generally non-human) and tissues that are available. Progress in human tissue engineering is making tissue constructs available for both clinical and research purposes. Over the past ten years, the use of tissue constructs has led to major advances in toxicology, particularly in the study of ocular and dermal effects. Based on these successes, specialized constructs for the study of oral mucosa, airway epithelium, vaginal epithelium, and pigmented skin models are now commercially available. However, the tissue is only part of the modeling process. Selecting the appropriate exposure kinetics and assay end points is critical and remains a constant challenge. This presentation will focus on current and prospective commercial human tissue models, their application to toxicology, and some lessons learned over the past ten years.

A3: The Use of Reconstructed Human Epidermis EPISKIN^TM in the Assessment of Efficacy and Local Tolerance of Cosmetics and Chemicals
Roland Roguet. L'Oréal Life Sciences, 90 rue du gl roguet, 92583 Clichy Cedex, France. rroguet@recherche.loreal.com.

Reconstructed human epidermis is one of the most promising tools in the in vitro evaluation of cosmetic, pharmaceutical, and chemical compounds. Its recent production as kits offers users a practical tool for the assessment of safety and efficacy of ingredients and finished products. However, the evaluation of both their reproducibility and the relevance of the data obtained with regard to cutaneous response of cosmetics or chemicals will be a crucial requirement for use.

The EPISKIN^TM model is a human-reconstructed epidermis, produced on an industrial scale in a quality-controlled environment. The industrial manufacturing process of EPISKIN^TM model and its quality control results will also be presented. The various pharmaco-toxicological applications of the model (percutaneous absorption of compounds and effects of formulations, metabolism studies (Phase I and II enzymes and the modulation of their activities), and in vitro assessment of tolerance of cosmetics or chemicals) will be presented using results obtained in the multi-center studies performed with the model. Results of new investigations using genomic technology for studying the biological effects of actives or the tolerance of products will also be presented.

Finally, an overview of the future development of reconstructed human epidermis systems in the alternative methods field will be presented.

A3: Full Thickness Epiderm^TM: A Dermal-Epidermal Skin Model to Study Epithelial-Mesenchymal Interactions
M. Klausner, J. Sheasgreen, J. Kubilus, S. Ayehunie, P. Hayden, S. Lamore, and K. Bellavance. MatTek Corporation, 200 Homer Avenue, Ashland, MA 01721, USA. jsheasgreen@mattek.com.

In contradistinction to previous dermal/epidermal models, Full Thickness EpiDerm is cultured in an easily manipulated cell culture insert, and the tissue extends from wall-to-wall. In terms of ease of use, these characteristics greatly facilitate testing of potential allergens or irritants in that direct topical application is possible. Topical exposure to the common surfactant, 1% Triton X-100, results in MTT tissue viability dose response curves that fall within the normal range of the keratinocyte-only tissue, EpiDerm. Currently, in order to produce a standardized, reproducible organotypic tissue, all lots of EpiDerm are compared to a reference database of Effective Time-50 (ET-50) values, i.e. the time of exposure, after which, viability is reduced to 50% following exposure to 100 L of Triton X-100. The database average (184 tissue lots) is 6.74 ± 0.99 hours (± 1 standard deviation); initial lots of the full thickness tissue, tested in an identical manner, average 7.14 ± 1.33 hours (n = 3). Histological cross-sections of the full thickness tissue show an epidermal layer that is very similar to EpiDerm and native epidermis atop a fibroblast-containing collagen matrix dermis-like layer. Based on these initial results, investigation of the tissue response to stimuli specifically affecting the dermis or epidermal/dermal "crosstalk" will likely prove rather informative.

A3: cdna Array-based Analysis of Gene Expression Profile in Three Different Models of Reconstructed Human Epidermis
F. Bernerd and B. Mehul. L'Oréal Advanced research, Life Sciences, 90, rue du General Roguet, Clichy, La Garenne, France. bmehul@recherche.loreal.com.

Organotypic skin models allowing epidermal differentiation can be performed on different substrates. The aim of this study was to compare the profile of gene expression, using cDNA array technology of three reconstructed epidermis obtained on different substrates, i.e. i) the de-epidermized dermis (DED) (Régnier et al., 1981), ii) a dermal equivalent containing fibroblasts embedded in a collagen gel (Lattice) (Asselineau et al., 1985), and iii) an amorphous polymerized collagen I substrate recovered by a thin layer of collagen IV (Episkin) (Tinois et al., 1991). The study also included normal human sub-confluent keratinocytes (NHK) cultured on plastic. Total RNA from epidermal keratinocytes were reversed transcribed and hybridized to membranes composed of 504 known genes related to cutaneous biology. Results calculated from the average of triplicate samples were analyzed for minimal expression level (> 1000 counts), or for differential expression ratio (> 3 fold). Correlation coefficients between the models indicate striking similarities. Only 81 genes out of the 504 were not expressed in all the three models, and among the remaining 423 genes, 368 were expressed in all the three models. The analysis of expression ratio revealed that 36 genes were over-expressed in one or two models compared to the third. Epidermis grown on DED showed a high expression of keratin 2e, loricrin, or CLSP mRNA. Epidermis obtained on lattice or DED shared highly expressed genes, such as filaggrin or caspase 14. Commonly over-expressed genes between the lattice and Episkin were identified, such as annexins, calgranulin a and b, psoriasin, or laminin 5. Moreover, gene expression profile of the three models was compared to that of NHK. Twenty-two genes were identified to be over-expressed at least in one of the three organotypic models. Six were drastically over-expressed in all the 3D models (keratin 1, loricrin, filaggrin, caspase 14, corneodesmosin, and CLSP), therefore representing a set of irrefutable markers of epidermal differentiating conditions. In contrast, 18 genes were over-expressed in KHN compared to at least one 3D model, but none of them were common for all the three models.

These results allow a better characterization of the three reconstructed epidermis models, i.e. DED, LATTICE, or EPISKIN, and illustrate their usefulness for further investigations regarding epidermal differentiation or proliferation.

A3: Advanced Skin Test 2000 (Ast-2000): Reconstructed Human Skin Designed for Dermatological and Pharmaceutical Research
J. J. Hoffmann, P. Peters, P. Frost, and H. W. Fuchs. CellSystems Biotechnologie Vertrieb GmbH, D-53562 St. Katharinen, Germany. jens.hoffmann@cellsystems.de.

The use of excised human skin has so far been considered to be one of the most suitable tools for pharmacotoxicology and dermatology research. However, the variability, and limited availability of donor material made it necessary to develop alternative test systems that allow reproducible results and the possibility to predict the effects of test substances on native human skin. In this study, we present data on "Advanced-Skin-Test-2000" (AST-2000), a reconstructed human skin, consisting of a dermal layer with fibroblasts overlayed by an epidermal layer with proliferating, differentiating, and cornified keratinocytes. Histological studies showed the expression of several markers, such as keratin 1/10, 5/14, involucrin, integrin-1, and laminin 1, correlating with the expression in native skin. Application of different test substances, e.g. SDS or NiSO₄, provoked a specific dose- and time-dependent induction of cell inflammation mediators, such as PGE2, GM-CSF, IL1, and IL 8, and a reduction of cell viability. These results strongly correlate with data obtained from the Human Patch Test and the Rabbit Skin Test. Additionally, first data showed significant up-regulation of KFG release in response to artificial lesions. In conclusion, AST-2000 proved to be a suitable model to reduce or replace animal tests.