Automation and validation of micronucleus detection in the 3D EpiDerm™ human reconstructed skin assay and correlation with 2D dose responses

Mutagenesis vol. 29 no. 3 pp. 165–175 Advance Access publication 27 March 2014 doi:10.1093/mutage/geu011 Automation and validation of micronucleus detection in the 3D EpiDerm™ human reconstructed skin assay and correlation with 2D dose responses K. E. Chapman1,†, A. D. Thomas1,†, J. W. Wills1, S. Pfuhler2, S. H. Doak1 and G. J. S. Jenkins1,* 1 DNA Damage Research Group, Institute of Life Science, College of Medicine, Swansea University, Singleton Park, Swansea, Wales, SA2 8PP, UK,2The Procter and Gamble Company, 11810 East Miami River Road, Cincinnati, OH 45252, USA † These authors contributed equally to the manuscript. *To whom correspondence should be addressed. Tel: +44 1792 602512; Fax: +44 1792 513430; Email: Received on November 12, 2013; revised on February 24, 2014; accepted on February 25, 2014 Recent restrictions on the testing of cosmetic ingredients in animals have resulted in the need to test the genotoxic potential of chemicals exclusively in vitro prior to licensing. However, as current in vitro tests produce some misleading positive results, sole reliance on such tests could prevent some chemicals with safe or beneficial exposure levels from being marketed. The 3D human reconstructed skin micronucleus (RSMN) assay is a promising new in vitro approach designed to assess genotoxicity of dermally applied compounds. The assay utilises a highly differentiated in vitro model of the human epidermis. For the first time, we have applied automated micronucleus detection to this assay using MetaSystems Metafer Slide Scanning Platform (Metafer), demonstrating concordance with manual scoring. The RSMN assay’s fixation protocol was found to be compatible with the Metafer, providing a considerably shorter alternative to the recommended Metafer protocol. Lowest observed genotoxic effect levels (LOGELs) were observed for mitomycin-C at 4.8 µg/ ml and methyl methanesulfonate (MMS) at 1750 µg/ml when applied topically to the skin surface. In-medium dosing with MMS produced a LOGEL of 20 µg/ml, which was very similar to the topical LOGEL when considering the total mass of MMS added. Comparisons between 3D medium and 2D LOGELs resulted in a 7-fold difference in total mass of MMS applied to each system, suggesting a protective function of the 3D microarchitecture. Interestingly, hydrogen peroxide (H2O2), a positive clastogen in 2D systems, tested negative in this assay. A non-genotoxic carcinogen, methyl carbamate, produced negative results, as expected. We also demonstrated expression of the DNA repair protein N-methylpurine-DNA glycosylase in EpiDerm™. Our preliminary validation here demonstrates that the RSMN assay may be a valuable followup to the current in vitro test battery, and together with its automation, could contribute to minimising unnecessary in vivo tests by reducing in vitro misleading positives. Introduction An understanding of the genotoxic effects of chemicals in humans is imperative for the prediction of genotoxic potential resulting from chemical exposure. This is particularly important for chemicals utilised in products and treatments with repeated human exposure, such as cosmetics, pharmaceuticals and food. Further, following the 7th Amendment to the Cosmetics Directive in 2009, use of laboratory animals in cosmetics testing has been banned in the European Union (1). Due to the fact that in vivo testing has previously been heavily relied upon to confirm positive results of in vitro genotoxicity assays, it is now vital that in vitro tests alone can accurately identify human carcinogens (2). However, in vitro tests often exaggerate toxic effects when compared with in vivo results, possibly due to lack of normal human tissue structure or dosing strategies being unrealistic. Further, some in vitro tests have used p53-deficient cell lines, which tend to be over-sensitive for genotoxic endpoints compared to normal human cells, contributing to misleading positives (3). Therefore, improvement of the current in vitro test battery is urgently required, to allow chemicals with safe, or even beneficial, human exposure levels to be identified and subsequently utilised in products and treatments (2). 3D EpiDerm™ reconstructed human skin models An alternative to the current genotoxicity test battery is the inclusion of in vitro methods that mimic human tissues, such as EpiDerm™ 3D reconstructed human skin models (1). Such models are predicted to better reflect the microarchitecture of human tissues and more accurately recapitulate human metabolism than cell lines. They are also predicted to exhibit near-normal DNA repair and cell cycle control (1,4). It has been demonstrated that 87% of tested xenobiotic metabolising enzymes were expressed consistently between the EpiDerm™ model and human skin, further supporting the relevance to the human condition (5). As skin is an organ frequently exposed directly to chemicals, particularly via dermal application of cosmetics and occupational exposure, the EpiDerm™ models represent one of the most common human chemical exposure routes (1). The stratum corneum of EpiDerm™ provides more relevant exposure conditions for target cells during topical dosing, avoiding the non-physiological concentrations of test chemical that are often used in in vitro tests (1,4). Therefore, an advantage of such models is that realistic concentrations of chemicals may be tested, with any associated kinetic effects that may occur as the test article diffuses from the epidermal surface to the basal layer of proliferating keratinocytes (4). Additionally, the reconstructed skin micronucleus (RSMN) assay has been shown to detect some chemicals that require metabolic activation (6). Previous studies utilising this assay have established that it is reproducible between laboratories in the USA and Europe, and also that it correctly classifies chemicals into either genotoxic or non-genotoxic carcinogens (1,2,5). Application of test chemicals to the medium instead of topically may also serve as a basic model of systemic exposure to chemicals. The hair dye ingredient, p-phenylenediamine (PPD), has previously been added to medium of EpiDerm™ models by Hu et al. (7), who noted cytotoxicity and evidence of PPD metabolites (7). An aim of this study was to expand the database for the RSMN assay in another European laboratory and to explore medium exposure © The Author 2014. Published by Oxford University Press on behalf of the UK Environmental Mutagen Society. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/ Downloaded fromby/3.0/), https://academic.oup.com/mutage/article-abstract/29/3/165/2883265 which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. by guest on 22 April 2018 165 K. E. Chapman et al. in more depth. To our knowledge, this is the first publication exploring the use of models fro (...truncated)