Modifications induced by chemical skin allergens on the metabolome of reconstructed human epidermis: A pilot high‐resolution magic angle spinning nuclear magnetic resonance study

Background High‐resolution magic angle spinning (HRMAS) is a nuclear magnetic resonance (NMR) technique that enables the characterization of metabolic phenotypes/metabolite profiles of cells, tissues, and organs, under both normal and pathological conditions, without resorting to time‐consuming extraction techniques. Objectives To assess the impact of chemical skin sensitizers vs non‐sensitizers on the metabolome of three‐dimensional reconstructed human epidermis (RHE) by HRMAS NMR. Methods Based on the SENS‐IS assay, 12 skin sensitizers and five non‐sensitizing chemicals were investigated and applied on EpiSkin RHE at the published maximal non‐irritating concentrations under the conditions of the test. The metabolome of RHE samples was then analyzed by HRMAS NMR. Results A total of 32 different metabolites were identified; 20 of these were quantified for all samples. Statistical univariate analysis showed that the tissue content of most measured metabolites (with the exception of acetate and glucose) was different in the untreated, treated with non‐sensitizers, and treated with sensitizers samples. In RHE samples in contact with sensitizing chemicals, concentrations of 18 metabolites were significantly decreased. Alanine and tyrosine could not discriminate between sensitizer‐ and non‐sensitizer‐treated groups. A multivariate partial least‐squares‐discriminant analysis was performed on the two treated groups, discriminating sensitizing and non‐sensitizing chemicals with a very good R2Y value of 0.87 and a good Q2Y value of 0.70. Conclusions Data suggest that HRMAS NMR could be used to monitor the impact of chemicals, skin allergens vs non‐sensitizers, on the metabolome of three‐dimensional RHE. Highlights A total of 32 different metabolites were identified in reconstructed human epidermis (RHE) by high‐resolution magic angle spinning nuclear magnetic resonance, among which 20 could be quantified for all samples. Statistical univariate analysis showed that the measured metabolites were different in the RHE samples treated with sensitizers vs non‐sensitizers. A multivariate partial least‐squares‐discriminant analysis was performed on the two treated groups, discriminating sensitizing and non‐sensitizing chemicals.