Employment of cytology for in vitro skin irritation test using a reconstructed human epidermis model, Keraskin™

Skin irritation tests using reconstructed human epidermis (RhE) employ viability as an endpoint, but color interference or borderline results are often problematic. We examined whether the cytology of cells from treated RhE could determine skin irritancy. Six chemicals (three irritants; DnP, 1-B, PH, three non-irritants; DP, APA, HS) were evaluated in a RhE, Keraskin™. DP, HS, and PH were clearly classified with viability, but DnP, 1-B, and APA were often falsely determined, due to borderline values falling near the cutoff, 50%. In histology, the tissues treated with DnP, 1-B, and PH showed erosion of the stratum corneum, vacuolization, and necrosis in the basal layer. DP- and HS-treated tissues showed relatively normal morphology but APA induced necrosis similar to irritants. Cytology revealed that DnP, 1-B or PH depleted cells and induced irregular and abnormal cell shapes. In contrast, relatively regular and normal shapes and clear distinction between the nucleus and cytoplasm was observed for DP, APA and HS. To further confirm it, additional 10 substances, including false positives from OECD TG 439, were tested. Overall (16 substances in total), cytology: total area predicted the skin irritancy of test chemicals with the highest accuracy (87.5%) followed by cytology: cell count (81.3%), histology (75%) and viability (68.8%), confirming the utility of cytology as an alternative endpoint in the skin irritation test using RhE.restrictio