Improved Tool for Predicting Skin Irritation on Reconstructed Human Epidermis Models Based on Electrochemical Impedance Spectroscopy

Biosensors (Basel). 2023 Jan 20;13(2):162. doi: 10.3390/bios13020162.

Abstract

The rabbit skin irritation test has been the standard for evaluating the irritation potential of chemicals; however, alternative methods that do not use animal testing are actively encouraged. Reconstructed human epidermis (RhE) models mimic the biochemical and physiological properties of the human epidermis and can be used as an alternative method. On RhE methods, the metabolic activity of RhE models is used to predict skin irritation, with a reduction in metabolic activity indicating a reduced number of viable cells and linking cell death to skin irritation processes. However, new challenges have emerged as the use of RhE models increases, including the need for non-invasive and marker-free methodologies to assess cellular states. Electrochemical impedance spectroscopy (EIS) is one such methodology that can meet these requirements. In this study, our results showed that EIS can differentiate between irritant and non-irritant chemicals, with a significant increase in the capacitance values observed in the irritant samples. A ROC curve analysis showed that the prediction method based on EIS met OECD TG 439 requirements at all time points and had 95% within-laboratory reproducibility. Comparison with the MTT viability assay showed that prediction using EIS achieved higher sensitivity, specificity, and accuracy. These results suggest that EIS could potentially replace animal testing in the evaluation of irritation potential and could be a valuable addition to in vitro testing strategies.

Keywords: alternative method; capacitance; electrochemical impedance spectroscopy; reconstructed human epidermis; safety assessment; skin irritation.

MeSH terms

  • Animal Testing Alternatives
  • Animals
  • Dielectric Spectroscopy*
  • Epidermis
  • Humans
  • Rabbits
  • Reproducibility of Results
  • Skin Irritancy Tests* / methods