INTRODUCTION: In vivo cosmetics donor testing remains the most expensive stage of the development chain to creating advanced safe cosmetics. Translation of lab-testing data to final product is not easy and can fail. Solutions to making safer cosmetics requires better testing which is why we have developed the world’s first 3D Bioprinted oily skin model which can measure the same non-invasive electrical activity as on real humans, while at the same time giving normal laboratory read-outs for cellular, matrix and oil development kinetics. Bioimpedance, which has long been used for general body composition testing can also be applied at the surface skin level, to evaluate changes in the local skin environment. Here we aimed to combine laboratory and in vivo parameters for better cosmetics testing.
METHODS: Human keratinocytes, fibroblasts and sebocytes were expanded from donations of human adult male and female skin donated following elective surgery and ethical consent. Cells at passage 2 or 3 were selected for optimal growth and mixed with a bioink (CELLINK, Sweden), into which epidermal adhesion proteins were added and cartridged into a CELLINK pneumatic 3D bioprinting system, after modelization with Sketchup and SLIC3R slicing software. Full thickness epidermal-dermal, and separate dermal models, with and without sebocytes were printed, along with an endothelial model and an additional immunized model (described at IFSCC2020, Henry Maso award 2022). Bioimpedance analysis chips were made with culture plates and supernatant holders and attached to a bioimpedance Electric Bridge Resistance Impedance Capacitance Inductance Digital LCR Meter calibrated in the 1kHz range. Bioimpedance readings were made following maturation of models and after treatments with modulators of the oil production known in sebaceous glands: 1mM Linoleic Acid, 50µM TOFA. Changes in cellular viability were made with live-dead measurements and oil production was further evaluated by the standard Oil Red O quantification, spectrophotometry and imaging over 96 hours. Both supernatants and printed models were evaluated. Simultaneously, we evaluated the bioimpedance directly on donor skin which was either untreated, or treated with a coconut oil derivative.
RESULTS: Differences in bioimpedance were seen in untreated models with different cellular components, with endothelial models having the lowest impedance and full-thickness models having the highest. Treated models containing sebocytes had reproducible oil production which was increased by Linoleic acid and reduced by TOFA and remarkably this was characterized by significant changes in bioimpedance in both the printed models and the supernatant surrounding them which demonstrated the accuracy of the technique to non-invasively measure differences in skin oil activity and was consistent with our own in vivo bioimpedance testing and that of the literature after hydration and oil application to donors. Donor testing revealed a mirror effect on the hydrated treated skin with coconut derivatives.
DISCUSSION AND CONCLUSION: Non-invasive methods to understand live donor testing have been seriously discussed for many years for ethical reasons. However, linking such data to the laboratory to make the development chain of cosmetics more effective has not been easy. Bioimpedance, linked to the oil production data provides an important step forward to help create effective and safe cosmetics, since the full-thickness models described here and linked with a simple chip system, accurately mirrored changes within the skin model and on live donors. Therefore, our models allow advanced translational lab-to-donor data to help bring sophisticated cosmetics ingredients to market faster, safely and more affordably.