Electrical Impedance Spectroscopy Quantifies Skin Barrier Function in Organotypic In Vitro Epidermis Models

Abstract

3 D human epidermal equivalents (HEEs) are a state-of-the-art organotypic culture model in pre-clinical investigative dermatology and regulatory toxicology. Here, we investigated the utility of electrical impedance spectroscopy (EIS) for non-invasive measurement of HEE epidermal barrier function. Our setup comprised a custom-made lid fit with 12 electrode pairs aligned on the standard 24-transwell cell culture system. Serial EIS measurements for seven consecutive days did not impact epidermal morphology and readouts showed comparable trends to HEEs measured only once. We determined two frequency ranges in the resulting impedance spectra: a lower frequency range termed EIS^diff correlated with keratinocyte terminal differentiation independent of epidermal thickness and a higher frequency range termed EIS^SC correlated with stratum corneum thickness. HEEs generated from CRISPR/Cas9 engineered keratinocytes that lack key differentiation genes FLG, TFAP2A, AHR or CLDN1 confirmed that keratinocyte terminal differentiation is the major parameter defining EIS^diff. Exposure to pro-inflammatory psoriasis- or atopic dermatitis-associated cytokine cocktails lowered the expression of keratinocyte differentiation markers and reduced EIS^diff. This cytokine-associated decrease in EIS^diff was normalized after stimulation with therapeutic molecules. In conclusion, EIS provides a non-invasive system to consecutively and quantitatively assess HEE barrier function and to sensitively and objectively measure barrier development, defects and repair.

Keywords: Human epidermal equivalents; TEER; electrical impedance spectroscopy; epidermal barrier; in vitro epidermis model; reconstructed human epidermis; skin barrier function.

Figure 1:. Design and function principle of a custom–made EIS device fitting the HEE culture system.

(a) Schematic overview of the EIS setup on HEE cultures. (b) Impedance and (c) phase spectrum of a fully–developed HEE culture after 8 days of air exposure. (d) Extended electrical equivalent circuit of an epidermal monolayer culture made up of the capacitance of the electrodes C_El, paracellular resistance R_P, transcellular resistance of cytoplasm R_Cyt, apical and basolateral membrane R_A, R_B as well as their capacitance C_A, C_B and the resistance of the medium R_Medium (adapted from (Yeste et al. 2018)). (e) Simplified electrical equivalent circuit of a HEE with the capacitance of both membranes taken together as C_Cell which together with R_P extends to a series of n parallel circuits in multi–layered 3 D culture systems (adapted from (Srinivasan et al. 2015) and (Groeber et al. 2015)). (f) Schematic overview indicating the contribution of individual electrical circuit parameters to the impedance spectrum (adapted from (Benson et al. 2013)). (g, h) Simulated impedance spectra illustrating the influence of changes in (g) paracellular flux (R_P) and (h) transcellular flux (C_Cell) (adapted from (Yeste et al. 2018)). (i) EIS impedance spectrum displaying EIS^diff (127–2212 Hz) and EIS^SC (28072–100000 Hz).

Figure 2:. Relative EIS measurements are reproducible and do not impair HEE development.

(a, b) EIS impedance spectra during HEE development with constructs being harvested (a) directly after measurements (endpoint measurements) and (b) at day 10 of air exposure (serial measurements). Each timepoint averages three biological replicates. (c) Comparison of EIS^diff and (d) EIS^SC between endpoint and serial measurements. (e) Histological comparison of HEEs undergoing endpoint or serial EIS measurements based on general morphology (H&E staining), differentiation status (FLG, IVL expression), proliferation (Ki67) and stress response (SKALP, KRT16). Pictures represent three biological replicates at day 10 of air exposure and taken at either 20x (Ki67) or 40x magnification. Size bars indicate 100 μm. (f) Impedance spectrum of HEEs (n = 3) measured 6 times within 1 hour at day 6 of air exposure. (g) Histological comparison of HEEs measured 6 times in 1 hour, either harvested directly or 24 hours after EIS measurements. Pictures represent three biological replicates and were taken at 40x magnification. Size bar indicates 100 μm.

Figure 3:. During HEE development, EIS^diff correlates with keratinocyte differentiation and epidermal thickness while EIS^SC correlates with stratum corneum thickness.

(a) Endpoint–measured impedance spectra, (b) EIS^diff and (c) EIS^SC during HEE development. Each timepoint represents three biological replicates and EIS^diff and EIS^SC were compared to 10 day air–exposed cultures. (d – g) Correlation of epidermal thickness with (d) EIS^diff and (e) EIS^SC and stratum corneum thickness with (f) EIS^diff and (g) EIS^SC. Each timepoint averages three biological replicates, R² values and significances indicate the correlation of individual replicates. (h) Staining of general morphology (H&E), keratinocyte differentiation (FLG, IVL) and cell–cell adhesions (CLDN1, CLDN4) during HEE development. Pictures represent three biological replicates and were taken at 40x magnification. Size bars indicate 100 μm. (i – l) Correlation of (i) FLG, (j) IVL, (k) CLDN1 and (l) CLDN4 protein expression with EIS^diff. Each timepoint averages three biological replicates, R² values and significances indicate the correlation of individual replicates.

Figure 4:. Stimulation with cytokines prove EIS^diff–determined barrier function to be independent of epidermal thickness.

(a) Endpoint–measured impedance spectra, (b) EIS^diff and (c) EIS^SC of cytokine–stimulated HEEs at day 8 of air exposure. Each condition represents three biological replicates and EIS^diff and EIS^SC were compared to control. (d) Epidermal thickness of cytokine–stimulated HEEs as compared to control. Correlation of (e) epidermal thickness to EIS^diff and (f) stratum corneum thickness to EIS^SC. Each condition averages three biological replicates, R² values and significances indicate the correlation of individual replicates. (g) HEEs stained for differentiation (FLG, IVL) and cell–cell adhesion (CLDN4) and proliferation (Ki67) proteins. Pictures represent three biological replicates and were taken at 20x (Ki67) or 40x magnification. Size bars indicate 100 μm.

Figure 5:. Knockout of genes involved in keratinocyte differentiation and cell–cell adhesion decreases EIS^diff.

(a) Endpoint–measured impedance spectrum, (b) EIS^diff and (c) EIS^SC of HEEs with knockout of target gene at day 10 of air exposure. Each condition represents three biological replicates and EIS^diff and EIS^SC are compared to control. Each (d) HEEs stained for differentiation (FLG, IVL) and cell–cell adhesion (CLDN1, CLDN4) proteins. Pictures represent three biological replicates and were taken at 40x magnification. Size bars indicate 100 μm.

Figure 6:. EIS detects therapeutic AHR response in a pro–inflammatory epidermis model.

(a) Endpoint–measured impedance spectrum, (b) EIS^diff and (c) EIS^SC of HEEs stimulated with IL–4 and IL–13 cytokines alone and in combination with AHR activating therapeutic compounds at day 8 of air exposure. Each condition represents three biological replicates and EIS^diff and EIS^SC of control conditions and AHR–binding compounds are compared to IL–4 + IL–13 stimulation.