A complex cytokine environment underlies the onset and progression of skin inflammatory diseases, including psoriasis and atopic dermatitis (AD). Although these diseases manifest with specific clinical features, they present common aspects, like disruption of the epidermal permeability barrier (EPB) and infiltration and activation of T helper (Th) cells (Chen et al., 2022). Th1 and Th17 cells play a key pathogenic role in psoriasis through the release of several cytokines including tumor necrosis factor alpha (TNF-α), and IL-17A, which, in turn, trigger the release of IL-6 and IL-1α by keratinocytes. The Th2-type cytokines IL-4 and IL-13 are primarily implicated in AD pathogenesis (Bieber, 2020). Keratinocytes are the major cell constituents of the epidermis serving both structural and immune functions (Jiang et al., 2020). Cytokine-induced abnormalities in the keratinocyte differentiation process are linked to the alteration of cutaneous homeostasis and to the impairment of the EPB function (Huang et al., 2022). In the pathogenesis of AD it is unclear whether the abnormal immune response precedes or follows the perturbation of the EPB. Mutations in the filaggrin (FLG) gene, reduced expression of structural epidermal proteins, and lipid abnormalities in response to Th2 inflammatory cytokines are all linked to a dysfunctional EPB (Weidinger et al., 2018; Bieber, 2020). Psoriasis begins with the activation of the immune system in response to environmental triggers and/or dysregulated immune tolerance. In psoriatic skin, the epidermal granular layer is reduced or almost absent, while the stratum corneum (SC) is enriched in keratinocytes with incomplete cornification. The epidermis appears thickened due to hyperplasia (Kok et al., 2023). A fine-tuned lipid composition and distribution are determinant of the proper EPB functionality. Lipid manufacture in keratinocytes is a highly orchestrated process accomplished along an articulated differentiation program (Chieosilapatham et al., 2021). Epidermal lipids (Weidinger et al., 2018) are composed by ceramides, free fatty acids (FFAs) and cholesterol in equimolar concentration (Kendall and Nicolaou, 2022). The composition and the lamellar organization of lipids in the EPB are altered in both AD and psoriasis. Lipidomics studies have supported the discovery of abnormal lipid composition in psoriatic skin (Li et al., 2016). Psoriatic keratinocytes show increased levels of multiple ceramide classes (NS, NP, AS, EOS) and decreased amounts of the NDS ceramide class compared to healthy controls (Luczaj et al., 2020). In AD, the content of ceramide species with long chain (more than 50 carbon atoms) in the NS and NDS classes is severely reduced, while species with shorter chain (less than 40 carbon atoms) present higher concentration level in the SC (Bhattacharya et al., 2019; Ishikawa et al., 2010). An increase in the relative abundance of unsaturated FFAs and decrease in C22-28 FFAs, which introduces a conformational disorder in the lipid matrix, has been described in the SC of lesional AD skin (Danso et al., 2017). An impairment of HMGCoA reductase activity and cholesterol synthesis contributes to the deterioration of the EPB function and the epidermal hyperproliferation in psoriasis (Kuwatsuka et al., 2021).

Calcium is a key regulator of keratinocyte differentiation in vivo and in vitro. Increased intracellular calcium levels trigger the recruitment of kinases and phospholipases, and, in turn, activate second messengers crucial for the differentiation program (Bikle et al., 2012; Xie et al., 2005). High concentration of calcium occurs primarily in the stratum granulosum (SG) (Bikle et al., 2012). Due to the pivotal role of calcium in the epidermal homeostasis and in the formation of the EPB, elevation of calcium concentration is commonly adopted to induce keratinocyte differentiation in vitro (Borowiec et al., 2013; Lamb and Ambler, 2013). The use of primary keratinocytes is preferred. However, it has drawbacks such as limited life span and heterogeneity because of donor-to-donor variability. In this study, we used the commercially available human keratinocyte cell line, Ker-CT, as a suitable system to reproduce the differentiation programme that occurs in primary keratinocytes upon elevation of calcium concentration (Smits et al., 2017; Reijnders et al., 2015; Beckert et al., 2019; Kovacs et al., 2020; Teshima et al., 2023). To validate the use of Ker-CT cells as a model, we compared their response to Th1/Th17 and Th2 cytokine stimulation in high calcium conditions with that obtained in a pool of human primary keratinocytes. We then analysed key markers of early and late keratinocyte differentiation and regulators of lipid metabolism. Nevertheless, the two-dimensional (2D) models are limited in reflecting the intricate structure and functions of the epidermis, including the EPB (Piasek et al., 2023). Thus, we used three-dimensional (3D) human epidermal equivalents (HEEs) to favour the EPB formation, which enabled us to investigate the interplay between immune signals conveyed by Th1/Th17 and Th2 cytokines and lipogenic pathways in an epidermal-like model.

HPLC/MS-grade acetonitrile, methanol and 2-propanol, were purchased from Biosolve (Chimie SARL, Dieuze, France; BV, Valkenswaard, Netherlands), while HPLC/MS-grade ethyl acetate was from Carlo Erba (Milan, Italy). UHPLC/MS-grade water was purchased from LiChrosolv by Merck (Darmstadt, Germany). Dimethyl sulfoxide (DMSO), the antioxidant butylhydroxytoluene (BHT) and the mobile phase modifiers ammonium fluoride (NH4F) and ammonium formate (NH4COOH) were purchased from Sigma Aldrich (Milan, Italy). The Q-TOF calibration solution was prepared in acetonitrile from Agilent Technologies Tuning mix (HP0321 solution, Agilent Technologies, CA, United States). Deuterated Ceramide LIPIDOMIX® (PN 330713) Mass Spec Standard Solution, EquiSPLASH™ LIPIDOMIX® (PN 330731) Mass Spec Standard Solution, N-palmitoyl-d31-D-erythro-sphingosine (d31-Cer16:0, MW 569) and deuterated cholesterol sulfate sodium salt (d7-CHS, MW 495) were purchased from Avanti Polar Lipids (Alabaster, Alabama, US). Deuterated cholesterol-2,2,3,4,4,6-d6 (d6-CH, MW 392) was purchased from Toronto Research Chemicals (Toronto, Ontario, Canada). Hexadecanoic-9,9,10,10,11,11,12,12,13,13,14,14,15,15,16,16,16-d17 acid (d17-PA, MW 273), glyceryl trihexadecanoate-d98 (d98TG 48:0, MW 906) and n-hexadecyl-1,1,2,2-d4 hexadecanoate-16,16,16-d3 (d7WE, MW 488) were purchased from CDN Isotopes Inc. (Pointe-Claire, Quebec, Canada). Additional information on internal standards is reported in Supplementary Table S1.

The immortalized human keratinocyte cell line Ker-CT (CRL-4048™) from American Type Culture Collection (ATCC®, Manassas, VA, United States) and a pool of neonatal primary keratinocytes from three different donors were cultured in defined Medium 154 with HKGS, L-glutamine (2 mM), penicillin (100 µ/mL), and streptomycin (100 µg/mlL, and Ca2+ (0.1 mM) at 37 °C under 5% CO2. For routine cell cultivation, cells were passaged when they reached 70%–90% of confluence. Cells were plated in fresh medium, in accordance with the experimental design (Supplementary Material S1). The cells maintained in Medium 154 in low calcium condition (0.1 mM) were used as control. Keratinocytes were induced to differentiate by switching the concentration of calcium in the culture medium from low (0.1 mM) to high (1.8 mM). Cells were stimulated with the pro-inflammatory cytokines, Th1/Th17 and Th2 type, in the pro-differentiative culture conditions. The Th1/Th17 cytokines, TNF-α (5 ng/mL), and IL-17A (10 ng/mL) with the addition of IL-6 (5 ng/mL) and IL-1α (10 ng/mL), from PeproTech (Rocky Hill, NJ, United States) were added to the fresh medium. IL-4 (10 ng/mL) and IL-13 (10 ng/mL) were used to investigate the Th2-type stimulus on keratinocytes. The medium was replaced every alternate day. Cells were harvested 2-, 4-, and 7-days post stimulation for gene and protein expression and lipid analysis. The Ker-CT cell line was used to generate 3D human epidermal equivalents (HEEs) as previously described (Flori et al., 2024). Briefly, Ker-CT were seeded on cell culture inserts (Thermo Scientific, Roskilde, Denmark; 0.4 µm pore size), maintained submerged for 3 days in CnT-Prime Epithelial Culture Medium (CnT-PR) (CellnTEC, Bern, Switzerland), and switched to CnT-Prime 3D Barrier Medium (CnT-PR-3D) in an air-liquid condition for 12 days. Fresh medium was replaced every alternate day. Mixtures of Th1/Th17 or Th2 cytokines, were added during the last 5 days of air-liquid culture. HEEs samples were processed for lipidomic profile. For histological analysis samples were formalin-fixed, paraffin-embedded and hematoxylin and eosin (H&E) stained.

Total RNA was isolated from keratinocytes using the Aurum™ Total RNA Mini kit, according to the manufacturer’s instructions. Total RNA samples were stored at −80°C until use. Following DNAse I treatment, cDNA was synthesized using a mix of oligo-dT and random primers and RevertAid™ First Strand cDNA synthesis kit according to the manufacturer’s instructions. Real-time RT-PCR was performed in a total volume of 10 μL with SYBR Green PCR Master Mix and 200 nM concentration of each primer. Sequences of all primers used are shown in Supplementary Table S2. Reactions were carried out in triplicate using a CFX96 Real Time System (Bio-Rad Laboratories S. r.l.). Melt curve analysis was performed to confirm the specificity of the amplified products. The mRNA expression was normalized to the mRNA expression of GAPDH by the change in the Δ cycle threshold (ΔCt) method and calculated based on 2−ΔCt. Results of the differentiated keratinocytes cultured in high calcium concentration were expressed as the fold change (FC) of control (taken as 1-fold). Data derived from differentiated keratinocytes treated with Th1/Th17 and Th2 cytokines were expressed as the FC relative to the high calcium condition. Results were represented as the mean ± SD of three independent experiments.

GCMS analysis of lipid extracts enabled the determination of FFAs and cholesterol; in epidermal equivalent models, sterol-like compounds were also identified and semi-quantified. The instrument used was an 8890 GC System combined with the 5977B Series MSD single quadrupole (Agilent Technologies, CA, United States). The analysis was performed on the HP-5MS UI fused silica column, chemically bonded with a 5%-phenyl-methylpolysiloxane phase (30 m × 0.250 mm internal diameter x 0.25 µm film thickness (Agilent Technologies CA, United States). Carrier gas (helium) flow rate was set at 1.2 mL/min. The GC oven program started at 80°C, reaching 280°C in 33 min and 310°C to final run time of 49 min. Samples were acquired in scan mode following EI (Ludovici et al., 2018) using MassHunter GC/MSD 5977B acquisition software (version 3.1.199). Data were processed by means of MassHunter Workstation Software Quantitative Analysis (version 10.1).

The chromatographic apparatus consisted of an Infinity II 1260 series HPLC equipped with a degasser, a quaternary pump, an autosampler and a column compartment (Agilent Technologies, CA, United States). RP-HPLC separation was conducted on a Zorbax Eclipse Plus C18 column (2.1 × 50 mm, 1.8 µm particle size, Agilent Technologies, CA, United States). The lipid extracts were eluted with a binary gradient of (A) 0.2 mM NH4F in water (18.2 Ω) and (B) 0.2 mM NH4F in methanol/2-propanol 80/20 (Maiellaro et al., 2024). The gradient elution was: 40% B, 0–2.0 min; 40%–99% B, 2.0–36.0 min, 99% B 36.0–46.0 min, 99%–40% B 46.0–48.0 min. A 10 min post-run time of 40% B was set for column re-equilibration. The column was thermostated at 60°C, the flow rate was 0.3 mL/min and the injection volume was 1 µL. HILIC separation was performed with a HALO HILIC column, 2.1 × 50 mm, 2.7 µm particle size (Advanced Materials Technology, AZ, United States), thermostated at 40°C. The mobile phase consisted of (A) 5 mM NH4COOH in water (18.2 Ω) and (B) acetonitrile. The flow rate was maintained at 0.4 mL/min during run time (22 min) and post-run time (10 min). The injection volume was 0.4 µL. Lipid extracts were eluted as follows: 98% B, 0–1.0 min; 98%–80% B, 1.0–18.0 min; 80% B, 18.0–20.0 min; 80%–98% B, 20.0–21.0 min, 98% B, 21.0–22.0 min.

Statistical analysis was performed by Agilent MassHunter Mass Profiler Professional (MPP) (version 15.1). A template (.csv) containing the final concentrations for the 107 annotated species in 2D keratinocyte lipid profiles expressed in pmol/mg protein was imported into the software. Samples were divided into groups according to the treatment, i. e., Ctr, ↑Ca2+, ↑Ca2++Th1/Th17, ↑Ca2++Th2 to determine significant differences of compounds across groups. The data in linear scale were log-transformed before performing statistical analysis. The log2 transformed concentration of lipid analytes is provided in Supplementary Table S3. An unpaired Student’s t-test was applied using Benjamini-Hochberg and false discovery rate (FDR) correction. The corrected p-value (p) and fold change (FC) cut-off were 0.05 and 1.5, respectively. Differences were considered statistically significant with p ≤ 0.05 and FC ≥ 1.5.

The values corresponding to the concentration expressed as pmol/mg protein of the 300 species data annotated in HEEs were log-transformed before performing statistical analysis. One-way analysis of variance (one-way ANOVA) was performed to determine relevant differences according to the treatment, i. e., Ctr, Th1/Th17, and Th2. Benjamini-Hochberg was used for multiple testing correction, and Tukey HSD as the post hoc test.

To analyse the effects induced by Th1/Th17 and Th2 cytokines on the expression of genes related to the calcium-induced differentiation process, the immortalized human keratinocyte cell line Ker-CT was exposed to mixtures of Th1 (TNF-α)/Th17 (IL-17A) plus IL-6 and IL-1α, or Th2 (IL-4, IL-13) cytokines for 2, 4, and 7 days (Supporting information). A time-dependent increase in gene expression of K10, FLG, LOR, CASP 14, IVL, and TGM1 was observed (Supplementary Figure S1). Th1/Th17 cytokines induced a significant decrease of the K10 expression at 4 and 7 days. The Th2 cytokines increased K10 expression only at 7 days. The induced expression of FLG remained unaffected following both Th1/Th17 and Th2 challenges. In contrast, both Th1/Th17 and Th2 cytokine mixtures decreased the LOR and CASP14 mRNAs. The early effects on LOR mRNA expression perdured up to 7 days, while effects on CASP14 mRNA were observed exclusively at the latest time point. The induction of IVL and TGM1 mRNA expression during differentiation was further increased by Th1/Th17, while it was decreased by Th2 cytokines (Supplementary Figure S1).

Consistent with the mRNA data, the Th1/Th17 cytokines reverted the calcium-induced K10 protein expression (Figures 1A, C), while, no significant change was observed after Th2 cytokines (Figures 1B, C). IVL protein expression was not affected by stimulation with both cytokine mixtures (Figures 1A, B). In contrast, both cytokine types counteracted the calcium-induced LOR and FLG protein expression (Figures 1A–C).

Figure 1. Western blot and corresponding densitometric analysis of K10, IVL, LOR, FLG (monomer, 37 kDa) expression in Ker-CT cells in differentiating conditions with ↑Ca2+ and after treatment with Th1/Th17 (A) and Th2 (B) cytokines at 7 days. GAPDH or HSP70 were used as the endogenous loading controls. Results are expressed as the fold change of the low calcium control cells (Ctr). Data represent the mean ± SD of three independent experiments (significance vs. low calcium control or vs. stimulated cells with ↑Ca2+ are marked with * and °, respectively; *p < 0.05, **p < 0.01 vs. low calcium cells; °p < 0.05, °p < 0.01 vs. differentiated cells). (C) Parallel immunofluorescence analysis of K10, LOR and FLG following stimulation with Th1/Th17 or Th2 cytokine types. Nuclei are counterstained in DAPI. Arrows point at FLG positive cells. Scale bar: 50 μm.

To validate the use of Ker-CT as a suitable model for studying the effects of Th1/Th17 and Th2 cytokines, we compared their response with that induced in human primary keratinocytes. Treatment of primary keratinocytes with the cytokine mixtures modulated the expression of the differentiation markers induced by high calcium condition in a fashion similar to that observed in Ker-CT cells (Supplementary Figure S2).

Genes of lipid metabolism were modulated in differentiated Ker-CT. De novo synthesis and desaturation of fatty acids (FAs) was repressed at 7 days in high calcium, as demonstrated by the lower mRNA levels of FAS, FADS2, and SCD1 (Figure 2). Both Th1/Th17 and Th2 cytokine mixtures contrasted the decrease of SCD1 mRNA levels, while only Th2 cytokines demonstrated to counteract the effects on FAS and FADS2 mRNA expression induced in the differentiation process (Figure 2A).

Figure 2. (A) Quantitative real time PCR analysis of genes involved in the synthesis of FFAs, ceramides and cholesterol in Ker-CT cells treated with ↑Ca2+, ↑Ca2++Th1/Th17 and ↑Ca2++Th2 for 2, 4 and 7 days. All mRNA values were normalized against the expression of GAPDH and were reported relative to low calcium control (Ctr). Data represent the mean ± SD of three independent experiments (significance vs. low calcium control or vs. stimulated cells with ↑Ca2+ are marked with * and °, respectively; *p < 0.05, **p < 0.01 and ***p < 0.001 vs. low calcium cells; °p < 0.05, °p < 0.01 vs. differentiated cells). Results are expressed as fold change relative to the value of cells grown in low calcium, which was set as 1 by definition. Quantitative real time PCR, Western blot with corresponding densitometry, and immunofluorescence analysis of (B) SLC27A4 and (C) ELOVL3 in Ker-CT cells in differentiating conditions with ↑Ca2+ and after treatment with Th1/Th17 and Th2 cytokines at 7 days. All mRNA values were normalized against the expression of GAPDH and were reported relative to low calcium control. Data represent the mean ± SD of three independent experiments (*p < 0.05, ***p < 0.001 vs. low calcium cells; °p < 0.05 vs. differentiated cells). Results are expressed as fold change relative to the value of cells grown in low calcium, which was set as 1 by definition. GAPDH was used as the endogenous loading controls for Western blot analysis. Results are expressed as the fold change respect to low calcium control cells. Data represent the mean ± SD of three independent experiments (*p < 0.05 vs. low calcium cells; °p < 0.05 vs. differentiated cells). Nuclei are counterstained with DAPI. Scale bar: 50 μm.

The mRNA levels of HMGCR, gene of the mevalonate pathway in the cholesterol synthesis (Chen et al., 2023), significantly increased upon elevation of calcium concentration. Th1/Th17 and Th2 cytokines reverted the induction of HMGCR mRNA. ALOX12B plays a key role in the skin barrier function, serving oxidation of acyl-ceramides and protein crosslink (Meyer et al., 2023). Its mRNA level significantly increased in high calcium at all time points and was further augmented after treatment with both Th1/Th17 and Th2 cytokine mixtures at 4 and 7 days. DEGS1, responsible for the conversion of dihydroceramide into ceramide (Planas-Serra et al., 2023), significantly increased upon 7 days culture in differentiating conditions. Th1/Th17 cytokines decreased, whereas Th2 further increased DEGS1 mRNA expression. Similar results were observed for the mRNA levels of DEGS2, a gene involved in the synthesis of sphingoid bases (Ota et al., 2023). In keratinocytes, ABCA12 serves the transmembrane lipid transport, particularly of ceramides, to form extracellular lipid layers (Proksch et al., 2008; Akiyama, 2014) and SLC27A4 (Schmuth et al., 2005) acts as a FFA trasporter. The mRNA levels of ABCA12 and SLC27A4 significantly increased upon elevation of calcium concentration (Figure 2B). The rise of ABCA12 levels started at 4 days and further increased at 7 days, while SLC27A4 was increased at 7 days (Figure 2B). The treatment with both cytokine types significantly decreased the ABCA12 mRNA expression at 7 days. The slight decrease of the SLC27A4 mRNA concided with more pronounced effects on the protein expression (Figure 2B). Th1/Th17 stimulation reverted the prominent induction of both ELOVL3 mRNA and protein levels at 7 days. Effects of Th2 cytokines were observed only on the reduction of ELOVL3 protein levels (Figure 2C). The Supplementary Table S4 reports the mean fold changes in the expression of the evaluated genes involved in the lipid metabolism. Analysis of key genes of the lipid metabolism in human primary keratinocytes after stimulation with Th1/Th17 and Th2 cytokines revealed a behaviour comparable to that of Ker-CT cells (Supplementary Figure S2), further supporting the suitability of the employed model.

Statistically significant differences were examined using pairwise comparisons. The volcano plot in Figure 3A shows the changes induced in the abundance of lipid species in high calcium differentiated keratinocytes. As reported in Table 1, 53 species were modulated at a statistically significant level (FC ≥ 1.5; p-value ≤0.05). The overall effect was an increase of lipid amounts in differentiated conditions. Specifically, 46 compounds were upregulated, i.e., cholesterol and CHS, most Cer [NS], including the short-chain Cer [NDS]34:0, some SMs, MUFAs, and PCs with 36–40 carbon atoms. Long-chain PEs (38–42 carbon atoms) were upregulated, while 7 short-chain PEs (31–36 carbon atoms) were downregulated. Overall, the presence of Th1/Th17 or Th2 cytokines modified the abundance of lipid species observed in high calcium concentration (Figures 3B, C). In differentiated keratinocytes exposed to Th1/Th17 cytokines, 25 and 5 species were significantly decreased and increased, respectively (Figure 3B). Th1/Th17 cytokines decreased the abundance of SMs, short-chain ceramides, FA 22:1, FA 24:1, PEs (except PE 32:2 and PE 34:0), and that of some PCs with different numbers of carbon atoms and double bonds (Table 2). The keratinocytes treated with Th2 cytokines showed a significant modulation of 27 lipid species (Figure 3C). These cytokines showed a suppressive effect, especially against FA synthesis. In fact, although minimally affected by differentiation, saturated FAs (SFAs) showed a strong decrease. Furthermore, Th2 cytokines caused the decrease of Cer [NS]46:1, highly unsaturated PCs, PE 34:2, and PE 42:5. The abundance of a few species, such as Cer [NS]46:2, PE 34:0 and PE 36:5, was upregulated (Table 3). Effects of Th1/Th17 and Th2 cytokines were compared in the volcano plot in Figure 3D, where 31 lipid species presented with significant different concentration (Table 4). The Th2 signal caused the upregulation of 20 compounds, i. e., CHS, FA 22:1, FA 24:1, short-chain ceramides, and some phospholipid. Eleven (11) compounds, i.e., 6 saturated FFAs, Cer [NS]42:1, Cer [NS]46:1 and 3 PCs, were downregulated. The Venn’s diagram in Figure 3E supports visualizing the effects consequent to Th1/Th17 and Th2 challenges compared to the high calcium condition. The table embedded in Figure 3 reports the individual lipid species at the intersection of the three subsets in the Venns’ diagram along the direction of their modification. The abundance of about half of the 53 species modulated by high calcium was modified by the cytokine environments. The majority of the 30 lipid species affected by the Th1/Th17 coincided with those characterizing the pro-differentiating conditions. In contrast only 9 out of the 27 species modified upon Th2 challenge, were common to the high calcium conditions. The Venn’s diagram shows that 5 species, i.e., Cer [NS]46:2, PC 36:6, PC 38:6, PC 38:5, and PE 42:5, were modulated in all three conditions. The number of compounds whose concentration was modified upon challenge of differentiated keratinocytes with Th1/Th17 or Th2 cytokines, was 18 and 4, respectively. Both Th1/Th17 and Th2 cytokine mixtures resulted in a decrease of PE 34:2 and an increase of PE 34:0 levels. The abundance of these compounds was not affected by differentiation alone. Further analyses were conducted to investigate the role of cholesterol in the in vitro system. CEs were detected in the keratinocytes lipid extracts in positive ion mode. MS data mining supported the presence of two CE species, i.e., CE (16:1) and CE (18:1). The abundance of both CE species was induced by high calcium at a significant extent. Th2 signals caused a significant increase in the concentration of CE (18:1) (Supplementary Figure S3).

Figure 3. Results of the lipidomic analysis of Ker-CT cells under the different conditions at 7 days. Volcano plots depicting the following paiwise comparisons: (A) ↑Ca2+ vs. low calcium control (Ctr), (B) (↑Ca2++Th1/Th17) vs. ↑Ca2+, (C) (↑Ca2++Th2) vs. ↑Ca2+(D) (↑Ca2+ +Th2) vs. ↑Ca2+ +Th1/Th17. Each dot corresponds to a lipid species: Blue and red dots indicate species that are statistically (p ≤ 0.05) decreased or increased (FC ≥ 1.5), respectively. Upregulated and downregulated species that did not pass the significance cut-off are in orange and light blue, respectively. Compounds that passed significance p ≤ 0.05 and FC ≥ 1.5 cut-offs are reported in Table 1. (E) Venn diagram and corresponding table showing the number of lipid species commonly and specifically modified by Th1/Th17 and Th2 cytokines.

Table 1. 53 Lipid species upregulated and downregulated in differentiated Ker-Ct compared to control (the cut-offs of fold change >1.5 and p ≤ 0.05).

Table 2. 30 Lipid species upregulated and downregulated in [↑Ca2+ +Th1/Th17] Vs. [↑Ca2+] (the cut-offs of fold change >1.5 and p ≤ 0.05).

Table 3. 27 Lipid species upregulated and downregulated in [↑Ca2+ +Th2] Vs. [↑Ca2+]. (the cut-offs of fold change >1.5 and p ≤ 0.05).

Table 4. 31 Lipid species upregulated and downregulated in [↑Ca2+ +Th2] Vs. [↑Ca2+ +Th1/Th17] (the cut-offs of fold change >1.5 and p ≤ 0.05).

We next investigated the association between the mRNA expression of specific ceramide synthesis-related enzymes and the levels of the corresponding lipids. Five types of ceramide synthase 2–6 (CERS 2–6) have been identified in human keratinocytes and are known to be involved in the skin barrier alteration associated with AD (Ito et al., 2017). Figure 4 displays the gene expression of CERS3, 4, and 6, which are involved in the synthesis of the Cer [NS] class, and the abundance profile of their corresponding member with specific FA chain lengths (Park et al., 2010; Cingolani et al., 2016). CERS3 mRNA expression was promoted in differentiated keratinocytes. A pronounced increase was observed upon treatment with Th1/Th17 cytokines. The elevation of CERS3 due to differentiation remained unchanged following Th2 challenge. The amount of the Cer [NS]42:1, which is the prominent product of CERS3, resulted unmodified by increasing calcium concentration, whereas it was significantly induced by Th1/Th17 stimulation (Figure 4A). The expression of CERS4 augmented in response to high calcium and this effect was not modified by the co-treatment with both cytokine types. In contrast, Th2 stimulation decreased the concentration of the corresponding metabolite Cer [NS]40:1 compared to high calcium (Figure 4B). The comparison between CERS6 and its product is depicted in Figure 4C. A significant elevation of CERS6 mRNA and Cer [NS]34:1 concentration was observed in differentiated conditions compared to the control. This effect was counteracted by both Th1/Th17 and Th2 cytokine mixtures.

Figure 4. Comparison between gene expression and corresponding synthesized lipid species. Quantitative real time PCR analysis of CERS3 (A), CERS4, (B) CERS6 (C) and respective products in differentiated Ker-CT cells, co-treated with Th1/Th17 or Th2, for 7 days. All mRNA values were normalized against the expression of GAPDH and were expressed relative to low calcium control cells (Ctr). The data are expressed as mean ± SD of three independent experiments (significance vs. low calcium control or vs. stimulated cells are marked with * and °, respectively; *p < 0.05 and **p < 0.01 vs. low calcium control; °p < 0.05 and °p < 0.01 vs. stimulated cells with high calcium).

To gain insights into the influence of the cytokines under investigation, we opted to integrate the findings from the 2D system with the lipidomic analyses conducted in 3D organotypic cultures. In contrast to conventional 2D cell culture, the reconstruction of 3D equivalents can mimic human epidermis in terms of layering, differentiation and barrier formation. The exposure of HEEs to mixtures of Th1 (TNF-α)/Th17 (IL-17A) plus IL-6 and IL-1α, or Th2 (IL-4, IL-13) cytokines during the last 5 days of the air-liquid interphase culture, induced dissimilar phenotypes. The HEEs showed a well multilayered and differentiated epidermis with the presence of the stratum corneum (Supplementary Figure S4A). Stimulation with psoriasis-associated cytokines affected epidermal morphology, resulting in loss of the stratum granulosum, thickening of the stratum corneum and parakeratosis (Supplementary Figure S4B). Instead, AD-associated cytokines induced signs of spongiosis, as evidenced by the presence of intercellular spaces between adjacent keratinocytes (Supplementary Figure S4C). The lipids were extracted from HEEs and analyzed with the same procedure applied to 2D cultures. HEEs cultures enabled the identification of a wide range of ceramides and a distinct group of hexosylceramides. Moreover, the list of analysed species was expanded to include the determination of TGs and DGs, in addition to FFAs, CHS, and phospholipids, resulting in a total of 300 targeted species. One-way analysis of variance (ANOVA) of the lipidomics data supported comparation among the groups (i.e., control, Th1/Th17 and Th2). The test retrieved 141 lipid species that showed statistically significant differences (Supplementary Table S5). Filtering out the species with FC < 2 versus the unchallenged samples, retrieved 64 lipid species. To explore relationships and correlations between lipids, the expression profiles of the 64 species were organized in a hierarchical clustering (Figure 5). The three columns represent the data for each experimental condition, expressed as the mean of six samples per group. The colour intensity indicates the logarithmically transformed abundance of each lipid species, ranging from dark blue (low abundance) to red (high abundance). The dendrogram supports the identification of shared behaviours and interrelationships among the lipids in the three conditions (Supplementary Table S6). The hierarchical clustering illustrates that the 12 most abundant entities in the 3D system (predominantly TGs and Cer [NS]34:1) are grouped together. These species are distinguished by an upregulation induced by Th1/Th17, whereas Th2 appears to exert a minimal influence. A second cluster of 13 species, i.e., TGs and ceramides of different subclasses, displays a parallel behavioural response to the two cytokines. Nevertheless, these species are distinguished by a relatively lower abundance. Two distinct groups are evident in the central region of the heatmap. A cluster of five species demonstrates a downregulation induced by both Th1/Th17 and Th2 cytokines, which includes palmitoleic acid (FA 16:1n-7). The second cluster encompasses 12 lipids, including members of the TG class and long-chain ceramides. This group exhibits a notable increase in response to Th1/Th17 cytokines and a mild decrease in response to Th2. In the lower region of the heatmap, which is characterised by the less abundant species in the 3D system, the dendrogram highlights a significant depletion of HexCers with chain length in the range 34–46 carbon atoms, induced by Th1/Th17. The depletion of HexCers observed with Th2 cytokines was statistically significant only for the species with chain length in the range 42–46 carbon atoms (Supplementary Figure S5). Furthermore, an additional group can be identified which corroborates the observed increase in TGs and ceramides, which is promoted by Th1/Th17. Lignoceric acid (FA 24:0) is a constituent of this group. All results of expression analyses and lipidomics were summarized in Figure 6.

Figure 5. Hierarchical clustering of 64 lipid species, whose abundance was modulated upon treatment with Th1/Th17 or Th2 cytokines of 3D epidermal equivalents. The three columns in the heatmap represent the averaged data of six samples per experimental condition. The color intensity indicates the logarithmically transformed abundances of each lipid species, ranging from dark blue (lowest abundance) to red (highest abundance). The dendrogram supports the identification of shared behaviours and interrelationships among the lipids in the three conditions. The 64 lipid species are listed in cluster order in the Supplementary Table S6.

Figure 6. Schematic representation of the effects of Th1/Th17 and Th2 cytokines on protein and gene expression, and lipid modifications in differentiated keratinocytes. Blue and dark pink arrows indicate the effects of Th1/Th17 and Th2 cytokine mixtures respectively. The genes of the lipid pathways depicted in the keratinocyte cartoon on the right are in red or in black, respectively, when their expression was modulated or unmodulated by the Th1/Th17 or the Th2 cytokine signals.

Th1/Th17 and Th2 cytokines are primary contributors to the pathogenesis of psoriasis and AD, which exhibit deranged differentiation and lipid abnormalities in the epidermis. The differentiation program of keratinocytes is tightly intertwined with the generation of the hydrophobic EPB (Berdyshev et al., 2018; Pavel et al., 2022). The gradient of the calcium concentration in the epidermis regulates several intracellular events deploying the differentiation program leading to the functional EPB. The events of the abnormal epidermal differentiation characteristic of the Th1/Th17 and Th2 environments are not fully elucidated. Several models have been optimized to mimic in vitro the complex inflammatory milieus of AD or psoriasis using different cytokine mixtures (Ubago-Rodríguez et al., 2024; Alsabbagh, 2024; Quílez et al., 2024). The Th2 cytokines IL-4 and IL-13 are commonly used to model AD. For the psoriasis model, the most studied cytokines are TNF-α, IL-6, IL-1α, IL-17A, and IL-22 (Ubago-Rodríguez et al., 2024; Alsabbagh, 2024; Quílez et al., 2024). Some authors also evaluated the effects of the Th1-type IFN-γ (Chiricozzi et al., 2014; Nograles et al., 2008). In this study, we investigated the distinctive and shared responses to Th1/Th17 and Th2 cytokine mixtures on 2D cultures induced to differentiate, and organotypic models of epidermis using Ker-CT cells. The comparison of the responses to Th1/Th17 and Th2 cytokines provided matching results in terms of differentiation and lipid metabolism modulation in Ker-CT and human primary keratinocytes. In 2D cultures of Ker-CT, the calcium-induced protein expression of IVL, LOR, and FLG was tendentially decreased by both cytokine types, whereas K10 expression was significantly decreased and increased in keratinocytes treated with Th1/Th17 and Th2 cytokines, respectively. These results are in agreement with the decreased expression of K10 in psoriasis (Totsuka et al., 2017), in contrast to what observed in AD (Bovenschen et al., 2005). IVL expression is induced by IL-13, IL-17A, ET-1, TNF-α, and IFN-γ in both psoriatic and normal keratinocytes (Chen et al., 2013). In our study, opposite to the IVL protein levels, the IVL mRNA expression promoted by calcium was further elevated upon Th1/Th17 challenge. This divergent regulation may suggest the promotion of IVL removal consequent to Th1/Th17 cues (Zieba et al., 2017). Consistent to what observed for the LOR protein levels, both Th1/Th17 and Th2 cytokine mixtures decreased the LOR mRNA expression, as previously reported (Rabeony et al., 2014; Kim et al., 2008). The FLG protein expression was decreased by both cytokine types along a prominent downregulation of CASP14, which processes profilaggrin to FLG (Hoste et al., 2011). This result supports the detrimental effects of inflammatory cytokines exerted directly on the EPB (Hvid et al., 2011). TGM1 plays an essential role in skin barrier formation by cross-linking proteins in differentiated keratinocytes. In our model, Th1/Th17 and Th2 stimulation induced opposite effects on the TGM1 mRNA expression promoted by high calcium, in agreement with its pronounced expression in psoriatic lesions (Surbek et al., 2023) and no significant changes in AD lesion (Liedén et al., 2012).

Both the formation and the integrity of the EPB are governed by the build-up of epidermal lipids, i.e., cholesterol, ceramides, and FFAs (Bhattacharya et al., 2019). In healthy SC, lipids are arranged predominantly in a dense lamellar formation (Danso et al., 2017). Several lipogenic pathways were regulated in the differentiated Ker-CT. The downregulation of FAS, FADS2, and SCD1 genes and the upregulation of genes