High-salt diet aggravates skin inflammation of psoriasis-like mouse model with CCL20‒CCR6 axis further activation

Abstract

Epidermal keratinocytes are important for maintaining the skin barrier by interacting with immune cells and environmental stimuli. In this study, we reveal that a high-salt diet aggravated imiquimod-induced skin inflammation in a psoriasis-like mouse model, characterized by activation of the CCL20–CCR6 axis and increased infiltration of CCR6⁺ γδT cells in skin lesions. In the presence of additional sodium, the mRNA expression of Ccl20 was upregulated in keratinocytes, mediated by JNK/p38–SGK1 pathway. Consistent with our experimental findings, a positive correlation between SGK1 and CCL20 gene expression was also observed in the skin of patients with psoriasis. Taken together, these data demonstrated that a high-salt diet aggravated skin inflammation in a psoriasis-like mouse model through further activation of the CCL20‒CCR6 axis.

Introduction

Previous studies have shown that sodium accumulates in the skin of rats fed on a high-salt diet (HSD) (Machnik et al, 2009). Furthermore, sodium accumulation in the skin has been positively correlated with the severity of psoriasis (Maifeld et al, 2022). However, the influence of HSD on psoriasis and its underlying immunity-dependent mechanisms remains largely unknown. It has been reported that mice fed an HSD develop more severe autoimmune disease owing to enhanced polarization of pathogenic T helper 17 (Th17) cells (Kleinewietfeld et al, 2013; Wu et al, 2013). Inflammatory diseases, including psoriasis, are driven by the cytokine IL-17 (Nestle et al, 2009), and the CCR6–CCL20 axis is critical for recruiting IL-17–producing CCR6⁺ T cells into the lesional skin (Cai et al, 2011; Liu et al, 2022). Keratinocytes (KCs) are important for maintaining the skin barrier by interacting with immune cells and the external environmental inputs, which also play a critical role in producing proinflammatory cytokine CCL20 (Harper et al, 2009; Jin et al, 2022).

Western diets with excessive salt are associated with an elevated incidence of autoimmune and inflammatory diseases (Anzai et al, 2019; Ezzati and Riboli, 2013). As a serine/threonine kinase, SGK1 is widely expressed in various cell types (Firestone et al, 2003) and can be regulated by extracellular sodium ion (Binger et al, 2015). Multiple studies highlighted the critical role of SGK1 signaling pathway in immune modulation in the context of sodium chloride (NaCl). Previous studies demonstrated that HSD led to more severe symptoms of autoimmune encephalomyelitis in animal model, which is dependent on activating SGK1 in Th17 cells (Kleinewietfeld et al, 2013; Wu et al, 2013). Moreover, SGK1 also mediates salt-driven pathogenic program in regulatory T cells and T helper 2 cells (Côrte-Real et al, 2023; Matthias et al, 2019; Yang et al, 2020).

We became interested in CCL20‒CCR6 axis because increasing sodium can promote chemokine production in epithelial cells, including CX3CL1 and CCL2 (Berry et al, 2017). Thus, we hypothesize that SGK1 might upregulate CCL20 in KCs to modulate psoriasis-like dermatitis in the context of NaCl. The results of our study revealed that HSD aggravates psoriasis-like skin inflammation with CCL20‒CCR6 axis activation, and NaCl-induced CCL20 expression in KCs is regulated through the JNK/p38–SGK1 signaling pathway. To our knowledge, the effect of sodium on CCL20‒CCR6 axis has not been identified.

Results

HSD aggravates skin inflammation in psoriasis-like mouse model

We first investigated whether HSD contributes to psoriasis pathogenesis. Imiquimod (IMQ)-induced psoriasis-like mice fed an HSD (HSD/IMQ mice) developed typical psoriasis-like features on day 3, presenting more pronounced erythema and swelling (Figure 1b) and a higher Psoriasis Area and Severity Index (PASI) score (Figure 1c) than ND/IMQ mice. Histological analysis showed a significant increase in epidermal thickness in HSD/IMQ mice compared with that in ND/IMQ mice (Figure 1d and e). The percentage of Ly6G⁺ neutrophils in the skin was higher in HSD/IMQ mice than in ND/IMQ mice (Figure 1f). Consistently, proportion of Gr1⁺ CD11b⁺ neutrophils was significantly increased in the lesional skin of HSD/IMQ mice (Figure 1g). No obvious skin lesions were observed in the mice fed a normal diet (ND/control) or HSD (HSD/control).

Figure 1.

HSD exacerbates psoriasis-like phenotype in the IMQ-induced mouse model. (a) Schematic representation for mouse model establishment (C57BL/6J WT mice received the ND [black line] or HSD [red line] for 60 days before daily IMQ application [blue line] for 3 days on ND/IMQ mice and HSD/IMQ mice). (b) Representative images of psoriasis-like phenotypes in each group. (c) PASI scores were calculated daily. ∗P < .05 comparing HSD/IMQ mice with ND/IMQ mice on day 3. (d) Representative H&E-stained skin sections and quantification of epidermal thickening in each group. Bar = 100 μm. (e) Quantitated epidermis thickness of skin in each group. (f) Representative IHC-stained sections of Ly6G in skin in each group. Bar = 100 μm. (g) Flow plots gated on CD45⁺ cells from dorsal skin showing Ly6G⁺ and Gr1⁺ cell populations. Data are shown as mean ± SEM (n = 3 for each group). For Figure 1c, P-values were determined by repeated-measures 2-way ANOVA with interaction. For Figure 1e–g, P-values were determined by 2-way ANOVA with interaction. ∗P < .05, ∗∗P < .01, ∗∗∗P < .001, and ∗∗∗∗P < .0001. HSD, high-salt diet; IHC, immunohistochemistry; IMQ, imiquimod; ND, normal salt diet; WT, wild-type.

HSD exacerbates psoriasis-like skin inflammation through the CCL20‒CCR6 axis

We then explored whether the CCL20–CCR6 axis was affected in HSD/IMQ mice. Increased mRNA expression of S100a8, Ccl20, Il23, and Il17a was found in the skin lesions of HSD/IMQ mice, compared with that in ND/IMQ mice (Figure 2a). Because CCL20 is the ligand of CCR6, we found that the percentage of TCRγδ^int T cells, CCR6⁺ γδT cells, IL-17A–secreting CD3⁺ T cells, and IL-17A–secreting TCRγδ^int T cells was significantly higher in skin lesions of HSD/IMQ mice than in those of ND/IMQ mice (Figure 2b–e). Intriguingly, although HSD/control mice lacked overt psoriasis-like manifestations, they exhibited a trend toward increased Ccl20 mRNA expression (Figure 2a) and significantly elevated proportions of TCRγδ^int T cells and CCR6⁺ γδT cells, compared with ND/control mice (Figure 2b and c).

Figure 2.

The CCL20‒CCR6 axis is further activated in psoriasis-like mouse model fed with HSD. (a) qPCR analysis of mRNA encoding chemokines and cytokines of the dorsal skin in each group. (b) Flow plots gated on CD3⁺ cells derived from dorsal skin showing γδTCR^int cells populations. (c) Flow plots gated on γδTCR^int cells derived from dorsal skin showing CCR6⁺ cells populations. (d) Flow plots gated on CD3⁺ cells derived from dorsal skin showing IL-17A⁺ cells populations. (e) Flow plots gated on γδTCR^int cells derived from dorsal skin showing IL-17A⁺ cells populations. Data are shown as mean ± SEM (n = 3 for each group). P-values are determined by 2-way ANOVA with interaction. ∗P < .05, ∗∗P < .01, ∗∗∗P < .001, and ∗∗∗∗P < .0001. HSD, high-salt diet; IMQ, imiquimod; ND, normal salt diet; SSC-A, side scatter area.

Environmental sodium upregulates Ccl20 expression in KCs through the JNK/p38–SGK1 pathway in vitro

Given that increased sodium intake intensifies psoriasis-like skin inflammation with upregulated Ccl20 mRNA expression and considering prior studies demonstrating that p38–SGK1 pathway regulated NaCl-treated Th17 cells (Wu et al, 2013), we hypothesized that sodium may regulate chemokines, such as CCL20, in KCs. Thus, we cultured primary murine KCs with additional NaCl and observed elevated expression of Ccl20, Sgk1, Cxcl1 as well as increased phosphorylation of p38 and JNK (Figure 3a and b). Furthermore, treatment with a JNK inhibitor (SP600125), p38 inhibitor (SB203580), and SGK1 inhibitor (GSK650394) suppressed NaCl-induced Ccl20 mRNA expression (Figure 3c). Notably, JNK and p38 inhibitors also downregulated Sgk1 mRNA expression in additional NaCl-treated KCs (Figure 3c), suggesting that the JNK/p38–SGK1 pathway mediates NaCl-induced CCL20 expression.

Figure 3.

Increasing sodium concentration upregulates Ccl20 expression in KCs through the JNK/p38–SGK1 pathway. (a) qPCR analysis of selected chemokines, cytokines, and Sgk1 in KCs (n = 5 for each group). (b) Immunoblot analysis of p38, phosphorylated p38, JNK, and phosphorylated JNK in KCs for different time (0 min, 30 min, 60 min), and tubulin was used as control. The relative densitometry of all bands (arbitrary unit) was normalized to the level of 0 min (n = 3 for each group). (c) qPCR analysis of Ccl20 and Sgk1 RNA expression in KCs treated with pharmacological inhibitors (n = 3 for each group). Data are shown as mean ± SEM. For Figure 3a, P-values were determined by 2-tailed Student’s t-test. For Figure 3c, P-values were determined by 1-way ANOVA. ∗∗P < .01, ∗∗∗P < .01, and ∗∗∗∗P < .0001. KC, keratinocyte; min, minute.

Positive correlation between the gene expression of SGK1 and CCL20 in the skin of patients with psoriasis

To assess human relevance, we turned to psoriasis-related datasets and gene expression profiles from GSE13355 were obtained. Compared with those in healthy and nonlesional skin, the expression of CCL20 and SGK1 were significantly upregulated in lesional skin of patients with psoriasis (Figure 4a and b). Moreover, the expressions of CCL20, CXCL1, CXCL2, CXCL8, ICAM1, IFNG, IL17a, IL1b, IL22, IL23a, IL26, IL 6 and TNF were positively correlated with SGK1 (Figure 4c). Of note, CCL20 expression showed the strongest correlation with SGK1 among selected cytokines and chemokines (Figure 4d). In addition, we analyzed the RNA sequencing of lesional skin from ND/IMQ and HSD/IMQ mice. Most of the significant Kyoto Encyclopedia of Genes and Genomes terms in the lesional skin were IL-17–related immune responses (Figure 4e). Gene set enrichment analysis identified upregulated Kyoto Encyclopedia of Genes and Genomes terms, including FoxO and ErbB signaling pathway, in HSD/IMQ mice, along with increased expression of genes such as Sgk1, Mapk8 and Mapk9 (Figure 4f and g).

Figure 4.

Expression of selective chemokines genes verified in psoriasis-related Gene Expression Omnibus dataset and transcriptomic changes in lesional skin of ND/IMQ and HSD/IMQ mice. (a, b) Gene expression levels of CCL20 and SGK1 in skin from healthy individuals, nonlesional skin from patients with psoriasis (cases unaffected), and lesional skin from patients with psoriasis (cases affected) in the GSE13355 dataset. (c) Correlation analysis between SGK1 and selected psoriasis-associated genes using the GSE13355 database. (d) Correlation analysis shows a positive correlation (r = 0.76, P < .0001) between CCL20 and SGK1 genes in the GSE13355 dataset. (e) Kyoto Encyclopedia of Genes and Genomes terms determined by gene set enrichment analysis of ND/IMQ and HSD/IMQ mice. (f, g) Selected gene sets and enrichment plots of genes expressed in the lesional skin of ND/IMQ and HSD/IMQ mice. The heatmaps show the log₂ fold change relative to the geometric mean fragments per kilobase of transcript per million mapped reads + 0.01. Data are shown as mean ± SEM. P-values are determined by 1-way ANOVA. ∗∗∗∗P < .0001. HSD, high-salt diet; IMQ, imiquimod; ND, normal salt diet.

Discussion

Previous study has demonstrated the role of skin in the pathophysiology of salt-sensitive hypertension, in which the accumulation of skin interstitial sodium was caused by excessive sodium intake (Machnik et al, 2009). Furthermore, it has been suggested that sodium accumulated in the skin of patients with psoriasis is positively correlated with disease severity (Maifeld et al, 2022). Psoriasis is mainly driven by proinflammatory cytokines IL-23 and IL-17 (Harden et al, 2015) and involves interaction between the innate and adaptive immune systems (Perera et al, 2012). KCs, influenced by the environment signals, such as microbial ligands or physicochemical substances, interact with immune cells to regulate the function of skin barrier (Abdallah et al, 2017; Bosko, 2019). Our studies showed that IMQ-induced mice fed with HSD could develop a more severe psoriasis-like phenotype with further activation of the CCL20‒CCR6 axis. In vitro, high sodium upregulated the expression of Ccl20 through p38–SGK1 pathway in KCs.

To investigate the effect of HSD on psoriasis-like skin inflammation, we fed IMQ-treated mice with HSD. We observed that HSD/IMQ mice developed more severe skin inflammation, characterized by upregulated expression of Ccl20 and increased infiltration of CCR6⁺ T lymphocytes. CCL20, mainly secreted by KCs, is a key chemokine mediating the migration of CCR6⁺ T cells in psoriasis (Homey et al, 2000; Kim et al, 2014). Importantly, dermal CCR6⁺ γδT17 cells are recruited to psoriasis-like skin lesions through CCL20 (Cai et al, 2013, 2011). We observed significantly higher expression of Ccl20 and frequency of CCR6⁺ γδT cells in the skin of HSD/IMQ mice than in that of ND/IMQ mice. Although HSD/control mice showed no obvious skin lesions, the mRNA expression of Ccl20 in the skin tended to increase, accompanied with significantly increased infiltration of CCR6⁺ γδT cells, indicating that high sodium intake can activate the CCL20–CCR6 axis in the skin of unimmunized mice.

Previous studies demonstrated that CCL20 production in KCs (Li et al, 2017) and macrophages (Heo et al, 2020) is regulated by p38 signaling pathway. Consistent with our data in vivo, elevated expression of Sgk1 and Ccl20 mRNA and activation of p38 and JNK were observed in the KCs cultured with additional NaCl. Moreover, NaCl-induced upregulation of Sgk1 and Ccl20 expression was abrogated by p38/JNK inhibitor. In this study, only the SGK1/p38 inhibitor was used for preliminary validation. GSK650394, an inhibitor of SGK, shows significantly higher selectivity for SGK than protein kinase B and other related kinases (Sherk et al, 2008). However, GSK650394 was found to have comparable potency of targeting MNK1, AMPK, PHK, CDK2, GCK, and CAMKKβ (Maestro et al, 2020). Thus, more molecular experiments are demanded to explore the role of SGK1 in regulating CCL20 in KCs. In different scenarios, SGK1 is reported as an active or negative regulator of inflammation. In adaptive immune system, salt-driven SGK1 signaling pathway promotes the generation of pathologic Th17 cells in autoimmune animal model (Wu et al, 2018, 2013) and upregulates T helper 2 proinflammatory cytokines (Matthias et al, 2019). Conversely, in macrophages, SGK1 negatively modulates inflammation and has inhibitory effect in psoriasis-like inflammation (Meng et al, 2023). Given these different roles of SGK1 in different cell types, the impact of inhibitors targeting SGK1 on regulating skin inflammation in patients with psoriasis warrants further investigation. Although our findings highlight the essential role of KCs in propagating psoriasis pathogenesis under HSD conditions, additional immune cells, such as Th17 cells, are likely involved.

Furthermore, we analyzed skin mRNA profiles of patients with psoriasis from Gene Expression Omnibus dataset. Genes, including SGK1, CCL20, CXCL1, CXCL2, CXCL8, ICAM1, IFNG, IL17a, IL1b, IL22, IL23a, IL26, IL6, and TNF, were positively correlated with SGK1 expression in lesional skin, with strongest correlation between SGK1 and CCL20 among them. However, dietary patterns were not documented in the Gene Expression Omnibus dataset, making it unclear whether sodium directly upregulates SGK1 expression in human skin. To address this, we analyzed transcriptome of mice fed on HSD, demonstrating that HSD elicits Sgk1 upregulation in our animal model. The data indicate that high sodium potentiates Ccl20 mRNA expression in the murine KCs in an SGK1-dependent manner and, therefore, exacerbates psoriasis-like skin inflammation.

In conclusion, we mainly demonstrate that CCL20–CCR6 axis is further activated in lesional skin of HSD/IMQ mice. Mechanistically, elevated NaCl concentrations upregulated Ccl20 expression through the JNK/p38–SGK signaling pathway in KCs. There are also limitations of this study; we used IMQ-induced psoriasis-like mouse model, which only mimic acute inflammation of plaque psoriasis (Hawkes et al, 2017). The influence of HSD on other types of psoriasis requires clarification in the future. Overall, our data underscore the need to access causality of HSD and incidence of psoriasis through clinical trials. Clinically, restricting dietary salt intake should be required to modulate skin inflammation in patients with psoriasis.

Materials and Methods

Experimental animals

C57BL/6J wild-type mice (aged 6–8 weeks) were purchased from Zhejiang Charles River. Neonatal C57BL/6J mice were used to obtain primary KCs. Mice in these experiments were maintained under specific pathogen-free conditions and then randomly assigned to each group before each experiment. The animal study was reviewed and approved by Shanghai Jiao Tong University School of Medicine Animal care and Use Program (JUMC2023-195-B).

IMQ-induced psoriasis-like mouse model

Before the modeling, C57BL/6J wild-type mice were housed in a specific pathogen-free environment for few days to adapt to the experimental environment. The HSD with 8% NaCl was chosen, according to previous literature (Pajtók et al, 2021), in which an HSD with 8% NaCl was fed to IMQ-induced psoriasis-like mouse model. Normal diet or HSD was fed to the mice for 60 days before IMQ cream (5%, Aldara, 3M Pharmaceuticals) application. The hair on the dorsal skin (2 × 2 cm²) was shaved off, and depilatory cream was applied to remove the remaining hair, and then 31.25 mg of IMQ was applied to the dorsal skin daily for a total of 3 days. We used PASI to score the severity of psoriasis-like skin lesion. Skin lesions are graded on the basis of 3 criteria, including redness, thickness, and scaliness. Scores range from 0 to 4 (0 for clear; scores 1–4 for increasing severity). Then, the total PASI scores range from 0 to 12, with higher scores indicating more severe disease. The body weight was recorded in every 10 days (Figure 5). The mice were killed, and samples were collected 1 day after the last IMQ application.

Figure 5.

Effect of 8% HSD on body weight. During the period of ND or HSD feeding time, the body weight was measured every 10 days for 60 days. Data are shown as mean ± SEM (n = 3 for each group). P-values were determined by repeated-measures 2-way ANOVA with interaction. HSD, high-salt diet; IMQ, imiquimod; ND, normal salt diet.

Primary murine KC culture

After isolated from newborn pups, skin was incubated in 4 mg/ml Dispase II (Roche) for 8 hours at 4 °C, and then the epidermis was separated from the dermis, subsequently digested with 0.05% trypsin for 20 minutes at room temperature, and filtered through 100-mm strainer. Isolated KCs were cultured in 6-well plates with K-SFM medium (Life Technologies) at 37 °C and 5% carbon dioxide. As previously described (Machnik et al, 2009), to mimic the NaCl accumulation in the skin caused by HSD, culture medium with additional 40 mM NaCl concentration was chosen. Briefly, for experiment using the specific inhibitors, the cultured KCs were starved overnight in growth supplements-free K-SFM medium prior to treatment with inhibitors. SP600125 (Cell Signaling Technology, catalog number 8177S) as JNK inhibitor, SB203580 (Cell Signaling Technology, catalog number 5633S) as p38 inhibitor, or GSK650394 (MedChemExpress, catalog number HY 15192) as SGK1 inhibitor at concentrations of 5 μM each were added to the cultures, followed by stimulation with additional 40 mM NaCl for 24 hours. KCs were harvested for RT-PCR analysis. To perform western blotting analysis, we cultured KCs with or without additional 40 mM NaCl in K-SFM medium for 0.5 hours, 1 hour before collecting for western blotting analysis.

Histological assessment

Mice were killed on day 4, and then skin tissues were collected. Dorsal skins were removed, and a portion of the skin was fixed in 4% paraformaldehyde and embedded in paraffin. H&E staining was conducted on the skin sections. Pathologic sections were scanned and analyzed using NDP.view2 (Hamamatsu). The epidermis thickness was defined as the vertical distance from the uppermost layer of the stratum corneum to the basal layer and recorded as shown in Supplementary Table S1. For detail, 4 measurements at randomly selected points were taken per nonoverlapping field, across 3 fields on each section (yielding 12 measurements per section for each sample). For immunohistochemistry staining, sections of paraffin-sliced skin tissue were subjected to immunohistochemical staining for Gr1. In brief, the sections were incubated with anti-Ly6G antibody (1:1000, Abcam, catalog number ab238132) for 1 hour at room temperature after fixation and blocking and then incubated at 4 °C overnight. After washing, the sections were incubated with secondary antibody (Abcam, catalog number ab6721) at 37 °C for 1 hour. The 3,3'-diaminobenzidine staining was conducted to visualize the reaction. To measure Ly6G protein expression in immunohistochemistry, ImageJ analysis software was used, and quantitative results were reported as the percentage of the Ly6G stained cells relative to the total cell nucleuses.

Flow cytometry

After divided into pieces and digestion, the skin tissues were passed through cell strainer (40 μm) to obtain single-cell suspension. To perform intracellular cytokine staining, cells were stimulated at 37 °C and 5% carbon dioxide for 4 hours with Cell Activation Cocktail (BioLegend, catalog number 423304). After blocking with anti-mouse CD16/32 antibody (BioLegend, catalog number 101302), surface staining was performed by adding a mixture of the antibodies Fixable Viability Dye eFluor 780 (eBioscience, catalog number 65-0865-14), AF700 antimouse CD45.2 antibody (BioLegend, catalog number 109822), PerCP/Cy5.5 antimouse CD3 antibody (BioLegend, catalog number 100218), and allophycocyanin antimouse TCRγ/δ antibody (BioLegend, catalog number 118119) and then incubated at 4 °C for 20 minutes in the dark. For CCR6 staining, the phycoerythrin antimouse CCR6 antibody (BioLegend, catalog number 129804) was separately added and incubated at 37 °C for 20 minutes in the dark. After fixing and permeabilizing, intracellular staining was performed by adding intracellular antibody phycoerythrin/cyanine7 antimouse IL-17A antibody (BioLegend, catalog number 506922) and then was incubated at 4 °C in the dark for 1 hour. For detection of neutrophils, cell suspension was stained with mixture of antibodies containing Fixable Viability Dye eFluor 780 (eBioscience, catalog number 65-0865-14), FITC antimouse CD45 antibody (BioLegend, catalog number 103107), phycoerythrin antimouse Gr-1 antibody (BioLegend, catalog number 108407), and allophycocyanin antimouse/human CD11b antibody (BioLegend, catalog number 101211). Flow cytometry was performed within 24 hours after staining. Samples were harvested with BD LSRFortessa (Becton Dickinson, San Jose, CA) and analyzed with FlowJo software (Tree Star). Gating strategies are showed in Figure 6.

Figure 6.

Gating strategy. (a) Gating strategy for CCR6⁺γδ T cells (live⁺, CD45⁺, CD3⁺, γδTCR^int, CCR6⁺), IL-17–producing γδ T cells (live⁺, CD45⁺, CD3⁺, γδTCR^int, IL-17A⁺), and IL-17–producing T cells (live⁺, CD45⁺, CD3⁺, IL-17A⁺). Dermal γδ T cells typically have intermediate γδTCR expression than epidermal γδ T cells and usually fall into the γδTCR versus SSC defined gate (red). Dermal IL-17–producing γδ T cells are derived from red gate. Several T cell populations fall into the gate for murine skin CD3⁺ T cells (green), including epidermal γδ T cells (high CD3 expression) and dermal γδ T cells (intermediate CD3 expression). IL-17–producing CD3⁺ T cells are derived from green gate. (b) Gating strategy for neutrophils (live⁺, CD45⁺, CD11b⁺, Gr-1⁺). SSC, side scatter.

Western blot analysis

The KCs were fully lysed in RIPA buffer containing protease inhibitors and phosphatases on ice. After quantification, protein samples were separated through 10% SDS-PAGE and then transferred to membranes. After being blocked, the proteins were incubated sequentially at room temperature for 1 hour with a primary antibody and a secondary antibody and then finally visualized with enhanced chemiluminescence solution. JNK Rabbit mAb (Cell Signaling Technology, catalog number 5292), phosphorylated JNK Rabbit mAb (Cell Signaling Technology, catalog number 4668), p38 Rabbit mAb (Cell Signaling Technology, catalog number 8690), phosphorylated p38 Rabbit mAb (Cell Signaling Technology, catalog number 4511), mouse anti-Alpha Tubulin mAb (Proteintech, catalog number HRP-66031) were used as primary antibodies, and goat antirabbit antibody (Abcam, catalog number ab6721) was used as secondary antibody.

RT-qPCR

Total RNA isolation System (Promega, catalog number Z31001) and TRIzol (Thermo Fisher Scientific) were used for tissue and cell RNA extraction. After RNA was converted to cDNA using a reverse transcription kit (Takara, catalog number RR036A), SYBR (Qiagen, catalog number 208052) was used to perform RT-qPCR on the Quantstudio Q6 system (Thermo Fisher Scientific, Applied Biosystems). Primer sequences are provided in Table 1, but the mouse IL-17A primer (Qiagen, catalog number 249900) was purchased. All gene expression levels were normalized to the mouse endogenous housekeeping gene b-2 microglobulin and quantified using the 2−ΔΔCt method.

Table 1.

List of Primer Sequences Used for RT-qPCR Analysis

Gene	Primer Sequence (5′–3′)
S100a8	Fw: AGTGTCCTCAGTTTGTGCAG
	Re: ACTCCTTGTGGCTGTCTTTG
S100a9	Fw: ATACTCTAGGAAGGAAGGACACC
	Re: TCCATGATGTCATTTATGAGGGC
Mcp1	Fw: TTAAAAACCTGGATCGGAACCAA
	Re: GCATTAGCTTCAGATTTACGGGT
Cxcl1	Fw: GAAAGCTTGCCTCAATCCTG
	Re: CTTCCTCCTCCCTTCTGGTC
Il23	Fw: TATCCAGTGTGAAGATG GTTGTG
	Re: CACTAAGGGCTCAGTCAGAGTTG
Sgk1	Fw: GAGCCGGAGCTTATGAACG
	Re: AGTGAAAGTCGGAGGGTTTGG
Ccl20	Fw: CGACTGTTGCCTCTCGTACA
	Re: GAGGAGGTTCACAGCCCTTT

Abbreviations: Fw, forward; Re, reverse.

Gene Expression Omnibus database analysis

The correlation analysis was performed on the Gene Expression Omnibus database (GSE13355). First, the Pearson correlation coefficient (r) between variables was determined through the Corrplot package. To assess the linear correlation between the variables, the Corrplot package was subsequently used to generate a correlation heatmap, which was used to further visualize the strength of correlations between the numbers of variables.

Construction of RNA-sequencing library

RNA was extracted and purified from the skin sample. The total RNA concentration, integrity (28S/18S), and purity (OD260/OD280, OD260/OD230) was assessed through Frament Analyse. RNA Quality Number was also calculated. The library was prepared using Optimal Dual-mode mRNA Library Prep Kit (BGI Genomics) and amplified with rolling circle amplification to make DNA nanoball. The DNA nanoballs sequenced on G400/T7/T10 platform (BGI Genomics) to generate PE 100/150 base reads.

Statistical analysis of gene expression

Low-quality adapter sequences were removed from raw data using SOAPnuke (version 1.5.6) (Li et al, 2008). Clean data were aligned to the mm10 mouse reference genome using HISAT (version 2.1.0) (Kim et al, 2015) and were used to calculate the expression level of genes. Then, statistically significant differentially expressed genes were defined when log₂ fold change ≥1 and q-value ≤0.05 using DESeq2 (version1.4.5) (Love et al, 2014).

Gene set enrichment analysis

Gene set enrichment analysis was performed using BGI omics Explorer Platform, provided by the BGI Genomics (https://biosys.bgi.com). The differentially expressed genes were functionally classified, according to the Kyoto Encyclopedia of Genes and Genomes annotation results and classifications. To determine Kyoto Encyclopedia of Genes and Genomes enrichment, we used Gene Set Enrichment Analysis (version 4.1.0) and analyzed with the Molecular Signatures Database (version 7.5.1). Threshold was set with q-value ≤0.05.

Statistical analysis

Statistical analysis was performed using GraphPad Prism 9.0 software (GraphPad Software, San Diego, CA). Statistical significance between 2 groups was evaluated by Student’s t-test, and statistical significance among ≥3 groups was evaluated by 1-way ANOVA. For most animal experiments, 2-way ANOVA with interaction was used. Repeated-measures 2-way ANOVA with interaction was only for PASI and body weight analysis. For each sample, all the measurements of epidermal thickness in 1 section represent technical replicates. To confirm differences of epidermis thickness between groups, we constructed a linear mixed-effects model using lmer package in R. The sample size (n = 3 for each group) was chosen in our animal experiment on the basis of the previous literature using n = 3 for each group in the IMQ-induced psoriasis mouse model (Cai et al, 2019a, 2019b). Missing data were not imputed. All analysis included data derived from valid experimental samples. The level of statistical significance was set to P < .05; ∗P < .05, ∗∗P < .01, ∗∗∗P < .001, and ∗∗∗∗P < .0001. Data were expressed as mean ± SEM.

Ethics Statement

The animal study was reviewed and approved by Shanghai Jiao Tong University School of Medicine Animal care and Use Program (JUMC2023-195-B).

Data Availability Statement

RNA-sequencing data have been deposited into National Center for Biotechnology Information Sequence Read Archive databases (https://www.ncbi.nlm.nih.gov/sra/PRJNA1394367). Other data of this study are included in this article and its supplementary material.

ORCIDs

Mengyan Hu: http://orcid.org/0000-0001-6217-5146

Han Cao: http://orcid.org/0000-0001-5190-7014

Li Zhang: http://orcid.org/0000-0002-2714-5045

Yuhan Xia: http://orcid.org/0000-0003-1862-6135

Jiayi Zhang: http://orcid.org/0000-0002-1079-4188

Feng Xue: http://orcid.org/0000-0003-2630-8460

Xia Li: http://orcid.org/0000-0002-3977-800X

Jie Zheng: http://orcid.org/0000-0002-7961-6427

Conflict of Interest

The authors state no conflict of interest.

accepted manuscript published online XXX; corrected proof published online XXX

Supplementary Material

Supplementary Table S1

Measurements of Epidermal Thickness