Topical emollient prevents the development of atopic dermatitis and atopic march in mice

To investigate the effect of emollient on atopic march in a murine model of atopic dermatitis (AD). Following induction of AD with topical calcipotriol (MC903) and ovalbumin (OVA), one group of mice was treated topically with a linoleic acid‐ceramide‐containing emollient, while mice without emollient treatment served as disease controls. After 28 days, clinical, histological and transcriptomic analyses were performed in the skin lesions and the lung as well as serum cytokine levels. Treatments of mice with MC903 and OVA induced a typical phenotype of AD, accompanied by increased expression levels of Th2 and basophil‐related genes in the lung. Topical emollients markedly decreased the severity of skin lesions and inflammatory cell infiltration. Moreover, emollient treatments significantly downregulated expression levels of AD‐related genes (286 of 1450 differentially expressed genes), including those related to innate inflammation (S100a8/a9, Il1b, Defb3/6, Mmp12), chemokines (Cxcl1/3, Ccl3/4) and epidermal permeability barrier (Krt2/6b/80, Serpinb12, Lce3e, Sprr2), etc. Downregulated genes were enriched in mitochondrial OXPHOS‐related pathways, while upregulated genes were mainly enriched in axon guidance and tight junctions. Moreover, topical emollient treatments decreased total serum levels of IL‐4, along with substantial reductions in IgE and thymic stromal lymphopoietin (TSLP) levels. Furthermore, 187 of 275 upregulated genes in lung tissue were also significantly downregulated, including those involved in leucocyte chemotaxis (Ccl9, Ccr2, Retnlg, Ccl3, Cxcl10, Il1r2, etc.) and basophil activation (Mcpt8, Cd200r3, Fcer1a, Ms4a2). In conclusion, topical emollient not only reduces skin inflammation, but also mitigates systemic inflammation by decreasing TSLP and IgE levels. Moreover, topical emollient reduces chemokine production and basophil infiltration and activation in the lung.


| INTRODUC TI ON
The prevalence of allergic diseases, including atopic dermatitis (AD), food allergy, allergic asthma (AA) and allergic rhinitis (AR), has increased significantly worldwide in recent decades, affecting 20%-30% of the world's population. 1,2 These diseases usually develop in a certain way known as atopic march, starting with AD in early childhood and progress to AA and AR in children. 3 Several prospective cohort studies have shown that early AD increases the risk of developing AA (OR 2.06, 95% CI 1.09-3.90) or AR (OR 4.32, 95% CI 2.04-9.12) later in life. 4 as basophils 8 and innate lymphoid cells, 9 which then activate dendritic cells or Langerhans cells by producing type 2 cytokines. Activated dendritic cells or Langerhans cells take up antigen and migrate to the draining lymph nodes to promote Th2 differentiation. When the sensitized T cells are re-exposed to the allergen through the respiratory and digestive tracts, these T cells secrete large amounts of type 2 cytokines, among which IL-4/13 can disrupt the tight junction of epithelium, promote the antibody class switching to IgE and thus activate mast cells, increase mucin secretion and vascular permeability, while IL-5 induces the chemotaxis of eosinophils. 10 Taken together, a variety of cells and cytokines are involved in the atopic march and that epithelial damage plays a key role in this process.
Emollients are widely used in the basic treatment of AD patients because of their ability to repair the skin barrier and lower the risk of disease recurrence. But whether topical emollients can prevent the atopic march is inconclusive. An Australian clinical randomized controlled trial (RCT) trial showed that using emollient on a daily basis in newborns aged 0-6 months reduced the incidence of AD and food allergy at age 1. 11 However, a recent RCT in the UK did not find a reduction in the risk of eczema or food allergy at age 2 after the use of emollient in the first year of life. 12 The aim of the present study was to assess the effect of emollient on AD and later atopic march in a mouse model of AD.

| Induction of AD in mice
Specific pathogen free (SPF) grade 6-week-old female C57BL/6 mice were provided by Beijing Vitalriver Laboratory Animal Technology Co. The study protocol was approved by the  Table S1.
AD mice without emollient treatment served as controls, and mice without any treatment were used as normal controls.

| Scoring skin lesions
On Day 28, the gross appearance was scored according to a 0-to 3point scale (0, none; 1, mild; 2, moderate; 3, severe). And the scores of erythema, desquamation, oedema and scratch marks were totalled, with maximal score of 12.

| RNA sequencing library construction and analysis
Mouse tissue RNA was extracted using the Eastep Super Total RNA Extraction Kit (Promega, Shanghai), and sequencing libraries were synthesized using the SMARTER mRNA-Seq Library Prep Kit following manufacturer's protocols. The samples were pair-end sequenced using the Illumina sequencing platform to obtain raw data in FASTQ format. Splice removal and quality assessment were performed using FastQC (v0.10.1) software. The raw sequences obtained from sequencing were aligned to the mouse reference genome GRCm38 using Hisat2 (v2.0.1). Sequenced samples with a read alignment rate >80% were used in subsequent analysis. Transcripts were quantified and expression matrices were constructed using the HTSeq software (v0.6.1). Analysis of differentially expressed genes (DEGs) was performed using DESeq2 package (v1.30.1). Genes were selected using the BH multiple testing correction with a corrected p-value <0.05 and log 2 fold change ≥1. Heatmaps were performed using the ComplexHeatmap package (v2.6.2) for the top 100 absolute values of log 2 fold change up-and downregulated DEGs under different conditions and visualized by ggvenn (v0.1.9) package.

| KEGG enrichment analysis
The KEGG enrichment analysis was performed using the ClusterProfiler package (v3.18.1) on the DEGs screened between groups and tested based on the hypergeometric test. The p-values were corrected by the BH multiple testing correction. Pathways with q-values <0.2 were defined as significantly enriched pathways. Enrichment results were visualized using the ggplot2 (v3.3.5) package.

| Cytokine assays
Serum TSLP, IL-4 and IgE levels were measured with respective mouse ELISA kit (Thermo Fisher Scientific) according to the manufacturer's instructions.

| Real-time fluorescent quantitative PCR (RT-qPCR)
Total tissue RNA was extracted using the Total RNA Extraction Kit (Promega, Shanghai), and cDNA was obtained by reverse transcription, followed by RT-PCR using SYBER Green as a fluorescent marker to detect the expression of Il-4, Tslp and Mcpt8 in the tissues. β-actin was used as the internal reference. The results were expressed as 2 −dCt . Primers were synthesized by Shanghai Genewiz Co. The primer sequences are listed in Table S2.

| Statistical analysis
Data were analysed using Graph Prism v8 and represented as mean ± standard error of the mean (SEM). Comparisons between multiple groups were analysed using one-way analysis of variance (ANOVA) tests, followed by Dunnett's multiple comparisons.

| Topical emollient reduces OVA+MC903induced AD-like changes
As expected, repeated applications of OVA+MC903 to the skin induced AD-like skin lesions, including erythema, desquamation, oedema and scratch marks in mice ( Figure 1A). The severity of lesions was significantly decreased following topical applications of the emollient compared to the AD mice ( Figure 1A). The histology of AD lesions was characterized by increased skin thickness and inflammatory cell infiltration. Topical emollient largely restored the histological alterations in the skin, evidenced by reductions in epidermal thickening and inflammatory cell infiltration ( Figure 1B), particularly mast cells ( Figure 1C) and basophils ( Figure 1D) in the dermis. Moreover, expression levels of mRNA for Th2-related cytokines such as IL-4, IL-5 and IL-13 were increased in the skin and to a less extent in the lungs of AD mice although the increases were not significant except IL-13 in the lungs (p < 0.05) (Figures S2 and S3). Apparently, topical emollient tends to lower expression levels of mRNA for Th2 cytokines, IFNγ and TSLP in the skin of AD mice, while these cytokines were not altered by emollient in the lungs, which may be due to small range of change and insufficient sample size ( Figure S3).

| Topical emollient downregulates genes associated with mitochondrial oxidative phosphorylation and upregulates genes related to tight junctions
RNA sequencing of the skin showed a total of 2705 DEGs between AD and controls, including 1255 downregulated and 1450 upregulated genes (Figure 2A,B). The upregulated genes included innate immunity and inflammation-related genes such as S100a8/ a9, Defb6, Il1b, Cxcl2, Ccl3, Mmp12, epithelial alarmin gene Il1rl1 and Sema4a ( Figure 2C; Table S3), which may be involved in epidermal nerve fibre elongation and regulation of immune cells. These results demonstrate that topical emollient significantly attenuates alterations in epidermal gene expression in AD mice, particularly in those related to oxidative stress and permeability homeostasis.

| Topical emollient reduces serum TSLP and total IgE levels in AD mice
To determine whether emollient affects systemic inflammation in AD mice, we measured the levels of cytokines and IgE in the serum.

| Correlation analysis between DEGs in the skin and the lung as well as biomarkers and clinical scores
As shown in Figure 5, genes in the skin that are associated in epidermal barrier, protease inhibitors (e.g. Serpina/b, Stfa, Sprr2, Krt family), oxidative phosphorylation, type II immunity and leucocyte migration were positively correlated with each other. Allergic, inflammatory, basophil genes in the lung (e.g . Stfa2l1, Fcer1a, Cstdc4, Ms4a2, Stfa2) were positively correlated with genes related to epidermal barrier, protease inhibitors, oxidative phosphorylation, type II immunity, serum type II cytokines and clinical scores. The highest correlations were found between allergy/inflammation genes in the lung and leukocyte chemotaxis/innate immunity in the skin, suggesting an important role of the cutaneous function in the development of atopic march.

| DISCUSS ION
AD can be a precedent of the atopic march. Early-onset AD is usually more severe and persistent, representing a risk factor for asthma development. 4,14 In addition, patients with AD are more likely to develop food sensitization than healthy controls with an OR (odd ratio) up to 6 at 3 months of age. 15 Thus, early intervention of AD can be critical to prevent infants from developing lifelong chronic atopic diseases and reduce disease burden. Although biological agents and other immune suppressors can effectively alleviate AD, 16 can induce Th2-type responses and is involved in a variety of allergic diseases. Overexpression of epithelial-derived TSLP can induce or exacerbate AD or asthma-like phenotypes in mice. 24,25 In contrast, knockdown of TSLP blocks the development of allergic disease. 26 Moreover, TSLP can stimulate basophils to rapidly infiltrate inflammatory sites and collaborate with dendritic cells to promote Th2 differentiation. 8 IL-3 can stimulate basophil F I G U R E 5 Correlation analysis of emollient-regulated genes in the skin and lung. Genes' names are listed on the left with 's' indicating genes in the skin and 'l' indicating genes in the lung. degranulation and histamine production in response to FCER1 cross-linking signals. 27 Thus, it is likely that TSLP induces basophil proliferation, activation and infiltration into the lung, causing microinflammation. When the allergen is encountered again, FCER1 cross-linking induces basophil activation to produce IL-4 and interacts with innate lymphoid cells and dendritic cells to cause Th2 drift in the lung environment, triggering the development of the atopic march. Additionally, multiple sensitization and high IgE levels are strongly linked to the development of AA and AR from AD. 7,28 Hence, upregulation of Sema4A and downregulation of TSLP and total IgE are another mechanism by which topical emollient benefits AD and atopic march.
Finally, although basophils make up <1% of peripheral blood leukocytes, they can produce large amounts of Th2-type cytokines (including IL-4) early in the development of allergic disease. 29,30 In an

ACK N O WLE D G E M ENTS
None.

CO N FLI C T O F I NTE R E S T S TATE M E NT
The authors declare no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are openly available