GE1111 was synthesised from O2h laboratory, Ahmedabad India. CST-14 was custom synthesised from GenScript. 2,4-dinitrofluorobenzene (DNFB) was purchased from Shanghai Macklin Biochemical (No. M24218031; Shanghai, China). All other chemicals used in this study were of commercial grade. GE1111 was dissolved in dimethyl sulfoxide (DMSO, Sigma, St. Louis, MO, USA), then diluted to desired concentrations in buffer, saline or media.

The Laboratory of Allergic Disease 2 (LAD-2) human mast were cultured in StemPro-34 medium supplemented with 10 mL/L StemPro nutritional supplements, penicillin-streptomycin (1:100), 2 mmol/L glutamine and 100 ng/mL human stem cell factor and incubated at 37°C in 5% CO₂ incubator. Hemi-depletions of media were performed weekly and cell proliferation was determined through weekly cell number measurements. The immortalized human keratinocyte HaCaT cell line were routinely maintained in Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% Fetal Bovine Serum (FBS), and penicillin and streptomycin (1:100). The RAW 264.7 macrophage cell line were routinely maintained in DMEM with 10% FBS and 1 % penicillin-streptomycin.

We developed in vitro cell culture models to mimic the in vivo immune cell environment by using MC, keratinocytes, and macrophage cell lines. To investigate the effect of MRGPRX2 antagonists on mast cell and keratinocyte interaction we used a potent MRGPRX2 agonist, the neuropeptide CST-14 and GE1111 to manipulate MRGPRX2-MC. This was followed by treating HaCaT keratinocytes with MC supernatant (from different experimental groups). HaCaT cells were seeded at a density of 3 x 10⁵ cells/well in a six well plate overnight at 37°C in 5% CO₂ incubator. 1–2 hour prior to experimentation, LAD-2 MC were seeded at a density of 1.0 x 10⁶ cells/well. LAD-2 MC were then treated with 0.1% DMSO or 50µM GE1111 for 30 minutes, followed by stimulation with 5.81 µM CST-14 and 2 hours incubation at 37°C in 5% CO₂ incubator. LAD-2 MC and their supernatant were then collected into microcentrifuge tubes and centrifuged at 500g for 5 minutes to isolate the supernatant and transferred onto the keratinocytes and macrophage cells. LAD-2 MC were, at this point, harvested for protein or gene expression analysis. For western blotting and real-time PCR experiments, we transferred 800 µL of MC supernatant or media to the keratinocytes. This supernatant was incubated with the keratinocytes for 2 hours; keratinocytes were then harvested for gene or protein analysis. For the macrophage phagocytosis assay, the 150 µL of LAD-2 mast cell supernatant was instead transferred onto the macrophages and incubated for 14–16 hours.

LAD-2 MC and HaCaT keratinocytes were treated according to the method described above. Cells were washed with phosphate-buffered saline (PBS) to remove any debris or media residue. Subsequently, the cells were fixed using a fixative solution, 4% paraformaldehyde in PBS, for 10–15 minutes at room temperature. After washing with PBS three times, cells were permeabilized with 0.1 % Triton X-100 for 5–10 minutes. Following permeabilization, the cells were blocked with BSA for one hour at room temperature followed by incubation with primary antibodies specific to claudin-1 and TSLP (1:200) for 1–2 hours at room temperature. The cells were subsequently incubated with fluorescently labelled secondary antibodies such as Alexa Flu 594 donkey anti Rabbit IgG (red for TSLP). and Alexa Flu 488 donkey anti Rabbit IgG (green for claudin 1) Finally, the stained cells were mounted onto glass slides using an appropriate mounting medium containing DAPI. The immunofluorescence staining was visualised using a Fluorescent Microscope, Nikon DS-Ri2 camera. Images were captured at suitable magnifications to observe the cellular localisation and expression levels of TSLP and claudin-1 within the HaCaT keratinocytes.

The phagocytosis assay was performed to evaluate the phagocytic activity of RAW264.7 macrophage cells in response to the treatment of different groups of MC supernatant. FluoSpheres™ Carboxylate-Modified Microspheres yellow beads (Catalog number: F8823), and CellTracker™ Orange CMTMR Dye (Catalog number: C2927) were used to stain the cells and visualise the phagocytosed particles. The macrophage cells (1 x 10⁵ cells/ml (400 µL/well) were seeded onto ibidi 8 Well Chamber, removable slides (Cat.No:80841). MC were treated, and the supernatant was isolated as described above. The media was removed, and RAW264.7 macrophage cells were incubated with the mast cell supernatant (150 µL) for 14–16 hours. After the co-incubation period, FluoSpheres™ Carboxylate-Modified Microspheres were added to the cell culture medium at a 10:1 ratio (beads:cell). The beads were opsonized with 10% FBS to facilitate their recognition and engulfment by the macrophages for 1 h at 37°C. Following the incubation period, 100 µL of beads were added to the cells and cells were incubated for 90 minutes at 37°C. Cells were then carefully washed with PBS to remove any unbound beads. To visualise the macrophages and the phagocytosed particles, the cells were stained with CellTracker™ Orange CMTMR Dye. 100 µL of 5µM dye was added to the cell culture medium at an appropriate concentration and incubated for 15–30 minutes, allowing it to label the cell membranes and cytoplasm of the macrophages. The phagocytic activity of the RAW264.7 macrophage cells was visualised and quantified using The ECLIPSE Ti2 inverted microscope. The number of phagocytosed beads per macrophage and the percentage of phagocytic cells were determined by analysing the acquired images using Image J analysis software.

Western immunoblotting was used to separate, identify, and measure proteins of interest in LAD-2 MC, HaCaT keratinocytes, and mouse skin samples. In-vitro samples were lysed using RIPA cell lysis and extraction buffer (with phosphatase and protease inhibitor cocktail (Thermo-Fisher) and cell scrapers, then collected into microcentrifuge tubes following 30 min incubation on ice. Mouse samples were processed and homogenized through agitation in tubes with both RIPA and the Fisherbrand™ Bead Mill Homogenizer Accessory. These samples were then centrifuged at 13000 RPM for 15 min, and the supernatant was transferred to a new microcentrifuge tube. The protein concentration of these samples was then measured using the Bradford protein assay (Thermo-Fisher). 10%, and 15% acrylamide SDS-PAGE resolving gels (pH 8.8) were hand-casted, with a 4% acrylamide SDS-PAGE stacking gel. 30µg of protein was loaded into each well, then run at 100V for 2 hours in running buffer (gel electrophoresis). A wet transfer was performed at 100V for 2 hours onto a nitrocellulose membrane (Bio-Rad). The membrane was then incubated with blocking buffer [5% Bovine Serum Albumin (BSA) in 1x Tris-Buffered Saline with 0.1% Tween 20 Detergent (TBS-T)] for 2 hours at room temperature. Primary antibodies specific to proteins of interest, including STIM1 (CST-D88E10), TSLP (PA5–89013), claudin-1 (Ab15098), GAPDH (CST-141CO), phospho-AKT (CST-9271S), AKT (CST-9272S), phospho-ERK _½ (CST-4370T), and ERK _½ (CST-4695T) were prepared at dilutions of 1:1000 in 5% BSA and incubated with the membrane for 14–16 hours at 4°C. Goat anti-Rabbit IgG Secondary antibody (CST-7074) was prepared at dilutions of 1:5000–10000 in 5% BSA and incubated with the membrane for 2 hours at room temperature. Immunoreactive proteins were developed using UVITEC Alliance LD (UVITEC Ltd., Cambridge, UK) with SuperSignal Technology (Thermo Fisher Scientific, Inc., Waltham, MA, USA). Band intensities were quantified using and analysed using ImageJ software (37). Representative results from three or more independent experiments are shown.

Treatment of MC and keratinocytes was conducted as described above. RT-qPCR was performed to determine the gene expression of various inflammatory cytokines, chemokines and mediators in LAD-2 human MC, HaCaT keratinocytes and mouse skin samples. RNA extraction was performed using and according to the protocol of the FastPure Cell/Tissue Total RNA Isolation Kit V2 RC112 (Vazyme). Mouse samples were additionally processed and homogenized through agitation in tubes with both Buffer RL (from the FastPure Cell/Tissue Total RNA Isolation Kit) and the Fisherbrand™ Bead Mill Homogenizer Accessory. RNA to cDNA conversion was performed using HiScript II 1st Strand cDNA Synthesis Kit R211 (Vazyme) according to kit protocol. PCR was then performed using ChamQ SYBR Color RT-qPCR Master Mix (High ROX Premixed) Q441 (Vazyme) according to kit protocol. Primers used and their sequences have been included in Supplementary Table 1 .

Wild-type BALB/c adult male mice, 6 to 8 weeks old, were purchased from the Laboratory Animal Unit of the University of Hong Kong (accredited by the Association for Assessment and Accreditation of Laboratory Animal Care International). Mice were housed under a 24-hour light/dark cycle, with food and water ad libitum. The experimental protocols for the mouse experiment were approved by the Committee on the Use of Live Animals in Teaching and Research (CULATR) of the University of Hong Kong. All animal procedures were performed under ketamine/xylazine anesthesia (CULATR approved no 23–230).

DNFB-Induced Mouse Model of AD: Young adult male mice (BALB/c aged 6–8 weeks old, n=6–7/group) were used to develop a DNFB-induced AD on dorsal skin and ears in mice. Mice were randomly divided into 4 groups, vehicle control group, disease control group and GE1111 lower dose 10mg/kg treatment and higher dose treatment groups of 20 mg/kg. Mice were anesthetized and shaved dorsally two days prior to beginning of experimentation, and as required throughout the experimental time course of 4 weeks. The novel MRGPRX2 antagonist GE1111 or vehicle (<0.1 % DMSO in saline) was injected at a dose of 10 and 20 mg/kg body weight via intraperitoneal injection into experimental group. This was followed by painting 50 µL vehicle (acetone and olive oil) or 0.3% DNFB along the shaved dorsal skin and ears. The DFNB and antagonist treatment was performed once weekly, for a total of 4 weeks; the body weight and condition of the mice were monitored throughout the experimental time course. Furthermore, the degree of skin inflammation was assessed and scored by three individuals after 24 hours of the last DFNB and antagonist/vehicle treatment (38). Based on a modified clinical scoring system, the scores ranging from 0 to 3 [0 (none), 1 (mild), 2 (moderate), and 3 (severe)] were assigned to different aspects of inflammation, including thickening, scaling and erythema (38). Then, mice were anesthetized via intraperitoneal injection of ketamine and xylazine, and blood, skin and ear tissue were collected for downstream analysis by ELISA, histopathology western blot, and RT-qPCR. Both skin and ear samples were snap frozen in liquid nitrogen for later protein and gene expression analysis. Skin and ear samples for histology were stored in 10% formalin prior to processing for histological analysis.

Blood monocyte chemoattractant protein 1 (MCP-1) levels of experimental mice were assessed using the MCP-1 ELISA kit from Thermo Fischer Scientific (Catalog number BMS6005), following the manufacturer's instructions (the serum samples were diluted four times).

Mouse skin and ear samples harvested from mice were dehydrated in graded concentrations of ethanol (30%-95%), then embedded within paraffin blocks. Skin and ear samples were then sectioned at 5µm stained with H&E according to our previous method (36). For toluidine blue staining, slides were similarly de-paraffinized and hydrated, then stained with toluidine blue according to previous method (36). Degranulated and non-degranulated MC in the 20X magnified ear sections of mice from each experimental group were counted and analysed. Degranulated MC were distinct from non-degranulated MC by reduced toluidine blue staining intensity and dispersed visible cytoplasmic granules. Images were captured under 10X and 20X magnification using a Nikon Eclipse Ni-U, upright microscope, and Nikon DS-Ri2 microscopic camera (Nikon, Tokyo, Japan).

The paraffin sections were used for analysing Involucrin and periostin expression in the skin tissue by immunohistochemistry. We used DAB Substrate Kit, Peroxidase (HRP), with Nickel, (3,3'-diaminobenzidine) (SK-4100). Briefly, the IHC staining protocol involved deparaffinization, rehydration, blocking of endogenous peroxidase activity, and incubation with primary antibodies against Involucrin and periostin. After washing, the sections were incubated with HRP linked secondary antibodies. The sections were then counterstained by H&E, dehydrated, and mounted. The quantitative analysis of the stained sections (% area) was done by using ImageJ (37).

All comparisons were performed by GraphPad Prism (GraphPad Software, La Jolla, Calif) or Microsoft Excel (Microsoft, Redmond, Wash) software. The graphs were plotted by GraphPad Prism 9.3.1; data are presented as means ± SEMs of at least 3 independent experiments unless otherwise specified. P values are shown on top of the corresponding columns, as determined by 1-way ANOVA followed by Tukey’s multiple comparisons post-hoc test.

MC, a major contributor to immune reactions, releases various inflammatory cytokines and mediators following activation and subsequent degranulation process. Ligands like the canonical MRGPRX2 agonist CST-14 have been found to activate MC, mobilise calcium, and subsequently cause MC degranulation (25). Our previous research demonstrated how the MRGPRX2 antagonist GE1111 changed the EC₅₀ and E_max response of CST-14 on MC degranulation and MRGPRX2 activation (36). The IC₅₀ value of GE1111 was 16.24 µM (MC degranulation assay) and 35.34 µM (MRGPRX2 activation assay). Using the MC degranulation assay, we found that GE1111 changed the EC₅₀ of CST-14 from 1.71 ± 1.12 µM to 27.17 ± 2.73 µM (36). In addition, GE1111 changed the EC₅₀ of CST-14 from 0.127 ± 1.2 µM to 1.720 ± 1.36 µM in the MRGPRX2 activation assay (36). These data demonstrated the ability of GE1111 to inhibit MC degranulation and MRGPRX2 activation in vitro. Moreover, we looked at the downstream signalling pathway of CST-14 MRGPRX2 mediated MC degranulation and inflammatory cytokine gene expression. Figure 1A demonstrates the representative western blot images and protein quantification results of downstream signalling pathways (39, 40). We found a significant elevation in the ERK_1/2 (^***P < 0.001) and STIM1 (^***P < 0.001) signalling in CST-14 treated MC as compared to vehicle control MC ( Figures 1Ai–iii ). GE1111 treated MC (50 µM GE1111+CST-14) showed a significant reduction in the expression of ERK _1/2 (^****P < 0.0001) and STIM1 (^***P < 0.001) as compared to solely CST-14 treated MC. AD is characterised by a surge of inflammatory cytokines such as IL-13, IL-31, MCP-1, and TNF-α released from MC and keratinocytes. To assess the effect of GE1111 on CST-14 induced gene expression of inflammatory cytokines in MC, we quantified the gene expression of these cytokines by RT-qPCR. We measured the gene expression of IL-13, IL-31, MCP-1, and TNF-α, which were overexpressed in AD. Figure 1B showed a significant increase in the gene expression of IL-13 (^*P < 0.05), IL-31 (^****P < 0.0001), MCP-1 (^****P < 0.0001), and TNF-α (^****P < 0.0001) in CST-14 control MC as compared to vehicle control MC. However, GE1111 treated MC showed a significant reduction in the gene expression of IL-13 (^*P < 0.05) and other cytokines (IL-31, MCP-1, and TNF-α, ^****P < 0.0001) as compared to CST-14 control MC. This demonstrates the ability of GE1111 to inhibit MRGPRX2 activation by the canonical agonist CST-14, inhibit subsequent degranulation of MC, maintain consistent downstream signalling and reduce inflammatory cytokine gene expression.

In AD, both the immune and non-immune cells, including even keratinocytes, release type 2 inflammatory cytokines, further aggravating the AD symptoms, such as defects in the skin integrity (41). MC is one of the effector cells that remain in close contact with keratinocytes, potentially interacting with and modulating keratinocyte activity. We treated human HaCaT cells with MC supernatant (MC were treated with CST-14 or vehicle in the presence or absence of GE1111) ( Figure 2A ). TSLP has been found to play an essential role in the persistent MRGPRX2 activation seen in chronic skin allergies and is integral in maintaining sustained MRGPRX2 agonist sensitivity (42–46). Claudin 1, a tight junction protein, is an essential protein in epidermal barrier integrity; decreases in claudin represent a disruption of the epidermal barrier, as seen in atopic dermatitis (47–50). In the immunofluorescence staining and western blotting of keratinocytes, CST-14 MC supernatant-treated keratinocytes showed a significantly less ( Figures 2B, Ci, ii , ^*P <0.05) expression of tight junction protein claudin 1 (green colour) compared to negative control (no mast cell supernatant) keratinocytes. There was no significant difference in the claudin 1 expression in keratinocytes, which were challenged with MC supernatant without any stimulus. In line with the previous results, the GE1111 treatment reversed the change in claudin 1 expression ( Figures 2B, Ci, ii , ^*P <0.05). In addition to this, we found a significant increase in the type 2 cytokine TSLP (red colour) in CST-14 MC supernatant-treated keratinocytes as compared to the negative control (no mast cell supernatant) keratinocytes ( Figures 2B, Ci, iii , ***P <0.001). However, we found that CST-14+GE1111 MC supernatant-treated keratinocytes had significantly decreased TSLP (**P <0.01) expression as compared to CST-14 MC supernatant-treated keratinocytes (disease control).

Next, we measured the inflammatory cytokines gene expression in different experimental groups of MC supernatant-treated keratinocytes by RT-qPCR. Figure 2D showed a significant upregulation of the inflammatory cytokines, including MCP-1 (^***P < 0.001), TNF-α (^**P < 0.01), TSLP (^**P < 0.01), and IL-1ß (^*P < 0.05) in the CST-14 MC supernatant treated keratinocytes compared to the negative control (no mast cell supernatant) keratinocytes. There was no significant difference in the gene expression of inflammatory cytokines keratinocytes challenged with MC supernatant without any stimulus. In addition to the previous experimental results, we found that keratinocytes challenged with CST-14+GE1111 MC supernatant exhibited significant decreases in CST-14 induced gene expression of inflammatory cytokines as compared to the group treated only with CST-14 MC supernatant ( Figure 2Di MCP-1, ^****P < 0.0001 and Figure 2Dii–iv ), ^**P < 0.01). Each of these inflammatory cytokines plays a crucial role in the pathogenesis of AD. MCP-1 is an inflammatory chemokine that is essential in the migration and infiltration of both monocytes and macrophages to sites of inflammation or injury. MCP-1 has been linked to the pathophysiology of both AD and psoriasis (51). Tumour necrosis factor alpha (TNF-α) is a pro-inflammatory cytokine that plays a prominent role in multiple chronic inflammatory diseases, including AD (52). IL-1ß is a Th1-related inflammatory cytokine that has long been implicated in AD pathogenesis (53, 54). Collectively, our data demonstrated that the novel MRGPRX2 antagonist GE1111 can antagonise MRGPRX2-MC mediated CST-14 induced inflammatory cytokine expression in keratinocytes.

Besides MC, macrophages are one of the key immune cells that play a significant role in AD pathogenies (55, 56). We proposed that MC-macrophage crosstalk, mediated by MC-MRGPRX2 and its ligands CST-14, may play an essential role in modulating macrophage activity. To test this hypothesis, we treated RAW 264.7 cells with MC supernatant (MC were treated with CST-14 or vehicle in the presence or absence of GE1111) ( Figure 3A ). We measured the macrophage phagocytosis response in different experimental groups via a fluorescence-based assay. To evaluate the efficacy of GE1111 in treating macrophage dysregulation (compromise of macrophage capability) characteristic of atopic dermatitis, we performed macrophage phagocytosis assay using the in-vitro co-culture model with the RAW 264.7 macrophage cell line (56). The phagocytosis targets, represented by the yellow beads, were incubated with the RAW 264.7 macrophages in the presence and absence of MC supernatant from different experimental groups. We found a significant decrease in the number of phagocytosed beads in the macrophages within the CST-14 MC supernatant treatment as compared to negative control (no mast cell supernatant) macrophages ( Figures 3Bi, ii , ^****P <0.0001). However, macrophages challenged with CST-14+GE1111-treated MC supernatant significantly reversed the decrease in the phagocytosed beads compared to CST-14 MC supernatant-treated macrophages. ( Figures 3Bi, ii , ^**P < 0.01). Furthermore, we observed a notable reduction (^**P < 0.01) in the phagocytic activity of macrophages when treated with mast cell supernatant (without stimulation), in comparison to the negative control. This could potentially be attributed to the presence of inflammatory mast cell mediators, which are released spontaneously without any external stimulation. The findings from this experiment are consistent with previous clinical findings in AD patients, which found a decrease in the phagocytic response by mononuclear phagocytes (55, 57). The novel small molecule MRGPRX2 antagonist GE1111 significantly restored macrophage phagocytotic activity, which can protect the skin from the Staphylococcus aureus colonisation characteristic of AD and AD pathophysiology (58).

To further evaluate the protective effect of GE1111 in AD, we developed an in-vivo, 2,4-dinitrofluorobenzene (DNFB)-induced mouse model ( Figure 4A ). DNFB/vehicle was painted along the experimental mice's shaved dorsal skin and ears weekly for four weeks. GE1111 was injected via intraperitoneal injection at 10mg/kg (T10) or 20mg/kg (T20) body weight. We measured the clinically related symptoms, such as thickening, scaling and erythema of skin in mice at the end of week 4. Figure 4B depicts representative images of mice skin showing changes in the skin on day seven and day 28 of the treatment. Figures 4Ci–iii showed the quantitative score of thickening, scaling and erythema of the skin, respectively, in experimental mice. DFNB-treated mice (disease control) showed a significant increase in the thickening (^****P < 0.0001), scaling (^****P < 0.0001) and erythema score (^****P < 0.0001) as compared to vehicle control mice. GE1111 treated mice, at both doses, experienced significant reductions in the severity of these phenotypic changes (thickening and erythema of skin, ^****P < 0.0001). However, we found no significant change in the skin scaling score at 20 mg/kg GE1111, while there was a significant change in the lower dose at 10 mg/kg GE1111 (^***P < 0.001). We then measured the serum MCP-1 level in mice to relate the serum levels of this cytokine to the AD symptoms and severity. As can be seen in Figure 4D , we found a significantly higher level (**P < 0.01) of MCP-1 in DFNB-treated mice (disease control) as compared to vehicle-control mice. Mice treated with GE1111 10 mg/kg dose showed a reduction in serum MCP-1 level but was not significant as compared to DFNB-treated mice. However, mice treated with GE1111 20 mg/kg demonstrated a significant reduction in serum MCP-1 levels compared to the disease-control mice (^**P < 0.01). To measure and quantify the inflammation of experimental mice's skin and ear tissue, we performed hematoxylin and eosin (H&E) staining ( Figure 4E ). Our findings were consistent with the phenotypic changes observed in the skin and ear (scaling and thickness of skin/ear); we found a significant increase in skin and ear’s epidermal thickness ( Figures 4Eii, iii , ^****P < 0.0001) and immune cell infiltration of DFNB treated mice (disease control) as compared to vehicle control mice. Notably, in the skin H&E, we found a significant hyperkeratosis or scaling around the epidermis. However, upon treatment with GE1111, the epidermal thickness was significantly reduced in 10 mg/kg and 20 mg/kg dosage groups ( Figures 4Eii, iii , ^****P < 0.0001). Collectively, our in vivo study demonstrated the promising protective effect of the novel MRGPRX2 antagonist GE11111 in reducing AD-dependent/related phenotypic changes and inflammation.

To gain further insights into the molecular mechanism and the effect of GE1111 in mice with AD, we conducted toluidine blue staining, immunohistochemistry, quantitative gene expression, and protein expression analysis on the skin tissue. MC degranulation and release of inflammatory mediators is a crucial event in the pathogenesis of AD. Figures 5Ai, ii shows the increase in the degranulated MC in the ear tissue of DFNB-treated mice as compared to vehicle control mice (^****P < 0.0001). We have tested the antagonistic activity of GE1111 on in vitro MC degranulation and expected a similar response in the in vivo MC degranulation (36). As expected, we found a significant reduction in the degranulated MC in GE1111-treated mice compared to DFNB-treated mice ( Figure 5Aii , ^****P < 0.0001). Figure 5B illustrates the representative immunohistochemistry images of the quantitative expression of Involucrin (59) and periostin (60) in the epidermal and dermal region of the skin tissue, respectively, from the experimental mice. Involucrin is a tight junction protein expressed constitutively around the epidermis of mice skin. In AD, the expression of Involucrin has been reported to decrease clinically, as well as in mouse models of AD (60, 61). DFNB-induced AD mice showed a significant reduction in the expression of Involucrin around the epidermal area as compared to vehicle control mice ( Figures 4Bi, ii ), ^***P < 0.001). GE1111 10 mg/kg (^*P < 0.05) and 20 mg/kg (^**P < 0.01) treated mice showed a significant dose-dependent protection of this tight junction protein in the epidermis region of skin tissue. We also looked at the expression of periostin, which is overexpressed in chronic skin inflammatory conditions, including AD (60, 62). The DFNB-treated mice exhibited a significantly higher expression of periostin in the epidermis compared to the vehicle control mice ( Figures 4Bi, iii , ^****P < 0.0001). However, mice treated with GE1111 at 10 mg/kg (^**P < 0.01) and 20 mg/kg (^***P < 0.001) showed a significant dose-dependent reduction in the periostin expression ( Figures 4Bi, iii ), indicating a potential therapeutic effect of GE1111 in attenuating periostin-associated inflammation in AD (60).

Furthermore, to assess the effect of GE1111 on inflammatory cytokines such as TSLP, IL-13 and IL-1ß, we measured the gene expression of these cytokines in mice with AD skin lesions. Figure 5C demonstrated a significant upregulation of TSLP (^**P < 0.01), IL-13 (^**P < 0.01), and IL-1ß (^*P < 0.05) in DFNB-treated mice skin compared to vehicle control mice skin. On the other hand, mice treated with GE1111 at 10 mg/kg and 20 mg/kg body weight significantly reduced the gene expression of TSLP (^**P < 0.01), IL-13 (^**P < 0.01) and IL-1ß (^****P < 0.0001) compared to DFNB-treated mice. These findings are consistent with our in-vitro findings with MC and keratinocytes. We also looked at the protein expression of claudin 1, TSLP and the downstream signalling pathway of GE1111 in mice model of AD ( Figure 5D ). We measured the protein expression of the effect of STIM1 and AKT pathway; DFNB disease control mice skin showed a higher expression of STIM1 and p-AKT as compared to vehicle control mice skin (^*P < 0.05 & ^**P < 0.01 respectively). In comparison, GE1111 treatment significantly reduced the expression of STIM1 (at both dosages, ^*P < 0.05) and p-AKT (at 10 mg/kg, ^*P < 0.05 and 20 mg/kg, ^**P < 0.01). Furthermore, similar to the in vitro and immunohistochemistry results, we found a significant increase in the TSLP expression in DFNB control mice ( Figures 5Di, v , ^**P < 0.01). TSLP expression was significantly reduced in GE1111 20 mg/ kg treated mice (^**P < 0.01), while no significant difference was found in the lower dose group. In addition, we did not find a significant change in the claudin 1 expression in DFNB control mice, although it reduced a little. Surprisingly, GE1111-treated mice showed significantly higher claudin 1 expression (^*P < 0.05) than DFNB control mice. Collectively, our experimental data on the protein and gene expression of key inflammatory cytokines and downstream signalling pathways showed a promising protective effect of GE1111 in the AD mouse model.