Ta-Xi-San Suppresses Atopic Dermatitis Involved in Multitarget Mechanism Using Experimental and Network Pharmacology Analysis

Atopic dermatitis (AD) is a relapsing and chronic skin inflammation with a common incidence worldwide. Ta-Xi-San (TXS) is a Chinese herbal formula usually used for atopic dermatitis in clinic; however, its active compounds and mechanisms of action are still unclear. Our study was designed to reveal the pharmacological activities, the active compounds, and the pharmacological mechanisms of TXS for atopic dermatitis. Mice were induced by 2,4-dinitrocluorobenzene (DNCB) to build atopic dermatitis model. The pathological evaluation, enzyme-linked immunosorbent assay (ELISA), and hematoxylin and eosin (H&E) assay were performed. The UPLC-Q-Exactive-MSE and network pharmacology analysis were performed to explore active ingredients and therapeutic mechanisms of TXS. TXS treatment decreased levels of immunoglobulin E (IgE), interleukin-4 (IL-4), and tumor necrosis factor-α (TNF-α) in serum induced by DNCB. TXS reduced scratching behavior and alleviated inflammatory pathology of skin and ear. Meanwhile, TXS decreased the spleen index and increased spleen index. The UPLC-Q-Exactive-MSE results showed that 65 compounds of TXS were detected and 337 targets were fished. We collected 1371 AD disease targets, and the compound-target gene network reveled that the top 3 active ingredients were (−)-epigallocatechin gallate, apigenin, and esculetin, and the core target genes were PTGS2, PTGS1, and HSP90AA1. The KEGG pathway and GO analysis showed that TXS remedied atopic dermatitis via PI3K-Akt signaling pathway, mitogen-activated protein kinase (MAPK) signaling pathway, and Toll-like receptor (TLR) signaling pathway with the regulation of inflammatory response and transcription. Further, we found that the targets of PTGS2 and HSP90AA1 were both elevated in ears and skin of AD model mouse; however, TXS decreased the elevated expressions of PTGS2 and HSP90AA1. Our study revealed that TXS ameliorated AD based on (−)-epigallocatechin gallate, apigenin, and esculetin via targeting PTGS2 and HSP90AA1.


Introduction
Atopic dermatitis (AD) is a relapsing, intensely pruritic inflammatory disordering of the skin [1]. It is one of the most common health problems and affects 2.1%-4.9% of human population worldwide [2]. Meanwhile, the morbidity of atopic dermatitis is high in both developing countries and developed countries [3,4]. Atopic dermatitis is characterized by rash, erythema, edema, hemorrhage, lichenification, and severe itching, and it is always accompanied by destruction in skin barriers [5]. e pathogenesis of AD is closely related to the imbalance of T helper 2 ( 2) and T helper 1 ( 1) factors [6]. 2-based allergic inflammation plays a dominant role in the early stages of an acute outbreak of AD, which leads to the raised levels of IL-4, IL-13, and IgE [7]. However, 1-dominant allergic inflammation is predominated in the chronic phase of AD and mainly releases IL-2, TNF-α, and IFN-c [8,9]. Moreover, multiple inflammatory cells are found in the epidermis and dermis of AD patients including mast cells, monocytes/macrophages, eosinophils, and T lymphocytes [10,11].

TXS Preparation.
TXS (160 g, Shaanxi Academy of Traditional Chinese Medicine) was soaked with 1000 mL in water for 30 min, decocted for 5 min again after boiled, and then filtrated. e decoction filtrated was concentrated to 160 mL (1 g crude drug/mL). e solution was filtered using 0.22 mm membranes. e filtrate was used and stored at 4°C before its use for animal assays.

Building Atopic Dermatitis Mice Model and Treatment.
BALB/c mice (male, 18-20 g) were purchased from Xi'an Jiao Tong University (Xian, China). All the mice were fed under the standard experimental conditions (−25 ± 2°C, humidity 60 ± 5%, and 12/12 h day/night cycle). e AD mice model was induced by DNCB (dissolved in acetone/ olive oil (3 : 1)), which was applied to dorsal skin and ear. Mice were segmented into 6 groups (n � 6 per group): (1) normal group, (2) matrix group, (3) DNCB model group, (4) DNCB + matrix + 0.65 mg crude drug/mL TXS, (5) DNCB + matrix + 1.30 mg crude drug/mL TXS, and (6) Hydrocortisone Butyrate Cream (Hyd B). Firstly, the mice were shaved with an aesthetic state (ethyl carbamate) using an electric shaver. en, 100 μL 0.5% DNCB or acetone/olive oil (3 : 1) was applied to the shaved dorsal skin for 2 days, and there was no treatment in the 3rd∼6th days . 70 μL 0.5% DNCB on dorsal skin and 20 μL 0.5% DNCB on ear every 2 days were applied from the 7th day. e AD model was finished on the 28th day for a total of 10 times (on days 1, 2, 7, 10, 13, 16, 19, 22, 25, and 28). DNCB was applied on the same site r. TXS was applied from the 7th day, 500 μL on dorsal skin and 200 μL on ears, twice a day until the 28th day. Hyd B was also applied from the 7th day, 300 mg on dorsal skin and 70 mg on ears, twice a day until the 28th day. e flowchart of the experimental program we studied is shown in Figure 1. Our animal study followed the guidelines (Care and Use of Laboratory in Xi'an Jiao Tong University Health Science Center, license approval no. SCXK (Shan) 2017-003).

Evaluation of Skin Lesion and Measurement of Ear
Weight.
e relative severity of AD was evaluated weekly via scoring the lesions of the dermal tissues with the scoring procedure [29]: (1) erythema and hemorrhage, (2) erosion (excoriation), (3) edema, and (4) scaling (dryness). In each score, 0 was defined as exhibiting no symptoms, 1 as mild symptoms, 2 as moderate symptoms, and 3 as severe symptoms. Only one researcher participated in scoring the apparent AD symptoms throughout the experiment to 2 Evidence-Based Complementary and Alternative Medicine minimize technique variations. e ears of each mouse were harvested and weighted.

e ymus Index and Spleen Index
Measurement. e spleen and thymus of mice in each group were removed integrally. e thymus index and spleen index were defined as the organ/body ratio (organ mass (mg)/corresponding animal mass (g)).
e organ and the body weight were calculated using an electronic balance.

Histological Analysis.
e dorsal skin and ear samples were taken 24 h after final herbal administration and then fixed in 4% paraformaldehyde buffer, followed by embedding of the tissues in paraffin wax. e tissue sections were 2 μm thick, and then they were stained with H&E.

Cytokines Quantitation.
Total IgE in serum was analyzed using an ELISA Quantitation Kit according to the manufacturer's protocol. e serum concentrations of the cytokines, including TNF-α and IL-4, were also quantified using a mouse cytokine enzyme immunoassay kit, respectively.

Network Construction.
For better dissecting the latent active ingredients and key targets of TXS, we established the herb-ingredient-target-pathway network using Cytoscape 3.7.1 software. en, the quantitative "degree" was analyzed by plugin network analyzer. According to the descending order of "degree," we selected the vital ingredients and targets of TXS against AD.

Functional Enrichment Assay.
e Gene Ontology (GO) enrichment and the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway (https://www.genome.jp/kegg/ pathway.html) assay was performed based on the cluster-Profiler software package on R platform and analyzed using visualization. Afterwards, the GO interactive network and the bubble diagram of KEGG pathway were structured using the top GO packet of R platform. P < 0.05 was calculated meaningfully in these two enrichment analyses.

TXS Alleviates the Lesions in Ear and
Back. e skin inflammation of the back and ear of mice in each group was observed by visual method (Figure 2(a)). e skin of the ear and back of normal mice was smooth and healthy. Compared with the normal mice, the ear of mice in DNCB group was obviously swollen, rough, desquamate, and auricle with scab. Meanwhile, the back skin of mice in model group was thickened, rough, erythema, desquamate, and scab, which showed a typical dermatitis feature. However, TXS dramatically decreased the excoriation, striking hemorrhage, and erosion and suppressed the scratching behavior compared to the DNCB-induced mice. In addition, in HYB treatment group, the inflammatory response was lighter, the damage and scratches just existed in edge of the ear, the back skin was thinner, and the vascular lines were clear.
At the same time, the lesion of back skin was scored according to the "inflammatory score standard." Compared with the normal mice, the back inflammatory score of mice in DNCB group was significantly enhanced, and TXS (1.3 mg crude drug/mL) significantly reduced the inflammatory score ( Figure 2(b)). ere was a significant increase in ear weight of mice in DNCB group due to swelling of the ear compared with the mice in the normal group (Figure 2(c)). TXS and HBY administration decreased the ear weight and back of mice, respectively. In conclusion, both TXS and HYB promoted inflammation subsidence and skin healing of mice. (c) e ear weight value. Values are expressed as mean ± SD (n � 6). * P < 0.05 and * * P < 0.01 compared with the normal group;^P < 0.05 and P < 0.01 compared with the matrix group; # P < 0.05 and ## P < 0.01 compared with the DNFB model group.

TXS Improved the ymus Index and Spleen
memorably reduced; TXS administration could improve the thymus index of mice. e spleen index of DNCB-stimulated mice was significantly higher than that in the normal mice. High-dose TXS administration effectively reduced the spleen index.

Histological Analysis Results.
Hematoxylin study was used to recognize whether TXS regulates pathological symptoms of AD-like skin lesions and ear injury in DNFBinduced AD mouse model. In model group, skin tissues of mice showed typical inflammatory pathological changes including parakeratosis, thickening of the spinous layer, telangiectasia, and infiltration of inflammatory cells in the dermis (Figure 4(a)). In mice in TXS and HYB treatment group, there was no parakeratosis in skin tissue, and the spinous layer was thickened; however, a trifling infiltration of inflammatory cell was still in the dermis (Figure 4(a)).
Meanwhile, ear histological analysis showed that TXS decreased the epidermal thickening and inflammatory cells accumulation in the lesions (Figure 4(b)).

TXS Regulated the Level of Cytokines.
Elevated levels of inflammatory cytokines and inflammatory cell infiltration are a typical characteristic in AD [30]. us, we estimated the cytokines level in the serum of DNFB-sensitized mice. Subsequently, we observed that the TXS has an effect on the release of inflammatory factors. e levels of IL-4, TNF-α, and IgE in serum were significantly increased in the model group's mice induced by DNFB compared to the normal group. However, production of IL-4, TNF-α, and IgE was inhibited by TXS and Hyd B treatment (Figures 5(a)-5(c)).

e Ingredients Analysis of TXS Extracts.
e herb components of TXS extracts were detected via aligning each detectable compound mass data for each detectable compound with structure and confirmed by MS E substructure data. e Base Peak Chromatogram (BPC) of TXS extracts in positive ion mode and negative ion mode is shown in Figures 6(a) and 6(b). TXS mainly contained alkaloids and flavonoid glycosides, and these ingredients presented a better MS response in positive than in the positive ion mode. In this research, 65 compounds extracted in TXS were identified through comparing and analyzing the primary and secondary mass spectrometry information, combining with databases (Metlin, Massbank, and Human Metabolome Database) and combining with retention behavior, database, Values are expressed as mean ± SD (n � 6). * P < 0.05 and * * P < 0.01 compared with the normal group;^P < 0.05 and^P < 0.01 compared with the matrix group; # P < 0.05 and ## P < 0.01 compared with the DNFB model group.

Normal
Matrix e 65 compounds were confirmed as the principal compounds for the following assay and are detailedly displayed in Table 1.

Potential Target Genes and Network Analysis.
e 337 targets were collected form the 65 compounds in TXS based on UPLC-Q-Exactive-MS E analysis after eliminating the duplicate targets. Simultaneously, 1371 AD-related targets were collected via searching CTD, GeneCards, and OMIM databases, and 1241 targets were decontaminated after removing duplicates. en the compound targets and disease-related targets were merged, and we got the compound-target gene network. e ingredient-target gene network is presented in Figure 7. is network consisted of 190 nodes (1 TXS formula node, 52 compounds, and 137 target genes) and 522 edges. Compounds in the network were classified as catechin class, flavonoids, coumarin, isoflavone, and alkaloids. Particularly, as shown in Tables 2 and 3

e Results of KEGG and GO Analysis.
KEGG pathway enrichment showed that 199 pathways were harvested and the remarkable 30 related pathways are presented in Figure 8(a). Pathways in PI3K-Akt signaling pathway, MAPK signaling pathway, and TNF signaling pathway may be the latent mechanism to exert effects of TXS against AD. Subsequently, the results of GO evaluation of the top 10 enriched results are shown in Figure 8(b). e enriched biological process ontologies were dominated by positive regulation of transcription from RNA polymerase II promoter, negative regulation of apoptotic process, response to drug, inflammatory response, and positive regulation of transcription. e molecular function of targets was mainly enriched in ontologies including protein binding, identical protein binding, protein homodimerization activity, and enzyme binding. Meanwhile, cell component result showed that cytoplasm is the largest constitution.   Evidence-Based Complementary and Alternative Medicine regulatory targets for inflammatory response in AD and are core targets screened via network pharmacology. us, we investigate the regulatory effect of TXS on activation of PTGS2 and HSP90AA1. In our study, we found that the protein expressions of PTGS2 and HSP90AA1 in ears and skin of model mice were significantly enhanced compared to the normal mice; however, TXS decreased the elevated expressions of PTGS2 and HSP90AA1 (Figures 9(a)-9(d)).
Our results revealed that TSX might be a useful drug for AD therapy via regulating PTGS2 and HSP90AA1 expressions.

Discussion
e DNCB-induced skin lesion in mice is a classical animal model for AD research [31]. us, we established DNCBinduced mouse model to assess the effects of TXS on AD. IL-4 and TNF-α are risk factors for developing atopic dermatitis [32]. Previous studies showed that cytokines including IL-4, TNF-α, and IL-1β-were enhanced in serum of AD patients [33]. Higher level of IFN-c and IL-6 was detected in skin lesions of AD patients [34]. In our study, the levels of IL-4, TNF-α, INF-c, and IgE in serum were reduced; meanwhile,       Evidence-Based Complementary and Alternative Medicine 9 the inflammation of back skin, the skin inflammation score, and the ear swelling of mice were alleviated after taking the TXS. Hence, TXS might be effective to balance atopic and inflammatory disorders in the progress of atopic dermatitis.
In addition, to explore the effects of TXS on skin and ear inflammation, we performed H&E assay and examined thymus index and spleen index. e mice induced by DNFB showed typical AD features, such as epidermal hyperplasia and massive inflammatory cells infiltration, which were coincided with the typical symptoms of AD. TXS treatment effectively improved the skin and ear status. e thymus is a major immune organ, the index of which is an important index for assessing immunity [35]. Meanwhile, as another lymphatic organ of the human body, the spleen palys a core role in immune system and is used for assessing immunomodulatory effects of drugs [36]. TXS treatment lowered the spleen index and enhanced the thymus index. erefore, TXS regulated pathological changes in the development of AD.
UPLC-Q-Exactive-MS E as a rapid, efficient, and sensitive method is always used for analyzing extracts and bioactive      ingredients from medicine and herbs [37,38]. erefore, we applied UPLC-Q-Exactive-Ms E to analyze the information on compounds in herbal medicines of TXS. 65 components in TXS were identified, including alkaloids, phenols, flavonoids, and acids. Among them, various ingredients showed anti-inflammatory activity. Matrine, a quinolizidine alkaloid extracted from Sophora flavescens, has the capacity to effectively suppress experimental autoimmune encephalomyelitis and optic neuritis [39,40]. Oxyberberine prevented lipopolysaccharide-induced acute lung injury and improved colitis [41,42]. Epicatechin from garden burnet, epiberberine from golden cypress, and atractylenolide I from Rhizoma atractylodis showed anti-inflammatory activity [43][44][45]. us, we predict that TXS has an anti-inflammatory material basis. e network pharmacology study as a major tool systematically revealed the effects and mechanism of multicomponent and multitarget system, such as traditional Chinese medicine [46,47]. Here, we researched the pharmacological mechanisms of TXS against atopic dermatitis through using the network analysis method. e ingredienttarget gene network pharmacological analysis of TXS identified 52 ingredients and 137 target genes connected with AD. at means that one chemical component acted with an average of 9.03 targets, while one target was bound with an average of 3.43 compounds. e compound-target gene network of TXS revealed that (−)-epigallocatechin gallate (degree � 61, CC � 0.4797, BC � 0.3317), apigenin (degree � 41, CC � 0.4355, BC � 0.1750), and esculetin (degree � 34, CC � 0.4219, BC � 0.1296) were the major pharmacodynamic compositions of TXS according to the degree. Epigallocatechin gallate is a single ingredient from TCM and shows effective functions against AD due to the antioxidant and immunomodulatory effects [48]. Apigenin inhibits inflammatory cytokines (IL6, TNF-α, IL-5, and IL-13) production to decrease allergic responses in RBL-2H3 Values are expressed as mean ± SD (n � 6). * P < 0.05 and * * P < 0.01 compared with the normal group;^P < 0.05 and^P < 0.01 compared with the matrix group; # P < 0.05 and ## P < 0.01 compared with the DNFB model group.
cells and repairs the physical barrier of the skin in HaCaTcell model [49]. Esculetin attenuates atopic skin inflammation by inhibiting IFN-c, IL-13, IL-31, and IL-17 release via regulating nuclear factor-κB pathway [50]. Furthermore, in this network, the first 3 targets of degree values were PTGS2 (degree � 41, CC � 0.4846, BC � 0.1297), PTGS1 (degree � 27, CC � 0.3873, BC � 0.0358), and HSP90AA1 (degree � 21, CC � 0.3691, BC � 0.0224). Meanwhile, the three targets with higher BC and CC were recognized as critical targets of TXS for AD treatment. us, our results indicated that TXS as the complex system of herbal compounds displayed multipharmacological and superimposed effects.
Based on the enrichment results of KEGG, we determined that the effects of TSX against AD may be due to pathways like PI3K-Akt signaling pathway, MAPK signaling pathway, 17 cell differentiation, and TLR signaling pathway. e mitogen-activated protein kinase (MAPK) pathway is responsible for inflammatory and immune response and is widely expressed in multiple tissues [51]. Previous study showed that Solanum nigrum Linne possessed anti-inflammatory effects in vitro through regulating MAPK/NF-κB pathway in HaCaT cell model [52]. It has been reported that inhibiting TLR signaling pathway was a critical approach for chronic allergic skin inflammation treatment [53]. Artemisia argyi Folium extract ameliorates DNCB-induced skin lesions in BALB/c mice via inhibiting PI3K/Akt pathway [54]. In this study, target genes enrichment was mainly associated with the biological processes including positive regulation of transcription from RNA polymerase II promoter, negative regulation of apoptotic process, inflammatory response, and positive regulation of transcription. Further, enriched MF ontologies were dominated through protein binding, enzyme binding, identical protein binding, and protein homodimerization activity. In addition, CC analysis displayed that cytoplasm was the largest constitution. Further, we found that TXS downregulated the expressions of PTGS2 and HSP90AA1. More importantly, the center target HSP90AA1 selected by the compound-target gene network was involved in the PI3K-Akt signaling pathway and PTGS2 was the major target for anti-inflammation.

Conclusions
In conclusion, our results revealed that TXS attenuated IL-4, IgE, and TNF-ɑ levels in serum, increased thymus index, and decreased spleen index. TXS ameliorated behavioral changes and alleviated pathological changes of skin and ear in mice with the DNCB-induced atopic dermatitis. Additionally, TXS was mainly consisted of 65 chemical components. Among them, 52 components were closely associated with the effect of TXS against AD. TXS downregulated PTGS2 and HSP90AA1, which were associated with the PI3K-Akt signaling pathway. Our results revealed the effects, components, and pharmacological mechanism of TXS for AD, which provided scientific basis for clinical application.

Data Availability
All the data used to support the findings of this study are submitted and are available in this published article.

Ethical Approval
All animal experiments were performed according to the protocol approved by Animal Ethics Committee and Institution Animal Care (license approval no.. SCXK (Shan) 2017-003).

Conflicts of Interest
e authors declare that they have no conflicts of interest.