Abstract: 【Abstract】 Objective To investigate the effects of quercetin on gut microbiota homeostasis and inflammatory factors in a 2,4-dinitrofluorobenzene （DNFB）-induced atopic dermatitis （AD）-like mouse model, and to provide a new natural active ingredient option and an experimental basis for the clinical treatment of AD. Methods A total of 70 specific pathogen-free BALB/c female mice were randomly divided into 7 groups （10 mice per group） by block randomization: blank control group, model group, dexamethasone （5 mg/kg） group, Stattic （a signal transducer and activator of transcription 3 inhibitor, 5 mg/kg） group, and low-dose （12.5 mg/kg）, medium-dose （25 mg/kg）, and high-dose （50 mg/kg） quercetin groups. Except for the blank control group, all other groups were treated with DNFB to establish the AD-like mouse model. After successful modeling, each drug treatment group was given the corresponding drug by gavage for 2 weeks, while the blank control group and model group received an equal volume of 0.9% sodium chloride solution by gavage. At the end of the experiment, the Eczema Area and Severity Index （EASI） was evaluated; transepidermal water loss （TEWL） and spleen index were measured; mouse dorsal skin tissues were excised for histopathological examination and mast cell staining; enzyme-linked immunosorbent assay and real-time quantitative PCR were performed to detect the levels of inflammatory factors in sera and skin lesions, including Th2-related factors （interleukin ［IL］-4, IL-5, IL-13, IL-31）, the Th1-related factor interferon-γ （IFN-γ）, the Th17-related factor IL-17A, IL-10, and transforming growth factor-β （TGF-β）. The diversity of gut microbiota and the abundance of key phyla in mice were analyzed using 16S rRNA high-throughput sequencing. During the experiment, some animals were excluded due to modeling failure, death, or other reasons, and finally 6 mice in each group were included for data analysis. Statistical analysis included one-way analysis of variance and Spearman correlation analysis. Results After 2 weeks of continuous administration of drugs, the model group showed obvious erythema, infiltration, and edema, accompanied by marked desquamation and crusting; the dexamethasone group showed substantial remission of skin lesions; skin lesions in the quercetin groups were improved to varying degrees, with marked improvement in the high-dose quercetin group. There was a significant difference in EASI scores among the groups （F = 88.21, P < 0.001）; the EASI score was significantly higher in the model group than in all other groups （all P < 0.05）, but significantly lower in the medium- and high-dose quercetin groups than in the dexamethasone group （both P < 0.05）. The TEWL level and spleen index were significantly higher in the model group than in the blank control group （both P < 0.05）; the TEWL level was significantly lower in all drug treatment groups than in the model group, while the spleen index was significantly lower in the dexamethasone group, high-dose quercetin group, and Stattic group than in the model group （all P < 0.05）. Compared with the blank control group, the model group showed thickened epidermis and dermis, increased mast cell infiltration, and higher levels of IL-4, IL-5, IL-13, IL-31, IFN-γ, IL-17A, TGF-β, and IL-10 in both sera and skin lesion tissues （all P < 0.05）; compared with the model group, all drug treatment groups showed varying degrees of attenuation in dermal mast cell infiltration and lower levels of the aforementioned inflammatory factors （all P < 0.05）; among them, the serum levels of IL-4, IL-5, IL-13, IL-31, IL-17A, and TGF-β were significantly lower in the high-dose quercetin group than in the dexamethasone group （all P < 0.05）. Gut microbiota analysis showed that the microbiota structure of the model group was clearly separated from that of the blank control group, while the microbiota structures of the quercetin groups approached those of the blank control group. At the phylum level, compared with the blank control group, the model group showed significantly decreased relative abundance of Firmicutes （0.32 ± 0.06 vs. 0.78 ± 0.25） and Firmicutes/Bacteroidetes ratio （0.89 ± 0.21 vs. 4.83 ± 0.62）, but significantly increased relative abundance of Bacteroidetes （0.36 ± 0.05 vs. 0.16 ± 0.04） （all P < 0.05）; compared with the model group, all drug treatment groups showed significantly higher relative abundance of Firmicutes （all P < 0.05）; regarding Bacteroidetes, the low- and medium-dose quercetin groups and the dexamethasone group showed significantly lower relative abundance compared with the model group, whereas the high-dose quercetin group and the Stattic group showed significantly higher relative abundance （all P < 0.05）. At the genus level, the blank control group was dominated by Lactobacillus and Muribaculum; compared with the blank control group, the model group showed significantly lower relative abundance of Lactobacillus, but significantly higher relative abundance of Escherichia-Shigella （all P < 0.05）; compared with the model group, all drug treatment groups showed significantly lower relative abundance of Escherichia-Shigella （all P < 0.05）; regarding Lactobacillus, the low- and medium-dose quercetin groups and the dexamethasone group showed significantly higher relative abundance compared with the model group （all P < 0.05）, while the high-dose quercetin group and the Stattic group showed significantly lower relative abundance （both P < 0.05）. The absolute abundance of Proteobacteria was positively correlated with the EASI scores and TEWL levels （rs = 0.37, 0.33, P = 0.015, 0.034, respectively）. Conclusion Quercetin may improve AD-like skin lesions, and its mechanisms may involve synergistic effects through regulating the balance of serum Th1/Th2 inflammatory factors, repairing the skin barrier function, and modulating gut microbiota homeostasis.

Key words: Dermatitis, atopic, Quercetin, Models, animal, Gut microbiota, Th1/Th2 balance, Inflammatory factors