Abstract Background Lianling Tang (LLT), a traditional Chinese herbal decoction, is utilized for managing inflammatory skin conditions in China. This study explores the therapeutic effects and molecular mechanisms of LLT on atopic dermatitis (AD) in a mouse model, employing proteomics techniques. Methods We evaluated the therapeutic efficacy of LLT on AD in mice by assessing skin inflammation scores, conducting histopathological examinations, and measuring serum levels of IgE, IL-4, and IL-6. Using Tandem Mass Tag (TMT) technology, we performed proteomics analysis of skin samples from Control, Model, LLT-treated, and positive control drug-treated mice, and identified differentially expressed proteins (DEPs). These proteins were then analyzed via GO and KEGG pathways to elucidate LLT's mechanisms in treating AD. Results LLT markedly reduced skin inflammation scores, epidermal thickening, and mast cell infiltration, and decreased serum IgE, IL-4, and IL-6 levels in AD model mice, highlighting its therapeutic potential. Proteomics analysis identified Staphylococcus aureus infection as a pivotal AD pathogenesis mechanism and revealed critical proteins such as C1QA, C4B, C5, and FCGR3. LLT's mechanisms, such as modulating the complement activation, and influencing phagocytosis, effectively reduce inflammation and alleviate symptoms caused by Staphylococcus aureus in AD. Conclusion LLT shows potential in AD treatment by modulating pathways associated with Staphylococcus aureus infection, effectively alleviating symptoms. Keywords: Traditional Chinese medicine, Atopic dermatitis, Proteomics, Staphylococcus aureus infection, Complement activation, Phagocytosis 1. Introduction Atopic dermatitis (AD), a prevalent inflammatory dermatological condition, manifests through severe itching, repetitive eczematous eruptions, and xerosis [[29]1]. This disorder frequently initiates in childhood and persists into adulthood for many individuals. In pediatric populations, AD's prevalence spans from 15 % to 20 % [[30]2], whereas in adults, it approaches 10.2 % [[31]3]. According to the Global Burden of Disease Study, AD ranks as the 15th leading non-lethal disease globally, underscoring it as the foremost burdensome dermatological condition with substantial impacts on patients' life quality and psychological well-being [[32]4]. Present therapeutic strategies, including topical corticosteroids, immunosuppressants like Janus kinase inhibitors, interleukin inhibitors (IL-13 and IL-31), and biological therapies like dupilumab, demonstrate partial success [[33]5,[34]6]. Nonetheless, concerns about their prolonged safety, and efficacy, coupled with potential side effects continue. Therefore, developing therapies that target the fundamental pathophysiology of AD remains a critical priority [[35][7], [36][8], [37][9]]. Contemporary studies indicate that AD's pathogenesis is intricately linked to immune dysfunction, impaired skin barrier integrity, and a disrupted microbial ecosystem [[38]10]. Notably, in AD-affected individuals, there's a significant uptick in colonization by pathogenic Staphylococcus aureus (S. aureus) [[39]11]. This bacterium compromises the skin barrier by releasing virulence factors, triggering immune reactions and cytokine production, which in turn amplifies inflammatory responses [[40]12]. Additionally, post-S. aureus skin infection, the bacteria lodge within keratinocytes [[41]13]. They activate the complement and phagocytic systems, forming terminal complement complexes on the cellular surface, thereby exacerbating the inflammatory response [[42]14]. This complex interaction between the skin's microbiome and the immune system underlines the multifaceted nature of AD. It also emphasizes the challenge of developing treatments that comprehensively target the root causes of the disease. The integration of Traditional Chinese Medicine (TCM) into therapeutic strategies for AD reveals promising avenues, particularly through its holistic action on multifarious pathways and biological targets. Clinically, TCM formulations have consistently proven effective in mitigating AD symptoms, notably diminishing dependency on corticosteroid therapies [[43]15,[44]16]. Among these, Lianling Tang (LLT) stands out as a pivotal herbal concoction, reputed for its capacity to address inflammatory dermatological conditions, including AD [[45]17]. Despite its recognized potential and traditional use, research into LLT's biological impact and specific mechanisms against AD has been limited. To address this gap, the current investigation leverages proteomics technologies to dissect the nuanced therapeutic effects and underlying actions of LLT on AD in model mice. This approach promises a complementary and potentially superior treatment paradigm, aiming to significantly enhance patient outcomes by providing a scientifically substantiated alternative to traditional steroid-based treatments. 2. Materials and methods 2.1. LLT preparation LLT, acquired from Dongfang Hospital affiliated with the Beijing University of Chinese Medicine, encompasses components such as Rhizoma Coptidis, Poria, Cornu Bubali, Cortex Moutan, Atractylodes, Sophora flavescens, Kochia scoparia, Lonicera japonica, and Curcuma aromatica. These ingredients were meticulously blended in a precise 2: 5: 5: 5: 5: 4: 5: 5: 4 ratio, with each herb's authenticity verified through The World Flora Online. In the preparation phase, the herbs were initially soaked in water to quintuple their mass, attended by boiling for 1 h. The first extraction was then separated through filtration, while the residual herbs underwent a second round of boiling in water amounting to 2.5 times their weight to maximize extract yield. Post filtration, the combined extracts were concentrated via rotary evaporation and subsequently freeze-dried, producing a fine, stable powder stored at 4 °C. For experimental application, the powder was dissolved in distilled water to create an LLT solution with a concentration of 4.11 g/kg, ready for administration to the model mice. This meticulous preparation method guarantees the integrity and efficacy of the LLT for therapeutic evaluation in the study. 2.2. UPLC-MS/MS sample preparation Compound identification can be affected by the timing and methods of sample collection, storage, and processing. Therefore, carefully considering the sample preparation process in ultra-performance liquid chromatography coupled with tandem mass spectrometry (UPLC-MS/MS) analysis is crucial to ensure data stability and reproducibility. Proper sample preparation is essential for compatibility with UPLC-MS/MS, providing reliable analysis of LLT's bioavailability and in vivo absorption. For UPLC-MS/MS sample preparation, a 0.2 ml dose of the extract at 4.11 g/kg was orally administered to a healthy mouse. One hour after administration, the mouse was anesthetized to facilitate blood collection from the posterior orbital vein. The blood sample was then placed in a tube without anticoagulant agents and left at room temperature for an hour to allow coagulation. After coagulation, the sample underwent centrifugation at 3000 rpm for 15 min, enabling the separation of serum, which was subsequently transferred into a new, sterile centrifuge tube. The serum sample was stored at −80 °C, ready for subsequent UPLC-MS/MS analysis to identify LLT components in the serum. 2.3. UPLC-MS/MS serum component identification UPLC-MS/MS is widely used to analyze pharmaceutical compounds from both traditional Chinese and Western medicines, including their metabolites and endogenous substances. We employed UPLC-MS/MS to analyze serum compounds after LLT treatment, aiming to precisely identify the components affected by LLT. This analysis helps identify which LLT compounds are systemically absorbed and potentially linked to their therapeutic effects. The qualitative assessment of LLT was performed utilizing an advanced UPLC setup (NexeraX 2, Shimadzu, Japan) in conjunction with a tandem mass spectrometry (MS/MS) system (6500 QTRAP, Applied Biosystems, MA, USA). This analysis was carried out using an Agilent chromatography column, maintained at ambient temperature, featuring a 1.8 μm particle size, a 2.1 mm diameter, and a 100 mm length. The chromatographic operation harnessed mobile phases A and B; phase A comprised pure water enhanced with 0.1 % formic acid, while phase B consisted of acetonitrile containing 0.1 % formic acid, propelled at a stable flow rate of 0.35 ml/min. The gradient program commenced with a 5 % concentration of phase B, incrementally raised to 95 % across 9 min, and held steady for an additional minute. Subsequently, from 10 to 11.10 min, the concentration of phase B was dialed back to 5 %, where it remained for the duration until the 14-min mark. The column temperature was regulated at 40 °C, with a precise injection volume of 2 μL. The mass spectrometry parameters included a turbo spray ion source heated to 550 °C, with the spray voltage set to 5500 V for positive ion mode and −4500 V for negative ion mode. Gas pressures were configured to 50 psi for gas 1 (GSI), 60 psi for gas 2 (GSII), and 25 psi for the curtain gas (CUR), ensuring optimal operational conditions. 2.4. Animals and groups Female BALB/c mice, aged 8 weeks and weighing 18–20 g, were acquired from Beijing Sibeifu Biotechnology Co., Ltd. (Beijing, China). Before the study, these mice underwent a 7-day acclimatization to familiarize themselves with the laboratory environment. Following this period, the mice were divided into four groups: The Control group, the Model group, the Positive drug group (treated with prednisolone (PD) at 8 mg/kg/day), and the LLT group (treated with LLT at 4.11 g/kg/day). During the experiment, animals were maintained in a controlled habitat, ensuring a constant temperature of 22 ± 2 °C, humidity levels at 55 ± 2 %, and a consistent 12-h light-dark cycle. Moreover, they had unrestricted access to a standard diet and water. The experimental procedures received approval by the Animal Ethics Committee of the Beijing University of Chinese Medicine, affirming our commitment to ethical research. 2.5. AD model and treatment Induction of the AD model began with the removal of dorsal fur from the mice one day before the experiment began. During the sensitization phase, on days one and four, 100 μL of a 2 % 2,4-dinitrochlorobenzene (DNCB) solution was applied topically to the backs of the mice. The DNCB solution, prepared in a 3:1 ratio with acetone and olive oil, was designed to provoke a DNCB-specific immune reaction. In the subsequent elicitation phase from day nine to the twenty-two, 100 μL of 0.5 % DNCB solution was alternately applied to the back and 20 μL to the right ear of each mouse to reinforce the immune response. During this phase, mice assigned to treatment groups received oral LLT at 4.11 g/kg/day or Prednisolone at 8 mg/kg/day, administered at a volume of 0.2 ml per 20 g of body weight. Conversely, mice in the Control and Model groups were given an equivalent volume of saline via gavage across 14 days. For control integrity, mice in the Control group were treated with a similar volume of the acetone and olive oil mixture without DNCB ([46]Fig. 2A). Fig. 2. [47]Fig. 2 [48]Open in a new tab AD Animal Model and LLT Treatment. (A) Schematic of the animal experiment. Before the start of the experiment, mice underwent 7 days of acclimatization. On day 0 of the experiment, the mice were shaved and then sensitized on the back skin with a 2 % DNCB solution on days 1 and 4. From day 9 to day 22, we challenged the mice with a 0.5 % DNCB solution on their back and right ear every other day. During this period, we also administered LLT or a positive drug daily via gavage. On day 23, the mice were euthanized. (B) Clinical scoring of lesion characteristics in AD mice, with severity scores graded from 0 to 3 (representing none, mild, moderate, and severe symptoms). (C) Photographs of the back skin lesion characteristics in mice, including erythema, edema, excoriation, and epidermal peeling. Data are presented as mean ± SD (n = 6). Compared to normal mice (Control), ^###p < 0.001; compared to model mice (Model), ∗∗∗p < 0.001. 2.6. Histopathology of skin tissue To assess the microstructural alterations in the epidermis and dermis comprehensively, this study conducted a detailed histological examination. Skin specimens were initially preserved in a 4 % paraformaldehyde solution (Biosharp, Shanghai, China) for 48 h. This step was crucial to maintain the structural integrity of the samples and facilitate subsequent processing. Following fixation, the specimens were encased in paraffin and sectioned to prepare for staining, which elucidates tissue architecture and cellular details. Employing Hematoxylin and Eosin (H&E) alongside Toluidine Blue (TB) staining methods not only revealed the overarching tissue structures but also pinpointed specific cellular entities. These staining techniques were instrumental in identifying and evaluating changes within the skin tissue. 2.7. ELISA assay In the ELISA assay, blood samples were collected from the experimental animals using the retro-orbital blood collection technique. The samples were then centrifuged at 3000 rpm for 15 min to separate the serum components. Following this, a quantitative analysis of specific biomarkers in the serum was conducted according to the ELISA kit's instructions. 2.8. Tandem mass tag proteomics analysis 2.8.1. Protein pretreatment for proteomics Proteomic preprocessing is essential to ensure consistent handling and high-quality protein extraction across all samples, thereby guaranteeing reliable data and accurate analysis results. For proteomics analysis, skin tissue samples, from each group—Control, Model, LLT, and PD were processed to extract proteins using the Total Protein Extraction Kit for Mammalian Tissues (AP0601-50, Beijing Bangfei Bioscience Co., Ltd, Beijing, China). Three samples from each group were selected. Following extraction, protein concentrations were quantified using the BCA Protein Assay Kit, strictly following the manufacturer's instructions. This step was crucial for determining the exact protein amount necessary for subsequent analysis. For electrophoretic separation, 20 μg of protein from each sample was subjected to SDS-PAGE. The electrophoresis was conducted at a constant voltage of 120V for 60–90 min to ensure optimal separation of proteins. After electrophoresis, proteins were visualized using Coomassie Brilliant Blue staining. A digital gel imaging system was utilized to evaluate the purity and concentration of the proteins, marking the completion of the preparatory phase for proteomics analysis. 2.8.2. Filter-aided sample preparation and TMT labeling After protein samples are prepared, they undergo enzymatic digestion and Tandem Mass Tag (TMT) labeling for quantitative proteomic analysis. The preparation process began with the reduction of proteins using 10 mM DTT at 37 °C for an hour, followed by alkylation with 50 mM IAA in darkness at room temperature for another hour. This step was crucial for breaking disulfide bonds and stabilizing the proteins for digestion. Subsequently, contaminants such as detergents, DTT, and low molecular weight impurities were removed using an 8M urea UA buffer, and filtered through a 10kD ultrafiltration membrane. The filtration process involved thorough washing, twice with 100 μl UA buffer and then thrice with 100 μl of 0.5M TEAB buffer, ensuring the removal of any residual contaminants. Proteins were then digested at 37 °C for 12–16 h using an enzyme-to-protein ratio of 1:50, yielding digested peptides ready for the next phase. The resultant peptides were purified of salts and other contaminants using a C18 Cartridge, a critical step to ensure their suitability for labeling. After purification, TMT labeling was performed as per Thermo Fisher Scientific's protocol, enabling the peptides from each sample to be uniquely marked for quantitative proteomics analysis. 2.8.3. Reversed-phase chromatographic separation Reversed-phase high-performance liquid chromatography (RP-HPLC) is employed to separate the TMT-labeled peptides, enhancing detection efficiency and ensuring peptide purity for subsequent mass spectrometric analysis. The separation process was conducted using a high-performance liquid chromatography (HPLC) system. Peptide samples were diluted in buffer A (2 % acetonitrile at pH 10.0) and loaded onto an xBridge BEH 130C18 column, known for its efficiency in peptide separation. Following loading, peptides were eluted using buffer B (98 % acetonitrile at pH 10.0), ensuring efficient peptide collection. The flow rate was maintained at 0.7 ml/min to optimize the separation and recovery of the peptides. After elution, the peptides were subjected to vacuum centrifugation at 45 °C, effectively removing solvent and yielding dry peptides ready for mass spectrometry analysis. 2.8.4. Easy nLC Easy nanoLC-MS/MS is primarily utilized for the preliminary separation of samples, providing a cleaner sample input for subsequent mass spectrometric analysis through efficient peptide separation, ensuring comprehensive resolution of complex samples. Peptide samples were analyzed using a nanoLC-MS/MS system with a Thermo Fisher Scientific C18 reversed-phase analytical column. Separation was achieved with a 0.1 % formic acid water solution (buffer A) and elution through a linear gradient of buffer B (0.08 % formic acid in 80 % acetonitrile) at 600 nL/min. The gradient began with 94 % A, transitioning to 91 % A at 0–2 min, 87 % A at 2–10 min, 74 % A at 10–50 min, and concluding with a steep gradient to 0 % A from 71 to 78 min. 2.8.5. LC-MS/MS analysis LC-MS/MS analysis is used for mass spectrometric detection and data collection, optimized for peptide separation to ensure high resolution and sensitivity. This technique offers precise and comprehensive quantitative and identification data for complex proteomic samples, enabling the detection of protein expression levels and modification states. In this study, LC-MS/MS analysis was performed using an Orbitrap Fusion mass spectrometer (Thermo Fisher Scientific) coupled to an Ultimate 3000 (Thermo Fisher Scientific) for 78min. The mass spectrometer operated in positive ion mode, adopting a data-dependent top10 strategy for the dynamic selection of the most abundant precursor ions within the m/z range of 350–1400 for MS data acquisition. The primary scan was configured to a resolution of 120,000, accompanied by an Automatic Gain Control (AGC) target of 5e5 and a maximum injection time (IT) of 100 ms. Concurrently, the MS2 scan resolution was set to 50,000, with a maximum IT of 86 ms, an isolation window precisely of 1.6 m/z, and a collision energy of 32eV. 2.8.6. Data analysis MS/MS raw files were processed using the MASCOT engine (Matrix Science, London, UK; version 2.6) embedded into Proteome Discoverer 2.5, and searched against the UniProt-UP000000589_10090.fasta database. Search parameters permitted up to 2 missed cleavages, using trypsin to generate peptides. A precursor mass tolerance of 10 ppm was specified and 0.02 Da tolerance for MS2 fragments. Except for TMT labels, Carbamidomethyl (C) was set as a fixed modification, while Oxidation (M) and Acetyl (Protein N-term) were set as variable modifications. A peptide and protein false discovery rate of 1 % was enforced using a reverse database search strategy. Proteins with a fold change >1.5 and p-value <0.05 (Student's t-test) were considered differentially expressed proteins. Initially, all protein sequences were aligned to the database downloaded from NCBI (ncbi-blast-2.2.28+-win32.exe), retaining only the sequences in the top 10 with E-value ≤ 1e-3. The GO term (database version: go_201504.obo) of the sequence with the top Bit-Score by Blast2GO was selected. The annotation from GO terms to proteins was completed by Blast2GO Command Line. After the elementary annotation, InterProScan was used to search the EBI database by motif, adding the functional information of motifs to proteins to improve annotation. Further improvement of annotation and connection between GO terms was carried out by ANNEX. Fisher's Exact Test was used to enrich GO terms by comparing the number of differentially expressed proteins and total proteins correlated to GO terms. Potential targets influenced by LLT in the treatment of AD were analyzed for gene/protein interactions using the STRING 12.0 database. The resulting Protein-Protein Interaction (PPI) network was visualized using Cytoscape 3.9.1 software. We employed the Molecular Complex Detection (MCODE) plugin to identify critical subnetworks and genes within the network, focusing on the connections between edges and nodes to determine the most influential components. Pathway analysis was performed using the KEGG database. Fisher's Exact Test was used to identify the significantly enriched pathways by comparing the number of differentially expressed proteins and total proteins correlated to pathways. 2.9. Statistical analysis The statistical analysis results are displayed as means ± standard deviations (SD). Data were analyzed and visualized using GraphPad Prism version 10.1.1. One-way analysis of variance (ANOVA) with Dunnett's adjustment was utilized to evaluate statistical significance, considering results with a p-value less than 0.05 as indicative of statistically significant differences. 3. Results 3.1. Analysis of LLT serum components Utilizing UPLC-MS/MS analysis, this study identified the primary serum components of LLT. As depicted in [49]Fig. 1 and elaborated in Supplementary Table Sheet 1, the study identified 125 compounds by matching with the database, performing accurate mass measurements, applying cleavage rules, analyzing MS-MS data, and observing chromatographic behavior. These compounds comprised 29 amino acids and their derivatives, 23 alkaloids, 20 lipids, 11 nucleotides and their derivatives, 11 phenolic acids, 9 organic acids, 4 flavonoids, 4 terpenes, and 14 compounds of various categories. Fig. 1. [50]Fig. 1 [51]Open in a new tab The total ion chromatograms obtained in (A) negative ionization mode and (B) positive ionization mode. 3.2. Effect of LLT on dermatitis scoring in AD mouse model This study aimed to evaluate LLT's therapeutic effect on a mouse model of AD. Experimental mice were divided into four groups: Control, AD model, LLT treatment, and Positive drug treatment. The severity of skin lesions was assessed using a dermatitis scoring system, ranging from 0 (no symptoms) to 3 (severe symptoms), based on erythema, edema, excoriation, and epidermal desquamation. Results indicated that mice in the Control group exhibited healthy skin without inflammation or damage. In contrast, mice in the AD model group showed significant dermatitis symptoms, including redness, swelling, epidermal peeling, and scab formation, which intensified with the number of DNCB applications. The dermatitis scores in the AD model group were significantly higher than those in the Control group (mean difference = 8.500, 95 % CI: 7.096 to 9.904, p < 0.001), successfully establishing the AD model. Mice in the LLT treatment group displayed significantly lower dermatitis scores than those in the Model group, particularly in erythema, scaling, and epidermal desquamation, indicating LLT's significant improvement in AD symptoms (mean difference = 4.000, 95 % CI: 2.596 to 5.404, p < 0.001, [52]Fig. 2B and C). The Positive drug treatment group also showed significant improvement, confirming the control drug's efficacy in treating AD (mean difference = 3.167, 95 % CI: 1.762 to 4.571, p < 0.001, [53]Fig. 2B and C). Overall, LLT demonstrated potential as a treatment for AD, providing scientific evidence of its effectiveness. 3.3. Effect of LLT on epidermal thickness and mast cell infiltration in the ear and back skin of AD mice Histological analysis of ear and back skin sections, conducted using Hematoxylin & Eosin (H&E) and toluidine blue staining, demonstrated augmented epidermal thickness ([54]Fig. 3A–C) and increased mast cell infiltration ([55]Fig. 3E–G) in the model group. Conversely, LLT administration markedly mitigated these pathological changes. Specifically, in comparison to the control group, the model group showed a 5.1-fold increase in ear epidermal thickness (mean difference = 55.27, 95 % CI: 45.36 to 65.18, p < 0.001, [56]Fig. 3B) and a 10.8-fold increase in back skin thickness (mean difference = 200.3, 95 % CI: 158.6 to 242.1, p < 0.001, [57]Fig. 3D). Moreover, the number of mast cells infiltrating the ears and back skin rose by 4.0-fold (mean difference = 19.00, 95 % CI: 13.20 to 24.80, p < 0.001, [58]Figs. 3F) and 5.3-fold (mean difference = 32.67, 95 % CI: 24.11 to 41.23, p < 0.001, [59]Fig. 3H), respectively. Remarkably, LLT treatment led to a significant reduction in epidermal thickening in both ears and back skin by 67.7 % (mean difference = 46.50, 95 % CI: 36.59 to 56.41, p < 0.01) and 70.4 % (mean difference = 155.6, 95 % CI: 113.8 to 197.3, p < 0.01), respectively, when compared to the DNCB-treated model group. Additionally, LLT treatment halved the number of infiltrated mast cells in both ear and back skin by 50.0 % (mean difference = 12.67, 95 % CI: 6.867 to 18.47, p < 0.01) and 50.4 % (mean difference = 20.33, 95 % CI: 11.77 to 28.89, p < 0.01), respectively. These findings underscore LLT's substantial therapeutic potential in mitigating skin lesions within the AD mouse model, thus supporting its clinical utility in treating AD. Fig. 3. [60]Fig. 3 [61]Open in a new tab Effect of LLT on Epidermal Thickness and Mast Cell Infiltration in DNCB-Treated Mouse Ears and Back Skin. Ear (A) and back skin (C) sections were stained with Hematoxylin & Eosin (H&E) and analyzed for epidermal thickness (n = 3, scale bar = 200 μm). Ear (E) and back skin (G) sections were stained with toluidine blue and analyzed for mast cell infiltration (n = 3, scale bar = 200 μm). Data are presented as mean ± SD (n = 3). Compared to normal mice (Control), ^###p < 0.001; compared to model mice (Model), ∗∗∗p < 0.001. 3.4. Effect of LLT on serum IgE, IL-4, and IL-6 levels in AD mice [62]Fig. 4A demonstrates that DNCB sensitization precipitated a notable 2.5-fold escalation in serum IgE levels among BALB/c mice. In contrast, mice subjected to LLT treatment exhibited a substantial decrement in serum IgE levels compared to those only exposed to DNCB (mean difference = 3.259, 95 % CI: 1.922 to 4.596, p < 0.001). Moreover, IL-4 and IL-6 levels in the serum witnessed a significant surge in mice sensitized with DNCB. LLT administration, however, efficaciously mitigated the elevation of IL-4 and IL-6 in the serum of DNCB-treated mice (mean difference = 22.77, 95 % CI: 6.419 to 39.17, p < 0.01 and mean difference = 34.26, 95 % CI: 6.959 to 61.57, p < 0.05, respectively), as depicted in [63]Fig. 4B and C. These insights signify that LLT treatment can significantly attenuate serum IgE levels and curtail the expression of the pro-inflammatory cytokines IL-4 and IL-6 in the AD mouse model, underscoring its therapeutic potential against AD. Fig. 4. [64]Fig. 4 [65]Open in a new tab LLT's effects on serum IgE and inflammatory mediators in DNCB-treated mice. LLT significantly affected the levels of (A) serum IgE, (B) serum IL-4, and (C) serum IL-6 in these mice. The data are expressed as mean ± standard deviation (SD) with n = 3. Compared to normal mice (Control), ^###p < 0.001; compared to model mice (Model), ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001. 3.5. Analysis of general characteristics of AD skin proteomics Volcano plot analysis revealed identified differentially expressed proteins (DEPs) between the LLT-treated and the model groups, highlighting 213 proteins as upregulated and 186 as downregulated ([66]Fig. 5A, Supplementary Table Sheet 2). Cluster analysis, utilizing significantly different proteins, showed a pronounced separation between the DNCB and LLT groups, demonstrating minimal crossover, and underscoring the distinctive impact of LLT on the AD mouse model ([67]Fig. 5B). Principal component analysis (PCA) illustrated a distinct division in skin proteome data between the control (pink triangles) and model groups (purple crosses), with the LLT-treated (green crosses) and PD groups (blue diamonds) displaying protein levels intermediate between them. These observations suggest that LLT or PD treatment aims to normalize protein expression levels ([68]Fig. 5C and D). Fig. 5. [69]Fig. 5 [70]Open in a new tab DEPs in proteomics, cluster analysis, and principal component analysis. (A) The volcano plot between the LLT and Model groups shows the differences in protein expression. The x-axis represents the log2 fold change, while the y-axis represents the log10 p-value for significance. Proteins that are significantly downregulated are indicated in green (FC ≤ 0.667 and p < 0.05), those that are significantly upregulated are in red (FC ≥ 1.5 and p < 0.05), and non-significant proteins are marked in black (0.667 < FC < 1.5 or p > 0.05). (B) Cluster analysis of the LLT and Model groups. The x-axis contains sample information from both groups, and the y-axis lists proteins with significant differences between groups. The expression levels of these significantly different proteins across samples are standardized using the Z-Score method and represented in the heatmap with various colors, where red indicates upregulation and green indicates downregulation. (C) Principal component analysis (PCA) of the Model and LLT groups. (D) Partial least squares discriminant analysis (PLS-DA) of the Model and LLT groups. 3.6. GO enrichment analysis of biological process, PPI network analysis, and KEGG pathway enrichment analysis GO functional enrichment analysis identified 2898 GO terms, of which 2217 (76.5 %) pertained to biological processes, 326 (11.3 %) to cellular components, and 355 (12.2 %) to molecular functions. The Biological Process (BP) classification of GO analysis plays a crucial role in understanding the biological activities associated with genes and proteins. It facilitates the identification of biological pathways that are activated or suppressed under specific pathological conditions, aiding in the discovery of potential biomarkers and therapeutic targets [[71]18,[72]19]. Consequently, our study concentrated on the BP classification to explore the effects of LLT on AD. The BP classification in GO and PPI network analysis of upregulated and downregulated targets elucidated significant biological functions influenced by LLT. Compared to the AD model group, proteins that were downregulated in the LLT-treated group are associated with immune response and defense response, including Complement activation, Phagocytosis, Inflammatory response, Regulation of cell adhesion, and Cell migration. In contrast, proteins upregulated by LLT treatment are involved in Skin development and Intermediate filament cytoskeleton organization, highlighting the potential mechanisms through which LLT mitigates the symptoms of AD ([73]Fig. 6A and B). Fig. 6. [74]Fig. 6 [75]Open in a new tab (A) GO enrichment analysis for Biological process: The x-axis represents the number of differentially expressed proteins, with the length of the bars reflecting the number of proteins in that GO category. Larger numbers result in longer bars. The gradient from light to dark represents the P-values, with darker shades indicating a higher significance of enrichment in that GO category. (B) PPI network analysis for LLT and AD Groups: Different colors represent various protein or gene functional categories, with proteins or genes of similar functions clustered together in the network. (C) KEGG pathway enrichment analysis: The x-axis displays the enrichment factor, with each bubble representing the number of differentially expressed proteins in that KEGG category. Larger numbers result in bigger bubbles. The color gradient reflects the P-values, with deeper shades of red indicating higher levels of significance in the enrichment of that KEGG category. The KEGG pathway enrichment analysis was conducted on the top 20 significant pathways from a total of 242 pathways, as depicted in [76]Fig. 6C. This analysis revealed significant enrichment of differentially expressed proteins across multiple pathways, particularly those associated with S. aureus infection, complement, and coagulation cascades, formation of neutrophil extracellular traps, estrogen signaling pathway, and Fc gamma R-mediated phagocytosis. The S. aureus infection pathway may play a central role in the LLT treatment mechanism for AD, as it is closely related to processes such as Skin development, Regulation of cell adhesion, Complement activation, Phagocytosis, and Cell migration. 3.7. Changes in protein expression related to S. aureus infection, LLT could reverse these changes S. aureus infection is broadly considered a central mechanism in AD pathogenesis. Proteomic results indicated that proteins linked to S. aureus infection, such as KRT10 (Keratin), exhibited a significant reduction in expression in the AD group, which was restored following LLT treatment. Concomitantly, the expression of FGG (Fibrinogen gamma chain), C1QA (Complement C1q subcomponent subunit A), and other proteins was significantly increased in the AD group but normalized post-LLT treatment ([77]Fig. 7). Fig. 7. [78]Fig. 7 [79]Open in a new tab Schematic diagram of LLT-regulated S. aureus infection pathway and the involved DEPs. Proteins that are significantly downregulated are indicated in green (FC ≤ 0.667 and p < 0.05), and those that are significantly upregulated are in red (FC ≥ 1.5 and p < 0.05). Specifically, the Skin development-associated protein KRT10 was downregulated in the AD group, while the Regulation of cell adhesion-associated protein FGG experienced an upregulation. However, these effects were reversed after LLT treatment ([80]Fig. 8A). Complement activation-associated proteins, including C1QA, C1QB (Complement C1q subcomponent subunit B), C1S (Complement C1s-1 subcomponent), C4B (Complement C4-B), C5 (Complement C5), BF (Complement factor B), MASP2 (Mannan-binding lectin serine protease 2), which were upregulated in the AD group, had their expression levels restored following LLT treatment ([81]Fig. 8B). Proteins involved in Phagocytosis and Cell migration, including FCGR3 (Fc-gamma RIII) and ITGB2 (Integrin beta), also displayed increased expression in the AD group, which was reversed following LLT treatment ([82]Fig. 8C). Fig. 8. [83]Fig. 8 [84]Open in a new tab Evaluating the therapeutic effects of LLT on AD through multi-pathway regulation of differential proteins to reduce S. aureus infection. (A) Levels of Skin development and Regulation of cell adhesion-related proteins in each group. (B) Levels of Complement activation-related proteins in each group. (C) Levels of Phagocytosis and Cell migration-related proteins in each group. Data are presented as mean ± SD (n = 3). Compared to normal mice (Control), ^#p < 0.05, ^##p < 0.01, ^###p < 0.001; compared to model mice (Model), ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001. 3.8. Verification of key proteins by ELISA Serum ELISA tests were conducted to validate key proteins identified in our proteomics analysis. The ELISA results were consistent with our initial findings. Specifically, proteins associated with complement activation, namely C1QA, C4B, and C5, were significantly upregulated in the AD model group. However, these levels were reduced to near control levels following treatment with LLT (mean difference = 29.68, 95 % CI: 9.030 to 50.32, p < 0.01, mean difference = 23.24, 95 % CI: 3.192 to 43.29, p < 0.05, and mean difference = 18.49, 95 % CI: 5.110 to 31.88, p < 0.01, respectively), as shown in [85]Fig. 9A, B, and C. Additionally, the level of the FCGR3 protein, which is associated with phagocytosis and cell migration, was significantly elevated in the AD group but was similarly reduced to near control levels in the LLT-treated group (mean difference = 152.0, 95 % CI: 118.8 to 185.2, p < 0.001), as depicted in [86]Fig. 9D. Fig. 9. [87]Fig. 9 [88]Open in a new tab Quantification of key proteins of ELISA in AD. LLT significantly affected the levels of (A) serum C1QA, (B) serum C4B, (C) serum C5 and (D) serum FCGR3 in these mice. The data are expressed as mean ± standard deviation (SD) with n = 6. Compared to normal mice (Control), ^#p < 0.05, ^##p < 0.01, ^###p < 0.001; compared to model mice (Model), ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001. 4. Discussion This study observed that LLT effectively reduced skin inflammation scores in AD model mice, ameliorated epidermal thickening and mast cell infiltration, and significantly decreased serum levels of IgE, IL-4, and IL-6, thus validating its efficacy in treating AD. Epidermal barrier damage triggers the activation of inflammatory dendritic epidermal cells and a Th2-dominated immune response. During this process, activated Th2 cells secrete IL-4 and IL-13, promoting the class switching of B cells to IgE [[89]20]. Additionally, the accumulation of inflammatory cytokine IL-6, produced by mast cells, is closely linked to the pathogenesis of AD [[90]21,[91]22]. Utilizing UPLC-MS technology, this study successfully identified 125 compounds in LLT. Subsequently, the study explored how LLT aids in alleviating AD symptoms through proteomic analysis. Analysis indicated that S. aureus infection constitutes a key mechanism, with C1QA, C4B, C5, FCGR3, identified as critical targets. Ultimately, it was found that LLT exerts its therapeutic effects by regulating the complement activation and affecting phagocytosis, thereby influencing inflammatory and immune responses. Using UPLC-MS/MS, we detected 125 compounds in the serum of mice after LLT administration, including 29 amino acids and their derivatives, 23 alkaloids, 20 lipids, 11 phenolic acids, 11 nucleotides and their derivatives, 9 organic acids, 4 flavonoids, 4 terpenoids, and 14 various other compounds. The primary active components identified were categorized into four groups: amino acids (such as L-proline and L-valine), alkaloids (such as betaine and L-carnitine), lipids (such as α-hydroxy linoleic acid), and phenolic acids (such as ferulic acid). Research by Rui Pang has highlighted that amino acid metabolites, particularly L-proline and L-valine, play a critical role in combating infections by inhibiting arginase activity [[92]23]. This inhibition enhances the phagocytic and bactericidal capabilities against S. aureus in both mouse and human blood, suggesting a novel therapeutic avenue for S. aureus infections. Additionally, α-hydroxy linoleic acid, a primary component of Gynostemma pentaphyllum, exhibits significant anti-inflammatory properties by regulating the expression of genes like TNF-α, Myd88, CD14, and TLR4 [[93]24]. Betaine has been shown to protect cells, exhibit antioxidant properties, assist in osmoregulation, and interfere with bacterial membrane functions, contributing to its anti-inflammatory and antibacterial activities [[94]25]. L-carnitine is effective in reducing inflammatory cytokine levels, such as TNF-α, IL-6, and IL-1β, and may help mitigate UVA-induced oxidative stress and inflammation, thus reducing skin damage in rats exposed to UVA [[95]26,[96]27]. The synergistic effects of betaine and L-carnitine could potentially enhance cytokine suppression, control inflammation, and boost resistance to bacterial infections [[97]28]. Ferulic acid, known for its antioxidant and anti-inflammatory properties, has demonstrated therapeutic effects in DNCB-induced AD models by lowering IgE and Th2 cytokine levels and inhibiting the NF-κB pathway, thereby alleviating inflammation [[98]29]. Overall, the protective effects of LLT on AD likely result from the antibacterial and anti-inflammatory actions of these compounds, enhanced by the synergistic effects of the natural herbs used in the formulation. Episodes of AD are frequently associated with S. aureus infections, which play a critical role in disrupting the balance of the skin's microbiome [[99]30]. Upon infecting the skin, S. aureus can persist within keratinocytes and activate the complement system on the cell surface through the formation of terminal complement complexes [[100]11]. This activation aids pathogen recognition, promotes chemotaxis of immune cells, or facilitates pathogen clearance through direct destruction [[101]31]. The complement system can be activated via the classical, lectin, or alternative pathways [[102]32]. Activation of the classical pathway begins with C1q recognizing antibodies or apoptotic cells bound to antigens or microbial surfaces [[103]33]. This activation leads to the activation of serine proteases C1r and C1s, which mediate the cleavage of C4, generating anaphylatoxin C4a and opsonin C4b. C4b then combines with C2, which C1 further cleaves, forming a C3 convertase, cleaving C3 into C3b and releasing C3a, while C3b can bind to adjacent cells or microbial surfaces, activating downstream complement. The lectin pathway is activated by lectins bound to pathogens and MBL-associated serine proteases (MASP1/2) [[104]34]. Alternatively, the alternative pathway can initiate the complement system even in the absence of antibody recognition [[105]35]. All these pathways lead to the cleavage of C3, producing C3a and C3b, with C3b promoting the formation of a C5 convertase that cleaves C5 into C5a and C5b, amplifying the complement response [[106]36]. Furthermore, prior research has indicated that increased complement components, complement cleavage products, and regulatory proteins may exacerbate inflammation and tissue damage in AD [[107][37], [108][38], [109][39]]. For example, C5a receptor inhibitors have been shown to markedly reduce skinfold thickness in AD mice and decrease the infiltration of leukocytes and mast cells within skin tissues [[110]40]. Such interventions also lower serum concentrations of IL-4, IFN-γ, histamine, and IgE. Furthermore, during inflammation, increased vascular permeability through various proteolytic events can be mitigated by using a C1 inhibitor, reducing serum protein accumulation in inflamed skin areas [[111]41]. Notably, Breathnach R. et al. have found an elevated conversion rate of complement C3 to C3b in dogs with AD [[112]42]. As such, inhibitors targeting the complement system are seen as promising therapeutic strategies for AD [[113]36,[114]40]. In this study, proteomics analysis identified significant upregulation of complement components such as C1QA, C1QB, C1S, C4B, C5, BF, and MASP2 in the AD model, corroborating prior research. ELISA results further confirmed elevated serum levels of C1QA, C4B, and C5 in the AD model mice. Remarkably, treatment with LLT significantly reduced these levels, suggesting that LLT may mitigate inflammatory responses and control S. aureus-induced infections by downregulating the expression and activity of these complement proteins, offering a potential strategy for treating AD. Phagocytosis is an essential process for eliminating microbes, apoptotic cells, and other foreign particles [[115]43]. It involves significant cytoskeletal rearrangement, crucial for tissue homeostasis and immune responses [[116]43]. Fc receptors (FcRs), particularly targeted in the immune system, play roles in conditions like cancer, inflammation, infections, and autoimmune diseases [[117]44]. Among these, Fcγ receptors (FcγRs) are the most diverse, interacting with IgG complexes to stimulate immune cells and trigger various cellular functions [[118]45]. Inhibiting Fcγ receptor-mediated phagocytosis signal has shown potential in alleviating lesions and reactive oxygen species (ROS) in AD [[119]46]. FCGR3, a receptor found on natural killer cells and macrophages, facilitates the immune system's ability to recognize and eliminate bacteria tagged with IgG via antibody-dependent cell-mediated phagocytosis and cytotoxicity [[120][47], [121][48], [122][49]]. In our study, the upregulation of FCGR3 protein expression in AD mice corroborates previous research findings. Notably, the observed effect was mitigated by the application of LLT. Therefore, LLT may provide therapeutic benefits for AD by modulating FCGR3-mediated phagocytosis, reducing inflammation caused by infections, and enhancing the body's ability to control S. aureus infections effectively. Further investigation is needed to confirm this theory. There are some limitations to our study. On one hand, our study employed only a single concentration of LLT. Although this was sufficient to demonstrate the potential efficacy of LLT, future studies should explore gradient concentrations to establish a detailed dose-response relationship. Such research would help identify the most effective dosage range and optimize therapeutic protocols, thereby enhancing the overall reliability and applicability of LLT as a treatment for AD. On the other hand, while our study analyzed the proteins downregulated by LLT compared to the model group, the mechanisms behind the proteins that were upregulated by LLT require further exploration. These upregulated proteins may represent potential therapeutic targets, making them valuable subjects for future research to deepen our understanding of LLT's mechanisms of action. 5. Conclusion Our study highlights the effectiveness of LLT in treating AD in a mouse model. Proteomic analysis revealed that LLT could mitigate DNCB-induced AD by downregulating the S. aureus infection pathway. This therapeutic effect appears to be associated with improvements in the regulation of complement activation and modulation of phagocytosis. Notably, proteins such as C1QA, C4B, C5, and FCGR3 have been identified as potential targets for AD treatment, suggesting that LLT may offer a promising approach to managing this condition. Availability of data and materials The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. Ethics statement The animal study was reviewed and approved by the Ethics Committee of the Beijing University of Traditional Chinese Medicine (BUCM-2024031905-1202). Consent for publication All authors consent to the publication of this work in Heliyon. Funding This work was supported by the State Administration of Traditional Chinese Medicine of China under the National Clinical Excellent Talents Advanced Study and Training Program (Traditional Chinese Medicine Education Letter [2022] No. 1), the National Natural Science Foundation of China (Grant No. 82274535), the Hebei Provincial Administration of Traditional Chinese Medicine Scientific Research Project (Grant No. B2025030), and the Qihuang Elite Talent · Famous Doctor Cultivation Program of Beijing University of Chinese Medicine (Grant No. Y2023A05). CRediT authorship contribution statement Lili Zhang: Writing – review & editing, Writing – original draft, Visualization, Validation, Software, Methodology, Conceptualization. Ziyi Fan: Investigation, Formal analysis. Linxian Li: Data curation. Zhanxue Sun: Writing – review & editing, Supervision, Resources, Project administration, Funding acquisition, Conceptualization. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgments