https://orcid.org/0000-0003-0967-3079
Figures
Abstract
Atopic dermatitis (AD), a chronic inflammatory disorder, poses significant therapeutic challenges owing to its complex pathophysiology, involving disrupted epidermal barrier function and immune dysregulation. This study investigated the therapeutic potential of Kochiae Fructus in AD treatment using bioinformatics, including network pharmacology and molecular docking techniques. We identified 19 key phytochemicals from Kochiae Fructus and 268 potential targets using the Traditional Chinese Medicine Systems Pharmacology Database (TCMSP) and SwissTarget Prediction. Using GeneCards, 1786 AD-related genes were retrieved, resulting in 116 intersecting gene targets for further analysis. Protein-protein interaction (PPI) networks and Molecular Complex Detection (MCODE) analyses highlighted 78 anti-AD key targets, including SRC, MAPK3, MAPK1, JUN, PIK3CA, ESR1, PTGS2, PTPN11, IL-6, and ALOX5, among the top ten anti-AD core targets. Gene ontology (GO) enrichment analysis revealed that Kochiae Fructus affects biological processes and molecular functions, such as positive regulation of the apoptotic response, inflammatory response, and hormone-mediated signaling pathways, which may be associated with its anti-AD effects. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis showed that the C-type lectin receptor signaling pathway is the main pathway involved in the anti-AD effects of Kochiae Fructus, which interacts with a notably larger number of anti-AD core targets and plays a direct role in intensifying crucial inflammatory and immune responses in the heart of AD pathogenesis. Molecular docking demonstrated robust binding affinities of key phytochemicals, particularly ecdysterone and 11,14-eicosadienoic acid, to the anti-AD core targets. Molecular dynamics simulations of over 1000 ns confirmed the stability and potential efficacy of these interactions. Hence, this study underscores the therapeutic potential of Kochiae Fructus in AD management, offering a mechanistic basis for its clinical application and paving the way for novel anti-AD strategies that leverage TCM phytochemicals.
Citation: Khan SA (2025) Network pharmacology and in silico analysis reveal Kochiae Fructus as a potential therapeutic against atopic dermatitis through immunomodulatory pathway interactions. PLoS ONE 20(4): e0320818. https://doi.org/10.1371/journal.pone.0320818
Editor: Chandrabose Selvaraj, Saveetha University - Poonamallee Campus: SIMATS Deemed University, India
Received: October 21, 2024; Accepted: February 24, 2025; Published: April 3, 2025
Copyright: © 2025 Shakeel Ahmad Khan. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: The author(s) received no specific funding for this work.
Competing interests: The authors have declared that no competing interests exist.
1. Introduction
Atopic dermatitis (AD) is a pervasive inflammatory dermatological condition characterized by compromised epidermal barrier function and diminished moisture retention capacity [1,2]. This pathology is characterized by clinical manifestations including pruritus [3], xerosis, fissuring [4], erythema [5], lichenification [6], and cutaneous inflammation [7], concomitant with keratinocyte apoptosis [8]. Epidemiological evidence delineates a progressive escalation in AD prevalence, affecting up to 20% of the pediatric population [9] and up to 10% of adults globally [10], positioning it as the 15th most common nonfatal dermatological disorder [1, 10–12]. AD exhibits a predilection for early onset, frequently identified during infancy and early childhood, with potential persistence into adulthood or an initial manifestation in later life stages. Notably, prevalence rates exhibit variability across demographic groups, influenced by racial and ethnic distinctions. In the United States, for instance, a higher incidence was observed in African American children (19.3%) than in their Caucasian counterparts (16.1%) [1, 12,13]. Similarly, in Hong Kong, the prevalence is approximately 20% across the lifespan, with a significant pediatric burden affecting approximately 30% of children [14–16]. The etiological attribution of the rising AD incidence has been tentatively linked to environmental factors, particularly in affluent and industrialized locales, where exposure to air pollutants and domestic hygiene products prevail [1]. This correlation necessitates a comprehensive etiological exploration to unravel the complexities of AD and to develop new targeted therapeutic and prophylactic strategies [17,18].
Pathogenetically, AD is ascribed to disruption of the integrity of the epidermal barrier, which facilitates allergenic penetration and subsequent immunologic activation [1]. This disruption triggers the activation of dendritic and innate lymphoid cells, which subsequently recruit and activate Th2 lymphocytes. The activation of these lymphocytes leads to the dysregulated secretion of pro-inflammatory cytokines, such as interleukin-4 (IL-4), interleukin-13 (IL-13), and interleukin-31 (IL-31), which precipitate the inflammatory cascade in AD through the Janus kinase (JAK) signaling pathways. This cascade further propagates inflammation and AD pathology by activating plasma cells and B lymphocytes, thereby inducing the production of antigen-specific immunoglobulin E (IgE) and amplifying the inflammatory response [1,18–21].
The therapeutic landscape for AD, devoid of a definitive cure, is predominantly anchored by the use of topical anti-inflammatory agents such as corticosteroids and calcineurin inhibitors [18,22]. While corticosteroids have proven effective in alleviating AD symptoms, their chronic use is marred by potential adverse effects, including irreversible skin atrophy and systemic repercussions, such as adrenal suppression [18,23]. Calcineurin inhibitors, such as tacrolimus and pimecrolimus, offer symptomatic relief by inhibiting the transcription of critical cytokines (IL-2, IL-4, and IL-5) involved in AD pathogenesis, albeit with lingering concerns regarding their oncogenic potential [18, 24–26]. In this therapeutic milieu, Traditional Chinese Medicine (TCM) has emerged as a noteworthy alternative that offers immunomodulatory benefits for chronic conditions with a favorable safety profile [18].
Kochiae Fructus, the desiccated fruit of Kochia scoparia (L.) Schrad., revered as a “top grade” herb in the seminal “Shennong Ben Cao Jing,” has been a staple in TCM for millennia [27,28]. Its therapeutic applications, documented in historical pharmacopeias such as the Herbal Canon and Compendium of Materia Medica [28,29], encompass a range of conditions, including eczema [28], dermatitis [29,30], pruritus [3], and psoriasis [31]. Phytochemical investigations have identified over 150 compounds in Kochiae Fructus, encompassing a diverse phytochemical spectrum with demonstrated pharmacological activities, including anticancer, antimicrobial, anti-edema, anti-allergic, anti-inflammatory, anti-dermatitic, anti-pruritogenic, and hypoglycemic effects [28].
Contemporary TCM practices often incorporate Kochiae Fructus into polyherbal formulations for AD treatment [32–34]. However, the specific phytochemicals and their mechanistic pathways in AD therapy remain elusive. In light of these gaps, the application of bioinformatics, a burgeoning field intertwining big data and artificial intelligence, holds promise for identifying active pharmaceutical ingredients and elucidating underlying mechanisms of drugs [35]. Thus, this study leveraged bioinformatics methodologies such as network pharmacology and molecular docking to decipher the biological phytochemicals, molecular targets, and pathways integral to the anti-AD effects of Kochiae Fructus. Fig 1 shows a schematic workflow of the current study.
2. Materials and methods
2.1. Software/databases
Network pharmacology for the phytochemicals of Kochiae Fructus, its common targets, and AD targets was accomplished using different web tools and software (Table 1).
2.2. Retrieval of phytochemicals from Kochiae Fructus
The phytochemicals of Kochiae Fructus were retrieved from the Traditional Chinese Medicine Systems Pharmacology Database (TCMSP, Version 2.3, available at https://old.tcmsp-e.com/tcmsp.php, accessed on March 10, 2024) [37]. A total of 19 phytochemicals were identified and are listed in S1 Table.
2.3. SwissTargetPrediction for phytochemicals of Kochiae Fructus
The SwissTargetPrediction database (http://www.swisstargetprediction.ch/, accessed on March 10, 2024) was used to ascertain the potential biological targets of Kochiae Fructus phytochemicals [40]. Only targets with a probability score of> zero were earmarked for further examination.
2.4. Identification of atopic dermatitis-related genes
Genes implicated in AD were extracted from GeneCards®: The Human Gene Database (accessible at https://www.genecards.org/, accessed on March 10, 2024) using search terms such as “atopic dermatitis,” “atopic dermatitis inhibiting,” and “anti-atopic dermatitis” [38]. Only AD-related genes with a relevance score of ≥ 5 were earmarked for further analysis [52].
2.5. Delineation of intersection gene targets
The confluence of gene targets implicated in both the potential targets of Kochiae Fructus’s phytochemicals and the identified AD-related genes was ascertained using Venny 2.1, an online software tool (https://bioinfogp.cnb.csic.es/tools/venny/; accessed on March 10, 2024) [39].
2.6. Protein-protein interaction (PPI) analysis
Protein-protein interaction networks for the intersecting gene targets were generated using the STRING database (https://string-db.org/, version 12.0, accessed on March 10, 2024), with the highest confidence score threshold (0.900), restricted to Homo sapiens, [18,41]and the disconnected nodes were removed from the network. Subsequently, the PPI network from STRING in the form of tab-separated values (tsv.) file format was imported into Cytoscape software (version 3.10.1) [42]. Subsequent topological analyses of network mapping were performed using the analyze network option in the Cytoscape software tool, which facilitated the identification of anti-AD core targets based on their degree centrality (DC) values. The identified anti-AD core targets were further verified using the CytoHubba plug-in in the Cytoscape software, and the top 10 anti-AD core targets with the largest DC values were ranked out and further used for subsequent molecular docking.
2.7. Molecular Complex Detection (MCODE) analysis
The critical modules in the PPI network of 78 potential anti-AD key targets were identified using the Cystoscope MCODE plug-in [42]. The conditions for MCODE analysis were as follows: find clusters in the whole network, degree cutoff = 2, node score cutoff = 0.2, k-core = 2, and Maximum Depth = 100 [53].
2.8. Network construction between AD-related genes and phytochemicals
An integrated network analysis encompassing intersecting AD-related genes and phytochemicals of Kochiae Fructus was conducted using the Cytoscape software (version 3.10.1) [42] to elucidate the complex interplay and potential therapeutic pathways.
2.9. Enrichment analysis
Gene Ontology (GO) functional annotation and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses were performed on 78 intersecting gene targets using DAVID (version 6.8, available at https://david.ncifcrf.gov/, accessed on March 10, 2024) [54,55]. GO terms were categorized into cellular components (CC), biological processes (BP), and molecular functions (MF). The top-ranked terms and pathways were visualized as an enrichment dot bubble chart, employing a bioinformatics platform for a comprehensive presentation. The classical hypergeometric test, adjusted for multiple hypothesis testing using the Benjamini–Hochberg method, was applied to ascertain statistical significance, with an adjusted p-value threshold of less than 0.05 [18].
2.10. Molecular docking
Maestro 13.5, from the Schrödinger Suite 2023 version 1, was utilized for the preparation of proteins and ligands [56]. Glide within the same suite was used for molecular docking, whereas Desmond-2020 version 1 facilitated molecular dynamic (MD) simulation studies [51,56].
2.10.1. Anti-AD core targets: preparation.
The top ten potential anti-AD core targets (SRC, MAPK3, MAPK1, JUN, PIK3CA, ESR1, PTGS2, PTPN11, IL-6, and ALOX5), represented by PDB IDs (1O46, 4QTB, 4FUX, 1A02, 6VO7, 1XP1, 5KIR, 4RDD, 1IL6, and 3V92) respectively, were retrieved from the Protein Data Bank (PDB) database (https://www.rcsb.org/). These protein structures were imported into the Protein Preparation Wizard (PPW) of the Schrödinger Suite 2023-1 with several modifications [50]. These modifications include the addition of hydrogen, assignment of bond orders, optimization of hydrogen positions, and correction of side chains. Energy minimization was strategically performed to maintain the rigidity of the backbone atoms while allowing the flexibility of the side chains [57].
The procedure corrected missing structural elements, including loops and disulfide bonds, with the pH settings adjusted to 7.4 to mirror physiological conditions. The Prime module of the Schrödinger suite was employed to address missing side chains. Water molecules positioned further than 5 Å were removed and the metals were modeled with zero-order bonds. A secondary water removal step ensured that the remaining water molecules were at least 3 Å away from the hydrogen atoms of the non-water entities. The OPLS_2005 force field, alongside an optimization algorithm, was utilized for further energy minimization, with a heavy atom root mean square deviation (RMSD) threshold of 0.30 Å, ensuring the refinement process achieved the desired level of precision [51,57,58].
2.10.2. Ligand preparation.
The ligands MOL002038 (9E,12Z-octadecadienoic acid), MOL002211 (11,14-eicosadienoic acid), MOL001739 (Zoomaric acid), MOL002212 (Ecdysterone), MOL000263 (Oleanolic acid), MOL000675 (Oleic acid), and MOL002215 (Oleanic acid) were sourced from the TCMSP database in.mol2 file format [37]. The stabilization and preparation of these ligand structures were meticulously conducted using the LigPrep module of the Schrödinger suite 2023-1 [49]. During this process, 32 stereoisomers, various ionization states, and tautomeric forms were generated at pH 7.4 ± 2.0, for each ligand to ensure comprehensive representation. The preparation steps included the addition of hydrogen atoms, energy minimization, and the application of an appropriate force field (OPLS_2005) to refine their three-dimensional (3D) coordinates and optimize them for subsequent computational analyses [57,59].
2.10.3. Receptor grid generation.
The receptor grid generation tool within the Schrödinger Suite 2023-1 was employed to construct a receptor grid with grid resolution, center coordinates, and dimensions derived from this utility [48]. This process delineates the active site in proximity to the specified coordinates (x, y, z), with grid dimensions tailored to the active site ligands (reference ligands) of particular proteins. In cases where proteins lacked ligands in their active sites, the potential binding surfaces were predicted using the Sitemap function of the Schrödinger Suite 2023-1. This function not only predicts binding sites, but also provides a sitemap score, offering detailed insights into the binding site with clearly delineated regions. The utilization of the Sitemap results facilitated the identification of potential alternative binding sites or the validation of known binding sites. The grid details are listed in S2 Table.
2.10.4. Glide docking of prepared Anti-AD core targets and ligands.
The Glide module of the Schrödinger Suite 2023-1 was utilized to perform molecular docking between the ten pre-prepared anti-AD core targets and the seven key phytochemicals from Kochiae Fructus [48]. Glide facilitates three docking precision modes with rigid receptor docking: high-throughput virtual screening (HTVS), standard precision (SP), and extra precision (XP). For this study, extra precision (XP) docking was selected following the upload of the receptor grid file for the respective anti-AD core targets, with all additional parameters and scoring functions maintained at their default configurations. The parameter set for each docking execution included the number of generated poses and chosen sampling methods [60]. The initial docking computation yielded various ligand poses within the binding site, which were evaluated using the applied scoring function. During the docking process, the key conformations of the protein were kept constant, whereas the ligands were allowed to be flexible [57]. The docked complexes were further visualized using Maestro and PyMOL software [47], and 2D and 3D images were generated.
2.11. Molecular dynamic simulation studies
MD simulations were performed using Desmond 2020.1, from Schrödinger [51]. Two complexes, specifically MOL002212MAPK1 and MOL002211ESR1, were chosen for MD simulation analysis due to their higher binding affinity and immune modulation roles, respectively. The OPLS-2005 force field and explicit solvent model employing SPC water molecules within a 10 × 10 × 10 Å periodic boundary solvation box were used [61]. Preprocessing, optimization, and minimization of protein-ligand complexes were performed using a protein preparation wizard. The system was constructed using a system builder tool [57]. Equilibration and minimization in the NPT (Nuclear Non-Proliferation Treaty) ensemble for 12 ns. The temperature and pressure were maintained at 310 K and 1 atm, respectively, using the Nose–Hoover chain coupling scheme [62]. Long-range electrostatic interactions were calculated using the particle mesh Ewald method with a 9 Å coulomb interaction cut-off. The simulation employed a 2-fs time step and the Martyna–Tuckerman–Klein chain coupling scheme for pressure regulation [57,63]. A 1000 ns simulation run was completed to ensure the stability of the protein-ligand complexes, monitored through metrics such as root mean square deviation (RMSD), root mean square fluctuations (RMSF), interaction fractions, ligand-protein contacts, radius of gyration (rGyr), solvent-accessible surface area (SASA), and polar surface area (PSA) [64].
2.12. Molecular mechanics to determine free binding energies
Molecular mechanistic studies using the Generalized Born Solvent Accessibility model (MM-GBSA) were performed to calculate the free binding energies (ΔGBind) of the complexes (MOL002212MAPK1 and MOL002211ESR1). Prime MM-GBSA from Schrodinger Suite 2023-1 was used for performing MM-GBSA in the present work, since it integrates the energies of molecular mechanics (EMM) with the optimized force field (OPLS_2005) potential for liquid simulations. The VGSB solvent model was used in the present work because of its capability to produce better solvent-free energy with the solvent generalized born model compared to the chloroform and solvation models [65]. The binding free energy (ΔGBind) for specific proteins (MAPK1 and ESR1) and ligands (MOL002212 and MOL002211) was calculated using the formula,
where Gcomplex represents the energy of the receptor-ligand complex, Gprotein is the energy of the receptor, and Gligand is the energy of the unbound ligand.
3. Results
3.1. Identification of potential targets for phytochemicals in Kochiae Fructus
Potential targets of Kochiae Fructus phytochemicals were acquired using the SwissTargetPrediction online database. A total of 1248 potential targets of 19 phytochemicals were identified, with a probability score of > 0. After eliminating redundancy, 268 potential targets were obtained and further analyzed (Fig 2A).
(B) Relationship between Kochiae Fructus phytochemicals and AD-related gene targets as well as their intersecting gene targets. (C) STRING protein-protein interaction (PPI) network of potential 78 anti-AD key targets of Kochiae Fructus.
3.2. AD-related gene targets
A total of 3407 AD-related genes with a relevance score > 5 were obtained from the GeneCards database by searching the keywords “anti-atopic dermatitis,” “atopic dermatitis inhibiting,” and “atopic dermatitis.” After eliminating redundancy, 1786 gene targets were identified (Fig 2A) and investigated for further analysis.
3.3. Intersection gene targets analysis
The intersection gene targets were analyzed between the phytochemical potential targets and the AD-related genes using the Venny 2.1.0 web tool, and a total of 116 intersecting gene targets were identified and further analyzed (Fig 2B). Intersecting gene targets were considered as potential anti-AD key targets.
3.4. PPI network analysis of anti-AD key targets
The PPI network was constructed by loading potential 116 anti-AD key targets of Kochiae Fructus into the STRING database version 12.0, as shown in Fig 2C. The results demonstrated that the network comprised 78 nodes and 152 edges. Moreover, 38 nodes were disconnected and eliminated from the PPI network. The average node degree, average local clustering coefficient, and average PPI enrichment p-values were 2.62, 0.427, and p < 0.00001, respectively.
As shown in Fig 3A, Cytoscape analysis results demonstrated that the PPI network had 78 nodes and 152 edges, with a characteristic path length of 3.962 between all node pairs. Furthermore, the density, diameter, average number of neighbors, clustering coefficient, network heterogeneity, network centralization, and network radius are 0.070, 10, 4.462, 0.346, 0.816, 0.186, and 5, respectively. Thirteen nodes appeared to be disconnected from the parent network and were therefore eliminated. Finally, the 65 nodes mapped in the network were arranged based on their DC values. The color of each node changed from yellow to purple as the degree increased.
The 28 nodes that complied with the criterion of DC> average value (4.46) were further extracted and designated as potential anti-AD core targets of Kochiae Fructus (Fig 3A). The 28 potential anti-AD core targets ranked by their DC values are shown in Fig 3 (B) as a bar graph. Based on their DC values in the network, the top ten potential anti-AD core targets were SRC, MAPK3, MAPK1, JUN, PIK3CA, ESR1, PTGS2, PTPN11, IL-6, and ALOX5. The potential anti-AD core targets were further verified using the CytoHubba plug-in Cytoscape software, which ranked the same top eight targets based on their DC value in the network, as shown in Fig 3 (C) and 3 (D). Consequently, the top ten potential anti-AD core targets were chosen for molecular docking studies with the phytochemicals of Kochiae Fructus.
3.5. MCODE cluster analysis of the PPI network of anti-AD key targets
The PPI network, encompassing 78 potential anti-AD key targets, was subjected to an advanced cluster network analysis using the MCODE plug-in for Cytoscape. As depicted in Fig 3(E), this analysis revealed five distinct cluster networks within the anti-AD PPI landscape. Specifically, Cluster Network 1 comprised six nodes and 15 edges, achieving a score of 6, while Cluster Network 2 was characterized by ten nodes and twenty-six edges, yielding a score of 5.778. Conversely, Cluster Networks 3, 4, and 5 each manifested three nodes and three edges, each securing a score of 3.0.
Within Cluster Network 1, the nodes representing PTGS2, ALOX12, CYP2C19, PTGS1, ALOX5, and ALOX15 exhibited a high degree of interconnectivity with multiple gene targets, each attaining a DC value of 5.0. In the context of Cluster Network 2, the MAPK1, MAPK3, ESR1, PRKCA, MAPK14, JUN, ESR2, PRKCB, PRKCD, and EGFR nodes demonstrated extensive interconnectivity, with respective DC values of 8.0, 8.0, 6.0, 5.0, 5.0, 4.0, 4.0, and 2.0. Furthermore, in Cluster Network 3, PIK3CA, SYK, and SRC were interconnected with two additional gene targets, each with a DC value of 2.0. Similar connectivity patterns were observed in Clusters 4 and 5, where RXRA, NR1H2, and NR1H3, as well as ALDH2, ADH1C, and MAOA, were interconnected with two other gene targets, maintaining a DC value of 2.0, for each node.
Significantly, Cluster Networks 1, 2, and 3 corroborated the identification of the top eight quintessential anti-AD core targets, namely SRC, MAPK3, MAPK1, JUN, PIK3CA, ESR1, PTGS2, and ALOX5, as illustrated in Fig 3(E). This coherence between the cluster and PPI network analyses substantiates the robustness and validity of the identified anti-AD key targets, underscoring their potential therapeutic relevance in the context of atopic dermatitis.
3.6. Network of Kochiae Fructus’s phytochemicals and anti-AD targets
A network map elucidating the interactions between Kochiae Fructus phytochemicals and 78 potential anti-AD key targets was constructed using the Cytoscape software, as shown in Fig 4 (A). This structured network map comprises 97 nodes and 387 edges, manifesting a network diameter and radius of four and two, respectively. The topological parameters included a network density of 0.079, heterogeneity coefficient of 1.088, and centralization index of 0.302. Additionally, network map analysis revealed an average number of neighbors of 7.567, a characteristic path length of 2.881, a clustering coefficient of 0.000, and a single connected component. Each edge within this network represents a biochemical interaction between the phytochemicals of Kochiae Fructus and the delineated anti-AD key targets. The degree of a node reflects its connectivity within the network, with the nodal color gradient transitioning from orange to purple to indicate varying degrees of node centrality, thus providing insights into the network’s structural complexity.
Fig 4(B) presents a hub network map delineated using Cytoscape, illustrating the molecular interactions between Kochiae Fructus phytochemicals and 28 potential anti-AD core targets (determined in Section 3.4). This intricate hub network map consisted of 44 nodes and 101 edges with network diameters, radii, and densities of 6, 4, and 0.144, respectively. Analysis of the hub network map demonstrated that all 16 Kochiae Fructus phytochemicals interacted with the 28 potential anti-AD core targets to varying degrees. Notably, MOL002038 (9E,12Z-octadecadienoic acid), MOL002211 (11,14-eicosadienoic acid), MOL001739 (Zoomaric acid), MOL002212 (Ecdysterone), MOL000263 (Oleanolic acid), MOL000675 (Oleic acid), and MOL002215 (Oleanic acid) are the top seven key phytochemicals, ranked by their DC> average value (6.3125), that interact with more than six potential anti-AD core targets in the hub network. These phytochemicals are acknowledged as key phytochemicals in Kochiae Fructus and are depicted in the form of a bar graph ranked by the DC in the hub network (Fig 4C). Further validation using the CytoHubba plugin in Cytoscape corroborated the prominence of these seven phytochemicals based on their DC values (Fig 4 (D)). Consequently, our results demonstrate that the synergistic mode of interaction exists between the multiple key phytochemicals present in Kochiae Fructus and multiple anti-AD core targets in the therapeutic milieu of AD.
3.7. GO enrichment analysis of Kochiae Fructus’s phytochemicals in the treatment of atopic dermatitis
The 78 potential anti-AD key targets implicated in AD were meticulously subjected to GO enrichment analysis. As depicted in Fig 5, the outcomes enumerate the top 10 enrichment terms across the molecular function (MF), cellular component (CC), and biological process (BP) categories pertinent to these targets. This analysis revealed that the anti-AD key targets of Kochiae Fructus’s phytochemicals within the BP category are intricately involved in an array of physiological pathways, including the hormone-mediated signaling pathway, response to xenobiotic stimuli, positive regulation of transcription, DNA-templated, intracellular signaling pathway, positive regulation of apoptotic response, inflammatory response, positive regulation of ERK1, and ERK2 cascade, etc. In the realm of CC, the analysis delineated that anti-AD key targets of Kochiae Fructus’s phytochemicals are predominantly localized within critical cellular structures such as the mitochondrion, nucleoplasm, cytosol, plasma membrane, cytoplasm, endoplasmic reticulum membrane, etc. This distribution suggests a comprehensive involvement of these anti-AD key targets of Kochiae Fructus phytochemicals in various cellular functions and processes that are pivotal in the pathophysiology of AD. Moreover, enrichment within the MF category was found to be dominated by functions such as RNA polymerase II transcription factor activity, ligand-activated sequence-specific DNA binding, enzyme binding, zinc ion binding, protein serine/threonine/tyrosine kinase activity, sequence-specific DNA binding, etc.
3.8. KEGG pathway analysis of Kochiae Fructus’s phytochemicals in the treatment of atopic dermatitis
KEGG pathway elucidation was performed to decipher the pharmacological mechanisms by which Kochiae Fructus phytochemicals exert ameliorative effects on AD. The uploading of 78 potential anti-AD key targets to the DAVID platform resulted in the identification of 138 statistically significant pathways (p < 0.05). Fig 6 shows the 30 most critical KEGG pathways ascertained through this analysis, which revealed that the anti-AD key targets of the Kochiae Fructus phytochemicals might be implicated in pathways in cancer, inflammatory mediator regulation of TRP channels, AGE-RAGE signaling pathway in diabetic complications, C-type Lectin Receptor Signaling Pathway, Th17 cell differentiation, chemical carcinogenesis-receptor activation, EGFR tyrosine kinase inhibitor resistance, proteoglycans in cancer, VEGF signaling pathway, lipid and atherosclerosis, T cell receptor signaling pathway, HIF-1 signaling pathway, and ErbB signaling pathway. This comprehensive analysis accentuates the complex and multifaceted pharmacological milieu wherein the phytochemicals of Kochiae Fructus ostensibly modulate a range of pathophysiological pathways that are implicated in AD. The elucidation of these pathways not only augments our understanding of the pharmacodynamic underpinnings of Kochiae Fructus’s therapeutic potential in AD treatment but also heralds the advent of innovative therapeutic modalities targeting these molecular pathways, thereby augmenting the precision and efficacy of AD management.
3.9. Core pathway analysis and determination
To elucidate the core pathways implicated in the anti-AD properties of the phytochemicals of Kochiae Fructus within the therapeutic milieu of atopic dermatitis, a network map was established by integrating the top ten anti-AD core targets and the top 30 KEGG pathways (Fig 7A). The results showed that the network map comprised 40 nodes and 148 edges, with a network diameter of 5, a radius of 3, and a density of 0.158. Network analysis indicated that all 30 pathways interacted with the ten anti-AD core targets at different DC values. The pathway with the highest degree (8.0) was the C-type lectin receptor signaling pathway, whereas arachidonic acid metabolism had the lowest degree (2.0). The color of the pathway nodes transitioned from pink to light pink depending on the degree.
Furthermore, the pathways in the network were ranked by DC> average value of (4.933), resulting in the identification of 19 core pathways, as illustrated in Fig 7 (B). Eight of the ten anti-AD core targets (JUN, SRC, PTPN11, PTGS2, IL6, PIK3CA, MAPK1, and MAPK3) followed the C-type lectin receptor signaling pathway. In contrast, seven out of ten anti-AD core targets (PTGS2, MAPK1, MAPK3, JUN, ESR1, IL6, and PIK3CA) followed the pathways in cancer. Consequently, this study suggests that these 19 core pathways may play a role in the anti-AD effects of Kochiae Fructus phytochemicals.
3.10. Molecular docking analysis
The top seven key phytochemicals of Kochiae Fructus were further analyzed for molecular docking with the top ten anti-AD core targets, including SRC, MAPK3, MAPK1, JUN, PIK3CA, ESR1, PTGS2, PTPN11, IL-6, and ALOX5. The results of the molecular docking analysis in terms of binding affinity (kcal/mol) scores are presented in Fig 8.
Each cell color indicates the strength of the binding energy, with cooler colors (blues) indicating stronger binding and warmer colors (reds) indicating weaker binding.
Furthermore, the nine docked complexes, specifically MOL000675SRC, MOL002212SRC, MOL002212MAPK3, MOL002212MAPK1, MOL002212PTPN11, MOL002212IL6, MOL002211JUN, MOL002211ESR1, and MOL002211PTGS2, demonstrated binding affinities exceeding -7.0 kcal/mol. These complexes are depicted in both 2D and 3D formats in Fig 9 (A-E) and Fig 10 (A-D). Notably, compared to other phytochemicals, ecdysterone (MOL002212) exhibited exceptional binding affinities (-7.9 ± 0.04, -8.3 ± 0.03, -5.4 ± 0.03, -7.5 ± 0.02, -7.1 ± 0.04, and -5.1 ± 0.03 kcal/mol) for six anti-AD core targets including MAPK3 (H-Bond: LYS71, GLN122, ASP123, SER170), MAPK1 (H-Bond: GLU31, GLN103, ASP104, LYS149, SER151, ASP165), PIK3CA (H-Bond; HIE160, SER161, GLU218, LYS264, LYS290), PTPN11 (H-Bond: TYR279, TRP423, PRO424, ASP425, ARG465), IL-6 (H-Bond: GLU60, LYS67, GLU173), and ALOX5 (H-Bond: TRP147, ARG411), respectively. Conversely, 11,14-eicosadienoic acid (MOL002211) demonstrated notably high affinities (-7.1 ± 0.03, -7.3 ± 0.02, and -7.4 ± 0.02 kcal/mol) for JUN (H-Bond: ARG282; Salt bridge: ARG537), ESR1 (H-Bond: ASN532, LEU536), and PTGS2 (H-Bond: ARG120; Salt bridge: ARG513) respectively. Furthermore, oleic acid (MOL000675) displayed the highest binding affinity of -7.6 ± 0.03 kcal/mol against SRC and interacted with ARG34, GLU37, and CYS 44 through three hydrogen bond interactions and with ARG14, and ARG34 through two salt bridges. In conclusion, ecdysterone (MOL002212) and 11,14-eicosadienoic acid (MOL002211) demonstrated the best binding affinities with several anti-AD core targets compared with the other five key phytochemicals of Kochiae Fructus. Therefore, docked complexes of these two phytochemicals were further evaluated using MD simulations.
3.11. Molecular dynamic simulation analysis
Two complexes, specifically MOL002212MAPK1 and MOL002211ESR1, were chosen for the MD simulation analysis owing to the higher binding affinity and immune modulation roles, respectively. The results for each complex are shown in Fig 11(A-H) and 12 (A-I). Molecular dynamics simulations of the MOL002212MAPK1 and MOL002211ESR1 complexes over a 1000 ns period demonstrated consistent and stable molecular conformations, as substantiated by root-mean-square deviation (RMSD) analyses. For MOL002212MAPK1, the RMSD associated with the Cα backbone exhibited an ascending trend until approximately 800 ns, subsequently achieving pronounced stabilization that persisted throughout the remainder of the simulation interval, as shown in Fig 11(A). The observed RMSD values for the MOL002212MAPK1 complex ranged between 1.5 and 3 Å, conforming impeccably to the established thresholds for acceptability, while the ligand RMSD varied from roughly 0.50 to 1.20 Å, indicative of substantial stability (Fig 11B). For MOL002211ESR1, an increase in RMSD was discernible until approximately 500 ns, after which a heightened level of stability was attained until the conclusion of the 1000 ns period. The RMSD values for MOL002211ESR1 fell within the 2–4 Å range, consistent with acceptability standards [57], and the ligand RMSD ranged from approximately 1 to 3.75 Å, suggesting high stability (Fig 12A and 12B).
(A) RMSD of the Cα backbone of MAPK1 bound to ligand MOL002212. (B) RMSD of ligand MOL002212 bound to MAPK1. (C) Cα backbone RMSF of MAPK1 bound to the ligand MOL002212. (D) Ligand MOL002212 RMSF bound to MAPK1. (E) Interaction factions of amino acid residues MAPK1 with the ligand MOL002212. (F) Two-dimensional (2D) ligand-protein interaction diagrams (MOL002212MAPK1). (G) Radius of gyration (rGyr) and (H) polar surface area (PSA) of MOL002212 bound to MAPK1.
(A) Cα backbone RMSD of ESR1 bound to the ligand MOL002211. (B) RMSD of ligand MOL002211 bound to ESR1. (C) Cα backbone RMSF of ESR1 bound to ligand MOL002211. (D) Ligand MOL002211 RMSF bound to ESR1. (E) Interaction of amino acid residues of ESR1 with ligand MOL002211. (F) Two-dimensional (2D) ligand-protein interaction diagrams (MOL002211ESR1). (G) Radius of gyration (rGyr), (H) polar surface area (PSA), and (I) solvent-accessible surface area (SASA) of MOL002211 bound to ESR1.
Variations in the protein root-mean-square fluctuation (RMSF) spanned from 0.5 to 2.75 Å for MOL002212MAPK1 and from approximately 0.75 to 5.0 Å for MOL002211ESR1, with both complexes experiencing numerous ligand interactions. Minimal fluctuations were primarily noted below the 1.5 Å level for MOL002212MAPK1 and below the 2.0 Å threshold for MOL002211ESR1, highlighting the relative stability of both complexes (Fig 11C and 12C). The spatial congruence between the ligands and proteins was assessed to be less than 5.0 Å for MOL002212MAPK1 and less than 4.0 Å for MOL002211ESR1, reflecting minimal alterations in ligand conformation throughout the simulations (Fig 11D and 12D).
Interaction analysis of both complexes revealed that hydrogen bonds and water bridges were predominant, with fewer hydrophobic and ionic interactions throughout the simulation. Notably, hydrogen bonds were critical for maintaining the binding conformation of both complexes. This was corroborated by two-dimensional (2D) ligand-protein interaction diagrams, which illustrated that significant residue interactions occurred upon binding. For MOL002212MAPK1, five amino acid residues (GLU31, GLN103, ASP104, MET106, and SER151) formed direct hydrogen bonds with small molecules, whereas ASP109 facilitated water-mediated hydrogen bonding, with hydrogen bond proportions of 55%, 68%, 108%, 61%, and 62%, respectively for the direct interactions (Fig 11F). For MOL002211ESR1, LYS529 and LEU536 formed direct hydrogen bonds, and GLU380 formed a water-mediated bond, with hydrogen bond proportions of 73% and 38% for the direct interactions (Fig 12F). Interestingly, coherence was observed in the results of residue interactions in the MD simulation analysis and molecular docking for both complexes.
Ligand properties such as the radius of gyration (rGyr), solvent accessible surface area (SASA), and polar surface area (PSA) further underscored the complex’s molecular dynamics, with MOL002212MAPK1 showing rGyr between 4.6 and 5.3 Å and PSA between 208 and 249 Ų, along with SASA ranging from 167.03 to 333.02 Ų (Fig 11G, 11H, and S1 Fig). The fluctuations in PSA observed for MOL002212MAPK1 (Fig 11H) might be influenced by interactions with key residues such as GLU31, ASP104, GLN103, and ASP109, which form hydrogen bonds and water-mediated interactions with the ligand. While MOL002211ESR1 exhibited rGyr between approximately 4.75 and 7.25 Å and PSA between approximately 78 and 94 Ų, along with SASA ranging from 1.50 to 190 Ų (Fig 12G, 12H, and 12I).
The SASA of MOL002211 bound to ESR1 (Fig 12I) revealed the dynamic conformational changes of the ligand–protein complex throughout the MD simulation. The marked fluctuations observed in the early phase (0–200 ns) signify structural rearrangements as the ligand establishes stable interactions with critical binding site residues. Such deviations are typical while the system adapts to achieve optimal binding. After 200 ns, the SASA values stabilize, indicating that the ligand is well-accommodated within the binding pocket and that the complex has reached a stable conformation. This stabilization underscores the adaptability of the ligand–protein interaction and confirms the reliability of the binding pose over the simulation period.
3.12. Molecular mechanics generalized Born surface area calculations
The stability of the MOL002212MAPK1 and MOL002211ESR1 complexes was further examined by determining their Gibbs binding energies (ΔGBind). The results revealed that the stability of these complexes was primarily driven by Coulomb interactions (ΔGBindCoulomb), van der Waals forces (ΔGBindvdW), and lipophilic interactions (ΔGBindLipo), as shown in Table 2. In contrast, covalent interactions (ΔGBindCovalent) and solvation-free energy contributions (ΔGBindSolvGB) were identified as factors contributing to the instability of these complexes. Notably, both complexes (MOL002212MAPK1 and MOL002211ESR1) exhibited significantly higher binding free energies, highlighting their robust and stable protein-ligand interactions. This suggests that the ligands in these complexes form highly effective and long-lasting binding interactions with their respective targets.
4. Discussion
The current research work investigated the key phytochemicals and elucidated the underlying mechanisms of Kochiae Fructus in treating AD disease. In all, nineteen molecules were retrieved from TCMSP, and the top seven key phytochemicals ranked by DC ≥ average value of (6.3125) interacting with more than six anti-AD core targets include MOL002038 (9E,12Z-octadecadienoic acid), MOL002211 (11,14-eicosadienoic acid), MOL001739 (Zoomaric acid), MOL002212 (Ecdysterone), MOL000263 (Oleanolic acid), MOL000675 (Oleic acid), and MOL002215 (Oleanic acid). The network results also showed that single key phytochemicals may interact with multiple anti-AD core targets and that multiple key phytochemicals can also interact synergistically with a single anti-AD core target. Thus, our findings indicate a synergistic mode of interaction between multiple anti-AD core targets and multiple key phytochemicals in Kochiae Fructus for treating AD.
PPI network analysis identified several key genes involved in the etiology of AD, including SRC, MAPK3, MAPK1, JUN, PIK3CA (PI3K), ESR1, PTGS2, PTPN11, IL-6, and ALOX (Fig 3). Lu et al. found a positive correlation between SRC expression and inflammatory progression in AD, suggesting that SRC is a potential target for AD therapy [66]. Additionally, IL-17 has been implicated in activating the MAPKs and P38/ERK MAPK pathways, crucial in immunopathology, cellular proliferation, and inflammatory dermatological disorders [67–69]. The AP-1 complex, activated by JUN, drives inflammation, making c-Jun/AP-1 a viable target for managing psoriatic skin inflammation via immune modulation [70,71]. The PI3K/Akt signaling pathway, regulated by PIK3CA, is pivotal for various cellular functions, including metabolism, growth, and survival. In AD, the aberrant activation of this pathway in peripheral T cells induces T cell proliferation and cytokine release (IL-6 and IL-10), thereby contributing to the pathology of AD [72]. ESR1 (ERα) activation, as shown by Iwano et al., leads to IL-23 secretion from dendritic cells, exacerbating inflammatory and psoriasis-like dermatitis in a mouse model [73]. Moreover, Niu et al. reported that ESR1 enhances the Th2-immune response and Th2 cytokine levels in AD mouse models [74]. Recent studies have revealed that achieving an equilibrium between IL-6 and IL-17 is crucial for bolstering the anti-inflammatory properties of IL-4. Upregulation of IL-6 stimulates IL-4 production, leading to the acute phase of AD and pain. Furthermore, they promote the proliferation of T-cells and B cells, resulting in the formation of Th-17 cells. Elevated IL-4 expression suppresses IgE production, causing induced skin inflammation [75–78]. In contrast, PTGS2 plays a role in maintaining homeostasis but has also been implicated in various clinical pathologies, including pain and inflammation [79,80]. Although the implication of PTPN11 in AD remains unclear, Wang et al. showed that PTPN11 (SHP2) activation by Trichomide A has posed immunosuppressive effects on activated T lymphocytes, thereby mitigating contact dermatitis [81]. ALOX5, a member of the lipoxygenase enzyme family, is implicated in converting polyunsaturated fatty acids into leukotrienes, is critical in allergic inflammatory responses, and is notably active in AD-related inflammation [82]. Hence, this study suggests that modulating the expression levels of these ten anti-AD core targets could mitigate inflammation and ameliorate the symptoms associated with AD.
GO enrichment analysis showed that the ameliorative effects of Kochiae Fructus phytochemicals on AD might be attributed to multiple gene targets implicated in multiple BP, such as positive regulation of apoptotic response, inflammatory response, hormone-mediated signaling pathway, response to xenobiotic stimuli, positive regulation of transcription, DNA-templated, intracellular signaling pathway, positive regulation of ERK1, and ERK2 cascade, etc. Reports demonstrate that inflammation and apoptosis significantly contribute to the pathogenesis of AD. Patients afflicted with AD demonstrate elevated Fas and FasL expression, which are integral components of the apoptosis pathway. Moreover, induction of apoptosis in keratinocytes triggers AD progression, leading to cellular death and compromising the integrity of the skin barrier. Furthermore, the exacerbation of apoptosis is potentiated by proinflammatory cytokines such as TNF-α and IFN-γ via the Fas/FasL pathway [83,84]. Hormone-mediated signaling pathways have been shown to profoundly influence the pathogenesis and clinical manifestation of AD, implicating stress responses and immune modulation in their effects. Insights into these pathways are pivotal for understanding AD mechanisms and developing targeted therapeutic interventions [85,86]. The ERK1 and ERK2 cascades integral to the mitogen-activated protein kinase (MAPK) pathway have been implicated in the generation of inflammatory cytokines in epidermal keratinocytes, which promotes the recruitment of immune cells such as T cells and dendritic cells, thereby intensifying the inflammatory milieu characteristic of AD [87]. Consequently, therapeutic strategies aimed at modulating the ERK1/2 cascade are viable approaches to mitigating AD. The anti-AD effects of Kochiae Fructus phytochemicals are presumably linked to an array of gene targets predominantly located in cellular compartments such as the mitochondrion, nucleoplasm, cytosol, plasma membrane, cytoplasm, and endoplasmic reticulum membrane. Enriched MF ontology analysis further reveals that these gene targets engage in diverse functions, encompassing RNA polymerase II transcription factor activity, ligand-activated sequence-specific DNA binding, enzyme binding, zinc ion binding, protein serine/threonine/tyrosine kinase activity, sequence-specific DNA binding, etc.
KEGG pathway enrichment analysis revealed that the anti-AD effects of the Kochiae Fructus phytochemicals could be associated with pathways in cancer, inflammatory mediator regulation of TRP channels, AGE-RAGE signaling pathway in diabetic complications, C-type lectin receptor signaling pathway, Th17 cell differentiation, EGFR tyrosine kinase inhibitor resistance, VEGF signaling pathway, lipid and atherosclerosis, T cell receptor signaling pathway, HIF-1 signaling pathway, and ErbB signaling pathway. Moreover, network analysis of the top 30 KEGG pathways and top ten anti-AD core targets identified 19 core pathways ranked by DC ≥ average value of (4.933), as shown in Fig 7 (B). Among them, C-type lectin receptor signaling pathway, pathways in cancer, Endocrine resistance, Estrogen signaling pathway, EGFR tyrosine kinase inhibitor, VEGF signaling pathway, T cell receptor signaling pathway, and ErbB signaling pathway are the core pathways followed by eight of the ten anti-AD core targets (JUN, SRC, PTPN11, PTGS2, IL6, PIK3CA, MAPK1, and MAPK3), seven of ten anti-AD core targets (PTGS2, MAPK1, MAPK3, JUN, ESR1, IL6, and PIK3CA), six of ten anti-AD core targets JUN, PIK3CA, SRC, MAPK1, ESR1, and MAPK3), six of ten anti-AD core targets (JUN, PIK3CA, SRC, MAPK1, ESR1, and MAPK3), five of ten anti-AD core targets (IL6, PIK3CA, SRC, MAPK1, and MAPK3), five of ten anti-AD core targets (PIK3CA, SRC, MAPK1, PTGS2, and MAPK3), five of ten anti-AD core targets (JUN, PIK3CA, MAPK1, PTPN11, and MAPK3), and five of ten anti-AD core targets (JUN, PIK3CA, SRC, MAPK1, and MAPK3), respectively.
Numerous studies have demonstrated the contribution of the C-type lectin receptor signaling pathway to AD pathogenesis, which exacerbates enhanced inflammatory (TNF-α, IL-6, and IL-1β) and Th2 immune responses [88]. Therefore, targeting the CLR signaling pathway may offer new therapeutic strategies for managing AD. The results showed that pathways in cancer regulation may contribute to the modulation of anti-AD targets. Targeting cancer-related pathways may also affect their expression in AD. Hyperactivation of the estrogen signaling pathway enhances the activities of Th2/regulatory T cells (Tregs) but suppresses Th1/Th17, leading to fortification of the skin permeability barrier [89]. Modulation of the estrogen signaling pathway may offer a promising approach for mitigating AD. Sismanopoulos et al reported that the VEGF signaling pathway facilitates the release of Vascular Endothelial Growth Factor from mast cells via IL-9 in AD, contributing to inflammation and vascular changes. Elevated IL-9 and its receptor expression in the affected skin underscores potential therapeutic targets in this pathway for disease management [90]. It has been demonstrated that the signaling pathways of T-cell receptors, in conjunction with the ligation of co-stimulatory molecules, facilitate T-cell-mediated inflammation. The inflammatory response plays a crucial role in the pathophysiology of AD [91,92]. Considering the well-established role of VEGF and T-cell receptor signaling pathways in the progression of AD, targeting these pathways presents a promising therapeutic approach for the disease. The ErbB receptor, from the tyrosine kinase family, includes ErbB1/EGFR/HER1 cell surface receptors that are integral to inflammation and AD, and primarily affect skin barrier integrity and immune response dynamics. Moreover, EGFR and MAPK are key targets of the ErbB signaling pathway. Dysregulation of this pathway exacerbates susceptibility to allergens and inflammatory reactions, contributing to the disease’s characteristic skin barrier dysfunction and heightened immune responses [93]. Therefore, the inhibition of the ErbB signaling pathway could serve as a potential therapeutic strategy for the treatment of AD. Hence, the results of the current study indicate that while several pathways may influence the anti-AD effects of Kochiae Fructus phytochemicals, the C-type lectin receptor signaling pathway is particularly significant. It not only interacts with a notably larger number of anti-AD core targets but also plays a more direct role in intensifying the crucial inflammatory and immune responses at the heart of AD pathogenesis (Fig 13). Consequently, the C-type lectin receptor signaling pathway is the most prominent pathway affected by Kochiae Fructus phytochemicals in the management of AD.
The diagram highlights the relevant pathway anti-AD core targets (SRC, SHP2 (PTPN11), ERK (MAPK1 and MAPK2), AP-1 (JUN), PI3K (PIK3CA), IL-6, and Cox-2 (PTGS2)) through a discernible pink hue.
Our molecular docking study highlighted key phytochemicals from Kochiae Fructus, particularly ecdysterone (MOL002212) and 11,14-eicosadienoic acid (MOL002211), which exhibited significant binding affinities to anti-AD core targets. These findings suggest their potential use as therapeutic agents. Ecdysterone interactions, facilitated by extensive hydrogen bonding with targets, such as MAPK3 and MAPK1, suggest robust modulation capabilities against AD-related targets. Similarly, 11,14-eicosadienoic acid, supported by hydrogen bonds and salt bridges, displays high binding affinities to PTGS2, ESR1, and JUN, enhancing both stability and specificity. Moreover, it is noteworthy that both ecdysterone and 11,14-eicosadienoic acid exhibited strong binding affinities to all anti-AD core targets associated with the C-type lectin receptor signaling pathway, as shown in Fig 8 and 13.
MD simulations of the MOL002212–MAPK1 and MOL002211–ESR1 complexes confirmed stable interactions over 1000 ns, demonstrating robust conformational integrity throughout the simulation. By validating the docking predictions, these results underscore the resilience of each complex under physiological conditions, a key consideration for advancing therapeutic candidates. Prolonged stability in silico suggests these molecules may effectively bind their respective protein targets during extended exposure, increasing the likelihood of sustained pharmacological action. Such findings provide a solid foundation for subsequent in vitro and in vivo investigations aimed at assessing clinical viability and refining these agents for potential AD management. Moreover, Gibbs free energy (ΔGBind) calculations underscored the spontaneity and stability of the binding interactions driven by favorable Coulombic interactions (ΔGBindCoulomb), van der Waals forces (ΔGBindvdW), and lipophilic interactions (ΔGBindLipo). This energetic analysis suggested that these complexes could exhibit favorable pharmacokinetic properties, which are essential for further therapeutic development.
Hence, this study advances traditional medicine research by showcasing how modern bioinformatics can uncover the therapeutic potential of Kochiae Fructus, a traditional herbal remedy for AD. Beyond validating its historical applications, the findings bridge the gap between traditional knowledge and contemporary drug discovery. By providing a robust, mechanistic framework, this work not only highlights the relevance of Kochiae Fructus in managing AD but also serves as a foundation for integrating other traditional medicinal herbs into modern therapeutic strategies. This integration marks a significant step forward in leveraging natural products for innovative, evidence-based therapies.
5. Conclusion
In this study, we systematically delineated the pharmacodynamic profile of Kochiae Fructus, a cornerstone of TCM, revealing its potential as a multifaceted therapeutic agent against AD. Utilizing state-of-the-art bioinformatics approaches, including network pharmacology and molecular docking, we identified 19 phytochemicals in Kochiae Fructus. Among these, seven phytochemicals demonstrated profound binding affinities to the top ten pharmacologically important anti-AD core targets, which were elucidated through robust PPI network analyses. Our findings underscore the profound impact of Kochiae Fructus on pivotal biological pathways that are integral to the pathophysiology of AD. GO and KEGG pathway enrichment analyses demonstrated that the anti-AD effects of Kochiae Fructus are intricately linked to the modulation of pathways, such as the C-type lectin receptor signaling pathway. This pathway’s engagement with multiple core anti-AD targets significantly influences inflammatory and immune processes central to AD pathogenesis, thereby shaping the scope of therapeutic approaches. Further substantiation from MD simulations and MM-GBSA calculations not only confirmed the stability and binding efficacy of these phytochemical-target interactions but also highlighted the potential of these compounds (ecdysterone and 1114-eicosadienoic acid) to act as potent modulators of AD pathology.
The results of this study provide a robust foundation for the therapeutic application of Kochiae Fructus in AD. By elucidating the molecular mechanisms and key targets, this study paves the way for the development of novel anti-AD agents derived from Kochiae Fructus, thereby contributing to more effective and targeted AD management strategies. These findings advance our understanding of the pharmacodynamic properties of Kochiae Fructus, including its target interactions, pathway modulation, binding affinity, and binding stability. The integration of network pharmacology with molecular docking and dynamic simulations offers a powerful framework for exploring the multifaceted interactions and therapeutic potential of traditional medicinal herbs in modern health care. Despite the comprehensive insights gained through network pharmacology and bioinformatics, the in-silico predictions presented here warrant experimental validation to confirm their efficacy, safety, and real-world applicability. Consequently, future in-vitro and in-vivo investigations—alongside clinical studies—are indispensable to substantiate and refine these findings for therapeutic application.
Supporting information
S1 Fig. Solvent-accessible surface area (SASA) of MOL002212 bound to MAPK1.
https://doi.org/10.1371/journal.pone.0320818.s001
(DOCX)
S1 Table. List of phytochemicals identified with a probability score of > 0.
https://doi.org/10.1371/journal.pone.0320818.s002
(DOCX)
S2 Table. Grid details of proteins used for molecular docking analysis.
https://doi.org/10.1371/journal.pone.0320818.s003
(DOCX)
Acknowledgments
We thank the PolyU Distinguished Postdoctoral Fellowship Scheme at The Hong Kong Polytechnic University, Hong Kong, for helping to make this research successful.
References
- 1. Ständer S. Atopic dermatitis. Ropper AH, editor. New england journal of medicine. 2021;384:1136–43.
- 2. Ring J. Terminology of allergic phenomena. Chem Immunol Allergy. 2014;100:46–52. pmid:24925383
- 3. Chen X-L, Li B, Zhao X, Li L-H, Wang H-L, Zhang K, et al. Kochiae Fructus: evaluation on the antioxidant properties and oral safety of its water decoction. Toxicon. 2022;219:106934. pmid:36195155
- 4. Correale CE, Walker C, Murphy L, Craig TJ. Atopic dermatitis: a review of diagnosis and treatment. Am Fam Physician. 1999;60:1191–98. Available: https://www.aafp.org/pubs/afp/issues/1999/0915/p1191.html
- 5.
Wan J. Atopic dermatitis and erythroderma. Inpatient Dermatology. 2018; 299–302. https://doi.org/10.1007/978-3-319-18449-4_60
- 6. Nam YK, Kim MH, Ha IJ, Yang WM. Derma-Hc, a new developed herbal formula, ameliorates cutaneous lichenification in atopic dermatitis. Int J Mol Sci. 2021;22(5):2359. pmid:33652999
- 7. Man M-Q, Hupe M, Sun R, Man G, Mauro TM, Elias PM. Topical apigenin alleviates cutaneous inflammation in murine models. Evid Based Complement Alternat Med. 2012;2012:912028. pmid:23304222
- 8. Trautmann A, Akdis M, Schmid-Grendelmeier P, Disch R, Bröcker EB, Blaser K, et al. Targeting keratinocyte apoptosis in the treatment of atopic dermatitis and allergic contact dermatitis. J Allergy Clin Immunol. 2001;108(5):839–46. pmid:11692113
- 9. Nutten S. Atopic dermatitis: global epidemiology and risk factors. Ann Nutr Metab. 2015;66(1):8–16. pmid:25925336
- 10. HOME | Global Atopic Dermatitis Atlas. [cited 20 Dec 2024]. Available from: https://www.atopicdermatitisatlas.org/en/
- 11. Arents BWM, van Zuuren EJ, Vermeulen S, Schoones JW, Fedorowicz Z. Global Guidelines in Dermatology Mapping Project (GUIDEMAP), a systematic review of atopic dermatitis clinical practice guidelines: are they clear, unbiased, trustworthy and evidence based (CUTE)?. Br J Dermatol. 2022;186(5):792–802. pmid:34984668
- 12. Laughter MR, Maymone MBC, Mashayekhi S, Arents BWM, Karimkhani C, Langan SM, et al. The global burden of atopic dermatitis: lessons from the Global Burden of Disease Study 1990-2017. Br J Dermatol. 2021;184(2):304–9. pmid:33006135
- 13. McKenzie C, Silverberg JI. The prevalence and persistence of atopic dermatitis in urban United States children. Ann Allergy Asthma Immunol. 2019;123(2):173–8.e1. pmid:31128232
- 14. The Complete Guide to Eczema in Hong Kong | Healthy Matters. [cited 21 Dec 2021]. Available from: https://www.healthymatters.com.hk/eczema-in-hong-kong-causes-symptoms-treatment/
- 15. “Seeing The Invisible” – Hong Kong’s first community-wide program for children and families affected by childhood eczema - All News - Media - HKU. [cited 21 Dec 2021]. Available: https://www.hku.hk/press/news_detail_16241.html
- 16. Cheng NS, Chau PCJ, Hon KLE, Choi KC, Kung JSC, Ng WG, et al. Measuring the quality of life of the families of children with eczema in Hong Kong. Asia Pac Allergy. 2019;9(3):e26. pmid:31384581
- 17. Boguniewicz M, Leung DYM. Atopic dermatitis: a disease of altered skin barrier and immune dysregulation. Immunol Rev. 2011;242(1):233–46. pmid:21682749
- 18. Chu M, Tsang MS-M, He R, Lam CW-K, Quan ZB, Wong CK. The active compounds and therapeutic mechanisms of pentaherbs formula for oral and topical treatment of atopic dermatitis based on network pharmacology. Plants (Basel). 2020;9(9):1166. pmid:32916837
- 19. Langan SM, Irvine AD, Weidinger S. Atopic dermatitis. Lancet. 2020;396(10247):345–60. pmid:32738956
- 20. Jiao D, Wong C-K, Qiu H-N, Dong J, Cai Z, Chu M, et al. NOD2 and TLR2 ligands trigger the activation of basophils and eosinophils by interacting with dermal fibroblasts in atopic dermatitis-like skin inflammation. Cell Mol Immunol. 2016;13(4):535–50. pmid:26388234
- 21. Liu F-T, Goodarzi H, Chen H-Y. IgE, mast cells, and eosinophils in atopic dermatitis. Clin Rev Allergy Immunol. 2011;41(3):298–310. pmid:21249468
- 22. Alexander T, Maxim E, Cardwell LA, Chawla A, Feldman SR. Prescriptions for atopic dermatitis: oral corticosteroids remain commonplace. J Dermatolog Treat. 2018;29(3):238–40. pmid:28789575
- 23. Williams HC. Atopic dermatitis. N Engl J Med. 2009;352: 2314–24.
- 24. Ormerod AD. Topical tacrolimus and pimecrolimus and the risk of cancer: how much cause for concern?. Br J Dermatol. 2005;153(4):701–5. pmid:16181449
- 25. Nghiem P, Pearson G, Langley RG. Tacrolimus and pimecrolimus: from clever prokaryotes to inhibiting calcineurin and treating atopic dermatitis. J Am Acad Dermatol. 2002;46(2):228–41. pmid:11807435
- 26. Ashcroft DM, Dimmock P, Garside R, Stein K, Williams HC. Efficacy and tolerability of topical pimecrolimus and tacrolimus in the treatment of atopic dermatitis: meta-analysis of randomised controlled trials. BMJ. 2005;330(7490):516. pmid:15731121
- 27. Matsuda H, Li Y, Yamahara J, Yoshikawa M. Inhibition of gastric emptying by triterpene saponin, momordin Ic, in mice: roles of blood glucose, capsaicin-sensitive sensory nerves, and central nervous system. J Pharmacol Exp Ther. 1999;289(2):729–34. pmid:10215646
- 28. Zou W, Tang Z, Long Y, Xiao Z, Ouyang B, Liu M. Kochiae Fructus, the Fruit of Common Potherb Kochia scoparia (L.) Schrad: a review on phytochemistry, pharmacology, toxicology, quality control, and pharmacokinetics. Evid Based Complement Alternat Med. 2021;2021:5382684. pmid:33603816
- 29. Xiao Z, Xiao S, Zhang Y, Pan T, Ouyang B. The anti-inflammatory effect of fructus kochiae on allergic contact dermatitis rats via pERK1/2/TLR4/NF-κB pathway activation. Evid Based Complement Alternat Med. 2018;20181096920. pmid:29507585
- 30. Choi YY, Kim MH, Lee JY, Hong J, Kim S-H, Yang WM. Topical application of Kochia scoparia inhibits the development of contact dermatitis in mice. J Ethnopharmacol. 2014;154(2):380–5. pmid:24751363
- 31. Luo M, Zeng B, Wang H, Yang Z, Peng Y, Zhang Y, et al. Kochia scoparia Saponin momordin Ic modulates HaCaT Cell proliferation and apoptosis via the Wnt/β-catenin pathway. Evid Based Complement Alternat Med. 2021;2021:5522164. pmid:34326883
- 32. Wang L, Xian Y-F, Hu Z, Loo SKF, Ip SP, Chan WY, et al. Efficacy and action mechanisms of a Chinese herbal formula on experimental models of atopic dermatitis. J Ethnopharmacol. 2021;274:114021. pmid:33716079
- 33. Chang JB, Lane ME, Yang M, Heinrich M. Disentangling the complexity of a hexa-herbal chinese medicine used for inflammatory skin conditions-predicting the active components by combining LC-MS-based metabolite profiles and in vitro pharmacology. Front Pharmacol. 2018;9:1091. pmid:30344490
- 34.
Koh DJ, Kim YH, Kim DG, Lee JY, Lee KT. Evaluation of the atopic dermatitis-mitigating and anti-inflammatory effects of Kyung Hee Allergic Disease Herbal Formula (KAHF). Food Science and Biotechnology. 2008. pp. 1272–8.
- 35. Xia X. Bioinformatics and drug discovery. Curr Top Med Chem. 2017;17(15):1709–26. pmid:27848897
- 36. Papp KA, Hong C-H, Lansang MP, Turchin I, Adam DN, Beecker JR, et al. Practical management of patients with atopic dermatitis on dupilumab. Dermatol Ther (Heidelb). 2021;11(5):1805–28. pmid:34510403
- 37. Ru J, Li P, Wang J, Zhou W, Li B, Huang C, et al. TCMSP: a database of systems pharmacology for drug discovery from herbal medicines. J Cheminform. 2014;6:13. pmid:24735618
- 38. Rebhan M, Chalifa-Caspi V, Prilusky J, Lancet D. GeneCards: integrating information about genes, proteins and diseases. Trends Genet. 1997;13(4):163. pmid:9097728
- 39. Oliveros JC. Venny 2.1.0. 2007 [cited 17 Mar 2022]. Available: https://bioinfogp.cnb.csic.es/tools/venny/
- 40. Daina A, Michielin O, Zoete V. SwissTargetPrediction: updated data and new features for efficient prediction of protein targets of small molecules. Nucleic Acids Res. 2019;47(W1):W357–64. pmid:31106366
- 41. von Mering C, Huynen M, Jaeggi D, Schmidt S, Bork P, Snel B. STRING: a database of predicted functional associations between proteins. Nucleic Acids Res. 2003;31(1):258–61. pmid:12519996
- 42. Lopes CT, Franz M, Kazi F, Donaldson SL, Morris Q, Bader GD. Cytoscape Web: an interactive web-based network browser. Bioinformatics. 2010;26(18):2347–8. pmid:20656902
- 43. DAVID Functional annotation bioinformatics microarray analysis. [cited 19 Apr 2022]. Available: https://david.ncifcrf.gov/.
- 44. Weishengxin. Weishengxin - data analysis and visualization experts around you. [cited 17 Mar 2022]. Available from: http://www.bioinformatics.com.cn/.
- 45. RCSB PDB: Search. [cited 11 Feb 2022]. Available: https://www.rcsb.org/search
- 46. PubChem. [cited 8 Jun 2022]. Available: https://pubchem.ncbi.nlm.nih.gov/
- 47.
The PyMOL Molecular Graphics System V 3. 0 SLLC. PyMOL.
- 48.
Schrödinger Release 2023-1. Glide, Schrödinger, LLC, New York, NY, 2023.
- 49.
Schrödinger Release 2023-1. LigPrep, Schrödinger, LLC, New York, NY, 2023.
- 50.
Schrödinger Release 2023-1. Protein Preparation Wizard; Epik, Schrödinger, LLC, New York, NY, 2023; Impact, Schrödinger, LLC, New York, NY; Prime, Schrödinger, LLC, New York, NY, 2023.
- 51. Chow E, Rendleman CA, Bowers KJ, Dror RO, Hughes DH, Gullingsrud J, et al. Desmond performance on a cluster of multicore processors hardware and operating environment benchmark systems and simulation parameters. 2008. Available: https://www.deshawresearch.com/publications/Desmond-GPU_Performance_April_2021.pdf
- 52. Tang L, Gao J, Li X, Cao X, Zhou B. Molecular mechanisms of luteolin against atopic dermatitis based on network pharmacology and in vivo experimental validation. Drug Des Devel Ther. 2022;16:4205–21. pmid:36530790
- 53. Cao F, Cheng Y-S, Yu L, Xu Y-Y, Wang Y. Bioinformatics analysis of differentially expressed genes and protein-protein interaction networks associated with functional pathways in ulcerative colitis. Med Sci Monit. 2021;27e927917. pmid:33462173
- 54. Zhang Y-Z, Yang J-Y, Wu R-X, Fang C, Lu H, Li H-C, et al. Network pharmacology-based identification of key mechanisms of Xihuang pill in the treatment of triple-negative breast cancer stem cells. Front Pharmacol. 2021;12:714628. pmid:3473769
- 55. Huang DW, Sherman BT, Lempicki RA. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc. 2009;4(1):44–57. pmid:19131956
- 56. Beard H, Cholleti A, Pearlman D, Sherman W, Loving KA. Applying physics-based scoring to calculate free energies of binding for single amino acid mutations in protein-protein complexes. PLoS One. 2013;8(12):e82849. pmid:24340062
- 57. Srinivasan M, Gangurde A, Chandane AY, Tagalpallewar A, Pawar A, Baheti AM. Integrating network pharmacology and in silico analysis deciphers Withaferin-A’s anti-breast cancer potential via hedgehog pathway and target network interplay. Brief Bioinform. 2024;25(2):bbae032. pmid:38446743
- 58. He Q, Liu C, Wang X, Rong K, Zhu M, Duan L, et al. Exploring the mechanism of curcumin in the treatment of colon cancer based on network pharmacology and molecular docking. Front Pharmacol. 2023;14:1102581. pmid:36874006
- 59. Alzain AA, Mukhtar RM, Abdelmoniem N, Elbadwi FA, Hussien A, Garelnabi EAE, et al. Computational insights into natural antischistosomal metabolites as SmHDAC8 inhibitors: molecular docking, ADMET profiling, and molecular dynamics simulation. Metabolites. 2023;13(5):658. pmid:37233699
- 60. Owoloye AJ, Ligali FC, Enejoh OA, Musa AZ, Aina O, Idowu ET, et al. Molecular docking, simulation and binding free energy analysis of small molecules as PfHT1 inhibitors. PLoS One. 2022;17(8):e0268269. pmid:36026508
- 61. Shivakumar D, Williams J, Wu Y, Damm W, Shelley J, Sherman W. Prediction of absolute solvation free energies using molecular dynamics free energy perturbation and the OPLS force field. J Chem Theory Comput. 2010;6(5):1509–19. pmid:26615687
- 62. Martyna GJ, Klein ML, Tuckerman M. Nosé–Hoover chains: the canonical ensemble via continuous dynamics. J Chem Phys. 1992;97(4):2635–43.
- 63. Kagami LP, das Neves GM, Timmers LFSM, Caceres RA, Eifler-Lima VL. Geo-measures: a PyMOL plugin for protein structure ensembles analysis. Comput Biol Chem. 2020;87107322. pmid:32604028
- 64. Sastry GM, Adzhigirey M, Day T, Annabhimoju R, Sherman W. Protein and ligand preparation: parameters, protocols, and influence on virtual screening enrichments. J Comput Aided Mol Des. 2013;27(3):221–34. pmid:23579614
- 65. Setlur A, Karunakaran C, Anusha V, Shendre A, Uttarkar A, Niranjan V. Investigating the molecular interactions of quinoline derivatives for antibacterial activity against Bacillus subtilis: computational biology and in vitro study interpretations. Mol Biotechnol. 2023;11–22.
- 66. Lu J, Yin R, Fu Z, Lee VM, Nelson JS, Tan W. Activation of RhoA, Smad2, c-Src, PKC-βII/δ and JNK in atopic dermatitis. Australas J Dermatol. 2018;59(4):e258–61. pmid:29498756
- 67. Tan Q, Yang H, Liu E, Wang H. P38/ERK MAPK signaling pathways are involved in the regulation of filaggrin and involucrin by IL‑17. Mol Med Rep. 2017;16(6):8863–7. pmid:28990053
- 68. Johansen C, Kragballe K, Westergaard M, Henningsen J, Kristiansen K, Iversen L. The mitogen-activated protein kinases p38 and ERK1/2 are increased in lesional psoriatic skin. Br J Dermatol. 2005;152(1):37–42. pmid:15656798
- 69. Yu X-J, Li C-Y, Dai H-Y, Cai D-X, Wang K-Y, Xu Y-H, et al. Expression and localization of the activated mitogen-activated protein kinase in lesional psoriatic skin. Exp Mol Pathol. 2007;83(3):413–8. pmid:17599830
- 70. Qin Q, Qin L, Xie R, Peng S, Guo C, Yang B. Insight into biological targets and molecular mechanisms in the treatment of arsenic-related dermatitis with vitamin a via integrated in silico approach. Front Nutr. 2022;9:847320. pmid:35685889
- 71. Uluçkan Ö, Guinea-Viniegra J, Jimenez M, Wagner EF. Signalling in inflammatory skin disease by AP-1 (Fos/Jun). Clin Exp Rheumatol. 2015;33(4 Suppl 92):S44–9. pmid:26458100
- 72. Teng Y, Fan Y, Ma J, Lu W, Liu N, Chen Y, et al. The PI3K/Akt pathway: emerging roles in skin homeostasis and a group of non-malignant skin disorders. Cells. 2021;10(5):1219. pmid:34067630
- 73. Iwano R, Iwashita N, Takagi Y, Fukuyama T. Estrogen receptor α activation aggravates imiquimod-induced psoriasis-like dermatitis in mice by enhancing dendritic cell interleukin-23 secretion. J Appl Toxicol. 2020;40(10):1353–61. pmid:32368827
- 74. Niu J, Zhou H, Tian R, Wang X. Estrogen Receptor Alpha Gene (ESR1) facilitates Th2-immune response and enhances Th2 cytokines in experimental atopic dermatitis mice. Iran J Immunol. 2023;20(2):167–76. pmid:37246779
- 75. Lee K-J, Ratih K, Kim G-J, Lee Y-R, Shin J-S, Chung K-H, et al. Immunomodulatory and anti-inflammatory efficacy of hederagenin-coated maghemite (γ-Fe2O3) nanoparticles in an atopic dermatitis model. Colloids Surf B Biointerfaces. 2022;210:112244. pmid:34896691
- 76. Chiricozzi A, Maurelli M, Peris K, Girolomoni G. Targeting IL-4 for the treatment of atopic dermatitis. Immunotargets Ther. 2020;9:151–6. pmid:33062619
- 77. Hou W, Jin Y-H, Kang HS, Kim BS. Interleukin-6 (IL-6) and IL-17 synergistically promote viral persistence by inhibiting cellular apoptosis and cytotoxic T cell function. J Virol. 2014;88(15):8479–89. pmid:24829345
- 78. Onderdijk AJ, Baerveldt EM, Kurek D, Kant M, Florencia EF, Debets R, et al. IL-4 downregulates IL-1β and IL-6 and induces GATA3 in psoriatic epidermal cells: route of action of a Th2 cytokine. J Immunol. 2015;195(4):1744–52. pmid:26170379
- 79. Liana D, Eurtivong C, Phanumartwiwath A. Boesenbergia rotunda and Its Pinostrobin for atopic dermatitis: dual 5-lipoxygenase and cyclooxygenase-2 inhibitor and its mechanistic study through steady-state kinetics and molecular modeling. Antioxidants (Basel). 2024;13(1):74. pmid:38247498
- 80. Khan SA, Wu Y, Li AS-M, Fu X-Q, Yu Z-L. Network pharmacology and molecular docking-based prediction of active compounds and mechanisms of action of Cnidii Fructus in treating atopic dermatitis. BMC Complement Med Ther. 2022;22(1):275. pmid:36261841
- 81. Wang X, Zhang A, Gao J, Chen W, Wang S, Wu X, et al. Trichomide A, a natural cyclodepsipeptide, exerts immunosuppressive activity against activated T lymphocytes by upregulating SHP2 activation to overcome contact dermatitis. J Invest Dermatol. 2014;134(11):2737–46. pmid:24933319
- 82. Duroudier NP, Tulah AS, Sayers I. Leukotriene pathway genetics and pharmacogenetics in allergy. Allergy. 2009;64(6):823–39. pmid:19416143
- 83. Trautmann A, Akdis M, Kleemann D, Altznauer F, Simon HU, Graeve T, et al. T cell-mediated Fas-induced keratinocyte apoptosis plays a key pathogenetic role in eczematous dermatitis. J Clin Invest. 2000;106(1):25–35. pmid:10880045
- 84. Tai L-R, Chiang Y-F, Huang K-C, Chen H-Y, Ali M, Hsia S-M. Hinokitiol as a modulator of TLR4 signaling and apoptotic pathways in atopic dermatitis. Biomed Pharmacother. 2024;170116026. pmid:38128179
- 85. Buske-Kirschbaum A, Gierens A, Höllig H, Hellhammer DH. Stress-induced immunomodulation is altered in patients with atopic dermatitis. J Neuroimmunol. 2002;129(1–2):161–7. pmid:12161032
- 86. Searing DA, Leung DYM. Vitamin D in atopic dermatitis, asthma and allergic diseases. Immunol Allergy Clin North Am. 2010;30(3):397–409. pmid:20670821
- 87. Zeze N, Kido-Nakahara M, Tsuji G, Maehara E, Sato Y, Sakai S, et al. Role of ERK pathway in the pathogenesis of atopic dermatitis and its potential as a therapeutic target. Int J Mol Sci. 2022;23(7):3467. pmid:35408826
- 88. Tang C, Makusheva Y, Sun H, Han W, Iwakura Y. Myeloid C-type lectin receptors in skin/mucoepithelial diseases and tumors. J Leukoc Biol. 2019;106(4):903–17. pmid:30964564
- 89. Gratton R, Del Vecchio C, Zupin L, Crovella S. Unraveling the role of sex hormones on keratinocyte functions in human inflammatory skin diseases. Int J Mol Sci. 2022;23(6):3132. pmid:35328552
- 90. Sismanopoulos N, Delivanis DA, Alysandratos KD, Angelidou A, Vasiadi M, Therianou A, et al. IL-9 induces VEGF secretion from human mast cells and IL-9/IL-9 receptor genes are overexpressed in atopic dermatitis. PLoS One. 2012;7(3):e33271. pmid:22413008
- 91. Croft M. Co-stimulatory members of the TNFR family: keys to effective T-cell immunity?. Nat Rev Immunol. 2003;3(8):609–20. pmid:12974476
- 92. Croft M, Esfandiari E, Chong C, Hsu H, Kabashima K, Kricorian G. OX40 in the pathogenesis of atopic dermatitis—A new therapeutic target. Am J Clin Dermatol. 2024;2024:1–15.
- 93. Xia T, Liang X, Liu C-S, Hu Y-N, Luo Z-Y, Tan X-M. Network pharmacology integrated with transcriptomics analysis reveals ermiao wan alleviates atopic dermatitis via suppressing MAPK and activating the EGFR/AKT signaling. Drug Des Devel Ther. 2022;16:4325–41. pmid:36578822