The Efficacy and Underlying Pathway Mechanisms of ShiDuGao Treatment for Anus Eczema Based on GEO Datasets and Network Pharmacology

Anal eczema is an allergic skin condition that affects the perianal region and mucosa, exhibiting various clinical manifestations¹. The characteristic symptoms include anal erythema, papules, blisters, erosion, exudates, and crusting. These symptoms mostly arise due to scratching, thickening, and roughness of the affected area².

Anal eczema, characterized by a prolonged duration of the disease, recurrent attacks, and challenging treatment, can have adverse effects on patients' physical and mental health³. The pathogenesis of anal eczema is not yet clear, and modern medicine suggests that it may be related to local anal lesions, diet, environment, genetics, and other factors⁴. In addition to avoiding contact with irritants and potential allergens, the treatment of anal eczema mainly focuses on methods such as inhibiting inflammation, anti-allergy, and relieving itching⁵.

SDG has been extensively utilized for the treatment of anal eczema and other anal conditions. SDG regulates anal skin exudation, reduces moisture, repairs anal skin, and effectively addresses pruritus^6,7,8. Furthermore, SDG has the potential to regulate perianus microbiota, thereby improving anus eczema^9,10.

Network pharmacology, a novel and interdisciplinary, cutting-edge bioinformatic approach in the realm of artificial intelligence and big data, provides an in-depth exploration of traditional Chinese medicine. This discipline emphasizes the systemic expounding of molecular correlation rules between drugs and diseases from an ecological network perspective. It has been extensively adopted for various aspects, including identifying key active ingredients in herb extracts, deciphering their global mechanisms of action, formulating drug combinations, and studying prescription compatibility. Traditional Chinese prescriptions exhibit the attributes of multi-component and multi-target, signifying their substantial adaptability to the realm of network pharmacology. Driven by this methodology, fresh perspectives have emerged in the examination of complex traditional Chinese medicine systems, furnishing robust technical support for clinical application rationalization and drug innovation^11,12,13,14.

This study aims to explore the mechanism of effectiveness of SDG in the treatment of anal eczema. This investigative effort sought to elucidate the mechanism of topical drug administration using a synergistic integration of network pharmacology and GEO datasets. The findings provide valuable insights into the efficacy and underlying mechanisms of SDG in the management of anus eczema, indicating its potential as an effective therapeutic approach for this condition.The detailed workflow diagram of the study is presented in Figure 1.

This study does not refer to ethical approval and consent to participate. The data used in this study was obtained from gene databases.

1. Prediction of disease targets

1. Access the GeneCards database (https://www.genecards.org) and online Mendelian inheritance in man database (OMIM, https://www.omim.org), utilizing "anus eczema" as the search term for disease targets.
2. Download the spreadsheets of the disease targets. Delete the repeated targets to obtain the anus eczema targets.

2. Selection of active components

1. Search the keyword "indigo naturalis, golden cypress, calcined gypsum, calamine, and Chinese Gall" on the Traditional Chinese Medicine system's pharmacology database (TCMSP; http://tcmspw.com/tcmsp.php) to obtain the list of the candidate active ingredients and targets of SDG.
2. Entrust the component onto the Swiss ADME database (http://www.swissadme.ch/index.php), extracting details of those exhibiting "high" GI absorption, coupled with at least two "Yes" DL values as active elements.
   NOTE: Normally, only ingredients with drug-like (DL) values ≥0.18 in the database are included as active ingredients.

3. Construction of the PPI network and screening of the core proteins

1. In Venny2.1( https://bioinfogp.cnb.csic.es/tools/venny/index.html), enter the targets of SDG and anus eczema into LIST1 and LIST2, respectively. A visual representation of the intersection is generated instantly. Click on the shared area to reveal the common targets in the Results section.
2. Access the STRING database (https://string-db.org/). Enter the targets in the List of Names field. Then Select Homo sapiens as the Organism and proceed with Search > Continue.
3. When the results are available, open Advanced Settings and select the hide disconnected nodes in the network. In the Minimum Required Interaction Score, set the highest confidence (0.900) and then click on Update.
4. Click on Exports to download the text of the protein-protein interaction (PPI) network in .png and .tsv format.

4. Construction of a drug-component-disease-target network

1. Open Cytoscape 3.9.1 and import the .tsv file mentioned in step 3.4. Click on the Style bar in the control panel to optimize the color, font, and side of the network nodes.
2. For network topology analysis, employ the Analyze Network function. To obtain hub genes, use CytoHubba in Cytoscape software. Establish the drug-component-disease-target network.

5. GO and KEGG enrichment analysis

1. Access the Metascape website (https://metascape.org/). Select a file or paste a gene list into the dialog box and click the Submit button. Then select H. sapiens in both Input as Species and Analysis as Species; after that, enable the Custom Analysis function.
2. In the enrichment option, select GO Molecular Functions, GO Biological Processes, GO Cellular Components, and the KEGG Pathway database. Check Pick Selective GO Clusters, then click on the Enrichment Analysis button. Upon completion of the progress bar, initiate an Analysis Report Page click to retrieve the enrichment results.

6. GEO gene chip dataset analysis

1. Search and analyze the GEO gene chip dataset (GDS3806) using the GEO2R tool (https://ncbi.nlm.nih.gov/geo/geo2r/) to investigate the expression of central genes in different data groups (control group-non-atopic dermatitis; experimental group-atopic dermatitis).
2. Enter the GEO database website (https://www.ncbi.nlm.nih.gov/geo/). Input keyword or GEO Accession, and click on the Search button. Select the best matching result. Find the Reference Series (GSE26952).
3. Enter the GEO2R tool website (https://ncbi.nlm.nih.gov/geo/geo2r/), enter the reference series in the GEO Accession box, and click the Set button. Select Atopic Dermatitis as the experimental group, select Nonatopic Control as the control group, and click the Analyze button. After the calculation is completed, the result will appear.

7. Molecular docking

1. Open the TCMSP database and download the 3D structure of the selected ingredients. Use the Chemical Name search box and search the selected ingredient names to download the corresponding 3D structure files in mol2 format.
2. Open the RCSB protein database (http://www.pdb.org/) and download the crystal structures of the key targets. In the search box, search the target names and download the corresponding crystal structure files in pdb format.
3. Import ingredients and target structure files into the analysis software. Delete water molecules by clicking on Edit > Delete Water. Add hydrogens by clicking on Edit > Hydrogens > Add. Set the ingredients as the ligand, select whole targets as the receptor, and perform blind docking.
4. Determine the range of molecular docking.
   1. Select the receptor and ligand in sequence. Click on Grid > Grid Box to adjust the grid box to include the entire model. Click on File > Close saving current to save the grid box status. Save files in gpf format.
   2. Click on Run > Run Autogrid4 > Parameter Filename > Browse, select the gpf file, then click the Launch button.
5. Use AutoDock 4 to perform molecular docking.
   1. Click on Docking > Macromolecule > Set Rigid Filename to select the receptor. Click on Docking > Ligand > Open/ Choose to select the ligand.
   2. Click on Docking > Search Parameters to set operation algorithms and Docking > Docking Parameters to set docking parameters. Select the dpf file, then click the Launch button. Save files in the dpf format.
   3. Click on Analyze > Docking > Open, select the dlg file, click on Analyze > Macromolecule to open the receptor, click on Analyze > Conformations > Play, Ranked by Energy to analyze the results. Click Set Play > Write Complex to save the results in pdbqt format.
6. Import the docking files into PyMOL software to construct further visualization.
   1. Select the ligand, and click on Action > Find > Polar Contacts > To Other Atoms in Object to display hydrogen bonds between ligands and the external environment. Click on c to change color.
   2. Click on Action > Extract Object. Click on Show > Sticks to show the stick structure of the receptor. Identify the residues connected to ligands and show the stick structure.
   3. Click on Hide > Sticks to hide the stick structure of the receptor. Click on Wizard > Measurement and click on two atoms in sequence. Click on Label > Residue to show the label of the residues. Adjust the background color and transparency if necessary. Click on File > Export Image as to save the picture.

Anus eczema-related genes, SDG target genes, and common targets
A total of 958 potential gene candidates were screened in Genecards and 634 in OMIM databases, while duplicates were excluded. To gain a comprehensive understanding of anal eczema-related genes, the findings from multiple databases were amalgamated, yielding a total of 958 distinct genes. Consequently, a protein-protein interaction network (PPI) specific to anal eczema was meticulously formulated. SDG is composed of five traditional Chinese medicines, namely indigo naturalis, golden cypress, calcined gypsum, calamine, and Chinese Gall^15,16. The main component of calcined gypsum is anhydrous calcium sulfate (CaSO₄), while the main component of calamine is zinc carbonate (ZnCO₃). Indigo naturalis, golden cypress, and Chinese Gall have complex ingredients. From the TCMSP database, the drugs contain 92 compound components, obtaining a total of 867 reliable drug targets (Table 1).

Through the overlaying of both target gene datasets, a total of 149 frequently co-occurring target genes were pinpointed (Figure 2A), followed by the construction of an essential target protein-protein interaction (PPI) network (Figure 2B). Through a median-based screening method for degree, closeness, and betweenness, 59 key targets were selected as potential anal eczema drug targets. The median degree, closeness, and betweenness scores for the key targets were 49, 40.31947, and 0.522, respectively. The top 10 genes with a high degree score included AKT1, TNF, TP53, EGFR, STAT3, SRC, JUN, CASP3, HRAS, and PTGS2 (Table 2). These genes are highly relevant to anal eczema.

Pathways and networks involving common targets
KEGG and GO enrichment methods were utilized to analyze 59 key targets, revealing 218 associated pathways and over 3000 associated biological processes. Analysis uncovered pathways that strongly correlate with SDG and anal eczema proteins, including Cherry simplex virus 1 infection, Shigellosis, TNF signaling pathway, EGFR tyrosine kinase inhibitor resistance, Human cytomegalovirus infection, and T cell receptor signaling pathway (Figure 3A). These pathways primarily relate to genes such as AKT1, TNF, TP53, STAT3, SRC, EGFR, and CASP3. Figure 3B provides a detailed depiction of target genes and pathways. GO analysis was performed on biological processes (BP), cell composition (CC), and molecular function (MF) (Figure 4A). Results suggest that this study primarily focuses on common targets for SDG and anal eczema in biological processes, with a few relevant to CC and MF. Biological functions that were particularly relevant include peptidyl-tyrosine phosphorylation, peptidyl-tyrosine modification, regulation of cell-cell adhesion, positive regulation of cell adhesion, T cell activation, regulation of leukocyte cell-cell adhesion (Figure 4B-D).

Predicting the binding of SDG active components to anus eczema targets
Based on the median values of degree, closeness, and betweenness, 59 key targets were screened, including AKT1, TNF, TP53, EGFR, STAT3, SRC, JUN, CASP3, HRAS, and PTGS2. Further analysis of the GEO database revealed upregulation of PPARG, EGFR, and TNF, while PTPRC, MMP9, MAPK14, and CASP3 were downregulated in the experimental group (atopic dermatitis) (Figure 5). Through the analysis of common gene pathway enrichment, it was determined that these genes predominantly participated in the TNF signaling cascade and the MAPK signaling pathway. In the TNF signaling pathway, TNF expression was upregulated, while MMP9, MAPK14, and CASP3 expression were downregulated. In the MAPK signaling pathway, EGFR and TNF expression were upregulated, while MAPK14 and CASP3 were downregulated (Figure 6). Based on these findings, TNF, MAPK14, and CASP3 were considered as potential targets in SDG therapy.

To validate candidate targets in active components of SDG, docking analysis was used to test the accuracy between the active component structure and potential target proteins. These target proteins are involved in various functional connections and are the high nodes in the network, suggesting that they play a crucial role in the SDG response to anal eczema. The negative value of docking binding energy indicates the ability of SDG to dock with disease targets in vivo, with a more negative value indicating easier docking. In this investigation, the successful molecular docking of core active components with the key target was achieved, and the docking binding energy was negative, with values less than -1 kcal/mol. Indigo and berberrubine have good binding activity, with binding energy less than -5 kcal/mol (Table 3, Figure 7). Taken together, these results provide further evidence that these proteins corresponding to gene loci can act as SDG targets in anus eczema.

Figure 1: Network pharmacology analysis workflow. GO, Gene Ontology; KEGG,Kyoto Encyclopedia of Genes and Genomes; TCMSP, Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform; GEO, Gene Expression Omnibus. Please click here to view a larger version of this figure.

Figure 2: Venn diagram and PPI network of the common targets. (A) Venn diagram of intersection of drug target and disease target. (B) Common target PPI network by STRING. Please click here to view a larger version of this figure.

Figure 3: KEGG pathway enrichment analysis. (A) KEGG pathway enrichment analysis. The top 10 KEGG pathways are ranked according to the P-values in ascending order. (B) The connection between the pathway and the target: pathway (yellow), targets (red). Please click here to view a larger version of this figure.

Figure 4: GO enrichment analysis. (A) GO results of three ontology. (B) Biological process (BP) bubble chart. (C) Cell component (CC) bubble chart. (D) Molecular function (MF) bubble chart. Please click here to view a larger version of this figure.

Figure 5: Predicting potential targets result. (A) Heatmap of hub gene expression in GEO database, group A is the experimental group (atopic dermatitis), and group B is the control group (non-atopic dermatitis); (B) PPI network nodes represent proteins, edge represent the relationships. Please click here to view a larger version of this figure.

Figure 6: The signaling pathway. (A) MAPK signaling pathway. (B) TNF signaling pathway. Please click here to view a larger version of this figure.

Figure 7: Molecular docking of core genes and ingredients. Magenta represents the core components of SDG, and blue represents the residues of the core genes. Please click here to view a larger version of this figure.

Table 1: Active ingredients in SDG.

Table 2: Characteristics of the top 10 hub genes.

Table 3: The molecular docking binding energy of the ingredients and core genes.

Atopic dermatitis is a specific form of eczema that shares underlying mechanisms with eczema. Hub genes believed to be related to this condition are TNF, MAPK14, and CASP3. The therapeutic effects of SDG on anal eczema are mainly attributed to its action on the TNF and MAPK signaling pathways via these three hub genes¹⁷.

SDG includes five distinct drugs: indigo naturalis, golden cypress, calcined gypsum, calamine, and Chinese Gall. In traditional Chinese medicine, calcined gypsum and calamine can play a role in promoting wound healing and drying dampness, while indigo naturalis, golden cypress, and Chinese Gall can clear heat, detoxify, and dry dampness. The combination of these herbs can achieve the effect of draining moisture, promoting wound healing, clearing heat, and dispelling wind¹⁸.

Previous studies have indicated that the main components of SDG have anti-inflammatory properties. Indigo naturalis (IN) has been shown to treat colitis, psoriasis, and acute promyelocytic leukemia^19,20,21. IN's function may be related to its inhibition of TLR4/MyD88/NF-kB signal transduction, which reduces inflammation and promotes the healing of the intestinal mucosa in patients with ulcerative colitis (UC). It can also regulate intestinal flora, as demonstrated in the DSS-induced UC mouse model^22,23. Recent research highlights that UC is often coupled with an imbalance in the intestinal microbiome. IN can effectively rebalance the intestinal ecology and protect the gastrointestinal system, depending on the intestinal flora²⁴. By shifting proinflammatory cytokines to anti-inflammatory cytokines, golden cypress reduces the proliferation of T lymphocytes and DC-induced T cell and IL-12p70 cytokine secretions, promoting the interaction between DC and Treg²⁵. Saponarin and campesterol act as natural anti-inflammatory agents with anti-allergic effects^26,27,28. Tryptanthrin exhibits an antimicrobial action²⁹. Melianonen exhibits substantial inhibitory effects on both fungi and microbial flora that may contribute to the treatment of anal eczema^30,31.

Studies have found that the severity of skin diseases such as acne, irritant contact dermatitis, and allergic contact dermatitis are related to the microbial flora in the gut. Comparing the microflora distribution of acute and chronic anus eczema, the results showed that the staphylococcus microflora of acute anus eczema patients was more abundant in the chronic group³². Infants with atopic eczema and lower gut microbiome diversity demonstrate a correlation between microbial abundance and skin diseases³³. Based on the effects of various components in SDG on the intestinal flora, the possibility that SDG can improve anus eczema by regulating microflora cannot be ruled out. In addition, the melianone in SDG can also act on fungi to prevent anus eczema.

Mechanism research is acknowledged as the most intricate aspect of herbal prescription investigation. Network pharmacology currently permeates diverse aspects of the pharmaceutical field, marking a paradigm shift from conventional to contemporary biomedicine and redefining traditional Chinese medicine (TCM) development^34,35,36. It utilizes network targets as a foundation, constructing a network diagram linking TCM, active ingredients, targets, and disorders to anticipate relevant therapeutic targets. Network pharmacology comprehensively elucidates the interactions between drugs and disease targets and systematically examines associative network mechanisms, thereby forecasting pivotal metabolic pathways. Its usage has been strategically implemented for investigating the mechanisms of action of various herbals. Furthermore, by establishing a disease drug target PPI network, along with the construction of KEGG and GO enriched pathways, network pharmacology has facilitated the prediction of the complex mechanism by which Chinese herbs influence diseases and probes into the pathogenesis of afflictions^37,38,39. This research combined network pharmacology with GEO data sets to discern topical drug mechanisms.

Network pharmacology analysis merely predicts drug components and targets, verifying precise mechanisms necessitating animal experimentation or clinical trials. This study only used molecular docking simulation verification without conducting animal or clinical experiments to verify. The proposed network pharmacology framework for traditional Chinese medicine combines the predicted targets of individual herbs, albeit with a lower accuracy. The incorporation of GEO datasets substantially enhances this precision.

In this study, the pure data generation method was used to maximize data utilization by combining multiple databases. Especially for some diseases that are difficult to build animal models for, the online data are used primarily to predict and verify diseases and drug targets so as to guide the research direction and lay a good foundation for subsequent experimental verification.