Integrating Transcriptomics, Network Pharmacology, Molecular Docking and Experimental Validation to Explore the Pharmacological Mechanisms of Chinese Medicinal Honey in Treating Diabetic Wounds

ABSTRACT

Chinese medicinal honey (CMH) is traditionally used for wounds, but its mechanisms in diabetic wounds are unclear. This study aimed to investigate the active components of CMH and the molecular mechanisms underlying its promotion of wound healing in individuals with diabetes to provide a preliminary experimental basis for its clinical application. The main components of CMH were analysed using ultra‐performance liquid chromatography–quadrupole time‐of‐flight mass spectrometry (UPLC‐Q‐TOF/MS). An integrated strategy combining network pharmacology, transcriptomics, molecular docking and animal experiments was applied. A db/db diabetic mouse model was used to evaluate CMH in diabetic wound healing. Histopathology, immunohistochemistry, qRT‐PCR and western blotting were performed to assess anti‐inflammatory, antioxidant and pro‐angiogenic effects. Fifty‐two major CMH components, mainly flavonoids and phenolic acids, were identified. Network pharmacology and transcriptomics showed that CMH targets epidermal growth factor receptor (EGFR), interleukin (IL)‐1β, matrix metallopeptidase 9 (MMP9) and prostaglandin‐endoperoxide synthase 2 (PTGS2), with significant enrichment in the IL‐17 signalling pathway. Molecular docking validated a strong binding affinity between CMH components and key targets, including IL‐17A and NF‐κB. In vivo findings suggest that CMH markedly accelerates wound healing in diabetic mice, reduces inflammation and oxidative stress, promotes collagen deposition and angiogenesis and suppresses the activity of the IL‐17A/NF‐κB pathway. This study provides a preliminary experimental basis of CMH. CMH supports diabetic wound healing through a multi‐component, multi‐target mechanism, primarily by inhibiting the IL‐17A/NF‐κB signalling pathway, thereby attenuating inflammation and oxidative stress and promoting angiogenesis and tissue repair.

Abbreviations

BP

biological process

CC

cellular component

CMH

Chinese medicinal honey

DEGs

differentially expressed genes

DFU

diabetic foot ulcer

DW

diabetic wound

EGFR

epidermal growth factor receptor

GAPDH

glyceraldehyde‐3‐phosphate dehydrogenase

GO

Gene Ontology

H&E

haematoxylin and eosin

IF

immunofluorescence

IHC

immunohistochemistry

IL

interleukin

KEGG

Kyoto Encyclopedia of Genes and Genomes

MF

molecular function

MMP

matrix metallopeptidase

NF‐κB

nuclear factor‐κB

NK

natural killer cells

PPI

protein–protein interaction

PTGS2

prostaglandin‐endoperoxide synthase 2

qRT‐PCR

quantitative real‐time polymerase chain reaction

ROS

reactive oxygen species

TCM

traditional Chinese medicine

TNF‐α

tumour necrosis factor‐α

TRAF6

tumour necrosis factor receptor‐associated factor 6

UPLC‐Q‐TOF/MS

ultra‐performance liquid chromatography coupled with quadrupole time‐of‐flight mass spectrometry

WB

Western blot

WGCNA

weighted gene co‐expression network analysis

1. Introduction

With ongoing global changes in lifestyle and dietary patterns, diabetes has emerged as a major public health concern, seriously affecting human health [1]. According to the International Diabetes Federation, the number of adults living with diabetes worldwide reached 536.6 million in 2021, and this figure is expected to increase to 783.2 million by 2045 [2]. Persistent hyperglycaemia in diabetes causes systemic physiological dysfunction, resulting in complications such as retinopathy, neuropathy, renal failure, cardiovascular diseases and diabetic foot ulcers (DFUs) [3]. Among these, DFUs represent a classic form of diabetic wounds (DWs), affecting about 18.6 million individuals globally each year. Characterised by high disability, recurrence and mortality rates, DFUs impose a substantial burden on patients' quality of life, psychological health and socioeconomic status [4, 5]. Consequently, the effective treatment of DW remains an urgent clinical challenge.

The impaired healing of DW involves complex mechanisms, mainly including chronic inflammation, excessive reactive oxygen species (ROS), insufficient angiogenesis and reduced growth factor secretion [6, 7]. Traditional Chinese medicine (TCM), recognised for its proven efficacy, minimal side effects and broad availability, is increasingly viewed as a valuable complementary approach for DW management [8]. Honey, a viscous supersaturated sugar solution, has been used since ancient times for wound care [9]. Rich in flavonoids, honey demonstrates diverse biological activities, including antibacterial, anti‐inflammatory, antioxidant, immunomodulatory, autolytic debridement, nutritional and deodorising effects [10]. Numerous clinical studies have shown that honey can effectively support the healing of DFUs, reinforcing its unique potential in DW therapy [11].

Currently, the most extensively studied honey is Manuka honey from New Zealand. However, the Chinese Pharmacopoeia has long reported the use of Chinese medicinal honey (CMH) for treating non‐healing wounds and burns [12]. Nevertheless, research on CMH in promoting DW healing remains limited, and its underlying mechanisms and associated molecular pathways require further detailed investigation. Therefore, this study aims to integrate ultra‐high‐performance liquid chromatography–quadrupole time‐of‐flight mass spectrometry (UPLC‐Q‐TOF/MS) component analysis, network pharmacology, transcriptomics and molecular docking to systematically explore the mechanisms by which CMH promotes DW healing. Experimental validation was undertaken using a mouse model of DW, with the goal of providing a preliminary experimental basis for its clinical application and future research. The detailed design and strategy of the study are shown in Figure 1.

FIGURE 1.

Schematic diagram of this study.

2. Materials and Methods

2.1. Identification of Main Components of CMH

2.1.1. Preparation of the CMH Aqueous Extract

A total of 5.00 g of CMH (China Hunan Dehai Pharmaceutical Co. Ltd.; drug approval No. Z43020566; Batch No. 2403001; tolerance ±0.05 g) was weighed and dissolved in 10 mL of purified water. The solution was transferred to a 50 mL centrifuge tube, brought to volume with purified water and vortexed until fully dissolved. The mixture was centrifuged at 12,000 rpm for 10 min; the supernatant was collected and passed through a 0.22 μm membrane filter for subsequent analyses.

2.1.2. Chemical Composition Analysis of CMH

To elucidate and qualitatively characterise the chemical constituents of CMH, UHPLC‐Q‐TOF/MS (Waters, USA) was employed, with raw data acquired in MS^E mode. Raw files were processed in Progenesis QI (small‐molecule workflow). An in‐house library of honey‐related compounds was prepared from the literature, PubChem, BATMAN‐TCM and TCMBank, including compound names and corresponding .mol structures. The precursor‐ion mass tolerance was set to < 10 ppm, with a defined retention time matching window. Feature lists were matched against the in‐house library to generate a data matrix containing m/z, retention time and ion response values. In addition, fragment‐ion information inherent to MS^E acquisitions was extracted and matched within specified error tolerances in Progenesis QI to enhance the accuracy of compound identification and annotation.

2.1.3. Target Prediction for CMH Constituents

Putative targets of individual CMH constituents were predicted using SwissTargetPrediction (http://www.swisstargetprediction.ch/) and BATMAN‐TCM (http://bionet.ncpsb.org.cn/batman‐tcm/#/home) [13, 14]. Constituents without target information were excluded. Prediction results were filtered using database‐specific thresholds: probability > 0 for SwissTargetPrediction and score cutoff > 0.84 for BATMAN‐TCM. Targets from both databases were merged and deduplicated to obtain the final candidate target set for CMH analyses.

2.2. Identification of Differentially Expressed Genes (DEGs) in DFUs

Using the keyword ‘diabetic foot ulcer, DFU’, gene expression microarray datasets were retrieved from the Gene Expression Omnibus database (http://www.ncbi.nlm.nih.gov/geo). Dataset GSE80178 was used for differential expression analysis (normal, n = 3; DFU, n = 9; Homo sapiens ), and GSE134431 was used for weighted gene co‐expression network analysis (WGCNA) and immune infiltration analysis (normal, n = 8; DFU, n = 13; Homo sapiens ). DEGs were identified from GSE80178 using the ‘limma’ R package (version 4.3.3). Genes with |Log₂Fold Change| > 0.5 and adjusted p < 0.05 were considered statistically significant DEGs. Visualisation of DEGs was performed using the ‘ggplot2’ R package to generate volcano plots and heatmaps.

2.3. Immune Infiltration Analysis

The CIBERSORT algorithm was employed to quantify the proportions of 22 immune cell types in patients with different immune patterns within the dataset. Differences in immune cell fractions were assessed using the Wilcoxon rank‐sum test; p < 0.05 were considered statistically significant. Results were visualised using packages such as ‘ggplot2’ and ‘pheatmap’.

2.4. Construction of Weighted Gene Co‐Expression Network

A weighted gene co‐expression network was constructed using the ‘WGCNA’ R package. First, hierarchical clustering of study samples was performed to detect and remove outliers. Subsequently, a scale‐free network was constructed by selecting a soft threshold power of β = 20 (with scale‐free topology fit index R ² = 0.85) using the pickSoftThreshold function. An adjacency matrix was then constructed and converted to a topological overlap matrix. A gene dendrogram and module colours were generated based on dissimilarity measures. Finally, correlations between the identified modules and differential samples were calculated.

2.5. Protein–Protein Interaction (PPI) Network Construction

Potential therapeutic targets of CMH for DFU treatment were identified by intersecting the DEGs of DFUs, the module genes identified by WGCNA and CMH component targets. These potential targets were imported into the STRING database (https://cn.string‐db.org/) to construct a PPI network, with the species set as Homo sapiens and a minimum required interaction score of 0.4. The resulting data file ‘string_interactions_short.tsv’ was downloaded for further analysis. The ‘string_interactions_short.tsv’ file was imported into Cytoscape 3.8.2 for network visualisation. Node label size, colours depth and node size were configured according to Degree values, where higher Degree values correspond to larger labels, darker colours and bigger nodes, indicating greater importance of the node within the PPI network.

2.6. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) Enrichment Analysis

Enrichment analyses and visualisations were performed using R packages, including org. Hs.eg.db, colourspace, stringi, DOSE, clusterProfiler, pathview, ggplot2 and limma.imma.mma (limma.d limma.apvalueCutoff = 0.05 and qvalueCutoff = 0.05). The GO functional analysis primarily covered three categories: biological process (BP), cellular component (CC) and molecular function (MF). The top 10 entries with the smallest p values (p < 0.05) were selected for visualisation. The bubble size or bar length represents the number of genes enriched in GO/KEGG terms, while the colour indicates the significance of enrichment.

2.7. Molecular Docking Analysis

To evaluate the reliability of predicted interactions between the main components of CMH and the core targets, this study employed molecular docking. First, the molecular structures of the main components of CMH were obtained from the PubChem database (https://pubchem.ncbi.nlm.nih.gov/), and the core targets were identified. Subsequently, the crystal structures of the core target proteins were retrieved from the Protein Data Bank in the RCSB (https://www.rcsb.org/). Water molecules and extraneous atoms were removed using PyMOL software (https://pymol.org/2/). Finally, docking calculations were performed using AutoDock tools (https://autodock.scripps.edu/) to complete the molecular docking. The interaction modes between the core molecular components and their targets were visualised using PyMOL.

2.8. Animal Experiments

Male db/db mice (n = 12) and littermate db/m controls (n = 6), 8–10 weeks of age, were obtained and housed at the Animal Experiment Centre of Capital Medical University, Beijing, China. Research animals were treated humanely, and all animal procedures were approved by the Institutional Animal Care and Use Committee of Capital Medical University (AEEI‐2025‐149). After 1 week of acclimation, a full‐thickness excisional wound model was established. Mice were anaesthetised with 1% pentobarbital; dorsal hair was removed and the skin disinfected, followed by creation of a circular full‐thickness wound (10‐mm diameter) at the mid‐dorsal region; db/db mice were randomly assigned to either the db/db group or the db/db + CMH group (n = 6 per group). In the db/db + CMH group, 200 μL CMH was topically applied to the wound every other day; the db/m and db/db groups received an equal volume of phosphate‐buffered saline. Wounds were covered with sterile gauze, and mice were single‐housed. Digital photographs were taken on postoperative days 0, 3, 7, 10, 14 and 21 to document wound appearance; fasting blood glucose and body weight were recorded. Wound area and percent closure at each time point were quantified using ImageJ. On Day 21, peri‐wound full‐thickness skin was harvested and either snap‐frozen in liquid nitrogen and stored at −80°C or immediately fixed in 4% paraformaldehyde for subsequent experiments.

2.9. Histopathological Examination, Immunohistochemical Staining and Immunofluorescence (IF) Staining

On Day 21 post‐intervention, murine skin tissues were harvested, fixed in 4% paraformaldehyde and embedded in paraffin. Sequential sections of 5 μm thickness were prepared and subjected to haematoxylin and eosin (H&E) staining, Masson's trichrome staining, immunohistochemistry (IHC) and IF staining. Collagen deposition areas in Masson‐stained sections were quantified using ImageJ software and expressed as a percentage of the total tissue area. Epidermal thickness was measured at multiple sites within the wound bed using ImageJ, and the mean epidermal thickness was calculated for each group. The expression levels of tumour necrosis factor‐α (TNF‐α), CD31 and DHE were detected by IHC and IF to evaluate wound inflammation, angiogenesis and oxidative stress, respectively. Quantitative analysis of stained areas was conducted using ImageJ.

2.10. Quantitative Real‐Time Polymerase Chain Reaction (qRT‐PCR) Analysis

Total RNA was isolated from wound tissues using TRIzol reagent and reverse‐transcribed into cDNA with the NovoScript Plus All‐in‐one 1st Strand cDNA Synthesis SuperMix. qPCR was performed with a SYBR premix on a real‐time PCR system, with each sample run in technical triplicate. The primers were synthesised by Shanghai Sangon Biotech (Shanghai, China) and are presented in Table S1. Gene expression was normalised to glyceraldehyde‐3‐phosphate dehydrogenase (GAPDH), and relative mRNA levels were calculated using the 2^−ΔΔC _t method.

2.11. Western Blot (WB) Analysis

Total protein was extracted from murine wound tissues and quantified. Proteins were separated by 8% SDS‐PAGE gel electrophoresis and subsequently transferred onto polyvinylidene difluoride membranes. The primary antibodies (Table S2) diluted in TBST containing 5% BSA were added, and the membrane was incubated overnight at 4°C. After three 10‐min washes with TBST, the membrane was incubated with the Anti‐Mouse‐HRP secondary antibody (1/5000; CST, USA) or Anti‐Rabbit‐HRP secondary antibody (1/5000; CST, USA) at room temperature for 1 h. The membrane was then washed three times with TBST for 10 min each wash. After washing, protein bands were visualised using an ECL Plus chemiluminescence detection system. Quantitative analysis of the blots was performed using ImageJ, with protein expression levels normalised to GAPDH as the internal control.

2.12. Statistical Analysis

All statistical analyses were conducted using GraphPad Prism 9 (GraphPad Software, San Diego, CA, USA). Data were expressed as mean ± standard deviation from at least three independent experiments. For comparisons among multiple groups, one‐way analysis of variance followed by Tukey's post hoc test was applied. A p < 0.05 was considered statistically significant.

3. Results

3.1. Identification of Components in CMH

A total of 52 major constituents were identified in CMH by UHPLC‐Q‐TOF‐MS analysis (Figure 2), primarily flavonoids and phenolic acids, including kaempferol, quercetin, baicalein, hesperetin and luteolin. Detailed information and structural characteristics of these compounds are provided in Table S3.

FIGURE 2.

Identification of Chinese medicinal honey (CMH) decoction components by ultra‐performance liquid chromatography–quadrupole time‐of‐flight mass spectrometry (UPLC‐Q‐TOF/MS). (A) UPLC‐HRMS base peak ion current map: Negative ion mode of HH sample. (B) UPLC‐HRMS base peak ion current map: Positive ion mode of CMH sample.

3.2. Identification of DEGs in DFUs

In this retrospective study, data from the GSE80178 dataset were analysed, yielding 2297 DEGs. Among these, 967 genes were significantly upregulated, and 1130 genes were significantly downregulated (Figure 3A,B).

FIGURE 3.

Transcriptomic analysis of diabetic foot ulcers (DFUs). (A) Volcano plot of differentially expressed genes (DEGs) in DUFs. (B) Heatmap of DEGs in DUFs. (C) Proportion of 22 immune cell types in different DFU samples analysed by immune infiltration assessment. (D) Differential expression of 22 immune cell types between normal and DFU groups. *p < 0.05, **p < 0.01. (E) Weighted gene co‐expression network analysis (WGCNA) for identifying immune‐related module genes. (F) Hierarchical clustering analysis in WGCNA. (G) Correlation analysis of module genes between control and DFU groups in WGCNA.

3.3. Immune Infiltration Analysis

Immunoinfiltration analysis was conducted using the GSE134431 dataset. First, the proportions of various immune cell types across different samples were examined. The findings indicated that resting mast cells, M2 macrophages and resting CD4⁺ T memory cells constituted relatively high proportions in the samples (Figure 3C). Next, differences in immune cell infiltration between the normal and DFU groups were assessed. The analysis showed statistically significant differences in activated mast cells, neutrophils, activated natural killer (NK) cells and CD8⁺ T cells between the DFU and normal groups. Specifically, mast cells were activated, and neutrophils were upregulated in the DFU group, while NK cells were activated, and CD8⁺ T cells were downregulated (Figure 3D). Accordingly, these four immune cell types were selected as phenotypic traits, and WGCNA was applied to identify genes closely associated with these significantly altered immune cells.

3.4. Construction of the Weighted Gene Co‐Expression Network

For the scale‐free network, a soft threshold power of β = 20 (R ² = 0.85) was selected (Figure 3E). Similar modules were then merged using a minModuleSize of 500, yielding six distinct modules (Figure 3F). The results showed significant correlations between immune cells and specific modules: activated mast cells showed a significant negative correlation with genes in the grey module; neutrophils exhibited significant positive correlations with the sky blue and blue modules; NK cells were positively correlated with the grey module; and CD8⁺ T cells showed significant positive correlations with the grey and green modules (Figure 3G). Considering that the genes in the yellow and brown modules did not correlate significantly with any of the analysed immune cells, they were excluded from further analysis. The genes from the remaining modules were selected for subsequent investigation.

3.5. PPI Network Construction

UPLC‐Q‐TOF/MS identified a total of 52 compounds in CMH. Target prediction via BATMAN‐TCM and Swiss Target Prediction databases yielded 572 and 677 targets, respectively. After merging, 1082 potential CMH targets were obtained. Intersection analysis of these targets with 2297 DFU DEGs and 4651 WGCNA module genes identified 56 overlapping genes as the key gene set underlying CMH treatment of DFUs (Figure 4A). A PPI network was constructed using the STRING database to analyse functional associations among these genes. Topological analysis revealed that EGFR, IL1B, MMP9 and PTGS2 served as hub nodes in the PPI network (Figure 4B).

FIGURE 4.

Network pharmacology analysis of CMH in treating DFU. (A) Venn diagram illustrating the intersection of CMH targets, DFU DEGs and WGCNA module genes. (B) Protein–protein interaction (PPI) network of potential therapeutic targets of CMH for DFU. (C–E) Gene Ontology (GO) enrichment analysis of biological processes, cellular components and molecular functions. (F) Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis.

3.6. GO and KEGG Enrichment Analysis

To elucidate the therapeutic mechanisms of CMH in DFUs, functional enrichment analysis was conducted on the 56 intersecting genes. GO analysis indicated significant enrichment in 1161 BP terms, 44 CC terms and 96 MF terms. Highly enriched BP terms included neuroinflammatory response, female pregnancy, response to mechanical stimulus, multicellular organism processes and responses to ROS (Figure 4C–E). KEGG pathway analysis identified 109 significantly enriched pathways, among which the IL‐17 signalling pathway exhibited the highest significance. Other notable pathways included prostate cancer, bladder cancer, HIF‐1 signalling pathway, lipid and atherosclerosis, relaxin signalling pathway, renal cell carcinoma, TNF signalling pathway and human T‐cell leukaemia virus 1 infection (Figure 4F).

3.7. Molecular Docking Analysis

Based on the key genes and functional analysis results predicted by network pharmacology, the therapeutic effect of CMH on DFUs primarily involves modulation of inflammatory responses. Therefore, we selected key genes from the IL‐17 signalling pathway—IL‐17A, NF‐κB, IL‐1β and MMP9—and conducted molecular docking experiments with the main bioactive components of CMH: kaempferol, quercetin, baicalein, hesperetin, luteolin, apigenin, chrysin, vanillic acid, chlorogenic acid and ferulic acid. Binding affinity values were used to assess molecular docking results, with lower values indicating stronger binding. Typically, a binding affinity value < −5 kcal/mol suggests good binding activity. The data showed that all these compounds exhibited favourable binding activity with the targets (Figure 5A), with hesperetin displaying particularly strong binding, thus confirming its superior binding capability (Figure 5B–E).

FIGURE 5.

Molecular docking analysis. (A) Heatmap of binding energies between principal CMH components and core targets. (B–E) Representative molecular docking conformations of hesperetin with core targets, illustrating ligand–receptor interactions and binding conformations.

3.8. CMH Accelerates DW Healing In Vivo by Suppressing the Inflammatory Response via Modulation of the IL‐17A/NF‐κB Pathway

Compared with the db/m group, db/db mice displayed significantly delayed wound healing. In contrast, the CMH group exhibited higher wound healing rates at multiple time points, with statistically significant differences evident from Day 7 onward (Figure 6A–E). H&E and Masson staining revealed thinner epidermis, marked inflammatory infiltration, disorganised tissue architecture and diminished collagen deposition in db/db mice. Conversely, the CMH group demonstrated restored epidermal integrity, significantly reduced inflammatory cell infiltration and markedly enhanced collagen deposition, approaching levels observed in the db/m group (Figure 6F–I). During the experiment, there was no change in the blood sugar or body weight of the mice in each group (Figures S1 and S2).

FIGURE 6.

CMH facilitates diabetic wound healing in vivo. (A) Schematic illustration of the in vivo experimental protocol used to evaluate diabetic wound healing. (B) Representative wound images at various time points post‐wounding (scale bar = 5 mm). (C) Schematic overlays of wounds at different time points to visualise healing progression. (D) Wound size rate at different time points. (E) Wound closure rate at different time points. (F) Haematoxylin and eosin (H&E) staining of skin tissues from the wound area. Scale bars: 1 mm (upper panel); 200 μm (lower panel). (G) Quantitative analysis of epidermal thickness. (H) Masson staining of skin tissues from the wound area. Scale bars: 1 mm (upper panel); 200 μm (lower panel). (I) Quantitative analysis of collagen deposition. Data are expressed as mean ± standard deviation (SD) (n = 6 per group). Statistical significance was determined by one‐way analysis of variance followed by Tukey's post hoc test. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001; ns, not significant.

To explore the mechanism of CMH in DW healing, we collected tissue samples from the wound margins and conducted IF, IHC, qRT‐PCR and WB analyses. IF and IHC results showed that CMH treatment markedly reduced TNF‐α and ROS expression in DW, while enhancing CD31 expression (Figure 7A–F). qRT‐PCR validated the differential expression of key targets predicted via network pharmacology, including EGFR, IL‐1β, MMP9 and PTGS2. The results demonstrated that CMH significantly downregulated IL‐1β, MMP9 and PTGS2 expression in DW. EGFR expression increased slightly but was not statistically significant (Figure 7G–J). Additionally, WB analysis revealed that CMH significantly inhibited the IL‐17A/NF‐κB pathway (Figure 7K–L).

FIGURE 7.

Effects of CMH on inflammatory responses and angiogenesis. (A) Immunohistochemistry staining of TNF‐α; scale bars: 200 μm. (B) Quantitative analysis of TNF‐α in different groups (n = 6 per group). (C) Immunohistochemistry staining of CD31; scale bars: 200 μm. (D) Quantitative analysis of CD31 in different groups (n = 6 per group). (E) Immunofluorescence staining of reactive oxygen species (ROS); scale bars: 200 μm. (F) Quantitative analysis of ROS in different groups (n = 6 per group). (G–J) Quantitative analysis of epidermal growth factor receptor (EGFR), interleukin‐1β (IL‐1β), matrix metallopeptidase 9 (MMP9) and prostaglandin‐endoperoxide synthase 2 (PTGS2) mRNA (n = 3 per group). (K) Western blot (WB) analysis of IL‐17A/NF‐κB signalling pathway‐related targets. (L) Quantitative analysis of protein/GAPDH (n = 3 per group). Statistical significance was determined by one‐way analysis of variance followed by Tukey's post hoc test. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001; ns, not significant.

Collectively, these findings suggest that CMH modulates the IL‐17A/NF‐κB signalling pathway, attenuates wound inflammation and ROS, promotes angiogenesis and thereby expedites DW healing.

4. Discussion

A growing body of evidence highlights the unique advantages of TCM in wound management [15]. Honey, with a history of topical application spanning 5000 years, was successfully used by Hippocrates—the father of ancient Greek medicine—to treat skin diseases, ulcers and burns [16]. In ancient China, Li Shizhen's Compendium of Materia Medica (Bencao Gangmu) documented that honey alleviates pain immediately when applied to burns and scalds, reinforcing its long‐standing role in external therapy [17]. With advances in modern science, the therapeutic effectiveness of honey in wound healing has been further verified. Honey has also been developed into several innovative honey‐based dressings, which are now widely used for diverse wound types [18, 19]. However, the complexity of honey's composition presents challenges for research on its pharmacological mechanisms. To clarify the efficacy and mechanism of CMH in DW therapy, we adopted an integrated research strategy combining UPLC‐Q‐TOF/MS‐based component analysis, network pharmacology, transcriptomics, molecular docking and animal experiments.

Honey exhibits distinctive therapeutic strengths in wound management, particularly for individuals with DW characterised by chronic inflammation and a high susceptibility to infection, due to its low pH and abundance of phenolic acids and flavonoids, which confer antibacterial, anti‐inflammatory and antioxidant effects [20, 21]. Phenolic and flavonoid compounds in honey act as natural antioxidants that reduce ROS levels and suppress inflammatory responses in wounds by inhibiting lipid peroxidation and scavenging free radicals and peroxyl radicals [22, 23]. UPLC‐Q‐TOF/MS analysis revealed that CMH contains 52 major components, including flavonoids such as kaempferol, quercetin, baicalein, hesperetin and luteolin, as well as phenolic acids, including chlorogenic acid and ferulic acid. These compounds have been widely reported for their anti‐inflammatory, antioxidant and antimicrobial activities, which align with the recognised bioactivities of honey in wound management [24]. However, a small number of atypical honey‐related compounds, such as 2,3,5‐trichloro‐4‐(propylsulfonyl)pyridine and piperonyl butoxide, were also detected in this study. These compounds are not commonly recognised as natural metabolites of honey. Their presence may be associated with environmental exposure during raw material collection, exogenous contamination during processing or storage, packaging‐related background signals, analytical background noise, or possible database misannotation. Therefore, these compounds were retained in Table S3 only as tentative annotations generated from the non‐targeted screening workflow to preserve the completeness and transparency of the chemical profiling results. Importantly, they were not regarded as major bioactive constituents of CMH and were not used for mechanistic interpretation. Further studies using authentic reference standards and detailed MS/MS fragment matching are required to confirm their true origin and structural identity.

Network pharmacology has been widely applied in TCM research, enabling the construction of ‘drug‐target‐disease’ networks to screen for core targets by predicting interactions between drug components and disease‐related targets [25]. However, this approach carries a risk of false‐positive results. To improve the precision of target identification, we integrated network pharmacology with transcriptomics. By intersecting predicted component targets, DEGs and module genes identified through WGCNA, we identified several core targets—including EGFR, IL‐1β, MMP9 and PTGS2—that function as hub nodes within the PPI network. Both GO functional analysis and KEGG pathway enrichment indicated that CMH is involved in the regulation of inflammatory processes, with the IL‐17 signalling pathway emerging as the most significantly enriched. IL‐17, particularly IL‐17A, is a potent pro‐inflammatory cytokine that plays a key role in the pathogenesis of various inflammatory skin diseases [26, 27]. Upon binding to its receptor, IL‐17A initiates the recruitment of the adaptor protein Act1 and tumour necrosis factor receptor‐associated factor 6 (TRAF6), leading to activation of the nuclear factor‐κB (NF‐κB) pathway. This cascade upregulates the expression of inflammatory mediators such as IL‐1β, TNF‐α and MMP9, thereby amplifying pro‐inflammatory responses [28, 29]. Several TCM formulations have been shown to exert anti‐inflammatory effects by inhibiting the IL‐17 signalling pathway. For example, Wang et al. [30] demonstrated that Huayuwendan Decoction alleviates inflammation in diabetic rats by suppressing the IL‐17/NF‐κB axis. Moreover, targeting the IL‐17 pathway has recently emerged as a novel therapeutic strategy for DW repair. By disrupting the vicious cycle of ‘chronic inflammation–impaired healing’, such interventions can establish a regenerative microenvironment conducive to wound closure [31, 32]. In support of this, Zhang et al. [33] developed Zn‐DHM nanozymes that downregulate IL‐17 signalling, reduce inflammatory responses to restore immune homeostasis, thereby accelerating early wound healing in diabetic mice.

In vivo experiments confirmed that CMH alleviates wound inflammation, promotes collagen deposition and angiogenesis, thereby accelerating DW healing. These findings indicate that CMH treatment suppresses excessive inflammatory responses, enabling the healing process to progress beyond the inflammatory phase into the tissue repair stage. Molecular biological analyses revealed a chronic inflammatory state in DW, characterised by hyperactivation of the IL‐17A/NF‐κB pathway and elevated expression of inflammatory factors, including IL‐1β, TNF‐α and MMP9. These results are consistent with previous studies [34, 35]. CMH intervention effectively inhibited the IL‐17A/NF‐κB pathway, attenuated inflammation and facilitated wound repair, further underscoring the critical role of IL‐17 signalling in DW healing.

Molecular docking analysis further validated this mechanism, showing that key components of CMH exhibit strong binding affinity to core targets within the IL‐17A/NF‐κB signalling pathway, including IL‐17A, NF‐κB, IL‐1β and MMP9. Hesperetin demonstrated the most prominent binding affinity, which may be attributed to its unique chemical structure. It contains multiple hydroxyl (–OH) and a methoxy (–OCH₃) group, which can act as hydrogen bond donors or acceptors to form a stable hydrogen bond network with amino acid residues in target proteins, thereby enhancing binding affinity [36]. Moreover, as a dihydroflavone, hesperetin possesses a C2–C3 single bond, conferring molecular flexibility that enables conformational adjustment to optimise fit into target binding sites [37]. Previous studies have also confirmed that hesperetin exerts antioxidant and anti‐inflammatory effects, reduces ferroptosis, mitigates inflammation and accelerates DW healing [38].

Our findings identify the IL‐17A/NF‐κB pathway as a central mechanistic axis. KEGG enrichment analysis further revealed significant involvement of the HIF‐1 signalling pathway and the AGE‐RAGE signalling pathway in diabetic complications—both critically associated with angiogenesis and oxidative stress regulation, respectively [39]. Additionally, observed reductions in wound ROS levels and enhanced angiogenesis following CMH treatment suggest participation of the VEGF and Nrf2 pathways [24]. Collectively, these results indicate that CMH exerts pleiotropic effects across multiple healing phases, spanning inflammation and oxidative stress resolution to angiogenesis and tissue remodelling. Future studies should experimentally validate these additional pathways to fully delineate the therapeutic scope of CMH.

It is noteworthy that although Manuka honey has been extensively investigated, whereas CMH remains a traditional yet underexplored therapeutic agent. This study not only validates its efficacy but also elucidates the underlying molecular mechanisms supporting its therapeutic application. Nonetheless, this research has certain limitations. First, while the db/db mouse model is widely utilised, it may not fully replicate the complexity of human DFUs. Second, although several bioactive components have been identified in CMH, their individual roles and possible synergistic effects remain to be clarified. Future studies should isolate and recombine these compounds in vitro to systematically evaluate their individual and synergistic contributions to DW healing.

This study demonstrates that CMH facilitates DW healing through a multi‐component, multi‐target mechanism, with the IL‐17A/NF‐κB pathway acting as a central regulatory axis. These findings provide a preliminary experimental basis for the clinical application of CMH and highlight the value of integrating traditional medicinal knowledge with contemporary omics and bioinformatics strategies. Further clinical trials are essential to corroborate these findings and guide the optimisation of CMH‐based dressings for DW care.

5. Conclusions

In summary, this study helped to illustrate the principal components of CMH and its mechanism in facilitating DW healing through an integrated approach combining UPLC‐Q‐TOF/MS, network pharmacology, transcriptomics, molecular docking and experimental validation. CMH promotes DW healing through a multi‐component and multi‐target mechanism. It may alleviate inflammatory responses and oxidative stress by inhibiting the IL‐17A/NF‐κB pathway, thereby promoting angiogenesis and accelerating wound healing. These findings not only clarify the therapeutic potential of CMH in DW care but also provide a preliminary experimental basis for its clinical application.

Author Contributions

Yungang Hu: writing – original draft, conceptualisation, investigation and formal analysis. Yiwen Wang: methodology, investigation and formal analysis. Lin Zhi: software, visualisation and data curation. Xiaohua Hu: validation and resources. Yuming Shen: project administration and funding acquisition. Weili Du: supervision, project administration, methodology, writing – review and editing.

Funding

This work was supported by Beijing Natural Science Foundation—Changping Innovation Joint Fund, L234067 and Jiangxi Provincial Natural Science Foundation, 20232BAB216055.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Table S1: Primer sequences.

Table S2: Details of the primary antibody product.

Figure S1: The changes in blood sugar levels of the mice in each group during the experiment.

Figure S2: The Weight changes of the mice in each group during the experiment.

Table S3: The main chemical components of CMH.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.