Screening of GD targets and DF disease targets

Searches of the TCMSP and BATMAN-TCM databases and published literature revealed that GD contains 25 traditional Chinese medicine active ingredients. Among these active ingredients, 17 compounds had no known potential targets, while 8 compounds had many targets. Some molecules are targets of many kinds of compounds (Additional file 1: Table S2), and 202 targets were identified after removing duplicates. DF disease targets were identified by searching the GeneCards databases. After merging and deleting duplicate targets, a total of 3601 disease targets remained. By overlapping the predicted targets of GD and DF with a Venn diagram, we ultimately identified 140 potential targets of GD in the treatment of DF (Fig. 2).

Fig. 2

Target Venn diagram of GD and DF

Construction of a drug-active ingredient-target network of GD in the treatment of diabetic foot

To understand the multicomponent pharmacological mechanisms underlying the effects of GD, we constructed an herb-compound-target (H-C-T) network. Information about drug components and effective targets as well as other information about GD was imported into Cytoscape 3.7.2 software to construct a drug-active ingredient-effective target network diagram (Fig. 3). This network consisted of 153 nodes and 180 edges. The compounds with the highest degree values were quercetin, uridine and beta-sitosterol. These compounds may be the key compounds for the treatment of DF.

Fig. 3

Herb-Compound-Targets network. The inverted V-shaped pattern is the drug molecule, and the ellipse-shaped is the drug targets, and its color is the same as the color of the component it belongs to

Construction of the PPI network

In further analysis, the 140 targets identified by the intersection described above were uploaded to the String database (Additional file 1: Table S3), and the species "Homo sapiens" was selected to construct a PPI network (Fig. 4A). The downloaded file was imported into Cytoscape to calculate the degree value and reconstruct the PPI network. The network included 139 key nodes and 4196 edges, and the highest node value was 94. Degree value and betweenness centrality (BC) were used to perform the screening. The top core genes were used to generate as a histogram (Fig. 4B). The screening process is shown in Fig. 4C. The top 6 core network proteins that were identified included AKT1, TP53, IL6, CASP, TNF and VEGFA.

Fig. 4

Identification of core targets for GD against DF. A Target protein interaction network (PPI). B The top 9 hub genes histograms. C The process of topological screening for the PPI network. The yellow nodes represent the core targets, and the other nodes represent the noncore targets

GO and KEGG pathway enrichment analysis

GO and KEGG enrichment analyses of the 140 common targets were performed using Metascape. A total of 516 significantly related GO enrichments were screened, including 272 biological processes (BP), 106 cellular components (CC), and 138 molecular functions (MF). The results were ranked according to P < 0.01, count > 3 and Rich factor > 1.5, and 20 significantly enriched BP, CC and MF were identified. The top 10 enriched items are shown in Fig. 5A–C. According to the aforementioned method, the top 20 pathway enrichment results were determined. Subsequently, these pathways were classified and summarized according to KEGG pathway analysis, and the results are shown in Fig. 6. These pathways mainly include pathways in cancer, the AGE-RAGE signaling pathway in diabetic complications, lipids and atherosclerosis, and the most relevant pathway in DF disease is the AGE-RAGE signaling pathway in diabetic complications.

Fig. 5

GO enrichment analysis. A Biological Process. B Cell Component. C Molecular Function

Fig. 6

KEGG pathway enrichment analysis

Molecular docking

We selected the key genes AKT1, TP53, IL-6, CASP3, TNF-α and VEGEA that were identified in the previous steps and carried out molecular docking with uridine and quercetin, the main pharmaceutical components in GD. The docking energy results are shown in Fig. 7A. The results suggested good docking activity (binding energy less than 1.2 − kcal/mol). All the components mainly interacted with the corresponding targets through hydrogen bonds. Quercetin interacted with GLU-9, LYS-8 and GLU-98 of AKT1 (Fig. 7B). Quercetin interacted with ASN-239, ASN241 and LYS-240 of TP53 (Fig. 7C). Quercetin interacted with LEU-64, LEU-62, THR-162 and GLN-154 of IL-6 (Fig. 7D). Quercetin interacted with HIS-121, GLN-161, SER-205 and ARG-207 of CASP3 (Fig. 7E). Quercetin interacted with THR-218, GLU-137, ALA-133 and ASN-106 of TNF-α (Fig. 7F). Quercetin interacted with THR-145, VAL-135 and VAL-147 of VEGFA (Fig. 7G). Uridine interacted with LEU-52, GLU-40, GLN-47 and ALA-50 of AKT1 (Fig. 7H). Uridine interacted with GLN-98, TRP-76, LYS-168 and ASN-75 of TP53 (Fig. 7I). Uridine interacted with ARG-104, GLU-42, LYS-46, ASP-160 and SER-47 of IL-6 (Fig. 7J). Uridine interacted with LYS-242, ARG-241 and ARG-238 of CASP3 (Fig. 7K). Uridine interacted with GLN-267, VAL-233, GLY-26 and GLU-237 of TNF-α (Fig. 7L). Uridine interacted with TYR-21, SER-24, ASN-62 and CYS-61 of VEGFA (Fig. 7M). The abovementioned ligands could be well embedded in the active pockets of the receptor target proteins.

Fig. 7

Molecular binding mode diagram. A The results of binding energy. B–G Molecular docking results of Quercetin and core protein. H–M Molecular docking results of Uridine and core protein

GD extract accelerates wound healing in rats with DF at different time points

On the 1st, 4th, 7th, 14th, and 21st days after model establishment, the foot wounds of the rats were photographed for statistical analysis, and the results are shown in Fig. 8. After 21 days of gavage, compared with those in the model group, the foot wounds in the normal group were basically completely healed (P < 0.001). The medium and high doses of GD improved the diabetic foot wound healing rate at each time point (P < 0.05, vs. DF group), while the low-dose extract had no significant effect on the wound healing rate. In addition, as we expected, all dose groups of GD could reduce the blood glucose of diabetic rats (Additional file 1: Table S4).

Fig. 8

Wound healing of different time points in DF rats. A Images of rat foot wounds at different time points. B The healing rate of foot wounds at different time points in DF rats. *P < 0.05, **P < 0.01, ***P < 0.001 vs. DF

GD extract reduces serum levels of inflammatory factors in rats with DF

According to the network pharmacology results, GD may affect the levels of IL-6 and TNF-α to inhibit the occurrence and development of inflammation. Compared with those in the control group, the levels of the serum inflammatory factors IL-6 and TNF-α in the DF group were increased. Oral administration of GD significantly reduced the serum levels of IL-6 and TNF-α in the rats (Fig. 9).

Fig. 9

Effects of GD on Inflammatory Factors IL-1 and TNF-α in DF Rats. A IL-6 levels in serum. B TNF-α levels in serum. ###P < 0.001, vs. CON, **P < 0.01, vs. DF

GD extract enhances the mRNA and protein expression of VEGF in rats with DF

Increased expression of VEGF can enhance the growth of vascular endothelium and promote the healing of ulcers. Naturally, we examined the mRNA and protein levels of VEGF. As shown in Fig. 10, compared with the control group, the mRNA and protein expression of VEGF in the DF group was increased. After treatment with GD, the expression of VEGF in the high-, medium- and low-dose GD groups was significantly increased.

Fig. 10

Effects of GD on the expression of VEGF mRNA and protein in lower limb ischemic tissue of DF rats. A The mRNA expression of VEGF. B Western blot results of VEGF and GAPDH. C Quantitative Statistics of VEGF/GAPDH. #P < 0.05, vs. CON, *P < 0.05, **P < 0.01, vs. DF