The quality control of PZH has been performed in accordance with Chinese Pharmacopoeia Commission, 2020; Chinese Pharmacopoeia Commission, 2020). Briefly, take PZH powder (batch number: 2107111), pass through a 100-mesh sieve, weigh accurately 1 g, add internal standard solution (patchouli alcohol, 0.2 mg/mL) 2 mL, and then add 3 mL of anhydrous ethanol, ultrasonic treatment (power 300 W, frequency 40 kHz) for 10 min, weigh again and make up the lost weight with anhydrous ethanol after 2 h, 1 μL continuous filtrate was injected into the high performance liquid chromatograph. Gas chromatography was performed with a TRACE GC Ultra (Thermo Fisher Scientific, Dreieich, Germany). An HP-5 (5%phenyl/95%methyl-polysiloxane) capillary column (30 m, 0.32 mm ID, 0.25 μm film thickness) was used for separation. The PZH samples were detected under this condition: the temperature was initially 150°C for 30 min, rising to 250°C at 20°C/min for 15 min. Besides, the injection port temperature was 250°C and the detector was held at 300°C. The standard of patchouli alcohol and muscone (≥98% purity) were provided by Sichuan Weikeqi Biological Technology Co., Ltd. (Sichuan, China).

Male Sprague-Dawley rats (105 ± 15 g) were obtained from Beijing Huafukang Biotechnology Co., Ltd [SCXK (BJ) 2019-0008]. All animal care and experimental procedures were conducted according to the protocol approved by the Institute of Basic Theory for Chinese Medicine, China Academy of Chinese Medical Sciences. The project identification code is No. IBTCMCACMS21-2109-01, and the approval date of the Ethics Committee is 8 September 2021. All rats were group-housed with free access to water and food in the vivarium, which had a 12-h light/dark cycle and a temperature-controlled environment. After 3-days acclimatization, rats were fed a high-fat diet with free access to water for 4 weeks. Following 12 h of fasting, rats were injected intraperitoneally with 35 mg/kg streptozotocin (STZ) to induce diabetes. DM was diagnosed as fasting blood glucose (FBG) levels higher than 16.7 mmol/L but lower than 33.3 mmol/L. Two 2-cm diameter circular full-thickness wounds were created on each side of the dorsal surface (back of forelimb) of rats, and different drugs to the left and right holes. The rats were divided into five groups: Control, Model, PZH-L, PZH-H, and positive drug. The control group and model group were applied with physiological saline. The PZH-L and PZH-H groups were applied with PZH, at a dose of 0.05 g/cm² and 0.15 g/cm². The rhEGF was used as the positive control, at a dose of 40 IU/cm².

At 0, 6, 10, and 14 days during the treatment, the wounds were photographed and the wound closure rate was calculated as follows: A =  (B−C)/B × 100%. Where A is the rate of wound closure reduction on day 0, day 6, day 10, and day 14 post-wounding, B is the initial area of the wound, and C represents the wound area at the time of measurement. Then, the rats were anesthetized and placed on a two-photon microscope (SHG) to observe the collagen growth on the edges of the wounds.

Tissue samples were fixed in 4% paraformaldehyde for 3 days at room temperature. Tissue blocks dehydrated in graded ethanol series, cleared in xylene, embedded in molten paraffin, and sectioned (6 μm) for H&E and MTC stains. H&E staining and MTC staining were consistent with our previous study (Zhang et al., 2023). Briefly, for H&E staining and Masson staining, the skin tissue slices were carried out according to the instructions of the manufacturer’s protocol.

The skin tissue was poured into a mortar precooled and the liquid nitrogen was added to fully grind the sample into powder. Total RNA was extracted from skin tissue using the TRIzol reagent according to the manufacturer’s instructions, and the quality of RNA was evaluated by the 2100 Bioanalyzer. RNA-seq was performed in the platform of Illumina HiSeq 6000 in Novogene Bioinformatics Technology Co., Ltd. (Beijing, China). The raw data have been uploaded into the NCBI database SRA (PRJNA875293). The differentially expressed genes (DEGs) were selected according to p < 0.05 and |log2 (fold change)| ≥1 and through the volcano plot showing. The expression patterns were clustered by hierarchical clustering. Enrichment analyses were performed using DAVID, and the network of biological processes and DEGs was then constructed. The network of upregulated DEGs and downregulated DEGs after PZH treatment was constructed and enrichment analysis was performed, respectively. Besides, P. notoginseng is the main raw material of PZH, and notoginsenoside R1(NR1) is a major active component of P. notoginseng (Sun et al., 2018). Our previous research showed that NR1 facilitated wound healing by alleviating apoptosis and inflammation, at the same time increasing ECM remodeling and promoting angiogenesis (Cao et al., 2022). Besides, our previous research has proved that berberine (BBR) has a good therapeutic effect on diabetic ulcers. BBR promotes wound closure by decreasing ROS, oxidative stress, apoptosis, and increasing cell proliferation (Zhou et al., 2021). Importantly, the study of NR1, BBR, and PZH is consistent in the establishment of the type II diabetic ulcers model. To explore the mechanism of PZH against diabetic ulcers, we further compared and analyzed the RNA-seq data of NR1, BBR, and PZH. The RNA-seq data of NR1 and BBR were downloaded from the NCBI platform (BioProject PRJNA757790). DEGs were defined with the same conditions as PZH, and then the enrichment analyses were performed by Metascape. Finally, a network of DEGs and the unique biological process of PZH was constructed by STRING and visualized by Cytoscape 3.9.1 software.

Skin tissue was lysed with RIPA buffer, the protein concentration was measured by BCA protein assay, and equal protein amounts were added electrophoretically separated on precast 4%–20% Precast Gels (Meilun) and then transferred onto PVDF membranes. Following incubation with 5% nonfat milk for 2 h, the membranes were treated with primary antibodies overnight at 4°C. The primary antibodies were used as follows: RelB (10544S, CST, 1:1000), IκBα (ab76429, Abcam, 1:1000), and NKG2-D (sc-515599, Santa Cruz, 1:1000). After washing three times, the corresponding secondary antibody were added and incubated at room temperature for 2 h. Finally, the protein bands were visualized with a Gel Doc imaging system and analyzed with the ImageJ software.

We evaluated the expression of transforming growth factor-β1 (TGF-β1) (E-EL-0162c, Elabscience), matrix metallopeptidase 9 (MMP9) (E-EL-R3021, Elabscience), interleukin-1β (IL-1β) (E-EL-R0012c, Elabscience), toll-like receptor 2 (TLR2) (E-EL-R0907c, Elabscience), Interferon Gamma (IFN-γ) (E-EL-R0009c, Elabscience) in rat skin tissue by ELISA. Briefly, skin tissue was homogenized by a tissue homogenizer in RIPA lysis buffer on ice. After centrifugation at 12,000 rpm for 15 min at 4°C, supernatants were collected and prepared for the ELISA experiment. The samples were incubated on microtiter plates for 1.5 h, followed by biotinylated secondary antibodies for 1 h. Next, the slides were incubated with a Horseradish Peroxidase (HRP) conjugate peroxidase working solution for 30 min. Finally, tetramethylbenzidine (TMB) substrate solution was added, and after 15 min incubation in the dark, stop solution was then added to the microtiter plates. Absorbance was measured at 450 nm using a microplate reader.

Briefly, the sections had been repaired with the antigen repair solution, observed by using 0.5% Triton X-100 permeabilized, and blocked with 5% (w/v) bovine serum albumin (BSA). Then the sections were incubated with primary antibodies at 4°C overnight. Primary antibodies used were as follows: TGF-β1 (ab215715, Abcam, 1:200), MMP9 (ab58803, Abcam, 1:200), Platelet endothelial cell adhesion molecule-1 (CD31) (M1511-8, Huabio, 1:200), Cleaved caspase-3 (ab179517, Abcam, 1:200), NKG2-D (sc-515599, Santa Cruz, 1:200), and nucleotide-binding oligomerization domain containing 2 (NOD2) (sc-56168, Santa, 1:200). The slides were incubated with secondary antibodies for 60 min. Finally, nuclei were counterstained with 4′,6-diamidino-2-phenylindole (DAPI) (c0060, Solarbio), and the slides were visualized by fluorescence microscopy.

The chemical composition of PZH and the drug targets were obtained from the BATMAN-TCM database 2.0 (http://bionet.ncpsb.org.cn/batman-tcm/) and related research (Cao et al., 2021). The score cutoff  ≥ 80.88 and p adj < 0.05 were considered as indicators when screening the active compounds of PZH. Then, the drug targets related to PZH-regulated targets were obtained through the STRING database 12.0 (https://string-db.org/cgi/input.pl). Next, the network of active ingredients-drug targets-PZH-regulated targets was constructed and the active components were identified.

All the data have been subjected to a one-way analysis of variance (One-Way ANOVA) with the use of GraphPad Prism 9.0. Values are expressed as mean ± SD. p < 0.05 was considered statistically significant.

The characteristic peaks of the patchouli alcohol and muscone of standard and PZH samples are shown in Figures 1A, B. The content of muscone was 0.435mg, which confirmed the rules of 2020 Chinese pharmacopeia.

The experimental flow chart is shown in Figure 2A. And the wound healing efficacy of PZH was established by investigating the wound closure rate and collagen expression. The results showed that the wound closure rate of the hdfSTZ group was significantly decreased on days 6, 10, and 14. On postoperative days 6, 10, and 14, the wound closure rate of the diabetic rats was significantly increased by PZH treatment (Figures 2B, C). As shown in Figure 2D, the SHG result showed that collagen expression around the edge of the healing wounds drastically accelerated in the hfdSTZ+PZH-H group and hfdSTZ+rhEGF group. These results indicated that topical PZH treatment enhanced wound healing in diabetic rats.

ECM secretion and collagen expression in response to PZH administration were evaluated by H&E staining and Masson’s trichrome staining. H&E staining results demonstrated that the re-epithelialization of the hfdSTZ group was lower in the hfdSTZ group than in the control group (red arrow), whereas the re-epithelialization was increased after treatment with PZH and rhEGF (black arrow) (Figures 3A, C). In addition, the results indicated significantly higher blood capillary number of the diabetic rats in the hfdSTZ+PZH-L group, hfdSTZ+PZH-H group, and hfdSTZ+rhEGF group than in the hdfSTZ group (black triangle) (Figures 3A, D). Furthermore, the ECM secretion in the hfdSTZ group was lower than that in the control group, but ECM secretion was significantly increased in diabetic rats treated with PZH (Figures 3A, E). As shown in Figure 3B, MTC staining showed the largest amounts of collagen expression in PZH-treated wounds compared to untreated hfdSTZ wounds.

To investigate the mechanism underlying PZH against diabetic ulcers, RNA-seq was performed. RNA-seq analysis results showed that the hfdSTZ group had 899 upregulated DEGs and 1035 downregulated DEGs compared with the control group (Figure 4A). The hfdSTZ+PZH group had 839 upregulated DEGs and 900 downregulated DEGs compared with the hfdSTZ group (Figure 4B). A heatmap analysis illustrated that the hfdSTZ+PZH groups were more similar to the control group than to the STZ group (Figure 4C). As shown in Figures 4D, E, the DEGs of the hfdSTZ group were enriched in the inflammatory response, extracellular matrix, TNF signaling pathway, NF-kappa B signaling pathway, apoptotic process, angiogenesis, wound healing, etc. Whereas biological processes such as immune response, inflammatory response, apoptotic process, NF-kappa B signaling pathway, chemotaxis, PI3K-AKT signaling pathway, wound healing, angiogenesis, and extracellular matrix were significantly enriched after PZH treatment. Then, the network of DEGs enriched in biological processes associated with diabetic ulcers was constructed, such as apoptotic process, inflammatory response, wound healing, extracellular matrix, and angiogenesis (Figure 4F). The hfdSTZ group rats had different gene expression profiles compared to the control group rats, and the hfdSTZ+PZH group had more similar gene expression to the control group, respectively. The enrichment analysis showed that DEGs were not only associated with these biological processes, but also had a close connection with programmed cell death, interleukin-6 production, tissue remodeling, and blood vessel development, respectively.

Next, the key biological processes associated with diabetic ulcers were validated. The levels of MMP9, Cleaved caspase-3, TGF-β1, and CD31 were measured. The IF staining results demonstrated that TGF-β1 and CD31 were significantly decreased while the expression levels of MMP9 and Cleaved caspase-3 were significantly increased in diabetic rats. After PZH treatment, the expression of TGF-β1 and CD31 was significantly increased and the expression of MMP9 and Cleaved caspase-3 was significantly decreased in the tissue of diabetic rats (Figures 5A–H). Besides, compared with the control group, the expression level of IL-1β in the hfdSTZ group was significantly increased. Compared with the hfdSTZ group, the level of IL-1β in the hfdSTZ+PZH-L group and hfdSTZ+PZH-H group were significantly reduced (Figure 5I). The results showed that PZH treatment inhibited apoptosis and inflammation, and increased the ECM remodeling and angiogenesis to accelerate wound healing.

To identify the vital targets of PZH against diabetic ulcers, the network of upregulated DEGs and downregulated DEGs after PZH treatment was constructed. As shown in Figures 6A, B, the targets of Rps27a, Gfap, and Nes are with the highest degree in the hfdSTZ group, and the targets of Il10, Il1b, Ifng, Il6, Tlr2, Klrk1, and Nod2 are with the highest degree in the hfdSTZ+PZH group. As shown in Figures 6C, D, the upregulated DEGs were associated with processes related to the extracellular matrix, whereas the downregulated DEGs were associated with processes related to inflammation. Through the comparative analysis of the network, it can be seen that the number of downregulated targets was more than the number of upregulated targets. Besides, the network of downregulated DEGs showed an increased probability of protein-protein interaction. Therefore, downregulated DEGs are more important than upregulated DEGs in PZH against diabetic ulcers. The results reaffirm the importance of inflammation response. We further compared the effect of PZH on diabetic wounds with that of BBR and NR1, since the same diabetic model was used. As shown in Figure 6E, 184 DEGs were found in common between the PZH group and NR1 group, 126 DEGs were found in common between the PZH group and BBR group, and 270 DEGs were found in common between the NR1 group and the BBR group. Furthermore, 55 DEGs were found in common between the three drugs. As shown in Figures 6F, G, compared with the BBR or NR1, 3 unique GO terms were found in the PZH group, such as defense response to the bacterium, immune response-regulating signaling pathway, and regulation of response to biotic stimulus. The network of these unique PZH-based biological processes network and enriched DEGs was constructed (Figure 6H). We hypothesized that targets involved in more of these three biological processes may be relatively more important because it has more functions. Thus, 9 targets including Klrk1, Nod2, Tlr2, Oas1a, Ifng, Nlrc4, Tlr9, Oas3, and Irgm were identified as potential vital targets (Figure 6I). The result of the Wayne diagram showed that 5 DEGs overlapped in the downregulated DEGs and unique DEGs in PZH (Figure 6J). Therefore, the targets of klrk1, Nod2, Tlr2, Ifng, and Tlr9 were identified as vital targets for PZH intervention in diabetic ulcers.

Numerous studies demonstrate that Klrk1, Nod2, Tlr2, and Ifng are associated with various skin diseases, and excessive production of Klrk1, Nod2, Tlr2, and Ifng affects the prognosis and treatment of skin diseases (de Oliveira et al., 2016; Jiao et al., 2016; Jacquemin et al., 2020). As shown in Figures 7A–C, Figures 7E, F, and Figures 7I, J, the results showed that NKG2-D, NOD2, TLR2, and IFN-γ were markedly increased in the hfdSTZ group, whereas topical PZH treatment significantly decreased the expression of NKG2-D, NOD2, TLR2, and IFN-γ. Importantly, the increased expression of NOD2, TLR2, and NKG2-D could lead to downstream activation of NF-κB signaling pathway (Figure 8). As shown in Figures 7D, E, Figures 7G, H, RELB and IκBα were significantly decreased and nucleus-p65/cytoplasm-p65 was significantly increased in the hfdSTZ group. However, the expression of RELB and IκBα were significantly increased after PZH treatment, while the expression of nucleus-p65/cytoplasm-p65 was decreased.

To identify the active ingredients of PZH against diabetic ulcers, network analysis was performed. There were 609 drug targets of the 318 compounds of PZH were collected. And the network of active ingredient-drug targets-PZH-regulated targets was constructed, which included 19 active ingredients, 42 drug targets, and 12 PZH-regulated targets (Figure 7K). And there were 7 intersection targets between drug targets and PZH-regulated, including CASP3, IL1B, MMP9, IFNG, RELA, NFKBIA, and PECAM1. Besides, the active ingredients mainly included NR1, L-Lysine, Deoxycorticosterone, β-sitosterol, Ursolic acid, etc.