The curative effects of the traditional Chinese herbal medicine “Jinchuang ointment” on excisional wounds

Chemically synthesized borneol (purity: 98%) and camphor (purity: 96%) in compliance with the Chinese Pharmacopoeia were bought from Cheng Yi Chemical Co., Ltd. (Taipei, Taiwan). Frankincense (resin from Boswellia sacra Flueck) was the product of Daily Health Co., Ltd. (Taipei, Taiwan) (Batch number: BCGY161229). Myrrh (resin from Commiphora myrrha Engler) was the product of Hou Chuia Biopharm Co., Ltd. (Tainan, Taiwan) (Batch number: ULS031). “Bao chu Brand” dragon’s blood (dried fruit powder from Daemonorops draco) was the product of Meida Co., Ltd. (Singapore). Catechu (dried evaporated decoction from Uncaria gambir Roxb) was the product of Hing Zong Co., Ltd. (Kaohsiung, Taiwan) (Batch number: 213N-107-02). Identifications of above plant-based materials by HPLC were reported below. Vascular endothelial growth factor (VEGF) was bought from B&D Systems (Minneapolis, MN, USA). Neomycin ointment was obtained from Genuine Chemical Pharmaceutical Co., Ltd. (Taoyuan, Taiwan). Food-grade lard was obtained from President Nisshin Corporation (Tainan, Taiwan). Food-grade sesame oil was obtained from Fwusow Industry Co. Ltd. (Taichung, Taiwan). Beeswax (white, technical grade) was obtained from KahlWax (Trittau, Germany). Paraffin wax was obtained from Chenyi Chemical Co. (Taipei, Taiwan). The composition of the three different types of Jinchuang ointment that used in this study is listed in Table 1.

All experiments were carrying out on a Hitachi L-7000 HPLC system, equipped with L-7100 quaternary gradient pump and a L-7450 photo diode array detector. Hitachi HSM software was used for machine controlling, data collecting and processing. A μBondapak™ C18 Column, 125 Å, 10 μm, 3.9 × 300 mm, analytic column (Waters Corporation, Milford, Massachusetts, USA) was used for analysis. Sample preparation, and HPLC analytical conditions were reported previously [1].

The clinical case that presents in this study took place in an emergency room. Because using this herb medicine to treat wounds is a routine treatment method in this hospital, the Institution granting permission for the treatment is not required. Both the consent form from the patient and the Institution granting permission (Number: 1090114004) from China Medical University Beigang Hospital for using exist data in case report were included.

The Institutional Animal Care and Use Committee, National Laboratory Animal Center, Taiwan, ROC, approved all experimental animal procedures (Permission number: NLAC(TN)-107-M-010). Three female Lee-Sung pigs without skin disease were used in this study. They were obtained from the Department of Animal Science and Technology, National Taiwan University, Taiwan. The pigs were 4.5 months old with similar body shapes and an average weight of 15.8 kg. Three weeks before the experiments, the pigs were moved from an animal room located in the Department of Animal Science and Technology, National Taiwan University, by trunk to a laboratory. Each animal was maintained in an isolated room (1.56 m²) with polished concrete floors at 21 ± 2 °C and 30–70% humidity under a 12-h light and dark cycle. The room was cleaned twice a day. The animal filed was sterilized by 0.06% sodium hypochlorite once per week. Animals will be sacrificed prior to, during, or after the experiments, if animals cannot eat for 5 days, cannot stand for 24 h, or body temperature is lower than 35 °C.

General anesthesia was maintained by isoflurane via inhalation. Zoletil (5 mg/kg) and Xylazine (2.2 mg/kg) were intramuscularly injected and Atropine (0.05 mg/kg) was subcutaneously injected for sedation. Five minutes later, Isoflurane (0.4–2.0%) inhalation and Lidocaine (8%) spray anesthesia were started and continued for the duration of the operation (Fig. 1a). Subcutaneous injections of Buprenorphine (0.04 mg/kg) and oral administration of Meloxicam (0.4 mg/kg) were given for the relief of pain. Enrofloxacin (5 mg/kg) was intramuscularly injected and Chalexin (1.5 mg/kg) was administrated orally. The whole procedure was performed in an aseptic manner with uniform dermal wounds created among three minimally diseased adult female Lee-Sung pigs in the morning. A biopsy punch with a 10 mm diameter a 7 mm depth (Lot no: 18271, Robbins Instruments, Chatham, New Jersey, USA) was employed to create full-thickness dermal excisions from four different dorsal sites on each animal. Each pig received the same 12 wounds and wounds were separated from each other by 2 cm (Fig. 1b). Sterile gauze was used to stop bleeding and keep wounds clean. Test ointments (TA1, TA2, and TA3) and the neomycin ointment control (CA) were directly applied to the wound surface, and dressings were replaced daily (Table 2). The animals were sacrificed with pentobarbital overdose (120 mg/kg) at day 28. The wound sites were photographed at 7-, 14-, and 28 days post-surgery. The unhealed wound diameter was recorded and the blood flow flux at all wound sites was measured by High-Resolution Laser Doppler Imaging moorLDI2-HIR (Moor Instruments Inc., Wilmington, DE, USA). Results are expressed as the mean ± SD from three different wound sites. Differences between groups were assessed by one-tailed test. p value less than 0.05 was considered statistically significant.

The dermal wound tissue was sampled via a 6 mm biopsy punch (Lot no: 17L13, Integra LifeSciences, Plainsboro, NJ, USA) and preserved in 10% neutral buffered formalin (NBF) at 7-, 14- and 28-days post-surgery. After fixation, the tissues were trimmed, embedded, and sectioned 4–6 mm thick onto glass slides (Immuno Coated slide, MUTO, Japan). For the hematoxylin and eosin (H&E) stain, the procedures were carried out using automated staining equipment (DRS-2000, Sakura, Japan) with routine protocols. Meanwhile, the modified Masson’s trichrome (MT) stain was conducted manually following the manufacturer’s instructions (TRM-1, Scytek Laboratories, USA). A five-phase scoring system for semi-quantitative evaluation of histopathological observations was employed in this study (Additional file 1: Table S1) [8]. For dermal wound healing, a five-phase categorization was used (Additional file 1: Table S2) [9]. Finally, a five-point scoring system for recording histopathological observations was employed (Additional file 1: Table S3) [10].

For the immunohistochemistry (IHC) stain, the slides were dewaxed in xylene and rehydrated in PBS. The antigen retrieval was performed by incubating the slides in a pressure cooker utilizing commercial unmasking buffer (H-3300, Vector lab, USA). Sections were treated with endogenous peroxidase and ALKP blocking solution (SP-6000, Vector lab, USA) in order to block endogenous peroxidase activity for 10 min at room temperature and with protein blocker (K405-50-2, Biovision, USA) for 30 min. Slides were incubated with 10 μg/mL rabbit anti-VEGF serum (PAA143Po01, Cloud-Clone Corp, USA) in primary antibody diluent (K405-50-3, Biovision, USA) for 1 h at room temperature, followed by 15 min incubation with a goat anti-rodent/rabbit polymer visualization system (K405-50-4, Biovision, USA). Additionally, 3, 3′-diaminobenzidine (DAB, Biovision, USA) was used as a chromogen for color development for 6 min. Slides were counterstained with hematoxylin and observed under light microscopy. Slides were scored based on the signal intensity in the section: − (negative), + (weak), ++ (moderate) and +++ (strong). A modified four-point scoring system for qualitatively recording IHC observations was used (Additional file 1: Table S4) [11].

Cytotoxicity assay was performed using a 24-well plate. 4 * 10⁴ C2C12 cells were seeded into each well in complete medium followed by required treatment. Cell numbers were measured by counting after “Jinchuang ointment” treatment for 24 h. Relative survival rate was calculated and shown as mean ± SD, taking the value of DMSO sample as 100%. Through one tailed test analysis, * and ** show statistical significance (p < 0.05 and p < 0.005, respectively) compared with DMSO and represents two reproducible results.

The participant, Miss Su, was a 25-year-old female nurse working at China Medical University Beigang Hospital. On September 5, 2015, 1:47 PM in the hospital during work, her four fingers were accidentally and simultaneously cut by a scalpel (Fig. 2a). Surgical suturing was performed in the emergency room at 2:19 PM. After suturing, her index, middle, and ring finger were received three, five, and three stitches (Fig. 2b), respectively. Topical neomycin antibiotic ointment was then applied over the sutures to prevent infection (Fig. 2c). Neomycin is the first-line antibiotic used in wound treatment [12], as requested by the National Health Insurance Administration, Ministry of Health and Welfare, Taiwan.

On September 7th, instead of using neomycin, “Jinchuang ointment” was applied over the stitches on her middle, ring, and little finger. However, neomycin ointment was still applied over the suture on her index finger (Fig. 2d). After 11 days, the wounds on her middle, ring, and little finger were healed entirely, while the wound on the index finger was not (Fig. 2e). This treatment is the first clinical case to show that “Jinchuang ointment” significantly promotes wound healing when compared to the standard surgical treatment for traumatic cutaneous wounds.

The content of plant materials varies from batches to batches. We had to use HPLC to analyze the content of reference standards in dragon’s blood, catechu, frankincense, and myrrh that were used in this study, even though the method was published [1]. The mass percentage of dracorhodin, catechin and epicatechin, acety-11-keto-β-boswellic acid, and (E)-guggulsterone in the dragon’s blood, catechu, frankincense, and myrrh used in this study is 0.06%, 3.12% and 1.79%, 2.08%, and 0.02% (Additional file 1: Table S5). The separation of reference standards in aforementioned four herbal products by HPLC was shown in Additional file 1: Figure S1.

No animals were found dead or moribund during the study period. Figure 3a shows the wound appearance of control (CA) and Jinchuang ointment-treated (TA1, TA2, TA3) groups. The wound closure rate was assessed by tracing the wound on day 7, 14, and 28. Changes in wound area were evaluated, taking the initial size of the wound on day 0 as 100%. The average remaining unhealed wound area in the CA, TA1, TA2, and TA3 groups on day 14 was 38%, 3%, 2%, and 1%, respectively. These results revealed that at an intermediate time, day 14, the Jinchuang ointment treated wounds were significantly smaller compared to the control (Fig. 3b). Both treated and untreated wounds appeared to be almost closed on day 28.

A laser Doppler burns imager was used to monitor the blood flow flux in the excisional wounds of the swine skin (Fig. 4a). Taking the initial blood flow flux at each wound site on day 0 as 100%, the average blood flow flux of TA1, TA2, and TA3 was about 114%, 100%, and 74% higher than that of CA on day 28 (Fig. 4b)

The created wounds did not become clinically infected throughout the entire period of this study. The overall histopathologic scores for different treatments are presented in Additional file 1: Tables S6–S9.

The results of the Masson Trichrome staining, H&E staining and IHC staining are shown in Figs. 5, 6 and 7, respectively. On day 7, epithelial migration and proliferation were observed in TA1, TA2, and TA3 with 50% less epithelization. The CA group neither induced remarkable epithelial modification nor neovascularization in the wound area. At this time point, every sample in each group was considered to be in the “Inflammation” phase with diffuse polymorphonuclear (PMNL) cells and active fibroblast infiltration. All of the samples from TA1, TA2, and TA3 had brownish particles in the granulation tissue with sporadic and scattered distribution (Fig. 6). On the other hand, the Masson Trichrome (MT, Fig. 5) stain in this stage did not label significant collagen deposition within the wound area and the IHC stain showed that VEGF had been secreted from the fibroblasts (Fig. 7).

On day 14, the epithelium had initiated its restoration with over 50% migration from the edge in each treatment. The results of the MT stain indicated that there was no significant difference in the formation of collagen in all groups (Fig. 5). Meanwhile, sporadically distributed brownish particles could still be found in all groups. As shown in Fig. 6, the area of granulation tissue in CA group was higher than in the other three test groups at day 14. In contrast, the area of the granulation tissue in the TA1 group was the lowest among all groups, suggesting that the wounds in the TA1 group healed better at this stage. The IHC results showed that the VEGF in the CA group was the lowest among all groups (Fig. 7).

On day 28, the epithelium was completely regenerated in TA1 and TA2 with an intact epidermis and establishment of an epidermal ridge (Fig. 5). However, in the CA and TA3 groups, each had one sample which had not keratinized completely and only showed surface bridging. When comparing the TA1 and TA3 treatments under MT stain, the TA1 wounds showed more collagen deposits, similar to normal skin, whereas CA wounds secreted less collagen than TA2 or TA3 in the same categories. Meanwhile, the IHC results indicated the VEGF in all three test groups was higher than that in the CA group (Fig. 7). In accordance with these results, the average blood flow in all three test groups was higher than that in the CA group (Fig. 4b). Therefore, at this time point, the healing progression was TA1 > TA2 > TA3 = CA. Neither lard-containing Jinchuang ointment (TA1) nor sesame oil-reconstituted Jinchuang ointment (TA2) possesses cytotoxicity in muscle precursor cell line, C2C12 (Fig. 8).