To establish a reliable skin wound array model, techniques to create wounds, parameters of optimal wound locations, wound sizes, the distance between adjacent wounds, and analysis methods need to be explicitly defined and comprehensively evaluated. Establishing multiple wounds in an array using a singular animal can increase the consistency of results and resolve errors induced by individual animal differences. This can offer researchers a stable platform, with the required reproducibility of wound healing profiles, to reliably screen the effectiveness of various treatments.
Previously20, we demonstrated that Lanyu pigs are a more suitable model than Landrace pigs for evaluating the skin wound healing process, owing to their body size and body weight approaching that of humans, with stable wound healing features, especially after puberty. Studies have shown that porcine skin is similar to human skin15 in terms of epidermal thickness, subcutaneous composition, and healing processes; this was confirmed in our present study using Lanyu pigs (Fig. 2C–F). Our histological findings demonstrated the homogeneity of the skin structure and thickness at different locations on the dorsum of Lanyu pigs (Fig. 2), confirming its value as an experimental platform for wound healing.
Meeh’s formula35: body surface area (BSA) = 8.58 × body weight2/3.
According to the above Meeh’s formula, the body surface of a Lanyu pig ranges between 0.73 and 1.46 m2 after puberty, weighing between 25 kg at 5 months to 70 kg at 2 years of age. One side of the dorsal skin available for wounds is approximately 4.1–7.7% of the BSA, measuring approximately 25 cm × 12 cm at 5 months and 45 cm × 25 cm at 2 years of age. The wound-array coverage is ~ 1.3 to 2.0% of BSA if wounds measuring 4 cm × 4 cm were created, with 4 cm of distance between adjacent wounds, in which 3 × 2 = 6 and 6 × 3 = 18 wounds could be established for 5-month-old and 2-year-old pigs, respectively. There is no specific restriction for allowable wound area in the Animal Use Protocol only when the study complies with the animal ethical guidelines that animals undergo procedures that cause no pain or distress, or only momentary or slight pain or distress. There are criteria for severity of wound area occupying the whole BSA in humans, that is, < 2% for minor burn, 2–10% for moderate burn, and > 10% for major burn with 3rd degree burn wounds. Accordingly, less than 2% of the whole body surface, which complies with the Animal Protection Act, was used to create an optimized wound-array model.
A limited number of wounds can be created in a mouse, requiring a reference wound for each mouse. For a comprehensive investigation, numerous wounds are necessary when using small animals; the number of animals utilized further increases due to wound numbers and individual differences. Conversely, only a singular large animal might be essential to complete a study, especially for screening or comparing various treatments. As mentioned above, 5–17 treatments can be initially screened and compared with a control open wound, or 1–5 treatments can be screened and compared with an open control in triplicate. Accordingly, one or two Lanyu pigs are sufficient to complete a study when compared with 6–30 mice that may be needed (2 wounds/mouse including a control and treatment; 1–5 treatments; n ≥ 6 for individual variants). In contrast to smaller and limited wound sizes in small animals, larger wounds are possible in larger animals. Moreover, the pig skin system is tight, resembling that of humans, unlike the looser skin system in mice, rats, and rabbits12.
Accordingly, the wound array on the dorsum of Lanyu pigs can be used to compare and evaluate various treatments, reducing the number of animals required while minimizing inter-individual variations and further decreasing sample sizes. Therefore, in accordance with the welfare of animals and the 3Rs of replacement, reduction, and refinement36 for preclinical trials, this wound array model is crucial for basic and clinical research.
A well-established anesthetic procedure should be employed during surgery, along with proper postoperative care, such as administering antibiotics or analgesia. One side of the dorsal skin is considered at a time to create wounds and is also suggested to allow the pig to lay down on the other dorsal side for rest or sleep. To consider the performance of wound healing, although the thoracolumbar region between the boundaries of both armpits accounts for dorsal skin, some indentations at both ends of approximately 5 cm were considered to prevent skin movement near the front and back legs (Fig. 3A). This also assists in dressing with sterile gauze, elastic bandage, and mesh bandage to prevent them from being stripped off. Large amounts of wound fluid are usually secreted during the initial 7 days, and highly absorbent materials should be applied and changed every other day to cover wounds.
A pig with an array of large wounds could allow more efficient testing and reduce animal usage in the study. However, there are also issues associated with this technique. First, creating a large array of wounds could induce unwanted systemic changes or host responses that would impact wound healing. Second, it would be impossible to characterize the systemic host response to different treatments. Accordingly, mouse models may still be used for initial screening and testing in addition to the availability, cost, and convenience.
Unless a given treatment is particularly relevant to the immune system, which affects the wound healing process, most current therapies and treatments promote skin regeneration; hence, this wound array model is a powerful platform for high-throughput screening of dressings or therapies. In Fig. 4, our results indicate that a distance of at least a 4-cm distance should be maintained between adjacent wounds to prevent wound interaction. This minimum 4-cm distance can be established as the gold standard between adjacent wounds to effectively differentiate various interventions and is first reported in the present study.
Considering the cost and time required for experiments, an optimal wound size of at least 4 cm × 4 cm was suggested to fit multiple wounds in a limited skin area in Lanyu pigs (Fig. 5). Larger wounds may be required to distinguish the efficacy of investigational therapies or dressings. In contrast, smaller-sized wounds may be used, for example, to assess the gene expression of healing profiles, necessitating small amounts of samples. In a study by Elgharably et al.21, 32 punch wounds were applied on one side of the pig dorsum with a diameter of 3 mm and a depth of 7 mm. Indeed, each animal study has unique aims and should be uniquely designed to test the theories or clarify specific questions. In other words, a wound size that is optimized for one study may not be ideal for another, depending on the study objectives.
Although complete wound closure was observed after the initial 2 months, the tensile strength of healed skin in the 2 cm × 2 cm group was only approximately 50% of that observed in the normal skin at 6 months post-surgery. Restoration of tensile strength to that of normal skin may require a year or longer, possibly after tissue remodeling. Histologically, the healed matrix at month 6 marginally resembled that of normal skin, as shown in Fig. 7; however, the extent of extracellular matrix remodeling was incomplete.
In order to obtain better blood flow signals and wound closure profiles, different observation time points in Fig. 6A,B were purposely designed. Since there is depth limitation for Laser Doppler imaging and it is elaborate to perform wound observation and images, day 0, 3, and followed with every other day were chosen because of no significant changes in the early time points. However, prominent blood flow changes happened for partial-thickness wounds in the early time, therefore day 0, 1, 2, and followed with every other day were chosen for wound observation and image acquiring. This is not only the reason why distinct pig was chosen for different depth of wound study following 3R principle, but also distinct observation time points could be exemplary for different depth of wound study. The quantitative data of profile comparisons in blood flow, wound closure, and contraction of different depth of wounds are presented in Fig. 6C,D,E, respectively.
Comparing the time points in Fig. 6C,D, we observed that the decreased blood flow signals correlated with wound closure. During the initial 4 days, the wound closure rate in partial-thickness wounds was considerably faster than that in full-thickness wounds (Fig. 6D), correlating with the high blood flow during the granulation phase (Fig. 6C). As the percentage of wound contraction was proportional to time-lapse (Fig. 6E), the second phase of wound closure demonstrated a flat period of the inflammatory phase up to day 7 (Fig. 6D) in full-thickness wounds. After day 7, fibroplasia, extracellular matrix deposition and re-epithelialization by keratinocytes occurred simultaneously. An extraordinarily steep curve of wound closure was observed in the full-thickness group versus a smooth curve in the partial-thickness group. This corresponds to fibroblast-initiated collagen matrix contraction (Fig. 6E). The last phase of wound closure correlated with wound epithelialization, and the slopes were similar in both groups (Fig. 6D).
Ehrlich et al.37 have reported that re-epithelialization of partial-thickness wounds could be attributed to basal keratinocyte recruitment from wound edges and dermal glands or hair follicles during the early phase of healing, which would stimulate the migration of epithelial cells to form a new epithelial layer. Conversely, cells for re-epithelialization of full-thickness wounds can only be derived from the wound edges, as the dermis has been completely abraded. Therefore, re-epithelialization of partial-thickness wounds is achieved considerably faster than full-thickness wounds. The healing of full-thickness wounds requires dermal regeneration of fibroblasts derived from adipose tissues or wound edges. Berry et al.38 have reported that fibroblasts are converted to myofibroblasts, and collagen or neosynthesized collagen undergoes contraction; hence, the magnitude of wound contraction in full-thickness wounds was higher than that in partial-thickness wounds (Fig. 6E).
Although there is a common process of wound healing, rate and extent of tissue regeneration in wounds may vary depending on the pig strain, age, wound size, and sampling location within a wound20. The thickness of the epidermis and dermis layer, as well as collagen and elastin content, may also differ during any specific stage of wound healing. Nevertheless, qualitative information in Fig. 7 stands for an instruction of the wound healing process. During the transition of epidermal regeneration, the epidermal layer became thicker after wound closure and its basal layer changed from flat to obvious protrusion, finally modified to normal rete ridges. During the process of skin maturity, it was found that the newly formed epidermis and dermal papillae were relatively thicker and flatter than the normal skin and rete ridges developed subsequently. The thickness ratios of rete ridges to dermal papillae turned from 2.0 for normal skin to 1.6 at month 2 and 2.9 at month 6 post-operatively. While the contents of collagen and elastin increased along the process of wound healing, their fibrils developed to the extent in the normal skin. It was found that collagen content was close to normal skin at month 6 postoperatively. The data in Fig. 7 were obtained by sampling from central locations in wounds, indicating that the remodeling phase of wounds was still undergoing at month 6 postoperatively, although wounds closed at approximately 2.5 weeks.
In aggregate, we determined a set of optimal parameters, including the minimal wound size of 4 cm × 4 cm and separation distance of 4 cm, for wound arrays on the dorsal skin, at the 5 cm indentation of the thoracolumbar region between the boundaries of both armpits and the paravertebral region above the rib tips of mature Lanyu minipigs. This model follows the 3R principles and provides a high-throughput platform for precisely screening the effectiveness of different interventional treatments, examining histological changes, and clarifying the underlying molecular mechanisms involved in the wound healing process. This study establishes a golden standard for wound creation and wound care in miniature pigs. In addition, the speed and quality of wound healing may be influenced by the strain of pigs selected, as well as the regeneration capacity of the animals used. In the future, the wound healing efficacy of various biomaterials needs further dedicated investigation.
