The effect of various concentrations of human recombinant epidermal growth factor on split‐thickness skin wounds

Abstract

Epidermal growth factor (EGF) is a potent stimulant of epithelialisation. However, topical application of EGF to achieve facilitated re‐epithelialisation in partial thickness wounds has been controversial. A total of 10 pigs, each with eight 4 × 4 cm partial thickness wounds, were treated twice a day for 10 days to observe the effect of human recombinant EGF in concentrations of 0·1, 1, 5, 10, 25 ug/g, vehicle only and two controls. The control and the vehicle‐only wounds each demonstrated 100% healing time (HT₁₀₀) of 9·31 ± 1·34 and 8·5 ± 1·12 while the wounds treated with EGF ointment with concentrations of 0·1 (HT₁₀₀ = 6·4 ± 0·71), 1 (HT₁₀₀ = 5·2 ± 0·63), 5 (HT₁₀₀ = 5·8 ± 0·85), 10 (HT₁₀₀ = 7·1 ± 1·45) and 25 ug/g (HT₁₀₀ = 7·4 ± 0·57) demonstrated significant reduction in time to achieve re‐epithelialisation. Among the EGF‐treated wounds, the wounds treated with EGF concentrations of 1 and 5 ug/g achieved the fastest re‐epithelialisation with evidence of substantial increase in basal keratinocyte activity observed through Ki‐67 activity. In conclusion, this article demonstrates the efficacy of human recombinant EGF in facilitating re‐epithelialisation of partial thickness wounds with the most efficient healing found in EGF concentrations of 1 and 5 ug/g.

Introduction

It has long been recognised that growth factors contribute to various pattern of cell activation during wound healing. Physiologically, wound healing is divided into three phases: inflammation, proliferation and remodelling (1, 2). Wound repair results from these complex cellular and molecular events. Growth factors control many of the key cellular activities involved in the healing process including cell division, cell migration, neovascularisation and synthesis of extracellular matrix molecules. Some of the principle growth factors involved in this process are epidermal growth factor (EGF), platelet‐derived growth factor (PDGF), fibroblast growth factor, keratinocyte growth factor, transforming growth factor‐β, vascular endothelial growth factor and granulocyte colony‐stimulating factor (3, 4). Currently, the use of recombinant human PDGF‐BB (rh‐PDGF‐BB, bearplermin) and EGF has been reported to be successful in increasing the incidence of complete healing and decreases the time to completely heal inpatients with lower extremity diabetic ulcer 5, 6, 7, 8, 9.

EGF was discovered in mouse salivary glands in 1962 and was the first growth factor to be described (10). It interacts with the EGF receptor on epidermal cells and fibroblasts (11). It is produced by platelets, macrophages and monocytes, and its primary role is to stimulate epithelial cells to grow across the wound but also acts on fibroblasts and smooth muscle cells 11, 12, 13. EGF has been reported to significantly accelerate epidermal regeneration of partial and full thickness skin wounds in pigs and continuous or prolonged exposure of EGF to increase tensile strength in rat skin wounds (13, 14). But despite its reports, the ideal concentration to enhance healing in acute wound remains debatable due to the subjective conclusion based on observation of epithelialisation.

This study was designed to objectively evaluate the ideal concentration of EGF in a pig split‐thickness wound model through morphometric analysis till 100% heal time (HT₁₀₀), histological analysis and anti‐Ki‐67 antibody immunohistochemical staining.

Materials and Methods

Experiment model

All animal procedures were approved by the institutional review board of Ulsan University School of Medicine, Asan Medical Center. Total of 10 domestic pigs (Yorkshire species) weighing 18–22 kg between ages of 50 and 60 days were used. The pigs were kept in isolated cages with room temperature of 28·5°C. Under anaesthesia with ketamine (20 mg/kg, Yuhan Pharmaceutical, Seoul, Korea) and xylazine (2 mg/kg, Rompun^®, Bayer Korea, Seoul, Korea), each pig was wounded with eight 4 × 4 cm sized split‐thickness (20/1000 inches) wounds using a Zimmer dermatome (Zimmer, IN, USA). For each pig, the wounds were divided into seven groups. The first five wounds were treated with human recombinant EGF ointment (Daewoong^® Pharmaceutical Company, Seoul, Korea) with concentrations of 0·1, 1, 5, 10 and 25 ug/g each. The sixth wound was treated with a vehicle‐only ointment, and the last two were not treated and remained as controls (Figure 1). The treatments were changed in sequence to minimise any anatomical bias of the wound. On each wound, with exception of the control wounds, ointment of 0·4 g was applied evenly over the wounds twice a day with a 12‐h interval. To avoid loss of ointment over the wound, hydrocolloid dressing (Duoderm^®, Convatec, New York, USA) was applied around the margin of the wound, then a film (Opsite^®, Smith and Nephew, London, UK) dressing was applied over the wound and the hydrocolloid dressing (Figure 2). The dressings were changed twice a day also. The point of re‐epithelialisation was determined when there were no more discharge and when a new epithelium was visible through two times loupe magnification.

Figure 1.

Porcine split‐thickness model. A total of ten pigs were used in each experiment. Recombinant human epidermal growth factor doses were alternated differently by position from one pig to the next to eliminate any possible differences in healing based on postion. 1, rh‐EGF (0·1 ug/g); 2, rh‐EGF (1 ug/g); 3, rh‐EGF (5 ug/g); 4, rh‐EGF (10 ug/g); 5, rh‐EGF (25 ug/g); 6, vehicle; 7, non treated; 8, non treated.

Figure 2.

To avoid loss of ointment over the wound, hydrocolloid dressing (Duoderm^®, Convatec) was applied around the margin of the wound, then a film (Tegaderm^®, 3M) dressing was applied over the wound and the hydrocolloid dressing.

Analysis

100% healing time

All wounds were evaluated by digital photography until 100% epithelialisation of the wounds was achieved. An image analyser (image measurement standard v4·01, Bersoft, Puerto Plata, Dominican Republic) and single‐blinded opinion of two researchers were used to determine the point of full epithelialisation.

Hematoxylin and eosin and anti‐Ki‐67 antibody immunohistochemical staining

On the 11th day, when full epithelialisation of all the wounds were noted, the pigs were euthanised, and the wounds were harvested including the periwound margin and base. Specimens were fixed in 10% neutral‐buffered formalin and embedded in paraffin. The 3 µm thickness sections were stained with hematoxylin and eosin (H&E) stain to evaluate the quality of the regenerated epidermis. Immunohistochemical staining using mouse anti‐human Ki‐67 antibody (Dako, Zug, Switzerland, 1:250) was used to evaluate the proliferation of basal keratinocytes (12, 15, 16). The sections were evaluated by the intensity of uptake under ×200 magnification using an image analyser (image measurement standard v4·01, Bersoft).

All values were expressed as means ± standard error, and statistical analysis was performed by analysis of variance (anova) and student t‐test analysis.

Results

Ht₁₀₀

The time till complete epithelialisation was measured. The opinions of two researchers were used to determine the point of full epithelialisation where the wound demonstrated no discharge and when a new epithelium was visible through two times loupe magnification. The control‐ and the vehicle‐only wounds each demonstrated healing time of 9·31 ± 1·34 and 8·5 ± 1·12. There was no statistical significance between these groups. However, the wounds treated with EGF ointment with concentrations of 0·1 (HT₁₀₀ = 6·4 ± 0·71), 1 (HT₁₀₀ = 5·2 ± 0·63), 5 (HT₁₀₀ = 5·8 ± 0·85), 10 (HT₁₀₀ = 7·1 ± 1·45) and 25 ug/g (HT₁₀₀ = 7·4 ± 0·57) did show significantly reduced time till complete healing compared with the control‐ and vehicle‐only wounds (Table 1). The EGF ointment with concentrations of 0·1, 1 and 5 ug/g each had reduction of 2·1, 3·2, and 2·7 days each compared with the vehicle‐only wound (P < 0·01). The wounds treated with EGF ointment concentrations of 10 and 25 ug/g had reduction of 1·4 and 1·1 days each compared with the vehicle‐only group (P < 0·05) (3, 4).

Table 1.

The 100% healing time (HT₁₀₀) of split thickness wounds treated with various concentration of h‐EGF

Treatment (N = 10)	HT100 (day) (mean ± SD)	ΔHT100 (day) (treatment — vehicle)
Non treated	9·3 ± 1·34	−0·5
Ointment vehicle	8·5 ± 1·12	0·0
EGF ointment (0·1 ug/g)	6·4 ± 0·71*, †	2·1
EGF ointment (1 ug/g)	5·2 ± 0·63*, †	3·2
EGF ointment (5 ug/g)	5·8 ± 0·85*, †	2·7
EGF ointment (10 ug/g)	7·1 ± 1·45†, ‡	1·4
EGF ointment (25 ug/g)	7·4 ± 0·57†, ‡	1·1

^* P < 0·01.

^† P < 0·01, significantly different from non treated group.

^‡ P < 0·05, significantly different from vehicle‐treated group.

Figure 3.

The effects of recombinant human epidermal growth factor (rh‐EGF) on the healing of porcine split‐thickness wounds at 5 (A) and 6 (B) day post‐surgery. At 5 days post‐surgery, re‐epithelialisation was considerably increased in rh‐EGF‐treated wounds compared with vehicle‐treated wound and non treated wounds (A). At 6 days post‐surgery, full epithelialisation was markedly observed in rh‐EGF‐treated wounds compared with vehicle‐treated wound and non treated wounds (B). 1, rh‐EGF (0·1 ug/g); 2, rh‐EGF (1 ug/g); 3, rh‐EGF (5 ug/g); 4, rh‐EGF (10 ug/g); 5, rh‐EGF (25 ug/g), 6, vehicle; 7, non treated; 8, non treated.

Figure 4.

Comparison of 100% healing time (HT₁₀₀) of various rh‐EGF concentrations in porcine split‐thickness wounds. Treatment of rh‐EGF 0·1, 1, 5, 10, 25 ug/g concentrations significantly decreased wound size and shortened the HT₁₀₀ more than vehicle‐treated wounds and non treated wounds. About 1 and 5 ug/g rh‐EGF considerably accelerated the rate of healing of split‐thickness wounds. No difference was observed in the HT₁₀₀ between vehicle and non treated wounds. Each bar is mean ± standard deviation of the mean (N = 10). *P < 0·05, significantly different from vehicle‐treated group; **P < 0·01.

H&E and Anti‐Ki‐67 antibody immunohistochemical staining

After epithelialisation when specimens were observed at the 11th day, histological H&E staining did not reveal any significant microscopical differences. Once re‐epithelialisation was achieved, all wounds were covered with stratum corneum and maturing epidermal cells. However, when examined using the cell proliferation marker (anti‐Ki‐67 antibody immunohistochemical staining), it revealed increased uptake of basal keratinocytes in the specimens treated with EGF ointment compared with the control‐ and vehicle‐only wound groups. The expression of the Ki‐67 activity was measured in arbitrary units using the image analyser (image measurement standard v4·01, Bersoft). The control‐ and vehicle‐only wound groups showed values of 388 ± 69 and 417 ± 73, respectively, without any significant differences between the two groups. The EGF‐treated wounds with concentrations of 0·1, 1, 5, 10 and 25 ug/g all showed significantly increased activity compared with the control‐ and vehicle‐only wound group (Table 2, Figure 5). The h‐EGF‐treated wound with concentrations of 1 and 5 ug/g had 7·2 and 6·3 times activity compared with the vehicle‐only groups with statistical significance (P < 0·01). The wounds treated with 0·1, 10 and 25 ug/g EGF ointment had increased activity of 1·6, 2·4 and 2·2 times the vehicle‐only wound group (P < 0·05) (Figure 6).

Table 2.

Ki‐67‐positive cells of split‐thickness wounds treated with rh‐EGF

Treatment (N = 10)	Arbitrary unit (mean ± SD)	Relative ratio (treatment/vehicle)
Non treated	388 ± 69	0·9
Ointment vehicle	417 ± 73	1·0
EGF ointment (0·1 ug/g)	660 ± 54*, †	1·6
EGF ointment (1 ug/g)	2986 ± 187*, †	7·2
EGF ointment (5 ug/g)	2676 ± 314*, †	6·3
EGF ointment (10 ug/g)	1016 ± 197*, †	2·4
EGF ointment (25 ug/g)	900 ± 132*, †	2·2

^* P < 0·01, significantly different from vehicle‐treated group.

^† P < 0·01, significantly different from non treated group.

Figure 5.

Expression of Ki‐67 in the cytoplasm of keratinocytes in the epidermal basal cell layer (image measurement standard 4·1, ×200). Ki‐67 reactivity was displayed within the basal layer of the epidermis. The greater number of Ki‐67‐positive cells (red) were observed in 1 ug/g and 5 ug/g rh‐EGF‐treated group (D, E) than other groups. (A) Non treated group, (B) vehicle‐treated group, (C) 0·1 ug/g rhEG‐treated group, (D) 1 ug/g rh‐EGF‐treated group, (E) 5 ug/g rh‐EGF‐treated group, (F) 10 ug/g rh‐EGF‐treated group, (G) 25 ug/g rh‐EGF‐treated group

Figure 6.

Comparison of number of Ki‐67‐positive cells in rh‐EGF‐treated group, vehicle‐treated group and non treated group. Wounds treated with rh‐EGF 0·1, 1, 5, 10, 25 ug/g concentrations had significantly the greater of Ki‐67‐positive cells compared with vehicle‐treated wound. Each bar is mean ± standard deviation of the mean (N = 10). *P < 0·01, significantly different from vehicle‐treated group.

Discussion

Topical application of growth factors for wound healing has been well described 5, 6, 7, 8, 9. But these reports have been focusing on the application to chronic wounds where physiology and the process of healing may differ from that of acute wounds. The acute wound can be divided into full thickness and partial thickness wounds. The healing process differs, in that full thickness wound requires granulation formation, contraction and formation of extracellular matrix and remodelling while the partial thickness wound requires a re‐epithelialisation process which is the most important factor to achieve healing. Thus the reduction in time of re‐epithelialisation can be the key to facilitate healing in partial thickness wounds (13, 17). EGF interacts with epidermal cells and fibroblasts, and its primary role is to stimulate epithelial cells to grow across the wound but also to act on fibroblasts and smooth muscle cells 11, 12, 13.

Designing a wound‐healing model for ointment application can be a very difficult task. The ointment can frequently evaporate or be easily washed away. Thus, a protective barrier should be designed to maintain the ointment on the wounds as well as protect the wound from external environment. Using a hydrocolloid dressing around the wound, the ointment is refrained from being washed away from the wound, and the use of film over the hydrocolloid and the wound surface treated with ointment can be maintained as an isolated area from other wounds in a standardised manner. The clear film does not adhere to the wound because of the ointment, and the transparent film allows the observer to observe the wounds easily. This model successfully isolated each wound in a standardised manner and provided a reliable atmosphere for each wound.

EGF stimulates epithelialisation in early human wound repair and has been demonstrated in reports of partial thickness wound models in pigs (13, 17). There has been a number of attempts to show enhanced epidermal regeneration using EGF but have been unsuccessful (18, 19). But with advanced technology in cloning and large scale manufacturing growth peptides leading to large scale studies, a general consensus of facilitating effect of human recombinant EGF in early epithelialisation process has been carefully recognised despite the early unclear and varying reports. However, there still lies some confusion on which range of concentration exerts the most efficient response in acute wound healing. Brown et al. (13) reported the range of positive therapeutic response for h‐EGF in doses of 0·1–10 ug/ml in partial thickness burn model in pig. However, Breuing et al. reported that 10 ug/ml of recombinant human EGF (rh‐EGF) in partial thickness excision wound model actually delays healing compared with the facilitated healing observed in treatment between rh‐EGF concentrations of 0·01 and 1 ug/ml (17). The varying results among studies may be due to the subjective observations of the wound. In this study, porcine wounds having similar physiology and anatomy were used to closely resemble human donor site injuries and an objective as well as subjective investigation of healing of partial thickness wounds related to h‐EGF were performed.

The range of concentration of EGF was decided on previous studies and effects of each concentration were compared with the control, vehicle only as well as other concentrations of EGF. In the wounds treated with h‐EGF, all wounds achieved statistically faster re‐epithelialisation compared with the control‐ and vehicle‐only wounds. In wounds treated with concentrations of 10 and 25 ug/g, it still achieved faster re‐epithelialisation compared with the control‐ and vehicle‐only wounds. This finding is supported by the increased uptake of anti‐Ki‐67 antibodies. It can be presumed that EGF up to 25 ug/g lead to enhanced epithelialisation of partial thickness wounds. However, when looking closer at the results between the wounds treated with different concentrations of EGF, there was a tendency to achieve faster healing in the wound group treated with 1 and 5 ug/g. In these groups, re‐epithelialisation was achieved between 5 and 6 days compared with 8·5 days in the vehicle‐only group and 7 days in the 10 and 25 ug/g wound groups. This finding coincided with the findings seen from cell proliferation marker where the 1 and 5 ug/g groups showed most reaction with anti‐Ki‐67 antibodies. The Ki‐67 marker, a cell proliferation marker, can be used to access the proliferation activity of basal keratinocytes. The increased activity found in the anti‐Ki‐67 antibody immunohistochemical staining in the h‐EGF‐treated specimens indicate active proliferation of keratinocytes leading to facilitated healing. This finding coincides with one of the two major mechanisms of EGF, which is its mitogenic action and increased rate of cell migration (20, 21). However, there is a fairly different behaviour in healing when observing the wound treated with 0·1 ug/g EGF. Despite the low activity of Ki‐67 at the basal layer, the rate of re‐epithelialisation remains to be quite fast. When comparing the wounds of 0·1 ug/g treated wounds with 10 or 25 ug/g treated wounds, the 0·1 ug/g treated wounds seem to heal at a slightly faster rate despite of the low Ki‐67 positive cells. There might be a different mode of action to enhance healing at this low concentration. Further study should be performed to determine the mechanism behind wound healing at low concentrations of h‐EGF.

The h‐EGF concentration between 1 and 5 ug/g can be seen as the ideal concentration to achieve the most efficient results for acute wounds with partial thickness defects. But when findings from H&E staining were observed at the 11th day, there were no significant differences in histological appearance at this point of healing in all wounds. It can be presumed that h‐EGF exerts a clear histological effect on the epithelialisation at early period of healing but more time will be required to evaluate the effect of h‐EGF beneath the epithelium.

This article demonstrates the efficacy of human recombinant EGF in facilitating re‐epithelialisation of partial thickness wounds. Although all wounds treated with h‐EGF between the concentrations of 0·1 and 25 ug/g facilitated re‐epithelialisation compared with the control‐ and vehicle‐only treated wounds, the most efficient healing can be found with concentrations of 1 and 5 ug/g in ointment form which corresponded with the highest proliferative activity of basal keratinocytes.

Acknowledgements

This research has been performed with grant supported by Daewoong^® Pharmaceutical Company.