Local Insulin-Zinc Injection Accelerates Skin Donor Site Wound Healing

Journal of Surgical Research 142, 90 –96 (2007) doi:10.1016/j.jss.2006.10.034

Local Insulin-Zinc Injection Accelerates Skin Donor Site Wound Healing Xiao-jun Zhang, M.D.,*,‡ Xiaowu Wu, M.D.,‡,1 Steven E. Wolf, M.D.,‡,1 Hal K. Hawkins, M.D., Ph.D.,†,§ David L. Chinkes, Ph.D.,*,‡ and Robert R. Wolfe, Ph.D.*,‡,2 *Metabolism Unit and †Electron Microscopy Laboratory, Shriners Hospital for Children; ‡Department of Surgery, and §Department of Pathology, University of Texas Medical Branch, Galveston, Texas Submitted for publication July 13, 2006

out major systemic side effects, demonstrating its potential usefulness in burn treatment. © 2007 Elsevier Inc. All

Background. Systemically administered insulin has been shown to accelerate wound healing. To minimize the hypoglycemic and hypokalemic effects of insulin, we investigated a new route of insulin administration by local injection into skin wounds. Materials and methods. Partial thickness skin donor site wounds were created on the backs of adult rabbits with a dermatome set at 0.015 inch. The wounds were covered by Aquaphor gauze (Smith and Nephew, Largo, FL), and OpSite membrane (Smith and Nephew, Hull, United Kingdom) and protected by rabbit jackets. Long-acting insulin-zinc suspension was selected for local injection. In study 1, insulin was injected into the wound at different doses, and the concentrations of blood glucose and wound insulin were measured to determine the proper dose and injection frequency. In study 2, wound healing days were compared between two groups (n ⴝ 7 each) receiving local injection of either insulin-zinc or zinc alone as control. Based on the results from study 1, a dose of 0.25 units of longacting insulin-zinc suspension was injected into the wound every other day in the insulin group. Results. After injection, 0.25 units of insulin decreased blood glucose concentration (minimum 60 mg/ dL) during the first 3 h, which then returned to the preinjection level (80 mg/dL). One injection maintained wound insulin concentration above 50 ␮U/mL for more than 24 h. With local injection of 0.25 units insulin-zinc every other day, the wound healing time was 11.2 ⴞ 2.3 d, which was faster (P ⴝ 0.02) than 15.1 ⴞ 4.1 d in the control group. Conclusion. Local injection of long-acting insulinzinc suspension accelerated skin wound healing with-

rights reserved.

Key Words: skin wound; rabbit; insulin; zinc. INTRODUCTION

More than 70 years have passed since insulin was recognized to be important in wound healing [1]. Whereas a large body of evidence has demonstrated beneficial effects of insulin on wound healing [1–7], no method of administering insulin has been reported that is suitable for routine clinical use in wound healing. Systemic insulin administration has limited clinical applicability because of significant side effects, including hypoglycemia, hypokalemia, and hypoaminoacidemia [4, 8, 9]. Topical insulin application, although optimal in theory, has been hindered by the lack of a carrier vehicle that can reliably deliver insulin into the wound at a controllable rate. The doses of topical insulin that have been studied vary from 5 to 15 units in mice [6], 20 units in horses [5], to drops of insulin solution in rats and humans [7]. Recently, it has been demonstrated that insulin receptors are present in cell bodies of almost all tissues including keratinocytes and fibroblasts [10], and that insulin is able to activate the phosphatidylinositol 3-kinase and mitogen-activated protein kinase signal transduction pathways in the skin keratinocytes and fibroblasts [10 –15]. These findings support the notion that local insulin application could be an effective treatment if insulin concentration in the wound can be maintained at a sufficiently high level. The present study was designed to test the hypothesis that local insulin injection into the skin wound maintains sufficiently high insulin concentration in the wound, thereby accelerating wound healing with minor systemic effects. The local injection method al-

1 Current address: U.S. Army Institute of Surgical Research, 3400 Rawley E. Chambers Ave, Fort Sam Houston, TX 78234. 2 To whom correspondence and reprint requests should be addressed at UAMS, 4301 Markham Slot 806, Little Rock, AR 72205. E-mail: [email protected]

0022-4804/07 $32.00 © 2007 Elsevier Inc. All rights reserved.

90

ZHANG ET AL.: INSULIN-ZINC ACCELERATES WOUND HEALING

lows precise control of insulin dose and localizes the effects of insulin in the wound. Thus, we anticipated that this method might combine the advantages of systemic and topical application of insulin and avoid their disadvantages. MATERIALS AND METHODS Animals Male New Zealand white rabbits (Myrtle’s Rabbitry, Tompson Station, TN), weighing approximately 4.5 kg, were used for this study. The rabbits were housed in individual cages and were given 150 g of Lab Rabbit Chow HF 5326 (Purina Mills, St. Louis, MO) per day for weight maintenance. This protocol complied with NIH guidelines and was approved by the Animal Care and Use Committee of The University of Texas Medical Branch.

Surgical Procedures After an overnight fast with free access to water, the hair on the back was removed by clippers and application of a depilatory lotion. The animals were anesthetized with ketamine and xylazine as we previously described [16, 17]. Using aseptic technique, two partial thickness skin graft donor wounds were created using an electric dermatome (Padgett Instruments Inc., Kansas City, MO) set at 0.015 in., on the left and right sides of the back. The wounds were covered with Aquaphor gauze (Smith and Nephew, Largo, FL) and OpSite membrane (Smith and Nephew, Hull, United Kingdom), and protected with a rabbit jacket (Harvard Apparatus, Boston, MA). In study 1, additional surgical procedures were performed to place indwelling vascular catheters for monitoring blood glucose concentration in conscious rabbits. Using aseptic technique under anesthesia, an incision was made on the left neck and medical vinyl tubes (size 3A; Scientific Commodities, Inc., Lake Havasu City, AZ) with Luer stub adapters (gauge 20; Becton Dickinson, Sparks, MD) were placed in the left carotid artery and jugular vein. The tubes were filled with heparin solution (1000 unit/mL) and exited from the back between the ears through a subcutaneous tunnel. Jelco intermittent injection caps (Johnson and Johnson, Medical Division of Ethicon, Inc., Arlington, TX) were connected to the adapters of the catheters for repeated injection or sampling. The caps were secured onto the scalp by suturing. The incision was closed by layered suturing. A single dose of bicillin (50,000 U/kg; Wyeth Laboratories, Philadelphia, PA) was injected into thigh muscle to prevent infection. When the animals had awakened from anesthesia, buprenorphine (0.015 mg/kg) was injected intramuscularly twice a day for 2 d as an analgesic.

Local Insulin Injection Extended insulin zinc suspension (Humulin; Eli Lilly and Company, Indianapolis, IN) was added to albumin-saline solution (75 mg albumin per mL 0.9% sodium chloride) at a final insulin concentration of 0.25 or 1.0 unit/mL. The injection volume was 1 mL, which was evenly injected into the wound at 5 points at sites 1 to 1.5 cm from the corners of the wound and in the center of the wound. Only one of the two wounds (on left and right sides of the back) was injected. To ensure intradermal injection, we used a 1 mL syringe with a 27gauge, 1/2 in. needle (Becton Dickinson, Franklin Lakes, NJ), and clamped the tip of the needle with a mosquito hemostat placed so that only 3 mm of the tip could be inserted into the wound. In the control group, the injection solution contained the same dose of zinc (as zinc sulfate) and albumin, but insulin was not included.

91

Experimental Design Study 1 (n ⫽ 8) was performed on the day 2 post-injury to determine a proper insulin dose (n ⫽ 4) and injection frequency (n ⫽ 4). The proper insulin dose was defined as one that could cause a mild and temporary decrease in blood glucose concentration that was well tolerated by the rabbit without exogenous glucose replacement. This definition assured the effectiveness and safety of the local insulin injection. To determine the proper insulin dose, 0.25 or 1.0 unit of the long-acting insulin was injected into the donor wound and glucose concentration in the arterial blood was monitored for 6 to 12 h. These two doses were selected from our pilot studies in which different doses were tested. During the 6 to 12 h observation period, physiological saline was infused into the jugular vein tube at 20 to 30 mL/h to prevent dehydration; and no nutrition was given, so that the measured glucose concentration reflected the lowest possible level. When the proper dose was determined, this dose of insulin (0.25 units) was locally injected into the wound; at 4, 8, 12, and 24 h after the local injection, one rabbit was sacrificed by intravenous injection of 5 mL of saturated KCl under general anesthesia, and wound tissue was taken for measurement of insulin concentration. The injection frequency was determined from the changes of wound insulin concentration. Study 2 included an insulin group and a control group (n ⫽ 7 each). Skin donor site wounds were created on both sides of the back using the dermatome set at 0.015 inch. The weight (mg) and area (cm 2) of the removed split skin were measured by a technician who did not know the treatments. A ratio of weight to surface area (mg/cm 2) of between 30 to 50 mg/cm 2 in the removed skin was considered to be proper, as it took approximately 2 wk for this depth of wound to heal. The wound on the left side of the back was used for study in each rabbit. Originally, we planned to use the wound on the left for local insulin injection and the wound on the right as control. This plan was given up when we found that insulin concentration was increased not only in the injected wound but also in the uninjected wound. However, creation of two wounds in each rabbits was continued. Based on the results from study 1, 0.25 units of the long-acting insulin preparation were injected into one of the wounds in the insulin group every other day, starting immediately after injury throughout the entire healing period. In the control group, the vehicle solution, which contained zinc and albumin but no insulin, was used for local wound injection at the same injection volume and frequency as in the insulin group. Daily consumption of rabbit chow was recorded by providing 150 g rabbit chow at 8:00 every day and checking the amount of chow left the following morning at 8:00. The wound dressings were changed every other day for the first 6 d and once daily thereafter. Photos were taken at the time of wound decreasing changes and assessed by surgeons for wound healing. As soon as the wound was judged to be healed, skin biopsies were taken under general anesthesia (ketamine and xylazine) for histological observation.

Analysis Blood glucose concentration was measured on a glucose/lactate analyzer (model 2300; Yellow Spring Instruments, Yellow Springs, OH). To measure insulin concentration in the wound fluid, 0.3 g of wound tissue was cut into small pieces on dry ice and homogenized at 4°C in 1 mL of saline containing 2.3 mg ethylenediamine tetraacetic acid and 8 mg albumin. After centrifugation, the supernatant was used for insulin assay. The measured insulin concentration was multiplied by [(1 ⫹ 0.225)/0.225] to calculate the original concentration in the wound, where 1 is the volume of buffer added to 0.3 g of wound tissue and 0.225 (mL) is the water volume in 0.3 g of wound tissue. The water volume in the wound was estimated from the weight difference before and after drying in an oven (80°C) for 24 h. This sample processing is based on the notion that insulin can be degraded in the wound fluid so that the samples have to be kept cold

92

JOURNAL OF SURGICAL RESEARCH: VOL. 142, NO. 1, SEPTEMBER 2007

FIG. 1. (A) Effect of local insulin injection on arterial blood glucose concentration. Local injection of 0.25 units of long-acting insulin caused moderate decreases in blood glucose concentration during the first 3 h and returned to the basal value thereafter. In contrast, local injection of 1.0 unit of insulin caused severe hypoglycemia. (B) Injection of 0.25 units of long acting insulin in another two rabbits caused reproducible changes in arterial glucose concentration. Thus, the dose of 0.25 units was selected for local injection.

[18]. Insulin concentration in the plasma and in the wound (mixture of intracellular fluid and interstitial fluid) was measured using the microparticle enzyme immunoassay technique [19]. The skin biopsy samples were fixed in 10% phosphate-buffered formalin, embedded in paraffin, and sections were stained with hematoxylin and eosin (H&E). Masked H&E-stained slides were examined using a microscope with a 10⫻ objective and an ocular scale that was calibrated using a stage micrometer. The average thickness of the epidermis was estimated to two significant figures [20].

centration in the arterial blood was monitored for 12 h. Local injection of 0.25 units of the insulin caused a moderate decrease in blood glucose concentration (blood glucose ⱖ 60 mg/dL) during the first 3 h, which thereafter returned to the preinjection level (Fig. 1A). In contrast, local injection of 1 unit of the insulin preparation resulted in severe hypoglycemia, although the rabbit survived the injection (Fig. 1A). To confirm that 0.25 units is a proper dose for local insulin injection, we repeated the local injection of 0.25 units of the longacting insulin into the wound in another two rabbits and monitored the arterial blood glucose level for 6 h. The changes of blood glucose level (Fig. 1B) were comparable with that of 0.25 units of insulin injection in Fig. 1A. Therefore, we selected 0.25 units as the injection dose. In these two rabbits, arterial blood was collected at 0.5, 1, 2, 4, 6, 8, 10, and 12 h after the insulin injection for measurement of plasma insulin concentration. The concentrations of insulin in the plasma were all below the detectable level even after local injection of 1.0 unit of insulin. Four rabbits were used for measurement of insulin concentrations in the wound at 4, 8, 12, and 24 h after local injection of 0.25 units of long-acting insulin. Insulin concentrations in the wounds were increased to ⱖ50 ␮U/mL for at least 24 h (Fig. 2), so we selected the injection frequency of every other day. Study 2

The body weights of the rabbits in the control and insulin groups were 4.44 ⫾ 0.18 kg and 4.33 ⫾ 0.08 kg (n ⫽ 7 each; P ⫽ 0.13), respectively. The skin donor wounds created with the electric dermatome at 0.015 in. consistently spared a part of the dermis (Fig. 3A). Histological slides showed that the epidermis and papillary dermis were absent, and that the tips of hair follicles remained (Fig. 3B), indicating that these were partial thickness wounds. Fourteen rabbits received skin donor wounds on both sides of the back. The wound on the left side of the body was used for study

Statistics Data are expressed as means ⫹ SD; medians. The differences between two groups were evaluated using Wilcoxon test. One tail tests were used because increased insulin concentration will either improve healing time or have no effect. P ⬍ 0.05 was considered as statistically significant. The relation of epidermal growth rate and days for wound healing was examined with Pearson’s correlation.

RESULTS Study 1

Eight rabbits were used to determine the proper dose and injection frequency. To determine the proper dose, 0.25 or 1.0 unit of the long-acting insulin was injected into the donor wound in two rabbits and glucose con-

FIG. 2. Insulin concentration in the skin wound fluid after local insulin of 0.25 units of long-acting insulin. The concentration of insulin in the wound remained ⱖ50 ␮U/mL for at least 24 h. Thus, the injection frequency was selected as every other day.

ZHANG ET AL.: INSULIN-ZINC ACCELERATES WOUND HEALING

93

who were blinded to the treatments; the mean number of days for complete healing was used as an indicator of the rate of wound healing. In the insulin group, wounds were healed after a mean of 11.2 ⫾ 2.3 d (median ⫽ 12 d), earlier than in the control group, which required a mean of 15.1 ⫾ 4.1 d for complete healing (median ⫽ 13 d) (P ⫽ 0.03, Wilcoxon test, 1-tail test). Representative wounds in the two groups are illustrated in Fig. 5A and B. The thickness of the epidermis was measured by a pathologist who was masked regarding the treatments. The measured epidermal thickness in millimeters was divided by the number of days to complete healing to provide an estimate of the growth rate of the epidermis (mm/d). In the insulin group we missed biopsies in the first two rabbits, so results from five rabbits were compared with those from seven rabbits in the control group. The growth rate of the epidermis in the insulin group (0.0166 ⫾ 0.0047 mm/d, median ⫽ 0.0167 mm/d) was greater (P ⫽ 0.012) than in the control group (0.0092 ⫾ 0.0025 mm/d, median ⫽ 0.0083 mm/d). When data from the two groups were pooled, the rate of epidermal growth was negatively correlated (r ⫽ ⫺0.66, P ⫽ 0.02) with the number of d for complete wound healing (Fig. 6). DISCUSSION

FIG. 3. (A) the appearance of the donor wound created by the electric dermatome at 0.015 in. There was dermal tissue on the wound surface, characteristic of a partial thickness wound. (B) light micrograph (H&E-stained) of the skin donor site wound sampled immediately after wounding shows deep fascia, panniculus carnosus muscle, and fibrous connective tissue. The epidermis is absent, and no follicles or other epidermal appendages are seen, except for the tips of a few follicles (arrows point at residuals of hair follicles). These findings are typical of a partial thickness wound. (Color version of figure is available online.)

and the wound on the right side was not used except in one rabbit in the control group the wound on the right was used for study because the wound on the left side was too deep (depth: 53 mg/cm 2). The size and depth of the experimental wound (reflected by the ratio of weight to area of removed split skin) used for study were 40 ⫾ 6 cm 2 (median ⫽ 38 cm 2) and 37.8 ⫾ 7.4 mg/cm 2 (median ⫽ 37.6 mg/cm 2) in the insulin group. These values were comparable to the corresponding values of 44 ⫾ 6 cm 2 (median ⫽ 43 cm 2; P ⫽ 0.17) and 38.0 ⫾ 3.7 mg/cm 2 (median ⫽ 37.7 mg/cm 2; P ⫽ 0.42) in the control group. Daily chow intake after injury was comparable (P ⬎ 0.05) between two groups except that on day 5 after injury the control group consumed more chow than the insulin group (Fig. 4). The wound photos were assessed independently by two burn surgeons

In the present experiment, we have successfully used local injection of long-acting insulin to accelerate skin wound healing. The local injection of insulin reduced mean healing time from 15.1 d in the control group to 11.2 d in the insulin group, representing a reduction of healing time by ⬃25%. Histological observations indicate that there was stimulation of epidermal growth (i.e., re-epithelialization) in the insulin-injected wound. The negative correlation between epidermal growth rate and days for wound healing indicates that epidermal growth is a determinant of wound healing. These results are consistent with reports that insulin stimulates keratinocyte proliferation [21]. The local injection

FIG. 4. Daily chow intake in rabbits. The amounts of chow consumption after injury were not significantly different (P ⬎ 0.05) between the two groups except on day 5 the insulin group consumed less (P ⬍ 0.05) chow than the control group.

94

JOURNAL OF SURGICAL RESEARCH: VOL. 142, NO. 1, SEPTEMBER 2007

FIG. 5. (A) A representative wound in the control group. On day 9 after injury, the wound was not healed. There was a large open wound. This wound was fully healed on day 20. (B) A representative wound in the local insulin group. The wound was fully healed on post-injury day 9. (Color version of figure is available online.)

method is believed to be safe, convenient, and effective because: (1) the dose of insulin was only 0.25 units, much smaller than the doses previously reported for topical insulin of 5 to 20 units [5, 6], and did not increase plasma insulin concentration; (2) insulin was injected every other day, rather than several times a day, as when it was used as topical application [7]; and (3) the treatment significantly accelerated wound healing. After local injection of 0.25 units of long-acting insulin, the concentration of insulin in the wound was maintained at ⱖ50 ␮U/mL for at least 24 h (see Fig. 2). In our previous experiment, we found that systemic insulin infusion at a low dose (0.6 mU ·kg ⫺1·min ⫺1), which raised plasma insulin to 35 ␮U/mL, was effective in improving net protein balance in the skin wound and muscle in anesthetized rabbits [22]. Therefore, the insulin concentration of ⱖ50 ␮U/mL in the skin wound should be sufficient to stimulate wound anabolism and healing in the rabbit. According to the manufacturer’s package insert for the extended insulin-zinc suspen-

sion, the insulin activity lasts at least 36 h after a subcutaneous injection, suggesting that the injected insulin is absorbed slowly into the circulation. Insulin is normally injected subcutaneously for systemic therapy, not intradermally as in this study. It is likely that after local injection into the remaining dermal tissue within the wounds the absorption rate is even slower than that after a subcutaneous injection in normal skin, though the precise mechanism is not clear. This may explain why plasma insulin concentration was not detected even after local injection of 1 unit of insulin in study 1. Because the insulin injected into the wound was human insulin and the microparticle enzyme immunoassay technique is sensitive to human insulin [19], there should have been no problem detecting the systemic appearance of the injected insulin. The insulin concentration in the wound was successfully measured after a 5.4-fold dilution (1 mL buffer added to 0.3 g wound tissue which had 0.225 mL water), supporting that the insulin assay was sensitive to the injected human insulin. Thus, the undetected insulin level in plasma was most likely due to the low plasma concentration after local insulin injection. Because of the increased insulin concentration in the wound but not in plasma, the decrease in arterial blood glucose level was probably due to an increased glucose uptake predominately by the wound tissue. Originally, we planned to inject insulin into the wound on one side of the back and use the wound on the opposite side as a paired control. However, we found that in the insulin group, the wound on the opposite side also had an increased concentration of insulin, due to the diffusion of insulin into the adjacent tissues. Thus, we used only the injected wound (either insulin or vehicle solution) for comparison of healing

FIG. 6. Correlation (r ⫽ ⫺0.66, P ⫽ 0.02) between epidermal growth rate and days for wound healing. The faster the epidermal growth was, the less the number of days required for wound healing.

ZHANG ET AL.: INSULIN-ZINC ACCELERATES WOUND HEALING

rate in both the control and insulin groups. The diffusion of insulin into the adjacent tissues indicates that local injection of 0.25 units of insulin affected more than 40 cm 2 of wound area. We estimate that the diffusion distance from the center of the injected wound to the distal border of the opposite uninjected wound was 10 cm and, thus, the total covered area was at least 300 cm 2, assuming that injection of 0.25 units of insulin led to uniform diffusion of insulin into the adjacent wound tissues in all directions. According to the Meeh’s equation and coefficient [23], a rabbit with body weight of 4.33 kg should have a total surface area of 2663 cm 2. Therefore, the local injection method could cover ⬃10% of total body surface area in the rabbit. If an adult burn patient can tolerate injection of 3 units of insulin, the local injection method could cover up to 3600 cm 2 wound, which is approximately 20% of total body surface area, if the body surface area is 2 m 2. Because zinc is important in wound healing [24 –26], the role of zinc in the extended insulin-zinc suspension had to be evaluated. The mean tissue concentration of zinc in the untreated wound was 8 to 13 ␮g per g of tissue [24]. The total zinc in the 0.25 units of insulin was 0.4 ␮g, which was injected in an area of 40 cm 2 and diffused into a larger area estimated to be ⬃300 cm 2. However, given the potential interactive effects of zinc and insulin [27], and the accelerated loss of zinc from wounds [28], it was considered worthwhile to assess the possible role of zinc in the healing. The results showed that local injection of the small amount of zinc alone, as was conducted in the control group, did not accelerate wound healing. Using the same wound model without any local injection, it took 14.8 ⫾ 3.1 d for the wound to heal [17], virtually identical to the 15.1 ⫾ 4.1 d required for wound healing in the zincinjection control group of the present study. Thus, the acceleration of wound healing observed in the insulin-zinc group can be attributed to insulin action rather than to the small amount of zinc in the insulin preparation. Whereas the present experiment demonstrated beneficial effects of local insulin injection on wound healing, further studies are warranted in preparation for possible application to human skin wounds. First, modifications of the insulin solution, such as addition of local anesthetics (e.g., lidocaine) to reduce pain could make the local injection more acceptable. Second, the distance of insulin diffusion needs to be clarified because it is important to determine the optimal injection sites and distances for a wound larger than 40 cm 2 or for multiple wounds. Third, it is important to know whether this method is effective in patients with severe burns, in whom insulin resistance in muscle and liver tissues is a significant problem. The effectiveness of injected insulin in second degree burn wounds and grafted third degree burn wounds also remains to be

95

investigated. Even if insulin injection were effective in speeding healing only in donor site wounds, this would provide a significant benefit and might reduce the required length of hospitalization for treatment of patients with large burns. In summary, local injection of insulin provides a potentially useful approach to accelerate skin wound healing in patients. The local injection method is safer and more convenient than systemic insulin administration because it has only minor and transient systemic effects. In comparison with topical application of insulin, local injection eliminates the uncertainty of the rate of delivery of insulin to the tissues in the wound bed, and requires less frequent application. This method could become a broadly applicable approach to accelerate skin wound healing in patients. Further investigations are needed to improve the formulation, clarify the minimal effective injection dose and optimal injection spacing, and to confirm the effectiveness of this treatment in the presence of severe burns. ACKNOWLEDGMENTS The authors thank Dr. Peiyao Fang for valuable advice and The Animal Research Center of The University of Texas Medical Branch, Galveston, TX for professional care of rabbits. The experiment was sponsored by Shriners Hospitals for Children grants 8630, 8490, and 8790.

REFERENCES 1. 2. 3. 4. 5.

6.

7.

8.

9.

10.

11.

Grewal RS, Gupta SC, Singhal GM, et al. Wound healing in relation to insulin. Int Surg 1972;57:229. Rosenthal SP. Acceleration of primary wound healing by insulin. Arch Surg 1968;96:53. Rosenthal SP, Enquist IF. The effect of insulin on granulating wounds in normal animals. Surgery 1968;64:1096. Pierre EJ, Barrow RE, Hawkins HK, et al. Effects of insulin on wound healing. J Trauma Inj Infect Crit Care 1998;44:342. Edmonds T. Evaluation of the effects of topical insulin on wound healing in the distal limb of the horse. Vet Med Small Animal Clin 1976;71:451. Hanam SR, Singleton CE, Rudek W. The effect of topical insulin on infected cutaneous ulcerations in diabetic and nondiabetic mice. J Foot Surg 1983;22:298. Greenway SE, Filler LE, Greenway LF. Topical insulin in wound healing: A randomized, double-blind, placebo-controlled trial. J Wound Care 1999;8:526. Zhang XJ, Chinkes DL, Irtun ØO, et al. Anabolic action of insulin on skin wound protein is augmented by exogenous amino acids. Am J Physiol Endocrinol Metab 2002;282:E1308. Sakurai Y, Aarsland A, Herndon DN, et al. Stimulation of muscle protein synthesis by long-term insulin infusion in severely burned patients. Ann Surg 1995;222:283. Pelegrinelli FF, Thirone AC, Gasparetti AL, et al. Early steps of insulin action in the skin of intact rats. J Invest Dermatol 2001;117:971. Li LF, Fiedler VC, Kumar R. Induction of hair growth by skin irritants and its relation to skin protein kinase C isoforms. Br J Dermatol 1999;140:616.

96 12.

13.

14.

15.

16.

17. 18.

19.

20.

JOURNAL OF SURGICAL RESEARCH: VOL. 142, NO. 1, SEPTEMBER 2007 Mani I, Iversen L, Ziboh VA. Evidence of nuclear PKC/MAPkinase casecade in guinea pig model of epidermal hyperproliferation. J Invest Dermatol 1999;112:42. Dieckgraefe BK, Weems DM, Santoro SA. ERK and p38 MAP kinase pathways are mediators of intestinal epidermal woundinduced signal transduction. Biochem Biophy Res Commun 1997;233:389. Wang E, Zhao M, Forrester JV, et al. Electric fields and MAP kinase signaling can regulate early wounding in lens epithelium. Invest Opbtalmol Vis Sci 2003;44:244. Hirano S, Rees RS, Gilmont RR. MAP kinase pathways involving Hsp27 regulate fibroblast-mediated wound contraction. J Surg Res 2002;102:7. Zhang X-J, Sakurai Y, Wolfe RR. An animal model for measurement of protein metabolism in the skin. Surgery 1996; 119:326. Zhang X-J, Wu X, Wolf S, et al. Local insulin injection accelerates skin wound healing. J Burn Care Rehab 2003;24:S112. Shearer JD, Coulter CF, Engeland WC, et al. Insulin is degraded extracellularly in wounds by insulin-degrading enzyme (EC 3.4.24.56). Am J Physiol 1997;273:E657. Anderson L, Dinesen B, Jørgensen PN, et al. Enzyme immunoassay for intact human insulin in serum or plasma. Clin Chem 1993;39:578. Gambicher T, Boms S, Stucker M, et al. Acute skin alterations following ultraviolet radiation investigated by optical coherence tomography and histology. Arch Dermatol Res 2005;297:218.

21.

O’Keefe EJ, Chiu ML. Stimulation of thymidine incorporation in keratinocytes by insulin, epidermal growth factor, and placental extract: Comparison with cell number to assess growth. J Invest Dermatol 1988;90:2.

22.

Zhang X-J, Chinkes DL, Wolf SE, et al. Insulin, but not growth hormone, has an anabolic effect on protein metabolism in scalded skin and muscle. Am J Physiol 1999;276:E712.

23.

Hu RH, Yu YM, Costa D, et al. A rabbit model for metabolic studies after burn injury. J Surg Res 1998;75:153.

24.

Nezu R, Takagi Y, Ito I, et al. The importance of total parenteral nutrition-associated tissue zinc distribution in wound healing. Jpn J Surg 1999;29:34.

25.

Cario E, Jung S, Harder-d’Heureuse J, et al. Effects of exogenous zinc supplementation on intestinal epithelial repair in vitro. Eur J Clin Invest 2000;30:419.

26.

Iwata M, Takebayashi T, Ohta H, et al. Zinc accumulation and metallothionein gene expression in the proliferating epidermis during wound healing in mouse skin. Histochem. Cell Biol 1999;112:283.

27.

She QB, Huang JS, Mukherjee JJ, et al. The possible mechanism of synergistic effects of ethanol, zinc, and insulin on DNA synthesis in NTH 3T3 fibroblasts. FEBS Lett 1999;460:199.

28.

Larson DL, Maxwell R, Abston S, et al. Zinc deficiency in burned children. Plast Reconstr Surg 1970;46:13.