Temporal Effects of Topical Morphine Application on Cutaneous Wound Healing

A cutaneous wound healing model was utilized to evaluate wound closure rates in rats. A total of 78 male Sprague-Dawley rats (Harlan, Indianapolis, IN) approximately 8 weeks of age (200–220 g body weight) were randomly assigned to one of five treatment groups. Rats were subsequently anesthetized by intraperitoneal administration of 65 mg/kg ketamine hydrochloride and 5.5 mg/kg xylazine hydrochloride and the midperiscapular region clipped and shaved. A skin biopsy punch was used to excise a 4-mm-diameter full-thickness skin flap from the midline below the scapulae to a depth just above the panniculus carnosus muscle.21All rats were then housed individually to prevent cage mates from grooming or otherwise disturbing the wound. Animal facilities were temperature- and humidity-controlled with a 12-h dark-light cycle and food and water ad libitum . All surgical procedures and animal handling were performed in accordance with National Institutes of Health laboratory care standards and were approved by the University of Kansas Medical Center Animal Care and Use Committee (Kansas City, Kansas).

Morphine sulfate (25 mg/ml) (Abbott Laboratories, North Chicago, IL) was infused into IntraSite Gel (amorphous hydrogel; Smith+Nephew, Hull, United Kingdom) at a concentration of 5 mm. Gel and drug were combined in 3-ml syringes by repeated passage through a Luer Lock stopcock (PPG Industries, Pittsburgh, PA). Beginning 1 h postsurgery, 150-μl IntraSite Gel alone or morphine-infused IntraSite Gel was applied topically to the wound twice daily. Animals were divided into four treatment groups. Control animals received applications of IntraSite Gel alone throughout the entire time course (n = 37). Saline was added to control gel to match the saline content of the morphine-infused gel. A second treatment group received 5 mm morphine sulfate on days 0–3 postwounding and gel-only treatments for the remaining 7 days (n = 5). The third group received gel-only applications on wound days 0–3 and morphine sulfate-infused gel on days 4–10 postwounding (n = 6). A fourth group was treated with 5 mm morphine sulfate throughout the entire time course or until euthanasia (n = 25). The final treatment group received the opioid receptor antagonist naltrexone (1 mm) throughout the entire time course (n = 5).

Wound images were captured each morning prior to treatment using a handheld digital camera. A bar attached to the camera provided a fixed focal distance target for wound imaging. A size standard with known surface area was affixed to the target bar and included in each image. Wound surface area was determined using a computerized planimetric program (Scion Image, Fredrick, MD). The area occupied by the wound was defined by the boundary created by the granulation tissue or scab/intact tissue interface. Wound area data generated by Scion Image were converted from pixels to area units of mm²by comparison to the known area of the fixed size standard.

Rats were decapitated and wound tissue including approximately 1.0 cm of surrounding intact skin was harvested. Skin was dissected from rats in each treatment group on postwound day 0 (n = 3), 1 (n = 8), 3 (n = 6), 5 (n = 6), 8 (n = 6), and 18 (n = 9). Tissue was embedded in tissue-freezing medium (Electron Microscopy Sciences, Hatfield, PA), frozen on dry ice and cryosectioned serially throughout the wound site at 14 μm thickness. Sections at 10-section intervals were placed on slides and stored at −80°C until staining.

Sections from day 18 postwounding were postfixed at room temperature with 4% paraformaldehyde for 5 min and stained with hematoxylin and eosin. Slides were mounted with Permount (Fisher, St. Louis, MO). Images were captured using light microscopy at a magnification of ×20 (Nikon Eclipse 80i microscope, Nikon Digital Sight Fi1 camera; Nikon, Melville, NY). Epidermal and full-skin thicknesses were quantified in three sections equally spaced throughout each wound. In each analyzed section, two regions were selected randomly. All measurements were performed blind with ImageJ (National Institutes of Health, Bethesda, MD) analysis software.

Tissue sections from postwound days 0, 1, 3, 5, and 8 were postfixed in 4% paraformaldehyde for 5 min, blocked in 5% goat serum for 1 h, and directly immunostained (90 min at room temperature) with antibodies for the macrophage marker macrosialin (rat polyclonal immunoglobulin IgG, fluorescein isothiocyanate conjugated, 1:200, clone FA-11, Serotec, Oxford, United Kingdom) and for the myofibroblast marker α-smooth muscle actin (α-SMA, mouse monoclonal IgG 2A, Cy3 conjugated, 1:200, clone 1A4, Sigma, St. Louis, MO). Rat serum was added to the blocking serum of sections stained with macrosialin antisera to negate the impact of nonspecific staining. In addition, a rat IgG antibody conjugated to fluorescein isothiocyanate (1:100, Serotec, clone YTH71.3) was utilized as a negative control, and no specific staining was detected with this approach. In pilot studies, ED1 mouse monoclonal IgG (Chemicon, Temecula, CA), CD68 goat polyclonal (Santa Cruz Biotechnology, Santa Cruz, CA), ED1 mouse monoclonal IgG (Serotec), and CD68 mouse monoclonal IgG (Biomeda, Foster City, CA) antibodies were evaluated, and they demonstrated similar staining patterns. However, they exhibited reduced sensitivity compared with the antibody selected and were thus not utilized for quantification.

Slides were mounted with Fluoromount G (Fisher) and viewed using a Nikon Eclipse 80i microscope. Images were captured with a Nikon Digital Sight Fi1 camera and adjusted for brightness and contrast. Immunoreactive (-ir) morphologic structures were quantified in three sections equally spaced throughout each wound. In each analyzed section, four regions, each equaling an area of 0.35 mm², were selected randomly; two images from within the wound and two at the edge of the lesion were analyzed.

Numbers of cells macrosialin-ir were averaged and expressed as counts per unit area. The α-SMA-ir area was obtained by threshold discrimination, divided by the total area analyzed and expressed as a percent of field area. The α-SMA-ir vasculature was excluded from analysis and only cells demonstrating a spindle-shaped myofibroblast morphology were included. All images were evaluated by a blind observer using ImageJ analysis software.

Statistical analyses were performed using SigmaStat (San Jose, CA). Data are reported as mean ± SEM. The effects of drug treatment and time on wound closure, macrophage infiltration, and myofibroblast activation were evaluated using separate two-way repeated measures analyses of variance. Differences between treatment groups and within treatment groups over time were identified using Tukey tests. The long-term effects of morphine application on skin thickness and residual scar area were analyzed using unpaired Student t  tests. Differences between means were considered significant when P < 0.05.

The effects of topical morphine sulfate application at different times throughout healing on cutaneous wounds were assessed using a standardized model of cutaneous wound healing in rats. A previous concentration-response study conducted in this lab demonstrated that 5 mm morphine sulfate causes a significant, but not maximal, delay in wound closure and was therefore chosen as the concentration to be used in subsequent experiments. Topical morphine treatment on wound days 0–3 significantly delayed wound closure when compared with controls (P = 0.017). Animals treated with 5 mm morphine sulfate on days 0–3 postwounding had significantly larger wounds on days 1, 2, 3, 4, 6, and 7 when compared with gel-only treated controls (fig. 1). Total wound area over the complete time course of animals receiving 5 mm morphine sulfate treatments for the first 4 days was 14% larger than both gel-only treated control rats and rats treated with morphine sulfate for the last 7 days. In addition, no significant differences were seen in animals treated with 5 mm morphine sulfate on days 4–10 postwounding when compared with gel-only treated controls. In rats treated topically with gel infused with 1 mm naltrexone, the area of the wounds treated with the opioid receptor antagonist did not differ significantly from control wounds (data not shown).

Macrophage infiltration into the closing wound was assessed with immunohistochemistry using the macrophage marker macrosialin. Very few macrosialin-ir cells were observed in control skin adjacent to the periwound area. Macrosialin-ir cells in gel-only treated wounds were greatest on day 1 postwounding (fig. 2). A slight decrease in the number of macrophages in the wound was seen on wound day 3 followed by a return to near-baseline levels on days 5 and 8 postwounding. Appearance of macrosialin-ir cells was significantly delayed in morphine-treated wounds (P = 0.01). Macrophage concentrations in wounds treated with morphine sulfate did not peak until wound day 3. A significant decrease in macrosialin-ir cell numbers of 51% in morphine-treated wounds was observed on wound day 1 when compared with controls treated with gel only. While macrosialin-ir cell numbers declined in morphine-treated wounds on day 5 postwounding, they were significantly higher than control levels and remained elevated through wound day 8 compared with controls treated with gel only.

The α-SMA-immunoreactivity was utilized to measure myofibroblast density in the healing wounds. The α-SMA-ir cells first appear on wound day 3, although at negligible levels. In gel-only treated wounds, α-SMA-ir cell density increased rapidly at day 5 postwounding and still further at wound day 8 (fig. 3). However, this increase was markedly reduced in wounds treated with topical morphine sulfate (P < 0.001). The α-SMA-ir cell density was significantly lower in morphine-treated wounds on days 5 and 8 postwounding compared with controls, with an 84% and 79% reduction seen on days 5 and 8, respectively.

The long-term impact of topical morphine sulfate treatment during cutaneous wound healing on the structural architecture of healed skin was assessed at postwound day 18 using measures of epidermal and dermal thickness and the extent of scar tissue formation in rats receiving topical 5 mm morphine sulfate applications days 0–18 postwounding. Control wounds treated with gel alone had an average epidermal thickness of 71.5 μm, while the full thickness of the skin (epidermis and dermis) was 1.6 mm. Closed wounds of rats treated with topical morphine had significantly thinner (35%, P = 0.019) epidermis over the wound compared with controls treated with gel only (fig. 4A). The full-skin thickness was 19% (P = 0.014) less in closed morphine-treated wounds as well (fig. 4B). In addition, topical morphine treatment resulted in significantly larger (41%, P = 0.023, fig. 4C) scar tissue surface area on day 18 postwounding.

Previous experimental studies in rats have demonstrated that treatment with topical morphine results in a delay in the closure of cutaneous wounds via  inhibition of peripheral neuropeptide release from primary afferent neurons.18The onset of the delay begins within the first 24 h after wounding and continues for the first few days of healing. After approximately 4 days following wounding, wound closure accelerates with wounds ultimately closing fully at times not significantly different than controls. These results suggest a time-dependent effect of topical morphine treatment on wound closure. Therefore, these studies were designed to begin to characterize the temporal characteristics of morphine's impact on wound healing. A standardized model of cutaneous wound healing was used that is sensitive enough to detect age-related variations in wound closure as well as the negative impact of partial sensory denervation on wound healing.17,21 

The results of the current study demonstrate that the closure of wounds treated with morphine on days 0–3 postwounding only is nearly identical to that of wounds treated with morphine throughout the entire healing process,18with an immediate delay in closure followed by acceleration and full closure at approximately day 10 postwounding. However, wounds treated with morphine beginning on day 4 postwounding and continuing through the remainder of the time course were not significantly different from controls treated with gel alone. These results demonstrate that physiologic events in wound healing that facilitate morphine-induced delays in closure occur early in the healing process.

Neuropeptides have proinflammatory actions and serve as a link between the nervous and immune systems. For example, substance P promotes chemotaxis and increased function of macrophages,22as well as cytokine release from human monocytes.23,24Macrophages are essential for normal wound healing and are the predominant cell type in the wound in the early inflammatory phase of wound healing.25Macrophages not only clear the wound of bacteria and cellular debris but also secrete a number of growth factors and other cytokines that induce angiogenesis and stimulate keratinocytes and fibroblasts.26–30Accordingly, the effect of topical morphine treatment on macrophage density within the healing wound was assessed.

Activation of opioid receptors located on primary afferent neurons inhibits the antidromic release of neuropeptides from these terminals. Neuropeptides both attract and activate macrophages in states of inflammation. Therefore, decreased concentrations of neuropeptides within the periwound area may result in reduced macrophage infiltration. Results from this study demonstrate the inhibitory effect of topical morphine on macrophage infiltration into the closing wound. Macrophage density within control wounds peaked 1 day following injury. The number of macrophages in the wound began to decline on day 3 postwounding and continued through days 5 and 8 after wounding. The number of macrophages in the wound decreased to near-baseline levels by day 5 after wounding, a time consistent with the resolution of the inflammatory phase in normal wound healing. However, macrophage migration was significantly delayed in morphine-treated wounds. In this group, macrophage numbers were significantly less on day 1 postwounding and a maximum was not reached until wound day 3. Decreased macrophage density would result in reduced induction of the parenchymal cells essential for wound closure.29–31Hence, the results of this study suggest that topical morphine delays wound closure, at least in part, by inhibiting the recruitment of macrophages into the closing wound. In addition to delaying macrophage infiltration, topical morphine application prolonged the presence of macrophages within the wound. Macrophage density in morphine-treated wounds was significantly higher at later time points (wound days 5 and 8) than in controls treated with gel only. This finding suggests that morphine inhibits resolution of the inflammatory phase, which may result in aberrant healing, propagate chronic inflammation, and result in undue tissue damage.

Wound healing is a dynamic process consisting of a complex, overlapping pattern of events that have been artificially divided into three phases: the inflammatory phase, including coagulation and inflammation; the proliferative phase, consisting of angiogenesis, the formation of granulation tissue, and reepithelialization; and the maturation phase, involving matrix formation and tissue remodeling.32Myofibroblasts are transdifferentiated fibroblasts that express α-SMA, giving them contractile capabilities.33,34Myofibroblasts are the predominant cells recruited by macrophages to initiate the proliferative phase of wound healing.35,36Closure of excision wounds occurs primarily through the contractile actions of myofibroblasts. The α-SMA-immunoreactive myofibroblasts appeared within the wound in this study on day 3 postwounding in control animals. The density of myofibroblasts was significantly increased through day 8. This trend is consistent with the pattern of normal wound healing. However, the presence of myofibroblasts was drastically reduced in morphine-treated wounds, with only a small increase occurring by wound day 8. These data are consistent with the idea that topical morphine application prolongs the inflammatory phase of wound closure and concomitantly delays progression into the proliferation phase of wound healing.

Examining the effects of exogenous opioids on the closure of cutaneous wounds led us to question whether endogenous opioids play a role in regulating their closure. Previous studies have demonstrated that endogenous opioids produced by resident immune cells produce antinociceptive effects in inflamed tissues.19,37Accordingly, production of local endogenous opioids could impact wound closure rates. The possibility of delayed wound closure by activation of peripheral opioid receptors by endogenous opioids during normal wound closure was tested in this study, when the opioid antagonist naltrexone was applied topically to wounds at a concentration of 1 mm; however, no significant differences in wound closure rates resulted from this treatment (data not shown). This suggests that while local endogenous opioids can contribute to peripheral analgesia, their actions do not significantly impact wound closure and, therefore, do not modulate the effects of topical morphine on wound-closure dynamics.

Although morphine-treated wounds ultimately closed at times similar to control wounds, it was reasonable to suspect that morphine could have longer-term effects than that of delaying wound closure. Alterations in the cellular content of the healing wound or modulation of other mechanisms of wound closure by morphine could result in aberrant organization of newly formed skin despite apparent wound closure. Evaluation of the architecture of healed wound skin on day 18 demonstrated a significant decrease in the thickness of the healed skin over the wound following morphine treatment. Topical morphine applied throughout the entire time course of wound closure also resulted in a significant increase in the residual scar area. Analysis of the myofibroblast density during the healing process revealed a significant decrease in the numbers of activated myofibroblasts in morphine-treated wounds, which supports the likelihood that a larger residual scar is formed due to the lack of contractility of the newly formed tissue. These results demonstrate that topical morphine has lasting effects beyond the duration of its application to the healing wound and suggest that its use as a topical analgesic agent could produce persistent detrimental effects that would reduce the strength of healed skin.

Taken together, these studies provide evidence that the delays in wound closure seen with topical morphine application can be attributed to alterations in the initiation and duration of essential, early processes during wound healing. Morphine-induced delays in the onset of inflammation produce a shift in the timing of essential subsequent events (such as myofibroblast activation) in the healing process. Alterations in the temporal processes of wound healing not only result in delayed wound closure, but also result in long-term architectural deficits, jeopardizing the integrity of the healed skin following topical morphine administration. Topical morphine application is most widely used clinically on wounds of elderly patients with pressure ulcers or burn patients with highly inflamed wounds. It is important to acknowledge that the current studies were conducted on young, healthy animals with aseptic wounds. Further study is required to characterize fully the effects of topical morphine on wounds with comorbid disease states.

The authors would like to thank Michelle Winter, B.S. (Research Associate, Department of Pharmacology, Toxicology and Therapeutics, University of Kansas Medical Center, Kansas City, Kansas), and Donald Warn, Ph.D. (Imaging Center Managing Director, University of Kansas Medical Center, Kansas City, Kansas), for their expert technical assistance and the contributions of the University of Kansas Medical Center Kansas Intellectual and Developmental Disabilities Research Center core facilities.