Abstract: Acemannan, a complex carbohydrate consisting of a b-(1,4)-linked polymannose acetate, is a major carbohydrate component of aloe vera leaf pulp. To test the effectiveness of acemannan in a wound environment, we utilized a unique mouse footpad model. Chemically pure acemannan was applied topically to incised mouse tarsal wounds either in solution at three different concentrations or in the form of a gel. The wounds were scored by a double-blind procedure that permitted a controlled comparison between treated and control wounds. Statistical analysis determined wounds treated topically with acemannan resulted in significantly better wound healing as reflected by lower wound scores in treated feet compared to controls. This effect was dependent upon the dose of acemannan applied to the wound and was detectable when acemannan was applied in solution or in a gel. Other carbohydrate solutions were tested but showed no effect on wound healing. Because acemannan was suspected to stimulate macrophage cytokine production, mouse peritoneal macrophages were exposed to different doses of acemannan and other carbohydrates in vitro and resulted in a dose-dependent release of IL-1b, TNFa, and IL-6. We hypothesize the effect of acemannan on wound healing is probably mediated through local cytokine release from wound macrophages. Disclosure: This research was funded in part by Carrington Laboratories, Irving, Texas. Dr. Chinnah was employed by Texas A & M University during the period this research was performed. Dr. Tizard received a grant from Carrington Laboratories.

Introduction

Acemannan is a complex carbohydrate isolated from the clear gel in the center of the aloe vera leaf, which consists of polymerized beta-(1,4)-linked acetylmannose. Molecular weight after purification varies from 10-1000kDa with an average of 200kDa. The precise size depends on the degree of degradation during the manufacturing process. Acemannan does not appear to be mitogenic for monocytes, macrophages, or fibroblasts and is noncytotoxic even at high concentrations.[1–3]

Many wound dressings, such as alginates and hydrocolloids, contain various forms of carbohydrates.[4] Differences in carbohydrate chemical structure can lend different properties and functions to wound dressings, such as improvement of absorption, increased binding, or improved retention of moisture. At a cellular level, some carbohydrates appear to stimulate cell functions of proliferation, migration, and cytokine production.[5–10]

Early reports in the literature indicate carbohydrates can enhance wound repair by promoting early mobilization of macrophages to a wounded area.[11] Periwound injections of macrophages activated by the carbohydrate glucan caused an increase in fibroblast proliferation, fibrogenesis, collagen synthesis, epithelialization, and an increase in the tensile strength of the wound.[11] This was presumed to be due to the effect of glucan-induced interleukin-1 (IL-1), and tumor necrosis factor alpha (TNF-apha) on the wound fibroblasts.[11,12] Topical application of glucans to wounds seems to result in more rapid angiogenesis and reepithelialization compared to controls.[12] Other carbohydrate extracts have been shown to accelerate healing of open wounds and burns.[13,14] These extracts stimulate oxygen consumption, increase angiogenesis, and increase collagen synthesis in the wounds.[13]

Because aloe vera sap and acemannan gel have been reported to enhance wound healing, the following experiments were designed to determine if acemannan could enhance healing and to investigate possible mechanisms of the activity of acemannan. The wound model used in this study was adopted and modified from the methods developed previously by others.[11,15] The method developed by these authors has the twin merits of simplicity and speed. The reader chooses which wound of a pair has healed (if any) as well as gives each wound a subjective score. When read blind, the presence of a control on the opposite foot provides an additional measure of wound healing by providing a basis for comparison as well as a control for any bias in the reader’s scoring. This model is simple, yet quick and reliable. The reproducibility and reliability come from the use of blinds and multiple controls. The basic experimental unit consisted of two groups, each of 30 mice. One group of mice in each experiment had controls in both feet, while the other group received control in one foot and treatment in the opposite foot. All experiments were performed blind.

Materials and Methods

Mice. Male Swiss-Webster mice weighing approximately 20g were obtained from Harlan Sprague Dawley, Inc. (Indianapolis, Indiana). Groups of 30 mice were wounded on the posterior tarsus of both feet after being anesthetized with 0.6mL of a mixture of sodium pentothal, propylene glycol, 90-percent ethanol, and sterile distilled water administered intraperitoneally. A 5mm incision was made through the thickness of the skin between the Achilles tendon and the external vestigial finger parallel to the posterior tarsal vein. Immediately after making the incision, one wound was treated with one drop of test material. A 2mm x 4mm piece of sterile parafilm (Parafilm™, American National Can, Chicago, Illinois) was inserted into the wounds to ensure even distribution of the solutions and to prevent rapid adhesion of the wound edges. The parafilm fell out within 24 hours without further intervention. All experiments consisted of one treatment group and one control group of 30 mice.

The wounds were examined daily and measured by a subjective scoring system on a scale of 5 to 0. The basis of each score is as follows: 5 is open and totally unhealed; 4 is a small slit; 3 is a closed wound with a scab; 2 is a closed wound with scab shed; 1 is a healed wound with a visible scar; 0 is a healed wound and no visible scar. Wounds with represented scores are shown in Figure 1. Comparison of both feet of each animal was made at the same time. All wounds were made by the same individual and read by a different investigator. The wounds were read blind; therefore, the reader was unaware of treatment applied, which wound was the control, or if both wounded feet were treated with control material.

Figure 1

The scoring system for mouse tarsal wounds. This ranges from 0 (completely healed) to 5 (completely unhealed).

Treatments. See Table 1 for a complete summary of test articles and concentrations used in the experiments.

Table 1

Solutions. The control solution was sterile 0.9-percent saline. The test solution was made from freeze-dried acemannan from lot numbers PD1COO2 and PD1COO3 (Carrington Laboratories Inc., Irving, Texas). Acemannan was provided as a freeze-dried pellet that was reconstituted with sterile saline to 1µg/mL. All acemannan dilutions were made from this stock solution. Sodium hydroxide was used to hydrolyze and deactivate acemannan. The acemannan was hydrolyzed and column purified by chemists. No visible toxic effects from the hydrolyzed acemannan were seen in cellular bioassays indicating little if any sodium hydroxide residue was present in the preparation. Hydrolyzed acemannan was reconstituted to 200 and 20µg/mL using sterile saline. Mannose, 99-percent pure (Aldrich, Milwaukee, Wisconsin), was diluted to 200 and 20µg/mL using sterile saline.

Gels. A gel containing 0.1-percent acemannan in polyacrylic acid (Carbopol, Merck, West Point, Pennsylvania) was used. Polyacrylic acid gel without acemannan was used as a control and referred to as excipient.

Injectables. One milliliter of saline or acemannan solution was injected into the peritoneal cavity of mice immediately after tarsal wounding. All injectable solutions were made from a single freshly constituted stock solution of acemannan

Statistical analysis. All scores from these experiments had the variance and standard deviation calculated. Analysis of variance (ANOVA) was used to determine if the scores of treated and control wounds were significantly different. Statistical analysis was performed using Analyse-It for Microsoft Excel, version 1.62 (Analyse-It Software, Ltd., Leeds, England, UK).

Cytokine assays. Harvesting wound macrophages can be difficult. Since peritoneal macrophages behave similarly to macrophages found in wounds, these cells were used to determine if specific carbohydrates could induce proinflammatory cytokine production. Naïve mice were killed and 3mL of RPMI 1640 supplemented with 50U/mL penicillin and 50µg/mL streptomycin (P/S) was injected into the peritoneal cavity and the abdominal region was massaged. The solution was removed and the peritoneal cavity was washed twice more with 3mL of RPMI. These washes were pooled with the original wash. The cells were centrifuged at 1000rpm for 20 minutes and the pellet was resuspended in 2mL of RPMI 1640 supplemented with 10-percent FBS and P/S. The cells were counted using a hemocytometer.

The cell suspension was diluted in RPMI 1640 supplemented with 10-percent FBS and P/S to give a final concentration of 5 x 106 macrophages mL-1. A 12-well tissue culture plate was used for each cell concentration, and the cells were incubated at 37 degrees C in 5-percent CO2 for four hours. At the end of that time, the media was removed and replaced by RPMI 1640 without FBS containing 0.5, 10, 50, 100, or 200µg/mL acemannan or 0.5, 10, 20, 50, 100, or 200µg/mL hydrolyzed (deactivated) acemannan or 200µg/mL mannose. The cells were incubated for 24 hours; the media from each well was harvested and centrifuged at 2000rpm for 20 minutes; and the supernatants were frozen at –70 degrees C. The supernatant fluid was analyzed for the presence of TNF-alpha, IL-1beta, and IL-6 by ELISA assays (Genzyme, Cambridge, Massachusetts). Each fluid sample was assayed in duplicate. All test kits were used following the manufacturer’s recommendations with appropriate controls.

Histology. Samples for histological examination were taken on alternate days after wounding. Two animals from each group were used resulting in eight specimens. Tissue was fixed within five minutes of death in 10-percent buffered neutral formalin (pH7). The tissue was decalcified with equal portions of A) 50g sodium citrate in 250mL distilled water and B) 125mL of 90-percent formic acid and 125mL distilled water for two to three weeks. Serial paraffin sections were stained with hematoxylin and eosin (H&E) for general differences and with Masson’s trichrome connective tissue stain to distinguish collagen and muscle tissue.[16] Twenty-five slides per specimen per day post-wounding were evaluated. Slides were randomized and read by a blinded evaluator. The evaluator used a descriptive analysis for each slide. In addition to general comments, the descriptive analysis utilized a 0 to 3 plus scale to consistently describe the epithelium migration and thickness, granulation tissue composition, and collagen deposition and thickness in order to objectively determine what, if any, descriptive differences were recorded as opposed to determining statistical differences.

Results

Control wounds. Throughout the study, the control group receiving controls in both feet consistently had the right foot score slightly better (1 score better) than the left foot. Because this occurred in every control group, it is assumed, therefore, the handedness bias of the same surgeon occurred in the treated groups of mice and the score difference was adjusted accordingly (no less than 0.83 and no more than 0.93 to each left foot score depending on control group differences) per discussions with a statistician.

Response of wounds to aqueous acemannan. Three different doses of aqueous acemannan (5, 20, or 100µg/mL) or sterile saline were applied once to the wound at the time of injury. Wounds were read and scored daily for seven days. In all mice that received saline in one wound and acemannan solution in the other, the saline control wounds consistently healed more slowly. Wounds treated with 20µg/mL acemannan solution had better healing scores indicating the wounds healed rapidly compared to controls (Figure 2). The wound scores for the 20µg/mL solution were statistically significant for all seven days (p < 0.0001, p = 0.0112, p = 0.0027, p = 0.0257, p = 0.0007, p = 0.0246, and p = 0.0172, respectively). The differences between control wounds and wounds treated with 100µg/mL acemannan solution were not as pronounced as the 20µg/mL differences. Statistically significantly better scores were noted on days 2, 3, 5, 6, and 7 (p = 0.0002, p = 0.0003, p = 0.0288, p = 0.0402, and p = 0.0141, respectively). The wounds treated with 5µg/mL acemannan solution also showed significantly enhanced healing scores compared to saline controls at days 1, 6, and 7 (p < 0.0001, p = 0.0436, p = 0.0425, respectively).

Figure 2

Figure 2. Results of mouse tarsal wounds treated with 0, 5, 20, or 100µg/mL of acemannan at the time of injury. Wounds treated with 20µg/mL acemannan solution had statistically significant lower healing scores than controls.

Histological examination as described in the Methodology section suggested faster reepithelialization as well as greater collagen deposition, which occurred in all acemannan-treated wounds as compared to control wounds. Relatively little repair was detectable in saline control wounds by day 2 after wounding. At this time, the acemannan-treated wound epithelium had begun migrating under the scab. By day 4 after wounding, epithelium had begun migrating under the scab in saline control wounds. The acemannan-treated wounds at this time had a complete epithelium, and changes in the dermis had become apparent. By six days after wounding, the saline control wounds had complete epithelium and the dermis remained largely devoid of substantial collagen deposition. These saline wounds showed some granulation tissue formation with a small amount of collagen deposition. The wounds treated with acemannan had complete and thick epidermal layers by day 6. A large amount of granulation tissue had formed and it contained evenly deposited collagen fibers. By eight days, the saline control wounds had an incomplete dermis and granulation tissue formation. This granulation tissue contained evenly distributed collagen. Acemannan-treated wounds at this time had small amounts of granulation tissue and large amounts of collagen. This granulation tissue appeared to have more collagen fibers than the saline control wounds (Figure 3). Because there was less granulation tissue and it appeared more compact than the previous two days, it is possible that matrix remodeling and contraction had begun. The granulation tissue in saline control wounds at eight days after wounding was not as compact and did not contain as many collagen fibers in acemannan-treated wounds. Overall, acemannan-treated wounds showed complete epithelialization and collagen deposition approximately two days earlier than saline controls. The collagen fibrils in acemannan-treated wounds appeared to be thicker than the control wound fibers.

Figure 3

A: Saline control mouse tarsal wound at day 8, Masson’s, 200X. The epidermis is complete with all cell layers present (see black arrow). The dermis now has a purple appearance indicating the pink granulation tissue is being replaced with blue collagen fibers. At the white open arrow, a distinct blue collagen fibril can be seen next to the loosely packed pink/purple granulation tissue. B: 20µg/mL-acemannan–treated mouse tarsal wound at day 8, Masson’s, 200X. The epidermis is complete with all cell layers present (see black arrow). The dermis now dark purple from abundant blue collagen and has relatively little pink granulation tissue. Densely packed collagen fibers can be seen at the white open arrow.

Response of wounds to acemannan gel. To determine if acemannan delivered in a gel enhanced wound healing compared to excipient control gel, groups of 30 mice were wounded as described above, and control wounds were given excipient gel. Thus, one group of 30 mice received excipient gel in both wounds, while the other group received excipient gel in one foot and 0.1-percent acemannan gel in the other. Wounds were read and scored daily for eight days.

Wounds treated with acemannan gel consistently had lower scores than excipient gel wounds throughout the experiment (Figure 4). The wounds in the acemannan gel-treated group had statistically significantly better scores on days two through seven (p = 0.0024, p = 0.0012, p = 0.0119, p = 0.0003, p = 0.0003, p = 0.0151, respectively). By day eight, both types of wounds were essentially completely healed, and differences between them were insignificant.

Figure 4

Figure 4. Results of mouse tarsal wounds treated with excipient gel or acemannan gel.

Systemic effect. To determine if acemannan exerted a systemic effect on the rate of tarsal wound healing, a slightly different form of experiment was devised. Wounded mice were injected intraperitoneally with 1mL of saline or with one of three doses of acemannan (25, 50, or 100µg/mL) immediately after wounding. Wounds were made on both feet and scored daily as described above. No statistical differences in the scores of wounds of saline controls or acemannan-treated mice were seen irrespective of the dose of acemannan administered.

Other carbohydrate solutions. To determine if other carbohydrates could have an effect on healing, 20µg/mL solutions of mannose and hydrolyzed (deactivated) acemannan were used in the animal model. Neither 20µg/mL solution of mannose (Figure 5) or hydrolyzed acemannan (Figure 6) had an effect on wound healing scores. Both figures illustrate the reproducibility of the wound model and the scoring system.

Figure 5

Results of mouse tarsal wounds treated with 0 and 20µg/mL mannose solution.

Figure 6

Results of mouse tarsal wounds treated with 0 and 20µg/mL-hydrolyzed (deactivated) acemannan solution.

Cytokine release assays. Mouse peritoneal macrophages showed a dose-dependent response to acemannan in vitro (Figure 7). TNFa was released in greatest quantity with 1900pg/mL released following exposure to 200µg/mL acemannan. IL-1 was also released in a dose-dependent fashion with 569pg/mL being released after being exposed to 200µg/mL acemannan. IL-6 was released in an inverse response with 420pg/mL released after being exposed to 100µg/mL and 220pg/mL. released after being exposed to 200µg/mL acemannan. Macrophages exposed to 200µg/mL hydrolyzed acemannan did not release a detectable amount of TNF-alpha. Macrophages exposed to 200µg/mL mannose released 276pg/mL TNF-alpha.

Figure 7

Mouse peritoneal macrophage cytokine release after exposure to increasing doses of acemannan measured by ELISA.

Discussion

The results obtained from these experiments suggest that acemannan applied topically as a solution or as a gel enhanced wound healing in linear incisions in the tarsi of mice as depicted in the low healing scores. All acemannan-treated groups had lower scores, which represent faster wound closure than saline controls. Changing the chemical structure of acemannan rendered the wound healing effects of the molecule ineffective.

Histological examinations correlated well with the healing scores. Histological samples indicated the acemannan-treated wounds appeared to reepithelialize and deposit collagen approximately two days sooner than saline control wounds. Wounds treated with acemannan had thicker collagen fibrils than saline control wounds, and it could be speculated that collagen I may be replacing collagen III earlier in these wounds. Experiments are under way to determine if this is the case. The collagen deposition in acemannan-treated wounds appeared to be more evenly distributed than the saline control wounds.

Complex carbohydrates are believed to cause macrophage activation because they act as foreign bodies within the cell.[17] Activated macrophages release cytokines that may promote wound repair. Glucan-treated mouse serum has been shown to contain elevated colony-stimulating activity and macrophage colony-stimulating activity levels.[18] In addition, supernatant fluids from glucan, activated macrophage cultures, or topical glucan appears to increase the breaking strength of mouse wounds.[12,19] Acemannan is rapidly phagocytosed by mouse macrophages.[1,20] Although acemannan appears to be soluble and forms a clear solution when reconstituted, it is probable that small particles remain in suspension and may be readily ingested by macrophages. For example, acemannan labeled with 14C and injected intravenously or intraperitoneally was deposited in the liver and spleen of dogs within 48 hours.[21] Thus, the distribution of acemannan was consistent with removal of foreign particles by the mononuclear phagocytic system, since these organs are known to play major roles in host defenses by removing particulate antigen from the blood. The presence of acemannan may activate the macrophage foreign body response and, in turn, enhance wound healing. Alternatively, molecular aspects of acemannan may possibly bind to mannose receptors on the macrophage cell surface and subsequently trigger a signal that activates the macrophage similar to Pneumocystis cranii uptake.[22] The presence of the activated macrophages and the cytokines released may enhance the wound healing. In aged animals, macro-phage function is known to be a limiting factor in wound healing and the replacement of impaired macrophages improved wound healing.[23] Injection of acemannan into periwound tissue has been shown to reduce time to healing in aged animals.[24] These findings support the theory that acemannan interacts with macrophages within the wound, but these findings do not elucidate the mechanism of the interaction.

Acemannan enhances both macrophage phagocytic activity and nonspecific cytotoxicity as well as the expression of MHC Class II antigens.[1] It is possible that this increased expression of MHC Class II molecules may enhance antigen presentation in acemannan-treated animals.[25] From the results shown here, it is now apparent that acemannan, once within macrophages, causes their activation and promotes the synthesis of TNF-alpha, IL-1, and IL-6.[1,9,10,26,27] All three of these cytokines have been shown to be involved in the wound healing process.[28–30] Indeed, creams that contain IL-1 have been shown to accelerate wound healing.[30]

Systemic acemannan in the concentrations tested did not show an affect on healing, but acemannan applied topically does. The authors conclude acemannan must have been acting locally in the wound. Because acemannan and other carbohydrates can stimulate cytokine production by the macrophage, the authors hypothesize the increased synthesis of TNF-alpha, IL-1, and IL-6 in acemannan-treated wounds resulted in lower healing scores. Carbohydrates, like mannose and hydrolyzed acemannan, when tested at similar concentrations, did not stimulate cytokine production or have an effect on healing scores as acemannan did. This would indicate acemannan does not bind to the mannose receptor on the macrophage causing activation as previously thought. Further studies are underway to elucidate the mechanism of action for acemannan.

Acknowledgments

The authors would like to thank Bryan Maxwell and Yawei Ni for technical support. The gel containing 0.1-percent acemannan in polyacrylic acid was provided by Carrington Laboratories, Inc., Irving, Texas.