Medical Honey and Silver Dressings Do Not Interfere with Each Other’s Key Functional Attributes

Objective. This study sought to determine whether silver-containing dressings and medical-grade honey gel interfere with one another in measurable ways. Materials and Methods. Dressings applied together in clinical use were tested using in vitro and ex vivo methods to determine whether the combined modalities maintain their individual properties. In order to determine if the presence of silver dressings interfere with honey’s osmotic strength, which is a key physical property of medical honey, changes in honey’s 2 primary sugars were measured, as well as changes in its overall osmotic strength. Finally, the antibacterial barrier activity of the dressings were tested individually and in honey/silver pairs in 2 in vitro models with 2 clinical strains of bacteria. Results. The data demonstrate that honey with silver dressings resulted in an increased osmolarity, since both the concentration of the 2 primary sugars in honey as well as its overall osmolarity increased. The data also demonstrate that the in vitro antibacterial barrier activity seen with silver-containing dressings does not decrease with the addition of medical honey and in some instances increased. Conclusion. Altogether, these data suggest that these 2 classes of dressings do not interfere with each other. Clinical evidence is still required to fully validate these findings.

Honey has been used medicinally for centuries and more recent investigations have begun to characterize the properties of honey responsible for its apparent benefits in wound healing.¹ Of particular interest is honey’s currently accepted mechanism of the promotion of natural healing by maintaining a moist wound environment conducive to healing and promoting autolytic debridement of necrotic tissues.^2-5 This activity is based on honey’s naturally high solute content, and thus a low water content, which produces an osmotic pressure gradient. This pressure gradient draws fluids (exudates) out of the wound bed to the wound surface, which aids in the removal of non-vital tissues.

  Since medical honey is regulated as a medical device in the United States without antimicrobial claims, the product is frequently combined with other secondary dressings that may contain silver compounds to manage wound bioburden. Given the findings in vitro that silver was demonstrated to interfere with the activity of enzymatic debriding agents,⁶ there is a need to evaluate other combinations, including honey, for any measurable impact.

  While honey is a complex natural material, its primary constituents by mass are a variety of sugars with glucose and fructose comprising, on average, 70% of honey’s total mass (~80% of total sugars), depending upon its source.⁷ The high osmotic strength of honey is due to it having more sugars and less water than human tissue; any factors that alter either of these 2 contributing components to osmotic effects would be expected to alter this mode of action, for better or worse. Osmotic strength is a consequence of the number of freely diffusing solubilized molecules with respect to the amount of solvent (ie, water), and anything affecting this number or the amount of water would be expected to weaken or further strengthen a solution’s osmotic strength.

  Ionic silver (the monovalent silver ion) is thought to derive its primary antimicrobial activities from chemically reacting with, and forming adducts with amino acid side chains in microbial proteins.^8-12 This mechanism requires that the silver remain ionized, that it remain capable of diffusing to its microbial targets within the dressing, and that it not form adducts with other agents en route to the target bacteria. Agents or conditions that hinder silver’s diffusion or provide a surrogate adduction target would be expected to reduce silver’s antimicrobial activity.

  Given silver’s mechanism of forming protein adducts, even if silver does react with the sugars in honey, it would not be expected to decrease the number of modified sugar molecules and would therefore not be expected to interfere with the osmotic strength of honey. However, it remains a possibility that the silver might interfere by some unexpected mechanism. Honey, on the other hand, would be expected to provide a very difficult environment for diffusion of the silver as well as many possible surrogate targets for the silver to adduct, and could potentially attenuate the antimicrobial activity of silver-containing dressings.

  A series of experiments were performed to determine if either dressing interferes with the other. First, the authors tested whether 5 different silver-containing dressings combined with medical honey impacted the amount of measurable sugars in medical honey. Next, they tested whether the combination of the same silver-containing dressings with honey impacted the measured osmolarity of medical honey. Finally, they tested whether combinations of honey and a silver-containing dressing had any impact on the antibacterial barrier activity of the silver-containing dressings using in vitro zone of inhibition assays and an ex vivo porcine skin wound model of bacteria colonization using 2 clinically relevant bacteria.

In order to ascertain whether the addition of silver-containing dressings interferes with the primary sugars or osmotic strength of medical honey, 5 silver containing dressings (1 soluble glass powder [Arglaes Powder], 1 hydrogel [SilvaSorb Gel], 1 collagen [Puracol Plus Ag+], 1 foam [Optifoam Ag+ Non-Adhesive], and 1 alginate [Maxorb Extra Ag+]) were combined with a 100% medical honey gel dressing (TheraHoney Gel, Manuka Honey, no additives) and compared to the dressings or medical honey alone (control). All products were manufactured my Medline Industries, Inc, Mundelein, IL.

  Biochemical assessment of honey sugars. For these experiments, the honey was diluted (1/10 fold) to ensure reproducible amounts of honey were introduced to the test products (n = 6); to enable consistent mixing, thus ensuring maximum exposure of honey’s sugars to the potential interacting substances; and for compatibility with the assays and equipment used. Test products were also incubated in phosphate buffered saline (PBS) (n = 5) to control for changes due to the PBS diluent.

  Fructose and glucose were measured with a well-established mix-and-read assay (FA-20, Sigma Aldrich, St. Louis, MO). The assay has 2 components. The first is a dye that changes color in the presence of glucose; the second is an enzyme that turns fructose into glucose. Two separate assays were performed for each sample, 1 to measure glucose only, the other treated by the enzyme and then used to measure the glucose. The first value represents the total glucose, the second represents total glucose and fructose, and by subtracting the 2 values, the individual concentrations of fructose and glucose can be determined. Any chemical modifications to the fructose or glucose would be expected to preclude the ability to be quantified by this assay.

  Assessment of osmotic strength. The authors used vapor pressure osmometry to determine any changes in gross osmolarity. A vapor pressure osmometer (OSMOMAT 070, Gonotec, Berlin, Germany) was calibrated with normal saline (308 mOsmol/kg). The osmometer was then used to measure the diluted honey stocks and the honey post-incubation with the silver-containing products.

  Agar plate zone of inhibition assay. The agar plate zone of inhibition assay was performed using standard techniques adapted from clinical microbiology testing for antibiotic sensitivity of clinical isolates.¹³ Medical-grade honey gel and 5 silver dressings were applied alone and combined to planktonic cultures of Pseudomonas aeruginosa (PA, strain PA01) or methicillin-resistant Staphylococcus aureus (MRSA, strain SA35556) grown as confluent lawns of planktonic bacterial on agar plates. Plates using cotton gauze dressing without honey or silver were tested as a negative control. Briefly, 300 mL of 1.5% tryptic soy agar (TSA) was autoclaved and allowed to cool to 45ºC-50ºC, then inoculated with 1 mL of a suspension culture of PA01 or SA35556 containing a total of 10⁸ colony forming units (CFUs). The inoculated TSA agar was poured into 24 cm x 24 cm rectangular petri plates and allowed to gel at room temperature. The honey gel was applied as a thin continuous layer to 8 mm disks of oxidized regenerated cellulose/collagen dressing and the disks were placed onto the inoculated TSA agar plates with the honey gel in direct contact with the agar surface. Then, 8 mm disks or regions of application of the 5 silver dressings alone were also directly applied to surface of the inoculated TSA agar plates. To test the combination of honey gel and silver dressings, a thin, continuous layer of the honey gel was applied to 1 side of 8 mm disks of the silver dressings, or applied to the surface of the powder and gel dressings, then the silver disks were placed onto the surface of the inoculated TSA agar plates with the honey gel side in direct contact with agar. Fifty µL of sterile water was then applied to the top surface of each of the dressings and the TSA agar plates were placed in humidified incubators at 37ºC. After 24 hours of exposure, diameters of the zones of killing on the agar plates were measured visually using a millimeter ruler to the nearest 1 mm diameter distance. Triplicate replicates were performed for each test condition.

  Ex vivo pig skin explant model. The authors also tested the combinations in a pig skin explant model, using minor modifications of the previously described procedure, to test the antimicrobial efficacy of these devices in a model more closely mimicking a true wound.¹⁴ Briefly, large sheets of fresh pig skin approximately 20 cm by 30 cm were obtained from a commercial meat processing company and thoroughly cleaned with hair closely trimmed. The subcutaneous fat layer was trimmed to leave approximately 1-2 mm thickness of subcutaneous fat. A large, partial-thickness wound approximately 0.508 mm deep was mechanically created using an electric Paget’s dermatome. Individual explants 12 mm in diameter were punched from the wound area using 12 mm skin biopsy punch. The pig skin explants were sterilized using chlorine gas for 45 minutes then washed 3 times in sterile saline. Seven explants were transferred to 90 mm diameter petri dishes containing 0.5% soft TSA supplemented with antibiotics (50 µg gentamicin per ml for P. aeruginosa or 20 µg doxycycline per mL for S. aureus) to limit overgrowth of bacteria to the bottom of the explants. Sterile explants were inoculated with 50 µL of 107 CFUs of PA01or SA35556, and the explants were incubated for 12 hours at 37ºC. Explants were then transferred to fresh 90 mm petri dishes with 0.5% soft TSA plus antibiotics, again, to limit overgrowth of bacteria to the bottom of the explants. A thin, continuous layer of medical honey gel ~1 mm thick was then applied to the top surface of pig skin explants alone and covered with gauze or with 1 of the 5 silver dressings; 1.5 mL of sterile water was added to the surface of the top dressing. A sterile glass slide (25 mm x 75 mm x 1 mm) was placed on top of the gauze or silver dressings to reduce drying of the dressings and to enhance contact of the honey gel with the surface of the pig skin explants. After 24 hours of incubation, the honey gel and silver dressings were removed, explants were placed in 24 well culture plates, then washed 3 times for 5 minutes with sterile distilled water. Explants were transferred into 5 mL of PBS containing 5 ppm Tween 20, sonicated 5 times for 90 seconds with 1 minute between sonications. Suspensions were vortexed for 10 seconds, triplicate serial dilutions were plated, and CFUs were measured after 24 hours of incubation at 37ºC. Three to five replicate explants were performed for each test condition and the mean was calculated to represent the average CFUs remaining after each treatment condition.

  Ethical considerations. No animals or humans were subjects to the experiments described herein. All of the work described was done in vitro.

Statistical Analysis
All experiments were performed with either 3 to 5 replicates per test condition and all measurements were technically replicated 3 to 5 times per measurement. For the ex vivo microbiology model, the total colony forming units were log₁₀ normalized prior to statistical testing by Student t test. In the event that no bacteria were detected (eg, zero CFUs in nondiluted media), the replicate was treated as possessing only a single bacteria for the purposes of statistical testing since log₁₀ of 0 is undefined. Three different analyses were performed: 1) the antibacterial barrier effects of the silver dressings tested against the gauze control; 2) the antibacterial barrier of the silver dressing tested against the silver + honey combinations; and 3) silver + honey combinations tested against the gauze control. For all tests, a 2-tailed Student t test was used to determine statistical significance with a P-value of ≤ 0.05 chosen as the threshold for statistical significance.

Honey and silver sugars. In all tests, the concentration of sugars increased, with an average increase of 31.9% for the total sugars assay and 25.1% for the glucose-only assay (Figure 1A and 1B). The observed increase was statistically significant for all samples except the change in glucose concentration for the alginate dressing (Figure 1B, P = 0.29). None of the dressings were anticipated to contribute any sugars based on the knowledge of their compositions, and the empirical data provided by the saline extracted controls confirmed these expectations (Figure 1A and 1B).

  Honey and silver osmotic strength. The diluted honey gel stocks’ osmolarity ranged from 753 to 1005 mOsmol/kg prior to incubation. After incubation, the total osmolarity ranged from 909 to 3169 mOsmol/kg. In all instances, the total osmolarity of the diluted honey gel increased (range of 10.9% - 240%) after being incubated with 1 of the silver-containing dressings when compared to the diluted honey stock prior to incubation (Figure 1C).

  Honey and silver antimicrobial activity: Agar plate zone of inhibition assay. Treatment with the honey gel alone (covered by moist gauze) produced a clear zone of inhibition 14 mm in diameter for P. aeruginosa (Figure 2B) and 15 mm diameter for S. aureus (Figure 2B). Each of the silver dressings alone (covered by a glass slide) also produced large zones of inhibition for both bacteria, with slightly larger diameters for P. aeruginosa than for S. aureus. When the honey gel was combined with each of the 3 silver dressings, the diameters of the zones of inhibition were larger than the diameters of the zones of inhibition for the silver dressings alone or for the honey gel alone. The negative control dressing did not generate any zone of inhibition for either P. aeruginosa or S. aureus. Given the low spatial resolution used herein (1 mm), there was no observed variation for some of the tested conditions precluding the ability to perform statistical testing.

  Pig skin explant model. After 24 hours, the explants presented with ~1 x 108 CFUs of total bacteria for each germ when they were individually incubated with saline-moistened gauze. In all instances the log₁₀-normalized bacterial counts of both germs were significantly reduced (Figure 3, black bars vs control, P ≤ 0.002).

  In the presence of honey, the S. aureus count was further decreased beyond the effects seen with the silver dressing alone for all dressings tested (Figure 3A, black vs grey bars, P ≤ 0.01). This observed pattern was not apparent in the P. aeruginosa innoculated explants. For 2 of the tested dressings—the soluble glass power and the collagen dressing—the observed bacterial counts were higher with the addition of honey (Figure 3B), but the difference was only significant for the collagen dressing (P ≤ 0.02). Of the other 3 dressings, the observed additional reduction in bacteria was only significant for the foam dressing (P ≤ 0.0002). Overall, these data indicate that honey can interfere with at least 1 dressing, the collagen dressing. However, even with this interference, the observed reduction of both bacteria for all honey and silver combinations tested were statistically significant compared to the gauze control (Figure 3B, control vs grey bars, P ≤ 0.0003). The P-values for all of the tested conditions can be found in Table 1.

Given that the PBS extracted controls did not contribute to the generation of a detectable signal in the sugar assay and the increase in concentration was also apparent in the absence of the isomerization step, it appears that the dressings removed water from the diluted honey through nonspecific hydrogen bonding effects akin to a dialysis-like effect. Additional support for this type of mechanism is the osmotic strength data, which demonstrated an across-the-board increase in total osmolarity. In retrospect, overall, this finding is not counterintuitive, as the use of additional, nonhydrated dressings are expected to complement the osmotic mechanism of medical honey in that they provide additional capacity for wound fluid absorption. In future studies, the concept of total osmotic capacity may enable one to predict how a non-honey dressing paired with a honey-based dressing would contribute to, or interfere with, honey dressings, based on the level of hydration and hydration limit of the non-honey dressing.

  All of the silver-containing dressings were able to create a zone of inhibition. When honey was added to the dressings, the zone of inhibition was not decreased in any case, on the contrary, the combination of silver and honey consistently created a larger zone of inhibition than either dressing alone (Figure 2). While statistical testing was precluded by a lack of variation in some samples, the zone of inhibition test is usually interpreted qualitatively and all dressings and combinations still possessed an antimicrobial barrier effect.

  From a purely mechanistic perspective, combining silver-containing dressings with medical honey gel does not decrease honey’s osmotic mechanism, but instead, the typically dry dressings actually contribute some osmotic capacity. When the antimicrobial efficacy of the combined dressings was tested in vitro, the honey possessed obvious antimicrobial activity. In the zone of inhibition assay, the combined silver and honey treatment reduced the bacteria to levels below that of the honey treatment alone indicating that the silver dressing’s antimicrobial activity was not impaired by the presence of honey. In the ex vivo pig skin model, a much stronger contrast of effects were seen, where the type of dressing possessing the silver and the type of bacteria treated revealed differences amongst the tested dressings. However, in all but the case of the collagen-based dressing, the presence of honey did not lead to less antimicrobial activity than the silver dressing alone. The authors conclude that for the purposes of autolytic debridement, the evidence in this study suggests that silver dressings will not interfere with medical honey’s osmotic mechanism, and that the combination will still possess at least the same antimicrobial activity of a non-collagen silver dressing if used.

Affiliation: The authors are from the University of Florida, Department of OB/GYN, The Institute for Wound Research, Gainesville, FL.

Address correspondence to:
Daniel J. Gibson, PhD
1600 SW Archer Road M323C
Gainesville, FL 32610
gibsondj@ufl.edu

Disclosure: The authors disclose that Medline Industries, Inc, Mundelein, IL, provided financial and material support for the work described herein. In addition, Daniel J. Gibson, Qingping Yang, and Gregory S. Schultz disclose the receipt of consultant fees and research grants, research and travel grants, and speaker and consultant fees, respectively.