Authors: Zhiyuan Li & Rohan Parikh
Peer Reviewer: Janice Rateshwar
Professional Reviewer: Zachary Minchow-Proffit
Introduction: Chronic wounds and ulcers disproportionately affect the world’s diabetic, aging, and socioeconomically-disadvantaged populations, while antibiotic-heavy treatments contribute to global antimicrobial resistance. One of the most promising solutions to this issue is honey. Studies show that when antibiotics fail, topical application of honey supports healthy cell growth and fights infections, allowing chronic wounds to fully heal.
Natural honeys contain glucose oxidase (GOX), an enzyme that contributes to honey’s antibacterial activity through hydrogen peroxide (H2O2) synthesis. However, natural honeys are difficult to sterilize, can be allergens, and vary in antibacterial potency across floral sources and seasons.
Purpose: This research aims to integrate GOX activity enhancement with gelling hydrocolloids to create a synthetic honey that mitigates natural honey’s issues.
Methodology: Antibacterial potency against E. coli and B. cereus was evaluated for all honey treatments. Synthetic treatments were also tested for H2O2 accumulation using a GOX activity assay.
Results: The synthesized treatments were statistically comparable to or better than penicillin (p > 0.05) against both bacteria. Modified blueberry honey did not have significantly greater H2O2 accumulation than its semi-synthetic counterpart, showing that pectin can stabilize GOX without impacting its activity. Cost analysis revealed that the synthetic treatment is less than a tenth of the cost of its commercial alternative.
This research established a novel and reproducible method for developing a sustainable synthetic honey with the physical and chemical properties of natural honey. Combining GOX stabilization and antibacterial activity maximization with hydrocolloids enables the emergence of a low-cost wound-healing agent with immense potential to revolutionize treatment for infections and chronic wounds.
One of the largest global health issues is chronic wounds, which affect the world’s aging and increasingly diabetic population. These non-healing wounds require expensive and extensive treatment heavily reliant on antibiotics, putting low-income communities at a disadvantage and contributing to global antimicrobial resistance (Situm et al., 2014). Honey presents itself as a potentially effective alternative treatment due to its unique composition. Its high sugar content contributes to its high osmolarity, which both promotes healing by increasing the outflow of lymph in wounds, and contributes to antimicrobial activity of honey, as the sugars in honey can draw water out of bacterial cells. Furthermore, its pH of 3.2-4.5 makes for an acidic environment that supports tissue regeneration and inhibits bacterial growth; honey’s diverse flavonoids and phenolic compounds exhibit antioxidant, antibacterial, and anti-inflammatory effects; honey’s natural viscosity creates a moist wound-healing condition. These physical and chemical properties of honey are characteristic of an ideal wound-healing agent that can support the growth of healthy cells while fighting infections (M. Mandal & S. Mandal, 2011; Molan & Rhodes, 2015).
While honey’s acidity, high osmolarity, phytochemicals and phenolic compounds contribute to honey’s multifaceted antimicrobial activity, the majority of natural honeys are antibacterial due to the antimicrobial peptide bee defensin-1, and hydrogen peroxide synthesized by the enzyme glucose oxidase (GOX) from glucose, water, and oxygen (M. Mandal & S. Mandal, 2011). These components in honey make it potent against dermatologically relevant and multi-drug resistant bacteria, and have been found to reverse antimicrobial resistance and reduce microbial pathogenicity (Wasihun & Kasa, 2016; McLoone et al., 2016). However, the composition of honey is dependent on many factors—including floral sources, geographical distribution, and season—and thus the accumulation of hydrogen peroxide in most natural honeys is not concentrated enough for them to be effectively antibacterial (M. Mandal & S. Mandal, 2011). In contrast, current medical-grade honeys accumulate much more hydrogen peroxide, but are expensive—Revamil honey, a current option that has a maximum H2O2 accumulation of 4.4 mM, is 25 USD for 15 grams (Kwakman et al., 2011).
In previous research, we enhanced the activity of GOX in natural blueberry honey, developing a medical-grade honey with three times the accumulation of hydrogen peroxide as current commercial medical-grade honeys for only one-tenth of the cost (Li & Parikh, 2018). However, the honey developed was sub-optimal for wound healing applications for multiple reasons. The filtration method previously used–while protecting GOX from heat degradation–required that the honey be diluted. This compromises the viscosity of the honey, while the method itself was insufficient in completely sterilizing honey, even with 0.22 µm micron-filters (Li & Parikh, 2018). Further, despite finding that natural blueberry honey had great potential for maximizing H2O2 accumulation, blueberry honey is not globally available, nor is its composition consistent from season to season. Using natural honey therefore creates intrinsic limitations of scope on the derived medical-grade honeys. This continuation study thus developed a synthetic honey that mimics physical and chemical properties of raw honey to create an antimicrobial wound-healing agent that addresses the previously encountered issues of attaining sterility, maintaining viscosity, mitigating batch-to-batch variability, and maximizing global availability.
Extracted proteins and antioxidants were added to a hydrocolloid base that created the properties of acidity, antimicrobial activity, viscosity, and antioxidant activity. The synthetic honey and controls were tested for bacterial growth inhibition of Bacillus cereus ATCC 14579 (gram-positive) and Escherichia coli K-12 (gram-negative) bacteria. The raw and synthetic honeys were then characterized and protein concentration was adjusted for the synthetic variety to be comparable to current medical-grade and natural honeys—further testing of the adjusted synthetic honey composition on bacteria completed the iterative design process of a honey-inspired antimicrobial wound-healing agent.
Methods and Materials
UV Sterilization: Previous research showed that filtration of honey solution with 0.22 micron filters was insufficient to fully sterilize honey. To investigate whether UV-treatment is a viable alternative, raw blueberry honey solution was diluted in a 1:1 honey to diH2O ratio, then exposed to UV radiation for 5 minutes. The honey was then used for enzyme activity testing with a fluorescence assay to determine if the UV treatment would impact GOX activity. The resulting H2O2 accumulation was compared to that of raw, non-UV treated blueberry honey and UV-treated blueberry honey with EGCG added.
Protein Extraction: The protein extraction procedure was performed by spinning the diluted blueberry honey solution through the 100 kDa Centricon® filter (Millipore Sigma; Product #: UFC710008). The supernatant was collected and spun through the 3 kDa Centricon® filter (Millipore Sigma; Product #: UFC700308). The resulting protein concentrate was collected, aliquoted, and stored at -80°C until synthetic honey was needed. In addition to capturing the antimicrobial components of natural honey—GOX and bee defensin-1 with molecular weights of ~80 kDa and ~5.5 kDa, respectively—the protein extraction procedure also served as a sterilizing mechanism, as microorganisms are typically greater than 4,000 kDa in size (Baron, 1996).
Experimental Groups: In addition to synthetic honey, five other treatments were created to act as controls and investigate possible synergistic effects. Raw blueberry honey (RAW BB) was made by diluting honey in a 1 gram of honey:1 mL of distilled water ratio, then treating with UV light for 5 minutes. This acted as a baseline for all treatments that contained honey or honey proteins. Its modified version (MOD BB) featured EGCG (Sigma Aldrich; Product #: E4268), a tea catechin and antioxidant previously found to maximize the hydrogen peroxide accumulation in natural blueberry honey. EGCG was added to the honey solution to a final concentration of 400 µM before treating with UV-light and incubating for 48 hours.
Semi-synthetic honey (SEMI) was made by combining the modified blueberry honey with 1% pectin (CPKelco GENU® pectin type LM-104; Product #: 120012) to create a honey solution which maintained the viscosity of honey, which raw modified honey lost due to dilution with water. 1% Pectin (PEC) was tested on its own as a control for subsequent pectin-containing solutions, as pectin has been shown to have antimicrobial properties on its own.
Extracts and antioxidants (EXOX) was meant to be a synthetic honey without the addition of hydrocolloids. The EXOX treatment was created by adding protein extract to a final concentration of 1.5%, EGCG to a concentration of 400 µM, and D-glucose (Sigma-Aldrich; Product #: G8270) to a concentration of 30% to a volume of distilled water. Concentrations were derived from previous literature on the composition of honey, although the protein concentration was subject to later change after characterization testing. Synthetic honey (SYNTH) was the same as extracts and antioxidants, only with a hydrocolloid base rather than using water as the solvent. The hydrocolloid pectin-gellan base was created by heating water, then sifting in pectin to a final concentration of 1%, and gellan gum (CPKelco KELCOGEL® CG-LA; Product #: 20005621) to a final concentration of 0.5%.
By adding only glucose to EXOX and SYNTH, we were able to provide the enzyme with its required sugars without having other sugars that are typically present in raw honey, which can feed bacteria and infected cells. Additionally, the innate sterility of the honey proteins and the heating of the pectin-gellan base were able to kill any microbes prior to adding the extracted peptides. Lastly, two controls, a blank disk negative control (Carolina Biological Supply Company; Product #: 806491) and a positive penicillin disk control (Carolina Biological Supply Company; Product #: 806496) were also included as treatments.
Bacterial Inhibition Trials: Bacterial inhibition trials were done as double-blind disk diffusion assay to evaluate antimicrobial potency and reduce bias. The assay was conducted by first streaking a bacterial culture on nutrient agar plates. Blank sterile disks were then saturated with treatments and placed on the plate with the positive and negative controls using a randomization procedure that changed the order of the negative control, positive control, and treatments on any given plate. In the pre-optimization bacterial trials, the sequence in which various treatments were plated and the position of the plates in the incubator were also randomized. Thus, measurement of zones of inhibition was done without knowledge of what treatment was being measured, reducing possible bias.
The plates were incubated for 48 hours to allow treatments to diffuse through the agar and for the bacterial culture to fully grow. Its growth was theoretically inhibited where the treatment diffused, with the radius of the zone of inhibition indicating the treatment’s antibacterial potency. Measurements for the zones of inhibition were normalized to values
rs is the radius of the sample’s zone, rp is the radius of the positive control, and rn is the radius of the negative control. Normalization accounted for unforeseen differences across plates in any given set of trials.
GOX Activity Assay: To verify the presence of GOX in our extracted proteins and confirm the antibacterial activity of the synthetic honey was due to the enzyme, a fluorescence assay to quantify expressed GOX activity by H2O2 accumulation was run on a Gen 5 Synergy HT microplate reader (Bio-Tek; Part #: 70910000) as per assay specifications of the Amplex Red Glucose/Glucose Oxidase Assay Kit (Thermofisher Scientific; Product #: A22189). One change that was made to the assay directions was that the solution that was put in the well plate was 50 µL, but broken down as 25 µL honey solution and 25 µL 1X reaction buffer. A known concentration of H2O2 was also assayed in addition to testing GOX activity of treatments to help convert fluorescence measures into concentrations of H2O2 accumulated.
Bee Defensin ELISA: An enzyme-linked immunosorbent assay (ELISA) was used to quantify and compare the amount of defensin-1 in a known concentration of extracted protein and in raw honey (Genway Biotech; Product #: GWB-SKR002). Informed by the results of the GOX activity assay, the concentration of protein in the synthetic honey was increased from 1.5% to 5% for the ELISA. A standard curve was used to extrapolate the concentration needed to match the total protein content between synthetic and raw honeys.
Optimized Bacterial Inhibition Trials: For our optimized inhibition trails, the initial bacterial trials were modified in three distinct ways. First, the experimental groups that acted as controls for more effective treatments were eliminated, and thus the sole focus was on the treatments we wanted to compare our synthetic honey against. Second, the protein concentration of synthetic honey was increased to 5%, based on the results of the ELISA and GOX activity assay, to match the total protein content of raw honey. Third, the plating procedure was changed to include all three treatments (RAW, SEMI, & SYNTH) on any given plate, along with a negative control and positive control. The order of the treatments and controls were again randomized within a given trial of 18 plates; each trial was performed three times.
Data and Results
Effect of UV light on GOX activity: As shown in Figure 1, treating the blueberry honey with UV light for 5 minutes was not significantly different in the expressed glucose oxidase activity of the blueberry honey compared to the untreated (n = 9; Kruskal Wallis and Dunn’s Multiple Comparison; p > 0.05). Contrary to results from previous research, the baseline activity of blueberry honey this year was already higher than that of Revamil honey in current literature (~4.4 mM H2O2), while EGCG actually reduced the H2O2 accumulation in the honey, thus showing the inherent variability of natural honeys.
Preliminary Bacterial Trials: A Kruskal-Wallis and Dunn’s Multiple Comparisons test was used to determine significance (n = 18). As shown in Figures 2 and 3, all experimental groups had greater inhibitory effects on E. coli (gram-negative) than B. cereus (gram-positive), which agrees with prior research. For E. coli, there was no significant difference between pectin and negative controls, while all other groups were significantly better than the negative controls (p < 0.01) (Figure 2). The remaining treatments and positive control had no significant differences amongst them (p > 0.05). Pectin-containing solutions exhibited statistically significant inhibitory effects on B. cereus (p < 0.05) (Figure 3). While penicillin generally had greater zones of inhibition than the other treatments, the differences were not statistically significant (p > 0.05) with the exception of EXOX. All other pairwise comparisons had no significant differences (p > 0.05).
Protein Characterization: Modified blueberry honey did not have significantly greater GOX activity than the semi-synthetic honey, showing that pectin makes little difference on expressed GOX activity (n = 9; Kruskal Wallis and Dunn’s Multiple Comparison; p > 0.05). Further activity testing shows that the synthetic varieties accumulated H2O2, indicating a presence of GOX, which presumably is the source of antibacterial activity. The EXOX and SYNTH groups have significantly lower expressed GOX activity than MOD and RAW groups (p < 0.05), indicating that the SYNTH treatment has lower concentrations of GOX than the raw honey (Figure 4). The ELISA confirmed the presence of defensin-1 in the extracted protein solution, and showed that despite variability in the assay measurements, synthetic honey with 5% protein concentration had greater concentration of bee-defensin-1 than the raw honey. This result, shown in Table 1, was used to inform the protein composition of the next iteration of the modified synthetic honey that we tested.
Optimized Bacterial Trials: In the optimized bacterial trials for E. coli, shown in Figure 5, there was no significant difference between the antibacterial activity of raw blueberry honey, semi-synthetic honey, and the negative control (p > 0.05). In contrast, the synthetic honey was significantly more effective than the negative control, raw honey, and semi-synthetic honey (p < 0.001) and comparable to penicillin (p > 0.05). For the optimized trials on B. cereus, as shown in Figure 6, there was no significant difference between the raw and semi-synthetic honey (p > 0.05). Both raw and semi-synthetic exhibited significantly greater activity than the negative control (p < 0.001) and significantly less activity than the positive penicillin control (p < 0.001). The second-iteration synthetic honey’s effectiveness against B. cereus was much closer to that of penicillin, with no significant difference between the antibacterial activity of the two (p > 0.05).
This research established a novel synthetic honey with the physical and chemical properties of natural honey—the first of its kind in published literature. Our synthetic honey addresses issues with sterilization in natural honeys—individual components can be either heat-sterilized or filtered extensively before combination without jeopardizing GOX stability. Thus, full sterilization of synthetic honey is feasible. Qualitatively, in raw and semi-synthetic solutions, plates often exhibited microbial contamination from the solutions themselves. In contrast, the synthetic honey was never contaminated, indicating that it is much more feasible to sterilize and keep sterile on infections and wounds as a dressing for long periods of time.
In addition, modified synthetic honey may be more effective than penicillin. Synthetic honey was significantly more effective than its honey-containing counterparts and comparable to penicillin in inhibiting bacterial growth. Literature has shown that honey tends to display more potent antimicrobial activity against gram-positive bacteria like E. coli rather than gram-negative bacteria, such as B. cereus (M. Mandal & S. Mandal, 2011). While this was observed in the pre-optimization bacterial trials, optimized synthetic honey showed equal effectiveness on gram-positive bacteria and gram-negative bacteria. Anecdotally, it was observed that within 48 hours, the zones of inhibition for the penicillin disks were shrinking, suggesting that the penicillin treatment was becoming less effective over time—this did not appear to occur for the honey treatments.
Synthetic honey is also much more cost-effective than commercial medical-grade honey. A preliminary cost analysis, which did not account for aspects of commercialization such as overhead costs and profit margins, showed that the synthetic honey developed is 8.00% of the cost of Revamil. Finally, the composition of synthetic honey can be readily reproduced or adjusted to allow for efficient optimization, as was demonstrated in this research’s iterative design process. The synthetic honey’s effectiveness is thus not subject to seasonal changes, geographical differences, and floral composition, all of which are factors that increase the variation in the composition of natural honey.
Future work concerning this research includes obtaining a patent, conducting in-vivo testing, and testing synthetic honey on bacteria present on the skin. In conclusion, GOX stabilization and antibacterial activity maximization through the use of hydrocolloid hydrogels enable the emergence of a low-cost, sustainable, and effective wound-healing agent with immense potential to revolutionize treatment for infections and chronic wounds.