Antibacterial Activity of Different Blossom Honeys: New Findings

Antibacterial activity is the most investigated biological property of honey. The goal of this study was to evaluate the antibacterial activity of 57 Slovak blossom honeys against Staphylococcus aureus and Pseudomonas aeruginosa and investigate the role of several bioactive substances in antibacterial action of honeys. Inhibitory and bactericidal activities of honeys were studied to determine the minimum inhibitory and bactericidal concentrations. The contents of glucose oxidase (GOX) enzyme, hydrogen peroxide (H2O2), and total polyphenols (TP) were determined in honeys. We found that honey samples showed different antibacterial efficacy against the tested bacteria as follows: wildflower honeys > acacia honeys > rapeseed honeys. Overall antibacterial activity of the honeys was statistically-significantly correlated with the contents of H2O2 and TP in honeys. A strong correlation was found between the H2O2 and TP content. On the other hand, no correlation was found between the content of GOX and level of H2O2. Antibacterial activity of 12 selected honeys was markedly reduced by treatment with catalase, but it remained relatively stable after inactivation of GOX with proteinase-K digestion. Obtained results suggest that the antibacterial activity of blossom honeys is mainly mediated by H2O2 levels present in honeys which are affected mainly by polyphenolic substances and not directly by GOX content.


Introduction
Honey, a bee product, offers a broad-spectrum of biological properties including antimicrobial and wound healing activity [1]. Importantly, antibacterial and anti-biofilm potentials of honey have been considered as an exclusive criterion of its wound healing properties. Therefore, the antibacterial activity of this matrix from different floral sources has been intensively studied over the past few decades. It has been found that the botanical origin of honey influences it, due to its content of

Acacia Honeys
Melissopalynological analysis of acacia samples ( Figure 1A) documented the presence of R. pseudoacacia pollen with a frequency of 15-45% in all samples, as expected for a hypo-represented monofloral honey. Moreover, in these samples, Rubus sp. and Brassicaceae pollens were also detected with a frequency of <3% in 62.5 and 12.5% of the honeys, respectively; a frequency between 3 and 15% in 25% of the samples (for both pollen types); a frequency between 15 and 45% in 12.5 and 50.0% of the samples; and for only Brassicaceae, a frequency of >45% in 12.5% of the honeys. Nectariferous taxa with early flowering, for example Salix sp. and Prunus sp. (respectively present in 75.0 and 62.5% of the samples with variable percentage), Acer sp. and fruit bearing species (e.g., Malus/Pyrus sp.) were identifiable in this class of honeys. In particular, pollen granules, typically associated to Alpine area, like Vicia sp. and Loranthus sp., were also found in R. pseudoacacia honeys, as documented by Persano Oddo et al. [19].

Rapeseed Honeys
These samples ( Figure 1C) showed the presence of Brassicaceae pollen with a frequency of always >45% as expected, since it is dominant and hyper-represented in this plant species. Consequently, these rapeseed honeys appeared poor of other taxa, which could be detected only in traces with a frequency of <20%. We observed the presence of early flowering plants, such as Myosotis sp., (in 12.5% of samples with a percentage of <20%), Salix sp. (in 25% of the honeys with a frequency between 3-15%, and in 75% with <3%), Prunus sp. (detected in 50% of the samples with a frequency between 3-15%, and in 37.5% with <3%), Rubus sp. (found in 25% of the samples with a frequency between 3-15%, and in 62.5% with <3%), Malus/Pyrus sp. (present in 12.5% of the samples with a frequency between 3-15% and in 50% with a frequency <3%) and Acer sp. (present in 87.5% of the samples with a frequency <3%).

Raspberry and Phacelia Honeys
Rubus idaeus and Phacelia sp. pollen are present in the respective samples with a frequency of >45%, coherently with the definition of monofloral honeys ( Figure 1D). Rubus idaeus and Phacelia sp. are plant species typical of the Eastern and Northern Europe [19]. Characteristic spring-like taxa were also found: Trifolium repens (in the samples with a mean frequency <15%), Myosotis sp. (in raspberry honeys with a frequency between 3-15%), Vicia sp. and Prunus sp. (both present in the samples with a frequency of <3%).

Antibacterial Activity of Blossom Honey Samples
The minimal inhibitory concentration (MIC) and minimal bactericidal concentration (MBC) values of 57 blossom honey samples against two tested bacterial species are shown in Figure 2. For S. aureus, the MIC values of the honey samples ranged from 6.0 to 45.0% (concentrations of honeys in solution), whereas the MICs of honeys against P. aeruginosa ranged from 8.0 to 40.0%. Wildflower honeys (W16-W47), together with raspberry (Ra48) and phacelia (P49) honeys, exhibited the strongest antibacterial activities giving MIC average values of 13.9 and 15.0% against S. aureus and P. aeruginosa, respectively. On the other hand, acacia (A1-A15) and rapeseed honeys (R50-R57) exhibited weaker antibacterial activities, giving an average MIC values between 18.7-19.1% versus of 21.2-22.5% against S. aureus and P. aeruginosa, respectively ( Figure 3). Overall, wildflower honeys showed statistically-significantly higher antibacterial efficacy in comparison to acacia and rapeseed honeys ( Figure 3).
All honey types contained samples showing low antibacterial activity (MIC values lower than 20%). The MIC values of three honey samples (W19, W42, and R56) were above 30%, suggesting that these samples were most likely overheated or stored under unsuitable conditions. Six samples (W16, W22, W26, W27, W43, and W44) were the most effective against both tested bacteria, giving average values below 10%.
The profile of MBC values was similar to that of MIC values ( Figure 2). In particular, MBC values were either identical or slightly higher (one dilution above) than MIC values.

Antibacterial Activity of Blossom Honey Samples
The minimal inhibitory concentration (MIC) and minimal bactericidal concentration (MBC) values of 57 blossom honey samples against two tested bacterial species are shown in Figure 2. For S. aureus, the MIC values of the honey samples ranged from 6.0 to 45.0% (concentrations of honeys in solution), whereas the MICs of honeys against P. aeruginosa ranged from 8.0 to 40.0%. Wildflower honeys (W16-W47), together with raspberry (Ra48) and phacelia (P49) honeys, exhibited the strongest antibacterial activities giving MIC average values of 13.9 and 15.0% against S. aureus and P. aeruginosa, respectively. On the other hand, acacia (A1-A15) and rapeseed honeys (R50-R57) exhibited weaker antibacterial activities, giving an average MIC values between 18.7-19.1% versus of 21.2-22.5% against S. aureus and P. aeruginosa, respectively ( Figure 3). Overall, wildflower honeys showed statistically-significantly higher antibacterial efficacy in comparison to acacia and rapeseed honeys ( Figure 3).
All honey types contained samples showing low antibacterial activity (MIC values lower than 20%). The MIC values of three honey samples (W19, W42, and R56) were above 30%, suggesting that these samples were most likely overheated or stored under unsuitable conditions. Six samples (W16, W22, W26, W27, W43, and W44) were the most effective against both tested bacteria, giving average values below 10%.
The profile of MBC values was similar to that of MIC values ( Figure 2). In particular, MBC values were either identical or slightly higher (one dilution above) than MIC values. Manuka honey exhibited comparable or stronger antibacterial activity than blossom Slovak honey samples. Manuka honey was, similarly to blossom honeys, more effective against S. aureus than P. aeruginosa.
Molecules 2019, 24, x 5 of 20 Manuka honey exhibited comparable or stronger antibacterial activity than blossom Slovak honey samples. Manuka honey was, similarly to blossom honeys, more effective against S. aureus than P. aeruginosa.   Activity was determined with a minimum inhibitory concentration (MIC). Red line represents the mean of all measured data. Differences among groups of samples belonging to different honey types were analyzed by ANOVA. Asterisks indicate a significant difference between the examined honey types, * P < 0.05, ** P < 0.01, *** P < 0.001, NSnon-significant.

GOX Content and Antibacterial Activity of Honey Samples
The method for quantification GOX in honeys by the wet immunoblot assay based on prepared recombinant GOX (rGOX) and rabbit polyclonal anti-honeybee GOX antibody was developed. Both specificity and affinity of the antibody used was previously determined [20]. No cross-reaction with other honey proteins separated by SDS-PAGE could be detected. Prepared rGOX in denaturing solution (4 M urea, 20 mM Tris pH 8.0) had 70% purity (determined by SDS-PAGE). A strong linear relationship (r = 0.984) between rGOX amount and band volume intensity was observed in the calibration curve created for GOX quantification in honeys ( Figure 4). Activity was determined with a minimum inhibitory concentration (MIC). Red line represents the mean of all measured data. Differences among groups of samples belonging to different honey types were analyzed by ANOVA. Asterisks indicate a significant difference between the examined honey types, * P < 0.05, ** P < 0.01, *** P < 0.001, NS-non-significant.

GOX Content and Antibacterial Activity of Honey Samples
The method for quantification GOX in honeys by the wet immunoblot assay based on prepared recombinant GOX (rGOX) and rabbit polyclonal anti-honeybee GOX antibody was developed. Both specificity and affinity of the antibody used was previously determined [20]. No cross-reaction with other honey proteins separated by SDS-PAGE could be detected. Prepared rGOX in denaturing solution (4 M urea, 20 mM Tris pH 8.0) had 70% purity (determined by SDS-PAGE). A strong linear relationship (r = 0.984) between rGOX amount and band volume intensity was observed in the calibration curve created for GOX quantification in honeys ( Figure 4).
The GOX content in analyzed blossom honey samples, determined by the established assay, ranged from 5.0 to 56.4 µg/g with an average value of 29.7 µg/g. Acacia honey samples (A1-A15) contained the highest amount of GOX, with an average value of 33.7 µg/g of honey ( Figure 5A). The lowest content, with average value of 24.3 µg/g of honey, was determined in rapeseed samples.
No correlation was found between GOX content and MIC values ( Figure 2) for both tested bacteria, among the blossom honey samples (r = 0.022, P = 0.87 for S. aureus and r = 0.001, P = 0.99 for P. aeruginosa). This observation suggested that the level of GOX in honey was not a critical factor for overall honey antibacterial activity. The GOX content in analyzed blossom honey samples, determined by the established assay, ranged from 5.0 to 56.4 μg/g with an average value of 29.7 μg/g. Acacia honey samples (A1-A15) contained the highest amount of GOX, with an average value of 33.7 μg/g of honey ( Figure 5A). The lowest content, with average value of 24.3 μg/g of honey, was determined in rapeseed samples.
No correlation was found between GOX content and MIC values ( Figure 2) for both tested bacteria, among the blossom honey samples (r = 0.022, P = 0.87 for S. aureus and r = 0.001, P = 0.99 for P. aeruginosa). This observation suggested that the level of GOX in honey was not a critical factor for overall honey antibacterial activity.

H2O2 and Total Polyphenol Contents in Honey Samples
The content of H2O2 in blossom honey samples was determined in 40% honey solutions (pH 7.0) immediately after preparation of homogenous solutions. The content ranged from 2.4 to 47.2 μg/g of Figure 5. Glucose oxidase (GOX), hydrogen peroxide (H 2 O 2 ) and total polyphenol contents (TP) in blossom honey samples (n = 57). (A) GOX content was determined using a semi-quantitative immunoblot assay (B) H 2 O 2 content was measured in diluted honey samples (40% (w/w), without previous incubation) with a modified GOX assay kit (C) TP content was determined with a Folin Ciocalteu Phenolic Content Quantification Assay Kit in 20% (w/v) honey solutions. Gallic acid (GAE) was used as the TP reference standard compound and results are expressed as GAE equivalents (µg GAE/g of honey). The data represent the mean values with standard deviation (SD).

H 2 O 2 and Total Polyphenol Contents in Honey Samples
The content of H 2 O 2 in blossom honey samples was determined in 40% honey solutions (pH 7.0) immediately after preparation of homogenous solutions. The content ranged from 2.4 to 47.2 µg/g of honey ( Figure 5B). The average value of H 2 O 2 content was 13.5 µg/g of honey. Polyfloral honeys (W16-W47) showed the greatest amount of H 2 O 2 , with an average value of 18.1 µg/g of honey with respect to acacia and rapeseed honeys which had average values of 6.8 and 7.8 µg/g of honey, respectively. Correlation test showed no significant correlation between the values of H 2 O 2 and GOX content among the honey samples (r = −0.087, P = 0.521). These results suggest that there is no direct relationship between the level of GOX and the level of H 2 O 2 in blossom honeys.
One of the important groups of biologically active molecules in honey are polyphenols/flavonoids. It is considered that these molecules can elevate the levels of H 2 O 2 and thus improve the antibacterial activity of honey. TP content in blossom honey samples was determined in the 20% (w/v) honey solutions and it is shown in Figure 5C. The content in honey samples ranged from 91 to 452 µg GAE/g of honey with an average value of 228 µg GAE/g of honey. Acacia honey samples (A1-A15) showed the lowest content, with an average value of 126 µg GAE/g of honey ( Figure 5C). The highest content, with an average value of 274 µg GAE/g of honey, was determined, as expected, in wildflowers honey samples (W16-W47). High TP contents were also found in raspberry and phacelia honeys.
In contrast to results from statistical analysis between the GOX and H 2 O 2 contents, a strong statistically-significant correlation (r = 0.66, P < 0.0001) between the TP and H 2 O 2 content was found in blossom honey samples.

Role of Honey Proteins and H 2 O 2 in the Antibacterial Effect of Blossom Honeys
Our recent findings showed that bee-derived protein components, including GOX and Def-1, are not responsible for the pronounced antibacterial activity of the honeydews honey [10]. Therefore, we investigated the contribution of the proteins in selected blossom honeys showing the strong antibacterial activity (four acacia (A1, A2, A8, and A9), four wildflower (W21, W22, W26, and W27), and four rapeseed (R52, R53, R55, and R57) honeys) to the overall honey antibacterial effect. Upon proteinase K treatment of 50% (w/w) honey solutions, diluted with Mueller-Hinton broth (MHB), proteins including GOX were digested ( Figure 6). This digestion of GOX in honey solution resulted in a loss of its enzymatic activity. We indirectly verified the loss of activity using a partially purified native GOX from honey. Its enzymatic activity was completely abolished with proteinase-K treatment performed in the same conditions as in the case of honey solution (data not shown). and four rapeseed (R52, R53, R55, and R57) honeys) to the overall honey antibacterial effect. Upon proteinase K treatment of 50% (w/w) honey solutions, diluted with Mueller-Hinton broth (MHB), proteins including GOX were digested ( Figure 6). This digestion of GOX in honey solution resulted in a loss of its enzymatic activity. We indirectly verified the loss of activity using a partially purified native GOX from honey. Its enzymatic activity was completely abolished with proteinase-K treatment performed in the same conditions as in the case of honey solution (data not shown). After confirming the proteinase K inactivation effect on GOX, antibacterial activities of proteinase K-untreated and -treated honey samples were studied. Two kinds of untreated samples were used; non-incubated and 2-h-incubated (catalase treatment time) at room temperature after honey dilution in MHB. Differences in antibacterial activity were observed between non-incubated and 2-h-incubated untreated honeys depending on certain types of honey. In all acacia honey samples, the 2-h incubation caused increases of MIC values in comparison with non-incubated samples. In contrast, 2-h incubation led to decrease of MIC values in all wildflower and most rapeseed honeys. The changes were most pronounced in the case of S. aureus than P. aeruginosa (Figure 7). This finding suggests a significant catalase action and/or weaker H2O2 production capacity in diluted acacia honeys (having high GOX and low TP content) than it was in other two types of honey. No changes (S. aureus) or weak increasing of MIC values (P. aeruginosa, except A1 honey sample) were observed between the untreated 2-h-incubated and proteinase K-treated honey samples. The last results suggest that action of protein components GOX and Def-1 in diluted honeys do not markedly contribute to the antibacterial activity of the blossom honeys. After confirming the proteinase K inactivation effect on GOX, antibacterial activities of proteinase K-untreated and -treated honey samples were studied. Two kinds of untreated samples were used; non-incubated and 2-h-incubated (catalase treatment time) at room temperature after honey dilution in MHB. Differences in antibacterial activity were observed between non-incubated and 2-h-incubated untreated honeys depending on certain types of honey. In all acacia honey samples, the 2-h incubation caused increases of MIC values in comparison with non-incubated samples. In contrast, 2-h incubation led to decrease of MIC values in all wildflower and most rapeseed honeys. The changes were most pronounced in the case of S. aureus than P. aeruginosa (Figure 7). This finding suggests a significant catalase action and/or weaker H 2 O 2 production capacity in diluted acacia honeys (having high GOX and low TP content) than it was in other two types of honey. No changes (S. aureus) or weak increasing of MIC values (P. aeruginosa, except A1 honey sample) were observed between the untreated 2-h-incubated and proteinase K-treated honey samples. The last results suggest that action of protein components GOX and Def-1 in diluted honeys do not markedly contribute to the antibacterial activity of the blossom honeys.
The above findings supported the idea that H 2 O 2 , accumulated during the ripening of honey, has a major role in antibacterial activity of blossom honeys. It seems that the further production of H 2 O 2 in diluted honeys incubated at room temperature for 2 h contributed only partially to honey antibacterial activity. Moreover, production of H 2 O 2 was not connected with GOX action.
To demonstrate the role of H 2 O 2 in antibacterial activity of honey samples, a catalase treatment of honeys was performed. Catalase treated honeys had significantly lower (P < 0.001) antibacterial activity, with MIC values of 25-42% and 25-40% against S. aureus and P. aeruginosa, respectively, compared to non-treated samples having MIC values 3-27% and 8-25%, respectively. As expected, the MIC values of manuka honey following catalase or proteinase-K treatment did not change (Figure 7). activity. Moreover, production of H2O2 was not connected with GOX action.
To demonstrate the role of H2O2 in antibacterial activity of honey samples, a catalase treatment of honeys was performed. Catalase treated honeys had significantly lower (P < 0.001) antibacterial activity, with MIC values of 25-42% and 25-40% against S. aureus and P. aeruginosa, respectively, compared to non-treated samples having MIC values 3-27% and 8-25%, respectively. As expected, the MIC values of manuka honey following catalase or proteinase-K treatment did not change (Figure 7). The obtained results proved that H2O2 was major antibacterial factor of blossom honeys. This finding is also supported by a statistical analysis of the relationship between the level of total H2O2 and overall antibacterial activity of blossom honeys. Significant correlations between these two parameters were found in case of S. aureus (r = −0.48, P < 0.001) and P. aeruginosa (r = −0.65, P < 0.0001) Figure 7. Antibacterial activity of selected blossom honeys (n = 12) and medical-grade manuka honey following catalase and proteinase K treatment against (A) Staphylococcus aureus and (B) Pseudomonas aeruginosa isolates. The 50% (w/v) honey solutions were treated with catalase at room temperature (RT) for 2 h or with proteinase K at 37 • C for 30 min. As controls, antibacterial activity of untreated 50% (w/w) honey solutions are given, both not incubated (0 h) and incubated at RT for 2 h. The antibacterial activity was determined with a MIC assay. The MIC was defined as the lowest concentration of honey solution (%) inhibiting bacterial growth. M, manuka honey. The data are expressed as the mean values with standard deviation (SD).
The obtained results proved that H 2 O 2 was major antibacterial factor of blossom honeys. This finding is also supported by a statistical analysis of the relationship between the level of total H 2 O 2 and overall antibacterial activity of blossom honeys. Significant correlations between these two parameters were found in case of S. aureus (r = −0.48, P < 0.001) and P. aeruginosa (r = −0.65, P < 0.0001) ( Figure 8A,B). However, in spite of the mentioned overall correlations, there are a number of honey samples at which a weak correlation or no correlation can be observed. These honeys showed either similar MICs, 7-10%, at different concentrations of H 2 O 2 , 15-31 µg/g of honey; or markedly different MIC, 11-43%, at low concentrations of H 2 O 2 , 3.5-9 µg/g of honey ( Figure 8A).
( Figure 8A,B). However, in spite of the mentioned overall correlations, there are a number of honey samples at which a weak correlation or no correlation can be observed. These honeys showed either similar MICs, 7-10%, at different concentrations of H2O2, 15-31 μg/g of honey; or markedly different MIC, 11-43%, at low concentrations of H2O2, 3.5-9 μg/g of honey ( Figure 8A). To provide evidence of direct antibacterial activity of H2O2 against tested bacterial strains, we determined the MICs of artificial honey (AH) and AH enriched with H2O2 (at concentration of 50 μg/g of AH). A significant decrease of MIC values of AH enriched with H2O2 was demonstrated (Figure 9). This result suggests that H2O2 accumulated during the ripening of blossom honey can itself (not through the substance(s) derived by action of H2O2 and some other honey components) significantly affect the overall honey antibacterial activity. To provide evidence of direct antibacterial activity of H 2 O 2 against tested bacterial strains, we determined the MICs of artificial honey (AH) and AH enriched with H 2 O 2 (at concentration of 50 µg/g of AH). A significant decrease of MIC values of AH enriched with H 2 O 2 was demonstrated ( Figure 9). This result suggests that H 2 O 2 accumulated during the ripening of blossom honey can itself (not through the substance(s) derived by action of H 2 O 2 and some other honey components) significantly affect the overall honey antibacterial activity. Figure 9. Antibacterial activity of artificial honey (AH) and AH enriched in hydrogen peroxide H2O2 at concentration of 50 μg/g of AH against Staphylococcus aureus and Pseudomonas aeruginosa isolates. The antibacterial activity was determined with a MIC assay. The MIC was defined as the lowest concentration of honey solution (%) inhibiting bacterial growth. The data are expressed as the mean values with standard deviation. Differences between honey groups were analyzed by t-test. Asterisks indicate a significant difference between control AH and AH enriched in H2O2, *** P < 0.001.

Role of Polyphenols in the Antibacterial Effect of Blossom Honeys
Statistical analysis revealed a significant correlation between the TP content and overall antibacterial activity of blossom honeys against S. aureus (r = −0.40, P < 0.01) and P. aeruginosa (r = −0.47, P < 0.01) ( Figure 8C,D). Nevertheless, similarly as it is in the case of H2O2, there are honey samples at which weak or no correlation exists among these parameters. Taking all obtained results in the work, it seems that the concentration of polyphenols in blossom honeys together with the level of H2O2 are critical factors in determining the honey antibacterial potential.

Discussion
Antibacterial activity is the most evaluated and investigated biological activity of honey. Nevertheless, the mechanism of honey antibacterial action has not been elucidated in a sufficient manner. In this study, we characterized the inhibitory and bactericidal activity of 57 Slovak blossom honeys; determined contents of GOX (by an established a semi-quantitative assay), H2O2, and TP in honeys; and evaluated the role of these components in antibacterial activity.
Numerous studies have examined the antibacterial activity of honey from different botanical and geographical origin. Some types of honeys (e.g., buckwheat, fir honeydew honey) were more potent than others, but all honeys, except manuka honey, contained the same antibacterial substance, H2O2. Currently, it is believed that H2O2 is mainly produced by GOX-mediated conversion of glucose to gluconic acid under aerobic conditions in diluted honey [15]. However, the results presented here, together with our previous recent observations [10]-as well as results from other studies [21][22][23]support the theory by which H2O2 in honey is also produced via an alternative non-enzymatic pathway. Two facts were revealed in this study, (1) absence of a correlation between the contents of GOX and H2O2 occurring in blossom honeys, and (2) determination of the similar MIC values in diluted honeys not incubated and incubated with proteinase-K (destroying activity of GOX and obviously also other proteinous substances) provide evidence that GOX as well as Def-1 have no significant effect on the antibacterial activity of undiluted and diluted blossom honeys. Similar results were found in our recent study [10] where 23 honeydew honeys were analyzed and overall antibacterial activity was evaluated. In the recent study, statistical analysis did not show any The antibacterial activity was determined with a MIC assay. The MIC was defined as the lowest concentration of honey solution (%) inhibiting bacterial growth. The data are expressed as the mean values with standard deviation. Differences between honey groups were analyzed by t-test. Asterisks indicate a significant difference between control AH and AH enriched in H 2 O 2 , *** P < 0.001.

Role of Polyphenols in the Antibacterial Effect of Blossom Honeys
Statistical analysis revealed a significant correlation between the TP content and overall antibacterial activity of blossom honeys against S. aureus (r = −0.40, P < 0.01) and P. aeruginosa (r = −0.47, P < 0.01) ( Figure 8C,D). Nevertheless, similarly as it is in the case of H 2 O 2 , there are honey samples at which weak or no correlation exists among these parameters. Taking all obtained results in the work, it seems that the concentration of polyphenols in blossom honeys together with the level of H 2 O 2 are critical factors in determining the honey antibacterial potential.

Discussion
Antibacterial activity is the most evaluated and investigated biological activity of honey. Nevertheless, the mechanism of honey antibacterial action has not been elucidated in a sufficient manner. In this study, we characterized the inhibitory and bactericidal activity of 57 Slovak blossom honeys; determined contents of GOX (by an established a semi-quantitative assay), H 2 O 2 , and TP in honeys; and evaluated the role of these components in antibacterial activity.
Numerous studies have examined the antibacterial activity of honey from different botanical and geographical origin. Some types of honeys (e.g., buckwheat, fir honeydew honey) were more potent than others, but all honeys, except manuka honey, contained the same antibacterial substance, H 2 O 2 . Currently, it is believed that H 2 O 2 is mainly produced by GOX-mediated conversion of glucose to gluconic acid under aerobic conditions in diluted honey [15]. However, the results presented here, together with our previous recent observations [10]-as well as results from other studies [21][22][23]-support the theory by which H 2 O 2 in honey is also produced via an alternative non-enzymatic pathway. Two facts were revealed in this study, (1) absence of a correlation between the contents of GOX and H 2 O 2 occurring in blossom honeys, and (2) determination of the similar MIC values in diluted honeys not incubated and incubated with proteinase-K (destroying activity of GOX and obviously also other proteinous substances) provide evidence that GOX as well as Def-1 have no significant effect on the antibacterial activity of undiluted and diluted blossom honeys. Similar results were found in our recent study [10] where 23 honeydew honeys were analyzed and overall antibacterial activity was evaluated. In the recent study, statistical analysis did not show any correlation between antibacterial activity of honeydew honeys and level of H 2 O 2 . In contrast to these findings, the significant statistical correlation between these two parameters was observed within blossom honeys investigated in this work. A stronger correlation between MIC values against P. aeruginosa and H 2 O 2 content than against S. aureus was found. Nevertheless, some of the examined blossom honey samples did not show any correlation among MICs and H 2 O 2 content . We suppose that in these samples, the other factors associated with production of hydroxyl radicals and with variations in Cu(I) and Fe(II), polyphenols and catalase levels and other yet unknown factors affected their final antibacterial efficacy. Similarly, Grecka and co-workers (2018) determined the overall antibacterial activity and concentration of H 2 O 2 in 144 honey samples from Northern Poland [12]. Although a significant correlation between the level of accumulated H 2 O 2 and antibacterial activity of Polish honey samples was also documented, the samples with low MIC values (the most potent samples) were characterized with low concentration of H 2 O 2 . In contrast to our study, they determined H 2 O 2 concentrations in 25% (w/v) honey solutions in deionized water after 1 h of incubation at 37 • C. Even though the different methodological approach in H 2 O 2 determination resulted in statistically-significant correlations in both studies, we consider that the deeper understanding the role of H 2 O 2 in honey antibacterial action will require a study of the level of generated H 2 O 2 levels at individual MIC values of each honey.
One of the important observations in this study is that the initial content of H 2 O 2 in undiluted blossom honey is able to inhibit the growth of both tested bacteria. Despite some published data reporting that cell death of bacteria requires concentrations of H 2 O 2 higher than 50 mM [14] and the content of H 2 O 2 in honey is 900-fold lower than the H 2 O 2 content used in medical disinfectants [24], we showed here that H 2 O 2 at concentration about 1.0 mM (ca. 50 µg/g of honey) in undiluted AH was able to inhibit bacterial growth which was documented by a decrease in the MIC values of artificial honey from 45 to 14% for S. aureus and from 25 to 17% for P. aeruginosa.
Depending on several factors but mainly on its concentration, H 2 O 2 may exert either bacteriostatic or bactericidal effects. H 2 O 2 at low concentrations, that are more stable than higher concentrations, exerts weak antibacterial activity; however, extended slow-release of H 2 O 2 may improve its antibacterial activity and overwhelm the effect of bacterial/plant catalase [25].
Questions as to how H 2 O 2 is generated in diluted honeys and other factors that affect H 2 O 2 antibacterial action remain unanswered. We assumed that plant-derived polyphenolic compounds, often reported in honey samples, might be one factor responsible for H 2 O 2 production and elevating honey antibacterial activity. Indeed, as shown in this study, a significant correlation between the concentration of TP and antibacterial activity of tested blossom honey was documented. These findings are in agreement with another study, where a strong correlation between the antibacterial activity of various honey types and phenolic content was documented [26]. Another critical factor affecting the content of H 2 O 2 in honey can be presence of pollen-derived catalase, a plant enzyme responsible for degradation of H 2 O 2 . Interestingly, catalase has occasionally been reported in honey samples and the most proteomic studies have not been able to identify this enzyme in honey samples [27][28][29][30][31]. A more objective approach seems to be the identification of catalase in honey by determining its enzymatic activity [32]. Nowadays, it is difficult to evaluate the content of catalase in particular honey types because a reliable quantitative method is missing. In this study, we observed a strong catalase activity destroying H 2 O 2 in the case of acacia honeys. The sharp association between the effect and the type of honey is interesting and will require further study.
Polyphenols, mostly flavonoids and phenolic acids that are used as markers of honey botanical origin and for its authentication [33], attract honey consumers' interest because of their health-promoting and biological effects. The concentration of polyphenols and flavonoids in honey varies between 56 and 500 mg/kg and 0.6 and 6.4 mg/kg of honey, respectively [34][35][36][37][38]. A direct antibacterial effect of polyphenols in honey is negligible due to their very low concentration in honey. In fact, they may work in two ways to promote antimicrobial activity: 1) by producing H 2 O 2 and 2) by reducing Fe (III) to Fe (II), which triggers the Fenton reaction and lead to creation of more potent reactive oxygen species such as hydroxyl radicals. Recently, Grzesik and co-workers [39] tested 54 natural antioxidants, including polyphenols, and revealed that 27 of them were able to generate H 2 O 2 by autoxidation of polyphenols. No correlation between the structure of the studied compounds and generation of H 2 O 2 was disclosed.
It has been reported that Def-1, an antibacterial bee-derived peptide, is one of the major antibacterial compounds found in every type of honey [4,8,40,41]. Apart from its antibacterial/antibiofilm properties, Def-1 contributes to cutaneous wound closure and promotes wound re-epithelization [42]. Nevertheless, we showed in this study, and elsewhere [10], that destroying of Def-1 in honey by protease digestion did not affect the overall antibacterial activity of blossom and honeydew honeys. Similarly, Stagos et al. (2018) demonstrated that antibacterial activity of proteinase K-treated Greece, mainly polyfloral, honeys did not differ from the untreated ones [13]. These findings suggest that Def-1 and other protein components of bee or botanical origin do not contribute (or their contribution is negligible) to the antibacterial effects of the honey. Generally, these observations undoubtedly have some advantages from a clinical point of view and support the use of medical-grade honey in clinical practice for healing of chronic and infected wounds. Chronic wounds remain in the inflammatory stage of healing, where degradation of the extracellular matrix by proteases, particularly matrix metalloproteinases, occurs [43]. In addition, chronic wounds are commonly colonized by invading bacteria, which also secrete proteases that contribute to the overall high proteolytic activity in the wound environment. Defensins and other antimicrobial peptides are susceptible to degradation and inactivation by wound or bacterial proteases [44]. Our findings suggest that proteolytic degradation and inactivation of biologically active enzymes and peptides in honey, after its application to wounds, will most likely not be connected with decreasing honey antibacterial efficacy.
Currently, the use of honey in wound care is less mainstream, mainly due to: (a) the low quality of clinical evidence supporting the use of honey in wound and burn treatment [45]; (b) the unclear overall impact of honey on wound healing mechanisms [46]; and (c) the unclear mechanisms of action of different types of honey in wound healing. The present study, together with our recent observations [10], showed that antibacterial effects of honeydew and blossom honeys is H 2 O 2 -dependent, and their antibacterial efficacy can be-at least for honeydew, polyfloral honeys, and some monofloral honeys-as effectiveas or more effective than manuka honey. Despite manuka honey having some advantages compared to other honeys, such as the existing knowledge of the mechanism of its action through methylglyoxal (it enables its standardization). It is important to select another source of honey with proven antibacterial activity that will serve as a base for new kinds of medical-grade honey.
In conclusion, 57 tested blossom honey samples showed different antibacterial efficacy against S. aureus and P. aeruginosa as follows: wildflowers honeys > acacia honeys > rapeseed honeys. Four selected honeys of each honey type were subjected to detailed characterization of the antibacterial activity. This activity was markedly reduced by treatment with catalase but remained stable after complete protein digestion, suggesting that the antibacterial activity of blossom honeys is not directly dependent on GOX-mediated production of H 2 O 2 or the presence of Def-1. Lastly, and most importantly, we demonstrated that the overall antibacterial activity of honeys was correlated with the level of total H 2 O 2 present already in ripened honeys and with total polyphenol content.

Honey Samples
Honey samples (n = 57) collected in 2016 were received from beekeepers from several regions of Slovakia. Upon receiving, they were immediately stored in plastic containers at 4 • C in the dark. Identification of the floral source of the honey was carried out by a melissopalynological analysis, as reported above. Medical-grade honey, manuka honey MGO 550, was purchased from Manuka Health (Newmarket, Auckland, New Zealand). Artificial honey was prepared as described elsewhere [47] by dissolving 39 g d-fructose, 31 g d-glucose, 8 g maltose, 3 g sucrose, and 19 g distilled water.

Microorganisms
The antibacterial activity of honey samples was assessed against the isolates Pseudomonas aeruginosa CCM1960 and Staphylococcus aureus CCM4223, obtained from the Department of Medical Microbiology, Slovak Medical University (Bratislava, Slovakia).

Strains, Vectors, Enzymes, and Reagents Used for GOX Gene Cloning
Escherichia coli strain DH5α (ThermoFisher Scientific, Loughborough, UK) was used as the host for GOX cDNA cloning. E. coli Rosetta-gami (DE3) was purchased from Novagen (Darmstadt, Germany) and used as the host for expression of the heterologous protein. Bacteria were grown in Luria-Bertani (LB) medium or LB agar plates with appropriate antibiotics at 37 • C. E. coli cells were transformed using a standard heat-shock method. Plasmid pET28b(+) (Novagen) was chosen for the construction of expression plasmid. Phusion High-Fidelity PCR Master Mix, T4 DNA ligase, and restriction enzymes were purchased from New England BioLabs (Hitchin, UK), while other chemicals were purchased from Sigma-Aldrich (Taufkirchen, Germany) or Promega (Madison, WI, USA).

Construction of the GOX Expression Plasmid and Expression of Recombinant GOX (rGOX)
The cDNA encoding mature honeybee GOX were amplified using following primers: sense primer 5 -GGCGCGCGCCCATGGAGCCGTGCCAGCGC-3 ; antisense primer 5 -GGCGCGCGCGGCCGCCAGTAATTGAAGTAATTCTCC-3 . The amplified cDNA fragment was purified, digested with NcoI and NotI, then ligated into the pET28b(+) vector, which was digested with the same restriction enzymes, to construct the expression plasmid pET28-GOX. The resulting plasmid was transformed into DH5α E. coli and cDNA verified by DNA sequencing (GATC Biotech, Constance, Germany). The correct plasmid encoded a translational fusion protein containing the mature protein sequence of bee GOX, followed by a 6x His tag.
To express rGOX, an E. coli/pET28-GOX clone was inoculated into 15 mL LB medium supplemented with kanamycin (30 µg/mL) and chloramphenicol (50 µg/mL), then cultured at 37 • C overnight with shaking at 250 rpm. The overnight culture was diluted 100-fold into fresh LB medium (the total volume for expression of rGOX was 1200 mL) with mentioned antibiotics. When the cell culture reached the absorbance at 600 nm about 0.5 (mid-exponential phase), the expression of rGOX was induced by the addition of IPTG to a final concentration of 1.0 mM. After 4 h of incubation at 37 • C and 250 rpm, the harvested cells were immediately centrifuged at 4000 rpm for 10 min.

Purification of rGOX
The rGOX was isolated from crude bacterial lysate under denaturing conditions by affinity chromatography employing Ni-NTA (Qiagen, Hilden, Germany). A soluble protein fraction from the harvested cells was extracted using B-Per Reagent (ThermoFisher Scientific), according to the manufacturer's instructions. After centrifugation at 14,000 rpm for 20 min, the supernatant was removed and the pellet (insoluble protein fraction) was solubilized in denaturation buffer (6 M urea, 0.02 M Tris-HCl, 0.5 M NaCl, pH 8.0). The obtained solution was loaded onto the Ni-NTA column pre-equilibrated in denaturation buffer. The column was washed in denaturing buffer supplemented with 50 mM imidazole and pure rGOX was eluted with denaturing buffer supplemented with 250 mM imidazole. After Ni-NTA chromatography, purified rGOX was subjected to exchange buffer using a PD-10 desalting column (GE Healthcare, Little Chalfont, UK), eluted with 4 M urea solution and stored at −20 • C. The purity of the prepared rGOX was determined by 12% SDS PAGE, and total protein content was measured using the Quick Start Bradford protein assay (Bio-Rad, Hercules, CA, USA) as described in the instruction manual.

Melissopalynological Analysis
Qualitative and quantitative analyses of the honey sediment were performed according to the method proposed by Von der Ohe et al. [48]. Briefly, ten grams of honey were dissolved in 20 mL of distilled water and then centrifuged for 5 min at 3000× g. The pellet, after washing with distilled water, was spread on a microscope slide, dried at 40 • C, and colored by the addition of some drops of glycerin jelly enriched with 0.1% basic fuchsine ethanol solution. The slide was covered with a cover slip and examined under optical microscope. Pollen grains were identified and counted in groups of 100, following five parallel equidistant lines uniformly distributed from one edge of the cover slip to the other. For each pollen type, the relative frequency, or rather the percentage amount compared to the total number of counted pollen grains, was calculated. Moreover, non-identified grains and honeydew indicators were counted. Honey botanical origin was determined on the basis of the relative frequencies of pollen types identified in the microscopic analysis, as reported by Di Marco and co-workers [49].

Determination of Honey Antibacterial Activity
The antibacterial efficacy of the honey samples was evaluated with a MIC assay as described by Bucekova et al. (2014) [20]. Briefly, overnight bacterial culture was suspended in phosphate-buffered saline (PBS), pH 7.2, and the turbidity of the suspension was adjusted to 10 8 colony-forming units (CFU)/mL and diluted with Mueller-Hinton broth (MHB) medium (pH 7.3 ± 0.1) to a final concentration of 10 6 CFU/mL. Then, 10 µL aliquots of suspension were inoculated into each well of a sterile 96-well polystyrene U-shape plates (Sarstedt, Nümbrecht, Germany). The final volume in each well was 100 µL, consisting of 90 µL of sterile medium or diluted honey and 10 µL of bacterial suspension. After 18 h of incubation at 37 • C and 1000 rpm, bacterial growth inhibition was determined by monitoring the optical density at 490 nm. The MIC was defined as the lowest concentration of honey that completely inhibits bacterial growth. All tests were performed in triplicate and repeated three times.
MBC values of honey samples were also evaluated [50]. The viability of bacteria in wells with no turbidity was determined by spreading 100 µL onto an MHB agar plate and incubating at 37 • C for 24 h. The lowest concentration of honey that resulted in no survival of viable bacteria was recorded as MBC.

Determination of H 2 O 2 Content
H 2 O 2 content in honey samples was determined using Megazyme GOX assay kit (Megazyme International Ireland Ltd, Bray, Ireland), which is based on H 2 O 2 release. As a standard, H 2 O 2 diluted to 9.8-312.5 µM was used. 40% (w/w) honey solutions in 0.1 M potassium phosphate buffer (pH 7.0) were prepared and immediately measured. Each honey sample and H 2 O 2 standard was tested in duplicate in a 96-well microplate. The absorbance of reaction was then measured at 510 nm using a Synergy HT microplate reader (BioTek Instruments, Winooski, VT, USA).

Determination of Total Polyphenolic Content
Total polyphenolic content was determined with a Folin-Ciocalteu Phenolic Content Quantification Assay Kit (BioQuoChem, Llanera, Spain) in a 20% (w/v) honey solution in a 96-well microplate according to the manufacturer's instructions. Gallic (GAE) acid was used as the reference standard and results were expressed as GAE equivalents (µg GAE/g of honey). Absorbance was measured at 700 nm at 37 • C.

Enzymatic Treatment of Honey Samples with Catalase and Proteinase K
Diluted honey samples (50% w/w in MHB medium) were treated with catalase (2000-5000 U/mg protein; Sigma-Aldrich) at a final concentration of 1000-2500 U/mL at room temperature for 2 h, or proteinase K (30 U/mg; Promega) at a final concentration of 50 µg/mL at 37 • C for 30 min. Aliquots (15 µL) of proteinase-K treated samples were separated by SDS-PAGE using a Mini-Protean II electrophoresis cell (Bio-Rad) and treatment efficacy assessed after gel staining with Coomassie Brilliant Blue R-250. Catalase-and proteinase K-treated honey samples were then immediately used in the antibacterial assay to determine MIC values against S. aureus and P. aeruginosa.

Statistical Analysis
One-way ANOVA and Tukey's multiple comparison tests were used to evaluate antibacterial efficacy of different botanical types of honey. The Pearson correlation test was used for correlation analysis between antibacterial activity/GOX content and H 2 O 2 content or total polyphenol content in honeys. The data are expressed as mean values with the standard deviation (SD). Data with P values smaller than 0.05 were considered as statistically significant. All statistical analyses were performed using GraphPad Prism (GraphPad Software Inc., version 5, La Jolla, CA, USA).