Mr. Ogihara of Ogihara yohoen, Inc. (Nagano Prefecture) kindly donated nine types of Japanese honey, as well as chestnut leaves and chestnuts, to us. Honeys obtained from some plants (horse chestnut, blueberry, bush clover, clover, apple, mixed flowers, cherry, acacia, and chestnut) were used for this study.

Bis-(2,4,6-trichlorophenyl)oxalate (TCPO), bis-(N-methyl acridinium) (lucigenin), and isoluminol manufactured by Tokyo Chemical Industry Co., Ltd. (TCI, Tokyo) were used for the chemiluminescence assays. 8-anilino-naphthalene-1-sulfonic acid (ANS) and micro-peroxidase were acquired from Wako. MGO manufactured by Sigma and dihydroxyacetone, H₂O₂, glucose, and fructose manufactured by Wako were used as standard bactericidal components. The enzymes catalase manufactured by Nacalai Teque and superoxide dismutase (SOD) manufactured by Wako, each prepared at concentrations ranging from 100 to 10,000 U/assay, were used to check for enzymatic generation of active oxygen. 2-bromo-4’-nitroacetophenone utilized as internal standards for HPLC was obtained from TCI. 5,5-dimethyl-1-pyrroline N-oxide (DMPO) was obtained from Dojindo Laboratories. A pH paper (pH indicator strips, pH 0 - 14 universal indicator) manufactured by Merck was used for pH measurement. The water purified in a Milli-Q system (Merck Millipore) was used for all experiment.

The nine types of Japanese honey and the manuka honey within one year of collecting them were used. Some of the horse chestnut within one year of collection, while some has been stored for ten years were used. 0.1 - 0.5 g of each type of honey was dissolved in 1 mL of water (aqueous solutions from 10% to 50%) and these solutions passed through a hydrophilic 0.45 µm membrane filter (Kurabo Industries Ltd., Osaka) were used as samples. We have included values on the component concentrations in the various types of honey, converted into units of concentration (mol/L) in 100% aqueous solution of honey. In enzyme reactions, each solution of catalase and SOD was formulated by dissolving the respective enzyme (100 - 10,000 units/assay) in analytical grade water. A honey sample was combined with an enzyme solution and the mixture was incubated for 30 min at 37˚C.

100 μL of 0.05 mmol/L TCPO dissolved in ethyl acetate and 100 μL of ANS solution (0.02% ANS, 0.1% BSA, 0.2 mol/L barbital) was mixed with 10 μL of sample. The mixture was allowed to react for 10 seconds, and then the luminescence (integrated value) was measured during 10 seconds with a BLR-301 luminescence reader (Aloka).

270 μL of a spin trapping solution (containing 50 mmol/L phosphate buffer (PB) (pH 7.4), 0.5 mmol/L FeSO₄, 1 mmol/L diethylene triamine pentaacetic acid, and H₂O) and 5 μL of DMPO were mixed with a 30 μL sample aliquot, and the ESR spectrum was measured with a JES-FR30 ESR spectrometer (JEOL Resonance, Inc.). The ESR conditions included a microwave power of 4 mW, a modulation frequency of 9420.0 MHz, a field center of 335.250 mT (width 5.0 × 1 mT), a mod of 100 kHz (width 1.0 × 0.1 mT), an amplitude of 5.0 × 10, a time constant of 0.03 sec, and a sweep time of 1 min.

100 μL of lucigenin luminescence reagent solution (2.4 × 10⁻⁵ mol/L lucigenin, 0.16 mol/L KOH, and 0.02% Triton X-100) was mixed with a 10 μL aliquot of the sample. This was allowed to react for 10 seconds, and the subsequent luminescence (integrated value) was measured during 10 seconds with a BLR-301 luminescence reader (Aloka).

200 μL of a spin trapping solution (2.4 × 10⁻⁵ mol/L lucigenin, 0.16 mol/L KOH, and 0.02% Triton X-100) and 5 μL of DMPO were mixed into 30 μL of the sample, and the ESR spectrum was immediately measured afterward. The xanthine-xanthine oxidase reaction was used as an O 2 − control, and 20 μL of 20 mmol/L xanthine, 20 μL of 1 U/mL xanthine oxidase, and 200 μL of 0.1 mol/L phosphate buffer (pH 7.4) were mixed with 5 μL of DMPO. Then an ESR spectrum of the mixture was measured. The ESR conditions were the same as those used to measure hydroxy radicals.

80 μL of 0.8 mol/L bicarbonate buffer (pH 9.5), 200 μL of the isoluminol luminescence solution (containing 2.5 × 10⁻⁴ mol/L isoluminol and 1.0 × 10⁻⁶ mol/L micro-peroxidase dissolved in 0.8 mol/L bicarbonate buffer), and 5 μL of 1.0 × 10⁻⁵ mol/L H₂O₂ was mixed with a 50 μL sample aliquot. This was incubated for 10 seconds to allow the reaction to attain completion, followed by measurement of the generated luminescence (integrated value) during 10 seconds with the Luminescence-PSM AB-200. We calculated the antioxidative potential relative to H₂O₂ from the following formula:

Antioxidative   potential   ( % ) = 100 − ( X Y × 100 )

X indicates “luminescence intensity when the sample was added”. Y indicates “luminescence intensity when the sample was not added”.

An intelligent HPLC pump (Jasco Co.) and a CLD-110 chemiluminescence detector (manufactured by Tohoku Electronic Industrial) were used for the HPLC system. A two linked Inertsil NH₂ columns (4.6 mm × 250 mm, 5 µm particle size, manufactured by GL Sciences) was used for the separation column. 90% Acetonitrile was used for the mobile phase. The lucigenin reagent (2.4 × 10⁻⁶ mol/L lucigenin, 0.016 mol/L KOH, and 0.002% Triton X-100) diluted by a factor of 10 was used for HPLC as the luminescence reagent solution. The flow rate was maintained at 1 mL/min. 5 μL of the sample was injected into the HPLC system, and a chromatogram was obtained using the generated luminescence intensity. 4.1 × 10⁻³ mol/L 2-bromo-4’-nitroacetophenone (2B) was used as an internal standard.

In order to detect the presence of H₂O₂ in Japanese honey, a 10% aqueous solution of each honey was prepared and the TCPO/ANS-CL assay was performed. As shown in Figure 1, the amount of luminescence varied greatly depending on the type of honey; hence, we verified that the amount of H₂O₂ was different, depending on the honey plant from which the honey was derived. Chestnut honey showed the greatest value (13.8) for the signal-to-noise ratio (S/N ratio). Also, since the luminescence values of each honey on catalase treatment decreased, it was verified that H₂O₂ was included in all these types of honey. Next, the amount of H₂O₂ in each honey was calculated from an H₂O₂ calibration curve obtained with TCPO/ANS-CL. Chestnut honey showed the highest value of 3.2 × 10⁻⁴ mol/L, followed by the other types with concentrations of at least 2.0 × 10⁻⁵ mol/L. In order to verify whether the H₂O₂ in chestnut honey was derived from the honey plant, the source of H₂O₂ present in chestnut leaves and fruits were measured using TCPO/ANS-CL. 1 g each of chestnut leaves and chestnuts were grounded up in a mortar, were added 50 mL of water, were incubated for 30 minutes at 37˚C, and then were filtered the supernatant liquid with a 0.45 μm filter to prepare samples. Chestnut leaves and chestnuts were shown to have an S/N ratio of 10.7 and 11.8, respectively; which proves that H₂O₂ is found in honey plants.

H₂O₂ generates hydroxy radicals in the presence of FeSO₄, thereby becoming a DMPO adduct. This leads to a characteristic four-line ESR spectrum with an intensity ratio of 1:2:2:1. In this work, The H₂O₂ present in chestnut honey characterized as the one with strongest luminescence among all types of Japanese honey was identified as hydroxy radicals using ESR. The samples at 50% concentration were prepared, the hydroxy radical content of chestnut honey was independently verified. The results of the ESR measurement were shown in Figure 2(a). Next, a complete absence of the hydroxy radical spectrum was observed after reaction with catalase, as shown in Figure 2(b).

The authors have previously reported on the production of O 2 − from glucose, fructose, and MGO in manuka honey [17]. Japanese honey could include components other than H₂O₂ that are capable of generating active oxygen species. Therefore, in this study, a 10% aqueous solution of each type of honey was prepared and the lucigenin-CL assay was used to investigate the detection of O 2 − derived from bactericidal components. The results were shown in Figure 3, all honey types showed strong luminescence. Among them, acacia honey had the largest S/N ratio of 1350, due to its rather strong luminescence. On the other hand, the S/N ratio of chestnut honey was 208, which was the lowest among the types of honey measured at this time. The results of TCPO/ANS-CL (Figure 1) for chestnut honey showed the inclusion of a large amount of H₂O₂, but the lucigenin-CL results estimated the fraction of components that could generate species such as O 2 − tended to be small. From the above one could suppose that all types of honey generate O 2 − from some component.

Next, acacia honey, which showed the greatest luminescence with lucigenin-CL was carried out its reaction with catalase and SOD to investigate how the lucigenin chemiluminescence value would change. The amount of luminescence was weakened to almost a blank value. Therefore, it could be assumed to be due to the disappearance of O 2 − and H₂O₂ derived from some component, such that the chemiluminescence with lucigenin was weakened.

The O 2 − generated from some components of Japanese honey was identified using ESR. A sample of 50% chestnut honey was used in this study. A xanthine-xanthine oxidase reaction was used as a control for the O 2 − spectrum. As a result, the ESR spectrum by O 2 − generated from a component in chestnut honey was detected. When the chestnut honey was incubated with SOD and the O 2 − spectrum in the ESR plot was found to disappear. This allowed an independent verification that O 2 − was generated from a component included in chestnut honey. Figure 4(a) shows the O 2 − spectrum at 50% chestnut honey, and Figure 4(b) shows O 2 − spectrum at 50% chestnut honey obtained by the SOD treatment.

In this study, we used the isoluminol chemiluminescence method to examine the antioxidative potential of each type of honey. The antioxidative potential based on the weakening of the amount of luminescence due to H₂O₂ was determined, when each type of honey was mixed with 1 × 10⁻⁵ mol/L H₂O₂. The strength of the antioxidative potential to that of 1 × 10⁻⁵ to 10⁻³ mol/L ascorbic acid was compared as control. As shown in Figure 5, chestnut honey had the greatest antioxidative potential, followed by acacia and then horse chestnut (stored for ten years) honey. From these results, it appears that the higher H₂O₂ concentration, as measured by TCPO/ANS-CL and shown in Figure 1, leads to a stronger antioxidative potential. Prior research has mainly studied the active oxygen species included in honey. It is held that plants generate active oxygen on their own for its antimicrobial (oxidative) action, in order to protect themselves from external enemies. On the other hand, data from the present study showed that the measured antioxidation potential is contrary to that of antimicrobial action. Plants are credited to produce antioxidants, such as polyphenol, in order to eliminate internally generated active oxygen, and thereby prevent oxidation of the plant. Since our data linked higher H₂O₂ content in a type of honey to greater antioxidative potential, certain components of honey must be capable of H₂O₂ production potential. This could be the same component as the factor that shows antioxidative potential. With regard to this relationship between H₂O₂ generation potential and antioxidative potential, similar results have been obtained for aroma oil [25]. Furthermore, in the results obtained with lucigenin-CL (Figure 3), large amounts of reducing agents and O 2 − were detected in acacia, horse chestnut (stored for ten years), cherry, and apple honey. These results may also influence the antioxidative potential. This could translate to increase in active oxygen on reduction of oxygen by an electron from some other component. It is considered that the remaining radical, that has lost an electron, could show antioxidative potential. Therefore, more detailed analysis of components that show antioxidative potential is needed.

Simple analytical methods for the simultaneous detection of antibacterial components in honey like MGO, DHA, fructose, glucose, and H₂O₂, have not been reported. We have focused on reactive oxygen species (ROS) produced from these five components to permit easy detection of these compounds in honey. Lucigenin-CL is used primarily for the measurement of ROS such as O 2 − . The authors have previously developed lucigenin-CL-HPLC as a new method to simply and simultaneously detect five bactericidal components (H₂O₂, MGO, DHA, glucose, and fructose) in manuka honey [17]. We applied this lucigenin-CL-HPLC method to the analysis of Japanese honey used in this study. The samples of 40% honey were prepared and the nine types of honey were checked with the lucigenin-CL-HPLC method. The four species H₂O₂, glucose, fructose, and MGO have been detected in all types of honey. MGO is a bactericidal component believed to be specific to manuka honey, but the results showed that it is also included in Japanese honey. On the other hand, DHA, which was included in manuka honey, was not detected in Japanese honey. Figure 6 shows two examples of chromatograms (corresponding to chestnut and acacia honey). In addition, the concentrations calculated from calibration curves for respective standard products of the included components were listed in Table 1. The H₂O₂ concentration for each type of honey measured with lucigenin-CL-HPLC was roughly the same as the value measured with TCPO/ANS-CL. Chestnut honey had the largest content at 1.8 × 10⁻⁴ mol/L. Glucose and fructose showed roughly the same values as in a previous report [22]. Even though the amount of MGO was smaller than that in manuka honey, its inclusion in all types of Japanese honey represented a novel finding. MGO increases on storage of manuka honey due to its gradual conversion from DHA [17]. On the other hand, since Japanese honey did not contain DHA, it was assumed that the MGO is generated from a different component.