Chemical Profiling of Polyfloral Belgian Honey : Ellagic Acid and Pinocembrin as Antioxidants and Chemical Markers

Chemical profiling of northern Belgian polyfloral honeys was performed to analyse their phenolic compound content (flavonoids and phenolic acids). First, samples were subjected to a standard analysis of their physicochemical properties, and then, the phenolic fraction was isolated and analysed using a HPLC/PAD method. All of the tested honeys showed a common and specific phenolic profile that could be the basis for the differentiation of polyfloral honeys of the Antwerp region from other polyfloral honeys. Chromatographic data indicated a high content of ellagic acid (9.13–13.66mg/100 g honey), as well as the flavonoid pinocembrin (1.60–1.85mg/100 g honey) in these honeys. Ellagic acid, a compound with well-defined prohealth activities, might be used as a chemical marker for these honeys. With respect to total phenolic and flavonoid contents, 1,1-diphenyl-2-picrylhydrazyl (DPPH) assays were determined spectrophotometrically. The honey exhibited a moderate antioxidant activity, typical for light honeys.


Introduction
Because of their health-promoting activities, polyphenolic compounds are always of interest among substances of natural origin.Until now, the biological activity of a variety of phenolic acids and flavonoids of plant origin has been described in some detail.The most important feature of these compounds is their antibacterial and antioxidant activity [1][2][3][4][5].Dietary studies have clearly shown a correlation between the content of phenolic compounds in plant foods and a decrease in the incidence of lifestyle diseases [6,7].From a wide variety of honeys, consumers frequently choose unifloral honeys because their specific medicinal properties are well defined.On the other hand, polyfloral honey, called thousand flowers honey, is still the most popular and most widely consumed variety.Polyphenolic compounds are present in honeys due to their direct transfer from the nectar of plants by bees.The chemical patterns of these compounds, namely, their composition and relative concentration, have already been used for the authentication of honey bars [8,9].However, most studies focusing on unifloral honey are devoted to the identification of individual markers of their botanical origin [9][10][11].
In this study, we focused on the determination of the chemical profiles, namely, the identification of phenolic acids and flavonoids, of Belgian polyfloral honeys from three apiaries located in the Antwerp province to determine the utility of this approach for the determination of the geographical origin of honey.Furthermore, the antioxidant properties of these honeys were determined and correlated with the content of phenolic compounds.

Honey Samples.
Twenty-seven Belgian honey samples from the Antwerp province were studied.They were obtained from three beekeepers operating in the vicinity of Antwerp, namely, in Mortsel (M, the beekeeper guild was established there in 1695), Lichtaart (L), and Kasterlee (K).Honey sample originates from 3 successive years (2007, 2008, and 2009).All of the samples, after being acquired from the beekeepers, were stored in the dark at 4 ∘ C. Before they were studied, they were characterized by the determination of their moisture content, electrical conductivity, specific rotation, and invertase activity as well as by using classical visual and organoleptic testing.All samples of honeys were certified from their producers.The botanical origin of the analysed honeys was confirmed and certified by pollen analysis.The proper documents have been presented by beekeepers supplying the honey.

Isolation of Phenolic Compounds.
Extraction was carried out as described previously [9] with some modifications.The studied polyfloral honey samples (5 g) were thoroughly mixed with five parts (25 mL) of distilled water and sonicated for 1 hour until completely homogenized.The aqueous sample was then passed through filter paper to remove any solid particles.The filtrate was mixed with 40 g of Amberlite XAD-2 (pore size 9 nm, particle size 0.3-1.2mm) and stirred with a magnetic stirrer for 10 min to absorb phenolic compounds.The Amberlite particles were then packed into a glass column (3.5 × 65 cm), and the column was washed with a diluted hydrochloric acid (0.05 M, 120 mL) and then with distilled water (100 mL).The phenolic compounds were retained on the column, while all sugars and other polar compounds eluted.The phenolic fraction was then eluted from the column with methanol (150 mL) and evaporated under reduced pressure (40 ∘ C).The obtained residue was redissolved in a mixture of distilled water (1 mL) and HPLC-grade methanol (1 mL), and the mixture was analysed with the use of a HPLC-PDA system.The applied extraction method enabled the recovery values for analysed compounds to be higher than 85%.

Total Phenolic Content (TPC).
To determine the TPC of the honey extracts, the Folin-Ciocalteu method was applied [9,12,13].Briefly, 5 mL of aqueous eluate of honey (0.5 g honey/50 mL of distillate water) was placed into a 10 mL volumetric flask.Folin-Ciocalteu (FC) reagent (0.5 mL) and 1.5 mL of a 20% Na 2 CO 3 solution were added to the eluate.After filling the flask with distilled water to the mark and thoroughly agitating the reaction mixture, it was left to stand for 120 min at room temperature, and the TPC was determined spectrophotometrically (Hitachi U-2810) at 760 nm against the blank (water).The TPC was expressed in milligrams of gallic acid equivalents (GAE/100 g) and was an average value from three parallel measurements.

Total Flavonoid Content (TFC).
The total flavonoid content was determined colometrically according to the method described by Lin and Tang [14] with minor modifications.Five millilitres of a honey solution (0.5 g/mL) was mixed with 5 mL of 2% aluminium chloride (AlCl 3 ).A flavonoidaluminium complex was formed after 30 min of incubation time at room temperature.The formation of the complex was measured at 415 nm using a Hitachi U-2810 double beam spectrophotometer.Quercetin (0-100 mg/L) was used as a standard for obtaining the calibration curve.The TFC was calculated as the mean value of triplicate assays and expressed as mg of quercetin equivalents (QE) in 100 g of honey.

Determination of the Radical Scavenging Activity (DPPH).
The scavenging activity of the honey was determined according to a slightly modified procedure described by Turkmen et al. [15] against the radical 1,1-diphenyl-2-picrylhydrazyl (DPPH).All honey samples were prepared by dissolving 1 g of honey in 5 mL of distilled water.The samples were homogenized by shaking.Then, 1 mL of the aqueous solution from each sample was transferred into a centrifuge tube and centrifuged for 10 min at 20 ∘ C (10,000 rpm).Then, 0.25 mL of the solution was mixed with 0.75 mL of 0.1 mmol/L DPPH in methanol.A control test was performed using distilled water instead of the honey solution.The reaction mixtures were vortex-mixed and incubated for 1 hour at room temperature in darkness.After incubation, the mixtures were centrifuged for 10 min at 20 ∘ C (10,000 rpm), and the absorbance was measured at 517 nm against methanol using a Hitachi U-2810 double beam spectrophotometer.The antioxidant activity is expressed as a percentage of inhibition of the DPPH radical and calculated from the equation: where AA (%) is the percentage of disappearing DPPH radical; Absc is the absorbance of the control sample.

HPLC Analysis. HPLC analyses of the honey extracts
were performed using an Ultimate 3000 Dionex HPLC system with a photodiode array detector (PDA) and the Chromeleon 6.8 software (Dionex, Sunnyvale, CA, USA).Separations were carried out on a reversed-phase Gemini 5 m C-18 column (Merck, Darmstadt, Germany; 250 × 4.60 146 mm, particle size 5 m).The mobile phase consisted of 0.01 mol/dm 3 phosphate buffer at pH 2.5 (solvent A) and methanol (solvent B).A constant solvent flow rate of 1 mL/min was applied.The optimized gradient elution scheme is reported in Table 1.The temperature of the column oven was set at 30 ∘ C. The eluate was monitored at 214, 280 nm and 320, 340 nm because most natural phenolic compounds show their UV absorption maxima around these LOD, limit of detection at the signal three times at the baseline noise; LOQ, limit of quantification.
wavelengths.The comparison of the UV spectra and retention times with standard compounds allowed the identification of the phenolic acids and flavonoids presented in the analysed honey extracts.They were quantified using external standards.Parameters of calibration equations and validation characteristics of phenolic compounds are listed in Table 2.

Statistical Analysis.
All tests were repeated in triplicate, except for those applying the Folin-Ciocalteu method, which were repeated twice.The values obtained in experiments were expressed as the mean ± standard deviation (or relative standard deviation, RSD).The standard deviations and RSD were calculated using spreadsheet software (Excel5) [9].

Results and Discussion
Nine Belgian honey samples were purchased from 3 different beekeepers, which are indicated by the name of the neighbouring city, namely, Mortsel (M), Lichtaart (L), and Kasterlee (K).All samples were classified as polyfloral honeys.Honeys were preliminary characterized by determining their water content, electrical conductivity, specific rotation, and activity of invertase as well as by visual and organoleptic testing (Table 3).Samples of 3 successive years, 2007, 2008, and 2009, were studied.

Total Phenolic Content (TPC).
Phenolic compounds are an important group of compounds that influence the appearance and functional properties of honey.Thus, identifying and quantifying the total phenolic content are of significance, considering the initial assessment of the properties of honey.These substances can also be used as indicators of botanical and/or geographical origin and the source of honey [16].The total phenolic contents of the analysed honeys related to gallic acid (mg of GAE/100 g of honey;  2 = 0.9970) are presented  4).The total content of phenolic compounds was found to be relatively different among the honey samples collected in the consecutive years and were found to be year-dependent.The lowest content of TPC was observed in all samples collected during the production season of 2008, while the highest content was found in the honeys harvested in 2009 (Table 4).It is worth mentioning that the variation in TPC was not observed during honey storage.The quantified levels of the phenolic compounds in Belgian honeys were significantly higher than those reported by Socha et al. [13] and Wieczorek et al. [17] in Polish polyfloral honeys.Similar values of TPC were observed by Gašić et al. [18] in Serbian polyfloral honey, whereas significantly higher values were found in the case of honey from north western Spain [19] and in honey samples from an Atlantic European area [20].

Total Flavonoid Content (TFC).
The flavonoid content was determined by using a method that used aluminium chloride, which is based on the formation of a yellow complex between the aluminium ion, Al (III), and the carbonyl and hydroxyl groups of flavones and flavonols [21].Using a standard curve generated for quercetin ( 2 = 0.989), the total flavonoid content in the honey samples (mg of QE/100 g) varied from 3.11 to 5.12 mg, with a mean value of 4.08 ± 0.21 mg (Table 4).The highest TFC values were recorded in all of the samples of honey harvested in 2009.

Antioxidant Activity.
The standard DPPH method was used for the determination of the antioxidant activity.It is a rapid, easy, and widely used screening method for the quantification of antioxidants in foods, including honey [22].The radical scavenging activity of Belgian polyfloral honeys varied from 36.77% to 44.28% in the DPPH test (Table 4).Thus, their antioxidant activities are comparable to the values reported previously [13,23].
Furthermore, a significant correlation among the TPC and TFC values and the antioxidant activity was observed ( 2 = 0.951 and  2 = 0.872, respectively).The phenolic content of the honey samples is partially responsible for their antioxidant activity, which supports the relevance of this type of honey as an important source of antioxidant compounds and its possible use as a natural, therapeutically valued product.In addition, the values obtained by the DPPH studies have shown that this activity does not undergo significant changes between samples collected in subsequent years, which confirms that the activity depends on the composition of the phenolic fraction.Overall, our study confirms that all of the investigated polyfloral honey samples are substantially good sources of phenolic acids and flavonoids, which is reflected in their good antioxidant potential.

Profiles of Phenolic Compounds.
One of the emerging methods of differentiating between honeys of various origins is chemical profiling.Thus, the presence and relative level of a chosen set of phenolic acids in all of the samples have been determined by means of HPLC.Eighteen commercially available compounds were chosen, including 11 phenolic acids (3-hydroxybenzoic acid, 3,4-dihydroxybenzoic acid, 4hydroxybenzoic acid, caffeic acid, vanillic acid, syringic acid, p-coumaric acid, chlorogenic acid, rosmarinic acid, ferulic acid, and ellagic acid), six flavonoids (galangin, kaempferol, chrysin, quercetin, myricetin, and pinocembrin), and an isoprenoid, (±)-abscisic acid, which is a plant hormone that has been previously recognised as a marker of heather honey [9].The levels of the individual phenolic compounds in the analysed honeys are shown in Table 4.
It is obvious that the composition of phenolic acids in honey mainly results from the composition of plant nectar collected by bees.The most abundant compounds found in all studied samples are five acids, including ellagic, syringic, caffeic, vanillic, and chlorogenic acids.A comparison of the relative quantities of phenolic acids in the honey samples allowed the construction of the profiles of their phenolic compounds.As observed from Figure 1, these profiles are significantly similar to each other.Thus, a HPLC fingerprint of phenolic acids could be considered as one of the rational means to determine the identity of polyfloral honeys produced in the northern part of Belgium.Moreover, the examination of Figure 1 indicates that the similarities between these profiles are even more substantial when comparing honeys from different sources, taking into consideration the year of production.Thus, they also reflect the influence of  weather conditions on honey composition.Specifically, it is worth emphasizing that a high content of ellagic acid is found in all of the analysed honey samples.The relative amount of this component is between 47% and 88% among the 12 studied phenolic compounds (Figure 1).Therefore, ellagic acid might be considered to be a geographical origin marker of Belgian polyfloral honeys.Ellagic acid, a structural unit of ellagitannins, has been found in several types of honey.More specifically, it has only been proposed as a floral marker of heather honey to date [24,25].On the other hand, ellagic acid is a characteristic component in green tea and is also found in other natural sources, such as pomegranate, blueberries, blackberries, raspberries, strawberries, and walnuts.This compound has potential antimutagenic, antiviral, and antioxidative properties.Because it exhibits well-acknowledged biological activities, including antioxidant, anti-inflammatory, and anticancer properties, northern Belgian honeys, as they are a rich source of this acid, may be utilized as a prohealth food.
Less indicative are the profiles of flavonoids (Figure 2).However, it is worth mentioning that they are characterized by high levels of chrysin and pinocembrin, compounds that may be derived from propolis [26].We also found that the chemical profiles of the unidentified compounds observed in the HPLC chromatograms of the analysed Belgian polyfloral honey extracts are also very similar to each other and also might be useful as a part of their "HPLC fingerprint" (Figure 3).

Conclusions
Polyfloral honey samples from the Antwerp vicinity were studied to determine if their chemical profiling could be used for the identification of their geographical origin.Studies performed using the HPLC/PAD method indicated that the profiles of the chosen phenolic compounds do not vary considerably within samples from year to year.All of the tested honeys were characterized by a very high content of ellagic acid and a significant level of syringic, caffeic, and chlorogenic acids.The chemical profiling of flavonoids and unidentified components of the phenolic fraction of honeys, although less indicative, might also be useful as indicators of their geographical origin.Flavonoids and phenolic acids might be considered to significantly contribute to the total antioxidant activity in honey samples.Samples collected from the Antwerp province showed relatively high TPC and TFC values and are characterized by a good radical scavenging activity (DPPH).These activities correlate well with the total content of phenolic compounds.The obtained data are of particular interest in defining the effect of botanical origin in the biological activity of honey and to confirm its importance on the availability of phytochemistry compounds.Moreover, the high content of ellagic acid present in the northern Belgian polyfloral honeys may support its use as a beneficiary, prohealth food.

Table 4 :
The level of the individually identified phenolic compounds and abscisic acid (mg/100 g of product), the content of total phenolic compounds (GAE: gallic acid equivalent), and the content of total flavonoids (QE: quercetin equivalent) in polyfloral Belgian honeys.identified compounds of tested honeys [mg/100 g of product] (Av.,  = 3)

Table 1 :
Gradient elution profile for HPLC-PAD analysis.

Table 2 :
Calibration curves and detection limits of studied flavonoids and phenolic acids for HPLC-PAD analysis.

Table 4 .
It varied from 26.98 to 48.54 mg, with a mean of 36.56 ± 1.78 mg (Table