Physicochemical characterization and antioxidant activity of Palestinian honey samples

Abstract Physicochemical characteristics, main minerals, and antioxidant activity were determined for Palestinian honey samples belonging to different floral and geographical origins. The features of the analyzed samples were within the established international standards for honey quality control. One clear exception was the hydroxymethylfurfural (HMF) of the Ziziphus sample purchased from the Jericho region, which is the lowest city in the word characterized by a hot desert climate. The observed HMF value was 81.86 ± 2.64 mg/kg being two folds the maximum allowed in honey samples (40 mg/kg). As a second objective of the present work, the parameters were divided into two groups with different discriminatory power. The assessed physicochemical parameters, and the antioxidant activities, specific to the botanical origin discrimination, were used to run the first PCA. A strong correlation could be seen between the bioactive compounds and the antioxidant activities despite the geographical origin of the samples. Thyme and Ziziphus samples were the best samples, while citrus sample presented the lowest activity. Regarding the geographical discrimination, Ash and mineral contents in addition to the electrical conductivity were used. The output PCA conserved high represent ability of the data in the two‐first components being 82.72% and 9.60%. A little discrimination between the samples produced in the north and those produced in the south of the country, but it was not perfect. The intervention of the botanical variability could be the reason.

a supersaturated sugar solution in combination with minerals, enzymes, vitamins, flavoring organic compounds, free amino acids, and numerous volatile compounds (Gorjanović et al., 2013;Kayode & Oyeyemi, 2014). The verity of its sources subject its composition to high variability, which require standardization procedures for customer's protection (Albaba, 2015). In addition to the floral origin, other factors may be determinant in the final quality of honey such as the geographical and climate characteristics as well as the processing and storage conditions (Aazza, Lyoussi, Antunes, & Miguel, 2013;Imtara, Elamine, & Lyoussi, 2018).
The sensorial, chemical, physical, and microbiological characteristics of honey determine its quality (Khalil et al., 2012). EC Directive 2001/110 has specified the criteria for ensuring honey quality (European Community, 2004), concerning mainly, the electrical conductivity, moisture content, reducing and non-reducing sugars, pH, free acidity, diastase activity, ash content, HMF, and protein content.
At the best of author's knowledge, no previous study aimed a detailed characterization of commercialized honey samples in Palestine. Therefore, the main aim of the present work was to illustrate the quality characteristics of honey samples purchased from different areas of Palestine. The samples belong to different botanical origin and were characterized using a panel of known physicochemical parameters. In addition, ABTS, DPPH, iron reducing ability, and phosphomolybdenum reducing ability were assessed for the estimation of honey antioxidant activities. The entire data were used to study the correlations between the evaluated parameters, and to run the principal component analysis (PCA) for the discrimination of honey samples. The results were compared to the established quality standards, and to the reported honey samples belonging to the same botanical origin when it is possible.

| MATERIAL AND ME THODS
Ten local Palestinian honey samples were purchased from beekeeper, stored at room temperature (22-24°C) in airtight plastic containers until analysis, and labeled based on the commercial descriptions (Table 1). Visually, no sample of honey showed signs of fermentation or granulation before the characterization. Each assay was performed in triplicate, and the results were expressed as means ± SD.

| pH, free acidity, moisture, electrical conductivity, ash and proline content
The standardized methods of the International Honey Commission (IHC) were followed to assess the mentioned parameters (Bogdanov, 2009).

| Colour and melanoidins content estimations
The color was determined with a spectrophotometer by reading the absorbance of honey aqueous solutions at 635 nm (50% W/V) (Naab, Tamame, & Caccavari, 2008). The obtained absorbance was used to estimate the color in mmPfund following the algorithm: mmPfund = −38.7 + 371.39 × absorbance.
Honey color was also determined spectrophotometrically by measuring the difference between two net absorbances at 560 and 720 nm. Melanoidins content was estimated based on the browning index (net absorbance at A450-A720) (Brudzynski & Miotto, 2011), and the results were expressed as absorption units (AU).

| Hydroxymethylfurfural
The HMF content was determined followed the spectrophotometric procedure described in (Elmer, 2015).

| Determination of mineral elements
A 5 ml of nitric acid 0.1 M were added to the ashes, and the mixture was stirred on a heating plate to almost complete dryness. Then, 10 ml of the same acid was added for the solubilization and made up to 25 ml with distilled water. The mineral components were determined by atomic absorption spectrometry (Silva, Videira, Monteiro, Valentão, & Andrade, 2009

| Estimation of total antioxidant capacity by phosphomolybdate assay (TAC)
The TAC was estimated by the phosphomolybdenum method according to the reported procedure (Prieto, Pineda, & Aguilar, 1999).
The assay is based on the reduction of Mo (VI)-Mo (V) by the honey solutions and subsequent formation of a green phosphate/Mo (V) complex in acid medium. Briefly, 25 μl of honey solution was combined with 1 ml of reagent solution (0.6 M sulfuric acid, 28-mM sodium phosphate and 4-mM ammonium molybdate). The tubes containing the reacting medium were capped and incubated in a boiling water bath at 95°C for 90 min. After cooling to room temperature, the absorbance of the solution was measured at 695 nm. The TAC of each sample was expressed as mg of ascorbic acid equivalent/g (mgAAE/g).

| Total polyphenolic content
The total polyphenolic content estimation was based on the Folin-Ciocalteu protocol (Singleton & Rossi, 1964). A volume of 100 μl of honey solution was mixed with the 0.5 ml of Folin-Ciocalteu phenol reagent (1:10 dilution with distilled water) and 400 μl of 0.7 M Na 2 CO 3 solution. The reaction mixture was incubated for 2 hr and in darkness; and the absorbance was measured at 760 nm. The total content of each sample was expressed as mg gallic acid equivalent/100 g (mg GAE/100 g).

| Total flavone and flavonol content
The evaluation of flavone and flavonol content was carried out as previously described (Miguel, Nunes, Dandlen, Cavaco, & Antunes, 2014). Briefly, a volume of 500 μl of honey dilution was mixed with the 500 μl of AlCl 3 (5%) and incubated for 1 hr at room temperature.
The absorbance of the resulting solution was measured at 420 nm.
The calibration curve was performed using quercetin dissolved in 96% ethanol with serial dilutions. Total flavone and flavonol content of each sample was expressed as the quercetin equivalent/100 g (QE/100 g).

| Statistical analysis
The statistical analysis were performed by ANOVA through the GraphPad Prism 6 program and using the Tukey's post hoc test at p < 0.05. Correlations between phenol and flavonoid contents and antioxidant activity were achieved by Pearson correlation coefficient (r) at a significance level of 99% (p < 0.01). The data preprocessing and the PCA were accomplished using MultBiplot64 running in MATLAB R2017a.

| Quality control analysis
The analyzed honey samples presented acidic pH values, between 3.66 ± 0.01 in S2 and 4.25 ± 0.01 in S4 (Table 2). Such values are within the range accepted for honey (Bogdanov, Ruoff, & Oddo, 2004), and were similar to those found in Algerian, Portuguese, and Morocco honeys (Aazza et al., 2013;Elamine et al., 2017;Khalil et al., 2012). The acid pH inhibits the growth of microorganisms (Terrab, Díez, & Heredia, 2002). The free acidity of honey can be explained by the presence of organic acids in equilibrium with lactones, esters, and some inorganic ions such as phosphate. A high acid value indicates the fermentation of sugars into organic acids (Abselami et al., 2018). None of these samples exceeded the permitted acidity limit (50 mEq/kg) indicating the absence of undesirable fermentation process (European Community, 2004). The maximum value was seen in sample S5 (32.67 mEq/kg), while sample S4 presented the minimal value (11.67 mEq/kg) ( Table 2). Note. Values in the same column followed by the same letter are not significant different (p < 0.05) by the Tukey's multiple range test.

TA B L E 2 Physicochemical characterization of the analyzed samples
The moisture content of a honey sample depends on the environmental conditions and the manipulation by the beekeepers, which explain its usual year to year variations (Acquarone, Buera, & Elizalde, 2007). The moisture of the studied honey samples was within the standards (not more than 20%) (Codex Alimentarius Commission, 2001;European Community, 2004), except the Ziziphus honey (S10) with a moisture values of 20.2%. This value is similar to the Moroccan Ziziphus honey (Aazza, Lyoussi, Antunes, & Miguel, 2014) and higher than the Sudanese and Algerian Ziziphus honeys (Idris, Mariod, & Hamad, 2011;Zerrouk, Seijo, Escuredo, & Rodríguez-Flores, 2018), which explain the governance of the environmental conditions on determining this parameter. 16.9% was the minimum value, seen in the case the sample S4. High moisture content allows the fermentation of honey by undesirable osmo-tolerant yeasts and thus the formation of ethyl alcohol and carbon dioxide.
In addition, ethyl alcohol can in turn oxidize to acetic acid and water giving a bitter taste to the honey (Chirife, Zamora, & Motto, 2006).
The Ash content of honey samples, determining the mineral richness and the resulting electrical conductivity are important parameters in determining the botanical origin of a honey sample (Aazza et al., 2013). In addition, the mentioned parameters serve as differentiating features between nectar and honeydew honeys (Louveaux, 1959). The Ash content of the analyzed samples was between 0.065 ± 0.01% (S2) and 0.208 ± 0.01% (S1), being below 0.6%, the determined threshold for honey samples (Codex Alimentarius Commission, 2001). The results of our study show that the electrical conductivity values of the honey samples vary between 261.2 ± 1.15 μS/cm in sample S4 (Citrus) and a maximum of 533.67 ± 3.06 μS/cm in sample S7 (hairy fleabane) ( Table 2). The electrical conductivity measures the ionizable organic and inorganic substances and is not suitable to surpass 800 μS/cm, from a quality control point of view (Codex Alimentarius Commission, 2001). The values were similar to other Palestinian honey samples (Imtara et al., 2018), published earlier by the same group for other purposes, and to other samples from different botanical and geographical origins (Aazza et al., 2013;Elamine et al., 2017;Imtara et al., 2018). Other criteria used to determine the nutritional value of honey, with direct relation with the ash content and electrical conductivity is the mineral content (  (Table 3). All values found in the samples were within the ranges reported for honeys from other study (Aazza et al., 2013;Fernández-Torres et al., 2005;Imtara et al., 2018). The mineral composition of honey samples is also a potential indicator of its geographical origin, as well as a biomarker of possible pollution by toxic metals (Alves, Ramos, Gonçalves, Bernardo, & Mendes, 2013;Pohl, 2009).
The correlation matrix of some analyzed physicochemical parameters and the mineral compositions are illustrated in Table 6. Ash content has a strong positive correlation with potassium (r = 0.708, p < 0.05), explaining the prevalence of potassium in all analyzed honey samples (Table 3). The same correlation was reported previously (Hazali et al., 2017). As the Ash content determines the electrical conductivity of honeys (Guler, Bakan, Nisbet, & Yavuz, 2007), a strong positive correlation was also seen between the potassium levels and the electrical conductivity (r = 0.847, p < 0.001).
Proline, an essential free amino acid used for quality control of honey samples (Paramás, Bárez, Marcos, García-Villanova, & Sánchez, 2006). Values below 180 mg/100 g may indicate the none ripeness of a honey sample and/or adulteration (Bogdanov et al., 1999). None of the analyzed samples presented less amount with the maximum proline content found in hairy fleabane sample (S7) (720.87 ± 5.18 mg/kg) coming from Qalqilya. This value was three folds higher than the minimum value seen in Rocky Mountain honey sample (S9) coming from Bethlehem (229.44 ± 3.24 mg/kg) ( As honey color is also governed by the polyphenolics and melanoidin content (Aazza et al., 2013(Aazza et al., , 2014

| Bioactive compounds and antioxidant activity
The results of this section are illustrated in Table 5 the highest value was obtained in thyme honey (S1) from Al-Khalil (70.73 ± 0.71 mg/100 g). This value is similar to that found in thyme honey from Morocco (Aazza et al., 2014). The highest content of flavones and flavanol was found in S10 honey with a value of 8.23 ± 0.59 mg QE/100 g, while a minimum value of 0.18 ± 0.04 mg QE/100 g was seen in samples S3 (Table 5). ****Correlation is significant at the P < 0.0001; ***Correlation is signification at the P < 0.001; **Correlation is significant at the P < 0.01;*Correlation is significant at the P < 0.05.

TA B L E 4 Pearson correlation coefficients among compounds and antioxidant activity
The ability of the analyzed samples to scavenge DPPH free radicals, expressed as IC 50 mg/ml, was also evaluated. The lowest IC 50 was seen in the case of samples S1 and S10, being, so, the most efficient samples regarding the DPPH free radicals scavenging. Their   (Table 4). Both r values were negative, but the significant level was reached only in the case of flavones and flavonol contents (r = −0.738; p < 0.05).
Similar results, and correlation behavior were obtained by other groups when analyzing honeys samples from different botanical and geographical origins (Bertoncelj, Doberšek, Jamnik, & Golob, 2007;Khalil et al., 2012). BHT was used as positive control with a very low IC 50 in comparison to honey samples (0.009 ± 0.0001 mg/ml).
Antioxidant activity was also assessed by the ABTS assay (Table 5), through which, we found that sample S1 was the most active presenting an IC 50 of 3.26 ± 0.20 mg/ml. This results concordat the ones of the DPPH assay, which also explain the negative correlation between the IC 50 of ABTS and the polyphenolic content (r = −0.619; p < 0.05). Honey sample S9 honey presented the highest IC 50 16.28 ± 1.25 mg/ml, being the less active sample. Trolox was used as positive control with IC 50 of 0.019 ± 0.003 mg/ml.
The reducing power of the studied honey samples is dosedependent. The results illustrated in Table 5 shows that the sample Sample S6 presented the highest total antioxidant activity with a value of 120.03 ± 1.59 mg AAE/g, while sample S10 honey had the lowest activity (83.98 ± 1.35 mg AAE/g honey).

| Multivariate analysis
To further understand the distribution of the analyzed samples, based on the assessed parameters, principal component analysis was used (PCA). PCA is known to be a good tool for information extraction from multivariate matrices and concentrate it in only few components (Bevilacqua, Bucci, Magrì, Magrì, & Nescatelli, 2013).
The scores of the obtained components are then used to plot the data in an interpretable way.
In the present work, the evaluated parameters were divided into two main groups. The first group was formed by all parameters except the mineral content and was used as matrix to extract the information resulting from the botanical origin effect. The purpose was to cluster the Palestinian samples by their similarities in terms of physicochemical properties and antioxidant features. A second one was formed by the contents of minerals, the ash content, and the electrical conductivity. It is well established that this group of parameters, and besides being influenced by the botanical origin, it may indicate the geographical origins. It is then important to illustrate if there is a finger print characterizing samples produced in a specific Palestinian region. The results of both PCAs were illustrated in Figure 1a and b, successively. Thyme (S1) and Ziziphus (S10) honey samples shared the features regarding the bioactive compounds, and color intensity, being so the most antioxidant samples. This feature is already reported for both botanical origins (Aazza et al., 2014). The variability of the secondary plants, which may be specific to Palestine, seems to do not be significantly influencing, and both botanical origins seem to be a good option when honey antioxidant ability is desired. Honey samples labeled as hairy fleabane, Multifloral, Cornflower and Rocky Mountain shared the lowest pH values and high proline content in comparison to the remaining samples.
Low pH value is a property that inhibits the growth of undesirable microbial entities. In addition, the authors of the present work reported that low pH values is favorable parameters when a synergetic affect with essential oils against microbial strains is targeted (Imtara et al., 2018).
Regarding the study of the geographical component in discriminating the analyzed honey samples, the given data (ash, mineral contents and electrical activity) was highly conserved in the first two principal components explaining 82.72% and 9.06%, respectively. Two main clusters could be distinguished regarding the first principal component. The first cluster was formed by S1, S5, S6, S7, and S10 and were characterized by high potassium content (the most abundant element among the assessed minerals [table]) and electrical conductivity. Among the mentioned honey samples, S6 was the only one harvested in the north part of Palestine. The second cluster was formed by S9 and S2 produced in the south of the country, and the remaining samples (S3, S4, and S8) provided from the north part. The five samples presented less potassium content and electrical conductivity and, relatively, less amounts of the remaining parameters.
The geographical clustering was not perfect, and exception could be seen. This may be due to the intervention of the botanic origin, as it is well documented to be also crucial in determining the mineral profile of honey samples (Karabagias et al., 2017). However, the discrimination of samples using the mineral profile, the ash content and the electrical conductivity was clearer than in the case of the parameters used in Figure 1a.

| CON CLUS ION
Except the high HMF content of the honey sample originating from Jericho, no abnormal feature could be highlighted about the analyzed Palestinian honey. As it is the lowest city above the sea level in the world, the resulting climate may be the reason of the HMF increase. Such a feature needs to be a central interest in a possible study with extended sampling to other botanical honey from the same region. This will discriminate the possible effect of the botanical source and highlight at which level extreme climate of the region affects the quality of the produced honey samples.

CO N FLI C T O F I NTE R E S T
The authors of the present work declare no conflicts of interest in relation to published information. The authors are responsible for the content and writing of the article.