Antioxidant properties, element contents and antimicrobial activities of bee pollen collected by Apis mellifera L. in Türkiye

INTRODUCTION

Türkiye has a particularly rich biodiversity, with a characteristic climate, topography, and geomorphology, and is divided into three phytogeographic regions: Euro-Siberian, Irano-Turanian, and Mediterranean. This biodiversity enables the beekeepers to perform their beekeeping activities with both traditional and modern methods. Many melissopalynological studies have stated that plant species belonging to the families Fabaceae, Asteraceae, Lamiaceae, Rosaceae and Brassicaceae are important for beekeeping (De Souza et al., 2019; Lau et al., 2019). In Türkiye, there are 71 genera and 1013 species of which 400 are endemic in the Fabaceae family; 140 genera and 1186 species (446 endemic) in the Asteraceae family; 45 genera and 574 species (256 endemic) in the Lamiaceae family; 37 genera and 297 species (58 endemic) in the Rosaceae family; and 606 species (226 endemic) in the Brassicaceae family (Davis, 1965-1985; Erik & Tarıkahya, 2004; Mutlu & Karakuş, 2015).

Botanical richness enables beekeepers to produce bee products with various chemical and physical properties. Since ancient times, bee products have been among the most valuable products, which are indispensable for people because of their nutritional and therapeutic properties. Pollen, as one of these bee products, contains many bioactive components (phenolic and flavonoid compounds), vitamins, minerals, and pigments, as well as primary metabolites such as carbohydrates, proteins, and lipids (Denisow & Denisow-Pietrzyk, 2016). These phenolic and flavonoid components vary directly with the botanical origin, and indirectly depend on factors such as climate and soil (Komosinska-Vassev et al., 2015; Temizer et al., 2018). Additionally, the differences in the antioxidant activity properties of pollen are closely related to the chemical structure of these components (Rzepecka-Stojko et al., 2015). Furthermore, many researchers claimed that pollen has high antibacterial (Basim et al., 2006; Özkalp & Özcan, 2010), antifungal (Özcan, 2004), anti-inflammatory (Di Paola-Naranjo et al., 2004) and antimutagenic (Pascoal et al., 2014) effects.

The food supply needed due to the rapid increase in the world population, the use of synthetic chemical pesticides, mineral fertilizers, growth regulators and hormones in agricultural activities is increasing and as a result, pollen is also affected by the increasing environmental pollution. Many studies have emphasized that pollens are used as a bioindicator in environmental pollution (Temizer et al., 2018; Aldgini et al., 2019). The substances that are required at a minimum level to sustain the survival of living things, especially plants, are called trace elements. When the amounts of these elements exceed the tolerable limit values, they are defined as toxic contaminants. Essential and toxic elements that cause adverse effects for human health in pollen have been investigated in many countries, such as Türkiye, Bulgaria, Poland, Brazil, Romania, and Jordan (Dinkov & Stratev, 2016; Roman et al., 2016; Temizer et al., 2018; Aldgini et al., 2019).

The production and nutritional value of bee products other than honey such as pollen are ignored, although beekeeping is an important and common activity around the world. This study investigates the botanical origin, total phenolic/flavonoid components, antioxidant, and antimicrobial activities of the pollen samples. The levels of elements in the pollen were also determined to evaluate the health risks and environmental pollution.

MATERIAL AND METHODS

Pollen sample analysis

Pollen samples were obtained at the end of the pollen flow season in 2018. In this study, each mixed pollen sample was randomly collected from different beekeepers. They were collected from four different locations: sample S1 from İspir-Erzurum, which has a semi-humid climate and is located in the Irano-Turanian phytogeographical region (40° 29’ 3.2424’’ N - 41° 0’ 11.0268’’ E); samples S2, S3, and S4 from Bingöl, which has a semi-humid climate and is located in the Irano Turanian phytogeographical region (38° 53’ 7.2564” N - 40° 29’ 53.8476” E); sample S5 from Bulancak-Giresun, which has a humid climate and is located in the Euro-Siberian phytogeographical region (40° 56’ 15.0396” N - 38° 13’ 55.3584” E); and samples S6, S7, S8, S9 and S10 from Fethiye-Muğla, which has semi-drought and humid climate and is located in the Mediterranean phytogeographical region (36° 37’ 34.04” N - 29° 6’ 33.23” E) (Davis, 1965-1985; TSMS, 2022).

All pollen samples (2 g) were dissolved in 50 mL of absolute ethanol and pollen slides were elaborated according to Wodehouse (1935). At least 500 pollen grains were counted on each slide to determine its composition (Freire et al., 2012).The description of pollens as form or type was conducted according to the method of Mateo & Bosch-Reig (1998). A Nikon Eclipse Ci model microscope was used for examinations, including taxon identification at x1000 magnification and counting at x400 magnification. Monofloral and heterofloral properties of pollen were classified according to Freitas et al. (2013).

Pollen extract preparation

Five grams of pollen sample were added to 60 mL of absolute ethanol and this suspension was stirred at room temperature for 24 h by using a magnetic stirrer. This extraction solution was then filtered through Whatman no: 4 filter paper and stored at 4 °C.

Total phenolics content (TPC) and total flavonoids content (TFC)

It has been reported that ethanol is the best solvent for pollen extraction for measuring TPC/TFC and antioxidant activities (Karkar et al., 2018). We used ethanol as a solvent in determining these parameters in the samples. TPC of pollen extracts was determined according to the Folin-Ciocalteu reagent and Slinkard & Singleton´s (1977) method. Absorbances of the compounds were measured at 760 nm. TPC of the samples was calculated using the gallic acid calibration curve, which was used as a standard (R² = 0.9995).

The TFC of pollen extracts was determined according to the aluminium chloride colorimetric method (Chung et al., 2002). The absorbances were measured using a spectrophotometer (Optizen Pop UV/Vis Single Beam) at 415 nm. TFC of the samples were calculated using catechin’s calibration curve, which was used as a standard (R² = 0.9979).

Determination of phenolic and flavonoid components

Fragmentation profiles for the identified phenolic compounds were performed in the Agilent 1260 High Performance Liquid Chromatography (HPLC) system, and these compounds were determined by matching standard peak retention times run under equal HPLC conditions. A reverse phase column, Inertsil ODS-2 GL Sciences Inc. 5 μm (4.6×250 mm) C18 and Shimadzu SPD-M10 Avp PDA detector (270 nm) was used during fragmentation in a Shimadzu Prominence HPLC (Liu et al., 1997). System flow was 1.5 mL/min and system temperature was 25ºC. The following segmented gradient elution was used: 0-6.5 min, 90% A; 6.5-7.5 min, 89% A; 7.5-9 min, 87% A; 9-15 min, 20% A; 15-17 min, 95% A; 17-35 min, B (Miura et al., 2002). The mobile phase includes A, which is 97.5 water: 2.5 phosphate (v:v), B is acetonitrile. The measurements were repeated at least twice.

Antioxidant capacity assays

— DPPH (2,2-diphenyl-1-picrylhydrazyl) radical scavenging activity. The DPPH radical scavenging activity of pollen extracts was determined according to Blois (1958) with minor modifications. Sample solutions (3.0 mL) were mixed with 0.1 mM DPPH (1.0 mL) prepared by adding absolute ethanol. The mixture was incubated for 30 min at room temperature in a dark environment and the absorbance was measured at 517 nm using a spectrophotometer. The free radical scavenging activities of the reaction mixtures were calculated using the absorbance values after 30 min. The decline in absorbance is an indicator of the high rate of free radical scavenging activity in the samples, which was expressed in SC₅₀ (μg/mL).

— ABTS (2,2’-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid)) radical scavenging activity. The ABTS radical was dissolved in 7 mM potassium persulfate solution (2:1) and incubated for 16 h at room temperature in a dark environment. The prepared solution was diluted using 5 mM phosphate buffer (pH 7.4) so that the absorbance value reached 0.700 ± 0.020. Pollen extract samples (20 mL) were homogenized with 2 mL of the solution, kept for 5 min and measured in the spectrophotometer (Gökce et al., 2019).

— Hydrogen peroxide scavenging activity (HPSA). The HPSA of pollen samples (using 0.04 M phosphate buffer, pH 7.4) was performed according to Ruch et al. (1989). The HPSA of the samples was expressed in SC₅₀ (μg/mL).

— Ferric reducing antioxidant power (FRAP). The FRAP of pollen extracts and standard antioxidant solutions prepared with absolute ethanol was determined according to Oyaizu (1986). The absorbance values of the samples were measured using a spectrophotometer at 700 nm. The increase in absorbance values in the reaction mixture is an indicator of the high level of reducing power in extracts and standard antioxidant substances (Eq. 1):

where As: extracts or standard materials absorbance values, Ac: control’s absorbance values.

— Ferrous ion chelating activity (FICA). The FICA of the solutions prepared with pollen extracts and standard antioxidant substances in absolute ethanol was performed according to Dinis et al. (1994). The absorbance values of the mixtures were measured at 562 nm. FICA of the extracts and standard antioxidant substances was calculated according to Eq. (2):

Analysis of elements

In this study, Mg (microelement), Fe, Zn, Mn, Ni and Cu (trace elements), Se, Cr and Co (ultra-trace elements), and Cd (toxic element) were examined by inductively coupled plasma mass spectrometry (ICP-MS) (Model Bruker 820-MS). The microwave digestion method was used for the preparation of pollen samples: 0.5 g of each sample was digested in CEM MARS for 5 min with HN03 (ultrapure) and in 2 mL of HCl (ultra-purification) for 15 min at 1600 Watt in 210 mL. These solutions, which were placed in a cooling sample lid falcon tube and completed to 50 mL with distilled water, were analysed by Bruker 820-MS ICP-MS by filtering through a membrane filter (0.45 µm). The calibration curve was obtained using a certified multi-element standard. An intermediate stock of 10 mg/L was prepared from the main stock solution and a calibration curve was plotted from standard stocks of 5, 10, 20, 50, 100, and 250 μg/L. Samples were prepared in triplicate and ten readings were taken on each parallel ICP-MS. The blank sample prepared in 1% HNO₃ solution was measured 20 times. The standard slope, LOD (limit of detection) and LOQ (limit of quantification) were determined three, three and ten times, respectively (Temizer et al., 2018).

Antimicrobial activity

Antimicrobial activities of pollen extracts were tested against Staphylococcus aureus (ATCC 29213), Escherichia coli (ATCC 25922), Bacillus cereus (ATCC 10876) and Candida albicans (ATCC 10231) strains.

The strains were stored in a medium containing 15% glycerol and kept at -80 °C. Microbial cultures were incubated in sterile Mueller-Hinton Broth (MHB) for 6 h at 37 °C. Each culture was then transferred to a fresh MHB and incubated at 37 °C for 24 h. A medium containing 0.5X McFarland solution (approx. a culture density of 1.5 × 10⁸ cells/mL) was prepared.

Antimicrobial activities of pollen extracts were studied using the agar diffusion test method (Rios et al., 1988). Pollen extracts were prepared at a concentration of 8.33 g/mL in ethanol. Ethanol was used as a negative control. Each strain was inoculated onto Muller-Hinton Agar (MHA) medium with sterile swabs. Each pollen extract solution (100 µL) and controls were dropped into 8 mm diameter wells opened in the medium. Petri plates were incubated at 37 °C for 24 h. After incubation, the diameter of the inhibition zone around each well was measured and recorded. Antimicrobial activity was expressed as the diameter of the regions of inhibition generated by the tested extract. Each test was run three times.

The minimum inhibitory concentration (MIC) was determined using the microplate method (Wiegand et al., 2008). Pollen extract (8.33 g/mL) was serially diluted to 50% with MHB medium and 50 µL of a microorganism culture solution was added to 12 wells of a 96-well microplate. The microplates were incubated at 37 °C for 24 h. The absorbance values of the samples after incubation were measured at 450 nm using an automated microplate reader (Multiskan FC, Thermo).

Estimated daily mineral intake

The intake of elements in the diet is considered 0.04 mg/day for adults and 0.02 mg/day for children (Zafeiraki et al., 2022). The equation below was used to get estimated daily intake (EDI):

where c refers to the element level in bee pollen; AFC is the amount of food consumption (mg/day), calculated as 0.04 mg for adult consumption and 0.02 mg for child consumption (Zafeiraki et al., 2022); and BW is the consumer´s mean body weight (70 kg for male adults, 60 kg for female adults, and 15 kg for children) (Azanu et al., 2018; Tutun et al., 2022; Zafeiraki et al., 2022).

Health risk assessment

This analysis investigated the potential health risks associated with consuming pollen as a food. As a result, only oral consumption was used to determine the THI (target hazard index) and SHI (sum of hazard index) evaluations; THI (Eq. 4) being used to describe the health hazards associated with just one chemical, and SHI (Eq. 5) the risk of exposure to two or more chemicals (Zafeiraki et al., 2022):

where CF is contact frequency (365 days/year); CD is contact duration (70 years for adults and 4 years for children); RfD (mg/kg·day) is the oral reference dose for which the health-based recommendation value; and TA is the mean lifetime (365 days/year 70 years). The RfD levels for Mn, Fe, Co, Ni, Cu, Zn, Cr, Cd are defined in Demir et al. (2020) and for Se in Singh et al. (2019). There is no noticeable health risk when the computed THI and SHI values (Eq. 5) are both lower than 1.

where N is the total number of distinct elements found in each sample of bee pollen.

Statistical analysis

Statistical evaluations were performed from several methodological perspectives to analyse the pollen samples in terms of their antioxidant activity, contents of the elements and phenolics they contain.

The linear relationships between element, antioxidant, and phenolic contents in the pollen samples were examined. The Pearson correlation coefficient was used for variables complying with the normality condition as tested by the Shapiro & Wilk (1965) test. For those not complying, the non-parametric Spearman’s Rank Correlation Coefficient (SRC) (Spearman, 1904) was used.

Pollen samples were submitted to complete-linkage clustering analysis to be classified in terms of element, phenolic, and antioxidant contents (Defays, 1977). The findings of the clustering analysis have been visualized in the form of a dendrogram.

PCA has been widely used as an exploratory tool for data analysis (Pearson, 1895; Hotelling, 1933). Principal component analysis (PCA) was also performed based on element, antioxidant, and phenolic contents to determine the patterns in the data structure. To determine the number of principal components, two different statistical perspectives such as eigenvalues and the ratio of explaining the total variance can be considered. In this study, both perspectives were evaluated to select components.

RESULTS AND DISCUSSION

Pollen spectrum

The floral origin of bee pollen samples (n = 10) was determined, and the taxa found are given in Table 1. Sample S2 was classified as monofloral, and the other samples as heterofloral because each taxon was represented by less than 90% (Table 1). In this study, although some pollen samples (S2-S3-S4 and S6-S7-S8-S9-S10) were collected from different beekeepers in the same locality, their profiles differed from each other. This can be explained by the rich plant diversity in Türkiye (over 11,500 taxa, near to all Europe) (Alimoğlu et al., 2021).

The Apiaceae family was the predominant type of pollen in S1; Rosaceae family was the predominant type of pollen in S2 and S3; Fabaceae family was predominant type of pollen in S4 (Trifolium genus) and S5 (Onobrychis genus). The Pistacia genus was dominant in S6, S7, and S10 samples collected in Muğla. Many studies have reported that these taxa are crucial for the survival of bees (Freire et al., 2012; Freitas et al., 2013; Temizer et al., 2019). This vegetation also affects the phenolic compound, antioxidant, and antimicrobial activities of bee pollen.

TPC and TFC

TPC, TFC and antioxidant activities of the pollens are given in Table 2 and the concentration of phenolic substances in Table 3. Rosmanic acid, p-coumaric acid, quercetin, apigenin and naringin were found in all pollen samples.

TPC and TFC of the pollen samples ranged 271.42-601.85 mg GAE/100 g and 23.53-34.50 mg CAE/100 g, respectively (Table 2). Özcan et al. (2019) collected pollen samples from different regions of Türkiye and determined the total content of phenolic substances in the pollen in 434.17-719.58 mg GAE/100 g, values quite compatible with the results of our study. De-Melo et al. (2018) analysed the pollen of different botanical origins in Brazil and reported total phenolic and flavonoid contents of 560-2970 mg GAE/100 g and 25.74-1630.39 mg CAE/100 g, respectively. These values were quite higher than our results. The total flavonoid contents of the pollen samples in Malaysia (15.28-31.80 mg CAE/100 g) were very similar to our results (Harif Fadzilah et al., 2017).

Hierarchical clustering analysis from phenolic contents of pollen samples

Pollen samples were classified by a hierarchical clustering method in terms of phenolic contents. The dendrogram indicated that the pollen samples can be split into three main clusters (Fig. S1 [suppl]). PCA was also performed (Fig. S2a [suppl]) to determine the patterns in the data structure based on phenolic contents. Fig. S2b [suppl] represents the score plot of samples for PC1 and PC2, where S2 significantly dissociates from all other samples in terms of phenolic contents, parallel to the clustering analysis results.

Antioxidant capacity assays

HPSA and DPPH radical scavenging activity

HPSA and DPPH radical scavenging activities of pollen samples are ranged 20.43-150.94 and 22.19-23.78 μg/mL, respectively (Table 2). HPSA values of pollen samples collected from different regions of Türkiye by Temizer et al. (2018) were in the range 25.56-30.28 mg/mL. The HPSA values were similar to this study except for S2. The level of p-coumaric acid determined in S2 was also lower than in the rest of samples. Therefore, HPSA is also considered to be caused by p-coumaric acid. S5 and S9 had a high HPSA activity, but these samples were found to have low p-coumaric acid levels. A high HPSA activity value may be related to the amount of gallic acid and apigenin. The DPPH radical scavenging activity of the pollen samples was higher than that reported by Harif Fadzilah et al. (2017) and Sun et al. (2017), but similar to that of the pollen samples collected from Brazil (Freire et al., 2012).

FRAP and FICA

The FICA and FRAP of the pollen samples varied between 74.84-91.79% and 97.26-99.83%, respectively (Table 3). Kao et al. (2011) and Sardar et al. (2014) found metal-chelating activities in pollen samples in the range 11.27-73.50% and 41.40-58.80%, respectively. Almost all pollen samples showed an effective reduction in power activity. Temizer et al. (2017, 2018) reported that the iron reduction power in pollen samples was 71.92-73.86% and 72.29-72.58%, respectively. When compared with our study, it was determined that our pollen samples showed effective chelating activity. Analysing the components of all samples, p-coumaric acid, rosmanic acid, quercetin, apigenin, and naringin were the substances with a more prominent role in metal-chelating activity. The ability of these components to exhibit ferrous reducing power activity usually depends on substituted groups and phenolic structures.

ABTS radical scavenging activity

ABTS radical scavenging activities ranged 6.50-78.40 µg/mL. Freire et al. (2012) reported ABTS radical scavenging activities of pollen samples in Brazil as 6.0-97.20 µg/mL, quite similar to our study. Sun et al. (2017) determined ABTS radical scavenging activities as 3.19-3.85 µg/mL, lower than the values in our study. In our study, ABTS radical scavenging activity of S1 was higher than that of the other pollen samples (Table 2). The fact that S1 contains more hesperidin and quercetin than the other phenolic substances indicate an important role for these substances in ABTS radical scavenging activity. The phenolic groups in both compounds may be closely related to the substituted groups at the o-, m- and p- positions in the aromatic ring.

Hierarchical clustering analysis from antioxidant contents of pollen samples

A hierarchical clustering method was used to classify different locations according to the antioxidant contents of pollen samples. Fig. S3 [suppl] shows a dendrogram which illustrates the relationships among all pollen samples from the perspective of antioxidant content based on the pre-determined measure of similarity. Fig. S4a [suppl] shows a map for the antioxidant contents (PCA, PC1 and PC2). Fig. S4a [suppl] indicates that the positive part of PC1 was related to TPC, ABTS, FRAP and TFC, while the positive part of PC2 was related to HPSA, TPC and ABTS. For both PCs, the negative parts were related with these elements. Fig. S4b [suppl] shows that S2 significantly dissociates from the rest of samples in terms of antioxidant contents, parallel to the clustering analysis results. These findings demonstrate that the antioxidant contents can be used to classify these pollen samples.

Relationships between antioxidant and phenolic contents of pollen samples

The linear relationships between the antioxidant and phenolic contents of the pollen samples are shown in Table S1 [suppl]. Chlorogenic acid had a significant positive linear correlation with hesperidin (r=0.91; p<0.001) and TFC (r=0.57; p=0.085), at 0.05 and 0.10 significance levels, respectively; p-coumaric acid, at 0.05 significance level, had a significant positive linear correlation with TPC (r=0.91; p<0.001) and a significant negative linear correlation with HPSA (r=-0.74; p=0.013); hesperidin and TFC had a significant positive linear correlation (r=0.65; p=0.042), at 0.05 significance level.

Analysis of elements

Table 4 shows the results of the heavy metals detected in the pollen samples: Mg > Fe > Zn > Mn > Cu > Se > Cr > Ni > Cd > Co (Table 4). Adaškevičiūtė et al. (2019) found in all bee pollen samples Mg (644-1009 mg/kg), Fe (45.04-76.40 mg/kg), Mn (15.34-66.49 mg/kg), Zn (20.04-31.99 mg/kg), Co (0.000-0.105 mg/kg) and Cu (0.00-6.05 mg/kg), but Cr was found in none of the samples. Tutun et al. (2022) detected in bee pollen samples from Türkiye were Fe (12.6-130 mg/kg), Ni (0.35-3.70 mg/kg) Zn (12.7-53.5 mg/kg), Mn (4.74-19.0 mg/kg), Cu (3.19-18.5 mg/kg), Co (<LOD-0.006 mg/kg) and Cr (0.05-0.91 mg/kg). Zafeiraki et al. (2022) detected in bee pollen samples from Greece Mg (892-6098 μg/g), Fe (78-1496 μg/g), Mn (13-324 μg/g), Ni (0.012-1.4 μg/g), Zn (40-451 μg/g), Cu (11-57 μg/g), Se (LOQ<-0.76 μg/g), Cd (0.03-1.1 μg/g). The Mg level obtained from this study was higher than in previous studies (Adaškevičiūtė et al., 2019; Zafeiraki et al., 2022). In our analysis, Fe values were found in a wide range (Table 4). These values were greater than in Adaškevičiūtė et al. (2019) and similar to those found by Zaferaki et al. (2022) and Tutun et al. (2022). The Zn levels detected were found to be much lower than those reported by Zafeiraki et al. (2022) and closely similar to those reported by Tutun et al. (2022). In this study, Mn values are similar to Tutun et al. (2022), and lower than those values reported by Zafeiraki et al. (2022) and Adaškevičiūtė et al. (2019). Cu levels in our samples were similar to the values reported by Adaškevičiūtė et al. (2019), but lower than those found by Tutun et al. (2022) and Zafeiraki et al. (2022). In this study, Co (0.006 mg/kg) was detected only in S6 and it is within the range reported by Adaškevičiūtė et al. (2019) and Tutun et al. (2022). Adaškevičiūtė et al. (2019) never detected Cr in samples, and the Cr level reported by Tutun et al. (2022) was low compared to our study. While the Cd level in our study was similar to that found by Zaferaki et al. (2022), the Se level was higher than the one reported by these authors. Tutun et al. (2022) detected Se and Cd below the detection limit in the pollen samples.

Correlations between element contents of pollen samples

The findings related to the relationships between elements are shown in Table S2 [suppl]. The linear correlation between Cr and Co (r=0.588; p=0.074), and between Fe and Co (r=-0.583; p=0.077) showed a significant value at 0.10 significance level, in a positive and negative way respectively. Moreover, both Cu (r=0.758; p=0.011) and Zn (r=0.745; p=0.013) had a significant positive linear correlation with Mn at 0.05 significance level.

Hierarchical clustering analysis of element contents in the pollen samples

Fig. S5 [suppl.] shows the results of the clustering analysis with a dendrogram, which illustrates the relationships among all pollen sample locations based on element contents according to Euclidean distance. The dendrogram explained that the pollen samples can be split into three main clusters. S4 dissociated from all other samples and forms the third cluster alone. Fig. S6a [suppl] illustrates the element map. The positive part of PC1 was related to Cu, Zn, Mn, Fe, and Se, while the positive part of PC2 was related to Co, Cr, Ni, Cd, Cu, Zn, Mn. For each PC, the negative parts were related with these elements. For each PC, the negative parts were related with these elements. Fig. 6Sb [suppl] represents the score plot of samples for PC1 and PC2 and shows that S4 significantly dissociated from all other samples for element contains, parallel to the clustering analysis results. Besides, for PC1 and PC2, the proximity of the positions of S6, S7, and S9 on the plot indicates that these samples had a similar structure in terms of their element content.

Antimicrobial activity

The antimicrobial activity of ten pollen samples was investigated for three bacteria and one yeast. Different levels of antimicrobial activity in pollen samples were determined against B. cereus ATCC 10876 and S. aureus ATCC 29213 (Table 5). Sample S9 showed the highest antibacterial activity against B. cereus (14.00 mm, MIC= 4.17 mg/mL), and S2 the highest against S. aureus (13.00 mm, MIC=8.33 g/mL). No antimicrobial activity was observed against E. coli and C. albicans in any of the pollen samples. S. aureus and B. cereus are common human pathogens that cause serious infections in humans (Bottone, 2010; Campanile et al., 2015). All pollen samples, except S3 and S4, showed antibacterial activity against S. aureus. When compared with Nikolaieva et al. (2019) and Kaškonienė et al. (2020), who investigated the antimicrobial activities of natural and fermented pollens, our values were very close to values of the natural pollens, whereas they were at a very low level compared to fermented pollens.

All pollen samples had the same characteristics and properties regarding MIC and inhibition zones of C. albicans, B. cereus and E. coli. When S. aureus is considered, all samples exhibited similar behaviour except for S3 and S4, which could not eliminate the bacteria. These findings can be statistically supported by a cluster analysis (Fig. S5 [suppl]), where S3 and S4 formed a cluster (70.97% similarity), and the other samples formed separate clusters.

EDI and health risk assessment

The risk to human health was assessed based on bee pollen intake, considering the RfD values together with EDI (Tables 6 and 7). In this study, the calculation of risk was made according to the worst-case scenarios, and it was determined that the THI values of the essential and toxic elements examined were not greater than 1, except for Se for men, women and children (Table 7). However, due to the high Se content, SHI values were found to be greater than 1 for men, women, and children. SHI values were greater than 1 due to Se, which is naturally found in the soil and is an essential element for humans. In addition, THI values of the toxic Cd element were found to be less than 1 for men, women, and children.

CONCLUSIONS

Türkiye has suitable ecological conditions for beekeeping activities because it is at the intersection of three phytogeographical regions and has its own topographic and climatic conditions, which cause the blooms of plants located in different geographic locations to occur at different periods of the year. The taxa detected in pollen samples were found to belong to many plants regardless of whether the flowering period was long or short in this study. The antioxidant activities of pollen samples were evaluated using methods with different principles. The chemical structure of phenolic compounds was effective on antioxidant activity, although the total phenol and flavonoid levels in the pollen did not directly affect the antioxidant activity in our study. Our results showed as well that the amount and combinations of these compounds were effective on the antioxidant and antimicrobial activities of pollen.

Pollen samples were found to be rich in the essential element (Se); however, all pollen samples had toxic Cd contamination. According to the health risk assessment, if adults consume daily 0.04 mg of pollen and children 0.02 mg, there may be a possibility of toxic effects due to Se. Pollens have potential bioactive properties, its content needs to be studied extensively to be used for cosmetic, pharmacy, and food industry.