Fatty Acid and Proximate Composition of Bee Bread

Pollen an d nectar are essential components of honeybee, Apis mellifera L., diet. Nectar provides carbohydrates, while pollen supplies protein, lipid and vitamins. Pollen collected by foraging worker bees is combined with honeybee secretions (1). Bee bread is processed pollen stored and packed in the honeycomb cells following the addition of various enzymes and nectar or honey as it undergoes lactic acid fermentation. Generally, the methods employed for quantifi cation of nutritional dissimilarities amongst the levels of hive-stored and collected pollen have been proven diffi cult. There is a limited number of studies in the literature regarding the nutritional properties att ributed to the stored pollen. The reported results are contradictory, indicating either no signifi cant change or marginally increased nutrition (2). Bee pollen collection is a fairly new development. The pollen trap is used to scrape off the pollen from the legs of bees as they enter the hive. The scientifi c studies revealed various benefi cial therapeutic and nutritional properties of the bee pollen and enabled the scientists to identify its antimicrobial, antioxidant, antiradical, anticancer, and anti-infl ammatory activities (3). The main constituents of the bee pollen are carbohydrates (13–55 %), crude proteins (10–40 %), crude fi bre (0.3–20 %) and lipids (1–10 %) (4–6).


Introduction
Pollen an d nectar are essential components of honeybee, Apis mellifera L., diet. Nectar provides carbohydrates, while pollen supplies protein, lipid and vitamins. Pollen collected by foraging worker bees is combined with honeybee secretions (1). Bee bread is processed pollen stored and packed in the honeycomb cells following the addition of various enzymes and nectar or honey as it undergoes lactic acid fermentation. Generally, the methods employed for quantifi cation of nutritional dissimilarities amongst the levels of hive-stored and collected pollen have been proven diffi cult. There is a limited number of studies in the literature regarding the nutritional properties att ributed to the stored pollen. The reported results are contradictory, indicating either no signifi cant change or marginally increased nutrition (2). Bee pollen collection is a fairly new development. The pollen trap is used to scrape off the pollen from the legs of bees as they enter the hive. The scientifi c studies revealed various benefi cial therapeutic and nutritional properties of the bee pollen and enabled the scientists to identify its antimicrobial, antioxidant, antiradical, anticancer, and anti-infl ammatory activities (3). The main constituents of the bee pollen are carbohydrates (13-55 %), crude proteins (10-40 %), crude fi bre (0.3-20 %) and lipids (1-10 %) (4-6).
Fatt y acids are of high importance in fertility and health of the honeybees. Unsaturated FAs have also many benefi cial health eff ects such as reducing triglyceride (11) and cholesterol levels in blood and show anti-infl ammatory and antithrombotic activities (12). Current literature suggests that pollen and bee bread are good sources of polyunsaturated FAs (PUFAs) that are crucial for human nutrition. PUFAs cannot be synthesized in human body endogenously and must be obtained from food. In this respect, bee bread can be considered as a potential source of PUFAs in human diet. However, in particular, scientifi c research exploring various properties of bee bread is scarce and additional research into this topic is highly required. Therefore, the aim of the current study is to obtain and compare data on the FA content, pollen and proximate composition of bee bread samples harvested in Turkey.

Bee bread samples
Eight bee bread samples were obtained from apiaries located in diff erent monofl oral honey production regions in Turkey between June and October of 2014. The pooled samples were collected from minimum three beehives in apiaries with 50-100 colonies. Bee bread samples were hand collected from honeycombs and kept at -20 o C before the analyses. The type of fl ora and sampling locations were as follows: cott on from Adana and Urfa, citrus from Adana and Mersin, chestnut from Zonguldak, sunfl ower from Edirne and clover from Urfa and Adıyaman.

Pollen analysis
A mass of 10 g of bee bread sample was weighed into a centrifuge tube and mixed thoroughly with 20 mL of distilled water. The solution mixture was centrifuged at 1000×g for 10 min and the liquid part was discarded. The sediment was redissolved in 20 mL of distilled water and centrifuged. Then the sediment was placed on an absorbent paper to remove excess water, spread on a slide covering an area of about 20 mm and dried on a heating plate at 40 °C. The cover slips (22 mm×22 mm) containing a drop of glycerine jelly that liquefi ed by warming to 40 °C were warmed by a heating plate and then placed on the slide. The light microscope Eclipse E600 (Nikon, Tokyo, Japan) was used to visualize the pollen grain exine and shapes. Pollen grains were identifi ed using reference collection and the microphotographs from the literature.

Chemical analysis
Chemical analysis of bee bread (ash, crude fat and crude protein) was carried out using standard AOAC methods 920.153, 991.36 and 960.52, respectively (13)(14)(15). Moisture content was measured using a vacuum oven model VO200 (Memmert GmbH+Co. KG, Schwabach, Germany) at 60 °C and weighing until a constant mass. The ash content was measured gravimetrically aft er incineration at 550 °C and weighing. The total protein content was calculated by multiplying the nitrogen content by nitrogen to protein conversion factor of 6.25. All analyses were made in triplicate and the results were expressed in g per 100 g of fresh bee bread.
Determination of oil content was carried out using the ISO 659:2009 standard method (16). The bee bread samples were homogenized using a stainless steel blender (Waring, Atlanta, GA, USA). A mass of 2 g of sample was weighed accurately into a glass beaker and mixed with 100 mL of 4 M HCl. Then the content was heated at 100 °C and stirred for 15 min. Aft er cooling to room temperature the solution was washed three times with 25 mL of distilled water. The sample was fi ltered through a fi lter paper, which was dried at 105 °C in an oven for 1 h. The extraction of oil from bee bread samples was carried out with diethyl ether at 50 °C for 3 h by automated Soxhlet extractor (VELP Scientifi ca, Usmate (MB), Italy). The oil extracts were stored in amber vials prior to analysis of fatt y acids.

Fatt y acid analysis
Determination of FAMEs was performed using the ISO 12966-2:2011 standard method (17). Briefl y, 0.1 g of bee bread oil was weighed into a test tube. Aft er the addition of 5 mL of heptane and 0.5 mL of methanolic 2 M KOH, the tube content was mixed by vortexing for 1 min at room temperature. Then, the upper layer was dried with anhydrous sodium sulphate for gas chromatography analysis.
Chromatographic analysis was carried out by a gas chromatography (GC) system Clarus 500 (PerkinElmer, Shelton, CT, USA) equipped with an autosampler, split--splitless injector and a fl ame ionization detector. A 100-metre Supelco 2380 capillary column (Sigma-Aldrich, Belle-fonte, PA, USA) with an internal diameter of 0.25 mm and 0.2 μm fi lm thickness was used for chromatographic separation. Helium carrier gas fl ow rate was set at 1.2 mL/min. The injector and detector temperatures were set at 250 and 260 °C, respectively. The initial GC oven temperature was 165 °C, held for 5 min, increased to 240 °C at 5 °C/min and held at 240 °C for 10 min. A volume of 1.0 μL of sample was injected using the split injection mode (1:50). The peaks were identifi ed by comparison of their relative retention times with a standard FAME mixture. The results were expressed as percentage of total FAMEs.
The resulting FAMEs were also confi rmed by GC-MS through comparison of retention time and mass spectrometry data using the authentic reference standards. Confi rmation analyses of individual FAMEs were performed under identical conditions. Chromatographic separation of compounds was carried out using an Agilent GC-MS system (Agilent Technologies, Palo Alto, CA, USA) equipped with Agilent 6890 gas chromatograph and Agilent 5973 mass spectrometer.
Chromatographic separation of fatt y acids was achieved on a 30-metre DB-WAX capillary column (0.25 mm i.d., fi lm thickness 0.25 μm; Agilent Technologies, Folsom, CA, USA). The carrier gas (helium) fl ow rate was 1.5 mL/ min. The injection port temperature was set at 250 °C. The volume of the injected sample was 1 μL (split ratio 1:10). Initially, the GC oven temperature was maintained at 180 °C for 3 min. Then it was increased to 210 °C at a rate of 2 °C/min and aft er 20-minute isothermal run at 210 °C, finally increased to 240 °C at 10 °C/min and held for 5 min. The mass spectra were acquired in an electron-impact (EI) ionization mode at 70 eV in the mass scan range of m/ z=35-550. The temperatures of electron ionization source and mass quadrupole analyser were 150 and 280 °C, respectively.
The mass spectra of compounds were identifi ed by comparing the mass spectra obtained from their related chromatographic peaks with the Wiley and NIST mass spectral libraries (18,19).

Statistical analysis
All chemical assays were performed in triplicate. The obtained data were expressed as mean value±standard deviation. The data were compared using one-way analysis of variance (ANOVA) followed by least signifi cant dif-ference (LSD) test. Diff erences between the mean values at the 95 % confi dence interval (p<0.05) were considered statistically signifi cant.

Results and Discussion
Pollen content of the samples Botanical origin of the bee bread samples was identifi ed by pollen analysis. The results are presented in Table  1. All of the eight samples studied were unifl oral: cott on (two), clover (two), citrus (two), chestnut (one) and sunfl ower (one). Chestnut bee bread contained 94.4 % Castanea sativa pollen, while the clover bee bread samples con- Characterization of honeybee products such as honey, bee pollen and bee bread is important for consumers. Honey with a pollen frequency >45 % is considered to be monofl oral or unifl oral. In monofl oral honey, underrepresented pollen (e.g. citrus) frequency is minimum 10-20 % or 20-30 %, and overrepresented pollen (such as chestnut, eucalyptus) frequency is minimum 70-90 % (20). Similarly, the results of pollen analysis revealed that more than 45 % of total pollen detected in the bee bread samples was from monofl oral source. In particular, chestnut and clover samples had the monofl oral pollen contents of over 85 %. Castanea sativa is an important nectar and pollen source for a pollen forager bee (21) as it is abundantly available and easy to manage. The anemophilous plants like chestnut produce large quantities of small pollen grains and nectar foraging honeybees actively collect them mainly earlier in the day during the fl owering season to strengthen and improve the colony life (22,23). All the above information explains the preference of chestnut pollen by honeybees and high rate of Castanea sativa representation in bee bread.
Frequency of Trifolium spp. in clover bee bread samples from Adıyaman and Urfa was 85.6 and 86.2 % respectively. Trifolium species classifi ed under the Fabaceae family is among the most important pollen sources for honey-  (24). This could be a reasonable explanation for the higher Trifolium spp. pollen content in the tested clover bee bread samples.
Citrus, cott on and sunfl ower pollen is not represented as much as chestnut and clover. Cott on (Gossypium hirsutum L., Malvaceae) is a valuable plant from which fl oral and extrafl oral nectaries are collected by honeybees for honey production. However, honeybees seldom visit and obtain pollen from cott on plants (25,26). Cott on pollen is covered with sticky material (27) which makes its grooming from the body of bees painstakingly hard and it explains the avoidance of cott on pollen by honeybees (28). In another study, the repellency of cott on is att ributed to the gossypol, which is a dimeric sesquiterpenoid (25).
Sunfl ower bee bread contained the lowest pollen content from Helianthus annuus (45.4 %). This could be explained by the fact that although honeybees are the most frequent visitors of the sunfl owers, they rather collect nectar from sunfl owers and are less att racted to their pollen compared to other pollen types (29). Furthermore, it is reported that the protein content of sunfl ower pollen is low in both quality and quantity, so it is considered to be poor pollen source for honeybees (30).
Although citrus trees are considered as the most signifi cant fl oral source for the production of honey, they are rarely a good pollen source due to their low protein level. This could be one of the reasons for low representation of Citrus spp. pollen in the citrus bee bread samples (31).
Moreover, it is known that the nectar and pollen collecting behaviours of honeybees are diff erent. There are several factors which have to be taken into consideration regarding pollen collection by honeybees; the need of the colonies for pollen, brood production, the rhythm of the colony life throughout the season, biological value of bee pollen for honeybees, age of the foragers, handling time and factors related to pollen (size, colour, fl oral shape and symmetry, pigmentation patt erns, att ractiveness, etc.) (32).

Protein, fat, moisture and ash contents
The proximate compositions of the studied samples are given in Table 2. The moisture fractions of the samples were between 11.4 and 15.9 %. The mass fractions of ash were 1.9 to 2.5 %, the fat from 5.9 to 11.5 % and protein from 14.8 to 24.3 %.
The bee bread samples studied were obtained from regions with diff erent climatic conditions. For example, while the Mersin and Adana regions are under the infl uence of Mediterranean climate, continental climate is found in the other regions. Moreover, the sample collection points were at diff erent altitudes from the sea level. Therefore, changes determined in the moisture levels of the samples could be ascribed to the altitude and diff erent climatic conditions. Clover bee bread samples had the highest protein content (22.6 and 24.2 g/100 g) and cott on appeared to have the lowest content of protein (14.8 and 15 %). Clover bee bread also had the highest fat content along with citrus. Our results showed that the protein and lipid content varies according to the botanical origin of the bee bread. Herbert Jr and Shimanuki (33) reported similar fi ndings for the seven bee bread samples they studied but their data spread out over a wider range than ours. They found moisture content ranging between 18.8 and 28.0 %, protein content between 19.3 and 26.5 %, ash content between 2.1 and 3.2 %, and lipid content between 3.9 and 6.7 %.
A total of 35 and 32 fatt y acids were identifi ed in cotton (Gossypium hirsutum L.) bee bread samples from Adana and Urfa, respectively. Undecanoic and tridecanoic acids were not detected in the cott on bee bread, while hexanoic, octanoic and pentadecanoic acids were only present in the cott on sample from Adana. The ratios of major FAs found in the cott on bee bread samples were The groups in the same column with diff erent lett ers in superscript are statistically diff erent (p<0.05) *geographical and botanical origin of the samples are given in Table 1  Table 3. Fatt y acid composition of bee bread samples Saturated fatt y acid (Z)-tetradec-9-enoic signifi cantly diff erent from each other. (9Z,12Z)-octadeca--9,12-dienoic, hexadecanoic and (Z)-octadec-9-enoic acids were detected in the cott on sample from Adana at a higher level than in that from Urfa. On the contrary, (9Z,12Z,15Z)--octadeca-9,12,15-trienoic acid content of the cott on sample from Urfa was extremely high (40.7 %) and in that from Adana it was found only in traces (0.17 %). The dominant fatt y acid in the cott on sample from Adana was (9Z,12Z)-octadeca-9,12-dienoic acid (36.9 %), while (9Z,12Z,15Z)-octadeca-9,12,15-trienoic acid was the most abundant in the cott on sample from Urfa (40.7 %), which also had the second highest unsaturated fatt y acid content (67.5 %), aft er the clover sample from Urfa (70.3 %). Total number of FAs identifi ed in the citrus bee bread samples from Adana and Mersin was 33 and 34, respectively. Citrus bee bread from Adana was the only sample containing higher mass fraction of saturated FAs (51.6 %) than unsaturated FAs (48.3 %). Hexadecanoic acid fraction was the highest (38.7 %) in these samples, while tricosanoic acid was present only in the citrus bee bread sample from Mersin at 5.6 %.
Each of the clover bee bread samples obtained from the Urfa and Adana provinces contained 31 FAs that are mostly unsaturated. Fatt y acid profi les of the samples were slightly diff erent from one another. Undecanoic and heptadecanoic acids were barely detected in the clover bee bread sample from Urfa, while two of the saturated fatt y acids, octanoic and tricosanoic acids, were found only in the clover bee bread sample from Adana.
Omega-3 and ω-6 polyunsaturated fatt y acids are both required for the body to function. Humans cannot synthesize them and therefore they must be obtained from the diet (36). Omega-3 fatt y acids provide many benefi cial eff ects such as anti-infl ammatory function and prevention of cardiovascular diseases. Omega-6 fatt y acids are also benefi cial to human health. However, they have opposite eff ects on infl ammatory response and cardiovascular health. Because they compete for the same enzymes to produce signalling molecules, they have opposing physiological functions. For example, while ω-6-derived molecules are proinfl ammatory, ω-3-derived signalling molecules are anti-infl ammatory. Furthermore, they compete to incorporate into cell membranes. Therefore, the balance of ω-6/ω-3 fatt y acids is important for human health. Modern Western diets have ω-6/ω-3 ratio of 15:1 or 20:1. It was concluded that while very high ω-6/ω-3 ratio promotes the pathogenesis of many diseases, a reduced ω-6/ω-3 ratio can prevent these diseases. In addition to the ratio 2:1, the ratio 3:1 suppressed infl ammation in patients with rheumatoid arthritis, and the ratio 5:1 had a benefi cial eff ect on asthma (37). Therefore, the optimal ratio may vary because chronic diseases are multigenic and multifunctional. Simopoulos (38) concluded in his review that a lower ratio of ω-6/ω-3 fatt y acids is more desirable for reducing the risk of many diseases.

Conclusion
The pollen content, fatt y acid composition, and chemical composition of bee bread samples from diff erent botanical origins vary. Preferred or readily available plants for the bees as pollen source are also present in the bee bread samples, whereas others can be found in smaller amounts as a result of selective low preference.
The total amount of unsaturated fatt y acids (FAs) is higher than the sum of saturated FAs found in all the samples except citrus sample from the Adana region. The results obtained in the current study confi rmed that the bee bread can be considered as a good source of unsaturated FAs. The fatt y acid content of bee bread is very important for the honeybees and PUFAs are essential for a healthy body development and productivity. However, unsaturated FAs are not essential just for the bees but also for the human nutrition. The unique results of this study can thus be used as a reference for research into the bee and also human health. The fi ndings can also provide a scientifi c basis for the nutritional value assessment of the bee bread, thereby making contribution to the food composition database.