THE CHEMICAL COMPOSITION OF POLLEN, STAMINATE CATKINS, AND HONEY OF CASTANEA SATIVA MILL

The chemical composition of pollen and honey primarily depends on the botanical and geographical origin of the species, as well as other factors – climatic conditions, soil type, plant species, etc. The present study was to knowledge the biochemical profile of pollen, staminate catkins, and honey samples of Castanea sativa Mill. which were examined under conditions of Ukraine. Proteins are the major components of pollen and staminate catkins (169.0 ±1.60 g.kg-1 and 69.8 ±1.67 g.kg-1, respectively), while saccharides are predominant in honey samples (38.0 ±1.32 g.kg-1 fructose, 32.5 ±0.68 glucose g.kg−1 and 6.1 ±0.06 g.kg-1 sucrose). Glutamic acid (13.30 g.kg-1), aspartic acid (13.05 g.kg-1), and proline (12.45 g.kg-1) were predominant nonessential amino acids in the chestnut pollen. The content of macro and microelements was found in the pollen and staminate catkins much higher than in the honey. All Castanea sativa samples are a very valuable source of potassium as the main mineral element contained in pollen (7400 mg.kg-1), staminate catkins (7760 mg.kg-1), and honey (981 mg.kg-1). Microelements such as manganese and iron prevailed in pollen (478 mg.kg-1 Mn and 461 mg.kg-1 Fe), and staminate catkins (247 mg.kg-1 Mn and 109 mg.kg-1 Fe), and heavy metals (Hg, Sr, Sn, Sb, Li) are present only in the pollen samples with the most abundant Sr (12.8 mg.kg-1) and Sn (1.9 mg.kg-1) content and can be used as indicator suggesting the environmental pollution status in the region. Regarding the vitamin content, vitamin C was the most represented in all samples. Obtained results indicate that chestnut is species with important constituents such as amino acids and vitamins, with low content of heavy metals and high content of biogenic elements that may be used in phytotherapy and phytopharmacology.


INTRODUCTION
Castanea Mill. (Chestnut), belonging to the family Fagaceae is distributed in over ten species in temperate climate regions, including Asia, Europe, and North America. Among chestnut trees, C. sativa, C. crenata, C. dentata, and C. mollissima are important species consumed as a crop. Chestnut trees are valuable not only as food but also as a honey source plant, are already classified as major honey plants in many countries. The inflorescence is composed either of male or female flowers, or staminate and pistillate catkins occur together. Only the male flowers produce nectar that is secreted on the surface of the dislocated at the base of filaments in staminate catkins, from which is possible to produce 27.2 kg of honey per 100 trees (Farkas and Zajácz, 2007). Chestnut pollen, bee pollen, and bee bread were also identified to contain functional and specific substances Major honey components are carbohydrates (mainly monosaccharides, fructose, glucose, and sucrose) and water. In small amounts, honey also contains minerals, proteins, organic acids, volatile components, phenolics, vitamins, and pigments. Although present in small amounts, these components are very important for honey characterization and nutritive properties (White, 1975). A much better parameter for botanical and geographical origin identification was the analysis of free amino acids in honey, which was reported by several authors (Hermosín, Chicón and Cabezudo, 2003;Cotte et al., 2004;Nozal et al., 2004;González Paramás et al., 2006). Pollen contains an average of 1500 mg.100g -1 of amino acids, which means approximately ten times more than the average concentration in honey (Davies, 1975). The major amino acid in honey is proline (80 -90% of the total amino acids) except lavender, thyme, and rosemary kinds of honey where the major amino acid is phenylalanine (Hermosín, Chicón and Cabezudo, 2003;Cotte et al., 2004). The presence of proline is of animal origin and the quantity of proline depends on the time spent by the nectar in the bee's crop (Petrov, 1974). The proline content was proposed by the International Honey Commission as a honey ripeness and possible sugar adulteration indicator, with a minimum value of 180 mg.kg -1 for fresh, genuine honey (Bogdanov et al., 1999). Besides proline, 26 amino acids (both essential and nonessential) were also identified in honey (Hermosín, Chicón and Cabezudo, 2003). Due to ecological-geographical conditions very important are not only differences in some compounds (biogenic elements, heavy metals), which can increase/decrease the quality of pollen or honey but also evaluation of pollen sizes ( The chemical composition of honey primarily depends on the botanical origin of species or cultivars. However, climatic conditions and/or geographical origin can also affect the chemical composition even within the same honey type (Anklam, 1998).
The aim of the present study was to knowledge the biochemical profile of pollen, staminate catkins, and honey samples of Castanea sativa cultivar introduced under conditions of Ukraine, such as determination of saccharides, carotenoids, vitamins, amino acids, minerals (macro-and microelements) and heavy metals were examined. Comparative data of the mineral and amino acid composition of pollen and staminate catkins from different ecological-geographical zones little reflected in the literature.

Scientific hypothesis
Castanea sativa is a major honey plant in many countries is a potential source of phytochemical compounds used in food and health products against less-researched and lessused staminate catkins and chestnut pollen. The scientific hypothesis of this study was to examine the pollen, staminate catkins, and honey samples of Castanea sativa species and compared them due to their phytochemical, amino acid, and mineral profiles as appropriate substances of phytotherapeutic and phytopharmacology preparations.

Biological Material
We examined the biological material of C. sativa cultivar (staminate catkins, chestnut dried pollen, and monofloral chestnut honey) introduced in M. M. Gryshko National Botanical Garden of Ukraine in Kyiv (Figure 1 and Figure  2). Staminate catkins and pollen were harvested in June 2020 in dry conditions and with no precipitation. Samples were obtained from one cultivar of three trees.

Instruments
HPLC system with an ELSD detector (Agilent Technologies 1260 Infinity, Santa Clara, CA, USA).
HPLC system with ninhydrin and a VIS detector (Model AAA-400 amino acid analyzer, Ingos, Czech Republic).

Determination of dry matter, ash, and protein content
Total dry matter, ash, and protein content were determined according to the EN method (ČSN EN 12145, 1997). Total lipid content was determined according to methods specified in the ISO method (ISO 659, 1998).

Determination of saccharides
For the determination of saccharides, 1 g of sample was extracted with 10 mL of extraction solution (ultrapure water and ethanol mixed in ration 4:1) in a 50 mL centrifugation tube placed on a vertical shake table (GFL, Germany). After 1 h of extraction, samples were centrifuged for 4 min at 6000 rpm in a centrifuge (EBA 21, Hettich, Germany); the supernatant was filtered using a filter with 0.45 µm pore size (Labicom, Czech Republic) and filled up to 50 mL in a volumetric flask with ultrapure water. An Agilent Infinity 1260 liquid chromatography (Agilent Technologies, USA) equipped with an ELSD detector was used for the determination of saccharides. A Prevail Carbohydrates ES column (250/4.6 mm) was used as a stationary phase and acetonitrile (VWR) mixed with water in a 75:25 volume ratio was used as the mobile phase.

Determination of carotenoid
Total carotenoid content expressed as beta-carotene was analyzed at a wavelength of 445 nm spectrophotometrically (VIS spectrophotometer UV Jenway Model 6405 UV/VIS). Sample (1 g) was disrupted with sea sand and extracted with acetone until complete discoloration. Petroleum-ether was added and then water, in purpose to the separation of phases. After the separation, the petroleum ether-carotenoid phase was obtained and the absorbance was measured (ČSN 560053, 1986).

Determination of mineral contents
Sample for elemental analysis was prepared using the wet ashing method in a microwave oven (Milestone 1200, Milestone, Italy). A total of 0.25 g sample matrix was decomposed in a mixture of nitric acid (6 mL) (Analytika Praha Ltd, Czech Republic) and hydrochloric acid (2 mL) (Analytika Praha Ltd, Czech Republic).
After the decomposition sample was filtered using a filter with 0.45 µm pore size and filled up to 25 mL in a volumetric flask with ultrapure water. Elemental analysis was performed using ICP-OES (Ultima 2, Horiba Scientific, France) according to the procedure described by

Determination of amino acids
Amino acids were determined by ion-exchange liquid chromatography (Model AAA-400 amino acid analyzer, Ingos, Czech Republic) using post-column derivatization with ninhydrin and a VIS detector. A glass column (inner diameter 3.7 mm, length 350 mm) was filled manually with a strong cation exchanger in the LG ANB sodium cycle (Laboratory of Spolchemie) with average particles size 12 µM and 8% porosity. The column was tempered within the range of 35 to 95 °C. The elution of the studied  amino acids took place at a column temperature set to 74 °C. A double-channel VIS detector with the inner cell volume of 5 µL was set to two wavelengths: 440 and 570 nm. A solution of ninhydrin (Ingos, Czech Republic) was prepared in 75% v/v methyl cellosolve (Ingos, Czech Republic) and in 2% v/v 4 M acetic buffer (pH 5.5). Tin chloride (SnCl2) was used as a reducing agent. The prepared solution of ninhydrin was stored in an inert atmosphere (N2) in darkness at 4 °C. The flow rate was 0.25 (mL.min -1 ) and the reactor temperature was 120 °C.

Determination of vitamins Determination of Vitamin C concentration
Vitamin C concentration was determined according to the method described by Vikram, Ramesh and Prapulla (2005). Precisely 1 g of the sample was mixed with metaphosphoric acid and centrifuged at 10,000 rpm for 10 min. at 4 °C in a refrigerated centrifuge (Model H-2000C). The supernatant was sieved through Whatman No 4 filter paper. The filtrate was diluted with 1 mL of 0.8% metaphosphoric acid and filtered with a 0.450 Millipore filter and 20 μL of the sample injected into the HPLC (Shimadzu 20 A-Series, with column -NH2 -LUNA-100A (250X4.6mm), Diameter -5uLand Refractive Index Detector.

Determination of Vitamin A content
The method described by Okwu (2004) was used to determine vitamin A. Sample (1 g) was weighed into a beaker and macerated with a 10 mL mixture of acetone and n-hexane (1:1) and filtered. Thereafter, 10 mL of 50% (NH4)2SO4 solution was added, vigorously shaken, and allowed to settle. The upper layer was collected, and the absorbance read in Spectrophotometer (Spectro21D, Pec Medicals, USA) at 450 nm against hexane as blank betacarotene served as standard. Retinol equivalent was obtained by the conversion factor: 1 mg retinol equal to 12 mg of dietary beta-carotene.

Determination of Vitamin A content
Analysis Tocopherols and tocotrienols were analyzed by normal-phase HPLC using a Merck-Hitachi system (Merck, Darmstadt, Germany) with fluorescence detection (excitation: 292 nm, emission: 330 nm). Chromatographic separation was achieved within 45 min at 35 ±10 °C using a Eurospher-100 DIOL column (250 × 4.0 mm, 7 μm), preceded by a Eurospher-100 DIOL guard column (5×4.0 mm, 7 μm) (both from Knauer, Berlin, Germany). An isocratic mobile phase of n-hexane/MTBE (98:2, v/m) was used at a flow rate of 1.5 mL.min -1 . Individual tocochromanols were identified by comparing their retention times with those of external standards and quantified by 5-point calibration curve of externalstandards.

Sample
Preparation. Each sample was weighed and triturated into fine powder. The powdered sample was suspended in 50 mL of methanol, sonicated for 15 min, and dilute to 100 mL with methanol. 50 μL of each was for HPLC analysis.
Chromatographic Conditions Shimadzu LC system (LC 20A pump) equipped with UV/Visible detector was used for analysis. A single wavelength of 254 nm was used for simultaneous detection. A Promosil HPLC column C18 was used at 35 ºC. Methanol -0.05 M sodium acid phosphate (80 + 20, v/v) at pH 7.0 was used as mobile phase. The flow rate was adjusted to 0.8 mL.min -1 . 50 μL of sample and standard solutions were injected in a single injection.

Sample preparation
Freshly flowers (male catkins) were collected randomly from the three genotypes at the ballon stage (June 2020). Pollen samples released from male flowers (catkins) were further dried under laboratory conditions at 20 -25 °C and prepared for analysis. Honey samples A total of 5 honey samples were collected from Ukrainian beekeepers. The samples were stored at room temperature in the dark and analyzed within the month after extraction. To confirm the beekeeper's denomination, the samples were subjected to preliminary analyses. Uniflorality of samples was confirmed according to electrical conductivity. Number of samples analyzed: 18 Number of repeated analyses: 3 Number of experiment replication: 1

Statistical Analysis
Basic statistical analyses were performed using PAST 2.17. Data were presented as mean ±standard error of the mean (SEM) and differences between means were considered at p <0.05. The variability of all these parameters was evaluated using descriptive statistics.

RESULTS AND DISCUSSION
Bee pollen contains high amounts of carbohydrates, amino acids, fatty acids, minerals, and vitamins. The contents of these ingredients depend on the botanical origin of the pollen. The presence of these components shows that bee pollen can be used for human nutrition (Campos et al., 2008).
In honey, the content of proteins, including the enzymes, is relatively low and has a minor nutritive significance. On the other hand, the proteins, including the enzymes, are usually used as honey quality evaluation parameters. This is because protein content and enzyme activities vary regarding the botanical and geographical origin of the honey (Rodriguez-Otero et al., 1994; Flanjak   study presented that chestnut honey had the highest protein content (the mean was 94.8 mg.100g -1 ) followed by sage (the mean was 79.2 mg.100g -1 ), honeydew honey (the mean was 59.4 mg.100g -1 ). Black locust honey had the lowest protein content (the mean was 30.4 mg.100g −1 ). These differences are probably due to the different plant species and different environmental conditions. The protein content may be a bioactive indicator that contributes to the pharmacological activities of honey Three free sugars, namely, fructose, glucose, and sucrose, were detected in all samples, while the honey samples have the highest content of carbohydrates (fructose 38 g.kg -1 , glucose 32.5 g.kg -1 , and sucrose 6.1 g.kg -1 ). Saccharides represent the main components of honey, and many papers have been published for using sugars as an indication of adulteration but cannot be used to identify the floral or geographical origins of honey (Sivakesava and Irudayaraj, 2002). Kim et al. (2020) presented free sugar content and sugar content per catkin in floral nectar of various Castanea spp. Fructose was present in the interval from 61.5 (C. mollissima) to 74.8% (C. crenata), glucose in the range of 33.7 (C. dentata) to 20.5% (C. crenata), sucrose ranged from 3.2 (C. sativa) to 13.4% (C. dentata), free sugar content varied from 22.6 (C. crenata) to 82.5 µg.µL -1 (C. mollissima), sugar content per catkin was determined in the range from 0.31 (C. sativa) to 1.35 mg.catkin -1 (C. crenata) and nectar volume per catkin in the interval from 8.7 (C. dentata) -42.7 µL.catkin -1 (C. Regarding the vitamin content, vitamin C was the most represented in all samples in the following amounts in pollen 95 mg.kg -1 , staminate catkins 71 mg.kg -1 , and honey 55 mg.kg -1 .
The fat content predominates in pollen and staminate catkins against chestnut honey in amounts 55.4, 30.0, and 0.3 g.kg -1 , respectively.
The  Note: mean -arithmetic mean; SD -standard error of the mean. Figure 3 showed that eighteen (10 essential and 8 non-essential) amino acids were found in pollen and staminate catkins samples. Pollen had the greatest content of amino acids in comparison with staminate catkins and honey. Glutamic acid (13.30 g.kg -1 ), aspartic acid (13.05 g.kg -1 ), and proline (12.45 g.kg -1 ) were the predominant nonessential amino acids in pollen of C. sativa. Their content represents 38.41% of total amino acids. High amounts of leucine (11.44 g.kg -1 ) are also found in the pollen of C. sativa where its content is almost comparable to proline one.
The most abundant amino acid in staminate catkins was found to be aspartic acid (6.16 g.kg -1 ), slightly below glutamic acid (4.35 g.kg -1 ), which is 13.19 and 9.30% of the sum of total amino acid.

Essential amino acids Nonessential amino acids
The main amino acids found for honey were proline (0.59 g.kg -1 ), which is 22.52% of the sum of total amino acids. Lower but also important amounts of glutamic acid (0.34 g.kg -1 ), aspartic acid (0.30 g.kg -1 ), tyrosine (0.24 g.kg -1 ) and valine (0.21 g.kg -1 ) were present (Figure 4). Among the component methionine, tryptophan and cysteine are found to be present at the trace level (<0.01).
Macro-elements (Na, K, Ca, Mg, P) are the most represented group, dominated by potassium contained in pollen (7400 mg.kg -1 ), staminate catkins (7760 mg.kg -1 ) and honey (981 mg.kg -1 ). Potassium is the main mineral element with an average of one-third of the total. Some research has suggested that the trace element content in honey is mainly based on the botanical origin of honey, for instance, light blossom honey having a lower content than dark honey such as honeydew, chestnut, and heather Calcium is the 2 nd most abundant element in the staminate catkins (5940 mg.kg -1 ) and honey (170 mg.kg -1 ), while in pollen it is only in a small amount (0.27 mg.kg -1 ). Calcium was the prominent element in the honey sample in the amount of 170 mg.kg -1 . According to the data, the Ca contents of the honey sample in the present study was higher than the ones established in Serbia (20.1 mg.kg We can compare the percentages of individual components, which show that potassium was the most represented element in the honey (72%), pollen (53%), and staminate catkins (43%). Phosphorus was the second most abundant element in pollen (28%) of the total minerals, while calcium represented in staminate catkins 33% and the honey samples 13% of the total minerals. Microelements (Mn, Zn, Cu, Fe, Se, and Ni) are the second represented group of biogenic elements, where the content of manganese and iron prevailed in all samples. The content of manganese and iron predominate in pollen (478 mg.kg -1 Mn and 461 mg.kg -1 Fe) and staminate catkins (247 mg.kg -1 Mn and 109 mg.kg -1 Fe) than in the honey samples (13.6 mg.kg -1 Fe and 3.7 mg.kg -1 Mn).
Kızılpınar Temizer et al. (2018) determined heavy metal content in three unifloral honey samples of Castanea sativa from the Black Sea region in Turkey. Fe contents in honey samples were found in amounts from 6.280 mg.kg -1 to 20.515 mg.kg -1 and showed similarity to the other study in Turkey (Yücel and Sultanoğlu, 2013), and higher than performed in Portugal (0.70 -7.06 mg.kg -1 ) (Alves et al., 2013), in Serbia (0.58 -4.21 mg.kg -1 ) (Jevtić et al., 2012), and New Zealand (0.67 -3.39 mg.kg -1 ) (Vanhanen, Emmertz and Savage, 2011). Kızılpınar Temizer et al. (2018) determined Mn contents in amounts from 21.58 to 26.52 mg.kg -1 , which are higher compared to our honey sample. The mineral content in honey is also related to the geographical origin, therefore it is possible to be used to differentiate between honey of different geographical origins (Davies, 1975;Wang and Li, 2011). Fernandez-Torres et al. (2005) reported that the concentrations of Zn, Mn, Mg, and Na in eucalyptus, heather, orange blossom, and rosemary honey from Spain were strongly dependent on their botanical origin. Heavy metals (Hg, Sr, Sn, Sb, Li) are present only in the pollen samples with the most abundant of Sr (12.8 mg.kg -1 ), Sn (1.9 mg.kg -1 ), Hg (0.0027 mg.kg -1 ) content and other in the trace level (<0.05 mg.kg -1 ).
Heavy metals such as Cd, Pb, Cr, and Ni in honey can be used as indicators suggesting the environmental pollution status in the region (Porrini et al., 2003;Wang and Li, 2011). It has been reported that Pb contamination of honey in polluted and nonpolluted areas was not significantly different because of the high variability of the data, whereas the highest values were often found in polluted areas (Bogdanov, 2006).

CONCLUSION
The chemical composition, amino acid content, and minerals determined in pollen, staminate catkins, and honey obtained from the Castanea sativa cultivar were analyzed due to evaluate the representation of individual components in selected chestnut parts and the chestnut product. Total amino acids predominate in the pollen than in the staminate catkins and honey. Glutamic acid, aspartic acid, and proline were the most abundant. The fat content predominates in pollen and staminate catkins. The content of macro and microelements was found in the pollen and staminate catkins much higher than in the honey. All Castanea sativa samples are a very valuable source of potassium, which is necessary for the water and electrolyte balance. Vitamin analyses showed the presence of vitamin C in predominant quantity. Heavy metals present only in the pollen samples can be used as indicators suggesting the environmental pollution status in the region. These results showed significant amounts of biogenic substances in dried pollen, staminate catkins, and honey samples, which indicated that various plant parts are appropriate for phytotherapy and phytopharmacology.