Phenolic compound, nutritional and antioxidant profile of pollen collected by the genus melipona in North Eastern Brazil

The pollen collected by eusocial bees is often reported as being healthy food due to its important nutritional and therapeutic properties. However, studies reporting such properties are rare, especially for pollen collected by the genus Melipona in northeastern Brazil, which is the focus of this research. Pollen from seven species of stingless bees was analysed for its nutritional composition (sugar, lipid, protein and amino acids). The phenolic compound profile was described based on fourteen phenolic compounds (apigenin, kaempferol, luteolin, naringin, rutin, gallic acid, ferulic acid, caffeic acid, p-coumaric acid, chlorogenic acid, abscisic acid, protocatechuic acid, vanillic acid and trans-cinnamic acid). The antioxidant property was analysed by quantifying of total phenolic compounds, total flavonoids and DPPH. Chromatographic methods were used to identify and quantify the phenolic compounds and amino acids. The pollen samples from the bees under study showed good concentrations of proteins and amino acids and good antioxidant potential. The phenolic compounds luteolin, trans-cinnamic acid and apigenin were identified and described in pollen for the first time. Of the amino acids analysed, asparigine, glutamic acid, leucine and proline showed the highest concentrations. The research related to the theme showed that this is one of the first studies to identify and quantify the phenolic compounds and amino acids in stingless bee pollen, reflecting its importance in therapeutic use and as a food supplement.


Introduction
Pollen is collected by worker bees during foraging, and later mixed with nectar and substances released from their hypo-pharyngeal glands. Bees of the genus Apis store pollen (also known as bee pollen) in the brood cells, which serve as a source of food for all the individuals (Almeida-Muradian et al., 2005;Sattler et al., 2015). Several studies have reported the nutritional and chemical composition of pollen from bees of the genus Apis, characterizing the product as high in protein, minerals, carbohydrates, lipids, vitamins A, B, C, D and E, and also rich in phenolic compounds, flavonoids and phytosterols (Bogdanov, 2014). Thus bee pollen possesses several beneficial properties such as its antioxidant potential, and its anti-inflammatory, antibiotic, immunomodulator and antimutagenic properties (Graikou et al., 2011;Pascoal et al., 2014).
Pollen collected by species of the genus Melipona is deposited and stored in beehive pots where it undergoes fermentation, resulting in decreases in the pH and oxygen tension, after which it becomes known as Samburá or Sabura in Brazil (Freitas, 2003). The characterization of this type of pollen is difficult, since there are large numbers of native stingless bee species and huge variations in the pollen types (Silva et al., 2006;Bogdanov, 2014). Also, studies evaluating the chemical and nutritional composition of the pollen from these native bee species of Brazil are only incipient. The few manuscripts published have reported the sugar, lipid, moisture and hydroxymethylfurfural contents, the hydrogen ion potential and some other physicochemical characteristics (Alves et al., 2005;Silva et al., 2006), properties that assess the quality and shelf life of the product. However, studies that deal specifically with the identification of the phenolic compounds and amino acid are rare (Silva et al., 2006;Silva et al., 2013).
This study presents the phenolic compound profile, nutritional values (sugars, lipids, proteins and amino acids) and the antioxidant potential (total phenolic compounds, total flavonoids and DPPH) of twenty one pollen samples from seven different stingless bee species from north eastern Brazil, which can contribute to several subsequent studies searching for new functional and therapeutic food products. To the authors' knowledge, this is the first study that identifies and quantifies phenolic compounds and amino acids (essential and non-essential) in the pollen of bees of the genus Melipona. 2 Material and methods
The pollen samples were collected in 50 ml sterile falcon tubes and kept at -20 ° C for 96 hours. They were then freeze-dried for 24 hours using a CHRIST ALPHA-1-4LSC freeze dryer and stored in a desiccator to prevent moisture absorption by the samples.

Sugars and lipids
The sugars were broken down by acid hydrolysis and the reducing sugars determined using Fehling solution by the Lane and Eynon method (Association of Official Analytical Chemists,1984;Instituto Adolfo Lutz, 2008) and the lipids extracted and determined by the Soxhlet method (Instituto Adolfo Lutz, 2008).

Proteins and amino acids
The total protein content was quantified by the Kjeldahl method (American Association of Cereal Chemists, 1986;Yasuhara & Nokihara, 2001). To determine the total amino acid composition, 5-50 mg defatted solids were placed in 10 × 150 mm glass vials pre-pyrolyzed at 400 °C for 8 hours. 500 μl twice distilled 6N HCl was then added to each glass vial at 104 °C and the vials subjected to vacuum, sealed and maintained at 110 °C for 22 h. The HCl was then evaporated off in a rotary concentrator (SpeedVac AS160) and the hydrolysed samples resuspended in 1.0 ml MilliQ water. 5μL aliquots of the supernatant were transferred to another test tube, dried and subjected to pre-column derivatization with phenylisothiocyanate (PITC). The phenylthiocarbamide -amino acid derivative (PTC -aa) was separated on a C18 reverse phase column (Pico-Tag Waters, 3.9 × 150 mm) with monitoring at a wavelength of 254 nm. The amino acids in the sample were quantified based on their peak areas, comparing with those of fourteen amino acid standards (asparagine, glutamic acid, serine, glycine, histidine, arginine, threonine, alanine, proline, tyrosine, valine, methionine, cysteine and leucine) of known concentration (Bidlingmeyer et al., 1984).

Total phenolic compounds, total flavonoids and the DPPH analysis
The total phenolic compound content was estimated using the Folin-Ciocalteu spectrometric method (Beretta et al., 2005). The total flavonoid content was determined using a spectrophotometer according to the protocols described by Meda et al. (2005) and Ahn et al. (2007). The standard protocol for 2.2-diphenyl-2picrylhydrazyl (DPPH) was used to determine the antioxidant activity (Tominaga et al., 2005;Brand-Williams et al., 1995). The DPPH radical was quantified by spectrophotometry with absorption at 517 nm (Kulisic et al., 2006).

High Performance Liquid Chromatography (HPLC)
The phenolic compounds were extracted using 1g of pollen dissolved in 50 mL of methanol (Andrade et al., 1997). Due to the presence of sugars in the pollen, the samples were treated with Amberlite XAD-2 (Silva et al., 2013). The components in the samples were separated by HPLC using Prominence Shimadzu chromatograph (Kyoto, JAPAN) equipped with a diode array detector (SPDM2O), a Phenomenex Luna C18 reverse phase column (250 mm × 4.6 mm × 5 mm), a Phenomenex C18 pre-column (4 mm × 3 mm) and an oven (at 35 °C). The mobile phase consisted of a mixture of 1% aqueous formic acid (A) and methanol (B) with a flow rate of 1 mL/min and the following solvent gradient: 0-10 min 10% of B, 10-40 min 55% of B, 40-46 min 55% of B, 46-60 min 75% of B, 60-65 min 75% of B, 65-68 min 10% of B, and 68-70 min 10% of B. The injection volume was 20 μL. The phenolic compounds were identified based on their retention times and the UV/Vis spectra scan (245 nm to 370 nm) which was compared with analytical standards. Gallic, abscisic and trans-cinnamic acids were evaluated at 270 nm; protocatechuic acid, vanillic acid, luteolin, naringin and apigenin at 290 nm; chlorogenic, p-coumaric, and ferulic acids at 310nm; quercetin, caffeic acid, rutin and kaempferol at 370 nm.

Statistical analysis
All the analyses were carried out in triplicate. The means were calculated and the results presented as the mean ± standard deviation (SD). The differences between the compounds in each pollen sample were analysed using the analysis of variance followed by Tukey's test, with α = 0.05, using the Assistat software (Statistical Assistance beta version 7.6) and the PAST software 3.4 (Paleontological statistics package for education and data analysis) to carry out the multivariate cluster analysis and Principal Component Analysis (Hammer et al., 2001).

Sugars, lipids, proteins and amino acids
Normative instruction number 3 of January 19 th , 2001 (Brasil, 2001), of the Brazilian Health Surveillance Agency (ANVISA) states that bee pollen should present 145 g to 555 g kg -1 of sugar on a dry weight basis. As shown in Table 1, the total sugar contents varied from 0 to 27.2 ± 0.15 g kg -1 and hence the total sugar contents found in the bee pollen in the present study were lower than those referenced above. The low sugar contents found in the pollen samples from Melipona could be due to their consumption by microorganisms during fermentation.
The lipid content varied between 18.8 ± 0.8 and 47.9 ± 1.0 g kg -1 , values higher than those established by Normative Instruction 3, of January 19 th , 2001 of ANVISA (Brasil, 2001), which determined a maximum of 18 g kg -1 of lipids on a dry weight basis for bee pollen. The lipid contents found in the present study were lower than those found by Souza et al. (2004), which ranged from 70 to 300 g kg -1 in stingless bee pollen from the Amazon.
It should be mentioned that the amounts of sugar, lipid and many other nutritional compounds, established by Brazilian law, refer to bee pollen collected by the genus Apis and there are no nutritional values to use as a reference for Melipona, highlighting the need to establish a legal framework for products generated by this genus of stingless bees. The reduced amount of sugar decreases the energetic value of Melipona pollen, but the value is compensated by an increased lipid concentration, which is an important source of nutritional energy, due to innumerable functions lipids carry out in the human organism (Philippi, 2008).
The total protein contents ranged from 199.2 ± 8.1 to 444.1 ± 9.1 g kg -1 , above those referenced by the Brazilian legislation (Normative Instruction 3, of January 19 th , 2001 of ANVISA; Brasil, 2001), which determined a minimum of 80g kg -1 of protein on a dry weight basis. Since all the pollen samples analysed showed values above the minimum required by the legislation, it can be said that Melipona pollen is a proteinrich product. The pollen produced by Melipona compressipes showed the highest protein concentration, and should be further studied for its use as a food supplement. The variation amongst the pollen samples with respect to the concentrations of nutritional components can be explained by the seven different species of bees, the multiple floral origin, and other biological, ecological and geographical factors related to their collection, and also to the handling, storage and freeze-drying of the final product (Ratti, 2001;Orzáez Villanueva et al., 2002). The amino acids (6.60 ± 0.03 and 10.40 ± 0.02 mg g -1 protein) identified and quantified by HPLC are shown in Tables 1 and 2. Asparigine was found in the highest concentration followed by glutamic acid, proline and leucine. Their concentrations are significant according to the Food and Agriculture Organization (1985). All the pollen samples analysed in this study could be used as an alternative source of amino acids for human consumption, however further studies are required with respect to the quantity of pollen produced, its economic viability and other quality parameters established for food.

Total phenolic compounds, total flavonoids and DPPH
The total phenolic compound contents found in the pollen samples varied between 109.77 ± 1.25 and 1715.20 ± 1.13 mg GAE g -1 (gallic acid equivalents), values much higher than those described for bee pollen (Morais et al., 2011). The total flavonoid contents varied between 152.59 ± 1.07 and 14.57 ± 3.12 mg QE g -1 (Quercetin equivalents) as shown in Table 1, values different from those described by Silva et al. (2006) in their study with M. subnitida. The DPPH values of the phenolic extracts varied between 12.38 ± 1.05 and 55.54 ± 0.85 µg mL -1 , as also shown in Table 1. The values obtained in the DPPH analysis and the total phenolic compound and flavonoid contents referenced in this study confirm the high antioxidant potential of all the pollen samples analysed. The antioxidant potential presented confirmed the beneficial health potential of all the pollen samples under study (Pérez et al., 2015;Silva et al., 2013;Silva et al., 2014).
According to Morais et al. (2011), the antioxidant potential can vary in some foods, including bee pollen, depending on the chemical structure of the compounds, the reactivity with other antioxidants, the chelation potential with Fe and Cu metals, and also their botanical origin (Lachman et al., 2010).

Phenolic compound profile
Fourteen different phenolic compounds with known health benefits were identified and quantified, as shown in Table 3. Thirteen of these (except naringin) were identified and quantified for the first time in pollen samples. The flavonoids luteolin and apigenin were found in the highest concentrations, being distinguished from the other flavonoids found. The phenolic compounds trans-cinnamic acid, abscisic acid and p-coumaric acid, stood out in relation to the other phenolic compounds found, as shown in Table 3 and figure 1. Note that trans-cinnamic acid was the most prominent compound in the pollen samples from Melipona q. anthidioides, differentiating these samples from the others. It is also noteworthy that the pollen samples from Melipona q. quadrifasciata presented high concentrations of the flavonoid luteolin ( Figure 1). The other phenolic compounds were common in one or other samples of the bee species, in varying concentrations.
Abscisic acid, apigenin and kaempferol were found as the main components in the pollen samples from Melipona asilvai; luteoline, gallic acid and trans-cinnamic acid in M. q. anthidioides; luteoline and abscisic acid in M. q. quadrifasciata; galic acid, p-coumaric acid, abscisic acid and trans-cinnamic acid in Melipona mandacaia; apigenin in Melipona scutellaris and Melipona compressipes; and p-coumaric acid and abscisic acid in Melipona subnitida. Table 3 also shows that each pollen sample belonging to a different Melipona species had its own phenolic compound profile, which was possibly related to the botanical origin of the samples.
The phenolic compounds identified and quantified determined that all the pollen samples from stingless bees studied here showed therapeutic and beneficial health properties, since these compounds are responsible for the antioxidant potential (independent of the compound found) (Basim et al., 2006;Abdella et al., 2009;Carvalho et al., 2009;Fatrcová-Šramková et al., 2013).  1. Principal components analysis of the phenolic compounds found in the different species of bees in this study. The trend lines represent the major phenolic compounds present in the pollen samples and the circles illustrate the species in which these compounds are most strongly associated.