Flavonoid Composition and Antitumor Activity of Bee Bread Collected in Northeast Portugal

Bee bread (BB) is a fermented mixture of plant pollen, honey, and bee saliva that worker bees use as food for larvae, and for young bees to produce royal jelly. In the present study, five BB samples, collected from Apis mellifera iberiensis hives located in different apiaries near Bragança, in the northeast region of Portugal, and one BB commercial sample were characterized by high-performance liquid chromatography coupled to a diode array detector and electrospray mass spectrometry (HPLC-DAD-ESI/MS) in terms of phenolic compounds, such as flavonoid glycoside derivatives. Furthermore, the samples were screened, using in vitro assays, against different human tumor cell lines, MCF-7 (breast adenocarcinoma), NCI-H460 (non-small cell lung cancer), HeLa (cervical carcinoma) and HepG2 (hepatocellular carcinoma), and also against non-tumor liver cells (porcine liver cells, PLP2). The main phenolic compounds found were flavonol derivatives, mainly quercetin, kaempferol, myricetin, isorhamnetin and herbacetrin glycoside derivatives. Thirty-two compounds were identified in the six BB samples, presenting BB1 and BB3 with the highest contents (6802 and 6480 µg/g extract, respectively) and the highest number of identified compounds. Two isorhamnetin glycoside derivatives, isrohamnetin-O-hexosyl-O-rutinoside and isorhamnetin-O-pentosyl-hexoside, were the most abundant compounds present in BB1; on the other hand, quercetin-3-O-rhamnoside was the most abundant flavonol in BB3. However, it was not possible to establish a correlation between the flavonoids and the observed low to moderate cytotoxicity (ranging from >400 to 68 µg/mL), in which HeLa and NCI-H460 cell lines were the most susceptible to the inhibition. To the authors’ knowledge, this is the first report characterizing glycosidic flavonoids in BB samples, contributing to the chemical knowledge of this less explored bee product.


Introduction
The nutritional requirements of honeybees, Apis mellifera, are met by the collection of pollen, nectar, and water. Nectar is the primary source of carbohydrates, while pollen provides proteins, lipids, vitamins and minerals [1]. Bee bread (BB) is a fermented mixture of plant pollen, honey, and bee saliva that worker bees use as food for larvae, and for young bees to produce royal jelly. Pollen collected by bees is mixed with a small amount of honey and saliva, and packed into the cells of the honeycomb where it undergoes a chemical change to form a product called bee bread [2]. This mixture undergoes different chemical processes due to the action of distinct enzymes from glandular secretions, microorganisms, moisture and temperature (35-36 • C chamber temperature offspring), allowing the transformation, improvement and preservation of the stored pollen, which is called bee bread after two weeks of initial storage [3,4].
Despite the role of BB as the main source of protein for the bees, its functional properties have been correlated, as well as its flavonoid content, with the BB's floral origin [4]. In particular, BB has demonstrated in vitro antibacterial [5,6], antioxidant [3], and antitumor [2,7] properties. For the last activity, ethanolic extracts were screened against tumor cell lines (human glioblastoma cell line U87MG) and the normal human astroglia cell line SVGp12 (CRL-08621) using in vitro assays [2,7].
The BB composition varies according to the origin of the pollen but is mainly composed of water, proteins, carbohydrates, lipids, inorganic elements and various other minor components such as decanoic acid, gamma globulin, nucleic acids, vitamins B and C, pantothenic acid, biopterin, neopterin, acetylcholine, and reproductive hormones, among others [8].
The quality information available on the literature for bee bread remains limited, with few reports on the phenolic composition of this mixture. Some phenolic compounds were previously identified in BB samples from Poland, Russia, Latvia and Georgia [9,10]. Other reports on BB samples from Spain and Poland mentioned only total phenolics measured by the Folin-Ciocalteu colorimetric assay [2,4] and did not provide detailed characterization in terms of individual phenolic compounds.
In the present study, five BB samples, collected from Apis mellifera iberiensis hives located in different apiaries near Bragança, in the northeast region of Portugal, and one sample of commercial BB were characterized by HPLC-DAD-ESI/MS in terms of their phenolic profile. Furthermore, the samples were screened against different human tumor cell lines, as well as against non-tumor liver cells.

Chromatographic Profile of the BB
The chromatographic profile of BB1 recorded at 370 nm can be observed in Figure 1; the peak characteristics, tentative identities and quantification of all the samples are presented in Table 1 and the  quantification results are presented in Table 2. The main phenolic compounds found in bee bread were flavonol derivatives, mainly quercetin, kaempferol, myricetin, isorhamnetin and herbacetrin glycosides. The phenolic composition of bee bread has hardly been explored, only having been reported by a few authors [2,9,10], but using different analytical approaches. Tavdidishvili et al. [10] used HPLC-UV-Vis to study Georgian bee bread samples, reporting the presence of three flavonoids, naringin, rutin and quercetin. Isidorou et al. [9], using GC-MS, identified four phenolic acids (4-hydroxybenzoic, p-coumaric, ferulic and caffeic acids) and six flavonoids (chrysin, naringenin, kaempferol, isorhamnetin, apigenin and quercetin), whereas Markiewicz-Zukowska et al. [2], also using GC-MS, reported the presence of just two flavonoids, kaempferol and apigenin.
In our samples, up to 32 different flavonoids were detected (Table 1). Myricetin-3-O-glucoside (peak 7), quercetin-3-O-rutinoside (peak 10), kaempferol-3-O-rutinoside (peak 18), quercetin-3-O-glucoside (peak 19), isorhamnetin-3-O-rutinoside (peak 25) and isorhamnetin-3-O-glucoside (peak 28) were positively identified according to their retention, mass and UV-vis characteristics in comparison with commercial standards (Figure 2). Among them, peaks 10, 18 and 19 were found in all the studied samples, while compounds 7 (BB3), 25 (BB1) and 28 (BB1, BB3, BB5 and BBC) were only detected in some bee bread samples (  , kaempferol and isorhamnetin 3-O-rutinosides may point to peak 1 also a 3-O-rutinoside,  and thus it was tentatively assigned as myricetin-3-O- Tables 1 and 2. Peaks 2, 4, 8, 9 and 26 were identified as quercetin derivatives owing to the product ion observed at m/z 301 and UV spectra (λmax around 350-358 nm). Peaks 2 and 9 presented the same pseudomolecular ion [M − H] − at m/z 771. Peak 2's MS 2 fragments revealed the alternative loss of hexosyl (m/z at 609; −162 u) and deoxyhexosyl-hexoside (m/z at 463; −308 u) residues, indicating the location of each residue on different positions of the aglycone. Nevertheless, for peak 9 the observation of only one MS 2 fragment suggested that the three sugars were linked together. For both peaks 2 and 9, no information about the identity of the sugar moieties and the location on the aglycone could be obtained, so the compounds were tentatively identified as quercetin-O-hexosyl-O-(deoxyhexosyl-hexoside) and quercetin-O-hexosyl-deoxyhexosyl-hexoside, respectively. Nevertheless, the positive identification of different rutinosides, including quercetin-3-O-rutinoside, in the analyzed samples may suggest a rutinose identity for the deoxyhexosyl-hexose residue present in peaks 2 and 9. The mass characteristics of peak 4 ([M − H] − at m/z 625) indicated that it corresponds to a quercetin derivative bearing two hexosyl residues. The observation of MS 2 fragments at m/z 463 (−162 u) and 301 (−162 u) also indicated the alternative loss of each of the hexosyl moieties, respectively, pointing to their location on different positions of the aglycone. Thus, this compound was tentatively identified as quercetin-O-hexosyl-Ohexoside. Figure 3a exemplifies the fragmentation pattern of these types of compounds, and gives a tentative identification for peak 4. This compound was the majority flavonoid in the commercial sample and it was present in all the samples, with the exception of BB3 (Table 2). showed a similar fragmentation pattern as peak 9, only releasing one fragment at m/z 301, from the respective losses of pentosyl-hexoside (294 u) and deoxyhexose (146 u) moieties, and thus was tentatively assigned as a quercetin-Opentosyl-hexoside and a quercetin-O-deoxyhexoside. This latter compound was detected in all the samples, being the majority flavonoid found in BB3 and BB5. Although MS analysis does not allow concluding about the nature and position of the substituting sugar, owing to its relatively high abundance in the analyzed samples, peak 26 can be speculated to correspond to quercitrin (i.e., quercetin-3-O-rhamnoside), a widespread flavonoid in plants.  Figure 3a exemplifies the fragmentation pattern of these types of compounds, and gives a tentative identification for peak 4. This compound was the majority flavonoid in the commercial sample and it was present in all the samples, with the exception of BB3 (Table 2). showed a similar fragmentation pattern as peak 9, only releasing one fragment at m/z 301, from the respective losses of pentosyl-hexoside (294 u) and deoxyhexose (146 u) moieties, and thus was tentatively assigned as a quercetin-O-pentosyl-hexoside and a quercetin-O-deoxyhexoside. This latter compound was detected in all the samples, being the majority flavonoid found in BB3 and BB5. Although MS analysis does not allow concluding about the nature and position of the substituting sugar, owing to its relatively high abundance in the analyzed samples, peak 26 can be speculated to correspond to quercitrin (i.e., quercetin-3-O-rhamnoside), a widespread flavonoid in plants.      Peaks 6, 11, 13, 14 and 23 were tentatively identified as methyl-herbacetin glycosides based on their UV spectra and the production of two MS 2 fragments at m/z 315 and 300. This assignment was supported by the previous identification of similar compounds in bee pollen samples [11][12][13]. Compound identities were assigned based on their pseudomolecular ions, as methyl-herbacetin-O-deoxyhexosylhexoside (peak 13), methyl-herbacetin-O-hexoside (peak 23), methyl-herbacetin-O-hexosyl-deoxyhexosylhexoside (peak 11) and two methyl-herbacetin-O-dihexoside isomers (peaks 6 and 14). Peaks 13 and 23 can be speculated to correspond respectively to methyl-herbacetin-3-O-rutinoside and methylherbacetin-3-O-glucoside, taking into account the presence in the samples of equivalent glycosides derived from other flavonols. With the exception of peak 11, which was only found in sample BB5, the rest of the methyl-herbacetin derivatives were detected in most of the analyzed samples (Table 2).
Finally, peak 24 ([M − H] − at m/z 477), only detected in sample BB3, showed a similar fragmentation pattern as the methyl-herbacetin derivatives, releasing in this case two MS 2 fragment ions at m/z 331 and 315, and was tentatively identified as a methylmyricetin derivative, possibly laricitrin (i.e., 3 -O-methylmyricetin). According to its pseudomolecular ion, it would be a laricitrin-O-deoxyhexoside, which can also be speculated to correspond to laricitrin-3-Orhamnoside owing to the detection of other rhamnosides in the sample (peaks 26 and 30). To the authors' knowledge, this is the first report on the presence of flavonol glycoside derivatives in bee bread.

Antitumoral Activity of the BB Samples
The BB samples showed some toxicity against the four human tumor cell lines were used: MCF-7 (breast adenocarcinoma), NCI-H460 (non-small cell lung cancer), HeLa (cervical carcinoma) and HepG2 (hepatocellular carcinoma). However, at the tested concentrations, the samples inhibited less than 50% of the growth of the tumor cells, the results being expressed in terms of GI 25 values (sample concentration providing 25% of growth inhibition) ( Table 3). BBC was selective for the HeLa cell line, while BB3 inhibited the growth of all the tested human tumor cell lines, being the only one that was able to inhibit HepG2 growth. Besides the mentioned data, BB1 and mostly BB2 were also active against MCF-7, BB4 and BB5 against NCI-H460, and BB1, BB5 and principally BB4 against HeLa. It should be highlighted that up to 400 µg/mL, none of the BB samples showed toxicity for normal cells (non-tumor porcine liver primary cells). Despite the reports on the antitumor properties of different phenolic compounds, including flavonoids [14], it was not possible to establish a positive correlation between the concentration of flavonoids in each sample and the corresponding cytotoxicity. Therefore, these properties could be attributed to specific individual flavonoids, synergism/antagonism dynamics in the samples and, moreover, the presence of other compounds rather than flavonoids.

Samples Collection
The samples of BB were collected in 2012 from Apis mellifera iberiensis hives located in different apiaries near Bragança, in the northeast region of Portugal; specifically located in Bragança (BB1), Montesinho (BB2), Rio de Onor (BB3), Vinhais (BB4), and Castrelos (BB5). For extraction, the frames next to the bee brood were removed from the hives, freeze and the wax crushed mechanically to collect the bee bread. An additional commercial sample (BBC), collected from Apis dorsata on the Himalayan region, was courtesy of Bee Healthy Farms (Springfield, MO, USA).
Once extracted, the samples were lyophilized (FreeZone 4.5 model 7750031, Labconco, Kansas City, MO, USA), reduced to a fine dried powder (20 mesh), mixed to obtain homogenous samples and stored in a desiccator, protected from light, until further analysis.

Extracts Preparation
A methanol:water (80:20, v/v) extract was obtained from the lyophilized material. Each sample (1 g) was extracted twice by stirring (25 • C at 150 rpm) with 30 mL of methanol:water (80:20, v/v) for 1 h and subsequently filtered through a Whatman No. 4 paper. The combined methanol:water extracts were evaporated at 40 • C (rotary evaporator Büchi R-210, Flawil, Switzerland) to remove the methanol and further frozen and lyophilized.

Characterization of the Extracts by HPLC-DAD-ESI/MS
The phenolic compounds were analysed using a Hewlett-Packard 1100 chromatograph (Hewlett-Packard 1100, Agilent Technologies, Santa Clara, CA, USA) with a quaternary pump and a diode array detector (DAD) coupled to an HP Chem Station (rev. A.05.04) data-processing station. A Waters Spherisorb S3 ODS-2 C18, 3 µm (4.6 mm × 150 mm) column thermostatted at 35 • C was used and 10 µL of each sample was injected. The solvents used were: (A) 0.1% formic acid in water, (B) acetonitrile. The elution gradient established was 15% B for 5 min, 15% B to 20% B over 5 min, 20%-25% B over 10 min, 25%-35% B over 10 min, 35%-50% B for 10 min, and re-equilibration of the column, using a flow rate of 0.5 mL/min. Double online detection was carried out in the DAD using 280 nm and 370 nm as preferred wavelengths and in a mass spectrometer (MS) connected to HPLC system via the DAD cell outlet.
MS detection was performed in an API 3200 Qtrap (Applied Biosystems, Darmstadt, Germany) equipped with an ESI source and a triple quadrupole-ion trap mass analyser that was controlled by the Analyst 5.1 software. Zero grade air served as the nebulizer gas (30 psi) and turbo gas for solvent drying (400 • C, 40 psi). Nitrogen served as the curtain (20 psi) and collision gas (medium). The quadrupoles were set at unit resolution. The ion spray voltage was set at −4500 V in the negative mode. The MS detector was programmed for recording in two consecutive modes: Enhanced MS (EMS) and enhanced product ion (EPI) analysis. EMS was employed to show full scan spectra, so as to obtain an overview of all of the ions in sample. Settings used were: declustering potential (DP) −450 V, entrance potential (EP) −6 V, collision energy (CE) −10 V. EPI mode was performed in order to obtain the fragmentation pattern of the parent ion(s) in the previous scan using the following parameters: DP −50 V, EP −6 V, CE −25 V, and collision energy spread (CES) 0 V. Spectra were recorded in negative ion mode between m/z 100 and 1000.

Evaluation of In Vitro Cytotoxic Properties of the Extracts
The BB extracts were dissolved in water at 8 mg/mL and then submitted to further dilutions from 400 to 1.56 µg/mL. Four human tumor cell lines were used: MCF-7 (breast adenocarcinoma), NCI-H460 (non-small cell lung cancer), HeLa (cervical carcinoma) and HepG2 (hepatocellular carcinoma). Cells were routinely maintained as adherent cell cultures in RPMI-1640 medium containing 10% heat-inactivated FBS and 2 mM glutamine (MCF-7, NCI-H460 HeLa and HepG2 cells), at 37 • C, in a humidified air incubator containing 5% CO 2 . Each cell line was plated at an appropriate density (7.5 × 10 3 cells/well for MCF-7 and NCI-H460 or 1.0 × 10 4 cells/well for HeLa and HepG2) in 96-well plates. Sulforhodamine B assay was performed according to a procedure previously described by the authors [15].
For evaluation of the cytotoxicity in non-tumor cells, a cell culture was prepared from a freshly harvested porcine liver obtained from a local slaughterhouse, according to a procedure established by the authors [16]; it was designed as PLP2. Cultivation of the cells was continued with direct monitoring every two to three days using a phase contrast microscope. Before confluence was reached, cells were subcultured and plated in 96-well plates at a density of 1.0 × 10 4 cells/well, and commercial in DMEM medium with 10% FBS, 100 U/mL penicillin and 100 µg/mL streptomycin. Ellipticine was used as positive control and all the results were expressed in GI25 values (concentration that inhibited 25% of the net cell growth).

Statistical Analysis
Data on phenolic compounds quantification were expressed as mean ± standard deviation. Differences in the phenolic compounds levels among each of the assayed bee bread samples were analyzed through one-way analysis of variance (ANOVA). The fulfilment of the one-way ANOVA requirements, specifically the normal distribution of the residuals and the homogeneity of variance, was tested by means of the Shapiro Wilk's and the Levene's tests, respectively. All dependent variables were compared using Tukey's honestly significant difference (HSD) or Tamhane's T2 multiple comparison tests, when homoscedasticity was verified or not, respectively. For all compounds detected in only two bee bread samples, statistically significant differences among means were classified according to t-student test. All statistical tests were performed at a 5% significance level using SPSS software, version 23.0 (IBM Corp., Armonk, NY, USA).

Conclusions
Overall, bee bread is a very recent bee product and at this stage it can be collected and consumed as a food supplement named "bee bread". To the authors' knowledge, this is the first report characterizing glycosidic flavonoids in bee bread samples, contributing to the chemical knowledge of this less explored bee product. Thirty-two flavonol glycoside derivatives, such as quercetin, kaempferol, myricetin, isorhamnetin and herbacetrin derivatives, were identified in the six samples of bee bread. BB1 and BB3 were the bee bread samples that presented the highest content and the highest number of identified compounds (19 compounds). Two isorhamnetin glycoside derivatives, isrohamnetin-O-hexosyl-O-rutinoside and isorhamnetin-O-pentosyl-hexoside, were the most abundant compounds present in BB1; on the other hand, quercetin-3-O-rhamnoside was the most abundant flavonol in BB3. BB samples showed moderate antitumor activity; however, none of the BB samples have shown toxicity for normal cells.