PAPER 248 Ital. J. Food Sci., vol. 27 - 2015 - Keywords: antioxidant activity, pollen, bioactive compounds, chemical composition, fluorescence, Tuscany - PHYTOCHEMICAL COMPOSITION AND ANTIOXIDANT ACTIVITY OF TUSCAN BEE POLLEN OF DIFFERENT BOTANIC ORIGINS M. GABRIELEa, E. PARRIb, A. FELICIOLIc, S. SAGONAc, L. POZZOd, C. BIONDIe, V. DOMENICIb* and L. PUCCIa aNational Research Council (CNR), Institute of Biology and Agricultural Biotechnology (IBBA), Pisa Unit, Research Area of Pisa, Via Moruzzi 1, 56124, Pisa, Italy bChemistry and Industrial Chemistry Department, University of Pisa, Via Moruzzi 13, 56124, Pisa, Italy cDepartment of Veterinary Sciences, Viale delle Piagge 2, 56124, Università di Pisa, Italy dNational Research Council (CNR), Institute of Sciences of Food Production (ISPA), Turin Unit, Largo Braccini 2, 10095, Grugliasco (TO), Italy eVia Carlini S. 73, 56127 Pisa, Italy *Corresponding author: Tel. +39 050 2219215/267, Fax +39 050 2219260, email: valentina.domenici@unipi.it ABSTRACT Within the apicultural products, the honey bee-pollen is growing in commercial interest due to its high nutritional properties. For the first time, bee-pollen samples from Tuscany (Italy) were studied to evaluate botanical origin, phytochemical composition and antioxidant activity. The in- vestigated pollen loads were composed of three botanical families: Castanea, Rubus and Cistus. The highest levels of proteins and lipids were detected in Rubus pollen. Castanea pollen contained greater polyphenols, flavonoids and anthocyanins content, while the highest flavonols level was detected in Cistus pollen. These results were also confirmed by front-face fluorescence spectros- copy, used here, for the first time, as a fast tool to characterize bee-pollens. mailto:valentina.domenici@unipi.it Ital. J. Food Sci., vol. 27 - 2015 249 1. INTRODUCTION For centuries the apicultural products have been used in phytotherapy as well as in diet for their health positive implications (KROYER and HEGEDUS, 2001; FERRERES et al., 2010; ABOU- DA et al., 2011). Bee-gathered pollen (bee-pollen) is an apicul- tural product of great commercial interest owing to its high nutritional value and physiological properties representing an important source of energy and proteins for human nutrition (ABOU- DA et al., 2011). Worker-bees gather pollen from flowers and, after processing it with some proteic gland-secre- tion, they package it on their curbicula forming pollen loads before returning to the hive (PINZAU- TI et al., 2002; SCARSELLI et al., 2005). General- ly, a single pollen load is a one colored little pel- let reflecting a homogeneous and monospecific pollen content. Bee-pollen loads are stilled from the worker bees at their entrance to the hive by special pollen loads traps. In accordance to the season, to the timing of collection by the bee- keepers and to the post collection management is possible to obtain mono- or poly-pollen spe- cies loads. Amounts of pollen loads ready to be commercialized and consumed by human and animals is called bee-pollen. So it is possible to find bee-pollen of one specific flower (monoflo- ra) or belonging to several flower species (polyflo- ra), as well is possible to blend different mono- floral in order to create mixtures of bee-pollen with characteristic organoleptic properties and quality attributes. Nowadays, bee-pollen represents the richest and most complete natural food supplying high levels of carbohydrates (13-55%), proteins (10- 40%), particularly free aminoacids, enzymes, co- factors, lipids (1-13%), including fatty acids and sterols, minerals, trace elements and vitamins, especially B group, A, C and E. Fresh and dry bee-pollen loads hold a different water content, ranging from 20-30% in the original form and 4-10% if dried, affecting organoleptic and “shelf- life time” properties (CAMPOS et al., 2008; PAS- COAL et al., 2014). Moreover, it is also an excellent source of bio- active compounds, such as phytosterols, carot- enoids and polyphenols (especially flavonoids), that exert antioxidant, anti-inflammatory, anti- microbial, anti-allergic and antitumoral effects (MORAIS et al., 2011; PÉREZ- PÉREZ et al., 2012; FRATINI et al., 2014). Very recently, several re- search works were published showing that bee products, such as propolis and pollens, possess a sedative effect and may be effective in protect- ing humans against depression and similar dis- eases (YILDIZ et al. 2014); additionally, prelimi- nary studies show that pollens have a hepatopro- tective potential (YILDIZ et al. 2013). Pollens ef- fects on improving immune, cardiovascular and digestive systems as well as their therapeutic ef- fects have been mainly related to the polyphenol content and chemical composition (PASCOAL et al., 2014). In particular, the phenolic profile of bee-pollen consists of flavonol, glycosides and aglycones, and hydroxycinnamic acids, that can be present in free forms or combined with other pollen components (CHANTARUDEE et al., 2012; FANALI et al., 2013). However, as well as chem- ical composition, the phytochemical profile is affected by soil type, beekeeping management, climatic and preservation conditions, and es- pecially by botanical origin (ALMEIDA-MURADI- AN et al., 2005; ARRUDA et al., 2013; CAMPOS et al., 2008). Aromatic aminoacids, many polyphe- nols, some enzyme cofactors, but also some wa- ter- or lipid-soluble vitamins and pigment’s de- rivatives contained in bee-pollen are fluorescent intrinsic compounds (JØRGENSEN et al., 1992; LAKOWICZ, 2006). Front-face (FF) fluorescence spectroscopy is a non-destructive, rapid and sensitive technique suitable for complex and opaque samples, oth- erwise traditional right-angle florescence (KA- ROUI et al., 2007; ZANDOMENEGHI et al., 2005). In particular, this technique has been proved to be very effective in studying powders, crystal- line or amorphous samples, and complex ma- trix, such as food (ZANDOMENEGHI, 1999; ZAN- DOMENEGHI and ZANDOMENEGHI, 2009; AIRA- DO-RODRÍGUEZ et al., 2011; KULMYRZAEV et al. 2005). In this context, FF fluorescence could be a useful instrument to obtain a fingerprint of dif- ferent bee-pollen types, advantageous to study and compare them. In this study, we examined the botanical or- igin, the chemical composition (moisture, pro- teins, carbohydrates, lipids, ash) and the anti- oxidant profile (the total polyphenols and the flavonoids, flavonols and anthocyanins sub- classes) of different color fractions of an organ- ic Tuscan (Italian) bee-pollen sample. Moreover, we proposed for the first time a rapid qualitative evaluation of emission, excitation and synchro- nous spectra, obtained by front-face spectros- copy, of bulk state and ethanolic extracts of the different pollen types showing the main class- es of identified fluorescent molecules. Lastly, to evaluate the synergic effect of the bee-pollen bi- oactive components we analyzed by DPPH and ORAC assay the free radical scavenging activi- ty and the antioxidant capacity of both separate and mixed color fractions. 2. MATERIALS AND METHODS 2.1 Chemicals and reagents All standards and reagents were of analytical grade. Absolute ethanol, methanol, hydrochlo- ric acid, trichloroacetic acid, diethyl ether, so- dium carbonate, sodium idrosside, potassium chloride, sodium acetate, Folin-Ciocalteu rea- 250 Ital. J. Food Sci., vol. 27 - 2015 gent, catechin, gallic acid, 6-hydroxy-2,5,7,8- tetramethylchromane-2-carboxylic acid (Trolox), 1,1-diphenyl-2-picrylhydrazyl (DPPH) and fluo- rescein sodium salt were purchased from Flu- ka-Sigma-Aldrich, Inc. (St. Louis, MO), as well as the solid standards used in spectroscop- ic analysis tryptophan, b-carotene, gallic acid, 3,4-dihydroxybenzoic acid, 4-hydroxybenzoic acid, vanillic acid, caffeic acid, ferulic acid, p- coumaric acid, quercetin dihydrate and the vi- tamins C, B3, B2, B6 and B9. Hydroxide perox- ide, sulfuric acid, boric acid and Kjeldahl tab- lets were purchased from Merck (Readington, NJ). Sodium nitrite and aluminum chloride were purchased from Carlo Erba (Milan, IT), while 2,2’-azobis (2-amidinopropane) dihydrochloride (AAPH) was purchased from Polysciences, Inc. (Warrington, PA). 2.2 Plant materials Bee-pollen resulted from the blend of pollen loads collected during sunny days by the bee- keeper using 10 beehives equipped with bottom- fitted pollen traps located in Massa Macinaia (Latitude 43.80638- Longitude 10.54213) in Luc- ca Province (Tuscany, Italy) between April and July 2013. In total 1 kg of bee-pollen was col- lected. The blended fresh bee-pollen was stored at -20°C in the dark until further analysis. 2.3 Palynological analysis Within the bee-pollen sample the pollen loads were divided by colour into three groups (Fig. 1). Twenty single pollen loads of each colour were analyzed by microscope. Each single pollen load prepared by washing the pollen with distilled wa- ter and using glycerin jelly for permanent prep- arations. Pollen grains identification was per- formed by optical microscope with total magni- fication (400X and 1000X). A reference collection of Pisa University and different pollen morphology guides were used for the recognition of the pollen types. 2.4 Chemical composition Dry matter, proteins, lipids and ash quan- tification were performed for the pollen-load samples according to AOAC guidelines (AOAC, 2000) and values were expressed as percentage on fresh matter basis. Dry matter content was made through gravimetry until constant weight, using oven at 105°C. Protein content was ob- tained using Kjeldahl method, while lipid con- tent was determined by Soxhlet extractor using diethyl ether as solvent. Ash content was deter- mined by gravimetry until constant weight, us- ing oven at 550°C for one day. Moisture con- tent was obtained by subtracting the dry mat- ter from 100, while total carbohydrates content was determined according to the following for- mula: Carbohydrates (g) = 100 - [protein (g) + fat (g) + moisture (g) + ash (g)] (Ketkar, Rathore, Lohidasan, Rao, Paradkar, and Mahadik, 2014). 2.5 Bee-pollen extraction and phytochemical characterization Pollen-loads were separated in three color groups and the mixed pollen sample was made blending the three pollen type in equal part. Bee-pollen extracts were obtained after 1 hour incubation at room temperature in 95% ethanol while being shaken gently. Then samples were centrifuged 10 minutes at 3500 rpm at 4°C and the supernatants were collected and kept in the dark at 4°C. Total polyphenols were determined by the Fo- lin-Ciocalteu colorimetric method (SINGLETON et al.,1999). Briefly, 100 μL of each natural ex- tract were mixed with 500 μL of 0.2 N Folin-Cio- calteu reagent and incubated in the dark for 5 minutes. Then, 400 μL of 0.7 M sodium carbon- ate (Na2CO3) were added. The absorbance was measured at 760 nm, after 2 hours incubation at room temperature in the dark. Five serial di- lutions of gallic acid (0.009, 0,017, 0,043, 0,085, 0.17 mg/mL) were used to obtain the standard calibration curve with the following equation: conc=0.079Abs-0.004 (R2=0.996). Total polyphe- nols were expressed as mg of gallic acid equiv- alents (GAE)/g dry weight (dw). The aluminum chloride colorimetric meth- od was used for the total flavonoids determina- tion (Kim D.O., Chun, Kim Y.J., Moon, and Lee, 2003). Briefly, 200 μL of extracts were mixed with 800 μL of dH 2 O and 60 μL of 5% NaNO 2 , fol- lowed by incubation of 5 minutes at room tem- perature. Then 60 μL of 10% AlCl 3 were added, incubated for 6 min and finally reactions were neutralized with 400 μL of 1M NaOH. Absorb- Fig. 1 - Picture of the original blended bee-pollen sample and Castanea sp. (yellow), Rubus sp. (green) and Cistus sp. (ochre) pollen. Ital. J. Food Sci., vol. 27 - 2015 251 ance was measured at 430 nm after 30 minutes of incubation and catechin was used to make the calibration standard curve (0.016, 0.032, 0.063, 0.125, 0.25, 0.5 mg/mL). Flavonoids concentra- tion was obtained from the following calibration curve: conc=0.464Abs-0.005 (R2=0.998) and ex- pressed as mg catechin equivalent (CE)/g dw. Flavonols content was measured according to the method described by Romani, Mancini, Tatti and Vincieri (1996). Briefly, 25 μL of ex- tracts were mixed with 225 μL of 10% EtOH, 250 μL of 0.1% HCl in 95% EtOH and 1 mL of 2% HCl. After incubation at room temperature for 30 minutes, the absorbance of the reaction mixture was measured at 360 nm and querce- tin was used to make the calibration standard curve (0.0006, 0.00125, 0.0025, 0,0050, 0.0075, 0.01, 0.0125, 0.015 mg/mL). Flavonols con- tent, derived from the calibration curve equa- tion: conc=0.536Abs-0.0001 (R2=0.997), was ex- pressed as mg quercetin equivalent (QE)/g dw, Total monomeric anthocyanins were deter- mined according to the pH differential method described by Lee, Durst, and Wrolstad (2005), a spectrophotometric method based on the change in pigmentation pH-dependent of antho- cyanins. Absorbance was measured at 520 and 700 nm and anthocyanin concentration was ex- pressed as mg cyanidin-3-glucoside equivalents (C3GE)/g dw (cyd-glu, molar extinction coeffi- cient of 26,900 L cm-1 mol-1 and molecular weight of 449.2 g mol-1). 2.6 Front-Face Fluorescence spectroscopy analysis Each pollen sample was studied previous in its bulk state by FF fluorescence spectroscopy, then their ethanolic extracts were studied by UV- Vis absorption and FF fluorescence spectrosco- pies. The precipitate obtained after the extrac- tion was also studied by FF fluorescence spec- troscopy. To recognize the main classes of fluo- rescence compounds in pollen loads, standard solutions of the reagents reported in paragraph 2.1 were studied by Uv-Vis Absorbance and FF fluorescence spectroscopy. For bulk analysis pollen loads were finely pow- dered with a pestle in a mortar. A little amount of powder was put between two quartz windows of 1 mm optical path with the help of few water drops to homogenized the sample (about 1:1 mg/ ml sample: water). These quartz windows are held against a support in the spectrofluorometer by a laminar spring. For extraction procedure 2 mL of ethanol were added to 100 mg of pollen loads finely powdered. They were stirred for 30 minutes with a magnetic stir bar and then filtered through filter paper 0.45 mm pore size (Sartorious), to ob- tain a clear solution. To do it, Spectrum S-25-10 stirred cell device for ultra-filtration was used. A quartz cell with 2 mm optical path was used to re- cord absorbance spectra and low capacity quartz cells with 5 mm optical path was used to record FF fluorescence spectra. The residual material was kindly removed from the filter paper and it was put between two quartz windows of 1 mm optical path. Fluorescence spectra were recorded using a ISA Fluoromax II photon counting spec- trofluorometer, with Xenon arc lamp and a de- vice for front-face measurements with a cell hold- er designed to set the incident angle of the exci- tation beam at 31°, eliminating or reducing self- absorption effects, light-reflected and scattering (ZANDOMENEGHI et al., 2005). The excitation and the emission slits were 2 and 5 nm, respectively. The integration constant time was 0.5 s and the wavelength increment was 1 nm. The intensity of the spectra was determined as the ratio between the emisson signal (counts per second, cps) and the intensity of light from the excitation mono- chromator (mA), measured by means of a pho- tomultiplier and a photodiode, respectively. For each sample emission spectra (280 nm < lamb- da_ex < 550 nm, with a step of 10 nm), excita- tion spectra (at the wavelength of the maximum position of fluorescence emission spectra, l em ) and synchronous spectra (20 nm < Deltalambda < 120 nm, with a step of 10 nm) were recorded. For bulk analysis emission spectra with l ex =650 nm were recorded. Absorbance spectra were re- corded with a Jasco V-550 spectrophotomer, be- tween 200 nm and 750 nm with scanning speed of 400 nm/s, band width and data pitch of 1 nm and 0.5 nm, respectively. 2.7 Antioxidant activity 2.7.1 DPPH radical scavenging assay The free-radical scavenging activity of etha- nolic bee-pollen extracts was evaluated using the 1,1-diphenyl-2-picrylhydrazyl (DPPH) assay (FRASSINETTI et al. 2011). The reduction of DPPH radicals was recorded at 517 nm and the radi- cal scavenging activity (RSA) was calculated as percentage of DPPH inhibition according to the following equation: % RSA = [(A DPPH -A S )/A DPPH ] x 100, where A DPPH is the absorbance of DPPH so- lution and A S the absorbance of sample. The ex- tract concentration corresponding 50% of DPPH inhibition (EC 50 ) was measured by interpolation from the graph of RSA percentage versus bee- pollen concentration (Morais et al., 2011). Lower EC 50 values indicate higher antioxidant activities. 2.7.2 Oxygen Radical Absorbance Capacity (ORAC) Assay The antioxidant capacity of ethanolic bee-pol- len extracts was quantified using the oxygen rad- ical absorbance capacity (ORAC) assay, modify- ing some reagent concentration adapted to our requirements (NINFALI et al., 2005). The final reaction mixture of our assay con- tained 0·04 mM fluorescein sodium salt in 0.075 252 Ital. J. Food Sci., vol. 27 - 2015 M phosphate buffer, pH 7.4, at diluted sample or 5 mM Trolox. The control was 0.075 M phos- phate buffer, pH 7.4. AAPH was used as perox- yl radicals generator and fluorescein as probe. Fluorescein fluorescence decay was read at 485 nm excitation and 514 nm emission using a Vic- torTM X3 Multilabel Plate Reader (Waltham, MA) and Trolox was used as antioxidant standard. ORAC values were expressed as micromoles of Trolox equivalents (TE)/g dw. 2.8 Statistical analysis The statistical analysis was performed us- ing GraphPad Prism, version 5.00 for Windows (GraphPad software, San Diego, CA). Assays were carried out in triplicate and results were expressed as mean values ± standard deviation (SD). Differences between bee-pollen samples were analyzed by one-way analysis of variance (ANOVA) followed by Tukey’s post test. A p-val- ue lower than 0.05 is considered as statistical- ly significant. Interdependence between the an- tioxidant capacity and the phytochemical pro- file was evaluated by Pearson’s correlation co- efficient (r). 3. RESULTS 3.1 Palynological analysis The blended bee-pollen sample resulted to be a mixture of Castanea sp. (yellow), Rubus sp. (green) and Cistus sp. (ochre). Each pollen load owned a homogeneous and monospecific pollen content. The Castanea sp. was the most repre- sentative (70%), followed by Rubus sp. (23%) and Cistus sp. (7%). 3.2 Chemical composition The chemical composition of Castanea, Cistus and Rubus pollen samples is listed in Table 1 and values are expressed as percentage on fresh matter basis. The nutritional content measured is in agreement with literature values (Balkan- ska and Ignatova, 2012; Carpes, Mourão, Alen- car, and Masson, 2009; Nogueira, Iglesias, Feás, and Estevinho, 2012). Significant variations among samples were showed, with the highest protein (p<0.01) and lipid (p<0.001) content in Rubus compared to all other pollens. No differ- ence among groups was found for moisture, dry matter, ash and carbohydrates content (p=ns). 3.3 Phytochemicals profile of ethanolic pollen-load extracts Pollen-load extracts were screened for total polyphenols, flavonoids, flavonols and monomer- ic anthocyanins content. Phytochemical profile of Castanea, Cistus and Rubus pollen-load sam- ples is listed in Table 2 and significant differenc- es were found (p<0.001). In particular, Castanea pollen extracts contained the highest levels of polyphenols (24.75±0.78 mg GAE/g fw), flavo- noids (15.86±0.62 mg CE/g fw) and anthocya- nins (77.37±2.55 mg C3GE/L), while the high- est levels of flavonols (4.93±0.05 mg QE/g fw) were detected in Cistus pollen samples. Other- wise, Rubus pollen extracts showed the lowest Table 1 - Moisture, dry matter, proteins, lipids, ash and carbohydrates of pollen-load samples. a,b,c Different superscript letters indicate statistical differences among the bee-pollen extracts (p<0.001 ANOVA). Assays were carried out in triplicate and results were expressed as mean values ± SD. Ital. J. Food Sci., vol. 27 - 2015 253 content of polyphenols, flavonoids and flavonols. The discrepancies in phytochemical composi- tion observed among the different pollen sam- ples might depend on their botanical origin (AR- RUDA et al., 2013; CAMPOS et al., 2008; MORAIS et al., 2011). 3.4 Fluorescence spectroscopy 3.4.1 Bulk analysis Pollen is a complex matrix and, as a conse- quence, the fluorescence spectra are character- ized by broad and overlapped bands, caused by the presence of many fluorophores, which lim- it the quantification and identification of all of them. However, it is possible to get several infor- mation as well as qualitative features, which can be used as fingerprint of the investigated pollen. The fluorescence spectra are indeed character- ized by three main intervals of excitation wave- length: 280-290 nm, 320-370 nm and 420-480 nm. In Fig. 2a the emission spectra obtained with excitation wavelength (l ex ) of 280 nm are re- ported. Two bands can be recognized: the first one, less intense, is centered at 340 nm for Ru- bus and at 360 nm for Castanea and Cistus pol- lens. This band is probably due to the aromatic aminoacids, which can be residues of proteins or free. The second one is the dominant band and its shape is different between Rubus and Castanea/Cistus pollens: it is more intense in the range 420-620 nm for Rubus and red shift- ed of about 40 nm for Castanea and Cistus pol- lens. This emission band is probably due to hy- droxycinnamic acids, compounds belonging to polyphenols family, fluorescent water soluble vi- tamins such as B6, B9 and B2, and flavoinoid compounds. All these classes of compounds have two bands of absorption (three in the case of B2 vitamin), the first one centered in the 280-290 nm interval. This hypothesis is supported by the emission spectra obtained at highest l ex 320-370 nm, where the above mentioned substances are excited in their second absorption band; these pollens’ bands are indeed less intense but retain the same shape of previous broad band. It is in- teresting to observe that Castanea and Cistus’s spectra are again much different from Rubus (Fig. 2b), probably due to the presence of char- acteristic fluorophores in the latter pollen type. For all pollens, fluorescence intensity decreas- es until λ ex =400 nm; thereafter, it slightly in- creases again, to define a new band centered at 530 nm for Cistus and Castanea and at 510 nm for Rubus (λ ex =450 nm). At this wavelength, B2 vitamin has its third absorption band, moreover, other fluorophores are excited, such as xantho- phylls and carotenoids derivatives, although they are very weak emitter (JØRGENSEN et al., 1992). Changing the λ ex to 550 nm, only Cistus pollen presents a weak fluorescence at about 600 nm, probably caused by the presence of other po- lar carotenoids derivatives. At higher excitation wavelength, collected to investigate the eventual presence of chlorophyll derivatives (λ ex =650 nm), no relevant fluorescence is recorded. The synchronous spectra obtained with 60 nm offset (Fig. 2c) underlines the presence of three different bands of absorption and they are use- ful to compare the pollen types. The fluorescence profile of Rubus pollen is much different than that of Castanea/Cistus’ one, especially in the second absorption bands (IIR and IIC respective- ly). In this spectrum, also the third band differs between Rubus and Castanea/Cistus, centered Table 2 - Total polyphenols, flavonoids, flavonols and anthocyanins concentration of ethanolic pollen-loads extracts. a,b,c Different superscript letters indicate statistical differences among the bee-pollen extracts (p<0.001 ANOVA). Assays were carried out in triplicate and results were expressed as mean values ± SD. 254 Ital. J. Food Sci., vol. 27 - 2015 Fig. 2 - Comparison between fluorescence spectra of the different pollen types in their bulk state. (a) Emission spectra obtained with l ex =280 nm; the arrow points out the first emis- sion band. The peak at 560 nm is related to lamp; the peaks between wavelength 460-490 nm and at 765 nm are artefacts of lamp. (b) Emission spectra obtained with l ex =350 nm. (c) Synchronous spectrum obtained with Dl ex =60 nm. The roman numerals point out the princi- pal absorption bands, as explained in the text. This spectrum is normalized. at 440 nm (IIIR) and 480 nm (IIIC), respectively. Moreover, the Cistus pollen is the only one pre- senting an absorption centered at 550 nm (IV). All above data are consistent with the emission spectra. Vitamins C and B 3 did not show rele- vant fluorescence so they resulted undetectable. 3.4.2 Extracts analysis In order to facilitate the spectral interpreta- tion, an extraction procedure to separate the wa- ter soluble fraction from the lipid soluble one was adopted. The solvent used are not strong and the treatment of samples is such to limit the altera- tion of the matrix, and to use the potentiality of a rapid and direct spectroscopic technique. In the present work ethanolic extracts were studied. The most interesting region of the UV-Vis ab- sorption spectra is in the visible range (400-500 nm). Here, only Cistus extract presents an ab- sorption spectrum with the typical shape of ca- rotenoid or xanthophyll pigments, as shown in Fig. 3a. Considering fluorescence, the emission spec- tra at λ ex =280 nm (Fig. 3b) are characterized by a broad band, extended to the interval 340-520 nm. Castanea and Rubus pollen show maximum emission at about 420-430 nm, while the Cis- tus one shows a different profile with an emis- sion band centered at about 400-410 nm and a shoulder at 520 nm. In this region, the fluores- cence is probably due to the hydroxycinnamic acid, that at this wavelength begin to absorb. The polyphenols belonging to the hydroxyben- zoic family (e.g. gallic acid, 3,4-dihydroxybenzo- ic acid, 4-hydroxybenzoic acid, vanillic acid) are not the mainly fluorophores. In fact, they have their absorption and emission maxima includ- ed in the interval 280-297 nm and 325-370 nm, respectively. In Fig. 3b emission spectra of two represent- ative compounds of the hydroxybenzoic’s fami- ly (gallic and 4-hydroxybenzoic acids) compared with pollen’s spectra are reported. The hypoth- esis of the presence of hydroxycinnamic acid is supported by the emission spectra recorded with highest λ ex (310-340 nm), characterized by the same shape, but higher intensity. In fact, poly- phenols belonging to hydroxycinnamic family (e.g. caffeic acid, ferulic acid, p-cumaric acid) have their absorption and emission maxima in- cluded in the interval 310-350 nm and 410-440 nm, respectively. Emission spectra of two repre- sentative compounds of this group (caffeic and p–coumaric acid) recorded with λ ex =310 nm are reported in Fig. 3c and compared with pollen’s spectra. Castanea pollen spectrum differs from the two others for the higher intensity. Ital. J. Food Sci., vol. 27 - 2015 255 Considering the excitation spectra recorded with the emission wavelength fixed in the max- imum of emission (430 nm for Castanea/Rubus and 410 nm for Cistus) it is clear that their fluo- rescence is due to different fluorophores. In par- ticular, they are a single fluorophore, with an ab- sorption band centered at 330 nm for Rubus, a single fluorophore, with two absorption bands centered at 285 and 330 nm for Castanea, and two different fluorophores, centered at 290 and 325 nm excitation wavelength for Cistus. This latter hypothesis is confirmed by synchronous spectra obtained with offset 100 nm, in which it is possible to recognize two different contri- butions with absorption at 295 and 320 nm, re- spectively. Changing the λ ex to 350 nm (Fig. 3d), the fluo- rophores involved are others, thus confirming the complexity of the pollen matrix. Castanea and Rubus samples present the same profile, while Cistus’ spectrum shows a particular shape characterized by three shoul- ders, most likely due to three different fluoro- phores, confirmed by excitation spectra record- ed with λ em =460nm. These fluorophores are probably other hydroxycinnamic compounds (in the cases of emission at 430 nm and 460 Fig. 3 - Spectra of ethanolic extracts of the different pollen types. (a) Uv-Vis absorpotion spectra (optical path=1 cm) of pol- lens compare to a b-carotene solution (3.2 mg/ml). Series of emission spectra of pollens compare to: (b) solutions of gallic acid and 4-hydroxybenzoic acid (0.01 mg/ml 90% H 2 O-10% CH 3 OH) (l ex =280 nm); (c) solutions of caffeic acid and p-couma- ric acid (0.01 mg/ml 90% H 2 O-10% CH 3 OH) (l ex =310 nm); (d) solution of quercetin (20 mg/ml in CH 3 OH) (l ex =370 nm). All the fluorescence spectra are normalized. Caffeic and p-cumaric’ spectra intensity are multiplied by 2.4. nm) and flavonoids compounds (in the case of emission at 500-520 nm). For example, the flavonol quercetin has a weak emission, but the shape of its fluorescence spectrum is com- patible with that of Cistus pollen, as report- ed in Fig. 3d. The richness in fluorescent compounds of Cistus pollen is confirmed by the synchronous spectrum obtained with offset 120 nm, where it is possible to recognize four bands of absorp- tion, centered at 295-320 nm, 350 nm, 370 nm and 400 nm. In the case of the first broad band it is possible to discriminate the contribution of two fluorophores with offset 100 nm, like previ- ously discussed. With higher excitation wavelength, collected to check the presence of carotenoids or xantho- phyll with front-face fluorescent of liquid extract, no relevant fluorescence was recorded. It is im- portant to underline that there are several xan- thophylls and carotenoids derivatives, and only some of them have weak emission, as previous discussed in section “Bulk analysis”. Finally is interesting to analyze the fluores- cence spectra obtained for solid samples. For example in Fig. 4 the comparison between emis- sion spectra (λ ex = 280 nm) of Rubus in its bulk 256 Ital. J. Food Sci., vol. 27 - 2015 state with the precipitate after ethanol extrac- tion (dotted line) is reported. The first band due to water soluble aminoacids and proteins, not extracted in ethanol, is present into both spec- tra. Instead the shape of main broad emission band centered at 500 nm, differs from each oth- er. This can be due to the lack of the classes of compounds extracted with the solvent: hydrox- ycinnamic acid and flavonol compounds. It is important to remember that fluorescence sensi- tive changes in relation of the physical state of compounds and their environment (LAKOWICZ, 2006). Although the broad emission band, these results are coherent with extracts’ spectra. 3.5 DPPH radical-scavenging activity The DPPH assay has been widely used to test the free radical scavenging activity of apicultural products, either honey, propolis or bee-pollen. As reported in literature the radical scavenging ac- tivity of bee-pollen is very dissimilar among dif- ferent flower species with a value range of 0-97% that also depends on their chemical composition and solvent extraction (BASUNY et al.,2013; LEB- LANK et al., 2009; SILVA et al., 2006). We eval- uated both percentage of DPPH radical scav- enging activity (% RSA) and EC50, the extract concentration providing 50% DPPH inhibition. As shown in Table 3, ethanolic bee-pollen ex- tracts exhibited high free radical scavenging ac- tivity with DPPH inhibition values ranged from 37.95±0.19% (EC 50 = 641.3±11.4 μg/mL) of Ru- bus extracts and 94.45±0.01% (EC 50 = 215.2±2.7 μg/mL) of Castanea extracts. ANOVA with Tukey post-test showed signifi- cant differences among samples with the high- est and lowest activity detected in Castanea and Rubus extracts. As well as radical quenching ca- pacity, Castanea, Cistus and mixed bee-pollen Fig. 4 - Emission spectra obtained with l ex =280 nm for solid Rubus samples differently processed. The dotted line shows the pre- cipitate after ethanol extraction (spectrum intensity is divided by 1.7). The peak at 560 nm is re- lated to lamp; the peaks between wavelength 460-490 nm and at 765 nm are artefacts of lamp. Table 3 - ORAC and DPPH assay results of ethanolic pollen-load extracts expressed as μmol TE/g fw and % RSA and EC 50 (μg/mL), respectively. a,b,c Different superscript letters indicate statistical differences among the bee-pollen extracts (p<0.001 ANOVA). Assays were carried out in triplicate and results were expressed as mean values ± SD. Ital. J. Food Sci., vol. 27 - 2015 257 possessed similar EC 50 values, three times low- er than Rubus extracts, suggesting a lower rad- ical scavenging activity of Rubus bee-pollen as antioxidant. Our results also revealed a strong relation be- tween the antiradical activity and the total phe- nolics, flavonoids and flavonols content, result- ing in a significant positive correlation (r=0.9645, r=0.9888, and r=0.9847, respectively). Lastly, a moderate correlation was obtained between the anthocyanins and the DPPH scavenging activi- ty (r=0.5541). 3.6 Oxygen Radical Absorbance Capacity (ORAC) The antioxidant capacity of ethanolic bee-pol- len extracts was also screened using ORAC as- say and expressed as ORAC units (μmol Trolox equivalents/g fw). The ORAC values were list- ed in Table 3 and ranged from 519.45±15.07 μmol TE/g of Rubus fraction and 677.70±12.92 μmol TE/g of mixed bee-pollen. One-way anal- ysis of variance with Tukey’s post test showed a significant increase of ORAC values in mixed bee-pollen extracts respect to each one sepa- rate fractions (p<0.001), suggesting a synergic or additive effect among Castanea, Cistus and Rubus antioxidant compounds. In fact, besides to their specific effects, many antioxidants can interact in synergistic ways, maybe protecting another against oxidative degradation, exhib- iting greater antioxidant effects (MĂRGHITAŞ et al., 2009). The results showed no strong cor- relation between the ORAC values and the an- thocyanin (r=0.1524) and flavonoid (r=0.4586) compounds; however, a moderate interdepend- ence was obtained between the antioxidant ca- pacity and the polyphenols (r=0.6638) and fla- vonols (r=0.5572) content. 4. CONCLUSIONS In this study, an organic bee-pollen sample from Tuscany was analysed for the first time investigating the botanical origin, the chemical composition, the phytochemicals profile and the antioxidant activity. Different techniques were used. In particular, we propose an original ap- plication of FF fluorescence spectroscopy, a promising approach to put in evidence differ- ences and analogies among pollens with differ- ent floral origin. This unconventional technique presents the advantage to require no particular sample pre-treatment; moreover, it is economic, fast and easy to use and it could be useful for both further scientific researches and commer- cial applications. The FF fluorescence results are coherent and uphold the spectrophotomet- ric data obtained in this investigation. Specifical- ly, the differences among Castanea/Cistus and Rubus fluorescence profiles, arisen from bulk study, are comparable and in agreement with the significant differences found in lipids and proteins composition. Furthermore, the ethanol- ic extracts fluorescence analysis, as well spec- trophotometric results, displays a higher con- tent of flavonols and polyphenols in Cistus and Castanea, respectively. Moreover, the FF fluo- rescence analysis shows the greater presence of hydroxycinnamic acids than hydroxybenzoic acids, in agreement with literature data (FANA- LI et al., 2013; KETKAR et al., 2014). Lastly, UV- Vis ethanolic extracts’ spectra reveal the pres- ence of carotenoid or xanthophyll pigments only in Cistus, confirmed also by bulk state’s emis- sion and synchronous spectra λ ex =550 nm and offset 60 nm, respectively). Besides nutritional and phytochemical com- position, the antioxidant and free radical scav- enging activity of Tuscan bee-pollen and its monofloral groups was measured. The redox properties of phenolic compounds, especial- ly flavonoid components, play a key role in de- composing peroxides and quenching oxygen, as well as in absorbing and neutralizing free radi- cals (MĂRGHITAŞ et al., 2009). Moreover, specific bioactive compounds or a combination of them can exert a different antioxidant activity, strong- ly dependent on structure and polyphenols com- position, rather that the phytochemical concen- tration (MĂRGHITAŞ et al., 2009). In particular, we showed that mixed bee-pol- len exhibit a much better antioxidant activi- ty than the separate fractions with an ORAC value significantly greater than other samples; whereas, Castanea, Cistus and mixed bee-pol- len showed a comparable DPPH radical scav- enging activity greater than Rubus antiradical capacity. These results are in agreement with literature data that strongly associate the high ability to neutralize reactive oxygen species to the phenolic compounds structure, mainly fla- vonoids and cinnamic acid derivatives (LEJA et al., 2007), maybe more representative in Cistus and Castanea pollen. Furthermore, according to MĂRGHITAŞ et al. (2009) the antioxidant effects of bioactive compounds change differently depend- ing on the antioxidant method used. Therefore, we suppose a synergic or additive effect among Castanea, Cistus and Rubus antioxidant com- pounds or resulting from new antioxidant sub- stances with greater antioxidant activity. In conclusion, the use of spectroscopic tech- niques applied to bee-pollen samples is a suit- able tool to underline differences and analogies in their micronutrient composition. The data obtained with different or complementary tech- niques are coherent and in agreement with lit- erature concerning the variability in chemical composition and antioxidant activity of pollens from different floral sources (LEBLANC et al., 2009; PASCOAL et al., 2014). Further investigations should be performed to identify and quantify the main fluorescence 258 Ital. J. Food Sci., vol. 27 - 2015 compounds present, as well as to investigate on the nature of Cistus’ pigments. Moreover, fu- ture analysis is required to separate and identi- fy the specific profile of antioxidant compounds in mixed pollen sample in order to elucidate, strengthen and confirm the hypothesis of the an- tioxidant compound’s synergic or additive effect or the presence of new antioxidant compounds. Finally, the results obtained show for the first time the high nutritional value and the good an- tioxidant activity of Tuscan bee-pollen, which make it an excellent food supplement and a good candidate for a potential nutraceutical product that could be useful in the prevention of free rad- ical associated diseases. As suggested by find- ings in a previous paper (LUCCHESI et al., 2014), cellular studies could be useful to investigate the intracellular pathways involved in the bee-pol- len antioxidant response. ACKNOWLEDGEMENTS The authors wish to express their thanks to Apicoltura Bi- ologica Aldo Metalori [Massa Macinaia (LU), Italy] who sup- plied the Tuscan bee-pollen sample. REFERENCES Abouda Z., Zerdani I., Kalalou I., Faid M. and Ahami M.T. 2011. The antibacterial activity of moroccan bee bread and bee-pollen (Fresh and Dried) against pathogenic bacteria. Research Journal of Microbiology, 6, 376-384. Almeida-Muradian L.B., Pamplona L.C., Coimbra S. and Bar- th O.M., 2005. Chemical composition and botanical eval- uation of dried bee pollen pellets. Journal of Food Com- position and Analysis 18, 105-111. Airado-Rodrígueza D., Durán-Merása I., Galeano-Díaza T. and Petter Wold J. 2011. Front-face fluorescence spec- troscopy: A new tool for control in the wine industry. Journal of Food Composition and Analysis, 24, 257-264. Arruda V.A.S., Pereira A.A.S., Freitas A.S., Barth O.M. and Almeida-Muradian L.B. 2013. Dried bee-pollen: B com- plex vitamins, physicochemical and botanical compo- sition. Journal of Food Composition and Analysis, 29, 100-105. Balkanska R.G. and Ignatova M.M. Chemical composition of multifloral bee-pollen from Bulgaria. CEFood2012, 6th Central European Congress on Food. Basuny A.M., Arafat S.M. and Soliman H.M. 2013. Chem- ical analysis of olive and palm pollen: Antioxidant and antimicrobial activation properties. Wudpecker Journal of Food Technology, 1(2), 14-21. Campos M.G.R., Bogdanov S., Almeida-Muradian L.B., Szc- zesna T., Mancebo Y., Frigerio C. and Ferreira F. 2008. Review article: Pollen composition and standardization of analytical methods. Journal of Apicultural Research and Bee World, 47(2), 156-163. Carpes S.T., Mourão G.B., Alencar S.M. and de Masson M.L. 2009. Chemical composition and free radical scavenging activity of Apis mellifera bee-pollen from Southern Brazil. Brazilian Journal of Food Technology, 12(1/4), 220-229. Chantarudee A., Phuwapraisirisan P., Kimura K., Okuyama M., Mori H., Kimura A., and Chanchao C. 2012. Chem- ical constituents and free radical scavenging activity of corn pollen collected from Apis mellifera hives compared to floral corn pollen at Nan, Thailand. BMC Complemen- tary and Alternative Medicine, 12(45). Fanali C., Dugo L. and Rocco A. 2013. Nano-liquid chro- matography in nutraceutical analysis: Determination of polyphenols in bee-pollen. Journal of Chromatography A, 1313, 270-274. Ferreres F., Pereira D.M., Valentão P. and Andrade P.B. 2010. First report of non-coloured flavonoids in Echium plantagineum bee pollen: differentiation of isomers by liq- uid chromatography/ion trap mass spectrometry. Rapid Communication in Mass Spectrometry, 24(6), 801-806. Frassinetti S., Caltavuturo L., Cini M., Della Croce C.M. and Maserti B.E. 2011. Antibacterial and Antioxidant Activ- ity of Essential Oils from Citrus spp. Journal of Essen- tial Oil Research, 23(1), 27-31. Fratini F.,  Turchi B.,  Gasperini M.,  Torracca B.,  Giusti M., Sagona S., Felicioli A. and Cerri D. 2014. Bee-gath- ered pollen loads suspension: preliminary assessment of interaction with microbial growth for a potential em- ployment as a natural food additive. Journal of Microbi- ology., Biotechnology and Food Sciences, 3(6), 467-469. Jørgensen K., Stapelfeldt H. and Skibsted L.H. 1992. Flu- orescence of carotenoids. Effect of oxygenation and cis/ trans isomerization. Chemical Physics Letters, 190(5), 514-519. Karoui R., Dufour E., Bosset J-O. and De Baerdemaeker J. 2007. The use of front face fluorescence spectroscopy to classify the botanical origin of honey samples produced in Switzerland. Food Chemistry, 101, 314-323. Ketkar S.S., Rathore A.S., Lohidasan S., Rao L., Paradkar A.R. and Mahadik K.R. 2014. Investigation of the nutra- ceutical potential of monofloral Indian mustard bee-pol- len. Journal of Integrative Medicine, 12, 379-389. Kim D.O., Chun O.K., Kim Y.J., Moon H.Y. and Lee C.Y. 2003. Quantification of polyphenolics and their antioxi- dant capacity in fresh plums. Journal of Agricultural and Food Chemistry, 51(22), 6509-6515. Kroyer G. and Hegedus N. 2001. Evaluation of bioactive prop- erties of pollen extracts as functional dietary food supple- ment. Innovative Food Science and Emerging Technolo- gies, 2(3), 171–174. Kulmyrzaev A.A., Levieux D. and Dufour E. 2005. Front- face fluorescence spectroscopy allows the characteriza- tion of mild heat treatments applied to milk. Relations with the denaturation of milk proteins. J. Agric. Food Chem., 53, 502-507. Lakowicz J.R. 2007. Principles of fluorescence spectrosco- py. (3rd ed.). New York: Kluwer Academic/Plenum Pub- lishers, (Chapter 3). LeBlanc B.W. Davis O.K., Boue S., DeLucca A. and Deeby T. 2009. Antioxidant activity of Sonoran Desert bee-pol- len. Food Chemistry, 115, 1299-1305. Lee J., Durst R.W. and Wrolstad R.E. 2005. Determination of total monomeric anthocyanin pigment content of fruit juices, beverages, natural colorants, and wines by the pH differential method: collaborative study. Journal of AOAC International, 88(5), 1269-78. Leja M., Mareczek A., Wyz . golik G., Klepacz-Baniak J. and Czekońska K. 2007. Antioxidative properties of bee-pol- len in selected plant species. Food Chemistry, 100(1), 237-240. Lucchesi D., Russo R., Gabriele M., Longo V., Del Prato S., Penno G. and Pucci L. 2014. Grain and bean lysates im- prove function of endothelial progenitor cells from human peripheral blood: involvement of the endogenous antioxi- dant defenses. PLoS One 17;9(10):e109298. doi Mărghitaş L.A., Stanciu O.G., Dezmirean  D. S., Bobiş  O., Popescu O., Bogdanov S. and Campos M.G. 2009. In vit- ro antioxidant capacity of honeybee-collected pollen of se- lected floral origin harvested from Romania. Food Chem- istry, 115, 878-883. Morais M., Moreira L., Feás X. and Estevinho L. M. 2011. Honeybee-collected pollen from five Portuguese Natural Parks: palynological origin, phenolic content, antioxidant properties and antimicrobial activity. Food Chemical Tox- icology, 49(5), 1096-1101. Ninfali P., Mea G., Giorgini S., Rocchi M. and Bacchiocca M. 2005. Antioxidant capacity of vegetables, spices and dressings relevant to nutrition. British Journal of Nutri- tion, 93(2), 257-66. Nogueira C., Iglesias A., Feás X. and Estevinho L. M. 2012. Ital. J. Food Sci., vol. 27 - 2015 259 Commercial bee-pollen with different geographical ori- gins: a comprehensive approach. International Journal of Molecular Sciences, 13(9), 11173-11187. Pascoal A., Rodrigues S., Teixeira A., Feás X., and Estevin- ho L.M. 2014. Biological activities of commercial bee-pol- lens: antimicrobial, antimutagenic, antioxidant and anti- inflammatory. Food Chemical Toxicology, 63, 233-239. Pérez-Pérez E.M., Vit P., Rivas E., Sciortino R., Sosa A., Teja- da D. and Rodríguez-Malaver A.J. 2012. Antioxidant ac- tivity of four color fractions of bee-pollen from Mérida, Venezuela. Archivos Latino Americanos de Nutrición, 62(4), 375-380. Pinzauti M., Rondinini T., Niccolini L. and Felicioli A.  2002.  Investigation of the germinating potential of pollen transported by some bees.  Insect Social Life, 4, 107-114. Romani A., Mancini P., Tatti S. and Vincieri F. 1996. Polyphe- nols and polysaccharidies in Tuscan grapes and wines. Italian Journal of Food Science, 8(1), 13-24. Scarselli R., Donadio E., Giuffrida M.G., Fortunato D., Conti A., Balestreri E., Felicioli, R., Pinzauti M., Sabatini A.G. and Felicioli A. 2005. Towards royal jelly proteome. Pro- teomics, 5(3), 769-776. Silva T.M.S., Camara C.A., Lins A.C.S, Barbosa J.M., Silva E.M.S., Freitas B.M. and Santos F.A.R. 2006. Chemical composition and free radical scavenging activity of pollen Paper Received December 10, 2015 Accepted February 6, 2015 loads from stingless bee Melipona subnitida Ducke. Jour- nal of Food Composition and Analysis, 19 (6-7), 507–511. Singleton V.L., Orthofer R. and Lamuela-Raventos R. 1999. Analysis of total phenols and other oxidation sub- strates and antioxidants by mean of Folin-Ciocalteu re- agent. Methods in Enzymology, 299, 152-178. Zandomeneghi G. and Zandomeneghi M. 2009. Determina- tion of holo- and apo-riboflavin binding protein in avian egg whites through circular dichroism and fluorescence spectroscopy. Journal of Agricultural and Food Chemis- try, 57, 6510-6517. Zandomeneghi M., Carbonaro L. and Caffarata C. 2005. Flu- orescence of vegetable oils. Journal of Agricultural and Food Chemistry, 53, 759-766. Zandomeneghi M. 1999. Fluorescence of Cereal Flours. J. Agric. Food Chem., 1999, 47, 878-882. Yıldız O., Can Z., Saral O., Yuluğ E., Öztürk E., Aliyazıcıoğlu R., Canpolat S. and Kolaylı S. 2013. Hepatoprotective Potential of Chestnut Bee Pollen on Carbon Tetrachlo- ride-Induced Hepatic Damages in Rats. Evidence-Based Complementary and Alternative Medicine. 2013, Article ID 461478. doi.org/10.1155/2013/461478 Yildiz O., Karahalil F., Can Z., Sahin H. and Kolayli S. 2014. Total monoamine oxidase (MAO) inhibition by chestnut honey, pollen and propolis. Journal of enzyme inhibition and medicinal chemistry. 29, 690-694.