Fractionation of Phenolic Compounds Extracted from Propolis and Their Activity in the Yeast Saccharomyces cerevisiae

We have here investigated the activities of Slovenian propolis extracts in the yeast Saccharomyces cerevisiae, and identified the phenolic compounds that appear to contribute to these activities. We correlated changes in intracellular oxidation and cellular metabolic energy in these yeasts with the individual fractions of the propolis extracts obtained following solid-phase extraction. The most effective fraction was further investigated according to its phenolic compounds.


Introduction
Propolis is a resinous substance that is collected from certain plants by bees. The bees use it as a sealer in their hives and to prevent the decomposition of creatures that invade the hive and the bees can kill, but cannot remove. Although the composition of propolis varies, depending on the place and time of its collection [1], in general it contains resins and balsams (50%), waxes (30%), aromatic and essential oils (10%), pollen (5%) and other organic matter (5%) [2,3]. Propolis has a broad spectrum of biological activities, including antioxidative, anti-inflammatory, immunomodulatory, anticancer, antibacterial, antiviral, antifungal and antiparasitic effects [4,5]. Although propolis is a mixture of compounds, its pharmacological activities are reported to arise from its flavonoids and phenolic acids, and their esters [6].
In the present study, the activity of propolis was investigated using stationary phase Saccharomyces cerevisiae yeast as the model organism. In this system in our previous study, we showed that propolis decreases intracellular oxidation, with its antioxidative activity in yeast arising from only a part of it [7]. To better understand this antioxidative activity of propolis in yeast, we have here further investigated the activities of the phenolic compounds of propolis. Therefore, the objectives of this study were to: (1) fractionate a crude propolis extract using polarity-based solidphase extraction; (2) determine any correlations between the total phenolic content of individual propolis fractions and its antioxidant activity in vitro/in vivo; (3) determine any correlations between the total phenolic content of individual fractions of propolis and the cellular metabolic energy; and (4) identify the phenolic compounds of any fractions that promote changes in intracellular oxidation and cellular metabolic energy in yeast.

Ethics Statement
No specific permits were required for the described field studies. The location is not privately-owned or protected in any way. The field studies did not involve endangered or protected species.

Sample Preparation
Propolis was collected from bee hive in the Savinjska Valley in Slovenia during the autumn of 2010. The propolis (10 g) was extracted with 70% ethanol (100 mL) by mixing for 1 h at room temperature. The crude extract was recovered by centrifugation (3000 g, 5 min) and concentrated under vacuum using a rotary evaporator.

Determination of Total Phenolic Content
The total phenolic contents of the cleaned propolis EE96 and the eluates (EL30-EL70) were determined using the Folin-Ciocalteu method [8]. Each sample (50 mL) was mixed with distilled water (700 mL) and Folin-Ciocalteu reagent (125 mL). After 5 min, 20% (w/v) Na 2 CO 3 (125 mL) was added and the samples were mixed. Absorbance was measured after 90 min incubation in the dark at room temperature, using a Safire II microplate reader (Tecan) (l = 765 nm). The data expressed as g gallic acid equivalents (GAE)/L were estimated from the calibration curve using gallic acid as standard (absorbance at 765 nm).

Determination of DPPH Free Radical Scavenging Activity
The in vitro antioxidative activities of the cleaned propolis EE96 and the eluates (EL30-EL70) were evaluated using the 2,2diphenyl-1-picryl-hydrazyl (DPPH) free radical scavenging method [9]. Each sample (25 mL) was mixed with distilled water (225 mL) and DPPH solution (0.05 mg/L in 96% ethanol; 1 mL). The absorbance was measured after 30 min incubation in the dark at room temperature, using a Safire II microplate reader (Tecan) (l = 517 nm). The data expressed as g GAE/L were estimated from the calibration curve using gallic acid as standard (absorbance at 517 nm).

Yeast Strain and Cultivation
The yeast Saccharomyces cerevisiae ZIM 2155 was obtained from the Culture Collection of Industrial Microorganisms (ZIM) of the Biotechnical Faculty, University of Ljubljana, Ljubljana, Slovenia.
The yeast were cultivated in yeast extract (10 g/L), peptone (20 g/L), glucose (20 g/L) (YEPD) medium at 28uC and with agitation at 220 rpm, until their stationary phase. The yeast were then centrifuged at 4000 g for 3 min, washed once with phosphate-buffered saline (PBS) and resuspended in PBS at 1610 8 cells/mL. The yeast was further incubated at 28uC and 220 rpm, for 96 h.

Yeast Treatment
Following the 96 h of incubation in PBS, the yeasts were treated with either the cleaned propolis EE96 or with the particular eluates (EL30-EL70). After a further 2 h of incubation at 28uC and 220 rpm, samples were taken for the determination of: (1) intracellular oxidation; (2) cellular metabolic energy; (3) cell viability; and (4) cellular uptake of individual phenolic compounds.

Determination of Intracellular Oxidation
Intracellular oxidation was estimated using 29,79-dichlorodihydrofluorescein (H 2 DCF), which reacts with oxidants, thus revealing the presence of reactive oxygen species (ROS). This was given to the yeast as H 2 DCF diacetate, which penetrates the plasma membrane and is hydrolyzed inside cells by nonspecific esterases. The nonfluorescent H 2 DCF can then be oxidized to fluorescent 29,79-DCF, the levels of which are measured fluorimetrically [10].
After 2 h of treatment, the yeast from 2 mL cell suspensions were sedimented by centrifugation (14,000 g, 5 min), and washed three times with 50 mM potassium phosphate buffer (pH 7.8). The cell pellets were resuspended in 9 volumes of 50 mM potassium phosphate buffer (to 10%, v/v) and incubated at 28uC for 5 min. The ROS-sensing dye H 2 DCF diacetate was added  from a 1 mM stock solution in 96% ethanol, to a final concentration of 10 mM. After a 20 min incubation at 28uC and 220 rpm, the fluorescence of the yeast suspension was measured, using the kinetic mode of a Safire II microplate reader (Tecan).
The excitation and emission wavelengths of DCF were 488 nm and 520 nm, respectively. The data are expressed as fluorescence relative to control.

Determination of Cellular Metabolic Energy
The cellular metabolic energy was determined using the BacTiter-Glo TM Microbial Cell Viability Assay (Promega), according to the manufacturer's instructions. After 2 h of treatment, 100 mL cell suspension (washed three times with filtered PBS) at 1610 7 cells/mL, and 100 mL BacTiter-Glo TM reagent were placed in 96-well microplates and mixed. After 5 min the luminescence was measured using a Safire II microplate reader (Tecan). The data are expressed as luminescence relative to control.

Cell Viability Determination
Cell viability was measured as the cell membrane integrity, using LIVE/DEADH Funga Light TM Yeast Viability kits (Molecular Probes), according to the manufacturer's instructions. After 2 h of treatment, the yeast from 1 mL cell suspension were sedimented by centrifugation (14,000 g, 5 min), and washed three times with filtered PBS. A cell suspension in filtered PBS at 1610 6 cells/mL was prepared. Then 1 mL SYTOH 9 and 1 mL propidium iodide were added in the dark to 1 mL of cell suspension, and the samples were vortexed and incubated at 37uC for 30 min, with vortexing every 10 min. After the incubations, the fluorescence was measured using a Safire II microplate reader (Tecan). The excitation/emission wavelengths for these two dyes are 480/500 nm for SYTOH 9, and 490/635 nm for propidium iodide. The data are expressed as fluorescence normalized to OD 650 relative to the control.

Determination of Cellular Uptake of Phenolic Compounds
To study the cellular uptake of the phenolic compounds, the phenolic profile was determined in the PBS buffer immediately after the addition of 1% eluates (i.e. before yeast exposure) and after 2 h of exposure of yeast cells to eluates. The samples were centrifuged (4000 g, 3 min) and the supernatants obtained were first cleaned using SPE (according to procedure described for EE96) and then analyzed using LC-DAD, to obtain the phenolic profile.

Liquid Chromatography-diode Array Detection Analysis
The samples were diluted 20-fold in 1% formic acid in 50% methanol and analyzed by liquid chromatography-diode array detection (LC-DAD). The LC system consisted of an Agilent 1100

Liquid Chromatography-mass Spectrometry Analysis
To identify the phenolic compounds, liquid chromatographymass spectrometry (LC-MS) was used. The samples were diluted 20-fold in 1% formic acid in 50% methanol and analyzed by LC-MS. A Micromass Quattro Micro mass spectrometer equipped with an electrospray ionization source was operated in negative ion mode (Waters, Milford, MA, USA). The mass spectra were recorded with the following operating parameters: capillary voltage, 3.0 kV; cone voltage, 25 V; extractor, 5 V. The source temperature was 100uC, and the desolvation temperature was 350uC. The cone gas flow was set at 50 L/h, while the desolvation gas flow was set to 400 L/h, and the collision energy at 20 V. The

Statistical Analysis
The data are expressed as means 6S.D., as determined from triplicate analysis. Duncan's multiple range tests, at P,0.05, determined the significant differences among the means.

Results and Discussion
The present study is a continuation of our previous study [7], where we showed that with propolis, the antioxidative activity is not related to the whole propolis extract, but to only a part of it. To better understand which phenolic compounds are responsible for the antioxidative activity of propolis in cells, here fractionation of phenolic compounds in a crude ethanolic propolis extract was performed using SPE (Fig. 1). Five sequential eluates that varied in polarity were obtained, according to the ethanol used for the elution (30% to 70%): EL30, EL40, EL50, EL60 and EL70, from more to less polar, respectively. For the cleaned ethanolic extract of propolis (EE96) as well for the EL30 to EL70 eluates, the phenolic profiles were determined using LC-DAD (Fig. 2). As expected, these profiles differed between the eluates.
Here, as well as the phenolic profile, the total phenolic content and free radical scavenging activity of EE96 and EL30 to EL70 were determined ( Table 1). The highest total phenolic content was seen for eluates EL60 and EL70 (0.82 g GAE/L and 0.83 g GAE/ L, respectively), which also showed the highest in vitro antioxidative activities, measured as the free radical scavenging activity (1.20 g GAE/L and 1.22 g GAE/L, respectively). Thus, eluates EL30 to EL50 showed lower total phenolic contents (0.26 g GAE/L to 0.39 g GAE/L), which was reflected in their lower in vitro antioxidative activities (0.18 g GAE/L to 0.33 g GAE/L). The sum of the total phenolic content in these eluates was higher compared to the total phenolic content of EE96. This was expected, as the total volume of ethanol used for elution of the eluates (EL30 to EL70) was five-fold higher than the volume of ethanol used for elution of the cleaned propolis EE96 (Fig. 1). Therefore, some of the phenolic compounds from the crude propolis extract were lost in the process of obtaining EE96.
Fractionation of these compounds extracted from propolis has been reported previously [11,12], where nanofiltration [11] and supercritical fluid extraction [12] were used. In the case of nanofiltration, the compounds from an ethanolic extract of propolis were fractionated according to their molecular weights. The DPPH test was performed, and it was shown that the antioxidative activity was proportional to the flavonoid content of the fractions. The same was concluded for the fractions obtained using supercritical fluid extraction.
On the bases of in vitro studies, predictions about the activities of the phenolic compounds in the cell are usually done. However, this extrapolation can be misleading and additional in vivo studies need to be conducted. Therefore, the antioxidative activities of the cleaned propolis EE96 and the eluates (EL30-EL70) that showed free radical scavenging activities were here investigated using an appropriate model organism. These antioxidative activities were determined by measuring intracellular oxidation using the yeast S. cerevisiae. The yeast was treated for 2 h with EE96 or the EL30 to EL70 eluates. With EE96 used in the yeast suspensions at concentrations from 0.00125 g GAE/L to 0.02 g GAE/L, there was a dosedependent decrease in the intracellular oxidation (Fig. 3). This decrease continued to 0.01 g GAE/L EE96, after which it reached a plateau.
Next, the yeast suspensions were treated for 2 h with the eluates EL30 to EL70 at 1% (v/v). Preliminary experiments showed that the solvent of 1% (v/v) ethanol has no effects on intracellular oxidation in these yeast suspensions (data not shown). The greatest decreases in the intracellular oxidation were seen when the EL70 eluate was added, followed by eluates EL60 and EL50. In contrast, eluate EL30 showed a trend to an increase in intracellular oxidation, whereas no statistical difference compared to the control was seen for EL40 (Fig. 4). Eluates EL60 and EL70 have the same total phenolic content and free radical scavenging activities (Table 1), although they have different in vivo antioxidative activities determined by 29,79-DCF. Eluate EL70 decreased the intracellular oxidation by 66%, whereas eluate EL60 decreased it by 42%. The diversity of the phenolic compounds in these eluates would be a reason for this difference. Eluates EL30 and EL50 both have free radical scavenging activity, although they show different effects on yeast intracellular oxidation. In the case of eluate EL50, the intracellular oxidation decreased by 16%, whereas eluate EL30 increased it by 18%. Therefore, these data indicate that in these yeasts, the free radical scavenging activities of these eluates do not always correlate with their antioxidative activities.
As well as the antioxidative activities, insights into other biological activities are also of importance. Therefore, we investigated the cellular metabolic energy and the viability of these treated yeast.
For the cellular metabolic energy, eluates EL60 and EL70 again showed differences compared to the others, as seen in Figure 5. An increase in cellular metabolic energy was seen when eluates EL60 and EL70 were added to the yeast suspensions at the final concentration of 1% (v/v). Despite the same concentration of total phenolics in these two eluates (Table 1), eluate EL70 showed a greater increase in cellular metabolic energy (21%) than seen for EL60 (6%). This difference in effects between these eluates was already seen in the case of intracellular oxidation. With eluates EL30 to EL50, there were small, but significant, decreases in the cellular metabolic energy compared to the control (Fig. 5). As also seen in Figure 5, the cellular metabolic energy increased (19%) in the yeast treated with the cleaned propolis EE96 at the same final concentration of 1% (v/v).
Cell viability was measured as cell-membrane integrity using the nucleic-acid-specific dyes SYTOH9 and propidium iodide in the LIVE/DEADH Funga Light TM Yeast Viability kits (Molecular Probes). These data showed no significant changes in the viability of the yeast treated with the cleaned propolis EE96 and the EL30 to EL70 eluates at this final concentration of 1% (v/v) (Fig. 6). It was previously demonstrated that Brazilian propolis induces cell death in S. cerevisiae [13]. As stated, no such effect was observed in our study, where stationary yeast cells as a model organism were used. We also tested activity of eluate EL70 in the exponential growth phase, where the same effect (decreased intracellular oxidation and no changes in cell viability) as in stationary growth phase was observed (data not shown). Thus, activity of eluate EL70 is growth-phase independent. Therefore, different results might be due to the differences in the composition of propolis and concentration of its phenolic compounds, since it is known that European propolis differ from Brazilian propolis [1].
As the antioxidative activities of phenolic compounds in the cell have been shown to be connected with their cellular uptake [7], the uptake from the particular eluates by these yeasts was also investigated here. The LC-DAD profile of the incubation medium before and after exposure of these yeast to the various eluates was determined. Here, some of the LC-DAD peaks of the eluate EL70 samples were decreased after treatment (Fig. 7: retention 19.6160.01 min), whereas no changes in these LC-DAD profiles were seen before and after the treatments with eluates EL30 to EL50 (data not shown). These decreased LC-DAD peaks in eluate EL70 were also seen in EL60, but to a lesser extent. This might be a reason for the higher antioxidative activity of eluate EL70 compared to eluate EL60 (Fig. 4). Despite no apparent cellular uptake from eluate EL30, it increased intracellular oxidation  (Fig. 4). This might be explained in terms of many phenolic compounds being unstable in cell-culture media, whereby they can undergo rapid oxidation, to generate hydrogen peroxide and other ROS [14][15][16][17]. Among these products, hydrogen peroxide can pass through cell membranes by passive diffusion, which would contribute to an increase in intracellular oxidation [18].
Thus, we have shown initially that there is cellular uptake, a decrease in intracellular oxidation, and an increase in cellular metabolic energy in yeast treated with eluate EL70. This led to further investigation of the individual phenolic compounds in this EL70 fraction of the crude propolis extract. Here we used LC-MS/MS and focussed on the LC-DAD retention times from 17 min to 20 min.
To our knowledge, this is the first study where solid-phase extraction has been used for fractionation of phenolic compounds from crude propolis, with the investigation of the various activities of the individual eluates investigated in yeast as a model organism: intracellular oxidation, cellular metabolic energy, and cell viability. Additionally, we have identified the cellular uptake of individual phenolic compounds that may contribute to the antioxidative and cellular metabolic energy effects of propolis in yeast.