Antioxidant and dual dose-dependent antigenotoxic and genotoxic properties of an ethanol extract of propolis

Propolis is a resinous product made by honeybees from plant-derived materials, with high content of polyphenols associated to several beneficial bioactivities with potential use as a natural food additive for preservation and as a functional food ingredient. A Portuguese propolis ethanol extract (C.EE) protected Saccharomyces cerevisiae cells from loss of viability upon exposure to H2O2, both in coand in preincubation experiments. Results obtained with the comet assay suggest that lower concentrations are antigenotoxic while at higher concentrations a genotoxic effect prevails, which correlates with the cytotoxicity of high concentrations of C.EE. Flow cytometry analysis with dichlorofluorescein indicates that C.EE induced intracellular antioxidant activity in vivo. Overall the results suggest that C.EE is antigenotoxic but is also toxic at higher concentrations. This dual effect could be explained by the presence of compounds known to interfere with DNA synthesis and/or cell proliferation, such as caffeic acid phenethyl ester (CAPE) and chrysin, together with antioxidants, like kaempferol, pinobanksin and pinocembrin.


Introduction
Propolis is a resinous mixture produced by honeybees from exudates of buds and the bark of plants such as poplar (Populus spp.), birch (Betula spp.), beech (Fagus spp.), horse chestnut (Aesculus hippocastanum), alder (Alnus spp.), Brazilian rosemary (Baccharis dracunculifolia), eucalyptus (Eucalyptus spp.) and Brazilian pine (Araucaria angustifolia).It is generally accepted that propolis is used in the hive, mainly as a construction and antiseptic material, for repairing mechanical damage and avoiding microbial infections.
Propolis is chemically diverse and its composition varies geographically according to the ora and climate, the season of collection, as well as with the species of the producing bees. 1 Propolis compounds belong to very different chemical groups, such as polyphenols (avonoids, including avones, avonones, avonols, dihydroavonols and chalcones), phenolic acids and their esters, terpenoids, steroids and amino acids. 2 Most of these compounds, in particular phenolics such as caffeic acid derivatives and avonoids, have been associated with propolis biological activities, namely antimicrobial, 3 cytotoxic and hepatoprotective, 2 radioprotective, antimutagenic, 4 antioxidant 5 and as scavenger of free radicals. 2he widely recognized properties of propolis have been promoting its extensive use in nutraceutics, cosmetics and health care.In addition, the consumers' widely good acceptance of the incorporation of natural products in foods and medicines increases the potential use of propolis as food functional ingredient and as preservative.Hence, propolis has been attracted the attention of researchers to formulate new natural food functional ingredients 6,7 and was also included in several food products with benecial effects as preservative. 8,9Still in many countries, honey producers have disregarded propolis due to low yields and lack of knowledge of its economic potential as a valuable co-product.Hence, the demonstration of bioactivities but also cell-protective properties for the development of functional foods and health-care products based on propolis are a major goal.
The antioxidant activity of avonoids of propolis, in samples from very different provenances, has been reported as a mechanism of protection of genomic DNA against reactive oxygen species. 10,11Also, genotoxic effects of propolis extracts and of some of their constituents can be found in the literature, 12 having been attributed to the pro-oxidant activity of avonoids. 13This pro-oxidant activity has been shown to be directly dependent on concentration and is mediated by increased a Centre for the Research and Technology of Agro-Environmental and Biological Sciences (CITAB), Department of Biology, University of Minho, Campus de Gualtar, 4710-057 Braga, Portugal.E-mail: ruipso@bio.uminho.pt;Fax: +351 253604319; Tel: +351 253601512 b Centro de Apoio Tecnológico Agro Alimentar (CATAA), Zona Industrial de Castelo Branco, Rua A, 6000-459 Castelo Branco, Portugal c Chemistry Research Center (CQVR), University of Trás-os-Montes e Alto Douro, Quinta de Prados, 5000-801 Vila Real, Portugal

Experimental
Yeast strain, media and growth conditions In all experiments the haploid Saccharomyces cerevisiae strain BY4741 (MATa his3D1 leu2D0 met15D0 ura3D0) was used.Cultures were grown on liquid YPD medium (1% w/v yeast extract, 1% w/v peptone and 2% w/v glucose), in an orbital shaker at 30 C and 200 rpm.Growth was monitored by optical density at 600 nm (OD 600 ).

Propolis ethanol extract
Raw propolis was collected in August 2010 from an apiary set in the region of Côa (Beira Alta, Portugal).A 29 g propolis sample was incubated with 100 mL absolute ethanol in an orbital shaker at 100 rpm, at room temperature, in the dark, for 24 h.The solution was then ltered (Whatman nr.4) and the residue was re-dissolved in 100 mL absolute ethanol and extracted again.This procedure was repeated three times.The ltrates were pooled and dried in a rotary evaporator (Buchi RE 121), at 40 C, under vacuum and gentle rotation (50 rpm), yielding the ethanol extract of Côa propolis (C.EE), which was stored in the dark at 4 C until use.Working solutions at the desired concentrations were prepared in ethanol immediately before use.

Viability assays
Yeast cells from exponentially growing cultures were harvested by 2 min centrifugation at 6000 Â g, 4 C, washed twice with the same volume of sterilized deionized H 2 O at 4 C and suspended in the same volume of S buffer (1 M sorbitol, 25 mM KH 2 PO 4 , pH 6.5).The suspension was incubated at 30 C, 200 rpm, in the presence of 5 mM H 2 O 2 , aer which aliquots were harvested along time, serially diluted to 10 À4 , spread on YPD plates and incubated at 30 C for 48 h.The percentage of colony-forming units (CFUs) was calculated at each time-point taking as reference the number of colonies obtained before the addition of H 2 O 2 .In pre-incubation experiments, cells were incubated with C.EE for 20 min, at 30 C, 200 rpm and were washed and suspended in S buffer, as described above.

Comet assay
The yeast comet assay was performed as described before. 15riey, cells from exponentially growing cultures were harvested by centrifugation at 18 000 Â g, for 2 min at 4 C, washed twice with ice-cold deionized H 2 O, suspended in S buffer containing 2 mg mL À1 zymolyase (20 000 U g À1 ; ImmunO™-20T) and 50 mM b-mercaptoethanol and incubated at 30 C, 200 rpm for 40 min, in order to obtain spheroplasts.Aer washing with S buffer, spheroplasts were suspended in S buffer containing 10 mM H 2 O 2 , incubated for 20 min at 4 C and subsequently washed with S buffer.Spheroplasts were embedded in 1.5% low melting agarose (w/v in S buffer) at 35 C, spread onto glass slides previously layered with 0.5% (w/v) normal melting agarose, covered with a cover slip and incubated on ice in order to solidify the agarose.The cover slips were then removed and the unwinding of DNA and protein denaturation were made with ice-cold lysing buffer (30 mM NaOH, 1 M NaCl, 50 mM ethylenediamine tetraacetic acid (EDTA), 10 mM Tris-HCl, 0.05% (w/v) lauroylsarcosine, pH 10) for 20 min.Samples were subsequently stabilized in ice-cold electrophoresis buffer (30 mM NaOH, 10 mM EDTA, 10 mM Tris-HCl, pH 10) for 20 min.Glass slides were placed in an electrophoresis chamber and samples were exposed to 0.7 V cm À1 for 10 min, at 4 C, to allow unwound DNA to move out of the nucleoids.The gels were neutralized with 10 mM Tris-HCl buffer, pH 7.4, for 10 min at 4 C, samples were xed, rstly in 76% (v/v) ethanol and then in 96% (v/v) ethanol, both for 10 min, and the slides were dried at room temperature.Aer staining with 10 mL of GelRed™ (diluted 3300 fold from the stock solution; Biotium) comets were visualized by uorescence microscopy (Leica DMB 5000 with a P&B, Leica, DFC 350FX digital camera) and tail length was measured from at least 20 comets per treatment with the CometScore soware.Pre-, co-or post-incubation with C.EE were performed in the spheroplasts suspension as described for viability assays, before embedding in low melting agarose.Controls were also prepared as described in the same experimental procedure.

Flow cytometry
Cells from exponentially growing cultures were harvested as above, washed twice with the same volume of ice-cold PBS (137 mM NaCl, 2.7 mM KCl, 4.3 mM Na 2 HPO 4 , 1.47 mM KH 2 PO 4 , pH 7.4), diluted to an OD 600 of 0.02 and 500 mL were collected for measurement of autouorescence.Cells were loaded with 50 mM dichlorouorescein diacetate (H 2 DCFDA) by incubation at 30 C, 200 rpm, during 1 h in the dark, aer which they were washed twice with the same volume of ice-cold PBS.In coincubation experiments, aliquots of 1 mL were mixed with C.EE and 10 mM H 2 O 2 , for 20 min, at 30 C, 200 rpm, in the dark.In pre-and post-incubation experiments cells were incubated, respectively, with the extract and subsequently with H 2 O 2 or with H 2 O 2 and subsequently with C.EE. Washing steps were included between each incubation step, as described above.
Approximately twenty thousand cells of each sample were analysed by ow cytometry in an Epics® XLTM cytometer (Beckman Coulter) equipped with an argon-ion laser emitting a 488 nm beam at 15 mW.Green uorescence was collected through a 488 nm blocking lter, a 550 nm long-pass dichroic and a 225 nm band-pass lter.Data were analysed and histograms were made with the Flowing Soware.Controls were included by replacing C.EE by the same volume of ethanol and H 2 O 2 by PBS.

Chemical analysis of C.EE
Quantication of total polyphenols.The content in total polyphenols of the extract was determined by the Folin-Ciocalteu colorimetric method, 16 with some modications.Three hundred milligrams of C.EE solution (0.5 mg g À1 nal concentration) were mixed with 2.0 g deionized H 2 O, 200 mg Folin-Ciocalteu reagent (Panreac, Barcelona, Spain), 2.0 g of 10% (w/v) Na 2 CO 3 , and deionized H 2 O to complete 10.0 g nal mass of the mixture.The reducing power of the polyphenols in the mixture was evaluated using the OD 760 measured aer 1 h incubation at room temperature.C.EE polyphenol content was calculated using gallic acid as standard and results were expressed as milligrams of gallic acid equivalents per gram of C.EE (mg GAE per g).
Quantication of avonoids.Total avonoid content in the extract was determined using a method described by Woisky and Salatino. 17Five hundred milligrams of 2% (w/v) AlCl 3 $6H 2 O were added to 300 mg of C.EE and ethanol was used to complete 10.0 g of mixture nal mass.Aer 30 min incubation at room temperature, the OD 420 of the mixture was measured and avonoids content was calculated by comparison with the standard quercetin (1.2 mg g À1 ) and expressed as milligrams of quercetin equivalents per gram of C.EE (mg QE per g).
LC-MS analysis.One hundred milligrams of C.EE were dissolved with 1 mL of 80% ethanol at 70 C and ltered through a 0.22 mm nylon lter prior to injection.Standards for gallic acid, protocatechuic acid, chlorogenic acid, vanillic acid, caffeic acid, syringic acid, ferulic acid, o-coumaric acid, apigenin, and kaempferol were acquired from Sigma-Aldrich Co. LLC.Luteolin and gentisic acid standards were acquired from Extrasynthese, France.The chromatographic system consisted of an Agilent 1200 series equipped with a triple quadrupole mass spectrometer Agilent 6400.A Sorbax SB-C18 (50 mm Â 4.6 mm i.d.Â 1.8 mm particle diameter -Agilent Technologies) column was used for the separation at a ow rate of 0.7 mL min À1 , at 30 C. Elution was performed by means of a gradient of 0.1% formic acid (eluent A) and acetonitrile (eluent B).The gradient was as follows: started at 10% B, 20% B at 10 min, 40% B at 40 min, 60% B at 60 min, 90% B at 80 min, at 81 min return to initial conditions and stabilization for 9 min.Electrospray ionization (ESI) was performed with a nitrogen ow of 10 L min À1 at 300 C and the MS detector was operated in MS2-Scan scan type in the range 80-1000 Da, and negative mode was selected.The capillary voltage was set to 4.0 kV, the quadrupole temperatures 100 C, fragmentor energy was 145 V, and cell accelerator voltage was 7 V. Data were acquired and analysed using Masshunter Workstation Soware (version B.04.00) Agilent Technologies.
For MS/MS conrmation the same equipment and chromatographic conditions were used.The MS detector was operated in product ion scan type, selecting the precursor ions and performing a scan of the fragments in the range 80-500 Da and negative mode was selected.The capillary voltage set to 4.0 kV, the quadrupole temperatures were 100 C, fragmentor energy was 135 V, cell accelerator voltage was 7 V and collision energy was 15 V. Compounds were identied based on standards retention times and by comparison of the ESI-MS/MS data with the MS/MS data published in the literature.

Statistical analysis
All experiments were done in triplicate and the results are presented as mean AE standard deviation (SD).For the comet assay, the mean was obtained from the mean of three independent experiments.One-way analysis of variance (1-way ANOVA) was used to evaluate treatment effects and Tukey's test was used to perform comparisons between each treatment with the respective control.Asterisks (or § §) indicate statistically signicant differences: * means 0.01 < p # 0.05, ** means 0.001 < p # 0.01, and *** means p # 0.001.

C.EE polyphenolic content and chemical prole
Polyphenols constitute an important group of biologically active compounds abundant in propolis samples.Particularly, the avonoids have been associated with the antioxidant properties exhibited by plant and plant-based products, such as propolis, 5 but also some phenolic acids and their esters possess antioxidant activity. 18Hence, to assess C.EE bioactivity potential, an initial quantitative characterization involved the determination of total polyphenols and total avonoids contents (Table 1).
Total phenols of C.EE were in the range reported in the literature (120-443 mg GAE per g extract), 19 but in the lower third of the rank.The values are similar to those found for Portuguese samples from the centre of Portugal 20 but are also comparable to those of very distant places, such as India; 5 Anhui, China; 21 and to the red propolis from Brazil. 22Although still in the range for total avonoids (25-140 mg QE per g extract) also low levels were obtained for C.EE (Table 1).Again, this content is similar to that of some Portuguese samples from Alentejo, 20 but also to red propolis from Cuba 19 and to some samples collected in Anhui, China. 21he sample was further characterized qualitatively by LC-MS analysis (Table 2 and ESI Fig. 1 †).][25] Similarly to what has been described for other propolis samples, C.EE is mostly constituted by avonoids, phenolic acids and their esters, being avones, avonols and avanones the main avonoid groups as previously reported. 26In this sample 43 phenolic compounds were identied: benzoic and hydroxybenzoic acids,  24 which is the most common in Europe, China and Argentina, 25 indicating that, as expected, the C.EE sample ts into this propolis type.In particular, high relative amounts of chrysin, pinocembrin and galangin denote that poplar is an important source of raw material. 19owever, the phenolic composition of this sample displays some distinct features like the presence of uncommon phenolic acids, such as 3,4-dihidroxy vinylbenzene, and the rare non-avonoid phenolic ellagic acid, reported recently for Portuguese propolis. 24espite the low total avonoid content (Table 1), that may partially be explained by an underestimation of avanones and dihydroavonols by the use of the aluminium chloride-ethanol method, the results obtained from chemical characterization, revealing avonoid diversity and in high relative abundance, suggest that C.EE may exhibit some of the widely reported propolis biological activities, namely the antioxidant.Indeed, avonoids (such as pinobanksin, kaempferol, pinocembrin and galangin) and phenolic acids (caffeic acid and CAPE) have been the classes of phenolic substances most extensively reported to have antioxidant activity.On the other hand, a close correlation between antioxidant potential and the total content in avonoids is not always observed (see the reviews 19,21 ), the presence of specic compounds in propolis composition being more indicative of its bioactivity prole.
Relevant bioactivities have been described for most of these compounds, such as anti-proliferative, anti-tumor, antimicrobial and antioxidant, but also pro-oxidant effects were reported for some avonoids, depending on the redox state of the environment. 27For this a set of assays were performed to study the antioxidant properties of C.EE at the cell and DNA levels using the yeast model and different technical approaches.

Effects of C.EE on cell viability under oxidative stress conditions
To test the hypothesis that C.EE is capable of protecting cells against oxidative stress, yeast cells were pre-incubated with C.EE and their viability was assessed, as CFUs, in the presence of H 2 O 2 for 90 min.Also, co-incubation and post-incubation experiments were performed to investigate the antioxidant activity of the extract in the presence of the stressor agent and its participation in cell recovery from oxidative damage, respectively.
When cells were pre-incubated with C.EE at 100 mg mL À1 (Fig. 1A) and 300 mg mL À1 (Fig. 1B), viability loss was reduced when compared to control cells (pre-treated with ethanol) (p ¼ 0.0018 and p ¼ 0.0011 aer 60 min incubation, respectively), during the 90 min of exposure to 5 mM H 2 O 2 .Cells treated only with S buffer or pre-treated with ethanol or C.EE, and aerwards with water, displayed a nearly constant viability throughout the experiment (Fig. 1A and B).These results suggest that C.EE triggers a protection mechanism in yeast cells that allows increased resistance against oxidative stress.
To investigate a direct effect on the oxidant agent (H 2 O 2 ) and/or early antioxidant effects of C.EE we have determined the viability in co-incubation experiments.As depicted in Fig. 1D, 100 mg mL À1 C.EE signicantly decreased the rate of viability loss (p ¼ 0.0097 aer 60 min incubation) of cells exposed to 5 mM H 2 O 2 , while 25 mg mL À1 and 300 mg mL À1 C.EE had no signicant effect (Fig. 1C and E, respectively).In addition, unlike cells treated with 2% ethanol or with S buffer, cells incubated with 100 mg mL À1 C.EE (p ¼ 0.0029) or 300 mg mL À1 C.EE (p ¼ 0.005) displayed increased loss of viability (Fig. 1D  and E).The differences regarding pre-incubation experiments with the same concentrations of C.EE are consistent with the fact that while cells were in contact with C.EE for 20 min in preincubation experiments, in co-incubation experiments cells contacted with C.EE for 90 min.Together these observations suggest that there is a range of concentrations where C.EE protects cells against oxidative stress and a concentration threshold above which C.EE has a toxic effect on yeast cells.
The hypothesis that propolis could also improve recovery of cells aer oxidative shock was tested by performing postincubation experiments, which are based on a previous treatment with H 2 O 2 for 20 min, wash of the cells to remove the toxicant and a subsequent 20 min treatment with C.EE.As expected, considering the short term incubation, cells without any treatment or treated only with C.EE or 2% ethanol (solvent control) showed a nearly constant survival rate throughout the experiment (Fig. 1F-H).Aer treatment with 5 mM H 2 O 2 for 20 min, when yeast cells were incubated with 2% ethanol or C.EE (see Fig. 1F-H aer minute 20), the viability loss rate did not change signicantly (Fig. 1F and G), except in the experiment with 300 mg mL À1 C.EE, which promoted a reproducible although not statistically signicant faster loss of viability when compared with the respective control (Fig. 1H).These results proved that C.EE could not improve recovery from oxidative stress but also that at the highest concentration tested (300 mg mL À1 ; Fig. 1H) C.EE increased the loss of viability.In fact, when comparing viability loss aer 20 min in the presence of 300 mg mL À1 C.EE in all experiments (time-point 40 min in Fig. 1H), in the post-incubation experiment the C.EE effect was higher (Fig. 1B, E and H).These results may be explained by an exacerbated pro-oxidant effect of C.EE in cells that were exposed previously to oxidative challenge by H 2 O 2 (Fig. 1H).It should not be excluded however non-oxidative mechanisms of C.EE triggering general stress response in yeast cells.
Tsai et al. 13 proposed that avonoids from propolis could be pro-oxidant and genotoxic by reaction with metal ions.Besides C.EE avonoid total content being relatively low, it is also known that some avonoids, like quercetin, are more prone to oxidation than others, especially at physiological pH. 28Therefore the composition must also be taken into account when analysing these properties and considering that quercetin is one of C.EE components (Table 2) it is tempting to explain the decrease of cell viability by a signicant pro-oxidant activity of C.EE.This is also supported by post-incubation experiments with the highest concentration tested (300 mg mL À1 ; Fig. 1H), where the increased rate in viability loss with C.EE aer a previous incubation period with H 2 O 2 could be explained by further accumulation of oxidative damage in the cells.In studies also with propolis of Portuguese origin, cytotoxicity was observed in human tumour cell lines.Our results could help to explain such activity.In all cases however, a link between cytotoxicity and pro-oxidant activity was not reported, being the mechanism not studied 29,30 or associated with a disturbance of the glycolytic metabolism. 31lobally, the results suggest that C.EE can protect yeast cells against an induced oxidative stress, possibly by direct scavenging and reduction of free radicals and by improving adaptation of cells to stress.In addition C.EE can have cytotoxic effects at higher concentrations, especially in already compromised yeast cells by a previous exposure to oxidative stress.

C.EE protects yeast cells from DNA damage induced by H 2 O 2
It is conceivable that at least some of the antioxidant effects reported for propolis 5,32 have an impact in genomic DNA integrity under oxidative challenges by avoiding DNA damage.As the propolis sample herein studied also displayed antioxidant activity, it was decided to investigate its antigenotoxicity.
Yeast spheroplasts were pre-treated for 20 min with 25 mg mL À1 , 100 mg mL À1 or 300 mg mL À1 C.EE in S buffer, to maintain osmotic protection, and then exposed to 10 mM H 2 O 2 .Exposure to H 2 O 2 , aer pre-incubation with 2% ethanol (extract solvent), increased dramatically comet tail length (Fig. 2A) conrming that the H 2 O 2 concentration used in the experiments was genotoxic.However, when yeast spheroplasts were pre-treated with C.EE before exposure to H 2 O 2 , a statistically signicant decrease in comet tail length was observed even for the lowest concentration tested (Fig. 2A).This result indicates that C.EE has a potent antigenotoxic activity by protecting cells against oxidative-induced DNA damage and is consistent with the increase in yeast viability previously observed in the preincubation experiment (Fig. 1A and B).
Acknowledging the cytotoxic properties of C.EE, a possible explanation for its antigenotoxic effect is that a mild genotoxic effect of C.EE could trigger adaptation of cells against oxidative stress by H 2 O 2 .To investigate this hypothesis, spheroplasts were incubated with different C.EE concentrations and then with S buffer instead of H 2 O 2 .A statistically signicant increase in comet tail length was observed only with 300 mg mL À1 C.EE when compared with cells treated with 2% ethanol (Fig. 2B), indicating that C.EE has also genotoxic activity.DNA damage induced by 300 mg mL À1 C.EE alone (Fig. 2B) is similar to that found in C.EE pre-treated cells when subjected to H 2 O 2 (Fig. 2A), suggesting that the potential for genoprotection is constrained by its genotoxicity.These results seem to support the hypothesis that C.EE may exert its antioxidant/antigenotoxic activity through a mild geno-insult enabling cells to adapt to subsequent genotoxic oxidative stresses like that induced by H 2 O 2 .
Similarly to what was done in the cell viability assays, coincubation experiments using yeast spheroplasts were performed to investigate if C.EE can protect DNA from damage under oxidative stress conditions, presumably by direct ROS detoxication.In this experiment spheroplasts were incubated with C.EE and 10 mM H 2 O 2 for 20 min, aer which DNA damage was analysed as before.As observed in pre-incubation experiments, H 2 O 2 increased dramatically comet tail length (Fig. 2C) and when yeast spheroplasts were treated with H 2 O 2 and C.EE simultaneously, a signicant decrease of DNA damage was observed (Fig. 2C).Hence these results suggest that reducing/scavenging activities by some C.EE compounds may be protecting cells against H 2 O 2 -induced oxidative DNA damage.Genotoxicity of C.EE was also investigated aer incubation of spheroplasts with propolis extract without H 2 O 2 and omitting the subsequent step of incubation of 20 min that allows DNA damage repair.A signicant increase in tail length was observed for all concentrations tested (with S buffer) when compared with the ethanol control (Fig. 2D), indicating that C.EE acted as a genotoxic agent to S. cerevisiae cells.These results together with the decrease of viability observed for 100 mg mL À1 and 300 mg mL À1 C.EE (Fig. 1D and E) indicate that DNA damage is possibly involved in C.EE-induced loss of cell viability.In fact, propolis avonoids, as effective scavengers of free radicals and other reactive species in vitro, may explain a decrease in oxidative DNA damage in vivo. 33So, antigenotoxicity and genotoxicity were observed in both pre- incubation and co-incubation experiments, which strongly supports a dual activity of C.EE propolis regarding DNA integrity.A similar behaviour regarding genotoxic and antigenotoxic dual role of propolis has been previously reported by Tavares and co-workers 34 for a Brazilian propolis sample.However, in that study, the antigenotoxic activity was observed against the DNA intercalating chemotherapeutic drug doxorubicin.It is interesting to note that lower concentrations of C.EE (25 mg mL À1 and 100 mg mL À1 ) displayed genotoxicity only in coincubation experiments where damage was analysed immediately aer incubation unlike the pre-incubation experiments, which included an additional incubation with S buffer (Fig. 2B  and D).Repair of DNA damage during this incubation may be the reason explaining the difference observed in both experiments.In fact, DNA repair activity during incubation of yeast cells with S buffer, as has been reported before, 15 could eliminate DNA injuries caused by C.EE in pre-incubation up to a maximum C.EE concentration, aer which the amount of damage caused being presumably in excess to the repairing capacity of cells.Genotoxicity detected in the alkaline comet assay (Fig. 2B  and D) argues also in favour of a pro-oxidant activity.Longer comet tails are caused by single strand and double strand breaks, the former being one of the DNA lesions typically detected in the alkaline version of the comet assay as a result of oxidative DNA damage. 35Nevertheless, the possibility of a different non-oxidative stressing activity by propolis cannot be disregarded since these complex mixtures might contain genotoxic compounds that could lead to similar results as those presented in this work.The nding that propolis extract can enhance the antitumor activity of irinotecan, which induces DNA double-strand breaks by a non-oxidative mechanism 36 provides support to this idea.Also, many propolis compounds, namely most of those found in C.EE, have been reported to have anti-proliferative, anti-tumour and anticancer effects.For instance, caffeic acid, CAPE, quercetin, apigenin, kaempferol, chrysin and galangin all exhibit antitumor activity (for a thorough review see 33 ).Quercetin 3-methyl ether, a methoxylated avonoid, has potent anticancer-promoting activity by inducing cell cycle G2-M phase accumulation. 37Chrysin inhibits DNA synthesis by G1 cell cycle arrest in C6 glioma cells. 38Even at low concentrations ellagic acid interacts synergistically with quercetin enhancing the anticarcinogenic activity of the individual counterparts. 39Pro-oxidant activity of C.EE may be responsible for its toxicity, but could also have an indirect role in protection due to induction of antioxidant defences and xenobiotic metabolizing enzymes by the imposition of a mild oxidative stress, which may contribute to a more effective cytoprotection. 40

C.EE decreases intracellular oxidation in pre-incubation and co-incubation experiments
To investigate if the antioxidant effect of C.EE in the presence of H 2 O 2 is mediated by a decrease in intracellular oxidation level, cells under co-incubation conditions were analysed using ow cytometry and H 2 DCFDA as the uorescent redox-sensitive probe.This lipophilic compound permeates the cells and is deacetylated to dichlorouorescein by intracellular esterases.The deacetylated form is hydrophilic and becomes trapped inside the cells.In the presence of oxidants, it oxidizes and uoresces with a maximum of excitation at 485 nm and of emission at 530 nm.As depicted in Fig. 3 the presence of H 2 O 2 induced a signicant peak displacement towards higher levels of uorescence (Fig. 3A and D), revealing an increase in intracellular oxidation.Treatment with C.EE decreased this intracellular uorescence in a dose-dependent manner (Fig. 3G, J  and M).Inspection of cells by bright-eld (Fig. 3B, E, H, K and  N) and uorescence (Fig. 3C, F, I, L and O) microscopy conrmed both, the intracellular origin of uorescence and the dose-dependent effect on uorescence decrease by the extract.
The same approach was used to investigate the antioxidant activity of C.EE in pre-incubation and post-incubation conditions, similarly to experiments of viability and the comet assay (Fig. 4).The antioxidant activity of the extract was still present in pre-incubation experiments in all concentrations tested (Fig. 4C-E).Interestingly, prior incubation of cells with the extract completely abolished the oxidative effect of H 2 O 2 as the intracellular uorescence of dichlorouorescein was similar (25 mg mL À1 C.EE; Fig. 4C) or lower (100 and 300 mg mL À1 C.EE; Fig. 4D and E, respectively) when compared to the negative control.These results correlate with the viability assays with pre-incubation of C.EE, where a decrease in the rate of loss of viability was observed (Fig. 1A and B) when compared with cells pre-incubated with the solvent (2% ethanol).As suggested by the loss of viability in co-incubation experiments in cells incubated only with C.EE (Fig. 1D and E), a pro-oxidant activity of C.EE is compatible with a mechanism of induction of the cellular response against oxidative stress.In addition, preincubation with 300 mg mL À1 C.EE did not decrease further the intracellular oxidation as it should be anticipated in a dosedependent activity (Fig. 4E), suggesting that at this concentration the pro-oxidant activity might be too strong to yield an increase in the protective effect, which correlates with the absence of increase in viability protection when the concentrations of 100 and 300 mg mL À1 are compared (Fig. 1A and B).In post-incubation experiments an increase in intracellular oxidation was observed for all concentrations tested (Fig. 4H-J), which is in line with a pro-oxidant activity of C.EE.
The dose-dependent intracellular antioxidant activity in coincubation experiments is in accordance with the observed protective effect in cell viability (Fig. 1) and the antigenotoxic activity upon oxidative shock (Fig. 2) and with the view of some compounds acting directly in quenching oxidative species. 2 The evaluation of propolis antioxidant potential is generally performed by in vitro assays (e.g.2][43] Here, the antioxidant effects of C.EE were evaluated in vivo using an eukaryotic cell model, as others did before, 44,45 which is of higher signicance when considering applications in human cells and tissues or in a whole-body physiological context.However, in our study the antioxidant activity was investigated in cells under highly challenging conditions by exogenously added H 2 O 2 , which highlights the antioxidant potency of this propolis extract.

Conclusions
Globally, these results t in the so-called "Janus" effect, which is used to classify compounds or mixtures that have a dual effect, one positive and one negative. 46These dual and opposite effects of propolis have been also reported before by Tavares et al. 34 for Brazilian green propolis, which also acts as antigenotoxic against the DNA intercalating chemotherapeutic drug doxorubicin at low concentration and as genotoxic at high concentration.Here we provide evidence that the Janus effect of propolis is also present when cells are challenged with an oxidationmediated DNA damaging agent such as H 2 O 2 and that detection can be made with the alkaline version of the comet assay.In conclusion, depending of the dose, the studied propolis extract exhibits both genotoxic and antigenotoxic activities and acts as an intracellular antioxidant in cells challenged with H 2 O 2 .These conclusions highlight the need for careful formulation of propolis-based food and medical products and for biological monitoring of these products to avoid undesirable harming effects.

Fig. 1
Fig. 1 Influence of pre-incubation (A and B), co-incubation (C, D and E) and post-incubation (F, G and H) with C.EE on the kinetics of loss of viability of S. cerevisiae cells exposed to oxidative stress.Yeast cells were incubated with C.EE (25 mg mL À1 , (C and F); 100 mg mL À1 , (A, D and G); or 300 mg mL À1 , B, E and H) and with 5 mM H 2 O 2 (see Experimental for details).At each time-point an aliquot was collected, diluted and spread on YPD plates.Colonies were counted after 48 h of incubation at 30 C and viability was calculated as percentage, taking time 0 min as reference (100% viability).Data are the mean AE SD of three independent experiments.** means 0.001 < p # 0.01 between samples treated with H 2 O 2 and samples treated with H 2 O 2 and C.EE (pre-incubation and co-incubation).§ § means 0.001 < p # 0.01 between C.EE-treated samples and those treated with the solvent of C.EE (pre-incubation).

Fig. 2
Fig. 2 Antigenotoxicity of C.EE in S. cerevisiae cells exposed to oxidative stress (A and C) and genotoxicity of C.EE in S. cerevisiae cells (B and D).Yeast spheroplasts were pre-treated with C.EE (25 mg mL À1 , 100 mg mL À1 or 300 mg mL À1 ) for 20 min, washed with S buffer and incubated with 10 mM H 2 O 2 (A) or S buffer (B) for further 20 min.In co-incubation experiments, spheroplasts were incubated with C.EE (25 mg mL À1 , 100 mg mL À1 or 300 mg mL À1 ) and 10 mM H 2 O 2 for 20 min (C) or with C.EE and S buffer for 20 min, which was used instead of H 2 O 2 (D).DNA damage was analysed with the yeast comet assay (see Experimental section).Controls included untreated cells as well as cells treated with the solvent of C.EE (ethanol; 2% final concentration as in the assays with C.EE).Mean AE SD values are from three independent experiments (* represents 0.01 < p # 0.05, ** represents 0.001 < p # 0.01 and ***p # 0.001).

Fig. 3
Fig. 3 Intracellular oxidation of S. cerevisiae cells exposed to H 2 O 2 is decreased in the presence of C.EE.Cells were loaded with H 2 DCFDA and treated simultaneously with 2% ethanol (A-C), 10 mM H 2 O 2 (D-F) or with 10 mM H 2 O 2 and C.EE ((G-I) 25 mg mL À1 ; (J-L) 100 mg mL À1 ; (M-O) 300 mg mL À1 ) simultaneously for 20 min and analysed for fluorescence by flow cytometry (A, D, G, J and M), bright-field microscopy (B, E, H, K and N) and fluorescence microscopy (C, F, I, L and O).Data are from a representative experiment from three independent experiments.Bar ¼ 10 mm.

Fig. 4
Fig. 4 Intracellular oxidation of S. cerevisiae cells pre-incubated with C.EE and exposed to H 2 O 2 is decreased while post-incubation with C.EE aggravates intracellular oxidation.Cells were loaded with H 2 DCFDA, incubated with 2% ethanol (A and B), or C.EE ((C) 25 mg mL À1 ; (D) 100 mg mL À1 ; (E) 300 mg mL À1 ), washed and incubated with H 2 O (A) or 10 mM H 2 O 2 (B, C, D and E).Alternatively, H2DCFDA-loaded cells were incubated with H 2 O (F), or 10 mM H 2 O 2 (G, H, I and J), washed and incubated with 2% ethanol (F and G) or C.EE ((H) 25 mg mL À1 ; (I) 100 mg mL À1 ; (J) 300 mg mL À1 ).All cells were analysed for fluorescence by flow cytometry.Data are from a representative experiment from three independent experiments.
In co-incubation experiments C.EE and H 2 O 2 were added to the suspension simultaneously.In post-incubation experiments cells were incubated for 20 min with H 2 O 2 followed by a washing step, suspension in S buffer and incubation with C.EE for further 20 min.Controls were included by replacing C.EE by the same volume of the solvent (ethanol) and/or H 2 O 2 by the same volume of S buffer.

Table 1
Chemical characterization of C ôa propolis ethanol extract (C.EE).The content of total polyphenols and total flavonoids was expressed, respectively, in equivalents of gallic acid (mg GAE) and in equivalents of quercetin (mg QE) per gram of propolis extract C.EE 160.40 AE 16.56 30.21AE 0.52 This journal is © The Royal Society of Chemistry 2016 RSC Advances Paper
a [M À H] À m/z a + indicates that TIC peak is not pure.bStandardfor compounds identied with standards, or m/z (abundance percent) -for compounds conrmed by comparison of MS/MS fragmentation with bibliography.See chromatogram in ESI Fig. 1.caffeic