ORIGINAL ARTICLE

Quercetin supplementation: insight into the potentiallyharmful outcomes of neurodegenerative prevention

Maja Jazvinšćak Jembrek & Ana Čipak Gašparović &Lidija Vuković & Josipa Vlainić & Neven Žarković &Nada Oršolić

Received: 26 July 2012 /Accepted: 2 October 2012 /Published online: 17 October 2012# Springer-Verlag Berlin Heidelberg 2012

Abstract Dietary antioxidant supplements have been con-sidered for the prevention of neuronal oxidative injury anddeath. Recent studies indicate that excessive antioxidantscould exert adverse effects, thereby questioning the safetyof prolonged supplementation. The aim of our study was toinvestigate the effects of quercetin (up to 150 μM), theubiquitous plant-derived flavonoid and highly potent scav-enger of reactive oxygen species (ROS) on healthy P19neurons, in order to assess the efficacy and safety of itslong-term use in neurodegenerative prevention. Althoughexposure for 24 h to quercetin did not compromise neuronalsurvival, morphological examination revealed diminishedneuronal branching, a finding probably related to an

observed decrease in lactate dehydrogenase activity. Using2′,7′-dichlorofluorescin diacetate and dot–blot analysis, wefound reduced basal levels of ROS and 4-hydroxy-2-none-nal, a biomarker of lipid peroxidation, confirming the anti-oxidative mechanism of quercetin action. Unexpectedly,quercetin also depleted intracellular glutathione content.Reverse transcriptase PCR and western blot analysisshowed depletion of total RNA amount and changes in theexpression of cell survival regulating genes Bcl-2, p53, andc-fos. Nuclear condensation and caspase-3/7 activity, phe-nomena related to programmed cell death cascade, were notaffected. The potential risk of observed changes indicatesthat quercetin-enriched supplements should be taken withcaution. The diversity of quercetin effects and complexity ofpossible intracellular interactions between affected genespointed out the necessity for additional pharmacologicaland toxicological studies in order to better elucidate themechanisms of quercetin action and to recognize its poten-tial side effects at higher doses and during long-termadministration.

Keywords Quercetin . P19 neurons . Antioxidantsupplementation . Glutathione depletion . Bcl-2,p53 and c-fos expression

AbbreviationsATRA all-trans retinoic acidDCF-DA 2′,7′-dichlorofluorescin diacetateDIV Days in vitroHNE 4-Hydroxy-2-nonenalMTS 3-(4,5-Dimethyl-2-yl)-5-(3-carboxymethoxy-

phenyl)-2-(4-sulfophenyl)-2H-tetrazoliumROS Reactive oxygen speciesRFU Relative fluorescence unit

M. Jazvinšćak Jembrek (*) : J. VlainićLaboratory for Molecular Neuropharmacology,Division of Molecular Medicine, Rudjer Boskovic Institute,Bijenicka 54,10 000 Zagreb, Croatiae-mail: jazvin@irb.hr

A. Čipak Gašparović :N. ŽarkovićLaboratory for Oxidative Stress, Division of Molecular Medicine,Rudjer Boskovic Institute,Bijenicka 54,Zagreb, Croatia

L. VukovićLaboratory for Genotoxic Agents, Division of Molecular Biology,Rudjer Boskovic Institute,Bijenicka 54,Zagreb, Croatia

N. OršolićDepartment of Animal Physiology, Faculty of Science,University of Zagreb,Rooseveltov trg 6,Zagreb, Croatia

Naunyn-Schmiedeberg's Arch Pharmacol (2012) 385:1185–1197DOI 10.1007/s00210-012-0799-y

Introduction

In physiological concentrations, reactive oxygen species(ROS) have many important functions, particularly as sig-naling molecules in various transduction pathways. On theother hand, excessive ROS generation has deleteriouseffects on biological macromolecules (nucleic acids, pro-teins, lipids, and carbohydrates) that end in alterations oftheir structure and function, frequently leading to neuronaldeath. The condition of oxidative stress, induced by elevatedROS accumulation, has been implicated in brain dysfunc-tion in physiological aging and various neurodegenerativediseases (Emerit et al. 2004; Slemmer et al. 2008).

It is hypothesized that ROS-lowering actions, like intakeof drugs that are able to act as ROS scavengers, may exertbeneficial effects in the prevention of age-related cognitiveand motor impairments in humans. A great interest has beendirected to natural antioxidants and their potential to regainredox homeostasis and prevent or delay oxidative injury(Zhang et al. 2010). Common sources of natural antioxi-dants include fruits, vegetables, wine, and tea. The risk oftoxicity from food supply is relatively low, but numerousreports of beneficial effects of natural compounds on humanhealth resulted in many nutraceutical supplements and puri-fied herbal extracts with doses and bioavailability of anti-oxidants highly beyond levels associated with a typical diet(Martin and Appel 2010). Since dietary supplements are notmarketed as drugs, but as supplements increasing the overallwell-being, they are in principle not subjected to rigoroustoxicological and pharmacological studies. Additionally, al-though dietary intervention studies demonstrated an abilityof phytochemicals to improve neurological dysfunction,some clinical trials did not detect benefit from antioxidantsupplements. Even more, some studies suggested that excesssupplementation with certain antioxidants may be harmful,presumably through prooxidative actions, leading to the con-clusion that the effects of antioxidants, either natural or syn-thetic, can be considered as a double-edged sword (Mennen etal. 2005; Bjelakovic et al. 2007; Boots et al. 2008). Recentfindings also revealed reduced health-promoting effects ofphysical exercise in subjects consuming antioxidant supple-ments (Ristow et al. 2009). Longevity-promoting factors, suchas physical exercise, cause an activation of mitochondrialoxygen consumption and promote the increased formationof ROS (Antoncic-Svetina et al. 2010). Generated ROS, inturn, induce downstream signaling and evoke endogenousdefense mechanisms, especially those associated with ROSdetoxification, which result in increased antioxidant defenseand oxidative stress resistance that might be further enhancedby mild prooxidants (Ristow and Schmeisser 2011;Yelisyeyeva et al. 2012). This concept of mitochondrial horm-esis questions Harman’s free radical theory of aging andimplicates that a mild increase in mitochondrial ROS actually

promotes metabolic health. Since exogenous antioxidants arecapable to interfere with this beneficial ROS signaling, andtaking into account that there is still concern about unrecog-nized toxic side effects of concentrated natural products athigher doses and during long-term administration, antioxidantsupplementation considered for a potential treatment in theprevention of oxidative stress-related diseases is doubtednowadays.

The objective of our study was to investigate the effectsof quercetin, the ubiquitous natural antioxidant and one ofthe most potent scavengers of ROS from the flavonoidfamily, on healthy P19 neurons, in order to better understandthe mechanisms of its action in physiological conditions.Besides its antioxidative action, it is known that quercetinmay affect different signaling pathways that control geneexpression and cell fate (Spencer et al. 2003; Kelsey et al.2010). The effects of quercetin are still puzzling since dif-ferent modes of action have been observed. Cytotoxiceffects were mostly described on tumor cells (Choi andKim 2010; Vargas and Burd 2010), while cytoprotectiveeffects were predominantly demonstrated on cells exposedto different types of oxidative injury (Jung et al. 2010;Pavlica and Gebhardt 2010; Jazvinšćak Jembrek et al.2012). Based on such findings, Boots et al. (2008) alreadyindicated that favorable effects of quercetin appear to bemore pronounced when the basal level of oxidative damageis high. In addition, protective concentrations of quercetincould become toxic after prolonged exposure (Ossola et al.2008, 2009); under certain conditions, quercetin may act asa prooxidant (Filipe et al. 2004; Martins et al. 2009) and isable to display genotoxic effects in vitro (Boots et al. 2008).Thus, we analyzed on an in vitro model the possible mech-anisms of the undesirable effects of quercetin supplementa-tion intended for neurodegenerative prevention.

Materials and methods

Chemicals and reagents

Quercetin dihydrate (98 %) was obtained from Aldrich Ch.Co. Inc. (Milwaukee, WI, USA). Cytosine–arabinofurano-side, Hoechst 33342, 2′,7′-dichlorofluorescin diacetate(DCF-DA), and all chemicals used for maintaining anddifferentiation of P19 cells were purchased from Sigma-Aldrich Chemicals (St. Louis, MO, USA). Other chemicalsused were of analytical grade.

P19 cell culturing and P19 neuronal differentiation

Undifferentiated P19 cells (pluripotent mouse embryonal car-cinoma cell line) were maintained in high-glucose Dulbecco’smodified Eagle’s medium (DMEM) supplemented with 10 %

1186 Naunyn-Schmiedeberg's Arch Pharmacol (2012) 385:1185–1197

heat-inactivated fetal bovine serum (FBS), 2 mM L-glutamine, 100 units/ml penicillin G, and 100 μg/mlstreptomycin (growth medium) in a humidified atmo-sphere of 5 % CO2 at 37 °C. They were passaged at1:12 dilution every 2 days using 0.05 % trypsin and1 mM EDTA.

Neuronal differentiation of P19 cells was induced byexposure to 1 μM all-trans retinoic acid (ATRA) for 4 days.Near-confluent cultures were trypsinized and seeded (1×106) into 10-cm bacteriological grade Petri dishes contain-ing 10 ml of DMEM medium (high glucose) supplementedwith a reduced concentration of FBS (5 %), 2 mM L-gluta-mine, and antibiotics (induction medium). ATRA is addedimmediately after plating. Aggregates (embryonal bodies)of P19 cells were formed after 1–2 days. After 48 h, the oldmedia was replaced with the fresh ATRA-containing medi-um. Aggregated cultures were grown in a humidified atmo-sphere for an additional 2 days.

After the 4-day ATRA treatment, P19 embryonal bodieswere harvested, washed with phosphate buffer saline (PBS),dissociated by trypsinization and pipetting, collected bycentrifugation (1,250 rpm, 5 min), and finally resuspendedin growth medium. For optimal neuronal differentiation,single cells at a density 105 cells/cm2 were plated onto 96-well plates or 35-mm Petri culture dishes (Sarstedt, Newton,NC, USA, and NUNC, Roskilde, Denmark) and grown ingrowth medium for 2 days. Since serum inhibits neuronalproduction and favors the growth of astrocytes and fibro-blasts, after 2 days the growth medium was changed toserum-free differentiation medium containing DMEM sup-plemented with 5 μg/ml insulin, 30 μg/ml transferrin,20 μM ethanolamine, 30 nM sodium selenite, 0.5 mM L-glutamine, and antibiotics (neuron-specific medium), andthe cells were differentiated for an additional 2 days.Mitotic inhibitor cytosine-arabinofuranoside (AraC) at a10-μM concentration was added to inhibit the proliferationof non-neuronal cells. Complete neuronal differentiationwas confirmed with monoclonal anti-tubulin β-III mouseIgG, clone TU-20 conjugated with Alexa Fluor®488(Millipore, Temecula, CA, USA). Differentiated cellsexpressing neuronal marker β-III tubulin were visualizedby fluorescence microscopy (data not shown).

Drug treatment

P19 neurons at DIV8 were used for all drug treatments.Each batch of cultured cells was divided into control (vehi-cle-treated) group and drug-treated groups that were ex-posed to different concentrations of quercetin (3–150 μM)in neuron-specific medium for 24 h. Concentrations of quer-cetin were chosen based on other in vitro studies. Quercetinwas dissolved in DMSO at a final concentration of 30 mg/ml and diluted in culture medium.

Determination of P19 neuron viability

Celltiter 96® Aqueous One Solution Cell Proliferation Assay(Promega, WI, USA) was used to assess the viability of P19neurons following quercetin treatment for 24 h. The solutionreagent contains a tetrazolium compound MTS [3-(4,5-dimethyl-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, inner salt] that is reduced into acolored formazan product soluble in culture medium.Conversion is accomplished by NADPH or NADH producedby dehydrogenases present in metabolically active cells. P19neurons were seeded onto 96-well microplates and incubatedfor 24 h with various concentrations of quercetin in 100 μl ofculture medium. Assays were performed by adding 20 μl ofreagent solution directly to culture wells. After 2 h of incuba-tion at 37 °C, absorbance was recorded at 492 nm with amicroplate reader (Easy-Reader 400 FW, SLT LabInstruments GmbH, Austria). Cell viability is expressed as theratio between the values obtained in treated versus control cells.Morphological examination was performed by light microsco-py (KRÜSS Optronic GmbH, Germany).

Measurement of lactate dehydrogenase release

The effect of quercetin treatments on P19 neurons plasmamembrane permeability was determined my measuring therelease of a soluble cytosolic enzyme lactate dehydrogenase(LDH) into the culture medium using CytoTox-ONE™Homogeneous Membrane Integrity Assay (Promega, WI,USA). LDH activity is an indicator of cell membrane integ-rity and serves as a general mean to assess the cytotoxicityresulting from exposure to chemical compounds. Following24 h of treatment, aliquots (100 μl) of culture medium werecollected from culture dishes (P19 neurons were treated in1.5 ml of culture medium), and an equal volume ofCytoTox-ONE Reagent containing lactate, NAD+, and resa-zurin as substrates in the presence of diaphorase was addedand incubated at room temperature (RT) for 10 min.Generation of the fluorescent resorufin product is propor-tional to the amount of LDH released. Fluorescence wasrecorded with an excitation wavelength of 560 nm and anemission wavelength of 590 nm (Varian Cary Eclipse fluo-rescence spectrophotometer). Data were calculated aftersubtraction of background fluorescence obtained from neu-ronal growth medium.

Measurement of intracellular ROS production

The antioxidant and radical scavenging ability of quercetinwas determined by the assay that utilizes non-fluorescentpermeable compound DCF-DA. After deacetylation, itreacts with ROS and forms the fluorescent product dichlor-ofluorescein. By measuring the fluorescence, it is possible to

Naunyn-Schmiedeberg's Arch Pharmacol (2012) 385:1185–1197 1187

quantify the overall oxidative stress in the cells. At the endof the treatment period, P19 neurons (from at least fourwells per group in one experiment) were incubated with100 μM DCF-DA in PBS for 1 h, rinsed, and incubatedfor an additional 1 h in PBS. Results are obtained withVarian Cary Eclipse fluorescence spectrophotometer withan excitation wavelength (λex) of 504 nm and an emissionwavelength (λemm) of 529 nm. The results were calculatedand represented as percentage of fluorescence intensityobtained on vehicle-treated control cells.

Measurement of HNE production

Cells were cultivated as described earlier. In order to mea-sure HNE production, P19 neurons were harvested at 1, 3,and 24 h from the beginning of quercetin treatment andlysed by three cycles of snap freezing, thawing, and vortex-ing. Afterwards, cell lysates were centrifuged at 14,000g for10 min. An aliquot of the supernatant was used for proteinquantitation according to Bradford. For dot–blot analysis,protein extracts (each) were spotted onto nitrocellulosemembranes (0.2 μm; Amersham). The membranes wereincubated in blocking solution (0.5 % non-fat milk powderin PBS) at room temperature for 60 min and subsequentlyincubated overnight with mouse monoclonal antibodies di-rected against HNE–His adducts (1:45 in 1 % BSA in PBS).Blots were then washed and incubated for 2.5 h withperoxidase-labeled secondary antibody (EnVision, DAKO,Denmark). Immunocomplexes were visualized using 3,3-diaminobenzidine tetrahydrochloride in organic solvent(DAKO, Denmark) and scanned for signal quantification.

Measurement of reduced glutathione

Changes in the intracellular level of reduced glutathione(GSH) were monitored by using a GSH-Glo™ GlutathioneAssay (Promega, Madison, WI, USA) according to themanufacturer’s instructions. Briefly, the assay is based onthe conversion of a luciferin derivative into luciferin byglutathione S-transferase in the presence of GSH. The signalgenerated in a coupled reaction with luciferase is propor-tional to the amount of GSH. Following treatment, themedium is removed and 100 μl of GSH-Glo™ reagent isadded per well. After incubation for 30 min, 100 μl ofluciferin detection reagent was added, and following incu-bation for 15 min, emitted light was measured with a lumin-ometer (Fluoroskan Ascent FL, Thermo Scientific).

Nuclear Hoechst staining

Control and quercetin-treated P19 neurons were stainedwith nuclear dye Hoechst 33342 to examine nuclear mor-phology. Following treatment, attached cells and cells

floating in the culture medium were collected, centrifugedat 250g, resuspended in 100 μl of culture medium, andstained with 5 μM Hoechst 33342 for 10 min at RT. Cellswere analyzed under a fluorescent microscope. The percent-age of cells with condensed nuclei in relation to the totalnumber of cells was determined. In each experiment, at least500 neurons were counted in randomly selected microscopicfields.

Caspase-3/7 assay

The activity of caspase-3/7 was determined by Apo-ONE®Homogeneous Caspase-3/7 assay (Promega, Madison, WI,USA). These members of the caspase family play key ef-fector roles in apoptosis. To perform the assay, buffer andnon-fluorescent caspase substrate Z-DEVD-rhodamine 110were mixed (50 μl) and added to treated P19 neurons grownin an equal amount of culture medium. Upon sequentialcleavage and removal of the DEVD peptides by caspase-3/7 activity, rhodamine 110 leaving group becomes fluores-cent. The amount of generated fluorescent product is pro-portional to the amount of caspase-3/7 cleavage activitypresent in the sample. The excitation wavelength for detec-tion was 485 nm. Emission was recorded at a wavelength of538 nm (Fluoroskan Ascent FL, Thermo Scientific).

Determination of Bcl-2, Bax, c-fos, and p53 mRNA levelsby semi-quantitative RT-PCR

Expressions of Bcl-2, Bax, c-fos, and p53 mRNA wereexamined by semiquantitative RT-PCR analysis accordingto the method previously described by Jazvinšćak Jembreket al. (2012). cDNAs were amplified and analyzed duringtwo consecutive cycles in the log phase of PCR reactions.PCR primers, annealing temperatures, and numbers ofcycles are listed in Table 1. The reactions were performedin a Perkin Elmer 9600 thermocycler and amplified products(10 μl) were electrophoretically separated on 1.5 % agarosegel and stained with ethidium bromide (0.5 μg/ml). Opticaldensities of detected bands were analyzed with ImageJ NIHsoftware 1.0. Expression of housekeeping gene TATA-boxbinding protein (TBP) mRNA was used as an internal stan-dard for normalization.

Western blot analysis of Bcl-2 and pBad expression

Following treatment, P19 neurons were washed twice withPBS and whole-cell extracts were obtained by scrappinginto 100 μl of hot S buffer (20 mM Tris–HCl, pH 8.5,1 mM EDTA, 5 % glycerol, 1 mM dithiothreitol, and0.5 mM phenylmethylsulfonyl fluoride) and sonicated.Equal aliquots of cell lysates were fractionated by electro-phoresis (2 h/80 V) on 12 % sodium dodecyl sulfate (SDS)–

1188 Naunyn-Schmiedeberg's Arch Pharmacol (2012) 385:1185–1197

polyacrylamide gels and transferred (1 h/400 mA) onto anitrocellulose membrane (Hybond C, AmershamPharmacia, USA). Non-specific binding was blocked by5 % non-fat milk in Tris-buffered saline containing 0.05 %Tween 20 for 1 h at RT. Membranes were incubated over-night (at 4 °C) with the polyclonal (Bcl-2) or monoclonalrabbit antibodies (pBad and ERK) diluted at 1:1,000 (SantaCruz Biotechnology, USA). Detection of protein expressionwas enabled by incubation of membranes (2 h at RT) withanti-rabbit secondary antibodies (diluted to 1:5,000,Amersham Pharmacia, USA). The levels of Bcl-2 andpBad were determined using enhanced chemiluminescencedetection reagent (Amersham Pharmacia, USA). Incubationwith monoclonal rabbit extracellular signal-regulatedkinases (ERK) antibody was performed to control for equalloading. Relative intensities of the bands were analyzed withthe ImageJ NIH software.

Statistical analysis

The results are expressed as mean ± SEM from at least threeindependent experiments. The obtained data were analyzedusing GraphPad statistical software. One-way ANOVA andDunnett’s multiple-comparison test were used to determinethe effects of quercetin and the differences among treatedgroups. The differences were considered significant whenthe obtained P value was below 0.05.

Results

Effects of quercetin on survival and morphologicalappearance of P19 neurons

Differentiation of P19 embryonal carcinoma cells resulted ina mixed culture consisting mostly of neurons and, to a lesserextent, fibroblasts and astrocytes. The number of the latterwas reduced by exposure to antimitotic AraC. Resultsobtained by MTS-based assay indicated that exposure to

quercetin for 24 h failed to affect neuronal viability in abroad range of concentrations (Fig. 1a). We also performedmorphological analysis by light microscopy. As expected,P19 neurons from the vehicle-treated group have round cellbodies aggregated in tight clusters and elongated axonal anddendritic processes making complex patterns of neuritebranching and interconnections, often grouped in bundles(Fig. 1b). Following quercetin treatment, we failed to ob-serve visible changes in neuronal body appearance.However, the network of thick bundles as well as the num-ber of fine, thin neuronal dendritic processes was reduced atthe end of 24 h of exposure to quercetin (Fig. 1c).

Exposure to quercetin diminished LDH release

LDH measurements were conducted in order to examine thepreservation of membrane integrity after exposure to quer-cetin. The obtained results have shown that following treat-ment with the highest concentration of quercetin, P19neurons diminished the release of LDH into the surroundingmedium (Fig. 2). While in the control group the measuredactivity of LDH was 18.33±0.59 (expressed in RFU), in thegroup treated with 150 μM quercetin, the obtained valuewas 12.25±1.89, which represents a decrease of 33 %.

Exposure to quercetin decreased the basal levels of ROSand 4-hydroxy-2-nonenal and depleted intracellular GSHcontent

While high concentrations of ROS are undoubtedly detri-mental, physiologically low levels of ROS exert desirableeffects, triggering signaling cascades that actually promotemetabolic health. In the culture of P19 neurons, quercetincaused a concentration-dependent decrease in intracellularaccumulation of ROS as measured by 2′,7′-dichlorofluores-cin diacetate (Fig. 3a). With the highest concentration ofquercetin applied, the level of ROS was decreased to ap-proximately 40 % of control group. As presented in Fig. 3b,the antioxidative action of quercetin was also demonstrated

Table 1 Primer sequences andconditions used for PCRamplifications

Gene Primer sequence (5′→ 3′) Productlength (bp)

Annealingtemperature (°C)

Numberof cycles

Bcl-2 F: GGAGATCGTGATGAAGTACATAC 373 58 28–29R: CCTGAAGAGTTCCTCCACCACC

Bax F: ATCGAGCAGGGAGGATGGCT 470 62 28–29R: CTTCCAGATGGTGAGCGAGG

c-fos F: GAATGGTGAAGACCGTGTCAGG 456 60 25–26R: CGTTGCTGATGCTCTTGACTGG

p53 F: AGAGACCGCCGTACAGAAGA 231 62 35–36R: CTGTAGCATGGGCATCCTTT

TBP F: ACCCTTCACCAATGACTCCTATG 190 60 29–30R: ATGATGACTGCAGCAAATCGC

Naunyn-Schmiedeberg's Arch Pharmacol (2012) 385:1185–1197 1189

by its ability to decrease the level of the major bioactiveend-product of lipid peroxidation, 4-hydroxynonenal(HNE). The effect of quercetin on the production of HNEprotein adducts was time dependent. Thus, the highest de-crease of HNE was observed in the group treated with

150 μM quercetin for 24 h. Very similar effects wereobtained with 30 μM quercetin in all time-points, sug-gesting that the effect of quercetin on lipid peroxidationwas not concentration dependent. On the other hand,exposure to the highest concentration of quercetin for24 h, as shown in Fig. 3c, also induced a significantdecrease by 30 % in the cellular level of GSH, one ofthe major cellular antioxidants.

Exposure to quercetin failed to initiate programmed celldeath cascade

Hoechst 33342 dye (final concentration 5 μM) was used forcounterstaining the cell nuclei of living cells. Exposure to 30and 150 μM quercetin for 24 h failed to induce apoptosis-related changes in nuclear morphology, such as chromatincondensation and/or nuclear fragmentation. As representedin Fig. 4, treatment with quercetin did not result in anymeasurable alterations in chromatin condensation; the numberof condensed nuclei was not different between the examinedgroups. Exposure of P19 neurons to quercetin also failed toaffect caspase-3/7 activity. No differences were found in fluo-rescent signals obtained from all investigated groups, reflect-ing the same ability of quercetin-treated cells to cleavecaspase-3/7-specific substrate Z-DEVD-rhodamine 110 incomparison to control neurons.

Exposure to quercetin induced changes in the expressionof genes involved in intracellular signaling and inductionof apoptosis

In order to perform semiquantitative RT-PCR analysis, weextracted total cellular RNA from treated P19 neurons andfound that exposure to quercetin in a dose-dependent man-ner decreased the amounts of total RNA extracted (Fig. 5).

Fig. 1 Effect of quercetin on viability and morphological appearanceof P19 neurons. P19 neurons were incubated with various concentra-tions of quercetin for 24 h. Neuronal viability was determined by MTS-based assay (a). Data are expressed as means ± SEM from threeindependent experiments. Statistical analysis with one-way ANOVApointed out that exposure to quercetin did not compromise the survivalof P19 neurons. Morphological examination revealed that quercetinreduced the network of neuronal branching and interconnections. Pho-tographs are obtained by light microscope and represent P19 neuronsexposed to vehicle (b) and 150 μM quercetin (c). Scale bars, 50 μM

Fig. 2 Effect of quercetin on LDH release. P19 neurons were exposedto 3, 30, and 150 μM quercetin for 24 h. Release of LDH, a measure ofmembrane permeability, was determined as described in “Materials andmethods”. Values represent the mean ± SEM of four independentexperiments performed in triplicate. *P

From the vehicle-treated group, we obtained 6.83±0.13 μgof total RNA; in the presence of 30 μM quercetin, 4.90±0.41 μg of total RNAwas isolated, while from P19 neuronstreated with 150 μM quercetin, 4.47±0.67 μg of total RNAwas obtained.

As indicated in Fig. 6, densitometric analysis of ampli-fied PCR products revealed that exposure to quercetindownregulated the expression of antiapoptotic protein Bcl-2 at the transcriptional level by 37.3 % and the expression ofc-fos mRNA by 40.2 %. Transcriptional expression of Bax,

Fig. 4 Lack of effects of quercetin on nuclear condensation andcaspase-3/7 activity in P19 neurons. P19 neurons were exposed todifferent concentrations of quercetin for 24 h and then stained with5 μM Hoechst 33342 dye for 10 min. The morphological appearanceof nuclear chromatin was examined with fluorescence microscopy. Ineach experiment, the number of condensed nuclei was determined bycounting at least 500 neurons in randomly selected fields (a). Theactivity of caspase-3/7 was measured following quercetin treatmentusing Apo-ONE® Homogeneous Caspase-3/7 assay (b). Values areexpressed as means ± SEM from at least three independent experi-ments. According to one-way ANOVA followed by Dunnett’smultiple-comparison test, exposure to quercetin for 24 h failed toinitiate apoptotic cascade in P19 neuronsFig. 3 Effect of quercetin on ROS production, HNE-protein adducts

formation, and GSH content. P19 neurons were exposed to 3, 30, and150 μM quercetin for 24 h. Generation of ROS was quantified by theassay that utilizes the cell permeable substrate 2′,7′-dichlorofluorescindiacetate and indicated a strong antioxidative action of quercetin (a).Results representing decreased formation of HNE adducts as a measureof lipid peroxidation level are shown in b. Exposure to quercetin alsoreduced intracellular GSH content (c). Data represent the mean ± SEMfrom at least three independent experiments. *P

a typical proapoptotic member, was not affected, while the p53mRNAwas increased by 81.1 % in the presence of the highestquercetin concentration. Altered expression of Bcl-2 mRNAalso resulted in the overall decrease of Bcl-2/Bax ratio by 23 %(Fig. 6c). In all analyses, the level of housekeeping gene TBPwas used as an internal loading control. Similar results wereobtained when data were normalized to the expression of β-actin, another housekeeping gene (data not shown).

In addition, we analyzed the expression of Bcl-2 at theprotein level by Western blot method. We also looked forchanges in the expression of pBad, an important proapoptoticmember of the Bcl-2 family in neuronal cells. As demonstrat-ed in Fig. 7, exposure to quercetin induced a decrease (36.5%)in the relative intensity of immunodetectable Bcl-2 bands. Theexpression of the phosphorylated form of Bad was not affect-ed by quercetin treatment.

Discussion

Numerous studies have shown that flavonoids, powerful poly-phenolic antioxidants, could exert a remarkable spectrum ofbiochemical and pharmacological activities with beneficialeffects on diverse pathologies (Montenegro et al. 2010;Oršolić et al. 2010, 2011). At the same time, a growingnumber of evidence suggests that the effects of antioxidantsare not exclusively beneficial, particularly at higher doses.Expectedly, this duality resulted in an ongoing intricate debateover the efficacy and safety of antioxidant supplementation inthe prevention of cellular oxidative damage.

We studied the effects of quercetin, one of the most abun-dant and very potent natural antioxidants, on healthy P19neurons in order to better characterize the mechanisms of itsaction, assess the efficacy and safety of its long-term use, and

Fig. 6 Effects of quercetin onBcl-2, Bax, c-fos, and p53mRNA expression. P19 neu-rons were exposed to 30 and150 μM quercetin for 24 h. To-tal RNA was extracted andreverse-transcribed into cDNA.The obtained cDNA is furtheramplified using specific pri-mers. Following densitometricquantification, intensities ofBcl-2 (a), Bax (b), p53 (d), andc-fos (e) bands were normalizedto the expression of house-keeping gene TBP. Representa-tive agarose gel electrophoresesare also shown (f). Exposure toquercetin decreased Bcl-2 andc-fos mRNAs expression, aswell as the ratio of Bcl-2/Bax,and induced the overexpressionof p53 mRNA (c). The data areexpressed as means ± SEMfrom at least six independentRT-PCR analyses. *P

estimate if it is indeed a candidate for neurodegenerativeprevention. Studies of the effects of quercetin on healthyneuronal cells in physiological conditions are not numerousbut are crucial for the valuable evaluation of its potential as aneuroprotective agent.

A culture of P19 neurons, consisting mostly of neurons, andto a lesser extent of astrocytes and fibroblasts, possesses phys-iological and pharmacological characteristics of neurons invivo (Turetsky et al. 1993). In a broad range of concentrations,quercetin failed to affect the viability of P19 neurons, indicatingthat they tolerate relatively high concentrations of this antiox-idant. Neuronal SH-SY5Y cell line also tolerates relatively highconcentrations of quercetin (Ossola et al. 2008), but cerebellargranule and cortical neurons exert much greater sensitivity, withneurotoxic effects observed in the low micromolar range(Spencer et al. 2003; Arredondo et al. 2010).

Although quercetin did not compromise neuronal surviv-al in our study, it reduced the network of neuronal intercon-nections, leading to the overall reduction of cell surface andlowering the ability of LDH to enter into the surroundingculturing medium. Hence, we suggest that the decrease inLDH activity is not related to changes in membrane integritybut simply reflects changes in neuronal branching. On primaryrat neuronal cells, Jakubowicz-Gil et al. (2008) also noticeddiminished neuronal arborization with a non-toxic concentra-tion of quercetin and concluded that such morphologicalchanges may affect proper contacts and signal transmissionand cause unfavorable brain dysfunction. Interestingly, in P19neurons exposed to hydrogen peroxide, the presence of

quercetin reduced cellular damage and improved neuronalbranching (Jazvinšćak Jembrek et al. 2012), supporting theopinion that the effects of quercetin might be different inphysiological conditions and during oxidative injury.

In accordance with the described antioxidative mecha-nism of quercetin action through the literature, in P19 neu-rons quercetin effectively diminished accumulation of ROSbelow the basal level. The antioxidative effect was alsoconfirmed by its ability to decrease lipid peroxidation, asevidenced by the decreased 4-hydroxy-2-nonenal formation.Although for a long time ROS and lipid peroxidation end-products have been considered predominantly as cytotoxicagents, recent findings provided evidence for their importantrole in intracellular signaling (Zarkovic 2003; Guéraud et al.2010). Physiological concentrations of HNE regulate geneexpression (including p53 and c-fos) and modulate intracel-lular pathways involved in cell cycle arrest, differentiation,and apoptosis (Čipak Gašparović et al. 2010). Bringing tomind mitochondrial hormesis and the importance of mildoxidative stress in redox homeostasis, the obtained resultsimplicate that quercetin could interfere with cellular adapta-tion to oxidative stress by decreasing redox signaling, theHNE-related transduction pathways, and finally antioxida-tive defense. Hence, a prolonged decrease in the basal levelsof ROS and HNE has undesirable consequences on cellularphysiology at the end. Namely, the adaptive cellular re-sponse induced by HNE is mediated by the gene expressionof cytoprotective proteins via NF-E2-related factor 2/Kelch-like-ECH-associated protein 1 (Nrf2/Keap-1) pathway

Fig. 7 Effects of quercetin on Bcl-2 and pBad expression in P19neurons. At the end of quercetin treatment, cell extracts were preparedand loaded onto a 12% SDS–polyacrylamide gel. Expressions of selectedproteins were determined using Western blot method with the rabbit anti-Bcl-2 (a) and anti-pBad (b) antibodies. Protein band intensities were

quantified using ImageJ NIH software and normalized according toloading control (ERK). The images (c) represent one of the three inde-pendent experiments. The results shown are expressed as means ± SEMfrom three independent experiments. *P

(Noguchi 2008). Consequently, in case of lower HNE level,the expression of genes encoding cytoprotective proteinsinduced by Nrf2/ARE signaling pathway, which is activatedby HNE, would be reduced, thus making cells less viableand less adaptive to oxidative stress. Such undesirable effectof quercetin might consequently affect the overall capacityof the cellular antioxidants and eventually affect the viabil-ity of the cells. However, in our study, quercetin did notexert so negative effects; therefore, the observed decrease ofROS and HNE could be considered mostly as a desirable,early antioxidant effect of quercetin.

On the other hand, in addition to its desirable ability toreduce concentrations of ROS and HNE acting as antioxi-dant, quercetin also decreased intracellular GSH content,demonstrating its potential to induce overall misbalance ofcellular antioxidant mechanisms. Alongside with the alreadymentioned Nrf2/ARE signaling pathway physiologically in-duced by HNE and possibly attenuated by quercetin reduc-ing the cellular HNE levels, there are two other mechanismsthat could contribute to the observed GSH depletion. First,GSH depletion could be a cellular answer to the exogenousantioxidant quercetin. Quercetin replaces GSH as a cellularantioxidant, so P19 neurons attenuate GSH synthesis andreduce endogenous antioxidative defense to maintain redoxhomeostasis. These processes are achieved through the re-dox signaling—lack of ROS and HNE diminished the stim-ulation of intracellular antioxidative system, and cellsattenuate their own antioxidative defense by decreasingGSH production. Second, the formation of thiol-reactiveoxidation products of quercetin could also participate inthe observed drop in GSH level. When quercetin reacts withROS, it is converted into quercetin quinones that are veryreactive towards thiols, particularly GSH, and can instanta-neously generate mono- and di-glutathionyl adducts (Bootset al. 2003, 2008). As in some experimental cellular sys-tems, the decrease in GSH content might induce an over-shooting response that leads to excess glutathione (Darley-Usmar et al. 1991); in future studies, treatments longer than24 h are needed to clarify the effect of quercetin on GSHlevel.

Because reduced glutathione serves as the major cytosol-ic antioxidant and participates in the detoxification of manyelectrophilic xenobiotics (Pastore et al. 2003), its presence isoften taken as a marker of cellular health. Previous studieson healthy neurons have shown that quercetin either had noeffect on GSH level (Lavoie et al. 2009) or increased GSHamount (Arredondo et al. 2010). To our knowledge, this isthe only report on neuronal cells to demonstrate that expo-sure to quercetin could induce GSH depletion in physiolog-ical conditions, suggesting that the effects of quercetin maydepend on the type of cells under study and the treatmentschedules. Quercetin-induced GSH decrease was alsoevidenced in rat lung epithelial cells (Boots et al. 2007)

and in liver homogenate of rats that were given oral quer-cetin for 6 weeks (Choi et al. 2005).

The toxic effects of quercetin observed on primary neu-ronal cells involved the activation of apoptotic machinery(Spencer et al. 2003). Studies on genetically modified micehave revealed a significant role for the Bcl-2 family in aninitiation of apoptosis and in certain neurological diseaseprocesses (Akhtar et al. 2004). As evidenced by the obser-vation of chromatin structure and measurement of caspaseactivity, quercetin did not initiate a cell death cascade in P19neurons. Moreover, we failed to find changes in the expres-sion of phosphorylated forms of proapoptotic protein pBad.This finding also support the lack of neuronal apoptoticstimuli since dephosphorylation of pBad promotes the re-lease of mitochondrial cytochrome c and induction of celldeath cascade (Ham et al. 2005). However, quercetin de-creased the expression of Bcl-2, a prototypical antiapoptoticmember of Bcl-2 family. Although the transcriptional ex-pression of Bax, a prototypical proapoptotic member, wasnot affected, the overall ratio of Bcl-2/Bax was thus changedin the favor of apoptotic events. Since the execution of thecell death stimulus may depend on the intracellular balancebetween various Bcl-2 subfamily members, it is possiblethat in P19 neurons some other antiapoptotic member(s) ofBcl-2 family take over the role of Bcl-2 protein. We alsoobserved an overexpression of p53 mRNA following expo-sure to the highest concentration of quercetin. In principle,transcription factor p53 has a crucial role in eliciting neuro-nal cell death after exposure to a range of stressors, fre-quently by altering the expression of Bcl-2 family membersin a proapoptotic manner (Steckley et al. 2007; Geng et al.2010). As reviewed by Tedeschi and Di Giovanni (2009),besides this general point of view, post-translational mod-ifications could allow p53 to mediate a wide range ofdifferent functional outcomes. While phosphorylation ispredominantly associated with the induction of neuronalapoptosis, acetylation is related to prosurvival effects, suchas neuronal outgrowth and regeneration, and perhaps couldexplain why the highly increased expression of p53 does nothave detrimental consequences on P19 neurons (Fig. 8). Inaddition to these proapoptotic events, potentially antiapop-totic pathways were simultaneously activated in P19 neu-rons in response to quercetin stimulus. In accordance with afinding of Kley et al. (1992), showing that wild-type p53can inhibit c-fos gene expression in a dose-dependent man-ner, we found a decreased transcriptional expression of c-fosin quercetin-treated P19 neurons. Protein c-fos is a part oftranscription factor activator protein 1 (AP-1) that controls anumber of cellular processes, including apoptosis. Sincemany studies have reported a simultaneous increase in c-fos expression and rate of neuronal apoptosis (Estus et al.1994), it is possible that the quercetin-induced decrease in c-fos mRNA expression also has a role in the attenuation of

1194 Naunyn-Schmiedeberg's Arch Pharmacol (2012) 385:1185–1197

apoptotic machinery. Additionally, it is known that HNEparticipates in the activation of caspases and c-Jun N-terminal kinases and stimulation of AP-1 binding(Dianzani 2003). Hence, quercetin-induced decrease in thebasal level of HNE could also partially prevent initiations ofpro-death events. The coexistence of pro- and antiapoptoticpathways was also observed by Spencer et al. (2003) onmouse cortical neurons. Since in that study overall neuronaldeath results, the interplay between these pathways could bedetrimental in determining neuronal fate. Finally, in P19neurons, quercetin diminished the overall RNA content(Fig. 5), although at the protein level we did not observechanges (data not shown). Since quercetin might interactwith RNA (Marinić et al. 2006), the observed reduction isperhaps caused by the decreased ability of RNA to bind tothe glass fiber matrix of the columns for RNA extraction.Further studies are needed to clarify if quercetin treatmentindeed induces RNA down-regulation and/or changes inRNA stability and, if it does, whether these changes arespecific or non-specific.

Absorption, metabolism, and bioavailability of quercetinhave been extensively discussed regarding its potential usein the prevention of neurodegenerative changes. A quiteslow elimination of quercetin (half-life as long as 20–72 h)favors its accumulation in human plasma during repeatedconsumption. Moreover, quercetin absorption and bioavail-ability can be increased by administration in purified,delivery-improved form such as nanoparticles. Quercetin ismainly present in food as hydrophilic glycosides. Afterintake, they are hydrolyzed to quercetin aglycone, whoseabsorption is estimated to be 65–81 % (Bischoff 2008).Following hydrolysis, aglycone is rapidly converted intoconjugated metabolites due to the very efficient phase IImetabolism in the small intestine and liver. Hence, after oraladministration, quercetin is found exclusively in the form ofsulfated or glucuronated conjugates, often O-methylated

(Ossola et al. 2009). The main circulating compounds aregenerally glucuronides. Evidence exists that certain querce-tin glucuronides may cross the BBB, indicating that theremay be a specific uptake mechanism for glucuronides invivo (Ishisaka et al. 2011). β-Glucuronidase activity hasbeen confirmed in the brain. Although glia cells exert great-er β-glucuronidase activity than neurons, they also possessefficient machinery for the secretion of β-glucuronidase intothe extracellular space. Thus, released β-glucuronidase canbe taken up by neurons in receptor-mediated endocytosis.Enzymes with β-glucuronidase activity can be releasedunder certain physiologic conditions such as inflammationand, together with lysosomal β-glucuronidase enzymes, canconvert glucuronides directly to aglycone (de Boer et al.2005). Hence, our study performed with unconjugated agly-cone molecules gives valuable, pharmacologically relevantinformation regarding the potential mechanisms and possi-ble side effects of quercetin action after oral intake thatcould be an incentive for future studies in vivo.

In conclusion, the effects of quercetin on P19 neuronscan be summarized as seemingly preserved homeostasis dueto unchanged viability and no signs of apoptotic events.However, quercetin provoked morphological changes, de-creased levels of ROS and 4-hydroxy-2-nonenal, depletedintracellular glutathione content, and caused transcriptionalchanges of genes involved in intracellular signaling andinduction of apoptosis (Bcl-2, c-fos and p53). Indicationsof quercetin-induced changes, such as ability to interferewith cellular adaptation to oxidative stress by interactionwith redox signaling and HNE-related transduction path-ways, should not be neglected when considering its admin-istration for the prevention of neurodegenerative changes.Due to the diversity of quercetin actions on P19 neurons inphysiological conditions and the complexity of potentialintracellular interactions between affected genes, precisepharmacological and toxicological studies for all

Fig. 8 Schematic overview of the quercetin effects on healthy P19neurons. Exposure to quercetin for 24 h maintains cell homeostasis butcauses hormesis: quercetin decreased the basal levels of ROS and HNE,depleted intracellular glutathione content, and caused transcriptional

changes of genes involved in intracellular signaling and apoptosis. Post-translational modifications of p53, such as acetylation, could be involvedin the attenuation of apoptotic machinery

Naunyn-Schmiedeberg's Arch Pharmacol (2012) 385:1185–1197 1195

quercetin-containing supplements are needed in order tooptimize quercetin concentrations and time window of ad-ministration that will be efficient in neuroprotection, butwithout side effects. As pointed out by Boots et al. (2008),and confirmed by our studies, the beneficial effects of quer-cetin seem to be more pronounced in oxidative stress, proba-bly suggesting that quercetin supplementation should bedelayed until the first onset of oxidative stress-induced events,with desirable monitoring of the possible side effects.

Acknowledgments Undifferentiated P19 cells were kindly providedby Dr. J. Pachernik (Prague, Czech Republic). The skilful technicalassistance of Sanjica Rak and Lidija Milković is gratefully acknowl-edged. This study was supported by the Croatian Ministry of Science,Education, and Sports and the COST Action CM1001.

References

Akhtar RS, Ness JM, Roth KA (2004) Bcl-2 family regulation ofneuronal development and neurodegeneration. Biochim BiophysActa 1644:189–203

Antoncic-Svetina M, Sentija D, Cipak A, Milicic D, Meinitzer A,Tatzber F, Andrisic L, Zelzer S, Zarkovic N (2010) Ergometryinduces systemic oxidative stress in healthy human subjects.Tohoku J Exp Med 221:43–48

Arredondo F, Echeverry C, Abin-Carriquiry JA, Blasina F, Antúnez K,Jones DP, Go Y-M, Liang Y-L, Dajas F (2010) After cellularinternalization, quercetin causes Nrf2 nuclear translocation,increases glutathione levels, and prevents neuronal death againstan oxidative insult. Free Radic Biol Med 49:738–747

Bischoff SC (2008) Quercetin: potentials in the prevention and therapyof disease. Curr Opin Clin Nutr Metab Care 11:733–740

Bjelakovic G, Nikolova D, Gluud LL, Simonetti RG, Gluud C (2007)Mortality in randomized trials of antioxidant supplements forprimary and secondary prevention. Systematic review and meta-analysis. JAMA 297:842–857

Boots AW, Kubben N, Haenen GRMM, Bast A (2003) Oxidizedquercetin reacts with thiols rather than with ascorbate: implicationfor quercetin supplementation. Biochem Biophys Res Commun308:560–565

Boots AW, Li H, Schins RPF, Duffin R, Heemskerk JWM, Bast A,Haenen GRMM (2007) The quercetin paradox. Toxicol ApplPharmacol 222:89–96

Boots AW, Haenen GRMM, Bast A (2008) Health effects of quercetin:from antioxidant to nutraceutical. Eur J Pharmacol 585:325–337

Choi EJ, Kim G-H (2010) Quercetin accumulation by chronic admin-istration causes the caspase-3 activation in liver and brain of mice.Biofactors 36:216–221

Choi EJ, Lee BH, Lee K, Chee K-M (2005) Long-term combinedadministration of quercetin and daidzein inhibits quercetin-induced suppression of glutathione antioxidant defenses. FoodChem Toxicol 43:793–798

Čipak Gašparović A, Lovaković T, Žarković N (2010) Oxidative stressand antioxidants: biological response modifiers of oxidative ho-meostasis in cancer. Period Biol 4:433–439

Darley-Usmar VM, Severn A, O’Leary VJ, Rogers M (1991) Treatmentof macrophages with oxidized low-density lipoprotein increasestheir intracellular glutathione content. Biochem J 278:429–434

de Boer VCJ, Dihal AA, van der Woude H, Arts ICW, Wolffram S,Alink GM, Rietjens IMCM, Keijer J, Hollman PCH (2005) Tissuedistribution of quercetin in rats and pigs. J Nutr 135:1718–1725

de Martins RP, de Braga HC, da Silva AP, Dalmarco JB, de Bem AF,dos Santos AR, Dafre AL, Pizzolatti MG, Latini A, Aschner M,Farina M (2009) Synergistic neurotoxicity induced by methylmer-cury and quercetin in mice. Food Chem Toxicol 47:645–649

Dianzani MU (2003) 4-Hydroxynonenal from pathology to physiolo-gy. Mol Asp Med 24:263–272

Emerit J, Edeas M, Bricaire F (2004) Neurodegenerative diseases andoxidative stress. Biomed Pharmacother 58:39–46

Estus S, Zaks W, Freeman R, Gruda M, Bravo R, Johnson E (1994)Altered gene expression in neurons during programmed celldeath: identification of c-jun as necessary for neuronal apoptosis.J Cell Biol 127:1717–1727

Filipe P, Haigle J, Silva JN, Freitas J, Fernandes A, Mazière J-C,Mazière C, Santus R, Morlière P (2004) Anti- and pro-oxidanteffects of quercetin in copper-induced low density lipoproteinoxidation. Quercetin as an effective antioxidant against pro-oxidant effects of urate. Eur J Biochem 271:1991–1999

Geng Y, Walls KC, Ghosh AP, Akhtar RS, Klocke BJ, Roth KA (2010)Cytoplasmic p53 and activated Bax regulate p53-dependent,transcription-independent neural precursor cell apoptosis. JHistochem Cytochem 58:265–275

Guéraud F, Atalay M, Bresgen N, Cipak A, Eckl PM, Huc L, Jouanin I,Siems W, Uchida K (2010) Chemistry and biochemistry of lipidperoxidation products. Free Radic Res 44:1098–1124

Ham J, Towers E, Gilley J, Terzano S, Randall R (2005) BH3-onlyproteins: key regulators of neuronal apoptosis. Cell Death Differ12:1015–1020

Ishisaka A, Ichikawa S, Sakakibara H, Piskula MK, Nakamura T, KatoY, Ito M, Miyamoto K, Tsuji A, Kawai Y, Terao J (2011)Accumulation of orally administered quercetin in brain tissueand its antioxidative effects in rats. Free Radic Biol Med51:1329–1336

Jakubowicz-Gil J, Rzeski W, Zdzisinska B, Dobrowolski P, Gawron A(2008) Cell death and neuronal arborization upon quercetin treat-ment in rat neurons. Acta Neurobiol Exp 68:139–146

Jazvinšćak Jembrek M, Vuković L, Puhović J, Erhardt J, Oršolić N(2012) Neuroprotective effect of quercetin against hydrogenperoxide-induced oxidative injury in P19 neurons. J MolNeurosci 47:286–299

Jung SH, Kim BJ, Lee EH, Osborne NN (2010) Isoquercitrin is themost effective antioxidant in the plant Thuja orientalis and able tocounteract oxidative-induced damage to a transformed cell line(RGC-5 cells). Neurochem Int 57:713–721

Kelsey NA, Wilkins HM, Linseman DA (2010) Nutraceutical antiox-idants as novel neuroprotective agents. Molecules 15:7779–7814

Kley N, Chung RY, Fay S, Loeffler JP, Seizinger BR (1992)Repression of the basal c-fos promoter by wild-type p53. NuclAcid Res 20:4083–4087

Lavoie S, Chen Y, Dalton TP, Gysin R, Cuénod M, Steullet P, DoKQ (2009) Curcumin, quercetin, and tBHQ modulate glutathi-one levels in actrocytes and neurons: importance of the gluta-mate cysteine ligase modifier subunit. J Neurochem 108:1410–1422

Marinić M, Piantanida I, Rusak G, Žinić M (2006) Interactions ofquercetin and its lanthane complex with double stranded DNA/RNA and single stranded RNA: spectrophotometric sensing ofpoly G. J Inorg Biochem 100:288–298

Martin KR, Appel CL (2010) Polyphenols as dietary supplements: adouble-edged sword. Nutr Diet Suppl 2:1–12

Mennen LI, Walke R, Bennetau-Pelissero C, Scalbert A (2005) Risksand safety of polyphenol consumption. Am J Clin Nutr 81(suppl):326S–329S

Montenegro MF, Neto-Neves EM, Dias-Junior CA, Ceron CS, CastroMM, Gomes VA, Kanashiro A, Tanus-Santos JE (2010) Quercetinrestores plasma nitrite and nitroso species levels in renovascular

1196 Naunyn-Schmiedeberg's Arch Pharmacol (2012) 385:1185–1197

hypertension. Naunyn-Schmiedeberg’s Arch Pharmacol 4:293–301

Noguchi N (2008) Role of oxidative stress in adaptive responses inspecial reference to atherogenesis. J Clin Biochem Nutr 43:131–138

Oršolić N, Benković V, Lisičić D, Đikić D, Erhardt J, Horvat-KneževićA (2010) Protective effects of propolis and related polyphenolic/flavonoid compounds against toxicity induced by irinotecan. MedOncol 27:1346–1358

Oršolić N, Gajski G, Garaj-Vrhovac V, Đikić D, Prskalo ZŠ, SirovinaD (2011) DNA-protective effects of quercetin or naringenin inalloxan-induced diabetic mice. Eur J Pharmacol 656:110–118

Ossola B, Kääriäinen TM, Raasmaja A, Männistö PT (2008) Time-dependent protective and harmful effects of quercetin on 6-OHDA-induced toxicity in neuronal SH-SY5Y cells. Toxicology 250:1–8

Ossola B, Kääriäinen TM, Männistö PT (2009) The multiple faces ofquercetin in neuroprotection. Expert Opin Drug Saf 8:397–409

Pastore A, Federici G, Bertini E, Piemonte F (2003) Analysis ofglutathione: implication in redox and detoxification. Clin ChimActa 333:19–39

Pavlica S, Gebhardt R (2010) Protective effects of flavonoids and twometabolites against oxidative stress in neuronal PC12 cells. LifeSci 86:79–86

Ristow M, Schmeisser S (2011) Extending life span by increasingoxidative stress. Free Radic Biol Med 51:327–336

Ristow M, Zarse K, Oberbach A, Klöting N, Birringer M, KiehntopfM, Stumvoll M, Kahn CR, Blüher M (2009) Antioxidants preventhealth-promoting effects of physical exercise in humans. ProcNatl Acad Sci USA 106:8665–8670

Slemmer JE, Shacka JJ, Sweeney MI, Weber JT (2008) Antioxidantsand free radical scavengers for the treatment of stroke, traumaticbrain injury and aging. Curr Med Chem 15:404–414

Spencer JPE, Rice-Evans C, Williams RJ (2003) Modulation of pro-survival Akt/protein kinase B and ERK1/2 signaling cascades byquercetin and its in vivo metabolites underlie their action onneuronal viability. J Biol Chem 278:34783–34793

Steckley D, Karajgikar M, Dale LB, Fuerth B, Swan P, Drummond-MainC, Poulter MO, Ferguson SSG, Strasser A, Cregan SP (2007) Pumais a dominant regulator of oxidative stress induced Bax activationand neuronal apoptosis. J Neurosci 27:12989–12999

Tedeschi A, Di Giovanni S (2009) The non-apoptotic role of p53 inneuronal biology: enlightening the dark side of the moon. EMBORep 10:576–583

Turetsky DM, Huettner JE, Gottlieb DI, Goldberg MP, Choi DW(1993) Glutamate receptor-mediated currents and toxicity in em-bryonal carcinoma cells. J Neurobiol 24:1157–1169

Vargas AJ, Burd R (2010) Hormesis and synergy: pathways andmechanisms of quercetin in cancer prevention and management.Nutr Rev 68:418–428

Yelisyeyeva O, Semen K, Zarkovic N, Kaminskyy D, Lutsyk O,Rybalchenko V (2012) Activation of aerobic metabolism by ama-ranth oil improves heart rate variability both in athletes and patientswith type 2 diabetes mellitus. Arch Physiol Biochem 118:1–11

Zarkovic N (2003) 4-Hydroxynonenal as a bioactive marker of patho-physiological processes. Mol Asp Med 24:281–291

Zhang S, Ye J, Dong G (2010) Neuroprotective effect of baicalein onhydrogen peroxide-mediated oxidative stress and mitochondrialdysfunction in PC12 cells. J Mol Neurosci 40:311–320

Naunyn-Schmiedeberg's Arch Pharmacol (2012) 385:1185–1197 1197

Quercetin supplementation: insight into the potentially harmful outcomes of neurodegenerative preventionAbstractIntroductionMaterials and methodsChemicals and reagentsP19 cell culturing and P19 neuronal differentiationDrug treatmentDetermination of P19 neuron viabilityMeasurement of lactate dehydrogenase releaseMeasurement of intracellular ROS productionMeasurement of HNE productionMeasurement of reduced glutathioneNuclear Hoechst stainingCaspase-3/7 assayDetermination of Bcl-2, Bax, c-fos, and p53 mRNA levels by semi-quantitative RT-PCRWestern blot analysis of Bcl-2 and pBad expressionStatistical analysis

ResultsEffects of quercetin on survival and morphological appearance of P19 neuronsExposure to quercetin diminished LDH releaseExposure to quercetin decreased the basal levels of ROS and 4-hydroxy-2-nonenal and depleted intracellular GSH contentExposure to quercetin failed to initiate programmed cell death cascadeExposure to quercetin induced changes in the expression of genes involved in intracellular signaling and induction of apoptosis

DiscussionReferences