Fitoterapia 80 (2009) 327–332

Contents lists available at ScienceDirect

Fitoterapia

j ourna l homepage: www.e lsev ie r.com/ locate / f i to te

Flavonoid-induced autophagy in hormone sensitive breast cancer cells

Elisa Brunelli, Giulia Pinton, Paolo Bellini, Alberto Minassi, Giovanni Appendino, Laura Moro⁎Dipartimento di Scienze Chimiche, Alimentari, Farmaceutiche e Farmacologiche and Drug and Food Biotechnology Center, Università degli Studi del Piemonte Orientale“Amedeo Avogadro”, Via Bovio 6, 28100 Novara, Italy

a r t i c l e i n f o

⁎ Corresponding author. Tel.: +39 0321 375820; faxE-mail address: moro@pharm.unipmn.it (L. Moro)

0367-326X/$ – see front matter © 2009 Elsevier B.V.doi:10.1016/j.fitote.2009.04.002

a b s t r a c t

Article history:Received 10 March 2009Received in revised form 31 March 2009Accepted 1 April 2009Available online 14 April 2009

The activity of 8-prenylapigenin (8-PA) and its 3'-methoxylated analogue isocannflavin B (IsoB)was investigated in estrogen-dependent T47-D and estrogen-independent MDA-MB-231human breast cancer cell lines. 8-PA showed a biphasic effect on T47-D cell proliferation,while no significant effect was observed on MDA-MB-231 cells. Conversely, IsoB exhibited onlyan inhibitory effect on T47-D cell proliferation, accompanied by the appearance of an intenseintracytoplasmic vacuolization of autophagic origin. Moreover, biochemical analysis showedthat IsoB reduced Akt phosphorylation and p21Cip1 expression in T47-D cells. These data showthat the prenylflavone moiety is a versatile platform for the induction and modulation ofbioactivity.

© 2009 Elsevier B.V. All rights reserved.

Keywords:8-Prenylapigenin8-PrenylnaringeninIsocannflavin BBreast cancer cellsCell signallingAutophagy

1. Introduction

Over the past few decades, a remarkable range ofbiological activities affecting various cellular systems hasbeen demonstrated for C-prenylated flavonoids, includingbinding to several hot targets of current biomedical research,like estrogen receptors (ERs) [1,2]. While in most cases it isnot clear whether the prenyl group has a specific role forbioactivity, some interesting findings emerged during thestudy of the estrogenic activity of 8-prenylnaringenin (8-PN,Fig. 1, 1), a hop and beer constituent [3]. Thus, the 8-prenylgroup is apparently engaged in specific hydrophobic interac-tions with the ER, since a shift to C-6 resulted in a loss ofestrogenic binding activity. The affinity of 8PN for ERs is atleast three orders of magnitude lower than that of oestradiol,an important factor when considering the biological transla-tion of its dietary intake [4]. Studies on oestrogen-dependenthuman breast cancer cell lines have shown that someflavonoids, depending on the concentration used, exhibit acombination of proliferative and antiproliferative effects [5].

: +39 0321 375821..

All rights reserved.

Upon entering cells, estrogens bind nuclear receptors, indu-cing their dimerization and interaction with DNA sequencesthat regulate gene transcription. Alternatively, rather thaninteracting directly with DNA, ERs bind DNA-associatedtranscription factors stimulating or repressing transcription[6,7]. Furthermore, very rapid and non-genomic actions havealso been associated to estrogens, possibly depending on theirinteraction with Src and activate the Src/Shc/Ras/Erk path-way [8]. Interference with this activation pathway abolishesthe hormone-dependent growth. Finally, estrogens can alsoactivate the PI3-kinase/Akt pathway, a critical component ofthe cell growth regulation, while the finding that p85α- andAkt K-expression in MCF-7 cells suppresses oestradiol activa-tion of the cyclin D1 promoter, suggests that oestradiol canalso increase cyclin D1 expression by stimulating cyclin D1transcription via PI3-kinase/Akt pathway [9].

We have recently demonstrated that the hop phytoestro-gen 8-PN inhibits the PI3-kinase/Akt pathway, inducingapoptosis in estrogen-dependent breast cancer cell lines[10,11]. In continuationwith these studies, we have comparedthe activity of 8-prenylnaringenin and various natural orsemi-synthetic analogues where the critical functionalizationof ring A is maintained, whereas rings B and C are varied, inestrogen-dependent T47-D and in the estrogen-independentMDA-MB-231 human breast cancer cell lines. In the course of

Fig. 1. Structural formula of 8-prenylnaringenin (8-PN, 1), 8-prenylapigenin(8-PA, 2a) and Isocannflavin B (IsoB, 2b).

Fig. 2. Effects of 8-PA and IsoB on T47-D cells andMDA-MB-231 proliferation.T47-D cells (panel A) or MDA-MB-231 (panel B) made quiescent as indicatedin Experimental section for 24 h, were incubated with different concentra-tions of 8-PA or IsoB, and the effect on cell growthwas examined after 24 h oftreatment. Results are expressed as cell number fold induction above thecontrol and given as means±SEM of triplicate determinations.

328 E. Brunelli et al. / Fitoterapia 80 (2009) 327–332

these studies, we discovered that isocannflavin B (IsoB, Fig. 1,2a), the unnatural C-8 regioisomer of the COX-inhibitingprinciple of marijuana (cannflavin B) [12], potently inducedautophagy in breast cancer cells, a type of activity not shownby either 8-PN or its flavone analogue 8-PA (Fig. 1, 2b).

2. Experimental

2.1. Reagents and antibodies

Polyclonal antibodies to Erk-1/2 MAP kinase, cyclin D1,p21CIP1and tubulin were from Santa Cruz Biotechnology(Santa Cruz, CA). Monoclonal antibody to Akt and phospho-specific polyclonal antibodies to Erk 1 (Thr 202 and Tyr 204),Erk2 (Thr 185 and Tyr 187) MAP kinases, and to Akt (Ser 473and Thr 308) were from Cell Signalling Technology (Beverly,MA). Nitrocellulose membranes and protein assay kits werefrom Bio-Rad (Hercules, CA). Anti mouse and anti rabbit IgGperoxidase conjugated antibodies and chemical reagentswere from Sigma-Aldrich (St Louis, MO). All reagents wereof analytical grade. 8-PA (1) and IsoB (2) were synthesizedaccording to literature [13,14]. Culture media, sera andantibiotics were from Invitrogen (Karlsruhe, Germany).

2.2. Cell cultures

Human breast cancer derived cell lines T47-D and MDA-MB-231 were purchased from ATCC. Breast cancer cells weregrown at 37 °C in a 5% CO2 in air atmosphere in RPMIsupplemented with phenol red, L-glutamine (2 mM), peni-cillin (100 U/ml), streptomycin (100 mg/ml), and 10% foetalcalf serum (FCS). Prior to the experiments, the cells werepassaged in the same media lacking phenol red and contain-ing charcoal-stripped FCS.

2.3. Proliferation assay by cell count

Breast cancer cells were seeded at a density of 5×104

cells/well on 6-well plates in growth medium with FCS andincubated overnight at 37 °C in a humidified environmentcontaining 5% CO2 to allow adherence. After 24 h treatment,cells were trypsinized and stained with Trypan blue, bymixing (1:1) with 4% Trypan blue (diluted in 1× phosphate-buffered saline). The number of cells which excluded Trypanblue (considered viable) was counted in a Burker chamberwithin 5 min after staining.

2.4. Cell morphologic analysis andMDC staining of cytoplasmaticvacuoles

T47-D and MDA-MB-231 cells growing exponentially oncover slips were treated respectively, with 25 µM of eachmolecule for 24 h. After treatment cells were fixed withmethanol and stained with Hemacolor® or Hoechst accordingto the supplier's recommended protocol. Slides weremounted and observed under a light or fluorescence micro-scope. Cytoplasmic vacuoles were stained with MDC accord-ing to the method described elsewhere [15]. After treatments,cells were exposed to 50 µM of MDC for 1 h at 37 °C and fixedwith 4% paraformaldehyde. Slides were observed under afluorescence microscope (Nikon, Japan).

2.5. Cell cycle analysis

Cell cycle/apoptosis analyses were performed usingpropidium iodide staining with subsequent FACS analysis.5×105 cells/well were cultured on tissue culture plates withor without 25 or 100 µM 8-PA and IsoB treatment for 24 h at37 ° C in a 5% CO2 atmosphere. After incubation, adherent cellswere detached with trypsin (0.5% trypsin/0.1% EDTA in PBS).Detached and suspended cells were harvested in completeDMEM medium and centrifuged at 500 g for 10 min. Pelletswere washed with PBS and fixed with ice cold 75% ethanolovernight at 4 °C, treated with 100 µg/ml RNAse A (Sigma),and subsequently stained with 25 µg/ml propidium iodide(Sigma). Then they were analyzed by using a flow cytometerFACS (Becton Dickinson, Franklin Lakes, NJ) and Modfitsoftware (Verity Software House, Inc., Topsham, Maine).

329E. Brunelli et al. / Fitoterapia 80 (2009) 327–332

2.6. Cell lysis and immunoblotting

Cells were extracted with 1% NP-40 lysis buffer (1% NP-40,150 mM NaCl, 50 mM Tris–HCl pH 8, 5 mM EDTA, 10 mM NaF,10 mMNa4P2O7, 0.4 mMNa3VO4,10 µg/ml leupeptin, 4 µg/mlpepstatin and 0.1 U/ml aprotinin). Cell lysates were centri-

Fig. 3. Staining of 8-PA and IsoB treated T47-D and MDA-MB-231 cells. T47-D and MDand then stained with Hemacolor (A–D) MDC (E–H) or Hoechst (I–N). Light or fluo

fuged at 13,000 ×g for 10 min and the supernatants werecollected and assayed for protein concentration with the Bio-Rad protein assay method (Bio-Rad). Proteins were separatedby SDS–PAGE under reducing conditions. Following SDS–PAGE, proteins were transferred to nitrocellulose, reactedwith specific antibodies and then detected with peroxidase-

A-MB-231 cells were left untreated or treated 24 h with 25 mM 8-PA or IsoBrescence microscopy photos (400× magnification) are reported.

330 E. Brunelli et al. / Fitoterapia 80 (2009) 327–332

conjugate secondary antibodies and chemioluminescent ECLreagent. When appropriate, the nitrocellulose membraneswere stripped according to manufacturer's recommendationsand re-probed. For Cyclin D1 and p21CIP1 expression, cellswere extracted in RIPA Buffer (1% Triton X-100, 0.1% SDS, 1%Na-deoxycholate,150mMNaCl, 50mMTris–HCl pH 7, 0.4mMNa3VO4, 10 µg/ml leupeptin, 4 µg/ml pepstatin and 0.1 U/mlaprotinin) and analysed as indicated above.

2.7. Statistical analysis

Results are expressed as arithmetic means (S.D.). Tocompare treatments, Student's t-test was applied. Differenceswere considered to be statistically significant when pb0.05.

Fig. 4. Cell cycle analysis of T47-D cells treated with 8-PA or IsoB. Cell cycle analysis wof treatment cells were harvested, fixed, stained with PI and analyzed by fluoresceneach phase of cell cycle. Each histogram plots cell count versus DNA content. In the hisecond peak represents cells in G2/Mwith 4N DNA content. Cells traversing S phasethe hypodiploid peak correspond to apoptotic cells. All experiments were performe

3. Results and discussion

The effects of increasing concentrations of 8-PA and IsoBon proliferation of estrogens sensitive ER (+) T47-D andinsensitive ER (−) MDA-MB-231 cells were investigated. Tothis purpose, exponentially growing cells were treated with abroad concentration range (1–100 µM) of 8-PA or IsoB for 24and 48 h and cells were counted. The dose–response profileobtained at 24 h of treatment (reported in Fig. 2A) shows that8-PA exerted a biphasic effect on T47-D cells proliferation,expressed by a stimulatory effect at low concentrations andan inhibitory effect at high concentrations. The stimulatoryeffect was mainly observed for concentrations rangingbetween 1 and 25 µM and was lost and replaced by a growth

as performed on T47-D cells treatedwith 25 or 100 µM 8-PA or IsoB. After 24 hce-activated cell sorting analysis. The data indicate the percentage of cells instogram, the first peak represents cells in G0/G1with 2N DNA content and theare between the two peaks with DNA content ranging from 2N to 4N whereasd in triplicate and gave similar results.

331E. Brunelli et al. / Fitoterapia 80 (2009) 327–332

inhibitory effect when treatments were performed withconcentrations greater than 25 µM, independently of theexposure time used. To determine if the biphasic effectdetected on ER (+) cells proliferation was equally present inan ER (−) breast cancer cell line, MDA-MB-231 cells wereexposed to the same serial range of concentrations for 24 and48 h. The dose–response profile obtained at 24 h (Fig. 2B)shows that treatments with low concentrations did notstimulate cell proliferation and only the inhibitory effectassociated with high concentrations (N50 µM) was observedin these cells. This absence of growth promoting effect in theER (−) breast cancer cell line suggests an involvement of ERsin the proliferative effect of both 8-PA and IsoB, which isconsistent with the known estrogenic activity of C-8 prenylflavonoids [16,17]. On the other hand, IsoB caused a cellgrowth arrest in T47-D cells starting from 1 µM concentration,while a toxic effect was evident from 25 µM dosage. No effectson MDA-MB-231 cell proliferation were observed, suggestingthat the prenyl group at C8 determines ER dependent effects,while the methoxy group is responsible of the differences inthe biological profile of these compounds. Surprisingly, IsoBinduced inhibition of T47-D cell proliferation was accompa-nied by the appearance of an intense intracytoplasmicvacuolization (Fig. 3, A–D), an observation that prompted us

Fig. 5. Effects of 8-PA and IsoB on signal transduction pathways. A) T47-D and MDAwere treated with 10, 25 or 100 µM 8-PA or IsoB for 15 min and detergent extracted.187) MAP kinases, and Akt (Ser 473 and Thr 308) were analyzed by immunoblottingequal amount of loading. B) T47-D cells in the same conditions as in A) were treattransferred to nitrocellulose, and blotted with anti-cyclin D1 or p21CIP1. The bands weout of three distinct experiments.

to undertake a more detailed characterization of the cyto-plasmic vacuoles detected. To this purpose, after IsoB treat-ment, cells were stained with monodansylcadaverine (MDC),a specific marker of autophagic vacuoles, and then analyzedby fluorescent microscopy (Fig. 3, E–H). Fluorescent dot-likestructures, characteristic of MDC-positive cells, were visua-lized in T47-D cells treated with IsoB, suggesting that thismolecule could induce autophagic death in ER (+) breastcancer cells. To further characterize this process, we analyzedthe presence of chromatin condensation and nuclear frag-mentation by fluorescence microscopy using the DNA-bind-ing fluorescent dye Hoechst 33258. In the presence of 25 µM8-PA or IsoB cultures presented nuclei with homogeneouschromatin distribution similar to control cells (Fig. 3, I–N).Moreover, in order to evaluate the effects of 8-PA and IsoBtreatment on cell cycle, we determined the ploidy status ofthe cells by flow cytometry. T47-D cells, exposed 24 h to 25 or100 µM 8-PA and IsoB, were labelled with propidium iodideand analysed by DNA flow cytometry using the MODFITsoftware. The results showed that the antiproliferativeactivity of IsoB was due to a G0/G1 cell-cycle arrest, with amaximal effect at 100 µMwithout any induction of apoptosis.Conversely, treatment with 100 µM 8-PA increased the peakarea corresponding to a hypodiploid (apoptotic) DNA content

-MB-231 cells made quiescent as indicated in Experimental section for 24 h,Levels of phosphorylated Erk 1 (Thr 202 and Tyr 204), Erk2 (Thr 185 and Tyr. Membranes were also blotted with antibodies to Erk 1/2 and Akt to show aned with 10 or 100 µM 8PN for 4 h. Cell extracts were run on 10% SDS–PAGE,re normalized to tubulin. Data reported here are a representative experiment

332 E. Brunelli et al. / Fitoterapia 80 (2009) 327–332

(Fig. 4). Since estrogens promote c-Src/MAPK and PI(3)K/Aktsignal transduction pathways involved in the control of cellproliferation, we investigated if 8-PA or IsoB treatments couldindeed interfere with the activation of these pathways in T47-D and MDA-MB-231 cells. When administered 15 min afterthe treatment, 100 µM 8-PA, and 25 and 100 µM IsoB reducedAkt phosphorylation in T47-D cell, while no effects were seenin MDA-MB-231 cells. Moreover, Erk 1/2 phosphorylationwas not affected in all tested conditions. Cyclin D1 regulatesthe cell-cycle progression through CDK4/6, is induced byhormones, and requires both Erk 1/2 and PI(3)K/Aktactivated pathways for induction in ER positive cells. Asshown in Fig. 5B, neither 8-PA nor IsoB could induce cyclin D1expression, but, interestingly, p21CIP1 expression was abol-ished by treatment with IsoB. The p21CIP1 factor was originallyidentified as a universal inhibitor of cyclin-dependent kinases(CDKs) belonging to the CIP/KIP family of CDK inhibitors [18].Recently, growing evidence has shown that p21CIP1 hasfunctions in addition to CDK inhibition. In particular, down-regulation of p21CIP1 was recently shown to be critical for theinduction of autophagy [19]. Both T47-D and MDA-MB-231cells express mutant forms of p53 [20] and even if p21CIP1 is agene that acts downstream of p53, in our system, aspreviously demonstrated in others, its expression could beregulated in a p53-independent way. Taken together, ourresults show that the hop phytoestrogen 8PN is an interestingmolecular framework to investigate estrogen-related andestrogen-independent biological responses of prenylflavo-noids, an important class of ubiquitous plant secondarymetabolites. In particular, the introduction of a methoxygroup on ring B in the analogue from the flavone series, leadsto autophagy induction in ER (+) cells. While themechanisticdetails of this activity need further investigation, the growinginterest for the biological role of autophagy make theidentification of new inducers an finding worth pursuingwith systematic structure–activity studies.

Acknowledgements

We thank Dr. Vallaro (DiSCAFF) for FACS analysis.This work was supported by Miur and Regione Piemonte.

References

[1] Kitaoka M, Kadokawa H, Sugano M, Ichikawa K, Taki M, Takaishi S, et al.Prenylflavonoids: a new class of non-steroidal phytoestrogen (part 1).

Isolation of 8-isopentenylnaringenin and an initial study on itsstructure–activity relationship. Planta Med 1998;64(6):511–5.

[2] MiyamotoM,Matsushita Y, KiyokawaA, Fukuda C, IijimaY, SuganoM, et al.Prenylflavonoids: a new class of non-steroidal phytoestrogen (part 2).Estrogenic effects of 8-isopentenylnaringenin on bonemetabolism. PlantaMed 1998;64(6):516–9.

[3] Milligan SR, Kalita JC, Pocock V, Van De Kauter V, Stevens JF, Deinzer ML,et al. Identification of a potent phytoestrogen in hops (Humus lupulus L.)and beer. J Clin Endocrinol Metabol 2000;85:4912–5.

[4] Schaefer O, Huempel M, Fritzemeier KH, Bohlmann R, Schleuning WD.8-Prenylnaringenin is a potent ERalpha selective phytoestrogen presentin hops and beer. J Steroid Biochem Mol Biol 2003;84:359–60.

[5] Matsumura A, Ghosh A, Pope GS, Darbre PD. Comparative study ofoestrogenic properties of eight phytoestrogens in MCF7 human breastcancer cells. J Steroid Biochem Mol Biol 2005;94(5):431–43.

[6] Marino M, Acconcia F, Ascenzi P. Estrogen receptor signalling: basesfor drug actions. Curr Drug Targets Immune Endocr Metabol Disord2005;5(3):305–14.

[7] Acconcia F, Kumar R. Signaling regulation of genomic and nongenomicfunctions of estrogen receptors. Cancer Lett 2006;238(1):1–14.

[8] Migliaccio A, Piccolo D, Castoria G, Di Domenico M, Bilancio A, LombardiM, et al. Activation of the Src/p21ras/Erk pathway by progesteronereceptor via cross-talk with estrogen receptor. EMBO J 1998;17(7):2008–18.

[9] Castoria G, Migliaccio A, Bilancio A, Di Domenico M, de Falco A,Lombardi M, et al. PI3-kinase in concert with Src promotes the S-phaseentry of estradiol-stimulated MCF-7 cells. EMBO J 2001;20(21):6050–9.

[10] Brunelli E, Minassi A, Appendino G, Moro L. 8-Prenylnaringenin, inhibitsestrogen receptor-alpha mediated cell growth and induces apoptosis inMCF-7 breast cancer cells. J Steroid Biochem Mol Biol 2007;107(3–5):140–8.

[11] Brunelli E, Pinton G, Chianale F, Graziani A, Appendino G, Moro L. 8-Prenylnaringenin inhibits epidermal growth factor-induced MCF-7breast cancer cell proliferation by targeting phosphatidylinositol-3-OHkinase activity. J Steroid Biochem Mol Biol 2009;113(3–5):163–70.

[12] Minassi A, Giana A, Ech-Chahad A, Appendino G. A regiodivergentsynthesis of ring A C-prenylflavones. Org Lett 2008;10:2267–70.

[13] Dong X, Fan Y, Yu L, Hu Y. Synthesis of four natural prenylflavonoids andtheir estrogen-like activities. Arch Pharm 2007;340:372–6.

[14] Minassi A, Giana A, Ech-Chahad A, Appendino G. A regiodivergentsynthesis of ring A C-prenylflavones. Org Lett 2008;10:2267–70.

[15] Biederbick A, Kern HF, Elsässer HP. Monodansylcadaverine (MDC) is aspecific in vivomarker for autophagic vacuoles. Eur J Cell Biol 1995;66(1):3–14.

[16] Zava DT, Blen M, Duwe G. Estrogenic activity of natural and syntheticestrogens in human breast cancer cells in culture. Environ HealthPerspect 1997;105(Suppl 3):637–45.

[17] Zava DT, Duwe G. Estrogenic and antiproliferative properties ofgenistein and other flavonoids in human breast cancer cells in vitro.Nutr Cancer 1997;27(1):31–40.

[18] Sherr CJ, Roberts JM. CDK inhibitors: positive and negative regulators ofG1-phase progression. Genes Dev 1999;13:1501–12.

[19] Fujiwara K, Daido S, Yamamoto A, Kobayashi R, Yokoyama T, Aoki H, etal. Pivotal role of the cyclin-dependent kinase inhibitor p21WAF1/CIP1in apoptosis and autophagy. J Biol Chem 2008;283(1):388–97.

[20] Wang YA, Johnson SK, Brown BL, McCarragher LM, Al-Sakkaf K, RoydsJA, et al. Enhanced anti-cancer effect of a phosphatidylinositol-3 kinaseinhibitor and doxorubicin on human breast epithelial cell lines withdifferent p53 and oestrogen receptor status. Int J Cancer 2008;123(7):1536–44.