TheProstate

Trimethoxy-ResveratrolandPiceatannolAdministeredOrallySuppress and Inhibit Tumor Formation andGrowth in

ProstateCancerXenografts

Steven J. Dias,1 Kun Li,1 Agnes M. Rimando,2 Swati Dhar,1 Cassia S. Mizuno,2

Alan D. Penman,3 and Anait S. Levenson1,4*1Cancer Institute,UniversityofMississippiMedical Center, Jackson,Mississippi

2United StatesDepartmentof Agriculture, Agricultural Research Service,Natural Products,Utilization ResearchUnit,University,Mississippi

3Centerof Biostatisticsand Bioinformatics,UniversityofMississippiMedical Center, Jackson,Mississippi4Departmentof Pathology,UniversityofMississippiMedical Center, Jackson,Mississippi

BACKGROUND. Resveratrol (Res) is recognized as a promising cancer chemopreventiondietary polyphenol with antioxidative, anti-inflammatory, and anticancer properties. Howev-er, the role of its analogues in prostate cancer (PCa) chemoprevention is unknown.METHODS. We synthesized several natural and synthetic analogues of Res and character-ized their effects on PCa cells in vitro using a cell proliferation assay. A colony formationassay and in vitro validation of luciferase (Luc) activity was done for LNCaP-Luc cells thatwere consequently used for in vivo studies. The efficacy of Res, trimethoxy-resveratrol (3M-Res) and piceatannol (PIC) was studied in a subcutaneous (s.c.) model of PCa using oralgavage. Tumor progression was monitored by traditional caliper and bioluminescent imag-ing. The levels of cytokines in serum were examined by ELISA, and the levels of compoundsin serum and tumor tissues were determined by gas chromatography-mass spectrometry.RESULTS. We examined the anti-proliferative activities of Res/analogues in three PCa celllines. We further compared the chemopreventive effects of oral Res, 3M-Res, and PIC inLNCaP-Luc-xenografts. We found that 2 weeks pretreatment with the compounds dimin-ished cell colonization, reduced tumor volume, and decreased tumor growth in the xeno-grafts. Both 3M-Res and PIC demonstrated higher potency in inhibiting tumor progressioncompared to Res. Notably, 3M-Res was the most active in inhibiting cell proliferation andsuppressing colony formation, and its accumulation in both serum and tumor tissues was thehighest.CONCLUSIONS. Our findings offer strong pre-clinical evidence for the utilization of die-tary stilbenes, particularly 3M-Res, as novel, potent, effective chemopreventive agents inPCa. Prostate # 2013 Wiley Periodicals, Inc.

KEY WORDS: prostate cancer; chemoprevention; dietary stilbenes; xenografts; oralgavage

INTRODUCTION

Nutritional chemoprevention is a particularlypromising approach for prostate cancer (PCa) due toits slow progression and predominance in elderlymen. Epidemiological data indicate a reduced PCarisk associated with red wine consumption, attribut-able mostly to a higher resveratrol (Res) content [1].Resveratrol (trans-3,5,40-trihydroxystilbene) is one of

Additional supporting information may be found in the online ver-sion of this article.

Grant sponsor: Intramural Research Support Program (IRSP).

Steven J. Dias and Kun Li contributed equally to this work.

*Correspondence to: Anait S. Levenson, Cancer Institute, Universityof Mississippi Medical Center, 2500 North State Street, Jackson, MS39216. E-mail: alevenson@umc.eduReceived 24 September 2012; Accepted 30 January 2013DOI 10.1002/pros.22657Published online in Wiley Online Library(wileyonlinelibrary.com).

� 2013WileyPeriodicals,Inc.

the best characterized stilbenes and is known to pos-sess a wide range of biological activities includingantioxidative and anti-inflammatory activity, cardio-vascular protection and cancer preventive, and thera-peutic effects. Although numerous studies supportthe use of Res in human chemoprevention, low bio-availability, and rapid metabolism has hindered itsprogression into practice [2,3]. Other diet-derived stil-benes chemically related to Res with the same back-bone, but differing in the type, number, and positionof substituents are of interest since they may possessgreater biopotency [4]. For example, substitution ofa hydroxyl (OH) group with a methoxy (OCH3)increases both the transport into cells and the meta-bolic stability of the molecule [5]. It is reasonable,therefore, to hypothesize that Res analogues mayshow higher biological activity and anticancer efficacyand may be developed as natural product drugs(dietary supplements) for PCa chemoprevention.

In the current study, we sought to compare and con-trast Res as a lead stilbene to its natural and syntheticanalogues (Fig. 1) on anticancer and chemopreventiveeffects in PCa in vitro and in vivo. The analoguesinvestigated in the current study commonly exhibitanti-inflammatory and anticancer activities. Trime-thoxy-resveratrol (trans-3,5,40-trimethoxystilbene; 3M-

Res) found in Pterobolium hexapetallum has demonstrat-ed anticancer effects by controlling cell cycle and in-ducing apoptosis [6–8]. 3,4,30,50-tetrahydroxylstilbene,piceatannol (PIC), another naturally occurring Res de-rivative found in grapes and red wine which containsone extra hydroxyl group, has been recognized as apowerful tyrosine kinase inhibitor in different types ofcancer [9–14]. Pterostilbene (trans-3,5-dimethoxy-40-hydroxystilbene; PTER), naturally found in blueberriesand grapes has therapeutic and anti-metastatic proper-ties in melanoma and hepatocellular carcinoma [15,16].We also synthesized 3,5,40-triacetylstilbene (3Ac-Res),3,5,-diacetylstilbene (2Ac-Res), and (E)-4-(3,5-dime-thoxystyryl)aniline (DMSA). Among these new com-pounds, 3Ac-Res has been reported to show anticancereffects in PCa and protection from g-irradiation-relateddeath in mice [7,17]. Interestingly, 2Ac-Res modulatedmitochondrial function in Leydig cells and inhibitedsteroidogenesis, which may be a potential therapeuticinterest for PCa [18]. Finally, there are preliminaryreports on greater in vivo absorption and metabolicstability of DMSA and its ability to inhibit cell migra-tion and invasion in breast cancer [19].

In the current study, we present a comparativeanalysis of Res analogues in inhibiting PCa cell prolif-eration and anchorage-independent growth in vitro

Fig. 1. Chemical structuresofnatural (firstrow)andsynthetic (secondrow)stilbenesusedin this study.Resveratrol (Res), trans-3,5,40 -trihy-droxystilbene; Trimethoxy-Resveratrol (3M-Res), trans-3,5,40 -trimethoxystilbene; Piceatannol (PIC), trans-3,5,3040 -tetrahydroxystilbene;Pterostilbene (PTER), trans-3,5-dimethoxystilbene; Diacetyl-Resveratrol (2Ac-Res), trans-3,5-diacetylstilbene; Triacetyl-Resveratrol(3Ac-Res), trans-3,5,40 -triacetylstilbene; (E)-4-(3,5-dimethoxystyryl)aniline (DMSA), trans-4-(3,5-dimethoxystyryl)aniline

2 Dias et al.

The Prostate

and, for the first time, we demonstrate in vivo the che-mopreventive activities of orally administered 3M-Res and PIC utilizing LNCaP-Luc xenografts.

MATERIALSANDMETHODS

Materials

Resveratrol and PIC were purchased from Sigma–Aldrich (St. Louis, MO) and Calbiochem–Novabio-chem (San Diego, CA), respectively. The other ana-logues were chemically synthesized at the USDA,ARS, National Products Utilization Research Unit inOxford, MS. Pterostilbene was synthesized as pre-viously described [20]. Trimethoxy-resveratrol andDMSA were synthesized according to previously pub-lished methods [21]. Triacetyl-Res and 2Ac-Res weresynthesized by acetylating Res. Resveratrol (18 mg in500 ml of MeOH), was treated with 10 drops of pyri-dine and 500 ml of acetic anhydride, and stirred over-night at room temperature. Ethyl acetate (EtOAc) wasadded to the reaction mixture and then partitionedwith water. The EtOAc layer was concentrated under astream of nitrogen, and spotted on a thin-layer chro-matography plate. The plate was developed using asolvent system of EtOAc:hexane (3:7). The acetylatedproducts, 3Ac-Res (Rf 0.8) and 2Ac-Res (Rf 0.6), werescraped from the plate and extracted with EtOAc. Thestructures of 3Ac-Res and 2Ac-Res, as well as otheranalogues synthesized, were confirmed by nuclearmagnetic resonance spectroscopy and mass spectrome-try. All compounds had �99% purity. Compoundswere dissolved either in high purity 100% ethanol orpure DMSO and stored in the dark at�208C.

CellCulture

LNCaP, Du145 (American Type Culture Collection,Manassas, VA), PC3M cells (a gift from Dr. R. Bergan,Northwestern University, Chicago, IL) [22,23] andLNCaP cells tagged with luciferase (LNCaP-Luc; a giftfrom Dr. R. Graeser, Tumor Biology Center, Freiburg,Germany) [24,25] were grown in RPMI 1640 containing10% FBS and 1X penicillin–streptomycin–amphotericinB at 378C/5% CO2 as previously described [26,27]. Forexperiments involving treatment with Res and ana-logues, cells were first plated in conventional mediawhich was replaced with phenol red-free RPMI 1640containing 5% charcoal-stripped serum 16–18 hr priorto treatment to eliminate steroid-background that caninterfere with their action. RPMI 1640 was obtainedfrom Gibco BRL (Gaithersburg, MD).

Cell ProliferationAssay

We used the CellTiter 96 AQueous Non-RadioactiveCell Proliferation Assay kit from Promega (Madison,

WI) according to the protocol provided by the manu-facturer (MTS method). Cells were seeded at 1–1.5 � 103 cells/well density (PC3M, Du145) or5 � 103 (LNCaP or LNCaP-Luc) in 96-well platesusing usual media. The media was changed to phenolred-free with 5% charcoal-stripped serum 16 hr priorto treatments. Fresh media containing different con-centrations of Res or its analogues was added everyother day for 6 days, after which the absorbanceof the formazan at 490 nm was measured using aSynergy-4 plate reader (BioTeK Instruments, Inc.,Winooski, VT). IC50 values were calculated by theChou–Martin method using the CompuSyn for DrugCombinations and for General Dose-Effect Analysissoftware (ComboSyn, Inc., Paramus, NJ).

ColonyFormationAssay

To make the hard agar layer, 1.2% agarose wasdissolved in PBS, boiled and immediately added to asix-well plate and kept in a 378C incubator to be so-lidified. To make the soft agar layer containing thecells, 2 � 104 LNCaP-Luc cells were suspended inthe mixture of 0.7% agarose solution with media con-taining compounds in different concentrations, and1 ml of this mixture was inoculated on top of thehard layer as described [28]. The media was replacedwith a fresh media containing compounds every2 days during the observation time. After 2 weeks ofincubation, when colonies were freely visible(>50 cells/colony), cells were fixed with formalde-hyde and stained with 0.01% of crystal violet solu-tion overnight at 48C. Colonies were analyzed at 4�magnification using an Olympus CKX41 microscopewith U-CMAD3 camera. The images were acquiredusing the Infinity Analyze software, version 5.0.3(Lumenera Corporation). Image Tool software wasused for counting the number of colonies in five toseven fields in each dish. The experiments weredone in triplicates twice. One-tailed Student’s t-testwas used to verify the differences between thegroups.

InVitroand InVivoBioluminescent (BL) Imaging

Bioluminescence was performed using an IVISSpectrum (Caliper Life Sciences, Hopkinton, MA).Images and detection of BL signals were acquired andanalyzed using Living Image Software V. 4.1. For invitro imaging, LNCaP-Luc cells were serially dilutedin a black, clear bottom 96-well plate (Costar, Corn-ing, NY). Following the addition of D-luciferin(15 mg/ml; Caliper), the plate was incubated at 378C,5% CO2 for 10 min after which images were taken(Supplementary Fig. 1). For in vivo BL imaging, anes-thetized mice (2% isoflurane sustained sedation) were

PotentDietaryStilbenes forProstateCancerChemoprevention 3

The Prostate

injected intraperitoneally (i.p.) with 150 mg D-lucifer-in/kg body weight (bw) and placed inside the camerabox for 10 min. Acquisition time was a maximum of3 min with auto-exposure depending on tumor biolu-minescence. Measurements of emitted light were per-formed for regions of interest (ROI) and quantified asPhoton Flux (photons/sec/cm2/sr). Normalizationwas done for all images at the end of the experiment.Gray scale images of mice were also collected at eachsession.

Subcutaneous (s.c.) PCaXenograftMouseModel

Five to six-week-old male athymic nude mice(Fox n1nu, Harlan Laboratory, Indianapolis, IN)were housed in a barrier room at the University ofMississippi Medical Center (UMMC) animal facility.Upon arrival, all animals were given a phytoestro-gen-free AIN-76A diet ad libitum. Animals wererandomly assigned to the following four groups,with six mice in each: (1) control vehicle-treated(sham); (2) treated with Res; (3) treated with 3M-Res;and (4) treated with PIC. After a week, the treat-ments groups started to receive their correspondingcompound by oral gavage (OG) every other day for2 weeks. The concentration of Res and analogueswas 50 mg/kg bw, based on published effectivedose in mice [29,30]. Resveratrol and analogues weredissolved in pure DMSO and given by OG in totalvolume of 200 ml. After 2 weeks of a preventive pro-tocol, the animals were injected s.c. with humanLNCaP-Luc cells (2.5 � 106) in 50% suspension withMatrigel (BD Biosciences, Bedford, MA; 200 ml totalvolume), and treatment was continued for another5 weeks. The cell inoculation was done for all ani-mals at the same time with the same batch of cells.Animals were monitored daily for signs of toxicity.Measurements of tumor volume by traditional Ver-nier digital caliper were carried out twice per week,and tumor volume was calculated as before [27].Body weight and BL signals were also monitoredtwice per week. The mice were sacrificed at day 38after tumor cell inoculation. At sacrifice, tumorswere excised, weighed, a portion was snap frozenunder liquid nitrogen for gas chromatography-massspectrometry (GC-MS) analysis, and the remainingwas placed in 10% formalin for histological andimmunohistochemical (IHC) analysis. Due to thesmall volume of compound-treated tumors, thelatter was only possible for a limited number oftumors. Blood was collected at sacrifice, and serumsamples were collected and stored at �808C. Theexperiments were conducted according to an Institu-tional Animal Care and Use Committee (IACUC)approved protocol.

AnalysisofRes, 3M-ResandPICContent inMouseSerumandTumorSamplesbyGC-MS

Serum samples were thawed on ice and centri-fuged at 408C for 15 min and treated with 80 ml of b-glucuronidase (1,250 U/ml potassium phosphatebuffer 75 mM, pH 6.8 at 378C). The mixture was incu-bated at 378C with shaking at 750 rpm for 17.5 hr andcooled to room temperature, then partitioned with75 ml EtOAc. The EtOAc layer was dried under astream of nitrogen, and derivatized with 30 ml of 1:1mixture of N,O-bis[trimethylsilyl]trifluoroacetamideand dimethylformamide (BSTFA:DMF; Pierce Bio-technology, Inc., Rockford, IL), with heating at 708Cfor 40 min and used for analysis of Res, 3M-Res, andPIC by GC-MS.

Prostate tissues, kept in �808C refrigerator untilused, were lyophilized. The freeze-dried tissues werehomogenized manually in a sodium acetatebuffer, pH 5 (1.0 ml/100 mg tissue), then treated withan enzyme solution (793.67 U/ml sulfatase and666.66 U/ml b-glucuronidase in sodium acetate buff-er, 300 ml/100 mg tissue), pH 5. The enzyme mixturewas incubated at 378C while shaking at 750 rpm for16 hr. 0.1 N HCl was added to the mixture (100 ml/100 mg tissue), vortex mixed, then partitioned withethyl acetate (200 ml � 2). The ethyl acetate layer wasdried under a stream of nitrogen. The dried samplewas derivatized with 30 ml BSTFA:DMF (1:1), withheating at 708C for 40 min and used for GC-MS analy-sis for determination of levels of Res, 3M-Res, andPIC.

We performed GC-MS using a GCMate II Instru-ment (JEOL, USA, Inc.) with a J&W DB-5 capillarycolumn (0.25 mm internal diameter, 0.25 mm filmthickness, 30 mm length; Agilent Technologies, FosterCity, CA). The GC temperature program was: initial1908C held for 1 min, increased to 2448C at the rate of308C/min and held at this temperature for 0.5 min,increased to 2468C at the rate of 0.58C/min and heldfor 0.5 min, increased to 2808C at the rate of 208C/min and held for 2 min, then finally increased to3008C at the rate of 308C/min and held for 0.8 min.The carrier gas was ultrahigh purity helium, at 1 ml/min flow rate. The injection port was kept at 2508C,the GC-MS interface and the ionization chamber at2308C. The volume of injection was 2 ml (splitless in-jection). The mass spectrum was acquired in selectedion-monitoring mode, electron impact 70 eV. Trime-thoxy-Res was monitored using m/z 271, 258, 240(retention time 8.3 min). Resveratrol was monitoredusing m/z 445, 430, and 413 (retention time 10.3 min).Piceatannol was monitored using m/z 531, 517, and430 (retention time 11.9 min). Quantitation was doneusing external standards of the analytes.

4 Dias et al.

The Prostate

Enzyme-Linked ImmunosorbentAssay (ELISA)

The levels of cytokines in murine serum was esti-mated by using mouse IL-6 OptEIATM ELISA kit (BDBiosciences, San Diego, CA) in accordance with themanufacturer’s instructions. Serum was diluted (1:10)at the time of analysis and the assay was performedusing 96-well EIA/RIA plates (Costar, Corning, NY).The reaction was stopped by TMB substrate (3,30,5,50-TetraMethylBenzidine), and absorbance was read ona Biotek Synergy-4 plate reader.

StatisticalAnalysis

Each experiment was reproduced at least threetimes. Values are expressed as the mean � SE of trip-licate measurements unless otherwise stated. Stu-dent’s one-tailed paired t-test was used to analyzedifferences between treatment groups in vitro. Differ-ences in tumor BL intensity in treatment groups wereevaluated as repeated measurements, data were natu-ral log-transformed and analyzed by Kruskal–Wallistest. The differences were considered significant atP � 0.05.

RESULTS

Resveratroland ItsAnalogues InhibitPCaCell Proliferation

We examined the effect of Res and its three natural(3M-Res, PIC, and PTER) and three synthetic (2Ac-Res, 3Ac-Res, and DMSA) analogues on the growth ofLNCaP, Du145, and PC3M cells. Cells were plated inphenol red-free media supplemented with 5% char-coal-stripped FBS and cultured in the presence of var-ious concentrations of compounds (1–100 mM) for6 days. As shown in Figure 2, Res and analoguesinhibited the growth of all three cell lines in a dose-dependent manner, however, with different potency.The cell growth inhibitory effects of analogues werecell-specific. IC50 values, the compound concentrationcausing 50% maximal inhibition of cells, were calcu-lated for each compound in each cell line by theChou–Martin method (see Materials and MethodsSection). The IC50 of Res for LNCaP was 12.7 mMwhile for Du145 and PC3M were 42.3 and 31.5 mM,respectively. All six analogues displayed greater cellgrowth inhibition than Res in Du145 cells, while fouranalogues, namely 3M-Res, 3Ac-Res, DMSA, andPTER showed higher potency than Res in PC3M cells,and two, 3M-Res and 3Ac-Res, in LNCaP cells. 3M-Res exhibited the most potent inhibition in LNCaP(IC50 ¼ 2.5 mM) and Du145 (IC50 ¼ 9.7 mM) cells, fol-lowed by 3Ac-Res (IC50 ¼ 10.5 and 14.2 mM, respec-tively; Fig. 2A and B). 3M-Res also showed more

potent activity than Res in PC3M cells (IC50 ¼ 23.3 vs.31.5 mM; Fig. 2C). It is apparent that replacement ofall three hydroxyl substituents with methoxyl (OCH3)or acetyl (COCH3) groups resulted in considerablegrowth inhibitory activity. This effect was also seen inall three cell lines with DMSA, which has an aminogroup (NH2) in place of the hydroxyl group on ringA, and 3,5-dimethoxy substitution on ring B.

3M-ResandPICTreatmentResultedinaSignificantInhibitionofLNCaP-LucColonyFormation

To predict the effects of Res and its two naturalanalogues, 3M-Res and PIC, on in vivo cell survival,colonization, and growth, we analyzed anchorage-in-dependent growth of LNCaP-Luc cells treated withthree doses (5, 50, and 100 mM) of each compoundusing a colony formation assay in soft agar. As shownin Figure 3, there was a dose-dependent decrease inthe colony formation ability of LNCaP-Luc cellstreated with all three compounds compared to un-treated cells. However, the extent of colony formationinhibition was much greater in 3M-Res-treated cellswith major reductions in size and number of coloniescompared to Res- and PIC-treated cells, which wasmost evident at the low 5 mM concentration.

Oral 3M-Res andPICInhibitTumorGrowthinLNCaP-LucXenografts

To evaluate the in vivo protective and anticancereffects of selected natural analogues (3M-Res andPIC) compared to Res, we used s.c. xenograft modelwith LNCaP-Luc cells. The protocol was designed todetermine the chemopreventive capabilities of thenatural analogues and is outlined in SupplementaryFigure 2. The animals were randomized into fourgroups (n ¼ 6/each), and treatment groups began re-ceiving oral administration of Res, 3M-Res, or PIC forthe next 2 weeks, whereas sham mice receivedDMSO-vehicle pretreatment. After 2 weeks of pre-treatment, LNCaP-Luc cells were implanted, and ad-ministration of compounds or vehicle by OG wascontinued for the next 5 weeks. As shown inFigure 4A and B, Res, 3M-Res, and PIC inhibited tu-mor formation, development, and progression com-pared to sham control (caliper measurements).Indeed, although tumor intake was 100%, there wereconsiderably diminished initial tumor volumes andsuppressed subsequent growth in animals pretreatedwith compounds. The differences between sham con-trol and treatment groups reached statistical signifi-cance at day 18 (P < 0.05). There was no significantchange in body weights or visible signs of toxicityover the course of the experiment although the shammice showed a tendency to lose weight in the first

PotentDietaryStilbenes forProstateCancerChemoprevention 5

The Prostate

Fig. 2. Effects ofRes andits analogues onPCacellproliferation.A: 3M-Res and3Ac-Resdisplay themostpotentgrowth inhibitory activityin LNCaP cells.B: 3M-Res and 3Ac-Res display the most potent growth inhibitory activity in Du145 cells.C: DMSA and PTER display themost potent growth inhibitory in PC3M cells. Cells were grown in phenol red-free plus 5% charcoal stripped serum media containingvarious concentrations of Res/analogues (1^100 mM).Cell proliferation assay (MTS) was performedwith formazanmeasurements at day 6.Left, viable cells are plotted as fraction of untreated cells (Ctrl, horizontal line) which is set to1.Data represent themean � SE of three tosix independent experiments in which each treatment (data point) was performed in triplicates. P-values were calculated using one-tailStudent’s t-test. Significantly different from the Ctrl,*P � 0.05, #P � 0.01, �P � 0.001.Right, the IC50 values were calculatedby the Chou^MartinmethodusingCompuSynsoftware.

6 Dias et al.

The Prostate

2 weeks after cell injection, during colonization(Fig. 4C). We did not detect appreciable differencesamong Res and analogues-treated groups by conven-tional caliper measurements. Auspiciously, we alsomonitored tumors by the more sensitive BL imaging,which allowed us to detect subtle differences betweenRes-treated and analogues-treated xenografts. Suchimaging revealed that the tumors went through a re-gression period up to 10–12 days of cell colonizationfollowed by regrowth and subsequent responses totreatments at week 3–5 (Fig. 4D). Representative micefrom each group are shown, and different bar scalesfor sham group (�107) and treatments groups (�106)

are indicated. Significant differences between treat-ments groups were detected at day 38 (week 5;Fig. 4E). As can be seen in the quantitative Total Fluxanalysis, there was a suppressed tumor progressionin 3M-Res and PIC compared to Res group. Specifical-ly, while tumors in Res-treated group showed no ap-parent changes in growth, tumors in PIC-treated and3M-Res-treated mice were reduced in volume andtended towards decreased growth (PIC) and volumeregression (3M-Res; Fig. 4E). Although 3M-Res wasthe only treatment that showed clear and convincingdecrease in tumor growth, the differences betweenRes-treated tumors and both 3M-Res-treated an PIC-

Fig. 3. Resveratrol-, 3M-Res-, and PIC-treated LNCaP-Luc cells havereduced tumorigenicity in soft agar. Anchorage-independentgrowthofLNCaP-Luccellsin theabsence(Ctrl)orpresenceofdifferentdosesofcompoundswasdeterminedafter2weeksincubation.Representativeimages arepresented (40�).Thenumberofcolonieswithdiameter�3 mmwerecounted, andresults are shownasmeans � SEof threeinde-pendentexperiments (topright).P-valueswerecalculatedusingone-tailedStudent’st-test.���P � 0.001comparedwithuntreatedCtrl.

PotentDietaryStilbenes forProstateCancerChemoprevention 7

The Prostate

treated tumors were statistically significant at the endof the experiment (P < 0.05).

Analysisof Steady-StateLevelsofRes, 3M-Res,andPICinLNCaP-LucXenografts

We next analyzed the concentrations of Res,3M-Res, and PIC in serum and tumor samples takenat sacrifice (Table I). Even though the same doses(50 mg/kg) of compounds were administered by OGevery other day starting with pretreatment and

through the duration of experiment (total 52 days, 26dosing), we found that the average serum Res concen-tration was 0.02 � 0.01 mg/ml (mean � SE, n ¼ 5),while the average serum for 3M-Res was 0.94 �0.55 mg/ml (mean � SE, n ¼ 3). Res and 3M-Res werealso detectable in tumor tissues, although due to thesmall volume of tumors, tissues from the same treat-ment group were combined and measured together.Once again, 3M-Res demonstrated greater accumula-tion than Res (3.32 mg/g vs. 0.01 mg/g). Piceatannolwas not detected in serum or in tumor tissues.

Fig. 4. Pretreatment of mice with oral Res, 3M-Res, or PIC reduced and inhibited tumor formation and growth in xenografts.Male nudemicewere injected s.c.with 2.5 � 106 LNCaP-Luc cells after 2weeks ofpretreatmentbyoral administrationofcompounds (50 mg/kg) or ve-hicleDMSO(sham).Treatmentswerecontinuedeveryotherday for another5weeks (seeSupplementaryFig. 2 fordetails).A: Imagesofrepre-sentativemice from each group at day 22 are shown. At the endpoint, micewere sacrificed and tumors were excised and photographed.B:Tumor growthwasmonitoredusing a digital caliper twice aweek. Significant tumor growth inhibitionwas detected in all compound-treatedversus control sham group starting at day18.Themeans � SE are shown (n ¼ 6/each). �P � 0.05; ��P � 0.01.C: Body weights of sham andcompound-treatedgroupsduring the study.Notereversible drop in averageweights of shammice soonaftercancercell inoculation.D:Tumorgrowthwasmonitored by BL imaging.Representative images of nudemice bearing LNCaP-Luc tumors and showing the trend in each treat-mentgroupare shown.Theminimumandmaximumphotons/secofdifferentgroups areindicatedineachrainbowbar scale.Imageswere takentwiceperweekby injecting D-luciferin andusing IVIS Imaging System.E: Quantitative analysis of BL signals for compound-treatedgroupsbe-ginning on day 2 after cancer cell injections. Light emission data inTotal Flux (photons/sec) are log-transformed and plotted on a linear scalewithlog trendlines against time.P-valueswerecalculatedbyKruskal^Wallis test.�P � 0.05.

8 Dias et al.

The Prostate

Resveratrol, 3M-Res, andPICInhibitIL-6 SecretioninMurineSerum

In an attempt to elucidate the mechanism underly-ing the more potent chemopreventive effects of 3M-Res and PIC compared to Res, we analyzed the tumorinhibitory effects in relation to their anti-inflammato-ry activity. We examined levels of pro-inflammatorycytokine IL-6 in mice serum using ELISA (Fig. 5) andfound that average levels of IL-6 in serum from micetreated with compounds were significantly lowerthan in control sham mice (P < 0.001). Although thedifferences between IL-6 levels were not significantamong the treatment groups, there was a trend to-wards lower IL-6 in 3M-Res- and PIC-treated micecompared to Res-treated mice.

DISCUSSION

The goal of this study was to evaluate and comparethe bioactivity of Res and its analogues to find chemi-cally related agents with better pharmacokineticproperties and superior pharmacological potency foruse as dietary chemopreventive supplements in PCa.

Resveratrol and analogues demonstrated growthinhibitory effects in a panel of PCa cell lines repre-senting different stages of PCa progression. Anti-pro-liferative effects of Res in PCa cells have beenreported previously by us and others [31–33]. In thisstudy, we demonstrated that although analogues hadcell-specific growth inhibitory potency, 3M-Res wasconsistently more potent than all the other analoguestested. In fact, it showed the most potent anti-prolifer-ative activity in LNCaP and Du145 cells and higherthan Res activity in PC3M cells.

We chose to use LNCaP-Luc cells for our in vivostudies since these androgen-responsive cells repre-sent earlier stages of PCa and are most appropriatefor chemoprevention protocol. Since 3M-Res and PICdemonstrated significant inhibition of anchorage-in-dependent growth of LNCaP-Luc cells in the colonyformation assay, we proceeded to compare preventiveand anticancer effects of these compounds in vivo.For this, we used s.c. PCa xenografts treated 2 weeksprior to cancer cell inoculation with orally adminis-tered Res, 3M-Res, and PIC to mimic aspects of che-moprevention. The current chemopreventive protocolwas developed as an extension to our encouragingprevious study of LNCaP-Luc orthotopic xenograftmodel, in which we provided Res in diet (�200 mg/kg bw/day) and documented diminished coloniza-tion and reduced development of tumors in Res-treated group. Regrettably, we also detected variableand uncontrolled diet consumption that affected sta-tistically reliable analysis of the data (not shown). Inthe current study, we used lower doses of Res andcompounds (50 mg/kg bw/day) and ensured admin-istration and direct delivery of agents by oral gavage.Our results demonstrated that tumor formation,development, and progression were significantlydelayed and decreased in mice treated with com-pounds compared to sham-treated mice. Althoughthe same amount of cells was injected into all mice,soon after cell inoculation it was evident that com-pound-pretreated mice had a different predispositionfor cell colonization, tumor formation, and develop-ment. This was consistent with our earlier observationwith colony formation assay, in which we showedless clonogenic abilities of cells treated with com-pounds versus untreated cells. In addition, consider-ing the well-established strong connection betweenchronic inflammation and cancer, these anti-

TABLE I. Res, 3M-Res, and PIC Steady-State Levels inLNCaP-Luc Xenografts Serum and TumorTissues After aTotal 52-DayDosingbyOralGavage

StilbeneDose

(mg/kg/day)Serum

(mg/ml)Tumor

(mg/g, FW)

Res 50 0.022 � 0.01 0.014a

3M-Res 50 0.937 � 0.55 3.324a

PIC 50 NI NI

Serum (n ¼ 3–5) and tumor tissues were isolated at sacrifice.Res, 3M-Res, and PIC levels were measured and quantified byGC-MS method. FW, fresh weight.aTwo repeated measurements of combined tumors within eachgroup: NI, not identified.

Fig. 5. Reduction of pro-inflammatory IL-6 cytokine by Res,3M-Res, and PIC in murine serum analyzed by ELISA.Total bloodwascollectedfrommice fromeachgroup;Ctrl (n ¼ 3),Res (n ¼ 5),3M-Res(n ¼ 3),andPIC(n ¼ 5)at sacrifice,andserumwasisolatedandstoredat�8008C.Serumwasdiluted (1:10) at the timeof analy-sis and sandwich ELISA was performed. P-values were calculatedusing Student’s t-test. ���P � 0.001 compared to untreated Ctrlsham.Insert:Differencesof IL-6 levels among treatmentgroups.

PotentDietaryStilbenes forProstateCancerChemoprevention 9

The Prostate

inflammatory agents evidently contributed to the dif-ferences in tumor microenvironment, resulting in aninitial diminished tumor volume and reduced suscep-tibility to develop tumors [34,35]. Indeed, we detectedsignificantly lower levels of the pro-inflammatory cy-tokine IL-6 in serum from compound-treated micecompared to untreated sham mice. Interestingly, a re-cent report suggested the inhibition of IL-6 secretionin PCa cells as one of the mechanisms by which PICinhibits STAT3 activation, which in turn reduces theexpression of angiogenesis-related proteins resultingin inhibition of migration and invasion [13]. Due tothe small volumes of 3M-Res and PIC-treated tumors,IHC analysis was performed with a limited amountof tumors. Immunohistochemistry showed signifi-cantly higher levels of Ki67 in control sham mice com-pared to available Res-treated mice (data not shown).Taken together, our results indicate that pretreatmentof mice with dietary phytochemicals established a mi-croenvironment in which tumor initiation and subse-quent development and progression was diminished.Studies have suggested that the tumor microenviron-ment play an essential role in tumor progression [36].An important observation of the current study is thatthe microenvironment is important not only for tumorprogression but also for tumor initiation, which is theessence of chemoprevention.

One of the goals of our study was to distinguishbetween the effects of Res and its potent analogues.While conventional caliper measurements were effec-tive in measuring growing tumors in the vehicle-treated group, it did not allow for an appreciable dis-tinction among small, compounds-treated tumors(Fig. 4B). In contrast, BL imaging had an advantageover the traditional method of enabling us to deter-mine differences between Res, 3M-Res, and PICeffects but exhibited its own limitations towards largetumors. Specifically, data obtained with BL imagingof sham mice were inconsistent: the tumor boundaryvisualized by eye and image-processing software (agray scale body surface image) was not always corre-lated with tumor luciferase signals: large untreatedtumors had un-matching photon emission. This mayresult because of at least two reasons: first, only meta-bolically active cells emit luciferase activity. Largetumors often consist of living and dead cells in addi-tion to necrotic areas and peritumoral edema that allcontribute to caliper-based measurements but mostlikely interfere with BL imaging [37–39]. Secondly,substrate availability for large s.c. tumors is much lessthan for organs in the peritoneal cavity due to theirremote location from the site of the D-luciferin i.p. in-jection. This latter observation in our experience maysupport an advantage of BL imaging for orthotopicmodels and small s.c tumors.

In this study we found that 3M-Res was the mostpotent compound tested. The increased biopotencyand anticancer effects of 3M-Res could be explainedby our findings of its superior anti-proliferative andanti-clonogenic properties and its higher accumula-tion in serum and tumor tissue compared to Res. Inagreement, others also demonstrated greater potencyfor 3M-Res compared to Res in reducing colony for-mation, promoting cell cycle arrest, and inducing p53-dependent apoptosis in PCa and breast cancer cells[7,30,40]. Additional studies have also shown 3M-Resto be much more potent than Res in inhibiting cellgrowth: it was 20 times more active in pulmonary ar-tery smooth muscle cells [41] and 100 times more ac-tive in human colon cancer cells [42]. To the best ofour knowledge, there are only two reports on the invivo anti-tumor activity of 3M-Res, albeit using hu-man colorectal carcinoma cells in SCID mice [29,43].Pan et al. [29] demonstrated therapeutic efficacy of3M-Res at a concentration of 50 mg/kg by i.p. injec-tions, while Paul et al. [43] administered i.p. 10 mg/kg/bw once daily for 3 weeks. We are the first toshow potent chemopreventive efficacy using the50 mg/kg dose by oral administration in PCa xeno-grafts. We detected high levels of 3M-Res in serumand tumor tissue, which is consistent with reportedlonger half-life, greater plasma exposure, and lowerclearance of 3M-Res compared to Res [44]. The favor-able pharmacokinetic property of 3M-Res is related toits structural modification of having methoxy groups,which increase its metabolic stability and subsequentbiopotency. The methoxy groups also make 3M-Resmore lipophilic and highly distributed into fat tissues.In general, 3M-Res distribution in major organs hasbeen shown to be higher than in plasma [45]. Picea-tannol, on the other hand, was undetectable by GC-MS in our samples despite exhibiting strong preven-tive and anti-tumor effects. Piceatannol, having an ad-ditional hydroxyl group, has a short plasma half-life(4.23 hr), and undergoes extensive and rapid glucuro-nidation after a single intravenous administration of10 mg/kg in rats [46]. Piceatannol also had a high vol-ume of distribution value (10.76 L/kg) indicating thatit is very much distributed into tissue [46]. Thesepharmacokinetic properties of PIC may explain unde-tectable serum and prostate levels in our study. Nev-ertheless, a broad spectrum of biological activity ofPIC has been reported. It has been shown that PIC hasanti-proliferative and apoptotic effects in a wide vari-ety of tumor cells mediated by various mechanismsincluding regulation of cell cycle, induction of pro-apoptotic factors, and regulation of signaling path-ways implicated in inflammation and cancer [47]. Thechemopreventive and therapeutic potential of PICwas also demonstrated in animal models including a

10 Dias et al.

The Prostate

recent report on the inhibition of lung metastasis inan experimental metastasis model of PCa [48].

In summary, for the first time, our pre-clinicalstudy using PCa xenograft model and oral adminis-tration of dietary stilbenes demonstrated successfulreduction of tumor volume and inhibition of tumorformation, development, and progression. We foundthat anti-proliferative, anti-inflammatory, and anti-clonogenic properties of the compounds played anessential role in modulating tumor microenvironmentand predisposition of cancer cells to a less aggressivephenotype with diminished ability to colonize andgrow. Notably, both 3M-Res and PIC demonstratedmore potent activities in vitro and in vivo comparedto Res. In conclusion, we suggest that 3M-Res andPIC are potential chemopreventive agents for PCaand anticipate that our findings will facilitate thedevelopment of potent stilbenes as natural productdrugs for cancer prevention and treatment.

ACKNOWLEDGMENTS

We are grateful to Drs. Gomez, Espinoza, andRaucher (Cancer Institute, UMMC) for their contribu-tions to the study. We are also grateful to Dr. Miele(Cancer Institute, UMMC) for valuable discussions.The study was partially supported by the IntramuralResearch Support Program (IRSP) grant from UMMCto A.S.L.

REFERENCES

1. Schoonen WM, Salinas CA, Kiemeney LA, Stanford JL. Alcoholconsumption and risk of prostate cancer in middle-aged men.Int J Cancer 2005;113(1):133–140.

2. Patel KR, Brown VA, Jones DJ, Britton RG, Hemingway D,Miller AS, West KP, Booth TD, Perloff M, Crowell JA, BrennerDE, Steward WP, Gescher AJ, Brown K. Clinical pharmacologyof resveratrol and its metabolites in colorectal cancer patients.Cancer Res 2010;70(19):7392–7399.

3. Syed DN, Khan N, Afaq F, Mukhtar H. Chemoprevention ofprostate cancer through dietary agents: Progress and promise.Cancer Epidemiol Biomarkers Prev 2007;16(11):2193–2203.

4. Aggarwal BB, Bhardwaj A, Aggarwal RS, Seeram NP, Shisho-dia S, Takada Y. Role of resveratrol in prevention and therapyof cancer: Preclinical and clinical studies. Anticancer Res2004;24(5A):2783–2840.

5. Lin HS, Yue BD, Ho PC. Determination of pterostilbene in ratplasma by a simple HPLC-UV method and its application inpre-clinical pharmacokinetic study. Biomed Chromatogr 2009;23(12):1308–1315.

6. Wang TT, Schoene NW, Kim YS, Mizuno CS, Rimando AM.Differential effects of resveratrol and its naturally occurringmethylether analogs on cell cycle and apoptosis in humanandrogen-responsive LNCaP cancer cells. Mol Nutr Food Res2010;54(3):335–344.

7. Hsieh TC, Huang YC, Wu JM. Control of prostate cell growth,DNA damage and repair and gene expression by resveratrolanalogues, in vitro. Carcinogenesis 2011;32(1):93–101.

8. Weng CJ, Yang YT, Ho CT, Yen GC. Mechanisms of apoptoticeffects induced by resveratrol, dibenzoylmethane, and theiranalogues on human lung carcinoma cells. J Agric Food Chem2009;57(12):5235–5243.

9. Kang CH, Moon DO, Choi YH, Choi IW, Moon SK, Kim WJ,Kim GY. Piceatannol enhances TRAIL-induced apoptosis inhuman leukemia THP-1 cells through Sp1- and ERK-depen-dent DR5 up-regulation. Toxicol In Vitro 2011;25(3):605–612.

10. Vo NT, Madlener S, Bago-Horvath Z, Herbacek I, Stark N,Gridling M, Probst P, Giessrigl B, Bauer S, Vonach C, Saiko P,Grusch M, Szekeres T, Fritzer-Szekeres M, Jager W, KrupitzaG, Soleiman A. Pro- and anticarcinogenic mechanisms of picea-tannol are activated dose dependently in MCF-7 breast cancercells. Carcinogenesis 2010;31(12):2074–2081.

11. Thakkar K, Geahlen RL, Cushman M. Synthesis and protein-tyrosine kinase inhibitory activity of polyhydroxylated stilbeneanalogues of piceatannol. J Med Chem 1993;36(20):2950–2955.

12. Fleming I, Fisslthaler B, Busse R. Calcium signaling in endothe-lial cells involves activation of tyrosine kinases and leads toactivation of mitogen-activated protein kinases. Circ Res1995;76(4):522–529.

13. Su L, David M. Distinct mechanisms of STAT phosphorylationvia the interferon-alpha/beta receptor. Selective inhibition ofSTAT3 and STAT5 by piceatannol. J Biol Chem2000;275(17):12661–12666.

14. Roupe KA, Remsberg CM, Yanez JA, Davies NM. Pharmaco-metrics of stilbenes: Seguing towards the clinic. Curr ClinPharmacol 2006;1(1):81–101.

15. Pan MH, Chiou YS, Chen WJ, Wang JM, Badmaev V, Ho CT.Pterostilbene inhibited tumor invasion via suppressing multi-ple signal transduction pathways in human hepatocellular car-cinoma cells. Carcinogenesis 2009;30(7):1234–1242.

16. Pan MH, Chang YH, Tsai ML, Lai CS, Ho SY, Badmaev V, HoCT. Pterostilbene suppressed lipopolysaccharide-induced up-expression of iNOS and COX-2 in murine macrophages.J Agric Food Chem 2008;56(16):7502–7509.

17. Koide K, Osman S, Garner AL, Song F, Dixon T, GreenbergerJS, Epperly MW. The Use of 3,5,40-Tri-O-acetylresveratrol as apotential pro-drug for resveratrol protects mice from gamma-irradiation-induced death. ACS Med Chem Lett 2011;2(4):270–274.

18. Svechnikov K, Spatafora C, Svechnikova I, Tringali C, Soder O.Effects of resveratrol analogs on steroidogenesis and mitochon-drial function in rat Leydig cells in vitro. J Appl Toxicol2009;29(8):673–680.

19. Kondratyuk TP, Park EJ, Marler LE, Ahn S, Yuan Y, Choi Y, YuR, van Breemen RB, Sun B, Hoshino J, Cushman M, JermihovKC, Mesecar AD, Grubbs CJ, Pezzuto JM. Resveratrol deriva-tives as promising chemopreventive agents with improved po-tency and selectivity. Mol Nutr Food Res 2011;55(8):1249–1265.

20. Paul S, DeCastro AJ, Lee HJ, Smolarek AK, So JY, Simi B,Wang CX, Zhou R, Rimando AM, Suh N. Dietary intake ofpterostilbene, a constituent of blueberries, inhibits the beta-cat-enin/p65 downstream signaling pathway and colon carcino-genesis in rats. Carcinogenesis 2010;31(7):1272–1278.

21. Rimando AM, Cuendet M, Desmarchelier C, Mehta RG, Pez-zuto JM, Duke SO. Cancer chemopreventive and antioxidantactivities of pterostilbene, a naturally occurring analogue ofresveratrol. J Agric Food Chem 2002;50(12):3453–3457.

22. Lakshman M, Xu L, Ananthanarayanan V, Cooper J, TakimotoCH, Helenowski I, Pelling JC, Bergan RC. Dietary genisteininhibits metastasis of human prostate cancer in mice. CancerRes 2008;68(6):2024–2032.

PotentDietaryStilbenes forProstateCancerChemoprevention 11

The Prostate

23. Xu L, Ding Y, Catalona WJ, Yang XJ, Anderson WF, JovanovicB, Wellman K, Killmer J, Huang X, Scheidt KA, MontgomeryRB, Bergan RC. MEK4 function, genistein treatment, and inva-sion of human prostate cancer cells. J Natl Cancer Inst2009;101(16):1141–1155.

24. Graeser R, Chung DE, Esser N, Moor S, Schachtele C, Unger C,Kratz F. Synthesis and biological evaluation of an albumin-binding prodrug of doxorubicin that is cleaved by prostate-specific antigen (PSA) in a PSA-positive orthotopic prostatecarcinoma model (LNCaP). Int J Cancer 2008;122(5):1145–1154.

25. Jantscheff P, Ziroli V, Esser N, Graeser R, Kluth J, SukolinskayaA, Taylor LA, Unger C, Massing U. Anti-metastatic effects ofliposomal gemcitabine in a human orthotopic LNCaP prostatecancer xenograft model. Clin Exp Metastasis 2009;26(8):981–992.

26. Kai L, Samuel SK, Levenson AS. Resveratrol enhances p53 acet-ylation and apoptosis in prostate cancer by inhibiting MTA1/NuRD complex. Int J Cancer 2010;126(7):1538–1548.

27. Kai L, Wang J, Ivanovic M, Chung YT, Laskin WB, Schulze-Hoepfner F, Mirochnik Y, Satcher RL Jr, Levenson AS. Target-ing prostate cancer angiogenesis through metastasis-associatedprotein 1 (MTA1). Prostate 2011;71(3):268–280.

28. Aziz MH, Nihal M, Fu VX, Jarrard DF, Ahmad N. Resveratrol-caused apoptosis of human prostate carcinoma LNCaP cells ismediated via modulation of phosphatidylinositol 3’-kinase/Akt pathway and Bcl-2 family proteins. Mol Cancer Ther2006;5(5):1335–1341.

29. Pan MH, Gao JH, Lai CS, Wang YJ, Chen WM, Lo CY, WangM, Dushenkov S, Ho CT. Antitumor activity of 3,5,40-trime-thoxystilbene in COLO 205 cells and xenografts in SCID mice.Mol Carcinog 2008;47(3):184–196.

30. Narayanan NK, Nargi D, Randolph C, Narayanan BA. Lipo-some encapsulation of curcumin and resveratrol in combina-tion reduces prostate cancer incidence in PTEN knockout mice.Int J Cancer 2009;125(1):1–8.

31. Kai L, Levenson AS. Combination of resveratrol and antiandro-gen flutamide has synergistic effect on androgen receptor inhi-bition in prostate cancer cells. Anticancer Res 2011;31(10):3323–3330.

32. Narayanan BA, Narayanan NK, Re GG, Nixon DW. Differen-tial expression of genes induced by resveratrol in LNCaP cells:P53-mediated molecular targets. Int J Cancer 2003;104(2):204–212.

33. Mitchell SH, Zhu W, Young CY. Resveratrol inhibits the ex-pression and function of the androgen receptor in LNCaP pros-tate cancer cells. Cancer Res 1999;59(23):5892–5895.

34. Paul S, Rimando AM, Lee HJ, Ji Y, Reddy BS, Suh N. Anti-inflammatory action of pterostilbene is mediated through thep38 mitogen-activated protein kinase pathway in colon cancercells. Cancer Prev Res (Phila) 2009;2(7):650–657.

35. Son PS, Park SA, Na HK, Jue DM, Kim S, Surh YJ. Piceatannol,a catechol-type polyphenol, inhibits phorbol ester-induced NF-{kappa}B activation and cyclooxygenase-2 expression in hu-man breast epithelial cells: Cysteine 179 of IKK{beta} as a po-tential target. Carcinogenesis 2010;31(8):1442–1449.

36. Sottnik JL, Zhang J, Macoska JA, Keller ET. The PCa tumormicroenvironment. Cancer Microenviron 2011;4(3):283–297.

37. Rehemtulla A, Stegman LD, Cardozo SJ, Gupta S, Hall DE,Contag CH, Ross BD. Rapid and quantitative assessment ofcancer treatment response using in vivo bioluminescence im-aging. Neoplasia 2000;2(6):491–495.

38. Zhang C, Yan Z, Arango ME, Painter CL, Anderes K. Advanc-ing bioluminescence imaging technology for the evaluation ofanticancer agents in the MDA-MB-435-HAL-Luc mammary fatpad and subrenal capsule tumor models. Clin Cancer Res2009;15(1):238–246.

39. Kim JB, Urban K, Cochran E, Lee S, Ang A, Rice B, Bata A,Campbell K, Coffee R, Gorodinsky A, Lu Z, Zhou H, Kishi-moto TK, Lassota P. Non-invasive detection of a small numberof bioluminescent cancer cells in vivo. PLoS ONE2010;5(2):e9364.

40. Hsieh TC, Wong C, John Bennett D, Wu JM. Regulation of p53and cell proliferation by resveratrol and its derivatives inbreast cancer cells: An in silico and biochemical approach tar-geting integrin alphavbeta3. Int J Cancer 2011;129(11):2732–2743.

41. Gao G, Wang X, Qin X, Jiang X, Xiang D, Xie L, Hu J, Gao J.Effects of trimethoxystilbene on proliferation and apoptosis ofpulmonary artery smooth muscle cells. Cell Biochem Biophys2012;64(2):101–106.

42. Schneider Y, Chabert P, Stutzmann J, Coelho D, FougerousseA, Gosse F, Launay JF, Brouillard R, Raul F. Resveratrol analog(Z)-3,5,4’-trimethoxystilbene is a potent anti-mitotic druginhibiting tubulin polymerization. Int J Cancer 2003;107(2):189–196.

43. Paul S, Mizuno CS, Lee HJ, Zheng X, Chajkowisk S, RimoldiJM, Conney A, Suh N, Rimando AM. In vitro and in vivo stud-ies on stilbene analogs as potential treatment agents for coloncancer. Eur J Med Chem 2010;45(9):3702–3708.

44. Lin HS, Ho PC. A rapid HPLC method for the quantification of3,5,40-trimethoxy-trans-stilbene (TMS) in rat plasma and its ap-plication in pharmacokinetic study. J Pharm Biomed Anal2009;49(2):387–392.

45. Ma N, Liu WY, Li HD, Jiang XY, Zhang BK, Zhu RH, Wang F,Xie YL, Zhou XQ, Wu X, Xiang DX. RP-HPLC study of resvera-trol derivative (BTM-0512) in rat plasma and tissue distribu-tion. Yao Xue Xue Bao 2007;42(11):1183–1188.

46. Roupe KA, Yanez JA, Teng XW, Davies NM. Pharmacokineticsof selected stilbenes: Rhapontigenin, piceatannol and pinosyl-vin in rats. J Pharm Pharmacol 2006;58(11):1443–1450.

47. Boutegrabet L, Fekete A, Hertkorn N, Papastamoulis Y, Waffo-Teguo P, Merillon JM, Jeandet P, Gougeon RD, Schmitt-Kop-plin P. Determination of stilbene derivatives in Burgundy redwines by ultra-high-pressure liquid chromatography. AnalBioanal Chem 2011;401(5):1513–1521.

48. Lin LL, Lien CY, Cheng YC, Ku KL. An effective sample prepa-ration approach for screening the anticancer compound picea-tannol using HPLC coupled with UV and fluorescencedetection. J Chromatogr B Anal Technol Biomed Life Sci2007;853(1–2):175–182.

12 Diaset al.

The Prostate