Carcinogenesis 2003 Singh 555 63

Carcinogenesis vol.24 no.3 pp.555–563, 2003

Inositol hexaphosphate inhibits growth, and induces G1 arrest and apoptotic death ofprostate carcinoma DU145 cells: modulation of CDKI-CDK-cyclin and pRb-relatedprotein-E2F complexes

Rana P.Singh1, Chapla Agarwal1 and Rajesh Agarwal1–3

1Department of Pharmaceutical Sciences, School of Pharmacy and2University of Colorado Cancer Center, University of Colorado HealthSciences Center, Denver, CO 80262, USA3To whom correspondence should be addressedEmail: [email protected]

Cancer chemopreventive effects of inositol hexaphosphate(IP6), a dietary constituent, have been demonstrated againsta variety of experimental tumors, however, limited studieshave been done against prostate cancer (PCA), and molecu-lar mechanisms are not well defined. In the present study,we investigated the growth inhibitory effect and associatedmechanisms of IP6 in advanced human PCA cells.Advanced human prostate carcinoma DU145 cells wereused to study the anticancer effect of IP6. Flow cytometricanalysis was performed for cell cycle progression andapoptosis studies. Western immunoblotting, immunopre-cipitation and kinase assay were performed to investigatethe involvement of G1 cell cycle regulators and theirinterplay, and end point markers of apoptosis. A significantdose- as well as time-dependent growth inhibition wasobserved in IP6-treated cells, which was associated withan increase in G1 arrest. IP6 strongly increased the expres-sion of CDKIs (cyclin-dependent kinase inhibitors), Cip1/p21 and Kip1/p27, without any noticeable changes in G1CDKs and cyclins, except a slight increase in cyclin D2.IP6 inhibited kinase activities associated with CDK2, 4 and6, and cyclin E and D1. Further studies showed theincreased binding of Kip1/p27 and Cip1/p21 with cyclinD1 and E. In down-stream of CDKI–CDK/cyclin cascade,IP6 increased hypophosphorylated levels of Rb-relatedproteins, pRb/p107 and pRb2/p130, and moderatelydecreased E2F4 but increased its binding to both pRb/p107 and pRb2/p130. At higher doses and longer treatmenttimes, IP6 caused a marked increase in apoptosis, whichwas accompanied by increased levels of cleaved PARPand active caspase 3. IP6 modulates CDKI–CDK–cyclincomplex, and decreases CDK–cyclin kinase activity, pos-sibly leading to hypophosphorylation of Rb-related proteinsand an increased sequestration of E2F4. Higher doses ofIP6 could induce apoptosis and that might involve caspasesactivation. These molecular alterations provide an insightinto IP6-caused growth inhibition, G1 arrest and apoptoticdeath of human prostate carcinoma DU145 cells.

Introduction

Prostate cancer (PCA) is the major non-skin cancer in elderlymen and second most common male malignancy in the United

Abbreviations: CDKs, cyclin-dependent kinases; CDKIs, cyclin-dependentkinase inhibitors; IP6, inositol hexaphosphate; FACS, fluorescence-activatedcell sorting; PARP, poly (ADP-ribose) polymerase; PCA, prostate cancer; PI,propidium iodide; pRb/p107, retinoblastoma-related protein/p107; pRb2/p130,retinoblastoma-related protein 2/p130.

© All rights reserved. Oxford University Press 2003 555

States (1). Similar to other epithelial malignancies, autocrineand paracrine growth factor-receptor interactions and associ-ated mitogenic signalling are the major contributors to deregu-lated PCA cell growth (2,3). Since, in advanced stage, PCAgrowth and development become androgen independent thatrenders androgen ablation therapy ineffective (4), in thissituation, control of PCA through chemoprevention and inter-vention strategies is highly desirable. Cancer chemopreventionis the use of natural or pharmacologic agents to prevent thedevelopment of invasive malignancy by inhibiting early stagesof carcinogenesis and suppression or reversal of premalignantlesions (5,6). Several epidemiological and laboratory studieshave shown that many vegetables, fruits and grains as well asphytochemicals from non-dietary sources offer significantprotection against various cancers (3,5,6).

Inositol hexaphosphate (IP6) is a natural dietary ingredientand constitutes 0.4–6.4% (w/w) of most cereals, legumes, nuts,oil seeds and soybean (7). IP6 is abundant in cereals andlegumes, the consumption of which has been associated withreduction in mammary, colon and prostate cancer (7). Recentstudies have reported anti-cancer effects of IP6 in variouscancer models, both in vitro and in vivo (7–9). IP6 has beenshown to: (i) inhibit DMBA-induced rat mammary cancer andhuman breast cancer cell growth (9,10); (ii) suppress largeintestinal cancer in rats (8); (iii) inhibit growth and reversetransformed phenotype of HepG2 liver cancer cells (11), andregress liver cancer xenotransplant (12); (iv) prevent pulmonaryadenomas in mice (13); and (v) prevent skin tumorigenesis(14), inhibit growth of mouse fibrosarcoma FSA-1 tumorxenografts (15), and inhibit cell transformation in mouseepidermal JB6 cells (16). However, limited studies are donewith IP6 in PCA, and that molecular mechanisms of its anti-cancer effect are not well defined. Shamsuddin and Yang (17)showed that IP6 inhibits growth and induces differentiation ofhuman PCA PC-3 cells. Recently, we showed that IP6 impairserbB1 receptor and fluid-phase endocytosis, and associatedmitogenic signalling as its growth inhibitory effect in humanPCA DU145 cells (18).

The high expression of growth factor receptors (e.g. membersof erbB family) and associated ligands, and their interactionsare known to cause autocrine and paracrine loops for bothmitogenic and anti-apoptotic signalling leading to uncontrolledgrowth and metastasis of PCA (19–21). The activation of thesesignalling cascades alters the activation of several cell cycleregulators including cyclin-dependent kinases (CDKs), cyclins,CDK inhibitors (CDKIs), retinoblastoma protein (pRb) andpRb-related proteins such as pRb/p107 and pRb2/p130, sub-sequently resulting in activation of E2F family of transcriptionfactors commanding cell growth, proliferation, and cell survival(22–24). CDKIs are believed to play a critical role in cyclin-CDK-CDKI complex and negatively regulate cell cycle pro-gression from one stage to another (22). Furthermore, unphos-phorylated forms of pRb and pRb-related proteins have beenshown to inhibit cell proliferation by sequestering E2F familyof transcription factors (24).

by guest on June 15, 2013http://carcin.oxfordjournals.org/

Dow

nloaded from

http://carcin.oxfordjournals.org/

R.P.Singh, C.Agarwal and R.Agarwal

Fig. 1. Growth inhibitory effects of IP6 on DU145 cells. For the studiesassessing the effect of IP6 on exponentially growing DU145 cell growth,5000 cells/cm2 were plated in 60 mm dishes and next day, cells were fedwith fresh medium either with vehicle (distilled water) alone or 0.25–4 mMof IP6 in medium. After 24, 48 and 74 h of these treatments, total cellswere counted using hemocytometer. The cell growth data shown weremean � SE of three independent plates; each sample was counted induplicate. *P � 0.05, **P � 0.01, $P � 0.001 versus control.

Taken together, developing strategies that could halt cellcycle progression by targeting CDKIs or Rb-related proteinscould be useful for the prevention and/or intervention of theadvance stage of PCA. In an effort to develop IP6 as amolecular mechanism-based anticancer agent against PCA,here we focused our attention on the effect of IP6 on (a) cellgrowth and cell cycle progression; (b) cell cycle regulatorsand their interplay; and (c) CDKs and cyclin D1 associatedkinase activity. Since, in advance prostate carcinoma DU145cells, p53 and pRb genes are mutated and their proteins arenon-functional, we reasoned that CDKIs and pRb-relatedproteins could be the likely tumor suppressor targets for IP6in modulating the cell cycle progression in advanced PCAcells, and accordingly mechanistic aspects related to thesemolecules were explored in this investigation. At higher dosesand longer treatment time apoptotic effect of IP6 was alsoinvestigated.

Materials and methodsCell line and reagentsHuman prostate carcinoma cell line DU145 was obtained from AmericanType Culture Collection (Manassas, VA). Cells were cultured in RPMI 1640with 10% fetal bovine serum (Hyclone) under standard culture conditions(37°C, 95% humidified air and 5% CO2). IP6 was from Sigma-AldrichChemical Co. (St Louis, MO). RPMI 1640 and other culture materials werefrom Life Technologies, Inc. (Gaithersburg, MD). Anti-Cip1/p21 antibodywas from Calbiochem (Cambridge, MA), and anti-Kip1/p27 and cyclin D2antibodies were from Neomarkers (Fremont, CA). Antibodies to CDK2, 4and 6; cyclin D1, D3, E and A; pRb/p107, pRb2/p130, E2F3, 4 and 5, andRB-GST fusion protein were from Santa Cruz Biotechnology (Santa Cruz,CA). Histone H1 was from Boehringer Mannheim Corp. (Indianapolis, IN).[γ-32P] ATP (sp. act. 3000 Ci/mmol) and ECL detection system were fromAmersham Corp. (Arlington Heights, IL). Other chemicals were obtained intheir commercially available highest purity grade.

Cell culture and IP6 treatmentDU145 cells were grown in RPMI 1640 medium under standard cultureconditions. Cells were then treated with various doses (0.25–4 mM) of IP6(pH 7.4) dissolved in distilled water for different time points (24–72 h). Thedoses of IP6 selected in this study are those used by others and us in severalstudies (16–18). Following IP6 treatments, cell lysates were prepared in non-denaturing lysis buffer (10 mM Tris–HCl, pH 7.4, 150 mM NaCl, 1% TritonX-100, 1 mM EDTA, 1 mM EGTA, 0.3 mM phenyl methyl sulfonyl fluoride,0.2 mM sodium orthovanadate, 0.5% NP-40, 5 U/ml aprotonin). For lysate

556

preparation, medium was aspirated and cells were washed twice with ice-coldPBS followed by incubation in lysis buffer for 20 min. Then, cells werescrapped and kept on ice for an additional 30 min, and finally cell lysateswere cleared by centrifugation at 4°C for 30 min in a table-top centrifuge.Protein concentration in lysates was determined using Bio-Rad DC proteinassay kit (Bio-Rad laboratories, Hercules, CA) by the Lowry method.

Cell growth assay

DU145 cells were plated at 5000 cells/cm2 in 60 mm plates under the standardculture conditions detailed earlier and after 24 h, fed with fresh medium andtreated with different concentrations (0.25–4 mM) of IP6. After 12, 24, 48and 72 h, both floating and attached cells were collected by a brief trypsiniz-ation, and counted in duplicate with a hemocytometer. Each treatment andtime point had three independent plates. The representative data shown in thisstudy were reproducible in three independent experiments.

Flow cytometry analysis for cell cycle distribution

DU145 cells were grown in serum supplemented RPMI 1640 medium. At~30% confluence, cells were treated with 0.25–2 mM IP6 for 12–72 h and atthe end of each treatment time, cells were collected after a brief incubationwith trypsin-EDTA followed by processing for cell cycle analysis. Briefly,0.5 � 105cells were suspended in 0.5 ml of saponin/PI solution [0.3% saponin(w/v), 25 µg/ml PI (w/v), 0.1 mM EDTA and 10 µg/ml RNase (w/v) in PBS],and incubated overnight at 4°C in dark. Cell cycle distribution was thenanalyzed by flow cytometry using FACS core facility of University of ColoradoCancer Center. Finally, percentage of cells in different phases of cell cyclewas determined by ModFit LT cell cycle analysis software.

Western blotting

For cell cycle regulatory molecules CDKs, CDKIs and cyclins, pRb-relatedproteins and E2Fs, 40–80 mg of protein lysates per sample were denaturedin 2X SDS–PAGE sample buffer and subjected to SDS–PAGE on 8 or 12%Tris–glycine gel. The separated proteins were transferred on to nitrocellulosemembrane followed by blocking with 5% non-fat milk powder (w/v) in TBS(10 mM Tris, 100 mM NaCl, 0.1% Tween 20) for 1 h at room temperatureor overnight at 4°C. Membranes were probed for the protein levels of Cip1/p21, Kip1/p27, CDK2, CDK4, CDK6, cyclin D1, D2, D3, E and A, pRb/p107, pRb2/p130, E2F3, 4 and 5, PARP, caspase 3 and β-actin using specificprimary antibodies followed by peroxidase-conjugated appropriate secondaryantibody, and visualized by ECL detection system.

Cyclin-CDKI and pRb-related protein-E2F binding studies

For binding study, DU145 cells were treated with 2 mM IP6 for 24 h andwhole cell lysates were prepared as detailed above. 200 µg protein lysateswere cleared with protein A/G-plus agarose beads (Santa Cruz) for 45 min at4°C. Cyclin D1 and E were immunoprecipitated from protein lysates, usingspecific antibodies (2 µg) for 6 h incubation followed by addition of 25 µl ofprotein A/G-plus agarose beads and rocking overnight at 4°C. Non-specificantiserum was also used as a control to observe the non-specific immunoprecipi-tation in similar conditions. Immunoprecipitates were washed three times withlysis buffer, and samples were boiled in 2X sample buffer for 5 min followedby centrifugation. The resulting clear supernatants were subjected to SDS–PAGE on 12% gel. The separated proteins were electrophoretically transferredon to nitrocellulose membrane followed by blocking with 5% non-fat milkpowder (w/v) in TBS (10 mM Tris, 100 mM NaCl, 0.1% Tween 20) for 1 hat room temperature. Membrane was probed and visualized for Cip1/p21,Kip1/p27, cyclin D1 and E as detailed above. Similarly, pRb/p107and pRb2/p130 were immunoprecipitated from 400 mg protein lysates using specificantibodies and subjected to SDS-PAGE and western blotting, and membraneswere probed with E2F3, 4 and 5 antibodies and visualized by ECL detec-tion system.

Kinase assays

CDK2 and cyclin E-associated H1 histone kinase activity was determined asdescribed by us recently (25) with some modifications. Briefly, 200 µg ofprotein lysates from each sample was pre-cleared with protein A/G-plusagarose beads and CDK2 protein was immunoprecipitated using anti-CDK2and cyclin E antibodies (2 µg) and protein A/G-plus agarose beads. Beadswere washed three times with lysis buffer and finally once with kinase assaybuffer (50 mM Tris–HCl, pH 7.4, 10 mM MgCl2 and 1 mM DTT).Phosphorylation of histone H1 was measured by incubating the beads with30 µl of ‘hot’ kinase solution [0.25 µl (2.5 µg) of histone H1, 0.5 µl (5 mCi)of [γ-32P]ATP, 0.5 µl of 0.1 mM ATP and 28.75 µl of kinase buffer] for 30min at 37°C. The reaction was stopped by boiling the samples in SDS samplebuffer for 5 min. The samples were analyzed by 12% SDS–PAGE, and thegel was dried and subjected to autoradiography. Similarly, to determine theCDK4-, 6- and cyclin D1-associated Rb kinase activities, CDK4, 6 and cyclinD1 proteins were immunoprecipitated using specific antibodies, and beadsconjugated with antibody and proteins were washed three times with Rb-lysis


Dow

nloaded from


IP6 induces G1 arrest and apoptosis in PCA cells

Fig. 2. Effect of IP6 on cell cycle progression in DU145 cells. Cells were cultured in RPMI 1640 supplemented with 10% FBS, and treated with eithervehicle or 0.25–4 mM concentrations of IP6. After 24 h of these treatments, cells were collected, washed with PBS, digested with RNase and then cellularDNA stained with propidium iodide as detailed in ‘Materials and methods’. Flow cytometric analysis was then performed for cell cycle distribution. Thepercentage of cell distribution data for each treatment group shown is the mean of two plates and representative of three independent experiments.

buffer (50 mM HEPES–KOH, pH 7.5, 150 mM NaCl, 1 mM EDTA, 2.5 mMEGTA, 1 mM DTT, 80 mM β-glycerophosphate, 1 mM NaF, 0.1 mM sodiumorthovanadate, 0.1% Tween 20, 10% glycerol, 1 mM PMSF, and 10 µg/mlleupeptin and aprotonin) and twice with Rb-kinase assay buffer (50 mMHEPES–KOH, pH 7.5, 2.5 mM EGTA, 10 mM β-glycerophosphate, 1 mMNaF, 10 mM MgCl2, 0.1 mM sodium orthovanadate and 1 mM DTT).Phosphorylation of Rb was measured by incubating the beads with 30 µl of‘hot’ Rb-kinase solution (2 µg of Rb-GST fusion protein, 5 µCi of [γ-32P]ATP,0.1 mM ATP in Rb-kinase buffer) for 30 min at 37°C. Reaction was stoppedby boiling the samples in 5X SDS sample buffer for 5 min. Sampleswere analyzed by SDS–PAGE, and the gel was dried and subjected toautoradiography. Non-specific anti-serum control was also used to monitor ifthere were any non-specific immunoprecipitation in all these studies.

Apoptotic cell death assay

IP6 caused cell death at higher doses and longer treatment times and that wasanalyzed for apoptosis. To quantify IP6-induced apoptotic death of humanprostate carcinoma DU145 cells, annexin V and PI staining was performedfollowed by flow cytometry, as described recently (26). Briefly, after IP6treatment (1, 2 and 4 mM IP6 for 48 and 72 h), both floating and attachedcells were collected and subjected to annexin V and PI staining using VybrantApoptosis Assay Kit2 (Molecular Probes, Eugene, OR) and following thestep-by-step protocol provided by the manufacture. The kit contains recombin-ant annexin V conjugated to fluorophores and the Alexa fluoro 488 dye,providing maximum sensitivity. In apoptotic cells, annexin V binds tophosphatidylserine, which is translocated from inner to outer leaflet of theplasma membrane. The apoptotic cells stained with annexin V showed greenfluorescence, dead cells stained with both annexin V and PI showed red andgreen fluorescence, and live cells showed little or no fluorescence.

Statistical and densitometric analyses

The data were analyzed using the Jandel Scientific SigmaStat 2.0 software.For all measurements, one way ANOVA followed by Student’s t-test wasemployed to assess the statistical significance of difference between controland IP6-treated groups. A statistically significant difference was consideredto be present at P � 0.05. Autoradiograms/bands were scanned with AdobePhotoshop 6.0 (Adobe Systems, San Jose, CA), and adjusted for brightnessand contrast for minimum background. The mean density of each band wasanalyzed by the Scanimage program (NIH, Bethesda, MD). Densitometric

557

data presented below bands are fold change compared with control for eachtreatment time.

Results

IP6 inhibits growth of human prostate carcinoma DU145 cellsAs shown in Figure 1, IP6 treatment inhibited the growth ofhuman prostate carcinoma DU145 cells in a dose- and time-dependent manner. IP6 treatment at 0.25, 0.5, 1 and 2 mMdoses for 24–72 h resulted in 13–22 (P � 0.05), 22–29(P � 0.05), 31–40 (P � 0.01–0.001) and 35–54 (P � 0.01–0.001) % inhibition of cell growth, respectively (Figure 1).Much stronger cell growth inhibition (54–77%, P � 0.001)was observed at 4 mM dose of IP6 treatment for 24–72 h(Figure 1), and virtually the cell number did not increase after48 and 72 h following addition of treatment media to cultureplates. We reasoned that apoptosis might also be a contributorin IP6-caused growth inhibition of DU145 cells, which wasinvestigated later.

IP6 induces G1 arrest in DU145 cellsG1 arrest can prevent the replication of damaged DNA and,therefore, is helpful in checking the uncontrolled proliferationof cancer cells. Based on the growth inhibitory response ofIP6 in DU145 cells, we next examined its effect on cell cycleprogression. Consistent with its effect on cell growth inhibition,IP6 induced significant (P � 0.05–0.001) G1 arrest in DU145cells (Figure 2). IP6 treatment (0.25–4 mM IP6) for 24 hresulted in accumulation of 51–60% cells in G1 phase comparedwith control showing 38%. The observed increase in G1 cellpopulation was accompanied by a decrease in the number ofcells in both S phase as well as G2-M phase. In time kinetics,


Dow

nloaded from



Fig. 3. Effect of IP6 on cell cycle regulators in DU145 cells. Cells werecultured in RPMI 1640 medium with 10% FBS, and treated with 0.5, 1 and2 mM doses of IP6 for 24 and 48 h as described in ‘Materials andmethods’. At the end of treatments, total cell lysates were prepared andsubjected to SDS–PAGE followed by western immunoblotting. Membraneswere probed with anti-Cip1/p21, Kip1/p27, CDK2, 4, and 6, cyclin D1, D2,D3, E and A, and β-actin antibodies followed by peroxidase-conjugatedappropriate secondary antibodies, and visualized by ECL detection system.The experiments were repeated three times with similar results.

the maximum G1 arrest was observed at 24 h of IP6 treatment,which was sustained significantly (P � 0.001) up to 72 h (datanot shown). Induction of tumor suppressor genes is closelyassociated with the induction of G1 arrest that was investigatedin subsequent studies.

Effect of IP6 on G1 cell cycle regulatorBased on the data showing that IP6 induces strong G1 arrestin DU145 cells, we assessed the effect of IP6 on the cell cycleregulatory molecules that play important roles in G1 phase ofcell cycle progression. p53 is the most important tumorsuppressor protein associated with the G0–G1 arrest in cellcycle. Since DU145 cells do not have functional p53 protein,our focus was to assess the protein expression of CDKIs, theenhanced levels of which have been shown to be associatedwith G1 arrest. Western blot analysis showed that IP6 treatment(0.5, 1 and 2 mM IP6 for 24 and 48 h) of DU145 cells stronglyinduced the protein levels of Cip1/p21 and Kip1/p27 (Figure 3)

558

in a dose-dependent manner. In the time kinetics, maximuminducible effect of IP6 on CDKIs was evident at 24 h and thatwas also positively correlated to its effect on G1 arrest.Compared with controls, at the investigated doses and timepoints, 2 mM IP6 caused a maximum induction in the proteinlevels of both Cip1/p21 (by 2.5-fold) and Kip1/p27 (by 9-fold) after 24 h of treatment (Figure 3). The observed stronginduction in Cip1/p21 and Kip1/p27 protein levels by IP6 wasnot due to an overall change in protein levels as confirmed byprobing the same membranes with actin antibody (Figure 3).

Perturbations in cell cycle regulation have been demonstratedas one of the most common features in cancer cells. Thesealterations are generally associated with uncontrolled cellgrowth involving a lack of CDKI or loss of its function andincreased expression of CDKs and cyclins (27). Accordingly,we next assessed the effect of IP6 on the protein levels ofCDKs and cyclins involved in G1 phase and G1 to S phasetransition of cell cycle progression. As shown in Figure 3, IP6treatment of DU145 cells (0.5, 1 and 2 mM IP6 for 24 and48 h) did not show any noticeable change in the protein levelsof CDK2, 4 and 6, and cyclin D1, D3, E and A except a slightincrease in cyclin D2 after 24 h of treatment (Figure 3).

IP6 increases cyclin-CDKI binding and decreases kinaseactivities associated with CDK2, 4 and 6, and cyclin D1 and ESince up-regulation in CDKI level was observed maximallyat 2 mM IP6 treatment for 24 h, this was selected for CDK-CDKI binding and kinase studies. As shown in Figure 4A–C,IP6 showed an increased binding Cip1/p21 and Kip1/p27 withcyclin D1 and E. The bound level of Kip1/p27 with thesecyclins was strongly increased following IP6 treatment thatwas correlated with increased expression of these CDKIs.Densitometric analysis (Figure 4B,C) showed 1.4-fold increasein the bound protein level of Cip1/p21 whereas in the case ofKip1/p27 the increase was 6–10-fold compared with control.Since increased binding of CDKIs to CDK-cyclin complexsubsequently decreases kinase activity of CDK-cyclin complex,we next investigated the effect of IP6 on CDK2, 4 and 6, andcyclin D1 and E associated kinase activity. As shown inFigure 4D, IP6 treatment of DU145 cells at 2 mM dose for24 h strongly decreased the histone H1-associated kinaseactivities of CDK2 and cyclin E immunoprecipitates. Similarly,IP6 also decreased Rb-associated kinase activities of CDK4,and 6, and cyclin D1 immunoprecipitates (Figure 4D). In allthese studies, specificity of immunoprecipitation and kinaseassays was compared with antiserum control that did not showany such specific bands observed in assays (data not shown).Densitometric analysis showed that kinase activities wereinhibited by 50–80%, which might be due to up-regulation ofCDKIs and their increased binding with CDK-cyclin com-plexes. Together, these results suggest that growth inhibitoryeffect of IP6 in DU145 cells could be via an induction ofCDKIs leading to a decrease in kinase activity of CDKsfollowed by modulation of further down-stream targets suchas pRb-related proteins and E2F family of transcription factors.This suggestion, therefore, was next explored.

IP6 increases hypophosphorylated levels of pRb/p107 andpRb2/p130, and their binding with E2F4One of the down-stream targets of the cyclin kinases is thepRb-related proteins in the cells lacking functional pRb protein(28). Since IP6 showed a strong decrease in kinase activityassociated with CDK2, 4 and 6, and cyclin D1 and E, ournext focus was to investigate the phosphorylation status of


Dow

nloaded from



Fig. 4. Effect of IP6 on cyclin-CDKI complex, and CDK and cyclin-associated kinase activities. Cells were cultured in RPMI 1640 medium with 10% FBS,and treated with either vehicle or 2 mM doses of IP6 for 24 h as described in ‘Materials and methods’. At the end of treatments, total cell lysates wereprepared. (A) In binding study, cyclin D1 and E were immunoprecipitated using specific antibody from 200 mg of total protein lysates followed by SDS–PAGE and western immunoblotting. Membranes were probed for Cip1/p21, Kip1/p27, cyclin D1 and E. (B and C) Densitometric data of bound levels ofCip1/p21 and Kip1/p27 with cyclin D1 and E, presented as fold change of control after IP correction. (D) Effect of IP6 on CDK2, 4 and 6, and cyclin D1and E-associated kinase activities. Treatments and cell culture conditions were the same as those in binding studies. CDK2 and cyclin E kinase activities weredetermined by in-bead histone H1 kinase assay using immunoprecipitated CDK2 and cyclin E from total cell lysates (200 µg protein) using specific antibody;and CDK4, 6 and cyclin D1-associated kinase activities were determined by in-bead RB–GST fusion protein kinase assay using immunoprecipitated CDK4, 6and cyclin D1 from total cell lysates (200 µg protein) using specific antibodies as described in ‘Materials and methods’. After the assay, the labelledsubstrates were subjected to SDS–PAGE, and the gel was dried and exposed to X-ray film. The bottom panel of each band indicates the fold change in bandintensity compared with that of control at each treatment time. The experiments were repeated 2–3 times with similar results. IB, western immunoblotting; IP,immunoprecipitation.

559


Dow

nloaded from



Fig. 5. Effect of IP6 on Rb-related proteins, E2F transcription factor andRb-related proteins-F2F complex in DU145 cells. (A) Cells were cultured inRPMI 1640 medium with 10% FBS, and treated with 0.5, 1 and 2 mMdoses of IP6 for 24 h as described in ‘Materials and methods’. At the endof treatments, total cell lysates were prepared in Rb-lysis buffer andsubjected to SDS–PAGE (8 or 12% gel) followed by westernimmunoblotting as detailed in ‘Materials and methods’. Membranes wereprobed with anti-pRb/p107, pRb2/p130, E2F4 and β-actin antibodiesfollowed by peroxidase-conjugated appropriate secondary antibodies, andvisualized by ECL detection system. (B) In binding studies, only 2 mMdose of IP6 for 24 h was used as described in ‘Materials and methods’.pRb/p107 and pRb2/p130 were immunoprecipitated using specific antibodiesfrom 400 µg of total protein lysates followed by SDS–PAGE and westernblotting. In pRb/p107 IP, membrane was probed for E2F4 and pRb/p107protein levels and in pRb2/p130 IP, membrane was probed for E2F4 andpRb/p130; and (C) bound level of E2F4 was extrapolated by densitometricanalysis after IP correction for both Rb-related proteins. Experiments wererepeated twice with similar results.

560

Fig. 6. Apoptotic effect of IP6 on DU145 cells. Cells were grown in 10%FBS condition and treated with 1–4 mM IP6 for 48 and 72 h. (A) At theend of treatments, total cells were collected and stained with annexin V/PIas mentioned in ‘Materials and methods’ followed by flow cytometricanalysis. Data presented are mean � SE of percent annexin V/PI stainedcells of two independent samples in each treatment group. (B) In similartreatments as detailed above, total cells were collected and cell lysates wereprepared as described in ‘Materials and methods’. SDS–PAGE and westernblot analysis were performed for PARP, caspase 3 and β-actin using specificantibodies as described in ‘Materials and methods’. The experiments wererepeated with similar results. *P � 0.05.

pRb-related proteins pRb/p107 and pRb2/p130. IP6 treatmentof DU145 cells at 0.5–2.0 mM doses for 24 h, showed increasein hypophosphorylated levels of pRb/p107 and pRb2/p130 ina dose-dependent manner (Figure 5A). Next, we examined theprotein levels of E2F3, 4 and 5, as E2Fs are the targettranscription factors in this pathway that regulate the transcrip-tion of growth responsive genes. Similar IP6 treatment ofDU145 cells did not show any observable change in E2F3 and5 protein levels (data not shown), but moderately decreasedthe level of E2F4 (Figure 5A). Protein loading was checkedby re-probing of the same membrane with actin antibody(Figure 5A). Since, oncogenic activity of E2Fs is dependenton their free/bound ratio with pRb-related proteins in pRblacking cells, next we determined the bound levels of E2F3,4, and 5 with pRb/p107 and pRb2/p130 after IP6 treatment.As shown in Figure 5B, IP6 showed an increased binding ofE2F4 with pRb/p107 and pRb2/p130, which accounted foralmost 2–2.5 fold increase as compared with that of control(Figure 5C). There was no noticeable change in the boundlevels of E2F3 and 5 with pRb/p107 and pRb2/p130 (data notshown). The bound level of E2F4 with pRb2/p130 was moreprominent compared with pRb/p107. Since pRb2/p130-E2F4has been reported as the most abundant complex in the restingor quiescent cells in G0 (29), the observed effect of IP6 onan induction of this complex level could be of greater signific-ance in defining a mechanistic rationale involved in IP6-causedG1 arrest and cell growth inhibition, and possibly also inapoptotic death which was investigated next.

IP6 causes apoptotic death of DU145 cellsWe investigated any possible involvement of apoptosis induc-tion by IP6 that might have contributed in its strong growth


Dow

nloaded from



inhibitory effect at higher doses and longer treatment times.A quantitative apoptotic cell death study was performed toanswer this question. IP6-caused apoptotic death of DU145cells was analyzed by annexin V/PI staining which showedthat higher dose (4 mM) treatment of IP6 for 48–72 hsignificantly (P � 0.05) increased the percentage of apoptoticcells up to 3–4 fold accounting for 9–14% compared with thatof control showing 3–4% (Figure 6A). During same treatmenttime the lower doses (1 and 2 mM) of IP6 could maximallyinduce up to 1.5 fold apoptotic cell death (Figure 6A). Tofurther confirm the apoptotic effect of IP6, western blot analysiswas performed to analyze PARP cleavage and caspase 3 (activefragment). As shown in Figure 6B, IP6 increased the levelsof both cleaved PARP and caspase 3, and therefore suggestedthe possible involvement of caspases activation as one of thepossible mechanisms of apoptosis induction, which is yet tobe investigated.

Discussion

The major finding of the present study is that IP6 stronglyinduces Cip1/p21 and Kip1/p27 and inhibits CDK2, 4, and 6,and cyclin D1 and E-associated kinase activities followed byan increase in hypophosphorylated levels of pRb/p107 andpRb2/p130, and their increased binding with E2F4. Thesemolecular effects of IP6 could be one of the possible underlyingmechanisms that resulted in inhibition of cell growth and G1arrest in DU145 cell cycle progression. Furthermore, IP6causes apoptotic cell death in human prostate carcinoma cellsthat was evidenced by the cleavage of PARP and caspase 3.

The results obtained in the present study provide convincingevidence that IP6 exerts its effects on cell cycle progressionmainly via an up-regulation of Cip1/p21 and Kip1/p27. Whencell cycle phase distributions are compared with alterations incell cycle regulatory molecules, a strong increase in CDKIscan be attributed as one of the major causes of IP6-inducedG1 arrest and cell growth inhibition. This finding is consistentwith earlier reports in which IP6 has been shown to induceCip1/p21 and G1 arrest in breast cancer cells (30,31). However,it should also be noted that IP6 induces G2-M arrest inleukemia cell lines by down-regulation of CSK2 (cdc 28 proteinkinase 2) and Kip2/p57, suggesting a different mechanism ofcell cycle arrest by IP6 and that could be most likely due todissimilar origin of these cells (32). The mammalian cell cycleis regulated by complex machinery, in which CDKs, CDKIs andcyclins play essential roles (27). CDKIs are tumor suppressorproteins that down regulate the cell cycle progression bybinding with active CDK-cyclin complexes and thereby inhibit-ing their kinase activities (33,34). The important CDKIs includeCip1/p21, a universal inhibitor of CDKs whose expression ismainly regulated by the p53 tumor suppressor protein (35);and Kip1/p27 that is also up regulated in response to antiprolif-erative signals (36). The increased expression of G1 cyclinsin cancer cells provide them an uncontrolled growth advantagebecause most of these cells either lack CDKI or possess non-functional CDKI or have low expression of CDKI (37). Theincreased expression of CDKIs by IP6 is both encouragingand important in the sense that decreased Kip1/p27 expressionin prostatic carcinomas has been associated with aggressivephenotype, and loss of Cip1/p21 function has been implicatedin the failure of irradiation response (38). Additionally, ourdata suggest that IP6-caused CDKIs up-regulation involvesp53-independent pathway, as DU145 cells lack functional p53.

561

The pRb-related proteins, pRb/p107 and pRb2/p130cooperate to regulate cell cycle progression through G1 phase ofcell cycle with a common goal of G1-S checkpoint regulation. Ithas been reported that pRb2/p130 and Kip1/p27 are mutuallyinvolved in a negative regulation of cellular proliferation (24).The Rb family members are nuclear phospho-proteins and areregulated in a cell cycle-dependent manner by phosphorylation,and are critical targets for inactivation by transforming oncopro-teins (24,28). pRb/p107 and pRb2/p130 show considerablehomology with pRb and classified as a sub-family of pRbfamily members. These proteins bind to and modulate theactivity of the E2F family of transcription factors that inducethe transcription of genes needed for the cell cycle progressionthrough S phase (28). The pRb2/p130-E2F4 is the mostabundant E2F complex found in a resting or quiescent cellsin G0, and is suggested to help in maintaining a state oftranscriptional silence (29). Consistent with these reports, IP6-caused a decrease in CDK-cyclin kinase activity, and increasedhypophosphorylated levels of Rb-related proteins and theirincreased binding with E2Fs may be attributed as decreasedexpression of growth responsive genes and subsequent growthinhibition of PCA cells by IP6. E2Fs have been identified asdown-stream targets for the action of the tumor suppressorproteins p53 and pRb (39). An increase in E2F transcriptionalactivity as a consequence of loss of p53 and pRb, that is thecase in DU145 cells, is considered to be a critical factor inderegulated cell cycle progression, a common phenomenon intumorigenesis (29). The Rb-related proteins and associatedfactors are suggested to be critical targets in cancer therapy.Accordingly, the observed biological effects of IP6 in DU145cells and their associated mechanisms are encouraging, andneed more detailed study to justify the modulation of thesemolecular events towards anticancer effect of IP6 againsthuman PCA.

It is well established that apoptosis and, the associatedsignalling pathways and cellular events controlling it have aprofound effect on the progression of benign to malignantphenotype and can be targeted for the therapy of variousmalignancies including PCA (40). Accordingly, our resultsshowing induction of apoptotic death of DU145 cells by IP6could be of greater significance in identifying another (togetherwith growth inhibitory) anticancer effect of IP6 in PCA. Sinceinduction of CDKIs has been reported in anticancer agent-induced apoptosis in human PCA cells (41), the IP6-causedincrease in CDKIs could be in part responsible for the observedapoptotic death of DU145 cells. PARP cleavage and increasein the level of active caspase 3 suggested that caspasesactivation could have an important role in IP6-inducedapoptosis induction. However, the detailed studies are neededto describe IP6-induced apoptotic death of DU145 cells.

The IP6 doses (0.25–4 mM) used in the present study werebased on earlier published studies, which in most cases usedup to 5 mM IP6 concentration in cell culture treatments (16–18,30,31). The physiological and/or pharmacological signific-ance of these levels of IP6 in cell culture studies needs to beestablished in future in both animal studies as well as inhumans. In this regard, it is important to re-emphasize thatendogenous mammalian cellular concentration of IP6 rangesfrom 0.01–1 mM (7). However, most effects reported in thepresent study are at 2 mM dose in cell cycle studies and areat 4 mM dose in both cell cycle and apoptosis studies.Similarly, other biological effects of IP6 reported in literatureare also at 2–5 mM doses in cell culture (16–18,30,31).


Dow

nloaded from



Comparing these concentrations, it is obvious that the endogen-ous cellular/physiological level of IP6 is not sufficient for itsanti-cancer activity at least in cell culture, and that higherlevels of IP6 are needed for its pharmacological efficacy. Whenthese cell culture concentrations of IP6 used in the presentstudy and in earlier reports (with only one treatment) werecompared with published literature showing its anti-canceractivity in animal models by using up to 15 mM dose incontinuous feeding in drinking water (reviewed in ref. 10),concentrations in cell culture studies are much lower, whichclearly establish their relevance in defining molecular mechan-ism of efficacy of IP6 at pharmacological concentrationsrelated to pre-clinical studies. The clinical efficacy of IP6against any cancer is an open area, and has to be investigatedand established in future. However, a recent study on oralabsorption of IP6 in humans has reported that IP6 concentrationin plasma could go up by 3–5-fold (~0.3 mg/l) by ingestionof IP6-normal diet when compared with the ingestion of IP6-poor diet (42). This study also reported that humans becomedeficient in IP6 if they consume IP6-poor diet for as little as2 weeks, and that the normal level of IP6 could be easilyachieved by IP6 dietary supplement. These findings suggestthat a dose-escalation study with IP6 is needed to assess thepharmacologically achievable concentrations of IP6 as well asassociated toxicity, if any.

Several studies suggest that androgen-independent prostatetumors contain an altered form of androgen receptor, evolvingfrom gene rearrangement, point mutation or deletion, leadingto the presence of a receptor that remains (constitutively)active even in the absence of androgen or presence of anti-androgen (43). This hormone refractory PCA phenotype isassociated with several changes at molecular level includingmutations of androgen receptor protein, alteration in mitogenicand cell survival signalling, inactivation of p53 and loss ofpRb (44). There is no effective standard chemotherapy for thepatients diagnosed with hormone-refractory PCA, which oftenresults from androgen ablation therapy that invariably fails toprevent the progressive metastatic disease. In such patients,medial survival time is only 6–9 months (45). One possiblepractical and translational approach in controlling the growthand metastatic potential of PCA in those patients as well asincreasing the survival time and quality of life could be throughdietary agents. In this regard, IP6 has already shown promisingresults in several animal tumor studies as well as epidemio-logical reports related to mammary, colon and prostate cancers.Based on these reports, recent studies by us, and the data ofthe present work, a recommendation could be made that furtherstudies are needed to establish the relevance of observedin vitro anticancer effects of IP6 under in vivo pre-clinicalprostate cancer models, and may be helpful in developing IP6as an effective cancer preventive agent against advancedhuman PCA.

Acknowledgement

This work was supported by USPHS grant CA83741 from the National CancerInstitute, NIH.

References

1.Greenlee,R.T., Murray,T., Bolden,S. and Wingo,P.A. (2000) Cancerstatistics. CA Cancer J. Clin., 50, 7–33.

2.Gioeli,D., Mandell,J.W., Petroni,G.R., Frierson Jr,H.F. and Weber,M.J.(1999) Activation of mitogen-activated protein kinase associated withprostate cancer progression. Cancer Res., 59, 279–284.

562

3.Agarwal,R. (2000) Cell signalling and regulators of cell cycle as moleculartargets for prostate cancer prevention by dietary agents. Biochem.Pharmacol., 60, 1051–1059.

4.Aquilina,J.W., Lipsky,J.J. and Bostwick,D.G. (1999) Androgen deprivationas a strategy for prostate cancer chemoprevention. J. Natl Cancer Inst.,89, 689–696.

5.Hong,W.K. and Sporn,M.B. (1997) Recent advances in chemopreventionof cancer. Science, 278, 1073–1077.

6.Kellof,G.J., Lieberman,R., Steele,V.E, Boone,C.W, Lubet,R.A.,Kopelovitch,L., Malone,W.A., Crowell,J.A. and Sigman,C.C. (1999)Chemoprevention of prostate cancer: concepts and strategies. Eur. Urol.,35, 342–350.

7.Shamsuddin,A.M., Vucenik,I. and Cole,K.E. (1997) IP6: a novel anticanceragent. Life Sci., 61, 343–354.

8.Shamsuddin,A.M., Elsayed,A.M. and Ullah,A. (1988) Suppression of largeintestinal cancer in F344 rats by inositol hexaphosphate. Carcinogenesis,9, 577–580.

9.Vucenik,I., Yang,G.Y. and Shamsuddin,A.M. (1997) Comparison of pureinositol hexaphosphate and high-bran diet in the prevention of DMBA-induced rat mammary carcinogenesis. Nutr. Cancer, 28, 7–13.

10.Shamsuddin,A.M. and Vucenik,I. (1999) Mammary tumor inhibition byIP6: a review. Anticancer Res., 19, 3671–3674.

11.Vucenik,I., Tantivejkul,K., Zhang,J.S., Cole,K.E., Saied,I. andShamsuddin,A.M. (1998) IP6 treatment of liver cancer. I: IP6 inhibitsgrowth reverses transformed phenotype in HepG2 human liver cancer cellline. Anticancer Res., 18, 4083–4090.

12.Vucenik,I., Zhang,J.S. and Shamsuddin,A.M. (1998) IP6 treatment of livercancer. II: intratumoral injection of IP6 regresses pre-existing human livercancer xenotransplanted in nude mice. Anticancer Res., 18, 4091–4096.

13.Wattenberg,L.W. (1999) Chemoprevention of pulmonary carcinogenesisby myo-inositol. Anticancer Res., 19, 3659–3661.

14. Ishikawa,T., Nakatsuru,Y., Zarkovic,M. and Shamsuddin,A.M. (1999)Inhibition of skin cancer by IP6 in vivo: initiation-promotion model.Anticancer Res., 19, 3749–3752.

15.Vucenik,I., Tomazik,V.J., Fabian,D. and Shamsuddin,A.M. (1992) Anti-tumor activity of phytic acid (inositol hexaphosphate) in murinetransplanted and metastatic fibrosarcoma, pilot study. Cancer Lett., 65,9–13.

16.Huang,C., Ma,W.I., Hecht,S.S. and Dong,Z. (1997) Inositol hexaphosphateinhibits cell transformation and activator protein 1 activation by targetingphosphatidylinositol-3 kinase. Cancer Res., 57, 2873–2878.

17.Shamsuddin,A.M. and Yang,G.Y. (1995) Inositol hexaphosphate inhibitsgrowth and induces differentiation of PC-3 human prostate cancer cells.Carcinogenesis, 16, 1975–1979.

18.Zi,X., Singh,R.P. and Agarwal,R. (2000) Impairment of erbB1 receptorand fluid phase endocytosis and associated mitogenic signalling by inositolhexaphosphate in human prostate carcinoma DU145 cells. Carcinogenesis,21, 2225–2235.

19.Hofer,D.R., Sherwood,E.R., Bromberg,W.D., Mendelsohn,J., Lee,C. andKozlowski,J.M. (1991) Autonomous growth of androgen-independenthuman prostatic carcinoma cells: role of transforming growth factor alpha.Cancer Res., 51, 2780–2785.

20.Fong,C.J, Sherwood,E.R., Mendelsohn,J., Lee,C. and Kozlowski,J.M.(1992) Epidermal growth factor receptor monoclonal antibody inhibitsconstitutive receptor phosphorylation, reduces autonomous growth, andsensitizes androgen-independent prostatic carcinoma cells to tumor necrosisfactor alpha. Cancer Res., 52, 5887–5892.

21.Bostwick,D.G. and Aquilina,J.W. (1996) Prostatic intra-epithelial neoplasia(PIN) and other prostatic lesions as risk factors and surrogate endpointsfor cancer chemoprevention trials. J. Cell Biochem., 25S, 156–164.

22.Grana,X. and Reddy,P. (1995) Cell cycle control in mammalian cells: roleof cyclins, cyclin-dependent kinases (CDKs), growth suppressor genesand cyclin-dependent kinase inhibitors (CDKIs). Oncogene, 11, 211–219.

23.Morgan,D.O. (1995) Principles of CDK regulation. Nature, 374, 131–134.24.Howard,C.M., Claudio,P.P., Luca,A.D., Stiegler,P., Jori,F.P., Safdar,N.M.,

Caputi,M., Khalili,K. and Giordano,A. (2000) Inducible pRb2/p130expression and growth-suppressive mechanisms: evidence of pRb2/p130,p27kip1, and cyclin E negative feedback regulatory loop. Cancer Res.,60, 2737–2744.

25.Zi,X., Grasso,A.W., Kung,H. and Agarwal,R. (1998) A flavonoidantioxidant, silymarin, inhibits activation of erbB1 signalling and inducescyclin-dependent kinase inhibitors, G1 arrest, and anticarcinogenic effectsin human prostate carcinoma DU145 cells. Cancer Res., 58, 1920–1929.

26.Agarwal,C., Sharma,Y. and Agarwal,R. (2000) Anticarcinogenic effect ofa polyphenolic fraction isolated from grape seeds in human prostatecarcinoma DU145 cells: modulation of mitogenic signalling and cell cycle


Dow

nloaded from



regulators and induction of G1 arrest and apoptosis. Mol. Carcinog., 28,129–138.

27.Hunter,T. and Pine,J. (1994) Cyclins and cancer II: cyclin D and CDKinhibitors come of age. Cell, 79, 573–582.

28.Paggi,M.G., Baldi,A., Bonetto,F. and Giordano,A. (1996) Retinoblastomaprotein family in cell cycle and cancer: a review. J. Cell Biochem., 62,418–430.

29.Moberg,K., Starz,M.A. and Lees,J.A. (1996) E2F4 switches from p130 top107 and pRb in response to cell cycle reentry. Mol. Cell Biol., 16,1436–1449.

30.Shamsuddin,A.M. (1999) Metabolism and cellular functions of IP6: areview. Anticancer Res., 5, 3733–3736.

31.El-Sherbiny,Y.M., Cox,M.C., Ismail,Z.A., Shamsuddin,A.M. and Vucenik,I.(2001) G0/G1 arrest and S phase inhibition of human cancer cell lines byinositol hexaphosphate (IP6). Anticancer Res., 4, 2393–2403.

32.Deliliers,G.L., Servida,F., Fracchiolla,N.S., Ricci,C., Borsotti,C.,Colombo,G. and Soligo,D. (2002) Effect of inositol hexaphosphate (IP6)on human normal and leukaemic haematopoietic cells. Br. J. Haematol.,117, 577–587.

33.Toyoshima,H. and Hunter,T. (1994) P27, a novel inhibitor of G1 cyclin-cdk protein kinase activity, is related to p21. Cell, 78, 67–74.

34.Tsai,L.H., Lees,E., Faha,B., Harlow,E. and Riabowol,K. (1993) The CDK2kinase is required for the G1 to S transition in mammalian cells. Oncogene,8, 1593–1602.

35.Xiong,Y., Hannon,G.J., Zhang,H., Casso,D., Kobayashi,R. and Beach,D.(1993) P21 is a universal inhibitor of cyclin kinases. Nature, 366, 701–704.

36.Polyak,K., Lee,M.H., Erdjument-Bromage,H., Koff,A., Roberts,J.M.,Tempst,P. and Massague,J. (1994) Cloning of p27/kip1, cyclin-dependentkinase inhibitor and a potential mediator of extracellular antimitogenicsignals. Cell, 78, 59–66.

563

37.Dulic,V., Lees,E. and Reed,S.I. (1992) Association of human cyclin E witha periodic G1-S phase protein kinase. Science, 257, 1958–1961.

38.Cheng,L., Lloyd,R.V., Weaver,A.L., et al. (2000) The cell cycle inhibitorsp21/waf1 and p27/kip1 are associated with survival in patients treated bysalvage prostatectomy after radiation therapy. Clin. Cancer Res., 6,1896–1899.

39.Slansky,J.E. and Farnham,P.J. (1996) Introduction to the E2F family:protein structure and gene regulation. Curr. Top. Microbiol. Immunol.,208, 1–30.

40.Lowe,S.W. and Lin,A.W. (2000) Apoptosis in cancer. Carcinogenesis, 21,485–495.

41.Don,M.J., Chang,Y.H., Chen,K.K., Ho,L.K. and Chau,Y.P. (2001) Inductionof CDK inhibitors (p21/waf1 and p27/kip1) and Bak in the β-lapachone-induced apoptosis of human prostate cancer cells. Mol. Pharmacol., 59,784–794.

42.Grases,F., Simonet,B.M., Vucenik,I., Prieto,I.M., Costa-Bauza,A.,March,J.G. and Shamsuddin,A.M. (2001) Absorption and excretion oforally administered inositol hexaphosphate (IP6 or phytate) in humans.BioFactors, 15, 53–61.

43.Koivisto,P., Kolmer,M., Visakorpi,T. and Kallioniemi,O.P. (1998) Androgenreceptor gene and hormonal therapy failure of prostate cancer. Am. J.Pathol., 152, 1–9.

44.Pilat,M.J., Kamradt,J.M. and Pienta,K.J. (1999) Hormone resistance inprostate cancer. Cancer Metastasis, 17, 373–381.

45.Kantoff,P.W. (1995) New agents in the therapy of hormone-refractoryprostate cancer. Semin. Oncol., 22, 32–34.

Received October 24, 2002; revised December 2, 2002;accepted December 10, 2002


Dow

nloaded from


Carcinogenesis 2003 Singh 555 63

Documents

Transcript of Carcinogenesis 2003 Singh 555 63