Activation of PI3K-Akt signaling pathway promotes prostate cancer cell invasion

Sanjeev Shukla1, Gregory T. MacLennan

2,3, Douglas J. Hartman

2, Pingfu Fu

3,4, Martin I. Resnick

1,3and Sanjay Gupta

1,3*

1Department of Urology, Case Western Reserve University and University Hospitals Case Medical Center, Cleveland, OH2Department of Pathology, Case Western Reserve University and University Hospitals Case Medical Center, Cleveland, OH3Case Comprehensive Cancer Center, Cleveland, OH4Department of Epidemiology and Biostatistics, Case Western Reserve University and University HospitalsCase Medical Center, Cleveland, OH

Activated phosphoinositide 3-kinase (PI3K) and its downstreamtarget Akt/PKB are important signaling molecules and key sur-vival factors involved in the control of cell proliferation, apoptosisand oncogenesis. We investigated the role of the PI3K-Akt signal-ing pathway in the invasion of prostate cancer cell lines and acti-vation of this pathway in primary human prostate tumors. Treat-ment of human prostate cancer cells viz. LNCaP, PC-3 and DU145with PI3K pharmacological inhibitor, LY294002, potentially sup-pressed the invasive properties in each of these cell lines. Restora-tion of the PTEN gene to highly invasive prostate cancer PC-3 cellsor expression of a dominant negative version of the PI3K target,Akt also significantly inhibited invasion and downregulated pro-tein expression of urokinase-type plasminogen activator (uPA)and matrix metalloproteinase (MMP)-9, markers for cell invasion,indicating a central role of the PI3K-Akt pathway in this process.Immunoblot analysis of PI3K and total/activated levels of Aktshowed increased protein levels of catalytic (p110a/b) and regula-tory (p85) subunits of PI3K and constitutive Akt activation inhigh-grade tumors compared to low-grade tumor and benign tis-sue. Immunohistochemical analyses further confirmed a progres-sive increase in p-Akt (p-Ser473) levels but not of total-Akt (Akt1/2) in cancer tissues compared to benign specimens. A successiveincrease in p-Akt expression was further noted in specimens seri-ally obtained from individuals with time-course disease progres-sion. Taken together, these results suggest that aberrant activationof PI3K-Akt pathway may contribute to increased cell invasive-ness and facilitate prostate cancer progression.' 2007 Wiley-Liss, Inc.

Key words: Akt/protein kinase B; PI3K; prostate cancer; PTEN; cellinvasion; cell survival

During prostate cancer progression, tumor cells undergo a vari-ety of molecular alterations that lead to the acquisition of uncon-trolled growth properties.1 One such set of molecular alterationsmay be mediated by the PI3K-Akt signaling pathway.2,3 Phospho-inositide 3-kinase (PI3K) is a heterodimeric protein composed of acatalytic subunit (p110a/b/g/d) and a regulatory subunit (p85a/b)that participate in multiple cellular processes, including cellgrowth, transformation, migration and differentiation.4 The PI3Kpathway has been shown to be an essential survival mechanism ina number of cell types and in some forms of human cancer.2 Thispathway is upregulated by several different mechanisms. Amplifi-cation of the gene coding for the catalytic subunit of PI3K hasbeen observed in cervical and ovarian cancers.5,6 Following acti-vation of cells by growth factors or cytokines, PI3K is recruited tothe plasma membrane, where it catalyzes the conversion of mem-brane phosphoinositide 4,5-biphosphate (PIP2) in the D3 positionto generate phosphoinositide 3,4,5-triphosphate (PIP3). The accu-mulation of PIP3 creates a docking site for Akt at the plasmamembrane, which binds to PIP3 via the pleckstrin homology do-main. PIP3 binding induces a conformational change in Akt,exposing the critical Thr308/309 residue in the activation loopto phosphorylation by phosphotidylinositol-dependent kinase 1(PDK-1). For full length activation, Akt is subsequently phospho-rylated at Ser473/474 by an as yet unidentified kinase termedPDK-2.7 Activation of PI3K leads to the generation of PIP3 whichcan be counterbalanced by the action of PTEN/MMAC1/TEP1, alipid phosphatase and tumor suppressor that dephosphorylates

PIP3 back to PIP2, controlling the activation of Akt.2,7 PTEN isfrequently lost in glioblastoma, breast cancer, endometrial cancer,melanoma and prostate cancer.8–11

Akt/protein kinase B/RAC-PK is an essential serine/threoninekinase and a core component of the PI3K signaling pathwaywhose activation has been implicated in the genesis or progressionof various human malignancies.2,7 The genes of PI3K targets,AKT1 and AKT2, are amplified and over-expressed in breast, gas-tric and ovarian cancers.12,13 Akt3 activity is often increased inprostate and breast cancer.14 In experimental systems, constitu-tively active PI3K or Akt is oncogenic in cell culture systems andanimal tumor models.15,16 Several studies have shown that Akt/PKB activates the transcription of a wide variety of genes, espe-cially those involved in immune activation, cell proliferation, apo-ptosis and cell survival.2,7,17 Activated Akt protects cells from ap-optotic death by phosphorylating substrates such as BAD, procas-pase-9, NF-jB and fork-head transcription family members.17 Aktactivation affects cell cycle progression, through regulation ofcyclin D stability and inhibition of p27/Kip1and p21/WAF1 pro-tein levels.18 Akt has been shown to prolong cell survival bydelaying p53-dependent apoptosis through MDM2 phosphoryla-tion.19 Akt has also been shown to inhibit the Raf-MEK-ERKpathway through phosphorylation of Raf-1 and to overcome con-stitutively activated MAPK-induced cell cycle arrest.20 It has beenproposed that Akt regulates permeability transition pore openingswithin the mitochondrial membrane by increasing the coupling ofglucose metabolism to oxidative phosphorylation.21 Based on itsrole as a key regulator of cell survival, Akt has emerged as an im-portant factor in tumorigenesis.22

Several studies have shown that various growth factors, cyto-kines and oncogenes, exert their effects via the PI3K signalingpathway, which, in turn, leads to Akt activation.23,24 IncreasedAkt activity facilitates prostate cancer progression through down-regulation of the cyclin-dependent kinase inhibitor, p27/Kip1.25

Akt has been shown to suppress androgen-induced apoptosis byphosphorylation and inhibition of androgen receptor.26 Condi-tional activation of Akt has been shown to promote androgen-independent progression and is essential for neuroendocrine differ-entiation of prostate cancer.27,28 Development of hormone-insen-sitivity in patients who have been on long-term androgen ablationtherapy for prostate cancer is associated with reinforcement of the

Grant sponsor: United States Public Health Services; Grant numbers:RO1 CA108512, RO3 CA107806, RO3 CA094248, RO3 CA099049;Grant sponsor: Cancer Research and Prevention Foundation.*Correspondence to: Department of Urology, The James and Eilleen

Dicke Research Laboratory, Case Western Reserve University and Univer-sity Hospitals Case Medical Center, 10900 Euclid Avenue, Cleveland, OH44106, USA. Fax:1216-368-0213. E-mail: sanjay.gupta@case.eduReceived 27 December 2006; Accepted after revision 30 March 2007DOI 10.1002/ijc.22862Published online 5 June 2007 in Wiley InterScience (www.interscience.

wiley.com).

Abbreviations: BSA, bovine serum albumin; NF- jB, nuclear factor-kappaB; NHPE, normal human prostate epithelial cells; PDK, phosphoti-dylinositol-dependent kinase; PI3K, phosphatidylinositol 30 kinase; PIP2,phosphoinositide 4,5-biphosphate; PIP3, phosphoinositide 3,4,5-triphos-phate; PKB, protein kinase B; PTEN, phosphatase and tensin homologdeleted on chromosome 10.

Int. J. Cancer: 121, 1424–1432 (2007)' 2007 Wiley-Liss, Inc.

Publication of the International Union Against Cancer

PI3K-Akt pathway.29 Akt/PKB activation has been shown to cor-relate with increased angiogenesis and metastasis throughhypoxia-inducible factor-1a.30 Increase of p-Akt expression, par-ticularly at serine 473, has been shown to correlate with higherGleason score and is an excellent predictor of poor clinical out-come in prostate cancer patients.31,32 Although high levels ofactivated Akt expression have been demonstrated in prostatecancer,31–33 it is unclear if PI3K-Akt activation plays an essentialrole during prostate cancer progression. We investigated the roleof the PI3K-Akt signaling pathway in this process in prostate can-cer cell lines and in primary human prostate cancers.

Material and methods

Tissue samples

Discarded benign and malignant prostate tissue from patientswithout any previous form of adjuvant therapy and who underwentsurgery was obtained from the Tissue Procurement Facility ofUniversity Hospitals Case Medical Center and the MidwesternDivision of the Cooperative Human Tissue Network. The Gleasongrade and score of adenocarcinoma specimens were assigned by asurgical pathologist experienced in genitourinary pathology. Im-mediately after procurement, samples were snap frozen in liquidnitrogen and stored at 280�C till further use. In addition, 4-lmparaffin-embedded tissue sections were obtained from 15 benignprostate specimens and from 46 prostate cancer specimens alongwith 6 cases of retrospective needle biopsy specimens. These stud-ies were approved by the Institutional Review Board at CaseWestern Reserve University.

Cell lines and reagents

Androgen-responsive human prostate cancer (LNCaP and22Rv1) and androgen-refractory (DU145 and PC-3) cells wereobtained from ATCC (Manassas, VA). Normal human prostateepithelial cells (NHPE) and culture medium were obtained fromClonetics1 (Walkersville, MD). RPMI 1640 medium and all othercell culture materials were obtained from Life Technologies(Gaithersburg, MD). Dominant negative Akt in pUSEamp(K179M mutant) (DN-Akt) and the empty vector pUSEamp (1)were purchased from Upstate Cell Signaling (Lake Placid, NY).LipofectamineTM 2000 and Geneticin (G-418 sulfate) were pur-chased from Invitrogen (Carlsbad, CA). The PI3K inhibitor,LY294002 was purchased from Sigma Chemical (St Louis, MO).Anti-Akt1/2, anti-PI3K (p110a/b), anti-MMP-9 and anti-uPAantibodies were purchased from Santa Cruz Biotechnologies(Santa Cruz, CA). Antibodies for anti-p-Akt (Ser473), anti-p-Akt(Thr308), anti-PI3K (p85), anti-PTEN and anti-p-GSK-3a/b werepurchased from Cell Signaling Technology1 (Danvers, MA).

Cell culture treatments

NHPE cells were cultured in PrEBM media and supplementsspecially designed to support growth of human primary derivedcells. Human prostate cancer cells were cultured under recom-mended condition in RPMI 1640 culture medium with 10% fetalbovine serum (FBS) and 1% penicillin–streptomycin. The cellswere maintained at 37�C and 5% CO2 in a humid environment. At60% confluence, cultures were switched to serum-free medium for16 hr, and then treated with specified doses of PI3K inhibitor,LY294002 (5–20 lM), in complete cell culture medium for vari-ous time intervals. After preferred treatments, medium was aspi-rated, cells were harvested by the addition of Trypsin-EDTA andcell extracts were prepared as previously described.34

For the transfection experiments, PC-3 cells were transfectedusing LipofectamineTM 2000 (Invitrogen, Carlsbad, CA) accord-ing to the vendor’s protocol. Cells in 100-mm dishes were co-transfected with 5 lg of pEGFP-C1-CMV (Clontech, MountainView, CA) for green fluorescent protein (GFP) expression togetherwith 1 lg of pCMV-XL5 vector or pCMV-XL5 vector encodinghuman cDNA clone PTEN (NM_000314) (OriGene, Rockville,

MD) with pEGFP-C1-CMV G-418 resistance vector. Cells weresubcultured at a 1:4 dilution 24 hr after transfection and main-tained for 5 days in 500 lg/mL G-418-containing medium to elim-inate untransfected cells. Transfection efficiency was determinedby observing transfected cells under a UV microscope, 48 hr post-transfection and counting bright cells. For invasion assays, cellswere serum starved in RPMI, 0.1% BSA during the last 24 hr ofG-418 selection to exclude any effects of growth suppression oncell invasion.

Cell proliferation assay

The effect of LY294002 and DN-Akt on the viability of cellswas determined by MTT (3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl tetrazoliumbromide) assay and the absorbance wasrecorded on a microplate reader at the wavelength of 540 nm. Theeffect of LY294002 and DN-Akt on cell growth inhibition wasassessed as percent cell viability where vehicle-treated cells weretaken as 100% viable.

Phospho-Akt (Ser473) ELISA assay

Following treatment with PI3K pharmacological inhibitor, theendogenous levels of phosphorylated Akt protein were determinedby the PathScanTM Phospho-Akt (Ser473) sandwich ELISA kit(no. 7160) (Cell Signaling Technology) according to vendor’s pro-tocol. The PathScan Phospho-Akt ELISA kit detects endogenouslevels of Akt only when it is phosphorylated at Serine 473. Themagnitude of optical density of this assay is proportional to thequantity of phosphorylated Akt1 protein.

Invasion assay

Invasion was assayed using Biocoat Matrigel Invasion Cham-bers (24-well size; Becton-Dickinson, Franklin Lakes, NJ) accord-ing to manufacturer’s protocol. All experiments were performedin triplicate. Cells were serum-starved in RPMI 1640, 0.1% BSAand seeded into the upper well of the invasion chamber in RPMI1640, 0.1% BSA. RPMI 1640, 10% FBS was added to the bottomwell of the chamber to serve as a chemoattractant. For pharmaco-logical inhibition assays, 10,000 cells/well were seeded in thepresence (both top and bottom chamber) or absence of inhibitors(treated with DMSO only), and invasion assessed after 24 hr bystaining membranes with Diff-Quick cell staining kit (Fisher Sci-entific, Pittsburgh, PA). At least 5 independent fields were countedfor each chamber. For transfection experiments, 10,000 G-418-selected cells were seeded into each chamber. Invasion wasassessed after 24 hr by staining membranes with Diff-Quick;again, at least 5 independent fields were counted for each cham-ber. Invasion of GFP transfected cells was normalized to 1 andresults for others were reported relative to GFP cells (relativeinvasion).

Immunoblot analysis

Frozen tissues (benign or cancer) and cultured cells were proc-essed for tissue/cell extract and the protein content was deter-mined using the DC Bio-Rad protein assay kit, as previouslydescribed.34 For immunoblot analysis, 40-lg protein was resolvedusing 4–20% polyacrylamide gels (Novex, Carlsbad, CA) andtransferred to a nitrocellulose membrane. The blot was blocked inblocking buffer (5% nonfat dry milk/1% Tween 20; in 20 mMTBS, pH 7.6) for 2 hr at room temperature, incubated with appro-priate primary antibody in blocking buffer for 2 hr at room tem-perature or overnight at 4�C, followed by incubation with theappropriate IgG secondary antibody conjugated to horseradishperoxidase (Amersham-Pharmacia, Piscataway, NJ) and detectedby ECL-chemiluminescence and autoradiography using XAR-5film (Eastman Kodak, Rochester, NY).

Immunohistochemistry

Immunohistochemical staining for p-Akt was performed usingthe SignalStainTM phospho-Akt (Ser473) IHC detection kit (Cell

1425ROLE OF PI3K-Akt IN PROSTATE CELL INVASION

Signaling Technology). Briefly, 4-lm-thick paraffin-embeddedsections from benign and cancer tissues were deparaffinized, rehy-drated, immersed in target retrieval solution, and blocked for en-dogenous peroxidase activity. The sections were permeabilized inTNB-BB (100 mM Tris, pH 7.5, 150 mM NaCl, 0.5% blockingagent, 0.3% Triton-X and 0.2% saponin) and incubated with pri-mary antibodies of prediluted p-Akt overnight at 4�C. Controlsections were incubated with antisera in the presence of 10-foldexcess of these antibodies or with isotype-matched IgG normalgoat serum. After washing 3 times in TBS, sections were incu-bated for 2 hr at room temperature with biotinylated secondaryantibody. Immunoreactive complexes were detected using diamino-benzidine substrate-chromagen. Slides were then counterstained inMayer’s hematoxylin, mounted in crystal mount media and driedovernight on a level surface. For fluorescence assay of p-Akt(Ser473), sections were subsequently incubated with 4.0 lg/mLTexas Red fragment of donkey anti-mouse IgG (Santa Cruz) andimages were captured using a fluorescent microscope (Olympus,BX51).

Immunostaining evaluation

The immunostained sections were examined independently by 3of the authors (SS, SG and GTM) using light microscopy. Com-puter-assisted morphometric methods were applied for analysisand nuclear staining of p-Akt in the high-grade cancer cells wereused as positive control. Sections were examined with an invertedOlympus BX51 microscope and images were acquired with Olym-pus MicroSuiteTM Five Software (Soft Imaging System, Lake-wood, CO). The intensity of staining was graded semiquantita-tively and each specimen was assigned a score on a scale from 0to 3 designated as 0 (negative), 11 when 10–20% of cells stain(weak), 21 when 20–50% of cell stain (moderate) and 31 when>50% of cell stain (strong). The immunoreactive score was desig-nated by the percentage of positive cells and the staining intensity,as previously described.32 The percent of staining for p-Akt wasscored by counting the positive-stained cells and total number ofcells quantified in random microscopic fields (4003 magnifica-tion) with the assistance of the software program.

Statistical analysis

Tumor grade (Gleason score) and p-Akt expression determinedat different times were summarized by mean, median, range andSEM. The difference in tumor grade, p-Akt expression at varioustime intervals was examined by paired t-test. All tests were two-sided and p-values less than 0.05 were considered statisticallysignificant.

Results

Constitutive PI3K expression and Akt phosphorylationis higher in some prostate cancer cells

To determine whether the PI3K-Akt signaling pathway is acti-vated in human prostate cancer, we evaluated a panel of 5 humanprostate-derived cell lines that included normal prostate epithelial(NHPE) cells, androgen-responsive prostate cancer (LNCaP and22Rv1) and androgen-refractory prostate cancer (DU145 and PC-3) cells. The cells were deprived of serum and/or supplementsovernight, and subsequently transferred to complete cell culturemedium for 3 hr; following this, expression of PI3K and p-Aktwas evaluated in total cell lysate. As shown in Figure 1a, proteinlevels of catalytic (p110a/b) and regulatory (p85) subunits ofPI3K were modestly elevated in androgen responsive-LNCaP andrefractory-PC-3 cells compared to NHPE and other human pros-tate cancer cells. Both LNCaP and PC-3 cells were highly phos-phorylated at Ser473, compared to other cell lines examined.In vitro kinase assay using glycogen synthase kinase-3a/b as asubstrate was performed to verify that phosphorylated Akt/PKB isenzymatically active in these cells. The highest level of Akt/PKBenzymatic activity was observed in PC-3 and LNCaP cells, which

have shown higher levels of p-Akt previously (Figs. 1a and 1b).These results suggest that phosphorylation of Akt at Ser473 corre-lates precisely with Akt kinase activity.

For further studies we selected 3 primary cancer cell lines,LNCaP, PC-3 and DU145, which are widely used and representa-tive of advanced prostate cancer and are highly tumorigenic.35

LNCaP cells harbor PTEN mutations, and PC-3 cells harborPTEN deletion.36 However, despite upregulated Akt status, thedegree of dependence on the PI3K-Akt pathway for invasivenessvaries between LNCaP and PC-3 cells (Table I). Therefore, todetermine whether activation of PI3K-Akt may lead to prostatecancer progression, we used another androgen-refractory humanprostate cancer cell line, DU145, which expresses modest proteinexpression of PI3K and p-Akt levels, without mutation of PTEN.As control, we used normal human prostate epithelial (NHPE)cells in these experiments.

FIGURE 1 – Protein expression of (a) catalytic (p110a/b) and regu-latory (p85) subunits of PI3K, total Akt, p-Akt and PTEN in normalhuman prostate epithelial (NHPE), and carcinoma cells (LNCaP,22Rv1, DU145 and PC-3). Cells were grown in serum/supplement-depleted medium for 16 hr, and switched to complete culture mediumfor 3 hr. Total cell extracts were prepared and electrophoresed bySDS-PAGE, followed by immunoblotting with anti-p110a/b, anti-p85, anti-Akt, anti-p-Akt (Ser473) and anti-PTEN in the total celllysate. In vitro kinase assay of Akt immunoprecipitates from normaland carcinoma cells were determined using Akt kinase assay with gly-cogen synthase kinase (GSK)-3a/b as a substrate. To ensure equalprotein loading the blots were stripped and reprobed with anti-a-tubu-lin antibody. (b) Densitometic analysis of each protein expressed per-cent relative to normal prostate epithelial cells. The details aredescribed in Material and methods section.

1426 SHUKLA ET AL.

Pharmacological inhibition of the PI3K-Akt pathway reducesinvasiveness of human prostate cancer cells

To determine whether activated Akt/PKB promotes invasive-ness of prostate cancer cells, we used a pharmacological approachto alter PI3K-Akt activity and assessed whether this inhibitionleads to diminished cell invasion ability. We used the Matrigelinvasion-chamber assay, which measures 2 important factors thatcontribute to tumor aggressiveness, motility and invasiveness.Cells must first degrade a reconstituted basement membrane thatoccludes the pores of a filter and then move through these clearedpores to adhere to the other side. As expected, NHPE cells werecompletely noninvasive using this assay, while the LNCaP, PC-3and DU145 cells were invasive. A highly specific pharmacologicalinhibitor of PI3K catalytic activity, LY294002, was used as an ini-tial approach to assess the significance of the PI3K pathway in

invasion. To initiate these experiments, for each cell line an opti-mal concentration had to be established with LY294002 that maxi-mally suppressed its target pathway, but maintained cellular via-bility over the course of a 24-hr invasion assay. Thus, we deter-mined a dose–response curve with PI3K inhibitor and examinedpathway inhibition and effects on cellular viability. In the experi-ments shown in Figures 2a and 2b, lysates were prepared from all3 cell lines after 16-hr treatment with LY294002 (5–20 lM) andwere blotted with anti-Akt1/2, anti-p-Akt and anti-PI3K antibod-ies. While LY294002 inhibited Akt phosphorylation as expected,it had no effect on the total Akt1/2 levels in all 3 cancer cell lines.Treatment of all cancer cell lines at these doses of LY294002 upto 16 hr did not exhibit any significant change in cell viability;time chosen for a cut-off point (data not shown).

In the final experiments, the concentration of PI3K inhibitorwas chosen for each cell line that inhibited the target pathway by>90% while maintaining >90% cellular viability. Figure 2cshows the effect of LY294002 on cell invasion. While the totalAkt level remains unchanged, inhibition of Akt phosphorylationshowed positive correlation with decreased invasiveness in all 3cancer cell lines. Compared to untreated controls, decrease ininvasiveness and p-Akt levels were in the order PC-3 > LNCaP >DU145, respectively (Figs. 2b and 2c).

Biochemical manipulation of the PI3K-Akt pathway reducesinvasiveness in human prostate cancer PC-3 cells

Previous results demonstrated strong evidence for an essentialrole of the PI3K pathway in prostate cancer cell invasion. We

TABLE I – CORRELATION OF PTEN, PI3-KINASE AND P-AKT LEVELS WITHINVASIVENESS IN VARIOUS HUMAN PROSTATE EPITHELIAL CELLS

Cell line PTEN PI3K p-Akt Invasion

NHPE Normal 22 22 22LNCaP Mutated 11 111 1122Rv1 Normal 22 22 11DU145 Normal 22 22 111PC-3 Deleted 11 111 111

NHPE, normal prostate epithelial cells; LNCaP and 22Rv1, androgen-responsive cancer cells; DU145 and PC-3, androgen-refractorycancer cells.

FIGURE 2 – Effect of LY294002 treatment on the protein expression of (a) regulatory (p85) subunit of PI3K, total Akt, p-Akt (Ser473) inhuman prostate cancer cells (LNCaP, PC-3 and DU145). Cells were grown in complete culture medium along with increasing doses ofLY294002 for 16 hr. Total cell extracts were prepared and electrophoresed by SDS-PAGE, followed by immunoblotting with anti-p85, anti-Aktand anti-p-Akt (Ser473) in the total cell lysate. To ensure equal protein loading, the membrane was stripped and reprobed with anti-a-tubulinantibody. (b) Akt phosphorylation at Ser473 by ELISA in NHPE and human prostate cancer LNCaP, PC-3 and DU145 cells after treatment withincreasing doses of LY294002. (c) For invasion assay, cells were either treated with DMSO (control) or cultured for 16 hr in the presence ofincreasing concentrations of LY294002. The details are described in Material and methods section. Bars 6 SE, *p < 0.05, **p < 0.001, com-pared to control.

1427ROLE OF PI3K-Akt IN PROSTATE CELL INVASION

sought to confirm these results independently by inhibiting thePI3K pathway with the expression of 2 negative regulators; the an-tagonist PTEN or a dominant negative version of Akt and its cor-relation to invasion markers.

PC-3 cells were co-transfected with pEGFP-C1-CMV (for GFPexpression) vector encoding G-418 resistance with pCMV-XL5human cDNA clone PTEN or DN-Akt pUSEamp(1). Cells wereplaced in G-418 containing media for 5 days to select cellsexpressing both plasmids, and then the same numbers of cellswere seeded into invasion chambers in triplicate. G-418 selectioneliminates untransfected cells, resulting in a pure population ofexpressing cells. As the expression of PTEN or DN-Akt in PC-3cells may inhibit their growth, cells were serum-starved during thelast 24 hr of G-418 selection to suppress growth, excluding contri-butions of growth suppression in the invasion assay as previouslydemonstrated.37 After selection, transfected cells were also seededonto coverslips and stained with DAPI, which revealed no evi-dence of condensed or fragmented nuclei (morphological changesindicative of apoptosis) in the GFP, GFP-PTEN or GFP/DN-Akttransfectants (data not shown). In addition, in parallel experi-ments, it was determined that transfection of PTEN or DN-Akthad no significant effect on cell number or viability of selectedcells during the 24-hr period of the invasion assay (data notshown). In a separate study, adenoviral-mediated overexpressionof MMAC/PTEN in PC-3 cells reduced cell numbers by 46% at 6days after infection.36 Increasing levels of MMAC/PTEN wereassociated with more growth inhibition.36,38 While there was nogrowth inhibition by PTEN during the 24-hr period used it is likelythat the expression levels of PTEN achieved did not approachthose obtained with adenoviral-mediated overexpression.

As shown in Figure 3a, the effect of expression of PTEN andDN-Akt on Akt phosphorylation in PC-3 cells was visualized byWestern blotting. The expression of PTEN coincides with abol-ished Akt phosphorylation similar to that observed with 20 lMLY294002 treatment for 24 hr, while expression of DN-Aktreduced Akt phosphorylation by 85%. The extent of inhibition ofinvasion by PTEN, DN-Akt and LY294002 treatment was quanti-fied, with GFP invading cells normalized to 1 (Fig. 3b). Theexpression of PTEN inhibited the invasive properties of PC-3 cellsby 95%, expression of DN-Akt inhibited invasion by 92% andLY294002 inhibited invasion by 80%. This experiment wasrepeated 3 times and the results shown in Figure 3b are representa-tive. Interestingly, inhibition of p-Akt expression correlated withdecreased protein expression of urokinase-type plasminogen acti-

vator (uPA) and matrix metalloproteinase (MMP)-9, markers forcell invasion. These results confirm the observations from ourpharmacological inhibition studies, together with biochemicalmanipulation indicating that the PI3K-Akt pathway is essential inthe invasive behavior of prostate cancer cells.

Constitutive PI3K expression and Akt phosphorylationare elevated in human prostate cancer specimens

Next we sought to ascertain whether the PI3K-Akt pathway,Akt phosphorylation in particular, was overactive in human pros-tate adenocarcinoma. We performed immunoblot analysis forPI3K (catalytic subunit p110a/b and regulatory subunit p85),total- and activated-Akt levels in benign prostate tissue and pros-tate cancer specimens. As shown in Figure 4a, immunoblot analy-sis for catalytic and regulatory subunits of PI3K exhibited a signif-icant increase in high-grade cancer tissue (Gleason score 514;and 515) compared with low-grade cancer (Gleason score 313)and benign tissue. A similar pattern of protein expression for acti-vated Akt (p-Akt; Ser473) was observed as previously shown forPI3K with significantly higher levels in high-grade cancer tissuecompared to low-grade cancer and benign tissue. However, no sig-nificant variation in the protein expression of total Akt (Akt1/2)was observed in these tissues (Fig. 4a). Similar results wereobserved for p-Akt (Thr308) in these tissues (data not shown). Wealso compared the protein levels of PTEN in these tissues. For ourstudies, we used monoclonal antibody produced by immunizingmice with a synthetic peptide (KLH coupled) derived from thecarboxyl terminus sequence of human PTEN recognizing a 54-kDa protein. The endogenous PTEN protein was detected in alltissue specimens examined. Compared to benign tissue, PTENprotein expression was higher in low-grade cancer with a loss ofexpression observed in high-grade cancer tissues (Fig. 4a). Ourresults are in agreement with previous studies reporting loss ofPTEN protein expression in advanced grade cancer, which subse-quently leads to prostate cancer progression.39,40 The densitomet-ric analysis for these proteins are shown in Figure 4b.

Next we examined the protein expression of p-Akt (Ser473) byimmunohistochemical analysis in benign prostate tissue and prostatecancer specimens. Cancers were analyzed according to Gleasongrade and tissue specimens were assigned to 4 subgroups consist-ing of benign tissue, low-grade cancer (Gleason score 2–4), me-dium-grade cancer (Gleason score 5–7) and high-grade cancer(Gleason score 8–10). The degree of immunoreactivity of these

FIGURE 3 – Inhibition of PC-3 cell invasion by PTEN overexpression, PI3K inhibitor (LY240092) treatment and Akt dominant-negativeapproach. (a) Cells were transfected with either pEGFP-C1-CMV (for GFP expression) vector encoding G-418 resistance with pCMV-XL5human cDNA clone PTEN or DN-Akt pUSEamp (1) or treated with 20 lM LY294002 for 24 hr. Total cell extract were prepared and electro-phoresed by SDS-PAGE, followed by immunoblotting with anti-PTEN, anti-p-Akt (Ser473), anti-uPA and anti-MMP-9 antibodies. For proteinloading the blots were stripped and reprobed with anti-a-tubulin antibody. (b) PTEN- and DN-Akt transfected and LY294002 treated cells wereseeded into invasion chambers in triplicate and invasion was measured. The relative number6 SE of invading cells is plotted for each condition.Bars 6 SE, **p < 0.001.

1428 SHUKLA ET AL.

FIGURE 4 – Protein expression of (a) catalytic (p110a/b) and regulatory (p85) subunits of PI3K, total Akt, p-Akt (Ser473) and PTEN in benignprostate tissue and prostate cancer specimens. Total cell extracts from tissues were prepared and electrophoresed by SDS-PAGE, followed by immu-noblotting with anti-p110a/b, anti-p85, anti-Akt, anti-p-Akt (Ser473) and anti-PTEN antibodies. Cancer tissue is subdivided into low-grade cancer(Gleason score <7) and high-grade cancer (Gleason score 7–10). (b) Densitometic analysis of each protein normalized to epithelial content by prob-ing the blots with anti-cytokeratin (CK) 18 antibodies. Bars 6 SE, *p < 0.05, **p < 0.001, compared to benign tissue. (c) Immunostaining for acti-vated Akt (p-Akt, Ser473) and its evaluation in representative samples of benign prostate tissue and prostate cancer specimens of various Gleasongrades. Magnification 3200. Evaluation of total sections based on staining intensity scored as weak, moderate or strong. A detailed staining patternof benign and prostate cancer specimens is shown below the figures. The details are described in Material and methods section. [Color figure can beviewed in the online issue, which is available at www.interscience.wiley.com.]

1429ROLE OF PI3K-Akt IN PROSTATE CELL INVASION

tissues was assessed by evaluation of the amount and intensity oftissue immunostaining in the sections. As shown in Figure 4c, be-nign tissues did not exhibit any significant p-Akt expression. Aprogressive increase in the staining intensity of p-Akt (Ser473)with increasing tumor grade was observed in the prostate cancerspecimens. Of 15 benign specimens that were analyzed, 2 exhib-ited weak expression and 13 were negative for p-Akt. In 6 low-grade tumors, moderate p-Akt expression was observed in 1, 1exhibited weak staining and 4 showed no p-Akt staining. In 18medium-grade tumors, strong p-Akt expression was observed in 6,4 exhibited moderate staining, 3 exhibited weak staining and 5showed no p-Akt staining. In 22 high-grade tumors, 10 exhibitedstrong staining, 6 exhibited moderate staining, 3 exhibited weakstaining and 3 were negative for p-Akt staining (Fig. 4c).

Additionally, in 6 cases we had an opportunity to evaluate se-quential samples of prostate tissue obtained at different periods ofthe patient’s life; for example, 2 initial prostate biopsies followedwithin 5 years by a transuretheral resection of the prostate. Thesespecimens were studied to address whether levels of p-Aktincreases with increased malignancy. An increase in the levels ofp-Akt (Ser473) was observed in all these specimens, which corre-lated with increasing Gleason score (Figs. 5a–5c). Since activationof Akt is essential for efficient prostate cancer cell invasion, and

tumor invasion correlates strongly with poor prognosis in patients,it seems probable that activation of the PI3K-Akt pathway is criti-cal in the progression of some prostate tumors.

Discussion

Prostate cancer cells utilize multiple molecular pathways to pro-liferate and invade tissue during the course of tumor progression.1

Among several independent cell survival signaling pathways,upregulation of PI3K-Akt signaling through mutations in thePTEN gene and constitutive activation of growth factor receptorsare particularly important.3,7 Lipid products of PI3K provide ananchor for assembling signaling proteins at specific locations inthe membrane in response to cell stimulation. These signaling pro-teins coordinate complex events that lead to changes in cell me-tabolism, cell growth, cell motility, invasiveness and survival.7

The PI3K-Akt signaling pathway is believed to play an importantrole in the genesis of some human cancers.2 However, the role ofthe PI3K-Akt signaling pathway in the progression of prostatecancer has not been yet established. Because Akt provides strongcell survival signals in response to external stimuli and is activatedin numerous human malignancies,5,6,12–14 we hypothesized that

FIGURE 5 – Time-course evaluation of p-Akt (Ser473) expression and its correlation with tumor progression. (a) Immunoflourescence stainingfor p-Akt (Ser473) in needle biopsy specimens and resected tissue from same patient (no. 1). The H&E staining is shown above each panel.Magnification 3100. (b) Evaluation of total sections based on staining intensity and its correlation with tumor grade. (c) Statistical analysis oflongitudinal data of tumor grade and p-Akt expression in consecutive specimens along with the p-values. [Color figure can be viewed in theonline issue, which is available at www.interscience.wiley.com.]

1430 SHUKLA ET AL.

PI3K-Akt signaling plays a role in prostate carcinogenesis. Inthis study, we have shown not only that PI3K and Akt are consti-tutively activated in human prostate adenocarcinoma, but alsothat they enhance invasiveness resulting in tissue invasion. Inaddition, we have shown that increasing Akt activation correlateswith increasing grade of prostate cancer, suggesting that Akt im-munostaining may be useful as a prognostic indicator of tumoraggressiveness.

Accumulating evidence indicates a potential role for the PI3Kpathway in prostate cancer progression. Mutation and/or loss offunction in the negative regulator PTEN has been observed inadvanced stage human prostate cancer9,39,40 and in xenograft mod-els.36,41 Loss of PTEN function through PTEN mutations in mu-rine models has been shown to be associated with neoplasia inmultiple organ sites, including endometrium, liver, gastrointestinaltract, thyroid, thymus and prostate.42 Loss of PTEN function inprostate cancer cells has been shown to be associated withincreased proliferation, angiogenesis and tumorigenesis.43,44 Innearly 50% of prostate cancers, the PI3K-Akt survival pathwayhas been shown to be constitutively upregulated because of loss offunction and/or mutations of tumor suppressor PTEN, which func-tions as a negative regulator of PI3K through its lipid phosphataseactivity.39,40 Previous studies have demonstrated that loss ofPTEN function in some prostate cancer cells leads to higher Aktexpression,41 a feature that may promote prostate cancer progres-sion. We have additionally observed that PTEN protein expressiondiminishes with increasing grade of cancer, with maximum loss ofexpression observed in high-grade cancer specimens of Gleasonscore 8–10. These observations suggest that aberrant activation ofthe PI3K-Akt pathway by any number of mechanisms may con-tribute to increased tumor invasiveness and cancer progression.

We evaluated the role of constitutive Akt activation in cell inva-sion by studying a PI3K pharmacological inhibitor, LY294002, aswell as a dominant negative mutant construct of Akt (DN-Akt).Highly invasive LNCaP and PC-3 prostate cancer cells harbormutations or deletions of PTEN, and each has a high level of PI3Kactivity, judging from the LY294002-mediated inhibition of Aktphosphorylation that we observed. LY294002 suppressed the inva-sive properties of each of these cell lines, and reconstitution ofintact PTEN into PC-3 cells also efficiently suppressed invasion.This indicates that the 3-phosphorylated lipid products of PI3Kare necessary for cellular invasion. In support of our findings, aprevious study has demonstrated that treatment of prostate cancercells with PI3K inhibitor, LY294002, resulted in cell cycle-medi-ated arrest and induction of apoptosis.45 In addition, androgen re-

ceptor-mediated Akt activation has also been shown to enhancecell growth and survival of prostate cancer cells, through the nu-clear b-catenin signaling pathway.46 More recent studies havedemonstrated that conditional Akt activation promotes androgen-independent progression and is essential for neuroendocrine dif-ferentiation of prostate cancer.27,28 While the PI3K-Akt pathwayis critical for prostate cancer invasion, several other pathways areknown to be involved in tumor cell invasion. These include focaladhesion kinase, C-Jun-N terminal kinase, phospholipase C-d andRas/ERK1/2, which are involved in invasion of many cancer typesincluding prostate cancer and their activation is sufficient toinduce invasive behavior.47,48

Activation of the PI3K/Akt pathway confers chemotherapeuticresistance in numerous tumor types including cancers of lung, cer-vix, ovary, pancreas, bladder and breast.2–5,49 Adenoviral-medi-ated expression of PTEN inhibits proliferation and metastasis inhuman prostate cancer PC-3 cells.37 Overexpression of PTEN anddiminution of Akt phosphorylation restores doxorubicin sensitivityto the doxorubicin-resistant prostate cancer PC-3 cells.50 All ofthe aforementioned studies suggest that PI3K and Akt may bepromising molecular targets in the management of prostate cancer.As the pathway appears to be involved in several cellular proc-esses it will be important to identify biochemical and gene targetsof PI3K-Akt activation that specifically lead to increased tumoraggressiveness and chemotherapeutic resistance.

In summary, our studies suggest that constitutive PI3K-Akt acti-vation actively contributes to the progress of prostate cancer fromorgan-confined disease to highly invasive and potentially meta-static disease. Our results also point to the possibility that p-Aktmight be used as a marker for those low-grade tumors that are atrisk of progression to high-grade invasive tumors. We found thatphosphorylation and activation of Akt increases tumor invasive-ness, which increased in parallel with increasing grade of cancer,with maximum activation observed in high-grade cancer speci-mens (Gleason score 8–10). Thus, PI3K-Akt activation may be animportant prognostic indicator of tumor aggressiveness. Further-more, PI3K-Akt and its associated regulatory signaling pathwaysare potential targets for therapeutic intervention and molecular-based approaches for management of prostate cancer in humans.

Acknowledgements

This work was conducted in The James and Eilleen DickeResearch Laboratory, Department of Urology, Case WesternReserve University, Cleveland, OH.

References

1. Martin GS. Cell signaling and cancer. Cancer Cell 2003;4:167–74.2. Vivanco I, Sawyers CL. The phosphatidylinositol 3-Kinase AKT

pathway in human cancer. Nat Rev Cancer 2002;2:489–501.3. Paez J, Sellers WR. PI3K/PTEN/AKT pathway. A critical mediator of

oncogenic signaling. Cancer Treat Res 2003;115:145–67.4. Fresno Vara JA, Casado E, de Castro J, Cejas P, Belda-Iniesta C,

Gonzalez-Baron M. PI3K/Akt signalling pathway and cancer. CancerTreat Rev 2004;30:193–204.

5. Ma YY, Wei SJ, Lin YC, Lung JC, Chang TC, Whang-Peng J, LiuJM, Yang DM, Yang WK, Shen CY. PIK3CA as an oncogene incervical cancer. Oncogene 2000;19:2739–44.

6. Zhang L, Yang N, Katsaros D, Huang W, Park JW, Fracchioli S,Vezzani C, Rigault de la Longrais IA, Yao W, Rubin SC, Coukos G.The oncogene phosphatidylinositol 30-kinase catalytic subunit alphapromotes angiogenesis via vascular endothelial growth factor in ovar-ian carcinoma. Cancer Res 2003;63:4225–31.

7. Blume-Jensen P, Hunter T. Oncogenic kinase signalling. Nature2001;411:355–65.

8. Guldberg P, thor Straten P, Birck A, Ahrenkiel V, Kirkin AF, ZeuthenJ. Disruption of the MMAC1/PTEN gene by deletion or mutationis a frequent event in malignant melanoma. Cancer Res 1997;57:3660–3.

9. Li J, Yen C, Liaw D, Podsypanina K, Bose S, Wang SI, Puc J, Milia-resis C, Rodgers L, McCombie R, Bigner SH, Giovanella BC, et al.PTEN, a putative protein tyrosine phosphatase gene mutated in humanbrain, breast, and prostate cancer. Science 1997;275:1943–7.

10. Risinger JI, Hayes AK, Berchuck A, Barrett JC. PTEN/MMAC1mutations in endometrial cancers. Cancer Res 1997;57:4736–8.

11. Wang SI, Puc J, Li J, Bruce JN, Cairns P, Sidransky D, Parsons R.Somatic mutations of PTEN in glioblastoma multiforme. Cancer Res1997;57:4183–6.

12. Bellacosa A, de Feo D, Godwin AK, Bell DW, Cheng JQ, AltomareDA, Wan M, Dubeau L, Scambia G, Masciullo V. Molecular altera-tions of the AKT2 oncogene in ovarian and breast carcinomas. Int JCancer 1995;64:280–5.

13. Staal SP. Molecular cloning of the akt oncogene and its human ho-mologues AKT1 and AKT2: amplification of AKT1 in a primaryhuman gastric adenocarcinoma. Proc Natl Acad Sci USA 1987;84:5034–7.

14. Nakatani K, Thompson DA, Barthel A, Sakaue H, Liu W, Weigel RJ,Roth RA. Up-regulation of Akt3 in estrogen receptor-deficient breastcancers and androgen-independent prostate cancer lines. J Biol Chem1999;274:21528–32.

15. Chang F, Lee JT, Navolanic PM, Steelman LS, Shelton JG, BlalockWL, Franklin RA, McCubrey JA. Involvement of PI3K/Akt pathwayin cell cycle progression, apoptosis, and neoplastic transformation: atarget for cancer chemotherapy. Leukemia 2003;17:590–603.

16. Shioi T, McMullen JR, Kang PM, Douglas PS, Obata T, Franke TF,Cantley LC, Izumo S. Akt/protein kinase B promotes organ growth intransgenic mice. Mol Cell Biol 2002;22:2799–809.

17. Nicholson KM, Anderson NG. The protein kinase B/Akt signallingpathway in human malignancy. Cell Signal 2002;14:381–95.

1431ROLE OF PI3K-Akt IN PROSTATE CELL INVASION

18. Muise-Helmericks RC, Grimes HL, Bellacosa A, Malstrom SE, Tsi-chlis PN, Rosen N. Cyclin D expression is controlled post-transcrip-tionally via a phosphatidylinositol 3-kinase/Akt-dependent pathway.J Biol Chem 1998;273:29864–72.

19. Ogawara Y, Kishishita S, Obata T, Isazawa Y, Suzuki T, Tanaka K,Masuyama N, Gotoh Y. Akt enhances Mdm2-mediated ubiquitinationand degradation of p53. J Biol Chem 2002;277:21843–50.

20. Rommel C, Clarke BA, Zimmermann S, Nunez L, Rossman R, ReidK, Moelling K, Yancopoulos GD, Glass DJ. Differentiation stage-specific inhibition of the Raf-MEK-ERK pathway by Akt. Science1999;286:1738–41.

21. Gottlob K, Majewski N, Kennedy S, Kandel E, Robey RB, Hay N. In-hibition of early apoptotic events by Akt/PKB is dependent on the firstcommitted step of glycolysis and mitochondrial hexokinase. GenesDev 2001;15:1406–18.

22. Testa JR, Bellacosa A. AKT plays a central role in tumorigenesis.Proc Natl Acad Sci USA 2001;98:10983–5.

23. Jiang BH, Zheng JZ, Aoki M, Vogt PK. Phosphatidylinositol 3-kinasesignaling mediates angiogenesis and expression of vascular endothe-lial growth factor in endothelial cells. Proc Natl Acad Sci USA2000;97:1749–53.

24. Tu Y, Gardner A, Lichtenstein A. The phosphatidylinositol 3-kinase/AKT kinase pathway in multiple myeloma plasma cells: roles incytokine-dependent survival and proliferative responses. Cancer Res2000;60:6763–70.

25. Graff JR, Konicek BW, McNulty AM, Wang Z, Houck K, Allen S,Paul JD, Hbaiu A, Goode RG, Sandusky GE, Vessella RL, NeubauerBL. Increased AKT activity contributes to prostate cancer progressionby dramatically accelerating prostate tumor growth and diminishingp27Kip1 expression. J Biol Chem 2000;275:24500–5.

26. Lin HK, Yeh S, Kang HY, Chang C. Akt suppresses androgen-induced apoptosis by phosphorylating and inhibiting androgen recep-tor. Proc Natl Acad Sci USA 2001;98:7200–5.

27. Li B, Sun A, Youn H, Hong Y, Terranova PF, Thrasher JB, Xu P.Conditional Akt activation promotes androgen-independent progres-sion of prostate cancer. Carcinogenesis 2007;28:572–83.

28. Wu C, Huang J. PI3 kinase-AKT-mTOR pathway is essential for neu-roendocrine differentiation of prostate cancer. J Biol Chem, in press.

29. Pfeil K, Eder IE, Putz T, Ramoner R, Culig Z, Ueberall F, Bartsch G,Klocker H. Long-term androgen-ablation causes increased resistanceto PI3K/Akt pathway inhibition in prostate cancer cells. Prostate2004;58:259–68.

30. Jiang BH, Jiang G, Zheng JZ, Lu Z, Hunter T, Vogt PK. Phosphatidyl-inositol 3-kinase signaling controls levels of hypoxia-inducible factor1. Cell Growth Differ 2001;12:363–9.

31. Kreisberg JI, Malik SN, Prihoda TJ, Bedolla RG, Troyer DA, Kreis-berg S, Ghosh PM. Phosphorylation of Akt (Ser473) is an excellentpredictor of poor clinical outcome in prostate cancer. Cancer Res2004;64:5232–6.

32. Malik SN, Brattain M, Ghosh PM, Troyer DA, Prihoda T, Bedolla R,Kreisberg JI. Immunohistochemical demonstration of phospho-Akt in high Gleason grade prostate cancer. Clin Cancer Res 2002;8:1168–71.

33. Ayala G, Thompson T, Yang G, Frolov A, Li R, Scardino P, Ohori M,Wheeler T, Harper W. High levels of phosphorylated form of Akt-1in prostate cancer and non-neoplastic prostate tissues are strongpredictors of biochemical recurrence. Clin Cancer Res 2004;10:6572–8.

34. Gupta S, Hastak K, Afaq F, Ahmad N, Mukhtar H. Essential role ofcaspases in epigallocatechin-3-gallate-mediated inhibition of nuclearfactor jB and induction of apoptosis. Oncogene 2004;23:2507–22.

35. Mitchell S, Abel P, Ware M, Stamp G, Lalani E. Phenotypic andgenotypic characterization of commonly used human prostatic celllines. BJU Int 2000;85:932–44.

36. Davies MA, Kim SJ, Parikh NU, Dong Z, Bucana CD, Gallick GE.Adenoviral-mediated expression of MMAC/PTEN inhibits prolifera-tion and metastasis of human prostate cancer cells. Clin Cancer Res2002;8:1904–14.

37. Tanaka M, Koul D, Davies MA, Liebert M, Steck PA, Grossman HB.MMAC1/PTEN inhibits cell growth and induces chemosensitivity todoxorubicin in human bladder cancer cells. Oncogene 2000;19:5406–12.

38. Davies MA, Koul D, Dhesi H, Berman R, McDonnell TJ, McConkeyD, Yung WK, Steck PA. Regulation of Akt/PKB activity, cellulargrowth, and apoptosis in prostate carcinoma cells by MMAC/PTEN.Cancer Res 1999;59:2551–6.

39. Whang YE, Wu X, Suzuki H, Reiter RE, Tran C, Vessella RL, SaidJW, Isaacs WB, Sawyers CL. Inactivation of the tumor suppressorPTEN/MMAC1 in advanced human prostate cancer through loss ofexpression. Proc Natl Acad Sci USA 1998;95:5246–50.

40. McMenamin ME, Soung P, Perera S, Kaplan I, Loda M Sellers WR.Loss of PTEN expression in paraffin-embedded primary prostate can-cer correlates with high Gleason score and advanced stage. CancerRes 1999;59:4291–6.

41. Vlietstra RJ, van Alewijk DC, Hermans KG, van Steenbrugge GJ,Trapman J. Frequent inactivation of PTEN in prostate cancer celllines and xenografts. Cancer Res 1998;58:2720–3.

42. Podsypanina K, Ellenson LH, Nemes A, Gu J, Tamura M, YamadaKM, Cordon-Cardo C, Catoretti G, Fisher PE, Parsons R. Mutation ofPten/Mmac1 in mice causes neoplasia in multiple organ systems. ProcNatl Acad Sci USA 1999;96:1563–8.

43. Koul D, Shen R, Garyali A, Ke LD, Liu TJ, Yung WK. MMAC/PTEN tumor suppressor gene regulates vascular endothelial growthfactor-mediated angiogenesis in prostate cancer. Int J Oncol 2002;21:469–75.

44. Wang S, Gao J, Lei Q, Rozengurt N, Pritchard C, Jiao J, Thomas GV,Li G, Roy-Burman P, Nelson PS, Liu X, Wu H. Prostate-specific dele-tion of the murine Pten tumor suppressor gene leads to metastaticprostate cancer. Cancer Cell 2003;4:209–21.

45. Gao N, Zhang Z, Jiang BH, Shi X. Role of PI3K/AKT/mTOR signal-ing in the cell cycle progression of human prostate cancer. BiochemBiophys Res Commun 2003;310:1124–32.

46. Chesire DR, Ewing CM, Gage WR, Isaacs WB. In vitro evidence forcomplex modes of nuclear b-catenin signaling during prostate growthand tumorigenesis. Oncogene 2002;21:2679–94.

47. Sliva D. Signaling pathways responsible for cancer cell invasion astargets for cancer therapy. Curr Cancer Drug Targets 2004;4:327–36.

48. Price JT, Thompson EW. Mechanisms of tumour invasion and metas-tasis: emerging targets for therapy. Expert Opin Ther Targets 2002;6:217–33.

49. West KA, Castillo SS, Dennis PA. Activation of the PI3K/Akt path-way and chemotherapeutic resistance. Drug Resist Update 2002;5:234–48.

50. Grunwald V, DeGraffenried L, Russel D, Friedrichs WE, Ray RB,Hidalgo M. Inhibitors of mTOR reverse doxorubicin resistance con-ferred by PTEN status in prostate cancer cells. Cancer Res 2002;62:6141–5.

1432 SHUKLA ET AL.