Cancer Research Tuberous Sclerosis Complex 1: An Epithelial … · Tumor and Stem Cell Biology Tuberous Sclerosis Complex 1: An Epithelial Tumor Suppressor Essential to Prevent Spontaneous

Tum

TubSupPro

RaleigJason

Abst

Intro

Humgressiprogretions isitidemammas a cevationsupprlogue

AuthorsMedicinCancerof Mediof CaliDepartmSchoolSanford

Note: SOnline

CorresAvenueFax: 31

doi: 10

©2010

www.a

Dow

Published OnlineFirst October 12, 2010; DOI: 10.1158/0008-5472.CAN-10-1646

Canceresearch

or and Stem Cell Biology

erous Sclerosis Complex 1: An Epithelial Tumorpressor Essential to Prevent Spontaneous

R

state Cancer in Aged Mice

h D. Kladney1, Robert D. Cardiff5, David J. Kwiatkowski6, Gary G. Chiang7,
D. Weber1, Jeffrey M. Arbeit2,3,4, and Zhi Hong Lu2
ractThe

plays astreamgrowthactivapotenbut Tsby 6 mold muand atcontaidespitregula

essor of thdeleted o

' Affiliationse, 2DivisionCenter, andcine, St. Loufornia, Davent of Med, Boston, M-Burnham M

upplementa(http://cance

pondingAu, Campus B4-362-3246

.1158/0008-

American A

acrjourna

nloaded

phosphoinositide 3-kinase (PI3K) pathway regulates mammalian cell growth, survival, and motility andmajor pathogenetic role in human prostate cancer (PCa). However, the oncogenic contributions down-of the PI3K pathway made by mammalian target of rapamycin complex 1 (mTORC1)–mediated cellsignal transduction in PCa have yet to be elucidated in detail. Here, we engineered constitutive mTORC1

tion in prostate epithelium by a conditional genetic deletion of tuberous sclerosis complex 1 (Tsc1), at negative regulator of mTORC1 signaling. Epithelial inactivation was not immediately tumorigenic,c1-deficient mice developed prostatic intraepithelial neoplasia (mPIN) in lateral and anterior prostatesonths of age, with increasing disease penetrance over time. Lateral prostate lesions in 16- to 22-month-tant mice progressed to two types of more advanced lesions, adenomatous gland forming lesion (Type 1)ypical glands embedded in massively expanded reactive stroma (Type 2). Both Type 1 and Type 2 lesionsned multiple foci of microinvasive carcinoma. Epithelial neoplastic and atypical stromal lesions persistede 4 weeks of RAD001 chemotherapy. Rapalogue resistance was not due to AKT or extracellular signal-ted kinase 1/2 activation. Expression of the homeobox gene Nkx3.1 was lost in Tsc1-deficient mPIN, andperated with TSC1 loss in mPIN initiation in doubly mutant Tsc1:Nkx3.1 prostatic epithelial knockout
it coo
mice. Thus, TSC1 inactivation distal to PI3K and AKT activation is sufficient to activate a molecular signalingcascade producing prostatic neoplasia and focal carcinogenesis. Cancer Res; 70(21); 8937–47. ©2010 AACR.

to 70%and mgene iadenoPI3K sAm

motesof the(mTOmTOR

duction

an prostate cancer (PCa) develops via multistage pro-on over a particularly long duration, associated withssive accumulation of mutations and epigenetic altera-n oncogenes and tumor suppressors. The phosphoino-3-kinase (PI3K) pathway, which plays a major role inalian cell growth, survival, and motility, has emergedntral regulator of human PCa. PCa PI3K pathway acti-occurs via complete loss or haploinsufficiency in the

is pathway, phosphatase and tensin homo-n chromosome 10 (PTEN), occurring in 20%

tuberodimerfor bo(GAP)RHEBribosoand pLike

sors. Oinherimesenhave band hhumaseemsthan P

: 1Division of Molecular Oncology, Department of Internalof Urologic Surgery, Department of Surgery, 3Siteman4Program in Cell Biology, Washington University Schoolis, Missouri; 5Center for Genomic Pathology, Universityis, California; 6Division of Translational Medicine,icine, Brigham and Women's Hospital, Harvard Medicalassachusetts; and 7Signal Transduction Program,edical Research Institute, La Jolla, California

ry data for this article are available at Cancer Researchrres.aacrjournals.org/).

thors:ZhiHongLuor JeffreyM.Arbeit, 660SouthEuclidox 8242, St. Louis, MO 63110. Phone: 314-362-3222;; E-mail: [email protected] or [email protected].

5472.CAN-10-1646

ssociation for Cancer Research.

ls.org

on April 7, 2020. © 2cancerres.aacrjournals.org from

of primary PCa and 30% to 80% of locally advancedetastatic disease (1–3). Conditional deletion of the Ptenn mouse prostate epithelium leads to invasive prostatecarcinoma (4), further supporting a key etiologic role ofignaling in prostate neoplasia.ong many proposed functions, the PI3K pathway pro-mammalian cell growth through constitutive activationAKT and mammalian target of rapamycin complex 1RC1) protein kinase signaling module. AKT activatesC1 in part through inhibitory phosphorylation ofus sclerosis complex 2 (TSC2; refs. 5–8). TSC2 hetero-izes with TSC1, and heterodimer formation is requiredth TSC2 protein stabilization and activation of GTPasefunction on the small G protein, RHEB (9–12). GTP-is a direct activator of mTORC1, which in turn promotesmal biogenesis, enhances protein and lipid synthesis,revents autophagy (reviewed in ref. 13).upstream PTEN, TSC1 and TSC2 are tumor suppres-riginally identified through their involvement in an

ted tumor syndrome affecting multiple soft tissues ofchymal origin (14), TSC1 and TSC2 genetic alterationseen identified in sporadic carcinomas of the bladderead and neck, implicating the genes as broad spectrumn tumor suppressors (15–18). Nonetheless, TSC1/TSC2
to be much less commonly involved in human cancerTEN or LKB1. There is limited data on mutation or
8937

010 American Association for Cancer Research.

http://cancerres.aacrjournals.org/

deletiosuggesditionby extphospvancedTSC1/(21). Atinct scauseof mTTo d

ogeneof prentransgthat eattainewas inHer

pathwnovelHomohyperformsboth alial neactivaand fa

Mate

MousAll a

AnimaTsc1 abeen dPB-Creepithe(mixedin a twTsc1fl/f

Tsc1fl/

i.e., pebetweline NNkx3.1therapwereby gav22-mowere r

TissueMal

perfusa stan(AP) atates

and emtransv5-μmobtain(Olymware (

ImmuSec

(pH7.NorthDeclocookeDilue(1:500Abcamtal Stucytokefrom JSignal(1:250ERK1/Ki67 (S6S235/

cubatewith Aand cmountrogena SoftBiologstaininprotoc

ImmuSna

munosodiumton X(1:50,itor Coμg) weand tMembTweenwashwash bylatedtein SAKTS4

1:1,000tibodywashemol/LbufferIgG-h

Kladney et al.

Cance8938

Dow


n of TSC1 or TSC2 in human PCa, but strong evidencets TSC1/2 functional inactivation in this disease. In ad-to AKT, TSC1/2 GAP activity toward RHEB is inhibitedracellular signal-regulated kinase 1/2 (ERK1/2) pathwayhorylation (19, 20). Moreover, loss of LKB1, seen in ad-PCa, produces reduced AMPK activity, diminution of

TSC2 suppression, and consequent mTORC1 activations AKT, ERK1/2, andAMPK all phosphorylate TSC2 at dis-ites (21, 22), multiple aberrant signaling pathways maycombinatorial TSC1/2 inactivation, enabling gradationsORC1 dysregulation during human PCa progression.ate, the direct effect of TSC1/2 inactivation on PCa path-sis has not been tested in a preclinical model. Inductioneoplastic hyperplasia without neoplastic progression inenic mouse prostates overexpressing RHEB suggestedither an insufficient level of mTORC1 activation wasd by RHEB overexpression, ormTORC1 activation alonesufficient to produce carcinogenesis (23).e, we tested the role of persistent constitutive mTORC1ay activation in prostate carcinogenesis using amouse model of TSC1 deletion in prostate epithelium.zygous Tsc1 loss of function and consequent mTORC1activation produced two histologic and molecularof spontaneous prostate carcinogenesis in aged mice,ssociated with precursor (mouse prostatic intraepithe-oplasia) mPIN lesions. Thus, TSC1/2-mTORC1 pathwaytion alone is sufficient to initiate prostate neoplasiacilitate malignant transformation.

rials and Methods

e experimentsnimal workwas approved by theWashingtonUniversityl Studies Committee. Mice harboring a loxP-flankedllele (Tsc1fl) and PB-Cre4 (PBCre) transgenic mice haveescribed previously (4, 14), and the expression of the4 transgene has been shown to be specific for prostatelial cells (4). PBCre males (FVB/n) and Tsc1fl/fl femalesBALB/c, C57BL/6, and 129/SvJae) were intercrossedo-step breeding strategy to generate the target PbCre+-l genotype. The mixed genetic background of the PbCre+-fl mice could underlie differences in tumor phenotype,netrance and time of onset of tumor developmenten our study and that of others in the literature. Germ-kx3.1 knockout mice used to generate Pb-Cre+Tscfl/flko/ko have been described previously (24). In the chemo-y trial, six PbCre+Tsc1fl/fl mice (16–22 month of age)treated with 7.5 μg/g of RAD001 (Novartis Corp.)age three times per week for 4 weeks. A cohort ofnth-old mice was castrated, and genitourinary tractsemoved 4 weeks later.

harvest and histologye genitourinary tracts were harvested from formalin-ed mice, fixed by microwave, and dissected followingdard procedure (25, 26). The isolated anterior prostates
nd urethra containing dorsal, lateral, and ventral pros-were processed through graded alcohols and xylenes
for 30times

r Res; 70(21) November 1, 2010

on April 7, 2020. © 2cancerres.aacrjournals.org nloaded from

bedded in paraffin so that coronal/sagittal (APs) anderse (urethra) sections were obtained. For histology,sections were stained with H&E, and images wereed with a DP70 color Bayer mosaic digital camerapus America) with MicroSuite Biological Suite v5 soft-Olympus Soft Imaging Solutions).

nofluorescence and immunohistochemistrytions were deparaffinized, rehydrated, washed in PBS4), andblockedwith serum-freeDakoProteinBlock (DakoAmerica). Antigen retrieval was performed in 1× Revealaker reagent (pH6.0; Biocare Medical) with a pressurer. Primary antibodies were diluted in Dako Antibodynt. Primary antibodies included rabbit anti-α-SMA, Abcam), rabbit anti-desmin (1:250 IF; 1:4,000 IHC,), rat anti-cytokeratin 8 (TROMA-1; 1:500, Developmen-dies Hybridoma Bank, University of Iowa), rabbit anti-ratin 5 (1:1,000, Covance), rabbit anti-Nkx3.1 (1:100, a gift. Milbrandt, WUSM), rabbit anti-E-cadherin (1:200, Celling Technology), rabbit anti-phosphorylated AKTS473

, Cell Signaling Technology), rabbit anti-phosphorylated2T202/Y204 (1:100, Cell Signaling Technology), rabbit anti-1:2,000, Vector Labs), and rabbit anti-phosphorylated236 (1:500, Cell Signaling Technology). The slides were in-d with corresponding secondary antibodies conjugatedlexa Fluor 488 and Alexa Fluor 594 (1:400; Invitrogen)ounterstained for nuclei with SlowFade Gold Antifadeting reagent with 4′,6-diamidino-2-phenylindole (Invi-). Fluorescence microscope images were collected usingImaging Solutions FVII digital camera with MicroSuiteical Suite software (Olympus). Immunohistochemicalg was carried out following our previously publishedol (27).

noblottingp-frozen prostate tissues were homogenized in radioim-precipitation assay buffer (1× PBS, 1% Nonidet-P40, 1%deoxycholate, 0.1% SDS, 1 mmol/L EDTA, and 1% Tri-

-100) supplemented with Protease Inhibitor CocktailSigma-Aldrich), 5 mmol/L NaF, and Phosphatase Inhib-cktail 1 and 2 (1:50, Sigma-Aldrich). Protein lysates (100re separated on polyacrylamide gels by electrophoresisransferred to polyvinylidene difluoride membranes.ranes were blocked in TBS (pH 7.6) containing 0.5%20 (TBST) and 10% nonfat dry milk, briefly washed in

buffer (0.5% nonfat dry milk in TBST) and incubated inuffer containing rabbit antibodies for TSC1, phosphor-S6KT389, S6K, phosphorylated S6S235/236, ribosomal pro-6, phosphorylated eIF4GS1108, eIF4G, phosphorylated73, AKT, phosphorylated ERK1/2T202/Y204, ERK1/2 (all, Cell Signaling Technology), and a rabbit polyclonal an-for β-tubulin (1:35,000, Abcam). Membranes wered once with high-salt buffer (TBST containing 0.5NaCl and 0.5% nonfat dry milk), two times with wash, and incubated in wash buffer with goat anti-rabbitorseradish peroxidase conjugate (1:1,000, Santa Cruz)
minutes. After one wash with high-salt buffer and threewith wash buffer, peroxidase activity was revealed by
Cancer Research



enhanand imtity On

StatisDat

signifStudenANOVGraph

Resu

Tsc1 iconstCon

tein leand ATsc1 Kbut we(LP, Ffrom 5of pho

phospactivaevidenFig. 1Auniforin conshowntary Fshownluminactivaevidenat this

AdultprostNec

nary t3 andwere nto 5 m

Figureattenua(P-S6)C, the imutantLow ma

Epithelial Tsc1 Inactivation Causes Prostate Cancer

www.a

Dow


ced chemiluminescence using ECL Plus (GEHealthcare)aged using a Chemidoc XRS imaging systemwith Quan-e one-dimensional analysis software (Bio-Rad).

tical analysisa were expressed as the mean ± SD, and statisticalicance was determined using two-tailed unpairedt's t test, Mann-Whitney U test, χ2 test, or one-wayA test followed by Bonferroni's multiple tests usingPad PRISM software (GraphPad Prism).

lts

nactivation in mouse prostate epithelium induceditutive mTORC1 activationsistent with conditional gene inactivation, TSC1 pro-vels were drastically reduced in lateral prostates (LP)Ps of 5-month-old PbCre+Tsc1fl/fl mice (designated asO) compared with Tsc1fl/fl (wild-type, WT) littermatesre similar to WT in heart, lung, liver, spleen, and testisig. 1A; AP and nonprostate, data not shown). Extracts
-month-old Tsc1 KO prostates exhibited elevated levelssphorylated S6S235/236, phosphorylated eIF4GS1108, and
logic amTOR

mice aged. D, H&E staining of normal AP and LP glands of a 17-mo-old WT mognification views of the same fields are shown in Supplementary Fig. S1A and B

acrjournals.org


horylated S6KT389, consistent with mTORC1 pathwaytion, and diminution of phosphorylated AKTS473,ce for concomitant mTORC2-negative feedback (LP,; AP, data not shown). Phosphorylated S6S235/236 wasmly distributed in each TSC1 KO luminal epithelial cell,trast to sporadic activation in WT counterparts asby cytokeratin 8 (keratin-8) colocalization (Supplemen-ig. S1A; high-magnification views of the same fields arein Fig. 1B), indicating that Tsc1 inactivation in prostateal cells led to cell autonomous constitutive mTORC1tion. Notably, there was neither histopathology norce of mTORC1 activation in the stroma of Tsc1 KOmiceage (see below).

Tsc1 KO mice developed spontaneousate carcinogenesisropsy and histopathologic examination of genitouri-racts were performed on adult Tsc1 KO mice between22 months of age (n = 45). Age-matchedWT littermatesegative controls (n = 27). Prostates of Tsc1 KO mice uponths of age did not show any morphologic or histo-
bnormality (Fig. 1C, 3–5 months, n = 8) despite chronicC1 activation. At 6 months of age, mutant mice began
1. Adult Tsc1 KO mice develop spontaneous PCa. A, loss of TSC1 protein, mTORC1 pathway activation, and concomitant mTORC2 signalingtion in LPs of 5-mo-old Tsc1 KO mice (n = 3) compared with WT littermate controls (n = 2). B, uniform distribution of phosphorylated S6S235/236

immunofluorescence in K8(+) LP luminal epithelium of a 4-mo-old Tsc1 KO mouse compared with sporadic expression in a WT counterpart.ncidence of mPIN increased in APs and LPs of Tsc1 KO mice over time, and the incidence of Type 1 and Type 2 PCa also increased in LPs as

use and AP and LP mPIN lesions of 15-mo-old Tsc1 KO mice.. Bars, 50 μm (B) and 100 μm (D).

Cancer Res; 70(21) November 1, 2010 8939



to devwith dn = 14monthTsc1 Kconsisatypia(Suppsamemice dcysticanalys(Fig. 2LP t

pathotwo tymulatpattermorsinto resels (dcancerof agecells wders astrommalitihypertpromineath(insets

ponenType(Fig. 2Type 222 mo

ProgrprostTo

carcinprolif(Fig. 3was eevidenBoth A(4.1–4restrictop leand dtive ceType 1maligneratio2 canc

EarlytransprostNex

Kladney et al.

Cance8940

Dow


elop mPIN in LP and AP sites (6–11 months, n = 11)isease penetrance increasing over time (12–15 months,), reaching 100% in APs and 75% in LPs in the 16- to 22--old group (n = 12; henceforth called the aged group).O AP and LP mPIN lesions were often multifocalting of markedly enlarged epithelial cells with nuclearthat partially or completely filled the glandular lumenlementary Fig. S1B; high-magnification views of thefields are shown in Fig. 1D). Seven of 12 aged mutanteveloped visible LP enlargement with two also showingAP dilation (Supplementary Fig. S1C and D). Furtheres focused on the LP because those lesions progressedA–D) in contrast to mPIN persistence in the AP.umor histology revealed complex, multicompartmentallogy (Supplementary Fig. S1D; Fig. 2A–D). There werepes of advanced lesions. One contained a marked accu-ion of adenomatous glands, designated as Type 1 tumorn (Supplementary Fig. S1D, left; Fig. 2A). As Type 1 tu-contained multiple foci of epithelial cell microinvasionactive stroma (Fig. 2B, yellow arrowheads) or microves-ata not shown), these lesions were designated as Type 1s, which were detected in 4 of 12 mice of 16–22 months(33%). Type 2 lesions consisted of neoplastic epithelialith nuclear atypia arranged in glands with irregular bor-nd embedded in a massive accumulation of hyperplastica (Supplementary Fig. S1D, right; Fig. 2C). Stromal abnor-es possessed two components. One was elongated androphic fibroblasts with nuclear enlargement, nucleolarnence, and frequent mitotic figures, localized just be-
the basement membrane of atypical neoplastic glands cers w


t contained eosinophilic mature fibroblasts (Fig. 2C).2 tumors also contained focal epithelial cell invasionD, yellow arrowheads) and, as such, were designatedcancers, which were detected in 6 of 12 mice of 16 to

nths of age (50%).

essive proliferative activation during Tsc1 KOate carcinogenesisfurther determine mechanisms of Tsc1 multistageogenesis, we determined the level and localization oferating cells in knockout prostates during aging). As expected, the Ki67 index in normal prostatesxtremely low (1–2%). Preneoplastic Tsc1 KO prostatesced equivalently low proliferation (data not shown).P and LP mPIN lesions displayed significant elevation.7%) compared with WT prostates, with Ki67(+) cellsted to neoplastic epithelium (Fig. 3A, top right versusft). There was a differential progression in Ki67 indexisparate tissue compartmental localization of prolifera-lls between Type 1 (9.3%) and Type 2 (16.3%) cancers.proliferating cells were restricted to the neoplastic andant epithelial cells, whereas there was increased prolif-n in both stroma and neoplastic epithelium of Typeers (Fig. 3A, bottom left versus bottom right).

-stage epithelial dysplasia, malignantformation, and stromal remodeling in Tsc1 KOate carcinogenesist, we interrogated mPIN and Type 1 and Type 2 can-
ith a panel of antibodies specific for prostate epithe-
(keratin-5, keratin-8, and

ureeloproinma15wednomndugnifthelioplalowinophistoor ouse.s o

ndlobe

mbrndsds).tainsio, 500 μm (A and C).

of Fig. 2D, green arrowheads). The second stromal com- lial differentiation dysregulation

Figdevmicof aof ashoadeglamaepicyt(yeleosC,tummoviewspijustmeglaheaconinvaBar

010 American Association for

2. Aged Tsc1 KO miceed two types ofvasive PCa. A, histologyssive Type 1 LP tumor-mo-old Tsc1 KO mousemultiple enlargedatous glands and cystic

lar dilatation. B, higherication views of A revealedal cells with enlargedsm and nuclear atypiaarrowheads) invadinghilic reactive stroma.logy of a huge Type 2 LPf a 16-mo-old Tsc1 KOD, higher magnificationf C revealed enlargedid cells with nuclear atypianeath the basementane of atypical neoplastic(insets of D, green arrowType 2 tumors alsoed multiple foci of epithelialn (yellow arrowheads).

Cancer Research

Cancer Research.


E-cadhstromactin;marketate dutin-5 edistriblost inof neonancielial celepithewas reType 2data ntheseunifor

Fig. S2originby duaphorytumorthe mstrommusclneoplaSimiladistribwas abType 2test thfromor absmesenof zinrepreeffortsepithedetectof ma

Tsc1 KrespoTo

the pr(16–22hibitoof rapmTORS4, P-a promtary Fthe prwas pwith astromMorphthe prstromwas nplyingresistaenlargfibrosilesionwill beInh

mediawhichDiscuswas sby immedia

FigurecarcinoWT LP(+) cellsrestricteshowedB, progcarcinoType 2


www.a

Dow


erin), mTORC1 activation (phosphorylated S6), andal remodeling/activation (desmin and smooth muscleFig. 4; Supplementary Fig. S2). Keratin-5, a basal cellr, was circumferentially localized around eachWT pros-ctule (Fig. 4; Supplementary Fig. S2, K5/K8, WT). Kera-xpression was markedly attenuated and sporadicallyuted within neoplastic cells of Tsc1 KO mPIN lesions,Type 1 cancers, but sporadically upregulated in clustersplastic and malignant epithelial cells of Type 2 malig-s (Fig. 4; Supplementary Fig. S2). E-cadherin, an epithe-l-to-cell junctionmolecule, was uniformly present inWTlia cells as well as in Tsc1 KO mPIN neoplastic cells butduced or absent in expression in subsets of Type 1 andcancer cells (Fig. 4; Supplementary Fig. S2, E-CAD/K8;ot shown) consistent with the invasive phenotype of

genesis (WT, 1–2%; mPIN, 4.1–4.7%; Type 1, 9.3%;, 16.3%). Bar, 100 μm (A).

cells. The mPIN and Type 1 malignant epithelial cellsmly exhibitedmTORC1 activation (Fig. 4; Supplementary

aged m2-mon

acrjournals.org


, P-S6/K8), indicating a Tsc1-deficient luminal epithelial. Type 1 malignant microinvasion was strikingly evidentl phosphorylated S6 and K8 labeling. In contrast, phos-lated S6 was markedly elevated in stromal cells of Type 2s with areas of microinvasive K8(+) cells also expressingTORC1 marker (Supplementary Fig. S3). mPIN inducedal remodeling evidenced by loss of desmin andα-smoothe actin expression in the fibromuscular layer encirclingstic glands (Fig. 4,Tsc1 KOmPIN, DES/K8, and SMA/K8).r to phosphorylated S6, tissue compartmental desminution distinguished Type 1 and Type 2 cancers; desminsent in Type 1 lesions but was markedly upregulated incancer stroma (Fig. 4 and Supplementary Fig. S2). Toe possibility that the Type 2 stromal cells were derivedsubsets of Type 2 microinvasive cells with reducedent E-cadherin expression through an epithelial-to-chymal transition (EMT), we investigated expressionc finger protein Snail, an E-cadherin transcriptionalssor and a potent EMT inducer. Despite extensive, we could not detect Snail upregulation in invadinglial cells or in desmin (+) stromal cells, nor could wekeratin-8/desmin colocalization in mPIN or either typelignant tumor (data not shown).

O prostate lesions showed partial therapeuticnses to mTORC1 inhibitor RAD001determine whether mTORC1 signaling was required forostate disease maintenance, six aged Tsc1 KO micemonths) were treated for 4 weeks with mTORC1 in-

r RAD001 (everolimus), an orally bioavailable derivativeamycin. As expected, the treatment produced a markedC1 pathway inhibition in testes (Supplementary Fig.S6) and prostates (Fig. 5A, P-S6/K8). RAD001 inducedinent antiproliferative effect in the testis (Supplemen-ig. S3, Ki67) but had no effect on the proliferation inostate lesions (Fig. 5A, Ki67/K8; data not shown). mPINersistent in both the APs (six of six; Fig. 5B, AP) and LPssubset of LP lesions containing additional desmin (+)a (five of six and three of six, respectively; Fig. 5B, LP).ologically, RAD001 decreased but did not eliminateevalence of atypical giant cell epithelium in mPIN oral lesions, whereas microinvasive adenocarcinomaotably undetectable in LPs of the treated mice, im-that mPIN and Type 2 stromal lesions were morent to RAD001 than microinvasion. The regression ofed mPIN glands was often accompanied by stromals (Fig. 5B). Whereas the rapamycin resistance of theses is compelling, an expanded study with larger cohortsneeded for adequate testing of this hypothesis.

ibition of mTORC1/S6K pathway may abrogate S6K-ted negative feedback on AKT and ERK1/2 activation,has been linked to tumor cell rapalogue resistance (seesion). However, phosphorylation of AKT and ERK1/2imilar between 15-month-old WT and Tsc1 KO micemunoblotting (Fig. 5C), indicating that mTORC1/S6K-ted negative feedback was not elicited in prostates of

3. Progressive proliferative activation during Tsc1 KO prostategenesis. A, immunofluorescence analysis of Ki67 expression inglands, Tsc1 KO mPIN, and Type 1 and Type 2 PCa lesions. Ki67were localized to neoplastic epithelium in AP and LP mPINs,d to the Type 1 neoplastic and malignant epithelial cells, butbi-compartmental proliferative induction in Type 2 cancers.ressive increase in Ki67 index during Tsc1 KO prostate

ice. As a positive AKT activation control, LPs from ath-old PbCre+PTENfl/fl (Pten KO) mouse, as expected,




exhibi(Fig. 5RAD0low-bP-AKTsporadmatorstromLP UnRAD0phoryLP, RAthat acfor the

Loss ocarcininitiaAs

proteand mcoopeesis (prostacells swheredetect

Figureprogresand nemalignamTORC h muscglands ated in

Kladney et al.

Cance8942

Dow


ted marked phosphorylated AKT immunofluorescenceD, P-AKT, Pten KO LP). In contrast, both untreated and01-treated Tsc1 KO LP tumor cells showed similarlyackground phosphorylated AKT signal (Fig. 5D,, Tsc1 KO LP, RAD001 versus Untreated). In addition,ic ERK1/2 activation was detected in stroma inflam-y cells of aged WT LPs and in both epithelial andal cells in Tsc1 KO mPIN lesions (Fig. 5D, P-ERK, WTtreated versus Tsc1 KO LP Untreated). Importantly,01 treatment did not change the levels of ERK phos-lation in Tsc1 KO prostate tumors (Fig. 5D, Tsc1 KOD001 versus Untreated). Together, our data suggested

1 marker. Early-stage mPIN lesions progressively lost desmin and α-smoot(arrows). Desmin was absent in Type 1 lesions but was markedly upregul

tivation of either mTORC2 or ERK was not responsibleRAD001 resistance of Tsc1 KO prostate tumors.

was pcells b



f Nkx3.1 protein expression occurred in Tsc1 KOogenesis and played a cooperative role in thetion of mPINthe prostate-specific tumor suppressor, homeoboxin NKX3.1, is frequently downregulated in humanouse PCas and loss of a single Nkx3.1 allele canrate with Pten loss of function in prostate carcinogen-28), we analyzed its protein expression in Tsc1 KOte carcinogenesis. As expected, WT prostate luminalhowed abundant nuclear NKX3.1 immunoreactivity,as age-matched Nkx3.1 KO counterparts lackedable staining (Fig. 6A, WT and Nkx3.1 KO). NKX3.1

le actin expression in the fibromuscular layer encircling neoplasticType 2 cancer stroma. Bar, 100 μm.

4. Epithelial and stromal marker alterations during Tsc1 KO prostate carcinogenesis. Progressive loss of keratin-5 (K5) expression during Tsc1 KOsion to Type 1 microinvasive cancer with stochastic focal reexpression in Type 2 tumors. E-cadherin was uniformly present in WT epithelial cellsoplastic cells of Tsc1 KO mPINs, reduced or absent in expression in a subgroup of Type 1 and Type 2 malignant cells. Neoplastic and Type 1nt cells uniformly exhibited mTORC1 activation (P-S6), whereas stromal cells and microinvasive K8(+) cells of Type 2 tumors expressed the

resent in morphologically normal Tsc1 KO luminalut undetectable in all mPINs and carcinomas (Tsc1

Cancer Research



KO mwas asTsc1 KmunolesionNKX3RAD00that logrowthTo t

cinogemice hKO bawith Npropodevelo

eithermice.mice (displa5 monhighermiceKO (z7–11 mteristithe DKvelopecountment

FiguretreatedC, LPsPbCre+

ERK1/2B, bars


www.a

Dow


PIN, Fig. 6A; PCa, data not shown). Thus, NKX3.1 losssociated with the onset of mPIN, a precursor event inO prostate carcinogenesis. Remarkably, NKX3.1 im-reactivity remained undetectable in RAD001-treateds (Fig. 6A, Tsc1 KO LP, RAD001), showing that.1 loss in neoplastic cells was not regulated by1-mediated mTORC1 inactivation. Thus, it is possibless of NKX3.1 may provide the mTORC1-independentsignal to Tsc1 KO tumors treated with RAD001.

est a functional role for NKX3.1 in Tsc1 KO prostate car-nesis, disease incidence and severity were tested inomozygous for a germline Nkx3.1 deletion in the Tsc1ckground (designated as DKO for double KO) comparedkx3.1 and Tsc1 single KO mice. A significantly higher
rtion of 4- to 5-month-old DKO mice (eight of nine) loss-o
(P-ERK, Tsc1 KO LP untreated versus RAD001) in Tsc1 KO prostate mPIN lesion, 200 μm (left and left middle panels) and 100 μm (right middle and right panels).

acrjournals.org


Nkx3.1 (three of seven) or Tsc1 single KO (zero of seven)Unlike the low-grade mPINs exhibited by the Nkx3.1 KOFig. 6C, 5 months, AP, Nkx3.1 KO, inset), DKOmPIN cellsyed enlarged cytoplasm and nuclear atypia (Fig. 6C,ths, AP, DKO, inset). Furthermore, a significantlyproportion (six of seven) of 7- to 8-month-old DKO

developed LP mPIN lesions than age-matched Nkx3.1ero of five) and Tsc1KO (one of five) mice (Fig. 6B,onths, LP). DKO LP mPIN contained Tsc1 KO charac-

c atypical cells (Fig. 6C, 8 months, LP, DKO, inset). AsO mice further aged, Type 1 and Type 2 cancers de-d, albeit at similar frequencies compared with Tsc1 KOerparts with no evidence of metastatic disease develop-(data not shown). Thus, combinatorial Nkx3.1 and Tsc1
f-function accelerated prostatic neoplastic initiation
ped mPIN in APs (Fig. 6B, 4–5 months, AP) than did but failed to affect malignant conversion.

5. Prostate lesion persistence in RAD001-treated Tsc1 KO mice. A, marked reduction in P-S6 immunoreactivity in Tsc1 KO prostate AP glandswith 4 wk RAD001, without significant effect on Ki67 index. B, mPIN and stromal hyperplastic persistence in RAD001-treated Tsc1 KO prostates.of 15-mo-old WT (n = 2) and Tsc1 KO (n = 3) mice displayed similar level of AKT and ERK1/2 phosphorylation. D, the LP glands of a 4-mo-oldPTENfl/fl mouse exhibited drastically activated AKT (left). RAD001 treatment did not activate AKT (P-AKT, Tsc1 KO LP untreated versus RAD001) or

s. A, bars, 200 μm (left two panels) and 100 μm (right two panels).D, bar, 100 μm.




Tsc1 KandroThe

therapde novdepenthreewas chistoplesion(SupplesionFig. Smarkegressiolocalimarkeout, aexpreswith npersis

mal mthe prbeforegest,was pconda

Discu

HermiceTsc1 dtate eneopla proNKX3of thistion ifuncti

Figurecells shmorphoRAD00KO, TsWT, Nkx3.1 KO, Tsc1 KO, and DKO mouse. Insets are a higher (40×) magnification view of the images. Arrows, presence of markedly enlarged epithelialcells w . Bars,

Kladney et al.

Cance8944

Dow


O lesions display differential cell typegen dependencyresistance of mPIN lesions during chronic RAD001y and the persistent loss of NKX3.1 expression ino and RAD001-treated mice led us to explore androgendency of Tsc1 KO prostate neoplasias. A small group ofTsc1 KO, one Tsc1:Nkx3.1 DKO, and one control mouseastrated, and genitourinary tracts were analyzed byathology and immunofluorescence. Multiple mPINs persisted in both the lateral and anterior prostateslementary Fig. S5; Supplementary Table S1). Theses also retained mTORC1 activation (Supplementary5). Type 2 lesions, confined to the LP, underwentd regression (Supplementary Fig. S5, Type 2 stroma re-n). Even massive lesions evidenced extensive necrosiszed to in the stroma. High power views revealedd inflammatory cell infiltration, stromal cell dropnd fibrosis (Supplementary Fig. S5). Castrate desminsion response was heterogeneous, sporadic in lesions

ith nuclear atypia in the early phase of mPIN formation in the DKO mouse

ecrosis and stromal infiltration, and diffuse in a smalltent tumor that also displayed both epithelial and stro-

predonounc



TORC1 hyperactivation. These data, along with bothogression kinetics wherein epithelial mPIN developedstromal alterations and the mPIN persistence, sug-

but do not prove, that Tsc1 KO epithelial neoplasiaredominant and that stroma lesions were likely se-ry to chronic epithelial activation.

ssion

e, we showed multistage prostate carcinogenesis inwith prostate luminal cell Tsc1 deletion. As expected,eletion produced mTORC1 activation in luminal pros-pithelial cells of young male mice. However, prostateasia development and microinvasive cancer requiredlonged latency. mPIN onset was associated with.1 loss of expression, and the functional importancesecondary event was supported by disease accelera-

n Nkx3.1:Tsc1 DKO mice. Interestingly, TSC1 loss-of-on produced two distinct histologic malignancies, one

100 μm (A) and 200 μm (C).

6. Loss of Nkx3.1 protein expression in Tsc1 KO mPIN with cooperative neoplasia induction in double knockout mice. A, WT prostate luminalowed uniform NKX3.1 nuclear expression, whereas age-matched Nkx3.1 KO counterparts lacked detectable staining. NKX3.1 was present inlogically normal Tsc1 KO luminal cells (arrowheads) but undetectable in all mPINs and carcinomas (arrows). NKX3.1 remained undetectable in1-treated lesions (RAD001). B, incidence of mPIN in APs of 4- to 5-mo-old Nkx3.1 KO, Tsc1 KO, and DKO mice and in LPs of 7- to 11-mo-old Nkx3.1c1 KO, and DKO mice. C, histology of AP glands of 5-mo-old WT, Nkx3.1 KO, Tsc1 KO, and DKO mouse. Histology of LP glands of 8-mo-old

minantly glandular and the other associated with pro-ed stromal activation and hyperplasia. These data

Cancer Research



show tprogrethe kiPI3K

sivelycondittion afunctiway lebialleltivelyalbeitFVB gthat AtivitiesterminpathwNon

AKT, iprostaa singal TSCalso sTSC1/hyperp(23). Wthe RHto fulRHEBThe

Tsc1 Kary geWe shsion ashowetate tuprostasomesubseexprestion dhumanmousenormabeen sprostaFurtheand pPten+/

tiated(37). Athe lacin Nkxwith tPten lin turmechand lo

AnoationmPINtive, imtive stPCa-e(38). Crived flular menhanPCa-ation wexperimaligntheliaconsiseratedtwo dhyperpcularin Typbeendeed,in retigain oalso pthrougpressiand wEMTcells oFurthemechaAdd

in Tsccontra2 weealternausingcausedS6K-PERK1/tion (2PI3K-ATsc1 Kto AKcomplTSC

in thaseen ituresThis ifrom lvatedtivatioleads


www.a

Dow


hat TSC1 loss-of-function produces a form of neoplasticssion and malignant conversion remarkably similar tonetics and molecular biology of human disease.-mTORC1 pathway perturbations have been exten-

investigated in prostate luminal epithelial cells throughional PTEN (4, 29) or TSC1 (current study) loss of func-nd through targeted AKT (30) or RHEB (23) gain ofon, offering the opportunity for a comparison of path-vel perturbation on PCa pathogenesis. Whereas Ptenic deletion caused invasive PCa, expression of constitu-activated myristoylated AKT (MPAKT) produced mPINwith a more severe phenotype in the C57BL/6 versusenetic background (31). Consistent with the notionKT activation alone did not possess all malignant ac-of PTEN inactivation, PTEN loss activated c-Jun NH2

al kinase along with AKT in human PCa and the twoays synergized in cell transformation (32).etheless, given the many downstream substrates oft is surprising that AKT activation had a more benignte phenotype than loss of TSC1, which represents onlyle AKT target. This may reflect recruitment of addition-1/TSC2 functions in response to AKT activation. It isurprising that RHEB overexpression, the canonicalTSC2 target, produced only preneoplastic epitheliallasia in mice, despite constitutive mTORC1 activationhether the absence of malignant transformation inEB model was due to insufficient transgene expressionly activate mTORC1 or was indicative of potential-independent functions of TSC1/2 is uncertain.stochastic and delayed appearance ofmPIN in prostaticOmice suggests a requirement for spontaneous second-netic or epigenetic alterations for neoplastic induction.owed a tight correlation between loss of NKX3.1 expres-nd disease progression in the TSC1 model and alsod by genetic means that loss of NKX3.1 enhanced pros-morigenesis consequent to TSC1 deletion. NKX3.1 is ate-specific homeobox gene located on human chromo-8p21, a region undergoing monoallelic deletion in at of PIN and PCa (33). Additionally, NKX3.1 proteinsion is progressively lost through epigenetic inactiva-uring progression of human PCa found in ∼70% ofPCa metastatic cancers (34) and as well as in severalPCa models (35, 36). The critical role of NKX3.1 inl prostate development and tumor suppression hashown by Nkx3.1 deletion in mice resulting in defectivete epithelial differentiation and mPIN formation (33).rmore, Nkx3.1 germline loss increased PCa incidenceromoted androgen independence and metastasis in− mice, mainly through AKT and ERK1/2 activation ini-by PTEN inactivation and synergized by NKX3.1 lossbsence of AKT or ERK1/2 activation could underliek of frank local cancer induction or metastatic spread3.1:Tsc1 DKOmice. Nonetheless, our data are consistenthe notion that PI3K pathway dysregulation throughoss downregulates Nkx3.1 (35) and NKX3.1 deficiencyn promotes prostate tumorigenesis through multiple
anisms, including AR transcriptional dysregulation,ss of p53 function (35).
progrepatien

acrjournals.org


ther notable feature seen in Tsc1 KO mice was alter-in stromal cell terminal differentiation adjacent toand the marked subepithelial activation of prolifera-mature, stromal fibroblasts in Type 2 PCa. The reac-roma in Type 2 lesions also has clinical relevance aslicited reactive stroma has been described in humansarcinoma-stromal interactions are thought to be de-rom “dysregulated” tissue repair consisting of extracel-atrix accumulation, activation of myofibroblasts, andced fibroblastic collagen synthesis. The critical role ofssociated stroma in tumor cell malignant transforma-as clearly shown by a series of tissue recombinationments wherein PCa-associated stromal cells producedant transformation of nontumorigenic prostate epi-

l cells (ref. 38, and reference therein). Our data weretent with the notion that the mPIN/cancer lesions gen-two types of paracrine signals activating the stroma inistinct ways. Thus, epithelial neoplasia produced eitherlastic, immature stroma of Type 2 cancer or fibromus-layer dedifferentiation without hyperplastic expansione 1 cancers. Type 2 hyperplastic stroma could havederived from epithelial cancer cells through EMT. In-PCa-associated stromal abnormalities were describednoblastoma (pRb) knockout (39) and transgenic Fgfr1f function PCa models (40). Atypical stromal cells wereroduced following p53 inactivation in the pRb KO orh EMT in the Fgfr1 model. However, FGFR1 or p53 ex-on was unchanged in Tsc1 KO mice (data not shown),e could not detect convincing molecular evidence forvia Snail induction in Tsc1 KO prostatic epithelialr in underlying reactive fibroblasts (data not shown).r study will ultimately determine potential molecularnisms for the reactive stroma induction.itionally, bothmPIN lesions and reactive stroma persisted1 KO mice despite 4 weeks of RAD001 treatment. Inst, MPAKT mPIN was efficiently reversed followingks of RAD001 (30). As such, Tsc1 KO lesions manifestedtive proliferation and survival signals. In preclinical trialsdouble heterozygous Pten:Nkx3.1 KO mice, rapamycinERK1/2 activation through a signaling loop involving

I3K-ERK1/2 (41), with inhibition of both mTORC1 and2 producing combinatorial prostate tumor growth inhibi-0). Likewise, AKT activation by rapamycin through S6K-KT signaling was also documented (42–44). However,O prostate lesion persistence was surprisingly not dueT or ERK1/2 activation, thus revealing a further level ofexity in the PCa responses to rapamycin.is a relatively unique tumor suppressor gene syndrome,

t although many tumors in different sites are commonlyn TSC patients, progression to classic pathologic fea-of malignancy is quite rare (lifetime risk, <5%; ref. 14).s likely due in part to feedback pathways emanatingoss of the functional TSC1/TSC2 complex through acti-mTORC1 that lead to suppression of PI3K and AKT ac-n (42–44). In TSC mouse models, loss of TSC1 or TSC2to renal carcinoma through a cyst-adenoma-carcinoma
ssive series (45–47), but this development is rare in TSCts. This is the first epithelial tissue outside of the kidney



in whito fullOur daaboveIn c

plex fuprostaTSC1expanopmentumorcompotivity(revieTsc1 Kpotentreatin

Discl

No p

Ackn

WeRAD001

Grant

NIHgrant 1P

Theof pageaccorda

Refe1. Mc

loscan

2. Sirassref

3. YoPTpriep

4. WaPteCa

5. TeeJ.getdo

6. Inoinh20

7. Inocon

8. Matiopropa

9. ChbitJ B

10. TeerossigRh

11. Gaa m2.

12. InoGA182

13. La20

14. Kwand

15. PlaMAtio

Kladney et al.

Cance8946

Dow


ch TSC1 or TSC2 loss alone has been shown to progresscarcinoma without other engineered genetic events.ta suggest that loss of Nkx3.1 expression, as discussed, seems to be the key step in this progression.onclusion, we show that the TSC1/TSC2 protein com-nctions as a critical cell growth checkpoint in mousete luminal epithelial cells, such that isolated loss ofleads to spontaneous PCa formation. This observationds our concepts of PI3K pathway function in PCa devel-t but generates new questions about the differentialigenic phenotypes produced by downstream pathwaynent dysregulation. The incomplete rapalogue sensi-of mTORC1 (48), the parallel signaling of mTORC2wed in ref. 49), and the partial rapalogue efficacy inO prostate neoplasia motivate further study of more
t mTOR catalytic site or alternative PI3K inhibitors in Rece
tt FM, Hurst CD, Taylor CF, Gregory WM, Harnden P, Knowles. Spectrum of phosphatidylinositol 3-kinase pathway gene altera-ns in bladder cancer. Clin Cancer Res 2009;15:6008–17.

16. Pytumfun

17. ChofmTCa

18. HeHyare

19. ChKR11

20. KinmTca

21. Pede20

22. Maphfor17

23. NamTon

24. Abin a20

25. Lufacviru

26. Suouma

27. LucinanRe

28. KimlosNa

29. Mavaacred



osure of Potential Conflicts of Interest

otential conflicts of interest were disclosed.

owledgments

thank Rebecca Sohn for mouse husbandry, Novartis for the gift of, and Leonard B. Maggi, Jr., for the critical reading of the manuscript.

Support

grants R01 CA10101 and R01 CA120436, National Cancer Institute01CA120964-01A1, and the Beatrice Roe fund for Urological Research.costs of publication of this article were defrayed in part by the paymentcharges. This article must therefore be hereby marked advertisement innce with 18 U.S.C. Section 1734 solely to indicate this fact.

ived 05/10/2010; revised 08/16/2010; accepted 09/01/2010; published
g PCa with PI3K pathway a ctivation (50). OnlineFirst 10/12/2010.
rencesCall P, Witton CJ, Grimsley S, Nielsen KV, Edwards J. Is PTENs associated with clinical outcome measures in human prostatecer? Br J Cancer 2008;99:1296–301.car K, Yoshimoto M, Monzon FA, et al. PTEN genomic deletion isociated with p-Akt and AR signalling in poorer outcome, hormoneractory prostate cancer. J Pathol 2009;218:505–13.shimoto M, Cutz JC, Nuin PA, et al. Interphase FISH analysis ofEN in histologic sections shows genomic deletions in 68% ofmary prostate cancer and 23% of high-grade prostatic intra-ithelial neoplasias. Cancer Genet Cytogenet 2006;169:128–37.ng S, Gao J, Lei Q, et al. Prostate-specific deletion of the murinen tumor suppressor gene leads to metastatic prostate cancer.ncer Cell 2003;4:209–21.AR, Fingar DC, Manning BD, Kwiatkowski DJ, Cantley LC, Blenis

Tuberous sclerosis complex-1 and -2 gene products function to-her to inhibit mammalian target of rapamycin (mTOR)-mediatedwnstream signaling. Proc Natl Acad Sci U S A 2002;99:13571–6.ki K, Li Y, Zhu T, Wu J, Guan KL. TSC2 is phosphorylated andibited by Akt and suppresses mTOR signalling. Nat Cell Biol02;4:648–57.ki K, Zhu T, Guan KL. TSC2 mediates cellular energy response totrol cell growth and survival. Cell 2003;115:577–90.nning BD, Tee AR, Logsdon MN, Blenis J, Cantley LC. Identifica-n of the tuberous sclerosis complex-2 tumor suppressor geneduct tuberin as a target of the phosphoinositide 3-kinase/aktthway. Mol Cell 2002;10:151–62.ong-Kopera H, Inoki K, Li Y, et al. TSC1 stabilizes TSC2 by inhi-ing the interaction between TSC2 and the HERC1 ubiquitin ligase.iol Chem 2006;281:8313–6.AR, Manning BD, Roux PP, Cantley LC, Blenis J. Tuberous scle-

is complex gene products, Tuberin and Hamartin, control mTORnaling by acting as a GTPase-activating protein complex towardeb. Curr Biol 2003;13:1259–68.rami A, Zwartkruis FJ, Nobukuni T, et al. Insulin activation of Rheb,ediator of mTOR/S6K/4E-BP signaling, is inhibited by TSC1 and

Mol Cell 2003;11:1457–66.ki K, Li Y, Xu T, Guan KL. Rheb GTPase is a direct target of TSC2P activity and regulates mTOR signaling. Genes Dev 2003;17:9–34.

plante M, Sabatini DM. mTOR signaling at a glance. J Cell Sci09;122:3589–94.iatkowski DJ. Animal models of lymphangioleiomyomatosis (LAM)tuberous sclerosis complex (TSC). Lymphat Res Biol 2010;8:51–7.

mar LS, Platt FM, Askham JM, Morrison EE, Knowles MA. Bladderour-derived somatic TSC1 missense mutations cause loss ofction via distinct mechanisms. Hum Mol Genet 2008;17:2006–17.akraborty S, Mohiyuddin SM, Gopinath KS, Kumar A. InvolvementTSC genes and differential expression of other members of theOR signaling pathway in oral squamous cell carcinoma. BMCncer 2008;8:163.bert C, Norris K, Parashar P, Ord RA, Nikitakis NG, Sauk JJ.poxia-inducible factor-1α polymorphisms and TSC1/2 mutationscomplementary in head and neck cancers. Mol Cancer 2006;5:3.o NY, Choi M, Kim BH, Cho YM, Moon KC, Kang GH. BRAF andAS mutations in prostatic adenocarcinoma. Int J Cancer 2006;9:1858–62.kade CW, Castillo-Martin M, Puzio-Kuter A, et al. Targeting AKT/OR, ERK MAPK. signaling inhibits hormone-refractory prostatencer in a preclinical mouse model. J Clin Invest 2008;118:3051–64.arson HB, McCarthy A, Collins CM, Ashworth A, Clarke AR. Lkb1ficiency causes prostate neoplasia in the mouse. Cancer Res08;68:2223–32.L, Chen Z, Erdjument-Bromage H, Tempst P, Pandolfi PP. Phos-

orylation and functional inactivation of TSC2 by Erk implicationstuberous sclerosis and cancer pathogenesis. Cell 2005;121:

9–93.rdella C, Chen Z, Salmena L, et al. Aberrant Rheb-mediatedORC1 activation and Pten haploinsufficiency are cooperativecogenic events. Genes Dev 2008;22:2172–7.dulkadir SA, Magee JA, Peters TJ, et al. Conditional loss of Nkx3.1dult mice induces prostatic intraepithelial neoplasia. Mol Cell Biol02;22:1495–503.ZH, Wright JD, Belt B, Cardiff RD, Arbeit JM. Hypoxia-inducibletor-1 facilitates cervical cancer progression in human papilloma-s type 16 transgenic mice. Am J Pathol 2007;171:667–81.wa T, Nyska A, Haseman JK, Mahler JF, Maronpot RR. Spontane-s lesions in control B6C3F1 mice and recommended sectioning ofle accessory sex organs. Toxicol Pathol 2002;30:228–34.ZH, Shvartsman MB, Lee AY, et al. Mammalian target of rapamy-activator RHEB is frequently overexpressed in human carcinomasd is critical and sufficient for skin epithelial carcinogenesis. Cancers 2010;70:3287–98.MJ, Cardiff RD, Desai N, et al. Cooperativity of Nkx3.1 and Pten

s of function in a mouse model of prostate carcinogenesis. Proctl Acad Sci U S A 2002;99:2884–9.X, Ziel-van der Made AC, Autar B, et al. Targeted biallelic inacti-

tion of Pten in the mouse prostate leads to prostate cancercompanied by increased epithelial cell proliferation but not byuced apoptosis. Cancer Res 2005;65:5730–9.

Cancer Research



30. Madeapo

31. XumoNa

32. VivingCa

33. KimmicRe

34. Beinad8p

35. LeitioCa

36. Beade

37. GaAbingPro

38. Frafibr

39. HilmeCe

40. Acacep

41. Caleaba

42. Maanha17

43. ZhmTPD

44. MaCatum

45. Evca19

46. On(+/an10

47. Kocaca59

48. Chenrep10

49. GuSig


www.a

Dow


jumder PK, Febbo PG, Bikoff R, et al. mTOR inhibition reverses Akt-pendent prostate intraepithelial neoplasia through regulation ofptotic and HIF-1-dependent pathways. Nat Med 2004;10:594–601.Q, Majumder PK, Ross K, et al. Identification of prostate cancerdifier pathways using parental strain expression mapping. Proctl Acad Sci U S A 2007;104:17771–6.anco I, Palaskas N, Tran C, et al. Identification of the JNK signal-pathway as a functional target of the tumor suppressor PTEN.ncer Cell 2007;11:555–69.MJ, Bhatia-Gaur R, Banach-Petrosky WA, et al. Nkx3.1 mutante recapitulate early stages of prostate carcinogenesis. Cancers 2002;62:2999–3004.thel CR, Faith D, Li X, et al. Decreased NKX3.1 protein expressionfocal prostatic atrophy, prostatic intraepithelial neoplasia, andenocarcinoma: association with gleason score and chromosomedeletion. Cancer Res 2006;66:10683–90.Q, Jiao J, Xin L, et al. NKX3.1 stabilizes p53, inhibits AKT activa-n, and blocks prostate cancer initiation caused by PTEN loss.ncer Cell 2006;9:367–78.thel CR, Bieberich CJ. Loss of Nkx3.1 expression in the transgenicnocarcinoma of mouse prostate model. Prostate 2007;67:1740–50.o H, Ouyang X, Banach-Petrosky WA, Gerald WL, Shen MM,ate-Shen C. Combinatorial activities of Akt and B-Raf/Erk signal-in a mouse model of androgen-independent prostate cancer.c Natl Acad Sci U S A 2006;103:14477–82.nco OE, Shaw AK, Strand DW, Hayward SW. Cancer associatedoblasts in cancer pathogenesis. Semin Cell Dev Biol 2010;21:33–9.l R, Song Y, Cardiff RD, Van Dyke T. Selective evolution of stromalsenchyme with p53 loss in response to epithelial tumorigenesis.ll 2005;123:1001–11.
evedo VD, Gangula RD, Freeman KW, et al. Inducible FGFR-1tivation leads to irreversible prostate adenocarcinoma and anithelial-to-mesenchymal transition. Cancer Cell 2007;12:559–71.
50. WoPI320

acrjournals.org


rracedo A, Ma L, Teruya-Feldstein J, et al. Inhibition of mTORC1ds to MAPK pathway activation through a PI3K-dependent feed-ck loop in human cancer. J Clin Invest 2008;118:3065–74.L, Teruya-Feldstein J, Behrendt N, et al. Genetic analysis of Pten

d Tsc2 functional interactions in the mouse reveals asymmetricalploinsufficiency in tumor suppression. Genes Dev 2005;19:79–86.ang H, Cicchetti G, Onda H, et al. Loss of Tsc1/Tsc2 activatesOR and disrupts PI3K-Akt signaling through downregulation ofGFR. J Clin Invest 2003;112:1223–33.nning BD, Logsdon MN, Lipovsky AI, Abbott D, Kwiatkowski DJ,ntley LC. Feedback inhibition of Akt signaling limits the growth ofors lacking Tsc2. Genes Dev 2005;19:1773–8.

eritt JI, Goldsworthy TL, Wolf DC, Walker CL. Hereditary renal cellrcinoma in the Eker rat: a rodent familial cancer syndrome. J Urol92;148:1932–6.da H, Lueck A, Marks PW, Warren HB, Kwiatkowski DJ. Tsc2-) mice develop tumors in multiple sites that express gelsolind are influenced by genetic background. J Clin Invest 1999;4:687–95.bayashi T, Minowa O, Kuno J, Mitani H, Hino O, Noda T. Renalrcinogenesis, hepatic hemangiomatosis, and embryonic lethalityused by a germ-line Tsc2 mutation in mice. Cancer Res 1999;:1206–11.oo AY, Yoon SO, Kim SG, Roux PP, Blenis J. Rapamycin differ-tially inhibits S6Ks and 4E-BP1 to mediate cell-type-specificression of mRNA translation. Proc Natl Acad Sci U S A 2008;5:17414–9.ertin DA, Sabatini DM. The pharmacology of mTOR inhibition. Scinal 2009;2:e24.rkman P, Clarke PA, Raynaud FI, van Montfort RL. Drugging the
kinome: from chemical tools to drugs in the clinic. Cancer Res
10;70:2146–57.




2010;70:8937-8947. Published OnlineFirst October 12, 2010.Cancer Res Raleigh D. Kladney, Robert D. Cardiff, David J. Kwiatkowski, et al. in Aged Mice

CancerSuppressor Essential to Prevent Spontaneous Prostate Tuberous Sclerosis Complex 1: An Epithelial Tumor

Updated version

10.1158/0008-5472.CAN-10-1646doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2010/10/11/0008-5472.CAN-10-1646.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/70/21/8937.full#ref-list-1

This article cites 50 articles, 20 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/70/21/8937.full#related-urls

This article has been cited by 5 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/70/21/8937To request permission to re-use all or part of this article, use this link

on April 7, 2020. © 2010 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from


http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-10-1646

http://cancerres.aacrjournals.org/content/suppl/2010/10/11/0008-5472.CAN-10-1646.DC1

http://cancerres.aacrjournals.org/content/70/21/8937.full#ref-list-1

http://cancerres.aacrjournals.org/content/70/21/8937.full#related-urls

http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://cancerres.aacrjournals.org/content/70/21/8937


Documents

Cancer Research Tuberous Sclerosis Complex 1: An Epithelial … · Tumor and Stem Cell Biology Tuberous Sclerosis Complex 1: An Epithelial Tumor Suppressor Essential to Prevent Spontaneous