The Prostate 70:239 ^251 (2010)

ArachidonicAcidActivationof Intratumoral SteroidSynthesisDuring ProstateCancer Progressionto

CastrationResistance

Jennifer A. Locke,1 Emma S. Tomlinson Guns,1 Melanie L. Lehman,1

Susan Ettinger,1 Amina Zoubeidi,1 Amy Lubik,4 Katia Margiotti,1,3

Ladan Fazli,1 Hans Adomat,1 Kishor M. Wasan,2 Martin E. Gleave,1

and Colleen C. Nelson1,4*1DepartmentofUrologic Sciences,Universityof British Columbia,Vancouver,BC,Canada

2Departmentof Pharmaceutical Sciences,Universityof British Columbia,Vancouver,BC,Canada3LaboratoryofMolecularMedicineandBiotechnology,University‘Campus Bio-Medico of Rome’,Rome, Italy4Australian Prostate Cancer ResearchCentre-Queensland, Institute ofHealthandBiomedical Innovation,

QueenslandUniversityof Technology, Brisbane,QLD, Australia

BACKGROUND. De novo androgen synthesis and subsequent androgen receptor (AR)activation has recently been shown to contribute to castration-resistant prostate cancer (CRPC)progression. Herein we provide evidence that fatty acids (FA) can trigger androgen synthesiswithin steroid starved prostate cancer (CaP) tumor cells.METHODS. Tumoral FA and steroid levels were assessed by GC–MS and LC–MS,respectively. Profiles of genes and proteins involved in FA activation of steroidogenesis wereassessed by fluorescence microscopy, immunohistochemistry, microarray expression profilingand Western blot analysis.RESULTS. In human CaP tissues the levels of proteins responsible for FA activation of steroidsynthesis were observed to be altered during progression to CRPC. Further investigating thismechanism in LNCaP cells, we demonstrate that specific FA, arachidonic acid, is synthesized inan androgen-dependent and AR-mediated manner. Arachidonic acid is known to inducesteroidogenic acute regulatory protein (StAR) in steroidogenic cells. When bound to hormonesensitive lipase (HSL), StAR shuttles free cholesterol into the mitochondria for downstreamconversion into androgens. We show that arachidonic acid induces androgen production insteroid starved LNCaP cells coincidently in the same conditions that HSL and StAR arepredominantly localized in the mitochondria. Furthermore, their activities are verified by afunctional increase in mitochondrial uptake of cholesterol in this steroid starved environment.CONCLUSIONS. We propose that this characterized arachidonic acid induced steroido-genesis mechanism significantly contributes to the activation of AR in CRPC progression andtherefore recommend that fatty acid pathways be targeted therapeutically in progressing CaP.Prostate 70: 239–251, 2010. # 2009 Wiley-Liss, Inc.

KEY WORDS: arachidonic acid; castration-resistant prostate cancer; cholesterol;androgens

INTRODUCTION

Most men with metastatic prostate cancer (CaP)initially respond to androgen deprivation therapy(ADT) but unfortunately, castration-resistant prostatecancer (CRPC) ultimately prevails. Despite reducedlevels of circulating androgens in these patients after

*Correspondence to: Colleen C. Nelson, PhD, Institute of Health andBiomedical Innovation, Queensland University of Technology, 60Musk Avenue Kelvin Grove Urban Village Kelvin Grove, Queens-land, 4059 Australia. E-mail: colleen.nelson@qut.edu.auReceived 9 June 2009; Accepted 18 August 2009DOI 10.1002/pros.21057Published online 29 September 2009 in Wiley InterScience(www.interscience.wiley.com).

) 2009 Wiley-Liss, Inc.

ADT, androgen receptor (AR) activation remains acentral mechanism in CRPC progression [1–6]. We andothers have recently demonstrated that tumors adaptto an environment deprived of circulating androgensto synthesize their own androgens for AR activation[7–9]. In clinical trials evaluating androgen synthesisinhibitors in patients with CRPC, AR-mediated PSAresponses have been achieved in more than 50% ofthe patients supporting the role of these intratumoralandrogens in disease progression [10,11]. Futureexploration into the mechanisms underlying intra-tumoral androgen synthesis is therefore of centralimportance to developing and improving treatmentsfor this disease.

Many enzymes responsible for androgen synthesisare under the control of AR-regulated transcriptionfactors, sterol regulatory element binding proteins(SREBPs) and are increased in expression during CRPCprogression including those responsible for the regu-lation of endogenous fatty acids (FAs) (fatty acidsynthase, FASN) and cholesterol, the central precursorof androgens [2,9,12–16]. FAs and cholesterol havemany regulatory functions in the cell [17–21] andare the main precursors for several lipids that havebeen implicated in CaP development and progression[19,22,23]. Additionally, in steroidogenic ovarian,adrenal and testicular cells FAs have been shown to

trigger steroid synthesis from cholesterol [24,25]. Wehypothesize that FAs also play a role in mediatingintratumoral androgen synthesis in CaP cells.

In this study, increased AR-mediated synthesis ofarachidonic acid (AA) was observed in CaP cells.Interestingly, AA is known to be the most efficient FA toinduce steroid synthesis [25]. Several groups haveexplored the detailed mechanisms involved in AAactivation of steroid synthesis from cholesterol in otherorgans [24–26]. Herein, we have characterized the keyproteins involved in AA activation of steroidogenesisin CaP cells (Fig. 1, Table I) and discuss how thismechanism may contribute to the progression ofthe disease to CRPC (Fig. 1). Mechanistically, inthe presence of exogenous androgens in vitro or beforecastration in vivo, CaP cells preferentially producelarge endogenous stores of non-toxic cholesteryl esters(CEs) from cholesterol and FAs by an enzyme known asacyl-CoA:cholesterol acyltransferase-2 (ACAT2) [27](Fig. 1, AD). After castration in vivo or in the absence ofexogenous androgens in vitro, CEs are cleaved by anenzyme known as hormone sensitive lipase (HSL) toproduce free cholesterol and free FAs [28] (Fig. 1, N).In turn, these FAs including AA are known to beactivated by long-chain acyl-CoA synthetase (ACSL)[29] (Fig. 1, CRPC1). The isoform that is highly expres-sed in the human prostate, ACSL3, is preferentially

The Prostate

Fig. 1. Overallschematicdiagramof theproposedmechanisminCaPcellsduringprogressiontoCRPC.SchematicofclinicalCaPprogression:initialtumorisandrogen-dependent(AD)andafterandrogendeprivationtherapy(ADT)thetumorentersanadir(N)stateuntilitprogresses toa castration-resistant (CRPC) form. In the presence of androgens (AD) increased FASN production of arachidonic acid (E) leads to ACAT-2inducedaccumulationof storageCE’s.UponADTthe tumor isdeprivedof androgens (N)andHSLcleavesCE’s to formfreeCand freeE.FreeEis activatedbyACSL3 in thecytosol toproduceactivatedarachidonic acid (E-CoA)whichthenmigrates to theoutermitochondrialmembranefor transfer to the inner mitochondria via an ACSL3 ^ACBP^PBR complex (CRPC1). In the mitochondria E-CoA undergoes thioesterasereactiontoproducefreeEwhichis thenmetabolizedandactivatestranscriptionofStARgeneleading toincreasedaccumulationofStARproteinat the mitochondrial membrane. Free C is then transported into the mitochondria by newly synthesized StAR where it undergoes sidechain cleavagebyCYP11A1to formpregnenolone for subsequentconversion to downstream androgens therebycontributing toAR-mediateddiseaseprogression (CRPC2).

240 Locke et al.

responsible for AA activation [30,31]. Activated AA-CoA can be transported through a series of reactionswith acyl-CoA binding protein (ACBP) and peripheral-type-benzodiazepine-receptor (PBR) to ultimatelycross the mitochondrial membrane [25,29]. In themitochondria, AA-CoA undergoes another reactionwith mitochondrial acyl-CoA thioesterase-9 (ACOT9)to once again produce free AA which can then undergometabolism and activate the transcription of steroidacute regulatory protein (StAR) in the nucleus [24–26](Fig. 1, CRPC2). New StAR protein accumulates at themitochondrial membrane and catalyzes the rate-limit-ing step involved in steroid synthesis [24–26,32–35].Through an interaction between mitochondrial StARand cytosolic HSL, free cholesterol is shuttled intothe mitochondria and subsequently converted byCYP11A1 into downstream steroid pregnenolone[24–26,32–34,36,37]. This in turn triggers a cascade ofreactions to produce androgens [38] which havepreviously been shown to be synthesized in CRPCtumors in levels sufficient for AR activation [7–9].

We propose that this characterized AA inducedsteroidogenesis mechanism significantly contributes toCRPC progression and may therefore be of significancein developing and improving therapies targeting thisdisease.

MATERIALSANDMETHODS

Materials

Fatty acid methyl ester (FAMEs) standards werereconstituted in hexane (Sigma Aldrich, Oakville, ON).R1881 (Dupont, Boston, MA), arachidonic acid and

casodex (Sigma–Aldrich, Oakville, Ontario, Canada)were prepared in 95% EtOH. Radioactively labeled14C-cholesterol was obtained from GE HealthcareBio-Sciences (Piscataway, NJ).

InVitroModel:LNCaPCells

LNCaP cells (passage 40–48; American Type Cul-ture Collection, Rockville, MD) were cultured in RPMI-1640 (without phenol red) with L-glutamin, PS and 5%fetal bovine serum (FBS, Hyclone, Logan, UT) or 5%charcoal stripped serum (CSS, Hyclone). LNCaP cellswere maintained in 5% FBS, however, 48 hr prior totreatment with R1881 or arachidonic acid, cells werecultured in 5% CSS. After 72 hr of treatment with 0, 0.1,1, and 10 nM R1881 in the presence/absence of 25 mMcasodex, or 48 hr of treatment with 1 mM arachidonicacid, media was removed, cells were washed withPBS, scraped, pelleted at 2,000g for 5 min and storedat �808C.

InVivoModel:LNCaPTumor Progression toCastration-Resistance

All animal experimentation was conducted inaccordance with accepted standards of the UBCCommittee on Animal Care. LNCaP xenograft tumorswere grown in athymic nude mice at four sites;progression of the disease was monitored by tumorvolume and PSA measurements [2]. Tumors wereobtained at predetermined time points based on PSAprofiles: androgen-dependence (AD, pre-castration),nadir (N, 8 days post-castration) and castration-resistance (CRPC, 35 days post-castration) [7].

The Prostate

TABLE I. Proteins Involvedin FAActivationof Steroid Synthesis

Symbol Full name Function

ACAT2 Acetyl-coenzyme A acetyltransferase 2 Catalyzes formation of cholesteryl esters from free cholesteroland free fatty acids

ACBP Acyl-coenzyme A binding protein Involved in transport of activated fatty acid into the mitochondriaby PBR

ACOT9 Acyl-coenzyme A thioesterase-9 Catalyzes reaction which converts activated fatty acid into afree fatty acid in the mitochondria

ACSL3 Long-chain acyl-CoA synthetase-3 Catalyzes activation of fatty acid in the cytosol and initiates transferof activated fatty acid across the mitochondrial membrane

FASN Fatty acid synthase Responsible for endogenous fatty acid synthesisHSL Hormone sensitive lipase Catalyzes the breakdown of cholesteryl esters into free cholesterol

and free fatty acid; carries free cholesterol to outer mitochondrialmembrane

PBR Peripheral benzodiazepine receptor Catalyzes the transport of activated fatty acid from the outerto inner mitochondria membrane

StAR Steroidogenic acute regulatory protein Catalyzes the transport of free cholesterol into the mitochondriafor steroid synthesis

FattyAcids and Steroidogenesis in Prostate Cancer 241

Human Prostate Samples

Prostastic tissues included 14 primary CaPs frompatients undergoing radical prostatectomy with notherapy before surgery, 12 primary CaPs after 1–3 months of Neoadjuvant Hormone Therapy (NHT),5 primary CaPs after 5–6 months NHT, 4 primary CaPsafter 8–9 months NHT, and 3 CRPCs. Tissues obtainedwere flash frozen in OCT Compound (Tissue-Tek) untilprocessing.

FattyAcidDerivatization and Extraction

Fatty acids in cell pellets and xenograft tumorhomogenates were derivatized with 2% H2SO4 inMeOH for 2 hr at 378C in glass tubes. Hexane (200 ml)was added to tubes for extraction of FAMEs and phaseswere separated by centrifugation at 1,000g for 5 min.One hundred microliter of the top layer was transferredto a GC–MS vial for analysis.

FAMEAnalysis byGasChromatography^MassSpectrometry (GC^MS)

The GC–MS method for FAME analysis wasadapted from [39] on a Varian 210 ion trap massspectrometer. All MS data were collected in electronimpact positive (EIþ) mode with capillary voltage at3 kV, source and desolvation temperatures of 1208Cand 3508C, respectively and N2 gas flow of 450 L/hr.Chromatographic separations were carried out using atemperature gradient of 808C held at for 1 min, thenincreased at a rate of 208C/min to 2508C final temper-ature which was held at for another 6 min; totalrun time of 15.50 min. Flow rate was 0.3 mL/min,column temperature 358C and 0.05% formic acidwas present throughout the run. MS scan data wascompared to NCIS Library for metabolite identifica-tion. Peak area comparison to a 6-point calibrationcurve was used to determine overall FAME yield.

Steroid Extraction andDerivatization

Cell pellets were extracted with ethyl acetate (1:1)and dried down. They were then reconstituted in 50%MeOH and derivatized in 0.2 M hydroxylamine–HClat 658C for 1 hr.

SteroidAnalysis by Liquid Chromatography^MassSpectrometry (LC^MS)

Samples were analyzed for steroids using a similarmethod to that reported by Kalhorn et al. [40]. Masstransitions specific to the individual steroids wereidentified to create multiple reaction monitoring(MRM) similar to those outlined in the reference. A7-point calibration curve was used to quantify levels of

pregnenolone, testosterone and dihydrotestosteroneusing deuterated testosterone as an internal standardwith detection limits of approximately 100, 5, and25 pg/ml, respectively. Levels of steroids are normal-ized to initial cell pellet weight.

Laser CaptureMicrodissection andMicroarrayAnalysis

Laser capture microdissection (LCM) was per-formed on cancer cells using the PALM Microlasersystem (P.A.L.M. Microlaser Technologies, Germany).Samples were catapulted into sterile caps of 0.5 mleppendorf tubes (RNAse-free) containing 40 ml extrac-tion buffer and total RNA was isolated according tomanufacturers’ instructions (PicoPure RNA IsolationKit, KIT0204, ARCTURUS) and subjected to DNasetreatment using Qiagen RNase-Free DNase kit (Qiagen,Inc). The RiboAmp HS RNA Ampification kit (KIT0215,ARCTURUS) was used to generate amplified aminoallyl-modified antisense RNA according to manufac-turer’s instructions; amplified aRNA was quantifiedby spectrophotomer and the quality was assessed byrunning a 1% denaturing agarose gel. Amino allyl-modified aRNA was labeled using the Amino AllylMessage Amp IIa RNA Amplicfication Kit (Ambion),and the labeled aRNA was fragmented with RNAFragmentation Reagents (Ambion) prior to hybrid-ization. Microarrays of 34,580 (70-mer) human oligosrepresenting 24,650 genes and 37,123 gene transcripts(Human Operon V3.0, Operon Technologies, Hunts-ville, Al) were printed on slides (Matrix Technologies,Hudson, NH). Microarrays were competitively hybri-dized with 2 mg amplified aRNA from the micro-dissected samples labeled with Cy5 and 2 mg ofamplified Universal Human Reference RNA (Strata-gene) labeled with Cy3 fluors (Amersham Bioscience).Following overnight hybridization and washing,arrays were scanned on a Scan Array Express Micro-array Scanner (Perkin Elmer). Signal quality andquantity were assessed using ImaGene 8.0 software(BioDiscovery, San Diego, CA). Feature data extractedby ImaGene were subjected to background correction,print-tip lowess within-array normalization and quan-tile between-array normalization using the LIMMABioconductor package [41]. A moderated t-statistic(corrected for a false discovery rate of 10%) from linearmodels built using LIMMA was used to determinedifferential expression between the treatment groups.

Immunohistochemistry

Immunohistochemical staining was conducted byVentana autostainer model Discover XT TM (VantanaMedical System, Tuscan, AZ) using enzyme labeledbiotin streptavidin system and solvent resistant DAB

The Prostate

242 Locke et al.

Map kit. Antibodies: mouse monoclonal ACBP (1:100,Abcam, Cambridge, MA), rabbit polyclonal ACSL3(1:1,000, Abgent, San Diego, CA), rabbit polyclonalFASN (1:1,000, Santa Cruz Biotechnologies, Santa Cruz,CA), rabbit polyclonal HSL (1:1,000, Abcam) andpolyclonal rabbit StAR (1:100; kindly donated fromDr. D. Hales, Chicago, IL) were used for immunohis-tochemical staining.

Mitochondrial Fractionation

Mitochondrial fractionation of LNCaP cells wasconducted using a Mitosciences cell fractionation kit(Eugene, OR). Verification of mitochondrial isolationwas conducted by Western blot analysis of cytosolicspecific protein GAPDH (1:5,000, Abcam) and mito-chondrial specific protein cytochrome c (1:100; BDBiosciences, San Jose, CA).

Western BlotAnalysis

Cell pellets were reconstituted in 100 ml of RIPAbufferþprotease inhibitor, sheared and spun down at13,000g for 5 min. Supernatants from pellet extractionsand mitochondrial isolations were analyzed for totalprotein content using the BCA protein determinationkit (Sigma, Oakville, Ontario, Canada). Fifteen micro-gram of total protein was loaded for each sample onto a9% acrylamide gel. Antibodies previously mentionedin immunohistochemistry section were used to identifyand quantify respective proteins.

FluorescenceMicroscopy

Cells were fixed in ice-cold methanol completedwith 3% acetone for 10 min at 48C, then washed withPBS and incubated with 0.2% Triton/PBS for 10 min,followed by washing and 30 min blocking in 3% nonfatmilk before the addition of antibodies overnight todetect cytochrome c, DAPI (1:100; Vector laboratories,Inc., Burlingame, CA) and StAR. Antigens werevisualized using anti-rabbit or anti-mouse antibodiescoupled to FITC or rhodamine (1:500; 30 min). Photo-micrographs were taken at 20� magnification usingZeiss Axioplan II fluorescence microscope, followed byanalysis with imaging software (Northern Eclipse,Empix Imaging, Inc., Mississauga, ON).

Mitochondrial14C-CholesterolUptakeAssay in LNCaPCells

Steroid deprived LNCaP cells were treatedwith �1 nM R1881 in combination with 10 mCi of14C-cholesterol for 2 and 48 hr. As previously docu-mented, cells were pelleted and mitochondria wereisolated. Fractions of mitochondria, cytosol and entirecell media were analyzed by a scintillation counter.

RESULTS

FattyAcid Levels IncreaseDuringProgression toCRPC

To expand on the observation that increased FASNexpression occurs during CaP progression [42–44], weanalyzed tumors obtained using the LNCaP xenograftmodel at various stages of the disease for FA levels byGC–MS. Results indicate that tumor levels of the 9 FAsanalyzed in Figure 2a slightly decrease after castration(N (n¼ 5) compared to AD (n¼ 5)) and then increase inCRPC tumors (n¼ 6).

Specif|c FattyAcidsAre Producedin anAndrogen-Dependent andAndrogen-Receptor

MediatedMechanism

To investigate the specific FAs that are endo-genously synthesized in an androgen-dependent man-ner, we conducted dose-titration studies with syntheticandrogen, R1881, in steroid starved LNCaP cells.Results indicate that three particular FAs are producedby LNCaP cells in a dose-dependent androgen-medi-ated manner: myristic acid (r¼ 0.899, P¼ 0.101), arach-idonic acid (r¼ 0.910, P¼ 0.089), and eicosapentaenoicacid (r¼ 0.950, P¼ 0.050) (Fig. 2b). In the presence ofantiandrogen, casodex, which inhibits AR signaling,the androgen-mediated production of myristic acid,arachidonic acid (AA) and eicosapentanoic acid aredepleted (Fig. 2c).

These results indicate that select FAs are producedby CaP cells in an AR-mediated manner. AA haspreviously been shown to be the most efficient FA toinduce steroid synthesis from cholesterol [25].

Steroid Levels Increase inAndrogenDeprivedCaPCellsTreatedWithArachidonic Acid

The functional implication of AA induced steroidsynthesis was assessed by LC–MS analysis. AA in aconcentration of 1 mM significantly increases the pro-duction of androgen precursor pregnenolone as well asdownstream androgens testosterone and dihydrotes-tosterone in steroid starved LNCaP cells as comparedto ethanol control (P< 0.05) (Fig. 3).

Proteins Responsible forArachidonicAcidActivationof Steroid Synthesis Are Expressedin

HumanCaPTissues

Drawing from findings in other steroidogenic celllines (Fig. 1, Table I), we have identified five keyproteins responsible in mediating AA activation ofsteroid synthesis: ACBP, ACSL3, FASN, HSL, andStAR. To determine if this mechanism may occur

The Prostate

FattyAcids and Steroidogenesis in Prostate Cancer 243

The Prostate

Fig. 2. a: FA levels in LNCaP xenograft tumors.GC^MS analysis for FAcontent in LNCaP xenograft tumors at androgen-dependent (AD,n¼ 5), nadir (N, n¼ 5) and castration-resistant (CRPC, n¼ 6) stages of the disease.Mean FA levels depicted as normalized initially to tumorweight (g) and then foldchangeinADlevels of the samemouse.b:FAlevels inandrogen treatedLNCaPcells.GC^MSanalysis forFAcontentinsteroidstarvedLNCaPcellsinthepresenceof0(EtOH),0.1,1,and10nMR1881(n¼ 3each).MeanFAlevelsdepictedasnormalizedtopelletweight(g) andfoldchangeinEtOHalonelevelsof the sameexperiment.Linearregressionstatisticswereemployedtoidentifydose-dependent trendsin FAs levels and are depicted as r-values and respective P-values. c: FA levels in androgen/casodex treated LNCaP cells.GC^MS analysis formyristic acid, arachidonic acid and eicosapentanoic acid content in steroid starved LNCaP cells in the presence of 0 (EtOH), 0.1,1, and10nMR1881in theabsenceorpresenceof25mMantiandrogen, casodex(n¼ 3each).MeanFAlevels, r andP-valuesdepictedasin (b).

244 Locke et al.

within CaP tumor cells, we initially analyzed humantissues obtained from CaP patients by immunohisto-chemistry for these proteins (Fig. 4). We confirm ourpreviously reported result of ACBP expression inhuman CaP (Fig. 4a) [2]. ACSL3 is found to be presentin human CaP tumors and expressed in a nuclearmanner while HSL is expressed in human CaP tissuesin a cytoplasmic pattern (Fig. 4b,d). Furthermore, StARexhibits immunoreactivity in both benign and cancercells (Fig. 4e) while FASN appears to be over-expressedin cancer cells as compared to benign cells (Fig. 4c).These results verify that ACBP, ACSL3, FASN, HSL,and StAR are expressed in human CaP tissues.

Expressions of Proteins Responsible forArachidonicAcidActivation of Steroid Synthesis Are Increased

During Progression toCRPC

We assessed the transcriptional expression of theindicated enzymes needed for AA activation (Fig. 1,Table I) during progression to CRPC in patient tissuesamples in relation to ADT treatment response(untreated androgen-dependent (AD) cancer, patientsundergoing neoadjuvant hormone therapy (NHT) for1–3 months (1–3m), 5–6 months (5–6m), 8–9 months(8–9m) and at castration-resistance (CRPC) as definedby recurring PSA). We confirm that in human CaPtumors ACBP mRNA levels significantly decrease aftercastration (neoadjuvant hormone therapy) (P¼ 0.002)and appear to increase in tumors from patients at CRPC(Fig. 5a-i) [2]. Furthermore, human tumor ACSL3mRNA levels significantly decrease immediately aftercastration (P< 0.001), remain low up to 9 months(NHT) (P¼ 0.003) and then increase in tumors fromCRPC patients (Fig. 5a-iii). Human tumor HSL mRNAexpression increases immediately post-castration

The Prostate

Fig. 3. Effect of exogenous arachidonic acid (AA) on steroidproduction in steroid starved LNCaPcells.Levels of pregnenolone,testosterone and dihydrotestosterone in steroid starved LNCaPcells�1mMarachidonic acidaredepictedasnormalizedtocellpelletweight (g). Letter (a) indicates statistically significant from EtOHcontrol (P< 0.05).

Fig. 4. Expression of proteins in humanCaP tissues. Immunohis-tochemistryanalysis for thepresenceofACBP(a),ACSL3(b),FASN(c), HSL (d), and StAR (e) in CaP tumor specimens collected frompatients afterradicalprostatectomy.

FattyAcids and Steroidogenesis in Prostate Cancer 245

(1–3 months) (P¼ 0.048) and then gradually decreasein time leading to CRPC (Fig. 5a-v) while aftercastration human tumor PBR mRNA levels decreasegradually to CRPC (Fig. 5a-vi). Human tumor ACOT9mRNA expression appears to decrease gradually aftercastration to 8–9 months NHT (P¼ 0.037) and thenincrease once again at CRPC (P¼ 0.022) (Fig. 5a-ii)while tumor StAR mRNA levels remain the same

immediately after castration and then increase at CRPC(P¼ 0.040) (Fig. 5a-vii). As predicted by previousstudies, FASN tumor mRNA expression decreasesafter 1–3 months of NHT (P¼ 0.017) and then FASN’sexpression re-appears during progression of thedisease to CRPC (P¼ 0.008) (Fig. 5a-iv).

Due to the limited availability of human tissues toanalyze protein levels of these enzymes, we utilized

The Prostate

Fig. 5. a: Profiles ofproteins duringCaPprogression.Microarray analysis formeanmRNAlevels ofACBP (i),ACOT9 (ii),ACSL3 (iii),FASN(iv),HSL(v),PBR(vi),andStAR(vii) in tumors fromuntreatedandrogen-dependent(AD)patients,patients treatedwithneoadjuvanthormonetherapy(NHT)for1^3months(1^3m),5 ^ 6months(5 ^ 6m),8 ^9months(8 ^9m)andatcastration-resistance(CRPC)asdefinedbyrecurringPSA (�SEM).Letter (a) indicates statistically significant fromAD (P< 0.05)while letter (b) indicates statistically significant from1to 3monthsNHT(P< 0.05).b:ProfilesofproteinsintheLNCaPxenograftmodel.Westernblotanalysis formeanproteinlevelsofACBP(i),ACSL3(ii),FASN(iii),HSL (iv), and StAR (v) inAD (n¼ 3),N (n¼ 3), andCRPC (n¼ 3) tumors from the LNCaP xenograftmodelnormalized toGAPDHof thesameblotandADof the samemouse(�SEM)aswellasrepresentativeWesternblotdepictedbelow.Letter (a) indicates statistically significantfromAD(P< 0.05)whileletter (b) indicates statistically significantfromN(P< 0.05).

246 Locke et al.

tumors obtained at various stages of the disease fromthe LNCaP xenograft model. Previously, ACBP andStAR protein levels at AD, N, and CRPC in the LNCaPxenograft model have been reported [2,7]. In replica-tion of these results, we show that ACBP protein levelssignificantly decrease at N (P< 0.001) and increaseonce again during progression of the disease to CRPC(P< 0.001) (Fig. 5b-i) and StAR protein levels appearto increase from AD to N to CRPC (Fig. 5b-v). Inthis study, ACSL3 protein levels appear to decreaseimmediately after castration and significantly increaseonce again at CRPC (P¼ 0.005) (Fig. 5b-ii). HSL proteinlevels increase in N tumors as compared to AD(P< 0.001) and remain relatively high in CRPC tumors(Fig. 5b-iv). Finally, FASN expression is high at AD,decreases at N and then once again emerges at CRPC(P¼ 0.002) (Fig. 5b-iii).

The mRNA and protein profiles of these enzymesdetermined from CaP tissues obtained from menundergoing various NHT treatment time courses andtumors obtained using the LNCaP xenograft model ofCaP progression are consistent with each other.

Proteins Responsible forArachidonicAcidActivationof Steroid Synthesis Are Localized inTheirActive

Site inAndrogenDeprived CaPCells

We evaluated the localization of these proteinswithin the CaP cell’s mitochondria in the absence andpresence of androgens in order to infer their active/inactive involvement in this AA mediated steroido-genesis mechanism. Initially, in order to verify that the1nM R1881 treatment was effective, the expression ofthese proteins were analyzed in whole cell lysates byWestern blot analysis; all (except StAR) are signifi-cantly induced by 1nM R1881 treatment in steroidstarved LNCaP cell lysates (Fig. 6a-i–vi). Mitochon-drial and cytosolic extracts from these cells were alsoanalyzed by Western blot for protein expression,alongside untreated whole cell lysates. ACBP, ACSL3,FASN, HSL and StAR are expressed in the mitochon-dria of steroid starved LNCaP cells (Fig. 6b); howeverdistinct differences arise in the presence and absenceof androgens. ACBP expression is predominantlycytosolic (P¼ 0.006); however, in the absence ofandrogens, localization of ACBP in the mitochondriais significantly increased as compared to in thepresence of androgens (P¼ 0.01) (Fig. 6b-i). ACSL3 islocalized mainly in the mitochondria in the absenceof androgens whereas in the presence of androgensACSL3 appears to be localized mainly in the cytosol(Fig. 6b-ii). HSL expression appears to be higher in themitochondria regardless of the presence of androgens(Fig. 6b-iv). FASN expression increases in both cytosoland mitochondrial fractions in the presence of andro-

gen as compared to the vehicle control (EtOH alone)(P¼ 0.01). This result is predicted because unlike theother enzymes being studied, the localization of FASNin the mitochondria is not linked to its overall activity(Fig. 6b-iii). StAR expression remained low in allfractions except during the absence of androgens whenits expression became elevated in the mitochondriafraction (P¼ 0.01) (Fig. 6b-v). To confirm the novelfinding that in the absence of androgens StAR becomesactive, we also used a fluorescence labeling techniqueto determine if StAR co-localized with the mitochon-dria marker, cytochrome c, in steroid starved LNCaPcells. Fluorescence microscopy after cytochrome c,StAR and nuclear DAP1 staining (Fig. 6c) indeed showsco-localization of cytochrome c and StAR in LNCaPcells in the absence of androgens.

These unique and novel findings suggest that theproteins necessary for AA activation of steroid syn-thesis are present and active in CaP cells in an androgendeprived environment. PBR and ACOT9 subcellularlocalization were not evaluated in this study as bothare known to reside in the mitochondria regardlessof activity.

IncreasedMitochondrialuptake of Cholesterol inAndrogenDeprivedCaPCells

Drawing from previous reports, StAR induction byAA leads to the uptake of cholesterol into themitochondria by an HSL–StAR shuttling interaction[36]. We evaluated the localization of cholesterol, theendpoint of this mechanism, in fractionated LNCaPcells previously treated with radioactively labeledcholesterol �1 nM R1881 for 2 and 48 hr. In both thepresence and absence of androgens, most of theradioactively labeled cholesterol remained in the media(supernatant) of the cells after 2 hr (Fig. 7a). Over time(48 hr) in the absence of androgens significantly morecholesterol was taken into the cell (P< 0.001) andspecifically migrated to the mitochondria (P< 0.001)than in the presence of androgens (Fig. 7b).

DISCUSSION

Previously, we and others have shown that CaP cellsexpress and utilize all of the enzymes necessary tosynthesize androgens from cholesterol in an environ-ment deprived of a circulating source of androgens[3,7,8,12]. In this study, we detail a novel mechanismwhereby androgen dependent and AR-mediated pro-duction of AA can trigger mitochondrial cholesteroluptake for subsequent androgen synthesis in CaP cellsin the absence of circulating androgens after castration(Fig. 1).

We first demonstrate that in the presence ofandrogens, or prior to castration, CaP tumors produce

The Prostate

FattyAcids and Steroidogenesis in Prostate Cancer 247

large amounts of FAs through the induction of FASN.Previously, we have shown that in the presence ofandrogens CaP cells store these FAs in the form ofnon-toxic cholesteryl esters (CEs) through a reactionmediated by ACAT2 [27]. Herein, we provide evidencethat after castration these CE stores can be utilized byCaP cells to synthesize androgens for AR activation.HSL is an enzyme responsible for cleavage of CEs toform free FAs and free cholesterol. The expression ofHSL increases immediately after castration in tumorsas compared to pre-castrate levels suggesting thatcleavage of these CEs by HSL is increased immediatelyafter castration when the supply of exogenous andro-gens is removed. Furthermore, in CaP cells, it appears

that AA is one of the main FAs involved in CEformation by ACAT2 [27] and thus also one of themain FAs involved in dissociation by HSL immediatelyafter castration. Analysis of CaP cells and xenografttumors by GC–MS demonstrate that AA is producedinitially in an androgen-dependent and AR-mediatedmanner. Furthermore, AA’s activating enzyme,ACSL3, significantly increases in expression in CRPCtumors as compared to immediately post-castration.After activation, AA is normally transported by ACSL3to the outer mitochondrial membrane so that it can bindACBP in a complex with PBR allowing for it to shuttleinto the mitochondria [25,26,29]. We have verifiedthat ACSL3 mitochondrial localization is increased in

The Prostate

Fig. 6. a:Effectof androgensonproteinexpression.Westernblotanalysis formeanproteinlevelsofACBP(i),ACSL3(ii),FASN(iii),HSL(iv),StAR(v), andGAPDH(vi) inwholecell lysates fromLNCaPcells�1nMR1881.GAPDHwasusedas aproteinloadingcontrol.b:Effectof andro-gensonlocalizationofproteins.Westernblotanalysis formeanproteinlevels ofACBP(i),ACSL3 (ii),FASN(iii),HSL(iv), StAR(v), andGAPDH(vi) incytosolic (cyt)andmitochondrial (mit) fractionsandwholecelllysates(LNCaP)ofsteroidstarvedLNCaPcells�1nMR1881(n¼ 3each)aswellasrepresentativeWesternblotdepictedbelow.GAPDHproteinlocalizationsolelyinthecytosolic andwholecelllysatefractionsconfirmedefficiencyof fractionationexperiment.Letter(c)indicatesstatisticallydifferentfromcytosolwithsametreatment(�R1881)(P< 0.05)andletter(d) indicates statisticallydifferentfromR1881treatmentinthesamefraction(cytormit) (P< 0.05).c:Co-localizationofStARandcytochromecinthemitochondria.FluorescencemicroscopyanalysisforStAR,cytochromec (mitochondrialmarker)andDAP1(nuclearmarker)cellularlocal-izationinsteroidstarvedLNCaPcells(n¼ 2).Cytochromec localizationvisualizedingreen,StARlocalizationvisualizedinred,DAP1localizationvisualizedinblue;co-localizationofcytochromec andStARvisualizedinyellow.

248 Locke et al.

steroid deprived LNCaP cells as compared to steroidreplete cells, demonstrating that the trafficking ofactivated AA into the mitochondria is a probable eventin CaP cells. Upon entering the mitochondria activatedAA undergoes a thioesterase reaction to produce freeAA which then undergoes metabolism leading tothe nuclear transcription of StAR, the rate-limitingenzyme in steroid synthesis [24,25]. Nuclear StARtranscription in turn leads to increased production ofStAR protein at the mitochondrial membrane [34,35].In our fractionation experiment, protein expression ofStAR increases in the mitochondria in the absence ofandrogens but does not change in any of the otherfractions. Because only newly synthesized StAR caninduce steroidogenesis [34,35], we propose that theobserved increase in mitochondrial protein StARexpression is responsible for subsequent steroido-genesis in these androgen deprived LNCaP cells.Furthermore, HSL also appears to accumulate in themitochondria with StAR. Evidence that the mitochon-drial uptake of radioactively labeled cholesterol isincreased in steroid starved LNCaP cells as comparedto steroid replete cells was also demonstrated, furthersuggesting that in the absence of androgens, StAR andHSL work together to shuttle cholesterol into themitochondria for steroid synthesis to occur [36]. Oncecholesterol has entered the mitochondria, it readily

undergoes reactions with the downstream steroido-genesis enzyme, CYP11A1, to produce pregnenolone,the steroid precursor of androgens [7,34,45]. Weconfirm that AA treatment of steroid starved LNCaPcells does indeed induce the production of pregneno-lone as well as downstream androgens, testosteroneand dihydrotestosterone. Our results from in vitro,in vivo and clinical samples support the hypothesisthat CaP cells use synthesized AA to initiate de novoandrogen synthesis after castration in an environmentdeprived of circulating androgens.

This study highlights interplay between two lipidpathways (FA and cholesterol) in progression to CRPC.Expression of genes regulating FA synthesis (FASN)and cholesterol synthesis (HMG-CoA synthase, squa-lene monooxygenase and farnesyl diphosphate syn-thase) have been shown to be androgen-regulatedand linked to CaP progression [2,12]. There are manyreported cases implicating FAs and cholesterol inbiological mechanisms contributing to disease pro-gression. Some of these triggered events include cellmaintenance and energy storage for rapidly growingcancer cells, production of membrane lipid raftsfor various cell signaling events and metabolism tomolecules involved in these cell signaling events[13,17,46–49]. We provide data describing a mecha-nism whereby CaP cells can develop an ability to evadecastration-induced apoptosis through the intratumoralproduction of androgens and subsequent AR activa-tion. This mechanism in combination with many othersincluding increased anti-apoptotic and survival path-way signaling [50–52] and AR activation by alternativegrowth factor pathways and co-regulators [53–56],ultimately culminate in a castration-resistant pheno-type, the fatal form of CaP. By developing a betterunderstanding of how CaP progresses to CRPCthrough the elucidation of mechanisms such as intra-tumoral androgen synthesis described here, new andmore promising targets including ACAT2, ACSL3,FASN, and HSL can be evaluated in hope to providebetter treatments for patients and prolong progressionto CRPC.

REFERENCES

1. Chen CD, Welsbie DS, Tran C, Baek SH, Chen R, Vessella R,Rosenfeld MG, Sawyers CL. Molecular determinants of resist-ance to antiandrogen therapy. Nat Med 2004;10:33–39.

2. Ettinger SL, Sobel R, Whitmore TG, Akbari M, Bradley DR,Gleave ME, Nelson CC. Dysregulation of sterol responseelement-binding proteins and downstream effectors in prostatecancer during progression to androgen independence. CancerRes 2004;64:2212–2221.

3. Stanbrough M, Bubley GJ, Ross K, Golub TR, Rubin MA,Penning TM, Febbo PG, Balk SP. Increased expression of genesconverting adrenal androgens to testosterone in androgen-independent prostate cancer. Cancer Res 2006;66:2815–2825.

The Prostate

Fig. 7. Effect of androgens onmitochondrial cholesterol uptake.14C-cholesterolcontentincellularcytosolic (cyt) andmitochondrial(mit) fractions as well as supernatant media (sup) from steroidstarvedLNCaPcells�1nMR1881at2hr(a)and48hr(b) timepoints(n¼ 3 each) portrayed as fraction of total counts as determinedbyscintillation count. Letter (a) indicates statistically significant fromR1881treatmentinthesamefraction(cytormit)(P< 0.05)andletter(b) indicates statistically significant from2hr incubation in the samefraction(cytormit) (P< 0.05).

FattyAcids and Steroidogenesis in Prostate Cancer 249

4. Mostaghel EA, Nelson PS. Intracrine androgen metabolism inprostate cancer progression: Mechanisms of castration resist-ance and therapeutic implications. Best Pract Res Clin Endo-crinol Metab 2008;22:243–258.

5. Gregory CW, Hamil KG, Kim D, Hall SH, Pretlow TG, Mohler JL,French FS. Androgen receptor expression in androgen-independent prostate cancer is associated with increasedexpression of androgen-regulated genes. Cancer Res 1998;58:5718–5724.

6. Mohler JL, Gregory CW, Ford OH III, Kim D, Weaver CM,Petrusz P, Wilson EM, French FS. The androgen axis in recurrentprostate cancer. Clin Cancer Res 2004;10:440–448.

7. Locke JA, Guns ES, Lubik AA, Adomat HH, Hendy SC, WoodCA, Ettinger SL, Gleave ME, Nelson CC. Androgen levelsincrease by intratumoral de novo steroidogenesis duringprogression of castration-resistant prostate cancer. Cancer Res2008;68:6407–6415.

8. Dillard PR, Lin MF, Khan SA. Androgen-independent prostatecancer cells acquire the complete steroidogenic potential ofsynthesizing testosterone from cholesterol. Mol Cell Endocrinol2008;295:115–120.

9. Montgomery RB, Mostaghel EA, Vessella R, Hess DL, KalhornTF, Higano CS, True LD, Nelson PS. Maintenance of intra-tumoral androgens in metastatic prostate cancer: A mechanismfor castration-resistant tumor growth. Cancer Res 2008;68:4447–4454.

10. Attard G, Reid AH, Yap TA, Raynaud F, Dowsett M, Settatree S,Barrett M, Parker C, Martins V, Folkerd E, Clark J, Cooper CS,Kaye SB, Dearnaley D, Lee G, de Bono JS. Phase I clinical trial of aselective inhibitor of CYP17, abiraterone acetate, confirms thatcastration-resistant prostate cancer commonly remains hormonedriven. J Clin Oncol 2008;26:4563–4571.

11. Small EJ, Halabi S, Dawson NA, Stadler WM, Rini BI, Picus J,Gable P, Torti FM, Kaplan E, Vogelzang NJ. Antiandrogenwithdrawal alone or in combination with ketoconazole inandrogen-independent prostate cancer patients: A phase III trial(CALGB 9583). J Clin Oncol 2004;22:1025–1033.

12. Holzbeierlein J, Lal P, LaTulippe E, Smith A, Satagopan J, ZhangL, Ryan C, Smith S, Scher H, Scardino P, Reuter V, Gerald WL.Gene expression analysis of human prostate carcinoma duringhormonal therapy identifies androgen-responsive genes andmechanisms of therapy resistance. Am J Pathol 2004;164:217–227.

13. Swinnen JV, Heemers H, van de Sande T, de Schrijver E,Brusselmans K, Heyns W, Verhoeven G. Androgens, lipogenesisand prostate cancer. J Steroid Biochem Mol Biol 2004;92:273–279.

14. Ashida S, Nakagawa H, Katagiri T, Furihata M, Iiizumi M,Anazawa Y, Tsunoda T, Takata R, Kasahara K, Miki T, Fujioka T,Shuin T, Nakamura Y. Molecular features of the transition fromprostatic intraepithelial neoplasia (PIN) to prostate cancer:Genome-wide gene-expression profiles of prostate cancers andPINs. Cancer Res 2004;64:5963–5972.

15. Kuhajda FP. Fatty acid synthase and cancer: New application ofan old pathway. Cancer Res 2006;66:5977–5980.

16. Di Vizio D, Sotgia F, Williams TM, Hassan GS, Capozza F, FrankPG, Pestell RG, Loda M, Freeman MR, Lisanti MP. Caveolin-1 isrequired for the upregulation of fatty acid synthase (FASN),a tumor promoter, during prostate cancer progression. CancerBiol Ther 2007;6:1263–1268.

17. Patel MI, Kurek C, Dong Q. The arachidonic acid pathway and itsrole in prostate cancer development and progression. J Urol2008;179:1668–1675.

18. Nie D, Che M, Grignon D, Tang K, Honn KV. Role of eicosanoidsin prostate cancer progression. Cancer Metastasis Rev 2001;20:195–206.

19. Swinnen JV, Brusselmans K, Verhoeven G. Increased lipogenesisin cancer cells: New players, novel targets. Curr Opin Clin NutrMetab Care 2006;9:358–365.

20. Brusselmans K, De Schrijver E, Verhoeven G, Swinnen JV. RNAinterference-mediated silencing of the acetyl-CoA-carboxylase-alpha gene induces growth inhibition and apoptosis of prostatecancer cells. Cancer Res 2005;65:6719–6725.

21. De Schrijver E, Brusselmans K, Heyns W, Verhoeven G, SwinnenJV. RNA interference-mediated silencing of the fatty acidsynthase gene attenuates growth and induces morphologicalchanges and apoptosis of LNCaP prostate cancer cells. CancerRes 2003;63:3799–3804.

22. Hughes-Fulford M, Li CF, Boonyaratanakornkit J, Sayyah S.Arachidonic acid activates phosphatidylinositol 3-kinase signal-ing and induces gene expression in prostate cancer. Cancer Res2006;66:1427–1433.

23. Liu Y. Fatty acid oxidation is a dominant bioenergetic pathwayin prostate cancer. Prostate Cancer Prostatic Dis 2006;9:230–234.

24. Castilla R, Maloberti P, Castillo F, Duarte A, Cano F, Maciel FC,Neuman I, Mendez CF, Paz C, Podesta EJ. Arachidonic acidregulation of steroid synthesis: New partners in the signalingpathway of steroidogenic hormones. Endocr Res 2004;30:599–606.

25. Duarte A, Castillo AF, Castilla R, Maloberti P, Paz C, Podesta EJ,Cornejo Maciel F. An arachidonic acid generation/exportsystem involved in the regulation of cholesterol transport inmitochondria of steroidogenic cells. FEBS Lett 2007;581:4023–4028.

26. Maloberti P, Maciel FC, Castillo AF, Castilla R, Duarte A, ToledoMF, Meuli F, Mele P, Paz C, Podesta EJ. Enzymes involved inarachidonic acid release in adrenal and Leydig cells. Mol CellEndocrinol 2007;265–266:113–120.

27. Locke JA, Wasan KM, Nelson CC, Guns ES, Leon CG. Androgen-mediated cholesterol metabolism in LNCaP and PC-3 cell lines isregulated through two different isoforms of acyl-coenzymeA:Cholesterol Acyltransferase (ACAT). Prostate 2008;68:20–33.

28. Yeaman SJ. Hormone-sensitive lipase—New roles for an oldenzyme. Biochem J 2004;379:11–22.

29. Sandoval A, Fraisl P, Arias-Barrau E, Dirusso CC, Singer D,Sealls W, Black PN. Fatty acid transport and activation and theexpression patterns of genes involved in fatty acid trafficking.Arch Biochem Biophys 2008;477:363–371.

30. Kang MJ, Fujino T, Sasano H, Minekura H, Yabuki N, Nagura H,Iijima H, Yamamoto TT. A novel arachidonate-preferring acyl-CoA synthetase is present in steroidogenic cells of the ratadrenal, ovary, and testis. Proc Natl Acad Sci USA 1997;94:2880–2884.

31. Attard G, Clark J, Ambroisine L, Mills IG, Fisher G, Flohr P, ReidA, Edwards S, Kovacs G, Berney D, Foster C, Massie CE, FletcherA, De Bono JS, Scardino P, Cuzick J, Cooper CS, TransatlanticProstate Group. Heterogeneity and clinical significance of ETV1translocations in human prostate cancer. Br J Cancer 2008;99:314–320.

32. Wang X, Walsh LP, Reinhart AJ, Stocco DM. The role ofarachidonic acid in steroidogenesis and steroidogenic acuteregulatory (StAR) gene and protein expression. J Biol Chem 2000;275:20204–20209.

33. Jefcoate C. High-flux mitochondrial cholesterol trafficking, aspecialized function of the adrenal cortex. J Clin Invest 2002;110:881–890.

The Prostate

250 Locke et al.

34. Miller WL. Steroidogenic acute regulatory protein (StAR), anovel mitochondrial cholesterol transporter. Biochim BiophysActa 2007;1771:663–676.

35. Manna PR, Dyson MT, Stocco DM. Regulation of the steroido-genic acute regulatory protein gene expression: Present andfuture perspectives. Mol Hum Reprod 2009;15:321–333.

36. Shen WJ, Patel S, Natu V, Hong R, Wang J, Azhar S, Kraemer FB.Interaction of hormone-sensitive lipase with steroidogenic acuteregulatory protein: Facilitation of cholesterol transfer in adrenal.J Biol Chem 2003;278:43870–43876.

37. Kraemer FB. Adrenal cholesterol utilization. Mol Cell Endocri-nol 2007;265–266:42–45.

38. Miller WL. Androgen synthesis in adrenarche. Rev EndocrMetab Disord 2009;10:3–17.

39. Dodds ED, McCoy MR, Rea LD, Kennish JM. Gas chromato-graphic quantification of fatty acid methyl esters: Flameionization detection vs. electron impact mass spectrometry.Lipids 2005;40:419–428.

40. Kalhorn TF, Page ST, Howald WN, Mostaghel EA, Nelson PS.Analysis of testosterone and dihydrotestosterone from bio-logical fluids as the oxime derivatives using high-performanceliquid chromatography/tandem mass spectrometry. RapidCommun Mass Spectrom 2007;21:3200–3206.

41. Smyth GK. Linear models and empirical Bayes methods forassessing differential expression in microarray experiments. StatAppl Genet Mol Biol 2004;3:Article 3.

42. Menendez JA, Lupu R. Fatty acid synthase and the lipogenicphenotype in cancer pathogenesis. Nat Rev Cancer 2007;7:763–777.

43. Lu S, Archer MC. Fatty acid synthase is a potential moleculartarget for the chemoprevention of breast cancer. Carcinogenesis2005;26:153–157.

44. Migita T, Ruiz S, Fornari A, Fiorentino M, Priolo C, Zadra G,Inazuka F, Grisanzio C, Palescandolo E, Shin E, Fiore C, Xie W,Kung AL, Febbo PG, Subramanian A, Mucci L, Ma J, Signoretti S,Stampfer M, Hahn WC, Finn S, Loda M. Fatty acid synthase: Ametabolic enzyme and candidate oncogene in prostate cancer.J Natl Cancer Inst 2009;101:519–532.

45. Miller WL. Androgen biosynthesis from cholesterol to DHEA.Mol Cell Endocrinol 2002;198:7–14.

46. Swinnen JV, Vanderhoydonc F, Elgamal AA, Eelen M, VercaerenI, Joniau S, Van Poppel H, Baert L, Goossens K, Heyns W,Verhoeven G. Selective activation of the fatty acid synthesispathway in human prostate cancer. Int J Cancer 2000;88:176–179.

47. Brusselmans K, Timmermans L, Van de Sande T, Van VeldhovenPP, Guan G, Shechter I, Claessens F, Verhoeven G, Swinnen JV.Squalene synthase, a determinant of Raft-associated cholesteroland modulator of cancer cell proliferation. J Biol Chem 2007;282:18777–18785.

48. Freeman MR, Solomon KR. Cholesterol and prostate cancer.J Cell Biochem 2004;91:54–69.

49. Hager MH, Solomon KR, Freeman MR. The role of cholesterolin prostate cancer. Curr Opin Clin Nutr Metab Care 2006;9:379–385.

50. Lin Y, Fukuchi J, Hiipakka RA, Kokontis JM, Xiang J. Up-regulation of Bcl-2 is required for the progression of prostatecancer cells from an androgen-dependent to an androgen-independent growth stage. Cell Res 2007;17:531–536.

51. Wang Y, Kreisberg JI, Ghosh PM. Cross-talk between theandrogen receptor and the phosphatidylinositol 3-kinase/Aktpathway in prostate cancer. Curr Cancer Drug Targets 2007;7:591–604.

52. Miyake H, Nelson C, Rennie PS, Gleave ME. Testosterone-repressed prostate message-2 is an antiapoptotic gene involvedin progression to androgen independence in prostate cancer.Cancer Res 2000;60:170–176.

53. Culig Z. Androgen receptor cross-talk with cell signallingpathways. Growth Factors 2004;22:179–184.

54. Craft N, Shostak Y, Carey M, Sawyers CL. A mechanism forhormone-independent prostate cancer through modulation ofandrogen receptor signaling by the HER-2/neu tyrosine kinase.Nat Med 1999;5:280–285.

55. Urbanucci A, Waltering KK, Suikki HE, Helenius MA, VisakorpiT. Androgen regulation of the androgen receptor coregulators.BMC Cancer 2008;8:219.

56. Yanase T, Fan W. Modification of androgen receptor function byIGF-1 signaling implications in the mechanism of refractoryprostate carcinoma. Vitam Horm 2009;80:649–666.

The Prostate

FattyAcids and Steroidogenesis in Prostate Cancer 251