Journal of Cellular Biochemistry 93:233–241 (2004)

Prostate Specific Antigen Gene Regulation byAndrogen Receptor

Joshua Kim and Gerhard A. Coetzee*

Departments of Molecular Microbiology and Immunology, Urology and Preventive Medicine,USC/Norris Comprehensive Cancer Center, Keck School of Medicine, University of Southern California,Los Angeles, California

Abstract Prostate specific antigen (PSA) is a serine protease that is synthesized by both normal and malignantepithelial cells of the human prostate. PSA expressed by malignant cells, however, are released into the serum at anincreased level, which can be detected to diagnose and monitor prostate cancer. Moreover, increases in serum PSAfollowing local and systemic treatments are highly correlatedwith tumor recurrence and progression, and this associationhas further established PSAas a clinically important biomarker. The expression of PSA ismainly inducedby androgens andregulated by the androgen receptor (AR) at the transcriptional level. Extensive research on the regulation of PSA geneexpressionhas provided significant information about the functionofAR,which is a crucial transcription factor involved inall phases of prostate cancer. Still, the molecular mechanism(s) by which the transcription of the PSA gene escapesregulation in advanced prostate cancer has yet to be clearly defined. Accumulating evidence suggests that a number ofprocesses including androgen-independent activation of AR are involved. Lacking an effective treatment, advancedprostate cancer is almost invariably fatal, which highlights the importance of elucidating mechanisms of tumorprogression. Insights into AR activity at the PSA gene could be extended to transcriptional regulation of other AR targetgenes, which may be crucial in discerning prostate cancer progression. Ultimately, our improved understanding of AR-regulated PSA expression could aid in developing viable therapies in treating and/or preventing advanced prostate cancer.J. Cell. Biochem. 93: 233–241, 2004. � 2004 Wiley-Liss, Inc.

Key words: PSA; prostate cancer; androgen receptor; transcription; gene regulation

Androgen action, mediated by the androgenreceptor (AR), is essential for normal growth,differentiation, and maintenance of the pros-tate epithelium. The majority of prostate can-cers are also androgen-dependent and respondinitially to androgen ablation therapy, which isthe only effective systemic treatment currentlyavailable [Denmeade and Isaacs, 2002]. Despitean initial favorable response, androgen ablationis essentially palliative, anddiseaseprogressioneventually ensues [Kozlowski et al., 1991].

Prostate cancers that recur after androgenablation are often responsive to subsequentalternative hormone manipulations that de-pend on a functional AR. However, tumorssurviving multiple ablation therapies even-tually progress to androgen-independence,at which point the tumors are far more ag-gressive and are invariably fatal. Althoughthese hormone-refractory prostate cancers areoperationally defined as being androgen-independent, they are thought to still dependon AR signaling for growth and continue toexpress high levels of AR and AR-regulatedgenes, such as prostate specific antigen (PSA).The mechanisms of AR activation in theseclinically androgen-independent tumors areunclear and may represent critical processesin disease progression. Nevertheless, there area number of proposed pathways that cancercells may utilize to stimulate AR-mediatedtranscription under low androgen condition,and these include AR gene mutation andamplification, co-regulator alterations, and

� 2004 Wiley-Liss, Inc.

Grant sponsor: NIH/NCI; Grant number: R01CA84890;Grant sponsor: United States Department of Defense;Grant number: DAMD17-00-1-0102; Grant sponsor: Pros-tate Cancer Foundation.

*Correspondence to: Gerhard A. Coetzee, USC/NorrisCancer Center, NOR 6411, 1441 Eastlake Ave, Los Angeles,CA 90089. E-mail: coetzee@usc.edu

Received 4 June 2004; Accepted 7 June 2004

DOI 10.1002/jcb.20228

androgen-independent activation of the AR[Weber and Gioeli, 2004]. Although thesemechanismsmay operate simultaneously, emer-ging evidence suggests that the androgen-inde-pendent activation of the AR through cross-talkwith various signal transduction pathways is amajor contributor to this process. The key tounderstanding the molecular progression toandrogen-independent prostate cancer may beto delineate the mechanisms involved in activa-tion of AR-regulated genes such as PSA.

ANDROGEN RECEPTOR STRUCTUREAND MECHANISM OF ACTION

Substantial evidence amassed over the past30 years since its identification indicates thatAR is one of the putative susceptibility genes forprostate cancer. The AR gene encodes thetranscription factor that regulates the expres-sion of androgen-responsive genes involved inprostate epithelial cell division and differentia-tion. Moreover, both normal and malignantprostate epithelial cells are dependent on theARsignaling axis for growth, andARexpressionis maintained in all forms of prostatic tumors[Buchanan et al., 2001; Feldman and Feldman,2001]. The AR gene is located on the long arm ofthe X chromosome (Xq11-12) and compriseseight exons that encode an mRNA of 11 kb. ThemRNAhas a 2.8 kb open reading frame, a 1.1 kb50 untranslated region (UTR), anda6kb30 UTR.AR protein has four domains: an N-terminaldomain (TAD) involved principally in transcrip-tional activation, aDNA-binding domain (DBD)that is required for interaction with specificgene sequences, a so-called hinge region, and aC-terminal ligand-binding domain (LBD) thatbinds androgens.

Testosterone is the main circulating andro-gen inmales that is secreted primarily by testesand circulates in the blood, where it is bound tosex hormone binding globulin (SHBG). Uponentering prostate cells, 90% of testosterone isconverted to the more active hormone, dihydro-testosterone (DHT), by the enzyme 5 alpha-reductase [Feldman and Feldman, 2001]. Simi-lar to othernuclear receptors, theAR is bound toa multi-protein (including heat shock proteins)chaperone complex in its inactive state. Andro-gen binding induces a conformational change inthe AR that leads to dissociation from thechaperone complex and phosphorylation of thereceptor. This conformation change also facil-

itates dimerization of the AR, which then bindsto the major groove of specific DNA sequencescalled androgen response elements (AREs) inthe 50 regulatory regions of target genes.

Subsequently, the ARE-bound AR homodi-mer recruits coactivators, such as the p160 andp300/CBP, which bridge interactions with thegeneral transcription machinery and modifyhistones, thus activating gene expression[Rosenfeld and Glass, 2001]. Although manyAR target genes have been identified to date,none has been as extensively studied as PSA.

PSA STRUCTURE AND FUNCTION

ThePSA gene is amember of a family of genesencoding kallikrein-like serine proteases. Atotal of 15 tissue kallikrein genes have beenidentified through sequencing of the humangenome, and they are clustered in the kallikreinlocus at chromosome 19q13.3-13.4 [Diamandisand Yousef, 2001]. All of the kallikrein genesencode five exons of similar size and have signi-ficant homology with each other (40–80%).Many of these genes, such as KLK2, KLK3(PSA), and KLK4, are regulated by steroidhormones.

The PSA gene encodes a 33 kDa glycoprotein,which consists of a single polypeptide chain of240 amino acids. In the normal prostate, PSA issecreted into the glandular ducts where itfunctions to degrade high molecular weightproteins synthesized in the seminal vesicles toinhibit coagulation of the semen [Balk et al.,2003]. PSA only enters the serum throughleakage into the extracellular fluid of thenormal prostate. However, during tumorigen-esis, serum PSA levels are elevated due to theloss of the normal glandular architecture. As aresult of being complexed to protease inhibitorssuch as alpha 1-antichymotrypsin, PSA in thecirculation is usually inactive [Stenman et al.,1991]. The proteolytic activity of PSA may gobeyond its normal function in prostate tumorsand could possibly contribute to cancer progres-sion. PSA has been shown to reduce the affinityof insulin-like growth factor binding protein-3(IGFBP-3) for insulin-like growth factor-1 (IGF-1) through the proteolysis of IGFBP-3, whichresults in increased availability of mitogenicIGF-1 [Cohen et al., 1994]. Furthermore, theelevated levels of IGF-1 increased proliferationof cultured prostate epithelial cells. Also, PSAitself may increase the growth of primary tissue

234 Kim and Coetzee

cultures of prostatic epithelial cells. However,other in vitro experiments have failed todemonstrate a correlation between PSA func-tion and tumor promoting activities [Balk et al.,2003]. In addition, human prostate cancer celllines that do not express PSA, such as PC-3 andDU-145, are more tumorigenic and invasivethan the LNCaP cell line, which produces PSA.Such conflicting results make it difficult toestablish whether PSA has an active biologicalrole in promoting the growth or progression ofprostate cancer, and the significance of thesemechanisms remains to be determined.

ANDROGEN REGULATEDEXPRESSION OF PSA

The expression of PSA is primarily activatedby androgens and regulated by the AR at thetranscriptional level. AR regulated expressionof PSA is mediated through AREs in the pro-ximal promoter (�600 to þ12) and the 50

upstream enhancer (�3875 to �4325) of thePSA gene (Fig. 1). The proximal promotercontains a TATA box (�28/�23), ARE I (�170/�156), ARE II—also known as androgenresponse region (ARR;�395/�376), and SP-1(�43/�54, �84/�93, �113/�123) [Sadar et al.,1999]. ARE I is a high affinity AR bindingsequence (AGAACAgcaAGTGCT), which is clo-sely related to the ARE consensus sequence(GGTACAnnnTGTTCT) [Riegman et al., 1991].In contrast, ARR is a weak nonconsensus AREsequence (GGATCAgggAGTCTC), which maycooperate with other AREs to increase thetranscriptional potential of the promoter[Cleutjens et al., 1996]. The upstream enhancerwas shown to be required for high androgen-stimulated PSA expression and was found tocontain a single strong consensusARE (AREIII;GGAACAtatTGTATC) [Cleutjens et al., 1996]

and multiple additional weak nonconsensusAREs, which may also contribute to highandrogen activity [Huang et al., 1999]. Anotherregulatory site present at the enhancer isthe putative cAMP response element (CRE;TGACGTCA) located at �3196/�3189 [Sadaret al., 1999]. Although cAMP strongly inducesthe expression of the PSA gene by an AR-dependent pathway, it is not known whethertranscription factors that are responsive tocAMP, such as cAMP responsive element bind-ingprotein (CREB)andactivating transcriptionfactor (ATF), regulate PSA gene expression bybinding to this site.

Activation of PSA gene transcription involvesmany different proteins that are recruiteddirectly (primary coactivators) and indirectly(secondary coactivators and general transcrip-tion machinery) by the AR. Primary coactiva-tors include three related genes that encode thep160 coactivators, referred toasSRC-1,GRIP-1/TIF2, and pCIP/ACTR/RAC/AIB-1/TRAM-1, aswell as Swi/Snf and TRAP/DRIP [Rosenfeldand Glass, 2001]. In addition, a number of ARcoactivators have been identified that interactdirectly with AR, including a series of AR asso-ciated proteins (ARAs), gelsolin, and b-catenin[Rahman et al., 2004]. Also, it was shown thatthe tumor suppressor, BRCA1, can enhanceAR-dependent transactivation by directly bindingto AR and GRIP-1, which suggests an interest-ing interplay between a tumor suppressor andAR [Park et al., 2000]. Thus, the primarycoactivators generally function by stabilizingthe receptor complex and recruiting othercoactivators. On the other hand, secondarycoactivators, such as CREB binding protein(CBP)/p300 and coactivator-associated argininemethyltransferase 1(CARM1), promote tran-scription by binding to the p160 coactivatorsand covalently modifying the local histones,

Fig. 1. The proximal promoter and upstream enhancer comprise the PSA 50 regulatory regions that arerequired for high androgen-stimulated expression.

PSA Gene Regulation 235

which allow access to the basal transcrip-tional machinery [Rosenfeld and Glass, 2001].These coactivators may also stabilize the pre-initiation complex by interacting directly withthe components of the basal transcriptionalmachinery (Fig. 2). Consistent with this model,it was recently reported that androgen inducesrobust recruitment of AR, p160 coactivators,and CBP/p300 specifically to the PSA enhancer[Louie et al., 2003]. Furthermore, RNA poly-merase II (pol II) was shown to be recruitedto the enhancer independent of the proximalpromoter, which taken together, support amodel in which AR and associated coactivatorsmediate transcription initiation by serving as astaging platform for Pol II recruitment.

CHROMATIN MODIFICATIONAT THE PSA GENE

The molecular processes that lead to activa-tion or repression of transcription are emergingas the foundation for gene regulation. There hasbeen a rapid accumulation of compelling evid-ence that points toward chromatin modifica-tion, a process that alters the transcriptionpotential of nucleosomes, as a key determinantin gene regulation. It is becoming apparent thataccessibility and activation of genes are largelydependent on diverse post-translational mod-ifications of histone amino-termini. These mod-ifications include acetylation, phosphorylation,

and methylation which covalently add acetyl-,phospho-, or methyl-groups to specific residuesof the N-terminal tail of core histones [Zhangand Reinberg, 2001]. The best characterized ofthemodifications is the acetylation of lysines onhistones H3 and H4, which is correlated totranscriptional activation. Acetylation, cata-lyzed by histone acetyltransferases (HAT) suchas p300/CBP, is reversed by the function ofhistone deacetylases (HDAC), which mediatetranscriptional repression [Johnstone, 2002].Other covalent modifications including phos-phorylation of histone H3 at serine 10 andmethylation of histones H3 and H4 are emerg-ing as important factors that determine tran-scriptional status. Furthermore, the complexinteractions between the different histone tailmodificationshave led to theproposal of ‘histonecode hypothesis,’ which suggests that specifichistone modifications affect and interact withother histone modifications and serve as marksfor the recruitment of associated factors thatregulate chromatin functions [Strahl and Allis,2000].

While most of the studies on chromatinmodifications have been performed on lowerorganisms such as Saccharomyces cerevisiaeand Drosophila melaganoster, only recentlychromatin modifications at the PSA gene inhuman were investigated. It was shown thathistone H3 acetylation followed AR occupancyat the PSA promoter and enhancer, which is

Fig. 2. Activation of PSA gene transcription involves AR-mediated recruitment of coactivators, generaltranscription factors (GTFs), and RNA polymerase II (RNAP II). The PSA promoter and enhancer maycooperate in proximity to activate transcription by bridged AR/coactivators/RNA polymerase II complexes.

236 Kim and Coetzee

consistent with p300/CBP recruitment to thoseregions by AR/p160 complex [Jia et al., 2003;Louie et al., 2003]. Also, AR-mediated tran-scription was accompanied by rapid decreasesin di- and tri-methylated H3 at lysine 4 (K4) atboth 50 regulatory regions [Kim et al., 2003],which is contrary to the data in Saccharomycescerevisiae, where tri-methylated H3-K4 isexclusively associated with active chromatin[Santos-Rosa et al., 2002]. The results maysuggest a novel role for methylated H3-K4 intranscriptional regulation, possibly pointing toa yet to be discovered histone demethylase. Analternative hypothesis may be that methylatedH3-K4 acts as a binding site for a proteinassociated with active transcription (possibly acoactivator), and thus the decrease in methy-lated H3-K4 observed with transcriptionalactivation by androgen is actually due to anti-body epitope masking. Still another hypothesisis that the adjacent threonine 3 (H3-T3) could bephosphorylated, which may sterically hinderthe methylation at H3-K4 (personal commu-nication, C.D. Allis and J. Rice). Thus, thefunctional consequences of specific histonemodifications at the PSA gene still remain tobe determined. However, it is clear that ourunderstanding of the histone modificationswould provide important clues to the transcrip-tional regulation at the PSA gene.

PSA POLYMORHISMS ANDPROSTATE CANCER RISK

The inevitable failure of androgen ablationtherapy has led to strategies of developingmethods to aggressively screen for cancers thatare still confined locally to the prostate andso are potentially curable by specific treat-ment. Consequently, the correlation that existsbetween PSA expression level and prostatecancer development has established PSA asthe most widely used biomarker for detectingand monitoring prostate cancer. However, therole of PSA in screening and staging of prostatecancer remains limited because of the broad andoverlapping ranges in serum PSA levels thatexist in many patients with either localized oradvanced disease. Any condition that increasesthe volume of the prostate or disrupts theprostatic architecture, including benign pro-static hyperplasia and prostatitis, can elevateserum PSA levels [Balk et al., 2003]. Further-more, factors such as age and race also have

been shown to be associated with serum PSAlevels, possibly due to differences in prostatevolume.

Accruing evidence suggests that geneticvariation in the promoter of the PSA gene mayalso contribute to individual variation in serumPSA levels. For example, serum PSA levels inhealthy men are associated with a G/A singlenucleotide polymorphism (SNP) at ARE I (atposition �158) and/or the number of CAGrepeats in exon 1 of the AR gene [Xue et al.,2001]. It was found that PSA levels are higheramong healthy men with the AA genotype (atARE I) compared to men with the AG or GGgenotypes. The same SNP was reported to beassociated with prostate cancer risk, with theGG genotype at significantly increased risk foradvanced cancer [Xue et al., 2000]. Further-more, shorter AR exon CAG repeats hadincreasing PSA effect on those genotypes ofboth healthy men and prostate cancer patients,suggesting a probable gene–gene interaction inthe etiology of prostate cancer. However, otherstudies, involving separate groups of men, haveshown that either the PSA AA genotype (asopposed to GG) was associated with increasedcancer risk or not associated with prostatecancer at all [Medeiros et al., 2002; Xu et al.,2002]. These divergent resultsmaybe due to thedifferent populations that exhibit disparatelevels of genetic variation, environmental fac-tors, and gene–environment interactions.Alternatively, a possible explanation comesfrom a study, which reported that additionalSNPs present at the PSA enhancer (G/A at�4643, C/T at �5412, and T/G at �5429) areassociated with serum PSA levels in healthymen, suggesting that there may be a linkagedisequilibrium of the PSA promoter polymorph-ism with those at the enhancer region [Crameret al., 2003]. Still, further studies are neededto clarify the potential association betweengenetic variation and prostate cancer risk.

ANDROGEN-INDEPENDENTINDUCTION OF PSA

It has been hypothesized that continuedsignaling of the AR in a castrate hormoneenvironment could result from overexpressionof the receptor, gain-of-function AR genesomatic mutations, coactivator overexpression,and ligand-independent activation of the AR[Buchanan et al., 2001; Feldman and Feldman,

PSA Gene Regulation 237

2001; Balk, 2002]. As a result, the tumor cellsmay continue to proliferate and avert apoptosisby exploiting other available steroid hormonesand growth factors in low androgen conditions.

Prostate tumors could survive and proliferatein the altered hormonal environment possiblyby interactions between growth factor- andcytokine-activated pathways and AR signal-ing. Indeed, recent evidence indicates that anumber of growth factors and cytokines suchas insulin-like growth factor 1 (IGF-1), kerati-nocyte growth factor (KGF), epidermal growthfactor (EGF), forskolin (FSK), and interleukin 6(IL-6) have been shown to stimulate AR signal-ing in the absence of androgen [Culig et al.,1994; Nazareth and Weigel, 1996; Ueda et al.,2002]. Various signal transduction pathwaysand mechanisms of activation seem to beinvolved in the function of these alternative

ligands (Fig. 3). Both EGF and FSK increaseAR phosphorylation at serine 650, which couldstimulate AR transactivation [Gioeli et al.,2002]. These signaling molecules may also actthrough divergent mechanisms. EGF signal-ing through mitogen activated protein kinase(MAPK) pathway was shown to increase TIF2/GRIP1 coactivation of AR transactivation inrecurrent prostate cancer [Gregory et al., 2004],whileFSK is able to inducePSAgene expressionin LNCaP cells by activating the amino-termi-nus of the AR and increasing protein–DNAcomplex formation between AR and PSA-ARE[Sadar et al., 1999]. Another growth factor, IGF-1, has been implicated in prostate cancer cellproliferation as mentioned above. Consistentwith a positive correlation between elevatedserum IGF-1 levels and increased prostatecancer risk [Chan et al., 1998], IGF-1 enhanced

Fig. 3. Androgen-independent activation of PSA gene transcription via various signal transductionpathways.

238 Kim and Coetzee

the magnitude of the AR response in thepresence of low levels of androgen, suggestingthat the cross-talk between AR and growthfactor signaling pathways may sensitize AR tosuboptimal stimulation by low levels of andro-gens [Orio et al., 2002].Besides growth factors, cytokines, such as

IL-6, are involved in prostate cancer progres-sion. IL-6 levels in patients with therapy re-sistant prostate tumors were elevated [Twillieet al., 1995], and moreover, IL-6 elevation over7 pg/ml was found to be associated with poorprognosis in multivariate analysis [Nakashimaetal., 2000]. Spurredby these clinical findings, anumber of groups reported that IL-6 enhancesAR transactivation. IL-6 was shown to enhanceAR transactivation via either signal transdu-cers and activators of transcription-3 (STAT3)or MAPK pathways [Chen et al., 2000; Uedaet al., 2002]. It was further proposed that CBP/p300 mediates this IL-6 effect [Debes et al.,2002]. Moreover, Chang et al. found thatwhereas IL-6 enhanced AR transactivationactivity via either the STAT3 or MAPK path-ways, it suppressed AR transactivation activityvia the Akt pathway [Yang et al., 2003]. Still,another study indicates that IL-6 can induceandrogen responsiveness in prostate cancercells through upregulation of AR expression[Lin et al., 2001]. However, recent findingsindicate that IL-6 can also inhibit PSA expres-sion and prostate cancer cell proliferationthrough inhibition of p300 recruitment to thePSAenhancer andpromoter [Jia et al., 2003; Jiaet al., 2004]. Furthermore, the inhibitory activ-ity of IL-6 is, at least in part, mediated throughthe STAT3 pathway without MAPK involve-ment. These conflicting reports may be explain-ed by a finding that long-term treatment ofthe prostate cancer cell line, LNCaP, withIL-6 abolishes its inhibitory growth response[Hobisch et al., 2001]. Thus, the stimulatoryeffects reported may be due to the transition ofIL-6 from a paracrine growth inhibitor to anautocrine growth stimulator [Chung et al.,1999].

CONCLUSIONS

PSA has been widely investigated as a modelgene for determining the mechanism by whichAR-mediated transactivation occurs in normalandmalignant prostate cells. It is clear, though,that many molecular aspects of AR-mediated

expression of PSA remain to be elucidated inorder to delineate the mechanisms of prostatecancer progression. Furthermore, the high levelof PSA expression associated with tumor pro-gression, taken together with its proteolyticactivity, suggests the possibility that PSA mayplay a functional role in progression itself, byeither direct or indirect stimulation of cancercell growth or invasion. Continued expansion ofour understanding of functions of AR and PSAwill undoubtedly contribute to more effectivetherapies, as current therapies are usuallybeneficial in the earlier stages of cancer butare essentially palliative in more advanced orrecurring androgen-insensitive tumors. More-over, as the relationships between variouscytokines, growth factors, and AR as well asthe physiologic manifestations of androgenaction are understood, the possible targets forpharmacologic intervention will be clarified.These imperative tasks represent an excitingchallenge in prostate cancer research.

ACKNOWLEDGMENTS

We thank Dr. Michael Stallcup, Dr. LucioComai, and Dr. Grant Buchanan for criticallyreading the manuscript.

REFERENCES

Balk S. 2002. Androgen receptor as a target in androgen-independent prostate cancer. Urology 60:132–138.

Balk SP, Ko YJ, Bubley GJ. 2003. Biology of prostate-specific antigen. J Clin Oncol 21:383–391.

Buchanan G, Irvine RA, Coetzee GA, Tilley WD. 2001.Contribution of the androgen receptor to prostate cancerpredisposition and progression. Cancer Metastasis Rev20:207–223.

Chan JM, Stampfer MJ, Giovannucci E, Gann PH, Ma J,Wilkinson P, Hennekens CH, Pollak M. 1998. Plasmainsulin-like growth factor-I and prostate cancer risk: Aprospective study. Science 279:563–566.

Chen T, Wang LH, Farrar WL. 2000. Interleukin 6activates androgen receptor-mediated gene expressionthrough a signal transducer and activator of transcrip-tion 3-dependent pathway in LNCaP prostate cancercells. Cancer Res 60:2132–2135.

Chung TD, Yu JJ, Spiotto MT, Bartkowski M, Simons JW.1999. Characterization of the role of IL-6 in theprogression of prostate cancer. Prostate 38:199–207.

Cleutjens KB, van Eekelen CC, van der Korput HA,Brinkmann AO, Trapman J. 1996. Two androgen re-sponse regions cooperate in steroid hormone regulatedactivity of the prostate-specific antigen promoter. J BiolChem 271:6379–6388.

Cohen P, Peehl DM, Graves HC, Rosenfeld RG. 1994.Biological effects of prostate specific antigen as aninsulin-like growth factor binding protein-3 protease.J Endocrinol 142:407–415.

PSA Gene Regulation 239

Cramer SD, Chang BL, Rao A, Hawkins GA, Zheng SL,Wade WN, Cooke RT, Thomas LN, Bleecker ER,Catalona WJ, Sterling DA, Meyers DA, Ohar J, Xu J.2003. Association between genetic polymorphisms in theprostate-specific antigen gene promoter and serumprostate-specific antigen levels. J Natl Cancer Inst 95:1044–1053.

Culig Z, Hobisch A, Cronauer MV, Radmayr C, Trapman J,Hittmair A, Bartsch G, Klocker H. 1994. Androgenreceptor activation in prostatic tumor cell lines byinsulin-like growth factor-I, keratinocyte growth factor,and epidermal growth factor. Cancer Res 54:5474–5478.

Debes JD, Schmidt LJ, Huang H, Tindall DJ. 2002. p300mediates androgen-independent transactivation of theandrogen receptor by interleukin 6. Cancer Res 62:5632–5636.

Denmeade SR, Isaacs JT. 2002. A history of prostate cancertreatment. Nat Rev Cancer 2:389–396.

Diamandis EP, Yousef GM. 2001. Human tissue kallikreingene family: A rich source of novel disease biomarkers.Expert Rev Mol Diagn 1:182–190.

Feldman BJ, Feldman D. 2001. The development ofandrogen-independent prostate cancer. Nat Rev Cancer1:34–45.

Gioeli D, Ficarro SB, Kwiek JJ, Aaronson D, Hancock M,Catling AD, White FM, Christian RE, Settlage RE,Shabanowitz J, Hunt DF, Weber MJ. 2002. Androgenreceptor phosphorylation. Regulation and identificationof the phosphorylation sites. J Biol Chem 277:29304–29314.

Gregory CW, Fei X, Ponguta LA, He B, Bill HM, French FS,Wilson EM. 2004. Epidermal growth factor increasescoactivation of the androgen receptor in recurrentprostate cancer. J Biol Chem 279:7119–7130.

Hobisch A, Ramoner R, Fuchs D, Godoy-Tundidor S,Bartsch G, Klocker H, Culig Z. 2001. Prostate cancercells (LNCaP) generated after long-term interleukin 6(IL-6) treatment express IL-6 and acquire an IL-6partially resistant phenotype. Clin Cancer Res 7:2941–2948.

Huang W, Shostak Y, Tarr P, Sawyers C, Carey M. 1999.Cooperative assembly of androgen receptor into anucleoprotein complex that regulates the prostate-spe-cific antigen enhancer. J Biol Chem 274:25756–25768.

Jia L, Kim J, Shen H, Clark PE, Tilley WD, Coetzee GA.2003. Androgen receptor activity at the prostate specificantigen locus: Steroidal and non-steroidal mechanisms.Mol Cancer Res 1:385–392.

Jia L, Choong CS, Ricciardelli C, Kim J, Tilley WD, CoetzeeGA. 2004. Androgen receptor signaling: Mechanism ofinterleukin-6 inhibition. Cancer Res 64:2619–2626.

Johnstone RW. 2002. Histone-deacetylase inhibitors: Noveldrugs for the treatment of cancer. Nat Rev Drug Discov1:287–299.

Kim J, Jia L, Tilley WD, Coetzee GA. 2003. Dynamicmethylation of histone H3 at lysine 4 in transcriptionalregulation by the androgen receptor. Nucleic Acids Res31:6741–6747.

Kozlowski JM, Ellis WJ, Grayhack JT. 1991. Advancedprostatic carcinoma. Early versus late endocrine therapy.Urol Clin North Am 18:15–24.

Lin DL, WhitneyMC, Yao Z, Keller ET. 2001. Interleukin-6induces androgen responsiveness in prostate cancer cells

through up-regulation of androgen receptor expression.Clin Cancer Res 7:1773–1781.

Louie MC, Yang HQ, Ma AH, Xu W, Zou JX, Kung HJ,Chen HW. 2003. Androgen-induced recruitment of RNApolymerase II to a nuclear receptor-p160 coactivatorcomplex. Proc Natl Acad Sci USA 100:2226–2230.

Medeiros R, Morais A, Vasconcelos A, Costa S, Pinto D,Oliveira J, Carvalho R, Lopes C. 2002. Linkage betweenpolymorphisms in the prostate specific antigen ARE1gene region, prostate cancer risk, and circulating tumorcells. Prostate 53:88–94.

Nakashima J, Tachibana M, Horiguchi Y, Oya M, OhigashiT, Asakura H, Murai M. 2000. Serum interleukin 6 as aprognostic factor in patients with prostate cancer. ClinCancer Res 6:2702–2706.

Nazareth LV, Weigel NL. 1996. Activation of the humanandrogen receptor through a protein kinase A signalingpathway. J Biol Chem 271:19900–19907.

Orio F, Jr., Terouanne B, Georget V, Lumbroso S, AvancesC, Siatka C, Sultan C. 2002. Potential action of IGF-1 andEGF on androgen receptor nuclear transfer and trans-activation in normal and cancer human prostate celllines. Mol Cell Endocrinol 198:105–114.

Park JJ, Irvine RA, Buchanan G, Koh SS, Park JM, TilleyWD, Stallcup MR, Press MF, Coetzee GA. 2000. Breastcancer susceptibility gene 1 (BRCAI) is a coactivator ofthe androgen receptor. Cancer Res 60:5946–5949.

Rahman M, Miyamoto H, Chang C. 2004. Androgenreceptor coregulators in prostate cancer: Mechanismsand clinical implications. Clin Cancer Res 10:2208–2219.

Riegman PH, Vlietstra RJ, van der Korput JA, BrinkmannAO, Trapman J. 1991. The promoter of the prostate-specific antigen gene contains a functional androgenresponsive element. Mol Endocrinol 5:1921–1930.

Rosenfeld MG, Glass CK. 2001. Coregulator codes of tran-scriptional regulation by nuclear receptors. J Biol Chem276:36865–36868.

Sadar MD, Hussain M, Bruchovsky N. 1999. Prostatecancer: Molecular biology of early progression to andro-gen independence. Endocr Relat Cancer 6:487–502.

Santos-Rosa H, Schneider R, Bannister AJ, Sherriff J,Bernstein BE, Emre NC, Schreiber SL, Mellor J,Kouzarides T. 2002. Active genes are tri-methylated atK4 of histone H3. Nature 419:407–411.

Stenman UH, Leinonen J, Alfthan H, Rannikko S,Tuhkanen K, Alfthan O. 1991. A complex betweenprostate-specific antigen and alpha 1-antichymotrypsinis the major form of prostate-specific antigen in serum ofpatients with prostatic cancer: Assay of the compleximproves clinical sensitivity for cancer. Cancer Res 51:222–226.

Strahl BD, Allis CD. 2000. The language of covalent histonemodifications. Nature 403:41–45.

Twillie DA, Eisenberger MA, Carducci MA, Hseih WS,Kim WY, Simons JW. 1995. Interleukin-6: A candidatemediator of human prostate cancer morbidity. Urology45:542–549.

Ueda T, Mawji NR, Bruchovsky N, Sadar MD. 2002.Ligand-independent activation of the androgen receptorby IL-6 and the role of the coactivator SRC-1 in prostatecancer cells. J Biol Chem 277:38087–38094.

Weber MJ, Gioeli D. 2004. Ras signaling in prostate cancerprogression. J Cell Biochem 91:13–25.

240 Kim and Coetzee

Xu J, Meyers DA, Sterling DA, Zheng SL, Catalona WJ,Cramer SD, Bleecker ER, Ohar J. 2002. Associationstudies of serum prostate-specific antigen levels and thegenetic polymorphisms at the androgen receptor andprostate-specific antigen genes. Cancer Epidemiol Bio-markers Prev 11:664–669.

XueW, Irvine RA, YuMC, Ross RK, Coetzee GA, Ingles SA.2000. Susceptibility to prostate cancer: Interactionbetween genotypes at the androgen receptor and pros-tate-specific antigen loci. Cancer Res 60:839–841.

Xue WM, Coetzee GA, Ross RK, Irvine R, Kolonel L,Henderson BE, Ingles SA. 2001. Genetic determinants ofserum prostate-specific antigen levels in healthy men

from a multiethnic cohort. Cancer Epidemiol BiomarkersPrev 10:575–579.

Yang L, Wang L, Lin HK, Kan PY, Xie S, Tsai MY, WangPH, Chen YT, Chang C. 2003. Interleukin-6 differentiallyregulates androgen receptor transactivation via PI3K-Akt, STAT3, and MAPK, three distinct signal pathwaysin prostate cancer cells. Biochem Biophys Res Commun305:462–469.

Zhang Y, Reinberg D. 2001. Transcription regulation byhistone methylation: Interplay between different cova-lent modifications of the core histone tails. Genes Dev15:2343–2360.

PSA Gene Regulation 241