Androgen Action in Prostate Cancer

Androgen Action in Prostate Cancer Donald Tindall l James Mohler Editors Androgen Action in Prostate Cancer Edi...

Author: Donald Tindall | James Mohler (Editors)

166 downloads 1340 Views 8MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

Androgen Action in Prostate Cancer

Donald Tindall

l

James Mohler

Editors

Androgen Action in Prostate Cancer

Editors Donald Tindall 200 1st Street SW., Rochester MN 55905 USA [email protected]

James Mohler Elm and Carlton Streets Buffalo NY 14263 USA [email protected]

ISBN 978-0-387-69177-0 e-ISBN 978-0-387-69179-4 DOI: 10.1007/978-0-387-69179-4 Springer Dordrecht Heidelberg London New York Library of Congress Control Number: 2008942188 # Springer Science+Business Media, LLC 2009 All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. Printed on acid-free paper Springer is part of Springer Science+Business Media (www.springer.com)

Contents

Introduction: Proper Nomenclature Facilitates Clinical and Translational Research in Prostate Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 James L. Mohler and Donald J. Tindall Part I: The Role of Androgens in Developement of the Normal Prostate and Progression of Prostate Cancer Androgen Action and Modulation of Prostate and Prostate Cancer Growth: An Historical Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Shutsung Liao, John M. Kokontis, Chih-Pin Chuu, and Richard A. Hiipakka Clinical Progression to Castration-Recurrent Prostate Cancer . . . . . . . . . . . . . 57 Mark Pomerantz and Philip Kantoff Differential Roles of Androgen Receptor in Prostate Development and Cancer Progression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 Shuyuan Yeh, Yuanjie Niu, Hiroshi Miyamoto, Tamin Chang, and Chawnshang Chang Imaging Androgen Receptor Function In Vivo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 Michael Carey and Lily Wu Part II: Androgen Metabolism Increased Expression of Genes Converting Adrenal Androgens to Testosterone in Castration-Recurrent Prostate Cancer . . . . . . . . . . . . . . . . 123 Steven P. Balk

v

vi

Contents

Androgen-Metabolic Genes in Prostate Cancer Predisposition and Progression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141 Juergen K.V. Reichardt and Ann W. Hsing Effect of Steroid 5a-Reductase Inhibitors on Markers of Tumor Regression and Proliferation in Prostate Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . 155 Lynn N. Thomas and Roger S. Rittmaster 5a-Reductase Isozymes in Castration-Recurrent Prostate Cancer . . . . . . . 175 Mark A. Titus and James L. Mohler Improving Intermittent Androgen-Deprivation Therapy: OFF Cycle and the Role of Steroid 5a-Reductase Inhibitors . . . . . . . . . . . . . 187 Shubham Gupta, Daniel Shevrin, and Zhou Wang Part III: Androgen Receptor Structure/Function Relationships Insights from AR Gene Mutations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207 Grant Buchanan, Eleanor F. Need, Tina Bianco-Miotto, Norman M. Greenberg, Howard I. Scher, Margaret M. Centenera, Lisa M. Butler, Diane M. Robins, and Wayne D. Tilley Functional Motifs of the Androgen Receptor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241 Elizabeth M. Wilson The Role of the Androgen Receptor Polyglutamine Tract in Prostate Cancer: In Mice and Men . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269 Diane M. Robins The Androgen Receptor Coactivator-Binding Interface . . . . . . . . . . . . . . . . . . . 297 Eva Este´banez-Perpin˜a´ and Robert J. Fletterick Part IV: Co-Regulators of the Androgen Receptor Co-regulators and the Regulation of Androgen Receptor Action in Prostate Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315 Irina U. Agoulnik and Nancy L. Weigel Androgen Receptor Coregulators and Their Role in Prostate Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341 Latif A. Wafa, Robert Snoek, and Paul S. Rennie Interaction of the Androgen Receptor Ligand-Binding Domain with the N-Terminal Domain and with Coactivators . . . . . . . . . . . . . 375 Jan Trapman

Contents

vii

Multitasking and Interplay Between the Androgen Receptor Domains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385 F. Claessens, T. Tanner, and A. Haelens Chromatin Remodeling and Androgen Receptor-Mediated Transcription . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 405 Li Jia, Omar Khalid, Baruch Frenkel, and Gerhard A. Coetzee Part V: Ligand-Independent Activation of the Androgen Receptor Ligand-Independent Androgen Receptor Activity . . . . . . . . . . . . . . . . . . . . . . . . . 427 Scott M. Dehm and Donald J. Tindall Role of IL-6 in Regulating the Androgen Receptor . . . . . . . . . . . . . . . . . . . . . . . . 451 Zoran Culig and Alfred Hobisch The Role of Cyclic AMP in Regulating the Androgen Receptor . . . . . . . . . . 465 Marianne D. Sadar Part VI: Role of the Androgen Receptor in Prostate Cancer During Castration Cellular and Molecular Signatures of Androgen Ablation of Prostate Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 507 Clifford G. Tepper and Hsing-Jien Kung Tissue Levels of Androgens in Castration-Recurrent Prostate Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 553 James L. Mohler and Mark A. Titus Unique Effects of Wnt Signaling on Prostate Cancer Cells: Modulation of the Androgen Signaling Pathway by Interactions of the Androgen Receptor Gene and Protein with Key Components of the Canonical Wnt Signaling Pathway . . . . . . . . . . . . . . . . . . . . 569 Matthew J. Tanner, Elina Levina, Michael Shtutman, Mengqian Chen, Patrice Ohouo, and Ralph Buttyan The Role of Foxa Proteins in the Regulation of Androgen Receptor Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 587 David J. DeGraff, Xiuping Yu, Qian Sun, Janni Mirosevich, Ren Jie Jin, Yongqing Wang, Aparna Gupta, Srinivas Nandana, Thomas Case, Manik Paul, Hong-Ying Huang, Ellen Shapiro, Susan Logan, Kichiya Suzuki, Marie-Claire Orgebin-Crist, and Robert J. Matusik

viii

Contents

Part VII: Androgen-Regulated Genes during Prostate Cancer Progression Androgen Receptor as a Licensing Factor for DNA Replication . . . . . . . . . 619 Donald J. Vander Griend and John T. Isaacs Androgen-Regulated Genes in the Prostate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 631 Nigel Clegg and Peter S. Nelson Mapping the Androgen Receptor Cistrome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 663 Qianben Wang and Myles Brown Differential Regulation of Clusterin Isoforms by the Androgen Receptor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 681 Tanya K. Day, Colleen C. Nelson, and Martin E. Gleave Androgen Regulation of Prostate Cancer Gene Fusions . . . . . . . . . . . . . . . . . . . 701 Rou Wang, Scott A. Tomlins, and Arul M. Chinnaiyan Androgens and the Lipogenic Switch in Prostate Cancer . . . . . . . . . . . . . . . . . 723 Johannes V. Swinnen, Koen Brusselmans, Hannelore V. Heemers, and Guido Verhoeven Part VIII: The Androgen Receptor as a Drugable Target Molecular Biology of Novel Targets Identified Through Study of Castration-Recurrent Prostate Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . 743 Philip A. Watson and Charles L. Sawyers Selenium and Androgen Receptor in Prostate Cancer . . . . . . . . . . . . . . . . . . . . . 755 Nagalakshmi Nadiminty and Allen C. Gao Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 781

Introduction: Proper Nomenclature Facilitates Clinical and Translational Research in Prostate Cancer James L. Mohler and Donald J. Tindall

This book serves as a tribute to the work of the many investigators who have attempted to understand the role of the androgen receptor (AR) in the development and progression of prostate cancer. The relationship between prostate cancer and androgen status was initially appreciated in the late 1800s, and then awareness reawakened in the 1940s (Huggins and Hodges 2002) by the studies of Charles Huggins and colleagues, for which the Nobel prize was awarded in 1966. Androgen deprivation therapy by surgical or medical means using methods that remove the source of testicular androgens or inhibit the production of testicular androgens, respectively, produced remission in most, but not all, men. Moreover, clinical studies have cast doubt on whether ‘‘combined androgen blockade’’ (Labrie et al. 1982) is any more effective than androgen deprivation monotherapy (Prostate Cancer Trialists’ Collaborative Group, 1995), except when an antiandrogen is used to block testosterone flair in men with bulky metastatic disease. Nonetheless, responses to androgen deprivation therapy vary widely. In order to explain this clinical variation, it was proposed initially that prostate cancer consists of ‘‘androgen-dependent’’ and ‘‘androgen-independent’’ cells (Carter and Isaacs 1988; Isaacs and Coffey 1981). The argument went as follows: The relative distribution of prostate cancer cells between these two phenotypes determine whether there is a clinical response to androgen deprivation therapy and the duration of any response. If the tumor consists of androgen-dependent cells entirely, androgen deprivation therapy would prove curative since all androgen-dependent cells would undergo apoptosis upon androgen deprivation. If a prostate cancer is composed exclusively of androgenindependent cells, then no clinical response to androgen deprivation therapy would occur and that patient would ‘‘resist’’ androgen deprivation therapy and succumb to his prostate cancer as if androgen deprivation therapy had never been administered. Almost, if not, all patients fall between these two extremes and exhibit clinical remissions, which may be complete and sustained if the androgen-dependent cells far outnumber the androgenindependent cells. Because apoptosis does not occur in as high a proportion of cells as predicted, a third phenotype of prostate cancer cell was proposed, the androgen-sensitive cell (Pollack et al. 1983; Grossman 1986). The ‘‘androgen-sensitive’’ cell is responsive to androgen deprivation therapy by becoming quiescent until such time that androgens become available. Thus, the extent of regression of prostate cancer after androgen

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4, # Springer Science + Business Media, LLC 2009

1

2

J.L. Mohler, D.J. Tindall deprivation therapy is influenced by the relative proportion of androgen-dependent, androgen-sensitive and androgen-independent cells. The eventual emergence of the ‘‘androgen-independent phenotype’’ is predicted by the proportion of androgen-independent cells.

The development of prostate cancer cell lines, prostate cancer cell line xenografts, and human prostate cancer xenografts has enabled translational researchers to model manifestations of clinical disease progression. The availability of these models has resulted in increased complexity in terminology. Many commercially available tissue culture media contain levels of androgens sufficient to activate even a normal androgen receptor. However, a truly androgen-independent state can be achieved in a cell culture environment by careful charcoal stripping of media and the avoidance of inclusion of phenol red. Hence, investigators should prove that their androgen-free media is truly androgen free lest results be compromised. In the in vitro situation, proper terminology allows for androgen-dependent, androgensensitive, or androgen-independent growth. An androgen-dependent cell line is one that is comprised exclusively of androgen-dependent cells, which undergo apoptosis upon androgen withdrawal. Restoration of androgen after a certain period would have no effect since all cells would have died. Even the LNCaP cell line does not meet this definition. LNCaP cells vary their growth rate many fold depending on androgen levels and exhibit a well-known bimodal growth response to androgens; androgen levels either too low or too high inhibit growth (Gregory et al. 2001). Thus, LNCaP should be considered an androgen-sensitive cell line in which the vast majority undergoes apoptosis but some cells become quiescent, i.e., growth ceases until such time that androgens are restored in the media. LNCaP cells grown in progressively androgen-depleted media selects for ‘‘androgen-independent’’ variants. But androgen response may be even more complicated; although xenografted cells grow equally well in castrate or intact hosts, they decrease growth rate approximately twofold upon androgen stimulation (Lee et al. 2007). Other ‘‘androgen-independent’’ cells, such as CWR-R1 and LNCaP-C4-2, increase growth twofold when androgens are added to the media (Gregory et al. 2001). Even greater difficulties ensue when the terms androgen-dependent, androgensensitive, and androgen-independent are applied to xenografts and patients. The mouse has widely fluctuating levels of testosterone, and most investigators control testosterone levels by surgical castration and implantation of sustained-release pellets to provide physiologically relevant (human) androgen stimulation of xenografted cell lines or clinical specimens. However, the laboratory mouse differs markedly from the human. Testicular androgen levels in the castrate mouse are below the level of detection using radioimmunoassay and routine mass spectrometric measurement (Van Steenbrugge 1991; Jin et al. 2004). However, clinically significant circulating DHEA levels may be similar in intact and castrate BALB/c nude mice (Jin et al. 2004). Removal of sustained release androgen pellets creates an androgen deprivation therapy situation different than in the human due to the availability in the human of very high levels of circulating low-affinity andrenal androgens. In addition, many androgen-sensitive cell lines (LNCaP, LAPC4), androgen-dependent xenografts (CWR22, LUCaP), or clinical specimens (fresh

Introduction: Proper Nomenclature Facilitates Clinical and Translational Research

3

tissue samples) require orthotopic seeding, which is challenging technically, or coadministration with Matrigel1 for establishment and growth, even in the face of normal circulating levels of testicular androgens. In the human patient, advanced prostate cancer is treated with androgen deprivation therapy. In rare cases, no clinically measureable response will occur, and the tumor should be labeled appropriately ‘‘castration-resistant prostate cancer’’ (Scher and Sawyers 2005). However, most men experience a clinical and/or biochemical regression and, hence, their tumor should be considered either androgen-dependent or -sensitive. Unfortunately, an androgen-dependent tumor cannot be distinguished from an androgen-sensitive tumor in a human because of the inability to assess accurately the degree of apoptosis vs. cell quiescence, since biopsy specimens are not obtained routinely to measure apoptosis directly, and neither spatial nor functional imaging is reliable for accurate response assessment. Thus, the treatmentnaive prostate cancer is best called ‘‘androgen-stimulated,’’ thus avoiding an assumption of androgen dependence, sensitivity, or resistance. Almost no men with advanced prostate cancer are cured by androgen deprivation therapy. Thus, the relative distribution of malignant epithelial cells among the androgen-dependent, androgen-sensitive, and androgen-independent compartments is really only of academic interest. Of more practical interest is the proper terminology for prostate cancer that returns clinically during androgen deprivation therapy. What should the tumor be called when it recurs after some degree of response to androgen deprivation therapy? ‘‘Androgen-independent’’ is an inappropriate term since it is now well established that prostate cancer, which recurs during surgical or medical androgen deprivation therapy, has tissue levels of testicular androgens that are sufficient for activation of even a normal androgen receptor (Geller et al. 1979; Mohler et al. 2004; Titus et al. 2005). In fact, a recent publication has suggested that benign prostate tissue also contains levels of testicular androgens sufficient to activate the androgen receptor within 4 weeks after initiation of androgen deprivation therapy (Page et al. 2006). ‘‘Hormone refractory’’ is an equally imprecise terminology since the prostate cancer dependence upon androgens far outweighs the magnitude of its dependence of any of the other steroid hormones and, furthermore, some recurrent prostate cancers will respond to secondary treatment with nonandrogenic steroid hormones. When prostate cancer recurs after radiation therapy, it is often called ‘‘radiationrecurrent.’’ When prostate cancer recurs after radical prostatectomy, it is often deemed recurrent, which should more appropriately be termed ‘‘radical prostatectomy-recurrent.’’ Hence, a nomenclature is proposed when a patient with unsuccessfully treated, early-detected prostate cancer may eventually be described as having Gleason grade 4 + 3 = 7, radical prostatectomy-recurrent, salvage radiationrecurrent, castration-recurrent, anti-androgen-recurrent, taxotere-recurrent, and prednisone-recurrent prostate cancer. A similar nomenclature could be applied to xenografted cell lines or human tissue specimens. In contrast, when modeling prostate cancer in vitro, use of the term ‘‘androgen-independent’’ is appropriate when media has been properly charcoal-stripped by the investigator or androgen levels have been confirmed below the limit of detection using appropriate mass

4

J.L. Mohler, D.J. Tindall

Table 1 The terms used and preferred (bold) for clinical prostate cancer and its treatment Treatment Naı¨ve Treatment Clinically recurrent cancer Androgen-dependent Androgen ablation therapy Hormone-relapsed Androgen-sensitive Hormone therapy Androgen-independent Hormone-dependent Androgen depletion therapy Castration-resistant Hormone-stimulated Castration Recurrent Androgen-stimulated Combined androgen Hormone-resistant blockade Total androgen blockade Androgen-deprivation therapy current Hormone withdrawal Lethal phenotype Androgen withdrawal Androgen depletion-independent Androgen deprivation Castration-recurrent therapy

spectrometry techniques. More effective communication between translational scientists, clinical scientists, clinicians, and patients will be facilitated by proper description of androgen status, and response to treatment, of prostate cancer. Thus, the editors recommend a uniform nomenclature (see Table 1) that hopefully allows the reader to more easily grasp the excellent reviews provided by the contributors.

References Carter, H. B., Isaacs, J. T.: Experimental and theoretical basis for hormonal treatment of prostatic cancer. Semin Urol, 6: 262, 1988 Geller, J., Albert, J., Loza, D.: Steroid levels in cancer of the prostate-markers of tumour differentiation and adequacy of anti-androgen therapy. J Steroid Biochem, 11: 631, 1979 Gregory, C. W., Johnson, R. T., Jr., Mohler, J. L. et al.: Androgen receptor stabilization in recurrent prostate cancer is associated with hypersensitivity to low androgen. Cancer Res, 61: 2892, 2001 Grossman, H. B.: Hormonal therapy of prostatic carcinoma: is there a rationale for delayed treatment? Urology, 27: 199, 1986 Huggins, C., Hodges, C. V.: Studies on prostatic cancer: I. The effect of castration, of estrogen and of androgen injection on serum phosphatases in metastatic carcinoma of the prostate. 1941. J Urol, 168: 9, 2002 Isaacs, J. T., Coffey, D. S.: Adaptation versus selection as the mechanism responsible for the relapse of prostatic cancer to androgen ablation therapy as studied in the Dunning R-3327-H adenocarcinoma. Cancer Res, 41: 5070, 1981 Jin, R. J., Wang, Y., Masumori, N. et al.: NE-10 neuroendocrine cancer promotes the LNCaP xenograft growth in castrated mice. Cancer Res, 64: 5489, 2004 Labrie, F., Dupont, A., Belanger, A. et al.: New hormonal therapy in prostatic carcinoma: combined treatment with an LHRH agonist and an antiandrogen. Clin Invest Med, 5: 267, 1982 Lee, S. O., Dutt, S. S., Nadiminty, N. et al.: Development of an androgen-deprivation induced and androgen suppressed human prostate cancer cell line. Prostate, 67: 1293, 2007 Mohler, J. L., Gregory, C. W., Ford, O. H., III et al.: The androgen axis in recurrent prostate cancer. Clin Cancer Res, 10: 440, 2004

Introduction: Proper Nomenclature Facilitates Clinical and Translational Research

5

Page, S. T., Lin, D. W., Mostaghel, E. A. et al.: Persistent Intraprostatic Androgen Concentrations after Medical Castration in Healthy Men. J Clin Endocrinol Metab, 2006 Pollack, A., Block, N. L., Stover, B. J. et al.: In vivo selection of androgen-insensitive cells in R3327-G rat prostate tumors: diethylstilbestrol diphosphate treatment versus orchiectomy. J Natl Cancer Inst, 70: 907, 1983 Prostate Cancer Trialists’ Collaborative Group: Maximum androgen blockade in advanced prostate cancer: an overview of 22 randomised trials with 3283 deaths in 5710 patients. Lancet, 346: 265, 1995 Scher, H. I., Sawyers, C. L.: Biology of progressive, castration-resistant prostate cancer: directed therapies targeting the androgen-receptor signaling axis. J Clin Oncol, 23: 8253, 2005 Titus, M. A., Schell, M. J., Lih, F. B., et al.: Testosterone and dihydrotestosterone tissue levels in recurrent prostate cancer. Clin Cancer Res, 11: 4653, 2005 Van Steenbrugge, G. J.: The Nude Mouse in Oncology Research. In: Hormones. Edited by E. Boven and B. Winograd. London: CRC, p. 224, 1991

Androgen Action and Modulation of Prostate and Prostate Cancer Growth: An Historical Perspective Shutsung Liao, John M. Kokontis, Chih-Pin Chuu, and Richard A. Hiipakka

Abstract Early models of steroid hormone action emphasized potential effects of these hormones on metabolic pathways to explain their effects on cells. However these models were abandoned soon after the discovery of the pathway for information transfer from DNA to protein through RNA. Testosterone (T), the major circulating androgen in blood, increases mRNA levels in target tissues and is metabolized to 5a-dihydrotestosterone (DHT) by 5a-reducatse in many target organs. DHT is selectively retained as a protein complex, the androgen receptor (AR), in nuclei, the site of RNA synthesis. Metabolism of T to DHT is critical for androgen action in certain tissues based on the phenotype of individuals with mutations in the gene for 5a-reductase. The cloning of the cDNA for AR has revealed its primary structure and its similarity to other steroid receptors, all members of a superfamily of transcription factors controlled by small lipophilic molecules. Various mutations in the gene for AR are responsible for androgeninsensitivity in men and a potential cause of prostate cancer progression. Certain natural products, like the polyunsaturated fatty acid, g-linoleic acid and the green tea catechin, epigallocatechin gallate (EGCG) are inhibitors of 5a-reductase and may be useful for treatment of disorders dependent on DHT. EGCG also affects appetite and may have a role in the treatment of obesity. Clinical prostate cancer progression can be mimicked in vitro using the LNCaP human prostate cancer cell line. These cells become hypersensitive to androgens, elevate expression of AR and are repressed by physiological doses of androgens after long-term androgen deprivation or treatment with the antiandrogen, Casodex. Although androgen receptor signaling is important for prostate cancer growth and progression, and a target of current therapies, other nuclear receptor signaling pathways may have utility in the treatment of prostate cancer. Activation of liver X

S. Liao(*) The Ben May Department for Cancer Research, The University of Chicago, The Gordon Center for Integrative Science, 929 East 57th Street, Chicago, IL, 60637, USA, E-mail: sliao@huggins. bsd.uchicago.edu

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4_1, # Springer Science + Business Media, LLC 2009

9

10

S. Liao et al.

receptor signaling modulates the growth and progression in human prostate tumor xenografts.

1 Introduction This chapter provides an historical account of many studies regarding the mechanism of action of androgens in the prostate and efforts toward uncovering the molecular and cellular mechanisms controlling prostate cancer growth and function. Some of these studies have been described in more detail in earlier reviews (Liao and Fang 1969; Liao 1975, 1994; Liao et al. 1989, 2001a; Hiipakka and Liao 1998; Kokontis and Liao 1999; Chuu et al. 2007). A time line for many of the discoveries cited in this account is presented in Fig. 1. The studies described in this chapter were performed over a period of about 50 years and were carried out at the Ben May Laboratory for Cancer Research. Dr. Charles B. Huggins established this laboratory at the University of Chicago in 1950 with generous support from Ben May, a businessman and philanthropist from Mobile, Alabama. The Ben May Laboratory became the Ben May Institute for Cancer Research in 1985 and the Ben May Department for Cancer Research in 2007. In September 1957, I (S.L.) was a graduate student in the Biology Department at the Illinois Institute of Technology and had accepted a fellowship from Cornell University at Ithaca, NY, and planned to transfer to the laboratory of Robert Holley (Nobel Laureate 1968, shared with M.W. Nirenberg and H.G. Khorana) to work on a research project dealing with nucleic acids. About a week before my scheduled departure for Ithaca, I visited the University of Chicago campus in the Hyde Park neighborhood on the south side of Chicago. After admiring the very impressive gothic university buildings, I visited, without a prior appointment, Earl Evans, then chairman of the Department of Biochemistry. Dr. Evans kindly showed me around his laboratory and discussed a few research projects in his laboratory. Later that afternoon, Dr. Evans called and suggested that I visit the next morning with Paul Talalay, an associate professor in the Department of Biochemistry, who was working in the Ben May Laboratory. Drs. Evans and Talalay had had a short meeting prior to my meeting with Dr. Talalay, and they made all the necessary arrangements, in one afternoon, for me to become a graduate student at the University of Chicago, even though I never requested or applied for admission. I met Dr. Huggins on my first day in the Ben May Laboratory. His first question was: ‘‘What did you discover today?’’ He then remarked: ‘‘You must make a discovery before you go home.’’ Huggins was a founding member of the University of Chicago Medical School, a surgeon specializing in urology, and a medical researcher who pioneered hormonal therapy for prostate and breast cancer. He received the Nobel Prize in Medicine in 1966 for his work on prostate cancer. He was my mentor until passing away in 1997 (Liao 1997, 2002; Talalay 1997). A photograph of Dr. Huggins and myself (S.L.) is presented in Fig. 2.

Androgen Action and Modulation of Prostate and Prostate Cancer Growth 1957 1961 1962 1965 1966 1968 1969 1969 1969 1971 1973 1976 1978 1979 1980 1982 1982 1985 1987 1987 1988 1988 1089 1990 1991 1991 1992 1993 1994 1994 1995 1995 1995 1996 1996 1998 2000 2000 2001 2002 2004 2005 2006 2008

11

Liao begins graduate studies at the University of Chicago Ph. D. thesis: Discovery, purification and characterization of a new FAD enzyme Testosterone, in vivo, increases mRNA level associated with prostate ribosomes Testosterone, in vivo, increase mRNA content in prostate nuclei Evidence for selective gene transcription by androgen Discovery of nuclear retention and protein (AR) binding of 5α-DHT in prostate Discovery of multiple forms of RNA polymerase in prostate and liver. Antiandrogens, in vivo, reduce nuclear retention of 5α-DHT Androgen specificity of AR Steroid and tissue-specific nuclear retention of 5α-DHT-AR complex Steroid structure and androgenic activity Use of anti-steroid antibodies to assay steroid receptors Discovery of androgen-regulated spermine-binding protein (SBP) Purification and characterization of SBP RNA-dependent release of AR from DNA Discovery of prostate α-protein as a cholesterol binding protein AR recycling in the prostate cells is energy-dependent Discovery of autoimmune antibodies to AR in prostate cancer patients SBP cloned and sequenced Glutathion-S-transferase is androgen-repressed in rat ventral prostate 3-Deoxyadenosine inhibits recycling of AR AR cDNA cloned and sequenced SGP-2 identified as an androgen-repressed in rat ventral prostate 5α-reductase antibodies in prostate cancer patients Mutations in AR are responsible for androgen-insensitivity in men Mutation in AR in LNCaP cells Unsaturated fatty acids are 5α-reductase inhibitors and suppress sebum production Production of antibodies to 5α-reductase Increased AR expression after long term androgen deprivation of LNCaP cells Discovery of UR (the 1st LXR receptor, later called LXRβ), cloned and sequenced Chromosomal localization and immunocytochemical localization of UR Green tea EGCG inhibits 5α-reductase and suppresses sebum production Inhibition of prostate cancer and breast cancer xenografts by EGCG Testosterone suppression of androgen-independent LNCaP xenografts Finasteride stimulation of the growth of androgen-independent LNCaP xenografts Androgen repression of LNCaP cells after long term androgen-deprivation Suppression of endocrine systems, food intake and obesity by EGCG Many hydroxy steroids are ligands for LXR Many topical 5α-reductase inhibitors suppress hamster flank organ growth Many polyphenols inhibit 5α-reductase Anti-proliferative effects of LXR agonists on prostate cancer cells Androgen (A) suppression and reversion of A-independent to A-dependent LNCaP Inhibition of LNCaP xenograft growth and progression by LXR agonist Anti-atherosclerotic effects of a novel synthetic LXR agonist

Fig. 1 Time line of important events and discoveries related to the authors’ laboratory

Although joining the Ben May Laboratory was an accident, it led to a wonderful scientific journey into studies of the mechanism of androgen action and the role of androgens in the growth of benign and malignant prostate. This journey was possible because of the research environment and scientific atmosphere in the

12

S. Liao et al.

Fig. 2 The author and Dr. Charles B. Huggins in 1992. Dr. Huggins graduated from Harvard Medical School in 1924 and did his residency in general surgery at University of Michigan. In 1927, he married Margaret Wellman and joined the University of Chicago Medical School where he quickly became the leading urologist in the world. In 1941, Huggins published the classic paper showing that the deprivation of male hormone could lead to regression of metastatic prostate cancer in men. In the 1950s, he showed that removal of female hormones could lead to substantial regression of advanced breast cancer in many women. For these discoveries he was awarded the 1966 Nobel Prize. His discoveries showed that cancer cells were not autonomous and selfperpetuating but are dependent on chemical signals such as hormones to grow, and depriving those signals could restore the health of patients with widespread metastasis. Huggins’s fundamental discoveries led to hormonal therapy and provided an immense stimulus to research in cancer chemotherapy. In 1950, with a generous contribution from Mr. Ben May, Huggins established the Ben May Laboratory for Cancer Research and declared ‘‘Discovery is our business’’ its motto. A proponent of ‘‘small is better,’’ he invited a few young scientists to join him with unambiguous advice: ‘‘Do not go to meetings or the library, do not write books or become a head of anything, they are a waste of your time.’’ ‘‘Discover first’’ he insisted, read afterward to find out whether you can connect your new discovery to established ideas. This creative scientist made many fundamental discoveries (see text). Huggins worked 7 days a week for science and medicine until he was 92. He passed away on January 12, 1997, at his home near the University of Chicago, when he was 95 years old

Ben May Laboratory and at the University of Chicago, which played critical roles in forming the philosophy and scientific approach in my laboratory. During the 1950s, Huggins had already mentored several young scientists, who made many fundamental discoveries, such as identification of mitochondria as the sites of cellular energy production (Albert Lehninger) and elucidation of the biochemical steps involved in phospholipid synthesis (Eugene Kennedy). One of Huggins’ earliest students was William W. Scott, who built the distinguished urology department at the Johns Hopkins University Medical School. In the 1950s, Dr. Talalay, who had been a medical student in the laboratory of Dr. Huggins in the 1940s, was a pioneer in the study of steroid metabolizing

Androgen Action and Modulation of Prostate and Prostate Cancer Growth

13

enzymes. He fed testosterone to Pseudomonas and other bacteria and isolated enzymes that could metabolize testosterone and other steroid hormones. Dr. H. Guy Williams-Ashman, a British biochemist with considerable experience in reproductive biology, joined the Ben May Laboratory during the early 1950s. He was an assistant professor in the Department of Biochemistry in 1957 when I joined Talalay’s laboratory. At that time Talalay and Williams-Ashman were evaluating the possibility that the biological effects of certain steroid hormones, such as estradiol may be related to coupled oxidation and reduction of the hormone. They proposed that a steroid hormone might mediate a pyridine nucleotide-dependent transhydrogenation without consumption of the hormone itself (Talalay and Williams-Ashman 1960). For estrogen action they proposed and showed that the following two reactions could occur in test tubes: (1) Estradiol + NADP+Ô Estrone + NADPH + H+ and (2) Estrone + NADH + H+Ô Estradiol + NAD+. In combination, these two reactions give the following net reaction: NADP+ + NADH Ô NADPH + NAD+. One important implication of this proposed mechanism was that hydrogen produced in the form of NADH by mitochondrial oxidation would be transformed into NADPH that could participate in various synthetic reactions in a cell. However, the steroid hormone-dependent pyridine nucleotide transhydrogenation theory of Talalay and Williams-Ashman was soon found not to be the mechanism involved in steroid hormone action, because (1) diethylstilbestrol, a potent nonsteroidal synthetic estrogen, could not carry out the transhydrogenation, and (2) 17b-estradiol appeared to function without cycling through reduction and oxidation in estrogen-sensitive organs. Gordon Tomkins had proposed another model for steroid hormone action. He showed that a steroid hormone could act as an allosteric effector and modify the function of the active site of an enzyme (Tomkins et al. 1965). Diethylstilbestrol and certain other steroid hormones were shown to alter the quaternary structure of pure glutamic dehydrogenase and caused enzyme heteromers to disaggregate into monomers. This change decreased glutamic dehydrogenase activity and increased alanine dehydrogenase activity of the same enzyme. However, this effect required high concentrations of enzyme and hormone, and so was not likely to occur in vivo. In the Ben May Laboratory, Elwood Jensen, an organic chemist (who was a founding member of the Ben May Laboratory with Huggins, Lehninger, and Talalay) and his associate, Herbert Jacobson, initiated an entirely different approach toward understanding estrogen action in target organs. Their idea was to trace the fate of estrogen in estrogen-sensitive organs. They planned to do this by synthesizing highly radioactive [3H]17b-estradiol and administering this compound to rats. The idea of using tracer technology was not surprising given the University of Chicago’s recent involvement in radioisotope methodologies after the first successful demonstration of controlled nuclear fission at the university. Jensen and colleagues showed clearly that estrogen-sensitive organs, such as uterus and ovary, but not estrogen-insensitive organs, had a specific mechanism to retain estrogen against a concentration gradient between blood and target organ (Jensen and Jacobson 1962; Jensen 1978). This discovery led to the novel idea that estrogen-sensitive cells had receptor(s) that were responsible for retention of

14

S. Liao et al.

estrogens and presumably for some aspects of the biological effects of the hormone. The biological significance of this discovery was supported by the demonstration that antiestrogens hindered estrogen retention. Jensen’s pioneering studies led other investigators to discover and study other steroid hormone receptors and ultimately uncover the superfamily of nuclear receptors. Another very important discovery that was made in the Ben May Laboratory was by Samuel Weiss. He discovered that liver cell nuclei contain an enzyme that uses DNA as template to make RNA. This was the first demonstration of a DNAdependent RNA polymerase (transcriptase). Although Weiss’ very important original discovery with a mammalian system was ahead of other reports for bacterial systems, his contribution is rarely acknowledged. Weiss showed that DNase can abolish RNA synthesis and also demonstrated by nearest neighbor frequency analysis that the RNA product synthesized reflected the DNA template sequence (Weiss and Nakamoto 1961a, b). Good scientists making important discoveries provided a fertile research environment in the Ben May Laboratory. I began my studies and research career as a graduate student under Dr. Williams-Ashman’s guidance and after my graduation in 1961, Dr. Huggins persuaded me to stay on in the Ben May Laboratory as an independent researcher.

2 Androgen Regulation of the Synthesis of RNA and Protein 2.1

Hormone–Gene Theory

With the discovery of various enzymes and hormones in the 1930s and 1940s, biochemical studies of the mechanism of action of steroid hormones in the next few decades concentrated on induction of specific enzymes and proteins by different hormones. This approach became very popular after it was understood how genetic information in DNA is transmitted through mRNA to direct specific protein synthesis. Thus, the term, ‘‘Hormone–Gene Theory’’ evolved. One important discovery in this regard was the demonstration that injection of the insect hormone ecdysone into Chironomus tentans (midge) larvae caused a puff to form in specific regions of the chromosomes of the salivary gland (Clever and Karlson 1960). It was suggested that ecdysone induced puff formation by activation of a gene locus to produce RNA. Many subsequent studies of hormonal responses at the cellular and organ level also showed that steroid hormones induce RNA synthesis and subsequently specific protein synthesis.

2.2

Androgen and mRNA Synthesis

To investigate the mechanism by which androgens control RNA and protein synthesis, the rat ventral prostate was used as the target organ for our studies. It was observed that the androgen-dependent increase in protein-synthesizing activity

Androgen Action and Modulation of Prostate and Prostate Cancer Growth

15

of isolated ribosomal particles from the rat ventral prostate was due to an increase in the level of mRNA associated with ribosomes (Liao and Williams-Ashman 1962). Castration led to an increase in the proportion of ribosomes that were not occupied by native mRNA and, therefore, could utilize externally added synthetic polyribonucleotides as mRNA templates to guide radioactive amino acid incorporation into polypeptide linkages. Thus, with prostate ribosomes isolated from castrated animals, poly U stimulated phenylalanine incorporation, whereas poly UG stimulated radioactive valine incorporation into polypeptide linkages. The extent of this synthetic polymer-dependent incorporation was reduced by androgen injection into castrated animals, which suggested that androgen increased the proportion of ribosomes that were occupied by endogenous mRNA. These experiments suggested that androgen increases the amount of mRNA associated with ribosomes and provided an explanation for androgen stimulation of protein synthesis. These experiments were performed shortly after the discovery of the genetic code by Nirenberg and Matthaei (Nirenberg and Matthaei 1961; Nirenberg et al. 1962) and Ochoa and colleagues (Lengyel et al. 1962; Crick 1963), who used synthetic ribopolymers, synthesized by polynucleotide phosphorylase, to determine that bacterial protein synthesis uses a three-base codon to specify amino acid incorporation into proteins. Nirenberg actually visited our laboratory and discussed with us the use of these synthetic polymers. Our findings with a prostate ribosomal system were consistent with his important discoveries and supported the universality of the amino acid codons proposed at that time. In order to show that androgens increase the mRNA level in the nuclei of target cells, nuclear RNA was isolated from the prostate of rats, and a bacterial ribosomal system was used to determine whether this nuclear RNA could act as an mRNA template to stimulate the incorporation of radioactive amino acids into peptide linkages. This method allowed demonstration that androgen in vivo could increase the mRNA template activity of nuclear RNA isolated from the ventral prostate of castrated rats (Liao 1965). The RNA-synthesizing activity of isolated prostate cell nuclei was enhanced within 1 h after androgen injection into castrated rats (Liao et al. 1965).

2.3

Discovery of Multiple Forms of RNA Polymerases

Since very little was known about RNA polymerase at that time, we decided to study the behavior of RNA polymerase in liver and prostate. Cell nuclei from these organs, prepared with buffers containing hypertonic sucrose to prevent leakage of nuclear enzymes, retained a large quantity of RNA polymerase that is not tightly associated with chromatin (Liao et al. 1968). Two different forms of RNA polymerase were identified (Liao et al. 1969). These different forms of RNA polymerase synthesized different types of RNA in vitro. The existence of multiple forms of liver RNA polymerase (RNA polymerases I, II, and III) was later shown by other investigators to be responsible for synthesizing different classes of RNA in mammalian cells.

16

2.4

S. Liao et al.

Androgen and Modulation of Specific Proteins in Rat Ventral Prostate

In the process of regulating the growth and function of target organs, androgens stimulate the expression of certain genes, but also simultaneously repress the expression of other genes. In the rat ventral prostate two new proteins, a-protein and spermine-binding protein (SBP) were discovered that were positively regulated by androgen. Two other proteins, glutathione S-transferase (GST) and sulfated glycoprotein 2 (SGP-2, also known as clusterin and apo J), were identified that were negatively regulated by androgen. a-Protein was originally discovered in our laboratory in 1971 (Fang and Liao 1971; Chen et al. 1982) but was rediscovered later by other investigators and called prostate-binding protein or prostatein. a-Protein is a cholesterol-binding protein composed of four polypeptide subunits that is a major (>40%) component of the secretory fluid of the rat ventral prostate. One of the smallest subunits that was purified and sequenced inhibited the binding of androgen receptor (AR) to prostate nuclear fractions (Shyr and Liao 1978), but its biological function remains unclear. SBP was discovered in 1977 as a specific polyamine-binding protein (Liang, et al. 1978; Mezzetti et al. 1979; Hiipakka et al. 1984). cDNA for SBP was isolated and the protein sequenced by employing fast atom bombardment mass spectrometry and tandem mass spectrometry because the amino terminus of SBP is blocked (Anderegg et al. 1988). The function of SBP remains unclear, although prostates are rich in polyamines. One of at least four androgen-repressed proteins identified in the ventral prostate is the Yb1 (GST M1-1) subunit of GST that colocalizes with uridylic acid-rich small nuclear RNAs at interchromatinic regions in the cell nucleus (Saltzman et al. 1987; Chang et al. 1987). SGP-2 is another androgen-repressed protein that was identified by cloning and sequencing (Bettuzzi et al. 1989). SGP-2 mRNA increases 17-fold 4 days after castration and decreases rapidly within 6 h after androgen administration.

3 Importance of the Conversion of Testosterone to 5a-Dihydrotestosterone (DHT) 3.1

Selective Retention of DHT

Since androgens rapidly affect nuclear RNA synthesis in the rat ventral prostate, it was reasonable to assume that androgens act in the cell nuclei of the prostate. Based on studies by Jensen, 17b-estradiol is retained by cell nuclei in estrogen target organs, such as uterus without metabolic conversion. However, testosterone produced in testis and circulating in blood is metabolized to many metabolites in liver, prostate, and other organs. Therefore, three simple questions were asked:

Androgen Action and Modulation of Prostate and Prostate Cancer Growth

17

(1) Which metabolite(s) of testosterone are retained by prostate cell nuclei? (2) Is nuclear retention specific for androgen-sensitive cells? and (3) Is nuclear retention of androgens inhibited by antiandrogens? Studies were carried out first in vivo by injecting [3H]testosterone intraperitoneally into castrated rats. Later studies relied on in vitro incubation of isolated rat ventral prostate tissue and other organs with [3H]testosterone. Whether using whole animals or isolated organs, the ventral prostate, but not other organs, such as liver, brain, or thymus that are normally not considered androgen target organs, retained [3H]5a-DHT in nuclei. Retention of 5a-DHT in nuclei lasted for about 12 h. The cytoplasmic soluble (cytosol) fraction of prostate tissue had more than six radioactive metabolites, including a large quantity of steroids more polar than testosterone. When [3H]4-androsten-3,17dione was used instead of [3H]testosterone, the radioactive steroid retained by isolated cell nuclei was exclusively [3H]5a-DHT. A steroid chemist, Josef Fried, verified our identification of the steroids retained by prostate nuclei. Prior to joining the Ben May Laboratory, he had synthesized many fluorinated steroidal drugs in industry. The radioactive steroid retained by the prostate cell nuclei could not be removed easily by incubating the nuclei at 0–4 C with high concentrations of nonradioactive testosterone or 5a-DHT, or washing nuclei with isotonic sucrose containing 1% deoxycholate or 0.4% Triton X-100, which remove cytoplasmic membrane contaminants and nuclear membranes. Proteases, but not RNases, effectively released the [3H]5a-DHT from nuclei. However, [3H]5a-DHT retained in nuclei could be extracted with buffers containing 0.4 M KCl as a protein-bound complex that could be characterized by gel filtration or by sedimentation analysis using sucrose gradient centrifugation. These findings strongly suggested that 5a-DHT was retained in nuclei as a specific 5a-DHT–protein complex that was tightly bound to chromatin. The original findings were submitted to several journals in the United States for publication, but for more than 6 months our manuscript was not accepted for publication. Most of the criticism centered on a perceived lack of evidence to support the biological significance of our observations. For example, one referee’s comment was: ‘‘Authors did not succeed in showing that the phenomena observed were related to growth and function of the ventral prostate.’’ Therefore, the manuscript was submitted to Nature, which accepted the manuscript but with a request to remove the word ‘‘receptor’’ (Anderson and Liao 1968). The use of the term ‘‘androgen receptor’’ was taboo at this time. In the proof, the use of ‘‘androgen receptor’’ was reduced. This research work also was submitted as K.M. Anderson’s Ph.D. thesis in the Department of Biochemistry at the University of Chicago. A meeting abstract for a clinical conference, submitted by N. Bruchovsky and J. Wilson, came to our attention after submission of our manuscript to Nature. The abstract was cited when we received our galley proofs of the manuscript. In this abstract, these investigators reported that although [3H]5a-DHT, androstane3,17-diol and testosterone are found in prostate cytoplasm within 1 min after administration of [3H]testosterone, only [3H]5a-DHT and testosterone are found in prostate nuclei after 2 h (Bruchovsky and Wilson 1968).

18

S. Liao et al.

In 1967, Dr. Williams-Ashman, who had moved to Johns Hopkins University Medical School in 1963, was informed of our findings. A set of slides was prepared for him and he presented our findings at a ‘‘Symposium on Hormone Action’’ in Europe in early 1968. Several participants of the meeting communicated with us before our study was published in Nature. Our findings elated two investigators in the Ben May Laboratory. One was Charles Huggins, who had already observed that 5a-DHT is more active than testosterone in stimulating the growth of prostate-like organs in the urinary ducts in female rats. The other was Elwood Jensen. His hypothesis that a specific receptor plays a role in estrogen action was strengthened by another example with a different steroid receptor. A distinct difference between the two sex steroids is that 17b-estradiol interacts with its receptor without metabolism, whereas testosterone needs, at least in some target organs, to be converted to 5a-DHT. Nuclear retention of the 5a-DHT–protein complex was reduced considerably in vivo by coadministration of cyproterone or its acetate, both known antiandrogens that suppress male hormone action, which provided additional evidence that retention of 5a-DHT by prostate nuclei is biologically important (Fang and Liao 1969). Both antiandrogens also suppressed retention of 5a-DHT when they were incubated with minced ventral prostate tissue. A nonsteroidal antiandrogen, flutamide, suppressed nuclear retention of 5a-DHT in vivo, but not when prostate tissue was incubated with androgen and antiandrogen in vitro (Liao et al. 1974). The lack of effect in vitro suggested that the active form of flutamide might be a metabolite. Neri et al. (1979) subsequently described the potent antiandrogenic activity of hydroxyflutamide, a metabolite of flutamide. These studies strongly supported our contention that 5a-DHT was bound to a specific protein, the androgen receptor. In collaboration with the Stumpf laboratory at the University of Chicago, nuclear retention of 5a-DHT in rat seminal vesicles and prostates was demonstrated by dry mount autoradiography (Sar et al. 1970).

3.2

Different Biological Functions of Testosterone and 5a-DHT

Clear biological evidence that the conversion of testosterone to 5a-DHT is important for androgen action was provided several years after our observation that 5aDHT is retained selectively by target cell nuclei. J. Imperato-McGinley and her associates discovered that a 5a-reductase deficiency is the cause of a form of male pseudohermaphroditism, where testosterone levels are adequate for development of the Wolffian ducts but are inadequate for complete virilization of the urogenital sinus from which the prostate and external genitalia develop (Imperato-McGinley et al. 1974; Siiteri and Wilson 1974; Andersson et al. 1991; Imperato-McGinley and Gautier 1992). Like testosterone, certain other natural androgens, such as 4-androstenedione or androstanediols, can be converted to 5a-DHT in androgen target organs and produce androgenic effects. 5a-DHT produced in these organs may be a source of

Androgen Action and Modulation of Prostate and Prostate Cancer Growth

19

circulating 5a-DHT. 5a-DHT in blood, however, is bound to blood proteins and may not be as effective as intracellularly formed 5a-DHT. Some androgen-sensitive tissues, such as muscle, contain little or no 5a-reductase. In these tissues testosterone may function as the active androgen. Specific aspects of embryonic male urogenital tract differentiation and certain testicular and brain functions, which include events related to sexual behavior, may also utilize testosterone or its estrogenic metabolite, 17b-estradiol. Testosterone negatively regulates luteinizing hormone (LH) release from the pituitary in males. Feedback inhibition involves both testosterone and estrogen signaling in the pituitary and hypothalamus (Pitteloud et al. 2008). Negative feedback of LH secretion by testosterone does not appear to require 5a-reduction in humans (Iranmanesh and Veldhuis 2005). In rats there is evidence for (Liang et al. 1984) and against (Nagamoto et al. 1994) a role for 5a-DHT in feedback regulation of LH secretion. In rams there is evidence that 5a-DHT may be more effective than testosterone in feedback control (Schanbacher et al. 1987). Testosterone can also be converted by 5b-reductase to 5b-DHT. 5b-DHT is not androgenic and does not bind to AR. However, liver 5b-reductase is required for the synthesis of bile acids, which are secreted in bile and into the small intestine to facilitate lipid absorption. Some 5b-androstanes are active as hematopoietic agents. They increase the production of heme, hemoglobin synthesis, and erythropoiesis, perhaps by interacting with the nuclear receptor, CAR, which regulates synthesis of the mRNA for delta-aminolevulinic acid synthase, the rate-limiting enzyme for heme synthesis (Yamamoto et al. 2003). Testosterone and 5a-DHT also enhance erythropoiesis by stimulating the synthesis of erythropoietin by the kidney. There are two 5a-reductase isozymes that exhibit strikingly different biochemical behavior and expression patterns (Andersson et al. 1989, 1991; Thigpen et al. 1993; Russell and Wilson 1994). The type 2 reductase has been proposed to play an anabolic role in target organ development, whereas type 1 reductase may function more in catabolism of testosterone (Normington and Russell 1992). Many mutations in the gene for the type 2 reductase have been described, which cause a form of male pseudohermaphroditism. Mutations in the gene for the type 1 reductase are not normally related to androgen insensitivity (reviewed in Hiipakka and Liao 1998; Kokontis and Liao 1999) but do cause defects in murine parturition and increase murine embryonic lethality, which in the latter case appears to be due to high intrauterine estrogen levels (Mahendroo et al. 1997). Testosterone is the precursor for both 17b-estradiol and 5a-dihydrotestosterone, and competition for substrate may limit excessive production of estrogen in certain tissues.

3.3

Medical Uses for 5a-Reductase Inhibitors

If the biological action of testosterone in certain tissues is dependent on its conversion to 5a-DHT and binding of 5a-DHT to a specific receptor, certain medical conditions such as benign prostatic hyperplasia (BPH); baldness; acne;

20

S. Liao et al.

female hirsutism; and even androgen-dependent cancers of the prostate, breast, and liver may be controlled by a 5a-reductase inhibitor or by an antiandrogen that prevents androgen receptor binding of 5a-DHT. In 1967, before we submitted our paper on selective retention of 5a-DHT by prostate nuclei for publication, we requested that the University of Chicago file a patent for medical uses of 5areductase inhibitors. However, the university, at that time, was not interested in commercialization of new research findings and rejected our request. However, Merck hired Tehming Liang, who was a Ph.D. student and then a research associate in our laboratory, in 1974. At Merck, Liang was part of a group that identified a synthetic compound, finasteride, which is an effective 5a-reductase inhibitor (Fig. 3). Finasteride became the drug Proscar for the control of BPH. Years later the same compound was also used as a drug (Propecia) to treat male pattern baldness. Finasteride is a 4-azasteroid and is a better inhibitor of the type 2 than type 1 reductase. Dutasteride (Avodart), another drug used for treatment of BPH, is also a 4-azasteroid like finasteride, but it is an effective inhibitor of both isozymes (Bramson et al. 1997). Because diet is often an important component in disease susceptibility and since diet may modulate androgen signaling in unknown ways, we thought it was important to analyze dietary components for effects on 5a-reductase. Our laboratory looked for natural compounds that could function as 5a-reductase inhibitors. Several unsaturated fatty acids were found to be excellent 5a-reductase inhibitors. One of the more potent compounds we identified is g-linolenic acid (Fig. 3). Saturated fatty acids like stearic acid are much less active. To determine if these

Fig. 3 Structures of testosterone, 5a-dihydrotestosterone, and various 5a-reductase inhibitors

Androgen Action and Modulation of Prostate and Prostate Cancer Growth

21

compounds could be used for treatment of a medical condition modulated by androgens, such as acne, their effects were investigated using the hamster flank organ, a modified sebaceous gland dependent on androgen for growth and pigmentation. Male hamsters were castrated to reduce the size of flank organs and testosterone in ethanol with or without g-linolenic acid was applied topically. Flank organ growth was stimulated by topically applied testosterone, but we found that this stimulation was suppressed by 5a-reductase inhibitors, such as g-linolenic acid and other unsaturated fatty acid, but not by saturated analogs, such as stearic acid (Liang and Liao 1997). The catechins, epigallocatechin-3-gallate (EGCG) and epicatechin3-gallate (ECG), isolated from green tea were also potent 5a-reductase inhibitors and suppressed flank organ growth and pigmentation, whereas other green tea catechins were less active or not active (Liao et al. 2001b). Both g-linolenic acid and EGCG suppressed sebum production from human male forehead skin, which suggested that they could be utilized to control sebum production, which in excess contributes to acne. Various acne medicines are available and they work by opening blocked follicles (retinoids) or reduction of skin microflora (antibiotics). Since androgens stimulate sebum production, 5a-reductase inhibitors, such as g-linolenic acid and EGCG, may be useful as a new type of acne medicine (Fig. 3). A variety of polyphenolic compounds are also inhibitors of 5a-reductase (Hiipakka et al. 2002). Some of these compounds are part of a normal diet. It remains to be determined whether any of these compounds are useful for treating androgen-dependent disorders or whether they have any adverse affects on androgen signaling in vivo.

4 Modulation of Prostate Tumor Growth, Endocrine Systems, and Food Intake by EGCG Green tea catechins have cancer chemopreventative activity in many laboratory models of cancer initiation and progression (Liao et al. 2001a). We have shown that EGCG, but not other green tea catechins, repress prostate and breast tumor xenograft growth in nude mice (Liao et al. 1995). Green tea catechins also inhibit the development of prostate and breast cancer in TRAMP and C3(1) SV40 transgenic mice that express oncogenic SV-40 large T antigen under the control of androgenregulated promoters (Gupta et al. 2001; Leong et al. 2008). Green tea catechins may have similar properties in humans, since tea catechins have been shown to prevent progression of prostate intraepithelial neoplasia to more advance cancer when given orally to men (Bettuzzi et al. 2006). In contrast to these laboratory studies, some epidemiological studies have been equivocal in their support of cancer chemopreventative effects of green tea (Kuriyama et al. 2006). Since green tea catechins have been advocated for health promotion, we studied the acute physiological effects of green tea catechins in rats. EGCG but not other green tea catechins reduced food intake, body weight, blood levels of testosterone,

22

S. Liao et al.

estradiol, leptin, insulin, IGF-I, glucose, cholesterol, and triglycerides. The growth of the prostate, uterus, and ovary was also reduced. Similar effects were observed in normal and leptin receptor-defective Zucker rats. Therefore, EGCG does not cause these affects by modulating leptin signaling. Decreased food intake is known to modulate many endocrine systems (Schreihofer et al. 1993). How EGCG regulates appetite is not clear, but EGCG is an inhibitor of fatty acid synthase (FAS; Tian 2006). Inhibition of FAS elevates the level of malonyl CoA in the hypothalamus and this metabolic intermediate can regulate signaling mechanism controlling appetite perhaps by regulating the activity of carnitine palmitoyl-transferase-1 (Wolf 2006).

5 Identification and Biochemical Characterization of Androgen–AR Complexes 5.1

Radioactive Androgens for Identification and Characterization of AR

Highly radioactive 5a-DHT was not commercially available in the late 1960s as a tool to identify and characterize AR in various target tissues. Therefore, [3H] testosterone was incubated with a prostate cytoplasmic membrane fraction or prostate cell nuclear fraction to obtain [3H]5a-DHT for our early studies. Fortunately, [3H]5a-DHT was soon commercially available. Another androgen identified as a potent ligand for AR (Liao et al. 1973a) that has been widely used in characterization of the AR is 17a-methyl-17b-hydroxy-estra-4,9,11-trien-3-one, now commonly known as R1881 or methyltrienolone. R1881 is stable to metabolic conversion and binds to AR with high affinity. Another synthetic androgen, 7a, 17a-dimethyl-19-nortestosterone, DMNT or mibolerone, also proved useful for study of AR (Liao et al. 1973a; Schilling and Liao 1984). Besides its metabolic stability and low affinity for TeBG, DMNT is more receptor selective than R1881.

5.2

Sedimentation Analysis of AR by Gradient Centrifugation

Introduction of gradient centrifugation for the study of estrogen receptor (Toft and Gorski 1966) stimulated the use of this technology in the characterization of various other steroid hormone receptors. When the cytosolic fraction of rat ventral prostate was incubated with [3H]5a-DHT, the radioactive complexes formed migrated with sedimentation coefficients (Svedberg value) of 3–9 S. (Fang et al. 1969; Liao and Fang 1969; Fang and Liao 1971). The larger complexes dissociated into smaller 3–4 S units in the presence of high salt, which suggested that AR was present as a dimer or associated with other cytosolic components. Adventitious interactions of

Androgen Action and Modulation of Prostate and Prostate Cancer Growth

23

AR with other cellular components in tissue extracts can occur. In addition protease action may alter the native receptor structure and alter biochemical characteristics of the receptor. [3H]5a-DHT complexes extracted by buffers containing 0.4 M KCl from cell nuclei of the ventral prostate of rats injected with [3H]testosterone or from prostate tissue incubated with [3H]testosterone or [3H]5a-DHT migrated as 3–4 S entities. Similar complexes were formed when other radioactive androgens (R1881 or DMNT) were used instead of [3H]5a-DHT.

5.3

Assay of AR Using Insolubilized Antisteroid Antibodies

The gradient centrifugation technique is not easily used for quantitative determination of steroid receptors because it is a laborious procedure and because receptors specifically and nonspecifically interact with other cellular components, which makes accurate quantification of steroid-bound receptors difficult because they are dispersed in many fractions of the gradient. Gel filtration of the steroid-receptor complexes and methods using charcoal for removal of free steroids also have limitations when used in quantification of steroid receptors. Anti-5a-DHT antibodies coupled to dextran-agarose beads were used to separate AR complexes from free steroid. The mixture is simply centrifuged and receptor-bound radioactivity in the supernatant measured. This method (Castaneda and Liao 1974) works well because the radioactive DHT is bound tightly in the androgen-binding cavity of AR, and the complex has a very low rate of dissociation. The complex is also not recognized by 5a-DHT antibodies. The method is applicable to the quantitative assay of other steroid receptors because commercially available steroid antibodies can bind to their respective steroids, in free form or weakly bound to nonreceptor proteins, but not steroids in the ligand-binding pocket of the receptor (Castaneda and Liao 1975a, b). The method is very simple and can handle many samples.

5.4

Using a Hydroxylapatite Filtration Assay to Measure AR

A very simple method was developed to assay the androgen-receptor complex using hydroxylapatite adsorption and washing the adsorbed receptor free of unbound steroid on filters (Liao et al. 1984). Radioactive androgen–AR complexes were mixed with slurry of hydroxylapatite and the mixture was transferred onto glass filters that were mounted on a multiple sample filter manifold and washed with a cold buffer solution. The washed filters were transferred into vials for measurement of radioactivity. Protease inhibitors were added to samples to minimize proteolysis. The procedure is economical, simple, fast and requires only small amounts of sample. More than 50 samples could be processed in 30 min. Other steroid receptors, such as glucocorticoid or estrogen receptors in different tissue or cell

24

S. Liao et al.

extracts, also could be measured using this method. This method has been employed in many of our studies of AR.

5.5

Structural Requirements for Interaction of Androgens with AR

Various steroid-binding studies have shown that 5a-DHT interacts differently than testosterone with AR. This must be due to differences in steroid structure at the A and B ring junction where these two steroids differ. Dreiding stick and CPK space-filled molecular models were used to compare the structures of these two steroids (Liao et al. 1972, 1973a). These molecular models show that the thickness of the A/B ring area including the C19 methyl angular group at C10 is approxi˚ for testosterone and 3.2 A ˚ for 5a-DHT. At that time, several very active mately 4 A synthetic androgens called ‘‘super androgens’’ were known, and they did not have a C19-methyl group attached to C10. Without an angular C19-methyl group, the structure of the A/B ring area is less bulky. Although these synthetic super androgens have a double bond at C4, as in testosterone, they can interact with the prostate AR even more tightly than 5a-DHT. This observation strongly supported our contention that spatial geometry is an important characteristic of androgens that bind strongly to AR. These studies suggested that the need for 5a-reductase to convert testosterone to 5a-DHT is simply to make the A/B ring area of the molecule less bulky so that it could be accommodated in the androgen-binding cavity of AR. The orientation of methyl groups at other position is clearly very important, since 7b-methyl derivatives are poor competitors, while 7a-methyl derivatives compete well for binding to AR. In addition, cis-testosterone, which has a 17a-OH, does not compete for binding to AR and is not androgenic, an indication of the importance of the configuration of the 17-hydroxyl group for binding activity.

5.6

Androgen and Antiandrogen Binding to the Hydrophobic Cavity of AR

Structural studies on AR binding to androgens and the results from antiandrogen antibody experiments strongly suggest that androgens interact with AR as if the androgen is being totally enveloped in a hydrophobic cavity of AR. This should reduce the rate of dissociation, increase the binding affinity, and increase the biological activity of the androgen. Certain chemicals may resemble in part active androgens and may be capable of interacting with the androgen-binding cavity with sufficient strength to exhibit androgenic or antiandrogenic activity. A transparent plastic box having nine faces was made to illustrate the required geometric

Androgen Action and Modulation of Prostate and Prostate Cancer Growth

25

structure, sufficient to accommodate from all sides a CPK molecular model of 5aDHT, but not testosterone (Liao et al. 1972). Two holes were made to accommodate the 7a- and 17a- methyl groups of methylated androgens. A light and battery was placed at a corner so that the light would go on when the box cover was closed completely indicating a good fit for the ligand. For example, the CPK model of cimetidine (Tagamet), a histamine H2-receptor antagonist prescribed for gastric and duodenal ulcers, fits very well inside the box. In line with this analysis, cimetidine inhibits [3H]R1881 binding to AR, which may explain why cimetidine exhibits antiandrogenic activity in some male patients. The ability of various small molecules to compete with [3H]R1881 for binding to AR in the cytosolic fraction of rat ventral prostate was investigated to explore the possibility that the steroid-binding site of a steroid receptor may have a significant affinity toward small molecules representing portions of the natural ligand. More than 80 mono- and polycyclic aromatic hydrocarbons were tested and certain reduced phenanthrenes are strikingly active at 20 mM, while many anthracenes are inactive at 2 mM. Interestingly, 9,10-dihydrophenanthrene is about 20 times more effective than phenanthrene or phenanthrenes reduced in the two end aromatic rings. The 9- and 10-hydrogenation may alter the conjugated phenanthrene in such a way that the nonconjugated molecule is more bulky, and this decreases the rate of dissociation from the AR ligand-binding pocket (Chang and Liao 1987). These comparisons are biologically meaningful because 9,10-dihydrophenanthrene, but not phenanthrene, injected into rats inhibits the androgen-dependent growth of accessory reproductive organs. 9,10-Dihydrophenanthrene did not compete with [3H]estradiol for binding to estrogen receptor of rat uterus or from human breast tumor MCF-7 cells or glucocorticoid receptor of rat liver (Liao et al. 1983; Chang and Liao 1987). These studies indicate that while the hormonal action of a steroid may be dependent on the interaction of a functional group of the hormone with the receptor, the presence of such a group might not be required for antagonistic activity of a compound that can physically block hormone binding to the receptor. These studies also led to a similar conclusion for estrogen and glucocorticoid receptors. These observations suggest that certain environmental compounds might affect the actions of various steroid hormones through interactions with their receptors (Chang and Liao 1987).

6 Intracellular Transformation and Recycling of AR It is widely accepted that steroid-receptor complexes act by recognizing and interacting with specific promoter regions in target genes through the DNA-binding region of the steroid receptors. Precisely how this occurs is not yet very clear especially because of the multitude of proteins present at chromatin loci where RNA synthesis and processing are carried out. One subject often discussed is how a steroid-receptor complex is recycled or processed after its involvement in transcription initiation. In 1969, we proposed that the androgen–AR complex may bind to

26

S. Liao et al.

RNA and is involved in the processing and transport of newly synthesized RNA (Liao and Fang 1969). Ribonucleoprotein particles in rat ventral prostate or calf uterus interacted with androgen–AR complex or estrogen–estrogen receptor complexes, respectively (Liao et al. 1973b; Liang and Liao 1974). Certain polyribonucleotides, such as poly UG, are much better that other polyribonucleotides, such as poly C, poly A, or poly U, in removing androgen–AR complex from DNA-cellulose or from prostate chromatin aggregates, suggesting that androgen–AR complex can interact specifically, depending on nucleotide sequences, with ribopolymers (Liao et al. 1980). Therefore, steroid receptors could play an important role in removing specific newly synthesized RNA from the DNA template for post-transcriptional processing and allow access to DNA for further RNA synthesis. Androgen–AR complex appears to be trapped and accumulates in the nuclei and is not recycled back to the cytoplasm, if RNA synthesis is inhibited by actinomycin D. In the presence of 30 -dA, a potent inhibitor of RNA polyadenylation and nucleocytoplasmic RNA transport, androgen–AR complex is maintained in an RNA-bound form and recycling of androgen–AR complex is prevented. The level of chromatin-bound androgen–AR complex is therefore reduced, and the androgen–AR complex accumulates in the cytoplasm compartment (Rossini and Liao 1982; Hiipakka and Liao 1988; Liao et al. 1989). Some of our other studies also suggest that AR is controlled in target cells by an energy-dependent activation process. A significant portion of AR in prostate cells is in a form that cannot interact with 5a-DHT unless it is reactivated by an energy-dependent process. In the presence of respiratory poisons such as 2,4dinitrophenol, AR is rapidly deactivated (half-life, 2 min) to an inactive form that does not bind 5a-DHT unless it is reactivated by an energy-dependent process to the androgen-binding form. A majority of 5a-DHT–AR complexes formed do not bind to DNA/chromatin unless they are ‘‘transformed’’ by a temperature-dependent process. Isotope-chasing experiments with labeled androgen indicate that the steps involved in androgen–AR complex recycling between chromatin-bound and cytosolic forms are slow compared to other steps in the recycling process and that the 5a-DHT–AR complex has a half-life of more than 50 min in this second part of the cycle (Rossini and Liao 1982). This slow process may reflect the participation of 5a-DHT–AR complex in a time-consuming mechanism that is essential for the hormonal response. Work from many other laboratories has shown subsequently that heat-shock protein/chaperone interactions with nuclear receptors during receptor folding and maturation are dependent upon ATP hydrolysis (see review by Pratt and Toft 1997), and may underlie some of our observations of energy-dependent AR activation.

7 Cloning of AR cDNA Our laboratory has always been small and initially did not attempt the cloning of the cDNA for AR. In April 1987, I was invited to the ‘‘International Symposium on Hormonal Therapy of Prostatic Diseases: Basic and Clinical Aspects’’ in

Androgen Action and Modulation of Prostate and Prostate Cancer Growth

27

Milan, Italy. During the meeting, a group presented an interesting lecture and detailed experimental results in a poster claiming success in the cloning of a cDNA for AR. The AR cDNA was transcribed and then the RNA translated in an in vitro system to generate AR protein. The synthesized AR protein was used to show specific steroid binding and other interesting properties, some of which led me to question whether AR cDNA had in fact been cloned. After returning from the meeting, I decided that it was time to attempt the cloning of AR cDNA in our own laboratory because the current claim would discourage other investigators to take on the task. The cDNAs for estrogen and glucocorticoid receptors had been cloned and it was apparent that the cDNAs for these steroid receptors contained regions with similar DNA sequences. In particular, the DNA encoding the DNA-binding domain of these receptors was very similar. Degenerate oligonucleotide probes that were complementary to the DNA-binding domain of these receptors were generated and used to probe cDNA libraries from the testis and prostate. Several unique cDNAs were isolated. After sequencing various clones, one composite cDNA contained an open reading frame encoding a protein that was homologous to other steroid receptors (Chang et al. 1988a, b). When RNA was transcribed from this cDNA and translated it in a rabbit reticulocyte lysate, the product bound [3H]R1881 and the sedimentation rate of the labeled complex was shifted by an IgG fraction from the serum of a prostate cancer patient known to have antibodies against AR. Our sequence for AR cDNA was confirmed by Dr. D.P. Ma at Texas A&M University. Several other laboratories also cloned cDNAs for AR at this time or shortly thereafter (Lubahn et al. 1988; Brinkmann et al. 1989; Tilley et al. 1989).

8 AR Structure and Mutations The cDNA for both rat and human AR were cloned and sequenced in our laboratory (Chang et al. 1988b; Liao et al. 1989). The cDNA for human AR encodes a protein containing 918 amino acids with a molecular weight of about 99,000. Human AR clones from different laboratories encode proteins that vary in length due to variation in the length of polyglutamine and polyglycine repeats in AR (Fig. 4). Like other steroid receptors, AR is composed of three functional domains, which include an amino-terminal domain, a central DNA-binding domain, and a carboxylterminal ligand-binding domain. The amino-terminal domain of human AR consists of 555 amino acids and contains a ligand-independent transactivation function. This domain contains polyglutamine and polyglycine repeats that are 17- and 27-residue long in the clones we sequenced. A number of other smaller repeats are present including repeats of alanine, glutamine, leucine, proline, and serine (Fig. 4). Some of these repeats are also present in mouse (He et al. 1990) and rat AR. Expansion of the polyglutamine repeat to greater than 40 residues is associated with X-linked spinal and bulbar

28

S. Liao et al.

Fig. 4 Amino acid repeats present in human (h), rat (r), and mouse (m) AR (top) and two mutations in AR causing androgen insensitivity and one mutation altering the ligand specificity of AR in LNCaP prostate cancer cells (bottom)

muscular atrophy (Kennedy’s disease; La Spada et al. 1991). Expansion of polyglutamine repeats in other proteins has been linked to several other neurological diseases (Shao and Diamond 2007). Repeat expansion appears to endow the protein with a gain of function that leads to neuron-specific toxicity. Affected proteins are

Androgen Action and Modulation of Prostate and Prostate Cancer Growth

29

found often as insoluble aggregates within nuclei and cytoplasm. Toxicity may be the result of a disruption of the function of cell chaperones (Brignull et al. 2007). Although direct effects of the abnormal protein on neurons are likely, tissuespecific overexpression of wild-type AR in muscle leads to neuronal degeneration similar to that found in mice expressing an AR transgene containing a long polyglutamine repeat (Yu et al. 2006; Monks et al. 2007). Polyglutamine repeats are present in a variety of proteins, which include several transcriptional regulators, but a function for this structure has not been determined clearly (Butland et al. 2007). The DNA-binding domain of AR is composed of about 70 amino acids. This domain shares 56–79% identity with other steroid receptors. This domain contains two zinc atoms that are coordinated to the sulfurs of eight cysteines. The amino acids of this domain are present in a helix–loop–helix structure that interacts with specific DNA sequences called ‘‘androgen response elements.’’ Three amino acid residues within the amino-terminal helix that make critical contacts with the DNA response element are conserved among androgen, progestin, glucocorticoid, and mineralocorticoid receptors, which bind to a similar sequence of DNA. The carboxyl-terminal ligand-binding domain contains about 290 amino acids. The ligand-binding domains of ARs from humans, rats, and mice are identical, and homology to other steroid receptor ligand-binding domains is 15–45%. Like other steroid receptors, this domain is composed of 12 alpha helices that form three layers that create a hydrophobic ligand-binding pocket. The ligand-binding domain of AR bound to several agonists and antagonists has been studied using x-ray crystallography (Matias et al. 2000; Sack et al. 2001; Bohl et al. 2005a, b; Askew et al. 2007). Helices 3, 4, 5, 7, 11, and 12 form the ligand-binding pocket. Polar residues within this pocket form hydrogen bonds to the 3-keto and 17-hydroxy groups. Side chains of nonpolar amino acids form van der Waals contacts with various carbons of the steroid nucleus. Upon ligand binding, helix 12 moves to cover the ligand pocket and creates a new surface for coactivator recruitment or for interactions with the amino-terminal domain of AR (Langley et al. 1995). This process activates the second activation domain (AF-2) of AR. Differences in the ability of testosterone and dihydrotestosterone to activate AR may be related to differences in the ability of these ligands to modulate AF-2 activity (Askew et al. 2007). Cloning of the AR cDNA allowed us and many other investigators to define mutations in AR that are responsible for androgen-insensitivity in humans (Marcelli et al. 1990; Ris-Stalpers et al. 1990; Sai et al. 1990; Trifiro et al. 1991; McPhaul and Marcelli 1992; Bruggenwirth et al. 1996; Gottlieb et al. 1997). More than 300 different naturally occurring germ line and somatic mutations in the AR gene have been described that alter AR’s response to potential ligands (Gottlieb B. The Androgen Receptor Gene Mutations Database World Wide Web Server http:// androgendb.mcgill.ca). Most of these mutations reside in the ligand-binding domain of AR. Mutations in the AR ligand-binding domain cause premature termination or transitions and transversions (Fig. 4) that are responsible for androgen insensitivity (Sai et al. 1990; Ris-Stalpers et al. 1991). Androgen insensitivity in

30

S. Liao et al.

the Tfm mouse is due to a mutation (deletion) causing a frame shift mutation leading to premature termination of AR synthesis (He et al. 1991). Our laboratory and others found that human prostate cancer LNCaP cells contain a mutation T877A (Fig. 4) that alters ligand specificity of AR and allows the antiandrogen flutamide to act as an agonist (Veldscholte et al. 1990; Kokontis et al. 1991). This threonine in helix 11 is responsible for hydrogen bonding to the 17b-hydroxy group of active androgens. Mutation to alanine allows other steroids, like progesterone and estradiol, to fit into the binding pocket and activate AR. The prostate cancer cell line CWR22 has the AR mutation H874Y, which also affects ligand specificity (Tan et al. 1997). In contrast to LNCaP cells, this mutation does not directly affect a residue contacting ligand. This amino acid residue lies in a cavity formed between helices 11 and 12. Interactions between these helices are important for coactivator recruitment and the mutation may influence binding of coactivators and modulate their effects on transcription (McDonald et al. 2000).

9 Anti-AR Antibodies 9.1

Autoantibodies to AR in Prostate Cancer Patients

Jensen and colleagues were the first to obtain antibodies against a steroid receptor by immunization of rabbits with estrogen receptor isolated from a very large quantity of calf uteri (Greene et al. 1977). Similar attempts by our group to isolate enough AR for immunization were not successful. One method utilized was to covalently link 5a-DHT through its 17 position to beads by a thioether bridge. This material was used as an affinity adsorbent to bind and isolate AR. Unfortunately, the amount of AR in target tissue extracts is only on the order of 1–100 fmol (0.1–10 ng) per mg protein; therefore, affinity chromatography was not practical for isolating AR for further analysis or for immunization to obtain antibodies. In 1985, we thought that prostate cancer patients, especially those with metastases, might have autoantibodies against AR. About 200 serum samples from prostate cancer patients were obtained from a clinical chemistry laboratory and more than 30% of the samples had antibodies to AR. Three serum samples with very high titers were from prostate cancer patients with metastatic prostate cancers (Liao and Witte 1985). Gradient centrifugation was used to show that purified IgG from one of the patient’s sera interacted with 4 S cytosolic androgen–AR complexes of rat ventral prostate to form 9–12 S complexes. Other serum samples formed 14–19 S units, which suggested that different classes of immunoglobulins might be involved in AR binding or multiple epitopes were being recognized. These antibodies interacted with both cytosolic and nuclear AR complexes, but not with estrogen, progestin, or glucocorticoid receptors from a variety of sources.

Androgen Action and Modulation of Prostate and Prostate Cancer Growth

31

Antibodies also did not interact with many nonreceptor proteins. Later, these autoantibodies played an important role in identification of AR made in a cellfree system through transcription of cloned human AR cDNA and translation of the AR mRNA (see earlier). This provided strong evidence for the identity of the cDNA clone isolated. Tindall and colleagues generated monoclonal antibodies against AR by immortalizing lymphocytes from men with antibodies to AR in their sera (Young et al. 1988).

9.2

Monoclonal Antibodies to AR

Antibodies to AR are useful tools to immunohistochemically localize AR and for techniques like Western blotting. The cloned AR cDNA was used to generate fusion proteins representing different domains of AR linked to the bacterial protein TrpE. Rabbits and rats were immunized with purified fusion proteins to produce rabbit polyclonal and rat monoclonal antibodies to AR (Chang et al. 1989). These antibodies were used to localize AR in various cell types (West et al. 1990).

10

10.1

Development of the LNCaP 104 Model of Prostate Tumor Cell Progression Progression of Androgen-Sensitive Prostate Cancer Cells to Androgen-Independent and Androgen-Repressed Cells

After isolation and characterization of the AR cDNA, interest turned to exploring the role of AR in the growth and progression of prostate cancer. Our strategy was to subject androgen-sensitive prostate cancer cells to androgen deprivation in vitro to derive androgen-independent cells, and try to understand what was happening with regard to AR in the progression process. This would mimic the clinical situation in which prostate cancer recurs during androgen deprivation and/or antiandrogen therapy. At the time, around 1990, two choices for androgensensitive cells were the LNCaP human prostate cancer cell line (Horoszewicz et al. 1983) or the Shionogi carcinoma 115 mouse mammary tumor cell line (Minesita and Yamaguchi 1965). The LNCaP cell line was chosen because of its greater applicability to human prostate cancer. A hormone-dependent clonal isolate would minimize the possibility of selection of pre-existent independent clones from a potentially heterogenous population during androgen deprivation. The initial panel of about 15 LNCaP clonal isolates showed that the parental population was indeed quite heterogenous with respect to growth response to

32

S. Liao et al.

androgen (Kokontis et al. 1994; Kokontis and Liao 1999). The induction of cell growth, measured by simple cell counts of trypsinized cells after a week of incubation in the presence or absence of an optimal concentration of androgen, 0.1 nM of the synthetic androgen R1881, ranged from 1.5- to 12-fold. The clone that was most androgen sensitive, called 104-S, was used in all subsequent androgen-deprivation experiments. After about 20 passages in medium supplemented with dextran-coated charcoal-stripped fetal bovine serum, LNCaP 104-S cells gave rise to cells, called 104-I cells, that proliferated very slowly in the absence of androgen and showed greatly reduced responsiveness to exogenously supplied androgen (Kokontis et al. 1994). After about 20–25 additional passages, cells that we initially called 104-R cells emerged that grew much more rapidly in the absence of androgen. While 104-R cell growth was stimulated slightly by 0.01 nM R1881, cell growth was unexpectedly repressed by concentrations of androgen (0.1 nM R1881) that were optimal for 104-S cell proliferation. After about 60 additional passages in androgen-depleted medium, LNCaP 104-R cells gave rise to cells, called 104-R2 cells, that proliferated in the absence of androgen at a rate comparable to the proliferation rate of 104-S cells grown in androgen and continued to be repressed by 0.1 nM R1881 (Kokontis et al. 1998). The original 104-R cells were referred to as 104-R1 cells. During the transition of 104-S cells to 104-R1 and 104-R2 cells, AR mRNA and protein level increased dramatically. AR transcriptional activity, measured by androgen induction of prostate-specific antigen (PSA) mRNA or expression of an androgen-dependent reporter gene transfected into cells, also increased up to 20fold during the progression. LNCaP 104-R1 and 104-R2 cells were proliferatively insensitive to the antiandrogen bicalutamide (Casodex). Casodex, unlike flutamide and cyproterone acetate, acts as a true antiandrogen in the LNCaP cell line (Veldscholte et al. 1992a, b). Casodex did not exhibit agonist activity in 104-R1 and 104-R2 cells, as has been reported by other investigators for androgen-independent LNCaP cells (Culig et al. 1999; Chen et al. 2004). Casodex repressed androgen target genes, such as PSA in 104-R1 and 104-R2 cells, and Casodex blocked androgenic repression of growth (Kokontis et al. 1998). This observation suggests that 104-R1 and 104-R2 cells are truly androgen independent and are not just more efficient at scavenging residual androgen in the medium. The basal level of PSA expression did not rise appreciably in 104-R1 and 104-R2 cells, which indicates that AR transactivation of PSA expression continues to be androgen sensitive, even when proliferation is androgen independent. When 104-R1 cells were incubated for several weeks in a high concentration of R1881 (20 nM), cells adapted after a period of growth arrest to grow at a rate equivalent to the parental 104-R1 cells (Kokontis et al. 1998). These cells, called 104-R1Ad cells, grew optimally in 10 nM R1881 and Casodex inhibited R1881-induced growth. 104-R1Ad cells exhibit drastically reduced levels of AR, which suggests that elevated AR expression is responsible for the repressive effect of androgen in 104-R1 and 104-R2 cells.

Androgen Action and Modulation of Prostate and Prostate Cancer Growth

10.2

33

Progression of Prostate Cancer Cells to Androgen-Insensitive Cancer Cells

LNCaP 104-S cells were incubated with 5 mM Casodex to more closely mimic the clinical situation of combined androgen deprivation and antiandrogen therapy. After three to four weeks, Casodex-resistant colonies appeared at low frequency as most of the cells appeared to undergo senescent cell death (Kokontis et al. 2005). This frequency increased with increasing 104-S passage number, which suggests that a subpopulation of androgen-independent cells emerges spontaneously over time even in a clonal population of cells grown in the presence of androgen (Fig. 5). The presence of this subpopulation was manifested by selection in antiandrogen. Six independent Casodex-resistant clonal sublines were derived from these colonies, designated CDXR1 through CDXR6. Like 104-R1 and 104-R2 cells, CDXR cells had increased AR expression and activity, and were repressed by androgen (Kokontis et al. 2005). Unlike 104-R1 cells, most CDXR1, CDXR2, and CDXR3 cells grown in 10 nM R1881 underwent apoptotic cell death starting 6–8 days after R1881 exposure. At low frequency, however, R1881-insensitive colonies appeared that were not repressed by R1881 or Casodex. These sublines, designated IS1, IS2, and IS3 (derived from the corresponding CDXR subline), showed greatly reduced

Fig. 5 Frequency of Casodex-resistant cells increases with LNCaP 104-S passage number. Equal numbers of LNCaP 104-S cells at increasing passage numbers were plated in 12-well plates and grown in medium supplemented with 10% charcoal-stripped FBS and 5 mM Casodex. After 2 weeks, cells were trypsinized and replated to allow expansion of rare colonies. After 4 additional weeks of growth in 5 mM Casodex, wells were scored as growth or nongrowth. Values for frequency were calculated from Poisson’s formula: a = (ln p0)/N, where a is frequency of CDXR cells in the population, p0 is fraction of wells with no Casodex-resistant cell growth, and N is initial cell number/well

34

S. Liao et al.

Fig. 6 Progression scheme of LNCaP 104-S cells after androgen (A) deprivation or Casodex (CDX) treatment, and androgen replacement

AR expression. But unlike 104-R1Ad cells, the growth of IS cells was not stimulated by R1881. 104-R2 cells, like CDXR cells and unlike 104-R1 cells, gave rise to androgen-insensitive cells when challenged by high concentration R1881. During progression from 104-R1 to 104-R2 stages, therefore, the cells appear to pass a point where cells can no longer recover responsiveness to androgen, but instead progress to androgen insensitivity (Liao et al. 2005). Direct progression of 104-S cells to the CDXR stage by selection in antiandrogen seems to bypass this intermediate 104-R1 stage. This suggests that the more aggressive combined treatment may cause a more rapid and irreversible selection of independent cells that have lost the potential to reacquire androgen responsiveness and become controllable by endocrine strategies. Androgen-deprivation therapy alone, on the other hand, may promote a slow adaptation to androgen independence that is reversible up to an as yet undefined point during the transition from 104-R1 to 104-R2 cells. This progression scheme is summarized in Fig. 6. Knockdown of AR expression in CDXR3 cells by shRNAi methodology, either constitutive or conditional, relieved androgenic repression of growth and did not affect cell growth in the absence of androgen (Kokontis et al. 2005). These data suggested that elevated AR expression is not required for androgen-independent growth, contrary to reports by others (Zegarra-Moro et al. 2002; Haag et al. 2005; Yuan et al. 2006). Retroviral overexpression of AR in IS2 and IS3 cells, on the other hand, restored the androgen-repressed phenotype in these cells, which confirms that elevated AR expression alone was sufficient for conferring this response in LNCaP cells. Conditional overexpression of AR in 104-S cells also caused androgeninduced growth repression and did not confer androgen-independent growth. Constitutive overexpression of AR is not tolerated in 104-S cells because the cells cannot proliferate either in the presence or absence of androgen. These observations suggest that enforced overexpression of AR may have therapeutic benefit to repress prostate cancer cell growth. The ability of CDXR cells to down-regulate AR expression during progression to the IS stage and rapidly proliferate in the absence of AR expression may indicate that therapeutic strategies based on reduction of AR expression may be easily bypassed by many prostate cancer cells.

Androgen Action and Modulation of Prostate and Prostate Cancer Growth

11

11.1

35

Androgen Suppression of Androgen-Independent Prostate Cancer Androgen Induction of G1 Arrest in Androgen-Repressed Prostate Cancer Cells

In an effort to determine the mechanisms by which androgen modulates LNCaP 104-S, 104-R1, and 104-R2 cell proliferation, the expression of known cdk inhibitors (CKIs; including p15, p16, p18, p19, and p21waf1/cip1, p27Kip1, p57Kip2) was examined in cells treated with or deprived of androgen. Flow cytometric analysis of androgen-treated cells revealed that androgen treatment of 104-S cells relieved a G1 arrest induced by androgen deprivation. Conversely, R1881 induced G1 arrest in 104-R1 and 104-R2 cells beginning after about 24 h of exposure (Kokontis et al. 1998). Casodex could block the effect of androgen in all cell lines. Of all tested CKIs, only p21waf1/cip1 and p27Kip1 levels were induced by androgen (Kokontis et al. 1998). p21waf1/cip1 was induced transiently in 104-R1 cells only, while p27Kip1 was induced persistently about threefold in both 104-R1 and 104-R2 cells. In contrast, expression of p21waf1/cip1 and p27Kip1 was repressed by androgen in 104-S cells, which suggests that these inhibitors function in androgen-deprivation-induced G1 arrest. Cdk2 was the most androgen-sensitive cdk. Similar results were obtained with the CDXR sublines (Kokontis et al. 2005). Androgens regulate expression of the F-box protein Skp2 that binds phosphorylated p27Kip1 (Carrano et al. 1999; Tsvetkov et al. 1999; Lu et al. 2002) leading to its ubiquitination and proteolysis. Androgen also down-regulates Skp2 in 104-R1, 104-R2, and CDXR cells, which leads to accumulation of p27Kip1. Androgen also rapidly down-regulates c-Myc at the mRNA level in 104-R1 cells, and enforced retroviral overexpression of c-Myc blocks androgenic repression of 104-R1 growth (Kokontis et al. 1994). c-Myc may have an indirect effect on p27 Kip1 expression through the induction of Cks1, a component of the SCFSkp2 complex responsible for p27 Kip1 degradation (Keller et al. 2007).

11.2

Androgen Suppression of Prostate Cancer Xenografts

The tumorigenicity of LNCaP 104-S and 104-R2 cells was examined in normal and castrated athymic mice (Umekita et al. 1996). LNCaP 104-S cells grow well in normal male mice but do not form tumors in castrated male mice, and this reflects their androgen dependence. On the other hand, 104-R2 cells grow well in castrated hosts but not in normal male mice, consistent with their androgen independence and sensitivity to the repressive effects of androgen. In fact, finasteride reversed the repressive effects of testosterone implants on 104-R2 tumor growth, similar to the blockage of androgen repression by Casodex in vitro. When mice bearing 104-R1

36

S. Liao et al.

Fig. 7 Growth of 104-R1 and 104-R1Ad tumors in castrated athymic mice. Open symbols represent tumors grown without testosterone replacement and closed symbols represent tumors grown in mice subcutaneously implanted with testosterone propionate pellets (20 mg). Circles represent 104-R1 tumors grown without initial testosterone replacement and squares represent 104-R1Ad tumors grown with initial testosterone replacement

tumors are treated with testosterone, reversion to an androgen-stimulated phenotype occurs following a period of growth repression (Chuu et al. 2005), similar to the effect of androgen on 104-R1 cells in vitro. If testosterone pellets are removed from mice bearing 104-R1Ad tumors, tumor growth ceases until pellets are reimplanted, at which time growth resumes (Fig. 7). On the other hand, CDXR3 tumors are repressed by androgen treatment, and tumors either regress or relapse after about 90 days of treatment (Kokontis et al. 2005). The relapsed tumors now show diminished expression of AR, but no longer require androgen for growth, essentially identical to the behavior of IS3 cells that emerged after androgen exposure in vitro.

11.3

AR as Tumor Suppressor

Other groups have derived androgen-repressed sublines from the LNCaP cell line after androgen deprivation (Joly-Pharaboz et al. 1995, 2000; Soto et al. 1995; Geck et al. 1997; Culig et al. 1999; Shi et al. 2004). Abreu-Martin et al. (1999) reported that a constitutively activated form of MEKK1 could induce apoptosis in AR-positive cells but not in AR-negative cells. Lin et al. (2006) showed that AR was necessary for Bax-induced apoptosis in LNCaP cells and that androgen enhanced Bax-induced cell death. Recently AR was reported to repress expression of c-Met, the receptor for hepatocyte growth factor, in LNCaP cells by interfering with Sp1 binding to the c-Met promoter region (Verras et al. 2007). Increased c-Met signaling has been associated with prostate cancer progression (Knudsen et al. 2002; Knudsen and Edlund 2004). Evidence is increasing for the androgenrepressed phenotype in addition to that from the LNCaP cell line. The ARCaP cell line (Zhau et al. 1996) is another androgen-repressed prostate cancer cell line. Hara et al. (2003) reported suppression of MDA PCa 2b-hr prostate cancer cell

Androgen Action and Modulation of Prostate and Prostate Cancer Growth

37

growth by high-dose androgen after long term androgen deprivation. PC-3 cells ectopically expressing AR are repressed by androgen (Yuan et al. 1993; Heisler et al. 1997; Litvinov et al. 2004). Bruchovsky et al. (2000) observed possible androgenic repression of serum PSA in an interim analysis of the Canadian Prospective Trial of intermittent androgen-deprivation therapy. Decreased androgen levels during aging may contribute to the onset of prostate cancer in some men, and androgen replacement may benefit many patients with castration-recurrent prostate cancer as well as improve quality of life (Prehn 1999; Algarte-Genin et al. 2004). In an early study of androgen replacement therapy in prostate cancer patients, 7 of 52 patients experienced symptomatic benefit, while the remainder had unfavorable responses (Fowler and Whitmore 1981). Since this study predated convenient quantification of AR expression levels and analysis of AR expression was not performed, the seven patients with favorable responses may represent a patient subgroup with tumors expressing higher levels of AR than the patients who showed unfavorable responses. A more recent phase I study demonstrated no adverse effects of testosterone replacement administered by skin patch or topical gel in patients with castration-recurrent and metastatic disease (Morris et al. 2004). Phase II studies are planned. A recently published case study has findings similar to our animal studies and supports the use of testosterone to treat certain prostate cancer patients (Mathew 2008). In this case study, a prostate cancer patient had undergone radical prostatectomy and received LH-RH therapy. Testosterone levels remained at castrated levels and PSA was undetectable for 15 years. PSA levels then began to rise and the patient was given parenteral testosterone replacement therapy to attain a normal range of serum testosterone. After an initial flare, PSA levels gradually declined over 18 months. After 27 months, PSA level started to increase. When testosterone replacement therapy was discontinued, PSA levels dropped. Testosterone therapy also improved the patient’s quality of life. In the Prostate Cancer Prevention Trial, treatment with finasteride was reported to decrease the incidence of low-grade prostate cancer, but increase the incidence of high-grade prostate cancer (Thompson et al. 2003). The latter finding, however, may have resulted from detection bias due to finasteride-induced reduction in prostate volume (Cohen et al. 2007; Lucia et al. 2007). In experimental systems, androgen has been reported to repress cell growth during androgen-induced differentiation of immortalized human and rodent prostate epithelial cells in vitro (Ling et al. 2001; Whitacre et al. 2002). Bruchovsky et al. (1975) also described the repression by androgen on rat ventral prostate epithelial cell proliferation in the hormonal regulation of prostate homeostasis. In testosterone-induced rat ventral prostate regeneration, secretory epithelial cells enter a state of quiescence after an initial burst of cell proliferation (Waltregny et al. 2001). The proliferative status of cells correlated inversely with p27Kip1 expression. Normal differentiated prostate luminal secretory cells express high levels of p27Kip1 (Cote et al. 1998), and a strong correlation exists between low p27Kip1 expression in tumor cells and poor prognosis clinically (Guo et al. 1997; Cote et al. 1998; Yang et al. 1998; Vis et al. 2000). Recently, a low testosterone level in Nkx3.1:Pten mutant mice was reported to accelerate the progression of prostate cancer compared with a normal testosterone level

38

S. Liao et al.

(Banach-Petrosky et al. 2007). AR has been suggested to play a direct inhibitory role in the licensing of DNA for replication (Litvinov et al. 2006; Vander Griend et al. 2007). In this role, AR degradation in prostate tumor cells is hypothesized to be required during mitosis to allow origins of replication to acquire the capacity to fire in the next round of DNA replication. Overexpression of AR may prevent complete degradation of all AR and thereby repress cell cycle progression. This mechanism may operate parallel to the action of AR in regulating c-Myc and p27 levels in prostate cancer cells. c-Myc itself has recently been reported to function nontranscriptionally in the activation of origins of replication (Dominguez-Sola et al. 2007). Much evidence therefore exists for AR functioning as a liganddependent tumor suppressor in prostate cancer cells when it is expressed at high levels and is fully activated.

12

12.1

Modulation of Liver X Receptor Signaling as Novel Therapy for Prostate Cancer Discovery and Characterization of Ubiquitous Receptor (UR or LXR)

The cloning of cDNAs for various steroid receptors led to the realization that conserved structure and sequence existed within this protein family and that new members of this family could be identified by using techniques relying on cDNA sequence homology especially within the DNA-binding domain. In 1993 as part of a project to identify a receptor for the weak androgen D5-androstenediol that produces estrogen-like effects on the rat vaginal epithelium (Shao et al. 1975), a novel member of the nuclear receptor superfamily was discovered (Song et al. 1994). This new receptor was named ubiquitous receptor or UR because of its expression in many tissues. cDNAs for the same receptor or paralogs of this receptor were isolated around this time or shortly after including LXR (Willy et al. 1995), NER (Shinar et al. 1994), RIP15 (Seol et al. 1995), RLD-1 (Apfel et al. 1994), and OR-1 (Teboul et al. 1995). Since one isoform of this receptor was present in liver at high levels, it was called liver X receptor (LXR). Eventually, the isoform most abundant in liver became known as LXRa (NR1H3), and the more ubiquitously expressed isoform became known as LXRb(NR1H2). LXRa is also abundant in other tissues besides liver, which include adipose tissue, adrenal gland, brain, intestine, kidney, testis, spleen, and macrophages (Bookout et al. 2006). LXRb is expressed in many mouse tissues and is expressed at higher levels than LXRa in most tissues (Bookout et al. 2006). LXRb is expressed in many cultured cell lines and most of the developing tissues of the mouse embryo based on immunocytochemistry (Song et al. 1995). The human gene for LXRb was localized to chromosome 19q13.3 using fluorescence in situ hybridization, (Le Beau et al. 1995). The gene for LXRa is located at 11p11.2.

Androgen Action and Modulation of Prostate and Prostate Cancer Growth

12.2

39

UR/LXR is an Oxysterol Receptor

LXRa and LXRb form heterodimers with the obligate partner retinoid X receptor that binds the ligand 9-cis retinoic acid (Song et al. 1994; Willy et al. 1995). The LXR/RXR heterodimers can be activated with either LXR or RXR agonists. Naturally occurring oxysterols, such as 22(R)-hydroxy-, 24-hydroxy-, 25-hydroxy-, 27-hydroxy-, and 24(S), 25-epoxycholesterol, bind to and activate transcription by LXR (Janowski et al. 1996, 1999; Lehmann et al. 1997; Fu et al. 2001; Spencer et al. 2001). Cholestenoic acid is also a natural ligand for LXR (Song and Liao 2000). A few synthetic LXR agonists have been developed, which include nonsteroidal compounds, such as T0901317 (Schultz et al. 2000) and GW3965 (Collins et al. 2002). Analogs of 6a-hydroxy bile acids activate LXR (Song et al. 2000) and oral administration of a 6a-hydroxy bile acid analog, hypocholamide (3a, 6a-dihydroxy-5b-cholanic acid-N-methyl-N-methoxy-24-amide), produces hypolipidemic effects in hypercholestermic rats, mice, and hamsters (Song and Liao 2001). Hypocholamide also enhances cholesterol and phospholipid transport to apoE in cultured mouse astrocytes (Peng et al. 2003). The compound 3a, 6a, 24trihydroxy-24, 24-di-(trifluoromethyl)-5b-cholane is a potent agonist for LXR, reduces atherosclerosis in LDLR-/- mice fed a Western diet, and does not induce triglyceridemia, a side effect of synthetic LXR agonists, like T00901317 that has complicated development of LXR agonists as drugs to treat cardiovascular disease (Peng et al. 2008). The auto-oxidized cholesterol sulfates, 5a, 6a-epoxycholesterol3-sulfate and 7-ketocholesterol-3-sulfate, often present in atherosclerotic plaques are antagonists of LXR (Song et al. 2001). LXRs are important regulators of cholesterol, fatty acid, and glucose homeostasis and modulate inflammation (Zelcer and Tontonoz 2006). LXR has an important role in reverse cholesterol transport, since it regulates the expression of the cholesterol transporters ABCA1, ABCG1, ABCG5, and ABCG8 and the lipid acceptor lipoprotein apoE (Zelcer and Tontonoz 2006). Composition of lipoproteins may also be controlled by LXR, since it regulates the expression of cholesterol ester transfer protein and phospholipid transfer protein, which are involved in lipid transfer between different lipoproteins (Zelcer and Tontonoz 2006). Treatment with an LXR agonist lowers the cholesterol level in serum and liver and inhibits the development of atherosclerosis in murine models of this disease (Peet et al. 1998; Alberti et al. 2001; Joseph et al. 2002). LXR also regulates fatty acid synthesis by modulating the expression of sterol regulatory element-binding protein-1c (SREBP-1c) and downstream lipogenic genes, which include acetyl CoA carboxylase and fatty acid synthase (Repa et al. 2000; Yoshikawa et al. 2001; Liang et al. 2002). LXR agonists suppress gluconeogenesis, induce expression of glucokinase in liver, stimulate insulin secretion, reduce plasma glucose, and improve glucose tolerance and insulin resistance in murine and rat obesity models (Cao et al. 2003; Laffitte et al. 2003; Efanov et al. 2004). Glucose is a low-affinity ligand for LXR and can modulate its transcriptional activity at physiological concentrations (Mitro et al. 2006). LXR agonists can have an anti-inflammatory effect (Joseph et al. 2003), and LXR signaling is important for brain function (Wang et al. 2002; Andersson et al. 2005).

40

12.3

S. Liao et al.

Antiproliferative Effect of LXR Agonists

Production and secretion of cholesterol increases in prostate cancer compared to healthy prostate tissue (Sporer et al. 1982). Expression of fatty acid synthase, the enzyme responsible for converting acetate to fatty acids, is up-regulated in early stage prostate cancer and further increases in advanced prostate cancer (Epstein et al. 1995; Shurbaji et al. 1996; Swinnen et al. 2002). Since LXR signaling regulates both cholesterol and fatty acid homeostasis, it is possible that LXR agonists may be good candidates for treatment or management of prostate cancer. Since 2000, our laboratory has had projects focusing on identification of new LXR agonists for cardiovascular and Alzheimer’s disease therapy and novel therapies against prostate cancer. In 2004, several natural compounds including oxysterols used in the LXR project were tested in an assay of prostate cancer progression. Unexpectedly, we found that 22(R)-hydroxycholesterol, 24(S)-hydroxycholesterol, and T0901317 suppressed the proliferation of both androgen-dependent 104-S cells and androgen-independent 104-R1 cells (Fukuchi et al. 2004). These LXR agonists also suppressed the proliferation of AR-negative prostate cancer cell lines PC-3 and DU145. The suppression of cell proliferation using LXR agonists was via induction of G1 cell cycle arrest. T0901317 decreased the percentage of cells in S-phase, increased the percentage of cells in G1-phase, suppressed the expression of Skp2, and promoted the accumulation of cell cycle inhibitor p27Kip1. In vivo, administration of T0901317 suppressed prostate tumor growth in athymic mice. T0901317 (10 mg/kg/day) inhibited the growth of androgen-sensitive LNCaP 104-S tumors, which resulted in a twofold difference in mean tumor volume between the control and the LXR agonist-treated mice (Fig. 8). T0901317 was found later to suppress the proliferation of many commonly used cancer cell lines (Chuu et al. 2007).

12.4

Inhibition of Prostate Cancer Progression Using LXR Agonists

Since expression of LXRa and its target gene ABCA1 was lower in androgenindependent 104-R1 and 104-R2 cells compared to androgen-dependent 104-S cells (Fukuchi et al. 2004), down-regulation of LXR signaling-related genes may assist the progression of prostate cancer cell toward androgen independency. A prostate cancer progression model was developed in athymic mice using LNCaP cells that mimic the cell culture model. LNCaP 104-S cells were inoculated into athymic mice to form tumors. Androgen-sensitive 104-S tumor xenografts progressed to androgen-independent tumors (104-Rrel) in athymic mice after castration (Chuu et al. 2006). The growth of 104-Rrel tumors was suppressed using testosterone. AR expression in tumors and serum PSA increased during progression from 104-S to 104-Rrel (Chuu et al. 2006). Expression of LXRa, ABCA1, and another LXRa target gene, sterol 27-hydroxylase (CYP27), decreased during progression from

Androgen Action and Modulation of Prostate and Prostate Cancer Growth

41

Fig. 8 Inhibition of proliferation and progression of prostate cancer by the LXR agonist T0901317. (a) Inhibition of androgen-sensitive LNCaP 104-S tumor growth in intact mice by treatment with the LXR agonist T0901317. Mice were administered 10 mg/kg T0901317 ( filled square, 10 mice with 13 tumors) or vehicle alone ( filled circle, 10 mice with 15 tumors) using gavage once a day during the experimental period. Relative tumor volumes were expressed as mean SE. (b) Inhibition of progression of androgen-sensitive LNCaP 104-S tumors toward androgen-independent 104-Rrel tumors in castrated mice by treatment with the LXR agonist T0901317. After castration of control mice that were used in Fig. 5a, mice were administered 10 mg/kg T0901317 ( filled square, 9 mice with 15 tumors) or vehicle alone ( filled circles, 9 mice with 13 tumors) by gavage five times a week during the experimental period. Relative tumor volumes were expressed as mean SE

104-S to 104-Rrel (Chuu et al. 2006). The development of androgen-independent relapsed tumors in athymic mice bearing 104-S tumors after castration was delayed by 4 weeks in mice treated with T0901317 (10 mg/kg; Fig. 8) compared to the control group (Chuu et al. 2006). Therefore, LXR agonists can retard progression of prostate cancer in vivo.

12.5

Phytosterols and LXR Signaling

Phytosterols, plant equivalents of mammalian cholesterol, are essential components of all plants. Phytosterols are abundant in plant oil, seeds of legumes and nuts, as well as vegetables and fruits. The most common phytosterols include b-sitosterol, campesterol, and stigmasterol. Dietary phytosterols are believed to be beneficial to health and are reported to have anticancer activity. Treatment with b-sitosterol also retards the growth and metastasis of ARnegative PC-3 tumors in mice (Awad et al. 2001) and the proliferation of ARpositive LNCaP cells in culture (Awad et al. 2000). The mechanism involved in the anticancer activity of phytosterols is not clear. However, plant sterols and stanols

42

S. Liao et al.

from the 4-desmethylsterol family activate LXRa and LXRb (Plat et al. 2005) with EC50 of 30–150 nM. Phytosterols, therefore, may suppress growth and metastasis of prostatic cancer through activation of LXR signaling. LXR agonists inhibit proliferation and progression of prostate cancer cells both in vitro and in vivo. LXR agonists also suppress the proliferation of breast cancer cells in vitro. Phytosterols, a recently discovered group of LXR agonists, inhibit proliferation of several prostate and breast cancer cell lines. Synthetic nonsteroidal LXR agonists T0901317 and GW3965 exhibit more potent agonistic effect than natural oxysterols. Although T0901317 and GW3965 were reported to be beneficial for treatment of atherosclerosis, diabetes, and Alzheimer’s disease in murine disease models, both T0901317 and GW3965 have been reported to increase plasma and liver triglycerides in some mice models. The synthetic steroidal LXR agonist hypocholamide, a 6a-hydroxylated analog of bile acids, shows an overall hypolipidemic effect but does not increase the serum triglyceride level (Song and Liao 2001). YT-32, a synthetic LXR agonist developed by modifying the phytosterol structure, selectively activated intestinal ABC transporters in mice without increasing plasma triglyceride levels (Kaneko et al. 2003). These findings with hypocholamide and YT-32 suggest the possibility of developing potent and useful LXR agonist drugs through modification of bile acid and phytosterol structure. Although the exact mechanism responsible for inhibition of prostate cancer progression by LXR agonists requires further study, modulation of LXR signaling may be a novel and potentially important therapy for prostate and other cancers.

13

Conclusion

The journey into the realm of androgen biology has taken us to many different areas, and much has been learned about the mechanism of action of this steroid hormone over the past 50 years. Many individuals and many different laboratories have contributed in this effort. Even though our knowledge in this field has increased enormously, much more remains to be learned and perhaps some surprises await us.

References Abreu-Martin, M. T., Chari, A., Palladino, A. A., Craft, N. A. and Sawyers, C. L. 1999. Mitogenactivated protein kinase kinase kinase 1 activates androgen receptor-dependent transcription and apoptosis in prostate cancer. Mol. Cell. Biol. 19:5143–5154. Alberti, S., Schuster, G., Parini, P., Feltkamp, D., Diczfalusy, U., Rudling, M., Angelin, B., Bjorkhem, I., Pettersson, S. and Gustafsson, J. A. 2001. Hepatic cholesterol metabolism and resistance to dietary cholesterol in LXRbeta-deficient mice. J. Clin. Invest. 107:565–573. Algarte-Genin, M., Cussenot, O. and Costa, P. 2004. Prevention of prostate cancer by androgens: experimental paradox or clinical reality. Eur. Urol. 46:285–295.

Androgen Action and Modulation of Prostate and Prostate Cancer Growth

43

Anderegg, R. J., Carr, S. A., Huang, I. Y., Hiipakka, R. A. and Liao, S. 1988. Correction of the cDNA-derived protein sequence of prostatic spermine binding protein: pivotal role of tandem mass spectroscopy in sequence analysis. Biochemistry 27:4214–4221. Anderson, K. M. and Liao, S. 1968. Selective retention of dihydrotestosterone by prostatic nuclei. Nature 219:277–279. Andersson, S., Bishop, R. W. and Russell, D. W. 1989. Expression cloning and regulation of steroid 5 alpha-reductase, an enzyme essential for male sexual differentiation. J. Biol. Chem. 264:16249–16255. Andersson, S., Berman, D. M., Jenkins, E. P. and Russell, D. W. 1991. Deletion of steroid 5 alphareductase 2 gene in male pseudohermaphroditism. Nature 354:159–161. Andersson, S., Gustafsson, N., Warner, M. and Gustafsson, J. A. 2005. Inactivation of liver X receptor beta leads to adult-onset motor neuron degeneration in male mice. Proc. Natl. Acad. Sci. USA 102:3857–3862. Apfel, R., Benbrook, D., Lernhardt, E., Ortiz, M. A., Salbert, G. and Pfahl, M. 1994. A novel orphan receptor specific for a subset of thyroid hormone-responsive elements and its interaction with the retinoid/thyroid hormone receptor subfamily. Mol. Cell. Biol. 14:7025–7035. Askew, E. B., Gampe, R. T., Jr., Stanley, T. B., Faggart, J. L. and Wilson, E. M. 2007. Modulation of androgen receptor activation function 2 by testosterone and dihydrotestosterone. J. Biol. Chem. 282:25801–25816. Awad, A. B., Gan, Y. and Fink, C. S. 2000. Effect of beta-sitosterol, a plant sterol, on growth, protein phosphatase 2A, and phospholipase D in LNCaP cells. Nutr. Cancer 36:74–78. Awad, A. B., Fink, C. S., Williams, H. and Kim, U. 2001. In vitro and in vivo (SCID mice) effects of phytosterols on the growth and dissemination of human prostate cancer PC-3 cells. Eur. J. Cancer Prev. 10:507–513. Banach-Petrosky, W., Jessen, W. J., Ouyang, X., Gao, H., Rao, J., Quinn, J., Aronow, B. J. and Abate-Shen, C. 2007. Prolonged exposure to reduced levels of androgen accelerates prostate cancer progression in Nkx3.1; Pten mutant mice. Cancer Res. 67:9089–9096. Bettuzzi, S., Hiipakka, R. A., Gilna, P. and Liao, S. 1989. Identification of the androgen-repressed mRNA for a 48-kilodalton prostate protein as sulfated glycoprotein-2 by cDNA cloning and sequence analysis. Biochem. J. 257:293–296. Bettuzzi, S., Brausi, M., Rizzi, F., Castagnetti, G., Peracchia, G. and Corti, A. 2006 Chemoprevention of human prostate cancer by oral administration of green tea catechins in volunteers with high-grade prostate intraepithelial neoplasia: a preliminary report from a one-year proofof-principle study. Cancer Res. 66:1234–1240. Bohl, C. E., Gao, W., Miller, D. D., Bell, C. E. and Dalton, J. T. 2005a. Structural basis for antagonism and resistance of bicalutamide in prostate cancer. Proc. Natl. Acad. Sci. USA 102:6201–6206. Bohl, C. E., Miller, D. D., Chen, J., Bell, C. E. and Dalton, J. T. 2005b. Structural basis for accommodation of nonsteroidal ligands in the androgen receptor. J. Biol. Chem. 280:37747– 37754. Bookout, A. L., Jeong, Y., Downes, M., Yu, R. T., Evans, R. M. and Mangelsdorf, D. J. 2006. Anatomical profiling of nuclear receptor expression reveals a hierarchical transcriptional network. Cell 126:789–799. Bramson, H. N., Hermann, D., Batchelor, K. W., Lee, F. W., James, M. K. and Frye, S. V. 1997. Unique preclinical characteristics of GG745, a potent dual inhibitor of 5AR. J. Pharmacol. Exp. Ther. 282:1496–1502. Brignull, H. R., Morley, J. F. and Morimoto, R. I. 2007. The stress of misfolded proteins: C. elegans models for neurodegenerative disease and aging. Adv. Exp. Med. Biol. 594:167–189. Brinkmann, A. O., Klassen, P., Kuiper, G. G. J. M., van der Korput, J. A. G. M., Bolt, J., de Boer, W., Smit, A., Faber, P. W., van Rooij, H. C. J., Geurts van Kessel, A., Voorhorst, M. M., Mulder, E. and Trapman, J. 1989. Structure and function of the androgen receptor. Urol. Res. 17:87–93. Bruchovsky, N. and Wilson, J. D. 1968. Evidence that dihydrotestosterone is the active form of testosterone. Clin. Res. 16:74.

44

S. Liao et al.

Bruchovsky, N., Lesser, B., Van Doorn, E. and Craven, S. 1975. Hormonal effects on cell proliferation in rat prostate. In Vitamins and Hormones, eds. Munson, P. L., Glover, J., Diczfalusy, E. and Olson, R. E. pp. 61–102. New York: Academic Press. Bruchovsky, N., Klotz, L. H., Sadar, M., Crook, J. M., Hoffart, D., Godwin, L., Warkentin, M., Gleave, M. E. and Goldenberg, S. L. 2000. Intermittent androgen suppression for prostate cancer: Canadian prospective trial and related observations. Mol. Urol. 4:191–200. Bruggenwirth, H. T., Boehmer, A. L., Verleun-Mooijman, M. C., Hoogenboezem, T., Kleijer, W. J., Otten, B. J., Trapman, J. and Brinkmann, A. O. 1996. Molecular basis of androgen insensitivity. J. Steroid Biochem. Mol. Biol. 58:569–575. Butland, S. L., Devon, R. S., Huang, Y., Mead, C. L., Meynert, A. M., Neal, S. J., Lee, S. S., Wilkinson, A., Yang, G. S., Yuen, M. M., Hayden, M. R., Holt, R. A., Leavitt, B. R. and Ouellette, B. F. 2007. CAG-encoded polyglutamine length polymorphism in the human genome. BMC Genomics 8:126. Cao, G., Liang, Y., Broderick, C. L., Oldham, B. A., Beyer, T. P., Schmidt, R. J., Zhang, Y., Stayrook, K. R., Suen, C., Otto, K. A., Miller, A. R., Dai, J., Foxworthy, P., Gao, H., Ryan, T. P., Jiang, X. C., Burris, T. P., Eacho, P. I. and Etgen, G. J. 2003. Antidiabetic action of a liver x receptor agonist mediated by inhibition of hepatic gluconeogenesis. J. Biol. Chem. 278:1131– 1136. Carrano, A. C., Eytan, E., Hershko, A. and Pagano, M. 1999. SKP2 is required for ubiquitinmediated degradation of the CDK inhibitor p27. Nat. Cell Biol. 1:193–199. Castaneda, E. and Liao, S. 1974. A new method for the characterization of androgen receptors by use of a steroid antibody. Endocr. Res. Commun. 1:271–281. Castaneda, E. and Liao, S. 1975a. Assay of cellular steroid receptors using steroid antibodies. Methods Enzymol. 36:52–58. Castaneda, E. and Liao, S. 1975b. The use of anti-steroid antibodies in the characterization of steroid receptors. J. Biol. Chem. 250:883–888. Chang, C. and Liao, S. 1987. Topographic recognition of cyclic hydrocarbons and related compounds by receptors for androgens, estrogens, and glucocorticoids. J. Steroid Biochem. 27:123–131. Chang, C., Saltzman, A. G., Sorensen, N. S., Hiipakka, R. A. and Liao, S. 1987. Identification of glutathione S-transferase Yb1 mRNA as the androgen-repressed mRNA by cDNA cloning and sequence analysis. J. Biol. Chem. 262:11901–11903. Chang, C., Kokontis, J. M. and Liao, S. 1988a. Molecular cloning of human and rat complementary DNA encoding androgen receptors. Science 240:324–326. Chang, C., Kokontis, J. M. and Liao, S. 1988b. Structural analysis of complementary DNA and amino acid sequences of human and rat androgen receptors. Proc. Natl. Acad. Sci. USA 85:7211–7215. Chang, C. S., Whelan, C. T., Popovich, T. C., Kokontis, J. and Liao, S. 1989. Fusion proteins containing androgen receptor sequences and their use in the production of poly- and monoclonal anti-androgen receptor antibodies. Endocrinology 125:1097–1099. Chen, C., Schilling, K., Hiipakka, R. A., Huang, I. Y. and Liao, S. 1982. Prostate alpha-protein. Isolation and characterization of the polypeptide components and cholesterol binding. J. Biol. Chem. 257:116–121. Chen, C. D., Welsbie, D. S., Tran, C., Baek, S. H., Chen, R., Vessella, R., Rosenfeld, M. G. and Sawyers, C. L. 2004. Molecular determinants of resistance to antiandrogen therapy. Nat. Med. 10:33–39. Chuu, C. P., Hiipakka, R. A., Fukuchi, J., Kokontis, J. M. and Liao, S. 2005. Androgen causes growth suppression and reversion of androgen-independent prostate cancer xenografts to an androgen-stimulated phenotype in athymic mice. Cancer Res. 65:2082–2084. Chuu, C.-P., Hiipakka, R. A., Kokontis, J. M., Fukuchi, J., Chen, R.-Y. and Liao, S. 2006. Inhibition of tumor growth and progression of LNCaP prostate cancer cells in athymic mice by androgen and liver X receptor agonist. Cancer Res. 66:6482–6486. Chuu, C. P., Kokontis, J. M., Hiipakka, R. A. and Liao, S. 2007. Modulation of liver X receptor signaling as novel therapy for prostate cancer. J. Biomed. Sci. 14:543–553.

Androgen Action and Modulation of Prostate and Prostate Cancer Growth

45

Clever, U. and Karlson, P. 1960. [Induction of puff changes in the salivary gland chromosomes of Chironomus tentans by ecdysone]. Exp. Cell Res. 20:623–626. Cohen, Y. C., Liu, K. S., Heyden, N. L., Carides, A. D., Anderson, K. M., Daifotis, A. G. and Gann, P. H. 2007. Detection bias due to the effect of finasteride on prostate volume: a modeling approach for analysis of the Prostate Cancer Prevention Trial. J. Natl. Cancer Inst. 99:1366– 1374. Collins, J. L., Fivush, A. M., Watson, M. A., Galardi, C. M., Lewis, M. C., Moore, L. B., Parks, D. J., Wilson, J. G., Tippin, T. K., Binz, J. G., Plunket, K. D., Morgan, D. G., Beaudet, E. J., Whitney, K. D., Kliewer, S. A. and Willson, T. M. 2002. Identification of a nonsteroidal liver X receptor agonist through parallel array synthesis of tertiary amines. J. Med. Chem. 45:1963– 1966. Cote, R. J., Shi, Y., Groshen, S., Feng, A. C., Cordon-Cardo, C., Skinner, D. and Lieskovosky, G. 1998. Association of p27Kip1 levels with recurrence and survival in patients with stage C prostate carcinoma. J. Natl. Cancer Inst. 90:916–920. Crick, F. H. C. 1963. The recent excitement in the coding problem. Prog. Nucleic Acid Res. 1:163– 217. Culig, Z., Hoffmann, J., Erdel, M., Eder, I. E., Hobisch, A., Hittmair, A., Bartsch, G., Utermann, G., Schneider, M. R., Parczyk, K. and Klocker, H. 1999. Switch from antagonist to agonist of the androgen receptor bicalutamide is associated with prostate tumour progression in a new model system. Br. J. Cancer 81:242–251. Dominguez-Sola, D., Ying, C. Y., Grandori, C., Ruggiero, L., Chen, B., Li, M., Galloway, D. A., Gu, W., Gautier, J. and Dalla-Favera, R. 2007. Non-transcriptional control of DNA replication by c-Myc. Nature 448:445–451. Efanov, A. M., Sewing, S., Bokvist, K. and Gromada, J. 2004. Liver X receptor activation stimulates insulin secretion via modulation of glucose and lipid metabolism in pancreatic beta-cells. Diabetes 53(Suppl. 3):S75–S78. Epstein, J. I., Carmichael, M. and Partin, A. W. 1995. OA-519 (fatty acid synthase) as an independent predictor of pathologic state in adenocarcinoma of the prostate. Urology 45:81–86. Fang, S. and Liao, S. 1969. Antagonistic actions of anti-androgens in the formation of a specific dihydrotestosterone-receptor protein complex in rat ventral prostate. Mol. Pharmacol. 5:428– 431. Fang, S. and Liao, S. 1971. Androgen receptors: steroid- and tissue-specific retention of a 17bhydroxy-5a-androstan-3-one protein complex by the cell nuclei of ventral prostate. J. Biol. Chem. 246:16–24. Fang, S., Anderson, K. M. and Liao, S. 1969. Receptors for androgens: on the role of specific proteins in selective retention of 17b-hydroxy-5a-androstan-3-one by rat ventral prostate in vivo and in vitro. J. Biol. Chem. 244:6584–6595. Fowler, J. E. and Whitmore, W. F. 1981. The response to metastatic adenocarcinoma of the prostate to exogenous testosterone. J. Urol. 126:372–375. Fu, X., Menke, J. G., Chen, Y., Zhou, G., MacNaul, K. L., Wright, S. D., Sparrow, C. P. and Lund, E. G. 2001. 27-Hydroxycholesterol is an endogenous ligand for liver X receptor in cholesterolloaded cells. J. Biol. Chem. 276:38378–38387. Fukuchi, J., Kokontis, J. M., Hiipakka, R. A., Chuu, C. P. and Liao, S. 2004. Antiproliferative effect of liver X receptor agonists on LNCaP human prostate cancer cells. Cancer Res. 64:7686–7689. Geck, P., Szelei, J., Jimenez, J., Lin, T. M., Sonnenschein, C. and Soto, A. M. 1997. Expression of novel genes linked to the androgen-induced, proliferative shutoff in prostate cancer cells. J. Steroid Biochem. Mol. Biol. 63:211–218. Gottlieb, B., Trifiro, M., Lumbroso, R. and Pinsky, L. 1997. The androgen receptor gene mutations database. Nucleic Acids Res. 25:158–162.

46

S. Liao et al.

Greene, G. L., Closs, L. E., Fleming, H., DeSombre, E. R. and Jensen, E. V. 1977. Antibodies to estrogen receptor: immunochemical similarity of estrophilin from various mammalian species. Proc. Natl. Acad. Sci. USA 74:3681–3685. Guo, Y., Sklar, G. N., Borkowski, A. and Kyprianou, N. 1997. Loss of the cyclin-dependent kinase inhibitor p27(Kip1) protein in human prostate cancer correlates with tumor grade. Clin. Cancer Res. 3:2269–2274. Gupta, S., Hastak, K., Ahmad, N., Lewin, J. S. and Mukhtar, H. 2001. Inhibition of prostate carcinogenesis in TRAMP mice by oral infusion of green tea polyphenols. Proc. Natl. Acad. Sci USA 98:10350–10355. Haag, P., Bektic, J., Bartsch, G., Klocker, H. and Eder, I. E. 2005. Androgen receptor down regulation by small interference RNA induces cell growth inhibition in androgen sensitive as well as in androgen independent prostate cancer cells. J. Steroid Biochem. Mol. Biol. 96:251– 258. Hara, T., Nakamura, K., Araki, H., Kusaka, M. and Yamaoka, M. 2003. Enhanced androgen receptor signaling correlates with the androgen-refractory growth in a newly established MDA PCa 2b-hr human prostate cancer cell subline. Cancer Res. 63:5622–5628. He, W. W., Fischer, L. M., Sun, S., Bilhartz, D. L., Zhu, X. P., Young, C. Y., Kelley D. B. and Tindall, D. J. 1990. Molecular cloning of androgen receptors from divergent species with a polymerase chain reaction technique: complete cDNA sequence of the mouse androgen receptor and isolation of androgen receptor cDNA probes from dog, guinea pig and clawed frog. Biochem. Biophys. Res. Commun. 171:697–704. He, W. W., Kumar, M. V. and Tindall, D. J. 1991. A frame shift mutation in the androgen receptor gene causes complete androgen insensitivity in the testicular feminized mouse. Nucleic Acids Res. 19:2373–2378. Heisler, L. E., Evangelou, A., Lew, A. M., Trachtenberg, J., Elsholtz, H. P. and Brown, T. J. 1997. Androgen-dependent cell cycle arrest and apoptotic death in PC-3 prostatic cell cultures expressing a full-length human androgen receptor. Mol. Cell. Endocrinol. 126:59–73. Hiipakka, R. A. and Liao, S. 1988. Intracellular inhibition of chromatin binding and transformation of androgen receptor by 30 -deoxyadenosine. J. Biol. Chem. 263:17590–17595. Hiipakka, R. A. and Liao, S. 1998. Molecular mechanism of androgen action. Trends Endocrinol. Metab. 9:317–324. Hiipakka, R. A., Chen, C., Schilling, K., Oberhauser, A., Saltzman, A. and Liao, S. 1984. Immunochemical characterization of the androgen-dependent spermine-binding protein of the rat ventral prostate. Biochem. J. 218:563–571. Hiipakka, R. A., Zhang, H. Z., Dai, W., Dai, Q. and Liao, S. 2002. Structure-activity relationships for inhibition of human 5alpha-reductases by polyphenols. Biochem. Pharmacol. 63:1165– 1176. Horoszewicz, J. S., Leong, S. S., Kawinski, E., Karr, J., Rosenthal, H., Chu, T. M., Mirand, E. A. and Murphy, G. P. 1983. LNCaP model of human prostatic carcinoma. Cancer Res. 43:1809– 1818. Imperato-McGinley, J. and Gautier, T. 1992. Inherited 5a-reductase deficiency in man. Trends Genet. 2:130–133. Imperato-McGinley, J., Guerrero, L., Gautier, T. and Peterson, R. E. 1974. Steroid 5alphareductase deficiency in man: an inherited form of male pseudohermaphroditism. Science 186:1213–1215. Iranmanesh, A. and Veldhuis, J. D. 2005. Combined inhibition of types I and II 5 alpha-reductase selectively augments the basal (nonpulsatile) mode of testosterone secretion in young men. J. Clin. Endocrinol. Metab. 90:4232–4237. Janowski, B. A., Willy, P. J., Devi, T. R., Falck, J. R. and Mangelsdorf, D. J. 1996. An oxysterol signalling pathway mediated by the nuclear receptor LXR alpha. Nature 383:728–731. Janowski, B. A., Grogan, M. J., Jones, S. A., Wisely, G. B., Kliewer, S. A., Corey, E. J. and Mangelsdorf, D. J. 1999. Structural requirements of ligands for the oxysterol liver X receptors LXRalpha and LXRbeta. Proc. Natl. Acad. Sci. USA 96:266–271.

Androgen Action and Modulation of Prostate and Prostate Cancer Growth

47

Jensen, E. V. 1978. Interaction of steroid hormones with the nucleus. Pharmacol. Rev. 30:477–491. Jensen, E. V. and Jacobson, H. I. 1962. Basic guide to the mechanism of estrogen action. Rec. Prog. Horm. Res. 18:387–414. Joly-Pharaboz, M.-O., Soave, M.-C., Nicolas, B., Mebarki, F., Renaud, M., Foury, O., Morel, Y. and Andre, J. G. 1995. Androgens inhibit the proliferation of a variant of the human prostate cancer cell line LNCaP. J. Steroid Biochem. Mol. Biol. 55:67–76. Joly-Pharaboz, M., Ruffion, A., Roch, A., Michel-Calemard, L., Andre, J., Chantepie, J., Nicolas, B. and Panaye, G. 2000. Inhibition of growth and induction of apoptosis by androgens of a variant of LNCaP cell line. J. Steroid Biochem. Mol. Biol. 73:237–249. Joseph, S. B., McKilligin, E., Pei, L., Watson, M. A., Collins, A. R., Laffitte, B. A., Chen, M., Noh, G., Goodman, J., Hagger, G. N., Tran, J., Tippin, T. K., Wang, X., Lusis, A. J., Hsueh, W. A., Law, R. E., Collins, J. L., Willson, T. M. and Tontonoz, P. 2002. Synthetic LXR ligand inhibits the development of atherosclerosis in mice. Proc. Natl. Acad. Sci. USA 99:7604–7609. Joseph, S. B., Castrillo, A., Laffitte, B. A., Mangelsdorf, D. J. and Tontonoz, P. 2003. Reciprocal regulation of inflammation and lipid metabolism by liver X receptors. Nat. Med. 9:213–219. Kaneko, E., Matsuda, M., Yamada, Y., Tachibana, Y., Shimomura, I. and Makishima, M. 2003. Induction of intestinal ATP-binding cassette transporters by a phytosterol-derived liver X receptor agonist. J. Biol. Chem. 278:36091–36098. Keller, U. B., Old, J. B., Dorsey, F. C., Nilsson, J. A., Nilsson, L., MacLean, K. H., Chung, L., Yang, C., Spruck, C., Boyd, K., Reed, S. I. and Cleveland, J. L. 2007. Myc targets Cks1 to provoke the suppression of p27Kip1, proliferation and lymphomagenesis. EMBO J. 26:2562–2574. Knudsen, B. S. and Edlund, M. 2004. Prostate cancer and the met hepatocyte growth factor receptor. Adv. Cancer Res. 91:31–67. Knudsen, B. S., Gmyrek, G. A., Inra, J., Scherr, D. S., Vaughan, E. D., Nanus, D. M., Kattan, M. W., Gerald, W. L. and Vande Woude, G. F. 2002. High expression of the Met receptor in prostate cancer metastasis to bone. Urology 60:1113–1117. Kokontis, J. M. and Liao, S. 1999. Molecular action of androgen in the normal and neoplastic prostate. In Vitamins and Hormones, ed. Litwack, G. pp. 219–308. New York: Academic Press. Kokontis, J., Ito, K., Hiipakka, R. A. and Liao, S. 1991. Expression and function of normal and LNCaP androgen receptors in androgen-insensitive human prostatic cancer cells: altered hormone and antihormone specificity in gene transactivation. Receptor 1:271–279. Kokontis, J., Takakura, K., Hay, N. and Liao, S. 1994. Increased androgen receptor activity and altered c-myc expression in prostate cancer cells after long-term androgen deprivation. Cancer Res. 54:1566–1573. Kokontis, J. M., Hay, N. and Liao, S. 1998. Progression of LNCaP prostate tumor cells during androgen deprivation: hormone-independent growth, repression of proliferation by androgen and role for p27Kip1 in androgen-induced cell cycle arrest. Mol. Endocrinol. 12:941–953. Kokontis, J. M., Hsu, S., Chuu, C. P., Dang, M., Fukuchi, J., Hiipakka, R. A. and Liao, S. 2005. Role of androgen receptor in the progression of human prostate tumor cells to androgen independence and insensitivity. Prostate 65:287–298. Kuriyama, S., Shimazu, T., Ohmori, K., Kikuchi, N., Nakaya, N., Nishino, Y., Tsubono, Y. and Tsuji, I. 2006. Green tea consumption and mortality due to cardiovascular disease, cancer, and all causes in Japan: the Ohsaki study. JAMA 296:1255–1265. Laffitte, B. A., Chao, L. C., Li, J., Walczak, R., Hummasti, S., Joseph, S. B., Castrillo, A., Wilpitz, D. C., Mangelsdorf, D. J., Collins, J. L., Saez, E. and Tontonoz, P. 2003. Activation of liver X receptor improves glucose tolerance through coordinate regulation of glucose metabolism in liver and adipose tissue. Proc. Natl. Acad. Sci. USA 100:5419–5424. Langley, E., Zhou, Z. X. and Wilson, E. M. 1995. Evidence for an anti-parallel orientation of the ligand-activated human androgen receptor dimer. J. Biol. Chem. 270:29983–29990.

48

S. Liao et al.

La Spada, A. R., Wilson, E. M., Lubahn, D. B., Harding, A. E. and Fischbeck, K. H. 1991. Androgen receptor gene mutations in X-linked spinal and bulbar muscular atrophy. Nature 352:77–79. Le Beau, M. M., Song, C., Davis, E. M., Hiipakka, R. A., Kokontis, J. M. and Liao, S. 1995. Assignment of the human Ubiquitous Receptor (UNR) to 19q13.3 using fluorescence in situ hybridization. Genomics 26:166–168. Lehmann, J. M., Kliewer, S. A., Moore, L. B., Smith-Oliver, T. A., Oliver, B. B., Su, J. L., Sundseth, S. S., Winegar, D. A., Blanchard, D. E., Spencer, T. A. and Willson, T. M. 1997. Activation of the nuclear receptor LXR by oxysterols defines a new hormone response pathway. J. Biol. Chem. 272:3137–3140. Lengyel, P., Speyer, J. F., Basilio, C. and Ochoa, S. 1962. Synthetic polynucleotides and the amino acid code. III. Proc. Natl. Acad. Sci. USA 48:282–284. Leong, H., Mathur, P. S. and Greene, G. L. 2008. Inhibition of mammary tumorigenesis in the C3(1)/SV40 mouse model by green tea. Breast Cancer Res. Treat. 107:359–369. Liang, T. and Liao, S. 1974. Association of the uterine 17beta-estradiol-receptor complex with ribonucleoprotein in vitro and in vivo. J. Biol. Chem. 249:4671–4678. Liang, T. and Liao, S. 1997. Growth suppression of hamster flank organs by topical application of g-linolenic and other fatty acid inhibitors of 5a-reductase. J. Invest. Dermatol. 109:152–157. Liang, T., Mezzetti, G., Chen, C., Liao, S. 1978. Selective polyamine-binding proteins: spermine binding by an androgen-sensitive phosphoprotein. Biochim. Biophys. Acta 542:430–441. Liang, T., Brady, E. J., Cheung, A. and Saperstein, R. 1984. Inhibition of luteinizing hormone (LH)-releasing hormone-induced secretion of LH in rat anterior pituitary cell culture by testosterone without conversion to 5 alpha-dihydrotestosterone. Endocrinology 115: 2311–2317. Liang, G., Yang, J., Horton, J. D., Hammer, R. E., Goldstein, J. L. and Brown, M. S. 2002. Diminished hepatic response to fasting/refeeding and liver X receptor agonists in mice with selective deficiency of sterol regulatory element-binding protein-1c. J. Biol. Chem. 277:9520–9528. Liao, S. 1965. Influence of testosterone on template activity of prostatic ribonucleic acids. J. Biol. Chem. 240:1236–1243. Liao, S. 1975. Cellular receptors and mechanisms of action of steroid hormones. Int. Rev. Cytol. 41:87–172. Liao, S. 1994. Androgen action: molecular mechanism and medical application. J. Formos. Med. Assoc. 93:741–51. Liao, S. 1997. Obituary. J Am Med Assoc. 278:1545. Liao, S. 2002. Charles Brenton Huggins. American National Biography Online. http://www.anb. org/articles/12/12-02114.html?a=1&n=Charles%20Brenton%20Huggins. Liao, S. and Fang, S. 1969. Receptor-proteins for androgens and the mode of action of androgens on gene transcription in ventral prostate. Vitam. Horm. 27:17–90. Liao, S. and Williams-Ashman, H. G. 1962. An effect of testosterone on amino acid incorporation by prostatic ribonucleoprotein particles. Proc. Natl. Acad. Sci. USA 48:1956–1964. Liao, S. and Witte, D. 1985. Autoimmune anti-androgen receptor antibodies in human serum. Proc. Natl. Acad. Sci. USA 82:8345–8348. Liao, S., Leininger, K. R., Sagher, D. and Barton, R. W. 1965. Rapid effect of testosterone on ribonucleic acid polymerase activity of rat ventral prostate. Endocrinology 77:763–765. Liao, S., Sagher, D. and Fang, S. M. 1968. Isolation of chromatin-free RNA polymerase from mammalian cell nuclei. Nature 220:1336–1337. Liao, S., Sagher, D., Lin, A. H. and Fang, S. 1969. Magnesium and manganese specific forms of soluble liver RNA polymerase. Nature 223:297–298. Liao, S., Liang, T. and Tymoczko, J. L. 1972. Structural recognitions in the interactions of androgens and receptor proteins and in their association with nuclear acceptor components. J. Steroid Biochem. 3:401–408.

Androgen Action and Modulation of Prostate and Prostate Cancer Growth

49

Liao, S., Liang, T., Fang, S., Castaneda, E. and Shao, T. C. 1973a. Steroid structure and androgenic activity: specificities involved in the receptor binding and nuclear retention of various androgens. J. Biol. Chem. 248:6154–6162. Liao, S., Liang, T. and Tymoczko, J. 1973b. Ribonucleoprotein binding of steroid-‘‘receptor’’ complexes. Nat. New Biol. 241:211–213. Liao, S., Howell, D. K. and Chang, T.-M. 1974. Action of a nonsteroidal antiandrogen, flutamide, on receptor binding and nuclear retention of 5a-dihydrotestosterone in rat ventral prostate. Endocrinology 94:1205–1209. Liao, S., Smythe, S., Tymoczko, J., Rossini, G. P., Chen, C. and Hiipakka, R. A. 1980. RNAdependent release of androgen and other steroid receptor complexes from DNA. J. Biol. Chem. 245:5545–5551. Liao, S., Chang, C. and Saltzman, A. G. 1983. Androgen-receptor interaction – an overview. In Steroid Hormone Receptors: Structure and Function, Nobel Symposium, Sweden, eds. Eriksson, H. and Gustafsson, J. A. pp. 123–135. New York: Elsevier Liao, S., Witte, D., Schilling, K. and Chang, C. 1984. The use of a hydroxylapatite-filter steroid receptor assay method in the study of the modulation of androgen receptor interaction. J. Steroid Biochem. 20:11–17. Liao, S., Kokontis, J., Sai, T. and Hiipakka, R. A. 1989. Androgen receptors: structures, mutations, antibodies and cellular dynamics. J. Steroid Biochem. 34:41–51. Liao, S., Umekita, Y., Guo, J., Kokontis, J. M. and Hiipakka, R. A. 1995. Growth inhibition and regression of human prostate and breast tumors in athymic mice by tea epigallocatechin gallate. Cancer Lett. 96:239–243. Liao, S., Kao, Y. H. and Hiipakka, R. A. 2001a. Green tea: biochemical and biological basis for health benefits. Vitam. Horm. 62:1–94. Liao, S., Lin, J., Dang, M. T., Zhang, H., Kao, Y. H., Fukuchi, J. and Hiipakka, R. A. 2001b. Growth suppression of hamster flank organs by topical application of catechins, alizarin, curcumin, and myristoleic acid. Arch. Dermatol. Res. 293:200–205. Liao, S., Kokontis, J. M., Chuu, C. P., Hsu, S., Fukuchi, J., Dang, M. T. and Hiipakka, R. A. 2005. Four stages of prostate cancer: suppression and eradication by androgen and green tea epigallocatechin gallate. In Hormonal Carcinogenesis IV, eds. Li, J. J. and Li, S. A. pp. 211– 220. New York: Springer. Lin, Y., Kokontis, J., Tang, F., Godfrey, B., Liao, S., Lin, A., Chen, Y. and Xiang, J. 2006. Androgen and its receptor promote Bax-mediated apoptosis. Mol. Cell. Biol. 26:1908–1916. Ling, M. T., Chan, K. W. and Choo, C. K. 2001. Androgen induces differentiation of a human papillomavirus 16 E6/E7 immortalized prostate epithelial cell line. J. Endocrinol. 170:287– 296. Litvinov, I. V., Antony, L. and Isaacs, J. T. 2004. Molecular characterization of an improved vector for evaluation of the tumor suppressor versus oncogene abilities of the androgen receptor. Prostate 61:299–304. Litvinov, I. V., Vander Griend, D. J., Antony, L., Dalrymple, S., De Marzo, A. M., Drake, C. G. and Isaacs, J. T. 2006. Androgen receptor as a licensing factor for DNA replication in androgen-sensitive prostate cancer cells. Proc. Natl. Acad. Sci. USA 103:15085–15090. Lu, L., Schulz, H. and Wolf, D. A. 2002. The F-box protein SKP2 mediates androgen control of p27 stability in LNCaP human prostate cancer cells. BMC Cell Biol. 3:22. Lubahn, D. B., Joseph, D. R., Sullivan, P. M., Willard, H. F., French, F. S. and Wilson, E. M. 1988. Cloning of human androgen receptor complementary DNA and localization to the X chromosome. Science 240:327–330. Lucia, M. S., Epstein, J. I., Goodman, P. J., Darke, A. K., Reuter, V. E., Civantos, F., Tangen, C. M., Parnes, H. L., Lippman, S. M., La Rosa, F. G., Kattan, M. W., Crawford, E. D., Ford, L. G., Coltman, C. A., Jr. and Thompson, I. M. 2007. Finasteride and high-grade prostate cancer in the Prostate Cancer Prevention Trial. J. Natl. Cancer Inst. 99:1375–1383. Mahendroo, M. S., Cala, K. M., Landrum, D. P. and Russell, D. W. 1997. Fetal death in mice lacking 5alpha-reductase type 1 caused by estrogen excess. Mol. Endocrinol. 11:917–927.

50

S. Liao et al.

Marcelli, M., Tilley, W. D., Wilson, C. M., Wilson, J. D., Griffin, J. E. and McPhaul, M. J. 1990. A single nucleotide substitution introduces a premature termination codon into the androgen receptor gene of a patient with receptor-negative androgen resistance. J. Clin. Inves. 85:1522–1528. Mathew, P. 2008. Prolonged control of progressive castration-resistant metastatic prostate cancer with testosterone replacement therapy: the case for a prospective trial. Ann. Oncol. 19:395–396. Matias, P. M., Donner, P., Coelho, R., Thomaz, M., Peixoto, C., Macedo, S., Otto, N., Joschko, S., Scholz, P., Wegg, A., Basler, S., Schafer, M., Egner, U. and Carrondo, M. A. 2000. Structural evidence for ligand specificity in the binding domain of the human androgen receptor. Implications for pathogenic gene mutations. J. Biol. Chem. 275:26164–26171. McDonald, S., Brive, L., Agus, D. B., Scher, H. I. and Ely, K. R. 2000. Ligand responsiveness in human prostate cancer: structural analysis of mutant androgen receptors from LNCaP and CWR22 tumors. Cancer Res. 60:2317–2322. McPhaul, M. J. and Marcelli, M. 1992. Molecular defects in the androgen receptor causing androgen resistance. J. Invest. Dermatol. 98:97S–99S. Mezzetti, G., Loor, R. and Liao, S. 1979. Androgen-sensitive spermine-binding protein of rat ventral prostate. Purification of the protein and characterization of the hormonal effect. Biochem. J. 184:431–440. Minesita, T. and Yamaguchi, K. 1965. An androgen-dependent mouse mammary tumor. Cancer Res. 25:1168–1175. Mitro, N., Mak, P. A., Vargas, L., Godio, C., Hampton, E., Molteni, V., Kreusch, A. and Saez, E. 2006. The nuclear receptor LXR is a glucose sensor. Nature 445:219–223. Monks, D. A., Johansen, J. A., Mo, K., Rao, P., Eagleson, B., Yu, Z., Lieberman, A. P., Breedlove, S. M. and Jordan, C. L. 2007. Overexpression of wild-type androgen receptor in muscle recapitulates polyglutamine disease. Proc. Natl. Acad. Sci. USA 104:18259–18264. Morris, M. J., Kelly, W. K., Slovin, S., Sauter, N., Eicher, C., Regan, T., Curley, A., Delacruz, A., Reuter, V. and Scher, H. I. 2004. Phase I trial of exogenous testosterone (T) for the treatment of castrate metastatic prostate cancer (PC). Amer. Soc. Clin. Oncol. Meeting Abstracts 22:4560. Nagamoto, A., Noguchi, K., Murai, T. and Kinoshita, Y. 1994. Significant role of 5 alphareductase on feedback effects of androgen in rat anterior pituitary cells demonstrated with a nonsteroidal 5 alpha-reductase inhibitor ONO-3805. J. Androl. 15:521–527. Neri, R., Peets, E. and Watnick, A. 1979. Anti-androgenicity of flutamide and its metabolite Sch 16423. Biochem. Soc. Trans. 7:565–569. Nirenberg, M. W. and Matthaei, J. H. 1961. The dependence of cell-free protein synthesis in E. coli upon naturally occurring or synthetic polyribonucleotides. Proc. Natl. Acad. Sci. USA 47:1588–1602. Nirenberg, M. W., Matthaei, J. H. and Jones, O. W. 1962. An intermediate in the biosynthesis of polyphenylalanine directed by synthetic template RNA. Proc. Natl. Acad. Sci. USA 48:104– 109. Normington, K. and Russell, D. W. 1992. Tissue distribution and kinetic characteristics of rat steroid 5 alpha-reductase isozymes. Evidence for distinct physiological functions. J. Biol. Chem. 267:19548–19554. Peet, D. J., Turley, S. D., Ma, W., Janowski, B. A., Lobaccaro, J. M., Hammer, R. E. and Mangelsdorf, D. J. 1998. Cholesterol and bile acid metabolism are impaired in mice lacking the nuclear oxysterol receptor LXR alpha. Cell 93:693–704. Peng, D., Song, C., Reardon, C. A., Liao, S. and Getz, G. S. 2003. Lipoproteins produced by ApoE-/- astrocytes infected with adenovirus expressing human ApoE. J. Neurochem. 86:1391–1402. Peng, D., Hiipakka, R. A., Dai, Q., Guo, J., Reardon, C. A., Getz, G. S. and Liao, S. 2008. Antiatherosclerotic effects of a novel synthetic tissue-selective steroidal liver x receptor agonist in low-density lipoprotein receptor-deficient mice. J. Pharmacol. Exp. Therap. 327:332–342.

Androgen Action and Modulation of Prostate and Prostate Cancer Growth

51

Pitteloud, N., Dwyer, A. A., Decruz, S., Lee, H., Boepple, P. A., Crowley, W. F., Jr. and Hayes, F. J. 2008. Inhibition of LH secretion by testosterone in men requires aromatization for its pituitary but not its hypothalamic effects: evidence from the tandem study of normal and gonadotropin-releasing hormone-deficient men. J. Clin. Endocrinol. Metab. 93:784–791. Plat, J., Nichols, J. A. and Mensink, R. P. 2005. Plant sterols and stanols: effects on mixed micellar composition and LXR (target gene) activation. J. Lipid Res. 46:2468–2476. Pratt, W. B. and Toft, D. O. 1997. Steroid receptor interactions with heat shock protein and immunophilin chaperones. Endocr. Rev. 18:306–360. Prehn, R. T. 1999. On the prevention and therapy of prostate cancer by androgen administration. Cancer Res. 59:4161–4164. Repa, J. J., Liang, G., Ou, J., Bashmakov, Y., Lobaccaro, J. M., Shimomura, I., Shan, B., Brown, M. S., Goldstein, J. L. and Mangelsdorf, D. J. 2000. Regulation of mouse sterol regulatory element-binding protein-1c gene (SREBP-1c) by oxysterol receptors, LXRalpha and LXRbeta. Genes Dev. 14:2819–2830. Ris-Stalpers, C., Kuiper, G. G., Faber, P. W., Schweikert, H. U., van Rooij, H. C., Zegers, N. D., Hodgins, M. B., Degenhart, H. J., Trapman, J. and Brinkmann, A. O. 1990. Aberrant splicing of androgen receptor mRNA results in synthesis of a nonfunctional receptor protein in a patient with androgen insensitivity. Proc. Natl. Acad. Sci. USA 87:7866–7870. Ris-Stalpers, C., Trifiro, M. A., Kuiper, G. G. J. M., Jenster, G., Romalo, G., Sai, T., van Rooij, H. C. J., Kaufman, M., Rosenfeld, R. L., Liao, S., Schweikert, H.-U., Trapman, J., Pinsky, L. and Brinkmann, A. O. 1991. Substitution of aspartic acid-686 by histidine or asparagine in the human androgen receptor leads to a functionally inactive protein with altered hormone-binding characteristics. Mol. Endocrinol. 5:1562–1569. Rossini, G. P. and Liao, S. 1982. Intracellular inactivation, reactivation and dynamic status of prostate androgen receptors. Biochem. J. 208:383–392. Russell, D. W. and Wilson, J. D. 1994. Steroid 5a-reductase: two genes/two enzymes. Annu. Rev. Biochem. 63:25–61. Sack, J. S., Kish, K. F., Wang, C., Attar, R. M., Kiefer, S. E., An, Y., Wu, G. Y., Scheffler, J. E., Salvati, M. E., Krystek, S. R., Jr., Weinmann, R. and Einspahr, H. M. 2001. Crystallographic structures of the ligand-binding domains of the androgen receptor and its T877A mutant complexed with the natural agonist dihydrotestosterone. Proc. Natl. Acad. Sci. USA 98:4904–4909. Sai, T., Seino, S., Chang, C., Trifiro, M., Pinsky, L., Mhatre, A., Kaufman, M., Lambert, B., Trapman, J., Brinkmann, A. O., Rosenfield, R. L. and Liao, S. 1990. An exonic point mutation of the androgen receptor gene in a family with complete androgen insensitivity. Am. J. Hum. Genet. 46:1095–1100. Saltzman, A. G., Hiipakka, R. A., Chang, C. and Liao, S. 1987. Androgen repression of the production of a 29-kilodalton protein and its mRNA in the rat ventral prostate. J. Biol. Chem. 262:432–437. Sar, M., Liao, S. and Stumpf, W. E. 1970. Nuclear concentration of androgens in rat seminal vesicles and prostate demonstrated by dry-mount autoradiography. Endocrinology 86:1008– 1011. Schanbacher, B. D., Johnson, M. P. and Tindall, D. J. 1987. Androgenic regulation of luteinizing hormone secretion: relationship to androgen binding in sheep pituitary. Biol. Reprod. 36:340–350. Schilling, K. and Liao, S. 1984. The use of radioactive 7a, 17a-dimethyl-19-nortestosterone (mibolerone) in the assay of androgen receptors. Prostate 5:581–588. Schreihofer, D. A., Amico, J. A. and Cameron, J. L. 1993. Reversal of fasting-induced suppression of luteinizing hormone (LH) secretion in male rhesus monkeys by intragastric nutrient infusion: evidence for rapid stimulation of LH by nutritional signals. Endocrinology 132:1890–1897.

52

S. Liao et al.

Schultz, J. R., Tu, H., Luk, A., Repa, J. J., Medina, J. C., Li, L., Schwendner, S., Wang, S., Thoolen, M., Mangelsdorf, D. J., Lustig, K. D. and Shan, B. 2000. Role of LXRs in control of lipogenesis. Genes Dev. 14:2831–2838. Seol, W., Choi, H.-S. and Moore, D. D. 1995. Isolation of proteins that interact specifically with retinoid X receptor: two novel orphan receptors. Mol. Endocrinol. 9:72–85. Shao, J. and Diamond, M. I. 2007. Polyglutamine diseases: emerging concepts in pathogenesis and therapy. Hum. Mol. Genet. 16(Spec No. 2):R115–R123. Shao, T.-C., Castaneda, E., Rosenfield, R. L. and Liao, S. 1975. Selective retention and formation of a D5-androstenediol-receptor complex in cell nuclei of the rat vagina. J. Biol. Chem. 250:3095–3100. Shi, X. B., Ma, A. H., Tepper, C. G., Xia, L., Gregg, J. P., Gandour-Edwards, R., Mack, P. C., Kung, H. J. and DeVere White, R. W. 2004. Molecular alterations associated with LNCaP cell progression to androgen independence. Prostate 60:257–271. Shinar, D. M., Endo, N., Rutledge, S. J., Vogel, R., Rodan, G. A. and Schmidt, A. 1994. NER, a new member of the gene family encoding the human steroid hormone nuclear receptor. Gene 147:273–276. Shurbaji, M. S., Kalbfleisch, J. H. and Thurmond, T. S. 1996. Immunohistochemical detection of a fatty acid synthase (OA-519) as a predictor of progression of prostate cancer. Hum. Pathol. 27:917–921. Shyr, C. and Liao, S. 1978. A protein factor that inhibits binding and promotes the release of the androgen-receptor complex from nuclear chromatin. Proc. Natl. Acad. Sci. USA 75:5969– 5973. Siiteri, P. K. and Wilson, J. D. 1974. Testosterone formation and metabolism during male sexual differentiation in the human embryo. J. Clin. Endocrinol. Metab. 38:113–125. Song, C. and Liao, S. 2000. Cholestenoic acid is a naturally occurring ligand for liver X receptor a. Endocrinology 141:4180–4184. Song, C. and Liao, S. 2001. Hypolipidemic effects of selective liver X receptor alpha agonists. Steroids 66:673–681. Song, C., Kokontis, J. M., Hiipakka, R. A. and Liao, S. 1994. Ubiquitous receptor: a receptor that modulates gene activation by retinoic acid and thyroid hormone receptors. Proc. Natl. Acad. Sci. USA 91:10809–10813. Song, C., Hiipakka, R. A., Kokontis, J. M. and Liao, S. 1995. Ubiquitous receptor: structures, immunocytochemical localization, and modulation of gene activation by receptors for retinoic acids and thyroid hormones. Ann. NY Acad. Sci. 761:38–49. Song, C., Hiipakka, R. A. and Liao, S. 2000. Selective activation of liver X receptor alpha by 6alpha-hydroxy bile acids and analogs. Steroids 65:423–427. Song, C., Hiipakka, R. A. and Liao, S. 2001. Autooxidized cholesterol sulphates are antagonistic ligands of liver X receptors: implications for the development and treatment of atherosclerosis. Steroids 66:473–479. Soto, A. M., Lin, T.-M., Sakabe, K., Olea, N., Damassa, D. A. and Sonnenschein, C. 1995. Variants of the human prostate LNCaP carcinoma cell line as tools to study discrete components of the androgen-mediated proliferative response. Oncol. Res. 7:545–558. Spencer, T. A., Li, D., Russel, J. S., Collins, J. L., Bledsoe, R. K., Consler, T. G., Moore, L. B., Galardi, C. M., McKee, D. D., Moore, J. T., Watson, M. A., Parks, D. J., Lambert, M. H. and Willson, T. M. 2001. Pharmacophore analysis of the nuclear oxysterol receptor LXRalpha. J. Med. Chem. 44:886–897. Sporer, A., Brill, D. R. and Schaffner, C. P. 1982. Epoxycholesterols in secretions and tissues of normal, benign, and cancerous human prostate glands. Urology 20:244–250. Swinnen, J. V., Roskams, T., Joniau, S., Van Poppel, H., Oyen, R., Baert, L., Heyns, W. and Verhoeven, G. 2002. Overexpression of fatty acid synthase is an early and common event in the development of prostate cancer. Int. J. Cancer 98:19–22. Talalay, P. 1997. Charles Brenton Huggins (1901–1997). Cancer Res. 57:cover.

Androgen Action and Modulation of Prostate and Prostate Cancer Growth

53

Talalay, P. and Williams-Ashman, H. G. 1960. Participation of steroid hormones in the enzymatic transfer of hydrogen. Recent Prog. Horm. Res. 16:1–47. Tan, J., Sharief, Y., Hamil, K. G., Gregory, C. W., Zang, D. Y., Sar, M., Gumerlock, P. H., deVere White, R. W., Pretlow, T. G., Harris, S. E., Wilson, E. M., Mohler, J. L. and French, F. S. 1997. Dehydroepiandrosterone activates mutant androgen receptors expressed in the androgendependent human prostate cancer xenograft CWR22 and LNCaP cells. Mol. Endocrinol. 11:450–459. Teboul, M., Enmark, E., Li, Q., Wikstrom, A. C., Pelto-Huikko, M. and Gustafsson, J. A. 1995. OR-1, a member of the nuclear receptor superfamily that interacts with the 9-cis-retinoic acid receptor. Proc. Natl. Acad. Sci. USA 92:2096–2100. Thigpen, A. E., Silver, R. I., Guileyardo, J. M., Casey, M. L., McConnell, J. D. and Russell, D. W. 1993. Tissue distribution and ontogeny of steroid 5 alpha-reductase isozyme expression. J. Clin. Invest. 92:903–910. Thompson, I. M., Goodman, P. J., Tangen, C. M., Lucia, M. S., Miller, G. J., Ford, L. G., Lieber, M. M., Cespedes, R. D., Atkins, J. N., Lippman, S. M., Carlin, S. M., Ryan, A., Szczepanek, C. M., Crowley, J. J. and Coltman, C. A., Jr. 2003. The influence of finasteride on the development of prostate cancer. N. Engl. J. Med. 349:215–224. Tian, W. X. 2006. Inhibition of fatty acid synthase by polyphenols. Curr. Med. Chem. 13:967–977. Tilley, W. D., Marcelli, M., Wilson, J. D. and McPhaul, M. J. 1989. Characterization and expression of a cDNA encoding the human androgen receptor. Proc. Natl. Acad. Sci. USA 86:327–331. Toft, D. and Gorski, J. 1966. A receptor molecule for estrogens: isolation from the rat uterus and preliminary characterization. Proc. Natl. Acad. Sci. USA 55:1574–1581. Tomkins, G. M., Yielding, K. L., Curran, J. F., Summers, M. R. and Bitensky, M. W. 1965. The dependence of the substrate specificity on the conformation of crystalline glutamate dehydrogenase. J. Biol. Chem. 240:3793–3798. Trifiro, M., Gotlieb, B., Pinsky, L., Kaufman, M., Prior, L., Belsham, D. D., Wrogemann, K., Brown, C. J., Wilard, H. F., Trapman, J., Brinkman, A. O., Chang, C., Liao, S., Sergovich, F. and Jung, J. 1991. The 56/58 kDa androgen-binding protein in male genital skin fibroblasts with a deleted androgen receptor gene. Mol. Cell. Endocrinol. 75:37–47. Tsvetkov, L. M., Yeh, K. H., Lee, S. J., Sun, H. and Zhang, H. 1999. p27(Kip1) ubiquitination and degradation is regulated by the SCF(Skp2) complex through phosphorylated Thr187 in p27. Curr. Biol. 9:661–664. Umekita, Y., Hiipakka, R. A., Kokontis, J. M. and Liao, S. 1996. Human prostate tumor growth in athymic mice: inhibition by androgens and stimulation by a 5a-reductase inhibitor. Proc. Natl. Acad. Sci. USA 93:11802–11807. Vander Griend, D. J., Litvinov, I. V. and Isaacs, J. T. 2007. Stabilizing androgen receptor in mitosis inhibits prostate cancer proliferation. Cell Cycle 6:647–651. Veldscholte, J., Ris-Stalpers, C., Kuiper, G. G. J. M., Jenster, G., Berrevoets, C., Claassen, E., van Rooij, H. C. J., Trapman, J., Brinkmann, A. O. and Mulder, E. 1990. A mutation in the ligand binding domain of the androgen receptor of human LNCaP cells affects steroid binding characteristics and response to anti-androgens. Biochem. Biophys. Res. Commun. 173:534–540. Veldscholte, J., Berrevoets, C. A., Brinkmann, A. O., Grootegoed, J. A. and Mulder, E. 1992a. Anti-androgens and the mutated androgen receptor of the LNCaP cells: differential effects on binding affinity, heat-shock protein interaction, and transcription activation. Biochemistry 31:2393–2399. Veldscholte, J., Berrevoets, C. A., Ris-Stalpers, C., Kuiper, G. G. J. M., Jenster, G., Trapman, J., Brinkmann, A. O. and Mulder, E. 1992b. The androgen receptor in LNCaP cells contains a mutation in the ligand binding domain which affects steroid binding characteristics and response to antiandrogens. J. Steroid Biochem. Mol. Biol. 41:665–669.

54

S. Liao et al.

Verras, M., Lee, J., Xue, H., Li, T. H., Wang, Y. and Sun, Z. 2007. The androgen receptor negatively regulates the expression of c-Met: implications for a novel mechanism of prostate cancer progression. Cancer Res. 67:967–975. Vis, A. N., Noordzij, M. A., Fitoz, K., Wildhagen, M. F., Schroder, F. H. and van Der Kwast, T. 2000. Prognostic value of cell cycle proteins p27kip1 and Mib-1, and the cell adhesion protein CD44s in surgically treated patients with prostate cancer. J. Urol. 164:2156–2161. Waltregny, D., Leav, I., Signoretti, S., Soung, P., Lin, D., Merk, F., Adams, J. Y., Bhattacharya, N., Cirenei, N. and Loda, M. 2001. Androgen-driven prostate epithelial cell proliferation and differentiation in vivo involve the regulation of p27. Mol. Endocrinol. 15:765–782. Wang, L., Schuster, G. U., Hultenby, K., Zhang, Q., Andersson, S. and Gustafsson, J. A. 2002. Liver X receptors in the central nervous system: from lipid homeostasis to neuronal degeneration. Proc. Natl. Acad. Sci. USA 99:13878–13883. Weiss, S. B. and Nakamoto, T. 1961a. On the participation of DNA in RNA biosynthesis. Proc. Natl. Acad. Sci. USA 47:694–697. Weiss, S. B. and Nakamoto, T. 1961b. The enzymatic synthesis of RNA: Nearest-neighbor base frequencies. Proc. Natl. Acad. Sci. USA 47:1400–1405. West, N. B., Chang, C. S., Liao, S. and Brenner, R. M. 1990. Localization and regulation of estrogen, progestin and androgen receptors in the seminal vesicle of the rhesus monkey. J. Steroid Biochem. Mol. Biol. 37:11–21. Whitacre, D. C., Chauhan, S., Davis, T., Gordon, D., Cress, A. E. and Miesfeld, R. L. 2002. Androgen induction of in vitro prostate cell differentiation. Cell Growth Differ. 13:1–11. Willy, P. J., Umesono, K., Ong, E. S., Evans, R. M., Heyman, R. A. and Mangelsdorf, D. J. 1995. LXR, a nuclear receptor that defines a distinct retinoid response pathway. Genes Dev. 9:1033–1045. Wolf, G. 2006. The regulation of food intake by hypothalamic malonyl-coenzyme A: the MaloA hypothesis. Nutr. Rev. 64:379–383. Yamamoto, Y., Kawamoto, T. and Negishi, M. 2003. The role of the nuclear receptor CAR as a coordinate regulator of hepatic gene expression in defense against chemical toxicity. Arch. Biochem. Biophys. 409:207–211. Yang, R. M., Naitoh, J., Murphy, M., Wang, H. J., Phillipson, J., deKernion, J. B., Loda, M. and Reiter, R. E. 1998. Low p27 expression predicts poor disease-free survival in patients with prostate cancer. J. Urol. 159:941–945. Yoshikawa, T., Shimano, H., Amemiya-Kudo, M., Yahagi, N., Hasty, A. H., Matsuzaka, T., Okazaki, H., Tamura, Y., Iizuka, Y., Ohashi, K., Osuga, J., Harada, K., Gotoda, T., Kimura, S., Ishibashi, S. and Yamada, N. 2001. Identification of liver X receptor-retinoid X receptor as an activator of the sterol regulatory element-binding protein 1c gene promoter. Mol. Cell. Biol. 21:2991–3000. Young, C. Y., Murthy, L. R., Prescott, J. L., Johnson, M. P., Rowley, D. R., Cunningham, G.R. Killian, C. S., Scardino, P. T., VonEschenbach, A. Tindall D. J. and Tindall D. J. 1988. Monoclonal antibodies against the androgen receptor: recognition of human and other mammalian androgen receptors. Endocrinology 123:601–610. Yu, Z., Dadgar, N., Albertelli, M., Gruis, K., Jordan, C., Robins, D. M. and Lieberman, A. P. 2006. Androgen-dependent pathology demonstrates myopathic contribution to the Kennedy disease phenotype in a mouse knock-in model. J. Clin. Invest. 116:2663–2672. Yuan, S., Trachtenberg, J., Mills, G. B., Brown, T. J., Xu, F. and Keating, A. 1993. Androgeninduced inhibition of cell proliferation in an androgen-insensitive prostate cancer cell line (PC-3) transfected with a human androgen receptor complementary cDNA. Cancer Res. 53:1304–1311. Yuan, X., Li, T., Wang, H., Zhang, T., Barua, M., Borgesi, R. A., Bubley, G. J., Lu, M. L. and Balk, S. P. 2006. Androgen receptor remains critical for cell-cycle progression in androgenindependent CWR22 prostate cancer cells. Am. J. Pathol. 169:682–696.

Androgen Action and Modulation of Prostate and Prostate Cancer Growth

55

Zegarra-Moro, O. L., Schmidt, L. J., Huang, H. and Tindall, D. J. 2002. Disruption of androgen receptor function inhibits proliferation of androgen-refractory prostate cancer cells. Cancer Res. 62:1008–1013. Zelcer, N. and Tontonoz, P. 2006. Liver X receptors as integrators of metabolic and inflammatory signaling. J. Clin. Invest. 116:607–614. Zhau, H. Y., Chang, S. M., Chen, B. Q., Wang, Y., Zhang, H., Kao, C., Sang, Q. A., Pathak, S. J. and Chung, L. W. 1996. Androgen-repressed phenotype in human prostate cancer. Proc. Natl. Acad. Sci. USA 93:15152–15157.

Clinical Progression to Castration-Recurrent Prostate Cancer Mark Pomerantz and Philip Kantoff

1 Introduction Since Huggins and Hodges proved unequivocally that prostate cancer (CaP) regresses in the castrate state, androgen deprivation therapy (ADT) has been the cornerstone of treatment for advanced CaP (Huggins and Hodges 1941). This treatment strategy has taken many forms. Bilateral orchiectomy reliably diminishes androgen production and was the mainstay of treatment in the 1940s. In 1941, Huggins and Hodges published their seminal work showing that therapy with estrogen, in the form of diethylstilbestrol (DES), was as effective as orchiectomy (Huggins and Hodges 1941). Chemical castration via DES was widely used until the mid-1960s, when the estrogen formulation was directly compared to other strategies, including bilateral orchiectomy. DES was associated with an improved cancer-related mortality but worse overall survival, due primarily to increased cardiovascular side effects (The Veterans Administration Co-operative Urological Research Group 1967). Despite attempts to find a less toxic but effective dose, DES fell out of favor as an alternative to orchiectomy. In the 1980s, long-acting, synthetic luteinizing hormone-releasing hormone (LHRH) agonists were developed (Tolis et al. 1982). These agents have emerged as the predominant method for achieving castrate levels of androgen. ADT is highly effective, capable of inducing regression of disease in over 90% of CaP patients (Prostate Cancer Trialists’ Collaborative Group 2000; The Medical Research Council Prostate Cancer Working Party Investigators Group 1997). However, ADT is rarely curative. While the time in remission can vary markedly between patients, CaP invariably becomes refractory to surgical or chemical castration. This chapter will examine clinical progression from androgen dependence to androgen independence, focusing on the clinical factors associated with the

M. Pomerantz(*) Lank Center for Genitourinary Oncology, Dana-Farber Cancer Institute, Dana Bldg, Room 1230, Boston, MA 02115, USA, E-mail: [email protected]

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4_2, # Springer Science + Business Media, LLC 2009

57

58

M. Pomerantz, P. Kantoff

emergence of resistance, and the natural history of and treatment strategies for castration-resistant prostate cancer (CRPC).

2 Definition of Androgen-Independent Prostate Cancer Castration recurrent prostate cancer (CRPC) denotes progression of disease despite chemical or surgical castration. Various terms have been used to describe this disease state, and most have limitations. ‘‘Hormone refractory prostate cancer’’ is a term commonly used but is often inaccurate. Many patients who progress on ADT may still respond to further hormonal manipulation, such as inhibition of ligand– androgen receptor interaction or suppression of adrenal androgens (Scher et al. 1995). The term androgen-independent prostate cancer (AIPC) also can be misleading. AIPC is ‘‘androgen independent’’ in the sense that CaP flourishes at castrate levels of androgen, but the disease remains to some extent dependent on signaling of the androgen receptor. ‘‘CRPC’’ has been adopted as the best descriptor, is becoming more widely used, and will serve as the term describing castrateresistant disease in this chapter. The Prostate Specific Antigen Working Group was organized in 1999 to establish a standard definition for progression of CaP in the castrate state and maintain consistency across clinical trials (Bubley et al. 1999). The group defined progression as an increase in radiographic measurements of soft tissue CaP metastases, an increase in the number of metastatic sites on bone scan, or two consecutive increases in prostate-specific antigen (PSA). Disease progression must occur after bilateral orchiectomy or in the setting of a serum testosterone of 50 ng/dl. These criteria broadly characterize CRPC, yet they do not neatly categorize all men with CRPC. The biology and natural history of CRPC can vary widely from patient to patient. Further categorization is necessary to better assess prognosis and optimal treatment for CRPC patients. Prior to the era of widespread PSA testing, there was considerably more uniformity among CRPC patients. Since biomarkers such as PSA that herald the onset of advanced disease did not exist, surgical or chemical castration was rarely performed prior to the onset of CaP-related symptoms. Thus, virtually all men who developed CRPC reached this state in the setting of metastatic disease. In the PSA era there is considerably more clinical diversity among CRPC patients, and the variation in clinical presentation at the onset of CRPC has important implications. Much of this heterogeneity is related to the substantial stage migration that has occurred since the introduction of PSA testing. The majority of PSA-screened CaP patients are treated with radical prostatectomy (RP) or radiation therapy (RT) for localized disease (Catalona et al. 1993; Farkas et al. 1998), and up to 40% of these patients experience biochemical recurrence, defined as a serial rise in PSA in the absence of clinical metastases (Moul 2000). A rise in PSA is their first manifestation of recurrence, usually occurring well before the onset of metastatic CaP. Despite a lack of convincing evidence of benefit

Clinical Progression to Castration-Recurrent Prostate Cancer

59

of early intervention, at least one-third of patients are treated with ADT within 5 years of PSA-only recurrence without evidence of metastases (Mehta et al. 2004). Analysis of the Cancer of the Prostate Strategic Urologic Research Endeavor (CaPSURE) database, a network of urologists at more than 30 community and academic urology practice sites, showed that the median time from biochemical recurrence to the initiation of ADT is approximately 9 months (Mehta et al. 2004). As a result of starting ADT early in the course of disease, a considerable number of patients develop CRPC before the onset of clinical progression. The natural history of this group of patients is different from that of CRPC patients with metastatic disease, as will be discussed later. Another important factor dividing CRPC patients into separate biologic categories is responsiveness to secondary hormonal manipulations. One group of patients progresses despite castrate levels of testosterone but is still capable of responding to further hormonal manipulation. A second group is deemed insensitive to all forms of hormonal therapy and is truly ‘‘hormone refractory.’’ Finally, it is now well established that ADT in the neoadjuvant and adjuvant setting with radiation therapy (RT) improves outcomes compared with RT alone for those with intermediate- or high-risk CaP. Should these patients recur after treatment, ADT is often reinitiated. This introduces another factor distinguishing one group at risk for CRPC patient from another: those who have been exposed to ADT in the past versus those who have not. Some evidence suggests that such patients progress to CRPC more rapidly. In short, CRPC can be loosely defined as progression of disease despite castrate levels of androgen. However, a more precise definition must take into account other factors. As will be discussed later, responsiveness to other hormonal manipulations, extent of disease, and timing of initial ADT affect progression from androgen-naı¨ve to castration-resistant CaP. These distinctions also affect prognosis and treatment options.

3 Clinical Predictors for the Development of CRPC In the PSA era, clinical progression from treatment for localized disease to prostate cancer-specific mortality (PCSM) typically begins with PSA recurrence, ultimately followed by development of metastases and death (Pound et al. 1999). Several retrospective and prospective studies have examined time to reach these three endpoints (The Medical Research Council Prostate Cancer Working Party Investigators Group 1997; Moul 2000; Messing et al. 1999). Meanwhile, sometime during this series of events, ADT is started and CRPC develops. The length of time between the start of ADT and the development of CRPC, as gauged by PSA levels, may influence the time to clinical progression or PCSM; there is evidence that the time to PSA progression on ADT has prognostic significance (Shulman and Benaim 2004; Svatek et al. 2006). Moreover, evaluating time to PSA progression on ADT may provide important insight into the biology of CaP, as PSA elevation is the first

60

M. Pomerantz, P. Kantoff

indication of CaP growth in the castrate state. However, few studies have comprehensively examined time to ADT resistance as a clinical endpoint, and fewer still have completely captured the heterogeneity of patients developing CRPC. A retrospective series of 553 patients treated with ADT at the Dana-Farber Cancer Institute (DFCI) is among the largest studies to date analyzing time to CRPC (Ross et al. 2008). Most in this cohort were treated in the setting of relapse after local treatment with RP or RT. Approximately half of the patients had developed metastatic disease by the time ADT was initiated, and 16% had previously received ADT as part of neoadjuvant and/or adjuvant treatment. With a median follow-up of over 5 years, the median time to PSA progression from the start of ADT was 23.7 months (Ross et al. 2008). Other, smaller studies have reported time to CRPC among ADT patients (Oefelein et al. 2002; Fowler et al. 1995). One retrospective analysis consisted of 245 patients with localized CaP and 78 patients with metastatic CaP followed at the University of Mississippi. Those with localized CaP were ineligible for definitive treatment with RP or RT due to advanced disease, age, or comorbidity. ADT was the initial CaP treatment for all patients. As in the DFCI cohort, PSA progression was analyzed as a clinical endpoint. Median time to CRPC was only 15 months in the nonmetastatic subgroup and 10 months in the metastatic subgroup. The differences in outcome among these datasets reflect the heterogeneity of patients receiving ADT. Though Gleason scores appear similar across the two studies, shorter time to CRPC is related to the extent of disease at the start of ADT. Virtually all patients in the University of Mississippi cohort, for example, were diagnosed due to physical findings or CaP-related symptoms. The majority of patients classified as having ‘‘localized’’ disease had locally advanced disease, such as lymph node or seminal vesicle involvement (Fowler et al. 1995). The median pre-ADT PSA in the locally advanced group was 49 ng/mL, compared with 8.4 ng/mL in the DFCI subgroup without metastatic CaP. Overall, the differences across studies demonstrate how the development of CRPC is influenced by host factors. This observation was clearly demonstrated within each series, when patients who started ADT in the setting of metastases were directly compared with those who started before objective evidence of metastases. In each study, the extent of disease at the time of initiation of ADT significantly effected the time to development of CRPC (Ross et al. 2008; Fowler et al. 1995). The biological basis for this is not known, but it may be that advanced, metastatic tumors contain a higher number and/or proportion of cells whose growth is androgen independent, accounting for a more rapid detection of CRPC as these cells flourish. Clinical parameters at the time of ADT initiation also predicted time to resistance in these studies. In the DFCI cohort, PSA > 16.3 ng/mL was a predictor of CRPC (p < 0.05) (Ross et al. 2008). A higher PSA level, like metastatic disease itself, perhaps reflects a higher volume of disease, a larger volume of androgenindependent cells, and therefore shorter time to progression. However, this association was limited to only those patients without evidence of metastases at the ADT start date. A similar finding was reported in the University of Mississippi cohort. A direct correlation was noted between pretreatment PSA levels and time to PSA

Clinical Progression to Castration-Recurrent Prostate Cancer

61

elevation in the group with localized disease but not in the group with metastases (Fowler et al. 1995). Why this effect should be limited to those without evidence of metastases is less clear. Perhaps, once CaP tumors reach a volume detectable by CT or bone scan, absolute PSA value is a less meaningful surrogate for the size of the androgen-independent cell population. The presence of detectable metastatic tumors in itself may indicate that an appreciable number of castrate-resistant cells exist, regardless of PSA level. Other pretreatment PSA parameters may provide a more accurate reflection of the biology of the disease and how quickly CRPC will emerge in the metastatic setting. In particular, a shorter PSA doubling time (PSADT) prior to ADT is associated with a shorter time to progression to CRPC. In the DFCI cohort, across both metastatic and nonmetastatic patients, those with a PSADT < 3 months prior to ADT experienced PSA progression sooner than those with a longer PSADT (median 12 months vs. 33 months, p = 0.02) (Ross et al. 2008). As will be discussed later, PSA kinetics are also strongly associated with clinical outcome in patients with CRPC. Another PSA parameter predicting development of CRPC is PSA nadir. A lower PSA nadir (0.2 ng/mL in the DFCI study) is associated with longer time to progression to CRPC when using PSA endpoints (Ross et al. 2008; Oefelein et al. 2002; Fowler et al. 1995). Of note, Gleason score greater than 7 was associated with shorter time to CRPC among patients with metastatic disease in the DFCI cohort (Ross et al. 2008). This suggests that high-grade, androgen-naı¨ve, metastatic CaP becomes resistant sooner than low- or intermediate-grade disease (Morote et al. 2005). Yet, this biologic event may not have significant clinical consequences. Gleason score is not consistently associated with survival in patients with metastatic disease starting ADT (Yossepowitch et al. 2007; Figg et al. 2004). A particularly intriguing finding from the DFCI study suggests that prior ADT for local disease affects the development of CRPC. The median duration of neoadjuvant and adjuvant ADT was 8 months, and the median duration between completion of ADT as a part of local therapy and the beginning ADT for recurrence was about 22 months. Prior therapy was associated with shorter time to CRPC in metastatic and nonmetastatic patients, even when taking other established prognostic factors into account (Ross et al. 2008). This suggests at least two possible models for CRPC. One is that time on ADT is ‘‘fixed’’ no matter when the ADT is administered. If a patient’s disease is ‘‘calibrated’’ to become resistant after a biologically specified period of ADT, the duration of effective ADT can be divided into different segments, for example, a period of adjuvant therapy and a later period of salvage therapy. Time to CRPC is then dependent on the total time treatment was administered. Another possibility is that the time disease remains androgen-sensitive is ‘‘set’’ from the time of first exposure to ADT until the time of CRPC, and includes periods off treatment. In this scenario, adjuvant ADT and ADT after biochemical recurrence are, in effect, part of an intermittent ADT regimen, with a long interval between treatments. Experience with intermittent androgen deprivation (IAD) informs this issue. Several phase II and phase III studies have examined the efficacy of administering

62

M. Pomerantz, P. Kantoff

chemical castration for a proscribed period of time, then withholding treatment until disease progresses to a predetermined PSA threshold. It has been shown that the time from start of hormonal therapy until the development of CRPC is similar for both IAD and continuous ADT (Stewart et al. 2005). This is consistent with the aforementioned latter scenario. On the other hand, these data are not consistent with data from Radiation Therapy Oncology Group (RTOG) trial 86-10. In this trial, men with extensive local involvement of CaP were randomized to 4 months of neoadjuvant ADT followed by RT or RT alone (Pilepich et al. 2001). Secondary analysis was performed to assess overall and disease-specific survival among men in both treatment arms who ultimately required salvage ADT (Shipley et al. 2006). Outcomes were not significantly different between those who received neoadjuvant treatment and those who did not. Unlike the DFCI analysis, the study did not examine time to PSA recurrence and therefore did not capture the earliest marker for CRPC. The findings may not contradict one another. However, they do raise the possibility that there is a threshold duration of prior ADT, which does not compromise outcome. The median duration of neoadjuvant/adjuvant ADT in the DFCI series was twice as long as the length of ADT in the RTOG study. Serum testosterone levels after castration levels may also be a significant predictor for development of CRPC. In a recent study, serial serum testosterone levels were measured over the first 6 months of ADT in nonmetastatic CaP patients. Patients were treated with LHRH agonist alone or in combination with an antiandrogen (Morote et al. 2007). Among those treated with LHRH agonist alone, time to CRPC was significantly shorter for those with a testosterone level greater than 32 ng/dL (88 months vs. 137 months). Interestingly, for patients receiving combined androgen blockade (LHRH agonist and antiandrogen), this difference was not seen, even for patients with a testosterone level greater than 50 ng/dL. This suggests that post-ADT testosterone levels can predict time to CRPC. This study also nicely demonstrates that CRPC patients are not necessarily ‘‘hormone refractory.’’ Further manipulation of the androgen axis, in this case with an antiandrogen, may provide further benefit to the CRPC patient. In summary, patients are started on ADT at widely varying stages of their disease. This significantly influences the development of resistance to castrate levels of testosterone. The extent of disease, PSA parameters, Gleason score, previous exposure to ADT, and adrogen levels are factors that may impact the time to CRPC.

4 CRPC Natural History and Predictors of Clinical Outcome Once CRPC develops, insight into its natural history and predictors of clinical outcome are valuable and may help direct treatment decisions. Extent of disease is, again, a critical consideration and influences the natural history of castrate-resistant disease. Patients with PSA-only CRPC appear to have a better outcome compared than patients with CRPC that develops in the setting of metastases. However, PSA-

Clinical Progression to Castration-Recurrent Prostate Cancer

63

only CRPC is a relatively new entity, a result of changing practice patterns in the PSA era. Few studies have focused directly on CRPC patients with no clinical evidence of disease. The most thorough examination of this patient population was performed as part of an aborted study evaluating the effect of zoledronic acid on time to bone metastases in men with PSA-only CRPC (Smith et al. 2005). The study was closed due to a lower than expected event rate, but valuable insight was gained into the natural history of CRPC in the absence of clinical disease. The study enrolled 201 CRPC patients and at 2 years, only 33% developed bone metastases. Median metastasis-free survival was 907 days, and median overall survival was not reached at 2-years follow-up. The relatively indolent nature of disease in this study stands in contrast to CRPC in the setting of metastasis, as discussed later. It should be noted, however, that interventions with secondary hormonal therapy or chemotherapy were not documented in this cohort. It is therefore unknown if further therapeutic interventions contributed to outcome. PSA levels were well annotated in this study and served as useful predictors of outcome. A longer PSA doubling time, for example, was associated with metastasis-free survival – a parameter that is also a strong predictor of outcome in hormone-sensitive CaP (D’Amico et al. 2003). In multivariate analysis, PSA > 10 ng/mL and increased PSA velocity were associated with shorter time to first metastasis and were predictive of overall survival (Smith et al. 2005). Notably, PSA velocity was also found to be associated with mortality in a similar cohort of CRPC patients analyzed by D’Amico et al. (2005). Patients in this cohort, assembled from two multi-institutional databases, started ADT for PSA recurrence after RP or RT and ultimately developed CRPC. PSA velocity greater than 1.5 ng/mL/year after the start of ADT was associated with increased risk for all-cause and prostate cancer-specific mortality. The natural history of metastatic CRPC has been more widely studied, though it is important to note that even within this population of CRPC patients, there is considerable clinical variability. Much of the data concerning the metastatic CRPC is derived from several studies enrolling a high percentage of patients with significantly advanced disease (Scher et al. 1999; Vollmer et al. 1998; Kantoff et al. 1999; Tannock et al. 2004; Petrylak et al. 2004). Two groups collected data from several such trials to describe the clinical course of metastatic CRPC and to define predictors of outcome (Smaletz et al. 2002; Halabi et al. 2003). The first analysis was derived from clinical series at Memorial Sloan-Kettering Cancer Center (MSKCC). A total of 409 patients were included in a discovery set and 433 in a validation set. Median survival across both groups was less than 16 months (Smaletz et al. 2002). The second analysis was based on a collection of six studies conducted by the Cancer and Leukemia Group B (CALGB). A total of 760 patients were included in the learning sample and 341 in a validation sample. Median survival across these groups was 13 months after a follow-up of 37 months (Halabi et al. 2003). Predictors for survival after multivariable analysis from each study were used to develop nomograms. Both studies found poor performance status, lower

64

M. Pomerantz, P. Kantoff

hemoglobin, higher LDH, and higher alkaline phosphatase to be associated with poorer overall survival. The CALGB study also incorporated Gleason score >7 and presence of visceral metastases into its nomogram, while the MSKCC study did not find that these parameters added information to their model. The MSKCC nomogram included age and levels of albumin, while the CALGB model did not. A higher PSA prior to treatment for CRPC correlated with worse outcome in the MSKCC study, but the effect was very slight. In fact, when comparing the 25% and 75% PSA quartiles, no difference in survival was seen (Smaletz et al. 2002). In an analysis of 143 CRPC cases from DFCI, higher PSA levels predicted improved survival for men with bone metastases and a normal alkaline phosphatase (Xie et al. 2007). PSA in this setting may be a surrogate for two competing biologic processes. On the one hand, a higher PSA represents progressive disease, greater tumor volume and, as a result, worse outcome. As such, PSA also may be associated with worse performance status, higher LDH, and lower hemoglobin – parameters that directly correlate with outcome. On the other hand, a higher PSA may be associated with a more differentiated tumor, and a tumor with greater proportion of androgen-dependent cells. At an advanced stage of disease, this could represent a more favorable biology. Also, PSA itself is reported to have antiangiogenic properties (Balk et al. 2003). Finally, PSA has been implicated in the degradation of PTHrP (Balk et al. 2003), perhaps accounting for the normal alkaline phosphatase observed in the subset of patients in whom PSA appeared protective. The MSKCC and CALGB studies only analyzed single, static PSA levels (Smaletz et al. 2002; Halabi et al. 2003). Since the publication of these studies, numerous groups have demonstrated the predictive value of PSA kinetics. In one series of 160 CRPC patients diagnosed in the PSA era, PSA doubling time was the variable most strongly associated with prostate cancer-specific survival (Shulman and Benaim 2004). Investigators divided the cohort into risk groups and highlighted the heterogeneity in presentation and prognosis among CRPC patients with advanced disease. The group with the most favorable PSA parameters, men with a PSADT >6 months, had a median cancer-specific survival of 89.1 months. Those in the highest risk grouping, which included men with a PSADT <6 months, had a median cancer-specific survival of 14.0 months. Unlike the MSKCC and CALGB studies, no patient in this cohort received cytotoxic therapy, suggesting that castration-resistant disease was discovered relatively early in its course. Likewise, in a similar cohort of 129 men with untreated CRPC, median overall survival was 52 months, and PSADT remained a statistically significant predictor of survival in multivariate models (Svatek et al. 2006). Two recent studies evaluated survival in more advanced CRPC patients, those about to receive chemotherapy (Oudard et al. 2007; Armstrong et al. 2007). These series are particularly informative since most patients received docetaxel-based chemotherapy, the contemporary standard of care. The first study involved a cohort of 250 patients. For patients with a PSADT <45 days, median survival was 16.5 months and with a PSADT of >45 days it was 26.4 months (Oudard et al. 2007). The second study analyzed the cohort from the TAX327 trial, comparing docetaxel-based regimens to mitoxantrone. PSADT was separated into quintiles.

Clinical Progression to Castration-Recurrent Prostate Cancer

65

Median survival for patients with a PSADT <1 month was 13.3 months, and for patients with a PSADT >6 months median survival had not yet been reached (mean survival 22.5 months) (Armstrong et al. 2007).

5 Treatment of CRPC As described earlier, the clinical course of CRPC may vary considerably. As such, treatment should be individualized based on a variety of factors including the presence or absence of metastases, performance status, and PSA kinetics. However, choosing an appropriate treatment for a particular patient can be challenging since there is an absence of survival data for most CRPC therapies. Moreover, CRPC is a difficult condition to evaluate via clinical trials. Overall survival is the most meaningful clinical parameter, but may not be a realistic study endpoint given the long natural history of many with CRPC, particularly those with PSA-only recurrence. Surrogates for survival in CaP have limitations. Compared with most other common cancers, CaP develops few soft tissue lesions measurable by radiograph. The most common site of CaP metastasis is bone, and bone scans are ill suited for accurately measuring changes in disease and are subject to interobserver variation. PSA would seem a worthy surrogate marker in light of the data emerging from CRPC series, as shown earlier. In addition, there are clinical data suggesting that treatment-induced PSA decreases 50% confer a survival advantage (Scher et al. 1997; Small et al. 2004). Yet there remains a lack of consensus regarding the best use of PSA, and no prospective trial data are available to resolve the issue. There may be a patient population in which a watchful waiting approach is appropriate. CRPC patients with PSA-only recurrence after ADT may experience prolonged metastasis-free survival, particularly in the setting of favorable predictors, such as a long PSA doubling time (Smith et al. 2005). At the opposite extreme, cytotoxic therapy may be indicated for patients with rapidly progressive, symptomatic disease. In particular, docetaxel-based therapy has been clearly shown to improve survival and quality of life in this population (Tannock et al. 2004; Petrylak et al. 2004). For the large number of CRPC patients with a disease profile somewhere in between these two scenarios, a secondary hormonal manipulation may be the treatment of choice. When initiating secondary hormonal treatment, it is helpful to consider the pathophysiology of CRPC. Despite the use of the term ‘‘androgen independence’’ to describe this disease state, the androgen axis remains involved in disease progression. In fact, the androgen axis often appears hyperactive in this disease state, and androgen receptor is consistently overexpressed in CRPC (Chen et al. 2004). Thus, relatively low levels of serum and intraprostatic androgen are required to promote disease progression. Moreover, mutations in the androgen receptor have been described, and certain mutations may confer a growth advantage to a given CaP clone (Taplin et al. 1995). The androgen axis hyperactivity may also result from alterations in cofactor expression, increased ligand promiscuity on the part of

66

M. Pomerantz, P. Kantoff

the androgen receptor, or complete independence from ligand-receptor binding (Feldman and Feldman 2001). Because the androgen axis remains very active in CRPC and is helping drive disease progression, ADT must be maintained no matter which treatment course is chosen, including watchful waiting. Though prospective data are lacking, there appears to be benefit in minimizing the ligand available to the androgen receptor. In a retrospective analysis of 341 patients treated in four randomized controlled trials for CRPC, continued androgen suppression was an independent predictor of survival (Taylor et al. 1993). While secondary hormonal therapies have several mechanisms of action, there are two main methods by which these treatments take advantage of the cancer’s continued dependency on the androgen pathway: antagonism of the androgen receptor (AR) and decrease in adrenal androgen production. For patients progressing after LHRH agonist or bilateral orchiectomy alone, AR antagonism with nonsteroidal antiandrogens is the strategy of choice. In a series of over 200 patients receiving the antiandrogen flutamide for metastatic CRPC, response rate was 34.5% and mean duration of response was 24 months (Labrie et al. 1988). Among patients progressing while receiving an antiandrogen (either as part of initial treatment or as salvage treatment) up to 20% experience a greater than 50% decrease in PSA in response to withdrawal of the antiandrogen (Small and Carroll 1994; Small and Vogelzang 1997). The withdrawal effect occurs because the antiandrogen, which once served as an antagonist of the androgen receptor, begins to act as an agonist. It has been shown that over time antiandrogens select for androgen receptor mutations, fundamentally changing their effects. However, this may not be the mechanism of action in all cases (Taplin et al. 1999). Response, when it occurs, generally lasts 3–5 months before further progression. Despite the low rates of withdrawal response, almost all CRPC patients should be given a trial period off antiandrogen before new therapy is started. This transformation from inhibition to activation appears to be a molecular event specific to each individual antiandrogen agent. This is evidenced by the fact that many patients respond to a second antiandrogen after failing a prior one. High doses of the antiandrogen bicalutamide (150–200 mg) are capable of inducing responses in patients who progressed on flutamide (Scher et al. 1997; Joyce et al. 1998). The antiandrogen nilutamide is also effective as a second line antiandrogen. In one study, this agent induced a nearly 50% reduction in PSA in 27% of patients whose disease had progressed on flutamide or bicalutamide (Kassouf et al. 2003). Since the androgen receptor remains active in the setting of CRPC, further suppression of circulating androgen has therapeutic potential. This has been demonstrated with the administration of cytochrome P450 inhibitors, agents that block steroidogenesis in the adrenal gland as well as the testes (Pont et al. 1982). The adrenal gland is the source of approximately 10% of androgen production. The antifungal drug ketoconazole is the most widely used regimen in this class. A phase III trial randomized 260 CRPC patients to antiandrogen withdrawal or antiandrogen withdrawal followed by high-dose ketoconazole 400 mg three times daily with hydrocortisone (Small et al. 2004). Hydrocortisone is routinely included in this

Clinical Progression to Castration-Recurrent Prostate Cancer

67

regimen to prevent adrenal insufficiency. Those progressing on withdrawal alone arm were allowed to receive deferred ketoconazole. PSA response rates were 27% in the ketoconazole arm and 32% when including the patients receiving deferred treatment. Other trials using intermediate to high doses of ketoconazole have shown 50% reduction in PSA in 28–62% of patients (Small et al. 1997; Millikan et al. 2001; Nakabayashi et al. 2006). The agent abiraterone is also a potent inhibitor of adrenal androgen steroidogenesis. It is more focused than ketoconazole, specifically targeting 17-alpha hydroxylase and C17,20-lyase. Early clinical data suggest a profound effect on androgen production, and clinical trials utilizing this agent in patients with CRPC are ongoing (Taplin 2007). Estrogens have known activity in CRPC, a phenomenon reported as early as the 1950s, when Nesbit and Baum reported a 17% response rate in this setting (Nesbit and Baum 1950). Several studies have since demonstrated estrogen’s effects in CRPC, including several in the PSA era (Oh 2002). Estrogens are administered in several forms – diethylstilbestrol (DES), transdermal estradiol, conjugated estrogens (Premarin), and, though no longer available, PC-SPES. Estrogens exert their effects though multiple mechanisms. There is estrogen-mediated inhibition of the pituitary release of gonadotropins, which blocks testosterone production by the testes. It appears unlikely, however, that further decrease in circulating testosterone beyond that achieved by an LHRH agonist is clinically meaningful in CRPC. There is also evidence that estrogen treatment leads to a decrease in adrenal androgen precursors, suggesting a mechanism similar to ketoconazole (Takezawa et al. 2001; Kitahara et al. 1997). The side effects associated with estrogen therapy can be significant. Thomboembolic and cardiovascular events are not uncommon, and even attempts to utilize a lower dose of estrogen (DES 1 mg) result in event rates between 5 and 18% (Oh 2002). Nonetheless, estrogens can induce responses in CRPC, and side effects may be manageable when patients are carefully monitored and treated with antithrombosis prophylaxis. In the PSA era, Smith et al. administered 1-mg DES and reported a 43% PSA response rate in 21 CRPC patients with a 5% rate of deep venous thrombosis (Smith et al. 1998). More recently, 90 CRPC patients were randomized in a phase II crossover study to 3-mg DES or the estrogenic herbal formulation, PCSPES. PSA response to DES was 24% with a thromboembolism rate of 9% after low-dose warfarin prophylaxis (Oh et al. 2004). Less data are available for other estrogen formulations. Transdermal estrogens have an improved safety profile, and no thromboembolic events were noted across two studies following a total of 44 patients (Ockrim et al. 2003; Bland et al. 2005). However, PSA response rates were poor, only 12.5% in one study (Bland et al. 2005). Premarin is composed of conjugated estrogens and has been widely studied as hormone replacement therapy in postmenopausal women. In a phase II trial including 45 CRPC patients, PSA declines of greater than 50% were seen in 25% receiving high-dose Premarin (1.25 mg three times daily) (Pomerantz et al. 2007). Efforts are ongoing to improve the effectiveness of secondary hormonal manipulations. Several agents were developed to better utilize the mechanisms of action described earlier. For example, new androgen receptor antagonists are in early

68

M. Pomerantz, P. Kantoff

phase clinical trials. Other novel agents alter downstream functions of AR. Hsp-90 inhibitors, for example, block a protein–protein interaction necessary for proper AR protein folding (Taplin 2007).

6 Summary The emergence of CRPC marks a critical juncture in the clinical course of the disease. Host factors such as extent of disease significantly impact the development of CRPC. Several clinical parameters, such as PSA kinetics, are associated with time to CRPC and the natural history of CRPC once it develops. For patients whose disease is not rapidly progressive or symptomatic, CRPC may be susceptible to secondary hormonal treatment, a therapy generally less toxic that chemotherapy. Secondary hormonal treatments take advantage of CaP’s continued dependence on the androgen axis. The markers for CRPC development and prognosis and secondary hormonal treatments become useful as increasing number of patients develop CRPC in the absence of any other signs of clinical progression.

References Armstrong AJ, Garrett-Mayer ES, Yang YC, de Wit R, Tannock IF, Eisenberger M. A contemporary prognostic nomogram for men with hormone-refractory metastatic prostate cancer: a TAX327 study analysis. Clin Cancer Res. 2007;13(21):6396–6403. Balk SP, Ko YJ, Bubley GJ. Biology of prostate-specific antigen. J Clin Oncol. 2003;21(2):383– 391. Bland LB, Garzotto M, DeLoughery TG, Ryan CW, Schuff KG, Wersinger EM, Lemmon D, Beer TM. Phase II study of transdermal estradiol in androgen-independent prostate carcinoma. Cancer. 2005;103(4):717–723. Bubley GJ, Carducci M, Dahut W, Dawson N, Daliani D, Eisenberger M, Figg WD, Freidlin B, Halabi S, Hudes G, Hussain M, Kaplan R, Myers C, Oh W, Petrylak DP, Reed E, Roth B, Sartor O, Scher H, Simons J, Sinibaldi V, Small EJ, Smith MR, Trump DL, Wilding G, et al. Eligibility and response guidelines for phase II clinical trials in androgen-independent prostate cancer: recommendations from the Prostate-Specific Antigen Working Group. J Clin Oncol. 1999;17(11):3461–3467. Catalona WJ, Smith DS, Ratliff TL, Basler JW. Detection of organ-confined prostate cancer is increased through prostate-specific antigen-based screening. JAMA. 1993;270(8):948–954. Chen CD, Welsbie DS, Tran C, Baek SH, Chen R, Vessella R, Rosenfeld MG, Sawyers CL. Molecular determinants of resistance to antiandrogen therapy. Nat Med. 2004;10(1):33–39. D’Amico AV, Moul JW, Carroll PR, Sun L, Lubeck D, Chen MH. Surrogate end point for prostate cancer-specific mortality after radical prostatectomy or radiation therapy. J Natl Cancer Inst. 2003;95(18):1376–1383. D’Amico AV, Moul J, Carroll PR, Sun L, Lubeck D, Chen MH. Surrogate end point for prostate cancer specific mortality in patients with nonmetastatic hormone refractory prostate cancer. J Urol. 2005;173(5):1572–1576.

Clinical Progression to Castration-Recurrent Prostate Cancer

69

Farkas A, Schneider D, Perrotti M, Cummings KB, Ward WS. National trends in the epidemiology of prostate cancer, 1973 to 1994: evidence for the effectiveness of prostate-specific antigen screening. Urology. 1998;52(3):444–448; discussion 448–449. Feldman BJ, Feldman D. The development of androgen-independent prostate cancer. Nat Rev Cancer. 2001;1(1):34–45. Figg WD, Franks ME, Venzon D, Duray P, Cox MC, Linehan WM, Van Bingham W, Eastham JA, Reed E, Sartor O. Gleason score and pretreatment prostate-specific antigen in survival among patients with stage D2 prostate cancer. World J Urol. 2004;22(6):425–430. Fowler JE, Jr., Pandey P, Seaver LE, Feliz TP. Prostate specific antigen after gonadal androgen withdrawal and deferred flutamide treatment. J Urol. 1995;154(2 Pt 1):448–453. Halabi S, Small EJ, Kantoff PW, Kattan MW, Kaplan EB, Dawson NA, Levine EG, Blumenstein BA, Vogelzang NJ. Prognostic model for predicting survival in men with hormone-refractory metastatic prostate cancer. J Clin Oncol. 2003;21(7):1232–1237. Huggins C, Hodges CV. The effect of castration of estrogen and of androgen injection on serum phosphatases in metastatic carcinoma of the prostate. Cancer Res. 1941;1:293–297. The Medical Research Council Prostate Cancer Working Party Investigators Group. Immediate versus deferred treatment for advanced prostatic cancer: initial results of the Medical Research Council Trial. Br J Urol. 1997;79(2):235–246. Joyce R, Fenton MA, Rode P, Constantine M, Gaynes L, Kolvenbag G, DeWolf W, Balk S, Taplin ME, Bubley GJ. High dose bicalutamide for androgen independent prostate cancer: effect of prior hormonal therapy. J Urol. 1998;159(1):149–153. Kantoff PW, Halabi S, Conaway M, Picus J, Kirshner J, Hars V, Trump D, Winer EP, Vogelzang NJ. Hydrocortisone with or without mitoxantrone in men with hormone-refractory prostate cancer: results of the cancer and leukemia group B 9182 study. J Clin Oncol. 1999;17(8):2506– 2513. Kassouf W, Tanguay S, Aprikian AG. Nilutamide as second line hormone therapy for prostate cancer after androgen ablation fails. J Urol. 2003;169(5):1742–1744. Kitahara S, Yoshida K, Ishizaka K, Kageyama Y, Kawakami S, Tsujii T, Oshima H. Stronger suppression of serum testosterone and FSH levels by a synthetic estrogen than by castration or an LH-RH agonist. Endocr J. 1997;44(4):527–532. Labrie F, Dupont A, Giguere M, Borsanyi JP, Lacourciere Y, Monfette G, Emond J, Bergeron N. Benefits of combination therapy with flutamide in patients relapsing after castration. Br J Urol. 1988;61(4):341–346. Prostate Cancer Trialists’ Collaborative Group. Maximum androgen blockade in advanced prostate cancer: an overview of the randomised trials. Lancet. 2000;355(9214):1491–1498. Mehta SS, Lubeck DP, Sadetsky N, Pasta DJ, Carroll PR. Patterns of secondary cancer treatment for biochemical failure following radical prostatectomy: data from CaPSURE. J Urol. 2004;171(1):215–219. Messing EM, Manola J, Sarosdy M, Wilding G, Crawford ED, Trump D. Immediate hormonal therapy compared with observation after radical prostatectomy and pelvic lymphadenectomy in men with node-positive prostate cancer. N Engl J Med. 1999;341(24):1781–1788. Millikan R, Baez L, Banerjee T, Wade J, Edwards K, Winn R, Smith TL, Logothetis C. Randomized phase 2 trial of ketoconazole and ketoconazole/doxorubicin in androgen independent prostate cancer. Urol Oncol. 2001;6(3):111–115. Morote J, Esquena S, Abascal JM, Trilla E, Cecchini L, Raventos CX, Orsola A, Planas J, Catalan R, Reventos J. Usefulness of prostate-specific antigen nadir as predictor of androgen-independent progression of metastatic prostate cancer. Int J Biol Markers. 2005;20(4):209–216. Morote J, Orsola A, Planas J, Trilla E, Raventos CX, Cecchini L, Catalan R. Redefining clinically significant castration levels in patients with prostate cancer receiving continuous androgen deprivation therapy. J Urol. 2007;178(4 Pt 1):1290–1295. Moul JW. Prostate specific antigen only progression of prostate cancer. J Urol. 2000;163(6):1632– 1642.

70

M. Pomerantz, P. Kantoff

Nakabayashi M, Xie W, Regan MM, Jackman DM, Kantoff PW, Oh WK. Response to low-dose ketoconazole and subsequent dose escalation to high-dose ketoconazole in patients with androgen-independent prostate cancer. Cancer. 2006;107(5):975–981. Nesbit RM, Baum WC. Endocrine control of prostatic carcinoma; clinical and statistical survey of 1,818 cases. J Am Med Assoc. 1950;143(15):1317–1320. Ockrim JL, Lalani EN, Laniado ME, Carter SS, Abel PD. Transdermal estradiol therapy for advanced prostate cancer–forward to the past? J Urol. 2003;169(5):1735–1737. Oefelein MG, Ricchiuti VS, Conrad PW, Goldman H, Bodner D, Resnick MI, Seftel A. Clinical predictors of androgen-independent prostate cancer and survival in the prostate-specific antigen era. Urology. 2002;60(1):120–124. Oh WK. The evolving role of estrogen therapy in prostate cancer. Clin Prostate Cancer. 2002;1 (2):81–89. Oh WK, Kantoff PW, Weinberg V, Jones G, Rini BI, Derynck MK, Bok R, Smith MR, Bubley GJ, Rosen RT, DiPaola RS, Small EJ. Prospective, multicenter, randomized phase II trial of the herbal supplement, PC-SPES, and diethylstilbestrol in patients with androgen-independent prostate cancer. J Clin Oncol. 2004;22(18):3705–3712. Oudard S, Banu E, Scotte F, Banu A, Medioni J, Beuzeboc P, Joly F, Ferrero JM, Goldwasser F, Andrieu JM. Prostate-specific antigen doubling time before onset of chemotherapy as a predictor of survival for hormone-refractory prostate cancer patients. Ann Oncol. 2007;18 (11):1828–1833. Petrylak DP, Tangen CM, Hussain MH, Lara PN, Jr., Jones JA, Taplin ME, Burch PA, Berry D, Moinpour C, Kohli M, Benson MC, Small EJ, Raghavan D, Crawford ED. Docetaxel and estramustine compared with mitoxantrone and prednisone for advanced refractory prostate cancer. N Engl J Med. 2004;351(15):1513–1520. Pilepich MV, Winter K, John MJ, Mesic JB, Sause W, Rubin P, Lawton C, Machtay M, Grignon D. Phase III radiation therapy oncology group (RTOG) trial 86–10 of androgen deprivation adjuvant to definitive radiotherapy in locally advanced carcinoma of the prostate. Int J Radiat Oncol Biol Phys. 2001;50(5):1243–1252. Pomerantz M, Manola J, Taplin ME, Bubley G, Inman M, Lowell J, Beard C, Kantoff PW, Oh WK. Phase II study of low dose and high dose conjugated estrogen for androgen independent prostate cancer. J Urol. 2007;177(6):2146–2150. Pont A, Williams PL, Azhar S, Reitz RE, Bochra C, Smith ER, Stevens DA. Ketoconazole blocks testosterone synthesis. Arch Intern Med. 1982;142(12):2137–2140. Pound CR, Partin AW, Eisenberger MA, Chan DW, Pearson JD, Walsh PC. Natural history of progression after PSA elevation following radical prostatectomy. JAMA. 1999;281(17):1591– 1597. Ross RW, Xie W, Regan MM, Pomerantz M, Nakabayashi M, Daskivich TJ, Sartor O, Taplin ME, Kantoff PW, Oh WK. Efficacy of androgen deprivation therapy (ADT) in patients with advanced prostate cancer: association between Gleason score, prostate-specific antigen level, and prior ADT exposure with duration of ADT effect. Cancer. 2008;112(6):1247–1253. Scher HI, Steineck G, Kelly WK. Hormone-refractory (D3) prostate cancer: refining the concept. Urology. 1995;46(2):142–148. Scher HI, Liebertz C, Kelly WK, Mazumdar M, Brett C, Schwartz L, Kolvenbag G, Shapiro L, Schwartz M. Bicalutamide for advanced prostate cancer: the natural versus treated history of disease. J Clin Oncol. 1997;15(8):2928–2938. Scher HI, Kelly WM, Zhang ZF, Ouyang P, Sun M, Schwartz M, Ding C, Wang W, Horak ID, Kremer AB. Post-therapy serum prostate-specific antigen level and survival in patients with androgen-independent prostate cancer. J Natl Cancer Inst. 1999;91(3):244–251. Shipley WU, Desilvio M, Pilepich MV, Roach M, 3rd, Wolkov HB, Sause WT, Rubin P, Lawton CA. Early initiation of salvage hormone therapy influences survival in patients who failed initial radiation for locally advanced prostate cancer: A secondary analysis of RTOG protocol 86–10. Int J Radiat Oncol Biol Phys. 2006;64(4):1162–1167.

Clinical Progression to Castration-Recurrent Prostate Cancer

71

Shulman MJ, Benaim EA. The natural history of androgen independent prostate cancer. J Urol. 2004;172(1):141–145. Smaletz O, Scher HI, Small EJ, Verbel DA, McMillan A, Regan K, Kelly WK, Kattan MW. Nomogram for overall survival of patients with progressive metastatic prostate cancer after castration. J Clin Oncol. 2002;20(19):3972–3982. Small EJ, Carroll PR. Prostate-specific antigen decline after casodex withdrawal: evidence for an antiandrogen withdrawal syndrome. Urology. 1994;43(3):408–410. Small EJ, Vogelzang NJ. Second-line hormonal therapy for advanced prostate cancer: a shifting paradigm. J Clin Oncol. 1997;15(1):382–388. Small EJ, Baron AD, Fippin L, Apodaca D. Ketoconazole retains activity in advanced prostate cancer patients with progression despite flutamide withdrawal. J Urol. 1997;157(4):1204– 1207. Small EJ, Halabi S, Dawson NA, Stadler WM, Rini BI, Picus J, Gable P, Torti FM, Kaplan E, Vogelzang NJ. Antiandrogen withdrawal alone or in combination with ketoconazole in androgen-independent prostate cancer patients: a phase III trial (CALGB 9583). J Clin Oncol. 2004;22(6):1025–1033. Smith DC, Redman BG, Flaherty LE, Li L, Strawderman M, Pienta KJ. A phase II trial of oral diethylstilbesterol as a second-line hormonal agent in advanced prostate cancer. Urology. 1998;52(2):257–260. Smith MR, Kabbinavar F, Saad F, Hussain A, Gittelman MC, Bilhartz DL, Wynne C, Murray R, Zinner NR, Schulman C, Linnartz R, Zheng M, Goessl C, Hei YJ, Small EJ, Cook R, Higano CS. Natural history of rising serum prostate-specific antigen in men with castrate nonmetastatic prostate cancer. J Clin Oncol. 2005;23(13):2918–2925. Stewart AJ, Scher HI, Chen MH, McLeod DG, Carroll PR, Moul JW, D’Amico AV. Prostatespecific antigen nadir and cancer-specific mortality following hormonal therapy for prostatespecific antigen failure. J Clin Oncol. 2005;23(27):6556–6560. Svatek R, Karakiewicz PI, Shulman M, Karam J, Perrotte P, Benaim E. Pre-treatment nomogram for disease-specific survival of patients with chemotherapy-naive androgen independent prostate cancer. Eur Urol. 2006;49(4):666–674. Takezawa Y, Nakata S, Kobayashi M, Kosaku N, Fukabori Y, Yamanaka H. Moderate dose diethylstilbestrol diphosphate therapy in hormone refractory prostate cancer. Scand J Urol Nephrol. 2001;35(4):283–287. Tannock IF, de Wit R, Berry WR, Horti J, Pluzanska A, Chi KN, Oudard S, Theodore C, James ND, Turesson I, Rosenthal MA, Eisenberger MA. Docetaxel plus prednisone or mitoxantrone plus prednisone for advanced prostate cancer. N Engl J Med. 2004;351(15):1502–1512. Taplin ME. Drug insight: role of the androgen receptor in the development and progression of prostate cancer. Nat Clin Pract Oncol. 2007;4(4):236–244. Taplin ME, Bubley GJ, Shuster TD, Frantz ME, Spooner AE, Ogata GK, Keer HN, Balk SP. Mutation of the androgen-receptor gene in metastatic androgen-independent prostate cancer. N Engl J Med. 1995;332(21):1393–1398. Taplin ME, Bubley GJ, Ko YJ, Small EJ, Upton M, Rajeshkumar B, Balk SP. Selection for androgen receptor mutations in prostate cancers treated with androgen antagonist. Cancer Res. 1999;59(11):2511–2515. Taylor CD, Elson P, Trump DL. Importance of continued testicular suppression in hormonerefractory prostate cancer. J Clin Oncol. 1993;11(11):2167–2172. Tolis G, Ackman D, Stellos A, Mehta A, Labrie F, Fazekas AT, Comaru-Schally AM, Schally AV. Tumor growth inhibition in patients with prostatic carcinoma treated with luteinizing hormonereleasing hormone agonists. Proc Natl Acad Sci U S A. 1982;79(5):1658–1662. The Veterans Administration Co-operative Urological Research Group. Treatment and survival of patients with cancer of the prostate. Surg Gynecol Obstet. 1967;124(5):1011–1017. Vollmer RT, Dawson NA, Vogelzang NJ. The dynamics of prostate specific antigen in hormone refractory prostate carcinoma: an analysis of cancer and leukemia group B study 9181 of megestrol acetate. Cancer. 1998;83(9):1989–1994.

72

M. Pomerantz, P. Kantoff

Xie W, Nakabayashi M, Regan MM, Oh WK. Higher prostate-specific antigen levels predict improved survival in patients with hormone-refractory prostate cancer who have skeletal metastases and normal serum alkaline phosphatase. Cancer. 2007;110(12):2709–2715. Yossepowitch O, Bianco FJ, Jr., Eggener SE, Eastham JA, Scher HI, Scardino PT. The natural history of noncastrate metastatic prostate cancer after radical prostatectomy. Eur Urol. 2007;51(4):940–947; discussion 947–948.

Differential Roles of Androgen Receptor in Prostate Development and Cancer Progression Shuyuan Yeh, Yuanjie Niu, Hiroshi Miyamoto, Tamin Chang, and Chawnshang Chang

Abstract The androgen depletion therapy (ADT) has become the major treatment for the cancer patients through the use of chemical castration and/or antiandrogens, yet the therapy eventually fails and cancers progress to more advanced stages. The mutation, amplification, overexpression of AR, and cross-talk between AR, AR coregulators, and other growth factor pathways have provided potential explanations for the failure of androgen ablation therapies in some cases. However, whether the differential AR roles in different types of prostate cells could contribute to the failure of ADT remains unclear and will be the focus of this review. AR expresses in both stromal and epithelial compartments of prostate. It has been shown that there are three basic types of prostatic epithelial cells: (i) cytokeratin 8 (CK8)-positive, CK5-negative luminal cells, (ii) CK5/CK8-double positive intermediate cells, and (iii) CK8-negative, CK5-positive basal cells. In addition to prostatic stromal cells, AR expression could be detected in some basal cells, some intermediate cells, and all luminal cells in prostate. By Cre-LoxP strategy, the prostate epithelium-specific AR knockout (pes-ARKO) and inducible-cre ARKO mice were recently established and have allowed the field to address the differential and distinct AR roles in different types of prostatic cells. These ARKO mice were bred with TRAMP prostate cancer model, and results from these models suggest that (i) prostatic epithelial AR plays dual roles as a suppressor of basal cell proliferation and as a survival factor for luminal cells, and (ii) the stromal AR plays a proliferator role to support the epithelial cell survival and proliferation. Using microarray analysis of primary tumor cells isolated from the prostate tumors of pes-ARKO-TRAMP mice, it was found that a series of metastatic genes were altered and responsible for the higher invasiveness and metastatic rates. These recent ARKO animal studies not only advance our understanding of the differential roles of AR in different type of prostatic cells, but also closely reflect the pathological changes for the patients undergoing the ADT. Together, these findings provide new evidences to support the potential beneficial effects of C. Chang(*) George Whipple Lab for Cancer Research, Departments of Pathology and Urology, University of Rochester Medical Center, Rochester, NY, USA, E-mail: [email protected]

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4_3, # Springer Science + Business Media, LLC 2009

73

74

S. Yeh et al.

intermittent ADT therapy, and they also urge the development of cell type and stage selective anti-AR therapies for the prostate cancer patients.

1 Introduction Prostate cancer remains a major factor of cancer morbidity and mortality in man (Jemal et al. 2008). Nearly 200,000 new prostate cancer cases would be diagnosed in the USA and close to 30,000 men would die of the disease yearly. Androgen and androgen receptor (AR) signals are required for prostate development, differentiation, and normal functioning as well as cancer initiation and progression (Huggins and Hodges 1972; Niu et al. 2008a, b). Androgen exerts its biological effects through binding to the AR and inducing AR transcriptional activity. The androgen-induced AR transactivation is modulated by the interaction of the AR with various coregulators in response to various growth factors (Buchanan et al. 2001b; Heinlein and Chang 2002, 2004; Heemers and Tindall 2007). Initial stages of prostate cancer growth could be suppressed by reducing the availability of androgens to cancer cells via surgical or chemical androgen deprivation therapy (ADT). However, ADT ultimately fails, and prostate cancer progresses to a castration-recurrent (androgen-independent) state that is frequently metastatic and almost always fatal. The AR is expressed throughout prostate cancer progression as well as in the majority of patients with castrationrecurrent disease (Chodak et al. 1992; Hobisch et al. 1996; Mohler et al. 1996; Sadi et al. 1991; van der Kwast et al. 1991). The reasons for failure of ADT are not well understood and require further investigation. This chapter discusses findings that the AR can function to stimulate tumor growth or suppress tumor metastasis depending on its expression in different prostatic cells. These differential functions of the AR in different types of prostatic cells could lead to emerging new concepts to treat prostate cancer.

2 The Classic Concept for an AR Role in Stimulating the Growth and Differentiation of Epithelial Cells During Prostate Development The prostate develops from the endodermal urogenital sinus (UGS) (Cunha et al. 2004). The UGS consists of an outer layer of embryonic connective tissue called the urogenital sinus mesenchyme (UGM), and an inner layer called the urogenital sinus epithelium (UGE). UGM expresses the AR, which could mediate the effect of androgens in the initial step of prostatic development. In response to androgen secreted from fetal testis, epithelial buds emerge and grow into the surrounding UGM. In rodents, solid prostate buds start to elongate and undergo branching morphogenesis in the prenatal period, and the process continues until maturation

Differential Roles of Androgen Receptor in Prostate Development

75

at the end of puberty (Cunha et al. 2004; Kinbara and Cunha 1996; Sugimura et al. 1986a). Beginning at the neonatal period, the epithelial cords undergo ductal canalization during which the epithelial cells differentiate into the luminal and basal cells (Cunha et al. 2004; Prins and Putz 2008). Prostatic epithelial differentiation is accompanied by differentiation of the mesenchyme into fibroblasts and smooth muscle cells (Hayward et al. 1996a, 1997). In 1978, Cunha (Cunha and Lung 1978) reported that when the UGM compartment of UGS from wild-type mouse embryo was combined with epithelium from the neonatal bladder of wild-type or testicular feminized mouse (Tfm, a mouse model lacking functional AR) and transplanted into the kidney capsules of recipient mice, the tissues developed into glandular structures histologically resembling prostatic glands. In contrast, when the UGM of embryonic UGS from Tfm mice was combined with epithelium from bladder of wild-type or Tfm mice, the combined tissues developed into structures histologically resembling vaginal epithelium. These results suggested that (a) the AR in the UGM regulates signals for ductal morphogenesis for the initiation of prostatic development through a paracrine mechanism, and (b) Tfm bladder epithelium is capable of participating in an androgenic response from wild-type UGM cells. In spite of the importance of stromal AR signals, the postnatal development and maturation of the prostate are believed to depend on a reciprocal interaction between the stromal (mesenchymal) and epithelial compartments (Hayward et al. 1998). Indeed, experimental evidence showed that the epithelial AR is required for epithelial cell differentiation, expression of some prostatic secretory proteins (Cunha and Young 1991; Donjacour and Cunha 1993), and differentiation of stromal fibroblasts into the smooth muscle cells (Cunha et al. 1992; Donjacour and Cunha 1988; Hayward et al. 1996b). Other experiments indicated that castration of the adult mice resulted in prostate involution, which was attributed to epithelial cell apoptosis (Isaacs 1984), and androgen supplementation in castrated animals resulted in restoration of prostate size and ductal morphology (Kyprianou and Isaacs 1988; Sugimura et al. 1986b). These experiments indicated that the growth of adult prostate epithelial cells is dependent on AR signaling. Further reported data indicated that apoptosis of prostate tissues recombined from rat UGM + wild-type epithelium or rat UGM + Tfm epithelium was similar in castrated hosts, and androgen supplementation inhibited apoptosis of these recombinants in a similar manner (Kurita et al. 2001). Together, those observations suggested that the regeneration of prostate epithelial cells is dependent on stromal, but not epithelial, AR signaling. Our recent study of transgenic mice, in which the AR was specifically knocked out from stromal smooth muscle cells or fibroblasts, indicated that reducing AR signaling in the prostatic stromal cells resulted in the development of smaller prostate glands due to a moderate decrease in proliferation and substantial increase in apoptosis in the epithelium (Yeh and Chang et al. unpublished observations). These observations strongly support the view that prostatic stromal AR is an important modulator for the epithelial cell survival, proliferation, and differentiation.

76

S. Yeh et al.

3 Prostatic AR as a Suppressor of Basal Cells and a Survival Factor for Luminal Cells Although results from the embryonic tissue recombination studies suggested that stromal AR plays a dominant role over the epithelial AR, the studies using embryonic tissues would allow only 4–6 weeks of observation. Thus, tissue recombination studies were not adequate to assess the relative contributions of epithelium and stroma in the adult prostate. To study AR functions in adult prostate, a temporal and tissue-specific strategy to control AR gene expression was necessary, and the CreloxP recombination system (Lakso et al. 1992; Nagy 2000; Yeh et al. 2002) is very useful for generating transgenic animal models with conditional AR gene knockout. Prostatic epithelial cells develop from stem cells through proliferation stepwise into basal and intermediate cells that finally differentiate into epithelial luminal cells (Litvinov et al. 2003). Earlier studies have shown that there are three basic types of prostatic epithelial cells: (a) cytokeratin 8 (CK8)-positive, CK5-negative luminal cells, (b) CK5/CK8-double positive intermediate cells (van Leenders and Schalken 2003), and (c) CK8-negative, CK5-positive basal cells (van Leenders et al. 2001). AR expression is detected in some epithelial basal cells, some intermediate cells, and all luminal cells (Mirosevich et al. 1999). By crossing floxed AR mice (Yeh et al. 2002) with probasin-Cre mice (Wu et al. 2001), we have generated prostate epithelium-specific AR knockout (pes-ARKO) mice to study the role of the epithelial AR in prostatic tissue homeostasis (Wu et al. 2007). Results suggest that prostatic epithelial AR plays dual roles as a suppressor of basal cell proliferation and a survival factor for luminal cells. Within the epithelium, the probasin promoter transgene is expressed increasingly from 2 to 7 weeks and expression is sustained throughout life (Wu et al. 2001). Consistent with increasing probasin-Cre expression, our data showed that the epithelial AR expression gradually decreased in the prostate of pes-ARKO mice beginning at 6 weeks of age and was undetectable at 24 weeks of age (Wu et al. 2007). Histological analysis of the ventral prostate in pesARKO mice suggested a progressive decrease in epithelial height, loss of glandular infolding, and increase in luminal epithelial cell apoptosis at 6–32 weeks of age. As the pes-ARKO animals matured, the p63-positive epithelial basal cell population increased during puberty and remained elevated, while the CK8/CK18-positive epithelial luminal cell population declined. The loss of epithelial luminal cells was accompanied by decreased expression of the differentiation markers, probasin, PSP-94, and Nkx3.1. At 24 weeks of age, the glands of pes-ARKO mice exhibited one layer of undifferentiated epithelial cells with increased incorporation of 5-bromodeoxyuridine (BrdU) suggesting an increase in epithelial cell proliferation. Those BrdU-positive cells primarily colocalized with CK5-positive basal cells. To rescue the phenotype, we then generated the AR(T857A)-pes-ARKO double transgenic mice by crossing pes-ARKO mice with AR(T857A) transgenic mice carrying an androgen-responsive AR mutant (Han et al. 2005). These mice exhibited normal prostatic morphology and glandular histology. These results suggested that the increase in epithelial basal proliferation and the loss of epithelial luminal

Differential Roles of Androgen Receptor in Prostate Development

77

cells in pes-ARKO mice are attributable to the loss of epithelial AR signaling. These observations not only support the view that the AR in prostatic epithelial cells is required for epithelial cell differentiation (Donjacour and Cunha 1993), but also suggest that the epithelial AR functions as a suppressor for basal cell proliferation and a survival regulator for differentiated epithelial luminal cells.

4 Differential Roles of AR in Prostate Cancer Progression: The Stromal AR as a Tumor Proliferation Stimulator and Epithelial AR as a Tumor Suppressor Through In Vitro and In Vivo Evidences The success in generating cell-type selective ARKO mice using the cre-loxP strategy (Yeh et al. 2002) permitted further investigation of the role of epithelial AR in prostate cancer progression by creating a pes-ARKO in a mouse TRAMP prostate cancer model. The TRAMP mouse develops spontaneous prostate tumors starting at 10–12 weeks of age (Gingrich et al. 1997). There are three basic types of prostate epithelial cells: luminal, intermediate, and basal cells. We have bred TRAMP mice with the floxed AR transgenic and probasin-Cre mice to generate pes-ARKO-TRAMP mice (Niu et al. 2008a). The knockdown of the AR from the epithelial cells in pes-ARKO-TRAMP mice proceeded gradually with age in a time course very similar to that of pes-ARKO mice (Wu et al. 2007). The prostate of pes-ARKO-TRAMP mice exhibited elevated epithelial cell apoptosis compared to the prostates of wild-type AR (WtAR) TRAMP mice at 16 weeks of age. Double immunostaining indicated that the apoptotic cells were CK8-positive epithelial luminal cells. In contrast, there was an increasing number of proliferating cells in the CK5-positive epithelial basal compartment as indicated by the expression of Ki 67 (a proliferation marker) and BrdU incorporation in pes-ARKO-TRAMP mice compared with that in WtAR TRAMP mice (Niu et al. 2008a). Increased apoptosis in the epithelial luminal cells and increased proliferation in the epithelial basal cells resulted in an increase in a population of cells that were CK5/CK8-double positive intermediate cells. Interestingly and unexpectedly, we observed that pes-ARKO-TRAMP mice developed larger and less differentiated primary tumors in the ventral prostate than TRAMP mice at 16 weeks of age (Niu et al. 2008a). CD44, an early progenitor cell marker, expresses in basal and intermediate prostatic cells (Liu et al. 2004). In addition, it has been found that CD44-positive and AR-negative prostate cancer cells purified from human prostate cancer xenografts are enriched in tumorigenic and metastatic progenitor cells (Patrawala et al. 2006). The primary tumors of pes-ARKO-TRAMP exhibited a higher population of CK5/CK8- and CD44-positive cells than WtAR TRAMP mice. These results suggest that knockout of AR in the prostatic epithelium of TRAMP mice resulted in cell population changes with expansion of intermediate-like tumor cells and

78

S. Yeh et al.

decrease of secretory epithelial luminal cells in the prostate of pes-ARKO-TRAMP mice (Niu et al. 2008a). This is in agreement with a prior report that CK5-positive intermediate cells increased from 29% to 75% when prostate cancer patients received ADT (van Leenders et al. 2001). Moreover, these results confirm the observations in pes-ARKO mice that epithelial AR functions both as a proliferation suppressor of basal intermediate cells and as a survival factor for epithelial luminal cells. Since prostate cancer arises from epithelial cells in TRAMP mice, the dual opposing functions of the epithelial AR might also influence prostate cancer development by favoring the survival of differentiated luminal cells and suppress the expansion of CK5/CK8-positive basal intermediate epithelial cancer cells. Using animal cancer models, it was also found the AR signaling might influence metastasis of prostate cancer. The size of metastatic tumors in pelvic lymph nodes (PLN) in pes-ARKO-TRAMP mice was increased compared with that in their WtAR TRAMP littermates at 24 weeks of age. We observed that primary cultured PLN tumor cells from pes-ARKO-TRAMP mice were more invasive than those from WtAR TRAMP mice in the matrigel invasion assay. Furthermore, restoring the functional AR by transfection of human AR cDNA reduced the invasiveness of PLN tumor cells from pes-ARKO-TRAMP mice (Niu et al. 2008a). It was further examined whether the AR status affected the invasiveness of several human prostate cancer cell lines, which include PC-3 and CWR22rv1 cells. The AR-negative PC-3 cells were isolated from a bone metastasis and the Ar-positive CWR22rv1 cells were isolated from a prostate tumor growing despite ADT (Nagabhushan et al. 1996). Stable AR-positive transfectants of PC-3 cells (PC-3-AR9 cells) also exhibited less invasiveness than control vector-transfected PC-3 (PC-3-v) cells. AR was knocked down in CWR22rv1 cells using a homologous AR gene recombination strategy (Yeh et al. 2003) to obtain CWR22rv1 cells with knockdown (KD) of AR expression (CWR22rv1-ARKD cells). CWR22rv1ARKD cells express much less AR and are more invasive than parental cells in the Boyden chamber invasion assay. Since PC-3 cells were isolated from a bone metastasis, these cells were injected into the tibia of athymic nude mice according to Corey’s method (Corey et al. 2002). We found that PC-3-v tumors grew more aggressively and more invasively than PC-3-AR9 tumors as determined by X-ray analysis and measurement of tumor weights in the mouse tibia. These data from knockin of the functional human AR show that absence of AR signaling in prostate cancer cells promotes invasion both in vitro and in vivo. PC-3-v or PC-3-AR9 cells were also orthotopically inoculated into the anterior prostate of nude mice. Results showed that mice inoculated with PC-3-v cells developed bigger primary tumors that were less differentiated and larger PLN metastatic tumors than mice inoculated with PC-3-AR9 cells (Niu et al. 2008a). Nelius et al. transfected PC-3 cells with an inducible AR-expressing transgene and observed that induction of AR expression resulted in decreased invasion in vitro and decreased tumorigenecity due to decreased microvescular density and increased tumor cell apoptosis (Nelius et al. 2007). These results showed that loss of prostatic epithelial AR expression leads to the development of more invasive and metastatic prostate cancers and restoration of

Differential Roles of Androgen Receptor in Prostate Development

79

a functional AR reverses these characteristics. Together, these results suggest that the epithelial AR is defined as a suppressor of prostate tumor growth and metastasis. Tumor microenvironment and stromal–epithelial interaction remain important for tumor initiation, progression, and metastasis of prostate cancer (Bhowmick and Moses 2005; Condon 2005). Our recent data also indicated that the AR in prostatic stroma could function as a stimulator of prostate cancer progression and metastasis (Niu et al. 2008b). Furthermore, human prostatic stromal WPMY1 cells (Webber et al. 1999) were stably transfected with AR short hairpin RNA (AR shRNA) that can be used to silence or knock down endogenous AR to generate WPMY1-ARsh cells. The influence of WPMY1-ARsh and vector-transfected WPMY1-v cells on invasion abilities of PC-3-v and PC-3-AR9 cells was studied using coculture: stromal cells were placed in the bottom chamber and cancer cells were placed in the upper chamber of a Boyden chamber. The results indicated that both PC-3-v and PC-3-AR9 cells were less invasive when cocultured with WPMY1-ARsh cells than WPMY1-v cells. These results indicated that AR signaling in stromal cells functions as a stimulator of cancer cell invasion regardless of the expression of the endogenous AR in cancer epithelial cells. Furthermore, PC-3-v or PC-3-AR9 cells and WPMY1-v or WPMY-ARsh cells were orthotopically coinoculated into the anterior prostate of nude mice to investigate the role of the stromal AR in tumor progression and metastasis (Niu et al. 2008b). Both cancer cell lines produced smaller primary tumors and PLN metastases when combined with WPMY1-ARsh compared with WPMY1-v, although tumors derived from PC-3-AR9 cells were smaller than those derived from PC-3-v cells in either combination. These observations suggest that AR signaling in stromal cells can provide growth or paracrine factor(s) for growth of AR-positive and AR-negative prostate cancer cells and influence their metastasis. In contrast, the smaller tumors observed with PC-3AR9 cells compared with PC3-v cells suggest that AR expression in prostate cancer epithelial cells promotes differentiation, controls growth, and inhibits metastasis of epithelial prostate tumors regardless of the presence of stromal stimulation effects. These observations support the concept that stromal AR functions as a stimulator of prostate cancer epithelial cells by promoting proliferation and inhibiting apoptosis. In addition, the stromal AR appears prometastatic as opposed to the epithelial AR, which appears antimetastatic. An inducible ARKO model of TRAMP (ind-ARKO-TRAMP) (Niu et al. 2008b) was produced by mating female TRAMP mice carrying heterozygous floxed AR with male Mx1-Cre mice (Kuhn et al. 1995). Following injection of pI-pC to induce cre expression and knockout of AR gene at 12 weeks of age, we found that the AR mRNA levels in the prostate of ind-ARKO-TRAMP mice at 16- and 20 weeks of age were knocked down 40–50% when AR mRNA expression was assessed in prostate epithelium and stroma isolated with laser capture microdissection. Knockdown of AR expression resulted in smaller and less differentiated primary prostate tumors in ind-ARKO-TRAMP than in control Wt-AR-TRAMP mice 16–24 weeks of age. The tumors of ind-ARKO-TRAMP mice had lower proliferation rates and higher apoptosis rates than tumors of the control TRAMP mice. The tumors from

80

S. Yeh et al.

ind-ARKO-TRAMP mice had decreased CK8-positive epithelial luminal cells and increased CD44-positive and CK5/CK8-double positive basal intermediate cells compared with control TRAMP mice. Although pes-ARKO-TRAMP and ind-ARKO-TRAMP mice both demonstrated increased epithelial apoptosis and expanded intermediate cells, pes-ARKO-TRAMP mice produced larger and indARKO-TRAMP mice produced smaller metastases compared with control mice. Together, those data support the concept that epithelial luminal ARKO would alter the cell population of prostatic epithelial cells by reducing more differentiated luminal cells and increasing CD44-positive and CK5/CK8-double positive intermediate cells. Epithelial ARKO or AR signal blocking may reduce the survival of luminal epithelial tumor cells, yet would promote the expansion of more malignant intermediate tumor cells. In addition, metastatic tumors were also evaluated when primary tumors reached 1 cm3 in TRAMP (20 weeks of age), pes-ARKO-TRAMP (18 weeks of age), and ind-ARKO-TRAMP (36 weeks of age) mice and found that the well-differentiated primary tumors of TRAMP mice developed small metastases in the PLN. The poorly differentiated tumors of pes-ARKO-TRAMP mice developed much larger PLN metastases and metastasized to multiple organs, whereas those of ind-ARKO-TRAMP mice were smaller and more often invaded into the seminal vesicle and migrated to the liver. Thus, loss of the epithelial AR could promote prostate cancer growth and metastasis. The concurrent knockdown of the stromal and epithelial AR could override partially the affects of the epithelial AR knockout, thereby retarding growth of the primary prostate tumor and suppressing metastasis. Furthermore, ind-ARKO-TRAMP mice with early induction of ARKO had longer survival time than wild-type TRAMP and pes-ARKO-TRAMP mice (Niu et al. 2008b). However, the apparent dominance of stromal AR over epithelial AR function diminished if induction of ARKO happened after the primary tumor had progressed for a period of time in ind-ARKO-TRAMP mice. Together, these observations support the concept that the epithelial AR can function as a differentiation factor and a tumor metastasis suppressor and also suggest that the stromal AR functions as a stimulator of prostate cancer growth and metastasis.

5 Molecular Mechanisms Promoting Metastases Following Depletion of AR Function In addition to characterizing the differential AR roles in the prostate cancer metastasis using cell and mouse models, gene microarrays have been applied to understand the mechanistic changes in metastatic tumors of pes-ARKO-TRAMP and of PC-3-v cancer cell xenografts, and found that several prometastasis genes such as cyclooxgenase-2 (Cox-2) (Attiga et al. 2000), matrix metaloproteinase-9 (MMP-9) (Aalinkeel et al. 2004; Corey et al. 2002; Saleem et al. 2006), interleukin-6 (IL-6)

Differential Roles of Androgen Receptor in Prostate Development

81

(Hammacher et al. 2005; Saleem et al. 2006), and tumor necrosis factor-alpha (TNF-a) (Hammacher et al. 2005; Michalaki et al. 2004) were elevated, while antimetastasis genes such as neutral endopeptidase (NEP) (Nelson and Carducci 2000; Tanimura et al. 2005) and P27(kip1) (Baldassarre et al. 2005; Belletti et al. 2005; Papandreou et al. 1998) were decreased in the tumors of pes-ARKO-TRAMP mice and PC3-v orthotopic grafts compared with the tumors of their AR-expressing counterparts (Niu et al. 2008a). Thus, due to the dual functions of the AR, treatment of prostate cancer patients with ADT should have good responses at an early stage of cancer but the prognosis may worsen when ADT is given at later stages when tumors are less dependent on stromal AR-derived signals for growth, a scenario apparently observed in castrated TRAMP mice (Johnson et al. 2005). These findings may also explain why some patients can benefit from intermittent cycles of ADT and androgen supplementation (Akakura et al. 1993; Feltquate et al. 2006), which may allow the tumor cells to remain sensitive to ADT without pushing the adaptive phenotype changes.

6 Adaptive Phenotypic Changes via AR Somatic Mutations in Prostate Cancer After ADT In addition to managing androgen activity on the prostate cancer patients, the AR is subjected to somatic mutations in prostate cancer, and most of the mutations involve single base change substitution of an amino acid residue and are found more frequently in castration-recurrent, metastatic bone tumors than in primary tumors (Bentel and Tilley 1996; Buchanan et al. 2001a, c; Chen et al. 2005; Gottlieb et al. 2004; Linja and Visakorpi 2004; Marcelli et al. 2000; Taplin et al. 1995; Tilley et al. 1996). Many of these AR mutants display gain-of-function with equal sensitivity toward androgen, increased sensitivity toward dehydroepiandrosterone (DHEA), and a broader range of ligand activation by 17b-estradiol, progesterone, corticosteroids, and/or antiandrogens (Yeh et al. 1997, 1998; Buchanan et al. 2001c; Chen et al. 2005; Fenton et al. 1997; Monge et al. 2006; Shi et al. 2002). Other AR mutants exhibit decreased transactivation, no activity (Shi et al. 2002), or altered interaction with AR coactivatiors (Bentel and Tilley 1996; Duff and McEwan 2005; Li et al. 2005). Some AR mutations produce nonsense codons that lead to expression of truncated AR proteins with significantly altered transcriptional activity (Lapouge et al. 2007). The nonsense AR mutants were detected at higher frequency in metastatic cells, where several of the nonsense mutations coexist with other AR mutants, particularly the promiscuous T877A mutation, and one of the nonsense mutants, Q640X, appeared capable of androgen-independent stimulation of the activity of the AR(T877A) mutant (Alvarado et al. 2005). Somatic mutation of AR also occurs spontaneously in TRAMP mouse prostate tumors at a high rate and the rate of mutation increases further after castration (Han et al. 2001), which suggests an adaptive response. In clinical prostate cancer, AR

82

S. Yeh et al.

mutants are detected more frequently in castration-recurrent disease and after treatment with antiandrogens (Linja and Visakorpi 2004), which also suggests an adaptive change. Expression of the murine AR mutant, AR (E231G), in mouse prostate resulted in development of prostate intraepithelial neoplasia (PIN) (Pinkas and Teicher 2006), which progressed to invasive and metastatic disease (Han et al. 2005;Yeh et al. 1997, 1998). In view of this observation and those gain-of-function AR mutants, which are considered as providing growth advantages for prostate cancer cells, AR may be regarded as a proto-oncogene. Although the linkages of phenotypic changes via AR somatic mutations in prostate cancer cells remain unclear, these changes may contribute to prostate cancer progression and influence the response to ADT. In addition, AR interacts with a group of proteins, AR coregulators, including coactivators and corepressors. To date, there are numerous AR coregulators identified (Heinlein and Chang 2002; Heemers and Tindall 2007). AR requires the proper interaction with its associated proteins to activate the target gene expression. Several AR coregulators are upregulated in advanced prostate cancer (Culig and Bartsch 2006; Fujimoto et al. 2007; Heemers and Tindall 2007; Hu et al. 2004; Kahl et al. 2006; Nishimura et al. 2003; Yang et al. 2007a, b) and increase androgen sensitivity and ligand promiscuity of wild-type AR and some AR mutants (Yeh et al. 1997, 1998; Heinlein and Chang 2002; Rahman et al. 2004). AR-interacting proteins could alter the AR functions both in the prostate cancer cells and in cancerassociated stromal cells. Regulating the AR protein complex in prostatic stromal or epithelial cells could possibly be applied as an alternative therapeutic strategy in combination with ADT to treat prostate cancer. Together, the studies on AR mutation and AR-associated proteins have added more complexity to the understanding of AR functions in prostate cancer initiation and progression.

7 Clinical Implication and New Concept for Prostate Cancer Treatment There are different stages of prostate cancer development: nonmalignant stage, prostatic intraepithelial neoplasia PIN, invasive prostatic adenocarcinoma, and metastatic tumor (Fig. 1). The classic clinical evidence for AR to function as a proliferation stimulator is the observation by Huggins and Hodges (1972) that treatment of prostate cancer patients with castration or injection of estrogen, to reduce levels of circulating testosterone, resulted in reduction of tumor sizes. The success of ADT for treating prostate cancer patients at earlier stages appears to support the classic view that AR functions as a proliferator that stimulates the growth of prostate cancer. However, ADT treatment eventually fails in most patients, and the cancers recur with high rates of metastasis and almost certain mortality. Current evidence of different AR roles in prostate stromal and epithelial cells using multiple in vitro and in vivo strategies indicates that AR signaling does have differential roles on prostate cancer cell proliferation, survival, and metastasis

Differential Roles of Androgen Receptor in Prostate Development

83

Fig. 1 Histopathology of the human prostate cancer progression. (a) normal/nonmalignant prostate, (b) prostatic intraepithelial neoplasia (PIN), (c) invasive adenocarcinoma, and (d) metastatic tumor in lymph node (See Color Insert)

(Wu et al. 2007; Niu et al. 2008a, b). These findings also provide explanations for the fact that ADT will have some positive effects on earlier prostate cancer but will eventually fail via the dual roles of AR signaling in prostate cancer progression. Treatment regimens targeting solely the AR signaling, such as intermittent ADT (Crook et al. 1999; Egawa et al. 2000; Higano et al. 1996; Hurtado-Coll et al. 2002), or secondary hormonal therapy (Sharifi et al. 2008), may have additional benefits, but may eventually fail. Therefore, new treatment regimens should be developed to use in combination with ADT for the management of prostate cancer. Strategies that selectively target the tumor stromal cells (Bouzin and Feron 2007; Hofmeister et al. 2008), and not epithelial cells, should benefit prostate cancer patients. Furthermore, patients with castration-recurrent prostate cancer usually die of metastatic disease, but their survival can be extended if tumor metastasis is delayed or prevented through additional treatment regimens. Treatments targeting metastasisrelated genes downstream of the epithelial AR signaling should improve results from adjunct ADT alone. For example, treatments targeting TGFb1 and/or its receptor (Pinkas and Teicher 2006), Akt, COX-2, MMP-9 (Miyamoto et al.

84

S. Yeh et al.

2005), or other relevant targets should produce clinical benefits for patients with prostate cancer, when used in conjunction with ADT. The discovery of the differential roles of the stromal and epithelial AR on prostate development, cancer progression, and metastasis may facilitate the development of new therapeutic approaches to battle this deadly disease.

References Aalinkeel, R., Nair, M.P., Sufrin, G., Mahajan, S.D., Chadha, K.C., Chawda, R.P. and Schwartz, S. A. (2004) Gene expression of angiogenic factors correlates with metastatic potential of prostate cancer cells. Cancer Res, 64, 5311–5321. Akakura, K., Bruchovsky, N., Goldenberg, S.L., Rennie, P.S., Buckley, A.R. and Sullivan, L.D. (1993) Effects of intermittent androgen suppression on androgen-dependent tumors. Apoptosis and serum prostate-specific antigen. Cancer, 71, 2782–2790. Alvarado, C., Beitel, L.K., Sircar, K., Aprikian, A., Trifiro, M. and Gottlieb, B. (2005) Somatic mosaicism and cancer: a micro-genetic examination into the role of the androgen receptor gene in prostate cancer. Cancer Res, 65, 8514–8518. Attiga, F.A., Fernandez, P.M., Weeraratna, A.T., Manyak, M.J. and Patierno, S.R. (2000) Inhibitors of prostaglandin synthesis inhibit human prostate tumor cell invasiveness and reduce the release of matrix metalloproteinases. Cancer Res, 60, 4629–4637. Baldassarre, G., Belletti, B., Nicoloso, M.S., Schiappacassi, M., Vecchione, A., Spessotto, P., Morrione, A., Canzonieri, V. and Colombatti, A. (2005) p27(Kip1)-stathmin interaction influences sarcoma cell migration and invasion. Cancer Cell, 7, 51–63. Belletti, B., Nicoloso, M.S., Schiappacassi, M., Chimienti, E., Berton, S., Lovat, F., Colombatti, A. and Baldassarre, G. (2005) p27(kip1) functional regulation in human cancer: a potential target for therapeutic designs. Curr Med Chem, 12, 1589–1605. Bentel, J.M. and Tilley, W.D. (1996) Androgen receptors in prostate cancer. J Endocrinol, 151, 1–11. Bhowmick, N.A. and Moses, H.L. (2005) Tumor–stroma interactions. Curr Opin Genet Dev, 15, 97–101. Bouzin, C. and Feron, O. (2007) Targeting tumor stroma and exploiting mature tumor vasculature to improve anti-cancer drug delivery. Drug Resist Updat, 10, 109–120. Buchanan, G., Greenberg, N.M., Scher, H.I., Harris, J.M., Marshall, V.R. and Tilley, W.D. (2001a) Collocation of androgen receptor gene mutations in prostate cancer. Clin Cancer Res, 7, 1273–1281. Buchanan, G., Irvine, R.A., Coetzee, G.A. and Tilley, W.D. (2001b) Contribution of the androgen receptor to prostate cancer predisposition and progression. Cancer Metastasis Rev, 20, 207–223. Buchanan, G., Yang, M., Harris, J.M., Nahm, H.S., Han, G., Moore, N., Bentel, J.M., Matusik, R. J., Horsfall, D.J., Marshall, V.R., Greenberg, N.M. and Tilley, W.D. (2001c) Mutations at the boundary of the hinge and ligand binding domain of the androgen receptor confer increased transactivation function. Mol Endocrinol, 15, 46–56. Chen, G., Wang, X., Zhang, S., Lu, Y., Sun, Y., Zhang, J., Li, Z. and Lu, J. (2005) Androgen receptor mutants detected in recurrent prostate cancer exhibit diverse functional characteristics. Prostate, 63, 395–406. Chodak, G.W., Kranc, D.M., Puy, L.A., Takeda, H., Johnson, K. and Chang, C. (1992) Nuclear localization of androgen receptor in heterogeneous samples of normal, hyperplastic and neoplastic human prostate. J Urol, 147, 798–803. Condon, M.S. (2005) The role of the stromal microenvironment in prostate cancer. Semin Cancer Biol, 15, 132–137.

Differential Roles of Androgen Receptor in Prostate Development

85

Corey, E., Quinn, J.E., Bladou, F., Brown, L.G., Roudier, M.P., Brown, J.M., Buhler, K.R. and Vessella, R.L. (2002) Establishment and characterization of osseous prostate cancer models: intra-tibial injection of human prostate cancer cells. Prostate, 52, 20–33. Crook, J.M., Szumacher, E., Malone, S., Huan, S. and Segal, R. (1999) Intermittent androgen suppression in the management of prostate cancer. Urology, 53, 530–534. Culig, Z. and Bartsch, G. (2006) Androgen axis in prostate cancer. J Cell Biochem, 99, 373–381. Cunha, G.R. and Lung, B. (1978) The possible influence of temporal factors in androgenic responsiveness of urogenital tissue recombinants from wild-type and androgen-insensitive (Tfm) mice. J Exp Zool, 205, 181–193. Cunha, G.R. and Young, P. (1991) Inability of Tfm (testicular feminization) epithelial cells to express androgen-dependent seminal vesicle secretory proteins in chimeric tissue recombinants. Endocrinology, 128, 3293–3298. Cunha, G.R., Donjacour, A.A., Cooke, P.S., Mee, S., Bigsby, R.M., Higgins, S.J. and Sugimura, Y. (1987) The endocrinology and developmental biology of the prostate. Endocr Rev, 8, 338–362. Cunha, G.R., Battle, E., Young, P., Brody, J., Donjacour, A., Hayashi, N. and Kinbara, H. (1992) Role of epithelial–mesenchymal interactions in the differentiation and spatial organization of visceral smooth muscle. Epithelial Cell Biol, 1, 76–83. Cunha, G.R., Ricke, W., Thomson, A., Marker, P.C., Risbridger, G., Hayward, S.W., Wang, Y.Z., Donjacour, A.A. and Kurita, T. (2004) Hormonal, cellular, and molecular regulation of normal and neoplastic prostatic development. J Steroid Biochem Mol Biol, 92, 221–236. Donjacour, A.A. and Cunha, G.R. (1988) The effect of androgen deprivation on branching morphogenesis in the mouse prostate. Dev Biol, 128, 1–14. Donjacour, A.A. and Cunha, G.R. (1993) Assessment of prostatic protein secretion in tissue recombinants made of urogenital sinus mesenchyme and urothelium from normal or androgen-insensitive mice. Endocrinology, 132, 2342–2350. Duff, J. and McEwan, I.J. (2005) Mutation of histidine 874 in the androgen receptor ligandbinding domain leads to promiscuous ligand activation and altered p160 coactivator interactions. Mol Endocrinol, 19, 2943–2954. Egawa, S., Takashima, R., Matsumoto, K., Mizoguchi, H., Kuwao, S. and Baba, S. (2000) A pilot study of intermittent androgen ablation in advanced prostate cancer in Japanese men. Jpn J Clin Oncol, 30, 21–26. Feltquate, D., Nordquist, L., Eicher, C., Morris, M., Smaletz, O., Slovin, S., Curley, T., Wilton, A., Fleisher, M., Heller, G. and Scher, H.I. (2006) Rapid androgen cycling as treatment for patients with prostate cancer. Clin Cancer Res, 12, 7414–7421. Fenton, M.A., Shuster, T.D., Fertig, A.M., Taplin, M.E., Kolvenbag, G., Bubley, G.J. and Balk, S. P. (1997) Functional characterization of mutant androgen receptors from androgen-independent prostate cancer. Clin Cancer Res, 3, 1383–1388. Fujimoto, N., Miyamoto, H., Mizokami, A., Harada, S., Nomura, M., Ueta, Y., Sasaguri, T. and Matsumoto, T. (2007) Prostate cancer cells increase androgen sensitivity by increase in nuclear androgen receptor and androgen receptor coactivators; a possible mechanism of hormoneresistance of prostate cancer cells. Cancer Invest, 25, 32–37. Gingrich, J.R., Barrios, R.J., Kattan, M.W., Nahm, H.S., Finegold, M.J. and Greenberg, N.M. (1997) Androgen-independent prostate cancer progression in the TRAMP model. Cancer Res, 57, 4687–4691. Gottlieb, B., Beitel, L.K., Wu, J.H. and Trifiro, M. (2004) The androgen receptor gene mutations database (ARDB): 2004 update. Hum Mutat, 23, 527–533. Hammacher, A., Thompson, E.W. and Williams, E.D. (2005) Interleukin-6 is a potent inducer of S100P, which is up-regulated in androgen-refractory and metastatic prostate cancer. Int J Biochem Cell Biol, 37, 442–450. Han, G., Foster, B.A., Mistry, S., Buchanan, G., Harris, J.M., Tilley, W.D. and Greenberg, N.M. (2001) Hormone status selects for spontaneous somatic androgen receptor variants that demonstrate specific ligand and cofactor dependent activities in autochthonous prostate cancer. J Biol Chem, 276, 11204–11213.

86

S. Yeh et al.

Han, G., Buchanan, G., Ittmann, M., Harris, J.M., Yu, X., Demayo, F.J., Tilley, W. and Greenberg, N.M. (2005) Mutation of the androgen receptor causes oncogenic transformation of the prostate. Proc Natl Acad Sci USA, 102, 1151–1156. Hayward, S.W., Baskin, L.S., Haughney, P.C., Foster, B.A., Cunha, A.R., Dahiya, R., Prins, G.S. and Cunha, G.R. (1996a) Stromal development in the ventral prostate, anterior prostate and seminal vesicle of the rat. Acta Anat, 155, 94–103. Hayward, S.W., Cunha, G.R. and Dahiya, R. (1996b) Normal development and carcinogenesis of the prostate. A unifying hypothesis. Ann N Y Acad Sci, 784, 50–62. Hayward, S.W., Rosen, M.A. and Cunha, G.R. (1997) Stromal–epithelial interactions in the normal and neoplastic prostate. Br J Urol, 79 Suppl 2, 18–26. Hayward, S.W., Haughney, P.C., Rosen, M.A., Greulich, K.M., Weier, H.U., Dahiya, R. and Cunha, G.R. (1998) Interactions between adult human prostatic epithelium and rat urogenital sinus mesenchyme in a tissue recombination model. Differentiation, 63, 131–140. Heemers, H.V. and Tindall, D.J. (2007) Androgen receptor (AR) coregulators: a diversity of functions converging on and regulating the AR transcriptional complex. Endocr Rev, 28, 778–808. Heinlein, C.A. and Chang, C. (2002) Androgen receptor (AR) coregulators: an overview. Endocr Rev, 23, 175–200. Heinlein, C.A. and Chang, C. (2004) Androgen receptor in prostate cancer. Endocr Rev, 25, 276–308. Heitzer, M.D. and DeFranco, D.B. (2006) Hic-5/ARA55, a LIM domain-containing nuclear receptor coactivator expressed in prostate stromal cells. Cancer Res, 66, 7326–7333. Higano, C.S., Ellis, W., Russell, K. and Lange, P.H. (1996) Intermittent androgen suppression with leuprolide and flutamide for prostate cancer: a pilot study. Urology, 48, 800–804. Hobisch, A., Culig, Z., Radmayr, C., Bartsch, G., Klocker, H. and Hittmair, A. (1996) Androgen receptor status of lymph node metastases from prostate cancer. Prostate, 28, 129–135. Hofmeister, V., Schrama, D. and Becker, J.C. (2008) Anti-cancer therapies targeting the tumor stroma. Cancer Immunol Immunother, 57, 1–17. Hu, Y.C., Yeh, S., Yeh, S.D., Sampson, E.R., Huang, J., Li, P., Hsu, C.L., Ting, H.J., Lin, H.K., Wang, L., Kim, E., Ni, J. and Chang, C. (2004) Functional domain and motif analyses of androgen receptor coregulator ARA70 and its differential expression in prostate cancer. J Biol Chem, 279, 33438–33446. Huggins, C. and Hodges, C.V. (1972) Studies on prostatic cancer. I. The effect of castration, of estrogen and androgen injection on serum phosphatases in metastatic carcinoma of the prostate. CA Cancer J Clin, 22, 232–240. Hurtado-Coll, A., Goldenberg, S.L., Gleave, M.E. and Klotz, L. (2002) Intermittent androgen suppression in prostate cancer: the Canadian experience. Urology, 60, 52–56; discussion 56. Isaacs, J.T. (1984) Antagonistic effect of androgen on prostatic cell death. Prostate, 5, 545–557. Jemal, A., Siegel, R., Ward, E., Hao, Y., Xu, J., Murray, T. and Thun, M.J. (2008) Cancer statistics, 2008. CA Cancer J Clin, 58, 71–96. Johnson, M.A., Iversen, P., Schwier, P., Corn, A.L., Sandusky, G., Graff, J. and Neubauer, B.L. (2005) Castration triggers growth of previously static androgen-independent lesions in the transgenic adenocarcinoma of the mouse prostate (TRAMP) model. Prostate, 62, 322–338. Kahl, P., Gullotti, L., Heukamp, L.C., Wolf, S., Friedrichs, N., Vorreuther, R., Solleder, G., Bastian, P.J., Ellinger, J., Metzger, E., Schule, R. and Buettner, R. (2006) Androgen receptor coactivators lysine-specific histone demethylase 1 and four and a half LIM domain protein 2 predict risk of prostate cancer recurrence. Cancer Res, 66, 11341–11347. Kinbara, H. and Cunha, G.R. (1996) Ductal heterogeneity in rat dorsal-lateral prostate. Prostate, 28, 58–64. Kuhn, R., Schwenk, F., Aguet, M. and Rajewsky, K. (1995) Inducible gene targeting in mice. Science, 269, 1427–1429. Kurita, T., Wang, Y.Z., Donjacour, A.A., Zhao, C., Lydon, J.P., O’Malley, B.W., Isaacs, J.T., Dahiya, R. and Cunha, G.R. (2001) Paracrine regulation of apoptosis by steroid hormones in the male and female reproductive system. Cell Death Differ, 8, 192–200.

Differential Roles of Androgen Receptor in Prostate Development

87

Kyprianou, N. and Isaacs, J.T. (1988) Activation of programmed cell death in the rat ventral prostate after castration. Endocrinology, 122, 552–562. Lakso, M., Sauer, B., Mosinger, B., Jr., Lee, E.J., Manning, R.W., Yu, S.H., Mulder, K.L. and Westphal, H. (1992) Targeted oncogene activation by site-specific recombination in transgenic mice. Proc Natl Acad Sci USA, 89, 6232–6236. Lapouge, G., Erdmann, E., Marcias, G., Jagla, M., Monge, A., Kessler, P., Serra, S., Lang, H., Jacqmin, D., Bergerat, J.P. and Ceraline, J. (2007) Unexpected paracrine action of prostate cancer cells harboring a new class of androgen receptor mutation – a new paradigm for cooperation among prostate tumor cells. Int J Cancer, 121, 1238–1244. Li, W., Cavasotto, C.N., Cardozo, T., Ha, S., Dang, T., Taneja, S.S., Logan, S.K. and Garabedian, M.J. (2005) Androgen receptor mutations identified in prostate cancer and androgen insensitivity syndrome display aberrant ART-27 coactivator function. Mol Endocrinol, 19, 2273–2282. Linja, M.J. and Visakorpi, T. (2004) Alterations of androgen receptor in prostate cancer. J Steroid Biochem Mol Biol, 92, 255–264. Litvinov, I.V., De Marzo, A.M. and Isaacs, J.T. (2003) Is the Achilles’ heel for prostate cancer therapy a gain of function in androgen receptor signaling? J Clin Endocrinol Metab, 88, 2972–2982. Liu, A.Y., Roudier, M.P. and True, L.D. (2004) Heterogeneity in primary and metastatic prostate cancer as defined by cell surface CD profile. Am J Pathol, 165, 1543–1556. Marcelli, M., Ittmann, M., Mariani, S., Sutherland, R., Nigam, R., Murthy, L., Zhao, Y., DiConcini, D., Puxeddu, E., Esen, A., Eastham, J., Weigel, N.L. and Lamb, D.J. (2000) Androgen receptor mutations in prostate cancer. Cancer Res, 60, 944–949. Michalaki, V., Syrigos, K., Charles, P. and Waxman, J. (2004) Serum levels of IL-6 and TNFalpha correlate with clinicopathological features and patient survival in patients with prostate cancer. Br J Cancer, 90, 2312–2316. Mirosevich, J., Bentel, J.M., Zeps, N., Redmond, S.L., D’Antuono, M.F. and Dawkins, H.J. (1999) Androgen receptor expression of proliferating basal and luminal cells in adult murine ventral prostate. J Endocrinol, 162, 341–350. Miyamoto, H., Altuwaijri, S., Cai, Y., Messing, E.M. and Chang, C. (2005) Inhibition of the Akt, cyclooxygenase-2, and matrix metalloproteinase-9 pathways in combination with androgen deprivation therapy: potential therapeutic approaches for prostate cancer. Mol Carcinog, 44, 1–10. Mohler, J.L., Chen, Y., Hamil, K., Hall, S.H., Cidlowski, J.A., Wilson, E.M., French, F.S. and Sar, M. (1996) Androgen and glucocorticoid receptors in the stroma and epithelium of prostatic hyperplasia and carcinoma. Clin Cancer Res, 2, 889–895. Monge, A., Jagla, M., Lapouge, G., Sasorith, S., Cruchant, M., Wurtz, J.M., Jacqmin, D., Bergerat, J.P. and Ceraline, J. (2006) Unfaithfulness and promiscuity of a mutant androgen receptor in a hormone-refractory prostate cancer. Cell Mol Life Sci, 63, 487–497. Nagabhushan, M., Miller, C.M., Pretlow, T.P., Giaconia, J.M., Edgehouse, N.L., Schwartz, S., Kung, H.J., de Vere White, R.W., Gumerlock, P.H., Resnick, M.I., Amini, S.B. and Pretlow, T.G. (1996) CWR22: the first human prostate cancer xenograft with strongly androgendependent and relapsed strains both in vivo and in soft agar. Cancer Res, 56, 3042–3046. Nagy, A. (2000) Cre recombinase: the universal reagent for genome tailoring. Genesis, 26, 99–109. Nelius, T., Filleur, S., Yemelyanov, A., Budunova, I., Shroff, E., Mirochnik, Y., Aurora, A., Veliceasa, D., Xiao, W., Wang, Z. and Volpert, O.V. (2007) Androgen receptor targets NFkappaB and TSP1 to suppress prostate tumor growth in vivo. Int J Cancer, 121, 999–1008. Nelson, J.B. and Carducci, M.A. (2000) Small bioactive peptides and cell surface peptidases in androgen-independent prostate cancer. Cancer Invest, 18, 87–96. Nishimura, K., Ting, H.J., Harada, Y., Tokizane, T., Nonomura, N., Kang, H.Y., Chang, H.C., Yeh, S., Miyamoto, H., Shin, M., Aozasa, K., Okuyama, A. and Chang, C. (2003) Modulation

88

S. Yeh et al.

of androgen receptor transactivation by gelsolin: a newly identified androgen receptor coregulator. Cancer Res, 63, 4888–4894. Niu, Y., Altuwaijr, S., Lai, K.-P., Wu, C.-T., Ricke, W.A., Messing, E.M., Yao, J., Yeh, S. and Chang, C. (2008a) Androgen receptor is a tumor suppressor and stimulator in prostate cancer metastasis. Proc Natl Acad Sci USA, 105, 12182–12187. Niu, Y., Altuwaijri, S., Yeh, S., Lai, K.-P., Yu, S., Chuang, K.-H., Huang, S.-P., Lardy, H. and Chang, C. (2008b) Targeting the stromal androgen receptor in primary prostate tumors at earlier stages. Proc Natl Acad Sci USA, 105, 12188–12193. Papandreou, C.N., Usmani, B., Geng, Y., Bogenrieder, T., Freeman, R., Wilk, S., Finstad, C.L., Reuter, V.E., Powell, C.T., Scheinberg, D., Magill, C., Scher, H.I., Albino, A.P. and Nanus, D. M. (1998) Neutral endopeptidase 24.11 loss in metastatic human prostate cancer contributes to androgen-independent progression. Nat Med, 4, 50–57. Patrawala, L., Calhoun, T., Schneider-Broussard, R., Li, H., Bhatia, B., Tang, S., Reilly, J.G., Chandra, D., Zhou, J., Claypool, K., Coghlan, L. and Tang, D.G. (2006) Highly purified CD44 + prostate cancer cells from xenograft human tumors are enriched in tumorigenic and metastatic progenitor cells. Oncogene, 25, 1696–1708. Pilatus, U., Ackerstaff, E., Artemov, D., Mori, N., Gillies, R.J. and Bhujwalla, Z.M. (2000) Imaging prostate cancer invasion with multi-nuclear magnetic resonance methods: the metabolic Boyden chamber. Neoplasia, 2, 273–279. Pinkas, J. and Teicher, B.A. (2006) TGF-beta in cancer and as a therapeutic target. Biochem Pharmacol, 72, 523–529. Prins, G.S. and Putz, O. (2008) Molecular signaling pathways that regulate prostate gland development. Differentiation, 76, 641–659. Rahman, M., Miyamoto, H. and Chang, C. (2004) Androgen receptor coregulators in prostate cancer: mechanisms and clinical implications. Clin Cancer Res, 10, 2208–2219. Roy, A.K., Lavrovsky, Y., Song, C.S., Chen, S., Jung, M.H., Velu, N.K., Bi, B.Y. and Chatterjee, B. (1999) Regulation of androgen action. Vitam Horm, 55, 309–352. Sadi, M.V., Walsh, P.C. and Barrack, E.R. (1991) Immunohistochemical study of androgen receptors in metastatic prostate cancer. Comparison of receptor content and response to hormonal therapy. Cancer, 67, 3057–3064. Saleem, M., Kweon, M.H., Johnson, J.J., Adhami, V.M., Elcheva, I., Khan, N., Bin Hafeez, B., Bhat, K.M., Sarfaraz, S., Reagan-Shaw, S., Spiegelman, V.S., Suri, V. and Mukhtar, H. (2006) S100A4 accelerates tumorigenesis and invasion of human prostate cancer through the transcriptional regulation of matrix metalloproteinase 9. Proc Natl Acad Sci USA, 103, 14825– 14830. Sharifi, N., Dahut, W.L. and Figg, W.D. (2008) Secondary hormonal therapy for prostate cancer: what lies on the horizon? BJU Int, 101, 271–274. Shi, X.B., Ma, A.H., Xia, L., Kung, H.J. and de Vere White, R.W. (2002) Functional analysis of 44 mutant androgen receptors from human prostate cancer. Cancer Res, 62, 1496–1502. Sugimura, Y., Cunha, G.R. and Donjacour, A.A. (1986a) Morphogenesis of ductal networks in the mouse prostate. Biol Reprod, 34, 961–971. Sugimura, Y., Cunha, G.R. and Donjacour, A.A. (1986b) Morphological and histological study of castration-induced degeneration and androgen-induced regeneration in the mouse prostate. Biol Reprod, 34, 973–983. Tanimura, Y., Kokuryo, T., Tsunoda, N., Yamazaki, Y., Oda, K., Nimura, Y., Naing Mon, N., Huang, P., Nakanuma, Y., Chen, M.F., Jan, Y.Y., Yeh, T.S., Chiu, C.T., Hsieh, L.L. and Hamaguchi, M. (2005) Tumor necrosis factor alpha promotes invasiveness of cholangiocarcinoma cells via its receptor, TNFR2. Cancer Lett, 219, 205–213. Taplin, M.E., Bubley, G.J., Shuster, T.D., Frantz, M.E., Spooner, A.E., Ogata, G.K., Keer, H.N. and Balk, S.P. (1995) Mutation of the androgen-receptor gene in metastatic androgen-independent prostate cancer. N Engl J Med, 332, 1393–1398.

Differential Roles of Androgen Receptor in Prostate Development

89

Tilley, W.D., Buchanan, G., Hickey, T.E. and Bentel, J.M. (1996) Mutations in the androgen receptor gene are associated with progression of human prostate cancer to androgen independence. Clin Cancer Res, 2, 277–285. van der Kwast, T.H., Schalken, J., Ruizeveld de Winter, J.A., van Vroonhoven, C.C., Mulder, E., Boersma, W. and Trapman, J. (1991) Androgen receptors in endocrine-therapy-resistant human prostate cancer. Int J Cancer, 48, 189–193. van Leenders, G.J. and Schalken, J.A. (2003) Epithelial cell differentiation in the human prostate epithelium: implications for the pathogenesis and therapy of prostate cancer. Crit Rev Oncol Hematol, 46 Suppl, S3–S10. van Leenders, G.J., Aalders, T.W., Hulsbergen-van de Kaa, C.A., Ruiter, D.J. and Schalken, J.A. (2001) Expression of basal cell keratins in human prostate cancer metastases and cell lines. J Pathol, 195, 563–570. Webber, M.M., Trakul, N., Thraves, P.S., Bello-DeOcampo, D., Chu, W.W., Storto, P.D., Huard, T.K., Rhim, J.S. and Williams, D.E. (1999) A human prostatic stromal myofibroblast cell line WPMY-1: a model for stromal–epithelial interactions in prostatic neoplasia. Carcinogenesis, 20, 1185–1192. Wu, C.T., Altuwaijri, S., Ricke, W.A., Huang, S.P., Yeh, S., Zhang, C., Niu, Y., Tsai, M.Y. and Chang, C. (2007) Increased prostate cell proliferation and loss of cell differentiation in mice lacking prostate epithelial androgen receptor. Proc Natl Acad Sci USA, 104, 12679–12684. Wu, X., Wu, J., Huang, J., Powell, W.C., Zhang, J., Matusik, R.J., Sangiorgi, F.O., Maxson, R.E., Sucov, H.M. and Roy-Burman, P. (2001) Generation of a prostate epithelial cell-specific Cre transgenic mouse model for tissue-specific gene ablation. Mech Dev, 101, 61–69. Yang, Z., Chang, Y.J., Miyamoto, H., Ni, J., Niu, Y., Chen, Z., Chen, Y.L., Yao, J.L., di Sant’Agnese, P.A. and Chang, C. (2007a) Transgelin functions as a suppressor via inhibition of ARA54-enhanced androgen receptor transactivation and prostate cancer cell growth. Mol Endocrinol, 21, 343–358. Yang, Z., Chang, Y.J., Miyamoto, H., Yeh, S., Yao, J.L., di Sant’Agnese, P.A., Tsai, M.Y. and Chang, C. (2007b) Suppression of androgen receptor transactivation and prostate cancer cell growth by heterogeneous nuclear ribonucleoprotein A1 via interaction with androgen receptor coregulator ARA54. Endocrinology, 148, 1340–1349. Yeh, S., Miyamoto, H. and Chang, C. (1997) Hydroxyflutamide may not always be a pure antiandrogen. Lancet, 349, 852–853. Yeh, S., Miyamoto, H., Shima, H. and Chang, C. (1998) From estrogen to androgen receptor: new pathway for sex hormones in prostate. Proc Natl Acad Sci USA, 95, 5527–5532. Yeh, S., Tsai, M.Y., Xu, Q., Mu, X.M., Lardy, H., Huang, K.E., Lin, H., Yeh, S.D., Altuwaijri, S., Zhou, X., Xing, L., Boyce, B.F., Hung, M.C., Zhang, S., Gan, L. and Chang, C. (2002) Generation and characterization of androgen receptor knockout (ARKO) mice: an in vivo model for the study of androgen functions in selective tissues. Proc Natl Acad Sci USA, 99, 13498–13503. Yeh, S., Hu, Y.C., Wang, P.H., Xie, C., Xu, Q., Tsai, M.Y., Dong, Z., Wang, R.S., Lee, T.H. and Chang, C. (2003) Abnormal mammary gland development and growth retardation in female mice and MCF7 breast cancer cells lacking androgen receptor. J Exp Med, 198, 1899–1908.

Imaging Androgen Receptor Function In Vivo Michael Carey and Lily Wu

Abstract Bioluminescence imaging (BLI) is a facile method for studying androgen receptor (AR) signaling during tumor progression in xenografts and genetic models of prostate cancer in mice. This chapter summarizes work where BLI and positron emission tomography coupled with CT were used to analyze AR-mediated transcriptional activity using gene expression-based imaging approaches.

1 Introduction This chapter will focus on approaches for visualizing androgen receptor (AR) function in preclinical models of prostate cancer (PCa) using bioluminescence imaging (BLI). Two themes that will be discussed include whether AR activates transcription in both the androgen-dependent (AD) as well as the recurrent and hormone refractory forms of PCa in standard tumor models [reviewed in Dehm and Tindall (2007)], and how imaging can be employed to study AR signaling during cancer progression. These are key issues in the PCa field as AR function remains a major target of therapeutic intervention (Taplin 2007). There are numerous methods for visualizing AR location and function. Fluorescent-AR fusion proteins can be employed to directly visualize AR dynamics, subcellular localization, and response to various ligands in cell culture (see Marcelli et al. 2006). AR-responsive reporter genes can also be introduced into animals or xenograft tumors, and transcription can be measured by monitoring the activity of the reporter gene (Wu et al. 2003). The reporter gene can be a fluorescent protein or bioluminescent protein that can be visualized using cryogenically cooled, charge-coupled device (CCD) cameras (Contag et al. 2000; Massoud and Gambhir 2003). Alternately, the reporter protein can be designed to bind a probe molecule that can be visualized in some manner. These methods include probes that can be studied using positron emission tomography (PET) (Gambhir 2002). Finally, AR M. Carey(*) Department of Biological Chemistry, David Geffen School of Medicine at UCLA, Los Angeles, CA 90095, USA, E-mail: [email protected]

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4_4, # Springer Science + Business Media, LLC 2009

91

92

M. Carey, L. Wu

itself is a ligand-binding molecule, and a new generation of PET probes have recently been synthesized that bind directly to AR and permit its localization within small animals and man by PET (Dehdashti et al. 2005; Jacobson et al. 2005; Parent et al. 2006, 2007; Schoder and Larson 2004).

2 Gene Expression-Based Imaging 2.1

Overview

Gene expression-based BLI has emerged as a relatively facile approach for studying signaling pathways during tumor progression in live animals (Contag and Bachmann 2002; Contag et al. 2000; Helms et al. 2006; Massoud and Gambhir 2003). In gene expression-based BLI, a promoter is placed upstream of a bioluminescence reporter gene (i.e. firefly luciferase – FLuc or Renilla luciferase – RLuc) (Fig. 1). The reporter cassette is introduced into tumor cells in a live animal either genetically or using DNA/viral vectors. To measure the bioluminescence signal emitted from the animal, the mouse is sedated and injected with D-luciferin (substrate for FLuc) or coelenterazine (for RLuc), placed in a light tight chamber and

Fig. 1 Bioluminescence imaging of animal models of prostate cancer (PCa). The diagram illustrates the general approach for imaging promoter activity in live animals. A promoter is attached to a bioluminescence reporter gene, FLuc, introduced into animals and imaged in a Xenogen IVIS. The output is a photo of the mouse superimposed with a color pseudoimage representing the emitted light in photons/s/cm2/steradian (See Color Insert)

Imaging Androgen Receptor Function In Vivo

93

imaged using a cryogenically cooled, CCD camera. In many of the currently available devices, several mice can be imaged simultaneously.

2.2

Methodology

A popular and widely used device is the In Vivo Imaging System (IVIS1) manufactured by Caliper Life Sciences (Hopkinton MA). IVIS measures bioluminescence generated by the reaction of luciferase with its substrate or measures fluorescence upon irradiation of animals bearing a fluorescent reporter protein. A computer interprets and integrates the light signals over a few seconds to five minutes acquisition period. A program (Living Image1) then superimposes a color pseudoimage, representing the quantity of photons emitted by the tissue, over a grayscale photograph of the animal. The optical signals over a particular region of interest can be specifically queried. As of this writing, there are several models of IVIS, the most recent of which allow 3D reconstruction of fluorescence and bioluminescence signals. The imaging technology is linear with respect to dose of luciferase and light output. Additionally, the substrate clears from the system quickly – although D-luciferin and coelenterazine display different pharmacokinetics – and the animal can be reinjected subsequently to generate a time course of promoter activity over days or weeks. Indeed, the short half-life of FLuc in cells, 2 h (Ignowski and Schaffer 2004), permits multiple measurements in a single day. The ability to repetitively image a single animal has advantages over traditional animal studies requiring sacrifice of animals at each time point followed by pathological and molecular analyses to measure transcription. In effect, each animal can serve as its own internal control over time. Hence, repetitive imaging significantly reduces the number of animals needed for a statistically meaningful study.

2.3

Drawbacks

There are several drawbacks of the technology. To date, in the most widely used versions of IVIS, the scans are two-dimensional and relatively of low resolution. Hence, the scans do not accurately reveal the depth of the cells emitting the light. The newest models of IVIS (Spectrum1 and the 3D series) have partially overcome that drawback via 3D reconstruction techniques, although the reconstruction does have limitations. The reader should consult the Caliper Life Sciences website for additional details of the technology (www.caliperls.com). Nevertheless, when in vivo studies are accompanied by ex vivo imaging of tissues, immunohistochemistry, and pathological analyses, the precise location of the cells emitting light can be ascertained. This latter set of techniques requires sacrificing the animal, but is usually performed at an endpoint rather than during a time course. A second

94

M. Carey, L. Wu

challenge in BLI is detecting optical signals in dense tissues. Much of the light is scattered or absorbed by blood, tissue, fur, and skin pigmentation. The successful implementation of the technology requires high luciferase levels to image at depths of 1–2 cm (i.e. as in a tumor). Thus, the method functions most efficiently with small animals. Improvements of the technology using red-shifted reporter proteins should decrease some of the light absorption by blood and tissue. The technology was originally validated using potent viral promoter/enhancers, that is, CMV and SV40, which generate robust signals sufficient to image tissues deep within a mouse. However, cellular promoters are typically weaker than viral ones and more difficult to image. Yet cellular promoters that are AR-responsive are more desirable because they can be used to measure the output of specific signaling pathways.

2.4

Reporter Gene Delivery

The methods used to deliver the reporter gene largely depend upon the question being addressed. The use of cells stably transformed with an imaging vector allows the cells to be placed into the animal and tracked. In such cases, the vector can be constitutively expressed via a strong enhancer/promoter like CMV or SV40, or designed to measure particular signaling pathways in the cell. This method can be performed with cell lines or primary cells. In cancer, such approaches can be used to monitor cancer progression, test therapies on xenografts, and track metastatic lesions. A second approach utilizes viral vectors (e.g. adeno-associated virus, adenovirus, and lentiviruses) or encapsulation devices (e.g. nanoparticles) bearing an imaging cassette. These can be injected systemically to allow the vector to be targeted to a wide range of accessible tissues. In this case, it is preferable that the imaging reporter be under a tissue-specific promoter, otherwise the tissue that is best transduced will display the highest signal, which may or may not be desirable depending upon the tropism of the vector. Viral vectors also have an advantage in that their gene transduction capability is sufficiently robust that direct administration of vectors into the animal is feasible. Thus, they can be injected into the bloodstream or directly into a targeted organ or xenograft. Many cancer xenograft models do not grow as cell lines, and thus, direct infection of the tumor with a virus or other agent provides a means to monitor the xenograft tumor directly. Finally, the imaging reporter system can be placed under tissue-specific promoters and employed to generate transgenic mice to study the biology of a particular transcription pathway. This approach can be used to study signaling in an animal, an issue we discuss below with respect to AR. It is particularly useful when the transgenic reporter animal is crossed with a cancer model, that is, to generate a bigenic animal, or with a knockout model to study how loss of a gene affects reporter activity. In this case, the reporter cassette

Imaging Androgen Receptor Function In Vivo

95

can track cancer progression and metastasis in a natural environment and be used to study drugs that intervene at certain stages of cancer. The transgenic approach is important since many xenografts can grow quite rapidly, and it is frequently necessary to euthanize the animal before fully studying the metastatic potential of a tumor. Additionally, the cancer in transgenic models follows a natural progression, which may be desirable in order to study specific regulatory checkpoints in the process. Several of these issues will be explored in following text.

2.5

Reporter Gene Potency

There are two strategies that have been used to augment transcriptional activity for studying the AR signaling pathway in animals by imaging. First, the promoter and enhancer of an androgen-regulated gene are manipulated to artificially augment their activity – this is termed the chimeric or composite enhancer approach. The second approach is based on a concept termed two-step transcriptional amplification (TSTA). A cellular promoter expresses a potent chimeric activator, GAL4VP16. GAL4-VP16 is a fusion of the high-affinity yeast GAL4 DNA-binding domain to the potent herpes simplex virus VP16 activation domain (Emami and Carey 1992; Sadowski et al. 1988). GAL4-VP16 has a unique potency and specificity. No protein exists in mammalian cells that binds to the GAL4 recognition site, and few cellular activators, if any, match the strength of the VP16 activation domain. GAL4-VP16 then binds a GAL4-responsive reporter gene and generates high levels of gene product. The TSTA approach allows weak cellular promoters to be visualized and, hence, expands the repertoire of signaling pathways that can be measured via imaging. The TSTA system is not toxic in animal models allowing for continuous monitoring and the generation of transgenic models. Furthermore, as discussed below, the TSTA approach recapitulates the signal responsiveness of the promoter, permitting detailed investigation of the pathway.

3 Chimeric/Composite Enhancer Approach 3.1

The PSA Promoter and Enhancer

Prostate-specific antigen (PSA) is a kallikrein family serine protease. Many studies have demonstrated the prostate tissue specificity of PSA, and it is a widely used marker for evaluating treatment and progression of cancer (Bok and Small 2002). The PSA gene represents a model for studying AR-mediated gene expression during PCa because it is active in the AD and early recurrent stages of cancer growth. Biochemical and genetic studies have led to the cloning and genetic

96

M. Carey, L. Wu

dissection of the PSA promoter and enhancer by numerous groups with much of the work being completed over a decade ago. Both the enhancer and the promoter display androgen responsiveness consistent with the observation that both contain multiple androgen response elements (AREs), which bind AR. The promoter has been delineated to a 600-bp fragment upstream of the start site. It contains two AREs and other DNA sites important for PSA expression (Cleutjens et al. 1996; Reid et al. 2001; Riegman et al. 1991). The enhancer is centered in a region about 4.2 kb upstream of the transcription start site (Cleutjens et al. 1997; Schuur et al. 1996). The importance of the enhancer is highlighted by the presence of an androgeninducible DNase I hypersensitive site (DHS II) at 4.2 kb (Cleutjens et al. 1997). The DHS II region is within a minimal 440-bp core sequence ( 4,324 to 3,884), which provides strong enhancer activity and contains a high-affinity ARE (AREIII). Analysis of the enhancer by DNase I footprinting and mutagenesis indicated that AREIII is flanked by several lower affinity but functional AREs, all of which contribute to the synergistic activation of PSA gene expression (Huang et al. 1999). The reasons behind the development of an AR-responsive PSA gene imaging system are that (1) PSA is an established, albeit imperfect, marker for early and recurrent PCa; (2) the PSA enhancer and promoter have been incorporated into transgenic mice and shown to maintain tissue specificity; (3) the promoter and enhancer are well characterized as described earlier and hence amenable to manipulations to enhance potency; (4) since the PSA promoter and enhancer are critically dependent on AR function, an imaging system based on the PSA gene would provide a real-time readout of AR activity in vivo.

3.2

Chimeric and Composite Constructs

In a study performed by Wu et al. (2001), the synergistic nature of AR action was employed to improve the activity of the PSA regulatory region (PSE) bearing the larger PSA enhancer region ( 5,322 to 2,855) fused to the proximal promoter ( 541 to +12) (Fig. 2). Chimeric plasmid constructs were generated by three main strategies: (1) insertion of four tandem copies of the PSA promoter-proximal AREI element (ARE4) adjacent to the core enhancer (PBA), (2) duplication of core enhancer region (PBC), or (3) a combination of first and second strategies (PBAC). Some intervening sequences ( 3,744 to 2,855) were removed between the enhancer and promoter (PSE-B series) in all of the constructs. The most potent construct, PBC, was 20-fold more active than the baseline PSE in cell culture and retained activity and specificity when inserted into an adenovirus vector (Ad-PBC-FLuc). Other chimeric approaches have been employed to generate reporter constructs that accurately measure AR function. Latham et al. (2000) employed a larger version of the PSA enhancer in cell culture studies and found that duplication of

Imaging Androgen Receptor Function In Vivo

97

Fig. 2 The chimeric/composite enhancer approach. The diagram shows three chimeric constructs in which the potency of the PSA enhancer promoter was modulated by either addition of four copies of PSA AREI (PBA), duplication of the enhancer (PBC), or addition of four AREs plus duplication of the enhancer (PBAC). These were compared with the starting construct PSE and a version of PSE where additional sequences between the enhancer and promoter were removed (PSE-B). The bar graph shows relative activity of the constructs in LNCaP cells

a 1.2-kb enhancer fragment upstream of the promoter led to a synergistic increase in activity. These investigators inserted the duplicated enhancer construct into adenovirus and demonstrated that reporter gene expression maintained cell selectivity, androgen inducibility, and potency. Additionally, Xie et al. (2001) delimited the regulatory regions of the highly similar prostate-specific human kallikrein 2 gene (hK2) and identified the enhancer and promoter. Like the PSA gene, the hK2 enhancer had a significant effect upon androgen-mediated induction of transcriptional activity and an increase in the number of enhancers led to an increase in the fold induction. The authors generated adenovirus constructs bearing EGFP reporters under either CMV or hK2 regulatory regions. They found that hK2-EGFP virus maintained androgen inducibility and cell type specificity in a panel of cultured cell lines. Moreover, the hK2-EGFP adenovirus maintained tissue specificity when injected into xenograft tumors generated from subcutaneously implanted LNCaP cells versus injection of the virus into control mouse tissues, which included liver and brain (Xie et al. 2001). Finally, Matusik and coworkers performed extensive promoter and enhancer analyses to delimit the native mouse probasin (Pb) enhancer and promoter. They generated constructs with an additional ARE from the proximal Pb promoter (termed ARR in their studies) (Zhang et al. 2000, 2004). The ARR2Pb construct has been used to generate prostate-specific gene expression in several

98

M. Carey, L. Wu

transgenic mouse models. In the cases described above, the chimeric enhancers have been applied to develop noninvasive, gene expression-based, molecular imaging systems that measure AR function.

3.3

Imaging Metastasis with a Viral Chimeric Construct

A study by Wu et al. demonstrated the ability of Ad-PBC-FLuc to interrogate the AR signaling pathway in live animals bearing LAPC4 xenografts (Adams et al. 2002). The LAPC (Los Angeles Prostate Cancer) series of xenografts were derived by transplanting various PCa specimens from patients into the flanks of SCID mice (Klein et al. 1997). As described above, for the BLI (bioluminescence imaging) experiments to be described and illustrated herein, a computer-generated pseudoimage of the measured light is displayed over the grayscale photograph of the mouse. The color images will be shown in the online version of this book, while grayscale images will be displayed in the paper version. The animal results established both the specificity and potency of the PBC chimeric enhancer and validated its use in monitoring cancer progression. The study showed that whereas Ad-CMV-FLuc is constitutively active in tumors, liver, and other tissues, Ad-PBC-FLuc displayed prostate tumor specificity in both AD and AI (recurrent and hormone refractory) xenografts. The study also demonstrated that the adenovirus-based imaging cassette could detect distal metastatic lesions in the lung and spine, a result confirmed by immunohistochemical staining with antibodies directed against human cytokeratin (Fig. 3). The virus also detects lung metastases upon direct injection via tail vein. The coxsackie adenovirus receptor is overexpressed in advanced PCa, particularly in metastatic lesions. Although the reason for coxsackie adenovirus receptor overexpression is unknown, McCormick has suggested that this property of advanced PCa may make metastases particularly amenable to adenovector-based imaging and gene therapy (Rauen et al. 2002). Additionally, adenovirus displays lymphotropic properties allowing it to target a major route of metastasis (see below). The imaging results agreed well with other studies, which examined AR expression and function in xenografts and patient samples of advanced cancer (Gregory et al. 2001a, 2001b). In summary, the work established the ‘‘proof of principle’’ that a mechanistic approach of augmenting AR-dependent PSA expression could have direct application in imaging xenograft models of cancer. The capabilities of the system could be further improved because the chimeric enhancers, although more active than native PSA regulatory regions, were still transcriptionally weaker (20-fold) than Ad-CMV-FLuc in tumors. Ad-CMV has been employed as a benchmark in the imaging and gene therapy arenas due to its transcriptional potency, but it lacks tissue specificity (constitutive gene therapy has been used in clinics without clear success). A series of studies were initiated to improve the activity of the PSA enhancers, while maintaining their cell and tissue specificity. This approach led to the development of the TSTA imaging system.

Imaging Androgen Receptor Function In Vivo

99

Fig. 3 Imaging of metastasis. The figure shows a time course where AdPBC-FLuc is injected into the LAPC4 xenograft tumor but over time can detect distal metastases in the spine of mouse 1 and lungs of mouse 2 and mouse 3. Ex vivo imaging of the spine from mouse 1 is shown after sacrifice on day 21. To the right of the images are low- and high-resolution immunohistochemical analyses of the spine with an antibody against human cytokeratin, which was used to confirm the tumor origin of the metastatic lesions. Also shown is a confocal image of immunofluorescent staining of lung of mouse 2. It revealed a cluster of human cytokeratin+ metastatic cells and lectin stained blood vessels (See Color Insert)

4 The TSTA Imaging System 4.1

Cell Culture Studies

In the initial tests of the TSTA imaging system (Fig. 4), the baseline PSA regulatory region, PSE, was designed to express GAL4-VP16 (Emami and Carey 1992). GAL4-VP16 then activated high levels of reporter gene expression. The TSTA system was used to amplify expression of both FLuc and a PET reporter gene – mutant herpes simplex virus type 1 thymidine kinase (HSV1-sr39tk) – in LNCaP cells (Iyer et al. 2001). The TSTA expression studies demonstrated 50-fold (FLuc) and 12-fold (HSV1-sr39tk) enhancement over the conventional one-step vectors, which expressed FLuc and HSV1-sr39tk directly from the PSA enhancer–promoter, PSE. The TSTA system was observed to retain cell selectivity by analysis of a panel of cultured cell lines. Additionally, CCD imaging was employed to visualize the amplified FLuc expression in living mice implanted with transfected LNCaP cells. These imaging experiments revealed a significant gain in signal when comparing the TSTA system versus the one-step system (PSE-FLuc).

100

M. Carey, L. Wu

Fig. 4 The TSTA scheme. In the presence of AR and DHT, the PSA promoter/enhancer drives expression of the potent activator GAL4-VP16. GAL4-VP16 then binds to multiple GAL4 sites positioned upstream of a promoter expressing luciferase. This scheme can amplify luciferase up to1000-fold

Fig. 5 Creating a titratable TSTA system. The PSE or PBC constructs were used to express GAL4-VP16 derivatives fused to one, two, or four copies of the N-terminal VP16 activation domains. The activities of the reporters were measured on GAL4 reporters bearing either one, two, or five sites upstream of the adenovirus E4T promoter. By varying the activity of both the promoter and effector proteins, the activity of the TSTA system can be carefully modulated. [Figure adapted from Zhang et al. (2002)]

To further refine the system, the technologies from Emami and Carey (1992), Iyer et al. (2001), and Wu et al. (2001) were merged to develop a titratable, prostatespecific, androgen-responsive system for imaging (Zhang et al. 2002). The ‘‘chimeric TSTA’’ system employed normal and chimeric PSA regulatory regions (Wu et al. 2001) to express GAL4 derivatives fused to 1, 2, or 4 VP16 activation domains (GAL4-VP1, -VP2, and -VP4) (Emami and Carey 1992) (Fig. 5). The resulting activators were cotransfected into a panel of cell lines with another

Imaging Androgen Receptor Function In Vivo

101

Fig. 6 Results from titratable system in LNCaP cells. This bar graph shows the results of an LNCaP transfection experiment, in the presence and absence of R1881, comparing the various one-step and two-step vectors to the benchmark activity of the CMV enhancer/promoter. [Figure adapted from Zhang et al. (2002)] (See color Insert)

plasmid containing reporter templates bearing 1, 2, or 5 GAL4-binding sites (G1, G2, and G5) upstream of the adenovirus E4 promoter (E4T) driving FLuc (Emami and Carey 1992). Activity was monitored in parallel via FLuc assays on transfected cell extracts (via luminometry) and in live nude mice implanted with transfected cells (via the IVIS). Luciferase activities were relatively comparable in the luminometer and IVIS device. FLuc expression in PCa cells could be varied over an 800-fold range using the one- and two-step systems (Fig. 6). The optimal combination of components in LNCaP cells was shown to be the PBC regulatory region driving GAL4-VP2 with a separate plasmid containing the FLuc reporter bearing 5 GAL4-binding sites upstream of the adenovirus E4 core promoter. Remarkably, a single plasmid containing the optimized components expressed FLuc at higher levels than CMV-FLuc. This dramatic increase in activity, while maintaining specificity, again highlights the advantages of applying basic transcription principles toward the cancer imaging problem. On the basis of this result, the optimal TSTA construct was inserted into a replication-defective E1/E3deleted adenovirus and used in cell culture and animal studies (Fig. 7). Note that other variations of the TSTA system could be used. The PBC-GAL4-VP4 variant of TSTA is extremely potent, and the baseline expression in absence of ligand (Fig. 6) likely represents the ability of the system to detect residual androgen in the culture

102

M. Carey, L. Wu

Fig. 7 The optimal TSTA system inserted into adenovirus. The optimal TSTA construct was inserted into an E1/E3-deleted adenovirus. The figure shows the basic organization of the functional TSTA modules within adenovirus (See Color Insert)

medium. The ligand-inducible level is lower than with PBC-GAL4-VP2, probably because excess production of GAL4-VP4 is toxic. However, in principle, the system may be more sensitive in detecting trace levels of androgen in dedifferentiating tumors.

5 Imaging AR Function During PCa Progression in Xenografts Using TSTA 5.1

TSTA Animal Studies

An advantage of the AdTSTA approach is the ability to inject the virus into any xenograft animal without having to establish cell lines. Many xenografts are derived directly from patient tumors and do not exist as cell lines. The TSTA adenovirus was used to image human xenografts in SCID mice to address several questions about AR function in recurrent cancer (Zhang et al. 2003). As this study was one of the first to comprehensively use imaging to study the role of AR in PCa progression, it was essential to provide a correlation between the imaging and biology of AR function. This required a combination of in vivo, pathological, and molecular analyses. Thus, the CCD signals were compared to serum PSA levels and tumor AR levels. Immunohistochemistry was used to probe the localization of

Imaging Androgen Receptor Function In Vivo

103

Fig. 8 Androgen deprivation is specific to AdTSTA. A total of 107 pfu of AdCMV-FLuc or AdTSTA-FLuc were injected into LAPC9 tumors. A baseline image was taken on day 5 and the mice were castrated and imaged on day 10. AdCMV signal decreases less than 2-fold while AdTSTA decreases more than 20-fold. [Figure adapted from Zhang et al. (2003)] (See Color Insert)

AR in tumor cells, and chromatin immunoprecipitation (ChIP) was used to probe the binding of AR, pol II, and other factors to the endogenous PSA gene in tumors. Figure 8 shows an experiment where AdTSTA was injected into 0.5 cm tumors (diameter by caliper measurement) of male SCID mice bearing LAPC9 xenografts (Klein et al. 1997). LAPC9 is derived from a bone metastasis. It contains a wildtype AR gene and secretes PSA. LAPC9 xenograft tumors are unusual in that they do not regress substantially upon castration and eventually transition into a recurrent state. Five days after virus injection the tumors were imaged, and the mice were castrated. By the 10-day time point, the imaging signal decreased 20-fold in the castrate animals (Fig. 8) but was unchanged in noncastrated cohorts (not shown). The data strongly support the idea that the system is measuring androgen signaling because injection of an adenovirus bearing CMV-FLuc (AdCMV-FLuc) into the tumor revealed that upon castration there is little decrease in the luciferase signal versus the result observed with AdTSTA-FLuc (Fig. 8). It should be pointed out that unlike LAPC9, many xenografts derived from cell lines or true AD xenografts such as CWR22 tend to regress rapidly upon castration, making the aforementioned experiment very difficult to perform. Another advantage of the LAPC9 model is that it rapidly progresses to the recurrent state in animals. Thus, the TSTA system can be employed to monitor this transition. As shown in Fig. 9, the AdTSTA system detects the decrease in signal upon castration and the increased signal upon transition of the tumor to the recurrent state. In this experiment, a baseline image was taken from a single animal on day 4, after which the animal was castrated. The figure shows the decrease in optical signal several days later and then the rise again as the tumor transitions to a recurrent form within 2 weeks after castration. Note that the optical signal decreases and then rises in parallel with PSA. Additionally, the optical signal is more sensitive than PSA to castration, probably because of the short half-life of FLuc in live cells.

104

M. Carey, L. Wu

Fig. 9 Transition to the recurrent state. This figure shows the progression of the tumor signal in a single mouse as it transits from AD to the castrated state (ADc) and finally to recurrent cancer (AI) over a period of 3 weeks. Luciferase activity is in photons/s/cm2/steradian 106. PSAs are the serum PSA levels at the various stages in this same individual (mean of three values). The immunohistograms are stained with AR antibodies and show the localization of AR in animals at similar stages in cancer. The immunoblots of AR were normalized to actin and obtained from tumors at similar stages to the animal shown. [Figure adapted from Zhang et al. (2003)]. Note that the term AI refers to animals where the tumor grows after removal of the major source of androgen (i.e. the testis). The caveat is that androgen may be synthesized intratumorally using adrenal androgens. Thus, as discussed by the editors at the beginning of this book, the proper term is recurrent. This caveat applies to all figures shown here where AI is used in the label (See Color Insert)

The figure also shows representative AR nuclear localization in xenograft tumors obtained at different stages of cancer from different mice. AR was localized to the nucleus in AD animals, diffused between the cytoplasm and nucleus after castration, and reentered the nucleus in the recurrent stage. Additionally, the tumor levels of AR parallel its activity in the imaging assay. Collectively, the study argues that AR activity recovers during the transition to recurrent cancer and shows that the imaging signal correlates with biological benchmarks of AR function. In a separate study, the specificity of the virus was tested in a panel of ARcontaining and -lacking cell lines that have been classified as either AD or hormone refractory (AI). The TSTA system displayed robust activity in AD cell lines LNCaP and LAPC4, but was relatively inactive in PCa cell lines lacking AR such as PC3 and DU145. In contrast, the TSTA virus displayed robust androgen responsiveness in the hormone refractory, AR-positive cell line MDA PCa 2b. Indeed, in vivo, the

Imaging Androgen Receptor Function In Vivo

105

TSTA system seemed to display more activity in LAPC9 AI versus AD tumors, suggesting that it may respond to as yet uncharacterized aspects of AR function in recurrent tumors (Sato et al. 2005). This enhanced activity may be related to both the presence of intratumoral androgens and the hypersensitivity of AR to such androgens in recurrent cancer (Gregory et al. 2001b; Mohler et al. 2004). The data collectively argue that AR signaling remains active in recurrent cancer xenografts.

5.2

Correlating TSTA Activity with AR Binding to the PSA Enhancer in Tumors

A key issue was whether the activity of AR reported by the imaging constructs correlated with the actual binding of AR to its sites within genomic DNA (Zhang et al. 2003). Figure 10 shows an experiment, where AR binding to the PSA enhancer and promoter within the xenograft tumor was measured by ChIP. The multiplex nature of the ChIP PCR (four sets of primers in any given reaction) provided internal controls for any given sample. Thus, the middle regions between the promoter and the enhancer, as well as exon 5, were used as negative controls for AR binding in each sample. The figure shows that AR bound weakly to the

Fig. 10 Chromatin IP of tumor samples. Tumors from AD, AI, and ADc xenografts were minced and treated with formaldehyde. Formaldehyde cross-linked and sheared chromatin was immunoprecipitated with AR antibodies or IgG. The coprecipitated DNA was amplified with primer sets encompassing the PSA enhancer, a middle intervening region, the proximal promoter, and downstream exon 5. The multiplex DNA products were fractionated on polyacrylamide gels and autoradiographed. Input are input samples added to the IP reaction. A graph of a quantitative PCR experiment on the exact same samples using primers against the promoter is shown. [Figure adapted from Zhang et al. (2003)]

106

M. Carey, L. Wu

promoter and strongly to the enhancer in AD tumors, decreased upon castration, but bound again in the AI tumor. The result was confirmed by quantitative PCR. The data demonstrate that AR is bound to DNA and activates transcription in a recurrent cancer model, and that the imaging signals accurately report the binding status of AR in vivo.

5.3

TSTA Lentiviruses

A similar study that used direct intratumoral injection of lentiviral versions of TSTA generated very similar results to AdTSTA in monitoring tumor growth (Iyer et al. 2004). Indeed, the ability of lentivirus to stably transduce the tissues allowed visualization of the mouse prostate for up to 3 months postorthotopic injection. Moreover, a subsequent study employing systemic intravenous injection of lentivirus revealed that the virus could stably transduce LAPC9 xenografts within 4 days, while maintaining expression for up to 3 weeks (Iyer et al. 2006). Small amounts of background signal were observed in the spleen and liver, and these signals correlated with increased uptake of the virus as measured by qPCR. The background expression may result from basal expression of TSTA due to the lentivirus regulatory elements or may be due to an inherent background in the TSTA system, which at high MOI (of either adenovirus or lentivirus) can generate aberrant signals in a relatively nontissue-specific manner. The background signal issue is addressed in a later section. Nevertheless, this study showed conclusively that the integrating lentivirus could stably deliver the PSA promoter-based TSTA imaging reporter system into a tumor in a live animal and allow for persistent imaging of AR function.

5.4

Detecting the Effects of PCa Therapy by Imaging in Animal Models

A major advantage of the TSTA imaging system is its ability to directly probe AR function during drug treatment. A good example is the effect of flutamide, a nonsteroidal, antiandrogen prodrug (Ilagan et al. 2005). Figure 11 shows that flutamide (60 mg/kg/day) pellets implanted into the mice greatly decrease the imaging signal in AD but not in AI tumors. The inability of antiandrogen therapy to work in AI cancer contributes significantly to PCa mortality. The data support the ability of an AR-dependent imaging system to measure specific antiandrogen drug effects in a xenograft model. Because the imaging is performed repetitively, it permits measurement of AR signaling cues and their response to inhibitors in live animals over the course of a treatment without sacrificing animals at various points to obtain the information.

Imaging Androgen Receptor Function In Vivo

107

Fig. 11 Effect of flutamide on AD and AI xenografts. AdTSTA was injected directly into AD and AI tumors. On day 3, a baseline image was taken and either flutamide or placebo pellets were implanted onto the backs of the mice. Only the flutamide is shown as placebo displayed no changes. The mice were then imaged every 3 days. The image at the endpoint on day 18 is shown. [Figure adapted from Ilagan et al. (2005)] (See Color Insert)

The TSTA system can also be utilized to deliver PET reporter genes. One such gene, sr39tk, also serves as a potential therapeutic gene. In a recent study, AdTSTA-sr39tk was used in a PET imaging-guided therapy in a preclinical model with mice bearing AR- and PSA-positive LAPC4 tumors (Johnson et al. 2005). The study exploited the dual capability of HSV-sr39tk as a PET reporter gene and as a prodrug (ganciclovir)-activated suicide gene (Fig. 12). 18F-FHBGPET was used to monitor the magnitude and location of vector-mediated HSVsr39tk gene expression prior to and after ganciclovir treatment. The suicide therapy with the prostate-specific AdTSTA-sr39tk was compared to that with the constitutive AdCMV-sr39tk. AdTSTA-sr39tk was superior to AdCMV-sr39tk in that it achieved effective tumor destruction while preventing system toxicity (i.e. liver damage). Tumor-directed injection of AdCMV-sr39tk resulted in leakage of the vector into the systemic circulation and liver transduction. Hepatic cellular damage was observed in the AdCMV-sr39tk-treated animals but not in the AdTSTA-sr39tktreated cohorts. In a subsequent study which exploited the lymphotropic features of adenovirus, the AdTSTA-sr39tk and AdTSTA-FLuc viruses were employed to

108

M. Carey, L. Wu

Fig. 12 MicroPET/CT imaging of prostate cancer (PCa) gene therapy using TSTA. A comparison of AdCMV-sr39tk and AdTSTA-sr39tk in LAPC4 xenografts by PET. The scheme indicates treatment time course in days which includes intratumoral virus injection, followed by baseline PET measurement (preRx), followed by treatment with Ganciclovir, and followed by an endpoint PET measurement (postRx). The location of the tumor and liver signals are indicated. The PET signal in the abdomen area of the AdCMV-sr39tk-injected animal is due to accumulation of 18 F-FHBG [9-(4-[18F]-fluoro-3-hydroxymethylbutyl)guanine] after leakage of AdTSTA-sr39tk from the tumor and constitutive expression of the PET reporter in liver. H&E and Tunel staining indicate apoptosis in treated tumors and micrographs show liver damage in the AdCMV-sr39tktreated animal by Tunel. [Figure adapted from Iyer et al. (2005b)] (See Color Insert)

detect sentinel lymph node metastases in xenografts using MicroPET (Fig. 13; Burton et al. 2008). Macroscopic and occult nodal lesions from both subcutaneous and orthotopic xenograft tumor models were visualized. These findings also point out that PSA expression and AR function remain active in nodal metastasis in the LAPC-9 model and correlates with findings in clinical scenarios as determined by rapid autopsy (Roudier et al. 2003; Shah et al. 2004). These data raise the possibility of combining specific imaging of the AR pathway in metastatic lesions with therapy in a clinical environment.

5.5

Modifications of AdTSTA

Both an advantage and a drawback of the TSTA strategy is that the close head-tohead orientation of the PBC-GAL4-VP2 gene and the GAL4-responsive FLuc reporter can generate a feed forward loop, where small amounts of AR signaling

Imaging Androgen Receptor Function In Vivo

109

Fig. 13 The TSTA system detects lymph node metastases. AdTSTA-sr39tk or AdTSTA-FLuc was injected via the paw to detect lymph node metastases resulting from subcutaneously implanted LAPC9 xenografts. The figure shows images of axillary lymph node metastases as detected by bioluminescence (BLI) upon injection of luciferin, and MicroPET after injection of 18F-FHBG. The accompanying panels show gross pathological and microscopic evaluation of the lymph nodes (See Color Insert)

generate GAL4-VP16, which further increases the activity of the potent PBC promoter/enhancer. An advantage is enhanced sensitivity, while a disadvantage is that the system is possibly overly sensitive to very small changes in activity that will be driven forward by an autoregulatory loop. Indeed, high MOI with adenoviral and lentiviral TSTA vectors generates activity in nonprostate cell lines. Studies have shown that such background can be lowered when the PBC-GAL4-VP2 effector and GAL4-responsive FLuc reporter are placed in separate adenoviruses (Sato et al. 2003). However, simultaneous delivery of two vectors into one cell is very inefficient in vivo, making such an approach unfeasible when designing strategies for targeted gene therapy or systemic injection of imaging vectors. To overcome this limitation, the effector and reporter were placed at separate locations within the adenovirus genome. In the original TSTA configuration, the effector and reporter were cloned head to head in the E1 portion of the genome. In the optimized versions, the effector (PBC-GAL4-VP2) is placed in E3, and the reporter or therapeutic gene (G5E4FLuc or G5E4sr39tk) is placed in E1, 25 kb away (Sato et al. 2008). This modified TSTA system displays greater cell selectivity and more

110

M. Carey, L. Wu

robust androgen induction. The E1/E3 showed significant selectivity and little background in lung tissues of mice upon systemic injection.

5.6

Imaging AR Activity in Combination with MAPK

The advantages of imaging are clearly important when attempting to visualize a key signaling event that may drive the tumor into the recurrent state. To this point we have focused on AR signaling. However, one advantage of the TSTA system is its binary nature. Both the tissue-specific promoter as well as the GAL4-derived activator can be changed to alter the properties of the system. It is possible to modify the system to simultaneously detect both AR and mitogen-activated kinase (MAPK) signals. Elevated MAPK has been proposed to contribute to the transition to recurrent PCa (Gioeli et al. 1999; Weber and Gioeli 2004). The system was modified to detect MAPK by utilizing the serum response transcription factor Elk1, which contains three functional domains – the N-terminal and central domains mediate DNA-binding and the C-terminal domain activates transcription (Cruzalegui et al. 1999; Marais et al. 1993). The activation domain from Elk1 has been shown to function when fused to the GAL4 DNA-binding domain (Hexdall and Zheng 2001). GAL4-Elk1 responds to serum stimulation and to MEKK, which activates ERK. ERK, in turn, phosphorylates multiple sites in the Elk-1 C-terminal region. Upon phosphorylation, Elk-1 becomes transcriptionally active (Fig. 14). The TSTA-MAPK system replaces GAL4-VP16 with GAL4-Elk1 (Ilagan et al. 2006). This binary system has the unique ability to measure the oncogenic potential of PCa where both AR and MAPK are active. When tested in cell culture, the system responds synergistically to the combination of R1881 to induce GAL4-Elk1 and EGF or dominant active MEKK to activate the MAPK pathway. As predicted, the signals are inhibited by both EGF receptor and AR inhibitors. In the LAPC9 xenograft model, the system detects systemic injection of EGF within 4 h of treatment (Fig. 14). Similarly, the system detects the much higher MAPK activity found in the CWR22 PCa xenografts versus the LAPC9 model (Ilagan et al. 2006). Since activation domains of many transcription factors respond selectively to extracellular signals and kinase cascades, it is possible that varying the promoter and the GAL4-fusion protein could permit cell-specific measurement of many different signaling pathways.

6 Imaging AR Function in Transgenic Mouse Models The use of transgenic mice for imaging of AR function in combination with other transgenic mice that possess the ability to develop prostate malignancies (i.e. transgenic adenocarcinoma mouse prostate – TRAMP) allows for the noninvasive, long-term monitoring of PCa progression and metastasis. The strategies developed may also be useful for transgenic research in general by allowing for amplified tissue-specific gene expression.

Imaging Androgen Receptor Function In Vivo

111

Fig. 14 The GAL4-Elk1 system to measure MAPK. The scheme shows that GAL4-Elk1 is cloned in place of GAL4-VP16. In the presence of AR, GAL4-Elk1 is synthesized in prostate cancer cells. GAL4-Elk1 is transcriptionally inactive until it is phosphorylated by ERK1/2. Phosphorylated GAL4-Elk activates transcription of the GAL4-responsive FLuc reporter gene. In the imaging experiment below, LAPC9 xenografts, which have low levels of active MAPK, were injected with AdTSTA-Elk1 and imaged before and 4 h after systemic administration of EGF. The result demonstrates that the TSTA-Elk1 system responds to MAPK activation in live animals. [Figure adapted from Ilagan et al. (2006)] (See Color Insert)

6.1

The TSTA Transgenic

A transgenic mouse was generated to determine the selectivity of TSTA in an animal (Iyer et al. 2005a). One potential drawback of TSTA was that GAL4-VP16 might be toxic to the animals. However, transgenic mice displayed normal physical characteristics and developmental behavior, indicating that the high level of GAL4-VP16mediated expression is tolerated. Detailed pathological examination of the prostates from transgenic and wild-type animals revealed that the structures of the lobes of the transgenic prostate were normal and there was no microscopic evidence of tissue or gland abnormality. The results further showed that the bioluminescence signal was largely, although not exclusively, localized to the lower abdominal region (Fig. 15). It should be noted that it is difficult to use BLI to decisively identify the exact tissue origin of a signal unless it is a xenograft implanted subcutaneously. This drawback

112

M. Carey, L. Wu

Fig. 15 Bioluminescence imaging of TSTA transgenic animals. (a) Male and female littermates bearing the TSTA-FLuc system were imaged after injection with D-luciferin. (b) The effects of castration on TSTA transgenics. (c) Harvesting and ex vivo imaging of select individual organs demonstrate prostate specificity of the signal. Note that in one of the transgenic lines there was a weak signal observed in heart probably due to an integration site effect (See Color Insert)

requires that, at early stages in an imaging study, the animal has to be sacrificed and the various organs imaged ex vivo in a petri dish (Fig. 15). Nevertheless, once a tissue origin is determined, the subsequent studies can utilize that knowledge to evaluate the properties of the imaging system in the live animal. In the case of TSTA, the system was male-specific and AR-dependent, that is blocking androgen availability by castration led to downregulation of FLuc expression in the prostate (Fig. 15). These findings suggest that the GAL4-VP16 transactivator can be used to amplify reporter gene expression from a relatively weak promoter in a transgenic mouse model and raise the possibility that the approach can be applied to study other pathways.

6.2

The ARR2Pb Transgenic

In another example of how chimeric constructs can be used to facilitate imaging, Sawyers and coworkers developed a transgenic mouse in which FLuc was driven directly by Matusik’s ARR2Pb construct (Ellwood-Yen et al. 2006). These mice expressed robust levels of luciferase as measured by imaging. Analysis of individual tissues after dissection revealed prominent expression of FLuc in the

Imaging Androgen Receptor Function In Vivo

113

dorsolateral and ventral prostate consistent with Matusik’s earlier characterization of an ARR2PbCAT transgenic model (Ellwood-Yen et al. 2006). Additionally, imaging of FLuc expression showed that it responded appropriately to castration as well as systemic treatment with the antiandrogen bicalutamide. A similar animal was developed by Spencer and coworkers and will be described below.

6.3

The GAL4-AR Transgenic

O’Malley and coworkers developed a special approach that allows AR activity to be measured directly in any tissue where the transgene cassette is active (Ye et al. 2005). The mouse contains a GAL4-AR fusion protein coupled with a GAL4responsive Luciferase reporter cassette. The GAL4-AR was constructed by replacing the DNA-binding domain of AR with that of GAL4 in a Bac construct bearing the native AR locus. The construct contained a 15-kb region upstream of the gene, most of the 195-kb gene (except where the DNA-binding domain had been swapped) and a 10-kb region downstream of the gene. In vivo BLI showed that the reporter gene was expressed largely in the testis, although pathological and immunohistochemical analyses showed that it was also expressed in skin and cerebellum. The reason for the lack of prostate-specific expression is unknown but the authors speculated that certain transcription regulatory regions might be absent from the Bac clone or were present in the intergenic region between exons 2 and 3, which was swapped out. Nevertheless, the model allowed the authors to cross the animals with TIF2 heterozygous knockout mice and show that this particular AR coactivator was important for AR function in the Sertoli cells of the testis. This is an exciting approach and emphasizes how imaging can be combined with knockout models of mice to study very specific aspects of transcription factor signaling pathways.

6.4

6.4.1

Imaging AR Function During Metastasis and Cancer Progression The sPSA Model

Several other transgenic models have been generated and shown to detect metastatic lesions. Chung’s laboratory developed the sPSA model (Hsieh et al. 2005). The sPSA construct contains the core PSA enhancer linked to a minimal region containing the PSA TATA and surrounding sequences. The sPSA FLuc construct displayed strong ventral prostate expression with lower levels in the dorsolateral prostate. Expression was also noted in the cauda epididymis and a small background level of expression was noted in bone marrow of both male and female littermates. Similar

114

M. Carey, L. Wu

to the TSTA and ARR2Pb models, the sPSA model responded well to castration with optical signals reaching a baseline 7 days postcastration in 10-week male mice. In a follow-up study, Chung’s group crossed the sPSA transgenic with TRAMP mice to determine the utility of in vivo imaging for monitoring AR activity during tumor development and metastasis (Hsieh et al. 2007). TRAMP mice contain a minimal AR-responsive Pb promoter linked to SV40 T antigen (Tg). The promoter expresses Tg in the dorsolateral and ventral lobes of the prostate (Gingrich et al. 1996). The mice develop adenocarcinoma and display a pattern of metastasis that parallels in several respects the pattern observed in human PCa. Chung’s group combined imaging with pathological examination to identify two groups of mice with different degrees of pathology. In group 1, the signal intensity was restricted to the prostate over 24-week course of the study and correlated with PIN upon gross pathological examination. Group 2 showed strong signals in the prostate and lower abdomen that spread to the groin lymph node by week 16. Interestingly, the signal appeared to decrease substantially by week 20. Histopathological analysis indicated that the majority of the tumors in group 2 were poorly differentiated and, as cancer progressed, AR expression was apparently lost, which was consistent with the decrease in luminescence signal, although, paradoxically, immunohistochemical analyses showed that the cancer expressed Tg from the AR-dependent Pb promoter. With respect to the lesions, molecular analysis of the lymph node metastasis confirmed the presence of luciferase, which disappeared concomitant with increased dedifferentiation of the tumor cells. Bone metastases were also observed in some animals. This particular study emphasizes the need to combine in vivo imaging with ex vivo imaging and pathological analyses to fully exploit the benefits of imaging metastatic lesions in animal models of cancer. It would be very difficult to conclusively distinguish groin metastasis from the signal in the prostate without ex vivo analysis. The study also reveals an interesting property of the TRAMP model. The AR signal decreases rather rapidly and unexpectedly, suggesting that AR function is diminishing rapidly during cancer progression and cellular dedifferentiation.

6.4.2

The EZC Models

Spencer and coworkers performed perhaps the most thorough studies on AR imaging in transgenic mice (Seethammagari et al. 2006). Spencer’s group developed transgenic lines using three different chimeric constructs based on the triplicated enhancer of the human kallikrein (hK2-E3-P) gene, the duplicated enhancer of PSA (PSA-E2-P), and the ARR2Pb construct. They termed these transgenic models EZC1, EZC2, and EZC3. All three models express luciferase in the prostate as determined by in vivo and ex vivo imaging. EZC2 and 3 display much lower expression in other tissues versus EZC1. In vivo imaging revealed high levels of expression in the lower abdomen, which correlated largely with ventral prostate expression when analyzed ex vivo. Minor levels of expression (0.1–2% vs ventral prostate) were detected in various tissues such as dorsolateral prostate, epididymis,

Imaging Androgen Receptor Function In Vivo

115

and testes. These values contrasted with the higher background signals and wider tissue distribution observed in the EZC1 model. Interestingly, the authors were able to observe weak signals in utero of developing male and female animals. Additionally, the signals allowed in vivo and ex vivo analyses of AR function in numerous tissues during early development up to and past puberty. The luminescence signals like all other models were highly responsive to castration in adult males and recovered in castrated males implanted with testosterone pellets. The signals also responded well to treatments that lower circulating levels of testosterone including a GnRH modulator. The study established the utility of the EZC models in monitoring AR during cancer progression and metastasis through the generation of bigenic cancer models. In one model, EZC was crossed with the JOCK-1 model, which employs a prostatetargeted form of FGFR1 that is attached to a ligand-inducible dimerization domain. Addition of the ligand causes receptor dimerization and enhanced tyrosine kinase activity. This system generates hyperplasia at 2 weeks, PIN at 24 weeks, and widespread adenocarcinoma by 40 weeks. Interestingly, despite tumor progression, the luminescence signal intensity did not increase concomitantly. Indeed, AR activity decreased on a per cell basis, and the authors speculated that this may be due to numerous factors including dedifferentiation, hypoxia, tumor burden, etc. Similarly, studies in the TRAMP model, which developed adenocarcinomas by 12 weeks and metastases by 12–18 weeks, did not reveal significant increases in signal intensity and in some cases showed decreases despite the obvious progression of the cancer. Analysis of the EZC3 mice, which displayed the least amount of extraprostatic background signals, allowed facile detection of metastases. Precise localization of the lesions is quite difficult in vivo but subsequent dissection and ex vivo analysis of the signals revealed high efficiency of metastasis to the pancreas, lymph nodes, kidney, adrenal glands, liver, spleen, and salivary glands. In summary, these animals serve as excellent models to systematically analyze various PCa treatments in a preclinical setting.

7 Conclusions and Future Prospects The emphasis on understanding AR-responsive promoters and signaling pathways in the 1990s has led to the development of tissue-specific vectors that allow the detection of AR-dependent transcriptional activity via BLI. Numerous investigators in the AR field have demonstrated the utility of BLI in understanding AR signaling in xenograft tumors (Adams et al. 2002; Zhang et al. 2003), during mouse development (Ellwood-Yen et al. 2006; Hsieh et al. 2005; Iyer et al. 2005a; Seethammagari et al. 2006) and in mouse bigenic models of PCa (Hsieh et al. 2007; Seethammagari et al. 2006). Studies on subcutaneous and orthotopic PCA xenografts using viral transduction have revealed that AR is indeed transcriptionally active in the AD, recurrent, and metastatic models of PCa (Sato et al. 2005; Zhang et al. 2003). Indeed, in some models, AR seems to be hyperactive in recurrent cancer. Systemic

116

M. Carey, L. Wu

injection of viral vectors bearing imaging cassettes has also revealed the ability of these vectors to infect xenograft tumors and metastatic lesions resulting from the xenograft (Adams et al. 2002; Iyer et al. 2006; Sato et al. 2005). Studies on transgenic mice bearing AR-responsive imaging cassettes have shown that AR activity can be tracked in utero and throughout development providing new clues to AR signaling during development (Hsieh et al. 2007; Seethammagari et al. 2006). The models have been validated in terms of their tissue specificity, which varies by the model but is mostly prostate specific. More importantly, the ARmediated light signals in the mouse models respond appropriately to various treatments such as castration and drugs, which include GnRH modulators and nonsteroidal antiandrogens. Studies of bigenic models, particularly TRAMP mice, have revealed the loss of AR function during cancer progression and demonstrated the ability to visualize metastasis (Hsieh et al. 2007; Seethammagari et al. 2006). These studies have also demonstrated the limitations of the imaging methodology and the need for ex vivo imaging and detailed pathological analyses in order to specifically localize the lesions and track the route of dissemination. Future studies will incorporate the various imaging systems to study the effects of new drugs and to study in more detail the role of AR during cancer progression and dedifferentiation. New reporter genes such as red-shifted reporter proteins are already allowing higher resolution localization of light-emitting cells. PET coupled with CT in the micro-PET scanners allows significantly higher resolution than bioluminescence in animal models but is more suited for laboratories with specialized facilities. We demonstrated that the use of AR-responsive promoters to express imaging reporter genes can be an effective means to interrogate AR activity in prostate tumors in living subjects. However, this form of imaging is an indirect measurement of AR function. Recently, PET probes derived from substrates of nuclear receptors that can bind to AR have been synthesized. For instance, F18-DHT has already been successfully employed to monitor metastatic lesions in patients and their response to treatment (Dehdashti et al. 2005). A new generation of probes should permit more specific assessment of various treatments on AR-containing lesions in man.

References Adams, J. Y., Johnson, M., Sato, M., Berger, F., Gambhir, S. S., Carey, M., Iruela-Arispe, M. L., and Wu, L. 2002. Visualization of advanced human prostate cancer lesions in living mice by a targeted gene transfer vector and optical imaging. Nat Med 8: 891–897. Bok, R. A., and Small, E. J. 2002. Bloodborne biomolecular markers in prostate cancer development and progression. Nat Rev Cancer 2: 918–926. Burton, J. B., Johnson, M., Sato, M., Koh, S. B. S., Mulholland, D., Stout, D., Chatziioannou, A. F., Phelps, M. E., Wu, H., and Wu, L. 2008. Adenovirus mediated gene expression imaging to detect sentinel lymph node metastasis of prostate cancer. Nat Med 14: 882–888. Cleutjens, K. B., van der Korput, H. A., van Eekelen, C. C., van Rooij, H. C., Faber, P. W., and Trapman, J. 1997. An androgen response element in a far upstream enhancer region is essential

Imaging Androgen Receptor Function In Vivo

117

for high, androgen-regulated activity of the prostate-specific antigen promoter. Mol Endocrinol 11: 148–161. Cleutjens, K. B., van Eekelen, C. C., van der Korput, H. A., Brinkman, A. O., and Trapman, J. 1996. Two androgen response regions cooperate in steroid hormone regulated activity of the prostate-specific antigen promoter. J Biol Chem 271: 6379–6388. Contag, C. H., and Bachmann, M. H. 2002. Advances in in vivo bioluminescence imaging of gene expression. Annu Rev Biomed Eng 4: 235–260. Contag, C. H., Jenkins, D., Contag, P. R., and Negrin, R. S. 2000. Use of reporter genes for optical measurements of neoplastic disease in vivo. Neoplasia 2: 41–52. Cruzalegui, F. H., Cano, E., and Treisman, R. 1999. ERK activation induces phosphorylation of Elk-1 at multiple S/T-P motifs to high stoichiometry. Oncogene 18: 7948–7957. Dehdashti, F., Picus, J., Michalski, J. M., Dence, C. S., Siegel, B. A., Katzenellenbogen, J. A., and Welch, M. J. 2005. Positron tomographic assessment of androgen receptors in prostatic carcinoma. Eur J Nucl Med Mol Imaging 32: 344–350. Dehm, S. M., and Tindall, D. J. 2007. Androgen receptor structural and functional elements: role and regulation in prostate cancer. Mol Endocrinol 21: 2855–2863. Ellwood-Yen, K., Wongvipat, J., and Sawyers, C. 2006. Transgenic mouse model for rapid pharmacodynamic evaluation of antiandrogens. Cancer Res 66: 10513–10516. Emami, K. H., and Carey, M. 1992. A synergistic increase in potency of a multimerized VP16 transcriptional activation domain. Embo J 11: 5005–5012. Gambhir, S. S. 2002. Molecular imaging of cancer with positron emission tomography. Nat Rev Cancer 2: 683–693. Gingrich, J. R., Barrios, R. J., Morton, R. A., Boyce, B. F., DeMayo, F. J., Finegold, M. J., Angelopoulou, R., Rosen, J. M., and Greenberg, N. M. 1996. Metastatic prostate cancer in a transgenic mouse. Cancer Res 56: 4096–4102. Gioeli, D., Mandell, J. W., Petroni, G. R., Frierson, H. F., Jr., and Weber, M. J. 1999. Activation of mitogen-activated protein kinase associated with prostate cancer progression. Cancer Res 59: 279–284. Gregory, C. W., He, B., Johnson, R. T., Ford, O. H., Mohler, J. L., French, F. S., and Wilson, E. M. 2001a. A mechanism for androgen receptor-mediated prostate cancer recurrence after androgen deprivation therapy. Cancer Res 61: 4315–4319. Gregory, C. W., Johnson, R. T., Jr., Mohler, J. L., French, F. S., and Wilson, E. M. 2001b. Androgen receptor stabilization in recurrent prostate cancer is associated with hypersensitivity to low androgen. Cancer Res 61: 2892–2898. Helms, M. W., Brandt, B. H., and Contag, C. H. 2006. Options for visualizing metastatic disease in the living body. Contrib Microbiol 13: 209–231. Hexdall, L., and Zheng, C. F. 2001. Stable luciferase reporter cell lines for signal transduction pathway readout using GAL4 fusion transactivators. Biotechniques 30: 1134–1138, 1140. Hsieh, C. L., Xie, Z., Liu, Z. Y., Green, J. E., Martin, W. D., Datta, M. W., Yeung, F., Pan, D., and Chung, L. W. 2005. A luciferase transgenic mouse model: visualization of prostate development and its androgen responsiveness in live animals. J Mol Endocrinol 35: 293–304. Hsieh, C. L., Xie, Z., Yu, J., Martin, W. D., Datta, M. W., Wu, G. J., and Chung, L. W. 2007. Noninvasive bioluminescent detection of prostate cancer growth and metastasis in a bigenic transgenic mouse model. Prostate 67: 685–691. Huang, W., Shostak, Y., Tarr, P., Sawyers, C., and Carey, M. 1999. Cooperative assembly of androgen receptor into a nucleoprotein complex that regulates the prostate-specific antigen enhancer. J Biol Chem 274: 25756–25768. Ignowski, J. M., and Schaffer, D. V. 2004. Kinetic analysis and modeling of firefly luciferase as a quantitative reporter gene in live mammalian cells. Biotechnol Bioeng 86: 827–834. Ilagan, R., Pottratz, J., Le, K., Zhang, L., Wong, S. G., Ayala, R., Iyer, M., Wu, L., Gambhir, S. S., and Carey, M. 2006. Imaging mitogen-activated protein kinase function in xenograft models of prostate cancer. Cancer Res 66: 10778–10785.

118

M. Carey, L. Wu

Ilagan, R., Zhang, L. J., Pottratz, J., Le, K., Salas, S., Iyer, M., Wu, L., Gambhir, S. S., and Carey, M. 2005. Imaging androgen receptor function during flutamide treatment in the LAPC9 xenograft model. Mol Cancer Ther 4: 1662–1669. Iyer, M., Salazar, F. B., Lewis, X., Zhang, L., Carey, M., Wu, L., and Gambhir, S. S. 2004. Noninvasive imaging of enhanced prostate-specific gene expression using a two-step transcriptional amplification-based lentivirus vector. Mol Ther 10: 545–552. Iyer, M., Salazar, F. B., Lewis, X., Zhang, L., Wu, L., Carey, M., and Gambhir, S. S. 2005a. Noninvasive imaging of a transgenic mouse model using a prostate-specific two-step transcriptional amplification strategy. Transgenic Res 14: 47–55. Iyer, M., Salazar, F. B., Wu, L., Carey, M., and Gambhir, S. S. 2006. Bioluminescence imaging of systemic tumor targeting using a prostate-specific lentiviral vector. Hum Gene Ther 17: 125–132. Iyer, M., Sato, M., Johnson, M., Gambhir, S. S., and Wu, L. 2005b. Applications of molecular imaging in cancer gene therapy. Curr Gene Ther 5: 607–618. Iyer, M., Wu, L., Carey, M., Wang, Y., Smallwood, A., and Gambhir, S. S. 2001. Two-step transcriptional amplification as a method for imaging reporter gene expression using weak promoters. Proc Natl Acad Sci U S A 98: 14595–14600. Jacobson, O., Bechor, Y., Icar, A., Novak, N., Birman, A., Marom, H., Fadeeva, L., Golan, E., Leibovitch, I., Gutman, M., et al. 2005. Prostate cancer PET bioprobes: synthesis of [18F]radiolabeled hydroxyflutamide derivatives. Bioorg Med Chem 13: 6195–6205. Johnson, M., Sato, M., Burton, J., Gambhir, S. S., Carey, M., and Wu, L. 2005. Micro-PET/CT monitoring of herpes thymidine kinase suicide gene therapy in a prostate cancer xenograft: the advantage of a cell-specific transcriptional targeting approach. Mol Imaging 4: 463–472. Klein, K. A., Reiter, R. E., Redula, J., Moradi, H., Zhu, X. L., Brothman, A. R., Lamb, D. J., Marcelli, M., Belldegrun, A., Witte, O. N., and Sawyers, C. L. 1997. Progression of metastatic human prostate cancer to androgen independence in immunodeficient SCID mice. Nat Med 3: 402–408. Latham, J. P., Searle, P. F., Mautner, V., and James, N. D. 2000. Prostate-specific antigen promoter/enhancer driven gene therapy for prostate cancer: construction and testing of a tissue-specific adenovirus vector. Cancer Res 60: 334–341. Marais, R., Wynne, J., and Treisman, R. 1993. The SRF accessory protein Elk-1 contains a growth factor-regulated transcriptional activation domain. Cell 73: 381–393. Marcelli, M., Stenoien, D. L., Szafran, A. T., Simeoni, S., Agoulnik, I. U., Weigel, N. L., Moran, T., Mikic, I., Price, J. H., and Mancini, M. A. 2006. Quantifying effects of ligands on androgen receptor nuclear translocation, intranuclear dynamics, and solubility. J Cell Biochem 98: 770–788. Massoud, T. F., and Gambhir, S. S. 2003. Molecular imaging in living subjects: seeing fundamental biological processes in a new light. Genes Dev 17: 545–580. Mohler, J. L., Gregory, C. W., Ford, O. H., 3rd, Kim, D., Weaver, C. M., Petrusz, P., Wilson, E. M., and French, F. S. 2004. The androgen axis in recurrent prostate cancer. Clin Cancer Res 10: 440–448. Parent, E. E., Carlson, K. E., and Katzenellenbogen, J. A. 2007. Synthesis of 7alpha-(fluoromethyl)dihydrotestosterone and 7alpha-(fluoromethyl)nortestosterone, structurally paired androgens designed to probe the role of sex hormone binding globulin in imaging androgen receptors in prostate tumors by positron emission tomography. J Org Chem 72: 5546–5554. Parent, E. E., Dence, C. S., Sharp, T. L., Welch, M. J., and Katzenellenbogen, J. A. 2006. Synthesis and biological evaluation of a fluorine-18-labeled nonsteroidal androgen receptor antagonist, N-(3-[18F]fluoro-4-nitronaphthyl)-cis-5-norbornene-endo-2,3-dicarboxylic imide. Nucl Med Biol 33: 615–624. Rauen, K. A., Sudilovsky, D., Le, J. L., Chew, K. L., Hann, B., Weinberg, V., Schmitt, L. D., and McCormick, F. 2002. Expression of the coxsackie adenovirus receptor in normal prostate and in primary and metastatic prostate carcinoma: potential relevance to gene therapy. Cancer Res 62: 3812–3818.

Imaging Androgen Receptor Function In Vivo

119

Reid, K. J., Hendy, S. C., Saito, J., Sorensen, P., and Nelson, C. C. 2001. Two classes of androgen receptor elements mediate cooperativity through allosteric interactions. J Biol Chem 276: 2943–2952. Riegman, P. H., Vlietstra, R. J., van der Korput, J. A., Brinkmann, A. O., and Trapman, J. 1991. The promoter of the prostate-specific antigen gene contains a functional androgen responsive element. Mol Endocrinol 5: 1921–1930. Roudier, M. P., True, L. D., Higano, C. S., Vesselle, H., Ellis, W., Lange, P., and Vessella, R. L. 2003. Phenotypic heterogeneity of end-stage prostate carcinoma metastatic to bone. Hum Pathol 34: 646–653. Sadowski, I., Ma, J., Triezenberg, S., and Ptashne, M. 1988. GAL4-VP16 is an unusually potent transcriptional activator. Nature 335: 563–564. Sato, M., Figueiredo, M. L., Burton, J. B., Johnson, M., Chen, M., Powell, R., Gambhir, S. S., Carey, M., and L., W. 2008. Configurations of a two-tiered amplified gene expression system vector to improve the specificity of in vivo prostate cancer imaging. Gene Therapy (under revision). Sato, M., Johnson, M., Zhang, L., Gambhir, S. S., Carey, M., and Wu, L. 2005. Functionality of androgen receptor-based gene expression imaging in hormone refractory prostate cancer. Clin Cancer Res 11: 3743–3749. Sato, M., Johnson, M., Zhang, L., Zhang, B., Le, K., Gambhir, S. S., Carey, M., and Wu, L. 2003. Optimization of adenoviral vectors to direct highly amplified prostate-specific expression for imaging and gene therapy. Mol Ther 8: 726–737. Schoder, H., and Larson, S. M. 2004. Positron emission tomography for prostate, bladder, and renal cancer. Semin Nucl Med 34: 274–292. Schuur, E. R., Henderson, G. A., Kmetec, L. A., Miller, J. D., Lamparski, H. G., and Henderson, D. R. 1996. Prostate-specific antigen expression is regulated by an upstream enhancer. J Biol Chem 271: 7043–7051. Seethammagari, M. R., Xie, X., Greenberg, N. M., and Spencer, D. M. 2006. EZC-prostate models offer high sensitivity and specificity for noninvasive imaging of prostate cancer progression and androgen receptor action. Cancer Res 66: 6199–6209. Shah, R. B., Mehra, R., Chinnaiyan, A. M., Shen, R., Ghosh, D., Zhou, M., Macvicar, G. R., Varambally, S., Harwood, J., Bismar, T. A., et al. 2004. Androgen-independent prostate cancer is a heterogeneous group of diseases: lessons from a rapid autopsy program. Cancer Res 64: 9209–9216. Taplin, M. E. 2007. Drug insight: role of the androgen receptor in the development and progression of prostate cancer. Nat Clin Pract Oncol 4: 236–244. Weber, M. J., and Gioeli, D. 2004. Ras signaling in prostate cancer progression. J Cell Biochem 91: 13–25. Wu, L., Johnson, M., and Sato, M. 2003. Transcriptionally targeted gene therapy to detect and treat cancer. Trends Mol Med 9: 421–429. Wu, L., Matherly, J., Smallwood, A., Adams, J. Y., Billick, E., Belldegrun, A., and Carey, M. 2001. Chimeric PSA enhancers exhibit augmented activity in prostate cancer gene therapy vectors. Gene Ther 8: 1416–1426. Xie, X., Zhao, X., Liu, Y., Young, C. Y., Tindall, D. J., Slawin, K. M., and Spencer, D. M. 2001. Robust prostate-specific expression for targeted gene therapy based on the human kallikrein 2 promoter. Hum Gene Ther 12: 549–561. Ye, X., Han, S. J., Tsai, S. Y., DeMayo, F. J., Xu, J., Tsai, M. J., and O’Malley, B. W. 2005. Roles of steroid receptor coactivator (SRC)-1 and transcriptional intermediary factor (TIF) 2 in androgen receptor activity in mice. Proc Natl Acad Sci U S A 102: 9487–9492. Zhang, J., Gao, N., Kasper, S., Reid, K., Nelson, C., and Matusik, R. J. 2004. An androgendependent upstream enhancer is essential for high levels of probasin gene expression. Endocrinology 145: 134–148.

120

M. Carey, L. Wu

Zhang, J., Thomas, T. Z., Kasper, S., and Matusik, R. J. 2000. A small composite probasin promoter confers high levels of prostate-specific gene expression through regulation by androgens and glucocorticoids in vitro and in vivo. Endocrinology 141: 4698–4710. Zhang, L., Adams, J. Y., Billick, E., Ilagan, R., Iyer, M., Le, K., Smallwood, A., Gambhir, S. S., Carey, M., and Wu, L. 2002. Molecular engineering of a two-step transcription amplification (TSTA) system for transgene delivery in prostate cancer. Mol Ther 5: 223–232. Zhang, L., Johnson, M., Le, K. H., Sato, M., Ilagan, R., Iyer, M., Gambhir, S. S., Wu, L., and Carey, M. 2003. Interrogating androgen receptor function in recurrent prostate cancer. Cancer Res 63: 4552–4560.

Increased Expression of Genes Converting Adrenal Androgens to Testosterone in Castration-Recurrent Prostate Cancer Steven P. Balk

Abstract Androgen deprivation is still the standard systemic therapy for locally advanced or metastatic prostate cancer (PCa), but patients invariably relapse with a more aggressive form of PCa that has been termed castration-recurrent PCa (CRPCa). The androgen receptor (AR) is expressed at high levels in most cases of CRPCa, and these tumors resume their expression of multiple AR regulated genes, which indicates that AR transcriptional activity becomes reactivated at this stage of the disease. Mechanisms that may contribute to AR reactivation in CRPCa include increased AR protein expression, AR mutations, increased expression of transcriptional coactivator proteins, and activation of signal transduction pathways that can enhance AR responses to low levels of androgens. Recent data indicate that a further mechanism for AR reactivation in CRPCa cells may be through increased intracellular synthesis of testosterone and 5a-dihydrotestosterone (DHT). The enzymes that mediate androgen synthesis and metabolism in normal prostate and in PCa, and evidence indicating that their increased expression contributes to the development of CRPCa, are outlined in this chapter. The early use of therapies that more aggressively block androgen production may enhance responses to androgen deprivation therapy, and prevent or delay the adaptations that eventually lead to CRPCa.

1 Introduction Androgen receptor (AR) is expressed at low levels in most tissues and its transcriptional activity is stimulated by circulating testosterone, which in males is produced primarily by the testes. AR is also expressed at low levels by a subset of stromal

S.P. Balk Cancer Biology Program, Hematology Oncology Division, Beth Israel Deaconess Medical Center and Harvard Medical School, 330 Brookline Avenue, Boston, MA 02215, USA, E-mail: sbalk@ bidmc.harvard.edu

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4_5, # Springer Science + Business Media, LLC 2009

123

124

S.P. Balk

cells in prostate, but is expressed at high levels by the luminal epithelial cells in adult prostate. The majority of prostate cancers (PCa) similarly express high levels of AR, and their growth can be suppressed by androgen-deprivation therapy, which involves removal of testicular androgens (surgical or medical castration) or treatment with an AR antagonist (or a combination of both). Androgen deprivation is still the standard systemic therapy for locally advanced or metastatic PCa, but patients invariably relapse with a more aggressive form of PCa that has been termed hormone-refractory, androgen-independent, or castration-recurrent PCa (CRPCa). Significantly, the AR is expressed at high levels in most cases of CRPCa, and these tumors resume their expression of multiple AR-regulated genes (including TMPRSS2-Ets fusion genes), which indicates that AR transcriptional activity becomes reactivated at this stage of the disease. Moreover, studies in cell line and xenograft models of CRPCa indicate that AR remains essential for tumor growth, which suggests that the reactivated AR provides a therapeutic target in CRPCa. However, the molecular basis for this reactivation remains unclear. Mechanisms that may contribute to AR reactivation in CRPCa include increased AR protein expression, AR mutations (which can enhance AR activation by weak androgens, other steroid hormones, and drugs), increased expression of transcriptional coactivator proteins, and activation of signal transduction pathways that can enhance AR responses to low levels of androgens. Recent data indicate that a further mechanism for AR reactivation in CRPCa cells may be through increased intracellular synthesis of testosterone and 5a-dihydrotestosterone (DHT). The enzymes that mediate androgen synthesis and metabolism in normal prostate and in PCa, and evidence indicating that their increased expression contributes to the development of CRPCa, are outlined in this chapter.

2 Intraprostatic Androgen Production in Normal Prostate and Prostate Cancer The majority of circulating testosterone in healthy adult males is synthesized in the testes, with only about 5–10% being produced by the adrenal glands. However, the adrenal gland in humans is a major source of weak androgens, which include androst-4-ene-3,17 dione (androstenedione), dehydroepiandrosterone (DHEA), and DHEA-sulfate (DHEA-S). These adrenal androgens serve as precursors that can be converted to testosterone and DHT in some peripheral tissues, with the levels of circulating DHEA-S being 100-fold higher than those of testosterone. Interestingly, the adrenal glands of many other species, including mouse, rat, and dog, do not produce significant levels of these weak androgens (Belanger et al. 1989). Prostate is one of the peripheral tissues that can convert DHEA-S and DHEA to androstenedione, and further convert androstenedione to testosterone, although this may not be a major source of intraprostatic testosterone in the setting of normal circulating levels (see later).

Increased Expression of Genes Converting Adrenal Androgens to Testosterone

125

Although testosterone is a high-affinity ligand for AR, AR transcriptional activity in prostate is further stimulated through the intraprostatic reduction of testosterone to the higher-affinity ligand DHT (approximately eightfold higher affinity). This reaction in benign prostate is mediated primarily by the type 2 5a-reductase, and inhibition of this enzyme with the drug finasteride decreases AR transcriptional activity by 50% based on expression of PSA. The prostatic expression of enzymes that regulate testosterone and DHT synthesis and metabolism presumably reflects the need to maintain DHT at levels that are optimal for maximal AR activity.

2.1

Intraprostatic Androgen Levels in Benign Prostate and Untreated Primary PCA

While intraprostatic synthesis may not be a major source of testosterone in men with intact testes and normal circulating testosterone, it does appear to be a substantial source when testicular production is blocked. In one study of healthy men, treatment for one month with a GnRH antagonist caused a 94% decline in serum testosterone, but only a 70–80% decline in prostate tissue levels of testosterone and DHT (Page et al. 2006). Several other studies similarly have found that intraprostatic androgen levels do not decline as markedly as serum levels after androgen-deprivation therapy. One early study found that DHT levels declined by only 75% after castration, and declined by 86% after ketoconazole to suppress both testicular and adrenal androgen production (Geller and Albert 1987). More recent studies also found that prostate tissue levels of DHT in PCa patients undergoing neoadjuvant androgen-deprivation therapy were only 75% lower than control levels (Belanger et al. 1989; Mizokami et al. 2004; Nishiyama et al. 2004). Taken together, these data show that intraprostatic synthesis can be a significant source of testosterone and DHT in benign prostate in the absence of testicular androgens.

2.2

Intraprostatic Androgen Levels in CRPCA

Several studies have examined intraprostatic androgen levels in men with CRPCa and intact prostates. An early study in men with CRPCa found that intraprostatic DHT levels were elevated to >50% of control levels (noncastrated men) in 2/20 men treated with surgical castration and in 4/9 men who received estrogen treatment (Geller et al. 1984; Geller 1985). A more recent larger study was done in 22 castrated men with CRPCa who had tissue obtained from transurethral resections performed to relieve obstruction. Importantly, intraprostatic testosterone levels were not reduced significantly relative to controls with normal serum androgen levels, while DHT levels were reduced to 18% of controls (Mohler et al. 2004; Titus et al. 2005b). Overall, these data indicate that the progression to CRPCa may be

126

S.P. Balk

associated with increased intratumoral accumulation or synthesis of testosterone, while DHT levels appear to be similar in primary PCa shortly after castration and in subsequent CRPCa.

3 Expression of Enzymes Mediating Androgen Synthesis in Normal Prostate Consistent with the relatively high tissue levels of testosterone and DHT in benign prostate after castration, the enzymes that can generate these potent androgens from weak adrenal androgen precursors are expressed in benign prostate. The major enzymes that appear to regulate androgen synthesis and metabolism in benign prostate are outlined in Fig. 1.

3.1

3 b-Hydroxysteroid Dehydrogenase 1 (HSD3B1)

The enzyme HSD3B1 is expressed in prostate basal epithelial cells and converts DHEA to androstenedione, which can then be reduced to testosterone (see later) (Labrie et al. 2000). DHEA-S is the major androgen precursor synthesized by the adrenal gland and circulates at relatively high levels. DHEA-S is converted to DHEA in prostate, so 3b-HSD1 has the potential to provide the prostate with substantial substrate for intraprostatic androgen production. A study using cultured benign prostate epithelial cells (PrEC, which are predominantly basal cells) found that HSD3B1 expression could be induced rapidly by IL-4 and IL-13, which suggests a mechanism by which inflammatory cells may enhance intraprostatic androgen synthesis (Gingras and Simard 1999).

β

α

β

α

Fig. 1 Outline of enzymes mediating androgen metabolism described in the text

Increased Expression of Genes Converting Adrenal Androgens to Testosterone

3.2

127

Aldo-Keto Reductase Family 1, Member C3 (AKR1C3)

AKR1C3 is the major enzyme in prostate that functions as a 17-ketoreductase to reduce androstenedione to testosterone (Lin et al. 1997; El Alfy et al. 1999; Dufort et al. 1999; Penning et al. 2000; Lin et al. 2004). This enzyme has also been referred to as 17b-hydroxysteroid dehydrogenase type 5 (17b-HSD5) or 3a-hydroxysteroid dehydrogenase type 2. The AKR1C family in humans has four linked and highly homologous (>86% amino acid identity) members, AKR1C1–4, which function primarily as ketosteroid reductases because the reverse oxidative reactions are inhibited strongly by NADPH (Ratnam et al. 1999; Penning et al. 2000; Rizner et al. 2003; Steckelbroeck et al. 2004). AKR1C1 and AKR1C2 are also expressed in prostate, but in contrast to AKR1C3, the AKR1C1 and C2 enzymes function primarily to inactivate DHT (see later). Testosterone synthesis by Leydig cells in the testes is mediated by a distinct enzyme, type 3 17b-hydroxysteroid dehydrogenase (Andersson et al. 1995). In normal prostate, AKR1C3 has been localized to stromal, endothelial, and perineural cells using in situ hybridization and immunohistochemistry (Pelletier et al. 1999; Lin et al. 2004; Fung et al. 2006; Penning et al. 2006). A biological function for AKR1C3 in normal prostate may be to increase testosterone levels in prostate stroma and/or to generate increased testosterone for the epithelia. However, AKR1C3’s importance as a source of testosterone in men with normal testicular function is not clear, and the enzyme may have alternative biological functions in normal prostate. Other reactions catalyzed by AKR1C3 include the reduction of estrone to 17b-estradiol and the reduction of prostaglandin D2 to prostaglandin F2 (Matsuura et al. 1998; Komoto et al. 2004). This latter reaction could decrease the production of prostaglandin J2 ligands for PPARg, which are normally derived from prostaglandin D2 (Desmond et al. 2003). AKR1C3 also can function as a 3-ketosteroid reductase and convert DHT to the weak androgen, 5a-androstane-3a,17b-diol (3a-androstanediol), similar to AKR1C2 (see later), but this activity does not appear significant in vivo (Penning et al. 2000).

3.3

Type 1 and 2 5a-Reductases (Steroid Reductase 5a1 And 5a2, SRD5A1, And SRD5A2)

Both SRD5A1 and SRD5A2 are expressed in normal prostate epithelium and mediate the conversion of testosterone to DHT, which has an eightfold higher affinity for AR (Russell and Wilson 1994). However, the type 2 enzyme is the major form expressed in normal prostate. The importance of this enzyme for prostate function has been clearly established based on the failure of prostate development in men with SRD5A2 deficiency (Wilson et al. 1993). In addition, treatment of healthy men with finasteride, a relatively specific inhibitor of the type 2 enzyme, results in a substantial 50% reduction in expression of the androgenregulated PSA gene and can decrease prostate size (Thompson et al. 2003).

128

S.P. Balk

Moreover, a major study in healthy men showed that PCa incidence decreased significantly in men taking finasteride versus placebo (Thompson et al. 2003).

3.4

Aldo-Keto Reductase Family 1, Member C2 (AKR1C2)

AKR1C2 has also been termed 3a-hydroxysteroid dehydrogenase type 3 (3a-HSD type 3). AKR1C2 is closely related to AKR1C3, but it acts primarily as a 3ketosteroid reductase and reduces DHT to 5a-androstane-3a,17b-diol (3a-androstanediol), which is an extremely weak androgen (Penning et al. 2000; Rizner et al. 2003; Ji et al. 2003). This compound is subsequently glucuronidated to form 3aandrostanediol-glucuronide, which is eliminated into the circulation. In contrast to AKR1C3, AKR1C2 is expressed at comparable or higher levels in the luminal epithelium versus the stroma of benign prostate (Ji et al. 2003; Bauman et al. 2006a). Consistent with a biological role in the inactivation of DHT, expression of AKR1C2 is increased by DHT treatment of LNCaP PCa cells (Ji et al. 2003). Moreover, LNCaP cells expressing increased levels of transfected AKR1C2 have a diminished response to DHT, but not to the synthetic agonist R1881, with similar findings in transfected PC3 and LAPC-4 cells (Rizner et al. 2003; Ji et al. 2003).

3.5

Aldo-Keto Reductase Family 1, Member C1 (AKR1C1)

AKR1C1 is the third member of the AKR1C family that is expressed in normal prostate. Similar to AKR1C2, it can function as a 3-ketosteroid reductase to reduce DHT, and increased expression by transfection in LNCaP cells decreases DHT stimulated growth. However, AKR1C1 is less active as a DHT reductase, and in contrast to AKR1C2, it converts DHT primarily to 5a-androstane-3b,17b-diol (3bandrostanediol) (Steckelbroeck et al. 2004). 3b-androstenediol has been identified as a potential endogenous ligand for the estrogen receptor b, which provides another link between androgen metabolism and estrogens (Weihua et al. 2001; Weihua et al. 2002). AKR1C1 also functions as a 20a-ketosteroid reductase, and converts progesterone to 20a-hydroxyprogesterone, a weak progestin, which has its major activity in breast (Penning et al. 2000).

3.6

Retinol Dehydrogenase like 3a-Hydroxysteroid Dehydrogenase (HSD17B6)

While DHT is synthesized primarily from testosterone, an alternative source is through oxidation of 3a-androstanediol. This is the reverse of the reaction catalyzed

Increased Expression of Genes Converting Adrenal Androgens to Testosterone

129

by AKR1C2, but AKR1C2 does not carry out this reaction efficiently in vivo. Recent studies of enzymes with 3a-HSD activity in human prostate have shown that the major enzyme mediating this reaction is retinol dehydrogenase like 3a-hydroxysteroid dehydrogenase (HSD17B6) (Bauman et al. 2006b). Based on real-time RT-PCR of cultured prostate stromal and epithelial cells, HSD17B6 is expressed primarily in stromal cells and is increased in BPH (Bauman et al. 2006a; Penning et al. 2007).

3.7

UDP Glycosyltransferase 2, B15 (UGT2B15) and UDP Glycosyltransferase 2, B17 (UGT2B17)

Both DHT and 3a-androstanediol can be inactivated irreversibly by glucuronidation. The major metabolite in prostate is generated by addition of a glucuronyl group to the 17-hydroxy position of 3a-androstanediol to generate 5a-androstane3a,17b-diol glucuronide (3a-androstanediol-17 glucuronide) (Belanger et al. 2003). This reaction can be carried out by UGTB15 and UGTB17, which are expressed in human prostate. In situ hybridization and immunohistochemistry indicate that UGT2B15 is expressed primarily in the luminal epithelium of normal prostate, while UGT2B17 is in the basal cells (Chouinard et al. 2004). Both enzymes are expressed in LNCaP PCa cells, but their expression is downregulated by androgen stimulation, suggesting that they do not regulate androgen metabolism by a straightforward negative feedback loop (Guillemette et al. 1996; Chouinard et al. 2006). In contrast, UGT2B15 expression is increased strongly by estrogen in breast cancer cells (Harrington et al. 2006).

4 Enzymes Mediating Androgen Metabolism in Untreated Primary PCA Altered expression of the aldo-keto reductases and 5a-reductases has been reported in several studies of untreated primary PCa.

4.1

Altered Expression of AKR1C2 and AKR1C3

In an initial study using RT-PCR to assess AKR1C family expression in matched normal versus PCa tissue, one group found a variable (approximately 3 to 10-fold) and significant reduction in AKR1C2 message levels, but only small and variable changes in AKR1C1 and AKR1C3 (Ji et al. 2003). In a subsequent study, this group found decreased AKR1C1 expression (Ji et al. 2007). Moreover, they also found that DHT levels were increased in PCa tissue compared to paired benign tissue

130

S.P. Balk

(42% higher) and that metabolism of 3H-DHT was reduced in freshly isolated malignant versus benign tissues. Consistent with this result, an early study also found increased DHT and decreased 3a-androstanediol levels in PCa (Krieg et al. 1979). Other studies using epithelial cells cultured from PCa versus benign prostate have not detected significant changes in AKR1C2, AKR1C1, or AKR1C3 transcript levels, but it is not clear whether expression levels in the cultured cells are similar to those in tumor cells in vivo (Bauman et al. 2006a). Indeed, immunohistochemical studies using a monoclonal antibody against AKR1C3 have shown that AKR1C3 protein is detectable in primary PCa cells, but absent in benign luminal epithelial cells (Lin et al. 2004; Fung et al. 2006). AKR1C3 protein expression was also increased in endothelial cells in PCa. Taken together, these results suggest that DHT levels may be increased in primary PCa through decreased catabolism (due to lower AKR1C2 expression) and increased tumor cell synthesis of testosterone (mediated by AKR1C3).

4.2

Increased Expression of SRD5A1 and Decreased SRD5A2

While type 2 5a-reductase (SRD5A2) is the predominant form in benign prostate, several studies have shown that expression of the type 1 5a-reductase (SRD5A1) is increased in primary PCa based on transcript and protein levels (Bonkhoff et al. 1996; Iehle et al. 1999; Thomas et al. 2003, 2005; Titus et al. 2005a). In contrast, SRD5A2 levels were not increased and were found to be decreased in some studies (Bjelfman et al. 1997; Soderstrom et al. 2001; Luo et al. 2003; Thomas et al. 2005; Titus et al. 2005a). This apparent switch from expression of SRD5A2 in normal prostate to SRD5A1 in PCa may have important clinical implications, as the commonly used 5a-reductase inhibitor (finasteride) is relatively specific for the type 2 enzyme.

5 Enzymes Mediating Androgen Metabolism in CRPCA Two studies of global gene expression in CRPCa using Affymetrix oligonucleotide microarrays have found altered expression of enzymes mediating androgen synthesis and catabolism. The first study found increased expression of multiple enzymes in the cholesterol biosynthetic pathway, which may provide increased cholesterol for de novo steroid biosynthesis (Holzbeierlein et al. 2004). The second found increased expression of enzymes that mediate testosterone and DHT synthesis from adrenal precursors, and that mediate DHT inactivation (Stanbrough et al. 2006). Significantly, both studies showed marked increases in AR message levels and found that expression of AR-regulated genes was restored, consistent with reactivation of AR transcriptional activity in CRPCa.

Increased Expression of Genes Converting Adrenal Androgens to Testosterone

5.1

131

Increased Expression of Enzymes Mediating Cholesterol Synthesis

The first study used Affymetrix oligonucleotide microarrays to assess global changes in gene expression in 23 untreated primary PCa, 17 primary PCa after 3 months of androgen-deprivation therapy, six cases of untreated metastatic PCa, and three cases of metastatic CRPCa (Holzbeierlein et al. 2004). Of the 650 genes that were downregulated in the group after 3 months of androgen deprivation, 25% appeared to be direct AR targets. Significantly, the overall pattern of gene expression in the three CRPCa samples was more similar to the untreated tumors than the androgen-deprived tumors, consistent with the reactivation of AR transcriptional activity. A comparison of gene expression in the untreated primary and metastatic PCa versus the metastatic CRPCa revealed 100 genes that were at least threefold differentially expressed in the CRPCa. One of these genes with increased expression in CRPCa was AR, which was increased approximately tenfold. A series of genes whose products regulate cholesterol synthetase were also increased (HMG-CoA synthase, squalene synthtase, squalene monooxygenase, and lanosterol synthase), and may provide increased endogenous substrate for steroid and lipid synthesis. Previous studies have shown that androgens activate both fatty acid and cholesterol biosynthetic pathways by increasing expression of sterol response-element-binding proteins 1 and 2 (SREBP), which are the transcription factors regulating multiple genes that control fatty acid synthesis (such as fatty acid synthetase) and cholesterol synthesis (such as HMG-CoA synthase) (Swinnen et al. 1997; Heemers et al. 2001). Androgens further enhance the activation of SREBP by increasing expression of SREBP cleavage-activating protein (SCAP), which binds to SREBP in the endoplasmic reticulum and facilitates its cleavage and subsequent nuclear translocation (Heemers et al. 2001). Importantly, another study showed that SREBP and SCAP levels were increased in LNCaP xenografts that recurred after castration (Ettinger et al. 2004). Moreover, in clinical samples examined using immunohistochemistry, this study found that SREBP-1 levels decreased after androgen-deprivation therapy, but were increased in CRPCa. Therefore, the apparent increase in cholesterol synthesis in CRPCa is consistent with AR reactivation, and may provide a positive feedback loop for further AR activation by providing cholesterol for de novo androgen synthesis in addition to lipid synthesis.

5.2

Increased Expression of Enzymes Mediating Androgen Synthesis from Adrenal Precursors

Affymetirx oligonucleotide microarrays were used in a larger study of 33 CRPCa bone metastases, which were compared with 22 laser capture microdissected primary PCa from patients who had not received androgen-deprivation therapy

132

S.P. Balk

(and control bone marrow biopsies that did not contain tumor) (Stanbrough et al. 2006). AR message levels were increased consistently, with the average increase being approximately sixfold. Multiple AR-regulated genes were also highly expressed, although their levels were still approximately threefold lower than in the primary tumors, which indicates that AR transcriptional activity was only partially restored. Significant changes also were found in the expression of multiple enzymes mediating androgen metabolism. The largest change was in AKR1C3, which reduces androstenedione to testosterone and was increased >5-fold in the CRPCa samples. This increase was also observed using real-time RT-PCR and immunohistochemistry, with increased immunostaining observed in 11/19 CRPCa samples. Importantly, the increase in AKR1C3 is consistent with data outlined in Sect. 2.2, which showed that intraprostatic testosterone levels are increased in CRPCa (relative to levels shortly after androgen deprivation) and do not differ significantly from the levels in benign prostate with normal systemic androgen levels. An alternative reaction catalyzed by AKR1C3 is the reduction of DHEA to androst-5-ene-3b,17bdiol (5-androstenediol), which has weak AR agonist activity (Miyamoto et al. 1998). This activity may be significant, as a recent study showed that levels of 5-androstenediol in prostate were not decreased by castration (Mizokami et al. 2004). Although 5-androstenediol is 100-fold less potent than DHT as an AR agonist, its potency is markedly increased on the T877A mutant AR expressed by LNCaP cells (Mizokami et al. 2004). This study also found that type 1 5a-reductase expression was increased (approximately twofold) in CRPCa, while expression of the type 2 5a-reductase was reduced to 50% of the levels in primary untreated prostate cancers (Stanbrough et al. 2006). This shift from the type 2 to the type 1 5a-reductase is also observed in primary untreated PCa (Sect. 4.2), and increases in the type 1 protein and activity in CRPCa have been observed in other studies (Thomas et al. 2005; Titus et al. 2005a). Overall, these results indicate that there is a progressive shift to the type 1 5a-reductase that occurs during the development of PCa and continues as the tumors adapt to androgen deprivation to become CRPCa. The selective advantages conferred on the tumor cells by this shift are not clear, but these observations suggest that a dual type 1 and type 2 5a-reductase inhibitor may have additional activity in PCa relative to a type 2 specific inhibitor (Thomas et al. 2007). A final enzyme mediating androgen synthesis that was increased (1.8-fold) in CRPCa was 3b-hydroxysteroid dehydrogenase 2 (HSD3B2), which converts DHEA to androstenedione. In humans there are two 3b-hydroxysteroid dehydrogenase isoforms (HSD3B1 and HSD3B2), and these are also required for the synthesis of all other steroid hormones (Simard et al. 2005). As noted in Sect. 3.1, HSD3B1 is the isoform that normally is expressed in prostate basal cells and in other peripheral tissues such as breast and skin, while HSD3B2 is primarily expressed in adrenal, ovary, and testes (Simard et al. 2005). Therefore, as observed for 5a-reductases, both an increase and a shift to a distinct isoform for HSD3B activity appear to occur in CRPCa.

Increased Expression of Genes Converting Adrenal Androgens to Testosterone

5.3

133

Increased Expression of Enzymes Mediating Androgen Inactivation

Surprisingly, expression of AKR1C2 and AKR1C1, which function to inactivate DHT (Sects. 3.4 and 3.5), also was increased significantly in the CRPCa tumors (each approximately threefold) (Stanbrough et al. 2006). Moreover, there was a strong correlation between these increases and the increased AKR1C3 expression in individual tumors. The increases in AKR1C1 and AKR1C2 were in contrast to their decreased expression in untreated primary PCa (Sect. 4.2). One interesting possibility is that this increase in CRPCa is in response to elevated intracellular testosterone and increased DHT synthesis, as AKR1C2 expression is stimulated by DHT in PCa cells (Sect. 3.4). In this case, increased androgen synthesis would represent a primary adaptation to androgen deprivation in CRPCa, with the increases in AKR1C1 and AKR1C2 being secondary. Finally, expression of UDP glycosyltransferase 2, B15 (UGT2B15) was increased 3.5-fold. This enzyme mediates the glucuronidation of DHT metabolites, in conjunction with UGT2B17 (Sect. 3.7). The selective increase in UGT2B15 vs. UGT2B17 may be related to UGT2B15 expression in the luminal epithelia of normal prostate, with UGT2B17 being expressed in basal cells (Sect. 3.7). While the increase in UGT2B15 could be in response to increased androgen synthesis (as suggested for AKR1C2), androgens do not stimulate UGT2B15 in LNCaP cells and instead decrease its expression (Sect. 3.7). Increased expression of another enzyme mediating androgen inactivation, 17bhydroxysteroid dehydrogenase type 2 (HSD17B2), also was observed in another study of CRPCa. This study examined a small group of selected genes using realtime RT-PCR in primary untreated PCa versus 13 CRPCa (the latter tissues were obtained from transurethral resections) and found an increase of approximately fourfold in HSD17B2 in CRPCa (Fromont et al. 2005). The major activity of this enzyme with respect to androgens is inactivation of testosterone by converting it to androstenedione (the reverse of the reaction mediated by AKR1C3). This result contrasts with findings in LNCaP cells, in which HSD17B2 was decreased markedly when the cells progressed to an androgen-independent stage (Harkonen et al. 2003). An early study also found that HSD17B2 activity was decreased in primary PCa compared to BPH (Elo et al. 1996).

5.4

Implications of Increased Androgen Synthesis for AR Antagonists

A poorly understood feature of CRPCa is that it generally does not respond to treatment with AR antagonists, including high-dose (150–200 mg) bicalutamide, which can block the AR when used as single agents for initial hormone therapy

134

S.P. Balk

(Fowler et al. 1995; Scher et al. 1997; Joyce et al. 1998). One explanation may be that bicalutamide and hydroxyflutamide (the active metabolite of flutamide), as relatively weak competitive antagonists, no longer can compete effectively for AR binding due to increased intracellular testosterone and DHT. Moreover, other molecular changes in CRPCa may increase markedly the AR affinity foragonist ligands (‘ ‘hypersensitive’ ’ AR). If this is correct, then the efficacy of AR antagonists in CRPCa may be enhanced by agents that decrease androgen synthesis. However, an alternative general hypothesis is that bicalutamide can compete for AR binding in CRPCa cells, but no longer functions as an antagonist due to increased AR expression or relative increases in coactivator versus corepressor proteins (Shang et al. 2002; Cheng et al. 2002; Masiello et al. 2002; Chen et al. 2004; Zhu et al. 2006; Baek et al. 2006; Hodgson et al. 2007).

6 Summary and Clinical Implications Taken together, these studies indicate that residual weak androgens, which in castrated men are produced primarily by the adrenal gland, may be an important source of ligand for AR in CRPCa. Consistent with this hypothesis, early studies found that about one-third of patients with CRPCa had objective responses (tumor shrinkage) to adrenalectomy or hypophysectomy, and that the majority of patients improved symptomatically (Mahoney and Harrison 1972). This surgical approach was later replaced by treatment with aminoglutethimide or ketoconazole, which both suppress adrenal androgen synthesis. However, responses are generally partial and transient, and it appears clear that even total elimination of adrenal androgen production will have limited efficacy. This may reflect additional sources of androgen precursors or other alternative ligands, including the de novo synthesis of androgens from cholesterol by CRPCa cells, but may also reflect the eventual emergence of tumor cells that are either truly ligand independent or are no longer dependent on AR transcriptional activity. In any case, it is clear that current androgen-deprivation therapies do not ablate androgen synthesis and that a subset of CRPCa may respond to inhibitors that can decrease further intracellular androgen levels. Abiraterone is one such promising inhibitor of adrenal androgen production that is currently in clinical trials, and dutasteride may have some value as a dual type 1 and type 2 5a-reductase inhibitor. Importantly, the efficacy of available competitive AR antagonists also may be enhanced by therapies that decrease intracellular androgen levels. Finally, it is possible that the early use of therapies that more aggressively block androgen production will enhance responses to androgen-deprivation therapy, and prevent or delay the adaptations that eventually lead to CRPCa.

Increased Expression of Genes Converting Adrenal Androgens to Testosterone

135

References Andersson, S., Geissler, W.M., Patel, S., and Wu, L. (1995). The molecular biology of androgenic 17 beta-hydroxysteroid dehydrogenases. J. Steroid Biochem. Mol. Biol. 53, 37–39. Baek, S.H., Ohgi, K.A., Nelson, C.A., Welsbie, D., Chen, C., Sawyers, C.L., Rose, D.W., and Rosenfeld, M.G. (2006). Ligand-specific allosteric regulation of coactivator functions of androgen receptor in prostate cancer cells. Proc. Natl. Acad. Sci. U. S. A 103, 3100–3105. Bauman, D.R., Steckelbroeck, S., Peehl, D.M., and Penning, T.M. (2006a). Transcript profiling of the androgen signal in normal prostate, benign prostatic hyperplasia, and prostate cancer. Endocrinology 147, 5806–5816. Bauman, D.R., Steckelbroeck, S., Williams, M.V., Peehl, D.M., and Penning, T.M. (2006b). Identification of the major oxidative 3alpha-hydroxysteroid dehydrogenase in human prostate that converts 5alpha-androstane-3alpha,17beta-diol to 5alpha-dihydrotestosterone: a potential therapeutic target for androgen-dependent disease. Mol. Endocrinol. 20, 444–458. Belanger, A., Pelletier, G., Labrie, F., Barbier, O., and Chouinard, S. (2003). Inactivation of androgens by UDP-glucuronosyltransferase enzymes in humans. Trends Endocrinol. Metab. 14, 473–479. Belanger, B., Belanger, A., Labrie, F., Dupont, A., Cusan, L., and Monfette, G. (1989). Comparison of residual C-19 steroids in plasma and prostatic tissue of human, rat and guinea pig after castration: unique importance of extratesticular androgens in men. J. Steroid Biochem. 32, 695–698. Bjelfman, C., Soderstrom, T.G., Brekkan, E., Norlen, B.J., Egevad, L., Unge, T., Andersson, S., and Rane, A. (1997). Differential gene expression of steroid 5 alpha-reductase 2 in core needle biopsies from malignant and benign prostatic tissue. J. Clin. Endocrinol. Metab 82, 2210–2214. Bonkhoff, H., Stein, U., Aumuller, G., and Remberger, K. (1996). Differential expression of 5 alpha-reductase isoenzymes in the human prostate and prostatic carcinomas. Prostate 29, 261– 267. Chen, C.D., Welsbie, D.S., Tran, C., Baek, S.H., Chen, R., Vessella, R., Rosenfeld, M.G., and Sawyers, C.L. (2004). Molecular determinants of resistance to antiandrogen therapy. Nat. Med. 10, 33–39. Cheng, S., Brzostek, S., Lee, S.R., Hollenberg, A.N., and Balk, S.P. (2002). Inhibition of the dihydrotestosterone-activated androgen receptor by nuclear receptor corepressor. Mol. Endocrinol. 16, 1492–1501. Chouinard, S., Pelletier, G., Belanger, A., and Barbier, O. (2004). Cellular specific expression of the androgen-conjugating enzymes UGT2B15 and UGT2B17 in the human prostate epithelium. Endocr. Res. 30, 717–725. Chouinard, S., Pelletier, G., Belanger, A., and Barbier, O. (2006). Isoform-specific regulation of uridine diphosphate-glucuronosyltransferase 2B enzymes in the human prostate: differential consequences for androgen and bioactive lipid inactivation. Endocrinology 147, 5431–5442. Desmond, J.C., Mountford, J.C., Drayson, M.T., Walker, E.A., Hewison, M., Ride, J.P., Luong, Q. T., Hayden, R.E., Vanin, E.F., and Bunce, C.M. (2003). The aldo-keto reductase AKR1C3 is a novel suppressor of cell differentiation that provides a plausible target for the non-cyclooxygenase-dependent antineoplastic actions of nonsteroidal anti-inflammatory drugs. Cancer Res. 63, 505–512. Dufort, I., Rheault, P., Huang, X.F., Soucy, P., and Luu-The, V. (1999). Characteristics of a highly labile human type 5 17beta-hydroxysteroid dehydrogenase. Endocrinology 140, 568–574. El Alfy, M., Luu-The, V., Huang, X.F., Berger, L., Labrie, F., and Pelletier, G. (1999). Localization of type 5 17beta-hydroxysteroid dehydrogenase, 3beta-hydroxysteroid dehydrogenase, and androgen receptor in the human prostate by in situ hybridization and immunocytochemistry. Endocrinology 140, 1481–1491.

136

S.P. Balk

Elo, J.P., Akinola, L.A., Poutanen, M., Vihko, P., Kyllonen, A.P., Lukkarinen, O., and Vihko, R. (1996). Characterization of 17beta-hydroxysteroid dehydrogenase isoenzyme expression in benign and malignant human prostate. Int. J. Cancer 66, 37–41. Ettinger, S.L., Sobel, R., Whitmore, T.G., Akbari, M., Bradley, D.R., Gleave, M.E., and Nelson, C. C. (2004). Dysregulation of sterol response element-binding proteins and downstream effectors in prostate cancer during progression to androgen independence. Cancer Res. 64, 2212–2221. Fowler, J.E., Jr., Pandey, P., Seaver, L.E., and Feliz, T.P. (1995). Prostate specific antigen after gonadal androgen withdrawal and deferred flutamide treatment. J. Urol. 154, 448–453. Fromont, G., Chene, L., Vidaud, M., Vallancien, G., Mangin, P., Fournier, G., Validire, P., Latil, A., and Cussenot, O. (2005). Differential expression of 37 selected genes in hormone-refractory prostate cancer using quantitative taqman real-time RT-PCR. Int. J. Cancer 114, 174–181. Fung, K.M., Samara, E.N., Wong, C., Metwalli, A., Krlin, R., Bane, B., Liu, C.Z., Yang, J.T., Pitha, J.V., Culkin, D.J., Kropp, B.P., Penning, T.M., and Lin, H.K. (2006). Increased expression of type 2 3alpha-hydroxysteroid dehydrogenase/type 5 17beta-hydroxysteroid dehydrogenase (AKR1C3) and its relationship with androgen receptor in prostate carcinoma. Endocr. Relat. Cancer 13, 169–180. Geller, J. (1985). Rationale for blockade of adrenal as well as testicular androgens in the treatment of advanced prostate cancer. Semin. Oncol. 12, 28–35. Geller, J. and Albert, J. (1987). Effects of castration compared with total androgen blockade on tissue dihydrotestosterone (DHT) concentration in benign prostatic hyperplasia (BPH). Urol. Res. 15, 151–153. Geller, J., Albert, J.D., Nachtsheim, D.A., and Loza, D. (1984). Comparison of prostatic cancer tissue dihydrotestosterone levels at the time of relapse following orchiectomy or estrogen therapy. J. Urol. 132, 693–696. Gingras, S. and Simard, J. (1999). Induction of 3beta-hydroxysteroid dehydrogenase/isomerase type 1 expression by interleukin-4 in human normal prostate epithelial cells, immortalized keratinocytes, colon, and cervix cancer cell lines. Endocrinology 140, 4573–4584. Guillemette, C., Hum, D.W., and Belanger, A. (1996). Regulation of steroid glucuronosyltransferase activities and transcripts by androgen in the human prostatic cancer LNCaP cell line. Endocrinology 137, 2872–2879. Harkonen, P., Torn, S., Kurkela, R., Porvari, K., Pulkka, A., Lindfors, A., Isomaa, V., and Vihko, P. (2003). Sex hormone metabolism in prostate cancer cells during transition to an androgenindependent state. J. Clin. Endocrinol. Metab. 88, 705–712. Harrington, W.R., Sengupta, S., and Katzenellenbogen, B.S. (2006). Estrogen regulation of the glucuronidation enzyme UGT2B15 in estrogen receptor-positive breast cancer cells. Endocrinology 147, 3843–3850. Heemers, H., Maes, B., Foufelle, F., Heyns, W., Verhoeven, G., and Swinnen, J.V. (2001). Androgens stimulate lipogenic gene expression in prostate cancer cells by activation of the sterol regulatory element-binding protein cleavage activating protein/sterol regulatory element-binding protein pathway. Mol. Endocrinol. 15, 1817–1828. Hodgson, M.C., Astapova, I., Hollenberg, A.N., and Balk, S.P. (2007). Activity of androgen receptor antagonist bicalutamide in prostate cancer cells is independent of NCoR and SMRT corepressors. Cancer Res. 67, 8388–8395. Holzbeierlein, J., Lal, P., LaTulippe, E., Smith, A., Satagopan, J., Zhang, L., Ryan, C., Smith, S., Scher, H., Scardino, P., Reuter, V., and Gerald, W.L. (2004). Gene expression analysis of human prostate carcinoma during hormonal therapy identifies androgen-responsive genes and mechanisms of therapy resistance. Am. J. Pathol. 164, 217–227. Iehle, C., Radvanyi, F., Gil Diez de, M., Ouafik, L.H., Gerard, H., Chopin, D., Raynaud, J.P., and Martin, P.M. (1999). Differences in steroid 5alpha-reductase iso-enzymes expression between normal and pathological human prostate tissue. J. Steroid Biochem. Mol. Biol. 68, 189–195. Ji, Q., Chang, L., Stanczyk, F.Z., Ookhtens, M., Sherrod, A., and Stolz, A. (2007). Impaired dihydrotestosterone catabolism in human prostate cancer: critical role of AKR1C2 as a prereceptor regulator of androgen receptor signaling. Cancer Res. 67, 1361–1369.

Increased Expression of Genes Converting Adrenal Androgens to Testosterone

137

Ji, Q., Chang, L., VanDenBerg, D., Stanczyk, F.Z., and Stolz, A. (2003). Selective reduction of AKR1C2 in prostate cancer and its role in DHT metabolism. Prostate 54, 275–289. Joyce, R., Fenton, M.A., Rode, P., Constantine, M., Gaynes, L., Kolvenbag, G., DeWolf, W., Balk, S., Taplin, M.E., and Bubley, G.J. (1998). High dose bicalutamide for androgen independent prostate cancer: effect of prior hormonal therapy. J. Urol. 159, 149–153. Komoto, J., Yamada, T., Watanabe, K., and Takusagawa, F. (2004). Crystal structure of human prostaglandin F synthase (AKR1C3). Biochemistry 43, 2188–2198. Krieg, M., Bartsch, W., Janssen, W., and Voigt, K.D. (1979). A comparative study of binding, metabolism and endogenous levels of androgens in normal, hyperplastic and carcinomatous human prostate. J. Steroid Biochem. 11, 615–624. Labrie, F., Luu-The, V., Lin, S.X., Simard, J., and Labrie, C. (2000). Role of 17 beta-hydroxysteroid dehydrogenases in sex steroid formation in peripheral intracrine tissues. Trends Endocrinol. Metab. 11, 421–427. Lin, H.K., Jez, J.M., Schlegel, B.P., Peehl, D.M., Pachter, J.A., and Penning, T.M. (1997). Expression and characterization of recombinant type 2 3 alpha-hydroxysteroid dehydrogenase (HSD) from human prostate: demonstration of bifunctional 3 alpha/17 beta-HSD activity and cellular distribution. Mol. Endocrinol. 11, 1971–1984. Lin, H.K., Steckelbroeck, S., Fung, K.M., Jones, A.N., and Penning, T.M. (2004). Characterization of a monoclonal antibody for human aldo-keto reductase AKR1C3 (type 2 3alpha-hydroxysteroid dehydrogenase/type 5 17beta-hydroxysteroid dehydrogenase); immunohistochemical detection in breast and prostate. Steroids 69, 795–801. Luo, J., Dunn, T.A., Ewing, C.M., Walsh, P.C., and Isaacs, W.B. (2003). Decreased gene expression of steroid 5 alpha-reductase 2 in human prostate cancer: implications for finasteride therapy of prostate carcinoma. Prostate 57, 134–139. Mahoney, E.M. and Harrison, J.H. (1972). Bilateral adrenalectomy for palliative treatment of prostatic cancer. J. Urol. 108, 936–938. Masiello, D., Cheng, S., Bubley, G.J., Lu, M.L., and Balk, S.P. (2002). Bicalutamide functions as an androgen receptor antagonist by assembly of a transcriptionally inactive receptor. J. Biol. Chem. 277, 26321–26326. Matsuura, K., Shiraishi, H., Hara, A., Sato, K., Deyashiki, Y., Ninomiya, M., and Sakai, S. (1998). Identification of a principal mRNA species for human 3alpha-hydroxysteroid dehydrogenase isoform (AKR1C3) that exhibits high prostaglandin D2 11-ketoreductase activity. J. Biochem. 124, 940–946. Miyamoto, H., Yeh, S., Lardy, H., Messing, E., and Chang, C. (1998). Delta5-androstenediol is a natural hormone with androgenic activity in human prostate cancer cells. Proc. Natl. Acad. Sci. U. S. A 95, 11083–11088. Mizokami, A., Koh, E., Fujita, H., Maeda, Y., Egawa, M., Koshida, K., Honma, S., Keller, E.T., and Namiki, M. (2004). The adrenal androgen androstenediol is present in prostate cancer tissue after androgen deprivation therapy and activates mutated androgen receptor. Cancer Res. 64, 765–771. Mohler, J.L., Gregory, C.W., Ford, O.H., III, Kim, D., Weaver, C.M., Petrusz, P., Wilson, E.M., and French, F.S. (2004). The androgen axis in recurrent prostate cancer. Clin. Cancer Res. 10, 440–448. Nishiyama, T., Hashimoto, Y., and Takahashi, K. (2004). The influence of androgen deprivation therapy on dihydrotestosterone levels in the prostatic tissue of patients with prostate cancer. Clin. Cancer Res. 10, 7121–7126. Page, S.T., Lin, D.W., Mostaghel, E.A., Hess, D.L., True, L.D., Amory, J.K., Nelson, P.S., Matsumoto, A.M., and Bremner, W.J. (2006). Persistent intraprostatic androgen concentrations after medical castration in healthy men. J. Clin. Endocrinol. Metab 91, 3850–3856. Pelletier, G., Luu-The, V., Tetu, B., and Labrie, F. (1999). Immunocytochemical localization of type 5 17beta-hydroxysteroid dehydrogenase in human reproductive tissues. J. Histochem. Cytochem. 47, 731–738.

138

S.P. Balk

Penning, T.M., Bauman, D.R., Jin, Y., and Rizner, T.L. (2007). Identification of the molecular switch that regulates access of 5alpha-DHT to the androgen receptor. Mol. Cell Endocrinol. 265–266, 77–82. Penning, T.M., Burczynski, M.E., Jez, J.M., Hung, C.F., Lin, H.K., Ma, H., Moore, M., Palackal, N., and Ratnam, K. (2000). Human 3alpha-hydroxysteroid dehydrogenase isoforms (AKR1C1AKR1C4) of the aldo-keto reductase superfamily: functional plasticity and tissue distribution reveals roles in the inactivation and formation of male and female sex hormones. Biochem. J. 351, 67–77. Penning, T.M., Steckelbroeck, S., Bauman, D.R., Miller, M.W., Jin, Y., Peehl, D.M., Fung, K.M., and Lin, H.K. (2006). Aldo-keto reductase (AKR) 1C3: role in prostate disease and the development of specific inhibitors. Mol. Cell Endocrinol. 248, 182–191. Ratnam, K., Ma, H., and Penning, T.M. (1999). The arginine 276 anchor for NADP(H) dictates fluorescence kinetic transients in 3 alpha-hydroxysteroid dehydrogenase, a representative aldoketo reductase. Biochemistry 38, 7856–7864. Rizner, T.L., Lin, H.K., Peehl, D.M., Steckelbroeck, S., Bauman, D.R., and Penning, T.M. (2003). Human type 3 3alpha-hydroxysteroid dehydrogenase (aldo-keto reductase 1C2) and androgen metabolism in prostate cells. Endocrinology 144, 2922–2932. Russell, D.W. and Wilson, J.D. (1994). Steroid 5 alpha-reductase: two genes/two enzymes. Annu. Rev. Biochem. 63, 25–61. Scher, H.I., Liebertz, C., Kelly, W.K., Mazumdar, M., Brett, C., Schwartz, L., Kolvenbag, G., Shapiro, L., and Schwartz, M. (1997). Bicalutamide for advanced prostate cancer: the natural versus treated history of disease. J. Clin. Oncol. 15, 2928–2938. Shang, Y., Myers, M., and Brown, M. (2002). Formation of the androgen receptor transcription complex. Mol. Cell 9, 601–610. Simard, J., Ricketts, M.L., Gingras, S., Soucy, P., Feltus, F.A., and Melner, M.H. (2005). Molecular biology of the 3beta-hydroxysteroid dehydrogenase/delta5-delta4 isomerase gene family. Endocr. Rev. 26, 525–582. Soderstrom, T.G., Bjelfman, C., Brekkan, E., Ask, B., Egevad, L., Norlen, B.J., and Rane, A. (2001). Messenger ribonucleic acid levels of steroid 5 alpha-reductase 2 in human prostate predict the enzyme activity. J. Clin. Endocrinol. Metab. 86, 855–858. Stanbrough, M., Bubley, G.J., Ross, K., Golub, T.R., Rubin, M.A., Penning, T.M., Febbo, P.G., and Balk, S.P. (2006). Increased expression of genes converting adrenal androgens to testosterone in androgen-independent prostate cancer. Cancer Res. 66, 2815–2825. Steckelbroeck, S., Jin, Y., Gopishetty, S., Oyesanmi, B., and Penning, T.M. (2004). Human cytosolic 3alpha-hydroxysteroid dehydrogenases of the aldo-keto reductase superfamily display significant 3beta-hydroxysteroid dehydrogenase activity: implications for steroid hormone metabolism and action. J. Biol. Chem. 279, 10784–10795. Swinnen, J.V., Ulrix, W., Heyns, W., and Verhoeven, G. (1997). Coordinate regulation of lipogenic gene expression by androgens: evidence for a cascade mechanism involving sterol regulatory element binding proteins. Proc. Natl. Acad. Sci. U. S. A 94, 12975–12980. Thomas, L.N., Douglas, R.C., Vessey, J.P., Gupta, R., Fontaine, D., Norman, R.W., Thompson, I. M., Troyer, D.A., Rittmaster, R.S., and Lazier, C.B. (2003). 5alpha-reductase type 1 immunostaining is enhanced in some prostate cancers compared with benign prostatic hyperplasia epithelium. J. Urol. 170, 2019–2025. Thomas, L.N., Lazier, C.B., Gupta, R., Norman, R.W., Troyer, D.A., O’Brien, S.P., and Rittmaster, R.S. (2005). Differential alterations in 5alpha-reductase type 1 and type 2 levels during development and progression of prostate cancer. Prostate 63, 231–239. Thomas, L.N., Douglas, R.C., Lazier, C.B., Too, C.K., Rittmaster, R.S., and Tindall, D.J. (2007). Type 1 and Type 2 5alpha-reductase expression in the development and progression of prostate cancer. Eur. Urol. 53, 244-252. Thompson, I.M., Goodman, P.J., Tangen, C.M., Lucia, M.S., Miller, G.J., Ford, L.G., Lieber, M. M., Cespedes, R.D., Atkins, J.N., Lippman, S.M., Carlin, S.M., Ryan, A., Szczepanek, C.M., Crowley, J.J., and Coltman, C.A., Jr. (2003). The influence of finasteride on the development of prostate cancer. N. Engl. J. Med. 349, 215–224.

Increased Expression of Genes Converting Adrenal Androgens to Testosterone

139

Titus, M.A., Gregory, C.W., Ford, O.H., III, Schell, M.J., Maygarden, S.J., and Mohler, J.L. (2005a). Steroid 5alpha-reductase isozymes I and II in recurrent prostate cancer. Clin. Cancer Res. 11, 4365–4371. Titus, M.A., Schell, M.J., Lih, F.B., Tomer, K.B., and Mohler, J.L. (2005b). Testosterone and dihydrotestosterone tissue levels in recurrent prostate cancer. Clin. Cancer Res. 11, 4653–4657. Weihua, Z., Lathe, R., Warner, M., and Gustafsson, J.A. (2002). An endocrine pathway in the prostate, ERbeta, AR, 5alpha-androstane-3beta,17beta-diol, and CYP7B1, regulates prostate growth. Proc. Natl. Acad. Sci. U. S. A 99, 13589–13594. Weihua, Z., Makela, S., Andersson, L.C., Salmi, S., Saji, S., Webster, J.I., Jensen, E.V., Nilsson, S., Warner, M., and Gustafsson, J.A. (2001). A role for estrogen receptor beta in the regulation of growth of the ventral prostate. Proc. Natl. Acad. Sci. U. S. A 98, 6330–6335. Wilson, J.D., Griffin, J.E., and Russell, D.W. (1993). Steroid 5 alpha-reductase 2 deficiency. Endocr. Rev. 14, 577–593. Zhu, P., Baek, S.H., Bourk, E.M., Ohgi, K.A., Garcia-Bassets, I., Sanjo, H., Akira, S., Kotol, P.F., Glass, C.K., Rosenfeld, M.G., and Rose, D.W. (2006). Macrophage/Cancer cell interactions mediate hormone resistance by a nuclear receptor derepression pathway. Cell 124, 615–629.

Androgen-Metabolic Genes in Prostate Cancer Predisposition and Progression Juergen K.V. Reichardt and Ann W. Hsing

Abstract Prostate cancer is the most common nonskin cancer and the second leading cause of cancer deaths among men in most Western countries, including the US and Australia. Despite its high morbidity and mortality, the etiology of prostate cancer remains elusive. Longstanding clinical and compelling laboratory data suggest a role for androgens in prostate carcinogenesis. This chapter reviews the status of research on hormones, particularly androgens, and prostate cancer and focuses first on hormone-related genetic loci in constitutional (‘‘germline’’) DNA. This set of loci has been investigated in a number of studies to date that will undoubtedly expand further. These data provide insights into susceptibility for prostate cancer. This review next explores the emerging field of somatic mutations in tumor (‘‘somatic’’) DNA in androgen-metabolic genes, especially the androgen receptor and the type II steroid 5a-reductase. Integration of these forthcoming data with those on susceptibility may provide novel insights into the etiology and progression of prostate cancer. These lines of investigation may lead to the presymptomatic identification of high-risk individuals for active disease prevention, diagnostic improvements in affected men and personalized treatment. Current and future data on individual markers and genes should be integrated into a comprehensive, pathway-based picture that includes constitutional DNA (for prostate cancer susceptibility) and tumor DNA (for disease progression). These efforts may lead to a comprehensive genetically based risk and progression assessment algorithm.

1 Introduction Prostate cancer is the most common nonskin cancer among American men, with 218,890 new cases expected in 2007 alone, and it ranks second in the number of estimated male cancer deaths, behind only lung cancer (Jemal et al. 2007). Despite the J.K.V. Reichardt(*) Plunkett Chair of Molecular Biology (Medicine), University of Sydney, Bosch Institute, Camperdown, NSW, 2006, Australia, E-mail: [email protected]

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4_6, # Springer Science + Business Media, LLC 2009

141

142

J.K.V. Reichardt, A.W. Hsing

magnitude of prostate cancer incidence and mortality, few risk factors have been identified definitively other than age, race, and family history of prostate cancer. Prostate cancer is a hormone-dependent cancer. Longstanding biological and medical evidence supports the fact that androgens play a role in the growth and maintenance of the prostate, that prostate cancer regresses after androgen-deprivation or antiandrogen therapy, and that administration of testosterone (T) induces prostate tumors in laboratory animals (Huggins and Hodges 1941; Noble 1977; Richie 1999). Estrogenic compounds have long been used to control prostate cancer growth, though recognition of their serious cardiac and sexual side effects has diminished their use (Harkonen and Makela 2004). Similarly, strong laboratory evidence shows that vitamin D, another hormone, has strong antiproliferative and proapoptotic effects on prostate cancer (Zhao and Feldman 2001). In addition, obesity, which has been linked to elevated risk of aggressive prostate cancer and prostate cancer death (Freedland et al. 2006), is associated with lower levels of sex-hormone-binding globulin (SHBG) and presumably higher levels of free T (Pasquali et al. 1991). Prostate carcinogenesis, like most if not all cancers, is a multistep process involving multiple genes in both the susceptibility, initiation and progression phases along with environmental factors, including diet (Hsing et al. 2002a). This chapter focuses on the current status and implications of androgen-metabolic genes and their contributions to the susceptibility and progression of prostate cancer.

2 Androgen Metabolism Androgens play a critical role in normal and malignant development of the prostate. Studies of androgens and prostate cancer go back over 60 years (beginning with Huggins and Hodges (1941)) and led to a Nobel Prize for Charles Huggins. Normal prostate development is induced by DHT (5a-dihydrotestosterone or dihydrotestosterone), which is synthesized from T by the enzyme steroid 5a-reductase (e.g., Hsing et al. 2002a, b; Coffey 1993; SRD5A; Fig. 1). T is produced in large amounts by the testes and small amount by the adrenal gland (e.g., Hsing et al. 2002a, b; Coffey 1993). T is irreversibly metabolized intracellularly to DHT (Fig. 1). Intracellularly, DHT is bound by a cytosolic receptor, the androgen receptor (AR; Fig. 1). This complex is translocated to the cell nucleus where it activates transcription of genes with hormone-responsive elements in their promoters (e.g., Hsing et al. 2002a; Coffey 1993). DHT can be inactivated in the prostate by further reduction (Fig. 1) to 3a- or 3b-androstanediol (or ‘‘diol’’; Fig. 1), which circulate as glucuronide conjugates. DHT homeostasis is regulated by its biosynthesis and degradation (see Fig. 1). Both ends involve multiple enzymatic steps. Those of most interest include the reactions catalyzed by the cytochrome P450 17a-hydroxylase (CYP17), 17b-hydroxysteroid dehydrogenase (HSD17B3), SRD5A2, 3b-hydroxysteroid dehydrogenase (HSD3B2), and UGT2B15 (encoding the glucuronidase) gene products (Fig. 1). DHT binds directly to the AR to exert its signal transduction activity (Fig. 1).

6 Androgen-etabolic Genes in Prostate Cancer

143

Outside the Prostate

Within the Prostate 3β-Diol G 3α-Diol G

5-androstenediol CYP17 gene

HSD3B1/2/AKR gene

HSD3B1/2 gene

17α-Hydroxylase/ 17,20-Lyase

3β-Hydroxysteroid Dehydrogenase

3β-Hydroxysteroid Dehydrogenase

Preg

DHEA

Androstenedione

T

3β-Diol

HSD3B

T

17β-Hydroxysteroid Dehydrogenase

UGT2B 3α-Diol HSD3A/AKR

DHT Steroid 5α-Reductase

Androgen Receptor AR gene

SRD5A2 gene

HSD17B3 gene

DHT/AR CYP3A4/5 gene Complex OH-T

Transactivation Cell Division

Fig. 1 Androgen metabolism and androgen-metabolic genes 0. The abbreviations used are: preg pregnelone, DHEA dehydroepiandrosterone, DHT dihydrotestosterone, diol androstanediol, diol G androstanediol glucuronide, and OH-T hydroxy-testosterone

Animal studies have shown that androgen, in particular DHT, plays a pivotal role in prostate cancer development (Noble 1977); enzymes modulating DHT levels are usually targeted for the design of antiprostate cancer drugs. Finasteride is the first drug that was used in the chemoprevention of prostate cancer (Thompson et al. 2003). It acts by inhibiting the enzyme steroid 5a-reductase type II (encoded by the SRD5A2 gene; Fig. 1), hence blocking DHT synthesis from testosterone (T; Fig. 1). In the Prostate Cancer Prevention Trial (PCPT) finasteride was tested in healthy men 55 years of age or older. The prevention study resulted in a significant (25%) decrease in prostate cancer incidence over the seven-year study period (Thompson et al. 2003).

3 Androgen-Metabolic Genes Prostate cancer is an androgen-dependent tumor. Therefore, it is likely that genes whose products are involved in androgen metabolism are involved in prostate cancer etiology. Ross et al. (1998) first proposed a polygenic model to help explain the racial/ethnic difference in prostate cancer risk. This model triggered a series of studies that investigated the involvement of genes such as CYP17, aromatase (CYP19), steroid 5a-reductase (SRD5A2), (HSD3B2), and androgen receptor (AR) in prostate cancer (cf. Fig. 1). These candidate genes and their effects in constitutional (‘‘germline’’) DNA on risk are considered first (cf. Table 1). Data on acquired somatic mutations (i.e., de novo mutations in tumor tissue) are also accumulating (cf. Table 1). We will summarize and integrate data on constitutional and somatic DNA in androgen-metabolic genes here into a comprehensive picture of the combined role of these genes in prostate cancer.

144

J.K.V. Reichardt, A.W. Hsing

Table 1 Characteristics of androgen-metabolic genes analyzed in prostate cancer Chromosomal Constitutional DNA Somatic Gene Gene product location SNPs mutations Androgen biosynthesis CYP17 Cytochrome p450c17 10q24.3 Reviewed in None reported Chokkalingam et al. (2007) T metabolism SRD5A2 Steroid 5a-reductase

2p23

Reviewed in Chokkalingam et al. (2007)

CYP3A4

7q22.1

Reviewed in Chokkalingam et al. (2007) Reviewed in Chokkalingam et al. (2007)

CYP3A5

cytochrome P450, subfamily 3A, polypeptide 4 cytochrome P450, subfamily 3A, polypeptide 5

7q22.1

DHT metabolism HSD3B2 3B-hydroxysteroid dehydrogenase

1p13

HSD17B3

17B-hydroxysteroid dehydrogenase

9q22

SULT

sulfotransferase

various

UGT2B15

UDPglycosyltransferase

4q13

Androgen signaling AR Androgen receptor

Reviewed in Chokkalingam et al. (2007) Reviewed in Chokkalingam et al. (2007) Reviewed in Culig et al. (1997) Reviewed in Reichardt (2006)

Penning et al. (2004) and Fung et al. (2006) None reported

None reported

None reported

None reported

None reported None reported

Reviewed in Reviewed in Chokkalingam Culig et al. et al. (2007) and (1997) Culig et al. (1997) AIB1 Nuclear receptor 20q12 Reviewed in Culig None reported coactivator et al. (1997) Notes: The numbers in this table indicate references. The presence of constitutional and/or somatic mutations refers to publications referring to or reviewing data on DNA from either lymphocytes (constitutional or ‘‘germline’’ DNA) or prostate tumor tissue (somatic DNA). Absence of publications, e.g., on somatic mutations, may only indicate that this gene has not yet been investigated and hence somatic mutations may be found in the future

3.1

Xq11-12

CYP17

Cytochrome P450c17a hydroxylase, encoded by the CYP17 gene located in chromosome band 10q24.3, catalyzes early steps in the biosynthesis of testosterone (see Fig. 1; Table 1). A single nucleotide polymorphism (SNP), that changes the normal T to the polymorphic C, also known as the A2 allele in the 50 -untranslated region (UTR) of the CYP17 gene, has been linked to male pattern baldness (Carey et al.

6 Androgen-etabolic Genes in Prostate Cancer

145

1994), a putative risk factor for prostate cancer. However, the relationship between CYP17 and prostate cancer is inconclusive, which was recently reviewed in (Chokkalingam et al. 2007). These somewhat inconsistent results suggest that the effect of the CYP17 gene on prostate cancer, if any, is likely small. No reports on somatic mutations in this gene are available.

3.2

SRD5A2

The SRD5A2 gene, which encodes the prostatic or type II steroid 5a-reductase enzyme (Fig. 1; Table 1), came to the forefront of prostate cancer research after cross-sectional surveys showed that Western men have higher serum levels of the DHT metabolite, 3a-diol glucuronide (cf. Fig. 1), than native Japanese men (Ross et al. 1992). Serum hormone levels of 3a-diol glucuronide may reflect prostatic steroid 5a-reductase activity. These observations led to the hypothesis that population differences in steroid 5a-reductase activity, encoded by polymorphisms in the SRD5A2 gene, may be related to the development of prostate cancer and may contribute, at least in part, to racial/ethnic differences in risk (Makridakis et al. 1999, 2000). More than 22 mutations, including ten single amino acid missense substitutions, have been reported for the SRD5A2 gene (Makridakis et al. 2000) (Table 1). Four of these SNPs – A49T (substitution of threonine for the normal alanine at codon 49), V89L (replacement of leucine with valine at codon 89), R227Q (arginine-227 to glutamine), and a (TA)n dinucleotide repeat in the 30 UTR- have been investigated for their association with prostate cancer in 18 epidemiologic studies that have produced mixed results: see Chokkalingam et al. (2007) for a recent review. A case series reported that the A49T genotype was associated with more aggressive prostate cancer (Chokkalingam et al. 2007). However, of the 11 studies investigating the A49T marker in the SRD5A2 gene, which increases enzyme activity fivefold in vitro (Makridakis et al. 2000), just one reported a statistically significant association between the higher activity polymorphic variant and prostate cancer, and this association was only significant among African-American and Latino men (Makridakis et al. 1999). The null results of most epidemiologic studies may be due to the low prevalence of the variant T allele (about 1% in most populations), despite the large difference in enzymatic activity that results from this nucleotide change. In general, the results of the studies investigating the prostate cancer associations of the V89L marker (Chokkalingam et al. 2007), which modestly reduces enzyme activity modestly in vitro and in vivo (Makridakis et al. 2000), and the (TA)n repeat length marker (Chokkalingam et al. 2007), which has no known functional relevance, are mixed. The V89L polymorphic LL genotype, which is associated with lower serum levels of 3a-diol glucuronide (Makridakis et al. 2000), occurs in Asian men much more frequently than in African-American and Caucasian men. The R227Q mutation, which is also found in Asian male pseudohermaphrodites and almost abolishes enzyme activity (Makridakis et al. 2000), has been detected only

146

J.K.V. Reichardt, A.W. Hsing

in Asians thus far, and the only study investigating its role found no significant association with prostate cancer risk (Chokkalingam et al. 2007). The inconsistent findings in the SRD5A2 constitutional DNA markers in various epidemiologic studies for prostate cancer susceptibility may be due largely to the low frequency of most polymorphic alleles in the SRD5A2 gene. For example, other than the V89L mutation, the frequency of the polymorphic alleles in various markers (including A49T and R227Q) is less than 5%, limiting the power of detection. Nonetheless, a recent meta-analysis by Ntais et al. concluded that the polymorphic T allele of the A49T SNP in the SRD5A2 gene is associated with a modest risk increase, and that, although the evidence is less consistent, the LL genotype of the V89L polymorphism and longer (TA)n repeats may also be associated with increased risk for prostate cancer (Ntais et al. 2003). Although the epidemiologic data are less than convincing, the SRD5A2 gene and its variants are among the best characterized in cancer biology. Enzyme activity varies in vitro by about 200-fold variation, and there is also a 60-fold pharmacogenetic variation for finasteride among SNPs (Makridakis et al. 2000). This extensive variability is likely to have important consequences for the risk of prostate cancer, and it may also be relevant in active, personalized disease prevention, e.g., in the context of the PCPT (Thompson et al. 2003). Finally, detailed in vitro biochemical and pharmacogenetic investigations of SRD5A2 haplotypes (i.e., combinations of multiple SNPs) have been reported (Reichardt 2006). Such studies may become significant in the future as haplotypes move to the forefront of molecular epidemiology and pharmacogenetics. Somatic mutation of the SRD5A2 gene has been investigated in prostate cancer tissue (Akalu et al. 1999; Makridakis et al. 2004) (Table 1). These investigations have uncovered common somatic mutations in the SRD5A2 gene in prostate cancer, and their in vitro biochemical characterization (Makridakis et al. 2004) is consistent with driving tumor growth through increased DHT synthesis. These early encouraging data on somatic DNA remain to be confirmed. Furthermore, these data on somatic mutations should be integrated fully with data on constitutional DNA to develop a comprehensive picture of the role of the SRD5A2 gene in prostate cancer susceptibility and disease progression.

3.3

HSD Genes

Incomplete activation or slow catabolism of DHT within the prostate could lead to the accumulation of DHT and, perhaps, increased androgenic action (cf. Fig. 1). Thus, enzymes that catalyze the activation and inactivation of DHT may be of potential etiologic importance for prostate cancer. As shown in Fig. 1, at least four enzymes – 17b-hydroxysteroid dehydrogenase type I (encoded by the HSD17B1 gene), 17b-hydroxysteroid dehydrogenase type III (encoded by the HSD17B3 gene), 3a-hydroxysteroid dehydrogenase, and the 3b-hydroxysteroid dehydrogenases (encoded by the HSD3B1 and HSD3B2 genes located in chromosome

6 Androgen-etabolic Genes in Prostate Cancer

147

band 1p13.1) – may be involved in the catabolism of DHT in the prostate (Table 1). Three studies have examined these genes. One reported a significant association of the G289S polymorphism in the HSD17B3 gene (Margiotti et al. 2002) and another reported an association of the N367T marker within the HSD3B1 gene, but no prostate cancer associations of the C7062T marker in the HSD3B1 gene and the C7159G and C7474T SNPs in HSD3B2 (Chang et al. 2002). These early results are evidence for a role of these genes in prostate cancer risk, but further studies are needed. In contrast, the most recent study, a comprehensive examination of the HSD17B1 gene, examined cases and controls from a consortium of cohort studies (Kraft et al. 2005). A large sample size (>9,000 cases and 9,000 controls), multiple ethnicities investigated, and utilization of haplotype-tagging SNPs rather than single SNPs provide strong evidence for a lack of association of the HSD17B1 gene with prostate cancer risk. Finally, the list of genes involved in DHT inactivation (Fig. 1) may still be incomplete. The Aldo-Keto Reductase (AKR) 1C gene family includes AKR1C1, AKR1C2, AKR1C3, and AKR1C4 (Ji et al. 2007). These four genes share a high sequence homology (>80%). The AKR1C2, AKR1C3, and AKR1C4 genes are also known as 3a-hydroxysteroid dehydrogenase (HSD) III, HSD II, and HSDI, respectively. The gene product of AKR1C2 synthesizes the DHT metabolite, 3 a-diol, which is a more water-soluble metabolite, for elimination from the prostate (Ji et al. 2003; Penning et al. 2004; Fung et al. 2006). Increased expression of AKRC3 in prostate cancer cells has been reported (Fung et al. 2006). Few epidemiologic studies have investigated the role of AKR genes, but molecular studies have shown that expression of AKR1C2 was associated with higher tissue levels of DHT (Penning et al. 2004). Thus, further investigations of the AKR genes are warranted in epidemiologic analyses. Finally, no publications to date have systematically addressed these loci in prostate cancer tissue.

3.4

CYP3A Genes

The CYP3A (cytochrome P450, family 3, subfamily A) subfamily members CYP3A4 and CYP3A5 are located at a single locus in human chromosome band 7q22.1 (Table 1). The CYP3A4 gene product is involved in the oxidative conversion of testosterone to the biologically less active forms 2b-, 6b-, and 15b-hydroxytestosterone (Fig. 1). The variant *1B allele (of the A391G SNP in the 50 UTR of CYP3A4) has been linked to higher prostate cancer stage and grade (Rebbeck et al. 1998). However, while one case-control study showed significant excess risk of aggressive disease with the *1B allele among African-Americans, it did not find this risk among Caucasians (Bangsi et al. 2006). Several other studies, which were reviewed recently in Chokkalingam et al. (2007), reported a modest but nonsignificant reduced risk of prostate cancer in connection with the variant *1B allele. The low frequency of the polymorphic variant in most populations limited the statistical

148

J.K.V. Reichardt, A.W. Hsing

power of many of these studies. The CYP3A4 gene product may also be involved in the metabolism of finasteride and hence have relevance for prostate cancer chemoprevention, e.g., via the PCPT (Thompson et al. 2003). The CYP3A5 gene product, which is highly expressed in the prostate, is a major contributor to the metabolic clearance of many CYP3A substrates, including drugs and other xenobiotic compounds, as well as steroid hormones. A number of polymorphic markers have been identified in the CYP3A5 gene, two of which have been studied already in epidemiologic investigations of prostate cancer. The CYP3A5*1 marker, whose variant allele may promote inactivation of testosterone, was found to have a borderline significant association with reduced risk of prostate cancer in one study but not another (see Chokkalingam et al. 2007). The variant allele of the CYP3A5*3 marker, which results in a truncated and presumably less functional protein (reviewed recently in Ntais et al. 2003), supports the observation in a Japanese population that the CYP3A5*3 wild-type allele was associated with reduced prostate cancer risk (reviewed in Chokkalingam et al. 2007). To date, no systematic mutational investigations have been reported of the two CYP3A genes in prostate tumor tissue.

3.5

UGT Genes

Conjugation with UDP (uridinediphosphate) by UGT (UDP-glycosyltransferase) enzymes (i.e., a glucuronidation reaction) is one of the major routes of elimination of endogenous compounds. UGT2 B15 and B17 enzymes are expressed in the prostate, in particular in androgen-sensitive cell lines (Belanger et al. 1998). In vitro, these enzymes conjugate DHT and its metabolites, 3 alpha diol and androsterone, into a less active form (Belanger et al. 1998). The glucuronidation by UGT2B15 and B17 is an irreversible process and an important step in DHT catabolism. Data on the role of UGT2B15 and B17 in prostate cancer are limited, with three recent studies showing that the variant allele of the UGT2B15 gene, which resulted in higher transcription, was associated with a higher risk of prostate cancer (Chouinard et al. 2007; Park et al. 2004, 2006; Gallagher et al. 2007; Dalhoff et al. 2005).

3.6

SULT Genes

Sulfation is an important steroid inactivation process in human tissue. To date, 11 human cytosolic sulfotransferase (SULT) isoforms have been identified (Wilborn et al. 2006), with SULT2A1 being the major enzyme responsible for sulfating dehydroepiandrosterone (DHEA), a precursor of T. A recent series of studies revealed that sulfotransferase SULT2B1b selectively conjugates 3b-hydroxysteroids and is expressed in epithelial cells of normal and cancerous prostate tissues (He et al. 2004; He and Falany 2007; Table 1). Despite the importance of

6 Androgen-etabolic Genes in Prostate Cancer

149

these genes in steroid inactivation, data on these genes remain limited. Further studies are warranted, including constitutional DNA and tumor DNA.

4 Androgen-Signaling Genes 4.1

Androgen Receptor

Currently the most studied gene with regard to prostate cancer is that encoding the AR (Table 1). The AR is expressed in all histological types and stages of prostate cancer (Culig et al. 1997). Interestingly much of the focus in the AR has been on somatic mutations. In fact, numerous somatic mutations in the AR gene, located on the long arm of the X chromosome, have been reported in prostate cancer (Table 1). Most of these mutations have been detected in tumor tissue of late stage prostate cancer, with consistent findings showing that somatic mutation of the AR gene is involved in the progression and aggressiveness of prostate cancer (Culig et al. 1997). Most epidemiologic studies of the AR gene in constitutional DNA have focused on two trinucleotide repeat polymorphisms in exon 1, the (CAG)n and (GGN)n (where n stands for any nucleotide) repeats, which encode polyglutamine and polyglycine, respectively (Table 1). The (CAG)n repeat length ranges from 11 to 31 repeats in most men, and experimental data show that shorter (CAG)n repeat lengths are linked to increased AR trans-activation activity (Chamberlain et al. 1994; Edwards et al. 1992). In addition, (CAG)n repeat length has been linked to male pattern baldness (Sawaya and Shalita 1998; Hoffmann and Happle 2000), a medical condition associated with higher DHT levels and prostate cancer risk. Epidemiologic studies show, that in general, shorter (CAG)n repeat lengths are associated with increased risk in many but not all populations (Chokkalingam et al. 2007) (for a review). Evidence of association for the second AR repeat, (GGN)n, is less consistent than that reported for the (CAG)n repeat. The AR protein has a modular structure, consisting primarily of (a) a large amino-terminal transactivation domain, (b) a centrally located DNA-binding domain, and (c) a carboxy-terminal ligand-binding domain that confers high affinity and specificity for androgen binding. AR gene mutations have been reported in prostate cancer at a frequency of 5–50%, with a higher frequency reported in more advanced disease and following androgen-deprivation therapy, reviewed recently in Makridakis et al. (2005). In contrast to the inherited syndrome of androgen insensitivity, where AR somatic mutations result in a loss of receptor function, those detected in prostate cancer occur in distinctly different regions of the receptor, and generally confer promiscuity or increased sensitivity for receptor activation by other steroid hormones and/or by the specific antiandrogen used in clinical management of the disease (Makridakis et al. 2005). These observations led to the concept of ‘‘therapy-mediated selection pressure,’’ which is exemplified by the detection of AR gene mutations that confer enhanced receptor activity in response

150

J.K.V. Reichardt, A.W. Hsing

to flutamide but not to other antiandrogens in tumors from patients treated with hydroxyflutamide plus orchidectomy or LHRH agonists/antagonists, as part of a combined androgen blockade strategy (Makridakis et al. 2005). The rapidly evolving evidence that the AR is a key mediator of continued prostate cancer growth during androgen deprivation makes a compelling argument for the development of new strategies that directly target the receptor (Makridakis et al. 2005). Such new therapies are necessary for significant advances in the management of patients with advanced prostate cancer who will ultimately fail androgen-deprivation therapy. Several approaches to achieve this goal have been documented. Reducing the level of AR and suppressing human prostate cancer cell growth in vitro and in vivo used a variety of methods, including double-stranded RNA interference, antisense oligonucleotides, hammerhead ribozymes, and analogs of the ansamycin antibiotics such as 17-allylamino-17-demethoxygeldanamycin (Makridakis et al. 2005). Inhibition of AR function has been achieved with dominant negative AR inhibitors, microinjection of AR antibodies, ‘‘decoy’’ double-stranded DNA fragments containing specific AR response elements, and histone deacetylase inhibitors including suberoylanilide hydroxamic acid and phenylbutyrate (Makridakis et al. 2005). Provided issues such as the mode of delivery and the potential for disruption of multiple signaling pathways can be circumvented, these newer approaches that directly target the AR provide promise as therapeutic options for castration-resistant prostate cancer.

4.2

AR Coactivators

AR coactivators enhance transactivation of the AR several-fold (Yeh and Chang 1996) and therefore may increase the risk of prostate cancer as well. One AR coactivator, encoded by the AIB1 (Amplified in Breast Cancer 1) gene located in chromosome band 20q12, has a (CAG)n/(CAA)n (polyglutamine) trinucleotide repeat. Two epidemiologic studies have investigated the role of this repeat polymorphism in prostate cancer: one found a positive association between AIB1 (CAG)n repeat length and prostate cancer (Hsing et al. 2002b), and the other reported no association (Platz et al. 2000; Table 1). Future studies should be directed at clarifying the combined effects of AR and AR coactivators in prostate cancer risk, because of their importance in modulating the effects of androgenic action.

5 Current Status and Implications for Androgen-Metabolic Genes in Prostate Cancer Much current clinical, epidemiological, and molecular evidence support a central role for androgen-metabolic genes in prostate cancer etiology. Current data are very strong for the AR and fairly supportive on the biosynthetic end of the pathway

6 Androgen-etabolic Genes in Prostate Cancer

151

(leading to DHT; Fig. 1). These investigations should be extended to DHT inactivation which is currently underinvestigated. These combined studies (such as those proposed in Ross et al. (1998)) may be useful in identifying at-risk men presymptomatically and for developing personalized chemopreventive strategies. The evidence on somatic mutations is only now beginning to emerge, but it will probably play a significant role in explaining disease progression in the near future. Finally, the heterogeneity (or multifocality) of prostate tumors is an underappreciated and under-reported phenomenon that deserves more attention (Platz et al. 2000).

6 Future Directions Although much progress has been made on androgen-metabolic genes and prostate cancer, some areas deserve more attention in the future. First, DHT inactivation appears understudied to date (cf. Fig. 1). This area is critical to fully assess DHT homeostasis. Second, full functional (e.g., biochemical and pharmacogenetic) analyses of SNPs and especially haplotypes of all androgen-metabolic genes will be useful for rational marker selection. Third, prostatic tumor tissue should be investigated systematically for the emergence of somatic mutations in all androgenmetabolic genes (Fig. 1). Fourth, the full impact of chemopreventive strategies needs to be rigorously assessed molecularly. Finally, data on individual markers and genes must be fully integrated into a comprehensive, pathway-based picture that includes constitutional DNA (for prostate cancer susceptibility) and tumor DNA (for disease progression). This strategy will lead to an overall genetically based risk and progression assessment algorithm. Considerable work remains to achieve this goal. These integrated investigations will lead to a full picture of the multistep nature of prostate cancer development and progression (e.g., Hsing et al. 2002a, b; Ntais et al. 2003; Reichardt 2006; Dalhoff et al. 2005).

Acknowledgments This work was supported in part by NCI grant P01 CA108964 (project 1) to JKVR who is also a Medical Foundation Fellow at the University of Sydney and by NIH Intramural Support (AWH).

References Akalu, A., Elmajian, D. A., Highshaw, R. A., Nichols, P. W., Reichardt, J. K. V.: Somatic mutations at the SRD5A2 locus encoding prostatic steroid 5a-reductase during prostate cancer progression. Journal of Urology, 161: 1355–1358, 1999. Bangsi, D., Zhou, J., Sun, Y., Patel, N. P., Darga, L. L., Heilbrun, L. K., Powell, I. J., Severson, R. K., Everson, R. B.: Impact of a genetic variant in CYP3A4 on risk and clinical presentation of prostate cancer among white and African-American men. Urologic Oncology, 24: 21–27, 2006.

152

J.K.V. Reichardt, A.W. Hsing

Belanger, A., Hum, D. W., Beaulieu, M., Levesque, E., Guillemette, C., Tchernof, A., Belanger, G., Turgeon, D., Dubois, S.: Characterization and regulation of UDP-glucuronosyltransferases in steroid target tissues. The Journal of Steroid Biochemistry and Molecular Biology, 65: 301–310, 1998. Carey, A. H., Waterworth, D., Patel, K., White, D., Little, J., Novelli, P., et al. Polycystic ovaries and premature male pattern baldness are associated with one allele of the steroid metabolism gene CYP17. Human Molecular Genetics, 3: 1873–1876, 1994. Chamberlain, N. L., Driver, E. D., Miesfeld, R. L.: The length and location of CAG trinucleotide repeats in the androgen receptor N-terminal domain affect transactivation function. Nucleic Acids Research, 22: 3181–3186, 1994. Chang, B. L., Zheng, S. L., Hawkins, G. A., Isaacs, S. D., Wiley, K. E., Turner, A., Carpten, J. D., Bleecker, E. R., Walsh, P. C., Trent, J. M., Meyers, D. A., Isaacs, W. B., Xu, J.: Joint effect of HSD3B1 and HSD3B2 genes is associated with hereditary and sporadic prostate cancer susceptibility. Cancer Research, 62: 1784–1789, 2002. Chokkalingam, A. P., Stanczyk, F. Z., Reichardt, J. K. V., Hsing A. W.: Molecular Epidemiology of Prostate Cancer: Hormone-Related Genetic Loci. Frontiers in Bioscience, 12: 3436–60, 2007. Chouinard, S., Barbier, O., Belanger, A.: UDP-glucuronosyltransferase (UGT)2B15 and UGT2B17 enzymes are major determinants of the androgen response in prostate cancer LNCaP cells. The Journal of Biological Chemistry, 282: 33466–33474, 2007; [Epub ahead of print] Coffey, D. S.: The Molecular Biology of the Prostate. In: Prostate Diseases (Lepor, H. and Lawson, R. K., eds.), WB Saunders, Philadelphia, PA, 28–56, 1993. Culig, Z., Hobisch, A., Hittmair, A., Peterziel, H., Radmayr, C., Bartsch, G., et al. Hyperactive androgen receptor in prostate cancer: what does it mean for new therapy concepts? Histology and Histopathology, 12: 781–786, 1997. Dalhoff, K., Buus Jensen, K., Enghusen Poulsen, H.: Cancer and molecular biomarkers of phase 2. Methods Enzymology, 400: 618–627, 2005. Edwards, A., Hammond, H. A., Jin, L., Caskey, C. T., Chakraborty, R.: Genetic variation at five trimeric and tetrameric tandem repeat loci in four human population groups. Genomics, 12: 241–253, 1992. Freedland, S. J., Giovannucci, E., and Platz, E. A.: Are findings from studies of obesity and prostate cancer really in conflict? Cancer Causes Control, 17: 5–9, 2006. Fung, K. M., Samara, E. N., Wong, C., Metwalli, A., Krlin, R., Bane, B., Liu, C. Z., Yang, J. T., Pitha, J. V., Culkin, D. J., Kropp, B. P., Penning, T. M., Lin, H. K.: Increased expression of type 2 3alpha-hydroxysteroid dehydrogenase/type 5 17beta-hydroxysteroid dehydrogenase (AKR1C3) and its relationship with androgen receptor in prostate carcinoma. Endocrine Related-Cancer, 13: 169–180, 2006. Gallagher, C. J., Kadlubar, F. F., Muscat, J. E., Ambrosone, C. B., Lang, N. P., Lazarus, P.: The UGT2B17 gene deletion polymorphism and risk of prostate cancer A case-control study in Caucasians. Cancer Detection and Prevention, 31: 310–315, 2007. Harkonen, P. L. and Makela, S. I.: Role of estrogens in development of prostate cancer. The Journal of Steroid Biochemistry and Molecular Biology, 92: 297–305, 2004. He, D, Falany, C. N.: Inhibition of SULT2B1b expression alters effects of 3beta-hydroxysteroids on cell proliferation and steroid hormone receptor expression in human LNCaP prostate cancer cells. Prostate, 67: 1318–1329, 2007. He, D., Meloche, C. A., Dumas, N. A., Frost, A. R., Falany, C. N.: Different subcellular localization of sulphotransferase 2B1b in human placenta and prostate. Biochemistry Journal 379: 533–40, 2004. Hoffmann, R., Happle, R.: Current understanding of androgenetic alopecia. Part I: etiopathogenesis. European Journal of Dermatology, 10: 319–327, 2000. Hsing, A. W., Reichardt, J. K., and Stanczyk, F. Z.: Hormones and prostate cancer: current perspectives and future directions. Prostate, 52: 213–235, 2002a.

6 Androgen-etabolic Genes in Prostate Cancer

153

Hsing, A. W., Chokkalingam, A. P., Gao, Y. T., Wu, G., Wang, X., Deng, J. et al: Polymorphic CAG/CAA repeat length in the AIB1/SRC-3 gene and prostate cancer risk: a population-based case-control study. Cancer Epidemiology Biomarkers and Prevention, 11: 337–341, 2002b. Huggins, C. and Hodges, C. V.: Studies on prostatic cancer: Effect of castration, of estrogen, and of androgen injection on serum phosphatases in metastatic carcinoma of the prostate. Cancer Research, 1: 293–297, 1941. Jemal, A., Siegel, R., Ward, E., Murray, T., Xu, J., Smigal, C. and Thun, M. J.: Cancer statistics, 2007. CA: A Cancer Journal for Clinicians, 57: 43–66, 2007. Ji, Q., Chang, L., VanDenBerg, D., Stanczyk, F. Z., Stolz, A.: Selective reduction of AKR1C2 in prostate cancer and its role in DHT metabolism. Prostate, 2003; 54: 275–289. Erratum in: Prostate, 66: 445, 2006. Ji, Q., Chang, L., Stanczyk, F. Z., Ookhtens, M., Sherrod, A., Stolz, A.: Impaired dihydrotestosterone catabolism in human prostate cancer: critical role of AKR1C2 as a pre-receptor regulator of androgen receptor signaling. Cancer Research, 67: 1361–1369, 2007. Kraft, P., Pharoah, P., Chanock, S. J., Albanes, D., Kolonel, L. N., Hayes, R. B., Altshuler, D., Andriole, G., Berg, C., et al. Genetic variation in the HSD17B1 gene and risk of prostate cancer. PLoS Genet, 1: e68, 2005. Makridakis, N. M., Ross, R. K., Pike, M. C., Crocitto, L. E., Kolonel, L. N., Pearce, C. L., Henderson, B. E., J. K. Reichardt: Association of mis-sense substitution in SRD5A2 gene with prostate cancer in African-American and Hispanic men in Los Angeles, USA. Lancet, 354: 975–978, 1999. Makridakis, N. M., di Salle E., Reichardt, J. K.: Biochemical and pharmacogenetic dissection of human steroid 5 alpha-reductase type II. Pharmacogenetics, 10: 407–413, 2000. Makridakis, N. M., Akalu, A., Reichardt, J. K. V.: Identification and Characterization of Somatic Steroid 5a-Reductase (SRD5A2) Mutations in Human Prostate Cancer Tissue, Oncogene 23: 7399–7405, 2004. Makridakis, N. M., Buchanan, G., Tilley, W., Reichardt, J. K. V.: Androgen Metabolic Genes in Prostate Cancer Predisposition and Progression, Frontiers in Bioscience, 10: 2899–2903, 2005. Margiotti, K., Kim, E., Pearce, C. L., E. Spera, G. Novelli, J. K. Reichardt: Association of the G289S single nucleotide polymorphism in the HSD17B3 gene with prostate cancer in Italian men. Prostate, 53: 65–68, 2002. Noble, R. L.: The development of prostatic adenocarcinoma in Nb rats following prolonged sex hormone administration. Cancer Research, 37: 1929–1933, 1977. Ntais, C., Polycarpou, A., Ioannidis, J. P.: SRD5A2 gene polymorphisms and the risk of prostate cancer: a meta-analysis. Cancer Epidemiology Biomarkers and Prevention, 12: 618–624, 2003. Park, J., Chen, L., Shade, K., Lazarus, P., Seigne, J., Patterson, S., Helal, M., Pow-Sang, J.: Asp85tyr polymorphism in the udp-glucuronosyltransferase (UGT) 2B15 gene and the risk of prostate cancer. Journal of Urology, 171(6 Pt 1): 2484–2488, 2004. Park, J., Chen, L., Ratnashinge, L., Sellers, T. A., Tanner, J. P., Lee, J. H., Dossett, N., Lang, N., Kadlubar, F. F., Ambrosone, C. B., Zachariah, B., Heysek, R. V., Patterson, S., Pow-Sang, J.: Deletion polymorphism of UDP-glucuronosyltransferase 2B17 and risk of prostate cancer in African American and Caucasian men. Cancer Epidemiology Biomarkers and Prevention, 15: 1473–1478, 2006. Pasquali, R., Casimirri, F., Cantobelli, S., Melchionda, N., Morselli Labate, A. M., Fabbri, R. et al: Effect of obesity and body fat distribution on sex hormones and insulin in men. Metabolism, 40: 101–104, 1991. Penning, T. M., Jin, Y., Steckelbroeck, S., Lanisnik Rizner, T., Lewis, M.: Structure-function of human 3 alpha-hydroxysteroid dehydrogenases: genes and proteins. Molecular and Cellular Endocrinology, 215: 63–72, 2004. Platz, E. A., Giovannucci, E., Brown, M., Cieluch, C., Shepard, T. F., Stampfer, M. J., et al. Amplified in breast cancer-1 glutamine repeat and prostate cancer risk. Prostate Journal, 2: 27– 32, 2000.

154

J.K.V. Reichardt, A.W. Hsing

Rebbeck, T. R., Jaffe, J. M., Walker, A. H., Wein, A. J., Malkowicz, S. B.: Modification of clinical presentation of prostate tumors by a novel genetic variant in CYP3A4. Journal of National Cancer Institute, 90: 1225–1229, 1998. Reichardt, J. K. V.: Unexpected biochemical and pharmacogenetic consequences of SNPs and haplotypes: A cautionary tale for human molecular genetics and epidemiology. Genomics, 88: 673–674, 2006. Richie, J. P.: Anti-androgens and other hormonal therapies for prostate cancer. Urology, 54: 15–18, 1999. Ross, R. K., Bernstein, L., Lobo, R. A., Shimizu, H., Stanczyk, F. Z., Pike, M. C. & Henderson, B. E.: 5-alpha-reductase activity and risk of prostate cancer among Japanese and US white and black males. Lancet, 339: 887–889, 1992. Ross, R. K., Pike, M. C., Coetzee, G. A., Reichardt, J. K., Yu, M. C., Feigelson, H., et al: Androgen metabolism and prostate cancer: establishing a model of genetic susceptibility. Cancer Research, 58: 4497–4504, 1998. Sawaya, M. E., Shalita, A. R.: Androgen receptor polymorphisms (CAG repeat lengths) in androgenetic alopecia, hirsutism, and acne. Journal of Cutaneous Medicine and Surgery, 3: 9–15, 1998. Thompson, I. M., Goodman, P. J., Tangen, C. M., Lucia, M. S., Miller, G. J., Ford, L. G., Lieber, M. M., Cespedes, R. D., Atkins, J. N., Lippman, S. M., Carlin, S. M., Ryan, A., Szczepanek, C. M., Crowley, J. J., Coltman, C. A. Jr.: The Influence of Finasteride on the Development of Prostate Cancer, New England Journal of Medicine, 349: 215–224, 2003. Wilborn, T. W., Lang, N. P., Smith, M., Meleth, S., Falany, C. N.: Association of SULT2A1 allelic variants with plasma adrenal androgens and prostate cancer in African American men. The Journal of Steroid Biochemistry and Molecular Biology, 99: 209–214, 2006. Yeh, S., Chang, C.: Cloning and characterization of a specific coactivator, ARA70, for the androgen receptor in human prostate cells. Proceedings of the National Academy of Sciences of the United States of America, 93: 5517–5521, 1996. Zhao, X. Y. and Feldman, D.: The role of vitamin D in prostate cancer. Steroids, 66: 293–300, 2001.

Effect of Steroid 5a-Reductase Inhibitors on Markers of Tumor Regression and Proliferation in Prostate Cancer Lynn N. Thomas and Roger S. Rittmaster

Abstract Intracrine conversion of testosterone to the more potent dihydrotestosterone (DHT) by the enzymes 5a-reductase type 1 (5aR1) and type 2 (5aR2) is important for normal and pathological growth of the prostate. Consequently, inhibitors of 5aR have been developed to treat prostate disease. Finasteride, the first commercially available 5aR inhibitor, targets 5aR2 since this isoenzyme predominates in benign prostate tissue. Finasteride has been shown to cause a 20–30% reduction in prostate volume in men with BPH, through a combination of atrophy and apoptosis. It has further been shown to reduce the incidence of detectable prostate cancer by 25% in the Prostate Cancer Prevention Trial (PCPT), through prevention and/or treatment of subclinical carcinomas. However, development of prostate cancer is accompanied by a decrease in 5aR2 levels and an increase in 5aR1. The dual 5aR inhibitor dutasteride provides greater suppression of DHT formation than finasteride (93–97% vs. 68–86%) and may be even more effective in prevention of prostate cancer. The 4-year Reduction by Dutasteride of Prostate Cancer Events (REDUCE) trial is currently underway to test this hypothesis. Defining the role of 5aR inhibitors in treatment of established prostate cancer requires further investigation. However, the effect of such inhibitors on markers of tumor regression (apoptosis, proliferation, angiogenesis, and tumor histopathology) has been explored, in part. In several studies of men with BPH, who were treated with finasteride for 3 months to 4 years and who subsequently developed prostate cancer, the effect of finasteride on the histopathology of malignant cells was unclear. Atrophic changes such as pyknosis and vacuolization were reported in some studies, and no significant effect of finasteride was observed in others. Finasteride was not shown to have an effect on prostate cancer cell proliferation. Its effect on apoptosis and angiogenesis in prostate cancer has not been studied. In contrast, the dual 5aR inhibitor dutasteride has been shown to cause significant regressive changes in established prostate cancer, although some of the effect is time L.N. Thomas(*) Department of Biochemistry and Molecular Biology, Dalhousie University, 9D Sir Charles Tupper Building, Halifx, NS, Canada B3H 1X5, E-mail: [email protected]

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4_7, # Springer Science + Business Media, LLC 2009

155

156

L.N. Thomas, R.S. Rittmaster

dependent. Short-term treatment (6–10 weeks) caused a 40% reduction in tumor volume and resulted in significant increases in apoptosis, number of atrophic glands, and stromal:epithelial ratio. It was also associated with a nonsignificant decrease in microvessel density. After longer treatment (4 months), the regressive changes induced by dutasteride waned, although reductions in tumor volume continued to be evident. In summary, the results from the PCPT demonstrated that inhibition of 5aR may prevent or delay prostate cancer in some patients. From preprostatectomy trials of dutasteride, it is also clear that suppression of DHT formation with a 5aR inhibitor may cause regression of prostatic carcinomas. The fact that regressive changes were not observed to the same extent with finasteride may be for two reasons. First, since finasteride is a 5aR2 selective inhibitor, DHT formation can continue due to the action of 5aR1. Second, the timing of the observations in many of the studies with finasteride may not have been optimal.

1 Introduction In the prostate, normal growth, development, and secretory activity are androgen dependent. Intracrine conversion of circulating testosterone to the more potent androgen dihydrotestosterone (DHT) by the enzyme 5a-reductase (5aR) is a key step in both normal prostate function and development of prostate cancer (George et al. 1989; Imperato-McGinley et al. 1992; Petrow 1986). DHT has approximately a twofold higher intrinsic potency for stimulating proliferation and secretory function in the prostate (Wright et al. 1999), and it also binds the androgen receptor with approximately a tenfold greater affinity, providing a more stable receptorligand complex (Deslypere et al. 1992). Deficiency of 5aR results in male pseudohermaphroditism (Andersson et al. 1991). There are two well-characterized 5aR isoenzymes, designated type 1 (5aR1) and type 2 (5aR2). They are encoded by two distinct genes, SRD5A1 and SRD5A2, which are located on separate chromosomes, and they have differing kinetic properties (Jenkins et al. 1991, 1992; Russell and Wilson 1994; Span et al. 1996). In the normal prostate, 5aR2 is the predominant isoenzyme. It is present at high levels and has a higher binding affinity for testosterone than 5aR1 (Km value 11.8 vs. 1,995 nM) (Span et al. 1996). The 5aR1 isoenzyme was first detected in nonreproductive tissues such as the liver and skin (Normington and Russell 1992; Thigpen et al. 1993). However, studies now have shown that 5aR1 is present in the prostate, and its levels are elevated in prostate cancer (Bonkhoff et al. 1996; Iehle et al. 1999; Span et al. 1999; Thomas et al. 2003). Recent reports have identified a third 5aR isoenzyme, but the clinical significance of this isoenzyme has yet to be elucidated (Titus et al. 2007; Uemura et al. 2007). This chapter will focus on the relationship between 5aR and prostate cancer, with particular emphasis on the effect of inhibition of 5aR on markers of regression and proliferation in prostate cancer.

Effect of Steroid 5a-Reductase Inhibitors on Markers of Tumor Regression

157

2 5a-Reductase Inhibitors and BPH Because DHT formation is crucial for normal and abnormal prostate growth, 5aR inhibitors were developed for treatment of prostate disease. Finasteride was the first commercially available 5aR inhibitor. It targeted 5aR2, since it is the predominant isoenzyme in the benign prostate. It has been widely used for treatment of benign prostatic hyperplasia (BPH), because it suppresses intraprostatic DHT formation by 68–86% (Span et al. 1999; McConnell et al. 1992). This results in decreased prostate volume, improved symptoms and urinary flow rates, and a reduction in the risk of acute urinary retention or BPH-related surgery (McConnell et al. 1998). Prostate volume is decreased by 20–30% through a combination of increased atrophy and apoptosis (Marks et al. 1999; Montironi et al. 1996a; Rittmaster et al. 1996). Finasteride has been shown to cause a 55% decrease in benign epithelial cell content after 6 months of treatment (Marks et al. 1997). Despite its proven benefit, finasteride provides only partial suppression of serum and intraprostatic DHT (Span et al. 1999; Cote et al. 1998; McConnell et al. 1992). Residual DHT formation is likely due to the action of 5aR1. The dual 5aR inhibitor dutasteride provides greater and more consistent suppression of serum DHT than does finasteride (95 and 70%, respectively) (Clark et al. 2004). Like finasteride, it has been shown to be effective in relieving symptoms of BPH (Roehrborn et al. 2002). Studies of the efficacy of 5aR inhibitors in the treatment of prostate cancer are ongoing. The Prostate Cancer Prevention Trial (PCPT) has shown that finasteride can be a useful therapy, which may prevent development of prostate cancer and/or prevent progression of subclinical malignancies to clinically significant prostate cancer (Thompson et al. 2003). Further studies have examined the effect of 5aR inhibition on regression of existing prostate cancer.

3 Androgen-Deprivation Therapy for Treatment of Prostate Cancer Currently, androgen-deprivation therapy (ADT) is widely used as a treatment for prostate cancer. It is accomplished through the use of GnRH agonists/antagonists, alone or in combination with antiandrogens, which bind the androgen receptor. ADT results in characteristic changes in the histopathology of benign, PIN, and prostate cancer tissues. These changes include a reduction in tumor volume, as well as shrunken nucleoli, pyknosis, vacuolated cytoplasm, and smaller tumor glands (Bostwick et al. 2004; Civantos et al. 1996). In addition to these changes, ADT also causes increased apoptosis (Rittmaster et al. 1999), decreased proliferation (Armas et al. 1994), and decreased microvessel density (Marshall and Narayan 1993; Mazzucchelli et al. 2000). Figure 1 shows increased staining for two markers of

158

L.N. Thomas, R.S. Rittmaster

Fig. 1 Apoptosis in prostate adenocarcinoma after ADT. Panels a and b show representative TUNEL staining in untreated prostate cancer (a) and prostate cancer treated by ADT (b). Panel c shows representative tTG staining in untreated prostate cancer, while panel d shows prostate cancer treated by ADT. (Reproduced from Rittmaster et al. 1999, with permission from Elsevier)

apoptosis, TUNEL and tissue transglutaminase (tTG), in prostate cancer tissues after 3 months of neoadjuvant ADT. Although ADT initially is very effective in causing regression of prostate cancer, there are problems associated with this treatment. The first is the high risk of recurrence, since most patients eventually become refractory to androgen withdrawal. The second problem is side effects, which include impotence, hot flashes, depression, osteoporosis, weakness, and fatigue. For these reasons, it is important to discover new treatment options.

4 5a-Reductase Inhibition and Prostate Cancer Studies of nude mice bearing prostate cancer cell xenografts have demonstrated that DHT has a stronger stimulatory effect on prostate tumor growth than testosterone (Xu et al. 2006). A number of studies have shown that inhibition of 5aR can decrease proliferation of prostate cancer cells in vitro (Bologna et al. 1995) or reduce the growth of prostate cancer xenografts in nude mice (Lamb et al. 1992).

Effect of Steroid 5a-Reductase Inhibitors on Markers of Tumor Regression

159

Festuccia et al. (2005) have shown that both the 5aR2 inhibitor finasteride and the 5aR1 inhibitor MK386 decrease proliferation in primary cultures of BPH, PIN, and prostate cancer cells. More recently, these authors have shown that the dual 5aR inhibitor dutasteride decreases proliferation and increases apoptosis to a greater extent and in a higher proportion of prostate cancer cells in primary cultures than either finasteride or MK386. Dutasteride reduced growth in 78% of cases compared to 39% of cases for both finasteride and MK386 (Festuccia et al. 2008). Inhibitors of 5aR also have been shown to reduce the growth of the Dunning R3327 tumor in rats (Lamb et al. 1992; Zaccheo et al. 1998) and to inhibit growth of PC-82 xenografts in mice (Lamb et al. 1992). Finally, finasteride has been shown to prevent progression of prostate cancer from microscopic to macroscopic tumors in rats (Tsukamoto et al. 1998). It has been reported that 5aR activity levels may be correlated with risk of prostate cancer. Asian men, who have a lower risk of developing prostatic malignancies, have decreased levels of androstanediol glucuronide compared to Caucasians, suggesting lower 5aR activity (Lookingbill et al. 1991; Ross et al. 1992). In addition, genetic variants of 5aR with greater intrinsic activity are associated with prostate cancer progression (Makridakis et al. 2004). However, in studies examining the effect of 5aR inhibitors on prostate cancer prevention or treatment, results have been mixed. In one of the earliest studies of 120 men with detectable PSA levels (0.6–10 ng/ml) but no evidence of bone metastases following radical prostatectomy for localized prostate cancer, finasteride treatment did not significantly affect tumor recurrence, as defined by a positive bone scan or biopsy. However, treatment was associated with a delay of 14 months in the time to an increase in serum PSA (Andriole et al. 1995). Similarly, in a study by Cote et al. (1998) of 52 men with elevated PSA (>4) and a negative biopsy for prostate cancer, who received finasteride (n = 27) or no medication (n = 25) for 12 months prior to being rebiopsied, finasteride was found to have no effect on cancer incidence. However, this was a small study that was not appropriately powered to assess this endpoint. In contrast, the Prostate Cancer Prevention Trial (PCPT) was powered to assess prostate cancer incidence and, in this study of 18,882 men with serum PSA values 3 ng/ml, treatment with finasteride was associated with a 24.8% reduction in prostate cancer detection rates. This result has provided significant evidence of a role for 5aR inhibition in the prevention or treatment of prostate cancer (Thompson et al. 2003).

5 Markers of Tumor Regression 5.1

TUNEL

Apoptosis is an important component of prostate regression following ADT. Castration-induced apoptosis in benign human prostate xenografts has been

160

L.N. Thomas, R.S. Rittmaster

shown to account for the loss of nearly 90% of epithelial cells, even though the apoptotic index was relatively low (Staack et al. 2003). In men with prostate cancer, neoadjuvant ADT increased apoptosis in prostate cancer tissue (Rittmaster et al. 1999; Matsushima et al. 1999). However, sensitivity and specificity of apoptotic measurements can be problematic, since marker expression is dependent upon the stage of apoptosis. TUNEL staining, a widely accepted method for detecting apoptosis, identifies fragmentation of DNA, which is a hallmark of apoptosis (Gavrieli et al. 1992). However, TUNEL staining is evanescent, and many apoptotic cells are not labeled at any given time.

5.2

Tissue Transglutaminase

Tissue transglutaminase is an alternate marker for apoptosis (Piacentini et al. 1991), which persists longer than TUNEL, thereby providing more information regarding cumulative levels of apoptosis. Although it has not been as widely used as TUNEL, there is sufficient evidence to conclude that it is a useful marker of apoptosis. In a study in which dexamethasone, which induces apoptosis in the thymus, was administered to tTG knock-out mice and wild-type mice, cell viability was decreased to a greater extent in the knock-out mice, even though there was no difference in the rate of apoptosis. This suggests that there was increased clearing of apoptotic cells in the knock-out mice, which in turn suggests that tTG may be involved in stabilizing apoptotic cells before clearance (Nanda et al. 2001). Tissue transglutaminase may have more than one function in the cascade of events during apoptosis. Cells that overexpress tTG have been shown to have ultrastructural changes associated with hyperpolarization of the mitochondria and increased production of reactive oxygen species during staurosporine-induced apoptosis. Tissue transglutaminase may have a role in sensitizing cells for apoptosis through modification of mitochondrial homeostasis (Piacentini et al. 2002). However, tTG is not necessarily specific for apoptosis, since it is involved in other cellular processes, such as signal transduction, cell adhesion, and wound healing. Due to the limitations of these methods for evaluating apoptosis, it is prudent to assess both.

5.3

Microvessel Density and Vascular Endothelial Growth Factor

Angiogenesis is a potent prognostic indicator for prostate cancer patients, since it is a crucial factor for the growth of solid tumors beyond 2 cm2 (Margolin 2002). Since studies have shown that DHT may be necessary for induction of angiogenesis (Pan et al. 2003), microvessel density (MVD) or vascular endothelial growth factor (VEGF) may be useful markers of tumor regression. MVD is progressively

Effect of Steroid 5a-Reductase Inhibitors on Markers of Tumor Regression

161

increased in BPH, PIN, and prostate cancer, compared to normal prostate tissue (Bigler et al. 1993; Brawer et al. 1994; Montironi et al. 1996b), and it is increased in high-grade prostate cancer compared to low-grade prostate cancer (Bettencourt et al. 1998; Mydlo et al. 1998). Staining for vascular endothelial growth factor (VEGF) is decreased in benign, PIN, and prostate cancer tissues after 3 months of androgen ablation (Mazzucchelli et al. 2000).

5.4

Histopathology

Since atrophy is also a component of regression (Rittmaster et al. 1995, 1996), histopathological evaluation of tissues for characteristic changes of atrophy is valuable. Such an evaluation should include morphometrics measurements of epithelial cell height, duct width, and nuclear area, as well as assessment of mean tumor volume, stroma:gland ratio, and Gleason score.

6 Effect of 5a-Reductase Inhibitors on Prostate Cancer Histology 6.1

Finasteride

The earliest observations of the effect of 5aR on prostate cancer histopathology produced conflicting results. The studies were conducted in men who were treated with finasteride for BPH and subsequently developed cancer. In a study of cancer containing biopsy specimens from the Proscar Long-Term Efficacy and Safety Study (PLESS), which assessed the efficacy of finasteride in the treatment of benign prostatic hyperplasia (BPH), finasteride (4 years) was not shown to have an effect on the histopathological features of prostate cancer tissue compared to placebo (Yang et al. 1999). Likewise, in the study by Cote et al. (1998) described earlier, the authors reported no change in PIN and no change in the proliferation index, as assessed by proliferating cell nuclear antigen (PCNA) staining after treatment with finasteride for 1 year. Finally, Carver et al. evaluated prostate cancer tissue from men who received finasteride for at least 6 months in order to treat lower urinary tract problems and subsequently developed prostate cancer, which was treated by radical prostatectomy. Their aim was to determine whether finasteride caused sufficient change in tumor histopathology to prevent accurate Gleason-grade assignments, and they concluded that it did not (Carver et al. 2005). In contrast, Civantos et al. examined radical prostatectomy specimens from five men who had received finasteride for 3–24 months prior to being diagnosed with prostate cancer. They reported that finasteride induced similar, but less pronounced histopathological changes to androgen ablation. These changes included atrophy, vacuolization,

162

L.N. Thomas, R.S. Rittmaster

pyknosis, and smaller tumor glands. However, these changes were less common and milder with finasteride compared to androgen ablation (Civantos et al. 1997). Finasteride also has been used in preprostatectomy studies of men with prostate cancer, designed to assess the impact of 5aR inhibition on markers of regression in prostatic adenocarcinoma. In one such study, the effect of at least 6 months of finasteride treatment on prostate cancer morphology and clinical outcome was compared to 3 months’ androgen ablation with an LHRH agonist. Fifty-six patients received finasteride, 56 patients received surgery only, and 44 patients received an LHRH agonist. After radical prostatectomy, the prostate tissues were scored in a blinded manner on a scale of 0–2, in which 0 was assigned when there was no apparent effect, 1 was assigned when the tissue looked suspicious for hormonal effect, and 2 was assigned when the histopathological features were highly suggestive of hormonal treatment. The morphologic changes in the finasteride and surgery-only groups were comparable (mean scores of 0.5 and 0.4, respectively), whereas the mean score for the LHRHa group was fourfold higher (1.6). Likewise, clinical outcomes were similar for finasteride and control groups. (Figure 2 shows the distribution of scores in control, finasteride, and LHRHa groups.) (Rubin et al. 2005).

6.2

Dutasteride

Because there are two well-characterized 5aR isoenzymes, and finasteride at clinical doses inhibits only 5aR2, it is possible that dual inhibition of 5aR may

Fig. 2 Distribution of treatment score in biopsy and prostatectomy specimens from control, finasteride, and LHRHa treatment groups. No hormonal effect was evident in the biopsy specimens. However, in prostatectomy specimens, approximately 80% of LHRH patients showed suspicious or strong treatment effects, compared to 20% of control patients and 26% of finasteride patients. (Reproduced from Rubin et al. 2005, with permission from Elsevier)

Effect of Steroid 5a-Reductase Inhibitors on Markers of Tumor Regression

163

have greater efficacy in treatment of early stage prostate cancer. There is increasing evidence that 5aR1 expression is greater in prostate cancer compared with benign tissue at both the messenger RNA (mRNA) level (Iehle et al. 1999; Thomas et al. 2005) and at the protein level (Bonkhoff et al. 1996; Thomas et al. 2003), whereas 5aR2 expression and activity are decreased (Thomas et al. 2005; Bjelfman et al. 1997; Luo et al. 2003; Soderstrom et al. 2001; Titus et al. 2005). The potential role of 5aR1 in prostate pathology is supported clinically by evidence that dual inhibition with dutasteride provides greater and more complete suppression of serum DHT levels than finasteride in men with BPH (Clark et al. 2004). Furthermore, dutasteride inhibits growth of prostate cancer cells in vitro (Lazier et al. 2004) and growth of the Dunning R-3327H tumor in rats (Xu et al. 2006) to a greater extent than finasteride. Preprostatectomy trials to assess the effect of dutasteride on markers of regression in prostate cancer were undertaken. Two studies were conducted, both of which will be discussed in depth. In the first, dutasteride was administered for 6–10 weeks prior to prostatectomy, and in the second, dutasteride was administered for 4 months prior to prostatectomy. In both studies, the decision to undergo a prostatectomy was made independent of participation in the study.

6.3

ARIA2003 (Preradical Prostatectomy Study with Dutasteride)

ARIA2003 was a study designed to determine whether dutasteride causes regressive changes in prostate cancer in vivo. In a double-blind, multicenter trial, 46 men with clinically staged T1 or T2 prostate cancer were randomized to receive 5 mg/ day of dutasteride or placebo for 6–10 weeks prior to radical prostatectomy. Resected tissues were analyzed to determine the effect of dutasteride on intraprostatic androgen levels, and indices of apoptosis and MVD in malignant tissue, as well as degree of atrophy in benign tissue (Andriole et al. 2004a). Surgical margins, extraprostatic extension, stroma:gland ratio, percent atrophic epithelium, and PIN and tumor volumes were also evaluated (Iczkowski et al. 2005). Serum DHT decreased by 96.5% from 296 to 10.3 pg/ml after dutasteride treatment. Treatment with dutasteride also caused a 97.0% reduction in intraprostatic DHT – from 6,178.8 pg/g in the dutasteride group to 177.2 pg/g in the placebo group. There was no significant change in serum testosterone levels, but intraprostatic testosterone levels increased from 124.5 pg/g in the placebo group to 2,502 pg/g in the dutasteride group. The combined T and DHT concentrations in the prostate were 60% lower in the dutasteride group compared to placebo. TUNEL staining (number of positive cells/mm2 of tumor) showed a nonsignificant increase of 64% from 32.5 7.9 in the placebo group to 53.3 11.9 in the dutasteride group. Staining for tTG, which was measured using image analysis software, was doubled in the dutasteride group compared to placebo. When the men were grouped according to duration of treatment (<45 days or 45 days), tTG staining was quadrupled in the longer-term treatment group (P = 0.036).

164

L.N. Thomas, R.S. Rittmaster

Fig. 3 Apoptosis in prostate adenocarcinoma after treatment with dutasteride. Panels a and b show representative TUNEL staining in prostate cancer tissues from patients treated with placebo and dutasteride, respectively. Panels c and d show representative tTG staining in prostate cancer tissues from patients treated with placebo and dutasteride, respectively. (Reproduced from Andriole et al. 2004a, with permission from Elsevier)

Figure 3 shows representative staining for TUNEL and tTG in placebo and dutasteride-treated subjects. MVD was assessed by staining for the vascular endothelial antigen CD34. While there was no statistical difference overall between the two groups, MVD was decreased by 45% from 6.6 0.9 in the placebo group to 3.6 0.8 in men treated with dutasteride for 45 days. In addition there was a trend toward reduced MVD over time (Fig. 4). Histopathological assessment of cancerous tissue showed that tumor volume was decreased by 40%, PIN volume was decreased by 50%, stroma:gland ratio was doubled, and the percent of atrophic epithelium was doubled. The treatment alteration score, which was determined by a pathologist based on cytological and architectural changes in the cancer tissue, was also doubled in the dutasteride group compared to placebo (Table 1). The number of patients with positive surgical margins was not statistically different between the two groups, but the mean extent of margin involvement

Effect of Steroid 5a-Reductase Inhibitors on Markers of Tumor Regression

165

CD34 (% of tumor area staining positively)

30

25

20

15

10

5

0 30

35

40

45

50

55

60

65

70

75

80

Duration of Treatment Fig. 4 Trend toward reduced MVD over time in prostate cancer after finasteride treatment. This graph shows the negative correlation between duration of dutasteride treatment (up to 10 weeks) and MVD. (Reproduced from Andriole et al. 2004a, with permission from Elsevier) Table 1 Histopathological features of prostate cancer after treatment with dutasteride Placebo Dutasteride No. of subjects 19 18 Mean margin involvement 11.7 4.7 4.2 0.9 Mean extraprostatic extension 15.4 7.9 4.3 1.8 Mean PIN volume 0.67 0.16 0.31 0.08 (P = 0.052) Tumor volume:prostate weight ratio 0.09 0.02 0.06 0.02 Gleason score 6.9 0.75 6.9 0.9 % of specimen involved by cancer 24.2 12.4 15.2 12.6 (P = 0.025) Treatment alteration score 1.6 2.2 3.2 2.5 (P = 0.055) % of atrophic epithelium 26.7 31.3 52.9 38.9 (P = 0.041) Stroma:gland ratio 0.26 0.15 0.52 0.56 (P = 0.046)

was decreased from 11.7 4.7 in the placebo group to 4.2 0.9 in the dutasteride group. Similarly, the mean extent of extraprostatic extension was decreased from 15.4 7.9 to 4.3 1.8 in the placebo and dutasteride groups, respectively (Table 1).

166

L.N. Thomas, R.S. Rittmaster

Benign epithelial cell width decreased by 18% from 43.0 1.4 in the placebo group to 35.4 1.0 in the dutasteride group (P < 0.0001). The decrease was evident in both the transition zone (where BPH usually develops) and the peripheral zone (where the majority of prostate cancers arise). In this study, dutasteride significantly increased apoptosis in men treated for 45 days or more and resulted in a trend toward reduced microvessel density. It also caused significant histopathological changes in both benign and cancer tissues, which included increased atrophy, increased stroma:gland ratio, and decreased tumor volume. Clinically, there was a nonsignificant decrease in surgical margin involvement and mean extent of extraprostatic extension, although no reduction in incidence, in men taking dutasteride. These findings, in combination with the almost complete suppression of intraprostatic DHT by dutasteride, support a role for this compound in prevention or early stage treatment of prostate cancer.

6.4

ARI40010 (4-Month Preradical Prostatectomy Study with Dutasteride)

The follow-up study to ARIA2003 was ARI40010. In this study, 81 men with clinically localized prostate cancer were randomized to receive no treatment or dutasteride at a dose of 0.5 or 3.5 mg/day for 4 months prior to radical prostatectomy. Resected tissues were analyzed for intraprostatic androgen concentrations; indices of apoptosis and proliferation; MVD and histopathological features such as prostate volume, tumor volume, atrophy, Gleason grade, stroma:epithelial ratio, and nuclear/architectural changes (treatment alteration score). Serum PSA measurements were also made (Gleave et al. 2006). The results demonstrated that intraprostatic DHT was significantly reduced by 93.1% in the 0.5-mg dutasteride group and 98.8% in the 3.5-mg dutasteride group compared to the surgery-only group. Intraprostatic testosterone levels were significantly higher in both dutasteride groups compared to the surgery-alone group. In benign tissue, there was no significant effect of dutasteride on apoptosis, MVD, or stroma:epithelial ratio. The proportion of atrophic epithelium was higher in dutasteride-treated subjects, in both the transition and the peripheral zones. In malignant tissues, median tumor volume was decreased by 41 and 46% in the 0.5- and 3.5-mg dutasteride groups, respectively. Unexpectedly, TUNEL staining was decreased in the dutasteride groups, but tTG staining was nonsignificantly increased by a factor of 3.6 in the dutasteride-treated subjects as a whole. The trend was similar in both dutasteride groups. The results for MVD and proliferation measurements were opposite to the expected results. MVD was greater in the dutasteride groups compared to the surgery-alone group. The increase reached statistical significance in the 0.5-mg dutasteride group (P = 0.031). Similarly, Ki67 staining for cell proliferation was increased in both dutasteride groups, and the increase was significant in the 0.5-mg group compared to the surgery-alone group (P = 0.038).

Effect of Steroid 5a-Reductase Inhibitors on Markers of Tumor Regression

167

Table 2 Time-dependent changes in markers of tumor regression after treatment with dutasteride Fold change <7 weeks 7 weeks 16 weeks TUNEL 1.64 1.65 0.51 tTG 1.92 3.75 3.63 MVD 1.0 0.55 1.22

Treatment alteration scores were nonsignificantly greater in tissues from treated men. Mean serum PSA concentrations were 47.1 19.2% lower in men receiving 0.5 mg/day dutasteride and 58.0 17.9% lower in men receiving 3.5 mg/day dutasteride, compared to untreated men. Dutasteride treatment resulted in almost total suppression of intraprostatic DHT formation and a decrease in tumor volume. Serum PSA values were reduced to a similar extent as in men with BPH treated with 0.5 mg dutasteride for 6 months or longer (Andriole et al. 2006), suggesting that dutasteride suppressed PSA from both benign and cancer tissue. Taken together, these data indicate that 4 months of treatment with dutasteride was effective in causing regression of prostate cancer. Cumulative levels of apoptosis, as assessed by tTG staining, were nonsignificantly increased, and there was a trend toward decreased levels of vascular endothelial growth factor (VEGF), an important growth factor for promoting angiogenesis. Proliferation and MVD were increased after dutasteride treatment and the rate of apoptosis, as assessed by TUNEL, was decreased. Histopathological features such as proportion of atrophy and treatment alteration score were not significantly altered by dutasteride treatment. The apparent disparity in the results of ARIA2003 and ARI40010 may be because the remodeling induced by treatment with dutasteride occurs early after onset of treatment and levels off thereafter. By 4 months, the remodeling process may be near complete, and the cells that are left are demonstrating a reaction to androgen deprivation. This theory is supported by studies of ADT showing that apoptosis in prostate cancer tissues, as visualized by TUNEL, is increased in the first few days after initiation of treatment, and then returns to baseline (Staack et al. 2003; Ohlson et al. 2005). This phenomenon has also been shown to occur in benign prostate tissue after treatment with finasteride. In men treated for 6–18 days, mean TUNEL staining (counts/mm2) was increased from 0.4 0.2 in the placebo group to 2.8 0.9 (P = 0.0026). After treatment for 23–73 days, TUNEL staining was 1.7 0.5 (P = 0.10) and returned to near baseline levels (0.7 0.5) after 3 months of treatment. Staining for tTG, which has a longer duration of staining than TUNEL, peaked in men treated for 23–73 days and returned to baseline thereafter (Rittmaster et al. 1996). A similar time-dependent effect has been documented for Ki67 staining. A decrease in proliferation for a few days is followed by an increase to levels greater than baseline (Ohlson et al. 2005). The same may be true for MVD, since a trend toward a decrease in MVD was observed in ARIA2003. This is supported by the observation of a trend toward a decrease in VEGF in the dutasteride group in ARI40010. Alternately, treatment with a 5aR inhibitor may have a neutral or slight

168

L.N. Thomas, R.S. Rittmaster

effect on MVD, which is offset by the reductions in prostate and tumor volumes caused by inhibition of 5aR. The changes in markers of tumor regression over time are outlined in Table 2. It is clear from these studies that timing is crucial in any assessment of the effect of traditional ADT or 5aR inhibition on regression of prostate cancer. The likely explanation for the time dependence of regressive changes in growth characteristics is that androgen-sensitive cells respond early to treatment through increases in apoptosis and decreases in proliferation and angiogenesis. The treatment effect subsequently may wane when cells resistant to androgen deprivation emerge, as an adaptive or selective mechanism in response to low androgen levels.

7 Effect of 5a-Reductase Inhibitors on Gleason Scores Gleason scoring has been used for decades as a prognostic indicator in prostate cancer. In the PCPT, although finasteride reduced the overall incidence of prostate cancer, there were more Gleason 7–10 (high grade) tumors in the finasteride arm. Current evidence suggests that this was due partly to an ascertainment bias and partly to reduced efficacy of finasteride in shrinking high-grade tumors. However, some investigators have hypothesized that either finasteride-induced androgen deprivation made tumors look higher grade or that finasteride actually induced development of high-grade tumors. In the 4-month preradical prostatectomy study (ARI40010), the rate of Gleason upgrading between the initial biopsy and the radical prostatectomy was lower in the two dutasteride arms than in the control arm (Fig. 5). This result suggests that dutasteride does not lead to Gleason upgrading, either through creating a pathological artifact or actual induction of more aggressive tumors.

8 Conclusion Taken cumulatively, the data from these studies provide evidence that apoptosis and atrophy are important processes during androgen-deprivation induced regression of prostate cancer. Changes in proliferation and MVD may also be factors, but likely have a smaller impact on regression. The consensus of evidence indicates that finasteride treatment causes changes in prostate cancer histopathology that are similar to, but less pronounced than ADT. That such changes were not observed in all studies is likely because the timing of the observations was not optimal in some of the studies. It is also likely that changes could not be detected in the studies of biopsy specimens due to the incomplete and heterogeneous pattern of the changes induced by finasteride. Finasteride has been shown to have little impact

Effect of Steroid 5a-Reductase Inhibitors on Markers of Tumor Regression

169

Fig. 5 Rate of Gleason upgrading between initial biopsy and radical prostatectomy. This graph shows the percentage of prostate cancers with a decreased, same, or increased Gleason score between biopsy and radical prostatectomy. (Permission obtained)

on cancer cell proliferation in vivo, and its effect on apoptosis has not been studied in prostate cancer. Although finasteride has not been proven to have a significant impact on regression of clinically apparent prostate cancer, the PCPT has demonstrated that finasteride can reduce prostate cancer detection rates through prevention and/or through treatment of early stage, subclinical carcinomas. Since detection rates diverged early in the study, it seems likely that the latter possibility can account for at least part of the reduction in prostate cancer prevalence. Dutasteride has been shown to have an effect on markers of regression in prostate cancer, although some of the effect is time dependent. Treatment for 6–10 weeks is associated with an increase in apoptosis, a nonsignificant decrease in MVD, an increase in the proportion of atrophic tumor glands, an increase in stroma:gland ratio, and reductions in prostate and tumor volumes. Despite these results, the role of dutasteride in the treatment of prostate cancer has not yet been defined, as some of the effects were diminished after a longer duration of treatment. With regard to prevention or treatment of subclinical prostate cancer, it is likely that dutasteride may be even more effective than finasteride. The results from the two preprostatectomy trials, in combination with increased expression of 5aR1 in prostate cancer and the almost complete suppression of intraprostatic DHT formation by dutasteride, provide support for this hypothesis. The 4-year Reduction by Dutasteride of Prostate Cancer Events (REDUCE) trial is currently underway to test this hypothesis (Andriole et al. 2004b).

170

L.N. Thomas, R.S. Rittmaster

References Andersson S, Berman DM, Jenkins EP, Russell DW. Deletion of steroid 5 alpha-reductase 2 gene in male pseudohermaphroditism. Nature. 1991 354(6349):159–61. Andriole G, Lieber M, Smith J, Soloway M, Schroeder F, Kadmon D, et al. Treatment with finasteride following radical prostatectomy for prostate cancer. Urology. 1995 45(3):491–7. Andriole GL, Humphrey P, Ray P, Gleave ME, Trachtenberg J, Thomas LN, et al. Effect of the dual 5alpha-reductase inhibitor dutasteride on markers of tumor regression in prostate cancer. J Urol. 2004a 172(3):915–9. Andriole G, Bostwick D, Brawley O, Gomella L, Marberger M, Tindall D, et al. Chemoprevention of prostate cancer in men at high risk: rationale and design of the reduction by dutasteride of prostate cancer events (REDUCE) trial. J Urol. 2004b 172(4 Pt 1):1314–7. Andriole GL, Marberger M, Roehrborn CG. Clinical usefulness of serum prostate specific antigen for the detection of prostate cancer is preserved in men receiving the dual 5alpha-reductase inhibitor dutasteride. J Urol. 2006 175(5):1657–62. Armas OA, Aprikian AG, Melamed J, Cordon-Cardo C, Cohen DW, Erlandson R, et al. Clinical and pathobiological effects of neoadjuvant total androgen ablation therapy on clinically localized prostatic adenocarcinoma. Am J Surg Pathol. 1994 18(10):979–91. Bettencourt MC, Bauer JJ, Sesterhenn IA, Connelly RR, Moul JW. CD34 immunohistochemical assessment of angiogenesis as a prognostic marker for prostate cancer recurrence after radical prostatectomy. J Urol. 1998 160(2):459–65. Bigler SA, Deering RE, Brawer MK. Comparison of microscopic vascularity in benign and malignant prostate tissue. Hum Pathol. 1993 24(2):220–6. Bjelfman C, Soderstrom TG, Brekkan E, Norlen BJ, Egevad L, Unge T, et al. Differential gene expression of steroid 5a-reductase 2 in core needle biopsies from malignant and benign prostatic tissue. J Clin Endocrinol Metab. 1997 82(7):2210–4. Bologna M, Muzi P, Biordi L, Festuccia C, Vicentini C. Finasteride dose-dependently reduces the proliferation rate of the LnCap human prostatic cancer cell line in vitro. Urology. 1995 45 (2):282–90. Bonkhoff H, Stein U, Aumuller G, Remberger K. Differential expression of 5a-reductase isoenzymes in the human prostate and prostatic carcinomas. Prostate. 1996 29(4):261–7. Bostwick DG, Qian J, Civantos F, Roehrborn CG, Montironi R. Does finasteride alter the pathology of the prostate and cancer grading? Clin Prostate Cancer. 2004 2(4):228–35. Brawer MK, Deering RE, Brown M, Preston SD, Bigler SA. Predictors of pathologic stage in prostatic carcinoma. The role of neovascularity. Cancer. 1994 73(3):678–87. Carver BS, Kattan MW, Scardino PT, Eastham JA. Gleason grade remains an important prognostic predictor in men diagnosed with prostate cancer while on finasteride therapy. BJU Int. 2005 95 (4):509–12. Civantos F, Soloway MS, Pinto JE. Histopathological effects of androgen deprivation in prostatic cancer. Semin Urol Oncol. 1996 14(2 Suppl 2):22–31. Civantos F, Watson RB, Pinto JE, Korman HJ, Soloway MS. Finasteride effect on prostatic hyperplasia and prostate cancer. J Urol Pathol. 1997 6:1–13. Clark RV, Hermann DJ, Cunningham GR, Wilson TH, Morrill BB, Hobbs S. Marked suppression of dihydrotestosterone in men with benign prostatic hyperplasia by dutasteride, a dual 5a-reductase inhibitor. J Clin Endocrinol Metab. 2004 89(5):2179–84. Cote RJ, Skinner EC, Salem CE, Mertes SJ, Stanczyk FZ, Henderson BE, et al. The effect of finasteride on the prostate gland in men with elevated serum prostate-specific antigen levels. Br J Cancer. 1998 78(3):413–8. Deslypere JP, Young M, Wilson JD, McPhaul MJ. Testosterone and 5a-dihydrotestosterone interact differently with the androgen receptor to enhance transcription of the MMTV-CAT reporter gene. Mol Cell Endocrinol. 1992 88(1–3):15–22.

Effect of Steroid 5a-Reductase Inhibitors on Markers of Tumor Regression

171

Festuccia C, Angelucci A, Gravina GL, Muzi P, Vicentini C, Bologna M. Effects of 5a-reductase inhibitors on androgen-dependent human prostatic carcinoma cells. J Cancer Res Clin Oncol. 2005 131(4):243–54. Festuccia C, Gravina GL, Muzi P, Pomante R, Angelucci A, Vicentini C, et al. Effects of dutasteride on prostate carcinoma primary cultures: a comparative study with finasteride and MK386. J Urol. 2008 180(1):367–372. Gavrieli Y, Sherman Y, Ben-Sasson SA. Identification of programmed cell death in situ via specific labeling of nuclear DNA fragmentation. J Cell Biol. 1992 119:493–99. George FW, Johnson L, Wilson JD. The effect of a 5a-reductase inhibitor on androgen physiology in the immature male rat. Endocrinology. 1989 125(5):2434–38. Gleave M, Qian J, Andreou C, Pommerville P, Chin J, Casey R, et al. The effects of the dual 5areductase inhibitor dutasteride on localized prostate cancer–results from a 4-month pre-radical prostatectomy study. Prostate. 2006 66(15):1674–85. Iczkowski KA, Qiu J, Qian J, Somerville MC, Rittmaster RS, Andriole GL, et al. The dual 5a-reductase inhibitor dutasteride induces atrophic changes and decreases relative cancer volume in human prostate. Urology. 2005 65(1):76–82. Iehle C, Radvanyi F, Gil Diez de Medina S, Ouafik LH, Gerard H, Chopin D, et al. Differences in steroid 5a-reductase iso-enzymes expression between normal and pathological human prostate tissue. J Steroid Biochem Mol Biol. 1999 68(5–6):189–95. Imperato-McGinley J, Gautier T, Zirinsky K, Hom T, Palomo O, Stein E, et al. Prostate visualization studies in males homozygous and heterozygous for 5a-reductase deficiency. J Clin Endocrinol Metab. 1992 75(4):1022–6. Jenkins EP, Hsieh CL, Milatovich A, Normington K, Berman DM, Francke U, et al. Characterization and chromosomal mapping of a human steroid 5a-reductase gene and pseudogene and mapping of the mouse homologue. Genomics. 1991 11(4):1102–12. Jenkins EP, Andersson S, Imperato-McGinley J, Wilson JD, Russell DW. Genetic and pharmacological evidence for more than one human steroid 5a-reductase. J Clin Invest. 1992 89(1):293– 300. Lamb JC, Levy MA, Johnson RK, Isaacs JT. Response of rat and human prostatic cancers to the novel 5a-reductase inhibitor, SK&F 105657. Prostate. 1992 21(1):15–34. Lazier CB, Thomas LN, Douglas RC, Vessey JP, Rittmaster RS. Dutasteride, the dual 5areductase inhibitor, inhibits androgen action and promotes cell death in the LNCaP prostate cancer cell line. Prostate. 2004 58(2):130–44. Lookingbill DP, Demers LM, Wang C, Leung A, Rittmaster RS, Santen RJ. Clinical and biochemical parameters of androgen action in normal healthy Caucasian versus Chinese subjects. J Clin Endocrinol Metab. 1991 72(6):1242–8. Luo J, Dunn TA, Ewing CM, Walsh PC, Isaacs WB. Decreased gene expression of steroid 5areductase 2 in human prostate cancer: implications for finasteride therapy of prostate carcinoma. Prostate. 2003 57(2):134–9. Makridakis NM, Akalu A, Reichardt JKV. Identification and characterization of somatic steroid 5a-reductase (SRD5A2) mutations in human prostate cancer tissue. Oncogene. 2004 23:7399– 405. Margolin K. Inhibition of vascular endothelial growth factor in the treatment of solid tumors. Curr Oncol Rep. 2002 4(1):20–8. Marks LS, Partin AW, Gormley GJ, Dorey FJ, Shery ED, Garris JB, et al. Prostate tissue composition and response to finasteride in men with symptomatic benign prostatic hyperplasia. J Urol. 1997 157(6):2171–8. Marks LS, Partin AW, Dorey FJ, Gormley GJ, Epstein JI, Garris JB, et al. Long-term effects of finasteride on prostate tissue composition. Urology. 1999 53(3):574–80. Marshall S, Narayan P. Treatment of prostatic bleeding: suppression of angiogenesis by androgen deprivation. J Urol. 1993 149(6):1553–4.

172

L.N. Thomas, R.S. Rittmaster

Matsushima H, Goto T, Hosaka Y, Kitamura T, Kawabe K. Correlation between proliferation, apoptosis, and angiogenesis in prostate carcinoma and their relation to androgen ablation. Cancer. 1999 85(8):1822–7. Mazzucchelli R, Montironi R, Santinelli A, Lucarini G, Pugnaloni A, Biagini G. Vascular endothelial growth factor expression and capillary architecture in high-grade PIN and prostate cancer in untreated and androgen-ablated patients. Prostate. 2000 45(1):72–9. McConnell JD, Wilson JD, George FW, Geller J, Pappas F, Stoner E. Finasteride, an inhibitor of 5a-reductase, suppresses prostatic dihydrotestosterone in men with benign prostatic hyperplasia. J Clin Endocrinol Metab. 1992 74(3):505–8. McConnell JD, Bruskewitz R, Walsh P, Andriole G, Lieber M, Holtgrewe HL, et al. The effect of finasteride on the risk of acute urinary retention and the need for surgical treatment among men with benign prostatic hyperplasia. Finasteride Long-Term Efficacy and Safety Study Group. N Engl J Med. 1998 338(9):557–63. Montironi R, Valli M, Fabris G. Treatment of benign prostatic hyperplasia with 5a-reductase inhibitor: morphological changes in patients who fail to respond. J Clin Pathol. 1996a 49 (4):324–8. Montironi R, Diamanti L, Thompson D, Bartels HG, Bartels PH. Analysis of the capillary architecture in the precursors of prostate cancer: recent findings and new concepts. Eur Urol. 1996b 30(2):191–200. Mydlo JH, Kral JG, Volpe M, Axotis C, Macchia RJ, Pertschuk LP. An analysis of microvessel density, androgen receptor, p53 and HER-2/neu expression and Gleason score in prostate cancer. Preliminary results and therapeutic implications. Eur Urol. 1998 34(5):426–32. Nanda N, Iismaa SE, Owens WA, Husain A, Mackay F, Graham RM. Targeted inactivation of Gh/ tissue transglutaminase II. J Biol Chem. 2001 276(23):20673–8. Normington K, Russell DW. Tissue distribution and kinetic characteristics of rat steroid 5areductase isozymes. Evidence for distinct physiological functions. J Biol Chem. 1992 267 (27):19548–54. Ohlson N, Wikstrom P, Stattin P, Bergh A. Cell proliferation and apoptosis in prostate tumors and adjacent non-malignant prostate tissue in patients at different time-points after castration treatment. Prostate. 2005 62(4):307–15. Pan SL, Guh JH, Huang YW, Chern JW, Chou JY, Teng CM. Identification of apoptotic and antiangiogenic activities of terazosin in human prostate cancer and endothelial cells. J Urol. 2003 169(2):724–9. Petrow V. The dihydrotestosterone (DHT) hypothesis of prostate cancer and its therapeutic implications. Prostate. 1986 9(4):343–61. Piacentini M, Autuori F, Dini L, Farrace MG, Ghibelli L, Piredda L, et al. ‘‘Tissue’’ transglutaminase is specifically expressed in neonatal rat liver cells undergoing apoptosis upon epidermal growth factor-stimulation. Cell Tissue Res. 1991 263(2):227–35. Piacentini M, Farrace MG, Piredda L, Matarrese P, Ciccosanti F, Falasca L, et al. Transglutaminase overexpression sensitizes neuronal cell lines to apoptosis by increasing mitochondrial membrane potential and cellular oxidative stress. J Neurochem. 2002 81(5):1061–72. Rittmaster RS, Manning AP, Wright AS, Thomas LN, Whitefield S, Norman RW, et al. Evidence for atrophy and apoptosis in the ventral prostate of rats given the 5a-reductase inhibitor finasteride. Endocrinology. 1995 136(2):741–8. Rittmaster RS, Norman RW, Thomas LN, Rowden G. Evidence for atrophy and apoptosis in the prostates of men given finasteride. J Clin Endocrinol Metab. 1996 81(2):814–9. Rittmaster RS, Thomas LN, Wright AS, Murray SK, Carlson K, Douglas RC, et al. The utility of tissue transglutaminase as a marker of apoptosis during treatment and progression of prostate cancer. J Urol. 1999 162(6):2165–9. Roehrborn CG, Boyle P, Nickel JC, Hoefner K, Andriole G. Efficacy and safety of a dual inhibitor of 5a-reductase types 1 and 2 (dutasteride) in men with benign prostatic hyperplasia. Urology. 2002 60(3):434–41.

Effect of Steroid 5a-Reductase Inhibitors on Markers of Tumor Regression

173

Ross RK, Bernstein L, Lobo RA, Shimizu H, Stanczyk FZ, Pike MC, et al. 5a-Reductase activity and risk of prostate cancer among Japanese and US white and black males. Lancet. 1992 339 (8798):887–9. Rubin MA, Allory Y, Molinie V, Leroy X, Faucon H, Vacherot F, et al. Effects of long-term finasteride treatment on prostate cancer morphology and clinical outcome. Urology. 2005 66 (5):930–4. Russell DW, Wilson JD. Steroid 5a-reductase: two genes/two enzymes. Annu Rev Biochem. 1994 63:25–61. Soderstrom TG, Bjelfman C, Brekkan E, Ask B, Egevad L, Norlen BJ, et al. Messenger ribonucleic acid levels of steroid 5a-reductase 2 in human prostate predict the enzyme activity. J Clin Endocrinol Metab. 2001 86(2):855–8. Span PN, Benraad Th J, Sweep CG, Smals AG. Kinetic analysis of steroid 5a-reductase activity at neutral pH in benign prostatic hyperplastic tissue: evidence for type I isozyme activity in the human prostate. J Steroid Biochem Mol Biol. 1996 57(1–2):103–8. Span PN, Voller MC, Smals AG, Sweep FG, Schalken JA, Feneley MR, et al. Selectivity of finasteride as an in vivo inhibitor of 5a-reductase isozyme enzymatic activity in the human prostate. J Urol. 1999 161(1):332–7. Staack A, Kassis AP, Olshen A, Wang Y, Wu D, Carroll PR, et al. Quantitation of apoptotic activity following castration in human prostatic tissue in vivo. Prostate. 2003 54(3):212–9. Thigpen AE, Silver RI, Guileyardo JM, Casey ML, McConnell JD, Russell DW. Tissue distribution and ontogeny of steroid 5a-reductase isozyme expression. J Clin Invest. 1993 92 (2):903–10. Thomas LN, Douglas RC, Vessey JP, Gupta R, Fontaine D, Norman RW, et al. 5a-Reductase type 1 immunostaining is enhanced in some prostate cancers compared with benign prostatic hyperplasia epithelium. J Urol. 2003 170(5):2019–25. Thomas LN, Lazier CB, Gupta R, Norman RW, Troyer DA, O’Brien SP, et al. Differential alterations in 5a-reductase type 1 and type 2 levels during development and progression of prostate cancer. Prostate. 2005 63(3):231–9. Thompson IM, Goodman PJ, Tangen CM, Lucia MS, Miller GJ, Ford LG, et al. The Influence of Finasteride on the Development of Prostate Cancer. N Engl J Med. 2003 349(3):213–22. Titus MA, Gregory CW, Ford OH, III, Schell MJ, Maygarden SJ, Mohler JL. Steroid 5a-reductase isozymes I and II in recurrent prostate cancer. Clin Cancer Res. 2005 11(12):4365–71. Titus MA, Li Y, Kawinski E, Kozyreva O, Mohler JL. Biochemical and pharmacological evidence for a third isozyme of steroid 5a-reductase in prostate cancer [abstract no. 268]. Anaheim, CA: American Urology Association. 2007. Tsukamoto S, Akaza H, Onozawa M, Shirai T, Ideyama Y. A 5a-reductase inhibitor or an antiandrogen prevents the progression of microscopic prostate carcinoma to macroscopic carcinoma in rats. Cancer. 1998 82(3):531–7. Uemura M, Tamura K, Chung S, Honma S, Okuyama A, Nakamura Y, et al. Novel 5a-steroid reductase (SRD5A3, type 3) is overexpressed in hormone-refractory prostate cancer. Cancer Sci. 2007 99(1):81–6. Wright AS, Douglas RC, Thomas LN, Lazier CB, Rittmaster RS. Androgen-induced regrowth in the castrated rat ventral prostate: role of 5a-reductase. Endocrinology. 1999 140(10):4509–15. Xu Y, Dalrymple SL, Becker RE, Denmeade SR, Isaacs JT. Pharmacologic basis for the enhanced efficacy of dutasteride against prostatic cancers. Clin Cancer Res. 2006 12(13):4072–9. Yang XJ, Lecksell K, Short K, Gottesman J, Peterson L, Bannow J, et al. Does long-term finasteride therapy affect the histologic features of benign prostatic tissue and prostate cancer on needle biopsy? PLESS Study Group. Proscar Long-Term Efficacy and Safety Study. Urology. 1999 53(4):696–700. Zaccheo T, Giudici D, di Salle E. Effect of early treatment of prostate cancer with the 5areductase inhibitor turosteride in Dunning R3327 prostatic carcinoma in rats. Prostate. 1998 35(4):237–42.

5a-Reductase Isozymes in Castration-Recurrent Prostate Cancer Mark A. Titus and James L. Mohler

Abstract Intracrine biosynthesis of dihydrotestosterone is the final step in anabolic androgen metabolism in the prostate. The NADPH-dependent steroid 5a-reductase isozymes irreversibly catalyze 5a-reduction of intracellular testosterone to dihydrotestosterone. In castration-recurrent prostate cancer mean mRNA levels suggest relative gene expression gradients of 5a-reductase-3 > 5a-reductase-1 5a-reductase-2. Furthermore, sufficient levels of testosterone and dihydrotestosterone were observed in castration-recurrent prostate cancer to activate androgen receptor signaling pathway. In intact and recurrent CWR22 human xenografts, persistent dihydrotestosterone formation was observed after pretreatment with dutasteride. Improved inhibitors that target 5a-reductase-1, 2 and 3 isozymes may stop intraprostatic DHT biosynthesis and prevent the development of clinical prostate cancer or its progression.

1 Introduction The first androgen, androsterone, a 5a-reduced 19-carbon steroid, was isolated from human male urine in 1929 by Nobel Laureate, Adolf Butenandt. Androsterone was considered the active human male hormone until 1935 when testosterone (T) was purified from several tons of bull testes (Fieser and Fieser 1959). This discovery suggested that the 5a-reduced androgens, androsterone, dihydrotestosterone (DHT), and 5a-androstane-3a,17b-diol (Fig. 1) were inactive catabolic metabolites of T (Fieser and Fieser 1959). 5a-Reductase was discovered by incubating desoxycorticosterone with rat liver slices and characterizing the major metabolite, 5a-pregnan-3a,21-diol-3,20-dione (Schneider and Horstmann 1951). It was proposed that 5a-reduction directed steroids toward catabolic 3-ketone reduction, phase II sulfation and glucuronidation, and excretion. Further studies showed that

M.A. Titus(*) Department of Urologic Oncology, Roswell Park Cancer Institute, Elm and Carlton Streets, Buffalo, NY 14263, USA, E-mail: [email protected]

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4_8, # Springer Science + Business Media, LLC 2009

175

176

M.A. Titus, J.L. Mohler

Fig. 1 Testosterone is irreversibly metabolized to the more potent androgen, dihydrotestosterone, and dihydrotestosterone catabolites, 3a-androstanediol and androsterone. 5a-Reductase isozymes (SRD5A) and aldo–keto reductase isozymes (AKR1C)

5a-reductase catalyzed the irreversible reduction of C19 and C21 steroids (epitestosterone, androstenedione, progesterone, 17a-hydroxy-progesterone, 20a-hydroxy-progesterone), which suggested steroid 5a-reductase might be a regulatory enzyme (Tomkins 1957). The enzymatic reduction is NADPH dependent and catalyzes trans addition of a hydride ion at carbon-5 and a proton from water at carbon-4 (Wilton and Ringold 1968). Reversal of this reaction was not observed using mammalian enzymes due to the high energy barrier for hydride ion abstraction from carbon-5 (Wilton and Ringold 1968). Bjorkhem (1969) showed that the pro-S-hydrogen from NADPH is delivered stereospecifically to the less hindered alpha face of 7a-hydroxycholest-4-en-3-one in rat liver microsomes. The same mechanism was proven for the reduction of T to DHT using rat seminal vesicles (Suzuki and Tamaoki 1974). The importance of T to DHT metabolism was recognized first when rats were injected with radiolabeled T, and it was found that DHT, not T, was the predominant androgen isolated from prostate nuclei (Anderson and Liao 1968; Bruchovsky and Wilson 1968). The detection of DHT in the nucleus within 5–10 min of T injection led to complete revision of classical androgen action and promoted DHT to the status of primary AR ligand in prostate physiology. The 5a-reduction of T to the more potent androgen, DHT, suggested that T, the major circulating androgen, acts as a prohormone in androgen prostate cells, which express steroid 5a-reductase enzymes. Furthermore, Bruchovsky et al. (1971) showed that androgen target tissues convert 3a-androstanediol and androsterone back to DHT. 5a-reduction of the D4–5 double bond in T creates the planar and less polar DHT with a trans A–B ring system and decreased 3-keto enolization. Structure–function analysis by Liao and Fang (1969) concluded that the more planar DHT fit more tightly than T to the AR-ligand binding domain. Thus, most of the enhanced affinity is due to the decreased dissociation rate of DHT from the AR–DHT complex (Grino et al. 1990).

2 5a-Reductase Isozyme Molecular Genetics The differential role of T and DHT in androgen signaling was determined initially during clinical studies of androgen insensitivity, an X-linked recessive genetic disorder. Simpson et al. (1971) described a family with three affected brothers

5a-Reductase Isozymes in Castration-Recurrent Prostate Cancer

177

who had 46 XY karyotype ambiguous genitalia as children and masculinized at puberty. However, inheritance of this reproductive disorder is autosomal recessive, and T and estrogen levels are normal in affected children. Walsh et al. (1974) described these patients as pseudohermaphrodites (type 2) and suggested that DHT biosynthesis was defective. Further clinical evidence was documented when Imperato-McGinley et al. (1974) studied 24 male pseudohermaphrodites in the Dominican Republic. The affected males present with ambiguous genitalia at birth, attain normal male secondary characteristics and spermatogenesis at puberty but suffer decreased fertility. These men have decreased acne, facial hair, impaired conversion of T to DHT, and small seminal vesicles and prostates. The primary sequence of human 5a-reductase (SRD5A, EC 1.3.99.5) isozymes was determined using cDNA libraries constructed from surgical specimens of prostate tissue. 5a-Reductase 1 was the first isozyme to be cloned and characterized biochemically (Anderson and Russell 1990). The 5a-reductase 2 isozyme was cloned in 1992 and distinguished from 5a-reductase 1 using biochemical and pharmacological assays (Jenkins et al. 1992). 5a-Reductase 1 and 2 are 259 and 254 amino acids in length with unique enzyme activity pH optima of 8.0 and 5.0, respectively. The predicted molecular weight of 5a-reductase is 28,000–29,000 Da. The sequence identity between the two proteins is 47%, and over one-third of amino acids are classified as hydrophobic (Anderson et al. 1991). 5a-Reductase functional domains were studied by identifying mutations in the 5a-reductase 2 gene (Anderson et al. 1991; Thigpen et al. 1992a; Thigpen et al. 1992b). One set of mutations affected the ability of the enzyme to bind the substrate T, and the other set decreased affinity for the cofactor, NADPH (Russell et al. 1994). Amino acid substitutions near the N or C terminal ends of 5a-reductase 1 and 2 increase the Km for T binding. To confirm substrate-binding location, a chimeric rat N-terminus/ human C-terminus steroid 5a-reductase 1 protein demonstrated rat-like finasteride sensitivity. In contrast, amino acid substitutions in the carboxylic half of the protein reduced NADPH cofactor binding (Thigpen, Davis, Gautier et al., 1992). In addition, Bhattacharyya et al. (1995) isolated and sequenced an NADPH photoaffinity-labeled carboxy terminal peptide from rat steroid 5a-reductase 1. The sequence consisted of 11 amino acids and corresponded to carboxy terminal residues 170–180.

3 5a-Reductase Isozymes in Prostate Cancer 5a-reductase 1 and 2 are located in androgen-responsive tissues, such as prostate, and exhibit distinct biochemical parameters (Russell and Wilson 1994). Both 5areductase 1 and 2 are active in androgen-stimulated benign prostate and are expressed at higher levels in high grade compared to low-grade clinically localized prostate cancer (CaP) (Thomas et al. 2008). However, compared to 5a-reductase 2, 5a-reductase 1 protein (Bonkhoff et al. 1996; Titus et al. 2005a) and gene expression (Iehle et al. 1999) increase in CaP versus androgen-stimulated benign prostate. Two inhibitors have been used in CaP chemoprevention, finasteride (in the PCPT clinical trial) (Thompson et al. 2003) and dutasteride (in the REDUCE

178

M.A. Titus, J.L. Mohler

clinical trial) (Andriole et al. 2004). Finasteride selectively inhibits 5a-reductase 2 (IC50 = 69 nM) and dutasteride inhibits both 5a-reductase 1 (IC50 = 6 nM) and 2 (IC50 = 7 nM) (Frye 2006). The PCPT study was terminated early due to a 25% reduction in CaP occurrence over a 7-year period, although the proportion of highgrade CaP cases was found to be increased. Thorpe et al. (2007) suggested the increased high-grade CaP is due to small prostate volume and biopsy sampling error. However, Cussenot et al. (2007) and Li et al. (2003) suggested that the most common 5a-reductase 2 V89L polymorphism, which decreases 5a-reductase 2 activity (Makridakis et al. 2000), is associated with more aggressive CaP risk and may explain the increase in high-grade CaP associated with finasteride use. Hsing et al. (2001) and Boger-Megiddo et al. (2008) advise caution in focusing on a single polymorphism and suggest that the 5a-reductase 2 V89L variant may be part of a set of enzymatic polymorphisms, which act together to increase CaP risk. Xu et al. (2006) hypothesize that dutasteride will reduce CaP growth only if used in combination with medical or surgical castration, because 5a-reductase inhibition will increase T levels that will activate AR and stimulate growth.

4 5a-Reductase Isozymes in Castration-Recurrent Prostate Cancer Men who fail curative therapy of clinically localized CaP or advanced disease usually receive androgen-deprivation therapy (ADT) that causes regression of androgen-dependent CaP through programmed cell death (Denmeade et al. 1996). However, ADT is palliative (Huggins, 1941) since CaP almost always recurs and causes death. A molecular role for androgen receptor in the transition from androgen-stimulated CaP to castration-recurrent CaP is supported by androgen receptor expression (Culig et al. 1998; de Vere White, 1997; Mohler et al. 2004) and expression of androgen-regulated genes (Gregory et al. 1998; Mousses et al. 2001) in castration-recurrent CaP. Many laboratories have focused upon mechanisms for androgen receptor transactivation despite castrate levels of circulating testicular androgens, such as androgen receptor amplification (Visakorpi et al. 1995), gene mutations (Culig et al. 2000), promiscuous ligand binding (Tan et al. 1997), and/or post-translational modification through peptide growth factor signaling (Arnold and Isaacs 2002; Feldman and Feldman 2001). However, Mohler et al. (2004) and Titus et al. (2005b) and others (Belanger et al. 1989; Geller et al. 1979) have demonstrated that androgen-stimulated CaP and castration-recurrent CaP exhibit tissue levels of DHT sufficient for activation of even wild-type androgen receptor. Castration-recurrent CaP must synthesize tissue T from weak circulating adrenal androgens, androstenedione and DHEA, or cholesterol and metabolize T to DHT at levels sufficient for androgen receptor activation. Limited clinical experience has not demonstrated efficacy for finasteride added to ADT for treatment of advanced CaP (Fleshner and Trachtenberg 1995) or

5a-Reductase Isozymes in Castration-Recurrent Prostate Cancer

179

bispecific 5a-reductase inhibitor (LY320236) for secondary treatment of castrationrecurrent CaP (Eisenberger et al. 2004). Additionally, dutasteride was shown to decrease serum DHT to 95% of normal levels (Clark et al. 2004) and decrease intraprostatic DHT by 94% in human androgen-stimulated benign prostate (Wurzel et al. 2007) compared to placebo controls. But DHT persists at measurable levels despite high doses of bispecific 5a-reductase inhibition. These clinical observations led to speculation about the existence of a third 5areductase enzyme when the Human Genome Project identified a sequence consistent with a 5a-reductase isozyme that mapped to chromosome 4q12 (Ota et al. 2004; Strausberg et al. 2002). 5a-Reductase 1, 2, and 3 (Mohler et al. 2006) mRNA levels were quantified in androgen-stimulated benign prostate, androgen-stimulated CaP, and castration-recurrent CaP. Mean 5a-reductase 1 mRNA levels decreased 30% in androgen-stimulated CaP and 75% in castration-recurrent CaP normalized to androgen-stimulated benign prostate. 5a-Reductase 2 gene expression was negligible in androgen-stimulated CaP and was not detected in castration-recurrent CaP. 5aReductase 3 mRNA levels decreased 30% in androgen-stimulated CaP and 45% in castration-recurrent CaP versus androgen-stimulated benign prostate. Moreover, Uemura et al. (2008) observed upregulation of 5a-reductase 3 mRNA in castrationrecurrent CaP. These results suggested a relative gene expression gradient of 5areductase 1 5a-reductase 3 5a-reductase 2 in androgen-stimulated CaP specimens. The mean mRNA levels in castration-recurrent CaP specimens suggested an expression gradient of 5a-reductase 3 > 5a-reductase 1 5a-reductase 2 (Titus et al. 2007). The relative 5a-reductase isozyme mRNA contribution suggested that (1) 5a-reductase 1 and 5a-reductase 2 mRNA correlated with 5a-reductase isozyme protein levels and enzymatic activities (Titus et al. 2005a) and (2) 5a-reductase 3 mRNA expression levels were similar in castration-recurrent CaP and androgenstimulated CaP, and maintained at higher levels than either 5a-reductase 1 or 5areductase 2 in castration-recurrent CaP. EST analysis also demonstrated 5a-reductase 3 upregulation in CaP compared to normal prostate controls (Table 1, Unigene: Hs.552, 5a-reductase 1; Hs.458345, 5a-reductase 2; 5a-reductase 3, Hs.39311). 5a-Reductase 3 ESTs were not observed in cancer of the adrenal gland or bladder. To study 5a-reductase 3 enzyme activity the androgen-dependent CWR22 human CaP xenograft model was used. Persistent DHT formation was observed using TLC analysis after dutasteride (5 nM) pretreatment of protein lysates from CWR22 intact or recurrent tumors. Incubations at pH 7.0 using 1 mM labeled (1 mCi-3H) androstenedione and CWR22 intact or CWR22 recurrent protein lysates Table 1 5a-Reductase 1, 2, and 3 (5aR) express sequence tags (EST) measured as transcripts per million (TMP) in cancer and normal tissues Cancer tissue TPM Normal tissue TPM 5aR1 5aR2 5aR3 5aR1 5aR2 5aR3 Prostate 20 0 48 Prostate 21 48 31 Skin 40 0 64 Skin 47 0 47 Adrenal 78 0 0 Adrenal 90 0 0 Bladder 0 – 0 Bladder 0 0 0

180

M.A. Titus, J.L. Mohler

produced DHT at 19% (intact) or 17% (recurrent), and 5a-ASD at 15% (intact) or 19% (recurrent) of controls in absence of DHT, respectively (Titus et al. 2007). 5aReductase 3 purified recombinant protein expressed in CHO cells also exhibited T conversion to DHT by a tandem mass spectrometry assay. Hence, 5a-reductase 3 can convert both androstenedione and T to 5a-reduced androgens, 5a-androstanedione and DHT, and the bispecific inhibitor, dutasteride, does not block 5a-reductase 3 activity. This unique 5a-reductase may play a significant role in the persistent DHT levels observed in castration-recurrent CaP and should be inhibited to block completely DHT biosynthesis.

5 5a-Reductase Intracellular Location in Castration-Recurrent Prostate Cancer 5a-reductase isozyme location and biochemical properties are as important as expression levels in determining intracellular CaP DHT concentration. Moore and Wilson (1972) used rat prostate organelle fractionation to localize 5a-reductase enzyme activity to the nuclear membrane, and Houston et al. showed 5a-reductase activity was located in human prostate epithelial nuclei. Savory et al. (1995) immunolocalized 5a-reductase 1 to the nuclear periphery in rat prostate, and Aumu¨ller et al. (1996) observed 5a-reductase 1 in prostate epithelial and stromal nuclei by immunohistochemistry. 5a-Reductase 2 was immunolocalized to granular endoplasmic reticulum in prostate tissue (Eicheler et al. 1994); however, 5areductase 2 mRNA and enzyme activity were decreased in CaP specimens (Luo et al. 2003; Soderstrom et al. 2001). In castration-recurrent CaP, the majority of 5areductase 1 immunostaining was perinuclear, whereas 5a-reductase 2 immunostaining was decreased and dispersed in the cytoplasm (Titus et al. 2005a). 5a-Reductase 3 immunostaining was equally distributed between nuclei and cytoplasm in castration-recurrent CaP using a monoclonal antibody. The overall 5a-reductase isozyme distribution and expression suggest that cytoplasmic AR, bound to heat shock proteins, will compete with 5a-reductase 1, 2, and 3 for intracellular T in castration-recurrent CaP cells. The AR equilibrium dissociation constant (KD) for T is 0.49 nM (Wilson and French 1976), and the apparent 5a-reductase 1 and 2 dissociation constant (Km) (Fersht 1985) is 6.3 mM and 6.3 nM (Frye 2006), respectively (Km 6¼ KD, comparison is to illustrate concept). 5a-Reductase 3 kinetic parameters are unknown but may be similar to 5a-reductase 1 and 2. Hence, increased T levels in castration-recurrent CaP (Mohler et al. 2004; Titus et al. 2005b) should interact primarily with AR and then 5a-reductase 1 and 3. AR nuclear translocation will deliver T to the nucleus and concentrate T. Bruchovsky et al. (1975) suggested androgen levels can approach 200–250 nM. More recently, Gioeli et al. (2006) proposed that AR ligand-binding cycle occurs in the nucleus as well as the cytoplasm. Therefore, concentrated intranuclear T may be converted to DHT by nuclear 5a-reductase 1 and 3, and displace T as DHT binds to AR to promote

5a-Reductase Isozymes in Castration-Recurrent Prostate Cancer

181

stable AR gene activation. Taken together, the actual nuclear DHT level may be higher than levels measured in whole tissue using ELISA or tandem mass spectrometry assays.

6 Conclusion Clinical and laboratory observations support the existence of a novel third 5a-reductase enzyme in castration-recurrent CaP. The persistent DHT levels observed in serum and prostate specimens of men receiving dutasteride (5a-reductase 1 and 5a-reductase 2 dual inhibitor) suggest that 5a-reductase 3 in peripheral and androgen-responsive tissue metabolize T to DHT. Both the adrenal androgens and/or cholesterol may be precursors for intracellular DHT formation. Moreover, AR nuclear translocation may concentrate the low DHT levels observed in prostate specimens to the cancer epithelial nuclei and promote AR gene expression. The inability of finasteride or dutasteride to eliminate DHT suggests that 5a-reductase 3 is an important target in CaP, and possibly acne and male pattern baldness. Thus, a trispecific inhibitor may be required to block completely biosynthesis of DHT in CaP and castration-recurrent CaP.

References Anderson, K. M., & Liao, S. (1968). Selective retention of dihydrotestosterone by prostatic nuclei. Nature, 219(5151), 277–279. Anderson, S., & Russell, D. W. (1990). Structural and biochemical properties of cloned and expressed human and rat steroid 5a-reductase. Proc Natl Acad Sci USA, 87, 3640–3644. Anderson, S., Berman, D. M., Jenkins, E. P., & Russell, D. W. (1991). Deletion of steroid 5a-reductase 2 gene in male pseudohermaphroditism. Nature, 354, 159–161. Andriole, G., Bostwick, D., Brawley, O., Gomella, L., Marberger, M., Tindall, D., et al. (2004). Chemoprevention of prostate cancer in men at high risk: rationale and design of the reduction by dutasteride of prostate cancer events (REDUCE) trial. J Urol, 172(4 Pt 1), 1314–1317. Arnold, J. T., & Isaacs, J. T. (2002). Mechanisms involved in the progression of androgenindependent prostate cancers: it is not only the cancer cell’s fault. Endocr Relat Cancer, 9 (1), 61–73. Aumu¨ller, G., Eicheler, W., Renneberg, H., Adermann, K., Vilja, P., & Forssmann, W. G. (1996). Immunocytochemical evidence for differential subcellular localization of 5 alpha-reductase isoenzymes in human tissues. Acta Anat (Basel), 156, 241–252. Belanger, B., Belanger, A., Labrie, F., Dupont, A., Cusan, L., & Monfette, G. (1989). Comparison of residual C-19 steroids in plasma and prostatic tissue of human, rat and guinea pig after castration: unique importance of extratesticular androgens in men. J Steroid Biochem, 32(5), 695–698. Bhattacharyya, A. K., Chavan, A. J., Haley, B. E., Taylor, M. F., & Collins, D. C. (1995). Identification of the NADP(H) binding site of rat liver microsomal 5 alpha-reductase (isozyme-1): purification of a photolabeled peptide corresponding to the adenine binding domain. Biochemistry, 34(11), 3663–3669.

182

M.A. Titus, J.L. Mohler

Bjorkhem, I. (1969). Mechanism and stereochemistry of the enzymatic conversion of a delta 43-oxosteroid into a 3-oxo-5alpha-steroid. Eur J Biochem, 8(3), 345–351. Boger-Megiddo, I., Weiss, N. S., Barnett, M. J., Goodman, G. E., & Chen, C. (2008). V89L polymorphism of the 5alpha-reductase type II gene (SRD5A2), endogenous sex hormones, and prostate cancer risk. Cancer Epidemiol Biomarkers Prev, 17(2), 286–291. Bonkhoff, H., Stein, U., Aumuller, G., & Remberger, K. (1996). Differential expression of 5 alphareductase isoenzymes in the human prostate and prostatic carcinomas. Prostate, 29(4), 261–267. Bruchovsky, N. (1971). Comparison of the metabolites formed in rat prostate following the in vivo administration of seven natural androgens. Endocrinology, 89(5), 1212–1222. Bruchovsky, N., & Wilson, J. D. (1968). The intranuclear binding of testosterone and 5-alphaandrostan-17beta-ol-3-one by rat prostate. J Biol Chem, 243(22), 5953–5960. Bruchovsky, N., Rennie, P. S., &Vanson, A. (1975). Studies on the regulation of the concentration of androgens and androgen receptors in nuclei of prostatic cells. Biochim Biophys Acta, 394(2), 248–266. Clark, R. V., Hermann, D. J., Cunningham, G. R., Wilson, T. H., Morrill, B. B., & Hobbs, S. (2004). Marked suppression of dihydrotestosterone in men with benign prostatic hyperplasia by dutasteride, a dual 5alpha-reductase inhibitor. J Clin Endocrinol Metab, 89(5), 2179–2184. Culig, Z., Hobisch, A., Hittmair, A., Peterziel, H., Cato, A. C., Bartsch, G., et al. (1998). Expression, structure, and function of androgen receptor in advanced prostatic carcinoma. Prostate, 35(1), 63–70. Culig, Z., Hobisch, A., Bartsch, G., & Klocker, H. (2000). Expression and function of androgen receptor in carcinoma of the prostate. Microsc Res Tech, 51(5), 447–455. Cussenot, O., Azzouzi, A. R., Nicolaiew, N., Mangin, P., Cormier, L., Fournier, G., et al. (2007). Low-activity V89L variant in SRD5A2 is associated with aggressive prostate cancer risk: an explanation for the adverse effects observed in chemoprevention trials using 5-alpha-reductase inhibitors. Eur Urol, 52(4), 1082–1087. Denmeade, S. R., Lin, X. S., & Isaacs, J. T. (1996). Role of programmed (apoptotic) cell death during the progression and therapy for prostate cancer. Prostate, 28(4), 251–265. de Vere White, R. (1997). Human androgen receptor expression in prostate cancer following androgen ablation. Eur Urol, 31, 1–6. Eicheler, W., Tuohimaa, P., Vilja, P., Adermann, K., Forssmann, W. G., & Aumu¨ller, G. (1994). Immunocytochemical localization of human 5 alpha-reductase 2 with polyclonal antibodies in androgen target and non-target human tissues. J Histochem Cytochem, 42, 667–675. Eisenberger, M. A., Laufer, M., Vogelzang, N. J., Sartor, O., Thornton, D., Neubauer, B. L., et al. (2004). Phase I and clinical pharmacology of a type I and II, 5-alpha-reductase inhibitor (LY320236) in prostate cancer: elevation of estradiol as possible mechanism of action. Urology, 63(1), 114–119. Feldman, B. J., & Feldman, D. (2001). The development of androgen-independent prostate cancer. Nat Rev Cancer, 1(1), 34–45. Fersht, A. (1985). Enzyme Structure and Mechanism. New York: W. H. Freeman and Company. Fieser, L. F., & Fieser, M. (1959). Steroids. Reinhold, New York. Fleshner, N. E., & Trachtenberg, J. (1995). Combination finasteride and flutamide in advanced carcinoma of the prostate: effective therapy with minimal side effects. J Urol, 154(5), 1642–1645; discussion 1645–1646. Frye, S. V. (2006). Discovery and clinical development of dutasteride, a potent dual 5alphareductase inhibitor. Curr Top Med Chem, 6(5), 405–421. Geller, J., Albert, J., & Loza, D. (1979). Steroid levels in cancer of the prostate–markers of tumour differentiation and adequacy of anti-androgen therapy. J Steroid Biochem, 11(1B), 631–636. Gioeli, D., Black, B. E., Gordon, V., Spencer, A., Kesler, C. T., Eblen, S. T., et al. (2006). Stress kinase signaling regulates androgen receptor phosphorylation, transcription, and localization. Mol Endocrinol, 20(3), 503–515.

5a-Reductase Isozymes in Castration-Recurrent Prostate Cancer

183

Gregory, C. W., Hamil, K. G., Kim, D., Hall, S. H., Pretlow, T. G., Mohler, J. L., et al. (1998). Androgen receptor expression in androgen-independent prostate cancer is associated with increased expression of androgen-regulated genes. Cancer Res, 58(24), 5718–5724. Grino, P. B., Griffin, J. E., & Wilson, J. D. (1990). Testosterone at high concentrations interacts with the human androgen receptor similarly to dihydrotestosterone. Endocrinology, 126(2), 1165–1172. Hsing, A. W., Chen, C., Chokkalingam, A. P., Gao, Y. T., Dightman, D. A., Nguyen, H. T., et al. (2001). Polymorphic markers in the SRD5A2 gene and prostate cancer risk: a population-based case-control study. Cancer Epidemiol Biomarkers Prev, 10(10), 1077–1082. Huggins, C. S. (1941). Studies on prostatic cancer:2. the effects of castration on advanced carcinoma of the prostate gland. Arch Surg, 43, 209–212 Iehle, C., Radvanyi, F., Gil Diez de Medina, S., Ouafik, L. H., Gerard, H., Chopin, D., et al. (1999). Differences in steroid 5alpha-reductase iso-enzymes expression between normal and pathological human prostate tissue. J Steroid Biochem Mol Biol, 68(5–6), 189–195. Imperato-McGinley, J., Guerrero, L., Gautier, T., & Peterson, R. E. (1974). Steroid 5alphareductase deficiency in man: an inherited form of male pseudohermaphroditism. Science, 186(4170), 1213–1215. Jenkins, E. P., Andersson S., Imperato-McGinley, J., Wilson, J. D., & Russell, D. W. (1992). Genetic and pharmacological evidence for more than one human steroid 5 alpha-reductase. J Clin Invest, 89(1), 293–300. Li, Z., Habuchi, T., Mitsumori, K., Kamoto, T., Kinoshitu, H., Segawa, T., et al. (2003). Association of V89L SRD5A2 polymorphism with prostate cancer development in a Japanese population. J Urol, 169(6), 2378–2381. Liao, S., & Fang, S. (1969). Receptor-proteins for androgens and the mode of action of androgens on gene transcription in ventral prostate. Vitam Horm, 27, 17–90. Luo, J., Dunn, T. A., Ewing, C. M., Walsh, P. C., & Isaacs, W. B. (2003). Decreased gene expression of steroid 5 alpha-reductase 2 in human prostate cancer: implications for finasteride therapy of prostate carcinoma. Prostate, 57(2), 134–139. Makridakis, N. M., di Salle, E., & Reichardt, J. K. (2000). Biochemical and pharmacogenetic dissection of human steroid 5 alpha-reductase type II. Pharmacogenetics, 10(5), 407–413. Mohler, J. L., Gregory, C. W., Ford, O. H., III, Kim, D., Weaver, C. M., Petrusz, P., et al. (2004). The androgen axis in recurrent prostate cancer. Clin Cancer Res, 10(2), 440–448. Mohler, J. L., Titus, M. A., Kozyreva, O. G. Ford, O. H. III, Kawinski, E., & Li, Y. (2006). Novel steroid 5a-reductase. USA. Moore, R. J., & Wilson, J. D. (1972). Localization of the reduced nicotinamide adenine dinucleotide phosphate: 4-3-ketosteroid 5-oxidoreductase in the nuclear membrane of the rat ventral prostate. J Biol Chem, 247(3), 958–967. Mousses, S., Wagner, U., Chen, Y., Kim, J. W., Bubendorf, L., Bittner, M., et al. (2001). Failure of hormone therapy in prostate cancer involves systematic restoration of androgen responsive genes and activation of rapamycin sensitive signaling. Oncogene, 20(46), 6718–6723. Ota, T., Suzuki, Y., Nishikawa, T., Otsuki, T., Sugiyama, T., Irie, R., et al. (2004). Complete sequencing and characterization of 21,243 full-length human cDNAs. Nat Genet, 36(1), 40–45. Russell, D. W., & Wilson, J. D. (1994). Steroid 5 alpha-reductase: two genes/two enzymes. Annu Rev Biochem, 63, 25–61. Russell, D. W., Berman, D. M., Bryant, J. T., Cala, K. M., Davis, D. L., Landrum, C. H., et al. (1994). The molecular genetics of steroid 5a-reductase. Recent Prog Horm Res, 49, 275–284. Savory, J. G., May, D., Reich, T., La Casse, E. C., Lakins, J., Tenniswood, M., et al. (1995). 5 alpha-Reductase type 1 is localized to the outer nuclear membrane. Mol Cell Endocrinol, 110 (1–2), 137–147. Schneider, J. J., & Horstmann, P. M. (1951). Effects of incubating desoxycorticosterone with various rat tissues. J Biol Chem, 191(1), 327–338. Simpson, J. L., New, M., Peterson, R. E., & German, J. (1971). Pseudovaginal perineoscrotal hypospadias (PPSH) in sibs. Birth Defects Orig Artic Ser, 7(6), 140–144.

184

M.A. Titus, J.L. Mohler

Soderstrom, T. G., Bjelfman, C., Brekkan, E., Ask, B., Egevad, L., Norlen, B. J., et al. (2001). Messenger ribonucleic acid levels of steroid 5 alpha-reductase 2 in human prostate predict the enzyme activity. J Clin Endocrinol Metab, 86(2), 855–858. Strausberg, R. L., Feingold, E. A., Grouse, L. H., Derge, J. G., Klausner, R. D., Collins, F. S., et al. (2002). Generation and initial analysis of more than 15,000 full-length human and mouse cDNA sequences. Proc Natl Acad Sci USA, 99(26), 16899–16903. Suzuki, K., & Tamaoki, B. (1974). In vitro metabolism of testosterone in seminal vesicles of rats. JSteroid Biochem, 5(3), 249–256. Tan, J., Sharief, Y., Hamil, K. G., Gregory, C. W., Zang, D. Y., Sar, M., et al. (1997). Dehydroepiandrosterone activates mutant androgen receptors expressed in the androgen-dependent human prostate cancer xenograft CWR22 and LNCaP cells. Mol Endocrinol 11(4), 450–459. Thigpen, A. E., Davis, D. L., Gautier, T., Imperato-McGinley, J., & Russell, D. W. (1992a). Brief report: the molecular basis of steroid 5 alpha-reductase deficiency in a large Dominican kindred. N Engl J Med, 327(17), 1216–1219. Thigpen, A. E., Davis, D. L., Milatovich, A., Mendonca, B. B., Imperato-McGinley, J., Griffin, J. E., et al. (1992b). Molecular genetics of steroid 5 alpha-reductase 2 deficiency. J Clin Invest, 90 (3), 799–809. Thomas, L. N., Douglas, R. C., Lazier, C. B., Too, C. K., Rittmaster, R. S., & Tindall, D. J. (2008). Type 1 and type 2 5alpha-reductase expression in the development and progression of prostate cancer. Eur Urol, 53(2), 244–252. Thompson, I. M., Goodman, P. J., Tangen, C. M., Lucia, M. S., Miller, G. J., Ford, L. G., et al. (2003). The influence of finasteride on the development of prostate cancer. N Engl J Med, 349 (3), 215–224. Thorpe, J. F., Jain, S., Marczylo, T. H., Gescher, A. J., Steward, W. P., & Mellon, J. K. (2007). A review of phase III clinical trials of prostate cancer chemoprevention. Ann R Coll Surg Engl, 89 (3), 207–211. Titus, M. A., Gregory, C. W., Ford, O. H., III, Schell, M. J., Maygarden, S. J., & Mohler, J. L. (2005a). Steroid 5alpha-reductase isozymes I and II in recurrent prostate cancer. Clin Cancer Res, 11(12), 4365–4371. Titus, M. A., Schell, M. J., Lih, F. B., Tomer, K. B., & Mohler, J. L. (2005b). Testosterone and dihydrotestosterone tissue levels in recurrent prostate cancer. Clin Cancer Res, 11(13), 4653–4657. Titus, M. A., Li, Y., Kawinski, E., Kozyreva, O., & Mohler, J. L. (2007). Biochemistry and pharmacological evidence for a third isozyme of steroid 5a-reductase in prostate cancer. J Urol, 177 (suppl):268. Tomkins, G. M. (1957). The enzymatic reduction of delta 4–3-ketosteroids. J Biol Chem, 225(1), 13–24. Uemura, M., Tamura, K., Chung, S., Honma, S., Okuyama, A., Nakamura, Y., et al. (2008). Novel 5 alpha-steroid reductase (SRD5A3, type-3) is overexpressed in hormone-refractory prostate cancer. Cancer Sci, 99(1), 81–86. Visakorpi, T., Hyytinen, E., Koivisto, P., Tanner, M., Keinanen, R., Palmberg, C., et al. (1995). In vivo amplification of the androgen receptor gene and progression of human prostate cancer. Nat Genet, 9(4), 401–406. Walsh, P. C., Madden, J. D., Harrod, M. J., Goldstein, J. L., MacDonald, P. C., & Wilson, J. D. (1974). Familial incomplete male pseudohermaphroditism, type 2. Decreased dihydrotestosterone formation in pseudovaginal perineoscrotal hypospadias. N Engl J Med, 291(18), 944–949. Wilson, E. M., & French, F. S. (1976). Binding properties of androgen receptors. Evidence for identical receptors in rat testis, epididymis, and prostate. J Biol Chem, 251(18), 5620–5629. Wilton, D. C., & Ringold, H. J. (1968). Mechanism of enzymatic reduction of 4-3-ketosteroids by liver enzymes. Third International Congress of Endocrinology, 157, 105–166.

5a-Reductase Isozymes in Castration-Recurrent Prostate Cancer

185

Wurzel, R., Ray, P., Major-Walker, K., Shannon, J., & Rittmaster, R. (2007). The effect of dutasteride on intraprostatic dihydrotestosterone concentrations in men with benign prostatic hyperplasia. Prostate Cancer Prostatic Dis, 10(2), 149–154. Xu, Y., Dalrymple, S. L., Becker, R. E., Denmeade, S. R., & Isaacs, J. T. (2006). Pharmacologic basis for the enhanced efficacy of dutasteride against prostatic cancers. Clin Cancer Res, 12(13), 4072–4079.

Improving Intermittent Androgen-Deprivation Therapy: OFF Cycle and the Role of Steroid 5a-Reductase Inhibitors Shubham Gupta, Daniel Shevrin, and Zhou Wang

Abstract Intermittent androgen-deprivation therapy (IADT) consists of cycles of androgen-deprivation (ON cycle) and treatment-free periods (OFF cycle), and is under investigation as a means to improve quality of life and retard progression to castration-recurrent prostate cancer in patients with advanced prostate cancer. IADT appears to be as effective as continuous androgen-deprivation therapy, while improving the side-effect profile. During the OFF cycle of IADT, the androgen-deprived prostate cancer cells are re-exposed to androgens. This milieu of androgen-induced regrowth provides a unique window of opportunity for designing OFF cycle-specific interventions. Testosterone, relative to dihydrotestosterone (DHT), is more potent in the induction of growth suppressive androgen-response genes during regrowth of regressed prostate. These findings suggest that during the OFF cycle of IADT, steroid 5a-reductase (SRD5A) inhibitors can be used to enhance the expression of tumor suppressive androgen-response genes and retard tumor growth by blocking conversion of testosterone to DHT. This chapter discusses how addition of SRD5A inhibitors during the OFF cycle of IADT could improve survival under certain clinical conditions. The molecular changes that accompany treatment suggest that prolongation of the OFF cycle may not be advisable.

1 Introduction Prostate cancer survival has improved consistently over the past three decades. Prostate cancer is the most common noncutaneous malignancy in United States (US) men and accounts for about 30% of all such neoplasms (American Cancer Z. Wang(*) University of Pittsburgh School of Medicine, 5200 Centre Ave, Ste G03, Pittsburgh, PA, 15232, USA, E-mail: [email protected]

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4_9, # Springer Science + Business Media, LLC 2009

187

188

S. Gupta et al.

Society 2007). Prostate cancer is second only to lung cancer as a cause of cancer-specific mortality and accounts for 9% of cancer deaths. The 5-year survival was 69% during 1975–1977, 76% during 1984–1986, and 100% during 1996–2002. Current data suggest that while one man in six will be diagnosed with prostate cancer during his lifetime, only 1 in 35 will die of it (American Cancer Society 2007). The present survival figures are attributable to increased diagnosis and aggressive treatment in both the preprostate-specific antigen (PSA) and PSA era (Klein et al. 2007). Given the long natural history of prostate cancer, whether PSA screening by itself has reduced prostate cancer mortality remains unclear (Thompson and Ankerst 2007). However, PSA screening has led to significant clinical and pathological stage migration; more and more cancers are being diagnosed at early stages (Newcomer et al. 1997; Thompson and Ankerst 2007). The high incidence and the long natural history of prostate cancer make it a disease with enormous social and economic burdens. There are close to 2 million men living with prostate cancer in the US, and this number is expected to increase (American Cancer Society 2007). The medical expenditure for prostate cancer was $1.3 billion in 2000 (Penson and Chan 2007). In addition, the indirect costs due to premature mortality equal the direct medical costs of prostate cancer (Max et al. 2002). Prostate cancer is primarily a disease of old age. A multitude of factors including androgens, genetic predisposition, inflammation pathways, and inactivation of caretaker genes influence the development of prostate cancer. The role of androgens in prostate carcinogenesis is critical but incompletely understood. Androgens are required for the development of the prostate, and exposure to androgens over a long period of time is required for prostate carcinogenesis (Klein et al. 2007). In the PSA era, more than 90% of new prostate cancer cases are diagnosed when the cancer is confined locally or regionally (American Cancer Society 2007). Several options are available for managing these cases, which include watchful waiting, active surveillance, radical prostatectomy, external beam radiotherapy, androgen-deprivation therapy (ADT); and also newer treatments like cyroablation, radiofrequency ablation, and high-intensity focused ultrasound. Locally advanced prostate cancer may require combinations of two or three modalities. Recurrence after potentially curative therapy and metastatic disease at presentation are managed usually with ADT. The response of prostate cancer to ADT is consistent and dramatic. Most patients exhibit a strong and prolonged response manifested by improvement in symptoms and objective measures of disease response. Ever since Huggins’ seminal papers in 1941 (Huggins and Hodges 1941; Huggins et al. 1941), few cancer interventions have come close to ADT in terms of simplicity of concept and reliability of response. Options for ADT have evolved from Huggins’ orchiectomy to maximal androgen blockade using gonadotropin-releasing hormone (GnRH) agonists and antiandrogens. Shahinian et al. (2005) reported that the use of ADT has increased dramatically for all stages and grades of prostate cancer; being used even in circumstances where ADT has not been shown to be of benefit. For instance, in men over 80 years of age with localized stage and low- to moderate-grade tumors, primary GnRH use increased 9-fold from 3.7% in 1991 to 30.9% in 1999.

Improving Intermittent Androgen-Deprivation Therapy

189

An analysis of Medicare claims data estimated that one-third of all patients diagnosed with prostate cancer underwent ADT (Krupski et al. 2006). The annual costs for prostate cancer related costs are highest in patients treated with ADT (Wilson et al. 2007); men on ADT cost the health care system almost twice as much as those not on ADT (Krupski et al. 2007). In 2001, the Medicare reimbursements for GnRH agonists exceeded 1 billion dollars, which ranked as the second highest part B drug expenditure (Shahinian et al. 2005). Costs of ADT extend beyond the drug costs. Complications of ADT include osteoporosis, hot flashes, sexual dysfunction, cognitive dysfunction, anemia, gynecomastia, and changes in body habitus. The most dreaded and inevitable eventuality of ADT is progression to castration-recurrent prostate cancer, where the tumor continues to grow despite castrate levels of circulating testicular androgens. ADT has limitations, and strategies to reduce the economic burden, improve the side-effect profile, and delay progression to castration-recurrent prostate cancer would be embraced quickly.

2 Intermittent Androgen-Deprivation Therapy ADT as currently practiced is usually continuous and lifelong. Intermittent androgendeprivation therapy (IADT) has been proposed as a viable alternative for selected patients. The patients are treated cyclically, wherein periods of ADT (On treatment time, or ON cycle) are alternated with treatment-free periods (Time OFF period, or OFF cycle) when androgen levels recover (Fig. 1). For each cycle, ADT is continued until the PSA drops below a pre‐established threshold. Sometimes, ADT continues for a specified time period after the PSA drops below the threshold. Patients are then observed without treatment, and ADT is restarted after serum PSA reaches a preset arbitrary level which varies among protocols. Patients are no longer eligible for OFF cycle periods if PSA continues to increase, or if other evidence of disease progression develops despite ADT. In the Shionogi androgen-dependent mouse mammary carcinoma model, castration induces tumor regression, but also leads to androgen independence (Bruchovsky et al. 1990). Bruchovsky’s group expanded on these ideas to develop IADT. They proposed that the replacement of androgens at the end of a period of apoptotic regression might result in the regeneration of differentiated tumor cells with further apoptotic potential. IADT would promote tumor cytoreduction in the treatment period and the expansion of androgen-sensitive clones in the OFF treatment period (Akakura et al. 1993). These androgen-dependent cells, by competing with androgen-independent cells, would prevent occupation of the tumor by androgenindependent elements and would sustain androgen dependence for longer periods (Rennie et al. 1990; Akakura et al. 1993; Sato et al. 1996). These laboratory studies and the presumed benefits of OFF cycle periods, which include improved quality of life (QOL) and reduced costs, made a strong case for clinical trial of IADT. The notion that IADT would lead to improvement in overall QOL is based on the fact that many of the adverse effects of ADT are related to the lack of

190

S. Gupta et al.

Fig. 1 Schematic illustration of the phases of intermittent androgen-deprivation therapy (IADT) and the expected cycling of prostate-specific antigen (PSA) and testosterone. One full cycle of IADT consists of 1 ‘‘ON Cycle’’ + 1 ‘‘OFF Cycle.’’ Note the rapid fall in PSA and testosterone levels upon androgen deprivation, and the more gradual rise upon cessation of ADT. The graph is representative of the expected changes in serum PSA and testosterone levels, and does not depict any reported dataset. The OFF cycle of IADT provides the opportunity to design interventions that target androgen-deprived prostate cancer cells that are re-exposed to androgens. The inhibition of testosterone to DHT conversion by the addition of a steroid 5a-reductase (SRD5A) inhibitor (such as finasteride or dutasteride) during the OFF cycle may selectively enhance the expression of tumor suppressive genes and improve survival

androgens and are reversible upon androgen replenishment. During the OFF cycles of IADT, testosterone levels recover and side effects are alleviated (Albrecht et al. 2003; Mottet et al. 2005). IADT appears sound theoretically and firmly grounded on preclinical data; however, insufficient clinical data prevent/preclude unconditional endorsement. Klotz et al. (1986) reported the first clinical use of IADT using diethylstilbestrol (DES) and flutamide in 20 patients. Subsequently, several phase II (Goldenberg et al. 1995; Higano et al. 1996; Oliver et al. 1997; Crook et al. 1999; Strum et al. 2000; Sato et al. 2004; Peyromaure et al. 2005; Bruchovsky et al. 2006; BocconGibod et al. 2007) studies have evaluated IADT, and benefits appear more definitive on side effects than overall survival (OS). In a study of 250 men with locally advanced or metastatic prostate cancer treated with IADT, QOL was assessed every 3 months for 30 months using the European Organization for Research and Treatment of Cancer QOL Questionnaire (EORTC

Improving Intermittent Androgen-Deprivation Therapy

191

QLQ-C30) and EORTC prostate cancer module (QLQ-PR25) (Spry et al. 2006). Testosterone suppression led to a significant reduction in global health-related QOL (HrQOL) and deterioration in most function and symptom scales. During the OFF cycles, a trend of progressive improvement in HrQOL paralleled testosterone recovery. Maximum recovery of HrQOL occurred most frequently by months 9–12. Testosterone recovery was slower and less complete in older men, and led to concomitant poorer HrQOL recovery. However, Bouchot et al. (2000) reported no difference in QOL function assessed using EORTC QLQ-C30 between ON and OFF cycles in 43 patients on IADT. IADT appears beneficial for sexual function. Klotz et al. (1986) reported that 90% of the patients who were rendered impotent after initiation of ADT resumed sexual activity within 3 months of stopping treatment. Malone et al. (2005) reported the recovery of sexual function in 47% of the 38 IADT cycles that were evaluated. A phase III study by Da Silva et al. comparing IADT to continuous ADT showed that at 12 months, 29% of men receiving IADT reported sexual activity in the previous month, compared with 9% of men on continuous ADT (Boccon-Gibod et al. 2007). Among the other side effects of ADT, the loss of bone mineral density is reversible in part during the OFF cycle of IADT (Higano et al. 2004). Bruchovsky et al. (2007) reported the recovery of blood hemoglobin to normal levels during the OFF cycles of IADT. Increasing evidence supports that ADT leads to metabolic syndrome (Braga-Basaria et al. 2006), and cardiovascular mortality is more prevalent in patients on ADT (Tsai et al. 2007). Patients managed with IADT may experience fewer of these complications. A recent Cochrane review compared IADT and continuous ADT using data from five randomized studies involving 1,382 patients (Conti et al. 2007). All included studies involved men with advanced prostate cancer, had relatively small populations, and were of short duration. No data were available that allowed assessment of the relative effectiveness of IADT versus continuous ADT on overall survival, disease-specific survival, or disease progression. There was a trend toward lower adverse events in IADT groups versus continuous ADT groups, but accurate interpretation was limited due to inadequate sample sizes. The authors concluded that IADT may have reduced slightly adverse events, and that IADT was superior to continuous ADT for erectile function during the OFF cycle periods (Conti et al. 2007). Miller et al. (2007) reported on 335 patients with advanced prostate cancer who were randomized to continuous ADT or IADT. The median time to progression was longer in the intermittent arm (16.6 months) compared with the continuous arm (11.5 months), although the difference was not statistically significant (log rank test, p = 0.1758). Safety was similar and QOL indices were more favorable in the intermittent arm. The authors concluded that IADT was safe and feasible. Currently, several phase III trials that compare IADT with continuous ADT are ongoing, in the next few years, as the results from these trial accrue, (will be reported- this needs to be deleted) questions regarding time to progression and survival and the true efficacy of IADT should be resolved. The currently available data have established the safety of IADT. Notably, tumor ‘‘flare’’ was not identified after ADT was stopped. However, the trigger points for stopping and reinstituting ADT remain unclear.

192

S. Gupta et al.

In addition to the issues of efficacy, safety, and economics, IADT poses several unique questions. During the ON cycle, ADT leads to decreased cell proliferation and increased apoptosis. During the OFF cycle, the cancer cells are exposed to recovering androgens in most cases. The molecular pathways operational during this androgeninduced regrowth have not been fully illuminated and provide an important area for future research. The elucidation of these molecular mechanisms may also help settle the issue of when to start and stop the ON and OFF cycles. The androgen-deprived cancer cells at the onset of the OFF treatment cycle differ from ADT naı¨ve cancer cells, thereby allowing the possibility of designing interventions targeting the OFF Cycle to further enhance survival and slow progression to the lethal phenotype or castration-recurrent prostate cancer. An ongoing trial (NCT00283803) is evaluating the use of exisulind, a sulindac derivative, that inhibits cyclic guanosine monophosphate (cGMP) phosphodiesterase (PDE) and promotes apoptosis, to extend the duration of the OFF cycle of IADT. Another candidate for OFF cycle manipulation is the enzyme, steroid 5a-reductase (SRD5A), which has two established isoenzymes, SRD5A1 and SRD5A2. SRD5A catalyzes the conversion of testosterone to dihydrotestosterone (DHT), and an understanding of differential androgen function can suggest ways to improve IADT.

3 Differential Actions of Testosterone and Dihydrotestosterone Testosterone is the principal circulating androgen in men. Testosterone is synthesized by the Leydig cells in the testes and released into the systemic circulation (Snyder 2006). About 98% of the circulating testosterone is bound to steroid hormone-binding globulin (SHBG) or albumin. Testosterone is converted by the enzyme SRD5A to dihydrotestosterone (DHT) in many target tissues, which include the male reproductive organs and skin (Snyder 2006). There are two wellcharacterized synthetic inhibitors of SRD5A – finasteride, an inhibitor of SRD5A2; and dutasteride, which inhibits both SRD5A1 and SRD5A2 (Thomas et al. 2008). Testosterone can be converted by aromatase to estradiol at other sites such as liver, adipose tissue, central nervous system (Snyder 2006). Both testosterone and DHT are androgens and act via the androgen receptor (AR) (Liao et al. 1973; Wilson and French 1976; Grino et al. 1990). Using Chinese hamster ovary (CHO) cells that have low levels of SRD5A, Deslypere et al. (1992) showed that testosterone can induce an MMTV-CAT promoter to the same extent as DHT, albeit at a concentration that was 10-fold higher than the latter. DHT, relative to testosterone, exhibits a slower dissociation rate for the AR and achieves greater protection of AR against degradation (Grino et al. 1990; Zhou et al. 1995). It has been often proposed that testosterone and DHT differ in their potency, conversion of testosterone to DHT possibly being a metabolic amplification step (George 1997). However, in addition to being a substrate for the formation of DHT and estradiol in specific tissues, testosterone has several independent and unique functions. During embryogenesis, the development of male external genitalia, prostate, and seminal vesicles depends on the availability of DHT, while the development of the

Improving Intermittent Androgen-Deprivation Therapy

193

wolffian system depends on testosterone (Snyder 2006). In addition, androgendependent increase in muscle mass is affected by testosterone (Kumar et al. 1992; Snyder 2006). Similarly, DHT specific effects include the ability to turn down the production of Interleukin 4 (IL-4), Interleukin 5 (IL-5), and Interferon-g (IFN-g) from mouse T-cell hybridomas (Araneo et al. 1991), and the ability to induce the expression of its own biosynthetic enzyme – SRD5A (George et al. 1991). These observations make the case that the difference between testosterone and DHT goes beyond the notion of ligand preference, and has led many researchers to search for more areas of differential activity. Lin and Chang (1997) demonstrated that TDD5/N-myc downstream-regulated gene 1 (NDRG1) is suppressed differentially by testosterone and DHT. In the prostate, Avila et al. (1998) identified several genes that were modulated differently by castration (loss of testosterone and DHT) and finasteride treatment (loss of DHT). From the patterns of gene expression variations, the authors concluded that for some genes, testosterone and DHT exert fundamentally different effects. The differential effects of testosterone and DHT should be explored during prostate regrowth after castration. Under normal physiologic conditions in the adult, the presence of androgens maintains the architecture and turnover of the prostate but does not cause prostatic epithelial proliferation (Bruchovsky et al. 1975). After androgen deprivation, the prostate cells undergo massive apoptosis and prostate size and weight are reduced. Upon reintroduction of androgens, the regressed prostate cells undergo vigorous proliferation, and re-establish the precastration steady state (Bruchovsky et al. 1975). Clearly, the response of the regressed prostate cells to androgens is different from that of the cells in the intact prostate. A panel of over 20 androgen-response genes was identified using a polymerase chain reaction (PCR)-based complementary deoxyribonucleic acid (cDNA) subtraction assay based on upregulation during androgen-stimulated regrowth after castration in rat ventral prostate (Wang et al. 1997). Later, the difference between testosterone and DHT-induced gene expression was investigated in the rat prostate (Dadras et al. 2001). in intact animals, finasteride inhibited the conversion of testosterone to DHT and decreased the wet weight of the prostate (Fig 2). At the transcriptome level, the expression of only a few androgen- response genes was reduced slightly, a finding that was in agreement with previous reports (Rittmaster et al. 1991; Shao et al. 1993). Expectedly, the addition of finasteride decreased the wet weight of the regrowing prostate, which underlined the importance of DHT for the proliferation of prostate cells from the regressed state (Fig. 2). However, the patterns of gene expression during prostate regrowth were remarkable – several of the previously identified androgen-response genes showed higher expression with testosterone than with DHT during prostate regrowth after castration. This differential expression was in marked contrast to that seen in intact animals, where the addition of finasteride produced little effect on the gene expression profile. The most significant outcome was that many of the genes that were selectively enhanced by testosterone were markers of differentiation (prostatein C3) or

194

S. Gupta et al.

Fig. 2 Effects of hormonal manipulations on the prostate. Inhibition of SRD5A by finasteride or dutasteride leads to a modest decline in the size and weight of the prostate. Castration produces the most rapid and maximum decrease in the weight and size of the prostate. Androgen replacement with DHT (either directly with DHT or with testosterone, allowing for endogenous conversion to DHT by SRD5A) to the regressed prostate is strongly mitogenic; weight and size increase to precastrate levels. Replacement with testosterone (testosterone given with finasteride or dutasteride to prevent the endogenous synthesis of DHT) elicits a much weaker mitogenic response, and the enhancement in size/weight is retarded. During this phase of regrowth from a regressed state, selective exposure to testosterone preferentially enhances the expression of several androgen-response genes that have growth suppressive properties. This selective enhancement of growth suppressive/differentiation genes was not observed under any other conditions of hormonal manipulation

were growth inhibitory (calreticulin, adrenomedullin, human aci-reductone dioxygenase, ELL associated factor 2/upregulated 19 (EAF2/U19)) (Dadras et al. 2001). This phenomenon provides a strategy to preferentially enhance the expression of growth suppressive androgen-response genes, which may improve the efficacy of prostate cancer treatment.

4 SRD5A Inhibition During the OFF Cycle The differential regulation of androgen-response genes by testosterone and DHT during prostate regrowth provides an opportunity to improve IADT for advanced prostate cancer by OFF cycle inhibition of SRD5A. Huggins’ ‘‘biological syllogism’’ (Huggins et al. 1941) equated malignant epithelia to normal prostatic epithelium with

Improving Intermittent Androgen-Deprivation Therapy

195

respect to response to castration. During the regrowth of prostate cancer after a period of ADT, selective exposure to testosterone may enhance the upregulation of markers of differentiation and growth-suppressive genes. The clinical analog of this circumstance is met in IADT, where after a period of ADT (On Cycle), the treatment is discontinued and androgen levels are allowed to return to normal (OFF Cycle). During this OFF cycle, inhibition of SRD5A (by finasteride or dutasteride) should upregulate ‘‘good genes’’ – genes associated with differentiation and growth suppression, and drive the tumor phenotype toward a more differentiated one, and eventually improve survival (Fig. 1). Since the QOL benefits of the OFF Cycle are most often mediated by testosterone, QOL should improve also. Follow-up experiments have shown that addition of finasteride during the OFF cycle of IADT improved survival in vivo (Eggener et al. 2006). Nude mice with subcutaneously implanted androgen-sensitive LNCaP tumors were randomized to continuous ADT, IADT, continuous ADT + finasteride, and IADT + finasteride (finasteride was added during the OFF cycle). After one cycle of therapy, the IADT + finasteride group had significantly less tumor growth than the other treatment groups (Fig. 3). Mice treated with IADT + finasteride also had the longest survival (Fig. 4). The mice in the IADT + finasteride group were cycled similarly to the IADT group, i.e., the duration of ON and OFF cycles was similar relative to standard IADT. Given the upregulation of growth suppressive genes by testosterone during prostate regrowth, and the survival benefit of IADT + finasteride in a murine xenograft tumor model, the same genes may be upregulated during the OFF cycle in IADT upon inhibition of SRD5A. Indeed, SRD5A inhibition (by both finasteride and dutasteride) during the OFF cycle enhances the expression of EAF2/U19 (unpublished data). EAF2/U19 is an androgen-regulated tumor suppressor that frequently exhibits allelic loss and downregulation in high-grade prostate cancer (Xiao et al. 2003). Whether the selective upregulation of EAF2/U19 is indeed the effector of the better survival in the IADT + finasteride group, or the incidental side effect of other mechanisms, remains to be established. In addition, the transcriptome changes that occur during the OFF cycle of IADT, show temporal variation. Finasteride-induced differential gene expression changed during regrowth of the rat prostate (Dadras et al. 2001). In addition, the upregulation of EAF2/U19 in LNCaP tumors with OFF cycle finasteride/duatsteride (see earlier) was observed after 3 days of OFF Cycle intervention. Selective exposure to testosterone during the OFF cycle may lead to an initial enhanced expression of growth suppressive genes, and with time, the expression of these genes subsides to become similar to that in cases where there is no SRD5A inhibition during the OFF cycle. Therefore, the benefit of OFF cycle SRD5A inhibition will be present only while there is selective enhancement of tumor suppressive genes. After the gene expression returns to baseline, no advantages will accrue from continuation of the OFF cycle. In order to use SRD5A inhibitors during the IADT OFF Cycle, the issue of OFF Cycle duration must be addressed. Does OFF cycle duration correlate with prognosis? Does SRD5A inhibition affect the duration of the OFF cycle? Theoretically, the OFF cycle duration and QOL benefits should be correlated, since longer periods of high testosterone levels translate into better QOL function. However, no clear

196

S. Gupta et al.

Fig. 3 Effect of finasteride maintenance during the OFF cycle of intermittent ADT on the growth of androgen-sensitive LNCaP xenograft tumors. Nude mice with established LNCaP xenograft tumors were castrated and then randomized into following four different treatment groups: (1) continuous ADT (ADT); (2) continuous ADT + finasteride (ADT + F); (3) intermittent ADT (IADT); and (4) intermittent ADT with finasteride during the OFF cycle (IADT + F). More than 40% of the mice treated with finasteride during the OFF cycle of intermittent ADT (IADT + finasteride) showed no increase in tumor volume after one cycle of treatment (a). Tumor volume after 14 days of treatment increased the least in mice treated with IADT + F (b). Asterisk denotes statistical significance. From Eggener et al. (2006) (Reprinted with permission of John Wiley & Sons, Inc.)

correlation between progression to castration-recurrent prostate cancer and OFF cycle duration has been established. In a meta-analysis of 1446 patients, who underwent IADT, Shaw et al. (2007) reported that being OFF treatment for 2 or 3 years was associated with reduced overall mortality and a lower hazard of

Improving Intermittent Androgen-Deprivation Therapy

197

Fig. 4 Effect of finasteride maintenance during the OFF cycle of intermittent ADT on the survival of nude mice bearing androgen-sensitive LNCaP xenograft tumors. The experimental groups were the same as described in Fig. 3. (a) Kaplan–Meier survival curves for different treatments. Mice treated with IADT + F had the longest survival (p = 0.048). Mice treated with IADT + F were 3–5 times more likely alive 70 days after treatment initiation compared to other treatment regimens (b). From Eggener et al. (2006) (Reprinted with permission of John Wiley & Sons, Inc.)

castration-recurrent prostate cancer. Being able to stay at OFF treatment for 2 or 3 years may reflect less aggressive disease characteristics, and thus appear indicate delayed progression to castration-recurrent prostate cancer and improved overall survival. In a retrospective analysis of data from 101 patients treated with IADT with or without finasteride during the OFF cycle, Scholz et al. (2006) found that the addition of finasteride doubled the OFF cycle duration even after correcting for the expected 50% reduction in PSA levels due to finasteride. However, progression to castration-recurrent prostate cancer occurred with similar frequency in both treatment arms. The only variable that affected the progression to castration-recurrent prostate cancer in their analysis was clinical stage. While they did not perform any

198

S. Gupta et al.

QOL/adverse effect analyses, the authors surmised that increased time OFF treatment should lead to better QOL, and that OFF cycle finasteride may accomplish that. Even with the limitations of a retrospective study, these results suggest the potential for SRD5A inhibitors in prolongation of the OFF cycle duration. Currently, the use of dutasteride to prolong the OFF treatment period in IADT is being explored in two prospective phase II randomized trials (NCT00516815, NCT00553878). The next logical question would be whether there are any survival benefits or risks to extending the OFF cycle duration using SRD5A inhibition. No a priori assumptions should be made about the survival effects of OFF cycle prolongation by SRD5A inhibition because the effect of SRD5A inhibition on prostate cancer biology during IADT remains unclear. Serum PSA, which is used to guide decisions on timing in IADT, may have some problems. PSA is far from the ideal biomarker for prostate cancer detection (Bensalah et al. 2008), and performs less than perfectly as a predictor of disease progression and recurrence (Noguchi et al. 2000; Stamey 2001; Stamey et al. 2004; Chun et al. 2006; Mitchell et al. 2006; Steuber et al. 2006). Neoplastic cells produce less PSA than benign prostatic hyperplasia cells (Roehrborn et al. 1999a, b, 2000), and PSA expression may decrease with advancing Gleason grade (Aihara et al. 1994). In addition, the determination of PSA levels at the systemic level (in the serum) means that changes occurring in PSA-secreting tumor cells must be sustained and strong enough to be detected systemically. This introduces a selection bias and fails to address the issue of tumor heterogeneity. For instance, cells that do not produce a lot of PSA do not contribute substantially to the treatment decisions based on serum PSA levels, although they may be more aggressive than cells that produce more PSA. These limitations of serum PSA must be considered while evaluating the role of SRD5A inhibitors during the OFF cycle of IADT.

5 Challenges and Future Directions Prostate cancer can be targeted during the OFF cycle of IADT in several ways. Questions regarding safety and efficacy must be resolved before such an approach is practiced clinically. The current evidence allows several inferences and shows the path for exploring the answers to other issues. First, the issue of safety – finasteride and dutasteride are frequently used for indications other than prostate cancer, and are fairly safe drugs. Available data suggest that this safety profile is maintained during the OFF cycle. Questions regarding efficacy are trickier. Efficacy will depend on how the drugs are administered, how the response is measured, and what benchmarks are adopted for success. The final goal of any new intervention should be a survival advantage over previously established treatments, and the optimization of OFF cycle duration and the use of SRD5A inhibitors should be done with progression and survival results foremost in mind. Eggener et al. (2006) showed that OFF cycle finasteride can enhance survival in vivo when the OFF cycle duration remains similar to standard IADT; Scholz et al. (2006) showed that OFF cycle finasteride can double the OFF cycle duration

Improving Intermittent Androgen-Deprivation Therapy

199

without influencing progression to castration-recurrent prostate cancer. Evidently, there is a difference in outcome depending on how finasteride is administered – it seems that the survival benefit of OFF cycle finasteride in IADT is sustainable only when the OFF cycle durations remain similar in the IADT and IADT + finasteride groups. Finasteride is capable of increasing the OFF cycle duration; however, when the OFF cycle duration is increased, the survival advantage of IADT + finasteride over IADT is lost. This seems to suggest that one should not aim for an increase in the duration of the OFF cycle when using OFF cycle SRD5A inhibitors. The rationale for these observations may lie in the temporal expression of certain genes that are differentially regulated by testosterone and DHT. The molecular mechanisms that improve survival appear to have a different time course from those that are currently used to determine OFF cycle duration, i.e., PSA levels. Overall, OFF cycle SRD5A inhibitors may influence serum PSA differently than they do the androgen-response tumor suppressive genes at the cellular level in prostate cancer; by the time serum PSA levels increase to the threshold where androgen suppression is restarted, the enhancement in the expression of the tumor suppressive genes may have concluded. Therefore, the OFF cycle duration should be matched to the expression profile – the peaking and return to baseline – of growth suppressive genes like EAF2/U19. Of course, the need to follow gene expression profile in tumor cells necessitates repeated sampling of prostate cancer foci, which may not be possible in most patients. This obstacle may be overcome by tying the gene expression profiles to more readily measurable events and markers. One such candidate is serum testosterone. Since the preferential enhancement of tumor suppressive genes due to OFF cycle SRD5A inhibition is mediated by selective exposure to testosterone during the initial phase of regrowth, the OFF cycle durations could be tailored to testosterone levels. The time course of testosteroneinduced gene expression will help define the duration of the OFF cycle. When the enhanced expression of tumor suppressive genes subsides, further exposure to testosterone would be redundant. Therefore, testosterone levels need to be in the eugonadic range only for such period of time during which growth suppressive genes are elevated, and then androgen suppression could be reinitiated. The question of timing – when to start and stop the ON and OFF cycles – is crucial, and the answer will be critical to the success of SRD5A inhibition during the OFF treatment phase. OFF cycle SRD5A targeting is nascent yet exciting, and the near future may include SRD5A inhibition as an important component of prostate cancer IADT.

200

S. Gupta et al.

Acknowledgments We thank Minh Nguyen for critical reading, William H. Blair and Russell Gould for insightful discussion, and support from Department of Defense Prostate Cancer Research Program, DAMD 17-02-1-0113, National Institute of Health, R01 DK51193, and National Institute of Health Prostate Cancer Specialized Program of Research Excellence (SPORE), CA90386.

References Aihara, M., Lebovitz, R.M., Wheeler, T.M., Kinner, B.M., Ohori, M. and Scardino, P.T. 1994. Prostate specific antigen and Gleason grade: an immunohistochemical study of prostate cancer. J Urol, 151, 1558–64. Akakura, K., Bruchovsky, N., Goldenberg, S.L., Rennie, P.S., Buckley, A.R. and Sullivan, L.D. 1993. Effects of intermittent androgen suppression on androgen-dependent tumors. Apoptosis and serum prostate-specific antigen. Cancer, 71, 2782–90. Albrecht, W., Collette, L., Fava, C., Kariakine, O.B., Whelan, P., Studer, U.E., De Reijke, T.M., Kil, P.J. and Rea, L.A. 2003. Intermittent maximal androgen blockade in patients with metastatic prostate cancer: an EORTC feasibility study. Eur Urol, 44, 505–11. American Cancer Society. Cancer Facts and Figures 2007. Atlanta: American Cancer Society; 2007. Araneo, B.A., Dowell, T., Diegel, M. and Daynes, R.A. 1991. Dihydrotestosterone exerts a depressive influence on the production of interleukin-4 (IL-4), IL-5, and gamma-interferon, but not IL-2 by activated murine T cells. Blood, 78, 688–99. Avila, D.M., Fuqua, S.A., George, F.W. and McPhaul, M.J. 1998. Identification of genes expressed in the rat prostate that are modulated differently by castration and Finasteride treatment. J Endocrinol, 159, 403–11. Bensalah, K., Lotan, Y., Karam, J.A. and Shariat, S.F. 2008. New circulating biomarkers for prostate cancer. Prostate Cancer Prostatic Dis, 11(2), 112–20. Boccon-Gibod, L., Hammerer, P., Madersbacher, S., Mottet, N., Prayer-Galetti, T. and Tunn, U. 2007. The role of intermittent androgen deprivation in prostate cancer. BJU Int, 100, 738–43. Bouchot, O., Lenormand, L., Karam, G., Prunet, D., Gaschignard, N., Malinovsky, J.M. and Buzelin, J.M. 2000. Intermittent androgen suppression in the treatment of metastatic prostate cancer. Eur Urol, 38, 543–9. Braga-Basaria, M., Dobs, A.S., Muller, D.C., Carducci, M.A., John, M., Egan, J. and Basaria, S. 2006. Metabolic syndrome in men with prostate cancer undergoing long-term androgendeprivation therapy. J Clin Oncol, 24, 3979–83. Bruchovsky, N., Lesser, B., Van Doorn, E. and Craven, S. 1975. Hormonal effects on cell proliferation in rat prostate. Vitam Horm, 33, 61–102. Bruchovsky, N., Rennie, P.S., Coldman, A.J., Goldenberg, S.L., To, M. and Lawson, D. 1990. Effects of androgen withdrawal on the stem cell composition of the Shionogi carcinoma. Cancer Res, 50, 2275–82. Bruchovsky, N., Klotz, L., Crook, J., Malone, S., Ludgate, C., Morris, W.J., Gleave, M.E. and Goldenberg, S.L. 2006. Final results of the Canadian prospective phase II trial of intermittent androgen suppression for men in biochemical recurrence after radiotherapy for locally advanced prostate cancer: clinical parameters. Cancer, 107, 389–95. Bruchovsky, N., Klotz, L., Crook, J. and Goldenberg, S.L. 2007. Locally advanced prostate cancer – biochemical results from a prospective phase II study of intermittent androgen suppression for men with evidence of prostate-specific antigen recurrence after radiotherapy. Cancer, 109, 858–67.

Improving Intermittent Androgen-Deprivation Therapy

201

Chun, F.K., Briganti, A., Lebeau, T., Fradet, V., Steuber, T., Walz, J., Schlomm, T., Eichelberg, C., Haese, A., Erbersdobler, A., McCormack, M., Perrotte, P., Graefen, M., Huland, H. and Karakiewicz, P.I. 2006. The 2002 AJCC pT2 substages confer no prognostic information on the rate of biochemical recurrence after radical prostatectomy. Eur Urol, 49, 273–8; discussion 278–9. Conti, P.D., Atallah, A.N., Arruda, H., Soares, B.G., El Dib, R.P. and Wilt, T.J. 2007. Intermittent versus continuous androgen suppression for prostatic cancer. Cochrane Database Syst Rev, (4), CD005009. Crook, M., Szumacher, E., Malone, S., Huan, S. and Segal, R. 1999. Intermittent androgen suppression in the management of prostate cancer. Urology 53(3), 530–4. Dadras, S.S., Cai, X., Abasolo, I. and Wang, Z. 2001. Inhibition of 5alpha-reductase in rat prostate reveals differential regulation of androgen-response gene expression by testosterone and dihydrotestosterone. Gene Expr, 9, 183–94. Deslypere, J.P., Young, M., Wilson, J.D. and McPhaul, M.J. 1992. Testosterone and 5alphadihydrotestosterone interact differently with the androgen receptor to enhance transcription of the MMTV-CAT reporter gene. Mol Cell Endocrinol, 88, 15–22. Eggener, S.E., Stern, J.A., Jain, P.M., Oram, S., Ai, J., Cai, X., Roehl, K.A. and Wang, Z. 2006. Enhancement of intermittent androgen ablation by ‘‘off-cycle’’ maintenance with finasteride in LNCaP prostate cancer xenograft model. Prostate, 66, 495–502. George, F. 1997. Androgen metabolism in the prostate of the finasteride-treated, adult rat: a possible explanation for the differential action of testosterone and 5alpha-dihydrotestosterone during development of the male urogenital tract. Endocrinology, 138, 871–7. George, F.W., Russell, D.W. and Wilson, J.D. 1991. Feed-forward control of prostate growth: dihydrotestosterone induces expression of its own biosynthetic enzyme, steroid 5alpha-reductase. Proc Natl Acad Sci U S A, 88, 8044–7. Goldenberg, S.L., Bruchovsky, N., Gleave, M.E., Sullivan, L.D. and Akakura, K. 1995. Intermittent androgen suppression in the treatment of prostate cancer: a preliminary report. Urology, 45, 839–44; discussion 844–5. Grino, P.B., Griffin, J.E. and Wilson, J.D. 1990. Testosterone at high concentrations interacts with the human androgen receptor similarly to dihydrotestosterone. Endocrinology, 126, 1165–72. Higano, C.S., Ellis, W., Russell, K. and Lange, P.H. 1996. Intermittent androgen suppression with leuprolide and flutamide for prostate cancer: a pilot study. Urology, 48, 800–4. Higano, C., Shields, A., Wood, N., Brown, J. and Tangen, C. 2004. Bone mineral density in patients with prostate cancer without bone metastases treated with intermittent androgen suppression. Urology, 64, 1182–6. Huggins, C. and Hodges, C. V. 1941. Studies on prostatic cancer. I: the effect of castration, of estrogen and androgen injection on serum phosphatases in metastatic carcinoma of the prostate. Cancer Res 1, 293–7. Huggins, C., Stevens, R.E. and Hodges, C.V. 1941. Studies on prostatic cancer: II. The effects of castration on advanced carcinoma of the prostate gland. Arch Surg 43, 209–23. Klein, E.A., Platz, E.A. and Thompson, I.M. 2007. Epidemiology, etiology, and prevention of prostate cancer. In Campbell-Walsh Urology –9th ed, ed. A. J. Wein, pp. 2854–73. Philadelphia: Saunders Elsevier. Klotz, L.H., Herr, H.W., Morse, M.J. and Whitmore, W.F., Jr. 1986. Intermittent endocrine therapy for advanced prostate cancer. Cancer, 58, 2546–50. Krupski, T.L., Saigal, C.S., Hanley, J., Schonlau, M. and Litwin, M.S. 2006. Patterns of care for men with prostate cancer after failure of primary treatment. Cancer, 107, 258–65. Krupski, T.L., Foley, K.A., Baser, O., Long, S., Macarios, D. and Litwin, M.S. 2007. Health care cost associated with prostate cancer, androgen deprivation therapy and bone complications. J Urol. 178(4 Pt 1), 1423–8. Kumar, N., Didolkar, A.K., Monder, C., Bardin, C.W. and Sundaram, K. 1992. The biological activity of 7alpha-methyl-19-nortestosterone is not amplified in male reproductive tract as is that of testosterone. Endocrinology, 130, 3677–83.

202

S. Gupta et al.

Liao, S., Liang, T., Fang, S., Castaneda, E. and Shao, T.C. 1973. Steroid structure and androgenic activity. Specificities involved in the receptor binding and nuclear retention of various androgens. J Biol Chem, 248, 6154–62. Lin, T.M. and Chang, C. 1997. Cloning and characterization of TDD5, an androgen target gene that is differentially repressed by testosterone and dihydrotestosterone. Proc Natl Acad Sci U S A, 94, 4988–93. Malone, S., Perry, G., Segal, R., Dahrouge, S. and Crook, J. 2005. Long-term side-effects of intermittent androgen suppression therapy in prostate cancer: results of a phase II study. BJU Int, 96, 514–20. Max, W., Rice, D.P., Sung, H.Y., Michel, M., Breuer, W. and Zhang, X. 2002. The economic burden of prostate cancer, California, 1998. Cancer, 94, 2906–13. Miller, K., Steiner, U., Lingnau, A., Keilholz, U., Witzsch, U., Haider, A., Wachter, U., Russel, C. and Altweir, J. 2007. Randomised prospective study of intermittent versus continuous androgen suppression in advanced prostate cancer. J Clin Oncol, ASCO Annual Meeting Proceedings, #5015, p 238s, 2007. Mitchell, R.E., Desai, M., Shah, J.B., Olsson, C.A., Benson, M.C. and McKiernan, J.M. 2006. Preoperative serum prostate specific antigen remains a significant prognostic variable in predicting biochemical failure after radical prostatectomy. J Urol, 175, 1663–7; discussion 1667. Mottet, N., Lucas, C., Sene, E., Avances, C., Maubach, L. and Wolff, J.M. 2005. Intermittent androgen castration: a biological reality during intermittent treatment in metastatic prostate cancer? Urol Int, 75, 204–8. Newcomer, L.M., Stanford, J.L., Blumenstein, B.A. and Brawer, M.K. 1997. Temporal trends in rates of prostate cancer: declining incidence of advanced stage disease, 1974 to 1994. J Urol, 158, 1427–30. Noguchi, M., Stamey, T.A., McNeal, J.E. and Yemoto, C.M. 2000. Preoperative serum prostate specific antigen does not reflect biochemical failure rates after radical prostatectomy in men with large volume cancers. J Urol, 164, 1596–600. Oliver, R.T., Williams, G., Paris, A.M. and Blandy, J.P. 1997. Intermittent androgen deprivation after PSA-complete response as a strategy to reduce induction of hormone-resistant prostate cancer. Urology, 49, 79–82. Penson, D.F. and Chan, J.M. 2007. Prostate cancer. J Urol, 177, 2020–9. Peyromaure, M., Delongchamps, N.B., Debre, B. and Zerbib, M. 2005. Intermittent androgen deprivation for biologic recurrence after radical prostatectomy: long-term experience. Urology, 65, 724–9. Rennie, P.S., Bruchovsky, N. and Coldman, A.J. 1990. Loss of androgen dependence is associated with an increase in tumorigenic stem cells and resistance to cell-death genes. J Steroid Biochem Mol Biol, 37, 843–7. Rittmaster, R.S., Magor, K.E., Manning, A.P., Norman, R.W. and Lazier, C.B. 1991. Differential effect of 5alpha-reductase inhibition and castration on androgen-regulated gene expression in rat prostate. Mol Endocrinol, 5(7), 1023–9. Roehrborn, C.G., Boyle, P., Gould, A.L. and Waldstreicher, J. 1999a. Serum prostate-specific antigen as a predictor of prostate volume in men with benign prostatic hyperplasia. Urology, 53, 581–9. Roehrborn, C.G., McConnell, J.D., Lieber, M., Kaplan, S., Geller, J., Malek, G.H., Castellanos, R., Coffield, S., Saltzman, B., Resnick, M., Cook, T.J. and Waldstreicher, J. 1999b. Serum prostate-specific antigen concentration is a powerful predictor of acute urinary retention and need for surgery in men with clinical benign prostatic hyperplasia. PLESS Study Group. Urology, 53, 473–80. Roehrborn, C.G., McConnell, J., Bonilla, J., Rosenblatt, S., Hudson, P.B., Malek, G.H., Schellhammer, P.F., Bruskewitz, R., Matsumoto, A.M., Harrison, L.H., Fuselier, H.A., Walsh, P., Roy, J., Andriole, G., Resnick, M. and Waldstreicher, J. 2000. Serum prostate specific antigen

Improving Intermittent Androgen-Deprivation Therapy

203

is a strong predictor of future prostate growth in men with benign prostatic hyperplasia. PROSCAR long-term efficacy and safety study. J Urol, 163, 13–20. Sato, N., Gleave, M.E., Bruchovsky, N., Rennie, P.S., Goldenberg, S.L., Lange, P.H. and Sullivan, L.D. 1996. Intermittent androgen suppression delays progression to androgen-independent regulation of prostate-specific antigen gene in the LNCaP prostate tumour model. J Steroid Biochem Mol Biol, 58, 139–46. Sato, N., Akakura, K., Isaka, S., Nakatsu, H., Tanaka, M., Ito, H. and Masai, M. 2004. Intermittent androgen suppression for locally advanced and metastatic prostate cancer: preliminary report of a prospective multicenter study. Urology, 64, 341–5. Scholz, M.C., Jennrich, R.I., Strum, S.B., Johnson, H.J., Guess, B.W. and Lam, R.Y. 2006. Intermittent use of testosterone inactivating pharmaceuticals using finasteride prolongs the time off period. J Urol, 175, 1673–8. Shahinian, V.B., Kuo, Y.F., Freeman, J.L., Orihuela, E. and Goodwin, J.S. 2005. Increasing use of gonadotropin-releasing hormone agonists for the treatment of localized prostate carcinoma. Cancer, 103, 1615–24. Shao, T.C., Kong, A., Marafelia, P. and Cunningham, G.R. 1993. Effects of finasteride on the rat ventral prostate. J Androl, 14, 79–86. Shaw, G.L., Wilson, P., Cuzick, J., Prowse, D.M., Goldenberg, S.L., Spry, N.A. and Oliver, T. 2007. International study into the use of intermittent hormone therapy in the treatment of carcinoma of the prostate: a meta-analysis of 1446 patients. BJU Int, 99, 1056–65. Snyder, P.J. 2006. Androgens. In Goodman Gilman’s the Pharmacological Basis of Therapeutics – 11th ed, ed. L. L. Brunton, pp. 1573–85. New York: The McGraw-Hill. Spry, N.A., Kristjanson, L., Hooton, B., Hayden, L., Neerhut, G., Gurney, H., Corica, T., Korbel, E., Weinstein, S. and McCaul, K. 2006. Adverse effects to quality of life arising from treatment can recover with intermittent androgen suppression in men with prostate cancer. Eur J Cancer, 42, 1083–92. Stamey, T.A. 2001. Preoperative serum prostate-specific antigen (PSA) below 10 microg/l predicts neither the presence of prostate cancer nor the rate of postoperative PSA failure. Clin Chem, 47, 631–4. Stamey, T.A., Caldwell, M., McNeal, J.E., Nolley, R., Hemenez, M. and Downs, J. 2004. The prostate specific antigen era in the United States is over for prostate cancer: what happened in the last 20 years? J Urol, 172, 1297–301. Steuber, T., Erbersdobler, A., Graefen, M., Haese, A., Huland, H. and Karakiewicz, P.I. 2006. Comparative assessment of the 1992 and 2002 pathologic T3 substages for the prediction of biochemical recurrence after radical prostatectomy. Cancer, 106, 775–82. Strum, S.B., Scholz, M.C. and McDermed, J.E. 2000. Intermittent androgen deprivation in prostate cancer patients: factors predictive of prolonged time off therapy. Oncologist, 5, 45–52. Thomas, L.N., Douglas, R.C., Lazier, C.B., Too, C.K., Rittmaster, R.S. and Tindall, D.J. 2008. Type 1 and type 2 5alpha-reductase expression in the development and progression of prostate cancer. Eur Urol, 53, 244–52. Thompson, I.M. and Ankerst, D.P. 2007. Prostate-specific antigen in the early detection of prostate cancer. CMAJ, 176, 1853–8. Tsai, H.K., D’Amico, A.V., Sadetsky, N., Chen, M.H. and Carroll, P.R. 2007. Androgen deprivation therapy for localized prostate cancer and the risk of cardiovascular mortality. J Natl Cancer Inst, 99, 1516–24. Wang, Z., Tufts, R., Haleem, R. and Cai, X. 1997. Genes regulated by androgen in the rat ventral prostate. Proc Natl Acad Sci U S A, 94, 12999–3004. Wilson, E.M. and French, F.S. 1976. Binding properties of androgen receptors. Evidence for identical receptors in rat testis, epididymis, and prostate. J Biol Chem, 251, 5620–9. Wilson, L.S., Tesoro, R., Elkin, E.P., Sadetsky, N., Broering, J.M., Latini, D.M., DuChane, J., Mody, R.R. and Carroll, P.R. 2007. Cumulative cost pattern comparison of prostate cancer treatments. Cancer 109(3), 518–27.

204

S. Gupta et al.

Xiao, W., Zhang, Q., Jiang, F., Pins, M., Kozlowski, J.M. and Wang, Z. 2003. Suppression of prostate tumor growth by U19, a novel testosterone-regulated apoptosis inducer. Cancer Res, 63, 4698–704. Zhou, Z., Lane, M., Kemppainen, J., French, F. and Wilson, E. 1995. Specificity of liganddependent androgen receptor stabilization: receptor domain interactions influence ligand dissociation and receptor stability. Mol Endocrinol, 9, 208–18.

Insights from AR Gene Mutations Grant Buchanan, Eleanor F. Need, Tina Bianco-Miotto, Norman M. Greenberg, Howard I. Scher, Margaret M. Centenera, Lisa M. Butler, Diane M. Robins, and Wayne D. Tilley

Abstrat The accumulation of somatic mutations on the background of natural germline variation is one of the fundamental mechanisms underpinning both disease and the development and progression of perhaps all tumors. For example, germline inactivating mutations in one allele of a tumor suppressor gene (TSG) (e. g., p53 or BRCA1) predisposes an individual to a lifetime of increased risk of cancer in particular tissues. In those cases, tumors commonly arise after somatic inactivation of the second TSG allele, and are often characterized by a younger age of onset and a more aggressive phenotype than tumors arising in the same tissues without a common or dominant genetic predisposition. A small proportion of prostate cancers do indeed occur in a hereditary manner, but it has been difficult to attribute inheritance and risk to a single gene. Even for the monoallelic androgen receptor (AR) gene on the X chromosome, which exhibits variously traits of an atrisk allele, oncogene, and mediator of prostate cancer progression and therapy resistance, inactivating germline variation in the AR is essentially nonexistent in this disease. Instead, germline AR variants often underpin the relatively common inherited syndrome of androgen insensitivity (AIS). In contrast, however, somatic variation of the AR is potentially a frequent event during tumor progression. As a consequence of these complexities, unraveling the precise role of the AR at each stage of prostate cancer progression, and indeed in different prostatic compartments or populations of tumor cells is challenging. In this chapter, we detail how the identification and characterization of somatic AR variants arising in prostate cancer has provided crucial information on (1) the role of the receptor throughout disease etiology and the emergence of castrate-recurrent disease, (2) the

W.D. Tilley(*) Dame Roma Mitchell Cancer Research Laboratory, School of Medicine, University of Adelaide/ Hanson Institute, PO Box 14, Rundle Mall, SA 5000, Australia, E-mail: [email protected]

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4_10, # Springer Science + Business Media, LLC 2009

207

208

G. Buchanan et al.

fine functional subdomain structure of the AR, (3) the oncogenic potential of aberrant AR signaling, and (4) new approaches to targeting AR function in disease management.

1 AR and Prostate Cancer A functional androgen signaling axis is essential for the development, growth, and homeostasis of the normal prostate gland, and has been linked to all stages of prostate tumorigenesis including predisposition and progression (Culig et al. 2000; Feldman and Feldman 2001; Scher et al. 2004). The critical mediator of these effects is AR, a member of the superfamily of nuclear receptor transcription factors that regulates diverse cellular functions in the prostate. The requirement for sustained androgen signaling in prostate cancer underpins hormonal strategies [e.g., androgen-deprivation therapy (ADT) and/or AR antagonists] in disease treatment (Fig. 1, Androgen-deprivation therapy). While virtually all prostate cancers initially respond to ADT, the majority relapse and the disease progresses. Whereas initial studies using the Dunning animal model and with the human prostate cancer cell lines, DU145 and PC-3, suggested that loss of AR gene expression could be a mechanism for failure of ADT (Quarmby et al. 1990; Tilley et al. 1990), subsequent immunohistochemical studies of clinical prostate cancer have demonstrated that the AR is expressed in essentially all metastatic tumors, including those that continue to grow during ADT (Culig et al. 1998). Additionally, tissue levels of prostate-specific antigen (PSA) and other androgen responsive genes increase during castrationrecurrent tumor regrowth (Bentel and Tilley 1996; Culig et al. 1998; Hobisch et al. 1995). The failure of ADT strategies can be explained in many cases by selection for continued androgen and/or AR signaling in the castrate environment (Scher et al. 2004). The contributing mechanisms include increased AR levels, crosstalk between the AR and other signaling pathways, increased response to or level of AR coregulators, and the selection of permissive AR mutations (Buchanan et al. 2001b; Feldman and Feldman 2001; Scher et al. 2004).

2 Mutations in the AR Gene in Prostate Cancer Approximately 130 individual somatic missense or nonsense mutations have been reported in the AR gene in human prostate cancer, several more in latent disease at autopsy, and a significant number in cellular and animal models of the disease (Gottlieb et al. 2004; Scher et al. 2004). Whereas AR gene mutations have been found only rarely in early-stage or primary disease, the reported incidence increases with the stage of disease analyzed, and particularly for those tumors regrowing in a castrate setting (Buchanan et al. 2001b). Moreover, almost overwhelming evidence suggests that the majority of these mutations do not result in a loss of function, but

Insights from AR Gene Mutations

209

Fig. 1 AR signaling and points of therapeutic intervention in prostate cancer. A schematic of AR signaling in a prostate cancer cell highlights known and predicted effects of somatic AR variants detected in clinical disease and animal models (green boxes), and new and traditional approaches to therapeutic intervention (red boxes). After synthesis in a normal cell, the AR is maintained in equilibrium between immature and active states through association/dissociation with a complex that includes heat-shock proteins, p23 and a tetratricopeptide (TPR)-containing protein. The active state is available for binding high-affinity androgenic ligands, such as 5a-dihydrotestosterone (DHT). Ligand binding results in the dissociation of heat-shock proteins, dimerization and phosphorylation, nuclear transport, DNA binding, recruitment of the transcription machinery and cofactors, and activation of downstream gene pathways. AR activity also can be stimulated in the presence and absence of ligand by membrane-bound tyrosine kinase receptors, such as HER2/neu, and by signaling molecules, growth factors, and cytokines. Mutation effects: Promiscuous activation, broadened specificity or sensitivity to noncanonical ligands (e.g., V730M, H874Y, T877A [LNCaP]); Decreased control, altered equilibrium interaction with the chaperone complex leading to multiple modes of activation (e.g., altered interaction with the TPR proteins aSGT of FKBP52); Stimulation, altered phosphorylation status (e.g., creation, disruption, or hyperstimulation of a phosphorylation site (e.g., S515G)); Sensitization, altered physical stability or receptor turnover (e.g., polyQ2L or C619Y); Constitutive bypass, truncated AR variants circumvent the requirement for ligand and are constitutively active (e.g., Q640STOP); others exhibit leaky or increased basal activity (e.g., T877A, E236G); activation of the wild-type receptor in the absence of ligand by growth factor, cytokine or phosphorylation kinase signaling can mimic constitutive bypass mutations; Differential cofactor recruitment, altered ability to recruit coregulator molecules (e.g., I673T, E236G), Differential gene activation, altered capacity to activate downstream genes (e.g., E236G). Therapeutic intervention: Androgen-deprivation therapy,

210

G. Buchanan et al.

can be ascribed one of a range of phenotypes capable of maintaining AR signaling during ADT (Fig. 1). These observations support the theory of ‘‘Therapy-mediated selection pressure’’ for prostate cancer (Scher et al. 2004), which holds that the specific treatment strategy employed against the disease will select for tumor cells containing changes permissive for continued survival and growth during that treatment and/or in the castrate environment. We discuss later how the study of AR gene mutations from prostate cancer samples has provided key insights not only into the role of the receptor in this disease, but also into the function of each of its domains and how we might better target the AR signaling pathway for disease management.

3 The AR-Ligand-Binding Domain The AR is a member of a small number of steroid-activated nuclear transcription factors defined by conserved domains for DNA (DBD) and steroid-ligand binding (LBD) that are separated by a divergent hinge region, and a poorly conserved and highly variable amino-terminal domain (NTD) thought to direct transcriptional response and competency (Fig. 2a). Much of the initial understanding of AR function stems from studies of AIS, in which genetic XY individuals exhibit a phenotype ranging from ambiguous external genitalia and/or impaired masculinization to complete feminization (Quigley et al. 1995). Commonly, germline mutations in the AR-LBD that result in a partial or complete loss of steroid-binding capacity and an inactive receptor are the underlying cause of this syndrome. For this reason, and perhaps additionally for technical or logistical limitations, the vast majority of studies for AR mutations in prostate cancer have been performed only on the LBD, which comprises barely a quarter of the receptor’s coding sequence. Nonetheless, the intensive analysis of this relatively small domain has been particularly instructive in the dissection of AR functional subdomain structure, protein– protein interactions, and the molecular mechanisms of AR activation. In particular, more than 85% of missense mutations detected in clinical prostate tumors were Fig. 1 (Continued) traditional systemic ADT; Antagonists & SARMS, competitively inhibit ligand binding and slow movement of the receptor into and through the nucleus, Enzymatic inhibitors, e.g., 5a-reductase inhibitors such as finasteride or dutasteride prevent conversion of testosterone to the more potent DHT; Antisense, shRNA, and HDACIs, antisense AR oligonucleotides, short hairpin RNAs directed at the AR and histone deacetylase inhibitors reduce AR levels; Ansamycins and HDACIs, Ansamycin antibiotics (e.g., 17-allylamino-geldanamycin) and HDACIs act to disrupt Hsp90 activity or chaperone functioning; Ansamycins, act to inhibit chaperone-mediated trafficking; ARIs, AR inhibitors or dominant negative AR constructs act to prevent receptor dimerization and/or recruitment of coregulators; Response blockers, specific response-elementbinding proteins (e.g., polyamides, decoy response elements) block or compete with AR interactions on DAN; Antibodies, specific for AR to block activity directly or for growth factor receptors (e.g., Herceptin) to prevent crosstalk; Inhibitors, directed at MAPK, JAK-STAT, or Akt pathways to prevent receptor crosstalk; Recruitment inhibitors, act to prevent coactivator recruitment to the receptor (See Color Insert)

Insights from AR Gene Mutations

211

Fig. 2 Collocation of AR gene mutations. Schematic of the human AR (top) delineating aminoterminal (NTD), DNA-binding (DBD) and ligand-binding (LBD) domains. Green shaded areas represent regions where mutations collocate in clinical prostate cancer (see later and Fig. 2). Collocation of mutations in the LBD is shown in greater detail; the position and frequency of missense mutations identified in the inherited from of androgen insensitivity (AIS; upper) and human prostate cancer (hPCa; lower) were used to calculate a proportional frequency distribution of mutations. Vertical bars represent the percentage of mutations identified in the AR-LBD in either hPCa or AIS within a nine-amino acid region centered on individual codons. A cut-off of 4% was applied such that bars represent at least two mutations from prostate cancer within each nineamino-acid region. Areas of collocation in AIS (blue boxes) and hPCa (green boxes) were determined by delineating the actual position of mutations occurring in regions defined by the frequency distribution and required a minimum of four independent mutations. The amino acid residues encompassing these regions are indicated (See Color Insert)

located in five discrete regions of the LBD that collectively comprised less than 10% of the AR coding sequence (Fig. 2b) (Buchanan et al. 2001a; Scher et al. 2004). Of critical importance was the concomitant finding that inactivating mutations in the AR identified in AIS collocated to almost completely distinct regions (Fig. 2b) (Buchanan et al. 2001a). Each of the LBD regions where mutations collocate in prostate cancer has been ascribed to modulation of ligand-binding specificity, cofactor responses, and transactivation capacity of the receptor, thereby facilitating increased AR function and a survival advantage for prostate cancer

212

G. Buchanan et al.

cells. The insights gained from these studies have led to new directions in both defining AR function and investigating better ways to modulate its behavior.

3.1

Collocation of AR Gene Mutations at 874–911

Agonist binding by the AR results in conformational changes in the LBD and the formation of activation function 2 (AF-2), which is a conserved protein–protein binding site on the surface of the receptor formed by the juxtaposition of several distinct structural elements along the length of the LBD (Matias et al. 2000; Sack et al. 2001).The agonist-induced AF-2 surface is responsible for mediating interaction with both the NTD (the so-called amino-carboxyl [i.e., N/C] interaction) and with a set of essential transcriptional coregulators that include the three p160 family members, NCOA1/SRC-1/RIP160, NCOA2/GRIP1/TIF2/SRC-2, and NCOA3/ AIB1/TRAM-1/ACTR/RAC-3/SRC-3. The N/C interaction is thought essential for AR function; its abolition results in a receptor incapable of activating chromatin-integrated targets in vivo (Li et al. 2006). Mutations between residues 874 and 911 identified in prostate cancer lie proximal to the core sequence of AF-2 (amino acids 893–899) and typically result in a receptor that can be inappropriately activated by a range of nonclassical ligands and/or receptor antagonists (Fig. 1, Promiscuous activation). None of these mutations are located within the AF-2 core where missense mutations abolish N/C and/or p160 interactions and result in an inactive receptor (Berrevoets et al. 1998). The prototypical example of a permissive mutation is the AR-T877A variant identified in clinical prostate tumors and in the human prostate cancer cell line LNCaP, which exhibits increased basal transactivation activity and response to progesterone, 17b-estradiol, adrenal androgens, and hydroxyflutamide compared to wild-type AR (wtAR) (Veldscholte et al. 1992). The T877A AR variant inappropriately adopts an AF-2 surface permissive for N/C and p160 coactivator interaction for each of these noncanonical ligands (Kemppainen et al. 1999). Structural analysis has revealed that the T877A mutation increases the volume of the ligand-binding pocket, thereby facilitating accommodation of bulkier ligands and inappropriate formation of a correct AF-2 surface (Han et al. 2001). The selection for AR variants able to utilize endogenous noncanonical ligands somewhat negates the potential for disease treatment by enzymatic inhibitors of androgen biosynthesis such as 5a-reductase (Fig. 1, Enzymatic inhibitors). For other mutations in the vicinity of AF-2, detailed functional and crystal structure analysis has revealed a distinct mechanism of promiscuous activity (Askew et al. 2007; Bohl et al. 2007; He et al. 2006). These mutations slow dissociation of bound ligand, and increase receptor stabilization and recruitment of p160 or other coactivators (Fig. 1, Differential cofactor recruitment). The study of those mutations revealed that AR transcriptional potency depends on the interchange between structural states promoted by ligand and coactivator interactions (He et al. 2006).

Insights from AR Gene Mutations

3.2

213

Collocation of AR Gene Mutations at 715–730

One of the most characterized group of prostate-cancer-derived AR-LBD mutations lie between residues 715 and 730, and span the highly conserved signature sequence region of the steroid receptors that is responsible for ligand recognition and interaction, and contributes to the AF-2 surface (Wurtz et al. 1996). Critically, none of the residues mutated in prostate cancer corresponds to those known to be essential for direct contact with the ligand (Matias et al. 2000). Instead, these mutations result in an inappropriate AR response to estrogens, progestins, adrenal androgens, and other nonclassical ligands (Buchanan et al. 2001a, b). In AIS, the loss of ligand binding commonly is associated with mutations that collocate to codons 702–714 and make direct contact with the ligand (Gottlieb et al. 2004).

3.3

Collocation of AR Gene Mutations at 670–678

Mutations in amino acids 670–678 at the boundary of hinge and LBD, which have been found in both clinical prostate cancer and in the transgenic adenocarcinoma of the mouse prostate (TRAMP) model, exhibit a two- to fourfold greater transactivation activity in response to 5-a-dihydrotestosterone (DHT), nonclassical ligands, and hydroxyflutamide compared to wtAR (Buchanan et al. 2000). Unlike mutations in the signature sequence or AF-2, enhanced activity of the 670–678 variants occurred in the absence of changes to ligand-binding kinetics, receptor levels, or DNA-binding capacity. These findings argued that the 670–678 AR variants altered receptor interaction with a coactivator or corepressor protein; the hypothesis is supported by the identification of a putative coregulator-binding site using molecular modeling (Buchanan et al. 2000). Recent crystal studies provide direct evidence that the 670–678 region is contained within a coactivator interaction site distinct from AF-2, but where binding of proteins and other agents weakens the structural integrity of AF-2 and receptor interactions with p160 coactivators (Estebanez-Perpina et al. 2007). Attenuation of this allosteric effect by these prostate cancer mutations should accentuate coregulator function at AF-2, which provides a mechanism for increased function relevant to prostate cancer, particularly in response to increasing levels of p160s or other transcriptional coregulators that occur during progression (Chmelar et al. 2007). These findings have precipitated the search for coregulators that interact with the 670–678 region rather than AF-2 (see Sect. 4.6) (Buchanan et al. 2007), and for compounds that bind to this site and disrupt cofactor interactions here and at AF-2 (Fig. 1, Recruitment inhibitors) (Estebanez-Perpina et al. 2007).

3.4

Collocation of AR Gene Mutations at 741–757 and 791–798

Whereas much less is known about the two regions consisting of residues 741–757 and 791–798, the former is known from AIS to affect the N/C interaction and

214

G. Buchanan et al.

recruitment of transcriptional coactivators to AF-2. The consequences of mutations in these regions remains poorly understood for prostate cancer.

3.5

SARMS and the Syndrome of Antiandrogen Withdrawal

The latent promiscuous potential of the AR-LBD has prompted the search for better antagonists and/or selective AR modulators (Fig. 1, Antagonists & SARMS); the latter being compounds that exert agonist and/or antagonist effects on the AR in a tissue or target gene-specific manner. However, all clinically utilized AR antagonists (e.g., flutamide, nilutamide, and bicalutamide), as well as other hormonal therapies for prostate cancer (e.g., estrogens such as diethylstilbestrol, and progestational agents such as megestrol acetate) have been associated with the syndrome of steroid-hormone and anti-antiandrogen withdrawal (Kelly et al. 1997). This syndrome is characterized by tumor regression and decreasing serum levels of PSA when treatment with antiandrogens is discontinued at the time of clinical progression (Kelly et al. 1997). In many cases of antiandrogen withdrawal, tumors have been found to harbor somatic AR gene mutations permissive for receptor activation by the particular hormonal therapy received by the patient (Kelly et al. 1997). Due to the evolutionary diversity of ligands known to bind with high affinity to the conserved LBD of the steroid and nuclear-receptor superfamily, mediated in most cases by relatively minor differences in primary LBD sequence, permissive AR mutations are likely to arise for all new-generation SARMS. Nonetheless, broadening the arsenal of AR antagonists will facilitate cyclic or effective second and third-tier hormonal treatment options for better overall disease management (Scher et al. 2004).

3.6

Cochaperones as Critical Mediators of AR Function in Prostate Cancer

Molecular chaperones are essential for the activity of diverse signaling molecules, which include steroid and tyrosine kinase receptors, p53 and telomerase, and are emerging as critical players in the pathogenesis of human disease (Auluck et al. 2002; Cheung and Smith 2000; Fonte et al. 2002; Waza et al. 2005; Whitesell and Lindquist 2005). Whereas the Hsp70/Hsp90 chaperone system traditionally has been ascribed to folding and stabilization of these molecules, compelling evidence suggests additional roles in movement in both the cytoplasm and nucleus, signal and/or transcriptional competence after activation, and for the biologically diverse actions mediated by structurally related proteins (Cheung and Smith 2000; Pratt et al. 2004; Pratt and Toft 1997). Indeed, recent data suggest that the Hsp90 chaperone machinery contributes not only to maturation of the steroid receptors,

Insights from AR Gene Mutations

215

but also to receptor movement through the cytoplasm, across the nuclear pore complex, in the nucleus, and cycling between active sites of transcription and nucleoplasmic stores of cofactors (Pratt et al. 2004). An example of how collocation of AR mutations identified in prostate cancer has facilitated a greater understanding of normal receptor function as it relates to chaperones comes from studies of the putative coregulator-binding region encompassing mutations in amino acids 670–678 at the boundary of the hinge and LBD. Previous studies utilized the isolated but complete AR-LBD as bait in the presence of high levels of ligand to identify AR coregulators. In addition to biasing the outcome to the highly structured ligand-dependent AF-2 site, modeling studies and the more recently generated crystal structures of this region indicate that the putative coregulatorbinding site spanning the mutated 670–678 residues might not be fully recapitulated using the LBD alone (Buchanan et al. 2000; Estebanez-Perpina et al. 2007). Consequently, AR residues 618–754 that encompass the hinge and the amino-terminal third of the LBD were used to identify and characterize the evolutionarily conserved Hsp70/Hsp90 cochaperone, small glutamine-rich tetratricopeptide repeat containing protein alpha (aSGT/SGTA), as a critical AR-binding partner in prostate cancer. Distinct from the nuclear transcriptional coregulators identified by most studies, aSGT/SGTA acts in the cytoplasm to maintain precise control of AR sensitivity, specificity, and degree of activation, and is downregulated in advanced prostate cancer (Buchanan et al. 2007). The loss of aSGT/SGTA results in decreased control of AR function (Fig. 1, Decreased Control), and mimics the functional consequences on AR activity as mutations between residues 670 and 678 (see Sect. 4.3). Along with increased expression of the AR, each of these mechanisms is implicated in the sensitization of metastatic prostate tumor cells to androgen signaling in a castrate environment (Fig. 1, Sensitization) (Buchanan et al. 2007). aSGT/SGTA belongs to a small group of steroid-receptor interacting chaperones defined by a conserved tetratricopeptide repeat (TPR) that includes the FK506binding proteins, FKBP51 and FKPB52; the immunophilin, Cyp-40; and the protein phosphatase, PP5. The importance of these TPR proteins to AR signaling is highlighted by the phenotype of FKBP52 knockout mice, which exhibit severe defects in male reproductive tissues (Cheung-Flynn et al. 2005; Yong et al. 2007).

3.7

Ansamycin Antibiotics to Target AR Signaling in Prostate Cancer

Geldanamycin and its derivatives are a class of ansamycin antibiotic that target the ATPase activity of Hsp90 to prevent the energy-dependent maturation of client proteins that include AR, p53, and cell surface receptors such as Her2, and promote their subsequent degradation by the 26S proteasome (Segnitz and Gehring 1997; Solit et al. 2002, 2003). Consequently, at least four compelling reasons warrant evaluation of these agents as a ‘‘double-whammy’’ treatment option for prostate

216

G. Buchanan et al.

cancer (Fig. 1, Ansamycins). First, maturation of the AR is thought essential for activity of both wild-type and mutant ARs. Second, inhibition of a range of Hsp90 client proteins may inhibit direct growth-promoting effects of other signaling molecules in prostate cancer cells, as well as target the stimulation of AR and its coregulators and/or the sensitization of receptor by cytokines, growth factors, and tyrosine kinase receptors (Fig. 1, Stimulation and Sensitization) (Powers and Workman 2006; Saporita et al. 2007). These factors have also been implicated in mediating constitutive bypass activation of the AR in the absence of ligand (Fig. 1, Constitutive bypass; see Sect. 6.6) (Feldman and Feldman 2001; Scher et al. 2004). Third, gain-of-function mutations in signaling molecules such as v-src, bcr-abl, c-kit, and p53 can increase their dependence on Hsp90 compared to their wild-type counterparts and make them more sensitive to degradation by 17-AAG (An et al. 2000; Blagosklonny et al. 1996; Fumo et al. 2004; Grbovic et al. 2006; Schulte et al. 1998; Whitesell and Lindquist 2005). Finally, cancer cells have a highly primed and sensitized Hsp90 chaperone system that allows more rapid responses to insult, buffers cellular signaling pathways against genetic instability, and enhances sensitivity to extracellular signaling, but which makes them particularly susceptible to inhibitors such as geldanamycin and its derivatives (Kamal et al. 2003; Whitesell and Lindquist 2005). To investigate the first of those hypotheses, the geldanamycin derivative 17-allylamino-17-demethoxygeldanamycin (17-AAG) was tested against a range of AR variants with distinct phenotypes (Fig. 3a). MostAR variants tested were equally or more sensitive to inhibition by 17-AAG in a dose-dependent manner compared to the wild-type receptor, and none exhibited resistance (Fig. 3b). In support of a use for these agents, geldanamycin and 17-AAG are able to reduce the levels of AR and suppress the proliferation of human prostate cancer cells both in vitro and in vivo (Solit et al. 2002, 2003). Phase II clinical trials with 17-AAG in prostate cancer are ongoing (Heath et al. 2005; Lattouf et al. 2006), but clinical breast cancer studies suggest that the responses will not recapitulate those found for tumors growing in mice (Eiseman et al. 2007), and that there may be a subclass of AR variants resistant to these inhibitors (see Sect. 4.8).

3.8

Truncated Receptors and the Constitutive Bypass Mechanism

Historical in vitro studies established that truncation of the AR in the hinge or LBD before amino-acid 710 results in a receptor that is constitutively active on classic androgen responsive elements (ARE) (Jenster et al. 1991). However, only recently a somatic AR gene mutation leading to a truncated receptor (Q640STOP) was identified in human prostate cancer (Ceraline et al. 2004). By circumventing completely the requirement for ligand activation, this truncated receptor provides a direct mechanism for constitutive bypass growth of prostate cancer and escape

Insights from AR Gene Mutations

217

Fig. 3 Ansamycin antibiotics effectively block activity of multiple classes of AR variants identified in prostate cancer. (a) Schematic of the human AR showing relative position of a subset of AR variants identified in prostate cancer. (b) Effect of increasing concentrations of 17allylamino-geldanamycin on activity of AR variants shown in A. Human prostate cancer PC-3 cells seeded in 96-well plates at 1.5 104 cells/well for 24 h were transiently transfected for 4 h with AR-wt or variant AR expression vectors (2.5 ng) and with the minimal androgen responsive probasin luciferase reporter construct (tk81-PB3) (100 ng) After transfection, cells were treated for 36 h with 1 nM DHT and increasing concentrations of 17-AAG. Gray shading represents the normalized AR-wt luciferase activity (sem) from each of the component experiments. Luciferase activity of each AR variant is expressed as a percentage of AR-wt activity measured in the same experiment. Data represent the mean (sem) activity measured in six replicate wells (See Color Insert)

from hormonal control (Fig. 1, Constitutive bypass). Signaling by this variant may also impart a paracrine response that promotes constitutive activity of wild-type AR in adjacent cells (Monge et al. 2006). The activity of constitutive truncated AR variants is not inhibited by the ansamycin antibiotics unlike the wtAR or all other mutant ARs analyzed (Buchanan, Butler & Tilley, unpublished observations). The constitutive response of truncated ARs may stem from decreased Hsp90 chaperone control (see Sect. 4.6), probably via amino acids 704–758 (Pratt and Toft 1997). This mechanism may also be related to the inability of the truncated ARs to adopt an N/C interaction, the release of distinct transcriptional domains in the NTD (see Sect. 6.1), constitutively nuclear localization, and more rapid cycling on and off response elements compared to the wild-type receptor (Farla et al. 2004; Tepper

218

G. Buchanan et al.

et al. 2002). Truncation of the AR after amino acid 718 (i.e., retaining the Hsp90-binding site) results in an inactive receptor variant that supports an argument for chaperone involvement (Gottlieb et al. 2004). The options to target constitutive truncated ARs likely are limited to those that act directly on receptor synthesis to reduce AR levels in the cell (see Sect. 4.9).

3.9

A New Wave: Agents Targeting AR Synthesis

As a consequence of the diverse ways the AR protein appears able to circumvent inhibition by specific agents, and to maintain sufficient signaling for continued survival and/or growth of prostate cancer cells, it has long been argued that a better approach to preventing AR signaling may be to inhibit receptor synthesis (Scher et al. 2004). Three approaches have been shown somewhat effective: antisense oligonucleotides, short hairpin RNAs, and histone deacetylase inhibitors (Fig. 1, Antisense, shRNA & HDACIs). Prostate-restricted expression of an AR short hairpin RNA (shRNA) using a lentiviral vector inhibited prostate tumor growth in vivo and delayed tumor progression following castration (Cheng et al. 2006). Similarly, AR antisense oligonucleotides reduced AR level and expression of the androgen-regulated PSA gene in vitro and in vivo (Ko et al. 2004). The histone deacetylase inhibitors (HDACIs) suberoylanilide hydroxamic acid (SAHA/vorinostat) and phenylbutyrate suppressed the growth of prostate cancer cells in vitro and in vivo (Butler et al. 2000; Carducci et al. 2001; Gore and Carducci 2000; Kelly et al. 2003; Marks et al. 2001), with efficacy partially stemming from reduced AR expression and activity (Marrocco et al. 2007). Although it is not yet known whether these agents will perform equally well against known AR variants or select for new ones under treatment conditions, successful application of these compounds clinically depends on developing more suitable means of introducing and maintaining them in prostate cells.

4 The DNA-Binding Domain 4.1

Collocation of AR Gene Mutations at 574–586

The AR-DBD typically is ascribed only the function of binding to specific response elements in DNA, but it has also been implicated in receptor dimerization, forms a putative protein–protein interaction surface, and was found to interact with several coregulators in a ligand-dependent manner (Blanco et al. 1998; Moilanen et al. 1998, 1999; Poujol et al. 1997). A total of five somatic missense mutations were identified in the AR-DBD in clinical prostate cancer that collocate to the 12 amino

Insights from AR Gene Mutations

219

acids 574–586 at the carboxyl-terminal end of the first zinc finger motif (Marcelli et al. 2000; Tilley et al. 1996). None of these residues has been reported to contain mutations that cause receptor inactivation in AIS. Highlighting the potential of these variants to mediate prostate cancer growth, substitution in the DBD of T575A in conjunction with the LNCaP T877A variant results in an AR promiscuous for ligand binding and transactivation at non-canonical-binding sites (Monge et al. 2006). This supports previous studies that demonstrated that mutations in the AR DBD can affect selectively AR transactivation and transrepression functions on different promoters despite reduced DNA-binding ability (Aarnisalo et al. 1999; Bruggenwirth et al. 1998). In addition, it has been suggested that mutations in the DBD might result in AR variants that bind to response elements normally specific for other nuclear receptors, thereby leading to inappropriate activation or repression of growth-regulatory pathways (Poujol et al. 1997). Mutations in the DBD also might alter the affinity of the AR for particular response elements, either directly or via modification of specific coregulator interactions, to produce both qualitative and quantitative differences in the regulation of its target genes. In a converse finding that supports this hypothesis, AR response elements differing by only a single nucleotide were found to exert different allosteric effects on AR structure that may relate to transcriptional capacity (Lai et al. 2007).

4.2

Targeting DNA Binding and Response Elements

Blocking the interaction of the receptor with its specific response elements, or preventing the generation of an active transcription complex on DNA, provides an alternative means of inhibiting AR function. The former approach has been tested by transfecting double-stranded DNA ‘‘decoys,’’ which consist of an androgen response element (Kuratsukuri et al. 1999), and synthetic polyamide molecules designed to target the consensus AR-binding site throughout the genome (Nickols and Dervan 2007) (Fig. 1, Response element blockers). The latter approach has been attempted using dominant negative AR inhibitors (Fig. 1, ARIs) (Butler et al. 2006; Palvimo et al. 1993). Whereas each of these approaches has blocked AR activity on known response elements, slowed growth of prostate cancer cells in culture and/or induced apoptosis (Kuratsukuri et al. 1999; Nickols and Dervan 2007; Palvimo et al. 1993), two conceptual challenges limit further development. First, difficulties associated with delivering these types of molecules to whole animals have not been sufficiently resolved. Second, recent genome-wide AR-binding studies (ChIP-chip) demonstrate that (1) the AR interacts with DNA directly via response elements and indirectly via other factors, and (2) the full complement of AR-binding sites or the determinants of DNA binding remain unknown (Bolton et al. 2007; Takayama et al. 2007; Wang et al. 2007) (Buchanan G, Coetzee GA & Frenkel B, unpublished data). In particular, those studies had difficulty in defining a consensus ARE and suggest that direct binding to DNA is mediated only loosely by the response element, but is stabilized in a cell-specific manner by collaborating factors binding to neighboring sites.

220

G. Buchanan et al.

5 The AR Amino-Terminal Domain Transcriptional regulation by the AR and other nuclear receptors stems from the recruitment of a unique complement of cofactors that mediate nuclear-cytoplasmic movement, response element recognition, chromatin remodeling, and ultimately recruitment of basal transcription factors. Whereas in many nuclear receptors this process is mediated through the conserved AF-2 surface in the LBD (see Sect. 4.1), the structural constraints of high-affinity hormone binding have restricted evolutionary changes in AF-2, and limited interactions with this surface to a few common proteins (He et al. 2004b). Consequently, the evolution of the different steroid receptors for enhancer/promoter and cell-specific regulation has necessitated divergence and strengthening of the NTD for distinct cofactor interactions (Han et al. 2005; He et al. 2004a; Thornton and Kelley 1998). Nonetheless, the nature of the NTDs as large, evolutionarily divergent, and structurally disordered domains has meant that they have received less attention than the domains involved in DNA and ligand binding. Mutations identified in the AR-NTD in prostate cancer have demonstrated in vitro and in vivo the critical roles played by intrinsic and induced structural order of this domain for normal receptor function, provided insight into functional subdomain structures, and have challenged the paradigm of AR involvement in prostate cancer development and progression.

5.1

AR NTD Structure and Activation Functions

For most steroid receptors, the conformational changes wrought in the LBD by agonist binding results in the formation of the primary p160 coactivator interaction surface, AF-2. The binding of p160s to this highly structured and conserved surface via LxxLL-containing peptides in turn mediates receptor transactivation capacity by both direct and indirect mechanisms (He and Wilson 2003). AR activation is distinct from this generalized model in two related ways. First, the AR AF-2 surface has been modified evolutionarily to preferentially interact with the 23FQNLF27 peptide of the AR-NTD in the N/C interaction, thereby displacing p160 interactions to the AR-NTD (Chang and McDonnell 2002). Second, AR transactivation capacity is determined almost exclusively by the NTD (Jenster et al. 1995). Consequently, the AF-2 surface makes a smaller direct contribution to AR transactivation activity than the analogous region of the other steroid receptors, whereas the AR-NTD plays a far more critical role. Early deletion mapping broadly categorized the AR-NTD as containing two transactivation units (TAU; Fig. 4a). TAU-1, which consists of amino acids 111–361, was defined as the minimal region sufficient for normal ligand-dependent activation of the AR, and deletion of TAU-1 completely abrogates receptor transactivation. TAU-5, defined as amino acids 352–537, was found to be sufficient and

Insights from AR Gene Mutations

221

Fig. 4 Structure of the AR amino-terminal domain (AR-NTD) and consequences of somatic NTD mutations in human prostate cancer. (a) Schematic to-scale structure of the AR-NTD delineating activation functions TAU-1 and TAU-5, secondary a-helical structure elements (H) determined by chemical and/or in silico analysis (Reid et al. 2002) and recognized subdomain structures within the NTD including homopolymeric repeat sequences [polyglutamine, (Q)n; polyglycine, (G)n; polyproline, (P)n] and peptides implicated in interdomain communication (N/C interaction) and cofactor binding (i.e., 23FQNLF27 and 433WHTLF437). SIGNATURE refers to the evolutionarily conserved NTD signature sequences 234–247. (b) Amino acid conservation across species in the

222

G. Buchanan et al.

necessary for the constitutive activity of ARs truncated of the LBD (Callewaert et al. 2006; Chamberlain et al. 1996; Jenster et al. 1995). Further experiments are required to determine whether the two TAUs exhibit functional redundancy or are completely independent, and how this relationship changes with the cellular and environmental context. However, the relative contribution of TAU-1 and TAU-5 to AR activity depends on the LBD, which suggests that the N/C interaction is important. Chemical analysis revealed that the AR-NTD has only a small number of predicted secondary structure elements, adopts a disordered but flexible structure typical of transactivation domains, and exhibits increased structural order following recruitment of the transcriptional machinery and other transcriptional coregulators (Kumar et al. 2004; Reid et al. 2003). More recently, core sequences of the NTD have been found to exist in collapsed disordered conformation or molten globulelike state that has distinct implications for AR transcriptional function (Lavery and McEwan 2008a, b).

Fig. 4 (Continued) AR-NTD and relationship to somatic AR mutations in prostate cancer. The five most evolutionarily conserved regions of the AR-NTD (Thornton and Kelley 1998) are shown as blue bars. Orange inset represents the AR-NTD signature sequence, which is almost 100% conserved among all AR species, delineated by alignment of 47 full-length AR-NTD sequences from 42 different animals (note that several fish species have been found to have alpha, beta, and gamma AR isoforms) using the Clustal W algorithm in Vector NTI (Invitrogen). The position of each somatic AR mutation identified in the AR-NTD in prostate cancer is shown as a vertical bar, with length proportional to the number of individual reports of mutations at that residue. Numbers and boxing indicate regions to which mutations cluster/collocate. (c) AR polyQ length frequency data from control populations. The human AR-NTD contains a polymorphic uninterrupted glutamine tract (polyQ) of 6–39 residues with a median of 19–20 in African American, 21–22 in Caucasian, 22–23 in Asian, and 23 in Hispanic populations (Edwards et al. 1992; Irvine et al. 1995; Spurdle et al. 1999), with 91–99% of all alleles within the range Q16–Q29. In addition to germlinelength polymorphisms, recent data suggest that the AR-CAG tract is a target for somatic alterations in clinical prostate cancer and animal models of the disease, which include contractions in repeat length and missense mutations (Buchanan et al. 2004; Hyytinen et al. 2002; Wallen et al. 1999; Watanabe et al. 1997). (d) A novel dual missense mutation identified in human prostate cancer results in disruption of the polyglutamine repeat of the receptor with two leucine (L) residues as indicated (i.e., polyQ2L). In silico modeling and analysis of calculated energy suggest that the polyQ2L variation forces a more rigid structure on the AR-NTD (ribbon diagrams). This conclusion is supported by secondary structure prediction [NNpredict algorithm of Kneller et al. (1990)], which indicates an increase in the a-helical (H) content of the region. (e) The AR-NTD signature sequence. Multiple sequence alignment of ARs from different species was performed using the Clustal W algorithm. The region between residues 234–247 (human numbering) is the most highly conserved in all known AR species. This region contains the only six residues conserved in all 47 known full-length AR NTDs. Blue, identical; Yellow, conserved; White, nonhomologous. (f) Overlays of ten independent 3D molecular dynamic model structures of AR-wt and AR-E326G signature sequence peptides, with each solution colored according to positional fluctuation (blue, minimal displacement; red, maximum displacement). The lower average positional fluctuation (p < 0.05; Root Mean Square Deviation, RMSD) of the E236G compared to AR-wt peptide suggests a more stable secondary structure (See Color Insert)

Insights from AR Gene Mutations

5.2

223

AR Gene Mutations in the NTD

The study of the AR-NTD is complicated by the presence of two long GC-rich polymorphic trinucleotide repeat sequences. As a consequence, and due to the relative length of the NTD and other factors, very few studies have sequenced the AR-NTD in prostate cancer. Nonetheless, emerging evidence suggests a role of AR mutations in this domain in disease etiology, including risk, the oncogenic potential of the receptor, and survival of tumor cells during ADT. In particular, three studies suggest that surgical and/or medical castration results in the preferential accumulation of mutations in the AR NTD. This was first documented in the autochthonous transgenic adenocarcinoma of the mouse prostate (TRAMP) model (Han et al. 2001). In that study, 6/8 AR gene mutations identified in tumors from intact TRAMP mice at 24–28 weeks of age were located in the AR-LBD. In mice castrated at 12 weeks, 7/7 AR gene mutations identified in recurrent tumors at 24–28 weeks of age were located in the AR-NTD. In clinical prostate cancer, two independent studies have shown that AR gene mutations predominantly arise in the NTD in patients treated with combined androgen blockade, and moreover are located almost exclusively within the same 40 amino acid region at the C-terminal of TAU-5 (Hyytinen et al. 2002; Scher et al. 2004). A similar incidence and/or pattern has not been observed in primary disease or in patients treated with monotherapy (Scher et al. 2004). Until recently, insufficient AR-NTD mutations had been identified in human prostate cancer to formally test their collocation in this domain. However, those now listed in the AR gene mutations database (http://androgendb.mcgill.ca) and identified recently by us and others (Scher et al. 2004) (unpublished) demarcate four clusters of mutations in the AR-NTD, include two with sufficient numbers to be formally defined as ‘‘hotspots’’ for mutations (Fig. 4b). Three of the five short NTD sequences conserved during evolution (Thornton and Kelley 1998) almost precisely overlap with mutation clusters at 221–269, 433–455, and 497– 537 (Fig. 4b).

5.3

Collocation of AR Gene Mutations at 43–90

The first area of collocation in the AR-NTD is between amino acids 43 and 90, a region with poor evolutionary conservation that contains the polyglutamine tract (polyQ) (Fig. 4b). The AR-polyQ consists of a string of glutamine residues ranging in length from 7 to 40 repeats in healthy individuals, with a median length of 20–21 repeats in African American, 22–23 in Caucasian, 23–24 in Asian, and 24 in Hispanic populations (Balic et al. 2002; Bennett et al. 2002; Edwards et al. 1992; Giovannucci et al. 1997; Hsing et al. 2000; Jin et al. 2000; Platz et al. 2000; Sartor et al. 1999) (Fig. 4c). More than 91–99% of AR alleles across all racial-ethnic groups encode a polyglutamine tract within the range of Q16–Q29 (Buchanan et al.

224

G. Buchanan et al.

2004). Variation within the normal range has been associated with prostate cancer risk, age of onset and/or risk of advanced disease at diagnosis (Beilin et al. 2000; Giovannucci et al. 1997; Hsing et al. 2000; Stanford et al. 1997), and an extraordinary array of human diseases that include breast and endometrial cancers, schizophrenia, infertility, polycystic ovarian syndrome, and cryptorchidism (Palazzolo et al. 2008). The AR-polyQ tract is a target for somatic contraction or missense mutation in clinical prostate cancer and animal models of the disease (Fig. 4b) (Buchanan et al. 2001b; Hyytinen et al. 2002; Schoenberg et al. 1994; Wallen et al. 1999; Watanabe et al. 1997). Because ARs with shorter polyQ tracts are more active (Buchanan et al. 2004), most believe that the relationship between AR polyQ length and disease relates simply to a gradient of AR transcriptional competence. This simple interpretation has been challenged by the study of a dual somatic missense mutation identified in prostate cancer that disrupts the pure AR-polyQ tract with two leucine residues [i.e., from wtAR(Q)23 to AR-polyQ2L = AR(Q)12L(Q)6L(Q)3] (Fig. 4d) (Buchanan et al. 2004). The AR-polyQ2L variant has the following properties compared to wtAR: (1) twofold to fourfold greater ability to transactivate target genes despite decreased protein stability; (2) increased response to the p160 coactivators; (3) increased structural order and less flexibility across the polyQ repeat region (Fig. 4d); and (4) 50% reduced capacity to undergo N/C interaction (Buchanan et al. 2004). Increased p160 coactivator levels during prostate cancer progression (Chmelar et al. 2007; Gregory et al. 2001) provide an obvious means by which AR-polyQ2L might maintain prostate cancer cell growth following ADT. The N/C acts in a temporal and spatial manner to protect the receptor from inappropriate binding and timing of cofactor interactions (van Royen et al. 2007), the polyQ2L variant probably alters. Alternatively, because the dynamic relationship of the AR with N/C interaction and p160 recruitment. The polyQ2L AR variant is consequently predicted to respond quite differently to the cellular milieu of cofactors, which may translate to activation of distinct gene pathways. These studies provided both a novel mechanism for AR-mediated survival of prostate cancer cells during ADT, and a model for the relationship between transactivation capacity, NTD flexibility, and the vital in vivo N/C interaction. These findings also precipitated a more detailed analysis of wtAR N/C interaction and polyQ length, which revealed that maximal N/C is sustained only for 16–29 glutamine repeats (Buchanan et al. 2004). In addition to providing an evolutionary reason why AR-polyQ length has been maintained mostly within this same range across populations (Fig. 4c), these findings suggest that functionality of AR with a polyQ length either shorter or longer than the critical range of 16–29 residues may be a more important mediator of disease phenotype than a stepwise reduction in activity with increasing polyQ length across the entire range (Buchanan and Tilley 2000). In this model, ARs with 16–29 glutamines in the polyQ tract could be considered functionally equivalent, and those with either shorter or longer repeats as distinct. As most population studies have historically used the median AR-polyQ repeat length as a cutpoint, many relationships between AR genotype and disease may well have been missed.

Insights from AR Gene Mutations

5.4

225

Clustering of AR Gene Mutations at 221–269. The AR NTD-Signature Sequence

The NTD is the least conserved of the steroid receptor domains, sharing as little as 20% amino acid homology between even the closely related 3-ketosteroid receptors for, progestins, glucocorticoids, and mineralocorticoids (i.e., PR, GR, and MR, respectively). In comparison, these receptors share approximately 70–80% homology in the LBD and DBD regions. The AR NTD itself has diverged in sequence during evolution at almost twice the rate of the NTDs of the other steroid receptors, and only five small regions of the NTD have been conserved (Fig. 4b) (Thornton and Kelley 1998). Alignment of 47 informative full-length AR-NTD sequences from 42 different animals (data not shown) confirmed previous observations (Han et al. 2001; Thornton and Kelley 1998) that the most dramatically conserved region of the NTD resides between amino acids 234 and 247 (Fig. 4b, e). This small region, which our group has coined the AR-NTD signature, has no sequence homology to any other protein, and contains the only six amino acids conserved in all known AR-NTD sequences (Fig. 4e). Mutations have been identified recently proximal to the AR-NTD signature in clinical prostate cancers (DM Robins, unpublished) and TRAMP tumors (i.e., A234K & E236G, Fig. 4E) (Han et al. 2001). In silico modeling suggested that A234K and E236G substitutions stabilize the secondary structure about the NTD signature sequence, which may confer greater order to the disordered NTD (Fig. 4f). In support of this hypothesis, each of these AR variants was found to confer greater basal activity (i.e., in the absence of ligand), and increased response to androgenic and nonandrogenic ligands compared to wtAR (Fig. 5a). The basal activity of AR-E236G consistently was higher than wtAR, but cell and experimental variation in the comparative response of AR-E236G to ligand (Fig. 5b) suggests altered response or interaction with a variable pool of coregulators (Buchanan et al. 2004). AR-E236G exhibits a greater response to coactivators ARA70 (Fig. 5b) and ARA160 (Han et al. 2001), and more than a 40% reduced interaction with the C-terminal of Hsp70-interacting protein (CHIP) chaperone that acts as a negative regulator or AR activity (He et al. 2004a). The prostate and epithelial cell-specific probasin promoter was used to generate three transgenic mouse lines overexpressing either wtAR, the promiscuous ART877A LNCaP variant (see Sect. 4.1), or AR-E236G variant as a model of altered AR function (Han et al. 2005). Whereas transgenic mice overexpressing wtAR and AR-T877A had no discernable phenotype, 100% of mice expressing AR-E236G developed lesions similar to prostatic intraepithelial neoplasia (PIN) at 12 weeks of age, and advanced prostate cancer with lung metastases by 50 weeks of age (Fig. 5c) (Han et al. 2005). Unlike the TRAMP model, AR-E236G mice do not rely on overexpression of viral oncoproteins or supraphysiological levels of steroids to initiate tumorigenesis. Instead, these findings highlight the functional significance of mutations in the AR NTD, and argue that mutations resulting in aberrant AR function (in the case of E236G via altered interaction with coregulators), can turn

226

G. Buchanan et al.

Fig. 5 Mice expressing AR-E236G develop spontaneous prostate cancer with distant lung metastases. (a) Transactivation profiling of AR-wt and receptor variants identified in clinical (LNCaP T877A variant) prostate cancer and animal models (A234K and E236G) of the disease. Data show mean (sem) transactivation capacity of AR-wt and AR variants normalized for b-galactosidase activity in transfected PC-3 cells in the absence or presence of different steroids. (b) AR-E236G transactivation capacity is cell type and/or ligand dependent, and more responsive to cofactors compared to AR-wt. Transactivation analysis was performed in ultraresponsive PC-3 cells (Buchanan et al. 2004) under different conditions as A (left-hand panel), or in the absence or

Insights from AR Gene Mutations

227

the AR into an oncogene sufficient to trigger tumorigenesis in the prostate and promote metastatic progression. An important hypothesis arising from these observations is that a change in the level or composition of the AR cofactor milieu in a prostate cancer cell may be sufficient to promote the receptors oncogenic potential in the absence of structural changes to the receptor itself. Affymetrix 430 2.0 microarray analysis on prostate RNA from wtAR, ART877A, and AR-E236G overexpressing mice early in disease etiology has revealed two distinct facts: (1) overexpression of all three ARs resulted in increased or decreased expression of a unique subset of target genes; and (2) the greatest number of differentially expressed genes was observed in the comparison of wtAR with E236G, and the least for E236G with T875A (Fig. 5d). A subset of AR variants may promote cancer initiation and/or progression by qualitatively or quantitatively affecting expression of a unique complement of target genes (Fig. 1, Differential gene activation). The analysis of these particular AR variants (such as E236G) may allow identification of the pivotal genes capable of mediating oncogenic transformation of the prostate well before prostate cancer arises. This is a paradigm shift from defining those genes expressed at later stages of tumorigenesis as a consequence of malignant transformation.

5.5

Clustering of AR Gene Mutations at 431–455

The clustering of mutations at residues 431–455 intersects three documented structural and/or functional elements in the AR-NTD; TAU-5, and the polyglycine repeat (polyG) and 433WHTLF437 sequences encompassed by it (Fig. 4a). Whereas the individual mutations have not been investigated in detail, their clustering supports an emerging body of evidence that TAU-5 is a critical mediator of normal AR activity (Callewaert et al. 2006; Funderburk et al. 2008) (EF Need and Fig. 5 (Continued) presence of specific receptor coregulators as indicated (right-hand panel). (c) Pathobiology of prostate samples and tumors arising in transgenic mice overexpressing AR-E236G. Representative histological sections (5 um) were prepared from AR-wt and AR-E236G mice at 12 and 50 weeks of age, and analyzed as follows: H&E staining of 12-week-old ventral prostate samples demonstrates (i) normal architecture in AR-wt overexpressing mice, but (ii) the presence of PIN-like lesions in mice overexpressing AR-E236G. Mice overexpressing AR-T877A were indistinguishable from those overexpressing AR-wt (not shown). (iii–vi) Staining of locally advanced prostate cancers in AR-E236G mice at 50 weeks with H&E, and HA, AR and E-cadherin antibodies, respectively. (vii-ix) Staining of lung metastasis from AR-E236G mice at 50 weeks of age with H&E, HA, and NKX3.1 antibodies, respectively. Mice overexpressing the LNCaP AR variant, AR-T877A, exhibited similar phenotype to wtAR overexpressing mice. (d) Differential gene expression in prostate samples from AR-wt, AR-T877A, and AR-E236G overexpressing mice. For each mutant mouse line, RNA pooled from the ventral prostates of three animals at 12 weeks of age was analyzed using Affymetrix mouse genome 430 2.0 arrays. Shown on clock diagrams are the number of genes with significantly increased or decreased gene expression among animal lines. These data suggest that overexpression of each AR is affecting a unique subset of genes, with the largest difference between AR-wt and AR-E236G (See Color Insert)

228

G. Buchanan et al.

G Buchanan, unpublished). In particular, TAU-5 appears to be the primary mediator of transactivation by an AR lacking the LBD or activated by cytokine or growth factor pathways. In addition, TAU-5 has been implicated in DNA binding and as the primary NTD interaction site of the p160 coactivators, the transcriptional coregulator p300/CBP, and the corepressor SMRT (Alen et al. 1999; Berrevoets et al. 1998; Brodie and McEwan 2005; Callewaert et al. 2006; Chamberlain et al. 1996; Dotzlaw et al. 2002; Fronsdal et al. 1998; Ikonen et al. 1997; Ma et al. 1999). Of particular importance is the potential for p160s to form a bridge between the AR-NTD and LBD, thereby stabilizing the receptor for maximal activity (Ma et al. 1999; Shen et al. 2005). In certain stages of AR activation, this bridge may circumvent the N/C interaction and alter AR function (Shen et al. 2005). While a model that incorporates each of these effects is still lacking, the body of evidence argues that TAU-5 functions to ‘‘fine-tune’’ transcriptional response of the AR based on the cellular complement of coregulators. AR N/C interaction is determined primarily by interaction of the 23FQNLF27 sequence with the AF-2 surface in the LBD after agonist ligand binding and has at least three key effects: slowing the rate of ligand dissociation, excluding coregulator interactions from AF-2, and allowing activation of chromatin-integrated target genes (He et al. 2000, 2001; Li et al. 2006; Toumazou et al. 2007). However, N/C also depends partially on the structurally similar 433WHTLF437 peptide in TAU-5, which co-ordinates with 23FQNLF27 to slow ligand dissociation but interacts with the LBD in a ligand and AF-2 independent manner (He et al. 2000). Our unpublished data suggest that deletion of 433WHTLF437 leads to increased interaction of the AR-NTD with both the p160 coactivator, GRIP1 and the corepressor, SMRT (EF Need and G Buchanan, unpublished data). Blocking N/C also increases AR interaction with the p160 coactivators (van Royen et al. 2007). Missense mutations identified in or flanking the 433WHTLF437 peptide in clinical prostate cancer (Fig. 4b) are likely to have significant consequences for AR function and modulate N/C and/or coregulator interactions at either TAU-5 or AF-2. The AR polyglycine (polyG) repeat at amino acids 449–472 in TAU-5 exhibits polymorphic tendencies and ranges 10–27 repeats in the normal population, but its length is biased to just four common alleles of 19–20 (5.5%) or 23–24 (87.5%) residues (Platz et al. 1998). This bias may reflect a biological/functional requirement for a precise polyG length in the protein or greater replication fidelity of the encoding sequence. Nonetheless, longer repeats have been associated with increased AR activity, penile hypospadias, and cryptorchidism in men, and higher-grade endometrial cancers in women (Radpour et al. 2007; Rodriguez et al. 2006; Werner et al. 2006), while men with AR polyG repeat lengths shorter or longer than 23 residues may have a slight decreased risk of prostate cancer (Platz et al. 1998). The effect of polyG length variation may depend on sequences within the AR hinge (Werner et al. 2006), which suggests a link between mutations in this region and communication among AR domains. Four somatic missense mutations have been identified in the polyG tract that result in the missense replacement of glycine residues within the repeat (Sanchez et al. 2006), but the precise consequences for receptor function remain unknown.

Insights from AR Gene Mutations

5.6

229

Constitutive Bypass Via Ligand-Independent Activation

Growth factors, cytokines, protein kinase-A and overexpression of HER2/neu have been found to enhance AR transactivation capacity, an effect generally thought to stem from altered/induced states of receptor phosphorylation, and/or phosphorylation of AR cofactors (Fig. 1, Stimulation) (Craft et al. 1999; Culig et al. 2002; Gregory et al. 2004; Nazareth and Weigel 1996; Rowan et al. 2000a, b; Ueda et al. 2002a; b; Yeh et al. 1999). For example, high Her2/neu expression exacerbates the impact of high AR levels on disease progression and metastatic potential in clinical prostate cancer (Ricciardelli et al. 2008). Similar mechanisms appear to promote sufficient AR function in the absence of androgens in vitro or in vivo (Craft et al. 1999; Culig et al. 1994; Gregory et al. 2004; Grossmann et al. 2001; Nazareth and Weigel 1996; Ueda et al. 2002b) for activation of androgen-regulated genes and proliferation and survival of prostate cancer cells (Wen et al. 2000; Yeh et al. 1999). These constitutive bypass mechanisms do not involve an AR truncation mutation. Whereas a direct link between the phosphorylation status of particular AR or cofactor residues and effects on downstream receptor signaling have been difficult to demonstrate conclusively, the AR-NTD appears pivotal. Of the approximately 13 putative AR phosphorylation sites, 11 reside in the NTD and only one in each of the LBD and hinge regions (Gioeli et al. 2002, 2006; Guo et al. 2006; Lin et al. 2001; Mahajan et al. 2007; Weigel and Moore 2007). While none of the individual phosphorylation sites has been identified as mutated in prostate cancer, several lie within areas of mutation clustering and have been implicated in similar transcriptional functions. In particular, S256 and Y267 phosphorylation sites lie with mutations clustered about the NTD signature at residues 221–269; S81 and S94 sites are proximal to collocation with the polyQ at 43–90; S515 and Y534 are located within the hotspot for mutations in TAU-5 at residues 497–537. Mutation of the two NTD Erk2 phosphorylation sites in TAU-5 to alanine (i.e., S425A and S515A) impaired AR activity, decreased nuclear localization, and caused cytoplasmic aggregation of the receptor (Funderburk et al. 2008), which suggests an important link between AR phosphorylation and normal receptor function. Whether receptor NTD mutations identified in prostate cancer affect interaction with, or response of, the AR to kinase signaling has not been addressed specifically. Alternatively, AR activation by androgens initiates a rapid nongenomic phosphorylation cascade that has been implicated in a multitude of downstream consequences for AR activity and growth of prostate cancer cells. For example, Hsp27 is phosphorylated via MAPK14/p38 within minutes of androgen binding to the AR, displaces Hsp90 in the AR complex, and chaperones the receptor into the nucleus and onto AREs (Zoubeidi et al. 2007). Specific AR mutations may affect these events. Recognition of the importance of phosphorylation cascades in AR signaling should lead to the study of specific inhibitors (Fig. 1, Inhibitors), or the use of antibodies to cell-surface receptors that impact these processes (Fig. 1, Antibodies). AR antibodies, which have been shown to suppress AR activity and proliferation of androgen-independent prostate cancer cells, have the potential to circumvent both classical AR activity and nongenomic AR signaling (Zegarra-Moro et al. 2002).

230

5.7

G. Buchanan et al.

Collocation of AR Gene Mutations at 497–537

The second area of collocation in TAU-5 lies at the far C-terminus of the NTD between amino acids 497 and 537, in one of only three small areas of high amino acid sequence identify between species (Fig. 4b). Remarkably, more than 25% of all mutations detected in the AR-NTD in prostate cancer have been detected in the 497–537 region (Fig. 4b), with the majority identified in prostate tumors from men following complete androgen blockade (Need, Tilley, Buchanan, unpublished) (Hyytinen et al. 2002). Preliminary investigation of the 497–537 region, which lies at the far C-terminus of TAU-5, has revealed two dramatic results. First, several mutations identified in prostate cancer between residues 497 and 537 result in increased response to all three of the p160 coactivators compared to wtAR (EF Need, G Buchanan and WD Tilley, unpublished), which is consistent with previous studies that demonstrate a role for his region in maximal p160 response (Callewaert et al. 2006). Enhancement of AR response to coregulators, analogous to increased levels of the p160s in prostate cancer (Culig et al. 2004; Gregory et al. 2001), likely creates a permissive environment for AR activation in the low androgen environment or in response to nonclassical agonists. Second, deletion of residues 497–537 only moderately affected transcriptional response of the AR, but completely abolished receptor N/C and p160 interactions as determined using the mammalian two-hybrid system (EF Need, G Buchanan and WD Tilley, unpublished). Whether these two events are mutually exclusive and/or exhibit a spatial and temporal relationship with the AR remains to be defined. Nonetheless, as regions primarily responsible for p160 and N/C interactions lie elsewhere in the NTD, these findings highlight a structural or indirect functional requirement for residues 497–537 in normal AR function. Clustering of mutations to the 431–455 and 497–537 regions of TAU-5 thus constitutes two examples of where mutations in prostate cancer disrupt inter- or intramolecular communication between AR domains or cofactor interactions. Like other regions of mutation clustering in the NTD, these changes may be permissive for continued AR signaling during ADT and/or alter, AR function sufficiently to promote oncogenic transformation and disease progression as exemplified by the E236G mutation.

6 Conclusions There are several compelling genetic and functional reasons why germline variations in the AR that relate to disease are tolerated, and why the receptor itself appears to be particularly vulnerable to somatic mutations in prostate cancer. (1) The AR resides on the X chromosome and is present as a single allele in men, and as a single active allele in women due to X inactivation. Thus, somatic and germline alterations in the AR are more likely to exhibit functional consequences. (2) The AR itself is not essential for life but exhibits diverse functional roles from

Insights from AR Gene Mutations

231

apoptosis to growth stimulation, and impacts, or is impacted on, by a vast array of cellular signaling systems. This provides the necessary functional leeway by which interrupted androgen signaling can be tolerated and ultimately adapted to the particular needs of a prostate cancer cell, for example. (3) PolyQ and polyG repeat sequence expansion in higher-order mammals appears to represent a trade-off between the capacity for transcriptional activation and fine-tuning of functional control. Unfortunately, the nature of the encoding DNA sequence makes these repeats particularly susceptible to both germline and somatic variation. (4) By aggressively targeting the AR pathway in a multifocal prostate cancer, and particularly after the disease has acquired the necessary genetic heterogeneity for metastatic spread, mutant ARs that serendipitously confer resistance to treatment (or are stimulated by them) are selected. Given the protracted development and progression of prostate cancer, and that combinations of local and systemic therapies are now administered earlier in the course of disease, the characteristics of prostate cancer have changed. (5) Transcriptional regulation by the AR involves interaction with a unique complement of coregulators that mediate nuclear-cytoplasmic movement, recognition of response elements, chromatin remodeling, interdomain communication, and recruitment of basal transcription factors. As detailed in this chapter, perturbation of this finely balanced cascade of events is achieved easily by a broad range of single point mutations that can have dire functional consequences, which are exemplified by the AR-E236G variant. The collocation of AR gene mutations in prostate cancer has been instructive in defining the role of the receptor in disease etiology, but few mutations have been detected outside of the canonical LBD until recently. Sufficient mutations have now been identified in the AR-NTD to assign four regions of clustering/collocation to that domain, three of which coincide with the limited evolutionarily conserved NTD sequence. A unifying theme is emerging from the vast body of functional data on mutations that cluster/collocate throughout the AR in prostate cancer. In almost every case, the direct or indirect consequence is the disruption of AR-cofactor interactions or interdomain communication. Continued analysis of these AR variants allows delineation of the subdomain structure of the receptor and key coregulatorbinding sites; provides functional data to support and/or initiate structural analysis; defines the biology underpinning ARs role in the development, progression, and treatment resistance of prostate cancer; and suggests new ways of targeting the receptor for treatment. Acknowledgments The authors acknowledge the contributions of Drs. Rene Chmelar and Guangzhou Han (Fred Hutchinson Cancer Research Center, University of Washington) and technical assistance of Ms. Joanna Treloar. The authors are supported by the Prostate Cancer Foundation of Australia (GB, ID#Y102), the National Health and Medical Research Council of Australia (WDT, ID#453662), The United States Department of Defense (WDT, ID#PC060443), and a SPORE Grant from the National Cancer Institute (DMR, NCI P50 CA69568). GB holds a National Health and Medical Research Council of Australia CJ Martin Biomedical Fellowship. EFN is a recipient of a Freemasons Foundation Postdoctoral Fellowship.

232

G. Buchanan et al.

References Aarnisalo P, Santti H, Poukka H, Palvimo JJ, Janne OA (1999) Transcription activating and repressing functions of the androgen receptor are differentially influenced by mutations in the deoxyribonucleic acid-binding domain. Endocrinology 140:3097–3105 Alen P, Claessens F, Verhoeven G, Rombauts W, Peeters B (1999) The androgen receptor aminoterminal domain plays a key role in p160 coactivator-stimulated gene transcription. Mol Cell Biol 19:6085–6097 An WG, Schulte TW, Neckers LM (2000) The heat shock protein 90 antagonist geldanamycin alters chaperone association with p210bcr-abl and v-src proteins before their degradation by the proteasome. Cell Growth Differ 11:355–360 Askew EB, Gampe RT, Jr, Stanley TB, Faggart JL, Wilson EM (2007) Modulation of androgen receptor activation function 2 by testosterone and dihydrotestosterone. J Biol Chem 282:25801–25816 Auluck PK, Chan HY, Trojanowski JQ, Lee VM, Bonini NM (2002) Chaperone suppression of alpha-synuclein toxicity in a Drosophila model for Parkinson’s disease. Science 295:865–868 Balic I, Graham ST, Troyer DA, Higgins BA, Pollock BH, Johnson-Pais TL, Thompson IM, Leach RJ (2002) Androgen receptor length polymorphism associated with prostate cancer risk in Hispanic men. J Urol 168:2245–2248 Beilin J, Ball EM, Favaloro JM, Zajac JD (2000) Effect of the androgen receptor CAG repeat polymorphism on transcriptional activity: specificity in prostate and non-prostate cell lines. J Mol Endocrinol 25:85–96 Bennett CL, Price DK, Kim S, Liu D, Jovanovic BD, Nathan D, Johnson ME, Montgomery JS, Cude K, Brockbank JC, Sartor O, Figg WD (2002) Racial variation in CAG repeat lengths within the androgen receptor gene among prostate cancer patients of lower socioeconomic status. J Clin Oncol 20:3599–3604 Bentel JM, Tilley WD (1996) Androgen receptors in prostate cancer. J Endocrinol 151:1–11 Berrevoets CA, Doesburg P, Steketee K, Trapman J, Brinkmann AO (1998) Functional interactions of the AF-2 activation domain core region of the human androgen receptor with the amino-terminal domain and with the transcriptional coactivator TIF2 (transcriptional intermediary factor2). Mol Endocrinol 12:1172–1183 Blagosklonny MV, Toretsky J, Bohen S, Neckers L (1996) Mutant conformation of p53 translated in vitro or in vivo requires functional HSP90. Proc Natl Acad Sci U S A 93:8379–8383 Blanco JG, Minucci S, Lu J, Yang XJ, Walker KK, Chen H, Evans RM, Nakatani Y, Ozato K (1998) The histone acetylase PCAF is a nuclear receptor coactivator. Genes Dev 12:1638–1651 Bohl CE, Wu Z, Miller DD, Bell CE, Dalton JT (2007) Crystal structure of the T877A human androgen receptor ligand-binding domain complexed to cyproterone acetate provides insight for ligand-induced conformational changes and structure-based drug design. J Biol Chem 282:13648–13655 Bolton EC, So AY, Chaivorapol C, Haqq CM, Li H, Yamamoto KR (2007) Cell- and gene-specific regulation of primary target genes by the androgen receptor. Genes Dev 21:2005–2017 Brodie J, McEwan IJ (2005) Intra-domain communication between the N-terminal and DNAbinding domains of the androgen receptor: modulation of androgen response element DNA binding. J Mol Endocrinol 34:603–615 Bruggenwirth HT, Boehmer AL, Lobaccaro JM, Chiche L, Sultan C, Trapman J, Brinkmann AO (1998) Substitution of Ala564 in the first zinc cluster of the deoxyribonucleic acid (DNA)binding domain of the androgen receptor by Asp, Asn, or Leu exerts differential effects on DNA binding. Endocrinology 139:103–110 Buchanan G, Greenberg NM, Scher HI, Harris JM, Marshall VR, Tilley WD (2001a) Collocation of androgen receptor gene mutations in prostate cancer. Clin Cancer Res 7:1273–1281 Buchanan G, Irvine RA, Coetzee GA, Tilley WD (2001b) Contribution of the androgen receptor to prostate cancer predisposition and progression. Cancer Metastasis Rev 20:207–223

Insights from AR Gene Mutations

233

Buchanan G, Craft PS, Yang M, Cheong A, Prescott J, Jia L, Coetzee GA, Tilley WD (2004) PC-3 cells with enhanced androgen receptor signaling: A model for clonal selection in prostate cancer. Prostate 60:352–366 Buchanan G, Ricciardelli C, Harris JM, Prescott J, Yu ZC, Jia L, Butler LM, Marshall VR, Scher HI, Gerald WL, Coetzee GA, Tilley WD (2007) Control of androgen receptor signaling in prostate cancer by the cochaperone small glutamine rich tetratricopeptide repeat containing protein alpha. Cancer Res 67:10087–10096 Buchanan G, Tilley WD (2000) Androgen receptor structure and function in prostate cancer. In: JJ Li, JR Darling, SA Li (eds.), Hormonal Carcinogenesis III, III edn, pp. 333–341. New York: Springer-Verlag Buchanan G, Yang M, Cheong A, Harris JM, Irvine RA, Lambert PF, Moore NL, Raynor M, Neufing PJ, Coetzee GA, Tilley WD (2004) Structural and functional consequences of glutamine tract variation in the androgen receptor. Hum Mol Genet 13:1677–1692 Buchanan G, Yang M, Nahm SJ, Han G, Moore N, Bentel JM, Matusik RJ, Horsfall DJ, Marshall VR, Greenberg NM, Tilley WD (2000) Mutations at the boundary of the hinge and ligand binding domain of the androgen receptor confer increased transactivation function. Mol Endocrinol 15:46–56 Butler LM, Agus DB, Scher HI, Higgins B, Rose A, Cordon-Cardo C, Thaler HT, Rifkind RA, Marks PA, Richon VM (2000) Suberoylanilide hydroxamic acid, an inhibitor of histone deacetylase, suppresses the growth of prostate cancer cells in vitro and in vivo. Cancer Res 60:5165–5170 Butler LM, Centenera MM, Neufing PJ, Buchanan G, Choong CS, Ricciardelli C, Saint K, Lee M, Ochnik A, Yang M, Brown MP, Tilley WD (2006) Suppression of androgen receptor signaling in prostate cancer cells by an inhibitory receptor variant. Mol Endocrinol 20:1009–1024 Callewaert L, Van Tilborgh N, Claessens F (2006) Interplay between two hormone-independent activation domains in the androgen receptor. Cancer Res 66:543–553 Carducci MA, Gilbert J, Bowling MK, Noe D, Eisenberger MA, Sinibaldi V, Zabelina Y, Chen TL, Grochow LB, Donehower RC (2001) A Phase I clinical and pharmacological evaluation of sodium phenylbutyrate on an 120-h infusion schedule. Clin Cancer Res 7:3047–3055 Ceraline J, Cruchant MD, Erdmann E, Erbs P, Kurtz JE, Duclos B, Jacqmin D, Chopin D, Bergerat JP (2004) Constitutive activation of the androgen receptor by a point mutation in the hinge region: a new mechanism for androgen-independent growth in prostate cancer. Int J Cancer 108:152–157 Chamberlain NL, Whitacre DC, Miesfeld RL (1996) Delineation of two distinct type 1 activation functions in the androgen receptor amino-terminal domain. J Biol Chem 271:26772–26778 Chang CY, McDonnell DP (2002) Evaluation of ligand-dependent changes in AR structure using peptide probes. Mol Endocrinol 16:647–660 Cheng H, Snoek R, Ghaidi F, Cox ME, Rennie PS (2006) Short hairpin RNA knockdown of the androgen receptor attenuates ligand-independent activation and delays tumor progression. Cancer Res 66:10613–10620 Cheung J, Smith DF (2000) Molecular chaperone interactions with steroid receptors: an update. Mol Endocrinol 14:939–946 Cheung-Flynn J, Prapapanich V, Cox MB, Riggs DL, Suarez-Quian C, Smith DF (2005) Physiological role for the cochaperone FKBP52 in androgen receptor signaling. Mol Endocrinol 19:1654–1666 Chmelar R, Buchanan G, Need EF, Tilley W, Greenberg NM (2007) Androgen receptor coregulators and their involvement in the development and progression of prostate cancer. Int J Cancer 120:719–733 Craft N, Shostak Y, Carey M, Sawyers CL (1999) A mechanism for hormone-independent prostate cancer through modulation of androgen receptor signaling by the HER-2/neu tyrosine kinase. Nat Med 5:280–285

234

G. Buchanan et al.

Culig Z, Hobisch A, Cronauer MV, Radmayr C, Trapman J, Hittmair A, Bartsch G, Klocker H (1994) Androgen receptor activation in prostatic tumor cell lines by insulin-like growth factor-I, keratinocyte growth factor, and epidermal growth factor. Cancer Res 54:5474–5478 Culig Z, Hobisch A, Hittmair A, Peterziel H, Cato AC, Bartsch G, Klocker H (1998) Expression, structure, and function of androgen receptor in advanced prostatic carcinoma. Prostate 35:63–70 Culig Z, Hobisch A, Bartsch G, Klocker H (2000) Androgen receptor – an update of mechanisms of action in prostate cancer. Urol Res 28:211–219 Culig Z, Bartsch G, Hobisch A (2002) Interleukin-6 regulates androgen receptor activity and prostate cancer cell growth. Mol Cell Endocrinol 197:231–238 Culig Z, Comuzzi B, Steiner H, Bartsch G, Hobisch A (2004) Expression and function of androgen receptor coactivators in prostate cancer. J Steroid Biochem Mol Biol 92:265–271 Dotzlaw H, Moehren U, Mink S, Cato AC, Iniguez Lluhi JA, Baniahmad A (2002) The amino terminus of the human AR is target for corepressor action and antihormone agonism. Mol Endocrinol 16:661–673 Edwards A, Hammond HA, Jin L, Caskey CT, Chakraborty R (1992) Genetic variation at five trimeric and tetrameric tandem repeat loci in four human population groups. Genomics 12:241–253 Eiseman JL, Guo J, Ramanathan RK, Belani CP, Solit DB, Scher HI, Ivy SP, Zuhowski EG, Egorin MJ (2007) Evaluation of plasma insulin-like growth factor binding protein 2 and Her-2 extracellular domain as biomarkers for 17-allylamino-17-demethoxygeldanamycin treatment of adult patients with advanced solid tumors. Clin Cancer Res 13:2121–2127 Estebanez-Perpina E, Arnold LA, Nguyen P, Rodrigues ED, Mar E, Bateman R, Pallai P, Shokat KM, Baxter JD, Guy RK, Webb P, Fletterick RJ (2007) A surface on the androgen receptor that allosterically regulates coactivator binding. Proc Natl Acad Sci U S A 104:16074–16079 Farla P, Hersmus R, Geverts B, Mari PO, Nigg AL, Dubbink HJ, Trapman J, Houtsmuller AB (2004) The androgen receptor ligand-binding domain stabilizes DNA binding in living cells. J Struct Biol 147:50–61 Feldman BJ, Feldman D (2001) The development of androgen-independent prostate cancer. Nat Rev 1:34–45 Fonte V, Kapulkin V, Taft A, Fluet A, Friedman D, Link CD (2002) Interaction of intracellular beta amyloid peptide with chaperone proteins. Proc Natl Acad Sci U S A 99:9439–9444 Fronsdal K, Engedal N, Slagsvold T, Saatcioglu F (1998) CREB binding protein is a coactivator for the androgen receptor and mediates cross-talk with AP-1. J Biol Chem 273:31853–31859 Fumo G, Akin C, Metcalfe DD, Neckers L (2004) 17-Allylamino-17-demethoxygeldanamycin (17-AAG) is effective in down-regulating mutated, constitutively activated KIT protein in human mast cells. Blood 103:1078–1084 Funderburk SF, Shatkina L, Mink S, Weis Q, Weg-Remers S, Cato AC (2008) Specific N-terminal mutations in the human androgen receptor induce cytotoxicity. Neurobiol Aging 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) Androgen receptor phosphorylation. Regulation and identification of the phosphorylation sites. J Biol Chem 277:29304–29314 Gioeli D, Black BE, Gordon V, Spencer A, Kesler CT, Eblen ST, Paschal BM, Weber MJ (2006) Stress kinase signaling regulates androgen receptor phosphorylation, transcription, and localization. Mol Endocrinol 20:503–515 Giovannucci E, Stampfer MJ, Krithivas K, Brown M, Dahl D, Brufsky A, Talcott J, Hennekens CH, Kantoff PW (1997) The CAG repeat within the androgen receptor gene and its relationship to prostate cancer. Proc Natl Acad Sci U S A 94:3320–3323 Gore SD, Carducci MA (2000) Modifying histones to tame cancer: clinical development of sodium phenylbutyrate and other histone deacetylase inhibitors. Expert Opin Investig Drugs 9:2923–2934 Gottlieb B, Beitel LK, Wu JH, Trifiro M (2004) The androgen receptor gene mutations database (ARDB): 2004 update. Hum Mutat 23:527–533

Insights from AR Gene Mutations

235

Grbovic OM, Basso AD, Sawai A, Ye Q, Friedlander P, Solit D, Rosen N (2006) V600E B-Raf requires the Hsp90 chaperone for stability and is degraded in response to Hsp90 inhibitors. Proc Natl Acad Sci U S A 103:57–62 Gregory CW, He B, Johnson RT, Forsd OH, Mohler JL, French FS, Wilson EM (2001) A mechanism for androgen receptor-mediated prostate cancer recurrence after androgen deprivation therapy. Cancer Res 61:4315–4319 Gregory CW, Fei X, Ponguta LA, He B, Bill HM, French FS, Wilson EM (2004) Epidermal growth factor increases coactivation of the androgen receptor in recurrent prostate cancer. J Biol Chem 279:7119–7130 Grossmann ME, Huang H, Tindall DJ (2001) Androgen receptor signaling in androgen-refractory prostate cancer. J Natl Cancer Inst 93:1687–1697 Guo Z, Dai B, Jiang T, Xu K, Xie Y, Kim O, Nesheiwat I, Kong X, Melamed J, Handratta VD, Njar VC, Brodie AM, Yu LR, Veenstra TD, Chen H, Qiu Y (2006) Regulation of androgen receptor activity by tyrosine phosphorylation. Cancer Cell 10:309–319 Han G, Foster BA, Mistry S, Buchanan G, Harris JM, Tilley WD, Greenberg NM (2001) Hormone status selects for spontaneous somatic androgen receptor variants that demonstrate specific ligand and cofactor dependent activities in autochthonous prostate cancer. J Biol Chem 276:11204–11213 Han G, Buchanan G, Ittmann M, Harris JM, Yu X, DeMayo FJ, Tilley W, Greenberg NM (2005) Mutation of the androgen receptor causes oncogenic transformation of the prostate. Proc Natl Acad Sci U S A 102:1151–1156 He B, Wilson EM (2003) Electrostatic modulation in steroid receptor recruitment of LXXLL and FXXLF motifs. Mol Cell Biol 23:2135–2150 He B, Kemppainen JA, Wilson EM (2000) FXXLF and WXXLF sequences mediate the NH2terminal interaction with the ligand binding domain of the androgen receptor. J Biol Chem 275:22986–22994 He B, Bowen NT, Minges JT, Wilson EM (2001) Androgen-induced NH2- and COOH-terminal interaction inhibits p160 coactivator recruitment by activation function 2. J Biol Chem 276:42293–42301 He B, Bai S, Hnat AT, Kalman RI, Minges JT, Patterson C, Wilson EM (2004a) An androgen receptor NH2-terminal conserved motif interacts with the COOH terminus of the Hsp70interacting protein (CHIP). J Biol Chem 279:30643–30653 He B, Gampe RT, Jr, Kole AJ, Hnat AT, Stanley TB, An G, Stewart EL, Kalman RI, Minges JT, Wilson EM (2004b) Structural basis for androgen receptor interdomain and coactivator interactions suggests a transition in nuclear receptor activation function dominance. Mol Cell 16:425–438 He B, Gampe RT, Jr, Hnat AT, Faggart JL, Minges JT, French FS, Wilson EM (2006) Probing the functional link between androgen receptor coactivator and ligand-binding sites in prostate cancer and androgen insensitivity. J Biol Chem 281:6648–6663 Heath EI, Gaskins M, Pitot HC, Pili R, Tan W, Marschke R, Liu G, Hillman D, Sarkar F, Sheng S, Erlichman C, Ivy P (2005) A phase II trial of 17-allylamino-17- demethoxygeldanamycin in patients with hormone-refractory metastatic prostate cancer. Clin Prostate Cancer 4:138–141 Hobisch A, Culig Z, Radmayr C, Bartsch G, Klocker H, Hittmair A (1995) Distant metastases from prostatic carcinoma express androgen receptor protein. Cancer Res 55:3068–3072 Hsing AW, Gao YT, Wu G, Wang X, Deng J, Chen YL, Sesterhenn IA, Mostofi FK, Benichou J, Chang C (2000) Polymorphic CAG and GGN repeat lengths in the androgen receptor gene and prostate cancer risk: a population-based case-control study in China. Cancer Res 60:5111– 5116 Hyytinen ER, Haapala K, Thompson J, Lappalainen I, Roiha M, Rantala I, Helin HJ, Janne OA, Vihinen M, Palvimo JJ, Koivisto PA (2002) Pattern of somatic androgen receptor gene mutations in patients with hormone-refractory prostate cancer. Lab Invest 82:1591–1598

236

G. Buchanan et al.

Ikonen T, Palvimo JJ, Janne OA (1997) Interaction between the amino- and carboxyl-terminal regions of the rat androgen receptor modulates transcriptional activity and is influenced by nuclear receptor coactivators. J Biol Chem 272:29821–29828 Irvine RA, Yu MC, Ross RK, Coetzee GA (1995) The CAG and GGC microsatellites of the androgen receptor gene are in linkage disequilibrium in men with prostate cancer. Cancer Res 55:1937–1940 Jenster G, van der Korput HA, van Vroonhoven C, Van Der Kwast TH, Trapman J, Brinkmann AO (1991) Domains of the human androgen receptor involved in steroid binding, transcriptional activation, and subcellular localization. Mol Endocrinol 5:1396–1404 Jenster G, van der Korput HA, Trapman J, Brinkmann AO (1995) Identification of two transcription activation units in the N-terminal domain of the human androgen receptor. J Biol Chem 270:7341–7346 Jin B, Beilin J, Zajac J, Handelsman DJ (2000) Androgen receptor gene polymorphism and prostate zonal volumes in Australian and Chinese men. J Androl 21:91–98 Kamal A, Thao L, Sensintaffar J, Zhang L, Boehm MF, Fritz LC, Burrows FJ (2003) A highaffinity conformation of Hsp90 confers tumour selectivity on Hsp90 inhibitors. Nature 425:407–410 Kelly WK, Slovin S, Scher HI (1997) Steroid hormone withdrawal syndromes. Pathophysiology and clinical significance. Urol Clin North Am 24:421–431 Kelly WK, Richon VM, O’Connor O, Curley T, MacGregor-Curtelli B, Tong W, Klang M, Schwartz L, Richardson S, Rosa E, Drobnjak M, Cordon-Cordo C, Chiao JH, Rifkind R, Marks PA, Scher H (2003) Phase I clinical trial of histone deacetylase inhibitor: suberoylanilide hydroxamic acid administered intravenously. Clin Cancer Res 9:3578–3588 Kemppainen JA, Langley E, Wong CI, Bobseine K, Kelce WR, Wilson EM (1999) Distinguishing androgen receptor agonists and antagonists: distinct mechanisms of activation by medroxyprogesterone acetate and dihydrotestosterone. Mol Endocrinol 13:440–454 Kneller DG, Cohen FE, Langridge R (1990) Improvements in protein secondary structure prediction by an enhanced neural network. J Mol Biol 214:171–182 Ko YJ, Devi GR, London CA, Kayas A, Reddy MT, Iversen PL, Bubley GJ, Balk SP (2004) Androgen receptor down-regulation in prostate cancer with phosphorodiamidate morpholino antisense oligomers. J Urol 172:1140–1144 Kumar R, Betney R, Li J, Thompson EB, McEwan IJ (2004) Induced alpha-helix structure in AF1 of the androgen receptor upon binding transcription factor TFIIF. Biochemistry 43:3008–3013 Kuratsukuri K, Sugimura K, Harimoto K, Kawashima H, Kishimoto T (1999) ‘‘Decoy’’ of androgen-responsive element induces apoptosis in LNCaP cells. Prostate 41:121–126 Lai J, Kedda MA, Hinze K, Smith RL, Yaxley J, Spurdle AB, Morris CP, Harris J, Clements JA (2007) PSA/KLK3 AREI promoter polymorphism alters androgen receptor binding and is associated with prostate cancer susceptibility. Carcinogenesis 28:1032–1039 Lattouf JB, Srinivasan R, Pinto PA, Linehan WM, Neckers L (2006) Mechanisms of disease: the role of heat-shock protein 90 in genitourinary malignancy. Nat Clin Pract Urol 3:590–601 Lavery DN, McEwan IJ (2008a) Functional characterization of the native NH2-terminal transactivation domain of the human androgen receptor: binding kinetics for interactions with TFIIF and SRC-1a. Biochemistry 47:3352–3359 Lavery DN, McEwan IJ (2008b) Structural characterization of the native NH2-terminal transactivation domain of the human androgen receptor: a collapsed disordered conformation underlies structural plasticity and protein-induced folding. Biochemistry 47:3360–3369 Li J, Fu J, Toumazou C, Yoon HG, Wong J (2006) A role of the amino-terminal (N) and carboxylterminal (C) interaction in binding of androgen receptor to chromatin. Mol Endocrinol 20:776– 785 Lin HK, Yeh S, Kang HY, Chang C (2001) Akt suppresses androgen-induced apoptosis by phosphorylating and inhibiting androgen receptor. Proc Natl Acad Sci U S A 98:7200–7205

Insights from AR Gene Mutations

237

Ma H, Hong H, Huang SM, Irvine RA, Webb P, Kushner PJ, Coetzee GA, Stallcup MR (1999) Multiple signal input and output domains of the 160-kilodalton nuclear receptor coactivator proteins. Mol Cell Biol 19:6164–6173 Mahajan NP, Liu Y, Majumder S, Warren MR, Parker CE, Mohler JL, Earp HS, Whang YE (2007) Activated Cdc42-associated kinase Ack1 promotes prostate cancer progression via androgen receptor tyrosine phosphorylation. Proc Natl Acad Sci U S A 104:8438–8443 Marcelli M, Ittmann M, Mariani S, Sutherland R, Nigam R, Murthy L, Zhao Y, DiConcini D, Puxeddu E, Esen A, Eastham J, Weigel NL, Lamb DJ (2000) Androgen receptor mutations in prostate cancer. Cancer Res 60:944–949 Marks P, Rifkind RA, Richon VM, Breslow R, Miller T, Kelly WK (2001) Histone deacetylases and cancer: causes and therapies. Nat Rev Cancer 1:194–202 Marrocco DL, Tilley WD, Bianco-Miotto T, Evdokiou A, Scher HI, Rifkind RA, Marks PA, Richon VM, Butler LM (2007) Suberoylanilide hydroxamic acid (vorinostat) represses androgen receptor expression and acts synergistically with an androgen receptor antagonist to inhibit prostate cancer cell proliferation. Mol Cancer Ther 6:51–60 Matias PM, Donner P, Coelho R, Thomaz M, Peixoto C, Macedo S, Otto N, Joschko S, Scholz P, Wegg A, Basler S, Schafer M, Egner U, Carrondo MA (2000) Structural evidence for ligand specificity in the binding domain of the human androgen receptor: implications for pathogenic gene mutations. J Biol Chem 275:26164–26171 Moilanen AM, Poukka H, Karvonen U, Hakli M, Janne OA, Palvimo JJ (1998) Identification of a novel RING finger protein as a coregulator in steroid receptor-mediated gene transcription. Mol Cell Biol 18:5128–5139 Moilanen AM, Karvonen U, Poukka H, Yan W, Toppari J, Janne OA, Palvimo JJ (1999) A testisspecific androgen receptor coregulator that belongs to a novel family of nuclear proteins. J Biol Chem 274:3700–3704 Monge A, Jagla M, Lapouge G, Sasorith S, Cruchant M, Wurtz JM, Jacqmin D, Bergerat JP, Ceraline J (2006) Unfaithfulness and promiscuity of a mutant androgen receptor in a hormonerefractory prostate cancer. Cell Mol Life Sci 63:487–497 Nazareth LV, Weigel NL (1996) Activation of the human androgen receptor through a protein kinase A signaling pathway. J Biol Chem 271:19900–19907 Nickols NG, Dervan PB (2007) Suppression of androgen receptor-mediated gene expression by a sequence-specific DNA-binding polyamide. Proc Natl Acad Sci U S A 104:10418–10423 Palazzolo I, Gliozzi A, Rusmini P, Sau D, Crippa V, Simonini F, Onesto E, Bolzoni E, Poletti A (2008) The role of the polyglutamine tract in androgen receptor. J Steroid Biochem Mol Biol 108:245–253 Palvimo JJ, Kallio PJ, Ikonen T, Mehto M, Janne OA (1993) Dominant negative regulation of trans-activation by the rat androgen receptor: roles of the N-terminal domain and heterodimer formation. Mol Endocrinol 7:1399–1407 Platz EA, Giovannucci E, Dahl DM, Krithivas K, Hennekens CH, Brown M, Stampfer MJ, Kantoff PW (1998) The androgen receptor gene GGN microsatellite and prostate cancer risk. Cancer Epidemiol Biomarkers Prev 7:379–384 Platz EA, Rimm EB, Willett WC, Kantoff PW, Giovannucci E (2000) Racial variation in prostate cancer incidence and in hormonal system markers among male health professionals. J Natl Cancer Inst 92:2009–2017 Poujol N, Lobaccaro JM, Chiche L, Lumbroso S, Sultan C (1997) Functional and structural analysis of R607Q and R608K androgen receptor substitutions associated with male breast cancer. Mol Cell Endocrinol 130:43–51 Powers MV, Workman P (2006) Targeting of multiple signalling pathways by heat shock protein 90 molecular chaperone inhibitors. Endocr Relat Cancer 13 Suppl 1:S125–S135 Pratt WB, Toft DO (1997) Steroid receptor interactions with heat shock protein and immunophilin chaperones. Endocr Rev 18:306–360 Pratt WB, Galigniana MD, Harrell JM, DeFranco DB (2004) Role of hsp90 and the hsp90-binding immunophilins in signalling protein movement. Cell Signal 16:857–872

238

G. Buchanan et al.

Quarmby VE, Beckman WCJ, Cooke DB, Lubahn DB, Joseph DR, Wilson EM, French FS (1990) Expression and localization of androgen receptor in the R-3327 Dunning rat prostatic adenocarcinoma. Cancer Res 50:735–739 Quigley CA, De Bellis A, Marschke KB, el-Awady MK, Wilson EM, French FS (1995) Androgen receptor defects: historical, clinical, and molecular perspectives. Endocr Rev 16:271–321 Radpour R, Rezaee M, Tavasoly A, Solati S, Saleki A (2007) Association of long polyglycine tracts (GGN repeats) in exon 1 of the androgen receptor gene with cryptorchidism and penile hypospadias in Iranian patients. J Androl 28:164–169 Reid J, Kelly SM, Watt K, Price NC, McEwan IJ (2002) Conformational analysis of the androgen receptor amino-terminal domain involved in transactivation. Influence of structure-stabilizing solutes and protein–protein interactions. J Biol Chem 277:20079–20086 Reid J, Betney R, Watt K, McEwan IJ (2003) The androgen receptor transactivation domain: the interplay between protein conformation and protein–protein interactions. Biochem Soc Trans 31:1042–1046 Ricciardelli C, Jackson MW, Choong CS, Stahl J, Marshall VR, Horsfall DJ, Tilley WD (2008) Elevated levels of HER-2/neu and androgen receptor in clinically localized prostate cancer identifies metastatic potential. Prostate 68:830–838 Rodriguez G, Bilbao C, Ramirez R, Falcon O, Leon L, Chirino R, Falcon O, Jr., Diaz BP, Rivero JF, Perucho M, Diaz-Chico BN, Diaz-Chico JC (2006) Alleles with short CAG and GGN repeats in the androgen receptor gene are associated with benign endometrial cancer. Int J Cancer 118:1420–1425 Rowan BG, Garrison N, Weigel NL, O’Malley BW (2000a) 8-Bromo-cyclic AMP induces phosphorylation of two sites in SRC-1 that facilitate ligand-independent activation of the chicken progesterone receptor and are critical for functional cooperation between SRC-1 and CREB binding protein. Mol Cell Biol 20:8720–8730 Rowan BG, Weigel NL, O’Malley BW (2000b) Phosphorylation of steroid receptor coactivator-1. Identification of the phosphorylation sites and phosphorylation through the mitogen-activated protein kinase pathway. J Biol Chem 275:4475–4483 Sack JS, Kish KF, Wang C, Attar RM, Kiefer SE, An Y, Wu GY, Scheffler JE, Salvati ME, Krystek SR, Weinmann R, Einspahr HM (2001) Crystallographic structures of the ligandbinding domains of the androgen receptor and its T877A mutant complexed with the natural agonist dihydrotestosterone. Proc Natl Acad Sci U S A 98:4904–4909 Sanchez D, Rosell D, Honorato B, Lopez J, Arocena J, Sanz G (2006) Androgen receptor mutations are associated with Gleason score in localized prostate cancer. BJU Int 98:1320– 1325 Saporita AJ, Ai J, Wang Z (2007) The Hsp90 inhibitor, 17-AAG, prevents the ligand-independent nuclear localization of androgen receptor in refractory prostate cancer cells. Prostate 67:509– 520 Sartor O, Zheng Q, Eastham JA (1999) Androgen receptor gene CAG repeat length varies in a race-specific fashion in men without prostate cancer. Urology 53:378–380 Scher HI, Buchanan G, Gerald W, Butler LM, Tilley WD (2004) Targeting the androgen receptor: improving outcomes for castration-resistant prostate cancer. Endocr Relat Cancer 11:459–476 Schoenberg MP, Hakimi JM, Wang S, Bova GS, Epstein JI, Fischbeck KH, Isaacs WB, Walsh PC, Barrack ER (1994) Microsatellite mutation (CAG24–>18) in the androgen receptor gene in human prostate cancer. Biochem Biophys Res Commun 198:74–80 Schulte TW, Akinaga S, Soga S, Sullivan W, Stensgard B, Toft D, Neckers LM (1998) Antibiotic radicicol binds to the N-terminal domain of Hsp90 and shares important biologic activities with geldanamycin. Cell Stress Chaperones 3:100–108 Segnitz B, Gehring U (1997) The function of steroid hormone receptors is inhibited by the hsp90specific compound geldanamycin. J Biol Chem 272:18694–18701 Shen HC, Buchanan G, Butler LM, Prescott J, Henderson M, Tilley WD, Coetzee GA (2005) GRIP1 mediates the interaction between the amino- and carboxyl-termini of the androgen receptor. Biol Chem 386:69–74

Insights from AR Gene Mutations

239

Solit DB, Zheng FF, Drobnjak M, Munster PN, Higgins B, Verbel D, Heller G, Tong W, CordonCardo C, Agus DB, Scher HI, Rosen N (2002) 17-Allylamino-17-demethoxygeldanamycin induces the degradation of androgen receptor and HER-2/neu and inhibits the growth of prostate cancer xenografts. Clin Cancer Res 8:986–993 Solit DB, Scher HI, Rosen N (2003) Hsp90 as a therapeutic target in prostate cancer. Semin Oncol 30:709–716 Spurdle AB, Dite GS, Chen X, Mayne CJ, Southey MC, Batten LE, Chy H, Trute L, McCredie MR, Giles GG, Armes J, Venter DJ, Hopper JL, Chenevix-Trench G (1999) Androgen receptor exon 1 CAG repeat length and breast cancer in women before age forty years. J Natl Cancer Inst 91:961–966 Stanford JL, Just JJ, Gibbs M, Wicklund KG, Neal CL, Blumenstein BA, Ostrander EA (1997) Polymorphic repeats in the androgen receptor gene: molecular markers of prostate cancer risk. Cancer Res 57:1194–1198 Takayama K, Kaneshiro K, Tsutsumi S, Horie-Inoue K, Ikeda K, Urano T, Ijichi N, Ouchi Y, Shirahige K, Aburatani H, Inoue S (2007) Identification of novel androgen response genes in prostate cancer cells by coupling chromatin immunoprecipitation and genomic microarray analysis. Oncogene 26:4453–4463 Tepper CG, Boucher DL, Ryan PE, Ma AH, Xia L, Lee LF, Pretlow TG, Kung HJ (2002) Characterization of a novel androgen receptor mutation in a relapsed CWR22 prostate cancer xenograft and cell line. Cancer Res 62:6606–6614 Thornton JW, Kelley DB (1998) Evolution of the androgen receptor: structure-function implications. Bioessays 20:860–869 Tilley WD, Wilson CM, Marcelli M, McPhaul MJ (1990) Androgen receptor gene expression in human prostate carcinoma cell lines. Cancer Res 50:5382–5386 Tilley WD, Buchanan G, Hickey TE, Bentel JM (1996) Mutations in the androgen receptor gene are associated with progression of human prostate cancer to androgen independence. Clin Cancer Res 2:277–285 Ueda T, Bruchovsky N, Sadar MD (2002a) Activation of the androgen receptor N-terminal domain by interleukin-6 via MAPK and STAT3 signal transduction pathways. J Biol Chem 277:7076– 7085 Ueda T, Mawji NR, Bruchovsky N, Sadar MD (2002b) Ligand-independent activation of the androgen receptor by interleukin-6 and the role of steroid receptor coactivator-1 in prostate cancer cells. J Biol Chem 277:38087–38094 van Royen ME, Cunha SM, Brink MC, Mattern KA, Nigg AL, Dubbink HJ, Verschure PJ, Trapman J, Houtsmuller AB (2007) Compartmentalization of androgen receptor protein– protein interactions in living cells. J Cell Biol 177:63–72 Veldscholte J, Berrevoets CA, Brinkmann AO, Grootegoed JA, Mulder E (1992) Anti-androgens and the mutated androgen receptor of LNCaP cells: differential effects on binding affinity, heat-shock protein interaction, and transcription activation. Biochemistry 31:2393–2399 Wallen MJ, Linja M, Kaartinen K, Schleutker J, Visakorpi T (1999) Androgen receptor gene mutations in hormone-refractory prostate cancer. J Pathol 189:559–563 Wang Q, Li W, Liu XS, Carroll JS, Janne OA, Keeton EK, Chinnaiyan AM, Pienta KJ, Brown M (2007) A hierarchical network of transcription factors governs androgen receptor-dependent prostate cancer growth. Mol Cell 27:380–392 Watanabe M, Ushijima T, Shiraishi T, Yatani R, Shimazaki J, Kotake T, Sugimura T, Nagao M (1997) Genetic alterations of androgen receptor gene in Japanese human prostate cancer. Jpn J Clin Oncol 27:389–393 Waza M, Adachi H, Katsuno M, Minamiyama M, Sang C, Tanaka F, Inukai A, Doyu M, Sobue G (2005) 17-AAG, an Hsp90 inhibitor, ameliorates polyglutamine-mediated motor neuron degeneration. Nat Med 11:1088–1095 Weigel NL, Moore NL (2007) Steroid receptor phosphorylation: a key modulator of multiple receptor functions. Mol Endocrinol 21:2311–2319

240

G. Buchanan et al.

Wen Y, Hu MC, Makino K, Spohn B, Bartholomeusz G, Yan DH, Hung MC (2000) HER-2/neu promotes androgen-independent survival and growth of prostate cancer cells through the Akt pathway. Cancer Res 60:6841–6845 Werner R, Holterhus PM, Binder G, Schwarz HP, Morlot M, Struve D, Marschke C, Hiort O (2006) The A645D mutation in the hinge region of the human AR gene modulates AR activity depending on the context of the polymorphic glutamine and glycine repeats. J Clin Endocrinol Metab 101:1–10 Whitesell L, Lindquist SL (2005) HSP90 and the chaperoning of cancer. Nat Rev Cancer 5:761– 772 Wurtz JM, Bourguet W, Renaud JP, Vivat V, Chambon P, Moras D, Gronemeyer H (1996) A canonical structure for the ligand-binding domain of nuclear receptors. Nat Struct Biol 3:206 Yeh S, Lin HK, Kang HY, Thin TH, Lin MF, Chang C (1999) From HER2/Neu signal cascade to androgen receptor and its coactivators: a novel pathway by induction of androgen target genes through MAP kinase in prostate cancer cells. Proc Natl Acad Sci U S A 96:5458–5463 Yong W, Yang Z, Periyasamy S, Chen H, Yucel S, Li W, Lin LY, Wolf IM, Cohn MJ, Baskin LS, Sanchez ER, Shou W (2007) Essential role for co-chaperone Fkbp52 but not Fkbp51 in androgen receptor-mediated signaling and physiology. J Biol Chem 282:5026–5036 Zegarra-Moro OL, Schmidt LJ, Huang H, Tindall DJ (2002) Disruption of androgen receptor function inhibits proliferation of androgen-refractory prostate cancer cells. Cancer Res 62:1008–1013 Zoubeidi A, Zardan A, Beraldi E, Fazli L, Sowery R, Rennie P, Nelson C, Gleave M (2007) Cooperative interactions between androgen receptor (AR) and heat-shock protein 27 facilitate AR transcriptional activity. Cancer Res 67:10455–10465

Functional Motifs of the Androgen Receptor Elizabeth M. Wilson

Abstract Androgen receptor (AR) transcriptional activity stimulated by high affinity binding of testosterone or dihydrotestosterone involves a dynamic NH2and carboxyl-terminal (N/C) interaction between an AR NH2-terminal FXXLF motif (23FQNLF27) and a hydrophobic interaction surface in the ligand binding domain known as activation function 2 (AF2). The AR N/C interaction slows the dissociation rate of bound androgen, stabilizes AR and competitively inhibits activity from AF2. Functional significance of the androgen-dependent AR N/C interaction is supported by naturally occurring germline mutations in AF2 that decrease AR transcriptional activity and cause the androgen insensitivity syndrome by interfering with the N/C interaction without altering equilibrium androgen binding affinity. Gain-of-function AR somatic mutations in prostate cancer can enhance the AR N/C interaction in association with increased AR transcriptional activity. Melanoma antigen gene protein-A11 (MAGE-11) is an AR coactivator that competitively inhibits the AR N/C interaction by binding the AR FXXLF motif, but increases AR signaling through mechanisms that appear to contribute to the development and growth of prostate cancer. The androgen receptor (AR) is a ligand-activated nuclear receptor of the steroid receptor subfamily that mediates the developmental and behavioral effects of the biologically active, naturally occurring androgens, testosterone (T) and dihydrotestosterone (DHT). High-affinity androgen binding targets AR to the nucleus where it interacts with androgen response element DNA associated with androgen-regulated genes. The human syndrome of androgen insensitivity, in which 46XY genetic males have a mutation in the AR gene and are phenotypically female, demonstrates the functional requirement for AR to mediate the effects of androgen during normal male sex development. AR transactivation also contributes to prostate cancer development and progression. Structure and function relationships of AR interacting domains are the focus of this review.

E.M. Wilson Laboratories for Reproductive Biology and the Lineberger Comprehensive Cancer Center. Departments of Pediatrics, and the Department of Biochemistry and Biophysics, University of North Carolina at Chapel Hill, NC 27599, USA, E-mail: [email protected]

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4_11, # Springer Science + Business Media, LLC 2009

241

242

E.M. Wilson

AR has a multidomain structure characteristic of the steroid hormone family of nuclear receptors (Simental et al. 1991). The carboxyl-terminal ligand binding domain (LBD) binds T and DHT with similar high affinity (Wilson and French 1976). The central DNA binding domain binds androgen response element DNA associated with androgen-regulated genes. Naturally occurring mutations in the AR DNA binding domain result in the androgen insensitivity syndrome (AIS) and demonstrate that AR must bind DNA to mediate the biological effects of T and DHT (Quigley et al. 1995). A hinge region lies between the DNA binding domain and LBD and contains an AR nuclear-targeting signal (Zhou et al. 1994). The AR NH2-terminal region is the least conserved in AR between species and relative to other steroid receptors. The largely unstructured AR NH2-terminal region harbors a strong activation function 1 (AF1) that facilitates tissue-specific gene activation. AF1 is not influenced directly by ligand binding but depends on bound androgen for AR dimerization, nuclear transport, DNA binding, and transcriptional activation. AR transcriptional activity is also mediated by activation function 2 (AF2) in the LBD. AF2 bears no structural similarity to AF1, but both activation regions are involved in recruiting coactivator proteins that link to the transcriptional machinery of chromatin. Considerable research has focused on the structural and functional properties of AF2. Our understanding of AF2 derives in part from crystallographic studies made possible by the highly ordered arrangement of alpha helices in the LBD. AF2 is a hydrophobic surface of the LBD that lies contiguous to the ligand binding pocket and is structurally dependent on bound ligand (He et al. 2004). Functional studies on AR demonstrate an AF2 binding specificity unique among steroid receptors for a set of structurally related amphipathic alpha helical motifs.

1 AR N/C interaction Early studies on the structure and function of AR led to the identification and characterization of the androgen-dependent AR NH2-terminal and carboxyl-terminal (N/C) interaction (Langley et al. 1995, 1998; Wong et al. 1993; Zhou et al. 1995). The AR N/C interaction requires high-affinity androgen binding in the LBD and is mediated by an interaction between the AR NH2-terminal FXXLF motif sequence 23FQNLF27 in human AR and the AF2 region in the LBD (He et al. 1999, 2000, 2004) (Fig. 1). Original studies that led to the identification of the AR N/C interaction were based on determinants of ligand and DNA binding and AR stabilization.

1.1

Androgen Dissociation Kinetics

Of the two biologically active androgens, T dissociates about three times faster than DHT despite similar apparent equilibrium binding affinities determined from the ratios of association and dissociation rate constants (Wilson and French 1976).

Functional Motifs of the Androgen Receptor

243

Fig. 1 Schematic of the AR FXXLF and WXXLF motifs. Full-length human AR (amino acid residues 1–919) contains the NH2-terminal FXXLF (23FQNLF27) and WXXLF (433WHTLF437) motifs, activation function 1 (AF1) and activation function 2 (AF2), DNA binding domain (DBD), and ligand binding domain (LBD) (He et al. 2001; Wilson 2007). (Copyright 2007, The Endocrine Society)

Consistent with the domain function of steroid receptors, the AR LBD functions relatively independently to bind T and DHT with high affinity. However, early studies demonstrated that deletion of the NH2-terminal region increases the dissociation rate of bound androgen in a manner reminiscent of the faster dissociation of T than DHT (Zhou et al. 1995). The basis for the different dissociation rates of T and DHT lies with the AR N/C interaction and is discussed further (see Sect. 3.2). When an AR fragment containing the DNA-and ligand binding domains is coexpressed with an AR fragment containing the NH2-terminal and DNA binding domains, the N/C interaction causes bound androgen to dissociate more slowly (Langley et al. 1998). These studies provided the first evidence that the AR NH2-terminal region directly influences the kinetics of androgen binding. Small deletions within the AR NH2-terminal region helped to pinpoint the NH2-terminal interacting region to the AR FXXLF motif (He et al. 2000; Zhou et al. 1995). Functional studies of naturally occurring mutations that cause AIS without altering apparent equilibrium androgen binding affinity provided evidence that the FXXLF motif in the AR NH2terminus binds AF2 in the AR LBD and is the basis of the AR N/C interaction (He et al. 1999; Langley et al. 1998). For AR, ligand dissociation rate is directly related to ligand potency as an AR agonist. DHT dissociates more slowly than T, and DHT is the more potent androgen than T. However, T is the predominant androgen produced by Leydig cells of the testis and by theca cells of the ovary and has important physiological effects in muscle and other tissues. Potency differences between T and DHT are related to the inability of T to fully stabilize the AF2 surface for AR FXXLF motif binding, with the result that T dissociates faster than DHT. AR antagonists with moderate to high binding affinity dissociate faster, with dissociation half-times from full-length AR of 5 min for AR antagonist RU56187 (Kemppainen et al. 1999) compared to 1 h for T and 3 h for DHT determined in whole cell cultures at 37 C (Askew et al. 2007; He et al. 2006; Zhou et al. 1995). Dissociation rates of the more classical AR antagonists, hydroxyflutamide and casodex, have not been determined due to their unavailability in radiolabeled form, but are expected to be rapid. AR antagonists bind with only moderate affinity and do not induce the AR N/C interaction in wild-type AR (Langley et al. 1995).

244

1.2

E.M. Wilson

Ligand and DNA Binding

The AR N/C interaction is one of the most androgen-selective properties of AR. While other steroids, including 17b-estradiol, progesterone and the synthetic progestin, medroxyprogesterone acetate, bind AR with moderate to high affinity, cause AR nuclear localization and activate androgen-responsive reporter genes (Kemppainen et al. 1992), none of these ligands induces the AR N/C interaction, and none is a potent androgen in vivo. Indeed, ligand specificity of the AR N/C interaction closely parallels ligand potency in vivo, providing support for its physiological importance. The AR N/C interaction is also induced in response to synthetic steroids that have anabolic activity in vivo (Kemppainen et al. 1999). Anabolic steroids, such as oxandrolone and fluoxymesterone, induce the AR N/C interaction despite lower equilibrium binding affinities. In the presence of androgen, AR binds androgen response element DNA as a homodimer (Wong et al. 1993). However, in the absence of a crystal structure for full-length AR bound to DNA, the precise structural arrangement is not known. Early studies using AR NH2- and carboxyl-terminal fragments provided some evidence that the N/C interaction influences AR binding to DNA. An AR fragment containing the DNA- and ligand binding domains but lacking the NH2-terminal region binds androgen response element DNA in the absence and presence of androgen. Dimerization of this carboxyl-terminal AR fragment presumably occurs through the DNA binding domain dimerization region. Ligand-independent DNA binding by the carboxyl-terminal AR fragment that lacks the NH2-terminal region raises the possibility that under certain physiological or pathological conditions, AR may bind DNA in the absence of androgen. In striking contrast, when the same AR fragment containing the DNA binding domain and LBD is coexpressed with an AR fragment containing the NH2-terminal and DNA binding domains without the LBD, DNA binding by this stable AR dimer complex requires androgen bound in the LBD (Kemppainen et al. 1999; Zhou et al. 1995). The results indicate that the AR N/C interaction influences AR DNA binding specificity and/or affinity.

1.3

AR Stabilization

AR differs from other steroid receptors because it is stabilized against degradation by high-affinity agonist binding, such as T, DHT, and synthetic androgen agonist derivatives such as methyltrienolone (R1881) (Kemppainen et al. 1992; Zhou et al. 1995). Other steroid receptors are typically down-regulated by agonist binding. Agonist-dependent stabilization of AR is another consequence of the androgendependent AR N/C interaction mediated by AR FXXLF motif binding to AF2 that appears to be critical for AR transcriptional activity (He et al. 2001; Langley et al. 1998). The extent to which a ligand stabilizes AR directly correlates with its ability

Functional Motifs of the Androgen Receptor

245

to act as an agonist (Kemppainen and Wilson 1996). As noted earlier, ligands such as estradiol and progesterone, and the antagonist hydroxyflutamide, bind AR with moderate affinity, but do not stabilize AR against degradation due to their inability to induce the AR N/C interaction (Kemppainen et al. 1999). AR appears to be more stable in prostate cancer cell lines in the absence and presence of androgen, and in the CWR22 tumor after recurrence in the relative absence of circulating androgen (Gregory et al. 1998, 2001c). The mechanism for increased AR stabilization in castration-recurrent prostate cancer is not known, but occurs in association with increased AR nuclear localization and may be related to its interaction with coregulatory proteins. One mechanism that may contribute to increased AR stabilization in prostate cancer is the increased expression of the AR coregulator melanoma antigen gene protein-11 (MAGE-11 or MAGE-A11), which stabilizes AR in the absence or at low levels of androgen (see Sect. 5). Alternatively, there appears to be local androgen production in prostate cancer tissue that could stabilize AR (Mohler et al. 2004). On the other hand, local androgen production in prostate cancer cell lines cultured under serum-free conditions does not likely account for AR stabilization because of the lack of a biosynthetic substrate. The mechanisms that underlie increased AR stabilization in prostate cancer remain to be fully established and may be independent of the AR N/C interaction. Reports have suggested both direct and indirect mechanisms are the basis of the AR N/C interaction. Proposed indirect mechanisms involve coactivators such as CBP/p300, SRC1 and c-jun (Bubulya et al. 2001; Ikonen et al. 1997). A direct AR N/C interaction is supported by mammalian two-hybrid assays using GAL4 DNA binding domain and VP16 activation domain fusion proteins of the AR NH2-terminal region and LBD (Langley et al. 1995; Wilson et al. 2003). Since the original report, the interaction between the AR NH2-terminal domain and AF2 has been confirmed by several groups (Berrevoets et al. 1998; Hsu et al. 2005; Ikonen et al. 1997; Li et al. 2006; Schaufele et al. 2005). The biological significance of the AR N/C interaction is supported by its androgen specificity and inhibition by antagonists. Studies on AR AF2 mutations that cause AIS without altering apparent equilibrium androgen binding affinity (Langley et al. 1998), and more recently, experiments using FRET analysis of wild-type AR (Schaufele et al. 2005), support the possibility that the AR N/C interaction results in an antiparallel AR dimer. There is also evidence for intramolecular domain interactions within the AR monomer (Klokk et al. 2007; Schaufele et al. 2005). The likelihood for both models is indicative of domain swapping between the AR FXXLF motif and AF2 during transition from the ligand-bound inactive monomer to the ligand-bound active AR dimer (Bennett et al. 1995). In this self-association model of protein structure, the same interaction interface mediates self-complementation within a closed functional dimer unit arising from opening the monomer and rotation about a hinge loop (Bennett et al. 2006). Because the protein interaction interface is the same for monomer and dimer, only the hinge region conformation is expected to change. Altered structure of the hinge region could provide a new interaction surface that modulates AR transcriptional activity (Askew et al. 2007; Haelens et al. 2007; Wang et al. 2001) and contribute to higher-order aggregate formation (Liao et al.

246

E.M. Wilson

1999). Abnormal AR aggregation in Kennedy’s disease results from expansion of the AR NH2-terminal glutamine repeat at residues 58–78 positioned proximal to the AR FXXLF motif at residues 23–27 (La Spada et al. 1991; Taylor et al. 2003).

2 AR FXXLF and WXXLF Motifs The androgen-dependent AR N/C interaction is mediated by AR NH2-terminal FXXLF and WXXLF motif binding to the AF2 surface of the LBD in the carboxylterminal region of AR (He et al. 2000; He and Wilson 2002). However, it is primarily AR FXXLF motif binding to AF2 that is the basis of the androgendependent AR N/C interaction that slows the androgen dissociation rate and stabilizes AR.

2.1

Motif Binding Preferences

Small deletions within the AR NH2-terminal region that increase the dissociation rate of bound androgen (Zhou et al. 1995) led to the identification of the AR FXXLF motif sequence 23FQNLF27 (He et al. 2000) in the NH2-terminus of full-length AR amino acid residues 1–919 (Lubahn et al. 1988a, b). The AR FXXLF motif is the principal interaction site in the NH2-terminal domain for the AR N/C interaction (He et al. 2000). The AR FXXLF motif, where F is phenylalanine, X is any amino acid, and L is leucine, forms an amphipathic alpha helix when bound to AF2 in the AR LBD. The AR FXXLF motif binds the same AF2 hydrophobic cleft that binds the LXXLL motifs of SRC/p160 coactivators. Fluorescent binding measurements demonstrate binding of the AR-20–30 FXXLF motif peptide is five- to tenfold higher affinity than binding of a peptide containing the third and strongest interacting LXXLL motif of transcriptional intermediary factor 2 (TIF2, SRC2) (Askew et al. 2007; He et al. 2004), a member of the SRC/p160 family of coactivators. Higher binding affinity of the AR FXXLF motif compared to the coactivator LXXLL motifs results from amino acid sequence changes in the AR AF2 binding site during the evolution of AR (He et al. 2004). FXXLF and LXXLL motif binding to AF2 appears to be facilitated by flanking charged residues that are complementary to oppositely charged clusters on the LBD surface bordering AF2 (Darimont et al. 1998). The AF2 charged residues include the so-called charge clamp residues Lys-720 and Glu-897 (K720 and E897, Fig. 2), which correspond to the charge clamp residues first described for estrogen receptor binding of the coactivator signature LXXLL motif (Heery et al. 1997; Henttu et al. 1997; Le Douarin et al. 1996). There are, however, additional residues that make up two charge clusters surrounding AF2 that complement the charge distribution flanking the AR FXXLF motif and SRC/p160 coactivator LXXLL motifs (He and Wilson 2003). A greater percentage of positively charged residues lie NH2-terminal

Functional Motifs of the Androgen Receptor

247

Fig. 2 Charge clusters flanking AF2 in the AR ligand binding domain. Space-filled model of the DHT-bound AR ligand binding domain (Sack et al. 2001) illustrates positively charged residues Arg-726 (R726), Lys-717 (K717), Lys-720 (K720) (in red ), and negatively charged residues Glu709, Glu-893, and Glu-897 (E709, E893 and E897) (in blue) surrounding the AF2 hydrophobic floor (in yellow). Charge clamp residues are K720 and E897. AF2 charge clusters are complementary to the charge distribution flanking the FXXLF and LXXLL motifs, suggesting electrostatic forces facilitate orientation of motif binding to AF2 (He and Wilson 2003) (See Color Insert)

to the FXXLF and LXXLL motifs, and a greater proportion of negative charges are positioned carboxyl-terminal to the binding motifs. The complementary nature of the charge distribution suggests that electrostatic interactions orient FXXLF and LXXLL motif binding to AF2. High-affinity motif binding is subsequently mediated by hydrogen bonding to backbone atoms of charge clamp residues and extensive hydrophobic interactions with the AF2 floor. The characteristic charge polarity of both the FXXLF and LXXLL motifs may be a useful property in predicting new potential interacting motifs. On the other hand, in mammalian two-hybrid assays and crystallization studies, the minimal AR-20–30 region that optimally binds AF2 contains positively charged Arg-20, important for the AR N/C interaction (Steketee et al. 2002), but lacks the carboxyl-flanking negative charged residues. Superimposition of crystal structures of the AR LBD bound to R1881 (He et al. 2004) and T (Askew et al. 2007) indicates an overlapping, parallel alignment of the AR FXXLF motif relative to the coactivator LXXLL motif within the same hydrophobic groove of AF2. Some studies suggest a deeper hydrophobic groove in

248

E.M. Wilson

association with FXXLF motif binding compared to the bound coactivator LXXLL motif (Dubbink et al. 2004; Hur et al. 2004). Crystal structures of the AR FXXLF motif bound in the presence of R1881 suggest an induced fit movement of AF2 residue side chains while the LBD core remains relatively static (He et al. 2004). The AR FXXLF motif peptide forms hydrogen bonds to AF2 charge clamp residues Lys-720 and Glu-897. The SRC/p160 coactivator TIF2 LXXLL motif is phase ˚ away from Glu-897 toward Lys-720, placing it too distant to shifted by 2 A hydrogen bond to Glu-897, but maintaining hydrogen bonds to Lys-720. The results are in agreement with the demonstrated requirement for Glu-897 in the AR N/C interaction but not in SRC/p160 coactivator binding (Slagsvold et al. 2000). Loss of hydrogen bonding to Glu-897 likely contributes to the lower AF2 binding affinity for the LXXLL than FXXLF motif. AR preference for FXXLF over the LXXLL motif is supported by several lines of evidence. Peptide display screens using the AR LBD as bait have detected a variety of FXXLF-related motifs, but few if any LXXLL motif-containing peptides (Chang et al. 2005; Hsu et al. 2003; van de Wijngaart et al. 2006). A consequence of higheraffinity AR FXXLF motif binding is that the AR N/C interaction competitively inhibits SRC/p160 coactivator LXXLL motif binding to AF2 (He et al. 2001). Inhibition of AF2 recruitment of SRC/p160 coactivators by the AR N/C interaction appears to be one mechanism that shifts the predominant activation function of AR away from AF2 in the LBD to AF1 in the NH2-terminal region (He et al. 2004). Identification of FXXLF motifs in several putative AR coactivators in yeast twohybrid screens using the AR LBD as bait suggests that these coactivators regulate AR transcriptional activity by binding AF2 (He et al. 2002b; Hsu et al. 2003). The biological significance of FXXLF motif-containing AR coactivators is currently under investigation. The AR NH2-terminal WXXLF motif, where W is tryptophan, also interacts with AR AF2 in an androgen-dependent manner, but with lower affinity than the FXXLF motif (He et al. 2000, 2002a). The functional significance of the WXXLF motif remains to be fully established. Some recent studies suggest that the AR WXXLF motif contributes to ligand-independent AR transcriptional activity in castration-recurrent prostate cancer cells (Dehm et al. 2007). When expressed in multiple copies, but not as a single sequence as a GAL-4 fusion protein, the AR WXXLF motif region was transcriptionally active. The competitive relationship between AR FXXLF and coactivator LXXLL motif binding to AF2 raises the possibility that alternative forms of AR that lack the NH2terminal FXXLF motif region may be transcriptionally more active than full-length AR by providing greater access by AF2 for coactivator recruitment. This would be reminiscent of progesterone receptors A and B, where A lacks an extended LXXLL motif-related sequence (Tung et al. 2006). While an intriguing hypothesis, it is not strongly supported because of the questionable physiological relevance of shorter forms of AR. Previous studies have shown that AR is especially susceptible to proteolytic cleavage during isolation and storage (Wilson and French 1979). Shorter forms of AR that lack part of the AR NH2-terminal region (Gao and McPhaul 1998) can result from proteolytic degradation during isolation (Gregory et al. 2001a).

Functional Motifs of the Androgen Receptor

249

It therefore seems unlikely that endogenous alternative forms of AR lack the AR NH2-terminal FXXLF motif region. On the other hand, mutagenesis studies have demonstrated that removing or mutating the AR FXXLF motif in the AR NH2terminal region increases accessibility of AR AF2 for coactivator recruitment and can increase AR transactivation when SRC/p160 coactivators are overexpressed (He et al. 2001). As discussed further (see Sect. 5), the increase in AR transcriptional activity caused by a mutation of the AR FXXLF motif led to the prediction and subsequent identification of the AR coregulator, MAGE-11. MAGE-11 binds the AR FXXLF motif, allowing AR to escape competitive inhibition by FXXLF motif binding to AF2, with the result that AR transcriptional activity is increased by the SRC/p160 coactivators. MAGE-11 binding to AR appears to be an important mechanism to increase AR transcriptional activity in normal cells of the male and female reproductive tract and in prostate cancer.

2.2

Sequence Conservation Through Evolution

The AR amino acid sequence is relatively well conserved through evolution, with greatest sequence homology within the highly structured DNA and ligand binding domains (Choong et al. 1998). However, relative to other steroid receptors, a greater number of sequence changes occurred through evolution in the AR AF2 site that has resulted in increased binding specificity and affinity for the FXXLF motif (He et al. 2004). The AR NH2-terminal domain is the least conserved of the AR domains across species and relative to other steroid receptors. One notable region in the AR NH2-teminal domain is the polymorphic glutamine repeat at residues 58–78 (Lubahn et al. 1988a). The AR glutamine repeat varies in length in the normal human population from 7 to 34 residues, with a mean of 21 residues (Choong and Wilson 1998; Edwards et al. 1992). Slight expansions in AR glutamine repeat length have been linked to infertility (Casella et al. 2003; Mifsud et al. 2001) and genital abnormalities of otherwise unknown etiology (Lim et al. 2001). Glutamine repeat expansions greater than 40 residues cause spinal bulbar muscular atrophy, also known as Kennedy’s disease (La Spada et al. 1991). Contraction of the glutamine repeat has been correlated with prostate cancer (Linja and Visakorpi 2004; Schoenberg et al. 1994). The precise functional significance of the AR glutamine repeat remains unclear, but its length appears to influence AR levels. Longer CAG repeat lengths are associated with reduced AR levels, and shorter repeats with higher levels of AR (Choong et al. 1996). Glutamine repeat length increases during primate evolution, with longest repeats found in chimpanzees and humans (Choong et al. 1998). The polymorphic human AR glutamine repeat (residues 58–78) is located relatively close to the FXXLF motif (residues 23–27) (Lubahn et al. 1988a). However, glutamine repeat length does not appear to directly influence the AR N/C interaction (Langley et al. 1995). The human AR FXXLF motif sequence FQNLF is conserved among mammals but degenerates in fish (He et al. 2002a).

250

E.M. Wilson

Alternative FXXLF motif sequences in fish AR are FQNVF in rainbow trout (Takeo and Yamashita 1999), YQNVF in eel (Todo et al. 1999), and YQSVF in goldfish (AY090897), indicating conservation of key hydrophobic residues in the i + 1, 4 and 5 positions of the motif. This raises the probability that the AR N/C interaction contributes to AR function throughout the vertebrate lineage and that sequence differences may be complemented by changes in AF2. The WXXLF motif sequence WHTLF at human AR NH2-terminal residues 433–437 is also conserved among mammals through Xenopus but degenerates in fish to a greater extent than the FXXLF motif. The less well-conserved WXXLF motif may reflect its minor role in the AR N/C interaction (He et al. 2000). On the other hand, conservation of the WXXLF motif through mammals suggests it serves an additional function important in AR transcriptional activity that remains to be fully established.

3 Functional Effects of AR Motif Interactions A number of assays have provided evidence that the androgen-dependent AR N/C interaction is important for AR function in vivo. Modulation of AR transcriptional activity by the AR N/C interaction is complex and involves high-affinity androgen binding and multiple interacting motifs, activation functions, and coregulatory proteins.

3.1

Regulation of AR Transactivation

Promoter-specific effects of the AR N/C interaction are widely recognized, but the basis for these differences is not fully understood. The AR N/C interaction increases to different extents AR transcriptional activation of enhancer/promoter regions in reporter gene assays (Berrevoets et al. 1998). AR transactivation of the prostatespecific antigen (PSA) enhancer and most other natural promoter and enhancer regions is inhibited by AR NH2-terminal mutations that disrupt the N/C interaction (He et al. 2002a). In contrast, AR activation of the mouse mammary tumor virus (MMTV) long terminal repeat enhancer region is largely independent of the AR N/ C interaction. However, studies in which these androgen response regions were incorporated into chromatin indicate that both MMTV and PSA depend on the AR N/C interaction for activation by AR (Li et al. 2006). Furthermore, both the AR NH2- and carboxyl-terminal regions were required for an intranuclear compartmentalization pattern similar to transcriptionally active full-length AR (Saitoh et al. 2002). One possible explanation for reporter gene differences is that the MMTV enhancer comprises a series of stronger androgen response elements compared to PSA and most naturally occurring enhancer regions when assayed in transient

Functional Motifs of the Androgen Receptor

251

transfection assays. This is supported by transfection studies in prostate cancer cell lines where endogenous AR transactivates MMTV but not the PSA reporter gene (Gregory et al. 2004). Like the PSA enhancer, most naturally occurring androgen response elements contain multiple half-site sequences. It has been suggested that the AR N/C interaction enhances cooperativity between weak androgen response elements to increase selective AR gene activation (Scheller et al. 1998). Arrangement in chromatin may dampen the transcriptional response of MMTV making it more dependent on the stabilizing effects of the AR N/C interaction. The in vivo potency differences between T and DHT parallel their abilities to induce the AR N/ C interaction (Askew et al. 2007), providing further support that the N/C interaction is required for AR transcriptional activity in vivo. One of the transcriptional effects of the AR N/C interaction is to limit SRC/p160 coactivator recruitment by AF2 in the LBD. This was demonstrated in studies using full-length AR except with an internal deletion of the AF1 region in the NH2terminal domain (He et al. 2001). In the presence of androgen, the AR FXXLF motif competitively inhibits AF2 recruitment of SRC/p160 coactivator LXXLL motifs, and inhibition is relieved by loss-of-function mutations in the FXXLF motif. SRC/ p160 coactivator activation at AF2 is further limited by sequence changes in AF2 that result in higher-affinity FXXLF motif binding compared to the SRC/p160 coactivator LXXLL motifs. Androgen-dependent AR FXXLF motif inhibition of AF2 coactivator recruitment may provide a dynamic mechanism for AF1 to be the predominant activation domain (He et al. 2004) and for coordinate androgendependent gene regulation by AF1 and AF2. Discrepancies exist in the literature regarding the role of the AR NH2-terminal region in the recruitment of the SRC/p160 coactivators. Most reports concur that the AR NH2-terminal region interacts with members of the SRC/p160 coactivator family (Bevan et al. 1999; He et al. 1999). SRC/p160 coactivators are reported to interact with AR in a manner that is synergistic (Bevan et al. 1999), independent (Berrevoets et al. 1998), and dependent (He et al. 2002a) on LXXLL motif binding to AF2. Some of these differences may be attributed to promoter-selective effects for activation by AR. Our findings continue to support a dependence on SRC/p160 coactivator LXXLL motif binding to AF2 to mediate transcriptional effects of AR. The AR N/C interaction nevertheless contributes to coactivator interactions possibly through the formation of new interaction surfaces to increase AR transcriptional activity (Alen et al. 1999). The evolutionary diversity of the NH2-terminal AF1 region provides a variety of potential interaction sites for tissueand species-selective gene activation. In contrast, AF2 is highly constrained by the structural requirements of high-affinity androgen binding. Other steroid receptors such as the estrogen receptor lack an extended NH2-terminal AF1 region and are more dependent on AF2 for transcriptional activity. In prostate cancer, the AR AF2 region appears to gain activity in part through increased expression of SRC/p160 coactivators (Gregory et al. 2001b). Higher levels of SRC/p160 coactivators increase the activity of wild-type AR in the absence and presence of the AR N/C interaction (He et al. 2002a). Overexpression of FXXLF motif peptides was reported to suppress the growth of the recurrent

252

E.M. Wilson

CWR22 tumor xenograft (Hsu et al. 2005) further implicating the AF2 site in ARmediated prostate cancer cell proliferation. The AR coregulator MAGE-11 increases AR transcriptional activity in normal reproductive tissues and in prostate cancer by increasing exposure of AF2 as discussed further (see Sect. 5).

3.2

Potency Differences Between T and DHT

Fundamental to understanding the molecular basis of AR activation of gene transcription is to establish the structural and functional consequences of the AR N/C interaction. Original studies to characterize AR FXXLF motif binding to AF2 were based on changes in androgen dissociation rate (He et al. 2000; Zhou et al. 1995). While the AR N/C interaction slows the dissociation rate of high-affinity agonists that stabilize AR, SRC/p160 coactivator LXXLL motif binding has only a minimal effect on androgen dissociation rate even when the LXXLL motif is inserted in the AR NH2-terminal region to replace the FXXLF motif (He et al. 2001). The ineffectiveness of LXXLL motif binding to slow androgen dissociation reflects the lower binding affinity of AF2 for LXXLL motifs compared to the FXXLF motif in AR and AR coactivators (Askew et al. 2007; He et al. 2004). T and DHT bind AR with similar apparent equilibrium binding affinities (Wilson and French 1976), yet T is a less potent androgen in vivo. The weaker inherent activity of T is demonstrated by the 5a-reductase syndrome, in which 46XY genetic males have a female phenotype at birth because of insufficient levels of 5a-reductase, the enzyme that converts T to DHT (Imperato-McGinley et al. 1974). Despite T levels within the normal range, virilization of the newborn male with 5a-reductase deficiency does not occur due to inadequate levels of DHT. Recently, pharmaceutical companies have identified tissue-selective synthetic nonsteroidal AR agonist analogs (Ostrowski et al. 2007; Wilson 2007). This raises the possibility for ligandselective regulation of AR target genes through mechanisms that include ligandspecific coactivator recruitment. However, there is as yet no evidence that T and DHT regulate different genes because higher levels of T elicit the same response as lower concentrations of DHT (Grino et al. 1990; Maes et al. 1979; Mahendroo et al. 2001). Recent biochemical and structure studies investigating the molecular basis for potency differences between T and DHT indicate that T-bound AR exhibits greater conformational heterogeneity at the AF2 surface (Askew et al. 2007). Even slight destabilization of the AF2 surface appears to compromise the AR N/C interaction to an extent that accounts for the faster dissociation rate and reduced in vivo potency of T. T is a 19-carbon steroid, which differs from DHT by a ~4,5 double bond that increases the polarity of T (Fig. 3). Polarizing effects of the A ring double bond in T appear to be transferred through Met-745, a ligand binding pocket residue whose side chain extends toward Leu-712 in the AF2 floor. Crystal structures of the T-bound AR LBD with FXXLF or LXXLL peptide indicate two conformations of Leu-712, whereas similar structures of DHT-bound (Estebanez-Perpina et al. 2005; Hur et al. 2004; Sack et al. 2001) and R1881-bound AR LBD (unpublished results)

Functional Motifs of the Androgen Receptor

253

Fig. 3 Conversion of testosterone to dihydrotestosterone by 5a-reductase. Testosterone (T) and dihydrotestosterone are 19-carbon, biologically active androgens that differ by a ~4,5 double bond in ring A of T (Wilson 2007). (Copyright 2007, The Endocrine Society)

indicate a single conformation of Leu-712. The importance of Leu-712 in AR function and the AR N/C interaction is indicated by the naturally occurring AR-L712F mutation that causes complete AIS without altering high-affinity androgen binding (He et al. 1999, 2006). Functional differences between T and DHT are linked to their effects on AF2 binding of the FXXLF and LXXLL a-helical motif regions. The apparent slight destabilizing effect of bound T on the AF2 motif binding surface is supported by fluorescence polarization measurements. Fluorescein-labeled FXXLF and LXXLL motif peptides bind AF2 of the AR LBD with twofold lower affinity with bound T than DHT (Askew et al. 2007). Furthermore, in the absence of FXXLF motif binding, T and DHT dissociate with similar rates from an AR fragment containing the LBD, whereas T dissociates three times faster than DHT from full-length AR due to the FXXLF motif-mediated N/C interaction. These results indicate that the faster dissociation rate and weaker in vivo potency of T are a consequence of weaker AR FXXLF motif binding to AF2. Thus T is a weaker androgen than DHT in vivo because of the polarizing effects arising from ring A of T that cause local conformational heterogeneity at the AF2 surface. Weaker AR FXXLF motif binding causes faster dissociation of T from fulllength AR (Askew et al. 2007). Because AF2 is also the binding site for SRC/p160 coactivator LXXLL motifs, bound T similarly weakens LXXLL motif binding contributing to its weaker potency. The greater stabilizing effect of the less polar DHT increases AF2 binding affinity for the AR FXXLF and SRC/p160 coactivator LXXLL motifs making DHT a more potent androgen in vivo.

4 Natural AR Gene Mutations Alter Motif Binding Phenotypic expression of naturally occurring AR gene mutations occurs in part because AR function is not required for life (Goh et al. 2007; Quigley et al. 1995). Localization of the AR gene on the X chromosome (Brown et al. 1989) further confounds the potential detrimental effects of AR mutations on male reproductive function and health. In the 46XY genetic male, more than 300 different single amino acid AR mutations result in a range of phenotypes that reflect the degree to

254

E.M. Wilson

which AR function is disrupted (Quigley et al. 1995, 2004). The AIS phenotype spans normal male genital development with infertility, to partial and complete AIS characterized by different degrees of feminization of the external genitalia. Later in life, AR somatic mutations can arise in prostate cancer in association with the increasing genome instability of cancer. The frequency of AR mutations in early stage prostate cancer is low, but increases with progression to castration-recurrent disease after androgen-deprivation therapy (Gelmann 2002; Shi et al. 2002; Taplin et al. 1999). Unlike germline mutations that disrupt AR function and cause AIS, AR somatic mutations in prostate cancer can increase AR transcriptional activity through a gain-in-function. Many of the well-characterized AR mutations in prostate cancer are positioned near the AF2 site in the LBD. Like loss-of-function mutations that cause AIS, somatic AR LBD mutations in prostate cancer positioned near AF2 alter the AR N/C interaction through mechanisms that are independent of changes in equilibrium androgen binding affinity.

4.1

AR AF2 Germline Mutations in the Androgen Insensitivity Syndrome

The many different single amino acid AR mutations that cause AIS (http://ww2. mcgill.ca/androgendb/data.htm) localize predominantly in the highly structured regions of the AR DNA and ligand binding domains (Matias et al. 2000; Quigley et al. 1995). In contrast, the major AR NH2-terminal AF1 activation region is surprisingly free of missense mutations, and a naturally occurring mutation in the AR NH2-terminal FXXLF motif has not been reported. The detrimental effects of most AR LBD mutations result from disruption of equilibrium androgen binding affinity (Quigley et al. 1995). However, there are a few AR LBD mutations that cause partial or complete AIS without disrupting equilibrium androgen binding affinity. These have been of particular interest because they implicate a function for AR not directly related to ligand binding affinity. AR mutations that selectively inhibit AR FXXLF and coactivator LXXLL motif binding were helpful in early studies to characterize the region of the LBD involved in the AR interdomain N/C interaction (He et al. 1999; Langley et al. 1998). It soon became clear that the AF2 site for coactivator LXXLL motif binding was the same site required for the AR N/C interaction. Since then, a number of naturally occurring mutations have been identified that cause AIS by interfering with the AR N/C interaction (Ghali et al. 2003; He et al. 1999; Jaaskelainen et al. 2006; Langley et al. 1998; Quigley et al. 2004; Thompson et al. 2001). A universal finding regarding AF2 mutations that cause AIS is diminished AR FXXLF motif binding and increased dissociation rate of bound androgen. This is consistent with properties of the AR N/C interaction and the ability of the AR FXXLF motif to slow the androgen dissociation rate. The studies provide further evidence that the normally slow androgen dissociation rate from wild-type AR is a requirement for AR function in vivo. The near wild-type equilibrium androgen

Functional Motifs of the Androgen Receptor

255

binding affinities of the AR AF2 mutants, yet faster androgen dissociation rates, are consistent with the properties and potency differences between T and DHT described earlier. Equilibrium binding affinity can be determined from the ratio of association and dissociation rate constants. Thus, fast and slow dissociating ligands can display high-affinity binding if there is a reciprocal change in association rate (Wilson and French 1976). This suggests that AF2 mutations cause more rapid association rates, although this has not been rigorously measured. The results parallel our findings with T and DHT since these biologically active androgens have similar equilibrium binding affinities, but the slower dissociation rate of DHT renders it a more active androgen. When T is the only predominant androgen during male sex development in individuals with the 5a-reductase syndrome, the phenotypic effects are similar to AIS. But unlike AIS, rising T levels at puberty in a 5areductase syndrome patient compensate for the lack of DHT and virilization occurs. Details of the AR AF2 interaction surface for FXXLF and LXXLL motif binding were revealed through biochemical studies of AR mutants that cause AIS and from the crystal structure of the wild-type AR LBD (He et al. 2006). AR AF2 residues Gln-733, Phe-725, and Ile-737 form hydrophobic Cluster A that interacts with Phe-27 of the AR 23FQNLF27 motif (Fig. 4). Cluster A is relatively distant from the ligand binding pocket so that the predominant effect of a Cluster A mutation is decreased FXXLF and LXXLL motif binding with an indirect effect on the bound ligand. Decreased FXXLF motif binding accounts for the faster androgen dissociation rate from the full-length AR mutants, whereas dissociation rates of mutant and wild-type AR fragments containing the LBD were similar. A second AF2 hydrophobic Cluster B lies closer to the ligand binding pocket and is formed by Ile-737, Ile-898, and Leu-712. Cluster B interacts with Phe-23 of the AR 23 FQNLF27 motif. Mutations in Cluster B have direct effects on FXXLF and LXXLL motif binding and on the bound ligand, and cause faster androgen dissociation rates from the full-length AR mutants and from the carboxyl-terminal fragments containing the LBD. AR mutation V889M, which causes complete AIS, results in a very rapid androgen dissociation rate from both full-length AR and the carboxyl-terminal fragment, suggesting that Val-889 functions in a ligand gateway at the base of helices 10, 11, and 12. AF2 mutations that cause partial and complete androgen insensitivity by disrupting FXXLF and LXXLL motif binding to AF2 without altering equilibrium androgen binding affinity indicate that the AF2 site is important for AR function in vivo. However, it is difficult to differentiate the in vivo significance of AF2 in terms of AR FXXLF and SRC/p160 coactivator LXXLL motif binding since the two motifs interact at essentially the same AF2 binding site and are similarly disrupted by AF2 mutations that cause AIS (He et al. 2004, 2006).

4.2

AR Somatic Mutations in Prostate Cancer

AR transcriptional activity increases in prostate cancer through a number of mechanisms that include mitogen signaling (Gregory et al. 2004) and somatic

256

E.M. Wilson

Fig. 4 Structure of the AR AF2 binding surface with bound AR FXXLF motif. Wild-type AR residues that undergo germline mutations and cause partial or complete androgen insensitivity syndrome (AIS) (cyan) are distributed within the AF2 binding surface for the AR FXXLF (23FQNLF27) motif, where the latter is represented by Phe-27 (F27), Phe-23 (F23), and Leu-26 (L26) (magenta). AF2 residues mutated in individual cases of AIS form hydrophobic Cluster A comprising Gln-733 (Q733), Phe-725 (F725), and Ile-737 (I737) (cyan). Cluster A interacts with Phe-27 of the AR FXXLF motif and lies more distal than Cluster B to the ligand binding pocket. Naturally occurring mutations in Cluster A cause AIS by disrupting AR FXXLF motif binding, which increases the dissociation rate of bound androgen. Cluster B comprises Ile-737 (I737), Leu712 (L712) and Ile-898 (I898) (cyan). Cluster B interacts with Phe-23 of the AR FXXLF motif and lies closer to the ligand binding pocket than Cluster A. AIS mutations in Cluster B alter AR FXXLF motif binding and have direct effects on bound ligand. Note that Leu-712, positioned at the AF2 surface near the ligand binding pocket, was identified in two conformations in T-bound AR LBD crystal structures, indicating that bound T imparts greater conformational heterogeneity in AF2. Bound DHT and R1881 result in a single conformation of Leu-712. Val-889 (V889) (cyan) has a gatekeeper function in ligand binding. Charge clamp residues Lys-720 (K720) and Glu-897 (E897) are in yellow, and the bound synthetic androgen, methyltrienolone (R1881) is in green. Sites in wild-type AR somatic AF2 mutations identified in prostate cancer, His-874 (H874), Val-715 (V715), and Thr-877 (T877), are in gold (He et al. 2006) (See Color Insert)

mutations in the AR gene (Culig et al. 1993; Shi et al. 2002; Tan et al. 1997; Veldscholte et al. 1992). The frequency of AR mutations in prostate cancer is low, but increases during tumor progression to castration-recurrent disease in association with increased genome instability (Gelmann 2002; Taplin et al. 1999). Wellcharacterized somatic mutations that increase ligand-dependent AR transcriptional

Functional Motifs of the Androgen Receptor

257

activity include T877A in the LNCaP prostate cancer cell line (Veldscholte et al. 1992), H874Y in the CWR22 human prostate xenograft and CWR-R1 prostate cancer cell line derived from the castration-recurrent CWR22 xenograft (Gregory et al. 1998, 2001c; Tan et al. 1997), V715M (Culig et al. 1993), and V730M (Culig et al. 1993; He et al. 2004; Newmark et al. 1992; Peterziel et al. 1995) identified in prostate cancer specimens. Mechanisms underlying the increase in AR activity by these mutations differ even though each involves the AF2 motif binding site. Val-730 is in the hydrophobic core of AF2. The V730M somatic mutation in prostate cancer increases the AR AF2 interaction with the LXXLL motif of SRC1 (He et al. 2004). His-874, Val-715, and Thr-877 lie in a structural boundary ˚ from Trp-741, a ligand binding between AF2 and the ligand binding pocket, 5 A pocket residue whose side chain changes position depending on which ligand is bound (Askew et al. 2007; He et al. 2006). H874Y and V715M increase the AF2 interaction with the FXXLF and LXXLL motifs, which for H874Y, is independent of a change in motif binding affinity (Askew et al. 2007; Duff and McEwan 2005; He et al. 2006). T877A increases the intramolecular space around ring D of the bound steroid, allowing a broader spectrum of ligands to function as agonists, while maintaining an essentially wild-type response to T and DHT (He et al. 2006). AR-T877A enhances the AR N/C interaction in response to partial agonists and traditional antagonists primarily by changes in ligand binding specificity (Langley et al. 1995). A notable feature of the H874Y, V715M, and T877A somatic mutations that arose in prostate cancer is that each slows the dissociation rate of bound androgen from full-length AR and from an AR fragment that contains the DNA and ligand binding domains and lacks the NH2-terminal region (He et al. 2006). This result indicates that each of these mutations stabilizes androgen in the ligand binding pocket independent of the AR N/C interaction. The stabilizing effects of the H874Y, V715M, and T877A mutations on bound ligand may provide an explanation for previous reports suggesting transcriptional-growth-promoting effects of partial agonists in LNCaP cells that are not associated with the AR N/C interaction (Sathya et al. 2003). In addition, sufficiently high pharmacological concentrations of partial AR agonists and antagonists may contribute to the activity of wild-type and mutant AR (Kemppainen et al. 1992; Sathya et al. 2003; Song et al. 2004; Wong et al. 1993) as demonstrated by the flutamide withdrawal syndrome (Scher and Kelly 1993). Lack of a requirement for the AR N/C interaction for activation of the MMTV promoter enhancer region (He et al. 2002a) suggests further that some androgen-responsive genes may not require the AR N/C interaction. This may depend on the level of specific AR coregulatory proteins and the arrangement of specific transcription factor binding sites in the promoter and enhancer regions. Increased levels of SRC/p160 coactivators in prostate cancer tissue and cell lines provide a mechanism more broadly applicable to increases in AR transcriptional activity in response to androgens and other ligands (Gregory et al. 2001b, c). Crystal structures of the AR-H874Y LBD bound to T and AR FXXLF peptide indicate that the H874Y mutation in helix 10 results in the formation of new direct hydrogen bonds to residues in helices 4 and 5 that replace a water-mediated

258

E.M. Wilson

hydrogen bond network (Askew et al. 2007). In two-hybrid assays, AR FXXLF and LXXLL motif binding increases as a result of the H874Y mutation, but motif binding affinity is unchanged. The H874Y mutation appears to stabilize the LBD core and AF2 helix 12 through direct helix-to-helix hydrogen bonding that renders AF2 a more effective motif binding site. The predominant transcriptional effect of H874Y is to render DHT-like activity to T. Thus, several well-characterized AR somatic mutations in prostate cancer increase AR transcriptional activity through structural effects on AF2. These mutations are most likely selected during tumor progression by their ability to increase AR transcriptional activity and prostate cancer growth in response to a broader range of ligands. The results are consistent with a critical role for AR in prostate cancer development and progression and point to the AF2 binding surface as a mechanism for increased AR transcriptional activity in castration-recurrent prostate cancer that arises in an environment of low circulating androgen. Increased coactivator recruitment by AF2 in wild-type AR is also facilitated by increased levels of the SRC/p160 family members during prostate cancer progression (Gregory et al. 2001b).

5 AR Coregulator MAGE-11 Binds the AR FXXLF Motif AR FXXLF motif binding to AF2 is an important control mechanism regulating AR transcriptional activity. We discussed how the AR N/C interaction enhances the potency of endogenous androgens by stabilizing AR but inhibits SRC/p160 coactivator recruitment by AF2. These observations suggest that AR coregulators that alter the AR N/C interaction could either inhibit or enhance AR transcriptional activity. Inhibition of AR transcriptional activity by p53 and cyclin D1 has been linked to interference of the AR N/C interaction (Burd et al. 2005; Shenk et al. 2001). Two-hybrid interaction screens of a human testis library using the AR FXXLF motif region as bait identified melanoma antigen gene protein-11 (MAGE-11 or MAGE-A11) as an AR coregulator that binds the FXXLF motif and increases AR transcriptional activity by exposing AF2 for coactivator recruitment (Bai et al. 2005).

5.1

AR Coregulator MAGE-11

Identification of MAGE-11 as an AR coregulator that selectively binds the AR NH2-terminal FXXLF motif demonstrates that the AR FXXLF motif serves at least two functions in modulating AR transcriptional activity. In addition to binding AF2 in the androgen-dependent AR N/C interaction, the AR FXXLF motif serves as the principal interaction site for MAGE-11 (Bai et al. 2005). Mutagenesis studies demonstrated different flanking sequence requirements for AR FXXLF motif binding to AF2 and MAGE-11. In two-hybrid assays and crystal structures, the short

Functional Motifs of the Androgen Receptor

259

AR-20–30 amino acid peptide RGAFQNLFQSV is necessary and sufficient to bind AF2 (Askew et al. 2007; He et al. 2004; He and Wilson 2003) and requires androgen bound in the AR ligand binding pocket. In contrast, the minimal sequence required for MAGE-11 to interact with the AR FXXLF motif region is AR-16–36. AR residues Val-30 and Val-33 are required for MAGE-11 to bind the AR FXXLF motif region, whereas mutations at these sites have no effect on the AR N/C interaction (Bai et al. 2005). Thus, there is an overlapping but distinct FXXLF motif binding region for AF2 and MAGE-11. Furthermore, MAGE-11 does not interact with FXXLF motifs present in several putative AR coregulators that interact with the AR AF2 site. Thus, MAGE-11 binding to AR requires a longer alpha helical FXXLF motif region and is more selective than AR AF2 binding of the AR FXXLF motif. Another difference is that MAGE-11 binds the AR FXXLF motif in an androgen-independent manner. MAGE-11 more readily coimmunoprecipitates with AR in the absence than in the presence of androgen, although recent studies indicate that the interaction between AR and MAGE-11 in the presence of androgen is dynamic and influenced by growth factor signaling. EGF-dependent phosphorylation at MAGE-11 Thr-360 serves to target the monoubiquitinylation of MAGE-11 at Lys-240 and 245, which is required for MAGE-11 to interact with the FXXLF motif region of AR (Bai and Wilson 2008). The mechanisms whereby MAGE-11 increases AR transcriptional activity are being actively investigated. Thus far we know that binding of MAGE-11 to the AR NH2-terminal FXXLF motif in the presence of androgen increases AR transcriptional activity by interfering with the androgen-dependent AR N/C interaction and exposing AF2 for coactivator recruitment (Bai et al. 2005). MAGE-11 also increases AR transcriptional activity through mechanisms independent of AF2. This is evident using the AR NH2-terminal and DNA binding fragment AR-1– 660 that lacks the LBD and AF2 binding site. AR-1–660 is strongly activated by MAGE-11 in an FXXLF motif-dependent manner through mechanisms that remain to be established. A model for the interrelationship between the AR N/C interaction, MAGE-11 binding, and AR transactivation involving SRC/p160 coactivators is shown in Fig. 5. MAGE-11 binding to the AR FXXLF motif increases the turnover of AR in association with growth-factor-dependent phosphorylation and ubiquitinylation of MAGE-11 (Bai and Wilson 2008). The MAGE-11-dependent increase in AR turnover and transcriptional activity is reminiscent of agonist-induced down-regulation that characterizes other steroid receptors and demonstrates a link between receptor degradation and transcriptional activity (Dennis et al. 2001; Verma et al. 2004; Wu et al. 2007). In the presence of androgen, AR and MAGE-11 colocalize in a disperse pattern throughout the nucleus. This differs from the colocalization pattern of MAGE-11 with an AR DNA binding mutant within prominent subnuclear particles. Thus, AR and MAGE-11 are intimately linked within the nucleus dependent on the transcriptional status of AR. In the absence of androgen, MAGE-11 colocalizes with AR in the cytoplasm but is also detected in the nucleus unassociated with AR (Bai et al. 2005). Under these conditions, MAGE-11 binding to the AR FXXLF motif is unimpeded by the

260

E.M. Wilson

Fig. 5 Schematic of AR modulation by the N/C interaction, MAGE-11 and SRC/p160 coactivators. A dynamic equilibrium that regulates AR transcriptional activity includes the androgendependent AR N/C interaction and AR binding of MAGE-11 and SRC/p160 coactivators (Bai et al. 2005). In the absence of androgen, MAGE-11 (hatched rectangle) binds the AR FXXLF motif (Fx) and stabilizes AR. In the presence of bound testosterone, dihydrotestosterone (DHT) or synthetic AR agonist, the AR FXXLF motif binds AF2 in the AR ligand binding domain (LBD). The N/C interaction may occur within an inactive AR monomer and through domain swapping in the active AR dimer. The AR N/C interaction has an inhibitory effect on AR transactivation by competing with SRC/p160 coactivator LXXLL (Lx) motif binding to AF2. In the presence of androgen, AR transcriptional activity is increased by AR FXXLF motif binding to MAGE-11 through mechanisms that include, but are not limited to, increased AR AF2 recruitment of SRC/ p160 coactivators (Bai et al. 2005)

androgen-dependent AR N/C interaction. MAGE-11 stabilizes AR in the absence of androgen and at less than saturating levels of androgen, indicating at second mechanism involving the AR FXXLF motif that stabilizes AR and contributes to increased AR transcriptional activity.

5.2

Expression Pattern of MAGE-11

The physiological significance of MAGE-11 in modulating AR transcriptional activity remains to be fully established. MAGE-11 is a member of the MAGE gene superfamily of so-called cancer-testis or cancer-germline antigens and part of the 12 member MAGE-A subfamily coded at the Xq28 region of the human X chromosome (Chomez et al. 2001; Rogner et al. 1995; Simpson et al. 2005). MAGE-11 shares extensive sequence homology within its largest carboxyl-terminal exon with

Functional Motifs of the Androgen Receptor

261

other members of the MAGE family (Bai and Wilson 2008). However, the MAGE11 gene has three additional short 50 coding exons, one of which contains a nuclear localization signal (Bai et al. 2005; Irvine and Coetzee 1999). Although AR and MAGE-11 are X-linked genes, naturally occurring mutations in the MAGE-11 gene have yet to be identified. MAGE-11 is expressed in human and nonhuman primates but not in other mammalian species such as rats and mice. Since a functional homolog for MAGE-11 has not been identified in rodents, MAGE-11 may have evolved as part of a mechanism for higher-level control of AR function in primates. Recent studies demonstrate that MAGE-11 is expressed in reproductive tissues of the human male and female, with temporal expression in human endometrium during the menstrual cycle (Bai et al. 2008). MAGE-11 mRNA levels are up-regulated by cyclic AMP and down-regulated by estradiol in Ishikawa and ECC-1 human endometrial cell lines. Studies are ongoing in primate models to determine the functional effects of MAGE-11 in male reproductive function. In prostate cancer, the levels of MAGE-11 appear to increase based on studies in prostate cancer cell lines. MAGE-11 mRNA levels correlate with AR expression since both are detected in LNCaP, CWR-R1, and LAPC-4 prostate cancer cell lines but are essentially undetectable in DU145 and PC-3 cells. The functional role of MAGE-11 in prostate cancer remains to be fully established. However, MAGE-11 binding to the AR FXXLF motif contributes to increased SRC/p160 coactivator recruitment by AF2 and may contribute to ligand-independent AR activity through mechanisms that more directly involve the AR NH2-terminal AF1 transcriptional activation region.

References Alen P, Claessens F, Verhoeven G, Rombauts W, Peeters B (1999) The androgen receptor amino-terminal domain plays a key role in p160 coactivator-stimulated gene transcription. Mol Cell Biol 19:6085–6097 Askew EB, Gampe RT, Stanley TB, Faggart JL, Wilson EM (2007) Modulation of androgen receptor activation function 2 by testosterone and dihydrotestosterone. J Biol Chem 282:25801–25816 Bai S, Wilson EM (2008) EGF dependent phosphorylation and ubiquitinylation of MAGE-11 regulates its interaction with the androgen receptor. Mol Cell Biol 28:1947–1963 Bai S, He B, Wilson EM (2005) Melanoma antigen gene protein MAGE-11 regulates androgen receptor function by modulating the interdomain interaction. Mol Cell Biol 25:1238–1257 Bai S, Grossman G, Yuan L, Lessey BA, French FS, Young SL, Wilson EM (2008) Hormone control and expression of androgen receptor coregulator MAGE-11 in human endometrium during the window of receptivity to embryo implantation. Mol Hum Reprod 14:107–116 Bennett MJ, Schlunegger MP, Eisenberg D (1995) 3D domain swapping: a mechanism for oligomer assembly. Protein Sci 4:2455–2468 Bennett MJ, Sawaya MR, Eisenberg D (2006) Deposition diseases and 3D domain swapping. Structure 14:811–824

262

E.M. Wilson

Berrevoets CA, Doesburg P, Steketee K, Trapman J, Brinkmann AO (1998) Functional interactions of the AF-2 activation domain core region of the human androgen receptor with the amino-terminal domain and with the transcriptional coactivator TIF2 (transcriptional intermediary factor 2). Mol Endocrinol 12:1172–1183 Bevan CL, Hoare S, Claessens F, Heery DM, Parker MG (1999) The AF1 and AF2 domains of the androgen receptor interact with distinct regions of SRC1. Mol Cell Biol 19:8383–8392 Brown CJ, Goss SJ, Lubahn DB, Joseph DR, Wilson EM, French FS, Willard HF (1989) Androgen receptor locus on the human X chromosome: regional localization to Xq11–12 and description of a DNA polymorphism. Am J Hum Genet 44:264–269 Bubulya A, Chen SY, Fisher CJ, Zheng Z, Shen XQ, Shemshedini L (2001) c-Jun potentiates the functional interaction between the amino and carboxyl termini of the androgen receptor. J Biol Chem 276:44704–44711 Burd CJ, Petre CE, Moghadam H, Wilson EM, Knudsen KE (2005) Cyclin D1 binding to the androgen receptor (AR) NH2-terminal domain inhibits activation function 2 association and reveals dual roles for AR corepression. Mol Endocrinol 19:607–620 Casella R, Maduro MR, Misfud A, Lipshultz LI, Yong EL, Lamb DJ (2003) Androgen receptor gene polyglutamine length is associated with testicular histology in infertile patients. J Urol 169:224–227 Chang CY, Abdo J, Hartney T, McDonnell DP (2005) Development of peptide antagonists for the androgen receptor using combinatorial peptide phage display. Mol Endocrinol 19:2478–2490 Chomez P, De Backer O, Bertrand M, De Plaen E, Boon T, Lucas S (2001) An overview of the MAGE gene family with the identification of all human members of the family. Cancer Res 61:5544–5551 Choong CS, Wilson EM (1998) Trinucleotide repeats in the human AR: a molecular basis for disease. J Mol Endocrinol 21:235–257 Choong CS, Kemppainen JA, Zhou ZX, Wilson EM (1996) Reduced androgen receptor gene expression with first exon CAG repeat expansion. Mol Endocrinol 10:1527–1535 Choong CS, Kemppainen JA, Wilson EM (1998) Evolution of the primate androgen receptor: a structural basis for disease. J Mol Evol 47:334–342 Culig Z, Hobisch A, Cronauer MV, Cato AC, Hittmair A, Radmayr C, Eberle J, Bartsch G, Klocker H (1993) Mutant androgen receptor detected in an advanced-stage prostatic carcinoma is activated by adrenal androgens and progesterone. Mol Endocrinol 7:1541–1550 Darimont BD, Wagner RL, Apriletti JW, Stallcup MR, Kushner PJ, Baxter JD, Fletterick RJ, Yamamoto KR (1998) Structure and specificity of nuclear receptor-coactivator interactions. Genes Dev 12:3343–3356 Dehm SM, Regan KM, Schmidt LJ, Tindall DJ (2007) Selective role of an NH2-terminal WxxLF motif for aberrant androgen receptor activation in androgen depletion independent prostate cancer cells. Cancer Res 67:10067–10077 Dennis AP, Haq RU, Nawaz Z (2001) Importance of the regulation of nuclear receptor degradation. Front Biosci 6:D954–D959 Dubbink HJ, Hersmus R, Verma CS, van der Korput HA, Berrevoets CA, van Tol J, Ziel-van der Made AC, Brinkmann AO, Pike AC, Trapman J (2004) Distinct recognition modes of FXXLF and LXXLL motifs by the androgen receptor. Mol Endocrinol 18:2132–2150 Duff J, McEwan IJ (2005) Mutation of histidine 874 in the androgen receptor ligand-binding domain leads to promiscuous ligand activation and altered p160 coactivator interactions. Mol Endocrinol 19:2943–2954 Edwards A, Hammond HA, Jin L, Caskey CT, Chakraborty R (1992) Genetic variation at five trimeric and tetrameric tandem repeat loci in four human population groups. Genomics 12:241–253 Estebanez-Perpina E, Moore JM, Mar E, Delgado-Rodrigues E, Nguyen P, Baxter JD, Buehrer BM, Webb P, Fletterick RJ, Guy RK (2005) The molecular mechanisms of coactivator utilization in ligand-dependent transactivation by the androgen receptor. J Biol Chem 280:8060–8068

Functional Motifs of the Androgen Receptor

263

Gao T, McPhaul MJ (1998) Functional activities of the A and B forms of the human androgen receptor in response to androgen receptor agonists and antagonists. Mol Endocrinol 12: 654–663 Gelmann EP (2002) Molecular biology of the androgen receptor. J Clin Oncol 20:3001–3015 Ghali SA, Gottlieb B, Lumbroso R, Beitel LK, Elhaji Y, Wu J, Pinsky L, Trifiro MA (2003) The use of androgen receptor amino/carboxyl-terminal interaction assays to investigate androgen receptor gene mutations in subjects with varying degrees of androgen insensitivity. J Clin Endocrinol Metab 88:2185–2193 Goh KI, Cusick ME, Valle D, Childs B, Vidal M, Baraba´si AL (2007) The human disease network. Proc Natl Acad Sci USA 104:8685–8690 Gregory CW, Hamil KG, Kim D, Hall SH, Pretlow TG, Mohler JL, French FS (1998) Androgen receptor expression in androgen-independent prostate cancer is associated with increased expression of androgen-regulated genes. Cancer Res 58:5718–5724 Gregory CW, He B, Wilson EM (2001a) The putative androgen receptor-A form results from in vitro proteolysis. J Mol Endocrinol 27:309–319 Gregory CW, He B, Johnson RT, Ford OH, Mohler JL, French FS, Wilson EM (2001b) A mechanism for androgen receptor mediated prostate cancer recurrence after androgen deprivation therapy. Cancer Res 61:4315–4319 Gregory CW, Johnson RT, Mohler JL, French FS, Wilson EM (2001c) Androgen receptor stabilization in recurrent prostate cancer is associated with hypersensitivity to low androgen. Cancer Res 61:2892–2898 Gregory CW, Fei X, Ponguta LA, He B, Bill HM, French FS, Wilson EM (2004) Epidermal growth factor increases coactivation of the androgen receptor in recurrent prostate cancer. J Biol Chem 279:7119–7130 Grino PB, Griffin JE, Wilson JD (1990) Testosterone at high concentrations interacts with the human androgen receptor similarly to dihydrotestosterone. Endocrinology 126:1165–1172 Haelens A, Tanner T, Denayer S, Callewaert L, Claessens F (2007) The hinge region regulates DNA binding, nuclear translocation, and transactivation of the androgen receptor. Cancer Res 67:4514–4523 He B, Wilson EM (2002) The NH2-terminal and carboxyl-terminal interaction in the human androgen receptor. Mol Genet Metab 75:293–298 He B, Wilson EM (2003) Electrostatic modulation of steroid receptor recruitment of the LXXLL and FXXLF motifs. Mol Cell Biol 23:2135–2150 He B, Kemppainen JA, Voegel JJ, Gronemeyer H, Wilson EM (1999) Activation function 2 in the human androgen receptor ligand binding domain mediates interdomain communication with the NH2-terminal domain. J Biol Chem 274:37219–37225 He B, Kemppainen JA, Wilson EM (2000) FXXLF and WXXLF sequences mediate the NH2terminal interaction with the ligand binding domain of the androgen receptor. J Biol Chem 275:22986–22994 He B, Bowen NT, Minges JT, Wilson EM (2001) Androgen-induced NH2- and carboxyl-terminal interaction inhibits p160 coactivator recruitment by activation function 2. J Biol Chem 276:42293–42301 He B, Lee LW, Minges JT, Wilson EM (2002a) Dependence of selective gene activation on the androgen receptor NH2- and carboxyl-terminal interaction. J Biol Chem 277:25631–25639 He B, Minges JT, Lee LW, Wilson EM (2002b) The FXXLF motif mediates androgen receptorspecific interactions with coregulators. J Biol Chem 277:10226–10235 He B, Gampe RT, Kole AJ, Hnat AT, Stanley TB, An G, Stewart EL, Kalman RI, Minges JT, Wilson EM (2004) Structural basis for androgen receptor interdomain and coactivator interactions suggests a transition in nuclear receptor activation function dominance. Mol Cell 16:425–438 He B, Gampe RT, Hnat AT, Faggart JL, Minges JT, French FS, Wilson EM (2006) Probing the functional link between androgen receptor coactivator and ligand binding sites in prostate cancer and androgen insensitivity. J Biol Chem 281:6648–6663

264

E.M. Wilson

Heery DM, Kalkhoven E, Hoare S, Parker MG (1997) A signature motif in transcriptional coactivators mediates binding to nuclear receptors. Nature 387:733–736 Henttu PM, Kalkhoven E, Parker MG (1997) AF-2 activity and recruitment of steroid receptor coactivator 1 to the estrogen receptor depend on a lysine residue conserved in nuclear receptors. Mol Cell Biol 17:1832–1839 Hsu CL, Chen YL, Yeh S, Ting HJ, Hu YC, Lin H, Wang X, Chang C (2003) The use of phage display technique for the isolation of androgen receptor interacting peptides with (F/W)XXL (F/W) and FXXLY new signature motifs. J Biol Chem 278:23691–23698 Hsu CL, Chen YL, Ting HJ, Lin WJ, Yang Z, Zhang Y, Wang L, Wu CT, Chang HC, Yeh S, Pimplikar SW, Chang C (2005) Androgen receptor (AR) NH2- and COOH-terminal interactions result in the differential influences on the AR-mediated transactivation and cell growth. Mol Endocrinol 19:350–361 Hur E, Pfaff SJ, Payne ES, Gron H, Buehrer BM, Fletterick RJ (2004) Recognition and accommodation at the androgen receptor coactivator binding interface. PLoS Biol 2:E274 Ikonen T, Palvimo JJ, Janne OA (1997) Interaction between the amino- and carboxyl-terminal regions of the rat androgen receptor modulates transcriptional activity and is influenced by nuclear receptor coactivators. J Biol Chem 272:29821–29828 Imperato-McGinley J, Guerrero L, Gautier T, Peterson RE (1974) Steroid 5-alpha-reductase deficiency in man: an inherited form of male pseudohermaphroditism. Science 186:1213–1215 Irvine RA, Coetzee GA (1999) Additional upstream coding sequences of MAGE-11. Immunogenetics 49:585 Jaaskelainen J, Deeb A, Schwabe JW, Mongan NP, Martin H, Hughes IA (2006) Human androgen receptor gene ligand-binding-domain mutations leading to disrupted interaction between the N- and C-terminal domains. J Mol Endocrinol 36:361–368 Kemppainen JA, Wilson EM (1996) Agonist and antagonist activities of hydroxyflutamide and casodex relate to androgen receptor stabilization. Urology 48:157–163 Kemppainen JA, Lane MV, Sar M, Wilson EM (1992) Androgen receptor phosphorylation, turnover, nuclear transport and transcriptional activation: specificity for steroids and antihormones. J Biol Chem 267:968–974 Kemppainen JA, Langley E, Wong CI, Bobseine K, Kelce WR, Wilson EM (1999) Distinguishing androgen receptor agonists and antagonists: distinct mechanisms of activation by medroxyprogesterone acetate and dihydrotestosterone. Mol Endocrinol 13:440–454 Klokk TI, Kurys P, Elbi C, Nagaich AK, Hendarwanto A, Slagsvold T, Chang CY, Hager GL, Saatcioglu F (2007) Ligand-specific dynamics of the androgen receptor at its response element in living cells. Mol Cell Biol 27:1823–1843 Langley E, Zhou ZX, Wilson EM (1995) Evidence for an antiparallel orientation of the ligand activated human androgen receptor dimer. J Biol Chem 270:29983–29990 Langley E, Kemppainen JA, Wilson EM (1998) Intermolecular NH2-/carboxyl-terminal interactions in androgen receptor dimerization revealed by mutations that cause androgen insensitivity. J Biol Chem 273:92–101 La Spada AR, Wilson EM, Lubahn DB, Harding AE, Fischbeck KH (1991) Androgen receptor gene mutations in X-linked spinal and bulbar muscular atrophy. Nature 352:77–79 Le Douarin B, Nielsen AL, Garnier JM, Ichinose H, Jeanmougin F, Losson R, Chambon P (1996) A possible involvement of TIF1 alpha and TIF1 beta in the epigenetic control of transcription by nuclear receptors. EMBO J 15:6701–6715 Li J, Fu J, Toumazou C, Yoon HG, Wong J (2006) A role of the amino-terminal (N) and carboxylterminal (C) interaction in binding of androgen receptor to chromatin. Mol Endocrinol 20: 776–785 Liao M, Zhou Z, Wilson EM (1999) Redox-dependent DNA binding of the purified androgen receptor: evidence for disulfide-linked androgen receptor dimers. Biochemistry 38:9718–9727 Lim HN, Nixon RM, Chen H, Hughes IA, Hawkins JR (2001) Evidence that longer androgen receptor polyglutamine repeats are a causal factor for genital abnormalities. J Clin Endocrinol Metab 86:3207–3210

Functional Motifs of the Androgen Receptor

265

Linja MJ, Visakorpi T (2004) Alterations of androgen receptor in prostate cancer. J Steroid Biochem Mol Biol 92:255–264 Lubahn DB, Joseph DR, Sar M, Tan JA, Higgs HN, Larson RE, French FS, Wilson EM (1988a) The human androgen receptor: complementary DNA cloning, sequence analysis and gene expression in prostate. Mol Endocrinol 2:1265–1275 Lubahn DB, Joseph DR, Sullivan PM, Willard HF, French FS, Wilson EM (1988b) Cloning of human androgen receptor cDNA and localization to the X chromosome. Science 240:327–330 Maes M, Sultan C, Zerhouni N, Rothwell SW, Migeon CJ (1979) Role of testosterone binding to the androgen receptor in male sexual differentiation of patients with 5 alpha-reductase deficiency. J Steroid Biochem 11:1385–1392 Mahendroo MS, Cala KM, Hess DL, Russell DW (2001) Unexpected virilization in male mice lacking steroid 5 alpha-reductase enzymes. Endocrinology 142:4652–4662 Matias PM, Donner P, Coelho R, Thomaz M, Peixoto C, Macedo S, Otto N, Joschko S, Scholz P, Wegg A, Basler S, Schafer M, Egner U, Carrondo MA (2000) Structural evidence for ligand specificity in the binding domain of the human androgen receptor. Implications for pathogenic gene mutations. J Biol Chem 275:26164–26171 Mifsud A, Sim CK, Boettger-Tong H, Moreira S, Lamb DJ, Lipshultz LI, Yong EL (2001) Trinucleotide (CAG) repeat polymorphisms in the androgen receptor gene: molecular markers of risk for male infertility. Fertil Steril 75:275–281 Mohler JL, Gregory CW, Ford OH, Kim D, Weaver CM, Petrusz P, Wilson EM, French FS (2004) The androgen axis in recurrent prostate cancer. Clin Can Res 10:440–448 Newmark JR, Hardy DO, Tonb DC, Carter BS, Epstein JI, Isaacs WB, Brown TR, Barrack ER (1992) Androgen receptor gene mutations in human prostate cancer. Proc Natl Acad Sci USA 89:6319–6323 Ostrowski J, Kuhns JE, Lupisella JA, Manfredi MC, Beehler BC, Krystek SR Jr, Bi Y, Sun C, Seethala R, Golla R, Sleph PG, Fura A, An Y, Kish KF, Sack JS, Mookhtiar KA, Grover GJ, Hamann LG (2007) Pharmacological and x-ray structural characterization of a novel selective androgen receptor modulator: potent hyperanabolic stimulation of skeletal muscle with hypostimulation of prostate in rats. Endocrinology 148:4–12 Peterziel H, Culig Z, Stober J, Hobisch A, Radmayr C, Bartsch G, Klocker H, Cato AC (1995) Mutant androgen receptors in prostatic tumors distinguish between amino-acid-sequence requirements for transactivation and ligand binding. Int J Cancer 63:544–550 Quigley CA, De Bellis A, Marschke KB, El-Awady MK, Wilson EM, French FS (1995) Androgen receptor defects: historical, clinical and molecular perspectives. Endocr Rev 16:271–321 Quigley CA, Tan JA, He B, Zhou ZX, Mebarki F, Morel Y, Forest M, Chatelain P, Ritzen EM, French FS, Wilson EM (2004) Partial androgen insensitivity with phenotypic variation caused by androgen receptor mutations that disrupt activation function 2 and the NH2- and carboxylterminal interaction. Mech Ageing Dev 125:683–695 Rogner UC, Wilke K, Steck E, Korn B, Poustka A (1995) The melanoma antigen gene (MAGE) family is clustered in the chromosomal band Xq28. Genomics 29:725–731 Sack JS, Kish KF, Wang C, Attar RM, Kiefer SE, An Y, Wu GY, Scheffler JE, Salvati ME, Krystek SR Jr, Weinmann R, Einspahr HM (2001) Crystallographic structures of the ligandbinding domains of the androgen receptor and its T877A mutant complexed with the natural agonist dihydrotestosterone. Proc Natl Acad Sci USA 98:4904–4909 Saitoh M, Takayanagi R, Goto K, Fukamizu A, Tomura A, Yanase T, Nawata H (2002) The presence of both the amino- and carboxyl-terminal domains in the AR is essential for the completion of a transcriptionally active form with coactivators and intranuclear compartmentalization common to the steroid hormone receptors: a three-dimensional imaging study. Mol Endocrinol 16:694–706 Sathya G, Chang CY, Kazmin D, Cook CE, McDonnell DP (2003) Pharmacological uncoupling of androgen receptor-mediated prostate cancer cell proliferation and prostate-specific antigen secretion. Cancer Res 63:8029–8036

266

E.M. Wilson

Schaufele F, Carbonell X, Guerbadot M, Borngraeber S, Chapman MS, Ma AA, Miner JN, Diamond MI (2005) The structural basis of androgen receptor activation: intramolecular and intermolecular amino-carboxy interactions. Proc Natl Acad Sci USA 102:9802–9807 Scheller A, Hughes E, Golden KL, Robins DM (1998) Multiple receptor domains interact to permit, or restrict, androgen-specific gene activation. J Biol Chem 273:24216–24222 Scher HI, Kelly WK (1993) Flutamide withdrawal syndrome: its impact on clinical trials in hormone-refractory prostate cancer. J Clin Oncol 11:1566–1572 Schoenberg MP, Hakimi JM, Wang S, Bova GS, Epstein JI, Fischbeck KH, Isaacs WB, Walsh PC, Barrack ER (1994) Microsatellite mutation (CAG24!18) in the androgen receptor gene in human prostate cancer. Biochem Biophys Res Commun 198:74–80 Shenk JL, Fisher CJ, Chen SY, Zhou XF, Tillman K, Shemshedini L (2001) p53 represses androgen-induced transactivation of prostate-specific antigen by disrupting hAR amino- to carboxyl-terminal interaction. J Biol Chem 276:38472–38479 Shi XB, Ma AH, Xia L, Kung HJ, de Vere White RW (2002) Functional analysis of 44 mutant androgen receptors from human prostate cancer. Cancer Res 62:1496–502 Simental JA, Sar M, Lane MV, French FS, Wilson EM (1991) Transcriptional activation and nuclear targeting signals of the human androgen receptor. J Biol Chem 266:510–518 Simpson AJ, Caballero OL, Jungbluth A, Chen YT, Old LJ (2005) Cancer/testis antigens, gametogenesis and cancer. Nat Rev Cancer 5:615–625 Slagsvold T, Kraus I, Bentzen T, Palvimo J, Saatcioglu F (2000) Mutational analysis of the androgen receptor AF-2 (activation function 2) core domain reveals functional and mechanistic differences of conserved residues compared with other nuclear receptors. Mol Endocrinol 14:1603–1617 Song LN, Coghlan M, Gelmann EP (2004) Antiandrogen effects of mifepristone on coactivator and corepressor interactions with the androgen receptor. Mol Endocrinol 18:70–85 Steketee K, Berrevoets CA, Dubbink HJ, Doesburg P, Hersmus R, Brinkmann AO, Trapman J (2002) Amino acids 3–13 and amino acids in and flanking the 23FxxLF27 motif modulate the interaction between the N-terminal and ligand-binding domain of the androgen receptor. Eur J Biochem 269:5780–5791 Takeo J, Yamashita S (1999) Two distinct isoforms of cDNA encoding rainbow trout androgen receptors. J Biol Chem 274:5674–5680 Tan JA, Sharief Y, Hamil KG, Gregory CW, Zang DY, Sar M, Gumerlock PH, deVere White RW, Pretlow TG, Harris SE, Wilson EM, Mohler JL, French FS (1997) Dehydroepiandrosterone activates mutant androgen receptors expressed in the androgen dependent human prostate cancer xenograft CWR22 and LNCaP cells. Mol Endocrinol 11:450–459 Taplin ME, Bubley GJ, Ko YJ, Small EJ, Upton M, Rajeshkumar B, Balk SP (1999) Selection for androgen receptor mutations in prostate cancers treated with androgen antagonist. Cancer Res 59:2511–2515 Taylor JP, Tanaka F, Robitschek J, Sandoval CM, Taye A, Markovic-Plese S, Fischbeck KH (2003) Aggresomes protect cells by enhancing the degradation of toxic polyglutamine-containing protein. Hum Mol Genet 12:749–757 Thompson J, Saatcioglu F, Ja¨nne OA, Palvimo JJ (2001) Disrupted amino- and carboxyl-terminal interactions of the androgen receptor are linked to androgen insensitivity. Mol Endocrinol 15:923–935 Todo T, Ikeuchi T, Kobayashi T, Nagahama Y (1999) Fish androgen receptor: cDNA cloning, steroid activation of transcription in transfected mammalian cells, and tissue mRNA levels. Biochem Biophys Res Commun 254:378–383 Tung L, Abdel-Hafiz H, Shen T, Harvell DM, Nitao LK, Richer JK, Sartorius CA, Takimoto GS, Horwitz KB (2006) Progesterone receptors (PR)-B and -A regulate transcription by different mechanisms: AF-3 exerts regulatory control over coactivator binding to PR-B. Mol Endocrinol 20:2656–2670 van de Wijngaart DJ, van Royen ME, Hersmus R, Pike AC, Houtsmuller AB, Jenster G, Trapman J, Dubbink HJ (2006) Novel FXXFF and FXXMF motifs in androgen receptor cofactors

Functional Motifs of the Androgen Receptor

267

mediate high affinity and specific interactions with the ligand-binding domain. J Biol Chem 281:19407–19416 Veldscholte J, Berrevoets CA, Ris-Stalpers C, Kuiper GG, Jenster G, Trapman J, Brinkmann AO, Mulder E (1992) The androgen receptor in LNCaP cells contains a mutation in the ligand binding domain which affects steroid binding characteristics and response to antiandrogens. J Steroid Biochem Mol Biol 41:665–669 Verma S, Ismail A, Gao X, Fu G, Li X, O’Malley BW, Nawaz Z (2004) The ubiquitin-conjugating enzyme UBCH7 acts as a coactivator for steroid hormone receptors. Mol Cell Biol 24:8716– 8726 Wang Q, Lu J, Yong EL (2001) Ligand- and coactivator-mediated transactivation function (AF2) of the androgen receptor ligand-binding domain is inhibited by the cognate hinge region. J Biol Chem 276:7493–7499 Wilson EM (2007) Muscle bound? A tissue-selective nonsteroidal androgen receptor modulator. Endocrinology 148:1–3 Wilson EM, French FS (1976) Binding properties of androgen receptors: evidence for identical receptors in rat testis, epididymis, and prostate. J Biol Chem 251:5620–5629 Wilson EM, French FS (1979) Effects of proteases and protease inhibitors on the 4.5S and 8S androgen receptor. J Biol Chem 254:6310–6319 Wilson EM, He B, Langley E (2003) Methods for detecting domain interactions in nuclear receptors. In: Russell DW, Mangelsdorf DJ (eds) Methods in Enzymology, Nuclear Receptors, Vol. 364. Academic Press, New York, pp. 142–152 Wong CI, Zhou ZX, Sar M, Wilson EM (1993) Steroid requirement for androgen receptor dimerization and DNA binding: modulation by intramolecular interactions between the NH2terminal and steroid binding domains. J Biol Chem 268:19004–19012 Wu RC, Feng Q, Lonard DM, O’Malley BW (2007) SRC-3 coactivator functional lifetime is regulated by a phospho-dependent ubiquitin time clock. Cell 129:1125–1140 Zhou ZX, Sar M, Simental JA, Lane MV, Wilson EM (1994) A ligand dependent bipartite nuclear targeting signal in the human androgen receptor: requirement for the DNA binding domain and modulation by the NH2-terminal and carboxyl-terminal sequences. J Biol Chem 269:13115– 13123 Zhou ZX, Lane MV, Kemppainen JA, French FS, Wilson EM (1995) Specificity of ligand dependent androgen receptor stabilization: receptor domain interactions influence ligand dissociation and receptor stability. Mol Endocrinol 9:208–218

The Role of the Androgen Receptor Polyglutamine Tract in Prostate Cancer: In Mice and Men Diane M. Robins

Abstract The androgen receptor (AR) is critical in the initiation and progression of prostate cancer, and therefore may contribute to disease through its genetic variation. Particular scrutiny has focused on a polymorphic N-terminal glutamine (Q) tract (CAG repeat) that shows population heterogeneity. Abnormal expansion of this tract underlies late-onset neurodegeneration, and in vitro the length correlates inversely with transcriptional activity. Yet the question of whether length variation within the range of normal human alleles affects cancer has produced discordant epidemiological results, in part due to interacting genetic and environmental factors in human disease. To test Q tract length effects, the mouse AR gene was converted to the human sequence (h/mAr), creating alleles with 12, 21, or 48 CAG repeats. These mice were grossly normal, but molecular analysis revealed allele-dependent differences in target gene expression. Further, when crossed with mice transgenic for a prostate-directed oncogene (TRAMP), Q tract length-dependent differences in cancer initiation and progression were evident. TRAMP mice with short Q tract ARs exhibited earlier but more slowly progressing disease than mice with median or long Q tract ARs. Q tract length also affected disease progression after castration, but in directions opposite to those in intact mice – the AR12Q allele delayed tumor detection whereas mice with the AR48Q allele fared worse. These experiments provided evidence for a causal relationship between a human polymorphism and a cancer phenotype. In man, Q tract length effects may only be significant at extremes of variation within the normal range and may vary with stage of disease. The h/mAR mice provide an experimental paradigm in which to dissect mechanisms by which Q tract length affects development and progression of prostate cancer. Some of these mechanisms may lead to better predictors of response to therapy and new treatments targeted to the human AR.

D.M. Robins Department of Human Genetics, University of Michigan Medical School, Ann Arbor, MI, 481090618, USA, E-mail: [email protected]

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4_12, # Springer Science + Business Media, LLC 2009

269

270

D.M. Robins

1 Introduction Androgen signaling through its nuclear receptor (androgen receptor, AR) controls male differentiation, including the development and function of the prostate gland. Because of AR’s regulatory role in normal growth and homeostasis, it is also a pivotal component in diseases of the prostate. Prostate cancer in particular is a substantial health problem, with an etiology and course affected by many disparate factors. Risk of disease varies with population and with environment, and many cancers remain indolent while others progress rapidly to a lethal state. Underlying this heterogeneity in disease origin and trajectory is a nearly universal dependence on androgen in early stages of tumor growth and, despite therapeutic hormone depletion, on AR in later stages. Therefore, genetic variation in AR structure and expression may impact initiation and progression of disease. This chapter focuses on a highly polymorphic region of the AR, the polyglutamine tract (polyQ, Q tract, CAG repeat). Variable length of the CAG repeat first was noted in association with neurological disease and subsequently was suspected to be a factor in prostate cancer. Epidemiological studies, however, with different human populations and different patient criteria have produced discordant results. Dissection of glutamine tract effects in vitro has provided mechanisms underlying AR activity differences, but it is unclear whether these differences impact the function of AR during tumorigenesis. To address Q tract variation in vivo and to avoid the confounding genetic and environmental heterogeneity of man, a mouse model with variant human Ar alleles was created that represented extremes of Q tract length that were within the normal range. The mice demonstrated Q tract length effects in prostate cancer that varied with tumor stage and with strength of the androgen axis, which reflected well the complexity of human disease. Analysis of the Q tract in mice has lended support to a broader view of AR’s role in oncogenesis, which includes dictating context-dependent and opposing functions.

2 AR and the Polymorphic Polyglutamine Tract As a member of the nuclear receptor superfamily, AR consists of a highly conserved central DNA-binding domain (DBD), a moderately conserved C-terminal ligand-binding domain (LBD), and a variable N-terminal transactivation domain (NTD). The NTD of AR comprises over half of the encoded protein and bears little similarity to functionally homologous regions of other steroid receptors. Hormone binding to the LBD alters the stable association of AR with a cytoplasmic chaperone complex to a dynamic interaction that permits nuclear localization (Pratt et al. 2004). In the nucleus the DBD recognizes response elements in target genes. The NTD participates in recruiting diverse cofactors that orchestrate transcription, in part dependent on direct contact with the hormone-dependent activation function (AF-2) in the C-terminus (He and Wilson 2002). This N/C interaction is critical for

The Role of the Androgen Receptor Polyglutamine Tract in Prostate Cancer

271

optimal ligand binding and transactivation, and integrates differential effects of coactivators, chaperones, and post-translational modifications by ubiquitin ligases and kinases (Shen and Coetzee 2005). Promoter context-dependent interactions of the NTD with other DNA-binding proteins also enhance selectivity of AR for its targets (Robins 2004). Significant insight into nuclear receptor physiology is possible because the AR gene is located on the X-chromosome (Ar; Xp11-12). Since males are hemizygous for AR, mutations are phenotypically evident, and cases of partial to complete androgen insensitivity (AIS) have revealed structure/function correlations that are informative to the steroid receptor family in general. While mutations leading to androgen insensitivity mostly occur in the DBD and LBD (McPhaul 2002), abnormal expansion of an N-terminal CAG repeat encoding a polyglutamine tract underlies the androgen-dependent late-onset neurodegenerative disease, spinal and bulbar muscular atrophy (SBMA, or Kennedy disease) (La Spada et al. 1991). The Q tract is highly polymorphic, ranging from 9 to 37 repeats in the normal population, but occurrence of 40 or more glutamine residues leads to neurodegeneration. In contrast to the loss of AR function found in AIS, Kennedy disease, like other trinucleotide repeat expansion disorders, results from both a toxic gain of function due to misfolding and aggregation of the mutant protein, as well as a partial loss of normal protein activity (Lieberman and Robins 2008). Studies in vitro ascribe the loss of function to an inverse correlation between length of the Q tract and transcriptional strength of the AR (Mhatre et al. 1993; Chamberlain et al. 1994). Given the wide range in Q tract length in the population, whether its variation underlies other androgen-dependent pathologies, such as prostate cancer, should be investigated. In most individuals, the AR Q tract consists of between 15 and 30 contiguous glutamines from amino acid 58, with 21 CAG repeats in the reference human AR sequence (Zitzmann and Nieschlag 2003) (Fig. 1). In addition to the Q tract, there is a polymorphic polyglycine tract (G tract, GGN repeat) at amino acid 449, with 14–27 repeats. These trimeric repeats vary among ethnic populations in the United States, calling attention to them as possible genetic risk factors (Edwards et al. 1992). In an initial survey, the frequency of AR alleles with fewer than the median 22 CAGs was about 50% for Caucasians but much greater for African Americans and much less for Asian Americans, which parallels their differential risk of prostate cancer. These epidemiological observations suggest the hypothesis that the Q tract might be associated with prostate cancer risk, with higher risk deriving from greater activity of a shorter Q tract AR (Coetzee and Ross 1994). This was initially corroborated by finding a prevalence of short AR-CAG alleles in prostate cancer patients, especially those with advanced disease (Irvine et al. 1995), as well as somatic shortening of the CAG repeat in some tumors (Schoenberg et al. 1994). These findings are potentially important for risk assessment and prognosis, but subsequent research has produced conflicting results. Addressing the complexity requires first establishing that Q tract length differences in the normal range affect AR activity.

272

D.M. Robins

Fig. 1 Polyamino acid tracts within human, mouse, and dog ARs. The diagram shows the receptor domains for transactivation, DNA binding (DBD), and ligand binding (LBD); the former shares 85% amino acid identity between man and mouse whereas the DBD and LBD are identical, except for the intervening hinge (h) domain. The region of coding sequence that was swapped to humanize the mouse AR was stippled, and includes the polyglutamine tracts (QI and QIII) and the polyglycine tract (G). Key amino acid positions are below the diagram for scale. Range in number of codon repeats within the polyamino acid tracts in normal populations is shown for the three species

3 Functional Analysis of the Glutamine Tract Glutamine-rich regions are functionally important in several transcription factors besides AR, including Sp1, TATA-binding protein, and the glucocorticoid receptor (Gerber et al. 1994). These domains form b-sheets that act as polar zippers to enhance affinity with other proteins in a nonspecific manner, or, for expanded Q tracts, to promote self-aggregation (Perutz et al. 1994). More recent biophysical studies of full-length ARs revealed that the relatively disordered NTD shows Q tract length-dependent differences in secondary structure, with longer tracts conferring greater flexibility (Duff et al. 2006). Functional subdomains within the NTD encompassing distinct transactivation surfaces (e.g., TAU1, TAU5) and the N/C interaction motifs (23FxxLF, 435WxxLF) may be differentially masked or exposed dependent on hormone-binding, post-translational modification, and cellspecific coregulators (Dehm and Tindall 2007). The Q tract may modulate these intra- and intermolecular interactions in a length-dependent manner by functioning as a spacer between domains. Initial mechanistic studies ascribed the reduced activity of long Q-tract ARs in transfection assays to decreased AR expression at both the mRNA and protein levels (Choong and Wilson 1998). Trinucleotide repeats are inherently unstable in DNA, and CAG repeats can form stem-loop structures in RNA that could be the target of RNA-binding proteins. Altered affinity of RNA-binding proteins due to CAG repeat length could impact both mRNA stability and translation efficiency. Precedent exists for this mechanism although such binding proteins have yet to be identified for AR. In the gene responsible for myotonic dystrophy (Yeap et al. 2004), an expanded CUG repeat in the 30 untranslated region of the mRNA was

The Role of the Androgen Receptor Polyglutamine Tract in Prostate Cancer

273

bound by a protein (CUGBP1) that caused aberrant RNA processing, which contributes to neurotoxicity (Mankodi et al. 2002). For AR, gene expression is also subject to complex cell-type dependent regulatory mechanisms that may be influenced by Q tract length. Furthermore, while decreased mRNA and protein levels may contribute to reduced activity of expanded Q tract ARs, significant differences in expression levels have not been noted for CAG repeat lengths within the normal range. The first AR-associated protein found to be sensitive to Q tract length differences was Ran/ARA24, which showed decreased interaction and activation of ARs with expanded Q tracts (Hsiao et al. 1999). While RAN/ARA24 behaved as an AR coactivator in transfection assays, it is better known for its roles in cell cycle progression, maintenance of nuclear structure, and nuclear transit of RNA and protein. Differential AR activity impacted by Q tract length thus may be indirectly linked via efficiency of nuclear localization or nuclear retention. However, RAN/ ARA24 has yet to show sensitivity to repeat length differences within the range of normal alleles. Such sensitivity, however, has been found for p160 coactivators. The effect is more likely due to Q tract length effects on AR’s structure than to direct interaction of the tract with other proteins (Irvine et al. 2000; Wang et al. 2004). In experiments where protein expression levels were controlled carefully, AR transactivation decreased linearly with increasing tract lengths by 20% over a range of 9Q to 42Q, and more precipitously by 80% for 50Q. The presence of p160 coactivators accentuated these modest differences. Furthermore, the short tract 9Q AR responded in transfection assays to lower concentrations of androgens than moderate or long tract ARs, primarily due to enhanced N/C interaction (Wang et al. 2004). Short tract ARs increased association with coactivators as well as components of the SWI/SNF chromatin remodeling complex, which increased transactivation and androgen-dependent cell proliferation in prostate cancer cell lines. Thus modest Q tract effects caused by several molecular mechanisms ultimately combine to impact transcriptional output. The effect of Q tract length on transactivation is not a simple linear relationship (Buchanan et al. 2004; Ding et al. 2004). AR N/C interaction seems to be optimally maintained over a critical size range of 16–29Qs. This range encompasses more than 90% of AR alleles in most populations, which suggests that pathologies may be associated with lengths at the extremes of the normal range rather than with subtle variations. This is supported by analysis of a somatic AR mutation where two nonconsecutive leucines disrupt the Q tract, presumably reducing flexibility of the NTD (Buchanan et al. 2004). For this AR-polyQ2L mutant, both protein levels and N/C interaction were reduced, but transactivation activity was increased. Thus very short or long Q tracts disrupt N/C interaction, which destabilizes AR and likely underlies reduced protein levels. This N/C interaction effect is distinct from the incremental effect of Q tract length on transactivation, for which each additional glutamine residue within the normal range may reduce structural order of the NTD, which results in reduced ability to recruit coactivators and other components of the transcriptional machinery. In contrast, interaction with corepressors appears less sensitive to differences in Q tract length (Buchanan et al. 2004). These studies

274

D.M. Robins

support the notion that the Q tract does not possess an intrinsic function but instead enables contact between NTD motifs and AF-2 to position AF-1 for interaction with coregulators. How the multiple, alternative surfaces within AF-1 vary in their interactions dependent on Q tract length remains to be determined.

4 Epidemiology of AR Glutamine Tract Associations Prostate cancer has a significant genetic component, with twin studies suggesting that over 40% of cases are influenced by heredity, but whether this is due to common genetic variants or rare disease genes is unclear (Lichtenstein et al. 2000). Since AR is intimately involved in initiation and progression of prostate cancer, AR variation is a suspected risk factor. Although Q tract length differences within the range of normal alleles lead to demonstrable effects in vitro, studies to ascertain whether such variation impacts human physiology or cancer risk are difficult. Substantial evidence suggests that male fertility is impaired for longer Q tract lengths short of the neuropathological range (Casella et al. 2001; Davis-Dao et al. 2007). However, this association is less evident in European populations, which underscores a role for environmental factors (Yong et al. 2003). In hypogonadal men, such as those with Klinefelter’s syndrome, low androgen levels fail to saturate receptors, and therefore differences in AR activity may be more apparent (Zitzmann et al. 2004; Crabbe et al. 2007). In fact hypogonadal men with short CAG repeats were more responsive to testosterone replacement, which provides a physiological correlate to the hypersensitivity of short tract ARs in vitro (Wang et al. 2004). These syndromes provide support for the notion that Q tract length differences within the normal range affect AR activity in vivo. The hypothesis that short Q tract ARs increase prostate cancer risk was suggested by (1) the greater prevalence of these tracts in the high-incidence AfricanAmerican population, (2) identification of Q tract contraction in malignant but not benign prostate cells (Alvarado et al. 2005), and (3) in vitro evidence for greater transcriptional activity of short Q tract ARs, which might enhance oncogenic transformation. But association at the genetic level has been difficult to confirm for a variety of reasons. Out of nearly 100 studies examining Q tract length and prostate cancer, about half have found an association between short tract length and increased risk, earlier age of diagnosis or more advanced disease at diagnosis (e.g., Irvine et al. 1995; Giovannucci et al. 1997; Stanford et al. 1997). However, many studies, including some with large numbers of subjects, found no association (Zeegers et al. 2004; Freedman et al. 2005), and others found the opposite association of increased risk with long Q tract lengths (Edwards et al. 1999; Suzuki et al. 2002; Li et al. 2003; Lindstrom et al. 2006b). The G tract has not been examined as extensively, but there is a similar level of disagreement as to whether its variation is associated with risk of prostate cancer (e.g., Irvine et al. 1995; Stanford et al. 1997;

The Role of the Androgen Receptor Polyglutamine Tract in Prostate Cancer

275

Hsing et al. 2000; Chang et al. 2002; Zeegers et al. 2004). With regard to later stage disease and AR function, results also disagree on effects of CAG repeat length on response to androgen-deprivation treatment (Bratt et al. 1999; Suzuki et al. 2003). These conflicting conclusions are due to several factors. Smaller studies lack sufficient power to achieve statistical significance. The association with race is subject to differences in environment, socioeconomic status, admixture, and disparity in health care. Furthermore, while the median CAG repeat number is 19 in African Americans and 22 in Caucasian Americans (Edwards et al. 1992; Irvine et al. 1995), relatively few African Americans have tracts less than the 16 Qs that are the cut-off for transcriptional effects in vitro (Buchanan et al. 2004). Heterogeneity in both genetics and environment influences outcome, as evident in opposite associations found between Q tract length and prostate cancer risk in Swedish compared to Japanese populations (Li et al. 2003). Heterogeneity in prostate cancer progression complicates results, since some studies find associations with more advanced disease or earlier age at diagnosis but no association with overall risk (Hardy et al. 1996; Beilin et al. 2001; Cude et al. 2002; Santos et al. 2003; Shimbo et al. 2005; Sieh et al. 2006). Importantly, stronger associations are found in studies with populations either prior to or less reliant on PSA testing, since these patients present with more advanced tumors. Subsequent to PSA testing, malignancies that might never become symptomatic are diagnosed at greater frequency, increasing bias to the null hypothesis (Giovannucci 2002). Advances in genotyping technology have increased the number of genetic association studies for complex diseases such as prostate cancer, but have not resolved an association with Q tract length. A meta-analysis performed in 2004 detected a slightly increased risk associated with short CAG repeats, and this is probably an underestimation due to the inclusion of low-incidence populations (Zeegers et al. 2004). More recent studies focusing on the high-risk Swedish population found association of short Q tracts with advanced disease but disagreed with respect to overall risk. This may stem from differences in cut-off points – the study with a negative outcome defined short as 22 and long as >22 CAG repeats (Lindstrom et al. 2006b), whereas division of lengths by tertiles resolved a higher risk associated only with tracts 19 Qs (Andersson et al. 2006). Given the complexity of the epidemiological data, AR may be best viewed as a quantitative trait, the effect of which depends on variation at other loci that also vary in populations and are influenced by environment. Thus more compelling data include other genes in the androgen axis, including those encoding enzymes of testosterone synthesis, most notably cytochrome P450 (cyp17), and the enzyme that converts testosterone to the more active dihydrotestosterone, steroid-5-a-reductase type 2 (SRD5A2). Examining AR, cyp17, and SRD5A2 alleles together revealed haplotypes with a significant two-fold greater risk overall (Lindstrom et al. 2006a). This report underscores the multigenic nature of the androgen axis in the development of prostate cancer.

276

D.M. Robins

5 Comparative Biology of Glutamine Tract Variation Insights into complex diseases are often gained from comparison to animal models. Dogs are the only species other than primates known to develop age- and hormonedependent prostate cancer (Cunha et al. 1987). Canine prostate physiology and prostate cancer progression are similar to man, and both have polymorphic AR Q tracts. There are two major CAG repeats in mammalian ARs, with the N-terminal one (CAG-I) more extensive in primates and the C-terminal one (CAG-III, at human amino acid 193) more extensive in rodents (Choong and Wilson 1998; Lu et al. 2001) (Fig. 1). A very short CAG-II just proximal to CAG-I is less variable between species. While most species have preferentially amplified either CAG-I or CAG-II, dogs are unusual in having long tracts at both positions, although CAG-I is shorter than in man. Only primates have an extensive G tract. Given the rich diversity of canine breeds, Q tract variation was assessed in multiple breeds and an association with prostate cancer incidence sought using data on dog prostate cancer incidence from the Veterinary Medical Data Base of Purdue University via collaboration with Dr. Vilma Yuzbasiyan-Gurkan of Michigan State Veterinary Medical Center. Of over 200,000 cases presenting for any medical problem, approximately 0.5% of male dogs had prostate cancer. In addition, incidence varied significantly among breeds. For example, 3.5% of intact male beagles presented with prostate cancer compared to only 0.5% of golden retrievers. Therefore, polymorphism in CAG-I and –III within and between breeds was analyzed for ten individuals of each of five breeds with low prostate cancer incidence (golden retriever, Siberian husky, cocker spaniel, dalmatian, and greyhound) and five breeds with high incidence (beagle, Scottish terrier, Rhodesian ridgeback, doberman pinscher, and Shetland sheepdog). Both canine Q tracts are polymorphic, with 9–12 Qs in CAG-I and 21–24 Qs in CAG-III (Shibuya et al. 1993). Two allele lengths were observed for CAG-I and 4 for CAG-III. While some variation in both tracts occurred within both groups, association of short tracts with tumorigenesis by breed did not reach statistical significance. More compelling data came from analysis of individual dogs with spontaneous disease. Of 13 histologically confirmed prostate cancer samples, ten had the short allele at CAG-I and either the shortest of the four alleles at CAG-III, or an even shorter allele absent from normal dogs. Moreover, two tumors were heterogeneous in CAG-III, displaying yet shorter alleles indicative of contraction during tumor progression. This pilot study supports the hypothesis that ARs with short Q tracts, in men or dogs, confer greater prostate cancer risk, and further, that such receptors arise and are selected for during tumorigenesis. Mice are preferable to dogs as models of human disease pathogenesis because of smaller size, faster generation time, and the homogeneous genetic background of numerous inbred lines. Although mice have a lobular prostate and develop spontaneous prostate cancer rarely, xenograft and germline modification techniques compensate for these deficiencies. All stages of disease can be sampled, in contrast to clinical specimens in which early events are more difficult to observe. However,

The Role of the Androgen Receptor Polyglutamine Tract in Prostate Cancer

277

the mouse Q tract differs significantly from the human (Fig. 1). Rodents have an abbreviated CAG-I and while rats have 20 Qs in CAG-III, the mouse tract is disrupted by histidines (Q8HQ3HQ2HQ4), likely influencing its flexibility as an interdomain spacer. Exclusive of the Q tract, the amino acid sequence of the mouse and human NTDs differ by 15%, in contrast to the DBD and LBD that are identical except for eight amino acids in the hinge region. In transfection experiments, the reduced transactivation strength of the rodent AR relative to human maps to the NTD (Chamberlain et al. 1994). Therefore, to create a mouse model to assess Q tract length effects on phenotypes and disease, genetic engineering was used to ‘‘humanize’’ the mouse AR, by swapping sequences encoding the human NTD into the mouse gene.

6 Development of ‘‘Humanized’’ AR Mice to Study Glutamine Tract Function To convert the mouse AR to the human sequence, targeting vectors for germline homologous recombination were constructed that used a backbone of a 129/Sv mouse genomic fragment encompassing mAR exon 1, which encodes almost the entire NTD. Amino acids 31–484 (including a median length Q tract of 21 residues) from human AR cDNA were substituted for the equivalent mouse sequences via conserved restriction sites, and a neomycin selectable marker was introduced into intron 1, near the 50 splice junction (Albertelli et al. 2006). Additional vectors were constructed to create AR alleles with 12 or 48 Qs, which represented extreme lengths within the normal variation range. These DNAs were electroporated into mouse embryonic stem (ES) cells where homologous recombination occurs via the substantial 50 flanking and intronic chromosomal sequences in the vector. The selectable marker was excised from ES clones by transfection of a plasmid expressing cre-recombinase, which produced a transcription unit with the humanized AR under the control of mouse regulatory sequences. This ‘‘h/mAR’’ sequence differs from human in retaining mouse codons for one N-terminal amino acid, 14 residues just before the DBD (ten of which are conservative changes) and eight amino acids (four conservative) in the hinge region. The targeted ES cells were introduced into mouse blastocysts, and chimeric progeny were used to establish lines of h/mAR mice that have been backcrossed onto the C57BL/6 background. Both sexes of the h/mAR mice were indistinguishable from wild-type mAR littermates in gross physiology including behavior, growth rate, body weight, and lifespan (Albertelli et al. 2006). All three strains were fertile with similar frequency of litters and number of pups per litter. In particular, while longer Q tracts were associated with reduced fertility in man, the 48Q male mice remained as fertile as the other genotypes even at older ages. A preliminary analysis of the androgen axis indicated no significant differences between genotypes in serum testosterone, LH or FSH levels, although individual values varied over a broad range. Whether the

278

D.M. Robins

androgen axis is sensitive to Q tract length will be tested more stringently for mice that have been backcrossed additional generations and are individually housed to eliminate effects of social interaction. When seminal vesicle weight was used as a more direct indicator of androgen action, differences among genotypes became apparent; seminal vesicle weight was lower in h/mAR48Q mice at 6 months and greater by 2 years of age in h/mAR12Q mice. These data confirm an inverse correlation between AR activity and Q tract length, and suggest that minor effects may sum over time to detectable phenotypes. The h/mAR48Q mice were studied further since 48 CAGs may cause partial androgen insensitivity or Kennedy disease in man. However, gene defects often must be more severe in mice than man to model a human syndrome, perhaps due in part to differences in lifespan and, relevant to both trinucleotide repeats and cancer, differences in genomic instability. While transgenic ARs with 65Qs produced no phenotype in mice (Bingham et al. 1995), mice created by introducing 112Qs with our targeting vector replicated the pathology of Kennedy disease (Yu et al. 2006). The h/mAR48Q mice showed no evidence of neuromuscular deficiency, as determined by grip strength tests, even at advanced ages. Even though fertility remained normal, some variation in testis function was discernable at the molecular level. In particular, mRNA levels of Hsd17b3, an indicator of mature Leydig cell function, were significantly lower in 48Q than other mice, although levels were still within a normal expression range. Further, although AR protein levels were similar between genotypes, and no histological evidence of aggregation was observed in AR48Q cells as reported in Kennedy disease, this protein may be predisposed to aggregate since substantial amounts were found in the pellet fraction of whole cell lysates. Therefore, h/mAR48Q, while sufficient for male mouse differentiation and virility, may have suboptimal activity in some circumstances. This characterization confirmed that human and mouse ARs have sufficient conservation, as might their interactions with critical coregulators, to substitute functionally for one another at the organismal level. Furthermore, while minor phenotypic variations were detected for h/mARs with 12 or 48 Qs, these were mostly within the normal range of variation for these traits and did not produce pathology. These alleles in mice may model the extremes of Q tract length within the range of normal alleles similar to those found in man, which allows investigation of whether these variants influence prostate biology or disease. At a gross level, the three h/mAR strains appeared no different from wild-type mice in prostate morphology, with equivalent epithelial and nuclear AR at the immunohistochemical level. At a molecular level, however, differences in AR regulation were detected using quantitative RT-PCR. AR mRNA levels showed a trend to inverse correlation with Q tract length in 6-month-old mice. More significantly, the variation in expression of some AR target genes was consistent with greater transcriptional activity of the 12Q allele and reduced activity of the 48Q allele (Fig. 2). In particular, androgen-regulated probasin, a protein secreted from mature prostatic epithelium, was expressed at higher levels in 12Q than 21Q or 48Q h/mAR mice. Nkx3.1, a prostate-specific and AR-dependent homeobox gene, showed significantly lower levels in h/mAR48Q mice. In contrast, clusterin

The Role of the Androgen Receptor Polyglutamine Tract in Prostate Cancer

279

Fig. 2 Prostate gene expression was influenced by AR Q tract length. Prostate mRNAs from 3 to 8 mice at 6 months of age per h/mAR allele were quantified using real-time RT-PCR. Results are relative to levels in wild-type mAR mice. Probasin and Nkx3.1 were up-regulated by AR, clusterin was repressed. Asterisks indicate significant differences. Modified from Albertelli et al. (2006), Copyright 2006, The Endocrine Society

(an antiapoptotic protein also known as testosterone-repressed prostate message-2), which was down-regulated by AR (July et al. 2002), had higher expression in h/ mAR48Q mice. Q tract length appears to impact not just AR activated but also ARrepressed genes. Overall, differences in target gene expression due to h/mAR allelic variation were small, but consistently trended to greater activity of the 12Q allele and lesser activity of the 48Q allele, for both induction and repression. These results support the hypothesis that Q tract length impacts differential transcription of critical androgen-dependent genes, some of which may impact prostate cancer risk. A broader view of differential regulation by AR alleles was obtained from a preliminary comparison of prostate gene expression using Affymetrix oligonucleotide microarrays. Stringent statistical criteria did not reveal significant differences between AR alleles (including mAR), but when data were analyzed without adjusting for multiple comparisons, some intriguing patterns emerged (Albertelli 2007). For example, differentially expressed genes were frequently at highest levels in h/mAR48Q mice, which corroborated studies that reported more genes were repressed than activated by AR in the prostate (Desai et al. 2004). h/mAR12Q and h/mAR48Q expression often trended in the same direction relative to that of h/mAR21Q, which suggested that for many promoters the median Q tract length provides optimal activation (or repression), with deviation to either longer or shorter tracts having similar effects. Frequently only the 12Q or the 48Q differed from the other two alleles, which suggested that Q tract effects may be promotercontext dependent and encompass multiple mechanisms. Specific transcripts showing differential expression include some previously noted as either androgen-regulated or overexpressed in prostate cancer, such as several involved in inflammatory responses. Of particular interest are a few genes up-regulated in h/mAR48Q mice that are involved in Wnt signaling, especially Wnt5a, whose twofold greater expression was validated using Q-PCR. Crosstalk occurred between

280

D.M. Robins

AR and Wnt signaling pathways and Wnt overexpression has been associated with androgen-stimulated prostate cancer (Yang et al. 2006). In this case, higher expression of Wnt5a in h/mAR48Q mice may be a factor in their development of aggressive castration-recurrent cancer (see later). More definitive analyses using laser capture microdissection will provide additional information on genes and pathways influenced by AR Q tract length.

7 Glutamine Tract Length Effects in Mouse Prostate Cancer h/mAR mice were crossed to the engineered model TRAMP (transgenic adenocarcinoma of the mouse prostate) to investigate the effect of Q tract length variation on oncogenesis. TRAMP mice carry the SV40 T antigen (Tag) oncogene driven by the prostate epithelium-specific probasin promoter and enhancer (Greenberg et al. 1995). TRAMP males develop prostatic intraepithelial neoplasia (PIN) and welldifferentiated prostate cancer that seems histologically similar to the human disease (Kaplan-Lefko et al. 2003). PIN generally develops by 12 weeks of age, and tumors are abdominally palpable somewhat later; metastasis occurs predominantly to lymph nodes, lungs, and liver, but not bone. Development of prostate cancer in TRAMP males is fully penetrant and can be delayed by castration, but ultimately progresses, indicating that disease becomes androgen independent, as in man. While T antigen is a more potent inducer of disease than in human prostate cancer, abrogation of Rb and p53 function occurs in both. TRAMP mice have been invaluable for studying early events in prostate cancer and for evaluating treatment and prevention strategies, although more recent models that incorporate inactivation of the tumor suppressor PTEN may more accurately mimic the origins of human disease (Wang et al. 2003). Regardless of how these models induce cancer and despite the genetic uniformity of mice, there is heterogeneity in time of initiation and rate of progression, which demonstrates the stochastic nature of tumorigenic events. Some of these events are genetic, as they vary with strain background, and some may be epigenetic and subject to diet, inflammation, and hormonal parameters. h/mAR-TRAMP mice and mAR-TRAMP littermates were compared to determine whether differences in Q tract length affected the development of PIN (Albertelli et al. 2008). PIN in mice is the precursor to carcinoma and is characterized by features including epithelial tufting, nuclear hyperchromasia, amphophilic cytoplasm, apoptotic debris, and cribriform architecture (Shappell et al. 2004). At 12 weeks of age, each prostate from ten mice per AR allele had varying levels of PIN in all prostatic lobes. The amount of dorsal epithelium associated with PIN per dorsal lobe was quantified by scoring into categories 1–10, which approximated the percentage of involvement (Fig. 3). A compelling trend was evident, although within each cohort interindividual differences precluded reaching statistical significance. More 12Q mice clustered at higher levels of PIN (70% in category 6 or above), while the majority of 48Q mice exhibited lower PIN

The Role of the Androgen Receptor Polyglutamine Tract in Prostate Cancer

281

Fig. 3 The amount of PIN was inversely proportional to Q tract length. The amount of epithelium involved in PIN at 12 weeks of age was assessed in sections of dorsal lobes, ten prostates per allele, and categorized by approximate % PIN involvement (categories 1–10, where 1 represents about 10% and 10 about 100% PIN involvement). Distribution per genotype showed a trend to inverse correlation with Q tract length (median categories of 7, 6, and 4 for 12Q, 21Q, and 48Q, respectively). Areas of PIN identified by H&E staining were examined in sequential sections for AR and Tag expression using IHC. Overall, AR and Tag expression was similar among Q tract variants, with AR fairly uniform in epithelium and stroma, while Tag expression was highest in areas of PIN. Modified from Albertelli et al. (2008), Copyright 2008, Oxford University Press

levels (60% in category 4). Furthermore, a higher grade of PIN, distinguished by greater expansion of the gland, cribriform proliferation, and more pronounced hyperchromasia, was detected in some of the 12Q and 21Q, but not 48Q, mice. Therefore, AR Q tract length had a notable affect on PIN by 12 weeks of age, with shorter tracts promoting more and higher-grade PIN than longer tracts. The higher levels of PIN may reflect increased proliferation rates or decreased apoptosis, which would also influence subsequent oncogenesis; whether these are affected by Q tract length is under study.

282

D.M. Robins

The h/mAR mice expressed similar levels of prostatic AR mRNA in all genotypes, and immunohistochemistry (IHC) showed similar levels of AR protein in epithelial and stromal nuclei (Albertelli et al. 2006; Albertelli et al. 2008) (Fig. 3). AR in areas of PIN appeared slightly more intense than in normal epithelia, but this likely reflected the higher cell density in these regions. In contrast, although Tag did not show differential staining dependent on genotype, it was more prominent in areas involved in PIN than in normal epithelium. AR allele-dependent differences in Tag expression below the level of detection of IHC may exist, since expression of probasin, whose promoter drives the transgenic oncogene, varied with AR allele in the parental strains. Furthermore, functions of AR in stroma vs. epithelia may enter into higher Tag expression in regions of PIN compared to normal epithelia. Altogether, a cascade of events in neoplastic transformation may enhance Tag expression. In some regards, Tag is analogous to the translocation product TMPRSS2:ETS, in which an androgen-responsive promoter is fused to a transcriptional regulator of proliferation as a common and early event in human prostate oncogenesis (Tomlins et al. 2007). Fusions of this type may maximize the contribution of (or the sensitivity to) the androgen signaling axis in prostate cancer progression. Tumorigenesis was followed in h/mAR-TRAMP mice by abdominal palpation, which detects tumors only a few days later on average then their identification by more costly and time-consuming magnetic resonance imaging (Albertelli et al. 2008). Tumors become palpable at approximately 1 cm in diameter, at midstage disease. Kaplan-Meier analysis revealed a significant difference among genotypes in time to tumor detection (P value <0.0001; Fig. 4), with the median age of detection 10 weeks later for 48Q mice than for the other genotypes. Tumors were initially palpable in 12Q, 21Q, and mAR mice over a similar age range (median: 25 weeks for 12Q and 27 weeks for 21Q), but the kinetics of progression differed among genotypes. Comparing disease status at 29 weeks of age, by which time about half of the 21Q and mAR mice have a detectable tumor or have already died, overt disease was detected in 85% of the 12Q mice, and in less than one-third of the 48Q mice. Age of death further distinguished the genotypes by Kaplan-Meier analysis (P value <0.0001). mAR and 21Q mice survived to mean age of 30 weeks, compared to 42 weeks for 12Q and 55 weeks for 48Q mice. Taken together, the slightly greater PIN levels at 12 weeks, more tumors detectable by 29 weeks, and longer survival relative to 21Q mice suggest that in 12 Q mice tumors initiate earlier but progress more slowly than for the other AR alleles. This is corroborated by comparing ‘‘disease length,’’ noted as the time between initial tumor detection and death. Disease length is longer, and ranges more widely, in 12Q and 48Q mice than 21Q or mAR mice. Given the genetic homogeneity of these mice, stochastic events appear to increase heterogeneity in tumor progression as well as initiation. End-stage tumors were used to explore whether the different AR alleles played a role in heterogeneous tumor progression, using IHC and a tissue microarray. Shorter disease length corresponded to poorly differentiated or undifferentiated tumors with heterogeneous AR staining for all genotypes (Fig. 5). In contrast, the majority of 12Q and 48Q mice that had long disease lengths had well- or moderately

The Role of the Androgen Receptor Polyglutamine Tract in Prostate Cancer

283

Fig. 4 Q tract length affected age of tumor detection and survival in intact mice. The upper panel shows proportion of mice without a palpable tumor and the lower panel shows prostate cancer-free survival; P values are from logrank analysis. Number of mice per group (n) was: mAR = 18, 12Q = 13, 21Q = 19, 48Q = 19. Modified from Albertelli et al. (2008), Copyright 2008, Oxford University Press

differentiated tumors with high levels of AR expressed in most epithelial and stromal cells. IHC results were confirmed using Q-PCR and Western blot analysis of gross tumor samples. For all genotypes, greater differentiation and slower progression was associated with AR levels higher than benign prostate, and was influenced by Q tract length. In contrast, Tag levels were heterogeneous regardless of AR allele, level of expression, or disease length, which corroborated a general uncoupling from AR regulation that was evident in advanced tumors (KaplanLefko et al. 2003). These TRAMP tumors also differed in their extent of neuroendocrine phenotype, as indicated by synaptophysin expression. In man, neuroendocrine differentiation is associated with aggressive, castration-recurrent prostate cancer that mostly lacks AR expression (Shariff and Ather 2006). In TRAMP neuroendocrine cells are also associated with aggressive disease, but they can express AR (Kaplan-Lefko et al. 2003). Most 21Q tumors express moderate levels of both synaptophysin and AR. 48Q tumors showed the least synaptophysin expression, with only about half of the tumors positive for the marker and only half of those also showing AR expression. Tumors from 12Q mice were more similar to the human phenotype; synaptophysin and AR expression were mutually exclusive, with half of the tumors expressing synaptophysin and the other half expressing AR. Overall, synaptophysin expression in the TRAMP tumors, as in human disease, is most often associated with rapid

Fig. 5 Markers of prostate cancer progression in intact h/mAR-TRAMP mice. End-stage tumors were used to create a tissue microarray and sections were stained for AR (above) and synaptophysin (below) (with vertical rows from the same tumor). Two tumors are represented per genotype, the left one represents shorter disease length and the right one longer disease length; disease length in weeks is indicated in the black box in the upper right. 12Q and 48Q tumors of longer disease length had greater differentiation and AR expression, and less synaptophysin expression. Bar = 100 mM. Modified from Albertelli et al. (2008), Copyright 2008, Oxford University Press

284 D.M. Robins

The Role of the Androgen Receptor Polyglutamine Tract in Prostate Cancer

285

progression. A complementary pattern was shown for expression of clusterin, an antiapoptotic protein more often associated with more differentiated tumors (Caporali et al. 2004). Study of this mouse model of AR Q tract length polymorphism confirmed some expectations from epidemiological data. Disease initiation follows a linear trend from earlier occurrence in mice with shorter Q tracts to significant protection for longer Q tracts. Disease progression, however, follows a more complicated course. The effect on initiation likely stems from increased or earlier Tag activation by shorter Q tract ARs that hastens the time until some critical level for neoplasia is reached. Early androgen-dependent events in human prostate cancer may act similarly in the case of TMPRSS2 fusion genes (Tomlins et al. 2007). Tumors in mice with long Q tract ARs initiate later and progress slowly, perhaps modeling indolent disease more commonly found in aging men. In the 12Q mice, tumors progress along divergent paths – some show a more aggressive phenotype similar to that driven by the median Q tract AR, but most exhibit a slow-growing welldifferentiated disease. Some of this apparent complexity may reflect differences in the androgen axis in man and mouse and consequences of this in neoplasia. Since the AR Q tract length affects androgen-sensitive tumorigenesis in mice, h/mAR-TRAMP mice orchiectomized at 12 weeks of age were monitored to test whether this variation also influenced castration-recurrent cancer. Since adrenal androgen synthesis differs in mice and men, castration reduces serum testosterone to very low and perhaps undetectable levels. TRAMP tumors arise from preexisting PIN lesions and can be compared to those in men during androgendeprivation therapy. Remarkably, Q tract length elicited differences in tumor detection and progression in castrated mice, and in directions distinct from those in intact mice (Albertelli et al. 2008). The mice with short Q tract ARs trended to delayed tumor detection by Kaplan-Meier analysis (P value = 0.079) compared to those with median and long Q tracts, which followed a disease course very similar to each other in time to palpation and age of death (Fig. 6). This difference became apparent at 29 weeks of age when palpable tumors were present in only 15% of 12Q mice, whereas about half of the 21Q and 48Q mice had detectable tumors or had died. For all genotypes, the time from tumor palpation until death was shorter following castration than in intact animals, but was significantly longer for castrated 12Q mice than for the other alleles. In some animals, tumors either never became palpable or remained small, and these mice died of metastatic disease found at necropsy. Remarkably, half of the 12Q mice succumb in this manner, compared to only about 10% of the 21Q and 48Q mice. Furthermore, 2 of 14 castrated 12Q mice survived beyond the term of the experiment and were euthanized at the ages of 16 months and 2 years, with no gross evidence of disease upon necropsy. Thus the mice with short Q tracts followed distinct pathways of disease, one distinguished by delayed onset of large primary tumors and the other by late onset of aggressive metastatic disease. The prevalence of metastasis as cause of death in 12Q mice is reminiscent of epidemiological data indicating aggressive disease as a common feature of men with short Q tract ARs, such as found in the African-American population (Bennett et al. 2002).

286

D.M. Robins

Fig. 6 Q tract length inversely affected tumor detection and survival in castrated mice. The upper panel shows proportion of mice without a palpable tumor and the lower panel shows prostate cancer-free survival; P values from the log-rank statistics are shown. Number of mice per group (n) was: 12Q = 13, 21Q = 16, 48Q = 14. Modified from Albertelli et al. (2008), Copyright 2008, Oxford University Press

Tumors from castrated TRAMP mice were mostly undifferentiated by histological analysis, in accord with the short survival time once tumors become palpable. AR IHC staining was modest and heterogeneous, but generally nuclear despite the presumed absence of testosterone. However, the small primary tumors of 12Q mice that died of metastatic disease were more differentiated with cytoplasmic as well as nuclear AR immunostaining that was distinctly higher overall than in the other genotypes (Fig. 7a). These small tumors also failed to express synaptophysin, in contrast to the larger palpable tumors from this group. Interestingly, the 48Q tumors also lacked synaptophysin expression, unlike 21Q tumors, even though these mice had similar ages of tumor detection and death, and similar gross histology. These phenotypic markers further illustrate that alternative courses of tumor progression are influenced by ARs with different Q tract lengths. The effect of AR Q tract length was accentuated by comparing intact and castrated mice of the same genotype. h/mAR21Q-TRAMP mice behaved similar to mAR-TRAMP mice, with a large proportion of those castrated exhibiting delayed tumorigenesis and death, but also with a substantial number developing tumors and succumbing earlier than intact counterparts. These alternative responses to castration were polarized in the 12Q and 48Q mice (Fig. 7b). In 12Q mice, tumor detection was significantly delayed in castrated compared to intact mice (P value

The Role of the Androgen Receptor Polyglutamine Tract in Prostate Cancer

287

Fig. 7 Prostate cancer progression in castrated h/mAR-TRAMP mice. (a) Representative endstage tumors from 12Q and 48Q castrated mice, stained for AR, at low (left) and high (right) magnification. The 12Q tumor (*) was a nonpalpable tumor from a mouse that died of metastatic disease. The 48Q tumor had a 7-week disease length. High magnification showed significant cytoplasmic AR in the 12Q tumor, and more nuclear AR in the 48Q tumor. (b) The response to castration varied with Q tract length. The 12Q AR shifted the balance to tumors developing later after castration, but the 48Q AR responded poorly to treatment. Modified from Albertelli et al. 2008, Copyright 2008, Oxford University Press

<0.0001), although an effect on survival was modest due to the long disease length of intact 12Q mice. In contrast, while 48Q mice showed little difference in time of tumor detection between intact or castrated mice, tumor progression was more rapid following castration and survival times were significantly shorter (P value = 0.011). Thus the short Q tract was associated with a positive response to earlier castration in mice. Androgen-independent cells are thought to exist in TRAMP mice prior to castration, and castration may synchronize these cells and drive selection for aggressive growth (Johnson et al. 2005; Wikstrom et al. 2005). This may be analogous to results of the finasteride prevention trial in man, in which reduced dihydrotestosterone levels decreased the number of prostate cancer cases but more of those occurring were of higher grade (Thompson et al. 2003). Since tumors arise later in 12Q mice following castration, and survival is decreased in 48Q mice by castration, AR strength may impact response to castration, with a stronger AR producing a more favorable outcome. Some studies have noted a similar effect in man, where low testosterone prior to treatment, which may suggest a weaker androgen axis, as in 48Q mice, correlated with poor prognosis (San Francisco

288

D.M. Robins

et al. 2006). Overall, despite differences between the human and murine disease, the 12Q and 48Q mice provide a model in which response to hormonal treatment is genetically programmed. Further analysis may reveal tumor characteristics that could serve as early biomarkers in man for predicting response to androgendeprivation therapy and ultimately suggest distinct strategies for treatment.

8 Conclusions Humanized AR mice with alleles varying in Q tract length provide an in vivo model to test the role of this polymorphism in androgen-dependent traits and in the etiology of disease. In creating these mice, Q tract lengths were chosen at the extremes of those found in the normal human population to optimize detection of informative differences. The resulting strains are indistinguishable from wild-type mice at a gross level, but reveal AR allele-specific variation within the normal range in some physiological and molecular indicators of androgen action. Therefore modest phenotypic alterations, directly or indirectly due to differences in transcriptional activity of the variant Q tract ARs, are detectable and may be cumulative with age. These variations are accentuated in mice by their genetic and environmental homogeneity, but also may exist in man (Zitzmann and Nieschlag 2003). Our data and those of others suggest that they may be most penetrant for alleles with CAG repeat numbers at the extremes of the normal range. Polyglutamine-tract-dependent effects in the h/mAR mice are more pronounced in the context of cancer, with statistically significant differences in tumor detection and survival evident in the TRAMP model. The probasin promoter that drives the Tag oncogene is the likely initial oncogenic sensor of AR strength, but the critical differential gene activation may occur early or in a subset of cells since genotypic differences in Tag expression are not evident at a gross level. Moreover Tag expression is not the sole determinant of disease course, since significant differences occur in progression as well as tumor onset in these strains. Whether early hyperplasia becomes neoplasia depends not only on proliferative capacity of the epithelia cells but also on growth-promoting vs. inhibiting effects of stroma. These opposing forces are influenced overall by the androgen axis, and thus may be modulated directly and indirectly by AR Q tract effects and further exacerbated by neoplasia. Mechanisms that under homeostatic conditions limit the effects of genetic variation (e.g., AR alleles) and environmental change (e.g., hormone levels) may be overcome by the more global dysregulation accompanying oncogenesis. One such mechanism is sequestration of AR by chaperone molecules such as heatshock protein 90. Cryptic genetic variation and subclinical phenotypes may be kept in check by chaperones, but unmasked and amplified by stress or disease states. Cancer progression in TRAMP mice does not show a linear correlation with CAG repeat number, and short-term effects may differ from long term, complexities perhaps also underlying discordant results in man. Our studies suggest similar disease courses in intact 12Q and 48Q mice may occur through distinct mechanisms

The Role of the Androgen Receptor Polyglutamine Tract in Prostate Cancer

289

at the cellular level, due to disparate influences of these alleles on downstream targets involved in proliferation and differentiation. Early in disease, AR maintains its normal function as a tumor suppressor, so increased transcriptional activity associated with a short Q tract slows tumor progression by favoring differentiation. Later in disease, multiple events conspire to switch AR’s function to that of an oncogene and promote tumor growth (Litvinov et al. 2003). Thus reduced activity associated with the long Q tract delays tumor initiation and slows tumor progression in 48Q mice. Q tract variation is likely to affect AR activity differentially in stroma vs. epithelia due to the presence of differing amounts or types of cofactors and the array of active signaling pathways. Comparison of gene expression profiles in these two cell types and at different stages of disease may define multiple pathways and precise mechanisms by which these variant AR alleles influence androgen-dependent prostate cancer. In man, extremes within the normal range may favor divergent pathways of disease progression, although modest differences in CAG repeat number are likely insignificant. Comparison of disease progression in castrated TRAMP mice is complicated by the androgen dependence of the transgenic oncogene’s promoter. Nevertheless, striking differences highlight influences that may also operate in human disease, for instance via TMPRSS2 fusion genes. The divergent responses of the 12Q and 48Q alleles to castration suggest that an effect of Q tract length on ligand-independent AR activation may enter into castration-recurrent disease. Moreover, the slow growth and differentiated phenotype of 12Q tumors relative to the rapid progression of 48Q tumors after castration suggests some residual AR activity may be advantageous in treatment. This benefit pertains largely to the primary tumor in the TRAMP model since androgen-independent and probably AR-negative cells ultimately metastasize. Analysis of Q tract effects in clinical prostate cancer may be more difficult due to this dichotomy between slow tumor growth and aggressive metastasis. Detection of insignificant prostate cancer by PSA testing and disease heterogeneity may confound treatment effect further. These distinct pathways of progression following androgen-deprivation therapy suggest continued sensitivity to AR activity and the androgen axis. The role of AR during androgen-deprivation therapy will be clarified by examining the Q tract variant alleles in the context of an oncogenic model not directly driven by androgen, such as in mice with prostatespecific deletion of the Pten tumor suppressor. In summary, the Q tract variant h/mAR alleles provide a novel genetic paradigm that allows investigators to establish a causal relationship between a polymorphism and a phenotype. Extremes of the normal range of variation produce demonstrable differences in laboratory mice in a uniform genetic and environmental context. Some of these differences may be direct transcriptional effects and some may result indirectly from systemic modulation of the androgen axis. These differences are more apparent, and more complex, in cancer, which suggests that Q tract effects vary with cell type and disease stage. As a risk factor in man, extremes of the normal allelic range may have modest predictive value that take on more significance in combination with certain alleles for androgen metabolizing enzymes. Furthermore, extremes of Q tract length may affect cancer progression, and of more clinical

290

D.M. Robins

relevance, response to androgen-deprivation therapy. Recent evidence from mouse models suggests ablation of AR function may not be the best strategy for prostate cancer, but rather in at least some situations may select for more aggressive ARnegative cells (Johnson et al. 2005; Wikstrom et al. 2005; Banach-Petrosky et al. 2007). Elucidation of mechanisms underlying different progression pathways may ultimately provide diagnostic tools to distinguish which patients will benefit from androgen-deprivation therapy and which may require additional or alternative treatments. Distinguishing downstream AR functions that promote differentiation from those that promote proliferation will lead to development of more targeted treatments. In these endeavors, the Q tract variant mice are a valuable preclinical model for testing therapies directed at the human AR.

Acknowledgments This chapter summarizes work by members of the laboratory over many years, but particularly Megan A. Albertelli, D.V.M., Ph.D., Arno Scheller, Ph.D., Orla A. O’Mahony, Ph.D., and Michele Brogley, M.S. Numerous colleagues providing essential discussions included, from the University of Michigan, Drs. Andrew Lieberman, Mara Steinkamp, and Christopher Krebs; and, from the Fred Hutchinson Cancer Research Center, Norman Greenberg. The Robins laboratory has been supported by the NIH (NIDDK-RO156356, NCI-P50-CA69568, NIH-T32-RR07008) and the DOD (DOD17-02-1-0099, W81XWH05-1-0105), as well as Core support from the University of Michigan Cancer Center (5 P30 CA46592) and the Michigan Diabetes Research and Training Center (5 P60 DK20572).

References Albertelli, M. 2007. Genetic Variation in the Androgen Receptor Impacts Prostate Cancer Initiation and Progression in the Humanized AR Mouse. Department of Human Genetics. Ann Arbor, University of Michigan Medical School. Ph.D. Dissertation. Albertelli, M. A., Scheller, A., Brogley, M., and Robins, D. M. 2006. Replacing the mouse androgen receptor with human alleles demonstrates glutamine tract length-dependent effects on physiology and tumorigenesis in mice. Mol. Endocrinol. 20:1248–60. Albertelli, M. A., O’Mahony, O. A., Brogley, M., Tosoian, J., Steinkamp, M., Daignault, S., Wojno, K., and Robins, D. M. 2008. Glutamine tract length of human androgen receptors affects hormone-dependent and -independent prostate cancer in mice. Hum. Mol. Genet. 17:98–110. Alvarado, C., Beitel, L. K., Sircar, K., Aprikian, A., Trifiro, M., and Gottlieb, B. 2005. Somatic mosaicism and cancer: a micro-genetic examination into the role of the androgen receptor gene in prostate cancer. Cancer Res. 65:8514–8. Andersson, P., Varenhorst, E., and Soderkvist, P. 2006. Androgen receptor and vitamin D receptor gene polymorphisms and prostate cancer risk. Eur. J. Cancer 42:2833–7. Banach-Petrosky, W., Jessen, W. J., Ouyang, X., Gao, H., Rao, J., Quinn, J., Aronow, B. J., and Abate-Shen, C. 2007. Prolonged exposure to reduced levels of androgen accelerates prostate cancer progression in Nkx3.1; Pten mutant mice. Cancer Res. 67:9089–96. Beilin, J., Harewood, L., Frydenberg, M., Mameghan, H., Martyres, R. F., Farish, S. J., Yue, C., Deam, D. R., Byron, K. A., and Zajac, J. D. 2001. A case-control study of the androgen receptor gene CAG repeat polymorphism in Australian prostate carcinoma subjects. Cancer 92:941–9.

The Role of the Androgen Receptor Polyglutamine Tract in Prostate Cancer

291

Bennett, C. L., Price, D. K., Kim, S., Liu, D., Jovanovic, B. D., Nathan, D., Johnson, M. E., Montgomery, J. S., Cude, K., Brockbank, J. C., Sartor, O., and Figg, W. D. 2002. Racial variation in CAG repeat lengths within the androgen receptor gene among prostate cancer patients of lower socioeconomic status. J. Clin. Oncol. 20:3599–604. Bingham, P. M., Scott, M. O., Wang, S., McPhaul, M. J., Wilson, E. M., Garbern, J. Y., Merry, D. E., and Fischbeck, K. H. 1995. Stability of an expanded trinucleotide repeat in the androgen receptor gene in transgenic mice. Nat. Genet. 9:191–6. Bratt, O., Borg, A., Kristoffersson, U., Lundgren, R., Zhang, Q. X., and Olsson, H. 1999. CAG repeat length in the androgen receptor gene is related to age at diagnosis of prostate cancer and response to endocrine therapy, but not to prostate cancer risk. Br. J. Cancer 81:672–6. Buchanan, G., Yang, M., Cheong, A., Harris, J. M., Irvine, R. A., Lambert, P. F., Moore, N. L., Raynor, M., Neufing, P. J., Coetzee, G. A., and Tilley, W. D. 2004. Structural and functional consequences of glutamine tract variation in the androgen receptor. Hum. Mol. Genet. 13:1677–92. Caporali, A., Davalli, P., Astancolle, S., D’Arca, D., Brausi, M., Bettuzzi, S., and Corti, A. 2004. The chemopreventive action of catechins in the TRAMP mouse model of prostate carcinogenesis is accompanied by clusterin over-expression. Carcinogenesis 25:2217–24. Casella, R., Maduro, M. R., Lipshultz, L. I., and Lamb, D. J. 2001. Significance of the polyglutamine tract polymorphism in the androgen receptor. Urology 58:651–6. Chamberlain, N. L., Driver, E. D., and Miesfeld, R. L. 1994. The length and location of CAG trinucleotide repeats in the androgen receptor N-terminal domain affect transactivation function. Nucleic Acids Res. 22:3181–6. Chang, B. L., Zheng, S. L., Hawkins, G. A., Isaacs, S. D., Wiley, K. E., Turner, A., Carpten, J. D., Bleecker, E. R., Walsh, P. C., Trent, J. M., Meyers, D. A., Isaacs, W. B., and Xu, J. 2002. Polymorphic GGC repeats in the androgen receptor gene are associated with hereditary and sporadic prostate cancer risk. Hum. Genet. 110:122–9. Choong, C. S., and Wilson, E. M. 1998. Trinucleotide repeats in the human androgen receptor: a molecular basis for disease. J. Mol. Endocrinol. 21:235–57. Coetzee, G. A., and Ross, R. K. 1994. Re: Prostate Cancer and the Androgen Receptor. J. Nat. Cancer Inst. 86:872–873. Crabbe, P., Bogaert, V., De Bacquer, D., Goemaere, S., Zmierczak, H., and Kaufman, J. M. 2007. Part of the interindividual variation in serum testosterone levels in healthy men reflects differences in androgen sensitivity and feedback set point: contribution of the androgen receptor polyglutamine tract polymorphism. J. Clin. Endocrinol. Metab. 92:3604–10. Cude, K. J., Montgomery, J. S., Price, D. K., Dixon, S. C., Kincaid, R. L., Kovacs, K. F., Venzon, D. J., Liewehr, D. J., Johnson, M. E., Reed, E., and Figg, W. D. 2002. The role of an androgen receptor polymorphism in the clinical outcome of patients with metastatic prostate cancer. Urol. Int. 68:16–23. Cunha, G. R., Donjacour, A. A., Cooke, P. S., Mee, S., Bigsby, R. M., Higgins, S. J., and Sugimura, Y. 1987. The endocrinology and developmental biology of the prostate. Endocrinol. Rev. 8:338–62. Davis-Dao, C. A., Tuazon, E. D., Sokol, R. Z., and Cortessis, V. K. 2007. Male Infertility and Variation in CAG Repeat Length in the Androgen Receptor Gene: A Meta-analysis. J. Clin. Endocrinol. Metab. 92:4319–26. Dehm, S. M., and Tindall, D. J. 2007. Androgen receptor structural and functional elements: role and regulation in prostate cancer. Mol. Endocrinol. 21:2855–63. Desai, K. V., Michalowska, A. M., Kondaiah, P., Ward, J. M., Shih, J. H., and Green, J. E. 2004. Gene expression profiling identifies a unique androgen-mediated inflammatory/immune signature and a PTEN (phosphatase and tensin homolog deleted on chromosome 10)-mediated apoptotic response specific to the rat ventral prostate. Mol. Endocrinol. 18:2895–907. Ding, D., Xu, L., Menon, M., Reddy, G. P., and Barrack, E. R. 2004. Effect of a short CAG (glutamine) repeat on human androgen receptor function. Prostate 58:23–32.

292

D.M. Robins

Duff, J., Davies, P., Watt, K., and McEwan, I. J. 2006. Structural dynamics of the human androgen receptor: implications for prostate cancer and neurodegenerative disease. Biochem. Soc. Trans. 34:1098–102. Edwards, A., Hammond, H. A., Jin, L., Caskey, C. T., and Chakraborty, R. 1992. Genetic variation at five trimeric and tetrameric tandem repeat loci in four human population groups. Genomics 12:241–53. Edwards, S. M., Badzioch, M. D., Minter, R., Hamoudi, R., Collins, N., Ardern-Jones, A., Dowe, A., Osborne, S., Kelly, J., Shearer, R., Easton, D. F., Saunders, G. F., Dearnaley, D. P., and Eeles, R. A. 1999. Androgen receptor polymorphisms: association with prostate cancer risk, relapse and overall survival. Int. J. Cancer 84:458–65. Freedman, M. L., Pearce, C. L., Penney, K. L., Hirschhorn, J. N., Kolonel, L. N., Henderson, B. E., and Altshuler, D. 2005. Systematic evaluation of genetic variation at the androgen receptor locus and risk of prostate cancer in a multiethnic cohort study. Am. J. Hum. Genet. 76:82–90. Gerber, H. P., Seipel, K., Georgiev, O., Hofferer, M., Hug, M., Rusconi, S., and Schaffner, W. 1994. Transcriptional activation modulated by homopolymeric glutamine and proline stretches. Science 263:808–11. Giovannucci, E. 2002. Is the androgen receptor CAG repeat length significant for prostate cancer? Cancer Epidemiol. Biomarkers Prev. 11:985–6. Giovannucci, E., Stampfer, M. J., Krithivas, K., Brown, M., Dahl, D., Brufsky, A., Talcott, J., Hennekens, C. H., and Kantoff, P. W. 1997. The CAG repeat within the androgen receptor gene and its relationship to prostate cancer. Proc. Natl. Acad. Sci. U.S.A. 94:3320–3. Greenberg, N. M., DeMayo, F., Finegold, M. J., Medina, D., Tilley, W. D., Aspinall, J. O., Cunha, G. R., Donjacour, A. A., Matusik, R. J., and Rosen, J. M. 1995. Prostate cancer in a transgenic mouse. Proc. Natl. Acad. Sci. U.S.A. 92:3439–43. Hardy, D. O., Scher, H. I., Bogenreider, T., Sabbatini, P., Zhang, Z. F., Nanus, D. M., and Catterall, J. F. 1996. Androgen receptor CAG repeat lengths in prostate cancer: correlation with age of onset. J. Clin. Endocrinol. Metab. 81:4400–5. He, B., and Wilson, E. M. 2002. The NH(2)-terminal and carboxyl-terminal interaction in the human androgen receptor. Mol. Genet. Metab. 75:293–8. Hsiao, P. W., Lin, D. L., Nakao, R., and Chang, C. 1999. The linkage of Kennedy’s neuron disease to ARA24, the first identified androgen receptor polyglutamine region-associated coactivator. J. Biol. Chem. 274:20229–34. Hsing, A. W., Gao, Y. T., Wu, G., Wang, X., Deng, J., Chen, Y. L., Sesterhenn, I. A., Mostofi, F. K., Benichou, J., and Chang, C. 2000. Polymorphic CAG and GGN repeat lengths in the androgen receptor gene and prostate cancer risk: a population-based case-control study in China. Cancer Res. 60:5111–6. Irvine, R. A., Yu, M. C., Ross, R. K., and Coetzee, G. A. 1995. The CAG and GGC microsatellites of the androgen receptor gene are in linkage disequilibrium in men with prostate cancer. Cancer Res. 55:1937–40. Irvine, R. A., Ma, H., Yu, M. C., Ross, R. K., Stallcup, M. R., and Coetzee, G. A. 2000. Inhibition of p160-mediated coactivation with increasing androgen receptor polyglutamine length. Hum. Mol. Genet. 9:267–74. Johnson, M. A., Iversen, P., Schwier, P., Corn, A. L., Sandusky, G., Graff, J., and Neubauer, B. L. 2005. Castration triggers growth of previously static androgen-independent lesions in the transgenic adenocarcinoma of the mouse prostate (TRAMP) model. Prostate 62:322–38. July, L. V., Akbari, M., Zellweger, T., Jones, E. C., Goldenberg, S. L., and Gleave, M. E. 2002. Clusterin expression is significantly enhanced in prostate cancer cells following androgen withdrawal therapy. Prostate 50:179–88. Kaplan-Lefko, P. J., Chen, T. M., Ittmann, M. M., Barrios, R. J., Ayala, G. E., Huss, W. J., Maddison, L. A., Foster, B. A., and Greenberg, N. M. 2003. Pathobiology of autochthonous prostate cancer in a pre-clinical transgenic mouse model. Prostate 55:219–37.

The Role of the Androgen Receptor Polyglutamine Tract in Prostate Cancer

293

La Spada, A. R., Wilson, E. M., Lubahn, D. B., Harding, A. E., and Fischbeck, K. H. 1991. Androgen receptor gene mutations in X-linked spinal and bulbar muscular atrophy. Nature 352:77–9. Li, C., Gronberg, H., Matsuyama, H., Weber, G., Nordenskjold, M., Naito, K., Bergh, A., Bergerheim, U., Damber, J. E., Larsson, C., and Ekman, P. 2003. Difference between Swedish and Japanese men in the association between AR CAG repeats and prostate cancer suggesting a susceptibility-modifying locus overlapping the androgen receptor gene. Int. J. Mol. Med. 11:529–33. Lichtenstein, P., Holm, N. V., Verkasalo, P. K., Iliadou, A., Kaprio, J., Koskenvuo, M., Pukkala, E., Skytthe, A., and Hemminki, K. 2000. Environmental and heritable factors in the causation of cancer – analyses of cohorts of twins from Sweden, Denmark, and Finland. N. Engl. J. Med. 343:78–85. Lieberman, A. P., and Robins, D. M. in press. The androgen receptor’s CAG/glutamine tract in mouse models of neurological disease and cancer. J. Alzheimers Dis. 14:247–255. Lindstrom, S., Wiklund, F., Adami, H. O., Balter, K. A., Adolfsson, J., and Gronberg, H. 2006a. Germ-line genetic variation in the key androgen-regulating genes androgen receptor, cytochrome P450, and steroid-5-alpha-reductase type 2 is important for prostate cancer development. Cancer Res. 66:11077–83. Lindstrom, S., Zheng, S. L., Wiklund, F., Jonsson, B. A., Adami, H. O., Balter, K. A., Brookes, A. J., Sun, J., Chang, B. L., Liu, W., Li, G., Isaacs, W. B., Adolfsson, J., Gronberg, H., and Xu, J. 2006b. Systematic replication study of reported genetic associations in prostate cancer: Strong support for genetic variation in the androgen pathway. Prostate 66:1729–43. Litvinov, I. V., De Marzo, A. M., and Isaacs, J. T. 2003. Is the Achilles’ heel for prostate cancer therapy a gain of function in androgen receptor signaling? J. Clin. Endocrinol. Metab. 88:2972–82. Lu, B., Smock, S. L., Castleberry, T. A., and Owen, T. A. 2001. Molecular cloning and functional characterization of the canine androgen receptor. Mol. Cell. Biochem. 226:129–40. Mankodi, A., Takahashi, M. P., Jiang, H., Beck, C. L., Bowers, W. J., Moxley, R. T., Cannon, S. C., and Thornton, C. A. 2002. Expanded CUG repeats trigger aberrant splicing of ClC-1 chloride channel pre-mRNA and hyperexcitability of skeletal muscle in myotonic dystrophy. Mol. Cell 10:35–44. McPhaul, M. J. 2002. Androgen receptor mutations and androgen insensitivity. Mol. Cell. Endocrinol. 198:61–7. Mhatre, A. N., Trifiro, M. A., Kaufman, M., Kazemi-Esfarjani, P., Figlewicz, D., Rouleau, G., and Pinsky, L. 1993. Reduced transcriptional regulatory competence of the androgen receptor in Xlinked spinal and bulbar muscular atrophy. Nat. Genet. 5:184–8. Perutz, M. F., Johnson, T., Suzuki, M., and Finch, J. T. 1994. Glutamine repeats as polar zippers: their possible role in inherited neurodegenerative diseases. Proc. Natl. Acad. Sci. U.S.A. 91:5355–8. Pratt, W. B., Galigniana, M. D., Morishima, Y., and Murphy, P. J. 2004. Role of molecular chaperones in steroid receptor action. Essays Biochem. 40:41–58. Robins, D. M. 2004. Multiple mechanisms of male-specific gene expression: lessons from the mouse sex-limited protein (Slp) gene. Prog. Nucleic Acid Res. Mol. Biol. 78:1–36. San Francisco, I. F., Regan, M. M., Dewolf, W. C., and Olumi, A. F. 2006. Low age adjusted free testosterone levels correlate with poorly differentiated prostate cancer. J. Urol. 175:1341–5; discussion 1345–6. Santos, M. L., Sarkis, A. S., Nishimoto, I. N., and Nagai, M. A. 2003. Androgen receptor CAG repeat polymorphism in prostate cancer from a Brazilian population. Cancer Detect. Prev. 27:321–6. Schoenberg, M. P., Hakimi, J. M., Wang, S., Bova, G. S., Epstein, J. I., Fischbeck, K. H., Isaacs, W. B., Walsh, P. C., and Barrack, E. R. 1994. Microsatellite mutation (CAG24 –> 18) in the androgen receptor gene in human prostate cancer. Biochem. Biophys. Res. Commun. 198:74–80.

294

D.M. Robins

Shappell, S. B., Thomas, G. V., Roberts, R. L., Herbert, R., Ittmann, M. M., Rubin, M. A., Humphrey, P. A., Sundberg, J. P., Rozengurt, N., Barrios, R., Ward, J. M., and Cardiff, R. D. 2004. Prostate pathology of genetically engineered mice: definitions and classification. The consensus report from the Bar Harbor meeting of the Mouse Models of Human Cancer Consortium Prostate Pathology Committee. Cancer Res. 64:2270–305. Shariff, A. H., and Ather, M. H. 2006. Neuroendocrine differentiation in prostate cancer. Urology 68:2–8. Shen, H. C., and Coetzee, G. A. 2005. The androgen receptor: unlocking the secrets of its unique transactivation domain. Vitam. Horm. 71:301–19. Shibuya, H., Nonneman, D. J., Huang, T. H., Ganjam, V. K., Mann, F. A., and Johnson, G. S. 1993. Two polymorphic microsatellites in a coding segment of the canine androgen receptor gene. Anim. Genet. 24:345–8. Shimbo, M., Suzuki, H., Kamiya, N., Imamoto, T., Komiya, A., Ueda, T., Watanabe, M., Shiraishi, T., and Ichikawa, T. 2005. CAG polymorphic repeat length in androgen receptor gene combined with pretreatment serum testosterone level as prognostic factor in patients with metastatic prostate cancer. Eur. Urol. 47:557–63. Sieh, W., Edwards, K. L., Fitzpatrick, A. L., Srinouanprachanh, S. L., Farin, F. M., Monks, S. A., Kronmal, R. A., and Eaton, D. L. 2006. Genetic Susceptibility to Prostate Cancer: Prostatespecific Antigen and its Interaction with the Androgen Receptor (United States). Cancer Causes Control 17:187–97. Stanford, J. L., Just, J. J., Gibbs, M., Wicklund, K. G., Neal, C. L., Blumenstein, B. A., and Ostrander, E. A. 1997. Polymorphic repeats in the androgen receptor gene: molecular markers of prostate cancer risk. Cancer Res. 57:1194–8. Suzuki, H., Akakura, K., Komiya, A., Ueda, T., Imamoto, T., Furuya, Y., Ichikawa, T., Watanabe, M., Shiraishi, T., and Ito, H. 2002. CAG polymorphic repeat lengths in androgen receptor gene among Japanese prostate cancer patients: potential predictor of prognosis after endocrine therapy. Prostate 51:219–24. Suzuki, H., Ueda, T., Ichikawa, T., and Ito, H. 2003. Androgen receptor involvement in the progression of prostate cancer. Endocr. Relat. Cancer 10:209–16. Thompson, I. M., Goodman, P. J., Tangen, C. M., Lucia, M. S., Miller, G. J., Ford, L. G., Lieber, M. M., Cespedes, R. D., Atkins, J. N., Lippman, S. M., Carlin, S. M., Ryan, A., Szczepanek, C. M., Crowley, J. J., and Coltman, C. A., Jr. 2003. The influence of finasteride on the development of prostate cancer. N. Engl. J. Med. 349:215–24. Tomlins, S. A., Mehra, R., Rhodes, D. R., Cao, X., Wang, L., Dhanasekaran, S. M., KalyanaSundaram, S., Wei, J. T., Rubin, M. A., Pienta, K. J., Shah, R. B., and Chinnaiyan, A. M. 2007. Integrative molecular concept modeling of prostate cancer progression. Nat. Genet. 39:41–51. Wang, S., Gao, J., Lei, Q., Rozengurt, N., Pritchard, C., Jiao, J., Thomas, G. V., Li, G., RoyBurman, P., Nelson, P. S., Liu, X., and Wu, H. 2003. Prostate-specific deletion of the murine Pten tumor suppressor gene leads to metastatic prostate cancer. Cancer Cell 4:209–21. Wang, Q., Udayakumar, T. S., Vasaitis, T. S., Brodie, A. M., and Fondell, J. D. 2004. Mechanistic relationship between androgen receptor polyglutamine tract truncation and androgen-dependent transcriptional hyperactivity in prostate cancer cells. J. Biol. Chem. 279:17319–28. Wikstrom, P., Lindahl, C., and Bergh, A. 2005. Characterization of the autochthonous transgenic adenocarcinoma of the mouse prostate (TRAMP) as a model to study effects of castration therapy. Prostate 62:148–64. Yang, X., Chen, M. W., Terry, S., Vacherot, F., Bemis, D. L., Capodice, J., Kitajewski, J., de la Taille, A., Benson, M. C., Guo, Y., and Buttyan, R. 2006. Complex regulation of human androgen receptor expression by Wnt signaling in prostate cancer cells. Oncogene 25:3436–44. Yeap, B. B., Wilce, J. A., and Leedman, P. J. 2004. The androgen receptor mRNA. Bioessays 26:672–82. Yong, E. L., Loy, C. J., and Sim, K. S. 2003. Androgen receptor gene and male infertility. Hum. Reprod. Update 9:1–7.

The Role of the Androgen Receptor Polyglutamine Tract in Prostate Cancer

295

Yu, Z., Dadgar, N., Albertelli, M., Gruis, K., Jordan, C., Robins, D. M., and Lieberman, A. P. 2006. Androgen-dependent pathology demonstrates myopathic contribution to the Kennedy disease phenotype in a mouse knock-in model. J. Clin. Invest. 116:2663–72. Zeegers, M. P., Kiemeney, L. A., Nieder, A. M., and Ostrer, H. 2004. How strong is the association between CAG and GGN repeat length polymorphisms in the androgen receptor gene and prostate cancer risk? Cancer Epidemiol. Biomarkers Prev. 13:1765–71. Zitzmann, M., and Nieschlag, E. 2003. The CAG repeat polymorphism within the androgen receptor gene and maleness. Int. J. Androl. 26:76–83. Zitzmann, M., Depenbusch, M., Gromoll, J., and Nieschlag, E. 2004. X-chromosome inactivation patterns and androgen receptor functionality influence phenotype and social characteristics as well as pharmacogenetics of testosterone therapy in Klinefelter patients. J. Clin. Endocrinol. Metab. 89:6208–17.

The Androgen Receptor Coactivator-Binding Interface Eva Este´banez-Perpin˜a´ and Robert J. Fletterick

Abstract When hormone binds to the androgen receptor (AR), the ligand-binding domain (LBD) becomes ordered, displaying a new protein–protein interaction surface called AF2 (coactivator-binding pocket), which is a hydrophobic groove that fits AR coregulators. The association of coregulators with AR LBD is often a critical step for its transcriptional function. Existing pharmaceuticals block AR activity by disrupting AF2 surface’s ability to recruit coactivators. Such antagonists bind to the hormone-binding site inside the LBD core and perturb the structure of the most terminal helix of the LBD, distorting the AF2 surface. The AF2 pocket is also a potential candidate for pharmaceutical intervention by surface-directed small molecules that will directly block coactivator recruitment. Such molecules may be a novel generation of antiandrogens for treating prostate cancer.

1 Introduction The science addressed in this chapter is one snapshot in the cine´ma verite´ of the androgen receptor (AR) function: the nature of the interaction between its coregulatory partners and the coactivator-binding pocket present in its ligand-binding domain (LBD). The LBD is the heart of AR not only because it binds to steroid hormones, but because it also determines interactions with chaperones and partner transcription factors (Prescott and Coetzee 2006) (Fig. 1). It is believed that apo-AR (receptor in the absence of hormone), like other steroid receptors, is partly folded, and hormone binding is the pivotal folding key that triggers the formation of coregulator interaction surfaces (Feng et al. 1998). When hormone binds to the receptor, unknown conformational changes follow that drive disassembly of the AR-chaperone complex (Aranda and Pascual 2001). One key change in structure, hypothesized by R.J. Fletterick(*) Department of Biochemistry & Biophysics, University of California San Francisco, San Francisco, CA 94143-2240, USA, E-mail: [email protected]

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179_4_13, # Springer Science + Business Media, LLC 2009

297

298

E. Este´banez-Perpin˜a´, R.J. Fletterick

Fig. 1 Hypothesized AR complexes. AR domains are represented as LBD (medium oval), DBD (small gray circle oval), and NTD (big gray oval). AR forms different macromolecular complexes depending on what ligand is present in its LBP. (a) In the apo-AR absence of ligand is complexed to chaperones (larger light gray sphere). (b) Upon DHT binding (small diamond), AR undergoes a series of conformational changes and the chaperones (HSP, smaller light gray sphere) are released; then AR is targeted to the nucleus and is (c) able to form productive complexes with coactivators (i.e., p160 family, larger oval). (d) After coactivator binding to AR AF2 pocket, recruitment of the transcriptional machinery (RNA POL II ) occurs with transcription or AR-dependent genes (DNA helix shown as schematic double helix). (e) When AR LBP binds antiandrogens (lightning), a series of conformational changes disrupt AF2, impeding coactivator recruitment. (f) AR-coactivator recruitment can be blocked by AF2- interacting compounds (star). (g) Allosteric modulators of AF2 may also block AR action

analogy with other NRs (Bourguet et al. 2000; Moras and Gronemeyer 1998; Wurtz et al. 1996), is a conformational change of the C-terminal helix 12 (H12) present in its LBD. Upon agonist binding, H12 and the C-terminal 20 amino acids (F domain) of AR fold against the body of the LBD (Nichols et al. 1998). In its new position, the F domain blocks the dimer assembly site known for other NRs, and H12 completes the hydrophobic coactivator-binding pocket (also known as AF2 site,

The Androgen Receptor Coactivator-Binding Interface

299

Fig. 2 Structure of AR LBD. (a) AR LBD surface is represented in gray with the residues lining the AF-2 pocket highlighted in darker gray. (b) Secondary structure representation of AR LBD in gray with the buried DHT shown as a blue space-filling model. The position of helices 3, 5, and 12 is shown

Fig. 3 The AR AF2 pocket is a trifurcated groove formed by residues of helices H3 (Lys717, Val713), H5 (Asn733, Val730, Met734), and H12 (Glu897, Glu893, Met894). The floor of the pocket is formed by residues Phe725, Ile737, Asn738, Ile898, and Leu712

for trans-activation function 2), which is a charge-capped hydrophobic groove (Figs. 2–4) (Darimont 2003; Este´banez-Perpin˜a´ et al. 2005). This process is different from what happens when ligands bind to enzymes, which typically possess preformed and solvent accessible active sites. Binding of an agonist to AR completes the hydrophobic core of the receptor, reshaping its surface, which in turn

300

E. Este´banez-Perpin˜a´, R.J. Fletterick

Fig. 4 LXXLL (SRC2-3) and FXXLL (ARA70) peptides bound to AR LBD AF2 pocket. AR LBD surface is colored in pale gray. (a) AR LBD surface is shown in the absence of the LXXLL peptide. The three subsites S1, S2, and S3 are indicated. (b) Three hydrophobic residues of the hydrophobic motif of the SRC2-2 peptide, Leu 1 (L1), Leu 4 (L4), and Leu 5 (L5), are shown as sticks and spheres. The three leucine side chains of this biomotif interact with the correspondent subpockets of the AF2 groove. (c) SRC2-2 peptide is shown as sticks and spheres. (d) SRC22 peptide is shown as a ribbon running along the AR AF2 pocket surface. (e) AR LBD surface is shown in the absence of the FXXLF peptide. (f) The two phenylalanines and leucine side chains of the FXXLF biomotif (shown as sticks and spheres) interact with the correspondent subsites of the AF2 groove. (g) ARA70 peptide is shown as sticks and spheres. (h) ARA70 peptide is shown as a ribbon running along the AR AF2 pocket surface

allows assembly of coregulators on its surfaces (Este´banez-Perpin˜a´ et al. 2005; Hur et al. 2004). Both the N-terminal domain (NTD) and LBD recruit partnering proteins (Bevan et al. 1999; Zhou et al. 2002; McKenna et al. 1999; Hodgson et al. 2005; Lonard and O’Malley 2006). A curious confusion relates to competition for coactivator binding to the AR AF2 pocket. The AR NTD itself binds coactivators but also engages in a hormone-dependent interaction with its LBD, and this intramolecular interaction is known to regulate AR transcriptional activity by competing with coactivator recruitment to the LBD (He and Wilson 2002; Hsu et al. 2005). Apparently, both the NTD and coactivators bind to the AF2 surface (He et al. 2002). Coactivators may alter the self-assembly of AR and compete with AR NTD for binding to the AF2 groove (He et al. 2001). The NTD-LBD interaction also affects AR binding to chromatin (Li et al. 2006). Coregulators comprise coactivators and corepressors, which are structurally and functionally diverse proteins (Lonard and O’Malley 2006, 2007; McKenna and O’Malley 2002; Wang et al. 2005). Corepressors are believed to bind to the unliganded receptor (apo-AR) or to the antagonist-bound state, but are released from the receptor upon hormone binding (agonist-bound AR) (Lazar 2003; Burd et al. 2006). Several members of the p160 family (SRC, steroid receptor coactivators) and ARA family (androgen-receptor-associated proteins) of coactivators are

The Androgen Receptor Coactivator-Binding Interface

301

involved in normal transcription and in the development and progression of prostate cancer (Chmelar et al. 2007; Yeh and Chang 1996; Agoulnik et al. 2005). Once bound to AR, coactivators are part of multiprotein complexes that include the transcriptional machinery (Shang et al. 2002; Choudhry et al. 2006) (Fig. 1). These multisubunit complexes contain many different enzymatic activities such as ubiquitylation, chromatin remodeling, or DNA methylation among others (Lonard and O’Malley 2006, 2007). Several lines of evidence support the idea that AR AF2 pocket can be a bona fide pharmaceutical target (Arnold et al. 2005, 2006, 2007; Este´banez-Perpin˜a´ and Fletterick 2007; Chang and McDonnell 2005). Small molecules have been reported to bind weakly to AR AF2 by Xray crystallography and fluorescent polarization assays (Este´banez-Perpin˜a´ et al. 2007a). Such compounds would serve as starting point chemical scaffolds to design novel antiandrogens. Such antiandrogens will be surface directed and would work as either steric blockers of the AF2 pocket or as allosteric modulators, reshaping the AF2 surface or adjacent putative regulatory surfaces (Este´banez-Perpin˜a´ and Fletterick 2007; Este´banez-Perpin˜a´ et al. 2007a).

2 AR Coactivator-Binding Pocket The LBD is the best understood of AR functional domains (Table 1, Fig. 2). The elucidation of the crystal structure of AR LBD revealed the structural determinants of AR agonist recognition (Matias et al. 2000; Sack et al. 2001; Askew et al. 2007), AR antagonist resistance development (Bohl et al. 2005a, b), coactivator recognition and recruitment to AR surface (Este´banez-Perpin˜a´ et al. 2005; Hur et al. 2004; He et al. 2004, 2006) (Fig. 2). There is no structure of AR LBD in either the apoform, in the antagonist-bound conformation, or as a part of the multidomain/fulllength AR. All AR LDB structures deposited in the Protein Data Bank (PDB) for either wild-type or mutant AR LBD are in the agonist-bound conformation (Fig. 2). A structure of AR DBD domain in complex with its hormone-response element has also been reported (Shaffer et al. 2004). Two ligand-binding sites on the AR LBD contribute to successful AR function. First is the ligand-binding pocket (LBP), which is formally defined as an allosteric site since it ‘‘builds’’ the active site located some distance away. The LBP contains embedded the hydrophobic hormone: testosterone or dihydrotestosterone (DHT). Second, the coactivator-binding pocket (AF2) located at the surface, which is the true active site (Fig. 1). Recruitment of coactivator proteins to the AF2 pocket is a crucial step in AR function. The AR AF2 groove is the most well-defined solvent-exposed protein–protein interaction site on the AR surface (Figs. 2–4). Peptides derived from specific coactivators represent good models for studying binding of coregulators to AF2. Several crystal structures are known for coactivator peptides bound to AR AF2 pocket at atomic resolution (Este´banez-Perpin˜a´ et al. 2005; Hur et al. 2004; Askew et al. 2007; He et al. 2004, 2006). These structures reveal the molecular

302

E. Este´banez-Perpin˜a´, R.J. Fletterick

Table 1 Structures of the androgen receptor (DBD and LBD) with atomic coordinates deposited in the Protein Data Bank Protein PDB code Comments Reference AR (P10275) 1I37, 1I38 Rat, wf/T877A mutant, DHT, Sack et al. (2001) ˚ 2.0 A ˚ 1XNN Rat, T877A mutant, HYQ, 2.2 A Salvati et al. (2005) ˚ 1E3G Rat, Metribolone (R1881), 2.4 A Matias et al. (2000) 1GS4 Human AR mutant Matias et al. (2002) ˚ (ARccr), 1.95 A 1R4I DBD bound to a direct HRE Shaffer et al. (2004) ˚ 1T73 DHT, FXXFF motif, 2.2 A Hur et al. (2004) ˚ 1T74 DHT, WXXLF motif, 2.0 A Hur et al. (2004) ˚ 1T76 DHT, WXXVW motif, 2.1 A Hur et al. (2004) ˚ 1T79 DHT, FXXLW motif, 1.8 A Hur et al. (2004) ˚ 1T7F DHT, LXXLL motif, 1.6 A Hur et al. (2004) ˚ 1T7M DHT, FXXYF motif, 1.6 A Hur et al. (2004) ˚ 1T7R DHT, FXXLF motif, 1.4 A Hur et al. (2004) ˚ 1T7T DHT, no peptide, 1.7 A Hur et al. (2004) ˚ 1XOW R1881, NTD peptide, 1.8 A He et al. (2004) ˚ 1XQ3 R1881, 2.25 A He et al. (2004) ˚ 2A06 R1881, Tif2(Iii) peptide, 1.89 A He et al. (2004) ˚ 1T5Z DHT, ARA70 peptide, 2.3 A Este´banez-Perpin˜a´ et al. (2005) ˚ 1T63 DHT, SRC2-3 peptide, 2.07 A Este´banez-Perpin˜a´ et al. (2005) ˚ 1T65 DHT, SRC2-2 peptide, 1.66 A Este´banez-Perpin˜a´ et al. (2005) ˚ 1XJ7 DHT, RAC3 peptide, 2.7 A Este´banez-Perpin˜a´ et al. (2005) ˚ 1Z95 R-bicalutamide, 1.8 A Bohl et al. (2005b) 2AX6 T877A mutant, Bohl et al. (2005a) ˚ Hydroxyflutamide, 1.5 A ˚ 2AX7 T877A mutant, S-1, 1.9 A Bohl et al. (2005a) ˚ 2AX8 W741L mutant, S-1, 1.7 A Bohl et al. (2005a) ˚ 2AX9 R-3 compound, 1.65 A Bohl et al. (2005a) ˚ 2AXA S-1 compound, 1.80 A Bohl et al. (2005a) ˚ 2AM9 Testosterone, 1.64 A Pereira de et al. (2006) ˚ 2AMA DHT, 1.90 A Pereira de et al. (2006) ˚ 2AMB Tetrahydrogestrinone, 1.75 A Pereira de et al. (2006) 2PIO DHT, 2-methylindole in AF2 Este´banez-Perpin˜a´ et al. (2007a) 2PIP DHT, 1H-INDOLE-3Este´banez-Perpin˜a´ et al. (2007a) CARBOXYLIC ACID (ICO) and 1-TERT-BUTYL-3-(2,5DIMETHYLBENZYL)-1HPYRAZOLO[3,4-D] PYRIMIDIN-4-AMINE (K10) in AF2, 1.80 A 2PIQ DHT, 3-[(4-AMINO-1-TERTEste´banez-Perpin˜a´ et al. (2007a) BUTYL-1H-PYRAZOLO [3,4-D]PYRIMIDIN-3-YL) METHYL]PHENOL in AF2, 2.40 A 2PIR DHT, Salicylaldehyde, 2.10 A Este´banez-Perpin˜a´ et al. (2007a) (continued)

The Androgen Receptor Coactivator-Binding Interface Table 1 (continued) Protein PDB code Comments 2PIT DHT, [4-(4-HYDROXY-3-IODOPHENOXY)-3,5-DIIODOPHENYL]-ACETIC ACID, 1.76 A 2PIU DHT, with two [4-(4-HYDROXY3-IODO-PHENOXY)-3,5DIIODO-PHENYL]-ACETIC ACID, 2.12 A 2PIV DHT, 3,5,3’TRIIODOTHYRONINE (T3 hormone), 1.95 A 2PIW DHT, 3,5,3’TRIIODOTHYRONINE (T3 hormone), 2.58 A 2PIX DHT, 2-[[3(TRIFLUOROMETHYL) PHENYL]AMINO] BENZOIC ACID (FLF, Flufenamic acid), 2.4 A 2PKL DHT, [4-(4-HYDROXY-3-IODOPHENOXY)-3,5-DIIODOPHENYL]-ACETIC ACID, 2.49 A 2QPY DHT, [4-(4-HYDROXY-3-IODOPHENOXY)-3,5-DIIODOPHENYL]-ACETIC ACID, 2.50 A

303

Reference Este´banez-Perpin˜a´ et al. (2007a)

Este´banez-Perpin˜a´ et al. (2007a)

Este´banez-Perpin˜a´ et al. (2007a)

Este´banez-Perpin˜a´ et al. (2007a)

Este´banez-Perpin˜a´ et al. (2007a)

Este´banez-Perpin˜a´ et al. (2007a)

Este´banez-Perpin˜a´ et al. (2007a)

determinants of the regulatory complex formation and specificity. A crystal structure of AR LBD has also been reported with a corepressor peptide (SHP) bound to its AF2 (Jouravel et al. 2007). The coactivator peptides studied by Xray crystallography are of two types: the first type comprised the actual domains or protein fragments from the family of physiological coactivators known to interact with the AR LBD (p160 and ARA family of coactivators) (Este´banez-Perpin˜a´ et al. 2005), and the second type is formed by libraries of peptides obtained by phage display that may mimic the natural binding domains (Hur et al. 2004). Natural regulatory peptides usually contain either the FxxLF motif (where F is phenylalanine, and X is any amino acid) present in the coactivator ARA70, as well as the LxxLL motifs (where L is leucine, and X is any amino acid) present in the p160 family member of coactivators. Peptides identified by phage display contain hydrophobic sequences with homologous motifs have also been reported: FxxLF, FxxFF, FxxYF, FxxLW, WxxLF, WxxVW, and LxxLL (Y and W are tyrosine and tryptophan, respectively) (Hur et al. 2004; Hsu et al. 2003). These hydrophobic amino acid sequence motifs are referred to as NR boxes (NR stands for Nuclear Receptor) (Fig. 4). AR LBD is an a-helical protein with 12 helices arranged in three layers (Matias et al. 2000; Sack et al. 2001). The AR AF2 surface is formed by residues of helices

304

E. Este´banez-Perpin˜a´, R.J. Fletterick

H3, H5, and H12 (Fig. 2). The AF2 hydrophobic pocket is a trifurcated groove formed by three moderately deep subpockets, each accommodating one of the three hydrophobic side chains of the NR motif (F1XXL4F5 and L1XXL4L5, respectively) (Figs. 2–4). The S1 subsite is the narrowest and deepest of the three, and it hosts the first hydrophobic residue: Leu1/Phe1 is embedded within this subsite (Fig. 4). S1 is lined by H5 residues M734, Q738, and H12 residues E897 and I898 (Figs. 3 and 4). The S2 subsite is the broadest and shallowest of the three and hosts the second hydrophobic residue: Leu4 is embedded within this subsite (Fig. 4). S2 is lined by H3 residues K720 and V713, and H12 residue M894 (Figs. 3 and 4). The S3 subsite is intermediate in size in comparison to the other two and quite deep. S3 is lined by H5 residues M734 and H3 residue K720 (Figs. 3 and 4). Both NR FXXLF and LXXLL motifs bind to the AR AF2 region as amphipathic a-helices with hydrophobic surfaces complementary to AF2 (Fig. 4). There are two clusters of oppositely charged residues on the AR LBD surface, at the opposite ends of the AF2 hydrophobic groove, that assist in orienting and registering the bound NR motifs in the AF2 groove. K720 and E897 are interesting because they function as ‘‘helix-capping’’ residues, forming the electrostatic ‘‘charge clamp’’ residues and stabilizing binding interactions of the FXXLF motif with the AF2 core via hydrogen bonding. At the start of the helix, Glu897 of AR forms hydrogen bonds to the backbone amide (NH) of Phe1 and preceding Lys. At the end of the helix, Lys720 of AR does not contact the carbonyl group of Leu4 or Phe5; instead, it forms hydrogen bonds to the carbonyl groups flanking residues at positions 6 and 7. The alternative NR-binding motif, LxxLL, interacts just like the FxxLF with Lys720, but its Leu1 residue does not contact Glu897 of AR. The structural basis of AR preference for peptides containing FxxLF motif over the canonical LxxLL sequence has been determined (Este´banez-Perpin˜a´ et al. 2005; Hur et al. 2004; He et al. 2004). Structural comparisons of several AR LBDs bound to different regulatory peptides revealed that side chains forming the AF2 pocket rearrange to bind multiple variants of the NR boxes. There are specific structural features in the AF2 site of AR that allow accommodation of the bulkier FxxLF motif. For example, residues M434 and K720 of the AR LBD assume different conformations to make room for FxxLF peptide, allowing its interaction with I737 of AR helix H4. In general, coactivator peptides cause the reformable AF2 pocket to open upon binding, resulting in a remarkable complementary fit between the AF2 surface and the bound FxxLF motifs. Interestingly, both LxxLL and FxxLF containing helical NR peptides interact with the same set of residues on the surface of the AR LBD; however, these residues are engaged in different binding interactions. Furthermore, ˚ the major axes of the helices docked into the AF2 groove shift by more than 1 A between the two classes of regulatory peptides. The described studies focused on the central region of the interface between the regulatory peptides and the AR LBD. However, the interactions may be more extensive in vivo as neighboring features of coactivators find footholds on the AR surface. The sequences beyond the conserved hydrophobic motif hot spot help to confer specificity to the receptor–coactivator coupling (Este´banez-Perpin˜a´ et al. 2005).

The Androgen Receptor Coactivator-Binding Interface

305

Table 2 Mutations at the binding pocket (LBP), and AF2 pocket associated with prostate cancer and androgen-insensitivity syndromes (AIS) Residues Mutations in prostate cancer Mutations in AIS AF2 pocket V713 Not described Not described V716 Not described Not described K717 Not described Not described K720 K720E Not described R726 R726L R726L V730 V730M Not described Q733 Not described Q733H M734 Not described Not described I737 Not described I737I Q738 Not described Not described M894 Not described Not described E897 Not described Not described I898 Not described I898T

Structural analysis of the corepressor peptide SHP bound to the AR AF2 region reveals that SHP peptide adopts a conformation similar to that observed for the bound LxxLL coactivator peptides (Jouravel et al. 2007). Several mutations in the AF2 pocket that affect Lys720, Ile737, Val730, Gln733 are implicated in prostate cancer and androgen-insensitivity syndromes (http://androgendb.mcgill.ca/) (Table 2). These mutations affect coactivator recruitment to AR. Specifically, mutations affecting residues Ile737 and Phe725 have been shown to disrupt the AR NTD/AF2 interaction without affecting hormone binding (Quigley et al. 2004). Many residues in the AF2 pocket are conserved among several members of the nuclear receptor family (Table 3). There are nuclear receptor coactivators that bind to several nuclear receptors while others do not exhibit promiscuity and are selective to either one or another. The variety coactivator–nuclear receptor partnerships increase the complexity of nuclear receptors transcriptional regulation (Lonard and O’Malley 2007).

3 The AF2 Pocket may be a Drugable Interaction Surface Because the AF2 pocket plays a pivotal role in AR function, it may be a target for therapeutical intervention. In particular, blocking associations of AR with coregulatory partners might lead to alternative treatments for prostate cancer. Compounds targeting the regulatory AR LBD surfaces have been proposed as alternative antiandrogens (Este´banez-Perpin˜a´ et al. 2005, 2007a; Este´banez-Perpin˜a´ and Fletterick 2007; Chang and McDonnell 2005). Drugs that would bind to the ARcoregulator assembly site should be less susceptible to mutation of the AR protein by cancer cells because two proteins form the contact, and any mutation of the AF2 surface to render the drug ineffective might also affect coregulator binding.

AF2 residues

L712 V713 V716 K720 F725 R726 V730 M734 Q738 M894 E897 I898

V571 I572 V575 K579 F584 R585 L589 M593 Q597 M752 E755 I756

L354 V355 I358 K362 F367 V368 L372 V376 E380 D538 L541 E542

L726 L727 V730 K734 F739 R740 I744 I748 Q752 M908 E911 V912

M777 I778 V781 K785 F790 K791 L795 I799 Q803 M959 E962 I963

Table 3 Structurally conserved amino acids in AF2 among several nuclear receptors NR AR GR ERa PR MR PDB code 1T5Z 1P93 1ERE 1A28 2AA7 I226 T227 V230 K234 F239 S240 C244 I248 K252 L400 E403 V404

TRa 2H79 T281 R282 D285 K288 L290 P291 P297 Q301 K306 L454 E457 V458

TRb 1BSX L276 F277 V280 K284 F289 S290 L294 V298 R302 F450 E453 M454

RXRa 1FBY

V293 Q294 T297 K301 F306 V307 L309 V315 K319 L468 Q470 I472

PPARg 1PRG

I234 Q235 I238 K242 F247 R248 S252 I256 K260 L413 E416 V417

VDR 1RJK

306 E. Este´banez-Perpin˜a´, R.J. Fletterick

The Androgen Receptor Coactivator-Binding Interface

307

Surface targeted antiandrogens directly blocking the AR coregulator assembly may be therapeutically useful alone or paired with the current LBP-antiandrogens. Small molecules would have to block AR associations by preferentially populating the surface normally filled by the amphipathic a-helices provided to AR by coregulator proteins (Este´banez-Perpin˜a´ et al. 2005; Este´banez-Perpin˜a´ and Fletterick 2007) (Fig. 4). The first example of small molecules that block coactivator recruitment both in vitro and in vivo has been reported by Arnold et al. for the thyroid receptor b (TRb) (Arnold et al. 2005, 2006, 2007; Este´banez-Perpin˜a´ et al. 2007b). These compounds (aromatic b-amino ketones (BAK)), identified in large-scale chemical screens, are representing a novel class of potent TRb antagonists. As the crystal structure revealed, these compounds bind irreversibly to one of the exposed cysteine residues located in TRb AF2 pocket (Este´banez-Perpin˜a´ et al. 2007b). In addition, peptide antagonists for the AR-coactivator recruitment have been identified using phage display (Chang et al. 2005). The existence of peptides that target the AR AF2 pocket together with the discovery of TRb AF2-binding compounds set a precedent that could also be applied to AR for discovery of novel antiandrogens targeting its surface. The same research strategy that led to the discovery of TRb specific blockers also proved to be successful for AR, identifying its AF2 site as a pharmaceutical target. X ray crystallography combined with fluorescent polarization techniques have identified compounds that bind weakly to the AR AF2 surface. These compounds belong to different chemotypes including thyroid hormones (TRIAC and T3), purine analogs, and small heterocyclic molecules such as 2-methylindole. Most of these compounds exhibit weaker association at AF-2 site due to their poor fit. Subsite S1 of the AF2 that hosts F1 or L1 of the signature motif F1XXL4F5/L1XXL4L5 seems to be the hot spot for compounds that bind at AF2. Research strategies aiming at improving binding to the AF2 subsites S2 or S3 might yield higher affinity compounds.

4 AF2 is Allosterically Regulated by an Adjacent Surface-Exposed Hydrophobic Pocket It has been observed that the AF2 pocket is reshaped not only by ligand binding to the LBP but also by small molecules binding elsewhere on the LBD surface. Recent structural and mutational studies have provided evidence that a novel hydrophobic surface-exposed pocket termed BF-3 (Binding Function 3) is able to remodel the AF2 surface and affect coactivator recruitment (Fig. 2). This surface pocket is adjacent to the AF2 site and appears to modulate it allosterically (Este´banezPerpin˜a´ et al. 2007a). The BF-3 surface comprises H1 (Pro723, Phe673, and Ile672), H3–5 (Gly724 and Asn727) and H9 (Phe826, Glu829, Glu837, Arg840, and Asn833) (Fig. 2). The crystal structures of ternary complexes of DHT-bound AR LBD with coactivator peptides (containing either LXXLL or FXXLF motif) revealed that TRIAC (a thyroid hormone) can bind to BF-3, causing coregulator peptides bound to AF-

308

E. Este´banez-Perpin˜a´, R.J. Fletterick

2 to become disordered. Atomic resolution structural analysis revealed that binding of TRIAC to BF-3 remodels the adjacent AF-2 pocket weakening coactivator binding (Este´banez-Perpin˜a´ et al. 2007a). Mutating residues that form BF-3 inhibits AR function and its AF-2 activity in vitro and in vivo. Mutations at BF-3 can lead to prostate cancer and androgen-insensitivity syndrome. Specifically, mutations of Gln670, Ile672, and Leu830 are associated with prostate cancer (Buchanan et al. 2001a, b; Shi et al. 2002), and Leu830, Pro723, Gly724, and Arg840 are mutated in androgen-insensitivity syndrome (http://www.androgendb.mcgill.ca/) (McPhaul 2002). Targeted mutagenesis of Asn727 and Arg840 in the BF-3 site abolished the AR LBD activity, similar to inhibition observed with mutations in the AF2 region (Este´banez-Perpin˜a´ et al. 2005). Curiously, the same mutations resulted in a changed conformation of the bound TRIAC as revealed by Xray crystallography. Similar to mutating Asn727 and Arg840, substitutions of Phe673, Pro723, Glu724, Glu737 reduce AR activity. In contrast, mutations in the vicinity of the BF-3 site (affecting residues Gln670, Ile672, Glu829, and Asn833) increase the AR AF2 activity up to fivefold (Estebanez- 222 Perpina´ et al. 2007a). The BF-3 site could be present in other nuclear receptors. Consistent with this speculation, part of the site, the H3–H4 loop, contains a signature sequence (Brelivet et al. 2004). Superposition of available crystal structures reveals conservation of BF-3 residues in the subfamily of steroid nuclear receptors. Mutations in equivalent regions of estrogen and glucocorticoid receptor are shown to affect coactivator binding (Tanenbaum et al. 1998; Milhon et al. 1997). Acknowledgments We thank Elena Sablin, Debra Singer, and Leslie Cruz for their useful comments on the manuscript.

References Agoulnik I, Vaid A, Bingman WEIII, Erdeme H, Frolov A, Smith CL, Ayala G, Ittmann MM, Weigel NL: Role of SRC-1 in the promotion of prostate cancer cell growth and tumor progression. Cancer Res. 2005, 65:7976–7983. Aranda A, Pascual A: Nuclear hormone receptors and gene expression. Physiol Rev 2001, 81:1269–1304. Arnold LA, Estebanez-Perpina E, Togashi M, Jouravel N, Shelat A, McReynolds AC, Mar E, Nguyen P, Baxter JD, Fletterick RJ, Webb P, Guy RK: Discovery of small molecule inhibitors of the interaction of thyroid hormone receptor with transcriptional coregulators. J Biol Chem 2005, 280(52):43048–43055. Arnold LA, Estebanez-Perpina E, Togashi M, Shelat A, Ocasio CA, McReynolds AC, Nguyen P, Baxter JD, Fletterick RJ, Webb P, Guy RK: A high-throughput screening method to identify small molecule inhibitors of thyroid hormone receptor coactivator binding. Sci STKE 2006, 341:13. Arnold L, Kosinski A, Este´banez-Perpi–a E, Robert J, Fletterick Guy RK: Inhibitors of the interaction of a thyroid hormone receptor and coactivators: preliminary structure-activity relationships. J Med Chem 2007, 50:5269–5280.

The Androgen Receptor Coactivator-Binding Interface

309

Askew E, Gampe RT Jr, Stanley TB, Faggart JL, Wilson EM: Modulation of androgen receptor activation function 2 by testosterone and dihydrotestosterone. J Biol Chem 2007, 282:25801– 25816. Bevan CL, Hoare S, Claessens F, Heery DM, Parker MG: The AF1 and AF2 domains of the androgen receptor interact with distinct regions of SRC1. Mol Cell Biol 1999, 19:8383–8392. Bohl CE, Miller DD, Chen J, Bell CE, Dalton JT: Structural basis for accommodation of nonsteroidal ligands in the androgen receptor. J Biol Chem 2005a, 280:37747–37754. Bohl CE, Gao W, Miller DD, Bell CE, Dalton JT: Structural basis for antagonism and resistance of bicalutamide in prostate cancer. Proc Natl Acad Sci USA 2005b, 102:6201–6206. Bourguet W, Germain P, Gronemeyer H: Nuclear receptor ligand-binding domains: three-dimensional structures, molecular interactions and pharmacological implications. Trends Pharmacol Sci 2000, 21:381–388. Brelivet Y, Kammerer S, Rochel N, Poch O, Moras D: Signature of the oligomeric behaviour of nuclear receptors at the sequence and structural level. EMBO Rep 2004, 5:423–429. Buchanan G, Yang M, Harris JM, Nahm HS, Han G, Moore N, Bentel JM, Matusik RJ, Horsfall DJ, Marshall VR, et al: Mutations at the boundary of the hinge and ligand binding domain of the androgen receptor confer increased transactivation function. Mol Endocrinol 2001a, 15:46–56. Buchanan G, Greenberg NM, Scher HI, Harris JM, Marshall VR, Tilley WD: Collocation of androgen receptor gene mutations in prostate cancer. Clin Cancer Res 2001b, 7:1273–1281. Burd CJ, Morey LM, Knudsen KE: Androgen receptor corepressors and prostate cancer. Endocr Relat Cancer 2006, 13:979–994. Chang C, McDonnell DP: Androgen-receptor-cofactor interactions as targets for new drug discovery. Trends Pharmacol Sci 2005, 26:225–228. Chang CY, Abdo J, Hartney T, McDonnell DP: Development of peptide antagonists for the androgen receptor using combinatorial peptide phage display. Mol Endocrinol 2005, 19:2478–2490. Chmelar R, Buchanan G, Need EF, Tilley W, Greenberg NM: Androgen receptor coregulators and their involvement in the development and progression of prostate cancer. Int J Cancer 2007, 120:719–733. Choudhry M, Ball A, McEwan IJ: The role of the general transcription factor IIF in androgen receptor-dependent transcription. Mol Endocrinol 2006, 20:2052–2061. Darimont BD: Finding specificity within a conserved interaction site. Chem Biol 2003, 10:675– 676. Este´banez-Perpin˜a´ EJN, Fletterick RJ: Perspectives on designs of antiandrogens for prostate cancer. Expert Opin Drug Discov 2007, 2:1341. Este´banez-Perpin˜a´ E, Moore JM, Mar E, Delgado-Rodrigues E, Nguyen P, Baxter JD, Buehrer BM, Webb P, Fletterick RJ, Guy RK: The molecular mechanisms of coactivator utilization in ligand-dependent transactivation by the androgen receptor. J Biol Chem 2005, 280:8060–8068. Este´banez-Perpin˜a´ E, Arnold AA, Nguyen P, Rodrigues ED, Mar E, Bateman R, Pallai P, Shokat KM, Baxter JD, Guy RK, Webb P, Fletterick RJ: A surface on the androgen receptor that allosterically regulates coactivator binding. Proc Natl Acad Sci USA 2007a, 104:16074–16079. Este´banez-Perpin˜a´ E, Arnold LA, Jouravel N, Togashi M, Blethrow J, Mar E, Nguyen P, Phillips KJ, Baxter JD, Webb P, Guy RK, Fletterick RJ: Structural insight into the mode of action of a direct inhibitor of coregulator binding to the thyroid hormone receptor. Mol Endocrinol. 2007b, 21:2919–2928. Feng W, Ribeiro RC, Wagner RL, Nguyen H, Apriletti JW, Fletterick RJ, Baxter JD, Kushner PJ, West BL: Hormone-dependent coactivator binding to a hydrophobic cleft on nuclear receptors. Science 1998, 280:1747–1749. He B, Wilson EM: The NH(2)-terminal and carboxyl-terminal interaction in the human androgen receptor. Mol Genet Metab 2002, 75:293–298.

310

E. Este´banez-Perpin˜a´, R.J. Fletterick

He B, Bowen NT, Minges JT, Wilson EM: Androgen-induced NH2- and COOH-terminal interaction inhibits p160 coactivator recruitment by activation function 2. J Biol Chem 2001, 276:42293–42301. He B, Lee LW, Minges JT, Wilson EM: Dependence of selective gene activation on the androgen receptor NH2- and COOH-terminal interaction. J Biol Chem 2002, 277:25631–25639. He B, Gampe RT Jr, Kole AJ, Hnat AT, Stanley TB, An G, Stewart EL, Kalman RI, Minges JT, Wilson EM: Structural basis for androgen receptor interdomain and coactivator interactions suggests a transition in nuclear receptor activation function dominance. Mol Cell 2004, 16:425–438. He B, Gampe RT Jr, Hnat AT, Faggart JL, Minges JT, French FS, Wilson EM: Probing the functional link between androgen receptor coactivator and ligand-binding sites in prostate cancer and androgen insensitivity. J Biol Chem 2006, 281:6648–6663. Hodgson M, Astapova I, Cheng S, Lee LJ, Verhoeven MC, Choi E, Balk SP, Hollenberg AN: The androgen receptor recruits nuclear receptor CoRepressor (N-CoR) in the presence of mifepristone via its N and C termini revealing a novel molecular mechanism for androgen receptor antagonists. J Biol Chem 2005, 280:6511–6519. Hsu CL, Yeh S, Chen YL, Ting HJ, Hu YC, Lin H, Wang X, Chang C: The use of phage display technique for the isolation of androgen receptor interacting peptides with F/WXXLF/W and FXXLY new signature motifs. J Biol Chem 2003, 278:23691–23698. Hsu CL, Chen YL, Ting HJ, Lin WJ, Yang Z, Zhang Y, Wang L, Wu CT, Chang HC, Yeh S, Pimplikar SW, Chang C: Androgen receptor (AR) NH2- and COOH-terminal interactions result in the differential influences on the AR-mediated transactivation and cell growth. Mol Endocrinol 2005, 19:350–361. Hur E, Pfaff SJ, Sturgis PE, Hanne G, Buehrer BM, Fletterick RJ: Recognition and accommodation at the androgen receptor coactivator binding interface. PLoS 2004, 2:363. Jouravel N, Sablin E, Arnold LA, Guy RK, Fletterick RJ: Interaction between the androgen receptor and a segment of its corepressor SHP. Acta Crystallogr D Biol Crystallogr 2007, 63:1198–1200. Lazar M: Nuclear receptor corepressors. Nucl Recept Signal. 2003, 1:e001. Li J, Fu J, Toumazou C, Yoon HG, Wong J: A role of the amino-terminal (N) and carboxylterminal (C) interaction in binding of androgen receptor to chromatin. Mol Endocrinol 2006, 20:776–785. Lonard DM, O’Malley BW: The expanding cosmos of nuclear receptor coactivators. Cell 2006, 125:411–414. Lonard DM, O’Malley BW: Nuclear receptor coregulators: judges, juries, and executioners of cellular regulation. Mol Cell 2007, 27:691–700. Matias PM, Donner P, Coelho R, Thomaz M, Peixoto C, Macedo S, Otto N, Joschko S, Scholz P, Wegg A, et al: Structural evidence for ligand specificity in the binding domain of the human androgen receptor. Implications for pathogenic gene mutations. J Biol Chem 2000, 275:26164– 26171. Matias PM, Carrondo MA, Coelho R, Thomaz M, Zhao XY, Wegg A, Crusius K, Egner U, Donner P: Structural basis for the glucocorticoid response in a mutant human androgen receptor (AR (ccr)) derived from an androgen-independent prostate cancer. J Med Chem 2002, 45:1439– 1446. McKenna NJ, O’Malley BW: Minireview: nuclear receptor coactivators – an update. Endocrinology 2002, 143:2461–2465. McKenna N, Xu J, Nawaz Z, Tsai SY, Tsai MJ, O’Malley BW: Nuclear receptor coactivators: multiple enzymes, multiple complexes, multiple functions. J Steroid Biochem Mol Biol 1999, 69:3–12. McPhaul MJ: Androgen receptor mutations and androgen insensitivity. Mol Cell Endocrinol 2002, 198:61–67.

The Androgen Receptor Coactivator-Binding Interface

311

Milhon J, Lee S, Kohli K, Chen D, Hong H, Stallcup MR: Identification of amino acids in the tau 2-region of the mouse glucocorticoid receptor that contribute to hormone binding and transcriptional activation. Mol Endocrinol 1997, 11:1795–1805. Moras D, Gronemeyer H: The nuclear receptor ligand-binding domain: structure and function. Curr Opin Cell Biol 1998, 10:384–391. Nichols M, Rientjes JM, Stewart AF: Different positioning of the ligand-binding domain helix 12 and the F domain of the estrogen receptor accounts for functional differences between agonists and antagonists. EMBO J 1998, 17:765–773. Pereira de J, Jesus-Tran K, Cote PL, Cantin L, Blanchet J, Labrie F, Breton R: Comparison of crystal structures of human androgen receptor ligand-binding domain complexed with various agonists reveals molecular determinants responsible for binding affinity. Protein Sci 2006, 15:987–999. Prescott J, Coetzee GA: Molecular chaperones throughout the life cycle of the androgen receptor. Cancer Lett 2006, 231:12–19. Quigley C, Tan JA, He B, Zhou ZX, Mebarki F, Morel Y, Forest MG, Chatelain P, Ritze´n EM, French FS, Wilson EM: Partial androgen insensitivity with phenotypic variation caused by androgen receptor mutations that disrupt activation function 2 and the NH(2)- and carboxylterminal interaction. Mech Ageing Dev 2004, 125:683–689. Sack JS, Kish KF, Wang C, Attar RM, Kiefer SE, An Y, Wu GY, Scheffler JE, Salvati ME, Krystek SR, Jr., et al: Crystallographic structures of the ligand-binding domains of the androgen receptor and its T877A mutant complexed with the natural agonist dihydrotestosterone. Proc Natl Acad Sci USA 2001, 98:4904–4909. Salvati M, Balog A, Shan W, Wei DD, Pickering D, Attar RM, Geng J, Rizzo CA, Gottardis MM, Weinmann R, Krystek SR, Sack J, An Y, Kish K: Structure based approach to the design of bicyclic-1H-isoindole-1,3(2H)-dione based androgen receptor antagonists. Bioorg Med Chem Lett 2005, 15:271–276. Shaffer PL, Jivan A, Dollins DE, Claessens F, Gewirth DT: Structural basis of androgen receptor binding to selective androgen response elements. Proc Natl Acad Sci USA 2004, 101:4758– 4763. Shang Y, Myers M, Brown M: Formation of the androgen receptor transcription complex. Mol Cell 2002, 9:601–610. Shi XB, Ma AH, Xia L, Kung HJ, de Vere White RW: Functional analysis of 44 mutant androgen receptors from human prostate cancer. Cancer Res 2002, 62:1496–1502. Tanenbaum DM, Wang Y, Williams SP, Sigler PB: Crystallographic comparison of the estrogen and progesterone receptor’s ligand binding domains. Proc Natl Acad Sci USA 1998, 95:5998– 6003. Wang L, Hsu CL, Chang C: Androgen receptor corepressors: an overview. Prostate 2005, 63:117– 130. Wurtz JM, Bourguet W, Renaud JP, Vivat V, Chambon P, Moras D, Gronemeyer H: A canonical structure for the ligand-binding domain of nuclear receptors. Nat Struct Biol 1996, 3:206. Yeh S, Chang C: Cloning and characterization of a specific coactivator, ARA70, for the androgen receptor in human prostate cells. Proc Natl Acad Sci USA 1996, 93:5517–5521. Zhou ZX, He B, Hall SH, Wilson EM, French FS: Domain interactions between coregulator ARA (70) and the androgen receptor (AR). Mol Endocrinol 2002, 16:287–300.

Coregulators and the Regulation of Androgen Receptor Action in Prostate Cancer Irina U. Agoulnik and Nancy L. Weigel

Abstract The importance of androgen receptors in primary prostate cancer is well established and some form of androgen blockade is the primary treatment for metastatic prostate cancer. There is increasing evidence that the androgen receptor (AR) continues to play a role in castration resistant disease despite the decrease in serum androgens. Thus, factors that modulate AR activity are potential therapeutic targets. AR is a transcription factor that regulates its target genes by recruiting a complex of coregulators with multiple enzymatic activities. These coregulators remodel chromatin, modify receptor, other coregulators, and general transcription factors, as well as affect splicing decisions. Here we summarize current evidence for changes in expression of coregulators in prostate cancer and their function in prostate cancer cell lines. Many of these coregulators have pleiotropic functions and modulate transcription factors other then AR. Thus, they may have both AR dependent and AR independent roles in prostate cancer.

1 Introduction Androgens, testosterone and dihydrotestosterone (DHT), acting through the androgen receptor (AR) are required for the development and maintenance of the prostate. Moreover, prostate cancer is an androgen-dependent disease (Arnold and Isaacs 2002). Prostate cancers that have escaped the prostate capsule are treated with androgen-deprivation therapy that is effective initially. However, most of the prostate cancers recur and become resistant to androgen-deprivation therapy. Despite this, there is strong evidence that castration-recurrent tumors continue to depend on the AR signaling pathway. AR is a ligand-activated transcription factor that assembles a series of coactivators and corepressors on the promoter of its target

N.L. Weigel(*) Baylor College of Medicine, Houston, TX, USA. E-mail: [email protected]

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4_14, # Springer Science + Business Media, LLC 2009

315

316

I.U. Agoulnik, N.L. Weigel

genes to regulate transcription of genes, which include prostate-specific antigen (PSA). AR activity is regulated by coactivators, corepressors, cell signaling, and the concentration of available AR ligands. Evidence that AR is reactivated and plays a role in prostate cancer growth and survival in androgen-depleted conditions is derived from both in vitro and in vivo studies. Although C4-2 prostate cancer cells grow in castrated nude mice and in androgen-depleted medium in vitro, reducing AR expression using AR-specific siRNA strongly inhibits cellular proliferation and basal PSA expression (Agoulnik et al. 2005). Although these cells do not require exogenous androgens, they remain AR sensitive for cell growth and expression of PSA. In vivo, castration-recurrent disease is detected typically by an increase in serum PSA levels, which suggests that AR is reactivated under these conditions. The level of AR in primary tumors is associated with markers of more aggressive disease (Li et al. 2004). Moreover, AR expression typically increases during progression to recurrence after castration (Chen et al. 2004a). These findings led to increased interest in identifying the coregulators and cell signaling pathways that regulate androgen-dependent and androgen-independent AR action in prostate as these factors are potential therapeutic targets to more completely block AR action.

1.1

Androgen Receptor

After translation, AR is folded and maintained in a form capable of binding hormone by a series of chaperone proteins that include hsp90. AR dissociates from this complex upon ligand binding, forms homodimers, and translocates into the nucleus. AR can regulate transcription in multiple ways (Fig. 1) (Agoulnik and Weigel 2006). In the best understood mechanism of action, upon agonist action, AR dimerizes, binds specific DNA sequences named AR response elements (ARE), and recruits coactivator complexes that modify chromatin to facilitate recruitment of general transcription factors and Pol II to induce transcription. A well-known gene regulated in this manner is PSA produced by both normal epithelium and prostate cancer cells; serum levels of PSA are used as a marker to evaluate tumor burden in prostate cancer patients. Another important AR target gene regulated in this manner is TMPRSS2, a gene whose promoter is frequently fused to ETS factor oncogenes in prostate cancer (Tomlins et al. 2005). AR also represses transcription, as in the case of maspin, a tumor suppressor from the serpin family (He et al. 2005). Although this is also accomplished through direct DNA binding, the associated factors that cause repression rather than induction have not been identified. AR also binds directly to other transcription factors and regulates their transcriptional activity presumably functioning as a coactivator for these factors (Lu et al. 2000). Finally, recent reports indicate that hormone-bound AR interacts with modulator of nongenomic action of estrogen receptor (MNAR) and src and thus activates cell signaling pathways such as MAPK-p42/p44 and PI3K

Coregulators and the Regulation of Androgen Receptor Action in Prostate Cancer

317

Fig. 1 Modes of androgen receptor action AR is folded in the cytoplasm with the aid of heat-shock proteins (HSPs) and is maintained in a complex with HSPs in the absence of ligand. AR, upon binding androgen (triangles), sheds HSPs and modulates transcription in at least four different ways. (1) AR dimerizes in an antiparallel orientation, binds to AREs, recruits coactivator complexes that remodel chromatin, and facilitates binding of the general transcription machinery (GTF) and RNA Polymerase II (Pol) to initiate transcription. (2) Agonist-bound AR is recruited to repressive AREs, assembles coregulator complexes of as yet unknown content, and actively represses transcription. (3) Liganded AR binds heterologous transcription factors (TF) and presumably recruits coregulators to enhance transcriptional activity. (4) AR binds to src/MNAR complexes that activate kinase cascades, which results in altered transcriptional activity of transcription factors (TF) that do not necessarily physically interact with AR. AR, coregulators, GTFs, and polymerase II are all subjects to regulation by phosphorylation (phosphogroups are depicted as four pointed stars) that is initiated, in part, by receptor tyrosine kinases (YK) and/or G-protein-coupled receptors (GPCR)

(Castoria et al. 2003, 2004; Unni et al. 2004). In addition to the hormone-dependent pathways, there is increasing evidence that activation of some cell signaling pathways is sufficient to induce AR activity in the absence of hormone (Culig et al. 1994; Ueda et al. 2002). The exact mechanism of ligand-independent activation is not known but, as discussed later, it utilizes some of the same coactivators as liganddependent activation complexes. AR and many of its coactivators are phosphoproteins and their activity is regulated by phosphorylation.

318

1.2

I.U. Agoulnik, N.L. Weigel

Coregulators

Regulation of transcription requires a series of steps mediated in part by the enzymatic activities of AR-interacting proteins. These steps include chromatin remodeling, initiation of transcription, elongation, and termination. During the activation process, AR is post-translationally modified to modulate activity and subsequently to target it for degradation through a proteasome-mediated pathway. Inhibition of the proteasome pathway hinders AR-dependent transcription (Kang et al. 2002; Lin et al. 2002). Many modifications of the active chromatin proteins are reversed and in some cases DNA is methylated during transcriptional repression. Although many coregulators have intrinsic enzymatic activity, others serve as scaffolds to promote interactions between multiple proteins. The first steroid receptor coactivator, SRC-1, was described in 1995 (Onate et al. 1995). The number of candidate coregulators capable of regulating the activity of steroid receptors has surpassed 300 (http://www.nursa.org/). Clearly, a single receptor molecule cannot interact with so many proteins simultaneously. A partial explanation for the large number of proteins is that coactivators typically function as large multiprotein complexes, conduits of multiple signaling pathways in the cell. Multiple complexes of distinct compositions must be recruited sequentially to chromatin in order to regulate transcription. While some factors may be required for transcription of all target genes, others may be receptor, promoter, or cell type specific. From the perspective of prostate cancer, the challenge is to identify the coregulators that are important for target gene-specific regulation of androgendependent and androgen-independent AR activity.

2 Coregulators in Prostate Cancer Because of the importance of AR in prostate cancer and its significance as a therapeutic target, the contributions of various coactivators to AR activity and their relative expression levels in prostate cancer have been active areas of research. Interpretation of some of these results is complex. Most, if not all, of the AR coactivators identified to date also regulate other nuclear receptors and many of these receptors are expressed in the benign and malignant prostate. Moreover, many of the proteins participate in the regulation of a wide variety of transcription factors. Thus, while changes in AR-mediated gene transcription in response to overexpression or reduction in coregulator expression are likely due to direct effects on AR, caution is needed in interpreting changes in cell proliferation or the consequences of altered expression in prostate cancer as AR-dependent activities. In addition, the overall activity of coregulators depends not only on the level of expression but also on a number of post-translational modifications, which include acetylation, sumoylation, phosphorylation, and ubiquitination. The exact effect of each modification on AR activity and levels of modifications in tissue is impossible to assess at present. The expression of coactivators in benign prostate and prostate cancer has

Coregulators and the Regulation of Androgen Receptor Action in Prostate Cancer

319

been evaluated on both the transcriptional and the protein levels. In most cases, comparisons of the mRNA levels in samples of benign and cancer tissue do not consider the changed ratio between cells of stromal and epithelial origin; the stromal compartment is a major contributor in benign tissue, while cells of epithelial origin predominate in prostate cancer. Thus, it is difficult to determine accurately whether observed changes are prostate cancer specific. Because immunohistochemical studies can compare expression of the coregulators in specific cellular populations, they are more accurate in determining whether expression of a particular coregulator is changed in malignant compared to benign epithelial cells. Therefore, this chapter summarizes the available immunohistochemical data on coregulator expression in prostate cancer and benign prostate with emphasis on coregulators with enzymatic activity. In addition, current data are reported that describe a role for coregulators in AR action in prostate cancer cell lines.

2.1

Histone Acetyltransferases

The first nuclear receptor coactivator identified, SRC-1, is a histone acetyltransferase (HAT) (Spencer et al. 1997). Many of the subsequently identified AR coactivators are also HATs or function as scaffold proteins that recruit other HATs. Histones are major substrates for these enzymes. A change in global levels of histone acetylation is associated with cancer and is predictive of clinical outcome. The global histone modification pattern is an independent predictor of recurrence in patients with low-grade prostate cancer (Seligson et al. 2005). Although histones are important substrates, some of the HATs acetylate other coactivators and even AR itself.

2.1.1

The p160 Coactivators

This small family consists of three 160-kDa proteins that potentiate activities of nuclear receptors and some other transcription factors, although the range of transcription factors activated by these proteins is coactivator specific. The family includes SRC-1, SRC-2 (GRIP1, TIF2), and SRC-3 (AIB1, p/CIP, ACTR, RAC-3, TRAM-1) (Xu and O’Malley 2002). SRC-1 and SRC-3 coactivators are HATs, and all three recruit additional HATs to AR. The p160 coactivators have multiple LXXLL motifs, which interact with the ligand binding domains of AR and serve as platforms for recruitment of secondary coactivators such as methyltransferases and acetyltransferases. However, the most important p160 interaction site in AR is in the amino-terminal AF1 region, which interacts with a glutamine-rich region in the p160 coactivators (Bevan et al. 1999).

320

I.U. Agoulnik, N.L. Weigel

SRC-1 The first demonstration that SRC-1 plays a physiologically significant role in prostate came from analysis of SRC-1 null mice (Xu et al. 1998). In castrated mice, androgen-induced prostate growth was less than in wild-type littermates. In SRC-1 null mice the level of TIF2 was increased twofold, possibly as a compensatory response. If SRC-1 and TIF2 have overlapping functions, as might be predicted from molecular biology studies, the role for SRC-1 in prostate growth may be underestimated. SRC-1 is expressed in benign prostate epithelium and in all the commonly studied human prostate cancer cell lines. SRC-1 expression was reduced using siRNA to determine whether it is required for androgen-dependent gene induction. SRC-1 proved necessary for maximal expression of the androgen-induced target gene, PSA, in LNCaP cells (Agoulnik et al. 2005). Coactivators are thought generally to play roles in activation, rather than repression, of transcription. However, depletion of SRC-1 resulted in derepression of the androgen-repressed tumor suppressor, maspin. SRC-1 expression was reduced in C4-2 cells grown in androgen-depleted medium and PSA expression measured to evaluate the role of SRC-1 in androgen-independent activities of AR; PSA expression was reduced, which indicated that SRC-1 also plays a role in androgen-independent AR activity. Although PSA expression is used as a measure of AR action, AR-dependent growth in LNCaP cells was induced by a selective AR modulator (SARM) without induction of PSA (Sathya et al. 2003). Thus, induction of PSA may be an imperfect surrogate for AR activity. However, depletion of SRC-1 from androgen-sensitive LNCaP cells or androgen-independent C4-2 cells reduced proliferation measured using [3H]-thymidine incorporation (Agoulnik et al. 2005). In contrast, depletion of SRC-1 had no effect on proliferation of the AR-negative prostate cancer cell lines PC-3 and DU145, which suggests that the effect of SRC-1 on prostate cell proliferation is AR dependent. Several reports describe SRC-1 expression in prostate cancer. In a small study, Gregory et al. (2001) found that SRC-1 expression was increased in castrationrecurrent prostate cancer compared to androgen-stimulated clinically localized prostate cancer and benign prostatic hyperplasia. When SRC-1 expression was examined in a tissue microarray containing androgen-stimulated prostate cancer and corresponding benign samples from more than 500 patients, SRC-1 expression was highly variable (Agoulnik et al. 2005). The average SRC-1 expression levels in benign prostate versus prostate cancer did not differ, but higher expression correlated with several markers characteristic of more aggressive prostate cancer. In contrast, expression in a set of metastatic tumors was higher than in the primary tumors. An unexpected finding was a highly significant correlation between SRC-1 expression in benign prostate and prostate cancer in the same patient. This implies that higher levels of SRC-1 expression in benign prostate may contribute to the development of more aggressive prostate cancer. Although the correlations between SRC-1 expression and prostate cancer are not striking, SRC-1 activity is regulated by phosphorylation. Rowan et al. (2000)

Coregulators and the Regulation of Androgen Receptor Action in Prostate Cancer

321

identified multiple phosphorylation sites in SRC-1. At least two of these sites are phosphorylated by p42/p44 MAPK, and these sites play a role in the SRC-1dependent stimulation of AR activity induced by IL-6 (Ueda et al. 2002). IL-6 treatment induces p42/p44 MAPK activity and this activity is required for the induction of AR activity. Moreover, p42/p44 MAPK activity is increased in advanced prostate cancer (Gioeli et al. 1999); advanced prostate cancer may have elevated SRC-1 activity without significant changes in expression.

SRC-2 Accumulating evidence suggests that SRC-2 plays a role in prostate cancer and its relapse during androgen-deprivation therapy. Some of the concerns in inhibiting AR action in prostate cancer include factors that increase AR action at low androgen levels or facilitate activation by other ligands. SRC-2 is capable of both these activities. Overexpression of SRC-2 not only potentiates AR action, but the fold induction is higher at suboptimal levels of androgen (Agoulnik et al. 2006); depletion of SRC-2 using siRNA reduced androgen-dependent induction of several primary AR target genes, including PSA and TMPRSS2 (Agoulnik et al. 2006). Unlike SRC-1, SRC-2 does not play a role in androgen-dependent repression of maspin. Depletion of SRC-2 using siRNA reduced proliferation of not only the androgen-sensitive LNCaP cell line but also AR-negative PC-3 cells. Thus, SRC-2 modulates AR signaling, but it also has an AR-independent role in cellular proliferation. SRC-2 expression was repressed by androgens, which was a direct action of AR (Agoulnik et al. 2006). Analysis of the DNA regions to which AR is recruited in response to androgen treatment identified AR binding sites both in an SRC-2 intron and in the immediate upstream promoter. A small fragment of 50 flanking sequence linked to the coding sequence of luciferase was sufficient to recapitulate the repression caused by androgens. Repression required AR with an intact DNA and ligand binding domain and the presence of ligand (either agonist or antagonist). SRC-2 is also a target of the p42/ p44 MAPK pathway. In androgen-independent CWR-R1 cells, EGF treatment increased SRC-2 expression, phosphorylation, and interaction with AR, and increased transcriptional activity of AR (Gregory et al. 2004). SRC-2 expression also has been studied in prostate cancer. Gregory et al. examined 24 primary prostate specimens that included eight specimens of androgenstimulated benign prostatic hyperplasia, eight specimens of androgen-stimulated clinically localized prostate cancer, and eight specimens of castration-recurrent prostate cancer for SRC-2 expression. SRC-2 was elevated in castration-recurrent prostate cancer compared to either androgen-stimulated benign or malignant prostate (Gregory et al. 2001). In a large-scale tissue microarray study of primary prostate tumors and adjacent normal prostate from over 500 patients, Agoulnik et al. (2006) found that expression of SRC-2 increased with progression of prostate cancer. Increased expression correlated with a shorter time to biochemical recurrence (rising serum PSA levels). SRC-2 levels had no prognostic significance in patients with low levels of AR. Patients with higher levels of AR recurred sooner if they also expressed high levels of

322

I.U. Agoulnik, N.L. Weigel

SRC-2 compared to those with lower levels of SRC-2 expression. Thus, in primary tumors, SRC-2 likely facilitates AR action despite the fact that it also has an AR-independent effect on cellular proliferation. Analysis of samples taken from patients during failed androgen-deprivation therapy revealed maximal expression in all nine samples, which supports the idea that androgen-deprivation therapy leads to increased expression of SRC-2; increases in SRC-2 expression may activate pathways that lead to stimulation of AR-dependent and AR-independent proliferation of prostate epithelium.

SRC-3 As inferred from one of its names, AIB1 (amplified in breast cancer 1)/SRC-3 is overexpressed in breast cancer (Anzick et al. 1997). Genetic disruption of SRC-3 in mice results in pleiotropic phenotypes that include dwarfism, delayed sexual maturation of females, and smaller prostate size in males (Xu et al. 2000; Yan et al. 2006). SRC-3 regulates expression of multiple IGF and Akt signaling pathway proteins, at least in part, through coactivation of AP-1 transcription factors (Yan et al. 2006; Zhou et al. 2003, 2005). SRC-3 also is an AR coactivator and is recruited to the PSA promoter and enhancer (Wang et al. 2005). Depletion of SRC-3 using siRNA reduced proliferation of prostate cancer cells, including ARnegative prostate cancer cells, which confirms that SRC-3 has diverse functions in prostate cancer cells (Zhou et al. 2003). Moreover, prostate cancer cells artificially overexpressing SRC-3 increased in size and grew faster in nude mouse xenografts. Similar to other members of the p160 family, SRC-3 is a phosphoprotein. Six phosphorylation sites were identified that are important for AR coactivation and at least one of them is an ERK1/2 target (Wu et al. 2004). Immunohistochemical analysis of tissue microarrays containing tissues from more than 500 patients showed nuclear SRC-3 staining in basal and luminal epithelium, as well as in malignant cells. SRC-3 expression was higher in primary prostate cancer than benign prostate; higher SRC-3 expression in prostate cancer correlated with increased proliferation and reduced apoptosis, which are characteristics of aggressive prostate cancer, and shorter time to biochemical recurrence (Zhou et al. 2005). Consistent with cell-based studies, SRC-3 expression in clinical prostate cancer positively correlated with phospho-Akt staining. In a mouse prostate reconstitution model, artificial overexpression of AR and Akt, but not of either alone, was sufficient to induce prostate cancer (Xin et al. 2006). Thus, SRC-3 may potentiate AR activity not only through its role as a classical AR coactivator, but also by increasing Akt activity.

2.1.2

Acetyltransferases

The acetyltransferases discussed are sometimes called cointegrators or secondary coactivators. They are broad in their action, can interact with numerous transcription

Coregulators and the Regulation of Androgen Receptor Action in Prostate Cancer

323

factors and coregulators, and acetylate both histones and other nuclear proteins, which include AR and AR coregulators. Acetylation of AR on lysines in the hinge region selectively augments AR transcriptional activity on some target genes, enhances p300 binding, and reduces NCoR/HDAC/Smad3 corepressor binding (Fu et al. 2003). Transfection of AR mutants that mimic hyperacetylation in its hinge region into prostate cancer cells increases prostate cancer cell survival and proliferation in soft agar and nude mouse xenografts (Fu et al. 2003).

p300 The p300 coactivator acetylates histones, receptors, and coactivators (Fu et al. 2004; Kalkhoven 2004). p300, but not its homolog CBP (CREB binding protein), may have a role in nonsteroidal as well as steroid-mediated AR activation. Activation of AR by IL-6 requires p300 (Debes et al. 2002). Depletion of p300 from cells reduces AR activity and the level of AR acetylation. Treatment with bombesin, a neuropeptide, also activates AR in the absence of ligand by increasing p300 HAT activity and the AR acetylation level (Desai et al. 2006; Gong et al. 2006). Analysis of p300 expression in prostate cancer shows that expression of p300 correlates with increased proliferation, pathological characteristics of more aggressive disease, and decreased time to biochemical recurrence after treatment (Debes et al. 2002). The expression of p300 is subject to androgen regulation; androgen treatments reduces p300 protein levels without affecting its mRNA levels (Heemers et al. 2007). P/CAF (p300/CBP-Associated Factor) P/CAF is an acetyltransferase that interacts with AR (Fu et al. 2004), and AR coactivators such as CBP and the p160 coactivators, and thus can be recruited to AR through these coactivators. AR is acetylated by P/CAF potentiating the activity of AR through direct post-translational modification as well as through its actions on histones and other proteins/(Fu et al. 2004). Tip60 (Tat-Interacting Protein, 60 kDa) Tip60 is a HAT that has multiple substrates which include histones, AR, and p53 (Fu et al. 2004; Sykes et al. 2006). Acetylation of the AR’s hinge region is required for Tip60 coactivation of AR. While acetylation of AR and histones increases AR activity, cell proliferation, and survival (Fu et al. 2004; Popov et al. 2007), acetylation of p53 induced apoptosis in response to DNA damage (Sykes et al. 2006). Analysis of Tip60 expression in prostate cancer revealed a shift in its subcellular localization, from diffuse distribution in androgen-stimulated benign prostate and primary prostate cancer to nuclear accumulation in castrationrecurrent prostate cancer (Halkidou et al. 2003). Studies in prostate cancer cell lines showed recruitment of Tip60 to the PSA promoter in both androgen-sensitive and

324

I.U. Agoulnik, N.L. Weigel

androgen-independent cell lines, which suggests a role for Tip60 in both androgendependent and nonsteroidal activation of AR (Halkidou et al. 2003). Furthermore, androgen depletion increased Tip60 expression and nuclear accumulation, while androgen treatment decreased Tip60 levels.

2.1.3

Deacetylases

Acetylation of histones is an important reversible process in the regulation of transcriptional activation and silencing. The relative acetylation levels of histones and other proteins involved in transcription is maintained by the balancing action of HATs and histone deacetylases (HDAC). The three classes of HDACs are based on their homology to yeast transcriptional repressors. Classes I and II are the classical HDACs (Gregoretti et al. 2004), and Class III contains the newly discovered family of sirtuins (Michan and Sinclair 2007).

Classical HDACs Classes I and II are phylogenetically related to the yeast deacetylases, Rpd3 and Hda1, respectively. HDAC1, 2, 3, and 8 belong to class I and HDAC4, 5, 6, 7, 9, 10, and 11 belong to class II (Waltregny et al. 2004). HDACs from both classes are expressed in the prostate and can be inhibited by HDAC inhibitors, such as trichostatin A. HDAC1 is expressed predominantly in benign and malignant epithelium rather than stroma (Waltregny et al. 2004). HDAC8 is expressed uniquely in stromal cells of the prostate (Waltregny et al. 2004). Most HDACs are recruited to AR in multicomponent repressor complexes such as NCoR (nuclear receptor corepressor), SMRT (silencing mediator of retinoid and thyroid receptors), or DAX1 (dosage-sensitive sex-reversal, AHC, on the X-chromosome, gene 1) complexes. Unlike other deacetylases, HDAC7 binds to AR directly and inhibits receptor independent of its acetylation sites (Karvonen et al. 2006). Deacetylase activity is partially dispensable for HDAC7 repression of AR (Karvonen et al. 2006).

Sirtuins Class III HDACs are the Sir2 (silent mating type information regulation 2) family of NAD + dependent enzymes, SIRT1–7, and they are insensitive to class I and II HDAC inhibitors. SIRT1 deacetylates AR in a NAD-dependent fashion and inhibits its association with p300 and its transcriptional activity (Fu et al. 2006). As a result, SIRT1 inhibits NAD- and androgen-sensitive cellular proliferation (Fu et al. 2006). SIRT1 is thought to be a key protein in calorie-restriction-mediated life span extension and delay in carcinogenesis (Bordone and Guarente 2005; Cohen et al. 2004). However, expression of SIRT1 does not increase in the prostates of calorie-restricted mice (Huffman et al. 2007). Moreover, prostate cancer cells

Coregulators and the Regulation of Androgen Receptor Action in Prostate Cancer

325

express higher levels of SIRT1 than benign epithelial cells as determined using immunohistochemical analysis of human benign prostate and prostate cancer (Huffman et al. 2007). Similarly, immunohistochemical analysis revealed that TRAMP mice expressed higher levels of SIRT1 in poorly differentiated adenocarcinomas compared to those with less advanced disease (Huffman et al. 2007). The intriguing observation that expression of some HATs and HDACs increases with the progression of prostate cancer underscores the complexity of AR regulation by coactivators. The preinitiation complexes may be promoter specific; consequently, an increase in any one coregulator does not change all AR-regulated genes to the same extent.

Corepressors Some of the best characterized corepressors of AR are NCoR, SMRT, and DAX-1 (Agoulnik et al. 2003). Both NCoR and SMRT are recruited to the PSA promoter in an androgen-dependent manner (Shang et al. 2002). Though they have no enzymatic activities of their own, they act as scaffolding proteins that bring HDACs to the AR complexes and thus mediate repression. In addition, all three corepressors interfere with the carboxy- and amino terminal interaction of the AR dimer (Agoulnik et al. 2003). Limited information is available regarding expression of these proteins in prostate cancer. DAX-1 is a member of the nuclear receptor family, but lacks the classical DNA binding domain. DAX-1 is expressed abundantly in the luminal epithelium of the prostate and is localized to the nuclei. Expression is reduced drastically in benign prostatic hyperplasia, that may increase AR signaling (Agoulnik et al. 2003). These corepressors have diverse functions independent of AR signaling. SMRT facilitates DNA repair (Yu et al. 2006), NCoR sequesters p85-a, a regulatory subunit of PI3 kinase, which reduces its activity (Furuya et al. 2007), and DAX-1 directly binds to RNA and is found in polyribosomal complexes (Lalli et al. 2000).

2.2

Methyltransferases as AR Coactivators

Methyltransferases are enzymes that methylate specific arginine or lysine residues in histones. Either amino acid can be monomethylated or polymethylated and the biological consequences depend upon the specific methylation (Daniel et al. 2005). The pattern of methylation is indicative of transcriptional status of the chromatin (Seligson et al. 2005). While histone acetylation generally is associated with transcriptional activation, histone methylation plays roles both in stimulation and inhibition of transcription. Methylation of histones was considered an irreversible epigenetic signature until recently. A number of demethylases are now known to actively demethylate the repressed chromatin and balance methylase activity.

326

2.2.1

I.U. Agoulnik, N.L. Weigel

CARM1

CARM1 (Coactivator-associated arginine methyltransferase) methylates arginines 2, 17, and 26 of histone H3 (Chen et al. 1999). CARM1 is a secondary coactivator because it does not bind directly to AR but is recruited to receptor by the p160 coactivators (Chen et al. 1999). Thus, CARM1 activation of AR depends on the presence of p160 coactivators and their action is synergistic. In addition to modifying histones, CARM1 also methylates other proteins in the AR preinitiation complex, including p300 (Chen et al. 1999; Majumder et al. 2006). Depleting CARM1 in prostate cancer cells reduces AR transcriptional activity, diminishes proliferation, and induces apoptosis (Chen et al. 1999). CARM1 expression is higher in prostate cancer than benign prostate and is highest in castration-recurrent prostate cancer (Majumder et al. 2006).

2.2.2

G9a

The target for G9a methyltransferase activity is lysine 9 of histone H3, which is commonly associated with transcriptional repression. However, in the case of AR, G9a potentiates AR transcriptional activity synergistically with SRC-2, p300, and CARM1. G9a methyltransferase activity is dispensable for its role as an AR coactivator, which suggests that G9a functions as a scaffold rather than a methyltransferase (Lee et al. 2006).

2.2.3

PRMT1

Arginine 3 of histone H4 is the target for this arginine methyltransferase. PRMT1 binds to p160 coactivators and potentiates AR action like CARM1 (Wang et al. 2001).

2.3

Demethylases

Histone methylation originally was considered irreversible since the half-life of methyl lysine residues and histones are similar (Byvoet et al. 1972). Recent discoveries of demethylases show that histone methylation is dynamic in some regions of the genome (Metzger et al. 2005; Wissmann et al. 2007). Prostate cancer patients with a Gleason score of less than 7 have a lower 10-year recurrence rate if the percentage of cells with histone H3 Lysine 4 dimethylation staining is above the 60th percentile (Seligson et al. 2005). Increased methylation or decreased demethylation activity results in a more favorable prognosis for prostate cancer patients.

Coregulators and the Regulation of Androgen Receptor Action in Prostate Cancer

2.3.1

327

LSD1

Lysine-specific demethylase 1 (LSD1) interacts with AR in a ligand-dependent manner to demethylate the repressive histone modifications, mono- and dimethyl histone H3-lysine 9 (Metzger et al. 2005). Depletion of LSD1 in LNCaP cells reduces androgen-sensitive PSA induction and cellular proliferation. Immunohistochemical analysis of LSD1 expression in prostate cancer tissue microarrays revealed higher levels of LSD1 staining in the nuclei of patients who relapsed after radical prostatectomy and Kaplan–Meier analysis revealed shorter time to biochemical recurrence in patients with an LSD1 staining score of 8–12 compared to less than 8 (Kahl et al. 2006). 2.3.2

JARID1B

JARID1B (also called Plu-1) is a demethylase whose overexpression results in removal of mono-, di-, and trimethyl groups from histone H3 lysine 4 without changing H3K36me3 and H3K9me3 methylation (Xiang et al. 2007). JARID1B directly interacts with AR and its overexpression increases AR transcriptional activity on a PSA luciferase reporter (Xiang et al. 2007). JARID1B is undetectable in benign prostate tissue but expression is high in prostate cancer (Xiang et al. 2007). The role of endogenous JARID1B on various types of AR-regulated promoters requires further investigation since JARID1B is a strong transcriptional repressor in some promoter contexts; it interacts directly with the class I and II HDACs and indirectly with NCoR (Barrett et al. 2007). Alternatively, JARID1B’s primary activity in prostate cancer may be AR independent. Microarray analysis of the global knockdown and overexpression of JARID1B in breast cancer cells show that it represses the majority of its target genes (Scibetta et al. 2007).

2.3.3

JHDM2A

Unlike LSD1, which is constitutively associated with the PSA promoter (Yamane et al. 2006), JHDM2A interaction with AR and recruitment to AR responsive promoters is androgen dependent. JHDM2A can demethylate mono- and dimethyl H3 lysine 9. Knockdown of JHDM2A results in increased levels of dimethyl histone H3 lysine 9 at the promoters and decreased expression of the AR target genes PSA, TMPRSS2, and Nkx3.1 (Yamane et al. 2006).

2.3.4

JMJD2C

This demethylase interacts with LSD1 and is constitutively associated with ARregulated promoters (Wissmann et al. 2007). JMJD2C is a histone H3 lysine 9 tridemethylase that stimulates activity of AR cooperatively with LSD1 (Wissmann

328

I.U. Agoulnik, N.L. Weigel

et al. 2007). Depletion of JMJD2C in LNCaP cells reduced both androgen-sensitive PSA induction and cellular proliferation (Wissmann et al. 2007). JMJD2C is expressed ubiquitously and is also found in benign prostate and prostate cancer as determined using immunohistochemistry (Wissmann et al. 2007).

2.4

Ubiquitin Ligases

AR activity and turnover is regulated, in part, by ubiquitin ligases and ubiquitination of AR and histones at AR responsive genes. Ubiquitin ligases conjugate the small protein, ubiquitin, to the epsilon amino group of lysines in target proteins (Haas 2007). Ubiquitination plays a prominent role in the regulation of protein stability through targeting proteins for proteasome-dependent degradation. Degradation typically involves the formation of polyubiquitin chains linked to one or more lysines in target proteins. However, ubiquitination and particularly monoubiquitination play roles in a variety of other protein functions (Sigismund et al. 2004). E6AP ubiquitin ligase interacts directly with AR and activates AR transcriptional activity in a ligand-dependent manner (Khan et al. 2006). Consistent with this hypothesis, prostates are smaller in E6AP null mice compared to wildtype littermates. AR activity on the probasin promoter is decreased despite increased levels of AR protein in E6AP knockout mice (Khan et al. 2006). These data suggest that efficient turnover is necessary for optimal AR activity. Inhibition of the proteasome inhibits AR activity (Kang et al. 2002; Lin et al. 2002). However, E6AP expression is decreased in prostate cancer relative to benign tissue (Gao et al. 2005).

2.5

Deubiquitinases

Histone H2A was the first protein shown ubiquitinated on a highly conserved residue, lysine 119 (Goldknopf and Busch 1977). An H2A deubiquitinase JAMM/ MPN+ (H2A-DUB) is required for optimal activation of the AR in prostate cancer cells (Zhu et al. 2007). H2A-DUB interacts directly with p/CAF and is recruited to the PSA promoter in response to androgen treatment causing a coordinate increase in acetylation of histone H3 and a decrease in monoubiquitination of histone H2A. H2A-DUB overexpression increases AR transcriptional activity and its depletion from LNCaP cells reduces expression of AR target genes including PSA and Nkx3.1 (Zhu et al. 2007). Immunohistochemical analysis of tissue microarrays reveals a remarkable reduction in ubiquitinated H2A in prostate cancer compared to benign prostate (Zhu et al. 2007).

Coregulators and the Regulation of Androgen Receptor Action in Prostate Cancer

2.6

329

Small Ubiquitin-Like Modifier Proteins

A number of proteins in AR signaling are targets for sumoylation; these include AR, SRC-1, SRC-2, p300, and HDAC1. In contrast to ubiquitination, sumoylation does not target proteins for degradation. Mutations in SRC-2 sumoylation sites reduce its ability to activate AR (Kotaja et al. 2002). In contrast, mutation of the sumoylation site in the p300 carboxyl terminus increases p300-mediated transcriptional activity (Girdwood et al. 2003). AR is sumoylated in vivo on lysine residues 386 and 520 (Poukka et al. 2000). Mutation of these residues does not affect DNA binding but increases transcriptional activation by AR, which suggests a repressive role for small ubiquitin-like modifier (SUMO) conjugation in transcriptional activity. Multiple proteins are involved in AR sumoylation pathways. SUMO is an evolutionarily conserved protein; this suggests an important function in cells. There are four SUMO proteins in humans that vary slightly in length; only SUMO-1, -2, and -3 form covalent isopeptide bonds with the substrate proteins. SUMO-2 and -3 are most closely related (90% identity vs. about 50% identity with SUMO-1). SUMO-1 represses, while SUMO-2 and -3 increase AR activity (Callewaert et al. 2004; Poukka et al. 2000; Zheng et al. 2006). SUMO E2 ligase UBC-9 is an AR coactivator (Poukka et al. 1999), while other SUMO E3 ligases can either activate or repress AR activity (Kotaja et al. 2000). Very little is known about cell and endogenous promoter-specific actions of these proteins in benign prostate and prostate cancer. Sumoylation of proteins is dynamic and is reversed by SUMO proteases called SENPs. Humans express six SENPS that deconjugate SUMO from the large number of modified proteins. SENP1 is a SUMO protease that increases AR transcriptional activity (Cheng et al. 2004). The target protein for SENP1 is HDAC1, not AR. By desumoylating HDAC1, SENP1 reduces its histone deacetylase activity and therefore its ability to repress transcription (Cheng et al. 2004). SENP1 expression is upregulated in turn by androgens, which suggests a regulatory feedback mechanism (Bawa-Khalfe et al. 2007). SENP1 also is induced upon IL-6 treatment and androgens and IL-6 act synergistically to induce its expression (Cheng et al. 2004). Depletion of SENP1 causes a decrease in androgen-mediated cellular proliferation of LNCaP cells (Cheng et al. 2006). Cheng et al. (2004) investigated expression of SENP1 in normal prostate, high-grade prostate intraepithelial neoplasia (PIN), and prostate cancer using in situ hybridization and prostatectomy samples from 43 patients. Expression of SENP1 was increased in high-grade PIN and cancer.

2.7

Splicing Associated Factors

AR-induced transcription is a sequential multistep process involving receptor recruitment to promoters, chromatin remodeling, initiation, elongation, RNA splicing,

330

I.U. Agoulnik, N.L. Weigel

and termination. The coregulators described play a critical role in the initial steps of transcription. Recent data suggest an important role for AR in determining the splicing signature of the transcriptome in benign prostate and prostate cancer. Many AR target genes have numerous splicing variants. For example, the primary target genes KLK2 and KLK3 (PSA) have numerous splicing variants and protein products generated from the alternatively spliced forms, which are present in benign prostate and prostate cancer (David et al. 2002). Another example is a tumor suppressor and the earliest marker of prostate formation, Nkx3.1. Although initial expression of Nkx3.1 in prostate epithelium precedes that of AR, later expression of Nkx3.1 depends completely on androgen signaling. At least five alternatively spliced variants of Nkx3.1 exist in humans (Shen and Abate-Shen 2003). The effect of receptor-mediated transcription on splicing can be described as a combination of two synergistic models. The first model is based on the kinetics of transcriptional elongation and suggests that more rapid transcription favors exon skipping. The second model is based on the fact that AR can recruit various splicing factors to the RNA polymerase II (Pol II) complex. AR coactivators increase the rate of transcription via different mechanisms and thus could induce a higher rate of skipping. A number of splicing factors that interact with AR are recruited to AR responsive promoters to affect splicing according to the second model. Splicing factors can both activate and repress AR activity. For example, U5snRNP/ANT-1 increases AR transcriptional activity (Zhao et al. 2002) while COBRA1, PSF, and p54nrb repress it.

2.7.1

Cofactor of BRCA1 (COBRA1)

COBRA1 is a BRCA1-interacting protein and is a subunit in the negative elongation factor complex, NELF-B. COBRA1 interacts with AR directly and represses its transcriptional activity. Overexpression of COBRA1 induces exon skipping in the CD44 minigene driven by the androgen responsive MMTV promoter (Sun et al. 2007).

2.7.2

PTB-Associated Splicing Factor (PSF) and p54nrb

These two RNA splicing factors interact with the Pol II holoenzyme, bind to AR complexes, and inhibit AR transcriptional activity by recruiting HDAC complexes and destabilizing AR interaction with the ARE (Dong et al. 2007). A liganddependent recruitment of PSF to the PSA promoter occurs simultaneously with release from the PSA enhancer (Dong et al. 2007). Because PSF and p54nrb directly interact with the C terminal domain (CTD) of Pol II and AR, they can serve as a molecular link among the transcriptional machinery, splice site recognition, and splicing.

Coregulators and the Regulation of Androgen Receptor Action in Prostate Cancer

2.8 2.8.1

331

Signaling Molecules Kinases and Phosphatases

AR and its numerous coregulators are phosphoproteins and therefore their activity is subject to regulation by kinases and phosphatases. Tyrosine phosphorylation of AR is increased during progression to androgen independence; AR phosphorylation on tyrosine 534 is induced by activation of the Src kinase (Guo et al. 2006). AR phosphorylated on tyrosine 534 becomes more active at low androgen concentrations and mutation of this site abolished EGF-dependent AR activity (Guo et al. 2006). Phosphorylation of AR on tyrosine 534 is increased almost tenfold in castration-recurrent prostate cancer compared to benign prostate as determined using immunohistochemistry (Guo et al. 2006). Although some kinases and phosphatases influence receptor function through regulation of AR phosphorylation, others form more stable interactions with AR and function as coactivators.

2.8.2

Male Germ Cell-Associated Kinase

Male germ cell-associated kinase (MAK) is an AR coactivator. MAK is a direct target of AR and its expression is induced by androgens. Once expressed, MAK interacts directly with the AR, is recruited to AR-dependent promoters, and increases ligand-dependent transcriptional activity of AR and androgen-sensitive cellular proliferation (Ma et al. 2006).

2.8.3

Activated cdc42-Associated Tyrosine Kinase (Ack1)

Ack1 is a Her2-activated nonreceptor tyrosine kinase that phosphorylates AR on tyrosines 267 and 363 (Mahajan et al. 2007). Ack1 interacts directly with AR and is recruited to the PSA enhancer in an androgen-dependent manner (Mahajan et al. 2007). LNCaP cells expressing a constitutively active mutant of Ack1 formed tumors in nude mice faster than cells transfected with either vector or a kinase dead mutant of Ack1, and these tumors became resistant to castration after 2 weeks in animals (Mahajan et al. 2007). The Ack1 gene was amplified in primary prostate cancers and its transcript levels significantly increased in castration-recurrent prostate cancer (van der Horst et al. 2005). While of little importance for proliferation, overexpression of Ack1 significantly increases metastatic potential of prostate cancer cells (van der Horst et al. 2005). 2.8.4

Cdc25B

Along with multiple kinases, AR activity also is enhanced by the dual-specificity phosphatase, Cdc25B. Cdc25B interacts directly with AR and synergizes with

332

I.U. Agoulnik, N.L. Weigel

P/CAF and p300 in potentiating AR activity (Ma et al. 2001). Ngan et al. (2003) showed that Cdc25B expression assessed using immunohistochemistry increased in prostate cancers and correlated with Gleason score. 2.8.5

Protein-Kinase-C-Related-Kinase-1 (PRK1)

PRK1 interacts directly with the amino terminus of AR and increases its liganddependent transcriptional activation (Metzger et al. 2003). Inhibition of PRK1 using a chemical inhibitor or PRK1 depletion with siRNA represses AR transcriptional activity (Metzger et al. 2008). Androgen treatment causes its association with AR and concomitant recruitment to both the PSA promoter and enhancer. As PRK1 is recruited to the promoter, it phosphorylates H3T11 and causes increased demethylation of H3K9, which is associated with inactive chromatin. When PRK1 is inhibited or depleted, androgen treatment does not cause H3T11 phosphorylation and the androgendependent induction of AR target genes, PSA and KLK2, is reduced. Immunohistochemical analysis of PRK1 expression in tissue microarrays containing samples of 20 benign prostates and 111 prostate cancers revealed that higher expression of PRK1 and higher phosphorylation levels of H3T11 positively correlated with Gleason score and other aggressive characteristics of prostate cancer (Metzger et al. 2008). 2.8.6

Cyclin D1

Cyclin D1 interacts with AR and represses AR activity. Cyclin D1 is expressed as two splice variants, a and b; the D1a isoform strongly represses AR activity whereas the D1b is less effective (Burd et al. 2006). Cyclin D1a overexpression reduces proliferation of AR-dependent prostate cancer cell lines, while cyclin D1b mildly stimulates proliferation. Growth of prostate cancer cell lines that do not express AR is unaffected by overexpression of either cyclin D1a or cyclin D1b (Burd et al. 2006). A frequent polymorphism in the Cyclin D1 gene, G/A 870, is associated with increased risk of prostate cancer (Comstock et al. 2007). Cells with this polymorphic change in the gene have an increased tendency to express the cyclin D1b variant (Burd et al. 2006). Immunohistochemical analysis of cyclin D1 expression revealed that cytoplasmic localization of Cyclin D1 correlates with lower Ki-67 index (indicator of cellular proliferation), while nuclear staining positively correlated with higher Gleason score, increased proliferation, and nuclear localization of p21Cip1 (Comstock et al. 2007).

2.8.7

Cyclin D3/CDK11

Cyclin D3 forms a complex with the 58-kDa isoform of cyclin-dependent kinase 11 (CDK11p58) and AR, phosphorylates AR on serine 308, and represses AR induction of ARE, MMTV, and PSA-driven luciferase reporters, and endogenous PSA expression in LNCaP cells (Zong et al. 2007). Ectopic expression of cyclin D3 in

Coregulators and the Regulation of Androgen Receptor Action in Prostate Cancer

333

LNCaP cells increased accumulation of cells in the G1 phase of the cell cycle, but not in PC-3 cells, which suggests cyclin D3 may interfere with the mitogenic cellular response to androgens (Olshavsky et al. 2008); depletion of CDK11p58 resulted in a decline in BrdU incorporation in androgen-treated LNCaP cells and expression of cyclin D3 was necessary for this response (Zong et al. 2007). Total expression of cyclin D3 declined with prostate cancer progression and showed a negative correlation with cellular proliferation. (Olshavsky et al. 2008). 2.8.8

Isomerase and Wnt/b-Catenin Signaling

Pin1 is a peptidyl-prolyl isomerase that catalyzes cis/trans conversion of proteins upon binding to their phosphorylated Ser/Thr-Pro motifs. Multiple protein targets are regulated by Pin1 isomerase. Pin1 interferes with the ability of AR to inhibit the b-catenin/Tcf4 pathway that is in turn regulated by wnt signaling (Chen et al. 2006). b-catenin itself modestly increases AR activity and broadens ligand specificity (Truica et al. 2000). Wnt-1 and b-catenin are expressed at higher levels in lymph node and bone metastases than in primary cancers, and their expression positively correlates with Gleason score and serum PSA levels (Chen et al. 2004b). Expression analysis of Pin1 showed increased expression in prostate cancer compared to benign prostate tissue and predicted progression-free survival. Patients with high Pin1 expression have over eight times higher risk for recurrence than patients with low Pin1 expression (Ayala et al. 2003). A further increase in Pin1 expression was observed in metastatic prostate cancers compared to primary tumors using immunohistochemical evaluation of tissue microarrays (Chen et al. 2006). 2.8.9

PELP-1 (Proline-, Glutamic Acid-, and Leucine-Rich Protein-1)/ MNAR (Modulator of Nongenomic Activity of Estrogen Receptor)

MNAR is a scaffolding protein that brings together AR, general coactivators, such as p300 and CBP, and other coactivators, such as FHL2. In addition, MNAR complexes in the cytoplasm contain AR and src kinase and can activate downstream kinases, such as p42/p44 MAPK (Unni et al. 2004). In a limited MNAR expression analysis in prostate cancer, MNAR expression was higher in advanced prostate cancer specimens than in benign prostate tissue or low-grade prostate cancer (Nair et al. 2007).

3 Conclusion AR recruits a series of coactivator complexes to modulate transcription of target genes. Some of the coactivator/coregulators may be required regardless of the target, whereas others are needed to regulate subsets of target genes. Coregulators possess a wide range of enzymatic activities or serve as scaffold proteins to recruit other enzymes. Once recruited they covalently modify chromatin, other coregulators,

334

I.U. Agoulnik, N.L. Weigel

transcription factors, and Pol II thus modulating transcription initiation, elongation, and splicing. Therefore, there are a large number of coactivators that are required for AR-mediated transcription. Cell-based studies using siRNA that transiently deplete individual coactivators typically show reductions in AR transcriptional activity of greater than 50%. These findings suggest that cellular levels of coactivators are limiting and that in an acute situation, other closely related coactivators do not compensate for the loss of one coactivator. However, in vivo, the situation may be somewhat different. For example, the SRC-1 null mouse expresses twice the level of the closely related p160 coactivator, SRC-2/TIF2 (Xu et al. 1998). Receptors function as scaffolds to recruit a variety of coregulators and act to integrate the input from a variety of signaling pathways. The initial studies of the role of post-translational modifications in regulating coactivator function and transcription factor selectivity suggest that one explanation for the apparent redundancy in coactivators is that the individual coactivators respond to unique sets of signaling pathways (Wu et al. 2004). Thus, they may aid in coordinating cellular response to hormone with the level of activity of specific cell signaling pathways. The analyses of coactivator expression comparing benign tissue and prostate cancer or correlating expression with Gleason score often show a correlation between increased coactivator expression and prostate cancer, Gleason score, or time to biochemical recurrence. Global chromatin modifications are also somewhat predictive of patient outcome; however, the exact coregulators that cause these changes in chromatin are not known. The coregulators modulate activity of multiple transcription factors and, in some cases, also have activities distinct from their roles as coregulators, but the correlations between many of these, their ability to facilitate AR action, and altered levels in prostate cancer suggest that they are major factors in potentiating AR activity in prostate cancer. Some coactivators interact with AR through regions in the hormone binding domain and antagonists that alter the conformation of the hormone binding domain reduce these interactions limiting AR activity. However, the p160 coactivators interact with the amino terminal AF-1. Developing means to block interaction of coactivators with this important surface is a potential new target to block AR activity and prostate cancer progression. However, to our knowledge there has been no report showing that expression of one or even a combination of coactivators is an independent predictor of outcome. It is possible that expression of the enzymatic classes of coregulators will be more predictive.

References Agoulnik, I. U., and N. L. Weigel. 2006. Androgen receptor action in hormone-dependent and recurrent prostate cancer. J Cell Biochem 99: 362–72. Agoulnik, I. U., W. C. Krause, W. E. Bingman, III, H. T. Rahman, M. Amrikachi, G. E. Ayala, and N. L. Weigel. 2003. Repressors of androgen and progesterone receptor action. J Biol Chem 278: 31136–48.

Coregulators and the Regulation of Androgen Receptor Action in Prostate Cancer

335

Agoulnik, I. U., A. Vaid, W. E. Bingman, III, H. Erdeme, A. Frolov, C. L. Smith, G. Ayala, M. M. Ittmann, and N. L. Weigel. 2005. Role of SRC-1 in the promotion of prostate cancer cell growth and tumor progression. Cancer Res 65: 7959–67. Agoulnik, I. U., A. Vaid, M. Nakka, M. Alvarado, W. E. Bingman, III, H. Erdem, A. Frolov, C. L. Smith, G. E. Ayala, M. M. Ittmann, and N. L. Weigel. 2006. Androgens modulate expression of transcription intermediary factor 2, an androgen receptor coactivator whose expression level correlates with early biochemical recurrence in prostate cancer. Cancer Res 66: 10594–602. Anzick, S. L., J. Kononen, R. L. Walker, D. O. Azorsa, M. M. Tanner, X. Y. Guan, G. Sauter, O. P. Kallioniemi, J. M. Trent, and P. S. Meltzer. 1997. AIB1, a steroid receptor coactivator amplified in breast and ovarian cancer. Science 277: 965–8. Arnold, J. T., and J. T. Isaacs. 2002. Mechanisms involved in the progression of androgenindependent prostate cancers: it is not only the cancer cell’s fault. Endocr Relat Cancer 9: 61–73. Ayala, G., D. Wang, G. Wulf, A. Frolov, R. Li, J. Sowadski, T. M. Wheeler, K. P. Lu, and L. Bao. 2003. The prolyl isomerase Pin1 is a novel prognostic marker in human prostate cancer. Cancer Res 63: 6244–51. Barrett, A., S. Santangelo, K. Tan, S. Catchpole, K. Roberts, B. Spencer-Dene, D. Hall, A. Scibetta, J. Burchell, E. Verdin, P. Freemont, and J. Taylor-Papadimitriou. 2007. Breast cancer associated transcriptional repressor PLU-1/JARID1B interacts directly with histone deacetylases. Int J Cancer 121: 265–75. Bawa-Khalfe, T., J. Cheng, Z. Wang, and E. T. Yeh. 2007. Induction of the SUMO-specific protease 1 transcription by the androgen receptor in prostate cancer cells. J Biol Chem 282: 37341–9. Bevan, C. L., S. Hoare, F. Claessens, D. M. Heery, and M. G. Parker. 1999. The AF1 and AF2 domains of the androgen receptor interact with distinct regions of SRC1. Mol Cell Biol 19: 8383–92. Bordone, L., and L. Guarente. 2005. Calorie restriction, SIRT1 and metabolism: understanding longevity. Nat Rev Mol Cell Biol 6: 298–305. Burd, C. J., C. E. Petre, L. M. Morey, Y. Wang, M. P. Revelo, C. A. Haiman, S. Lu, C. M. Fenoglio-Preiser, J. Li, E. S. Knudsen, J. Wong, and K. E. Knudsen. 2006. Cyclin D1b variant influences prostate cancer growth through aberrant androgen receptor regulation. Proc Natl Acad Sci USA 103: 2190–5. Byvoet, P., G. R. Shepherd, J. M. Hardin, and B. J. Noland. 1972. The distribution and turnover of labeled methyl groups in histone fractions of cultured mammalian cells. Arch Biochem Biophys 148: 558–67. Callewaert, L., G. Verrijdt, A. Haelens, and F. Claessens. 2004. Differential effect of small ubiquitin-like modifier (SUMO)-ylation of the androgen receptor in the control of cooperativity on selective versus canonical response elements. Mol Endocrinol 18: 1438–49. Castoria, G., M. Lombardi, M. V. Barone, A. Bilancio, M. Di Domenico, D. Bottero, F. Vitale, A. Migliaccio, and F. Auricchio. 2003. Androgen-stimulated DNA synthesis and cytoskeletal changes in fibroblasts by a nontranscriptional receptor action. J Cell Biol 161: 547–56. Castoria, G., M. Lombardi, M. V. Barone, A. Bilancio, M. Di Domenico, A. De Falco, L. Varricchio, D. Bottero, M. Nanayakkara, A. Migliaccio, and F. Auricchio. 2004. Rapid signalling pathway activation by androgens in epithelial and stromal cells. Steroids 69: 517–22. Chen, D., H. Ma, H. Hong, S. S. Koh, S. M. Huang, B. T. Schurter, D. W. Aswad, and M. R. Stallcup. 1999. Regulation of transcription by a protein methyltransferase. Science 284: 2174–7. Chen, C. D., D. S. Welsbie, C. Tran, S. H. Baek, R. Chen, R. Vessella, M. G. Rosenfeld, and C. L. Sawyers. 2004a. Molecular determinants of resistance to antiandrogen therapy. Nat Med 10: 33–9. Chen, G., N. Shukeir, A. Potti, K. Sircar, A. Aprikian, D. Goltzman, and S. A. Rabbani. 2004b. Upregulation of Wnt-1 and beta-catenin production in patients with advanced metastatic prostate carcinoma: potential pathogenetic and prognostic implications. Cancer 101: 1345–56.

336

I.U. Agoulnik, N.L. Weigel

Chen, S. Y., G. Wulf, X. Z. Zhou, M. A. Rubin, K. P. Lu, and S. P. Balk. 2006. Activation of betacatenin signaling in prostate cancer by peptidyl-prolyl isomerase Pin1-mediated abrogation of the androgen receptor-beta-catenin interaction. Mol Cell Biol 26: 929–39. Cheng, J., D. Wang, Z. Wang, and E. T. Yeh. 2004. SENP1 enhances androgen receptor-dependent transcription through desumoylation of histone deacetylase 1. Mol Cell Biol 24: 6021–8. Cheng, J., T. Bawa, P. Lee, L. Gong, and E. T. Yeh. 2006. Role of desumoylation in the development of prostate cancer. Neoplasia 8: 667–76. Cohen, H. Y., C. Miller, K. J. Bitterman, N. R. Wall, B. Hekking, B. Kessler, K. T. Howitz, M. Gorospe, R. de Cabo, and D. A. Sinclair. 2004. Calorie restriction promotes mammalian cell survival by inducing the SIRT1 deacetylase. Science 305: 390–2. Comstock, C. E., M. P. Revelo, C. R. Buncher, and K. E. Knudsen. 2007. Impact of differential cyclin D1 expression and localisation in prostate cancer. Br J Cancer 96: 970–9. Culig, Z., A. Hobisch, M. V. Cronauer, C. Radmayr, J. Trapman, A. Hittmair, G. Bartsch, and H. Klocker. 1994. Androgen receptor activation in prostatic tumor cell lines by insulin-like growth factor-I, keratinocyte growth factor, and epidermal growth factor. Cancer Res 54: 5474–8. Daniel, J. A., M. G. Pray-Grant, and P. A. Grant. 2005. Effector proteins for methylated histones: an expanding family. Cell Cycle 4: 919–26. David, A., N. Mabjeesh, I. Azar, S. Biton, S. Engel, J. Bernstein, J. Romano, Y. Avidor, T. Waks, Z. Eshhar, S. Z. Langer, B. Lifschitz-Mercer, H. Matzkin, G. Rotman, A. Toporik, K. Savitsky, and L. Mintz. 2002. Unusual alternative splicing within the human kallikrein genes KLK2 and KLK3 gives rise to novel prostate-specific proteins. J Biol Chem 277: 18084–90. Debes, J. D., L. J. Schmidt, H. Huang, and D. J. Tindall. 2002. p300 mediates androgen-independent transactivation of the androgen receptor by interleukin 6. Cancer Res 62: 5632–6. Desai, S. J., A. H. Ma, C. G. Tepper, H. W. Chen, and H. J. Kung. 2006. Inappropriate activation of the androgen receptor by nonsteroids: involvement of the Src kinase pathway and its therapeutic implications. Cancer Res 66: 10449–59. Dong, X., J. Sweet, J. R. Challis, T. Brown, and S. J. Lye. 2007. Transcriptional activity of androgen receptor is modulated by two RNA splicing factors, PSF and p54nrb. Mol Cell Biol 27: 4863–75. Fu, M., M. Rao, C. Wang, T. Sakamaki, J. Wang, D. Di Vizio, X. Zhang, C. Albanese, S. Balk, C. Chang, S. Fan, E. Rosen, J. J. Palvimo, O. A. Janne, S. Muratoglu, M. L. Avantaggiati, and R. G. Pestell. 2003. Acetylation of androgen receptor enhances coactivator binding and promotes prostate cancer cell growth. Mol Cell Biol 23: 8563–75. Fu, M., C. Wang, X. Zhang, and R. G. Pestell. 2004. Acetylation of nuclear receptors in cellular growth and apoptosis. Biochem Pharmacol 68: 1199–208. Fu, M., M. Liu, A. A. Sauve, X. Jiao, X. Zhang, X. Wu, M. J. Powell, T. Yang, W. Gu, M. L. Avantaggiati, N. Pattabiraman, T. G. Pestell, F. Wang, A. A. Quong, C. Wang, and R. G. Pestell. 2006. Hormonal control of androgen receptor function through SIRT1. Mol Cell Biol 26: 8122–35. Furuya, F., C. J. Guigon, L. Zhao, C. Lu, J. A. Hanover, and S. Y. Cheng. 2007. Nuclear receptor corepressor is a novel regulator of phosphatidylinositol 3-kinase signaling. Mol Cell Biol 27: 6116–26. Gao, X., S. K. Mohsin, Z. Gatalica, G. Fu, P. Sharma, and Z. Nawaz. 2005. Decreased expression of e6-associated protein in breast and prostate carcinomas. Endocrinology 146: 1707–12. Gioeli, D., J. W. Mandell, G. R. Petroni, H. F. Frierson, Jr., and M. J. Weber. 1999. Activation of mitogen-activated protein kinase associated with prostate cancer progression. Cancer Res 59: 279–84. Girdwood, D., D. Bumpass, O. A. Vaughan, A. Thain, L. A. Anderson, A. W. Snowden, E. GarciaWilson, N. D. Perkins, and R. T. Hay. 2003. P300 transcriptional repression is mediated by SUMO modification. Mol Cell 11: 1043–54. Goldknopf, I. L., and H. Busch. 1977. Isopeptide linkage between nonhistone and histone 2A polypeptides of chromosomal conjugate-protein A24. Proc Natl Acad Sci USA 74: 864–8.

Coregulators and the Regulation of Androgen Receptor Action in Prostate Cancer

337

Gong, J., J. Zhu, O. B. Goodman, Jr., R. G. Pestell, P. N. Schlegel, D. M. Nanus, and R. Shen. 2006. Activation of p300 histone acetyltransferase activity and acetylation of the androgen receptor by bombesin in prostate cancer cells. Oncogene 25: 2011–21. Gregoretti, I. V., Y. M. Lee, and H. V. Goodson. 2004. Molecular evolution of the histone deacetylase family: functional implications of phylogenetic analysis. J Mol Biol 338: 17–31. Gregory, C. W., B. He, R. T. Johnson, O. H. Ford, J. L. Mohler, F. S. French, and E. M. Wilson. 2001. A mechanism for androgen receptor-mediated prostate cancer recurrence after androgen deprivation therapy. Cancer Res 61: 4315–9. Gregory, C. W., X. Fei, L. A. Ponguta, B. He, H. M. Bill, F. S. French, and E. M. Wilson. 2004. Epidermal growth factor increases coactivation of the androgen receptor in recurrent prostate cancer. J Biol Chem 279: 7119–30. Guo, Z., B. Dai, T. Jiang, K. Xu, Y. Xie, O. Kim, I. Nesheiwat, X. Kong, J. Melamed, V. D. Handratta, V. C. Njar, A. M. Brodie, L. R. Yu, T. D. Veenstra, H. Chen, and Y. Qiu. 2006. Regulation of androgen receptor activity by tyrosine phosphorylation. Cancer Cell 10: 309–19. Haas, A. L. 2007. Structural insights into early events in the conjugation of ubiquitin and ubiquitinlike proteins. Mol Cell 27: 174–5. Halkidou, K., V. J. Gnanapragasam, P. B. Mehta, I. R. Logan, M. E. Brady, S. Cook, H. Y. Leung, D. E. Neal, and C. N. Robson. 2003. Expression of Tip60, an androgen receptor coactivator, and its role in prostate cancer development. Oncogene 22: 2466–77. He, M. L., A. L. Jiang, P. J. Zhang, X. Y. Hu, Z. F. Liu, H. Q. Yuan, and J. Y. Zhang. 2005. Identification of androgen-responsive element ARE and Sp1 element in the maspin promoter. Chin J Physiol 48: 160–6. Heemers, H. V., T. J. Sebo, J. D. Debes, K. M. Regan, K. A. Raclaw, L. M. Murphy, A. Hobisch, Z. Culig, and D. J. Tindall. 2007. Androgen deprivation increases p300 expression in prostate cancer cells. Cancer Res 67: 3422–30. Huffman, D. M., W. E. Grizzle, M. M. Bamman, J. S. Kim, I. A. Eltoum, A. Elgavish, and T. R. Nagy. 2007. SIRT1 is significantly elevated in mouse and human prostate cancer. Cancer Res 67: 6612–8. Kahl, P., L. Gullotti, L. C. Heukamp, S. Wolf, N. Friedrichs, R. Vorreuther, G. Solleder, P. J. Bastian, J. Ellinger, E. Metzger, R. Schule, and R. Buettner. 2006. Androgen receptor coactivators lysine-specific histone demethylase 1 and four and a half LIM domain protein 2 predict risk of prostate cancer recurrence. Cancer Res 66: 11341–7. Kalkhoven, E. 2004. CBP and p300: HATs for different occasions. Biochem Pharmacol 68: 1145– 55. Kang, Z., A. Pirskanen, O. A. Janne, and J. J. Palvimo. 2002. Involvement of proteasome in the dynamic assembly of the androgen receptor transcription complex. J Biol Chem 277: 48366–71. Karvonen, U., O. A. Janne, and J. J. Palvimo. 2006. Androgen receptor regulates nuclear trafficking and nuclear domain residency of corepressor HDAC7 in a ligand-dependent fashion. Exp Cell Res 312: 3165–83. Khan, O. Y., G. Fu, A. Ismail, S. Srinivasan, X. Cao, Y. Tu, S. Lu, and Z. Nawaz. 2006. Multifunction steroid receptor coactivator, E6-associated protein, is involved in development of the prostate gland. Mol Endocrinol 20: 544–59. Kotaja, N., S. Aittomaki, O. Silvennoinen, J. J. Palvimo, and O. A. Janne. 2000. ARIP3 (androgen receptor-interacting protein 3) and other PIAS (protein inhibitor of activated STAT) proteins differ in their ability to modulate steroid receptor-dependent transcriptional activation. Mol Endocrinol 14: 1986–2000. Kotaja, N., U. Karvonen, O. A. Janne, and J. J. Palvimo. 2002. The nuclear receptor interaction domain of GRIP1 is modulated by covalent attachment of SUMO-1. J Biol Chem 277: 30283–8. Lalli, E., K. Ohe, C. Hindelang, and P. Sassone-Corsi. 2000. Orphan receptor DAX-1 is a shuttling RNA binding protein associated with polyribosomes via mRNA. Mol Cell Biol 20: 4910–21.

338

I.U. Agoulnik, N.L. Weigel

Lee, D. Y., J. P. Northrop, M. H. Kuo, and M. R. Stallcup. 2006. Histone H3 lysine 9 methyltransferase G9a is a transcriptional coactivator for nuclear receptors. J Biol Chem 281: 8476–85. Li, R., T. Wheeler, H. Dai, A. Frolov, T. Thompson, and G. Ayala. 2004. High level of androgen receptor is associated with aggressive clinicopathologic features and decreased biochemical recurrence-free survival in prostate: cancer patients treated with radical prostatectomy. Am J Surg Pathol 28: 928–34. Lin, H. K., S. Altuwaijri, W. J. Lin, P. Y. Kan, L. L. Collins, and C. Chang. 2002. Proteasome activity is required for androgen receptor transcriptional activity via regulation of androgen receptor nuclear translocation and interaction with coregulators in prostate cancer cells. J Biol Chem 277: 36570–6. Lu, S., G. Jenster, and D. E. Epner. 2000. Androgen induction of cyclin-dependent kinase inhibitor p21 gene: role of androgen receptor and transcription factor Sp1 complex. Mol Endocrinol 14: 753–60. Ma, Z. Q., Z. Liu, E. S. Ngan, and S. Y. Tsai. 2001. Cdc25B functions as a novel coactivator for the steroid receptors. Mol Cell Biol 21: 8056–67. Ma, A. H., L. Xia, S. J. Desai, D. L. Boucher, Y. Guan, H. M. Shih, X. B. Shi, R. W. deVere White, H. W. Chen, C. G. Tepper, and H. J. Kung. 2006. Male germ cell-associated kinase, a malespecific kinase regulated by androgen, is a coactivator of androgen receptor in prostate cancer cells. Cancer Res 66: 8439–47. Mahajan, N. P., Y. Liu, S. Majumder, M. R. Warren, C. E. Parker, J. L. Mohler, H. S. Earp, and Y. E. Whang. 2007. Activated Cdc42-associated kinase Ack1 promotes prostate cancer progression via androgen receptor tyrosine phosphorylation. Proc Natl Acad Sci USA 104: 8438–43. Majumder, S., Y. Liu, O. H. Ford, III, J. L. Mohler, and Y. E. Whang. 2006. Involvement of arginine methyltransferase CARM1 in androgen receptor function and prostate cancer cell viability. Prostate 66: 1292–301. Metzger, E., J. M. Muller, S. Ferrari, R. Buettner, and R. Schule. 2003. A novel inducible transactivation domain in the androgen receptor: implications for PRK in prostate cancer. EMBO J 22: 270–80. Metzger, E., M. Wissmann, N. Yin, J. M. Muller, R. Schneider, A. H. Peters, T. Gunther, R. Buettner, and R. Schule. 2005. LSD1 demethylates repressive histone marks to promote androgen-receptor-dependent transcription. Nature 437: 436–9. Metzger, E., N. Yin, M. Wissmann, N. Kunowska, K. Fischer, N. Friedrichs, D. Patnaik, J. M. Higgins, N. Potier, K. H. Scheidtmann, R. Buettner, and R. Schule. 2008. Phosphorylation of histone H3 at threonine 11 establishes a novel chromatin mark for transcriptional regulation. Nat Cell Biol 10: 53–60. Michan, S., and D. Sinclair. 2007. Sirtuins in mammals: insights into their biological function. Biochem J 404: 1–13. Nair, S. S., Z. Guo, J. M. Mueller, S. Koochekpour, Y. Qiu, R. R. Tekmal, R. Schule, H. J. Kung, R. Kumar, and R. K. Vadlamudi. 2007. Proline-, glutamic acid-, and leucine-rich protein-1/ modulator of nongenomic activity of estrogen receptor enhances androgen receptor functions through LIM-only coactivator, four-and-a-half LIM-only protein 2. Mol Endocrinol 21: 613–24. Ngan, E. S., Y. Hashimoto, Z. Q. Ma, M. J. Tsai, and S. Y. Tsai. 2003. Overexpression of Cdc25B, an androgen receptor coactivator, in prostate cancer. Oncogene 22: 734–9. Olshavsky, N. A., E. M. Groh, C. E. Comstock, L. M. Morey, Y. Wang, M. P. Revelo, C. Burd, J. Meller, and K. E. Knudsen. 2008. Cyclin D3 action in androgen receptor regulation and prostate cancer. Oncogene 27: 3111–21. Onate, S. A., S. Y. Tsai, M. J. Tsai, and B. W. O’Malley. 1995. Sequence and characterization of a coactivator for the steroid hormone receptor superfamily. Science 270: 1354–7. Popov, V. M., C. Wang, L. A. Shirley, A. Rosenberg, S. Li, M. Nevalainen, M. Fu, and R. G. Pestell. 2007. The functional significance of nuclear receptor acetylation. Steroids 72: 221–30. Poukka, H., P. Aarnisalo, U. Karvonen, J. J. Palvimo, and O. A. Janne. 1999. Ubc9 interacts with the androgen receptor and activates receptor-dependent transcription. J Biol Chem 274: 19441–6.

Coregulators and the Regulation of Androgen Receptor Action in Prostate Cancer

339

Poukka, H., U. Karvonen, O. A. Janne, and J. J. Palvimo. 2000. Covalent modification of the androgen receptor by small ubiquitin-like modifier 1 (SUMO-1). Proc Natl Acad Sci USA 97: 14145–50. Rowan, B. G., N. L. Weigel, and B. W. O’Malley. 2000. Phosphorylation of steroid receptor coactivator-1. Identification of the phosphorylation sites and phosphorylation through the mitogen-activated protein kinase pathway. J Biol Chem 275: 4475–83. Sathya, G., C. Y. Chang, D. Kazmin, C. E. Cook, and D. P. McDonnell. 2003. Pharmacological uncoupling of androgen receptor-mediated prostate cancer cell proliferation and prostatespecific antigen secretion. Cancer Res 63: 8029–36. Scibetta, A. G., S. Santangelo, J. Coleman, D. Hall, T. Chaplin, J. Copier, S. Catchpole, J. Burchell, and J. Taylor-Papadimitriou. 2007. Functional analysis of the transcription repressor PLU-1/JARID1B. Mol Cell Biol 27: 7220–35. Seligson, D. B., S. Horvath, T. Shi, H. Yu, S. Tze, M. Grunstein, and S. K. Kurdistani. 2005. Global histone modification patterns predict risk of prostate cancer recurrence. Nature 435: 1262–6. Shang, Y., M. Myers, and M. Brown. 2002. Formation of the androgen receptor transcription complex. Mol Cell 9: 601–10. Shen, M. M., and C. Abate-Shen. 2003. Roles of the Nkx3.1 homeobox gene in prostate organogenesis and carcinogenesis. Dev Dyn 228: 767–78. Sigismund, S., S. Polo, and P. P. Di Fiore. 2004. Signaling through monoubiquitination. Curr Top Microbiol Immunol 286: 149–85. Spencer, T. E., G. Jenster, M. M. Burcin, C. D. Allis, J. Zhou, C. A. Mizzen, N. J. McKenna, S. A. Onate, S. Y. Tsai, M. J. Tsai, and B. W. O’Malley. 1997. Steroid receptor coactivator-1 is a histone acetyltransferase. Nature 389: 194–8. Sun, J., A. L. Blair, S. E. Aiyar, and R. Li. 2007. Cofactor of BRCA1 modulates androgendependent transcription and alternative splicing. J Steroid Biochem Mol Biol 107: 131–9. Sykes, S. M., H. S. Mellert, M. A. Holbert, K. Li, R. Marmorstein, W. S. Lane, and S. B. McMahon. 2006. Acetylation of the p53 DNA-binding domain regulates apoptosis induction. Mol Cell 24: 841–51. Tomlins, S. A., D. R. Rhodes, S. Perner, S. M. Dhanasekaran, R. Mehra, X. W. Sun, S. Varambally, X. Cao, J. Tchinda, R. Kuefer, C. Lee, J. E. Montie, R. B. Shah, K. J. Pienta, M. A. Rubin, and A. M. Chinnaiyan. 2005. Recurrent fusion of TMPRSS2 and ETS transcription factor genes in prostate cancer. Science 310: 644–8. Truica, C. I., S. Byers, and E. P. Gelmann. 2000. Beta-catenin affects androgen receptor transcriptional activity and ligand specificity. Cancer Res 60: 4709–13. Ueda, T., N. R. Mawji, N. Bruchovsky, and M. D. Sadar. 2002. Ligand-independent activation of the androgen receptor by interleukin-6 and the role of steroid receptor coactivator-1 in prostate cancer cells. J Biol Chem 277: 38087–94. Unni, E., S. Sun, B. Nan, M. J. McPhaul, B. Cheskis, M. A. Mancini, and M. Marcelli. 2004. Changes in androgen receptor nongenotropic signaling correlate with transition of LNCaP cells to androgen independence. Cancer Res 64: 7156–68. van der Horst, E. H., Y. Y. Degenhardt, A. Strelow, A. Slavin, L. Chinn, J. Orf, M. Rong, S. Li, L. H. See, K. Q. Nguyen, T. Hoey, H. Wesche, and S. Powers. 2005. Metastatic properties and genomic amplification of the tyrosine kinase gene ACK1. Proc Natl Acad Sci USA 102: 15901–6. Waltregny, D., B. North, F. Van Mellaert, J. de Leval, E. Verdin, and V. Castronovo. 2004. Screening of histone deacetylases (HDAC) expression in human prostate cancer reveals distinct class I HDAC profiles between epithelial and stromal cells. Eur J Histochem 48: 273–90. Wang, H., Z. Q. Huang, L. Xia, Q. Feng, H. Erdjument-Bromage, B. D. Strahl, S. D. Briggs, C. D. Allis, J. Wong, P. Tempst, and Y. Zhang. 2001. Methylation of histone H4 at arginine 3 facilitating transcriptional activation by nuclear hormone receptor. Science 293: 853–7.

340

I.U. Agoulnik, N.L. Weigel

Wang, Q., J. S. Carroll, and M. Brown. 2005. Spatial and temporal recruitment of androgen receptor and its coactivators involves chromosomal looping and polymerase tracking. Mol Cell 19: 631–42. Wissmann, M., N. Yin, J. M. Muller, H. Greschik, B. D. Fodor, T. Jenuwein, C. Vogler, R. Schneider, T. Gunther, R. Buettner, E. Metzger, and R. Schule. 2007. Cooperative demethylation by JMJD2C and LSD1 promotes androgen receptor-dependent gene expression. Nat Cell Biol 9: 347–53. Wu, R. C., J. Qin, P. Yi, J. Wong, S. Y. Tsai, M. J. Tsai, and B. W. O’Malley. 2004. Selective phosphorylations of the SRC-3/AIB1 coactivator integrate genomic responses to multiple cellular signaling pathways. Mol Cell 15: 937–49. Xiang, Y., Z. Zhu, G. Han, X. Ye, B. Xu, Z. Peng, Y. Ma, Y. Yu, H. Lin, A. P. Chen, and C. D. Chen. 2007. JARID1B is a histone H3 lysine 4 demethylase up-regulated in prostate cancer. Proc Natl Acad Sci USA 104: 19226–31. Xin, L., M. A. Teitell, D. A. Lawson, A. Kwon, I. K. Mellinghoff, and O. N. Witte. 2006. Progression of prostate cancer by synergy of AKT with genotropic and nongenotropic actions of the androgen receptor. Proc Natl Acad Sci USA 103: 7789–94. Xu, J., Y. Qiu, F. J. DeMayo, S. Y. Tsai, M. J. Tsai, and B. W. O’Malley. 1998. Partial hormone resistance in mice with disruption of the steroid receptor coactivator-1 (SRC-1) gene. Science 279: 1922–5. Xu, J., L. Liao, G. Ning, H. Yoshida-Komiya, C. Deng, and B. W. O’Malley. 2000. The steroid receptor coactivator SRC-3 (p/CIP/RAC3/AIB1/ACTR/TRAM-1) is required for normal growth, puberty, female reproductive function, and mammary gland development. Proc Natl Acad Sci USA 97: 6379–84. Xu, J., and B. W. O’Malley. 2002. Molecular mechanisms and cellular biology of the steroid receptor coactivator (SRC) family in steroid receptor function. Rev Endocr Metab Disord 3: 185–92. Yamane, K., C. Toumazou, Y. Tsukada, H. Erdjument-Bromage, P. Tempst, J. Wong, and Y. Zhang. 2006. JHDM2A, a JmjC-containing H3K9 demethylase, facilitates transcription activation by androgen receptor. Cell 125: 483–95. Yan, J., C. T. Yu, M. Ozen, M. Ittmann, S. Y. Tsai, and M. J. Tsai. 2006. Steroid receptor coactivator-3 and activator protein-1 coordinately regulate the transcription of components of the insulin-like growth factor/AKT signaling pathway. Cancer Res 66: 11039–46. Yu, J., C. Palmer, T. Alenghat, Y. Li, G. Kao, and M. A. Lazar. 2006. The corepressor silencing mediator for retinoid and thyroid hormone receptor facilitates cellular recovery from DNA double-strand breaks. Cancer Res 66: 9316–22. Zhao, Y., K. Goto, M. Saitoh, T. Yanase, M. Nomura, T. Okabe, R. Takayanagi, and H. Nawata. 2002. Activation function-1 domain of androgen receptor contributes to the interaction between subnuclear splicing factor compartment and nuclear receptor compartment. Identification of the p102 U5 small nuclear ribonucleoprotein particle-binding protein as a coactivator for the receptor. J Biol Chem 277: 30031–9. Zheng, Z., C. Cai, J. Omwancha, S. Y. Chen, T. Baslan, and L. Shemshedini. 2006. SUMO-3 enhances androgen receptor transcriptional activity through a sumoylation-independent mechanism in prostate cancer cells. J Biol Chem 281: 4002–12. Zhou, G., Y. Hashimoto, I. Kwak, S. Y. Tsai, and M. J. Tsai. 2003. Role of the steroid receptor coactivator SRC-3 in cell growth. Mol Cell Biol 23: 7742–55. Zhou, H. J., J. Yan, W. Luo, G. Ayala, S. H. Lin, H. Erdem, M. Ittmann, S. Y. Tsai, and M. J. Tsai. 2005. SRC-3 is required for prostate cancer cell proliferation and survival. Cancer Res 65: 7976–83. Zhu, P., W. Zhou, J. Wang, J. Puc, K. A. Ohgi, H. Erdjument-Bromage, P. Tempst, C. K. Glass, and M. G. Rosenfeld. 2007. A histone H2A deubiquitinase complex coordinating histone acetylation and H1 dissociation in transcriptional regulation. Mol Cell 27: 609–21. Zong, H., Y. Chi, Y. Wang, Y. Yang, L. Zhang, H. Chen, J. Jiang, Z. Li, Y. Hong, H. Wang, X. Yun, and J. Gu. 2007. Cyclin D3/CDK11p58 complex is involved in the repression of androgen receptor. Mol Cell Biol 27: 7125–42.

Androgen Receptor Coregulators and Their Role in Prostate Cancer Latif A. Wafa, Robert Snoek, and Paul S. Rennie

Abstract The limiting factor in the survival of a patient with prostate cancer is the rate of progression to the noncurable androgen-independent (AI) stage of disease. The androgen receptor (AR) is a critical regulator of prostate cancer development and is involved in AI progression. Coregulators are proteins that interact directly with AR to enhance (coactivators) or reduce (corepressors) its transcriptional activity. Currently, over 165 AR coregulators have been discovered. In this chapter, we focus on a subset of the most well-characterized AR coregulators that are associated with prostate cancer. The first part of our review discusses the mechanisms by which classical type I and nonclassical type II AR coactivators/corepressors regulate AR transcriptional activity. The second section focuses on the role of coregulators in prostate cancer, including their expression profile in prostate cancer patients, tumor cell growth effects, and potential as therapeutic targets. In view of their involvement in prostate cancer progression, it is anticipated that further study of AR coregulators will provide more treatment options for increasing survival of patients with AI prostate cancer. Prostate cancer is the most commonly diagnosed nonskin cancer in men and the second leading cause of cancer death in North America (Jemal et al. 2004). While frequently curable in its early stages, many patients will present with locally advanced or metastatic disease for which there are currently no curative treatment options (Albertsen et al. 2005). Although androgen deprivation therapies, which block the growth promoting effects of androgens, are often used to treat advanced disease, progression to a lethal state variably referred to as castration-resistant, hormone refractory, castration-recurrent, or AI is the usual outcome, giving rise to a median survival of approximately 19 months (Albertsen et al. 2005; Kent and Hussain 2003; Martel et al. 2003). Some brief survival extensions can sometimes be achieved using current docetaxel-based chemotherapy (Petrylak 2005; Tannock et al. 2004). However, to have any major impact on current survival rates, more

P.S. Rennie(*) Department of Urologic Sciences, Faculty of Medicine, University of British Columbia, Prostate Centre at Vancouver General Hospital, Vancouver, BC, Canada, V6H 3Z6, E-mail: [email protected]

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4_15, # Springer Science + Business Media, LLC 2009

341

342

L.A. Wafa et al.

effective treatment strategies need to be developed. Because androgens mediate their effect on target cells through their receptor, it is thought that the AR is important in the development of AI disease. While the molecular mechanisms responsible for the development of AI prostate cancer are largely unknown, typically they do not appear to involve loss of AR expression (Taplin and Balk 2004). In over 80% of locally advanced AI prostate cancers, high levels of nuclear AR have been observed (Tilley et al. 1994), and in bone metastases, the amount of AR present is often higher than in primary tumors (Hobisch et al. 1995). There is evidence that in most cases, some form of inappropriate activation of AR is linked to recurrent, AI growth of prostate cancers (Taplin and Balk 2004; Rennie and Nelson 1998). In fact, in vivo knockdown of AR can delay AI progression (Cheng et al. 2006). One can envision a variety of molecular alterations that could lead to continued or amplified AR signaling following surgical or medical castration, such as gain of function AR mutations (Culig et al. 2003; Gaddipati et al. 1994; Marcelli et al. 2000; Newmark et al. 1992; Shi et al. 2002; Taplin et al. 2003) and amplification/overexpression of wild-type AR (Ford et al. 2003; Linja et al. 2001; Visakorpi et al. 1995). However, the available evidence suggests that the most commonly occurring mechanisms for progression to the AI stage are likely to be epigenetic, involving ligand-independent (or ligand-reduced) activation of AR either through convergence of cell signaling pathways and/or altered activity and expression of AR coregulators (Taplin and Balk 2004; Rennie and Nelson 1998; Sadar et al. 1999; Ueda et al. 2002). The latter is the focus of this chapter. Section 1 will describe the various types of AR coregulators that have been identified and Sect. 2 will discuss their role in prostate cancer and how they may provide therapeutic targets.

1 Coregulators of Androgen Receptor Activity 1.1

Overview of Androgen Receptor Coregulator Proteins

Transcriptional activity of androgen receptor (AR), as well as other members of the nuclear receptor superfamily, is modulated by coregulatory proteins. Generally, coregulators can be defined as proteins that interact directly with nuclear receptors to enhance (coactivators) or reduce (corepressors) transactivation of target genes, without significantly altering the basal transcriptional rate (McKenna et al. 1999). Coregulator proteins are not considered to possess specific DNA-binding ability (Xu et al. 1999). A list of more than 270 coregulators is available on the Nuclear Receptor Signaling Atlas (NURSA) website (http:www.nursa.org). Overall, a subset of 165-plus coregulators have been identified to interact with AR (Heemers and Tindall 2007). Many reviews have been written on coregulators (Heemers and Tindall 2007; Chmelar et al. 2007; Burd et al. 2006a; Kumar et al. 2005; Lonard et al. 2007; Dehm and Tindall 2007; Culig et al. 2005; Wang et al. 2005; Rosenfeld et al. 2006; Smith and O’Malley 2004) and this chapter will focus on a subset of the most well-characterized coregulators that are associated with prostate cancer.

Androgen Receptor Coregulators and Their Role in Prostate Cancer

343

Since coactivators use multiple mechanisms to influence AR transcription, they can be categorized, based on their functional characteristics, into two major types: classical/type I and nonclassical/type II coactivators (Heinlein and Chang 2002). Type I coactivators exert their function on AR while the receptor is at the target gene promoter and facilitate chromatin remodeling, DNA occupancy, or recruitment of basal transcription factors associated with the RNA polymerase (pol) II holocomplex (Sect. 1.2). Type II coactivators function primarily by modulating the appropriate folding of AR, ligand binding to AR, or facilitating the N-/ C-terminal interaction of AR. These events can contribute to AR protein stability, alter ligand-binding capability, or influence the subcellular distribution of AR, which ultimately leads to an increase in AR transcriptional activity. As discussed in Sect. 1.3, a wide array of type II coactivators has also been found for AR, which can act at many stages of receptor activation. In contrast to coactivators, corepressors suppress nuclear receptor transcriptional activity. Similar to coactivators, corepressors can be classified into two major types: classical/type I and nonclassical/type II corepressors (Sect. 1.4). Type I corepressors are characterized by their ability to bind nuclear receptors to repress transcription through the formation of nonproductive interactions with basal transcriptional machinery or through chromatin remodeling (Heinlein and Chang 2002). Type II corepressors function through other mechanisms such as inhibition of AR nuclear translocation/DNA-binding and disruption of both intermolecular interactions of AR and AR binding to coactivators. A partial list of the more established type I/II AR coactivators and corepressors is provided in Tables 1 and 3, respectively. A subset of reported AR coactivators that function through unique or unknown mechanisms is included in Table 2. More recently, involvement of coregulator proteins in the nongenomic actions of AR and signal transduction cascades has also been demonstrated (Sect. 1.5).

1.2 1.2.1

Type I Classical Coactivators of AR SRC/p160 Coactivators

The first identified and most extensively characterized of AR coregulators is the steroid receptor coactivator (SRC) family consisting of three 160-kDa proteins. Members of this family include SRC1, TIF2 (transcriptional intermediary factor 2) or SRC2, and AIB1 (amplified in breast cancer 1) or SRC3 (Onate et al. 1995; Ding et al. 1998; Alen et al. 1999). The SRC/p160 coactivators have been shown to interact with the AR-N-terminal domain (AR-NTD) and activation function 2 (AF2) surface of the AR-ligand-binding domain (AR-LBD) to enhance liganddependent transactivation of the receptor (Ikonen et al. 1997; Needham et al. 2000; Wang et al. 2001a; Callewaert et al. 2006). SRC1, SRC2, and SRC3 share similar structural organization that includes an N-terminal tandem basic helix-loophelix domain, a C-terminal glutamine rich region, and three LXXLL motifs in the

344 Table 1 Type I/II AR coactivators Name Interacting domain of AR Type I classical coactivators SRC/p160 family SRC1/NCoA-1 NTD, DBD, LBD SRC2/TIF2/GRIP1 NTD, DBD SRC3/AIB1/ACTR LBD and other domains Chromatin remodeling CBP/p300 NTD, DBD p/CAF NTD, DBD SRC1 and SRC3 NTD, DBD, LBD SWI/SNF Precise domain unknown Link to basal transcriptional machinerya RNA Pol II subunit LBD (RBP-2) TFIIB Interacts with SRC1 TFIIF NTD TFIIH NTD Type II nonclassical coactivators Molecular chaperonesa hop (p60) Precise domain unknown hsp40 (Ydj1) LBD hsp70 LBD hsp90 p23

LBD Precise domain unknown Modulation of ligand binding and AR stabilization ANPK (PKY) DBD ARA70 DBD, LBD

L.A. Wafa et al.

Selected references

Heinlein and Chang 2002; Onate et al. 1995 Heinlein and Chang 2002; Ding et al. 1998 Heinlein and Chang 2002; Ozers et al. 2007

Fronsdal et al. 1998 Reutens et al. 2001 Spencer et al. 1997; Chen et al. 1997 Marshall et al. 2003

Lee et al. 2003 Takeshita et al. 1996 McEwan and Gustafsson 1997 Lee et al. 2000

Chmelar et al. 2007; Freeman et al. 2000 Caplan et al. 1995; Fan et al. 2005 Veldscholte et al. 1992; Smith and Toft 1993; Shatkina et al. 2003 Marivoet et al. 1992; Fang et al. 1996 Freeman et al. 2000; Thomas et al. 2006 Moilanen et al. 1998 Heinlein and Chang 2002; Yeh and Chang 1996 Shatkina et al. 2003; Froesch et al. 1998

BAG-1L (hsp70 NTD, LBD cochaperone) Cellular trafficking of AR Caveolin-1 NTD, LBD Lu et al. 2001 Filamin Hinge domain Ozanne et al. 2000 Gelsolin DBD, LBD Nishimura et al. 2003 Ran/ARA24 NTD Hsiao et al. 1999 Supervillin NTD, DBD Ting et al. 2002 The organization of coactivators is based on their known functions and the mechanism by which they can enhance AR activity NTD N-terminal domain, DBD DNA-binding domain, LBD ligand-binding domain a These subcategories are not typically classified as coactivators but have been included here due to the close nature of their effects on AR with type I and II coactivators

Androgen Receptor Coregulators and Their Role in Prostate Cancer

345

Table 2 Other AR coactivators Name Mode of action Selected references b-Catenin Binds to AR-DBD and LBD to enhance AR Truica et al. 2000; transactivation, possibly by increasing receptor nuclear Mulholland et al. 2002 translocation ARA160 Binds to AR-NTD and can enhance AR transactivation Hsiao and Chang 1999 cooperatively with ARA70 coactivator in prostate cancer cells ARA267 Interacts with both N- and C-termini of AR to increase Wang et al. 2001b AR activity and can synergize with other coactivators, such as Ran/ARA24 and p/CAF ART-27 Binds to the NTD of AR and enhances transcription in Taneja et al. 2004 prostate cancer cells BRCA1 Tumor suppressor that binds to AR N-/C-termini and Yeh et al. 2000 synergistically enhances AR transcription with ARA70 CAK An AR-NTD interacting coactivator of AR and Lee et al. 2000 kinase moiety of the general transcription factor complex TFIIH HNF-3a DNA-binding protein that interacts with AR-DBD/hinge Gao et al. 2003 domain to promote assembly of AR-dependent transcription complexes, resulting in increased receptor activity hsp27 Binds to AR-NTD and LBD and enhances AR Zoubeidi et al. 2007 transcriptional activity Ku70 and Interact directly with AR-LBD, enhance AR transcription, Mayeur et al. 2005 Ku80 and can bind to prostate-specific antigen promoter in an androgen-dependent manner DDC Binds to AR-NTD and LBD to enhance AR transcriptional Wafa et al. 2003 activity through a ligand-dependent mechanism MAK Enhances AR transactivation potential in an androgenMa et al. 2006 and kinase-dependent manner in prostate cancer cells and can synergize with SRC3 RanBPM Binds to AR-NTD and DBD to enhance AR Rao et al. 2002b transcriptional activity Rb Tumor suppressor binds the NTD and LBD of AR Yeh et al. 1998 and enhances receptor transactivation in prostate cancer cells Tip60 Interacts with the LBD and hinge domain of AR and Brady et al. 1999 can increase ligand-dependent receptor transcription This list includes AR coactivators that have been shown to enhance receptor transactivation through unique or unknown mechanisms AR androgen receptor, DBD DNA-binding domain, LBD ligand-binding domain, NTD N-terminal domain, ART-27 AR-trapped clone-27, BRCA1 breast cancer-susceptibility gene 1, CAK Cdkactivating kinase, HNF-3a hepatocyte nuclear factor-3a, DDCL-Dopa decarboxylase, MAK male germ cell-associated kinase, Rb retinoblastoma protein, Tip60 TAT-interactive protein

central portion of the protein that are necessary for interaction with nuclear receptors (McKenna et al. 1999). Notably, binding of the LXXLL motifs of SRCs to the AR-LBD is weaker than with the LBD AF2 of other steroid receptors (e.g. glucocorticoid and estrogen receptors), possibly due to competition with the AR-NTD for

346

L.A. Wafa et al.

Table 3 Type I/II AR corepressors Name Interacting domain of AR Type I classical corepressors Histone deacetylase recruitment HDAC1 DBD, LBD NCoR LBD SMRT NTD, LBD

Selected references

Gaughan et al. 2002 Cheng et al. 2002 Dotzlaw et al. 2002; Liao et al. 2003

Type II nonclassical corepressors Inhibition of AR nuclear translocation and DNA binding Calreticulin DAX-1

DBD LBD

Dedhar et al. 1994 Holter et al. 2002

Disruption of AR N-/C-terminal interaction and coactivator binding Akt Precise domain unknown Lin et al. 2001 Cyclin D1 NTD, hinge domain Burd et al. 2006a; Reutens et al. 2001; Knudsen et al. 1999a GSK3b NTD, LBD Wang et al. 2004 p53 Precise domain unknown Shenk et al. 2001 SHP NTD, LBD Gobinet et al. 2001, 2005 Other corepressors of AR Alien NTD HBO1 DBD, LBD PSF Unknown PTEN DBD, LBD

Moehren et al. 2007 Sharma et al. 2000 Dong et al. 2007 Lin et al. 2004

The organization of corepressors is based on their known functions and the mechanism by which they can repress AR activity NTD N-terminal domain, DBD DNA-binding domain, LBD ligand-binding domain

binding with the C-terminal AF2 of AR (He et al. 1999, 2001). However, SRC coactivator LXXLL motif binding does play an important role in the regulation of AR activity and has been suggested to induce conformational changes in the AF2 of AR to modulate ligand-binding kinetics (He et al. 2006). The AR-LBD can also bind to FXXLF motifs contained in some AR coregulators, such as the type II AR-associated protein 70 (ARA70) coactivator, but this sequence is also present in the NTD of AR, which further adds to the competition between coactivators and the NTD for binding to the LBD of the receptor (He et al. 2002; Dubbink et al. 2004). SRCs can increase AR transactivation via their intrinsic histone acetyltransferase (HAT) activity, which allows maintenance of a transcriptionally open chromatin structure at the promoter of target genes (Spencer et al. 1997). Recently, SRC1 was shown to actually enhance the direct binding of AR to chromatin, presumably due to its HAT activity (Li et al. 2007). In addition, SRCs can act as platforms for the recruitment of secondary coactivators (Fig. 1), such as CREB-binding protein (CBP)/p300 and p300/CBP-associated factor (p/CAF) (see Sect. 1.2.2), which possess chromatin remodeling capabilities (Chakravarti et al. 1996; Blanco et al. 1998) and which can also bridge nuclear receptors to basal transcriptional machinery (see Sect. 1.2.3) (Takeshita et al. 1996).

Androgen Receptor Coregulators and Their Role in Prostate Cancer

347

Fig. 1 Type I AR coactivators. Type I/classical coactivators exert their function on androgen receptor (AR), while the receptor is at a target gene promoter. This is typified by the steroid receptor coactivator (SRC) family which binds to AR, contains intrinsic histone acetyltransferase activity, recruits secondary coactivators that possess chromatin remodeling capabilities, and can also bridge to general transcription factors (GTFs) associated with the RNA polymerase II holocomplex. ARR androgen responsive region, l dihydrotestosterone

1.2.2

Chromatin Remodeling Coactivators

Chromatin remodeling protein complexes act upon the nucleosome (the subunit of chromatin composed of a short length of DNA wrapped around a core of histone proteins) by either disrupting the histone–DNA interaction or controlling the acetylation status of histones. This disruption of chromatin structure allows transcription factors to bind more readily to DNA and thus facilitate transcriptional activation (Rosenfeld and Glass 2001; Rhodes 1997). AR coactivators that contain intrinsic HAT activity include CBP or its homologue p300 (CBP/p300), p/CAF, and members of the SRC/p160 family, SRC1 and SRC3 (Spencer et al. 1997; Chen et al. 1997; Blanco et al. 1998; Ogryzko et al. 1996; Heery et al. 1997). CBP/p300 has been shown to increase AR-mediated transcription by three- to fivefold (Aarnisalo et al. 1998) and p/CAF can reduce cyclin D1-mediated repression of AR transactivation (Reutens et al. 2001). AR itself has also been shown to be acetylated by the coactivators CBP/p300, p/CAF, and TAT-interactive protein (Tip60) (Fu et al. 2000; Gong et al. 2006; Gaughan et al. 2002). Acetylation of AR at the KXKK motif (amino acids 630–633) increases coactivator protein interactions and enhances AR transactivation (Fu et al. 2003). Also, members of the SRC family, such as SRC2, can be recruited simultaneously with CBP/p300 by AR to the prostate-specific antigen (PSA) promoter/enhancer for activation of transcription (Black et al. 2004; Kim et al. 2005). The switch/sucrose nonfermentable (SWI/SNF) multiprotein complex, which is able to perturb the conformation of the nucleosome by diminishing the interaction between DNA and histones, can also enhance transcription of AR at the PSA promoter (Marshall et al. 2003). The normal N/C-terminal interaction of AR is required to recruit the SWI/SNF complex, which in turn remodels chromatin to allow AR to bind to androgen response elements (Li et al. 2006). Both SWI/SNF

348

L.A. Wafa et al.

and CBP/p300 HAT activity may be required for androgen-dependent activation of AR. SWI/SNF can be targeted to chromatin by CBP/p300, which itself is recruited through interaction with SRC coactivators (Huang et al. 2003). Thus, multiple cofactors required for activation are not always recruited through direct interactions with steroid receptors but may involve a coordination of cofactor–cofactor interactions (Marshall et al. 2003; Chiba et al. 1994; Ratajczak 2001; Lambert and Nordeen 2003).

1.2.3

Role of Basal Transcriptional Machinery in Coactivation of AR

The transcriptional activation of steroid receptors ultimately requires the recruitment of RNA pol II to the promoter of target genes (Roeder 1996). RNA pol II is recruited through the assembly of the transcriptional preinitiation complex, formed by general transcription factors (GTFs). A large preinitiation complex consisting of three major components (RNA pol II, GTFs, and mediator), composed of 58 subunits, assembles before initiation of transcription occurs (Kornberg 2007). Coactivators of steroid receptors, such as SRC1, CBP/p300, and p/CAF, can also exert their activities by facilitating communication between the receptor and basal transcriptional machinery (Heinlein and Chang 2002). SRC1 interacts with TATAbinding protein and TFIIB and also recruits CBP/p300 to the transcriptional start site (Takeshita et al. 1996). In turn, CBP/p300, along with p/CAF, can bind directly with subunits of RNA pol II. AR can also interact directly with GTFs, such as TFIIF and TFIIH, through its NTD (McEwan and Gustafsson 1997; Lee et al. 2000). GTFs are not considered coregulator proteins since they can alter the basal transcription rate (McKenna et al. 1999). Analyses of AR-mediated transcription suggest that the orchestrated interaction of AR with TFIIF and TFIIH may increase efficiency of transcriptional elongation from androgen target genes, such as PSA (Lee and Chang 2003). AR may regulate transcription by enhanced assembly of the transcriptional initiation complex, by regulating promoter clearance, and during the elongation step of transcription (Lee and Chang 2003; Choudhry et al. 2006). Enhanced communication of AR with the basal transcriptional machinery, through other coactivators or direct GTF interactions, allows for more efficient transcription and an additional level of control in regulating AR activity.

1.3 1.3.1

Type II Nonclassical Coactivators of AR Role of Molecular Chaperones in Coactivation of AR

Chaperones prevent the irreversible aggregation of unfolded or partially folded proteins through recognition of and binding to their hydrophobic regions (Bukau and Horwich 1998). Although, chaperone proteins are not usually designated as coactivators, they interact with steroid receptors and modulate their activity. In the

Androgen Receptor Coregulators and Their Role in Prostate Cancer

349

absence of ligand in the cytoplasm, the structural conformation of the AR-LBD necessary for receptor activation is preserved by multiple cycles of binding and release with components of a multichaperone complex, which consist of heat-shock proteins (hsp) and cochaperone molecules (Prescott and Coetzee 2006). The minimal complex believed crucial for ligand-responsive signaling of steroid receptors consists of hsp70 (hsc70), hsp40 (ydj1), hop (p60; hsp organizer protein), hsp90, and p23 (Prescott and Coetzee 2006; Pratt and Toft 1997; Dittmar et al. 1998; Kosano et al. 1998). The interaction of AR with these proteins can maintain the receptor in a stable, partially unfolded conformation that is primed for high-affinity binding with androgens (Bohen et al. 1995). Upon binding of AR to ligand, the chaperone heterocomplex mediates trafficking of the receptor into the nucleus, possibly by facilitating interaction with dynein, a cytoplasmic protein that can drive active nuclear transport of the receptor along the cytoskeleton (Owens-Grillo et al. 1996; Davies et al. 2002). Hsp70, hsp40 (Rassow et al. 1995; Caplan and Douglas 1991), and hsp27 (Zoubeidi et al. 2007) have also been implicated in receptor translocation across intranuclear membranes. Within the nucleus, the chaperone complex dissociates from AR, but can still modulate receptor activity by promoting the dissociation of androgen from the receptor that returns AR to a primed state, which can be reactivated if ligand becomes available (Chmelar et al. 2007). However, in the absence of nuclear androgen, AR action is rapidly terminated and the receptor is degraded through the ubiquitin–proteosome pathway (Prescott and Coetzee 2006).

1.3.2

Coactivators of AR that Modulate Ligand Binding and Receptor Stability

The stability of AR protein depends on the ability of the AR-LBD to bind ligand (Krongrad et al. 1991). Binding of androgen to AR requires the proper folding of the receptor, which is regulated by the chaperone heterocomplex (see Sect. 1.3.1). Ligand binding induces the formation of an AF-2 coregulator-binding site (Dehm and Tindall 2007). Upon ligand binding, AR dimerizes, allowing the N- and C-termini of the receptor to interact. The occurrence of the N/C-terminal interaction of AR can result in decreased ligand dissociation and increased AR protein stability (Heinlein and Chang 2002). Hence, coactivators that can alter the folding of AR protein, receptor ligand binding, or the ability of the N- and C-termini to interact can potentially regulate AR transcriptional activity. Several proteins have been shown to coactivate AR transactivation through these mechanisms. A component of the hsp70 chaperone complex, Bcl-2-associated athagene-1 (BAG-1L), can increase AR transcription by interacting with the NTD and LBD of the receptor to facilitate proper folding (Shatkina et al. 2003; Froesch et al. 1998). Hsp90 can participate in the activation of AR by maintaining the receptor in a high-affinity ligand-binding conformation (Fang et al. 1996). The serine/threonine kinase AR-interacting nuclear protein kinase (ANPK) interacts with the AR-DBD and enhances AR-dependent transcription through stabilization

350

L.A. Wafa et al.

of AR protein levels (Moilanen et al. 1998). The most well-established coactivator of AR that modulates receptor function through stabilization of the ligand-bound receptor is ARA70 (Yeh and Chang 1996; Miyamoto et al. 1998). ARA70 has been implicated in changing the conformation of cytosolic AR so that it binds and/or retains androgen more easily, which leads to a faster rate of nuclear translocation (Hu et al. 2004). Furthermore, estrogen has been shown to activate AR and ARA70 retards the dissociation of estrogen from AR (Hu et al. 2004). ARA70 is a key example of how coactivators can regulate AR activity in the cytoplasm, prior to or during nuclear translocation. L-dopa decarboxylase (DDC) is another cytosolic AR-interacting protein that enhances receptor transactivation, possibly through modulating receptor androgen-binding ability (Wafa et al. 2003).

1.3.3

AR Coactivators Involved in Cellular Trafficking of the Receptor

The transactivation of steroid receptors can also be enhanced by coactivators that modulate cellular transport of the ligand-bound receptor. Both an increase in the rate of nuclear translocation and retention of AR in the nucleus can result in higher levels of transcription. Major structural components of cells such as filamentous actin (f-actin) play an important role in the cellular trafficking of proteins along the cytoskeleton (Stossel et al. 1985). Proteins that bind to f-actin mediate the process of actin-bundling, which influences cellular morphology and regulates actin polymerization and depolymerization. Several actin-binding proteins have been shown to alter AR function, including filamin, supervillin, and gelsolin (Ozanne et al. 2000; Nishimura et al. 2003; Ting et al. 2002). The f-actin cross-linking protein filamin interacts with the hinge domain of AR and facilitates receptor nuclear translocation (Ozanne et al. 2000). However, paradoxically, this protein has also been implicated in repression of AR transactivation (Loy et al. 2003). Supervillin binds directly to the NTD and DBD of AR, while gelsolin interacts with the AR-DBD and LBD (Nishimura et al. 2003; Ting et al. 2002); both increase AR transactivation. The association of AR with these actin-binding proteins is thought to serve as a mechanism by which the receptor can migrate along the cytoskeleton. The scaffolding protein, caveolin-1, is a principal component of caveolae membranes involved in many signal transduction pathways and is also another coactivator of AR that modulates transport of AR into the nucleus (Lu et al. 2001). High levels of caveolin-1 expression result in increased AR transactivation and nuclear localization of phosphorylated AR (Li et al. 2003). In contrast to the mechanism of AR coactivation by caveolin-1, Ras-related nuclear G-protein (Ran)/ARA24 has been suggested to enhance AR transcription by maintaining higher receptor levels in the nucleus (Heinlein and Chang 2002; Hsiao et al. 1999). In addition, a Ran-binding protein, RanBPM (ran-binding protein in the microtubule organizing center), has also been shown to enhance AR transcription (Rao et al. 2002b). Ran/ ARA24 is responsible for the nuclear export of importin proteins, which interact with nuclear localization signal-containing cargo proteins to dock protein complexes onto nuclear pores and allow nuclear import (Dasso 2001). Hence, an

Androgen Receptor Coregulators and Their Role in Prostate Cancer

351

Fig. 2 Type II androgen receptor (AR) coactivators. Type II/nonclassical coactivators can act at many stages of AR activation. These include modulation of AR folding by hsp chaperones, alteration of receptor ligand-binding capacity and AR protein stability in the cytosol, and influencing the trafficking of AR to the nucleus along the cytoskeleton. l dihydrotestosterone

increase in levels of Ran/ARA24 may result in rapid return of importins to the cytoplasm that in turn may result in more efficient translocation of proteins into the nucleus (Heinlein and Chang 2002). Figure 2 illustrates the various stages of AR activation at which type II coactivators can modulate receptor function.

1.4

Type I Classical and Type II Nonclassical Corepressors of AR

While the majority of AR coregulators identified have been coactivators, in recent years several corepressor proteins have been reported. The mechanisms by which corepressors suppress AR transcription seem to be as variable as that of coactivators. These include classical/type I histone deacetylase (HDAC) recruitment and nonclassical/type II mechanisms such as inhibition of the formation of active nuclear AR, inhibition of AR intermolecular interactions, or AR binding to coactivators and other unique processes (Wang et al. 2005). In addition to coactivators, corepressors may also be critical for regulation of AR transcription in a precise and efficient manner. A partial list of AR corepressors discussed here is provided in Table 3.

352

1.4.1

L.A. Wafa et al.

HDAC Recruitment (Type I)

The acetylation status of histones plays a critical role in the regulation of transcription as discussed previously for coactivators (Sect. 1.2). In contrast to coactivators that utilize HAT activity to disrupt chromatin structure and activate transcription, histone deacetylation of core histones can restabilize chromatin to suppress transcription by nuclear receptors (Xu et al. 1999). Silencing mediator for retinoid and thyroid hormone receptors (SMRT) corepressor and nuclear receptor corepressor (NcoR) are the best characterized nuclear receptor-binding proteins that suppress AR transcription, presumably through recruitment of HDACs and by competing with coactivators for interaction with the receptor (Wang et al. 2005). SMRT has been shown to interact with both the NTD and the LBD of AR (Dotzlaw et al. 2002; Liao et al. 2003), while NCoR only binds to the AR-LBD (Cheng et al. 2002). Moreover, HDAC1 has been shown to bind the AR-LBD/DBD and downregulate AR transcription (Gaughan et al. 2002). Recently, small interference RNA used to knock down SMRT and NCoR in prostate cancer cells enhanced the recruitment of the coactivators SRC1 and CBP/p300 by ligand-bound AR, which suggests that these corepressors can compete with coactivators for binding to the active receptor (Yoon and Wong 2006).

1.4.2

Inhibition of AR Nuclear Translocation and DNA-Binding (Type II)

AR can be targeted by corepressor proteins at multiple stages during receptor activation. DAX-1 (dosage-sensitive sex reversal adrenal hypoplasia congenita critical region on the X chromosome gene 1) corepressor interacts directly with the AR-LBD and sequesters the receptor in the cytoplasm (Holter et al. 2002). Calreticulin, a calcium-binding protein, has been shown to bind directly to the AR-DBD and thereby inhibit AR binding to androgen response elements and reduce receptor transactivation (Dedhar et al. 1994).

1.4.3

Disruption of AR N-/C-Terminal Interaction and Coactivator Binding (Type II)

The N-/C-terminal interaction of AR plays an important role in stabilizing AR protein and facilitating androgen binding for enhanced transactivation (Heinlein and Chang 2002). Glycogen synthase kinase 3b (GSK3b) is a serine/threonine kinase that phosphorylates a wide range of substrates, including the NTD of AR (Wang et al. 2004). Through direct binding to the AR-NTD, GSK3b is thought to suppress AR transcription by inhibiting the N-/C-terminal interaction of the receptor. The tumor suppressor protein p53 also can function as a corepressor of AR through disruption of the N-/C-terminal interaction, which may inhibit receptor homodimerization (Shenk et al. 2001).

Androgen Receptor Coregulators and Their Role in Prostate Cancer

353

Several corepressors have been found to alter AR transactivation through inhibiting AR association with coactivators. Cyclin D1, known to phosphorylate the cellular proliferation promoting Rb (retinoblastoma) protein, binds AR via its hinge domain and suppresses transactivation activity, possibly by disrupting association of AR with the p/CAF coactivator (Reutens et al. 2001; Knudsen et al. 1999a, 1999b). More recently, cyclin D1 has been shown to bind HDAC3, which suggests a role for this corepressor in chromatin remodeling (Petre-Draviam et al. 2005). Short heterodimer partner (SHP) is an orphan nuclear receptor that can repress AR transcriptional activity through competition with the SRC2 coactivator for binding to the receptor (Gobinet et al. 2001). More recently, SHP has also been shown to inhibit AR transactivation via recruitment of HDAC1 (Gobinet et al. 2005). The role of some AR corepressor proteins, such as Akt, is complex and controversial. Akt, an oncoprotein, is a serine/threonine kinase that plays a critical role in the phosphatidylinositol 3-kinase (PI3K)-mediated pathways. While some studies have shown Akt to be a positive modulator of AR transcriptional activity (Wen et al. 2000; Lin et al. 2003; Gregory et al. 2005), others have reported that Akt suppresses AR transactivation by binding and phosphorylating the receptor as well as inhibiting the interaction between AR and the coactivator ARA70 (Lin et al. 2001; Taneja et al. 2005).

1.4.4

Other Corepressors of AR (Type II)

Several other corepressors of AR have been reported, although the mechanism by which these corepressors suppress AR transactivation is not clear. PTEN (phosphatase and tensin homologue deleted on chromosome 10) tumor suppressor protein can interact with AR, inhibit nuclear translocation and promote AR protein degradation, which results in inhibition of receptor transactivation (Lin et al. 2004). HBO1 (histone acetyltransferase binding to the origin recognition complex-1 subunit protein) also binds to the DBD and LBD of AR to repress receptor transcriptional activity (Sharma et al. 2000). Recently, the NcoR Alien (Dressel et al. 1999) was shown to bind to the N-terminus of AR and inhibit endogenous PSA expression in LNCaP human prostate cancer cells (Moehren et al. 2007). RNA splicing factors, such as PSF (polypyrimidine tract-binding (PTB)-associated splicing factor), p54 nuclear RNA-binding protein (Dong et al. 2007), and COBRA1 (cofactor of BRCA1) (Sun et al. 2007), have also been shown to bind to AR and inhibit its transactivation.

1.5

Coregulator Involvement in Nongenomic and Rapid Signaling Events of AR

Addition of androgens to prostate cancer cells can result in activation of signaling pathways in a time frame too rapid to be accounted for by transcriptional activation of genes and thus has been termed nongenomic AR activity (Freeman et al. 2005).

354

L.A. Wafa et al.

The criteria for nongenomic actions are rapid onset (within minutes), resistant to inhibitors of RNA and protein synthesis, and activated at physiological concentrations of androgens (Boonyaratanakornkit and Edwards 2007). For AR to be involved in nongenomic events, it must be located outside the nucleus. There appear to be two types of receptors involved in nongenomic signaling. One is a putative membrane steroid receptor, which can bind to nonpermeable analogues such as bovine serum albumin-bound testosterone, to increase intracellular calcium, activate PI3K, phosphorylate focal adhesion kinase, and rearrange actin (Papakonstanti et al. 2003; Lyng et al. 2000; Kampa et al. 2002). However, the identity of this androgen-binding membrane protein remains unknown (Boonyaratanakornkit and Edwards 2007). The other type of receptor involved in nongenomic signaling is the bona fide AR. Cinar et al. (2007) have reported a subpopulation of AR found in lipid rafts in LNCaP cell membranes. The most likely mechanism for AR membrane localization is palmitoylation of the highly conserved nine-amino acid motif in the ligand-binding domain of AR (Pedram et al. 2007). Whether cytosolic AR physically associates with the cell membrane or is retained by a complex of proteins, addition of androgens leads to the AR associating with signaling proteins and rapid Src tyrosine kinase, mitogen-activated protein kinase (MAPK), and PI3K nongenomic steroid signaling (Boonyaratanakornkit and Edwards 2007). The AR coregulator PELP-1 (proline-, glutamic acid-, and leucine-rich protein-1), also known as MNAR (modulator of nongenomic activity of the estrogen receptor), facilitates AR interactions with cytoplasmic signaling components by forming a complex with Src-AR (Unni et al. 2004). PELP-1 also interacts with coregulators CBP/p300, cyclin D1, Rb (Vadlamudi and Kumar 2007), and FHL2 (four and a half LIMdomain protein 2) (Nair et al. 2007). Rapid signal transduction pathways can also activate AR in the absence of androgens or under androgen-deprived conditions (Edwards and Bartlett 2005a). Growth factors, cytokines, and neuropeptides can activate AR through the MAPK, janus kinase/signal transducers and activators of transcription (JAK/STAT), and PKA/PKC (protein kinase A and C) pathways, respectively [see recent reviews by Edwards and Bartlett (2005a, 2005b); Lange et al. (2007); and Weigel and Moore (2007)]. In addition, some of these pathways can also activate AR coregulators (Edwards and Bartlett 2005b; Lange et al. 2007; Weigel and Moore 2007; Gioeli 2005) (Fig. 3). Epidermal growth factor (EGF) activation of MAPK results in the phosphorylation of the coactivators SRC1 (Edwards and Bartlett 2005b) and SRC2 (Gregory et al. 2004). EGF also increases AR coactivator SRC2 (Gregory et al. 2004) and activated CREB (cAMP response element-binding protein), which increases ART-27 (Nwachukwu et al. 2007) expression. Levels of the AR corepressor cyclin D1 are increased by MAPK (Lange et al. 2007), and the growth factor, heregulin (Casimiro et al. 2007). Interleukin-6 cytokine activation of AR requires coactivators p300 (Debes et al. 2002) and SRC1 (Ueda et al. 2002). Src is a coactivator of AR and transiently phosphorylates the receptor at Tyr534 in response to EGF (Guo et al. 2006; Kraus et al. 2006). Interleukin-6 (Guo et al. 2006) and the neuropeptide bombesin (Desai et al. 2006) can also activate AR via Src. Moreover,

Androgen Receptor Coregulators and Their Role in Prostate Cancer

355

Fig. 3 Coregulator involvement in androgen receptor (AR) rapid signaling events. AR can be activated through signal transduction pathways that involve cytokines, growth factors, and neuropeptides. Only those pathways involving AR coregulators are depicted. Neuropeptide activation of PKA/PKC signaling can act upon AR itself or on coregulators through cross talk with the MAPK pathway. Growth factors activate MAPK, which results in activation of coactivators SRC1 and SRC2. Cytokine activation of AR requires the coactivators SRC1 and p300. Addition of androgens also leads to AR associating with Src tyrosine kinase and a rapid activation of both AR and Src, which can also activate the MAPK pathway. Boxes represent signal transduction proteins and circles depict AR coregulators. Arrows indicate connections but not necessarily direct interactions. l dihydrotestosterone

the Erk-2 (extracellular signal-regulated kinase) MAPK can phosphorylate AR at Ser514, which results in increased interaction with coactivator proteins (Yeh et al. 1999; Bakin et al. 2003; Rochette-Egly 2003; LaFevre-Bernt and Ellerby 2003).

1.6

Summary

AR activity is regulated by an ever enlarging number of coactivators and corepressors. However, the relative in vivo importance of modifying AR function via these coregulators and the relevance to pathological conditions remains to be demonstrated. The expression profile, tumor cell growth effects, and therapeutic potential of a number of AR coregulators in prostate cancer have been reported and are discussed in Sect. 2.

356

L.A. Wafa et al.

2 AR Coregulators in Prostate Cancer 2.1

Association of AR Coactivators with Prostate Cancer

Since coactivators enhance AR function in vitro, their expression may be deregulated in prostate cancer to provide a growth advantage. Recently, increased SRC1 expression was associated with increased tumor aggressiveness in clinically localized prostate cancer (Agoulnik et al. 2005). In the same study, reduction of SRC1 expression significantly reduced growth of androgen-sensitive LNCaP prostate cancer cells, whereas this had no effect on growth of AR-negative PC-3 and DU145 human prostate cancer cell lines. SRC3 has also been shown to be required for proliferation of androgen-sensitive prostate cancer cells and for tumor growth through controlling the expression of key cell-cycle genes (Zou et al. 2006). Moreover, SRC3 is overexpressed in prostate cancer patients and its expression correlates inversely with apoptosis of tumor cells (Zhou et al. 2005). The expression profile and tumor cell growth effects of several other AR coactivators have been reported. For example, CBP/p300 expression has been found to be upregulated during androgen deprivation therapy in patients with prostate cancer (Comuzzi et al. 2004). Knockdown of CBP/p300 transcripts using small interference RNA inhibits prostate cancer cell proliferation (Debes et al. 2003). The microRNA miR-26a for which CBP is a predicted target (Volinia et al. 2006) is also downregulated in prostate cancer (Mattie et al. 2006; Porkka et al. 2007), which could lead to increased expression of this coactivator. The core subunit of the SWI/SNF coactivator that mediates direct interactions with AR has been shown to be involved in the proliferation of AR-sensitive prostate cancer cells (Link et al. 2005). Expression of the ARA70 type II coactivator has been reported to increase in high-grade prostate cancer and in androgen-deprived prostate cancer cells (Hu et al. 2004; Chang et al. 2005). The AR cytosolic coactivator and neuroendocrine marker, DDC, has been shown to increase in expression with neo-adjuvant androgen deprivation therapy of prostate cancers (Wafa et al. 2007). Immunostaining of the AR chaperone hsp27 increases in androgen-independent (AI) tumors compared to untreated prostate cancer (Rocchi et al. 2004). In addition, expression of another AR coactivator, gelsolin, has been found to increase in LNCaP xenografts and human prostate cancers after androgen deprivation therapy (Nishimura et al. 2003). Recently, Tip60 AR coactivator was shown to increase in nuclear accumulation in 87% of AI prostate cancer specimens compared to benign prostatic hyperplasia (BPH) samples (Halkidou et al. 2003). Immunohistochemically, PELP-1 has been shown to increase in high-grade tumors (Nair et al. 2007) and HIP-1 (huntingtin interacting protein 1), a novel AR coactivator that prevents receptor protein degradation in prostate cancer cells (Mills et al. 2005), also increases in prostate cancers (Rao et al. 2002a). The above studies suggest that coactivators of AR not only facilitate receptor transactivation in vitro but also promote cell growth and play a crucial role in activation of AR in clinical prostate cancer.

Androgen Receptor Coregulators and Their Role in Prostate Cancer

2.2

357

Association of AR Corepressors with Prostate Cancer

Although effects of expression of the original AR corepressors SMRT and NCoR in prostate cancer are not clear, their ability to inhibit activated AR in prostate cancer cells suggests that loss of expression of these corepressors could facilitate tumor growth (Chmelar et al. 2007; Miyamoto et al. 2004). On the other hand, expression of cyclin D1 has been found to inhibit cell-cycle progression specifically in AR-dependent prostate cancer cells, which suggests that loss of this corepressor may promote proliferation of these cells (Burd et al. 2006b). However, the microRNA miR-20, whose predicted targets include cyclin D1 (Volinia et al. 2006), has been found to be downregulated in prostate cancer (Mattie et al. 2006). Indeed, Kaltz-Wittmer et al. (2000) found that the cyclin D1 gene was amplified in 25–37% of prostate cancer patients and Drobnjak et al. (2000) found cyclin D1 protein overexpression in metastatic relative to primary prostate cancer. While Akt has been shown to directly suppress AR activity, activation of AR by Akt through the PI3K-mediated pathway has been implicated in prostate cancer progression. Increased Akt activity can synergize with AR signaling to promote initiation and progression of prostate cancer to androgen independence in xenograft prostate cancer tumor models (Graff et al. 2000; Xin et al. 2006). Analysis of human prostate cancer tissues has demonstrated that, although there is neither Akt gene amplification nor enhanced protein expression in prostate cancer compared to normal prostate tissue, poorly differentiated tumors exhibit increased expression of a phosphorylated (activated) form of Akt compared to normal prostate tissue. This activated Akt may promote AI survival of prostate cancer cells through the activation of AR (Ghosh et al. 2003). The tumor suppressor proteins p53 and PTEN, both of which corepress AR activity, have been well documented to be mutated or lost in prostate cancer. PTEN is a critical regulator of prostate cancer growth and progression (Burd et al. 2006a). In fact, loss of PTEN, which occurs at a frequency of approximately 20–27% in prostate cancer, is associated with a decrease in survival rates (McMenamin et al. 1999; Halvorsen et al. 2003). It has also been shown that prostate-specific deletion of PTEN induces metastatic prostate cancer (Wang et al. 2003). Since the ability of PTEN to induce apoptosis in prostate cancer cells is reversible with androgen treatment, PTEN loss may result in enhancement of AR transactivation and resistance to cell death in prostate cancer cells (Li et al. 2001). Nevertheless, because PTEN plays an important role in regulation of many other cell survival pathways, such as PI3K/Akt, one cannot definitively conclude that its effects on prostate cancer growth are entirely due to loss of the PTEN corepressor function for AR (Paez and Sellers 2003). Mutations and loss of the p53 tumor suppressor protein are also crucial in cancer development and progression (Burd et al. 2006a). In prostate cancer, approximately 45% of tumors have been found to contain p53 mutations (Shi et al. 2004). It has been reported that there is a balance of AR and p53 expression during the androgenstimulated growth of prostate cancer, but this is lost during further progression

358

L.A. Wafa et al.

(Cronauer et al. 2004). Hence, similar to PTEN, it is tempting to speculate that loss of p53 function results in increased tumor growth, predominantly through an elevation in AR activity due to loss of corepressor function. However, it is difficult to separate this effect from the normal tumor suppressor role of p53. Nevertheless, it is possible that loss of AR corepressors, such as PTEN and p53, is at least in part responsible for increased activation of AR and subsequent enhancement of prostate cancer growth.

2.3

Expression Profile of AR Coregulators Associated with Progression of Clinical Prostate Cancer

Increasing data suggest that conventional androgen deprivation therapy does not completely eliminate intracellular androgens from prostate cancer cells (Mohler et al. 2004; Mostaghel et al. 2007a). After 9 months of androgen deprivation therapy, approximately 25% of the precastration androgen levels remain in prostatic tissues (Mostaghel et al. 2007b). Since coactivators increase AR transcriptional activity, their overexpression can potentially promote progression of prostate cancer by sensitizing the receptor to lower concentrations of androgens (Heinlein and Chang 2004). Moreover, the simultaneous loss of corepressor expression could further add to the unopposed AR activation effect of coactivators. Lonard et al. (2007) examined the ONCOMINE expression profiling metaanalysis database (www.oncomine.org) and found 117 overexpressed and 86 underexpressed coregulators in prostate cancer samples. In addition, they reviewed the literature and cite ten coregulators that were upregulated (ARA55, CARM1, Daxx, PDEF, PIAS1, RAF1, SENP1, SRC1, TSG101, and VAV3). Heemers and Tindall (2005) reviewed publications of protein expression in prostate cancer samples and found nine upregulated coactivators (SRC1, SRC2, SRC3, CBP, p300, Tip60, CARM-1, b-catenin, and ARA70). However, Chmelar et al. (2007) analyzed a number of studies that compared steady state levels of coregulator transcripts and found a lack of concordance due to variability in methods of analysis and heterogeneity of the samples. To assess the role of coregulators in prostate cancer progression to androgen independence, the literature was reviewed. A total of 18 studies analyzed clinical prostate cancer sample coregulator protein levels immunohistochemically and included samples from patients treated by androgen deprivation therapy. Heterogeneity of disease even within the same patient (Shah et al. 2004), different treatments (Chmelar et al. 2007), quality of tissue and availability, and quality of antibodies all combined to make evaluation of these studies difficult. As shown in Table 4, most were coactivators and most were overexpressed in AI prostate cancer. The only corepressor that also increased in AI prostate cancer was cyclin D1, which may be a consequence of its role in cell-cycle progression as opposed to that of an AR corepressor. Moreover, other immunohistochemical studies of prostate cancer

Androgen Receptor Coregulators and Their Role in Prostate Cancer Table 4 AR coregulators altered in clinical prostate cancer Name Number of patients Alteration ARA55 6 paired PCa and AI 2 patients had increase in AI disease BAG-1 32 untreated, 18 Increased nuclear preoperative staining for AI hormone treatment disease BAG-1L 190 untreated, 73 AI Increased nuclear staining for AI disease CARM1 66 untreated, 12 AI Increased staining (Coactivatorfor AI disease associated arginine methyltransferase 1) CBP 0 untreated, 26 AI High nuclear expression Cyclin D1 86 PCa (primary), Increased staining 22 AI bone for AI disease metastases DDC 21 PCa, 21 < 3 months Increased with >6 ADT, 28 3–6 months ADT and months ADT, in AI disease 28 > 6 months ADT, 13 AI DJ-1 21 PCa, 21 < 3 months Increased in patients ADT, 28 3–6 with >6 months months ADT, ADT 28 > 6 months ADT GAK 21 PCa, 49 6 months Increased with ADT ADT, 28 > 6 and in AI disease months ADT, 13 AI Gelsolin 11 PCa, 16 androgen Higher expression deprivation with ADT HIP-1 114 PCa (localized), Higher expression 14 AI metastatic with ADT hsp27 35 PCa, 58 < 3 months Increased with ADT ADT, 52 3–6 and in AI disease months ADT, 57 > 6 months ADT, 30 AI SRC1 205 PCa, 49 AI Increased staining in AI disease but not significant SRC1, SRC2 8 BPH, 8 PCa, 8 AI More intense in AI disease SRC2 518 total PCa samples More intense in AI disease including 9 AI Tip60 10 BPH, 43 PCa, 15 AI

359

References Fujimoto et al. 2007

Krajewska et al. 2006 Maki et al. 2007

Hong et al. 2004

Comuzzi et al. 2004 Drobnjak et al. 2000

Wafa et al. 2007

Tillman et al. 2007

Ray et al. 2006

Nishimura et al. 2003 Rao et al. 2002a Rocchi et al. 2004

Maki et al. 2006

Gregory et al. 2001 Agoulnik et al. 2006 Halkidou et al. 2003 (continued)

360

L.A. Wafa et al.

Table 4 (Continued) Name Number of patients

Alteration References Increased nuclear staining in AI disease (87%) vs untreated tumors (37%) aSGT (small 30 PCa, 54 metastatic AR/aSGT ratio Buchanan et al. 2007 glutamine-rich increased in tetratricopeptide metastatic vs PCa repeat and in AI versus containing nontreated protein) metastatic cochaperone disease b-Catenin 21 AI metastatic 24% (5/21) positive Chesire et al. 2002 for nuclear staining This list of coregulators is based on studies of immunohistochemically stained clinical samples including patients treated with androgen deprivation therapy PCa untreated prostate cancer, AI androgen-independent, ADT androgen deprivation therapy and BPH benign prostatic hyperplasia, GAK cyclin G-associated kinase

have shown that the coactivators SRC3 (Gnanapragasam et al. 2001), b-catenin (Aaltomaa et al. 2005; Assikis et al. 2004), caveolin-1 (Karam et al. 2007), lysinespecific histone demethylase-1, and FHL2 (Kahl et al. 2006) all predict aggressive prostate cancer. These studies indicate that dysfunctional expression of coregulators may play a critical role in progression of prostate cancer to AI disease. Although several coregulators have been shown to be upregulated in prostate cancer (Heemers and Tindall 2005; Culig and Bartsch 2006) (Table 4), whether these proteins contribute to prostate cancer progression remains unclear.

2.4

In Vivo Functional Studies of AR Coregulators in Prostate Cancer

There are only a few examples where a functional role for coregulators in prostate cancer has been demonstrated in animal models. PTEN is one of the best examples of this type of in vivo analysis. Prostate-specific PTEN null mice spontaneously develop prostate cancer (Sect. 2.2) (Wang et al. 2003). Studies with knockout mice (Lonard et al. 2007) have shown Zimp10 (a novel PIAS type coregulator) (Beliakoff and Sun 2006), p300 (Yao et al. 1998), ARIP4 [androgen receptor (AR)interacting protein 4] (Zhang et al. 2007), and NCoR (Hermanson et al. 2002) are all embryonically lethal, which suggests that they have a critical role in early development. Fkbp52 chaperone protein (the protein product of FKBP4)-null

Androgen Receptor Coregulators and Their Role in Prostate Cancer

361

mice have defective development of the prostate (Yong et al. 2007), which implies an essential physiological role for AR-mediated signaling. Moreover, induced overexpression of DDC in LNCaP xenografts increased PSA production, which demonstrates an in vivo coactivation effect of this AR-binding protein (Wafa et al. submitted for publication). Activated ErbB-2-induced cyclin D1 expression in the mouse prostate is associated with prostate intraepithelial neoplasia (Casimiro et al. 2007). Although SRC3 has been shown to be required for prostate cancer growth (Zou et al. 2006), SRC3-null mice show a reduction of IGF (insulin-like growth factor)/ Akt signaling in prostate glands (Yan et al. 2006), which may mean that the effects of SRC3 are not exclusively via AR but involve genes in the IGF/Akt pathway. However, SRC3-null TRAMP (transgenic adenocarcinoma of the mouse prostate) mice fail to progress to poorly differentiated adenocarcinoma, which supports the notion that SRC3 is required for tumor progression (Chung et al. 2007). Disruption of SRC1 in mice results in decreased growth and development of the prostate, but knockout of SRC2 has no effect on the gland (Xu et al. 1998; Gehin et al. 2002; Mark et al. 2004). The single and combinatorial knockout targeting of these coactivators suggest that they have partial functional redundancy in vivo, whereby the absence of one SRC can be compensated for by another family member or possibly by non-SRC coactivators.

2.5

Therapeutic Targeting of Coregulators in Prostate Cancer

One of the major stumbling blocks to prostate cancer treatment is being able to predict the aggressiveness of tumors. Specific coregulators may be useful as predictive markers. Moreover, since there is no prognostic test for the clinical outcome of androgen deprivation therapy, Ross (2007) postulated that focusing on coregulators may be predictive of AR pathway status and guide prostate cancer treatment. Since aberrant modulation of AR activity by coregulators has been suggested to be involved in progression of prostate cancer, some coregulators may serve as potential therapeutic targets (Heemers and Tindall 2005). A detailed understanding of the binding motifs of coregulators to AR and computational chemistry may allow design of agents that mimic binding surfaces of protein complexes to target protein–protein interactions (target biomotifs). The aim of this approach is to block coactivator and increase corepressor binding to AR (Fletterick et al. 2005) (Fig. 4). Migliaccio et al. (2007) created a ten-amino acid peptide that inhibited the AR–Src interaction, which reduced LNCaP xenograft growth. An alternative approach would be restoration of the function of a protein lost during cancer progression. For example, DOC-2/DAB2 (differentially expressed in ovarian cancer-2/disabled 2), which is expressed in normal prostate and primary prostate cancers but not in cell lines derived from AI cancer, has been shown to disrupt AR interaction with Src (Zhoul et al. 2005). Zhou et al. (2006) attached a proline-rich sequence derived from DOC-2/DAB2 to a cell-permeable

362

L.A. Wafa et al.

Fig. 4 Targeting AR coregulators in prostate cancer. Depicted are three different types of therapies targeting AR coregulators. In clinical trial, are therapies targeting chaperones (hsp) to decrease protein levels of AR and Src kinase inhibitors to prevent activation of AR. Target biomotifs to AR–coregulator interactions, aimed at blocking coactivator or increasing corepressor binding, to decrease AR transcriptional activity are under development. ARR androgen responsive region, l dihydrotestosterone, bound or unbound dihydrotestosterone

peptide delivery system, which restored the function of DOC-2/DAB2 and resulted in suppression of androgen-induced growth of prostate cancer cells. Moreover, AZD0530, a dual Src/v-Abl (Abelson murine leukemia viral oncogene homologue) kinase inhibitor is in multicenter phase II clinical trials for a variety of malignancies, which include prostate cancer (Chang et al. 2007). Molecular chaperones of AR, such as hsp, have also been targeted to abrogate AR activity. hsp90 is a therapeutic target in prostate cancer clinical drug trials, where inhibition of hsp90 has been shown to lead to increased degradation of AR (Neckers and Ivy 2003; Harashima et al. 2005). In addition, a multicenter phase I clinical study is underway targeting hsp27 (ClinicalTrials.gov identifier NCT00487786). High-throughput functional screening of known drugs is also being pursued to identify coregulator targets. Estebanez-Perpina et al. (2007) assessed more than 55,000 compounds for their ability to inhibit AR AF-2/SRC2 LXXLL binding. They found three off-patent aspirin derivatives already approved for human use and two types of thyroid hormones that could block binding of SRC2. In addition, they discovered a previously unrecognized allosteric regulatory site (binding function-3)

Androgen Receptor Coregulators and Their Role in Prostate Cancer

363

near the AF2 of AR, which could be a pharmaceutical target. Selective AR modulators (SARMs) are being designed to modulate coregulator recruitment (Gao and Dalton 2007). Targeting of a wide variety of classic and nonclassic AR coregulators, as part of combinatorial AR targeted therapy, may be efficacious in the treatment of advanced prostate cancer (Singh et al. 2006).

3 Conclusions The limiting factor in the survival of a patient with prostate cancer is the rate of progression to AI disease (Albertsen et al. 2005; Kent and Hussain 2003; Martel et al. 2003). In order to impact on mortality rates, prostate cancers that will rapidly progress to the AI stage must be identified and the AI phenotype must be prevented, delayed, or treated effectively. AR is a critical regulator of prostate cancer development and AI progression. Since recurrent AI prostate cancer cells generally retain AR activity, inhibition of AR function must be a central focus of treatment. Because of this importance of AR in prostate cancer, a plethora of studies have tried to identify specific coregulator proteins that modify AR action to allow growth in spite of therapeutic intervention. Coactivators of AR can contribute to aberrant AR activation under androgen-deprived conditions and can be predictive for aggressive prostate cancers. Moreover, loss of AR corepressor function may contribute to deregulated AR activity during prostate cancer progression. Hence, in addition to the targeting of AR directly and ligand-independent/ligand-reduced signaling pathways that activate the receptor, AR coregulator proteins may provide targets for therapy. Since numerous coregulators have been shown to modulate AR activity and since compensatory mechanisms of action exist, it is probably necessary to simultaneously target many of these receptor-binding proteins in order to achieve maximal suppression of AR activity. With the ability to target coregulators using new strategies and drugs that effectively shut down AR signaling, more effective treatment options may arise to improve the clinical outcome of patients with AI prostate cancer.

References Aaltomaa S. et al. 2005. Reduced alpha- and beta-catenin expression predicts shortened survival in local prostate cancer. Anticancer Res 25:4707–4712. Aarnisalo P. et al. 1998. CREB-binding protein in androgen receptor-mediated signaling. Proc Natl Acad Sci U S A 95:2122–2127. Albertsen P. C. et al. 2005. 20-Year outcomes following conservative management of clinically localized prostate cancer. JAMA 293:2095–2101. Alen P. et al. 1999. The androgen receptor amino-terminal domain plays a key role in p160 coactivator-stimulated gene transcription. Mol Cell Biol 19:6085–6097.

364

L.A. Wafa et al.

Agoulnik I. U. et al. 2005. Role of SRC-1 in the promotion of prostate cancer cell growth and tumor progression. Cancer Res 65:7959–7967. Agoulnik I. U. et al. 2006. Androgens modulate expression of transcription intermediary factor 2, an androgen receptor coactivator whose expression level correlates with early biochemical recurrence in prostate cancer. Cancer Res 66:10594–10602. Assikis V. J. et al. 2004. Clinical and biomarker correlates of androgen-independent, locally aggressive prostate cancer with limited metastatic potential. Clin Cancer Res 10:6770–6778. Bakin R. E. et al. 2003. Constitutive activation of the Ras/mitogen-activated protein kinase signaling pathway promotes androgen hypersensitivity in LNCaP prostate cancer cells. Cancer Res 63:1981–1989. Beliakoff J. and Sun Z. 2006. Zimp7 and Zimp10, two novel PIAS-like proteins, function as androgen receptor coregulators. Nucl Recept Signal 4:e017. Black B. E. et al. 2004. Transient, ligand-dependent arrest of the androgen receptor in subnuclear foci alters phosphorylation and coactivator interactions. Mol Endocrinol 18:834–850. Blanco J. C. et al. 1998. The histone acetylase PCAF is a nuclear receptor coactivator. Genes Dev 12:1638–1651. Bohen S. P. et al. 1995. Hold ‘em and fold ‘em: chaperones and signal transduction. Science 268:1303–1304. Boonyaratanakornkit V. and Edwards D. P. 2007. Receptor mechanisms mediating non-genomic actions of sex steroids. Semin Reprod Med 25:139–153. Brady M. E. et al. 1999. Tip60 is a nuclear hormone receptor coactivator. J Biol Chem 274:17599– 17604. Buchanan G. et al. 2007. Control of androgen receptor signaling in prostate cancer by the cochaperone small glutamine rich tetratricopeptide repeat containing protein alpha. Cancer Res 67:10087–10096. Bukau B. and Horwich A. L. 1998. The Hsp70 and Hsp60 chaperone machines. Cell 92:351–366. Burd C. J. et al. 2006a. Androgen receptor corepressors and prostate cancer. Endocr Relat Cancer 13:979–994. Burd C. J. et al. 2006b. Cyclin D1b variant influences prostate cancer growth through aberrant androgen receptor regulation. Proc Natl Acad Sci U S A 103:2190–2195. Callewaert L. et al. 2006. Interplay between two hormone-independent activation domains in the androgen receptor. Cancer Res 66:543–553. Caplan A. J. and Douglas M. G. 1991. Characterization of YDJ1: a yeast homologue of the bacterial dnaJ protein. J Cell Biol 114:609–621. Caplan A. J. et al. 1995. Hormone-dependent transactivation by the human androgen receptor is regulated by a dnaJ protein. J Biol Chem 270:5251–5257. Casimiro M. et al. 2007. ErbB-2 induces the cyclin D1 gene in prostate epithelial cells in vitro and in vivo. Cancer Res 67:4364–4372. Chakravarti D. et al. 1996. Role of CBP/P300 in nuclear receptor signalling. Nature 383:99–103. Chang H. C. et al. 2005. In vitro gene expression changes of androgen receptor coactivators after hormone deprivation in an androgen-dependent prostate cancer cell line. J Formos Med Assoc 104:652–658. Chang Y. M. et al. 2007. Nonreceptor tyrosine kinases in prostate cancer. Neoplasia 9:90–100. Chen H. et al. 1997. Nuclear receptor coactivator ACTR is a novel histone acetyltransferase and forms a multimeric activation complex with P/CAF and CBP/p300. Cell 90:569–580. Cheng H. et al. 2006. Short hairpin RNA knockdown of the androgen receptor attenuates ligandindependent activation and delays tumor progression. Cancer Res 66:10613–10620. Cheng S. et al. 2002. Inhibition of the dihydrotestosterone-activated androgen receptor by nuclear receptor corepressor. Mol Endocrinol 16:1492–1501. Chesire D. R. et al. 2002. In vitro evidence for complex modes of nuclear beta-catenin signaling during prostate growth and tumorigenesis. Oncogene 21:2679–2694.

Androgen Receptor Coregulators and Their Role in Prostate Cancer

365

Chiba H. et al. 1994. Two human homologues of Saccharomyces cerevisiae SWI2/SNF2 and Drosophila brahma are transcriptional coactivators cooperating with the estrogen receptor and the retinoic acid receptor. Nucleic Acids Res 22:1815–1820. Chmelar R. et al. 2007. Androgen receptor coregulators and their involvement in the development and progression of prostate cancer. Int J Cancer 120:719–733. Choudhry M. A. et al. 2006. The role of the general transcription factor IIF in androgen receptordependent transcription. Mol Endocrinol 20:2052–2061. Chung A. C. et al. 2007. Genetic ablation of the amplified-in-breast cancer 1 inhibits spontaneous prostate cancer progression in mice. Cancer Res 67:5965–5975. Cinar B. et al. 2007. Phosphoinositide 3-kinase-independent non-genomic signals transit from the androgen receptor to Akt1 in membrane raft microdomains. J Biol Chem 282:29584–29593. Comuzzi B. et al. 2004. The androgen receptor co-activator CBP is up-regulated following androgen withdrawal and is highly expressed in advanced prostate cancer. J Pathol 204:159–166. Cronauer M. V. et al. 2004. Inhibition of p53 function diminishes androgen receptor-mediated signaling in prostate cancer cell lines. Oncogene 23:3541–3549. Culig Z. and Bartsch G. 2006. Androgen axis in prostate cancer. J Cell Biochem 99:373–381. Culig Z. et al. 2003. Androgen receptors in prostate cancer. J Urol 170:1363–1369. Culig Z. et al. 2005. Mechanisms of endocrine therapy-responsive and -unresponsive prostate tumours. Endocr Relat Cancer 12:229–244. Dasso M. 2001. Running on Ran: nuclear transport and the mitotic spindle. Cell 104:321–324. Davies T. H. et al. 2002. A new first step in activation of steroid receptors: hormone-induced switching of FKBP51 and FKBP52 immunophilins. J Biol Chem 277:4597–4600. Debes J. D. et al. 2002. p300 mediates androgen-independent transactivation of the androgen receptor by interleukin 6. Cancer Res 62:5632–5636. Debes J. D. et al. 2003. p300 in prostate cancer proliferation and progression. Cancer Res 63:7638– 7640. Dedhar S. et al. 1994. Inhibition of nuclear hormone receptor activity by calreticulin. Nature 367:480–483. Dehm S. M. and Tindall D. J. 2007. Androgen receptor structural and functional elements: role and regulation in prostate cancer. Mol Endocrinol 21:2855–2863. Desai S. J. et al. 2006. Inappropriate activation of the androgen receptor by nonsteroids: involvement of the Src kinase pathway and its therapeutic implications. Cancer Res 66:10449–10459. Ding X. F. et al. 1998. Nuclear receptor-binding sites of coactivators glucocorticoid receptor interacting protein 1 (GRIP1) and steroid receptor coactivator 1 (SRC-1): multiple motifs with different binding specificities. Mol Endocrinol 12:302–313. Dittmar K. D. et al. 1998. The role of DnaJ-like proteins in glucocorticoid receptor.hsp90 heterocomplex assembly by the reconstituted hsp90.p60.hsp70 foldosome complex. J Biol Chem 273:7358–7366. Dong X. et al. 2007. Transcriptional activity of androgen receptor is modulated by two RNA splicing factors, PSF and p54nrb. Mol Cell Biol 27:4863–4875. Dotzlaw H. et al. 2002. The amino terminus of the human AR is target for corepressor action and antihormone agonism. Mol Endocrinol 16:661–673. Dressel U. et al. 1999. Alien, a highly conserved protein with characteristics of a corepressor for members of the nuclear hormone receptor superfamily. Mol Cell Biol 19:3383–3394. Drobnjak M. et al. 2000. Overexpression of cyclin D1 is associated with metastatic prostate cancer to bone. Clin Cancer Res 6:1891–1895. Dubbink H. J. et al. 2004. Distinct recognition modes of FXXLF and LXXLL motifs by the androgen receptor. Mol Endocrinol 18:2132–2150. Edwards J. and Bartlett J. M. 2005a. The androgen receptor and signal-transduction pathways in hormone-refractory prostate cancer. Part 1: Modifications to the androgen receptor. BJU Int 95:1320–1326.

366

L.A. Wafa et al.

Edwards J. and Bartlett J. M. 2005b. The androgen receptor and signal-transduction pathways in hormone-refractory prostate cancer. Part 2: Androgen-receptor cofactors and bypass pathways. BJU Int 95:1327–1335. Estebanez-Perpina E. et al. 2007. A surface on the androgen receptor that allosterically regulates coactivator binding. Proc Natl Acad Sci U S A 104:16074–16079 Fan C. Y. et al. 2005. The type I Hsp40 zinc finger-like region is required for Hsp70 to capture nonnative polypeptides from Ydj1. J Biol Chem 280:695–702. Fang Y. et al. 1996. Hsp90 regulates androgen receptor hormone binding affinity in vivo. J Biol Chem 271:28697–28702. Fletterick R. J. 2005. Molecular modelling of the androgen receptor axis: rational basis for androgen receptor intervention in androgen-independent prostate cancer. BJU Int 96 (Suppl 2):2–9. Ford O. H., 3rd et al. 2003. Androgen receptor gene amplification and protein expression in recurrent prostate cancer. J Urol 170:1817–1821. Freeman B. C. et al. 2000. The p23 molecular chaperones act at a late step in intracellular receptor action to differentially affect ligand efficacies. Genes Dev 14:422–434. Freeman M. R. et al. 2005. Membrane rafts as potential sites of nongenomic hormonal signaling in prostate cancer. Trends Endocrinol Metab 16:273–279. Froesch B. A. et al. 1998. BAG-1L protein enhances androgen receptor function. J Biol Chem 273:11660–11666. Fronsdal K. et al. 1998. CREB binding protein is a coactivator for the androgen receptor and mediates cross-talk with AP-1. J Biol Chem 273:31853–31859. Fu M. et al. 2000. p300 and p300/cAMP-response element-binding protein-associated factor acetylate the androgen receptor at sites governing hormone-dependent transactivation. J Biol Chem 275:20853–20860. Fu M. et al. 2003. Acetylation of androgen receptor enhances coactivator binding and promotes prostate cancer cell growth. Mol Cell Biol 23:8563–8575. Fujimoto N. et al. 2007. Prostate cancer cells increase androgen sensitivity by increase in nuclear androgen receptor and androgen receptor coactivators; a possible mechanism of hormoneresistance of prostate cancer cells. Cancer Invest 25:32–37. Gaddipati J. P. et al. 1994. Frequent detection of codon 877 mutation in the androgen receptor gene in advanced prostate cancers. Cancer Res 54:2861–2864. Gao W. and Dalton J. T. 2007. Expanding the therapeutic use of androgens via selective androgen receptor modulators (SARMs). Drug Discov Today 12:241–248. Gao N. et al. 2003. The role of hepatocyte nuclear factor-3 alpha (Forkhead Box A1) and androgen receptor in transcriptional regulation of prostatic genes. Mol Endocrinol 17:1484–1507. Gaughan L. et al. 2002. Tip60 and histone deacetylase 1 regulate androgen receptor activity through changes to the acetylation status of the receptor. J Biol Chem 277:25904–25913. Gehin M. et al. 2002. The function of TIF2/GRIP1 in mouse reproduction is distinct from those of SRC-1 and p/CIP. Mol Cell Biol 22:5923–5937. Ghosh P. M. et al. 2003. Akt in prostate cancer: possible role in androgen-independence. Curr Drug Metab 4:487–496. Gioeli D. 2005. Signal transduction in prostate cancer progression. Clin Sci (Lond) 108:293–308. Gnanapragasam V. J. et al. 2001. Expression of RAC 3, a steroid hormone receptor co-activator in prostate cancer. Br J Cancer 85:1928–1936. Gobinet J. et al. 2001. Characterization of the interaction between androgen receptor and a new transcriptional inhibitor, SHP. Biochemistry 40:15369–15377. Gobinet J. et al. 2005. SHP represses transcriptional activity via recruitment of histone deacetylases. Biochemistry 44:6312–6320. Gong J. et al. 2006. Activation of p300 histone acetyltransferase activity and acetylation of the androgen receptor by bombesin in prostate cancer cells. Oncogene 25:2011–2021.

Androgen Receptor Coregulators and Their Role in Prostate Cancer

367

Graff J. R. et al. 2000. Increased AKT activity contributes to prostate cancer progression by dramatically accelerating prostate tumor growth and diminishing p27Kip1 expression. J Biol Chem 275:24500–24505. Gregory C. W. et al. 2001. A mechanism for androgen receptor-mediated prostate cancer recurrence after androgen deprivation therapy. Cancer Res 61:4315–4319. Gregory C. W. et al. 2004. Epidermal growth factor increases coactivation of the androgen receptor in recurrent prostate cancer. J Biol Chem 279:7119–7130. Gregory C. W. et al. 2005. Heregulin-induced activation of HER2 and HER3 increases androgen receptor transactivation and CWR-R1 human recurrent prostate cancer cell growth. Clin Cancer Res 11:1704–1712. Guo Z. et al. 2006. Regulation of androgen receptor activity by tyrosine phosphorylation. Cancer Cell 10:309–319. Halkidou K. et al. 2003. Expression of Tip60, an androgen receptor coactivator, and its role in prostate cancer development. Oncogene 22:2466–2477. Halvorsen O. J. et al. 2003. Combined loss of PTEN and p27 expression is associated with tumor cell proliferation by Ki-67 and increased risk of recurrent disease in localized prostate cancer. Clin Cancer Res 9:1474–1479. Harashima K. et al. 2005. Heat shock protein 90 (Hsp90) chaperone complex inhibitor, radicicol, potentiated radiation-induced cell killing in a hormone-sensitive prostate cancer cell line through degradation of the androgen receptor. Int J Radiat Biol 81:63–76. He B. et al. 1999. Activation function 2 in the human androgen receptor ligand binding domain mediates interdomain communication with the NH(2)-terminal domain. J Biol Chem 274:37219–37225. He B. et al. 2001. Androgen-induced NH2- and COOH-terminal interaction inhibits p160 coactivator recruitment by activation function 2. J Biol Chem 276:42293–42301. He B. et al. 2002. The FXXLF motif mediates androgen receptor-specific interactions with coregulators. J Biol Chem 277:10226–10235. He B. et al. 2006. Probing the functional link between androgen receptor coactivator and ligandbinding sites in prostate cancer and androgen insensitivity. J Biol Chem 281:6648–6663. Heemers H. V. and Tindall D. J. 2005. Androgen receptor coregulatory proteins as potential therapeutic targets in the treatment of prostate cancer. Current Cancer Therapy Reviews 1:175– 186. Heemers H. V. and Tindall D. J. 2007. Androgen receptor (AR) coregulators: a diversity of functions converging on and regulating the AR transcriptional complex. Endocr Rev 28:778–808. Heery D. M. et al. 1997. A signature motif in transcriptional co-activators mediates binding to nuclear receptors. Nature 387:733–736. Heinlein C. A. and Chang C. 2002. Androgen receptor (AR) coregulators: an overview. Endocr Rev 23:175–200. Heinlein C. A. and Chang C. 2004. Androgen receptor in prostate cancer. Endocr Rev 25:276–308. Hermanson O. et al. 2002. N-CoR controls differentiation of neural stem cells into astrocytes. Nature 419:934–939. Hobisch A. et al. 1995. Distant metastases from prostatic carcinoma express androgen receptor protein. Cancer Res 55:3068–3072. Holter E. et al. 2002. Inhibition of androgen receptor (AR) function by the reproductive orphan nuclear receptor DAX-1. Mol Endocrinol 16:515–528. Hong H. et al. 2004. Aberrant expression of CARM1, a transcriptional coactivator of androgen receptor, in the development of prostate carcinoma and androgen-independent status. Cancer 101:83–89. Hsiao P. W. and Chang C. 1999. Isolation and characterization of ARA160 as the first androgen receptor N-terminal-associated coactivator in human prostate cells. J Biol Chem 274:22373–22379.

368

L.A. Wafa et al.

Hsiao P. W. et al. 1999. The linkage of Kennedy’s neuron disease to ARA24, the first identified androgen receptor polyglutamine region-associated coactivator. J Biol Chem 274:20229– 20234. Hu Y. C. et al. 2004. Functional domain and motif analyses of androgen receptor coregulator ARA70 and its differential expression in prostate cancer. J Biol Chem 279:33438–33446. Huang Z. Q. et al. 2003. A role for cofactor-cofactor and cofactor-histone interactions in targeting p300, SWI/SNF and mediator for transcription. Embo J 22:2146–2155. Ikonen T. et al. 1997. Interaction between the amino- and carboxyl-terminal regions of the rat androgen receptor modulates transcriptional activity and is influenced by nuclear receptor coactivators. J Biol Chem 272:29821–29828. Jemal A. et al. 2004. Cancer statistics, 2004. CA Cancer J Clin 54:8–29. Kahl P. et al. 2006. Androgen receptor coactivators lysine-specific histone demethylase 1 and four and a half LIM domain protein 2 predict risk of prostate cancer recurrence. Cancer Res 66:11341–11347. Kaltz-Wittmer C. et al. 2000. FISH analysis of gene aberrations (MYC, CCND1, ERBB2, RB, and AR) in advanced prostatic carcinomas before and after androgen deprivation therapy. Lab Invest 80:1455–1464. Kampa M. et al. 2002. The human prostate cancer cell line LNCaP bears functional membrane testosterone receptors that increase PSA secretion and modify actin cytoskeleton. Faseb J 16:1429–1431. Karam J. A. et al. 2007. Caveolin-1 overexpression is associated with aggressive prostate cancer recurrence. Prostate 67:614–622. Kent E. C. and Hussain M. H. 2003. The patient with hormone-refractory prostate cancer: determining who, when, and how to treat. Urology 62(Suppl 1):134–140. Kim J. et al. 2005. The role of protein kinase A pathway and cAMP responsive element-binding protein in androgen receptor-mediated transcription at the prostate-specific antigen locus. J Mol Endocrinol 34:107–118. Knudsen K. E. et al. 1999a. D-type cyclins complex with the androgen receptor and inhibit its transcriptional transactivation ability. Cancer Res 59:2297–2301. Knudsen K. E. et al. 1999b. The retinoblastoma tumor suppressor inhibits cellular proliferation through two distinct mechanisms: inhibition of cell cycle progression and induction of cell death. Oncogene 18:5239–5245. Kornberg R. D. 2007. The molecular basis of eukaryotic transcription. Proc Natl Acad Sci U S A 104:12955–12961. Kosano H. et al. 1998. The assembly of progesterone receptor-hsp90 complexes using purified proteins. J Biol Chem 273:32973–32979. Krajewska M. et al. 2006. Expression of BAG-1 protein correlates with aggressive behavior of prostate cancers. Prostate 66:801–810. Kraus S. et al. 2006. Receptor for activated C kinase 1 (RACK1) and Src regulate the tyrosine phosphorylation and function of the androgen receptor. Cancer Res 66:11047–11054. Krongrad A. et al. 1991. Androgen increases androgen receptor protein while decreasing receptor mRNA in LNCaP cells. Mol Cell Endocrinol 76:79–88. Kumar R. et al. 2005. The clinical relevance of steroid hormone receptor corepressors. Clin Cancer Res 11:2822–2831. LaFevre-Bernt M. A. and Ellerby L. M. 2003. Kennedy’s disease. Phosphorylation of the polyglutamine-expanded form of androgen receptor regulates its cleavage by caspase-3 and enhances cell death. J Biol Chem 278:34918–34924. Lambert J. R. and Nordeen S. K. 2003. CBP recruitment and histone acetylation in differential gene induction by glucocorticoids and progestins. Mol Endocrinol 17:1085–1094. Lange C. A. et al. 2007. Integration of rapid signaling events with steroid hormone receptor action in breast and prostate cancer. Annu Rev Physiol 69:171–199. Lee D. K. and Chang C. 2003. Molecular communication between androgen receptor and general transcription machinery. J Steroid Biochem Mol Biol 84:41–49.

Androgen Receptor Coregulators and Their Role in Prostate Cancer

369

Lee D. K. et al. 2000. From androgen receptor to the general transcription factor TFIIH. Identification of cdk activating kinase (CAK) as an androgen receptor NH(2)-terminal associated coactivator. J Biol Chem 275:9308–9313. Lee D. K. et al. 2003. The second largest subunit of RNA polymerase II interacts with and enhances transactivation of androgen receptor. Biochem Biophys Res Commun 302:162–169. Li J. et al. 2006. A role of the amino-terminal (N) and carboxyl-terminal (C) interaction in binding of androgen receptor to chromatin. Mol Endocrinol 20:776–785. Li J. et al. 2007. Structural and functional analysis of androgen receptor in chromatin. Mol Endocrinol [Epub ahead of print, doi:10.1210/me.2006–0221] Li L. et al. 2003. Caveolin-1 maintains activated Akt in prostate cancer cells through scaffolding domain binding site interactions with and inhibition of serine/threonine protein phosphatases PP1 and PP2A. Mol Cell Biol 23:9389–9404. Li P. et al. 2001. Antagonism between PTEN/MMAC1/TEP-1 and androgen receptor in growth and apoptosis of prostatic cancer cells. J Biol Chem 276:20444–20450. Liao G. et al. 2003. Regulation of androgen receptor activity by the nuclear receptor corepressor SMRT. J Biol Chem 278:5052–5061. Lin H. K. et al. 2001. Akt suppresses androgen-induced apoptosis by phosphorylating and inhibiting androgen receptor. Proc Natl Acad Sci U S A 98:7200–7205. Lin H. K. et al. 2003. Suppression versus induction of androgen receptor functions by the phosphatidylinositol 3-kinase/Akt pathway in prostate cancer LNCaP cells with different passage numbers. J Biol Chem 278:50902–50907. Lin H. K. et al. 2004. Regulation of androgen receptor signaling by PTEN (phosphatase and tensin homolog deleted on chromosome 10) tumor suppressor through distinct mechanisms in prostate cancer cells. Mol Endocrinol 18:2409–2423. Linja M. J. et al. 2001. Amplification and overexpression of androgen receptor gene in hormonerefractory prostate cancer. Cancer Res 61:3550–3555. Link K. A. et al. 2005. BAF57 governs androgen receptor action and androgen-dependent proliferation through SWI/SNF. Mol Cell Biol 25:2200–2215. Lonard D. M. et al. 2007. Nuclear receptor coregulators and human disease. Endocr Rev 28:575– 587. Loy C. J. et al. 2003. Filamin-A fragment localizes to the nucleus to regulate androgen receptor and coactivator functions. Proc Natl Acad Sci U S A 100:4562–4567. Lu M. L. et al. 2001. Caveolin-1 interacts with androgen receptor. A positive modulator of androgen receptor mediated transactivation. J Biol Chem 276:13442–13451. Lyng F. M. et al. 2000. Rapid androgen actions on calcium signaling in rat sertoli cells and two human prostatic cell lines: similar biphasic responses between 1 picomolar and 100 nanomolar concentrations. Biol Reprod 63:736–747. Ma A. H. et al. 2006. Male germ cell-associated kinase, a male-specific kinase regulated by androgen, is a coactivator of androgen receptor in prostate cancer cells. Cancer Res 66:8439–8447. Maki H. E. et al. 2006. Screening of genetic and expression alterations of SRC1 gene in prostate cancer. Prostate 66:1391–1398. Maki H. E. et al. 2007. Overexpression and gene amplification of BAG-1L in hormone-refractory prostate cancer. J Pathol 212:395–401. Marcelli M. et al. 2000. Androgen receptor mutations in prostate cancer. Cancer Res 60:944– 949. Marivoet S. et al. 1992. Interaction of the 90-kDa heat shock protein with native and in vitro translated androgen receptor and receptor fragments. Mol Cell Endocrinol 88:165–174. Mark M. et al. 2004. Partially redundant functions of SRC-1 and TIF2 in postnatal survival and male reproduction. Proc Natl Acad Sci U S A 101:4453–4458. Marshall T. W. et al. 2003. Differential requirement of SWI/SNF for androgen receptor activity. J Biol Chem 278:30605–30613.

370

L.A. Wafa et al.

Martel C. L. et al. 2003. Current strategies in the management of hormone refractory prostate cancer. Cancer Treat Rev 29:171–187. Mattie M. D. et al. 2006. Optimized high-throughput microRNA expression profiling provides novel biomarker assessment of clinical prostate and breast cancer biopsies. Mol Cancer 5:24. Mayeur G. L. et al. 2005. Ku is a novel transcriptional recycling coactivator of the androgen receptor in prostate cancer cells. J Biol Chem 280:10827–10833. McEwan I. J. and Gustafsson J. 1997. Interaction of the human androgen receptor transactivation function with the general transcription factor TFIIF. Proc Natl Acad Sci U S A 94:8485–8490. McKenna N. J. et al. 1999. Nuclear receptor coregulators: cellular and molecular biology. Endocr Rev 20:321–344. McMenamin M. E. et al. 1999. Loss of PTEN expression in paraffin-embedded primary prostate cancer correlates with high Gleason score and advanced stage. Cancer Res 59:4291–4296. Migliaccio A. et al. 2007. Inhibition of the SH3 domain-mediated binding of Src to the androgen receptor and its effect on tumor growth. Oncogene 26:6619–6629. Mills I. G. et al. 2005. Huntingtin interacting protein 1 modulates the transcriptional activity of nuclear hormone receptors. J Cell Biol 170:191–200. Miyamoto H. et al. 1998. Promotion of agonist activity of antiandrogens by the androgen receptor coactivator, ARA70, in human prostate cancer DU145 cells. Proc Natl Acad Sci U S A 95:7379–7384. Miyamoto H. et al. 2004. Molecular basis for the antiandrogen withdrawal syndrome. J Cell Biochem 91:3–12. Moehren U. et al. 2007. Alien interacts with the human androgen receptor and inhibits prostate cancer cell growth. Mol Endocrinol 21:1039–1048. Mohler J. L. et al. 2004. The androgen axis in recurrent prostate cancer. Clin Cancer Res 10:440– 448. Moilanen A. M. et al. 1998. Activation of androgen receptor function by a novel nuclear protein kinase. Mol Biol Cell 9:2527–2543. Mostaghel E. A. et al. 2007a. The basic biochemistry and molecular events of hormone therapy. Curr Urol Rep 8:224–232. Mostaghel E. A. et al. 2007b. Intraprostatic androgens and androgen-regulated gene expression persist after testosterone suppression: therapeutic implications for castration-resistant prostate cancer. Cancer Res 67:5033–5041. Mulholland D. J. et al. 2002. The androgen receptor can promote beta-catenin nuclear translocation independently of adenomatous polyposis coli. J Biol Chem 277:17933–17943. Nair S. S. et al. 2007. Proline-, glutamic acid-, and leucine-rich protein-1/modulator of nongenomic activity of estrogen receptor enhances androgen receptor functions through LIM-only coactivator, four-and-a-half LIM-only protein 2. Mol Endocrinol 21:613–624. Neckers L. and Ivy S. P. 2003. Heat shock protein 90. Curr Opin Oncol 15:419–424. Needham M. et al. 2000. Differential interaction of steroid hormone receptors with LXXLL motifs in SRC-1a depends on residues flanking the motif. J Steroid Biochem Mol Biol 72:35–46. Newmark J. R. et al. 1992. Androgen receptor gene mutations in human prostate cancer. Proc Natl Acad Sci U S A 89:6319–6323. Nishimura K. et al. 2003. Modulation of androgen receptor transactivation by gelsolin: a newly identified androgen receptor coregulator. Cancer Res 63:4888–4894. Nwachukwu J. C. et al. 2007. Transcriptional regulation of the androgen receptor cofactor androgen receptor trapped clone-27. Mol Endocrinol 21:2864–2876. Ogryzko V. V. et al. 1996. The transcriptional coactivators p300 and CBP are histone acetyltransferases. Cell 87:953–959. Onate S. A. et al. 1995. Sequence and characterization of a coactivator for the steroid hormone receptor superfamily. Science 270:1354–1357. Owens-Grillo J. K. et al. 1996. A model of protein targeting mediated by immunophilins and other proteins that bind to hsp90 via tetratricopeptide repeat domains. J Biol Chem 271:13468– 13475.

Androgen Receptor Coregulators and Their Role in Prostate Cancer

371

Ozanne D. M. et al. 2000. Androgen receptor nuclear translocation is facilitated by the f-actin cross-linking protein filamin. Mol Endocrinol 14:1618–1626. Ozers M. S. et al. 2007. The androgen receptor T877A mutant recruits LXXLL and FXXLF peptides differently than wild-type androgen receptor in a time-resolved fluorescence resonance energy transfer assay. Biochemistry 46:683–695. Paez J. and Sellers W. R. 2003. PI3K/PTEN/AKT pathway. A critical mediator of oncogenic signaling. Cancer Treat Res 115:145–167. Papakonstanti E. A. et al. 2003. A rapid, nongenomic, signaling pathway regulates the actin reorganization induced by activation of membrane testosterone receptors. Mol Endocrinol 17:870–881. Pedram A. et al. 2007. A conserved mechanism for steroid receptor translocation to the plasma membrane. J Biol Chem 282:22278–22288. Petre-Draviam C. E. et al. 2005. A central domain of cyclin D1 mediates nuclear receptor corepressor activity. Oncogene 24:431–444. Petrylak D. P. 2005. Future directions in the treatment of androgen-independent prostate cancer. Urology 65:8–12. Porkka K. P. et al. 2007. MicroRNA expression profiling in prostate cancer. Cancer Res 67:6130– 6135. Pratt W. B. and Toft D. O. 1997. Steroid receptor interactions with heat shock protein and immunophilin chaperones. Endocr Rev 18:306–360. Prescott J. and Coetzee G. A. 2006. Molecular chaperones throughout the life cycle of the androgen receptor. Cancer Lett 231:12–19. Rao D. S. et al. 2002a. Huntingtin-interacting protein 1 is overexpressed in prostate and colon cancer and is critical for cellular survival. J Clin Invest 110:351–360. Rao M. A. et al. 2002b. RanBPM, a nuclear protein that interacts with and regulates transcriptional activity of androgen receptor and glucocorticoid receptor. J Biol Chem 277:48020–48027. Rassow J. et al. 1995. Partner proteins determine multiple functions of Hsp70. Trends Cell Biol 5:207–212. Ray M. R. et al. 2006. Cyclin G-associated kinase: a novel androgen receptor-interacting transcriptional coactivator that is overexpressed in hormone refractory prostate cancer. Int J Cancer 118:1108–1119. Ratajczak T. 2001. Protein coregulators that mediate estrogen receptor function. Reprod Fertil Dev 13:221–229. Rennie P. S. and Nelson C. C. 1998. Epigenetic mechanisms for progression of prostate cancer. Cancer Metastasis Rev 17:401–409. Reutens A. T. et al. 2001. Cyclin D1 binds the androgen receptor and regulates hormonedependent signaling in a p300/CBP-associated factor (P/CAF)-dependent manner. Mol Endocrinol 15:797–811. Rhodes D. 1997. Chromatin structure. The nucleosome core all wrapped up. Nature 389:231–233. Rocchi P. et al. 2004. Heat shock protein 27 increases after androgen ablation and plays a cytoprotective role in hormone-refractory prostate cancer. Cancer Res 64:6595–6602. Rochette-Egly C. 2003. Nuclear receptors: integration of multiple signalling pathways through phosphorylation. Cell Signal 15:355–366. Roeder R. G. 1996. The role of general initiation factors in transcription by RNA polymerase II. Trends Biochem Sci 21:327–335. Rosenfeld M. G. and Glass C. K. 2001. Coregulator codes of transcriptional regulation by nuclear receptors. J Biol Chem 276:36865–36868. Rosenfeld M. G. et al. 2006. Sensors and signals: a coactivator/corepressor/epigenetic code for integrating signal-dependent programs of transcriptional response. Genes Dev 20:1405–1428. Ross J. S. 2007. The androgen receptor in prostate cancer: therapy target in search of an integrated diagnostic test. Adv Anat Pathol 14:353–357. Sadar M. D. et al. 1999. Prostate cancer: molecular biology of early progression to androgen independence. Endocr Relat Cancer 6:487–502.

372

L.A. Wafa et al.

Shah R. B. et al. 2004. Androgen-independent prostate cancer is a heterogeneous group of diseases: lessons from a rapid autopsy program. Cancer Res 64:9209–9216. Sharma M. et al. 2000. Androgen receptor interacts with a novel MYST protein, HBO1. J Biol Chem 275:35200–35208. Shatkina L. et al. 2003. The cochaperone Bag-1L enhances androgen receptor action via interaction with the NH2-terminal region of the receptor. Mol Cell Biol 23:7189–7197. Shenk J. L. et al. 2001. p53 represses androgen-induced transactivation of prostate-specific antigen by disrupting hAR amino- to carboxyl-terminal interaction. J Biol Chem 276:38472–38479. Shi X. B. et al. 2002. Functional analysis of 44 mutant androgen receptors from human prostate cancer. Cancer Res 62:1496–1502. Shi X. B. et al. 2004. A modified yeast assay used on archival samples of localized prostate cancer tissue improves the detection of p53 abnormalities and increases their predictive value. BJU Int 94:996–1002. Singh P. et al. 2006. Combinatorial androgen receptor targeted therapy for prostate cancer. Endocr Relat Cancer 13:653–666. Smith C. L. and O’Malley B. W. 2004. Coregulator function: a key to understanding tissue specificity of selective receptor modulators. Endocr Rev 25:45–71. Smith D. F. and Toft D. O. 1993. Steroid receptors and their associated proteins. Mol Endocrinol 7:4–11. Spencer T. E. et al. 1997. Steroid receptor coactivator-1 is a histone acetyltransferase. Nature 389:194–198. Stossel T. P. et al. 1985. Nonmuscle actin-binding proteins. Annu Rev Cell Biol 1:353–402. Sun J. et al. 2007. Cofactor of BRCA1 modulates androgen-dependent transcription and alternative splicing. J Steroid Biochem Mol Biol 107:131–139. Takeshita A. et al. 1996. Molecular cloning and properties of a full-length putative thyroid hormone receptor coactivator. Endocrinology 137:3594–3597. Taneja S. S. et al. 2004. ART-27, an androgen receptor coactivator regulated in prostate development and cancer. J Biol Chem 279:13944–13952. Taneja S. S. et al. 2005. Cell-specific regulation of androgen receptor phosphorylation in vivo. J Biol Chem 280:40916–40924. Tannock I. F. et al. 2004. Docetaxel plus prednisone or mitoxantrone plus prednisone for advanced prostate cancer. N Engl J Med 351:1502–1512. Taplin M. E. and Balk S. P. 2004. Androgen receptor: a key molecule in the progression of prostate cancer to hormone independence. J Cell Biochem 91:483–490. Taplin M. E. et al. 2003. Androgen receptor mutations in androgen-independent prostate cancer: Cancer and Leukemia Group B Study 9663. J Clin Oncol 21:2673–2678. Thomas M. et al. 2006. Pharmacologic and genetic inhibition of hsp90-dependent trafficking reduces aggregation and promotes degradation of the expanded glutamine androgen receptor without stress protein induction. Hum Mol Genet 15:1876–1883. Tilley W. D. et al. 1994. Detection of discrete androgen receptor epitopes in prostate cancer by immunostaining: measurement by color video image analysis. Cancer Res 54:4096–4102. Tillman J. E. et al. 2007. DJ-1 binds androgen receptor directly and mediates its activity in hormonally treated prostate cancer cells. Cancer Res 67:4630–4637. Ting H. J. et al. 2002. Supervillin associates with androgen receptor and modulates its transcriptional activity. Proc Natl Acad Sci U S A 99:661–666. Truica C. I. et al. 2000. Beta-catenin affects androgen receptor transcriptional activity and ligand specificity. Cancer Res 60:4709–4713. Ueda T. et al. 2002. Ligand-independent activation of the androgen receptor by interleukin-6 and the role of steroid receptor coactivator-1 in prostate cancer cells. J Biol Chem 277:38087– 38094. Unni E. et al. 2004. Changes in androgen receptor nongenotropic signaling correlate with transition of LNCaP cells to androgen independence. Cancer Res 64:7156–7168.

Androgen Receptor Coregulators and Their Role in Prostate Cancer

373

Vadlamudi R. K. and Kumar R. 2007. Functional and biological properties of the nuclear receptor coregulator PELP1/MNAR. Nucl Recept Signal 5:e004. Veldscholte J. et al. 1992. Hormone-induced dissociation of the androgen receptor-heat-shock protein complex: use of a new monoclonal antibody to distinguish transformed from nontransformed receptors. Biochemistry 31:7422–7430. Visakorpi T. et al. 1995. In vivo amplification of the androgen receptor gene and progression of human prostate cancer. Nat Genet 9:401–406. Volinia S. et al. 2006. A microRNA expression signature of human solid tumors defines cancer gene targets. Proc Natl Acad Sci U S A 103:2257–2261. Wafa L. A. et al. 2003. Isolation and identification of L-dopa decarboxylase as a protein that binds to and enhances transcriptional activity of the androgen receptor using the repressed transactivator yeast two-hybrid system. Biochem J 375:373–383. Wafa L. A. et al. 2007. Comprehensive expression analysis of L-dopa decarboxylase and established neuroendocrine markers in neoadjuvant hormone-treated versus varying Gleason grade prostate tumors. Hum Pathol 38:161–170. Wang L. et al. 2004. Suppression of androgen receptor-mediated transactivation and cell growth by the glycogen synthase kinase 3 beta in prostate cells. J Biol Chem 279:32444–32452. Wang L. et al. 2005. Androgen receptor corepressors: an overview. Prostate 63:117–130. Wang Q. et al. 2001a. Ligand- and coactivator-mediated transactivation function (AF2) of the androgen receptor ligand-binding domain is inhibited by the cognate hinge region. J Biol Chem 276:7493–7499. Wang S. et al. 2003. Prostate-specific deletion of the murine Pten tumor suppressor gene leads to metastatic prostate cancer. Cancer Cell 4:209–221. Wang X. et al. 2001b. Identification and characterization of a novel androgen receptor coregulator ARA267-alpha in prostate cancer cells. J Biol Chem 276:40417–40423. Weigel N. L. and Moore N. L. 2007. Kinases and protein phosphorylation as regulators of steroid hormone action. Nucl Recept Signal 5:e005. Wen Y. et al. 2000. HER-2/neu promotes androgen-independent survival and growth of prostate cancer cells through the Akt pathway. Cancer Res 60:6841–6845. Xin L. et al. 2006. Progression of prostate cancer by synergy of AKT with genotropic and nongenotropic actions of the androgen receptor. Proc Natl Acad Sci U S A 103:7789–7794. Xu J. et al. 1998. Partial hormone resistance in mice with disruption of the steroid receptor coactivator-1 (SRC-1) gene. Science 279:1922–1925. Xu L. et al. 1999. Coactivator and corepressor complexes in nuclear receptor function. Curr Opin Genet Dev 9:140–147. Yao T. P. et al. 1998. Gene dosage-dependent embryonic development and proliferation defects in mice lacking the transcriptional integrator p300. Cell 93:361–372. Yan J. et al. 2006. Steroid receptor coactivator-3 and activator protein-1 coordinately regulate the transcription of components of the insulin-like growth factor/AKT signaling pathway. Cancer Res 66:11039–11046. Yeh S. and Chang C. 1996. Cloning and characterization of a specific coactivator, ARA70, for the androgen receptor in human prostate cells. Proc Natl Acad Sci U S A 93:5517–5521. Yeh S. et al. 1998. Retinoblastoma, a tumor suppressor, is a coactivator for the androgen receptor in human prostate cancer DU145 cells. Biochem Biophys Res Commun 248:361–367. Yeh S. et al. 1999. From HER2/Neu signal cascade to androgen receptor and its coactivators: a novel pathway by induction of androgen target genes through MAP kinase in prostate cancer cells. Proc Natl Acad Sci U S A 96:5458–5463. Yeh S. et al. 2000. Increase of androgen-induced cell death and androgen receptor transactivation by BRCA1 in prostate cancer cells. Proc Natl Acad Sci U S A 97:11256–11261. Yong W. et al. 2007. Essential role for co-chaperone Fkbp52 but not Fkbp51 in androgen receptormediated signaling and physiology. J Biol Chem 282:5026–5036.

374

L.A. Wafa et al.

Yoon H. G. and Wong J. 2006. The corepressors silencing mediator of retinoid and thyroid hormone receptor and nuclear receptor corepressor are involved in agonist- and antagonistregulated transcription by androgen receptor. Mol Endocrinol 20:1048–1060. Zhang F. P. et al. 2007. An adenosine triphosphatase of the sucrose nonfermenting 2 family, androgen receptor-interacting protein 4, is essential for mouse embryonic development and cell proliferation. Mol Endocrinol 21:1430–1442. Zhou H. J. et al. 2005. SRC-3 is required for prostate cancer cell proliferation and survival. Cancer Res 65:7976–7983. Zhou J. et al. 2006. Inhibition of mitogen-elicited signal transduction and growth in prostate cancer with a small peptide derived from the functional domain of DOC-2/DAB2 delivered by a unique vehicle. Cancer Res 66:8954–8958. Zhoul J. et al. 2005. The role of DOC-2/DAB2 in modulating androgen receptor-mediated cell growth via the nongenomic c-Src-mediated pathway in normal prostatic epithelium and cancer. Cancer Res 65:9906–9913. Zou J. X. et al. 2006. ACTR/AIB1/SRC-3 and androgen receptor control prostate cancer cell proliferation and tumor growth through direct control of cell cycle genes. Prostate 66:1474–1486. Zoubeidi A. et al. 2007. Cooperative interactions between androgen receptor (AR) and heat-shock protein 27 facilitate AR transcriptional activity. Cancer Res 67:10455–10465.

Interaction of the Androgen Receptor Ligand-Binding Domain with the N-Terminal Domain and with Coactivators Jan Trapman

Abstract The ligand-binding domain of the androgen receptor not only binds ligands, but also contains a ligand-induced protein interaction surface, the cofactor-binding groove. The groove interacts with short amphipatic a-helices in cofactors composed of an FxxLF motif, or with LxxLL motifs at a lower affinity. Moreover, the cofactor-binding groove interacts with an FxxLF motif in the N-terminal domain of the androgen receptor. The groove is able to adapt its shape in complexes with interacting peptides. In the peptide motifs, an F at +1 seems essential for high-affinity binding. L+4 can be replaced by several other hydrophobic amino acid residues without losing activity. Although F at +5 has the highest activity, it can be substituted by tryptophane or tyrosine. Studies of the spatial and temporal distribution of the androgen receptor in the living cell indicates consecutive protein interactions, including intramolecular and intermolecular androgen receptor domain interactions and cofactor binding, depending on the cellular localization.

1 Introduction The structure of the ligand-binding domains (LBD) of essentially all nuclear receptors (NR) is composed of twelve a-helices, folded in a three-layered antiparallel sandwich conformation (Li et al. 2003). Ligands bind to a buried ligandbinding pocket formed by the cooperative action of specific amino acid residues scattered over many helices. The LBD is not only indirectly involved in transcription regulation by ligand binding, but also more directly, via the ligand-induced

J. Trapman Department of Pathology, Josephine Nefkens Institute, Erasmus Medical Center, PO Box 2040, 3000 CA Rotterdam, The Netherlands, E-mail: [email protected]

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4_16, # Springer Science + Business Media, LLC 2009

375

376

J. Trapman

activation function 2 (AF-2). Upon binding of an agonistic ligand, the position of helix 12 (H12) in the LBD is shifted in such a way that it directly interacts with the ligand. The agonist-induced repositioning of H12 not only completes the formation of the ligand-binding pocket, but also generates a hydrophobic protein interaction interface at the LBD surface, the cofactor-binding groove (Greschik and Moras 2003). The agonist-induced cofactor-binding groove facilitates the binding of many NR coactivating proteins. Antagonists induce an unfavorable conformation of the groove, which prohibits binding of coactivators. Corepressors appear to bind in the same cofactor groove or an overlapping binding site. Many NR coactivators bind to the LBD via LxxLL motifs, the so-called NR boxes, where L represents a leucine and x any amino acid (McKenna and O’Malley 2002; Rosenfeld et al. 2006). These peptide motifs form small amphipatic ahelices. Amino acid residues flanking the leucines modulate the specificity and affinity of LxxLL motifs. Cofactors might contain one or more LxxLL motifs. Best studied in this regard are the members of the p160 (SRC) family of cofactors, which possess three or four NR boxes. One p160 cofactor can bind simultaneously to a NR homodimer or heterodimer. Although the LBD of the androgen receptor (AR) has high sequence homology with the LBDs of other NRs and folds in a similar manner, the affinity for LxxLL motifs is relatively low. Instead, the AR favors FxxLF-like motifs (F is phenylalanine) for high-affinity interactions. The unique properties of the AR cofactorbinding groove induce specificity to interacting proteins (Fig. 1). Interestingly, it facilitates also the interaction between an FxxLF motif in the AR N-terminal domain (NTD) with the LBD, which plays a role in intramolecular and intermolecular interaction between the LBD and NTD (AR N/C interaction). AF-2 function of AR is weaker than that of other NRs. In this chapter the characteristics of the interaction of FxxLF- and LxxLL-like motifs with the AR LBD are described. Moreover, the in vivo relevance of interaction of AR LBD with the NTD and with coactivators is discussed.

Cofactors

AR NTD

DBD

LBD

Fig. 1 Schematic presentation of protein–protein interactions of the AR LBD. The AR LBD can interact with cofactors and with the NTD via FxxLF- and LxxLL-like amphipatic a-helix peptide motifs. NTD N-terminal domain, DBD DNA-binding domain, LBD ligand-binding domain

Interaction of the Androgen Receptor Ligand-Binding Domain

377

2 Peptides Interacting with the Cofactor-Binding Groove of the Androgen Receptor Yeast two-hybrid protein interaction assays have identified many protein fragments that can interact with AR in a ligand-dependent fashion (Heinlein and Chang 2002; Heemers and Tindall 2007). For most of these proteins the mode of interaction and the functional relevance of the interaction remains unknown. For five interacting proteins, ARA54, ARA70, Rad9, Gelsolin, and PAK6, it has been shown that their high, ligand-induced affinity for the AR LBD is mediated by a specific peptide motif (He et al. 2002; Hu et al. 2004; Wang et al. 2004; van de Wijngaart et al. 2006). None of the interacting peptides is composed of LxxLL motifs. Instead, they are composed of short amphipatic a-helices with phenylalanine residues at positions +1 and +5. Yeast protein–protein interaction assays showed not only interaction of AR LBD with other proteins, but also binding between the AR LBD and its NTD (Doesburg et al. 1997). This interaction is mediated by an FxxLF motif in the NTD (He et al. 2000; Steketee et al. 2002). The primary structures of the FxxLF motifs in interacting proteins and in AR NTD are summarized in Table 1. ARA54, ARA70, Rad9, and AR NTD possess an FxxLF motif that interacts with AR LBD (He et al. 2000, 2002; Steketee et al. 2002; Hu et al. 2004; Wang et al. 2004). Recently, the interacting motif of Gelsolin was established as FxxFF, and that of PAK6 as FxxMF (van de Wijngaart et al. 2006). Phage display has been used to select randomly for interacting peptide sequences with high affinity for the AR LBD (Hsu et al. 2003; Hur et al. 2004; Chang et al. 2005). Large libraries have been screened for peptides that interacted with AR LBD in the presence but not in the absence of ligand, and the interacting motifs were characterized. Peptides that interact with high affinity are mainly composed of FxxLF, FxxLY, or FxxLW motifs. Peptides showing high affinity with a different amino acid residue than phenylalanine on position +1 are hardly detected. No highaffinity peptides with LxxLL motifs were reported. It is unclear why in many cases at +4 a leucine was found, considering the more flexible position of this amino acid residue (see later).

Table 1 Composition of peptide motifs in proteins interacting with AR LBD +4 +1+5 AR NTD KTYRGAFQNLFQSVRE ARA54 DPGSPCFNRLFYAVDV ARA70 RETSEKFKLLFQSYNV RAD9 TPPPKKFRSLFFGSIL Gelsolin GGETPLFKQFFKNWRD PAK6 SLKRRLFRSMFLSTAA

378

J. Trapman

3 The Structure of the Cofactor-Binding Groove of the Androgen Receptor Computer modeling, comparing the primary structure of AR and ER LBD, predicted that FxxLF motifs have a similar interaction with the cofactor-binding groove of AR as LxxLL motifs have with the cofactor groove of the estrogen receptor (ER) (Dubbink et al. 2004) (Fig. 2). However, the cofactor-binding groove in AR is much deeper than that of ER and many other NRs. This deep groove facilitates docking of the bulky F residues in the groove. It also explains the weaker binding of many LxxLL motifs to the groove, because important hydrophobic interactions between the peptide motif and the amino acid residues lining the groove in the LBD cannot take place. Like other NRs the AR LBD contains two highly conserved amino acid residues, K720 in H3 and E897 in H12, which form a ‘‘charge clamp’’ that stabilizes the interaction with the a-helical structure of peptide motifs. Crystallographic studies of ligand-activated AR LBD complexed with peptide motifs provided detailed information about the structure of the coactivator-binding groove in the AR LBD and of the interactions with peptide motifs (He et al. 2004; Hur et al. 2004; Estebanez-Perpina et al. 2005). Not unexpected, the L-shaped AR cofactor-binding groove is formed by amino acid residues in helices H3, H4, H5, and H12. F+1 in the peptide motif binds to a hydrophobic part of the groove that is formed by L712, V716, M734, I737, Q738, M894, and I898. The F+5 binding site is formed by V712, K720, F725, V730, Q733, M734, and I737. L+4 interacts with V713, V716, and M894.

K720 F+5 V716

F+1

L+4 V713

E897

M894

Fig. 2 Model of the cofactor-binding groove complexed with an FxxLF peptide. In the groove charge clamp amino acid residues, and amino acid residues interacting with L+4 are indicated. Note that F+1 and F+5 of the peptide are buried in the groove and that L+4 lies on an edge of the groove (See Color Insert)

Interaction of the Androgen Receptor Ligand-Binding Domain

379

The coactivator groove is not a rigid structure, but it can adapt its form slightly upon peptide binding. Side chains of amino acid residues that line the coactivatorbinding groove rearrange upon binding of the peptide motif. The largest conformational changes were observed for K720, M734, M894, and E897. As expected, in FxxLF peptides F+1 and F+5 are directed to the AR LBD surface, whereas the amino acid residues at positions +2 and +3 are exposed to the solvent. L at +4 binds to a shallow hydrophobic depression and is partly solvent exposed (see Fig. 2). From the structure of the AR LBD complexed with an FxxLF peptide motif it can be predicted that substitutions of F+1 or F+5 are restricted, whereas position +4 in the peptide is more flexible. Crystal structures indicate that all interacting peptides bind in slightly different conformations to the coactivator groove. For some peptides it has been established clearly that flanking amino acid residues contribute to peptide binding. Differential binding was also deduced from functional studies. Mutations were introduced in the charge clamp residues K720 and E897. K720A and E897A substitutions both affect FxxLF interaction, but E897 is less important for interaction with some LxxLL motifs, due to the shifted position in the coactivator-binding groove (Dubbink et al. 2004; He et al. 2004; Hur et al. 2004; Estebanez-Perpina et al. 2005).

4 Mutational Analysis of Peptides Binding to the Coactivator Groove To study the role of individual amino acid residues the peptide motifs have been mutated extensively. First, mutational analysis of peptide motifs by alanine scanning of the AR FQNLF motif showed that not only the phenylalanine and leucine residues are essential for binding to AR LBD, but also that flanking amino acids modulate this interaction (He et al. 2000; Steketee et al. 2002). Similar findings were observed by alanine scanning of the FxxFF motif in Gelsolin and the FxxMF motif in PAK6 (van de Wijngaart et al. 2006). If F+1 and F+5 in the AR NTD FxxLF motif, FQNLF, were randomly mutated, none of the novel peptide motifs displayed a stronger interaction than the original FxxLF peptide (Dubbink et al. 2004). However, substitution of F+5 by M or W retained part of the activity. Substitution of F+1 by other amino acids resulted in inactive peptides, indicating that this amino acid residue is very important for highaffinity AR binding. Next, L+4 in the AR FQNLF motif was substituted by a complete panel of amino acids (van de Wijngaart et al. 2006). FQNFF and FQNMF had the same activity as the original peptide; all others interacted weakly or did not interact at all. This observation is in complete agreement with the structural data. Finally, the importance of Q+2 and N+3 in the AR FQNLF motif has been investigated by random mutagenesis (Dubbink et al. 2004). Many, but not all, amino acids can substitute Q and N without loss of activity. In general, an E at

380

J. Trapman

+2 and a charged or polar amino acid residue at +3, R,K, or Q, are preferred, probably because the stability of the amphipatic a-helical peptides is increased (Dubbink et al. 2006). Less is known about the preferred properties of amino acids flanking the FxxLF motif. Based on comparison of the composition of interacting peptides and limited mutation analysis, positively charged amino acids flanking F+1 seem preferred for stronger interaction (He and Wilson 2003). The optimal properties of amino acids flanking F+5 are not yet very clear. Complementary data came from F/L swapping experiments of FxxLF and LxxLL-based peptide motifs (Dubbink et al. 2006). Peptide interaction with AR LBD was systematically studied by substituting the phenylalanine residues by leucines in three FxxLF peptides, AR NTD, ARA54, and ARA70. In all three motifs, substitutions of F+1 and F+5 by leucine residues completely abrogated interaction with the AR LBD, clearly showing that in these motifs the bulky F+1 and F+5 are essential for interaction. Also, leucine residues at +1 and +5 were substituted with phenylalanine residues in nine LxxLL motifs, NR boxes of SRC1 and TIF2(SRC2) and the LxxLL peptides, D11 and D30 peptides (Dubbink et al. 2004, 2006). D11 and D30 were originally identified by phage display as highaffinity ER-binding peptides (Chang et al. 1999). The original LxxLL peptides showed variable interaction with AR LBD. Differential effects were observed for the L/F swaps. Binding of already interacting LxxLL motifs was unchanged or increased upon L to F substitutions. Interestingly, certain noninteracting LxxLL motifs turned into strong interacting peptides (SRC-1 NB-I and -IV); however, other inactive peptides remained inactive by L/F swaps. Mutational analysis of the AR LBD confirmed the essential role of K720 in interaction with FxxLF and LxxLL motifs. K717 and R726 modified this interaction in a peptide-specific manner. E897A substitution not only had less effect on LxxLL peptides, but also on some FxxLF peptides, extending previous observations that many peptides show different conformations if docked in the coactivator groove and that sequences flanking the core motif determine the specific mode of interactions. FxxLF peptides have also been assessed for interaction with LBDs of other NRs (Dubbink et al. 2006). Many NRs seem unable to interact with FxxLF peptides. Most information is available for the progesterone receptor (PR). L/F swapping indicates that PR LBD can interact with both LxxLL and FxxLF peptides; however, LxxLL interaction is clearly preferred above FxxLF. The obvious question is why the AR contains such a unique cofactor-binding site and why it contains an FxxLF motif in the NTD. Such questions remain to be answered. It has been shown that peptide binding can delay ligand dissociation; however, the effect is limited (He et al. 2000; Dubbink et al. 2004). It can be speculated that the combination FxxLF in AR NTD and a specific cofactor-binding groove in AR LBD protects the AR from unfavorable protein interactions by N/C interaction. Subsequently, the FxxLF motif in AR NTD and the cofactor-binding groove might be independently important for AR activity and specificity during transcription activation.

Interaction of the Androgen Receptor Ligand-Binding Domain

381

Other questions to be addressed are whether different ligands and whether naturally occurring mutations in AR LBD can affect peptide specificity. It has been reported that the hotspot T877A mutant in prostate cancer recruits peptides slightly different from wild-type AR (Ozers et al. 2007). Similarly, it has been shown that the AR H874Y mutation in prostate cancer improves the activity of testosterone (Askew et al. 2007). Recently, ligand specificity of AR LBD interaction with peptide motifs has been studied more extensively (Brooke et al. 2008). It was shown for AR T877A and other mutants that the partial antagonist cyproterone acetate preferentially stimulates interaction with LxxLL motifs, whereas the complete antagonist hydroxyflutamide stimulates interaction with an FxxFF or FxxLF peptide.

5 In Vivo Relevance of the Coactivator Groove of the Androgen Receptor An important question to be addressed concerns the relevance of interaction of the AR LBD with cofactors in the living cell. Which cofactors interact directly with the cofactor-binding groove under physiological conditions? Obviously, there can be competition between interaction of AR LBD with cofactors and with AR NTD. Although many cofactors with FxxLF-like motifs bind to the AR LBD under overexpression conditions, evidence for AR interaction under physiological conditions is limited. Many gene activation assays have been done with transiently transfected reporter gene constructs, lacking a proper chromatin structure. The establishment of in vivo relevant AR LBD interaction with cofactors binding to the cofactor-binding groove is complicated, because protein interactions might occur simultaneously between many proteins in large stabilized multicomponent protein complexes. In addition, one cofactor might interact with more than one AR protein interaction domain. Moreover, the cellular distribution is an essential element in facilitating protein–protein interactions. By far the most detailed information for in vivo interactions of AR LBD with FxxLF-like motifs is available for the FxxLF motif in the NTD (Fig. 3). There is ample evidence that the AR N/C interaction indeed takes place in living cells as shown by fluorescence resonance energy transfer (FRET) assays with doubletagged AR (YFP-AR-CFP) expressed at physiological levels (Schaufele et al. 2005; van Royen et al. 2007). Mutation of the FxxLF motif in AR NTD to AxxAA diminishes the interaction. AR N/C interaction is induced very rapidly following addition of ligand to cell cultures. Studies of the spatial and temporal organization of AR N/C interaction show that AR N/C interaction is intramolecular in the cytoplasm. Interestingly, subsequent AR transport to the nucleus causes a shift from intramolecular to mainly intermolecular N/C interaction. Extended interaction studies with FxxLF motifs indicate that this interaction is interrupted upon binding to DNA (van Royen et al. 2007). This raises the possibility that the

382

J. Trapman DHT Cytoplasm

NTD DBD LBD

Nucleus

AR

Cofactor

DNA ARE

Fig. 3 Model of consecutive interactions of the AR LBD with AR NTD and with cofactors in the cytoplasm and nucleus. Upon binding of the ligand in the cytoplasm, intramolecular AR N/C interaction is induced. In the nucleus, intramolecular AR N/C interaction is shifted to an intermolecular interaction. At DNA-binding the AR N/C interaction is replaced by interaction of the AR LBD with cofactors via the same cofactor-binding groove

cofactor-binding groove in a chromatine context is available for cofactors with lower affinity LxxLL or other interaction motifs.

6 Conclusions Although much progress has been made during the last few years in elucidation of the interaction of the AR LBD with the NTD and with other proteins via FxxLF and LxxLL motifs binding to the cofactor-binding groove, it is clear that still many gaps in our knowledge remain to be filled. Most important in this regard is the identification of the most relevant proteins binding to the coactivator groove during transcription regulation by AR in vivo. In addition, our knowledge of the specificity of the cofactor-binding groove in normal cells and in tumor cells ought to be extended. This should lead to answering the question of why the AR contains a cofactor-binding groove that deviates so much from that of other NRs. This knowledge will be instrumental in establishing whether the cofactor-binding groove is a therapeutic target for inhibition of AR function in prostate cancer.

Acknowledgment The author is indebted to Martin van Royen and Dennis van de Wijngaart for help with preparation of the manuscript.

Interaction of the Androgen Receptor Ligand-Binding Domain

383

References Askew EB, Gampe RT, Stanley TB, Faggart JL, Wilson EM (2007) Modulation of androgen receptor activation function AF2 by testosterone and dihydrotestosterone. J Biol Chem 282:25801–25816. Brooke GN, Parker MG, Bevan CL (2008) Mechanisms of androgen receptor activation in advanced prostate cancer: differential co-activator recruitment and gene expression. Oncogene 27:2941–2950. Chang C-Y, Norris JD, Gron H, Paige LA, Hamilton PT, Kenan DJ, Fowlkes D, McDonnell DP (1999) Dissection of the LxxLL nuclear receptor-coactivator interaction motif using combinatorial peptide libraries: discovery of peptide antagonists of estrogen receptors a and b. Mol Cell Biol 19:8226–8239. Chang C-Y, Abdo J, Hartney T, McDonnell DP (2005) Development of peptide antagonsists for the androgen receptor using combinatorial peptide phage display. Mol Endocrinol 19:2478– 2490. Doesburg P, Kuil CW, Berrevoets CA, Steketee K, Faber PW, Mulder E, Brinkmann AO, Trapman J (1997) Functional in vivo interaction between the amino-terminal transactivation domain and the ligand-binding domain of the androgen receptor. Biochemistry 36:1052–1064. Dubbink HJ, Hersmus R, Verma CS, van der Korput HAGM, Berrevoets CA, van Tol J, Ziel-van der Made ACJ, Brinkmann AO, Pike ACW, Trapman J (2004) Distinct recognition modes of FXXLF and LXXLL motifs by the androgen receptor. Mol Endocrinol 18:2132–2150. Dubbink HJ, Hersmus R, Pike ACW, Molier M, Brinkmann A, Jenster G, Trapman J (2006) Androgen receptor ligand-binding domain interaction and nuclear receptor specificity of FXXLF and LXXLL motifs as determined by L/F swapping. Mol Endocrinol 20:1742–1756. Estebanez-Perpina E, Moore JMR, Mar E, Delgado-Rodrigues E, Nguyen P, Baxter JD, Buehrer BM, Webb P, Fletterick RJ, Guy RK (2005) The molecular mechanisms of coactivator utilization in ligand-dependent transactivation by the androgen receptor. J Biol Chem 280:8060–8068. Greschik H, Moras D (2003) Structure–activity relationship of nuclear receptor-ligand interactions. Curr Top Med Chem 3:1573–1599. He B, Wilson EM (2003) Electrostatic modulation in steroid receptor recruitment of LXXLL and FXXLF motifs. Mol Cell Biol 23:2135–2150. He B, Kemppainen JA, Wilson EM (2000) FXXLF and WXXLF sequences mediate the NH2terminal interaction with the ligand binding domain of the androgen receptor. J Biol Chem 275:22986–22994. He B, Minges JT, Lee LW, Wilson EM (2002) The FXXLF motif mediates androgen receptorspecific interactions with coregulators. J Biol Chem 277:10226–10235. He B, Gampe RT, Kole AJ, Hnat AT, Stanley TB, An G, Stewart EL, Kalman RI, Minges JT, Wilson EM (2004) Structural basis for androgen receptor interdomain and coactivator interactions suggests a transition in nuclear receptor activation function dominance. Mol Cell 16:425–438. Heemers HV, Tindall DJ (2007) Androgen receptor (AR) coregulators: a diversity of functions converging on and regulating the AR transcriptional complex. Endocr Rev 28:778–808. Heinlein CA, Chang C (2002) Androgen receptor coregulators: an overview. Endocr Rev 23:175–200. Hsu C-L, Chen Y-L, Yeh S, Ting H-J, Hu Y-C, Lin H, Wang X, Chang C (2003) The use of phage display technique for the isolation of androgen receptor interacting peptides with (F/W)XXL (F?W) and FxxLY new signature motifs. J Biol Chem 278:23691–23698. Hu YC, Yan S, Yeh SD, Sampson ER, Huang J, Li P, Hsu CL, Ting HJ, Lin HK, Wang L, Kim E, Ni J, Chang C (2004) Functional domain and motif analyses of androgen receptor coregulator ARA70 and its differential expression in prostate cancer. J Biol Chem 279:33438–33446.

384

J. Trapman

Hur E, Pfaff SJ, Payne ES, Gron H, Buehrer BM, Fletterick RJ (2004) Recognition and accommodation at the androgen receptor coactivator binding interface. PLoS Biol 2:1301–1312. Li Y, Lambert MH, Xu HE (2003) Activation of nuclear receptors: a perspective from functional genomics. Structure 11:741–746. McKenna NJ, O’Malley BW (2002) Combinatorial control of gene expression by nuclear receptors and coregulators. Cell 108:465–474. Ozers MSS, Marks BD, Gowda K, Kupcho KR, Ervin KM, De Rosier T, Qadir N, Eliason HC, Riddle SM, Shekhani MS (2007) The androgen receptor T877A mutant recruits LXXLL and FXXLF peptides differently than wild-type androgen receptor in a time-resolved fluorescence resonance energy transfer assay. Biochemistry 46:683–695. Rosenfeld MG, Lunyak VV, Glass CK (2006) Sensors and signals: a coactivator/corepressor/ epigenetic code for integrating signal-dependent programs of transcriptional response. Genes Dev 20:1405–1428. Schaufele F, Carbonell X, Guerbadot M, Borngraeber S, Chapman MS, Ma AAK, Miner JN, Diamond MI (2005) The structural basis of androgen receptor activation: intramolecular and intermolecular amino–carboxy interactions. Proc Natl Acad Sci USA 102:9802–9807. Steketee K, Berrevoets CA, Dubbink HJ, Doesburg P, Hersmus R, Brinkmann AO, Trapman J (2002) Amino acids 3-13 and amino acids in and flanking the FXXLF motif modulate the interaction between the amino-terminal and ligand-binding domain of the androgen receptor. Eur J Biochem 269:5780–5791. van de Wijngaart DJ, van Royen ME, Hersmus R, Pike ACW, Houtsmuller AB, Jenster G, Trapman J, Dubbink HJ (2006) Novel FXXFF and FXXMF motifs in androgen receptor cofactors mediate high affinity and specific interactions with the ligand-binding domain. J Biol Chem 281:19407–19416. van Royen ME, Cunha SM, Brink M, Mattern KA, Nigg AL, Dubbink HJ, Verschure PJ, Trapman J, Houtsmuller AB (2007) Compartmentalization of androgen receptor protein–protein interactions in living cells. J Cell Biol 177:63–72. Wang L, Hsu CL, Ni J, Wang PH, Yeh S, Keng P, Chang C (2004) Human checkpoint protein hRad9 functions as a negative coregulator to repress androgen receptor transactivation in prostate cancer cells. Mol Cell Biol 24:2202–2213.

Multitasking and Interplay Between the Androgen Receptor Domains F. Claessens, T. Tanner, and A. Haelens

Abstract The androgen receptor, like the other nuclear receptors, consists of three canonical domains: the aminoterminal domain (NTD), the DNA-binding domain (DBD) and the ligand-binding domain (LBD). The flexible hinge between the DBD and LBD can also be considered as a separate entity. Each of these domains has multiple functions. The NTD harbors two interdependent transactivation functions Tau-1 and Tau-5, two SUMOylation sites that seem to control cooperativity of the AR, and an 23FQNLF27 motif that interacts with high affinity with the ligandbinding domain. The DBD is involved in the correct interactions of the AR with its response elements, but it also contains a nuclear export signal as well as a nuclear translocation signal. The hinge region controls the interactions of the AR with selective AREs. It harbors an acetylation and a phosphorylation acceptor site, overlaps with the nuclear translocation signal, and seems involved in the control of the steady state of the AR. The LBD binds its natural agonists with high affinity; it interacts with heat-shock protein complexes when unbound and with a series of coregulators when bound by agonists. Many of these coregulators harbor motifs that resemble the 23FQNLF27-motif of the NTD. Clearly, the domains of the AR do not function independently, but rather act in concert with each other and with other proteins during androgen activation of transcription. The overall activity of the AR as a transcription factor is codetermined by such communications, and also by the nature of the response element or enhancer it is binding to.

F. Claessens(*) Molecular Endocrinology Laboratory, Department of Molecular Cell Biology, K.U. Leuven, Campus GHB O/N1, Herestraat 49, 3000 Leuven, Belgium

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4_17, # Springer Science + Business Media, LLC 2009

385

386

F. Claessens et al.

1 Introduction 1.1

Molecular Mechanism of AR Action

The genomic action of the androgen receptor (AR) follows the classical ligandinduced pathway observed for most of the steroid receptors (reviewed in Gelman 2002; Gobinet et al. 2002 and Fig. 1a). In its inactive form the AR is sequestered in the cytoplasm as a complex with heat-shock proteins. This AR:Hsp complex maintains the AR in a suitable structure that allows the efficient binding of hormone to the ligand-binding domain (LBD) (Fang et al. 1996). Androgens will bind to the LBD of the AR. Once bound by androgen, the AR undergoes a conformational change, is released from its complex with the heat-shock proteins (Veldscholte et al. 1992), and acquires its active state. To exert its effects at the genomic level, the ligand-activated AR translocates to the nucleus. The active conformation of the AR also facilitates dimerization of the AR molecules that bind to response elements at the enhancer of a target gene (Luisi et al. 1991). In addition, the ligand-induced conformational changes facilitate the recruitment of coactivator complexes to the AR. These include members of the SRC family (Alen et al. 1999; Bevan et al. 1999), CBP/p300 (Huang et al. 2003), SWI/SNF (Link et al. 2005), and the Mediator complex (Vijayvargia et al. 2007). SRC coactivators interact directly with the AR and in turn recruit p300 and CARM1, as well as a series of other coactivators (Lee et al. 2007 and references therein). The SWI/SNF and Mediator complexes are then targeted to the chromatin by p300. These complexes remodel chromatin and alter DNA topology, and as a result the responsive genes become more accessible to other transcription factors (Huang et al. 2003). Subsequent recruitment of the basal transcription machinery to the promoter of the target gene results in the initiation of transcription (Lewis and Reinberg 2003). The product of transcription, mRNA, is transported to the cytoplasm where it is translated, and the resultant protein can then elicit the gene response.

1.2

Natural Mutations as Learning Tools

Several diseases are associated with AR dysfunction, the most well known are Androgen Insensitivity Syndrome (AIS), Kennedy’s disease, and prostate cancer. AIS is an X-linked genetic disease with a failure of normal masculinization in 46, XY (chromosomally male) individuals (Gottlieb et al. 1999). The pathophysiology of AIS results from the fact that cells, which would normally be androgen sensitive, are unable to respond to androgens. This end-organ resistance to androgens has been attributed to defects in the AR status. The AR has three major functional domains: the N-terminal domain (NTD), a DNA-binding domain (DBD), and an LBD (Fig. 1b). The majority of the AIS mutations occur in the DBD or the LBD, resulting in defective DNA binding/dimerization and hormone binding, respectively. Mutations

Multitasking and Interplay Between the Androgen Receptor Domains

387

a Cytoplasm

Nucleus

5 -Red DHT

A

Co

AR

C’s

T

AR AR

Hsps

AR

mRNA

CR

T

TA

RNA Pol II F’s TBP GTFs

ARE (Enhancer)

Gene Product

GOI

(Promoter)

Erk-2

AR

Raf-1

Response

Src

b

q11-12

Intron size (kb) AR gene

>20 >15

1

Exon size (bp)

1 AR protein

q

p

X chromosome

2

26

3

1613

152 117

(Gln)19-23

(Pro)9 (Gly)16 NTD

Tau-1 AF-1

5.6

4.8 0.8 0.7

4 288

537

5

6 7 8

145 131158155

625 669

DBD H

919 LBD

Tau-5 AF-2

Fig. 1 (a) The Mechanisms of AR Action. The mechanism is explained in Sect. 1.1. Testosterone (T) passively diffuses into the cell where it can be converted to 5a-dihydrotestosterone (DHT) by 5a-reductase (5a-Red) type 1 or 2. Inactive AR is sequestered in the cytoplasm through interaction with the heat-shock proteins (Hsps). Upon DHT binding, the Hsps are released and the AR has one of two effects. (1) For genomic activity, the DHT-bound AR translocates to the nucleus and binds as a homodimer to androgen response elements (AREs) where it recruits coactivator (CoA) proteins and chromatin-remodeling complexes (CRC’s). Subsequent recruitment of the basal transcription machinery (such as RNA-polymerase II, RNA Pol II; TATA box-binding protein, TBP; and TBP associating factors, TAF’s) and general transcription factors (GTFs) result in

388

F. Claessens et al.

in the NTD are less frequent with the majority resulting directly in a stop codon. Recently mutations in the NTD have been associated with male infertility (Zuccarello et al. 2007). Kennedy’s disease is a neurodegenerative disorder (reviewed in Greenland and Zajac 2004). The AR gene of affected males has an abnormal expansion of the CAG (cytosine, adenine, and guanine) triplet repeat located in exon 1. Depending on its length, the poly-Q stretch inversely regulates the transcriptional activity of the AR protein (Callewaert et al. 2003; Buchanan et al. 2004). The development and maintenance of the normal prostate is androgen dependent. Following androgen-deprivation therapy of advanced tumors, this initially androgen-dependent cancer progresses to one of androgen independence. Interestingly, the AR continues to play a role in many of these androgen-independent tumors. The link between AR gene mutations and prostate cancer progression is controversial, as the incidence of AR mutations in primary prostate cancer is low. On the other hand, the frequency of AR mutations in metastatic prostate cancer is much higher (Avila et al. 2001; Tilley et al. 1996), suggesting that as prostate cancers metastasize to local lymph nodes, the incidence of AR mutations rises. The description of these mutations provides important clues regarding the roles of these domains in AR function.

2 The Modular Structure of the AR Protein The human AR is composed of 919 amino acids and possesses the same modular structure as other members of the nuclear receptor superfamily. The separate modules/ domains are responsible for DNA binding, ligand binding (LBD), nuclear localization, and controlling transcriptional activity (Fig. 1) (Dehm and Tindall 2007).

2.1

The NH2-Terminal Domain (NTD)

The NTD contains three amino acid repeats, namely the poly-glutamine (poly-Q), poly-proline (poly-P), and poly-glycine (poly-G) stretches (Fig. 1b). The most striking of these is the poly-Q stretch, encoded by a polymorphic CAG repeat, Fig. 1 (Continued) transcription of the gene of interest (GOI). The resultant mRNA is transported to the cytoplasm, translated, and the protein product elicits the androgen response. (2) For nongenomic activity the DHT-bound AR interacts with either membrane-bound or cytoplasmic proteins. (b) The Genomic Organization of the AR Gene and Protein. The AR gene was mapped to the long arm of the X-chromosome at q11–12 and spans approximately 90 kb. The gene contains eight exons that code for a 2,757 bp open reading frame within a 10.6-kb mRNA. The resultant AR protein is composed of 919 amino acids and has a molecular mass of approximately 110 kDa. The corresponding shaded areas in the diagram of the protein structure demonstrate how the exon organization translates into discrete functional regions of the receptor protein. The three regions of repetitive DNA sequences in the first exon (NTD) are shown. The major activation functions (AF-1 and AF-2) are shown, where AF-1 is further delineated into Tau-1 and Tau-5

Multitasking and Interplay Between the Androgen Receptor Domains

389

which normally contains 19–23 repeats but can also contain as few as 14, or more than 38 repeats (Sleddens et al. 1992). The length of the poly-Q stretch affects AR activity. Shorter poly-Q stretches are associated with a higher AR transcriptional activity, whereas longer repeats are linked to a lower AR transcriptional activity (Callewaert et al. 2003; Buchanan et al. 2004). For the poly-G stretch, similar but less pronounced effects have been reported (Werner et al. 2006). The major transcription activation function AF-1 of the AR resides within the NTD. AF-1 has initially been shown to comprise two activation domains: transcription activation unit 1 (Tau-1) and transcription activation unit 5 (Tau-5) (Jenster et al. 1995). Both Tau-1 and Tau-5 are required for the full activity of the AR, as well as the intrinsic activity of the AR deleted of its LBD. The boundaries of these domains are indicated in Fig. 1b.

2.1.1

Transcription Activation Unit 5

The transcription activation unit 5 resides between aa 370 and aa 494 (Jenster et al. 1995). It acts as an independent activation function but is much stronger when fused to the ARDBD as compared to the Gal4-DBD (Callewaert et al. 2006), indicating a functional communication between these two AR domains. The AR-NTD recruits steroid receptor coactivators of the p160 family through interactions between Tau-5 and a conserved glutamine-rich region (Irvine et al. 2000; Alen et al. 1999). Initial truncation analysis of Tau-5 indicated that its complete sequence is probably involved in a structure needed for its interaction with the p160s (Christiaens et al. 2002; Callewaert et al. 2006). However, the 433WHTLF437 motif, first described as a possible interaction site with the liganded AR LBD (Alen et al. 1999; He et al. 1999), has now been revealed to be crucial for the actions of Tau-5 in androgenindependent prostate cancer (Dehm et al. 2007). Metzger et al. (2003) unraveled a signal transduction pathway that links the RhoA effector protein kinase C-related kinase PRK1 to Tau-5 and results in superactivation of AR by low concentrations of agonists or even by antagonists. This is an alternative AR activation mechanism possibly important in some cases of prostate cancer. Several prostate cancer point mutations in the AR gene reside in Tau-5, but the effect of these mutations on the p160 and/or PRK1 activation remains to be elucidated.

SUMOylation of the AR in Tau-5 A number of transcription factors can be modified by the small ubiquitin-related modifier (SUMO). The process of conjugating SUMO, a 101-amino acid polypeptide, to a target protein is termed SUMOylation. SUMO is covalently linked to lysine residues in target proteins. SUMOylation is often involved in directing the subcellular localization and stabilization of transcription factors (reviewed in Freiman and Tjian 2003). The conjugation pathway of SUMOylation is mediated by three enzymes: an activation enzyme, a conjugation enzyme, and a ligating enzyme.

390

F. Claessens et al.

Furthermore, SUMOylation is reversible. The E2 SUMO-conjugating enzyme Ubc9 was shown to interact with the AR via Tau-5 (Callewaert et al. 2004) as well as via the hinge region (Poukka et al. 2000a, b). The AR is SUMOylated at two lysine residues in Tau-5 (Poukka et al. 2000a). Substitution of these residues results in an increase in the transcriptional activity of the AR protein via an increased potency for cooperativity on composite response elements (Callewaert et al. 2004).

2.1.2

Transcription Activation Unit 1

Tau-1 has been delineated between amino acids 101 and 360 of the AR. Its deletion from the full-length AR results in a strongly impaired transactivation (Jenster et al. 1995), and this AR fragment displays transactivation properties when fused to a heterologous DBD. Several motifs and Tau-1-interacting proteins have been reported.

Core Tau-1 The AR segment between amino acids 177 and 194 has been defined as the core of Tau-1 based on its property as an autonomous activation function (Callewaert et al. 2006). Core Tau-1 coincides with the earlier described AF-1a (Chamberlain et al. 1996). It regulates the interaction between the other activation unit Tau-5 and the p160 coactivators, and also interacts with, as yet undefined, coactivators (Callewaert et al. 2006). The same AR motif has also been implicated recently in a possible mechanism by which IL1 beta converts an androgen antagonist to function as an agonist (Zhu et al. 2006). Indeed, increased IL1 beta signaling by infiltrating fibroblasts in prostate cancer seems to change the ability of a TAB2-containing NCoR complex to interact with the AR, which converts the antagonistic characteristics of ligands into agonistic characteristics.

The CHIP Interaction Site Immediately carboxy-terminal of core Tau-1, a highly conserved motif ‘‘AKELCKAVSVSMGL,’’ was discovered by the group of Wilson (He et al. 2004a) to interact with an E3 ubiquitin ligase CHIP (carboxy-terminus of the Hsp70-interacting protein). CHIP–AR interactions also can be mediated through the AR-hinge region, and thus these two AR domains could be bridged by CHIP (Rees et al. 2006). Functionally, the AR–CHIP interactions might control the steady-state levels of the AR, but the exact action mechanism is difficult to determine because of the many contact points between the heat-shock proteins and the AR.

Multitasking and Interplay Between the Androgen Receptor Domains

2.1.3

391

Structure of the AR-NTD

There is no atomic resolution structure of the AR-NTD available. Studies using circular dichroism, fluorescence spectroscopy, partial protease digestion, and secondary structure predictions have revealed that the AR AF-1 lacks stable secondary structure in aqueous solution but adopts a more a-helical structure in the presence of a structure-stabilizing solute. Furthermore, even though there is almost no amino acid sequence homology between the NTDs of the NRs, they appear to share some structural characteristics, where hydrophobicity seems to be of importance, e.g., in core Tau-1. Many of the other NRs also adopt the more a-helical conformation in the presence of structure-stabilizing solutes (reviewed in McEwan 2004; McEwan et al. 2007).

2.1.4

N/C Interactions

An interesting characteristic of the AR is the ligand-induced interaction between the amino- and carboxy-terminals of the protein (N/C interaction) (Langley et al. 1995; Doesburg et al. 1997). The first 30 amino acids of the NTD have been demonstrated to be of importance for this interaction. More specifically, the 23 FQNLF27 motif and its flanking residues have been shown to be critical for the N/C interaction (Berrevoets et al. 1998). The N/C interaction of the AR seems to be involved in many processes. The deletion of the 23FQNLF27 motif results in a less active AR in transient transfections, and several AIS mutations result in a reduction of the AR potency. This N/C interaction is also thought to play a role in the ligandinduced stabilization of the AR protein (reviewed in (He et al. 1999, 2004b). Surprisingly, Van Royen et al. (2007) demonstrated that the AR present in speckles inside the nucleus has no N/C interactions, while the mobile fraction of the AR has a closed conformation.

2.1.5

NTD-Interacting Proteins

The NTD was revealed to interact with a number of proteins. These proteins include proteins that form part of the basal transcription machinery (TFIIF, TFIIH, PTEFb); coactivator proteins (SRC-1, ARA67, ARA160, ART27, CBP, MAGE11, PRK1); corepressor proteins (SMRT, AES); transcription factors (SMAD3, Daxx, STAT3, ANT-1); and proteins that play a role in the cell cycle (BRCA1, cyclin E, cyclin D1, pRB, caveolin-1, ARNIP)(reviewed in Dehm and Tindall 2007). Several of these proteins have been proposed to affect the N/C interaction, either enhancing or inhibiting it (e.g., He et al. 2004a; Burd et al. 2005 and Fig. 3). Surprisingly, mutations near the 23FQNLF27 motif can strongly affect the AR activity, without affecting the N/C interactions, suggesting that this motif might have more than one function (Callewaert et al. 2003; Bai et al. 2005; Li et al. 2007).

392

2.2 2.2.1

F. Claessens et al.

The DNA-Binding Domain Defining Androgen Response Elements

The AR is a transcription factor, which either directly binds androgen response elements (ARE), or is tethered to DNA by other transcription factors, as has been postulated for serum response factor (Heemers et al. 2007). Indirect effects of androgens can also occur through AREs, e.g., the activation of the transcription factor SREBP via an ARE in the gene encoding the activating protease SCAP (Heemers et al. 2004) Analogous to what had been described for the glucocorticoid and progestagen receptors, the AR was shown to act through DNA motifs that contain the hexamer 50 -TGTTCT-30 present, e.g., in the MMTV enhancer (Cato et al. 1987; Ham et al. 1988) and the tyrosine aminotransferase enhancer (Denison et al. 1989). In parallel, similar motifs were described in androgen target genes, like those encoding the components of the rat-prostatic-binding protein (Claessens et al. 1989, 1990). The AR is dependent on cooperativity to elicit a clear androgen response in transient transfections as well as on chromatin-integrated templates. This cooperativity can happen between multiple ARs, binding to adjacent AREs, or with other transcription factors, binding to adjacent motifs (e.g., Celis et al. 1993; Wang et al. 2007). In general, the DBDs of the steroid receptors (excluding the estrogen receptors) recognize nonselective hormone response elements (HREs) composed of inverted repeats of the 50 TGTTCT-30 hexanucleotide core sequence, separated by a three-nucleotide spacer. Adjacent to such classical AREs (clAREs), the AR has been demonstrated to bind to a set of selective androgen response elements (selAREs), which are proposed to be arranged as direct repeats of the 50 TGTTCT-30 -like core sequences. These selAREs have been identified in the regulatory regions of a number of genes such as the rat probasin, human secretory component, mouse sex-limited protein, and the mouse homeobox protein Pem genes (Claessens et al. 1996; Verrijdt et al. 1999; Barbulescu et al. 2001). When compared to the clAREs, the selAREs appear to possess a highly conserved, right half-site and a more degenerated left half-site. However, specific residues within the left half-site have been shown to be important in conferring AR selectivity (Schoenmakers et al. 2000; Verrijdt et al. 2000).

2.2.2

Structure of the DBD

The DBD is composed of about 90 amino acids including nine cysteine residues, eight of which form two co-ordination complexes, each consisting of four cysteines and a Zn2+ ion (the zinc fingers) (Fig. 2; Shaffer et al. 2004). Each zinc finger is encoded by a separate exon (2 and 3, see Fig. 1b). The DBD is further organized into three a-helices. The first helix, located at the C-terminal of the first zinc finger, contains the proximal box (P-box) and other residues involved in response element recognition or sequence discrimination. This helix is inserted into the major groove

Multitasking and Interplay Between the Androgen Receptor Domains

Hinge

DBD

a Exon 2

S

G

C

H

K

L

I L

C

C Zn

C

T P P Q K

C

N

D

P-box

P

Zn S

S

C

C

D-box R D-box

G

R N

T

T

G

R

A

Exon 4

K

I

A D

F

D

G

E

...

Exon 3

Y

A

393

C

L K V F F K R A A E G K Q K Y

S C R

NLS

L R K C Y E A G M T L G A R K L K K L G N

...

CTE

Fig. 2 The DNA-Binding Domain of the AR. (a) The amino acid sequence (residues 554–636) for part of the DBD (encoded by exons 2 and 3) and hinge region (encoded by part of exon 4) is shown using the one-letter code. The two zinc fingers resulting from co-ordination of a zinc atom (Zn) by four cysteine residues (C) are depicted. The first zinc finger contains the P-box (white residues) that is involved in the recognition of the androgen response elements. The second zinc finger contains the D-box (speckled residues) that is involved in the interaction with a second receptor molecule to form the homodimer complex. The bipartite nuclear localization signal (NLS) spans residues 617–635. The carboxy-terminal extension (CTE) of 4 or 12 amino acid residues into the hinge region, that is required for binding to nonselective hormone response elements and selective androgen response elements respectively, are shown. (b) Overall architecture of an AR-DBD dimer bound to a direct repeat of the 50 TGTTCT-30 motif as described in Shaffer et al. 2004. The spheres indicate the position of the Zn atoms. The two monomers are oriented in a head-to-head conformation, and the first a-helix enters the major groove of the DNA double helix at the bottom

394

F. Claessens et al.

of the response element during DNA binding. The second zinc finger contains the distal box (D-box) at its N-terminal, the second helix at its tip, the third helix at its C-terminus (see Fig. 2b). It is the second zinc finger which contains most of the residues involved in DNA-dependent dimerization (Claessens and Gewirth 2004). Surprisingly, the carboxy-terminal part of Tau-5 has an inhibitory function on DNA binding when tested in vitro (Liu et al. 2003). The ability of the AR to bind to the selAREs (see Sect. 2.2.1.) provides a general mechanism of androgen specificity. Furthermore, it has been demonstrated that the second zinc finger and a twelve-amino acid C-terminal extension (CTE, Fig. 2a) of the AR DBD is required for high-affinity binding to these selective AREs (Schoenmakers et al. 1999; Haelens et al. 2003). The crystal structure of the AR-DBD (Shaffer et al. 2004) implicated two residues in the D-box in DNA-dependent dimerization. However, swapping the Ser597 or Thr602 between AR and GRDBDs did not affect binding to selAREs or clAREs. By contrast, swapping the CTEs did affect binding to selAREs, but not to clAREs (Verrijdt et al. 2006). In addition, the nucleotide sequence flanking these AREs has been demonstrated to play a role in controlling the activity of the AR (Ham et al. 1988; Haelens et al. 2003).

2.2.3

The Nuclear Export Signal in the AR-DBD

Nuclear export activity has been mapped to the DBD of a number of NRs, including the AR (Black et al. 2001). This NES lies between the two zinc fingers in the DBD and more specifically requires the two residues, 583FF584. It was shown that export using this signal requires binding of calreticulin and is Crm1 (chromosomal region maintenance 1) independent. It had already been determined that calreticulin binds to the same site in the AR, inhibiting binding to DNA response elements as well as transcriptional activity (Dedhar et al. 1994). For the AR, a second NES has been identified in the LBD within the region of amino acid residues 742–817 (Saporita et al. 2003). This NES is leptomycin B insensitive and therefore Crm1 independent. It is active in the absence of androgen and repressed upon ligand binding. The precise mechanism by which the AR is exported out of the nucleus via either of these NESs still needs to be determined.

2.3

The Ligand-Binding Domain

The crystal structures of several nuclear receptor LBDs have been determined. It was revealed that there is a striking conservation in the three-dimensional structure of these domains, regardless of the modest sequence homology between species, which in some cases is as low as 20%. In general, these LBDs are made up of 12 helices (H1–H12) and two b-sheets (S1/S2) that form the ligand-binding pocket. In the unliganded state, helix 12 is positioned away from the pocket in a conformation that prevents the binding of

Multitasking and Interplay Between the Androgen Receptor Domains

395

coactivators. When ligand binds the receptor, helix 12 folds over the pocket, encloses the ligand, and as a consequence exposes a groove to which coactivators can bind (Gelmann 2002; Gobinet et al. 2002; and references therein). The AR LBD is composed of 250 amino acid residues and is arranged as 11 helices, since helix 2 is absent. The crystal structure of the AR LBD was first determined in complex with both the natural ligand, DHT (Sack et al. 2001), and the synthetic ligand, methyltrienolone (R1881) (Matias et al. 2000). These results showed that the AR LBD possesses the typical nuclear receptor LBD fold. Eighteen amino acid residues of the AR LBD have been identified that directly interact with the bound ligand. Most of these are hydrophobic residues that mainly interact with the steroid scaffold. Few are polar residues that are thought to form hydrogen bonds with the polar atoms in the ligand (Matias et al. 2000). Upon ligand binding, the AR undergoes a conformational change similar to the other nuclear receptors, where helix 12 folds over the ligand pocket and exposes a groove for protein–protein interaction. For the AR, this groove interacts primarily with the 23FQNLF27 motif of the NTD (N/C interaction) as opposed to the LxxLL motif found in coactivator proteins. Although coactivators can also bind to this groove, the NTD of the AR is the predominant site for coactivator binding (see Sect. 2).

2.4

The Hinge Region

The hinge region, situated between the DBD and the LBD, was previously thought to be unstructured and to function simply as a physical linker between these two domains (Fig. 1b). However, recent studies indicate that for some nuclear receptors the amino-terminus of the hinge region contains a helix that can be disrupted upon DNA binding. Such alterations in structure appear to be important for adaptation to different DNA response elements (Nascimento et al. 2006; Rastinejad et al. 2000). Furthermore, analysis of the AR, as well as other nuclear receptors, suggests that the hinge region plays a role in controlling the transcriptional activity of these proteins (Xu et al. 2004; Iordanidou et al. 2005; Wang et al. 2001; Tanner et al. 2004). A number of studies have shown that the hinge is an interaction domain for several proteins, including SNURF, Tip60, Ubc9, and Cyclin D1 reviewed in Heinlein and Chang (2002), as well as PAK6 (Lee et al. 2002), SMRT (Liao et al. 2003), BRCA2 (Shin and Verma 2003), ARR19 (Jeong et al. 2004), and SGT-1 (Buchanan et al. 2007), among others. These proteins do not have to function as transcriptional coactivators per se, but could be involved in many other steps of the androgen responses. 2.4.1

The Hinge Region and DNA Binding

The hinge region of the AR has been implicated in DNA binding and more specifically has been shown to be imperative for binding to selAREs (Haelens et al. 2003; Schoenmakers et al. 1999). This has been discussed in Sect. 2.2.2.

396

2.4.2

F. Claessens et al.

The Hinge Region and Intracellular Localization

The hinge region contains the second half of the bipartite NLS (Fig. 2a) and, as a consequence, regulates the nuclear translocation of the AR protein (Zhou et al. 1994; Haelens et al. 2007). Similar to other steroid receptors, it has been suggested that the AR has two NLSs. The first, NLS-1, is a classical, basic NLS that overlaps the DBD and the hinge region (Fig. 2a). This is a bipartite NLS that is involved in the ligand-dependent nuclear translocation of the AR (Jenster et al. 1993). In accordance with the amino acid numbering used in defining the various domains of the AR in Fig. 2a, this NLS was defined as 617RKCYEAGMTLGARKLKKLG635. The second, NLS-2, resides in the LBD and is also ligand dependent (Poukka et al. 2000b). NLS-1 in the AR appears to be more potent and is responsible for rapid nuclear import, whereas for NLS-2 nuclear import seems to be incomplete and slower. In general, the NLS-2 activity of the steroid receptors is poorly defined, and as there is no evidence of a linear NLS, it is assumed that this import signal lies within the three-dimensional structure of the ligand-bound LBD.

2.4.3

The Hinge Region and Post-translational Modifications

The hinge region is a target site for post-translational modifications such as phosphorylation (Zhou et al. 1995; Blok et al. 1996) and acetylation (Fu et al. 2000). This section of our review does not provide an overview of all possible modifications (as, e.g., in Blok et al. 1996; Faus and Haendler 2007), but rather provides a discussion of those modifications for which more is known on their involvement in specific functions of the AR. The AR can be acetylated at a highly conserved lysine-rich motif (630KLKK633) in the hinge region by both p300 and p300/cAMP-response element-binding protein-associated factor (P/CAF) (Fu et al. 2000). In a subsequent study, acetylation-deficient AR mutants were shown to have reduced transcriptional activity at androgen-responsive reporter genes (Fu et al. 2002). In addition, the preferentially acetylated lysine residues were identified as K630 and K632, with an additional acetylation site at K633. Acetylation of the AR has also been shown to be mediated by Tat-interactive protein, 60 kDa (Tip60) (Gaughan et al. 2002). In this study, it was further demonstrated that AR activity is down-regulated by the histone deacetylase activity of histone deactylase-1 (HDAC1). Chromatin immunoprecipitation assays revealed that the AR, Tip60, and HDAC1 form a trimeric complex at the endogenous prostate-specific antigen (PSA) promoter, suggesting that acetylation and deacetylation of the AR is an important mechanism for regulating transcriptional activity. On the other hand, in a recent study we have demonstrated that mutation of these acetylation sites results in an AR protein that is more active than the wild-type AR protein. This increased activity is consistent with the fact that these mutations were identified in biopsies from patients with prostate cancer (Haelens et al. 2007). These results suggest that the acetylation status of the AR protein at the 630KLKK633 motif differentially

Multitasking and Interplay Between the Androgen Receptor Domains

397

regulates AR activity in a cell- or tissue-type specific manner. When the 630 KLKK633 residues are mutated individually, this affects the activity of the AR, but contradicting reports have been published. The data of Faus and Haendler (2007) seem to solve this issue, since they report differential effects of these mutations on different reporter genes. It remains difficult to assess the exact effects of each of these mutations, since on the one hand they reduce in vitro DNA binding, as well as nuclear translocation, and on the other hand they affect transactivation properties of the AR. In addition, the intranuclear mobility of the AR as measured by FRAP is also disturbed, since the mutant proteins seem to have lost most of their intranuclear DNA binding and hence are much more mobile (Tanner et al. in prep.).

2.4.4

Putative PEST Sequence in the Hinge Region

PEST (proline-, glutamate-, serine-, and threonine-rich) sequences are involved in targeting proteins for degradation by the 26S proteasome (Rechsteiner and Rogers 1996). The human AR contains a putative PEST sequence (638KLQEEGEASSTTSPTEETTQK658) in the hinge region. A similar PEST sequence has been identified in the hinge-LBD of the vitamin D receptor (Li et al. 1999). Furthermore, it is known that AR degradation is regulated by the ubiquitin–proteasome pathway, as ubiquitylation destabilizes the protein (Gaughan et al. 2002), and inhibition of proteasomes increases endogenous AR levels (Sheflin et al. 2000). In addition to the ubiquitin–proteasome pathway, a second proteolytic pathway has been identified that regulates AR degradation. This pathway engages PTEN (phosphatase and tensin homolog deleted on chromosome 10) tumor suppressor activity (Lin et al. 2004). PTEN directly interacts with the DBD hinge (amino acids 483– 651) of the AR, blocking nuclear translocation and promoting AR degradation. Evidence suggests that caspase 3 may mediate this PTEN-induced AR degradation. In summary, the hinge region, which first had the reputation of being a less important connector between the DBD and LBD, is now known to have important input/output functions.

3 Conclusions: Multitasking and Communications Between Different AR Domains The AR, like the other nuclear receptors, consists of three canonical domains: the NTD, the DBD, and the LBD (Fig. 1b). The flexible hinge between the DBD and LBD should also be considered as a separate domain, although it is most likely unstructured. These domains have multiple functions. The aminoterminal domain harbors two interdependent transactivation functions Tau-1 and Tau-5, two SUMOylation sites that seem to control cooperativity of the AR, and an 23 FQNLF27 motif that interacts with high affinity with the ligand-binding domain.

398

F. Claessens et al. MAGE-11

S

FQNLF

S

hAR Tau-1

Tau-5

NLS DBD hinge

LBD

p160

Fig. 3 Overview of the communications between different domains of the AR. The different domains and their interactions are depicted as described in the text and summarized in Sect. 3. The Ss indicates the SUMOylation sites, NLS stands for nuclear localization signal

The DNA-binding domain is involved in the correct interactions of the AR with its response elements, but it also contains a nuclear export signal as well as a nuclear translocation signal. The hinge region controls the interactions of the AR with selective AREs. It harbors an acetylation and a phosphorylation acceptor site, overlaps with the nuclear translocation signal, and seems involved in the control of the steady state of the AR. The ligand-binding domain binds with high-affinity agonists and antagonists, interacts with heat-shock protein complexes when unbound and with a series of coregulators when bound by agonists. Many of these coregulators harbor 23FQNLF27-like motifs. Clearly, these domains do not function independently, but rather act in concert with each other and with other proteins during androgen activation of transcription (Fig. 3), but the eventual AR activity as a transcription factor also depends on the nature of the response element or enhancer under study.

Acknowledgments We are grateful to the members of the Molecular Endocrinology Laboratory of Leuven for helpful discussions. This laboratory has been supported by the ‘‘Fonds voor Wetenschappelijk Onderzoek-Vlaanderen (FWO),’’ by a U.S. Army Prostate Cancer Research Program grant, by a ‘‘Geconcerteerde Onderzoeksactie K.U. Leuven,’’ and by a grant of the Association for International Cancer Research (AICR).

References Alen, P., Claessens, F., Verhoeven, G., Rombauts, W. and Peeters, B. 1999. The androgen receptor amino-terminal domain plays a key role in p160 coactivator-stimulated gene transcription. Mol. Cell. Biol. 19:6085–6097. Avila, D.M., Zoppi, S. and McPhaul, M.J. 2001. The androgen receptor (AR) in syndromes of androgen insensitivity and in prostate cancer. J. Steroid Biochem. Mol. Biol. 76:135–142.

Multitasking and Interplay Between the Androgen Receptor Domains

399

Bai S., He, B. and Wilson, E.M. 2005 Melanoma antigen gene protein MAGE-11 regulates androgen receptor function by modulating the interdomain interaction. Mol. Cell Biol. 25:12380–1257. Barbulescu, K., Geserick, C., Schu¨ttke, I., Schleuning, W. and Haendler, B. 2001. New androgen response elements in the murine pem promoter mediate selective transactivation. Mol. Endocrinol. 15:1803–1816. Berrevoets, C.A., Doesburg, P., Steketee, K., Trapman, J. and Brinkmann, A.O. 1998. Functional interactions of the AF-2 activation domain core region of the human androgen receptor with the amino-terminal domain and with the transcriptional coactivator TIF2 (transcriptional intermediary factor2). Mol. Endocrinol. 12:1172–1183. Bevan, C.L., Hoare, S., Claessens, F., Heery, D.M. and Parker, M.G. 1999. The AF1 and AF2 domains of the androgen receptor interact with distinct regions of SRC1. Mol. Cell. Biol. 19:8383–8392. Black, B.E., Holaska, J.M., Rastinejad, F. and Paschal, B.M. 2001. DNA binding domains of diverse nuclear receptors function as nuclear export signals. Curr. Biol. 11:1749–1758. Blok, L.J., de Ruiter, P.E. and Brinkmann, A.O. 1996. Androgen receptor phosphorylation. Endocr. Res. 22:197–219. Buchanan, G., Yang, M., Cheong, A., Harris, J.M., Irvine, R.A., Lambert, P.F., Moore, N.L., Raynor, M., Neufling, P.J., Coetzee, G.A. and Tilley, W.D. 2004. Structural and functional consequences of glutamine tract variation in the androgen receptor. Hum. Mol. Genet. 13:1677–1692. Buchanan, G., Ricciardelli, C., Harris, J.M., Prescott, J., Yu, Z.C., Jia, L., Butler, L.M., Marshall, V.R., Scher, H.I., Gerald, W.L., Coetzee, G.A. and Tilley, W.D. 2007. Control of androgen receptor signaling in prostate cancer by the cochaperone small glutamine-rich tetratricopeptide repeat containing protein a. Cancer Res. 67:1–10. Burd, C.J., Petre, C.E., Moghadam, H., Wilson, E.M. and Knudsen, K.E. 2005. Cyclin D1 binding to the androgen receptor NH2-terminal domain inhibits activation function 2 association and reveals dual roles for AR corepression. Mol. Endocrinol. 19:607–620. Callewaert, L., Christiaens, V., Haelens, A., Verrijdt, G., Verhoeven, G. and Claessens, F. 2003. Implications of a polyglutamine tract in the function of the human androgen receptor. Biochem. Biophys. Res. Commun. 306:46–52. Callewaert, L., Verrijdt, G., Haelens, A. and Claessens, F. 2004. Differential effect of small ubiquitin-like modifier (SUMO)-ylation of the androgen receptor in the control of cooperativity on selective versus canonical response elements. Mol. Endocrinol. 18:1438–1449. Callewaert, L., Van Tilborgh, N. and Claessens, F. 2006. Interplay between two hormoneindependent activation domains in the androgen receptor. Cancer Res. 66:543–553. Cato, A.C., Henderson, D. and Ponta, H. 1987. The hormone response element of the mouse mammary tumour virus DNA mediates the progestin and androgen induction of transcription in the proviral long terminal repeat region. EMBO J. 6:363–368. Celis, L., Claessens, F., Peeters, B., Heyns, W., Verhoeven, G. and Rombauts, W. 1993. Proteins interacting with an androgen-responsive unit in the C3(1) gene intron. Mol. Cell. Endocrinol. 94:165–172. Chamberlain, N.L., Whitacre, D.C. and Miesfeld, R.L. 1996. Delineation of two distinct type 1 activation functions in the androgen receptor amino-terminal domain. J. Biol. Chem. 271:26772–26778. Christiaens, V., Bevan, C.L., Callewaert, L., Haelens, A., Verrijdt, G., Rombauts, W. and Claessens, F. 2002. Characterization of two co-activator interacting surfaces of the androgen receptor and their relative role in transcriptional control. J. Biol. Chem. 277:49230–49237. Claessens, F. and Gewirth, D. 2004. DNA recognition by nuclear receptors. Essays Biochem. 40:59–72. Claessens, F., Celis, L., Peeters, B., Heyns, W., Verhoeven, G. and Rombauts, W. 1989. Functional characterization of an androgen response element in the first intron of the C3(1) gene of prostatic binding protein. Biochem. Biophys. Res. Commun. 164:833–840.

400

F. Claessens et al.

Claessens, F., Rushmere, N.K., Davies, P., Celis, L., Peeters, B. and Rombauts, W.A. 1990. Sequence-specific binding of androgen-receptor complexes to prostatic binding protein genes. Mol. Cell. Endocrinol. 74:203–212. Claessens, F., Alen, P., Devos, A., Peeters, B., Verhoeven, G. and Rombauts, W. 1996. The androgen-specific probasin response element 2 interacts differentially with androgen and glucocorticoid receptors. J. Biol. Chem. 271:19013–19016. Dedhar, S., Rennie, P.S., Shago, M., Hagesteijn, C.Y., Yang, H., Filmus, J., Hawley, R.G., Bruchovsky, N., Cheng, H., Matusik, R.J. and Gigue´re, V. 1994. Inhibition of nuclear hormone receptor activity by calreticulin. Nature 367:480–483. Dehm, S.M. and Tindall, D.J. 2007. Androgen receptor structural and functional elements: role and regulation in prostate cancer. Mol. Endocrinol. 21:2855–2863. Dehm, S.M., Regan, K.M., Schmidt, L.J. and Tindall, D.J. 2007. Selective role of an NH2terminal WxxLF motif for aberrant androgen receptor activation in androgen depletion independent prostate cancer cells. Cancer Res. 67:10067–10077. Denison, S.H., Sands, A. and Tindall, D.J. 1989. A tyrosine aminotransferase glucocorticoid response element also mediates androgen enhancement of gene expression. Endocrinology 124:1091–1093. Doesburg, P., Kuil, C.W., Berrevoets, C.A., Steketee, K., Faber, P.W., Mulder, E., Brinkmann, A. O. and Trapman, J. 1997. Functional in vivo interaction between the amino-terminal, transactivation domain and the ligand binding domain of the androgen receptor. Biochemistry 36:1052–1064. Fang, Y., Fliss, A.E., Robins, D.M. and Caplan, A.J. 1996. Hsp90 regulates androgen receptor hormone binding affinity in vivo. J. Biol. Chem. 271:28697–28702. Faus, H. and Haendler, B. 2007. Post-translational modifications of steroid receptors. Biomed. Pharmacother. 60:520–528. Freiman, R.N. and Tjian, R. 2003. Regulating the regulators: lysine modifications make their mark. Cell 112:11–17. Fu, M., Wang, C., Reutens, A.T., Wang, J., Angeletti, R.H., Siconolfi-Baez, L., Ogryzko, V., Avantaggiati, M.L. and Pestell, R.G. 2000. p300 and p300/cAMP-response element-binding protein-associated factor acetylate the androgen receptor at sites governing hormone-dependent transactivation. J. Biol. Chem. 275:20853–50860. Fu, M., Wang, C., Wang, J., Zhang, X., Sakamaki, T., Yeung, Y.G., Chang, C., Hopp, T., Fuqua, S. A., Jaffray, E., Hay, R.T., Palvimo, J.J., Ja¨nne, O.A. and Pestell, R.G. 2002. Androgen receptor acetylation governs trans activation and MEKK1-induced apoptosis without affecting in vitro sumoylation and trans-repression function. Mol. Cell. Biol. 22:3373–3388. Gaughan, L., Logan, I.R., Cook, S., Neal, D.E. and Robson, C.N. 2002. Tip60 and histone deacetylase 1 regulate androgen receptor activity through changes to the acetylation status of the receptor. J. Biol. Chem. 277:25904–25913. Gelmann, E.P. 2002. Molecular biology of the androgen receptor. J. Clin. Oncol. 20:3001–3015. Gobinet, J., Poujol, N. and Sultan, C. 2002. Molecular action of androgens. Mol. Cell. Endocrinol. 198:15–24. Gottlieb, B., Beitel, L.K., Lumbroso, R., Pinsky, L. and Trifiro, M. 1999. Update of the androgen receptor gene mutations database. Hum. Mutat. 14:103–114. Greenland, K.J. and Zajac, J.D. 2004. Kennedy’s disease: pathogenesis and clinical approaches. Intern. Med. J. 34:279–286. Haelens, A., Verrijdt, G., Callewaert, L., Chrsitiaens, V., Schauwaers, K., Peeters, B., Rombauts, W. and Claessens, F. 2003. DNA recognition by the androgen receptor: evidence for an alternative DNA-dependent dimerization, and an active role of sequences flanking the response element on transactivation. Biochem. J. 369:141–151. Haelens, A., Tanner, T., Denayer, S., Callewaert, L. and Claessens, F. 2007. The hinge region regulates DNA binding, nuclear translocation, and transactivation of the androgen receptor. Cancer Res. 67:4514–4523.

Multitasking and Interplay Between the Androgen Receptor Domains

401

Ham, J., Thomson, A., Needham, M., Webb, P. and Parker, M.G. 1988. Characterization of response elements for androgens, glucocorticoids and progestins in mouse mammary tumour virus. Nucleic Acids Res. 16:5263–5276. He, B., Kemppainen, J.A., Voegel, J.J., Gronemeyer, H. and Wilson, E.M. 1999. Activation function 2 in the human androgen receptor ligand binding domain mediates interdomain communication with the NH(2)-terminal domain. J. Biol. Chem. 274:37219–37225. He, B., Bai, S., Hnat, A.T., Kalman, R.I., Minges, J.T., Patterson, C. and Wilson, E.M. 2004a. An androgen receptor NH2-terminal conserved motif interacts with the COOH terminus of the Hsp70-interacting protein (CHIP). J. Biol. Chem. 279:30643–30653. He, B., Gampe, R.T., Kole, A.J., Mnat, A.T., Stanley, T.B., An, G., Stewart, E.L., Kalman, R.I., Minges, J.T. and Wilson, E.M. 2004b. Structural basis for androgen receptor interdomain and coactivator interactions suggests a transition in nuclear receptor activation function dominance. Mol. Cell 16:425–438. Heemers, H., Verrijdt, G., Organe, S., Claessens, F., Heyns, W., Verhoeven, G. and Swinnen, J.V. 2004. Identification of an androgen response element in intron 8 of the sterol regulatory element-binding protein cleavage-activating protein gene allowing direct regulation by the androgen receptor. J. Biol. Chem. 279:30880–30887. Heemers, H.V., Regan, K.M., Dehm, S.M. and Tindall, D.J. 2007. Androgen induction of the androgen receptor coactivator four and a half LIM domain protein-2: evidence for a role for serum response factor in prostate cancer. Cancer Res. 67:10592–10599. Heinlein, C.A. and Chang, C. 2002. Androgen receptor (AR) coregulators: an overview. Endocr. Rev. 23:175–200. Huang, Z.Q., Li, J., Sachs, L.M., Cole, P.A. and Wong, J. 2003. A role for cofactor-cofactor and cofactor-histone interactions in targeting p300, SWI/SNF and Mediator for transcription. EMBO J. 22:2146–2155. Iordanidou, P., Aggelidou, E., Demetriades, C. and Hadzopoulou-Cladaras, M. 2005. Distinct amino acid residues may be involved in coactivator and ligand interactions in hepatocyte nuclear factor-4a. J. Biol. Chem. 280:21810–21819. Irvine, R.A., Ma, H., Yu, M.C., Ross, R.K., Stallcup, M.R. and Coetzee, G.A. 2000. Inhibition of p160-mediated coactivation with increasing androgen receptor polyglutamine length. Hum. Mol. Genet. 9:267–274. Jenster, G., Trapman, J. and Brinkmann, A.O. 1993. Nuclear import of the human androgen receptor. Biochem. J. 293:761–768. Jenster, G., van der Korput, J.A., Trapman, J. and Brinkmann, A.O. 1995. Identification of two transcription activation units in the N-terminal domain of the human androgen receptor. J. Biol. Chem. 270:7341–7346. Jeong, B.C., Hong, C.Y., Chattopadhyay, S., Park, J.H., Gong, E.Y., Kim, H.J., Chun, S.Y. and Lee, K. 2004. Androgen receptor corepressor-19 kDa (ARR19), a leucine-rich protein that represses the transcriptional activity of androgen receptor through recruitment of histone deacetylase. Mol. Endocrinol. 18:13–25. Langley, E., Zhou, Z.X. and Wilson, E.M. 1995. Evidence for an anti-parallel orientation of the ligand-activated human androgen receptor dimer. J. Biol. Chem. 270:29983–29990. Lee, S.R., Ramos, S.M., Ko, A., Masiello, D., Swanson, K.D., Lu, M.L. and Balk, S.P. 2002. AR and ER interaction with a p21-activated kinase (PAK6). Mol. Endocrinol. 16:85–99. Lee, D.Y., Ianculescu, I., Purcell, D., Zhang, X., Cheng, X. and Stallcup, M.R. 2007. Surfacescanning mutational analysis of protein arginine methyltransferase 1: roles of specific amino acids in methyltransferase substrate specificity, oligomerization, and coactivator function. Mol. Endocrinol. 21:1381–1393. Lewis, B.A. and Reinberg, D. 2003. The mediator coactivator complex: functional and physical roles in transcriptional regulation. J. Cell. Sci. 116:3667–3675. Li, X.Y., Boudjelal, M., Xiao, J.H., Peng, Z.H., Asuru, A., Kang, S., Fisher, G.J. and Voorhees, J.J. 1999. 1,25-Dihydroxyvitamin D3 increases nuclear vitamin D3 receptors by blocking ubiquitin/proteasome-mediated degradation in human skin. Mol. Endocrinol. 13:1686–1694.

402

F. Claessens et al.

Li, J., Zhang, D., Fu, J., Huang, Z. and Wong, J. 2007. Structural and functional analysis of androgen receptor in chromatin. Mol. Endocrinol. Liao, G., Chen, L.Y., Zhang, A., Godavarthy, A., Xia, F., Ghosh, J.C., Li, H. and Chen, J.D. 2003. Regulation of androgen receptor activity by the nuclear corepressor SMRT. J. Biol. Chem. 278:5052–5061. Lin, H.K., Hu, Y.C., Lee, D.K. and Chang, C. 2004. Regulation of androgen receptor signaling by PTEN (phosphatase and tensin homolog deleted on chromosome 10) tumor suppressor through distinct mechanisms in prostate cancer cells. Mol. Endocrinol. 18:2409–2423. Link, K.A., Burd, C.J., Williams, E., Marshall, T., Rosson, G., Henry, E., Weissman, B. and Knudsen, K.E. 2005. BAF57 governs androgen receptor action and androgen-dependent proliferation through SWI/SNF. Mol. Cell. Biol. 25:2200–2215. Liu, G., Wang, H. and Wang, Z. 2003. Identification of a highly conserved domain in the androgen receptor that suppresses the DNA-binding domain-DNA interactions. J. Biol. Chem. 278:14956–14960. Luisi, B.F., Xu, W.X., Otwinowski, Z., Freedman, L.P., Yamamoto, K.R. and Sigler, P.B. 1991. Crystallographic analysis of the interaction of the glucocorticoid receptor with DNA. Nature 352(6335):497–505. Matias, P.M., Donner, P., Coelho, R., Thomaz, M., Peixoto, C., Mecedo, S., Otto, N., Joschko, S., Scholz, P., Wegg, A., Ba¨sler, S., Scha¨fer, M., Egner, U. and Carrondo, M.A. 2000. Structural evidence for ligand specificity in the binding domain of the human androgen receptor. J. Biol. Chem. 275:26164–26171. McEwan, I.J. 2004. Molecular mechanisms of androgen receptor-mediated gene regulation: structure-function analysis of the AF-1 domain. Endocr. Relat. Cancer 11:281–293. McEwan, I.J., Lavery, D., Fischer, K. and Watt, K. 2007. Natural disordered sequences in the amino terminal domain of nuclear receptors: lessons from the androgen and glucocorticoid receptors. Nucl. Recept. Signal. 5:e001. Metzger, E., Mu¨ller, J.M., Ferrari, S., Buettner, R. and Schu¨le, R. 2003 A novel inducible transactivation domain in the androgen receptor: implications for PRK in prostate cancer. EMBO J. 22:270–280. Nascimento, A.S., Dias, S.M.G., Nunes, F.M., Aparicio, R., Ambrosio, A.L.B., Bleicher, L., Figueira, A.C.M., Santos, M.A.M., Neto, M.O., Fischer, H., Togashi, M., Craievich, A.F., Garratt, R.C., Baxter, J.D., Webb, P. and Plikaripov, I. 2006. Structural rearrangements in the thyroid hormone receptor hinge domain and their putative role in the receptor function. J. Mol. Biol. 360:586–598. Poukka, H., Karvonen, U., Janne, O.A. and Palvimo, J.J. 2000a. Covalent modification of the androgen receptor by small ubiquitin-like modifier 1 (SUMO-1). Proc. Natl. Acad. Sci. USA 97:14145–14150. Poukka, H., Karvonen, U., Yoshikawa, N., Tanaka, H., Palvimo, J.J. and Ja¨nne, O.A. 2000b. The RING finger protein SNURF modulates nuclear trafficking of the androgen receptor. J. Cell. Sci. 113:2991–3001. Rastinejad, F., Wagner, T., Zhao, Q. and Khorasanizadeh, S. 2000. Structure of the RXRRAR DNA-binding complex on the retinoic acid response element DR1. EMBO J. 19:1045–1054. Rechsteiner, M. and Rogers, S.W. 1996. PEST sequences and regulation by proteolysis. Trends Biochem. Sci. 21:267–271. Rees, I., Lee, S., Kim, H. and Tsai, FTF. 2006. The E3 ubiquitin ligase CHIP binds the androgen receptor in a phosphorylation-dependent manner. Biochim. Biophys. Acta 1764:1073–1079. Sack, J.S., Kish, K.F., Wang, C., Attar, R.M., Kiefer, S.E., An, Y., Wu, G.Y., Scheffler, J.E., Salvati, M.E., Krystek, S.R., Weinmann, R. and Einsphar, H.M. 2001. Crystallographic structures of the ligand-binding domains of the androgen receptor and its T877A mutant complexed with the natural agonist dihydrotestosterone. Proc. Natl. Acad. Sci. USA. 98:4904–4909.

Multitasking and Interplay Between the Androgen Receptor Domains

403

Saporita, A.J., Zhang, Q., Navai, N., Dincer, Z., Hahn, J., Cai, X. and Wang, Z. 2003. Identification and characterization of a ligand-regulated nuclear export signal in androgen receptor. J. Biol. Chem. 278:41998–42005. Schoenmakers, E., Alen, P., Verrijdt, G., Peeters, B., Verhoeven, G., Rombauts, W. and Claessens, F. 1999. Differential DNA binding by the androgen and glucocorticoid receptors involves the second Zn-finger and C-terminal extension of the DNA-binding domains. Biochem. J. 341:515–521. Schoenmakers, E., Verrijdt, G., Peeters, B., Verhoeven, G., Rombauts, W. and Claessens, F. 2000. Differences in DNA binding characteristics of the androgen and glucocorticoid receptors can determine hormone-specific responses. J. Biol. Chem. 275:12290–12297. Shaffer, P., Jivan, A., Dollins, D.E., Claessens, F. and gewirth, D.T. 2004. Structural basis of androgen receptor binding to selective androgen response elements. Proc. Natl. Acad. Sci. 101:4758–4763. Sheflin, L., Keegan, B., Zhang, W. and Spaulding, S.W. 2000. Inhibiting proteasomes in human Hep3B and LNCaP cells increases endogenous androgen receptor levels. Biochem. Biophys. Res. Commun. 276:144–150. Shin, S. and Verma, I.M. 2003. BRCA2 cooperates with histone acetyltransferases in androgen receptor-mediated transcription. Proc. Natl. Acad. Sci. USA 100:7201–7206. Sleddens, H.F., Oostra, B.A., Brinkmann, A.O. and Trapman, J. 1992. Trinucleotide repeat polymorphism in the androgen receptor gene (AR). Nucleic Acids Res. 20:1427. Tanner, T., Claessens, F. and Haelens, A. 2004. The hinge region of the androgen receptor plays a role in proteasome-mediated transcriptional activation. Ann. NY Acad. Sci. 1030:587–592. Tilley, W.D., Buchanan, G., Hickey, T.E. and Bentel, J.M. 1996. Mutations in the androgen receptor gene are associated with progression of human prostate cancer to androgen independence. Clin. Cancer Res. 2:277–285. Van Royen, M.E., Cunha, S.M., Brink, M.C., Mattern, K.A., Nigg, A.L., Dubbink, H.J., Verschure, P.J., Trapman, J. and Houtsmuller, A. 2007. Compartmentalization of androgen receptor protein-protein interactions in living cells. J. Cell. Biol. 177:63–72. Veldscholte, J., Berrevoets, C.A., Zegers, N.D., van der Kwast, T.H., Grootegoed, J.A. and Mulder, E. 1992. Hormone-induced dissociation of the androgen receptor-heat-shock protein complex: use of a new monoclonal antibody to distinguish transformed from nontransformed receptors. Biochemistry 31:7422–7430. Verrijdt, G., Schoenmakers, E., Alen, P., Haelens, A., Peeters, B., Rombauts, W. and Claessens, F. 1999. Androgen specificity of a response unit upstream of the human secretory component gene is mediated by differential receptor binding to an essential androgen response element. Mol. Endocrinol. 13:1558–1570. Verrijdt, G., Schoenmakers, E., Haelens, A., Peeters, B., Verhoeven, G., Rombauts, W. and Claessens, F. 2000. Change of specificity mutations in androgen-selective enhancers: evidence for a role of differential DNA binding by the androgen receptor. J. Biol. Chem. 275:12298– 12305. Verrijdt, G., Tanner, T., Moehren, U., Callewaert, L., Haelens, A. and Claessens, F. 2006. Biochem. Soc. Trans. 34:1089–1094. Vijayvargia, R., May, M.S. and Fondell, J.D. 2007. A coregulatory role for the mediator complex in prostate cancer cell proliferation and gene expression. Cancer Res. 67:4034–4041. Wang, Q., Lu, J.H. and Yong, E.L. 2001. Ligand- and coactivator-mediated transactivation function 2 (AF2) of the androgen receptor ligand-binding domain is inhibited by the cognate hinge region. J. Biol. Chem. 276:7493–7499. Wang, Q., Li, W., Liu, X.S., Carroll, J.S., Ja¨nne, O.A., Keeton, E.K., Chinnaiyan, A.M., Pienta, K. J. and Brown, M. 2007. A hierarchical network of transcription factors governs androgen receptor-dependent prostate cancer growth. Mol. Cell 27:380–392. Werner, R., Holterhus, P.M., Binder, G., Schwartz, H.P., Morlot, M., Struve, D., Marschke, C. and Hiort, O. 2006. The A645D mutation in the hinge region of the human androgen receptor (AR)

404

F. Claessens et al.

gene modulates AR activity, depending on the context of the polymorphic glutamine and glycine repeats. J. Clin. Endocrinol. Metab. 91:3515–3520. Xu, P., Liu, Y., Shan, S., Kong, Y., Zhou, Q., Li, M., Ding, J., Xie, Y. and Wang, Y. 2004. Molecular mechanism for the potentiation of the transcriptional activity of human liver receptor homolog 1 by steroid receptor coactivator-1. Mol. Endocrinol. 18:1887–1905. Zhou, Z.X., Sar, M., Simental, J.A., Lane, M.V. and Wilson, E.M. 1994. A ligand-dependent bipartite nuclear targeting signal in the human androgen receptor. Requirement for the DNAbinding domain and modulation by NH2-terminal and carboxyl-terminal sequences. J. Biol. Chem. 269:13115–13123. Zhou, Z.X., Kemppainen, J.A. and Wilson, E.M. 1995. Identification of three proline-directed phosphorylation sites in the human androgen receptor. Mol. Endocrinol. 9:605 615. Zhu, P., Baek, S.H., Bourk, E.M., Garcia-Bassets, I., Sanjo, H., Akira, S., Kotol, P.F., Glass, C.K., Rosenfeld, R.M. and Rose, D.W. 2006. Macrophage/cancer cell interactions mediate hormone resistance by a nuclear receptor derepression pathway. Cell 124:615–629. Zuccarello, D., Ferlin, A., Vinanzi, C., Prana, E., Callewaert, L., Claessens, F., Brinkmann, O.A. and Foresta, C. 2007. Detailed functional studies on androgen receptor mild mutations demonstrate their association with male infertility. Clin. Endocrinol. 68:580–588.

Chromatin Remodeling and Androgen Receptor-Mediated Transcription Li Jia, Omar Khalid, Baruch Frenkel, and Gerhard A. Coetzee

Abstract It has become apparent that the expression of human genes in chromatin is regulated by post-translational structural changes in histones, which form the major protein component of nucleosomes in chromatin. The process is generally referred to as chromatin epigenetics. Recently, it was demonstrated that histone amino-terminal tails, which extend from the core of nucleosomes out of chromatin, are methylated or acetylated at lysine residues with profound effects on gene structure and function. Since some of these changes are inherited from cells to daughter cells, lineages are established with stable histone modifications. In this way the regulation of androgen receptor-mediated transcription of target genes and the phenotype of androgen receptor-mediated prostate cancer progression are affected. The detail of this novel level of regulation is being pursued by many investigators and is summarized in this chapter.

1 Chromatin Epigenetics Relative to classic genetics, epigenetics is a novel concept that describes the inheritance of genetic information not encoded in the primary DNA sequence. Conceptually, the molecular/mechanistic biology of epigenetics can be divided into two major groups, covalent modifications of histones and methylation of CpG dinucleotides of DNA. Histone modifications, along with regulatory proteins that bind to them, affect nucleosome structure and positioning in chromatin, and consequently gene expression. We refer to this type as ‘‘chromatin epigenetics.’’ The other arm of epigenetics, DNA methylation, is well studied, especially its role in cancer (Jones and Baylin 2007). The two aspects of epigenetics – chromatin modifications and DNA methylation – are related, because one may affect the other.

G.A. Coetzee(*) Departments of Preventive Medicine & Urology and Norris Cancer Center, Keck School of Medicine, University of Southern California, Los Angeles, CA, USA, E-mail: [email protected]

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4_18, # Springer Science + Business Media, LLC 2009

405

406

L. Jia et al.

Here, however, we will concentrate on chromatin epigenetics as it relates to prostate cancer (PCa) progression and androgen receptor (AR) signaling. Just as is the case with DNA methylation, chromatin epigenetics is intimately involved in normal and abnormal somatic cell development, such as cancer progression (Feinberg 2007a, b). Histone modifications or ‘‘marks’’ lead to the recruitment and docking of protein complexes that regulate transcription via effects on chromatin substructure. Although the modifications were initially considered to be a simple ‘‘code’’ regulating transcription and other processes, newer data indicate the existence of a nuanced chromatin ‘‘language’’ that demarcates chromatin into structural domains with dynamic functional consequences (Berger 2007). This language as it relates to PCa will be the main topic of this chapter. The molecular understanding of how chromatin remodeling modulates transcription and affects PCa development will undoubtedly have a major impact in the diagnosis, management, and treatment of the disease. This is because the initiation of the disease and its progression to castration-recurrent cancer depend to a large degree on the activity of the AR, a potent transcription factor that mediates expression control at many target loci across the entire human genome, which in turn may be modulated by chromatin epigenetics.

2 AR Transcriptional Mechanisms The AR structure and function can be defined in terms of four distinct domains: a large N-terminal domain (NTD) containing strong constitutive transactivation activity; a DNA-binding domain (DBD) mediating DNA binding to androgenresponse elements (AREs); a so-called hinge region involved in AR stability and chaperone interaction; and a C-terminal ligand-binding domain (LBD) that mediates receptor activation by androgens as well as a relatively weak, but highly conserved, ligand-dependent transactivation function (reviewed in Buchanan et al. 2001; Tanner et al. 2004). The normal functions of the AR require dihydrotestosterone (DHT) binding to the LBD, AR dimerization and phosphorylation at selected sites, translocation to the nucleus, and binding to AREs, resulting in transcriptional activation (or inhibition) of target genes. However, it is thought that AR activity is also modulated via other signals, in particular during the recurrence of PCa after failing androgen ablation therapy. The AR-NTD is an attractive target for the action of nonsteroidal modulators due to its strong transactivation potential, although its activity in isolation within chromatin seems to be compromised (Li et al. 2009). Very little is known about the molecular detail of this AR domain or precisely how it functions during both DHT and nonsteroidal activation of the receptor. Clues to its importance come from earlier work demonstrating its strong interaction with coactivators, such as the p160 group, previously thought to interact exclusively with the AR-LBD (Alen et al. 1999; Ma et al. 1999; Irvine et al. 2000; Shen et al. 2005) and the idea that it forms a relatively unstructured nonfolded domain that undergoes strong induced-fit structural

Chromatin Remodeling and Androgen Receptor-Mediated Transcription

407

molding during its binding to protein factors (Kumar and Thompson 2003). Many cofactors with histone-modifying activities are thus recruited to the AR-NTD, and the domain may be viewed as a sticky flycatcher of proteins that affect AR activity by modifying chromatin epigenetics. The following represent a sequence of molecular events that underlie the accepted process of AR-mediated transactivation activity. 1. In the absence of native ligand (DHT) the AR is associated in the cytoplasm with a multiprotein chaperone complex, which is essential for receptor maturation and the acquisition of ligand-binding competence (Pratt and Toft 1997). Several cochaperones modulate the structures and functions of such complexes (Buchanan et al. 2007; Zoubeidi et al. 2007). 2. Following hormone binding, the complex dissociates and the receptor is rapidly translocated into the nucleus (Tyagi et al. 2000). Androgen-induced activation involves an N- and C-terminal interaction (N/C interaction), which could be due to either intermolecular or, more likely, intramolecular interactions (He et al. 1999; Klokk et al. 2007). 3. In the nucleus, the AR dimer binds in the major groove of the DNA double helix at specific androgen response elements and collaborates with other transcription factors in a hierarchical network as elaborated later (Wang et al. 2007). 4. Subsequent transactivation depends on the relative abundance of many other specific cofactors in a given cell type (Rosenfeld and Glass 2001) and transcriptional collaborators (Wang et al. 2007). Coactivators may themselves be or recruit histone acetyltransferases and methyltransferases, resulting in chromatin decondensation (activation). Corepressors recruit histone deacetylases resulting in condensation (inhibition). Sustained AR occupancy and activity at certain loci may lead to dramatic histone alterations across relatively large chromatin domains (Jia et al. 2006). Many AR-occupied loci may remain poised for transcription depending on developmental and other cues (see later). 5. Following ligand dissociation, the AR undergoes dissociation from target promoters, followed by recycling between the cytoplasm and nucleus (Tyagi et al. 2000).

3 Histone Modifications and Gene Expression The genomic mapping of regions occupied by transcription factors (TFs), including nuclear receptors, as well as genome-scale mapping of histone modifications has undergone a revolution over the past few years, mainly due to the development of high-throughput mapping technologies such as ChIP-on-chip (ChIP–chip) and ChIP sequencing (ChIP-seq). In principle, crosslinked chromatin (protein–protein and protein–DNA) complexes are extracted from tissues or cells and sheared, typically by sonication, down to relatively short (200–1,000 bp) fragments. Antibodies to the transcription factors or modified histones are then added to immunoprecipitate the

408

L. Jia et al.

protein–DNA complexes. The DNA regions isolated by this procedure are then identified by hybridization to DNA microarrays or by mass sequencing. Unbiased ChIP–chip studies in mammalian cells have been undertaken for NF-kB (Martone et al. 2003), cMyc (Cawley et al. 2004), SP1 (Cawley et al. 2004), p53 (Cawley et al. 2004; Wei et al. 2006), and CREB (Euskirchen et al. 2004). Interestingly, more binding sites for these TFs were found than expected, and they were not restricted to annotated upstream promoter regions. With respect to nuclear receptors, Brown and colleagues mapped the association of the estrogen receptor (ER) with the complete nonrepetitive sequence of the entire human genome in breast cancer cells (Carroll et al. 2006) and the AR to sites in chromosomes 21 and 22 in PCa cells (Wang et al. 2007). ChIP–chip methods targeting (and restricted to) gene promoters and surrounding sequences for AR responsive genes identified the expected sites along with transcriptional collaborators (Bolton et al. 2007; Massie et al. 2007). Noncanonical response elements for AR may exist in vivo, since in vitro-defined elements occurred in only about 60% of AR-binding sites (Bolton et al. 2007). More recently a new technique, the so-called ChIP-seq, has been employed to identify histone modifications and transcription-factor-occupied areas (Barski et al. 2007; Mikkelsen et al. 2007; Robertson et al. 2007). ChIP-seq, currently employing the Solexa mass sequencing technology, seems to be the method of choice, and although it requires a significant investment in instrumentation, in the long run it is highly cost effective. It is definitely more accurate, quantifiable and requires less amplification of the starting material and thus introduces less distortion during the processing of the original DNA. As stated earlier, AR-regulated gene expression is mediated by recruitment of coregulatory complexes in a ligand-dependent and sometimes also in a ligandindependent manner. Coregulatory complexes formed on DNA exhibit diverse enzymatic activities and can be divided into two generic classes: enzymes capable of remodeling chromatin structure using ATP (see Sect. 5) and enzymes capable of covalently modifying histone tails. Although roles for most of these modifications remain ill defined, recent studies revealed that acetylated (ac) and methylated (me) lysines in histone tails serve as docking sites for transcriptional coregulators (Bernstein et al. 2007). Figure 1 summarizes the significance of several key histone marks. While lysine acetylation almost always correlates with chromatin accessibility and transcriptional activity, lysine methylation can have different effects depending on which residue is modified. While methylation of histone H3 lysine 4 (H3K4me) is associated with active chromatin, methylation of H3 lysine 9 (H3K9me), H3 lysine 27 (H3K27me), and H4 lysine 20 (H4K20me) generally correlates with repression (Strahl and Allis 2000; Bernstein et al. 2007). On the other hand, H3 lysine 36 (H3K36me) marks transcriptional elongation (Bannister et al. 2005). Additionally, histone lysine residues can be mono (me1)-, di-(me2), or trimethylated (me3), thereby extending the coding potential and complexity. In fact, several general concepts affecting chromatin epigenetics have recently emerged and may be listed as:

Chromatin Remodeling and Androgen Receptor-Mediated Transcription

409

Fig. 1 Histone H3 and 4 amino-terminal tail modifications that mark expression (or repression) of genes. Modified from Cell Snap-shot collection of posters

410

L. Jia et al.

1. It was found that in only a small subset of genes where transcriptional initiation occurred, productive mRNA synthesis was completed (Guenther et al. 2007). Productive synthesis depended on histone modifications that mark transcriptional elongation, such as H3K36me3. Thus, most protein-coding genes, even blatantly inactive ones, experience the hallmarks of transcriptional initiation. Nucleosomes with H3K4me3 and acetylation of H3 lysines 9 and 14 (H3K9,14ac) as well as RNA polII occupancy were found at 75% of protein-coding genes in embryonic stem cells, while only about 50% of them yielded detectable transcripts. A concept arose of abundant loci where RNA polII is poised for productive RNA synthesis by the assembly of a preinitiation complex at gene promoters followed by initiation of RNA synthesis and subsequent RNA polII stalling. This phenomenon seems to be widespread, occurring at hundreds of genes (at least in the fruit-fly), apparently ready to respond to stimuli such as developmental and environmental cues (Muse et al. 2007). The disparity between initiation and elongation was also evident from in vivo dynamic studies of RNA polymerase II transcriptional elongation (Darzacq et al. 2007). These results imply that transcriptional elongation control, such as mediated by chromatin structure, may be one of the main regulatory steps in gene expression (Jia et al. 2006; Li et al. 2007). 2. Systematic studies of chromatin modifications have revealed a complex landscape including ‘‘punctate’’ sites of modified histones (H3K9,14ac, and H3K4me) at transcription start sites and distal regulatory elements, and broad domains at gene clusters and developmental loci (Bernstein et al. 2005). Insights into the potential functions of specific modified histones were revealed by the observation that H3K4me3 attracts the general transcription factor TFIID (Vermeulen et al. 2007), thus regulating RNA polymerase II-mediated transcription and linking in a direct fashion histone modification with gene expression. 3. Heintzman et al. (2007) proposed that H3K4me may be used not only to mark active chromatin, but also to distinguish enhancers from promoters; promoters are marked by trimethylation of H3K4 (H3K4me3), whereas enhancers are marked by monomethylation of H3K4 (H3K4me1), but not trimethylation of H3K4 (H3K4me3). We have recently addressed this issue in relation to the development of androgen-ablation resistance in PCa (see later). 4. The linear linking of promoters with enhancers on DNA has been challenged, since it was found that single enhancers may interact with several gene promoters at a distance and in some cases even across chromosomes (Savarese and Grosschedl 2006). Such mechanisms may coordinate networks of gene expression via the interaction with a single enhancer ‘‘hub.’’ The regulation of such hubs is unknown at present, but conceivably would need sophisticated chromatin structural and functional domains to avoid ‘‘traffic jams.’’ One way to prevent this from happening is to demarcate relatively large chromatin domains by insulator structures mediated by CTCF binding (Kim et al. 2007; Xie et al. 2007). Such chromatin domains demonstrably result in the containment of specific histone marks and associated gene expression control to specific areas of the genome.

Chromatin Remodeling and Androgen Receptor-Mediated Transcription

411

5. Interestingly, bivalent chromatin structures, containing histone marks for both gene activation (H3K4me) and repression (H3K27me), were found at many loci in mouse embryonic stem cells (Bernstein et al. 2006). The authors proposed that such bivalent domains may silence developmental genes in stem cells while at the same time keeping them poised for activation, and in this way maintain cells in a pluripotent state. Lineage-specific cell fate decisions are then associated with genes with bivalent chromatin marks that become either active or permanently silenced. Therefore, histone modifications, such as the combination of, or separately H3K4,27me3, discriminate genes that are expressed, poised for expression, or stably repressed reflecting a cell’s lineage potential (Mikkelsen et al. 2007), and thus orchestrate gene expression, not only during development, but also perhaps cell proliferation in a lineage-dependent way, such as during carcinogenesis (see later). 6. As stated earlier H3K36me3 marks transcriptional elongation and is found across the entire gene body beginning immediately after the promoter and in some cases includes noncoding transcripts such as miRNAs (Mikkelsen et al. 2007). In the same study H3K9me3 and H4K20me3 were detected at satellite, telomeric, and active long terminal repeats. H3K4,9me3 marked imprinted control regions. The authors suggest that these results provide a framework to a priori identify and characterize diverse cell populations by their histone language. 7. Part of the histone language necessary for transcriptional control at selected loci is the dissociation of the linker histone H1. Histone H2A ubiquitination coordinates this dissociation and is augmented by histone hyperacetylation (Zhu et al. 2007). Histone deubiquitination is mediated by an enzyme complex that contains 2A-DUB (histone deubiquitinase activity) and p/CAF (histone acetylase activity). It is likely that H2A ubiquitin ligases/deubiquitinases are important regulatory gene expression complexes at specific gene loci mediating cohorts of regulatory transcription units. 8. Interestingly, a multienzyme complex exists that can both demethylate H3K27me3 (i.e., UTX/JMJD3) and methylate H3K4 (i.e., MLL) resulting in the unlocking of gene expression during reprogramming during development (Rivenbark and Strahl 2007). How do histone modifications relate to chromatin structural domains and gene expression on a global scale? Historically chromatin structure has been divided into euchromatin (active) and heterochromatin (silent). More recently a concept of facultative heterochromatin has emerged (Trojer and Reinberg 2007). Facultative heterochromatin is thought to be silent but readily converted to active euchromatin and distinguishable from constitutive heterochromatin, which remains silent and of which the silent X-chromosome is the prime example. Reversible H3K27me3 is a good example of a histone modification that marks facultative heterochromatin within genes that are silent, yet poised for transcriptional activation.

412

L. Jia et al.

The control of differential gene expression is often governed by the assembly and activity of different combinations of transcription factors on specific DNA sites (Remenyi et al. 2004). Both FoxA1 and C/EBP had been implicated previously in ER-mediated gene expression (Carroll et al. 2006). The Matusik laboratory further showed that FoxA proteins are expressed in the prostate and interact with the AR, thus modulating its activity on some but not all promoters (Yu et al. 2005). With respect to AR collaborators, the Brown lab has identified FoxA1, GATA2, and Oct1 as potential AR collaborators by analyzing AR-occupied regions on chromosomes 21 and 22 (Wang et al. 2007). Collaborators, such as FoxA1 and GATA family members, can ‘‘open’’ compact chromatin to allow AR occupancy, thus acting as pioneer transcription factors mediating AR binding in a hierarchical fashion. They also are able to bind AR directly thus stabilizing a larger transcriptional regulatory complex (Wang et al. 2007). Other AR collaborators include AP-1, RAR, ZNF42, HNF-4a, and EGR (Bolton et al. 2007) as well as ETS1 (Massie et al. 2007). Noteworthy, ETS1 is recruited to AR target genes in response to DHT administration and thus unlikely to function as a ‘‘pioneer’’ factor. However, since AR-mediated transcription of specific ETS transcription factors occurs as a consequence of TMPRSS2-ETS fusions in a significant proportion of prostate tumors, the fact that ETS factors are themselves also AR collaborators on a subgroup of additional target genes may indicate a potent positively reinforcement feedback system that drives advanced PCa in an AR-dependent fashion.

4 Histone-Modifying Enzymes Enzymes that covalently modify histones by adding or removing marks such as those depicted in Fig. 1 are the driving force for encrypting the chromatin epigenetic fingerprint of cell lineages as well as the clonal expansions of cancer (see later), and ultimately may determine, to a large degree, the outcome of AR signaling and PCa progression. During the last decade a large number of histonemodifying enzymes have been identified and characterized (reviewed in Kozlowski et al. 1991; Couture and Trievel 2006; Kouzarides 2007). A unifying nomenclature was recently proposed to restore order in the many different names for the same enzyme (Allis et al. 2007). Histones undergo many post-translational modifications causing significant changes to chromatin structure and function. The modifications include acetylation, methylation, phosphorylation, sumoylation, ubiquitination, and ADP-ribosylation, which all occur most often in the flexible N- and sometimes also in the C-terminal tails of histones. The modifications working individually or in combination may additionally affect DNA methylation culminating in modifying an entire epigenetic signature. In addition to the primary DNA sequence, an underlying basis for the regulation of chromatin structure is likely the control and activity of the enzymes that add and remove the covalent modifications. Although more information is known about enzymes that add acetyl and methyl groups to histones (Couture and Trievel 2006) than the enzymes involved in the removal of

Chromatin Remodeling and Androgen Receptor-Mediated Transcription

413

these groups (Holbert and Marmorstein 2005), the understanding of their combinatorial controls and relationships to chromatin-modulated gene expression is still at a relatively primitive stage. This is despite an emerging role of histone lysine demethylases in disease (Shi 2007). Only specific, often anecdotal, information is available, and no genome-wide comprehensive hypotheses as to how these enzymes coordinate their functions have emerged. Unexpectedly, oncogenic transcription factors may work by recruiting inhibitors of transcription. For example, it was shown that a full complement of chromatin machinery is involved including the inhibitory polycomb group of proteins to enable certain types of leukemia cells (acute promyelocytic leukemia) to proliferate and emerge in certain patients (Villa et al. 2007). Within the context of prostate cancer (PCa), overexpression of the polycomb group protein EZ-H2 (new name, KMT6), an H3K27 methyltransferase, was found in metastatic PCa, which presumably leads to widespread transcriptional repression (Varambally et al. 2002). KMT6 was originally identified as a transcriptional repressor and involved in PCa (Sellers and Loda 2002; Varambally et al. 2002). Thus, the polycomb group proteins play a central role in gene silencing and may functionally link stem cells, metastasis, and cancer survival. On the other hand, histone lysine demethylases (LSD1 and JHDM1, new names KDM1 and KDM2, respectively), are clearly involved in androgen receptor (AR)-mediated transcription by being recruited to the transcriptional complex and mediating methyl groups removal that are inhibitory to transcription (Kahl et al. 2006; Metzger et al. 2006). JHDM2A (new name, KDM3A), which specifically demethylates mono- and dimethyl-H3K9, is recruited to AR target genes by androgen treatment and results in H3K9 demethylation and subsequent transcriptional activation (Yamane et al. 2006). More recently it was shown that the JmjC-domain protein, Jmjd3 (new name, KDM6B) is an H3K27 demethylase and is expressed in macrophages in response to bacterial products and inflammatory cytokines (Swigut and Wysocka 2007), thereby inhibiting polycomb-mediated gene silencing. Enzymes that add or remove histone marks (Couture and Trievel 2006) may consequently play vital roles in health and disease and are increasingly becoming targets of pharmacological intervention to affect disease outcome.

5 ATP-Mediated Remodeling Apart from the enzymatic modification of histones, as elaborated earlier, a parallel system evolved to allow the lateral movement of nucleosomes to facilitate efficient transcription. In many studies histone depletion is observed at exactly the transcription start sites of genes (Liang et al. 2004; Heintzman et al. 2007). Thus, ChIP–chip results of H3Ac clearly show a dip in the peak of H3Ac (Fig. 2), indicating significant histone depletion at the initiation of transcription. The movement of histones requires the hydrolysis of ATP mediated via a multiprotein complex called SWI/SNF shown to be involved in proliferation and differentiation of mammalian cells (Seo and Kroll 2006). The SWI/SNF complex contains at least eight proteins,

414

L. Jia et al.

Fig. 2 Histone H3Ac ChIP–chip in MCF7 cells indicating the dip in the H3Ac peak at the transcription start site of a gene

two of which, BRG1 and hBRM, have ATPase activities. The complex is involved in regulating the actions of transcription factors by facilitating nucleosome condensation (repression) or dispersion (activation). Ligand-dependent activation of ARand ER target genes requires SWI/SNF activity in a promoter/enhancer-dependent manner (Marshall et al. 2003). For example, whereas AR-mediated PSA expression requires SWI/SNF for activity, the probasin promoter, another AR target locus, maintains a low level of activation even in the absence of SWI/SNF. Furthermore, the inclusion of the known PSA enhancer in the reporter construct bypasses the SWI/SNF requirement, indicating that different DNA sequences containing different cis-acting elements can impact SWI/SNF influence on gene expression. Precisely how these ATP-dependent remodeling factors interplay with enzyme-mediated chromatin covalent modifications in gene expression remains to be determined.

6 Abnormal AR-Mediated Transcription Causes Ablation-Resistant PCa Treatment of advanced prostate tumors generally targets components of the AR signaling axis (i.e., by androgen ablation), inhibiting tumor growth. However, despite an initial positive response, disease progression eventually ensues (Kozlowski et al. 1991). Resistance to androgen ablation is not necessarily due to a loss of androgen sensitivity but develops as a consequence of a deregulated androgen signaling axis resulting from AR gene mutations or amplification, altered interactions of coregulatory molecules during transcription, nonsteroidal activation of AR by cytokines and growth factors (reviewed in Jenster 1999; Buchanan et al. 2001; Feldman and Feldman 2001; Balk 2002; Debes and Tindall 2004) or is due to ‘‘open’’ chromatin at particular loci as measured by histone modifications, which in turn leads to overexpression of AR target genes (Jia et al. 2006). Aberrant AR signaling alone may cause PCa as demonstrated by expressing a mutant AR in the prostate of transgenic mice (Han et al. 2005). The mutant AR had

Chromatin Remodeling and Androgen Receptor-Mediated Transcription

415

higher-than-normal transcriptional activity in the absence of ligand and had increased sensitivity to coregulators. Mice carrying this mutant, but not a wildtype AR transgene, subsequently developed PCa in 100% of cases. Additionally, the Sawyers’ laboratory (Chen et al. 2004) reported that 2- to 5-fold increases in AR expression were consistently associated with the development of cells that had become supersensitive to androgen, resulting in an ablation-resistant phenotype. Furthermore, reactivation of the androgen responsive pathway in ablation-resistant human tumors was also the major finding in another expression array study (Holzbeierlein et al. 2004). Even an unbiased network biology approach, in which reverse-engineered gene networks were combined with expression profiles, identified the AR among the top genetic mediators and the AR signaling pathway as highly enriched in metastatic PCa (Ergun et al. 2007). All the data are in line with previous reports that identified amplification of the AR gene in 22% of prostate cancer metastases (Bubendorf et al. 1999) and in 23–28% of primary tumors following androgen deprivation (Koivisto et al. 1997). A twofold increase in both AR and PSA expression has been reported in prostate tumor samples, which contained AR gene amplification compared to samples where no AR amplification was found (Koivisto and Helin 1999; Linja et al. 2001). In fact, the percentage of AR nuclear area (by quantitative IHC staining) in malignant epithelial nuclei predicted nonorgan confined recurrence of PCa in a recent study (the higher the percentage, the poorer the outcome) (Ricciardelli et al. 2005). In another study, AR protein expression was significantly higher in PCa tumors of African-American men compared to white men (Gaston et al. 2003). This result, the authors believe, contributes to why cancer occurs at a younger age and more rapidly in AfricanAmerican men. We have demonstrated an average shorter CAG repeat length of the AR in African-American men potentially contributing to their increased risk for PCa (Irvine et al. 1995) due to a more active AR (Irvine et al. 2000; Buchanan et al. 2004). Work from the Wilson laboratory (Gregory et al. 2001) demonstrated that in cell models of recurrent PCa (CWR-R1 and C4-2) the AR is more stable, leading to increased steady state expression of the receptor. This phenotype was associated with hypersensitivity of cell proliferation to extremely low levels of DHT. Chromatin epigenetics have the potential to augment AR signaling by affecting the efficiency of transcription from chromatin templates. We have shown that locus-wide chromatin remodeling and enhanced ARmediated transcription occurred at certain loci in ablation-resistant PCa (Jia et al. 2006). We proposed that the nonabortive transcription at such loci would lead to increased AR sensitivity and participation in the ablation-resistant phenotype. Interestingly, the maintenance of the modified loci required the AR. Removal of AR, as accomplished through experimental reduction of AR protein levels, will break the link between the AR and modified chromatin, and will allow the gene locus to return to its pre-stimulated state. This mechanistic relationship implies that treatment strategies for ablation-resistant PCa, which currently focus on targeting the AR, will be more effective than perhaps previously realized, since targeting the AR will also desensitize AR-regulated gene loci and consequently the entire cancer phenotype. Such treatment strategies are presently being contemplated (Scher et al.

416

L. Jia et al.

2004) and, as more information about the role of chromatin in PCa becomes available, adjuvant treatment that additionally targets specific histone modifications may have a much better chance to effectively treat advanced PCa. Recently, the Ren laboratory (Heintzman et al. 2007) proposed that histone H3 lysine 4 (H3K4) methylation may be used to distinguish enhancers from promoters – enhancers were preferentially mono-methylated, whereas promoters were preferentially tri-methylated. Although this is an interesting and novel concept, and probably applies to many promoters and enhancers, exceptions exist. We have recorded mono-, di-, and tri-methylation levels of histone H3K4 at several sites in the PSA locus during PCa conversion to ablation resistance and made several interesting observations (Fig. 3). We assessed histone modifications at the PSA locus in androgen-dependent LNCaP and its ablation-resistant derivative C4-2B cell line. Levels of mono-, di-, and tri-methylation of histone H3K4 at the enhancer, the promoter, and gene body (introns 1 and 2) in the two cell lines were analyzed in relation to the expression of PSA, as modulated by DHT (Fig. 3). (Note: The relative modified methylation levels between mono-, di-, and tri-methylation cannot be compared, since the efficiency of the antibodies used to precipitate them may not be the same). All three types of H3K4 methylation levels were detected at the enhancer and promoter in both cell lines regardless of DHT stimulation. Di-and trimethylation levels (but not mono-methylation levels) in the gene body increased up to threefold in C4-2B versus LNCaP cells regardless of short-term hormone treatment. Such modified chromatin regions (domains), as found in the gene body, may provide a robust epigenetic memory to maintain expression of genes in a lineage-specific manner regardless of short-term changes. Their relatively large size across linear DNA ensures that each daughter chromosome inherits a significant proportion of the modified histones, which in turn acts as a seed to promote similar modifications of the newly assembled neighboring histones. We conclude that chromatin domains are epigenetically selected during androgen-dependent to ablation-resistant conversion of PCa cells, resulting in efficient AR-mediated gene expression and cell growth. In contrast to the prediction (Heintzman et al. 2007), H3K4 methylation may not distinguish all enhancers from promoters, but does seem to mark lineage specifically the gene body of PSA. Such signatures across chromatin domains may additionally indicate epigenetic, and thus inherited marks, for efficient transcription. We further tested whether histone modification marks spatially coexist with the AR in the nucleus of C4-2B cells. As shown by immunofluorescence (Fig. 4) H3ac, H3K4me2, H3K4me3, H3K27me3, and H4K20me2 were diffused throughout the nucleus. In contrast, H3K4me1 appeared to be enriched toward the nuclear periphery. While the significance of the unique nuclear sublocalization of H3K4me1 is interesting (Kosak and Groudine 2004), the important point is that all histone modification marks tested partially colocalized with the AR apparently in different nuclear locations. All evidence indicates that AR-mediated transcriptional control is qualitatively different in different cells under different conditions. That is, different genes are affected under different conditions. It is well known that the endpoints of AR

Chromatin Remodeling and Androgen Receptor-Mediated Transcription

417

Fig. 3 A comparison between PSA expression of LNCaP and C4-2B as well as histone H3K4 methylation levels at the PSA locus

418

L. Jia et al.

Fig. 4 Global localization of AR and histones with specific modification. DHT-treated C4-2B cells were probed with antibodies against the AR and the histone modification marks listed to the left. Co-localization of the AR and sites with each histone modification are demonstrated in the merged images. (See Color Insert)

signaling change, from targeting differentiation markers (such as PSA or hK2 (Murtha et al. 1993)) during normal prostate development, to genes that control proliferation during the carcinogenic process. Such critical ‘‘cancer genes’’ are able to augment cell growth via cell cycle control and are therefore considered important downstream targets of the AR in the context of tumor growth and progression

Chromatin Remodeling and Androgen Receptor-Mediated Transcription

419

(reviewed in Balk 2002). For example, it has been established that AR signaling in prostate epithelial cells, even immortalized ones, resulted in cell differentiation and growth inhibition via the upregulation of cytokeratins, PSA, p21, and p27, and downregulation of c-myc, bcl-2, and telomerase (Ling et al. 2001). On the other hand, AR signaling in PCa cell lines, such as LNCaP, stimulates cell proliferation (Schuurmans et al. 1991), and anti-androgens have well-established apoptotic effects on androgen-dependent PCa (Catz and Johnson 2003). A comprehensive catalogue of AR-mediated target gene expression under different physiological and pathological conditions is lacking and is needed to make sense of disease initiation and progression. The past decade has seen significant and important development of technologies to allow the in detail dissection of gene regulation. This has resulted in the emergence of promising new concepts to understand and integrate the regulatory networks as it pertains to AR-mediated processes.

7 Conclusions Steroid hormone receptor-mediated signaling intersects at the chromatin level with many cofactors and other transcriptional collaborators, resulting in significant histone alterations that allow productive transcription. Depending on time-and cell-specific cues, this differs qualitatively and quantitatively from locus to locus. In this manner a nuanced genomic response to such cues can be integrated to initiate and sustain a particular phenotype. In the case of PCa initiation and progression the AR seems to be pivotal, although it is important to realize that many other molecular players need to be considered. The fact that these players are being uncovered, bodes well for an in-depth understanding of the mechanisms governing the disease process. Armed with such knowledge, therapeutic agents can be contemplated that may have selective anti-PCa actions. Acknowledgments We thank Yankel Gabet for critically reviewing the manuscript. We acknowledge the following grant support: W81XWH-04-1-0823 and W81XWH-07-1-0067 to GAC, and W81XWH-05-1-0025 to BF from the US Department of Defense; CA109147 to GAC and DK071122 to BF from the National Institutes of Health; IRG-58-007-48 to LJ from the American Cancer Society; Awards from the Prostate Cancer Foundation to GAC; The J. Harold and Edna L. LaBriola Chair in Genetic Orthopaedic Research, held by BF.

References Alen, P., Claessens, F., Verhoeven, G., Rombauts, W., and Peeters, B. (1999). The androgen receptor amino-terminal domain plays a key role in p160 coactivator-stimulated gene transcription. Mol. Cell. Biol. 19:6085–6097. Allis, C. D., Berger, S. L., Cote, J., Dent, S., Jenuwien, T., Kouzarides, T., Pillus, L., Reinberg, D., Shi, Y., Shiekhatter, R., Shilatifard, A., Workman, J., and Zhang, Y. (2007). New nomenclature for chromatin-modifying enzymes. Cell 131:633–636.

420

L. Jia et al.

Balk, S. (2002). Androgen receptor as a target in androgen-independent prostate cancer. Urology 60:132–138. Bannister, A. J., Schneider, R., Myers, F. A., Thorne, A. W., Crane-Robinson, C., and Kouzarides, T. (2005). Spatial distribution of di- and tri-methyl lysine 36 of histone H3 at active genes. J. Biol. Chem. 280:17732–17736. Barski, A., Cuddapah, S., Cui, K., Roh, T. Y., Schones, D. E., Wang, Z., Wei, G., Chepelev, I., and Zhao, K. (2007). High-resolution profiling of histone methylations in the human genome. Cell 129:823–837. Berger, S. L. (2007). The complex language of chromatin regulation during transcription. Nature 447:407–412. Bernstein, B. E., Kamal, M., Lindblad-Toh, K., Bekiranov, S., Bailey, D. K., Huebert, D. J., McMahon, S., Karlsson, E. K., Kulbokas, E. J., III, Gingeras, T. R., Schreiber, S. L., and Lander, E. S. (2005). Genomic maps and comparative analysis of histone modifications in human and mouse. Cell 120:169–181. Bernstein, B. E., Mikkelsen, T. S., Xie, X., Kamal, M., Huebert, D. J., Cuff, J., Fry, B., Meissner, A., Wernig, M., Plath, K., Jaenisch, R., Wagschal, A., Feil, R., Schreiber, S. L., and Lander, E. S. (2006). A bivalent chromatin structure marks key developmental genes in embryonic stem cells. Cell 125:315–326. Bernstein, B. E., Meissner, A., and Lander, E. S. (2007). The mammalian epigenome. Cell 128:669–681. Bolton, E. C., So, A. Y., Chaivorapol, C., Haqq, C. M., Li, H., and Yamamoto, K. R. (2007). Celland gene-specific regulation of primary target genes by the androgen receptor. Genes Dev. 21:2005–2017. Bubendorf, L., Kononen, J., Koivisto, P., Schraml, P., Moch, H., Gasser, T. C., Willi, N., Mihatsch, M. J., Sauter, G., and Kallioniemi, O. P. (1999). Survey of gene amplifications during prostate cancer progression by high-throughput fluorescence in situ hybridization on tissue microarrays. Cancer Res. 59:803–806. Buchanan, G., Irvine, R. A., Coetzee, G. A., and Tilley, W. D. (2001). Contribution of the androgen receptor to prostate cancer predisposition and progression. Cancer Metastasis Rev. 20:207–223. Buchanan, G., Yang, M., Cheong, A., Harris, J. M., Irvine, R. A., Lambert, P. F., Moore, N. L., Raynor, M., Neufing, P. J., Coetzee, G. A., and Tilley, W. D. (2004). Structural and functional consequences of glutamine tract variation in the androgen receptor. Hum. Mol. Genet. 13:1677–1692. Buchanan, G., Ricciardelli, C., Harris, J. M., Prescott, J., Yu, Z. C., Jia, L., Butler, L. M., Marshall, V. R., Scher, H. I., Gerald, W. L., Coetzee, G. A., and Tilley, W. D. (2007). Control of androgen receptor signaling in prostate cancer by the cochaperone small glutamine rich tetratricopeptide repeat containing protein alpha. Cancer Res. 67:1008710096. Carroll, J. S., Meyer, C. A., Song, J., Li, W., Geistlinger, T. R., Eeckhoute, J., Brodsky, A. S., Keeton, E. K., Fertuck, K. C., Hall, G. F., Wang, Q., Bekiranov, S., Sementchenko, V., Fox, E. A., Silver, P. A., Gingeras, T. R., Liu, X. S., and Brown, M. (2006). Genome-wide analysis of estrogen receptor binding sites. Nat. Genet. 38:1289–1297. Catz, S. D., and Johnson, J. L. (2003). BCL-2 in prostate cancer: a minireview. Apoptosis 8:29–37. Cawley, S., Bekiranov, S., Ng, H. H., Kapranov, P., Sekinger, E. A., Kampa, D., Piccolboni, A., Sementchenko, V., Cheng, J., Williams, A. J., Wheeler, R., Wong, B., Drenkow, J., Yamanaka, M., Patel, S., Brubaker, S., Tammana, H., Helt, G., Struhl, K., and Gingeras, T. R. (2004). Unbiased mapping of transcription factor binding sites along human chromosomes 21 and 22 points to widespread regulation of noncoding RNAs. Cell 116:499–509. Chen, C. D., Welsbie, D. S., Tran, C., Baek, S. H., Chen, R., Vessella, R., Rosenfeld, M. G., and Sawyers, C. L. (2004). Molecular determinants of resistance to antiandrogen therapy. Nat. Med. 10:33–39. Couture, J. F., and Trievel, R. C. (2006). Histone-modifying enzymes: encrypting an enigmatic epigenetic code. Curr. Opin. Struct. Biol. 16:753–760.

Chromatin Remodeling and Androgen Receptor-Mediated Transcription

421

Darzacq, X., Shav-Tal, Y., de Turris, V., Brody, Y., Shenoy, S. M., Phair, R. D., and Singer, R. H. (2007). In vivo dynamics of RNA polymerase II transcription. Nat. Struct. Mol. Biol. 14:796– 806. Debes, J. D., and Tindall, D. J. (2004). Mechanisms of androgen-refractory prostate cancer. N. Engl. J. Med. 351:1488–1490. Ergun, A., Lawrence, C. A., Kohanski, M. A., Brennan, T. A., and Collins, J. J. (2007). A network biology approach to prostate cancer. Mol. Syst. Biol. 3:82. Euskirchen, G., Royce, T. E., Bertone, P., Martone, R., Rinn, J. L., Nelson, F. K., Sayward, F., Luscombe, N. M., Miller, P., Gerstein, M., Weissman, S., and Snyder, M. (2004). CREB binds to multiple loci on human chromosome 22. Mol. Cell. Biol. 24:3804–3814. Feinberg, A. P. (2007a). An epigenetic approach to cancer etiology. Cancer J. 13:70–74. Feinberg, A. P. (2007b). Phenotypic plasticity and the epigenetics of human disease. Nature 447:433–440. Feldman, B. J., and Feldman, D. (2001). The development of androgen-independent prostate cancer. Nat. Rev. Cancer 1:34–45. Gaston, K. E., Kim, D., Singh, S., Ford, O. H., III, and Mohler, J. L. (2003). Racial differences in androgen receptor protein expression in men with clinically localized prostate cancer. J. Urol. 170:990–993. Gregory, C. W., Johnson, R. T., Jr., Mohler, J. L., French, F. S., and Wilson, E. M. (2001). Androgen receptor stabilization in recurrent prostate cancer is associated with hypersensitivity to low androgen. Cancer Res. 61:2892–2898. Guenther, M. G., Levine, S. S., Boyer, L. A., Jaenisch, R., and Young, R. A. (2007). A chromatin landmark and transcription initiation at most promoters in human cells. Cell 130:77–88. Han, G., Buchanan, G., Ittmann, M., Harris, J. M., Yu, X., Demayo, F. J., Tilley, W., and Greenberg, N. M. (2005). Mutation of the androgen receptor causes oncogenic transformation of the prostate. Proc. Natl Acad. Sci. U. S. A. 102:1151–1156. He, B., Kemppainen, J. A., Voegel, J. J., Gronemeyer, H., and Wilson, E. M. (1999). Activation function 2 in the human androgen receptor ligand binding domain mediates interdomain communication with the NH(2)-terminal domain. J. Biol. Chem. 274:37219–37225. Heintzman, N. D., Stuart, R. K., Hon, G., Fu, Y., Ching, C. W., Hawkins, R. D., Barrera, L. O., Van Calcar, S., Qu, C., Ching, K. A., Wang, W., Weng, Z., Green, R. D., Crawford, G. E., and Ren, B. (2007). Distinct and predictive chromatin signatures of transcriptional promoters and enhancers in the human genome. Nat. Genet. 39:311–318. Holbert, M. A., and Marmorstein, R. (2005). Structure and activity of enzymes that remove histone modifications. Curr. Opin. Struct. Biol. 15:673–680. Holzbeierlein, J., Lal, P., LaTulippe, E., Smith, A., Satagopan, J., Zhang, L., Ryan, C., Smith, S., Scher, H., Scardino, P., Reuter, V., and Gerald, W. L. (2004). Gene expression analysis of human prostate carcinoma during hormonal therapy identifies androgen-responsive genes and mechanisms of therapy resistance. Am. J. Pathol. 164:217–227. Irvine, R. A., Yu, M. C., Ross, R. K., and Coetzee, G. A. (1995). The CAG and GGC microsatellites of the androgen receptor gene are in linkage disequilibrium in men with prostate cancer. Cancer Res. 55:1937–1940. Irvine, R. A., Ma, H., Yu, M. C., Ross, R. K., Stallcup, M. R., and Coetzee, G. A. (2000). Inhibition of p160-mediated coactivation with increasing androgen receptor polyglutamine length. Hum. Mol. Genet. 9:267–274. Jenster, G. (1999). The role of the androgen receptor in the development and progression of prostate cancer. Semin. Oncol. 26:407–421. Jia, L., Shen, H. C., Wantroba, M., Khalid, O., Liang, G., Wang, Q., Gentzschein, E., Pinski, J. K., Stanczyk, F. Z., Jones, P. A., and Coetzee, G. A. (2006). Locus-wide chromatin remodeling and enhanced androgen receptor-mediated transcription in recurrent prostate tumor cells. Mol. Cell. Biol. 26:7331–7341. Jones, P. A., and Baylin, S. B. (2007). The epigenomics of cancer. Cell 128:683–692.

422

L. Jia et al.

Kahl, P., Gullotti, L., Heukamp, L. C., Wolf, S., Friedrichs, N., Vorreuther, R., Solleder, G., Bastian, P. J., Ellinger, J., Metzger, E., Schule, R., and Buettner, R. (2006). Androgen receptor coactivators lysine-specific histone demethylase 1 and four and a half LIM domain protein 2 predict risk of prostate cancer recurrence. Cancer Res. 66:11341–11347. Kim, T. H., Abdullaev, Z. K., Smith, A. D., Ching, K. A., Loukinov, D. I., Green, R. D., Zhang, M. Q., Lobanenkov, V. V., and Ren, B. (2007). Analysis of the vertebrate insulator protein CTCFbinding sites in the human genome. Cell 128:1231–1245. Klokk, T. I., Kurys, P., Elbi, C., Nagaich, A. K., Hendarwanto, A., Slagsvold, T., Chang, C. Y., Hager, G. L., and Saatcioglu, F. (2007). Ligand-specific dynamics of the androgen receptor at its response element in living cells. Mol. Cell. Biol. 27:1823–1843. Koivisto, P. A., and Helin, H. J. (1999). Androgen receptor gene amplification increases tissue PSA protein expression in hormone-refractory prostate carcinoma. J. Pathol. 189:219–223. Koivisto, P., Kononen, J., Palmberg, C., Tammela, T., Hyytinen, E., Isola, J., Trapman, J., Cleutjens, K., Noordzij, A., Visakorpi, T., and Kallioniemi, O. P. (1997). Androgen receptor gene amplification: a possible molecular mechanism for androgen deprivation therapy failure in prostate cancer. Cancer Res. 57:314–319. Kosak, S. T., and Groudine, M. (2004). Form follows function: the genomic organization of cellular differentiation. Genes Dev. 18:1371–1384. Kouzarides, T. (2007). SnapShot: histone-modifying enzymes. Cell 131:822. Kozlowski, J. M., Ellis, W. J., and Grayhack, J. T. (1991). Advanced prostatic carcinoma. Early versus late endocrine therapy. Urol. Clin. North Am. 18:15–24. Kumar, R., and Thompson, E. B. (2003). Transactivation functions of the N-terminal domains of nuclear hormone receptors: protein folding and coactivator interactions. Mol. Endocrinol. 17:1–10. Li, B., Carey, M., and Workman, J. L. (2007). The role of chromatin during transcription. Cell 128:707–719. Li, J., Zhang, D., Fu, J., Huang, Z., and Wong, J. (2009). Structural and functional analysis of androgen receptor in chromatin. Mol. Endocrinol. [Epub ahead of print, doi:10.1210/me.20060221] Liang, G., Lin, J. C., Wei, V., Yoo, C., Cheng, J. C., Nguyen, C. T., Weisenberger, D. J., Egger, G., Takai, D., Gonzales, F. A., and Jones, P. A. (2004). Distinct localization of histone H3 acetylation and H3-K4 methylation to the transcription start sites in the human genome. Proc. Natl Acad. Sci. U. S. A. 101:7357–7362. Ling, M. T., Chan, K. W., and Choo, C. K. (2001). Androgen induces differentiation of a human papillomavirus 16 E6/E7 immortalized prostate epithelial cell line. J. Endocrinol. 170:287– 296. Linja, M. J., Savinainen, K. J., Saramaki, O. R., Tammela, T. L., Vessella, R. L., and Visakorpi, T. (2001). Amplification and overexpression of androgen receptor gene in hormone-refractory prostate cancer. Cancer Res. 61:3550–3555. Ma, H., Hong, H., Huang, S. M., Irvine, R. A., Webb, P., Kushner, P. J., Coetzee, G. A., and Stallcup, M. R. (1999). Multiple signal input and output domains of the 160-kilodalton nuclear receptor coactivator proteins. Mol. Cell. Biol. 19:6164–6173. Marshall, T. W., Link, K. A., Petre-Draviam, C. E., and Knudsen, K. E. (2003). Differential requirement of SWI/SNF for androgen receptor activity. J. Biol. Chem. 278:30605–30613. Martone, R., Euskirchen, G., Bertone, P., Hartman, S., Royce, T. E., Luscombe, N. M., Rinn, J. L., Nelson, F. K., Miller, P., Gerstein, M., Weissman, S., and Snyder, M. (2003). Distribution of NF-kappaB-binding sites across human chromosome 22. Proc. Natl Acad. Sci. U. S. A. 100:12247–12252. Massie, C. E., Adryan, B., Barbosa-Morais, N. L., Lynch, A. G., Tran, M. G., Neal, D. E., and Mills, I. G. (2007). New androgen receptor genomic targets show an interaction with the ETS1 transcription factor. EMBO Rep. 8:871–878. Metzger, E., Wissmann, M., and Schule, R. (2006). Histone demethylation and androgen-dependent transcription. Curr. Opin. Genet. Dev. 16:513–517.

Chromatin Remodeling and Androgen Receptor-Mediated Transcription

423

Mikkelsen, T. S., Ku, M., Jaffe, D. B., Issac, B., Lieberman, E., Giannoukos, G., Alvarez, P., Brockman, W., Kim, T. K., Koche, R. P., Lee, W., Mendenhall, E., O’Donovan, A., Presser, A., Russ, C., Xie, X., Meissner, A., Wernig, M., Jaenisch, R., Nusbaum, C., Lander, E. S., and Bernstein, B. E. (2007). Genome-wide maps of chromatin state in pluripotent and lineagecommitted cells. Nature 448:553–560. Murtha, P., Tindall, D. J., and Young, C. Y. (1993). Androgen induction of a human prostatespecific kallikrein, hKLK2: characterization of an androgen response element in the 50 promoter region of the gene. Biochemistry 32:6459–6464. Muse, G. W., Gilchrist, D. A., Nechaev, S., Shah, R., Parker, J. S., Grissom, S. F., Zeitlinger, J., and Adelman, K. (2007). RNA polymerase is poised for activation across the genome. Nat. Genet. 39:1507–1511. Pratt, W. B., and Toft, D. O. (1997). Steroid receptor interactions with heat shock protein and immunophilin chaperones. Endocr. Rev. 18:306–360. Remenyi, A., Scholer, H. R., and Wilmanns, M. (2004). Combinatorial control of gene expression. Nat. Struct. Mol. Biol. 11:812–815. Ricciardelli, C., Choong, C. S., Buchanan, G., Vivekanandan, S., Neufing, P., Stahl, J., Marshall, V. R., Horsfall, D. J., and Tilley, W. D. (2005). Androgen receptor levels in prostate cancer epithelial and peritumoral stromal cells identify non-organ confined disease. Prostate 63:1928. Rivenbark, A. G., and Strahl, B. D. (2007). Molecular biology. Unlocking cell fate. Science 318:403–404. Robertson, G., Hirst, M., Bainbridge, M., Bilenky, M., Zhao, Y., Zeng, T., Euskirchen, G., Bernier, B., Varhol, R., Delaney, A., Thiessen, N., Griffith, O. L., He, A., Marra, M., Snyder, M., and Jones, S. (2007). Genome-wide profiles of STAT1 DNA association using chromatin immunoprecipitation and massively parallel sequencing. Nat. Methods 4:651–657. Rosenfeld, M. G., and Glass, C. K. (2001). Coregulator codes of transcriptional regulation by nuclear receptors. J. Biol. Chem. 276:36865–36868. Savarese, F., and Grosschedl, R. (2006). Blurring cis and trans in gene regulation. Cell 126:248– 250. Scher, H. I., Buchanan, G., Gerald, W., Butler, L. M., and Tilley, W. D. (2004). Targeting the androgen receptor: improving outcomes for castration-resistant prostate cancer. Endocr. Relat. Cancer 11:459–476. Schuurmans, A. L., Bolt, J., Veldscholte, J., and Mulder, E. (1991). Regulation of growth of LNCaP human prostate tumor cells by growth factors and steroid hormones. J. Steroid Biochem. Mol. Biol. 40:193–197. Sellers, W. R., and Loda, M. (2002). The EZH2 polycomb transcriptional repressor – a marker or mover of metastatic prostate cancer? Cancer Cell 2:349–350. Seo, S., and Kroll, K. L. (2006). Geminin’s double life: chromatin connections that regulate transcription at the transition from proliferation to differentiation. Cell Cycle 5:374–379. Shen, H. C., Buchanan, G., Butler, L. M., Prescott, J., Henderson, M., Tilley, W. D., and Coetzee, G. A. (2005). GRIP1 mediates the interaction between the amino- and carboxyl-termini of the androgen receptor. Biol. Chem. 386:69–74. Shi, Y. (2007). Histone lysine demethylases: emerging roles in development, physiology and disease. Nat. Rev. Genet. 8:829–833. Strahl, B. D., and Allis, C. D. (2000). The language of covalent histone modifications. Nature 403:41–45. Swigut, T., and Wysocka, J. (2007). H3K27 demethylases, at long last. Cell 131:29–32. Tanner, T., Claessens, F., and Haelens, A. (2004). The hinge region of the androgen receptor plays a role in proteasome-mediated transcriptional activation. Ann. N. Y. Acad. Sci. 1030:587–592. Trojer, P., and Reinberg, D. (2007). Facultative heterochromatin: is there a distinctive molecular signature? Mol. Cell 28:1–13. Tyagi, R. K., Lavrovsky, Y., Ahn, S. C., Song, C. S., Chatterjee, B., and Roy, A. K. (2000). Dynamics of intracellular movement and nucleocytoplasmic recycling of the ligand-activated androgen receptor in living cells. Mol. Endocrinol. 14:1162–1174.

424

L. Jia et al.

Varambally, S., Dhanasekaran, S. M., Zhou, M., Barrette, T. R., Kumar-Sinha, C., Sanda, M. G., Ghosh, D., Pienta, K. J., Sewalt, R. G., Otte, A. P., Rubin, M. A., and Chinnaiyan, A. M. (2002). The polycomb group protein EZH2 is involved in progression of prostate cancer. Nature 419:624–629. Vermeulen, M., Mulder, K. W., Denissov, S., Pijnappel, W. W., van Schaik, F. M., Varier, R. A., Baltissen, M. P., Stunnenberg, H. G., Mann, M., and Timmers, H. T. (2007). Selective anchoring of TFIID to nucleosomes by trimethylation of histone H3 lysine 4. Cell 131:58–69. Villa, R., Pasini, D., Gutierrez, A., Morey, L., Occhionorelli, M., Vire, E., Nomdedeu, J. F., Jenuwein, T., Pelicci, P. G., Minucci, S., Fuks, F., Helin, K., and Di Croce, L. (2007). Role of the polycomb repressive complex 2 in acute promyelocytic leukemia. Cancer Cell 11:513– 525. Wang, Q., Li, W., Liu, X. S., Carroll, J. S., Janne, O. A., Keeton, E. K., Chinnaiyan, A. M., Pienta, K. J., and Brown, M. (2007). A hierarchical network of transcription factors governs androgen receptor-dependent prostate cancer growth. Mol. Cell 27:380–392. Wei, C. L., Wu, Q., Vega, V. B., Chiu, K. P., Ng, P., Zhang, T., Shahab, A., Yong, H. C., Fu, Y., Weng, Z., Liu, J., Zhao, X. D., Chew, J. L., Lee, Y. L., Kuznetsov, V. A., Sung, W. K., Miller, L. D., Lim, B., Liu, E. T., Yu, Q., Ng, H. H., and Ruan, Y. (2006). A global map of p53 transcription-factor binding sites in the human genome. Cell 124:207–219. Xie, X., Mikkelsen, T. S., Gnirke, A., Lindblad-Toh, K., Kellis, M., and Lander, E. S. (2007). Systematic discovery of regulatory motifs in conserved regions of the human genome, including thousands of CTCF insulator sites. Proc. Natl Acad. Sci. U. S. A. 104:7145–7150. Yamane, K., Toumazou, C., Tsukada, Y., Erdjument-Bromage, H., Tempst, P., Wong, J., and Zhang, Y. (2006). JHDM2A, a JmjC-containing H3K9 demethylase, facilitates transcription activation by androgen receptor. Cell 125:483–495. Yu, X., Gupta, A., Wang, Y., Suzuki, K., Mirosevich, J., Orgebin-Crist, M. C., and Matusik, R. J. (2005). Foxa1 and Foxa2 interact with the androgen receptor to regulate prostate and epididymal genes differentially. Ann. N. Y. Acad. Sci. 1061:77–93. Zhu, P., Zhou, W., Wang, J., Puc, J., Ohgi, K. A., Erdjument-Bromage, H., Tempst, P., Glass, C. K., and Rosenfeld, M. G. (2007). A histone H2A deubiquitinase complex coordinating histone acetylation and H1 dissociation in transcriptional regulation. Mol. Cell 27:609–621. Zoubeidi, A., Zardan, A., Beraldi, E., Fazli, L., Sowery, R., Rennie, P., Nelson, C., and Gleave, M. (2007). Cooperative interactions between androgen receptor (AR) and heat-shock protein 27 facilitate AR transcriptional activity. Cancer Res. 67:10455–10465.

Ligand-Independent Androgen Receptor Activity Scott M. Dehm and Donald J. Tindall

Abstract The androgen receptor (AR) is important for the growth and survival of normal and malignant prostate cells. As such, androgen-deprivation therapy is the current mainstay of systemic prostate cancer therapy. Invariably, prostate cancer will develop resistance to androgen deprivation and recur with a ‘‘castrationrecurrent’’ phenotype. The surprising finding that castration-recurrent prostate cancer is still reliant on AR activity indicates that novel means of targeting the AR could be developed to treat this stage of the disease. Several mechanisms, including ligand-independent activation, have been described as means by which the AR can achieve a critical level of activity in castration-recurrent prostate cancer. This chapter will explore the mechanisms of ligand-independent AR activation and highlight some recent findings generated in the Tindall laboratory.

1 Introduction 1.1

Androgens and the Androgen Receptor: The Androgen-Signaling Axis

Androgens are the male sex hormones, which drive the differentiation and maturation of male reproductive organs and development of male secondary sexual characteristics. Testosterone represents approximately 90% of total circulating androgens and is synthesized by Leydig cells in the testes. The remaining circulating androgens are synthesized in the adrenal cortex, and include dehydroepiandrosterone (DHEA), androstenediol, and androstenedione. These adrenal androgens can be converted to testosterone in peripheral tissues. Testosterone production is regulated tightly by the hypothalamic-pituitary-gonadal axis of the endocrine system, whereby Leydig cells are stimulated directly by leuteinizing hormone D.J. Tindall(*) Departments of Urology and Biochemistry & Molecular Biology, Mayo Clinic College of Medicine, 200 First Street SW, Rochester, MN 55905, USA, E-mail: [email protected]

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4_19, # Springer Science + Business Media, LLC 2009

427

428

S.M. Dehm, D.J. Tindall

(LH) secretions from the anterior pituitary. LH release is controlled by leuteinizing hormone releasing hormone (LHRH), which is secreted in a pulsatile manner by the hypothalamus. Testosterone feeds back on this signaling axis to inhibit LHRH release by the hypothalamus and decrease the sensitivity of the pituitary to LHRH. Approximately 98% of circulating testosterone is bound to globular carrier proteins in the bloodstream such as albumin and sex-hormone-binding globulin (Hammond et al. 2003). Testosterone is lipophilic and can freely diffuse into target cells. Upon entry into target tissues such as the prostate, 5a-reductase enzymes rapidly convert testosterone to dihydrotestosterone (DHT). Both testosterone and DHT exert their activity by binding the androgen receptor (AR), a 110-kDa phosphoprotein and member of the steroid hormone receptor transcription factor superfamily (Lamb et al. 2001). DHT is a more potent activator of the AR due to enhanced stabilization of the receptor and slower dissociation (Zhou et al. 1995). However, both these androgens are required for full virilization.

1.2

Role of Androgens in Normal Prostate Physiology

The prostate, a walnut-sized and -shaped exocrine gland that surrounds the neck of the bladder and first part of the urethra, is probably the best-studied androgen-dependent tissue. The prostate is composed of an epithelial cell compartment supported by a surrounding stroma of fibroblasts, smooth muscle cells, and blood vessels. The epithelial cell compartment contains mostly basal and luminal epithelial cells, which are arranged as glandular ascini and produce secretions that are a major component of seminal fluid. A complex signaling relationship exists between prostatic stromal and epithelial tissues, which results in a steady state where only 1–2% of epithelial cells are proliferating or undergoing apoptosis per day (reviewed in Sung and Chung 2002). For example, within the epithelial compartment, the AR directly inhibits epithelial cell proliferation, induces differentiation, and regulates prostate metabolic and secretory function (Sung and Chung 2002). Within the stromal compartment, the AR directly regulates the production of various peptide growth factors; these peptides subsequently act in a paracrine fashion on the epithelial layer to positively regulate the proliferation of basal cells and the survival of secretory luminal cells (Sung and Chung 2002). Current knowledge of the androgen-regulated genes that are regulated directly by the AR comes primarily from large-scale gene expression and ChIP-on-chip studies (Dehm and Tindall 2006b). These genes join a list of AR transcriptional targets that include the seminal protease, prostate-specific antigen (PSA), which is the best-defined androgen-regulated gene and an important clinical marker for the development and progression of prostate cancer. Many studies will continue to assess the importance of these androgen-regulated genes in mediating normal prostate function, as well as the growth and survival of prostate cancer cells.

1.3

Role of Androgens in Prostate Cancer

Prostate cancer is an important worldwide health concern, representing the sixth most common form of cancer overall, and third most common cancer in men. In

Ligand-Independent Androgen Receptor Activity

429

developed countries such as the United States, prostate cancer is the most prevalent form of male malignancy and second leading cause of cancer deaths (Jemal et al. 2003). The lifetime risk of developing prostate cancer is 1 in 6 for North American men; however, the chance of a man developing this disease before age 40 is low (Jemal et al. 2003). The typical treatment for locally confined prostate cancer consists of surgery and/or radiation therapy (Scherr et al. 2003). For clinically advanced, relapsed, or metastatic disease, androgen-deprivation therapy is the standard treatment (Scherr et al. 2003). The reduction in circulating testicular androgen levels effectively inhibits the AR and results in clinical regression. Ultimately however, tumors that initially responded to androgen-deprivation therapy attain the ability to grow in an androgen-depleted environment. When prostate cancer progresses to this castration-recurrent state, the disease is highly metastatic and further treatment is essentially palliative (Scherr et al. 2003). A major priority, therefore, is to understand the molecular mechanisms that drive prostate cancer from depending on circulating androgens to being able to grow in a castrate host. This chapter will focus on the AR during this phase of prostate cancer progression.

2 AR and Castration-Recurrent (Androgen-Independent/ Androgen-Refractory/Androgen Depletion-Independent) Prostate Cancer Like normal prostate tissue, almost all prostate cancers are initially androgen dependent. This critical feature of cells of prostatic origin is the foundation for the initial success of androgen-deprivation therapy for prostate cancer. An early hypothesis was that most castration-recurrent prostate cancers bypassed AR-regulated signaling pathways during androgen-deprivation therapy in order to resume growth, presumably associated with the loss of AR expression. This was supported by the fact that several commonly studied androgen-independent and highly metastatic prostate cancer cell lines, such as DU-145 and PC-3, did not express AR. Moreover, tumors derived from the Dunning rat prostate cancer model exhibited loss of AR expression during the transition to castration-recurrent disease (Buchanan et al. 2001b). These observations led to a hypothesis that the AR was no longer important for the growth and survival of prostate cancer when it recurred after medical or surgical castration.

2.1

AR and Castration-Recurrent Prostate Cancer

Several laboratories reported AR activation by growth factors, cytokines, and their activated downstream signaling pathways (Culig et al. 1994; Hobisch et al. 1998), which suggested that androgen-independent mechanisms of AR activation existed. These observations led our laboratory to hypothesize that the AR might still be an

430

S.M. Dehm, D.J. Tindall

important factor for the growth and survival of truly androgen-independent prostate cancer cells, in a manner that was uncoupled from normal androgen regulation (Grossmann et al. 2001). Indeed, several cell-based models of androgen-independent prostate cancer had been developed, and all expressed a functional AR (Wu et al. 1994). To test whether a role existed for the AR in castration-recurrent prostate cancer, LNCaP cells were employed as a model for androgen-dependent prostate cancer, and two LNCaP-derived sublines, LNCaP-Rf and LNCaP-C4, were used as models for androgen-independent prostate cancer. The DU-145 cell line was also employed as a model for androgen-independent and AR-null prostate cancer. The critical question of whether the AR played a role in these cell-based models was addressed by microinjecting AR-directed antibodies into their nuclei (Fig. 1). As a control, similar microinjections were performed with the same antibodies that had been heat inactivated. Antibodies were coinjected with Texas Red Dextran, which allowed monitoring of the microinjected cells for up to 9 days by using confocal laser scanning microscopy. Nuclear microinjection of AR-targeted antibodies, but not heat-inactivated antibodies, markedly reduced proliferation in all AR expressing androgen-dependent and androgen-independent prostate cancer cells. No effect of

Fig. 1 An AR-targeting approach to assess the role of AR in androgen-independent prostate cancer cells. Androgen-sensitive LNCaP and androgen-independent LNCaP-C4, LNCaP-Rf, and DU-145 cells were microinjected with antibodies targeting the AR or an AR-directed hammerhead ribozyme. Up to 9 days post-injection, proliferation and expression of the AR-regulated PSA gene were assessed

Ligand-Independent Androgen Receptor Activity

431

AR-targeted antibodies was observed in the AR-null DU-145 cell line. AR antibodies also inhibited PSA expression in all AR-positive cell lines, which further confirmed that the AR was able to achieve a critical level of activity in the androgen-independent prostate cancer cells. A similar set of microinjections using an AR-directed hammerhead ribozyme produced the exact same effects on prostate cancer cell growth. These results demonstrated that the AR had a functional role in androgen-independent prostate cancer cells. Since this initial report, several follow-up studies have been performed, and all have yielded similar results. Development of small-interfering RNA (siRNA) has permitted a more efficient and simple means of selectively knocking down AR expression, and demonstrated that targeted inhibition of the AR decreases PSA expression, cell proliferation, and survival in various cell-based models of castration-recurrent prostate cancer (Eder et al. 2000; Haag et al. 2005; Liao et al. 2005). Moreover, studies of clinical samples have demonstrated that most castrationrecurrent prostate cancer retains high levels of AR expression (Buchanan et al. 2001b), and the PSA gene continues to be expressed (Heinlein and Chang 2004). Taken together, these and other findings have solidified a role for the AR in castration-recurrent prostate cancer, and suggest that castration-recurrent cancer cells continue to proliferate and survive in an androgen-depleted environment through aberrant mechanisms of AR activation.

2.2

Mechanisms of AR Activation in Castration-Recurrent Prostate Cancer

Several mechanisms have been proposed whereby the AR signal transduction pathway could be activated during androgen-deprivation therapy and contribute to the growth and survival of castration-recurrent prostate cancer. When a prostate tumor progresses to this state, it is commonly referred to as androgen independent; however, it has not yet been established if prostate tumors are able to grow in the complete absence of androgens in vivo. One hypothesis is that castration-recurrent tumors still require androgens for growth and survival, but at much lower concentrations. In this case, the AR may be sensitized to castrate testosterone or DHT concentrations, or responds more robustly to adrenal androgens (Mohler et al. 2004). Alternatively, AR activation may be achieved in the absence of androgens. This theory is based on many studies with cultured cells, which have shown the AR can be activated via mutation, nonandrogenic steroidal ligands, antiandrogens, changes in coactivator expression patterns, and various peptide growth factor signaling pathways (Grossmann et al. 2001). 2.2.1

AR Mutations

Current studies have indicated that approximately 10% of castration-recurrent prostate cancers harbor somatic mutations in the AR gene. (Taplin et al. 2003). Most often these mutations impart a gain of function to the AR. A database has been

432

S.M. Dehm, D.J. Tindall

compiled to document the different AR mutations observed in prostate cancer as well as androgen insensitivity syndrome (Gottlieb et al. 2004). Currently, this database lists approximately 70 different AR mutations observed in prostate cancer. Interestingly, of these mutations, the majority (79%) are confined to three discrete regions which make up 8% of the AR coding sequence (Buchanan et al. 2001a). Mutations clustered within the AR signature sequence (amino acids 701–730) have been shown to alter the specificity of the AR, which allows inappropriate activation by alternative steroidal ligands, such as adrenal androgens, glucocorticoids, and progesterone (Culig et al. 1993; Culig et al. 1996; Zhao et al. 2000). Mutations within a region flanking the AR AF-2 region (amino acids 874–910) confer a similar property of altered ligand specificity on the AR (Bentel and Tilley 1996). One such mutation, T877A, has been observed in the LNCaP cell line, as well as castration-recurrent prostate cancer during antiandrogen therapy with hydroxyflutamide. The most recently defined region of the AR exhibiting clustering of mutations in prostate tumors is between the hinge and LBD domains of the AR (amino acids 670–678). Functional study of these AR mutants has shown an overall enhanced transactivation in response to DHT, as well as other steroidal and nonsteroidal ligands (Buchanan et al. 2001c).

2.2.2

AR Overexpression

Two RT-PCR studies reported higher mRNA expression levels in castration-recurrent compared with primary untreated tumors (Latil et al. 2001; Linja et al. 2001). These studies also showed that AR mRNA levels were similar in normal prostate tissue and primary untreated prostate cancer (Latil et al. 2001; Linja et al. 2001). Immunohistochemistry showed increased AR expression in approximately 35% of prostate cancer that had recurred during androgen-deprivation therapy (Edwards et al. 2003). A gene expression profiling study, using seven paired androgen sensitive and androgen-independent prostate cancer xenografts, demonstrated that expression of the AR gene was consistently up-regulated during progression to castrationrecurrent disease (Chen et al. 2004). Further functional assessment showed that forced AR overexpression in LNCaP cells sensitized these cells to low ligand levels and caused a more rapid progression to castration-recurrent disease (Chen et al. 2004). Based on these findings, an important issue has been to understand the mechanisms leading to AR overexpression during the progression of prostate cancer. One such mechanism involves amplification of the AR gene. The consensus from four independent studies is that AR amplification occurs at a rate of 20–33% in castrationrecurrent prostate cancer (Bubendorf et al. 1999; Edwards et al. 2003; Ford et al. 2003; Linja et al. 2001; Visakorpi et al. 1995). Another mechanism is up-regulated transcription of the AR gene. The AR gene expression is regulated by two genomic AREs, which are engaged by the AR in response to androgens (Dai and Burnstein 1996). In addition, a repressor complex has been identified that binds to the 50 genomic untranslated region of the AR gene and negatively regulates transcription (Wang et al. 2004).

Ligand-Independent Androgen Receptor Activity

2.2.3

433

Ligand-Independent AR Activation

A considerable body of evidence implicates activation of various growth factor signaling pathways as well as altered activity and/or expression of coactivator proteins in the maintenance of AR activity in castration-recurrent prostate cancer. Indeed, the AR can be activated by growth factors such as insulin-like growth factor (IGF), epidermal growth factor (EGF), and keratinocyte growth factor (KGF) (Culig et al. 1994), activated growth factor receptors such as HER-2/neu (Craft et al. 1999), as well as cytokines such as interleukin-4 (IL-4) and IL-6 (Hobisch et al. 1998; Lee et al. 2003b).

The role of IL-6 and p300 Our laboratory has been interested particularly in the role of IL-6 in prostate cancer progression. Cytokines, such as IL-6, are mediators of immune inflammatory responses, and also influence the growth of normal and cancer cells. Importantly, several reports have demonstrated elevated circulating levels of IL-6 in patients with castration-recurrent prostate cancer (Adler et al. 1999; Drachenberg et al. 1999; Lee et al. 2003a; Wise et al. 2000). IL-6 transactivated the endogenous AR in LNCaP cells, in the absence of androgens (Debes et al. 2002). The MAPK pathway had been shown to influence AR activity; therefore, a potential role for MAPK activity in IL-6mediated AR activation was explored (Hobisch et al. 1998; Ueda et al. 2002). The MAP/ERK kinase (MEK)-1 inhibitor PD98059 inhibited the transactivation of the AR by IL-6 (Debes et al. 2002). Previous studies had suggested that p300, an ARassociated coregulator, may be regulated by the MAPK pathway. Transfection of LNCaP cells with p300 abrogated the inhibitory effect of PD98059 on IL-6 mediated AR transactivation. Moreover, inhibiting p300 function, through cotransfection with E1A (an oncoprotein that sequesters p300 and inhibits its HAT activity) (Chakravarti et al. 1996), or p300-targeted siRNA, blocked IL-6-mediated transactivation of the AR. Together, these results suggested that IL-6-mediated transactivation of the AR occurred through the MAPK pathway and likely involved p300 as a target (Fig. 2). Because the data indicated that p300 may be involved in prostate cancer progression, p300 expression was assessed in 95 men with prostate cancer and correlated with a variety of clinicopathological parameters. p300 expression in diagnostic prostate cancer biopsies correlated positively with proliferative index, tumor size at prostatectomy, extraprostatic extension of prostate cancer, seminal vesicle involvement, nondiploid DNA content, and disease progression following surgery (Debes et al. 2003). These data confirmed the in vitro data, which demonstrated that p300 may play an important role in prostate cancer. These findings further suggested that p300 expression on biopsy may be a potential marker to predict aggressive features of prostate cancer in men who undergo radical prostatectomy for clinically localized prostate cancer. During the course of these studies, evidence was obtained that suggested expression of p300 was affected by androgens (Heemers et al. 2007). For example, when

434

S.M. Dehm, D.J. Tindall

Fig. 2 The IL-6/MAPK/p300/AR pathway. IL-6 binds the IL-6 receptor, which associates with the gp130 receptor and recruits JAK kinases in order to transduce signals to the cytosol. These signals induce MAPK activity, which in turn induces AR to engage with AREs in enhancers and promoters of target genes and activate their transcription in a ligand-independent manner. p300 is involved as an intermediary factor in this pathway and functions downstream of the MAPK pathway

LNCaP cells were cultured in the presence of androgens for 3 days, p300 protein expression was decreased markedly. Time-course studies demonstrated that the androgen-induced decrease in p300 expression first occurred 24 h after treatment and became more pronounced after longer periods of treatment. A maximal affect on p300 expression occurred following 72 h of androgen treatment. Dose-response studies showed that androgen concentrations of at least 1nM were required for reduced p300 expression. This androgen effect on p300 expression was AR dependent, because androgens had no effect when AR activity was blocked with the antiandrogen, casodex, or AR-targeted siRNAs. Androgens had no effect on p300 mRNA expression, but rather exerted their effect at the level of p300 protein expression. Further studies with chemical inhibitors suggested involvement of nuclear factor-kB (NF-kB) signaling in the androgen effects on p300 expression. However, the exact molecular machinery by which NF-kB activity might mediate the effects of changes in the androgenic milieu on p300 expression remains unclear.

Ligand-Independent Androgen Receptor Activity

435

Given the suppressive effects of androgens on p300 expression, it became evident that androgen deprivation would lead to an increase in the expression of p300 in prostate cancer cells. Short-term androgen deprivation produced a marked increase in p300 expression when LNCaP cells were cultured 3 days in androgendepleted medium. The effect of long-term androgen deprivation was studied using the androgen-independent, LNCaP-derived cell line, LNCaP-Rf. p300 protein levels in LNCaP-Rf cells were considerably higher than in parental LNCaP cells. Moreover, elevated p300 expression in LNCaP-Rf cells was critical for overall cell viability (Heemers et al. 2007). Together, these results suggest that an increase in p300 expression, fostered by androgen deprivation, could facilitate subsequent activation of the AR by IL-6 and thereby provide a protective growth advantage during prostate cancer therapy.

3 AR Structure/Function Identification and analysis of the signaling pathways that converge on the AR transcriptional complex and influence its activity – either in a ligand-dependent or ligand-independent manner – is of critical importance. However, to understand precisely the molecular mechanisms by which the AR activates transcription in an androgen-dependent and/or androgen-independent manner, a thorough understanding of AR structure and function is required. The AR shares an overall modular organization similar to other steroid receptors (Fig. 3). The AR consists of an N-terminal domain (NTD) that harbors an AR transcriptional activation function (AF)-1, a central DNA-binding domain (DBD), a short hinge region, and a COOH-terminal

Fig. 3 Modular domains and functional motifs of the AR. The AR is composed of modular domains that harbor activation function (AF)-1, the DNA-binding domain (DBD), a hinge region, and the ligand-binding domain (LBD)/AF-2 coactivator-binding surface. Amino acids within the LBD that hydrogen bond with the O-3 position in the steroid ring are indicated in gray text, and amino acids that hydrogen bond with the 17b-OH position are indicated in outlined text. Functional domains and motifs indicated in this diagram are described in the text

436

S.M. Dehm, D.J. Tindall

domain (CTD) that contains both the AR ligand-binding domain (LBD) and AF2 coactivator-binding surface (Bain et al. 2006). The 3-dimensional structures of peptides representing the LBD and AF-2 folds of the AR have been determined using X-ray crystallography, as has the 3-dimensional structure of a peptide representing the AR DBD (Estebanez-Perpina et al. 2005; He et al. 2004; Hur et al. 2004; Matias et al. 2000; Sack et al. 2001). Conversely, the AR NTD is unstructured in solution and thus has been resistant to structural determination.

3.1

Structure and Function of the AR COOH-Terminal Domain

All nuclear receptor CTDs perform a similar mechanistic function of binding ligand within a central LBD cavity, which induces formation of the AF-2 coactivatorbinding surface. The 3-dimensional structures of many nuclear receptor CTDs have been solved using X-ray crystallography, which reveal highly similar globular domains made up of 12 a-helices organized as three antiparallel helical sheets (Li et al. 2003). The crystal structure of the human AR LBD bound to the synthetic androgen R1881 was reported first in 2000, which revealed that the AR also forms a highly similar structure (Matias et al. 2000). However, the prototype globular ahelical sandwich organization is composed of 11 a-helices for the AR, as opposed to its family members’ 12, due to the absence of helix 2. These helices are numbered 1–12, with helix 2 omitted to reflect the overall positional identity of these CTD helices in relation to other steroid receptor family members. A total of 18 amino acid residues, situated within each of the 11 AR LBD helices, are responsible for direct interactions with the ligand, with helix 12 functioning as a flexible lid to stabilize the ligand within the LBD cavity. While most of these interactions are hydrophobic, the androgen O-3 hydrogen bonds with AR amino acids Q711, M745, and R752, and the 17b-OH hydrogen bonds with AR amino acids N705 and T877. Binding of ligand to the AR LBD causes a significant positional change in helix 12, as well as an overall conformational change in the AR CTD, which induces formation of the AF-2 coactivator-binding surface. The AF-2 surface is a hydrophobic groove flanked by concentrated regions of positive and negative charges. Similar to other nuclear receptors, the AF-2 surface serves as a docking site for LxxLL motifs present in prototype nuclear receptor coactivators and corepressors. The crystal structures of liganded AR in complex with LxxLL-containing peptides derived from the AR coactivators SRC-2, SRC-3, ARA70 have been described (He et al. 2004; Hur et al. 2004). These crystal structures revealed that K720 in helix 3 and E897 in helix 12 function as ‘‘charge-clamp’’ residues, which stabilize the LxxLL/AF-2 interaction. Prior to elucidation of the AR LBD crystal structure it was apparent from sequence alignment that K720 and E897 were structurally and functionally homologous to charge clamp residues in other nuclear receptors. In addition to serving as a docking site for coactivators, AR AF-2 is able to mediate interaction with the AR NTD in an intramolecular fashion (Klokk et al. 2006).

Ligand-Independent Androgen Receptor Activity

437

Biochemical evidence has demonstrated that this N/C interaction is mediated by direct binding of FxxLF and/or WxxLF motifs in the AR NTD with the AR AF2 coactivator-binding surface (He et al. 2000). The crystal structures of FxxLF- and WxxLF-containing peptides engaged with the AF-2 surface of ligand-bound AR have been described, and have revealed a similar overall mode of binding to AF2 as coactivator-derived LxxLL peptides (He et al. 2004; Hur et al. 2004). The AR AF-2 domain displays a higher affinity for NTD-derived FxxLF-containing peptides than coactivator-derived LxxLL-containing peptides, which suggests that N/C interaction, rather than direct transcriptional activation, may be the primary role for AF-2. From a functional standpoint, the N/C interaction is critical for the AR to bind a chromatin-integrated MMTV promoter but not a nonintegrated reporter (Li et al. 2006). Moreover, AF-2 preferentially binds the NTD when the AR is mobile, but preferentially binds coregulators when the AR engages with DNA (van Royen et al. 2007). Therefore, these findings suggest that N/C interaction may block inappropriate coregulator interaction until the AR is engaged with AREs in the promoter and enhancer regions of target genes. After AR is bound to DNA, the AF2 cleft may be more amenable to coregulator binding, which would enhance the transcriptional activity of the AR.

3.2

Structure and Function of the AR DNA-Binding and Hinge Domains

The AR DBD/hinge region plays important roles in mediating AR nuclear localization, receptor dimerization, and DNA binding. Hormone receptor DBDs are conserved and consist of two zinc fingers and a loosely structured carboxy-terminal extension (CTE) (Verrijdt et al. 2003). The first zinc finger contains the P-box, which is the recognition helix that binds the DNA major groove. By virtue of perfect homology in this P-box, AR, glucocorticoid receptor (GR), progesterone receptor (PR), and mineralcorticoid receptor (MR) are defined as class I nuclear receptors. The second zinc finger contains the D-box, which is the AR dimerization interface. A peptide fragment containing only the AR DBD and CTE can recapitulate perfectly the DNA-binding site selectivity and specificity as well as dimerization of full-length receptor. The recent crystal structure of the AR DBD complexed with a prototype ARE revealed that AR unexpectedly binds in a head-to-head conformation (Shaffer et al. 2004). This contrasts with other nuclear receptor DBDs, which adopt the orientation of the DNA target. The AR head-to-head binding orientation on a direct repeat ARE demonstrates that the strength of the AR dimer interface plays a major role in AR DNA-binding site specificity. This also indicates that the principles obtained from study of the AR DBD engaged with a prototype ARE likely extend to the binding of AR to other direct repeat and inverted repeat palindromes. The sequence RKcyeagmtlgaRKLKK, which overlaps the DBD D-box, the CTE, and the hinge region, encodes the bipartite AR nuclear localization signal

438

S.M. Dehm, D.J. Tindall

(Zhou et al. 1994). The basic motifs at either end of this sequence are critical for AR nuclear import, with the intervening sequence of little importance. Thus, AR nuclear transport may be allowed when ligand binding induces a conformational change that makes this NLS accessible. In addition, the lysine residues within the KLKK motif are direct acetylation targets of the p300 and Tip60 acetyltransferases (Fu et al. 2000; Gaughan et al. 2002).

3.3

Structure and Function of the AR NH2-Terminal Domain

The AR NTD is highly flexible and displays intrinsic disorder in solution, which has hampered elucidation of its 3-dimensional structure (Bain et al. 2006; Lavery and McEwan 2005; Lavery and McEwan 2006; McEwan 2004). Biophysical study of the NTD has revealed that it exists in a molten globule conformation (Lavery and McEwan 2006). Thus, the AR NTD could be envisioned as regions of rigid secondary structure, or ‘‘sticky patches’’ that are either buried or exposed in response to cell type, androgen level, post-translational modification, expression/ activity of binding partners, and chromatin environment. The AR NTD accounts for over 60% of the AR protein and functions as a potent transcriptional activator independent of the CTD. This NTD activity, generally referred to as AR AF-1, contrasts with AF-2, which is a relatively weak transcriptional activation domain in isolation (Alen et al. 1999; Bevan et al. 1999; He et al. 1999). AR AF-1 is thought to be the major domain responsible for mediating AR transcriptional activity. Several studies have made progress toward identifying functional regions within the AR NTD, and have provided glimpses into the overall mechanisms of AR regulation in prostate cancer and other cell types. The strong transcriptional activity of AF-1 in the AR NTD maps to two large domains, termed transactivation unit (TAU)1 and TAU5. TAU1 and TAU5 were identified originally as the primary transactivation units within the AR NTD following functional deletion analysis of wild-type and CTD-deleted AR in COS-1 and HeLa cells (Jenster et al. 1995). Further deletion studies with rat AR demonstrated that TAU1 could be divided into two discrete transcriptional activation domains, termed AF-1a and AF-1b (Chamberlain et al. 1996). Transcriptional activity of the AR AF-1a domain has been observed for both rat and human AR in cell lines of various origins (Callewaert et al. 2006; Chamberlain et al. 1996; Dehm and Tindall 2006a). Subsequent studies defined the core sequence 178LKDIL182, resident within a putative NTD helix, as a key motif that can mediate autonomously AR TAU1 transcriptional activity (Callewaert et al. 2006; Chamberlain et al. 1996). AR AF-1b resembles an acidic activation domain, and deletion of this region from the rat or human AR impairs transcriptional activity in CV-1 and LNCaP cells (Chamberlain et al. 1996; Dehm and Tindall 2006a). However, deleting this fragment has no effect on AR transcriptional activity in COS-7 cells (Callewaert et al. 2006; Dehm and Tindall 2006a), which suggests cell-type specific roles for this putative transcriptional activation domain. Similar conflicting findings have

Ligand-Independent Androgen Receptor Activity

439

been reported for TAU5, with deletion of this fragment either impairing (Callewaert et al. 2006; Chamberlain et al. 1996; Jenster et al. 1995) or not affecting (Christiaens et al. 2002; Metzger et al. 2003) AR transcriptional activity. TAU5 may be an important AR module through which SRC-1 and the protein kinase C-related kinase (PRK1) can exert their influence as coactivators (Callewaert et al. 2006; Metzger et al. 2003). Overall, these studies highlight the fact that several discrete transcriptional activation domains are located within the AR NTD. The amorphous nature of the NTD is likely to explain the conflicting reports on the relative roles of these transcriptional activation domains in different cell lines and on different promoters. In addition to mediating an N/C interaction with the AR AF-2 domain, the FxxLF motif in the AR NTD has been shown to bind the X-chromosome-linked melanoma antigen gene product, MAGE-11 (Bai et al. 2005). MAGE-11 has been defined as an AR coactivator that stabilizes AR protein and enhances its transcriptional activity. Mechanistically, this appears to result from MAGE-11 competition with AF-2 for binding to FxxLF. For example, MAGE-11 binding to FxxLF increases AF-2 accessibility to LxxLL-containing p160 coactivators such as TIF2 (Bai et al. 2005). This finding serves as an excellent example of how expression levels of specific coregulator molecules can influence drastically the relative architecture of the AR NTD, and thus the relative activities of AR transcriptional activation domains. Overlapping the AF-1a LKDIL motif is an LSEASTMQLL (LX7LL) motif, which is evolutionarily conserved among sex steroid receptors. LX7LL serves as a binding site for TAB2 as a component of an NCoR corepressor complex (Zhu et al. 2006). When AR is bound to antiandrogens such as bicalutamide, the TAB2/ NCoR corepressor complex binds AR LX7LL, which results in histone deacetylase (HDAC)-dependent transcriptional repression of AR-regulated promoters such as PSA. However, inflammatory signals initiated by IL-1b stimulate recruitment of MEKK1 kinase to the PSA promoter, which causes the TAB2/NCoR corepressor complex to dissociate from bicalutamide-bound AR (Zhu et al. 2006). As a result of these events, AR de-repression occurs, which allows aberrant recruitment of coactivators and subsequent transcriptional activation in the presence of bicalutamide (Baek et al. 2006). These findings suggest a mechanism by which infiltration of inflammatory cells ultimately influences how the AR NTD modulates the prostate cancer cell response to agonists or antagonists.

3.4

Domains Required for Regulated and Deregulated AR Activity

The majority of our work on androgen-independent AR activation has involved an ‘‘outside-in’’ approach, wherein cells were treated with exogenous compounds, such as IL-6, and effects on AR activity observed. The isogenic LNCaP/C4-2 cell model of prostate cancer progression was used to study androgen-independent AR activation in an alternative model system that would permit the comparison of

440

S.M. Dehm, D.J. Tindall

regulated and deregulated mechanisms of AR activation. C4-2 cells grow in an androgen-independent manner in vivo and in vitro, and were derived from androgen-sensitive LNCaP cells through serial passage as xenografts in castrated athymic male mice (Wu et al. 1994). C4-2 cells displayed constitutive PSA expression in androgen-free medium, which was on average threefold to four4 fold higher than in LNCaP cells. PSA mRNA expression was decreased when AR expression was inhibited using AR-targeted siRNAs. These experiments indicated that the AR was aberrantly active in the absence of androgens in C4-2 cells. Differential AR activity on the PSA gene occurred even though LNCaP and C4-2 cells exhibited similar AR gene sequence, expression levels, and post-translational modification. Overall, these results suggested that the C4-2 and LNCaP cellular milieu were different, and this difference must support aberrant AR activity in C4-2 cells.

3.4.1

Role of the AR CTD

These findings prompted study of AR structure/function relationships in LNCaP and C4-2 cells. This presented a challenge, because LNCaP and C4-2 express endogenous AR protein, and all AR functional studies to date had been carried out in AR-null cell models. An AR-Gal4 fusion strategy was developed to allow for AR structure/function studies in any cell line, regardless of whether it is AR positive or AR negative (Dehm and Tindall 2006a). This strategy was based on the premise that swapping AREs in PSA enhancer-based reporter constructs with binding sites for the Gal4 yeast transcription factor would render these constructs responsive to hARGal4, a hybrid protein wherein the zinc-finger AR DBD was swapped with the zinc-finger Gal4 DBD. To corroborate findings garnered from this highly engineered reporter system, an AR replacement strategy was designed that relied on the generation of plasmid-based, siRNA-resistant forms of the AR (Dehm et al. 2007). Cotransfection of AR-expressing prostate cancer cells with ARtargeted siRNA and siRNA-resistant versions of AR would allow knock-down of endogenous AR protein, and concurrent replacement with ectopic AR protein. Both these approaches recapitulated salient features of AR activity in LNCaP and C4-2 cells. First, employing the hARGal4 reporter system as well as using the AR replacement strategy, AR was inactive in the absence of androgens in LNCaP cells, but responded robustly following androgen treatment (Dehm et al. 2007; Dehm and Tindall 2006a). Second, using these two approaches, AR displayed constitutive activity in the absence of androgens in C4-2 cells, which could be further enhanced via treatment with androgens (Dehm et al. 2007; Dehm and Tindall 2006a). These systems were exploited to ask some fundamental questions about the nature of AR activity in LNCaP and C4-2 cells. First is aberrant AR activity in the C4-2 cell line truly ligand independent, or does it result from hypersensitivity to residual androgens? The function of the AR LBD was abolished by introducing mutations that block the ability of the AR to bind ligand (Fig. 3). These mutations blocked the ability of AR to activate transcription in a ligand-dependent manner in

Ligand-Independent Androgen Receptor Activity

441

LNCaP cells, but there was no effect of these mutations on aberrant AR activity in C4-2 cells (Dehm and Tindall 2006a). This indicated that the aberrant activity observed in this cell line was indeed ligand independent. Next, does the AR AF-2 coactivator-binding surface play a role in mediating aberrant AR activity in C4-2 cells? Mutations were introduced in the charge clamp of the AR AF-2 coactivatorbinding surface to block AF-2 function (Fig. 3). As expected, these mutations compromised androgen-induced AR activity in LNCaP cells. Conversely, these AF-2 mutations were unable to block ligand-independent AR activity in C4-2 cells (Dehm and Tindall 2006a). Taken together, these results demonstrated that constitutive AR activity was independent of the C-terminal LBD and AF-2 motifs in androgen-independent C4-2 cells. These findings had important ramifications, because both androgen deprivation (castration) and AR antagonists (bicalutamide, flutamide) inhibit AR activity through the CTD of the AR protein. These findings also raised the possibility that antiandrogens such as bicalutamide may be ineffective at blocking ligand-independent AR activity in castration-recurrent prostate cancer cells due to the AF-2 independent nature of this mode of activation. To probe this hypothesis, the effect of bicalutamide was tested on androgen-dependent and -independent AR activity in LNCaP and C4-2 cells, respectively. In LNCaP cells, bicalutamide inhibited the androgen-dependent induction of PSA expression. However, in C4-2 cells, constitutive PSA expression was unaffected or slightly increased in the presence of bicalutamide. Moreover, bicalutamide had no effect on the androgen-independent growth of C4-2 cells (Dehm and Tindall 2006a). Overall, our data suggested that C4-2 cells were androgen independent due to constitutive, ligand-independent, AF-2-independent AR activity.

3.4.2

Role of the AR NTD

These findings essentially ruled out a major role for the AR CTD in mediating aberrant AR activity in C4-2 cells. Attention was shifted to the large NTD of the AR, which makes up over 60% of the AR protein. Because the structure of the AR NTD is not known, and functional motifs had not been well defined, large deletions were generated within the AR NTD, and the effects of these deletions were tested on ligand-dependent and ligand-independent AR activity in LNCaP and C4-2 cells, respectively. Several AR NTD deletions impaired both ligand-independent and ligand-dependent AR activity (Dehm and Tindall 2006a). These studies focused attention on one deletion, which encompassed the TAU5 domain (Fig. 3). Deletion of the TAU5 domain from the AR resulted in a two-fold increase in liganddependent activity, but a twofold decrease in ligand-independent AR activity (Dehm et al. 2007; Dehm and Tindall 2006a). This finding suggested that TAU5 activity was required for ligand-independent AR transcriptional activity in androgen-independent C4-2 cells. Androgen concentrations studied were zero and 1 nM, which is considered to be a physiological circulating androgen level. However, studies have demonstrated that although a complete androgen-depleted environment can be achieved in vitro,

442

S.M. Dehm, D.J. Tindall

such a situation may not arise in vivo (Mohler et al. 2004). To address this issue, the effect of TAU5 deletion over a range of concentrations of androgens was studied. In LNCaP cells, no AR transcriptional activity was observed at DHT concentrations lower than 0.1 nM. Conversely, in C4-2 cells, AR transcriptional activity was high in the absence of androgens, as well as at concentrations lower than 0.1 nM. Deletion of TAU5 resulted in approximately 50% lower AR transcriptional activity over all DHT concentrations studied (Dehm et al. 2007). These results suggested more strongly that TAU5 played an important role in mediating ligand-independent AR activation, and showed that TAU5 also played a role in mediating ligandhypersensitive AR activity in androgen-independent prostate cancer cells.

The AR WHTLF Motif is a Novel Transcriptional Activation Domain What are the molecular mechanism(s) by which the TAU5 domain participates in aberrant AR activation in castration-recurrent prostate cancer cells? The finding appeared most consistent with the existence of a transactivation domain within AR TAU5, which could account for approximately 50% of the ligand-independent and/ or ligand-hypersensitive AR activity in androgen-independent prostate cancer cells. To more specifically define a putative region within TAU5 that could account for this activity, protein secondary structure prediction algorithms were used to guide smaller NTD deletions and alanine scanning mutagenesis. Together, these approaches revealed a span of predicted secondary structure with the core amino acid sequence 435WHTLF439 (WxxLF, Fig. 3), which was located within TAU5 (Dehm et al. 2007). This region demonstrated strong potential to be the sought-after core motif that was responsible for mediating the selective activity of the TAU5 domain. For example, mutation of this motif to AHTAA impaired ligand-independent AR activity to the same extent as TAU5 deletion in C4-2 cells, but had no effect on ligand-dependent AR activity in LNCaP cells. This motif was studied in more detail since manipulating the WHTLF core sequence and deleting TAU5 from the AR selectively impaired ligand-independent AR activity. Manipulating the AR in two separate ways (i.e., deleted TAU5 and mutated WHTLF to AHTAA) produced similar effects on ligand-independent AR activity. However, did both manipulations alter the same mechanism, or were two separate and unrelated mechanisms involved? To address this issue, TAU5 was deleted from the AR and a 21-amino acid cassette that contained either the core WHTLF motif or a mutated AHTAA motif was inserted. The presence of the 21-amino acid WHTLF insert, but not the AHTAA insert, restored full ligand-independent AR activity in androgen-independent prostate cancer cells (Dehm et al. 2007). These findings confirmed that the AR WHTLF motif was playing a direct role in mediating the aberrant activity of TAU5. The AR WHTLF motif had previously been shown to mediate an interaction with the AR AF-2 domain. Therefore, the possibility existed that aberrant AR activation could be facilitated by stabilization of the AR in the absence of ligand via an aberrant N/C interaction. However, mutations in the AR CTD abolished AF-

Ligand-Independent Androgen Receptor Activity

443

2 function but had no effect on ligand-independent AR activity. Therefore, the possibility that WHTLF could function as a transcriptional activation surface by engaging with specific coregulatory molecules or components of the basal transcriptional machinery was explored. In this scenario, if the function of WHTLF was to recruit accessory proteins, this motif would be able to perform this function in isolation. To test this possibility, multiple copies of a peptide encompassing WHTLF were fused to a heterologous DNA-binding domain, and the ability of fusion proteins to activate transcription was assayed. These fusions functioned as potent transcriptional activators. The same peptides containing an AHTAA core sequence had no transactivation property (Dehm et al. 2007). These critical experiments led to the conclusion that that the AR WHTLF motif was a novel transcriptional activation domain, which selectively participated in mediating aberrant AR transcriptional activity in androgen-independent prostate cancer cells.

4 Implications and Conclusions Data from our laboratory and others have implicated the AR as a master regulator of the growth and survival of androgen-dependent as well as androgen-independent prostate cancer cells. This is an interesting conundrum indicating that the AR remains an attractive molecular target for therapy of prostate cancer cells that have become resistant to AR-targeted therapies. Based on current knowledge, what are some of the pathways and/or mechanisms that may be targeted in order to inhibit AR activity in castration-recurrent prostate cancer?

4.1

IL-6 and p300

Treatment of prostate cancer cells with IL-6, a cytokine that is found at elevated levels in serum of patients with advanced prostate cancer, can activate the AR in a ligand-independent manner. The AR-associated coactivator, p300, is required for this effect, via activation of the MAPK pathway (Debes et al. 2002). Further supporting the potential importance of this mechanism is the finding that p300 protein levels are elevated following androgen-deprivation therapy and these elevations are found in castration-recurrent prostate cancer (Debes et al. 2003).

4.1.1

Implications for Therapy

Elucidation of the IL-6/MAPK/p300/AR pathway (Fig. 2) provides several molecules that could be targeted for inhibition, either alone or in combination. First, inhibition of the IL-6 receptor, via targeted monoclonal antibodies or specific druglike small molecules, could prevent IL-6-mediated activation of the AR. The

444

S.M. Dehm, D.J. Tindall

MAPK pathway and p300 are required downstream of IL-6 for AR activation (Debes et al. 2002). Both ERK and p300 are enzymes that carry out phosphorylation and acetylation reactions, respectively. Since enzyme active sites are attractive targets for small molecule drug design, these proteins have the potential to be inhibited in a specific fashion. The ERK inhibitor, PD98059, can block the effects of IL-6 on AR activity, thus providing proof of principle for such an approach (Debes et al. 2002). Moreover, p300 enzymatic activity is critical for IL-6 mediated AR activation, because overexpression of wild-type p300, but not a mutant form that lacks acetyltransferase activity, can overcome the adenovirus E1A-mediated p300 sequestration (Debes et al. 2002). Therefore, specific p300 inhibitors could be tested for their effects on prostate cancer cells. The potential effectiveness of this approach is supported by the finding that siRNA-mediated knock-down of p300 inhibits the proliferation of androgen-dependent as well as androgen-independent prostate cancer cells (Heemers et al. 2007).

4.2

Functional Role of Discrete AR Domains

In addition to these pathways upstream of the AR, specific domains of the AR mediate activity in androgen-dependent and androgen-independent prostate cancer cells. In androgen-dependent prostate cancer cells, both the LBD and AF-2 coactivator-binding surfaces are critical for ligand-stimulated AR activity. Antiandrogens, such as bicalutamide, rely on this CTD dependence to inhibit AR activity, and exert their effect by binding the LBD and blocking the activity of AF-2. In androgen-independent prostate cancer cells, the AR is active in the absence of androgens. Mechanistically, this activity is independent of the AR LBD and AF-2 modules. This mode of activity also underlies resistance to antiandrogens such as bicalutamide (Dehm and Tindall 2006a). Therefore, these results suggest that any CTD-directed therapies would be unable to inhibit ligand-independent AR activity. Instead of targeting the AR CTD, targeting discrete NTD motifs may represent a novel and potentially durable alternative. Specifically, a novel AR transcriptional activation domain, mapped to a core WHTLF sequence, is required for ligandindependent AR activity (Dehm et al. 2007). This AR motif may function as a protein–protein interaction interface, which is exposed and accessible in the absence of androgens in androgen-independent prostate cancer cells. Coregulators and/or components of the basal transcriptional apparatus that may interact with this motif remain unknown.

4.2.1

Implications for Therapy

The best therapy for castration-recurrent prostate cancer would be one that completely and specifically abolishes AR activity. Androgen-deprivation therapy inhibits ligand-dependent AR activity. However, agents that specifically inhibit

Ligand-Independent Androgen Receptor Activity

445

ligand-independent AR activity have not been developed. Inhibition of the AR NTD, in general, or the AR WHTLF motif, in particular, may provide targets for the development of such agents. This rationale is supported by recent findings that decoy peptides representing the entire AR NTD effectively blocked the growth of androgen-dependent prostate cancer cells in vitro and in vivo, and also delayed progression to androgen-independent disease (Quayle et al. 2007). An attractive avenue for specific inhibition of the AR WHTLF motif would be to identify the proteins that dock with this site and design compounds to block this interaction. In addition, if proteins docking with the AR WHTLF motif have enzymatic activities, these enzyme active sites could be targeted to effectively block activity of this AR domain.

4.3

Conclusions

In summary and in retrospect, our laboratory has been privileged to contribute to the exciting body of work regarding the role of the AR in prostate cancer. Studies by our laboratory and others have come full circle over the past decade to demonstrate that the AR is critical for prostate cancer growth both before and during androgendeprivation therapy. With all the outstanding work currently focused on the role of the AR and prostate cancer, it is likely that novel modes of AR inhibition will be developed in the laboratory for testing in clinical trials. An important consideration will be whether these new and effective modes of AR inhibition will fail eventually through various modes of therapy circumvention, or if these approaches will provide a durable response and perhaps even a cure for advanced prostate cancer.

References Adler HL, McCurdy MA, Kattan MW, Timme TL, Scardino PT, Thompson TC (1999). Elevated levels of circulating interleukin-6 and transforming growth factor-beta1 in patients with metastatic prostatic carcinoma. J Urol 161: 182–7 Alen P, Claessens F, Verhoeven G, Rombauts W, Peeters B (1999). The androgen receptor aminoterminal domain plays a key role in p160 coactivator-stimulated gene transcription. Mol Cell Biol 19: 6085–97 Baek SH, Ohgi KA, Nelson CA, Welsbie D, Chen C, Sawyers CLet alet al. (2006). Ligand-specific allosteric regulation of coactivator functions of androgen receptor in prostate cancer cells. Proc Natl Acad Sci U S A 103: 3100–5 Bai S, He B, Wilson EM (2005). Melanoma antigen gene protein MAGE-11 regulates androgen receptor function by modulating the interdomain interaction. Mol Cell Biol 25: 1238–57 Bain DL, Heneghan AF, Connaghan-Jones KD, Miura MT (2006). Nuclear receptor structure: implications for function. Annu Rev Physiol 69 Bentel JM, Tilley WD (1996). Androgen receptors in prostate cancer. J Endocrinol 151: 1–11 Bevan CL, Hoare S, Claessens F, Heery DM, Parker MG (1999). The AF1 and AF2 domains of the androgen receptor interact with distinct regions of SRC1. Mol Cell Biol 19: 8383–92

446

S.M. Dehm, D.J. Tindall

Bubendorf L, Kononen J, Koivisto P, Schraml P, Moch H, Gasser TCet alet al. (1999). Survey of gene amplifications during prostate cancer progression by high-throughput fluorescence in situ hybridization on tissue microarrays. Cancer Res 59: 803–6 Buchanan G, Greenberg NM, Scher HI, Harris JM, Marshall VR, Tilley WD (2001a). Collocation of androgen receptor gene mutations in prostate cancer. Clin Cancer Res 7: 1273–81 Buchanan G, Irvine RA, Coetzee GA, Tilley WD (2001b). Contribution of the androgen receptor to prostate cancer predisposition and progression. Cancer Metastasis Rev 20: 207–23 Buchanan G, Yang M, Harris JM, Nahm HS, Han G, Moore Net alet al. (2001c). Mutations at the boundary of the hinge and ligand binding domain of the androgen receptor confer increased transactivation function. Mol Endocrinol 15: 46–56 Callewaert L, Van Tilborgh N, Claessens F (2006). Interplay between two hormone-independent activation domains in the androgen receptor. Cancer Res 66: 543–53 Chakravarti D, LaMorte VJ, Nelson MC, Nakajima T, Schulman IG, Juguilon Het alet al. (1996). Role of CBP/P300 in nuclear receptor signalling. Nature 383: 99–103 Chamberlain NL, Whitacre DC, Miesfeld RL (1996). Delineation of two distinct type 1 activation functions in the androgen receptor amino-terminal domain. J Biol Chem 271: 26772–8 Chen CD, Welsbie DS, Tran C, Baek SH, Chen R, Vessella Ret alet al. (2004). Molecular determinants of resistance to antiandrogen therapy. Nat Med 10: 33–9 (Epub 2003 Dec 21) Christiaens V, Bevan CL, Callewaert L, Haelens A, Verrijdt G, Rombauts Wet alet al. (2002). Characterization of the two coactivator-interacting surfaces of the androgen receptor and their relative role in transcriptional control. J Biol Chem 277: 49230–7 (Epub 2002 Oct 4) Craft N, Shostak Y, Carey M, Sawyers CL (1999). A mechanism for hormone-independent prostate cancer through modulation of androgen receptor signaling by the HER-2/neu tyrosine kinase. Nat Med 5: 280–5 Culig Z, Hobisch A, Cronauer MV, Cato AC, Hittmair A, Radmayr Cet alet al. (1993). Mutant androgen receptor detected in an advanced-stage prostatic carcinoma is activated by adrenal androgens and progesterone. Mol Endocrinol 7: 1541–50 Culig Z, Hobisch A, Cronauer MV, Radmayr C, Trapman J, Hittmair Aet alet al. (1994). Androgen receptor activation in prostatic tumor cell lines by insulin-like growth factor-I, keratinocyte growth factor, and epidermal growth factor. Cancer Res 54: 5474–8 Culig Z, Stober J, Gast A, Peterziel H, Hobisch A, Radmayr Cet alet al. (1996). Activation of two mutant androgen receptors from human prostatic carcinoma by adrenal androgens and metabolic derivatives of testosterone. Cancer Detect Prev 20: 68–75 Dai JL, Burnstein KL (1996). Two androgen response elements in the androgen receptor coding region are required for cell-specific up-regulation of receptor messenger RNA. Mol Endocrinol 10: 1582–94 Debes JD, Schmidt LJ, Huang H, Tindall DJ (2002). p300 mediates androgen-independent transactivation of the androgen receptor by interleukin 6. Cancer Res 62: 5632–6 Debes JD, Sebo TJ, Lohse CM, Murphy LM, Haugen de AL, Tindall DJ (2003). p300 in prostate cancer proliferation and progression. Cancer Res 63: 7638–40 Dehm SM, Tindall DJ (2006a). Ligand-independent androgen receptor activity is activation function-2-independent and resistant to antiandrogens in androgen refractory prostate cancer cells. J Biol Chem 281: 27882–93 Dehm SM, Tindall DJ (2006b). Molecular regulation of androgen action in prostate cancer. J Cell Biochem 99: 333–44 Dehm SM, Regan KM, Schmidt LJ, Tindall DJ (2007). Selective role of an NH2-terminal WxxLF motif for aberrant androgen receptor activation in androgen depletion independent prostate cancer cells. Cancer Res 67: 10067–77 Drachenberg DE, Elgamal AA, Rowbotham R, Peterson M, Murphy GP (1999). Circulating levels of interleukin-6 in patients with hormone refractory prostate cancer. Prostate 41: 127–33 Eder IE, Culig Z, Ramoner R, Thurnher M, Putz T, Nessler-Menardi Cet alet al. (2000). Inhibition of LncaP prostate cancer cells by means of androgen receptor antisense oligonucleotides. Cancer Gene Ther 7: 997–1007

Ligand-Independent Androgen Receptor Activity

447

Edwards J, Krishna NS, Grigor KM, Bartlett JM (2003). Androgen receptor gene amplification and protein expression in hormone refractory prostate cancer. Br J Cancer 89: 552–6 Estebanez-Perpina E, Moore JM, Mar E, Delgado-Rodrigues E, Nguyen P, Baxter JDet alet al. (2005). The molecular mechanisms of coactivator utilization in ligand-dependent transactivation by the androgen receptor. J Biol Chem 280: 8060–8 (Epub 2004 Nov 24) Ford OH, 3rd, Gregory CW, Kim D, Smitherman AB, Mohler JL (2003). Androgen receptor gene amplification and protein expression in recurrent prostate cancer. J Urol 170: 1817–21 Fu M, Wang C, Reutens AT, Wang J, Angeletti RH, Siconolfi-Baez Let alet al. (2000). p300 and p300/cAMP-response element-binding protein-associated factor acetylate the androgen receptor at sites governing hormone-dependent transactivation. J Biol Chem 275: 20853–60 Gaughan L, Logan IR, Cook S, Neal DE, Robson CN (2002). Tip60 and histone deacetylase 1 regulate androgen receptor activity through changes to the acetylation status of the receptor. J Biol Chem 277: 25904–13 (Epub 2002 May 6) Gottlieb B, Beitel LK, Wu JH, Trifiro M (2004). The androgen receptor gene mutations database (ARDB): 2004 update. Hum Mutat 23: 527–33 Grossmann ME, Huang H, Tindall DJ (2001). Androgen receptor signaling in androgen-refractory prostate cancer. J Natl Cancer Inst 93: 1687–97 Haag P, Bektic J, Bartsch G, Klocker H, Eder IE (2005). Androgen receptor down regulation by small interference RNA induces cell growth inhibition in androgen sensitive as well as in androgen independent prostate cancer cells. J Steroid Biochem Mol Biol 24: 24 Hammond GL, Avvakumov GV, Muller YA (2003). Structure/function analyses of human sex hormone-binding globulin: effects of zinc on steroid-binding specificity. J Steroid Biochem Mol Biol 85: 195–200 He B, Kemppainen JA, Voegel JJ, Gronemeyer H, Wilson EM (1999). Activation function 2 in the human androgen receptor ligand binding domain mediates interdomain communication with the NH(2)-terminal domain. J Biol Chem 274: 37219–25 He B, Kemppainen JA, Wilson EM (2000). FXXLF and WXXLF sequences mediate the NH2terminal interaction with the ligand binding domain of the androgen receptor. J Biol Chem 275: 22986–94 He B, Gampe RT, Jr, Kole AJ, Hnat AT, Stanley TB, An Get alet al. (2004). Structural basis for androgen receptor interdomain and coactivator interactions suggests a transition in nuclear receptor activation function dominance. Mol Cell 16: 425–38 Heemers HV, Sebo TJ, Debes JD, Regan KM, Raclaw KA, Murphy LMet alet al. (2007). Androgen deprivation increases p300 expression in prostate cancer cells. Cancer Res 67: 3422–30 Heinlein CA, Chang C (2004). Androgen receptor in prostate cancer. Endocr Rev 25: 276–308 Hobisch A, Eder IE, Putz T, Horninger W, Bartsch G, Klocker Het alet al. (1998). Interleukin-6 regulates prostate-specific protein expression in prostate carcinoma cells by activation of the androgen receptor. Cancer Res 58: 4640–5 Hur E, Pfaff SJ, Payne ES, Gron H, Buehrer BM, Fletterick RJ (2004). Recognition and accommodation at the androgen receptor coactivator binding interface. PLoS Biol 2: E274 Jemal A, Murray T, Samuels A, Ghafoor A, Ward E, Thun MJ (2003). Cancer statistics, 2003. CA Cancer J Clin 53: 5–26 Jenster G, van der Korput HA, Trapman J, Brinkmann AO (1995). Identification of two transcription activation units in the N-terminal domain of the human androgen receptor. J Biol Chem 270: 7341–6 Klokk TI, Kurys P, Elbi C, Nagaich AK, Hendarwanto A, Slagsvold Tet alet al. (2006). Ligandspecific dynamics of the androgen receptor at its response element in living cells. Mol Cell Biol 27: 1823–43 Lamb DJ, Weigel NL, Marcelli M (2001). Androgen receptors and their biology. Vitam Horm 62: 199–230 Latil A, Bieche I, Vidaud D, Lidereau R, Berthon P, Cussenot Oet alet al. (2001). Evaluation of androgen, estrogen (ER alpha and ER beta), and progesterone receptor expression in human

448

S.M. Dehm, D.J. Tindall

prostate cancer by real-time quantitative reverse transcription-polymerase chain reaction assays. Cancer Res 61: 1919–26 Lavery DN, McEwan IJ (2005). Structure and function of steroid receptor AF1 transactivation domains: induction of active conformations. Biochem J 391: 449–64 Lavery DN, McEwan IJ (2006). The human androgen receptor AF1 transactivation domain: interactions with transcription factor IIF and molten-globule-like structural characteristics. Biochem Soc Trans 34: 1054–7 Lee SO, Lou W, Hou M, de Miguel F, Gerber L, Gao AC (2003a). Interleukin-6 promotes androgen-independent growth in LNCaP human prostate cancer cells. Clin Cancer Res 9: 370–6 Lee SO, Lou W, Hou M, Onate SA, Gao AC (2003b). Interleukin-4 enhances prostate-specific antigen expression by activation of the androgen receptor and Akt pathway. Oncogene 22: 7981–8 Li Y, Lambert MH, Xu HE (2003). Activation of nuclear receptors: a perspective from structural genomics. Structure 11: 741–6 Li J, Fu J, Toumazou C, Yoon HG, Wong J (2006). A role of the amino-terminal (N) and carboxylterminal (C) interaction in binding of androgen receptor to chromatin. Mol Endocrinol 20: 776–85 Liao X, Tang S, Thrasher JB, Griebling TL, Li B (2005). Small-interfering RNA-induced androgen receptor silencing leads to apoptotic cell death in prostate cancer. Mol Cancer Ther 4: 505–15 Linja MJ, Savinainen KJ, Saramaki OR, Tammela TL, Vessella RL, Visakorpi T (2001). Amplification and overexpression of androgen receptor gene in hormone-refractory prostate cancer. Cancer Res 61: 3550–5 Matias PM, Donner P, Coelho R, Thomaz M, Peixoto C, Macedo Set alet al. (2000). Structural evidence for ligand specificity in the binding domain of the human androgen receptor. Implications for pathogenic gene mutations. J Biol Chem 275: 26164–71 McEwan IJ (2004). Molecular mechanisms of androgen receptor-mediated gene regulation: structure-function analysis of the AF-1 domain. Endocr Relat Cancer 11: 281–93 Metzger E, Muller JM, Ferrari S, Buettner R, Schule R (2003). A novel inducible transactivation domain in the androgen receptor: implications for PRK in prostate cancer. Embo J 22: 270–80 Mohler JL, Gregory CW, Ford OH, 3rd, Kim D, Weaver CM, Petrusz Pet alet al. (2004). The androgen axis in recurrent prostate cancer. Clin Cancer Res 10: 440–8 Quayle SN, Mawji NR, Wang J, Sadar MD (2007). Androgen receptor decoy molecules block the growth of prostate cancer. Proc Natl Acad Sci U S A 104: 1331–6 Sack JS, Kish KF, Wang C, Attar RM, Kiefer SE, An Yet alet al. (2001). Crystallographic structures of the ligand-binding domains of the androgen receptor and its T877A mutant complexed with the natural agonist dihydrotestosterone. Proc Natl Acad Sci U S A 98: 4904–9 Scherr D, Swindle PW, Scardino PT (2003). National Comprehensive Cancer Network guidelines for the management of prostate cancer. Urology 61: 14–24 Shaffer PL, Jivan A, Dollins DE, Claessens F, Gewirth DT (2004). Structural basis of androgen receptor binding to selective androgen response elements. Proc Natl Acad Sci U S A 101: 4758–63 Sung SY, Chung LW (2002). Prostate tumor-stroma interaction: molecular mechanisms and opportunities for therapeutic targeting. Differentiation 70: 506–21 Taplin ME, Rajeshkumar B, Halabi S, Werner CP, Woda BA, Picus Jet alet al. (2003). Androgen receptor mutations in androgen-independent prostate cancer: Cancer and Leukemia Group B Study 9663. J Clin Oncol 21: 2673–8 Ueda T, Bruchovsky N, Sadar MD (2002). Activation of the androgen receptor N-terminal domain by interleukin-6 via MAPK and STAT3 signal transduction pathways. J Biol Chem 277: 7076– 85 (Epub 2001 Dec 19)

Ligand-Independent Androgen Receptor Activity

449

van Royen ME, Cunha SM, Brink MC, Mattern KA, Nigg AL, Dubbink HJet alet al. (2007). Compartmentalization of androgen receptor protein-protein interactions in living cells. J Cell Biol 177: 63–72 Verrijdt G, Haelens A, Claessens F (2003). Selective DNA recognition by the androgen receptor as a mechanism for hormone-specific regulation of gene expression. Mol Genet Metab 78: 175–85 Visakorpi T, Hyytinen E, Koivisto P, Tanner M, Keinanen R, Palmberg Cet alet al. (1995). In vivo amplification of the androgen receptor gene and progression of human prostate cancer. Nat Genet 9: 401–6 Wang LG, Ossowski L, Ferrari AC (2004). Androgen receptor level controlled by a suppressor complex lost in an androgen-independent prostate cancer cell line. Oncogene 23: 5175–84 Wise GJ, Marella VK, Talluri G, Shirazian D (2000). Cytokine variations in patients with hormone treated prostate cancer. J Urol 164: 722–5 Wu HC, Hsieh JT, Gleave ME, Brown NM, Pathak S, Chung LW (1994). Derivation of androgenindependent human LNCaP prostatic cancer cell sublines: role of bone stromal cells. Int J Cancer 57: 406–12 Zhao XY, Malloy PJ, Krishnan AV, Swami S, Navone NM, Peehl DMet alet al. (2000). Glucocorticoids can promote androgen-independent growth of prostate cancer cells through a mutated androgen receptor. Nat Med 6: 703–6 Zhou ZX, Sar M, Simental JA, Lane MV, Wilson EM (1994). A ligand-dependent bipartite nuclear targeting signal in the human androgen receptor. Requirement for the DNA-binding domain and modulation by NH2-terminal and carboxyl-terminal sequences. J Biol Chem 269: 13115– 23 Zhou ZX, Lane MV, Kemppainen JA, French FS, Wilson EM (1995). Specificity of liganddependent androgen receptor stabilization: receptor domain interactions influence ligand dissociation and receptor stability. Mol Endocrinol 9: 208–18 Zhu P, Baek SH, Bourk EM, Ohgi KA, Garcia-Bassets I, Sanjo Het alet al. (2006). Macrophage/ cancer cell interactions mediate hormone resistance by a nuclear receptor derepression pathway. Cell 124: 615–29

Role of IL-6 in Regulating the Androgen Receptor Zoran Culig and Alfred Hobisch

Abstract Androgen receptor (AR) is expressed in human prostate cancer in which it could be activated by steroid hormones and nonsteroidal compounds. Interleukin-6 (IL-6) levels are increased in tissues and sera from patients with advanced prostate cancer. AR-negative prostate cancer cell lines express high levels of IL-6. This cytokine may activate signaling pathways of Janus kinase/ signal transducers and activators of transcription factors or mitogen-activated protein kinase, thus contributing to tumor cell growth. IL-6 activation of the AR was described by several investigators in various cell lines. This regulation of AR activity is of importance for cellular differentiation, as evidenced by increase of prostate-specific antigen expression. It was shown that the N-terminal region of the AR is required for activation by IL-6. The effect of IL-6 also depends on functional coactivators SRC-1 and p300. Interestingly, cells exposed to a chronic treatment with IL-6 acquire a more malignant phenotype and downregulate AR expression. Further studies are necessary to evaluate whether IL-6/AR axis could be targeted by novel therapies in prostate cancer.

1 Introduction Progression of prostate cancer toward castration-recurrent stage is associated with heterogenous expression of the androgen receptor (AR). On the one hand, there is an increased expression of AR mRNA and protein that ultimately results in enhanced receptor activation. On the other hand, lack of AR expression in some prostate cancer cells could be explained by epigenetic changes in the gene promoter. Increased AR activation by low concentration of hormones may be important for tumor progression since residual androgens persist after androgen ablation therapy (Titus et al. 2005). AR increase has been identified as the most Z. Culig(*) Department of Urology, Innsbruck Medical University, Anichstrasse 35, A-6020 Innsbruck, Austria, E-mail: [email protected]

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4_20, # Springer Science + Business Media, LLC 2009

451

452

Z. Culig, A. Hobisch

consistent alteration during prostate cancer progression in a number of xenograft models (Chen et al. 2004). In LNCaP sublines, which express the AR but do not respond to antiandrogen treatment, downregulation of the AR inhibits proliferation (Eder et al. 2000; Zegarra-Moro et al. 2002). In several studies, it was shown that the function of the AR may be potentiated by nonsteroidal compounds such as forskolin, ErbB2, growth factors, thyroid hormone, and proinflammatory cytokines (Nazareth and Weigel 1996; Yeh et al. 1999; Culig et al. 1994; Hobisch et al. 1998; Lee et al. 2003). Although the AR is similar to human progesterone and glucocorticoid receptors that are activated synergistically with ligand, ligand-independent effects of various cellular regulators on AR were reported in the literature and will be discussed elsewhere in this volume in detail. Ligand-independent activation is therefore of considerable interest for prostate cancer development and progression. Most studies focused on molecular mechanisms including identification of specific sequences of AR that are required for nonsteroidal activation and interaction with receptor coactivators. At present, it is difficult to assess clinical relevance of nonsteroidal AR activation. The effects in vitro are cell type- and promoter-specific, and the genes whose expression is regulated in this way in vivo have to be further investigated. Most notably, it was demonstrated that overexpression of ErbB2 leads to LNCaP tumor growth and prostate-specific antigen (PSA) increase in the absence of androgen (Craft et al. 1999). Those findings opened the possibilities for testing tyrosine kinase inhibitors in prostate cancer for interference with AR-mediated prostate cancer growth. In case of AR activation by compounds that increase intracellular cAMP levels, gene expression was profiled after treatment with either androgen or forskolin (Wang et al. 2006). That analysis showed that the number of genes regulated in a similar way by both compounds is limited. Thus, evidence supporting the role of cAMP in AR-dependent tumor progression is missing. There is, however, a more significant overlap in regulation of genes by androgen and the neuropeptide bombesin (Desai et al. 2006). In that study, 72 genes were found to be similarly regulated by androgen and bombesin. Bombesin-stimulated in vitro growth of AR-positive prostate cancer cells is inhibited by the antiandrogen bicalutamide. Further experiments aimed to investigate the role of bombesin in prostate cancer growth in vivo are therefore justified. The focus on this chapter is on interaction between the pleiotropic and proinflammatory cytokine interleukin-6 (IL-6) and the AR. The studies on IL-6 and androgen signaling were carried out mostly in prostate cancer cell lines. They improved understanding of the mechanisms by which this cytokine regulates cellular events in cancer. It will be necessary to extend this work on early lesions in prostate carcinogenesis such as prostate inflammatory atrophy or high-grade prostate intraepithelial neoplasia in the future. There is an increased number of patients with small cancers that will not necessarily become clinically manifest. In several institutions, those cancers are treated by radical surgery or radiotherapy, although little is known about their malignant potential. Chronic inflammation may be associated with early prostate cancer, and common alterations such as deletion of

Role of IL-6 in Regulating the Androgen Receptor

453

tumor suppressors were observed both in inflammation regions and in cancer (De Marzo et al. 2007). Modulation of AR activity by IL-6 and other cytokines may be therefore of interest also for tumor initiation and prevention.

2 IL-6 and Its Receptor in Prostate Cancer IL-6 is expressed in basal cells in the benign prostate and produced in cancer cells (Hobisch et al. 2000). In conditioned media obtained from primary stromal cultures, IL-6 was also measured. High IL-6 amounts are secreted into supernatants from PC-3 or DU-145 cells, whereas LNCaP were found to be IL-6-negative (Twillie et al. 1995). Several nonexclusive mechanisms may be responsible for increased IL-6 production during prostate carcinogenesis: it is known that steroid hormones repress expression of IL-6 and its receptor in bone cells, and it is likely that the same mechanism is operative in prostate cancer cell lines (Bellido et al. 1995). High expression of nuclear factor-kB, which is an upstream regulator of expression of several proinflammatory cytokines, is frequently observed in prostate cancer specimens and correlated with bad prognosis (Zerbini et al. 2003; Fradet et al. 2004). For this reason, chemopreventive strategies targeting nuclear factor-kB have been proposed (Gupta et al. 2002). In addition, the tumor suppressor retinoblastoma, frequently lost in prostate cancer, represses the promoter of the IL-6 gene (Santhanam et al. 1991). IL-6 is upregulated by transforming growth factor-b, which inhibits cellular proliferation in vitro but increases angiogenesis, metastatic spread, and attenuates immune response in vivo (Park et al. 2003). Activating protein-1 complex members also contribute to IL-6 upregulation in prostate cancer (Zerbini et al. 2003). Loss of IL-6 receptor has not been reported in human prostate cancer, and it is assumed that most prostate tumor cells have an intact IL-6 signal transduction cascade.

3 IL-6 Signaling in Cancer The role of IL-6 in regulation of cellular events in different cancers has been investigated. Most studies were conducted in leukemia, myeloma, melanoma, renal, gastric, and breast cancers. IL-6 is a cytokine that regulates proliferation and apoptosis by many different ways. It has been identified as a target for novel therapies due to its growth-stimulatory and antiapoptotic effects. The mechanisms responsible for different regulations are complex and only partially understood (Culig et al. 2005). The ligand-binding subunit of the IL-6 gp80 is widely expressed in benign and malignant tissues. The signal transducing subunit gp130 is common in signal transmission cascades of cytokines that are related to IL-6, such as IL-11 (Campbell et al. 2001). gp130 subunits of the IL-6 receptor homodimerize, and this process is followed by formation of a hexameric complex consisting of two

454

Z. Culig, A. Hobisch

molecules each of IL-6, gp80, and gp130. IL-6 action in target cells is also regulated by the soluble receptor, and this process is known as trans-signaling. Interestingly, the presence of the soluble receptor in prostate cancer tissue specimens has been associated with poor prognosis (Shariat et al. 2001). In a variety of cell lines, treatment with IL-6 leads to activation of the signaling pathway of Janus kinase/signal transducers and activators of transcription (JAK/ STAT). JAK are phosphorylated after binding of IL-6 to the receptor. JAK in turn phosphorylate the gp130 subunit and induce nuclear translocation of STAT3. Rapid phosphorylation of STAT3 occurs in target tissues after treatment with IL-6. Permanent activation of STAT3 and perhaps a chronic inflammatory response are prevented by induction of suppressors of cytokine signaling (SOCS) and other endogenous inhibitors, such as protein inhibitors of activated STAT. There is an increasing interest in the role of SOCS-3 and -1 in cancer. Overexpression of SOCS-3 in lung, liver, head, and neck cancer led to retardation of tumor growth (He et al. 2003; Niwa et al. 2005; Weber et al. 2005). Those functional studies were supported by pathomorphological analyses that revealed downregulation of SOCS-3 in cancer cells and biopsy specimens. However, this trend was not observed in prostate cancer. In comparison to benign tissues, expression of SOCS-3 was found to be higher in tumors (Bellezza et al. 2006). An inverse correlation between the expression of SOCS-3 and phosphorylation of STAT3 was found in prostate cancer cell lines. SOCS-3 is enhanced by androgen and acts as a negative feedback factor. Its upregulation leads to a diminished induction of cell cycle regulatory proteins (Neuwirt et al. 2007). The knowledge on cell type-specific activation of signaling pathways of JAK/ STAT, mitogen-activated protein kinase (MAPK), and phosphotidylinositol 3-kinase is important for understanding of cell-specific AR regulation by IL-6.

4 Regulation of Activity of AR Expression and Activity by IL-6 The rationale for studies on AR regulation by IL-6 was the fact that the cytokine’s expression is enhanced and its receptor is present in the majority of prostate cancers. Serum levels of IL-6 higher than 4 pg/ml are associated with bad prognosis, as revealed in two independent studies from Japan and USA (Nakashima et al. 2000; George et al. 2005). Further increase of the cytokine in serum is observed in patients with metastatic disease (Twillie et al. 1995). Upregulation of IL-6 and its receptor in tumor tissue might be early events in carcinogenesis, since the increases were seen in extracts obtained from patients with organ-confined tumors who underwent radical prostatectomy (Giri et al. 2001). IL-6 is also measured in conditioned media from osteoblasts, and it may therefore contribute to the metastatic spread of prostate cancer to bone (Blasczyk et al. 2004). IL-6 activation of the AR was first reported in DU-145 cells in which AR cDNA was transiently overexpressed (Hobisch et al. 1998). Maximal ligand-independent effect of

Role of IL-6 in Regulating the Androgen Receptor

455

IL-6 on AR activity was approximately 50% of the levels induced by the highest concentration of the synthetic androgen R1881. Although partial agonism of nonsteroidal receptor antagonists was previously reported in cancers other than those of prostate in nonsteroidal receptor activation, switch of an antagonist to agonist was not observed in the case of AR blockers that diminished IL-6-induced AR activity (Sartorius et al. 1993). Ligand-independent AR activation by IL-6 was antagonized by protein kinase inhibitors (Hobisch et al. 1998). Importantly, induction of reporter gene activity by lower concentrations of androgens is enhanced by IL-6 in a synergistic manner. The role of residual androgens in prostate cancer progression is of particular interest because of increased sensitivity of the AR. Regulation of AR activity by IL-6 is promoter-specific varying from a stronger induction of naturally occurring PSA or probasin promoters to an absent effect on the synthetic promoter (Fig. 1; Ueda et al. 2002a). For that reason, genes induced specifically by IL-6/AR interaction need to be identified and classified. For example, androgen and IL-6 regulate 15-hydroxyprostaglandin dehydrogenase, an enzyme that controls the cellular level of prostaglandins and lypoxins, in a synergistic manner (Tong and Tai 2004). The results showing inhibition of synergistic AR activation by either a nonsteroidal antiandrogen or a protein kinase inhibitor were also confirmed in that study (Hobisch et al. 1998; Tong and Tai 2004). In LNCaP cells, IL-6 inhibits the expression of the AR coactivator CBP mRNA and protein, thus acting similarly as androgens (Comuzzi et al. 2004). In advanced prostate cancer in vivo, the levels of CBP are, however, high most probably as a consequence of androgen ablation. Implications of AR activation on regulation of cellular events in a cell line that expresses endogenous receptor are important. Inhibition of proliferation of LNCaP cells by IL-6 was associated with an AR-mediated increase in expression of PSA mRNA and protein (Fig. 2; Hobisch et al. 1998). Data from several laboratories on regulation of proliferation by IL-6 in LNCaP cells remain controversial since both positive and negative effects were published in the literature (Giri et al. 2001; Okamoto et al. 1997; Degeorges et al. 1996; Spiotto and Chung 2000a). Sensitivity of LNCaP cells to IL-6 may be passage number-dependent and cell culture conditions, such as serum, may account for the divergent results. However, it was reported that IL-6 diminishes the volume of LNCaP xenografts, in parallel with induction of STAT3 phosphorylation and neuroendocrine differentiation (Wang et al. 2004; Spiotto and Chung 2000b). Other researchers used STAT3 antisense oligonucleotides and demonstrated the inhibition of the tumor cell growth in an ARnegative tumor cell line (Mora et al. 2002). The issue whether those differences are directly caused by expression of the AR may be further investigated. Neuroendocrine cells are growth arrested and terminally differentiated. Their role in tumor progression is most frequently explained by paracrine secretion of mitogenic peptides. According to those data, AR regulation by IL-6 is associated with a prodifferentiation rather than proliferative effect. This is not surprising since it is evident that compounds such as phenylbutyrate and phenylacetate cause an inhibitory effect on proliferation in parallel with AR activation (Sadar and Gleave 2000). Studies in LNCaP cells were extended by Lin et al. (2001) who showed that

456

Fig. 1 Effects of IL-6 on the activities of androgeninducible reporters in LNCaP cells. LNCaP cells were transiently transfected with PSA (6.1 kb) (a), PB ( 286/ +28) (b), or ARR3-tkluciferase (c) (1 mg/well) for 24 h and treated with R1881 (10 nM), a mixture of R1881 (10 nM) and IL-6, or vehicle for an additional 48 h under serum-free conditions. The error bars represent the mean SE of three independent experiments (From Ueda et al. (2002a) with permission)

Z. Culig, A. Hobisch

Role of IL-6 in Regulating the Androgen Receptor

457

Fig. 2 Measurement of PSA mRNA levels in LNCaP cells by semiquantitative RT-PCR. PSA mRNA levels were determined in LNCaP cells after stimulation with IL-6 (50 ng/ml) for 24 h. mRNA levels are expressed as a PSA:b2-microglobulin ration. Results are shown as mean values of five independent experiments; bars, SE; *, P < 0.05; IL-6 treatment versus untreated control, Mann-Whitney U test (From Hobisch et al. (1998) with permission)

IL-6 also upregulates AR gene promoter activity and, consequently, mRNA expression.

5 Mechanism of AR Activation by IL-6 Association of AR with STAT3 observed in LNCaP cells may in part reveal the mechanism of regulation of AR activity. However, absence of phosphorylated STAT3 in DU-145 cells may be compensated by activation of other signaling pathways such as that of p44/p42 MAPK. It was thus proposed that the levels of pSTAT3 or MAPK positively influence AR activation, whereas phosphorylated Akt may act in both ways (Yang et al. 2003). Different effects of Akt on AR transcriptional activity in early and late passages of LNCaP cells were reported (Lin et al. 2004). In early passages, the effect is inhibitory, whereas AR activity is stimulated in higher passages. Matsuda et al. (2001) found that phosphorylation of STAT3 is positively influenced by androgenic hormones; thus, physical interaction between the AR and STAT3 may have bidirectional functional effects. The N-terminal region of the AR is required for activation by IL-6 (Ueda et al. 2002a). IL-6 enhanced Gal4-luciferase activity in a yeast system consisting of the AR N-terminal domain fused to the Gal4 DNA-binding domain. Amino acids 234–558 in the N-terminal domain of the AR interact with STAT3. The N-terminal region is activated in part by the MAPK pathway, the mechanism that may also account for the effect of ErbB2 on the AR.

458

Z. Culig, A. Hobisch

Cytoplasmic kinases Etk and Pim1 are also involved in AR regulation by IL-6 (Kim et al. 2004). Etk is an upstream kinase that induces tyrosine phosphorylation of Pim1 that regulates cell cycle progression and inhibits apoptosis upon IL-6 treatment. A positive feedback loop was proposed for those two kinases. The kinase-deficient mutant forms have a profound negative effect on IL-6 action on the AR. Interference with action of cytoplasmic kinases may be a potential therapeutic approach to abolish ligand-independent activation of the AR. AR activity is mediated by a number of coregulatory proteins, coactivators, and corepressors. Improved possibilities for detection and knock down of coactivators in target tissues led to identification of their specific functions. Two cofactors, SRC-1 and p300, have been analyzed in association with the IL-6–AR interaction. The effect of IL-6 on the AR was thus potentiated by SRC-1 in a dose-dependent manner (Ueda et al. 2002b). Mutations of SRC-1, which interacts with the N-terminal part of the AR, at Thr1179 and Ser1185 diminished AR activation by IL-6. Protein–protein interaction between the AR and the SRC-1 in LNCaP cells is enhanced by IL-6. SRC-1 expression in localized prostate cancer is associated with tumor aggressiveness, and an inhibitory siRNA approach yielded reduced proliferation of prostate cancer cells (Agoulnik et al. 2005). For that reason, potentiation of AR activation by SRC-1 may be clinically relevant. p300 is a transcriptional integrator whose levels are also elevated in prostate cancer and correlate with larger tumor volume, extraprostatic extension, and seminal vesicle invasion (Debes et al. 2003). Sequestration of p300 with E1A causes an inhibition of IL-6-regulated gene transcription (Debes et al. 2002). In experiments in which p300 levels were completely downregulated by siRNA, the effect of IL-6 was abolished. Taken together, the results from mechanistic studies highlighted the importance of the AR N-terminus and SRC-1 and p300 coactivators for activation by IL-6. Similar to IL-6, oncostatin M, which regulates cellular events in prostate stroma in an autocrine and in epithelium in a paracrine manner, is also implicated in the regulation of AR activity (Godoy-Tundidor et al. 2002; Royuela et al. 2004). Cotreatment of prostate cancer cells with oncostatin M and androgen resulted in additive rather than synergistic AR regulation. It is assumed that, in contrast to breast cancer, oncostatin M does not act as a growth inhibitor in prostate tumor cell lines (Mori et al. 1999).

6 Chronic IL-6 Treatment and AR Regulation Alterations in AR responsiveness to IL-6 were observed in models developed to monitor changes that may occur during prostate cancer progression. Such models were generated independently by two groups, and their phenotype seems to be similar (Hobisch et al. 2001; Steiner et al. 2003, 2004; Lee et al. 2007). Basically, LNCaP subjected to chronic treatment with IL-6 showed features of malignant growth, increased expression of cyclin-dependent kinases, downregulation of tumor suppressors, and increased endogenous production of IL-6. Interestingly, in one of

Role of IL-6 in Regulating the Androgen Receptor

459

the sublines, the AR is expressed at low levels, whereas AR activity was found to be higher in another subline. Interestingly, p300 was able to replace the AR in cells derived after prolonged exposure to IL-6 in induction of AR downstream genes PSA and Nkx 3.1 (Debes et al. 2005). Although in most publications upregulation of AR activity in response to IL-6 treatment was observed as discussed above, reports showing an opposite effect are also notable in the literature. Jia et al. (2003, 2004) found that IL-6 antagonized stimulatory effects of either androgen or forskolin on the PSA promoter. The effect of IL-6 was explained by inhibition of p300/CBP but not by p160 coactivator recruitment and histone H3 acetylation. The differences regarding AR activation following IL-6 treatment have not been clarified, but similar factors as those responsible for variable results in proliferation studies have to be considered.

7 IL-6 Therapies in Human Prostate Cancer AR activation by IL-6 should be discussed also in context of development of experimental therapies. For IL-6 inhibition, antisense oligonucleotides, superantagonists (Sant 7), and antibodies are available. In prostate cancer, the effects of the anti-IL-6 antibody CNTO 328 were demonstrated in AR-negative PC-3 and androgen-responsive LuCaP 35 xenografts (Smith and Keller 2001; Wallner et al. 2006). Presently, it is not clear whether the inhibitory effect in LuCaP 35 is mediated through the AR. Favorable effects of the glucocorticoid hormone dexamethasone in terms of pain relief and PSA decline were demonstrated in a subgroup of patients, and it was suggested that the effect is mediated through IL-6 inhibition (Akakura et al. 2003). In future studies, it will be important to identify the patients who will benefit from anti-IL-6 therapy, also in combination with other established treatments.

8 Summary and Future Directions IL-6 effects in prostate cancer are of interest because of its increased expression in tissues and sera. In several cell lines, it accelerates proliferation and promotes cellular survival. The mechanism of IL-6 activation of the AR is clarified in part, by its interactions between the N-terminal region with STAT factors and the coactivator SRC-1. However, the implications of AR activation by IL-6 at the cellular level are still not understood. It is expected that newly developed cellular models for prostate cancer could be used to learn more about the significance of ligandindependent activation of the AR in general and the role of IL-6 in this process in particular.

460

Z. Culig, A. Hobisch

References Agoulnik IU, Vaid A, Bingman WEr, Erdeme H, Frolov A, Smith CL, Ayala G, Ittmann M, Weigel NL (2005) Role of SRC-1 in the promotion of prostate cancer cell growth and tumor progression. Cancer Res 65: 7959–7967 Akakura K, Suzuki H, Ueda T, Komiya A, Ichikawa T, Igarashi T, Ito H (2003) Possible mechanism of dexamethasone therapy for prostate cancer: suppression of circulating level of interleukin-6. Prostate 56: 106–109 Bellezza I, Neuwirt H, Nemes C, Cavarretta IT, Puhr M, Steiner H, Minelli A, Bartsch G, Offner F, Hobisch A, Doppler W, Culig Z (2006) Suppressor of cytokine signaling-3 antagonizes cAMP effects on proliferation and apoptosis and is expressed in human prostate cancer. Am J Pathol 169: 2199–2208 Bellido T, Jilka RL, Boyce BF, Girasole G, Broxmeyer H, Dalrymple SA, Murray R, Manolagas SC (1995) Regulation of interleukin-6, osteoclastogenesis, and bone mass by androgens. J Clin Invest 95: 2886–2895 Blasczyk N, Masri BA, Mawji NR, Ueda T, McAlinden G, Duncan CP, Bruchovsky N, Schweikert HU, Schnabel D, Jones EC, Sadar MD (2004) Osteoblast-derived factors induce androgenindependent proliferation and expression of prostate-specific antigen in human prostate cancer cells. Clin Cancer Res 10: 1860–1869 Campbell CL, Jiang Z, Savarese DM, Savarese TM (2001) Increased expression of the interleukin11 receptor and evidence of STAT3 activation in prostate carcinoma. Am J Pathol 158: 25–32 Chen CD, Welsbie DS, Tran C, Baek SH, Chen R, Vess R, Rosenfeld MG, Sawyers CL (2004) Molecular determinants of resistance to antiandrogen therapy. Nat Med 10: 33–39 Comuzzi B, Nemes C, Schmidt S, Jasarevic Z, Lodde M, Pycha A, Bartsch G, Offner F, Culig Z, Hobisch A (2004) The androgen receptor co-activator CBP is up-regulated following androgen withdrawal and is highly expressed in advanced prostate cancer. J Pathol 204: 159–166 Craft N, Shostak Y, Carey M, Sawyers CL (1999) A mechanism for hormone-independent prostate cancer through modulation of androgen receptor signaling by the HER-2/neu tyrosine kinase. Nat Med 5: 280–285 Culig Z, Hobisch A, Cronauer MV, Radmayr C, Trapman J, Hittmair A, Bartsch G, Klocker H (1994) Androgen receptor activation in prostatic tumor cell lines by insulin-like growth factorI, keratinocyte growth factor, and epidermal growth factor. Cancer Res 54: 5474–5478 Culig Z, Steiner H, Bartsch G, Hobisch A (2005) Interleukin-6 regulation of prostate cancer cell growth. J Cell Biochem 95: 497–505 De Marzo AM, Platz EA, Sutcliffe S, Xu J, Gro¨nberg H, Drake CG, Nakai Y, Isaacs WB, Nelson WG (2007) Inflammation in prostate carcinogenesis. Nat Rev Cancer 7: 256–269 Debes JD, Schmidt LJ, Huang H, Tindall DJ (2002) p300 mediates interleukin-6-dependent transactivation of the androgen receptor. Cancer Res 62: 5632–5636 Debes JD, Sebo TJ, Lohse CM, Murphy LM, Haugen de AL, Tindall DJ (2003) p300 in prostate cancer proliferation and progression. Cancer Res 63: 7638–7640 Debes JD, Comuzzi B, Schmidt LJ, Dehm SM, Culig Z, Tindall DJ (2005) p300 regulates androgen receptor-independent expression of prostate-specific antigen in prostate cancer cells treated chronically with interleukin-6. Cancer Res 65: 5965–5973 Degeorges A, Tatoud R, Fauvel Lafeve F, Podgorniak MP, Millot G, de Cremoux P, Calvo F (1996) Stromal cells from human benign prostate hyperplasia produce a growth-inhibitory factor for LNCaP prostate cancer cells, identified as interleukin-6. Int J Cancer 68: 207–214 Desai SJ, Ma AH, Tepper CG, Chen HW, Kung HJ (2006) Inappropriate activation of the androgen receptor by nonsteroids: involvement of the Src kinase pathway and its therapeutic implications. Cancer Res 66: 10449–10459 Eder IE, Culig Z, Ramoner R, Thurnher M, Putz T, Nessler-Menardi C, Tiefenthaler M, Bartsch G, Klocker H (2000) Inhibition of LNCaP prostate cancer cells by means of androgen receptor antisense oligonucleotides. Cancer Gene Therapy 7: 997–1007

Role of IL-6 in Regulating the Androgen Receptor

461

Fradet V, Lessard L, Begin L, Karakiewicz P, Masson AM, Saad F (2004) Nuclear factor-kappa B nuclear localization is predictive of biochemical recurrence in patients with positive margin prostate cancer. Clin Cancer Res 10: 8460–8464 George DJ, Halabi S, Shepard TF, Sanford B, Vogelzang NJ, Small EJ, Kantoff PW (2005) The prognostic significance of plasma interleukin-6 levels in patients with metastatic hormonerefractory prostate cancer: results from cancer and leukemia group B 9480. Clin Cancer Res 11: 1815–1820 Giri D, Ozen M, Ittmann M (2001) Interleukin-6 is an autocrine growth factor in human prostate cancer. Am J Pathol 159: 2159–2165 Godoy-Tundidor S, Hobisch A, Pfeil K, Bartsch G, Culig Z (2002) Acquisition of agonistic properties of nonsteroidal antiandrogens after treatment with oncostatin M in prostate cancer cells. Clin Cancer Res 8: 2356–2361 Gupta S, Afaq F, Mukhtar H (2002) Involvement of nuclear factor-kappa B, Bax and Bcl-2 in induction of cell cycle arrest and apoptosis by apigenin in human prostate carcinoma cells. Oncogene 21: 3727–3738 He B, You L, Uematsu K, Zang K, Xu Z, Lee AY, Costello JF, McCormick F, Jablons DM (2003) SOCS-3 is frequently silenced by hypermethylation and suppresses cell growth in human lung cancer. Proc Natl Acad Sci USA 100: 14133–14138 Hobisch A, Eder IE, Putz T, Horninger W, Bartsch G, Klocker H, Culig Z (1998) Interleukin-6 regulates prostate-specific protein expression in prostate carcinoma cells by activation of the androgen receptor. Cancer Res 58: 4640–4645 Hobisch A, Rogatsch H, Hittmair A, Fuchs D, Bartsch GJ, Klocker H, Bartsch G, Culig Z (2000) Immunohistochemical localization of interleukin-6 and its receptor in benign, premalignant and malignant prostate tissue. J Pathol 191: 239–244 Hobisch A, Ramoner R, Fuchs D, Godoy-Tundidor S, Bartsch G, Klocker H, Culig Z (2001) Prostate cancer cells (LNCaP) generated after long-term interleukin-6 treatment express interleukin-6 and acquire an interleukin-6-partially resistant phenotype. Clin Cancer Res 7: 2941–2948 Jia L, Kim J, Shen H, Clark PE, Tilley WD, Coetzee GA (2003) Androgen receptor activity at the prostate specific antigen locus: steroidal and non-steroidal mechanisms. Mol Cancer Res 1: 385–392 Jia L, Choong CS, Ricciardelli C, Kim J, Tilley WD, Coetzee GA (2004) Androgen receptor signaling: mechanism of interleukin-6 inhibition. Cancer Res 64: 2619–2626 Kim O, Jiang T, Xie Y, Guo Z, Chen H, Qiu Y (2004) Synergism of cytoplasmic kinases in IL6induced ligand-independent activation of androgen receptor in prostate cancer cells. Oncogene 23: 1838–1844 Lee SO, Lou W, Hou M, Onate SA, Gao AC (2003) Interleukin-4 enhances prostate-specific antigen expression by activation of the androgen receptor and Akt pathway. Oncogene 22: 6037–6044 Lee SQ, Chun JY, Nadiminty N, Lou W, Gao AC (2007) Interleukin-6 undergoes transition from growth inhibitor associated with neuroendocrine differentiation to stimulator accompanied by androgen receptor activation during LNCaP prostate cancer cell progression. Prostate 67: 764–773 Lin DL, Whitney MC, Yao Z, Keller ET (2001) Interleukin-6 induces androgen responsiveness in prostate cancer cells through up-regulation of androgen receptor expression. Clin Cancer Res 7: 1773–1781 Lin HK, Hu YC, Yang L, Altuwaijri S, Chen YT, Kang HY, Chang C (2004) Suppression vs induction of androgen receptor functions by the phosphatidylinositol 3-kinase/Akt pathway in prostate cancer LNCaP cells with different passage numbers. J Biol Chem 278: 50902–50907 Matsuda T, Junicho A, Yamamoto T, Kishi H, Korkmaz K, Saatcioglu F, Fuse H, Muraguchi A (2001) Cross-talk between signal transducer and activator of transcription 3 and androgen receptor signaling in prostate carcinoma cells. Biochem Biophys Res Commun 283: 179–187

462

Z. Culig, A. Hobisch

Mora LB, Buettner R, Seigne J, Diaz J, Ahmad N, Garcia R, Bowman T, Falcone R, Fairclough R, Cantor A, Muro-Cacho C, Livingston S, Karras J, Pow-Sang J, Jove R (2002) Constitutive activation of Stat3 in human prostate tumors and cell lines: direct inhibition of Stat3 signaling induces apoptosis of prostate cancer cells. Cancer Res 62: 6659–6666 Mori S, Murakami-Mori K, Bonavida B (1999) Oncostatin M (OM) promotes the growth of DU 145 human prostate cancer cells, but not PC-3 or LNCaP, through the signaling of the OM specific receptor. Anticancer Res 19: 1011–1015 Nakashima J, Tachibana M, Horiguchi Y, Oya M, Ohigashi T, Asakura H, Murai M (2000) Serum interleukin-6 as a prognostic factor in patients with prostate cancer. Clin Cancer Res 6: 2702–2706 Nazareth LV, Weigel NL (1996) Activation of the human androgen receptor through a protein kinase A signaling pathway. J Biol Chem 271: 19900–19907 Neuwirt H, Puhr M, Cavarretta IT, Mitterberger M, Hobisch A, Culig Z (2007) Suppressor of cytokine signalling-3 is up-regulated by androgen in prostate cancer cell lines and inhibits androgen-mediated proliferation and secretion. Endocr Relat Cancer 14: 1007–1019 Niwa Y, Kanda H, Shikauchi Y, Saiura A, Matsubara K, Kitagawa T, Yamamoto J, Kubo T, Yoshikawa H (2005) Methylation silencing of SOCS-3 promotes cell growth and migration by enhancing JAK/STAT and FAK signalings in human hepatocellular carcinoma. Oncogene 24: 6406–6417 Okamoto M, Lee C, Oyasu R (1997) Interleukin-6 as a paracrine and autocrine growth factor in human prostatic carcinoma cells in vitro. Cancer Res 57: 141–146 Park JI, Lee MG, Cho K, Park BJ, Chae KS, Byun DS, Ryu BK, Park YK, Chi SG (2003) Transforming growth factor-beta1 activates interleukin-6 expression in prostate cancer cell through the synergistic collaboration of the Smad2, p38-NF-kappaB, JNK, and Ras signaling pathways. Oncogene 22: 4314–4332 Royuela M, Ricote M, Parsons MS, Garcia-Tunon I, Paniagua R, de Miguel MP (2004) Immunohistochemical analysis of the IL-6 family of cytokines and their receptors in benign, hyperplastic, and malignant human prostate. J Pathol 202: 41–49 Sadar MD, Gleave ME (2000) Ligand-independent activation of the androgen receptor by the differentiation agent butyrate in human prostate cancer cells. Cancer Res 60: 5825–5831 Santhanam U, Ray A, Sehgal PB (1991) Repression of the interleukin 6 gene promoter by p53 and the retinoblastoma susceptibility gene product. Proc Natl Acad Sci USA 88: 7605–7609 Sartorius CA, Tung T, Takimoto GS, Horwitz KB (1993) Antagonist-occupied human progesterone receptors bound to DNA are functionally switched to transcriptional agonists by cAMP. J Biol Chem 268: 9262–9266 Shariat SF, Andrews B, Kattan MW, Kim J, Wheeler T, Slawin KM (2001) Plasma levels of interleukin-6 and its soluble receptor are associated with prostate cancer progression and metastasis. Urology 58: 1008–1015 Smith PC, Keller ET (2001) Anti-interleukin-6 monoclonal antibody induces regression of human prostate cancer xenografts in nude mice. Prostate 48: 47–53 Spiotto MT, Chung TD (2000a) STAT3 mediates IL-6-induced growth inhibition in the human prostate cancer cell line LNCaP. Prostate 42: 88–98 Spiotto MT, Chung TD (2000b) STAT3 mediates IL-6-induced neuroendocrine differentiation in prostate cancer cells. Prostate 42: 186–195 Steiner H, Godoy-Tundidor S, Rogatsch H, Berger AP, Fuchs D, Comuzzi B, Bartsch G, Hobisch A, Culig Z (2003) Accelerated in vivo growth of prostate tumors that up-regulate interleukin-6 is associated with reduced retinoblastoma protein expression and activation of the mitogenactivated protein kinase pathway. Am J Pathol 162: 655–663 Steiner H, Berger AP, Godoy-Tundidor S, Bjartell A, Lilja H, Bartsch G, Hobisch A, Culig Z (2004) Vascular endothelial growth factor autocrine loop is established in prostate cancer cells generated after prolonged treatment with interleukin-6. Eur J Cancer 40: 1066–1072

Role of IL-6 in Regulating the Androgen Receptor

463

Titus MA, Schell MJ, Lih FB, Tomer KB, Mohler JL (2005) Testosterone and dihydrotestosterone tissue levels in recurrent prostate cancer. Clin Cancer Res 11: 4653–4657 Tong M, Tai HH (2004) Synergistic induction of the nicotinamide adenine dinucleotide-linked 15hydroxyprostaglandin dehydrogenase by an androgen and interleukin-6 or forskolin in human prostate cancer cells. Endocrinology 145: 2141–2147 Twillie DA, Eisenberger MA, Carducci MA, Hseih W-S, Kim WY, Simons JW (1995) Interleukin-6: a candidate mediator of human prostate cancer morbidity. Urology 45: 542–549 Ueda T, Bruchovsky N, Sadar MD (2002a) Activation of the androgen receptor N-terminal domain by interleukin-6 via MAPK and STAT3 signal transduction pathways. J Biol Chem 277: 7076–7085 Ueda T, Mawji NR, Bruchovsky N, Sadar MD (2002b) Ligand-independent activation of the androgen receptor by interleukin-6 and the role of steroid receptor coactivator-1 in prostate cancer cells. J Biol Chem 277: 38087–38094 Wallner L, Dai J, Escara-Wilke J, Zhang J, Yao Z, Lu Y, Trikha M, Nemeth JA, Zaki MH, Keller ET (2006) Inhibition of interleukin-6 with CNTO328, an anti-interleukin-6 monoclonal antibody, inhibits conversion of androgen-dependent prostate cancer to an androgen-independent phenotype in orchiectomized mice. Cancer Res 66: 3087–3095 Wang G, Jones SJ, Marra MA, Sadar MD (2006) Identification of genes targeted by the androgen and PKA signaling pathways in prostate cancer cells. Oncogene 25: 7311–7323 Wang Q, Horiatis D, Pinski J (2004) Interleukin-6 inhibits the growth of prostate cancer xenografts in mice by the process of neuroendocrine differentiation. Int J Cancer 111: 508–513 Weber A, Hengge UR, Bardenhauer W, Tischoff I, Sommerer F, Markwarth A, Dietz A, Wittekind C, Tannapfel A (2005) SOCS-3 is frequently methylated in head and squamous cell carcinoma and its precursor lesions and causes growth inhibition. Oncogene 24: 6699–6708 Yang L, Wang L, Lin HK, Kan PY, Xie S, Tsai M, Wang PH, Chen YT, Chang C (2003) Interleukin-6 differentially regulates androgen receptor transactivation via PI3K-Akt, STAT3, and MAPK, three distinct signal pathways in prostate cancer cells. Biochem Biophys Res Commun 305: 462–469 Yeh S, Lin HK, Kang HY, Thin TH, Lin MF, Chang C (1999) From HER2/Neu signal cascade to androgen receptor and its coactivators: a novel pathway by induction of androgen target genes through MAP kinase in prostate cancer cells. Proc Natl Acad Sci USA 96: 5458–5463 Zegarra-Moro OL, Schmidt LJ, Huang H, Tindall DJ (2002) Disruption of androgen receptor function inhibits proliferation of androgen-refractory prostate cancer cells. Cancer Res 62: 1008–1013 Zerbini LF, Wang Y, Cho JY, Libermann TA (2003) Constitutive activation of nuclear factor kappaB p50/p65 and Fra-1 and JunD is essential for deregulated interleukin 6 expression in prostate cancer. Cancer Res 63: 2206–2215

The Role of Cyclic AMP in Regulating the Androgen Receptor Marianne D. Sadar

Abstract The androgen receptor (AR) is suggested to play a predominant role in the recurrence of prostate cancer in patients receiving androgen ablation therapy. In the absence of androgens, the AR is activated by compounds that increase cyclic adenosine 3’,5’-monophosphate (cAMP) and stimulate cAMP-dependent protein kinase (PKA) activity in prostate cancer cells. Thus cross-talk between AR and cAMP/PKA pathways is suspected to be involved in the mechanism underlying castration-recurrent prostate cancer. Elucidation of the molecular mechanism(s) of how the AR can be activated by alternative pathways such as cAMP/PKA in the absence of androgen may yield new therapeutic targets for the improved clinical management of advanced prostate cancer.

1 Introduction The only effective treatment available for advanced prostate cancer is the withdrawal of androgens which are essential for the survival of prostate epithelial cells. Androgen deprivation therapy causes a temporary reduction in tumor burden concomitant with a decrease in serum prostate-specific antigen (PSA). Unfortunately prostate cancer will eventually begin to grow again in the absence of androgens to form castration-resistant disease (also called castration-recurrent, androgenindependent, or hormone-refractory disease) (Huber et al. 1987). Castrationresistant prostate cancer is biochemically characterized before the onset of symptoms by rising levels of serum PSA (Miller et al. 1992). Once the disease becomes castration-resistant, most patients succumb to their disease within 2 years. The androgen receptor (AR) has been suspected to play an important role in hormone-refractory disease because many of the same genes that are increased by androgens in androgen-dependent prostate cancer xenografts become elevated in prostate cancer xenografts in castrated hosts (Gregory et al. 1998). This suggests M.D. Sadar Genome Sciences Center, BC Cancer Agency, Vancouver, BC, Canada V5Z 1L3, E-mail: [email protected]

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4_21, # Springer Science + Business Media, LLC 2009

465

466

M.D. Sadar

that the AR can be activated in the absence of testicular androgens by alternative signal transduction pathways in androgen-independent disease. Such findings are consistent with detecting nuclear AR protein in recurrent prostate cancer (Kim et al. 2002; van der Kwast et al. 1991). Other data supporting the role of AR in androgenindependent disease include amplification of the AR gene which has been detected in 20–30% of androgen-independent tumors (Moul et al. 1995; Visakorpi et al. 1995), delayed onset of androgen independence by altering the timing and sequence of use of antiandrogens (Bruchovsky et al. 2001), and the necessity of AR for the proliferation and tumor growth of androgen-independent prostate cancer cells (Chen et al. 2004; Zegarra-Moro et al. 2002). Therefore, activation of the AR is implicated in the underlying molecular mechanism of hormone-refractory disease. The AR can be activated in the absence of androgen by stimulation of the cyclic adenosine 30 ,50 -monophosphate (cAMP)-dependent protein kinase (PKA) pathway, interleukin-6, and various growth factors (Culig et al. 1994; Nazareth and Weigel 1996; Sadar 1999; Ueda et al. 2002a, 2002b). The mechanism of ligandindependent transactivation of the AR has been shown to involve: (1) increased nuclear AR protein suggesting nuclear translocation, (2) increased AR complex formation with androgen response elements (AREs), and (3) the AR N-terminal domain (NTD) (Sadar 1999; Ueda et al. 2002a, 2002b). This chapter focuses on the components of the cAMP/PKA pathway and compounds that stimulate this pathway in prostate cancer cells, cross talk between the androgen and cAMP/PKA signaling pathways, and potential mechanisms that may be involved.

2 Cyclic AMP Pathway The cAMP pathway was discovered 50 years ago and is considered the prototypical second messenger system required for diverse processes in many different cells and organisms (Sutherland and Rall 1958). These processes include metabolism, secretion, ion channel conductance, inflammatory response, differentiation, apoptosis, and cell growth. Stimulation of plasma membrane receptors by hormones, such as epinephrine, activates adenylyl cyclase (AC) to increase the synthesis of the second messenger cAMP. For many years, it was thought that PKA was the principle effector of elevated levels of cAMP. This canonical cAMP signaling consists of AC, PKA catalytic (C) and regulatory (R) subunits, and phosphodiesterases (PDEs). In mammals, there are three types of effector proteins for the cAMP pathway: (1) PKA, (2) exchange proteins activated by cAMP (EPACs), and (3) cyclic nucleotide gated ion channels.

2.1

Adenylyl Cyclases

ACs are enzymes that generate cAMP from ATP and together with guanylyl cyclases, which generate cGMP from GTP, belong to the nucleotidyl cyclase

The Role of Cyclic AMP in Regulating the Androgen Receptor

467

family. There are six classes of nucleotidyl cyclases but only class III is expressed in eukaryotes (Defer et al. 2000). Two types of class III ACs are expressed by mammals and are the transmembrane adenylyl cyclase (tmAC) family and the soluble adenylyl cyclase (sAC).

2.1.1

Transmembrane Adenylyl Cyclase

The tmAC family is encoded by nine genes (type I to IX) (see Kamenetsky et al. 2006; Sunahara and Taussig 2002 for reviews). The different isoforms are expressed in different tissues and some are alternatively spliced, which increases complexity. Heterotrimeric G-proteins are the major regulators of the tmACs to generate cAMP in response to hormones and neurotransmitters that signal through G–protein-coupled receptors (GPCRs) (Taussig and Gilman 1995). Most tmAC isoforms are stimulated by the a-subunit of Gs to increase synthesis of cAMP, while Gia inhibits isoforms I, V, and VI tmACs and subsequent production of cAMP (see Kamenetsky et al. 2006 for a review). Changes in the expression of specific G-protein isoforms are associated with aging. For example, age-related increases in the level of Gi in the bladder may result in decreased cAMP. Calmodulin regulates isoforms I and VIII tmACs (Cali et al. 1996; Smigel 1986; Tang et al. 1991; Yeager et al. 1985) and possibly isoform III (Choi et al. 1992). All AC activities are blocked by high concentrations of calcium. tmAC activity can be regulated by posttranslational modifications such as phosphorylation by PKA (to inhibit isoforms V and VI) and PKC (isoforms II, V, and VI), S-nitrosylation (isoforms V and VI), and N-linked glycolysation (isoform VIII) (Kamenetsky et al. 2006). Regulator of G-protein signaling 2 (RGS2) protein, an inhibitor of GPCR function, directly regulates isoforms II, V, and VI. RGS2 is decreased significantly in prostate cancer and inhibits androgen-independent activation of the AR (Cao et al. 2006). Ga(s) activation of PKA activity is required for activation of the AR in prostate cancer cells (Kasbohm et al. 2005). Inhibition of activated GPCRs has shown therapeutic benefits for patients with prostate cancer (Nelson 2003). The diterpene, forskolin, activates all mammalian tmACs except tmAC IX (Premont et al. 1996; Yan et al. 1998). tmAcs are involved in a variety of processes including metabolism, proliferation, differentiation, and apoptosis. The concentration of intracellular cAMP has an effect on cell cycle in many cells. Both stimulation and inhibition of proliferation have been noted. This is true also for prostate cancer cells as described below. Levels of cAMP have also been reported to induce differentiation of prostate cancer cells to a neuroendocrine (NE) phenotype but the AC isoforms implicated have not been described. Neuronal differentiation of P19 cells in response to retinoic acid through the cAMP/PKA cascade was associated with increased expression of AC isoforms II, V, and VII (Lipskaia et al. 2000). Mesodermal differentiation of these same cells was associated also with increased expression of AC isoforms II, V, and VI (Lipskaia et al. 2000). Isoform II is the suspected AC required based upon its function to arrest proliferation, which was essential for differentiation (Lipskaia et al. 2000).

468

M.D. Sadar

Characterization of tmACs in prostate tissue has not been reported in spite of the role of specific isoforms in the proliferative and NE differentiation responses to elevation of cAMP.

2.1.2

Soluble Adenylyl Cyclase

sAC is a second class of ACs that was recently discovered (Buck et al. 1999). sAC is a single gene with many isoforms produced by alternative splicing and is expressed in most tissues in discrete subcellular locations within the cell. Testis has the highest expression of sAC and there have been no studies showing expression in the prostate or in prostate cancer cells. sAC plays a role in regulation of pH in the epididymis, activation of tumor necrosis factor in granulocytes, and activation of sperm (Esposito et al. 2004; Han et al. 2005; Hess et al. 2005; PastorSoler et al. 2003). sAC is regulated by bicarbonate and calcium to produce cAMP (Chen et al. 2000; Jaiswal and Conti 2003). Regulation of sAC has been shown not to involve regulatory proteins or posttranslational modifications and it is insensitive to forskolin.

2.2

PKA

PKA (EC 2.7.11.11) refers to a family of enzymes that depend on the level of cAMP in the cell for activity. When cellular levels of cAMP are below threshold, the PKA holoenzyme remains intact as a tetramer of two R and two C subunits and has no catalytic activity because these regulatory subunits block the catalytic center of the C subunits (Corbin et al. 1978). Activation of ACs by certain GPCRs or by inhibition of PDEs, which degrade cAMP, can increase cellular levels of cAMP. When these levels are above threshold, cAMP binds to the R subunits causing a change in conformation that results in dissociation of the C subunits. These free C subunits can catalyze the transfer of ATP-terminal phosphates to serine or threonine residues on substrates with compatible amino acid motifs (Songyang et al. 1994). Phosphorylation represents a posttranslational modification that can regulate the activity of the protein substrate or its interaction with other proteins. PKA activity can increase in response to a variety of external signals to phosphorylate enzymes, transcription factors, and other proteins important in a diverse number of functions in many different cells. 2.2.1

A-Kinase Anchoring Proteins

To produce cell-specific responses to elevated cAMP in various tissues and different pathways requires numerous mechanisms to regulate PKAs. The A-kinase anchoring proteins (AKAPs) provide one mechanism. AKAPs are important organizing components of PKA. These proteins regulate the spatiotemporal control of cAMP signaling and other transduction events (see Smith et al. 2006 for a review).

The Role of Cyclic AMP in Regulating the Androgen Receptor

469

The first AKAP identified was microtubule-associated protein MAP2 (Theurkauf and Vallee 1982). Many more AKAPs have since been identified and include ezrin, Rab32, and Wave-1. AKAPs that anchor PKA may also bind to other proteins involved in dephosphorylation, such as protein phosphatase 2B (AKAP79), and other kinases, such as PKC. These interactions may integrate the cAMP, Ca2+, and phospholipid signaling pathways to one subcellular site (Klauck et al. 1996). Thus, AKAPs synchronize compartmentalization of signal transduction events. The spatiotemporal control of cAMP flux requires the concerted action of ACs and compartmentalized pools of PDEs, which metabolize cAMP into 50-AMP. AKAPs cluster PKA with PDEs to terminate cAMP signals when they diffuse into cells.

2.2.2

PKA Isoforms

Isoforms and splice variants of the R and C subunits of PKA provide another means of regulation of PKA activity. In mammals, two types of PKAs, type I and type II, are distinguished by their regulatory subunits R-I and R-II and are encoded by four genes to produce four isoforms (RIa, RIb, RIIa, and RIIb) (Skalhegg and Tasken 2000). These R subunits share common C subunits, which are encoded by three genes (Ca, Cb, and PrKX) (Skalhegg and Tasken 2000). Splice variants of Ca are Ca1 are expressed ubiquitously, while CaS has tissue specificity with restricted expression in sperm (Reinton et al. 2000; San Agustin et al. 1998). Cb has at least ten splice variants with Cb1 expressed ubiquitously, Cb2 is associated with lymphoid tissue, Cb3 and 4 are found in neuronal cells (Kvissel et al. 2004; Orstavik et al. 2001, 2005). There are differences where these subunits are located intracellularly. PKA-RI is essentially cytosolic while PKA-RII is associated with particulate. The interaction with AKAPs immobilizes PKA-RII. This large family of AKAPs are expressed in a cell-specific manner and are associated with a diverse range of subcellular structures, thereby regulating the intracellular localization of PKA-RII and its function. The activity of PKA can be regulated by a feedback mechanism involving a PDE that is a substrate of PKA. When phosphorylated, PDE converts cAMP to AMP thereby reducing available cAMP to bind to the R subunits required for PKA activity. PKA activity is also regulated by phosphorylation of the C subunit. A selective activator of PKA is N6-benzoyl-cAMP (Christensen et al. 2003). 2.2.3

PKA Isoforms in Prostate Cancer

LNCaP prostate cancer cells express RIa, RIIa, RIIb, and Cb1,2,3,4-subunits and do not express RIb subunit (Table 1; Cho et al. 2000b; Kvissel et al. 2007; Nesterova et al. 2000). This expression pattern was observed in clinical samples of prostate tissue. Cb2 levels increased (n = 8, p < 0.005) and RIIb appeared to decrease in prostate cancer compared to matched benign prostate (Kvissel et al. 2007). The RI isoform was associated with transformation, poor prognosis, resistance of breast

H Inconclusive N=3

RIIa

H Inconclusive N=3

H H H No change (not shown) No change

Source: Kvissel et al. (2007), Cho et al. (2000b), and Nesterova et al. (2000) H means detected; numbers 1–4 under Cb represent isoforms detected

Clinical samples Prostate cancer compared to benign (clinical samples)

Table 1 PKA isoforms in prostate cancer cells Samples RIa PC-3 cells H DU145 cells H LNCaP cells H LNCaP plus androgen (65 h) No change LNCaP NE by androgen depletion No change Not detected Not detected

RIb Not detected Not detected Not detected Not detected Not detected H Decreased N=3

H H No change (not shown) Increased

RIIb Not detected

– –

H – –

H –

Ca

Cb 1 1 1,2,3,4 Increased (2) Increased (1,3,4) Decreased (2) 1,2,3,4 Increased (2) N=3

470 M.D. Sadar

The Role of Cyclic AMP in Regulating the Androgen Receptor

471

cancer to antiestrogens, and was elevated in a number of cancers (Gordge et al. 1996). This isoform was not increased in prostate cancer compared to benign (Kvissel et al. 2007). PC-3 and DU145 prostate cancer cells were similar in expression of the R-subunits, except PC-3 cells did not express RIIb (Kvissel et al. 2007). Both PC-3 and DU145 cells did not express the Cb2 subunit (Kvissel et al. 2007). Levels of Cb2 transcript increased in LNCaP cells using concentrations of androgen associated with decreased cell growth (1–10 nM R1881), while no changes were detected in cells treated with concentrations of androgen associated with proliferation (0.1 nM or less R1881) (Kvissel et al. 2007). No changes in the levels of transcript for R subunits were observed at any concentration of androgen. Observations from other systems determined that the ratios of PKA-RI to PKA-RII change significantly during differentiation and transformation. Generally, an increase of the RIa subunit is associated with proliferation of cells while the RII subunit is associated with differentiation of cells. NE differentiation has been suggested to play a role in the progression of prostate cancer. Androgen deprivation and compounds that stimulate cAMP and PKA cause NE differentiation of prostate cancer cells maintained in culture (Bang et al. 1994; Burchardt et al. 1999; Cox et al. 1999, 2000; Goodin and Rutherford 2002). Studies examining changes in PKA isoforms revealed that RIIb levels increased with NE differentiation of LNCaP cells (Kvissel et al. 2007). Examination of the C subunits revealed increased levels of Cb2 in response to androgen and decreased levels of Cb2 in growth-arrested LNCaP cells. Levels of Cb1, Cb3, and Cb4 variants increased in NE-like LNCaP cells (Kvissel et al. 2007). No changes in R subunits were measured in response to androgen (Kvissel et al. 2007). However, while castration reduced the total amount of PKA type I in the rat ventral prostate that was restored by daily administration of dihydrotestosterone (DHT), the amount of PKA type II was not altered (Fuller et al. 1978). Unfortunately, these studies described little about the Ca subunit, which has been associated with an extracellular PKA (ECPKA) in prostate cancer.

2.2.4

Extracellular PKA

ECPKA has been detected and reported to be elevated in the plasma from prostate cancer patients (Cvijic et al. 2000) and in the conditioned medium from PC-3, PC-3M, LNCaP, and DU145 prostate cancer cell lines (Cho et al. 2000b; Cvijic et al. 2000). Fifty percent of prostate cancer patients (n = 14, p = 0.04) have elevated levels of ECPKA in their plasma (Cvijic et al. 2000). Higher levels of ECPKA are associated with hormone-independence in breast cancer cells (Cho et al. 2000b). Similar studies examining androgen-independent prostate cancer have not been reported. ECPKA is present as an active free Ca subunit that is secreted and not influenced by the addition of cAMP (Cho et al. 2000a). However, cAMP and ATP were present in the extracellular space with elevated levels in the urine of prostate cancer patients with bone metastases and refractory disease (Murray et al. 2001; Buchs et al. 1998). Forced expression of Ca and RIa led to increased intracellular PKA and ECPKA levels but caused no change in cell morphology or proliferation. Overexpression of

472

M.D. Sadar

RIIb produced no change in intracellular levels of PKA activity but decreased levels of ECPKA, which were associated with reduced levels of PKA-I, reduced cell growth, and altered morphology. Levels of expression of the Ca subunit are suspected to play a role in cell adherence and invasion (Redegeld et al. 1997). Treatment of PC-3M cells with 8-Cl-cAMP (site-selective cAMP analogue) downregulated intracellular PKA-I and ECPKA and reduced proliferation (Cho et al. 2000b). The autoantibody for ECPKA was proposed as a universal biomarker for cancer (Nesterova et al. 2006). 2.2.5

Therapies Targeting PKA

Alteration of expression of PKA subunits was shown to effect cellular growth in prostate cancer cells. Increased levels of RIa, the regulatory subunit of PKAI, were associated with cancer growth and antisense targeting this subunit (GEM231) reduced proliferation and induced differentiation (Cho and Cho-Chung 2003). Antisense suppression of RIa, which increases RIIb, resulted in decreased expression of genes involved in proliferation and transformation and increased the expression of genes involved in differentiation of PC-3M tumors that underwent regression. Consistent with this observation, increased expression of RIIb increased expression of differentiation genes and decreased expression of genes involved in cell proliferation and transformation in contradiction to the effects of RIa and Ca (Cho-Chung and Nesterova 2005). GEM231, combined with hydroxycamptothecin and irinotecan, showed enhanced activity against PC-3 and DU145 tumors (Agrawal et al. 2002; Cho and Cho-Chung 2003; Wang et al. 2002). These prostate cancer cell lines lack functional AR and similar studies with GEM231 using prostate cancer cells that express AR remain untested.

2.2.6

Ribosomal S6 Kinase and Mitogen-Activated Protein Kinase/ERK

PKA activity is restricted by the R subunits, which were shown to interact with AKAPs. Other interactions between R and C subunits of PKA are being discovered. The small GTPase regulatory protein Rab13 interacts and inhibits the C subunit of PKA involved in tight junction assembly (Kohler et al. 2004). The R subunit interacted with the following to inhibit or modify their function: phosphorylase phosphatases (Khatra et al. 1985), cytochrome c oxidase (Yang et al. 1998), PAP7 involved in steroid biosynthesis (Li et al. 2001), replication factor c complex (Gupte et al. 2005), the activated epidermal growth factor (EGF)-receptor via Grb2 (Tortora et al. 1997), and p90-kDa ribosomal S6 kinase (RSK) (Chaturvedi et al. 2006). RSK binds to the PKA RI subunit to reduce interaction between this subunit and the C subunits, which could potentially increase the sensitivity of PKA to cAMP or allow cAMP-independent activation of PKA. Phosphorylation of RSK by extracellular signal-related kinase ERK switches RSK from binding inactive PKA R subunits to binding active PKA C subunits, thereby enhancing RI interaction with

The Role of Cyclic AMP in Regulating the Androgen Receptor

473

the C subunits to decrease the sensitivity of PKA. Activation of RSK by ERK facilitates translocation to the nucleus where it can phosphorylate substrates such as c-fos and histone 3. The accumulation of RSK in the cytosol leads to increased phosphorylation of substrates such as BAD which is involved in apoptosis. Levels of RSK1 and RSK2 are elevated in prostate cancer with comparable levels in LNCaP cells. RSK2 increased proliferation and PSA gene expression by a mechanism dependent upon the AR and interaction with p300 (Clark et al. 2005). Increased mitogen-activated protein kinase (MAPK)/ERK activity is associated with the progression of prostate cancer. The PKA and MAPK pathways intersect at various points. Elevation of cAMP can provide either an inhibitory signal (c-Raf) or stimulatory signal (B-Raf) to ERK activation. ERK can also regulate the activity of PDEs. The long PDE4 isoform is inhibited by ERK phosphorylation while the short isoform is activated. Thus, depending on the isoform present, the levels of cAMP and subsequent activity of PKA would be differentially regulated.

2.3

Phosphodiesterases

PDEs are important for the regulation of levels of cAMP within the cell. These enzymes degrade cyclic nucleotides to 50 -monophosphates and are regulated by localization and expression of many different isoforms. There are at least 11 families of PDEs with 21 related genes and splicing produces more than 50 PDE proteins that can be expressed in mammalian cells (Francis et al. 2001). One PDE gene may have multiple splice variants that have distinct expression in tissue and/or subcellular localizations (Beavo 1995; Bolger et al. 1996; Yan et al. 1996). The nomenclature consists of the family indicated by an Arabic number (e.g., PDE4), a capital letter indicating the gene within the family (e.g., PDE4C), and a last Arabic number representing the splice variant (e.g., PDE4C3). Classification is based upon similarity of sequence, substrate specificity, and sensitivity to inhibitors (Francis et al. 2001). Generally, PDEs have specificity for three types of substrates. PDEs 4, 7, and 8 specifically hydrolyze cAMP with differences in sensitivity to rolipram. PDEs 5, 6, and 9 specifically hydrolyze cGMP. PDE 1, 2, 3, 10, and 11 hydrolyze both substrates (dual). PDEs are regulated by phosphorylation by the calciumcalmodulin-dependent kinase and PKA or GMPs. Cellular localization of PDEs may be particulate or soluble, which also controls regulation of activity. PDEs can hydrolyze tenfold more cAMP and/or cGMP than is synthesized. Thus, inhibition of PDEs has profound effects on elevating cellular levels of cAMP and/or cGMP to reach threshold levels required for kinase activation. A number of inhibitors have been described for the PDE family and are used clinically. 2.3.1

PDE Isoforms in Prostate Cancer

All PDE families except PDE6 are expressed in the prostate (Table 2) (see Wheeler et al. 2005 for a review). The prostate contains high levels of the cGMP-specific

474

M.D. Sadar

Table 2 Phosphodiesterase isoforms (PDEs) present in the prostate PDE Prostate Cyclic nucleotide Inhibitor substrates 1 A,B Dual substrate Vinopocetine, 8-methoxy-IBMX 2 A cGMP-stimulated Erythro-9-(2-hydroxy-3-nonyl)adenine 3 A cGMP-inhibited Cilostamide, milrinone, zardaverine, olprinone, cilostazol, quazinone, siguazodan 4 A,B,C,D cAMP-specific Rolipram, Ro 20-1724, piclamilast, zardaverine, cilomilast, roflumilast 5 A1,2,3 cGMP-specific Zaprinast, sildenafil, vardenafil, tadalafil, E4021 6 – Photoreceptor, Papaverine, zaprinast, dipyridamole, sildenafil cGMP-specific 7 A cAMP high-affinity Papaverine, MIX, dipyridamole 8 A cAMP high-affinity Papaverine, dipyridamole, IBMX-insensitive 9 A cGMP high-affinity Papaverine, zaprinast, dipyridamole 10 A Dual substrate Papaverine, dipyridamole 11 A Dual substrate Dipyridamole, zaprinast, tadalafil Source: Wheeler et al. (2005), Juang (2004), and Uckert et al. (2006) PDE 9 and 11 were abundantly expressed as indicated by bold font

PDE9A, especially PDE9A12 (Rentero et al. 2003), and dual specific PDE11As (Fawcett et al. 2000), with almost exclusive expression of PDE11A4 (Yuasa et al. 2000). PDE11A antibody staining was localized in the glandular epithelium of the prostate (Loughney et al. 2005) and in prostate cancer samples (D’Andrea et al. 2005).

2.3.2

Therapies Targeting PDE

High levels of cAMP can cause cell cycle arrest to inhibit growth of various types of cancer cells in culture. Yet application of cAMP analogues or drugs to increase cAMP levels for cancer therapies has been plagued by cytotoxicity. Tissue- and inhibitor specificity of PDE isoforms provides an approach to intervene using PDE inhibitors. These inhibitors generally result in moderate increases in accumulation of cAMP that are still effective in blocking malignant cell proliferation. Nonspecific PDE inhibitors have been investigated in several human prostate cancer cell lines with impressive attenuation of growth and induction of terminal differentiation (Bang et al. 1994; Goto et al. 1999). The nonselective PDE inhibitor, 3-isobutyl-1-methylxanthine (IBMX), has little inhibitory effect on PDE11A4 that is highly expressed in the prostate while dipyridamole had the most effect. Thus, the efficacy of using IBMX in prostate cancer cells to elevate levels of cAMP must be questioned. Cell-specific differences also have been shown with the PDE inhibitor, papaverine. Papaverine increased cAMP, decreased proliferation, decreased invasion potential, and caused changes in morphology in LNCaP cells. However, PC-3 and DU145 cells exhibited no change in morphology upon treatment with the

The Role of Cyclic AMP in Regulating the Androgen Receptor

475

PDE inhibitors, papaverine, IBMX, or theophylline (Goto et al. 1999). Exisulind, a metabolite of sulindac, inhibits cGMP-specific PDEs and b-catenin. Metabolites of sulindac reduced growth and promoted apoptosis of prostate cancer cell lines in vitro and in vivo (Goluboff et al. 1999; Lim et al. 1999). LNCaP cells, which are AR positive, were more sensitive to sulindac sulfide and exisulind than PC-3 cells which are AR negative. Inhibition of growth and increased apoptosis in LNCaP cells was associated with decreased expression and function of the AR and possibly involved other transcription factors, including cAMP response element-binding protein (CREB) (Lim et al. 2003). Clinical trials with 96 men with increasing PSA after radical prostatectomy showed a 2.12 month increase in PSA doubling time with exisulind (Goluboff et al. 2001). However, combined therapy with exisulind and docetaxel in a phase II trial performed in prostate cancer patients with hormone-refractory disease was shown no better than the single agent docetaxel for either toxicity or efficacy (Ryan et al. 2005; Sinibaldi et al. 2006).

2.4

Exchange Proteins Activated by cAMP

All effects of cAMP were attributed to activation of PKA until recently. Two other targets of cAMP have been identified: cAMP-regulated ion channels and EPAC. Cyclic AMP guanine-nucleotide exchange factor (GEF)-I (EPAC1) and cAMP GEFII (EPAC2) are GEFs for small GTPAses, Rap1 and Rap2. Elevated levels of cAMP levels stimulated EPAC (de Rooij et al. 1998) and modulated signaling to the MAPK pathway independent of PKA (Vossler et al. 1997). EPACs were identified in a quest to reveal a mechanism of activation of GTPase Rap1 by cAMP that was insensitive to inhibitors of PKA (Klauck et al. 1996). Rap1 plays a role in the regulation of integrin-mediated cell adhesion and cadherin-mediated cell-junction formation (Bos 2003, 2005). EPAC2 interacted with GTP-bound Ras to be recruited to the plasma membrane (Bos 2003, 2005). 8-pCPT-20 OMe-cAMP (AKA 007) is a selective and efficient agonist for EPAC proteins (Enserink et al. 2002) and induced activation of Rap1 through EPAC2 (Kang et al. 2006a). Thus, both the canonical (PKA) and noncanonical (EPAC) signaling pathways should be delineated when considering the downstream effects of compounds that elevate cAMP.

2.5

CREB and TORC

Elevation of cAMP can activate CREB through phosphorylation at serine 133 by PKA and/or by nuclear translocation of transducer of regulated CREB activity (TORC) coactivators (see Siu and Jin 2007 for a review). CREB is a transcription factor belonging to the basic leucine zipper family (Shaywitz and Greenberg 1999) that can regulate the expression of more than 5000 target genes including c-fos (Ahn et al. 1998), cyclins D2, and A1 (Desdouets et al. 1995; White et al. 2006) and

476

M.D. Sadar

AR (Mizokami et al. 1994; Stubbs et al. 1996). CREB binds to conserved TGACGTCA sequences called cAMP response elements (CREs) in the promoter regions of genes. Phosphorylation of CREB on serine 133 promoted recruitment of p300/CREB-binding protein (CBP) to increase transcription through acetylation of histones and bridging properties with RNA polymerase II via RNA helicase A. More than 300 different stimuli have been reported to increase the phosphorylation of CREB (see Johannessen et al. 2004 for a review) including androgens (Kim et al. 2001; Unni et al. 2004). CREB has been suspected to be a target of nongenotropic signaling by androgen in LNCaP cells to prevent apoptosis (Unni et al. 2004). Cooperation between the AR and CREB binding to a CRE in the enhancer region of the PSA gene by interaction with CBP is suggested as a mechanism by which PSA gene expression may be influenced by the cAMP/PKA pathway (Kim et al. 2005). Levels of expression of coactivators of the AR are also regulated by CREB. One example is AR coactivator (ART27) that increased AR transactivation, yet may play a role in differentiation rather than proliferation (Taneja et al. 2004). In LNCaP cells, CREB was activated in response to EGF and was recruited to the ART-27 promoter to induce ART27 gene expression (Nwachukwu et al. 2007). Consistent with a role for CREB in prostate cancer, studies have measured increased expression, phosphorylation, and DNA-binding activity of CREB in prostate cancer compared with benign prostate ( p = 0.01) (Ghosh et al. 2007; Kumar et al. 2007). Levels of phosphorylated CREB were also increased in bone metastasis (Wu et al. 2007). CREB is phosphorylated by multiple kinases in addition to PKA including MAPK, which has increased activity in recurrent prostate cancer (Gioeli et al. 1999). Transactivation of CREB can also occur via a phosphorylation-independent event involving TORCS that enhanced the interaction of CREB with the TAF(II) 130 component of TFIID on the promoter (Conkright et al. 2003).

3 Compounds that Increase cAMP Levels of cAMP and the activity of AC or PDEs can be altered by numerous molecules and signaling pathways as well as show age-related changes (Carmena et al. 1990; Razzaboni and Terner 1988). Since prostate cancer is diagnosed at the mean age of 70 years, age-related changes in cAMP combined with age-related decreases in circulating androgens may play a role in the pathology of the prostate. Advanced prostate cancers often have increased numbers of NE cells that secrete neuropeptides. These neuropeptides can increase proliferation of prostate cancer cells or by acting through their GPCRs transactivate the AR to promote progression (Lee et al. 2001; Shah et al. 1994). Table 3 lists just a few compounds that elevate levels of cAMP in prostate cells to alter proliferation. Consistent with cell-specific effects of cAMP and differences in downstream signaling, enhanced levels of cAMP may have differential effects on proliferation. In addition to cell-specific effects, other factors appear to play a role in the proliferative response to elevated cAMP that include cell density, transient versus sustained levels of cAMP, and other culture conditions, such as presence of serum.

Table 3 Compounds that increase cAMP and alter proliferation of prostate cells Compound Sample Proliferation Calcitonin DU145, LNCaP, PC-3, Increases PC-3M Calcitonin geneALVA-31, PPC-1, benign Increases related peptide epithelial cells, LNCaP Dibutyryl cAMP PC-3 Inhibits Forskolin LNCaP Inhibits (10 mM forskolin) Increases (1 mM forskolin) Gastrin-releasing DU145, PC-3, LNCaP Increases in DU145, LNCaP and peptide/ CWR22R CWR22R; decreases in PC-3 bombesin Interferon-a PC-3 Inhibits LHRH DU145 Not measured Melatonin LNCaP, PC-3 Inhibits Nitroprusside PC-3 cells Inhibits (100 mM) PACAP PC-3, LNCaP Increases and decreases with sustained cAMP levels Prolactin PC-3 Increases Serotonin PC-3 Decreases Sex hormoneLNCaP Increases binding globulin VIP Benign prostate epithelial Increases cells, rat ventral prostate, LNCaP, CWR22RV1, PC-346C, PC-3 transfected with AR Carmena and Prieto 1983; Gkonos et al. 1995; Jongsma et al. 2000; Juarranz et al. 2001; Xie et al. 2007

Juang 2004 Jongsma et al. 2000 Nakhla et al. 1990; Rosner et al. 1991

Farini et al. 2003; Juarranz et al. 2001; Leyton et al. 1998

Okutani et al. 1991 Culig et al. 1997 Rimler et al. 2001; Sainz et al. 2005 Juang 2004

Jongsma et al. 2000; Lee et al. 2001

Okutani et al. 1991 Farini et al. 2003; Jongsma et al. 2000; Ueda et al. 2002a

References Chien and Shah 2001; Gkonos et al. 1995; Jongsma et al. 2000; Sabbisetti et al. 2005; Shah et al. 1994 Chien and Shah 2001; Gkonos et al. 1995

The Role of Cyclic AMP in Regulating the Androgen Receptor 477

478

3.1

M.D. Sadar

Neuropeptides

Calcitonin and its receptor are expressed by the epithelium of the prostate. Levels of this secreted protein are elevated in cancer compared with benign tissues. Calcitonin increases proliferation of prostate cancer cells through cAMP and calcium/ phospholipids pathways. Calcitonin causes a transient increase in PKA activity and increased invasion can be blocked by inhibitors of PKA (Sabbisetti et al. 2005). Gastrin-releasing peptide/bombesin receptors are expressed in LNCaP, PC-3, and DU145 cells (Aprikian et al. 1996; Bartholdi et al. 1998; Markwalder and Reubi 1999) and elevated levels of these receptors have been measured in prostate cancer tissues (Markwalder and Reubi 1999). Bombesin increases the proliferation of LNCaP, CWR22R, and DU145 cells in the absence of androgens (Jongsma et al. 2000; Lee et al. 2001). Bombesin stimulated increased cAMP (Jongsma et al. 2000) and increased transactivation of the AR (Dai et al. 2002; Lee et al. 2001). The mechanism of transactivation of the AR is suspected to involve three nonreceptor tyrosine kinases, FAK, Src, and Etk, which integrate signals generated by GPCRs and tyrosine kinases, to the AR (Lee et al. 2001). Through Src and PKCd pathways, bombesin increased the activity of histone acetyltransferase of the AR coregulator, p300, which resulted in enhanced acetylation and transcriptional activity of the AR (Gong et al. 2006). Luteinizing hormone-releasing hormone (LHRH) increased cAMP in DU145 cells and transactivated the AR (Culig et al. 1997). The LHRH receptor is expressed in LNCaP and DU145 cells. However, this receptor was reported to be coupled to Ga(i) protein-cAMP signal transduction pathway, which would inhibit forskolininduced cAMP rather than the Ga(q/11)phospholipase C signaling system like in the anterior pituitary gland (Limonta et al. 1999). Coupling of the LHRH receptor to different G-proteins provides a basis for different actions of LHRH in the anterior pituitary compared to prostate cancer. Melatonin is secreted from the pineal gland and levels decrease with age. Melatonin inhibits both the growth and activity of the AR in prostate cancer cells, an effect which is not mediated by PKA activation in spite of transient increases in cAMP (Sainz et al. 2005). Cell density was important for the observed increases (high density), or decreases (low density), in cAMP in response to melatonin in PC-3 cells (Gilad et al. 1999). Studies on pituitary adenylate cyclase-activating polypeptide (PACAP) in prostate cancer have revealed differences depending on transient versus sustained elevation of cAMP. PACAP generates a transient increase in cAMP to increase proliferation through PKA-dependent activation of the MAPK pathway. Chronic PACAP stimulation with sustained cAMP accumulation inhibited proliferation and caused NE differentiation (Farini et al. 2003). Long-term elevation of cAMP mimics steroid depletion in vitro and results in NE differentiation and increased density of calcium current through low-voltage activated T-type calcium channels. LNCaP cells only express the a(1H) calcium channel for which levels are elevated with NE differentiation (Mariot et al. 2002).

The Role of Cyclic AMP in Regulating the Androgen Receptor

479

Vasoactive intestinal peptide (VIP) had no effect on AC in PC-3 cells but increased activity in LNCaP cells (Hoosein et al. 1993). VIP transactivates the AR through a PKA/Rap1/ERK1/2 pathway to promote androgen-independent growth (Xie et al. 2007). VIP works through VPAC receptor to increase cAMP and induce NE differentiation in LNCaP cells through PKA, ERK1/2, and PI3K (Gutierrez-Canas et al. 2005). LNCaP cells express VIP/PACAP receptor subtypes PAC1, VPAC1, and VPAC2 with a major role for VPAC1 and AC stimulation (Juarranz et al. 2001).

3.2

Cyclic AMP Analogues and Activators of AC

Application of forskolin, with or without IBMX, epinephrine, isoproterenol, or cAMP analogues to elevate levels of cellular cAMP, in prostate cancer cells maintained in vitro generally decreased proliferation in most (Bang et al. 1994; Cox et al. 1999; Farini et al. 2003; Okutani et al. 1991) but not all reports (Mizokami et al. 1994; Ueda et al. 2002a). These discrepancies may be due to transient versus sustained levels of cAMP, application of different culture conditions (serum free or not), concentrations of forskolin employed (>10 mM, with or without IBMX, showed inhibition effects versus 1 mM of forskolin that showed proliferation), the assay used to measure proliferation, and/or the cell density. Application of dibutyryl-cyclic AMP analogue to elevate levels of cAMP requires caution in interpretation because the butyrate moiety has considerable effects on its own. These effects include inhibition of proliferation, induction of differentiation, altering gene expression, inhibition of histone deacetylase activity, and activation of the AR in prostate cancer cells (Sadar and Gleave 2000). Divergent effects of cAMP and dibutyryl-cyclic AMP have been noted for almost 40 years (Hilz and Tarnowski 1970; Solomon et al. 1970).

3.3

Androgens

Connection between the cAMP and androgen pathways in the prostate has been investigated for over 40 years. In the late 1960s, it was realized that castration of rats had a marked effect on the activities of prostatic hexokinase, phosphofructokinase, and glucose 6-phosphate dehydrogenase that could be reversed by administration of testosterone (Singhal and Ling 1969). The rat ventral prostate shows high expression for b-adrenergic receptors implying a function for neurotransmitters in this androgen-dependent tissue. b-Adrenergic receptors and AC have been reported to be altered by circulating androgens (Collins et al. 1988; Guthrie et al. 1990; Poyet et al. 1986; Purvis et al. 1986; Shima et al. 1980; Sutherland and Singhal 1974a; Thomas and Singhal 1973). Early studies measured decreased AC activity in the prostate after castration that was 37% of the normal values in mature male rats. Restoration of levels of testosterone or DHT restored AC activity (Purvis et al.

480

M.D. Sadar

1986; Singhal et al. 1971; Sutherland and Singhal 1974b). Antiandrogens also decreased AC activity and cAMP levels (Tsang and Singhal 1976). Theophylline enhanced the effects of testosterone on levels of prostatic enzymes (Singhal et al. 1971), which were blocked by the antiandrogen cyproterone acetate and the b-adrenergic blocking agent propranolol. The increase in AC activities in the prostate was specific to androgens since estradiol-17b, progesterone, or prednisolone had no effect (Sutherland and Singhal 1974a). Androgens have also been reported to regulate levels of PKA-type I in the rat ventral prostate (Fuller et al. 1978) and induce autophosphorylation of PKA type II (Liu et al. 1981). However, other studies contradicted these observations. Mangan et al. (1973) could not detect increased levels of cAMP in response to androgen. Liao et al. (1971) also did not detect any alteration in AC activity in prostatic nuclei after castration or administration of testosterone in vivo, which was consistent with Rosenfeld and O’Malley (1970). Thirty years later, a ninefold (p = 0.02) increase in cAMP was measured in LNCaP cells cultured in androgen-depleted media (Burchardt et al. 1999). The in vivo increase in NE cells after castration (Burchardt et al. 1999) could imply elevation of cAMP and PKA activity since this pathway plays a predominant role in promoting NE differentiation (Cox et al. 1999, 2000). Prostate cancer patients with bone metastases and refractory disease have elevated levels of cAMP in their urine (Buchs et al. 1998; Murray et al. 2001).

4 Androgen-Regulated Genes Are Induced by Stimulation of the cAMP/PKA Pathway 4.1

Prostate-Specific Antigen

Early studies investigated whether activation of the cAMP pathway would affect expression of genes normally regulated by androgens in prostate cancer cells. One gene that was identified as being of clinical importance was PSA. Measurement of levels of PSA in the serum of men has been a useful tool for the earlier detection of prostate cancer and monitoring of treatment response and recurrence. Most prostate cancer patients treated with androgen deprivation therapy show an initial response of decreased levels of serum PSA that correlate to decreased tumor burden. However, after this initial response, relapse occurs where the malignancy grows in spite of medical or surgical castration. The earliest indication of progression during androgen deprivation is a rising titer of serum PSA that is associated with poor prognosis. Interest in the transcriptional regulation of the PSA gene increased as its clinical applications for prostate cancer gained widespread use. Expression of PSA is regulated by androgens (Goldfarb et al. 1986) through the AR (Young et al. 1991) binding to AREs in the promoter and enhancer regions (Cleutjens et al. 1996, 1997; Riegman et al. 1991). High concentrations of forskolin (10 to 100 mM) inhibited (Blok et al. 1998; Sadar 1999) or did not alter (Andrews

The Role of Cyclic AMP in Regulating the Androgen Receptor

481

et al. 1992) levels of PSA mRNA induced by androgens in LNCaP cells. Lower concentrations of forskolin (1 mM) increased PSA mRNA in serum-free conditions devoid of androgens and enhanced androgen induction of PSA gene expression (Sadar 1999). These findings demonstrated cross talk between the androgen and cAMP/PKA signaling pathways. The discovery that other signaling pathways, such as interleukin-6 (Ueda et al. 2002a), could also increase PSA gene expression in the absence of androgens confirmed findings in the clinic of reexpression of PSA in patients failing androgen deprivation therapy. If the mechanism could be determined about how one androgen-regulated gene, such as PSA, can escape regulation by androgen, clues may be revealed about how androgen-dependent prostate cancer escapes dependency on androgen in advanced prostate cancer.

4.2

Genes Regulated by Both Androgen and Forskolin

Progress in technology has enabled evaluation of expression of many genes simultaneously. Many high-throughput approaches have been used to identify androgenregulated genes in prostate cancer cells including cDNA arrays, serial analysis of gene expression, and Affymetrix GeneChip Arrays (see Dehm and Tindall 2006b for a review). The importance of identifying androgen-regulated genes rests in the unique feature of the prostate being an androgen-dependent tissue that relies on androgen for function, differentiation, growth, and maintenance. Upon removal of androgens, prostate cells undergo apoptosis, which provides the rationale of androgen deprivation for the systemic treatment of advanced prostate cancer. Animal models of progression of prostate cancer in response to castration have revealed that some of the genes normally expressed in the presence of androgens become reexpressed upon recurrence (Gregory et al. 1998). The androgen-regulated PSA gene is thus not unique in this expression pattern. To reveal if PSA was unique in its regulation of expression by both androgen and in the absence of androgen by stimulation of the cAMP/PKA pathway, studies were completed using Affymetrix GeneChip arrays. Comparison of the transcriptomes in LNCaP cells in response to androgen and forskolin allowed the analyses of approximately 47,000 transcripts. Expression of 858 genes changed significantly (p < 0.01) in response to androgen and expression of 303 genes changed in response to forskolin (Wang et al. 2006a). These studies revealed an overlap in the expression of a subset of 22 genes that included PSA (KLK3), the related gene KLK2, and SESN1 (Wang et al. 2006a). Table 4 provides the identities of the genes that were common to both pathways. Androgen inhibited expression of PKIB, a competitive inhibitor of PKA (Wang et al. 2006a). Cross talk between these two pathways may play a role in progression as suggested from validation studies using in vivo samples (Wang et al. 2006a). The mechanism by which stimulation of cAMP/ PKA, in the absence of androgen, alters the expression of androgen-regulated genes has predominantly focused upon the AR.

Table 4 Expression of genes that significantly increased in LNCaP cells in response to R1881 and forskolin (Wang et al. 2006a) Symbol Affymetrix Fold change in p-value in Fold change p-value in Description ID R1881 R1881 in FSK FSK ACSL3 201661_s_at 6.34 0.00321 1.32 0.025 Acyl-CoA synthetase long-chain family member 3 C20orf155 223978_s_at 2.93 0.00039 – – Chromosome 20 open reading frame 155 225324_at 2.60 0.00078 – – 232118_at – – 1.77 0.0401 CAMKK2 210787_s_at 6.65 0.00112 1.50 0.0171 Calcium/calmodulin-dependent protein kinase kinase 2, b Dlc2 229106_at 2.10 0.00607 1.52 0.0255 Dynein light chain 2 EG1 225159_s_at 4.44 0.00328 – – Endothelial-derived gene 1 218438_s_at – – 1.32 0.00731 EVI5 209717_at 2.04 0.00479 1.32 0.0426 Ecotropic viral integration site 5 FADS1 208962_s_at 2.67 0.00156 1.34 0.0452 601066683F1 NIH_MGC_10 Homo sapiens cDNA clone IMAGE:3452925 50 , mRNA sequence FLJ22649 222753_s_at 3.77 0.00872 1.80 0.0111 Hypothetical protein FLJ22649 similar to signal peptidase SPC-22/23 GNAI3 201179_s_at 2.11 0.0044 – – Guanine nucleotide-binding protein (G-protein), a-inhibiting activity polypeptide 3 201181_at – – 1.64 0.0452 HMGCR 202540_s_at 3.59 0.00306 1.56 0.0256 3-Hydroxy-3-methylglutaryl-coenzyme A zreductase INSIG1 201625_s_at 3.83 0.00134 1.71 0.00353 Insulin-induced gene 1 201626_at 4.58 0.0023 1.68 0.0157 KIAA0690 216913_s_at 2.04 0.00851 1.53 0.00579 KIAA0690 KLK2 209854_s_at 31.36 0.00022 1.92 0.00744 Kallikrein 2, prostatic 210339_s_at 21.60 4.08E-05 1.95 0.0085 KLK3 204582_s_at 6.42 0.00172 2.06 0.0118 Kallikrein 3, PSA 204583_x_at 5.38 0.00405 2.03 0.0194 MAP7 215471_s_at 1.93 0.00846 1.35 0.00947 Microtubule-associated protein 7 NAV1 224772_at 2.40 0.00579 – – Neuron navigator 1 227584_at – – 2.07 0.0116 NGLY1 220742_s_at 1.96 0.00498 1.31 0.0274 N-glycanase 1

482 M.D. Sadar

218346_s_at 217766_s_at 211144_x_at 209813_x_at 217823_s_at 221844_x_at

228559_at 229975_at

UBE2J1 –

– –

SESN1 SMP1 TRGV9

23.68 5.23

1.99 1.62 11.44 – 6.16 2.87 0.00125 0.00227

0.00589 0.00744 0.0057 – 0.00074 0.00946 5.90 1.92

1.32 1.46 – 3.18 1.46 1.54 0.0138 0.0111

0.023 0.0228 – 0.0046 0.0439 0.00339

Sestrin 1 Small membrane protein 1 T-cell receptor (V-J-C) precursor; Human T-cell receptor gchain VJCI-CII-CIII region mRNA, complete cds Ubiquitin-conjugating enzyme E2, J1 (UBC6 homolog, yeast) Transcribed sequence with moderate similarity to protein sp: P39195 (H. sapiens) CDNA clone IMAGE:6043059, partial cds Transcribed sequences

The Role of Cyclic AMP in Regulating the Androgen Receptor 483

484

M.D. Sadar

5 Androgen Receptor Androgens mediate their action through the AR. The AR has distinct functional domains involved in ligand-binding, transcriptional activation and repression, and a conserved DNA-binding domain (DBD). In the cryptic basal state, the AR resides in the cytoplasm as a complex with heat-shock proteins and immunophilins. Binding of androgen (ligand) results in an activation step such that the receptor can translocate to the nucleus and bind to its respective DNA elements called AREs. The mechanism of androgen-induced transformation of the AR may include the following: 1. a change in protein conformation required for DNA binding, 2. the removal of associated heat-shock proteins which may act as a cytoplasmic anchor or inhibit DNA interactions, 3. the covalent modification of receptor (e.g. phosphorylation, acetylation) presumed to be required for efficient DNA-binding activity, 4. homodimer formation, and 5. DNA binding to AREs and recruitment of regulatory proteins and components of the basal transcriptional machinery. Many of the steps involved in ligand-induced transactivation of the AR have also been shown for the unliganded AR in response to stimulation of the cAMP/PKA pathway (see Sadar et al. 1999 for a review).

6 Ligand-Independent Activation of the AR by the cAMP/PKA Pathway Cross talk between the AR and cAMP/PKA signal transduction pathways occurs in androgen-depleted human prostate cancer cells maintained in culture (Culig et al. 1997; Nazareth and Weigel 1996; Sadar 1999). These studies have shown that antiandrogens can block cAMP/PKA induction of PSA mRNA (Sadar 1999) and androgen-responsive reporters (Culig et al. 1997; Nazareth and Weigel 1996; Sadar 1999). PKA induction of reporter gene constructs containing AREs required a functional AR as indicated using PC-3 cells transfected with and without wildtype AR (Sadar 1999). Activation of the AR by cAMP/PKA was not by a mechanism that involved increased expression of AR protein. Rather, enhanced levels of AR protein in the nucleus suggested nuclear translocation of the receptor and increased DNA-binding activity of the AR to AREs (Nazareth and Weigel 1996; Sadar 1999). More AR–ARE complex formation occurred in the presence of nuclear extracts from forskolin-treated cells than from androgen-treated cells, even though the nuclear levels of AR were approximately tenfold higher in the androgen-treated cells (Sadar 1999). This suggests that the forskolin-transformed AR may have a greater affinity for the PSA–ARE than the receptor activated by androgen. Such a theory is supported by the fact that the AR NTD is activated by

The Role of Cyclic AMP in Regulating the Androgen Receptor

485

PKA in LNCaP cells (Sadar 1999) and this region of the AR contributes to the stability of the receptor–DNA complex (Kallio et al. 1994). The earliest studies examining cross talk between phosphorylation pathways and the AR employed the rat AR. The rat AR has 100% homology in the DBD and ligand-binding domain (LBD) with the 74% homology in the NTD. Studies with the rat AR show that modulators of PKA activity have negligible effects on ligandindependent activation of the receptor using the MMTV-CAT reporter in CV1 cells maintained in serum (Ikonen et al. 1994). In serum-free conditions, these same cells demonstrated robust and significant ligand-independent action with the human AR using the GRE2E1b-CAT reporter (Nazareth and Weigel 1996). Together these data emphasize potential species-specific and reporter-specific response and/or the importance of cultures conditions. In yeast and COS1 cells, the stimulation of cAMP/PKA only had a ligand-dependent effect on AR activity (Ikonen et al. 1994; Rana et al. 1999). Forskolin promoted binding of the rat AR to an ARE but required androgen to initiate transcription in yeast cells (Rana et al. 1999), but this was not observed in COS1 cells (Ikonen et al. 1994). Forskolin did not alter ligand binding of the rat AR (Ikonen et al. 1994; Rana et al. 1999) nor did removal of 40–147 amino acids of the rat AR have any effect, which suggest that phosphorylation sites within this region were not involved (Ikonen et al. 1994; Rana et al. 1999). Gene specificity for induction of ARE-driven reporters by cAMP/PKA have been noted between the PSA (630/+12)-promoter, probasin (286/+28)-promoter, and the artificial ARR3-tk-promoter that has six AREs from three repeat regions of the probasin promoter linked in tandem in front of the minimal thymidine kinase promoter (Sadar 1999). Reporter specificity has been suggested to require recruitment of gene-specific factors. Specificity in response to cAMP/PKA depending on species, cells, and reporters emphasizes the potential importance of cellular factors, which may be expressed in specific tissues or species. The mechanism of ligand-independent activation of AR in response to stimulation of the cAMP/PKA pathway has not been clarified fully and may involve altering: (1) levels of expression of the AR, (2) nuclear translocation of the AR, (3) posttranslational modifications of the AR, and (4) altering protein–protein interactions with chaperones, coregulators, or other transcription factors.

7 Regulation of the AR Gene The promoter region of the AR gene does not have a TATA box or CCAAT box but contains a GC box to which SP-1 regulates the levels of expression of AR (Faber et al. 1993; Tilley et al. 1990). The human AR gene contains numerous potential cis elements in the promoter region for AP-1, SP-1, NF-1, and CREB (Mizokami et al. 1994). Most studies have not observed an increase in expression of AR protein in response to compounds that increase cAMP at the times relevant to the measurement of enhanced transactivation of the AR (Ikonen et al. 1994; Nazareth and Weigel 1996; Rana et al. 1999; Sadar 1999). However, a slight increase in AR mRNA and increased activity of an AR reporter gene construct in response to

486

M.D. Sadar

cAMP analogues in LNCaP cells have been reported (Mizokami et al. 1994; Stubbs et al. 1996), while the nonfunctional analogue of forskolin, 1,9-dideoxy forskolin, had no effect (Mizokami et al. 1994; Lindzey et al. 1993). CREB protein from reticulolysate bound to a putative CRE at 508 to 501 of the AR promoter (Stubbs et al. 1996). DHT suppressed basal expression of AR through a mechanism that was dependent on levels of AR protein (Lindzey et al. 1993). Although the AR gene has a putative ARE, deletion of this ARE did not eliminate the DHT inhibition of forskolin-induced AR gene reporter activity, thereby implying protein–protein interactions rather than DNA interaction (Lindzey et al. 1993).

8 Posttranslational Modifications Posttranslational modifications of the AR are important in the response to ligand and may include phosphorylation, acetylation, sumoylation, and ubiquitination. These modifications may result in changes in intracellular localization, turnover, and protein–protein interactions. In the absence of ligand, alternative pathways such as cAMP/PKA may induce posttranslational modifications to override the need for androgen to transactivate the receptor.

8.1

Phosphorylation

Phosphorylation of nuclear receptors and their coactivators is a well-documented mechanism involved in the control of their activities (reviewed in Rochette-Egly 2003; Shao and Lazar 1999; Weigel and Moore 2007). The protein sequence for the AR contains a number of consensus phosphorylation sites for many kinases including PKA and the majority of these sites are in the NTD which is thought to be targeted by the cAMP/PKA pathway (Sadar 1999). To date, no studies support that the AR is a direct substrate of PKA. Interactions between the AR with kinases and phosphatases involved in many cellular processes has been reviewed recently (Heemers and Tindall 2007). In regard to cAMP/PKA signaling and cross talk with the AR, RSK is perhaps one kinase of particular interest. RSK1 is a serine/ threonine protein kinase that interacts with the PKA subunit and is a downstream effector of MAPK. RSK increases the expression of androgen-regulated genes even in the absence of ligand by a mechanism that was dependent on the ability of RSK to interact with p300 (Clark et al. 2005). RSK phosphorylated serine 208 of the AR in an in vitro phosphorylation assay but in vivo studies are required for validation. The AR is phosphorylated in response to androgen in LNCaP cells on serines 16, 81, 256, 308, 424, and 650 (Gioeli et al. 2002). This correlation between androgenbinding to the AR and increased phosphorylation of NTD implies that changes in conformation of the AR or chaperone composition are causal to kinase access to phosphorylation sites. Alternatively, androgen may enhance the phosphorylation of

The Role of Cyclic AMP in Regulating the Androgen Receptor

487

the AR by either negatively regulating a phosphatase or inducing a conformation that is resistant to phosphatase action. Recently, ligand-binding of the AR was shown to induce a conformational change that regulates interaction with protein phosphatase 2A (PP2A) (Yang et al. 2007). PP2A interacts with the DBD-hingeLBD of the AR to decrease receptor activity, which correlated to decreased phosphorylation at five phosphoserines (81, 94, 256, 308, and 424) in the AR NTD (Yang et al. 2005). Four of these sites are phosphorylated in response to androgen while serine 94 is constitutively phosphorylated (Gioeli et al. 2002). Forskolin (50 mM), EGF, and phorbol 12-myristate 13-acetate increase phosphorylation at serine 650 (Gioeli et al. 2002). MAPK and Akt do not appear to phosphorylate the AR directly in LNCaP cells stimulated with high levels of forskolin (50 mM) or EGF (Gioeli et al. 2002). In addition, the PKA consensus site at serine 16 was determined using mass spectrometry analysis to not be phosphorylated by PKA (Gioeli et al. 2002). Unfortunately, the elegant studies by Gioeli did not examine concentrations of forskolin except 50 mM, although 1 mM forskolin is optimal to induce PSA mRNA, nuclear translocation, activation of the AR, and its DNAbinding activity, with 50 mM inhibiting AR activity in nontransfected LNCaP cells (Blok et al. 1998; Sadar 1999). Thus, whether PKA under specific cellular conditions directly alters phosphorylation of the AR remains uncertain.

8.2

Acetylation of the AR

The AR is acetylated on its 630KLKK633 motif in the hinge region by p300, P/CAF, and TIF60 (Fu et al. 2000). Mutations of these acetylation sites cause defects in AR trafficking, misfolding, and aggregation similar to expanded glutamine tracts (Thomas et al. 2004). Both bombesin, which increases cAMP in prostate cancer cells, and DHT stimulate acetylation of the AR to enhance transactivation of the AR on promoters of target genes (Fu et al. 2000; Gong et al. 2006). Gain of function mutations of the acetylated lysine residues enhance androgen-dependent transcription and enhance cellular growth by reducing apoptosis (Fu et al. 2002, 2003). AR interaction with p300 correlates to reduced binding of nuclear receptor corepressor-2 (NCoR)/HDAC/Smad3/Sin3A complex and acetylation of the AR regulates cellular growth through recruitment of HDAC/NCoR/Smad complexes to the promoters of genes such as cyclin D1 that regulate the cell cycle (Fu et al. 2006). Acetylation may regulate phosphorylation of the AR at some residues similar to what has been shown for other proteins such as histone H3 (see McManus and Hendzel 2006 for a review); such changes could alter the transcriptional activity of the AR. Serines 641 and 653 lie in proximity to the hinge region and KLKK motif and are suggested to be important in response to high levels of forskolin (Blok et al. 1998). Unfortunately, these serine residues have not been examined directly for phosphorylation in relation to acetylation (Fu et al. 2004).

488

8.3

M.D. Sadar

Sumoylation

Sumoylation is the process of conjugating the small ubiquitin-related modifier (SUMO) protein and has been reported for the AR in the NTD (Poukka et al. 2000) and also for proteins that interact with the AR, such as steroid receptor coactivators (SRC) (Chauchereau et al. 2003; Kotaja et al. 2002; Wu et al. 2006), the histone acetyltransferase p300 (Girdwood et al. 2003), and the NCoR (Tiefenbach et al. 2006). SUMO proteins are usually involved in repressing transcription by recruitment of corepressors or clearance-related mechanisms (Ghisletti et al. 2007; Gill 2005; Hay 2005; Pascual et al. 2005; Seeler and Dejean 2003; Verger et al. 2003). However, sumoylation of SRC-2 enhances the association with AR and promotes the colocalization of SRC-2 and AR in the nuclei and increases AR-mediated transcription (Kotaja et al. 2002). Phosphorylationdependent sumoylation has been shown for MEF2 family members (Flavell et al. 2006; Gregoire et al. 2006; Kang et al. 2006b; Shalizi et al. 2006), HSF-1 (Hietakangas et al. 2003), and peroxisome proliferator-activated receptor (PPAR)-g (Yamashita et al. 2004). Thus, changes in the phosphorylation state of the AR or interacting coregulatory proteins via stimulation of the cAMP/PKA pathway could lead to alterations in sumoylation to alter nuclear localization and transactivation of the AR.

9 Nuclear Translocation Transcriptional activity of the AR requires translocation from the cytoplasm to the nucleus. Relatively little is understood about the mechanisms required for translocation of the receptor to the nucleus. The AR has a ligand-dependent bipartite nuclear localization signal in the DBD and hinge region (Jenster et al. 1993; Zhou et al. 1994). Activation of the cAMP/PKA pathway also leads to increased nuclear levels of AR protein by an unknown mechanism. However, proteins involved in phosphorylation have been suggested to play important roles. The strongest evidence for the importance of phosphorylation of the AR in cellular localization can be drawn from the detection of increased phosphorylation of serine 81 in cells with mutants of the AR that are defective in nuclear export (Black et al. 2004). The tyrosine kinase, Scr, also appears to play a dominant role in nuclear translocation of the AR (Guo et al. 2006). Adaptor/scaffolding protein receptor for activated C kinase 1 (RACK1) interacts with and increases nuclear translocation of the AR in the absence of androgens in response to activation of PKC (Rigas et al. 2003). RACK1 is also involved with Scr phosphorylation of the AR (Kraus et al. 2006). The loss of phosphatase and tensin homologue (PTEN) is frequently associated with prostate cancer. PTEN interacts with the AR to inhibit its nuclear translocation and promote degradation of the AR (Lin et al. 2004). Translocation of the AR has also been shown to be altered by interacting proteins (see Heemers and

The Role of Cyclic AMP in Regulating the Androgen Receptor

489

Tindall 2007 for a review). Any of these interactions could potentially be modified by changes in the cAMP/PKA pathway and downstream events involving phosphorylation.

10

Coregulators

Transcriptional activity of the AR requires functional and structural interaction with coregulators and is modulated by interactions with other transcription factors. Approximately 170 proteins that are involved in a variety of processes have been reported to interact with the AR (see Heemers and Tindall 2007 for a review). Levels of expression, posttranslational modifications, or cellular localization may be altered for any of these coregulators by stimulation of the cAMP/PKA pathway leading to altered protein–protein interactions with the AR and subsequent effects on its transcriptional activity. Changes in protein–protein interactions of the AR with coregulators via stimulation of the cAMP/PKA pathway may mediate a transcriptional active receptor in the absence of ligand. Precedence for such a mechanism can be drawn from the estrogen-related receptors (ERRs) that are orphan receptors for which there are no known ligands. There are three isoforms of ERRs that share amino acid similarity in their NTD, which emphasizes the importance of this domain in transcriptional activation. ERRs can be activated in the absence of ligand by interactions with SRC and the PPAR-g coactivator 1 families (Hong et al. 1999; Huss et al. 2002; Schreiber et al. 2003; Xie et al. 1999). Crystal structure analyses of ERRa bound to a PPAR-g coactivator 1a peptide, and ERRg bound to a SRC-1 or RIP-140 peptides, have shown that both ERRs assume the conformation of ligand-activated nuclear receptors in the absence of a ligand (Greschik et al. 2002; Kallen et al. 2004; Wang et al. 2006b). This strongly suggests that the presence of an agonist ligand is not an essential requirement for activation of receptors. Similar results also have been obtained with estrogen receptor-b and chick progesterone receptor (PR) that are transactivated in the absence of ligand by the cAMP/PKA pathway involving enhanced protein– protein interactions with SRC-1 and CBP (Dutertre and Smith 2003; Rowan et al. 2000; Tremblay et al. 1999). The phosphorylation state of coregulators of the AR may potentially play a more prominent role in altering receptor activity in response to compounds that alter the cAMP/PKA pathway. One example is SRC that is covalently modified by sumoylation, acetylation, and phosphorylation (see Li and Shang 2007 for a review). The cAMP/PKA signaling cascade leads to the phosphorylation of SRC-1 at MAPK sites of Thr-1179 and Ser-1185 to facilitate its interaction with p300/CBP. This interaction elicits optimal activation of both ligand-dependent and ligandindependent transcriptional activity of the chick PR (Li and Shang 2007) without change in phosphorylation of the PR itself (Bai et al. 1997). The phosphorylation of SRC-1 also is required for ligand-dependent and ligand-independent activation of the AR (Ueda et al. 2002b). Ligand-independent activation of the AR did not occur in response to overexpression of either wild-type SRC-1 or overexpression of the

490

M.D. Sadar

mutant SRC-1 that mimics the phosphorylated form (Ueda et al. 2002b). This suggests that additional factors must still be required for transactivation of the unliganded AR. One such candidate mechanism may be the loss of interaction with a corepressor. Transfection of the catalytic subunit of PKA or addition of cAMP analogues inhibited the binding of NCoR2 to the AR NTD (Dotzlaw et al. 2002), which provides a potential mechanism of increased activity of the AR via the cAMP/PKA pathway.

11

Conclusions

cAMP/PKA may regulate the AR through multiple mechanisms (Table 5). These mechanisms could involve canonically cAMP signaling with activation of PKA. Nongenotropic signaling or rapid signaling in response to androgen results in transcriptionally active CREB. CREB regulates the expression of the AR and coregulators and cooperates with the AR through CBP to increase the transcription of genes. Activation of PKA can result in changes in the phosphorylation state of potentially the AR and any interacting protein to affect the assembly of the AR transcriptional complex. Changes in phosphorylation may directly alter protein– protein interactions or indirectly alter interactions through changes in other posttranslational modifications, such as sumoylation and acetylation, or lead to changes in degradation, expression, and cellular localization of essential proteins. The potential for the cAMP/PKA pathway to modulate transactivation of the AR by altering protein–protein interactions is an attractive model with precedent from other nuclear receptors that can be activated in the absence of ligand. Transactivation of the AR is predominantly through AF1 in the NTD (Jenster et al. 1995). This domain contains the region required for interaction with the LBD (see He and Wilson 2002 for a review). The NTD is posttranslationally modified, which can alter protein interactions and localization, and is flexible with a high degree of intrinsic disorder. This limited structure is thought to require interactions with other proteins to assume correct folding for further protein–protein interactions, which Table 5 Mechanisms of cross talk between the cAMP and androgen pathways Nongenotropic signaling in response to androgen leads to transcriptionally active CREB The AR gene contains a functional CRE to which CREB binds to increase levels of AR Levels of expression of coactivators of the AR are regulated by CREB AR and CREB cooperate by a mechanism involving CBP on the promoter and enhancer of genes Activation of cAMP/PKA leads to phosphorylation of coactivators required for the transcriptional activity of the AR Phosphorylation or its effects on other posttranslational modifications such as acetylation, sumoylation, ubiquitination of the AR or coregulators may alter interactions, levels of expression, or cellular localization of essential components required for the AR transcriptional complex cAMP/PKA inhibits binding of the corepressors such as NCoR2 to the AR NTD

The Role of Cyclic AMP in Regulating the Androgen Receptor

491

include interactions with bridging factors and the basal transcriptional machinery, to result in active transcription (Reid et al. 2002). The NTD is activated in the absence of androgen by alternative pathways, which include the cAMP/PKA pathway (Blaszczyk et al. 2004; Dehm and Tindall 2006a; Gregory et al. 2004; Sadar 1999; Sadar et al. 1999; Ueda et al. 2002a, 2002b). Evidence that the NTD of the AR plays a key role in progression of prostate cancer has been shown recently in vivo by application of decoy molecules (Quayle et al. 2007). These studies were premised upon competing away essential interacting proteins of the NTD using a decoy. In vivo expression of the decoys decreased tumor incidence and inhibited the growth of prostate cancer both in the presence and in the absence of androgen (Quayle et al. 2007). Such studies emphasize the importance of identifying and characterizing the posttranslational modifications of the AR and interacting proteins to develop new drugs to improve the clinical outcome of men with advanced prostate cancer.

Acknowledgments I apologize to those authors whose work were inadvertently overlooked or could not be included due to limitations of space. Thank you to Dr. Joanne Johnson for helping prepare the manuscript. This work was supported by the National Institutes of Health CA105304 and the Canadian Institutes for Health Research MOP79308.

References Agrawal S, Kandimalla ER, Yu D, Ball R, Lombardi G, Lucas T, Dexter DL, Hollister BA, Chen SF (2002) GEM 231, a second-generation antisense agent complementary to protein kinase A RIalpha subunit, potentiates antitumor activity of irinotecan in human colon, pancreas, prostate and lung cancer xenografts. Int J Oncol 21:65–72. Ahn S, Olive M, Aggarwal S, Krylov D, Ginty DD, Vinson C (1998) A dominant-negative inhibitor of CREB reveals that it is a general mediator of stimulus-dependent transcription of c-fos. Mol Cell Biol 18:967–977. Andrews PE, Young CY, Montgomery BT, Tindall DJ (1992) Tumor-promoting phorbol ester down-regulates the androgen induction of prostate-specific antigen in a human prostatic adenocarcinoma cell line. Cancer Res 52:1525–1529. Aprikian AG, Han K, Chevalier S, Bazinet M, Viallet J (1996) Bombesin specifically induces intracellular calcium mobilization via gastrin-releasing peptide receptors in human prostate cancer cells. J Mol Endocrinol 16:297–306. Bai W, Rowan BG, Allgood VE, O’Malley BW, Weigel NL (1997) Differential phosphorylation of chicken progesterone receptor in hormone-dependent and ligand-independent activation. J Biol Chem 272:10457–10463. Bang YJ, Pirnia F, Fang WG et al. (1994) Terminal neuroendocrine differentiation of human prostate carcinoma cells in response to increased intracellular cyclic AMP. Proc Natl Acad Sci USA 91:5330–5334.

492

M.D. Sadar

Bartholdi MF, Wu JM, Pu H, Troncoso P, Eden PA, Feldman RI (1998) In situ hybridization for gastrin-releasing peptide receptor (GRP receptor) expression in prostatic carcinoma. Int J Cancer 79:82–90. Beavo JA (1995) Cyclic nucleotide phosphodiesterases: Functional implications of multiple isoforms. Physiol Rev 75:725–748. Black BE, Vitto MJ, Gioeli D, Spencer A, Afshar N, Conaway MR, Weber MJ, Paschal BM (2004) Transient, ligand-dependent arrest of the androgen receptor in subnuclear foci alters phosphorylation and coactivator interactions. Mol Endocrinol 18:834–850. Blaszczyk N, Masri BA, Mawji NR et al. (2004) Osteoblast-derived factors induce androgenindependent proliferation and expression of prostate-specific antigen in human prostate cancer cells. Clin Cancer Res 10:1860–1869. Blok LJ, de Ruiter PE, Brinkmann AO (1998) Forskolin-induced dephosphorylation of the androgen receptor impairs ligand binding. Biochemistry 37:3850–3857. Bolger GB, McPhee I, Houslay MD (1996) Alternative splicing of cAMP-specific phosphodiesterase mRNA transcripts. Characterization of a novel tissue-specific isoform, RNPDE4A8. J Biol Chem 271:1065–1071. Bos JL (2003) EPAC: A new cAMP target and new avenues in cAMP research. Nat Rev Mol Cell Biol 4:733–738. Bos JL (2005) Linking rap to cell adhesion. Curr Opin Cell Biol 17:123–128. Bruchovsky N, Goldenberg SL, Mawji NR, Sadar MD. (2001) Evolving aspects of intermittent androgen blockage for prostate cancer: Diagnosis and treatment of early tumor progression and maintenance of remission. Andrology in the 21st Century, Proceedings of the VIIth International Congress of Andrology 609–623. Buchs N, Bonjour JP, Rizzoli R (1998) Renal tubular reabsorption of phosphate is positively related to the extent of bone metastatic load in patients with prostate cancer. J Clin Endocrinol Metab 83:1535–1541. Buck J, Sinclair ML, Schapal L, Cann MJ, Levin LR (1999) Cytosolic adenylyl cyclase defines a unique signaling molecule in mammals. Proc Natl Acad Sci USA 96:79–84. Burchardt T, Burchardt M, Chen MW, Cao Y, de la Taille A, Shabsigh A, Hayek O, Dorai T, Buttyan R (1999) Transdifferentiation of prostate cancer cells to a neuroendocrine cell phenotype in vitro and in vivo. J Urol 162:1800–1805. Cali JJ, Parekh RS, Krupinski J (1996) Splice variants of type VIII adenylyl cyclase: Differences in glycosylation and regulation by Ca2+/calmodulin. J Biol Chem 271:1089–1095. Carmena MJ, Prieto JC (1983) Cyclic AMP-stimulating effect of vasoactive intestinal peptide in isolated epithelial cells of rat ventral prostate. Biochim Biophys Acta 763:414–418. Carmena MJ, Hueso C, Recio MN, Prieto JC (1990) Beta-adrenergic stimulation of cyclic AMP accumulation in rat prostatic epithelial cells during sexual maturation. Mech Ageing Dev 52:79–86. Cao X, Qin J, Xie Y, Khan O, Dowd F, Scofield M, Lin MF, Tu Y (2006) Regulator of G-protein signaling 2 (RGS2) inhibits androgen-independent activation of androgen receptor in prostate cancer cells. Oncogene 25:3719–3734. Chaturvedi D, Poppleton HM, Stringfield T, Barbier A, Patel TB (2006) Subcellular localization and biological actions of activated RSK1 are determined by its interactions with subunits of cyclic AMP-dependent protein kinase. Mol Cell Biol 26:4586–4600. Chauchereau A, Amazit L, Quesne M, Guiochon-Mantel A, Milgrom E (2003) Sumoylation of the progesterone receptor and of the steroid receptor coactivator SRC-1. J Biol Chem 278:12335– 12343. Chen CD, Welsbie DS, Tran C, Baek SH, Chen R, Vessella R, Rosenfeld MG, Sawyers CL (2004) Molecular determinants of resistance to antiandrogen therapy. Nat Med 10:33–39. Chen Y, Cann MJ, Litvin TN, Iourgenko V, Sinclair ML, Levin LR, Buck J (2000) Soluble adenylyl cyclase as an evolutionarily conserved bicarbonate sensor. Science 289:625–628. Chien J, Shah GV (2001) Role of stimulatory guanine nucleotide binding protein (GSalpha) in proliferation of PC-3M prostate cancer cells. Int J Cancer 91:46–54.

The Role of Cyclic AMP in Regulating the Androgen Receptor

493

Cho YS, Cho-Chung YS (2003) Antisense protein kinase A RIalpha acts synergistically with hydroxycamptothecin to inhibit growth and induce apoptosis in human cancer cells: Molecular basis for combinatorial therapy. Clin Cancer Res 9:1171–1178. Cho YS, Lee YN, Cho-Chung YS (2000a) Biochemical characterization of extracellular cAMPdependent protein kinase as a tumor marker. Biochem Biophys Res Commun 278:679–684. Cho YS, Park YG, Lee YN, Kim MK, Bates S, Tan L, Cho-Chung YS (2000b) Extracellular protein kinase A as a cancer biomarker: Its expression by tumor cells and reversal by a myristate-lacking calpha and RIIbeta subunit overexpression. Proc Natl Acad Sci USA 97:835–840. Cho-Chung YS, Nesterova MV (2005) Tumor reversion: Protein kinase A isozyme switching. Ann NY Acad Sci 1058:76–86. Choi EJ, Xia Z, Storm DR (1992) Stimulation of the type III olfactory adenylyl cyclase by calcium and calmodulin. Biochemistry 31:6492–6498. Christensen AE, Selheim F, de Rooij J et al. (2003) cAMP analog mapping of Epac1 and cAMP kinase. Discriminating analogs demonstrate that EPAC and cAMP kinase act synergistically to promote PC-12 cell neurite extension. J Biol Chem 278:35394–35402. Clark DE, Errington TM, Smith JA, Frierson HF, Jr, Weber MJ, Lannigan DA (2005) The serine/ threonine protein kinase, p90 ribosomal S6 kinase, is an important regulator of prostate cancer cell proliferation. Cancer Res 65:3108–3116. Cleutjens KB, van Eekelen CC, van der Korput HA, Brinkmann AO, Trapman J (1996) Two androgen response regions cooperate in steroid hormone regulated activity of the prostatespecific antigen promoter. J Biol Chem 271:6379–6388. Cleutjens KB, van der Korput HA, van Eekelen CC, van Rooij HC, Faber PW, Trapman J (1997) An androgen response element in a far upstream enhancer region is essential for high, androgen-regulated activity of the prostate-specific antigen promoter. Mol Endocrinol 11:148–161. Collins S, Quarmby VE, French FS, Lefkowitz RJ, Caron MG (1988) Regulation of the beta 2-adrenergic receptor and its mRNA in the rat ventral prostate by testosterone. FEBS Lett 233:173–176. Conkright MD, Canettieri G, Screaton R, Guzman E, Miraglia L, Hogenesch JB, Montminy M (2003) TORCs: Transducers of regulated CREB activity. Mol Cell 12:413–423. Corbin JD, Sugden PH, West L, Flockhart DA, Lincoln TM, McCarthy D (1978) Studies on the properties and mode of action of the purified regulatory subunit of bovine heart adenosine 30 :50 monophosphate-dependent protein kinase. J Biol Chem 253:3997–4003. Cox ME, Deeble PD, Lakhani S, Parsons SJ (1999) Acquisition of neuroendocrine characteristics by prostate tumor cells is reversible: Implications for prostate cancer progression. Cancer Res 59:3821–3830. Cox ME, Deeble PD, Bissonette EA, Parsons SJ (2000) Activated 30,50 -cyclic AMP-dependent protein kinase is sufficient to induce neuroendocrine-like differentiation of the LNCaP prostate tumor cell line. J Biol Chem 275:13812–13818. Culig Z, Hobisch A, Cronauer MV, Radmayr C, Trapman J, Hittmair A, Bartsch G, Klocker H (1994) Androgen receptor activation in prostatic tumor cell lines by insulin-like growth factor-I, keratinocyte growth factor, and epidermal growth factor. Cancer Res 54:5474–5478. Culig Z, Hobisch A, Hittmair A, Cronauer MV, Radmayr C, Zhang J, Bartsch G, Klocker H (1997) Synergistic activation of androgen receptor by androgen and luteinizing hormone-releasing hormone in prostatic carcinoma cells. Prostate 32:106–114. Cvijic ME, Kita T, Shih W, DiPaola RS, Chin KV (2000) Extracellular catalytic subunit activity of the cAMP-dependent protein kinase in prostate cancer. Clin Cancer Res 6:2309–2317. Dai J, Shen R, Sumitomo M, Stahl R, Navarro D, Gershengorn MC, Nanus DM (2002) Synergistic activation of the androgen receptor by bombesin and low-dose androgen. Clin Cancer Res 8:2399–2405. D’Andrea MR, Qiu Y, Haynes-Johnson D, Bhattacharjee S, Kraft P, Lundeen S (2005) Expression of PDE11A in normal and malignant human tissues. J Histochem Cytochem 53:895–903.

494

M.D. Sadar

de Rooij J, Zwartkruis FJ, Verheijen MH, Cool RH, Nijman SM, Wittinghofer A, Bos JL (1998) EPAC is a Rap1 guanine-nucleotide-exchange factor directly activated by cyclic AMP. Nature 396:474–477. Defer N, Best-Belpomme M, Hanoune J (2000) Tissue specificity and physiological relevance of various isoforms of adenylyl cyclase. Am J Physiol Renal Physiol 279:F400–F416. Dehm SM, Tindall DJ (2006a) Ligand-independent androgen receptor activity is activation function-2-independent and resistant to antiandrogens in androgen refractory prostate cancer cells. J Biol Chem 281:27882–27893. Dehm SM, Tindall DJ (2006b) Molecular regulation of androgen action in prostate cancer. J Cell Biochem 99:333–344. Desdouets C, Matesic G, Molina CA, Foulkes NS, Sassone-Corsi P, Brechot C, Sobczak-Thepot J (1995) Cell cycle regulation of cyclin A gene expression by the cyclic AMP-responsive transcription factors CREB and CREM. Mol Cell Biol 15:3301–3309. Dotzlaw H, Moehren U, Mink S, Cato AC, Iniguez Lluhi JA, Baniahmad A (2002) The amino terminus of the human AR is target for corepressor action and antihormone agonism. Mol Endocrinol 16:661–673. Dutertre M, Smith CL (2003) Ligand-independent interactions of p160/steroid receptor coactivators and CREB-binding protein (CBP) with estrogen receptor-alpha: Regulation by phosphorylation sites in the A/B region depends on other receptor domains. Mol Endocrinol 17:1296–1314. Enserink JM, Christensen AE, de Rooij J, van Triest M, Schwede F, Genieser HG, Doskeland SO, Blank JL, Bos JL (2002) A novel EPAC-specific cAMP analogue demonstrates independent regulation of Rap1 and ERK. Nat Cell Biol 4:901–906. Esposito G, Jaiswal BS, Xie F et al. (2004) Mice deficient for soluble adenylyl cyclase are infertile because of a severe sperm-motility defect. Proc Natl Acad Sci USA 101:2993–2998. Faber PW, van Rooij HC, Schipper HJ, Brinkmann AO, Trapman J (1993) Two different, overlapping pathways of transcription initiation are active on the TATA-less human androgen receptor promoter. The role of Sp1. J Biol Chem 268:9296–9301. Farini D, Puglianiello A, Mammi C, Siracusa G, Moretti C (2003) Dual effect of pituitary adenylate cyclase activating polypeptide on prostate tumor LNCaP cells: Short- and longterm exposure affect proliferation and neuroendocrine differentiation. Endocrinology 144:1631–1643. Fawcett L, Baxendale R, Stacey P, McGrouther C, Harrow I, Soderling S, Hetman J, Beavo JA, Phillips SC (2000). Molecular cloning and characterization of a distinct human phosphodiesterase gene family: PDE11A. Proc Nat Acad Sci USA 28:3702–7. Flavell SW, Cowan CW, Kim TK, Greer PL, Lin Y, Paradis S, Griffith EC, Hu LS, Chen C, Greenberg ME (2006) Activity-dependent regulation of MEF2 transcription factors suppresses excitatory synapse number. Science 311:1008–1012. Francis SH, Turko IV, Corbin JD (2001) Cyclic nucleotide phosphodiesterases: Relating structure and function. Prog Nucleic Acid Res Mol Biol 65:1–52. Fu M, Wang C, Reutens AT, Wang J, Angeletti RH, Siconolfi-Baez L, Ogryzko V, Avantaggiati ML, Pestell RG (2000) p300 and p300/cAMP-response element-binding protein-associated factor acetylate the androgen receptor at sites governing hormone-dependent transactivation. J Biol Chem 275:20853–20860. Fu M, Wang C, Wang J et al. (2002) Androgen receptor acetylation governs trans activation and MEKK1-induced apoptosis without affecting in vitro sumoylation and trans-repression function. Mol Cell Biol 22:3373–3388. Fu M, Rao M, Wang C et al. (2003) Acetylation of androgen receptor enhances coactivator binding and promotes prostate cancer cell growth. Mol Cell Biol 23:8563–8575. Fu M, Rao M, Wu K, Wang C, Zhang X, Hessien M, Yeung YG, Gioeli D, Weber MJ, Pestell RG (2004) The androgen receptor acetylation site regulates cAMP and AKT but not ERK-induced activity. J Biol Chem 279:29436–29449.

The Role of Cyclic AMP in Regulating the Androgen Receptor

495

Fu M, Liu M, Sauve AA et al. (2006) Hormonal control of androgen receptor function through SIRT1. Mol Cell Biol 26:8122–8135. Fuller DJ, Byus CV, Russell DH (1978) Specific regulation by steroid hormones of the amount of type I cyclic AMP-dependent protein kinase holoenzyme. Proc Natl Acad Sci USA 75:223–227. Ghisletti S, Huang W, Ogawa S, Pascual G, Lin ME, Willson TM, Rosenfeld MG, Glass CK (2007) Parallel SUMOylation-dependent pathways mediate gene- and signal-specific transrepression by LXRs and PPARgamma. Mol Cell 25:57–70. Ghosh R, Garcia GE, Crosby K, Inoue H, Thompson IM, Troyer DA, Kumar AP (2007) Regulation of cox-2 by cyclic AMP response element binding protein in prostate cancer: Potential role for nexrutine. Neoplasia 9:893–899. Gilad E, Laufer M, Matzkin H, Zisapel N (1999) Melatonin receptors in PC3 human prostate tumor cells. J Pineal Res 26:211–220. Gill G (2005) Something about SUMO inhibits transcription. Curr Opin Genet Dev 15:536–541. Gioeli D, Mandell JW, Petroni GR, Frierson HF, Jr, Weber MJ (1999) Activation of mitogenactivated protein kinase associated with prostate cancer progression. Cancer Res 59:279–284. Gioeli D, Ficarro SB, Kwiek JJ et al. (2002) Androgen receptor phosphorylation. Regulation and identification of the phosphorylation sites. J Biol Chem 277:29304–29314. Girdwood D, Bumpass D, Vaughan OA, Thain A, Anderson LA, Snowden AW, Garcia-Wilson E, Perkins ND, Hay RT (2003) P300 transcriptional repression is mediated by SUMO modification. Mol Cell 11:1043–1054. Gkonos PJ, Lokeshwar BL, Balkan W, Roos BA (1995) Neuroendocrine peptides stimulate adenyl cyclase in normal and malignant prostate cells. Regul Pept 59:43–51. Goldfarb DA, Stein BS, Shamszadeh M, Petersen RO (1986) Age-related changes in tissue levels of prostatic acid phosphatase and prostate specific antigen. J Urol 136:1266–1269. Goluboff ET, Shabsigh A, Saidi JA, Weinstein IB, Mitra N, Heitjan D, Piazza GA, Pamukcu R, Buttyan R, Olsson CA (1999) Exisulind (sulindac sulfone) suppresses growth of human prostate cancer in a nude mouse xenograft model by increasing apoptosis. Urology 53:440–445. Goluboff ET, Prager D, Rukstalis D, Giantonio B, Madorsky M, Barken I, Weinstein IB, Partin AW, Olsson CA, UCLA Oncology Research Network (2001) Safety and efficacy of exisulind for treatment of recurrent prostate cancer after radical prostatectomy. J Urol 166:882–886. Gong J, Zhu J, Goodman OB, Jr, Pestell RG, Schlegel PN, Nanus DM, Shen R (2006) Activation of p300 histone acetyltransferase activity and acetylation of the androgen receptor by bombesin in prostate cancer cells. Oncogene 25:2011–2021. Goodin JL, Rutherford CL (2002) Identification of differentially expressed genes during cyclic adenosine monophosphate-induced neuroendocrine differentiation in the human prostatic adenocarcinoma cell line LNCaP. Mol Carcinog 33:88–98. Gordge PC, Hulme MJ, Clegg RA, Miller WR (1996) Elevation of protein kinase A and protein kinase C activities in malignant as compared with normal human breast tissue. Eur J Cancer 32A:2120–2126. Goto T, Matsushima H, Kasuya Y, Hosaka Y, Kitamura T, Kawabe K, Hida A, Ohta Y, Simizu T, Takeda K (1999) The effect of papaverine on morphologic differentiation, proliferation and invasive potential of human prostatic cancer LNCaP cells. Int J Urol 6:314–319. Gregoire S, Tremblay AM, Xiao L, Yang Q, Ma K, Nie J, Mao Z, Wu Z, Giguere V, Yang XJ (2006) Control of MEF2 transcriptional activity by coordinated phosphorylation and sumoylation. J Biol Chem 281:4423–4433. Gregory CW, Hamil KG, Kim D, Hall SH, Pretlow TG, Mohler JL, French FS (1998) Androgen receptor expression in androgen-independent prostate cancer is associated with increased expression of androgen-regulated genes. Cancer Res 58:5718–5724. Gregory CW, Fei X, Ponguta LA, He B, Bill HM, French FS, Wilson EM (2004) Epidermal growth factor increases coactivation of the androgen receptor in recurrent prostate cancer. J Biol Chem 279:7119–7130.

496

M.D. Sadar

Greschik H, Wurtz JM, Sanglier S, Bourguet W, van Dorsselaer A, Moras D, Renaud JP (2002) Structural and functional evidence for ligand-independent transcriptional activation by the estrogen-related receptor 3. Mol Cell 9:303–313. Guo Z, Dai B, Jiang T et al. (2006) Regulation of androgen receptor activity by tyrosine phosphorylation. Cancer Cell 10:309–319. Gupte RS, Weng Y, Liu L, Lee MY (2005) The second subunit of the replication factor C complex (RFC40) and the regulatory subunit (RIalpha) of protein kinase A form a protein complex promoting cell survival. Cell Cycle 4:323–329. Guthrie PD, Freeman MR, Liao ST, Chung LW (1990) Regulation of gene expression in rat prostate by androgen and beta-adrenergic receptor pathways. Mol Endocrinol 4:1343–1353. Gutierrez-Canas I, Juarranz MG, Collado B, Rodriguez-Henche N, Chiloeches A, Prieto JC, Carmena MJ (2005) Vasoactive intestinal peptide induces neuroendocrine differentiation in the LNCaP prostate cancer cell line through PKA, ERK, and PI3K. Prostate 63:44–55. Han H, Stessin A, Roberts J et al. (2005) Calcium-sensing soluble adenylyl cyclase mediates TNF signal transduction in human neutrophils. J Exp Med 202:353–361. Hay RT (2005) SUMO: A history of modification. Mol Cell 18:1–12. He B, Wilson EM (2002) The NH(2)-terminal and carboxyl-terminal interaction in the human androgen receptor. Mol Genet Metab 75:293–298. Heemers HV, Tindall DJ (2007) Androgen receptor (AR) coregulators: A diversity of functions converging on and regulating the AR transcriptional complex. Endocr Rev 28:778–808. Hess KC, Jones BH, Marquez B et al. (2005) The ‘‘soluble’’ adenylyl cyclase in sperm mediates multiple signaling events required for fertilization. Dev Cell 9:249–259. Hietakangas V, Ahlskog JK, Jakobsson AM, Hellesuo M, Sahlberg NM, Holmberg CI, Mikhailov A, Palvimo JJ, Pirkkala L, Sistonen L (2003) Phosphorylation of serine 303 is a prerequisite for the stress-inducible SUMO modification of heat shock factor 1. Mol Cell Biol 23:2953–2968. Hilz H, Tarnowski W (1970) Opposite effects of cyclic AMP and its dibutyryl derivative on glycogen levels in HeLa cells. Biochem Biophys Res Commun 40:973–981. Hong H, Yang L, Stallcup MR (1999) Hormone-independent transcriptional activation and coactivator binding by novel orphan nuclear receptor ERR3. J Biol Chem 274:22618–22626. Hoosein NM, Logothetis CJ, Chung LW (1993) Differential effects of peptide hormones bombesin, vasoactive intestinal polypeptide and somatostatin analog RC-160 on the invasive capacity of human prostatic carcinoma cells. J Urol 149:1209–1213. Huber PR, Schnell Y, Hering F, Rutishauser G (1987) Prostate specific antigen. Experimental and clinical observations. Scand J Urol Nephrol Suppl 104:33–39. Huss JM, Kopp RP, Kelly DP (2002) Peroxisome proliferator-activated receptor coactivator1alpha (PGC-1alpha) coactivates the cardiac-enriched nuclear receptors estrogen-related receptor-alpha and -gamma. Identification of novel leucine-rich interaction motif within PGC-1alpha. J Biol Chem 277:40265–40274. Ikonen T, Palvimo JJ, Kallio PJ, Reinikainen P, Janne OA (1994) Stimulation of androgenregulated transactivation by modulators of protein phosphorylation. Endocrinology 135:1359–1366. Jaiswal BS, Conti M (2003) Calcium regulation of the soluble adenylyl cyclase expressed in mammalian spermatozoa. Proc Natl Acad Sci USA 100:10676–10681. Jenster G, Trapman J, Brinkmann AO (1993) Nuclear import of the human androgen receptor. Biochem J 293 (Pt 3):761–768. Jenster G, van der Korput HA, Trapman J, Brinkmann AO (1995) Identification of two transcription activation units in the N-terminal domain of the human androgen receptor. J Biol Chem 270:7341–7346. Johannessen M, Delghandi MP, Moens U (2004) What turns CREB on? Cell Signal 16:1211–1227. Jongsma J, Oomen MH, Noordzij MA, Romijn JC, van Der Kwast TH, Schroder FH, van Steenbrugge GJ (2000) Androgen-independent growth is induced by neuropeptides in human prostate cancer cell lines. Prostate 42:34–44.

The Role of Cyclic AMP in Regulating the Androgen Receptor

497

Juang HH (2004) Nitroprusside stimulates mitochondrial aconitase gene expression through the cyclic adenosine 30 ,50 -monosphosphate signal transduction pathway in human prostate carcinoma cells. Prostate 61:92–102. Juarranz MG, Bodega G, Prieto JC, Guijarro LG (2001) Vasoactive intestinal peptide (VIP) stimulates rat prostatic epithelial cell proliferation. Prostate 47:285–292. Kallen J, Schlaeppi JM, Bitsch F, Filipuzzi I, Schilb A, Riou V, Graham A, Strauss A, Geiser M, Fournier B (2004) Evidence for ligand-independent transcriptional activation of the human estrogen-related receptor alpha (ERRalpha): Crystal structure of ERRalpha ligand binding domain in complex with peroxisome proliferator-activated receptor coactivator-1alpha. J Biol Chem 279:49330–49337. Kallio PJ, Palvimo JJ, Mehto M, Janne OA (1994) Analysis of androgen receptor-DNA interactions with receptor proteins produced in insect cells. J Biol Chem 269:11514–11522. Kamenetsky M, Middelhaufe S, Bank EM, Levin LR, Buck J, Steegborn C (2006) Molecular details of cAMP generation in mammalian cells: A tale of two systems. J Mol Biol 362:623–639. Kang G, Chepurny OG, Malester B, Rindler MJ, Rehmann H, Bos JL, Schwede F, Coetzee WA, Holz GG (2006a) cAMP sensor EPAC as a determinant of ATP-sensitive potassium channel activity in human pancreatic beta cells and rat INS-1 cells. J Physiol 573:595–609. Kang J, Gocke CB, Yu H (2006b) Phosphorylation-facilitated sumoylation of MEF2C negatively regulates its transcriptional activity. BMC Biochem 7:5. Kasbohm EA, Guo R, Yowell CW, Bagchi G, Kelly P, Arora P, Casey PJ, Daaka Y (2005) Androgen receptor activation by G(s) signaling in prostate cancer cells. J Biol Chem 280:11583–11589. Khatra BS, Printz R, Cobb CE, Corbin JD (1985) Regulatory subunit of cAMP-dependent protein kinase inhibits phosphoprotein phosphatase. Biochem Biophys Res Commun 130:567–573. Kim D, Gregory CW, French FS, Smith GJ, Mohler JL (2002) Androgen receptor expression and cellular proliferation during transition from androgen-dependent to recurrent growth after castration in the CWR22 prostate cancer xenograft. Am J Pathol 160:219–226. Kim J, Amano O, Wakayama T, Takahagi H, Iseki S (2001) The role of cyclic AMP response element-binding protein in testosterone-induced differentiation of granular convoluted tubule cells in the rat submandibular gland. Arch Oral Biol 46:495–507. Kim J, Jia L, Stallcup MR, Coetzee GA (2005) The role of protein kinase A pathway and cAMP responsive element-binding protein in androgen receptor-mediated transcription at the prostate-specific antigen locus. J Mol Endocrinol 34:107–118. Klauck TM, Faux MC, Labudda K, Langeberg LK, Jaken S, Scott JD (1996) Coordination of three signaling enzymes by AKAP79, a mammalian scaffold protein. Science 271:1589–1592. Kohler K, Louvard D, Zahraoui A (2004) Rab13 regulates PKA signaling during tight junction assembly. J Cell Biol 165:175–180. Kotaja N, Karvonen U, Janne OA, Palvimo JJ (2002) The nuclear receptor interaction domain of GRIP1 is modulated by covalent attachment of SUMO-1. J Biol Chem 277:30283–30288. Kraus S, Gioeli D, Vomastek T, Gordon V, Weber MJ (2006) Receptor for activated C kinase 1 (RACK1) and src regulate the tyrosine phosphorylation and function of the androgen receptor. Cancer Res 66:11047–11054. Kumar AP, Bhaskaran S, Ganapathy M, Crosby K, Davis MD, Kochunov P, Schoolfield J, Yeh IT, Troyer DA, Ghosh R (2007) Akt/cAMP-responsive element binding protein/cyclin D1 network: A novel target for prostate cancer inhibition in transgenic adenocarcinoma of mouse prostate model mediated by nexrutine, a phellodendron amurense bark extract. Clin Cancer Res 13:2784–2794. Kvissel AK, Orstavik S, Oistad P, Rootwelt T, Jahnsen T, Skalhegg BS (2004) Induction of cbeta splice variants and formation of novel forms of protein kinase A type II holoenzymes during retinoic acid-induced differentiation of human NT2 cells. Cell Signal 16:577–587.

498

M.D. Sadar

Kvissel AK, Ramberg H, Eide T, Svindland A, Skalhegg BS, Tasken KA (2007) Androgen dependent regulation of protein kinase A subunits in prostate cancer cells. Cell Signal 19:401–409. Lee LF, Guan J, Qiu Y, Kung HJ (2001) Neuropeptide-induced androgen independence in prostate cancer cells: Roles of nonreceptor tyrosine kinases Etk/Bmx, src, and focal adhesion kinase. Mol Cell Biol 21:8385–8397. Leyton J, Coelho T, Coy DH, Jakowlew S, Birrer MJ, Moody TW (1998) PACAP(6–38) inhibits the growth of prostate cancer cells. Cancer Lett 125:131–139. Li H, Degenhardt B, Tobin D, Yao ZX, Tasken K, Papadopoulos V (2001) Identification, localization, and function in steroidogenesis of PAP7: A peripheral-type benzodiazepine receptor- and PKA (RIalpha)-associated protein. Mol Endocrinol 15:2211–2228. Li S, Shang Y (2007) Regulation of SRC family coactivators by post-translational modifications. Cell Signal 19:1101–1112. Liao S, Lin AH, Tymoczko JL (1971) Adenyl cyclase of cell nuclei isolated from rat ventral prostate. Biochim Biophys Acta 230:535–538. Lim JT, Piazza GA, Han EK et al. (1999) Sulindac derivatives inhibit growth and induce apoptosis in human prostate cancer cell lines. Biochem Pharmacol 58:1097–1107. Lim JT, Piazza GA, Pamukcu R, Thompson WJ, Weinstein IB (2003) Exisulind and related compounds inhibit expression and function of the androgen receptor in human prostate cancer cells. Clin Cancer Res 9:4972–4982. Limonta P, Moretti RM, Marelli MM, Dondi D, Parenti M, Motta M (1999) The luteinizing hormone-releasing hormone receptor in human prostate cancer cells: Messenger ribonucleic acid expression, molecular size, and signal transduction pathway. Endocrinology 140:5250–5256. Lin HK, Hu YC, Lee DK, Chang C (2004) Regulation of androgen receptor signaling by PTEN (phosphatase and tensin homolog deleted on chromosome 10) tumor suppressor through distinct mechanisms in prostate cancer cells. Mol Endocrinol 18:2409–2423. Lindzey J, Grossmann M, Kumar MV, Tindall DJ (1993) Regulation of the 50 -flanking region of the mouse androgen receptor gene by cAMP and androgen. Mol Endocrinol 7:1530–1540. Lipskaia L, Defer N, Esposito G, Hajar I, Garel MC, Rockman HA, Hanoune J (2000) Enhanced cardiac function in transgenic mice expressing a ca(2+)-stimulated adenylyl cyclase. Circ Res 86:795–801. Liu AY, Walter U, Greengard P (1981) Steroid hormones may regulate autophosphorylation of adenosine-30 ,50 -monophosphate-dependent protein kinase in target tissues. Eur J Biochem 114:539–548. Loughney K, Taylor J, Florio VA (2005) 30 ,50 -Cyclic nucleotide phosphodiesterase 11A: Localization in human tissues. Int J Impot Res 17:320–325. Mangan FR, Pegg AE, Mainwaring IP (1973) A reappraisal of the effects of adenosine 30 :50 -cyclic monophosphate on the function and morphology of the rat prostate gland. Biochem J 134:129–142. Mariot P, Vanoverberghe K, Lalevee N, Rossier MF, Prevarskaya N (2002) Overexpression of an alpha 1H (Cav3.2) T-type calcium channel during neuroendocrine differentiation of human prostate cancer cells. J Biol Chem 277:10824–10833. Markwalder R, Reubi JC (1999) Gastrin-releasing peptide receptors in the human prostate: Relation to neoplastic transformation. Cancer Res 59:1152–1159. McManus KJ, Hendzel MJ (2006) The relationship between histone H3 phosphorylation and acetylation throughout the mammalian cell cycle. Biochem Cell Biol 84:640–657. Miller JI, Ahmann FR, Drach GW, Emerson SS, Bottaccini MR (1992) The clinical usefulness of serum prostate specific antigen after hormonal therapy of metastatic prostate cancer. J Urol 147:956–961. Mizokami A, Yeh SY, Chang C (1994) Identification of 30 ,50 -cyclic adenosine monophosphate response element and other cis-acting elements in the human androgen receptor gene promoter. Mol Endocrinol 8:77–88.

The Role of Cyclic AMP in Regulating the Androgen Receptor

499

Moul JW, Srivastava S, McLeod DG (1995) Molecular implications of the antiandrogen withdrawal syndrome. Semin Urol 13:157–163. Murray RM, Grill V, Crinis N, Ho PW, Davison J, Pitt P (2001) Hypocalcemic and normocalcemic hyperparathyroidism in patients with advanced prostatic cancer. J Clin Endocrinol Metab 86:4133–4138. Nakhla AM, Khan MS, Rosner W (1990) Biologically active steroids activate receptor-bound human sex hormone-binding globulin to cause LNCaP cells to accumulate adenosine 30 ,50 monophosphate. J Clin Endocrinol Metab 71:398–404. Nazareth LV, Weigel NL (1996) Activation of the human androgen receptor through a protein kinase A signaling pathway. J Biol Chem 271:19900–19907. Nelson JB (2003) Endothelin inhibition: Novel therapy for prostate cancer. J Urol 170:S65–S67; discussion S67–S68. Nesterova M, Noguchi K, Park YG, Lee YN, Cho-Chung YS (2000) Compensatory stabilization of RIIbeta protein, cell cycle deregulation, and growth arrest in colon and prostate carcinoma cells by antisense-directed down-regulation of protein kinase A RIalpha protein. Clin Cancer Res 6:3434–3441. Nesterova MV, Johnson N, Cheadle C, Bates SE, Mani S, Stratakis CA, Khan IU, Gupta RK, ChoChung YS (2006) Autoantibody cancer biomarker: Extracellular protein kinase A. Cancer Res 66:8971–8974. Nwachukwu JC, Li W, Pineda-Torra I, Huang HY, Ruoff R, Shapiro E, Taneja SS, Logan SK, Garabedian MJ (2007) Transcriptional regulation of the androgen receptor cofactor androgen receptor trapped clone-27. Mol Endocrinol 21:2864–2876. Okutani T, Nishi N, Kagawa Y, Takasuga H, Takenaka I, Usui T, Wada F (1991) Role of cyclic AMP and polypeptide growth regulators in growth inhibition by interferon in PC-3 cells. Prostate 18:73–80. Orstavik S, Reinton N, Frengen E, Langeland BT, Jahnsen T, Skalhegg BS (2001) Identification of novel splice variants of the human catalytic subunit cbeta of cAMP-dependent protein kinase. Eur J Biochem 268:5066–5073. Orstavik S, Funderud A, Hafte TT, Eikvar S, Jahnsen T, Skalhegg BS (2005) Identification and characterization of novel PKA holoenzymes in human T lymphocytes. FEBS J 272:1559–1567. Pascual G, Fong AL, Ogawa S, Gamliel A, Li AC, Perissi V, Rose DW, Willson TM, Rosenfeld MG, Glass CK (2005) A SUMOylation-dependent pathway mediates transrepression of inflammatory response genes by PPAR-gamma. Nature 437:759–763. Pastor-Soler N, Beaulieu V, Litvin TN, Da Silva N, Chen Y, Brown D, Buck J, Levin LR, Breton S (2003) Bicarbonate-regulated adenylyl cyclase (sAC) is a sensor that regulates pH-dependent V-ATPase recycling. J Biol Chem 278:49523–49529. Poukka H, Karvonen U, Janne OA, Palvimo JJ (2000) Covalent modification of the androgen receptor by small ubiquitin-like modifier 1 (SUMO-1). Proc Natl Acad Sci USA 97:14145– 14150. Poyet P, Gagne B, Labrie F (1986) Characteristics of the beta-adrenergic stimulation of adenylate cyclase activity in rat ventral prostate and its modulation by androgens. Prostate 9:237–245. Premont RT, Matsuoka I, Mattei MG, Pouille Y, Defer N, Hanoune J (1996) Identification and characterization of a widely expressed form of adenylyl cyclase. J Biol Chem 271:13900– 13907. Purvis K, Rui H, Gordeladze JO, Attramadal H (1986) Hormonal activation of the adenylyl cyclases of the rat and human prostate gland. Prostate 8:11–24. Quayle SN, Mawji NR, Wang J, Sadar MD (2007) Androgen receptor decoy molecules block the growth of prostate cancer. Proc Natl Acad Sci USA 104:1331–1336. Rana S, Bisht D, Chakraborti PK (1999) Synergistic activation of yeast-expressed rat androgen receptor by modulators of protein kinase-A. J Mol Biol 286:669–681. Razzaboni B, Terner C (1988) Cyclic adenosine 30 ,50 -monophosphate-phosphodiesterases in epididymis and prostate of castrate and of aged rats. Mech Ageing Dev 43:61–69.

500

M.D. Sadar

Redegeld FA, Smith P, Apasov S, Sitkovsky MV (1997) Phosphorylation of T-lymphocyte plasma membrane-associated proteins by ectoprotein kinases: Implications for a possible role for ectophosphorylation in T-cell effector functions. Biochim Biophys Acta 1328:151–165. Reid J, Kelly SM, Watt K, Price NC, McEwan IJ (2002) Conformational analysis of the androgen receptor amino-terminal domain involved in transactivation. Influence of structure-stabilizing solutes and protein-protein interactions. J Biol Chem 277:20079–20086. Reinton N, Orstavik S, Haugen TB, Jahnsen T, Tasken K, Skalhegg BS (2000) A novel isoform of human cyclic 30 ,50 -adenosine monophosphate-dependent protein kinase, c alpha-s, localizes to sperm midpiece. Biol Reprod 63:607–611. Rentero C, Monfort A, Puigdomenech P (2003) Identification and distribution of different mRNA variants produced by differential splicing in the human phosphodiesterase 9A gene. Biochem Biophys Res Commun 301:686–692. Riegman PH, Vlietstra RJ, van der Korput JA, Brinkmann AO, Trapman J (1991) The promoter of the prostate-specific antigen gene contains a functional androgen responsive element. Mol Endocrinol 5:1921–1930. Rigas AC, Ozanne DM, Neal DE, Robson CN (2003) The scaffolding protein RACK1 interacts with androgen receptor and promotes cross-talk through a protein kinase C signaling pathway. J Biol Chem 278:46087–46093. Rimler A, Culig Z, Levy-Rimler G, Lupowitz Z, Klocker H, Matzkin H, Bartsch G, Zisapel N (2001) Melatonin elicits nuclear exclusion of the human androgen receptor and attenuates its activity. Prostate 49:145–154. Rochette-Egly C (2003) Nuclear receptors: Integration of multiple signalling pathways through phosphorylation. Cell Signal 15:355–366. Rosenfeld MG, O’Malley BW (1970) Steroid hormones: Effects on adenyl cyclase activity and adenosine 30 ,50 -momophosphate in target tissues. Science 168:253–255. Rosner W, Hryb DJ, Khan MS, Nakhla AM, Romas NA (1991) Sex hormone-binding globulin: Anatomy and physiology of a new regulatory system. J Steroid Biochem Mol Biol 40:813–820. Rowan BG, Garrison N, Weigel NL, O’Malley BW (2000) 8-Bromo-cyclic AMP induces phosphorylation of two sites in SRC-1 that facilitate ligand-independent activation of the chicken progesterone receptor and are critical for functional cooperation between SRC-1 and CREB binding protein. Mol Cell Biol 20:8720–8730. Ryan CW, Stadler WM, Vogelzang NJ (2005) A phase I/II dose-escalation study of exisulind and docetaxel in patients with hormone-refractory prostate cancer. BJU Int 95:963–968. Sabbisetti VS, Chirugupati S, Thomas S, Vaidya KS, Reardon D, Chiriva-Internati M, Iczkowski KA, Shah GV (2005) Calcitonin increases invasiveness of prostate cancer cells: Role for cyclic AMP-dependent protein kinase A in calcitonin action. Int J Cancer 117:551–560. Sadar MD (1999) Androgen-independent induction of prostate-specific antigen gene expression via cross-talk between the androgen receptor and protein kinase A signal transduction pathways. J Biol Chem 274:7777–7783. Sadar MD, Gleave ME (2000) Ligand-independent activation of the androgen receptor by the differentiation agent butyrate in human prostate cancer cells. Cancer Res 60:5825–5831. Sadar MD, Hussain M, Bruchovsky N (1999) Prostate cancer: Molecular biology of early progression to androgen independence. Endocr Relat Cancer 6:487–502. Sainz RM, Mayo JC, Tan DX, Leon J, Manchester L, Reiter RJ (2005) Melatonin reduces prostate cancer cell growth leading to neuroendocrine differentiation via a receptor and PKA independent mechanism. Prostate 63:29–43. San Agustin JT, Leszyk JD, Nuwaysir LM, Witman GB (1998) The catalytic subunit of the cAMPdependent protein kinase of ovine sperm flagella has a unique amino-terminal sequence. J Biol Chem 273:24874–24883. Schreiber SN, Knutti D, Brogli K, Uhlmann T, Kralli A (2003) The transcriptional coactivator PGC-1 regulates the expression and activity of the orphan nuclear receptor estrogen-related receptor alpha (ERRalpha). J Biol Chem 278:9013–9018.

The Role of Cyclic AMP in Regulating the Androgen Receptor

501

Seeler JS, Dejean A (2003) Nuclear and unclear functions of SUMO. Nat Rev Mol Cell Biol 4:690–699. Shah GV, Rayford W, Noble MJ, Austenfeld M, Weigel J, Vamos S, Mebust WK (1994) Calcitonin stimulates growth of human prostate cancer cells through receptor-mediated increase in cyclic adenosine 30 ,50 -monophosphates and cytoplasmic Ca2+ transients. Endocrinology 134:596–602. Shalizi A, Gaudilliere B, Yuan Z, Stegmuller J, Shirogane T, Ge Q, Tan Y, Schulman B, Harper JW, Bonni A (2006) A calcium-regulated MEF2 sumoylation switch controls postsynaptic differentiation. Science 311:1012–1017. Shao D, Lazar MA (1999) Modulating nuclear receptor function: May the phos be with you. J Clin Invest 103:1617–1618. Shaywitz AJ, Greenberg ME (1999) CREB: A stimulus-induced transcription factor activated by a diverse array of extracellular signals. Annu Rev Biochem 68:821–861. Shima S, Kawashima Y, Hirai M, Asakura M (1980) Effects of androgens on isoproterenolsensitive adenylate cyclase system of the rat prostate. Mol Pharmacol 18:45–48. Singhal RL, Ling GM (1969) Metabolic control mechanisms in mammalian systems. IV. Androgenic induction of hexokinase and glucose-6-phosphate dehydrogenase in rat seminal vesicles. Can J Physiol Pharmacol 47:233–239. Singhal RL, Parulekar MR, Ling GM (1971) Streptozotocin-induced diabetes and regulation of hepatic glucose metabolism. Can J Physiol Pharmacol 49:1005–1007. Sinibaldi VJ, Elza-Brown K, Schmidt J et al. (2006) Phase II evaluation of docetaxel plus exisulind in patients with androgen independent prostate carcinoma. Am J Clin Oncol 29:395–398. Siu YT, Jin DY (2007) CREB – a real culprit in oncogenesis. FEBS J 274:3224–3232. Skalhegg BS, Tasken K (2000) Specificity in the cAMP/PKA signaling pathway. Differential expression, regulation, and subcellular localization of subunits of PKA. Front Biosci 5:D678– D693. Smigel MD (1986) Purification of the catalyst of adenylate cyclase. J Biol Chem 261:1976–1982. Smith FD, Langeberg LK, Scott JD (2006) The where’s and when’s of kinase anchoring. Trends Biochem Sci 31:316–323. Solomon SS, Brush JS, Kitabchi AE (1970) Divergent biological effects of adenosine and dibutyryl adenosine 30 ,50 -monophosphate on the isolated fat cell. Science 169:387–388. Songyang Z, Blechner S, Hoagland N, Hoekstra MF, Piwnica-Worms H, Cantley LC (1994) Use of an oriented peptide library to determine the optimal substrates of protein kinases. Curr Biol 4:973–982. Stubbs AP, Lalani EN, Stamp GW, Hurst H, Abel P, Waxman J (1996) Second messenger upregulation of androgen receptor gene transcription is absent in androgen insensitive human prostatic carcinoma cell lines, PC-3 and DU-145. FEBS Lett 383:237–240. Sunahara RK, Taussig R (2002) Isoforms of mammalian adenylyl cyclase: Multiplicities of signaling. Mol Interv 2:168–184. Sutherland DJ, Singhal RL (1974a) Alterations in adenylate cyclase activity of the rat prostate gland. Biochim Biophys Acta 343:238–249. Sutherland DJ, Singhal RL (1974b) Stimulation of prostatic adenyl cyclase by dihydrotestosterone. Horm Metab Res 6:89. Sutherland EW, Rall TW (1958) Fractionation and characterization of a cyclic adenine ribonucleotide formed by tissue particles. J Biol Chem 232:1077–1091. Taneja SS, Ha S, Swenson NK, Torra IP, Rome S, Walden PD, Huang HY, Shapiro E, Garabedian MJ, Logan SK (2004) ART-27, an androgen receptor coactivator regulated in prostate development and cancer. J Biol Chem 279:13944–13952. Tang WJ, Krupinski J, Gilman AG (1991) Expression and characterization of calmodulinactivated (type I) adenylylcyclase. J Biol Chem 266:8595–8603. Taussig R, Gilman AG (1995) Mammalian membrane-bound adenylyl cyclases. J Biol Chem 270:1–4.

502

M.D. Sadar

Theurkauf WE, Vallee RB (1982) Molecular characterization of the cAMP-dependent protein kinase bound to microtubule-associated protein 2. J Biol Chem 257:3284–3290. Thomas JA, Singhal RL (1973) Testosterone-stimulation of adenyl cyclase and cyclic 30 ,50 adnosine monophosphate-3 H formation in rat seminal vesicles. Biochem Pharmacol 22:507–511. Thomas M, Dadgar N, Aphale A, Harrell JM, Kunkel R, Pratt WB, Lieberman AP (2004) Androgen receptor acetylation site mutations cause trafficking defects, misfolding, and aggregation similar to expanded glutamine tracts. J Biol Chem 279:8389–8395. Tiefenbach J, Novac N, Ducasse M, Eck M, Melchior F, Heinzel T (2006) SUMOylation of the corepressor N-CoR modulates its capacity to repress transcription. Mol Biol Cell 17:1643–1651. Tilley WD, Marcelli M, McPhaul MJ (1990) Expression of the human androgen receptor gene utilizes a common promoter in diverse human tissues and cell lines. J Biol Chem 265:13776–13781. Tortora G, Damiano V, Bianco C, Baldassarre G, Bianco AR, Lanfrancone L, Pelicci PG, Ciardiello F (1997) The RIalpha subunit of protein kinase A (PKA) binds to Grb2 and allows PKA interaction with the activated EGF-receptor. Oncogene 14:923–928. Tremblay A, Tremblay GB, Labrie F, Giguere V (1999) Ligand-independent recruitment of SRC-1 to estrogen receptor beta through phosphorylation of activation function AF-1. Mol Cell 3:513–519. Tsang BK, Singhal RL (1976) Androgenic effects on protein kinases and cyclic AMP-binding protein in the ventral prostate. Res Commun Chem Pathol Pharmacol 13:697–712. Uckert S, Hedlund P, Andersson KE, Truss MC, Jonas U, Stief CG (2006) Update on phosphodiesterase (PDE) isoenzymes as pharmacologic targets in urology: Present and future. Eur Urol 50:1194–207; discussion 1207. Ueda T, Bruchovsky N, Sadar MD (2002a) Activation of the androgen receptor N-terminal domain by interleukin-6 via MAPK and STAT3 signal transduction pathways. J Biol Chem 277:7076–7085. Ueda T, Mawji NR, Bruchovsky N, Sadar MD (2002b) Ligand-independent activation of the androgen receptor by interleukin-6 and the role of steroid receptor coactivator-1 in prostate cancer cells. J Biol Chem 277:38087–38094. Unni E, Sun S, Nan B, McPhaul MJ, Cheskis B, Mancini MA, Marcelli M (2004) Changes in androgen receptor nongenotropic signaling correlate with transition of LNCaP cells to androgen independence. Cancer Res 64:7156–7168. van der Kwast TH, Schalken J, Ruizeveld de Winter JA, van Vroonhoven CC, Mulder E, Boersma W, Trapman J (1991) Androgen receptors in endocrine-therapy-resistant human prostate cancer. Int J Cancer 48:189–193. Verger A, Perdomo J, Crossley M (2003) Modification with SUMO. A role in transcriptional regulation. EMBO Rep 4:137–142. Visakorpi T, Hyytinen E, Koivisto P, Tanner M, Keinanen R, Palmberg C, Palotie A, Tammela T, Isola J, Kallioniemi OP (1995) In vivo amplification of the androgen receptor gene and progression of human prostate cancer. Nat Genet 9:401–406. Vossler MR, Yao H, York RD, Pan MG, Rim CS, Stork PJ (1997) cAMP activates MAP kinase and elk-1 through a B-raf- and Rap1-dependent pathway. Cell 89:73–82. Wang G, Jones SJ, Marra MA, Sadar MD (2006a) Identification of genes targeted by the androgen and PKA signaling pathways in prostate cancer cells. Oncogene 25:7311–7323. Wang H, Hang J, Shi Z, Li M, Yu D, Kandimalla ER, Agrawal S, Zhang R (2002) Antisense oligonucleotide targeted to RIalpha subunit of cAMP-dependent protein kinase (GEM231) enhances therapeutic effectiveness of cancer chemotherapeutic agent irinotecan in nude mice bearing human cancer xenografts: In vivo synergistic activity, pharmacokinetics and host toxicity. Int J Oncol 21:73–80. Wang L, Zuercher WJ, Consler TG, Lambert MH, Miller AB, Orband-Miller LA, McKee DD, Willson TM, Nolte RT (2006b) X-ray crystal structures of the estrogen-related receptor-

The Role of Cyclic AMP in Regulating the Androgen Receptor

503

gamma ligand binding domain in three functional states reveal the molecular basis of small molecule regulation. J Biol Chem 281:37773–37781. Weigel NL, Moore NL (2007) Steroid receptor phosphorylation: A key modulator of multiple receptor functions. Mol Endocrinol 21:2311–2319. Wheeler MA, Ayyagari RR, Wheeler GL, Weiss RM (2005) Regulation of cyclic nucleotides in the urinary tract. J Smooth Muscle Res 41:1–21. White PC, Shore AM, Clement M, McLaren J, Soeiro I, Lam EW, Brennan P (2006) Regulation of cyclin D2 and the cyclin D2 promoter by protein kinase A and CREB in lymphocytes. Oncogene 25:2170–2180. Wu D, Zhau HE, Huang WC, Iqbal S, Habib FK, Sartor O, Cvitanovic L, Marshall FF, Xu Z, Chung LW (2007) cAMP-responsive element-binding protein regulates vascular endothelial growth factor expression: Implication in human prostate cancer bone metastasis. Oncogene 26:5070–5077. Wu H, Sun L, Zhang Y et al. (2006) Coordinated regulation of AIB1 transcriptional activity by sumoylation and phosphorylation. J Biol Chem 281:21848–21856. Xie W, Hong H, Yang NN, Lin RJ, Simon CM, Stallcup MR, Evans RM (1999) Constitutive activation of transcription and binding of coactivator by estrogen-related receptors 1 and 2. Mol Endocrinol 13:2151–2162. Xie Y, Wolff DW, Lin MF, Tu Y (2007) Vasoactive intestinal peptide transactivates the androgen receptor through a protein kinase A-dependent extracellular signal-regulated kinase pathway in prostate cancer LNCaP cells. Mol Pharmacol 72:73–85. Yamashita D, Yamaguchi T, Shimizu M, Nakata N, Hirose F, Osumi T (2004) The transactivating function of peroxisome proliferator-activated receptor gamma is negatively regulated by SUMO conjugation in the amino-terminal domain. Genes Cells 9:1017–1029. Yan C, Zhao AZ, Bentley JK, Beavo JA (1996) The calmodulin-dependent phosphodiesterase gene PDE1C encodes several functionally different splice variants in a tissue-specific manner. J Biol Chem 271:25699–25706. Yan SZ, Huang ZH, Andrews RK, Tang WJ (1998) Conversion of forskolin-insensitive to forskolin-sensitive (mouse-type IX) adenylyl cyclase. Mol Pharmacol 53:182–187. Yang CS, Vitto MJ, Busby SA et al. (2005) Simian virus 40 small t antigen mediates conformation-dependent transfer of protein phosphatase 2A onto the androgen receptor. Mol Cell Biol 25:1298–1308. Yang CS, Xin HW, Kelley JB, Spencer A, Brautigan DL, Paschal BM (2007) Ligand binding to the androgen receptor induces conformational changes that regulate phosphatase interactions. Mol Cell Biol 27:3390–3404. Yang WL, Iacono L, Tang WM, Chin KV (1998) Novel function of the regulatory subunit of protein kinase A: Regulation of cytochrome c oxidase activity and cytochrome c release. Biochemistry 37:14175–14180. Yeager RE, Heideman W, Rosenberg GB, Storm DR (1985) Purification of the calmodulinsensitive adenylate cyclase from bovine cerebral cortex. Biochemistry 24:3776–3783. Young CY, Montgomery BT, Andrews PE, Qui SD, Bilhartz DL, Tindall DJ (1991) Hormonal regulation of prostate-specific antigen messenger RNA in human prostatic adenocarcinoma cell line LNCaP. Cancer Res 51:3748–3752. Yuasa K, Kotera J, Fujishige K, Michibata H, Sasaki T, Omori K (2000) Isolation and characterization of two novel phosphodiesterase PDE11A variants showing unique structure and tissuespecific expression. J Biol Chem 275:31469–31479. Zegarra-Moro OL, Schmidt LJ, Huang H, Tindall DJ (2002) Disruption of androgen receptor function inhibits proliferation of androgen-refractory prostate cancer cells. Cancer Res 62:1008–1013. Zhou ZX, Sar M, Simental JA, Lane MV, Wilson EM (1994) A ligand-dependent bipartite nuclear targeting signal in the human androgen receptor. Requirement for the DNA-binding domain and modulation by NH2-terminal and carboxyl-terminal sequences. J Biol Chem 269:13115– 13123.

Cellular and Molecular Signatures of Androgen Ablation of Prostate Cancer Clifford G. Tepper and Hsing-Jien Kung

Abstract Androgen ablative therapy is the cornerstone of treatment for metastatic prostate cancer and some cases of high-risk, localized disease. This is only palliative, however, since castration-recurrent disease typically occurs. Accordingly, intense research efforts focus upon achieving a better understanding of the cellular and molecular responses to androgen ablation and their roles in facilitating the transition to recurrence. In this chapter, we discuss how diminished androgen receptor (AR) signaling represents the pivotal mediator of a tightly coordinated signaling response that manifests in loss of AR expression, growth arrest, neuroendocrine differentiation (NED), and survival. Classic androgen-deprivation therapy has evolved to include approaches aimed at achieving complete suppression of AR signaling through the utilization of AR antagonists and inhibitors of androgen metabolism. This mediates repression of AR function at multiple levels by abrogating its transcriptional activity, increasing its turnover, and reducing translation of its transcript. Consequently, this leads to cellular trans-differentiation from an epithelial to neuroendocrine phenotype. NED cells figure critically in disease progression by virtue of secreting growth-promoting neurotrophic factors, possessing features of cancer stem cells, and surviving in the absence of androgen. Along these lines, phosphatidylinositol-3 kinase (PI3K)-Akt signaling is hyperactivated in response to androgen ablation and functions as a dominant antiapoptotic pathway, especially in the context of PTEN-mutant cancers such as LNCaP. Interestingly, mammalian target of rapamycin (mTOR) is implicated as a critical sensor of androgen signaling and an integrator of androgen ablation-induced AR down-regulation, PI3K-Akt hyperactivation, and NED. The marked effects of androgen ablation are due in large part to widespread changes in AR-regulated gene expression which produce a diagnostic androgen withdrawal expression signature. Although not well defined, we discuss potential mechanisms and gene product interactions that might explain how these translate into the molecular and biological features of androgen ablation.

H-J. Kung(*) UC Davis Cancer Center, University of California, Davis School of Medicine, UCDMC Research III, Room 2400B, 4645 2nd Avenue, Sacramento, CA 95817, USA, E-mail: [email protected]

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4_22, # Springer Science + Business Media, LLC 2009

507

508

C.G. Tepper, H.-J. Kung

1 Introduction Androgen ablation is the cornerstone in the treatment of metastatic prostate cancer (CaP) and some cases of high-risk, localized disease (Ryan and Small 2005). Although this effectively mediates regression of the cancer, it is only palliative since recurrence typically occurs as castration-recurrent disease that is also resistant to conventional chemotherapeutic agents. Sadly this is the principal cause of mortality due to prostate cancer. The transition period spanning from the commencement of androgen deprivation through the development of castration-recurrent CaP therefore represents a critical time for the disease with respect to its persistence. Moreover, it also provides a unique therapeutic window of opportunity for the implementation of novel interventions designed to delay the onset of recurrence or to completely eradicate the cancer. Androgen is a dominant and pleiotropic mediator of various aspects of CaP biology and progression. Conversely, androgen withdrawal also evokes significant cellular and molecular changes and can be viewed as a very active process. In this chapter, we discuss the cellular and molecular signatures engendered by androgen ablation and their relevance to the development of castration-recurrent CaP. It would be impossible to discuss androgen withdrawal of CaP without considering the action of androgen. The normal physiological role of androgen is to regulate the development, function, and homeostasis of the prostate gland by its impact upon cellular processes including proliferation, differentiation, morphogenesis, and survival. The pivotal mediator of these effects is the androgen receptor (AR), which accomplishes this primarily by its ability to engender a distinctive transcriptional program by effecting widespread changes in gene expression. The ultimate outcome of AR signaling is influenced by diverse intracellular signaling pathways initiated by the AR itself as well as by exogenous stimuli derived locally via direct interactions between neighboring epithelial and stromal/mesenchymal cells or from circulating factors. These are seamlessly coordinated and delicately balanced in order to maintain homeostatic integrity. In CaP, this is compromised by various molecular aberrations that tilt the biological balance in favor of (1) uncontrolled proliferation, (2) maintenance of an undifferentiated state, (3) migration, invasion, metastasis, (4) and survival. These will be discussed in more detail in the appropriate sections later.

2 Clinical Androgen-Deprivation Therapy Androgen ablative therapy is the standard of care for advanced prostate cancer and exploits the marked dependence upon testicular androgens exhibited by most CaPs. As our understanding of the molecular and biochemical mechanisms of androgen action has increased, it has evolved from basic androgen-deprivation therapy (ADT) into the incorporation of approaches to achieve a more complete blockade

Cellular and Molecular Signatures of Androgen Ablation of Prostate Cancer

509

of the androgen–androgen receptor signaling axis. In 1941 Huggins and Hodges demonstrated that androgenic hormones positively influence CaP and that elimination of these androgens could be an effective therapy for CaP (Huggins and Hodges 1941; Huggins et al. 1941). In their clinical studies, androgen withdrawal monotherapy was achieved by surgical castration (i.e., bilateral orchiectomy) or injection of high-dose estrogens (i.e., diethylstilbestrol, estradiol benzoate). Both approaches decreased cancer growth as indicated by a reduction in the serum levels of prostatic acid phosphatase (PAP), which served as a tumor marker at the time (Gutman and Gutman 1938). However, the use of these endocrine modalities was limited due to lethal cardiovascular and cerebrovascular complications encountered with estrogen treatment and the physical effects and psychological impact associated with surgical castration (Peeling 1989; Robinson and Thomas 1971). Medical, or chemical, castration, which has replaced surgery and high-dose estrogens, is now widely used for the treatment of metastatic disease and increasingly for high-risk localized disease. It is achieved through the administration of luteinizing hormone-releasing hormone (LHRH) agonists or analogs (e.g., leuprelide, goserelin) (Tolis et al. 1982; Labrie 2004). These drugs suppress the release of luteinizing hormone (LH) from the anterior lobe of the pituitary gland, which in turn suppresses production of androgen by the testes. LHRH is a naturally occurring hormone released in a pulsatile manner from the hypothalamus when it senses low levels of testosterone and 5a-dihydrotestosterone (DHT). In turn, LHRH binds to its cognate receptor on the surface of pituitary gonadotropic cells, which then secrete LH to stimulate testosterone production in the testes. In contrast, LHRH agonists disrupt the hypothalamic–pituitary–gonadal axis due to persistent (rather than pulsatile) stimulation leading to desensitization of the pituitary by subsequent down-regulation of its LHRH receptors. Additionally, a reduction in the biological activity of LH has been implicated as a mechanism of LHRH agonist action (St-Arnaud et al. 1986). These agents typically suppress androgen levels to 95% of the physiological values within 3 weeks thereby achieving castrate levels defined as 50 ng/dL (1.7 nM) (Bubley et al. 1999; Leuprolide Study Group 1984; Garnick 1986; Limonta et al. 2001). Combined androgen blockade (CAB) is utilized to achieve complete suppression of androgen signaling. For this, antiandrogens are used to directly block the activity of the AR. These antagonists are divided into two structural classes, steroidal and nonsteroidal. The synthetic progestins cyproterone acetate (Androcur) and megestrol acetate are members of the former class and also have weak progestational and glucocorticoid activity. In contrast, nonsteroidal, ‘‘pure antiandrogens’’ bind exclusively to the AR and are devoid of antigonadotropic, antiestrogenic, and progestational effects. These are more commonly used in current practice and include the anilides flutamide (Eulexin), nilutamide (Anandron, Nilandron), and bicalutamide (Casodex). Advantages of the nonsteroidals include higher specificity, selectivity, and better pharmacokinetic properties (Neri et al. 1979; Cockshott et al. 1990; Teutsch et al. 1994). Although the results in randomized trials comparing combination therapy to castration alone are variable, better overall survival is associated with the use of nonsteroidal vs. steroidal antiandrogens, and the use of bicalutamide

510

C.G. Tepper, H.-J. Kung

in particular might reduce the hazard ratio for mortality (Klotz and Schellhammer 2005). Antiandrogens have also been vital for the prevention of ‘‘hormone flare,’’ which is an adverse effect accompanying LHRH agonist therapy (Kuhn et al. 1989). This occurs during the first 7–10 days of treatment as a result of a transient increase in testosterone levels consequent to the initial surge in LH levels. This can be very painful and might even stimulate the growth of bone metastases. This can be alleviated by treatment with antiandrogens prior to medical castration. The roles of adrenal androgens and intracrine androgen metabolism represent critical targets of hormonal therapies since both serve as additional sources of AR stimulation. Of note, several studies have demonstrated that even after long-term ADT, intraprostatic DHT levels are present at levels sufficient to activate the AR in vitro and sustain expression of a subset of AR-regulated genes in vivo (Titus et al. 2005; Nishiyama et al. 2004; Mostaghel et al. 2007). Although measurements were made in the context of different study designs and using different techniques, comparable results were obtained in that ADT resulted in a 25% decrease in tissue DHT (4.65 pmol/g tissue) (Nishiyama et al. 2004), and levels in recurrent CaP specimens declined by 91% (i.e., to 1.25 pmol/g tissue) (Titus et al. 2005). The adrenal androgens include dehydroepiandrosterone (DHEA), DHEA sulfate, androstenediol, and androstenedione. These can be converted to testosterone through the action of the appropriate 17b- and 3b-hydroxysteroid dehydrogenases (17-HSDs). DHT is then generated from testosterone by 5a-reductase (5aR) type 1 and 2 isoenzymes (Prout et al. 1976). Although the total amount of DHT converted from adrenal androgen ranges only from 3 to 7% (vs. 50–70% for testosterone), it potentially accounts for up to one-sixth of intraprostatic DHT due to the higher concentrations of adrenal androgen substrates (i.e., fourfold to sixfold) compared to testosterone (Geller 1985). As a result, the 5aR inhibitors, specifically finasteride (5aR2; Proscar) and dutasteride (dual 5aR1/2; Avodart, Avidart), have received much attention for CaP therapy. Recent studies have revealed a positive correlation between their expression and CaP progression since high-grade Gleason pattern 4/5 tumors exhibited higher expression of 5aR types 1 and 2 in comparison to lowgrade cancers (Thomas et al. 2008a, b). In addition, benign tissue adjacent to highgrade lesions had elevated 5aR1. These findings might account for the decreased effectiveness of the 5aR1 inhibitor (finasteride) against high-grade CaP observed in the Prostate Cancer Prevention Trial (Thomas et al. 2008a) and led to the proposal that dutasteride might be more effective in delaying the development and progression of CaP (Thomas et al. 2008b). Another approach is aimed at suppressing the synthesis of adrenal androgens. To this end, a prime target has been cytochrome P450c17, a bifunctional enzyme that catalyzes 17a-hydroxylation and C17,20-lyase reactions (Zuber et al. 1986; Nakajin et al. 1981; Chung et al. 1987; Kitamura et al. 1991) and converts 17a-hydroxypregnenolone and 17a-hydroxyprogesterone to the C-19 androgens, DHEA and androstenedione, respectively. Ketoconazole, originally used as a broad-spectrum antifungal agent, has been the only inhibitor to be used for the treatment of advanced prostate cancer (Trachtenberg et al. 1983; Trachtenberg and Pont 1984), but has been withdrawn due to toxicity resulting from off-target effects on

Cellular and Molecular Signatures of Androgen Ablation of Prostate Cancer

511

other P450 enzymes (De Coster et al. 1996). However, more specific and higheraffinity steroidal and nonsteroidal P450c17 antagonists are being developed (Matsunaga et al. 2004; Grigoryev et al. 1999). Importantly, since P450c17 is expressed in both the adrenal and the testis, the use of these inhibitors has the potential to significantly contribute to achieving absolute androgen deprivation.

3 Experimental Androgen Ablation In Vivo Castration-induced involution, or regression, of the rat prostate gland is a striking example of the potent proapoptotic effects elicited by androgen withdrawal and is one of the classic examples of the physiological importance of apoptosis (Kerr and Searle 1973). Importantly, it lends strong support to the basis for endocrine therapy for CaP by inhibiting growth and inducing cell death. These models have also underscored the concept that castration leads to degeneration of the entire gland. Indeed, not only epithelial cells, but also stromal and vascular components regress. In fact, the intimate association between these components, which is required for proper morphogenesis and homeostasis, is demonstrated by the impact androgen withdrawal has on all of these rather than just one isolated component. At the same time, the basis for this has become better understood through better characterization of the cellular and molecular events occurring in each component consequent to androgen withdrawal. As a result, the rat model of castration-induced apoptosis has been valuable to our understanding of the critical roles played by androgen in the normal prostate gland and highlighting the differences in the androgen sensitivity of CaP. The prostate regresses to 17 and 5% of its intact weight by 14 and 30 days postcastration, respectively, and as much as 20% of the glandular cells are depleted per day between days 2 and 5 (Berges et al. 1993). Anatomically, the rodent prostate comprises several lobes: the ventral prostate (VP), lateral prostate (LP), and dorsal prostate (DP) (Lee 1987). Apoptosis is most prominent in the luminal and stromal cells of the VP following androgen withdrawal, but is negligible in the LP and DP (Banerjee et al. 1995, 2000). This has been confirmed by morphological criteria and multiple molecular approaches, including the detection of DNA degradation by agarose gel electrophoresis and in situ with terminal deoxynucleotidyl (TdT)mediated dUTP nick end labeling (TUNEL). However, the rat VP does not have a clear functional homolog in the human prostate gland (Price 1963). The human equivalent of the rodent dosolateral prostate is the transitional zone (TZ), the site of occurrence for 20% of CaPs, with 80% arising in the peripheral zone (PZ) (Grignon and Sakr 1994). This is further confirmed by the fact that the expression profile for the dorsolateral lobe of the mouse prostate is similar to that of the human TZ (Berquin et al. 2005). The biological differences between these zones are also reflected in the features of the CaPs arising from them in that PZ-derived CaP are generally of higher Gleason grade (6.7 vs. 5.6, respectively) and proliferative index (5.0% and 1.6%, respectively). Interestingly, in Brown Norway rats the prostates

512

C.G. Tepper, H.-J. Kung

from aging animals (24-month old) are more resistant to apoptosis than those in younger animals (4-month old) in that apoptosis is decreased by 50% (Banerjee et al. 2000, 2002). Therefore, lobe-specific and age-related differences in the prostate gland might help explain the origins of CaP. Androgen withdrawal-induced apoptosis of prostate luminal epithelial cells can be attributed to a variety of molecular changes that either increase the susceptibility of cells to apoptosis and/or activate apoptotic signaling pathways. The Bcl-2 family of apoptosis regulators has been intensely studied in this regard. At least two studies demonstrated that the ratio of Bax/Bcl-2 increased transiently and coincidently with the peak of apoptosis occurring 2 and 3 days following castration (Banerjee et al. 2002; Perlman et al. 1999). Subsequently, Bcl-2 expression continued to increase, while Bax decreased. Bcl-2 was restricted to the epithelial compartment. In contrast to the VP, the DP and LP exhibit a prosurvival Bcl-2/Bax signature in that Bcl2 levels are significantly higher than in the VP, Bax expression is lower, and both are unresponsive to androgenic regulation (Banerjee et al. 2002). Furthermore, prostatic Bax expression is reduced in older rats. Additional apoptosis-sensitizing events include up-regulation of the expression of PTEN (phosphatase and tensin homolog deleted on chromosome 10), and IGF-binding proteins-3/5 (Desai et al. 2004). Early molecular studies in the rat model identified a set of prostate apoptosis response (Par) genes, which are up-regulated in the rat VP following castration (Sells et al. 1994). These included several previously characterized genes with diverse functions including dual specificity phosphatase 1(DUSP1; par-1), heparin-binding EGF-like growth factor (HB-EGF; par-2;), and CYR61 (par-3; cysteine-rich, angiogenic inducer, 61/IGFBP10), as well as a novel gene Par-4, which is a proapoptotic gene encoding a leucine zipper protein that binds to the zinc finger of Wilm’s tumor protein WT1 (Sells et al. 1997). Androgen withdrawal also leads to the activation of the transforming growth factor-b (TGF-b)-Smad pathway (Brodin et al. 1999), where elevated TGF-b expression and levels of Smad2 phosphorylation are evident following castration. Activation of Fas/CD95 death receptor signaling has also been implicated as a mechanism contributing, but not required for, castration-induced apoptosis (Suzuki et al. 1996; de la Taille et al. 1999). In the mouse prostate, castration leads to increased Fas/CD95 and its activation as verified by the assembly of the death-inducing signaling complex containing the adaptor proteins Fas/APO-1-associated death domain protein (FADD) and the kinase receptor-interacting protein (RIP). Furthermore, apoptosis induced by the extrinsic pathway would be further favored since the FADD-like interleukin-1b-converting enzyme (FLICE)-inhibitory protein (FLIP) is down-regulated in rat prostate after castration (Cornforth et al. 2008). It is also worth noting that although extensive apoptosis is evident following castration, a subpopulation of cells does survive, and readdition of testosterone can restore the tissue. From a molecular standpoint, there is substantial evidence to support the concept that antiapoptotic signaling is coordinated coincidently with, or following, the major wave of apoptosis. This is illustrated by the elevation in Bcl2 expression in later stages of castration and the up-regulation of DUSP1, a prosurvival phosphatase, which can counteract p38 MAPK signaling. One of the

Cellular and Molecular Signatures of Androgen Ablation of Prostate Cancer

513

most prominent castration-induced genes is clusterin (CLU)/testosterone-repressed prostate message-2 (TRPM-2), which has been demonstrated to be a critical mediator of resistance to androgen ablation and whose expression peaks during 2–5 days (Brandstrom et al. 1994). Interestingly, CCAAT-enhancer binding protein-d represents another castration-induced gene (Desai et al. 2004), which may promote growth arrest (vs. apoptosis) and the maintenance of surviving prostate epithelial cells, based upon its function in promoting stem cell renewal by slowing cell cycle progression via p27(Kip1) and p57(Kip2) (Barbaro et al. 2007). Interestingly, protein biosynthesis in the epithelial and stromal compartments of the rat ventral prostate actually increases fivefold and sevenfold, respectively, one week following castration (Zhao et al. 1993). The effects of androgen ablation upon the stroma are as profound as that of the epithelium. In the normal prostate, this has significant secondary effects upon the epithelial cells since this is critical to maintaining homeostasis through paracrine stimulation and vascular support. Accordingly, the responses of the stroma influence progression of PIN and CaP arising in the proximity. Like the luminal epithelial cells, stromal cells express AR and are very androgen sensitive. Using a tissue recombination model, the critical function for stromal AR was demonstrated by the fact that development of glandular units was dependent upon the stromal component having functional AR (Kurita et al. 2001). Substitution with mesenchyme derived from tfm mice abrogated formation. In contrast, epithelial AR function was dispensable in this model. However, carcinoma-associated stroma, such as in Dunning tumors, is apparently androgen insensitive since AR was not down-regulated (Johansson et al. 2007). It is therefore believed that adjuvant targeting of stromal factors can enhance the efficacy of androgen ablation. Vascular regression within the prostate in response to castration is marked following androgen withdrawal and contributes to degeneration of the epithelium. This is characterized by endothelial cell apoptosis, widening of the vasculature, and the presence of immune infiltrates. The basis for this effect is multifactorial since androgens positively regulate VEGF (Joseph et al. 1997). At the same time, the ability of VEGF and neuropilin-1 (NRP1/VEGF165R CaP) to promote osteoblast activity (Kitagawa et al. 2005) coupled with the stromal expression of many osteogenic proteins implicates the potential for normal stroma to increase the propensity of CaP for bone metastasis. Pigment epitheliumderived factor (PEDF) is a 50-kDa glycoprotein and represents one of the more dominant mediators of androgen withdrawal’s effect upon the prostatic vasculature since it is upregulated by androgen ablation both experimentally and clinically and inhibits angiogenesis in part by inducing apoptosis of endothelial cells (Doll et al. 2003). Importantly, its expression is lost in both low- and high-grade CaPs. The importance of neovascularization is underscored by the fact that it precedes testosteronestimulated regrowth of the prostate (Franck-Lissbrant et al. 1998). Since the prostates of PEDF/ mice exhibit increased vascularity and hyperplasia, loss of PEDF might represent a step in the molecular etiology of CaP. Interestingly, it is expressed in both the stroma and epithelium. PEDF might also function as a modulator of the neuroendocrine phenotype since it is also known as a neurotrophin

514

C.G. Tepper, H.-J. Kung

for retinoblastoma cells and sustaining neuronal survival. The fact that aggressive CaP cells can exhibit vasculogenic mimicry illustrates the plasticity of CaP cells (Sharma et al. 2002). Taken together, prostatic epithelial cells not only regress due to the absence of androgen, but also as a result of hypoxia due to damaged blood supply and resulting inflammation. In contrast to the widespread apoptotic response that occurs in normal prostate, this is not generally observed in CaPs in vivo, including human CWR22 (Agus et al. 1999) and the Dunning rat model (Westin et al. 1993). On the other hand, it has been demonstrated in the human PC-82 and PC-EW xenograft models (van Weerden et al. 1993). In the case of the highly differentiated Dunning R3327-PAP tumor model, castration leads to growth inhibition; recurrent, androgen-insensitive tumors are thought to arise due to a reduction of cells being depleted by apoptosis and are also characterized as being dedifferentiated (Landstrom et al. 1994). The latter feature is particularly relevant with respect to one major function for androgen in driving prostatic epithelial differentiation. In this light, a cancer stem cell hypothesis might be a reasonable explanation for the surviving fraction in that these cells are reminiscent of a transit cell possessing a phenotype intermediate between that of a basal and mature luminal cell, while being capable of differentiating along a NE lineage. Despite the arsenal of existing therapeutic approaches to completely ablating androgen-AR signaling, castration- and antiandrogen-recurrent tumors typically arise by acquiring the ability to circumvent one or more growth-inhibitory and/or proapoptotic mechanisms. This is due in large part to reinstatement of AR signaling derived from mechanisms resulting in restoration of AR expression, acquisition of hypersensitivity to androgen, and signal transduction pathways that collaborate with the former or serve as alternative modes of AR activation. Although androgen withdrawal halts the growth of the CaP, it also triggers cellular and molecular events that promote its survival and eventual androgen-independent outgrowth locally or at distant sites. A dominant feature of androgen withdrawal that facilitates the transition to androgen independence is the transdifferentiation of the epithelial cells to a neuroendocrine phenotype. In the following sections, we discuss the cellular and molecular mechanisms triggered by androgen ablation in experimental model systems.

4 Androgen Ablation-Induced Neuroendocrine Differentiation As described in the previous sections, both androgen ablation therapy and experimental androgen ablation in animals induce rapid growth arrest and apoptosis of affected prostate cells. A striking, commonly observed feature during this process is the emergence of neuroendocrine or neuroendocrine-like cells, intermingled with prostate epithelial and stromal cells. These cells are post-mitotic and express neuronal markers such as chromogranin A, synaptophysin, neuron-specific enolase, and b-tubulin III. They also release neurotropic hormones and peptides such as

Cellular and Molecular Signatures of Androgen Ablation of Prostate Cancer

515

serotonin, neurotensin, and gastrin-releasing peptides (For reviews see (Yuan et al. 2007; Sciarra et al. 2003; Vashchenko and Abrahamsson 2005; di Sant’Agnese 1992)). These cells are mostly AR-negative. The neuroendocrine cells induced by androgen ablation however differ from neuroendocrine cells present in normal prostate in that they retain epithelial luminal cell markers (e.g., K18) (Vashchenko and Abrahamsson 2005; van Bokhoven et al. 2003), rather than basal cell markers (e.g., p63) (Wojno and Epstein 1995; Signoretti et al. 2000), suggesting that their origin might be epithelial luminal tumor cells, which undergo transdifferentiation triggered by decreased amounts of androgen receptor. The relevance of neuroendocrine differentiation to the development of hormone-refractory prostate cancer has been a subject of considerable interest and discussion, which will be reviewed in a later section. The evidence that androgen ablation induces neuroendocrine differentiation came from both clinical observations, animal xenograft models as well as in vitro cell line studies. It has been reported that the majority of specimens (71%) from patients who received long-term androgen ablation therapy displayed neuroendocrine differentiation phenotypes, compared to only 32% for patients without therapy (Hirano et al. 2004; Ito et al. 2001). In vivo human xenograft models implanted into intact or castrated animals clearly showed that the latter displayed increased numbers of neuroendocrine-like cells, defined by neuronal markers. These models include PC-295 (Jongsma et al. 1999), PC-310 (Jongsma et al. 2000, 2002), LNCaP (Burchardt et al. 1999), and CWR22 (Huss et al. 2004). Perhaps the most direct evidence that androgen withdrawal induces neuroendocrine differentiation is supplied by in vitro cell line studies. Abundant literature exists for LNCaP and its derivatives (Burchardt et al. 1999; Zelivianski et al. 2001; Zhang et al. 2003; Yang et al. 2005). This is not surprising as LNCaP is one of the very few androgen-dependent cells that express functional androgen receptor, which controls its growth, survival, and differentiation state. Indeed, depletion of androgen receptor by introducing small interfering RNA (siRNA) targeting androgen receptor leads to neuroendocrine phenotypes (Wright et al. 2003; Frigo and McDonnell 2008). In addition, the neuroendocrine phenotype of LNCaP cells induced by steroid-reduced conditions can be reversed to the original epithelial-like state by the addition of DHT, but exacerbated by the antiandrogen Casodex (Yuan et al. 2006). Neuroendocrine differentiation of LNCaP can also be induced by a number of other chemicals and biologicals such as interleukin-6 (IL-6) (Qiu et al. 1998), forskolin (Bang et al. 1994; Cox et al. 2000), insulinlike growth factor-1 (IGF-1) (Fan et al. 2007), HB-EGF (Kim et al. 2002), genistein (Pinski et al. 2006), and IL-1b (Chiao et al. 1999). Perhaps, noncoincidentally, IL-6 treatment of LNCaP cells attenuates the expression of AR and the lack of AR expression correlates with neuroendocrine phenotypes; LNCaP variants selected for their adaption to long-term culture in IL-6-supplemented medium are characterized by restored AR expression and reversion back to an epithelial phenotype (Lee et al. 2007). As described earlier, IL-6, IGF-1 and likely HB-EGF (in analogy to EGF, TGFa) destabilize AR via the PI3K/Akt/mTOR pathway. The PI3K/Akt/mTOR pathway was recently reported to be required for NED of LNCaP (Wu and Huang 2007). Thus, the activation of the PI3K pathway and the depletion of AR may be

516

C.G. Tepper, H.-J. Kung

common denominators of all the treatments that induce NED in LNCaP. These data taken together suggest that prostate cancer cells possess the propensity to undergo neuroendocrine or epithelial differentiation and underscore the critical role of androgen receptor in determining the ‘‘fate’’ of these cells.

5 Androgen Withdrawal-Induced Signaling 5.1

MultiLevel Repression of Androgen Receptor Signaling

Androgen-deprivation therapy (ADT) achieves a comprehensive suppression of androgen signaling by affecting the pathway at multiple levels. The pivotal mediator of ADT (i.e., removal of DHT) is of course the androgen receptor (AR). In addition to this leading to the desired loss in AR transcriptional activity, the expression of AR protein is almost completely diminished. This is unique to CaP since the analogous endocrine approach for breast cancer does not impact the expression of the estrogen receptor. Since most castration-recurrent and androgenindependent CaPs are characterized by the reappearance of AR expression and reinstatement of AR signaling, albeit incomplete (Amler et al. 2000), intense research efforts are focused upon better understanding the mechanisms and complexity of AR-mediated transcription and expression of the AR protein, especially in the presence of castrate levels of androgen. The purpose of this section is to discuss the influence of androgen withdrawal upon the androgen signaling pathway.

5.1.1

Androgen Receptor Transcription

It is worthwhile to discuss general mechanisms of AR transcription since they also influence AR expression. The AR is a ligand-activated transcription factor and member of the steroid hormone receptor superfamily, which includes the estrogen, progesterone, and glucocorticoid receptors. The AR consists of three structural domains: the N-terminal transactivation domain, DNA-binding domain (DBD), and ligand-binding domain (LBD). The latter confers specificity of the AR for DHT. After translation and prior to ligand binding, the inactive AR is retained in the cytoplasm and stabilized in a complex with the chaperone protein Hsp90. Binding of DHT causes a conformational change that leads to dissociation of the AR from Hsp90 and nuclear translocation, which is mediated by virtue of a putative bipartite nuclear localization sequence (NLS) contained in the DBD and hinge region. The AR subsequently binds to the enhancer/promoter regions of its target genes via the DBD’s zinc fingers. The conformation of DHT-bound AR promotes the assembly of a transcription complex containing the molecular machinery required for chromatin remodeling and RNA synthesis. The complexes can have varied composition, but are typically composed of members of the p160 coactivator family of histone

Cellular and Molecular Signatures of Androgen Ablation of Prostate Cancer

517

acetyltransferases (HATs), CREB-binding protein, p300, histone demethylases, and RNA polymerase II (Louie et al. 2003; Shang et al. 2002; Wissmann et al. 2007; Yamane et al. 2006). Many other proteins with diverse activities have also been identified as having AR coregulatory functions and have been reviewed (Lee and Chang 2003; Chmelar et al. 2007). Interestingly, the antiandrogen bicalutamide inhibits AR activity not only by antagonizing androgen binding, but also by functioning as a ligand that ineffectively recruits coactivators (Hodgson et al. 2007). Instead, it favors the formation of a corepressor complex through the recruitment of nuclear receptor corepressor (NCoR) and silencing mediator of retinoid and thyroid receptors (SMRT) and histone deacetylases (HDACs) (Shang et al. 2002; Hodgson et al. 2007). Similarly to androgen deprivation, bicalutamidemediated inhibition of AR transactivational function also leads to down-regulation of its expression (Furutani et al. 2002).

5.1.2

Androgen Receptor Turnover

In addition to decreased transcriptional activity, the down-regulation of AR levels is another critical event triggered by androgen ablation. This response is generally explained by the dramatic destabilizing effect upon the receptor due to removal of androgen. Although androgen has potent stabilizing effects upon AR, the mechanisms responsible for this are not entirely clear, but appear to be dependent upon the transcriptional and nongenotropic functions of the receptor. Androgen was originally implicated in modulating the stability and/or translational efficiency of the AR by the demonstration that treatment of LNCaP cells with DHT or the synthetic androgen methyltrienolone (mibolerone; R1881) increased AR protein levels by 30% while reversibly suppressing AR mRNA levels by as much as 67% (Krongrad et al. 1991; Wolf et al. 1993). The ability of androgen to enhance the stability of its receptor was directly shown in studies with transfected COS cells demonstrating that while the half-life (t1/2) of AR was 6 h in the presence of R1881 (100 nM), it was rapidly degraded in its absence and exhibited a t1/2 of only 1 h (Kemppainen et al. 1992). In the context of CaP, pulse-chase experiments in LNCaP cells demonstrated that hormone withdrawal reduced the long t1/2 of endogenous AR from >12 h in the presence of androgen to 3 h (Gregory et al. 2001). Further, the AR in androgen-dependent CWR22 xenografts growing in intact mice had a similar t1/2 to that of LNCaP, but decreased twofold following castration. As alluded to earlier, restoration of AR expression is a major feature of androgen independence and is exemplified in several experimental models of androgen independence that exhibit enhanced stability of the AR in the absence of hormone, including LNCaPC4-2 (t1/2 = > 7 h), CWR-R1 (t1/2 = > 6 h), and recurrent CWR22 xenograft tumors (t1/2 = > 12 h) (Gregory et al. 2001). The ubiquitin-proteasome pathway (UPP) is a major mechanism responsible for AR turnover. Sheflin et al. identified a strong PEST (proline-, glutamate-, serine-, and threonine-rich) sequence in the AR hinge region (Sheflin et al. 2000). PEST sequences serve as degradation signals via targeting proteins for the attachment of

518

C.G. Tepper, H.-J. Kung

ubiquitin and subsequent proteolysis by the proteasome. The role of the UPP in AR degradation was supported experimentally by treatment of LNCaP and HepG2 cells with the proteasome inhibitor MG-132, which resulted in increased levels of the 110-kDa AR as well as a dramatic accumulation of polyubiquitinated species (Sheflin et al. 2000). Additional studies implicated Akt-mediated phosphorylation of AR upon Ser210 and Ser790 as a signal for, or regulator of, ubiquitination via the recruitment of Mdm2 E3 ubiquitin ligase (Lin et al. 2002). Indeed, expression of constitutively active Akt led to enhanced polyubiquitination and degradation of the AR, both of which could be reduced by preventing Akt-mediated phosphorylation (i.e., mutant AR, PI3K inhibition), Mdm2 function, or proteasome activity. The AR can also be specifically targeted for the UPP through an interaction with the carboxyl terminus of Hsc70-interacting protein (CHIP) (Ballinger et al. 1999), which is a U-box E3 ligase and Hsp70/Hsp90 cochaperone (Jiang et al. 2001). CHIP binds to a 14-amino acid sequence in the amino terminus of the AR at amino acid residues 234–247 and promotes its degradation (He et al. 2004). Since CHIP contains a tetratricopeptide repeat (TPR) domain, it also serves as a link between Hsp90 function and the UPP. All the links in the molecular chain leading from androgen withdrawal to degradation of the AR have yet to be established. However, classic and emerging evidence implicate AR’s association with heat-shock protein 90 (Hsp90) as a critical point of androgenic regulation. Indeed, the ability of hormone to regulate the LBD-Hsp90 interaction is considered the earliest event in hormone action. Hsp90 is a molecular chaperone that functions to ensure proper protein folding upon translation and refolding after denaturing stress (Young et al. 2001). In the case of steroid hormone receptors, this is vital to maintaining a structure capable of binding ligands and protecting them from degradation. Hsp90 is unique among the various classes of chaperones and chaperonins in that its substrates are primarily signaling proteins, including steroid hormone receptors (Pratt and Toft 1997) and a variety of kinases, including Src (Schuh et al. 1985) and Akt (Sato et al. 2000). The obligatory role for Hsp90 in maintaining the stability and function of the AR protein is underscored by the ability of the benzoquinone ansamycin geldanamycin (GA), a specific pharmacological inhibitor of Hsp90, to markedly diminish ligand binding (Georget et al. 2002) and steady-state AR levels (Whitesell and Cook 1996; Segnitz and Gehring 1997). Conversely, earlier studies had demonstrated that the AR-stabilizing effects of transition metal oxyanions (e.g., molybdate, tungstate) were mediated through stabilization of receptor-Hsp90 complexes (Renoir et al. 1990; Wright et al. 1981; Trachtenberg et al. 1981; Gaubert et al. 1980; Thompson and Chung 1984). Hsp90 is involved in the latter stages of protein folding after the transfer of substrates from Hsp70. Substrate binding to Hsp90 is ATP dependent and regulated through the dynamic assembly of a multichaperone complex containing Hsc70 and a variety of hsp-associated proteins (Johnson and Toft 1994). Many of these are tetratricopeptide repeat (TPR) domain-containing proteins that interact with the carboxy-terminal residues of Hsp90 (MEEVD) (Young et al. 1998) and/or Hsc70

Cellular and Molecular Signatures of Androgen Ablation of Prostate Cancer

519

(GPTIEEVD). Initially, the AR interacts cotranslationally with Hsp70 and is then transferred to Hsp90 via the Hsc70–Hsp90-organizing protein (Hop), which serves as an adaptor molecule by virtue of it containing two TPR domains (TPR1, TPR2A) that bind in a specific manner to each Hsp (Scheufler et al. 2000). A mature Hsp90AR complex is formed by dissociation of Hsc70 and Hop followed by binding of ATP, the cochaperone p23, and one of four TPR domain-containing immunophilins, FK506-binding protein (FKBP) 52, FKBP51, cyclophilin-40 (Cyp40), or phosphoprotein phosphatase 5 (Chadli et al. 2000; Veldscholte et al. 1992). R1881 treatment of LNCaP cells results in dissociation of the complex and transformation to the activated, nuclear-binding form of the AR (Veldscholte et al. 1992). The occurrence of androgen-induced AR-Hsp90 complex dissociation presents a paradox in terms of Hsp90s stabilizing influence in that (1) although androgen potently stabilizes AR, the ligand-activated receptor is not bound to Hsp90 and (2) in the absence of ligand, AR would be expected to be stable due to its complex formation with Hsp90. However, recent findings might explain and reconcile these apparent inconsistencies. First, it is conceivable that androgen withdrawal impacts upon the functioning of the Hsp90 chaperone complex via transcriptional modulation of one or more of its constituents. This could be significant from the standpoint that TPR-containing proteins can compete for binding to Hsp90 (Young et al. 1998; Chen et al. 1998). Along these lines, the TPR domain-containing immunophilin FKBP51/FKBP5 is an AR target gene (Febbo et al. 2005); androgen withdrawal would therefore lead to the down-regulation of FKBP51/FKBP5 and eventual depletion from complexes. In the absence of FKBP51/FKBP5, conditions might be particularly favorable for CHIP to be incorporated into the Hsp90 complex by virtue of its TPR domain and ability to associate with the AR. In addition to CHIP possessing E3 ligase activity, UPP-mediated AR degradation would be promoted by the ability of CHIP to reduce overall chaperone efficiency by negatively regulating the forward reaction of the Hsc70–Hsp70 substrate-binding cycle, causing p23 dissociation, and decreasing net ATPase activity of Hsp90 (Connell et al. 2001). Similarly, the AR-inhibitory effects of antiandrogens (e.g., cyproterone acetate, phospho-flutamide, nilutamide, and bicalutamide) have been attributed to their ability to prevent dissociation of the AR from Hsp90 by blocking it in an intermediate complex (Georget et al. 2002; Veldscholte et al. 1992). Recent data suggest that androgen-mediated dissociation of the Hsp90-AR complex is critical to AR stability by facilitating its transfer to the small heatshock protein Hsp27 (Zoubeidi et al. 2007). This is very interesting since it entails a combination of transcriptional and nongenotropic mechanisms. R1881 stimulates p38 MAPK-dependent Hsp27 phosphorylation upon Ser78 and Ser82. In turn, phospho-Hsp27 interacts with ‘‘free’’ ligand-activated AR and/or actively displaces Hsp90. This event is critical to AR stability as evidenced by the ability of an Hsp27specific antisense oligodeoxynucleotide (OGX-427) to suppress Hsp27 levels and stimulate AR ubiquitination and proteasomal degradation. In addition, Hsp27 was translocated to the nucleus with the AR and enhanced its transcriptional activity by functioning as a coregulator to enhance binding of AR to androgen-response

520

C.G. Tepper, H.-J. Kung

elements (AREs). It would be expected that androgen deprivation would abrogate Hsp27 phosphorylation and its interaction with AR. At the same time, Hsp27 represents a major cytoprotective mechanism of CaP during androgen ablation and of persistence of AI CaP since its expression was markedly elevated in human CaP specimens obtained from patients after neoadjuvant hormonal therapy and hormone-refractory metastatic lesions (>fourfold) (Rocchi et al. 2004, 2005). A particularly troubling aspect of Hsp27 overexpression is that it confers resistance to the GA analog, 17-AAG (McCollum et al. 2006). AR proteolysis can also be mediated by caspases and calpain. These are interesting since they produce easily identifiable processing intermediates as a result of the sequence specificity intrinsic to each protease. The AR has six putative caspase cleavage sites, but appears to be cleaved in vivo at only one site that resides in the amino-terminus to produce fragments of 18- and 87 kDa (Wellington et al. 1998). In vitro, caspases-1, -3, -7, and -8 can utilize 35S-labeled AR as a substrate, but caspase-3 exhibits the highest activity and cleaves at least two additional sites. Compared to the prototypical caspase-3 death substrate poly(ADP-ribose) polymerase (PARP), AR is cleaved inefficiently. However, this becomes enhanced if the length of the polyglutamine repeat region is increased (Kobayashi et al. 1998). This is particularly relevant in the case of spinal bulbar and muscular atrophy (SBMA) which results from an AR mutation characterized by a pathogenic expansion of the AR gene CAG repeat from the normal (10–36 repeats) to 37–66 repeats. In contrast, a shortened AR polyglutamine repeat region is frequently found in CaP specimens and potentially decreases its susceptibility to apoptotic caspase cleavage. Calpain is one of the best-characterized calcium (Ca2+)-activated proteases and has emerged as a critical protease in AR processing (Pelley et al. 2006; Libertini et al. 2007). It was found associated with AR in a multiprotein complex containing calpastatin and calmodulin (Pelley et al. 2006). Consistent with the biochemical properties of calpain, increasing the concentration of Ca2+ in cell extracts or intact cells promoted the generation of AR cleavage products 76, 50, and 31/34 kDa and was inhibited by the addition of specific calpain inhibitors (e.g., calpastatin, calpepetin) and calcium chelators [e.g., EGTA, 1,2-bis-(O-aminophenoxy)-ethaneN, N, N0 , N0 -tetraacetic acid tetra-(acetoxymethyl)-ester (AM-BAPTA)]. The clinical relevance of this is supported by the presence of these cleavage products in extracts of CaP tumor specimens (Libertini et al. 2007). Interestingly, the data can be interpreted for calpain as having dual roles in CaP biology, one being as a mechanism of AR turnover (Pelley et al. 2006) and secondly as a mechanism of androgen independence (Libertini et al. 2007). With respect to the latter, the 76-kDa calpain-derived product represents a constitutively active (i.e., ligand-independent) AR and the basis for AI growth in the 22Rv1 AI cell model (Libertini et al. 2007; Tepper et al. 2002). In these cells, as well as the recurrent CWR22 xenograft from which they were derived, this product is preferentially generated as a consequence of an in-frame tandem duplication of exon 3 that sensitizes the extended, full-length AR protein to calpain proteolysis. Conversely, treatment of 22Rv1 cells with calpain inhibitors in vitro can promote apoptosis in the absence of androgen and retard their growth as tumors in castrated mice. Data also suggest that aspects of

Cellular and Molecular Signatures of Androgen Ablation of Prostate Cancer

521

calpain activation might be regulated by androgen withdrawal since the truncated AR in androgen-deprived 22Rv1 was selectively retained/generated in the nucleus and was processed to a doublet shortly after the removal of androgen (Tepper et al. 2002).

5.1.3

Transcriptional and Post-transcriptional Regulation

Two critical processes that factor significantly into the equation determining steady-state levels of AR protein are (1) transcription of the AR gene locus and (2) translation of the AR mRNA transcript. The importance of AR transcript regulation in the context of CaP progression was underscored by a microarraybased study that demonstrated that overexpression of the AR mRNA was the most consistent finding in castration-recurrent derivatives of the common human CaP xenograft models compared to their androgen-dependent counterparts (Chen et al. 2004). This is a particularly significant phenomenon since the AR is autoregulated in that AR activation normally down-regulates AR gene expression. For example, AR mRNA levels increased twofold to tenfold upon androgen withdrawal (i.e., castration) in the rat ventral prostate, kidney, brain, seminal vesicle, and epididymis (Quarmby et al. 1990). Conversely, androgenic stimulation with testosterone proprionate decreased AR mRNA expression in these tissues to below basal levels and androgen replacement led to a fourfold reduction in castration-induced AR mRNA (Quarmby et al. 1990). This effect is conserved in hormone-sensitive CaP, as well as AR-positive T47 breast cancer cells, since treatment of LNCaP cells with testosterone or the synthetic androgen R1881 decreased AR mRNA to 30% of control levels by 48 h and decreased its rate of transcription fourfold (Wolf et al. 1993; Quarmby et al. 1990). Since this can be partially inhibited by the antiandrogen hydroxyflutamide and was not observed in rats with androgen insensitivity, this further supports the role of AR activation in the negative feedback regulation of its own gene expression. The discordant response between AR mRNA and protein levels in response to hormonal manipulations is distinct from other steroid family members. The mechanisms responsible for the transcriptional control of the AR gene are not well understood. Cloning and sequencing of the AR genomic and coding sequences provided great insight and tools for studying this mechanism as well as translational control (Lubahn et al. 1988a, b; Baarends et al. 1990; Mizokami et al. 1994; Tilley et al. 1990). This work demonstrated that human, mouse, and rat prostates express a major AR mRNA species of 9.6–10 kb and a less prominent 7-kb species (Lubahn et al. 1988a, b). Two major transcription initiation sites (TIS I/II) were identified in the human mRNA 1.1-kb (1,115 and 1,126 bases) upstream of the translation initiation codon (Tilley et al. 1990) as well as in the rat AR at 1,023 and 1,010 (Baarends et al. 1990). A minimal promoter is located between 300 and +100 (relative to the TIS, designated as +1), but lacks typical TATA and CCAAT box elements (Baarends et al. 1990; Tilley et al. 1990), which could account for the heterogeneous initiation (Maniatis et al. 1987). This region contains two potential

522

C.G. Tepper, H.-J. Kung

regulatory elements in close proximity to one another spanning positions 117/-32: a homopurine stretch (purine-rich region) and a GC-box (Baarends et al. 1990). The latter element was shown to be critical for regulating transcription initiation from TIS II and required the binding of the Sp1 transcription factor to a consensus site within the GC-box (Faber et al. 1993). In addition, the promoter/50 -UTR could impart a tissue specific regulation with respect to the usage of initiation sites since there was a significant difference between the ratios of TIS-I- and -II-initiated transcripts in LNCaP prostatic and T47 mammary adenocarcinoma cells. Additional analyses of the AR promoter revealed multiple potential cis-acting elements upstream of the TISs, including glucocorticoid response element (GRE), 2,3,7,8- tetrachlorodibenzo-p-dioxin (TCDD), nuclear factor-1 (NF-1), nuclear factor-kB (NF-kB), activator protein-1 (AP-1), retinoic acid response element (RARE), interleukin-6 (IL-6), and cAMP response element (CRE) (Baarends et al. 1990; Mizokami et al. 1994; Song et al. 1995). Many of these sites represent downstream targets of signaling pathways highly relevant to CaP biology. For instance, both IL-6 and cAMP/protein kinase A (PKA) have been implicated in CaP progression via mechanisms including the induction of NED and steroidindependent AR activation (Lee et al. 2007; Hsieh et al. 1993; Wallner et al. 2006; Desai et al. 2005). Notably, treatment of LNCaP cells with dibutyryl cAMP up-regulated AR mRNA and induced chloramphenicol acetyltransferase (CAT) activity of a reporter driven by the CRE derived from the AR promoter (Mizokami et al. 1994). While these results shed light upon AR gene regulation, the roles of these elements in mediating increased AR transcription during androgen withdrawal have yet to be determined. While AR activation autoregulates AR gene transcription in a negative manner, it exerts a positive influence upon post-transcriptional events, namely mRNA stability and translation. A very early effect (i.e., 1 h) of testosterone treatment is to induce AR protein synthesis and resulting nuclear accumulation in vivo in rat ventral prostate (Mora et al. 1996). While it was expected that testosterone-induced AR protein levels would be diminished in the presence of the protein synthesis inhibitor cycloheximide (CHX), AR mRNA levels also decreased, suggestive that protein synthesis was also necessary for androgen-mediated enhancement of AR mRNA stability in autoregulation (Mora et al. 1996). The AR mRNA-stabilizing effect of DHT was also observed in LNCaP cells (Yeap et al. 1999). The proximal 30 -UTR of the mRNA has been implicated in modulating transcript stability since it contains a highly conserved UC-rich region that contains 50 -C(U)nC and 30 -CCCUCCC poly (C)-binding protein (CP) motifs and to which three RNA-bining proteins involved in the control of mRNA stabilization (HuR/ELAV1) and turnover and translation (CP1 and CP2) were found to interact (Yeap et al. 2002). Although reporter assaybased experiments suggest that the UC-rich region mediates a general reduction in AR mRNA levels, androgen apparently does not affect this. Translational regulation has not been extensively defined for the AR, but stands as a prominent mechanism for controlling its levels. Mizokami and Chang functionally characterized the 50 -UTR and identified a 180-bp region of (+21/+202 relative to transcription start site; 1,106/926 relative to initiator codon) that

Cellular and Molecular Signatures of Androgen Ablation of Prostate Cancer

523

was vital for translational induction of AR expression without having any effect upon transcription (Mizokami and Chang 1994). Translational efficiency was enhanced fivefold, which was due in part (44% of the activity) by the formation of a putative stem-loop secondary structure (21-bp) in the center of this region. AR synthesis was recently reported to be mediated by cap-dependent translation and regulated by mTOR signaling (Cinar et al. 2005). Paradoxically, HB-EGFmediated mTOR stimulation attenuated AR mRNA translation. This could be reversed by inhibition of mTOR with the pharmacological inhibitor rapamycin or by silencing with siRNAs, which resulted in enhanced interactions of the initiation factor eIF4E and scaffolding protein eIF4G with the m7G mRNA cap structure of the AR mRNA (Cinar et al. 2005). Our unpublished data demonstrate that other EGF family members (e.g., EGF, TGF-a, HRG-b1) and the inflammatory cytokine IL-6 can also trigger AR down-regulation. For the former stimuli, this occurs in an mTOR-dependent manner. Despite the capacity for androgen to maintain higher basal AR levels, polypeptide growth factors were still able to markedly down-regulate expression.

5.2

Androgen Withdrawal-Induced Kinase Signaling

Androgen ablation triggers a definitive signature of molecular and biochemical changes in CaP cells. This is quite understandable based upon the marked effects elicited by androgen stimulation. Importantly, these have helped to begin to explain the basis for the biological consequences of androgen withdrawal, namely growth arrest, NED, cell survival, and the eventual outgrowth and persistence of the disease. Since growth arrest and tumor regression in vivo are such prominent effects, androgen withdrawal can be misleading in terms of it leading to a cellular state that is seemingly static or nondynamic. This is actually very far from the truth since although the cells become quiescent, they remain highly active based upon various criteria, including signaling, transcription, and metabolism. This is due in large part to multiple stimuli provided by extracellular cues, such as growth factors, extracellular matrix, etc. The purpose of this section is to discuss some of these signaling events in order to illuminate the importance of the AR in regulating kinase signaling events and in determining their outcomes in hormone-dependent neoplasms. Inappropriate activation and persistence of kinase signaling pathways contribute greatly to the survival of CaP when challenged by androgen withdrawal and in fostering the development of castration-recurrent CaP. These have been studied extensively in the context of the androgen-sensitive LNCaP cell model, which do not undergo apoptosis in the absence of androgen (Berchem et al. 1995; Murillo et al. 2001). These cells have a distinct survival advantage by virtue of a frame-shift mutation in the PTEN/MMAC1 tumor suppressor gene that encodes a lipid phosphatase (Steck et al. 1997; Li et al. 1997; Vlietstra et al. 1998). PTEN inactivation manifests in elevated PI3‐K signaling as a result of the accumulation and sustained

524

C.G. Tepper, H.-J. Kung

levels of 3-phosphorylated inositides that it produces, including phosphatidylinositol 3,4,5-trisphosphate [PI(3,4,5)P3] and PI(3,4)P2. Notably, androgen deprivation of LNCaP triggers a gradual and persistent hyperactivation of the PI3K-Akt pathway and becomes hyperdependent upon PI3-K signaling for survival in the absence of androgen or when AR activity is inhibited with Casodex (Murillo et al. 2001; Tepper et al. 2001). This was evidenced by increased PI3-K and Akt activities in in vitro kinase assays as well as by elevated levels of Ser473-phosphorylated Akt beginning 16 h after the withdrawal of androgen. The potency of the antiapoptotic signals elicited by PI3-K was demonstrated by findings that apoptosis was markedly induced (i.e., 60–80%) by pharmacological inhibition of PI3-K activity with LY294002 or wortmannin or by restoring PTEN function by ectopic expression (Murillo et al. 2001; Lin et al. 1999). Furthermore, specific inhibition of Akt with perifosine, a synthetic membrane-permeable alkyl-lysophospholipid (Kondapaka et al. 2003), preferentially induced PARP cleavage in the absence of androgen (Vinall et al. 2007). In the presence of androgen, PI3K-Akt inhibition mediated growth arrest with no evidence of apoptosis. The importance of PI3K-Akt in this context is further highlighted by the fact that inhibition of p44/42 MAPK/ ERK, p38 MAPK, or protein kinase A (PKA) signaling by treatment with PD98059, SB203580, and H89, respectively, had negligible effects upon survival. In our experiments, PI3-K hyperactivation continued for at least 60 days following the initiation of androgen withdrawal (Tepper et al. 2001). This represents a critical molecular event in the transition to androgen independence since marked elevations in PI3K-Akt levels are a consistent feature of AI LNCaP sublines derived by longterm culture in the absence of androgen in vitro (Murillo et al. 2001; Shi et al. 2004) as well as from LNCaP xenografts (Graff et al. 2000). However, unlike their androgen-sensitive, parental cells, androgen-independent LNCaP-cds cell lines were remarkably resistant to apoptosis induced by PI3K inhibition. This was due in large part to the inability of the inhibitors to completely abrogate Akt phosphorylation; in fact, the remaining phospho-Akt levels in the three AI cell lines tested (Shi et al. 2004) were higher than basal levels observed in LNCaP. The clinical relevance of these findings was demonstrated in studies using patient samples, which revealed that increased phospho-Akt(S473) staining intensity correlated with high Gleason scores (Malik et al. 2002), and also with the occurrence of biochemical/prostate-specific antigen (PSA) failure (Kreisberg et al. 2004). Taken together, these results imply that adjuvant targeting of compensatory survival pathways during androgen ablation, but prior to the development of castrationresistant disease, would be more efficacious if implemented prior to the development of castration-resistant disease. Since Akt phosphorylates a number of target substrates that function in the regulation of apoptosis, androgen withdrawal-induced PI3K-Akt hyperactivation engenders a comprehensive survival program. Through several complementary mechanisms, PI3K-Akt signaling promotes Bcl-2-mediated survival by simultaneously fostering interactions between antiapoptotic Bcl-2 family proteins (e.g., Bcl-2, Bcl-XL) and suppressing the function of proapoptotic members, such as Bad (BCL2-anatagonist of cell death) and Bid (BH3-interacting domain death agonist).

Cellular and Molecular Signatures of Androgen Ablation of Prostate Cancer

525

The most direct action proceeds via inactivation of Bad by Akt phosphorylation; Ser136-phosphorylated Bad is sequestered in the cytoplasm by the tau form of 14-3-3 proteins and is unable to form heterodimers with Bcl-2 and Bcl-XL and neutralize their antiapoptotic effects. Phosphorylation-dependent Bad inactivation has indeed been demonstrated to be critical for LNCaP survival when subjected to androgen deprivation (Sastry et al. 2006a, b). Consistent with the data discussed earlier, LY294002-mediated apoptosis of serum-starved cells is accompanied by Bad dephosphorylation. Likewise, expression of a dominant-negative Bad phosphorylation-defective mutant (S112A/S136A) triggers apoptosis even in the presence of functional PI3K-Akt. Akt also attenuates death receptor-mediated apoptosis triggered by tumor necrosis factor (TNF)-related apoptosis-inducing ligand (TRAIL) by attenuating caspase-8-mediated Bid processing. Although TRAIL resistance correlates with Akt activity, Bid cleavage and sensitivity can be restored by inhibiting PI3‐K (Chen et al. 2001; Rokhlin et al. 2002). Interestingly, increased Bcl-2 expression is also observed during hormone withdrawal of LNCaP cells and implicated as a major mechanism of survival (Berchem et al. 1995; Raffo et al. 1995). PI3K-Akt signaling potentially contributes to this event by enhancing Bcl-2 transcription by activation of NF-kB (Catz and Johnson 2001; Ozes et al. 1999; Romashkova and Makarov 1999) and/or by inactivation of Par-4 (Goswami et al. 2005), both of which have binding sites in the BCL2 promoter (Cheema et al. 2003). The proapoptotic protein prostate apoptosis response-4 (Par4) negatively regulates Bcl-2 transcription by binding to a Wilms’ tumor 1 (WT1) site in the BCL2 promoter (Cheema et al. 2003; Chendil et al. 2002; Camandola and Mattson 2000). Akt-mediated phosphorylation prevents its nuclear translocation and inhibits apoptosis of LNCaP and PC-3 cell lines (Goswami et al. 2005). In addition to Akt relieving repression of BCL-2 transcription by Par-4, Par-4 inactivation would also lead to derepression of NF-kB activity and increased Bcl2 expression (Chendil et al. 2002). As alluded to earlier, Akt is generally associated with the inactivation of proapoptotic and/or growth-inhibitory transcription factors. The forkhead factors, FOXO1/FKHR and FOXO3/FKHRL1, figure prominently in this regard and are direct targets of Akt. Upon phosphorylation, both are predominantly localized to the cytoplasm due to a shift in the equilibrium of nuclear-cytoplasmic shuttling. The clinical relevance of this is underscored by clinical studies that demonstrated that cytoplasmic phosphorylated FKHR was a significant indicator of biochemical recurrence (Li et al. 2007). As expected, androgen withdrawal results in heightened Thr32-phosphorylated FOXO3 levels. Inhibition of PI3K activity with LY294002 and wortmannin leads to rapid dephosphorylation and nuclear localization preceding engagement of the execution phase of apoptosis. Consistently, the FKHR-AAA (T24A/S256A/S319A) phosphorylation mutant is constitutively retained in the nucleus and riggers apoptosis of LNCaP. In the LAPC CaP model, apoptosis can be induced by adenoviral-mediated overexpression of FOXO1/3. Forkhead factors directly and indirectly regulate the expression of many genes (Ramaswamy et al. 2002; Modur et al. 2002), suggesting that it is unlikely that a single gene is responsible for their proapoptotic activities. Interestingly, TRAIL/TNFSF10 is a

526

C.G. Tepper, H.-J. Kung

direct target gene and is up-regulated by FOXO factors through their binding to a response element in the proximal promoter (Modur et al. 2002). Taken together, the data suggest that combined androgen withdrawal plus PI3K inhibition promotes apoptosis through the establishment of a TRAIL-based autocrine/paracrine suicide pathway. Additional insight into this mechanism has been provided by the recent finding that under conditions of PI3K suppression, androgen antagonizes TRAILmediated apoptosis by up-regulating the expression of the FLIP/CFLAR gene (Cornforth et al. 2008) and subsequent inhibition of caspase-8/FLICE. This is mediated by the interaction of ligand-bound AR with FOXO3 and subsequent FLIP transcriptional activation via a novel Forkhead consensus binding site in its promoter. In the absence of androgen, the AR-FOXO3 interaction is greatly diminished, and FLIP mRNA and protein levels decline. As a consequence, TRAIL sensitivity is most likely restored by relief of caspase-8 inhibition and resumption of Bid processing. There is also strong evidence that implicates PI3K-Akt hyperactivation in further suppressing AR signaling after removal of androgen; this highlights the existence of a tightly regulated PI3K-AR signaling axis or loop in which each component negatively regulates or controls the other. Furthermore, the transition to androgen independence is characterized by a dissociation of the negative regulatory effects. Although it is difficult to construct a hierarchy of the androgen withdrawalinduced biochemical events, we can discuss them as they might occur in response to androgen ablation. While AR acts as an upstream negative regulator of PI3K, it is also a direct target for phosphorylation by Akt at Ser210 and Ser790 (Lin et al. 2001). Paradoxically, Akt phosphorylation mediates down-regulation of AR transactivation and might contribute to its destabilization in the absence of androgen as a result of phosphorylation-dependent ubiquitylation and proteasomal degradation (Lin et al. 2002). Indeed, we have observed that Akt immunoprecipitated from androgen-deprived cells can phosphorylate recombinant AR and that MG-132 can partially stabilize AR during androgen withdrawal. While this might be paradoxical, it might be necessary for development of androgen independence since AR signaling can inhibit growth and induce apoptosis in cells hyper-responsive to androgen (Shi et al. 2004; Lin et al. 2001; Kokontis et al. 1998). However, consistent with the earlier findings, Akt activation stimulates AR signaling in late passage LNCaP (Lin et al. 2003). Recent findings support a hypothesis that mammalian target of rapamycin (mTOR) acts as a critical sensor of androgen signaling and regulator of androgen withdrawal-induced AR down-regulation and PI3K-Akt hyperactivation. First, activation of the rapamycin-sensitive raptor/mTOR complex (mTORC1) by polypeptide growth factors (including EGF, TGF-a, HB-EGF, and IGF-1) has been demonstrated to decrease AR levels via suppression of AR mRNA translation (Cinar et al. 2005). Second, in the context of a hormone-dependent neoplasm such as LNCaP, androgen strongly stimulates, and is required to maintain full, basal activation of mTORC1 (Xu et al. 2006), thereby establishing an AR-mTOR negative feedback regulatory loop. Accordingly, inhibition of mTORC1 with rapamycin or LY294002 abrogates the effects of growth factors and increases

Cellular and Molecular Signatures of Androgen Ablation of Prostate Cancer

527

basal AR levels in the presence of androgen (Cinar et al. 2005; Tepper et al. 2007). Although androgen withdrawal results in a substantial reduction in mTORC1 activity, the residual mTORC1 signal derived from serum growth factors and nutrients continues to drive down AR expression. The significance of this is underscored by the ability of rapamycin to restore AR expression in androgen-deprived cells to levels equivalent to, or greater than, that found under basal, androgensupplemented conditions. mTORC1 down-regulation also provides an explanation for androgen withdrawal-mediated hyperactivation of the PI3K-Akt pathway. In L6 myocytes, amino acid-stimulated mTOR activity can suppress insulin-mediated PI3K and Akt activation as a result of mTOR/p70S6K1-mediated phosphorylation of insulin receptor substrate-1 (IRS-1) (Tremblay and Marette 2001; Werner et al. 2004; Harrington et al. 2004). Although we are speculating that androgen attenuates PI3K through a similar mechanism in LNCaP cells, we have observed that (1) rapamycin treatment recapitulates the elevation of Akt Ser473 phosphorylation observed in response to androgen withdrawal and (2) that further treatment of androgen-ablated cells does not increase phospho-Akt levels further. Importantly, the ability of rapamycin to augment two dominant survival signaling pathways (i.e., AR, PI3K-Akt) by rapamycin suggests the use of mTOR inhibitors for the treatment of PTEN mutant CaPs should be carefully considered since these are common features of AI CaP cell lines.

6 The Transcriptional Profile of Androgen Withdrawal The biological sequelae of androgen withdrawal are mediated in large part via dramatic changes in the transcriptional program controlled by the AR. The advent of comprehensive, genome-wide microarray technologies has greatly accelerated research efforts aimed at identifying androgen-regulated genes and has enabled the generation of molecular profiles that can serve as molecular signatures of AR function, or lack thereof, in response to different ligands, throughout various stages of CaP progression, and in response to selected therapeutics. Advances in the field of bioinformatics have contributed significantly to better utilizing the resulting large microarray datasets by facilitating the classification of the genes in terms of molecular function and biological processes and to further reveal signal transduction networks reflective of these gene expression changes (Doniger et al. 2003; The Gene Ontology Consortium 2001). We were specifically interested in gaining insight into the genetic mechanisms underlying the responses of LNCaP cells to androgen deprivation, including survival, growth arrest, and neuroendocrine differentiation. To this end, we performed gene expression profiling of LNCaP cells cultured in hormone-free medium for 0.5, 6, 24, and 96 h and conducted comparison analysis to identify all the genes that were differentially expressed with respect to LNCaP cells maintained in androgensupplemented medium. As expected from profiling studies defining androgeninduced genes, we observed a multitude of expression changes throughout this

528

C.G. Tepper, H.-J. Kung

time course and with varying kinetics. Since we are primarily interested in the stable effects of this treatment, we will limit our discussion to some of the most highly differentially expressed genes (2.5-fold) after 96 h. This yielded a profile composed of 232 genes, 111 up- and 121 down-regulated, that generally exhibited significant modulation 24 h post-androgen withdrawal. Many of the genes downregulated by androgen withdrawal have not been previously identified as AR target genes, but might indeed be regulated via collaboration of the AR with other transcription factors, such as FoxA1, GATA2, and Oct1, via noncanonical enhancer sites, or modulation of higher-order chromatin remodeling (Wang et al. 2007). Interestingly, the androgen withdrawal-induced genes most likely represent indirect, or secondary repressed targets of androgen action since agonist-bound AR does not recruit corepressor complexes. In the following sections, we discuss selected androgen withdrawal-regulated genes in the context of the cellular and molecular features of androgen deprivation.

6.1

Androgen Receptor Signaling Pathway

Androgen withdrawal impacts the AR at various levels including direct inhibition of its transcriptional activity, ligand metabolism, and its expression level. The transcriptional effects of androgen withdrawal can immediately be recognized by markedly diminished expression of twelve known androgen-responsive genes that either contain AREs or are strongly regulated by androgen. These include the kallikrein-related peptidases (KLK2, KLK3/PSA), the transmembrane serine protease TMPRSS2, NKX3.1 transcription factor, and the metabolic enzyme-encoding genes DHCR24 and ACSL3 (Nelson et al. 2002). Interestingly, several gene expression changes would be expected to ‘‘actively’’ facilitate the inactivation of AR signaling in the absence of androgen. Three members of the UDP glucuronosyltransferase 2 family (UGT2B11, UGT2B15, UGT2B17) are among the most highly androgen withdrawal-induced genes (3–11-fold). These encode steroid-conjugating enzymes that function in androgen catabolism, thereby leading to dissociation of DHT from the AR and its clearance from the cell (Chouinard et al. 2007; Belanger et al. 1998; Guillemette et al. 1997). AR transactivation (Febbo et al. 2005) and possibly its stability would further be compromised as a result of diminished expression of FK506-binding protein 5 (FKBP5; 6.06-fold), as discussed earlier. Aside from this, the androgen withdrawal expression signature does not contain any obvious candidates responsible for diminished AR translation or stability. Along these lines, AR transcript levels are actually slightly elevated (1.45-fold). While the AR occupies the highest position in the molecular hierarchy of the androgen withdrawal pathway, the preponderance of the resulting expression profile is potentially shaped in collaboration with auxiliary transcriptional regulatory proteins whose expression is also androgen withdrawal regulated. Indeed, approximately 14.7% (i.e., 34/232) of the differentially expressed genes encode transcription factors, components of the basal transcription apparatus, coregulators, and

Cellular and Molecular Signatures of Androgen Ablation of Prostate Cancer

529

chromatin remodeling proteins. One theme that emerges from this dataset is that AR activity is autoregulated to some extent by modulating the expression of its own coregulators. This is illustrated by increased and decreased expression of the coactivator BAF60C/SMARCD3 (Link et al. 2005; Huang et al. 2005) and corepressor PA2G4/EBP1 (Zhang et al. 2005), respectively, during this state of diminished AR activity. Interestingly, this inverse regulation of coactivator/repressor expression was also observed for the c-Jun (3.37-fold) and c-Fos (17.48-fold) dimerization partners, which can activate or inhibit AR function, respectively, by regulating the amino- to carboxyl-terminal (N-to-C) interaction necessary for AR homodimerization (Bubulya et al. 1996, 2001; Tillman et al. 1998). This also holds true for the well-known steroid receptor p160 coactivator SRC-1/NCOA1 (2.04-fold). Additional transcriptional regulators will be discussed in the following sections.

6.2

Neuroendocrine Differentiation

As alluded to earlier, NED is one of the most critical consequences of androgen withdrawal. In addition to dramatic morphological changes, the newly acquired functional properties of these cells have implications for fostering tumor cell survival and the development of castration-resistant CaP. This shift in differentiation, or possibly dedifferentiation, highlights the normal function of the AR in directing epithelial differentiation and suggests that these cells are possibly derived from disrupted transit cell differentiation and might represent cancer stem cells. The latter is supported by our findings that SEMA6A and CD24 were strongly induced by androgen withdrawal and are typically overexpressed in human embryonic stem cells (hESCs) (Assou et al. 2007). Increased SOX4 expression is another common feature of NED and hESCs (hESC expression atlas, http://amazonia. montp.inserm.fr). CD24-positivity also suggests that androgen-deprived cells still retain some epithelial character and are not the most immature stem cell progenitors (Lawson et al. 2007). However, marked down-regulation of NKX3.1, a homeobox transcription factor, represents a critical event in CaP dedifferentiation since this is a key, androgen-dependent mediator of prostatic terminal differentiation and is essential to prostate ductal morphogenesis and secretory protein production (Bieberich et al. 1996; Bhatia-Gaur et al. 1999; Matusik et al. 2008). NKX3.1 is considered a tumor suppressor (Bhatia-Gaur et al. 1999) and accordingly, loss of its expression is positively correlated with CaPs with high Gleason grades, advanced tumor stage, and with castration-recurrent cancers (Bowen et al. 2000). The androgen withdrawal transcriptional program is significantly biased in favor of NED. It is remarkable that nearly 10% of the genes (23/232) comprising the expression signature are associated with neuronal differentiation or nervous system development. Furthermore, all of these with the exception of one gene (FOS) exhibited increased expression (3.8–7.0-fold) and accounted for nearly 19% (22/ 111) of the androgen withdrawal-induced genes. These include genes that participate in the establishment of a neuronal transcriptional program, confer the characteristic morphological changes, and engender neurosecretory properties.

530

C.G. Tepper, H.-J. Kung

LUZP2, a leucine zipper-motif containing transcription factor, represents one upstream, positive modulator of NED and has expression restricted to the brain and spinal cord (Wu et al. 2003), possibly explained by an upstream neuron-specific enhancer (Stefan et al. 2005). It is interesting to note that the LUZP locus maps within a region of chromosome 11p13–11p14 where breakpoints are associated with the development of Wilms’ tumor-Aniridia-Genitourinary anomalies-mentalRetardation (WAGR) syndrome, in which the genitourinary anomalies include tumors of the gonads, such as testes and ovaries. Transducin-like Enhancer of split 1 (TLE1) expression is elevated 3.46-fold by androgen withdrawal and is a prominent regulator of mammalian neurogenesis through its action as a transcriptional repressor. TLE1 is the mammalian homolog of Drosophila Groucho, the prototypical member of this family of corepressors, and interacts with basic helixloop-helix (bHLH) proteins of the Hairy/Enhancer of split (HES) family, such as HES-1 (Grbavec and Stifani 1996). In this manner, HES-1 provides specific DNA binding and TLE1 provides the corepressor function. The latter is mediated in part via TLE1’s interaction with the amino-terminal domain of histone H3, which is associated with transcriptional silencing (Palaparti et al. 1997). TLE1 is expressed in neural progenitors and is generally considered a negative regulator of neurogenesis since it negatively regulates postmitotic neuronal differentiation (Yao et al. 2000). However, its up-regulation during neural determination of mouse P19 embryonic carcinoma cells and expression in the postmitotic cells of the cerebral cortex and spinal cord suggest that TLE1 expression is dynamically regulated and functions at various stages of neuronal differentiation (Dehni et al. 1995; Yao et al. 1998). TLE1 is also oncogenic since transgenic expression in mice leads to the development of mucinous adenocarcinomas of the lung with nearly complete penetrance and is overexpressed in some human lung cancers (Allen et al. 2006). The role of TLE1 in prostate cancer has not been defined, but it is possible that differential alternative splicing of TLE family transcripts might be important (Nakaya et al. 2007). Interestingly, the amino-terminal enhancer of split (AES/ TLE5) can repress AR-dependent transcription involving an interaction with transcription factor TFIIE (Yu et al. 2001). Taken together, this suggests that in addition to working with HES bHLH proteins, TLE1 potentially functions to promote NED by further suppressing AR function. The transmembrane proteins encoded by NOTCH3 and SEMA6A are both induced by androgen withdrawal and might regulate NED by transmitting extracellular signals from soluble ligands and establishing cell-cell interactions. Notch receptor signaling performs critical signaling functions in neural development, particularly in controlling cell fate decisions of neural progenitors by influencing transcription (Mason et al. 2005). Interestingly, HES-1 and HES-5 are transcriptional targets of activated Notch (Jarriault et al. 1995), thereby suggesting that engendering the Notch-HES-Groucho/TLE1 pathway might represent a significant proneural mechanism of androgen withdrawal. This also represents a convergence point for the IL-6 and androgen withdrawal NED pathways since IL-6 has been demonstrated in MCF-7 mammary carcinoma cells to signal via Notch3 to increase expression of the Notch ligand Jagged-1 as well as carbonic anhydrase IX (Sansone

Cellular and Molecular Signatures of Androgen Ablation of Prostate Cancer

531

et al. 2007). Increased expression of carbonic anhydrase IX is particularly significant in that it promotes invasive behavior and can confer resistance to hypoxia. It is worth mentioning that a key event in ligand-induced Notch activation involves a juxtamembrane cleavage event mediated by a disintegrin and metallopeptidase 17 (ADAM17; tumor necrosis factor-a converting enzyme). Although expression of another family member, ADAM7, was markedly down-regulated (129-fold) in our studies and was previously demonstrated to be an androgen-regulated gene (Oh et al. 2005), it is not known if it participates in the Notch pathway. Semaphorin 6A (Sema6A) is typically expressed in postmitotic neurons (Kerjan et al. 2005) and is a potential mediator of NED phenotypic changes and motility. It functions as the ligand for plexin-A2 and through its control of actin cytoskeleton dynamics (Renaud et al. 2008), it induces growth cone collapse of sympathetic neurons and functions in axon guidance and channeling by providing repulsive guidance cues (Xu et al. 2000; Leighton et al. 2001). Similarly, SEMA3B, a secreted member of the class 3 semaphorin family, was also up-regulated and is important in axon guidance by binding to neuropilins (NP-1/2). The extension of neuritic processes is one of the most characteristic features of the NED phenotype both in vitro and in vivo. While the genes discussed earlier that have axon guidance function can certainly be important in this regard, several additional candidates revealed in the androgen withdrawal signature include midkine (MDK), chromosome 5 open reading frame 13 (C5orf13), cordon-bleu homolog (COBL), and dihydropyrimidinase-like 2 (DPYSL2) and function through the regulation of cytoskeleton dynamics. The neurotrophin MDK, also referred to as neurite growth-promoting factor 2, is a 13.4-kDa cysteine- and basic amino acidrich protein, has pleiotropic effects, and mediates growth, differentiation, and survival of neurons (Muramatsu 1993; Michikawa et al. 1993). Interestingly, it promotes cell-substratum adhesion and can serve as an attractive guidance force in neurite outgrowth for rat embryonic neurons in vitro (Kaneda et al. 1996). Similar to MDK (Muramatsu and Muramatsu 1991), C5orf13/P311 is a retinoic acid (RA)induced gene (Ueda 2001). C5orf13/P311 mRNA encodes an 8-kDa phosphoprotein that promotes neuronal differentiation (Fujitani et al. 2004). It is postulated to reorganize the actin cytoskeleton at the cell periphery through the formation of a signaling complex with Filamin A (also up-regulated by androgen withdrawal approximately twofold) and activating Rac1 GTPase downstream of integrin-b1 (McDonough et al. 2005). COBL mediates the induction of neurites and neurite branching by acting as an actin nucleation factor (Ahuja et al. 2007; Carroll et al. 2003). While critical to nervous system development, it is one of the most frequently mutated coding region microsatellites (23.8%) identified in colorectal cancers (Mori et al. 2001). Growth factor receptor-bound protein 10 (GRB10) is biallelically expressed with COBL and is induced to a similar magnitude by androgen withdrawal (Hitchins et al. 2002). DPYSL2 is also known as collapsin response mediator protein hCRMP-2. DPYSL2 facilitates neurite extension, axon elongation, and branching (Yoshimura et al. 2005), and regulates axon/dendrite fate in that it determines axon outgrowth and axon-dendrite specification. Like COBL, it accomplishes these actions by regulating the actin cytoskeleton by complexing

532

C.G. Tepper, H.-J. Kung

with kinesin-1 and linking it to the specifically Rac1-associated protein 1 (Sra-1)/ WASP family verprolin-homologous protein 1 (WAVE1) complex (Kawano et al. 2005). DPYSL2 also promotes tubulin polymerization. This is negatively regulated by glycogen synthase kinase-3b (GSK-3b), which phosphorylates and inactivates it, rendering it unable to interact with tubulin (Yoshimura et al. 2005). The same study demonstrated that neurotrophin-3 (NT-3) and brain-derived neurotrophic factor (BDNF) promote axon outgrowth by PI3K/Akt-mediated inactivation of GSK-3b. Considering the findings that the PI3K-Akt-mTOR pathway is required for NED of CaP (Wu and Huang 2007), DPYSL2/hCRMP-2 might represent a molecular link to the observed phenotypic changes.

6.3

Apoptosis-Regulatory Pathways

As suggested earlier, a critical obstacle to the success of androgen-deprivation therapy in curing prostate cancer outright is that although marked growth arrest and tumor regression are consistently achieved, a significant subpopulation of CaP cells survive. LNCaP serves as an excellent model for studying this surviving fraction since apoptosis is not induced by androgen withdrawal. The resulting expression profile provides insight into this phenomenon since approximately 20 genes that function in the regulation of apoptosis were differentially expressed and 80% of the androgen withdrawal-induced transcripts (i.e., 12 of 15) encode antiapoptotic proteins. At the same time, several proapoptotic genes are induced, suggesting that the balance between opposing pro- and antiapoptotic actions is still critical. This is exemplified by the ability to potently induce apoptosis of androgen-deprived LNCaP cells by elimination of antiapoptotic PI3K-Akt signaling (Murillo et al. 2001). Consistent with this, several of the androgen withdrawalmodulated genes can be functionally linked into a signaling network that integrates PI3K-Akt. Several secreted proteins induced by androgen withdrawal play critical roles in paracrine/autocrine survival signaling via pleieotropic mechanisms, including PI3-K activation. These include clusterin (CLU)/ TRPM2, relaxin H2 (RLN2), and MDK. Importantly, the induction of clusterin and relaxin expression by androgen ablation has been observed in clinical CaP samples and associated with tumor progression to androgen independence and bone metastasis (July et al. 2002; Thompson et al. 2006). In the case of clusterin, staining intensity increased 2.0–2.55-fold and percentage staining rose from 23.24% to as much as 63.83% in CaP samples obtained from patients treated with neoadjuvant hormonal therapy compared to those from untreated patients. Consistent with a prosurvival role for clusterin, treatment of mice with specific antisense oligonucleotides (ASOs) accelerated castration-induced regression, increased latency of recurrence, and enhanced chemosensitivity to paclitaxel (Gleave et al. 2001). Recent data demonstrate that clusterin elicits an antiapoptotic signal via its putative receptor, megalin, and downstream activation of PI3K/Akt, which in turn phosphorylates and inactivates Bad. Relaxin H2 signals via relaxin/insulin-like family peptide receptor 1 (RXFP1)/ LGR7, a G protein-coupled receptor, and has similarly been implicated as a

Cellular and Molecular Signatures of Androgen Ablation of Prostate Cancer

533

paracrine/autocrine tumor progression factor. This was demonstrated by the ability of purified relaxin, as well as conditioned medium derived from androgen-independent, p53 mutant-bearing cells overexpressing relaxin, to support growth of LNCaP cells in androgen-deprived medium (Vinall et al. 2006). Relaxin mediates this by an AR-dependent mechanism, which relies on an Akt-mediated increase in b-catenin stability and subsequent association with the AR transcription complex (Liu et al. 2008). Through a similar mechanism, it is quite possible that relaxin can simultaneously modulate NED by regulating the function of Tcf/Lef transcription factors. By interacting with TLE1 (described earlier), Tcf/Lef will function as transcriptional repressors. However, b-catenin can displace TLE1 from binding, thereby converting them to transcriptional activators (Daniels and Weis 2005). Midkine is highly expressed in high-grade CaP (Gleason score 8–10) (You et al. 2008) and is associated with the metastatic phenotype of LNCaP-C4-2 cells (Trojan et al. 2005). As described earlier, midkine is a neurotrophic factor and promotes survival via PI3K/Akt signaling (Tong et al. 2007; Owada et al. 1999). In addition to androgen withdrawal, its expression is also induced by EGF, IGF-I, hepatocyte growth factor, and the cytokines TNF-a and IL-1b through the NF-kB pathway (You et al. 2008). The mechanism responsible for mediating apoptosis induced by PI3K inhibitors has not been well defined. However, an excellent candidate for the androgen withdrawalinduced trigger is represented by the proapoptotic secreted glycoprotein insulin-like growth factor binding protein 3 (IGFBP3). In our study, IGFBP3 was the sixth most highly up-regulated gene (14.03-fold) and was very highly expressed. These findings are consistent with those reported from other investigators using LNCaP and the androgen-independent LNCaP-C4-2 subline (Kojima et al. 2006). Of note, serum IGFBP3 levels were increased in patients receiving bicalutamide monotherapy (Boccardo et al. 2006). Although it exhibits an androgen-repressed expression pattern, growth-inhibitory concentrations of DHT (i.e., 10 nM) can also induce IGFBP3 through an ARE in its promoter (Peng et al. 2006). Based upon its ability to mediate growth arrest and induce apoptosis, IGFBP3 is considered a tumor suppressor. In addition to having the capacity to induce apoptosis independently, it can function as a mediator of the effects of apoptosis-inducing stimuli, such as TGF-b (Rajah et al. 1997), and sensitize cells to apoptosis by IFN-g (Fang et al. 2008). Although it was originally described to bind and sequester insulin-like growth factor (IGF)-I and -II, it is now clear that IGFBP3 can mediate its effects via IGF-dependent and -independent mechanisms; it can induce apoptosis of an IGFR-negative cell line, and radiolabeled IGFBP3 binds to a variety of proteins (Rajah et al. 1997). IGFBP3 can be internalized and bind to the retinoid X receptora (RXRa) (Liu et al. 2000), which interacts with the Nur77/NR4A1 orphan receptor. As a consequence, Nur77 is translocated to the mitochondria and induces the intrinsic pathway of apoptosis (Lee et al. 2005) by inducing a conformational change in Bcl-2 that converts it into a ‘‘killer’’ (Lin et al. 2004). Furthermore, IGFBP3 can decrease Akt phosphorylation and increase stress-activated protein kinase JNK activity (Lee et al. 2007). The expression changes of several intracellular proteins demonstrate connections to PI3K signaling, but also reveal novel survival mechanisms that challenge

534

C.G. Tepper, H.-J. Kung

existing paradigms, or at least prompt consideration for known signaling intermediates performing dual, contrasting roles in regulating apoptosis. As discussed earlier, mTOR is critical to CaP biology and integrates androgen-AR and growth factor/nutrient signals. A key antiapoptotic effector downstream of Akt-mTOR is hypoxia-inducible factor (HIF-1a) (Majumder et al. 2004). This was elucidated using a transgenic mouse model for prostatic intraepithelial neoplasia (PIN) based upon probasin promoter-driven expression of Akt (AKT1-Tg). Treatment of AKT1Tg mice with the mTOR inhibitor RAD001 diminished expression of HIF-1a targets and induced apoptosis. It is important to note that in addition to its protective effects, HIF-1a also induces the expression of several proapoptotic genes, including BCL2/adenovirus E1B 19-kDa interacting protein 3 (BNIP3), BNIP3L, and DNAdamage-inducible transcript 4 (DDIT4), all of which are down-regulated in response to androgen withdrawal in which we see hyperactivation of Akt, but markedly decreased mTOR activity. This finding is also entirely consistent with the results found by Majumder, P. K. et al. (Majumder et al. 2004) in that the highest expression of BNIP3 is found in the tumors of untreated AKT1-Tg mice (Akt high, mTOR high), but greatly reduced after RAD001 treatment (Akt high, mTOR low). In addition to BNIP3 down-regulation resulting from androgenmTOR-associated HIF-1a down-regulation, the androgen withdrawal profile revealed several other potential mechanisms. First, elevated expression of FOXO3 and CBP/p300-interacting transactivator with E/D-rich C-terminal domain (CITED2) transcripts demonstrates the operation of the HIF-1a negative feedback loop that is necessary for survival (Bakker et al. 2007a, b). FOXO3 transcripts are induced by hypoxia (Bakker et al. 2007a) and CITED2 is a FOXO3 target gene (Bakker et al. 2007b). Importantly, the 35-kDa p300/CBPbinding protein CITED2 interacts with and directly represses HIF-1a and the expression of BNIP3 and DDIT4 (Bakker et al. 2007a). While it is paradoxical that the gene expression changes observed still occur in the context of HIF-1a and FOXO3 repression by CITED and Akt, respectively, we can propose that their activities are attenuated, but not abolished, since a HIF-1a coactivator aryl-hydrocarbon receptor nuclear translocator 2 (ARNT2) (Maltepe et al. 2000) is markedly elevated and Akt-mediated FOXO3 phosphorylation does not completely eliminate it from nuclear translocation. Finally, any BNIP3 protein that is expressed would potentially be neutralized by acetyl-Coenzyme A acyltransferase 2 (ACAA2) (3.67fold induced), which interacts and colocalizes with BNIP3 in the mitochondria and antagonizes its apoptotic effects (Cao et al. 2008).

7 Concluding Remarks Androgen ablation is the standard of care for the treatment of metastatic CaP. Although it dependably mediates tumor regression, it is only palliative, and the disease typically recurs as castration-resistant prostate cancer. Reinstatement of the AR signaling pathway represents one of the most prominent mechanisms driving recurrence. Accordingly, contemporary approaches to achieving complete suppression of androgen-AR signaling include the utilization of agents to eliminate

Cellular and Molecular Signatures of Androgen Ablation of Prostate Cancer

535

intraprostatic DHT and the development of novel SARMs. As discussed earlier, a biological explanation for recurrence is the ability of a subpopulation of CaP cells to evade apoptosis when challenged by androgen ablation. This is further complicated by the occurrence of neuroendocrine trans-differentiation, which engenders multiple biological and molecular properties that can facilitate the transition of CaP cells to ablation resistance. In an optimistic tone, the efficacy of androgen ablation can be augmented by adjuvant targeting of the survival mechanisms during this transition phase. Indeed, antisense oligodeoxynucleotides developed against clusterin, Hsp27, and Bcl-2 are in various phases of clinical trials (Hadaschik and Gleave 2007). The purpose of this chapter was to discuss the molecular mechanisms that mediate the cellular responses to androgen ablation, with particular focus upon AR regulation, NED, and survival signaling. In order to gain additional insight these processes, we also present the results of microarray gene expression profiling and attempt to integrate the components of the androgen withdrawal expression signature into relevant signaling pathways, including those associated with AR and PI3K-Akt. Since the androgen withdrawal signature contained both anti- and proapoptotic genes, it is apparent that CaP cell survival during androgen ablation is determined by the outcome of their interaction. Although the expression profiles cannot explain the basis for all biological phenomena observed, expression signatures have been valuable tools in guiding the development of therapeutics. For instance, using the expression pattern derived from cells subjected to hormone depletion or treated with Casodex as a template, it was found that ablation of AR signaling can also be achieved through the use of Hsp90 inhibitors and was subsequently utilized to identify similarly acting molecules (Lamb et al. 2006; Hieronymus et al. 2006). This illuminates the concept that agents inducing AR degradation are of significant value in CaP therapy in that they represent a nonredundant, complementary approach to suppressing AR function. Along these lines, a variety of nutriceuticals, such as genistein combined polysaccharide (GCP), are generating excitement since they induce apoptosis by having dual effects as a mimetic of androgen ablation and inhibitor of survival kinase signaling, by triggering AR degradation and acting as a potent mTOR inhibitor (Tepper et al. 2007). In closing, intense research efforts and technological advancements have substantially increased our understanding of mechanisms underlying CaP persistence during androgen ablation. As a result, novel therapeutic strategies are now being developed and are beginning to be translated to the clinic.

References Agus, D.B., Cordon-Cardo, C., Fox, W., Drobnjak, M., Koff, A., Golde, D.W. & Scher, H.I. Prostate cancer cell cycle regulators: response to androgen withdrawal and development of androgen independence. Journal of the National Cancer Institute 91, 1869–1876 (1999). Ahuja, R., Pinyol, R., Reichenbach, N., Custer, L., Klingensmith, J., Kessels, M.M. & Qualmann, B. Cordon-bleu is an actin nucleation factor and controls neuronal morphology. Cell 131, 337–350 (2007).

536

C.G. Tepper, H.-J. Kung

Allen, T., van Tuyl, M., Iyengar, P., Jothy, S., Post, M., Tsao, M.S. & Lobe, C.G. Grg1 acts as a lung-specific oncogene in a transgenic mouse model. Cancer Research 66, 1294–1301 (2006). Amler, L.C., Agus, D.B., LeDuc, C., Sapinoso, M.L., Fox, W.D., Kern, S., Lee, D., Wang, V., Leysens, M., Higgins, B., Martin, J., Gerald, W., Dracopoli, N., Cordon-Cardo, C., Scher, H.I. & Hampton, G.M. Dysregulated expression of androgen-responsive and nonresponsive genes in the androgen-independent prostate cancer xenograft model CWR22-R1. Cancer Research 60, 6134–6141 (2000). Assou, S., Le Carrour, T., Tondeur, S., Strom, S., Gabelle, A., Marty, S., Nadal, L., Pantesco, V., Reme, T., Hugnot, J.P., Gasca, S., Hovatta, O., Hamamah, S., Klein, B. & De Vos, J. A metaanalysis of human embryonic stem cells transcriptome integrated into a web-based expression atlas. Stem Cells 25, 961–973 (2007). Baarends, W.M., Themmen, A.P., Blok, L.J., Mackenbach, P., Brinkmann, A.O., Meijer, D., Faber, P.W., Trapman, J. & Grootegoed, J.A. The rat androgen receptor gene promoter. Molecular and Cellular Endocrinology 74, 75–84 (1990). Bakker, W.J., Harris, I.S. & Mak, T.W. FOXO3a is activated in response to hypoxic stress and inhibits HIF1-induced apoptosis via regulation of CITED2. Molecular Cell 28, 941–953 (2007a). Bakker, W.J., van Dijk, T.B., Parren-van Amelsvoort, M., Kolbus, A., Yamamoto, K., Steinlein, P., Verhaak, R.G., Mak, T.W., Beug, H., Lowenberg, B. & von Lindern, M. Differential regulation of Foxo3a target genes in erythropoiesis. Molecular and Cellular Biology 27, 3839–3854 (2007b). Ballinger, C.A., Connell, P., Wu, Y., Hu, Z., Thompson, L.J., Yin, L.Y. & Patterson, C. Identification of CHIP, a novel tetratricopeptide repeat-containing protein that interacts with heat shock proteins and negatively regulates chaperone functions. Molecular and Cellular Biology 19, 4535–4545 (1999). Banerjee, P.P., Banerjee, S., Tilly, K.I., Tilly, J.L., Brown, T.R. & Zirkin, B.R. Lobe-specific apoptotic cell death in rat prostate after androgen ablation by castration. Endocrinology 136, 4368–4376 (1995). Banerjee, S., Banerjee, P.P. & Brown, T.R. Castration-induced apoptotic cell death in the Brown Norway rat prostate decreases as a function of age. Endocrinology 141, 821–832 (2000). Banerjee, P.P., Banerjee, S. & Brown, T.R. Bcl-2 protein expression correlates with cell survival and androgen independence in rat prostatic lobes. Endocrinology 143, 1825–1832 (2002). Bang, Y.J., Pirnia, F., Fang, W.G., Kang, W.K., Sartor, O., Whitesell, L., Ha, M.J., Tsokos, M., Sheahan, M.D., Nguyen, P., Niklinski, W.T., Myers, C.E. & Trepel, J.B. Terminal neuroendocrine differentiation of human prostate carcinoma cells in response to increased intracellular cyclic AMP. Proceedings of the National Academy of Sciences of the United States of America 91, 5330–5334 (1994). Barbaro, V., Testa, A., Di Iorio, E., Mavilio, F., Pellegrini, G. & De Luca, M. C/EBPdelta regulates cell cycle and self-renewal of human limbal stem cells. The Journal of Cell Biology 177, 1037–1049 (2007). Belanger, A., Hum, D.W., Beaulieu, M., Levesque, E., Guillemette, C., Tchernof, A., Belanger, G., Turgeon, D. & Dubois, S. Characterization and regulation of UDP-glucuronosyltransferases in steroid target tissues. The Journal of Steroid Biochemistry and Molecular Biology 65, 301–310 (1998). Berchem, G.J., Bosseler, M., Sugars, L.Y., Voeller, H.J., Zeitlin, S. & Gelmann, E.P. Androgens induce resistance to bcl-2-mediated apoptosis in LNCaP prostate cancer cells. Cancer Research 55, 735–738 (1995). Berges, R.R., Furuya, Y., Remington, L., English, H.F., Jacks, T. & Isaacs, J.T. Cell proliferation, DNA repair, and p53 function are not required for programmed death of prostatic glandular cells induced by androgen ablation. Proceedings of the National Academy of Sciences of the United States of America 90, 8910–8914 (1993). Berquin, I.M., Min, Y., Wu, R., Wu, H. & Chen, Y.Q. Expression signature of the mouse prostate. The Journal of Biological Chemistry 280, 36442–36451 (2005).

Cellular and Molecular Signatures of Androgen Ablation of Prostate Cancer

537

Bhatia-Gaur, R., Donjacour, A.A., Sciavolino, P.J., Kim, M., Desai, N., Young, P., Norton, C.R., Gridley, T., Cardiff, R.D., Cunha, G.R., Abate-Shen, C. & Shen, M.M. Roles for Nkx3.1 in prostate development and cancer. Genes Development 13, 966–977 (1999). Bieberich, C.J., Fujita, K., He, W.W. & Jay, G. Prostate-specific and androgen-dependent expression of a novel homeobox gene. The Journal of Biological Chemistry 271, 31779–31782 (1996). Boccardo, F., Rubagotti, A., Battaglia, M., Zattoni, F., Bertaccini, A., Romagnoli, A. & Conti, G. Influence of bicalutamide with or without tamoxifen or anastrozole on insulin-like growth factor 1 and binding proteins in prostate cancer patients. International Journal of Biological Markers 21, 123–126 (2006). Bowen, C., Bubendorf, L., Voeller, H.J., Slack, R., Willi, N., Sauter, G., Gasser, T.C., Koivisto, P., Lack, E.E., Kononen, J., Kallioniemi, O.P. & Gelmann, E.P. Loss of NKX3.1 expression in human prostate cancers correlates with tumor progression. Cancer Research 60, 6111–6115 (2000). Brandstrom, A., Westin, P., Bergh, A., Cajander, S. & Damber, J.E. Castration induces apoptosis in the ventral prostate but not in an androgen-sensitive prostatic adenocarcinoma in the rat. Cancer Research 54, 3594–3601 (1994). Brodin, G., ten Dijke, P., Funa, K., Heldin, C.H. & Landstrom, M. Increased smad expression and activation are associated with apoptosis in normal and malignant prostate after castration. Cancer Research 59, 2731–2738 (1999). Bubley, G.J., Carducci, M., Dahut, W., Dawson, N., Daliani, D., Eisenberger, M., Figg, W.D., Freidlin, B., Halabi, S., Hudes, G., Hussain, M., Kaplan, R., Myers, C., Oh, W., Petrylak, D.P., Reed, E., Roth, B., Sartor, O., Scher, H., Simons, J., Sinibaldi, V., Small, E.J., Smith, M.R., Trump, D.L., Wilding, G. & et alet al. Eligibility and response guidelines for phase II clinical trials in androgen-independent prostate cancer: recommendations from the Prostate-Specific Antigen Working Group. Journal of Clinical Oncology 17, 3461–3467 (1999). Bubulya, A., Wise, S.C., Shen, X.Q., Burmeister, L.A. & Shemshedini, L. c-Jun can mediate androgen receptor-induced transactivation. The Journal of Biological Chemistry 271, 24583– 24589 (1996). Bubulya, A., Chen, S.Y., Fisher, C.J., Zheng, Z., Shen, X.Q. & Shemshedini, L. c-Jun potentiates the functional interaction between the amino and carboxyl termini of the androgen receptor. The Journal of Biological Chemistry 276, 44704–44711 (2001). Burchardt, T., Burchardt, M., Chen, M.W., Cao, Y., de la Taille, A., Shabsigh, A., Hayek, O., Dorai, T. & Buttyan, R. Transdifferentiation of prostate cancer cells to a neuroendocrine cell phenotype in vitro and in vivo. The Journal of Urology 162, 1800–1805 (1999). Camandola, S. & Mattson, M.P. Pro-apoptotic action of PAR-4 involves inhibition of NF-kappaB activity and suppression of BCL-2 expression. Journal of Neuroscience Research 61, 134–139 (2000). Cao, W., Liu, N., Tang, S., Bao, L., Shen, L., Yuan, H., Zhao, X. & Lu, H. Acetyl-Coenzyme A acyltransferase 2 attenuates the apoptotic effects of BNIP3 in two human cell lines. Biochimica et Biophysica Acta (2008). Carroll, E.A., Gerrelli, D., Gasca, S., Berg, E., Beier, D.R., Copp, A.J. & Klingensmith, J. Cordonbleu is a conserved gene involved in neural tube formation. Developmental Biology 262, 16–31 (2003). Catz, S.D. & Johnson, J.L. Transcriptional regulation of bcl-2 by nuclear factor kappa B and its significance in prostate cancer. Oncogene 20, 7342–7351 (2001). Chadli, A., Bouhouche, I., Sullivan, W., Stensgard, B., McMahon, N., Catelli, M.G. & Toft, D.O. Dimerization and N-terminal domain proximity underlie the function of the molecular chaperone heat shock protein 90. Proceedings of the National Academy of Sciences of the United States of America 97, 12524–12529 (2000). Cheema, S.K., Mishra, S.K., Rangnekar, V.M., Tari, A.M., Kumar, R. & Lopez-Berestein, G. Par4 transcriptionally regulates Bcl-2 through a WT1-binding site on the bcl-2 promoter. The Journal of Biological Chemistry 278, 19995–20005 (2003).

538

C.G. Tepper, H.-J. Kung

Chen, C.D., Welsbie, D.S., Tran, C., Baek, S.H., Chen, R., Vessella, R., Rosenfeld, M.G. & Sawyers, C.L. Molecular determinants of resistance to antiandrogen therapy. Nature Medicine 10, 33–39 (2004). Chendil, D., Das, A., Dey, S., Mohiuddin, M. & Ahmed, M.M. Par-4, a pro-apoptotic gene, inhibits radiation-induced NF kappa B activity and Bcl-2 expression leading to induction of radiosensitivity in human prostate cancer cells PC-3. Cancer Biology and Therapy 1, 152–160 (2002). Chen, S., Sullivan, W.P., Toft, D.O. & Smith, D.F. Differential interactions of p23 and the TPRcontaining proteins Hop, Cyp40, FKBP52 and FKBP51 with Hsp90 mutants. Cell Stress Chaperones 3, 118–129 (1998). Chen, X., Thakkar, H., Tyan, F., Gim, S., Robinson, H., Lee, C., Pandey, S.K., Nwokorie, C., Onwudiwe, N. & Srivastava, R.K. Constitutively active Akt is an important regulator of TRAIL sensitivity in prostate cancer. Oncogene 20, 6073–6083 (2001). Chiao, J.W., Hsieh, T.C., Xu, W., Sklarew, R.J. & Kancherla, R. Development of human prostate cancer cells to neuroendocrine-like cells by interleukin-1. International Journal of Oncology 15, 1033–1037 (1999). Chmelar, R., Buchanan, G., Need, E.F., Tilley, W. & Greenberg, N.M. Androgen receptor coregulators and their involvement in the development and progression of prostate cancer. International Journal of cancer 120, 719–733 (2007). Chouinard, S., Barbier, O. & Belanger, A. UDP-glucuronosyltransferase 2B15 (UGT2B15) and UGT2B17 enzymes are major determinants of the androgen response in prostate cancer LNCaP cells. The Journal of Biological Chemistry 282, 33466–33474 (2007). Chung, B.C., Picado-Leonard, J., Haniu, M., Bienkowski, M., Hall, P.F., Shively, J.E. & Miller, W.L. Cytochrome P450c17 (steroid 17 alpha-hydroxylase/17,20 lyase): cloning of human adrenal and testis cDNAs indicates the same gene is expressed in both tissues. Proceedings of the National Academy of Sciences of the United States of America 84, 407–411 (1987). Cinar, B., De Benedetti, A. & Freeman, M.R. Post-transcriptional regulation of the androgen receptor by Mammalian target of rapamycin. Cancer Research 65, 2547–2553 (2005). Cockshott, I.D., Cooper, K.J., Sweetmore, D.S., Blacklock, N.J. & Denis, L. The pharmacokinetics of Casodex in prostate cancer patients after single and during multiple dosing. European Urology 18 Suppl 3, 10–17 (1990). Connell, P., Ballinger, C.A., Jiang, J., Wu, Y., Thompson, L.J., Hohfeld, J. & Patterson, C. The cochaperone CHIP regulates protein triage decisions mediated by heat-shock proteins. Nature Cell Biology 3, 93–96 (2001). Cornforth, A.N., Davis, J.S., Khanifar, E., Nastiuk, K.L., & Krolewski, J.J. FOXO3a mediates the androgen-dependent regulation of FLIP and contributes to TRAIL-induced apoptosis of LNCaP cells. Oncogene (2008). Cox, M.E., Deeble, P.D., Bissonette, E.A. & Parsons, S.J. Activated 30 ,50 -cyclic AMP-dependent protein kinase is sufficient to induce neuroendocrine-like differentiation of the LNCaP prostate tumor cell line. The Journal of biological chemistry 275, 13812–13818 (2000). Daniels, D.L. & Weis, W.I. Beta-catenin directly displaces Groucho/TLE repressors from Tcf/Lef in Wnt-mediated transcription activation. Nature Structural & Molecular Biology 12, 364–371 (2005). De Coster, R., Wouters, W. & Bruynseels, J. P450-dependent enzymes as targets for prostate cancer therapy. The Journal of Steroid Biochemistry and Molecular Biology 56, 133–143 (1996). Dehni, G., Liu, Y., Husain, J. & Stifani, S. TLE expression correlates with mouse embryonic segmentation, neurogenesis, and epithelial determination. Mechanisms of Development 53, 369–381 (1995). de la Taille, A., Chen, M.W., Shabsigh, A., Bagiella, E., Kiss, A. & Buttyan, R. Fas antigen/CD-95 upregulation and activation during castration-induced regression of the rat ventral prostate gland. The Prostate 40, 89–96 (1999).

Cellular and Molecular Signatures of Androgen Ablation of Prostate Cancer

539

Desai, K.V., Michalowska, A.M., Kondaiah, P., Ward, J.M., Shih, J.H. & Green, J.E. Gene expression profiling identifies a unique androgen-mediated inflammatory/immune signature and a PTEN (phosphatase and tensin homolog deleted on chromosome 10)-mediated apoptotic response specific to the rat ventral prostate. Molecular Endocrinology (Baltimore, Md 18, 2895–2907 (2004). Desai, S.J., Tepper, C.G. & Kung, H.J. Neuroendocrine differentiation and androgen independence in prostate cancer. in Prostate Cancer: Basic Mechanisms and Therapeutic Approaches (ed. Chang, C.) 157–190 (World Scientific, Singapore, 2005). di Sant’Agnese, P.A. Neuroendocrine differentiation in human prostatic carcinoma. Human Pathology 23, 287–296 (1992). Doll, J.A., Stellmach, V.M., Bouck, N.P., Bergh, A.R., Lee, C., Abramson, L.P., Cornwell, M.L., Pins, M.R., Borensztajn, J. & Crawford, S.E. Pigment epithelium-derived factor regulates the vasculature and mass of the prostate and pancreas. Nature Medicine 9, 774–780 (2003). Doniger, S.W., Salomonis, N., Dahlquist, K.D., Vranizan, K., Lawlor, S.C. & Conklin, B.R. MAPPFinder: using Gene Ontology and GenMAPP to create a global gene-expression profile from microarray data. Genome Biology 4, R7 (2003). Faber, P.W., van Rooij, H.C., Schipper, H.J., Brinkmann, A.O. & Trapman, J. Two different, overlapping pathways of transcription initiation are active on the TATA-less human androgen receptor promoter. The role of Sp1. The Journal of Biological Chemistry 268, 9296–9301 (1993). Fang, P., Hwa, V., Little, B.M. & Rosenfeld, R.G. IGFBP-3 sensitizes prostate cancer cells to interferon-gamma-induced apoptosis. Growth Hormone and IGF Research 18, 38–46 (2008). Fan, W., Yanase, T., Morinaga, H., Okabe, T., Nomura, M., Daitoku, H., Fukamizu, A., Kato, S., Takayanagi, R. & Nawata, H. Insulin-like growth factor 1/insulin signaling activates androgen signaling through direct interactions of Foxo1 with androgen receptor. The Journal of Biological Chemistry 282, 7329–7338 (2007). Febbo, P.G., Lowenberg, M., Thorner, A.R., Brown, M., Loda, M. & Golub, T.R. Androgen mediated regulation and functional implications of fkbp51 expression in prostate cancer. The Journal of Urology 173, 1772–1777 (2005). Franck-Lissbrant, I., Haggstrom, S., Damber, J.E. & Bergh, A. Testosterone stimulates angiogenesis and vascular regrowth in the ventral prostate in castrated adult rats. Endocrinology 139, 451–456 (1998). Frigo, D.E. & McDonnell, D.P. Differential effects of prostate cancer therapeutics on neuroendocrine transdifferentiation. Molecular Cancer Therapeutics 7, 659–669 (2008). Fujitani, M., Yamagishi, S., Che, Y.H., Hata, K., Kubo, T., Ino, H., Tohyama, M. & Yamashita, T. P311 accelerates nerve regeneration of the axotomized facial nerve. Journal of Neurochemistry 91, 737–744 (2004). Furutani, T., Watanabe, T., Tanimoto, K., Hashimoto, T., Koutoku, H., Kudoh, M., Shimizu, Y., Kato, S. & Shikama, H. Stabilization of androgen receptor protein is induced by agonist, not by antagonists. Biochemical and Biophysical Research Communications 294, 779–784 (2002). Garnick, M.B. Leuprolide versus diethylstilbestrol for previously untreated stage D2 prostate cancer. Results of a prospectively randomized trial. Urology 27, 21–28 (1986). Gaubert, C.M., Tremblay, R.R. & Dube, J.Y. Effect of sodium molybdate on cytosolic androgen receptors in rat prostate. Journal of Steroid Biochemistry 13, 931–937 (1980). Geller, J. Rationale for blockade of adrenal as well as testicular androgens in the treatment of advanced prostate cancer. Seminars in Oncology 12, 28–35 (1985). Georget, V., Terouanne, B., Nicolas, J.C. & Sultan, C. Mechanism of antiandrogen action: key role of hsp90 in conformational change and transcriptional activity of the androgen receptor. Biochemistry 41, 11824–11831 (2002). Gleave, M.E., Miyake, H., Zellweger, T., Chi, K., July, L., Nelson, C. & Rennie, P. Use of antisense oligonucleotides targeting the antiapoptotic gene, clusterin/testosterone-repressed prostate message 2, to enhance androgen sensitivity and chemosensitivity in prostate cancer. Urology 58, 39–49 (2001).

540

C.G. Tepper, H.-J. Kung

Goswami, A., Burikhanov, R., de Thonel, A., Fujita, N., Goswami, M., Zhao, Y., Eriksson, J.E., Tsuruo, T. & Rangnekar, V.M. Binding and phosphorylation of par-4 by akt is essential for cancer cell survival. Molecular Cell 20, 33–44 (2005). Graff, J.R., Konicek, B.W., McNulty, A.M., Wang, Z., Houck, K., Allen, S., Paul, J.D., Hbaiu, A., Goode, R.G., Sandusky, G.E., Vessella, R.L. & Neubauer, B.L. Increased AKT activity contributes to prostate cancer progression by dramatically accelerating prostate tumor growth and diminishing p27Kip1 expression. The Journal of Biological Chemistry 275, 24500–24505 (2000). Grbavec, D. & Stifani, S. Molecular interaction between TLE1 and the carboxyl-terminal domain of HES-1 containing the WRPW motif. Biochemical and Biophysical Research Communications 223, 701–705 (1996). Gregory, C.W., Johnson, R.T., Jr., Mohler, J.L., French, F.S. & Wilson, E.M. Androgen receptor stabilization in recurrent prostate cancer is associated with hypersensitivity to low androgen. Cancer Research 61, 2892–2898 (2001). Grignon, D.J. & Sakr, W.A. Zonal origin of prostatic adenocarcinoma: are there biologic differences between transition zone and peripheral zone adenocarcinomas of the prostate gland? Journal of Cellular Biochemistry. Supplement 19, 267–269 (1994). Grigoryev, D.N., Long, B.J., Nnane, I.P., Njar, V.C., Liu, Y. & Brodie, A.M. Effects of new 17alpha-hydroxylase/C(17,20)-lyase inhibitors on LNCaP prostate cancer cell growth in vitro and in vivo. British Journal of Cancer 81, 622–630 (1999). Guillemette, C., Levesque, E., Beaulieu, M., Turgeon, D., Hum, D.W. & Belanger, A. Differential regulation of two uridine diphospho-glucuronosyltransferases, UGT2B15 and UGT2B17, in human prostate LNCaP cells. Endocrinology 138, 2998–3005 (1997). Gutman, A.B. & Gutman, E.B. An ‘‘ Acid ’’ Phosphatase Occurring in the Serum of Patients with Metastasizing Carcinoma of the Prostate Gland. The Journal of Clinical Investigation 17, 473– 478 (1938). Hadaschik, B.A. & Gleave, M.E. Therapeutic options for hormone-refractory prostate cancer in 2007. Urologic Oncology 25, 413–419 (2007). Harrington, L.S., Findlay, G.M., Gray, A., Tolkacheva, T., Wigfield, S., Rebholz, H., Barnett, J., Leslie, N.R., Cheng, S., Shepherd, P.R., Gout, I., Downes, C.P. & Lamb, R.F. The TSC1– 2 tumor suppressor controls insulin-PI3K signaling via regulation of IRS proteins. The Journal of Cell Biology 166, 213–223 (2004). He, B., Bai, S., Hnat, A.T., Kalman, R.I., Minges, J.T., Patterson, C. & Wilson, E.M. An androgen receptor NH2-terminal conserved motif interacts with the COOH terminus of the Hsp70interacting protein (CHIP). The Journal of Biological Chemistry 279, 30643–30653 (2004). Hieronymus, H., Lamb, J., Ross, K.N., Peng, X.P., Clement, C., Rodina, A., Nieto, M., Du, J., Stegmaier, K., Raj, S.M., Maloney, K.N., Clardy, J., Hahn, W.C., Chiosis, G. & Golub, T.R. Gene expression signature-based chemical genomic prediction identifies a novel class of HSP90 pathway modulators. Cancer cell 10, 321–330 (2006). Hirano, D., Okada, Y., Minei, S., Takimoto, Y. & Nemoto, N. Neuroendocrine differentiation in hormone refractory prostate cancer following androgen deprivation therapy. European Urology 45, 586–592; discussion 592 (2004). Hitchins, M.P., Bentley, L., Monk, D., Beechey, C., Peters, J., Kelsey, G., Ishino, F., Preece, M.A., Stanier, P. & Moore, G.E. DDC and COBL, flanking the imprinted GRB10 gene on 7p12, are biallelically expressed. Mammalian Genome 13, 686–691 (2002). Hodgson, M.C., Astapova, I., Hollenberg, A.N. & Balk, S.P. Activity of androgen receptor antagonist bicalutamide in prostate cancer cells is independent of NCoR and SMRT corepressors. Cancer Research 67, 8388–8395 (2007). Hsieh, J.T., Wu, H.C., Gleave, M.E., von Eschenbach, A.C. & Chung, L.W. Autocrine regulation of prostate-specific antigen gene expression in a human prostatic cancer (LNCaP) subline. Cancer Research 53, 2852–2857 (1993). Huang, C.Y., Beliakoff, J., Li, X., Lee, J., Sharma, M., Lim, B. & Sun, Z. hZimp7, a novel PIASlike protein, enhances androgen receptor-mediated transcription and interacts with SWI/SNFlike BAF complexes. Molecular Endocrinology (Baltimore, Md) 19, 2915–2929 (2005).

Cellular and Molecular Signatures of Androgen Ablation of Prostate Cancer

541

Huggins, C. & Hodges, C.V. Studies of prostatic cancer: I. Effect of castration, estrogen and androgen injections on serum phosphatases in metastatic carcinoma of the prostate. Cancer Research 1, 293–307 (1941). Huggins, C., Stevens, R.E. & Hodges, C.V. Studies on prostate cancer: II. The effect of castration on advanced carcinoma of the prostate gland. Archives of Surgery 43, 209 (1941). Huss, W.J., Gregory, C.W. & Smith, G.J. Neuroendocrine cell differentiation in the CWR22 human prostate cancer xenograft: association with tumor cell proliferation prior to recurrence. The Prostate 60, 91–97 (2004). Ito, T., Yamamoto, S., Ohno, Y., Namiki, K., Aizawa, T., Akiyama, A. & Tachibana, M. Upregulation of neuroendocrine differentiation in prostate cancer after androgen deprivation therapy, degree and androgen independence. Oncology Reports 8, 1221–1224 (2001). Jarriault, S., Brou, C., Logeat, F., Schroeter, E.H., Kopan, R. & Israel, A. Signalling downstream of activated mammalian Notch. Nature 377, 355–358 (1995). Jiang, J., Ballinger, C.A., Wu, Y., Dai, Q., Cyr, D.M., Hohfeld, J. & Patterson, C. CHIP is a U-boxdependent E3 ubiquitin ligase: identification of Hsc70 as a target for ubiquitylation. The Journal of Biological Chemistry 276, 42938–42944 (2001). Johansson, A., Jones, J., Pietras, K., Kilter, S., Skytt, A., Rudolfsson, S.H. & Bergh, A. A stroma targeted therapy enhances castration effects in a transplantable rat prostate cancer model. The Prostate 67, 1664–1676 (2007). Johnson, J.L. & Toft, D.O. A novel chaperone complex for steroid receptors involving heat shock proteins, immunophilins, and p23. The Journal of Biological Chemistry 269, 24989–24993 (1994). Jongsma, J., Oomen, M.H., Noordzij, M.A., Van Weerden, W.M., Martens, G.J., van der Kwast, T. H., Schroder, F.H. & van Steenbrugge, G.J. Kinetics of neuroendocrine differentiation in an androgen-dependent human prostate xenograft model. The American Journal of Pathology 154, 543–551 (1999). Jongsma, J., Oomen, M.H., Noordzij, M.A., Van Weerden, W.M., Martens, G.J., van der Kwast, T. H., Schroder, F.H. & van Steenbrugge, G.J. Androgen deprivation of the PC-310 [correction of prohormone convertase-310] human prostate cancer model system induces neuroendocrine differentiation. Cancer Research 60, 741–748 (2000). Jongsma, J., Oomen, M.H., Noordzij, M.A., Van Weerden, W.M., Martens, G.J., van der Kwast, T. H., Schroder, F.H. & van Steenbrugge, G.J. Different profiles of neuroendocrine cell differentiation evolve in the PC-310 human prostate cancer model during long-term androgen deprivation. The Prostate 50, 203–215 (2002). Joseph, I.B., Nelson, J.B., Denmeade, S.R. & Isaacs, J.T. Androgens regulate vascular endothelial growth factor content in normal and malignant prostatic tissue. Clinical Cancer Research 3, 2507–2511 (1997). July, L.V., Akbari, M., Zellweger, T., Jones, E.C., Goldenberg, S.L. & Gleave, M.E. Clusterin expression is significantly enhanced in prostate cancer cells following androgen withdrawal therapy. The Prostate 50, 179–188 (2002). Kaneda, N., Talukder, A.H., Nishiyama, H., Koizumi, S. & Muramatsu, T. Midkine, a heparinbinding growth/differentiation factor, exhibits nerve cell adhesion and guidance activity for neurite outgrowth in vitro. Journal of Biochemistry 119, 1150–1156 (1996). Kawano, Y., Yoshimura, T., Tsuboi, D., Kawabata, S., Kaneko-Kawano, T., Shirataki, H., Takenawa, T. & Kaibuchi, K. CRMP-2 is involved in kinesin-1-dependent transport of the Sra-1/WAVE1 complex and axon formation. Molecular and Cellular Biology 25, 9920–9935 (2005). Kemppainen, J.A., Lane, M.V., Sar, M. & Wilson, E.M. Androgen receptor phosphorylation, turnover, nuclear transport, and transcriptional activation. Specificity for steroids and antihormones. The Journal of Biological Chemistry 267, 968–974 (1992). Kerjan, G., Dolan, J., Haumaitre, C., Schneider-Maunoury, S., Fujisawa, H., Mitchell, K.J. & Chedotal, A. The transmembrane semaphorin Sema6A controls cerebellar granule cell migration. Nature Neuroscience 8, 1516–1524 (2005).

542

C.G. Tepper, H.-J. Kung

Kerr, J.F. & Searle, J. Deletion of cells by apoptosis during castration-induced involution of the rat prostate. Virchows Archiv 13, 87–102 (1973). Kim, J., Adam, R.M. & Freeman, M.R. Activation of the Erk mitogen-activated protein kinase pathway stimulates neuroendocrine differentiation in LNCaP cells independently of cell cycle withdrawal and STAT3 phosphorylation. Cancer Research 62, 1549–1554 (2002). Kitagawa, Y., Dai, J., Zhang, J., Keller, J.M., Nor, J., Yao, Z. & Keller, E.T. Vascular endothelial growth factor contributes to prostate cancer-mediated osteoblastic activity. Cancer Research 65, 10921–10929 (2005). Kitamura, M., Buczko, E. & Dufau, M.L. Dissociation of hydroxylase and lyase activities by sitedirected mutagenesis of the rat P45017 alpha. 5, 1373–1380 (1991). Klotz, L. & Schellhammer, P. Combined androgen blockade: the case for bicalutamide. Clinical Prostate Cancer 3, 215–219 (2005). Kobayashi, Y., Miwa, S., Merry, D.E., Kume, A., Mei, L., Doyu, M. & Sobue, G. Caspase-3 cleaves the expanded androgen receptor protein of spinal and bulbar muscular atrophy in a polyglutamine repeat length-dependent manner. Biochemical and Biophysical Research Communications 252, 145–150 (1998). Kojima, S., Mulholland, D.J., Ettinger, S., Fazli, L., Nelson, C.C. & Gleave, M.E. Differential regulation of IGFBP-3 by the androgen receptor in the lineage-related androgen-dependent LNCaP and androgen-independent C4–2 prostate cancer models. The Prostate 66, 971–986 (2006). Kokontis, J.M., Hay, N. & Liao, S. Progression of LNCaP prostate tumor cells during androgen deprivation: hormone-independent growth, repression of proliferation by androgen, and role for p27Kip1 in androgen-induced cell cycle arrest. Molecular Endocrinology (Baltimore, Md) 12, 941–953 (1998). Kondapaka, S.B., Singh, S.S., Dasmahapatra, G.P., Sausville, E.A. & Roy, K.K. Perifosine, a novel alkylphospholipid, inhibits protein kinase B activation. Molecular Cancer Therapeutics 2, 1093–1103 (2003). Kreisberg, J.I., Malik, S.N., Prihoda, T.J., Bedolla, R.G., Troyer, D.A., Kreisberg, S. & Ghosh, P. M. Phosphorylation of Akt (Ser473) is an excellent predictor of poor clinical outcome in prostate cancer. Cancer Research 64, 5232–5236 (2004). Krongrad, A., Wilson, C.M., Wilson, J.D., Allman, D.R. & McPhaul, M.J. Androgen increases androgen receptor protein while decreasing receptor mRNA in LNCaP cells. Molecular and Cellular Endocrinology 76, 79–88 (1991). Kuhn, J.M., Billebaud, T., Navratil, H., Moulonguet, A., Fiet, J., Grise, P., Louis, J.F., Costa, P., Husson, J.M., Dahan, R. & et alet al. Prevention of the transient adverse effects of a gonadotropin-releasing hormone analogue (buserelin) in metastatic prostatic carcinoma by administration of an antiandrogen (nilutamide). The New England Journal of Medicine 321, 413–418 (1989). Kurita, T., Wang, Y.Z., Donjacour, A.A., Zhao, C., Lydon, J.P., O’Malley, B.W., Isaacs, J.T., Dahiya, R. & Cunha, G.R. Paracrine regulation of apoptosis by steroid hormones in the male and female reproductive system. Cell Death and Differentiation 8, 192–200 (2001). Labrie, F. Medical castration with LHRH agonists: 25 years later with major benefits achieved on survival in prostate cancer. Journal of Andrology 25, 305–313 (2004). Lamb, J., Crawford, E.D., Peck, D., Modell, J.W., Blat, I.C., Wrobel, M.J., Lerner, J., Brunet, J.P., Subramanian, A., Ross, K.N., Reich, M., Hieronymus, H., Wei, G., Armstrong, S.A., Haggarty, S.J., Clemons, P.A., Wei, R., Carr, S.A., Lander, E.S. & Golub, T.R. The Connectivity Map: using gene-expression signatures to connect small molecules, genes, and disease. Science 313, 1929–1935 (2006). Landstrom, M., Damber, J.E. & Bergh, A. Prostatic tumor regrowth after initially successful castration therapy may be related to a decreased apoptotic cell death rate. Cancer Research 54, 4281–4284 (1994). Lawson, D.A., Xin, L., Lukacs, R.U., Cheng, D. & Witte, O.N. Isolation and functional characterization of murine prostate stem cells. Proceedings of the National Academy of Sciences of the United States of America 104, 181–186 (2007).

Cellular and Molecular Signatures of Androgen Ablation of Prostate Cancer

543

Lee, C. Gross dissection of three lobes of the rat prostate. in Current Concepts and Approaches to the Study of Prostate Cancer (eds. Coffey, D.S., Bruchovsky, N., Gardner, W.A.J., Resnick, M. & Karr, J.P.) 577–582 (Alan R. Liss, Inc., New York, NY, 1987). Lee, H.J. & Chang, C. Recent advances in androgen receptor action. Cellular and Molecular Life Sciences 60, 1613–1622 (2003). Lee, K.W., Cobb, L.J., Paharkova-Vatchkova, V., Liu, B., Milbrandt, J. & Cohen, P. Contribution of the orphan nuclear receptor Nur77 to the apoptotic action of IGFBP-3. Carcinogenesis 28, 1653–1658 (2007). Lee, K.W., Ma, L., Yan, X., Liu, B., Zhang, X.K. & Cohen, P. Rapid apoptosis induction by IGFBP-3 involves an insulin-like growth factor-independent nucleomitochondrial translocation of RXRalpha/Nur77. The Journal of Biological Chemistry 280, 16942–16948 (2005). Lee, S.O., Chun, J.Y., Nadiminty, N., Lou, W. & Gao, A.C. Interleukin-6 undergoes transition from growth inhibitor associated with neuroendocrine differentiation to stimulator accompanied by androgen receptor activation during LNCaP prostate cancer cell progression. The Prostate 67, 764–773 (2007). Leighton, P.A., Mitchell, K.J., Goodrich, L.V., Lu, X., Pinson, K., Scherz, P., Skarnes, W.C. & Tessier-Lavigne, M. Defining brain wiring patterns and mechanisms through gene trapping in mice. Nature 410, 174–179 (2001). Li, J., Yen, C., Liaw, D., Podsypanina, K., Bose, S., Wang, S.I., Puc, J., Miliaresis, C., Rodgers, L., McCombie, R., Bigner, S.H., Giovanella, B.C., Ittmann, M., Tycko, B., Hibshoosh, H., Wigler, M.H. & Parsons, R. PTEN, a putative protein tyrosine phosphatase gene mutated in human brain, breast, and prostate cancer. Science 275, 1943–1947 (1997). Li, R., Erdamar, S., Dai, H., Wheeler, T.M., Frolov, A., Scardino, P.T., Thompson, T.C. & Ayala, G.E. Forkhead protein FKHR and its phosphorylated form p-FKHR in human prostate cancer. Human Pathology 38, 1501–1507 (2007). Libertini, S.J., Tepper, C.G., Rodriguez, V., Asmuth, D.M., Kung, H.J. & Mudryj, M. Evidence for calpain-mediated androgen receptor cleavage as a mechanism for androgen independence. Cancer Research 67, 9001–9005 (2007). Limonta, P., Montagnani Marelli, M. & Moretti, R.M. LHRH analogues as anticancer agents: pituitary and extrapituitary sites of action. Expert Opinion on Investigational Drugs 10, 709– 720 (2001). Lin, J., Adam, R.M., Santiestevan, E. & Freeman, M.R. The phosphatidylinositol 30 -kinase pathway is a dominant growth factor-activated cell survival pathway in LNCaP human prostate carcinoma cells. Cancer Research 59, 2891–2897 (1999). Lin, H.K., Yeh, S., Kang, H.Y. & Chang, C. Akt suppresses androgen-induced apoptosis by phosphorylating and inhibiting androgen receptor. Proceedings of the National Academy of Sciences of the United States of America 98, 7200–7205 (2001). Lin, H.K., Wang, L., Hu, Y.C., Altuwaijri, S. & Chang, C. Phosphorylation-dependent ubiquitylation and degradation of androgen receptor by Akt require Mdm2 E3 ligase. The EMBO Journal 21, 4037–4048 (2002). Lin, H.K., Hu, Y.C., Yang, L., Altuwaijri, S., Chen, Y.T., Kang, H.Y. & Chang, C. Suppression versus induction of androgen receptor functions by the phosphatidylinositol 3-kinase/Akt pathway in prostate cancer LNCaP cells with different passage numbers. The Journal of Biological Chemistry 278, 50902–50907 (2003). Lin, B., Kolluri, S.K., Lin, F., Liu, W., Han, Y.H., Cao, X., Dawson, M.I., Reed, J.C. & Zhang, X. K. Conversion of Bcl-2 from protector to killer by interaction with nuclear orphan receptor Nur77/TR3. Cell 116, 527–540 (2004). Link, K.A., Burd, C.J., Williams, E., Marshall, T., Rosson, G., Henry, E., Weissman, B. & Knudsen, K.E. BAF57 governs androgen receptor action and androgen-dependent proliferation through SWI/SNF. Molecular and Cellular Biology 25, 2200–2215 (2005). Liu, B., Lee, H.Y., Weinzimer, S.A., Powell, D.R., Clifford, J.L., Kurie, J.M. & Cohen, P. Direct functional interactions between insulin-like growth factor-binding protein-3 and retinoid X receptor-alpha regulate transcriptional signaling and apoptosis. The Journal of Biological Chemistry 275, 33607–33613 (2000).

544

C.G. Tepper, H.-J. Kung

Liu, S., Vinall, R.L., Tepper, C., Shi, X.B., Xue, L.R., Ma, A.H., Wang, L.Y., Fitzgerald, L.D., Wu, Z., Gandour-Edwards, R., deVere White, R.W. & Kung, H.J. Inappropriate activation of androgen receptor by relaxin via beta-catenin pathway. Oncogene 27, 499–505 (2008). Louie, M.C., Yang, H.Q., Ma, A.H., Xu, W., Zou, J.X., Kung, H.J. & Chen, H.W. Androgeninduced recruitment of RNA polymerase II to a nuclear receptor-p160 coactivator complex. Proceedings of the National Academy of Sciences of the United States of America 100, 2226– 2230 (2003). Lubahn, D.B., Joseph, D.R., Sar, M., Tan, J., Higgs, H.N., Larson, R.E., French, F.S. & Wilson, E. M. The human androgen receptor: complementary deoxyribonucleic acid cloning, sequence analysis and gene expression in prostate. Molecular Endocrinology (Baltimore, Md) 2, 1265– 1275 (1988a). Lubahn, D.B., Joseph, D.R., Sullivan, P.M., Willard, H.F., French, F.S. & Wilson, E.M. Cloning of human androgen receptor complementary DNA and localization to the X chromosome. Science 240, 327–330 (1988b). Majumder, P.K., Febbo, P.G., Bikoff, R., Berger, R., Xue, Q., McMahon, L.M., Manola, J., Brugarolas, J., McDonnell, T.J., Golub, T.R., Loda, M., Lane, H.A. & Sellers, W.R. mTOR inhibition reverses Akt-dependent prostate intraepithelial neoplasia through regulation of apoptotic and HIF-1-dependent pathways. Nature Medicine 10, 594–601 (2004). Malik, S.N., Brattain, M., Ghosh, P.M., Troyer, D.A., Prihoda, T., Bedolla, R. & Kreisberg, J.I. Immunohistochemical demonstration of phospho-Akt in high Gleason grade prostate cancer. Clinical Cancer Research 8, 1168–1171 (2002). Maltepe, E., Keith, B., Arsham, A.M., Brorson, J.R. & Simon, M.C. The role of ARNT2 in tumor angiogenesis and the neural response to hypoxia. Biochemical and Biophysical Research Communications 273, 231–238 (2000). Maniatis, T., Goodbourn, S. & Fischer, J.A. Regulation of inducible and tissue-specific gene expression. Science 236, 1237–1245 (1987). Mason, H.A., Rakowiecki, S.M., Raftopoulou, M., Nery, S., Huang, Y., Gridley, T. & Fishell, G. Notch signaling coordinates the patterning of striatal compartments. Development 132, 4247– 4258 (2005). Matsunaga, N., Kaku, T., Itoh, F., Tanaka, T., Hara, T., Miki, H., Iwasaki, M., Aono, T., Yamaoka, M., Kusaka, M. & Tasaka, A. C17,20-lyase inhibitors I. Structure-based de novo design and SAR study of C17,20-lyase inhibitors. Bioorganic and Medicinal Chemistry 12, 2251–2273 (2004). Matusik, R.J., Jin, R.J., Sun, Q., Wang, Y., Yu, X., Gupta, A., Nandana, S., Case, T.C., Paul, M., Mirosevich, J., Oottamasathien, S. & Thomas, J. Prostate epithelial cell fate. Differentiation 76, 682–698 (2008). McCollum, A.K., Teneyck, C.J., Sauer, B.M., Toft, D.O. & Erlichman, C. Up-regulation of heat shock protein 27 induces resistance to 17-allylamino-demethoxygeldanamycin through a glutathione-mediated mechanism. Cancer Research 66, 10967–10975 (2006). McDonough, W.S., Tran, N.L. & Berens, M.E. Regulation of glioma cell migration by serinephosphorylated P311. Neoplasia (New York, NY) 7, 862–872 (2005). Michikawa, M., Kikuchi, S., Muramatsu, H., Muramatsu, T. & Kim, S.U. Retinoic acid responsive gene product, midkine, has neurotrophic functions for mouse spinal cord and dorsal root ganglion neurons in culture. Journal of Neuroscience Research 35, 530–539 (1993). Mizokami, A. & Chang, C. Induction of translation by the 50 -untranslated region of human androgen receptor mRNA. The Journal of Biological Chemistry 269, 25655–25659 (1994). Mizokami, A., Yeh, S.Y. & Chang, C. Identification of 30 ,50 -cyclic adenosine monophosphate response element and other cis-acting elements in the human androgen receptor gene promoter. Molecular Endocrinology (Baltimore, Md) 8, 77–88 (1994). Modur, V., Nagarajan, R., Evers, B.M. & Milbrandt, J. FOXO proteins regulate tumor necrosis factor-related apoptosis inducing ligand expression. Implications for PTEN mutation in prostate cancer. The Journal of Biological Chemistry 277, 47928–47937 (2002).

Cellular and Molecular Signatures of Androgen Ablation of Prostate Cancer

545

Mora, G.R., Prins, G.S. & Mahesh, V.B. Autoregulation of androgen receptor protein and messenger RNA in rat ventral prostate is protein synthesis dependent. The Journal of Steroid Biochemistry and Molecular Biology 58, 539–549 (1996). Mori, Y., Yin, J., Rashid, A., Leggett, B.A., Young, J., Simms, L., Kuehl, P.M., Langenberg, P., Meltzer, S.J. & Stine, O.C. Instabilotyping: comprehensive identification of frameshift mutations caused by coding region microsatellite instability. Cancer Research 61, 6046–6049 (2001). Mostaghel, E.A., Page, S.T., Lin, D.W., Fazli, L., Coleman, I.M., True, L.D., Knudsen, B., Hess, D.L., Nelson, C.C., Matsumoto, A.M., Bremner, W.J., Gleave, M.E. & Nelson, P.S. Intraprostatic androgens and androgen-regulated gene expression persist after testosterone suppression: therapeutic implications for castration-resistant prostate cancer. Cancer Research 67, 5033– 5041 (2007). Muramatsu, T. Midkine (MK), the product of a retinoic acid responsive gene, and pleiotrophin constitute a new protein family regulating growth and differentiation. The International Journal of Developmental Biology 37, 183–188 (1993). Muramatsu, H. & Muramatsu, T. Purification of recombinant midkine and examination of its biological activities: functional comparison of new heparin binding factors. Biochemical and Biophysical Research Communications 177, 652–658 (1991). Murillo, H., Huang, H., Schmidt, L.J., Smith, D.I. & Tindall, D.J. Role of PI3K signaling in survival and progression of LNCaP prostate cancer cells to the androgen refractory state. Endocrinology 142, 4795–4805 (2001). Nakajin, S., Shively, J.E., Yuan, P.M. & Hall, P.F. Microsomal cytochrome P-450 from neonatal pig testis: two enzymatic activities (17 alpha-hydroxylase and c17,20-lyase) associated with one protein. Biochemistry 20, 4037–4042 (1981). Nakaya, H.I., Beckedorff, F.C., Baldini, M.L., Fachel, A.A., Reis, E.M. & Verjovski-Almeida, S. Splice variants of TLE family genes and up-regulation of a TLE3 isoform in prostate tumors. Biochemical and Biophysical Research Communications 364, 918–923 (2007). Nelson, P.S., Clegg, N., Arnold, H., Ferguson, C., Bonham, M., White, J., Hood, L. & Lin, B. The program of androgen-responsive genes in neoplastic prostate epithelium. Proceedings of the National Academy of Sciences of the United States of America 99, 11890–11895 (2002). Neri, R., Peets, E. & Watnick, A. Anti-androgenicity of flutamide and its metabolite Sch 16423. Biochemical Society Transactions 7, 565–569 (1979). Nishiyama, T., Hashimoto, Y. & Takahashi, K. The influence of androgen deprivation therapy on dihydrotestosterone levels in the prostatic tissue of patients with prostate cancer. Clinical Cancer Research 10, 7121–7126 (2004). Oh, J., Woo, J.M., Choi, E., Kim, T., Cho, B.N., Park, Z.Y., Kim, Y.C., Kim, D.H. & Cho, C. Molecular, biochemical, and cellular characterization of epididymal ADAMs, ADAM7 and ADAM28. Biochemical and Biophysical Research Communications 331, 1374–1383 (2005). Owada, K., Sanjo, N., Kobayashi, T., Mizusawa, H., Muramatsu, H., Muramatsu, T. & Michikawa, M. Midkine inhibits caspase-dependent apoptosis via the activation of mitogen-activated protein kinase and phosphatidylinositol 3-kinase in cultured neurons. Journal of Neurochemistry 73, 2084–2092 (1999). Ozes, O.N., Mayo, L.D., Gustin, J.A., Pfeffer, S.R., Pfeffer, L.M. & Donner, D.B. NF-kappaB activation by tumour necrosis factor requires the Akt serine-threonine kinase. Nature 401, 82– 85 (1999). Palaparti, A., Baratz, A. & Stifani, S. The Groucho/transducin-like enhancer of split transcriptional repressors interact with the genetically defined amino-terminal silencing domain of histone H3. The Journal of Biological Chemistry 272, 26604–26610 (1997). Peeling, W.B. Phase III studies to compare goserelin (Zoladex) with orchiectomy and with diethylstilbestrol in treatment of prostatic carcinoma. Urology 33, 45–52 (1989). Pelley, R.P., Chinnakannu, K., Murthy, S., Strickland, F.M., Menon, M., Dou, Q.P., Barrack, E.R. & Reddy, G.P. Calmodulin-androgen receptor (AR) interaction: calcium-dependent, calpainmediated breakdown of AR in LNCaP prostate cancer cells. Cancer Research 66, 11754– 11762 (2006).

546

C.G. Tepper, H.-J. Kung

Peng, L., Malloy, P.J., Wang, J. & Feldman, D. Growth inhibitory concentrations of androgens upregulate insulin-like growth factor binding protein-3 expression via an androgen response element in LNCaP human prostate cancer cells. Endocrinology 147, 4599–4607 (2006). Perlman, H., Zhang, X., Chen, M.W., Walsh, K. & Buttyan, R. An elevated bax/bcl-2 ratio corresponds with the onset of prostate epithelial cell apoptosis. Cell Death and Differentiation 6, 48–54 (1999). Pinski, J., Wang, Q., Quek, M.L., Cole, A., Cooc, J., Danenberg, K. & Danenberg, P.V. Genisteininduced neuroendocrine differentiation of prostate cancer cells. The Prostate 66, 1136–1143 (2006). Pratt, W.B. & Toft, D.O. Steroid receptor interactions with heat shock protein and immunophilin chaperones. Endocrine Reviews 18, 306–360 (1997). Price, D. Comparative aspects of development and structure in the prostate. National Cancer Institute Monograph 12, 1–27 (1963). Prout, G.R., Jr., Kliman, B., Daly, J.J., Maclaughlin, R.A. & Griffin, P.P. In vitro uptake of 3H testosterone and its conversion to dihydrotestosterone by prostatic carcinoma and other tissues. The Journal of Urology 116, 603–610 (1976). Qiu, Y., Robinson, D., Pretlow, T.G. & Kung, H.J. Etk/Bmx, a tyrosine kinase with a pleckstrinhomology domain, is an effector of phosphatidylinositol 30 -kinase and is involved in interleukin 6-induced neuroendocrine differentiation of prostate cancer cells. Proceedings of the National Academy of Sciences of the United States of America 95, 3644–3649 (1998). Quarmby, V.E., Yarbrough, W.G., Lubahn, D.B., French, F.S. & Wilson, E.M. Autologous downregulation of androgen receptor messenger ribonucleic acid. Molecular Endocrinology (Baltimore, Md) 4, 22–28 (1990). Raffo, A.J., Perlman, H., Chen, M.W., Day, M.L., Streitman, J.S. & Buttyan, R. Overexpression of bcl-2 protects prostate cancer cells from apoptosis in vitro and confers resistance to androgen depletion in vivo. Cancer Research 55, 4438–4445 (1995). Rajah, R., Valentinis, B. & Cohen, P. Insulin-like growth factor (IGF)-binding protein-3 induces apoptosis and mediates the effects of transforming growth factor-beta1 on programmed cell death through a p53- and IGF-independent mechanism. The Journal of Biological Chemistry 272, 12181–12188 (1997). Ramaswamy, S., Nakamura, N., Sansal, I., Bergeron, L. & Sellers, W.R. A novel mechanism of gene regulation and tumor suppression by the transcription factor FKHR. Cancer Cell 2, 81–91 (2002). Renaud, J., Kerjan, G., Sumita, I., Zagar, Y., Georget, V., Kim, D., Fouquet, C., Suda, K., Sanbo, M., Suto, F., Ackerman, S.L., Mitchell, K.J., Fujisawa, H. & Chedotal, A. Plexin-A2 and its ligand, Sema6A, control nucleus-centrosome coupling in migrating granule cells. Nature Neuroscience 11, 440–449 (2008). Renoir, J.M., Radanyi, C., Faber, L.E. & Baulieu, E.E. The non-DNA-binding heterooligomeric form of mammalian steroid hormone receptors contains a hsp90-bound 59-kilodalton protein. The Journal of Biological Chemistry 265, 10740–10745 (1990). Robinson, M.R. & Thomas, B.S. Effect of hormonal therapy on plasma testosterone levels in prostatic carcinoma. British Medical Journal 4, 391–394 (1971). Rocchi, P., So, A., Kojima, S., Signaevsky, M., Beraldi, E., Fazli, L., Hurtado-Coll, A., Yamanaka, K. & Gleave, M. Heat shock protein 27 increases after androgen ablation and plays a cytoprotective role in hormone-refractory prostate cancer. Cancer Research 64, 6595–6602 (2004). Rocchi, P., Beraldi, E., Ettinger, S., Fazli, L., Vessella, R.L., Nelson, C. & Gleave, M. Increased Hsp27 after androgen ablation facilitates androgen-independent progression in prostate cancer via signal transducers and activators of transcription 3-mediated suppression of apoptosis. Cancer Research 65, 11083–11093 (2005). Rokhlin, O.W., Guseva, N.V., Tagiyev, A.F., Glover, R.A. & Cohen, M.B. Caspase-8 activation is necessary but not sufficient for tumor necrosis factor-related apoptosis-inducing ligand (TRAIL)-mediated apoptosis in the prostatic carcinoma cell line LNCaP. The Prostate 52, 1–11 (2002).

Cellular and Molecular Signatures of Androgen Ablation of Prostate Cancer

547

Romashkova, J.A. & Makarov, S.S. NF-kappaB is a target of AKT in anti-apoptotic PDGF signalling. Nature 401, 86–90 (1999). Ryan, C.J. & Small, E.J. Early versus delayed androgen deprivation for prostate cancer: new fuel for an old debate. Journal of Clinical Oncology 23, 8225–8231 (2005). Sansone, P., Storci, G., Tavolari, S., Guarnieri, T., Giovannini, C., Taffurelli, M., Ceccarelli, C., Santini, D., Paterini, P., Marcu, K.B., Chieco, P. & Bonafe, M. IL-6 triggers malignant features in mammospheres from human ductal breast carcinoma and normal mammary gland. The Journal of Clinical Investigation 117, 3988–4002 (2007). Sastry, K.S., Karpova, Y. & Kulik, G. Epidermal growth factor protects prostate cancer cells from apoptosis by inducing BAD phosphorylation via redundant signaling pathways. The Journal of Biological Chemistry 281, 27367–27377 (2006a). Sastry, K.S., Smith, A.J., Karpova, Y., Datta, S.R. & Kulik, G. Diverse antiapoptotic signaling pathways activated by vasoactive intestinal polypeptide, epidermal growth factor, and phosphatidylinositol 3-kinase in prostate cancer cells converge on BAD. The Journal of Biological Chemistry 281, 20891–20901 (2006b). Sato, S., Fujita, N. & Tsuruo, T. Modulation of Akt kinase activity by binding to Hsp90. Proceedings of the National Academy of Sciences of the United States of America 97, 10832–10837 (2000). Scheufler, C., Brinker, A., Bourenkov, G., Pegoraro, S., Moroder, L., Bartunik, H., Hartl, F.U. & Moarefi, I. Structure of TPR domain-peptide complexes: critical elements in the assembly of the Hsp70-Hsp90 multichaperone machine. Cell 101, 199–210 (2000). Schuh, S., Yonemoto, W., Brugge, J., Bauer, V.J., Riehl, R.M., Sullivan, W.P. & Toft, D.O. A 90,000-dalton binding protein common to both steroid receptors and the Rous sarcoma virus transforming protein, pp60v-src. The Journal of Biological Chemistry 260, 14292–14296 (1985). Sciarra, A., Mariotti, G., Gentile, V., Voria, G., Pastore, A., Monti, S. & Di Silverio, F. Neuroendocrine differentiation in human prostate tissue: is it detectable and treatable? BJU International 91, 438–445 (2003). Segnitz, B. & Gehring, U. The function of steroid hormone receptors is inhibited by the hsp90specific compound geldanamycin. The Journal of Biological Chemistry 272, 18694–18701 (1997). Sells, S.F., Wood, D.P., Jr., Joshi-Barve, S.S., Muthukumar, S., Jacob, R.J., Crist, S.A., Humphreys, S. & Rangnekar, V.M. Commonality of the gene programs induced by effectors of apoptosis in androgen-dependent and -independent prostate cells. Cell Growth & Differentation 5, 457–466 (1994). Sells, S.F., Han, S.S., Muthukkumar, S., Maddiwar, N., Johnstone, R., Boghaert, E., Gillis, D., Liu, G., Nair, P., Monnig, S., Collini, P., Mattson, M.P., Sukhatme, V.P., Zimmer, S.G., Wood, D. P., Jr., McRoberts, J.W., Shi, Y. & Rangnekar, V.M. Expression and function of the leucine zipper protein Par-4 in apoptosis. Molecular and Cellular Biology 17, 3823–3832 (1997). Shang, Y., Myers, M. & Brown, M. Formation of the androgen receptor transcription complex. Molecular Cell 9, 601–610 (2002). Sharma, N., Seftor, R.E., Seftor, E.A., Gruman, L.M., Heidger, P.M., Jr., Cohen, M.B., Lubaroff, D. M. & Hendrix, M.J. Prostatic tumor cell plasticity involves cooperative interactions of distinct phenotypic subpopulations: role in vasculogenic mimicry. The Prostate 50, 189–201 (2002). Sheflin, L., Keegan, B., Zhang, W. & Spaulding, S.W. Inhibiting proteasomes in human HepG2 and LNCaP cells increases endogenous androgen receptor levels. Biochemical and Biophysical Research Communications 276, 144–150 (2000). Shi, X.B., Ma, A.H., Tepper, C.G., Xia, L., Gregg, J.P., Gandour-Edwards, R., Mack, P.C., Kung, H.J. & deVere White, R.W. Molecular alterations associated with LNCaP cell progression to androgen independence. The Prostate 60, 257–271 (2004). Signoretti, S., Waltregny, D., Dilks, J., Isaac, B., Lin, D., Garraway, L., Yang, A., Montironi, R., McKeon, F. & Loda, M. p63 is a prostate basal cell marker and is required for prostate development. The American Journal of Pathology 157, 1769–1775 (2000).

548

C.G. Tepper, H.-J. Kung

Song, C.S., Jung, M.H., Supakar, P.C., Chen, S., Vellanoweth, R.L., Chatterjee, B. & Roy, A.K. Regulation of androgen action by receptor gene inhibition. Annals of the New York Academy of Sciences 761, 97–108 (1995). St-Arnaud, R., Lachance, R., Kelly, S.J., Belanger, A., Dupont, A. & Labrie, F. Loss of luteinizing hormone bioactivity in patients with prostatic cancer treated with an LHRH agonist and a pure antiandrogen. Clinical Endocrinology 24, 21–30 (1986). Steck, P.A., Pershouse, M.A., Jasser, S.A., Yung, W.K., Lin, H., Ligon, A.H., Langford, L.A., Baumgard, M.L., Hattier, T., Davis, T., Frye, C., Hu, R., Swedlund, B., Teng, D.H. & Tavtigian, S.V. Identification of a candidate tumour suppressor gene, MMAC1, at chromosome 10q23.3 that is mutated in multiple advanced cancers. Nature Genetics 15, 356–362 (1997). Stefan, M., Claiborn, K.C., Stasiek, E., Chai, J.H., Ohta, T., Longnecker, R., Greally, J.M. & Nicholls, R.D. Genetic mapping of putative Chrna7 and Luzp2 neuronal transcriptional enhancers due to impact of a transgene-insertion and 6.8 Mb deletion in a mouse model of Prader-Willi and Angelman syndromes. BMC Genomics 6, 157 (2005). Suzuki, A., Matsuzawa, A. & Iguchi, T. Down regulation of Bcl-2 is the first step on Fas-mediated apoptosis of male reproductive tract. Oncogene 13, 31–37 (1996). Tepper, C.G., Boucher, D.L., Meekay, M.M., Shi, X.B., Li, L.F., Gandour-Edwards, R., Bold, R.J., deVere White, R.W. & Kung, H.J. Androgen withdrawal augments the PI3K-Akt pathway and increases the susceptibility of LNCaP prostate cancer cells to PI3K inhibitors prior to the development of androgen independence. in AACR-NCI-EORTC International Conference on Molecular Targets and Cancer Therapeutics (Miami Beach, FL, 2001). Tepper, C.G., Boucher, D.L., Ryan, P.E., Ma, A.H., Xia, L., Lee, L.F., Pretlow, T.G. & Kung, H.J. Characterization of a novel androgen receptor mutation in a relapsed CWR22 prostate cancer xenograft and cell line. Cancer Research 62, 6606–6614 (2002). Tepper, C.G., Vinall, R.L., Wee, C.B., Xue, L., Shi, X.B., Burich, R., Mack, P.C. & de Vere White, R.W. GCP-mediated growth inhibition and apoptosis of prostate cancer cells via androgen receptor-dependent and -independent mechanisms. The Prostate 67, 521–535 (2007). Teutsch, G., Goubet, F., Battmann, T., Bonfils, A., Bouchoux, F., Cerede, E., Gofflo, D., GaillardKelly, M. & Philibert, D. Non-steroidal antiandrogens: synthesis and biological profile of highaffinity ligands for the androgen receptor. The Journal of Steroid Biochemistry and Molecular Biology 48, 111–119 (1994). The Gene Ontology Consortium. Creating the gene ontology resource: design and implementation. Genome Research 11, 1425–1433 (2001). The Leuprolide Study Group. Leuprolide versus diethylstilbestrol for metastatic prostate cancer. The New England Journal of Medicine 311, 1281–1286 (1984). Thomas, L.N., Douglas, R.C., Lazier, C.B., Gupta, R., Norman, R.W., Murphy, P.R., Rittmaster, R.S. & Too, C.K. Levels of 5alpha-reductase type 1 and type 2 are increased in localized high grade compared to low grade prostate cancer. The Journal of Urology 179, 147–151 (2008a). Thomas, L.N., Douglas, R.C., Lazier, C.B., Too, C.K., Rittmaster, R.S. & Tindall, D.J. Type 1 and type 2 5alpha-reductase expression in the development and progression of prostate cancer. European Urology 53, 244–252 (2008b). Thompson, T.C. & Chung, L.W. Extraction of nuclear androgen receptors by sodium molybdate from normal rat prostates and prostatic tumors. Cancer Research 44, 1019–1026 (1984). Thompson, V.C., Morris, T.G., Cochrane, D.R., Cavanagh, J., Wafa, L.A., Hamilton, T., Wang, S., Fazli, L., Gleave, M.E. & Nelson, C.C. Relaxin becomes upregulated during prostate cancer progression to androgen independence and is negatively regulated by androgens. The Prostate 66, 1698–1709 (2006). Tilley, W.D., Marcelli, M. & McPhaul, M.J. Expression of the human androgen receptor gene utilizes a common promoter in diverse human tissues and cell lines. The Journal of Biological Chemistry 265, 13776–13781 (1990). Tillman, K., Oberfield, J.L., Shen, X.Q., Bubulya, A. & Shemshedini, L. c-Fos dimerization with c-Jun represses c-Jun enhancement of androgen receptor transactivation. Endocrine 9, 193–200 (1998).

Cellular and Molecular Signatures of Androgen Ablation of Prostate Cancer

549

Titus, M.A., Schell, M.J., Lih, F.B., Tomer, K.B. & Mohler, J.L. Testosterone and dihydrotestosterone tissue levels in recurrent prostate cancer. Clinical Cancer Research 11, 4653–4657 (2005). Tolis, G., Ackman, D., Stellos, A., Mehta, A., Labrie, F., Fazekas, A.T., Comaru-Schally, A.M., & Schally, A.V. Tumor growth inhibition in patients with prostatic carcinoma treated with luteinizing hormone-releasing hormone agonists. Proceedings of the National Academy of Sciences of the United States of America 79, 1658–1662 (1982). Tong, Y., Mentlein, R., Buhl, R., Hugo, H.H., Krause, J., Mehdorn, H.M. & Held-Feindt, J. Overexpression of midkine contributes to anti-apoptotic effects in human meningiomas. Journal of Neurochemistry 100, 1097–1107 (2007). Trachtenberg, J., Halpern, N. & Pont, A. Ketoconazole: a novel and rapid treatment for advanced prostatic cancer. The Journal of Urology 130, 152–153 (1983). Trachtenberg, J., Hicks, L.L. & Walsh, P.C. Methods for the determination of androgen receptor content in human prostatic tissue. Investigative Urology 18, 349–354 (1981). Trachtenberg, J. & Pont, A. Ketoconazole therapy for advanced prostate cancer. Lancet 2, 433– 435 (1984). Tremblay, F. & Marette, A. Amino acid and insulin signaling via the mTOR/p70 S6 kinase pathway. A negative feedback mechanism leading to insulin resistance in skeletal muscle cells. The Journal of Biological Chemistry 276, 38052–38060 (2001). Trojan, L., Schaaf, A., Steidler, A., Haak, M., Thalmann, G., Knoll, T., Gretz, N., Alken, P. & Michel, M.S. Identification of metastasis-associated genes in prostate cancer by genetic profiling of human prostate cancer cell lines. Anticancer Research 25, 183–191 (2005). Ueda, K. Detection of the retinoic acid-regulated genes in a RTBM1 neuroblastoma cell line using cDNA microarrayThe Kurume Medical Journal 48, 159–164 (2001). van Bokhoven, A., Varella-Garcia, M., Korch, C., Johannes, W.U., Smith, E.E., Miller, H.L., Nordeen, S.K., Miller, G.J. & Lucia, M.S. Molecular characterization of human prostate carcinoma cell lines. The Prostate 57, 205–225 (2003). van Weerden, W.M., van Kreuningen, A., Elissen, N.M., Vermeij, M., de Jong, F.H., van Steenbrugge, G.J. & Schroder, F.H. Castration-induced changes in morphology, androgen levels, and proliferative activity of human prostate cancer tissue grown in athymic nude mice. The Prostate 23, 149–164 (1993). Vashchenko, N. & Abrahamsson, P.A. Neuroendocrine differentiation in prostate cancer: implications for new treatment modalities. European Urology 47, 147–155 (2005). Veldscholte, J., Berrevoets, C.A., Brinkmann, A.O., Grootegoed, J.A. & Mulder, E. Anti-androgens and the mutated androgen receptor of LNCaP cells: differential effects on binding affinity, heat-shock protein interaction, and transcription activation. Biochemistry 31, 2393–2399 (1992). Veldscholte, J., Berrevoets, C.A., Zegers, N.D., van der Kwast, T.H., Grootegoed, J.A. & Mulder, E. Hormone-induced dissociation of the androgen receptor-heat-shock protein complex: use of a new monoclonal antibody to distinguish transformed from nontransformed receptors. Biochemistry 31, 7422–7430 (1992). Vinall, R.L., Tepper, C.G., Shi, X.B., Xue, L.A., Gandour-Edwards, R. & de Vere White, R.W. The R273H p53 mutation can facilitate the androgen-independent growth of LNCaP by a mechanism that involves H2 relaxin and its cognate receptor LGR7. Oncogene 25, 2082–2093 (2006). Vinall, R.L., Hwa, K., Ghosh, P., Pan, C.X., Lara, P.N., Jr. & de Vere White, R.W. Combination treatment of prostate cancer cell lines with bioactive soy isoflavones and perifosine causes increased growth arrest and/or apoptosis. Clinical Cancer Research 13, 6204–6216 (2007). Vlietstra, R.J., van Alewijk, D.C., Hermans, K.G., van Steenbrugge, G.J. & Trapman, J. Frequent inactivation of PTEN in prostate cancer cell lines and xenografts. Cancer Research 58, 2720– 2723 (1998). Wallner, L., Dai, J., Escara-Wilke, J., Zhang, J., Yao, Z., Lu, Y., Trikha, M., Nemeth, J.A., Zaki, M.H. & Keller, E.T. Inhibition of interleukin-6 with CNTO328, an anti-interleukin-6 monoclonal antibody, inhibits conversion of androgen-dependent prostate cancer to an androgenindependent phenotype in orchiectomized mice. Cancer Research 66, 3087–3095 (2006).

550

C.G. Tepper, H.-J. Kung

Wang, Q., Li, W., Liu, X.S., Carroll, J.S., Janne, O.A., Keeton, E.K., Chinnaiyan, A.M., Pienta, K. J. & Brown, M. A hierarchical network of transcription factors governs androgen receptordependent prostate cancer growth. Molecular Cell 27, 380–392 (2007). Wellington, C.L., Ellerby, L.M., Hackam, A.S., Margolis, R.L., Trifiro, M.A., Singaraja, R., McCutcheon, K., Salvesen, G.S., Propp, S.S., Bromm, M., Rowland, K.J., Zhang, T., Rasper, D., Roy, S., Thornberry, N., Pinsky, L., Kakizuka, A., Ross, C.A., Nicholson, D.W., Bredesen, D.E. & Hayden, M.R. Caspase cleavage of gene products associated with triplet expansion disorders generates truncated fragments containing the polyglutamine tract. The Journal of Biological Chemistry 273, 9158–9167 (1998). Werner, E.D., Lee, J., Hansen, L., Yuan, M. & Shoelson, S.E. Insulin resistance due to phosphorylation of insulin receptor substrate-1 at serine 302. The Journal of Biological Chemistry 279, 35298–35305 (2004). Westin, P., Bergh, A. & Damber, J.E. Castration rapidly results in a major reduction in epithelial cell numbers in the rat prostate, but not in the highly differentiated Dunning R3327 prostatic adenocarcinoma. The Prostate 22, 65–74 (1993). Whitesell, L. & Cook, P. Stable and specific binding of heat shock protein 90 by geldanamycin disrupts glucocorticoid receptor function in intact cells. Molecular Endocrinology (Baltimore, Md) 10, 705–712 (1996). Wissmann, M., Yin, N., Muller, J.M., Greschik, H., Fodor, B.D., Jenuwein, T., Vogler, C., Schneider, R., Gunther, T., Buettner, R., Metzger, E. & Schule, R. Cooperative demethylation by JMJD2C and LSD1 promotes androgen receptor-dependent gene expression. Nature Cell Biology 9, 347–353 (2007). Wojno, K.J. & Epstein, J.I. The utility of basal cell-specific anti-cytokeratin antibody (34 beta E12) in the diagnosis of prostate cancer. A review of 228 cases. The American Journal of Surgical Pathology 19, 251–260 (1995). Wolf, D.A., Herzinger, T., Hermeking, H., Blaschke, D. & Horz, W. Transcriptional and posttranscriptional regulation of human androgen receptor expression by androgen. Molecular Endocrinology (Baltimore, Md) 7, 924–936 (1993). Wright, W.W., Chan, K.C. & Bardin, C.W. Characterization of the stabilizing effect of sodium molybdate on the androgen receptor present in mouse kidney. Endocrinology 108, 2210–2216 (1981). Wright, M.E., Tsai, M.J. & Aebersold, R. Androgen receptor represses the neuroendocrine transdifferentiation process in prostate cancer cells. Molecular Endocrinology (Baltimore, Md 17, 1726–1737 (2003). Wu, C. & Huang, J. Phosphatidylinositol 3-kinase-AKT-mammalian target of rapamycin pathway is essential for neuroendocrine differentiation of prostate cancer. The Journal of Biological Chemistry 282, 3571–3583 (2007). Wu, M., Michaud, E.J. & Johnson, D.K. Cloning, functional study and comparative mapping of Luzp2 to mouse chromosome 7 and human chromosome 11p13–11p14. Mammalian Genome 14, 323–334 (2003). Xu, X.M., Fisher, D.A., Zhou, L., White, F.A., Ng, S., Snider, W.D. & Luo, Y. The transmembrane protein semaphorin 6A repels embryonic sympathetic axons. Journal of Neuroscience 20, 2638–2648 (2000). Xu, Y., Chen, S.Y., Ross, K.N. & Balk, S.P. Androgens induce prostate cancer cell proliferation through mammalian target of rapamycin activation and post-transcriptional increases in cyclin D proteins. Cancer Research 66, 7783–7792 (2006). Yamane, K., Toumazou, C., Tsukada, Y., Erdjument-Bromage, H., Tempst, P., Wong, J. & Zhang, Y. JHDM2A, a JmjC-containing H3K9 demethylase, facilitates transcription activation by androgen receptor. Cell 125, 483–495 (2006). Yang, X., Chen, M.W., Terry, S., Vacherot, F., Chopin, D.K., Bemis, D.L., Kitajewski, J., Benson, M.C., Guo, Y. & Buttyan, R. A human- and male-specific protocadherin that acts through the wnt signaling pathway to induce neuroendocrine transdifferentiation of prostate cancer cells. Cancer Research 65, 5263–5271 (2005).

Cellular and Molecular Signatures of Androgen Ablation of Prostate Cancer

551

Yao, J., Liu, Y., Husain, J., Lo, R., Palaparti, A., Henderson, J. & Stifani, S. Combinatorial expression patterns of individual TLE proteins during cell determination and differentiation suggest non-redundant functions for mammalian homologs of Drosophila Groucho. Development Growth and Differentiation 40, 133–146 (1998). Yao, J., Liu, Y., Lo, R., Tretjakoff, I., Peterson, A. & Stifani, S. Disrupted development of the cerebral hemispheres in transgenic mice expressing the mammalian Groucho homologue transducin-like-enhancer of split 1 in postmitotic neurons. Mechanisms of Development 93, 105–115 (2000). Yeap, B.B., Krueger, R.G. & Leedman, P.J. Differential posttranscriptional regulation of androgen receptor gene expression by androgen in prostate and breast cancer cells. Endocrinology 140, 3282–3291 (1999). Yeap, B.B., Voon, D.C., Vivian, J.P., McCulloch, R.K., Thomson, A.M., Giles, K.M., CzyzykKrzeska, M.F., Furneaux, H., Wilce, M.C., Wilce, J.A. & Leedman, P.J. Novel binding of HuR and poly(C)-binding protein to a conserved UC-rich motif within the 30 -untranslated region of the androgen receptor messenger RNA. The Journal of Biological Chemistry 277, 27183– 27192 (2002). Yoshimura, T., Kawano, Y., Arimura, N., Kawabata, S., Kikuchi, A. & Kaibuchi, K. GSK-3beta regulates phosphorylation of CRMP-2 and neuronal polarity. Cell 120, 137–149 (2005). Young, J.C., Obermann, W.M. & Hartl, F.U. Specific binding of tetratricopeptide repeat proteins to the C-terminal 12-kDa domain of hsp90. The Journal of Biological Chemistry 273, 18007– 18010 (1998). Young, J.C., Moarefi, I. & Hartl, F.U. Hsp90: a specialized but essential protein-folding tool. The Journal of Cell Biology 154, 267–273 (2001). You, Z., Dong, Y., Kong, X., Beckett, L.A., Gandour-Edwards, R. & Melamed, J. Midkine is a NFkappaB-inducible gene that supports prostate cancer cell survival. BMC Medical Genomics 1, 6 (2008). Yuan, T.C., Veeramani, S., Lin, F.F., Kondrikou, D., Zelivianski, S., Igawa, T., Karan, D., Batra, S.K. & Lin, M.F. Androgen deprivation induces human prostate epithelial neuroendocrine differentiation of androgen-sensitive LNCaP cells. Endocrine-Related Cancer 13, 151–167 (2006). Yuan, T.C., Veeramani, S. & Lin, M.F. Neuroendocrine-like prostate cancer cells: neuroendocrine transdifferentiation of prostate adenocarcinoma cells. Endocrine-Related Cancer 14, 531–547 (2007). Yu, X., Li, P., Roeder, R.G. & Wang, Z. Inhibition of androgen receptor-mediated transcription by amino-terminal enhancer of split. Molecular and Cellular Biology 21, 4614–4625 (2001). Zelivianski, S., Verni, M., Moore, C., Kondrikov, D., Taylor, R. & Lin, M.F. Multipathways for transdifferentiation of human prostate cancer cells into neuroendocrine-like phenotype. Biochimica et Biophysica Acta 1539, 28–43 (2001). Zhang, X.Q., Kondrikov, D., Yuan, T.C., Lin, F.F., Hansen, J. & Lin, M.F. Receptor protein tyrosine phosphatase alpha signaling is involved in androgen depletion-induced neuroendocrine differentiation of androgen-sensitive LNCaP human prostate cancer cells. Oncogene 22, 6704–6716 (2003). Zhang, Y., Akinmade, D. & Hamburger, A.W. The ErbB3 binding protein Ebp1 interacts with Sin3A to repress E2F1 and AR-mediated transcription. Nucleic Acids Research 33, 6024–6033 (2005). Zhao, G.Q., Bacher, M., Friedrichs, B., Schmidt, W., Rausch, U., Goebel, H.W., Tuohimaa, P. & Aumuller, G. Functional properties of isolated stroma and epithelium from rat ventral prostate during androgen deprivation and estrogen treatment. Experimental and Clinical Endocrinology 101, 69–77 (1993). Zoubeidi, A., Zardan, A., Beraldi, E., Fazli, L., Sowery, R., Rennie, P., Nelson, C. & Gleave, M. Cooperative interactions between androgen receptor (AR) and heat-shock protein 27 facilitate AR transcriptional activity. Cancer Research 67, 10455–10465 (2007). Zuber, M.X., Simpson, E.R. & Waterman, M.R. Expression of bovine 17 alpha-hydroxylase cytochrome P-450 cDNA in nonsteroidogenic (COS 1) cells. Science 234, 1258–1261 (1986).

Tissue Levels of Androgens in Castration-Recurrent Prostate Cancer James L. Mohler and Mark A. Titus

Abstract AR remains active in growth signaling despite castrate levels of circulating androgens. AR protein and AR-regulated proteins are expressed in CaP that recurs during ADT in both the primary and bone metastases. AR activation in recurrent CaP may occur by a variety of mechanisms that alter the sensitivity or specificity of AR. Recent studies using androgen-independent CaP cell lines and xenografts demonstrated that AR over-expression allowed recurrent CaP growth in the presence of castrate levels of circulating androgens. However, AR mutations that prevented ligand-binding prevented recurrent growth; over-expressed AR required ligand to confer recurrent growth. We showed that DHT levels were decreased by 91% in clinical specimens of castration-recurrent CaP (1.25 pmol/ gm tissue) compared to androgen-stimulated benign prostate and DHT levels were sufficient for AR activation in most specimens of recurrent CaP. Other investigators have supported these findings in prostatectomy specimens obtained after 3-6 months of ADT and benign prostate specimens after 1 month of ADT. Perturbations in androgen metabolism pathways during ADT may allow intracrine production of DHT from adrenal androgens and even cholesterol. The tissue levels of T and DHT support a new paradigm. CaP that recurs after medical or surgical castration is not “androgen-independent” because recurrent CaP usually has androgen levels sufficient to activate AR. New treatments should be directed at preventing intracrine synthesis of testicular androgens from adrenal androgens or cholesterol, degrading these androgens or, failing both, destroying AR.

1 Introduction An American man is diagnosed with prostate cancer every 3 min (Jemal et al. 2007). In spite of earlier detection (Catalona et al. 1991), approximately 30% of J.L. Mohler(*) Department of Urologic Oncology, Roswell Park Cancer Institute, Elm & Carlton Streets, Buffalo, NY 14263, USA, E-mail: [email protected]

D. Tindall and J. Mohler (eds.) Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4_23, # Springer Science + Business Media, LLC 2009

553

554

J.L. Mohler, M.A. Titus

men treated with curative intent will suffer tumor recurrence. These men as well as those who present with locally advanced or metastatic prostate cancer can be palliated by androgen deprivation therapy (ADT), a treatment that remains unimproved since its discovery more than 50 years ago (Huggins and Hodges 2002). Regardless of the initial responsiveness to ADT, almost all patients succumb to recurrent prostate cancer, and an American dies from the disease every 17 min (Jemal et al. 2007). Better understanding of the mechanism of development and growth of castration-recurrent prostate cancer is imperative since all known treatments for castration-recurrent prostate cancer produce short-lived responses, at best, and side effects are high.

2 The Androgen Receptor in Castration-Recurrent Prostate Cancer The androgen receptor (AR) was initially hypothesized to play a minimal role in the recurrence and growth of prostate cancer during ADT; castration-recurrent prostate cancer was hypothesized to consist of cells that lacked AR, since androgens circulated at castrate levels. Although AR expression is diminished following ADT that induces clinical remission in most patients, castration-recurrent prostate cancer expresses levels of AR protein similar to those found in androgen-stimulated prostate cancer and benign prostate. In addition, AR-regulated genes are expressed in both androgen-stimulated and castration-recurrent prostate cancer (Gregory et al. 1998, 1999, 2001c; Miyake et al. 2000; Mohler et al. 2002; Mousses et al. 2001; Sadar et al. 1999; Stewart et al. 2001). AR protein and AR-regulated proteins are expressed in prostate cancer that recurs during ADT in both the primary tumor (Edwards et al. 2003; Linja et al. 2001; Mohler et al. 2004; van der Kwast et al. 1991) and bone metastases (Brown et al. 2002; Hobisch et al. 1995). Thus, it should not have been surprising when quantitative immunoanalysis of 19 specimens of castration-recurrent prostate cancer and 16 specimens of androgenstimulated benign prostate revealed that nuclei were immunopositive for AR with similar frequency (83.7% 11.6% vs 77.3% 13.0%, respectively, p = 0.25) and immunopositive nuclei expressed similar levels of AR protein (mean optical density, 0.284 0.115 vs 0.315 0.044, respectively, p = 0.48) (Fig. 1; Mohler et al. 2004). The finding that AR protein levels were similar in archived specimens of castration-recurrent prostate cancer and androgen-stimulated benign prostate when measured using an AR monoclonal antibody and automated image analysis was similar to earlier reports that used qualitative methods (van der Kwast et al. 1991; Visakorpi et al. 1995). High levels of AR expression in castration-recurrent prostate cancer suggest a central role for AR in growth regulation of castration-recurrent prostate cancer (Feldman and Feldman 2001; Gelmann 2002; Grossmann et al. 2001; Sadar et al. 1999). A fundamental question is ‘‘How is AR activated after medical or surgical castration?’’

Tissue Levels of Androgens in Castration-Recurrent Prostate Cancer

555

Fig. 1 AR protein expression is similar in androgen-stimulated benign prostate (left) and castration-recurrent prostate cancer (right)

Other theories for prostate cancer recurrence during ADT have been based upon preclinical studies and clinical specimens. AR mutations were thought to play a central role based upon the seminal report of point mutations in 50% of ten castration-recurrent prostate cancer specimens (Taplin et al. 1995); however, the high rate of mutations proved due to cloning/PCR artifacts. The ‘‘true’’ frequency of AR mutations in castration-recurrent prostate cancer appears to be closer to 5% but may be as high as 30% if bone marrow metastases from patients who received prolonged treatment with antiandrogens are included (Taplin et al. 1999). Moreover, a review of the Quebec AR mutation repository (http://androgendb.mcgill.ca) reveals that most mutations occur in the ligand-binding domain and recreation of these mutations shows that they are almost always functional (de Vere White et al. 1997), which contrasts with androgen-insensitivity syndrome when most mutations produce stop codons and AR proteins that lack function. The suggestion using preclinical models that AR amplification is important clinically has been investigated by several laboratories. For example, Linja et al. (2001) demonstrated that AR gene amplification was uncommon in androgenstimulated benign prostate or prostate cancer but occurred frequently in castration-recurrent prostate cancer. Our group (Ford et al. 2003) showed that AR amplification occurred in 8 of 24 men with castration-recurrent prostate cancer, and AR amplification increased the amount of AR protein expressed per cell by approximately 50%. However, AR amplification did not affect survival of prostate cancer patients. Finally, several laboratories continue preclinical investigations of ligand-independent AR activation and AR-independent androgen-regulated gene expression, although little data suggests that either of these mechanisms for growth occur clinically. Hence, new theories for the development of castration-recurrent prostate cancer require examination. First, the AR responds to castration with molecular and biochemical alterations that cause hypersensitivity to low levels of ligand. Our group was among the first to demonstrate that AR in androgen-independent prostate

556

J.L. Mohler, M.A. Titus

cancer cell lines was approximately four logs more sensitive to androgens (Gregory et al. 2001b). A biochemical explanation was offered; the AR coactivator profile changed from SRC-1 toward TIF-2, a finding that was true both in vitro and in clinical specimens (Gregory et al. 2001a). However, the large number of coactivators and corepressors suggests that redundancy must be a characteristic of this complex (http://androgendb.mcgill.ca) and interventions directed at AR coregulators may be difficult. Another biochemical mechanism for hypersensitization of the AR in the face of low ligand availability has been suggested independently by two groups. Guo et al. (2006) reported that phosphorylation of Y534 by a Sarc tyrosine kinase hypersensitized AR to low levels of androgen and was a common characteristic of clinical specimens of castration-recurrent prostate cancer. In addition to hypersensitizing the AR to dihydrotestosterone (DHT), AR was also hypersensitized to epidermal growth factor (EGF), and the combination of EGF and DHT interacted synergistically to activate phosphorylated AR. Moreover, xenografted prostate cancer cell lines capable of AR phosphorylation grew faster than cell lines in which the phosphorylation site was mutated. Similar findings were found at Y267 and Y363, which were phosphorylated by ACK1 tyrosine kinase, and phosphorylation of these sites also was increased in clinical samples of castration-recurrent prostate cancer (Mahajan et al. 2007). Taken together, the presence of high levels of AR and multiple means of hypersensitization suggest that AR responds in every way possible to ensure its transactivation in castration-recurrent prostate cancer. A common denominator in the pursuit of a role for AR in castration-recurrent prostate cancer has been the relative absence of ligand. However, one must consider the possibility that the levels of androgens may be different in serum and prostate tissue; AR transactivation may occur through ‘‘normal’’ levels of AR ligand in the prostate cancer microenvironment in spite of castrate levels of circulating androgens. If true, a paradigm shift is required; prostate cancer that recurs during ADT may not be androgen-independent.

3 Androgen Levels in Castration-Recurrent Prostate Cancer Tissue Geller et al. (1979) published an analysis of tissue androgen levels in prostate and nonandrogen target tissues with a specific emphasis on tissue steroid levels as markers of tumor differentiation and adequacy of antiandrogen therapy. Tissues procured by transurethral resection of the prostate were assayed using radioimmunoassay (RIA). They reported that 1-mg DES did not adequately suppress tissue DHT levels, since levels remained intermediate between androgen-stimulated benign prostate and prostate tissue procured from men who had been castrated (Table 1). They believed that their findings ‘‘support the long-suspected theoretical role of adrenal cortical androgens as biologically important sources of DHT in relapse of prostate cancer.’’ The clinical importance of their findings were obscured

Tissue Levels of Androgens in Castration-Recurrent Prostate Cancer Table 1 Prostate Tissue Androgen levels RIA T DHT Geller et al. (1979) AS-BP (n = 17) – 17.6 CaP orch DES (n = 9) – 4.47 CaP DES 1 mg (n = 6) – 12.4 Belanger et al. (1989) human CaP (n = ?) – 18.6 orch (n = 5, 2–12 m) – 9.29 orch + flu (n = 4, 2 m) – ND Mohler et al. (2004) AS-BP (n = 30) 3.26 8.13 RCaP (n = 15, 37 m) 2.78 1.45 Page et al. (2006) AS-BP (n = 4) 1.84 9.26 LHRH + T (n = 4, 1 m) 1.38 6.8 LHRH (n = 4, 1 m) 0.56 1.94

557

LC/MS/MS Nishiyama et al. (2004) AS-BP (n = 30) CaP (n = 30, 6 m) Mizokami et al. (2004) AS-BP (n = 15) CaP (n = 15, 3–6 m) Titus et al. (2005) AS-BP (n = 18) RCaP (n = 18, 37 m)

T

DHT

– –

18.7 4.65

– –

8.53 2.13

2.75 3.75

13.7 1.25

by the assertion (Belanger et al. 1989) that coadministration of antiandrogens cured prostate cancer by reducing tissue DHT levels to 0, a finding that was initially based upon experimental data in a total of four men (Table 1). A study of tissue androgen levels during ADT measured using RIA suggested that prostate cancer tissue DHT levels decreased from 5.24 ng/g tissue in noncastrated men 55–68 years of age to 2.7 ng/g tissue in five men who were castrated 2–12 months before radical prostatectomy. Among four castrated men receiving flutamide 250 mg three times daily for 2 months prior to prostatectomy, tissue DHT was undetectable. It was postulated that flutamide, by competing for high-affinity DHT binding to AR, decreased prostate DHT levels by increasing its degradation. These data led to the use of ‘‘total androgen blockade,’’ where tissue DHT was eliminated using antiandrogens (Labrie et al. 1982). However, a meta-analysis of clinical trials comparing LH-RH agonists and antiandrogens versus LH-RH agonists alone (Prostate Cancer Trialists’ Collaborative Group 1995) and a study comparing orchiectomy and antiandrogens versus orchiectomy alone (Eisenberger et al. 1998) demonstrated no survival benefit to combination therapy. After this nearly three-decade detour, interest and attention on Geller’s original hypothesis has been rekindled by our measurements made in men with castration-recurrent prostate cancer using RIA and confirmed using mass spectrometry (MS). Twenty one patients aged 57–86 years demonstrated clinical evidence of castration-recurrent prostate cancer (Table 1; Mohler et al. 2004). All underwent transurethral prostatectomy for urinary retention from local recurrence that occurred from 7 to 92 months after medical (10 men) or surgical (11 men) ADT. Histological examination revealed poorly differentiated carcinoma (Gleason sum 8–10) that represented an average of 92% (range 72–99%) of the cross-sectional area of the tissue sections. In order to compare these tissues to androgen-stimulated prostate

558

J.L. Mohler, M.A. Titus

tissue, frozen specimens of benign prostate tissue were obtained from radical prostatectomy specimens. Although tissue levels of DHT, dehydroepiandrosterone (DHEA), and androstenedione (ASD) were lower in castration-recurrent prostate cancer from men undergoing ADT than in benign prostate from untreated men (p < 0.01), tissue levels of DHT averaged 1.45 nM in castration-recurrent prostate cancer and 8.14 nM in benign prostate. Tissue levels of testosterone (T) were similar in castration-recurrent prostate cancer (2.78 nM) and benign prostate (3.27 nM) (p = 0.21). Tissue levels of prostate-specific antigen (PSA) in castration-recurrent prostate cancer were approximately one-tenth the level measured in benign prostate. Castration-recurrent prostate cancer tissue levels of androgens, estradiol (E2), sex hormone-binding globulin, and PSA did not differ between 3 patients who received flutamide and 12 patients who did not. In particular, tissue levels of DHT were similar (p = 0.29) in the two groups (flutamide, 3.75 3.58 pmol/g tissue, range 0.40–7.53 pmol/g tissue; no flutamide, 0.87 0.53 pmol/g tissue, range 0.37–2.17 pmol/g tissue). These results obtained using RIA should not have been surprising in view of Geller et al. However, skepticism by many was addressed by using a second method of measurement (Titus et al. 2005a). A prostate tissue homogenation and androgen extraction protocol and a liquid chromatography (LC)/electrospray ionization (ESI)/MS analytic method were developed in collaboration with Fred Lih and Dr. Kenneth Tomer, Director, of the NIEHS Mass Spectrometry Facility. T levels were similar in castration-recurrent prostate cancer (3.75 pmol/g tissue) and benign prostate (2.75 pmol/g tissue, p = 0.30). DHT levels in castration-recurrent prostate cancer (1.25 pmol/g tissue) were less than in benign prostate (13.7 pmol/g tissue; p < 0.0001), although DHT levels in most specimens of castrationrecurrent prostate cancer were sufficient for AR activation. DHT levels in castration-recurrent prostate cancer compared with benign prostate decreased 91% measured using MS and 82% measured using RIA. These findings in castration-recurrent prostate cancer were supported by two recent reports that measured tissue androgen levels in prostatectomy specimens from men who received neoadjuvant ADT. Mizokami et al. (2004) showed average tissue DHT levels measured using LC/MS/MS decreased 75% in prostatectomy specimens obtained after 3–6 months of ADT, and Nishiyama et al. (2004) found tissue DHT concentration decreased 75% in prostate tissue from 30 men receiving ADT for 6 months. Page et al. (2006) reported that tissue levels of testicular androgens in benign prostate, measured using RIA, may be sufficient for AR activation as early as 1 month after castration! Twelve men underwent prostate biopsies on day 28 after four men each received either placebo, acyline [a long-acting luteinizing hormonereleasing hormone (LHRH) antagonist], or a combination of acycline and T. Prostate tissue levels of T and DHT decreased by 70% and 80%, respectively, in the men receiving acycline. In spite of this decrease in prostate tissue levels of androgens, immuno histo chemistry (IHC) revealed no detectable differences among the three groups in cellular proliferation, apoptosis, and PSA or AR expression. This report, if confirmed, is especially important for two reasons. First, it

Tissue Levels of Androgens in Castration-Recurrent Prostate Cancer

559

suggests that benign prostate has the ability to produce testicular androgens as soon as 1 month after institution of ADT. Second, normal tissue homeostasis was recovered 1 month after ADT. Yet, in benign prostate, prostate volume remained reduced forever; ADT cures benign prostate enlargement. In contrast, prostate cancer must develop the ability to use these tissue androgens to recur as castration-recurrent prostate cancer to kill the patient.

4 Clinical Relevance of AR-Activating Levels of Tissue Androgens Are levels of 2.78–3.75 nM T and 1.25–1.45 nM DHT measured in castrationrecurrent prostate cancer tissues (Mohler et al. 2004; Titus et al. 2005a) sufficient to activate AR? Simard et al. (1986) were the first to suggest that residual DHT in prostatic tissue after castration was androgenic. On the basis of traditional transient transfection experiments in prostate cancer cell lines, 1 nM T efficiently activates most androgen-regulated reporter genes. Our group (Gregory et al. 2001b) and others (Culig et al. 1999) have shown that the ‘‘supersensitive’’ AR is activated in castration-recurrent prostate cancer cell lines by pM DHT. The presence of PSA in these specimens of castration-recurrent prostate cancer and in serum of patients is consistent with the presence of an activated AR, although PSA levels in castrationrecurrent prostate cancer tissue were only 7.6% of levels in benign tissue. Stege et al. (1992) reported a PSA level of 4,973 mg/g tissue (assuming 1 mg DNA/g tissue) in aspirated benign prostate, which was similar to the level we found in benign prostate (3,198 mg/g tissue). They reported a tissue PSA level of 458 mg/g tissue in prostate cancer from noncastrated patients that was similar to the level we measured for castration-recurrent prostate cancer (297 mg/g tissue). Yang et al. (1992) reported tissue PSA levels in transurethral resection specimens of 1,952.27 mg/g protein in benign prostate and 583.75 mg/g protein in prostate cancer from noncastrated patients. Since we and others obtained similar PSA levels in androgenstimulated benign prostate, the similar PSA levels measured by us in castrationrecurrent prostate cancer and others in androgen-stimulated prostate cancer and benign prostate suggest AR is activated in all tissues in spite of castrate serum levels of androgens.

5 Potential Sources of Tissue Androgens in CastrationRecurrent Prostate Cancer The substrates and metabolic pathways (Hsing et al. 2002) responsible for maintaining functional tissue levels of T and DHT in castration-recurrent prostate cancer remain to be clarified. T and DHT in recurrent prostate cancer tissue may result

560

J.L. Mohler, M.A. Titus

Fig. 2 Biosynthesis of dihydrotestosterone in prostate tissue from adrenal androgens dehydroepiandrosterone sulfate, dehydroepiandrosterone and androstenedione

from metabolism of circulating adrenal androgens (Belanger et al. 1989; Geller et al. 1978) or plasma membrane cholesterol (Schaffner 1981) located in lipid rafts (Freeman and Solomon 2004). Adrenal androgens may provide a more likely source of DHT during ADT since three enzymes (3b-hydroxysteroid dehydrogenase D5,4-isomerase, 17b-hydroxysteroid dehydrogenase, and 5a-reductase) are required to convert adrenal androgens to DHT, whereas four additional enzymes are required to synthesize DHT from cholesterol (Fig. 2). Belanger et al. (1989) suggested that persistent levels of prostatic DHT after castration alone resulted from metabolism of adrenal-derived DHEA, DHEA-SO4, and ASD in prostate tissue. Serum DHEA-SO4 levels can be 300–500 times the concentration of DHEA, and the sulfatase present in human prostate converts DHEA-SO4 to DHEA (Bartsch et al. 1990). In the only report of tissue levels of DHEA, nonhyperplastic tissue specimens obtained by open prostatectomy contained 90 pmol/ mg DNA (equivalent to 90 nM DHEA assuming 1 mg DNA/g tissue) (Bartsch et al. 1990). These levels of DHEA cause detectable activation of AR in cotransfection assays (Gregory et al. 2001a). Moreover, small amounts of DHT, formed from DHEA and DHEA-SO4 in benign prostate, have been reported (Harper et al. 1974). Some preliminary experiments support the possibility that conversion of ASD, DHEA, and DHEA-SO4 to DHT in castration-recurrent prostate cancer contributes to AR activation. Thirty six tissue homogenates were prepared from 12 samples each of frozen operative specimens of benign prostate, androgen-stimulated prostate cancer, and castration-recurrent prostate cancer. Tritiated ASD appeared as tritiated DHT in all three tissue types, which suggests that the androgen metabolic enzymes present in androgen-stimulated benign prostate and prostate cancer remain in castration-recurrent prostate cancer. The metabolic pathway that converts adrenal androgens to DHT appears to be present in the CWR-R1 cell line that was generated from a castration-recurrent CWR22 human xenograft tumor. An average of 5% of 14C-labeled DHEA appeared as 14C-labeled DHT in three experiments. These observations are consistent with recent reports (Bauman et al. 2006; Penning et al. 2006; Stanbrough et al. 2006) of upregulation of androgen metabolic

Tissue Levels of Androgens in Castration-Recurrent Prostate Cancer

561

enzymes during ADT that allow T formation from adrenal androgens or androgen metabolites. Others have suggested that additional changes in androgen metabolism allow for alternative synthesis of DHT (Stanbrough et al. 2006). ASD can be 5a reduced to 5a-androstanedione (5a-ASD) and then converted to DHT by AKR1C3 17 keto-reduction.

6 Technical Barriers to Further Investigations of Tissue Androgens Measurement of tissue androgen levels is difficult for several reasons. First, fresh prostate samples must be rapidly frozen, since tissue androgens are degraded rapidly by the effects of ischemia and prostate proteases. Prostate biopsies provide a convenient source of tissue since they are less affected by these processes; however, they are smaller in amount. Prostate tissue samples may much larger but must be protected from warm ischemia and cautery by rapid acquisition and freezing. The technique of radical prostatectomy was altered to leave the vascular pedicles intact, following which the prostate was inked for margin assessment and specimens were obtained immediately in the operating room. Transurethral prostatectomy specimens must be retrieved from the bladder immediately, trimmed of cauterized edges and frozen in the operating room. Second, the prostate, whether benign or malignant, is heterogeneous; cancer is admixed with benign tissue and the distribution among epithelia, endothelia, stroma, and inflammatory cells varies widely. Differences in androgen levels may represent differences in type of tissue if assays require large tissue samples. For example, RIA requires approximately 30 mg of high-quality tissue if one wishes to assay tissue androgens and estrogens, about 10 mg for measurement of five androgens, and as little as 2–3 mg for measurement of only T and DHT (Mohler et al. 2004). The problem of tissue heterogeneity can be overcome using laser microdissection to obtain relatively pure cell samples. However, this technique may produce thermal injury that compromises accurate measurement of tissue androgen levels and laser caps interfere with MS. Laser cutting with gravity drop into extraction buffer may minimize these problems. Either microdissection method requires a highly trained technician, the regular participation of a urological pathologist, and is tedious and time-consuming. In order to measure androgen levels in tissue samples, our group invested 3 years adapting RIA methods for measuring androgens in serum, saliva, and urine for measuring androgens using RIA in prostate tissue. The primary obstacles were proper tissue acquisition and optimal steroid extraction (Mohler et al. 2004). Small samples (like prostate biopsies) of pure cells (such as procured using laser microdissection) restricted our goal of measuring seven androgens and E2 (to profile androgen levels and androgen metabolism pathways) from 50 cell samples that has yet to be realized after 4 years invested. The electrospray tandem MS (ESI-LCMS/MS) method was validated for measurement of prostate tissue

562

J.L. Mohler, M.A. Titus

levels of T and DHT and used to confirm the T and DHT levels measured using RIA in recurrent prostate cancer (Titus et al. 2005b; Mohler et al. 2004). ESI-LCMS/MS allowed sensitive and reproducible analysis of T and DHT in prostate cancer specimens. However, the limit of detection precluded analysis of small tissue specimens, such as laser-microdissected tissues or prostate biopsies. An alternate ionization source, atmospheric pressure photoionization (APPI), has been used to allow more sensitive, and now simultaneous, measurement of T, DHT, DHEA, ASD, 5a-ASD, 5-androstenediol, androsterone, and E2. While the ESI method is suitable for separation, detection, and quantitation of C19 steroids: T, DHT, DHEA, ASD, 5a-ASD, androsterone, and 5-androstenediol, the APPI method minimizes interference and matrix effects and increases sensitivity in quantitation of T, ASD, DHT, and 5a-ASD. However, a surprising finding was that the DHT mass transition, m/z 291.2 > 255.2, was significantly decreased using APPI compared to ESI; even though all other C19 steroid mass transitions were the same (Fig. 3). DHT was ionized in the APPI source using the original T and DHT assay, and an intense peak at m/z 305.4 and its 13C satellite peak at m/z 306.5 were observed (Fig. 4a). This unique molecular ion corresponded to the peak at 11.32 min in the total ion chromatogram (Fig. 4b). The molecular ion at m/z 305.4 corresponds to addition of 15 mass units to the DHT, which was hypothesized to represent methylation of DHT in the column. The predicted mechanism of methylation involved acid-catalyzed nucleophilic substitution of methanol at either carbon 3 or carbon 17 and loss of a single water molecule. DHT ionization using APPI showed a very weak product ion at m/z 255.2 supporting methylation reactions at either the 3-keto or the 17b-hydroxy functional groups (Fig. 5). Using this method, DHT was detected as a novel methylated adduct at m/z 305.4. Further, the peak at m/z 84.9 represented a production of DHT, and therefore this product can be used as an indicator of prostate DHT levels.

Fig. 3 Reverse phase HPLC elution profile of seven steroids in minutes; androstenedione (26.1), 3,17-androstenediol (28.7), testosterone (30.3), dehydroepiandrosterone (30.8), androstanedione (33.9), dihydrotestostereone (36.2) and androsterone (38.8)

Tissue Levels of Androgens in Castration-Recurrent Prostate Cancer

563

Fig. 4 Tandem mass spectrometry DHT chromatograms using photoionization source. DHT photoionization and scan using quadrupole 1 (A). Total ion chromatogram using 40 ng of DHT (B)

The sample requirements for detection of the DHT product at m/z 84.9 were minimal, and therefore, even laser microdissected prostate biopsy samples should be able to be used for detection and quantification of DHT levels. This method should provide an improvement over RIA or GC-MS for the quantitation of anabolic steroids in prostate tissue samples. A nonprovisional patent application has been filed on October 17, 2007 (Mohler et al. 2007).

7 Conclusions AR plays an important role in the development and progression of prostate cancer and appears critical to the growth of castration-recurrent prostate cancer in spite of castrate levels of circulating testicular androgens. The high tissue levels of T and DHT in castration-recurrent prostate cancer is unexpected and suggests that these testicular androgens present a target for novel therapies. However, further study of the role of AR and intracrine metabolism of testicular androgens is made difficult by several factors. The evaluation of AR function in vitro requires highly artificial

564

J.L. Mohler, M.A. Titus

Fig. 5 Tandem mass spectrometry DHT product ions using ESI or APPI. The DHT molecular ion m/z 291.2 forms product ion m/z 255 using ESI source (A). The DHT molecular ion m/z 305.4 forms product ion m/z 84.9 using APPI source (B)

situations where AR reporter constructs must be transfected, and wild-type AR is often cotransfected. Second, the literature is confounded by careless use of terminology and difficulties in measuring androgen levels. Androgen-sensitive and androgen-independent growth are often not used carefully; androgen absence and androgen depletion are confused; and different types of ADT are often used interchangeably when they may not be equivalent. It is much easier to obtain an androgen-absent environment in vitro. However, plasticware may contain steroid hormones within its pores that may influence results; some commercially available androgen-free mediae contains low levels of androgens; and many media additives contain low levels of androgens. The difficulty of accurate measurement of low levels of androgens and the lability of androgens make direct assessment of androgen levels difficult and expensive. In addition, measuring tissue levels of AR is difficult. AR is labile and subject to tissue ischemia and fixation differences that affect immmunostaining. Quantitation is made difficult by the different means of measuring AR expression. Visual scoring systems are numerous and complex. AR measurement using image analysis suffers the challenge of segmenting and classifying malignant nuclei from complex immunohistochemical images and accurate measurement of AR mean optical density. Study of clinical specimens

Tissue Levels of Androgens in Castration-Recurrent Prostate Cancer

565

too often is done with lack of precision that may make the recognition of subtle differences in tissue androgen levels and AR expression and activation difficult. Finally, difficulties in procuring specimens of castration-recurrent prostate cancer for study, especially serially during treatment of patients with advanced prostate cancer, show progress in the field. In spite of these pitfalls and difficulties, the preclinical and clinical data all suggest that AR plays a critical central role in castration-recurrent prostate cancer and that castration-recurrent prostate cancer may not be ‘‘androgen-independent.’’ Therefore, new treatment paradigms for castration-recurrent prostate cancer may include: (1) the prevention of synthesis of tissue androgens that may be required to activate a normal AR, (2) prevention of hypersensitization of AR so that it can be activated by subphysiological levels of androgens, (3) degradation of tissue androgens formed, (4) destruction or inactivation of the AR, and (5) prevention of AR activation with improved antiandrogens. Progress will require better methods of gene transfer, new small molecule antiandrogens, better drugs that alter androgen metabolism, and more sensitive methods for measuring changes in AR protein expression, AR-regulated gene expression, and tissue androgens in small tissue specimens.

References Bartsch, W., Klein, H., Schiemann, U., Bauer, H. W., & Voigt, K. D. (1990). Enzymes of androgen formation and degradation in the human prostate. Ann N Y Acad Sci, 595, 53–66. Bauman, D. R., Steckelbroeck, S., Williams, M. V., Peehl, D. M., & Penning, T. M. (2006). Identification of the major oxidative 3alpha-hydroxysteroid dehydrogenase in human prostate that converts 5alpha-androstane-3alpha,17beta-diol to 5alpha-dihydrotestosterone: a potential therapeutic target for androgen-dependent disease. Mol Endocrinol, 20(2), 444–458. Belanger, B., Belanger, A., Labrie, F., Dupont, A., Cusan, L., & Monfette, G. (1989). Comparison of residual C-19 steroids in plasma and prostatic tissue of human, rat and guinea pig after castration: unique importance of extratesticular androgens in men. J Steroid Biochem, 32(5), 695–698. Brown, R. S., Edwards, J., Dogan, A., Payne, H., Harland, S. J., Bartlett, J. M., et al. (2002). Amplification of the androgen receptor gene in bone metastases from hormone-refractory prostate cancer. J Pathol, 198(2), 237–244. Catalona, W. J., Smith, D. S., Ratliff, T. L., Dodds, K. M., Coplen, D. E., Yuan, J. J., et al. (1991). Measurement of prostate-specific antigen in serum as a screening test for prostate cancer. N Engl J Med, 324(17), 1156–1161. Culig, Z., Hoffmann, J., Erdel, M., Eder, I. E., Hobisch, A., Hittmair, A., et al. (1999). Switch from antagonist to agonist of the androgen receptor bicalutamide is associated with prostate tumour progression in a new model system. Br J Cancer, 81(2), 242–251. de Vere White, R., Meyers, F., Chi, S. G., Chamberlain, S., Siders, D., Lee, F., et al. (1997). Human androgen receptor expression in prostate cancer following androgen ablation. Eur Urol, 31(1), 1–6. Edwards, J., Krishna, N. S., Grigor, K. M., & Bartlett, J. M. (2003). Androgen receptor gene amplification and protein expression in hormone refractory prostate cancer. Br J Cancer, 89(3), 552–556. Eisenberger, M. A., Blumenstein, B. A., Crawford, E. D., Miller, G., McLeod, D. G., Loehrer, P. J., et al. (1998). Bilateral orchiectomy with or without flutamide for metastatic prostate cancer. N Engl J Med, 339(15), 1036–1042.

566

J.L. Mohler, M.A. Titus

Feldman, B. J., & Feldman, D. (2001). The development of androgen-independent prostate cancer. Nat Rev Cancer, 1(1), 34–45. Ford, O. H., 3rd, Gregory, C. W., Kim, D., Smitherman, A. B., & Mohler, J. L. (2003). Androgen receptor gene amplification and protein expression in recurrent prostate cancer. J Urol, 170(5), 1817–1821. Freeman, M. R., & Solomon, K. R. (2004). Cholesterol and prostate cancer. J Cell Biochem, 91(1), 54–69. Geller, J., Albert, J., Loza, D., Geller, S., Stoeltzing, W., & de la Vega, D. (1978). DHT concentrations in human prostate cancer tissue. J Clin Endocrinol Metab, 46(3), 440–444. Geller, J., Albert, J., & Loza, D. (1979). Steroid levels in cancer of the prostate – markers of tumour differentiation and adequacy of anti-androgen therapy. J Steroid Biochem, 11(1B), 631–636. Gelmann, E. P. (2002). Molecular biology of the androgen receptor. J Clin Oncol, 20(13), 3001–3015. Gregory, C. W., Hamil, K. G., Kim, D., Hall, S. H., Pretlow, T. G., Mohler, J. L., et al. (1998). Androgen receptor expression in androgen-independent prostate cancer is associated with increased expression of androgen-regulated genes. Cancer Res, 58(24), 5718–5724. Gregory, C. W., Kim, D., Ye, P., D’Ercole, A. J., Pretlow, T. G., Mohler, J. L., et al. (1999). Androgen receptor up-regulates insulin-like growth factor binding protein-5 (IGFBP-5) expression in a human prostate cancer xenograft. Endocrinology, 140(5), 2372–2381. Gregory, C. W., He, B., Johnson, R. T., Ford, O. H., Mohler, J. L., French, F. S., et al. (2001a). A mechanism for androgen receptor-mediated prostate cancer recurrence after androgen deprivation therapy. Cancer Res, 61(11), 4315–4319. Gregory, C. W., Johnson, R. T., Jr., Mohler, J. L., French, F. S., & Wilson, E. M. (2001b). Androgen receptor stabilization in recurrent prostate cancer is associated with hypersensitivity to low androgen. Cancer Res, 61(7), 2892–2898. Gregory, C. W., Johnson, R. T., Jr., Presnell, S. C., Mohler, J. L., & French, F. S. (2001c). Androgen receptor regulation of G1 cyclin and cyclin-dependent kinase function in the CWR22 human prostate cancer xenograft. J Androl, 22(4), 537–548. Grossmann, M. E., Huang, H., & Tindall, D. J. (2001). Androgen receptor signaling in androgenrefractory prostate cancer. J Natl Cancer Inst, 93(22), 1687–1697. Guo, Z., Dai, B., Jiang, T., Xu, K., Xie, Y., Kim, O., et al. (2006). Regulation of androgen receptor activity by tyrosine phosphorylation. Cancer Cell, 10(4), 309–319. Harper, M. E., Pike, A., Peeling, W. B., & Griffiths, K. (1974). Steroids of adrenal origin metabolized by human prostatic tissue both in vivo and in vitro. J Endocrinol, 60(1), 117–125. Hobisch, A., Culig, Z., Radmayr, C., Bartsch, G., Klocker, H., & Hittmair, A. (1995). Distant metastases from prostatic carcinoma express androgen receptor protein. Cancer Res, 55(14), 3068–3072. Hsing, A. W., Reichardt, J. K., & Stanczyk, F. Z. (2002). Hormones and prostate cancer: current perspectives and future directions. Prostate, 52(3), 213–235. Huggins, C., & Hodges, C. V. (2002). Studies on prostatic cancer: I. The effect of castration, of estrogen and of androgen injection on serum phosphatases in metastatic carcinoma of the prostate. 1941. J Urol, 168(1), 9–12. Jemal, A., Siegel, R., Ward, E., Murray, T., Xu, J., & Thun, M. J. (2007). Cancer statistics, 2007. CA Cancer J Clin, 57(1), 43–66. Labrie, F., Dupont, A., Belanger, A., Cusan, L., Lacourciere, Y., Monfette, G., et al. (1982). New hormonal therapy in prostatic carcinoma: combined treatment with an LHRH agonist and an antiandrogen. Clin Invest Med, 5(4), 267–275. Linja, M. J., Savinainen, K. J., Saramaki, O. R., Tammela, T. L., Vessella, R. L., & Visakorpi, T. (2001). Amplification and overexpression of androgen receptor gene in hormone-refractory prostate cancer. Cancer Res, 61(9), 3550–3555.

Tissue Levels of Androgens in Castration-Recurrent Prostate Cancer

567

Mahajan, N. P., Liu, Y., Majumder, S., Warren, M. R., Parker, C. E., Mohler, J. L., et al. (2007). Activated Cdc42-associated kinase Ack1 promotes prostate cancer progression via androgen receptor tyrosine phosphorylation. Proc Natl Acad Sci U S A, 104(20), 8438–8443. Prostate Cancer Trialists’ Collaborative Group (1995). Maximum androgen blockade in advanced prostate cancer: an overview of 22 randomised trials with 3283 deaths in 5710 patients. Lancet, 346(8970), 265–269. Miyake, H., Pollak, M., & Gleave, M. E. (2000). Castration-induced up-regulation of insulin-like growth factor binding protein-5 potentiates insulin-like growth factor-I activity and accelerates progression to androgen independence in prostate cancer models. Cancer Res, 60(11), 3058–3064. Mizokami, A., Koh, E., Fujita, H., Maeda, Y., Egawa, M., Koshida, K., et al. (2004). The adrenal androgen androstenediol is present in prostate cancer tissue after androgen deprivation therapy and activates mutated androgen receptor. Cancer Res, 64(2), 765–771. Mohler, J. L., Morris, T. L., Ford, O. H., 3rd, Alvey, R. F., Sakamoto, C., & Gregory, C. W. (2002). Identification of differentially expressed genes associated with androgen-independent growth of prostate cancer. Prostate, 51(4), 247–255. Mohler, J. L., Gregory, C. W., Ford, O. H., 3rd, Kim, D., Weaver, C. M., Petrusz, P., et al. (2004). The androgen axis in recurrent prostate cancer. Clin Cancer Res, 10(2), 440–448. Mohler, J. L., Titus, M., Lih, F., & Tomer, K. (2007). Method for determination of DHT levels in tissue samples. Mousses, S., Wagner, U., Chen, Y., Kim, J. W., Bubendorf, L., Bittner, M., et al. (2001). Failure of hormone therapy in prostate cancer involves systematic restoration of androgen responsive genes and activation of rapamycin sensitive signaling. Oncogene, 20(46), 6718–6723. Nishiyama, T., Hashimoto, Y., & Takahashi, K. (2004). The influence of androgen deprivation therapy on dihydrotestosterone levels in the prostatic tissue of patients with prostate cancer. Clin Cancer Res, 10(21), 7121–7126. Page, S. T., Lin, D. W., Mostaghel, E. A., Hess, D. L., True, L. D., Amory, J. K., et al. (2006). Persistent intraprostatic androgen concentrations after medical castration in healthy men. J Clin Endocrinol Metab. Penning, T. M., Steckelbroeck, S., Bauman, D. R., Miller, M. W., Jin, Y., Peehl, D. M., et al. (2006). Aldo-keto reductase (AKR) 1C3: role in prostate disease and the development of specific inhibitors. Mol Cell Endocrinol, 248(1–2), 182–191. Sadar, M. D., Hussain, M., & Bruchovsky, N. (1999). Prostate cancer: molecular biology of early progression to androgen independence. Endocr Relat Cancer, 6(4), 487–502. Schaffner, C. P. (1981). Prostatic cholesterol metabolism: regulation and alteration. Prog Clin Biol Res, 75A, 279–324. Simard, J., Luthy, I., Guay, J., Belanger, A., & Labrie, F. (1986). Characteristics of interaction of the antiandrogen flutamide with the androgen receptor in various target tissues. Mol Cell Endocrinol, 44(3), 261–270. Stanbrough, M., Bubley, G. J., Ross, K., Golub, T. R., Rubin, M. A., Penning, T. M., et al. (2006). Increased expression of genes converting adrenal androgens to testosterone in androgenindependent prostate cancer. Cancer Res, 66(5), 2815–2825. Stege, R., Tribukait, B., Lundh, B., Carlstrom, K., Pousette, A., & Hasenson, M. (1992). Quantitative estimation of tissue prostate specific antigen, deoxyribonucleic acid ploidy and cytological grade in fine needle aspiration biopsies for prognosis of hormonally treated prostatic carcinoma. J Urol, 148(3), 833–837. Stewart, R. J., Panigrahy, D., Flynn, E., & Folkman, J. (2001). Vascular endothelial growth factor expression and tumor angiogenesis are regulated by androgens in hormone responsive human prostate carcinoma: evidence for androgen dependent destabilization of vascular endothelial growth factor transcripts. J Urol, 165(2), 688–693. Taplin, M. E., Bubley, G. J., Shuster, T. D., Frantz, M. E., Spooner, A. E., Ogata, G. K., et al. (1995). Mutation of the androgen-receptor gene in metastatic androgen-independent prostate cancer. N Engl J Med, 332(21), 1393–1398.

568

J.L. Mohler, M.A. Titus

Taplin, M. E., Bubley, G. J., Ko, Y. J., Small, E. J., Upton, M., Rajeshkumar, B., et al. (1999). Selection for androgen receptor mutations in prostate cancers treated with androgen antagonist. Cancer Res, 59(11), 2511–2515. Titus, M. A., Gregory, C. W., Ford, O. H., 3rd, Schell, M. J., Maygarden, S. J., & Mohler, J. L. (2005a). Steroid 5alpha-reductase isozymes I and II in recurrent prostate cancer. Clin Cancer Res, 11(12), 4365–4371. Titus, M. A., Schell, M. J., Lih, F. B., Tomer, K. B., & Mohler, J. L. (2005b). Testosterone and dihydrotestosterone tissue levels in recurrent prostate cancer. Clin Cancer Res, 11(13), 4653–4657. van der Kwast, T. H., Schalken, J., Ruizeveld de Winter, J. A., van Vroonhoven, C. C., Mulder, E., Boersma, W., et al. (1991). Androgen receptors in endocrine-therapy-resistant human prostate cancer. Int J Cancer, 48(2), 189–193. Visakorpi, T., Hyytinen, E., Koivisto, P., Tanner, M., Keinanen, R., Palmberg, C., et al. (1995). In vivo amplification of the androgen receptor gene and progression of human prostate cancer. Nat Genet, 9(4), 401–406. Yang, Y., Chisholm, G. D., & Habib, F. K. (1992). The distribution of PSA, cathepsin-D, and pS2 in BPH and cancer of the prostate. Prostate, 21(3), 201–208.

Unique Effects of Wnt Signaling on Prostate Cancer Cells: Modulation of the Androgen Signaling Pathway by Interactions of the Androgen Receptor Gene and Protein with Key Components of the Canonical Wnt Signaling Pathway Matthew J. Tanner, Elina Levina, Michael Shtutman, Mengqian Chen, Patrice Ohouo, and Ralph Buttyan

Abstract Canonical Wnt is a ligand-driven signaling pathway that augments the intracellular stability of b-catenin. b-Catenin is transcriptional coactivator protein that is best known for its role in regulating the activity of the Tcf/LEF-1 transcription factor family but it binds and regulates several other types of transcription factors including the androgen receptor (AR). Aberrant Wnt signaling can contribute to oncogenesis and this is often a factor in human colon cancer where mutations in key Wnt regulatory genes promote constitutive b-catenin stabilization in the tumor cells. This chapter reviews the evidence that Wnt signaling might also play some role in prostate cancer through its effects on the expression of androgendependent genes. These effects are mechanistically driven through a diverse set of interactions between three key molecules in the Wnt pathway, b-catenin, glycogen synthase kinase-3b, and Tcf/LEF-1, with the human AR gene and protein. For the most part, these interactions enhance AR expression or its functional activity as a transcription factor and they have the potential of augmenting the growth of prostate cancer cells even when androgen levels are within castrate range. Although the common gene mutations that drive abnormal Wnt signaling in other human cancers are not as frequently observed in human prostate tumors, gene expression profiles of prostate tumors often show overexpression of wnt ligands in the cancer cells and this may have a comparable effect on prostate cancer development or progression. Given the important role of the canonical Wnt signaling pathway in promoting oncogenesis in some human tissues as well as the specific involvement of key Wnt molecules in regulating androgen action in prostate cancer cells, R. Buttyan(*) The Division of Urology, Albany College of Medicine, Albany, NY 12208, USA

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4_24, # Springer Science + Business Media, LLC 2009

569

570

M.J. Tanner et al.

effectors of the Wnt signaling pathway likely represent unique targets to bring metastatic and castration-resistant prostate cancers under control.

1 Introduction Prostate cancer is a disease that develops and progresses under the influence of androgenic steroids. When detected in the advanced stage, prostate cancer is treated by some form of androgen deprivation (hormone) therapy (ADT) and, for most patients, ADT provides relief from symptoms of the disease and may even induce partial regression of metastatic lesions. Unfortunately, ADT is not curative and the respite in disease progression afforded by hormone therapy is usually lost within a few years from initiation even though patient androgen levels remain low. When this occurs, the tumor is considered to be hormone-refractory or castration-resistant. This is the most lethal form of the disease and the typical cause of prostate cancer-related mortality. At this time, the best evidence supports the idea that the development of castration-resistant prostate cancer is a consequence of an abnormal hyperactivity of the androgen signaling system in the castration-resistant prostate cancer cells (Grossmann et al. 2001; Mohler 2008). Therefore, the development of better treatments for this disease requires a greater understanding of the androgen signaling process and the various means through which it continues to function in castration-resistant prostate cancer cells in the low androgen environment of the hormone-treated patient. Conceptually, androgen signaling in the prostate cancer cell encompasses a relatively simple cell signaling system whose most crucial component is a protein, the androgen receptor (AR). AR is a DNA-binding transcription factor that is functionally activated by the binding of an agonistic (androgenic) steroid ligand, typically dihydrotestosterone (Roy and Chatterjee 1995). Dihydrotestosterone binds within the C-terminal domain of AR and this binding induces a conformational change in the protein that displaces it from its cytoplasmic chaperones and exposes the internal nuclear translocation and DNA-binding domain regions. Subsequently, liganded AR enters the nucleus and bind to cis-acting androgen response elements on DNA that lie relatively proximal to androgen-responsive genes. Once bound at these sites, AR attracts a number of other proteins that modify chromatin structure and enable the assembly of an active transcription complex for those genes whose transcription is upregulated by androgens. The process by which the AR protein modulates gene expression is described much more extensively in other chapters of this tome. However, this chapter considers the remarkable manner in which the androgen signaling process is influenced by cross talk with other cell signaling pathways. Some signaling pathways that work through protein kinase cascades can alter AR function by phosphorylating specific amino acid residues of the AR protein as is exemplified by the effects of mitogen-activated protein kinase signaling activation on androgen signaling (Gioeli et al. 2006; Gregory et al. 2004; Heinlein and Chang 2002; Taneja et al. 2005). Sometimes this cross talk is mediated by interaction of AR with transcriptional coactivators that are more associated with other signaling systems as is exemplified by the stimulatory effects

Unique Effects of Wnt Signaling on Prostate Cancer Cells

571

of the cAMP response element-binding protein, CBP/p300, on AR function (Heemers et al. 2008). Here, we will specifically highlight and describe the interaction of the androgen signaling pathway with the canonical Wnt signaling pathway that is best known for its role in developmental processes and embryogenesis. One of the more remarkable aspects of the interaction of the androgen signaling pathway with the Wnt pathway is the multiple nodes in which these two, very distinct signaling pathways can interact. In this regards, we will discuss the experimental evidence that AR function is modified by a direct interaction with b-catenin protein, the key molecular target of Wnt signaling activation. Furthermore, the functionality of AR can be altered by phosphorylation mediated by glycogen synthase kinase-3b (GSK-3b), a protein kinase component of the Wnt signaling cascade. Finally, we will also describe the evidence that one potential consequence of Wnt signaling, increased transcription from the Tcf/LEF-1 transcription factor family, might directly regulate AR gene expression in prostate cancer cells. Since the interactions between these key components of the Wnt signaling pathway with the androgen signaling pathway are so consequential in altering AR function or its sensitivity to other steroid hormones, we will also discuss the potential for clinical consequences of these interacting signaling pathways for patients with prostate cancer, especially those with the castration-recurrent form of the disease.

2 The Wnt Signaling Pathway and Its Role in Cell Growth, Tissue Development, and Cancer The Wnt acronym that is used to describe this signaling pathway is a hybrid term that was originally derived from two remarkably different manifestations of genetic perturbations that are linked to abnormal expression of a homologue of a gene involved in Wnt signaling (wnt-1). The wingless mutant of drosophila suffers a genetic loss of wnt-1 expression that results in adult flies that lack mature wing structures (Clevers 2006; Nusslein-Volhard and Wieschaus 1980). In contrast, a genetic perturbation of a gene, originally referred to as int-1, in mice results from the insertion of the mammary tumor virus genome adjacent to the mouse wnt-1 gene and this insertion leads to overexpression of wnt-1 in breast epithelial cells and the development of mammary tumors (Nusse et al. 1984; Nusslein-Volhard and Wieschaus 1980). These two biologically diverse situations illustrate that Wnt signaling is involved in normal embryonic development and indicate the potential for dysregulated Wnt signaling to drive oncogenesis in mammals. Additionally, these findings emphasize the remarkable conservation of key elements of the Wnt signaling pathway in the evolutionary history of higher organisms and its importance to the normal development and health of most multicellular organisms (Clevers 2006). In this chapter, we will focus on Wnt action through the ‘‘canonical’’ Wnt signaling pathway that, in the end, regulates the intracellular stability of the b-catenin protein. b-Catenin is a multifunctional protein that is often found at the inner cell membrane where it is bound to the intracellular domain of the cadherin

572

M.J. Tanner et al.

complex (Gavert and Ben-Ze’ev 2007). When bound at this site, b-catenin contributes to the intracellular adhesive functions of cadherins by helping to anchor the cells’ internal actin filaments. However, any free intracellular b-catenin that escapes from the cadherin complex is rapidly destroyed by a multiprotein degradation complex whose activity is regulated by the Wnt signaling process. In the Wntunstimulated state, the b-catenin degradation complex is assembled around a core scaffold consisting of the axin and adenomatous polyposis (APC) proteins. In addition to free b-catenin, this scaffold recruits two protein kinases, casein kinase 2 (CK2) and GSK-3b, both of which contribute to the phosphorylation of captured b-catenin. Phosphorylated b-catenin then becomes a substrate for ubiquitination mediated by the SKP1, CUL1, and F-Box (SCF) protein, b-TrpCP. Ubiquitinated b-catenin is readily captured and degraded by proteasomes and this action prevents it from serving its secondary function as a regulator of nuclear transcriptional processes (Aberle et al. 1997; Yost et al. 1996). In developmental situations, Wnt is a ligand-dependent signaling pathway that is conditionally active only in the presence of a wnt ligand. wnt ligands consist of a diverse family of secreted glycoproteins that are referred to as ‘‘wnts’’ and the wnt-1 gene is but one member of this family. The 21 known wnt gene homologues in the human genome typically contain 4–6 coding exons. The protein products encoded by these genes share 35% amino acid sequence identity with molecular weights that average 39–46 kDa (Miller 2002). All wnts have either 23 or 24 conserved cysteine residues found at similar locations throughout the family of proteins and this conformation of cysteines is required for interaction with the wnt receptors. The canonical Wnt signaling process is initiated by the binding of a wnt ligand to the extracellular domain of a protein from the Frizzled (Fzd) gene family. This binding induces engagement of the Fzd with a heterogeneous surface coreceptor protein of the LDL-receptor-related protein (LRP) family to initiate the signaling process. The juxtaposition of these coreceptors that occurs as a consequence of wnt ligand binding promotes signal transduction by capturing the Disheveled protein at the cytoplasmic interface of ligand-linked coreceptor complex (Malbon and Wang 2006). Disheveled capture leads to its phosphorylation and the subsequent recruitment of Axin to the growing subsurface complex. This may be one key event leading to canonical Wnt signaling since the sequestration of cytoplasmic Axin by the active wnt receptor complex blocks the formation of the b-catenin degradation complex that also requires Axin for its assembly and activity. Without the Axinbased scaffold to form this degradation complex, b-catenin protein escapes the capture, phosphorylation, and ubiquitination process; its intracellular levels rise sufficiently to allow it access to the nucleus where it complexes with transcription factors and influences their ability to induce transcription of target genes. In summary, the ultimate function of canonical Wnt signaling action is directed toward blocking the activity of the intracellular b-catenin degradation complex, leading to the increased stability of free b-catenin protein and a subsequent change in the pattern of genes expressed by the cell. The expression or presence of any given wnt ligand is not sufficient evidence of activation of canonical Wnt signaling in a cell. There are several other factors that

Unique Effects of Wnt Signaling on Prostate Cancer Cells

573

affect the ability of the canonical Wnt signaling pathway to be conditionally activated by wnt ligands. These include the presence of certain polypeptides that can block the access of wnt ligands to their cell surface Fzd receptors (Kawano and Kypta 2003). Prominent among these Wnt inhibitory proteins are the soluble Frizzled-related proteins that consist of secreted, truncated forms of the membrane-anchored Fzd receptor proteins. They retain the wnt-binding site of the membrane-bound Fzd protein so, if present in sufficient concentration, they can sequester active wnt ligands within the extracellular space and prevent their ability to engage the functional Fzd proteins in the cell membrane. Likewise, secreted Cerberus or Wnt inhibitory factor serves similar function though they are not standard homologues of the Fzd gene family. Finally, another secreted protein, Dickkopf-1 (Dkk-1), can abrogate a cells’ ability to activate canonical Wnt signaling by stimulating the endocytosis of LRP protein needed to form the active wnt receptor complex (Bafico et al. 2001). In some situations, a reduction in the expression of these wnt inhibitors is sufficient to activate Wnt signaling (Niehrs 2006). Another consideration is related to the conformation of Fzd proteins that are present on the cell surface. Some proteins of the Fzd family are more likely to signal through a separate ‘‘noncanonical’’ pathway that is driven by changes in intracellular Ca2+ ion levels and phospholipase D activation (Widelitz 2005). The noncanonical Wnt pathway does not stabilize b-catenin and will not be discussed further here except to mention that an active noncanonical Wnt signaling process can suppress the ability of the canonical Wnt signaling process to be activated in the same cell. For this reason, the repertoire of expressed Fzd proteins can also influence whether a cell is susceptible to activation of canonical Wnt signaling. In developmental situations, Wnt signaling is an archetypal paracrine process wherein different cell types (epithelium or mesenchyme) in developing tissues cross talk by means of secretion of various forms of wnt ligands or wnt inhibitors. In contrast, abnormal activation of Wnt signaling in oncogenic situations is more often associated with mutations in key Wnt signaling genes in the cancer cell that constitutively drive Wnt signaling in an autocrine fashion (Lustig and Behrens 2003). This is best understood in colon cancer where frequent mutations in the b-catenin or APC gene of the cancer cell render b-catenin protein refractory to the phosphorylation and degradation process. These types of mutations constitutively increase b-catenin protein levels in the cancer cells leading to changes in gene expression driven by the abnormally high levels of b-catenin. Finally, stabilization of b-catenin can also be achieved by direct suppression of GSK-3b kinase activity without the participation of any of the upstream Wnt regulatory molecules (Kockeritz et al. 2006). This is the mechanism through which a simple salt (lithium chloride) stabilizes b-catenin protein (Grimes and Jope 2001; Liao et al. 2004) and mimics an active Wnt signaling process. Likewise, certain protein kinases associated with other signaling pathways, such as Akt or Lkb1, that recognize GSK-3b as a substrate and inactivate it also can induce b-catenin stabilization (Mulholland et al. 2006; Ossipova et al. 2003) through wnt ligand-independent mechanisms. Since these alternate b-catenin-stabilizing signaling pathways may have some role in prostate cancer, we will briefly discuss them in later sections.

574

M.J. Tanner et al.

As was mentioned, the b-catenin stabilization process driven by canonical Wnt signaling is a potent effector of gene expression due the ability of b-catenin to bind and modulate the functional activity of several nuclear transcription factor families. The most well-known transcription factor family that is influenced by bcatenin binding is that of the HMG box nuclear transcription factors of the T-cell factor (Tcf)/lymphoid-enhancing factor-1 (LEF-1) family (Arce et al. 2006; Lustig and Behrens 2003). Transcription factors of this family are nuclear DNA-binding proteins that are normally bound (even in the Wnt uninduced state) to a cis-acting Tcf response element (TRE) along with a protein of the Groucho gene family that acts as a potent suppressor of Tcf/LEF-1 transcriptional activity. Stabilized bcatenin protein displaces Groucho from this complex and transactivates Tcfmediated transcription by assisting in the capture of other proteins such as Brg-1 and CBP that are needed for chromatin relaxation and formation of an active transcription complex. The known multitude and prominent functions of the downstream gene targets of active Tcf/LEF transcription factors give insight into the reasons why the Wnt signaling pathway is so frequently invoked in both developmental and oncogenic processes. These targets include genes that influence cell growth or cell cycle progression (cyclin D1, c-myc, jun, FGF-2, -4, -9, and -18), cell migratory and invasive behavior (stromelysin, uPAR, MMP-7, and -26), differentiation processes (BMP-4, PPAR-d, CD44, Id2, Runx2, and Ret), cell survival (survivin), and other genes whose expression is associated with the aggressive behavior of tumor cells (endothelin-1, Cox-2) (Tetsu and McCormick 1999; Zhang et al. 2001). Wnt signaling that promotes Tcf transcription is also important for stem cell biology since other Tcf-target genes (Jagged 1, nanog, Sox-9, and -2) influence stem cell maintenance and self-reproduction (Lustig and Behrens 2003; Yi and Merrill 2007). More pertinent to prostate cancer, another downstream target gene of activated Tcf transcription is the AR gene itself (Yang et al. 2006) and we will discuss the evidence for this in a later section. The prominent role of the AR in prostate cancer biology then identifies one reason that the study of the Wnt signaling process is pertinent to this tumor system. Although most treatises on canonical Wnt signaling cite the primacy of bcatenin-driven, Tcf/LEF-1-dependent transcriptional processes for its effects, there are several other transcription factors whose activity can be modified by bcatenin binding. They include Pitx2, Prop1, FoxO, and even NF-kB (Kim et al. 2005; Kioussi et al. 2002; Olson et al. 2006). More relevant to this chapter is the extensive body of literature cited below that shows that b-catenin is also a potent coactivator of AR transcriptional function and a modulator of prostate cancer growth driven by androgenic steroids. Indeed, three of the most important components of the canonical Wnt signaling process, the b-catenin protein, the Tcf/LEF transcription factors, and GSK-3b, have each proven to interact independently with the androgen signaling system in prostate cancer cells and, by altering the expression and the activity of AR in these cells, they can modulate the androgenic responsiveness of prostate cancer cells and their ability to grow under androgendeprived conditions. These interactions have the potential for significant consequences leading to activation of the androgen signaling process in prostate cancer cells even when androgen levels are very low.

Unique Effects of Wnt Signaling on Prostate Cancer Cells

575

3 b-Catenin, a Coactivator of Liganded AR Transcriptional Function in Prostate Cancer Cells The end result of activation of the canonical Wnt signaling pathway is b-catenin stabilization and accumulation. This action makes b-catenin available for binding and transcriptional coactivation of Tcf/LEF-1 or any of the other transcription factors with which it interacts. AR is one of the alternate transcription factors whose activity is altered by interaction with b-catenin but all current evidence indicates that this interaction requires that the AR be already engaged by a steroid ligand. Direct binding between liganded AR and b-catenin has been demonstrated by yeast-2-hybrid, coimmunoprecipitation, and nuclear colocalization assays (Pawlowski et al. 2002; Truica et al. 2000; Yang et al. 2002). The b-catenin-binding domain on AR is within the C-terminal domain, specifically within the activation function-2 region (Song et al. 2003). Since ligand binding also occurs within the AR C-terminal domain, the selective interaction of b-catenin with liganded AR suggests the need for a conformational shift in the AR protein associated with ligand binding to expose the b-catenin-binding site. The AR-binding region on b-catenin has been localized to the N-terminus that contains first set of armadillo repeats (Yang et al. 2002). In contrast to Tcf/LEF transcription factors that require b-catenin for their transcriptional activity, AR is transcriptionally active when bound to androgen, with or without b-catenin. However, binding of b-catenin to liganded AR significantly enhances the basal transcriptional efficiency of AR and, consequently, increases the expression of androgen-regulated gene products such as PSA in prostate cancer cells even when androgen levels are low (Masiello et al. 2004). The ability of b-catenin to enhance AR-mediated gene expression has also been repeatedly confirmed by studies involving androgen reporter genes (such as luciferase) that have shown that the reporter gene is expressed at much higher levels in prostate cancer cells when they were cotransfected with a stabilized b-catenin expression vector (Chesire et al. 2002; Mulholland et al. 2002; Song and Gelmann 2005; Truica et al. 2000). With regards to the mechanism of this effect, the activator domain regions of b-catenin, when bound to AR, more strongly attract a repertoire of transcription activating accessory proteins to the AR DNA-binding site on DNA, including GRIP-1 and CARM-1 (Koh et al. 2002; Song and Gelmann 2005), so bcatenin-bound AR is much more effective in assembling an active transcription complex on androgen-regulated genes than dimerized AR alone. This biochemical evidence suggests that canonical Wnt action might have a role in the development of castration-resistant prostate cancers in standard ADT-treated patients. For these patients, high level of stabilized b-catenin might enable the cancer cells to adapt to the low levels of androgen that remain in these patients despite treatment. Moreover, the ability of b-catenin activation domains to capture transcriptional accessory proteins also drastically affects the ability of AR liganded to an androgen antagonist (bicalutamide) to induce transcription. In the absence of b-catenin, bicalutamide-liganded AR is transcriptionally inactive, but in the presence of

576

M.J. Tanner et al.

b-catenin, this complex becomes transcriptionally active (Truica et al. 2000; Varambally et al. 2005). This finding indicates that canonical Wnt activity in prostate cancer cells might limit the potential for the success of bicalutamide, or other antiandrogens as a therapy for advanced prostate cancer patients. The AR is not the only member of the nuclear steroid transcription factor that interacts with b-catenin; both glucocorticoid and estrogen receptors (Kouzmenko et al. 2004; Pawlowski et al. 2002; Takayama et al. 2006) are also similarly affected by b-catenin binding. Regardless, the extensive evidence that b-catenin enhances androgen signaling through its interaction with the AR supports the idea that Wnt signaling has particular consequences for prostate tumor development or its progression to a therapeutically resistant state that goes beyond its usual role in the development of other human tumor systems that are associated with Wnt activation. It is of further interest that a truncated b-catenin (lacking an activation domain) was shown to strongly inhibit AR-mediated transcription, presumably by binding to AR but preventing the formation of the active AR transcription complex on genes (Song and Gelmann 2005). This interaction might inevitably be exploited as a means to inhibit AR action in prostate cancer cells. Given the propensity of the AR protein to bind and sequester b-catenin, the idea that high levels of AR protein in prostate cancer cells might compete with Tcf/LEF or with other b-catenin-binding transcription factors is not unexpected. This was confirmed by a study showing that the basal nuclear activity of Tcf was diminished by increasing the expression levels of AR protein in prostate cancer cells (Chesire and Isaacs 2002). The significance of this latter observation is somewhat obscured by observations of a potential direct binding interaction between the AR protein and one particular member of the Tcf transcription factor family (Tcf-4) that blocks its ability to induce transcription of Tcf-dependent genes (Amir et al. 2003). The complex and competitive interactions between the AR, b-catenin, and Tcf transcription factor proteins indicate that the basal AR- and TCF/LEF transcriptional activities within any given prostate cancer cell depend upon the levels of expression of each of these three molecules at any given time as well as the regional concentrations of androgenic steroids that permit the binding of b-catenin to AR protein.

4 GSK-3b Protein Kinase, a Wnt Signaling Intermediate Molecule, and Its Effects on AR Activity in Prostate Cancer GSK-3b is a serine-threonine protein kinase that participates in the b-catenin degradation complex (Grimes and Jope 2001; Kockeritz et al. 2006). By phosphorylating b-catenin at the serine-6 residue, GSK-3b (in the Wnt-inactive state) allows the recognition of b-catenin protein by the cellular ubiquitinylation complex and leads to b-catenin degradation by the proteasome. The upstream activity of canonical Wnt signaling leads to GSK-3b phosphorylation and this critical modification blocks its kinase activity and its ability to phosphorylate b-catenin. Without phosphorylation by GSK-3b, the b-catenin protein is spared from the degradation process and b-catenin levels increase. Since b-catenin is a coactivator of AR, this is one of the

Unique Effects of Wnt Signaling on Prostate Cancer Cells

577

means through which GSK-3b affects AR activity in prostate cancer cells. It is important to remember, however, that b-catenin is just one of the substrates for GSK-3b kinase action. In fact, GSK-3b is an intermediate protein kinase in other cellular signaling pathways including PI-3 kinase (see below). Indeed, one of the other known downstream targets of GSK-3b kinase activity is the AR protein itself (Kockeritz et al. 2006; Salas et al. 2004). For the most part, GSK-3b-mediated AR phosphorylation has been reported to inhibit AR transcriptional activity and it was proposed that GSK-3b-mediated phosphorylation impairs the functional interaction between the N- and C-terminal domains of the AR protein that is required for transcriptional activation (Li et al. 2004; Salas et al. 2004; Wang et al. 2004). However, there are contradictory reports that GSK-3b phosphorylation of AR enhances its transcriptional activity (Hall et al. 2005; Mazor et al. 2004), so the final judgment of how GSK-3b action ultimately affects androgen signaling is controversial and remains an issue for further research. Regardless as to whether GSK-3b phosphorylation is driven by canonical Wnt or not, this issue is an important matter for prostate cancer since the disease is also frequently associated with highly active Akt/protein kinase B as a consequence of PTEN gene loss or, in androgen-deprived states, with hyperactivity of a relaxin-driven signaling pathway (Dong 2006; Liu et al. 2008; Mulholland et al. 2006). GSK-3b protein is one of the downstream targets of activated Akt, so one might expect that hyperactive Akt acts similar to an active Wnt signaling process by stabilizing b-catenin as well as by suppressing the GSK-3b-mediated phosphorylation of AR. If these two actions of GSK-3b suppression have opposing effects (i.e., high levels of b-catenin promote androgen signaling whereas phosphorylation of AR suppresses androgen signaling), the consequences may be far different than if they are synergistic. This area certainly deserves more research to clarify the effects of GSK-3b expression and action on androgen signaling in prostate cancer but it may require a cell or biochemical system that is not complicated by the presence of b-catenin to make this determination.

5 Wnt Signaling and the Expression of the AR Gene By stabilizing b-catenin, canonical Wnt signaling can induce the expression of gene products dependent on the Tcf/LEF family of transcription factors. The human AR gene is one of the target genes transcriptionally upregulated by active Tcf/LEF-1, at least in cultured human prostate cancer cells. The proximal 50 promoter of the human AR gene contains a canonical TRE (at 1,160 to 1,166) (Yang et al. 2006). This region of the hAR gene promoter was specifically immunoprecipitated in a CHIP assay using an antibody that recognizes b-catenin only from fixed chromatin of LNCaP cells that had active Wnt signaling or stabilized b-catenin. The ability of stabilized b-catenin to act on the hAR gene promoter was also confirmed with the use of a chimeric hAR promoter-luciferase reporter expression constructs. Mutations introduced into the TRE of these chimeric vectors blocked the ability of stabilized b-catenin to upregulate luciferase. The description of an active TRE within the 50 promoter region of hAR is also consistent with the ability

578

M.J. Tanner et al.

of stabilized b-catenin or Wnt activation to upregulate AR mRNA expression in the LNCaP cells. These observations suggest that activation of Tcf/LEF-1 transcriptional activity in prostate cancer cells should also upregulate AR protein levels. Despite this expectation, AR protein levels were found to be drastically reduced by Wnt activation or stabilized b-catenin in the LNCaP cells (Yang et al. 2006). The paradoxical effects of Wnt/stabilized b-catenin upregulating expression of AR mRNA but downregulating AR protein levels were associated with the activation of an MDM2- and proteasome-dependent AR protein degradation process that appears to be dependent upon active Tcf/LEF-1 transcription since a dominantnegative form of Tcf was able to block the AR protein degradation process in the presence of stabilized b-catenin. This effect in one of the most commonly utilized human prostate cancer cell lines lends support to the idea that there can be extensive antagonism between Tcf/LEF-1dependent and AR-dependent transcriptional processes in prostate cancer cells that is more complex than a direct competition of these two types of transcription factors for b-catenin binding. In the presence of androgens, AR preferentially captures available b-catenin and this action enhances AR-mediated transcription in the cancer cells while suppressing the transcriptional activity of endogenous Tcf/LEF-1. In the acute absence of androgens, however, Tcf/LEF-1 seems able to capture the available b-catenin and its transcriptional activation as a consequence of this capture drives a further reduction of AR protein expression that would severely disable the androgen signaling system in these cells. Here we also need to consider that, with chronic androgen deprivation, the diminished AR protein levels in cultured LNCaP cells eventually rise again to levels higher than in parental cells grown in the presence of androgens and that this reexpression of AR protein coincides with restoration of the cell proliferation process despite the continued lack of androgenic steroids. The cells that transition through the androgen-deprived period are referred to as ‘‘androgen-insensitive’’ and these cells are sometimes considered to be models for the castration-resistant tumors that develop in prostate cancer patients chronically treated with ADT. Given what we know now regarding the effects of active Wnt on AR protein expression, one might speculate that the transition to the androgen-insensitive state is associated with a suppression of the AR protein degradation process driven by Wnt signaling. Certainly, this is one novel area of androgen biology that, with further research, might help us better understand the process involved in the development of castration-resistant prostate cancer.

6 Wnt Signaling in Normal and Malignant Prostate Epithelial Cell Growth: A View from Animal Model Systems Here we consider the in vivo evidence obtained from animal model systems that the activity of the canonical Wnt signaling pathway influences the growth of normal and malignant prostate epithelial cells in conjunction with androgens. With regards to normal prostate cells, the evidence of this relationship derives from the study of rodent (rat and mouse) prostate systems that are often used as models of androgendependent prostate epithelial cell growth and survival. Rodent ventral prostates are

Unique Effects of Wnt Signaling on Prostate Cancer Cells

579

exquisitely sensitive to androgens and, in the adult state, will quickly involute following castration in conjunction with apoptosis of most of the prostate epithelial cells. Once regressed, the rudimentary rodent prostate tissues can be restimulated to grow back to their original state by replenishing androgens. In the rat, at least, this androgen-stimulated regrowth is accompanied by a significant increase in the nuclear b-catenin (Chesire et al. 2002) of prostate epithelial cells consistent with idea that androgen replenishment is activating a canonical Wnt signaling process in the growing distal buds that coincides with the proliferative activity of the epithelial cells in this region. In the mouse, where a transgenic Tcf-dependent reporter (b-galactosidase) variant is available, we have also recently learned that a Wnt signaling process is transiently induced in the prostate by castration, but in this situation, it is restricted to the proximal duct regions of the prostate that survive involution (Placencio et al. 2008). These observations are consistent with the notion that Wnt signaling contributes to the proliferative activity of normal androgendependent prostate cells and that Wnt supports the survival of an androgen-independent subset of prostate epithelial cells following castration. With regards to a potential role for aberrant Wnt signaling in prostate malignancy, genetically manipulated mice in which prostate-specific b-catenin stabilization is driven by a stabilizing mutation in b-catenin (Bierie et al. 2003; Gounari et al. 2002) or by knockout of APC or Lkb-1 genes (Bruxvoort et al. 2007; Pearson et al. 2008) develop a broad spectrum of prostate growth abnormalities including hyperplasia, neoplasia, and anaplasia. These mouse models then lend support to the idea that dysregulation of canonical Wnt signaling in the prostate is sufficient to promote the development of prostate cancer in a manner similar to that reported in other tissues. Unfortunately, these particular mouse models have not yet defined the extent to which chronic b-catenin stabilization might be driving the abnormal prostate growth process through its effects on AR or on other b-catenin-sensitive transcription factors (such as Tcf/LEF-1). For the future, it could be of great interest to determine whether the prostate AR overexpressing transgenic mouse model, when crossed to the b-catenin stabilized prostate mouse mentioned above, develops overt prostate cancer or whether a prostate-targeted transgenic model in which a constitutively active form of Tcf is overexpressed in prostate epithelial cells is prone to the development of neoplastic prostate pathology.

7 Abnormal Wnt Activity in Human Prostate Cancer The idea that dysregulated Wnt signaling functionally contributes to human prostate cancer is supported by the identification of b-catenin-stabilizing mutations in, at least, a minor fraction of human prostate cancer specimens. Mutations in the b-catenin or the APC gene that are similar to those found so frequently in colon cancer have been found in some human prostate tumors, but only in a small minority of specimens, typically 5% or less (Gerstein et al. 2002; Voeller et al. 1998). The relative infrequency with which these particular b-catenin-stabilizing mutations are detected in prostate cancer suggests that there may be different

580

M.J. Tanner et al.

classes of prostate tumors, some are Wnt-driven while others are not, or that there may be alternative means for stabilizing b-catenin in prostate cancer. One possibility is that b-catenin stabilization in prostate cancer occurs as a consequence of mutations in genes other than those frequently mutated in colon cancer. In this regards, there has been one report of a mutation in the b-TrpCP gene in prostate cancer (Gerstein et al. 2002) that might stabilize b-catenin and such a mutation has never been previously reported in colon cancer. Aberrant Wnt activation in some prostate cancers might also occur in the context of generalized hyperactivity of some of the wnt ligand genes in tumor cells. In this regards, there are several published reports that some of the wnt ligands (wnt1, 2, 5a, and 11) are overexpressed in human prostate tumor specimens and that an inhibitor of canonical Wnt signaling, Wnt inhibitory factor-1, is underexpressed when compared to normal human prostate epithelial cells (Chen et al. 2004; Dhanasekaran et al. 2001; Glinsky et al. 2004; Stanbrough et al. 2006; Tomlins et al. 2006; Wissmann et al. 2003). A search of the publicly available expression microarray databank, Oncomine, reveals additional oligonucleotide microarray-based studies that support these observations (Varambally et al. 2005; Yu et al. 2004). In summary, based upon the cumulative evidence to date, we know that there is, at least, a minority of human prostate tumors that have mutations in b-catenin-stabilizing genes and a seeming larger fraction of prostate tumors that might stabilize b-catenin through wnt ligand overexpression or through wnt inhibitor gene downregulation. Since the hallmark of a hyperactive canonical Wnt signaling pathway is increased b-catenin protein expression and nuclear localization, one might think that immunohistochemical surveys of human normal and neoplastic prostate tissues with a b-catenin antibody would be an effective means to establish a relationship between abnormal Wnt signaling and the natural history of human prostate cancer. Unfortunately, the literature describing these surveys can be confusing, especially since some of these surveys involve only localized prostate cancers whereas others include specimens of metastatic or hormone-refractory cancer tissues. There tends to be a trend of reports describing lowered overall b-catenin or nuclear b-catenin in localized prostate cancers with poor correspondence to tumor grade (Aaltomaa et al. 2005; Horvath et al. 2005; Jaggi et al. 2005; Whitaker et al. 2008) and this is contrasted to most of the studies on metastatic or hormone-refractory prostate cancers that report a relatively high level of nuclear b-catenin immunostaining (Chesire et al. 2002; de la Taille et al. 2003; Saha et al. 2008; Yardy et al. 2008). Given these overall trends, it is tempting to speculate that Wnt signaling is much more a factor involved in metastatic aggressive or hormone-refractory prostate cancers than for localized cancers in hormone untreated patients. This conclusion concurs with the vast body of basic research literature, some described in previous sections, that has identified a role for b-catenin stabilization in cell motility/ metastasis, survival, and hormonal-responsiveness. Despite the current uncertainty as to the extent to which b-catenin stabilization is involved in the development of human prostate cancer, there is also a growing body of experimental research that supports the idea that the activity of the Wnt signaling pathway contributes to one of the more important biological consequences of prostate

Unique Effects of Wnt Signaling on Prostate Cancer Cells

581

cancer, the propensity of prostate cancer cells to metastasize to bone and to influence regional bone development. Here, Keller and coworkers have identified a unique relationship between the expression of the wnt inhibitor, Dkk-1, by prostate cancer cells and the ability of these cells to colonize and subsequently induce a regional osteoblastic response that is a feature of metastatic prostate cancer in bone (Chesire et al. 2002; Hall et al. 2005). In their proposed schema, bone colonization (establishment of metastasis) requires initial osteoclast activation and osteolysis that is driven by localized Wnt suppression from seeding metastatic tumor cells that secrete high levels of the Wnt inhibitor, Dkk-1. However, once bone is successfully colonized by the cancer cells, cancer cell Dkk-1 expression falls and the initial Wnt-suppressive environment of the micrometastasis is changed to a Wnt-promoting environment that effectively stimulates the activity of the regional osteoblasts. Currently, the mechanism through which Dkk-1 expression becomes turned down during bone metastatic progression of prostate cancer is not known, but it may involve a paracrine signaling process in which reciprocal signals from bone cells alter the behavior of the invading prostate cancer cells (Liu et al. 2007).

Summary: The Significance of the Interaction Between the Wnt and Androgen Signaling Pathways for Prostate Cancer Patients We have described here the basic research evidence that the canonical Wnt signaling pathway has the potential for extensive cross talk with the androgen signaling pathway in prostate cancer cells and that the outcome of this complex cross talk regulates the expression of the AR gene as well as the activity of the AR protein (Fig. 1). The manner in which the Wnt/androgen cross talk affects the prostate cancer cell depends very much on the steroid milieu. In this context, we have discussed the possibility that Wnt signaling might contribute to the beneficial effects of ADT in advanced prostate cancer patients by transiently promoting AR protein degradation in the cancer cells. However, with chronic androgen deprivation, the suppressive effects of Wnt on AR protein expression appear to diminish and, as AR levels rise in the cancer cells, Wnt may become one of the more important factors for the development of castration-resistant prostate cancer by sensitizing the AR to the low androgen levels of an ADT-treated patient. Finally, Wnt seems to have a role in the process through which prostate cancer cells metastasize to bone and induce osteoblastic lesions. Rodent prostate models have provided evidence that canonical Wnt signaling increases during androgen-driven proliferation of normal prostate epithelial cells and that Wnt signaling also increases in prostate cells that survive castration. These model systems link the experimental studies of Wnt/androgen cross talk to the in vivo response of prostate cells to androgens. Finally, transgenic and knockout mouse models have established the potential for a role of abnormal Wnt signaling in prostatic neoplasia.

582

M.J. Tanner et al.

Cytosol DHT

G GSK-3b

AR 2 AR

MDM2

AR

MDM2 --P

Akt --P

3

1 AR -ub

Akt

AR

b Cat

Proteasome Degradation

b Cat Tcf TRE

AR mRNA (+)

b Cat

Androgen-Regulated mRNA (-/ + ?)

AR ARE

Androgen Receptor Gene

AR --P

AR AR P---P ARE

Androgen-Regulated mRNA (+)

Nucleus Fig. 1 Graphical characterization of the three known interactive nodes between the Wnt and the androgen signaling pathways in prostate cancer cells: (1) Describes the interaction between b-catenin protein, which is stabilized by Wnt signaling, with liganded AR protein that leads to increased transcription of AR target genes; (2) Describes the effects of Tcf/Lef-1 transcription factors that are activated by Wnt-stabilized b-catenin on AR gene transcription and the stability of the AR protein. Here Tcf-mediated transcription enhances the expression of AR mRNA but cross talk of Wnt to Akt leads to phosphorylation and activation of MDM2 that promotes degradation of the AR protein; (3) Describes the interaction of GSK-3b that is suppressed by Wnt signaling with AR protein that leads to AR modification (phosphorylation) and modulation of AR transcriptional activity

There remains the questions as to the actual extent to which Wnt signaling is abnormal in human prostate malignancy as well as the mechanisms that might be driving abnormal Wnt signaling in human prostate cancer. Mutations in key Wnt regulatory genes are not as common in prostate cancer as they are in other Wntdriven tumor systems. Perhaps this is because Wnt does not appear to be so prominent a factor for prostate cancer development as it does for progression to therapeutic resistant states. In this regards, one should consider the immunohistochemical evidence discussed earlier that nuclear b-catenin levels are generally reported to be lower in localized, untreated prostate cancer as compared to normal prostate epithelial cells as well as the recent reports that suppression of a Wnt signaling environment by stromal cells secreting wnt inhibitors actually increases the propensity of immortalized prostate epithelial cells to undergo malignant transformation. Furthermore, there are other, more frequent genetic aberrations in prostate cancer (exemplified by the loss of the PTEN gene), which, in the end, may serve a similar function (stabilization of b-catenin) to canonical Wnt. Regardless,

Unique Effects of Wnt Signaling on Prostate Cancer Cells

583

there is much stronger evidence that Wnt signaling is abnormally high in metastatically aggressive or therapeutic-resistant (castration-resistant) prostate cancers and this evidence supports the idea that abnormal Wnt signaling is a useful target for development of therapies to treat metastatic and castration-resistant prostate cancer.

References Aaltomaa, S., V. Karja, P. Lipponen, T. Isotalo, J. P. Kankkunen, M. Talja, and R. Mokka. 2005. Reduced alpha- and beta-catenin expression predicts shortened survival in local prostate cancer. Anticancer Res 25(6C): 4707–12. Aberle, H., A. Bauer, J. Stappert, A. Kispert, and R. Kemler. 1997. beta-Catenin is a target for the ubiquitin-proteasome pathway. Embo J 16(13): 3797–804. Amir, A. L., M. Barua, N. C. McKnight, S. Cheng, X. Yuan, and S. P. Balk. 2003. A direct betacatenin-independent interaction between androgen receptor and T cell factor 4. J Biol Chem 278(33): 30828–34. Arce, L., N. N. Yokoyama, and M. L. Waterman. 2006. Diversity of LEF/TCF action in development and disease. Oncogene 25(57): 7492–504. Bafico, A., G. Liu, A. Yaniv, A. Gazit, and S. A. Aaronson. 2001. Novel mechanism of Wnt signalling inhibition mediated by Dickkopf-1 interaction with LRP6/Arrow. Nat Cell Biol 3(7): 683–6. Bierie, B., M. Nozawa, J. P. Renou, J. M. Shillingford, F. Morgan, T. Oka, M. M. Taketo, R. D. Cardiff, K. Miyoshi, K. U. Wagner, G. W. Robinson, and L. Hennighausen. 2003. Activation of beta-catenin in prostate epithelium induces hyperplasias and squamous transdifferentiation. Oncogene 22(25): 3875–87. Bruxvoort, K. J., H. M. Charbonneau, T. A. Giambernardi, J. C. Goolsby, C. N. Qian, C. R. Zylstra, D. R. Robinson, P. Roy-Burman, A. K. Shaw, B. D. Buckner-Berghuis, R. E. Sigler, J. H. Resau, R. Sullivan, W. Bushman, and B. O. Williams. 2007. Inactivation of Apc in the mouse prostate causes prostate carcinoma. Cancer Res 67(6): 2490–6. Chen, G., N. Shukeir, A. Potti, K. Sircar, A. Aprikian, D. Goltzman, and S. A. Rabbani. 2004. Up-regulation of Wnt-1 and beta-catenin production in patients with advanced metastatic prostate carcinoma: potential pathogenetic and prognostic implications. Cancer 101(6): 1345–56. Chesire, D. R., C. M. Ewing, W. R. Gage, and W. B. Isaacs. 2002. In vitro evidence for complex modes of nuclear beta-catenin signaling during prostate growth and tumorigenesis. Oncogene 21(17): 2679–94. Chesire, D. R., and W. B. Isaacs. 2002. Ligand-dependent inhibition of beta-catenin/TCF signaling by androgen receptor. Oncogene 21(55): 8453–69. Clevers, H. 2006. Wnt/beta-catenin signaling in development and disease. Cell 127(3): 469–80. de la Taille, A., M. A. Rubin, M. W. Chen, F. Vacherot, S. G. de Medina, M. Burchardt, R. Buttyan, and D. Chopin. 2003. Beta-catenin-related anomalies in apoptosis-resistant and hormone-refractory prostate cancer cells. Clin Cancer Res 9(5): 1801–7. Dhanasekaran, S. M., T. R. Barrette, D. Ghosh, R. Shah, S. Varambally, K. Kurachi, K. J. Pienta, M. A. Rubin, and A. M. Chinnaiyan. 2001. Delineation of prognostic biomarkers in prostate cancer. Nature 412(6849): 822–6. Dong, J. T. 2006. Prevalent mutations in prostate cancer. J Cell Biochem 97(3): 433–47. Gavert, N., and A. Ben-Ze’ev. 2007. beta-Catenin signaling in biological control and cancer. J Cell Biochem 102(4): 820–8. Gerstein, A. V., T. A. Almeida, G. Zhao, E. Chess, M. Shih Ie, K. Buhler, K. Pienta, M. A. Rubin, R. Vessella, and N. Papadopoulos. 2002. APC/CTNNB1 (beta-catenin) pathway alterations in human prostate cancers. Genes Chromosomes Cancer 34(1): 9–16. Gioeli, D., B. E. Black, V. Gordon, A. Spencer, C. T. Kesler, S. T. Eblen, B. M. Paschal, and M. J. Weber. 2006. Stress kinase signaling regulates androgen receptor phosphorylation, transcription, and localization. Mol Endocrinol 20(3): 503–15.

584

M.J. Tanner et al.

Glinsky, G. V., A. B. Glinskii, A. J. Stephenson, R. M. Hoffman, and W. L. Gerald. 2004. Gene expression profiling predicts clinical outcome of prostate cancer. J Clin Invest 113(6): 913–23. Gounari, F., S. Signoretti, R. Bronson, L. Klein, W. R. Sellers, J. Kum, A. Siermann, M. M. Taketo, H. von Boehmer, and K. Khazaie. 2002. Stabilization of beta-catenin induces lesions reminiscent of prostatic intraepithelial neoplasia, but terminal squamous transdifferentiation of other secretory epithelia. Oncogene 21(26): 4099–107. Gregory, C. W., X. Fei, L. A. Ponguta, B. He, H. M. Bill, F. S. French, and E. M. Wilson. 2004. Epidermal growth factor increases coactivation of the androgen receptor in recurrent prostate cancer. J Biol Chem 279(8): 7119–30. Grimes, C. A., and R. S. Jope. 2001. The multifaceted roles of glycogen synthase kinase 3beta in cellular signaling. Prog Neurobiol 65(4): 391–426. Grossmann, M. E., H. Huang, and D. J. Tindall. 2001. Androgen receptor signaling in androgenrefractory prostate cancer. J Natl Cancer Inst 93(22): 1687–97. Hall, C. L., A. Bafico, J. Dai, S. A. Aaronson, and E. T. Keller. 2005. Prostate cancer cells promote osteoblastic bone metastases through Wnts. Cancer Res 65(17): 7554–60. Heemers, H. V., J. D. Debes, and D. J. Tindall. 2008. The role of the transcriptional coactivator p300 in prostate cancer progression. Adv Exp Med Biol 617535–40. Heinlein, C. A., and C. Chang. 2002. The roles of androgen receptors and androgen-binding proteins in nongenomic androgen actions. Mol Endocrinol 16(10): 2181–7. Horvath, L. G., S. M. Henshall, C. S. Lee, J. G. Kench, D. Golovsky, P. C. Brenner, G. F. O’Neill, R. Kooner, P. D. Stricker, J. J. Grygiel, and R. L. Sutherland. 2005. Lower levels of nuclear beta-catenin predict for a poorer prognosis in localized prostate cancer. Int J Cancer 113(3): 415–22. Jaggi, M., S. L. Johansson, J. J. Baker, L. M. Smith, A. Galich, and K. C. Balaji. 2005. Aberrant expression of E-cadherin and beta-catenin in human prostate cancer. Urol Oncol 23(6): 402–6. Kawano, Y., and R. Kypta. 2003. Secreted antagonists of the Wnt signalling pathway. J Cell Sci 116(Pt 13): 2627–34. Kim, J. H., B. Kim, L. Cai, H. J. Choi, K. A. Ohgi, C. Tran, C. Chen, C. H. Chung, O. Huber, D. W. Rose, C. L. Sawyers, M. G. Rosenfeld, and S. H. Baek. 2005. Transcriptional regulation of a metastasis suppressor gene by Tip60 and beta-catenin complexes. Nature 434(7035): 921–6. Kioussi, C., P. Briata, S. H. Baek, D. W. Rose, N. S. Hamblet, T. Herman, K. A. Ohgi, C. Lin, A. Gleiberman, J. Wang, V. Brault, P. Ruiz-Lozano, H. D. Nguyen, R. Kemler, C. K. Glass, A. Wynshaw-Boris, and M. G. Rosenfeld. 2002. Identification of a Wnt/Dvl/beta-Catenin ! Pitx2 pathway mediating cell-type-specific proliferation during development. Cell 111(5): 673–85. Kockeritz, L., B. Doble, S. Patel, and J. R. Woodgett. 2006. Glycogen synthase kinase-3–an overview of an over-achieving protein kinase. Curr Drug Targets 7(11): 1377–88. Koh, S. S., H. Li, Y. H. Lee, R. B. Widelitz, C. M. Chuong, and M. R. Stallcup. 2002. Synergistic coactivator function by coactivator-associated arginine methyltransferase (CARM) 1 and betacatenin with two different classes of DNA-binding transcriptional activators. J Biol Chem 277 (29): 26031–5. Kouzmenko, A. P., K. Takeyama, S. Ito, T. Furutani, S. Sawatsubashi, A. Maki, E. Suzuki, Y. Kawasaki, T. Akiyama, T. Tabata, and S. Kato. 2004. Wnt/beta-catenin and estrogen signaling converge in vivo. J Biol Chem 279(39): 40255–8. Li, H., J. H. Kim, S. S. Koh, and M. R. Stallcup. 2004. Synergistic effects of coactivators GRIP1 and beta-catenin on gene activation: cross-talk between androgen receptor and Wnt signaling pathways. J Biol Chem 279(6): 4212–20. Liao, X., J. B. Thrasher, J. Holzbeierlein, S. Stanley, and B. Li. 2004. Glycogen synthase kinase3beta activity is required for androgen-stimulated gene expression in prostate cancer. Endocrinology 145(6): 2941–9. Liu, S., R. L. Vinall, C. Tepper, X. B. Shi, L. R. Xue, A. H. Ma, L. Y. Wang, L. D. Fitzgerald, Z. Wu, R. Gandour-Edwards, R. W. deVere White, and H. J. Kung. 2008. Inappropriate activation of androgen receptor by relaxin via beta-catenin pathway. Oncogene 27(4): 499–505.

Unique Effects of Wnt Signaling on Prostate Cancer Cells

585

Liu, X. H., A. Kirschenbaum, S. Yao, G. Liu, S. A. Aaronson, and A. C. Levine. 2007. Androgeninduced Wnt signaling in preosteoblasts promotes the growth of MDA-PCa-2b human prostate cancer cells. Cancer Res 67(12): 5747–53. Lustig, B., and J. Behrens. 2003. The Wnt signaling pathway and its role in tumor development. J Cancer Res Clin Oncol 129(4): 199–221. Malbon, C. C., and H. Y. Wang. 2006. Dishevelled: a mobile scaffold catalyzing development. Curr Top Dev Biol 72153–66. Masiello, D., S. Y. Chen, Y. Xu, M. C. Verhoeven, E. Choi, A. N. Hollenberg, and S. P. Balk. 2004. Recruitment of beta-catenin by wild-type or mutant androgen receptors correlates with ligand-stimulated growth of prostate cancer cells. Mol Endocrinol 18(10): 2388–401. Mazor, M., Y. Kawano, H. Zhu, J. Waxman, and R. M. Kypta. 2004. Inhibition of glycogen synthase kinase-3 represses androgen receptor activity and prostate cancer cell growth. Oncogene 23(47): 7882–92. Miller, J. R. 2002. The Wnts. Genome Biol 3(1): 3001.1–3001.15. Mohler, J. L. 2008. A role for the androgen-receptor in clinically localized and advanced prostate cancer. Best Pract Res Clin Endocrinol Metab 22(2): 357–72. Mulholland, D. J., H. Cheng, K. Reid, P. S. Rennie, and C. C. Nelson. 2002. The androgen receptor can promote beta-catenin nuclear translocation independently of adenomatous polyposis coli. J Biol Chem 277(20): 17933–43. Mulholland, D. J., S. Dedhar, H. Wu, and C. C. Nelson. 2006. PTEN and GSK3beta: key regulators of progression to androgen-independent prostate cancer. Oncogene 25(3): 329–37. Niehrs, C. 2006. Function and biological roles of the Dickkopf family of Wnt modulators. Oncogene 25(57): 7469–81. Nusse, R., A. van Ooyen, D. Cox, Y. K. Fung, and H. Varmus. 1984. Mode of proviral activation of a putative mammary oncogene (int-1) on mouse chromosome 15. Nature 307(5947): 131–6. Nusslein-Volhard, C., and E. Wieschaus. 1980. Mutations affecting segment number and polarity in Drosophila. Nature 287(5785): 795–801. Olson, L. E., J. Tollkuhn, C. Scafoglio, A. Krones, J. Zhang, K. A. Ohgi, W. Wu, M. M. Taketo, R. Kemler, R. Grosschedl, D. Rose, X. Li, and M. G. Rosenfeld. 2006. Homeodomainmediated beta-catenin-dependent switching events dictate cell-lineage determination. Cell 125(3): 593–605. Ossipova, O., N. Bardeesy, R. A. DePinho, and J. B. Green. 2003. LKB1 (XEEK1) regulates Wnt signalling in vertebrate development. Nat Cell Biol 5(10): 889–94. Pawlowski, J. E., J. R. Ertel, M. P. Allen, M. Xu, C. Butler, E. M. Wilson, and M. E. Wierman. 2002. Liganded androgen receptor interaction with beta-catenin: nuclear co-localization and modulation of transcriptional activity in neuronal cells. J Biol Chem 277(23): 20702–10. Pearson, H. B., A. McCarthy, C. M. Collins, A. Ashworth, and A. R. Clarke. 2008. Lkb1 deficiency causes prostate neoplasia in the mouse. Cancer Res 68(7): 2223–32. Placencio, V. R., A. R. Sharif-Afshar, X. Li, H. Huang, C. Uwamariya, E. G. Neilson, M. M. Shen, R. J. Matusik, S. W. Hayward, and N. A. Bhowmick. 2008. Stromal transforming growth factor-beta signaling mediates prostatic response to androgen ablation by paracrine Wnt activity. Cancer Res 68(12): 4709–18. Roy, A. K., and B. Chatterjee. 1995. Androgen action. Crit Rev Eukaryot Gene Expr 5(2): 157–76. Saha, B., A. Arase, S. S. Imam, D. Tsao-Wei, W. Y. Naritoku, S. Groshen, L. W. Jones, and S. A. Imam. 2008. Overexpression of E-cadherin and beta-catenin proteins in metastatic prostate cancer cells in bone. Prostate 68(1): 78–84. Salas, T. R., J. Kim, F. Vakar-Lopez, A. L. Sabichi, P. Troncoso, G. Jenster, A. Kikuchi, S. Y. Chen, L. Shemshedini, M. Suraokar, C. J. Logothetis, J. DiGiovanni, S. M. Lippman, and D. G. Menter. 2004. Glycogen synthase kinase-3 beta is involved in the phosphorylation and suppression of androgen receptor activity. J Biol Chem 279(18): 19191–200. Song, L. N., and E. P. Gelmann. 2005. Interaction of beta-catenin and TIF2/GRIP1 in transcriptional activation by the androgen receptor. J Biol Chem 280(45): 37853–67.

586

M.J. Tanner et al.

Song, L. N., R. Herrell, S. Byers, S. Shah, E. M. Wilson, and E. P. Gelmann. 2003. Beta-catenin binds to the activation function 2 region of the androgen receptor and modulates the effects of the N-terminal domain and TIF2 on ligand-dependent transcription. Mol Cell Biol 23(5): 1674–87. Stanbrough, M., G. J. Bubley, K. Ross, T. R. Golub, M. A. Rubin, T. M. Penning, P. G. Febbo, and S. P. Balk. 2006. Increased expression of genes converting adrenal androgens to testosterone in androgen-independent prostate cancer. Cancer Res 66(5): 2815–25. Takayama, S., I. Rogatsky, L. E. Schwarcz, and B. D. Darimont. 2006. The glucocorticoid receptor represses cyclin D1 by targeting the Tcf-beta-catenin complex. J Biol Chem 281 (26): 17856–63. Taneja, S. S., S. Ha, N. K. Swenson, H. Y. Huang, P. Lee, J. Melamed, E. Shapiro, M. J. Garabedian, and S. K. Logan. 2005. Cell-specific regulation of androgen receptor phosphorylation in vivo. J Biol Chem 280(49): 40916–24. Tetsu, O., and F. McCormick. 1999. Beta-catenin regulates expression of cyclin D1 in colon carcinoma cells. Nature 398(6726): 422–6. Tomlins, S. A., M. A. Rubin, and A. M. Chinnaiyan. 2006. Integrative biology of prostate cancer progression. Annu Rev Pathol 1243–71. Truica, C. I., S. Byers, and E. P. Gelmann. 2000. Beta-catenin affects androgen receptor transcriptional activity and ligand specificity. Cancer Res 60(17): 4709–13. Varambally, S., J. Yu, B. Laxman, D. R. Rhodes, R. Mehra, S. A. Tomlins, R. B. Shah, U. Chandran, F. A. Monzon, M. J. Becich, J. T. Wei, K. J. Pienta, D. Ghosh, M. A. Rubin, and A. M. Chinnaiyan. 2005. Integrative genomic and proteomic analysis of prostate cancer reveals signatures of metastatic progression. Cancer Cell 8(5): 393–406. Voeller, H. J., C. I. Truica, and E. P. Gelmann. 1998. Beta-catenin mutations in human prostate cancer. Cancer Res 58(12): 2520–3. Wang, L., H. K. Lin, Y. C. Hu, S. Xie, L. Yang, and C. Chang. 2004. Suppression of androgen receptor-mediated transactivation and cell growth by the glycogen synthase kinase 3 beta in prostate cells. J Biol Chem 279(31): 32444–52. Whitaker, H. C., J. Girling, A. Y. Warren, H. Leung, I. G. Mills, and D. E. Neal. 2008. Alterations in beta-catenin expression and localization in prostate cancer. Prostate 68(11): 1196–205. Widelitz, R. 2005. Wnt signaling through canonical and non-canonical pathways: recent progress. Growth Factors 23(2): 111–6. Wissmann, C., P. J. Wild, S. Kaiser, S. Roepcke, R. Stoehr, M. Woenckhaus, G. Kristiansen, J. C. Hsieh, F. Hofstaedter, A. Hartmann, R. Knuechel, A. Rosenthal, and C. Pilarsky. 2003. WIF1, a component of the Wnt pathway, is down-regulated in prostate, breast, lung, and bladder cancer. J Pathol 201(2): 204–12. Yang, F., X. Li, M. Sharma, C. Y. Sasaki, D. L. Longo, B. Lim, and Z. Sun. 2002. Linking betacatenin to androgen-signaling pathway. J Biol Chem 277(13): 11336–44. Yang, X., M. W. Chen, S. Terry, F. Vacherot, D. L. Bemis, J. Capodice, J. Kitajewski, A. de la Taille, M. C. Benson, Y. Guo, and R. Buttyan. 2006. Complex regulation of human androgen receptor expression by Wnt signaling in prostate cancer cells. Oncogene 25(24): 3436–44. Yardy, G. W., D. C. Bicknell, J. L. Wilding, S. Bartlett, Y. Liu, B. Winney, G. D. Turner, S. F. Brewster, and W. F. Bodmer. 2008. Mutations in the AXIN1 Gene in Advanced Prostate Cancer. Eur Urol. Yi, F., and B. J. Merrill. 2007. Stem cells and TCF proteins: a role for beta-catenin–independent functions. Stem Cell Rev 3(1): 39–48. Yost, C., M. Torres, J. R. Miller, E. Huang, D. Kimelman, and R. T. Moon. 1996. The axisinducing activity, stability, and subcellular distribution of beta-catenin is regulated in Xenopus embryos by glycogen synthase kinase 3. Genes Dev 10(12): 1443–54. Yu, Y. P., D. Landsittel, L. Jing, J. Nelson, B. Ren, L. Liu, C. McDonald, R. Thomas, R. Dhir, S. Finkelstein, G. Michalopoulos, M. Becich, and J. H. Luo. 2004. Gene expression alterations in prostate cancer predicting tumor aggression and preceding development of malignancy. J Clin Oncol 22(14): 2790–9. Zhang, T., T. Otevrel, Z. Gao, Z. Gao, S. M. Ehrlich, J. Z. Fields, and B. M. Boman. 2001. Evidence that APC regulates survivin expression: a possible mechanism contributing to the stem cell origin of colon cancer. Cancer Res 61(24): 8664–7.

The Role of Foxa Proteins in the Regulation of Androgen Receptor Activity David J. DeGraff, Xiuping Yu, Qian Sun, Janni Mirosevich, Ren Jie Jin, Yongqing Wang, Aparna Gupta, Srinivas Nandana, Thomas Case, Manik Paul, Hong-Ying Huang, Ellen Shapiro, Susan Logan, Kichiya Suzuki, Marie-Claire Orgebin-Crist, and Robert J. Matusik

Abstract Activation of the androgen receptor is required for normal prostate physiology and in controlling the growth prostate cancer. However, the fact that multiple target organs express androgen receptor and are exposed to circulating androgens, yet fail to express prostate-specific markers and fail to develop androgen-dependent cancers, indicates that androgen receptor alone is not sufficient to dictate normal function and progression to cancer. Therefore, androgen action can be restricted in a given tissue by transcription factors that serve as co-regulators of androgen receptor. How androgen signaling acts in concert with other transcription factors, resulting in tissue-specific gene expression needs to be understood. The establishment of unique transcription factor regulatory networks is responsible, at least in part, to control androgen receptor action (1) in tissue-specific gene expression; (2) organ determination; and (3) cell differentiation. The identification of TF networks involved in these disparate events will allow researchers to elucidate the mechanisms that control prostate development, function, and pathology. Experimental evidence generated by our laboratory and others indicates that members of the Foxa subfamily of transcription factors play an important role in (1) normal prostate development; (2) the determination of prostatic cell fate; and (3) specific types of prostate pathology. This chapter reviews evidence generated by our laboratory and others regarding the important role of the Foxa transcription factors in the regulation of prostate-specific gene regulatory networks.

R.J. Matusik(*) Department of Urologic Surgery, Vanderbilt University Medical Center, A-1302, Medical Center North, 1161 21st Avenue south, Nastiville, TN 37232–2765, USA, E-mail: robert.matusik@ vanderbilt.edu

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4_25, # Springer Science + Business Media, LLC 2009

587

588

D.J. DeGraff et al.

1 Introduction Androgen signaling has a paramount role in normal prostate development, differentiation, and the control of growth of prostate cancer. The studies of Cunha demonstrated that the androgen receptor (AR) is a critical trigger for the onset of embryonic prostate induction during development (Cunha et al. 1980, 1983; Sugimura et al. 1986). Furthermore, these studies showed that recombination of wild-type urogenital mesenchymal cells with testicular feminized male (Tfm) prostate epithelial cells (with mutant AR) resulted in the formation of prostatelike organs. However, these organs failed to produce secretions such as probasin (PB), which indicates functional AR is required for epithelial cell secretory function (Donjacour and Cunha 1993). Several studies have shown that the production of a variety of secretory products by terminally differentiated prostate epithelial cells is under the direct control of androgen action on epithelial cells (for example, the production of rodent PB, and rodent and human spermine-binding protein (SPB), prostatic acid phosphatase (PAP), prostate-specific antigen (PSA) (Fong et al. 1991) and hk2 (Young et al. 1995) or the result of androgen mediated paracrine signaling by the stroma. The requirement of androgen for the production of these paracrine factors indicates the need for androgenic signaling for normal prostate gland development. The ability of AR to act as a transcription factor (TF), which changes gene expression, depends on AR’s ability to enter the nucleus. While the nuclear translocation of AR is normally ligand dependent, androgen-insensitive prostate cancer cells display high levels of nuclear AR even in the absence of androgen (Saporita et al. 2007). Further, other TFs serve as coactivators or corepressors of AR action (Agoulnik and Weigel 2006; Chmelar et al. 2007; Heinlein and Chang 2002). Coactivators may enhance ligand-induced transcription of AR by such general mechanisms as controlling cytoplasmic-nuclear trafficking, bridging AR with the basal transcriptional machinery, facilitating ligand binding, recruiting chromatin-modifying complexes, and/or enhancing DNA binding. In prostate cancer patients, androgen-deprivation therapy directly targets AR but patients become resistant to therapy to the extent that prostate cancer may grow in response to antiandrogens (Heinlein and Chang 2004; Stearns and McGarvey 1992). AR transactivation by antiandrogen can result from an acquired mutation in AR such as the AR mutation found in LNCaP (Tan et al. 1997). However, coregulators may play an important role in the increased sensitivity of AR to weak androgens or antiandrogens. For example, ARA70 increases AR activity for strong androgens, weak androgens, and antiandrogens (Miyamoto et al. 1998). Increased SRC-1 expression correlates with aggressive prostate cancer (Agoulnik et al. 2005), and increased TIF2 correlates with androgen independence (Agoulnik et al. 2006). ARA70 levels increase when the CWR22 prostate xenograft progresses during androgen deprivation to become castration recurrent (Stearns and McGarvey 1992). However, the CWR22 line has a mutated AR, which increases sensitivity to adrenal androgens (Tan et al. 1997). A mutation in b-catenin (S33F) has been

The Role of Foxa Proteins in the Regulation of Androgen Receptor Activity

589

identified in primary prostate cancer (Voeller et al. 1998). This mutated b-catenin S33F increases AR response to androstenedione and dehydroepiandrosterone, weak adrenal androgens (Truica et al. 2000). The action of ARA70 and mutant b-catenin S33F on AR suggests a general mechanism where overexpression of coregulators could explain how weak androgens or even antiandrogens can become androgenic in prostate cancer patients that fail androgen-deprivation therapy. Also, in breast cancer, amplification and over-expression of estrogen receptor (ER) coactivators such as SRC-3 and PBP/DRIP205/TRAP230 have been reported that supports coregulators as playing an important role in hormone-dependent cancers (Anzick et al. 1997; Zhu et al. 1999). For these reasons, AR interactions with its coregulators are now viewed as potential targets for newly developed drugs (Chang and McDonnell 2005). Activation of the AR is required for normal prostate physiology and development of prostate cancer. However, the fact that multiple target organs express AR and are exposed to circulating androgens, yet fail to express prostate-specific markers and fail to develop androgen-dependent cancers, indicates that AR alone is not sufficient to dictate normal function and progression to cancer. Therefore, androgen action can be restricted in a given tissue by a TF that serve as coregulators of AR. How androgen signaling acts in concert with other TFs to cause tissuespecific gene expression needs to be understood. The establishment of unique TF regulatory networks is responsible, at least in part, to control AR action (1) in tissue-specific gene expression; (2) organ determination; and (3) cell differentiation. The identification of TF networks involved in these disparate events will allow researchers to elucidate the mechanisms that control prostate development, function, and pathology. This chapter reviews evidence generated by our laboratory and others regarding the important role of the Foxa transcription factors in the regulation of prostate-specific gene regulatory networks.

2 Prostate Development 2.1

Overview

Prostate development is a complex biological event that involves extensive, androgen-dependent biochemical crosstalk between tissue compartments, to produce mature accessory sex organ. Prostate development has been studied extensively in the rodent, but the detailed molecular events leading to organ establishment remain uncertain. The mouse prostate is derived from the urogenital sinus (UGS), which derives from the cloaca. The UGS is composed of endoderm-derived epithelium (UGE) surrounded by mesoderm-derived mesenchymal layer (UGM). In response to increased androgen production by the fetal testes around embryonic (E) day 17–18, budding morphogenesis begins and results in UGE invading the surrounding UGM. In the following several embryonic days and after birth (up until

590

D.J. DeGraff et al.

postnatal day (P) 7), continued outgrowth of these prostatic buds gives rise to a number of solid epithelial cords, which forms the branch structure of anterior, ventral, and dorsolateral prostatic glands in rodents (Cunha et al. 2004). In the first 1–2 weeks following birth, the continually elongating epithelial cord begins to send out side branches, in a process referred to as branching morphogenesis, and ductal canalization occurs concurrently. Prostate development is complete when the mouse becomes sexually mature. The mature prostate is composed of a glandular structure with discrete epithelial and stromal compartments. The mature gland also depends upon the presence of androgens, and the crosstalk that occurs between the stromal and epithelial compartments.

2.2

Role of AR in Prostate Development

Budding morphogenesis during embryonic development requires the presence of circulating androgens. However, AR expression in the UGM both before and during prostate bud formation is critical for prostatic development. Tissue recombination experiments definitively showed the importance of UGM AR expression for normal prostate development (Cunha and Lung 1978). These experiments involved recombination of wild-type UGM with UGE from testicular feminized mice (Tfm), which harbor an inactivating AR mutation. While this procedure resulted in the development of normal glandular structures, the converse recombination of Tfm UGM with wild-type UGE failed to result in the formation of prostate tissue. The activity of UGM AR, but not UGE AR is required for normal prostate development. In addition to the importance of UGM-associated AR expression for normal prostate development, androgenic signaling in the UGM (and the subsequent production of paracrine signals which acted distally in the UGE) was important for epithelial tissue specification. For example, the recombination of UGM with various other endoderm-derived UGS components (bladder and urethra epithelial cells) resulted in differentiation of these cell types into prostate epithelial cells, complete with the expression of prostate-specific genes (Cunha and Donjacour 1989; Cunha et al. 1983; Donjacour and Cunha 1993). Although androgen responsive factors produced within UGM in response to androgens remain unknown, UGM AR expression and activity appear prime determinants of normal prostate development and cell fate.

2.3

Epithelium-associated Gene Expression Networks Determine Response to Inductive Mesenchyme

In contrast, while UGM recombination with other endoderm-derived components results in the formation of prostatic glands, UGM recombination with mesodermally derived tissue of the Wolffian duct (epididymis, ductus deferens, and seminal vesicles), which all express AR, fails to induce the formation of prostatic ducts.

The Role of Foxa Proteins in the Regulation of Androgen Receptor Activity

591

The inability of these Wolffian duct derivatives to respond to UGM indicates the existence of regional differences of mesenchymal tissue, and/or that the plasticity of target tissue is restricted by the germ layer of origin. Identifying unknown androgen-induced UGM factors, and increased understanding of the molecular mechanisms that underlie the ability of UGE to develop and differentiate in response to these factors, are areas of active research. The repertoire of TFs expressed within a given target epithelial tissue provides an obvious determinant of UGE tissue response to UGM-produced factors. These TFs may be expressed in a tissue-specific manner, limited to certain germ layer, or ubiquitously expressed. TF interactions provide a combinatorial control of gene expression through the formation of transcriptional networks. Experimental evidence generated by our laboratory and others indicates that members of the Foxa subfamily of TFs play an important role in (1) normal prostate development; (2) the determination of prostatic cell fate; and (3) specific types of prostate pathology. In addition to increasing knowledge of normal prostate development, these mechanisms that normally operate in development (i.e., proliferation, invasion, and apoptosis) may contribute to prostate tumorigenesis, metastasis, and androgen independence.

3 Molecular Biology of Fox proteins The Forkhead box (Fox) super family of TFs consists in humans of 17 gene subfamilies, with 41 currently identified genes. Relative to humans, lower organisms possess a more limited repertoire of Fox genes, which suggests that these genes play an important role in increasing tissue complexity and size (Myatt and Lam 2007). Fox protein family members have been reported to play significant roles in a wide variety of biological processes which include proliferation, apoptosis, invasion, migration, metabolism, and survival. The role of Foxa proteins in pathological states, such as cancer, is well recognized and has been reviewed elsewhere (Myatt and Lam 2007). Foxa proteins were the first identified mammalian forkhead type proteins (Lai et al. 1991). Although three Foxa proteins were identified first as liver-enriched transcription factors [termed hepatocyte nuclear factor-3 (HNF-3): HNF-3a or Foxa1, HNF-3b or Foxa2, and HNF-3g or Foxa3], they are expressed in other endoderm-derived organs such as lung, pancreas, and prostate. The three members of the Foxa subfamily are encoded by different genes on different chromosomes. Foxa1, Foxa2, and Fox3 have 85% sequence identity in the DNA-binding domain, and 82% similarity with the Drosophila forkhead protein. Foxa proteins bind to the same consensus DNA sequence ((A/C)AA(C/T)), but each individual member does so with different affinity. The three-dimensional structure of the FH domain mimics the structure of the linker histones (H1 and H5). While the linker histones aid in chromatin compacting to inhibit transcription, Foxa proteins activate transcription by displacing linker histones from nucleosomes, causing an unfolding

592

D.J. DeGraff et al.

Fig. 1 Distribution of Foxa1 and Foxa2 in mouse urogenital sinus (UGS). (a) Foxa1 is expressed in embryonic day 18 (E18) urogenital epithelium (UGE) (asterisk). (b) Cytokeratin 14 staining (Ck14) reveals peripheral basal epithelial layers. (c) Triple immunofluorescence merges from a and b, plus DAPI. (d) Nuclear Foxa staining in 6 week old mature prostate epithelium. (e) E18 UGE expresses Foxa2 (nuclear staining) with strongest layer in peripheral basal epithelium. Inset shows entire UGS counterstained with DAPI. (f) Same basal epithelial cells express Shh (signal on cell-membrane). (g) Merges of e and f. (h) In E21 UGS, Ptch1 is expressed in both UGE and mesenchymal cells (arrow). Arrowhead indicates Ptch1 expressing nascent prostatic buds. (i) Shh is co-expressed in E21 UGE and nascent epithelial buds (arrowhead). (j) Merges of h and i. (k)

The Role of Foxa Proteins in the Regulation of Androgen Receptor Activity

593

in chromatin structure and an increase in the accessibility of DNA to other TFs (Cirillo et al. 2002). Both amino and carboxy terminal ends of all Foxa proteins display transactivation activity (Pani et al. 1992) and Foxa proteins interact with various steroid hormone receptors, which implicates them in the combinatorial control of endocrine-regulated genes. For example, protein kinase A (PKA) stimulation improves recruitment of Foxa1 to a glucocorticoid responsive unit on the tyrosine aminotransferase promoter (Espinas et al. 1995). Foxa2 facilitates the binding of glucocorticoid receptor (GR) to the phosphoenolpyruvate carboykinase gene promoter (Stafford et al. 2001; Wang et al. 1996), and Foxa1 recruitment allows ER transactivation of the vitellogenin B1 promoter (Robyr et al. 2000). Furthermore, multiple ER responsive genes contain Foxa1-binding sites (Carroll et al. 2005).

4 Expression of Foxa Proteins in the Developing Murine and Human Prostate, and Adult Murine Prostate Normal prostate development requires the presence of inductive signals from UGM. However, in order for developing prostate tissue to respond in a manner allowing for normal morphogenesis, and result in normal cell determination, gene regulatory networks must exist within UGE-derived tissue that respond to signals from UGM. In fact, detailed analysis of Foxa family gene expression revealed several lines of evidence that Foxa1 is a component of such gene regulatory networks. Foxa1 is expressed during early prostate development and continues into adulthood (Gao et al. 2005). For example, immunohistochemical studies revealed that Foxa1 is expressed in cytokeratin 14 (marker for basal epithelium) positive and negative (luminal epithelium) in E18 murine UGS (Fig. 1). Immunohistochemical analysis of transverse sections from E21 murine prostates revealed strong Foxa1 staining in all epithelial cells of endodermal origin. Western blotting analysis of murine prostates from 1 to 15 weeks postnatal (Fig. 2) indicated that Foxa1 protein levels peaked between 4 and 5 weeks of age (Mirosevich et al. 2005). In addition, Foxa1 protein levels appear to decline in all lobes after 5 weeks of age, and appeared to be lower in samples of dorsolateral prostate, although this later observation may reflect changes in the ratio of stroma (Foxa negative) to epithelium. Foxa1 levels within the prostate appeared to equal or exceed the amounts Fig. 1 (Continued) E21 UGS double stained for AR and smooth muscle actin (SMA). SMA expressing cells are the same population expressing Ptch1 (arrow). (l) E21 UGS double-stained for AR and Ck14. (m) E21 UGS double-stained for Ck8 and p63. (n) beta catenin is expressed in both epithelium and surrounding mesenchyme in E18 UGS, with strongest staining detected in the peripheral epithelium (arrowhead). (o) E18 UGS stained for p63. (p) Magnified merges of N and O. (q and r) In situ hybridization, Foxa1 and Foxa2 are expressed in P1 prostate rudiments (arrows, parasagital sections). Insets show strongly stained epithelial buds (arrowheads in r). (s and t) Foxa1, but not Foxa2, is expressed in mature glands. (See Color Insert)

594

D.J. DeGraff et al.

Fig. 2 Foxa and AR expression in the murine prostate during pubertal growth and adult maturation. Western blot analysis of Foxa1, Foxa2, and AR proteins. Foxa1 protein levels peaked between 4 and 5 weeks of age. Foxa2 protein was not detected in any lobes of the mouse prostate using western blotting. AR protein levels increased in a similar fashion to Foxa1. b-Actin served as a loading control. Reprinted with permission of John Wiley & Sons, Inc., Prostate 62:339–352, (2005)

expressed in the liver (Fig. 2), the tissue in which Foxa1 (HNF-3a) was described first (Costa et al. 1989). While Foxa1 expression begins early in prostate development and continues on into adulthood, Foxa2 expression is relatively limited. Foxa2 expression appears localized to cytokeratin 14-positive basal epithelial cells adjacent to surrounding mesenchyme at E18 (Fig. 1). In addition, Foxa2-positive UGE basal cells also express Shh at E18, and by E21, both UGE and UGM are Shh positive (Fig. 1). Approximately equal expression of Ptch 1 in both UGE and UGM correlates with Shh expression and is consistent with reports that Shh induces Foxa2 expression (Sasaki et al. 1997). By E21, Foxa2 expression is strongest in newly forming prostatic buds (Mirosevich et al. 2005) (Fig. 3a–b), and in these rodent results have been replicated in samples derived from human fetal tissue (Fig. 3c–d), indicating that Foxa2 function during budding morphogenesis is conserved from mouse to human. Foxa3 expression has not been detected in the developing or adult murine prostate gland. Foxa2 proteins also are expressed in adult murine epididymis principal cells (Yu et al. 2005) (Fig. 4a–b).

5 Foxa1 Regulates Prostate Ductal Morphogenesis and Promotes Maturation The presence of Foxa1 expression during all stages of prostate development, as well as in the adult gland, and the relatively restricted pattern of Foxa2 expression to a subset of basal-like epithelial cells in the prostate, suggest an important role for the Foxa family in normal prostate development and physiology. Foxa1 null mice were created by deletion of Foxa1 DNA-binding domain to create an in-frame fusion with the Escherichia coli lacZ gene. Since Foxa1 null mice die neonatally, renal capsule organ rescue and tissue recombination techniques were used to assess the

The Role of Foxa Proteins in the Regulation of Androgen Receptor Activity

595

Fig. 3 Foxa protein expression in developing murine and human prostate. Transverse sections of urogenital sinus derived from 21-day-old embryonic mice were used for immunohistochemistry using antibodies specifically for (a) Foxa1 and (b) Foxa2. Strong Foxa1 staining was detected in all epithelial cells of endoderm origin. Foxa2 expression was strongest in the newly forming prostatic buds. Human fetal prostate shows identical spatial expression of Foxa1 (c) and Foxa2 (d), with all epithelial cells displaying positive Foxa1 staining, while only distal ascini are Foxa2 positive. Reprinted with permission of John Wiley & Sons, Inc., Prostate 62:339–352, (2005) (See Color Insert)

impact of Foxa1 deletion on prostate development (Gao et al. 2005) (Fig. 5). Prostate rudiments from wild-type and Foxa1 null P1 pups were identical histologically; prostate development was normal to this stage. Entire prostate rudiments from both wild-type and Foxa1 null pups were rescued by renal grafting, and recovered at various intervals between 2 and 15 weeks. Upon gross examination, Foxa1 null prostate tissues were smaller, whereas wild-type tissues were larger and contained secretions. Prostate components null for Foxa1 expression contained a lacZ fusion protein. Tissue sectioning followed by beta-galactosidase staining revealed epithelial cell disorganization in beta-galactosidase-positive cells. Further histological analysis of prostate tissue rescued from Foxa1 null mice for periods of time between 2 and 15 weeks revealed the development of solid epithelial cords with cribriform patterns, absence of glandular elements, and loss of cell polarity, as indicated by E-cadherin staining. These findings were consistent for tissue derived from all prostatic lobes of Foxa1 null mice following renal graft rescue, which

596

D.J. DeGraff et al.

Fig. 4 Foxa expression in the murine epididymis. Immunohistochemistry analysis of Foxa1 (a) and Foxa2 (b) proteins in epididymis (100). Foxa2 expression was localized to the nucleus of epididymis principal cells (b, inset, 400). Reprinted with permission of Wiley-Blackwell, Annals of the New York Academy of Sciences, ‘‘Foxa1 and Foxa2 Interact with the Androgen Receptor to Regulate Prostate and Epididymal Genes Differently’’1061:fig 1, pg 82, fig 2 pg 83, 12 (2005) (See Color Insert)

indicates Foxa1 expression is important for normal prostate development, regardless of the lobe. Tissue and cytological characterization of the Foxa1 null phenotype in 4-weekold rescued prostates revealed an expanded basal cell population. Basal-like cells become the predominant cell type in Foxa1 null animals and formed piles that extend into the lumen of the secretory glands. While basal and luminal epithelial cell markers p63 and cytokeratin 8, respectively, segregated completely in wildtype tissue recombinants, p63 and cytokeratin 8 colocalized in Foxa1 null prostates, which is observed usually only in normal embryonic UGE. Thus, the epithelium of the Foxa1 null prostate remains at an intermediate cell type normally found only in the embryonic prostate. The ability to engage in substantive secretion is a measure of a functional and differentiated prostate epithelial cell. In order to detect abnormalities associated with secretion, 13-week-old wild-type and Foxa1 null VP sections were compared using electron microscopy (Gao et al. 2005) (Fig. 6). Wild-type VP cells contained both secretory granules and a normal appearing apical surface, but Foxa1 null tissue did not. These observations provide direct evidence of the inability of Foxa1 null prostate cells to terminally differentiate. Foxa1 null animals exhibit an expanded smooth muscle actin (SMA) gamma-positive layer that contains terminally differentiated muscle cells (Fig. 7). The detection of a modified stromal pattern indicates a failure of UGE and UGM to communicate during development. Since Foxa1 null prostates failed to develop normally, Foxa1 must enable UGE to respond properly to the inductive signals from the mesenchyme. Tissue recombination experiments were performed to test this hypothesis. Recombination experiments utilizing rat (r) UGM in combination with normal murine bladder epithelium, followed by renal grafting, result in the formation of prostate

Fig. 5 Foxa1 regulates prostate ductal morphogenesis. (a) Upper panels: urogenitial organs (lateral, left, and dorsal, right, views) dissected from P1 pups. The bladder was removed as indicated by broken lines. Lower panels: prostate rudiments (asterisks) and seminal vesicle (SV) were grafted as renal rescue tissue. (b–d) Upper panels: 8-week-old rescued tissues, with indicated genotypes, developed in the host renal capsules. SVs are circled with broken lines. Rescued Foxa1-/- prostate (asterisk in d) is smaller then controls upon comparison after fine dissection (lower panels). (e–g) beta galactosidase staining on 8-week-old rescued prostates. (h) Twelveweek-old tissue recombinants derived from wild-type (left) or Foxa1-/- (right) epithelium that was recombined with E18 rat UGM (rUGM). (i) The Foxa1-/- recombinants have significantly lower weights than controls (n = 3, P < 0.01). (j–k) Hematoxylin and eosin staining of 4-week-old rescued Foxa1-/- and Foxa1+/+ ventral prostates (VPs). (l and m) Androgen receptor (AR) staining. (n, o) Toluidine blue staining on 1m thin section of 12-week-old rescued Foxa1-/- and wild-type VPs. (p, q) E-cadherin staining on 12-week-old rescued Foxa1-/- and wild-type VPs. Foxa1-null epithelial cell polarity is disrupted (arrowheads). (r, s) Hematoxylin and eosin staining of 12-week-old tissue recombinants from Foxa1-/- and control epithelium. Figure first published in Development 132, 3431-3443 by author) (See Color Insert)

598

D.J. DeGraff et al.

Fig. 6 No mature luminal cells in Foxa1/ prostate. (a–d) EM analysis on 12-week-old rescued Foxa1+/+ and Foxa1/ ventral prostates. Scale bars: 2 mm in a, b: 500 nm in c, d. Asterisk indicates the lumen. Wild-type cells (a, c) contain secretory materials in apical vesicles (arrow) and at luminal surface (arrowhead). Secretory material was absent in Foxa1/ cells (b, d). Figure first published in Development 132, 3431–3443 by author

glandular structures. Therefore, tissue recombination can be used to assay the ability of target epithelium to respond to signals from normal inductive mesenchyme. Tissue recombination experiments were performed with E18 rUGM in combination with Foxa1 null bladder epithelium. The combination of wild-type bladder epithelium and rat UGM was used as a control (Fig. 5). While wild-type bladder epithelium developed into prostate tissue after recombination with rUGM, Foxa1 null tissue failed to do so. These results confirm that Foxa1 is necessary for the ability of UGE to respond to inductive stromal signaling, which specifies final cell fate. In summary, Foxa1 null mice display a developmentally arrested phenotype. The presence of this developmentally arrested phenotype in Foxa1 null mice is

The Role of Foxa Proteins in the Regulation of Androgen Receptor Activity

599

Fig. 7 location of Foxa-binding sites in probasin promoter. (a) Mutation studies revealed two key cis-acting elements R1 and R2 (dashed lines). Each contains a Foxa1-binding motif (bold arrows). The direction of the arrow indicates the sense or the antisense strand that matched the consensus sequence. Two adjacent ARBSs (ARBS-1 and ARBS-2) are underlined. Sequences in italics indicate the nucleotide replacements in mutation assays. (b) Southwestern blot. Nuclear extract from LNCaP and PC-3 cells were separated on SDS–PAGE, trans blotted to nitrocellulose membranes, and probed with radiolabeled 2 R2 (lanes 1 and 3), 2 mR2 (lanes 2 and 4), or 2 AR-con probes (lane 5). A western blot using Foxa1 antibody was performed as a parallel control. Arrows indicate migration of proteins bound to respective probe or antibody. Copyright 2003, The Endocrine Society, Molecular Endocrinology, ‘‘The Role of Hepatocyte Nuclear Factor-3a (Forkhead BoxA1) and Androgen Receptor in Transcriptional Regulation of Prostatic Genes’’, Vol. 17, 1484–1507, (2003)

evident based upon gross histological analysis, pronounced epithelial disorganization, formation of solid epithelial cords with a cribriform pattern, absence of ductal canalization, complete lack of lumen formation, and lack of secretion. At the cellular level, these gross morphological alterations appear to result from an increased number of basal-like cells which contain markers for both basal and luminal cytokeratins, that represent an intermediate epithelial cell type. Taken together, these findings suggest that Foxa1 expression, and subsequent activity, is crucial for normal basal and luminal cell differentiation, normal morphogenesis, and determination of cell fate.

600

D.J. DeGraff et al.

6 Foxa1 Interacts with Androgen Responsive Promoters and Androgen Receptor to Modulate Gene Expression The paradigm of combinatorial gene control, which describes how expression of a limited repertoire of gene regulatory factors can be utilized in combinations, to produce a wide variety of physiological responses was proposed over 30 years ago (Gierer 1974). Combinatorial control of gene expression starts in the embryonic process that results in cell determination, the assignment of a given developmental fate, and final differentiation, which is defined as the realization of specific cellular function. The paradigm of combinatorial gene control has evolved further through understanding of transcriptional synergy (Brivanlou and Darnell 2002; Carey 1998; Fry and Peterson 2002). For example, it is estimated that 2,000–3,000 transcription factors are expressed in mammals. This set of transcription factors consists of (1) general transcription factors; (2) conditional, signal-dependent transcription factors; and (3) cell-specific transcription factors (Brivanlou and Darnell 2002). The combinatorial use of these transcription factors is thought to allow correct temporal and spatial control of gene expression, as well as allow gene-specific regulation (Carey 1998). Furthermore, the set of transcription factors for exquisite control of gene regulation is limited. For example, it was shown recently that mouse and human fibroblasts can be reprogrammed into pluripotent stem cells through the ectopic expression of as few as four transcription factors (Okita et al. 2007; Takahashi and Yamanaka 2006; Wernig et al. 2007; Yu et al. 2007). Our in vivo studies show that Foxa1 null bladder epithelium does not differentiate into secretory prostate epithelium following tissue recombination. The inability of Foxa1 null tissue to develop normally suggested that Foxa1 is an essential component of transcription factor networks responsible for the commitment of endodermal tissue to a prostate epithelial-specific lineage (Matusik et al. 2008). In addition to a role for Foxa1 expression as an important component of normal prostate development, Foxa consensus-binding sites often are associated closely with androgen response elements (AREs) (Wang et al. 2007). For example, in a study of 90 AREs on Chromosomes 21 and 22, 36 AREs were predicted to have binding sites for Foxa1, GATA2, and Oct 1. Chromatin immunoprecipitation (ChIP) studies on validated AREs revealed that Foxa1 bound to 17 sites, GATA2 to 14 sites, and Oct1 to 32 sites. This finding, in conjunction with the essential role of the AR in prostate-specific gene expression, indicates that the combinatorial activity of AR and Foxa family members is essential for regulation of gene expression during prostate cell determination and differentiation. The extensively characterized PB promoter was used to study the molecular mechanisms that underlie the Foxa family’s ability to engage in prostate-specific gene expression, to determine if Foxa family members interact with AR, and to determine how Foxa family members are recruited to responsive promoters for normal prostate development and determination of cell fate. PB is a 20 kDa androgen-induced secreted protein originally identified in dorsolateral rat prostate. In addition, PB is expressed

The Role of Foxa Proteins in the Regulation of Androgen Receptor Activity

601

in the murine prostate. Since PB expression is restricted to prostatic epithelial cells, the PB promoter can be used to drive prostate tissue-specific gene expression in transgenic mice. Our studies have shown that the PB proximal promoter region encompassing 286/+28 bp is sufficient to control prostate-specific gene expression in transgenic mice (Zhang et al. 2000). Furthermore, DNase I footprinting and EMSA studies revealed the presence of two AR-binding sites (ARBS) at 140 to 117 bp (ARBS1) and 236 to 223 bp (ARBS2) in this region. These ARbinding sites are necessary and sufficient for maximal androgenic induction of the PB promoter, and therefore this promoter region (244 to 96 bp) is referred to as an Androgen Responsive Region (ARR).

6.1

Identification of Foxa-Binding Sites in the Rat Probasin Promoter

To identify the presence of key cis-acting elements required for prostate-specific gene expression, extensive mutagenesis of the DNA sequences within the 286/ +28 bp PB promoter was created, and 31 continuous link-mutation constructs were transfected into both prostate and nonprostate cell lines. This process allowed the assay of the mutant promoter constructs to drive PB expression upon androgen stimulation. The mutation of two elements abolished promoter activity in prostate cells but not in nonprostate cells transfected with PB. These regions in the PB promoter were therefore referred to as Tissue-Specific elements 1 (TS1) and 2 (TS2) (Fig. 7a). Bioinformatic analysis indicated that TS1 and TS2 DNA elements contained Foxa-binding sites (depicted as R1 and R2 in Fig. 7a) (Gao et al. 2003). These tissue-specific elements are adjacent to AR-binding sites, and share similar DNA sequences, which match the consensus sequence for Foxa-binding sites (50 -TRTTRYTY-30 ). Southwestern blotting analysis of LNCaP and PC3 nuclear extracts with a probe matching the R2 consensus sequence revealed two proteins with molecular masses of approximately 46 and 52 kDa, and parallel western blotting analysis identified the 52-kDa band as Foxa1 (Fig. 7b).

6.2

Identification of Foxa-Binding Sites in the Human PSA Core Enhancer

The PSA core enhancer region (4.2/3.8 kb), which contains multiple AREs (ARE I–III), but not the proximal promoter, is necessary and sufficient for androgen regulation and prostate specificity. Of these, ARE III displays the highest AR affinity and biological activity. Two potential Foxa1-binding sites were identified using bioinformatics analysis adjacent to ARE III in the PSA core enhancer

602

D.J. DeGraff et al.

Fig. 8 Identification of Foxa1-binding sites in the PSA core enhancer. (a) Two Foxa1-binding sites (bold arrow) were identified in the PSA core enhancer (4.2/3.9 kb) region. The first element (4122/4109), designated as PSA1, is next to ARE III (underlined). The second one (4028/4005), designated as PSA2, is in the middle of ARE IIIA and ARE IIIB (underlined). (b) EMSA. In vitro synthesized wild-type Foxa1 and mutant proteins (DN, FH, D C) were able to bind PSA1 (lanes 3–6) as compared with TNT blank control (lane 2). The Foxa1 binding was supershifted by Foxa1 antibody (lane 3 vs. lane 7). (c) Concomitant DNA binding of AR and Foxa1. Radiolabeled oligonucleotides containing both ARE III and PSA1 were incubated with constant amount (0.1 mg) of a purified GST AR-DBD protein. A slow migrating ternary complex (AR/Foxa1/DNA) (lanes 5–8) was formed as addition of increasing amounts (1, 2, 3, and 4 ml) of in vitro synthesized Foxa1, in contrast to Foxa1 alone (lane 3) or AR alone (lane 4), which indicates that Foxa1 and AR can occupy DNA concomitantly. The ternary complex was disrupted by addition of Foxa1 antibody (lane 9). Copyright 2003, The Endocrine Society, Molecular Endocrinology, ‘‘The Role of Hepatocyte Nuclear Factor-3a (Forkhead BoxA1) and Androgen Receptor in Transcriptional Regulation of Prostatic Genes’’, Vol. 17, 1484–1507, (2003)

(Gao et al. 2003). These Foxa1-binding sites (Fig. 8a) are located at 4122/4109, and 4028/4005. In vitro synthesized Foxa1 bound to both elements (Fig. 8b), and the binding of Foxa1 was supershifted using Foxa1 antibody. Furthermore, Foxa1 and AR display concomitant DNA binding to their respective sites. In addition to this strong in vitro evidence regarding the ability of Foxa1 to bind directly to the PSA core enhancer, ChIP studies in the human prostate cancer cell line, LNCaP revealed the ability of Foxa1 to bind to the PSA core enhancer in the

The Role of Foxa Proteins in the Regulation of Androgen Receptor Activity

603

presence and absence of DHT (Fig. 9). In addition to the finding that Foxa1 binds to cognate cis elements on the PSA enhancer, GST pull down studies showed that Foxa1 interacts directly with AR.

6.3

Foxa-Binding Site in the mE-RABP Promoter

The epididymis must engage in specific local-regional patterns of gene expression for proper sperm maturation. Therefore, the epididymis provides an excellent model for study of local gene expression regulatory events. Immunohistochemical analysis revealed Foxa2 expression in the nucleus of murine principal cells (Fig. 4). Two AR-binding sites (ARBS-2 at 1.3 kb and ARBS-3 at 1.2 kb) were identified using 50 deletion analysis of the cis DNA regulatory elements of the principal cellspecific murine epididymal retinoic-acid-binding protein (mE-RABP) gene followed by transfection into DC2 epididymal cells and EMSA. Bioinformatics analysis revealed a potential Foxa-binding motif (Yu et al. 2006) (Fig. 10). The use of a probe representing this region in EMSA studies from DC2 extracts revealed the formation of three DNA–protein complexes, two of which were abolished through probe competition experiments (Fig. 10). In addition, further studies confirmed the presence of a Foxa2-binding site in murine DC3 nuclear extracts, and rat and human E-RABP promoters (Fig. 10).

6.4

Foxa1 and Foxa2 Differentially Modulate Gene Expression from Androgen Responsive Promoters

The interaction between Foxa1 and AR suggested functional consequences which were tested by transfecting a PSA reporter construct with different combinations of AR, Foxa1, and Foxa2 in the presence or absence of DHT (Mirosevich et al. 2006). For example, following transient transfection of AR and PSA, both Hela and LNCaP cells responded to androgen treatment by increasing PSA reporter activity (Fig. 11). The additional transfection of Foxa1 to AR and PSA producing Hela and LNCaP cells abrogated the ability of DHT to increase PSA reporter activity. However, the triple introduction of PSA/AR/Foxa2 inducted the PSA reporter in Hela and LNCaP in the presence or absence of DHT. Since androgen-sensitive LNCaP cells are Foxa2 null, and androgen-independent PC3 cells express high levels of Foxa2, this finding suggests a role for Foxa2 in androgen independence. Further, androgen-independent DU145 cells also fail to express Foxa2. Following transient transfection of epididymal cells, Foxa2 expression inhibits AR induction of the mE-RABP promoter. This contrasts with the ability of Foxa2 to activate prostate-specific gene expression in the absence of androgens. These data suggest

Fig. 9 Foxa1 occupies PSA enhancer in vivo. (a) PSA gene 50 upstream region. Six DNA fragments corresponding to the distal, ARE III, middle, ARE II, and ARE (2) regions were amplified in ChIP assays. (b) LNCaP cells were initially grown in RMPI 1640 with 5% charcoal/dextran-treated fetal bovine serum for 3 days. Cells were then incubated for another 48 h in the presence or absence of 108 M DHT. Western blot was performed to determine the expression levels of PSA and Foxa1. (c) After an overnight IP with AR or Foxa1 antibodies, ChIP analysis was performed. Copyright 2003, The Endocrine Society, Molecular Endocrinology, ‘‘The Role of Hepatocyte Nuclear Factor-3a (Forkhead BoxA1) and Androgen Receptor in Transcriptional Regulation of Prostatic Genes’’, Vol. 17, 1484–1507 (2003)

The Role of Foxa Proteins in the Regulation of Androgen Receptor Activity

605

Fig. 10 Identification of Foxa2-binding sites on the mE-RABP promoter. (a) Foxa2-binding site exists in the vicinity of AR-binding sites on mE-RABP promoter. Underlined are the Foxa-binding sites core sequence and androgen-receptor-binding sites. (b) EMSA. Oligomers containing potential Foxa-binding motif on mE-RABP promoter were labeled as probe. Foxa-binding sequence on TTR gene promoter was used as a competitor. EMSA was performed using nuclear extract from DC2 cells and labeled Foxa2 probe. Three DNA–protein complexes (bands a, b, c) were formed. 300-fold excess nonlabeled self-competitor (lane 3) or TTR competitor (lane 4), which contains consensus Foxa-binding sites, abolished bands a and b. (c–e) Identification of Foxa2 binding on mouse, rat, and human E-RABP promoters. (c) EMSA indicates in vitro synthesized Foxa2 protein bound to Lcn5 promoter. In vitro translated Foxa2 protein and 32p-labeled probe were used in the EMSA. (d) Foxa2’s binding to Lcn5 promoter was further confirmed by antibody supershift. EMSA was done using nuclear extract DC2 cells and 32p labeled. Foxa2 antibody resulted in a supershift band. (e) Foxa-binding motif was also found on rat E-RABP promoter and human ERBP gene promoters. Oligomers from human and rat E-RABP promoters competed off Foxa protein binding to probe, indicating that Foxa-binding motif is conserved among human, rat, and mouse ERABP promoters. Modified and reprinted with permission of The Endocrine Society, Molecular Endocrinology, ‘‘The Role of Forkhead BoxA2 (Foxa2) to Restrict Androgen Regulated Gene Expression of Lipocalin 5 in the Mouse Epididymis’’, Vol. 20, 2418–2431, (2006)

606

D.J. DeGraff et al.

Fig. 11 Influence of single and cotransfection of AR, Foxa1, and Foxa2 on PSA reporter activity in (a) Hela and (b) LNCaP cells in the presence or absence of DHT. Results are expressed as means SEM and are presented as relative luciferase activities that are representative of three independent experiments performed in triplicate. *Significantly increased compared to PSA alone without DHT treatment, P < 0.01. +Significantly increased compared to PSA alone with DHT treatment, P < 0.01. Foxa2 significantly increase PSA gene promoter activity in the absence of DHT, or AR in both cell lines. Reprinted with permission of John Wiley & Sons, Inc., Prostate 66:1013–1028 (2006)

The Role of Foxa Proteins in the Regulation of Androgen Receptor Activity

607

distinct regulatory differences between Foxa1 and Foxa2 interaction with AR and other cofactors in different cellular environment.

7 Foxa Family Member Expression in Prostate Cancer Our studies showed that the rescue of Foxa1 null prostate tissue using renal grafting results in tissue that is hyperplastic, displays disrupted prostate ductal morphogenesis and lacks epithelial cell secretions. In addition, rescued Foxa1 null prostates developed an expanded layer of basal-like epithelial cells, which colocalized p63 (basal cell marker) and cytokeratin 8 (luminal cell marker). Taken together, these results indicated development is arrested in Foxa1 null murine prostates. However, Foxa1 is implicated in pathogenesis of several organs, which include the breast, lung, and esophagus. For example, Foxa1 expression levels are reduced in invasive breast cancer, and increased levels of Foxa1 expression are correlated with favorable prognosis in breast cancer patients. Furthermore, Foxa1 interacts with ER (Carroll et al. 2005), which may explain that cotransfection of Foxa1 suppressed ER trans-activity in ER-positive breast cancer cells (Carroll et al. 2005). Since forced overexpression of Foxa1 in breast cancer cells results in reduced colony formation, Foxa1 may have an inhibitory function in regard to estrogen signaling, which results in growth inhibition (Wolf et al. 2007). In addition to the possibility that Foxa1 inhibits estrogen signaling, Foxa1 has been shown to interact with the cell cycle inhibitor p27, which may cause further growth inhibition. While Foxa1 is expressed in adult prostate, and throughout normal murine development, Foxa2 is expressed only in murine embryonic prostates during budding morphogenesis and appears in a sparse population of basally located epithelial cells in the adult organ. These Foxa2-positive cells are positive for Foxa1 and synaptophysin, a marker for the rare neuroendocrine (NE) cells present in the prostate (Fig. 12). In an effort to produce new models of prostate cancer, a large 12-kb portion of the PB promoter was linked to a Simian Virus 40 (SV40) large T antigen (Tag) deletion mutant (that removed the expression of small t antigen). Directed expression of large Tag specifically to the prostate epithelium of CD-1 mice produced our 12T-7 and 12T-10 murine models. These murine models reproducibly undergo pathological alterations associated with prostatic intraepithelial neoplasia (PIN) (12T-7) and NE prostate cancer (12T-10). In addition, the 12T-10 NE model also develops metastases to various organs which include the liver and lung that exhibit differential gene expression patterns (Fig. 13). Human ASH-1 (hash-1) colocalized with the NE marker Chromogranin-A and correlated with the use of androgen-deprivation therapy in human samples of NE prostate cancer (Rapa et al. 2008). These clinical findings support our previous results that metastatic lesions of the 12T-10 NE model also express the murine homolog mASH-1 and further indicates the relevance of this animal model of NE prostate cancer (Gupta et al. 2008). While our Foxa1 knockout studies were not designed to measure the influence of a loss of Foxa1 expression in adult

608

D.J. DeGraff et al.

Fig. 12 Foxa protein expression in the normal adult murine prostate. Immunohistochemical staining of adult murine prostates for Foxa1 in the (a) ventral, (b) anterior, and (c) dorsolateral prostates. The majority of epithelial cells stained positively for Foxa1, while the stromal cells were negative for Foxa1. Bar ¼ 168 mm. (d), Dual Foxa2 and synaptophysin (Syn) immunofluorescence. Bar ¼ 42 mm. Foxa2 expression was exclusively restricted to individual basal epithelial cells of the periurethral ducts, which also coexpressed the neuroendocrine marker synaptophysin (arrowheads). DAPI served as a counterstain. Reprinted by permission of John Wiley & Sons, Inc., Prostate 66:1013–1028 (2006) (See Color Insert)

prostatic tissue, studies with the 12T-7 transgenic prostate cancer model revealed decreased Foxa1 expression in (PIN) and carcinoma, but Foxa2 is expressed in the 12T-10 NE cancer. Thus, while researching the role of Foxa1 in androgen-regulated genes in the prostate, we discovered that Foxa2 was found as a biomarker of NE prostate cancer in both mice and men. Our data suggest that Foxa2 contributes to the progression of this subtype of castration-recurrent prostate cancer. Increased NE differentiation (NED) has been associated with decreased survival. While Foxa2 expression is limited to the rare NE cells in wild-type mice, Foxa2 expression is increased the 12T-10 NE model, as well as in the TRAMP model, which is now known to develop NE prostate cancer. (Chiaverotti et al. 2008; Mirosevich et al. 2006) (Fig. 14). In addition to their association with poor prognosis, NE tumors are typically AR null. A recent morphological and histological analysis of 95 cases of NE prostate cancer suggested that NE prostate cancer may be more prevalent than originally thought

The Role of Foxa Proteins in the Regulation of Androgen Receptor Activity

609

Fig. 13 Top panel: Serial sections of neuroendocrine liver metastasis for a 69-week-old 12T-10 mouse showing (a) Hematoxylin and eosin, (b) positive immunostaining for Foxa2, (c) synaptophysin, and (d) T-antigen. Also shows negative staining for (e) AR, positive staining for (f) mash1, negative staining for (g) neurogenin-3, and positive nuclear staining for (h) Nkx2.2. Bottom panel: Immunohistochemical analysis on serial sections from lung metastasis from a 69-week-old 12T-10 mouse showing (a) Hematoxylin and eosin, (b) positive staining for Foxa2, (c) synaptophysin, and (d) T-antigen, negative staining for AR (E), positive nuclear staining for (F) mash-1, and (g) neurogenin-3, and negative staining for (h) Nkx2.2. Bars, 50 mm. Reprinted with permission of Wiley-Liss, Inc., a subsidiary of John Wiley & Sons, Inc. Prostate 68:50–60 (2008) (See Color Insert)

(Wang and Epstein 2008). Thus, the expression of Foxa2 in NE tumors provides a potential additional immunohistological marker of NE prostate cancer and suggests this forkhead protein is involved in NED that occurs as human prostate cancer progresses during androgen-deprivation therapy. Our in vitro studies support this hypothesis: when Foxa2 is cotransfected with PSA-luciferase reporter, Foxa2 activates the PSA promoter in the absence of androgen, while Foxa1 has limited effect on the PSA promoter activity in transiently transfected cells in the absence of androgen. The ability of Foxa2 to activate PSA promoter activity independent of androgen suggests that Foxa2 may contribute to the androgen independence of NE prostate cancer, and indicates a need for the study of Foxa2 expression and activity during progression to castration-recurrent disease. The NE tumors that develop in TRAMP and 12T-10 (while AR negative) still express the Tag transgene that is under the control of androgen responsive PB promoter. Foxa2 may drive the PB promoter activity in the absence of AR expression.

610

D.J. DeGraff et al.

Fig. 14 Pathology and immunostaining of 12T-10 prostates. (a) Hematoxylin and eosin stained dorsolateral prostate tissue from 50-week-old mouse shows areas of PIN (asterisk) and undifferentiated or neuroendocrine (NE) carcinoma. Immunohistochemical staining for (b) T-antigen, (c) Foxa1, (d) Foxa2, (e) AR and (f) Synaptophysin. NE cells were positive for both synaptophysin and Foxa2, and weakly positive for AR (e, inset. Bar = 84 mm). Reprinted by permission of John Wiley & Sons, Inc., Prostate 66:1013–1028 (2006) (See Color Insert)

8 Summary and Conclusions Foxa gene expression is necessary for normal prostate development and is implicated in the development and progression of prostate cancer. Foxa-binding sites are located in the rat prostate PB promoter, the human prostate PSA core enhancer, and

The Role of Foxa Proteins in the Regulation of Androgen Receptor Activity

611

the murine epididymal mE-RABP promoter. Finally, Foxa family members operate on responsive promoters to modulate androgen-induced gene expression. Specifically, transfection of Foxa1 to AR and PSA producing Hela and LNCaP cells abrogates androgen-induced increases in PSA reporter activity, while ectopic Foxa2 expression induces the PSA reporter in Hela and LNCaP in the presence or absence of DHT. In contrast, when Foxa2 is overexpressed in epididymal cells, the androgen response of the mE-RABP promoter is suppressed. Therefore, in two organs controlled by androgens, the Foxa proteins are expressed differentially and not exchangeable functionally. In short, these studies indicate that Foxa1 and Foxa2 TFs are important for normal prostate development and physiology of the male genital tract, and indicate that Foxa2 may be involved in the development of NE prostate cancer. Foxa1 expression is detected early in prostate development (UGS E18), while the Foxa1 target gene PB is detected at high levels only after prostate organogenesis is complete. Here we confirm that the expression of Foxa2 in the mouse prostatic embryonic buds also is conserved to the human embryonic prostatic buds. Therefore the role of Foxa1 and Foxa2 in prostate development is highly conserved. However, other TF must operate in concert with Foxa1 to regulate target gene expression. The fact that Foxa1 is required for normal prostate development and that it is expressed in the embryonic prostate before luminal secretory target gene are expressed, suggests that intermolecular interactions of Foxa1 with other transcription factors are important. Experimental evidence generated by other laboratories indicates that Foxa1 plays a central role in estrogen and androgen-regulated genes (Carroll et al. 2005; Lupien et al. 2008). Foxa1 functions in a cell type-specific manner to attract lineage-specific transcription factors. Together with the methylation of histone H3, Foxa1 forms part of the epigenetic signature that alters chromatin structure allowing for lineage-specific gene regulation (Lupien et al. 2008). This is consistent with the role of Foxa1 to serve as a ‘‘pioneer factor’’ that displaces linker histones from nucleosomes, causing an unfolding of chromatin and an increase in the accessibility of DNA to other TFs (Cirillo et al. 2002). This fundamental function of Foxa1 should play a critical role in chromosome organization in response to ER and AR signaling. Specific ER target genes establish interactions between their promoters that are located on separate chromosomes to achieve a three-dimensional complex that share at transcriptional machinery (Nunez et al. 2008). Similarly, androgens induce the AR to regulate the PSA (chromosome 19) and TMPRSS2 (chromosome 21) promoters to form an interchromosomal complex that enhances transcription of these AR target genes (Nunez et al. 2008). Therefore, the identification of Foxa1 interacting partners will help unravel the critical TFs that control cell type specific gene expression. In the prostate, the role of Foxa1 during development and the control of gene expression will provide a more complete understanding of prostate development and pathology. In contrast to the robust expression of Foxa1 during normal prostate development, Foxa2 expression localizes to a small subset of basal epithelial cells, and colocalizes with the neuroendocrine marker, synaptophysin. This observation, in conjunction with the fact that Foxa2 expression is elevated in the 12T-10 transgenic

612

D.J. DeGraff et al.

neuroendocrine mouse model, and Foxa2 is capable of stimulating PSA reporter activity in the absence of androgen, suggests that Foxa2 expression and activity is important in the development of NE prostate cancer. For these reasons, the identification of Foxa2 target genes in models of NE prostate cancer should be pursued.

Acknowledgments This research was supported by National Institute of Health (NIH) grants to Robert J Matusik (R01-CA76142, R01-DK55748, and R01-AG023490) and Frances Williams Preston Laboratories of the T.J. Martell Foundation; and by an NIH grant to Marie-Claire Orgebin-Crist (R01-HD36900). Janni Mirosevich is a recipient of the Department of Defense (DOD) Postdoctoral David J. DeGraff was supported by the VUMC Multidisciplinary Training Grant in Molecular Endocrinology (5 T32 DK007563-21). Traineeship Award (W81XWH-04-1-0050), Qian Sun is the recipient of a DOD Predoctoral Traineeship Award (W81XWH-07-1-0042) and Srinivas Nandana is the recipient of a DOD Predoctoral Traineeship Award (W81XWH-07-1-0155). The authors wish to acknowledge the assistance of Sherri Tomlinson in the preparation of this manuscript.

References Agoulnik IU, Weigel NL. 2006. Androgen receptor action in hormone-dependent and recurrent prostate cancer. J Cell Biochem 99(2):362–372. Agoulnik IU, Vaid A, Bingman WE, III, Erdeme H, Frolov A, Smith CL, Ayala G, Ittmann MM, Weigel NL. 2005. Role of SRC-1 in the promotion of prostate cancer cell growth and tumor progression. Cancer Res 65(17):7959–7967. Agoulnik IU, Vaid A, Nakka M, Alvarado M, Bingman WE, III, Erdem H, Frolov A, Smith CL, Ayala GE, Ittmann MM, Weigel NL. 2006. Androgens modulate expression of transcription intermediary factor 2, an androgen receptor coactivator whose expression level correlates with early biochemical recurrence in prostate cancer. Cancer Res 66(21):10594–10602. Anzick SL, Kononen J, Walker RL, Azorsa DO, Tanner MM, Guan XY, Sauter G, Kallioniemi OP, Trent JM, Meltzer PS. 1997. AIB1, a steroid receptor coactivator amplified in breast and ovarian cancer. Science 277(5328):965–968. Brivanlou AH, Darnell JE, Jr. 2002. Signal transduction and the control of gene expression. Science 295(5556):813–818. Carey M. 1998. The enhanceosome and transcriptional synergy. Cell 92(1):5–8. Carroll JS, Liu XS, Brodsky AS, Li W, Meyer CA, Szary AJ, Eeckhoute J, Shao W, Hestermann EV, Geistlinger TR, Fox EA, Silver PA, Brown M. 2005. Chromosome-wide mapping of estrogen receptor binding reveals long-range regulation requiring the forkhead protein FoxA1. Cell 122(1):33–43. Chang CY, McDonnell DP. 2005. Androgen receptor-cofactor interactions as targets for new drug discovery. Trends Pharmacol Sci 26(5):225–228. Chiaverotti T, Couto SS, Donjacour A, Mao JH, Nagase H, Cardiff RD, Cunha GR, Balmain A. 2008. Dissociation of epithelial and neuroendocrine carcinoma lineages in the transgenic adenocarcinoma of mouse prostate model of prostate cancer. Am J Pathol 172(1):236–246. Chmelar R, Buchanan G, Need EF, Tilley W, Greenberg NM. 2007. Androgen receptor coregulators and their involvement in the development and progression of prostate cancer. Int J Cancer 120(4):719–733. Cirillo LA, Lin FR, Cuesta I, Friedman D, Jarnik M, Zaret KS. 2002. Opening of compacted chromatin by early developmental transcription factors HNF3 (FoxA) and GATA-4. Mol Cell 9(2):279–289.

The Role of Foxa Proteins in the Regulation of Androgen Receptor Activity

613

Costa RH, Grayson DR, Darnell JE, Jr. 1989. Multiple hepatocyte-enriched nuclear factors function in the regulation of transthyretin and alpha 1-antitrypsin genes. Mol Cell Biol 9 (4):1415–1425. Cunha GR, Donjacour AA. 1989. Mesenchymal-epithelial interactions in the growth and development of the prostate. Cancer Treat Res 46:159–175. Cunha GR, Lung B. 1978. The possible influence of temporal factors in androgenic responsiveness of urogenital tissue recombinants from wild-type and androgen-insensitive (Tfm) mice. J Exp Zool 205(2):181–193. Cunha GR, Lung B, Reese B. 1980. Glandular epithelial induction by embryonic mesenchyme in adult bladder epithelium of BALB/c mice. Invest Urol 17(4):302–304. Cunha GR, Sekkingstad M, Meloy BA. 1983. Heterospecific induction of prostatic development in tissue recombinants prepared with mouse, rat, rabbit and human tissues. Differentiation 24 (2):174–180. Cunha GR, Cooke PS, Kurita T. 2004. Role of stromal-epithelial interactions in hormonal responses. Arch Histol Cytol 67(5):417–434. Donjacour AA, Cunha GR. 1993. Assessment of prostatic protein secretion in tissue recombinants made of urogenital sinus mesenchyme and urothelium from normal or androgen-insensitive mice. Endocrinology 132(6):2342–2350. Espinas ML, Roux J, Pictet R, Grange T. 1995. Glucocorticoids and protein kinase A coordinately modulate transcription factor recruitment at a glucocorticoid-responsive unit. Mol Cell Biol 15 (10):5346–5354. Fong CJ, Sherwood ER, Sutkowski DM, Abu-Jawdeh GM, Yokoo H, Bauer KD, Kozlowski JM, Lee C. 1991. Reconstituted basement membrane promotes morphological and functional differentiation of primary human prostatic epithelial cells. Prostate 19(3):221–235. Fry CJ, Peterson CL. 2002. Transcription. Unlocking the gates to gene expression. Science 295 (5561):1847–1848. Gao N, Zhang J, Rao MA, Case TC, Mirosevich J, Wang Y, Jin R, Gupta A, Rennie PS, Matusik RJ. 2003. The role of hepatocyte nuclear factor-3 alpha (Forkhead Box A1) and androgen receptor in transcriptional regulation of prostatic genes. Mol Endocrinol 17(8):1484–1507. Gao N, Ishii K, Mirosevich J, Kuwajima S, Oppenheimer SR, Roberts RL, Jiang M, Yu X, Shappell SB, Caprioli RM, Stoffel M, Hayward SW, Matusik RJ. 2005. Forkhead box A1 regulates prostate ductal morphogenesis and promotes epithelial cell maturation. Development 132(15):3431–3443. Gierer A. 1974. Molecular models and combinatorial principles in cell differentiation and morphogenesis. Cold Spring Harb Symp Quant Biol 38:951–961. Gupta A, Wang Y, Browne C, Kim S, Case T, Paul M, Wills ML, Matusik RJ. 2008. Neuroendocrine differentiation in the 12T-10 transgenic prostate mouse model mimics endocrine differentiation of pancreatic beta cells. Prostate 68(1):50–60. Heinlein CA, Chang C. 2002. Androgen receptor (AR) coregulators: an overview. Endocr Rev 23 (2):175–200. Heinlein CA, Chang C. 2004. Androgen receptor in prostate cancer. Endocr Rev 25(2):276–308. Lai E, Prezioso VR, Tao WF, Chen WS, Darnell JE, Jr. 1991. Hepatocyte nuclear factor 3 alpha belongs to a gene family in mammals that is homologous to the Drosophila homeotic gene fork head. Genes Dev 5(3):416–427. Lupien M, Eeckhoute J, Meyer CA, Wang Q, Zhang Y, Li W, Carroll JS, Liu XS, Brown M. 2008. FoxA1 translates epigenetic signatures into enhancer-driven lineage-specific transcription. Cell 132(6):958–970. Matusik RJ, Jin RJ, Sun Q, Wang YQ, Gupta A, Nandana S, Case TC, Paul M, Mirsevich J, Oottamasathien S, Thomas J. 2008. Prostate epithelial cell fate. Differentiation 76(6):682–698. Mirosevich J, Gao N, Matusik RJ. 2005. Expression of Foxa transcription factors in the developing and adult murine prostate. Prostate 62(4):339–352. Mirosevich J, Gao N, Gupta A, Shappell SB, Jove R, Matusik RJ. 2006. Expression and role of Foxa proteins in prostate cancer. Prostate 66(10):1013–1028.

614

D.J. DeGraff et al.

Miyamoto H, Yeh S, Wilding G, Chang C. 1998. Promotion of agonist activity of antiandrogens by the androgen receptor coactivator, ARA70, in human prostate cancer DU145 cells. Proc Natl Acad Sci USA 95(13):7379–7384. Myatt SS, Lam EW. 2007. The emerging roles of forkhead box (Fox) proteins in cancer. Nat Rev Cancer 7(11):847–859. Nunez E, Kwon YS, Hutt KR, Hu Q, Cardamone MD, Ohgi KA, Garcia-Bassets I, Rose DW, Glass CK, Rosenfeld MG, Fu XD. 2008. Nuclear receptor-enhanced transcription requires motorand LSD1-dependent gene networking in interchromatin granules. Cell 132(6):996–1010. Okita K, Ichisaka T, Yamanaka S. 2007. Generation of germline-competent induced pluripotent stem cells. Nature 448(7151):313–317. Pani L, Overdier DG, Porcella A, Qian X, Lai E, Costa RH. 1992. Hepatocyte nuclear factor 3 beta contains two transcriptional activation domains, one of which is novel and conserved with the Drosophila fork head protein. Mol Cell Biol 12(9):3723–3732. Rapa I, Ceppi P, Bollito E, Rosas R, Cappia S, Bacillo E, Porpiglia F, Berruti A, Papotti M, Volante M. 2008. Human ASH1 expression in prostate cancer with neuroendocrine differentiation. Mod Pathol 21:700–707. Robyr D, Gegonne A, Wolffe AP, Wahli W. 2000. Determinants of vitellogenin B1 promoter architecture. HNF3 and estrogen responsive transcription within chromatin. J Biol Chem 275 (36):28291–28300. Saporita AJ, Ai J, Wang Z. 2007. The Hsp90 inhibitor, 17-AAG, prevents the ligand-independent nuclear localization of androgen receptor in refractory prostate cancer cells. Prostate 67 (5):509–520. Sasaki H, Hui C, Nakafuku M, Kondoh, H. 1997. Development. 124(7):1313–1322. Stafford JM, Wilkinson JC, Beechem JM, Granner DK. 2001. Accessory factors facilitate the binding of glucocorticoid receptor to the phosphoenolpyruvate carboxykinase gene promoter. J Biol Chem 276(43):39885–39891. Stearns ME, McGarvey T. 1992. Prostate cancer: therapeutic, diagnostic, and basic studies. Lab Invest 67(5):540–552. Sugimura Y, Cunha GR, Bigsby RM. 1986. Androgenic induction of DNA synthesis in prostatic glands induced in the urothelium of testicular feminized (Tfm/Y) mice. Prostate 9(3):217–225. Takahashi K, Yamanaka S. 2006. Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell 126(4):663–676. Tan J, Sharief Y, Hamil KG, Gregory CW, Zang DY, Sar M, Gumerlock PH, deVere White RW, Pretlow TG, Harris SE, Wilson EM, Mohler JL, French FS. 1997. Dehydroepiandrosterone activates mutant androgen receptors expressed in the androgen-dependent human prostate cancer xenograft CWR22 and LNCaP cells. Mol Endocrinol 11(4):450–459. Truica CI, Byers S, Gelmann EP. 2000. Beta-catenin affects androgen receptor transcriptional activity and ligand specificity. Cancer Res 60(17):4709–4713. Voeller HJ, Truica CI, Gelmann EP. 1998. Beta-catenin mutations in human prostate cancer. Cancer Res 58(12):2520–2523. Wang W, Epstein JI. 2008. Small cell carcinoma of the prostate. A morphologic and immunohistochemical study of 95 cases. Am J Surg Pathol 32(1):65–71. Wang JC, Stromstedt PE, O’Brien RM, Granner DK. 1996. Hepatic nuclear factor 3 is an accessory factor required for the stimulation of phosphoenolpyruvate carboxykinase gene transcription by glucocorticoids. Mol Endocrinol 10(7):794–800. Wang Q, Li W, Liu XS, Carroll JS, Janne OA, Keeton EK, Chinnaiyan AM, Pienta KJ, Brown M. 2007. A hierarchical network of transcription factors governs androgen receptor-dependent prostate cancer growth. Mol Cell 27(3):380–392. Wernig M, Meissner A, Foreman R, Brambrink T, Ku M, Hochedlinger K, Bernstein BE, Jaenisch R. 2007. In vitro reprogramming of fibroblasts into a pluripotent ES-cell-like state. Nature 448 (7151):318–324. Wolf I, Bose S, Williamson EA, Miller CW, Karlan BY, Koeffler HP. 2007. FOXA1: growth inhibitor and a favorable prognostic factor in human breast cancer. Int J Cancer 120 (5):1013–1022.

The Role of Foxa Proteins in the Regulation of Androgen Receptor Activity

615

Young CY, Andrews PE, Tindall DJ. 1995. Expression and androgenic regulation of human prostate-specific kallikreins. J Androl 16(2):97–99. Yu X, Gupta A, Wang Y, Suzuki K, Mirosevich J, Orgebin-Crist MC, Matusik RJ. 2005. Foxa1 and Foxa2 interact with the androgen receptor to regulate prostate and epididymal genes differentially. Ann NY Acad Sci 1061:77–93. Yu X, Suzuki K, Wang Y, Gupta A, Jin R, Orgebin-Crist MC, Matusik R. 2006. The role of forkhead box A2 to restrict androgen-regulated gene expression of lipocalin 5 in the mouse epididymis. Mol Endocrinol 20(10):2418–2431. Yu J, Vodyanik MA, Smuga-Otto K, Antosiewicz-Bourget J, Frane JL, Tian S, Nie J, Jonsdottir GA, Ruotti V, Stewart R, Slukvin II, Thomson JA. 2007. Induced pluripotent stem cell lines derived from human somatic cells. Science 318(5858):1917–1920. Zhang J, Thomas TZ, Kasper S, Matusik RJ. 2000. A small composite probasin promoter confers high levels of prostate-specific gene expression through regulation by androgens and glucocorticoids in vitro and in vivo. Endocrinology 141(12):4698–4710. Zhu Y, Qi C, Jain S, Le Beau MM, Espinosa R, III, Atkins GB, Lazar MA, Yeldandi AV, Rao MS, Reddy JK. 1999. Amplification and overexpression of peroxisome proliferator-activated receptor binding protein (PBP/PPARBP) gene in breast cancer. Proc Natl Acad Sci USA 96 (19):10848–10853.

Androgen Receptor as a Licensing Factor for DNA Replication Donald J. Vander Griend and John T. Isaacs

Abstract The Androgen Receptor (AR) is a steroid transcription factor, the activity of which is the primary focus of androgen ablation therapies for advanced prostate cancer. In prostate cancers, the AR acquires gain-of-function changes allowing it to drive prostate cancer cell survival and proliferation in a cell-autonomous manner. As part of this malignancy-associated gain-of-function, AR acquires a role in licensing for DNA replication in prostate cancer cells. In its role as a licensing factor, AR must be degraded during mitosis in order to allow re-licensing in the subsequent cell cycle. This conclusion is supported by the demonstration that acute enhanced expression of AR in prostate cancer cells results in its incomplete degradation in mitosis. This lack of mitotic AR degradation inhibits subsequent cell proliferation due to the inability to re-license all origins of replication needed for the next round of cell division. These data provide a unifying paradigm to clarify a number of unresolved observations in prostate cancer research. In addition, they provide a rationale for a new therapeutic approach for prostate cancer based upon stabilization of AR. Androgen deprivation therapy (ADT) has been the mainstay of prostate cancer treatment since the seminal discovery by Charles Huggins and Clarence Hodges in 1941 that castration or estrogen administration significantly aided patients with advanced prostate cancer (Huggins and Hodges 1941). Inevitably, however, there is a relapse to ADT due to the growth of resistant prostate cancer cells. There are a series of mechanisms for the development of ADT-resistant prostate cancer (Isaacs and Isaacs 2004). These mechanisms center on the role of the molecular target of ADT, the androgen receptor (AR). Inhibiting the intracellular signaling initiated within prostate cancer cells by AR has been a major focus of prostate cancer research. This research has produced a myriad of chemical inhibitors of such AR signaling that are used in the clinic (Isaacs 1994; Singh et al. 2006). Unfortunately, while all these inhibitors produce an initial therapeutic response, this response is

J.T. Isaacs (*) The Chemical Therapeutics Program and the Brady Urological Institute, The Johns Hopkins University School of Medicine, Cancer Research Building #1, Room 1M40, 1650 Orleans street, Baltimore, MD 21231, USA, E-mail: [email protected]

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4_26, # Springer Science + Business Media, LLC 2009

619

620

D.J.V. Griend, J.T. Isaacs

universally followed by relapse. A better understanding of the differences in the functions of the AR in benign vs. malignant prostate cells is needed to overcome the present limitations in AR targeted therapies. The AR gene was cloned in 1988 and is located on the long arm of the X chromosome (i.e., Xq11.2), and thus males are hemizygous for this important gene (Chang et al. 1988; Lubahn et al. 1988). The AR protein consists of three functional domains, an N-terminal transactivation domain, a DNA-binding domain, and a C-terminal transactivation domain that contains the ligand-binding domain (LBD) (Heinlein and Chang 2004). Circulating testosterone (T) is converted intracellularly to dihydrotestosterone (DHT) by 5-alpha reductase, which binds in the cytoplasm to the LBD of AR monomers associated with a series of chaperone proteins, which include Heat Shock Protein-90 (HSP-90). DHT binding to the LBD dissociates HSP-90 and allows N- to C-terminal intramolecular interaction in AR monomers (Klokk et al. 2006). This conformational change is associated with translocation to the nucleus, where AR dimerizes and binds to DNA. The best-characterized sites of such AR binding are to consensus androgen response elements (AREs) within the enhancers and promoters of AR-regulated genes like the prostate-specific antigen (PSA) gene. In the normal prostate, androgen binds to AR in the nuclei of stromal cells, which causes the production of diffusible growth factors, collectively termed ‘‘andromedins,’’ which diffuse from the stromal compartment across the basement membrane to enter the epithelial compartment (Kurita et al. 2001; Cunha et al. 2004). Once in the epithelial compartment, these andromedins bind to their cognate receptors on AR-negative prostate basal cells stimulating both proliferation and terminal differentiation into proliferatively quiescent, AR-positive secretory-luminal epithelial cells. Ligand binding to AR in the secretory-luminal epithelial cell is associated with suppression of cell proliferation and terminal differentiation, as demonstrated by the expression of AR-regulated genes, such as PSA; and cell-cycle inhibitors, such as p27 (De Marzo et al. 1998; Litvinov et al. 2003). Experimentally, ectopic expression of AR in prostate basal cells induces growth arrest of these cells with elevated p21 and p27 expression (Litvinov et al. 2003; Berger et al. 2006). While these terminally differentiated secretory-luminal cells do not proliferate in response to stromal andromedins, they require these andromedins for their survival (Kurita et al. 2001). Androgen depletion results in cessation of stromal production of andromedins, induces concomitant apoptosis of AR-positive epithelial luminal cells, and inhibits proliferation of AR-negative basal cells, which result in the regression of the benign prostate gland (Kyprianou and Isaacs 1988; Kurita et al. 2001). This prostate regression is reversible; readministering androgen reinitiates andromedin secretion by stromal cells thereby inducing prostate basal cell proliferation and maturation into secretory-luminal cells, which restores the prostate gland (Kyprianou and Isaacs 1988). Thus, in the benign prostate, AR-stimulated growth occurs via a stroma-dependent paracrine interaction, and AR displays a growth-suppressor function within the secretory-luminal epithelial cells. In direct contrast, androgen-sensitive (AS) prostate cancer cells express AR, and occupancy of AR by its ligand in their nucleus directly (i.e., cell autonomously)

Androgen Receptor as a Licensing Factor for DNA Replication

621

regulates their proliferation and survival (Gao et al. 2001). This conversion to a cell autonomous autocrine mechanism of AR-stimulated growth control in AS prostate cancer cells, independent of AR expression in supporting stromal cells, occurs early during prostatic carcinogenesis (Gao et al. 2001). As part of this malignant conversion, AR undergoes a molecular switch from its ability to suppress proliferation of normal prostatic epithelia to directly stimulating the proliferation of prostate cancer cells (Gao et al. 2001; Litvinov et al. 2003). Such a molecular switch involves gainof-function changes that produce novel AR activities in prostate cancer cells. One such novel gain-of-function change involves DNA rearrangement such that the promoter of the TMPRSS2 gene, which contains AREs, is translocated to confer androgen responsiveness upon select members of the ETS transcription factor family (Tomlins et al. 2005). Besides these malignancy-dependent transcriptional changes, additional molecular changes result in AR becoming a critical factor for DNA replication. AR becomes part of the protein complex required to ‘‘license’’ DNA replication in AS prostate cancer cells (Litvinov et al. 2006). An overview of the DNA licensing process is necessary to appreciate the significances of this newly identified malignancy-specific role of AR as a licensing factor in prostate cancer cells.

1 DNA Replication Licensing Chromosomal DNA must be precisely replicated once and only once during each cell cycle to maintain cellular viability and genetic stability. A process known as DNA replication licensing ‘‘marks’’ specific sites in the DNA, known as origins of replication (i.e., OR), which establish early in G1 where replication forks will initiate DNA synthesis during S-phase. Such licensing also prevents reinitiation of DNA replication until the next cycle (Blow and Tanaka 2005; Lei 2005; Cvetic and Walter 2006). DNA replication licensing is initiated by the ordered binding of a series of licensing factors at OR sites (Fig. 1). The first to bind are the Origins of Replication Complex (ORC) proteins. The ORC complex is composed of six subunits, numbered Orc 1–6. Once the ORC is bound to the OR, Cdc6 binds via interactions with Orc1. This DNA-bound ORC/Cdc6 complex is required for binding of Cdt1, which is needed for loading ‘‘minichromosome maintenance’’ (MCM) proteins complexes at the ORs. Multiple MCM complexes are loaded at each OR, and each MCM complex is composed of a heterohexamer of related proteins, termed Mcm2–7. Once each Mcm2–7 complex is bound during G1, it forms a ring structure around the DNA at the OR to complete the formation known as the pre-replication complex (pre-RC). The sequential order of assembly of the pre-RC is regulated by Cyclin-dependent kinases (CDKs) via phosphorylation of the licensing factors, Orc1, Cdc6, and Cdt1 (Blow and Tanaka 2005; Cvetic and Walter 2006). Completion of pre-RC formation renders the ORs fully licensed for DNA replication leading into S-phase (Blow and Tanaka 2005; Cvetic and Walter 2006). Upon entry into S phase, Mcm2–7 complex rings are activated and move bidirectionally to allow assembly of the DNA replisome (Cvetic and Walter 2006).

622

D.J.V. Griend, J.T. Isaacs

Fig. 1 DNA replication licensing by formation of the pre-replication complex (Pre-RC). The Pre-RC is assembled via a stepwise binding of DNA licensing factors to origins of DNA replication in early G1 of the cell cycle. Origin replication complex (ORC) binding recruits Cdc6, which then recruits Cdt1, which facilitates the formation of the MCM complex. During and after DNA replication in S-phase, DNA licensing factors are prevented by a series of redundant mechanisms to prevent relicensing and aberrant re-replication. Such re-replication events cause widespread DNA damage and repair and can promote carcinogenesis

The earliest clues to DNA licensing came about via cell fusion experiments (Rao and Johnson 1970). Fusion of a G1 cell to an S cell prematurely initiates DNA replication in the G1 cell nucleus, which indicates that a G1 cell is capable of DNA replication but lacks necessary factors expressed during S phase. In contrast, fusion of a G2 cell to an S cell does not initiate DNA replication in the G2 nucleus, which indicates that the G2 cell is not able to relicense its DNA once it has replicated. Thus, cells in G1 are ‘‘licensed’’ for DNA replication, and this license is removed after DNA replication occurs in order to prevent re-replication. The mechanism for inhibition of re-replication during the same cell cycle is controlled by Cyclindependent kinase (CDK) activity throughout the remainder of the cell cycle (S, G2, and M). CDK activity inhibits relicensing by stimulating degradation of Orc1 and Cdt1, inactivation and nuclear export of Cdc6, and inactivation of ORC proteins (Takeda and Dutta 2005; Cvetic and Walter 2006). In addition, although initiation of DNA replication occurs in S-phase, many of the members of the preRC remain associated with ORs during G2, and this occupancy prevents the

Androgen Receptor as a Licensing Factor for DNA Replication

623

sequentially ordered binding of licensing factors at OR sites needed for relicensing and thus reinitiation of DNA replication. These combined effects allow only one round of DNA replication per cycle (Takeda and Dutta 2005). This also requires that the licensing factors which remain bound to ORs in G2 be removed in mitosis or early G1 via mechanisms, which include ubiquitination and subsequent proteasomal degradation, to allow reinitiation of DNA replication in the next cell cycle (Takeda and Dutta 2005).

2 Regulation of DNA Licensing A variety of mechanisms inhibit re-replication in an array of experimental organisms. In mammals, Cdt1 is the major target that is repressed upon entry into S-phase, though Cdc6 and ORC also are regulated. Overexpression of Cdt1 alone or with Cdc6 causes re-replication in p53 / human cancer cells (Vaziri et al. 2003). Mammalian cells use two mechanisms to repress Cdt1 after the initiation of S-phase (Fig. 1). First, Geminin binds to Cdt1 and inhibits its activity until late M phase, when the Anaphase-Promoting Complex (APC) becomes activated and degrades Geminin (Arias and Walter 2007). Second, Cdt1 is degraded via two distinct E3 ubiqutin ligases, the SCFskp2 and Cul4DDB1Cdt2 (Takeda et al. 2005; Arias and Walter 2007). Thus, in mammals two redundant pathways fully ensure inhibition of Cdt1 upon entry into S phase. The mechanisms for Cdc6 inhibition are defined well in yeast, but many of these mechanisms have yet to be fully defined in mammalian systems (Hook et al. 2007). Cdc6 may be exported from the nucleus and sequestered in the cytoplasm upon entry into S phase (Takeda and Dutta 2005). The significance of controlling relicensing and subsequent re-replication is paramount to preventing DNA damage and carcinogenesis.

3 DNA Licensing and Cancer While DNA replication is an essential component of tumorigenesis, it was somewhat surprising that many factors involved in DNA licensing drive tumorigenesis and malignant transformation (Hook et al. 2007). Two defined mechanisms of transformation by DNA replication licensing factors depend upon DNA replicative functions or function independent of their role in DNA replication licensing. The first mechanism of transformation involves re-replication driving a DNA damage response. In many types of mammalian cells, Cdt1 overexpression alone can cause re-replication, which is augmented by Cdc6 overexpression. Up-regulation of these two proteins is observed in several models of tumorigenesis (Arentson et al. 2002; Karakaidos et al. 2004; Gonzalez et al. 2006; Tatsumi et al. 2006; Liontos et al. 2007). Re-replication of DNA replication causes replication bubbles (replication forks forming within replication forks) and replication fork collision

624

D.J.V. Griend, J.T. Isaacs

that generates DNA double-strand breaks. Both events activate the DNA damage checkpoint ATM and ATR kinases, which activate the Chk1 and Chk2 kinases to evoke G2/M cell cycle block. Such a halt in the cell cycle requires p53 activity, and deletion of p53 may synergize with overexpression of Cdt1 and Cdc6 to cause genomic instability (Iizuka et al. 2007). Furthermore, re-replication of DNA results in cell ploidy changes, thereby contributing to genomic instability, gene amplification, and chromosome rearrangement. A second mechanism of Cdt1 and Cdc6 transformation may be independent of their ability to induce re-replication. Rather, DNA replication licensing at origins of replication may result in epigenetic changes that lead to carcinogenesis. Such a novel mechanism has been defined whereby Cdc6 is recruited to the INK4/ARF tumor suppressor locus, recruits histone deacetylases, and stimulates heterochromatinization (Gonzalez et al. 2006). This mechanism may explain why tumors overexpressing Cdc6 also tend to repress the three tumor suppressors expressed from this locus (p15INK4b, ARF, and p16INK4a) (Gonzalez et al. 2006). In prostate cancer, MCM2 and MCM7 are overexpressed as markers of DNA replication (Meng et al. 2001; Padmanabhan et al. 2004; Dudderidge et al. 2007). Evaluation of decreased Geminin expression (a negative regulator of Cdt1) and increased patient survival revealed no statistical significance, although the decreased expression of MCM2 and increased patient survival was statistically significant. These data support the hypothesis that an increase in Cdt1 and Cdc6 expression would correlate with poorer patient prognosis. Such markers may reveal the subpopulations of prostate cancer patients at greater risk for disease progression, since Cdt1 and Cdc6 overexpression directly contribute to genomic DNA changes, instability and gene amplification, and inhibition of tumor suppressor genes.

4 The Androgen Receptor Functions to Coordinate DNA Replication Licensing The AR plays a critical role in initiating DNA replication and cell division in prostate cancers. Androgen-stimulated proliferation of androgen-sensitive prostate cancer cells can be blocked by coexposure to antiandrogens (e.g., bicalutamide) in early G1 before the formation of the pre-replicative complexes that license DNA replication (Cifuentes et al. 2003; Bai et al. 2005). In contrast, proliferation is not prevented if such antiandrogen coexposure is delayed until a later point in G1 (i.e., after pre-RC formation) (Cifuentes et al. 2003; Bai et al. 2005). These results document that the major point of androgen regulation in androgen-sensitive prostate cancer cell proliferation occurs in early G1, at the point when licensing for DNA replication occurs. AR coimmunoprecipitates with all three members of the DNA-licensing complex, namely ORC2, Cdt1, and Cdc6 (Fig. 2) in LNCaP prostate cancer cells. As a licensing factor, AR is degraded in mitosis of each cell cycle in these androgen-sensitive prostate cancer cells (Litvinov et al. 2006). This loss of AR during mitosis is a proteasome-dependent process. As a control for these results in

Androgen Receptor as a Licensing Factor for DNA Replication

625

Fig. 2 AR interacts with all three members of the DNA licensing complex. LNCaP prostate cancer cells are AR positive and androgen sensitive. As expected, LNCaP cells express AR as well as the members of the DNA licensing complex, ORC2, Cdt1, and Cdc6 (lane 1). Immunoprecipitation (lane 2 – IgG control, lane 3 – anti-AR antibody) of the AR from LNCaP and western blotting revealed that ORC2, Cdt1, and Cdc6 coimmunoprecipitated with AR. These data strongly support a role for the AR in modulating DNA replication licensing to drive prostate cancer cell proliferation

androgen-sensitive prostate cancer cells, similar analyses were performed on benign prostate stromal cells (PRSCs). These PRSCs express low levels of AR, but do not depend upon androgen for their survival and growth and AR has no role in DNA replication (Litvinov et al. 2006). AR acts as a ligand-dependent transcription factor for the production of epithelial growth factors (andromedins) in these PRSCs but does not control their proliferation (Gao et al. 2001; Litvinov et al. 2006). AR is not degraded in these prostate stromal cells during mitosis. Upon growth to confluence and cell cycle arrest in G0 due to contact inhibition, PRSCs continue to express detectable AR even though Mcm2 expression was undetectable (Litvinov et al. 2006). Thus, the function of AR as a licensing factor and its degradation in mitosis is prostate cancer-specific and not a universal phenomenon. The realization that AR functions in DNA replication licensing in malignant prostate cells provides an explanation for a longstanding experimental paradox with regard to the biphasic androgen dose-response of androgen-sensitive prostate cancer cells to androgen. Multiple studies have documented that there is an optimal level for androgen above which there is a paradoxical decrease in vitro growth of androgensensitive prostate cancer cell lines (Langeler et al. 1993; Kokontis et al. 1998). The optimal level of androgen has been documented to be inversely dependent upon the level of AR protein expressed by the prostate cancer cells. Thus, prostate cancer cells which express the highest levels of AR protein are the most inhibited in their growth when the level of androgen is raised acutely in the media (Langeler et al. 1993; Kokontis et al. 1998). In addition, the most consistent molecular change in prostate cancer progressing within an androgen depleted environment is a 2- to 4fold up-regulation of AR (Chen et al. 2004). These observations are consistent with an adaptive mechanism in androgen-sensitive prostate cancer cells to maintain AR protein at a level optimal for growth during periods of fluctuations in ligand level.

626

D.J.V. Griend, J.T. Isaacs

Along these lines, androgen is known to stabilize AR protein from degradation but to down-regulate AR transcription (Quarmby et al. 1990; Krongrad et al. 1991; Kemppainen et al. 1992). This raises the possibility that when androgen levels are raised acutely, AR protein is stabilized to a point where it is not degraded sufficiently during mitosis (Fig. 2). This would result in a fraction of the origin of replication sites which were licensed and used during the previous cell cycle still retaining AR, and thus these origins of replication would not able to be relicensed in G1 of the subsequent cell cycle. This situation would allow entry into S-phase but would prevent complete DNA replication, which would induce an arrest in early Sphase. This is significant since 10 mM of the antiandrogen bicalutamide produced the same overall degree of growth inhibition due to its ability to arrest cells in G1 rather than early S-phase (Cifuentes et al. 2003; Bai et al. 2005). This raises the issue of whether a similar disruption of DNA replication licensing, and thus growth inhibition, can be induced by increasing the level of AR directly without increasing ligand in androgen-sensitive prostate cancer cells.

5 Therapeutic Implications Defining the mechanism(s) by which AR promotes DNA licensing and cell division is critical to the understanding of prostate cancer initiation and defining appropriate antiandrogen therapies for prostate cancer patients. The most critical question is how the AR interacts with DNA licensing factors to promote DNA licensing. Many of the genomic areas of active transcription also serve as the earliest sites of DNA replication (Tabancay and Forsburg 2006). Therefore, AR binding to AREs in the promoters of androgen-regulated genes may serve to recruit licensing factors to ARregulated genes. The limited number of AR-regulated genes (i.e., dozens/genome), however, is not able to account for the large number of origins of replication (i.e., thousands/genome). Regardless of the mechanism for how the AR is captured as part of the DNA replication licensing process, such a gain of function requires that AR is degraded during mitosis since lack of such degradation results in subsequent S-phase arrest of androgen-sensitive prostate cancer cells. Indeed, experimental data demonstrate that androgen-sensitive prostate cancer cells have an autoregulatory ability to maintain AR protein level optimal for the particular ligand concentration (Quarmby et al. 1990; Krongrad et al. 1991; Kemppainen et al. 1992). This adaptation results in down-regulation of AR in the presence of high ligand levels and upregulation when androgen levels are low (Langeler et al. 1993; Kokontis et al. 1998). However, such adaptation requires a critical time frame. This conclusion is supported by the observations that acute increases in AR protein levels via either direct transcriptional up-regulation or ligand-induced AR protein stability results in growth inhibition. Such growth inhibition is due to insufficient AR degradation in mitosis, which prevents full relicensing of DNA replication in the next cycle (Fig. 3). Several novel therapeutic approaches for prostate cancer can be imagined based upon this new revelation concerning the role and consequence of AR as a licensing factor. For example, the time dependence for AR adaptation provides a theoretical

Androgen Receptor as a Licensing Factor for DNA Replication

627

Fig. 3 Schematic for AR protein expression during the cell cycle of prostate cancer cells. (a) Physiologic AR protein expression in androgen-sensitive prostate cancer cells, where AR is degraded during mitosis and re-expressed to promote DNA licensing for another round of DNA replication. (b) AR protein amplification in prostate cancer cells after ADT relapse. In the absence of androgen, AR protein is overexpressed but still degraded during mitosis to promote DNA licensing the subsequent proliferation cycle. (c) Stabilization of AR using androgen, as in the case of IADT, would prevent AR degradation during mitosis and thereby inhibit proper DNA licensing in the following cell cycle. Prevention of DNA licensing would prevent DNA replication and block prostate cancer cell proliferation

rationale for improving intermittent androgen deprivation therapy (IADT). ADT is given for a limited time followed by a relief period to allow recovery to a physiological level of androgen (Mottet et al. 2005). During each cycle of ADT, AR protein is up-regulated. When ADT is stopped, the rate at which tissue androgen levels return will determine whether adaptive changes have sufficient time to downregulate the elevated level of AR to prevent relicensing problems. The efficacy of IADT could be enhanced, therefore, by acutely administering exogenous androgen replacement at the time ADT is halted in each cycle. In this way, insufficient time would be provided to down-regulate AR so that DNA replication relicensing problems would occur, which would growth inhibit the prostate cancer cells. There are additional ways to maximize the relicensing problems in prostate cancer cells. One way is to take advantage of the fact that mitotic degradation of AR protein is prevented by proteasome inhibitors (Litvinov et al. 2006). The growth of AR-sensitive prostate cancer cell lines is inhibited when treated with proteasome

628

D.J.V. Griend, J.T. Isaacs

inhibitors (Williams et al. 2003; Papandreou et al. 2004). A novel approach would combine proteasome inhibitors with high-dose androgen in patients who have failed previous ADT treatments. A further approach is to design and screen small molecule libraries to identify compounds that stabilize the AR protein from mitotic degradation (Isaacs 1994).

References Arentson E, Faloon P, Seo J, Moon E, Studts JM, Fremont DH, and Choi K (2002). Oncogenic potential of the DNA replication licensing protein CDT1. Oncogene 21(8): 1150–1158 Arias EE and Walter JC (2007). Strength in numbers: preventing rereplication via multiple mechanisms in eukaryotic cells. Genes Dev 21(5): 497–518 Bai VU, Cifuentes E, Menon M, Barrack ER, and Reddy GP (2005). Androgen receptor regulates Cdc6 in synchronized LNCaP cells progressing from G1 to S phase. J Cell Physiol 204(2): 381–387 Berger R, Lin DI, Nieto M, Sicinska E, Garraway LA, Adams H, Signoretti S, Hahn WC, and Loda M (2006). Androgen-dependent regulation of Her-2/neu in prostate cancer cells. Cancer Res 66 (11): 5723–5728 Blow JJ and Tanaka TU (2005). The chromosome cycle: coordinating replication and segregation. Second in the cycles review series. EMBO Rep 6(11): 1028–1034 Chang CS, Kokontis J, and Liao ST (1988). Molecular cloning of human and rat complementary DNA encoding androgen receptors. Science 240(4850): 324–326 Chen CD, Welsbie DS, Tran C, Baek SH, Chen R, Vessella R, Rosenfeld MG, and Sawyers CL (2004). Molecular determinants of resistance to antiandrogen therapy. Nat Med 10(1): 33–39 Cifuentes E, Croxen R, Menon M, Barrack ER, and Reddy GP (2003). Synchronized prostate cancer cells for studying androgen regulated events in cell cycle progression from G1 into S phase. J Cell Physiol 195(3): 337–345 Cunha GR, Cooke PS, and Kurita T (2004). Role of stromal-epithelial interactions in hormonal responses. Arch Histol Cytol 67(5): 417–434 Cvetic CA and Walter JC (2006). Getting a grip on licensing: mechanism of stable Mcm2–7 loading onto replication origins. Mol Cell 21(2): 143–144 De Marzo AM, Meeker AK, Epstein JI, and Coffey DS (1998). Prostate stem cell compartments: expression of the cell cycle inhibitor p27Kipl in normal, hyperplastic, and neoplastic cells. Am J Pathol 153(3): 911–919 Dudderidge TJ, McCracken SR, Loddo M, Fanshawe TR, Kelly JD, Neal DE, Leung HY, Williams GH, and Stoeber K (2007). Mitogenic growth signalling, DNA replication licensing, and survival are linked in prostate cancer. Br J Cancer 96(9): 1384–1393 Gao J, Arnold JT, and Isaacs JT (2001). Conversion from a paracrine to an autocrine mechanism of androgen-stimulated growth during malignant transformation of prostatic epithelial cells. Cancer Res 61(13): 5038–5044 Gonzalez S, Klatt P, Delgado S, Conde E, Lopez-Rios F, Sanchez-Cespedes M, Mendez J, Antequera F, and Serrano M (2006). Oncogenic activity of Cdc6 through repression of the INK4/ARF locus. Nature 440(7084): 702–706 Heinlein CA and Chang C (2004). Androgen receptor in prostate cancer. Endocr Rev 25(2): 276–308 Hook SS, Lin JJ, and Dutta A (2007). Mechanisms to control rereplication and implications for cancer. Curr Opin Cell Biol 19: 663–671 Huggins C, Stevens RE, and Hodges CV (1941). Studies on prostate cancer: II. the effects of castration on advanced carcinoma of the prostate gland. Arch Surg 43: 209–223 Iizuka M, Sarmento OF, Sekiya T, Scrable H, Allis CD, and Smith MM (2007). Hbo1 Links p53-dependent stress signaling to DNA replication licensing. Mol Cell Biol 28(1): 140–153

Androgen Receptor as a Licensing Factor for DNA Replication

629

Isaacs JT (1994). Role of androgens in prostatic cancer. Vitam Horm 49: 433–502 Isaacs JT and Isaacs WB (2004). Androgen receptor outwits prostate cancer drugs. Nat Med 10(1): 26–27 Karakaidos P, Taraviras S, Vassiliou LV, Zacharatos P, Kastrinakis NG, Kougiou D, Kouloukoussa M, Nishitani H, Papavassiliou AG, Lygerou Z, and Gorgoulis VG (2004). Overexpression of the replication licensing regulators hCdt1 and hCdc6 characterizes a subset of non-small-cell lung carcinomas: synergistic effect with mutant p53 on tumor growth and chromosomal instability – evidence of E2F-1 transcriptional control over hCdt1. Am J Pathol 165(4): 1351–1365 Kemppainen JA, Lane MV, Sar M, and Wilson EM (1992). Androgen receptor phosphorylation, turnover, nuclear transport, and transcriptional activation. Specificity for steroids and antihormones. J Biol Chem 267(2): 968–974 Klokk TI, Kurys P, Elbi C, Nagaich AK, Hendarwanto A, Slagsvold T, Chang CY, Hager GL, and Saatcioglu F (2006). Ligand-specific dynamics of the androgen receptor at its response element in living cells. Mol Cell Biol 27(5): 1823–1843 Kokontis JM, Hay N, and Liao S (1998). Progression of LNCaP prostate tumor cells during androgen deprivation: hormone-independent growth, repression of proliferation by androgen, and role for p27Kip1 in androgen-induced cell cycle arrest. Mol Endocrinol 12(7): 941–953 Krongrad A, Wilson CM, Wilson JD, Allman DR, and McPhaul MJ (1991). Androgen increases androgen receptor protein while decreasing receptor mRNA in LNCaP cells. Mol Cell Endocrinol 76(1–3): 79–88 Kurita T, Wang YZ, Donjacour AA, Zhao C, Lydon JP, O’Malley BW, Isaacs JT, Dahiya R, and Cunha GR (2001). Paracrine regulation of apoptosis by steroid hormones in the male and female reproductive system. Cell Death Differ 8(2): 192–200 Kyprianou N and Isaacs JT (1988). Activation of programmed cell death in the rat ventral prostate after castration. Endocrinology 122(2): 552–562 Langeler EG, van Uffelen CJ, Blankenstein MA, van Steenbrugge GJ, and Mulder E (1993). Effect of culture conditions on androgen sensitivity of the human prostatic cancer cell line LNCaP. Prostate 23(3): 213–223 Lei M (2005). The MCM complex: its role in DNA replication and implications for cancer therapy. Curr Cancer Drug Targets 5(5): 365–380 Liontos M, Koutsami M, Sideridou M, Evangelou K, Kletsas D, Levy B, Kotsinas A, Nahum O, Zoumpourlis V, Kouloukoussa M, Lygerou Z, Taraviras S, Kittas C, Bartkova J, Papavassiliou AG, Bartek J, Halazonetis TD, and Gorgoulis VG (2007). Deregulated overexpression of hCdt1 and hCdc6 promotes malignant behavior. Cancer Res 67(22): 10899–10909 Litvinov IV, De Marzo AM, and Isaacs JT (2003). Is the Achilles’ heel for prostate cancer therapy a gain of function in androgen receptor signaling? J Clin Endocrinol Metab 88(7): 2972–2982 Litvinov IV, Vander Griend DJ, Antony L, Dalrymple S, De Marzo AM, Drake CG, and Isaacs JT (2006). Androgen receptor as a licensing factor for DNA replication in androgen-sensitive prostate cancer cells. Proc Natl Acad Sci U S A 103(41):15085–15090 Lubahn DB, Joseph DR, Sullivan PM, Willard HF, French FS, and Wilson EM (1988). Cloning of human androgen receptor complementary DNA and localization to the X chromosome. Science 240(4850): 327–330 Meng MV, Grossfeld GD, Williams GH, Dilworth S, Stoeber K, Mulley TW, Weinberg V, Carroll PR, and Tlsty TD (2001). Minichromosome maintenance protein 2 expression in prostate: characterization and association with outcome after therapy for cancer. Clin Cancer Res 7(9): 2712–2718 Mottet N, Lucas C, Sene E, Avances C, Maubach L, and Wolff JM (2005). Intermittent androgen castration: a biological reality during intermittent treatment in metastatic prostate cancer? Urol Int 75(3): 204–208 Padmanabhan V, Callas P, Philips G, Trainer TD, and Beatty BG (2004). DNA replication regulation protein Mcm7 as a marker of proliferation in prostate cancer. J Clin Pathol 57 (10): 1057–1062

630

D.J.V. Griend, J.T. Isaacs

Papandreou CN, Daliani DD, Nix D, Yang H, Madden T, Wang X, Pien CS, Millikan RE, Tu SM, Pagliaro L, Kim J, Adams J, Elliott P, Esseltine D, Petrusich A, Dieringer P, Perez C, and Logothetis CJ (2004). Phase I trial of the proteasome inhibitor bortezomib in patients with advanced solid tumors with observations in androgen-independent prostate cancer. J Clin Oncol 22(11): 2108–2121 Quarmby VE, Yarbrough WG, Lubahn DB, French FS, and Wilson EM (1990). Autologous downregulation of androgen receptor messenger ribonucleic acid. Mol Endocrinol 4(1): 22–28 Rao PN and Johnson RT (1970). Mammalian cell fusion: studies on the regulation of DNA synthesis and mitosis. Nature 225(5228): 159–164 Singh P, Uzgare A, Litvinov I, Denmeade SR, and Isaacs JT (2006). Combinatorial androgen receptor targeted therapy for prostate cancer. Endocr Relat Cancer 13(3): 653–666 Tabancay AP, Jr. and Forsburg SL (2006). Eukaryotic DNA replication in a chromatin context. Curr Top Dev Biol 76: 129–184 Takeda DY and Dutta A (2005). DNA replication and progression through S phase. Oncogene 24 (17): 2827–2843 Takeda DY, Parvin JD, and Dutta A (2005). Degradation of Cdt1 during S phase is Skp2independent and is required for efficient progression of mammalian cells through S phase. J Biol Chem 280(24): 23416–23423 Tatsumi Y, Sugimoto N, Yugawa T, Narisawa-Saito M, Kiyono T, and Fujita M (2006). Deregulation of Cdt1 induces chromosomal damage without rereplication and leads to chromosomal instability. J Cell Sci 119(Pt 15): 3128–3140 Tomlins SA, Rhodes DR, Perner S, Dhanasekaran SM, Mehra R, Sun XW, Varambally S, Cao X, Tchinda J, Kuefer R, Lee C, Montie JE, Shah RB, Pienta KJ, Rubin MA, and Chinnaiyan AM (2005). Recurrent fusion of TMPRSS2 and ETS transcription factor genes in prostate cancer. Science 310(5748): 644–648 Vaziri C, Saxena S, Jeon Y, Lee C, Murata K, Machida Y, Wagle N, Hwang DS, and Dutta A (2003). A p53-dependent checkpoint pathway prevents rereplication. Mol Cell 11(4): 997–1008 Williams S, Pettaway C, Song R, Papandreou C, Logothetis C, and McConkey DJ (2003). Differential effects of the proteasome inhibitor bortezomib on apoptosis and angiogenesis in human prostate tumor xenografts. Mol Cancer Ther 2(9): 835–843

Androgen-Regulated Genes in the Prostate Nigel Clegg and Peter S. Nelson

Abstract The androgen receptor (AR) and attendant signaling program regulates key components of prostate organogenesis, contributes to normal physiological functions, and influences organ-specific pathologies that include benign prostate hypertrophy and carcinoma. AR signaling regulates genetic programs in both epithelium and in cells comprising the stromal compartment of the prostate. Given that multiple cellular and tissue effects are attributable to AR signaling, increased knowledge of the AR-regulated gene expression network is central to an understanding of prostate function in health and disease. Androgen-responsive gene expression can be regulated at the level of transcription, RNA processing, RNA stability, protein translation, or protein stability. The products of these genes form part of a network of biochemical interactions leading to physiological consequences for prostate development and pathology. This review focuses on recent advances in the identification of genes regulated by androgens and the AR and provides context for their potential influence on normal prostate physiology and mechanisms of disease.

1 Introduction Signaling through the androgen receptor (AR) regulates normal prostate development and contributes to the progression of prostate cancer. Early in embryogenesis, AR signaling in the urogenital mesenchyme is required for prostate specification (Marker et al. 2003). Subsequent expression of the AR in prostate epithelial cells leads to a differentiated epithelial phenotype and the production of prostate-specific proteins. In the mature gland, androgens promote the survival and proliferation of epithelial cells; however, they also modulate a ‘‘proliferative shut-off’’ function that leads to cell quiescence (Geck et al. 1997; Isaacs et al. 1992). In prostate cancer, AR signaling is required for reactivating a proliferative state: AR-based signaling P.S. Nelson(*) Fred Hutchinson Cancer Research Center, Division of Human Biology, 1100 Fairview Ave N, MS D4-100, Seattle, WA 91809-1024, USA, E-mail: [email protected]

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4_27, # Springer Science + Business Media, LLC 2009

631

632

N. Clegg, P.S. Nelson

persists as the cancer recurs during androgen-deprivation therapy. Several mechanisms have been identified that may contribute to castration recurrence, including mutations in the AR that lead to ligand-independent activation or altered ligand specificity; AR activation by crosstalk with growth factor and cytokine signaling pathways; AR gene amplification; increased AR transactivation by AR cofactors, intracrine metabolism of adrenal androgens, or de novo androgen synthesis (Feldman and Feldman 2001; Mostaghel et al. 2007; Waghray et al. 2001). Given that multiple cellular and tissue effects are attributable to AR signaling, increased knowledge of the AR-regulated gene expression network is central to an understanding of prostate function in health and disease. The term ‘‘androgen-responsive gene’’ has been used to describe any gene with marked changes in expression that occurs following the withdrawal or addition of androgens to a tissue, organ, or cell line. This usage is extremely broad and in practice most studies report quantifiable changes in gene or protein expression within 72 h of manipulating AR ligands. Based on a review of extant gene expression data, 1.5–4.3% of genes expressed in the prostate are androgen regulated (Dehm and Tindall 2006). Androgen-responsive gene expression can be regulated at the level of transcription, RNA processing, RNA stability, protein translation, or protein stability. The products of these genes form part of a network of biochemical interactions leading to physiological consequences for prostate development and disease. A variety of high-throughput methods have been used to investigate this network, including large-scale sequencing, expression microarrays, proteomics screens, and large-scale chromatin immunoprecipitation assays (ChIP-on-chip). This review details recent advances in the identification of androgen-responsive genes and their potential physiological roles.

2 The AR and AR Target Sequences The AR is a member of the steroid subfamily of nuclear receptors. About 919 amino acids long (polymorphic triplet repeat lengths confer variation across the population), the AR is composed of an amino-terminal region, a central DNA-binding domain, and a carboxy-terminal ligand-binding domain (Agoulnik and Weigel 2006). In its inactive state, the AR is sequestered in a cytoplasmic multiprotein chaperone complex. Androgen binding induces a conformational change in the AR that leads to dissociation of the complex, phosphorylation of the receptor, and translocation to the nucleus. In the best-studied mode of androgen action, an AR homodimer binds directly to androgen-response elements (AREs) in the promoter of a target gene. DNA-independent protein–protein interactions at target genes may also regulate AR signaling (Lefstin and Yamamoto 1998; Murtha et al. 1997; Sato et al. 1997). The consensus DNA-binding motif for functional AREs comprises 15 nucleotides organized as imperfect hexamer palindromes separated by three nucleotides, AGAACAnnnTGTTCT (TRANSFAC database, 2002). However, functional AREs may differ by several nucleotides from the consensus sequence. Only the guanine residue in the 30 hexamer is invariant. In addition to consensus-like AREs, AR can bind to

Androgen-Regulated Genes in the Prostate

633

motifs that resemble direct repeats of the hexamer TGTTCT (Claessens et al. 1996; Schoenmakers et al. 1999). The ARE consensus is based on a relatively small number of androgen-regulated genes whose promoters have been mapped. Since a first step in ARE isolation often involves a directed search for the consensus motif, the frequency of other noncanonical motifs may be higher than reported. The ARE-like sequences, including those that have been shown to have functional activity, occur frequently in the human genome. The specific consensus sequence alone has 563 perfect matches (Horie-Inoue et al. 2004). The precise number of ARE-like sequences has not been determined, but there are probably over 70,000 ARE-like sequences based on calculations from the estrogen receptor, which has a similar binding motif and some sequence variability (Bourdeau et al. 2004). It is improbable that all of these sites are occupied by AR. Even where AR has been shown to bind identical ARE-like sequences, the sequences exhibit differing levels of AR occupancy depending on chromosomal location (Horie-Inoue et al. 2004). The latter observation may indicate an altered chromatin state, which is likely the case for instances of tissue-specific androgen-regulated gene expression. Functional AREs are not restricted to any particular location with respect to the start site of a gene. They may be located within the promoter region of the gene or act as enhancers at a distance. The functional AREs in the gene encoding prostatespecific antigen (PSA; also termed KLK3 for kallikrein-related peptidase 3) are located in the promoter and upstream regions of the gene (Cleutjens et al. 1996; Riegman et al. 1991). The AR gene contains exonic AREs (Dai and Burnstein 1996; Grad et al. 2001), and one gene may have multiple AREs. The sequence variability of functional AREs makes it possible that any candidate androgen-responsive gene will have a potential AR-binding motif nearby. This, along with the variable location of AREs, limits the utility of bioinformatics for identifying legitimate AREs. Instead, a wide range of biochemical tools is used to test for AR binding. In vitro methods include electrophoretic mobility shift assays, methylation-protection footprinting, dimethylsulfate footprinting, and DNase1 assays. Chromatin immunoprecipitation is used to test for ARE occupancy in vivo or in cell culture. Fragments of DNA containing the putative ARE have been used to interrogate androgen-sensitive expression of a reporter gene in cell culture to demonstrate ARE functionality (Bolton et al. 2007). The precise location and sequence flexibility of the ARE is determined using site-directed mutagenesis to abrogate androgen sensitivity. Since each of these methods has inherent limitations, two or more assays are typically employed to bolster the case for direct AR-mediated control of transcription. PSA encodes a member of the gene family encoding kallikrein-like serine proteases and is the most studied androgen-responsive gene in humans. The regulation of PSA expression provides an example of the complexity of direct transcriptional control by androgens and the AR (Cleutjens et al. 1997; Cleutjens et al. 1996; Riegman et al. 1991). AR-regulated expression of PSA is mediated through AREs in the proximal promoter (600 to +12) and in a 50 upstream enhancer region (centered around 4 kb). The AREs are numbered consecutively with the first ARE closest to the TATA box of the gene. Two AREs (AREI and AREII) are located

634

N. Clegg, P.S. Nelson

within the promoter region. AREI (AGAACAnnnAGTGCT) is a high affinity binding site, while AREII (GGATCCnnnAGTCTC) is a weak affinity site that may cooperate with other AREs to increase transcription from the promoter. The upstream enhancer is more complex with one dominant AR-binding site, AREIII (GGAACAnnnTGTATC), and multiple weak AREs that may contribute to high androgen activity. The enhancer and promoter regions act synergistically to increase PSA transcription. AR signaling also has been shown to occur in a nongenotropic fashion. In cell culture, ligand-bound AR interacts with the SH3 domain of the tyrosine kinase c-Src, which activates Src/Raf-1/ERK signaling (Kousteni et al. 2001; Migliaccio et al. 2000). Src activation is associated with a ternary protein complex composed of AR, estradiol receptor b (ERb), and Src; either antiandrogens or antiestrogens can block Src activation. MAPK signaling induces proliferation in LNCaP cells, while it plays a role in mediating the antiapoptotic roles of estrogens and androgens in osteoblasts (Kousteni et al. 2001; Migliaccio et al. 2000). The extent and clinical relevance of nongenotropic AR signaling remains unclear, but these experimental findings add an additional level of control and complexity to the androgen-response network, which have implications for pragmatic study of direct AR signaling. Until recently, early transcriptional activation or repression of an androgen-stimulated gene coupled with ARE-like motifs in the promoter has been used as reasonable, suggestive evidence for direct, AR-mediated transcription. This assumption may not be justified.

3 Methods and Approaches for the Identification of Androgen-Responsive Genes Numerous methods have been employed for the genome-wide discovery and characterization of androgen-responsive genes. An introduction to the most widely utilized approaches is provided followed by sections detailing the results of applying these technologies.

3.1

Digital Sequencing Approaches: Expressed Sequence Tags (ESTs) and Serial Analysis of Gene Expression (SAGE)

An expressed sequence tag (EST) is a short DNA sequence (200–800 bp in length) that identifies a cDNA (Adams et al. 1991). A cDNA is a chemically synthesized DNA copy of an mRNA transcript from a single gene. When a representative cDNA library is constructed from cells, the number of cDNAs within it is proportional to the number of stable transcripts synthesized by each gene. The relative abundance of each transcript type in the original tissue can be estimated by sequencing short stretches from thousands of cDNA library clones (ESTs) (Nagaraj et al. 2007). The Serial Analysis of Gene Expression (SAGE) method also counts the number of tags

Androgen-Regulated Genes in the Prostate

635

representing a cDNA/transcript, but the tags are much shorter (generally 10–27 bp in length) (Porter et al. 2006; Velculescu et al. 1995). Instead of sequencing individually amplified cDNAs, a pool of cDNAs is digested with a restriction enzyme, and short fragments representing the 30 ends of the original cDNAs are ligated. The ligated DNA is amplified, sequenced and the number of tags counted. The EST method is expensive and time consuming, but long sequence tags permit unambiguous assignment of cDNAs to a specific gene. The SAGE method is less expensive with higher throughput, which allows much deeper sampling (by about an order of magnitude in most published studies), but many of the short tags cannot be assigned unambiguously to a single gene.

3.2

Microarray Expression Profiling

Microarrays are composed of a series of single- or double-stranded DNA molecules spotted or synthesized in situ on a ‘‘chip’’ made from a glass slide, a silicon wafer, or a nylon membrane (Ness 2007; Schena et al. 1995; Trevino et al. 2007). The spots, arrayed in a stereotyped grid pattern, are used as hybridization targets for cDNA or amplified RNA synthesized in vitro using RNA extracted from a source tissue. For spotted cDNA microarrays, two samples are prepared: one is generally the test sample and the other is a direct comparator (e.g., androgen-stimulated versus androgen-starved LNCaP cells) or a reference control. The samples are labeled differentially with fluorophores that emit different light wavelengths and are combined and hybridized to the DNA chip. After stringent washing, only complementary sequences remain bound to the chip. The ratio of the intensities of the two fluorophores in any spot (and its associated gene) represents the difference in levels of transcript expression in the test and control sequences. Oligonucleotide microarrays have single-stranded oligonucleotides affixed to the chip. The oligonucleotides can be short (25 nucleotides) or long (70 nucleotides). Long oligonucleotide arrays can be analyzed much like spotted cDNA microarrays, while short oligonucleotide arrays rely on a series of internal controls to infer ‘‘absolute’’ levels of gene expression (http://www.affymetrix.com).

3.3

ChIP-on-Chip

DNA microarray technology can be used to perform thousands of simultaneous chromatin immunoprecipitation (ChIP) experiments (Wu et al. 2006). ChIP is used to identify sites of interaction between genomic DNA and associated proteins, such as transcription factors, repressors, or DNA-modifying enzymes. Briefly, DNAassociated proteins are chemically crosslinked to genomic DNA, the chromatin is isolated, and the DNA is sheared into small fragments. Subsequently, an antibody to the DNA-associated protein is used to capture the target chromatin complex from bulk chromatin by immunoprecipitation. The chromatin crosslinking is reversed

636

N. Clegg, P.S. Nelson

and the protein is digested, leaving fragments of DNA bound to the target protein. The DNA is amplified using PCR and compared to a similar amount of amplified control DNA to determine an enrichment ratio for the DNA–protein interaction. In ChIP-on-chip experiments, part or all of the genome is tiled on an oligonucleotide microarray. After differentially labeling and hybridizing the immunoprecipitated target and control DNA samples (as for spotted cDNA expression arrays), sheared DNA fragments that bind to arrayed genomic DNA are identified and quantitated by having a high signal-to-noise ratio.

3.4

Isotope-Coded Affinity Tags (ICAT)

The ICAT method uses a reagent that binds to the cysteinyl residues of proteins (Gygi et al. 1999). The reagent contains an affinity tag (biotin) and is synthesized with either a heavy or a light isotope of hydrogen. A test sample and a control sample are each labeled with one of the isotope-bearing ICAT reagents. The samples are combined and enzymatically cleaved to create peptide fragments. Peptide fragments that contain an ICAT tag are purified using avidin affinity chromatography, and the isolated fragments are separated using microcapillary liquid chromatography followed by electrospray ionization MS/MS. Computer software is used to deduce the protein from which a peptide fragment was derived. The ICAT method is technically demanding but is able to quantitatively analyze a large proportion of the proteins in a sample.

3.5

Fluorescence Two-Dimensional Differential Gel Electrophoresis (2-D DIGE)

In 2-D gel electrophoresis, proteins are separated in a gel based on their isoelectric point and separated by mass in a direction 90 from the first separation. In 2-D DIGE, two samples are labeled, each one with a different fluorophore, pooled, and separated on a 2-D gel (Tonge et al. 2001). The mass and the charge of the two fluorophores are matched to permit alignment of the fluorescent images. The spots are quantified, normalized, and compared for intensity (abundance) differences.

4 Model Systems for the Identification of Androgen-Regulated Genes 4.1

Human Studies

It is possible, but difficult, to study androgen-responsive gene expression in human subjects. Antiandrogens can be used to block hormone action followed by tissue biopsy or prostatectomy. However, relatively long treatment times confound the

Androgen-Regulated Genes in the Prostate

637

results because many gene expression changes occur that are temporally distant from the effects of direct androgen regulation. Although the biochemical action of several antiandrogens has been determined, they may have additional effects at the level of gene expression that are not related to AR blockade. The majority of high-throughput methods used to query androgen-regulated gene expression in the human prostate rely on cell culture to obtain sufficient material. LNCaP cells, derived from a prostate cancer metastasis, are used frequently because they are epithelial in origin, express the AR, and are androgen sensitive for growth and survival both in culture and as xenografts (Horoszewicz et al. 1983). Like many human prostate cancers, androgen deprivation temporarily affects LNCaP cell survival, but ultimately a subset of cells progress to an androgeninsensitive phenotype. One major concern with LNCaP cells as a model system is the T877A mutation in the AR, which confers responses to other steroid hormones, such as glucocorticoids and progestins, as well as some AR antagonists, such as flutamide (Sun et al. 2006). Several other cell lines have been used to study androgen-responsive genes in the prostate. For example, LAPC-4 cells were derived from a cancer metastasis and express the wild-type AR (Klein et al. 1997). Cancer cell lines that do not express AR, such as DU145, have been transfected with AR expression constructs. Cell lines derived from immortalized benign epithelial cells are another model. Finally, androgen-independent (AI) cell lines have been used to study androgen-sensitive gene expression. While these cells grow in the absence of androgens, a component of their genome may still respond to androgen stimulation. An unusual feature of androgen-sensitive cell lines that is important for evaluating AR-regulated gene expression, especially AR-regulated genes implicated in cell proliferation, is a biphasic response to androgens. In LNCaP and LAPC-4 cells proliferation is stimulated around 0.1 nM DHT and can be repressed by concentrations exceeding 1 nM (de Launoit et al. 1991; DePrimo et al. 2002). Many experiments evaluating AR-regulated transcriptional activity are performed in the 1–10 nM range where cell proliferation is low. Whether this response is meaningful physiologically in normal cells is unclear, but it may have implications for prostate treatment modalities that modify patients’ androgen levels. A common concern with all cell lines is that the process of immortalization, transformation, and adaptation to tissue culture alters patterns of gene expression. Additional changes may occur over time as cells are passaged. Also, most cell lines employed to study androgen-responsive genes in the prostate were derived from metastatic cancers. Metastatic cancers are often aneuploid with altered patterns of gene expression. Thus, while cell lines may represent one state of the androgenresponse network, that state may not represent the situation in normal tissue. Researchers have used other model systems to support conclusions derived from in vitro experiments. An in vivo model used for human studies of androgenresponse genes is the primary prostate cancer xenograft CWR22 (Amler et al. 2000). CWR22 is serially transplantable, and upon castration of the host mouse, CWR22 tumors regress and form a nonproliferative mass. This phenotype is reversible with readministration of androgens. As in human prostate cancer, some

638

N. Clegg, P.S. Nelson

CWR22 cells survive and grow under the pressure of androgen deprivation. These prostate cancer xenografts are subject to a more natural ambient environment than in tissue culture – albeit one from a different species – hence they may behave more like typical human cancer. The LuCaP35 xenograft also has similarly been employed to evaluate androgen-responsive gene expression (Corey et al. 2003).

4.2

Rodent Models

Rat and mouse strains have been used as in vivo models to probe the prostate androgen-response network. Rats have relatively large prostates, ideal for experimental manipulation, and a long history as pharmacological models. They develop prostate cancer spontaneously, albeit at low frequency. Though substantially smaller in size with correspondingly smaller prostate glands, mice offer the power of genetic manipulation. A number of genetically engineered models for prostate cancer have been developed in the mouse that are responsive to androgens (Kasper 2005). Rats and mice have prostates of similar structure composed of anatomically distinct lobes, the anterior, ventral, dorsal, and lateral lobes (Cunha et al. 2004; Marker et al. 2003). Each lobe has a distinct morphology and expresses different levels of a variety of gene transcripts (Abbott et al. 2003). This structure differs considerably from the human condition where the prostate is an encapsulated organ divided into three zones, the central zone, the transition zone, and the peripheral zone (McNeal 1981a, b). Several studies have used gene expression profiling methods to evaluate molecular differences among prostate zones. Microarray expression profiling identified 346 genes that were expressed preferentially in the transition or peripheral zone, but protein expression profiling identified fewer differences (Lexander et al. 2005; van der Heul-Nieuwenhuijsen et al. 2006) As in humans, androgens are critical for rodent prostate cell differentiation and proliferation, and overall prostate development. The ventral prostate is studied most commonly because the glandular epithelial compartment undergoes involution in response to androgen deprivation. Readministration of androgens stimulates proliferation and restores prostate size and function (Evans and Chandler 1987). However, replacement of androgens following castration only restores the prostate to precastrate size and structure. Continued exposure to androgen does not produce hyperproliferation or hyperplasia; the mechanisms responsible for the precise regulation of organogenesis have not been defined. Despite differences between rodents and humans, most assume that both species share a variety of fundamental biochemical and functional pathways that are regulated by AR signaling. Numerous studies support this assumption, although specific examples of androgen-regulated gene expression will be addressed in each species. One particular caveat to the interpretation of AR signaling in rodent (or human) tissues is that prostate tissue samples frequently contain multiple cell types. Each cell type may exhibit different AR-regulated gene expression programs, which can impede identification of cell type-specific androgen-responsive genes and lead to errant connections between AR-responsive networks and functional pathways.

Androgen-Regulated Genes in the Prostate

639

5 Androgen-Regulated Gene Expression Determined Using Analyses of Transcripts 5.1

Sequence Tag-Based Studies

Waghray et al. (2001) used SAGE to survey androgen-responsive genes in LNCaP cells 24 h after treatment with dihydrotestosterone (DHT) (Table 1). Sequence analysis identified tags from approximately 16,500 genes. A total of 351 genes were found androgen responsive (p < 0.05): 147 were stimulated by androgen and 204 were repressed by androgens. Genes, such as PSA and NKX3.1, and b-microseminoprotein were found up-regulated in response to androgens. A digital analysis of the androgen-response program using ESTs was performed in LNCaP cells (Clegg et al. 2002). cDNA libraries were constructed in parallel for androgen-stimulated and androgen-starved cells. A total of 4,400 ESTs, which represented 2,486 distinct transcripts, were sequenced. Statistical analysis of ESTs from the two libraries identified 17 genes with a high probability (p > 0.90) of being androgen regulated. The androgen-responsive transcription of several highly expressed genes, such as PSA and FKBP5, was confirmed using Northern blot analysis, and several novel androgen-regulated transcripts were identified. Numerous studies have shown that FKBP5 expression is induced robustly by androgen in the prostate (Amler et al. 2000; DePrimo et al. 2002; Mousses et al. 2001). FKBP5 is an immunophilin with peptidyl-prolyl isomerase activity; it is part of a multiprotein chaperone complex that associates with glucocorticoid (GR) and progesterone (PR) receptors prior to ligand binding (Barent et al. 1998; Davies et al. 2002). Febbo et al. (2005) showed that FKBP5-specific antibodies co-immunoprecipitate AR, which suggests physical interaction similar to that of GR and PR. LNCaP cells overexpressing an FKBP5 gene-construct consistently displayed higher activation of an androgen-responsive MMTV-luciferase gene than cells expressing the same gene construct with wild-type levels of FKBP5. Overexpression of FKBP5 caused an increase in PSA secretion relative to normal levels of FKBP5, which implies that increased expression of FKBP5 increases androgen-mediated transcription. These data support the existence of a feedforward regulatory loop in the androgen-response network. Experiments involving the mouse ventral prostate support a direct role for the AR in regulating Fkbp5 expression, although no evidence exists that human FKBP5 is a direct AR target gene. Magee et al. (2006) determined that mouse Fkbp5 is an androgen-responsive gene, and identified 7 potential AREs in and around the gene using a bioinformatics approach. In ChIP assays, only chromatin encompassing the putative AREs 6 and 7 (located 65 kb downstream of the transcription start site) could be immunoprecipitated by anti-AR antibodies. ARE-7 was capable of conferring androgen-responsive transcription to a reporter gene transfected into LAPC-4 cells. A statistical difference in H3 acetylation between castrated and uncastrated mice involving chromatin surrounding ARE-7 was identified. These data suggest a

640

N. Clegg, P.S. Nelson

Table 1 Microarray expression profiling of androgen-regulated genes in humans and rodents Study Tissue Agent Regulated Platform genesa Amler et al. (2000) CWR22, AI Castration 160 cDNA array subline Xu et al. (2001) LNCaP 10 nM R1881 351 SAGE Clegg et al. (2002) LNCaP 1 nM R1881 17 EST sequencing DePrimo et al. LNCaP, 1–1,000 nM 517 cDNA array (2002) MDAPCa, R1881 LAPC4 Nelson et al. (2002) LNCaP 1 nM R1881 146 cDNA array Segawa et al. (2002) LNCaP 0.1–10 nM 98 Oligonucleotide R1881 array Eder et al. (2003) LNCaP Antisense AR; 118 cDNA array bicalutamide Shi et al. (2004) LNCaP, AI 0.01–100 nM 32 Oligonucleotide subline R1881 array Velasco et al. (2004) LNCaP 10 nM DHT 692 Oligonucleotide array Febbo et al. (2005) LNCaP, 0.1 nM R1881 >50 Oligonucleotide CWRI4, array CWR22 Oosterhoff (2005) LNCaP, AI 0.1 nM R1881 >1,000 cDNA array subline Coutinho-Camillo LNCaP 1 nM DHT; 35 Differential display et al. (2006) bicalutamide Bebermeier et al. LNCaP 0.001–10 nM 382 cDNA array (2007) R1881 Bolton et al. (2007) HPr-1 1 nM R1881 >200 cDNA microarray Kim et al. (2007) RC-165N/ 1–100 nM DHT 239 Oligonucleotide hTERT array Li et al. (2007) LNCaP AR shRNA; 302 cDNA microarray DHT Louro et al. (2007) LNCaP 1 nM R1881 39 cDNA microarray Nickols and Dervan LNCaP 1 nM DHT 287 Oligonucleotide (2007) array Waghray et al. LNCaP 1 nM DHT 351 SAGE, 2-DGE/ (2001) MALDI-TOF MS Wright et al. (2003) LNCaP 10 nM R1881 350 ICAT LC MS/MS Martin et al. (2004) LNCaP 1 nM R188I 50 ICAT LC MS/MS Meehan and Sadar LNCaP 10 nM R1881 77 ICAT 2-D LC MS/ (2004) MS Rowland et al. LNCaP 10 nM R1881 107 2-D DIGE (2007) Whitaker et al. LNCaP 10 nM R1881 5 2-D PAGE/MALDI(2007) TOF MS Pang et al. (2002) Rat VP Castration; 230 cDNA microarray testosterone (continued )

Androgen-Regulated Genes in the Prostate Table 1 (continued) Study

641

Tissue

Agent

Jiang and Wang (2003)

Rat VP, DLP

Castration; testosterone

Desai et al. (2004)

Rat VP, DLP

Nantermet et al. (2004) Pfundt et al. (2005)

Rat VP

Castration; testosterone Castration; 5alpha-DHT Castration

234

10 nM R1881

66

Asirvatham et al. (2006) Ci et al. (2008) Wang et al. (2007)

Normal rat, Dunning tumors Rat VP epithelial cells Mouse prostate Mouse prostate

Regulated genesa 300

1,496,256

422

Platform Microarray/ subtractive hybridization Oligonucleotide array Oligonucleotide array Oligonucleotide array Oligonucleotide array

Castration; DHT 114 SAGE Castration; 64 Oligonucleotide testosterone array a Androgen-responsive genes reported according to the study authors’ emphasis or statistical criteria

regulatory role for the ARE-7 region since hyperacetylation occurs at the Fkbp5 proximal promoter but nowhere else in the gene. Since the same ARE is embedded in a highly conserved region of the human FKBP5 gene, it may also represent a direct AR-binding site.

5.2

AR Transcriptional Network Defined by Microarray Analyses

Numerous microarray experiments have investigated the AR transcriptional network, primarily in LNCaP cells (see Table 1 for study references and summaries of results). In many of these experiments, RNA from androgen-stimulated and androgenstarved cells was hybridized to spotted cDNA or oligonucleotide microarrays. Genes with high differential expression (activated or repressed) were grouped based on known biological functions. Additional studies have used microarrrays to investigate the downstream effects of known androgen-responsive genes. These surveys have led to a number of insights into differentiated prostate cell function, proliferation, and cell survival.

5.2.1

AR-Regulated Functional Categories: General Processes and Metabolic Pathways

The prostate is an exocrine organ that manufactures 30–50% of the volume of the seminal fluid (DePrimo et al. 2002). This material includes proteins, organic

642

N. Clegg, P.S. Nelson

solutes, lipids, and cholesterol. A high proportion of genes induced by androgens appear to play a role in this specialized function. Among the most highly expressed androgen-responsive genes are PSA, KLK2, and prostatic alkaline phosphatase, all of which are secreted. These genes are enriched in the prostate and are usually considered ‘‘tissue-specific’’ or ‘‘tissue enriched’’ genes. Both PSA and KLK2 are regulated by androgens at the level of transcription (Mitchell et al. 2000; Riegman et al. 1991). The volume of material secreted by the prostate requires supporting cellular machinery. DePrimo et al. (2002) analyzed the response of LNCaP cells to R1881 over the course of 72 h and found that almost 40% of androgen-regulated genes (approximately 200/517) contributed to the modification of secretory proteins. An additional 30 androgen-responsive genes had roles in protein trafficking, secretory vesicle formation, or transport of secretory vesicles. Androgen treatment also induced proteins involved in protein folding (e.g., FKBP5) and glycosylation (e.g., STCH). Prostatic fluid also is rich in the polyamines choline, spermine, and spermidine. These molecules have numerous functions, which include stress-protective activity (Rhee et al. 2007). Five genes encoding enzymes that regulate polyamine synthesis (ODC1, AMD1, SMS, SRM, and SAT) are induced coordinately by R1881, which demonstrates that androgens are key regulators of this pathway (Segawa et al. 2002). Spermine synthase (SMS) protein is also androgen regulated (Meehan and Sadar 2004). Androgen regulation of elements of the polyamine synthesis pathway has been observed in rat ventral prostate and is not just a peculiarity of in vitro cell culture (Nantermet et al. 2004). In accord with previous findings (Heemers et al. 2006), expression microarrays demonstrated that androgens regulate genes involved in lipid and fatty acid biosynthesis. Genes coordinately up-regulated by treatment with 0.1 or 10 nM R1881 include fatty acid synthase (FASN), fatty acid amide hydrolase (FAAH), lowdensity lipoprotein receptor (LDLR), 3-hydroxy-3-methylglutaryl-Coenzyme A reductase (HMGCR), and farnesyl-diphosphate farnesyltransferase (FDFT1) (Segawa et al. 2002). Elongation of long fatty acids 5(ELOVL5/HELO1), acylCoA synthetase long-chain family member 3 (ACSL3), and several others are implicated in the same pathways (DePrimo et al. 2002; Nelson et al. 2002). Inhibition of AR expression with an antisense AR oligonucleotide down-regulated a subset of lipogenic genes (Eder et al. 2003). Expression profiling in the rat prostate also implicates androgens in these processes (Pang et al. 2002). Testosterone regulates genes involved in lipid and fatty acid biosynthesis through an indirect mechanism involving the sterol response element binding transcription factor 1 (SREBF1; also known as SREBP) (Heemers et al. 2006). SREBF1 function is regulated by SCAP, a protein with a sterol-sensing domain. In response to changes in cellular sterol levels, SCAP binds to SREBP proteins to mediate their transport from the endoplasmic reticulum to the Golgi where they are cleaved and activated. SCAP is regulated by androgen and may be the key effector in steroid regulation of lipid and fatty acid biosynthesis. Androgen regulation of lipogenic gene expression likely occurs in vivo, although its physiological function is uncertain (Heemers et al. 2006).

Androgen-Regulated Genes in the Prostate

643

A recent microarray screen for androgen-responsive genes using samples from patients 3 days after medical castration provides preliminary evidence that AR regulates CYP3A5, a gene implicated in androgen biosynthesis (Moilanen et al. 2007). CYP3A5 catalyzes the 6b-hydroxylation of testosterone, which produces a less biologically active substrate. A 4.5-fold increase in CYP3A5 mRNA was detected after 48 h, with no obvious stimulation prior to 24 h. Transient transfection of a reporter construct containing deletions of the CYP3A5 promoter mapped an androgen-responsive region to the site of a computer program-selected ARE. Mutation of the putative ARE relieved androgen-regulated control of a promoter construct, which suggested that the ARE is functional in vitro. The CYP3A5 ARE also bound AR in a supershift assay. Taken together these data suggest that CYP3A5 is androgen regulated and participates in a negative feedback loop in which androgens restrict their own effects by inducing testosterone inhibition (Moilanen et al. 2007). While this hypothesis is intriguing, CYP3A5 catalyzes a variety of reactions involving drug substrates, lipids, and cholesterol, in addition to steroids, and thus CYP3A5 may play a more general role in prostate metabolism (de Wildt et al. 1999). Androgen-induced expression of the endoplasmic reticulum (ER) stress response genes has been observed in androgen-sensitive LNCaP cells (Segawa et al. 2002). The ER stress response is triggered by the accumulation of unfolded or misfolded proteins in the ER, a process termed the unfolded protein response (UPR). One element of the stress response is the transcriptional induction of genes encoding ERresident molecular chaperones and folding enzymes. Segawa et al. observed many androgen-responsive chaperones and folding enzymes consistent with the UPR, including NDRG1, HERPUD1, PDIR, and ORP150. Noting that androgens induce prostate cancer in rat models, Segawa et al. hypothesized that the stress response induced by androgens may help prostate cells alleviate the adverse effects of androgen. It remains to be seen whether this stress response functions in normal cells.

5.2.2

AR-Regulated Functional Categories: IGF Signaling

Several androgen-responsive genes belong to the IGF signaling pathway. Insulinlike growth factor-1 (IGF-I) binds to the IGF-I receptor (IGF-IR), to initiate prosurvival and mitogenic signaling, in part through the phospatidylinositol-3kinase (PI3K) pathway (Kulik and Weber 1998). The IGF-I gene is regulated at the level of transcription by androgens (Wu et al. 2007). Two AREs, approximately 25 bp apart, were found 0.14 kb upstream of the transcription start site. A reporter assay showed that the two AREs act additively in response to androgen. Oligonucleotides matching both AREs bound androgen in an electrophoretic mobility shift assay, and chromatin precipitation from LNCaP cells using anti-AR precipitated a fragment containing both AREs. Thus, androgens play a significant role in regulating IGF-I signaling. Insulin-like growth-factor-binding proteins (IGFBPs) alter IGF-I (and IGF-2) signaling by inhibiting the interaction of IGF-I and IGF-IR (Firth and Baxter 2002).

644

N. Clegg, P.S. Nelson

IGFBP3, which is modulator of cell growth and apoptosis via IGF-dependent and– independent pathways, was identified as a down-regulated, androgen-responsive gene in microarray expression studies of LNCaP and androgen-independent C4–2 cells (Kojima et al. 2006). While several other groups have reported the androgenresponsive nature of IGFBP3 in LNCaP cells, there is no consensus on whether it is up-regulated or down-regulated (Arnold et al. 2005; Goossens et al. 1999; Martin and Pattison 2000; Peng et al. 2006). This may reflect different treatment of the cells or alterations caused by cell passaging; and it underscores the need for caution in interpreting in vitro experiments. Contradictory results also have been noted in the rodent prostate. Rat IGFBP3 is induced by androgens in the rat prostate cell line NRP-152, but in vivo it is up-regulated following androgen deprivation and down-regulated upon androgen replacement (Desai et al. 2004; Jiang and Wang 2003; Nickerson et al. 1998; Wang et al. 2007b). Peng et al. (2006) found that growth inhibitory concentrations of androgens upregulated IGFBP-3 expression via an ARE in LNCaP cells. The ARE was mapped using promoter deletions. A single fragment was shown capable of conferring androgen sensitivity in a reporter assay. A point mutation in an ARE-like sequence about 2.8 kb 50 to the start of transcription abrogated the androgen response.

5.2.3

AR-Regulated Functional Categories: Transcription Factors

A variety of transcription factors which include MAF, CEBPD, NRF1, TAF3B2, THRA, PDEF, ETV1, and NKX3.1 were identified using microarray expression profiling for androgen-responsive genes. These transcription factors control a specific subset of androgen-responsive genes, which form key nodes in the ARresponse network. Most have not been studied extensively in the context of prostate development or cancer. Those that have are multifunctional, with roles in differentiation, proliferation, and basic cellular processes. The best-characterized androgen-responsive transcription factor is NKX3.1, a homeobox gene located on chromosome 8p21. Loss of a single copy of NKX3.1 is observed in a high proportion of castration-recurrent and metastatic prostate cancers (Bova et al. 1993; He et al. 1997; Macoska et al. 1995). Disruption of Nkx3.1 in mice causes epithelial hyperplasia and prostatic intraepithelial neoplasia (Abdulkadir et al. 2002; Bhatia-Gaur et al. 1999), and contributes to prostate carcinogenesis (Abate-Shen et al. 2003; Kim et al. 2002) Microarray analysis of Nkx3.1þ/þ and Nkx3.1þ/ mice identified diverse targets of Nkx3.1 action (Magee et al. 2003). Surprisingly, some target genes were highly sensitive to the loss of one copy of the gene, while others were not. A model has been proposed whereby Nkx3.1 haploinsufficiency results in the stochastic inactivation of target genes in the prostate, some of which contribute to cancer progression. It is not yet known which genes regulated by NKX3.1 contribute to neoplastic phenotypes. An AR-binding site has been reported near the promoter of NKX3.1; hence, it may be directly regulated by androgens (Masuda et al. 2005).

Androgen-Regulated Genes in the Prostate

645

Another transcription factor that is likely to be an indirect target of the AR is cAMP-responsive element-binding protein 3-like 4 (CREB3L4). CREB3L4 was originally isolated from a library of cDNA clones enriched for androgen-response genes (Qi et al. 2002). The precise function is unknown, but CREB3L4 is sequestered in the ER and processed into an active form in the Golgi in response to an intracellular signal, perhaps calcium depletion. The mode of activation is similar to SREBP and ATF6 which regulate part of the ER stress response. Microarray-based gene expression profiling has been used to investigate targets of CREB3L4, thereby extending the androgen-responsive network (Ben Aicha et al. 2007). Conditional overexpression of CREB3L4 in LNCaP cells revealed several genes also observed in the androgen stress response. In addition, TMPRSS2, a gene that is transcriptionally activated by AR, was up-regulated by CREB3L4. While it is tempting to speculate that CREB3L4’s primary role is in the ER stress response, CREB3L4 expression profiling suggests roles in many other processes. These data demonstrate the complexity and density of interactions ascribed to gene products in the androgen-response network. Ets variant gene 1 (ETV1) is a direct target of AR (Cai et al. 2007b). ETV1 is a member of the ETS family of transcription factors, which are involved in multiple processes that include cell proliferation and cancer cell invasion (Hsu et al. 2004). ETV1 was identified as a DHT-up-regulated gene in LNCaP cells, whose expression can be repressed by the antiandrogen bicalutamide. An ARE was detected approximately 0.15 kb upstream of exon 1. A 1-kb DNA fragment containing the ARE increased the promoter activity of a reporter gene in response to DHT, and ChIP assay showed that AR binds to the region of the ARE. ETV1 mRNA and protein levels were not responsive to DHT stimulation in androgen-independent cell lines derived from LNCaP cells, which suggested that AR ligand was no longer necessary for ETV1 expression. These findings may be relevant to castrationrecurrent prostate cancer. Finally, when LNCaP cells were transfected with an siRNA targeting ETV1, suppression of ETV1 protein expression was associated with attenuated cell invasion in an in vitro assay. ETV1 may have an important role in prostate cancer metastasis. ETV1 regulates the expression of several MMPs in the prostate, which may explain these results. However, at least one MMP, MMP2, is also a direct target of AR (Li et al. 2007). ETV1 also affected LNCaP cell proliferation (Cai et al. 2007b). ETV1 expression occurs in prostate cancer cell lines and malignant prostate cancer, but not in primary prostate epithelial cells, which indicates that the AR network is influenced depending on cell differentiation state.

5.2.4

AR-Regulated Functional Categories: Regulators of Cell Proliferation

Many experiments used to identify androgen-regulated genes are performed in cell culture using 1–10 nM concentration ranges of the synthetic androgen, R1881. Cell proliferation is low at these concentrations. Nevertheless, many genes that affect proliferation have been identified. One possible explanation is that these genes also

646

N. Clegg, P.S. Nelson

participate in other cellular processes. Even so, some cell cycle genes have been identified, such as cell division cycle 14B (Cdc14B), a gene that is essential for cell cycle progression in yeast (Nelson et al. 2002); cyclin A2 (CCNA2) (DePrimo et al. 2002). Nine genes involved in cell cycle or mitosis and meiosis were identified after castration in CWR22 xenografts (Amler et al. 2000). Whether any of these genes are regulated directly by androgens or are simply a consequence of activating the cell cycle machinery remains unknown. The protein encoded by soluble guanylyl cyclase a1 (sGCa1) affects proliferation and is regulated by androgens (Cai et al. 2007a). sGCa1 is a subunit of the nitric oxide receptor, soluble guanylyl cyclase (sGC). Nitric oxide binds to sGC, which activates the enzyme to catalyze the formation of cGMP (30 ,50 -cyclic guanosine monophosphate). cGMP subsequently activates a variety of cell signaling proteins. The a1 subunit of sGC was up-regulated in a microarray expression screen for DHT-responsive genes in LNCaP cells compared to PC-3 cells transfected with AR as a negative control (because their proliferation is repressed by DHT). Subsequently, sGCa1 was shown androgen responsive in LNCaP cells, both at the level of protein concentration and function (Cai et al. 2007a). A near-consensus ARE was found approximately 0.7 kb 50 of the first exon of sGCa1,and a DNA fragment containing the potential ARE was capable of inducing reporter gene expression in response to stimulation by androgens. An siRNA-directed against AR nearly eliminated the androgen induction of the sGCa1 promoter construct. AR was recruited to the sGCa1 promoter in a chromatin precipitation assay. Taken together, these results support the idea that sGCa1 is regulated by AR at the level of transcription. A role for sGCa1 in proliferation was demonstrated in both androgen-sensitive and androgen-independent cell lines (Cai et al. 2007a). An siRNA made to sGCa1 decreased androgen-induced LNCaP expression. In an androgen-independent cell line derived from LNCaP cells (C81), sGCa1 mRNA and protein levels were high and unaffected by the presence of DHT. Nevertheless, siRNA-mediated repression of sGCa1 reduced cell proliferation. Adenoviral-mediated overexpression of sGCa1 in LNCaP cells enhanced cell proliferation in LNCaP cells without DHT to a level comparable to that of DHT alone. Since sGCa1 is an androgen-regulated gene that loses androgen regulation in androgen-independent cell lines, it may be important in the growth of castration-recurrent prostate cancer. The GREB1 gene is another factor that affects prostate cell proliferation and that is regulated by androgens (Rae et al. 2006). GREB1 was identified in a microarray study searching for genes implicated in estrogen-stimulated proliferation in breast cancer and was tested for AR responsiveness because of potential steroid receptorbinding sites in its promoter region. GREB1 is stimulated by androgens (R1881, DHT, and testosterone) in a dose-dependent manner in the LNCaP and VCaP cell lines. ChIP using anti-AR antibodies identified a promoter fragment in LNCaP cells, which suggests that AR directly mediates GREB1 transcription. An siRNA against GREB1 reduced the gene’s mRNA levels and suppressed cell growth to basal levels, which demonstrates a role for GREB1 in cell proliferation.

Androgen-Regulated Genes in the Prostate

5.2.5

647

AR-Regulated Functional Categories: NonCoding RNAs

Another level of the androgen-responsive gene network involves noncoding RNAs (ncRNAs), which may in turn suppress one or more genes with diverse function (Fig. 1). MiR-125b, a member of the micro-RNA class of ncRNAs, was found regulated by AR (Shi et al. 2007). Levels of miR-125b transcripts are increased in cell lines treated with androgens, and two promoter-proximal sites bind AR. Increased expression of miR-125b (using a synthetic construct, miR125bm) in the absence of androgens stimulated cell growth in both androgen-sensitive LNCaP cells and androgen-independent cds1 cells. Anti-miR-125b was tested in cds1 cells and inhibited cell growth and promoted the accumulation of sub-G1 phase cells, which suggests that anti-miR-12b promotes apoptosis. Expression microarrays were used to identify potential targets of miR-125b. LNCaP cells were cultured in androgen-depleted medium for 2 days, and transfected with miR-125bm or a micro-RNA control. Sixty target genes were identified based on a twofold difference in expression: 13 were up-regulated and 47 were down-regulated. One of the down-regulated genes, BAK1, has a 30 UTR that is responsive to miR-125bm in a

Fig. 1 Regulation levels of the androgen responsive gene and protein network. The regulation of gene and protein expression by androgenic ligands acting through AR can occur at multiple levels that result in a cascade of interactions. Depicted are several documented examples that include (a) the direct transcriptional regulation of a structural or secreted protein (e.g., prostate-specific antigen); (b) the regulation of a transcription factor, that in turn, regulates another group of genes, some of which could also be transcription factors (e.g., NKX3.1); (c) the regulation of noncoding RNAs such as microRNAs that modulate the expression of other genes/proteins (e.g., miR-125b); and (d) the regulation of a protease or protease inhibitor, that in turn regulates the activation, shedding, or release of other target proteins (e.g., Spint1). Together, this AR-regulated gene network is responsible for directing a diverse range of cellular events and phenotypes

648

N. Clegg, P.S. Nelson

reporter-gene assay. Transfection of miR-125bm into LNCaP cells reduced the levels of both BAK1 mRNA and protein. BAK1 belongs to the BCL2 family of proteins and functions as a proapoptotic regulator. However, knockdown of BAK1 expression alone (by siRNA) is not sufficient to induce increased cell proliferation in LNCaP and cds1 cells, which suggests that other RNAs regulated by miR-125b also participate in androgen-independent growth. Nevertheless, a molecular link to the control of prostate cell proliferation and apoptosis has been established via an androgen-regulated microRNA. These studies open a new tier of indirect, androgen-regulated control at the level of translation (Fig. 1). Louro et al. (2007) used a spotted cDNA array enriched in intronic transcripts to find androgen-responsive noncoding genes in LNCaP cells exposed to R1881 for 6–48 h. Thirty-nine androgen-responsive noncoding RNAs (ncRNAs) were detected. From 13 ncRNAs selected for further characterization, 10 antisense RNAs were detected in the introns of previously described genes. The transcripts were long (0.5–5 kb), unspliced, and without apparent coding regions. A potential ARE was located approximately 0.8 kb upstream of the location where MYO5A binds to the chromosome. ChIP using anti-AR showed a twofold to threefold increase in AR binding to this region, which suggests direct AR regulation. The function of the 39 ncRNAs is unknown, but they may help regulate transcription of the gene on the opposite strand, since the levels of some intronic transcripts correlated either with the levels of transcription of the protein coding gene on the opposite strand or with the alternate usage of exons in the associated coding gene. 5.2.6

AR-Regulated Functional Categories: Spatially Restricted Genes

The rat ventral prostate (VP) responds to androgen deprivation with extensive apoptosis, unlike the human prostate or other rat prostate lobes (Kyprianou and Isaacs 1988). Desai et al. (2004) compared the response to castration in the VP and dorsolateral prostate (DLP) and identified several androgen-repressed genes that were associated previously with apoptosis, which include IGBP-3, CCAATenhancer binding protein-d (C/EBPd), and phosphatase and tensin homolog (PTEN). Overexpression of PTEN induces apoptosis in LNCaP cells (Yuan and Whang 2002); PTEN may play a significant role in apoptosis in the VP. In support of this idea, a nuclear localization of PTEN protein was observed exclusively in the VP of castrate animals. PTEN was induced 3 days postcastration, so it is unlikely to be a direct target of AR. Another response to androgen deprivation in the rat VP is the stimulation of genes involved in inflammatory and immune response pathways. Androgen deprivation in vivo induces IL-15 and IL-18, probably due to an influx of T cells, mast cells, and macrophages (Desai et al. 2004). In cell culture, normal rat epithelial cells stimulated with R1881 showed increased expression of at least six genes associated with the IFN inflammation pathway, although their expression returned to pretreatment levels within 12 h (Asirvatham et al. 2006). The authors of the study speculate that R1881 may mimic or modulate the IFN response in prostate epithelial cells. Whether this response occurs in vivo is uncertain, but in mouse organ culture,

Androgen-Regulated Genes in the Prostate

649

a variety of genes with diverse functions exhibit similar kinetics (C. Pritchard and P.S. Nelson, personal communication).

6 Androgen-Regulated Gene Targets Determined by Analyses of DNA Binding 6.1

ChIP-on-Chip Analyses

Chromatin immunoprecipitation (ChIP) analyses using microarray chips, termed ChIP-on-chip, has emerged as a powerful means of interrogating the full spectrum of binding sites for transcription factors in vivo (Wu et al. 2006). In conjunction with expression profiling, ChIP-on-chip is particularly powerful because it suggests that an androgen-responsive gene is regulated directly by androgens at the level of transcription. In 2007, several groups identified binding sites for the AR transcription complex (Table 2) (Bolton et al. 2007; Jariwala et al. 2007; Massie et al. 2007; Takayama et al. 2007; Wang et al. 2007a). These results have yet to be annotated comprehensively either with each other, or with previously published microarray expression data; however, several interesting features of genes with nearby androgen-responsive binding regions (ARBRs) have emerged.

6.2

Properties of ARBRs

Large-scale ChIP-on-chip studies confirm that ARBRs are not restricted to the proximal promoter region of genes. Most ARBRs associated with androgen-responsive genes appear to be over 10 kb from the start of transcription, but ARBRs are at least as likely to be downstream from the start of transcription as upstream (Bolton et al. 2007; Jariwala et al. 2007; Takayama et al. 2007; Wang et al. 2007a). Some Table 2 Androgen-binding regions in prostate cells Study Source No. of ARBRs

Methodology No. of Functional AREsa Bolton et al. (2007) HPr-1 524 20/21 ChIP-on-chip Masuda et al.(2005) LNCaP 13 – ChIP Horie-Inoue et al. (2004) LNCaP 8 – ChIP Massie et al. (2007) LNCaP 1532 – ChIP-on-chip Takayama et al. (2007) LNCaP 10 2/2 ChIP-on-chip Wang et al. (2007) LNCaP 90 10/10 ChIP-on-chip Jariwala et al. (2007) C4-2B, LNCaP 19 – ChIP-on-chip ARBR androgen-binding region; ARE androgen-response element; ChIP chromatin immunoprecipitation a ARBRs shown to be androgen responsive using in vitro promoter assays

650

N. Clegg, P.S. Nelson

ARBRs may be 50 or more kilobases distant from the gene. Chromatin encompassing, promoter-distant ARBRs were enriched over twofold for acetylated histones, which suggests function in vivo (Takayama et al. 2007). Nevertheless, AREs are statistically enriched in the vicinity of androgen-responsive genes (Bolton et al. 2007). Many of these features have been noted for the binding regions of other transcription factors, such as the estrogen receptor (Carroll et al. 2005; Carroll et al. 2006). The bipartite ARE consensus sequence frequently is absent in ARBRs. Bolton et al. (2007) found 31% of 524 ARBRs did not have a near-consensus motif. In a genome-wide screen Massie et al. (2007) reported that only 27% (410/1,532) of ARBRs had sequences resembling a near-consensus ARE, but 79% (1,212/1,532) contained a sequence similar to a half-site. The enrichment of half-sites relative to noncandidate promoters was statistically significant. Wang et al. (2007a) found only 10% of 90 ARBRs on chromosomes 20 and 21 were similar to the consensus, but 68% had noncanonical binding sites consisting of half-sites, altered spacing between half-sites, or altered orientations of half-sites. Given these data, AR may bind consensus-like half-sites and be stabilized in vivo by cooperative binding of other elements of the transcription machinery. In support of the idea that poor matches to the ARE consensus still function in vivo, a DNA fragment which contains a half-site from the UNQ9419 promoter-bound AR in a ChIP assay (Massie et al. 2007). Also, the noncanonical ARE (AGAACA[8n]TGTTCT) in the promoter of TMPRSS2 was stimulated by androgens in a promoter assay (Wang et al. 2007a), a response which was abrogated when this ARE sequence was mutated. In some cases, genes with functional ARBRs may be clustered. Bolton et al. (2007) found two such clusters by mapping validated ARBRs to the proximity of androgen-response genes. An 800-kb cluster on chromosome 4 contained 7 androgen-repressed genes among androgen unresponsive genes. A second cluster on chromosome 1 contained 12 androgen-response genes in 370 kb. Within these clusters, all the androgen-regulated genes are either up- or down-regulated, but not both, implying functional organization. A more general survey of genes directly regulated by androgens may help in understanding the frequency, mechanism, and functional relevance of this phenomenon. Finally, ARBRs and androgen-response genes may vary considerably between cell types. Eighty-nine percent (182) of 205 androgen-responsive genes identified in HPr-1 cells as having a nearby ARBR do not appear to be androgen sensitive in LNCaP cells (Bolton et al. 2007). Further, 9 of the 23 genes that were androgenresponsive in both studies behaved in a contrary fashion: repressed genes in one cell line were stimulated in the other and vice versa. Since HPr-1 was derived from normal epithelium and LNCaP cells were derived from a metastatic cancer, these differences may represent different androgen response programs inherited from their founder cells. Substantial differences in response to androgen have also been noted between LNCaP cells and another cell line derived from normal epithelium, RC-165N/hTERT (Kim et al. 2007).

Androgen-Regulated Genes in the Prostate

6.3

651

ARBRs and Interactions Between AR and Other Transcription Factors

An important aspect of the androgen-responsive gene network is the set of interactions that govern the activity of genes stimulated by androgens. In this context, ARBRs are enriched for other transcription factors that interact with AR. For example, a subset of ARBRs has been shown enriched for Ets-binding motifs (Massie et al. 2007). In LNCaP cells, AR coimmunoprecipitates with ETS1. Also, in a reporter assay, ETS1 transfection enhanced AR transactivation in an AR- and androgen-dependent fashion. These data suggest that ETS1 contributes to androgen-responsive expression of a subset of target genes. Binding motifs for the Forkhead, GATA, and OCT transcription factors are enriched in ARBRs relative to the overall background of the genome (Wang et al. 2007a). Quantitative ChIP assays using antibodies against FoxA1, GATA2, and OCT1 indicate that all three transcription factors can be found in some ARBRs; and all three physically interact with AR. Finally, deletion of the OCT or GATA motifs in the PSA enhancer region reduces androgen-induced expression of the PSA promoter in vitro, thus demonstrating a functional interaction between AR and the other factors. siRNA inhibition of either OCT1 or GATA2 expression reduced levels of PSA and TMPRSS2 transcription, and reduced polymerase II (Pol II) loading onto both promoters. AR recruitment to both promoters was reduced by an siRNA to GATA2, but not to one to OCT1, which suggests that GATA2 is involved in recruiting AR to the PSA and TMPRSS2 promoters.

7 Androgen-Regulated Gene Targets Determined by Proteome Analysis Large-scale microarray profiling identifies androgen-responsive genes at the level of transcription, and ChIP-on-chip screens can identify potential direct targets of AR gene regulation. However, the levels of mRNA in a cell do not necessarily reflect the amount of protein product (Anderson and Seilhamer 1997; Haynes et al. 1998), and post-transcriptional and post-translational events may comprise important nodes in the cellular AR-regulated network (Fig. 1). Investigating alterations in the proteome in response to androgens may make it possible to discern alterations in protein localization, for example recruitment of proteins to the membrane or increased levels of secretion. Proteomics also enables the identification of different isoforms and states (such as phosphorylation) of proteins that are of functional significance in the living cell. Rowland et al. (2007) found evidence for an androgen-sensitive stress response in LNCaP cells using a quantitative gel-based proteomic approach. Six independent samples were androgen starved for 3 days, followed by treatment with 10 nM

652

N. Clegg, P.S. Nelson

R1881 or control vehicle (ethanol). Protein from whole cell lysates was labeled with fluorescent dye and subjected to 2-D DIGE. Treatment with androgen elevated the level of 140 gel spots and decreased the level of 57 spots. After protein sequencing, unknown and duplicated proteins were excluded, and 68 distinct proteins were up-regulated and 39 were down-regulated. Stress response proteins were approximately twofold (p = 0.013) over-represented using the hypergeometric distribution. Six of seven proteins involved in oxidative stress were up-regulated in androgen-stimulated cells (EPHX2, GSS, HSPB1, LTA4H, NANS, PRDX6), and one (TXNRD1) was down-regulated. Androgen-induced oxidative stress in LNCaP cells, which was originally identified in biochemical studies, has also been suggested using microarray expression profiling analyses of androgen-regulated genes in both rats and humans (Pang et al. 2002; Pinthus et al. 2007; Ripple et al. 1997; Segawa et al. 2002; Xu et al. 2001). In the Rowland study, 5 of 7 proteins involved in ER stress were up-regulated (HSPB1, HYOU1, STIP1, TXNDC4, VCP), and 2 were down-regulated (HTRA2 and XRCC5). The observed effect on ER stress proteins was consistent with microarray expression profiling (Segawa et al. 2002). Wright et al. (2003) used isotope-coded affinity tags and mass spectroscopy to identify and quantitate the relative abundance levels of 1,064 proteins from crude microsomal preparations of LNCaP cells that were starved of androgen or stimulated with 10 nM R1881 for 72 h. In response to androgen, the protein concentrations of known androgen-responsive genes PSA, FAS, and LDLR were increased (>1.5-fold) in microsomes. In total, 350 (33%) proteins showed an androgenresponse change greater than or equal to 1.6-fold. The Gene Ontology classification for cellular processes and statistics based on the hypergeometric distribution were used to identify enrichment of proteins in 5 of 45 tested cellular processes among 1,064 proteins tested (p < 0.004): RNA splicing; energy generation; nucleotide metabolism; lipid, fat, and steroid metabolism; and RNA processing. Two categories were under-represented: vesicular transport and protein synthesis. The RNA splicing category is noteworthy; 12 of 14 proteins were more highly represented in microsomes from androgen-deprived than from androgen-stimulated cells, which suggests coordinated regulation of this pathway through AR signaling. Consistent with observations from RNA transcript profiling, protein levels of three enzymes central to the fatty acid synthesis pathway (ATP citrate lyase, Acetyl CoA carboxylase, and fatty acid synthase) were more abundant in androgenstimulated cells than in androgen-starved cells (Wright et al. 2003). Only 2 of 11 enzymes in the cholesterol pathway were recovered in the microsome samples, but both were up-regulated in response to androgens. Numerous proteins in the TCA cycle and the oxidative phosphorylation pathways were repressed coordinately by androgen, but none was up-regulated. The latter findings suggest that androgendepleted cells have a greater metabolic demand for ATP than androgen-stimulated cells. Martin et al. (2004) attempted to quantitate androgen-responsive, secreted proteins from LNCaP cells. Such proteins may play a significant role in vivo, by mediating interactions between epithelial and stromal cells. Isotope-coded affinity tag (ICAT) mass spectrometry identified 524 quantifiable proteins from LNCaP

Androgen-Regulated Genes in the Prostate

653

cells stimulated with R1881 for 72 h. Some enrichment for secreted proteins was achieved, although proteins localized to other cellular compartments were also identified. However, very few of the recovered nuclear and cytoplasmic proteins were androgen responsive (6/343). In contrast, 22% (15/68) of the extracellular proteins and 39% (29/74) of the cytoplasmic proteins were up- or down-regulated at least twofold after androgen exposure. The proteins are a heterogenous group, many of which have been shown androgen responsive in expression profiling studies (for example, IGFBP3, ALCAM, B2M, VEGFA, and PSA). Proteins of particular interest in the Martin et al. study are those that point to a role for AR signaling in stimulating Notch regulation (JAG1 and NOTCH2). Notch signaling is required for normal prostate development and has been implicated in carcinogenesis. Connections between the AR network and the Notch pathway are emerging. For example, activation of Notch signaling inhibits prostate cell proliferation (Shou et al. 2001). Another intriguing observation arose while using Western blots to validate the androgen responsiveness of Spint1, a Kunitz-type 1 protease inhibitor. Not only was Spint1 protein up-regulated by androgen, the addition of androgens appeared to stimulate the proteolysis of Spint1 into discrete fragments. Whether androgen-stimulated proteolysis is specific to Spint1 or can be generalized to other extracellular proteins (perhaps through g-secretase ectodomain shedding) remains to be established. In either case, a role for androgens is suggested in the regulation of the abundance and function of extracellular proteins.

8 Summary Large-scale searches for androgen-responsive genes have been conducted both in vitro and in vivo in mice, rats, and humans. When these genes are categorized into functional associations, a number of common pathways emerge. These include fatty acid and cholesterol biosynthesis, polyamine biosynthesis, and an endoplasmic reticulum stress response. Individual genes that are directly or indirectly regulated at the level of transcription by AR have been found representing virtually every major functional pathway. Of particular interest are those genes with regulatory roles, such as transcription factors (NKX3.1), cell signaling pathways (IGF1, NOTCH1), and genes that modulate the translation machinery (miR-125b). The actions of such genes may account for many of the other responses to androgen observed in expression profiling studies. To date, relatively few studies have been performed to find the targets and roles of these regulators in either the normal or diseased prostate. Recent efforts, using ChIP-on-chip and in vitro reporter assays, have been directed toward defining androgen-responsive genes that are regulated directly by AR. These genes form the top level of an AR-regulated cascade of molecular events. Many AR-regulated genes have other common binding sites in their promoters, such as ETS and GATA motifs, which suggest that they share additional levels of regulation with each other.

654

N. Clegg, P.S. Nelson

The androgen-response program is extraordinarily complex. Only a small subset of genes is known to be directly regulated by androgens. How these genes interact and how they regulate other downstream aspects of the program will require systematic studies of the impact of individual genes on the AR signaling network (Fig. 1). Whether some AREs constitutively bind androgen and whether there is a hierarchy of binding (and trans-activation) remains to be established. Further, the androgen-response network appears to be dynamic. Cultured cells have a biphasic response to androgens, which suggests alternate targets. How this effect is achieved remains unknown. Also, different cell lines appear to have different androgen response programs. A major deficiency in the characterization of the AR network involves a lack of information detailing AR-responsive genes in cells comprising the prostate stroma. As androgen effects through the prostate mesenchyme are critical for normal development, future studies should delineate AR targets in this tissue compartment during embryogenesis and at maturity. Similarly, little is known about the androgen-regulated network of genes in endothelium. Research should address the extent to which different prostate cell types (for example basal cells, luminal cells, and stromal subtypes) share the same program and how they differ. Ideally these surveys would be performed on prostate tissue, not cell lines. Currently this is challenging, but modification of new high-throughput technologies, such as ChIPSeq (Johnson et al. 2007), may make it possible to assay relatively small amounts of tissue in the mouse and eventually humans. Defining the contextual intracellular and extracellular network of genes regulated by the AR and the attendant cellular and organismal phenotypes dictated by AR interactions should provide an improved understanding of the biology and pathology of the prostate.

References Abate-Shen, C., Banach-Petrosky, W.A., Sun, X., Economides, K.D., Desai, N., Gregg, J.P., Borowsky, A.D., Cardiff, R.D., and Shen, M.M. (2003). Nkx3.1; Pten mutant mice develop invasive prostate adenocarcinoma and lymph node metastases. Cancer Research 63, 3886– 3890 Abbott, D.E., Pritchard, C., Clegg, N.J., Ferguson, C., Dumpit, R., Sikes, R.A., and Nelson, P.S. (2003). Expressed sequence tag profiling identifies developmental and anatomic partitioning of gene expression in the mouse prostate. Genome Biology 4, R79 Abdulkadir, S.A., Magee, J.A., Peters, T.J., Kaleem, Z., Naughton, C.K., Humphrey, P.A., and Milbrandt, J. (2002). Conditional loss of Nkx3.1 in adult mice induces prostatic intraepithelial neoplasia. Molecular and Cellular Biology 22, 1495–1503 Adams, M.D., Kelley, J.M., Gocayne, J.D., Dunnick, M., Polymeropoulos, M.H., Xiao, H., Merril, C.R., Wu, A., Olde, B., Moreno, R.F., et al. (1991). Complementary DNA Sequencing: Expressed Sequence Tags and Human Genome Project. Science 252, 1651–1656 Agoulnik, I.U., and Weigel, N.L. (2006). Androgen receptor action in hormone-dependent and recurrent prostate cancer. Journal of Cellular Biochemistry 99, 362–372 Amler, L.C., Agus, D.B., LeDuc, C., Sapinoso, M.L., Fox, W.D., Kern, S., Lee, D., Wang, V., Leysens, M., Higgins, B., et al. (2000). Dysregulated expression of androgen-responsive

Androgen-Regulated Genes in the Prostate

655

and nonresponsive genes in the androgen-independent prostate cancer xenograft model CWR22-R1. Cancer Research 60, 6134–6141 Anderson, L., and Seilhamer, J. (1997). A comparison of selected mRNA and protein abundances in human liver. Electrophoresis 18, 533–537 Arnold, J.T., Le, H., McFann, K.K., and Blackman, M.R. (2005). Comparative effects of DHEA vs. testosterone, dihydrotestosterone, and estradiol on proliferation and gene expression in human LNCaP prostate cancer cells. American Journal of Physiology 288, E573–584 Asirvatham, A.J., Schmidt, M., Gao, B., and Chaudhary, J. (2006). Androgens regulate the immune/inflammatory response and cell survival pathways in rat ventral prostate epithelial cells. Endocrinology 147, 257–271 Barent, R.L., Nair, S.C., Carr, D.C., Ruan, Y., Rimerman, R.A., Fulton, J., Zhang, Y., and Smith, D.F. (1998). Analysis of FKBP51/FKBP52 chimeras and mutants for Hsp90 binding and association with progesterone receptor complexes. Molecular Endocrinology (Baltimore, Md) 12, 342–354 Bebermeier, J.H., Brooks, J.D., DePrimo, S.E., Werner, R., Deppe, U., Demeter, J., Hiort, O., and Holterhus, P.M. (2006). Cell-line and tissue-specific signatures of androgen receptor-coregulator transcription. Journal of Molecular Medicine 84, 919–931 Ben Aicha, S., Lessard, J., Pelletier, M., Fournier, A., Calvo, E., and Labrie, C. (2007). Transcriptional profiling of genes that are regulated by the endoplasmic reticulum-bound transcription factor AIbZIP/CREB3L4 in prostate cells. Physiological Genomics 31, 295–305 Bhatia-Gaur, R., Donjacour, A.A., Sciavolino, P.J., Kim, M., Desai, N., Young, P., Norton, C.R., Gridley, T., Cardiff, R.D., Cunha, G.R., et al. (1999). Roles for Nkx3.1 in prostate development and cancer. Genes & Development 13, 966–977 Bolton, E.C., So, A.Y., Chaivorapol, C., Haqq, C.M., Li, H., and Yamamoto, K.R. (2007). Celland gene-specific regulation of primary target genes by the androgen receptor. Genes & Development 21, 2005–2017 Bourdeau, V., Deschenes, J., Metivier, R., Nagai, Y., Nguyen, D., Bretschneider, N., Gannon, F., White, J.H., and Mader, S. (2004). Genome-wide identification of high-affinity estrogen response elements in human and mouse. Molecular Endocrinology (Baltimore, Md) 18, 1411–1427 Bova, G.S., Carter, B.S., Bussemakers, M.J., Emi, M., Fujiwara, Y., Kyprianou, N., Jacobs, S.C., Robinson, J.C., Epstein, J.I., Walsh, P.C., et al. (1993). Homozygous deletion and frequent allelic loss of chromosome 8p22 loci in human prostate cancer. Cancer Research 53, 3869–3873 Cai, C., Chen, S.Y., Zheng, Z., Omwancha, J., Lin, M.F., Balk, S.P., and Shemshedini, L. (2007a). Androgen regulation of soluble guanylyl cyclasealpha1 mediates prostate cancer cell proliferation. Oncogene 26, 1606–1615 Cai, C., Hsieh, C.L., Omwancha, J., Zheng, Z., Chen, S.Y., Baert, J.L., and Shemshedini, L. (2007b). ETV1 is a novel androgen receptor-regulated gene that mediates prostate cancer cell invasion. Molecular Endocrinology (Baltimore, Md) 21, 1835–1846 Carroll, J.S., Liu, X.S., Brodsky, A.S., Li, W., Meyer, C.A., Szary, A.J., Eeckhoute, J., Shao, W., Hestermann, E.V., Geistlinger, T.R., et al. (2005). Chromosome-wide mapping of estrogen receptor binding reveals long-range regulation requiring the forkhead protein FoxA1. Cell 122, 33–43 Carroll, J.S., Meyer, C.A., Song, J., Li, W., Geistlinger, T.R., Eeckhoute, J., Brodsky, A.S., Keeton, E.K., Fertuck, K.C., Hall, G.F., et al. (2006). Genome-wide analysis of estrogen receptor binding sites. Nature Genetics 38, 1289–1297 Ci, M., Mayumi, Y., Andre, B., Pascal, B., Lin, G., Yasukazu, T., Fernand, L., and St-Amand, J. (2008). Prostate-specific genes and their regulation by dihydrotestosterone. Prostate 68, 241–254 Claessens, F., Alen, P., Devos, A., Peeters, B., Verhoeven, G., and Rombauts, W. (1996). The androgen-specific probasin response element 2 interacts differentially with androgen and glucocorticoid receptors. The Journal of biological chemistry 271, 19013–19016

656

N. Clegg, P.S. Nelson

Clegg, N., Eroglu, B., Ferguson, C., Arnold, H., Moorman, A., and Nelson, P.S. (2002). Digital expression profiles of the prostate androgen-response program. The Journal of Steroid Biochemistry and Molecular Biology 80, 13–23 Cleutjens, K.B., van der Korput, H.A., van Eekelen, C.C., van Rooij, H.C., Faber, P.W., and Trapman, J. (1997). An androgen response element in a far upstream enhancer region is essential for high, androgen-regulated activity of the prostate-specific antigen promoter. Molecular Endocrinology (Baltimore, Md) 11, 148–161 Cleutjens, K.B., van Eekelen, C.C., van der Korput, H.A., Brinkmann, A.O., and Trapman, J. (1996). Two androgen response regions cooperate in steroid hormone regulated activity of the prostate-specific antigen promoter. The Journal of Biological Chemistry 271, 6379–6388 Corey, E., Quinn, J.E., Buhler, K.R., Nelson, P.S., Macoska, J.A., True, L.D., and Vessella, R.L. (2003). LuCaP 35: a new model of prostate cancer progression to androgen independence. Prostate 55, 239–246 Coutinho-Camillo, C.M., Salaorni, S., Sarkis, A.S., and Nagai, M.A. (2006). Differentially expressed genes in the prostate cancer cell line LNCaP after exposure to androgen and antiandrogen. Cancer Genetics and Cytogenetics 166, 130–138 Cunha, G.R., Ricke, W., Thomson, A., Marker, P.C., Risbridger, G., Hayward, S.W., Wang, Y.Z., Donjacour, A.A., and Kurita, T. (2004). Hormonal, cellular, and molecular regulation of normal and neoplastic prostatic development. The Journal of Steroid Biochemistry and Molecular Biology 92, 221–236 Dai, J.L., and Burnstein, K.L. (1996). Two androgen response elements in the androgen receptor coding region are required for cell-specific up-regulation of receptor messenger RNA. Molecular Endocrinology (Baltimore, Md) 10, 1582–1594 Davies, T.H., Ning, Y.M., and Sanchez, E.R. (2002). A new first step in activation of steroid receptors: hormone-induced switching of FKBP51 and FKBP52 immunophilins. The Journal of Biological Chemistry 277, 4597–4600 de Launoit, Y., Veilleux, R., Dufour, M., Simard, J., and Labrie, F. (1991). Characteristics of the biphasic action of androgens and of the potent antiproliferative effects of the new pure antiestrogen EM-139 on cell cycle kinetic parameters in LNCaP human prostatic cancer cells. Cancer Research 51, 5165–5170 de Wildt, S.N., Kearns, G.L., Leeder, J.S., and van den Anker, J.N. (1999). Cytochrome P450 3A: ontogeny and drug disposition. Clinical Pharmacokinetics 37, 485–505 Dehm, S.M., and Tindall, D.J. (2006). Molecular regulation of androgen action in prostate cancer. Journal of Cellular Biochemistry 99, 333–344 DePrimo, S.E., Diehn, M., Nelson, J.B., Reiter, R.E., Matese, J., Fero, M., Tibshirani, R., Brown, P.O., and Brooks, J.D. (2002). Transcriptional programs activated by exposure of human prostate cancer cells to androgen. Genome Biology 3, RESEARCH0032 Desai, K.V., Michalowska, A.M., Kondaiah, P., Ward, J.M., Shih, J.H., and Green, J.E. (2004) Gene expression profiling identifies a unique androgen-mediated inflammatory/immune signature and a PTEN (phosphatase and tensin homolog deleted on chromosome 10)-mediated apoptotic response specific to the rat ventral prostate. Molecular Endocrinology (Baltimore, Md) 18, 2895–2907 Eder, I.E., Haag, P., Basik, M., Mousses, S., Bektic, J., Bartsch, G., and Klocker, H. (2003). Gene expression changes following androgen receptor elimination in LNCaP prostate cancer cells. Molecular Carcinogenesis 37, 181–191 Evans, G.S., and Chandler, J.A. (1987). Cell proliferation studies in the rat prostate: II. The effects of castration and androgen-induced regeneration upon basal and secretory cell proliferation. The Prostate 11, 339–351 Febbo, P.G., Lowenberg, M., Thorner, A.R., Brown, M., Loda, M., and Golub, T.R. (2005). Androgen mediated regulation and functional implications of fkbp51 expression in prostate cancer. The Journal of Urology 173, 1772–1777 Feldman, B.J., and Feldman, D. (2001). The development of androgen-independent prostate cancer. Nature Reviews 1, 34–45

Androgen-Regulated Genes in the Prostate

657

Firth, S.M., and Baxter, R.C. (2002). Cellular actions of the insulin-like growth factor binding proteins. Endocrine Reviews 23, 824–854 Geck, P., Szelei, J., Jimenez, J., Lin, T.M., Sonnenschein, C., and Soto, A.M. (1997). Expression of novel genes linked to the androgen-induced, proliferative shutoff in prostate cancer cells. The Journal of steroid biochemistry and molecular Biology 63, 211–218 Goossens, K., Esquenet, M., Swinnen, J.V., Manin, M., Rombauts, W., and Verhoeven, G. (1999). Androgens decrease and retinoids increase the expression of insulin-like growth factor-binding protein-3 in LNcaP prostatic adenocarcinoma cells. Molecular and Cellular Endocrinology 155, 9–18 Grad, J.M., Lyons, L.S., Robins, D.M., and Burnstein, K.L. (2001). The androgen receptor (AR) amino-terminus imposes androgen-specific regulation of AR gene expression via an exonic enhancer. Endocrinology 142, 1107–1116 Gygi, S.P., Rist, B., Gerber, S.A., Turecek, F., Gelb, M.H., and Aebersold, R. (1999). Quantitative analysis of complex protein mixtures using isotope-coded affinity tags. Nature Biotechnology 17, 994–999 Haynes, P.A., Gygi, S.P., Figeys, D., and Aebersold, R. (1998). Proteome analysis: biological assay or data archive? Electrophoresis 19, 1862–1871 He, W.W., Sciavolino, P.J., Wing, J., Augustus, M., Hudson, P., Meissner, P.S., Curtis, R.T., Shell, B.K., Bostwick, D.G., Tindall, D.J., et al. (1997). A novel human prostate-specific, androgenregulated homeobox gene (NKX3.1) that maps to 8p21, a region frequently deleted in prostate cancer. Genomics 43, 69–77 Heemers, H.V., Verhoeven, G., and Swinnen, J.V. (2006). Androgen activation of the sterol regulatory element-binding protein pathway: current insights. Molecular Endocrinology (Baltimore, Md) 20, 2265–2277 Horie-Inoue, K., Bono, H., Okazaki, Y., and Inoue, S. (2004). Identification and functional analysis of consensus androgen response elements in human prostate cancer cells. Biochemical and Biophysical Research Communications 325, 1312–1317 Horoszewicz, J.S., Leong, S.S., Kawinski, E., Karr, J.P., Rosenthal, H., Chu, T.M., Mirand, E.A., and Murphy, G.P. (1983). LNCaP model of human prostatic carcinoma. Cancer Research 43, 1809–1818 Hsu, T., Trojanowska, M., and Watson, D.K. (2004). Ets proteins in biological control and cancer. Journal of Cellular Biochemistry 91, 896–903 Isaacs, J.T., Lundmo, P.I., Berges, R., Martikainen, P., Kyprianou, N., and English, H.F. (1992). Androgen regulation of programmed death of normal and malignant prostatic cells. Journal of Andrology 13, 457–464 Jariwala, U., Prescott, J., Jia, L., Barski, A., Pregizer, S., Cogan, J.P., Arasheben, A., Tilley, W.D., Scher, H.I., Gerald, W.L., et al. (2007). Identification of novel androgen receptor target genes in prostate cancer. Molecular Cancer 6, 39 Jiang, F., and Wang, Z. (2003). Identification of androgen-responsive genes in the rat ventral prostate by complementary deoxyribonucleic acid subtraction and microarray. Endocrinology 144, 1257–1265 Johnson, D.S., Mortazavi, A., Myers, R.M., and Wold, B. (2007). Genome-wide mapping of in vivo protein-DNA interactions. Science (New York, NY) 316, 1497–1502 Kasper, S. (2005). Survey of genetically engineered mouse models for prostate cancer: analyzing the molecular basis of prostate cancer development, progression, and metastasis. Journal of Cellular Biochemistry 94, 279–297 Kim, K.H., Dobi, A., Shaheduzzaman, S., Gao, C.L., Masuda, K., Li, H., Drukier, A., Gu, Y., Srikantan, V., Rhim, J.S., et al. (2007). Characterization of the androgen receptor in a benign prostate tissue-derived human prostate epithelial cell line: RC-165N/human telomerase reverse transcriptase. Prostate Cancer and Prostatic Diseases 10, 30–38 Kim, M.J., Bhatia-Gaur, R., Banach-Petrosky, W.A., Desai, N., Wang, Y., Hayward, S.W., Cunha, G.R., Cardiff, R.D., Shen, M.M., and Abate-Shen, C. (2002). Nkx3.1 mutant mice recapitulate early stages of prostate carcinogenesis. Cancer Research 62, 2999–3004

658

N. Clegg, P.S. Nelson

Klein, K.A., Reiter, R.E., Redula, J., Moradi, H., Zhu, X.L., Brothman, A.R., Lamb, D.J., Marcelli, M., Belldegrun, A., Witte, O.N., et al. (1997). Progression of metastatic human prostate cancer to androgen independence in immunodeficient SCID mice. Nature Medicine 3, 402–408 Kojima, S., Mulholland, D.J., Ettinger, S., Fazli, L., Nelson, C.C., and Gleave, M.E. (2006). Differential regulation of IGFBP-3 by the androgen receptor in the lineage-related androgendependent LNCaP and androgen-independent C4-2 prostate cancer models. The Prostate 66, 971–986 Kousteni, S., Bellido, T., Plotkin, L.I., O’Brien, C.A., Bodenner, D.L., Han, L., Han, K., DiGregorio, G.B., Katzenellenbogen, J.A., Katzenellenbogen, B.S., et al. (2001). Nongenotropic, sex-nonspecific signaling through the estrogen or androgen receptors: dissociation from transcriptional activity. Cell 104, 719–730 Kulik, G., and Weber, M.J. (1998). Akt-dependent and -independent survival signaling pathways utilized by insulin-like growth factor I. Molecular and cellular biology 18, 6711–6718. Kyprianou, N., and Isaacs, J.T. (1988). Activation of programmed cell death in the rat ventral prostate after castration. Endocrinology 122, 552–562 Lefstin, J.A., and Yamamoto, K.R. (1998). Allosteric effects of DNA on transcriptional regulators. Nature 392, 885–888 Lexander, H., Franzen, B., Hirschberg, D., Becker, S., Hellstrom, M., Bergman, T., Jornvall, H., Auer, G., and Egevad, L. (2005). Differential protein expression in anatomical zones of the prostate. Proteomics 5, 2570–2576 Li, B.Y., Liao, X.B., Fujito, A., Thrasher, J.B., Shen, F.Y., and Xu, P.Y. (2007). Dual androgenresponse elements mediate androgen regulation of MMP-2 expression in prostate cancer cells. Asian Journal of Andrology 9, 41–50 Louro, R., Nakaya, H.I., Amaral, P.P., Festa, F., Sogayar, M.C., da Silva, A.M., VerjovskiAlmeida, S., and Reis, E.M. (2007). Androgen responsive intronic non-coding RNAs. BMC Biology 5, 4 Macoska, J.A., Trybus, T.M., Benson, P.D., Sakr, W.A., Grignon, D.J., Wojno, K.D., Pietruk, T., and Powell, I.J. (1995). Evidence for three tumor suppressor gene loci on chromosome 8p in human prostate cancer. Cancer Research 55, 5390–5395 Magee, J.A., Abdulkadir, S.A., and Milbrandt, J. (2003). Haploinsufficiency at the Nkx3.1 locus. A paradigm for stochastic, dosage-sensitive gene regulation during tumor initiation. Cancer Cell 3, 273–283 Magee, J.A., Chang, L.W., Stormo, G.D., and Milbrandt, J. (2006). Direct, androgen receptormediated regulation of the FKBP5 gene via a distal enhancer element. Endocrinology 147, 590–598 Marker, P.C., Donjacour, A.A., Dahiya, R., and Cunha, G.R. (2003). Hormonal, cellular, and molecular control of prostatic development. Developmental Biology 253, 165–174 Martin, D.B., Gifford, D.R., Wright, M.E., Keller, A., Yi, E., Goodlett, D.R., Aebersold, R., and Nelson, P.S. (2004). Quantitative proteomic analysis of proteins released by neoplastic prostate epithelium. Cancer Research 64, 347–355 Martin, J.L., and Pattison, S.L. (2000). Insulin-like growth factor binding protein-3 is regulated by dihydrotestosterone and stimulates deoxyribonucleic acid synthesis and cell proliferation in LNCaP prostate carcinoma cells. Endocrinology 141, 2401–2409 Massie, C.E., Adryan, B., Barbosa-Morais, N.L., Lynch, A.G., Tran, M.G., Neal, D.E., and Mills, I.G. (2007). New androgen receptor genomic targets show an interaction with the ETS1 transcription factor. EMBO Reports 8, 871–878 Masuda, K., Werner, T., Maheshwari, S., Frisch, M., Oh, S., Petrovics, G., May, K., Srikantan, V., Srivastava, S., and Dobi, A. (2005). Androgen receptor binding sites identified by a GREF_GATA model. Journal of Molecular Biology 353, 763–771 McNeal, J.E. (1981a). Normal and pathologic anatomy of prostate. Urology 17, 11–16 McNeal, J.E. (1981b). The zonal anatomy of the prostate. The Prostate 2, 35–49 Meehan, K.L., and Sadar, M.D. (2004). Quantitative profiling of LNCaP prostate cancer cells using isotope-coded affinity tags and mass spectrometry. Proteomics 4, 1116–1134

Androgen-Regulated Genes in the Prostate

659

Migliaccio, A., Castoria, G., Di Domenico, M., de Falco, A., Bilancio, A., Lombardi, M., Barone, M.V., Ametrano, D., Zannini, M.S., Abbondanza, C., et al. (2000). Steroid-induced androgen receptor-oestradiol receptor beta-Src complex triggers prostate cancer cell proliferation. The EMBO Journal 19, 5406–5417 Mitchell, S.H., Murtha, P.E., Zhang, S., Zhu, W., and Young, C.Y. (2000). An androgen response element mediates LNCaP cell dependent androgen induction of the hK2 gene. Molecular and Cellular Endocrinology 168, 89–99 Moilanen, A.M., Hakkola, J., Vaarala, M.H., Kauppila, S., Hirvikoski, P., Vuoristo, J.T., Edwards, R.J., and Paavonen, T.K. (2007). Characterization of androgen-regulated expression of CYP3A5 in human prostate. Carcinogenesis 28, 916–921 Mostaghel, E.A., Page, S.T., Lin, D.W., Fazli, L., Coleman, I.M., True, L.D., Knudsen, B., Hess, D.L., Nelson, C.C., Matsumoto, A.M., et al. (2007). Intraprostatic androgens and androgenregulated gene expression persist after testosterone suppression: therapeutic implications for castration-resistant prostate cancer. Cancer Res 67, 5033–5041 Mousses, S., Wagner, U., Chen, Y., Kim, J.W., Bubendorf, L., Bittner, M., Pretlow, T., Elkahloun, A.G., Trepel, J.B., and Kallioniemi, O.P. (2001). Failure of hormone therapy in prostate cancer involves systematic restoration of androgen responsive genes and activation of rapamycin sensitive signaling. Oncogene 20, 6718–6723 Murtha, P.E., Zhu, W., Zhang, J., Zhang, S., and Young, C.Y. (1997). Effects of Ca++ mobilization on expression of androgen-regulated genes: interference with androgen receptor-mediated transactivation by AP-I proteins. The Prostate 33, 264–270 Nagaraj, S.H., Gasser, R.B., and Ranganathan, S. (2007). A hitchhiker’s guide to expressed sequence tag (EST) analysis. Briefings in Bioinformatics 8, 6–21 Nantermet, P.V., Xu, J., Yu, Y., Hodor, P., Holder, D., Adamski, S., Gentile, M.A., Kimmel, D.B., Harada, S., Gerhold, D., et al. (2004). Identification of genetic pathways activated by the androgen receptor during the induction of proliferation in the ventral prostate gland. The Journal of Biological Chemistry 279, 1310–1322 Nelson, P.S., Clegg, N., Arnold, H., Ferguson, C., Bonham, M., White, J., Hood, L., and Lin, B. (2002). The program of androgen-responsive genes in neoplastic prostate epithelium. Proceedings of the National Academy of Sciences of the United States of America 99, 11890–11895 Ness, S.A. (2007). Microarray analysis: basic strategies for successful experiments. Molecular biotechnology 36, 205–219 Nickerson, T., Pollak, M., and Huynh, H. (1998). Castration-induced apoptosis in the rat ventral prostate is associated with increased expression of genes encoding insulin-like growth factor binding proteins 2,3,4 and 5. Endocrinology 139, 807–810 Nickols, N.G., and Dervan, P.B. (2007). Suppression of androgen receptor-mediated gene expression by a sequence-specific DNA-binding polyamide. Proceedings of the National Academy of Sciences of the United States of America 104, 10418-10424 Oosterhoff, J.K., Grootegoed, J.A., and Blok, L.J. (2005). Expression profiling of androgendependent and -independent LNCaP cells: EGF versus androgen signalling. Endocrine-related Cancer 12, 135–148 Pang, S.T., Dillner, K., Wu, X., Pousette, A., Norstedt, G., and Flores-Morales, A. (2002). Gene expression profiling of androgen deficiency predicts a pathway of prostate apoptosis that involves genes related to oxidative stress. Endocrinology 143, 4897–4906 Peng, L., Malloy, P.J., Wang, J., and Feldman, D. (2006). Growth inhibitory concentrations of androgens up-regulate insulin-like growth factor binding protein-3 expression via an androgen response element in LNCaP human prostate cancer cells. Endocrinology 147, 4599–4607 Pfundt, R., Smit, F., Jansen, C., Aalders, T., Straatman, H., van der Vliet, W., Isaacs, J., van Kessel, A.G., and Schalken, J. (2005). Identification of androgen-responsive genes that are alternatively regulated in androgen-dependent and androgen-independent rat prostate tumors. Genes, Chromosomes & Cancer 43, 273–283

660

N. Clegg, P.S. Nelson

Pinthus, J.H., Bryskin, I., Trachtenberg, J., Lu, J.P., Singh, G., Fridman, E., and Wilson, B.C. (2007). Androgen induces adaptation to oxidative stress in prostate cancer: implications for treatment with radiation therapy. Neoplasia (New York, NY) 9, 68–80 Porter, D., Yao, J., and Polyak, K. (2006). SAGE and related approaches for cancer target identification. Drug Discovery Today 11, 110–118 Qi, H., Fillion, C., Labrie, Y., Grenier, J., Fournier, A., Berger, L., El-Alfy, M., and Labrie, C. (2002). AIbZIP, a novel bZIP gene located on chromosome 1q21.3 that is highly expressed in prostate tumors and of which the expression is up-regulated by androgens in LNCaP human prostate cancer cells. Cancer Research 62, 721–733 Rae, J.M., Johnson, M.D., Cordero, K.E., Scheys, J.O., Larios, J.M., Gottardis, M.M., Pienta, K.J., and Lippman, M.E. (2006). GREB1 is a novel androgen-regulated gene required for prostate cancer growth. The Prostate 66, 886–894 Rhee, H.J., Kim, E.J., and Lee, J.K. (2007). Physiological polyamines: simple primordial stress molecules. Journal of Cellular and Molecular Medicine 11, 685–703 Riegman, P.H., Vlietstra, R.J., van der Korput, J.A., Brinkmann, A.O., and Trapman, J. (1991). The promoter of the prostate-specific antigen gene contains a functional androgen responsive element. Molecular Endocrinology (Baltimore, Md) 5, 1921–1930 Ripple, M.O., Henry, W.F., Rago, R.P., and Wilding, G. (1997). Prooxidant-antioxidant shift induced by androgen treatment of human prostate carcinoma cells. Journal of the National Cancer Institute 89, 40–48 Rowland, J.G., Robson, J.L., Simon, W.J., Leung, H.Y., and Slabas, A.R. (2007). Evaluation of an in vitro model of androgen ablation and identification of the androgen responsive proteome in LNCaP cells. Proteomics 7, 47–63 Sato, N., Sadar, M.D., Bruchovsky, N., Saatcioglu, F., Rennie, P.S., Sato, S., Lange, P.H., and Gleave, M.E. (1997). Androgenic induction of prostate-specific antigen gene is repressed by protein-protein interaction between the androgen receptor and AP-1/c-Jun in the human prostate cancer cell line LNCaP. The Journal of Biological Chemistry 272, 17485–17494 Schena, M., Shalon, D., Davis, R.W., and Brown, P.O. (1995). Quantitative Monitoring of Gene Expression Patterns with a Complementary DNA Microarray. Science 270, 467–470 Schoenmakers, E., Alen, P., Verrijdt, G., Peeters, B., Verhoeven, G., Rombauts, W., and Claessens, F. (1999). Differential DNA binding by the androgen and glucocorticoid receptors involves the second Zn-finger and a C-terminal extension of the DNA-binding domains. The Biochemical Journal 341 ( Pt 3), 515–521 Segawa, T., Nau, M.E., Xu, L.L., Chilukuri, R.N., Makarem, M., Zhang, W., Petrovics, G., Sesterhenn, I.A., McLeod, D.G., Moul, J.W., et al. (2002). Androgen-induced expression of endoplasmic reticulum (ER) stress response genes in prostate cancer cells. Oncogene 21, 8749–8758 Shi, X.B., Ma, A.H., Tepper, C.G., Xia, L., Gregg, J.P., Gandour-Edwards, R., Mack, P.C., Kung, H-J., and deVere White, R.W. (2004). Molecular alterations associated with LNCaP cell progression to androgen independence. Prostate 60, 257–271 Shi, X.B., Xue, L., Yang, J., Ma, A.H., Zhao, J., Xu, M., Tepper, C.G., Evans, C.P., Kung, H.J., and deVere White, R.W. (2007). An androgen-regulated miRNA suppresses Bak1 expression and induces androgen-independent growth of prostate cancer cells. Proceedings of the National Academy of Sciences of the United States of America 104, 19983–19988 Shou, J., Ross, S., Koeppen, H., de Sauvage, F.J., and Gao, W.Q. (2001). Dynamics of notch expression during murine prostate development and tumorigenesis. Cancer research 61, 7291– 7297 Sun, C., Shi, Y., Xu, L.L., Nageswararao, C., Davis, L.D., Segawa, T., Dobi, A., McLeod, D.G., and Srivastava, S. (2006). Androgen receptor mutation (T877A) promotes prostate cancer cell growth and cell survival. Oncogene 25, 3905–3913 Takayama, K., Kaneshiro, K., Tsutsumi, S., Horie-Inoue, K., Ikeda, K., Urano, T., Ijichi, N., Ouchi, Y., Shirahige, K., Aburatani, H., et al. (2007). Identification of novel androgen response

Androgen-Regulated Genes in the Prostate

661

genes in prostate cancer cells by coupling chromatin immunoprecipitation and genomic microarray analysis. Oncogene 26, 4453–4463 Tonge, R., Shaw, J., Middleton, B., Rowlinson, R., Rayner, S., Young, J., Pognan, F., Hawkins, E., Currie, I., and Davison, M. (2001). Validation and development of fluorescence two-dimensional differential gel electrophoresis proteomics technology. Proteomics 1, 377–396 Trevino, V., Falciani, F., and Barrera-Saldana, H.A. (2007). DNA microarrays: a powerful genomic tool for biomedical and clinical research. Molecular Medicine (Cambridge, Mass) 13, 527–541 van der Heul-Nieuwenhuijsen, L., Hendriksen, P.J., van der Kwast, T.H., and Jenster, G. (2006). Gene expression profiling of the human prostate zones. BJU International 98, 886–897 Velasco, A.M., Gillis, K.A., Li, Y., Brown, E.L., Sadler, T.M., Achilleos, M., Greenberger, L.M., Frost, P., Bai, W., and Zhang, Y. (2004). Identification and validation of novel androgenregulated genes in prostate cancer. Endocrinology 145, 3913–3924 Velculescu, V.E., Zhang, L., Vogelstein, B., and Kinzler, K.W. (1995). Serial analysis of gene expression. Science 270, 384–387 Waghray, A., Feroze, F., Schober, M.S., Yao, F., Wood, C., Puravs, E., Krause, M., Hanash, S., and Chen, Y.Q. (2001). Identification of androgen-regulated genes in the prostate cancer cell line LNCaP by serial analysis of gene expression and proteomic analysis. Proteomics 1, 1327– 1338 Wang, Q., Li, W., Liu, X.S., Carroll, J.S., Janne, O.A., Keeton, E.K., Chinnaiyan, A.M., Pienta, K.J., and Brown, M. (2007a). A hierarchical network of transcription factors governs androgen receptor-dependent prostate cancer growth. Molecular Cell 27, 380–392 Wang, X.D., Wang, B.E., Soriano, R., Zha, J., Zhang, Z., Modrusan, Z., Cunha, G.R., and Gao, W.Q. (2007b). Expression profiling of the mouse prostate after castration and hormone replacement: implication of H-cadherin in prostate tumorigenesis. Differentiation; Research in Biological Diversity 75, 219–234 Whitaker, H.C., Stanbury, D.P., Brinham, C., Girling, J., Hanrahan, S., Totty, N., and Neal, D.E. (2007). Labeling and identification of LNCaP cell surface proteins: a pilot study. Prostate 67, 943–954 Wright, M.E., Eng, J., Sherman, J., Hockenbery, D.M., Nelson, P.S., Galitski, T., and Aebersold, R. (2003). Identification of androgen-coregulated protein networks from the microsomes of human prostate cancer cells. Genome Biology 5, R4 Wu, J., Smith, L.T., Plass, C., and Huang, T.H. (2006). ChIP-chip comes of age for genome-wide functional analysis. Cancer Research 66, 6899–6902 Wu, Y., Zhao, W., Zhao, J., Pan, J., Wu, Q., Zhang, Y., Bauman, W.A., and Cardozo, C.P. (2007). Identification of androgen response elements in the insulin-like growth factor I upstream promoter. Endocrinology 148, 2984–2993 Xu, L.L., Su, Y.P., Labiche, R., Segawa, T., Shanmugam, N., McLeod, D.G., Moul, J.W., and Srivastava, S. (2001). Quantitative expression profile of androgen-regulated genes in prostate cancer cells and identification of prostate-specific genes. International Journal of Cancer 92, 322–328 Yuan, X.J., and Whang, Y.E. (2002). PTEN sensitizes prostate cancer cells to death receptormediated and drug-induced apoptosis through a FADD-dependent pathway. Oncogene 21, 319–327

Mapping the Androgen Receptor Cistrome Qianben Wang and Myles Brown

Abstract Androgen receptor (AR) is a ligand-dependent transcription factor that plays a key role in prostate cancer. Mapping the AR cistrome (cis-targets of a transacting factor across the genome) by ChIP-on-chip (chromatin immunoprecipitation on a microarray) discovers that the majority of AR-binding regions is far from the promoters of AR target genes and contains noncanonical AR responsive elements (ARE). Importantly, a noncanonical ARE is served as a cis-regulatory target of AR action in TMPRSS2, a gene fused to ETS transcription factors in the majority of prostate cancers. In addition other DNA sequence motifs are significantly enriched within the AR-binding regions and are bound by cooperating transcription factors FoxA1, GATA2, Oct1, and ETS1. These transcription factors cooperate in mediating the androgen response and offer potential new opportunities for therapeutic intervention.

1 Introduction Androgens, functioning through the androgen receptor (AR), are essential for the initiation and development of prostate cancer (Heinlein and Chang 2004). Current treatments for androgen-dependent advanced prostate cancer involve androgenablation therapy through surgical castration or the use of luteinizing hormonereleasing hormone (LHRH) agonists. While such therapies often lead to disease regression they are rarely curative. In general advanced prostate cancer ultimately progresses to an androgen-independent late stage that is refractory to current therapies (Debes and Tindall 2004). The molecular basis for the transition to androgen independence is poorly understood. AR is expressed in nearly all prostate cancers (Balk 2002; Scher and Sawyers 2005), and decreasing levels of AR protein expression significantly reduces both androgen-dependent and -independent prostate cancer cell growth in model systems (Eder et al. 2002; Zegarra-Moro et al.

M. Brown(*) Division of Molecular and Cellular Oncology, Department of Medical Oncology, Dana-Farber Cancer Institute and Harvard Medical School, Boston, MA 02115, USA, E-mail: myles_brown@ dfc.harvard.edu

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4_28, # Springer Science + Business Media, LLC 2009

663

664

Q. Wang, M. Brown

2002; Wright et al. 2003; Haag et al. 2005; Liao et al. 2005; Yuan et al. 2006). Thus it appears that AR signaling pathways play a major role in both androgen-dependent and androgen-independent prostate cancer. AR is a ligand-dependent transcription factor belonging to the nuclear hormone receptor (NR) superfamily that mediates the action of lipophilic ligands including steroids, retinoids, vitamin D3, and thyroid hormone (Tsai and O’Malley 1994; Mangelsdorf et al. 1995). NRs share common structure and functional domains including a variable N-terminal domain (NTD), a centrally located highly conserved DNA-binding domain (DBD), a hinge region, and a conserved C-terminal ligand-binding domain (LBD) (Parker 1993; Tsai and O’Malley 1994; Mangelsdorf et al. 1995). Transcriptional activation by NRs is mediated by both a poorly conserved constitutive activation function (AF-1) in the NTD and a highly conserved, ligand-inducible activation function (AF-2) in the LBD (Danielian et al. 1992; Nagpal et al. 1992; Barettino et al. 1994; Kato et al. 1995). NRs can be divided into four major classes based on their DNA-binding and ligand-binding properties in vitro. AR belongs to Class I NRs, which also includes glucocorticoid receptor (GR), progesterone receptor (PR), and mineralocorticoid receptor (MR). These receptors bind as homodimers and recognize very similar DNA responsive elements consisting of inverted repeats (Parker 1993; Tsai and O’Malley 1994; Mangelsdorf and Evans 1995; Mangelsdorf et al. 1995). Since the cloning of the first NR more than 20 years ago (Hollenberg et al. 1985; Miesfeld et al. 1986), considerable progress has been made in our understanding of how NRs translate hormonal signals to target gene expression. One of the major breakthroughs in NR-mediated gene transcription comes from the identification of NR coregulatory factors that include coactivators and corepressors. In general, unliganded steroid receptors are associated with heat shock proteins, whereas a subset of nonsteroid receptors recruits corepressor complexes such as the N-CoR/ SMRT complex to repress gene expression. Upon hormone binding, NRs recruit a number of coactivator complexes, including p160 family proteins, histone acetyltransferases (CBP/p300 complex), histone methyltransferases (CARM1 and PRMT1), RNA or RNA processing factors (SRA and PGC1), and Mediator complex, to enhance NR-mediated transcription (Lemon and Tjian 2000; McKenna and O’Malley 2002; Perissi and Rosenfeld 2005). The application of chromatin immunoprecipitation (ChIP) to study protein– DNA interaction has provided a wealth of information on temporal and spatial assembly of NR/coregulatory factor complexes on target gene regulatory regions in vivo. Recent ChIP studies have found that AR and its coactivators bind to the regulatory regions of the AR target gene PSA (prostate-specific antigen) following androgen stimulation. Interestingly these studies demonstrate that AR transcriptional enhancer activity is dependent on RNA polymerase II (Pol II) tracking from the PSA enhancer to its promoter through looped DNA and on Mediator complex (Shang et al. 2002; Wang et al. 2002, 2005; Louie et al. 2003; Jia et al. 2004). While androgen triggers the loading of AR and coactivators primarily to the PSA enhancer, antiandrogens facilitate AR and corepressor recruitment predominantly to the PSA promoter (Shang et al. 2002; Kang et al. 2004). Despite the well-characterized

Mapping the Androgen Receptor Cistrome

665

dynamics of AR transcription complex assembly, these data are restricted to a very limited number of genes, mainly the PSA gene and the KLK2 gene. Such a limited number of target genes greatly limit our understanding of how AR regulates the target gene network whose translation products are responsible for androgenstimulated prostate cancer growth and progression to androgen-independent prostate cancer. Genome-wide analysis of expression arrays over the past decade in prostate cancer cells treated with androgens has identified numerous other androgenregulated genes (DePrimo et al. 2002; Nelson et al. 2002; Segawa et al. 2002; Velasco et al. 2004); however, the underlying regulatory mechanisms remain elusive. The recent development of the ChIP-on-chip (ChIP on a microarray) technique allows the global identification of specific transcription factor regulatory regions across all nonrepetitive regions of the human genome. In this technique (Fig. 1), a standard ChIP is first performed. The ChIP-enriched, amplified DNA fragments are then labeled and hybridized to DNA microarrays that tile the entire genome or part of the genome. Combining the identification of transcription-factorbinding targets identified from ChIP-on-chip with differentially expressed genes (activated or repressed) from expression microarrays allows the first-order discovery of transcription factor direct target genes. Several groups including our own have used ChIP-on-chip to map AR binding in the genome. We have coined the term ‘‘cistrome’’ to describe the set of cis-targets of a trans-acting factor across the genome. In this chapter, we will summarize the conclusions drawn from these studies about where and how AR binds across the human genome in prostate cells to mediate transcription of AR target genes.

2 Identification of AR Binding Sites in the Human Genome Recent work from our laboratory identified 90 novel AR binding sites on human chromosomes 21 and 22 in a human prostate cancer cell line LNCaP by using ChIP combined with Affymetrix-tiled oligonucleotide microarrays that cover the entire nonrepetitive sequence of these two chromosomes (Wang et al. 2007). We tested 29 of these predicted AR-binding regions by conventional ChIP assays, and all were confirmed. As we identified AR-binding sites from ChIP-on-chip based on a stringent statistical cut-off ( p < 1 105) and the fold enrichment determined by quantitative ChIP is highly correlated with p value, it is likely that very few of the identified AR binding sites are false positives. Several other groups performed similar AR ChIP-on-chip using DNA microarrays covering distinct regions of the human genome. Takayama et al. (2007) conducted AR ChIP-on-chip in LNCaP cells using Affymetrix ENCODE (ENCyclopedia of DNA Elements) arrays that tile 1% of human genome to identify ten AR binding sites. All ten AR binding sites were verified to be authentic by directed ChIP. Work by Massie et al. (2007) performed AR ChIP-on-chip in LNCaP cells using a NimbleGen promoter array with a tiling coverage of 24,275 gene promoter regions. Their study identified AR binding to 1,532 promoter proximal regions with a false discovery rate (FDR) of

666

Q. Wang, M. Brown

Fig. 1 ChIP-on-chip technique. The crosslinked chromatin is sonicated and immunoprecipitated with a specific antibody targeting a transcription factor (TF). The crosslinked protein–DNA complex is then reversed. Purified DNA fragments are amplified, labeled, and hybridized to DNA microarrays that tile the whole genome or part of the genome.

approximately 13%. While these AR ChIP-on-chip studies utilized the LNCaP prostate cancer model, a study by Bolton et al. (2007) used an immortalized normal prostate cell line stably transfected with a wild-type AR (Hpr-1AR). In this study they hybridized AR ChIP-enriched DNA from Hpr-1AR cells to custom NimbleGen-tiled microarrays that cover 104-kb genomic regions centered on the transcription start sites (TSS) of 548 genes that are regulated by AR or GR in various cell lines. In this study less than 2% of the 524 identified AR-binding regions were false positives. In addition to these ChIP-on-chip studies, a sequencing based ‘‘ChIP Display’’ technology identified 19 novel AR-binding sites in an androgen-independent prostate cancer cell line C4-2B (Jariwala et al. 2007). Together, these studies provide a large amount of information of where AR binds to the human genome.

Mapping the Androgen Receptor Cistrome

667

More recently we have completed a whole genome AR ChIP-on-chip study with LNCaP cells (Wang, Q., Li, W., Zhang, Y., Yuan X., Xu, K., Yu, J., Chen, Z., Beroukhim R, Wang, H., Wu, T., Lupien M., Carroll J.S., Manrai A.K., Ja¨nne O.A., Balk S.P., Mehra R., Chinnaiyan A.M., Rubin M.A., True L., Fiorentino M., Fiore C., Loda, M., Kantoff P.W., Liu, X.S., and Brown, M, manuscript submitted). Using an FDR of 5% we have identified 8,700 AR-binding sites across the entire human genome. We compared the AR-binding sites defined by these other studies on subsets of the human genome and found a very high degree of concordance. A detailed analysis of the overlap between the various studies may provide interesting insights into AR action. These analyses are complicated however by the different array platforms and differing biological conditions. For example, it has been documented that using different androgens such as the natural androgen 5a-dihydrotestosterone (DHT) versus a synthetic androgen R1881 results in a significant degree of variation among the androgen-induced genes identified in different expression profiling studies (Velasco et al. 2004). It will be interesting to see whether these same types of effects of different androgens are seen in the binding pattern of AR across the entire genome.

3 Most AR-Binding Sites Are Far from the Promoters of AR Target Genes Recent studies on AR regulation of the PSA gene have found that AR is predominantly loaded on the PSA enhancer rather than the PSA promoter (Kang et al. 2004; Wang et al. 2005). However, this is only one example of how AR can regulate a target gene. In our chromosome-wide AR ChIP-on-chip study we correlated ARbinding sites with androgen-dependent gene expression and found that almost all AR-binding sites are far from the TSSs of androgen-regulated genes (>10 kb). For example, TMPRSS2, an androgen-regulated gene that has recently been reported to fuse with ETS transcription factors to mediate androgen-responsive overexpression of ERG, ETV1, and ETV4 in a majority of human prostate tumors (Tomlins et al. 2005, 2006), contains an AR-binding site 13.5 kb upstream of the TSS. PDE9A, a member of cyclic nucleotide phosphodiesterases (PDE) that plays a critical role in regulating intracellular concentration of cyclic nucleotides and is highly expressed in prostate (Fisher et al. 1998; Guipponi et al. 1998), contains two intronic ARbinding sites that are 18.4 and 77.8 kb downstream of the TSS, respectively (Wang et al. 2007). Consistent with our findings, two other unbiased AR ChIP-on-chip studies reported that most AR bindings are located in nonpromoter regions (>10 kb) of androgen-regulated genes (Bolton et al. 2007; Takayama et al. 2007). Interestingly, our recent work on ER mapping also found that ER-binding sites are distal from estrogen-regulated genes (Carroll et al. 2005, 2006). In parallel with AR and ER studies, So et al. found that the majority of GR binding are located >10 kb from their target promoters (So et al. 2007). These studies suggest that position-independent binding may be a general rule for steroid receptors. This low frequency of promoter proximal binding for steroid receptors is unlikely to be caused by a technical artifact as shown by effective recruitment of promoter-specific factor Pol II and standard ChIP analysis. For example, a systematic analysis of

668

Q. Wang, M. Brown

TMPRSS2 14 kb upstream sequence found that only the 14 to 13 kb region corresponding to the 13.5 kb binding site has significant AR-dependent enhancer activity. In addition, standard ChIP assays demonstrated that AR is significantly recruited to the 13.5 kb region, but only weakly binds to the TMPRSS2 promoter region that contains a canonical androgen responsive element (ARE) (Wang et al. 2007). Another parallel Pol II and ER ChIP-on-chip study showed that the vast majority of Pol II-binding sites are at promoter regions, whereas most ER-binding sites occur in distal regions (Carroll et al. 2006). Interestingly, two recent studies reported that significant numbers of Sp1, cMyc, and p53 binding sites are also located at distal regions (Cawley et al. 2004; Wei et al. 2006), suggesting that distal binding is not exclusive to steroid receptors. Several lines of evidence have suggested that many of these distal AR-binding sites function as enhancers. First, some AR-binding sites are occupied by acetylated histones H3 and H4, Pol II and p160 coactivators (Takayama et al. 2007; Wang et al. 2007) in the presence of androgen which is indicative of transcriptional activation. Second, most AR-binding sites tested can function as enhancers when cloned upstream of minimal promoters (Bolton et al. 2007; Takayama et al. 2007; Wang et al. 2007). Third, comparative genomic analysis demonstrated that the center of AR-binding sites is highly conserved between genomes of seven vertebrate species (chimp, dog, mouse, rat, chicken, fugu, and zebrafish) (Wang et al. 2007), supporting the concept that noncoding sequences, which have survived selection, play important roles in gene regulation (Hardison 2000). The finding of distal AR-binding sites raises the question as to how these enhancers communicate with proximal promoters. The looping model, which proposes that enhancer bound proteins directly interact with promoters with the looping out of the intervening DNA, is an attractive hypothesis (Blackwood and Kadonaga 1998; Bulger and Groudine 1999). Recently, the advent of ChIPcombined chromosome conformation capture (ChIP-3C) assay (Dekker et al. 2002; Carroll et al. 2005; Horike et al. 2005; Wang et al. 2005) allows us to test whether looping occurs in vivo (Fig. 2). Here, a ligation is conducted using fragments of the enhancer and promoter generated by digestion with the same restriction enzyme. If the distal enhancers are physically associated with promoter-proximal regions due to looping, digested enhancers/promoters should be able to be ligated and detected by PCR using one primer in the enhancer and the other in the promoter. Using ChIP-3C, we recently demonstrated that the PSA enhancer 4 kb upstream of the TSS comes in close vicinity to the PSA promoter (Wang et al. 2005). As one example for novel enhancers, our ChIP-3C data supported the formation of a chromosomal loop between the 13.5 kb AR-binding site and the proximal TMPRSS2 promoter. In contrast, no loop was formed between a 462 kb downstream AR binding site and the TMPRSS2 promoter (Wang et al. 2007). These results are consistent with the finding that the androgen responsive fusion genes comprised TMPRSS2 50 upstream rather than 30 downstream region and coding regions of ETS oncogenes (Tomlins et al. 2005, 2006). The identification of this distal TMPRSS2 enhancer provides a molecular explanation of how TMPRSS2ETS transcription factor fusions are driven.

Mapping the Androgen Receptor Cistrome

669

Fig. 2 ChIP combined with chromosome conformation capture (ChIP-3C) assay. Crosslinked chromatin are sonicated and digested with restriction enzymes that recognize both enhancer and promoter regions. Transcription factor (TF) ChIP is performed following by religation. The religated DNA is then purified, verified by PCR, and sequencing confirmed

While the ChIP-3C assays allowed us to assign the distal AR binding sites to the PSA and TMPRSS2 genes, it is not feasible to perform this assay for each binding site to identify its regulated gene in view of the large number of AR binding sites in the genome. The advent of a high-throughput adaptation of 3C, 3C-carbon copy (5C), may permit us to detect large-scale interactions of enhancers and promoters. This technique uses ligation-mediated amplification (LMA) to generate a ‘‘carbon copy’’ of a selected part of 3C products. The resulting products are then amplified and hybridized to microarrays or sequenced (Dostie et al. 2006; Dostie and Dekker 2007). Such global looping information may help to assign distal AR binding sites to promoters of all protein-coding and noncoding genes that show differential expression by AR induction. Nonetheless, it is worthy to point out that it is likely that only a subset of AR binding sites are functional in one cell line under the specific experimental conditions tested and may be functional in different cell types and/or different conditions. It is also possible that some AR binding sites are indeed nonfunctional, as previously suggested (van Steensel 2005).

4 The Majority of AR-Binding Regions Contain Noncanonical AREs The responsive elements of class I NR have been characterized in detail by gel shift and reporter gene assays and shown to be 3-nucleotide (nt)-spaced palindromes of TGTTCT half-sites (Evans 1988; Green and Chambon 1988). Surprisingly, our

670

Q. Wang, M. Brown

motif search on the 90 AR binding regions on chromosomes 21 and 22 found that only 10% of the binding regions have the above canonical class I NR-binding motif. Of all the AR binding sites, 68% have single or multiple noncanonical AREs that include isolated ARE half-sites, head-to-head AREs (AGAACA[0-8n]TGTTCT, n6¼ 3), tail-to-tail AREs (TGTTCT[0-8n]AGAACA), and ARE direct repeats (AGAACA[0-8n]AGAACA) (Wang et al. 2007). Similarly, Massie et al. (2007) found that only 26.8% of all 1,532 AR binding sites on promoters contained sequence that resembled the canonical AREs. The 6-bp half sites occur in 57.2% of the all AR binding sequences. A study using a tiled custom array in Hpr-1AR cells also revealed that the majority of AR binding sites contain a partial palindromic motif rather than the canonical motif (Bolton et al. 2007). In addition, 68% of the GR-binding elements identified from a GR ChIP-on-chip study were imperfect palindromes (So et al. 2007). These results are confirmed by our unpublished whole genome data in which only 15% of the binding sites contain canonical AREs. In fact, a few noncanonical AREs have been reported previously (Verrijdt et al. 2003, 2006). For example, the rat probasin (pb) promoter, the human secretory component (sc) upstream enhancer, and the mouse sex-limited protein (slp) enhancer all contain partial direct repeats (Verrijdt et al. 2003). Structural studies demonstrated that the AR DNA-binding domain can bind to direct repeats of the hexameric half-site (ADR3) as a head-to-head symmetrical dimmer that is similar to GR’s DBD binding to a canonical class I NR motif (Luisi et al. 1991; Shaffer et al. 2004). Consistent with the structural evidence supporting AR binding to noncanonical motifs, a biochemical study demonstrated that AR can bind directly to the ARE half-sites identified from ChIP-on-chip using an in vitro oligonucleotide pull-down assay (Massie et al. 2007). Furthermore, the noncanonical motifs within the AR binding regions are able to mediate AR-dependent transcription. For example, the 13.5 kb binding site upstream of the TMPRSS2 mRNA start site demonstrates enhancer activity in a reporter gene assay. Deletion of the single noncanonical ARE within this region completely abolishes its enhancer activity (Wang et al. 2007). These findings of noncanonical AREs within the majority of AR binding sites in the genome suggest that transcription-factor-binding motifs as defined in the pre-ChIP era are under scrutiny in the face of the ability to find additional binding sites using ChIP-on-chip.

5 Identification of AR-Collaborating Transcription Factor Complexes While prokaryotes use single proteins for gene transcriptional regulation, eukaryotic gene expression normally involves combinatorial regulation by multiple proteins (Remenyi et al. 2004). Current studies on combinatorial regulation of gene expression by NRs and other proteins focus on interaction of NR and non-DNA binding coregulatory factors (McKenna and O’Malley 2002; Perissi and Rosenfeld

Mapping the Androgen Receptor Cistrome

671

2005), though a few DNA-binding collaborating transcription factors have been reported. For example, AR has been found to interact with NFkB (Palvimo et al. 1996), AP1 (Sato et al. 1997), PDEF (Oettgen et al. 2000), SRY (Yuan et al. 2001), Oct1 (Gonzalez and Robins 2001), and FoxA1 (Gao et al. 2003). However, the interaction of AR and these transcription factors was demonstrated by in vitro interaction assays and/or coimmunoprecipitation, and the functional consequences of these interactions were analyzed by reporter gene assays. Thus, whether these transcription factors act together with AR on genuine chromatin and what transcription factors play a general collaborative role in AR-regulated endogenous target gene network remain to be determined. In search of transcription factor binding motifs enriched in genuine AR binding regions on chromosomes 21 and 22, we found that the co-occurrence of an AR halfsite with Forkhead, GATA, or Oct motifs is significantly enriched compared with the genomic background (Wang et al. 2007). A subsequent AR ChIP-on-chip study focusing on promoter regions confirmed our finding (Massie et al. 2007). Interestingly, these motifs are also enriched in ER-binding regions (Carroll et al. 2005, 2006), suggesting those transcription factors recognizing these motifs may play general cooperative roles rather than serve as specific AR collaborating factors. FoxA1, GATA2, and Oct1 were then selected as potential AR collaborating factors as they are highly expressed in prostate cancer, as reported in several human prostate cancer studies (Singh et al. 2002; Lapointe et al. 2004; Yu et al. 2004) available on the Oncomine database (Rhodes et al. 2004). FoxA1 is a member of the FoxA subfamily within the winged helix/forkhead family of transcription factors (Kaestner et al. 2000). All three members of FoxA subfamily (FoxA1, FoxA2, and FoxA3) share 95% identify within the DNAbinding domain, termed the forkhead domain or winged-helix domain that bind to the consensus sequence (G/A)A(G/A)(T/C)AAA(T/C)A (Kaufmann and Knochel 1996; Friedman and Kaestner 2006). The structure of FoxA forkhead domain is similar to that of globular domain of linker histone (Clark et al. 1993; Ramakrishnan et al. 1993). Functional studies demonstrated that FoxA proteins are able to open compact chromatin, which is mediated by high-affinity DNA-binding sites and the interactions of the C-terminal domain of FoxA with core histones H3 and H4 (Cirillo et al. 1998, 2002). Therefore, FoxA proteins have been proposed to function as pioneer factors that facilitate binding of other transcription factors. Consistent with this idea, recent ER ChIP-on-chip studies found that FoxA1 functions as an ER pioneer factor (Carroll et al. 2005; Laganiere et al. 2005). FoxA1 has also been shown to interact with AR and is required for PSA gene (Gao et al. 2003). In contrast to most transcription factor families, including the NR and Forkhead families, the vertebrate GATA family consists of only six members. GATA1, GATA2, and GATA3 constitute one subfamily that is thought to be expressed primarily in cells of the erythroid lineage, whereas GATA4, GATA5, and GATA6 form the second subfamily that is restricted to endodermally derived tissues (Lowry and Atchley 2000; Molkentin 2000; Patient and McGhee 2002). The GATA transcription factors contain two zinc-finger motifs that recognize the consensus

672

Q. Wang, M. Brown

sequence (A/T)GATA(A/G). Interestingly, the two zinc motifs of GATA mediate both transcription factor binding to and opening the compact chromatin (Boyes et al. 1998; Cirillo et al. 2002). GATA transcription factors have been shown to play essential roles in hematopoiesis, cardiovascular, neural, and urogenital development (Zhou et al. 1998; Lakshmanan et al. 1999; Molkentin 2000; Nemer and Nemer 2001; Patient and McGhee 2002). Interestingly, recent studies have found that GATA transcription factors are also involved in steroid receptor signaling pathways in hormone-related cancers. For example, GATA3 expression is highly correlated with ER expression in breast cancer (Gruvberger et al. 2001; Lacroix and Leclercq 2004). A recent study from our laboratory demonstrated that ER and GATA3 form a positive crossregulatory loop to accelerate breast cancer cell growth (Eeckhoute et al. 2007). In prostate cancer cells, GATA2 binds to the PSA enhancer in vitro and is required for androgen-induced PSA gene transcription (Perez-Stable et al. 2000). Oct1 is a member of the POU domain family of transcription factors. The POU DNA-binding domain is a bipartite domain that includes an N-terminal POUspecific domain and a C-terminal POU homeodomain. These two domains of Oct1 bind to opposite faces of an octamer motif ATGCAAAT in a tail-to-tail orientation (Verrijzer and Van der Vliet 1993; Ryan and Rosenfeld 1997). In contrast to positive regulatory roles of FoxA and GATA factors in NR function, the interaction between Oct1 can either cooperate or interfere with NR function, depending on the genespecific regulatory region. For instance, Oct1/GR and Oct1/AR synergistically bind to the mouse mammary tumor virus (MMTV) promoter and the mouse slp enhancer, respectively (Bruggemeier et al. 1991; Prefontaine et al. 1998; Gonzalez and Robins 2001). In contrast, functional interference between GR and Oct1 occurs during Oct1 binding to the histone 2B promoter (Kutoh et al. 1992). Several lines of evidence support the idea that FoxA1, GATA2, and Oct1 function as AR-collaborating partners, which globally regulate target genes in the genome (Wang et al. 2007). First, all three factors are recruited to 80% of the AR binding sites tested on chromosomes 21 and 22. Second, AR interacts with all three factors in solution and on chromatin. Third, the three factors have distinct functional roles in androgen-dependent gene expression and cell proliferation. Silencing of GATA2 and Oct1 significantly decreases AR target genes PSA, TMPRSS2, and PDE9A expression, whereas FoxA1 silencing does not affect expression of these three genes. While this may suggest that FoxA1 does not play a critical role in ARdependent gene expression, it is also possible that FoxA1 function is required for other AR targets or is redundant with other Forkhead family members. Consistent with the roles of GATA2 and Oct1 in AR-dependent transcription, silencing of these two factors significantly decreases androgen-stimulated cell cycle progression. Interestingly, while FoxA1 silencing significantly decreases estrogen-induced cell cycle progression (Laganiere et al. 2005), silencing of FoxA1 does not affect androgen-stimulated cell cycle progression, suggesting that FoxA1 is a pioneer factor for ER but not for AR at the genes critical for cell cycle progression. The observation that GATA2 and Oct1 are essential for androgen-dependent transcription and cell growth raises the question of how these two factors

Mapping the Androgen Receptor Cistrome

673

collaborate with AR. ChIP analyses found that silencing of GATA2 decreases both AR and Pol II binding to chromatin. In contrast, silencing of Oct1 does not affect AR binding but reduces Pol II binding to chromatin. In addition, Oct1 binding to chromatin is also attenuated by GATA2 silencing, whereas GATA2 loading on chromatin is not affected by Oct1 silencing (Wang et al. 2007). These findings suggest a hierarchical model of collaborating transcription factors in AR signaling: GATA2 acts as a pioneer factor that opens compact chromatin to allow AR and Oct1 binding, whereas Oct1 functions in conjunction with AR at a step subsequent to GATA2 action (Fig. 3). In addition to GATA2, Oct1, and FoxA1, a recent AR ChIP-on-chip study found that both ETS motifs and AR half-sites are present in 70% of AR-enriched

Fig. 3 Model for AR-mediated transcription. GATA2 first bind and open the compact chromatin. Subsequently, Oct1 and AR load on distal enhancers, which recruit Pol II and coactivator complexes including p160 coactivators, histone acetyltranferase (HAT), histone methyltransferase (HMT), and Mediator (Med). The distal AR-collaborating factor complex and coactivator complex communicate with the basal transcription apparatus on promoter through chromosomal looping

674

Q. Wang, M. Brown

promoters, suggesting that ETS transcription factors might also function as AR cooperative partners (Massie et al. 2007). The human ETS family of proteins comprises 27 members that share a conserved ETS DNA-binding domain. This winged helix-turn-helix DNA-binding domain recognizes a core GGAA/T sequence (ETS-Binding Site, EBS) (Wasylyk et al. 1993; Seth and Watson 2005). The ETS family members play critical roles in human carcinogenesis (Hsu et al. 2004; Seth and Watson 2005). Indeed, five general classes of chromosomal rearrangements activate ETS family members in prostate cancer (Tomlins et al. 2007). The aforementioned fusions between TMPRSS2 and ERG, ETV1, and ETV4 comprise a predominant class of rearrangements. Importantly LNCaP harbors a rearrangement resulting in the overexpression of ETV1. It has been suggested that ETS1 functions to increase AR nuclear accumulation in LNCaP cells (Massie et al. 2007).

6 Future Directions Significant progress has been made with the generation of the AR cistrome in the prostate. Future studies should map AR binding regions in the entire genome under different conditions (e.g., stimulation with various androgens and antiandrogens) and/or different cell lines (e.g., androgen-dependent and androgen-independent prostate cancer cell lines). By combining the AR cistrome with gene expression profiles under the same conditions we may begin to understand the first-order relationship of how AR regulates target gene networks. A future challenge will be to apply the genome-wide ChIP technique to clinical samples obtained from different stages of prostate cancer. This would allow identification of critical cisregulatory sequences contributing to prostate cancer progression. Thanks to recent technological advances in genome-wide ChIP assays such as ChIP combined with high throughput sequencing (ChIP-seq) (Barski et al. 2007; Johnson et al. 2007; Mikkelsen et al. 2007), it may be possible in the near future to perform genomewide ChIP assays with small amounts of DNA samples, without amplifications and thus avoid some of the inherent problems of hybridization. Finally, future studies should extend our view from transcriptional regulation by AR to wider transcriptional regulation, including AR-collaborating transcription factors, in prostate cancer. While the currently identified AR-collaborating transcription factors give us a general idea of cooperative regulation of targets between AR and other transcription factors, a major challenge ahead will be to define the cistromes of the collaborating transcription factors and integration of these data to unravel combinatorial transcriptional regulatory mechanisms. Finally a dynamic view of how occupancy of the genome varies over time will lead to an improved understanding of transcriptional regulation in prostate cancer and extend our current static snapshot of understanding to a full-length feature.

Mapping the Androgen Receptor Cistrome

675

Acknowledgments We thank Drs. Jason Carroll, Timothy Geistlinger, and Wei Li for critical reading of the manuscript.

References Balk, S.P. 2002. Androgen receptor as a target in androgen-independent prostate cancer. Urology 60: 132–8; discussion 138–9. Barettino, D., M.M. Vivanco Ruiz, and H.G. Stunnenberg. 1994. Characterization of the liganddependent transactivation domain of thyroid hormone receptor. EMBO J 13: 3039–49. Barski, A., S. Cuddapah, K. Cui, T.Y. Roh, D.E. Schones, Z. Wang, G. Wei, I. Chepelev, and K. Zhao. 2007. High-resolution profiling of histone methylations in the human genome. Cell 129: 823–37. Blackwood, E.M. and J.T. Kadonaga. 1998. Going the distance: a current view of enhancer action. Science 281: 61–3. Bolton, E.C., A.Y. So, C. Chaivorapol, C.M. Haqq, H. Li, and K.R. Yamamoto. 2007. Cell- and gene-specific regulation of primary target genes by the androgen receptor. Genes Dev 21: 2005–17. Boyes, J., P. Byfield, Y. Nakatani, and V. Ogryzko. 1998. Regulation of activity of the transcription factor GATA-1 by acetylation. Nature 396: 594–8. Bruggemeier, U., M. Kalff, S. Franke, C. Scheidereit, and M. Beato. 1991. Ubiquitous transcription factor OTF-1 mediates induction of the MMTV promoter through synergistic interaction with hormone receptors. Cell 64: 565–72. Bulger, M. and M. Groudine. 1999. Looping versus linking: toward a model for long-distance gene activation. Genes Dev 13: 2465–77. Carroll, J.S., X.S. Liu, A.S. Brodsky, W. Li, C.A. Meyer, A.J. Szary, J. Eeckhoute, W. Shao, E.V. Hestermann, T.R. Geistlinger, E.A. Fox, P.A. Silver, and M. Brown. 2005. Chromosome-wide mapping of estrogen receptor binding reveals long-range regulation requiring the forkhead protein FoxA1. Cell 122: 33–43. Carroll, J.S., C.A. Meyer, J. Song, W. Li, T.R. Geistlinger, J. Eeckhoute, A.S. Brodsky, E.K. Keeton, K.C. Fertuck, G.F. Hall, Q. Wang, S. Bekiranov, V. Sementchenko, E.A. Fox, P.A. Silver, T.R. Gingeras, X.S. Liu, and M. Brown. 2006. Genome-wide analysis of estrogen receptor binding sites. Nat Genet 38: 1289–97. Cawley, S., S. Bekiranov, H.H. Ng, P. Kapranov, E.A. Sekinger, D. Kampa, A. Piccolboni, V. Sementchenko, J. Cheng, A.J. Williams, R. Wheeler, B. Wong, J. Drenkow, M. Yamanaka, S. Patel, S. Brubaker, H. Tammana, G. Helt, K. Struhl, and T.R. Gingeras. 2004. Unbiased mapping of transcription factor binding sites along human chromosomes 21 and 22 points to widespread regulation of noncoding RNAs. Cell 116: 499–509. Cirillo, L.A., C.E. McPherson, P. Bossard, K. Stevens, S. Cherian, E.Y. Shim, K.L. Clark, S.K. Burley, and K.S. Zaret. 1998. Binding of the winged-helix transcription factor HNF3 to a linker histone site on the nucleosome. EMBO J 17: 244–54. Cirillo, L.A., F.R. Lin, I. Cuesta, D. Friedman, M. Jarnik, and K.S. Zaret. 2002. Opening of compacted chromatin by early developmental transcription factors HNF3 (FoxA) and GATA-4. Mol Cell 9: 279–89. Clark, K.L., E.D. Halay, E. Lai, and S.K. Burley. 1993. Co-crystal structure of the HNF-3/fork head DNA-recognition motif resembles histone H5. Nature 364: 412–20. Danielian, P.S., R. White, J.A. Lees, and M.G. Parker. 1992. Identification of a conserved region required for hormone dependent transcriptional activation by steroid hormone receptors [published erratum appears in EMBO J 1992 11(6): 2366]. EMBO J 11: 1025–33. Debes, J.D. and D.J. Tindall. 2004. Mechanisms of androgen-refractory prostate cancer. N Engl J Med 351: 1488–90.

676

Q. Wang, M. Brown

Dekker, J., K. Rippe, M. Dekker, and N. Kleckner. 2002. Capturing chromosome conformation. Science 295: 1306–11. DePrimo, S.E., M. Diehn, J.B. Nelson, R.E. Reiter, J. Matese, M. Fero, R. Tibshirani, P.O. Brown, and J.D. Brooks. 2002. Transcriptional programs activated by exposure of human prostate cancer cells to androgen. Genome Biol 3: RESEARCH0032 Dostie, J. and J. Dekker. 2007. Mapping networks of physical interactions between genomic elements using 5C technology. Nat Protoc 2: 988–1002. Dostie, J., T.A. Richmond, R.A. Arnaout, R.R. Selzer, W.L. Lee, T.A. Honan, E.D. Rubio, A. Krumm, J. Lamb, C. Nusbaum, R.D. Green, and J. Dekker. 2006. Chromosome Conformation Capture Carbon Copy (5C): a massively parallel solution for mapping interactions between genomic elements. Genome Res 16: 1299–309. Eder, I.E., J. Hoffmann, H. Rogatsch, G. Schafer, D. Zopf, G. Bartsch, and H. Klocker. 2002. Inhibition of LNCaP prostate tumor growth in vivo by an antisense oligonucleotide directed against the human androgen receptor. Cancer Gene Ther 9: 117–25. Eeckhoute, J., E.K. Keeton, M. Lupien, S.A. Krum, J.S. Carroll, and M. Brown. 2007. Positive cross-regulatory loop ties GATA-3 to estrogen receptor alpha expression in breast cancer. Cancer Res 67: 6477–83. Evans, R.M. 1988. The steroid and thyroid hormone receptor superfamily. Science 240: 889–95. Fisher, D.A., J.F. Smith, J.S. Pillar, S.H. St Denis, and J.B. Cheng. 1998. Isolation and characterization of PDE9A, a novel human cGMP-specific phosphodiesterase. J Biol Chem 273: 15559–64. Friedman, J.R. and K.H. Kaestner. 2006. The Foxa family of transcription factors in development and metabolism. Cell Mol Life Sci 63: 2317–28. Gao, N., J. Zhang, M.A. Rao, T.C. Case, J. Mirosevich, Y. Wang, R. Jin, A. Gupta, P.S. Rennie, and R.J. Matusik. 2003. The role of hepatocyte nuclear factor-3 alpha (Forkhead Box A1) and androgen receptor in transcriptional regulation of prostatic genes. Mol Endocrinol 17: 1484–507. Gonzalez, M.I. and D.M. Robins. 2001. Oct-1 preferentially interacts with androgen receptor in a DNA-dependent manner that facilitates recruitment of SRC-1. J Biol Chem 276: 6420–8. Green, S. and P. Chambon. 1988. Nuclear receptors enhance our understanding of transcription regulation. Trends Genet 4: 309–14. Gruvberger, S., M. Ringner, Y. Chen, S. Panavally, L.H. Saal, A. Borg, M. Ferno, C. Peterson, and P.S. Meltzer. 2001. Estrogen receptor status in breast cancer is associated with remarkably distinct gene expression patterns. Cancer Res 61: 5979–84. Guipponi, M., H.S. Scott, J. Kudoh, K. Kawasaki, K. Shibuya, A. Shintani, S. Asakawa, H. Chen, M.D. Lalioti, C. Rossier, S. Minoshima, N. Shimizu, and S.E. Antonarakis. 1998. Identification and characterization of a novel cyclic nucleotide phosphodiesterase gene (PDE9A) that maps to 21q22.3: alternative splicing of mRNA transcripts, genomic structure and sequence. Hum Genet 103: 386–92. Haag, P., J. Bektic, G. Bartsch, H. Klocker, and I.E. Eder. 2005. Androgen receptor down regulation by small interference RNA induces cell growth inhibition in androgen sensitive as well as in androgen independent prostate cancer cells. J Steroid Biochem Mol Biol 96: 251–8. Hardison, R.C. 2000. Conserved noncoding sequences are reliable guides to regulatory elements. Trends Genet 16: 369–72. Heinlein, C.A. and C. Chang. 2004. Androgen receptor in prostate cancer. Endocr Rev 25: 276–308. Hollenberg, S.M., C. Weinberger, E.S. Ong, G. Cerelli, A. Oro, R. Lebo, E.B. Thompson, M.G. Rosenfeld, and R.M. Evans. 1985. Primary structure and expression of a functional human glucocorticoid receptor cDNA. Nature 318: 635–41. Horike, S., S. Cai, M. Miyano, J.F. Cheng, and T. Kohwi-Shigematsu. 2005. Loss of silentchromatin looping and impaired imprinting of DLX5 in Rett syndrome. Nat Genet 37: 31–40. Hsu, T., M. Trojanowska, and D.K. Watson. 2004. Ets proteins in biological control and cancer. J Cell Biochem 91: 896–903.

Mapping the Androgen Receptor Cistrome

677

Jariwala, U., J. Prescott, L. Jia, A. Barski, S. Pregizer, J.P. Cogan, A. Arasheben, W.D. Tilley, H.I. Scher, W.L. Gerald, G. Buchanan, G.A. Coetzee, and B. Frenkel. 2007. Identification of novel androgen receptor target genes in prostate cancer. Mol Cancer 6: 39. Jia, L., C.S. Choong, C. Ricciardelli, J. Kim, W.D. Tilley, and G.A. Coetzee. 2004. Androgen receptor signaling: mechanism of interleukin-6 inhibition. Cancer Res 64: 2619–26. Johnson, D.S., A. Mortazavi, R.M. Myers, and B. Wold. 2007. Genome-wide mapping of in vivo protein–DNA interactions. Science 316: 1497–502. Kaestner, K.H., W. Knochel, and D.E. Martinez. 2000. Unified nomenclature for the winged helix/ forkhead transcription factors. Genes Dev 14: 142–6. Kang, Z., O.A. Janne, and J.J. Palvimo. 2004. Coregulator recruitment and histone modifications in transcriptional regulation by the androgen receptor. Mol Endocrinol 18: 2633–48. Kato, S., H. Endoh, Y. Masuhiro, T. Kitamoto, S. Uchiyama, H. Sasaki, S. Masushige, Y. Gotoh, E. Nishida, H. Kawashima, et al. 1995. Activation of the estrogen receptor through phosphorylation by mitogen-activated protein kinase. Science 270: 1491–4. Kaufmann, E. and W. Knochel. 1996. Five years on the wings of fork head. Mech Dev 57: 3–20. Kutoh, E., P.E. Stromstedt, and L. Poellinger. 1992. Functional interference between the ubiquitous and constitutive octamer transcription factor 1 (OTF-1) and the glucocorticoid receptor by direct protein–protein interaction involving the homeo subdomain of OTF-1. Mol Cell Biol 12: 4960–9. Lacroix, M. and G. Leclercq. 2004. About GATA3, HNF3A, and XBP1, three genes co-expressed with the oestrogen receptor-alpha gene (ESR1) in breast cancer. Mol Cell Endocrinol 219: 1–7. Laganiere, J., G. Deblois, C. Lefebvre, A.R. Bataille, F. Robert, and V. Giguere. 2005. From the Cover: Location analysis of estrogen receptor alpha target promoters reveals that FOXA1 defines a domain of the estrogen response. Proc Natl Acad Sci U S A 102: 11651–6. Lakshmanan, G., K.H. Lieuw, K.C. Lim, Y. Gu, F. Grosveld, J.D. Engel, and A. Karis. 1999. Localization of distant urogenital system-, central nervous system-, and endocardium-specific transcriptional regulatory elements in the GATA-3 locus. Mol Cell Biol 19: 1558–68. Lapointe, J., C. Li, J.P. Higgins, M. van de Rijn, E. Bair, K. Montgomery, M. Ferrari, L. Egevad, W. Rayford, U. Bergerheim, P. Ekman, A.M. DeMarzo, R. Tibshirani, D. Botstein, P.O. Brown, J.D. Brooks, and J.R. Pollack. 2004. Gene expression profiling identifies clinically relevant subtypes of prostate cancer. Proc Natl Acad Sci U S A 101: 811–6. Lemon, B. and R. Tjian. 2000. Orchestrated response: a symphony of transcription factors for gene control. Genes Dev 14: 2551–69. Liao, X., S. Tang, J.B. Thrasher, T.L. Griebling, and B. Li. 2005. Small-interfering RNA-induced androgen receptor silencing leads to apoptotic cell death in prostate cancer. Mol Cancer Ther 4: 505–15. Louie, M.C., H.Q. Yang, A.H. Ma, W. Xu, J.X. Zou, H.J. Kung, and H.W. Chen. 2003. Androgeninduced recruitment of RNA polymerase II to a nuclear receptor-p160 coactivator complex. Proc Natl Acad Sci U S A 100: 2226–30. Lowry, J.A. and W.R. Atchley. 2000. Molecular evolution of the GATA family of transcription factors: conservation within the DNA-binding domain. J Mol Evol 50: 103–15. Luisi, B.F., W.X. Xu, Z. Otwinowski, L.P. Freedman, K.R. Yamamoto, and P.B. Sigler. 1991. Crystallographic analysis of the interaction of the glucocorticoid receptor with DNA. Nature 352: 497–505. Mangelsdorf, D.J. and R.M. Evans. 1995. The RXR heterodimers and orphan receptors. Cell 83: 841–50. Mangelsdorf, D.J., C. Thummel, M. Beato, P. Herrlich, G. Schutz, K. Umesono, B. Blumberg, P. Kastner, M. Mark, P. Chambon, et alet al. 1995. The nuclear receptor superfamily: the second decade. Cell 83: 835–9. Massie, C.E., B. Adryan, N.L. Barbosa-Morais, A.G. Lynch, M.G. Tran, D.E. Neal, and I.G. Mills. 2007. New androgen receptor genomic targets show an interaction with the ETS1 transcription factor. EMBO Rep 8: 871–8.

678

Q. Wang, M. Brown

McKenna, N.J. and B.W. O’Malley. 2002. Combinatorial control of gene expression by nuclear receptors and coregulators. Cell 108: 465–74. Miesfeld, R., S. Rusconi, P.J. Godowski, B.A. Maler, S. Okret, A.C. Wikstrom, J.A. Gustafsson, and K.R. Yamamoto. 1986. Genetic complementation of a glucocorticoid receptor deficiency by expression of cloned receptor cDNA. Cell 46: 389–99. Mikkelsen, T.S., M. Ku, D.B. Jaffe, B. Issac, E. Lieberman, G. Giannoukos, P. Alvarez, W. Brockman, T.K. Kim, R.P. Koche, W. Lee, E. Mendenhall, A. O’Donovan, A. Presser, C. Russ, X. Xie, A. Meissner, M. Wernig, R. Jaenisch, C. Nusbaum, E.S. Lander, and B.E. Bernstein. 2007. Genome-wide maps of chromatin state in pluripotent and lineagecommitted cells. Nature 448: 553–60. Molkentin, J.D. 2000. The zinc finger-containing transcription factors GATA-4, -5, and -6. Ubiquitously expressed regulators of tissue-specific gene expression. J Biol Chem 275: 38949–52. Nagpal, S., M. Saunders, P. Kastner, B. Durand, H. Nakshatri, and P. Chambon. 1992. Promoter context- and response element-dependent specificity of the transcriptional activation and modulating functions of retinoic acid receptors. Cell 70: 1007–19. Nelson, P.S., N. Clegg, H. Arnold, C. Ferguson, M. Bonham, J. White, L. Hood, and B. Lin. 2002. The program of androgen-responsive genes in neoplastic prostate epithelium. Proc Natl Acad Sci U S A 99: 11890–5. Nemer, G. and M. Nemer. 2001. Regulation of heart development and function through combinatorial interactions of transcription factors. Ann Med 33: 604–10. Oettgen, P., E. Finger, Z. Sun, Y. Akbarali, U. Thamrongsak, J. Boltax, F. Grall, A. Dube, A. Weiss, L. Brown, G. Quinn, K. Kas, G. Endress, C. Kunsch, and T.A. Libermann. 2000. PDEF, a novel prostate epithelium-specific ets transcription factor, interacts with the androgen receptor and activates prostate-specific antigen gene expression. J Biol Chem 275: 1216–25. Palvimo, J.J., P. Reinikainen, T. Ikonen, P.J. Kallio, A. Moilanen, and O.A. Janne. 1996. Mutual transcriptional interference between RelA and androgen receptor. J Biol Chem 271: 24151–6. Parker, M.G. 1993. Steroid and related receptors. Curr Opin Cell Biol 5: 499–504. Patient, R.K. and J.D. McGhee. 2002. The GATA family (vertebrates and invertebrates). Curr Opin Genet Dev 12: 416–22. Perez-Stable, C.M., A. Pozas, and B.A. Roos. 2000. A role for GATA transcription factors in the androgen regulation of the prostate-specific antigen gene enhancer. Mol Cell Endocrinol 167: 43–53. Perissi, V. and M.G. Rosenfeld. 2005. Controlling nuclear receptors: the circular logic of cofactor cycles. Nat Rev Mol Cell Biol 6: 542–54. Prefontaine, G.G., M.E. Lemieux, W. Giffin, C. Schild-Poulter, L. Pope, E. LaCasse, P. Walker, and R.J. Hache. 1998. Recruitment of octamer transcription factors to DNA by glucocorticoid receptor. Mol Cell Biol 18: 3416–30. Ramakrishnan, V., J.T. Finch, V. Graziano, P.L. Lee, and R.M. Sweet. 1993. Crystal structure of globular domain of histone H5 and its implications for nucleosome binding. Nature 362: 219–23. Remenyi, A., H.R. Scholer, and M. Wilmanns. 2004. Combinatorial control of gene expression. Nat Struct Mol Biol 11: 812–5. Rhodes, D.R., J. Yu, K. Shanker, N. Deshpande, R. Varambally, D. Ghosh, T. Barrette, A. Pandey, and A.M. Chinnaiyan. 2004. ONCOMINE: a cancer microarray database and integrated datamining platform. Neoplasia 6: 1–6. Ryan, A.K. and M.G. Rosenfeld. 1997. POU domain family values: flexibility, partnerships, and developmental codes. Genes Dev 11: 1207–25. Sato, N., M.D. Sadar, N. Bruchovsky, F. Saatcioglu, P.S. Rennie, S. Sato, P.H. Lange, and M.E. Gleave. 1997. Androgenic induction of prostate-specific antigen gene is repressed by protein–protein interaction between the androgen receptor and AP-1/c-Jun in the human prostate cancer cell line LNCaP. J Biol Chem 272: 17485–94.

Mapping the Androgen Receptor Cistrome

679

Scher, H.I. and C.L. Sawyers. 2005. Biology of progressive, castration-resistant prostate cancer: directed therapies targeting the androgen-receptor signaling axis. J Clin Oncol 23: 8253–61. Segawa, T., M.E. Nau, L.L. Xu, R.N. Chilukuri, M. Makarem, W. Zhang, G. Petrovics, I.A. Sesterhenn, D.G. McLeod, J.W. Moul, M. Vahey, and S. Srivastava. 2002. Androgen-induced expression of endoplasmic reticulum (ER) stress response genes in prostate cancer cells. Oncogene 21: 8749–58. Seth, A. and D.K. Watson. 2005. ETS transcription factors and their emerging roles in human cancer. Eur J Cancer 41: 2462–78. Shaffer, P.L., A. Jivan, D.E. Dollins, F. Claessens, and D.T. Gewirth. 2004. Structural basis of androgen receptor binding to selective androgen response elements. Proc Natl Acad Sci U S A 101: 4758–63. Shang, Y., M. Myers, and M. Brown. 2002. Formation of the androgen receptor transcription complex. Mol Cell 9: 601–10. Singh, D., P.G. Febbo, K. Ross, D.G. Jackson, J. Manola, C. Ladd, P. Tamayo, A.A. Renshaw, A.V. D’Amico, J.P. Richie, E.S. Lander, M. Loda, P.W. Kantoff, T.R. Golub, and W.R. Sellers. 2002. Gene expression correlates of clinical prostate cancer behavior. Cancer Cell 1: 203–9. So, A.Y., C. Chaivorapol, E.C. Bolton, H. Li, and K.R. Yamamoto. 2007. Determinants of celland gene-specific transcriptional regulation by the glucocorticoid receptor. PLoS Genet 3: e94. Takayama, K., K. Kaneshiro, S. Tsutsumi, K. Horie-Inoue, K. Ikeda, T. Urano, N. Ijichi, Y. Ouchi, K. Shirahige, H. Aburatani, and S. Inoue. 2007. Identification of novel androgen response genes in prostate cancer cells by coupling chromatin immunoprecipitation and genomic microarray analysis. Oncogene 26: 4453–63. Tomlins, S.A., D.R. Rhodes, S. Perner, S.M. Dhanasekaran, R. Mehra, X.W. Sun, S. Varambally, X. Cao, J. Tchinda, R. Kuefer, C. Lee, J.E. Montie, R.B. Shah, K.J. Pienta, M.A. Rubin, and A.M. Chinnaiyan. 2005. Recurrent fusion of TMPRSS2 and ETS transcription factor genes in prostate cancer. Science 310: 644–8. Tomlins, S.A., R. Mehra, D.R. Rhodes, L.R. Smith, D. Roulston, B.E. Helgeson, X. Cao, J.T. Wei, M.A. Rubin, R.B. Shah, and A.M. Chinnaiyan. 2006. TMPRSS2:ETV4 gene fusions define a third molecular subtype of prostate cancer. Cancer Res 66: 3396–400. Tomlins, S.A., B. Laxman, S.M. Dhanasekaran, B.E. Helgeson, X. Cao, D.S. Morris, A. Menon, X. Jing, Q. Cao, B. Han, J. Yu, L. Wang, J.E. Montie, M.A. Rubin, K.J. Pienta, D. Roulston, R.B. Shah, S. Varambally, R. Mehra, and A.M. Chinnaiyan. 2007. Distinct classes of chromosomal rearrangements create oncogenic ETS gene fusions in prostate cancer. Nature 448: 595–9. Tsai, M.J. and B.W. O’Malley. 1994. Molecular mechanisms of action of steroid/thyroid receptor superfamily members. Annu Rev Biochem 63: 451–86. van Steensel, B. 2005. Mapping of genetic and epigenetic regulatory networks using microarrays. Nat Genet 37 Suppl: S18–24. Velasco, A.M., K.A. Gillis, Y. Li, E.L. Brown, T.M. Sadler, M. Achilleos, L.M. Greenberger, P. Frost, W. Bai, and Y. Zhang. 2004. Identification and validation of novel androgen-regulated genes in prostate cancer. Endocrinology 145: 3913–24. Verrijdt, G., A. Haelens, and F. Claessens. 2003. Selective DNA recognition by the androgen receptor as a mechanism for hormone-specific regulation of gene expression. Mol Genet Metab 78: 175–85. Verrijdt, G., T. Tanner, U. Moehren, L. Callewaert, A. Haelens, and F. Claessens. 2006. The androgen receptor DNA-binding domain determines androgen selectivity of transcriptional response. Biochem Soc Trans 34: 1089–94. Verrijzer, C.P. and P.C. Van der Vliet. 1993. POU domain transcription factors. Biochim Biophys Acta 1173: 1–21. Wang, Q., D. Sharma, Y. Ren, and J.D. Fondell. 2002. A coregulatory role for the TRAP-mediator complex in androgen receptor-mediated gene expression. J Biol Chem 277: 42852–8. Wang, Q., J.S. Carroll, and M. Brown. 2005. Spatial and temporal recruitment of androgen receptor and its coactivators involves chromosomal looping and polymerase tracking. Mol Cell 19: 631–42.

680

Q. Wang, M. Brown

Wang, Q., W. Li, X.S. Liu, J.S. Carroll, O.A. Janne, E.K. Keeton, A.M. Chinnaiyan, K.J. Pienta, and M. Brown. 2007. A hierarchical network of transcription factors governs androgen receptor-dependent prostate cancer growth. Mol Cell 27: 380–92. Wasylyk, B., S.L. Hahn, and A. Giovane. 1993. The Ets family of transcription factors. Eur J Biochem 211: 7–18. Wei, C.L., Q. Wu, V.B. Vega, K.P. Chiu, P. Ng, T. Zhang, A. Shahab, H.C. Yong, Y. Fu, Z. Weng, J. Liu, X.D. Zhao, J.L. Chew, Y.L. Lee, V.A. Kuznetsov, W.K. Sung, L.D. Miller, B. Lim, E.T. Liu, Q. Yu, H.H. Ng, and Y. Ruan. 2006. A global map of p53 transcription-factor binding sites in the human genome. Cell 124: 207–19. Wright, M.E., M.J. Tsai, and R. Aebersold. 2003. Androgen receptor represses the neuroendocrine transdifferentiation process in prostate cancer cells. Mol Endocrinol 17: 1726–37. Yu, Y.P., D. Landsittel, L. Jing, J. Nelson, B. Ren, L. Liu, C. McDonald, R. Thomas, R. Dhir, S. Finkelstein, G. Michalopoulos, M. Becich, and J.H. Luo. 2004. Gene expression alterations in prostate cancer predicting tumor aggression and preceding development of malignancy. J Clin Oncol 22: 2790–9. Yuan, X., M.L. Lu, T. Li, and S.P. Balk. 2001. SRY interacts with and negatively regulates androgen receptor transcriptional activity. J Biol Chem 276: 46647–54. Yuan, X., T. Li, H. Wang, T. Zhang, M. Barua, R.A. Borgesi, G.J. Bubley, M.L. Lu, and S.P. Balk. 2006. Androgen receptor remains critical for cell-cycle progression in androgen-independent CWR22 prostate cancer cells. Am J Pathol 169: 682–96. Zegarra-Moro, O.L., L.J. Schmidt, H. Huang, and D.J. Tindall. 2002. Disruption of androgen receptor function inhibits proliferation of androgen-refractory prostate cancer cells. Cancer Res 62: 1008–13. Zhou, Y., K.C. Lim, K. Onodera, S. Takahashi, J. Ohta, N. Minegishi, F.Y. Tsai, S.H. Orkin, M. Yamamoto, and J.D. Engel. 1998. Rescue of the embryonic lethal hematopoietic defect reveals a critical role for GATA-2 in urogenital development. EMBO J 17: 6689–700.

Differential Regulation of Clusterin Isoforms by the Androgen Receptor Tanya K. Day, Colleen C. Nelson, and Martin E. Gleave

Abstract Clusterin (CLU) was initially reported as an androgen-repressed gene which is now shown to be an androgen-regulated ATP-independent cytoprotective molecular chaperone. CLU binds to a wide variety of client proteins to potently inhibit stress-induced protein aggregation and chaperone or stabilise conformations of proteins at times of cell stress. CLU is an enigmatic protein, being ascribed both pro- and anti-apoptotic roles. Recent evidence has shown that both secreted (sCLU) and nuclear (nCLU) isoforms can be produced, and that protein function is dependent on the sub-cellular localisation. We and others have shown that sCLU is cytoprotective, while nCLU is pro-apoptotic. It now seems likely that the apparently dichotomous functions of CLU result from the expression of different but related CLU isoforms and splice variants, and that cell survival depends in part on the relative expression of pro- versus anti-apoptotic CLU proteins. In cancer cells, increased sCLU expression is associated with increased resistance to apoptotic triggers and treatment resistance. CLU is a stress-induced protein upregulated after apoptotic triggers like androgen ablation and chemotherapy. Treatment strategies targeting stress-associated increases in sCLU expression enhance treatmentinduced apoptosis and delay the emergence of androgen independence. Differential regulation of CLU isoforms and splice variants by androgens may be a pathway whereby cancer cells develop treatment resistance and evade apoptosis.

1 Introduction Clusterin (CLU) is a highly conserved gene and protein which is expressed almost ubiquitously in tissues and bodily secretions. CLU was originally isolated from ram rete testis fluid where it was found to act as a cell-aggregating factor and promote clustering of suspensions of Sertoli cells (Blaschuk et al. 1983; Fritz et al. 1983). CLU plays a role in carcinogenesis and tumour progression through its many M.E. Gleave(*) The Prostate Centre, Vancouver General Hospital; Department of Urologic Sciences, Vancouver, BC, Canada V6H 3Z6, E-mail: [email protected]

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4_29, # Springer Science + Business Media, LLC 2009

681

682

T.K. Day et al.

intra-cellular activities, including those related to lipid transportation, tissue remodelling, cell adhesion, cell-cycle regulation, DNA repair and suppression of apoptotic cell death. CLU also functions as an extracellular molecular chaperone involved in stress-related responses. Due to the many identified biological functions of CLU, it has many names, including Apolipoprotein J (ApoJ), testosteronerepressed prostate message 2 (TRPM-2), sulphated glycoprotein 2 (SGP-2), or serum protein-40, 40 (SP-40, 40). CLU was initially identified as an apoptosis-associated protein and has now been ascribed both pro- and anti-apoptotic functions (Wong et al. 1993). These apparently dichotomous functions may result from separate but related CLU isoforms or splice variants. The identification of new CLU isoforms and splice variants remains an emerging area, and so we have recently proposed a unifying nomenclature for CLU based on exon usage and expression in human cancer cells (Cochrane et al. 2007). This nomenclature will aid in clarification of expression and functions of individual CLU isoforms and splice variants. It now seems likely that control of apoptosis in human cells is related to the relative expression levels of pro- and anti-apoptotic CLU isoforms and splice variants. Future investigations into the mechanisms underlying the control of CLU expression are required to fully understand and utilise its role in healthy and disease states in humans. The progression to androgen independence in prostate cancer cells is an important pathway leading to treatment resistance, and novel therapeutic strategies targeting the molecular basis of androgen resistance are required to improve survival for men with hormone refractory prostate cancer. Resistance to androgen-withdrawal induced cell death is mediated in part by the adaptive up-regulation of anti-apoptotic pathways and/or proteins. A potential treatment strategy would utilise agents targeting stress-associated increases in gene expression precipitated by androgen withdrawal to enhance treatment-induced apoptosis and delay emergence of androgen independence. CLU is a stress-induced survival protein which is upregulated after apoptotic triggers like androgen ablation. Targeted CLU antisense inhibitors (OGX-011, OncoGenex Technologies Inc.) are currently progressing through clinical trials for use in prostate cancer treatment, and recent evidence has shown that OGX-011 down-regulates expression of the anti-apoptotic secreted CLU isoforms and promotes expression of the pro-apoptotic CLU splice variants, to enhance cancer cell apoptotic rates. Regulation of CLU alternative splicing is one of the underlying mechanisms of action of the CLU antisense inhibitor, OGX-011. The related regulation of CLU isoforms and splice variants by androgens is therefore an important area of research aimed at further elucidating the pathways to androgen independence in prostate cancer cells.

2 CLU Protein Structure CLU exhibits a high degree of conservation across species, with 70–80% conservation amongst mammals at the amino acid (aa) level. The crystal structure of CLU remains undetermined due to the difficulty of crystallising this inherently ‘sticky’

Differential Regulation of Clusterin Isoforms by the Androgen Receptor

683

protein. CLU is a highly processed protein which is present in both secreted and intracellular forms and appears as a range of sizes on western blots. The predominant CLU protein (isoform 2, NCBI Accession NM_203339, CLU-2) is an antiapoptotic secreted glycosylated heterodimer with an apparent molecular weight of 75–80 kDa. Essentially all published data relate to this isoform, which is translated from an AUG in exon 2, and is synthesised as a 449-aa chain which is processed in the endoplasmic reticulum (ER) to a high-mannose form (60 kDa) that remains cytoplasmic (Shannan et al. 2006). Subsequent proteolytic cleavage of the 449 aa peptide between amino acids 22/23 removes the signal peptide, which is essential for correct processing and function of CLU. Absence of the signal peptide results in intracellular CLU accumulation and renders cells susceptible to apoptosis (Zhang et al. 2006). Cleavage between residues 227/228 generates the a and b chains, which are similarly sized and approximately 40 kDa depending on the degree of glycosylation (Suuronen et al. 2007). During transport to the golgi, the a and b chains are glycosylated at six sites (three on each chain), with the attached N-linked carbohydrate comprising approximately 30% of the mass of the mature CLU protein (Kapron et al. 2000). The a and b chains are then assembled in anti-parallel and linked by five disulphide bridges near the cysteine-rich centre to form a heterodimer (Montpetit et al. 1986). The disulphide bridges are flanked by two predicted coiled–coil a-helices and three predicted amphipathic a-helices. nCLU was initially reported as CLU-24II in association with MCF-7 breast cancer cell death to be a pro-apoptotic CLU protein. This protein (CLU-24II) was truncated at the N terminus and was reported to be translated from an alternatively spliced CLU transcript utilising an in-frame AUG codon in exon 3 which would result in a 416-aa primary transcript (Yang et al. 1999, 2000; Leskov et al. 2003). At the protein level, nCLU was identified as a 49-kDa primary product which is inactive and accumulates in the cytoplasm. Some stressors trigger translocation of a mature 55-kDa protein to the nucleus where it accumulates and functions as a proapoptotic molecule (Yang et al. 2000; Leskov et al. 2003). nCLU-24II has disulphide links but the number and location remains to be determined. Analysis of recent updates to the human genomic databases identified an additional in-frame AUG codon in exon 1, unique to human, which is the first translation initiation site in CLU. Translation from the exon 1 AUG results in production of CLU isoform 1 (sCLU-1, NCBI Accession NM_001831, 501 aa), which has an additional 52 N-terminal amino acids compared to sCLU-2. The exon 1 sequence was analysed for consensus functional sequence motifs to aid predictions of the function, localisation or biological effects of CLU-1 (MotifFinder, http://motif. genome.jp). No common motifs were found, including nuclear localisation signals or endoplasmic reticulum signal peptide sequences. Therefore, the functional significance of the inclusion of exon 1 in the CLU-1 protein remains to be determined experimentally. CLU-1 also provides the possibility of a novel splice variant (CLU-14II, 459 aa) translated from the first in-frame AUG codon in exon 1 and spliced to exon 3, rather than a nuclear variant in which translation is initiated in exon 3, as previously reported (Leskov et al. 2003). Until recently, these CLU isoforms and splice variants have not been well characterised in human cells at the

684

T.K. Day et al.

Fig. 1 Effects of transient transfection of CLU-1 and CLU-14II. Mammalian expression vectors encoding each CLU isoform and splice variant were transiently transfected into LNCaP cells. An expression vector which encodes CLU isoform 1 with point mutations of the translation initiation sites in exon II was designated as CLU-1MUT. Seventy-two hours after transfection, whole cell lysates (a) and culture media (b) were harvested and analysed by Western blot using a CLU antibody. Vinculin and ponsouau red were used to demonstrate equal loading for cell lysate and cell media, respectively

molecular and functional levels. We recently demonstrated that at the protein level, the mature glycosylated protein is identical for sCLU-1 and sCLU-2 on a western blot (Fig. 1). We hypothesise that the ER-targeting signal in exon 2, which is present in both isoforms, results in localisation of both CLU-1 and CLU-2 to the ER, where the signal sequence is cleaved and degraded. The resultant cleavage, disulphide bonding, glycosylation and secretion then proceeds identically for the two proteins. We have also demonstrated that in human cancer cells, the CLU-14II splice variant is nuclear and pro-apoptotic, and that the CLU-24II splice variant is not expressed under stressed or unstressed conditions in human cancer cells.

3 CLU Genomic Structure CLU is a single copy gene which was localised by Southern analysis to 8p21-p12 in humans (Fink et al. 1993; Wong et al. 1994), chromosome 15 in rat and chromosome 14 in mouse. CLU is divided into nine exons and spans 17 kb in human (Wong et al. 1994), 14 kb in rat (Wong et al. 1993) and approximately 23 kb in mouse (Jordan-Starck et al. 1994). In all three species, the exon sizes are similar, and in humans the sizes range from 47 bp (exon 1) to 412 bp (exon 5). mRNA transcription from the CLU gene results in a major product of approximately 1.6-1.9 kb in all

Differential Regulation of Clusterin Isoforms by the Androgen Receptor

685

Fig. 2 The potential mRNA transcripts of CLU formed from alternate translational initiation sites and alternative splicing. Clusterin has nine exons, with in-frame translation initiation sites in exons I, II and III (filled triangle). Alternative splicing of exon II gives rise to two potential splice variants in addition to the two isoforms identified in the NCBI database (NM_001831, sCLU-1 and NM_203339, sCLU-2). CLU-14II is the splice variant derived from CLU-1, and CLU-24II is derived from CLU-2. There is an ER-targeting signal in exon II (vertical stripes), and nuclear localisation signals in exons 3, 7 and 8/9 (diagonal stripes)

species (de Silva et al. 1990; French et al. 1993; Tschopp et al. 1993; Wong et al. 1993). Early studies in mouse and rat identified the start of CLU translation in exon 2 (de Silva et al. 1990). More recent studies have identified an additional upstream in-frame AUG in exon 1 which is unique to human and primates (Cochrane et al. 2007) (Fig. 2) and allows the possibility of a second transcriptional isoform of human CLU (Isoform 1, NM_001831; Isoform 2, NM_203339 in GenBankE). The two isoforms result in the production of proteins with different N termini. In rodents and non-primate species, only the equivalent of sCLU-2 can be translated; hence, the only possible nuclear splice variant is CLU-24II. In primates, the nuclear CLU isoforms are generated as a result of alternative splicing. The alternative splice sites are not conserved in rodents (Shannan et al. 2006). The mechanism underlying the production of nCLU in rodents is currently unknown, but is likely to involve either differential proteolytic processing or alternative initiation of translation. Human CLU has an in-frame AUG in each of the first three exons, which allows for multiple splice variants and isoforms lacking exon 2. None of these splice variants will include the ER-targeting signal located in exon 2 (Fig. 2), and so they will not be processed in the ER or transported to the Golgi. These splice variants are not expected to undergo a and b cleavage or be subjected to the same

686

T.K. Day et al.

high degree of glycosylation as the isoforms containing exon 2. All isoforms and splice variants identified to date have three nuclear localisation sequences (NLS) (Fig. 2). Analysis of the relative nuclear localisation capacity conferred by the 50 and the 30 NLS using small CLU fragments in MCF-7 human breast cancer cells showed that the 30 NLS acts as a stronger NLS (Leskov et al. 2003). The role of the central NLS, and the interaction between other protein motifs present in each isoform and splice variant on nuclear localisation and post-translational processing remains to be determined.

4 CLU Promoter Analysis Several regulatory elements have been identified in the CLU genomic promoter region. Of particular relevance to investigating the androgen responsiveness of CLU transcription, eight consensus glucocorticoid/androgen response element half-sites were found in the first intron of rat clusterin (Rosemblit and Chen 1994). The human promoter region encompassing the first intron and part of the first exon of CLU has similar regulatory elements, including multiple consensus glucocorticoid/androgen response element half-sites (Wong et al. 1994). More recently, the human CLU promoter (intron 1) has been analysed using Consite software, which is a reliable predictor of validated AREs (Yeung et al. 2003). A cluster of AREs was predicted between +2,595 and +3,211 (numbering from exon 1 AUG), and it was demonstrated that this region was responsive to androgens in prostate cancer cells in vitro (Fig. 3) (Cochrane et al. 2007). Furthermore, it was shown that ligand activated androgen receptor bound to the promoter region between +2,443 and +2,761, which contained the predicted AREs. This study was the first to identify putative ARE sites in the CLU promoter in the context of the two CLU isoforms. The authors recognised that the promoter region was between the transcriptional start sites for the two CLU isoforms, which raised the possibility of differential regulation of the two CLU isoforms by androgens (Cochrane et al. 2007). The genomic context can play an important role in the regulation of gene expression through epigenetic factors such as DNA methylation and chromatin winding and unwinding. In gene promoter regions, hypermethylation of cytosine residues in CpG dinucleotides can lead to silencing of gene expression. Epigenetic factors such as DNA methylation are now thought to play a role in regulating CLU expression. There is a CpG island in the rat CLU gene between positions 534 and 99, which is hypermethylated in rat tissues with low levels of expression such as prostate, liver and lung, compared with tissues showing high levels of CLU expression such as testis or epididymus (Rosemblit and Chen 1994). In human, a similar CpG island exists in the CLU promoter region. Both DNA methylation and histone acetylation, common epigenetic regulatory mechanisms, are involved in regulation of CLU expression in human retinal pigment epithelial cells (Suuronen et al. 2007). Epigenetic alterations including DNA hypomethylation and increased

Differential Regulation of Clusterin Isoforms by the Androgen Receptor

687

Fig. 3 Ligand-activated AR interacts with an enhancer region in human CLU intron 1 to increase transcription of CLU-2 but not CLU-1 (adapted from Cochrane et al. 2007). (a). Partial human CLU gene showing the first two exons (black) and the intervening intron, with the predicted ARE region (dark grey) and the region tested for AR binding by ChIP assay (light grey). (b). Chromatin immunoprecipitation (ChIP) assay shows that the AR binds in the first intron of the CLU gene, between the region +2,443 to +2,761. PCR products from ChIP assay. LNCaP prostate cancer cells were treated for 48 h, with or without 1 nM R1881. Immunoprecipitation using rabbit serum was a negative control showing background levels of the PCR products. (c). Quantitative real-time RTPCR shows that at the mRNA level, CLU-1 is downregulated by androgens whereas CLU-2 is upregulated by androgens in LNCaP cells

histone acetylation together increased expression of the CLU protein precursor (sCLU-2) and the cleaved b chain, although secretion rather than intracellular accumulation of the b chain was promoted by DNA hypomethylation. In breast cancer cells, clusterin is regulated transcriptionally and post-transcriptionally by histone deacetylases (Ranney et al. 2007). It remains possible that epigenetic regulation of CLU may contribute to differential expression of pro- and antiapoptotic CLU isoforms, but this is yet to be determined experimentally.

5 CLU Biological Functions CLU expression is associated with normal and disease states, ranging from amyloidosis and other neurodegenerative diseases such as Alzheimer’s Disease (reviewed in Calero et al. 2000) to mammary (Itahana et al. 2007) and prostate gland involution (Montpetit et al. 1986) and treatment resistance in cancer

688

T.K. Day et al.

(Montpetit et al. 1986; Park et al. 2006). CLU expression is transcriptionally activated by stress-induced binding of heat-shock factor (HSF)-1 to a highly conserved 14-bp element in the CLU promoter (Michel et al. 1995, 1997) A diverse range of stressors therefore result in increased CLU expression, including cytotoxic chemotherapy (Miyake et al. 2000b), radiation (Zellweger et al. 2002), heat shock (Michel et al. 1997), and androgen (Bruchovsky et al. 1990; Miyake et al. 2000c) or oestrogen (Kyprianou et al. 1991) withdrawal in hormone-dependent tumours. Interestingly, although the role of HSF in CLU transcriptional activation was discovered for the cyto-protective, secreted CLU isoforms, recent evidence now also shows accumulation of nuclear CLU after lethal heat shock (Caccamo et al. 2006). Further study of the HSF-binding sites under various stressor conditions is required to elucidate potential specificity of HSF-mediated transcription for each CLU isoform and splice variant.

5.1

Secreted CLU

Both CLU-1 and CLU-2 isoforms are secreted and their expression exerts antiapoptotic effects on human cells (Cochrane et al. 2007). Virtually all published data refer to sCLU-2, which is a potent inhibitor of stress-induced protein aggregation (Jones and Jomary 2002) and exists in the extracellular matrix as a heterogeneous aggregate, similar to other heat-shock proteins which aggregate intracellularly. In the extracellular matrix, CLU is involved in the clearance of toxic substances by binding to unfolded proteins, cell debris, or immune complexes and preventing their uptake into otherwise healthy cells. CLU shares several common features with heat-shock proteins, including the abilities to interact with partially folded stressed proteins and to interact with a wide range of partner molecules (Poon et al. 2000). The increased expression of CLU in dying tissues implicated CLU as an apoptosisassociated protein, to the extent that now CLU is an accepted cell death marker (Connor et al. 1991; French et al. 1992; Koch-Brandt and Morgans 1996; Schwochau et al. 1998; Yang et al. 2000). Full-length intracellular CLU exerts its antiapoptotic effects in part by binding to conformation altered Bax in the cytosol which inhibits oligomerisation of activated Bax, cytochrome c release and apoptosis in response to apoptotic stimuli (Zhang et al. 2005).

5.2

Nuclear CLU

Nuclear CLU was identified as a stress-induced, pro-apoptotic CLU protein (Reddy et al. 1996; Yang et al. 1999, 2000). Subsequent studies have demonstrated a proapoptotic role in prostate epithelial cells (Caccamo et al. 2003, 2004, 2005) and shown that nCLU can also induce cell-cycle arrest (Scaltriti et al. 2004). Initial reports proposed that nCLU was formed by alternative splicing of CLU-2, with

Differential Regulation of Clusterin Isoforms by the Androgen Receptor

689

translation initiated in exon 3 (CLU-24II) (Leskov et al. 2003). Other mechanisms have since been proposed for the formation of nCLU, including alternative initiation of translation (Moretti et al. 2007). Our analysis of the CLU human gene structure now denotes that the pro-apoptotic properties of CLU in human cancer cells are due to expression of a splice variant, CLU-14II, which lacks exon 2 and localises to the nucleus. Accumulating evidence suggests that nuclear localisation is essential for the anti-proliferative and pro-apoptotic functions of the nuclear form of CLU (Leskov et al. 2003; Scaltriti et al. 2004; Moretti et al. 2007). Nuclear CLU binds to Ku70 (Yang et al. 1999), a DNA double-strand break repair protein, suggesting a potential role for nCLU in DNA repair. However, DNA repair is an early event, and the nCLU-Ku70 interaction occurs at a time after DNA repair has occurred (Yang et al. 2000), providing evidence against a direct role for nCLU is DNA repair. Ku70 binds to inactive Bax in the cytoplasm, preventing Bax activation in response to apoptotic stimuli (Gomez et al. 2007). One hypothesis linking nCLU to its proapoptotic functions is that nCLU-Ku70 binding reduces Ku70-Bax binding, allowing cells to undergo apoptosis in response to pro-apoptotic stimuli.

5.3

CLU and Cancer

sCLU-2 is highly expressed and associated with progression of many types of cancer, including bladder (Miyake et al. 2001), colon (Pucci et al. 2004), prostate (Miyake et al. 2000c, 2004; Zellweger et al. 2002), breast (Redondo et al. 2000), ovarian (Hough et al. 2001), lung (July et al. 2004), and kidney (Hara et al. 2001). Overexpression of sCLU-2 in LNCaP prostate cancer cells enhances resistance to tumour necrosis factor-a induced apoptosis (Sensibar et al. 1995). In prostate epithelial cells, sCLU-2 inhibits cell-cycle progression (Bettuzzi et al. 2002), and in prostate cancer cells, sCLU-2 plays a cytoprotective role. Pre-clinical studies demonstrate that its expression is associated with progression to androgen-independent (AI) tumour growth, enhanced cell survival and chemoresistant phenotypes (Sensibar et al. 1995; Sintich et al. 1999; Miyake et al. 2000b, c). Higher sCLU-2 expression is correlated with higher Gleason grade in prostate cancer (Steinberg et al. 1997), renders prostate cancer cells more resistant to androgen ablation, and helps to mediate progression to androgen independence after castration (Miyake et al. 2000c). Together, this suggests that sCLU-2 may play a role in tumour aggressiveness. Preclinical studies demonstrate that antisense oligonucleotide (ASO) or siRNA induced sCLU-2 knockdown enhances treatment-induced apoptosis in many varied cancer models (Zellweger et al. 2001, 2002; Gleave et al. 2003; July et al. 2004; Trougakos et al. 2004; Cao et al. 2005; So et al. 2005), marking sCLU-2 as an anticancer target. A CLU inhibitor (OGX-011) is currently in phase II clinical trials in prostate, breast and lung cancer (Chi et al. 2005).

690

T.K. Day et al.

6 Animal Models of Clusterin Roles for CLU in disease states such as cancer have been experimentally established using in vitro and other test systems. Targeted gene disruption technology was used to generate CLU-deficient mice for study of the function of CLU in normal tissue growth, development and maintenance. CLU knockout mice were phenotypically normal and their major organs also appeared morphologically normal (McLaughlin et al. 2000). Aging CLU-deficient mice develop progressive glomerulopathy characterised by the build up of immune complexes in the mesangium (Rosenberg et al. 2002). Studies utilising CLU knockout mice have confirmed a neurophysiological role for CLU in vivo. CLU knockout mice have more motor neuron cell death after axotomy than CLU wild-type mice (Wicher and Aldskogius 2005). CLU-deficient mice also have less brain tissue loss than CLU wild-type mice after neonatal hypoxic-ischaemic brain injury. Morphologically, neuronal death after hypoxic-ischaemic injury in CLU-deficient cells was via a necrotic pathway which by immunostaining did not involve the caspase-3 apoptosis pathway (Han et al. 2001). These studies provide evidence that multiple factors, including the type of stressor, are important in determining if CLU promotes cell survival or cell death. In a mouse model of Alzheimer’s disease, CLU knockout mice also had less fibrillar ab (amyloid) deposits and reduced neuritic dystrophy and toxicity associated with these deposits than wild-type mice (DeMattos et al. 2002). Young male and female CLU knockout mice were fertile and females delivered normal-sized litters (McLaughlin et al. 2000), and spermatogenesis was normal in males, except for some incomplete spermiation after stage VIII of spermatogenesis (Bailey et al. 2002). However, in testis, both apoptosis and loss of germ cells were delayed in CLU knockout mice compared to CLU wild type after heat shock but not other stressors (Bailey et al. 2002). Prostate development and androgen-regulated involution and regrowth of the mouse prostate gland proceed normally in CLU knockout mice (Fink et al. 2006).

7 CLU Receptor: Megalin The CLU receptor is megalin, which is also called glycoprotein 330 or Lipoprotein receptor-related protein 2 (LPR2) (Kounnas et al. 1995). Megalin has extracellular and intracellular domains linked by a single transmembrane domain and belongs to a family of receptor related to the low-density lipoprotein receptors, which mediate cellular internalisation and lysosomal degradation of their ligands. Megalin binds multiple ligands, including proteins and protein-bound vitamins, lipids, drugs, toxins and hormones (Moestrup and Verroust 2001). Megalin expression closely shadows that of CLU, and is predominantly restricted to epithelial cells of both embryonic and adult tissues. Expression is mostly confined to the cell surface, often in clathrin-coated pits, and is observed in the central nervous and male and female reproductive systems, and tissues including lung, kidney, eye and yolk sac epithelium (Zheng et al. 1994).

Differential Regulation of Clusterin Isoforms by the Androgen Receptor

691

Megalin is also overexpressed in some types of renal cell carcinomas (Schuetz et al. 2005). Megalin binds with a high affinity to CLU, which is subsequently internalised and subjected to lysosomal degradation to aid the previously discussed chaperone properties of CLU. There is no evidence for differential affinity of megalin for any known CLU isoforms or splice variants; however, this is an area which requires further study now that the structure of the expressed CLU isoforms and splice variants has been more clearly defined. In vivo, CLU is cleared from testicular fluid through a receptor-mediated interaction with megalin, which is expressed in efferent duct and epididymal epithelial cells, where it acts as an endocytic receptor to mediate CLU lysosomal degradation (Morales et al. 1996). CLU also aids in the degradation of misfolded proteins such as amyloid b-peptide in cerebrospinal fluid through megalin-binding endocytosis and degradation (Hammad et al. 1997). In addition, CLU interacts with megalin on non-professional phagocytes to increase endocytosis of cellular debris from the extracellular environment (Bartl et al. 2001).

8 Androgens, the Androgen Receptor and Gene Regulation Androgens are essential for prostate development and maintenance. The major circulating androgen is testosterone which is synthesised primarily in the Leydig cells in the testis. In target tissues such as prostate, 5-a-reductase metabolises testosterone to the more potent androgen, dihydroxytestosterone (DHT) (Andriole et al. 2004; Soronen et al. 2004). In prostate, both testosterone and DHT act through the androgen receptor (AR). In blood, almost all testosterone is bound by albumin or sex-hormone-binding globulin (SHBG), leaving very little free circulating testosterone. SHBG-bound testosterone enters prostate cells by receptor-mediated endocytosis, whereas free testosterone can diffuse directly into cells (Andreassen 2006). Interestingly, uptake of biologically active androgens bound to SHBG has recently been shown to occur via binding to megalin and subsequent receptor-mediated endocytosis (Hammes et al. 2005). This pathway is biologically functional, as transcription of steroid target genes was induced after androgen-SHBG internalisation (Hammes et al. 2005). Further studies are required to determine if androgens taken into the cells via the megalin receptor can act on CLU transcription. The AR gene is located on the X chromosome, is over 90 kb in length (Kuiper et al. 1989) and results in the production of three transcripts of 11, 8 and 5 kb (Trapman et al. 1988), and a protein of approximately 917 aa which is 110–117 kDa in size (Brinkmann et al. 1989a, b; Tilley et al. 1989). The AR is a member of the steroid family of nuclear receptors, which have structural and functional similarities. At the structural level, these similarities include a C-terminal ligand-binding domain, a hinge region, a DNA-binding domain and an N-terminal domain. Binding of androgens to the AR causes a conformational change to the receptor which allows it to dimerise, undergo nuclear translocation, bind to DNA and act as a transcriptional activator for target genes (Tenbaum and Baniahmad 1997). The AR binds to androgen response elements (AREs) in the DNA, which are inverted

692

T.K. Day et al.

palindromic sequences separated by a three nucleotide spacer. The idealised ARE sequence is 50 -GGA/TACANNNTGTTCT-30 (Roche et al. 1992), and although the AR can induce transactivation from a single ARE, transactivation is increased by cooperative binding to multiple adjacent AREs (Cleutjens et al. 1996; Zhang et al. 2004). AR transactivation is also influenced by post-translational modifications of the AR itself and by interactions with coregulators. Thus, AR can influence gene expression by directly binding to the promoter region of target genes, or indirectly by modulating expression of genes which affect entire downstream pathways.

9 CLU and Androgens Castration-induced prostate involution is characterised by apoptotic cell death of androgen-dependent secretory prostatic epithelial cells. CLU was identified in dying cells from regressing rat prostate (Leger et al. 1987), and initial studies of the response of CLU to androgen deprivation suggested that CLU was negatively regulated by androgens. For example, sCLU-2 is highly upregulated in the involuting prostate following androgen deprivation (Montpetit et al. 1986; Leger et al. 1987; Wong et al. 1993; Lakins et al. 1998; Woolveridge et al. 1998) and also in response to anti-androgen treatment (Leger et al. 1988), resulting in its designation as testosterone-repressed prostate message-2 (TRPM-2). In the normal prostate, sCLU-2 expression is specific to a sub-population of epithelial cells lining the proximal ducts, with no CLU expression in epithelial cells in other prostatic regions (Sensibar et al. 1993; Sutkowski et al. 1993). The castration-induced upregulation of sCLU-2 is marked by intensely positive CLU immunostaining of all epithelial cells in the ventral prostate (Sensibar et al. 1993). In human prostate cancer, immunohistochemistry and western blot analysis showed that sCLU-2 protein expression was significantly increased within weeks after androgen ablation therapy (July et al. 2002). sCLU-2 protein and mRNA expression levels increase in human androgen-dependent prostate cancer models (Kyprianou et al. 1990; Bubendorf et al. 1999). Functionally, over-expression of sCLU-2 in LNCaP prostate cancer cells renders them highly resistant to androgen deprivation in vivo (Miyake et al. 2000c). Inhibition of CLU expression by systemic administration of CLU antisense oligonucleotides (ASOs) in the androgendependent Shionogi tumour model resulted in earlier onset of castration-induced apoptosis, accelerated tumour involution following castration, and significantly delayed progression to the AI phenotype (Miyake et al. 2000c). Additionally, CLU ASO treatment of Shionogi or human AI PC3 tumours enhanced the cytotoxic effects of taxol or mitoxanthrone chemotherapy both in vitro and in vivo (Miyake et al. 2000a, b). Significantly, human clinical data from phase I trials in prostate cancer patients have shown that CLU ASO treatment (OGX-011) significantly suppressed castration-induced increases in sCLU-2 expression at both the mRNA and protein level (Chi et al. 2005). This increased sCLU-2 expression correlated with increased apoptotic rates of prostate cancer cells, further validating a role of CLU ASO inhibitors in prostate cancer cell death after apoptotic stimuli.

Differential Regulation of Clusterin Isoforms by the Androgen Receptor

693

Fig. 4 sCLU-2 expression is associated with apoptosis in Shionogi tumours. CLU mRNA expression was not upregulated 4, 7 or 10 days post-castration when calcium channel blockers were used to inhibit castration-induced apoptosis and Shionogi tumour regression (adapted from Miyake et al. 2000c)

An increasing number of studies have now provided evidence that sCLU-2 is not negatively regulated by androgens. CLU was not upregulated in the absence of testicular androgens in mice bearing Shionogi tumours when calcium channel blockers were used to inhibit castration-induced apoptosis and Shionogi tumour regression (Fig. 4) (Miyake et al. 2000c). When glucocorticoid treatment was used to reduce the rate of prostatic atrophy, CLU mRNA levels were reduced (Rennie et al. 1989). Therefore, it was subsequently proposed that induction of CLU expression is related to apoptosis, rather than being an androgen-repressed gene, as was initially hypothesised. However, the regulation of CLU by androgens has been primarily studied in vivo by removal of androgens in rats or mice bearing xenograft tumours (Leger et al. 1987, 1988; Rennie et al. 1988), making separation of the androgenic and the apoptotic stimuli difficult. A recent study has sought to directly assess the regulation of CLU by androgens (Fig. 3) (Cochrane et al. 2007). Interestingly, this study demonstrated that CLU mRNA and protein were upregulated by androgens in a time- and dose-dependent manner in both normal and tumour prostate cells. CLU regulation in prostate cells was specific to androgens, as the AR was required, anti-androgens reversed the regulation and other steroid hormones did not regulate CLU expression. Moreover, chromatin immunoprecipitation (ChIP) experiments demonstrated that ligand-activated AR directly interacts with androgen response elements (AREs) in an enhancer regulatory region in the first intron of the CLU gene to differentially regulate sCLU-1 and sCLU-2 (Cochrane et al. 2007). sCLU-2 was upregulated by androgens in vitro and as xenografts progressed to androgen independence, whereas sCLU-1 was downregulated by androgens and remained unchanged in the progression to androgen independence (Cochrane et al. 2007).

694

T.K. Day et al.

It is likely that the differential regulation of the two CLU isoforms relates to the opposing pro- and anti-apoptotic functions of different CLU family members. sCLU-2 is an anti-apoptotic protein and so its upregulation by androgens is in agreement with the known cytoprotective role of androgens in normal prostate physiology. In contrast, sCLU-1 can undergo alternative splicing to produce a proapoptotic protein, and so androgens down-regulate this isoform to maximise their anti-apoptotic effects. Reactivation of androgen signalling is important for maintaining cell viability in androgen-independent tumours. The increase in sCLU-2, but not sCLU-1, expression during progression to androgen independence is therefore also in agreement with androgen regulation of CLU. Androgen regulation of CLU may therefore underlie the anti-apoptotic function of sCLU-2, but not the splice variant of sCLU-1, in normal prostate tissue and in prostate cancer progression. Alternative splicing of CLU-1 to the pro-apoptotic splice variant CLU-14II is mediated by ASO or siRNA molecules targeting the exon 2 translation initiation site. The down-regulation of sCLU-1 by androgens raises the possibility of a collaborative role for androgens in the regulation of expression of the pro-apoptotic splice variant, CLU-14II, by regulating the expression level of the CLU-1 precursor mRNA. The regulation of heat-shock protein 27 (Hsp27) phosphorylation by androgens is another example of the regulation of small heat-shock proteins by androgens. Like CLU, Hsp27 is a cytoprotective chaperone expressed in response to stress which regulates downstream effectors of apoptosis. Ligand-activated AR induces rapid Hsp27 phosphorylation in the cytosol, which displaces Hsp90 from a complex with AR to chaperone AR into the nucleus where AR interacts with its response elements to enhance transcriptional activity of target genes (Zoubeidi et al. 2007). This generates a novel feed-forward loop involving cooperative interactions between ligand-activated AR and Hsp27 phospho-activation, which enhances AR stability, nuclear shuttling, and transcriptional activity of anti-apoptotic target genes such as CLU, thereby increasing prostate cancer cell survival. The regulation of CLU and Hsp27 small heat-shock proteins by androgens may represent a previously under-recognised mechanism by which androgens, signalling through the AR, are involved in regulating the apoptotic rheostat in cancer cells by upregulating expression of anti-apoptotic proteins such as sCLU-2 and phospho-Hsp27, while simultaneously decreasing expression of genes such as sCLU-1 which are pro-apoptotic or have the potential to undergo alternative splicing to produce proapoptotic proteins.

References Andreassen TK. 2006. The role of plasma-binding proteins in the cellular uptake of lipophilic vitamins and steroids. Horm Metab Res 38: 279–90. Andriole G, Bruchovsky N, Chung LW, Matsumoto AM, Rittmaster R, Roehrborn C, Russell D, Tindall D. 2004. Dihydrotestosterone and the prostate: the scientific rationale for 5alphareductase inhibitors in the treatment of benign prostatic hyperplasia. J Urol 172: 1399–403.

Differential Regulation of Clusterin Isoforms by the Androgen Receptor

695

Bailey RW, Aronow B, Harmony JA, Griswold MD. 2002. Heat shock-initiated apoptosis is accelerated and removal of damaged cells is delayed in the testis of clusterin/ApoJ knockout mice. Biol Reprod 66: 1042–53. Bartl MM, Luckenbach T, Bergner O, Ullrich O, Koch-Brandt C. 2001. Multiple receptors mediate apoJ-dependent clearance of cellular debris into nonprofessional phagocytes. Exp Cell Res 271: 130–41. Bettuzzi S, Scorcioni F, Astancolle S, Davalli P, Scaltriti M, Corti A. 2002. Clusterin (SGP-2) transient overexpression decreases proliferation rate of SV40-immortalized human prostate epithelial cells by slowing down cell cycle progression. Oncogene 21: 4328–34. Blaschuk O, Burdzy K, Fritz IB. 1983. Purification and characterization of a cell-aggregating factor (clusterin), the major glycoprotein in ram rete testis fluid. J Biol Chem 258: 7714–20. Brinkmann AO, Faber PW, van Rooij HC, Kuiper GG, Ris C, Klaassen P, van der Korput JA, Voorhorst MM, van Laar JH, Mulder E, et al. 1989a. The human androgen receptor: domain structure, genomic organization and regulation of expression. J Steroid Biochem 34: 307–10. Brinkmann AO, Klaasen P, Kuiper GG, van der Korput JA, Bolt J, de Boer W, Smit A, Faber PW, van Rooij HC, Geurts van Kessel A, et al. 1989b. Structure and function of the androgen receptor. Urol Res 17: 87–93. Bruchovsky N, Rennie PS, Coldman AJ, Goldenberg SL, To M, Lawson D. 1990. Effects of androgen withdrawal on the stem cell composition of the Shionogi carcinoma. Cancer Res 50: 2275–82. Bubendorf L, Kolmer M, Kononen J, Koivisto P, Mousses S, Chen Y, Mahlamaki E, Schraml P, Moch H, Willi N, Elkahloun AG, Pretlow TG, Gasser TC, Mihatsch MJ, Sauter G, Kallioniemi OP. 1999. Hormone therapy failure in human prostate cancer: analysis by complementary DNA and tissue microarrays. J Natl Cancer Inst 91: 1758–64. Caccamo AE, Scaltriti M, Caporali A, D’Arca D, Scorcioni F, Candiano G, Mangiola M, Bettuzzi S. 2003. Nuclear translocation of a clusterin isoform is associated with induction of anoikis in SV40-immortalized human prostate epithelial cells. Ann N Y Acad Sci 1010: 514–9. Caccamo AE, Scaltriti M, Caporali A, D’Arca D, Scorcioni F, Astancolle S, Mangiola M, Bettuzzi S. 2004. Cell detachment and apoptosis induction of immortalized human prostate epithelial cells are associated with early accumulation of a 45 kDa nuclear isoform of clusterin. Biochem J 382: 157–68. Caccamo AE, Scaltriti M, Caporali A, D’Arca D, Corti A, Corvetta D, Sala A, Bettuzzi S. 2005. Ca2+ depletion induces nuclear clusterin, a novel effector of apoptosis in immortalized human prostate cells. Cell Death Differ 12: 101–4. Caccamo AE, Desenzani S, Belloni L, Borghetti AF, Bettuzzi S. 2006. Nuclear clusterin accumulation during heat shock response: implications for cell survival and thermo-tolerance induction in immortalized and prostate cancer cells. J Cell Physiol 207: 208–19. Calero M, Rostagno A, Matsubara E, Zlokovic B, Frangione B, Ghiso J. 2000. Apolipoprotein J (clusterin) and Alzheimer’s disease. Microsc Res Tech 50: 305–15. Cao C, Shinohara ET, Li H, Niermann KJ, Kim KW, Sekhar KR, Gleave M, Freeman M, Lu B. 2005. Clusterin as a therapeutic target for radiation sensitization in a lung cancer model. Int J Radiat Oncol Biol Phys 63: 1228–36. Chi KN, Eisenhauer E, Fazli L, Jones EC, Goldenberg SL, Powers J, Tu D, Gleave ME. 2005. A phase I pharmacokinetic and pharmacodynamic study of OGX-011, a 20 -methoxyethyl antisense oligonucleotide to clusterin, in patients with localized prostate cancer. J Natl Cancer Inst 97: 1287–96. Cleutjens KB, van Eekelen CC, van der Korput HA, Brinkmann AO, Trapman J. 1996. Two androgen response regions cooperate in steroid hormone regulated activity of the prostatespecific antigen promoter. J Biol Chem 271: 6379–88. Cochrane DR, Wang Z, Muramaki M, Gleave ME, Nelson CC. 2007. Differential regulation of clusterin and its isoforms by androgens in prostate cells. J Biol Chem 282: 2278–87. Connor J, Buttyan R, Olsson CA, D’Agati V, O’Toole K, Sawczuk IS. 1991. SGP-2 expression as a genetic marker of progressive cellular pathology in experimental hydronephrosis. Kidney Int 39: 1098–103.

696

T.K. Day et al.

de Silva HV, Harmony JA, Stuart WD, Gil CM, Robbins J. 1990. Apolipoprotein J: structure and tissue distribution. Biochemistry 29: 5380–9. DeMattos RB, O’Dell MA, Parsadanian M, Taylor JW, Harmony JA, Bales KR, Paul SM, Aronow BJ, Holtzman DM. 2002. Clusterin promotes amyloid plaque formation and is critical for neuritic toxicity in a mouse model of Alzheimer’s disease. Proc Natl Acad Sci U S A 99: 10843–8. Fink TM, Zimmer M, Tschopp J, Etienne J, Jenne DE, Lichter P. 1993. Human clusterin (CLI) maps to 8p21 in proximity to the lipoprotein lipase (LPL) gene. Genomics 16: 526–8. Fink D, Fazli L, Aronow B, Gleave ME, Ong CJ. 2006. Clusterin is not essential for androgenregulated involution and regeneration of the normal mouse prostate. Prostate 66: 1445–54. French LE, Sappino AP, Tschopp J, Schifferli JA. 1992. Distinct sites of production and deposition of the putative cell death marker clusterin in the human thymus. J Clin Invest 90: 1919–25. French LE, Chonn A, Ducrest D, Baumann B, Belin D, Wohlwend A, Kiss JZ, Sappino AP, Tschopp J, Schifferli JA. 1993. Murine clusterin: molecular cloning and mRNA localization of a gene associated with epithelial differentiation processes during embryogenesis. J Cell Biol 122: 1119–30. Fritz IB, Burdzy K, Setchell B, Blaschuk O. 1983. Ram rete testis fluid contains a protein (clusterin) which influences cell–cell interactions in vitro. Biol Reprod 28: 1173–88. Gleave M, Nelson C, Chi K. 2003. Antisense targets to enhance hormone and cytotoxic therapies in advanced prostate cancer. Curr Drug Targets 4: 209–21. Gomez JA, Gama V, Yoshida T, Sun W, Hayes P, Leskov K, Boothman D, Matsuyama S. 2007. Bax-inhibiting peptides derived from Ku70 and cell-penetrating pentapeptides. Biochem Soc Trans 35: 797–801. Hammad SM, Ranganathan S, Loukinova E, Twal WO, Argraves WS. 1997. Interaction of Apolipoprotein J-amyloid beta-peptide complex with low density lipoprotein receptor-related protein-2/megalin. A mechanism to prevent pathological accumulation of amyloid beta-peptide. J Biol Chem 272: 18644–9. Hammes A, Andreassen TK, Spoelgen R, Raila J, Hubner N, Schulz H, Metzger J, Schweigert FJ, Luppa PB, Nykjaer A, Willnow TE. 2005. Role of endocytosis in cellular uptake of sex steroids. Cell 122: 751–62. Han BH, DeMattos RB, Dugan LL, Kim-Han JS, Brendza RP, Fryer JD, Kierson M, Cirrito J, Quick K, Harmony JA, Aronow BJ, Holtzman DM. 2001. Clusterin contributes to caspase-3independent brain injury following neonatal hypoxia-ischemia. Nat Med 7: 338–43. Hara I, Miyake H, Gleave ME, Kamidono S. 2001. Introduction of clusterin gene into human renal cell carcinoma cells enhances their resistance to cytotoxic chemotherapy through inhibition of apoptosis both in vitro and in vivo. Jpn J Cancer Res 92: 1220–4. Hough CD, Cho KR, Zonderman AB, Schwartz DR, Morin PJ. 2001. Coordinately up-regulated genes in ovarian cancer. Cancer Res 61: 3869–76. Itahana Y, Piens M, Sumida T, Fong S, Muschler J, Desprez PY. 2007. Regulation of clusterin expression in mammary epithelial cells. Exp Cell Res 313: 943–51. Jones SE, Jomary C. 2002. Clusterin. Int J Biochem Cell Biol 34: 427–31. Jordan-Starck TC, Lund SD, Witte DP, Aronow BJ, Ley CA, Stuart WD, Swertfeger DK, Clayton LR, Sells SF, Paigen B, et al. 1994. Mouse Apolipoprotein J: characterization of a gene implicated in atherosclerosis. J Lipid Res 35: 194–210. July LV, Akbari M, Zellweger T, Jones EC, Goldenberg SL, Gleave ME. 2002. Clusterin expression is significantly enhanced in prostate cancer cells following androgen withdrawal therapy. Prostate 50: 179–88. July LV, Beraldi E, So A, Fazli L, Evans K, English JC, Gleave ME. 2004. Nucleotide-based therapies targeting clusterin chemosensitize human lung adenocarcinoma cells both in vitro and in vivo. Mol Cancer Ther 3: 223–32. Kapron JT, Hilliard GM, Matsubara E, Tenniswood MP, Frangione B, Carr SA, Crabb JW. 2000. Identification and characterization of glycosylation sites in human serum clusterin) and Alzheimer’s disease. Microsc Res Tech 50: 2120–33. Koch-Brandt C, Morgans C. 1996. Clusterin: a role in cell survival in the face of apoptosis? Prog Mol Subcell Biol 16: 130–49.

Differential Regulation of Clusterin Isoforms by the Androgen Receptor

697

Kounnas MZ, Loukinova EB, Stefansson S, Harmony JA, Brewer BH, Strickland DK, Argraves WS. 1995. Identification of glycoprotein 330 as an endocytic receptor for Apolipoprotein J/clusterin. J Biol Chem 270: 13070–5. Kuiper GG, Faber PW, van Rooij HC, van der Korput JA, Ris-Stalpers C, Klaassen P, Trapman J, Brinkmann AO. 1989. Structural organization of the human androgen receptor gene. J Mol Endocrinol 2: R1–4. Kyprianou N, English HF, Isaacs JT. 1990. Programmed cell death during regression of PC-82 human prostate cancer following androgen ablation. Cancer Res 50: 3748–53. Kyprianou N, English HF, Davidson NE, Isaacs JT. 1991. Programmed cell death during regression of the MCF-7 human breast cancer following estrogen ablation. Cancer Res 51: 162–6. Lakins J, Bennett SA, Chen JH, Arnold JM, Morrissey C, Wong P, O’Sullivan J, Tenniswood M. 1998. Clusterin biogenesis is altered during apoptosis in the regressing rat ventral prostate. J Biol Chem 273: 27887–95. Leger JG, Montpetit ML, Tenniswood MP. 1987. Characterization and cloning of androgenrepressed mRNAs from rat ventral prostate. Biochem Biophys Res Commun 147: 196–203. Leger JG, Le Guellec R, Tenniswood MP. 1988. Treatment with antiandrogens induces an androgen-repressed gene in the rat ventral prostate. Prostate 13: 131–42. Leskov KS, Klokov DY, Li J, Kinsella TJ, Boothman DA. 2003. Synthesis and functional analyses of nuclear clusterin, a cell death protein. J Biol Chem 278: 11590–600. McLaughlin L, Zhu G, Mistry M, Ley-Ebert C, Stuart WD, Florio CJ, Groen PA, Witt SA, Kimball TR, Witte DP, Harmony JA, Aronow BJ. 2000. Apolipoprotein J/clusterin limits the severity of murine autoimmune myocarditis. J Clin Invest 106: 1105–13. Michel D, Chatelain G, Herault Y, Brun G. 1995. The expression of the avian clusterin gene can be driven by two alternative promoters with distinct regulatory elements. Eur J Biochem 229: 215–23. Michel D, Chatelain G, North S, Brun G. 1997. Stress-induced transcription of the clusterin/apoJ gene. Biochem J 328 (Pt 1): 45–50. Miyake H, Chi KN, Gleave ME. 2000a. Antisense TRPM-2 oligodeoxynucleotides chemosensitize human androgen-independent PC-3 prostate cancer cells both in vitro and in vivo. Clin Cancer Res 6: 1655–63. Miyake H, Nelson C, Rennie PS, Gleave ME. 2000b. Acquisition of chemoresistant phenotype by overexpression of the antiapoptotic gene testosterone-repressed prostate message-2 in prostate cancer xenograft models. Cancer Res 60: 2547–54. Miyake H, Nelson C, Rennie PS, Gleave ME. 2000c. Testosterone-repressed prostate message-2 is an antiapoptotic gene involved in progression to androgen independence in prostate cancer. Cancer Res 60: 170–6. Miyake H, Hara I, Kamidono S, Gleave ME. 2001. Synergistic chemsensitization and inhibition of tumor growth and metastasis by the antisense oligodeoxynucleotide targeting clusterin gene in a human bladder cancer model. Clin Cancer Res 7: 4245–52. Miyake H, Hara I, Gleave ME, Eto H. 2004. Protection of androgen-dependent human prostate cancer cells from oxidative stress-induced DNA damage by overexpression of clusterin and its modulation by androgen. Prostate 61: 318–23. Moestrup SK, Verroust PJ. 2001. Megalin- and cubilin-mediated endocytosis of protein-bound vitamins, lipids, and hormones in polarized epithelia. Annu Rev Nutr 21: 407–28. Montpetit ML, Lawless KR, Tenniswood M. 1986. Androgen-repressed messages in the rat ventral prostate. Prostate 8: 25–36. Morales CR, Igdoura SA, Wosu UA, Boman J, Argraves WS. 1996. Low density lipoprotein receptor-related protein-2 expression in efferent duct and epididymal epithelia: evidence in rats for its in vivo role in endocytosis of Apolipoprotein J/clusterin. Biol Reprod 55: 676–83. Moretti RM, Marelli MM, Mai S, Cariboni A, Scaltriti M, Bettuzzi S, Limonta P. 2007. Clusterin isoforms differentially affect growth and motility of prostate cells: possible implications in prostate tumorigenesis. Cancer Res 67: 10325–33.

698

T.K. Day et al.

Park DC, Yeo SG, Shin EY, Mok SC, Kim DH. 2006. Clusterin confers paclitaxel resistance in cervical cancer. Gynecol Oncol 103: 996–1000. Poon S, Easterbrook-Smith SB, Rybchyn MS, Carver JA, Wilson MR. 2000. Clusterin is an ATPindependent chaperone with very broad substrate specificity that stabilizes stressed proteins in a folding-competent state. Biochemistry 39: 15953–60. Pucci S, Bonanno E, Pichiorri F, Angeloni C, Spagnoli LG. 2004. Modulation of different clusterin isoforms in human colon tumorigenesis. Oncogene 23: 2298–304. Ranney MK, Ahmed IS, Potts KR, Craven RJ. 2007. Multiple pathways regulating the antiapoptotic protein clusterin in breast cancer. Biochim Biophys Acta 1772: 1103–11. Reddy KB, Jin G, Karode MC, Harmony JA, Howe PH. 1996. Transforming growth factor beta (TGF beta)-induced nuclear localization of Apolipoprotein J/clusterin in epithelial cells. Biochemistry 35: 6157–63. Redondo M, Villar E, Torres-Munoz J, Tellez T, Morell M, Petito CK. 2000. Overexpression of clusterin in human breast carcinoma. Am J Pathol 157: 393–9. Rennie PS, Bruchovsky N, Buttyan R, Benson M, Cheng H. 1988. Gene expression during the early phases of regression of the androgen-dependent Shionogi mouse mammary carcinoma. Cancer Res 48: 6309–12. Rennie PS, Bowden JF, Freeman SN, Bruchovsky N, Cheng H, Lubahn DB, Wilson EM, French FS, Main L. 1989. Cortisol alters gene expression during involution of the rat ventral prostate. Mol Endocrinol 3: 703–8. Roche PJ, Hoare SA, Parker MG. 1992. A consensus DNA-binding site for the androgen receptor. Mol Endocrinol 6: 2229–35. Rosemblit N, Chen CL. 1994. Regulators for the rat clusterin gene: DNA methylation and cisacting regulatory elements. J Mol Endocrinol 13: 69–76. Rosenberg ME, Girton R, Finkel D, Chmielewski D, Barrie A, III, Witte DP, Zhu G, Bissler JJ, Harmony JA, Aronow BJ. 2002. Apolipoprotein J/clusterin prevents a progressive glomerulopathy of aging. Mol Cell Biol 22: 1893–902. Scaltriti M, Santamaria A, Paciucci R, Bettuzzi S. 2004. Intracellular clusterin induces G2-M phase arrest and cell death in PC-3 prostate cancer cells. Cancer Res 64: 6174–82. Schuetz AN, Yin-Goen Q, Amin MB, Moreno CS, Cohen C, Hornsby CD, Yang WL, Petros JA, Issa MM, Pattaras JG, Ogan K, Marshall FF, Young AN. 2005. Molecular classification of renal tumors by gene expression profiling. J Mol Diagn 7: 206–18. Schwochau GB, Nath KA, Rosenberg ME. 1998. Clusterin protects against oxidative stress in vitro through aggregative and nonaggregative properties. Kidney Int 53: 1647–53. Sensibar JA, Qian Y, Griswold MD, Sylvester SR, Bardin CW, Cheng CY, Lee C. 1993. Localization and molecular heterogeneity of sulfated glycoprotein-2 (clusterin) among ventral prostate, seminal vesicle, testis, and epididymis of rats. Biol Reprod 49: 233–42. Sensibar JA, Sutkowski DM, Raffo A, Buttyan R, Griswold MD, Sylvester SR, Kozlowski JM, Lee C. 1995. Prevention of cell death induced by tumor necrosis factor alpha in LNCaP cells by overexpression of sulfated glycoprotein-2 (clusterin). Cancer Res 55: 2431–7. Shannan B, Seifert M, Leskov K, Willis J, Boothman D, Tilgen W, Reichrath J. 2006. Challenge and promise: roles for clusterin in pathogenesis, progression and therapy of cancer. Cell Death Differ 13: 12–9. Sintich SM, Steinberg J, Kozlowski JM, Lee C, Pruden S, Sayeed S, Sensibar JA. 1999. Cytotoxic sensitivity to tumor necrosis factor-alpha in PC3 and LNCaP prostatic cancer cells is regulated by extracellular levels of SGP-2 (clusterin). Prostate 39: 87–93. So A, Sinnemann S, Huntsman D, Fazli L, Gleave M. 2005. Knockdown of the cytoprotective chaperone, clusterin, chemosensitizes human breast cancer cells both in vitro and in vivo. Mol Cancer Ther 4: 1837–49. Soronen P, Laiti M, Torn S, Harkonen P, Patrikainen L, Li Y, Pulkka A, Kurkela R, Herrala A, Kaija H, Isomaa V, Vihko P. 2004. Sex steroid hormone metabolism and prostate cancer. J Steroid Biochem Mol Biol 92: 281–6.

Differential Regulation of Clusterin Isoforms by the Androgen Receptor

699

Steinberg J, Oyasu R, Lang S, Sintich S, Rademaker A, Lee C, Kozlowski JM, Sensibar JA. 1997. Intracellular levels of SGP-2 (Clusterin) correlate with tumor grade in prostate cancer. Clin Cancer Res 3: 1707–11. Sutkowski DM, Kasjanski RZ, Sensibar JA, Ney KG, Lim DJ, Kozlowski JM, Lee C, Grayhack JT. 1993. Effect of spermatocele fluid on growth of human prostatic cells in culture. J Androl 14: 233–9. Suuronen T, Nuutinen T, Ryhanen T, Kaarniranta K, Salminen A. 2007. Epigenetic regulation of clusterin/Apolipoprotein J expression in retinal pigment epithelial cells. Biochem Biophys Res Commun 357: 397–401. Tenbaum S, Baniahmad A. 1997. Nuclear receptors: structure, function and involvement in disease. Int J Biochem Cell Biol 29: 1325–41. Tilley WD, Marcelli M, Wilson JD, McPhaul MJ. 1989. Characterization and expression of a cDNA encoding the human androgen receptor. Proc Natl Acad Sci U S A 86: 327–31. Trapman J, Klaassen P, Kuiper GG, van der Korput JA, Faber PW, van Rooij HC, Geurts van Kessel A, Voorhorst MM, Mulder E, Brinkmann AO. 1988. Cloning, structure and expression of a cDNA encoding the human androgen receptor. Biochem Biophys Res Commun 153: 241–8. Trougakos IP, So A, Jansen B, Gleave ME, Gonos ES. 2004. Silencing expression of the clusterin/ Apolipoprotein J gene in human cancer cells using small interfering RNA induces spontaneous apoptosis, reduced growth ability, and cell sensitization to genotoxic and oxidative stress. Cancer Res 64: 1834–42. Tschopp J, Chonn A, Hertig S, French LE. 1993. Clusterin, the human apolipoprotein and complement inhibitor, binds to complement C7, C8 beta, and the b domain of C9. J Immunol 151: 2159–65. Wicher GK, Aldskogius H. 2005. Adult motor neurons show increased susceptibility to axotomyinduced death in mice lacking clusterin. Eur J Neurosci 21: 2024–8. Wong P, Pineault J, Lakins J, Taillefer D, Leger J, Wang C, Tenniswood M. 1993. Genomic organization and expression of the rat TRPM-2 (clusterin) gene, a gene implicated in apoptosis. J Biol Chem 268: 5021–31. Wong P, Taillefer D, Lakins J, Pineault J, Chader G, Tenniswood M. 1994. Molecular characterization of human TRPM-2/clusterin, a gene associated with sperm maturation, apoptosis and neurodegeneration. Eur J Biochem 221: 917–25. Woolveridge I, Taylor MF, Wu FC, Morris ID. 1998. Apoptosis and related genes in the rat ventral prostate following androgen ablation in response to ethane dimethanesulfonate. Prostate 36: 23–30. Yang CR, Yeh S, Leskov K, Odegaard E, Hsu HL, Chang C, Kinsella TJ, Chen DJ, Boothman DA. 1999. Isolation of Ku70-binding proteins (KUBs). Nucleic Acids Res 27: 2165–74. Yang CR, Leskov K, Hosley-Eberlein K, Criswell T, Pink JJ, Kinsella TJ, Boothman DA. 2000. Nuclear clusterin/XIP8, an x-ray-induced Ku70-binding protein that signals cell death. Proc Natl Acad Sci U S A 97: 5907–12. Yeung LH, Read JT, Sorenson P, Nelson CC, Jia W, Rennie PS. 2003. Identification and characterization of a prostate-specific androgen-independent protein-binding site in the probasin promoter. Biochem J 371: 843–55. Zellweger T, Miyake H, July LV, Akbari M, Kiyama S, Gleave ME. 2001. Chemosensitization of human renal cell cancer using antisense oligonucleotides targeting the antiapoptotic gene clusterin. Neoplasia 3: 360–7. Zellweger T, Chi K, Miyake H, Adomat H, Kiyama S, Skov K, Gleave ME. 2002. Enhanced radiation sensitivity in prostate cancer by inhibition of the cell survival protein clusterin. Clin Cancer Res 8: 3276–84. Zhang J, Gao N, Kasper S, Reid K, Nelson C, Matusik RJ. 2004. An androgen-dependent upstream enhancer is essential for high levels of probasin gene expression. Endocrinology 145: 134–48. Zhang H, Kim JK, Edwards CA, Xu Z, Taichman R, Wang CY. 2005. Clusterin inhibits apoptosis by interacting with activated Bax. Nat Cell Biol 7: 909–15.

700

T.K. Day et al.

Zhang Q, Zhou W, Kundu S, Jang TL, Yang X, Pins M, Smith N, Jovanovic B, Xin D, Liang L, Guo Y, Lee C. 2006. The leader sequence triggers and enhances several functions of clusterin and is instrumental in the progression of human prostate cancer in vivo and in vitro. BJU Int 98: 452–60. Zheng G, Bachinsky DR, Stamenkovic I, Strickland DK, Brown D, Andres G, McCluskey RT. 1994. Organ distribution in rats of two members of the low-density lipoprotein receptor gene family, gp330 and LRP/alpha 2MR, and the receptor-associated protein (RAP). J Histochem Cytochem 42: 531–42. Zoubeidi A, Zardan A, Beraldi E, Fazli L, Sowery R, Rennie P, Nelson C, Gleave M. 2007. Cooperative interactions between androgen receptor (AR) and heat-shock protein 27 facilitate AR transcriptional activity. Cancer Res 67: 10455–65.

Androgen Regulation of Prostate Cancer Gene Fusions Rou Wang, Scott A. Tomlins, and Arul M. Chinnaiyan

Abstract Recurrent chromosomal rearrangements were not well characterized in epithelial malignancies until the recent discovery of recurrent fusions of TMPRSS2 and ETS transcription factors in prostate cancer. Since the initial discovery of the TMPRSS2:ETS gene fusions, multiple other 50 and 30 fusion partners have been discovered, which led to the identification of five different classes of ETS gene rearrangements to date. These rearrangements demonstrate variable response to androgen stimulation, and ultimately, may help tailor androgen-deprivation therapy for patients with advanced prostate cancer. Initial studies have demonstrated that ETV1 overexpression mediates invasion in prostate cell lines, though secondary factors may be required for the development of frank malignancy. TMPRSS2:ETS gene fusions may confer a more aggressive phenotype, and fusion status has been linked to adverse clinical factors, such as higher tumor stage and nodal metastases. ETS factor fusions are likely an important contributor to human prostate cancer development and progression, though more research will be needed to elucidate their regulation by androgen and clinical functionality. Ultimately, identification of distinct gene fusion status may allow for risk stratification and tailored treatment of patients with prostate cancer.

1 Introduction Prostate cancer remains the most common noncutaneous malignancy of men in the United States, with an estimated 186,320 new cases diagnosed in 2007, and is the second leading cause of cancer-related deaths (Jemal et al. 2008). A role for androgen regulation of prostate cancer has been recognized since the 1940s, during which time Huggins and Hodges reported their seminal findings of an androgeninduced increase in serum phosphatase of men with metastatic prostate cancer, and A.M. Chinnaiyan(*) Department of Pathology and Urology, University of Michigan, 5316 CCGC, 1500 E. Medical Center Drive, Ann Arbor, MI 48109-0944, USA, E-mail: [email protected]

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4_30, # Springer Science + Business Media, LLC 2009

701

702

R. Wang et al.

subsequent marked reduction of this enzyme by castration or estrogen (Huggins and Hodges 2002). The importance of these findings was recognized by the bestowment of the Nobel Prize to Huggins in 1966. The normal development of the prostate gland, benign prostatic hyperplasia, and prostate cancer is now known to rely upon interactions of the androgen receptor (AR) with target genes to activate their transcription. However, while most prostate cancers are initially responsive to androgen-deprivation therapy, virtually all tumors, if given enough time, ultimately recur, and the different mechanisms allowing survival then growth continue to be studied intently. The aggressiveness of prostate cancer is assessed histologically using the Gleason grading system, which relies entirely upon the degree of loss of the normal glandular tissue architecture (pattern 1 is the most differentiated and pattern 5 is the least differentiated) (Gleason and Mellinger 1974); the overall Gleason score is the sum of the two most prevalent tumor patterns. While the Gleason score and clinical features help predict a patient’s prognosis, they cannot account entirely for the heterogeneity of biological and morphological behavior of prostate cancers. Significant efforts have been made to define the underlying genetic changes and biological processes involved in tumorigenesis and prostate cancer progression, with the ultimate goals of improving prostate cancer detection and risk stratification, and developing novel pharmacologic targets. Microarray-based gene expression methods allow for high-throughput screening of thousands of genes of interest simultaneously, and have accelerated the search for potential molecular markers for diagnosis and treatment in various malignancies. A primary goal of expression profiling analysis is identification of candidate genes whose expression and subsequent function are altered in tumorigenesis or cancer progression. These alterations can arise from different recurrent chromosomal aberrations, which include deletions of a tumor-suppressor gene, translocations, or inversions. Translocations can result in promoter and/or enhancer elements of one gene juxtaposed to a proto-oncogene, thus activating the proto-oncogene, or rearrangement fusing two genes (Rabbitts 1994). Gene fusions often result in the production of a chimeric fusion protein that has new or altered oncogenic activity. An example of this type of translocation is the BCR–ABL gene fusion in chronic myelogenous leukemia (CML) (Rowley 1973; de Klein et al. 1982), which led to the development of imatinib mesylate (Gleevec), a drug which targets ABL tyrosine kinase activity (Deininger et al. 2005). Historically, gene fusions have been found commonly in hematologic malignancies, but were described rarely in epithelial carcinomas. Currently, 264 gene fusions have been identified in hematological disorders vs. 70 gene fusions in malignant solid tumors (Mitelman et al. 2007). Traditionally, gene fusions were thought to play a minor role in carcinoma tumorigenesis, but the utilization of novel bioinformatic approaches has helped identify recurrent gene fusions that play a primary role in the development of common solid tumors, including prostate cancer. This chapter provides insight into the discovery of gene fusions in prostate cancer and the role of androgen regulation of these gene fusions.

Androgen Regulation of Prostate Cancer Gene Fusions

703

2 Discovery of Prostate Cancer Gene Fusions 2.1

Cancer Outlier Profile Analysis

The detection of prostate cancer gene fusions required a conceptual paradigm shift in the discovery process of outlier genes. Traditional analytical methods that search for common activation of genes across a class of cancer samples will miss the heterogeneous patterns of oncogene activation found in many cancers. A method was required to detect genes that demonstrate marked overexpression in only a subset of cases, and thus cancer outlier profile analysis (COPA) was developed as a biostatistical algorithm to analyze DNA microarray data for gene outliers (Tomlins et al. 2005). In essence, COPA accentuates outlier profiles by applying a simple numerical transformation based on the median and median absolute deviation of a gene expression profile. In the first step, gene expression values are median centered; each gene’s median expression value is set to zero. Next, the median absolute deviation (MAD) is calculated and scaled to one by dividing each gene expression value by its MAD. Finally, the 75th, 90th, and 95th percentiles of the transformed expression values are calculated, and the genes are rank ordered by their percentile scores. When applied to the Oncomine database (Rhodes et al. 2007), an online compendium of multiple gene expression data sets, COPA produced a list of prioritized genes in different cancer types known to involve a recurrent rearrangement or high-level amplification. In several independent prostate cancer microarray data sets (Welsh et al. 2001; Dhanasekaran et al. 2001; Dhanasekaran et al. 2005; Glinsky et al. 2004; Lapointe et al. 2004; Yu et al. 2004), COPA identified strong outlier gene profiles for v-ets erythroblastosis virus E26 oncogene (ERG, 21q22.3) and Ets variant gene 1 (ETV1, 7p21.2) in a subset of cases, with scores in the top 10 for six different prostate profiling studies (Table 1). Both ERG and ETV1 are genes that encode E26transcription specific (ETS) family transcription factors and are known to be involved in oncogenic translocations in Ewing’s sarcoma and myeloid leukemias (Oikawa and Yamada 2003; Hsu et al. 2004). In acute myeloid leukemia, ERG has Table 1 Cancer outlier profile analysis (COPA) of Prostate Cancer Data sets Rank % Score Gene Cancer Study 1 95 10.03795 ETV1 Prostate Lapointe et al. (2004) 1 75 5.4071 ERG Prostate Dhanasekaran et al. (2005) 1 75 4.3628 ERG Prostate Welsh et al. (2001) 1 75 3.4414 ERG Prostate Lapointe et al. (2004) 1 75 3.3875 ERG Prostate Dhanasekaran et al. (2001) 9 95 17.1698 ETV1 Prostate Glinsky et al. (2004) 9 75 2.2218 ERG Prostate Yu et al. (2004 ) COPA analysis revealed strong outlier profiles for ERG and ETV1 in different prostate cancer microarray data sets

704

R. Wang et al.

been shown to regulate expression of genes significant in cell proliferation, differentiation, and apoptosis (Marcucci et al. 2005). In Ewing’s sarcoma, these translocations are redundant functionally, and thus only one type of translocation is observed in each case. These results were recapitulated with COPA analysis of the Oncomine prostate cancer microarray data sets. ERG and ETV1 consistently showed mutually exclusive outlier profiles, both in larger transcriptome studies using grossly dissected prostate tissues and in our own work utilizing prostate samples obtained with laser capture microdissection (Tomlins et al. 2005).

2.2

Recurrent Gene Fusions of TMPRSS2 to ERG or ETV1 in Prostate Cancer

To further delineate the mechanism responsible for ERG and ETV1 overexpression, quantitative polymerase chain reaction (QPCR) was used to determine which prostate cancer cell lines overexpressed either ERG or ETV1. For ETV1, the LNCaP prostate cancer cell line and two specimens from castration-recurrent prostate cancer (MET26-RP, from residual primary prostate cancer, and MET26LN, from lymph node metastases) demonstrated overexpression. Another lymph node metastasis (MET28-LN) and the prostate cancer cell lines, VCaP and DuCaP, overexpressed ERG. However, neither ERG nor ETV1 was amplified consistently in samples with transcript overexpression, which suggested the possibility of DNA rearrangement. Exon-walking QPCR with primer pairs spanning ETV1 exons 2–7 was used in the LNCaP and two MET26 specimens overexpressing ETV1, and showed similar overexpression of all measured ETV1 exons in LNCaP but differential expression of ETV1 exons in the two MET26 samples (with >90% reduction in ETV1 exons 2 and 3, compared to exons 4 through 7). The reduced expression of the 50 regions of ERG and ETV1 in the context of overexpression of the 30 regions suggested a possible gene fusion. 50 RNA ligase-mediated rapid amplification of cDNA ends (RLM-RACE) was used to identify the 50 transcripts of ERG and ETV1 in cases with overexpression of these genes. Sequencing of the cloned products revealed fusions of the prostatespecific, androgen-responsive transmembrane serine protease gene TMPRSS2 (21q22.2) with ETV1 in MET26-LN and with ERG in MET28-LN (Fig. 1). Ultimately, four different initial gene fusion products were discovered, using QPCR and standard reverse-transcription PCR (RT-PCR), with specific forward primers for TMPRSS2 and reverse primers for different exons of ERG or ETV1, and electrophoresis of the QPCR products and sequencing of cloned RT-PCR products. The fusion products included TMPRSS2:ETV1a (fusion of the complete exon 1 of TMPRSS2 with the beginning of exon 4 of ETV1) and TMPRSS2:ETV1b (fusion of exons 1 and 2 of TMPRSS2 with exon 4of ETV1) in MET26-LN, TMPRSS2:ERGa (fusion of the complete exon 1 of TMPRSS2 with the beginning of exon 4 of ERG) in MET28-LN, and TMPRSS2:ERGb (fusion of the complete exon 1 of TMPRSS2

Androgen Regulation of Prostate Cancer Gene Fusions

705

ETV1

TMPRSS2 71

11

142

2

365

3

132

3194

14

1

1

TMPRSS2:ETV1

220

2

268

3

4

1

142

2

ERG

365

3

6154

12

12

TMPRSS2 71

322

4

37

3194

14

TMPRSS2:ERG

1

1

139 225

2

4

3

443

4

3097

11

11

Fig. 1 TMPRSS2 Fusion Transcripts. Schematic of initial fusion transcripts found with 50 RLMRACE for TMPRSS2 with ETV1 and ERG. Structures for these genes have their basis in the Gen Bank reference sequences. Boxes indicate exons, and the numbers above reflect the last base of each exon. Untranslated regions are illustrated in lighter shades, and coding exons not depicted are illustrated by hatched boxes

with the beginning of exon 2 of ERG) in the prostate cancer sample PCA4. Evidence for these fusions was found only in cases that overexpressed either ERG or ETV1, and these fusions were confirmed at the chromosomal level with interphase fluorescence in situ hybridization (FISH) on formalin-fixed paraffinembedded (FFPE) specimens from MET26 and 28 and on tissue microarrays. Of 29 initial cases represented on the tissue microarray, 23 (79%) demonstrated evidence of TMPRSS2:ETV1 fusion (seven cases) or ERG rearrangement (16 cases). The substantial proportion of fusion-positive cases has been recapitulated in multiple different studies demonstrating that 15–78% of prostate cancers harbor the TMPRSS2:ERG fusion (Soller et al. 2006; Wang et al. 2006; Demichelis et al. 2007; Perner et al. 2007), while only 1–10% of prostate cancer cases harbor the TMPRSS2:ETV1 fusion (Hermans et al. 2006; Mehra et al. 2007a, b; Tu et al. 2007).

2.3

Androgen Regulation of ERG in TMPRSS2:ERG Fusion

Cloned in 1997 by Paoloni-Giacobino et al. (1997), TMPRSS2 encodes a multimeric protein with a serine protease domain (Paoloni-Giacobino et al. 1997). TMPRSS2 was found highly expressed in benign and malignant prostate epithelium and strongly induced by androgen stimulation (Lin et al. 1999; Afar et al. 2001; Jacquinet et al. 2001). However, its role in prostate cancer development was unclear until the discovery of TMPRSS2:ERG and TMPRSS2:ETV1 gene fusions. To determine whether the TMPRSS2:ERG fusion also resulted in the androgen regulation of ERG, ERG expression was assessed using QPCR in androgen-treated VCaP

706

R. Wang et al.

cells, which express the TMPRSS2:ERG fusion, and LNCaP cells, which do not express this fusion (Tomlins et al. 2005). Both cell lines demonstrated increased expression of PSA upon stimulation with R1881, a synthetic androgen, and decreased expression with the androgen receptor antagonists, bicalutamide and flutamide. However, only fusion-positive VCaP cells responded to androgen stimulation with increased ERG expression, which was inhibited by the androgen receptor antagonists. A similar increase in ERG expression was noted in the DuCaP cell line known to express the TMPRSS2:ERG fusion, which suggests that gene fusion with TMPRSS2 results in androgen-regulated aberrant expression of ERG and ETV1 in specific subsets of prostate cancer.

3 Other Gene Fusion Subtypes The expression of all ETS family members was examined in various prostate cancer profiling studies in the Oncomine database. The ETS family member ETV4 (17q21) was identified as having marked overexpression in a single prostate cancer case from two different studies, and ETV4 (17q21) was regulated possibly by TMPRSS2 (Tomlins et al. 2006). A similar technique of exon-walking QPCR demonstrated overexpression of ETV4 in the prostate cancer sample PCA5, vs. pooled benign prostate tissue and other prostate cancers not harboring ETV4 overexpression. A >99% decrease in the expression of exon 2 of ETV4 relative to more distal regions suggested a possible gene fusion. This was further investigated with RLM-RACE, which revealed two transcripts, TMPRSS2:ETV4a and TMPRSS2:ETV4b, both of which contain 50 ends consisting of sequence located 8 kb upstream of TMPRSS2. Even though these ETV4 fusion isoforms do not contain known exons from TMPRSS2 as in the original TMPRSS2:ERG and TMPRSS2:ETV1 gene fusions, the structure of these fusion transcripts and the marked overexpression of ETV4 exons lend further support to the idea that regulatory elements upstream to TMPRSS2 actually drive dysregulation of the ETS family members. After the discovery of TMPRSS2:ERG and TMPRSS2:ETV1 gene fusions, fewer TMPRSS2:ETV1 cases were identified than would be expected based on the frequency of ETV1 outlier expression in microarray studies. This led to a search for novel 50 fusion partners with ETV1, and by RLM-RACE, unique 50 sequences other than from TMPRSS2 were found in ETV1 outlier cases. In PCa_ETV1_1, ETV1 exons 1–4 were replaced with two exons from 22q11.23 with homology to human endogenous retrovirus family K (HERV-K_22q11.23). In PCa_ETV1_2, ETV1 exon 1 was replaced with exon 1 of HNRPA2B1 (7p15). In PCa_ETV1_3, exons 1–4 of ETV1 were replaced with a 50 -extended exon 1 of SLC45A3 (1q32), and in MET23, exons 1–5 of ETV1 were replaced with exons 1–2 from C15orf21 (15q21) (Tomlins et al. 2007a, b). These four new fusion transcripts were confirmed using QPCR and genomic fusions confirmed using FISH. However, unlike the initial TMPRSS2:ETV1 fusion, none of these fusions contained translated sequences from the 50 partner, with the exception of HNRPA2B1, which contributed only two residues to the ultimate HNRPA2B1:ETV1 fusion protein. The tissue specificity

Androgen Regulation of Prostate Cancer Gene Fusions

707

and androgen regulation of these new 50 partners was characterized by query into microarray data sets accessed in the Oncomine database and massively parallel signature sequencing (MPSS) data sets. SLC45A3 and C15orf21 showed overexpression in prostate cancer similar to TMPRSS2. HNRPA2B1 showed high expression in prostate and other tumor types, and HERV-K_22q11.23 showed high expression in benign prostate vs. other benign tissues. Endogenous expression of SLC45A3 and HERV-K_22q11.23, assessed using QPCR, in the LNCaP prostate cancer cell line was increased by R1881, similar to TMPRSS2. However, the expression of C15orf21 was decreased by androgen stimulation, while the expression of HNRPA2B1 was not affected by R1881 administration (Fig. 2). Screening prostate cancer cell lines with ETV1 outlier expression identified another class of rearrangements in LNCaP and MDA-PCa 2B cell lines. A rearrangement was identified, using a split-probe FISH strategy; a minimal region around ETV1 was inserted cryptically into an intronic sequence from the MIPOL1 locus at 14q13.3–14q21.1 of the LNCaP cell line, known to harbor ETV1 overexpression. Another ETV1 rearrangement was identified in MDA-PCa 2B, a cell line known previously to demonstrate a balanced t(7;14)(p21;q21) translocation (van Bokhoven et al. 2003), which corresponds to the locations of ETV1 and MIPOL1 loci. The ETV1 locus was shown using FISH to be translocated to the derivative chromosome 14. The 14q13.3–14q21.1 region is the partner of this balanced translocation, with the telomeric 14q13.3–14q21.1 probe localizing to the derivative chromosome 7. Both these distinct rearrangements result in the localization of ETV1 to 14q13.3– 14q21.1 in prostate cancer cell lines with ETV1 overexpression, which suggests that enhancer elements in this region confer coordinated prostate specificity and androgen regulation and drive ETV1 overexpression. The expression of the four contiguous genes within this 1.5 MB 14q13.3–14q21.1 region (SLC25A21, MIPOL1, FOXA1, and TTC6) was assessed using Oncomine analysis. All four genes demonstrated significant overexpression in prostate cancer compared to other cancers. Subsequent stimulation of LNCaP with R1881 minimally increased the expression of FOXA1, but not SLC25A21, MIPOL1, and TTC6. However, R1881 stimulation of LNCaP, known to harbor an ETV1 rearrangement, increased ETV1 expression, while that did not occur in androgen-sensitive cell lines without ETV1 rearrangements, such as VCaP. This suggests that rearrangement to 14q13.3–14q21.1 drives ETV1 overexpression and allows for androgen responsiveness. To further study the relationship between ETV1 and the genes at 14q13.3– 14q21.1 in a model simulating progression to castration-recurrent disease, gene expression was investigated in androgen-sensitive LNCaP and its derivative C4-2B (Thalmann et al. 1994), both confirmed to harbor the same ETV1 rearrangement using FISH and QPCR. In cell lines with publicly available expression profiling data, ETV1 shows marked downregulation in LNCaP derivatives (Liu et al. 2004; Shi et al. 2004a, b; Oudes et al. 2005; Murillo et al. 2006). QPCR confirmed downregulation of ETV1 in C4-2B, 22,500 fold compared to LNCaP, downregulation of PSA, and upregulation of C15ORF21. C4-2B demonstrated decreased expression of the four genes at 14q13.3–14a21 compared to LNCaP, which

708

R. Wang et al.

Response to Androgen

Fusion Structure

ETV1 132

1

220 268

2 2-43

322

4

452 641 1299

6

5

7-11

6154

12

TMPRSS2 71

142

1

2

365

3194

4-13

3

14

TMPRSS2:ETV1a 71

1

12

5-11

4

SLC45A3 109

1

511

2

1297 1563

4

3

3382

5

SLC45A3:ETV1 −117

109

1

12

5-11

HERV-K_22q11.23 256

974

1

2

HERV-K:ETV1

−27

334

1

2 5-11

12

C15ORF21 70

336 456 581 669

1

2

3

4

1927

6

5

C15ORF21:ETV1

−8

336

1

2

6

7-11

12

HNRPA2B1 175

1

286

2

1134 1217

3-9

10

1714

11

HNRPA2B1:ETV1

175

1

2

3

4

5-11

12

Fig. 2 New Regulatory Elements Fused to ETV1. Structure of new 50 partners fused to ETV1 in outlier cases and corresponding androgen regulation. All genes are prostate-specific except HNRPA2B1, which is a ubiquitously active regulatory element

Androgen Regulation of Prostate Cancer Gene Fusions

709

supports the idea that coordinated regulation of these genes drives ETV1 overexpression in androgen-sensitive prostate cancer cell lines. ETV1 downregulation in androgen-independent LNCaP derivatives may occur through deletion, mutation, or transcriptional mechanisms. Selecting mutations that allow for bypass of ETV1dependent pathways may be an integral part of the development of androgen independence. Most recently, ETV5 (3q27) was identified as the fourth ETS family member involved in recurrent gene fusions in prostate cancer (Helgeson et al. 2008). Three different TMPRSS2:ETV5 isoforms were detected in the prostate cancer sample PCa_ETV5_1, and the SLC45A3:ETV5 gene fusion was found in a different prostate cancer sample, PCa_ETV5_2. SLC45A3 was shown previously to be a 50 androgen-induced fusion partner in ETV1 rearrangements (Tomlins et al. 2007a, b). COPA analysis demonstrated that ETV5 outlier expression is rare (1%). Identification of these fusions with exon-walking QPCR and RLM-RACE was confirmed using FISH split signal, fusion assays, and QPCR. Ectopic overexpression of ETV5 in the benign prostate cancer cell line, RWPE, induced invasion and an invasive transcriptional program, as determined using molecular concept modeling, analogous to overexpression with ERG and ETV1. In summary, five general classes of chromosomal ETS rearrangements have been identified in prostate cancer (Fig. 3). Fusions between TMPRSS2 and ETS family members, which include ERG, ETV1, ETV4, and ETV5, represent the predominant class (Class I) of ETS rearrangements. Class II consists of rearrangements involving fusions with untranslated regions from other prostate-specific, androgen-induced 50 partner genes (such as SLC45A3) and endogenous retroviral elements (HERV-K), similar to TMPRSS2:ETS rearrangements. Class III comprises prostate-specific, androgen-repressed 50 partner fusions, such as C15ORF21:ETV1. Conversely, HNRPA2B1:ETV1 represents a gene fusion in which nontissue-specific promoter elements drive ETS expression and androgen does not seem to regulate expression (Class IV). Similar fusions involving strong promoters of universally expressed genes such as HNRPA2B1 could result in oncogene overexpression across different tumor types. Finally, Class V rearrangements involve the entire ETS family gene being rearranged to androgen-regulated prostate-specific regions, such as 14q13.3–14q21.1. Other insights and variants to add to this general classification have been described by different groups. These include the discovery of TMPRSS2:ERG fusion mechanisms via translocation or interstitial deletion (Hermans et al. 2006). The vast majority of tumor cell populations demonstrate homogeneous gene fusion status (Perner et al. 2007). A homogeneous deletion site between ERG and TMPRSS2 on chromosome 21q22.2-3 was identified using FISH and oligonucleotide single nucleotide polymorphisms arrays. Two subclasses were distinguished by the start point of the deletion at 38.765 or 38.911 Mb (Perner et al. 2006). A variant TMPRSS2 isoform has been discovered as an alternative fusion partner of ERG by Lapointe et al. (2007) using RACE and array CGH (Lapointe et al. 2007). This novel sequence begins 4 kb upstream of the TMPRSS2 start site, and is also expressed in some prostate cancers (with 70% demonstrating either the known or

710

R. Wang et al.

Class

Fusion Structure

I

1

4

−1

1

Element Type

Induced

Proximal promoter

Serine protease

Induced

Proximal promoter

?Function; also known as prostein

Induced

Retroviral element

Human endogenous retrovirus

Repressed

Proximal promoter

?Function; also known as D-PCA-2

Unchanged

Proximal promoter

Heteronuclear ribonuclear protein; not prostate-specific

Induced

Enhancer

12

5-11

TMPRSS2:ETV1 TMPRSS2:ERG TMPRSS2:ETV4 TMPRSS2:ETV5

IIa

Androgen Regulation

12

5-11

SLC45A3:ETV1 SLC45A3:ETV5

IIb

1

2 5-11

12

5’ Fusion Partner Characterization

HERV-K:ETV1 1

III

2

6 7-11

12

C15ORF21:ETV1

IV

V

1

2

3

4

5-11

12

HNRPA2B1:ETV1

ETS

No fusion transcript

ins(7;14)(p21;q21) (LNCaP) t(7;14)(p21;q21) (MDA-PCa 2B)

Fig. 3 Distinct Classes of Chromosomal ETS Rearrangements in Prostate Cancer. Five classes of chromosomal rearrangements activate ETS family members in prostate cancer. For each class, the prototypical fusion structure and specific examples are listed. Response to androgen and type of genomic element responsible for aberrant ETS expression are indicated. Characterization of the 5 0 fusion partners and relevant caveats are also noted – HNRPA2B1 (Class IV) is the only fusion partner that is not prostate specific, and Class V rearrangements do not express a fusion transcript

novel variant, 44% expressing both variants, and 10% expressing the variant isoform only). The TMPRSS2 isoform 2-ERG fusion is likely under similar androgen regulation as it results in comparable pathogenic alterations in the subset of cancers expressing only this isoform. Wang et al. (2006) analyzed 59 different clinically localized prostate cancers and demonstrated eight different TMPRSS2ERG fusion types varying in fusion transcript junctions and length of PCR products (Wang et al. 2006). The initial TMPRSS2:ERGa and TMPRSS2:ERGb fusions discovered by Tomlins et al. (2005) are included in this set. The type of ERG fusion mRNA expressed seems to be an important associated factor in prostate cancer aggressiveness and progression, with certain subtypes [Types I (exon 1 of TMPRSS2 fused to the beginning of ERG exon 2), II (exon1 of TMPRSS2 fused to the beginning of ERG exon 3), and VI (exons 1 and 2 of TMPRSS2 fused to the beginning of ERG exon 4)] associated with aggressive disease, defined by early PSA recurrence and seminal vesicle invasion. Tumors that do not express alternate fusion isoforms tend to have higher overall expression of fusion mRNA, and this higher expression may also be associated with poorer outcome. The discovery of these classes may have important implications for future detection of additional gene fusions, may aid in prognosis assessment, and may

Androgen Regulation of Prostate Cancer Gene Fusions

711

guide treatment, especially in androgen-deprivation therapy for patients with advanced prostate cancer. Men with prostate cancer metastasis are treated primarily with androgen-deprivation therapy, which causes tumor regression in most men. However, the cancer inevitably recurs and usually causes death. Knowledge of these different fusions is important since they may help identify patients with androgen-insensitive fusions (such as HNRPA2B1:ETV1) who may not respond to androgen-deprivation therapy, for example. Androgen-deprivation therapy may actually increase ETS fusion expression and hasten cancer progression in men with Class III rearrangements (such as C15ORF21:ETV1).

4 ETS Genes and Relation to Prostate Cancer Invasion and Progression In addition to the discovery of gene fusions in prostate cancer, the functional importance of ETS overexpression in prostate cancer biology has been elucidated. Recapitulating overexpression of truncated ETV1 in vitro and in vivo supported an oncogenic role of aberrant ETS family member expression (Tomlins et al. 2007a, b). ETV1 was overexpressed in RWPE cells, using adenoviral and lentiviral constructs. ETV1 overexpression had no effect on proliferation, percentage of cells in S phase of the cell cycle, and was not sufficient for transformation. However, ETV1 overexpression in RWPE and primary prostatic epithelial (PrEC) cells increased invasion in a modified basement membrane invasion assay. Furthermore, ETV1 knockdown in LNCaP (accomplished by infecting LNCaP cells with lentiviruses expressing ETV1 shRNA) inhibited invasion compared to control LNCaP cells and LNCaP cells infected with lentivirus expressing a nontargeting shRNA. Transgenic mice were created to express ARR2Pb-ETV1, a truncated version of ETV1 under the control of the modified probasin promoter (ARR2Pb). Strong transgene expression in the prostate under androgen regulation simulated the gene fusions of ETV1 in human cancer. By 12–14 weeks, 75% of these mice demonstrated mouse prostatic intraepithelial neoplasia (mPIN), with strong ETV1 expression in mPIN foci but not benign epithelium. None of these mice demonstrated progression to invasive carcinoma; additional factors may be required for cancer development. Nevertheless, ETV1 induced a preneoplastic phenotype in mouse prostate that suggests ETS factor fusions may contribute to human prostate cancer development. Studies were performed to recapitulate TMPRSS2:ERG fusions in vitro and in vivo, and demonstrated similar effects as TMPRSS2:ETV1 fusions (Tomlins et al. 2008). Expression of the TMPRSS2:ERG fusion product resulted in the development of PIN in mouse prostate. However, ERG overexpression in RWPE and PrEC cells increased invasion, but did not result in transformation. Furthermore, as TMPRSS2:ERG fusions often occur in the context of pre-existing genetic lesions, TMPRSS2:ERG fusions were studied in a more realistic context. siRNA knockdown of ERG in the VCaP cell line inhibited invasion and modulated transcriptional programs that were associated with ERG-positive vs. ERG-negative prostate

712

R. Wang et al.

cancers and invasive cancer vs. PIN. VCaP may serve as a more useful model as these programs were not modulated in RWPE cells. However, in both cell line models, RWPE-ERG and VCaP, the plasminogen activator pathway was found crucial for ERG-mediated invasion. Specifically, downstream direct targets of ERG in RWPE-ERG cells were identified, which include urokinase plasminogen activator (PLAU) and metalloproteinase 3 (MMP3). Inhibition of PLAU blocks ERGmediated invasion. ChIP identified both PLAU and tissue plasminogen activator (PLAT) as direct targets of ERG in VCaP-siERG cells, and inhibition of PLAU inhibited invasion of the VCaP cell line. These findings suggest that the plasminogen activator pathway is important in fusion-mediated invasion and may serve as a future therapeutic target. Perner et al. (2007) also helped elucidate the role of the TMPRSS2-ERG fusion as a likely early molecular event associated with invasion (Perner et al. 2007). Using multicolor interphase FISH assays, the TMPRSS2:ERG fusion was detected in 48.5% of clinically localized prostate cancer, 30% of androgen-stimulated metastases, and 33% of castration-recurrent metastases. The fusion, when detected, was distributed homogeneously, which suggests that it is an early event. The fusion was not detected in benign prostate tissue or proliferative inflammatory atrophy, but was found in 19% of high-grade prostatic intraepithelial neoplasia (PIN) lesions. Furthermore, all but one of the high-grade PIN lesions with the TMPRSS2:ERG fusion were in close proximity to prostate cancer with the same fusion, which suggests that proliferative inflammatory atrophy and high-grade PIN may be precursors to prostate cancer, either through sequential or separate pathways. Chromosome copy number changes have been detected in clinically localized prostate cancer, but not in paired high-grade PIN lesions with the same fusion status; some groups believe that the TMPRSS2:ERG fusion is an early event that precedes chromosome-level alterations in prostate cancer development (Cerveira et al. 2006). In our own integrative analysis using the molecular concepts map (MCM), an analytic framework for exploring the network for relationships between biologically related gene sets and visualizing ‘‘enrichment networks’’ (Tomlins et al. 2007a, b), PIN, and prostate cancer demonstrate more similar gene expression profiles (1.2% differentially expressed genes) than between benign prostate epithelium and PIN (13.4%), or localized and metastatic prostate cancer (15.7%). Further MCM analysis of PIN and benign signatures reveals enrichment of increased androgen signaling and protein biosynthesis concepts containing ETS target genes such as ELK1, ETS1, and GABPA (or NRF-2), which suggests that a major process distinguishing PIN from benign epithelium is increased protein synthesis, likely through ETS target genes. Prostate cancer demonstrates frank overexpression of ETS genes, such as ERG, ETV1, or ETV4, which are not found in benign prostate or PIN, suggesting that these gene fusions may serve to lock in pathways previously activated in PIN. Additionally, in ETS-overexpressing vs. non-ETS tumors, 6q21 was the most enriched chromosome subarm, which suggests a cooperating amplification of 6q21 in ETS tumors or loss of 6q21 in non-ETS tumors.

Androgen Regulation of Prostate Cancer Gene Fusions

713

5 Androgen Signaling in Prostate Cancer Development MCM analysis also confirmed the critical role of androgen signaling in prostate cancer development and progression, with increased androgen signaling noted in PIN compared to benign tissue and decreased androgen signaling in localized prostate cancer vs. PIN, high vs. low Gleason score, and androgen-stimulated or castration-recurrent metastases vs. clinically localized prostate cancer. Possible mechanisms for metastases involve the selective downregulation of androgenresponsive genes (resulting in increased proliferation, dedifferentiation, or reduced apoptosis (Hendriksen et al. 2006) or the ability of distinct cell populations), with varying amounts of androgen signaling reflecting differentiation status, to drive increased proliferation. Additionally, even though androgen signaling concepts on MCM analysis and androgen-regulated genes are downregulated in castrationrecurrent metastatic cancer, these genes still act as important transcripts. Androgen-deprivation therapy is successful initially in the treatment of metastatic prostate cancer, but these therapies eventually fail. Androgen-deprivation therapy may force the selection of lesions that are able to restore some minimal level of androgen signaling, thus allowing for survival of malignant cells that require minimal androgen signaling. Mechanisms for castration-recurrent growth can be separated into categories, although most center on the AR. These mechanisms include DNA-based alterations, such as amplifications or mutations, in the AR gene. Chen and colleagues (2004) demonstrated that a modest increase in AR mRNA was the only change consistently associated with resistance to androgendeprivation therapy, and that this increase was necessary for developing androgen independence but still relied upon a functional ligand-binding domain. Increased AR levels amplify signal output from low levels of residual ligands (whether androgen or noncanonical ligands such as estrogen or hydrocortisone) (Chen et al. 2004). However, AR mutations and amplification occur only in a subset of prostate cancers. Other mechanisms have been postulated, and include altered activity of AR co-regulators that increases activation of the AR promoter, ligandindependent activation of AR by other kinases and substrates, and complete bypass of the AR by alternative signaling pathways (Kasper and Cookson 2006). Regardless of the exact mechanism, even minimal androgen signaling can result still in expression of ETS family members through TMPRSS2:ETS gene fusions, via the strong androgen promoter-enhancer elements regulating TMPRSS2 expression. The exact interactions of AR with TMPRSS2 and ETS family genes, as well as the role of gene fusions in castration-recurrent prostate cancer, remain to be fully elucidated. Recently, more insight has been gained into the mechanism of TMPRSS2 regulation by androgen via the AR. Ninety AR-binding sites on chromosomes 21 and 22 were mapped in the LNCaP prostate cancer cell line, utilizing chromatin immunoprecipitation (ChIP) combined with tiled DNA oligonucleotide microarray analysis (ChIP-on-chip) and gene expression profiling (Wang et al. 2007). Most of these sites were found to contain noncanonical AR-responsive elements (AREs), which are required for the enhancer function of the AR-binding region. Quantitative ChIP was used to confirm AR binding to site B39, which is

714

R. Wang et al.

13.5 kb upstream of the TMPRSS2 mRNA start site. This binding site serves as a TMPRSS2 enhancer which interacts with the downstream TMPRSS2 promoter region, to suggest an explanation for androgen regulation of TMPRSS2-ETS family member gene fusions. Three transcription factors were identified as collaborators in AR function, including FoxA1, GATA2, and Oct1, which are recruited to ARbinding regions and may form independent complexes with AR. AR interacts with GATA2 and Oct1 in an androgen-dependent manner (with GATA2 directly required for AR binding), and both these motifs are required for activity of the TMPRSS2 enhancer. These findings provide further insight into androgen regulation of these gene fusions and cooperating transcription factors. Of the many mechanisms described for castration-recurrent growth, TMPRSS2 has been shown to activate protease-activated receptor-2 (PAR-2), a member of the family of transmembrane G-protein coupled receptors (PAR-1–4) that play a role in the thrombin-mediated clotting cascade and cancer progression (Wilson et al. 2005). PAR-1 levels are increased in cell lines which harbor a small proportion of androgen-sensitive cells and thus continue to proliferate in the context of androgen deprivation (Salah et al. 2005); and PAR-1 activation may play a role in angiogenesis (Huang et al. 2001; Yin et al. 2003; Kaushal et al. 2006), prevention of apoptosis (Tantivejkul et al. 2005), and cell migration in prostate cancer (Shi et al. 2004a, b). Castration-recurrent metastatic prostate cancers with ETS gene fusions have been shown to either continue to overexpress the fusion transcript (MET26 and MET28, from the initial discovery experiments for TMPRSS2:ERG and TMPRSS2: ETV1 gene fusions) or not express the fusion gene (Hermans et al. 2006). This dichotomy may represent two different mechanisms – the former in which the tumor selects processes to restore androgen signaling to allow for continued gene fusion expression, and the latter in which the tumor selects for mutations that alleviate the need for gene fusion expression and instead uses alternative molecular pathways. In addition, TMPRSS2:ERG fusion has been demonstrated in NCI-H660, an AR-negative cell line derived from a metastatic site of an extrapulmonary small cell carcinoma of the prostate (Mertz et al. 2007). The presence of TMPRSS2:ERG fusion through intronic deletion on chromosome 21 was demonstrated using realtime QPCR, two-stage FISH assay testing, and single-nucleotide polymorphism oligonucleotide arrays. Furthermore, in comparison to the VCaP cell line, in which androgen stimulation results in increased ERG expression, NCI-H660 does not respond to androgen stimulation with increases in PSA or ERG, and is insensitive to AR antagonists. This provides a new cell line for future study of the role of TMPRSS2:ETS gene fusions in castration-recurrent AR-negative prostate cancer.

5.1

TMPRSS2-ETS Gene Fusions and Clinical Outcomes

Various studies have attempted to address the association of TMPRSS2:ERG fusion status and patient outcomes; results have been conflicting, in some degree due to small cohorts and heterogeneous patient populations. Several studies have demon-

Androgen Regulation of Prostate Cancer Gene Fusions

715

strated no statistical association of TMPRSS2:ERG gene fusion with tumor stage, Gleason grade (Perner et al. 2007), recurrence-free survival (Lapointe et al. 2007), or biochemical recurrence (Mehra et al. 2007a, b). Only one study to date has demonstrated associations between TMPRSS2:ERG fusion-positive prostate cancer, lower Gleason grade, and better survival, although this study was limited by its small sample size of 50 tumors, of which 18 (36%) were found to be fusion positive (Winnes et al. 2007). The majority of studies have reported associations between TMPRSS2:ETS family gene fusion status and poor outcome. TMPRSS2:ERG rearrangement through deletion was associated with higher tumor stage (p = 0.03) and nodal metastases (p = 0.02) in a study by Perner et al. (2006) (Perner et al. 2006), and with higher pathologic stage (p = 0.04) in our study in which fusion status was detected using FISH (Mehra et al. 2007a, b). Certain TMPRSS2:ERG isoforms have been associated with clinical outcome. Expression of a fusion isoform in which the TMPRSS2 ATG is in frame with exon 4 of ERG was associated with early PSA recurrence (p = 0.038) and seminal vesicle invasion (p = 0.038). Expression of isoforms with the native ERG ATG from exon 3 as the first in-frame ATG was associated with seminal vesicle invasion (p = 0.02). Expression of any of these specific isoforms was associated with early recurrence (p = 0.035) and seminal vesicle invasion (p = 0.005) (Wang et al. 2006). In a series of 111 men on watchful waiting for clinically localized prostate cancer, with a 15% incidence of TMPRSS2:ERG fusion, fusion-positive tumors were associated with higher Gleason score (p = 0.01) and prostate cancer-specific death [cumulative incidence ratio = 2.7, 95% confidence interval(CI): 1.3–5.8, p < 0.01] (Demichelis et al. 2007). High ERG expression was demonstrated using QPCR in these fusion-positive tumors, again suggesting that the more aggressive phenotype associated with positive fusion status is mediated through increased ERG expression. Another study investigated the presence of TMPRSS2:ERG gene fusion products using RT-PCR and direct sequencing in 165 prostatectomy specimens, and discovered that the fusion protein was the single most important prognostic factor for prostate cancer recurrence in multivariate analysis, independent of grade, stage, and PSA level (adjusted hazard ratio: 8.6, 95%CI: 3.6–20.6, p < 0.0001) (Nam et al. 2007). TMPRSS2:ERG fusion status has also been associated with distinct morphological features comprising a more aggressive phenotype (Mosquera et al. 2007). A total of 253 cases were assessed for TMPRSS2:ERG fusion using ERG break-apart FISH assays. After logistic regression analysis, five morphological features were found associated with TMPRSS2:ERG fusion prostate cancer: blue-tinged mucin, cribriform growth pattern, macronucleoli, intraductal tumor spread, and signet-ring cell features (all p < 0.05), many of which have been associated previously with more aggressive behavior. A model was constructed from these pathologic features to predict TMPRSS2:ERG fusion status. 24% of cases with none of these features demonstrated TMPRSS2:ERG positivity, vs. 55% of case with one feature, 86% of cases with two features, and 93% of cases with three or more features (p < 0.001). This model predicted fusion status with sensitivity 75% and specificity 71%, and may be important to identify prostate cancer in which a confirmatory test via FISH

716

R. Wang et al.

is needed to document TMPRSS2:ERG fusion positivity. Results should be confirmed in larger, more representative patient cohorts before fusion status can be utilized as a definitive biomarker or other risk-stratification tool.

5.2

Multifocal Prostate Cancer and Fusion Status

Recently, different groups have demonstrated the heterogeneity of TMPRSS2 gene rearrangements in multifocal prostate cancer. Our group has evaluated 93 multifocal prostate cancers from 43 radical prostatectomy specimens, and demonstrated using FISH that 70% of cases showed any TMPRSS2 rearrangement, and of these 30 cases, 70% showed discordant TMPRSS2 rearrangement foci and the largest (index) tumor was rearranged most of the time (83%) (Mehra et al. 2007a, b). Barry et al. (2007) confirmed this interfocal clonal heterogeneity in 41% of 32 cases, with all individual tumor foci demonstrating homogeneity for fusion status (Barry et al. 2007). These findings highlight the importance of molecular heterogeneity when assessing accuracy and adequacy of guided needle biopsy strategies, and developing diagnostic or prognostic tumor markers. Tumor heterogeneity may also explain the difficulty in defining a distinct molecular signature for prostate cancer. Further investigation of heterogeneity is required to determine which of the different foci within one prostate is the most aggressive. The molecular mechanisms of the most aggressive foci may drive prostate cancer progression for that entire gland.

6 TMPRSS2:ERG Gene Fusions as a Clinical Target An early study in men with prostate cancer demonstrated the concordance between FISH-detected TMPRSSS:ERG fusion in genomic and urine samples following digital rectal exam (Laxman et al. 2006). Two subsequent studies have combined fusion status with other markers detectable in urine. Hessels et al. (2007) found that the combination of prostate cancer antigen 3 (PCA3) and TMPRSS2:ERG fusion transcripts as biomarkers increased sensitivity of prostate cancer detection to 73% (Hessels et al. 2007). In our prospective study of 236 men presenting for prostate biopsy or prostatectomy, the incorporation of gene fusion status into a urine multiplex of a total of seven biomarkers improved on the testing characteristics of any single urine marker, and serum PSA (Laxman et al. 2008). Fusion status may be particularly useful as a prostate cancer biomarker for patients who have had multiple negative prior biopsies, although more studies will need to address its performance in comparison or in addition to conventional markers of prostate cancer, such as serum PSA. Traditional therapy for prostate cancer includes radical prostatectomy, external beam radiation, or brachytherapy for clinically localized disease, and androgendeprivation therapy, radiation therapy, and chemotherapy, for metastatic disease.

Androgen Regulation of Prostate Cancer Gene Fusions

717

While there are no distinct anti-ETS pharmacotherapies currently available, the discovery of these gene fusions in prostate cancer serves as a springboard for developing future therapeutic targets, as has occurred in other malignancies. One novel method of screening for existing drugs that may modulate known gene fusions in malignancies was described by Stegmaier et al. (2007), who searched for drugs targeting the EWS/FLI fusion protein in pediatric Ewing sarcoma. They identified differences between the gene signatures of the EWS/FLI ‘‘on’’ and ‘‘off’’ (EWS/FLI RNA interference) states using gene expression-based high-throughput screening. A small-molecule library enriched for FDA-approved drugs was screened using a high-throughput ligation-mediated amplification assay and a fluorescent, bead-based detection system. Cytosine arabinoside (ARA-C) was found to modulate EWS/FLI. ARA-C diminished cell viability and transformation, stopped tumor growth, and demonstrated a good safety profile in a xenograft model. The authors advocate for the initiation of clinical trials testing ARA-C in Ewing sarcoma. ARA-C treatment did not affect RWPE-ERG invasion in our studies (Tomlins et al. 2008). Meanwhile, other new pharmacotherapies continue to be developed to treat prostate cancer. For example, the CYP540c17 17a-hydroxylase/ C17–20-lyase activity inhibitor, abiraterone acetate (Cougar Biotechnology, Inc.), is an oral systemic inhibitor of adrenal androgen synthesis intended to enhance maximal androgen suppression in patients with advanced prostate cancer (Attard et al. 2005). Initial studies have demonstrated suppression of testosterone levels in both castrate and noncastrate males with prostate cancer (O’Donnell et al. 2004), and further studies are in progress. Though abiraterone and other drugs are promising, they treat advanced prostate cancer. Potentially, future drugs that better target the AR or its downstream targets, which include the TMPRSS2:ETS family gene fusions, may halt prostate cancer development or progression.

7 Conclusions The initial discovery of TMPRSS2:ETS gene fusions in prostate cancer provided support for a primary role of chromosomal rearrangements in common epithelial cancers. Since that discovery, multiple classes of gene rearrangements have been identified, and more is being learned about the functional role of these gene fusions in prostate cancer development and progression. TMPRSS2:ETS family gene fusions likely mediate invasion through the plasminogen activator pathway, and expression of the ERG gene fusion product under androgen regulation in mice results in the development of prostatic intraepithelial neoplasia. Secondary molecular lesions may be required for the development of prostate cancer. Additionally, the nuances of androgen regulation of these gene fusions are being elucidated. The variable response of these fusion subtypes to androgen may be important ultimately in tailoring androgen-deprivation therapy in advanced disease and better predicting prognosis. For example, prostate cancers harboring fusions with androgen-insensitive partners, such as HNRPA2B1, may not respond to

718

R. Wang et al.

androgen-deprivation therapy, whereas those with androgen-repressed fusion partners such as C15orf 21 may experience further fusion expression and faster progression. Ultimately, as has been accomplished in other cancers such as chronic myelogenous leukemia, both investigators and clinicians hope that these gene fusions may provide a tool for tailored risk stratification and a new therapeutic target in prostate cancer.

References Afar, D.E., Vivanco, I., Hubert, R.S., et al. 2001. Catalytic cleavage of the androgen-regulated TMPRSS2 protease results in its secretion by prostate and prostate cancer epithelia. Cancer Research 61:1686–1692. Attard, G., Belldegrun, A.S., and de Bono, J.S. 2005. Selective blockade of androgenic steroid synthesis by novel lyase inhibitors as a therapeutic strategy for treating metastatic prostate cancer. BJU International 96:1241–1246. Barry, M., Perner, S., Demichelis, F., et al. 2007. TMPRSS2-ERG fusion heterogeneity in multifocal prostate cancer: clinical and biologic implications. Urology 70:630–633. Cerveira, N., Ribeiro, F.R., Peixoto, A., et al. 2006. TMPRSS2-ERG gene fusion causing ERG overexpression precedes chromosome copy number changes in prostate carcinomas and paired HGPIN lesions. Neoplasia 8:826–832. Chen, C.D., Welsbie, D.S., Tran, C., et al. 2004. Molecular determinants of resistance to antiandrogen therapy. Nature Medicine10:33–39. Dhanasekaran, S.M., Barrette, T.R., Ghosh, D., et al. 2001. Delineation of prognostic biomarkers in prostate cancer. Nature 412:822–826. Dhanasekaran, S.M., Dash, A., Yu, J., et al. 2005. Molecular profiling of human prostate tissues: insights into gene expression patterns of prostate development during puberty. FASEB Journal 19:243–245. Deininger, M., Buchdunger, E., and Druker, B.J. 2005. The development of imatinib as a therapeutic agent for chronic myeloid leukemia. Blood 105:2640–2653. de Klein, A., van Kessel, A.G., Grosveld, G., et al. 1982. A cellular oncogene is translocated to the Philadelphia chromosome in chronic myelocytic leukaemia. Nature 300:765–767. Demichelis, F., Fall, K., Perner, S., et al. 2007. TMPRSS2:ERG gene fusion associated with lethal prostate cancer in a watchful waiting cohort. Oncogene 26:4596–4599. Gleason, D.F., and Mellinger, G.T. 1974. Prediction of prognosis for prostatic adenocarcinoma by combined histological grading and clinical staging. Journal of Urology 111:58–64. Glinsky, G.V., Glinskii, A.B., Stephenson, A.J., et al. 2004. Gene expression profiling predicts clinical outcome of prostate cancer. The journal of Clinical Investigation 113:913–923. Helgeson, B.E., Tomlins, S.A., Shah, N., et al. 2008. Characterization of TMPRSS2:ETV5 and SLC45A3:ETV5 gene fusions in prostate cancer. Cancer Research 68:73–80. Hendriksen, P.J., Dits, N.F., Kokame, K., et al. 2006. Evolution of the androgen receptor pathway during progression of prostate cancer. Cancer Research 66:5012–5020. Hermans, K.G., van Marion, R., van Dekken, H., et al. 2006. TMPRSS2:ERG fusion by translocation or interstitial deletion is highly relevant in androgen-dependent prostate cancer, but is bypassed in late-stage androgen receptor-negative prostate cancer. Cancer Research 66:10658– 10663. Hessels, D., Smit, F.P., Verhaegh, G.W., et al. 2007. Detection of TMPRSS2-ERG fusion transcripts and prostate cancer antigen 3 in urinary sediments may improve diagnosis of prostate cancer. Clinical Cancer Research 13:5103–5108.

Androgen Regulation of Prostate Cancer Gene Fusions

719

Hsu, T., Trojanowska, M., and Watson, D.K. 2004. Ets proteins in biological control and cancer. Journal of Cellular Biochemistry 91:896–903. Huang, Y.Q., Li, J.J., Hu, L., et al. 2001. Thrombin induces increased expression and secretion of VEGF from human FS4 fibroblasts, DU145 prostate cells and CHRF megakaryocytes. Thrombosis & Haemostasis 86:1094–1098. Huggins, C., and Hodges, C.V. 2002. Studies on prostatic cancer: I. The effect of castration, of estrogen and of androgen injection on serum phosphatases in metastatic carcinoma of the prostate. 1941. Journal of Urology 168:9–12. Jacquinet, E., Rao, N.V., Rao, G.V., et al. 2001. Cloning and characterization of the cDNA and gene for human epitheliasin. European Journal of Biochemistry 268:2687–2699. Jemal, A., Siegel, R., Ward, E., et al. 2008. Cancer statistics, 2008. CA: a Cancer Journal for Clinicians 58:71–96. Kasper, S., Cookson, M.S. 2006. Mechanisms leading to the development of hormone-resistant prostate cancer. Urologic Clinics of North America 33:201–210. Kaushal, V., Kohli, M., Dennis, R.A., et al. 2006. Thrombin receptor expression is upregulated in prostate cancer. Prostate 66:273–282. Lapointe, J., Li, C., Higgins, J.P., et al. 2004. Gene expression profiling identifies clinically relevant subtypes of prostate cancer. Proceedings of the National Academy of Sciences of the United States of America 101:811–816. Lapointe, J., Kim, Y.H., Miller, M.A., et al. 2007. A variant TMPRSS2 isoform and ERG fusion product in prostate cancer with implications for molecular diagnosis. Modern Pathology 20:467–473. Laxman, B., Tomlins, S.A., Mehra, R., et al. 2006. Noninvasive detection of TMPRSS2:ERG fusion transcripts in the urine of men with prostate cancer. Neoplasia 8:885–888. Laxman, B., Morris, D.S., Yu, J., et al. 2008. A first generation multiplex biomarker analysis of urine for the early detection of prostate cancer. Cancer Research 68:645–649. Lin, B., Ferguson, C., White, J.T., et al. 1999. Prostate-localized and androgen-regulated expression of the membrane-bound serine protease TMPRSS2. Cancer Research 59:4180–4184. Liu, A.Y., Brubaker, K.D., Goo, Y.A., et al. 2004. Lineage relationship between LNCaP and LNCaP-derived prostate cancer cell lines. Prostate 60:98–108. Marcucci, G., Baldus, C.D., Ruppert, A.S., et al. 2005. Overexpression of the ETS-related gene, ERG, predicts a worse outcome in acute myeloid leukemia with normal karyotype: a Cancer and Leukemia Group B study. Journal of Clinical Oncology 23:9234–9242. Mehra, R., Han, B., Tomlins, S.A., et al. 2007a. Heterogeneity of TMPRSS2 gene rearrangements in multifocal prostate adenocarcinoma: molecular evidence for an independent group of diseases. Cancer Research 67:7991–7995. Mehra, R., Tomlins, S.A., Shen, R., et al. 2007b. Comprehensive assessment of TMPRSS2 and ETS family gene aberrations in clinically localized prostate cancer. Modern Pathology 20:538–544. Mertz, K.D., Setlur, S.R., Dhanasekaran, S.M., et al. 2007. Molecular characterization of TMPRSS2-ERG gene fusion in the NCI-H660 prostate cancer cell line: a new perspective for an old model. Neoplasia 9:200–206. Mitelman, F., Johansson, B., and Mertens, F. 2007. The impact of translocations and gene fusions on cancer causation. Nature Reviews Cancer 7:233–245. Mosquera, J.M., Perner, S., Demichelis, F., et al. 2007. Morphological features of TMPRSS2-ERG gene fusion prostate cancer. Journal of Pathology 212:91–101. Murillo, H., Schmidt, L.J., Karter, M., et al. 2006. Prostate cancer cells use genetic and epigenetic mechanisms for progression to androgen independence. Genes, Chromosomes & Cancer 45:702–716. Nam, R.K., Sugar, L., Yang, W., et al. 2007. Expression of the TMPRSS2:ERG fusion gene predicts cancer recurrence after surgery for localised prostate cancer. The British Journal of Cancer 97:1690–1695.

720

R. Wang et al.

O’Donnell, A., Judson, I., Dowsett, M., et al. 2004. Hormonal impact of the 17alpha-hydroxylase/ C(17,20)-lyase inhibitor abiraterone acetate (CB7630) in patients with prostate cancer. British Journal of Cancer 90:2317–2325. Oikawa, T., and Yamada, T. 2003. Molecular biology of the Ets family of transcription factors. Gene 303:11–34. Oudes, A.J., Roach, J.C., Walashek, L.S., et al. 2005. Application of Affymetrix array and Massively Parallel Signature Sequencing for identification of genes involved in prostate cancer progression. BMC Cancer 5:86. Paoloni-Giacobino, A., Chen, H., Peitsch, M.C., et al. 1997. Cloning of the TMPRSS2 gene, which encodes a novel serine protease with transmembrane, LDLRA, and SRCR domains and maps to 21q22.3. Genomics 44:309–320. Perner, S., Demichelis, F., Beroukhim, R., et al. 2006. TMPRSS2:ERG fusion-associated deletions provide insight into the heterogeneity of prostate cancer. Cancer Research 66:8337–8341. Perner, S., Mosquera, J.M., Demichelis, F., et al. 2007. TMPRSS2-ERG fusion prostate cancer: an early molecular event associated with invasion. American Journal of Surgical Pathology 31:882–888. Rabbitts, T.H. 1994. Chromosomal translocations in human cancer. Nature 372:143–149. Rhodes, D.R., Kalyana-Sundaram, S., Mahavisno, V., et al. 2007. Oncomine 3.0: genes, pathways, and networks in a collection of 18,000 cancer gene expression profiles. Neoplasia 9:166–180. Rowley, J.D. 1973. Letter: A new consistent chromosomal abnormality in chronic myelogenous leukaemia identified by quinacrine fluorescence and Giemsa staining. Nature 243:290–293. Salah, Z., Maoz, M., Cohen, I., et al. 2005. Identification of a novel functional androgen response element within hPar1 promoter: implications to prostate cancer progression. FASEB Journal 19:62–72. Shi, X.B., Ma, A.H., Tepper, C.G., et al. 2004a. Molecular alterations associated with LNCaP cell progression to androgen independence. Prostate 60:257–271. Shi, X., Gangadharan, B., Brass, L.F., et al. 2004b. Protease-activated receptors (PAR1 and PAR2) contribute to tumor cell motility and metastasis. Molecular Cancer Research 2:395–402. Soller, M.J., Isaksson, M., Elfving, P., et al. 2006. Confirmation of the high frequency of the TMPRSS2/ERG fusion gene in prostate cancer. Genes, Chromosomes & Cancer 45:717–719. Stegmaier, K., Wong, J.S., Ross, K.N., et al. 2007. Signature-based small molecule screening identifies cytosine arabinoside as an EWS/FLI modulator in Ewing sarcoma. PLoS Medicine 4: e122. Tantivejkul, K., Loberg, R.D., Mawocha, S.C., et al. 2005. PAR1-mediated NFkappaB activation promotes survival of prostate cancer cells through a Bcl-xL-dependent mechanism. Journal of Cellular Biochemistry 96:641–652. Thalmann, G.N., Anezinis, P.E., Chang, S.M., et al. 1994. Androgen-independent cancer progression and bone metastasis in the LNCaP model of human prostate cancer. Cancer Research 54:2577–2581. Tomlins, S.A., Rhodes, D.R., Perner, S., et al. 2005. Recurrent fusion of TMPRSS2 and ETS transcription factor genes in prostate cancer. Science 310:644–648. Tomlins, S.A., Mehra, R., Rhodes, D.R., et al. 2006. TMPRSS2:ETV4 gene fusions define a third molecular subtype of prostate cancer. Cancer Research 66:3396–3400. Tomlins, S.A., Laxman, B., Dhanasekaran, S.M., et al. 2007a. Distinct classes of chromosomal rearrangements create oncogenic ETS gene fusions in prostate cancer. Nature 448:595–599. Tomlins, S.A., Mehra, R., Rhodes, D.R., et al. 2007b. Integrative molecular concept modeling of prostate cancer progression. Nature Genetics 39:41–51. Tomlins, S.A., Laxman, B., Varambally, R., et al. 2008. The role of the TMPRSS2-ERG gene fusions in prostate cancer. Neoplasia 10:177–188. Tu, J.J., Rohan, S., Kao, J., et al. 2007. Gene fusions between TMPRSS2 and ETS family genes in prostate cancer: frequency and transcript variant analysis by RT-PCR and FISH on paraffinembedded tissues. Modern Pathology 20:921–928.

Androgen Regulation of Prostate Cancer Gene Fusions

721

Wang, J., Cai, Y., Ren, C., et al. 2006. Expression of variant TMPRSS2/ERG fusion messenger RNAs is associated with aggressive prostate cancer. Cancer Research 66:8347–8351. Wang, Q., Li, W., Liu, X.S.X., et al. 2007. A hierarchical network of transcription factors governs androgen receptor-dependent prostate cancer growth. Molecular Cell 27:380–392. Welsh, J.B., Sapinoso, L.M., Su, A.I., et al. 2001. Analysis of gene expression identifies candidate markers and pharmacological targets in prostate cancer. Cancer Research 61:5974–5978. Wilson, S., Greer, B., Hooper, J., et al. 2005. The membrane-anchored serine protease, TMPRSS2, activates PAR-2 in prostate cancer cells. The Biochemical Journal 388:967–972. van Bokhoven, A., Caires, A., Maria, M.D., et al. 2003. Spectral karyotype (SKY) analysis of human prostate carcinoma cell lines. The Prostate 57:226–244. Winnes, M., Lissbrant, E., Damber, J.E., et al. 2007. Molecular genetic analyses of the TMPRSS2ERG and TMPRSS2-ETV1 gene fusions in 50 cases of prostate cancer. Oncology Reports 17:1033–1036. Yin, Y.J., Salah, Z., Maoz, M., et al. 2003. Oncogenic transformation induces tumor angiogenesis: a role for PAR1 activation. FASEB Journal 17:163–174. Yu, Y.P., Landsittel, D., Jing, L., et al. 2004. Gene expression alterations in prostate cancer predicting tumor aggression and preceding development of malignancy. Journal of Clinical Oncology 22:2790–2799.

Androgens and the Lipogenic Switch in Prostate Cancer Johannes V. Swinnen, Koen Brusselmans, Hannelore V. Heemers, and Guido Verhoeven

Abstract Androgens have a major impact on prostate cancer cell biology and modulate a variety of key cellular processes and functions. One of the processes that are most strikingly affected is lipid biosynthesis. Through a unique indirect mechanism that involves activation of the lipogenic transcription factor SREBP, androgens coordinately stimulate the expression of more than 20 enzymes involved in lipid synthesis, and in this way they affect the entire lipogenic program in prostate cancer cells. Through additional mechanisms, including the stimulation of an ubiquitin-specific protease that removes the degradation-tag ubiquitin from lipogenic enzymes such as fatty acid synthase, an even more complex network of regulatory control is created. Progressive deregulation of this network results in a marked overexpression of lipogenic enzymes, referred to as the lipogenic switch. This switch typically accompanies the development and progression of prostate cancer and is thought to play an active role in prostate cancer cell biology. In fact, interference with the lipogenic process impairs proper membrane formation and functioning, halts cell proliferation, and induces cell death selectively in cancer cells. These findings suggest that enhanced lipogenesis in cancer cells is an essential trait of prostate cancer progression and is a promising novel target for antineoplastic intervention.

1 Introduction To gain a better understanding of the impact of androgens on prostate cancer cell biology and to exploit this knowledge to improve the clinical management of this devastating disease it is important to identify key cellular processes that are J.V. Swinnen(*) Katholieke Universiteit Leuven, Laboratory for Experimental Medicine and Endocrinology, Gasthuisberg O&N1, Herestraat 49 bus 902, B-3000, Leuven, Belgium, E-mail: Johan.Swinnen@ med.kuleuven.be

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4_31, # Springer Science + Business Media, LLC 2009

723

724

J.V. Swinnen et al.

controlled by androgens in tumor cells, to elucidate the mechanistic networks that underlie this regulation, and to establish the role of these processes in cancer cell biology. One such process that has emerged from these studies is the biosynthesis of lipids. Lipids are essential components of every cell. As building blocks of membranes they play a vital role in establishing the interface between the cell and its environment and are essential for compartmentalization of key cellular functions. Lipids are also used as a medium to store cellular energy; they function as second messengers and are used as anchors to target proteins to membranes. In contrast to normal cells, which obtain the bulk of the required lipids from the circulation, cancer cells markedly overexpress key enzymes involved in lipid biosynthesis and tend to produce their lipids de novo. In prostate cancer cells androgens have been shown to elicit a major stimulatory effect on de novo lipogenesis. Through a unique cascade of events they activate a key lipogenic transcription factor that in turn coordinately stimulates the entire lipogenic program. Through additional mechanisms including the stimulation of an ubiquitin-specific protease that removes the degradation-tag ubiquitin from lipogenic enzymes such as fatty acid synthase, an even more complex network of regulatory control is created. Intriguingly, this network is further modified as cancer progresses toward a more advanced stage, leading to even more pronounced overexpression of lipogenic enzymes after acquisition of a castration recurrent stage. As interference with this pathway by RNA interference or by the use of small molecule inhibitors impairs proper membrane formation and functioning, halts cell proliferation, and induces cell death selectively in cancer cells, enhanced lipogenesis in cancer cells is thought to play a key role in cancer progression and is a promising novel target for antineoplastic intervention. In this chapter we will summarize the available evidence for androgen regulation of this key metabolic pathway in prostate cancer cells and describe our current understanding of the underlying mechanisms. Furthermore, we will discuss how enhanced lipogenesis affects cancer cell biology and by which way interference with lipogenesis may open new avenues for cancer treatment.

2 Lipid Biosynthesis: An Androgen-Regulated Process in Prostate Cancer Cells Seminal to the identification of the lipogenesis pathway as a major androgenregulated target in prostate cancer cells was the observation that in LNCaP prostate cancer cells (one of the most widely used in vitro models for androgen-responsive prostate cancer) exposure to androgens leads to a remarkable appearance of lipid droplets, which stain positive with the lipophilic dye Oil red O (Swinnen et al. 1996). In support of the involvement of the androgen receptor (AR), the AR antagonist Casodex (bicalutamide) abolished the stimulatory effects of androgens. Moreover, no induction of lipid accumulation was observed in AR-negative pros-

Androgens and the Lipogenic Switch in Prostate Cancer

725

tatic cell lines. Analysis of the nature and of the origin of the lipids accumulated in lipid droplets revealed that they consist mainly of triglycerides and cholesteryl esters. These represent storage products of fatty acids and cholesterol, respectively. Experiments with labeled lipid precursors revealed that both fatty acids and cholesterol are actively synthesized de novo in prostate cancer cells and that androgens markedly stimulate their synthesis. The majority of the newly synthesized fatty acids were found to be incorporated into phospholipids, which together with cholesterol, are the main lipid components of cellular membranes. Consistent with these findings androgens increased the cellular content of phospholipids and cholesterol by twofold. The excess of lipids that are not used for membrane synthesis is stored in lipid droplets. Biosynthesis of lipids (both phospholipids and cholesterol) is a complex metabolic process that involves several enzymes (Fig. 1). Interestingly, androgens have been found to stimulate the expression of all the enzymes tested (Swinnen et al. 1997a, b, Brusselmans et al. 2007). Similarly, androgen regulation of several other genes involved in lipid binding, uptake, metabolism, and transport has been observed in several independent screens including SAGE and DNA microarray studies (Swinnen et al. 1994; Nelson et al. 2002; Segawa et al. 2002). The enzyme that has been studied most intensively in terms of androgen regulation is fatty acid synthase (FASN), the main enzyme involved in the synthesis of saturated fatty acids (Swinnen et al. 1997a). Androgen stimulation of this enzyme has been demonstrated in a wide variety of androgen-responsive prostate cancer cell lines including LNCaP, MDA-PCa-2a, MDA-PCa-2b, PC-346c, 22rv1, and LAPC4 (Heemers et al. 2001; Myers et al. 2001; Pizer et al. 2001; Schmidt et al. 2007). Also, androgens have been found to stimulate lipogenic gene expression in prostate cancer cells in vivo. This is for instance the case in the androgen-responsive CWR22 and LNCaP prostate cancer xenografts (Myers et al. 2001; Ettinger et al. 2004) and in human prostate cancer specimens upon androgen ablation (Pizer et al. 2001; Ettinger et al. 2004). Also, androgens (as well as progestagens) have been shown to stimulate lipogenic gene expression in breast cancer cells (Joyeux et al. 1989; Chambon et al. 1989). Effects of androgens on lipogenesis are not limited to tumor cells, since several lipogenic enzymes are androgen regulated also in normal androgen-responsive tissues in vivo (Heemers et al. 2003).

3 Androgens Coordinately Stimulate Lipogenic Gene Expression Through Activation of the SREBP Pathway In classical lipogenic tissues including liver and adipose tissue, lipogenic genes have been shown to be coordinately regulated at the transcriptional level. Key players in this process are the Sterol Regulatory Element-Binding Proteins (SREBPs) (Horton et al. 2002). SREBPs constitute a family of three basic helixloop-helix leucine zipper lipogenic transcription factors (SREBP-1a, SREBP-1c, SREBP-2). All isoforms are synthesized as inactive precursor proteins that are

726

J.V. Swinnen et al. Phospholipids (membranes) Triglycerides (storage)

Cholesteryl esters (storage)

β-oxidation

Acylated proteins

Fatty acids Transport proteins Glucose

Elongases, desaturases

Palmitic acid NADP+ malic enzyme

FASN Citrate

NADPH Malonyl-CoA

ACL

ACC

Acetyl-CoA HMG - CoA-S HMG-CoA Acetate Amino acids Fatty acids

HMG - CoA-R Mevalonate FPS Farnesyl-PP SQS Isoprenoids

Prenylated proteins

Cholesterol Cholesteryl esters (storage) Membranes

Fig. 1 Metabolic pathways of lipid synthesis. The key substrate for both fatty acid synthesis and cholesterol synthesis is acetyl-CoA. Acetyl-CoA is derived from nutrients such as glucose, acetate, amino acids, and fatty acids. Acetyl-CoA carboxylase (ACC) extends acetyl-CoA to malonyl-CoA and is used by fatty acid synthase (FASN) to add in a cyclical manner C2 units to the acetyl-CoA precursor, ultimately yielding long-chain saturated fatty acids, predominantly palmitic acid. Malic enzyme is one of the enzymes providing NADPH. Palmitic acid is activated to palmitoyl-CoA, is elongated by elongases, and desaturated by desaturases. The resulting acyl chains are used to synthesize phospholipids or triglycerides, to serve as anchors to target proteins to membranes, or to be degraded by b-oxidation. Alternatively two acetyl-CoA molecules condense to form acetoacetylCoA which enters the mevalonate/cholesterol synthesis pathway. Key enzymes in this pathway are HMG-CoA synthase (HMG-CoA-S), HMG-CoA reductase (HMG-CoA-R), farnesyl diphosphate synthase (FPS), and squalene synthase (SQS). Intermediates of this pathway are used for protein prenylation. Cholesterol is incorporated into membranes or is converted to cholesteryl esters. All indicated enzymes have been shown to be regulated by androgens in prostate cancer cells

Androgens and the Lipogenic Switch in Prostate Cancer

727

anchored to the membranes of the endoplasmic reticulum (ER), where they interact with an SREBP-cleavage-activating protein (SCAP), that is retained in the ER by Insig retention proteins (Fig. 2). One typical feature of this complex is that it is stabilized by cholesterol. Upon cholesterol depletion, the SREBP–SCAP complex is released from the Insig retention protein and travels to the Golgi apparatus. There, by a two-step mechanism of regulated intramembrane proteolysis (Rip) an amino-terminal SREBP fragment is released (Brown et al. 2000). This fragment represents the active transcription factor. Upon release it is translocated to the nucleus and (depending on the isoform) binds to specific DNA sequences present in genes involved in the synthesis, metabolism, binding, and transport of fatty acids and cholesterol. There is substantial evidence that the lipogenic effects of androgens are mediated largely by activation of the SREBP pathway. Androgens stimulate nuclear accumulation of mature SREBP. Deletion or mutation of the SREBP-binding sites in the affected target genes abolishes androgen regulation, and introduction of a dominantnegative form of SREBP suppresses the stimulatory effects of androgens (Swinnen et al. 1997b; Heemers et al. 2003; Brusselmans et al. 2007). Androgens are thought to act on the SREBP pathway primarily by stimulating the expression of SCAP, which is a direct target of the AR. The AR binds to a functional ARE in intron 8 of the SCAP gene and stimulates its transcription in promoter-reporter assays (Heemers et al. 2004). In addition to modulation of SCAP expression, androgens also cause a switch in isoform expression of the insig retention protein (Heemers et al. 2003; 2006). Together, these changes result in a shift in the balance between the SREBP– SCAP complex and the retention protein, such that after androgen stimulation, not all SCAP can be retained by the retention protein. The excess of SCAP is free to escort the SREBP precursor to the Golgi apparatus, resulting in proteolytic maturation and enhanced lipogenic gene transcription. Androgens also stimulate the expression of SREBP-1c and SREBP-2 precursors by to positive feed-forward regulation, further modulating this pathway. The identification of the SREBPs as mediators of the effects of androgens on lipogenic gene expression has unveiled an important indirect mechanism of androgen action whereby androgens through activation of a secondary transcription factor coordinately regulate the expression of a large set of genes. Importantly, androgen activation of the SREBP pathway and coordinate regulation of lipogenic genes has also been observed in vivo in clinical prostate cancer tissue (Pizer et al. 2001; Ettinger et al. 2004). Moreover, progestagens have been shown to activate lipogenesis in adipocytes and in breast cancer cells through a similar mechanism (Lacasa et al. 2001).

4 Additional Layers of Control Besides coordinate regulation of lipogenic gene expression by activation of the SREBP pathway, androgens may fine-tune the expression of individual lipogenic enzymes by acting at additional levels of control. In fact it cannot be excluded that

728

J.V. Swinnen et al.

Androgens Akt protease SREBP

SCAP

INSIG

ER

GOLGI

NUCLEUS

enhanced transcription

mRNA increased copy number

SREBP-BS

FASN ACC ACL ....

Steroids mRNA stabilization

mTOR

Akt Androgens

enhanced translation

protein Androgens

USP2a

protein stabilization

FASN ACC ACL ...

Akt

activation by phosphorylation

Fig. 2 Androgen-dependent and independent mechanisms leading to high-level expression of lipogenic enzymes in prostate cancer cells. In prostate cancer cells, androgens stimulate the expression of the SREBP escort protein SCAP and cause an isoform switch of the insig retention proteins. This results in a shift in the balance between the SREBP–SCAP complex and the retention protein. The excess of SCAP is free to escort the SREBP precursor to the Golgi apparatus, resulting in proteolytic maturation and enhanced lipogenic gene transcription. In breast cancer cells, steroids have also been shown to stabilize the mRNA encoding FASN. They also

Androgens and the Lipogenic Switch in Prostate Cancer

729

the AR binds directly to specific individual lipogenic genes and affects their expression. Moreover, in the case of FASN, androgens and other steroid hormones have been shown to further enhance its expression by stabilization of the mRNA (Joyeux et al. 1989). Interestingly, both androgen- and progestagen-mediated induction of lipogenesis have been shown to be accompanied by a marked increase in the expression of the lipogenic activator Spot 14 in breast and prostate cancer cells, suggesting that steroid hormones may also enhance lipid synthesis in cancer cells via this factor (Heemers et al. 2000; Martel et al. 2006; Kinlaw et al. 2007). By crosstalk with growth factor signaling pathways and stimulation of mammalian target of rapamycin (mTOR) androgens may enhance the translation of mRNAs encoding lipogenic enzymes such as FASN and ACC (Yoon et al. 2007). An additional level of control is established by the androgen regulation of the isopeptidase USP2a. USP2a is a preproteasomal ubiquitin-specific protease that removes the proteasome degradation tag ubiquitin from FASN, resulting in stabilization of the enzyme (Graner et al. 2004). This way, by acting at several levels of lipogenic enzyme regulation, a hierarchical network of regulatory control is established that enables androgens to both coordinately regulate the entire lipogenic pathway and to cause fine-tuning of individual enzymes (Fig. 2).

5 Increased Lipogenic Enzyme Expression as a Trait of Cancer Development and Progression Under normal physiological conditions the majority of cells and tissues in our body obtain the bulk of the required lipids from the circulation. Accordingly, the expression and activity of lipogenic enzymes is low and is primarily related to specific functions and/or conditions (Kuhajda 2000; Swinnen et al. 2006). Active fatty acid synthesis is found mainly in liver and adipose tissue, particularly under low fat and high carbohydrate dietary conditions. Fatty acid synthesis is also active in the lungs where it is involved in surfactant production, in the lactating breast to produce fatty acids for milk lipids, and in the cycling endometrium particularly during the proliferative phase. Interestingly, in contrast to normal tissues, tumor tissues display a high rate of fatty acid synthesis. Several decades ago it was observed that in cancer cells most lipids are derived from de novo synthesis, irrespective of their supply via the circulation (reviewed in Swinnen et al. 2006; Menendez and Lupu 2007). These changes are now known to be related to the marked overexpression and activation of lipogenic enzymes that is seen in many cancer types. Again, the enzyme that has been studied most extensively in this respect is FASN. As shown Fig. 2 (Continued) enhance the expression of the ubiquitin-specific protease USP2a, stabilizing FASN at the protein level. Activation of growth factor signaling in cancer cells in part mediated by Akt, further enhances SREBP activation, and together with androgens may stimulate mTOR enhancing mRNA translation. Akt also directly activates lipogenic enzymes such as ATP-citrate lyase (ACL) by protein phosphorylation. Overexpression of USP2a in prostate cancer cells and an increase of the copy number of FASN may further enhance lipogenesis in prostate cancer cells

730

J.V. Swinnen et al.

by immunohistochemistry, FASN is overexpressed in cancer of the breast, prostate, endometrium, ovary, lung, colon, bladder, stomach, kidney, liver, pancreas, esophagus, oral cavity, oral tongue, and skin, in retinoblastoma, mesothelioma, glioma and nephroblastoma, in Paget’s disease of the vulva, and in chemically and hormonally induced liver tumors of the rat (Swinnen et al. 2006). In many cancer types, FASN expression is already elevated very early in cancer development. In prostate cancer, overexpression of FASN has been observed in low-grade and high-grade PIN (Swinnen et al. 2002). Depending on the antibody, the immunohistochemical technique and on the freshness of the tissue, increased FASN expression is detected in 50–90% of invasive prostatic carcinomas (Epstein et al. 1995; Shurbaji et al. 1996; Pizer et al. 2001; Swinnen et al. 2002; Rossi et al. 2003). In several studies the expression level correlates with the Gleason grade and predicts a poor outcome (Shurbaji et al. 1996; Swinnen et al. 2002). Overexpression of FASN can be observed not only at the protein level but also at the mRNA level (Swinnen et al. 2000b), as confirmed in several recent DNA microarray studies (Bull et al. 2001; Nelson et al. 2002; Prowatke et al. 2007). Besides FASN several other lipogenic enzymes, including ATP citrate lyase (ACL) and acetyl-CoA carboxylase-alpha (ACC-alpha), show enhanced expression and/or activation (Szutowicz et al. 1979; Milgraum et al. 1997; Swinnen et al. 2000b; Turyn et al. 2003; Yahagi et al. 2005).

6 Mechanisms Underlying Enhanced Lipogenesis in Cancer Several recent studies indicate that the mechanisms underlying the activation of lipogenic pathways in tumor cells or the so called lipogenic switch may be complex. Initially, enhanced activation of lipogenesis in tumor cells has been linked to aberrant growth factor signaling that is observed in many tumors (Fig. 2). Growth factors that activate lipogenic pathways in prostate cancer cells include epidermal growth factor (EGF) and insulin-like growth factor-1 (IGF-1) (Swinnen et al. 2000a, 2006). Growth factor activation of lipogenesis is reminiscent of the activation of lipogenesis by insulin in liver and adipose tissue and has been shown to involve activation of sterol regulatory element-binding proteins (SREBPs) (Swinnen et al. 2000a). Signaling pathways that are involved include the mitogenactivated protein kinase (MAPK) pathway and the phosphatidylinositol 30 -kinase (PI3K)/Akt pathway. Particularly this latter pathway is frequently activated in prostate tumors, in part through deletion or mutation of PTEN, the enzyme that catalyzes the reverse reaction of PI3K. The high expression of FASN and activation of the SREBP pathway in LNCaP cells is in part related to the loss of PTEN. Reintroduction of wild-type PTEN or chemical inhibition of PI3K reduced FASN levels (Van de Sande et al. 2002). Cotransfection with a constitutively active form of Akt restored FASN expression. Similarly, using an inducible form of Akt, Porstmann et al. demonstrated that activation of Akt resulted in overexpression of FASN through activation of SREBPs (Porstmann et al. 2005). In clinical prostate cancers FASN overexpression was associated with activation of Akt and inversely

Androgens and the Lipogenic Switch in Prostate Cancer

731

correlated with levels of PTEN (Bandyopadhyay et al. 2005; Van de Sande et al. 2005; Wang et al. 2005). Besides activation of the SREBP pathway, Akt may enhance lipogenesis by stimulating the translation of lipogenic enzymes as shown in breast cancer cells (Yoon et al. 2007). Also, by direct phosphorylation and activation of ACL, Akt may promote the supply of lipid precursors, thereby linking lipogenesis to the glycolytic pathway, which is also activated by Akt (Bauer et al. 2005). Interestingly, Akt also enhances protein synthesis via the mTOR pathway (Inoki et al. 2002; Hahn-Windgassen et al. 2005), suggesting that Akt-mediated stimulation of lipogenesis in cancer cells is part of a more general metabolic switch enabling cancer cells to enhance their anabolic activity and to thrive even under hypoxic conditions. Besides the Akt pathway several other mechanisms may contribute to the enhanced expression and/or activity of lipogenic enzymes in cancer cells. In this respect, it is interesting that the USP2a protein, which functions as an ubiquitinspecific protease with FASN as one of its substrates, is frequently overexpressed in prostate cancer, thereby enhancing FASN expression at the protein level (Graner et al. 2004). Moreover, in approximately 25% of prostate tumors examined the gene copy number of the FASN gene appears to be increased (Shah et al. 2006). Finally, as mentioned earlier, androgens may contribute substantially to the high level of lipogenic enzyme expression in prostate cancer. Other steroid hormones may play a comparable role in other steroid-responsive tumors (Chalbos et al. 1990). In clinical prostate cancer specimens, SREBP expression, which is higher in malignant sections compared with noncancerous prostate tissue, decreased after androgen ablation (Ettinger et al. 2004). Also in the LNCaP and the CWR22 xenograft models of prostatic adenocarcinoma, fatty acid synthase expression markedly decreases after castration (Pizer et al. 2001; Myers et al. 2001; Ettinger et al. 2004). Interestingly, expression of SREBP and of lipogenic enzymes increases with the transition to an androgen recurrent stage and reaches levels that may be even greater than precastrate levels, indicating that the SREBP pathway and hence lipogenesis and cholesterol synthesis are dysregulated during progression to androgen independence (Ettinger et al. 2004).

7 Role of Enhanced Lipogenesis in Cancer Development and Progression Despite the fact that overexpression of lipogenic enzymes is one of the most common molecular changes accompanying the development and progression of cancer, the role of tumor-associated lipogenesis is just starting to be elucidated. From studies with labeled lipid precursors it is clear that in most tumor cells examined the majority of newly synthesized fatty acids are incorporated into phospholipids, the major building blocks of membranes (Pizer et al. 2001; Swinnen et al. 2003). Based on these findings and a series of observations indicating that inhibition of lipogenesis by chemical inhibitors or RNA interference blocks cell

732

J.V. Swinnen et al.

proliferation and ultimately induces cell death, it is generally assumed that cancer cells display increased lipogenesis to serve the high rate of membrane synthesis in rapidly dividing cells. However, in most clinical prostate cancers only a relatively small fraction of the cancer cells are, at any given moment, actively proliferating, while nearly all cancer cells express high levels of fatty acid synthase (Swinnen et al. 2002). One attractive hypothesis is that cancer cells in contrast to normal cells show a less efficient uptake of lipids from the circulation. This may be due to a limited access to circulating lipids because of changing tissue architecture or to changes in the mechanisms involved in lipid uptake (Kinlaw et al. 2006). In this respect it is interesting that prostate tumors show a frequent loss of heterozygosity of the lipoprotein lipase (LPL) locus (Bova et al. 1993). Lipoprotein lipase is a secreted enzyme that releases fatty acids from circulating lipoprotein particles for uptake by neighboring cells. In support of this concept is the observation that inhibition of lipogenesis by small molecules or by RNA interference decreases cell size, causes endoplasmic reticulum stress, halts cell proliferation, and induces cell death selectively in cancer cells (De Schrijver et al. 2003; Brusselmans et al. 2005a; Beckers et al. 2007; Little et al. 2007). A potentially important consequence of this shift from fatty acid uptake to de novo fatty acid synthesis is a change in the degree of saturation of membrane lipids. In fact, in contrast to the lipids derived from the diet, which may be relatively rich in polyunsaturated fatty acids, de novo synthesized fatty acids in prostate cancer cells are mainly saturated, as human cells do not have the ability to synthesize polyunsaturated fatty acids de novo and as many prostate tumors show a loss of stearoyl-CoA desaturase (Moore et al. 2005). In support of this contention, inhibition of tumor-associated lipogenesis causes a decrease in saturated fatty acyl chains and enriches the cell with polyunsaturated fatty acids. This change in balance between saturated and polyunsaturated acyl chains may have major consequences for membrane structure and function. In fact, the degree of fatty acyl unsaturation significantly contributes to membrane fluidity and microdomain formation. Together with cholesterol, phospholipids with mainly saturated acyl chains tend to partition into detergent-resistant membrane microdomains. These are raft aggregates that harbor specific subsets of proteins, which also require saturated acyl chains for raft localization and are implicated in key cellular processes including intracellular trafficking, signal transduction and cell migration. In support of this contention, a substantial fraction of de novo synthesized phospholipids are found in detergent-resistant membrane microdomains, and inhibition of lipogenesis preferentially affects those phospholipids. Along these same lines, enhanced lipogenesis in cancer cells may affect the acylation of proteins, this way targeting these proteins to membrane microdomains. Besides changes in membrane microstructure, lipogenesis-induced saturation of acyl chains in membranes may affect the flip-flop rate and decrease the susceptibility to peroxidation, a process that results in the production of reactive radical molecules, which may damage proteins and DNA and may shorten the life of the cell. Besides changes in membranes, it has been proposed that tumor-associated lipogenesis plays a role in the maintenance of the redox balance in cancer cells with a high glycolytic rate under hypoxic conditions (Hochachka et al. 2002). Another hypoth-

Androgens and the Lipogenic Switch in Prostate Cancer

733

esis states that overexpression of FASN per se may affect cancer cell biology by modulating cellular energy regulation as has been demonstrated in hypothalamic cells (Chakravarthy et al. 2007). Hence it is expected that increased lipogenesis in cancer cells by causing both quantitative and qualitative changes in lipids and perhaps also through other complementary mechanisms may affect multiple key aspects of membrane and tumor cell biology, thus actively contributing to the development and the progression of cancer.

8 Lipogenesis as a Target for Antineoplastic Intervention One of the major hurdles in the clinical management of prostate cancer is the effective treatment of metastatic cancer. Although most patients initially respond favorably to androgen ablation, most relapse due to the transition to a lethal androgen-independent cancer. Interestingly, the sustained overexpression of lipogenic enzymes through all stages of the disease, together with the finding that inhibition of tumor-associated lipogenesis induces proliferative arrest and causes cell death selectively in cancer cells and not in normal cells, suggests that the lipogenic pathway is a viable target for cancer prevention and for anticancer treatment. Initial studies addressing this issue were carried out with cerulenin, a mycotoxin isolated from Cephalosporium caerulens. Cerulenin, which forms a covalent bond with the active site cysteine of the b-keto-acyl synthase subunit of fatty acid synthase and irreversibly inhibits its activity, was shown to inhibit FASN in human cancer cell lines, including prostate, in vitro and to induce growth arrest and apoptosis (Pizer et al. 2001). In contrast, nontumoral fibroblasts displaying low levels of endogenous fatty acid synthesis were hardly affected (Kuhajda et al. 1994; Pizer et al. 1998; Thupari et al. 2001). Because of its chemical instability and the presence of a highly reactive epoxide group (which may induce nonspecific binding), the potential of cerulenin for therapeutic application is, however, limited. More recently, a synthetic analog of cerulenin, C75 has become available. Also this compound inhibits growth and induces cell death in various cancer cells cultured in vitro. Interestingly, systemic treatment of nude mice bearing human cancer xenografts, including prostate, markedly reduces tumor growth in vivo (Kuhajda et al. 2000; Pizer et al. 2001; Zhou et al. 2007). Nevertheless, the selectivity and safety of this compound remains a matter of debate. In several studies C75 has been shown to induce severe weight loss and visceral illness (Loftus et al. 2000; Clegg et al. 2002). Besides these compounds, several natural polyphenolic compounds from plants have been shown to act as inhibitors of FASN and have been demonstrated to inhibit lipogenesis and to induce growth arrest and death in cancer cells (Wang and Tian 2001; Brusselmans et al. 2003; Yeh et al. 2003; Li and Tian 2004; Brusselmans et al. 2005b). Despite the many protein targets that these compounds have, the finding that the antiproliferative effect of these compounds tends to correlate with their ability to inhibit lipogenesis, suggests that inhibition of lipogenesis is at least part of the mechanism underlying the well-known antitumoral

734

J.V. Swinnen et al.

effect of many of these compounds (Brusselmans et al. 2005b). Also the antibacterial compound Triclosan has been shown to inhibit FASN, to be cytotoxic for cancer cells, and to suppress methylnitrosurea-induced carcinogenesis in Sprague–Dawley rats (Liu et al. 2002; Lu and Archer 2005). Furthermore, in a screening for inhibitors of serine hydrolases, the antiobesity drug Orlistat was identified as an inhibitor of FASN. Orlistat inhibited fatty acid synthesis and induced growth arrest and death in prostate cancer cells in vitro and delayed tumor progression of prostate cancer xenografts in mice (Kridel et al. 2004). The mechanisms by which these chemical inhibitors induce cell death selectively in cancer cells remain incompletely understood. It has been shown that inhibition of FASN in cancer cells results in an accumulation of the fatty acid precursor malonyl-CoA (Pizer et al. 2000; Thupari et al. 2001). This finding has been confirmed in several other cell types (De Schrijver et al. 2003; Bandyopadhyay et al. 2006) and is thought to contribute to induction of cell death. In fact, intracellular levels of malonyl-CoA inactivate fatty acid oxidation, activate the energy sensor AMPK, increase intracellular levels of ceramides, and induce expression of the proapoptotic genes BNIP3, TRAIL (tumor necrosis factor-related apoptosis-inducing ligand), and death-associated protein kinase-2 in breast cancer cells. The fact that cancer cells frequently show a coordinated upregulation of lipogenic enzymes is expected to result in higher levels of malonyl-CoA upon FASN inhibition in cancer cells and may explain the cancer cell-selective effects of FASN inhibitors. The recent observation that inhibition of FASN by RNAi-mediated or chemical inhibition of ACC attenuates proliferation and induces death both in prostate and breast cancer cells (Brusselmans et al. 2005a; Chajes et al. 2006; Beckers et al. 2007), indicates that besides malonyl-CoA accumulation, depletion of fatty acids per se may induce cell death upon inhibition of fatty acid synthesis in cancer cells. Tumor selectivity of the effects may be related to differences in uptake of circulating lipids as discussed earlier. Also inhibition of ACL, the enzyme that provides cytosolic acetyl-CoA for lipid synthesis, decreases lipogenesis and limits growth and survival of cancer cells in vitro and in vivo (Hatzivassiliou et al. 2005). As acetyl-CoA is not only a precursor for fatty acid synthesis but also a building block for the mevalonate pathway, inhibition of ACL may also affect cholesterol biosynthesis and protein isoprenylation. Interestingly in this respect it has been shown that the cholesterol-lowering drugs, statins, induce growth arrest and apoptosis in various tumor cell lines in vitro (Denoyelle et al. 2003; Zhong et al. 2003). For the statins it remains unclear to what extent these effects are related to reduced cholesterol synthesis or to inhibition of protein prenylation. The recent finding however that RNAi-mediated and chemical inhibition of squalene synthase (SQS), the first enzyme of the mevalonate/isoprenoid pathway that is exclusively committed to the synthesis of sterols, also limits cell proliferation and induces apoptosis in prostate cancer cells (Brusselmans et al. 2007), indicates that besides fatty acid synthesis cholesterol synthesis may be a target for anticancer therapy also. Consistent with this possibility is the observation that inhibition of cholesterol synthesis or removal of cholesterol from cellular membranes inhibits raft function and leads to growth retardation and death of cancer cells (Hager et al. 2006).

Androgens and the Lipogenic Switch in Prostate Cancer

735

Another interesting observation is that inhibition of lipogenesis markedly sensitizes cancer cells for classical antineoplastic agents including cisplatin and doxorubicin (Menendez et al. 2005; own unpublished data). It is obvious that inhibition of lipogenesis opens promising perspectives for application in the prevention and treatment of prostate cancer. Thus, there is an urgent need for more potent and more selective inhibitors and for increasing efforts in translational research.

9 Conclusions Enhanced lipogenesis is a major hallmark of cancer cells, and androgens, by acting at several levels in a regulatory network, substantially contribute to high-level activity of this process in prostate cancer cells. Interference with this process impairs proper membrane formation and functioning, halts cell proliferation, and induces cell death selectively in cancer cells. These findings suggest that enhanced lipogenesis is a fundamental aspect of prostate cancer cell biology and, provided that safer and more selective inhibitors will become available, may represent an exploitable target for cancer intervention. Acknowledgments The authors are supported by a grant ‘‘Concerted Research Action’’ by the K.U. Leuven and by research grants from the Research Foundation-Flanders (FWO) (Belgium). K. Brusselmans is a postdoctoral fellow of the Research Foundation-Flanders (FWO) (Belgium).

References Bandyopadhyay S, Pai SK, Watabe M, Gross SC, Hirota S, Hosobe S, Tsukada T, Miura K, Saito K, Markwell SJ, Wang Y, Huggenvik J, Pauza ME, Iiizumi M, Watabe K (2005) FAS expression inversely correlates with PTEN level in prostate cancer and a PI 3-kinase inhibitor synergizes with FAS siRNA to induce apoptosis. Oncogene 24: 5389–5395. Bandyopadhyay S, Zhan R, Wang Y, Pai SK, Hirota S, Hosobe S, Takano Y, Saito K, Furuta E, Iiizumi M, Mohinta S, Watabe M, Chalfant C, Watabe K (2006) Mechanism of apoptosis induced by the inhibition of fatty acid synthase in breast cancer cells. Cancer Res 66: 5934– 5940. Bauer DE, Hatzivassiliou G, Zhao F, Andreadis C, Thompson CB (2005) ATP citrate lyase is an important component of cell growth and transformation. Oncogene 24: 6314–6322. Beckers A, Organe S, Timmermans L, Scheys K, Peeters A, Brusselmans K, Verhoeven G, Swinnen JV (2007) Chemical inhibition of acetyl-CoA carboxylase induces growth arrest and cytotoxicity selectively in cancer cells. Cancer Res 67: 8180–8187. Bova GS, Carter BS, Bussemakers MJ, Emi M, Fujiwara Y, Kyprianou N, Jacobs SC, Robinson JC, Epstein JI, Walsh PC, et al. (1993) Homozygous deletion and frequent allelic loss of chromosome 8p22 loci in human prostate cancer. Cancer Res 53: 3869–3873. Brown MS, Ye J, Rawson RB, Goldstein JL (2000) Regulated intramembrane proteolysis: a control mechanism conserved from bacteria to humans. Cell 100: 391–398. Brusselmans K, De Schrijver E, Heyns W, Verhoeven G, Swinnen JV (2003) Epigallocatechin-3gallate is a potent natural inhibitor of fatty acid synthase in intact cells and selectively induces apoptosis in prostate cancer cells. Int J Cancer 106: 856–862.

736

J.V. Swinnen et al.

Brusselmans K, De Schrijver E, Verhoeven G, Swinnen JV (2005a) RNA interference-mediated silencing of the acetyl-CoA-carboxylase-alpha gene induces growth inhibition and apoptosis of prostate cancer cells. Cancer Res 65: 6719–6725. Brusselmans K, Vrolix R, Verhoeven G, Swinnen JV (2005b) Induction of cancer cell apoptosis by flavonoids is associated with their ability to inhibit fatty acid synthase activity. J Biol Chem 280: 5636–5645. Brusselmans K, Timmermans L, Van de Sande T, Van Veldhoven PP, Guan G, Shechter I, Claessens F, Verhoeven G, Swinnen JV (2007) Squalene synthase, a determinant of Raftassociated cholesterol and modulator of cancer cell proliferation. J Biol Chem 282: 18777– 18785. Bull JH, Ellison G, Patel A, Muir G, Walker M, Underwood M, Khan F, Paskins L (2001) Identification of potential diagnostic markers of prostate cancer and prostatic intraepithelial neoplasia using cDNA microarray. Br J Cancer 84: 1512–1519. Chajes V, Cambot M, Moreau K, Lenoir GM, Joulin V (2006) Acetyl-CoA carboxylase alpha is essential to breast cancer cell survival. Cancer Res 66: 5287–5294. Chakravarthy MV, Zhu Y, Lo´pez M, Yin L, Wozniak DF, Coleman T, Hu Z, Wolfgang M, VidalPuig A, Lane MD, Semenkovich CF (2007) Brain fatty acid synthase activates PPARalpha to maintain energy homeostasis. J Clin Invest 117: 2539–2552. Chalbos D, Joyeux C, Galtier F, Escot C, Chambon M, Maudelonde T, Rochefort H (1990) Regulation of fatty acid synthetase by progesterone in normal and tumoral human mammary glands. Rev Esp Fisiol 46: 43–46. Chambon M, Rochefort H, Vial HJ, Chalbos D (1989) Progestins and androgens stimulate lipid accumulation in T47D breast cancer cells via their own receptors. J Steroid Biochem 33: 915–922. Clegg DJ, Wortman MD, Benoit SC, McOsker CC, Seeley RJ (2002) Comparison of central and peripheral administration of C75 on food intake, body weight, and conditioned taste aversion. Diabetes 51: 3196–3201. De Schrijver E, Brusselmans K, Heyns W, Verhoeven G, Swinnen JV (2003) RNA interferencemediated silencing of the fatty acid synthase gene attenuates growth and induces morphological changes and apoptosis of LNCaP prostate cancer cells. Cancer Res 63: 3799–3804. Denoyelle C, Albanese P, Uzan G, Hong L, Vannier JP, Soria J, Soria C (2003) Molecular mechanism of the anti-cancer activity of cerivastatin, an inhibitor of HMG-CoA reductase, on aggressive human breast cancer cells. Cell Signal 15: 327–338. Epstein JI, Carmichael M, Partin AW (1995) OA-519 (fatty acid synthase) as an independent predictor of pathologic state in adenocarcinoma of the prostate. Urology 45: 81–86. Ettinger SL, Sobel R, Whitmore TG, Akbari M, Bradley DR, Gleave ME, Nelson CC (2004) Dysregulation of sterol response element-binding proteins and downstream effectors in prostate cancer during progression to androgen independence. Cancer Res 64: 2212–2221. Graner E, Tang D, Rossi S, Baron A, Migita T, Weinstein LJ, Lechpammer M, Huesken D, Zimmermann J, Signoretti S, Loda M (2004) The isopeptidase USP2a regulates the stability of fatty acid synthase in prostate cancer. Cancer Cell 5: 253–261. Hager MH, Solomon KR, Freeman MR (2006) The role of cholesterol in prostate cancer. Curr Opin Clin Nutr Metab Care 9: 379–385. Hahn-Windgassen A, Nogueira V, Chen CC, Skeen JE, Sonenberg N, Hay N (2005) Akt activates the mammalian target of rapamycin by regulating cellular ATP level and AMPK activity. J Biol Chem 280: 32081–32089. Hatzivassiliou G, Zhao F, Bauer DE, Andreadis C, Shaw AN, Dhanak D, Hingorani SR, Tuveson DA, Thompson CB (2005) ATP citrate lyase inhibition can suppress tumor cell growth. Cancer Cell 8: 311–321. Heemers H, Vanderhoydonc F, Heyns W, Verhoeven G, Swinnen JV (2000) Progestins and androgens increase expression of Spot 14 in T47-D breast tumor cells. Biochem Biophys Res Commun 269: 209–212.

Androgens and the Lipogenic Switch in Prostate Cancer

737

Heemers H, Maes B, Foufelle F, Heyns W, Verhoeven G, Swinnen JV (2001) Androgens stimulate lipogenic gene expression in prostate cancer cells by activation of the sterol regulatory element-binding protein cleavage activating protein/sterol regulatory element-binding protein pathway. Mol Endocrinol 15: 1817–1828. Heemers H, Vanderhoydonc F, Roskams T, Shechter I, Heyns W, Verhoeven G, Swinnen JV (2003) Androgens stimulate coordinated lipogenic gene expression in normal target tissues in vivo. Mol Cell Endocrinol 205: 21–31. Heemers H, Verrijdt G, Organe S, Claessens F, Heyns W, Verhoeven G, Swinnen JV (2004) Identification of an androgen response element in intron 8 of the sterol regulatory elementbinding protein cleavage-activating protein gene allowing direct regulation by the androgen receptor. J Biol Chem 279: 30880–30887. Heemers HV, Verhoeven G, Swinnen JV (2006) Androgen activation of the sterol regulatory element-binding protein pathway: current insights. Mol Endocrinol 20: 2265–2277. Hochachka PW, Rupert JL, Goldenberg L, Gleave M, Kozlowski P (2002) Going malignant: the hypoxia-cancer connection in the prostate. Bioessays 24: 749–757. Horton JD, Goldstein JL, Brown MS (2002) SREBPs: activators of the complete program of cholesterol and fatty acid synthesis in the liver. J Clin Invest 109: 1125–1131. Inoki K, Li Y, Zhu T, Wu J, Guan KL (2002) TSC2 is phosphorylated and inhibited by Akt and suppresses mTOR signalling. Nat Cell Biol 4: 648–657. Joyeux C, Rochefort H, Chalbos D (1989) Progestin increases gene transcription and messenger ribonucleic acid stability of fatty acid synthetase in breast cancer cells. Mol Endocrinol 10: 681–686. Kinlaw WB, Quinn JL, Wells WA, Roser-Jones C, Moncur JT (2006) Spot 14: a marker of aggressive breast cancer and a potential therapeutic target. Endocrinology 147: 4048–4055. Kridel SJ, Axelrod F, Rozenkrantz N, Smith JW (2004) Orlistat is a novel inhibitor of fatty acid synthase with antitumor activity. Cancer Res 64: 2070–2075. Kuhajda FP, Jenner K, Wood FD, Hennigar RA, Jacobs LB, Dick JD, Pasternack GR (1994) Fatty acid synthesis: a potential selective target for antineoplastic therapy. Proc Natl Acad Sci U S A 91: 6379–6383. Kuhajda FP (2000) fatty-acid synthase and human cancer: new perspectives on its role in tumor biology. Nutrition 16: 202–208. Kuhajda FP, Pizer ES, Li JN, Mani NS, Frehywot GL, Townsend CA (2000) Synthesis and antitumor activity of an inhibitor of fatty acid synthase. Proc Natl Acad Sci U S A 97: 3450– 3454. Lacasa D, Le Liepvre X, Ferre P, Dugail I (2001) Progesterone stimulates adipocyte determination and differentiation factor 1/sterol regulatory element-binding protein 1c gene expression. Potential mechanism for the lipogenic effect of progesterone in adipose tissue. J Biol Chem 276: 11512–11516. Li BH, Tian WX (2004) Inhibitory effects of flavonoids on animal fatty acid synthase. J Biochem (Tokyo) 135: 85–91. Little JL, Wheeler FB, Fels DR, Koumenis C, Kridel SJ (2007) Inhibition of fatty acid synthase induces endoplasmic reticulum stress in tumor cells. Cancer Res 67: 1262–1269. Liu B, Wang Y, Fillgrove KL, Anderson VE (2002) Triclosan inhibits enoyl-reductase of type I fatty acid synthase in vitro and is cytotoxic to MCF-7 and SKBr-3 breast cancer cells. Cancer Chemother Pharmacol 49: 187–193. Loftus TM, Jaworsky DE, Frehywot GL, Townsend CA, Ronnett GV, Lane MD, Kuhajda FP (2000) Reduced food intake and body weight in mice treated with fatty acid synthase inhibitors. Science 288: 2379–2381. Lu S, Archer MC (2005) Fatty acid synthase is a potential molecular target for the chemoprevention of breast cancer. Carcinogenesis 26: 153–157. Martel PM, Bingham CM, McGraw CJ, Baker CL, Morganelli PM, Meng ML, Armstrong JM, Moncur JT, Kinlaw WB (2006) S14 protein in breast cancer cells: direct evidence of regulation by SREBP-1c, superinduction with progestin, and effects on cell growth. Exp Cell Res 312: 278–288.

738

J.V. Swinnen et al.

Menendez JA, Lupu R (2007) Fatty acid synthase and the lipogenic phenotype in cancer pathogenesis. Nat Rev Cancer 7: 763–777. Menendez JA, Vellon L, Colomer R, Lupu R (2005) Pharmacological and small interference RNA-mediated inhibition of breast cancer-associated fatty acid synthase (oncogenic antigen519) synergistically enhances Taxol (paclitaxel)-induced cytotoxicity. Int J Cancer 115: 19–35. Milgraum LZ, Witters LA, Pasternack GR, Kuhajda FP (1997) Enzymes of the fatty acid synthesis pathway are highly expressed in in situ breast carcinoma. Clin Cancer Res 3: 2115–2120. Moore S, Knudsen B, True LD, Hawley S, Etzioni R, Wade C, Gifford D, Coleman I, Nelson PS (2005) Loss of stearoyl-CoA desaturase expression is a frequent event in prostate carcinoma. Int J Cancer 114: 563–571. Myers RB, Oelschlager DK, Weiss HL, Frost AR, Grizzle WE (2001) Fatty acid synthase: an early molecular marker of progression of prostatic adenocarcinoma to androgen independence. J Urol 165: 1027–1032. Nelson PS, Clegg N, Arnold H, Ferguson C, Bonham M, White J, Hood L, Lin B (2002) The program of androgen-responsive genes in neoplastic prostate epithelium. Proc Natl Acad Sci U S A 99: 11890–11895. Pizer ES, Chrest FJ, DiGiuseppe JA, Han WF (1998) Pharmacological inhibitors of mammalian fatty acid synthase suppress DNA replication and induce apoptosis in tumor cell lines. Cancer Res 58: 4611–4615. Pizer ES, Thupari J, Han WF, Pinn ML, Chrest FJ, Frehywot GL, Townsend CA, Kuhajda FP (2000) Malonyl-coenzyme-A is a potential mediator of cytotoxicity induced by fatty-acid synthase inhibition in human breast cancer cells and xenografts. Cancer Res 60: 213–218. Pizer ES, Pflug BR, Bova GS, Han WF, Udan MS, Nelson JB (2001) Increased fatty acid synthase as a therapeutic target in androgen-independent prostate cancer progression. Prostate 47: 102–110. Porstmann T, Griffiths B, Chung YL, Delpuech O, Griffiths JR, Downward J, Schulze A (2005) PKB/Akt induces transcription of enzymes involved in cholesterol and fatty acid biosynthesis via activation of SREBP. Oncogene 24: 6465–6481. Prowatke I, Devens F, Benner A, Grone EF, Mertens D, Grone HJ, Lichter P, Joos S (2007) Expression analysis of imbalanced genes in prostate carcinoma using tissue microarrays. Br J Cancer 96: 82–88. Rossi S, Graner E, Febbo P, Weinstein L, Bhattacharya N, Onody T, Bubley G, Balk S, Loda M (2003) Fatty acid synthase expression defines distinct molecular signatures in prostate cancer. Mol Cancer Res 1: 707–715. Schmidt LC, Ballman KV, Tindall DJ (2007) Inhibition of fatty acid synthase activity in prostate cancer cells by dutasteride. Prostate 67: 1111–1120. Segawa T, Nau ME, Xu LL, Chilukuri RN, Makarem M, Zhang W, Petrovics G, Sesterhenn IA, McLeod DG, Moul JW, Vahey M, Srivastava S (2002) Androgen-induced expression of endoplasmic reticulum (ER) stress response genes in prostate cancer cells. Oncogene 21: 8749–8758. Shah US, Dhir R, Gollin SM, Chandran UR, Lewis D, Acquafondata M, Pflug BR (2006) Fatty acid synthase gene overexpression and copy number gain in prostate adenocarcinoma. Hum Pathol 37: 401–409. Shurbaji MS, Kalbfleisch JH, Thurmond TS (1996) Immunohistochemical detection of a fatty acid synthase (OA-519) as a predictor of progression of prostate cancer. Hum Pathol 27: 917–921. Swinnen JV, Esquenet M, Heyns W, Rombauts W, Verhoeven G (1994) Androgen regulation of the messenger RNA encoding diazepam-binding inhibitor/acyl-CoA-binding protein in the human prostatic adenocarcinoma cell line LNCaP. Mol Cell Endocrinol 104: 153–162. Swinnen JV, VanVeldhoven PP, Esquenet M, Heyns W, Verhoeven G (1996) Androgens markedly stimulate the accumulation of neutral lipids in the human prostatic adenocarcinoma cell line LNCaP. Endocrinology 137: 4468–4474. Swinnen JV, Esquenet M, Goossens K, Heyns W, Verhoeven G (1997a) Androgens stimulate fatty acid synthase in the human prostate cancer cell line LNCaP. Cancer Res 57: 1086–1090.

Androgens and the Lipogenic Switch in Prostate Cancer

739

Swinnen JV, Ulrix W, Heyns W, Verhoeven G (1997b) Coordinate regulation of lipogenic gene expression by androgens: evidence for a cascade mechanism involving sterol regulatory element binding proteins. Proc Natl Acad Sci U S A 94: 12975–12880. Swinnen JV, Heemers H, Deboel L, Foufelle F, Heyns W, Verhoeven G (2000a) Stimulation of tumor-associated fatty acid synthase expression by growth factor activation of the sterol regulatory element-binding protein pathway. Oncogene 19: 5173–5181. Swinnen JV, Vanderhoydonc F, Elgamal AA, Eelen M, Vercaeren I, Joniau S, Van Poppel H, Baert L, Goossens K, Heyns W, Verhoeven G (2000b) Selective activation of the fatty acid synthesis pathway in human prostate cancer. Int J Cancer 88: 176–179. Swinnen JV, Roskams T, Joniau S, Van Poppel H, Oyen R, Baert L, Heyns W, Verhoeven G (2002) Overexpression of fatty acid synthase is an early and common event in the development of prostate cancer. Int J Cancer 98: 19–22. Swinnen JV, Van Veldhoven PP, Timmermans L, De Schrijver E, Brusselmans K, Vanderhoydonc F, Van de Sande T, Heemers H, Heyns W, Verhoeven G (2003) Fatty acid synthase drives the synthesis of phospholipids partitioning into detergent-resistant membrane microdomains. Biochem Biophys Res Commun 302: 898–903. Swinnen JV, Brusselmans K, Verhoeven G (2006) Increased lipogenesis in cancer cells: new players, novel targets. Curr Opin Clin Nutr Metab Care 9: 358–365. Szutowicz A, Kwiatkowski J, Angielski S (1979) Lipogenetic and glycolytic enzyme activities in carcinoma and nonmalignant diseases of the human breast. Br J Cancer 39: 681–687. Thupari JN, Pinn ML, Kuhajda FP (2001) Fatty acid synthase inhibition in human breast cancer cells leads to malonyl-CoA-induced inhibition of fatty acid oxidation and cytotoxicity. Biochem Biophys Res Commun 285: 217–223. Turyn J, Schlichtholz B, Dettlaff-Pokora A, Presler M, Goyke E, Matuszewski M, Kmiec Z, Krajka K, Swierczynski J (2003) Increased activity of glycerol 3-phosphate dehydrogenase and other lipogenic enzymes in human bladder cancer. Horm Metab Res 35: 565–569. Van de Sande T, De Schrijver E, Heyns W, Verhoeven G, Swinnen JV (2002) Role of the phosphatidylinositol 30 -kinase/PTEN/Akt kinase pathway in the overexpression of fatty acid synthase in LNCaP prostate cancer cells. Cancer Res 62: 642–646. Van de Sande T, Roskams T, Lerut E, Joniau S, Van Poppel H, Verhoeven G, Swinnen JV (2005) High-level expression of fatty acid synthase in human prostate cancer tissues is linked to activation and nuclear localization of Akt/PKB. J Pathol 206: 214–219. Wang X, Tian W (2001) Green tea epigallocatechin gallate: a natural inhibitor of fatty-acid synthase. Biochem Biophys Res Commun 288: 1200–1206. Wang HQ, Altomare DA, Skele KL, Poulikakos PI, Kuhajda FP, Di Cristofano A, Testa JR (2005) Positive feedback regulation between AKT activation and fatty acid synthase expression in ovarian carcinoma cells. Oncogene 24: 3574–3582. Yahagi N, Shimano H, Hasegawa K, Ohashi K, Matsuzaka T, Najima Y, Sekiya M, Tomita S, Okazaki H, Tamura Y, Iizuka Y, Nagai R, Ishibashi S, Kadowaki T, Makuuchi M, Ohnishi S, Osuga J, Yamada N (2005) Co-ordinate activation of lipogenic enzymes in hepatocellular carcinoma. Eur J Cancer 41: 1316–1322. Yeh CW, Chen WJ, Chiang CT, Lin-Shiau SY, Lin JK (2003) Suppression of fatty acid synthase in MCF-7 breast cancer cells by tea and tea polyphenols: a possible mechanism for their hypolipidemic effects. Pharmacogenomics J 3: 267–276. Yoon S, Lee MY, Park SW, Moon JS, Koh YK, Ahn YH, Park BW, Kim KS (2007) Up-regulation of acetyl-CoA carboxylase {alpha} and fatty acid synthase by human epidermal growth factor receptor 2 at the translational level in breast cancer cells. J Biol Chem 282: 26122–26131. Zhong WB, Wang CY, Chang TC, Lee WS (2003) Lovastatin induces apoptosis of anaplastic thyroid cancer cells via inhibition of protein geranylgeranylation and de novo protein synthesis. Endocrinology 144: 3852–3859. Zhou W, Han W, Landree LE, Thupari JN, Pinn ML, Bililign T, Kim EK, Vadlamudi A, Medghalchi SM, El Meskini R, Ronnett GV, Townsend CA, Kuhajda FP (2007) Fatty acid synthase inhibition activates AMP-activated protein kinase in SKOV3 human ovarian cancer cells. Cancer Res 67: 2964–2971.

Molecular Biology of Novel Targets Identified Through Study of Castration-Recurrent Prostate Cancer Philip A. Watson and Charles L. Sawyers

Abstract Prostate cancer that recurs in men after complete androgen deprivation therapy remains today a lethal disease. Reactivation of androgen receptor (AR) signaling in the setting of castrate levels of androgen is widely accepted to be the dominant factor leading to prostate cancer progression. The mechanisms driving AR activation in clinical castration-recurrent prostate cancer (CRPC) are poorly understood. A number of hypotheses have been put forth to explain castrate-activated AR based on examination of patient material and the use of prostate cancer model systems. AR ligand binding domain mutations that confer gain-of-function properties through expanded use of alternative steroid ligands and overexpression of the non-mutated AR, with or without concurring genomic amplification, are two of the most widely cited hypotheses. Recently, much attention has been paid to the hypothesis that CRPC acquires the capacity to synthesize testosterone directly. Perturbation of AR signaling remains the main therapeutic objective. New drugs currently in clinical trials may offer some improved management of CRPC through antagonism of AR activity or blockage of extratesticular androgen production. Ultimately, the identification of drugs that promote selective AR degradation may have the greatest impact against the continued action of AR in CRPC. Androgen deprivation therapy to reduce androgen receptor (AR) signaling is the mainstay treatment for prostate cancer patients who are not cured by primary local therapy (Hellerstedt and Pienta 2002). Typically, patients receive bilateral orchiectomy or medical castration through the use of luteinizing hormone-releasing hormone (LHRH) agonists to deplete 90% of serum testosterone. Often, androgen deprivation therapy includes the use of an antiandrogen (hydroxyflutamide, bicalutamide) to directly antagonize AR transactivation. Androgen deprivation therapy leads to a temporary remission in approximately 90% of patients with advanced prostate cancer, but uniformly a fatal form of the cancer will remerge in the castrate environment. At this point, median overall survival is only 23–37 months (Hellerstedt and Pienta 2002)

C.L. Sawyers(*) Memorial Sloan-Kettering Cancer Center, New York, NY, USA, E-mail: [email protected]

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4_32, # Springer Science + Business Media, LLC 2009

743

744

P.A. Watson, C.L. Sawyers

and further treatment is only palliative. The molecular events that lead to the progression of prostate cancer from a state of androgen stimulation to that of castration recurrence are poorly understood. However, it is clear that the AR plays a central role in the biology of castration-recurrent prostate cancer (CRPC). This chapter highlights the key lines of evidence derived from both clinical and preclinical observations that point to the continued activity of AR in CRPC and discuss the recent use of novel compounds designed to interfere with AR function.

1 AR Mutations and Prostate Cancer An association between prostate cancer and AR mutations was made first in the human prostate cancer cell line, LNCaP (Horoszewicz et al. 1980, 1983), derived from a metastatic lymph node. In this cell line, Veldscholte et al. (1990) identified a missense mutation in the ligand-binding domain of AR (T877A). The T877A mutation expands the range of ligands that can bind and activate AR-mediated transcription to include glucocorticoids (Chang et al. 2001), dehydroepiandrosterone (Tan et al. 1997), progestins, estrogens, and anti-androgens, such as hydroxyflutamide and nilutamide (Veldscholte et al. 1990, 1992). In contrast, the antiandrogen bicalutamide does not activate the mutant AR found in LNCaP and remains a growth antagonist (Veldscholte et al. 1992). AR mutations in the ligandbinding domain that confer increased affinity for noncanonical ligands are a relatively common occurrence in human prostate cancer cell lines. The CWR22 xenograft (Pretlow et al. 1993; Wainstein et al. 1994) harbors the H874Y mutation (Tan et al. 1997), while the MDA Pca 2a and 2b (Navone et al. 1997) cell lines contain both T877A and L701H mutations (Zhao et al. 1999). As with LNCaP, bicalutamide remains an AR antagonist against both CWR22 (Sirotnak et al. 2002) and MDA Pca lines (Krishnan et al. 2002; Hara et al. 2003), which indicates that these alternative AR ligand-binding domain mutations do not alter the properties of bicalutamide. Numerous reports have evaluated the frequency of AR mutations in clinical samples of prostate cancer (Taplin et al. 1995, 1999; Hyytinen et al. 2002; Wallen et al. 1999; Culig et al. 1993; Gaddipati et al. 1994; Ruizeveld de Winter et al. 1994; Elo et al. 1995; Evans et al. 1996; Tilley et al. 1996; Marcelli et al. 2000; Thompson et al. 2003) and an online database collects the various mutations that have been found in prostate cancer and other human diseases (The Androgen Receptor Gene Mutations Database World Wide Web Server, http://androgendb.mcgill.ca). In general, AR mutations have been associated with CRPC, while untreated androgen-stimulated primary cancers have negligible mutation frequencies (Evans et al. 1996; Marcelli et al. 2000; Newmark et al. 1992). However, two studies were exceptions; Tilley et al. (1996) and Thompson et al. (2003) reported mutations in 44% and 29% of primary cancers, respectively. The reported frequency of mutations has varied widely from close to 0% (Wallen et al. 1999) to upwards of 50% (Taplin et al. 1995). Most mutations have been identified in the ligand-binding

Study of Castration-Recurrent Prostate Cancer

745

domain, attributable in part to a greater degree of scrutiny for this region of the receptor. A prospective study of 48 metastatic CRPC samples obtained from bone marrow biopsies found AR mutations in just 10% of the patients (Taplin et al. 2003). The frequency of AR mutations is likely influenced by several parameters, which include patient selection, treatment history, and stage of disease. The T877A mutation originally identified in LNCaP has also been observed in subsets of metastatic prostate cancers (Gaddipati et al. 1994). Based upon the observed in vitro properties of the T877A and other similar AR mutants, it has been speculated that ligand-binding domain mutations convert AR into a promiscuous steroid receptor and this is one hypothesis for the emergence of CRPC. However, in the complex in vivo environment of the prostate cancer patient, it is difficult to know to what extent individual steroids or combination of steroids and/ or antiandrogens transactivate the mutated AR. One preclinical in vivo study found that neither estradiol and progesterone nor flutamide stimulated CWR22 xenograft tumor formation in castrated mice or induced PSA production (Shao et al. 2003). The expanded steroid specificity of the AR with ligand-binding domain mutations ascertained using in vitro data could be an oversimplification, but more in vivo studies using additional xenografts and other mutations will be required to reach such a conclusion. The molecular evolution of any given prostate cancer may be influenced by the particular therapy or combination of therapies given to the patient. The T877A mutation was found in 5 of 16 patients who had been treated with combined androgen blockade using flutamide. In contrast, none of 17 patients who received androgen deprivation monotherapy had T877A (Taplin et al. 1999). All of the patients with T877A mutations subsequently responded well to bicalutamide therapy, which remains antagonistic to the mutated AR (Taplin et al. 1999). More recently, AR ligand-binding mutations have been associated with bicalutamide therapy as well. Yoshida et al. (2005) obtained the KUCaP serially transplantable xenograft from a prostate cancer liver metastasis in a patient treated with bicalutamide. KUCaP contains a W741C mutation in the AR ligand-binding domain, which was confirmed present in the original patient material. Additionally, bicalutamide stimulated both PSA production and growth of the xenograft tumors. In another clinical cohort, the W741C mutation was found in patients treated with combined androgen blockade, which used bicalutamide (Haapala et al. 2001). The W741C mutation was identified in LNCaP cells selected in vitro for 12 weeks in the presence of bicalutamide (Hara et al. 2003). Further characterization of the resulting cells (LNCaP-cxD2) demonstrated that bicalutamide acted as both a growth and transcriptional agonist. Furthermore, in reporter transactivation assays in COS-7 cells, the W741C AR remained antagonized by hydroxyflutamide (Hara et al. 2003). Therefore, in these particular examples, the mode of therapy (flutamide or bicalutamide) may have selected for cancers with specific ligand-binding domain mutations that could accommodate the antiandrogen as an agonist. Could it be that rare cells with these AR mutations preexist in untreated cancers before the delivery of complete androgen blockade therapy? If so, such cells would be expected to gain a growth advantage after exposure to an antiandrogen. Such a scenario has a

746

P.A. Watson, C.L. Sawyers

precedent in some chronic myelogenous leukemia patients. In some cases, a population of cells with BCR-ABL tyrosine kinase domain mutations preexisted in the cancers that rendered them refractory to subsequent treatment with the BCRABL small molecule inhibitor imatinib (Shah et al. 2002). The association of specific AR mutations with particular therapies suggests that the frequencies and spectrum of mutations in CRPC should be revaluated with modern sensitive sequencing methodologies.

2 Amplification and Overexpression of AR in Prostate Cancer Visakorpi et al. (1995) were the first to identify genomic amplification of the AR gene in CRPC. Using a combination of comparative genomic hybridization and fluorescence in situ hybridization, the AR locus was found amplified in 30% (7/23) of CRPC, but not in any of the androgen-stimulated primary cancers matched to the same patients. The extent of amplification was heterogeneous within a given sample; some individual tumor cells displayed up to 40 copies of AR. Analysis of an additional clinical data set of matched primary androgen-stimulated and CRPC prostate cancers from the same research group revealed an identical stage-specific pattern and frequency of AR amplification (Koivisto et al. 1997). The presence of amplification correlated with increased AR mRNA expression. Patients who ultimately developed AR amplification had responded more favorably to androgen deprivation therapy and had twice longer postrecurrence survival compared to cases without amplification (Koivisto et al. 1997). Subsequently, other investigators have observed that AR amplification is a moderately frequent alteration in CRPC (KaltzWittmer et al. 2000; Miyoshi et al. 2000; Brown 2002 et al.; Edwards 2003 et al.). The acquisition of AR amplification has been proposed as one possible mechanism leading to prostate cancer progression. AR gene copy increases presumably lead to overexpression of AR mRNA and protein, which theoretically would sensitize the cancer cell to castrate levels of androgens. The cell with AR amplification should have a selective growth advantage. However, study of the molecular consequences of AR amplification has been difficult since no human prostate cancer cell lines exhibit natural increases in the native gene. Two human prostate cancer xenografts (LuCaP 35 and 69) derived from CRPC were reported to have AR amplification (Linja et al. 2001). Unfortunately, these tumors have not been reported to propagate in culture, which makes it challenging to further study their biology. Our laboratory isolated a mouse cell line (Myc-CaP) derived from an androgen-stimulated primary prostate cancer that was found to contain spontaneous high-level AR genomic amplification and a corresponding large degree of AR protein overexpression (Watson et al. 2005). However, in this model, AR amplification does not in and of itself cause androgen independence. Myc-CaP cells grafted into castrated mice demonstrated delayed tumor formation compared to testes-intact mice and tumors established in intact mice regressed after castration (Watson et al. 2005). Nevertheless, Myc-CaP cells retain the potential to evolve to

Study of Castration-Recurrent Prostate Cancer

747

CRPC, which suggests the need for additional, and as yet unknown, molecular alterations necessary for disease progression after castration. In support of this hypothesis, an evaluation of prostate cancer cases with AR amplification demonstrated no negative effect on survival (Ford et al. 2003). So, it is not possible yet to assign a strict cause and effect relationship to AR copy number changes and onset of CRPC. At a minimum, human prostate cancer cell lines isolated from tumors with AR amplification are needed to investigate AR-mediated signaling as a function of gene copy number. Even though the majority of CRPC cases do not contain AR gene amplification, AR gene overexpression without alteration in copy number is nevertheless a frequently observed event. Microarray analysis demonstrated that AR mRNA was upregulated 5.84-fold in prostate cancer bone metastases compared to primary cancers (Stanbrough et al. 2006). Another microarray study revealed an increase in AR mRNA by 9–11-fold in CRPC relative to primary prostate cancers with a corresponding overexpression of AR protein (Holzbeierlein et al. 2004). Microarray profiling of matched pairs of androgen-dependent and CRPC xenografts revealed that AR was the only gene upregulated uniformly across all CRPC lines (Chen et al. 2004). Is AR overexpression a driving force for progression to castration recurrence? Our laboratory addressed this question by increasing the expression of AR in AR-positive, androgen-sensitive LAPC4 and LNCaP prostate cancer cell lines using retroviral or lentiviral infection (Chen et al. 2004). A modest degree of AR overexpression was sufficient to enable tumor formation in castrated mice. The additional AR was necessary for this gain of function, as an AR shRNA prevented the growth of LNCaP/AR and LAPC4/AR cells in castrated mice. Overexpression of AR in LNCaP and LAPC4 also allowed antiandrogens to act as weak transcriptional agonists. Bicalutamide treatment of LNCaP and LAPC4 cells overexpressing AR (but not their parental lines) stimulated the production of PSA and other known androgen-regulated genes (Chen et al. 2004). The induced gene expression correlated with a redistribution of coactivators and corepressors recruited to the promoter regions of AR target genes (Chen et al. 2004). These findings suggest that one possible consequence of AR overexpression in CRPC is the inappropriate activation of AR signaling resulting from treatment with bicalutamide and other antiandrogens. This hypothesis is consistent with the wellrecognized antiandrogen withdrawal syndrome, where a subset of patients progressing during antiandrogen therapy experience a reduction in PSA coupled with clinical improvement when treatment is discontinued (Scher and Kelly 1993; Schellhammer et al. 1997). The discovery that prostate cancer cells with overexpression of AR can utilize bicalutamide as an agonist prompted an effort to identify next-generation antiandrogens that would be effective against prostate cancers with increased AR, while at the same time would not aberrantly activate AR signaling. In collaboration with Michael Jung at UCLA, our laboratory screened a series of chemical compounds designed to fit into the ligand-binding pocket of AR for in vitro antagonistic activity against LNCaP/AR cells. A lead compound identified from the screen, RD162, binds AR with tenfold greater affinity than bicalutamide and acts as a pure transcriptional

748

P.A. Watson, C.L. Sawyers

antagonist against androgen-regulated gene expression in LNCaP/AR (Clegg et al. 2008). This result can be explained in part by chromatin immunoprecipitation assays that demonstrated that RD162 fails to recruit AR or Pol II to the promoter/ enhancers of PSA and Klk2 in LNCaP/AR cells, in clear contrast to bicalutamide. Most importantly, RD162 displays greater therapeutic efficacy than bicalutamide against LNCaP/AR tumors (Clegg et al. 2008). Under the name MDV3100, RD162 recently entered phase I/II trials for patients with progressive CRPC. To date, MDV3100 has been well tolerated and has demonstrated sustained declines of serum PSA levels by greater than 50% in a high proportion of patients (Scher et al. 2008).

3 Extratesticular Sources of Androgen Production in CRPC Medical or surgical castration does not entirely deplete the body of androgens, as the adrenal gland remains a production source for the weak androgens, dehydroepiandrosterone (DHEA) and androstenedione, both of which are precursors for the biosynthesis of testosterone. DHEA and androstenedione have been shown to stimulate prostate growth in the castrated rat at plasma concentrations normally found in adult men (Labrie et al. et al. 1988). Therefore, surgical removal of the adrenal glands or chemical inhibition of adrenal androgen synthesis has long been used for the treatment of CRPC. The gene, CYP17, encodes a critical enzyme in the biosynthesis of steroids, including androgens, and has both 17a-hydroxylase and C17,20-lyase activity (Miller et al. 1997). Ketoconazole, originally identified as an antifungal agent, also nonselectively suppresses steroidogenesis through inhibition of CYP17 (Santen et al. 1983). In CRPC, high-dose treatment with ketoconazole achieved tumor size reduction and normal bone scans in 14% of patients (Trump et al. 1989). Ketoconazole retained the ability to reduce PSA production even in patients after antiandrogen withdrawal (Small et al. 1997). A novel steroidal inhibitor of CYP17 (Potter et al. 1995; Rowlands et al. 1995), abiraterone acetate (CB7630), and ketoconazole were tested in mice for physiological indication of androgen synthesis inhibition. Abiraterone acetate decreased plasma testosterone levels to 0.1 nM and reduced weights of androgen-dependent tissues, which included the prostate, seminal vesicles, and testes (Barrie et al. 1994). In contrast, ketoconazole did not affect these tissues (Barrie et al. 1994). Hence, abiraterone acetate appears to have greater potency as a CYP17 inhibitor than ketoconazole. Based on the in vitro and preclinical data, abiraterone acetate has been developed for clinical trials to test its efficacy against CRPC. An initial phase I study achieved suppression of serum testosterone in both castrate and intact prostate cancer patients (O’Donnell et al. 2004). Early phase II trial results from both the United States and Great Britain indicate that 2–3 months of CB7630 treatment lowered serum PSA values 50% in 40–61% of patients with progressive CRPC (Reid et al. 2007; Attard et al. 2007; Danila et al. 2008). Furthermore, the burden of circulating tumor cells was reduced in a subset of these patients (Reid et al. 2007; Attard et al.

Study of Castration-Recurrent Prostate Cancer

749

2007; Danila et al. 2008). Based on these initial results, phase III trials are planned for patients with CRPC who have progressed after docetaxel chemotherapy. In patients with CRPC, the tumor itself may serve as a reservoir of androgen despite castrate serum levels of testosterone. Prostate tumor samples were obtained either from men with CRPC or from untreated benign prostate controls. The intraprostatic level of testosterone was similar between CRPC and benign tissue (Mohler et al. 2004). Although dihydrotestosterone was significantly decreased in CRPC, it remained at a level (1.45 nM) that positively correlated with high expression of PSA (Mohler et al. 2004). In another study, prostate biopsies were collected from normal healthy volunteers after 1 month of medical castration or cancer patients after a regimen of neoadjuvant androgen deprivation therapy for up to 9 months. Despite a 75% drop in the intraprostatic androgen concentration, androgen-regulated gene expression was partially maintained (Page et al. 2006; Mostaghel et al. 2007). CRPC may actually acquire the capacity to synthesize its own endogenous testosterone (Stanbrough et al. 2006; Holzbeierlein et al. 2004). Coordinated upregulation of mRNA for multiple genes involved in testosterone synthesis and catabolism was observed in a microarray study profiling CRPC bone metastases and androgen-stimulated primary prostate cancers (Stanbrough et al. 2006). Namely, HSD3B2, which converts DHEA to androstenedione, and AKR1C3, which in turn reduces androstenedione to testosterone, were increased by 1.8- and 5.3-fold, respectively, in CRPC. Increased AKRIC3 protein in CRPC was confirmed using immunohistochemistry (Stanbrough et al. 2006). Another group recently reported similar results using qRT-PCR to interrogate the expression of enzymes involved in steroid production in CRPC and primary prostate cancers (Montgomery et al. 2008). In this case, numerous genes were upregulated in CRPC, which included CYP17. Increased expression of steroidogenic enzymes positively correlated with higher levels of testosterone within CRPC tissue compared to androgen-stimulated primary prostate cancer (Montgomery et al. 2008). An earlier finding identified several enzymes upregulated at the mRNA level for sterol synthesis in CRPC (Holzbeierlein et al. 2004), which could lead ultimately to the increased production of steroid hormones. Although steroidogenic enzymatic activity has not been confirmed in these various reports, they nonetheless establish that CRPC expresses many of the genes required for intracrine production of testosterone. These data also suggest that prostate cancers remain exposed to a level of androgens sufficient to maintain some degree of AR signaling, which could contribute to castration recurrence.

4 Downregulation of AR Several groups have addressed whether AR is required for the growth of human prostate cancer cell lines. The earliest efforts to downregulate AR used antisense oligonucleotides and achieved nearly complete loss of AR protein in LNCaP, which resulted in severe growth inhibition and downregulation of PSA production (Eder

750

P.A. Watson, C.L. Sawyers

et al. 2000). This observation was also made in LNCaP-abl cells, a subline adapted to grow long term in charcoal stripped serum (Culig et al. 1999). After the adoption of RNAi technology to mammalian cells, the AR dependency of prostate cancer cells has been reassessed. Effects of AR knockdown have been determined in androgen-sensitive LNCaP and LAPC4 (Klein et al. 1997) and androgen-independent variants of LNCaP and CWR22 (Cheng et al. 2006; Haag et al. 2005; Liao et al. 2005; Wright et al. 2003; Yuan et al. 2006). A common theme from these studies is growth inhibition of prostate cancer cell lines after loss of AR irrespective of their requirement for normal levels of androgen. In one report, cell cycle analysis showed an accumulation of androgen-independent CWR22R3 cells in G0/G1 after AR knockdown (Yuan et al. 2006). Liao et al. (2005) found that siRNAs targeted against AR in LNCaP triggered apoptosis as judged by multiple parameters, which included increased annexin V binding, activation of caspases, and cleavage of PARP. However, other groups have not reported a substantial induction of apoptosis in prostate cancer cell lines after AR knockdown. In any event, it seems that potentially all prostate cancer cells that naturally express AR remain dependent on AR for cell growth. Since androgen-independent prostate cancer cell lines remain dependent on AR protein for growth, might AR degradation be an effective therapy where ligand withdrawal and antagonism fail? And to what degree must degradation occur in order to achieve clinical efficacy? Preliminary studies in the Myc-CaP prostate cancer cell with AR amplification revealed that despite an extreme abundance of AR protein, growth and androgen regulated gene expression was inhibited by relatively modest AR shRNA knockdown (Watson et al. 2008). These findings point to the possibility that drugs that induce modest AR degradation could nevertheless achieve clinical efficacy in CRPC with elevated AR.

5 Conclusions Ever since the seminal observation by Huggins and Hodges et al. (1941) that castration was beneficial in advanced prostate cancer, abrogation of AR action has remained the therapeutic objective in the clinical management of this disease. It is apparent that the onset of CRPC coincides with renewed AR signaling. Restoration of AR-mediated gene expression is likely to occur as a result of several potential mechanisms, which include amplification or overexpression of AR, AR mutagenesis that results in a gain-of-function, and enhanced sensitivity of the cancer to residual amounts of androgen or intratumoral synthesis of testosterone. Given that metastatic prostate cancer is a molecularly heterogeneous disease even within a single patient (Shah et al. 2004), multiple mechanisms may simultaneously and ultimately give rise to a molecularly diverse group of CRPC. However, irrespective of how CRPC arises, squelching of AR signaling will remain essential. Further research is needed to better understand the complexity of AR-driven

Study of Castration-Recurrent Prostate Cancer

751

biology, especially in the context of the in vivo environment, to improve the design of drugs targeting AR.

References Attard, G., et al. Abiraterone, an oral, irreversible, Cyp450C17 enzyme inhibitor appears to have activity in post-docetaxel castration refractory prostate cancer (CRPC) patients (pts). In European Society for Medical Oncology 2007 Annual Meeting. 2007. Lugano, Switzerland. Barrie, S.E., et al. Pharmacology of novel steroidal inhibitors of cytochrome P450(17) alpha (17 alpha-hydroxylase/C17–20 lyase). J Steroid Biochem Mol Biol, 1994. 50(5–6): 267–73. Brown, R.S., et al. Amplification of the androgen receptor gene in bone metastases from hormonerefractory prostate cancer. J Pathol, 2002. 198(2): 237–44. Chang, C.Y., P.J. Walther, and D.P. McDonnell, Glucocorticoids manifest androgenic activity in a cell line derived from a metastatic prostate cancer. Cancer Res, 2001. 61(24): 8712–7. Chen, C.D., et al. Molecular determinants of resistance to antiandrogen therapy. Nat Med, 2004. 10(1): 33–9. Cheng, H., et al. Short hairpin RNA knockdown of the androgen receptor attenuates ligandindependent activation and delays tumor progression. Cancer Res, 2006. 66(21): 10613–20. Clegg, N.J., et al. Development of androgen receptor antagonists with a novel mechanism of action. In Nuclear Receptors: Steroid Sisters. 2008. Whistler, British Columbia: Keystone Symposia. Culig, Z., et al. Mutant androgen receptor detected in an advanced-stage prostatic carcinoma is activated by adrenal androgens and progesterone. Mol Endocrinol, 1993. 7(12): 1541–50. Culig, Z., et al. Switch from antagonist to agonist of the androgen receptor bicalutamide is associated with prostate tumour progression in a new model system. Br J Cancer, 1999. 81 (2): 242–51. Danila, D.C., et al. Abiraterone acetate and prednisone in patients (Pts) with progressive metastatic castration resistant prostate cancer (CRPC) after failure of docetaxel-based chemotherapy. In 2008 ASCO Annual Meeting. 2008. Chicago, IL. Eder, I.E., et al. Inhibition of LncaP prostate cancer cells by means of androgen receptor antisense oligonucleotides. Cancer Gene Ther, 2000. 7(7): 997–1007. Edwards, J., et al. Androgen receptor gene amplification and protein expression in hormone refractory prostate cancer. Br J Cancer, 2003. 89(3): 552–6. Elo, J.P., et al. Mutated human androgen receptor gene detected in a prostatic cancer patient is also activated by estradiol. J Clin Endocrinol Metab, 1995. 80(12): 3494–500. Evans, B.A., et al. Low incidence of androgen receptor gene mutations in human prostatic tumors using single strand conformation polymorphism analysis. Prostate, 1996. 28(3): 162–71. Ford, O.H. III, et al. Androgen receptor gene amplification and protein expression in recurrent prostate cancer. J Urol, 2003. 170(5): 1817–21. Gaddipati, J.P., et al. Frequent detection of codon 877 mutation in the androgen receptor gene in advanced prostate cancers. Cancer Res, 1994. 54(11): 2861–4. Haag, P., et al. Androgen receptor down regulation by small interference RNA induces cell growth inhibition in androgen sensitive as well as in androgen independent prostate cancer cells. J Steroid Biochem Mol Biol, 2005. 96(3–4): 251–8. Haapala, K., et al. Androgen receptor alterations in prostate cancer relapsed during a combined androgen blockade by orchiectomy and bicalutamide. Lab Invest, 2001. 81(12): 1647–51. Hara, T., et al. Enhanced androgen receptor signaling correlates with the androgen-refractory growth in a newly established MDA PCa 2b-hr human prostate cancer cell subline. Cancer Res, 2003. 63(17): 5622–8.

752

P.A. Watson, C.L. Sawyers

Hara, T., et al. Novel mutations of androgen receptor: a possible mechanism of bicalutamide withdrawal syndrome. Cancer Res, 2003. 63(1): 149–53. Hellerstedt, B.A. and K.J. Pienta. The current state of hormonal therapy for prostate cancer. CA Cancer J Clin, 2002. 52(3): 154-79. Horoszewicz, J.S., et al. The LNCaP cell line–a new model for studies on human prostatic carcinoma. Prog Clin Biol Res, 1980. 37: 115–32. Holzbeierlein, J., et al. Gene expression analysis of human prostate carcinoma during hormonal therapy identifies androgen-responsive genes and mechanisms of therapy resistance. Am J Pathol, 2004. 164(1): 217–27. Horoszewicz, J.S., et al. LNCaP model of human prostatic carcinoma. Cancer Res, 1983. 43(4): 1809–18. Huggins, C. and C.V. Hodges. The effect of castration, estogen and androgen injection on serum phosphatases in metastatic carcinoma of the prostate. Cancer Res, 1941. 1: 293–297. Hyytinen, E.R., et al. Pattern of somatic androgen receptor gene mutations in patients with hormone-refractory prostate cancer. Lab Invest, 2002. 82(11): 1591–8. Kaltz-Wittmer, C., et al. FISH analysis of gene aberrations (MYC, CCND1, ERBB2, RB, and AR) in advanced prostatic carcinomas before and after androgen deprivation therapy. Lab Invest, 2000. 80(9): 1455–64. Klein, K.A., et al. Progression of metastatic human prostate cancer to androgen independence in immunodeficient SCID mice. Nat Med, 1997. 3(4): 402–8. Koivisto, P., et al. Androgen receptor gene amplification: a possible molecular mechanism for androgen deprivation therapy failure in prostate cancer. Cancer Res, 1997. 57(2): 314–9. Krishnan, A.V., et al. A glucocorticoid-responsive mutant androgen receptor exhibits unique ligand specificity: therapeutic implications for androgen-independent prostate cancer. Endocrinology, 2002. 143(5): 1889–900. Labrie, C., A. Belanger, and F. Labrie. Androgenic activity of dehydroepiandrosterone and androstenedione in the rat ventral prostate. Endocrinology, 1988. 123(3): 1412–7. Liao, X., et al. Small-interfering RNA-induced androgen receptor silencing leads to apoptotic cell death in prostate cancer. Mol Cancer Ther, 2005. 4(4): 505–15. Linja, M.J., et al. Amplification and overexpression of androgen receptor gene in hormonerefractory prostate cancer. Cancer Res, 2001. 61(9): 3550–5. Marcelli, M., et al. Androgen receptor mutations in prostate cancer. Cancer Res, 2000. 60(4): 944–9. Miller, W.L., R.J. Auchus, and D.H. Geller. The regulation of 17,20 lyase activity. Steroids, 1997. 62(1): 133–42. Miyoshi, Y., et al. Fluorescence in situ hybridization evaluation of c-myc and androgen receptor gene amplification and chromosomal anomalies in prostate cancer in Japanese patients. Prostate, 2000. 43(3): 225–32. Mohler, J.L., et al. The androgen axis in recurrent prostate cancer. Clin Cancer Res, 2004. 10(2): 440–8. Montgomery, R.B., et al. Maintenance of intratumoral androgens in metastatic prostate cancer: a mechanism for castration-resistant tumor growth. Cancer Res, 2008. 68(11): 4447–54. Mostaghel, E.A., et al. Intraprostatic androgens and androgen-regulated gene expression persist after testosterone suppression: therapeutic implications for castration-resistant prostate cancer. Cancer Res, 2007. 67(10): 5033–41. Navone, N.M., et al. Establishment of two human prostate cancer cell lines derived from a single bone metastasis. Clin Cancer Res, 1997. 3(12 Pt 1): 2493–500. Newmark, J.R., et al. Androgen receptor gene mutations in human prostate cancer. Proc Natl Acad Sci U S A, 1992. 89(14): 6319–23. O’Donnell, A., et al. Hormonal impact of the 17alpha-hydroxylase/C(17,20)-lyase inhibitor abiraterone acetate (CB7630) in patients with prostate cancer. Br J Cancer, 2004. 90(12): 2317–25.

Study of Castration-Recurrent Prostate Cancer

753

Page, S.T., et al. Persistent intraprostatic androgen concentrations after medical castration in healthy men. J Clin Endocrinol Metab, 2006. 91: 3850–6 Potter, G.A., et al. Novel steroidal inhibitors of human cytochrome P45017 alpha (17 alphahydroxylase-C17,20-lyase): potential agents for the treatment of prostatic cancer. J Med Chem, 1995. 38(13): 2463–71. Pretlow, T.G., et al. Xenografts of primary human prostatic carcinoma. J Natl Cancer Inst, 1993. 85(5): 394–8. Reid, A., et al. Inhibition of androgen synthesis results in a high response rate in castration refractory prostate cancer (CRPC). In European Society for Medical Oncology 2007 Annual Meeting. 2007. Lugano, Switzerland. Rowlands, M.G., et al. Esters of 3-pyridylacetic acid that combine potent inhibition of 17 alphahydroxylase/C17,20-lyase (cytochrome P45017 alpha) with resistance to esterase hydrolysis. J Med Chem, 1995. 38(21): 4191–7. Ruizeveld de Winter, J.A., et al. Androgen receptor status in localized and locally progressive hormone refractory human prostate cancer. Am J Pathol, 1994. 144(4): 735–46. Santen, R.J., et al. Site of action of low dose ketoconazole on androgen biosynthesis in men. J Clin Endocrinol Metab, 1983. 57(4): 732–6. Schellhammer, P.F., et al. Prostate specific antigen decreases after withdrawal of antiandrogen therapy with bicalutamide or flutamide in patients receiving combined androgen blockade. J Urol, 1997. 157(5): 1731–5. Scher, H.I. and W.K. Kelly. Flutamide withdrawal syndrome: its impact on clinical trials in hormone-refractory prostate cancer. J Clin Oncol, 1993. 11(8): 1566–72. Scher, H.I., et al. Phase I/II study of MDV3100 in patients (pts) with progressive castrationresistant prostate cancer (CRPC). In 2008 ASCO Annual Meeting. 2008. Chicago, IL: American Society of Clinical Oncology. Shah, N.P., et al. Multiple BCR-ABL kinase domain mutations confer polyclonal resistance to the tyrosine kinase inhibitor imatinib (STI571) in chronic phase and blast crisis chronic myeloid leukemia. Cancer Cell, 2002. 2(2): 117–25. Shah, R.B., et al. Androgen-independent prostate cancer is a heterogeneous group of diseases: lessons from a rapid autopsy program. Cancer Res, 2004. 64(24): 9209–16. Shao, T.C., et al. In vivo preservation of steroid specificity in CWR22 xenografts having a mutated androgen receptor. Prostate, 2003. 57(1): 1–7. Sirotnak, F.M., et al. Studies with CWR22 xenografts in nude mice suggest that ZD1839 may have a role in the treatment of both androgen-dependent and androgen-independent human prostate cancer. Clin Cancer Res, 2002. 8(12): 3870–6. Small, E.J., et al. Ketoconazole retains activity in advanced prostate cancer patients with progression despite flutamide withdrawal. J Urol, 1997. 157(4): 1204–7. Stanbrough, M., et al. Increased expression of genes converting adrenal androgens to testosterone in androgen-independent prostate cancer. Cancer Res, 2006. 66(5): 2815–25. Tan, J., et al. Dehydroepiandrosterone activates mutant androgen receptors expressed in the androgen-dependent human prostate cancer xenograft CWR22 and LNCaP cells. Mol Endocrinol, 1997. 11(4): 450–9. Taplin, M.E., et al. Androgen receptor mutations in androgen-independent prostate cancer: Cancer and Leukemia Group B Study 9663. J Clin Oncol, 2003. 21(14): 2673–8. Taplin, M.E., et al. Mutation of the androgen-receptor gene in metastatic androgen-independent prostate cancer. N Engl J Med, 1995. 332(21): 1393–8. Taplin, M.E., et al. Selection for androgen receptor mutations in prostate cancers treated with androgen antagonist. Cancer Res, 1999. 59(11): 2511–5. Thompson, J., et al. Androgen receptor mutations in high-grade prostate cancer before hormonal therapy. Lab Invest, 2003. 83(12): 1709–13. Tilley, W.D., et al. Mutations in the androgen receptor gene are associated with progression of human prostate cancer to androgen independence. Clin Cancer Res, 1996. 2(2): 277–85.

754

P.A. Watson, C.L. Sawyers

Trump, D.L., et al. High-dose ketoconazole in advanced hormone-refractory prostate cancer: endocrinologic and clinical effects. J Clin Oncol, 1989. 7(8): 1093–8. Veldscholte, J., et al. A mutation in the ligand binding domain of the androgen receptor of human LNCaP cells affects steroid binding characteristics and response to anti-androgens. Biochem Biophys Res Commun, 1990. 173(2): 534–40. Veldscholte, J., et al. Anti-androgens and the mutated androgen receptor of LNCaP cells: differential effects on binding affinity, heat-shock protein interaction, and transcription activation. Biochemistry, 1992. 31(8): 2393–9. Visakorpi, T., et al. In vivo amplification of the androgen receptor gene and progression of human prostate cancer. Nat Genet, 1995. 9(4): 401–6. Wainstein, M.A., et al. CWR22: androgen-dependent xenograft model derived from a primary human prostatic carcinoma. Cancer Res, 1994. 54(23): 6049–52. Wallen, M.J., et al. Androgen receptor gene mutations in hormone-refractory prostate cancer. J Pathol, 1999. 189(4): 559–63. Watson, P.A., et al. Context-dependent hormone-refractory progression revealed through characterization of a novel murine prostate cancer cell line. Cancer Res, 2005. 65(24): 11565–71. Watson, P.A., et al. Evidence for an androgen receptor threshold dependency in prostate cancer cells with natural androgen receptor overexpression. In 99th Annual Meeting of the American Association for Cancer Research. 2008. San Diego, CA: AACR. Wright, M.E., M.J. Tsai, and R. Aebersold. Androgen receptor represses the neuroendocrine transdifferentiation process in prostate cancer cells. Mol Endocrinol, 2003. 17(9): 1726–37. Yoshida, T., et al. Antiandrogen bicalutamide promotes tumor growth in a novel androgendependent prostate cancer xenograft model derived from a bicalutamide-treated patient. Cancer Res, 2005. 65(21): 9611–6. Yuan, X., et al. Androgen receptor remains critical for cell-cycle progression in androgenindependent CWR22 prostate cancer cells. Am J Pathol, 2006. 169(2): 682–96. Zhao, X.Y., et al. Two mutations identified in the androgen receptor of the new human prostate cancer cell line MDA PCa 2a. J Urol, 1999. 162(6): 2192–9.

Selenium and Androgen Receptor in Prostate Cancer Nagalakshmi Nadiminty and Allen C. Gao

Abstract Prostate cancer chemoprevention has generated considerable interest in the last decade and selenium and combinations of selenium have been recognized as one of the most efficacious chemopreventive agents against prostate cancer. This chapter focuses on a discussion of the findings regarding the mechanisms of action of various selenium compounds and their effects on cellular processes and signaling pathways, particularly on androgen signaling. We also describe the clinical and preclinical studies that have contributed enormously to the knowledge about dose, duration of exposure, and the chemical form of selenium effective in different scenarios. The initial analysis of the SELECT study found that 200 mg of selenium and 400 IUs daily of vitamin E do not prevent prostate cancer even though a 2002 follow-up report showed that men who took selenium for more than 7½ years had about 52 percent fewer new cases of prostate cancer than men who took placebo. This raises concerns about the efficacy of the use of selenomethionine in the trial as the source of selenium and underscores the importance of choosing the chemical form and dosage of selenium supplementation with care. Even though popular opinion is still undecided about whether selenium can be used as a chemopreventive agent in the clinic and whether studies with cell lines and populations at low, medium, or high risk can adequately represent the physiological behavior of this micronutrient, it can safely be said to offer the most diverse spectrum of protective effects against prostate cancer which warrants for its future as a chemopreventive agent. A more diverse analysis of the effects of selenium on androgen receptor signaling and related pathways would shed more light on its role in prevention of prostate cancer progression rather than initiation.

A.C. Gao(*) Department of Urology and Cancer Center, Research III Bldg, Suite 1300, 4645, 2nd Avenue, Sacramento, CA, 95817, USA, E-mail: [email protected]

D. Tindall and J. Mohler (eds.), Androgen Action in Prostate Cancer, DOI: 10.1007/978-0-387-69179-4_33, # Springer Science + Business Media, LLC 2009

755

756

N. Nadiminty, A.C. Gao

1 Introduction The human prostate is uniquely dependent on the male hormones, androgens, in that androgen-deprivation therapy results in shrinking of the prostate and restoration of androgen supply results in prostate regrowth. Thus, androgens are an important survival factor for the prostate, and androgen signaling through the androgen receptor (AR) forms the mainstay of growth and survival signaling in the normal adult prostate (Grossmann et al. 2001). The AR belongs to the family of steroid hormone receptor transcription factors and is localized to secretory epithelial cells and stromal cells of the prostate (Chatterjee 2003). The AR without its ligand is localized in the cytoplasm and upon ligand binding translocates to the nucleus and transactivates target genes involved in diverse physiological processes like proliferation, differentiation, apoptosis, and secretion. Thus androgens are responsible for maintaining prostate homeostasis by balancing proliferation and death among prostatic cells via the AR. Prostate cancer like many other cancers arises from cells that accumulate a wide spectrum of genetic changes or mutations conferring survival advantage. Almost all prostate cancers are androgen dependent, and standard therapies for advanced prostate cancer always start with androgen-deprivation therapy that leads to tumor regression. But invariably almost all patients progress to an ‘‘androgen-independent’’ or castrationresistant stage, which may be the result of additional genetic changes, mutations in the AR gene, activation of alternative signaling pathways, and/or activation of oncogenes that render prostate cancer fatal. Castration resistance is the clinical stage in which prostate cancer cells are able to survive and proliferate without the required signals from circulating testicular androgens. Several strategies have been developed and are currently in clinical trials for the treatment of castration-resistant prostate cancer, like agents targeting the epidermal growth factor receptor pathway, Akt/PI3K pathway, platelet-derived growth factor receptor pathway, etc. But targeting the AR pathway is more likely to provide clinical benefit (Nieto et al. 2007), though the current targeting strategies for AR fail to block the signaling pathway fully. Newer strategies that block AR signaling are thus likely to be of additional clinical benefit while maintaining unique specificity for prostate cancer. Accomplishing this goal will require better understanding of the signaling pathways branching out from the AR and how those pathways may become activated during the androgen-insensitive state. Targeting such effector pathways/proteins alone or in combination with androgen-deprivation therapy may improve the ability to treat advanced prostate cancer.

2 Background Selenium (Se) is an essential micronutrient trace element and the concentration of Se in diet depends on the soil in the region, the types of food consumed, and other factors that facilitate or inhibit its uptake. Se, as a part of the human body’s

Selenium and Androgen Receptor in Prostate Cancer

757

antioxidant defense system, has been shown to inhibit tumorigenesis in a variety of experimental models and clinical studies. Its crucial role is underlined by the fact that it is the only trace element to be specified in the genetic code (Rayman 2005) as selenocysteine (SeCys), which protects tissues and membranes from oxidative stress and controls cell redox status when incorporated into selenoproteins. In addition to SeCys, Se can also replace sulfur in methionine, forming selenomethionine, which can be nonspecifically incorporated into proteins. Finally, Se can be tightly bound by certain proteins, named Se-binding proteins to distinguish them from real selenoproteins (Patrick 2004). If Se may protect against prostate cancer, an adequate intake of Se is necessary. However, intake of Se is inadequate for cancer prevention in many countries (Rayman 2005). Average dietary intake worldwide usually ranges 20–300 mg/day (WHO 1987). In some countries intake is even lower than the 50 mg/day that is believed required for the full expression of plasma glutathione peroxidase (Thomson et al. 1993; Duffield et al. 1999). More recent studies have shown that the Se intake requirement is higher for full selenoprotein expression than that required for plasma glutathione peroxidase expression (Xia et al. 2005). Increasing evidence points to requirement of supranutritional doses of Se for reduction in cancer risk (Combs 2001; Rayman 2000). The relationship between Se and prostate carcinogenesis is complex. More Se is not better since overaccumulation of Se is potentially dangerous, as indicated by the U-shaped dose-response curve of Se in association with increased DNA damage in aging dogs (Waters et al. 2003). Se tissue concentrations (free nonmercury bound Se) are highest in kidney and the pituitary gland followed by thyroid, liver, spleen, and cerebral cortex (Drasch et al. 2000). Many forms of Se are being used in in vitro and in vivo studies: inorganic forms like selenite, selenate and organic forms like selenomethionine, methylseleninic acid (MSeA), and selenomethylselenocysteine. Se is highly bioavailable with selenite being >80% bioavailable, whereas selenomethionine or selenate are >90% bioavailable (Combs and Combs 1984). Selenomethionine is the major organic compound in cereal grains, grassland legumes, and soybeans as well as in Se-enriched yeast used for Se supplementation (Whanger 2002). Selenomethionine is the most bioavailable form of Se and may be converted to SeCys via transulfuration or degraded to hydrogen selenide by b-lyase. Tissue storage of Se was found to be higher with selenomethionine compared to SeCys, selenite or selenate. Use of Se as a chemopreventive agent requires that it be considered a pharmacological agent and supranutritional doses may be administered, which raises concerns about the safety margin and potential side effects (Reid et al. 2004; Squires and Berry 2006). The current interest in Se stems from a trial that analyzed its effect on nonmelanoma skin cancer (Combs et al. 1997); this study randomized 1,312 participants to a dose of 200 mg of Se/day in the form of selenized yeast or placebo. The primary end point was not reached, but a reduction in the rate of prostate cancer incidence at 4.5 and 7.4 years was observed. The highest association between decrease in prostate cancer incidence and dose was noted in participants with initially low levels of Se and in men below 65 years with a prostate-specific

758

N. Nadiminty, A.C. Gao

antigen (PSA) level <4 ng/ml. Several studies of association between Se levels and prostate cancer risk have been published and several are still ongoing. Almost all the studies have reported a reduced risk of prostate cancer in healthy men or those with baseline PSA level of >4 ng/ml where a significant reduction in cancer burden in the highest category of Se concentration was noted (Rayman 2005; van den Brandt et al. 2003; Li et al. 2004). In a number of these studies, Se was found to have a stronger protective effect against advanced prostate cancer which suggests that the primary effect of Se was on prostate cancer progression and not initiation.

3 Clinical Trials and Epidemiological Studies While several large-scale phase III clinical trials using Se are ongoing, a number of small earlier phase studies have contributed to the planning of these trials. Studies conducted by Shamberger and Frost (1969), Schrauzer et al. (1977a, b), Clark and Marshall (2001), and Yu et al. (1988) have found a marginal but evident inverse relationship between Se intake and overall risk of cancer. In most studies, the most pronounced effect of Se intake was observed for gastrointestinal and prostate cancers and the lowest effect for pancreas, skin, and bladder cancers (Meuillet et al. 2004; Reid et al. 2004). More recent epidemiological data support the inverse correlation between Se level and prostate cancer risk (Vogt et al. 2007). Serum Se levels were measured in 212 men with prostate cancer and 233 age-matched controls. Se levels were inversely associated with risk of prostate cancer with similar patterns seen in both Caucasian and African-American men. This study also showed that vitamin E levels were highest in men in the highest Se quartile. An interesting study among the Inuit, whose diet is rich in omega-3 polyunsaturated fatty acids and Se, found only one case of prostate cancer and no latent cancers in 61 men dying of other causes (Dewaily et al. 2003). The Nutritional Prevention of Cancer study found a significant reduction (49%) in prostate cancer risk among study subjects. Kranse et al. (2005) showed that dietary intervention with a dietary supplement containing carotenoids, Se, and some other substances could reduce dihydrotestosterone (DHT) and testosterone levels and increase free PSA and total PSA doubling time. Venkateswaran et al. (2004) have shown that treatment of the 12T-10-Lady transgenic prostatic adenocarcinoma model with antioxidants (vitamin E, Se, and lycopene) decreased the incidence of prostate cancer and increased disease-free survival. They also found that prostate cancer developed in 75% and 100% of controls in the standard and high fat diets, respectively, whereas only 10% and 15% of animals in the antioxidant-treated group developed tumors, which was accompanied by increased levels of p27(kip1) and decreased levels of PCNA expression. On the other hand, two European studies (Lipsky et al. 2004; Allen et al. 2004) studied the correlation between toenail Se levels and risk of prostate cancer and found that Se may not be strongly associated with reduced prostate cancer risk since men in the highest Se quartile had only a slightly reduced risk compared to men in

Selenium and Androgen Receptor in Prostate Cancer

759

the lowest quartile (Nyman et al. 2004). Karunasinghe et al. (2004) studied levels of accumulated DNA damage in a group at high risk for prostate cancer and found that lower Se serum levels had a statistically significant inverse relationship with accumulated DNA damage. Se intake in this population was not sufficient for adequate repair of DNA damage, thus increasing susceptibility to cancer. Another study analyzed the Se levels and glutathione peroxidase activities in whole blood, plasma, and prostates of 32 prostate cancer and 40 patients with benign prostatic hyperplasia (BPH) and 39 healthy subjects. Se concentrations in blood and plasma were found to be lower in both prostate cancer and BPH patients, whereas prostate tissue Se level was higher in both groups compared to controls. In contrast, prostate tissue glutathione peroxidase activity was lower in prostate cancer patients compared to BPH patients, which indicates that higher levels of Se do not necessarily mean higher efficacy in anticancer effects (Zachara et al. 2005a). Further corroboration was provided by the finding that the level of Se in the prostate gland was similar in control subjects and BPH patients (156 ng/g wet wt), while in prostate cancer patients this level was higher (182 ng/g wet wt) (Zachara et al. 2005b). Several large prospective studies have been conducted to unravel the relationship between Se and prostate cancer (Stratton et al. 2003a, b; Knekt et al. 1990; Yoshizawa et al. 1998; Helzslouer et al. 2000; Nomura et al. 2000; Brooks et al. 2001; Goodman et al. 2001; Van den Brandt et al. 2003; Li et al. 2004). Most of these studies show a reduced risk of prostate cancer overall for the highest vs. lowest Se status, with significant reduction in risk being observed in some of them. When the analysis is confined to subjects who had advanced prostate cancer or a baseline PSA level >4 ng/ml, six of these studies showed a significant reduction in prostate cancer risk in subjects with the highest Se status. In a number of these studies (Yoshizawa et al. 1998; Nomura et al. 2000; van den Brandt et al. 2003; Li et al. 2004) the protective effect of Se was found to be stronger for advanced prostate cancer vs. localized disease. Li et al. (2004) found that the protective effect of Se was stronger for all types of prostate cancer, but strongest in men with a baseline PSA level >4 ng/ml. Taken together, these studies suggest an effect of Se on disease progression rather than on initiation (Rayman 2005; Navarro-Silvera and Rohan 2007). Intervention trials also have attempted to establish a correlation between Se supplementation and lower prostate cancer risk. A double blind placebo-controlled cancer prevention trial by Clark et al. (1996) showed a 63% reduction in prostate cancer incidence in men supplemented with 200 mg/day Se (selenized yeast). A follow-up of this study continues to show a significant reduction in the incidence of prostate cancer following Se supplementation (Duffield-Lillico et al. 2003). The positive outcome of this study encouraged the implementation of several other trials. A study in men with normal pretreatment serum Se levels found that 200 mg oral Se per day resulted in higher levels of Se in prostate tissue compared to placebo in 51 men who underwent transurethral resection of the prostate for BPH (Gianduzzo et al. 2003). This suggests that oral administration of Se is clinically relevant and concentrates Se in prostate tissue (Klein and Thompson 2004). A recent 4 1 week double-blind, randomized crossover study in which healthy

760

N. Nadiminty, A.C. Gao

young men supplemented their diets with selenate and Se-enriched yeast (300 mg/ day Se) and Se-enriched milk (480 mg/day Se) found that even though all sources of Se increased serum levels of Se, short-term supplementation did not modulate blood lipid markers or expression and activity of several selenoenzymes (RavnHaren et al. 2007). Another randomized, placebo-controlled double-blind crossover dietary intervention study treated 37 hormonally untreated prostate cancer patients with increased PSA levels with a dietary supplement (verum, containing Se in addition to plant estrogens, antioxidants, and other putative anticancer agents) for 6 weeks. The intervention reduced DHT and testosterone levels and increased free PSA and total PSA doubling time (Kranse et al. 2005). A report by Sabichi et al. (2006) has shown that orally administered Se can accumulate preferentially in the prostate gland as opposed to the seminal vesicles. These findings are from a randomized and controlled Southwest Oncology Group study, which showed that oral administration of 200 mg/day Se in the form of L-selenomethionine increased Se levels in the prostate tissues by 22% compared to controls. But almost none of these studies evaluated expression of AR or AR-regulated genes in their study subjects. It is critical that attempts be made to analyze AR status in either observational or intervention studies of large male populations with respect to Se status. These equivocal results after analysis of completely different populations with varying life styles and environmental conditions prompted the initiation of large, randomized, and controlled studies using Se like the Selenium and Vitamin E Cancer Prevention Trial (SELECT), the Watchful Waiting and the Negative Biopsy studies in the United States, the Prevention of Cancer by Intervention with Selenium (PRECISE) in three European countries, and the Australian Prostate Cancer Prevention Trial using Selenium (APPOSE). The SELECT trial sponsored by the NCI is a phase III randomized double-blind placebocontrolled trial designed to test the efficacy of Se (200 mg L-selenomethionine) and vitamin E (400 mg DL-a-tocopherol) alone and in combination in the prevention of prostate cancer (Klein 2004). Accrual of 32,400 volunteers has been completed and final results are expected in 2013. For prostate cancer, the androgen axis is an important risk factor in addition to modifiable risk factors like diet, obesity, and screening history and nonmodifiable risk factors like age, race, family history, and presence of some genetic polymorphisms. SELECT will also assess, in a nested case-control study, genetic polymorphisms of four genes–AR, 5a-reductase type II, cytochrome p450c, 17b and a-hydroxysteroid dehydrogenase–on prostate cancer incidence to identify potential targets for screening and intervention. If such biomarker associations with increased risk of prostate cancer are validated, targeted interventions can be developed to address disparities in prostate cancer risk (Pak et al. 2002). A large-scale phase III study is ongoing to examine the impact of Se on men with high-grade prostatic intraepithelial neoplasia (HGPIN) with the hypothesis that Se decreases the risk of subsequent diagnosis of prostate cancer (Clark and Marshall 2001). This trial, designed by the Southwest Oncology Group, tests the chemopreventive effect of Se in the form of selenomethionine (200 mg Se/day) to prevent a diagnosis of prostate cancer in men with HGPIN (Marshall et al. 2006). The APPOSE trial is a randomized, controlled chemoprevention trial with a cohort

Selenium and Androgen Receptor in Prostate Cancer

761

size of 2,000 men at increased risk for prostate cancer (due to family history of prostate cancer). Each group receives 200 mg Se daily or placebo (Costello 2001). The subjects will be observed over a period of ten years with annual digital rectal exam, serum PSA, and Se measurement. These studies use different forms of Se. Selenized yeast (200–800 mg/day) is used in several trials underway like the Negative Biopsy study (200 and 400 mg/day Se) (Stratton et al. 2003a), the preprostatectomy study (Marshall 2001), and the Watchful Waiting study (Stratton et al. 2003b), whereas the SELECT study uses selenomethionine (Thompson 2007). Earlier human clinical trials, which used organic forms of Se as chemopreventive agents against prostate cancer, have shown promising results (Clark et al. 1996; Clark et al. 1998; Clark and Jacobs 1998). These studies will shed light on the differences in bioavailability and metabolism of these forms of Se and will pave the way for the design of future prevention efforts in a clinical setting. Taken together these trials and studies provide a rationale for the mechanistic evaluation of different Se compounds in the chemoprevention of initiation and progression of prostate cancer. Note: An initial analysis in October 2008 of data from more than 35,000 men of age 50 and older in the SELECT trial showed that selenium and vitamin E supplements, taken either alone or together, did not prevent prostate cancer. Therefore, the National Cancer Institute (NCI) ordered the SELECT trial closed earlier than the scheduled end in late 2011. The data also showed two concerning trends: a small but not statistically significant increase in the number of prostate cancer cases among the over 35,000 men of age 50 and older in the trial taking only vitamin E and a small, but not statistically significant, increase in the number of cases of adult onset diabetes in men taking only selenium. Neither of these findings proves an increased risk from the supplements and both may be due to chance. Researchers will continue to track the men’s health for three years before the results of the data analysis will be published.

4 Se AntiCancer Metabolites After digestion, Se is metabolized to physiologically active methylselenol (CH3SeH) or incorporated into antioxidant enzymes and other selenoproteins. The generation of methylselenol from selenium compounds is mediated by enzymes like g-glutamyl-selenomethylselenocysteine synthetase, b-lyase, etc. Methylselenol may be formed directly from selenomethionine by the action of a b-lyase also called methioninase (Zhao et al. 2006). Previous studies have shown that the choice of chemical form and dose of Se can influence the observed biological effects. Inorganic Se compounds like selenite or selenate are known to produce genotoxic effects, whereas organic Se-containing compounds are better tolerated and exhibit anticancer activity. Organic Se is present in the forms of selenomethionine, SeCys, and selenomethylselenocysteine. The differences in anticarcinogenic activities of the various forms of Se can be attributed to their

762

N. Nadiminty, A.C. Gao

metabolism. Metabolism of selenite is tightly regulated and forms hydrogen selenide after reaction with reduced glutathione. Hydrogen selenide can then act as a precursor for the synthesis of selenoproteins or undergo enzymatic methylation to generate methylselenol (Ip 1998; Ohta and Suzuki 2008). Selenomethionine can either be incorporated into general body proteins, be converted to hydrogen selenide and follow the same pathway as selenite, or be converted to methylselenol via methioninase. Synthetic Se compounds like methylseleninic acid (MSeA) can generate monomethylated Se easily and have been used extensively in in vitro studies to generate the active anticancer metabolite, methylselenol. Several other Se metabolites have been shown to have anticarcinogenic effects in cell or animal model systems: selenodiglutathione (GSSeSG), hydrogen selenide (H2Se), and the methylated selenides ([CH3]xSe), which are the excretory forms of the element (Fig. 1). Monomethylated Se compounds that are precursors of methylselenol, like MSeA, induce apoptosis through biochemical and cellular processes that are distinct from those induced by forms of Se that enter the hydrogen selenide pool like sodium selenite. Methylseleninic acid (CH3SeO2H) was developed specifically for in vitro studies since cultured cells do not metabolize selenomethionine adequately due to very low levels of b-lyase activity (Ip et al. 2000). MSeA and other precursors of methylselenol induce apoptosis via the caspase pathway and without induction of DNA single-strand breaks, whereas sodium selenite is genotoxic with induction of single-strand breaks and induces apoptosis without activation of caspases. A recent study suggests that selenite metabolism leads to generation of reactive oxygen species, mainly superoxide and hydroperoxide, which lead to DNA strand breaks and to p53 phosphorylation, ultimately leading to mitochondrial leakage and caspase activation. This explains the higher toxicity of selenite in p53-wild-type prostate cancer cells like LNCaP compared to MSeA and less efficient induction of apoptosis in p53-null prostate cancer cells like DU145 and PC-3 (Li et al. 2007). Certain organoselenium compounds have promise in delaying both early and late events in cell cycle progression of prostate cancer cells. Chemoprevention using organoselenium compounds involves apoptosis as a critical event and inhibits angiogenic molecules in contrast to those that enter the

Selenomethionine Selenoproteins (Selenocysteine) Selenocysteine

GS-Se-SG Hydrogen selenide (H2-Se)

Methionine α, γ-lyase

Selenite

Methylselenol (CH3SeH)

GS-SeH Selenomethylcysteine Methylseleninicacid Methylselenocyanate

Fig. 1 Schematic representation of metabolites and intermediates in the metabolism of Se and incorporation into selenoproteins

Selenium and Androgen Receptor in Prostate Cancer

763

hydrogen selenide pool. Jiang et al. (2000b) suggested that different pathways might be activated by these two metabolite pools in exerting their antiproliferative effects on cancer cells. Corcoran et al. (2004) have shown that inorganic Se in the form of sodium selenate was more effective than selenomethionine, methylselenocysteine, and selenized yeast in inhibiting progression to castration-resistant prostate cancer.

5 Mechanisms of Se AntiCancer Effect Several mechanisms have been suggested to mediate the anticancer effects of Se. The major ones are reduction of DNA damage (Rayman 2005; Li et al. 2007; Zhao and Brooks 2007; Zu et al. 2006), oxidative stress (Drake 2006; Pinto et al. 2007), and inflammation; induction of phase II-conjugating enzymes that detoxify carcinogens; enhancement of immune response (Ryan-Harshman and Aldoori 2005; Eng et al. 2003); incorporation into selenoproteins; alterations in DNA methylation status of tumor suppressor genes (Yu et al. 2007; Ramachandran et al. 2007); inhibition of cell cycle (Klein 2004); and angiogenesis and induction of apoptosis (Wang et al. 2008). Se inhibits carcinogenesis at both the initiation and later stages of carcinogenesis in animal models. Se anticarcinogenic effect reversibly alters carcinogen metabolism, inhibits tumor cell proliferation, enhances apoptosis, and suppresses tumor angiogenesis (Ip 1998; Combs and Gray 1998). Mechanisms of Se action on prostate cancer are not understood completely, but evidence suggests that Se induces apoptosis in prostate cancer cells, while having no, or very low, detrimental effect on benign prostate epithelium (Ghosh 2004; Syed et al. 2007). Inhibition of AR signaling, reduction in AR mRNA and protein levels, recruitment of corepressors to the androgen response elements (AREs) in the promoters of androgen-responsive genes, inhibition of signaling pathways like NF-kB, IL-6, Stat3 and induction of apoptosis are mechanisms specific for prostate cancer. Molecular and cellular bases for Se action in cancer prevention continue to emerge and provide plausible rationale for clinical trials like SELECT. Precursors of methylselenol in in vitro systems like MSeA have been shown to block cell cycle, induce apoptosis and inhibit angiogenesis by induction of caspases 1, 8, 10, and 12 (Zu and Ip 2003; Dong et al. 2003). Se also exerts its effects at physiological levels by a dose-dependent inhibition of growth, cell cycle progression, and induction of apoptosis by downregulation of cell cycle-related genes in PC-3 cells (Zhao et al. 2004; Dong et al. 2003). In another study, Schlicht et al. (2004) found an association between upregulation of IGFBP3 and RXR-a and progression of prostate cancer using human PC-3 and rat PAII prostate cancer cell lines. The emerging picture of Se action includes the induction of pathways resulting in growth arrest, caspase-mediated apoptosis, reduced androgen signaling, and impaired angiogenesis by Se metabolites (Combs 2004).

764

N. Nadiminty, A.C. Gao

6 Selenoproteins The human selenoproteome consists of 25 selenoproteins (Kryukov et al. 2003). The main groups are glutathione peroxidases 1–5, iodothyronine deiodinases 1–3, thioredoxin reductases, selenoprotein P, and other proteins with mostly unknown functions (Alexander 2007). The incorporation of Se into selenoproteins has been studied extensively and the effects of selenoproteins can vary by cell type, physiologic status, or the presence or absence of incorporated SeCys. Genetic variations (SNPs) in selenoproteins have been reported (Moscow et al. 1994; Kumaraswamy et al. 2000; Al-Taie et al. 2002; Hu et al. 2001). Different selenoprotein SNPs may respond differently to Se supplementation, which provides an explanation for the pharmacogenetic differences in the preventive effects exerted by Se. Identification of a 15-kDa selenoprotein in the rat prostatic glandular epithelium and the discovery that it bound the UDP-glucose:glycoprotein glucosyltransferase that is involved in protein folding led to speculation that Se may have a role in the regulation of protein folding (Korotkov et al. 2001). Selenoproteins have received a great deal of attention due to the finding that selenoprotein-P is downregulated in prostate cancer cell lines and in the progression from high-grade prostatic intraepithelial neoplasia (PIN) to metastatic prostate cancer (Calvo et al. 2002). A recent study reported that selenoprotein deficiency accelerates prostate carcinogenesis in a transgenic model (with targeted expression of SV40 large T-antigen and an altered SeCys tRNA) (Diwadkar-Navsariwala et al. 2006). Selenoenzymes, like glutathione peroxidase and thioredoxin reductase, were shown to be important in the protective effects of Se against cancer. Another selenoprotein, Selenoprotein H, has been shown to be overexpressed in LNCaP cells and mouse lung cancer cells, LCC-1 (Novoselov et al. 2007). Selenoprotein synthesis is a complex and highly regulated process. Se is incorporated as SeCys at specific UGA codons in 25 human proteins. Selenoprotein mRNAs all contain a SECIS element needed for the incorporation of SeCys into selenoproteins. Se supplementation may also modulate selenoprotein expression in prostate cancer cells. In an analysis of the modulation of selenoprotein expression in nontumorigenic prostate epithelial cells (RWPE-1) vs. prostate cancer cells (LNCaP and PC-3) glutathione peroxidases-1and -4 were shown to be elevated to higher levels in LNCaP and PC-3 cells compared to RWPE-1 cells in response to supplemental Se (0–250 nM sodium selenite, sodium selenate, or selenomethionine). Significant differences are proposed to exist between prostate cancer cells and benign epithelial cells in their ability to utilize organic sources of Se. These results demonstrate that selenoproteins and Se metabolism are differentially regulated at multiple levels in prostate cells (Rebsch et al. 2006). Clinical observations, which suggest that Se supplementation is only effective when baseline Se is at a level where selenoprotein production is regulated, lend credence to the hypothesis that selenoprotein production is a likely mediator of the effects of Se chemoprevention (Rebsch et al. 2006).

Selenium and Androgen Receptor in Prostate Cancer

765

7 Effects of Se on Signaling Pathways and Cellular Processes 7.1

The Androgen Receptor

The AR gene was cloned in 1988 and is located at Xq11.2. The AR protein consists of three functional domains, an N-terminal transactivation domain, a DNA-binding domain, and a C-terminal ligand-binding domain. The ligand-binding domain interacts with HSP 90 which dissociates upon binding ligand. After ligand binding, the AR translocates to the nucleus and interacts with AREs in the promoters and enhancers of androgen-responsive genes, like PSA. The AR is normally tumor suppressive in prostate cells but, as part of malignant conversion, undergoes a switch from tumor suppressor to tumor promoter. Further evidence indicates that prostate cells in which AR protein is stabilized in mitosis undergo apoptosis (Vander Griend et al. 2007). The transactivation by AR is modulated by 300-plus AR-interacting coregulators that can either promote (coactivators) or inhibit (corepressors) AR function. These coregulators are broadly divided into four categories (1) molecular chaperones that coordinate AR maturation and movement, (2) histone modifiers (CBP/p300, NCoR), (3) co-ordinators of transcription (TRAP/DRIP/ ARC), and (4) DNA structural modifiers (SWI/SNF/BRG1) (Chmelar et al. 2006). The molecular interactions between the AR and many of these coregulators are not known precisely, knowledge of which will provide many molecular targets for AR-targeted therapies. Alterations in coregulator levels and function have been proposed to contribute to the emergence of castration resistance. When coregulator levels change, AR function will conceivably change as a consequence. For example, an environment enriched for coactivators could make the AR more sensitive or responsive to lower levels of agonists or allow promiscuous activation of the AR by abundant low-affinity ligands. However, it remains to be determined whether a causal relationship exists between cofactor expression levels and initiation or progression of prostate cancer. Ligand-associated AR signaling drives cell proliferation in benign and malignant prostate cells and many signaling pathways are constitutively active that allow persistent growth of prostate cancer cells. Prostate cells accumulate genetic and epigenetic changes leading to neoplastic transformation of benign cells and metastatic disease in the case of cancer cells. Several recent studies have suggested that androgen-deprivation therapy does not constitute ‘‘androgen absence,’’ and that low levels of androgen produced by the adrenal glands or the tumor itself are sufficient to activate AR signaling (Mohler et al. 2004, Titus et al. 2005). This has been postulated variously to be due to hypersensitivity of the receptor to low levels of androgens, AR gene amplifications, or mutations leading to gain of function of AR (Sommer and Haendler 2003, Scher and Sawyers 2005, McLeod and Srivastava 2006) as well as ligand-independent activation of the AR. The neuroendocrine component of the evolving prostate cancer can secrete factors able to regulate AR activity by enhancing sensitivity to lower androgen levels in a ligand-dependent manner, increasing AR nuclear levels, and/or directly activating

766

N. Nadiminty, A.C. Gao

the AR in a ligand-independent manner (Jin et al. 2004). These studies imply that AR signaling is important in the development as well as progression of prostate cancer and chemopreventive agents that can block AR signaling would be beneficial in many clinical settings. Progression to castration resistance is signaled frequently by a rise in serum PSA during androgen-deprivation therapy and confirmed often by increase of metastatic tumor volume. Amplification of the AR leading to increased sensitivity and response to very low levels of androgens was found in about 30% of tissues of castration-resistant prostate cancer and found to be positively related to increased progression measured by the proliferation marker, Ki-67 (Haapala et al. 2007). Hypersensitivity of the AR may also result from an imbalance of the increased number of receptor molecules and the available number of competing molecules of antiandrogen in the nucleus (de Jong et al. 1992). High androgen levels are preserved in tissues from metastases in patients with castration-resistant prostate cancer, which suggests a high level of intracrine activity. A study by Stanbrough et al. (2006) showed increased expression of a number of genes involved in the androgenic part of the steroid metabolism, 3bHSD2, AKR1C3, SRD5A1, AKR1C1, C2, and UGT2B15 along with overexpression of the AR. Another study (Montgomery et al. 2007) found that tissues from bone metastases from castration-resistant prostate cancer patients contained higher levels of testosterone and equal levels of DHT to tissues from nonandrogen-deprived patients. They also found upregulation of transcripts of enzymes involved in androgen synthesis, which shows that synthesis and degradation of androgens occur in castration-recurrent prostate cancer. In benign prostate cells, AR-stimulated growth occurs in a stromal cell-dependent manner with emphasis on the growth suppressor function of the AR. In prostate cancer, the AR acquires gain-of-function changes allowing it to drive prostate cancer cell survival and proliferation in a cell-autonomous manner (D’amico et al. 2007). Changes in AR nongenotropic signaling have been shown to induce androgen independence in LNCaP human prostate cancer cells (Unni et al. 2004). Recent studies also indicate that AR signaling has a major role in advanced prostate cancer (Chen et al. 2004). At this stage, even though circulating levels of androgens are low, mutations or amplification of the AR enable it to respond to other steroids and result in activation of the AR in a ligand-independent fashion. Several AR mutations have been shown to be present in cases of partial or complete androgen insensitivity syndromes as well as prostate cancer (Taplin et al. 2003; Taplin 2007). AR continues to play a major role in these situations and is therefore a viable target for therapy. Even though the prospect of discovery of antiandrogens that specifically target the AR without any side effects is attractive, the enormous amount of work extended thus far without success makes a search for effective inhibitors of androgen signaling using currently available therapeutic or preventive agents tempting. In this scenario, Se is one of the most promising.

Selenium and Androgen Receptor in Prostate Cancer

7.2

767

Se and Androgen Receptor

An oral agent that can reduce AR expression and signaling may prove beneficial against prostate cancer since AR signaling is often hyperactive or hypersensitive to low levels of androgen (Pienta and Bradley 2006). Mechanisms of Se chemopreventive effect specific for prostate cancer are the inhibition of AR signaling; inhibition of signaling pathways like NF-kB, IL-6, Stat3; and induction of apoptosis. Mechanisms involved in reduction of AR signaling by Se include reduction in AR mRNA transcription (Dong et al. 2003) and stability, increase in AR protein turnover, reduction in AR translocation and inhibition of coactivator recruitment followed by corepressor recruitment to promoters of AR target genes (Chun et al. 2006). Gene expression studies using cDNA microarrays in both AR-positive (LNCaP) and AR-negative (PC-3) prostate cancer cells have indicated that Se exerts its effects at physiological levels by a dose-dependent inhibition of growth, cell cycle progression, and induction of apoptosis (Dong et al. 2003; Zhao et al. 2004; Schlicht et al. 2004). These authors also found that Se downregulated genes like CYCLIN A, CDK1, CDK2, CDK4, DHFR, PCNA while upregulating genes like GADD153, CASPASE-9, CHK2, P19, P21, RXR, and several zinc finger proteins in PC-3 cells. On the other hand, Se also modulated AR signaling and expression and decreased expression of AR-regulated genes in AR-positive LNCaP cells, which was not seen in AR-negative PC-3 cells. Se inhibited AR activity in LNCaP cells which was accompanied by a reduction in cell proliferation (Morris et al. 2006). Dong et al. (2004) have shown that 10 mM Se in the form of MSeA downregulates the expression of AR target genes like PSA by disruption of AR signaling and inhibition of AR mRNA and protein expression. Se also inhibited the transactivating activity and DNA-binding ability of AR. When benign prostate epithelial cells were treated with 100–200 mM Se or selenomethionine, DNA synthesis was dose dependently reduced accompanied by a concomitant reduction in AR activity (Morris et al. 2006). Chun et al. (2006) have shown that Se (10 mM MSeA) reduces the recruitment of coactivators to the promoters of AR-dependent genes while simultaneously enhancing the recruitment of corepressors. They also showed that AR nuclear translocation was affected by treatment with MSeA. In vitro studies showed that MSeA can inhibit LNCaP, DU145, and PC-3 prostate cancer cell growth, and in vivo studies have shown that tumor growth is also decreased in xenografts (Chun et al. 2006, Lee et al. 2006). MSeA also regulated the expression of AR, PSA, and phase II detoxification enzymes, thereby inducing cellular defenses against carcinogens. The mechanisms through which different Se compounds inhibit the AR signaling pathway vary; for example, both MSeA and selenomethionine produce mixed effects on the expression of androgen-responsive genes. MSeA regulates the expression of AR, PSA, and phase II detoxification enzymes, thereby inducing cellular defenses against carcinogens. MSeA affects the expression of more androgen-responsive genes than selenomethionine and also to a greater degree. While 10 mM MSeA after 48 h can reduce the expression of both AR and PSA, 10 mM

768

N. Nadiminty, A.C. Gao

selenomethionine does not show any effect on these parameters. Higher concentrations or longer exposures are required to manifest the effect of selenomethionine. MSeA decreases transcription of genes from all phases of the cell cycle, which suggests that MSeA causes LNCaP cells to exit cell cycle rather than inhibit a specific phase of the cell cycle. On the other hand, selenomethionine enriches transcripts in the G1/S phase and downregulates transcripts involved in the G2/M progression, which suggests that selenomethionine induces a G2/M arrest in LNCaP cells (Menter et al. 2000, Ni et al. 2003). Wang et al. (2002) have shown that addition of methioninase increases the anticancer efficacy of selenomethionine by generating methylselenol and inducing apoptosis. As low as 1 mM selemethionine was sufficient to induce antiproliferative effects and inhibition of Akt phosphorylation in DU145 cells, which express high levels of methioninase. Androgen-regulated genes, like KLK3, NKX3.1, and GUCY1A3, were downregulated over different time points by MSeA but not by selenomethionine (Zhao and Brooks 2007). Both agents induce changes in the transcriptional program in LNCaP cells. In vitro studies have shown that micromolar concentrations of selenomethionine inhibit growth and induce apoptosis in LNCaP, PC-3, and DU145 prostate cancer cells but not in primary fibroblast cultures or primary prostate epithelial cells (Dong et al. 2005; Hu et al. 2005). MSeA induces apoptosis and inhibits growth of PC-3 cells at a concentration of 5–10 mM, and similar to selenomethionine the mechanisms of apoptosis induction include DNA fragmentation and PARP cleavage (Dong et al. 2005; El-Bayoumy and Sinha 2005). LNCaP cells are more sensitive to selenomethionine than the androgen-independent cells PC-3 and DU145. This has been suggested to be due to the inhibition of the interaction of AR with the early growth response protein 1 (EGR-1) by selenomethionine (Yang and Abdulkadir 2003). MSeA and selenite also inhibit AR signaling by different mechanisms (Husbeck et al. 2006a; Hu et al. 2006). Selenite reacts with reduced glutathione in cells and produces superoxide radicals. This finding was confirmed by use of the antioxidant, N-acetylcysteine, which blocked the downregulation of AR and PSA expression by selenite. MSeA did interact with reduced glutathione but did not produce superoxide radicals as shown by the fact that N-acetylcysteine did not inhibit AR downregulation by MSeA. They also reported that selenite decreased AR expression by decreasing Sp1 expression, which reduced the binding of Sp1 to its target site in the AR promoter, and this effect was not seen with MSeA. Cho et al. (2004) have shown that methylselenol or MSeA specifically and rapidly inhibit PSA expression through induction of PSA protein degradation and suppression of androgen-stimulated PSA transcription in LNCaP cells. This inhibitory effect was not observed with sodium selenite or selenomethionine. MSeA has also been shown to potentiate apoptosis induced by chemotherapeutic agents in androgen-independent prostate cancer cell lines (Alexander 2007). MSeA increased the potency of SN38 (topoisomerase I inhibitor), etoposide (topoisomerase II inhibitor), and paclitaxel in inducing apoptosis of DU145 and PC-3 cells, and this effect depended on

Selenium and Androgen Receptor in Prostate Cancer

769

interactions with JNK-dependent targets to amplify the caspase 8-initiated apoptotic cascades. Monomethylated Se in the form of methylselenocysteine at pharmacological doses is able to inhibit the growth of LNCaP xenografts in nude mice, which was accompanied by a decrease in AR and PSA expression. These results are interesting in light of the fact that Se has been shown to be able to act as an antiandrogen in addition to its role as a chemopreventive agent (Lee et al. 2006). Doses of Se that are too low to show full chemopreventive activity may be used to antagonize AR signaling in a clinical setting. Most of these studies focused on the effects of Se on AR signaling by studying expression of PSA, the prototypic androgen- and AR-responsive gene. But recent studies indicate that effects on PSA may not represent the entire spectrum of AR action, which includes effects on growth, survival, angiogenesis, and other signaling pathways via its interaction with their signaling mediators. Hence, better molecular markers and readouts of AR action in castration-recurrent prostate cancer should be developed to evaluate a possible responsiveness of AR to Se.

7.3

Cell Cycle

Deregulated cell cycle progression is one of the hallmarks of cancer cells. Cell cycle progression is regulated by the activity of CDKs, cyclins, and CDK inhibitors. CDKs are activated at different phases of the cell cycle, for example, CDK4 and CDK6 are activated in early and mid-G1 phase along with D-type cyclins, whereas CDK2-cyclin E and CDK1(cdc2)-cyclin B are activated during the S and G2/M phases, respectively (Singh and Agarwal 2006). Several genetic or epigenetic changes are known to deregulate cdc2 kinase activity in cancers including prostate cancer (Fu et al. 2004, Kawabe 2004). Many prostate cancers harbor mutations in at least one of the CDKs, most frequently CDK2, CDK4, CDK6, or CDK1. Taken together, agents that provide checks in the process of deregulated cell proliferation due to aberrant activation of CDKs warrant further attention as chemotherapeutic or chemopreventive agents. Se has been shown to modulate the expression levels and activities of cyclins and cyclin-dependent kinases. Treatment with MSeA or selenite causes G1 and S arrest in DU145 cells. G1 arrest induced by MSeA is accompanied by increased expression of cdk inhibitors, such as p27kip1 and p21cip1, and downregulation of cdk2 (Jiang et al. 2000b; Dong et al. 2003; Jiang et al. 1999). 2–10 mM MSeA was shown recently to induce G1 arrest in immortalized human microvascular endothelial cells and halt progression to S-phase (Wang et al. 2008). This study also showed that MSeA was able to reduce tumor microvessel density in DU145 xenografts. These effects of MSeA were attributed to induction of increased binding of CDK inhibitors to CDKs 2, 4, and 6, and inhibition of hyperphosphorylation of Rb, thereby increasing steady-state levels of Rb-E2F1 complexes.

770

7.4

N. Nadiminty, A.C. Gao

NF-kB

NF-kB has been implicated in the initiation and progression of several types of cancer, including prostate cancer (Karin et al. 2002). Targets of the NF-kB pathway include cytokines, chemokines, cell adhesion molecules, survival and growth factors, and genes involved in metastasis and angiogenesis. Aberrant activation of the NF-kB transcription factor is found in castration-recurrent prostate cancer (Suh and Rabson 2004) and agents that can inhibit the NF-kB pathway may prove beneficial for treatment of prostate cancer. Se inhibits NF-kB activity in androgen-sensitive and androgen-independent prostate cancer cell lines (Gasparian et al. 2002). A recent report has found that higher concentrations of Se (7.6 mM MSeA) compared to lower concentrations (50 nM MSeA) reduced NF-kB DNA binding and consequently reduced the rates of transcription and mRNA levels of NF-kBregulated genes (Christensen et al. 2007). These effects occurred in the absence of activation or translocation of the NF-kB transcription factor itself, which suggests that Se may interact with the protein itself or may affect the recruitment of coactivators.

7.5

Epidermal Growth Factor Receptor

EGFR has been identified as being involved in the pathogenesis of many solid tumors including prostate cancer (Liao et al. 2005). Many studies have shown that increased expression of ligands for the EGFR pathway and high levels of EGFR lead to the formation of a constitutively active mitogenic signaling loop in androgen-independent prostate cancer cells (Lorenzo et al. 2003). Human HGPIN, primary and metastatic prostate cancer show frequent overexpression of the EGFR family members (Mendelsohn and Baselga 2000). Few reports exist about the effects of Se on EGFR signaling. One study found that 5 mM organic Se in the form of selenomethionine (24 h) enhanced EGFR expression in androgen-sensitive but not androgen-independent prostate cancer cells (Pinto et al. 2007).

7.6

Generation of Reactive Oxygen Species

Reactive oxygen species and the intracellular redox state have emerged as important determinants of cell signaling (Husbeck et al. 2006a, b). Alterations in the intracellular redox state can affect the activity of redox-sensitive proteins via the oxidation of critical cysteine residues that may have downstream effects on signal transduction and gene transcription. The generation of reactive oxygen species is a well-known and important mechanism of Se- and ionizing radiation-induced cytotoxicity, in which reactive oxygen species are scavenged by antioxidants like

Selenium and Androgen Receptor in Prostate Cancer

771

glutathione. The intracellular level of glutathione plays a major role in the ability of cells to withstand oxidative stress induced by reactive oxygen species. Levels of glutathione are elevated normally in cancer cells compared to benign cells, and the antitumor activities of Se compounds are dependent on the dose and chemical form of Se which can react with intracellular levels of glutathione. The inorganic form of Se, selenite, undergoes thiol-dependent reduction to selenide, which supplies Se for the synthesis of selenoproteins (Combs and Gray 1998). At lower concentrations, selenite acts as a micronutrient, but at higher concentrations, selenite is a potent pro-oxidant by producing superoxide ions. The addition of superoxide dismutase abolishes the cytotoxicity induced by selenite, which confirms the importance of oxidative stress in the mechanism of seleniteinduced cell death (Menter et al. 2000; Spallholz 1994). Organic forms of Se have been shown to have a predominantly antioxidant effect due to the incorporation of Se into selenoproteins and antioxidant enzymes, like glutathione peroxidase and glutathione-s-transferase, in the form of selenocysteine (Combs and Gray 1998; Spallholz 1994). Se has also been shown to induce endoplasmic reticulum stress in p53-null prostate cancer cells by redox modification of thiol-disulfide interchange in proteins leading to protein unfolding (Zu et al. 2006; Wu et al. 2005). MSeA induced several markers of endoplasmic reticulum stress like phosphorylated forms of PERK, eIF2a, GRP-78, and CHOP/GADD153. 10 mM selenite has been shown to enhance the radiosensitization of the LAPC-4 and DU145 human prostate cancer cells (Husbeck et al. 2005). Glutathione-depleted cells will be unable to eliminate reactive oxygen species and repair DNA damage induced by ionizing radiation. Therefore, in addition to other mechanisms of cytotoxicity, selenite may have clinical implications as a radiosensitizer. Although selenite, selenomethionine, methylselenol, and other anticancer metabolites of Se induce cancer-preventive effects in cell culture, they are present at very low concentrations in plasma and tissues (Combs and Gray 1998; Whanger et al. 1996). Hence, whether these effects occur in vivo at bioavailable concentrations of Se and whether they can translate to clinically relevant findings remain unclear.

7.7

Angiogenesis

Very little information is available as to the precise targets of Se in inhibition of angiogenesis in prostate cancer. Inorganic Se (sodium selenate) has been reported to inhibit the progression of experimental castration-resistant prostate cancer, which was accompanied by evidence of inhibition of angiogenesis (Corcoran et al. 2004). Monomethyl-Se can inhibit MMP-2 and VEGF expression, which was not seen with selenite (Jiang et al. 2000a). A recent report has shown that treatment with an oral dose of 1–3 mg/kg wt of Se reduced tumor microvessel density in DU145 xenografts (Wang et al. 2008). VEGF is the most important mitogenic and survival factor for vascular endothelial cells in both benign and malignant microenvironments. Binding of VEGF to its receptors stimulates

772

N. Nadiminty, A.C. Gao

receptor dimerization and activation of a signaling cascade, which leads to the proliferation of endothelial cells. Targeting VEGF signaling may provide an attractive target for chemoprevention since angiogenesis is the cornerstone of solid tumor development. MSeA, at lower concentrations than needed to induce apoptosis, has been shown to inhibit VEGF expression both in the cellular and secreted forms in DU145 prostate cancer cells (Jiang et al. 2000b; Vadgama et al. 2000). Effects on VEGF expression may precede the effects on the subsequent signaling molecules and lead to a rapid and sustained reduction in tumor angiogenesis. The breakdown of tissue matrix to facilitate tumor cell migration is performed by matrix metalloproteinases and several tumor cell types, which include prostate cancer cells, have been shown to overexpress MMPs. MMP-2 and MMP-9 are overexpressed in invasive prostate cancer (Vadgama et al. 2000; Takaha et al. 2004). MSeA and methylselenocyanate have been shown to inhibit MMP-2 activity in a dose-dependent manner (Jiang et al. 2000a). Reduced MMP-2 activity resulted from decreased expression of MMP-2 and occurred after 0.5 h of exposure. Exposure to MSeA also decreased HUVEC cell tube formation ability in vitro. Information on the effect of Se on integrin signaling is lacking and needs to be addressed to adequately assess whether the observed effects of Se on VEGF and MMPs are specific to cancer cells or the results of a general inhibitory effect on cell migration and invasion.

8 Conclusions Whether and at which stages Se inhibits tumorigenesis remains unclear. Several in vitro studies and some clinical intervention trials like the NPC found that Se protection is due largely to reduction of local disease and inhibition of early steps in tumorigenesis. There has also been speculation about whether Se is involved in delaying or preventing transformation to cancer as opposed to inhibition or treatment of subclinical, microscopic cancer. Most reports agree that either effect would be largely beneficial with the alarming rate of latent disease found in men with PSA levels <4 ng/ml. Despite the significant advances made in elucidating the molecular pathology of prostate cancer, no single specific candidate gene or combination of genes has been assigned responsibility for tumorigenesis. Several studies have identified multiple, nonrandom events involving genetic as well as epigenetic alterations in multiple genes (Konishi et al. 2005). But whether these events are associated with either initiation or progression of prostate cancer remains unanswered. The AR is a multifunctional molecule in prostate cancer cells (Javidan et al. 2005); it transactivates many target genes and crosstalks with several signaling pathways, like growth factor receptors (Culig et al. 1994; Syed and Tolcher 2003; Reddy et al. 2006; Craft et al. 1999; Chan et al. 1998), MAPKs (Abreu-Martin et al. 1999; Fu et al. 2002; Uzgare et al 2003), Src kinase (Asim et al. 2008), cytokines like TNF-a (Aarnisalo et al. 1998; Harada et al. 2001), IL-6 (Ueda et al. 2002; Yang et al. 2003; Hobisch

Selenium and Androgen Receptor in Prostate Cancer

773

et al. 1998), and the Wnt pathway (Schweizer et al. 2008; Terry et al. 2006; Yang et al. 2006). Most of these interactions have been shown to promote cancer cell survival (Kaarbo et al. 2007; Schroder 2008). The ability of Se to affect either AR expression or activation can have important implications in this setting, and emphasis is needed in in vitro and in vivo studies on targeting and understanding the mechanisms of modulation of AR function by Se.

Acknowledgments This work was supported by NIH grants CA118887 and CA109441 (Gao AC).

References Aarnisalo P, Palvimo JJ, Janne OA (1998) CREB-binding protein in androgen receptor-mediated signaling. Proc Natl Acad Sci USA. 95: 2122–2127. Abreu-Martin MT, Chari A, Palladino AA, et al. (1999) Mitogen activated protein kinase kinase kinase 1 activates androgen receptor-dependent transcription and apoptosis in prostate cancer. Mol Cell Biol 19: 5143–5144. Alexander J (2007) Selenium. Novartis Found Symp 282: 143–149. Allen NE, Morris JS, Ngwenyama RA, et al. (2004) A case-control study of selenium in nails and prostate cancer risk in British men. J Cancer 90(7): 1392–1396. Al-Taie OH, Suefert J, Mork H, et al. (2002) A complex DNA-repeat structure within the selenoprotein P promoter contains a functionally relevant polymorphism and is genetically unstable under conditions of mismatch repair deficiency. Eur J Hum Genet 10(9): 499–504. Asim M, Siddiqui IA, Hafeez BB, et al. (2008) Src kinase potentiates androgen receptor transactivation function and invasion of androgen-independent prostate cancer C4–2 cells. Oncogene (EPub) 1–9. Brooks JD, Metter EJ, Chan DW, et al. (2001) Plasma selenium level before diagnosis and the risk of prostate cancer development. J Urol 166: 2034–2038. Calvo A, Xiao N, Kang J, et al. (2002) Alterations in gene expression profiles during prostate cancer progression: functional correlations to tumorigenicity and down-regulation of selenoprotein-P in mouse and human tumors. Cancer Res 62(18): 5325–5335. Chan JM, Stampfer MJ, Giovanucci E, et al. (1998) Plasma insulin-like growth factor-1 and prostate cancer risk: a prospective study. Science 279: 563–566. Chatterjee B (2003) The role of the androgen receptor in the development of prostatic hyperplasia and prostate cancer. Mol Cell Biochem 253(1–2): 89–101. Chen SD, Welsbie DS, Tran C, et al. (2004) Molecular determinants of resistance to antiandrogen therapy. Nat Med 10: 33–39. Chmelar R, Buchanan G, Need EF, et al. (2006) Androgen receptor coregulators and their involvement in the development and progression of prostate cancer. Int J Cancer 120: 719–733. Cho SD, Jiang C, Malewicz B, et al. (2004) Methyl selenium metabolites decrease prostate specific antigen expression by inducing protein degradation and suppressing androgen-stimulated transcription. Mol Cancer Ther 3(5): 605–611. Christensen MJ, Nartey ET, Hada AL, et al. (2007) High selenium reduces NF-kappaB-regulated gene expression in uninduced human prostate cancer cells. Nutr Cancer 58(2): 197–204. Chun JY, Nadiminty N, Lee SO, et al. (2006) Mechanisms of selenium down-regulation of androgen receptor signaling in prostate cancer. Mol Cancer Ther 5(4): 913–918. Clark LC, Jacobs ET (1998) Environmental selenium and cancer: risk or protection? Cancer Epidemiol Biomarkers Prev 7(10): 847–848.

774

N. Nadiminty, A.C. Gao

Clark LC, Marshall JR (2001) Randomized, controlled chemoprevention trials in populations at very high risk for prostate cancer: elevated prostate specific antigen and high-grade prostatic intraepithelial neoplasia. Urology 57: 185–187. Clark LC, Combs GF Jr, Turnbull BW, et al. (1996) Effects of selenium supplementation for cancer prevention in patients with carcinoma of the skin. A randomized controlled trial. Nutritional Prevention of Cancer Study Group. JAMA 276(24): 1957–1963. Clark LC, Dalkin B, Krongrad A, et al. (1998) Decreased incidence of prostate cancer with selenium supplementation: results of a double-blind cancer prevention trial. Br J Urol 81(5): 730–734. Combs GF Jr (2001) Considering the mechanisms of cancer prevention by selenium. Adv Exp Med Biol 492: 107–117. Combs GF (2004) Status of selenium in prostate cancer prevention. Br J Cancer 91: 195–199. Combs GF Jr, Combs SB (1984) The nutritional biochemistry of selenium. Annu Rev Nutr 4: 257–280. Combs GF, Gray WP (1998) Chemopreventive agents:selenium. Pharmacol Ther 79: 179–192. Combs GF, Clark LC, Turnbull BW (1997) Reduction of cancer risk with an oral supplement of selenium. Biomed Environ Sci 10(2–3): 227–234. Corcoran NM, Najdovska M, Costello AJ (2004) Inorganic selenium retards progression of experimental hormone refractory prostate cancer. J Urol 171: 907–910. Costello AJ (2001) A randomized, controlled chemoprevention trial of selenium in familial prostate cancer: Rationale, recruitment and design issues. Urology 57(4A): 182–184. Craft N, Shostak K, Carey M, et al. (1999) A mechanism for hormone-independent prostate cancer through modulation of androgen receptor signaling by the HER-2/neu tyrosine kinase. Nat Med 5:280–285. Culig Z, Hobisch A, Cronauer MV, et al. (1994) Androgen receptor activation in prostatic tumor cell lines by insulin-like growth factor-I, keratinocyte growth factor and epidermal growth factor. Cancer Res 54: 5474–5478. D’amico F, Biancolella M, Margiotti K, et al. (2007) Genomic biomarkers, androgen pathway and prostate cancer. Pharmacogenomics 8(6): 645–661. de Jong FH, Reuvers PJ, Bolt-de Vries J, et al. (1992) Androgens and androgen receptors in prostate tissue from patients with benign prostatic hyperplasia: effects of cyproterone acetate. J Steroid Biochem Mol Biol 52: 49–55. Dewaily E, Mulvad G, Sloth Pedersen H, et al. (2003) Inuit are protected against prostate cancer. Cancer Epidemiol Biomarkers Prev 12(9): 926–927. Diwadkar-Navsariwala V, Prins GS, Swanson SM, et al. (2006) Selenoprotein.deficiency accelerates prostate carcinogenesis in a transgenic model. Proc Natl Acad Sci USA 103: 8179–8184. Dong Y, Zhang H, Hawthorn L, et al. (2003) Delineation of the molecular basis for seleniuminduced growth arrest in human prostate cancer cells by oligonucleotide array. Cancer Res 63(1): 52–59. Dong Y, Lee SO, Zhang H, et al. (2004) Prostate specific antigen expression is down-regulated by selenium through disruption of androgen receptor signaling. Cancer Res 64(1): 19–22. Dong Y, Zhang H, Gao AC, et al. (2005) Androgen receptor signaling intensity is a key factor in determining the sensitivity of prostate cancer cells to selenium inhibition of growth and cancerspecific biomarkers. Mol Cancer Ther 4(7): 1047–1055. Drake EN (2006) Cancer chemoprevention: selenium as a prooxidant, not an antioxidant. Med Hypotheses 67(2): 318–322. Drasch G, Mail der S, Schlosser C, et al. (2000) Content of non-mercury associated selenium in human tissues. Biol Trace Elem Res 77: 219–230. Duffield AJ, Thomson CD, Hill KE, et al. (1999) An estimation of selenium requirements for New Zealanders. Am J Clin Nutr 70: 896–903. Duffield-Lillico AJ, Dalkin BL, Reid ME, et al. (2003) Nutritional Prevention of Cancer Study group, Selenium-supplementation, baseline plasma selenium levels and incidence of prostate

Selenium and Androgen Receptor in Prostate Cancer

775

cancer: an analysis of the complete treatment period of the Nutritional Prevention of Cancer Trial. BJU Int 91(7): 608–612. El-Bayoumy K, Sinha R (2005) Molecular chemoprevention by selenium: A genomic approach. Mutat Res 591: 224–236. Eng J, Ramsum D, Verhoef M, et al. (2003) A population-based survey of complementary and alternative medicine use in men recently diagnosed with prostate cancer. Integr Cancer Ther 2(3): 212–216. Fu M, Wang C, Wang J, et al. (2002) Androgen receptor acetylation governs transactivation and MEKK-1-induced apoptosis without affecting in vitro sumoylation and trans-repression function. Mol Cell Biol 22: 3373–3388. Fu M, Wang C, Li Z, et al. (2004) Minireview: Cyclin D1: Normal and abnormal functions. Endocrinology 145(12): 5439–5447. Gasparian AV, Yao YJ, Lu J, et al. (2002) Selenium compounds inhibit IkappaB kinase (IKK) and nuclear factor-kappa B (NF-kappa B) in prostate cancer cells. Mol Cancer Ther 1(12): 1079– 1087. Ghosh J (2004) Rapid induction of apoptosis in prostate cancer cells by selenium: reversal by metabolites of arachidonate 5-lipoxygenase. Biochem Biophy Res Commun 315: 624–635. Gianduzzo TR, Holmes EG, Tinggi U, et al. (2003) Prostatic and peripheral blood selenium levels after oral supplementation. J Urol 170(3): 870–873. Goodman GE, Schaffer S, Bankson DD, et al. & the Carotene and Retinol Efficacy Trial (CARET) Co-investigators (2001) Predictors of serum in cigarette smokers and the lack of association with lung and prostate cancer risk. Cancer Epidemiol Biomarkers Prev 10: 1069–1076. Grossmann ME, Huang H, Tindall TJ (2001) Androgen receptor signaling in androgen refractory prostate cancer. J Natl Cancer Inst 22: 1687–1697. Haapala K, Kuukasjarvi T, Hyytinen E, et al. (2007) Androgen receptor amplification is associated with increased cell proliferation in prostate cancer. Hum Pathol 38: 474–478. Harada S, Keller ET, Fujimoto N, et al. (2001) Long-term exposure of tumor necrosis factor alpha causes hypersensitivity to androgen and anti-androgen withdrawal phenomenon in LNCaP prostate cancer cells. Prostate 46: 319–326. Helzslouer KJ, Huang H-Y, Alberg AJ, et al. (2000) Association between alpha-tocopherol, gamma-tocopherol, selenium and subsequent prostate cancer. J Natl Cancer Inst 92: 2018– 2023. Hobisch A, Eder IE, Putz T, et al. (1998) Interleukin-6 regulates prostate specific protein expression in prostate carcinoma cells by activation of the androgen receptor. Cancer Res 58: 4640–4645. Hu YJ, Korotkov KV, Mehta R, et al. (2001) Distribution and functional consequences of nucleotide polymorphisms in the 3’-untranslated region of the human Sep15 gene. Cancer Res 61(5): 2307–2310. Hu H, Jiang C, Li G, et al. (2005) PKB/Akt and Erk regulation of caspase-mediated apoptosis by methylseleninic acid in LNCaP prostate cancer cells. Carcinogenesis 26(8): 1374–81. Hu H, Jiang C, Schuster T, et al. (2006) Inorganic selenium sensitizes prostate cancer cells to TRAIL-induced apoptosis through superoxide/p53/Bax-mediated activation of the mitochondrial pathway. Mol. Cancer Ther 5(7): 1873–1882. Husbeck B, Peehl DM, Knox SJ (2005) Redox modulation of human prostate carcinoma cells by selenite increases radiation-induced cell killing. Free Radic Biol Med 38(1): 50–57. Husbeck B, Bhattacharya RS, Feldman D, et al. (2006a) Inhibition of androgen receptor signaling by selenite and methylseleninic acid in prostate cancer cells: two distinct mechanisms of action. Mol Cancer Ther 5(8): 2078–2085. Husbeck B, Nonn L, Peehl DM, et al. (2006b) Tumor-selective killing by selenite in patientmatched pairs of normal and malignant prostate cells. Prostate 66(2): 218–225. Ip C (1998) Lessons from basic research in selenium and cancer prevention. J Nutr 128: 1845– 1854.

776

N. Nadiminty, A.C. Gao

Ip C, Thompson HJ, Zhu Z, et al. (2000) In vitro and in vivo studies of methylseleninic acid: evidence that a monomethylated selenium metabolite is critical for cancer chemoprevention. Cancer Res 60(11): 2882–2886. Javidan J, Deitch A, Shi XB, et al. (2005) The androgen receptor and mechanisms for androgen independence in prostate cancer. Cancer Invest 23(6): 520–528. Jiang C, Jiang W, Ip C, et al. (1999) Selenium-induced inhibition of angiogenesis in mammary cancer at chemopreventive levels of intake. Mol Carcinog 26(4): 213–225. Jiang C, Ganther H, Lu J (2000a) Monomethylselenium – specific inhibition of MMP-2 and VEGF expression: implications for angiogenic switch regulation. Mol Carcinog 29(4): 236–250. Jiang C, Wang Z, Ganther H, et al. (2000b) Caspases as key regulators of methyl selenium-induced apoptosis (anoikis) of DU 145 prostate cancer cells. Cancer Res 61(7): 3062–3070. Jin RJ, Wang Y, Masumori N, et al. (2004) NE-10 neuroendocrine cancer promotes the LNCaP xenografts growth in castrated mice. Cancer Res 64: 5489–5495. Kaarbo M, Klokk TI, Saatcioglu F (2007) Androgen signaling and its interactions with other signaling pathways in prostate cancer. BioEssays 29: 1227–1238. Karin M, Cao Y, Greten FR, et al. (2002) NF-kappaB in cancer: from innocent by-stander to major culprit. Nature Rev Cancer 2: 301–310. Karunasinghe N, Ryan J, Tuckey J, et al. (2004) DNA stability and serum selenium levels in a high risk group for prostate cancer. Cancer Epidemiol Biomarkers Prev 13(3): 391–397. Kawabe T (2004) G2 checkpoint abrogators as anticancer drugs. Mol Cancer Ther 3(4): 513–519. Klein EA (2004) Selenium and Vitamin E cancer prevention trial. Ann NY Acad Sci 1031: 234–41. Klein EA, Thompson IM (2004) Update on chemoprevention of prostate cancer. Curr Opin Urol 14: 143–149. Knekt P, Aromaa A, Maatela J, et al. (1990) Serum selenium and subsequent risk of cancer among Finnish men and women. J Natl Cancer Inst 82: 864–868. Konishi N, Shimada K, Ishida E, et al. (2005) Molecular pathology of cancer. Pathol Int 55: 531–539. Korotkov KV, Kumaraswamy E, Zhou Y, et al. (2001) Association between the 15-kDa selenoprotein and UDP-glucose:glycoprotein glucosyltransferase in the endoplasmic reticulum of mammalian cells. J Biol Chem 276(18): 15330–15336. Kranse R, Dagnelie PC, van Kemenade M, et al. (2005) Dietary intervention in prostate cancer patients: PSA response in a randomized double-blind placebo-controlled study. Int J Cancer 113(5): 835–840. Kryukov GV, Castellano S, Novoselov SV, et al. (2003) Characterization of mammalian selenoproteomes. Science 300: 1439–1443. Kumaraswamy E, Malykh A, Korotkov KV, et al. (2000) Structure-expression relationships of the 15 kDa selenoprotein gene. Possible role of the protein in cancer etiology. J Biol Chem 275(45): 35540–35547. Lee SO, Chun JY, Nadiminty N, et al. (2006) Monomethylated selenium inhibits growth of LNCaP human prostate cancer xenografts accompanied by a decrease in the expression of androgen receptor and prostate specific antigen (PSA). Prostate 66(10): 1070–1075. Li H, Stampfer MJ, Giovanucci EL, et al. (2004) A prospective study of plasma selenium levels and prostate cancer risk. J Natl Cancer Inst 96: 696–703. Li G-X, Hu H, Jiang C, et al. (2007) Differential involvement of reactive oxygen species in apoptosis induced by two classes of selenium compounds in human prostate cancer cells. Int J Cancer 120: 2034–2043. Liao Y, Abel U, Grobholz R, et al. (2005) Up-regulation of insulin-like growth factor axis components in human primary prostate cancer correlates with tumor grade. Human Pathology 36: 1186–1196. Lipsky K, Zigeuner R, Zischka M, et al. (2004) Selenium levels of patients with newly diagnosed prostate cancer compared with control group. Urology 63(5): 912–916. Lorenzo GD, Bianco R, Tortora G, et al. (2003) Involvement of growth factor receptors of the epidermal growth factor receptor family in prostate cancer development and progression to androgen independence. Clin Prostate Cancer 2: 50–57.

Selenium and Androgen Receptor in Prostate Cancer

777

Marshall JR (2001) Larry Clark’s legacy: randomized controlled, selenium-based prostate cancer chemoprevention trials. Nutr Cancer 40(1): 74–77. Marshall JR, Sakr W, Wood D, et al. (2006) Design and progress of a trial of selenium to prevent prostate cancer among men with high-grade prostatic intraepithelial neoplasia. Cancer Epidemiol Biomarkers Prev 15(8): 1479–1484. McLeod DG, Srivastava S (2006) Androgen receptor mutation (T877A) promotes prostate cancer cell growth and cell survival. Oncogene 25(28): 3905–3913. Mendelsohn J, Baselga J (2000) The EGF receptor family as targets for cancer therapy. Oncogene 19: 6550–6565. Menter DG, Sabichi AL, Lippman SM (2000) Selenium effects on prostate cell growth. Cancer Epidemiol Biomarkers Prev 9(11): 1171–1182. Meuillet E, Stratton S, Cherukuri DP, et al. (2004) Chemoprevention of prostate cancer with selenium: An update on current clinical trials and preclinical findings. J Cell Biochem 91: 443– 458. Mohler JL, Gregory CW, Ford OHIII, et al. (2004) The androgen axis in recurrent prostate cancer. Clin Cancer Res 10(2): 440–448. Montgomery B, Mostaghel E, Vessella R, et al. (2007) Androgen synthesis in castration-adapted metastatic prostate cancer. ASCO Meeting 2007: Orlando, Florida, USA, 22–24 February (Abstract No. 98). Morris JD, Pramanik R, Zhang X, et al. (2006) Selenium- or quercetin-induced retardation of DNA synthesis in primary prostate cells occurs in the presence of a concomitant reduction in androgen receptor activity. Cancer Lett 239(1): 111–122. Moscow JA, Schmidt L, Ingram DT, et al. (1994) Loss of heterozygosity of the human cytosolic glutathione peroxidase I gene in lung cancer. Carcinogenesis 15(12): 2769–2773. Navarro-Silvera SA, Rohan TE (2007) Trace elements and cancer risk: a review of the epidemiologic evidence. Cancer Causes Control 18: 7–27. Nieto M, Finn S, Loda M, et al. (2007) Prostate cancer: Re-focusing on androgen receptor signaling. The Intl J of Biochem Cell Biol 39: 1562–1568. Ni J, Chen M, Zhang Y, et al. (2003) Vitamin E succinate inhibits human prostate cancer cell growth via modulating cell cycle regulatory machinery. Biochem Biophys Res Commun 300 (2): 357–363. Nomura AM, Lee J, Stemmermann GN, et al. (2000) Serum selenium and subsequent risk of prostate cancer. Cancer Epidemiol Biomarkers Prev 9: 883–887. Novoselov SV, Kryukov GV, Xu XM, et al. (2007) Selenoprotein H is a nuclear thioredoxin-like protein with a unique expression pattern. J Biol Chem 282(16): 11960–11968. Nyman DW, Stratton SM, Kopplin MJ, et al. (2004) Selenium and selenomethionine levels in prostate cancer patients. Cancer Detect Prev 28(1): 8–16. Ohta Y, Suzuki KT (2008) Methylation and demethylation of intermediates selenide and methylselenol in the metabolism of selenium. Toxicol. Appl. Pharmacol. 226(2): 169–177. Pak RW, Lanteri VJ, Scheuch JR, et al. (2002) Review of vitamin E and Selenium in the prevention of prostate cancer: implications of the selenium and vitamin E chemoprevention trial. Integr Cancer Ther 1(4): 338–344. Patrick L (2004) Selenium biochemistry and cancer: A review of the literature. Altern Med Rev 9 (3): 239–258. Pienta KJ, Bradley D (2006) Mechanisms underlying the development of androgen-independent prostate cancer. Clin Cancer Res 12(6): 1665–1671. Pinto JT, Sinha R, Papp K, et al. (2007) Differential effects of naturally occurring and synthetic organoselenium compounds on biomarkers in androgen responsive and androgen independent human prostate carcinoma cells. Int J Cancer 120: 1410–1417. Ramachandran K, Navarro L, Gordian E, et al. (2007) Methylation-mediated silencing of genes is not altered by selenium treatment of prostate cancer cells. AntiCancer Res 27(2): 921–925. Rayman MP (2000) The importance of selenium to human health. Lancet 356: 233–241.

778

N. Nadiminty, A.C. Gao

Ravn-Haren G, Bugel S, Krath BN, et al. (2007) A short-term intervention trial with selenate, selenium-enriched yeast and selenium-enriched milk: effects on oxidative defence regulation. Br J Nutr Sep 21, 1–10. Rayman MP (2005) Selenium in cancer prevention: a review of the evidence and mechanism of action. Proc Nutr Soc 64: 527–542. Rebsch CM, Penna FJ, Copeland PR (2006) Selenoprotein expression is regulated at multiple levels in prostate cells. Cell Res 16(12): 940–948. Reddy GP, Barrack ER, Dou QP, et al. (2006) Regulatory processes affecting androgen receptor expression, stability and function: potential targets to treat hormone-refractory prostate cancer. J Cell Biochem 98: 1408–1423. Reid ME, Stratton MS, Lillico A, et al. (2004) A report of high dose selenium supplementation: Response and toxicities. J Trace Elem Biol Med 18: 69–74. Ryan-Harshman M, Aldoori W (2005) The relevance of selenium to immunity, cancer, and infectious/inflammatory diseases. Can J Diet Pract Res 66(2): 98–102. Sabichi AL, Lee JJ, Taylor RJ, et al. (2006) Selenium accumulation in prostate tissue during a randomized, controlled short-term trial of L-selenomethionine: a Southwest Oncology Group study. Clin Cancer Res 12(7): 2178–2184. Scher HI, Sawyers CL (2005) Biology of progressive, castration-resistant prostate cancer: directed therapies targeting androgen receptor signaling axis. J Clin Oncol 23(32): 8253–8261. Schlicht M, Matysiak B, Brodzeller T, et al. (2004) Cross-species global and subset gene expression profiling identifies genes involved in prostate cancer response to selenium. BMC Genomics 5: 58–69. Schroder FH (2008) Progress in understanding androgen-independent prostate cancer (AIPC): A review of potential endocrine-mediated mechanisms. Eur Urol doi:10.1016/j.eururo. 2008.01.049. Schweizer L, Rizzo CA, Spires TE, et al. (2008) The androgen receptor can signal through Wnt/ b-catenin in prostate cancer cells as an adaptation mechanism to castration levels of androgens. BMC Cell Biol 9:4. Schrauzer GN, White DA, Schneider CJ (1977a) Cancer mortality correlation studies-III: statistical associations with dietary selenium intakes. Bioinorganic Chem 7(1): 23–31. Schrauzer GN, White DA, Schneider CJ (1977b) Cancer mortality correlation studies-IV: associations with dietary intakes and blood levels of certain trace elements, notably Se-antagonists. Bioinorganic Chem 7(1): 35–56. Shamberger RJ, Frost DV (1969) Possible protective effect of selenium against human cancer. Can Med Assoc J 100(14): 682–692. Singh RP, Agarwal R (2006) Mechanisms of action of novel agents for prostate cancer chemoprevention. Endocr Related Cancer 13: 751–778. Sommer A, Haendler B (2003) Androgen receptor and prostate cancer: molecular aspects and gene expression profiling. Curr Opin Drug Discov Devel 6(5): 702–711. Spallholz JE (1994) On the nature of selenium toxicity and carcinostatic activity. Free Radic Biol Med 17: 45–64. Squires J, Berry MJ (2006) Selenium, Selenoproteins and cancer. Hawaii Med J 65(8): 239–240. Stanbrough M, Bubley GJ, Ross K, et al. (2006) Increased expression of genes converting adrenal androgens to testosterone in androgen-independent prostate cancer. Cancer Res 66: 2815–2825. Stratton MS, Reid ME, Schwartzberg G, et al. (2003a) Selenium and prevention of prostate cancer in high-risk men: the Negative Biopsy study. Anticancer Drugs 14(8): 589–594. Stratton MS, Reid ME, Schwartzberg G, et al. (2003b) Selenium and inhibition of disease progression in men diagnosed with prostate carcinoma: study design and baseline characteristics of the ‘‘Watchful Waiting’’ study. Anticancer Drugs 14(8): 595–600. Suh J, Rabson AB (2004) NF-kappaB activation in human prostate cancer: important mediator or epiphenomenon? J Cell Biochem 91: 100–117.

Selenium and Androgen Receptor in Prostate Cancer

779

Syed DN, Khan N, Afaq F, et al. (2007) Chemoprevention of prostate cancer through dietary agents: Progress and promise. Cancer Epidemiol Biomarkers Prev 16(11): 2193–2203. Syed S, Tolcher A (2003) Innovative therapies for prostate cancer treatment. Rev Urol 5 (suppl 3): S78–84. Takaha N, Resar LM, Vindivich D, et al. (2004) High mobility group protein HMGI(Y) enhances tumor cell growth, invasion and matrix metalloproteinase-2 expression in prostate cancer cells. Prostate 60: 160–167. Taplin M-E (2007) Drug insight: role of the androgen receptor in the development and progression of prostate cancer. Nature Clin Pract Oncol 4(4): 236–244. Taplin ME, Rajeshkumar B, Halabi S, et al. (2003) Androgen receptor mutations in androgenindependent prostate cancer: Cancer and leukemia group study 1993. J Clin Oncol 21: 2673–2678. Terry S, Yang X, Chen MW, et al. (2006) Multifaceted interaction between the androgen receptor and Wnt signaling pathways and the implication for prostate cancer. J Cell Biochem 99: 402–410. Thompson IM (2007) Chemoprevention of prostate cancer: Agents and study designs. J Urol 178 (3): S9–S13. Thomson CD, Robinson MF, Butler JA, et al. (1993) Long term supplementation with selenate and selenomethionine: selenium and glutathione peroxidase (EC 1.11.1.9) in blood components of New Zealand women. Br J Nutr 69: 577–588. Titus MA, Schell MJ, Lih FB, et al. (2005) Testosterone and dihydrotestosterone tissue levels in recurrent prostate cancer. Clin Cancer Res 11(13): 4653–4657. Ueda T, Bruchovsky N, Sadar MD (2002) Activation of the androgen receptor N-terminal domain by interleukin-6 via MAPK and Stat3 signal transduction pathways. J Biol Chem 277: 7076–7085. Unni E, Sun S, Nan B, et al. (2004) Changes in androgen receptor nongenotropic signaling correlate with transition of LNCaP cells to androgen independence. Cancer Res 64: 7156–7168. Uzgare AR, Kapaln PJ, Greenberg NM (2003) Differential expression and/or activation of p38MAPK, erk1/2, and jnk in the initiation and progression of prostate cancer. Prostate 55: 128–139. Vadgama JV, Wu Y, Shen D, et al. (2000) Effect of selenium in combination with adriamycin or taxol on several different cancer cells. Anticancer Res 20(3A): 1391–1414. van den Brandt PA, Zeegers MP, Bode P, et al. (2003) Toenail selenium levels and the subsequent risk of prostate cancer: a prospective cohort study. Cancer Epidemiol Biomarkers Prev 12: 866–871. Vander Griend DJ, Litvinov IV, Isaacs JT (2007) Stabilizing androgen receptor in mitosis inhibits prostate cancer proliferation. Cell Cycle 6(6): 647–651. Venkateswaran V, Fleshner NE, Sugar LM, et al. (2004) Antioxidants block prostate cancer in Lady transgenic mice. Cancer Res 64(16): 5891–5896. Vogt TM, Ziegler RG, Patterson BH, et al. (2007) Racial differences in serum selenium concentration: Analysis of US population data from the Third National Health and Nutrition Examination survey. Am J Epidemiol 166(3): 280–288. Wang Z, Jiang C, Lu J (2002) Induction of caspase-mediated apoptosis and cell-cycle G1 arrest by selenium metabolite methylselenol. Mol Carcinog 34(3): 113–120. Wang Z, Hu H, Li G, et al. (2008) Methylseleninic acid inhibits microvascular endothelial G(1) cell cycle progression and decreases tumor microvessel density. Int J Cancer 122: 15–24. Waters DJ, Shen S, Cooley DM, et al. (2003) Effects of dietary selenium supplementation on DNA damage and apoptosis in canine prostate. J Natl Cancer Inst 95(3): 237–241. Whanger PD (2002) Selenocompounds in plants and animals their biological significance. J Am Coll Nutr 21: 223–232. Whanger P, Vendeland S, Park YC, et al. (1996) Metabolism of subtoxic levels of selenium in animals and humans. Ann Clin Lab Sci 26: 99–113.

780

N. Nadiminty, A.C. Gao

World Health Organization. (1987) Selenium, Geneva World Health Organization, Environmental Health Criteria 58. Wu Y, Zhang H, Dong Y, et al. (2005) Endoplasmic reticulum stress signal mediators are targets of selenium action. Cancer Res 65(19): 9073–9079. Xia Y, Hill KE, Byrne DW, et al. (2005) Effectiveness of selenium supplements in a low selenium area of China. Am J Clin Nutr 81: 829–834. Yang SZ and Abdulkadir SA (2003) Early growth response gene 1 modulates androgen receptor signaling in prostate carcinoma cells. J Biol Chem 278(41): 39906–39911. Yang L, Wang L, Lin HK, et al. (2003) Interleukin-6 differentially regulates androgen receptor transactivation via PI3K-Akt, Stat3 and MAPK, three distinct signal pathways in prostate cancer cells. Biochem Biophys Res Commun 305: 462–465. Yang X, Chen MW, Terry S, et al. (2006) Complex regulation of human androgen receptor expression by Wnt signaling in prostate cancer cells. Oncogene 25: 3436–3444. Yoshizawa K, Willett WC, Morris SJ, et al. (1998) Study of prediagnostic selenium levels in toenails and the risk of prostate cancer. J Natl Cancer Inst 90:1219–1224. Yu SY, Chu YJ, Li WG (1988) Selenium chemoprevention of liver cancer in animals and possible human applications. Biol Trace Elem Res 15: 231–241. Yu YP, Yu G, Tseng G, et al. (2007) Glutathione peroxidase 3, deleted or methylated in prostate cancer, suppresses prostate cancer growth and metastasis. Cancer Res 67(17): 8043–8050. Zachara BA, Szewczyk-Golec K, Tyloch J, et al. (2005a) Blood and tissue selenium concentrations and glutathione peroxidase activities in patients with prostate cancer and benign prostate hyperplasia. Neoplasma 52(3): 248–254. Zachara BA, Szewczyk-Golec K, Wolski Z, et al. (2005b) Selenium level in benign and cancerous prostate. Biol Trace Elem Res 103(3): 199–206. Zhao H and Brooks JD (2007) Selenomethionine induced transcriptional programs in human prostate cancer cells. J Urol 177: 743–750. Zhao H, Whitfield ML, Xu T, et al. (2004) Diverse effects of methylseleninic acid on the transcriptional program of human prostate cancer cells. Mol Biol Cell 15(2): 506–519. Zhao R, Domann FE, Zhong W (2006) Apoptosis induced by selenomethionine and methioninase is superoxide mediated and p53 dependent in human prostate cancer cells. Mol Cancer Ther 5 (1): 3275–3284. Zu K, Ip C (2003) Synergy between selenium and vitamin E in apoptosis induction is associated with activation of distinctive initiator caspases in human prostate cancer cells. Cancer Res 63 (20): 6988–6995. Zu K, Bihani T, Lin A, et al. (2006) Enhanced selenium effect on growth arrest by BiP/GRP78 knockdown in p53-null human prostate cancer cells. Oncogene 25(4): 546–554.

Index

A Ablation-resistant prostate cancer AR-mediated transcriptional control, 416, 418–419 cell models, 415 H3K4 methylation and PSA expression, 416–417 locus-wide chromatin remodeling, 415–416 unbiased network biology approach, 415 Activated cdc42-associated tyrosine kinase (Ack1), 331 A disintegrin and metallopeptidase 17 (ADAM17), 531 Agoulnik, I.U., 315 A-kinase anchoring proteins (AKAPs), 468–469 Albertelli, M.A., 279, 281, 283, 284, 286, 287 Aldo-keto reductase family 1, member C3 (AKR1C3), 127 17-Allylamino-17-demethoxygeldanamycin (17-AAG), 150, 216 Amino-terminal domain. See also NH2terminal domain (NTD) 43–90 collocation, 223–224 221–269 collocation, 225–227 431–455 clustering, 227–228 497–537 collocation, 230 ligand-independent activation, 229 signature sequence, 225–227 structure and activation functions, 220–222 Amler, L.C., 640 Anderson, K.M., 17

Androgen ablative therapy. See also Androgen-deprivation therapy (ADT); Neuroendocrine differentiation (NED) androgen withdrawal-induced signaling androgen receptor signaling pathway, 528–529 apoptosis-regulatory pathways, 532–534 Bcl-2-mediated survival, 524–525 compensatory survival, 524 forkhead factors, 525–526 growth arrest and tumor regression, 523 neuroendocrine differentiation, 529–532 receptor turnover, 517–521 target of rapamycin (mTOR), 526–527 transcription, 516–517 transcriptional and post-transcriptional regulation, 521–523 clinical studies adrenal androgen synthesis suppression, 510–511 combined androgen blockade (CAB), 509–510 LHRH administration, 509 long-term ADT, 510 experimental models, in vivo apoptotic response, 514 castration-induced genes, 512–513 stromal factors, 513 vascular regression, 513–514 withdrawal-induced apoptosis, 511–512 neuroendocrine differentiation, 514–516

781

782

Androgen–androgen receptor (AR) complexes hydrophobic cavity binding, 24–25 hydroxylapatite filtration assay, 23–24 identification and characterization, 22 insolubilized antisteroid antibodies, 23 sedimentation analysis, 22–23 structural requirements, 24 Androgen-deprivation therapy (ADT). See also Intermittent androgendeprivation therapy (IADT) clinical predictors, 59–62 receptor turnover caspases and calpain, 520–521 destabilizing effect, 517 heat-shock protein 90 (Hsp90), 518 ubiquitin-proteasome pathway (UPP), 517–518 steroid 5a-reductase (5aR), 157–158 transcription, 516–517 transcriptional and post-transcriptional regulation down-regulation, 522–523 mRNA stability and translation, 522 transcription initiation sites (TIS I/II), 521–522 Androgen insensitivity syndrome (AIS) AF2 germline mutations, 254–255 Kennedy’s disease, 388 pathophysiology of, 386, 388 Androgen-metabolic genes characteristics, 144 CYP3A, 147–148 CYP17, 144–145 DHT, 142–143 HSD, 146–147 implications, 150–151 signaling pathway, AR coactivators, 150 functional inhibition, 150 (CAG)n polymorphism, 149 therapy-mediated selection pressure concept, 149–150 SRD5A2, 145–146 SULT, 148–149 UGT, 148 Androgen receptor (AR) ablation-resistant PCa

Index

AR-mediated transcriptional control, 418–419 cell models, 415 H3K4 methylation and PSA expression, 416–417 locus-wide chromatin remodeling, 415–416 unbiased network biology approach, 415 adaptive phenotypic changes mutations, 81–82 prostate intraepithelial neoplasia (PIN), 82 somatic mutation, 81–82 amplification and overexpression, 745–747 androgen-induced transformation, 484 anti-AR antibodies autoantibodies, 30–31 monoclonal antibodies, 31 ATP-mediated remodeling, 413–414 cAMP pathway adenylyl cyclases, 466–468 analogues and AC activators, 479 androgens, 479–480 CREB and TORC, 475–476 neuropeptides, 476–479 phosphodiesterases, 473–475 PKA, 468–473 protein exchange activation, 475 cAMP/PKA pathway acetylation, 487 coregulators, 489–490 ligand-independent activation, 484–485 nuclear translocation, 488–489 phosphorylation, 486–487 sumoylation, 488 transactivation, 466, 490 castration-recurrent prostate cancer androgen deprivation therapy (ADT), 554, 555 AR mutation, 431–432 AR-targeting approach, 430–431 DHT and EGF, 556 growth regulation, 554 ligand-independent AR activation, 433–435

Index

mRNA overexpression, 432 siRNA development, 431 transactivation, 556 castration resistance, 756–758 cDNA cloning, 26–27 clinical implication, 82–83 clinical trials and epidemiology DNA damage, 759 risk factors, 760 SELECT, 760 Se supplementation, 759 discoveries, time line, 10–11 DNA replication licensing antiandrogens, 624 mitosis, 625 origin of replication complex (ORC), 621–622 regulation, 622–623 therapeutic implications, 626–627 tumorigenesis, 623–624 downregulation, 749, 750 epithelial cells growth and differentiation, 74–75 Foxa proteins family member expression, 607–610 gene expression modulation, 600–607 molecular biology, 591, 593 murine and human prostate, 593–594 prostate ductal morphogenesis and maturation, 594, 596–599 functional motifs FXXLF and WXXLF motifs, 246–250 gene mutations, 253–258 interaction effects, 250–253 MAGE-11coregulator, 258–261 N/C interaction, 242–246 T and DHT role, 252–253 transactivation regulation, 250–252 gene mutations amino-terminal domain, 220–230 DNA-binding domain, 218–219 ligand-binding domain (LBD), 210–218 prostate cancer, 208–210 gene regulation, 485–486 Gleason grade, 3 histone modifications ChIP-on-chip, 407–408

783

ChIP sequencing, 408 coregulators, 408 H3 and 4 amino-terminal tail, 408–409 histone-modifying enzymes, 412–413 human genome 5a-dihydrotestosterone (DHT), 667 ChIP-on-chip, 665, 666 false discovery rate (FDR), 665 transcription start sites (TSS), 666 interleukin-6 (IL-6) cAMP and bombesin, 452 cofactors, 458 endogenous receptor expression, 455 Etk and Pim1, 458 nonsteroidal activation, 455 prodifferentiation and proliferation, 455, 457 prostate cancer, 453 STAT3, 457 intracellular transformation and recycling, 25–26 luminal cells, 76 metastases molecular mechanisms, 80–81 modular structure DNA-binding domain (DBD), 392–394 hinge region, 395–397 ligand-binding domain (LBD), 394–395 NH2-terminal domain (NTD), 388–391 molecular mechanism, 386 multilevel signal repression receptor turnover, 517–521 transcription, 516–517 transcriptional and post-transcriptional regulation, 521–523 natural mutation AIS mutation, 386–388 Kennedy’s disease, 388 prostate cancer, 388 nomenclature, 3–4 noncanonical AREs, 669–670 normal prostate physiology, 428 plasma glutathione peroxidase expression, 757 polyglutamine (polyQ) tract CAG repeats, 276–277 epidemiology, 274–275 functional analysis, 272–274

784

humanized development, 277–280 nuclear receptor superfamily, 270–272 transgenic adenocarcinoma of the mouse prostate (TRAMP), 280–288 prostate cancer, 428–429 budding morphogenesis, 590 deprivation therapy, 3 epithelium-associated gene expression networks, 590–591 progression and metastasis, 83 urogenital sinus (UGS), 589 xenografts, 2, 35–36 regulation and deregulation COOH-terminal domain, 440–441 NH2-terminal domain (NTD), 441–443 Se anticancer effect, mechanisms, 763 metabolites, 761–763 selenocysteine (SeCys), 757 selenomethionine, 758 selenoproteins, 764 signaling pathways and cellular processes androgen-regulated genes, 768 angiogenesis, 771–772 cell cycle, 769 cell proliferation, 765 chemotherapeutic agents, 768 coregulators, 765 epidermal growth factor receptor (EGFR), 770 gene expression studies, 767 hypersensitivity, 766 ligand-binding domain, 765 MSeA, 762–763 NFk-B transcription factor, 770 nongenotropic signaling, 766 prostate-specific antigen (PSA), 757–758 reactive oxygen species generation, 770–771 structure and mutations, 27–28 structure/function COOH-terminal domain, 436–437 DBD/hinge region, 437–438 modular domains and motifs, 435–436 NH2-terminal domain, 438–439 target genes and sequences androgen-response elements, 632, 634

Index

cyclic nucleotide phosphodiesterases, 667 cytoplasmic multiprotein chaperone complex, 632 ligation-mediated amplification (LMA), 669 looping model, 668 PSA gene, 667 Src activation, 634 testosterone, 427–428 transcriptional mechanisms, 406–407 transcription factor ETS family, 674 FoxA1, 671 GATA family, 671, 672 Oct1, 672 tumor growth and metastasis, 74 tumor proliferation and suppressor pelvic lymph nodes (PLN), 78 prostate epithelial cell types, 77 WPMY1 cells, 79 tumor suppressor gene (TSG) allele, 207 Androgen response elements (AREs), 96, 392, 686, 691–693 Androgen-responsive genes cell proliferation regulators, 645–646 ChIP-on-chip, 635–636 DNA binding analyses, 649–651 expressed sequence tag (EST), 634 general processes and metabolic pathways endoplasmic reticulum (ER) stress response genes, 643 prostatic fluid and testosterone, 642 tissue-specific/tissue enriched genes, 642 human studies cancer cell lines, 637 LNCaP cells, 639 metastatic cancers, 637 ICAT method, 636 IGF signaling IGFBP-3 expression, 644 Insulin-like growth factor-1 (IGF-I), 643 microarray expression profiling, 635 noncoding RNAs, 647–648 proteome analysis, 651–653 rat and mouse strains, in vivo models, 638

Index

sequence tag-based studies dihydrotestosterone (DHT), 639–641 FKBP5 expression, 639, 641 LNCaP cells, 639 Serial Analysis of Gene Expression (SAGE), 634–635 spatially restricted genes, 648–649 transcription factors CREB3L4, 645 NKX3.1, homeobox gene, 644 two-dimensional differential gel electrophoresis (2-D DIGE), 636 Androgen signaling, gene fusions castration-recurrent growth, 713, 714 FoxA1, GATA2, and Oct1, 714 multifocal prostate cancer and fusion status, 716 TMPRSS2-ETS fusion status and clinical outcomes, 714–716 Ansamycin antibiotics 17-allylamino-17demethoxygeldanamycin (17-AAG), 216 geldanamycin, 215–216 Antagonist dissociation rate, 243 Anti-antiandrogen withdrawal syndrome, 214 Antisense oligonucleotides (ASOs), 218, 689, 692 AR-E236G expression, 225 Arnold, L.A., 307 Asirvatham, A.J., 641 Aumu¨ller, G., 180

B Balk, S.P., 123 Barry, M., 716 Baum, W.C., 67 Bebermeier, J.H., 640 Belanger, B., 557, 560 Benign prostatic hyperplasia (BPH), 157, 759 Bhattacharyya, A.K., 177 Bianco-Miotto, T., 207 Bicalutamide, 134 Bioluminescence imaging (BLI)

785

cell culture studies, TSTA, 99–102 chimeric/composite enhancer approach metastasis, 98–99 plasmid constructs generation, 96–98 PSA promoter and enhancer, 95–96 gene expression-based imaging animal models, 92 drawbacks, 93–94 methodology, 93 reporter gene delivery, 94–95 reporter gene potency, 95 metastasis and cancer progression EZC models, 114–115 sPSA model, 113–114 transgenic mouse models ARR2Pb transgenic, 112–113 GAL4-AR transgenic, 113 TSTA transgenic, 111–112 in xenografts, TSTA AdTSTA modifications, 108–110 animal studies, 102–105 lentiviruses, 106 MAPK combination, 110 PSA enhancer, 105–106 therapy effects, 106–108 Bjorkhem, I., 176 Boger-Megiddo, I., 178 Bohl, C.E., 302 Bolton, E.C., 640, 650, 666 Bouchot, O., 191 5-Bromodeoxyuridine (BrdU), 76 Brown, M, 408 Bruchovsky, N., 17, 37, 176, 180, 191 Brusselmans, K., 723 Buchanan, G., 207 Butenandt, A., 175 Butler, L.M., 207 Buttyan, R., 569

C Calcitonin, 478 Calpain, 520–521 cAMP-responsive element-binding protein 3-like 4 (CREB3L4), 645 Cancer and Leukemia Group B (CALGB), 63–64

786

Cancer outlier profile analysis (COPA), 703–704 Carboxy-terminus of the Hsp70-interacting protein (CHIP), 390 Carcinogen metabolism, 763 Carey, M., 91, 100 Carver, B.S., 161 Case, T., 587 Castration-recurrent prostate cancer (CRPC) androgen production, extratesticular sources abiraterone acetate (CB7630), 748 CYP17, 748–749 dehydroepiandrosterone (DHEA) and androstenedione, 747 steroidogenic enzyme, 749 androgen receptor (AR) amplification and overexpression, 745–747 androgen deprivation therapy (ADT), 554, 555 DHT and EGF, 556 downregulation, 749, 750 growth regulation, 554 mutation, 431–432, 743–745 transactivation, 556 AR-targeting approach, 430–431 cholesterol synthesis regulation, 131 clinical progression androgen deprivation therapy, 57, 59–60 androgen-independent prostate cancer, 58–59 DFCI study, 60 history and predictors of, 62–65 intermittent androgen deprivation (IAD), 61–62 PSA parameters, 60–61 Radiation Therapy Oncology Group (RTOG) trial, 62 serum testosterone levels, 62 enzyme expression 5a-reductase and HSD3B2, 132 AKR1C2, 128 AKR1C3, 127 androgen inactivation, 133 antagonists, 133–134 HSD3B1, 126

Index

HSD17B6, 128–129 type 1 and 2 5a-reductases, 127–128 UGT2B15 and UGT2B17, 129 untreated primary PCA, 129–130 intraprostatic androgen level, 125–126 ligand-independent AR activation, IL-6 MAPK pathway, 443 p300 expression, 443–444 mRNA overexpression, 432 5a-reductase isozymes AR transactivation, 178 express sequence tags (EST), 179 inhibitors activity, 177–178 intracellular location, 176 molecular genetics, 176–177 mRNA levels, 179 testosterone metabolism, 176 siRNA development, 431 tissue androgen levels AR-activating levels, 559 DHT, 556–558 flutamide, 557, 558 potential sources, 559–561 prostate-specific antigen (PSA), 558 radioimmunoassay (RIA), 556 technical barriers, 561–563 transurethral prostatectomy, 557 treatment antiandrogen agent, 66 estrogens, 67 ketoconazole inhibitors, 66–67 secondary hormonal treatment, 65–66 b-Catenin archetypal paracrine process, 573 AR gene expression, 577–578 Frizzled (Fzd) gene, 572–573 Groucho gene, 574 GSK-3b protein kinase, 576–577 intracellular domain, cadherin, 571–572 mutation, 579–580 phosphorylation and ubiquitination, 572 prostate cancer cells, 575–576 S33F, 588–589 stabilization process, 573–574, 580 T-cell factor/lymphoid-enhancing factor-1 (Tcf/LEF-1), 574 Cell lines, 743, 744 Centenera, M.M., 207

Index

Cerulenin, 733 Chang, C., 73, 193, 395, 522 Chang, T., 73 Chen, C.D., 713 Chen, M., 569 Cheng, J., 329 Chinnaiyan, A.M., 701 Chmelar, R., 358 Chokkalingam, A.P., 144, 145, 147 Cho, S.D., 768 Chromatin epigenetics ATP-mediated remodeling, 413–414 histone modifications ChIP-on-chip, 407–408 coregulators, 408 H3 and 4 amino-terminal tail, 408–409 histone-modifying enzymes, 412–413 Chromatin immunoprecipitation (ChIP) assay, 635, 649, 687, 693 Chromatin remodeling coactivators, 347–348 Chromosome 5 open reading frame 13 (C5orf13), 531 Chun, J.Y., 767 Chuu, C.P., 9 Ci, M., 641 Cinar, B., 354 Civantos, F., 161 Claessens, F., 385 Clark, L.C., 760, 761 Clegg, N., 631, 640 Clusterin isoforms androgen antisense oligonucleotides (ASOs), 692 heat-shock protein 27 (Hsp27), 694 receptor and gene regulation, 691–692 Shionogi tumour model, 692–693 animal models, 690 biological functions Alzheimer’s disease, 687 cancer, 689 heat-shock factor (HSF), 688 nuclear CLU, 688–689 secreted CLU, 688 definition, 681 genomic structure exon size, 684–685 mRNA transcription, 684–685

787

nuclear localisation sequences (NLS), 686 splice variants, 685–686 megalin, 690–691 pro-and anti-apoptotic function, 682 promoter analysis, 686–687 protein structure a and b chains, 683 crystal structure, 682–683 exon 1 sequence, 683 transient transfection effects, 684 Cochaperones, 214–215 Coetzee, G.A., 405 Collapsin response mediator protein hCRMP-2, 531–532 COOH-terminal domain (CTD) regulation and deregulation, 440–441 structure/function, 436–437 Corcoran, N.M., 763, 771 Cordon-bleu homolog (COBL), 531 Coregulators BRCA1 (COBRCA1) cofactor, 330 demethylases JHDM2A, 327 JMJD2C, 327–328 LSD1 and JARID1B, 327 deubiquitinases, 328 Gleason score, 334 histone acetyltransferases (HAT) acetyltransferases, 322–324 deacetylases, 324–325 p160 coactivators, 319–322 methyltransferases, 325–326 modes of action, 316, 317 nongenomic and rapid signaling events, 353–355 other coactivators, 343, 345 prostate cancer coactivators association, 356 corepressors association, 357–358 expression profile, 358–360 therapeutic targeting, 361–363 in vivo functional studies, 360–361 prostate-specific antigen (PSA), 320–325 proteins, transcriptional activity, 342, 343 PTB-associated splicing factor (PSF) and p54nrb, 330 signaling molecules

788

activated cdc42-associated tyrosine kinase (Ack1), 331 Cdc25B, 331–332 cyclin D1, 332 cyclin D3/CDK11, 332–333 isomerase and wnt/ß-catenin signaling, 333 male germ cell-associated kinase (MAK), 331 PELP-1 and MNAR, 333 phosphatases and kinases, 331 protein-kinase-C-related-kinase-1 (PRK1), 332 small ubiquitin-like modifier proteins (SUMO), 329 steroid receptor coactivator (SRC) family, 346–347 type I classical and type II nonclassical corepressors histone deacetylase (HDAC) recruitment, 352 N-/C-terminal interaction and coactivator binding, 352–353 nuclear translocation and DNA-binding, 352 type I classical coactivators basal transcriptional machinery, 348 chromatin remodeling coactivators, 347–348 steroid receptor coactivator (SRC)/p160, 343, 346 type I/II AR coactivators, 343, 344 type I/II AR corepressors, 343, 346 type II nonclassical coactivators cellular trafficking, 350–351 ligand binding and receptor stability, 349–350 molecular chaperones, 348–349 ubiquitin ligases, 328 Cote, R.J., 159, 161 Coutinho-Camillo, C.M., 640 Culig, Z., 144, 451 Cunha, G.R., 75, 588 Cussenot, O., 178 Cyclic AMP pathway adenylyl cyclases soluble, 468 transmembrane, 467–468

Index

analogues and AC activators, 479 androgens forskolin, 481–483 prostate cancer, 479–480 prostate-specific antigen, 480–481 cAMP-dependent protein kinase (cAMP/PKA) pathway acetylation, 487 androgen-induced transformation, 484 coregulators, 489–490 ligand-independent activation, 484–485 nuclear translocation, 488–489 phosphorylation, 486–487 sumoylation, 488 transactivation, 466, 490 CREB and TORC activation, 475–476 neuropeptides, 478–479 phosphodiesterases isoforms, 473–474 targeting therapies, 474–475 protein exchange activation, 475 protein kinase A (PKA) A-kinase anchoring proteins (AKAPs), 468–469 extracellular PKA, 471–472 isoforms, 469–471 ribosomal S6 kinase and MAPK/ERK, 472–473 targeting therapies, 472 Cyclin-dependent kinase (CDK) activity, 622 Cytochrome P450 3A (CYP3A) genes, 147–148

D D’Amico, A.V., 63 Dana–Farber Cancer Institute (DFCI), 60–62 Day, T.K., 681 DBD. See DNA-binding domain DeGraff, D.J., 587 Dehm, S.M., 427 Demethylases JHDM2A, 327 JMJD2C, 327–328

Index

LSD1 and JARID1B, 327 DePrimo, S.E., 640, 642 Dervan, P.B., 640 Desai, K.V., 641, 648 Deslypere, J.P., 192 Deubiquitinases, 328 Dhanasekaran, S.M., 703 Dickkopf-1 (Dkk-1) b-catenin, 573 human prostate cancer, 581 Diethylstilbestrol (DES), 57, 67 Dihydropyrimidinase-like 2 (DPYSL2), 531–532 5a-Dihydrotestosterone (5a-DHT), 142–143 DNA-binding domain (DBD) 574–586 collocation, 218–219 androgen response elements (ARE), 392 nuclear export activity, 394 response elements, 219 structure, 392–394 DNA decoys, 219 DNA replication licensing factors antiandrogens, 624 cyclin-dependent kinase (CDK) activity, 622 genetic stability and cellular viability, 621 minichromosome maintenance (MCM) protein complex, 622 mitosis, 625 origin replication complex (ORC), 621 prostate stromal cells (PRSCs), 625 regulation, 622–623 therapeutic implications, 626–627 tumorigenesis, 623–624 Dong, Y., 767 Drobnjak, M., 357 Dutasteride, 162–163, 198 ARI40010, preradical prostatectomy study intraprostatic DHT, 166–167 time-dependent effect, 167–168 ARIA2003, preradical prostatectomy study clinical complications, 166 histopathological features, 165 serum DHT, 163 TUNEL and tTG staining, 164

789

E EAF2/U19 tumor suppressor, 195 Eder, I.E., 640 Eggener, S.E., 198 Emami, K.H., 100 Endoplasmic reticulum (ER), 683–685 Enzymes expression, CRPC normal prostate AKR1C2, 128 AKR1C3, 127 HSD3B1, 126 HSD17B6, 128–129 type 1 and 2 5a-reductases, 127–128 UGT2B15 and UGT2B17, 129 untreated primary PCA AKR1C2 and AKR1C3, 129–130 SRD5A1 and SRD5A2, 130 Epidermal growth factor (EGF), 556 Este´banez-Perpin˜a´, E., 297, 302, 303, 361 Estrogen-related receptors (ERRs), 489 Ets variant gene 1 (ETV1), 645 Evans, E., 10

F Fang, S., 176 Fatty acid synthase (FASN) antineoplastic intervention, 733–734 control layer, 729 enzyme expression, 729–730 lipid biosynthesis, 725, 726 lipogenesis, 732–733 mechanism, cancer, 730–731 SREBP–SCAP, 728 Faus, H., 397 Febbo, P.G., 640 Festuccia, C., 159 Finasteride, 157, 161–162, 195–197 Fletterick, R.J., 297 Fluorescence two-dimensional differential gel electrophoresis (2-D DIGE), 636 Flutamide, 150 FoxA proteins family member expression, 607–610 neuroendocrine liver metastasis, 609 pathology and immunostaining, 610 Simian Virus 40 (SV40), 607

790

synaptophysin, 607, 608 gene expression modulation, 600–607 chromatin immunoprecipitation (ChIP), 600 human PSA core enhancer, 601–603 luciferase activity, 606 murine epididymal retinoic-acidbinding protein (mE-RABP), 603 rat probasin promoter, 601 molecular biology liver-enriched transcription factors, 591 transactivation activity, 593 murine and human prostate budding morphogenesis, 594 FoXa1 and FoXa2 distribution, 592, 593 immunohistochemical analysis, 593 western blotting, 593, 594 prostate ductal morphogenesis and maturation E-cadherin staining, 595 Escherichia coli lacZ gene, 594 hematoxylin and eosin staining, 597 smooth muscle actin (SMA) gamma-positive layer, 596 toluidine blue staining, 597 Frenkel, B., 405 Frizzled (Fzd) gene, 572–573 Frost, D.V., 758 Functional motifs FXXLF and WXXLF motifs binding preferences, 246–249 charge distribution flanking, 246–247 coactivators, 248–249 MAGE-11 coregulator, 249 schematic representation, 243 sequence conservation, 249–250 gene mutations AF2 germline, 254–255 phenotypic expression, 253–254 somatic type, 255–258 MAGE-11 coregulator biological functions, 258–260 expression pattern, 260–261 N/C interaction dissociation kinetics, 242–243 ligand and DNA binding, 244 stabilization, 244–246

Index

T and DHT effects functional differences, 253 potency differences, 252 transactivation regulation MMTV and PSA, 250–251 SRC/p160 coactivators, 251–252 Fung, K.M., 144

G Gao, A.C., 755 Gastrin-releasing peptide/bombesin receptors, 478 Geldanamycin. See Ansamycin antibiotics Geller, J., 556, 557, 558 Gene expression-based BLI animal models, 92 drawbacks of, 93–94 methodology, 93 reporter gene delivery, 94–95 reporter gene potency, 95 Gene fusions androgen signaling castration-recurrent growth, 713, 714 FoxA1, GATA2, and Oct1, 714 multifocal prostate cancer and fusion status, 716 TMPRSS2-ETS fusion status and clinical outcomes, 714–716 chimeric fusion protein, 702 discovery cancer outlier profile analysis, 703–704 ERG androgen regulation, 705–706 fluorescence in situ hybridization (FISH), 705 quantitative polymerase chain reaction (QPCR), 704 TMPRSS2 fusion transcripts, 704, 705 ETS overexpression mouse prostatic intraepithelial neoplasia (mPIN), 711 plasminogen activator pathway, 712 VCaP cell line, 711–712 hematologic malignancies, 702 subtypes ETV4 (17q21) exons, 706 ETV5, 707, 709

Index

HNRPA2B1, 708–710 massively parallel signature sequencing (MPSS) data sets, 707 new regulatory elements, 707, 708 oncomine database, 706, 707 R1881, 707 TMPRSS2:ERG, clinical target, 716–717 Gene mutations AF2 germline clusters A and B, 255 equilibrium binding affinity, 254–255 amino-terminal domain 43–90 collocation, 223–224 221–269 collocation, 225–227 431–455 clustering, 227–228 497–537 collocation, 230 ligand-independent activation, 229 NH2-terminal domain (NTD), 223 transactivation units (TAU), 220–222 DNA-binding domain 574–586 collocation, 218–219 response elements, 219 ligand-binding domain (LBD) 715–730 and 670–678 collocation, 213 741–757 and 791–798 collocation, 213–214 874–911 collocation, 212 ansamycin antibiotics, 215–216 approach, 218 germline mutations, 210 molecular chaperones system, 214–215 syndrome of steroid-hormone and anti-antiandrogen withdrawal, 214 truncated receptors and constitutive bypass mechanism, 216–218 somatic type H874Y, V715M, and T877A, 258 V730M, 257 Germline modification techniques, 277 Gioeli, D., 180, 487 Gleason grading system, 702 Gleason scores, steroid 5aR inhibitor effects, 168–169 Gleave, M.E., 681 Glinsky, G.V., 703 Glutamine repeat length, 249 Glycoprotein 330. See Megalin receptor Greenberg, N.M., 207

791

Gregory, C.W., 320 Groucho gene, 574 Growth factor receptor-bound protein 10 (GRB10), 531 GSK-3b protein kinase, 576–577 Guanylyl cyclase a1 (sGCa1), 646 Guo, Z., 556 Gupta, A., 587

H Haelens, A., 385 Haendler, B., 397 Hara, T., 36 Health-related quality of life (HrQOL), 191 Heat-shock protein 27 (Hsp27), 694 Heat-shock protein 90 (Hsp90), 518, 619 He, B., 302 Heemers, H.V., 358, 723 Heinlein, C.A., 395 Heintzman, N.D., 410 Hessels, D., 716 High-grade prostatic intraepithelial neoplasia (HGPIN), 760 Hiipakka, R.A., 9 Hinge region DNA binding domain (DBD), 395 intracellular localization, 396 PEST sequence, 397 post-translational modifications, 396–397 Histone acetyltransferases (HAT) acetyltransferases P/CAF and p300, 323 Tip60, 323–324 deacetylases classical HDACs, 324 corepressors, 325 sirtuins, 324–325 p160 coactivators SRC-1, 320–321 SRC-2, 321–322 SRC-3, 322 Histone deacetylase inhibitors (HDACIs), 218 Histone modifications ChIP-on-chip, 407–408 ChIP sequencing, 408

792

coregulators, 408 H3 and 4 amino-terminal tail, 408–409 histone-modifying enzymes, 412–413 Hobisch, A., 451 Hodges, C.V., 57, 509, 619, 701, 750 Hormone response elements (HREs), 392 Houston, 180 Hsing, A.W., 141, 178 Hsp70/Hsp90 chaperone system, 214 Huang, H-Y., 587 Huggins, C., 1, 10, 12, 14, 57, 142, 194, 509, 619, 701, 702, 750 Hur, E., 302 Hydroxyflutamide, 134 3b-Hydroxysteroid dehydrogenase 1 (HSD3B1), 126 Hydroxysteroid dehydrogenase (HSD) genes aldo-keto reductase (AKR) 1C gene family, 147 DHT catabolism, 146–147

I Imperato-McGinley, J., 18, 177 Insulin-like growth-factor-binding proteins (IGFBPs), 643 Insulin-like growth factor-1 (IGF-I), 643 Interleukin-6 (IL-6) AR activation cAMP and bombesin, 452 cofactors, 458 Etk and Pim1, 458 STAT3, 457 AR regulation cancer progression, 458–459 cytokine expression, 454 neuroendocrine cells, 455–457 nonsteroidal activation, 455 prostate cancer, 459 signaling Janus kinase/signal transducers, 454 regulation, 453–454 Intermittent androgen-deprivation therapy (IADT) advantages, 198–199 ON cycle, 189–191 OFF cycle

Index

health-related quality of life (HrQOL), 191 steroid 5a-reductase (SRD5A), 194– 198 vs. continuous ADT, 191–192 Intraprostatic androgen level benign and untreated primary PCA, 125 CRPC, 125–126 In Vivo Imaging System (IVIS), 93 Isotope-coded affinity tags (ICAT) method, 636, 652 Iyer, M., 100

J Jensen, E., 13, 16, 18, 30 Jia, L., 405, 459 Jiang, C., 763 Jiang, F., 641 Jin, R.J., 587 Jung, M., 747

K Kaltz-Wittmer, C., 357 Kantoff, P., 57 Karunasinghe, N., 759 Kennedy, E., 12 Kennedy’s disease, 388 Khalid, O., 405 Kim, K.H., 640 Klotz, L.H., 191 Kokontis, J.M., 9 Kranse, R., 760 Kung, H.-J., 507

L Lapointe, J., 703, 709 Latham, J.P., 96 LBD. See Ligand-binding domain LDL-receptor-related protein (LRP), 572, 573 Lehninger, A., 12 Lentiviruses, TSTA, 106 Leucine zipper-motif (LUZP2), 530 Levina, E., 569

Index

Liang, T., 20 Liao, S., 9, 176, 480 Liao, X., 749 Li, B.Y., 640 Ligand-binding domain (LBD), 394–395, 743–745 AF2 coactivator-binding pocket allosteric regulation, 307–308 corepressor peptide (SHP), 303 drugable interaction surface, 305 hypothesized complexes, 298, 301 LXXLL and FXXLL peptides, 300, 301 mutations, 305 nuclear receptor family, 305–306 structures, 299, 301 trifurcated groove, 299, 301, 304 ansamycin antibiotics, 215–216 antiandrogens, 301, 305, 307 coactivators and corepressors, 300–301 715–730 and 670–678 collocation, 213 741–757 and 791–798 collocation, 213–214 874–911 collocation, 212 FxxLF and LxxLL motifs interaction, 378 germline mutations, 210 ligand-binding pocket (LBP), 301 molecular chaperones system, 214–215 NTD and cofactor-binding groove mutational analysis, 379–381 peptides interaction, 377 protein-protein interaction, 376 structure, 378 in vivo interactions, 381–382 signal preventing approaches, 218 syndrome of steroid-hormone and anti-antiandrogen withdrawal, 214 truncated receptors and constitutive bypass mechanism, 216–218 Ligand-binding pocket (LBP), 301 Ligand-independent AR activation cAMP/PKA pathway, 484–485 interleukin-6 (IL-6) MAPK pathway, 443 p300 expression, 443–444 Li, H., 759 Lin, D.L., 455 Linja, M.J., 555 Lin, T.M., 193

793

Lin, Y., 36 Lipogenic switch antineoplastic intervention acetyl-CoA, 734 antineoplastic agents, 735 cerulenin, 733 fatty acid synthase (FASN) inhibition, 733–734 malonyl-CoA accumulation, 734 control layers, 727, 729 enzyme expression, 729–730 lipid biosynthesis androgen receptor (AR), 724 fatty acids and cholesterol, 725 metabolic pathway, 725, 726 lipogenesis fatty acid synthase (FASN), 732–733 hypothalamic cells, 733 lipoprotein lipase (LPL), 732 phospholipids, 731–732 mechanism, 730–731 sterol regulatory element-binding proteins (SREBPs), 725, 727, 728 Lipoprotein 313 receptor-related protein 2 (LPR2). See Megalin receptor Liver X receptor (LXR) antiproliferative effect cell proliferation suppression, 40 proliferation inhibition, 41 discovery and characterization, 38 oxysterol receptor, 39 phytosterols, 41–42 prostate cancer progression inhibition, 40–41 signaling, novel therapy antiproliferative effect, 40 oxysterol receptor, 39 phytosterols, 41–42 prostate cancer progression inhibition, 40–41 UR discovery and characterization, 38 Living Image, 93 Li, Z., 178 LNCaP 104 model androgen-independent and repressed cells, 31–32 prostate cancer cells progression

794

androgen deprivation and antiandrogen therapy, 33 Casodex-resistant cells, 33–34 Logan, S.K., 587 Lonard, D.M., 358 Louro, R., 640, 648 Luteinizing hormone-releasing hormone (LHRH), 478 LXXLL coactivator. See Functional motifs

Index

Mitogen-activated protein kinase (MAPK), 110, 354–355 Miyamoto, H., 73 Mizokami, A., 522, 557, 558 Modulator of nongenomic activity of estrogen receptor (MNAR), 333 Mohler, J.L., 1, 175, 178, 553, 557 Molecular chaperones, 348–349 Moore, R.J., 180 Mouse mammary tumor virus (MMTV), 250–251

M Ma, D.P., 27 MAGE-11 coregulator expression pattern, 260–261 FXXLF motif, 258–260 transcriptional activity, 259 Magee, J.A., 639 Majumder, P.K., 534 Makridakis, N.M., 149 Male germ cell-associated kinase (MAK), 331 Mangan, F.R., 480 Marshall, J.R., 758 Martin, D.B., 640, 652 Massie, C.E., 650, 665, 670 Matias, P.M., 302 Matsuda, T., 457 Matthaei, J.H., 15 Matusik, R.J., 97, 112, 113, 587 May, B., 10, 12 McCormick, F., 98 Meehan, K.L., 640 Megalin receptor, 690–691 Melatonin, 478 Memorial Sloan-Kettering Cancer Center (MSKCC), 63–64 Methylseleninic acid (MSeA), 757, 762 Methylselenol (CH3SeH), 761–763 Methyltransferases, AR regulation, 325–326 Metzger, E., 389 Microarray analysis, 746 Microvessel density (MVD), 160–161 Midkine (MDK), 531 Migliaccio, A., 361 Miller, K., 191 Mirosevich, J., 587

N Nadiminty, N., 755 Nandana, S., 587 Nantermet, P.V., 641 N/C interaction dissociation kinetics, 242–243 ligand and DNA binding, 244 polyglutamine tract, 270–271 stabilization mechanisms, 245 protein structure, self-association model, 245–246 Need, E.F., 207 Nelius, T., 78 Nelson, C.C., 681 Nelson, P.S., 631, 640 Neri, R., 18 Nesbit, R.M., 67 Neuroendocrine differentiation (NED) human xenograft model, 515 in vitro cell line studies, 515–516 neuroendocrine cells, 514–515 transcriptional profile dihydropyrimidinase-like 2 (DPYSL2), 531–532 leucine zipper-motif (LUZP2), 530 midkine (MDK), 531 NKX3.1 downregulation, 529 transducin-like enhancer of split 1 (TLE1) expression, 530 transmembrane proteins, 530–532 Ngan, E.S., 332 NH2-terminal domain (NTD) AF2 pocket, 305

Index

functional role, 444–445 FxxLF and LxxLL motifs interaction, 378 interacting proteins, 391 LBD and cofactor-binding groove in vivo interactions, 381–382 mutational analysis, 379–381 peptides interaction, 377 protein-protein interaction, 376 structure, 378 N/C interaction, 391 regulation and deregulation, 441–443 structure, 391 transcription activation unit 1, 390 transcription activation unit 5, 389–390 Nickols, N.G., 640 Nirenberg, M.W., 15 Nishiyama, T., 557, 558 Niu, Y., 73 NKX3.1 downregulation, 529 Ntais, C., 146 N-terminal domain (NTD). See NH2terminal domain Nuclear clusterin (nCLU) Ku70 binding, 689 prostate epithelial cells, 688 protein structure, 683 Nuclear export signal (NES), 394 Nuclear localisation sequences (NLS), 686 Nuclear receptors (NRs), 375

O Ochoa, S., 15 OFF cycle, IADT health-related quality of life (HrQOL), 191 steroid 5a-reductase (SRD5A) dutasteride, 198 EAF2/U19, 195 finasteride, 195–197 serum PSA, 198 Ohouo, P., 569 O’Malley, B.W., 113, 480 Oncostatin M, 458 Oosterhoff, J.K., 640 Orgebin-Crist, M.C., 587

795

P Page, S.T., 557, 558 Pang, S.T., 640 Paoloni-Giacobino, A., 705 Paul, M., 587 Peng, L., 644 Penning, T.M., 144 Pereira de, J., 302 Perner, S., 712, 715 PEST (proline–glutamate–serine and threonine-rich) sequence, 397 Pfundt, R., 641 Phosphatidylinositol 3´-kinase (PI3K), 730 Phosphodiesterases (PDE) isoforms, 473–474 targeting therapies, 474–475 PI3K-Akt pathway hyperactivation, 524–525 Pigment epithelium–derived factor (PEDF), 413–415 Pituitary adenylate cyclase-activating polypeptide (PACAP), 478 Polyglutamine (polyQ) tract biological role, 223–224 CAG repeats, 276 epidemiology, 274–275 functional analysis Ran/ARA24 coactivator, 273–274 RNA-binding proteins, 272–273 germline modification techniques, 277 humanized mice development cre-recombinase, 277 differentially expressed genes, 279–280 h/mAR48Q characterization, 278–279 nuclear receptor superfamily N/C interaction, 270–271 Q tract risk, 271–272 transgenic adenocarcinoma of the mouse prostate (TRAMP) castration, 286–288 disease progression, 285–286 immunohistochemistry (IHC), 282–283 neuroendocrine phenotype, 283–285 prostatic intraepithelial neoplasia (PIN) role, 280–281 PTEN tumor suppressor, 280 Pomerantz, M., 57

796

Proline–glutamic acid and leucine-rich protein-1 (PELP-1), 333 Prostate Cancer Prevention Trial (PCPT), 143 Prostate-specific antigen doubling time (PSADT), 61, 64 Prostate-specific antigen (PSA), 757 AR transactivation, 250–251 cancer diagnosis, 188 promotor and enhancer, 95–96 TSTA activity, 105–106 Prostate stromal cells (PRSCs), 625 Prostatic intraepithelial neoplasia (PIN), 280–281 Protein Data Bank (PDB), 301–303 Protein kinase A (PKA) activity A-kinase anchoring proteins (AKAPs), 468–469 extracellular PKA, 471–472 isoforms prostate cancer cells, 469–471 R and C subunits, 469 ribosomal S6 kinase and MAPK/ERK, 472–473 targeting therapies, 472 Protein-kinase-C-related-kinase-1 (PRK1), 332

Index

and prostate cancer treatment, 158–159 tissue transglutaminase (tTG), 160 TUNEL, 159–160 isozymes, CRPC AR transactivation, 178 express sequence tags (EST), 179 inhibitors activity, 177–178 intracellular location, 176 molecular genetics, 176–177 mRNA levels, 179 testosterone metabolism, 176 SRD5A2 gene enzyme activity and somatic mutation, 146 mutation and SNPs, 145 Reichardt, J.K.V., 141 Rennie, P.S., 341 Rittmaster, R.S., 155 RNA and protein synthesis androgen and mRNA synthesis, 14–15 hormone-gene theory, 14 proteins modulation, 16 RNA polymerases multiple forms, 15 Robins, D.M., 207, 269 Rosenfeld, M.G., 480 Ross, J.S., 361 Ross, R.K., 143 Rowan, B.G., 320 Rowland, J.G., 640, 651

R Ran/ARA24 coactivator, 273–274 5a-Reductase (5aR). See also Steroid 5a-reductase (SRD5A) histology ARI40010, 166–168 ARIA2003, 163–166 dutasteride, 162–163 finasteride, 161–162 inhibitor effects in androgen-deprivation therapy (ADT), 157–158 benign prostatic hyperplasia (BPH), 157 Gleason scores, 168–169 histopathology, 161 MK386, 159 MVD and VEGF, 160–161

S Sabichi, A.L., 760 Sack, J.S., 302 Sadar, M.D., 465, 640 Salvati, M.E., 302 Savory, J.G., 180 Sawyers, C.L., 112, 743 Scher, H.I., 207 Schlicht, M., 763, 767 Scholz, M.C., 197 Schrauzer, G.N., 758 Scott, W.W., 12 Secreted clusterin (sCLU), 688 cancer, 689 expression, 692–694 protein structure, 683–684

Index

Segawa, T., 640, 643 Selenium and Vitamin E Cancer Prevention Trial (SELECT), 760, 761 Selenoenzymes, 764 Sex-hormone-binding globulin (SHBG), 691 Shaffer, P.L., 302 Shahinian, V.B., 188 Shamberger, R.J., 758 Shapiro, E., 587 Shaw, G.L. Sheflin, L., 517 Shionogi tumour model, 692–693 Shi, X.B., 640 Short hairpin RNA (shRNA), 218 Shtutman, M., 569 Simard, J., 559 Simpson, J.L., 176 Small ubiquitin-like modifier (SUMO), 329, 389–390 Smith, D.C., 67 Snoek, R., 341 Somatic mutation, 81–82 Spectrum, 93 Spencer, D.M., 113, 114 Spermine synthase (SMS) protein, 642 Spinal bulbar and muscular atrophy (SBMA), 520 SRC/p160 coactivators FXXLF motif, 248–249 transactivation regulation, 251–252 SRD5A2 gene, 145–146 SREBP-cleavage-activating protein (SCAP), 727, 728 Stanbrough, M., 766 Stege, R., 559 Stegmaier, K., 717 Steroid 5a-reductase (SRD5A) EAF2/U19, 195 inhibitors dutasteride, 198 finasteride, 195–197 serum PSA, 198 Steroid receptor coactivator-1 (SRC-1), 320–321 Steroid receptor coactivator-2 (SRC-2), 321–322 Steroid receptor coactivator-3 (SRC-3), 322

797

Sterol regulatory element-binding proteins (SREBPs), 131, 725, 727, 728 Stromal and epithelialandrogen receptors, 77–80 Sulfotransferase (SULT), 148–149 Sun, Q., 587 Suzuki, K., 587 Swinnen, J.V., 723 Synaptophysin expression, 283

T Takayama, K., 665 Talalay, P., 10, 12, 13 T877A mutations, 743–745 Tanner, M.J., 569 Tanner, T., 385 Tat-interactive protein (Tip60), 396 T-cell factor/lymphoid-enhancing factor-1 (Tcf/LEF-1), 574 Tepper, C.G., 507 Terminal deoxynucleotidyl (TdT)-mediated dUTP nick end labeling (TUNEL) marker, 158–160 Testosterone, 176, 588, 589 5a–DHT biological functions, 18–19 DHT selective retention nuclear retention, 17 prostate cell nuclei, 17–18 dihydrotestosterone (DHT) prostate regrowth, 193–194 specific effects, 193 luteinizing hormone (LH), 19 pseudohermaphroditism, 19 5a-reductase inhibitors epicatechin-3-gallate (ECG), 21 epigallocatechin-3-gallate (EGCG), 21–22 SRD5A, 192–193 structures, 20 TDD5/N-myc downstream-regulated gene 1 (NDRG1), 193 conversion 5a–DHT selective retention, 16–18 5a-reductase inhibitors medical use, 19–20 testosterone and 5a–DHT biological functions, 18–19

798

Testosterone-repressed prostate message2 (TRPM-2), 513 Therapy-mediated selection pressure, 210 Thomas, L.N., 155 Thompson, J., 744 Thorpe, J.F., 178 Tilley, W.D., 207, 744 Tindall, D.J., 1, 358, 427 Tissue androgen levels clinical relevance, AR-activating levels, 559 DHT, 556–558 flutamide, 557, 558 potential sources adrenal androgens, 560 5a-androstanedione (5a-ASD), 561 serum DHEA-SO4 level, 560 prostate-specific antigen (PSA), 558 radioimmunoassay (RIA), 556 technical barriers, 561–563 transurethral prostatectomy, 557 Tissue transglutaminase (tTG) marker, 158, 160 Titus, M.A., 175, 178, 553, 557 TMPRSS2:ERG gene fusion clinical target, 716–717 ERG androgen regulation, 705–706 Tomkins, G., 13 Tomlins, S.A., 701 Transactivation units (TAU), 220–222 Transcription activation unit 1 (Tau-1), 390 Transcription activation unit 5 (Tau-5), 389–390 Transcription factor complexes ETS family, 674 FoxA1, 671 GATA family, 671, 672 Oct1, 672 Transducin-like enhancer of split 1 (TLE1) expression, 530 Transgenic adenocarcinoma of the mouse prostate (TRAMP) castration, 286–288 disease progression, 285–286 immunohistochemistry (IHC), 282–283 neuroendocrine phenotype, 283–285 prostatic intraepithelial neoplasia (PIN) role, 280–281

Index

PTEN tumor suppressor, 280 Trapman, J., 375 Two-step transcriptional amplification (TSTA) cell culture studies adenovirus, 101–102 LNCaP cells, 101 titratable system, 100–101 transgenic mouse models, 111–112 xenografts AdTSTA modifications, 108–110 androgen deprivation, AdCMV-FLuc, 103 chromatin immunoprecipitation (ChIP), 103, 105 flutamide effect, 106, 107 lentiviruses, 106 lymph node metastases, 109 MAPK combination, 110 MicroPET/CT imaging, 108 PSA enhancer, 105–106 recurrent state transition, 104 sr39tk gene, 107

U Ubiquitin ligases, 328 Ubiquitin-proteasome pathway (UPP), 517–518 Ubiquitous receptor (UR), 38 UDP Glycosyltransferase 2, B15 (UGT2B15), 129 Uemura, M., 179 UR. See Ubiquitous receptor Urogenital sinus (UGS), 74–75

V Van Royen, M.E., 391 Vascular endothelial growth factor (VEGF), 160–161 Vasoactive intestinal peptide (VIP), 479 Velasco, A.M., 640 Veldscholte, J., 744 Venkateswaran, V., 758 Verhoeven, G., 723 Visakorpi, T., 745

Index

W Wafa, L.A., 341 Waghray, A., 639, 640 Walsh, P.C., 177 Wang, Q., 641, 650 Wang, R., 701 Wang, X.D., 641 Wang, Y., 587 Wang, Z., 763 Watson, P.A., 743 Weigel, N.L., 315 Weiss. S., 14 Welsh, J.B., 703 Western blotting, 593, 594 Whitaker, H.C., 640 Williams-Ashman, H.G., 13, 14, 18 Wilson, E.M., 241 Wilson, J.D., 17, 180 Wnt signaling pathway androgen deprivation therapy (ADT), 570 AR gene expression, 577–578 b-catenin archetypal paracrine process, 573 Frizzled (Fzd) gene, 572–573 Groucho gene, 574 intracellular domain, cadherin, 571–572 phosphorylation and ubiquitination, 572 prostate cancer cells, 575–576 stabilization process, 573–574 T-cell factor/lymphoid-enhancing factor-1 (Tcf/LEF-1), 574 DNA-binding transcription factor, 570 genetic perturbation, 571 graphical characterization, 581, 582 GSK-3b protein kinase, 576–577 human prostate cancer b-catenin-stabilizing mutations, 579–580 Dickkopf-1 (Dkk-1), 581

799

metastatic/hormone-refractory cancer tissues, 580 normal and malignant prostate epithelial cell growth, 578–579 Wright, M.E., 640, 652 Wu, L., 91, 96, 98, 100 WXXLF motifs. See Functional motifs

X Xenografts, TSTA AdTSTA modifications, 108–110 androgen deprivation, AdCMV-FLuc, 103 chromatin immunoprecipitation (ChIP), 103, 105 flutamide effect, 106, 107 lentiviruses, 106 lymph node metastases, 109 MAPK combination, 110 MicroPET/CT imaging, 108 PSA enhancer, 105–106 recurrent state transition, 104 sr39tk gene, 107 Xie, X., 97 Xu, L.L., 640 Xu, Y., 178

Y Yang, Y., 559 Yeh, S., 73 Yoshida, T., 745 Yu, S.Y., 758 Yu, X., 587 Yu, Y.P., 703

Z Zhou, J., 361