University of Southern California Dissertations and Theses

The androgen receptor: Its role in the development and progression of cancers of the prostate and breast

(USC Thesis Other)

The androgen receptor: Its role in the development and progression of cancers of the prostate and breast

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content INFORMATION TO USERS This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer. The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send UM I a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion. Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps. Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6” x 9” black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order. ProQuest Information and Learning 300 North Zeeb Road, Ann Arbor, M l 48106-1346 USA 800-521-0600 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. THE ANDROGEN RECEPTOR: ITS ROLE IN THE DEVELOPMENT AND PROGRESSION OF CANCERS OF THE PROSTATE AND BREAST ©2000 by Ryan Andrew Irvine A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment o f the Requirements for the Degree DOCTOR OF PHILOSOPHY (Molecular Microbiology and Immunology) August 2000 Ryan Andrew Irvine Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UM I Number: 3018090 Copyright 2000 by Irvine, Ryan Andrew Ail rights reserved. __ ___ __® UMI UMI Microform 3018090 Copyright 2001 by Bell & Howell Information and Learning Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code. Bell & Howell Information and Learning Company 300 North Zeeb Road P.O. Box 1346 Ann Arbor, Ml 48106-1346 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UNIVERSITY OF SOUTHERN CALIFORNIA THE GRADUATE SCHOOL UNIVERSITY P A R K LOS ANGELES. CALI FORNI A 9 0 0 0 7 This dissertation, written by Ryan Andrew Irvine under the direction of h is Dissertation Committee, and approved by all its members* has been presented to and accepted by The Graduate School, in partial fulfillment of re quirements for tiie degree of DOCTOR OF PHILOSOPHY June 16, 2000 DISSEI Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DEDICATION To Julie Zissimopoulos, my wife, for her love and unwavering commitment. To Terrance and Judith Irvine, my parents, for their unselfish support of family, education, and achievement. To Traci Irvine, my sister, for her unconditional love. To Constantine and Carole Zissimopoulos, my parents in-law, for their friendship and generosity. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ACKNOWLEDGMENTS I am greatly indebted to Dr. Gerhard Coetzee for his steadfast support and encouragement, his expert knowledge and professionalism, his genuine scientific curiosity, and his warm friendship. Under his tutelage, I have learned to ask meaningful scientific questions and in their pursuit, to persevere with confidence and grace. Furthermore, I am grateful to Dr. Peter Jones for demanding excellence from all who participate in his weekly lab meetings. Indeed, his crystalline understanding that competent scientific discovery must be tempered with equally competent scientific presentation has served me well. Sincerest thanks are due to Dr. Michael Stallcup for his constructive guidance and generosity. The continued support of his laboratory and the use of his precious reagents ensured the success o f my thesis research. Dr. Stanley Tahara is also deserving of much gratitude for being most effective as a teacher, a scientist, and as a student advocate. He is the proverbial ‘rock’ of the Department. I would also like to thank Dr. Lucio Comai for his steady participation in my thesis research. His exacting criticism and thoughtful advice have been crucial to my success. Finally, Drs. Mimi Yu, Ron Ross, and Sue Ingles must be recognized for their tremendous contributions to my work. They truly are visionary scientists. 111 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TABLE OF CONTENTS DEDICATION............................................................................................................ ii ACKNOW LEDGEMENTS..................................................................................... iii LIST OF FIGURES.................................................................................................... ix LIST OF TABLES..................................................................................................... xii ABSTRACT........................................................................................... xiii CHAPTER 1 The prostate: growth, differentiation, and malignancy................. 1 Morphology and development of the human prostate.................... 1 Cancer of the prostate: incidence and pathology.......................... 3 The natural history and androgen dependence of prostate cancer.................................................................................................. 4 Genetic predisposition to prostate cancer....................................... 7 Thesis perspective.............................................................................. 11 CHAPTER 2 Androgen receptor gene allelic variation and susceptibility to prostate cancer.............................................................................. 13 INTRODUCTION........................................................................................... 13 Putative prostate cancer susceptibility genes.................................. 14 The AR gene..................................................................................... 15 Kennedy’s disease and aberrant AR function................................. 16 A circumstantial link between AR CAG repeat length and prostate cancer risk........................................................................... 17 MATERIALS AND METHODS................................................................... 19 Prostate cancer patients.................................................................... 19 Control subjects................................................................................. 20 Microsatellite and polymorphism analyses..................................... 21 DNA extraction.................................................................... 21 CAG repeat size determination........................................... 22 GGC repeat size determination........................................... 22 PSA allele determination.................................................... 23 Statistical analyses............................................................... 24 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. RESULTS....................................................................................................... 25 AR CAG and GGC allele frequencies differ significantly among racial-ethnic groups.............................................................. 25 Prevalence of a high-risk CAG/GGC allelotype in White prostate cancer patients..................................................................... 26 The CAG and GGC microsatellites of the AR are in linkage disequilibrium in men with prostate cancer..................................... 26 Short AR CAG alleles are associated with increased risk for advanced prostate cancer.................................................................. 27 Interaction between genotypes at the AR and PSA loci................ 28 DISCUSSION................................................................................................ 40 The AR CAG microsatellite and prostate cancer risk.................... 40 The AR GGC microsatellite and the prostate cancer allelotype... 44 Parsimonious conclusions regarding AR microsatellite variation and prostate cancer risk.................................................... 46 Age at onset of prostate cancer and the AR CAG microsatellite... 48 Evidence o f gene-gene interactions between the AR and PSA loci in conferring prostate cancer risk.............................................. 49 AR microsatellite variation and correlation with other human disorders............................................................................................. 51 Some final considerations................................................................ 52 CHAPTER 3 Inhibition o f p i 60 coactivation with increasing AR poly-Q length.................................................................................................. 54 INTRODUCTION......................................................................................... 54 The nuclear receptor superfamily.................................................... 55 AR structure and function................................................................ 56 The NR coactivator complex and chromatin remodeling.............. 58 The pl60 coactivators...................................................................... 59 A novel NR-p 160 coactivator interaction....................................... 61 Analysis of a somatic AR mutant.................................................... 63 MATERIALS AND METHODS................................................................ 65 Plasmids............................................................................................. 65 Yeast two-hybrid assays.................................................................. 67 GST pull-down assays..................................................................... 67 Cell culture and transfections.......................................................... 68 Western blotting and scanning densitometry.................................. 69 Ligand binding assays....................................................................... 70 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. RESULTS....................................................................................................... 71 The AR NTD interacts with the C-terminus o f GRIP 1................... 71 GRIP1 C-terminus binds to the AR NTD downstream of the TAU-1 core region............................................................................. 72 Coactivation of the AR through the AR NTD-GRIP1 C- terminal interaction............................................................................ 72 AIBI and SRC-1 a also functionally interact with the AR NTD... 74 Variation in the AR poIy-Q tract affects AR transactivation activity................................................................................................. 74 Increasing poly-Q length inhibits pl60-mediated coactivation of the AR............................................................................................. 76 The AR (2xLeu) mutant has enhanced transactivation activity compared to wild type AR................................................... 77 DISCUSSION................................................................................................. 109 The 2xLeu insertion mutation confers a gain-in-function on the AR............................................................................................. 117 CHAPTER 4 BRCA1 is a coactivator o f the androgen receptor........................ 118 INTRODUCTION.......................................................................................... 118 A circumstantial link between hereditary breast cancer and the AR......................................................................................... 118 Androgens and the control of breast cell proliferation.................. 119 The genetics of breast cancer susceptibility and BRCA1.............. 121 Putative functions of BRCA 1............................................................ 122 A rationalization for how BRCA I mutations lead to breast cancer development................................................................. 124 Does BRCA1 influence AR signaling?............................................. 126 MATERIALS AND METHODS................................................................... 127 Plasmids................................................................................................ 127 Cell culture and transfections........................................................... 128 GST pull-down assays...................................................................... 129 RESULTS......................................................................................................... 130 BRCA1 coactivates AR signaling through AR AF-1..................... 130 BRCA1 and the p i60 coactivators synergistically potentiate AR signaling in prostate- and breast-derived cell lines.................. 131 BRCA1 enhances AR AF-2 activity in the presence of p 160 coactivator.................................................................................. 132 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. BRCA1 physically interacts with the AR and GRIP1 in GST pull-down assays...................................................................... 133 BRCA1 potentiates p i60 mediated ER signaling in prostate- and breast-derived cell lines............................................................. 134 Exogenous BRCA1 expression masks the AR poly-Q effect 135 DISCUSSION.................................................................................................. 155 BRCA1 is an AR coactivator........................................................... 155 The molecular basis for BRCA1 function in AR coactivation.... 156 BRCA1 inhibition o f ERa signaling?............................................ 157 Reconciling the differences............................................................. 159 To end at the beginning................................................................... 161 CHAPTER 5 Identification and cloning o f androgen receptor interacting factors................................................................................................ 163 INTRODUCTION......................................................................................... 163 The yeast two-hybrid system........................................................... 164 Proteins that interact with the AR.................................................. 166 The nature of hypothesis-generating research.............................. 168 MATERIALS AND METHODS............................................................... 169 Plasmids........................................................................................... 169 Titer determination and amplification of the pACT2 human testis cDNA library.......................................................................... 170 Yeast two-hybrid screening............................................................ 171 Yeast plasmid isolation and recovery in E. coli............................. 173 Yeast two-hybrid P-galactosidase assays....................................... 174 Yeast protein extract preparation................................................... 175 Western analysis................................................................................ 176 Northern analysis.............................................................................. 176 Cell culture and transfections.......................................................... 177 Sequence database searching and ORF analysis........................... 178 RESULTS....................................................................................................... 179 Cloning and characterization of GAL4-AR NTD yeast two- hybrid ‘bait’ proteins......................................................................... 179 Yeast two-hybrid screen of a human testis cDNA library 180 Preliminary characterization of ARIPs encoded at Xq28............. 181 Preliminary characterization of ARIP29........................................ 184 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DISCUSSION.............................................................................................. 211 Identification and confirmation of putative AR interacting proteins............................................................................................. 212 ARIPs encoded at the Xq28IMAGE-II locus............................... 213 A'fAGE-l 1 mRNA function in AR transactivation?...................... 214 ARIP29............................................................................................. 217 ARIP29/hASB-3 is expressed in the prostate and may modulate AR signaling.................................................................... 218 Some final considerations............................................................... 220 SUMMARY OF PRINCIPAL FINDINGS............................................................ 222 REFERENCES......................................................................................................... 224 viii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF FIGURES Figure Page 2.1 Autoradiograms o f AR CAG and GGC microsatellite PCR products...................................................................................... 30 2.2 Cumulative distribution of AR CAG allele sizes among White control and White prostate cancer cases with advanced or localized disease........................................................... 31 3.1 Schematic diagrams of the structural/functional domains of the AR and two AR mutants......................................................... 79 3.2 NR-mediated transcriptional activation............................................. 80 3.3 Schematic diagrams of the structural/functional domains of the p 160 coactivators..................................................................... 82 3.4 Schematic representation of the construction o f pcDNA- hAR(Q)n vectors.................................................................................. 84 3.5 Androgen regulated promoters.......................................................... 86 3.6 The AR NTD interacts with the GRIP1 C-terminus in yeast two-hybrid and GST pull-down assays............................................. 88 3.7 GRIP-1 C-terminus binds to the AR NTD downstream of the TAU-1 core domain...................................................................... 90 3.8 GRIP-1 mutants................................................................................... 92 3.9 Coactivation of the AR via the NTD-GR1P1 C-terminal interaction............................................................................................ 94 3.10 Functional interactions of pl60 coactivators with a constitutively active AR mutant........................................................ 96 3.11 Determination of AR(Q)n protein expression by specific ligand binding..................................................................................... 98 ix Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3.12 Determination of AR(Q)n protein expression by Western blot densitometry............................................................................... 100 3.13 AR transactivation activity decreases with increasing poIy-Q size......................................................................................... 102 3.14 Inhibition of p 160-mediated coactivation o f the AR with increasing poly-Q size............................................................. 104 3.15 Relative expression and transactivation activity of the AR (2xLeu) mutant.......................................................................... 106 4.1 BRCA1 potentiates AR transactivation activity............................. 138 4.2 BRCA1 works through AR AF-1..................................................... 140 4.3 Synergistic potentiation of AR signaling by BRCA1 and the p 160 coactivators................................................................ 142 4.4 Potentiation of AR signaling by BRCA1 occurs in both prostate- and breast-derived cell lines.................................... 144 4.5 BRCA1 potentiates GRIP-1 mediated coactivation o f A R A F-l and AF-2............................................................................ 146 4.6 BRCA1 interacts with the AR NTD and the GRIP 1 C-terminus......................................................................................... 148 4.7 BRCA1 does not potentiate or repress ER signaling in PC-3 cells...................................................................................... 150 4.8 BRCA1 potentiates pi 60-mediated coactivation of ER signaling in prostate- and breast-derived cell lines....................... 152 4.9 BRCA 1 masks the AR poly-Q effect............................................... 154 5.1 Schematic diagram indicating the amino acid boundaries of three AR NTD fragments used in the yeast two- hybrid system..................................................................................... 187 5.2 Background activities and expression levels of AR NTD- GAJL4 DBD proteins......................................................................... 189 x Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5.3 Representative screen to confirm interactions between AR NTD (5-156) and putative interacting proteins...................... 191 5.4 Interactions of AR NTD (5-156) with proteins encoded at the Xq2S/MAGE-ll locus.......................................................... 193 5.5 Presumed exonic structure o f the Xq28/MAGE-11 clones 195 5.6 Putative amino acid sequences for MAGE-11 and Xq28- encoded ARIPs................................................................................. 197 5.7 Expression of ARIP-GAL4 AD proteins determined by Western analysis............................................................................. 198 5.8 Assessment of MAGE-11 mRNA function on AR transactivation in mammalian cells................................................ 200 5.9 Interactions between ARIP29 and The ARNTD.......................... 202 5.10 Putative ARIP29 amino acid sequence.......................................... 204 5.11 ARIP29 mRNA expression in human tissues............................... 206 5.12 Assessment of ARIP29 function on AR transactivation and expression in mammalian cells................................................ 208 xi Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF TABLES Table Page 2.1 Distributions of AR CAG microsatellite allele frequencies (%) among control subjects according to racial-ethnic grouping............................................................................................ 32 2.2 Distributions of AR GGC microsatellite allele frequencies (%) among control subjects according to racial-ethnic grouping............................................................................................ 33 2.3 Allelic distributions of the AR CAG and GGC microsatellites in White prostate cancer patients and control subjects................. 34 2.4 Distributions of AR CAG and GGC microsatellite allele frequencies (%) in prostate cancer patients................................... 35 2.5 Frequency of AR CAG alleles among patients with localized and advanced disease compared to control subjects..................... 36 2.6 Allelic distributions of the AR CAG microsatellite and PSA A/G promoter polymorphism in prostate cancer patients and controls...................................................................................... 37 2.7 Odds ratio of prostate cancer for a CAG microsatellite repeat length decrement of six in the AR gene among men in the Physicians’ Health Study................................................................. 38 2.8 Summary of studies evaluating the roles of the AR CAG and/or GGC microsatellites in prostate cancer risk, progression, and age at onset........................................................... 39 3.1 List o f PCR primers......................................................................... 107 5.1 Summary of preliminary yeast two-hybrid results....................... 209 5.2 Results of yeast two-hybrid screen o f a human testis cDNA library with the AR NTD (5-156) bait protein.............................. 210 xii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ABSTRACT The androgen receptor: its role in the development and progression of cancers of the prostate and breast. Ryan Andrew Irvine The size o f the exon 1 CAG microsatellite of the androgen receptor (AR) gene was determined in men from different racial-ethnic groups in Los Angeles, CA. The frequency distribution of AR CAG alleles correlated inversely with prostate cancer risk among the various groups, such that short CAG alleles were most prevalent in high-risk African-Americans, less prevalent in intermediate-risk Caucasians, and least prevalent in low-risk Asian-Americans. In a case-control study of Caucasian men, furthermore, short CAG alleles conferred significantly increased risk of prostate cancer development, particularly of advanced disease. Since short AR CAG alleles encode receptors with increased relative transactivation potential, it was hypothesized that prostatic epithelial cells exposed to such receptors are vulnerable to malignant transformation due to increased androgen-dependent proliferation. In an attempt to reveal the molecular basis for this genotype-phenotype association, interactions between the AR and the p i60 coactivator, GRIP1, were investigated. The C-terminus o f GRIP1 physically interacted with the AR N-terminal domain (NTD), which contains the CAG microsatellite-encoded polyglutamine (poly-Q) repeat. In transactivation experiments in prostatic carcinoma cells, GRIP1 potentiated the activity of a constitutively active AR mutant, comprising xiii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. only the NTD and DNA binding domain (DBD). A GRIP1 NR box mutant, moreover, unable to bind the holo-AR ligand binding domain (LBD), coactivated wt AR, further demonstrating the functional importance of this novel GRIP 1 -AR NTD interaction. Lastly, GRIP1 coactivation o f the AR was inhibited by increasing AR poly-Q length, suggesting a mechanism for poly-Q dependent modulation o f AR transactivation activity. In a parallel study, the role of the BRCA1 tumor suppressor protein in AR signaling was investigated. BRCA1 enhanced ligand-dependent AR transactivation in both prostate and breast cancer cell lines. The effects of BRCA1 were mediated through the N-terminal activation function (AF-1) of the receptor. BRCA1 and the p i 60 coactivators, moreover, synergistically potentiated AR transactivation activity. In addition to functional interactions, BRCA1, through its N-terminus, physically interacted with the ARNTD and with the GRIP1 C-terminus. These findings suggest that BRCA1 may directly modulate AR signaling, and therefore, may affect cancer risk in both the breast and prostate. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 1 The prostate: growth, differentiation, and malignancy Morphology and development o f the human prostate Found only in mammals, the prostate gland is responsible for the production and secretion of various components o f semen. In humans, the prostate enspheres the urethra immediately below the bladder and lies between the symphysis pubis and the rectum (Van de Voorde, 1996). The human prostate comprises four morphologically distinct ‘zones’: the anterior fibromuscular stroma (AFMS), the peripheral zone (PZ), the central zone (CZ), and the transition zone (TZ) (McNeal, 1983). Analysis of the glandular architecture of the prostate has revealed that it consists o f a complex array of tubuloalveolar ducts surrounded by stromal tissues (reviewed by Cunha et al., 1987). Individual ducts are lined by secretory columnar epithelia, which are separated from the stromal compartment of the prostate by nonsecretory basal cells associated with a well- defined basement membrane. Organized in a collagenous extracellular matrix, smooth muscle cells, fibroblasts, vascular endothelia, connective tissue cells, nerve terminals, and lymphatic cells make up the prostatic stroma (Aumuller, 1983). The ratio of epithelial to stromal cells is nearly 1:1 in the human prostate (Bartsch and Rohr, 1980). In an androgen dependent process, the endodermal urogenital sinus (UGS) of the developing fetus gives rise to epithelial outgrowths that invade the surrounding mesenchyme (reviewed by Cunha et al., 1987). These outgrowths or ‘prostatic buds’ 1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. develop into the elaborate network o f secretory ducts in the adult prostate during two distinct periods of growth and differentiation (i.e., periods of ductal morphogenesis). The first period occurs prenatally, while the second occurs during puberty. Importantly, no net growth occurs in the normal adult prostate, an observation that is consistent with very low measured rates of epithelial cell tum-over (Tuohimaa and Niemi, 1974; Tuohimaa, 1980). Prostatic development is androgen dependent. Indeed, development of all the male internal accessory sexual structures requires androgens. Beginning in week 8 of gestation, Leydig cells within the fetal testes produce testosterone (T) which promotes differentiation of the Wolffian duct into the epididymis, vas deferens, and seminal vesicle (reviewed by Wiener et al., 1997). While Wolffian duct differentiation appears to be mediated directly by T, development of the prostate from the UGS requires the reduced T derivative, 5a-dihydrotestosterone (DHT). DHT is created intracellularly by the enzyme 5a-reductase, the expression of which, by cells of the UGS, coincides with T secretion by the fetal testes. Following the initial period of fetal growth and differentiation, the rudimentary prostate does not substantially enlarge until puberty, when a surge in T production causes development of the mature adult gland (Weiner et al., 1997). Continued exposure of the adult prostate to T, moreover, is required for the maintenance of its normal morphology and function. This fact is vividly demonstrated in castrated men, whose prostates sustain dramatic epithelial and stromal cell death leading to involution of the gland (Lindzey et al., 1994). 2 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Cancer o f the prostate: incidence and pathology Among men in the United States, prostate cancer is the most commonly found solid tumor and the leading cause of cancer deaths. In 2000, it is estimated that nearly 180, 000 men will be diagnosed with prostate cancer, and that some 32,000 will die of the disease (Greenlee et al., 2000). Prostate cancer is a proliferative disease that causes abnormal enlargement o f the gland due to aberrant ductal morphogenesis (Cunha et al., 1987). Nearly 70% o f all carcinomas of the prostate originate in the PZ, which surrounds the distal prostatic urethra (reviewed by Van de Voorde, 1996). The remainder occurs primarily in the TZ, which lies anterior to the PZ, surrounds the proximal prostatic urethra, and is the exclusive site of origin of benign prostatic hyperplasia (BPH). Adenocarcinoma accounts for 98% of all neoplasms in the prostate and is characterized histologically by nuclear anaplasia (i.e., large, round, and vesicular nuclei) and by perturbed glandular architecture (e.g., crowding o f acinar structures). Poorly differentiated adenocarcinomas, moreover, often show a complete absence o f basal cells with peripheral stroma invasion by ductal epithelia (Van de Voorde, 1996). Newly diagnosed prostate cancers are assessed according to histological grade and pathological/clinical stage. The histological grading system of Gleason et al. (1974) is widely used to evaluate prostate tumors. By this system, tumors are graded (i.e., I-V) based on the degree or ‘pattern’ of glandular differentiation. For example, grade I tumors are characterized by well differentiated, uniform glands with circumscribed boundaries. Grade V tumors, on the other hand, have minimal glandular differentiation with diffusely 3 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. infiltrated stroma. The sum of the Gleason grade numbers, assigned to the two most dominant histological patterns within a particular tumor, is called the Gleason score, which is a legitimate predictor of survival (Gleason et al., 1974; Sogani et al., 1985) and of tumor recurrence following radical prostatectomy (Humphrey et al., 1993). High Gleason scores are indicative of poorly differentiated, aggressive cancers. While histological grade reflects the cellular phenotype of a given tumor, pathological stage reveals its state of progression. Using the updated TNM (i.e., tumor, node, metastasis) classification system of Schroder et al. (1992), tumors are categorized according to whether they are confined to the prostate (i.e., localized disease), outside the prostatic capsule and invading locally (i.e., advanced disease), or outside the capsule and invading distant metastatic sites (i.e., metastatic disease). Tumors exhibiting both high histological grade and advanced pathological stage are associated with very poor patient prognosis (Gleason et al., 1974; Sogani et al., 1985). The natural history and androgen dependence o f prostate cancer While histological grading/pathological staging provides a ‘snapshot’ o f a tumor’s state of differentiation/progression and is a predictor of some defined endpoints (i.e., survival and recurrence), it does not reveal anything about the natural history of prostate cancer. Some localized tumors (i.e., low grade/stage), for instance, progress rapidly to metastatic disease, whereas, most others are slow-growing and never become life- threatening. Indeed, in one study it was shown that about 30% of men over the age of 50 4 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. die with occult prostatic lesions (Franks, 1954). Isaacs (1993), furthermore, estimated that 50% of men aged 70-80 years possess latent, histologically localized prostate cancers. Clearly, the primary clinical challenge today, with regard to prostate cancer management, is to identify those tumors that will progress to metastatic disease from among those that will not. One compounding problem is that tumors are often multifocal or display considerable histological heterogeneity (reviewed by Van de Voorde, 1996). Another is that some apparently ‘localized’ tumors have undetected metastatic components in distant sites. Nevertheless, clinicians must forecast the likely course of a particular tumor and prescribe an appropriate treatment. For localized prostate cancers, treatments range from aggressive (i.e., radical prostatectomy and radiotherapy) to conservative (i.e., no treatment, but rather a “wait and see” approach) (Kirkels, 1996). For advanced or metastatic disease, the principle and perhaps only efficacious treatment is hormone deprivation therapy; the suppression of testicular androgen production with estrogens or with gonadotropin-releasing hormone (Gn-RH) analogs, and perhaps the inhibition of androgen receptor (AR) function with antiandrogens (Dupont et al., 1993). Success of this treatment depends on the fact that most prostate tumors, at least initially, are dependent on androgens for growth, which was demonstrated 60 years ago when castration first was used to cause tumor regression in men with metastatic prostate cancer (Huggins et al., 1941). Hormone deprivation treatment is not a cure, however, and while 70-80% of patients experience initial tumor Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. regression, they inevitably die as a result of the outgrowth o f so-called ‘androgen- independent’ disease (Kozlowski et al., 1991; Wilson, 1995). The reasons for this transition from androgen-dependent to androgen-independent growth are not well understood, but there is considerable evidence that the AR [a ligand- activated transcription factor that transduces the growth-stimulatory effects of T on prostate cells through the transcriptional regulation o f target genes (refer to the Introduction sections o f Chapters 2 and 3 for a complete description of the AR gene and the molecular action o f the AR protein)] is culpable in some cases (reviewed by Bentel and Tilley, 1996). AR gene amplification, for example, has been demonstrated in a subset of hormone refractory tumors, potentially facilitating tumor growth in low androgen concentrations (Koivisto et al., 1997). In other androgen-independent tumors, mutations in the ligand-binding domain of the AR were found that result in broadened ligand specificity, allowing for AR activation by other steroid hormones or even by antiandrogens (e.g., Talpin et al., 1995; Fenton et al., 1997). Even in tumors harboring no AR alterations, ligand-independent activation of the AR by other signaling pathways, such as those mediated by insulin-like growth factor (Culig et al., 1994) and HER-2/neu (Craft et al., 1999), may allow for androgen independent growth. In summary, the natural history of prostate cancer is poorly understood. While some localized (low grade/stage) lesions progress over time to a more advanced (high grade/stage) state, others do not. The importance of the androgen signaling axis in disease progression is well established, though other signaling/growth pathways are 6 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. clearly at play. Elucidation o f the complex interplay between the various signaling pathways, that impact normal and malignant prostatic growth and differentiation, will be critical to the future understanding of the progression o f this obviously complex and heterogeneous disease. Genetic Predisposition to Prostate Cancer The three principle risk factors for prostate cancer are age, race-ethnicity, and family history of disease. While rarely seen in men before age 40, the rate of increase in prostate cancer incidence, as a function of age, is substantially greater than for any other cancer (Cook et al., 1969). In US Caucasians, prostate cancer incidence per 100,000 increases from 34 to 150 to 440 for men ages 60, 70, and 80 years, respectively (Kosary et al., 1995). With regard to race-ethnicity, prostate cancer incidence rates are highest in African American men worldwide (Muir et al., 1987). Rates in African Americans are 50-70% higher than in US Caucasians and are as much as 3-fold higher than rates in native Asian populations (Ross et al., 1998). While differences in population-specific incidence rates partially may be attributable to variation in diet, lifestyle, or environmental exposures, they most likely reflect a strong genetic contribution to prostate cancer risk (reviewed by Ross and Coetzee, 1996). Several studies have indicated that positive family history is an important risk factor for prostate cancer. Using a case-control approach, Steinberg et al. (1990) found, for example, that 15% of cases but only 8% of controls had a father or brother with 7 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. prostate cancer. In addition, men with an affected father or brother were twice as likely to have prostate cancer than were men with no affected first degree relatives. Risk, furthermore, was shown to increase with the number of affected family members. While men with one affected first degree relative had a 2-fold increased risk o f prostate cancer, men with 2 and 3 affected relatives had increased risks of 5 and 11-fold, respectively. These observations, which were largely confirmed in subsequent studies (e.g., Spitz et ai., 1991), indicate a strong familial component to prostate cancer etiology and provide a logical basis for investigations aimed at identifying heritable risk factors. It is useful to distinguish between the terms ‘familial’ and ‘hereditary’ when discussing the occurrence of cancer within families (Carter et al., 1993). Familial prostate cancer refers generally to the clustering of disease within families and may be attributable to a number of factors including shared environmental exposures, polygenic inheritance of multiple low risk susceptibility genes (see below), or to complex gene-gene or gene-environment interactions. Hereditary prostate cancer, on the other hand, refers specifically to cancer incidence within families that is consistent with Mendelian inheritance of a single gene. Interestingly, some segregation studies have concluded that the clustering of prostate cancer in some families is best explained by a rare (i.e., population frequency of 0.36-1.67%), highly penetrant (i.e., 63-89% of carriers will get the disease by age 85), dominant susceptibility gene which may account for 5-10% of all prostate cancers (Carter et al., 1992; Gronberg et al., 1997; Shaid et al., 1998). Studies of risk and transmission in other families, however, have found evidence consistent with an 8 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. X-linked or recessive prostate cancer susceptibility gene (Narod et al., 1995; Monroe et al., 1995). The basis of the X-linked/recessive hypothesis is that men with an affected brother are at twice the risk for developing prostate cancer than are men with an affected father. The disparate results obtained from these studies may indicate the existence of multiple susceptibility genes, each conferring high prostate cancer risk within a small subset of hereditary families. Indeed, genome-wide linkage studies have tended to support this assertion. Several linkage studies using high risk families have found evidence of putative prostate cancer susceptibility loci on chromosome 1. Smith et al. (1996), for instance, reported linkage to lq24~25 (HPC1), a region confirmed in multiple subsequent studies (e.g., Hsieh et al., 1997; Neuhausen et al., 1999), though, not in all (e.g., Mclndoe et al., 1997; Eeles et al., 1998). Linkage to a more distal region of lq (lq42.2-43; PCAP) also was observed (Berthon et al., 1998). The third chromosome 1 susceptibility locus was identified at lp36 {CAPE) by Gibbs and colleagues in a family with hereditary prostate and brain cancers (Gibbs et al., 1999). In addition to loci on chromosome I, Xu et al. (1998) found linkage to Xq27-28 (HPCX), supporting the existence o f an X-linked genetic component to hereditary prostate cancer risk. Most recently, susceptibility loci have been found on the long arm of chromosome of 20 (20ql3; HPC20) (Berry et al., 2000) and on chromosomes 1,8, 10, 12, 14, and 16 (Gibbs et al., 2000). It is important to note that none of the putative hereditary prostate cancer (HPC) genes has yet been identified. 9 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Linkage studies done within high-risk cancer families, such as those discussed here, are geared toward the identification o f highly penetrant, dominant traits that carry high absolute risk of developing the disease, but occur with low frequency in the general population. In other words, they attempt to find the gene that causes cancer in this highly selected population. Since most prostate cancers occur sporadically, and therefore, are not attributable to dominant hereditary factors, other susceptibility genes, those conferring low absolute risk but possibly high population-attributable risk, should also be considered in the genetic predisposition paradigm. These types o f genes do not cause cancer individually, but rather, contribute to a polygenic etiology o f cancer susceptibility. Such genes are generally identified based on a priori knowledge of their function within the cell type of interest (Bell, 1993). With respect to prostate cancer, genes involved in androgen biosynthesis and metabolism are obvious candidates in this regard, considering the critical role of the androgen signaling axis in prostate growth and differentiation. In the Introduction to Chapter 2 of this dissertation, the polygenic model of prostate cancer susceptibility developed by Ross and colleagues (1998) is presented with particular focus on the AR gene. Inclusion of the AR gene in this model was based on several criteria: (1) the AR protein transduces the growth stimulatory effects of androgens on the prostate, (2) AR expression is maintained in all forms of prostatic tumors (Sadi and Barrack, 1993; Ruizeveld de Winter et al., 1994), (3) the AR gene is X- linked, and therefore, individual prostate cells are under the influence of a single allele, 10 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. and (4) the AR gene is highly polymorphic in the general population, and therefore, amenable to epidemiologic studies. Thesis perspective This chapter provides a relatively general introduction to the prostate, surveying aspects of its developmental biology and malignancy. It is not meant to provide a comprehensive prostate review, but rather a context from which the principal themes of this dissertation will emerge. The primary focus of this dissertation is the AR. In Chapter 2, the AR gene is considered in a polygenic model of prostate cancer susceptibility. Using a basic case-control approach, an attempt is made to understand how AR allelic variation relates to differences in risk among different racial-ethnic populations. It is demonstrated that African American men, who are at highest risk for prostate cancer, have on average shorter AR alleles (i.e., based on an intragenic CAG microsatellite) than do lower risk Caucasians and Asians. It is further shown that Caucasian men with short AR alleles are at significantly increased risk for prostate cancer, especially for advanced disease. In Chapter 3, the effects of length variation of a homopolymeric glutamine stretch (i.e., encoded by the CAG microsatellite) on AR transcriptional activation are assessed in prostatic carcinoma cells. It is demonstrated that AR transactivation potential decreases with increasing polyglutamine size. It is further shown that increased AR polyglutamine length inhibits potentiation of AR signaling by a family of steroid receptor coactivators 11 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (i.e., the p i60 coactivators). In Chapter 4, the role of the BRCA1 tumor suppressor protein in endocrine signaling is investigated. BRCA1 is demonstrated to be a novel coactivator of the AR, its effects particularly potent in the presence of p i60 coactivator. Contrary to some reports, BRCA1 is also shown to enhance ER signaling, though only in conjunction with the p i60 coactivators. Finally, in Chapter 5, a study using the yeast two-hybrid system to screen a human testis cDNA library for putative interacting factors of the AR N-terminal transactivation domain is presented. In this largely descriptive work (i.e., Chapter 5), both physical and functional interactions of two principle candidates with the AR are demonstrated. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 12 CHAPTER 2 Androgen receptor gene allelic variation and susceptibility to prostate cancer INTRODUCTION Over the last several years a polygenic model o f prostate cancer susceptibility has been developed around some of the genes involved in androgen metabolism and signaling in the prostate (Ross et al., 1998). The rationale for pursuing such a model is grounded in the fact that the vast majority of prostate cancers are not attributable to single heritable traits that confer high absolute risk of disease development, like germ-line mutations in breast cancer susceptibility gene 1 (BRCA1) or the adenomatous polyposis coli ( .APC) gene (Isaacs et al., 1994). Instead, it appears likely that multiple susceptibility genes, each individually conferring low absolute risk but possibly high population-attributable risk, encourage cancer development in conjunction with one another, thereby establishing a polygenic etiology of disease in most cases. In this view, the susceptibility gene does not directly predispose to or ‘cause’ cancer, but rather ‘contributes’ to overall cancer risk. Thus, low- or high-risk polygenic profiles can be defined for a given individual based on consideration of allelic variation at perhaps several susceptibility loci (Ross and Coetzee, 1996). 13 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Putative prostate cancer susceptibility genes Due to the fundamental role of the androgen signaling axis in the normal growth, differentiation, and maintenance of the prostate, as well as to the fact that most early- stage prostatic tumors are dependent on androgens for growth, four relevant candidate genes were initially identified as putative susceptibility loci. These included: (1) the cytochrome p450cl7a (CYP17) gene, which encodes an enzyme critical to the regulation of T biosynthesis; (2) the steroid 5-a reductase type II (SRD5A2) gene, the product of which catalyzes the conversion of intracellular T to DHT; (3) the 3 p-hydroxy steroid dehydrogenase type II (HSD3B2) gene, which encodes an enzyme that inactivates intracellular DHT by converting it to P-androstanediol; and (4) the AR gene, which encodes the AR, a transcription factor activated by bound DHT that regulates the expression o f a repertoire of androgen-responsive genes ultimately leading to prostate epithelial cell division. In addition to their obvious physiologic importance, these candidate genes were selected because they are all polymorphic and because they show considerable allelic variation among racial-ethnic groups with disparate prostate cancer incidence and mortality rates (see Chapter 1, Introduction). In the case of the AR gene, two polymorphic markers lie within the coding sequence of exon 1 and consequently, variation in either marker results in changes in AR amino acid composition. For one of these markers, the resultant changes in the AR protein have been shown to influence its 14 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. function (see below), indicating that this marker is perhaps directly involved in modifying prostate cancer risk at the AR locus. The AR gene The AR gene, located on the long arm of the X chromosome (i.e., Xqll-12), comprises 8 exons that encode an mRNA of nearly 11 kb (Lubahn et al., 1988; 1989; Chang et al., 1988; Tilley et al., 1989; Brown et al., 1989). The AR mRNA species comprises a 2.8 kb open reading frame (ORF), a 1.1 kb 5’ untranslated region (UTR), and a 6 kb 3’ UTR. As alluded to above, exon 1 of the AR gene contains two polymorphic trinucleotide microsatellites, CAG and GGC, which code for variable-length polyglutamine (poly-Q) and polyglycine (poIy-G) tracts, respectively, in the AR protein. The normal size ranges for the CAG and GGC microsatellites are between 6-39 and 7-20 repeats with averages of about 21 and 16 repeats, respectively (Edwards et al., 1992; Giovannucci et al., 1997; Coetzee, unpublished results). Interestingly, the CAG and GGC microsatellites have expanded exponentially throughout primate evolution (Rubinsztein et al., 1995a; Choong et al., 1998). We have shown that the Old World marmoset, drill, and macaque monkeys, for example, possess only about 3, 8, and 7 uninterrupted CAG repeats, respectively (Rubinsztein et al., 1995a). The macaque and the prosimian lemur, furthermore, possess only about 6 and 2 uninterrupted GGC repeats, respectively (Choong et al., 1998). This phylogenetic microsatellite expansion may be reflective of a mutational bias in favor of longer repeat 15 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. lengths. Indeed, Rubinsztein et al. (1995b) have shown that human microsatellite repeats statistically are more likely to be longer than their primate counterparts. O f course, such a directional expansion of coding microsatellite repeats would be expected to be tolerated evolutionarily until, that is, it results in dramatic loss o f reproductive fitness (i.e., due to the functional disruption of the encoded protein). Kennedy’ s disease and aberrant AR function Expansion o f the CAG microsatellite to 40 or more repeats causes a rare, X- linked, adult onset, neurodegenerative disorder called spinal and bulbar muscular atrophy (SBMA) or Kennedy’s disease (La Spada et al., 1991; 1992). In addition to progressive muscle weakness and atrophy due to loss of brain stem and spinal cord motor neurons, men with this disorder frequently present with symptoms of partial androgen insensitivity (i.e., gynecomastia and testicular atrophy), indicative of aberrant AR function (Arbizu et al., 1983; Nagashima et al., 1988; Fischbeck, 1989). Indeed, in in vitro transfection experiments, mutant receptors encoded by SBMA AR alleles had normal androgen- binding activities but reduced transactivation activities compared to wild type AR (Mhartre et al., 1993; Chamberlain et al., 1994). In several subsequent studies, moreover, a clear inverse relationship between AR transactivation activity and CAG repeat length was established over an extended CAG size range encompassing that of normal AR alleles (Kazemi-Esfaijani et al., 1995; Tut et al., 1997; Irvine et al., 2000). Thus, AR poly-Q expansion in SBMA results in a loss of receptor function, in terms of 16 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. transactivation competency, leading to the manifestation o f partial androgen insensitivity in affected men. Accordingly, it is possible to view receptors with reduced poly-Q stretches (i.e., encoded by short CAG alleles) in ‘normal’ men as having a gain in AR function resulting in increased transactivation activity, and therefore, an apparent increased androgen sensitivity. A circumstantial link between AR CAG repeat length and prostate cancer risk Frequency distributions of AR alleles based on CAG size in U.S racial-ethnic populations were reported in 1992 as part of a large survey of genetic variation at trimeric and tetrameric tandem repeats (Edwards et al., 1992). In that study, it was found that the frequency of AR alleles with fewer than 22 CAG repeats was 65% in high-risk African- Americans compared to 53% and 34% in intermediate-risk Whites and low-risk Asians, respectively. Because the frequency distributions of arbitrarily defined ‘short’ CAG alleles correlated with the lifetime prostate cancer risks in these populations, Coetzee and Ross (1994) hypothesized that, in general, short AR CAG alleles would be enriched in men with prostate cancer compared to unaffected men because shorter AR CAG alleles encode more active receptors that promote tumorigenesis by causing increased prostate epithelial cell turnover. This hypothesis was directly tested in a small case/control study involving 68 prostate cancer patients and 123 control subjects (Irvine et al., 1995). In this study, which is described herein, a prevalence of short AR CAG alleles was observed among high-risk 17 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. African-American controls, corroborating the findings of Edwards et al. (1992). In addition, significant racial-ethnic differences were found for distributions of GGC microsatellite allele frequencies. Finally, a modest, though not significant, enrichment of short CAG alleles was observed among White prostate cancer patients. These essentially supportive findings were confirmed and extended in our follow-up study using the same prostate cancer patients and an expanded control population (Ingles et al., 1997). In this study, patients, especially those with advanced disease, had a significantly higher prevalence of short AR CAG alleles compared to control subjects. Also, in a recent collaboration with the Ingles laboratory, we further have shown that men with both short AR CAG alleles and a certain prostate-specific antigen (PSA) promoter variant are at particularly high risk of developing prostate cancer, suggestive of gene-gene interactions between these loci (Xue et al., 2000). While the epidemiologic studies presented here explore the apparent correlation between AR CAG repeat variation and prostate cancer risk, they do not address the molecular mechanisms which underlie changes in AR transactivation activity due to CAG repeat variation. In Chapter 3, a series of in vitro experiments are shown which examine the consequences of poly-Q expansion on functional interactions between the AR and the p i60 family of nuclear receptor coactivators. The results from these experiments offer the first mechanistic explanation for why increasing poly-Q length leads to blunted AR activity. 18 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. MATERIALS AND METHODS Prostate cancer patients The case patients analyzed in the present study (Irvine et al., 1995; Ingles et al., 1997; Xue et al., 2000) were taken from a 1992 pilot investigation of familial prostate cancer that was initiated among patients identified in 1991-1992 by the Cancer Surveillance Program, the population-based Surveillance, Epidemiology, and End Results (SEER) Cancer Registry of Los Angeles County. This registry has been in existence since 1972 and is estimated to be 99% complete in its goal of identifying all incident cancer cases among the approximately 9 million current residents of Los Angeles County (Berstein and Ross, 1991). Five hundred fifty-nine (53%) of the 1062 eligible patients responded to a mailed questionnaire that requested detailed information on family history of cancer. Of these 559 patients, those with one or more brothers currently residing in Los Angeles County were selected for participation. The primary goal was to enlist approximately equal numbers of these patients either negative or positive for a family history of prostate cancer. A sample of venous blood in EDTA was collected from all consenting patients and their brothers. The data presented in this report derive from the first 68 incident prostate cancer patients diagnosed in 1991-1992 who were recruited for the pilot study. O f these, 57 were non-Hispanic Whites (Whites), 9 were African-Americans, one was Egyptian, and one was Chinese-American. Twelve (18%) of these 68 patients had one or more first 19 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. degree relatives (father, brother, and/or son) with prostate cancer. The age at diagnosis of the patients ranged form 51-68 years with a mean age o f 57.7 years. Thirty-five patients had disease localized to the prostate gland, 32 had so-called advanced disease [i.e., with tumor invading and extending beyond the prostatic capsule and/or involving regional lymph nodes or distant metastatic sites (SEER 1995 clinical and pathological extent of disease codes 41-85)], and one had disease of unknown stage. Buffy coats of lymphocytes were prepared from the blood samples and stored at -70° C until analysis. Control subjects The control subjects analyzed in the present study were selected from a cross- sectional survey initiated in 1991 of apparently healthy adult males from several racial- ethnic groups in Los Angeles (Yu et al., 1994). Subjects were identified from a comprehensive list of registered drivers residing in Los Angeles County who were over the age of 35 years. One hundred thirty-three such men were recruited into the survey and venous blood samples were collected from 123 of them [39 non-Hispanic Whites, 45 African-Americans, and 39 Asians (i.e., Chinese and Japanese)]. This group comprised the control subjects for our original study (i.e., Irvine et al., 1995). For purposes o f the original Yu et al.(1994) study, 65 (53%) of these controls were lifelong nonsmokers, while the remaining 58 were current smokers of varying intensity. As with the cancer patients, buffy coats of lymphocytes prepared from the blood samples were stored at -70° C until analysis. 20 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. It is important to note that in our follow-up studies o f AR CAG repeat variation and its relation to prostate cancer risk (i.e., Ingles et al., 1997; Xue et al., 2000), an expanded non-Hispanic White control group was used to analyze the same 57 non- Hispanic White case patients described above. Thus, in addition to the original 39, another 132 non-Hispanic White male control subjects were identified based on their participation in the Yu et al. (1994) study. The control subjects were selected to be comparable to the prostate cancer patients in age and in socioeconomic status. The mean age of the control subjects was 58.2 years. O f these 171 controls, either 2 (Ingles et al., 1997) or 13 (Xue et al., 2000) were excluded because of PCR amplification failure or sample DNA depletion, leaving 169 or 156 control subjects in the respective studies. Written consent was obtained from all subjects. Microsatellite and polymorphism analyses DNA extraction DNA was extracted from blood lymphocytes according to the rapid extraction protocol o f Talmud et al. (1991). In brief, 100 pi thawed blood were mixed with 400 pi freshly prepared 0.17M NH4 C1 and incubated at room temperature for 20 min. Next, the sample was centrifuged for 30 s in a microfuge, the supernatant was discarded, and the pellet was washed extensively in 0.9% NaCl with vortexing. The washed pellet was resuspended in 200 pi 0.05M NaOH and boiled for 10 min. Finally, 25 pi 1M Tris (pH 8.0) were added to neutralize the NaOH and the DNA sample was stored at -20° C. 21 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CAG repeat size determination The exon 1 CAG repeat of the AR was amplified from about 10 ng o f undigested genomic DNA in two rounds of PCR using Taq DNA polymerase (Amersham-Pharmacia Biotech, Piscataway, NJ) and nested primers (Tsai et al., 1995). The outside or first round PCR primers were 5 ’ -GTGCGCGAAGTGATCCAGAA-3 ’ and 5 ’ -TCTGGGACG- CAACCTCTCTC-3’. The inside or second round PCR primers were 5’-AGAGGCC- GCGAGCGC AGC ACCTC-3 ’ and 5’-GCTGTGAAGGTTGCTGTTCCTCAT-3\ The first round o f PCR consisted of 17 cycles of the following steps: 94° C for 1 min; 55° C for 1 min; and 72° C for 1.5 min. The second round, performed with 1 pi of the first round reaction and 2 pCi [a-3 2 P]dCTP, consisted of 28 cycles of the following steps: 94° C for 1 min; 66° C for 1 min; and 72° C for 1.5 min. Second round PCR products were separated on 5% denaturing polyacrylamide gels that were subsequently exposed to X- OMAT film (Eastman Kodak, Rochester, NY). The number o f CAG repeats was determined by comparing the size o f the predominant PCR product (i.e., the middle radiographic band, Fig. 2.1 A) to a series of previously sequenced CAG size standards. All unknown samples from a particular autoradiogram were then ranked according to relative size and rerun a second time to confirm CAG sizes. GGC repeat size determination The exon 1 GGC repeat of the AR was amplified from about 10 ng of undigested genomic DNA in two rounds of PCR using nested primers. Due to the high G + C 22 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. content of this repeat, it was amplified with the more thermostable Pfu DNA polymerase (Stratagene, La Jolla, CA). The outside or first round PCR primers were 5’- T CCT GGC AC ACT CT CTT C AC-3 ’ and 5 ’-GCC AGGGTACC AC AC ATC AGGT-3 ’. The inside or second round PCR primers were 5-ACTCTCTTCACAGCCGAA- GAAGGC-3 ’ and 5 ’ - ATCAGGTGCGGTGAAGTCTCTTTCC-3 \ The first round of PCR consisted o f 17 cycles of the following steps: 98° C for 1 min and 70° C for 5 min. The second, performed with lpl of the first round reaction and 2 pCi of [a-3 2 P]dCTP, consisted of 34 cycles of the same steps used in round one. Second round PCR products were analyzed as described above for the CAG repeat. The number o f GGC repeats was determined by comparing the predominant PCR product (i.e., the top radiographic band, Fig. 2. IB) with a series of previously sequenced GGC size standards. As with the CAG repeat, all unknown samples from a particular autoradiogram were then ranked according to relative size and rerun a second time to confirm GGC sizes. PSA allele determination Alleles o f the G/A polymorphism at position -158 in the promoter region of the PSA gene can be distinguished by cutting with the Nhe I restriction enzyme. The polymorphic site was amplified from about 40 ng of undigested genomic DNA using primers 5 ’ -TTGTATGAAGAATCGGGGATCGT-3 ’ and 5-TCCCCCAGGAGCCCTAT- AAAA-3’. The PCR reaction consisted o f 35 cycles of the following steps: 94° C for 1 min; 59° C for 1 min; and 72° C for 40 sec. An initial melting step (94° C for 10 min) and a final polymerization step (72° C for 10 min) were also included. A 7/50 pi aliquot 23 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. was digested with 1.5 u o f Nhe I enzyme (New England Biolabs Inc., Beverly, MA) for 4 h at 37° C and then separated on 2.5% agarose gel. The three possible genotypes were defined by distinct banding patterns: AA (300 bp), AG (150, 300 bp) and GG (150 bp). Statistical Analyses The X 2 test o f association was used to examine the relationship o f CAG and GGC repeats by race, as well as the relationship between CAG and GGC repeats among prostate cancer subjects and controls separately. The binomial test was used to examine the difference in prevalence of specific CAG/GGC alleles between White prostate cancer subjects and White controls. All P values quoted are two-sided unless otherwise stated. Where indicated, odds ratios (ORs) and confidence intervals (CIs) for subgroups of case patients were estimated by use o f exact distributions with a mid-P correction (Hiiji et al., 1991) as implemented by the program StatXact Turbo 2.04 (Cytel Software Corp., Cambridge, MA). Because the results of significance tests may depend on the choice of cut points for a continuous variable (Altman et al., 1994) and because cut points were arbitrarily chosen in the case o f the CAG polymorphism, differences among the CAG allele size distributions were also tested for significance by use of the Wilcoxon test (Lehmann, 1975). A test of interaction between the AR and PSA genes was performed by adding an interaction term to the unconditional logistic regression model and computing the likelihood ratio statistic (Breslow and Day, 1980). 24 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. RESULTS AR CAG and GGC allele frequencies differ significantly among racial-ethnic groups The distribution of CAG repeat allele frequencies by racial-ethnic grouping among the control subjects is presented in Table 2.1. Using the median CAG repeat size in Asians (i.e., 22 repeats) as an arbitrary cutoff, 33 (75%) of 44 African-Americans compared with 24 (62%) of 39 Whites and only 19 (49%) of 39 Asians had so-called ‘short’ alleles (i.e., <22 repeats). These differences were statistically significant [P (difference in distribution by three racial-ethnic groups: <22 versus >22 CAG repeats) = 0.046], The distribution of GGC repeat allele frequencies was also significantly different among the racial-ethnic groupings of the control subjects as presented in Table 2.2 [P (difference in distribution by the three racial-ethnic groups: 16 versus not-16 GGC repeats) <0.0005]. Only 8 (20%) of 41 African-Americans had 16 GGC repeats compared to 21 (57%) of 37 Whites and 26 (70%) of 37 Asians. The prevalence of short GGC alleles (i.e., <16 repeats) was especially high in African-Americans with a frequency of 61% (25 subjects) compared to 11% (4 subjects) in Whites and 27% (10 subjects) in Asians. The prevalence of GGC alleles longer than 16 repeats was very low in Asians (i.e., 3%), while in Whites and African-Americans it was relatively higher (32% and 20%, respectively). 25 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Prevalence o f a high-risk CAG/GGC allelotype in White prostate cancer patients Among the racial-ethnic groups represented in this study, African-Americans are at highest risk for developing prostate cancer. Assuming that this risk is attributable, at least in part, to variation at the AR locus, so-called ‘high-risk’ alleles o f the CAG and GGC microsatellites (i.e., <22 CAG repeats; not-16 GGC repeats) were defined on the basis of their frequency distribution in African-American control subjects (Tables 2.1 and 2.2). The prevalence of high-risk alleles in White prostate cancer patients and in White control subjects is presented in Table 2.3. There were modest, though nonsignificant, excesses of short CAG alleles and not-16 GGC alleles in case patients compared to control subjects conferring relative risks of 1.25 and 1.18, respectively. When both the CAG and GGC high-risk alleles were considered simultaneously, there was a 2.1-fold increased risk for prostate cancer relative to other allelotypes, however, the association failed to reach statistical significance (one-sided P = 0.08). Among control subjects, Whites and Asians had intermediate (9/37 = 24%) and low (3/34 = 9%) prevalence, respectively, of the putative high-risk allelotype compared to African Americans (25/39 = 64%) and the differences were statistically significant (P <0.0005). The CAG and GGC microsatellites o f the AR are in linkage disequilibrium in men with prostate cancer A test for linkage disequilibrium between CAG and GGC repeats in all control subjects was negative (P = 0.53). Furthermore, no association between the two 26 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. microsatellites was observed in any one of the racial-ethnic subgroups when assessed separately. In contrast, the distribution of CAG repeats was highly correlated with that of GGC repeats among all (i.e., n = 68) prostate cancer patients (P = 0.008, Table 2.4). There was a close association between short CAG alleles (i.e., <22 repeats) and long GGC alleles (i.e., >16) such that of the 47 patients with short CAG alleles, 20 (43%) had long GGC alleles. In contrast, of those 20 patients with long CAG alleles (i.e., >22 repeats), only 3 (15%) had long GGC alleles. When the analysis was limited to White prostate cancer patients only, similar results were obtained though due to the reduced sample size (n = 57), the P value for the X 2 test o f association between CAG and GGC repeats was slightly higher at 0.015. Short AR CAG alleles are associated with increased risk fo r advanced prostate cancer Of the 57 White prostate cancer patients analyzed in this study, 31 (54%) had disease localized to the prostate, while 26 (46%) had advanced disease (as defined in Materials and Methods). When case patients were considered as a single group, short CAG alleles (i.e., <20 repeats) conferred a 2-fold increased risk of prostate cancer (Table 2.5). Moreover, when case patients were stratified by stage of disease, the short CAG alleles were more strongly associated with advanced disease than with localized disease. In fact, the genotype-disease associations were statistically significant for advanced but not for localized disease. Because cut points for the CAG polymorphism were arbitrarily chosen, these results were verified by comparing CAG allele size distributions by use of a 27 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. nonparametric test that did not depend on cut points. Compared with values for control subjects, the AR allele size distribution was found to be shifted toward smaller values for patients with advanced disease (P = 0.05) but not for patients with localized disease (P = 0.73) (Fig.2.2). Interaction between genotypes at the AR and PSA loci The PSA genotype frequencies among White control subjects were 25% AA, 49%, AG, and 26% GG, consistent with the expected Hardy-Weinberg equilibrium frequencies. Among White case patients, there was a nonsignificant excess of the GG genotype (data not shown). Subjects with the GG genotype were at significantly increased risk for advanced, but not for localized prostate cancer [OR = 2.90 (95% Cl = 1.24-6.78) versus 0.85 (95% Cl = 0.34-2.11)]. Subjects were cross classified by PSA and AR genotypes in Table 2.6. Those having both AR CAG long alleles and PSA AA or AG (i.e., not GG) alleles were selected as the referent group. Subjects with either AR CAG short (and not PSA GG) or PSA GG (and not AR CAG short) alleles had modest, though nonsignificant increases in prostate cancer risk overall relative to the referent group. In combination, however, a short CAG allele and a PSA GG allele conferred more than 5-fold increased risk. The statistical test for interaction was not statistically significant (p = 0.28), though the power for testing interaction in this data set was small. For case patients with advanced disease, all odds ratios were substantially greater. A statistically significant increase in risk was conferred 28 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. by short AR CAG alleles and a PSA GG allele, either alone or in combination; however, odds ratios were imprecisely estimated due to the small number of advanced cases. 29 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 2.1 A 21 21 24 18 20 25 18 23 19 13 AR CAG repeat size 19 16 17 15 16 15 17 15 14 10 16 AR GGC repeat size Autoradiograms of AR CAG (A) and GGC (B) microsatellite PCR products. The relative size o f each repeat is indicated. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. 100 n 90- | 70- a > a 60- < & 50- a > £ 40- a 1 30‘ J 20- 10- controls localized PCa advanced PCa 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 AR CAG microsatellite (no. of repeats) Cumulative distribution of AR CAG allele sizes among White control subjects and White prostate cancer cases with advanced or localized disease. FIGURE 2.2 TABLE 2.1* Distributions o f AR CAG microsatellite allele frequencies (%) among control subjects according to racial-ethnic grouping No. of CAG repeats African-American (n = 44)§ White (n — 39) Asian (* = 39) 9 2.3 0 0 10 0 0 0 11 0 0 0 12 2.3 0 0 13 6.8 0 0 14 2.3 0 0 15 6.8 0 2.6 16 9.1 0 0 17 13.6 5.1 0 18 4.5 12.8 7.7 19 6.8 12.8 10.3 20 11.4 23.1 12.7 21 9.1 7.7 15.4 22 13.6 5.1 15.4 23 6.8 12.8 5.1 24 0 12.8 10.3 25 2.3 2.6 5.1 26 0 2.6 7.7 27 2.3 2.6 5.1 28 0 0 0 29 0 0 2.6 Total 100 100 100 * Modified from Irvine et al., 1995. + Two-sided P (difference in distribution by the three racial-ethnic groups: <22 versus >22 repeats) = 0.046. § One sample was noninformative, therefore n = 122 for the analysis. 32 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TABLE 2.2* Distributions o f AR GGC microsatellite allele frequencies (%) among control subjects according to racial-ethnic grouping’ ’ No. of GGC repeats African-American (« = 41)§ White (« = 37)§ Asian (n = 37)§ 8 2.4 0 0 9 0 0 0 10 4.9 2.7 2.7 11 4.9 0 2.7 12 9.8 0 5.4 13 4.8 0 2.7 14 12.2 0 2.7 15 22.0 8.1 10.8 16 19.5 56.8 70.3 17 19.5 32.4 2.7 Total 100 100 100 * Modified from Irvine et aI., 1995. + Two-sided P (difference in distribution by the three racial-ethnic groups: 16 versus not- 16 repeats) <0.0005. § Some samples were noninformative, therefore n = 115 for the analysis. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TABLE 2.3* Allelic distributions o f the AR CAG and GGC microsatellites in white prostate cancer patients and control subjects No. of controls No. o f cases Relative Risk No. of CAG repeats >22 19 15 1.00 <22 38 24 1.25 No. of GGC repeats 16 30 21 1.00 not-16 27 16 1.18 No. of CAG/No. of GGC other 34 28 1.00 <22/not-16 23 9 2.10t * Modified from Irvine et al., 1995. + One-sided P = 0.08. 34 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TABLE 2.4* Distributions o f AR CAG and GGC microsatellite allele frequencies (%) in prostate cancer patients'* No. of GGC§ No. of CAG <16 16 >16 total <22 11 (23%) 16 (34%) 20 (43%) 47 (100%) >22 2 (10%) 15 (75%) 3 (15%) 20 (100%) * Modified from Irvine et al., 1995. f Two-sided P (test of association between no. of CAG and no. of GGC repeats) = 0.008. ^ One sample was noninformative for the GGC assay, therefore n = 67 for the analysis. 35 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. TABLE 2.5* Frequency o f ARC AG alleles among patients with localized and advanced disease compared to control subjects L o n g (%) AR CAG alleles^ Short (%) Total (%) Odds ratio (95% confidence interval) Control subjects 124 (73) 45(27) 169(100) 1.00 (referent) All case patients 33 (58) 24 (42) 57(100) 2.00(1.07-3.75) Patients with localized disease 19(61) 12(39) 31 (100) 1.74 (0.78-3.87) Patients with advanced disease 14(54) 12(46) 26(100) 2.36(1.02-5.49) * Modified from Ingles et al., 1997. * AR CAG allele size: short = <20 repeats; long = >20 repeats. U> O n Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. TABLE 2.6* Allelic distributions o f the AR CAG microsatellite and PSA A/G promoter polymorphism in prostate cancer patients and controls Genotype No. of No. of No. of AR CAGf PSA controls (%) cases (%) RR (95% Cl) advanced (%) Odds ratio (95% Cl) Long not GG 80 (51) 21 (37) 1.00 (referent) 5(19) 1.00 (referent) Long GG 34 (22) 12(21) 1.34 (0.60-3.04) 9(35) 4.24(1.32-13.57) Short not GG 36 (23) 16(28) 1.69 (0.79-3.62) 8(31) 3.56(1.09-11.62) Short GG 6 (4 ) 8(14) 5.08(1.59-16.24) 4(15) 10.67 (2.25-50.49) * Modified from Xue et al., 2000. * AR CAG allele size: short = <20 repeats; long = >20 repeats. TABLE 2.7* Odds ratio ofprostate cancer fo r a CAG microsatellite repeat length decrement o f six in the AR gene among men in the Physicians ’ Health Study Prostate Cancer Cases Odds ratio* (six decrement in CAG) P value Total 587 1.28 0.04 High grade/stage§ 269 1.64 0.001 Low grade/stage 309 1.02 0.86 High grade 210 1.59 0.007 Advanced stage 180 1.75 0.002 Metastatic (distant) 56 2.44 0.004 Fatal 43 2.08 0.04 * Modified from Giovannucci et al., 1997. + Odds ratio is calculated by modeling CAG as a continuous variable in an unconditional logistic model and computing the odds ratio for a six CAG decrement (difference from median of low to median of high tertile of CAG repeat length). ^ Includes tumors with Gleason grade >7 or high grade or advanced stage (C or D). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. TA BLE 2.8 Summary of studies evaluating the roles o f the AR CAG and/or GGC microsatellites in prostate cancer risk, progression, and age at onset AR CAG repeat correlation with PCa AR GGC repeat correlation with PCa Study Subjects risk stage/grade age at onset risk stage/grade age at onset Irvine et al. (1995) US Caucasian yes* N/A N/A yes* N/A N/A Hardy et al. (1996) US Caucasian N/A no yes N/A N/A N/A Ingles e ta l. (1997) US Caucasian yes yes N/A N/A N/A N/A G iovannucci et al. (1997) US Caucasian yes yes no N/A N/A N/A Stanford et al. (1997) US Caucasian yes* no yes yes* no yes Hakimi et al. (1997) US Caucasian yes yes no yes 1 1 0 no Ekman et al. (1999) Swedish White yes N/A N/A N/A N/A N/A Correa-Cerro et al. (1999) French/German White no no no no no no Edwards et al. (1999) British Caucasian no no N/A no no N/A Bratt et al. (1999) Swedish White no yes yes N/A N/A N/A N/A = not applicable * < 22/not-16 GGC allelotype (RO = 2.10). 1 <22/< 16 GGC allelotype (RO = 2.05), DISCUSSION In order to assess the putative role of the AR in prostate cancer risk, two polymorphic microsatellites in the AR gene were analyzed in prostate cancer patients and in healthy normal men. Based largely on circumstantial evidence, it was predicted that short AR CAG alleles would be found at higher frequencies in the case patients than in the control subjects. No predictions were made with regard to GGC allele frequencies since no previous study had investigated racial-ethnic variation at this microsatellite nor associated it with any human disease. Indeed, this report represents the first formal analysis of prostate cancer risk as it relates to allelic variation at the AR locus. The AR CAG microsatellite and prostate cancer risk Among the three racial-ethnic control groups analyzed in this study, a significant difference in the prevalence of short CAG alleles (i.e., <22 repeats) was observed. Short alleles were most common among African-Americans and least common among Asians who are at highest and lowest risk for prostate cancer, respectively. These findings are entirely consistent with those of Edwards et al. (1992), affirming the circumstantial link between racial-ethnic risk and AR CAG size variation. Initial assessment of the prevalence of short CAG alleles in White case patients (n = 57) and in White control subjects (n = 39), revealed a modest though nonsignificant excess among the patients (Table 2.3). When, however, an expanded control group (« = 169) was used in the analysis, short CAG alleles were found to be significantly enriched in patients, conferring 40 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. a 2-fold increase in risk (Table 2.5). Interestingly, this genotype-risk association was strongest among patients with advanced disease. It is important to note that in the latter analysis, a cutpoint of 20 instead o f 22 CAG repeats was used to define the alleles as short or long (i.e., short = <20 repeats; long = >20 repeats). This adjustment was made because no difference in risk was observed among patients with 20-22 CAG repeats (Ingles et al., 1997). While it was desirable for the purposes of this analysis to dichotomize the continuous variable of CAG repeats into the two arbitrary categories of short and long, this approach suffers from well-known limitations, including information loss that potentially could result in a failure to detect a real association (Altman et al., 1994). The results of such analyses, moreover, can vary if different outpoints are used. To address this important issue, an additional statistical test was done in which CAG repeats were considered as a continuous variable. The results of this test were confirmatory, showing that the distribution of AR alleles was significantly shifted toward smaller CAG sizes in patients with advanced disease (Fig. 2.2). Thus, the data reported here are in support of the hypothesis that short AR CAG alleles confer risk o f developing prostate cancer, particularly of advanced disease. In recent years, several additional studies evaluating the role of the AR CAG microsatellite in prostate cancer risk have been published (summarized in Table 2.8). Most of these studies have more or less confirmed and in some cases extended the conclusions drawn from the work presented in this chapter. Perhaps, the most profound 41 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. o f these studies was that conducted by Giovannucci et al. (1997), in which 587 prostate cancer cases and 588 control subjects were analyzed. The study population was selected from the Physician’s Health Study, a predominantly Caucasian (i.e., 95%) cohort of 22,071 U.S. male physicians, aged 40-84 years in 1982. Due to the large sample size of this study, it was possible to stratify the case patients into several groups based on cancer stage and grade. When CAG repeats were analyzed as a continuous variable, the authors observed a highly significant inverse correlation between CAG repeat length and prostate cancer risk (Table 2.7). Consistent with our results, short CAG alleles conferred risk of advanced disease (i.e., high grade and/or stage), but not of localized disease (i.e., low grade and/or stage). Indeed, according to the data presented in the Giovannucci et al. (1997) study, a man with a CAG repeat length of 19 is nearly 2.5 times more likely to develop metastatic prostate cancer, but no more likely to develop low grade/stage cancer, than a man with a repeat length o f 25. In a different study, Stanford et al. (1997) analyzed CAG and GGC repeat lengths in 301 Caucasian men with prostate cancer and in 277 like controls. Only a slight excess o f short CAG alleles (i.e., <22 repeats) was found among the case patients compared to control subjects. Nevertheless, an overall age-adjusted relative odds of prostate cancer (associated with the number of CAG repeats as a continuous variable) of 0.97 was observed, consistent with a 3% decrease in risk with each additional CAG repeat. Contrary to our findings and those of Giovannucci et al. (1997), men in this study with short CAG alleles had similar risks of developing advanced and localized disease. 42 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In yet another study of Caucasian men, CAG and GGC allele sizes were measured in 59 prostate cancer cases and in 370 control subjects (Hakimi et al., 1997). Six of 59 (10%) case patients had short CAG alleles (i.e., <18 repeats) compared with only 11 of 370 (3%) controls, and the difference was statistically significant (P = 0.02). Interestingly, 5 of 6 (83%) of the prostate cancer patients with short CAG alleles had advanced disease (defined as lymph node positive cancer). Thus, in 4 of 4 independent case/control studies of US Caucasian men, short CAG alleles were shown to confer significant risk of developing prostate cancer. In 3 of 4 of these studies, this genotype- risk association was largely, if not completely, associated with advanced disease. It is interesting to note, however, that recent investigations among men from populations other than US Caucasian, generally have found no correlation between short AR CAG alleles and prostate cancer risk. Except for Ekman et al. (1999), who found that CAG alleles were significantly shorter among Swedish prostate cancer patients compared to controls, no significant correlation between CAG repeat length and prostate cancer risk was found among French/German (Correa-Cerro et al., 1999), British Caucasian (Edwards et al., 1999), or a different cohort of Swedish men (Bratt et al., 1999). Only in two of these studies, moreover, were prostate cancer cases stratified by grade and stage. While Bratt et al. (1999) observed “ ...trends toward associations between short CAG repeats and high grade (P = 0.07) and high stage (P = 0.07) disease,” no such trends were found in the Edwards et al. (1999) study. Thus, despite the existence of compelling 43 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. evidence among US Caucasians, short AR CAG alleles may not confer risk of prostate cancer in other racial-ethnic populations. The AR GGC microsatellite and the prostate cancer risk allelotype Allelic distributions of the GGC microsatellite were also significantly different among the racial-ethnic groups analyzed in our study. The 16-GGC repeat allele was least prevalent in high-risk African-Americans (i.e., 20%) and most prevalent in low-risk Asians (i.e., 70%). This is suggestive of a putative protective role for this allele in prostate cancer risk. It is possible that the 16-GGC allele encodes an AR with an ‘optimal’ polyglycine tract, one that allows for normal receptor function in prostatic epithelial cell growth and maintenance. O f course, this is speculative since nothing is known about how variation in poly-G length modulates AR activity. Nevertheless, a slight though nonsignificant paucity o f the 16-GGC allele was observed among White prostate cancer patients compared to control subjects. In other words, there was an enrichment of putative risk alleles (i.e., not-16 GGC alleles) among the patients. Because the AR gene is X-linked, each male inheriting a single maternal copy, it was possible to define a putative AR prostate cancer risk allelotype of short CAG repeats (i.e., <22) and not-16 GGC repeats. As expected, distributions o f this allelotype were significantly different among control subjects, with African-Americans and Asians having the highest and lowest prevalence, respectively. Among White prostate cancer 44 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. patients, the <22 CAG/not-16 GGC allelotype conferred a 2-fold increase in risk, though statistical significance was not reached. Interestingly, among prostate cancer patients, a nonrandom distribution of CAG and GGC alleles was observed, such that 66% of patients with short CAG alleles also had not-16 GGC alleles. This was in comparison to only 25% of patients with long CAG alleles. Due to the fact that the two microsatellites are in close proximity at the AR locus, it was not particularly surprising to find evidence of linkage disequilibrium between the intragenic markers. When, however, linkage was assessed among the controls (i.e., all together or by race), the two microsatellites were found to be completely equilibrated. This indicates that in normal men, either one or both of the microsatellites are hypermutable, resulting in a random distribution CAG and GGC alleles at the AR locus. Indeed, when the rate of mutation at the CAG microsatellite was measured using single cell assays o f sperm, an exceptionally high rate of 1-4% was observed (Zhang et al., 1994). Our data, therefore, suggest that unlike in the general population, a nonrandom subset of AR alleles occurs in men with prostate cancer. AR GGC repeats were also analyzed in the Stanford et al. (1997) and Hakimi et al. (1997) studies o f prostate cancer risk in Caucasian men. In each case, GGC alleles were dichotomized into arbitrary short and long categories for analysis. In the Stanford et al. (1997) study, a significant enrichment of short GGC alleles (i.e., 16 or fewer repeats) was observed among men with prostate cancer compared to control subjects. When the putative high-risk allelotype o f short CAG repeats and short GGC repeats (i.e., <22 CAG 45 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. and <17 GGC repeats) was assessed, it conferred a significant 2-fold increase in risk. Contrary to our findings, no evidence o f linkage disequilibrium between the two microsatellites was found among the prostate cancer patients. In the Hakimi et al. (1997) study, men with prostate cancer had a 4-fold higher frequency of short GGC alleles (i.e., <15 repeats). As in the Stanford et al. (1997) study, the CAG and GGC microsatellites were not linked in men with prostate cancer. Interestingly, as with the CAG repeat, analyses in populations other than US Caucasian failed to correlate GGC repeat variation with prostate cancer risk (Correa-Cerro et al., 1999; Edwards et al., 1999). Parsimonious conclusions regarding AR microsatellite variation and prostate cancer risk While some notable differences are apparent, the primary conclusions drawn from the various studies described herein essentially are the same. Firstly, it is clear that shorter AR CAG repeats are associated with prostate cancer risk, at least in US Caucasian men. All studies report an increase in risk associated with arbitrarily defined short CAG alleles. Furthermore, due to the statistical power of the Giovannucci et al. (1997) study, this ‘CAG effect’ was shown to be a linear one, with risk increasing inversely with CAG size. Stanford et al. (1997) confirmed this by showing that each additional CAG repeat was associated with a 3% decrease in prostate cancer risk. Because very long AR CAG alleles cause Kennedy’s disease with its associated partial androgen insensitivity and because AR transactivation competency increases with decreasing poly-Q size in vitro, it seems very likely that changes in AR CAG length lead 46 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. directly to changes in AR protein function that in turn, predispose the prostatic epithelium to cancer development. Despite this logic, it is possible that the CAG microsatellite is in linkage disequilibrium with other disease-causing mutations in the AR gene itself or with other nearby genes that effect prostate cancer risk. Indeed, linkage between the CAG and GGC microsatellites was observed among case patients in our study. This observation was not confirmed in the larger studies of Stanford et al. (1997) and Hakimi et al. (1997), however, but this may be due to differences in patient populations. The second major conclusion to be gleaned from these studies is that short CAG repeats primarily are associated with advanced prostate cancer in US Caucasian men. With the exception o f the Stanford et al. (1997) study, which observed no differences in risk based on the aggressiveness of disease, the prevalence of short CAG alleles was significantly different from control subjects only among men with “advanced” (Ingles et al., 1997; Table 2.5), “lymph node positive” (Hakimi et al., 1997), and “high grade/stage” (Giovannucci et al., 1997) prostate cancer. This conclusion potentially has profound implications for treating men with prostate cancer since presently, there are no reliable markers for advanced disease. Measurement of CAG size may prove to be valuable in identifying that subset of men whose cancers will become life-threatening due to metastases. The third conclusion, though perhaps less well established, is that short GGC alleles confer risk of prostate cancer independently o f the CAG microsatellite among US Caucasian men. In both the Stanford et al. (1997) and Hakimi et al. (1997) studies, short 47 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. GGC alleles were significantly enriched in men with prostate cancer. In our study, the so-called not-16 GGC alleles, which include short alleles, were associated with risk. So, as with the poly-Q stretch, changes in poly-G length, due to GGC repeat variation, also may alter AR function and encourage epithelial cell transformation. To date, however, no study has systematically evaluated the relationship between poly-G length and AR transactivation activity in vitro. Interestingly, unlike short CAG alleles, short GGC alleles conferred similar risks of localized and advanced disease (Stanford et al., 1997; Hakimi et al., 1997). This may indicate that short AR poly-G stretches predispose to a different type of prostate cancer than short poly-Q stretches. It should be mentioned that changes in AR poly-G length may not be due to variation only in the GGC repeat, but also to variation in a flanking (GGT)2 repeat (Lumbroso et al., 1997). While appreciation of cryptic changes in this GGT sequence may modestly affect the risk associations discussed here, it is unlikely to alter the overall interpretation. After all, what probably will prove to be important is the absolute number o f glycine residues in the poly-G stretch, not which codon directs their synthesis. Nevertheless, it is possible that a distinct subset o f GGN alleles confers relatively high risk of prostate cancer, though further studies will be required to explore this possibility. Age at onset o f prostate cancer and the AR CAG microsatellite Whether or not short AR CAG alleles are associated with earlier age at onset of prostate cancer is a point o f some contention. In a 1996 study of 109 mostly Caucasian 48 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. prostate cancer patients (i.e., 104 Caucasian patients), Hardy and colleagues observed a statistically significant positive association between increased age at diagnosis and longer CAG repeats [P (CAG repeat length as a continuous variable) = 0.0156]. Consistent with this finding, Stanford et al. (1997) reported an elevated risk among patients under the age o f 60 associated with short CAG alleles. Likewise, Bratt et al. (1999) reported a statistically significant correlation between short CAG repeats and younger age at diagnosis among Swedish patients. In contrast to these studies, however, Giovannucci et al. (1997) and Hakimi et al. (1997) failed to reveal any link between age at onset and CAG repeat length. No correlation was found among French/German patients either (Correa-Cerro et al., 1999). Thus, while the role of the AR CAG repeat in the onset of prostate cancer is by no means defined, it may be that short CAG alleles modify (i.e., increase) the penetrance of other susceptibility traits in some populations. Evidence o f gene-gene interactions between the AR and PSA loci in conferring prostate cancer risk As mentioned previously, most prostate cancers are likely to have a polygenic etiology. Accordingly, no single gene is sufficient to produce a dominant Mendelian pattern o f disease inheritance. Instead, perhaps several ‘susceptibility’ genes influence disease risk, individually or in conjunction with one another (i.e., through gene-gene interactions). While our initial multigenic model of prostate cancer susceptibility involved four genes particularly relevant to androgen metabolism and signaling (Ross et 49 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. al., 1998), other genes, especially those regulated by liganded AR, are also potentially very important. One such gene encodes PSA, a secreted protease that targets the major insulin-like growth factor (IGF) binding protein, IGFBP-3 (Cohen et al., 1994). PSA- mediated cleavage of IGFBP-3 increases IGF-1 and IGF-II bioavailability and elevated serum IGF-1 has been associated with prostate cancer risk (Chan et al., 1998; Cohen et al., 1998). The promoter of the PSA gene contains three androgen response elements (AREs), the most proximal o f which, ARE1, has two allelic variants: AGAACAnnnAGTA/GCT (Rao and Cramer, 1999). We hypothesized that the AR might interact differently with the two ARE1 allelic variants, resulting in quantitatively differential regulation of PSA gene expression. This, in turn, could lead to differential prostate cancer risk. Analysis of White prostate cancer cases and controls revealed a significant enrichment of PSA GG alleles in patients with advanced but not localized disease (data not shown). Men, furthermore, possessing both PSA GG and AR CAG short alleles had particularly elevated prostate cancer risk compared to those with neither risk allele (i.e., with PSA not GG and AR CAG long alleles, Table 2.6). Among patients with advanced disease, this genotype-risk association was even more pronounced. Thus, while both PSA GG and AR CAG short alleles individually confer modest risk for prostate cancer, together they confer higher risk than that predicted by their combined or additive effects. In other words, the data presented here are indicative of synergistic interaction between the AR and PSA loci. Clearly, the implied functional relationship between PSA gene 50 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. promoter variation and AR transactivation activity needs to be investigated in vitro; nevertheless, our results suggest that the androgen-mediated etiologic pathway of prostate cancer susceptibility may act in conjunction with the IGF signaling pathway. Because this interaction seems particularly relevant to advanced disease, it may be that in the restricted androgen environment of advanced disease, the influence of other growth factors becomes critical. It must be stated that the AR/PS A study was primarily the work o f Dr. Sue Ingles and her doctoral student, Wendy Xue. I have included a detailed description of it in this thesis because it is directly relevant to my work on the AR and has profound implications with respect to prostate cancer risk. In no way do I wish to exaggerate my role in revealing these important observations. AR microsatellite variation and correlation with other human disorders In light o f the relatively large normal size ranges of the AR CAG and GGC microsatellites, it is clear that the AR protein more or less can structurally and functionally accommodate great variation in poly-Q and poly-G lengths. Despite the dynamic nature o f this protein, however, it is becoming increasingly obvious that some CAG and GGC alleles within the ‘normal’ size range are in fact deleterious. For example, short CAG repeats, in addition to conferring risk for prostate cancer development and perhaps for increased age at onset (i.e., Hardy et al., 1996), are also culpable in the development of benign prostatic hyperplasia (Giovannucci et al. 1999a; 51 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1999b; Mitsumori et al., 1999) and in the penetrance o f rheumatoid arthritis (Kawasaki et al., 1999). Furthermore, longer CAG repeat sizes, but still within the normal range, have been correlated with impaired sperm production and male infertility in several independent studies (Tut et al., 1997; Yong et al., 1999; Dowsing et al., 1999; Legius et al., 1999; Yoshida et al., 1999). Considering this, it may be wise to reexamine the definitions of ‘normal’ CAG and GGC repeat sizes. Some final considerations Population based case-control studies are powerful tools that can provide much insight into the underlying genetic basis of a complex phenotype. By suggesting a functional or mechanistic link between the phenotype and a genetic trait, such studies can inspire clinical and/or laboratory based research efforts aimed at confirming the link on a molecular level. It should be kept in mind, however, that epidemiological studies often suffer from well-known limitations, such as population sampling biases and other confounding factors, and therefore, should be interpreted with some caution. Nevertheless, when multiple independent studies reveal the same or similar results, the argument for or against a particular genotype-phenotype association is all the more compelling. In this chapter, several population-based epidemiological studies are presented/discussed that either support or deny an association between prostate cancer risk and certain alleles of the AR gene (Table 2.8). Every study of US Caucasian men 52 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. revealed an association between short AR CAG alleles and prostate cancer risk, albeit some more convincingly than others. Most o f these studies, furthermore, found a correlation with disease stage/grade and half found some evidence of an effect on age at onset. Thus, based on the cumulative power of these studies, the argument for a link between the AR gene and prostate cancer susceptibility seems strong, at least in this population. The fact that 3 of 4 European studies failed to find such an association may be explained by racial-ethnic differences in population groups, though this seems unlikely considering the recent European history of US Caucasians. Another, perhaps more plausible explanation is that the European studies were conducted on more recent incident cases. The argument here rests on the fact that widespread PSA screening has dramatically altered the spectrum of prostate cancer cases to include men whose tumors will never become life-threatening (Jacobsen et al., 1995). Thus, the AR CAG effect may be effectively diluted in these latter studies. 53 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 3 Inhibition o f p l6 0 coactivation with increasing A R poly-Q length INTRODUCTION The androgen receptor (AR) is a ligand-dependent transcription factor and a member of the superfamily of nuclear receptors (NRs) (Mangelsdorf et al., 1995). AR activity is required for the growth, differentiation, and maintenance of male reproductive tissues, including the prostate (see Introduction, Chapter 1). Once bound to testosterone (T) or its more active metabolite, 5a-dihydrotestosterone (DHT), the AR undergoes allosteric changes that result in its dissociation from chaperone proteins, homodimerization, cytoplasmic to nuclear translocation, DNA binding to hormone response elements (HREs), and the transcriptional regulation of AR target genes (reviewed by MacLean et al., 1997; Roy et al., 1999). In addition to activating transcription, including that of genes involved in prostatic epithelial cell proliferation (i.e., ECnudsen et al., 1998), the AR has been shown to mediate transcriptional repression either by binding to so-called negative HREs (Zhang et al., 1997) or by direct protein-protein interactions with other transcription factors in lieu of AR DNA binding (Schneikert et al., 1996). The AR can also participate in other signal transduction pathways, distinct from the classical androgen signaling axis, due to its ligand-independent activation by the protein kinase A (PKA) cascade (Nazareth and 54 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Weigel, 1996; Sadar, 1999), the HER-2/neu receptor tyrosine kinase pathway (Craft et al., 1999; Yeh et al., 1999), and by various peptide growth factors (Culig et al., 1994). The nuclear receptor superfamily The nuclear receptor superfamily claims more than 100 members and is the single largest family of metazoan transcription factors (McKenna et al., 1999). Members o f this superfamily are generally divided into three subclasses based on differences in ligand discrimination and function. Receptors for the known steroid hormones (i.e., androgens, progestins, estrogens, glucocorticoids, and mineralocorticoids) are grouped together into Class I. When activated by ligand, these ‘steroid hormone’ receptors form homodimers that bind to palindromic inverted repeats composed o f two hexanucleotide half-sites separated by three spacer nucleotides (Roy et al., 1999). While the same palindromic HRE (i.e., GGTACA N3 TGTTCT) promiscuously can be bound by several different Class I receptors, leading to the transcriptional activation o f a target gene, other, more complex and receptor-specific c/s-elements have been described. The rat probasin gene, for example, contains two androgen-specific response elements, only one of which has sequence similarity to the canonical HRE (Kasper et al., 1994). Receptors for thyroid hormone, all-trans retinoic acid, 9-cis retinoic acid, and vitamin D3 make up Class II. These receptors can bind DNA in the presence or absence of ligand at a hexameric direct repeat with variable spacing requirements (Tsai et al., 1999). Unlike Class I receptors, Class II receptors form heterodimers, often with the 9- 55 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. cis retinoic acid receptor (RXR). The third and final class of receptors includes the so- called orphan receptors, for which ligands remain to be identified. It is estimated that there are nearly 50 such receptors awaiting functional characterization (Tsai et al., 1999). AR structure and function The AR, like other members of the nuclear receptor superfamily, comprises three distinct structural/functional domains: a poorly conserved N-terminal domain (NTD) encoded by exon 1 of the AR gene; a cysteine-rich DNA-binding domain (DBD); and a C-terminal ligand-binding domain (LBD) (Mangelsdorf et al., 1995; Fig. 3.1). The NTD contains a ligand-independent, constitutive activation function (generally referred to as AF-1) that is repressed by the apo- or unliganded LBD in the wild type (wt) AR (Jenster et al., 1991). Deletion mapping studies have localized the NTD activation function to two overlapping but distinct transcriptional activation units (TAUs) called TAU-1 and TAU-5 (Jenster et al., 1995; Fig. 3.1). The rat AR TAU-1, moreover, subsequently has been shown to contain two noncontiguous transactivation regions (Chamberlain et al., 1996). These regions, termed A F-la and -lb, correspond to aa 172-185 and 296-359, respectively, in the human AR. In addition to its activation function, the AR NTD also contains homopolymeric stretches of glutamine, glycine, and proline. As discussed in Chapter 2, the glutamine and glycine stretches are polymorphic, whereas, the proline stretch is invariant. Interestingly, poly-Q and polyproline (poly-P) stretches are 56 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. commonly found in transcription factors, perhaps suggestive of important roles in modulating transcriptional activation (Gerber et al., 1994). NR DNA binding domains are highly conserved. In fact, among the steroid hormone receptors, the homology is nearly 80% (Tsai et al., 1999). The AR, like all NRs, has 9 invariant cysteines within this 66 aa domain, 8 of which are involved in the coordination of zinc ions and form the basis of two tetrahedral structures known as Zn2 * fingers (Berg, 1989). Based on mutational analyses and on crystallographic studies, it has been shown that the first or N-proximal Zn2 " finger makes major groove contacts with the NR recognition sequence (i.e., the HRE) while the second finger is involved in dimerization, making contacts with the DBD of the other receptor monomer (Green et al., 1988a; Luisi et al., 1991). The crystal structures of several NR LBDs have been described (Moras and Gronemeyer, 1998 and references therein). All more or less are similar, consisting of 12 a-helices (i.e., H1-H12) and intervening P-strands arranged in an anti-parallel structure that contains a hydrophobic ligand binding pocket (LBP). When not occupied by ligand, this domain assumes an ‘open’ configuration with the C-terminal H12 folded away from the mouth of the LBP. In the holo- or ligand-bound state, however, the configuration is ‘closed’, with H12 positioned over the LBP, effectively trapping the ligand within. The AR LBD, as do other NR LBDs, contains an activation function (i.e., AF-2), but unlike the NTD AF-1, it is ligand-dependent (Danielian et al., 1992). The so-called AF-2 core domain has been localized to H I2, and it has been shown that the 57 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. conformational changes induced by agonist binding position it appropriately for interactions with NR coactivators (Shiau et al., 1998). Interestingly, the effectiveness of some compounds to serve as NR antagonists apparently lies in their respective abilities to induce allosteric changes in LBD structure that preclude productive AF-2/coactivator interactions (Brzozowski et al., 1997). While it is clear that both AR AFs can act autonomously in in vitro assays, maximal transactivation o f AR-responsive genes requires the coordinated activities of both AF-1 and AP-2, though their relative contributions to this process can vary depending on the promoter and cellular context (Jenster et al., 1995; Snoek et al., 1998). In addition to ligand binding and transactivation, the AR LBD also mediates nuclear localization (Jenster et al., 1993) and intra-/inter-molecular interactions with the AR NTD (Langley et al., 1995; Doesburg et al., 1997). The NR coactivator complex and chromatin remodeling Much information has been accrued about how NRs mediate the transactivation of target genes. In general, liganded, dimerized, and DNA-bound NRs recruit to the target promoter a large, multisubunit coactivator complex that possesses histone acetyltransferase (HAT) activity (Torchia et al., 1998; Xu et al., 1999; Fig. 3.2). Reversible acetylation of the basic N-terminal tails of the core histones results in the destabilization of histone-DNA interactions, creating a less condensed chromatin configuration (reviewed by Collingwood et al., 1999). This relaxed or ‘open’ chromatin 58 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. state, in turn, allows for the binding and assembly of the Pol II-containing preinitiation complex (PIC) at the core promoter of the target gene (reviewed by Roeder et al., 1996). Once the PIC has gained access to DNA, it is stabilized through direct contacts with the NR coactivator complex, increasing the likelihood of a processive transcriptional event. In addition to HAT-containing complexes, some NRs can also recruit switch/sucrose non-fermentable (SWI/SNF) complexes [Brahma-related gene-1/BRGl associated factor (BRG1/BAF) complexes in mammalian cells] to target promoters, which disrupt histone-DNA contacts in an ATP-dependent manner (Collingwood et al., 1999). Thus, binding of the liganded NR dimer to its cognate cis-element precipitates a elaborate series of molecular events which ultimately leads to local chromatin remodeling and transcriptional activation (Fig. 3.2). The p l6 0 coactivators Central to the NR coactivator complex is a family of 160 kDa proteins with three genetically distinct members [i.e., (i) SRC-l/NcoA-1, (ii) GRIPl/TIF2/NcoA-2, and (iii) AIBl/pCIP/ACTR/TRAM-l/RAC3] (reviewed by McKenna et al., 1999; Fig. 3.3). These so-called pi 60 coactivators bind to the LBDs of NRs in a ligand-dependent maimer via leucine-rich motifs (i.e., LxxLL where x is any amino acid) called NR boxes (Heery et al., 1997; Torchia et al., 1997; Ding et al., 1998). These conserved motifs form amphipathic a-helices, with the leucines creating a hydrophobic surface on one side (Torchia et al., 1997), that occupy an AF-2-containing hydrophobic cleft in the holo-LBD 59 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (Feng et al., 1998). Three centrally-located NR boxes comprise the NR interaction domain (NED) o f each pl60 family member (Fig. 3.3). An isoform o f SRC-1 (i.e., SRC- la), moreover, has an additional, nonconserved NR box at its extreme C-terminus. While NR boxes are conserved features of the pi 60 coactivators, they are not exclusive to them. Indeed, a plethora of other NR-associated factors contain these motifs, including the promiscuous transcriptional cointegrator CREB binding protein (CBP)/p300 (Heery et al., 1997). Thus, the LxxLL motif appears to be a signature sequence found in proteins that directly contact and regulate NR AF-2. In addition to the NR box-containing NID, p i60 coactivators possess other characterized structural/functional domains, including two autonomous activation domains, ADI and AD2 (Fig. 3.3). It is through the centrally located ADI that interactions with CBP/p300 are mediated. CBP and p300 are highly related, though not entirely functionally redundant, HATs that act as coactivators of a diverse set of transcription factors that includes p53 and NF-kB, as well as the NRs (McKenna et al., 1999 and references therein). They interact directly with NRs through an N-terminal LxxLL motif, with p i60 coactivators through a C-terminal glutamine-rich region, and with another HAT called P/CAF (p300/CBP associated factor) via a C-terminal cysteine- rich domain. Thus, through ADI, the NR-anchored p i60 coactivator serves to localize the independent HAT activities of CBP/p300 and P/CAF to the target promoter. Through the C-terminal AD2, interestingly enough, another possible chromatin remodeling activity is recruited by the p i60 coactivator, namely the novel histone methyltransferase 60 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. activity o f CARM-1 (Chen et al., 1999a). The p i60 coactivator, therefore, can be viewed as a critical component of the NR coactivator complex, linking the promoter-specified NR to the cellular machinery putatively responsible for activated transcription. It is important to note that in addition to recruiting histone acetyltransferase and methyltransferase activities, some p i60 coactivators possess intrinsic HAT activity (Chen etal., 1997; Spencer et al., 1997). A novel NR-pl60 coactivator interaction Due to our interest in the AR and in understanding the molecular basis for its role in prostate cancer etiology, we initiated a two-tiered study to elucidate the consequences of poly-Q variation on AR function. Because the polymorphic poly-Q stretch resides in the AR NTD, we hypothesized that it influences AR function either through direct interactions with other modulatory factors (i.e., known NR-associated factors or previously undescribed proteins) or by exerting conformational stresses on the NTD tertiary structure that indirectly affect factor binding at proximal interaction surfaces. The rationale for looking for AR NTD protein partners comes from the work o f Perutz and colleagues (Stott et al., 1995) who have shown that poly-Q stretches can act as polar zipper protein-protein interaction motifs. Based on this, it was not difficult to envisage a protein factor, possessing a glutamine-rich domain or a homopolymeric glutamine stretch, that might differentially interact with the variable NTD poly-Q and modify AR transactivation activity. To identify novel modulatory proteins that bind directly to the 61 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. poly-Q stretch, a yeast two-hybrid screen was done using an N-tenninal fragment of the AR NTD as bait. A full description of this approach is presented in Chapter 5, along with the results o f the two-hybrid screen. To assess the potentially important role of the NTD in AR-pl60 coactivator interactions, a collaboration was initiated with the laboratory of Dr. Michael Stallcup (University of Southern California, Department o f Pathology). Using yeast two-hybrid and in vitro interaction assays, a novel AR NTD interaction domain was identified in the C-terminus of GRIP1, distinct from the GRIP1 NR box-containing NID (Ma et al., 1999). This finding is consistent with other studies that have reported p i60 interactions with the NTDs of the estrogen receptor (ERa) (Webb et al., 1998; Lavinsky et al., 1998) and the progesterone receptor (PR) (Onate et al., 1998). Interestingly, the NTD is poorly conserved among these steroid receptors and yet the p i60 interaction is a shared common feature. This suggests that while the NTD-pI60 interaction is likely to be important, its functionality may vary significantly from receptor to receptor. In the present study, we mapped on the AR NTD the interaction with the p i60 coactivator GRIP1. Furthermore, in a series of in vitro transactivation assays, we investigated the consequences of this interaction on AR AF-1 function. Finally, we assessed whether or not pl60-mediated coactivation of the AR was influenced by size variations in the NTD poly-Q stretch. 62 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Analysis o f a somatic AR mutant As was discussed in Chapter 2, germ-line variation in AR CAG repeat length confers prostate cancer risk at one extreme and causes SBMA, with its associated partial androgen insensitivity, at the other. In addition to the CAG polymorphism, many other germ-line alterations in the AR gene have been described, most of which result in either partial or complete androgen insensitivity syndrome (PAIS or CAIS) (reviewed by Quigley et al., 1995). Perhaps not surprisingly, most AR mutations associated with AIS are Toss-of-function’ missense mutations that cause deficits in DNA or ligand binding. Somatic alterations of the AR gene, on the other hand, tend to confer ‘gain-in-fiinction’ and have been reported to occur at varying frequencies in primary and metastatic prostate tumors (reviewed by Kokontis and Liao, 1999). In one study, Tilley et al. (1996) found AR mutations in 11/25 (44%) primary lesions. Because the growth of most early stage prostate tumors is androgen-dependent, somatic AR mutations are thought to provide a selective growth advantage to the cells that harbor them. Indeed, there is much evidence to support this notion. Perhaps the most notorious somatic AR mutation described to date is the codon 877 mutation of the LN-CaP cell line. This threonine to alanine substitution results in dramatically reduced ligand specificity such that the mutant receptor is activated by progestins, estrogens, and even the antagonist, hydroxyflutamide (Kokontis and Liao, 1999 and references therein). In a finding directly relevant to the work presented in this thesis, Schoenberg et al. (1994) identified in a prostate tumor a somatic AR mutant with a 63 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6 CAG codon deletion in the CAG microsatellite (i.e., from 24 to 18 CAGs). Based on our own in vitro experiments (Irvine et al., 2000) and those of others (e.g., Kazemi- Esfaqani et al., 1995), it is tempting to speculate that this mutant provided a cellular growth advantage due to its increased transactivation activity. While the physiologic relevance of the LN-CaP and CAG2 4 _,1 S mutants has been bome out by experimental investigations, other identified somatic mutants await functional characterization. One such mutant is the AR 2xLeu mutant cloned by Dr. Wayne Tilley and colleagues (Flinders Cancer Centre, Bedford Park South Australia) from an adenocarcinoma of the prostate (Tilley et al., unpublished results). This mutant has two CTG codons inserted into the CAG microsatellite [i.e., 5’-CTG(CAG),2 CTG- (CAG)6 CTG(CAG)2 CAA-3’ instead o f 5’-CTG(CAG)2 0 CAA-3’] and therefore, encodes a poly-Q disrupted by two interspersed leucine residues. Here we report some preliminary characterizations of this mutant with respect to its relative transcriptional activation potential. 64 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. MATERIALS AND METHODS Plasmids Mammalian expression vector pcDNA-hAR was constructed by inserting a pCMX.hAR-derived Kpnl-Xbal fragment encoding the full length human AR into the corresponding restriction endonuclease sites of pcDNA3.1(+) (Invitrogen, Carlsbad, CA). To construct pcDNA-hAR(Q)n vectors encoding ARs with different poly-Q lengths, Nhel-AflU fragments were PCR amplified (primers FI and R l; Table 3.1) from genomic DNA samples and inserted into pcDNA-hAR digested with the corresponding restriction enzymes (Fig. 3.4). To construct vector pcDNA-AR (NTD-DBD) encoding aa 1-647 of the AR, an Nhel-BamHl fragment was PCR amplified from pcDNA-hAR plasmid DNA (primers F2 and R2) and inserted into the reciprocal restriction sites of pcDNA3.1(+). Vector pcDNA-AR (DBD-LBD) encoding aa 1-6 and 538-919 of the AR was constructed in sequential cloning steps. First, an Nhel-Kpnl PCR fragment containing the AR Kozak sequence (primers F2 and R3) was inserted into the corresponding sites of pcDNA3.1(+). Second, a Kpnl-EcoRl PCR fragment (primers F3 and R4) was inserted into the restored Kpnl site and the downstream £coRI site of the pcDNA3.1(+) multiple cloning site. Mammalian expression vector pSG5.HA, which we used to express proteins fused with an N-terminal HA tag from an SV40 promoter and in vitro from a T7 promoter, was described previously, as were pSG5.HA-GRIPl (full length), pSG5.HA-GRIPl (1122- 1462), and pSG5-SRC-la (full length) (Chen et al., 1999a; Ma et al., 1999). Mammalian 65 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. expression vectors pSG5.HA-GRIPl (5-1121) and pSG5.HA-GRIPl (NR*) (Ma et al., 1999), pCMV-hAR (Tilley et al., 1989), pcDNA3.1-AIBl (Anzick et al., 1997), and pMMTV-CAT (Giguere et al., 1986), that encodes the chloramphenicol acetyltransferase (CAT) gene under the control of the murine mammary tumor virus (MMTV) long terminal repeat (LTR), were described previously (Fig. 3.5). Reporter plasmid ARR3 TK- CAT, which encodes CAT under the control of three identical fragments of the rat probasin gene promoter [i.e., -244/-96, the androgen responsive region (ARR)] and a minimal thymidine kinase (TK) promoter, was also described previously (Kasper et al., 1999; Fig. 3.5). Plasmid pCMV-hAR (2xLeu) that encodes a mutant AR [i.e., 5’- CTG(CAG)1 2 CTG(CAG)6 CTG(CAG)2 -CAA-3’ instead of 5’-CTG(CAG)2 0 CAA-3’ in exon 1] was cloned in the laboratory of Dr. Wayne Tilley of the Flinders Cancer Centre, Bedford Park, South Australia (Tilley et al., unpublished results). Yeast expression vectors encoding GAL4 AD fused with AR NTD (aa 2-538), AR NTD (aa 2-156), AR NTD (aa 141-366), and AR NTD (aa 351-538), were constructed by inserting BamHl-Xhol PCR fragments into the corresponding restriction sites of pACT2 (Clontech, Palo Alto, CA). The PCR primers used were the following: AR NTD (aa 2- 538), F4 and R5; AR NTD (aa 2-156), F4 and R6; AR NTD (aa 141-366), F5 and R7; and AR NTD (aa 351-538), F6 and R5 (Table 3.1). Yeast expression vectors pGBT9-GRIPl (5-765) (Hong et al., 1999) and pGBT9-GRIPl (1122-1462) (Chen et al., 1999a) were described previously. Plasmid pLAM5’ encoding aa 66-230 of human lamin C is commercially available (Clontech). 66 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Bacterial expression vector encoding the glutathione S-transferase (GST) fusion protein GST-AR NTD was constructed by inserting a BamHL-Xhol fragment encoding aa 1-555 of the AR into the reciprocal sites of pGEX-2TK (Amersham-Pharmacia Biotech, Piscataway, NJ). Yeast two-hybrid assays Transformations o f the Saccharomyces cerevisiae reporter strain CG-1945 (Clontech) were performed according to the LiAc method (Schiestl and Gietz, 1989; Hill et al., 1991; Gietz et al., 1992). Yeast transformants were grown in selective media deficient in tryptophan and/or leucine. Total cellular extracts were prepared in Z-buffer [60 mM Na2 HPO4.7H2 0, 40 mM NaH2 P04 , 10 mM KC1, 1 mM MgS04.7H2 0, (pH 7.0)] by repeated freezing and thawing. P-galactosidase (P-gal) assays were performed using the Luminescent P-galactosidase Detection Kit II (Clontech). P-gal activities were corrected for culture densities (cpm/OD6 0 0 ). Data presented are the mean + standard error (SE) from three different yeast transformants and are representative of at least two independent experiments. GST pull-down assays GST pull-down assays were performed by Han Ma of the laboratory of Dr. Michael Stallcup (University of Southern California, Los Angeles, CA) as described previously (Hong et al., 1996) except that 3 5 S-labeled proteins were produced with the 67 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TNT T7 coupled reticulocyte lysate system (Promega, Madison, WI), and GST fusion proteins were produced in E. coli strain BL21. Cell culture and transfections PC-3 (Ohnuki et al., 1980) and COS-7 cell lines, obtained from the American Type Culture Collection (Manassas, VA), were maintained in RPM3 (PC-3) or DMEM (COS-7) medium that contained 10% fetal bovine serum (FBS). Approximately 24 h prior to transfection, 106 (PC-3) or 3 x 10s (COS-7) cells were seeded into each 60 mm dish. Cells were transfected with Lipofectamine reagent (Life Technologies, Rockville, MD) according to the manufacturer’s protocol using 20 pi o f reagent per dish. When appropriate, the total amount o f transfected DNA per dish was held constant by the addition o f pcDNA3.l(+) vector. After transfection, cells were grown in RPMI media (without phenol red) that contained 10% charcoal/dextran-stripped FBS (Gemini Bio Products, Calabasas, CA) for 48 h before harvest. Where indicated, medium was supplemented with 1 or 10 nM DHT during the last 24 h of growth. Whole-cell extracts were prepared in 0.25 M Tris-HCl (pH 8.0) by repeated freezing and thawing. Total cellular protein was measured using the Bio-Rad (Hercules, CA) Protein Assay Kit and CAT assays were performed using the Quant-T-CAT Kit (Amersham-Pharmacia Biotech). CAT activities corrected for total cellular protein (cpm/OD5 9 5 ) are shown as the mean + SE o f three independent dishes, each transfected with a unique liposome/DNA preparation. 68 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Western blotting and scanning densitometry PC-3 and COS-7 cells were maintained and transfected as above. Following 24 h exposure to 10 nM DHT or no hormone, transfected cells were harvested in 100 pi RIP A buffer [10 mM sodium phosphate, 2 mM EDTA, 150 mM NaCl, 50 mM NaF, 0.1% SDS, 1% Nonidet NP-40, 1% sodium deoxycholate, 0.2 mM Na3 V 0 4 , (pH 7.2)] that contained mammalian protease inhibitors [protease inhibitor cocktail (Sigma, St. Louis, MO): aprotinin, leupeptin, pepstatin A, bestatin, 4-(2-aminoethyl) benzenesulfonyl fluoride, trans-epoxysuccinyl-L-leucyl-amido (4-guanidino) butane]. Total cellular protein was measured using the Bio-Rad Protein Assay Kit (Bradford, 1976) and equal amounts of each extract were analyzed by SDS-PAGE. Proteins were transferred to Hybond-P membrane (Amersham-Pharmacia Biotech) and probed with rabbit polyclonal anti-AR antibody AR (N20) (Santa Cruz Biotechnology, Santa Cruz, CA) at 1 pg/ml, rabbit polyclonal anti-HA antibody (Y-l 1) (Santa Cruz Biotechnology) at 1 pg/ml, or mouse monoclonal anti-cytokeratin peptide 8 (CK8) antibody (Sigma) at 10 pg/ml. Horseradish peroxidase-conjugated anti-rabbit IgG (Santa Cruz Biotechnology) at 80 ng/ml (1:5000 dilution) or goat anti-mouse IgG (Bio-Rad) at 1:3000 dilution was used as the secondary antibody. Proteins were visualized by membrane treatment with Luminol Reagent (Santa Cruz Biotechnology) and exposure to Hyperfilm ECL (Amersham-Pharmacia Biotech). Autoradiograms from 3 independent experiments were analyzed by scanning densitometry using a Bio-Rad Model GS-710 Imaging Densitometer and data are reported as mean optical density (OD) units + SE. 69 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Ligand binding assays PC-3 cells were maintained and transfected as above. Forty-eight hours after transfection, cells received fresh medium supplemented with 10% charcoal/dextran- stripped FBS and 2 nM [3 H]DHT. Following a 1 h incubation at 37° C/5% C 02 , cells were washed extensively with ice-cold PBS that contained 1% bovine serum albumin (BSA) and then, were dissolved in 500 pi 1% sodium dodecyl sulfate (SDS). Total bound cpm was determined by scintillation counting. Non-specific ligand binding [i.e., total bound cpm from cells transfected with 5.0 pg pcDNA3.1(+)] was subtracted from each value. Data presented are the mean + SE of three dishes, each transfected with a unique liposome/DNA preparation. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 70 RESULTS The AR NTD interacts with the C-terminus o f GRIP1 Protein-protein interactions between the AR NTD and the p i60 coactivator GRIP1 were investigated using the yeast two-hybrid system (Fields and Song, 1989; Chien et al., 1991). The AR NTD (aa 2-538) and two fragments of GRIP 1 (i.e., aa 5-765 and 1122-1462), were expressed as GAJL4 AD and GAL4 DBD fusion proteins, respectively, in the CG-1945 yeast strain, which harbors an integrated P-gal reporter gene under the control of GAL4 response elements (Giniger et al., 1985). While the AR NTD did not interact with the N-terminal GRIP1 fragment (i.e., aa 5-765) or an irrelevant target protein (i.e., LAMC), it did interact with the C-terminal GRIP1 fragment (i.e., aa 1122-1462) as evidenced by increased relative P-gal activity (Fig. 3.6A). Due to its considerable intrinsic transactivation activity in the yeast system, an internal GRIP1 fragment comprising aa 766-1120 was not tested for interactions with the AR NTD. This region of GRIP1 was assessed, however, in the glutathione S-transferase (GST) pull down experiments described below. To confirm the physical interaction between the AR NTD and GRIP1 C-terminus, GST pull-down experiments were performed by Han Ma of the laboratory of Dr. Michael Stallcup (University of Southern California, Los Angeles, CA) in which in vitro translated and 3 5 S-labeled GRIP1 (full-length), GRIP1 C-truncation (aa 5-1121), and GRIP1 C-terminus (aa 1122-1462) proteins were incubated with immobilized GST-AR 71 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. NTD (Fig. 3.6B). Both GRIP1 (full-length) and GRIP1 C-terminus were able to bind to GST-AR NTD while the GRIP1 C-truncation fragment (aa 5-1121) was not. Thus, the combined yeast two-hybrid and GST pull-down results confirm an AR NTD-GRIP 1 interaction that is mediated by the C-terminus of GRIP1. GRIP I C-terminus binds to the AR NTD downstream o f the TAU-l core region The AR NTD-GRIP1 C-terminal interaction, described above, was mapped on the AR NTD using the yeast two-hybrid system. Three overlapping fragments of the AR NTD (i.e., aa 2-156, 141-366, and 351-538), were expressed as GAL4 AD fusions in the yeast reporter strain along with the GRIP1 C-terminus (aa 1122-1462)-GAL4 DBD fusion protein (Fig. 3.7B). The GRIP1 C-terminal fragment interacted with the AR NTD (aa 351-538) fragment but did not bind to either the AR NTD (aa 2-156) or (aa 141-366) fragments. Thus, a novel GRIP1 interaction surface in the AR NTD is situated downstream of the TAU-l core region defined previously by Jenster et al. (1995). Coactivation o f the AR through the AR NTD-GRIP1 C-terminal interaction To investigate the functional relevance of the AR NTD-GRIP 1 C-terminal interaction in GRIP 1-mediated coactivation o f the AR, a series of transient cotransfection experiments were performed using a panel o f GRIP1 mutants (Fig. 3.8A). The GRIP1 (C-truncation) mutant comprises aa 5-1121 and is, therefore, missing the C-terminal region that binds to the AR NTD (Figs. 3.8A and 3.6A). The GRIP1 (NR*) mutant is 72 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. full-length, but harbors alanine substitutions in the second and third NR boxes (i.e., LXXAA instead of LXXLL). The GRIP1 (C) mutant includes only aa 1122-1462, the AR NTD-binding region of GRIP1. Relative expression levels of the GRIP1 mutants were determined by performing Western blot analyses on transfected COS-7 cell extracts (Fig. 3.8B). AR transactivation activities in the presence of the GRIP1 mutants were not corrected for expression levels since the reporter assays and immunoblots were done in different cell lines. As expected, when GRIP1 (full length) was cotransfected into the prostate adenocarcinoma cell line, PC-3, with either wild type AR or an AR mutant comprising only the AR DBD and LBD [i.e., AR (DBD-LBD); Fig. 3.1], it increased the DHT- dependent, AR-mediated transactivation of the MMTV-CAT reporter gene 3-5 fold (Fig. 3.9). Similarly, although to a lesser degree, the GRIP1 (C-truncation) mutant coactivated both AR forms in a DHT-dependent manner. Because this mutant is unable to bind to the AR NTD (see Fig 3.6B), this result likely reflects the functional interaction between the GRIP1 NR boxes and the AR LBD (i.e., AF-2 core domain). In contrast to GRIP1 (full length) and GRIP1 (C-truncation), GRIP1 (NR*) failed to coactivate AR (DBD-LBD) but increased AR (wt) transactivation nearly to that achieved by GRIP1 (full length). Coactivation by GRIP1 (NR*) was DHT-dependent even though the AR LBD was not engaged by the GRIP1 NR boxes. Finally, GRIP1 (C) which, as expected, did not coactivate AR (DBD-LBD), potentiated AR (wt) transactivation in a DHT-dependent manner (Fig. 3.9). Both GRIP1 (full length) and GRIP1 (NR*) somewhat increased 73 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. background reporter activity in the absence of exogenously added DHT. This was presumably due to the presence o f residual androgens (i.e., possibly sulfated androgen derivatives) in the charcoal/dextran-treated FBS. AIB1 and SRC-la also functionally interact with the AR NTD To more effectively isolate the AR NTD-GRIP 1 C-terminal interaction in a functional assay, a series of transient cotransfection experiments were performed using the relevant GRIP1 mutants and an AR mutant lacking the LBD (Fig. 3.10). Despite its relatively powerful, hormone-independent, constitutive activation activity on the MMTV promoter, AR (NTD-DBD) was markedly coactivated by GRIP1 (full length), GRIP1 (NR*), and to a lesser extent by GRIP1 (C). In addition to GRIP1, two other members of the p i60 coactivator family, namely AIB1 and SRC-la, similarly enhanced AR (NTD- DBD)-mediated transactivation of the MMTV-CAT reporter. Thus, functional interactions with the AR NTD are common among p i60 coactivator family members. Variation in the AR poly-Q tract affects AR transactivation activity To assess the impact of poly-Q size variation on AR transactivation activity, several unique vectors were constructed that expressed AR with 9, 21, 29, 42, or 50 glutamine residues in the poly-Q tract [i.e. AR(Q)9 , AR(Q)2,, etc.]. AR(Q)4 2 and AR(Q)5 0 are representative mutant SBMA receptors. Ligand binding assays were performed to estimate the steady-state expression levels of the different AR alleles in PC-3 cells (Fig. 74 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3.11). Extracts prepared from PC-3 cells transfected with equal amounts of AR expression vector demonstrated similar [3 H]DHT binding activities for receptors containing from 9-42 glutamines in the poly-Q tract. The ligand binding activity o f AR(Q)S 0 extracts, however, was significantly lower than the others (P <0.001; Student’s t test). To test whether the measured ligand binding activities correlated with relative AR(Q)„ expression levels, anti-AR Western blots were carried out on whole-cell extracts from transfected PC-3 cells, and the resultant autoradiograms were quantified by densitometric scanning (Fig. 3.12A). The relative expression levels determined by this technique for AR(Q)9 -AR(Q)4 2 correlated with ligand binding activities. In contrast, no full-length AR(Q)S 0 was detected by this method, suggesting that this protein is markedly unstable in PC-3 cells exposed to DHT for 24 h (see inset representative autoradiogram, Fig. 3.12A). Indeed, an immunopositive band of about 70 kDa, presumably a degradation product, was detected in protein extracts, its relative amount (i.e., relative to full-length AR) increasing with AR poly-Q size. The experiment depicted in Figure 3.12B demonstrates the linear relationship between increasing AR protein content and optical density (OD) of the autoradiographic bands under the conditions used for the quantitative Western analyses. Next, a series of transient transfection assays was performed to determine the effect, if any, of the NTD poly-Q on AR-mediated transactivation of the MMTV-CAT reporter in PC-3 cells (Fig. 3.13). After correction for expression levels as determined by 75 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ligand binding assays, a statistically significant inverse relationship between poly-Q length and AR transactivation activity was observed. There was an approximately 20% reduction in transactivation activity between AR(Q)9 and AR(Q)4 2 (P for linear trend = 0.0001; analysis o f variance) (Fig. 3.13). The mean transactivation activity of AR(Q)5 0 was 80% less than that of AR(Q)9 and the difference was statistically highly significant (P = 0.0001; analysis of variance). Increasing poly-Q length inhibits p!60-mediated coactivation o f the AR To determine the effect of poly-Q length on pl60-mediated coactivation of the AR, a succession of transient cotransfection experiments was performed in which the various AR alleles [i.e., AR(Q)9 -AR(Q)4 2 ] were co-expressed in PC-3 cells along with AIB1, SRC-la, or GRIP1 (NR*). In each case, DHT-dependent, pl60-mediated coactivation of the AR decreased with increasing AR poly-Q length (Fig. 3.14). For example, with AIB1, coactivation, measured as DHT-dependent AR activity in the presence of coactivator (black histograms) minus AR activity in the absence o f coactivator (gray histograms), was approximately 44% lower with AR(Q)4 2 than with AR(Q)9 . The comparable values for SRC-la and GRIP1 (NR*) were 53% and 45%, respectively. In all cases, the relative coactivation levels with AR(Q)4 2 and with AR(Q)2 9 were significantly lower than that with AR(Q)9 (Student’s t test, data not shown). Tests o f trend (analysis of variance) revealed a highly significant inverse relationship between AR 76 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. transactivation activity in the presence o f coactivator (black histograms) and AR poly-Q length [i5 (poly-Q effect, linear trend with AIB1) <0.001; P (poly-Q effect, linear trend with SRC-la) = 0.003; P (poly-Q effect, linear trend with GRIP1 NR*) = 0.005]. In these experiments GRIP1 (NR*) was used because this mutant effectively isolates the AR NTD-pl60 interaction. Similar results were observed with GRIP1 (full length) (data not shown). The AR (2xLeu) mutant has enhanced transactivation activity compared to wild type AR To assess the effect of the 2xLeu insertion mutation on AR function, the relative expression level and transactivation activity of the AR (2xLeu) mutant were investigated. Expression levels of AR (2xLeu) and AR (wt) were determined by anti-AR Western blots carried out on whole-cell extracts from transfected PC-3 cells, and the resultant autoradiograms were quantified by densitometric scanning (Fig. 3.15A). Following correction for CK8 expression, AR (2xLeu) levels were about 50% lower than AR (wt) levels. In subsequent transactivation experiments, the corrected, DHT-dependent activity of AR (2xLeu) was nearly 2-fold higher than that of AR (wt) on the ARR3 TK-CAT reporter (Fig. 3.15B). Furthermore, potentiation of AR (2xLeu) activity by pi 60- coactivators likewise was markedly enhanced relative to AR (wt). 77 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 3.1 LEGEND Schematic diagrams of the structural/functional domains of the AR and two AR mutants. (A) wt AR: NTD, N-terminal domain; DBD, DNA-binding domain; LBD, ligand-binding domain; Q, polyglutamine tract; G, polyglycine tract; P, polyproline tract; NLS, nuclear localization signal; TAU-l, transcriptional activation unit 1 core region; TAU-5, transcriptional activation unit 5; AF, activation function; relevant amino acid numbers are indicated. (B) AR (NTD-DBD) comprises aa 1-647, and therefore, does not include the AF-2-containing LBD. This mutant is constitutively active on androgen-responsive promoters. (C) AR (DBD-LBD) comprises aa 538-919, and therefore, does not include the AF-1-containing NTD. The transcriptional activity of this mutant is hormone- dependent. Both mutants contain the native NLS. 78 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 3.1 C Q 79 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. D B D 1 L B D A R (DBD-LBD) 538 919 FIGURE 3.2 Coactivator complex — * >1 P/CAF J I C B P /p 3 0 0 j'» _ C ^ iTbrcai ISW I/SNF ) W/stone acetylation m ethylation Activation Nuclear receptors Model of nuclear receptor-mediated transcriptional activation. Modified from Xu et al., 1999. 80 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 3.3 LEGEND Schematic diagrams of the structural/functional domains of the p i60 coactivators. NLS, nuclear localization signal; bHLH, basic helix-loop-helix motif; PAS, period/aryl hydrocarbon receptor/single minded; NID, nuclear receptor interaction domain (LxxLL motifs I, II, and III); CID, CBP/p300 interaction domain; ADI and AD2, two autonomous activation domains; Q, glutamine-rich region; NTD^.,, nuclear receptor interaction domain AF-1 (see text for explanation). Gaps in the schematic alignment of the coactivators are indicated by dotted lines. SRC-la has a fourth LxxLL motif at its extreme C-terminus, however, its isoform, SRC-le, does not. These diagrams were modified from Moras and Gronemeyer, 1998. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 81 FIGURE 3.3 C M Q < Q| < g | o » M ta M H § e u M 06 o E o □ V C e £ c / 3 O < o X X X _J X X v[ o < 2 < o □ CM Q < ! ! ? ; 'v 82 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 3.4 LEGEND Schematic representation of the construction of pcDNA-hAR(Q)n vectors. PCR fragments with flanking Nhel and AflQ. restriction sites were amplified from genomic DNA samples and inserted into pcDNA-hAR vector digested with the corresponding restriction enzymes. The CAG microsatellite regions o f the various pcDNA-hAR(Q)n constructs were sequenced to verify CAG repeat lengths. Genomic DNA samples from two SBMA patients were used to construct AR alleles with 42 and 50 CAG repeats. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 83 FIGURE 3.4 ,-** i" <’ r i t i » t i r t i m 111 s m i s m i r m w o' \ \ u o o o o tur t f W » ■ I i i ii t i t t t i i n t i M r i r t i m m i rr i‘ t h W f t I I I I I I I I I I I I M I I llllM H M H IJ jy « * : » * i 1 f m t t W J M t t i m m m m f j» ** r r r r r r r r r r r r r r r r m »rr® in ® * ' ... j* v.“ >'i1 «■ « // t i i 11 1 i n tjH m » » n t ii i r u t i t u e t .V iv y y jv y y y A n hAR cDNA ^ pcDNA-hAR (Q)so pcDNA-hAR (Q)42 pcDNA-hAR (Q)29 pcDNA-hAR (Q)2i - - * ■ pcDNA-hAR (Q)9 Nhel f c * - . S w j S Q. ^ 3: ^ oo g £ 5 S lo lo i -- o a it 2 .Q . Q (I) Q o V W J O - C «Q Q . c «tnmujUj^0D2>:S'CQ- pcDNA3.1 (+) 5.4 kb Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 3.5 LEGEND Androgen regulated promoters. (A) Schematic diagram of the hormone regulatory region (HRR) of the MMTV long terminal repeat. HRE, hormone response element; NF-1, nuclear factor I binding site; Oct, octomer binding site; TATA, TBP binding site. The HHR contains two palindromic HREs and two HRE half-sites. The double-stranded sequence of each HRE is presented with arrows indicating the orientation o f each hexanucleotide half-site. This diagram was modified from Truss and Beato, 1993. (B) Schematic diagram of the ARR3 TK promoter. This promoter comprises three identical fragments (-244/-96) of the androgen responsive region (ARR) o f the rat probasin gene promoter cloned in tandem and upstream of a minimal TK promoter. Each probasin promoter fragment contains two unique AR binding sites (ARBS-1 and -2) that confer selective androgen responsiveness. 85 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 3.5 > H S £ u: H < n 0 5 0 6 < m t m t o i UJ c c cn a > ti I < I m If 1 1 1 0 ) C M C C O C O C O X X o w Q» 86 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 3.6 LEGEND The AR NTD interacts with the GRIP1 C-terminus in yeast two-hybrid and GST pull down assays. (A) S. cerevisiae yeast reporter strain CG-1945 was transformed with pACT2 vector encoding ARNTD (aa 2-538)-GAL4 AD and/or pGBT9 vector encoding GRIP1 (aa 5-765)- or GRIP1 (aa 1122-1462)-GAL4 DBD as indicated. Yeast extracts were prepared from transformants selected in leucine- and/or tryptophan-deficient medium and were assayed for p-gal activity by chemiluminescent detection of a fluorescent substrate. Data presented are the mean + SE from three different yeast transformants and are representative of at least two independent experiments. (B) Glutathione-Sepharose bound GST or GST-AR NTD was incubated with 3 5 S-labeled GRIP1 (full-length), GRIP1 (aa 5-1121), or GRIP1 (aa 1122-1462) translated in vitro from pSG5.HA vectors. Bound proteins were eluted and analyzed by SDS-PAGE and autoradiography; shown for comparison is 10% of the total labeled protein incubated in each binding reaction (i.e., 10% input). 87 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 3.6 AR NTD GRIP1 "335 h © -755 GRIP1 1 1 2 2 1 4 6 2 GRIP1 T S ? AR NTD 538 GRIP1 1122 1462 LAMC AR NTD 2----------------535 AR NTD 538 0 100 200 relative (3-galactosidase activity B GRIP1 (fuii length) r?\°' oS cS < 3 ? <P / GRIP1 (5-1121) GRIP1 (1122-1462) - ' V . - * C O - , y ~ ' ■ Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 3.7 LEGEND GRIP1 C-terminus binds to the AR NTD downstream of the TAU-1 core domain. (A) Schematic diagram of the structural/functional domains of the AR. NTD, N-terminal domain; DBD, DNA-binding domain; LBD, ligand-binding domain; Q, polyglutamine tract; G, polyglycine tract; P, polyproline tract; NLS, nuclear localization signal; TAU-1, transcriptional activation unit 1 core region; TAU-5, transcriptional activation unit 5; AF, activation function; relevant amino acid numbers are indicated. (B) Mapping of the GRIP1 C-terminal interaction on the AR NTD in yeast two-hybrid assays. S. cerevisiae yeast reporter strain CG-1945 was transformed with pGBT9 vector encoding GRIP1 (aa 1122-1462) and/or pACT2 vector encoding AR NTD (aa 2-538)-, AR NTD (aa 2-156)-, AR NTD (aa 141-366)-, or AR NTD (aa 351-538)-GAL4 AD as indicated. Extract preparation and measurement of P~gal activity were carried out as described in Fig. 3.6A. Data presented are the mean + SE from three different yeast transformants and are representative o f at least two independent experiments. 89 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 3.7 © ® © _ ^ A F -la A F -lb ^ ^ ^ I ■ ' NTD | M I DBD | LBD AF-2 101____________________2ZQ 91 9 TAU-1 2§2______528 TAU-S B AR NTD ------------- 53S AR NTD 351 538 I GRIP1 U dfljftV ITS T T 5 Z GRIP1 TTS Ti5? AR NTD 5----------------- 53S GRIP1 TTS-- US5 - 4»IS H > I AR NTD I — 2 156 GRIP1 TTS T O S S ’ AR NTD TOi "366 GRIP1 1122 1462 AR NTD 351 538 100 200 300 400 relative (3-galactosidase activity Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 3.8 LEGEND GRIP1 mutants. (A) Schematic diagrams of the structural/functional domains of GRIP1 and three GRIP1 mutants. GRIP1 (C-truncation), comprises aa 5-1121, and therefore, contains both the NID and the CID/AD1; GRIP1 (NR*), is full-length, but has alanine substitutions in the second and third NR boxes; GRIP1 (C), comprises aa 1122-1462, and therefore, contains only AD2. (B) Relative expression levels of the GRIP1 mutants determined by Western analysis. COS-7 cells were transiently transfected with 3.0 pg of the indicated pSG5.HA-GRIPl expression vector and 2.0 pg of pcDNA3.1(+) vector. Forty-eight hours after transfection, whole-cell extracts were prepared and normalized for total protein. Equivalent amounts of extract were probed with anti-HA antibody. A representative autoradiogram from three independent experiments is shown with appropriate band sizes for each of the GRIP1 proteins. MWM, molecular weight protein marker; control, extract prepared from COS-7 cells transfected with pcDNA3.1(+). 91 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. N L S b H L H P A S dom ain N I D CID/AD1 AD2 FIGURE 3.8 o 0 o CM CD X X X _ t • X X CD m X X X X 2 * ” V * ‘ - - A . * * > * » v i ^ ^ j §i i i i § ^ Ji tl li I & K * * < ■ * * 92 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 3.9 LEGEND Coactivation of the AR via the NTD-GRIP1 C-terminal interaction. PC-3 cells were transiently cotransfected with 0.5 pg pcDNA-hAR (wt) (left graph) or pcDNA-AR (DBD-LBD) (right graph), 2.0 pg o f the indicated pSG5.HA-GRIPl expression vector, and 2.0 pg of the MMTV-CAT reporter plasmid. The total amount of DNA per transfection was held constant by the addition of pcDNA3.1(+) vector when appropriate. Twenty-four hours after transfection, cells received fresh medium that contained 10% charcoal/dextran-stripped FBS and 10 nM DHT or ethanolic vehicle. Twenty-four hours later, whole-cell extracts were prepared and assayed for CAT activity as described in Materials and Methods. Data presented are the mean + SE of three independent dishes. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 93 A R (wt) A R (DBD-LBD) FIGURE 3.9 94 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4 0 0 3 0 0 2 0 0 1 0 0 0 0 1 0 0 2 0 0 300 relative C A T activity relative C A T activity FIGURE 3.10 LEGEND Functional interactions of pi 60 coactivators with a constitutively active AR mutant. PC- 3 cells were transiently cotransfected with 0.1 pg pcDNA-AR (NTD-DBD), 2.0 pg of the indicated coactivator expression vector [pSG5.HA-GRIPl vectors, pSG5.HA-SRC-la, or pcDNA3.1-AIBl], and 2.0 pg of the MMTV-CAT reporter plasmid. The total amount of DNA per transfection was held constant by the addition of pcDNA3.1 (+) plasmid when appropriate. Twenty-four hours after transfection, cells received fresh media that contained 10% charcoal/dextran-stripped FBS and no hormone. Twenty-four hours later, whole-cell extracts were prepared and assayed for CAT activity as described in Materials and Methods. Data presented are the mean + SE of three independent dishes. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 95 FIGURE 3.10 AR (NTD-DBD) no GRIP1 GRIP1 (full length) GRIP1 (NR*) GRIP1 (C) SRC-1 a relative CAT activity Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 3.11 LEGEND Determination o f AR(Q)n protein expression by specific ligand binding. PC-3 cells were transiently cotransfected with 3.0 pg of the indicated pcDNA-hAR(Q)n expression vector and 2.0 pg of pcDNA3.1(+). Forty-eight hours after transfection, ceils received fresh medium that contained 10% charcoal/dextran-stripped FBS and 2 nM [3 H]DHT. Following a 1 h incubation, cells were harvested and total bound cpm was determined by scintillation counting. Non-specific ligand binding [i.e., total bound cpm from cells transfected with pcDNA3.1(+)] was subtracted from each value. Data presented are the mean + SE of three independent dishes. Relative expression levels of the different AR variants are indicated above each histogram. *Tests of trend were performed by linear regression of binding activity (cpm) on poly-Q size, P = 0.879. **Binding activity of AR(Q)5 0 compared to others (P <0.001; two-tailed, impaired Student’s t test). 97 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 3.11 30000 1.00 0.88 20000 ° 10000 9 * 21 * 29 * 42 * 50 ** AR (Q)n Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 3.12 LEGEND Determination of AR(Q)„ protein expression by Western blot densitometry. (A) PC-3 cells were transiently cotransfected with 3.0 pg of the indicated pcDNA-hAR(Q)n expression vector and 2.0 pg o f pcDNA3.1(+). Twenty-four hours after transfection, cells received fresh medium that contained 10% charcoal/dextran-stripped FBS and 10 nM DHT. Twenty-four hours later, whole-cell extracts were prepared and normalized for total protein. Equivalent amounts of extract were probed with anti-AR antibody. A representative autoradiogram is shown with full-length AR(Q)„ detected at 100-115 kDa. An approximately 70 kDa AR degradation product is indicated by the asterisk. Autoradiograms from 3 independent experiments were analyzed by scanning densitometry and data are reported as mean optical density (OD) units + SE. Relative expression levels of the different AR variants are indicated within each histogram. (B) Establishing the linearity o f the western blot densitometry assay. Increasing amounts of a single AR(Q)9 extract were probed with anti-AR antibody that recognizes an NTD- derived epitope. The autoradiogram was analyzed by scanning densitometry and data are reported as relative OD units per pg protein extract. Data were subjected to linear regression analysis [solid line through the origin (x,y = 0)]. Confidence intervals (95%) of the linear regression line are depicted as dashed lines. MWM, molecular weight protein marker; C (control), 15 pg protein extract from PC-3 cells transfected with pcDNA3.1(+). 99 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 3.12 < o 'E 3 d o o .> 4 > pcDNA-AR (Q)n 97 60 40 20 29 42 0.73 0.83 0.99 1.00 C O c 3 ■ Q ■ O < D > AR (Q)n 15 10 tig protein extract 5 116 97 66 0 0 5 10 15 20 protein extract (fig) 100 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 3.13 LEGEND AR transactivation activity decreases with increasing poly-Q size. PC-3 cells were transiently cotransfected with 0.5 pg of the indicated pcDNA-hAR(Q)n expression vector, 2.0 |ig of the MMTV-CAT reporter plasmid, and 2.0 pg of pcDNA3.1(+). Twenty-four hours after transfection, cells received fresh medium that contained 10% charcoal/dextran-stripped FBS and 10 nM DHT or ethanolic vehicle. Twenty-four hours later, whole-cell extracts were prepared and assayed for CAT activity as described in Materials and Methods. CAT activities were corrected for total cellular protein and for AR(Q)n protein expression levels as determined in ligand binding assays (see Fig. 3.11). Data presented are the mean + SE of three independent dishes. Transactivation activities relative to AR(Q)g are indicated within each histogram. *Two-way analysis o f variance comparing Q9 -Q4 2 in 8 experiments each comprising triplicate dishes: P (poly-Q effect, linear trend) = 0.0001, P (experimental effect) = 0.0001. **Two-way analysis o f variance comparing Q9 and Qs0 in 2 experiments each comprising triplicate dishes: P (poly-Q effect) = 0.0001, P (experimental effect) = 0.03. 101 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 3.13 > > > o C O c o s C O .> o C O C O c C O k - < - < 0 ) > J O 0 ) p (poly-Q effect 9-42, linear trend) = 0.0001 1.0 0.5 0.0 0.20 0.79 0.87 0.93 9 * 21 * 29 * 42 * 50 * AR (Q)n 102 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 3.14 LEGEND Inhibition of pl60-mediated coactivation of the AR with increasing poly-Q size. PC-3 cells were transiently cotransfected with 0.5 pg o f the indicated pcDNA-hAR(Q)n expression vector, 2.0 pg o f the indicated coactivator expression vector [(A) pcDNA3.1- AIB1, (B) pSG5.HA-SRC-la, or (C) pSG5.HA-GRIPl(NR*)], and 2.0 pg of the MMTV- CAT reporter plasmid. Twenty-four hours after transfection, cells received fresh medium that contained 10% charcoal/dextran-stripped FBS and 10 nM DHT or ethanolic vehicle. Twenty-four hours later, whole-cell extracts were prepared and assayed for CAT activity as described in Materials and Methods. CAT activities were corrected for total cellular protein and for AR(Q)n protein expression levels as determined in ligand binding assays (see Fig. 3.11). Data presented are the mean + SE of three independent dishes. White histograms, no coactivator, no hormone; gray histograms, no coactivator, + 10 nM DHT; black histograms, + coactivator, + 10 nM DHT. *Tests o f trend were performed by linear regression of relative CAT activities on poly-Q size. Bracketed values are mean DHT- dependent AR activities in the presence of coactivator (black histograms) minus AR activities in the absence o f coactivator (gray histograms). 103 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 3.14 AIB1 p (linear trend)<0.001* o 1 0 0 - Co B 21 29 AR (Q)n SRC-1 a p (linear trend)=0.003* a 10 - 21 29 AR (Q)n GRIP1 (NR*) p (linear trend)=0.005* S 2 0 - a 21 29 AR (Q)n 104 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 3.15 LEGEND Relative expression and transactivation activity of the AR (2xLeu) mutant. (A) PC-3 cells were transiently cotransfected with 3.0 pg pCMV-hAR or pCMV-hAR (2xLeu) and 2.0 pg pcDNA3.1(+). Twenty-four hours after transfection, cells received fresh medium that contained 10% charcoal/dextran-stripped FBS and 10 nM DHT. Twenty-four hours later, whole-cell extracts were prepared and normalized for total protein. Equivalent amounts of each extract were probed with anti-AR or anti-CK8 antibody. Autoradiograms from three independent experiments are shown that were analyzed by scanning densitometry. The relative expression level (mean ± SE) of AR (2xLeu) compared to AR (wt) following correction for CK8 expression was 0.45 ± 0.07. MWM, molecular weight protein marker; control, extract from PC-3 cells transfected with pcDNA3.1(+). (B) PC-3 cells were transiently transfected with 0.1 pg pCMV-hAR, 2.0 pg of the indicated coactivator expression vector [pSG5.HA-GRIP, pSG5.HA-SRC-la, or pcDNA3.1-AIBl], and 2.0 pg of the ARR3 TK-CAT reporter plasmid. The total amount o f DNA per transfection was held constant by the addition of pcDNA3.1(+) vector when appropriate. Twenty-four hours after transfection, cells received fresh medium that contained 10% charcoal/dextran-stripped FBS and 1 nM DHT or ethanolic vehicle. Twenty-four hours later, whole-cell extracts were prepared and assayed for CAT activity as described in Materials and Methods. Data presented are the mean + SE of three independent dishes. 105 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 3.15 3000 -2000 -1000 O ) s o E u o © s N © ’a. o es ■ U P S 0 3 ±HQ l!\|U L B Q 106 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. relative C A T activity List o f P C R primers u o a o a a u a a u a o < < < < E- E — a o u u u a < < < < o o c < < < o o < < < < u a o o a a u u u u < < E — H a o a o u u o o u u E- E — < < < < < 2 U u a o o y 6 q ^ o o a o p u o a s: o - C M n t m — — — — — — — — U. U* C£. C I . U . C l . U . U . U . C l . U . U . U . C l . U ^ U . s £ s- ■ — 1 N n ^ vi 'fl f' Q £ O S 0 £ 0 £ Q £ 0 £ O S 107 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. co co co co ■ O 4 > 3 r^ i E d -J 09 < H > S « N « % < 5 , s f t, ' J '•n ■ •v . R © y o a a o a u a 0 y o c j < o a o a o a < cj a y C J a a o a cj cj y y O u - > < S J R © ft; o — cn m < r > v o r - * O S Q i O i e ^ Q i O ^ a i O i Q i O i 108 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DISCUSSION In light o f the epidemiological findings with AR CAG genotype and prostate cancer risk (refer to Chapter 2), we endeavored to understand the potential molecular mechanism(s) by which apparently normal AR function can lead to prostate cancer development in some men. To this end, physical and functional interactions between the AR NTD, which contains the polymorphic poly-Q tract (i.e., encoded by the exon 1 CAG repeat), and a key component of the NR coactivator complex, the p i60 coactivator GRIP1, were investigated. In both yeast two-hybrid and in vitro pull-down experiments, the AR NTD interacted with the C-terminus of GRIP1 (aa 1122-1462), a region not implicated in the ligand-dependent interaction of GRIP1 with the AR LBD (Hong et al., 1997; Ding et al., 1998). These observations are indicative of a complex AR-GRIP1 interaction involving two distinct binding surfaces in each molecule. Moreover, they are suggestive of a mechanism for AR AF-l/AF-2 coordination whereby the p i60 coactivator bridges communications between the NTD and the LBD through direct contacts. Indeed, this notion is supported by experiments demonstrating pl60-mediated enhancement of the ligand-dependent, cooperative transactivation of a target gene by separate NR LBD and NTD polypeptides (Mclnemey et al., 1996; Ikonen et al., 1997; Alen et al., 1999). This putative mechanism, moreover, may be universal among the Class I steroid hormone receptors since recent reports have shown binding of p i60 coactivators to the NTDs of ERa (Webb et al., 1998; Lavinsky et al., 1998) and the PR (Onate et al., 1998). As with 109 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the AR, interactions between the ERa NTD and GRIP1 required the C-terminus of GRIP1 (aa 1122-1462) (Webb et al., 1998). In parallel experiments to the ones reported here, we demonstrated that an area of GRIP1 between ADI and AD2 (i.e., aa 1122-1305: see Fig. 3.3) was essential for AR NTD interactions and therefore, it was designated NR interaction domain AF-1 (NID^.,) (Ma et al., 1999). Consistent with our findings (i.e., Ma et al., 1999), He et al. (1999) recently have mapped AR NTD interactions to the C- termini of TIF-2 (the human homologue o f murine GRIP!) and SRC-la. The AR NTD-GRIP1 C-terminal interaction was localized on the NTD to aa 351- 538 (Fig. 3.7). This part of the AR NTD corresponds to the TAU-5 region (aa 360-528) identified by Jenster et al. (1995) as sufficient for transactivation by a constitutively active AR mutant lacking the LBD. In the same study, deletion of aa 244-528 from wt AR resulted in a nearly 80% decrease in ligand-dependent transactivation activity, although deletions of other more N-terminal sequences resulted in comparable deficits. From this and other investigations (Gao et al., 1996; Gast et al., 1998), it is clear that essentially the entire NTD is required for wt AR transactivation activity, including the region involved in GRIP 1 binding. This may be reflective of a complex tertiary structure adopted by this domain, one that possesses multiple, precisely oriented interaction surfaces to accommodate various regulatory factors. Such a structure would most likely be sterically, and therefore, functionally disrupted by sizable deletions anywhere except perhaps at its extreme N- or C- termini. O f course, crystallographic studies of the AR 110 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. NTD are needed to provide a clearer view o f the structural and function relationships that exist within this domain. It is intriguing that the newly identified NTD-GRIP1 interface also overlaps with regions of the NTD implicated in interactions with the holo-AR LBD (Langley et al., 1995; Doesberg et al., 1997; Berrevoets et al., 1998), the promiscuous coactivator CBP/p300 (Fronsdal et al., 1998), and the basal transcription factors TFIIF and TBP (McEwan and Gustafsson, 1997). This convergence of diverse binding activities on an overlapping region of the NTD suggests a critical function for this domain in the assembly and coordination of a transcriptional activation complex at a target promoter. Indeed, in a series of functional assays, the NTD-GRIP1 interaction was sufficient to enhance AR-mediated transactivation of the hormonally regulated MMTV-CAT reporter gene. Specifically, the GRIP1 (NR*) mutant, which was unable to functionally interact with the isolated holo-LBD, efficiently coactivated wt AR in a DHT-dependent manner in PC-3 cells (Fig. 3.9). Furthermore, this GRIP1 mutant, as well as full length GRIP1 and other p i60 coactivator family members, SRC-la and AIB1, potentiated the transactivation activity o f a constitutively active AR mutant lacking the LBD (Fig. 3.10). These observations recently were confirmed by Alen et al. (1999). Using virtually identical AR (NTD-DBD) and (DBD-LBD) constructs, they showed that both were potentiated by the p i60 coactivators TIF2 and SRC-le. A TIF2 NR-box mutant, moreover, that was unable to functionally interact with the holo-LBD, efficiently coactivated wt AR. I l l Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Based on these in vitro experiments, it appears, therefore, that AR AF-1 works, at least in part, by recruiting p i60 coactivators to the target promoter. Surprisingly, p i60 interaction with the holo-LBD is dispensable for pl60-mediated activation of AF-1, perhaps underscoring the important regulatory function played by the apo-LBD in inhibiting the constitutively active NTD. After all, in the context of wt AR, the LBD imposes strict androgen-dependence on AF-1. To this point, He et al. (1999) have shown wt AR activity to be dependent on a strong AR NTD-AR LBD (i.e., AF-2) interaction that occurs only in the presence of ligand. So, even though the AR NTD independently can recruit and functionally interact with p i60 coactivators, it probably serves to stabilize ligand-dependent p i60 interactions occurring at the LBD. In this way, the pi 60- containing coactivator complex is tightly anchored to the liganded receptor, maximizing its effectiveness in modifying the local chromatin environment. Despite the low homology shared among NR NTDs, it seems that the N-terminal AFs of certain other NRs work in a similar fashion. Webb et al. (1998) convincingly showed that ERa AF-1 is strongly enhanced by GRIP1 and SRC-la and that in the case of GRIP1, this enhancement is dependent on the GRIP1 C-terminus. Likewise, Onate et al. (1998) reported the functional enhancement of PR, glucocorticoid receptor (GR), and ERa AF-1 activities by both SRC-1 and TTF2. Taken in the context of these studies, our findings contribute to an emerging model for steroid hormone receptor AF-1 activity that is dependent on productive p i60 coactivator interactions. 112 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Based on experimental evidence, we have argued that AR interactions with the NR coactivator complex are mediated by p i60 coactivator contacts with both AR AFs. It follows, therefore, that disruption of one or both of these contacts could lead to reduced AR-coactivator complex association and a concomitant loss in AR transactivation potential. To address whether or not variation in the size of the poly-Q tract of the AR NTD affects pl60-mediated coactivation of the AR, we first assessed AR transactivation activity alone (i.e., without exogenously expressed coactivator) in PC-3 cells. Following correction for relative AR expression levels, we found a highly significant inverse relationship between poly-Q length and AR transactivation activity on the MMTV-CAT reporter gene at saturating DHT concentrations (Fig. 3.13). Our findings are consistent with those of others who have shown that despite normal ligand binding affinities, ARs with increased poly-Q length have decreased transactivation activity (Mahtre et al., 1993; Chamberlain et al., 1994; Kazemi-Esfaq'ani et al., 1995; Tut et al., 1997). In the Chamberlain et al. (1994) study, for example, AR(Q)4 9 was approximately 17% less efficient at transact!vating the MMTV promoter than ARfQ)^. In the Kazemi-Esfaq ani et al. (1995) study, moreover, a 30-40% drop in transactivation activity, also on the MMTV promoter, was observed with increasing poly-Q size from 12-40 residues. Our findings do not support those of Gao et al. (1996) who found that both expansion and contraction of the poly-Q from 20 residues resulted in blunted AR transactivation activity. Furthermore, we found no poly-Q-dependent differences in protein expression levels for ARs with poly-Q lengths from 9-42 (Fig. 3.12), as has been 113 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. reported previously (Choong et al., 1996). We did, however, note a dramatic lack of expression of AR(Q)S 0 in PC-3 cells treated with DHT for 24 hr (Fig. 3.12A). Since we measured appreciable levels o f AR(Q)S 0 by ligand binding (Fig. 3.11), in which cells were exposed to DHT for 1 h, we conclude that this protein undergoes ligand-dependent degradation in this cell type. The increasing presence, with increasing poly-Q size, o f an approximately 70 kDa degradation fragment in extracts from transfected PC-3 ceils supports the observations of Butler et al. (1998) showing increased production o f a 74 kDa C-terminally truncated form of an SBMA mutant AR [AR(Q)5 2 ] in COS-7 and NB2a/dl neuroblastoma cells. Evidence suggests that SBMA mutant ARs become sequestered and degraded in cytoplasmic and nuclear aggregates in a ligand- and poly-Q size-dependent manner (Butler et al., 1998; Stenoien et al., 1999), and this phenomenon has been offered as an explanation for the observed decreases in their relative transactivation activities (Butler et al., 1998). Our data with AR(Q)5 0 are in agreement with this hypothesis since its low relative transactivation activity is coincident with greatly reduced protein levels (Figs. 3.11-3.13). Lack of AR(Q)S 0 protein expression, on the other hand, partially may be due to decreased mRNA expression as proposed by Choong et al. (1996), though we did not directly address this issue in the present study. Our observations support the existence o f a moderate poly-Q size effect on AR transactivation activity in PC-3 cells transfected with AR(Q)9 -AR(Q)4 2 . Presumably, increased poly-Q length causes allosteric changes in the AR NTD that negatively 114 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. influence interactions with the NR coactivator complex and thus, result in reduced transactivation potency. Because the co-overexpression of exogenous p i60 coactivator (Fig. 3.9) or CBP/p300 (Fronsdal et al., 1998; Aamisalo et al., 1998) results in enhancement of AR-mediated transactivation, endogenous levels of these factors must be limiting. In light of this, measurement of relative AR transactivation activities in the absence of co-expressed coactivator may result in an underestimation o f poly-Q effects if the coactivator is itself involved in mediating them. Indeed, in transient cotransfection experiments, relative pl60-mediated enhancement of AR activity decreased with increasing poly-Q size more dramatically than AR transactivation activity measured alone. Moreover, levels of p i60 coactivation were statistically different between ARs with poly-Q sizes in the normal size range (i.e., 9 vs. 29). In total, these results suggest that the observed effects of the NTD poly-Q on AR transactivation activity are indeed mediated through functional interactions with the NR coactivator complex. With respect to the p i60 coactivator, these effects are most likely indirect consequences of steric hindrance imposed on AR-pl60 interactions by increased poly-Q size. Direct effects, however, mediated by other cell-, promoter-, and/or AR- specific cofactors cannot be completely ruled out. One such cofactor may be a Ras related G-protein (i.e., Ran or AR-associated protein 24), recently shown to bind to the poly-Q region of the AR NTD and to coactivate the AR in an apparently poly-Q size- dependent manner (Hsiao et al., 1999). Due to the well-characterized role of this GTPase in the nuclear transport o f proteins (Rush et al., 1996), it is difficult to comprehend its 115 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. involvement in AR-mediated coactivation. It may well be that Ran Iess-efficiently transports AR into the nucleus with increasing poly-Q size. If so, this could account for apparent differences in coactivation due to poly-Q variation. Obviously, further studies will be needed to clarify this conjecture. Though not directly addressed in the present study, it is interesting to note that the GRIP1 binding region of the AR NTD contains the poly-G and poly-P tracts (Fig. 3.1). As discussed in Chapter 2, the poly-G tract is encoded by a polymorphic GGC microsatellite, the size of which varies in the normal population from about 8-18 repeats. Men with certain AR GGC alleles, moreover, have a significantly increased risk for prostate cancer (see Discussion, Chapter 2). Very little is known about the role(s) of the poly-G in AR function, though its deletion resulted in an approximately 50% decrease in AR transactivation activity (Gao et al., 1996). Thus, as with poly-Q size, variation in poly-G size may have subtle but important effects on AR transactivation potency. We are presently investigating the effects of poly-G size variation on pl60-mediated coactivation of the AR. The functional relevance of the non-polymorphic poly-P tract also remains obscure, though homopolymeric proline tracts can activate transcription when linked to an autologous DNA binding domain (Gerber et al., 1994). This transcription potential is likely due to poly-P-mediated recruitment/stabilization of basal and/or accessory transcription factors. 116 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The 2xLeu insertion mutation confers a gain-in-function on the AR It is clear that poly-Q expansion results in a loss o f AR function, a fact that is vividly demonstrated in persons afflicted with SBMA. We have hypothesized and subsequently provided epidemiologic and functional data to support the idea that reduced poly-Q size confers a moderate gain-in-fimction on the AR that is expressed phenotypically in prostate cancer development. The crux o f this ‘gain-in-function’ seems to be an incremental increase in AR transactivation activity with decreasing poly-Q size. In fact, the complete removal of the inhibitory poly-Q resulted in nearly a 2-fold increase in activity compared to wt AR (Kazemi-Esfaxjani et al., 1995). Along with somatic mutations that relax ligand specificity (i.e., codon 877-,-.^, AR poly-Q variation provides a logical link between altered AR function and phenotype. Analysis of the 2xLeu mutant also provides such a link. This mutant, isolated from an adenocarcinoma o f the prostate, was twice as active as wt AR in in vitro transactivation assays. Moreover, in coactivation assays with p i60 coactivators, this 2-fold relative activity was preserved. It may well be that disruption o f the poly-Q with intervening leucine residues sufficiently relaxes its inhibitory effect on AF-1 function to allow for increased transactivation activity. Whatever the underlying cause may be, the results of this preliminary characterization o f AR 2xLeu are consistent with the widely held belief that somatic AR alterations in early stage prostate cancer at least, are gain-in-function in nature. 117 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 4 BRCA1 is a coactivator o f the androgen receptor INTRODUCTION A circumstantial link between hereditary breast cancer and the AR In a recent report by Rebbeck et al. (1999), an association was observed between AR CAG repeat length and breast cancer penetrance among women with germline BRCA1 mutations (see below for a discussion of BRCA1). In that study, women who carried at least one long AR CAG allele (i.e., >28 repeats), had a significantly earlier age at diagnosis than women possessing no long AR CAG allele. Interestingly, breast cancer penetrance was found to increase with increasing AR CAG length. For example, women with a long AR allele of >28, >29, or >30 CAG repeats developed breast cancer 0.8, 1.8, or 6.3 years earlier than women with only short alleles. Thus, the data presented in the Rebbeck et al. (1999) study strongly suggest a role for the AR in modifying breast cancer risk in some women. Because of the well-characterized negative effect of increasing poly-Q length on AR transactivation potential, it is tempting to speculate that reduced AR signaling encourages neoplastic transformation in breast epithelial cells harboring BRCA1 mutations. Reduced AR signaling has been implicated in the development of hereditary male breast cancer. Germline mutations in exon 3 of the AR gene were identified in three men 118 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. with breast cancer, all of which suffered from androgen insensitivity (Wooster et al., 1992; Lobaccaro et al., 1993). Two of the men were brothers and shared a codon 607A rg _K 3 ln mutation while the other unrelated man possessed a codon 608^^^;. mutation. Both mutations result in amino acid substitutions in the second zinc finger of the AR DBD that confer reduced DNA binding capacity and reduced transactivation potential on the mutant receptors (Poujol et al., 1997). In a related study, expression of an exon 3 deleted AR mRNA splice variant (i.e., A3AR) was found in female breast cancer tissues and cell lines, but not in normal tissues (Zhu et al., 1997). The encoded protein is predicted to lack the second zinc finger and to be functionally compromised due to an inability to bind DNA. Thus, decreased AR signaling, caused by poly-Q expansion, loss- of-function mutations, or by the aberrant expression of AR splice variants is correlated with breast cancer development in some individuals. A simple interpretation of this correlation is that androgen action is protective against or opposes malignant breast epithelial cell proliferation. As discussed below, there is some clinical and in vitro evidence to support this interpretation. Androgens and the control o f breast cancer cell proliferation Historically, androgens were used to treat advanced female breast cancers and generally produced a 20-30% objective regression rate among patients (Goldenberg et al., 1975). The effectiveness of such androgenic treatments presumably was due to AR- mediated inhibition of breast cancer cell growth. Indeed, AR expression has been 119 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. detected in nearly 80% of primary breast tumors (Lea et al., 1989; Isola, 1993). AR expression, moreover, has been positively correlated with that of the ER and PR, the presence o f which usually predicts a favorable response to hormonal therapy (Lea et al., 1989; Isola, 1993). The common co-expression o f receptors for androgens, estrogens, and progestins in primary breast tumors suggests that breast cancer cell proliferation is determined by a complex interplay of complementary and/or opposing endocrine signals. Studies o f breast cancer cell lines that express both the ER and AR, generally have demonstrated a stimulation of proliferation in the presence of estrogens and an inhibition o f proliferation in the presence of androgens. Basal or estrogen-stimulated proliferation of the ZR-75-1 cell line, for example, was strongly inhibited by physiological concentrations of androgens (Poulin et al., 1989; Birrell et al., 1995; Hackenberg and Schulz, 1996). This androgenic inhibition apparently is due to AR- mediated down-regulation of ER mRNA and protein levels in this cell line, an effect that is reversed by antiandrogens (Poulin et al., 1989). Androgens also have been shown to inhibit the proliferation of the ER and AR positive T47D cell line (Birrell et al., 1995). Interestingly, MCF-7 cells, which express relatively low levels of AR, were growth stimulated by androgens (Birrell et al., 1995). When, however, AR expression was markedly increased in these cells by the stable transfection of an AR expression vector, androgen-induced inhibition of proliferation was observed (Szelei et al., 1997). Thus, evidence from both clinical and in vitro laboratory studies tends to support the general 120 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. paradigm of breast cancer cell proliferation whereby ER signaling is stimulatory and AR signaling is inhibitory. The genetics o f breast cancer susceptibility and BRCA1 In the Western world, nearly 1 in 10 women develops breast cancer, and of these cases, about 5-10% are attributable to highly penetrant germline mutations in autosomal dominant susceptibility genes (reviewed by Rahman and Stratton, 1998). Women who inherit loss-of-function mutations in one such gene, BRCA1 (breast cancer susceptibility gene 1), have very high lifetime risks o f developing breast and ovarian cancers (i.e., nearly 80% and 60%, respectively) (Eeles and Kadouri, 1999). Most known germline BRCAl mutations result in prematurely truncated forms of the protein that are thought to be functionally inactive (Bertwistle and Ashworth, 1998 and references therein). Because of frequent loss of the wild-type allele in tumors from women of BRCA1-linked families (Comelis et al., 1995), BRCAl is considered a tumor suppressor gene in the classical sense. Despite this, however, somatic disease-causing mutations in BRCAl are rarely found in sporadic, non-hereditary breast cancers (Rahman and Stratton, 1998 and references therein). The BRCAl gene, located on the long arm of chromosome 17 (i.e., 17q21.3), comprises 24 exons from which a 7.8 kb mRNA is transcribed (Miki et al., 1994). BRCAl is a nuclear phosphoprotein of 1,863 amino acids, the expression of which is tightly regulated during the cell-cycle with a peak occurring in late G,/early S phase 121 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (reviewed by Bertwistle and Ashworth, 1998). BRCAl possesses an N-terminal RING finger domain and a C-terminal BRCT (BRCAl carboxy-terminal) domain (Rahman and Stratton, 1998). The RING finger is a cysteine-rich, zinc-binding motif, found in a diverse set of proteins, that is believed to function primarily in protein-protein interactions (Saurin et al., 1996). The BRCT domain, on the other hand, consists of two, non-identical, tandem repeats (i.e., BRCT repeats) and has been shown to activate transcription when fused to an autologous DNA binding domain (Chapman and Verma, 1996). Like RING finger motifs, BRCT repeats are found in a number of other proteins, several of which are involved in DNA repair and cell-cycle regulation. Putative functions o f BRCAl There is much evidence to implicate BRCAl function in DNA repair (reviewed by Irminger-Finger et al., 1998). Firstly, BRCAl interacts with Rad51, the eukaryotic homologue of RecA, a bacterial enzyme which mediates the ATP-dependent exchange of DNA strands during recombination. BRCAl and Rad51, furthermore, co-localize in discrete foci (i.e., within S phase nuclei) that are dispersed upon DNA damage (Scully et al., 1997a). Murine embryonic stem cells, moreover, that are nullizygous for Brcal (i.e., Brcal" /" ) are hypersensitive to DNA damaging agents [i.e., ionizing radiation (IR) and hydrogen peroxide], and compared to cells harboring a wt Brcal gene, are defective in their ability to carry out transcription-coupled repair of oxidative DNA damage (Gowen et al., 1998) as well as the homologous repair of double-strand breaks (DSBs) (Moynahan 122 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. et al., 1999). Finally, when wt BRCA1 expression was restored to BRCA1v' human breast cancer cells, an increased efficiency of DSB repair and a decreased sensitivity to IR was observed (Scully et al., 1999). A compelling argument for the involvement o f BRCA1 in cell-cycle control also has emerged in recent years. Aprelikova et al. (1999) have shown that BRCA1 interacts directly with pRb and that growth arrest induced by Brcal in mouse embryo fibroblasts occurs only in cells with an intact rb gene (i.e., rb+ /* or rb+ / '). Because BRCA1 binds preferentially to the hypophosphorylated form of pRb (Aprelikova et al., 1999) and to the histone deacetylases HDAC1 and HDAC2 (Yarden and Brody, 1999), it is possible that BRCA1 participates in growth arrest by facilitating the transcriptional repression of genes that promote cell proliferation (i.e., E2F-dependent genes). Ironically, mice nullizygous for Brcal die in utero (embryonic day 7.5) apparently from a failure o f proliferation (Feunteun, 1998 and references therein). Interestingly, Brcal'1 ' embryos demonstrate overexpression of the cyclin-dependent kinase p21W A F I/C IP 1 , the expression of which is regulated by p53. It has been suggested that in the absence of Brcal, cells fail to maintain genome integrity and accumulate DNA damage that ultimately elicits p53- dependent cell-cycle arrest (Feunteun, 1998). This notion is supported by the fact that the survival of Brcalv' embryos is significantly extended in the context of a p53~ *~ or p21'h background (Hakem et al., 1997). In addition to DNA repair and cell-cycle control, several lines of evidence suggest a direct role for BRCA1 in transcriptional regulation. As already mentioned, the BRCT- 123 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. containing C-terminus of BRCA1 has been shown to activate transcription in mammalian cells when fused to the GAL4 DBD (Chapman and Verma, 1996). A similar BRCA1 C- terminal-GAL4 DBD fusion protein, moreover, activated the transcription of a GAL4 responsive gene in cell-free reactions that required purified Pol II, basal transcription factors, and the coactivator PC4 (Haile and Parvin, 1999). BRCA1 has been co-purified with the Pol II holoenzyme in HeLa cell extracts (Scully et al., 1997b), most likely due to its interactions with RNA helicase A (Anderson et al., 1998). BRCA1, furthermore, has been shown to interact with p53 and to activate the transcription of p21W A F I /C T P I in a p53- dependent manner (Ouchi et al., 1998; Chai et al., 1998). Most recently, CBP/p300 has been shown to interact with BRCA1 and to potentiate its activation o f the RSV LTR in transfected HeLa cells (Pao et al., 2000). A rationalization for how BRCAI mutations lead to breast cancer development Based on both circumstantial and functional evidence, it seems clear that BRCAI participates in several diverse, though not unrelated, cellular processes including DNA repair, cell-cycle control, and transcriptional regulation. Partial as well as complete loss of wt BRCAI function, as evidently occurs with some frequency in BRCAI"'' heteroaygotic breast and ovarian cells due to loss of heterozygosity (Comelis et al., 1995), may predispose cells to mutagenesis by conferring on them a failure to recognize and repair DNA lesions. Concomitant loss of other genes critically involved in maintaining genome integrity and/or in controlling cell cycle progression (i.e., p53), due 124 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. to an accumulation o f DNA damage, could then lead to malignant transformation and unchecked cellular proliferation. In support of this, a high frequency of p53 mutations has been observed in BRCAI-associated breast cancers (Crook et al., 1997). In addition to predisposing cells to DNA damage, loss of BRCAI function would presumably increase cellular proliferation due to impaired pRb-mediated growth suppression and to reduced p21W A F 1 /C I P I expression by BRCAl-enhanced/p53-dependent transcriptional activation. In line with this logic is the fact that most germline BRCAI mutations result in C-terminally truncated proteins lacking the BRCT domain, which mediates interactions with p53 (Chai et al., 1999), the Pol II holoenzyme (Anderson et al., 1998), CBP/p300 (Pao et al., 2000), the histone deacetylases HDAC1 and HDAC2, as well as the Rb associated proteins RbAp46 and RbAp48 (Yarden and Brody, 1999). One puzzling feature of germline BRCAI mutations is that they result exclusively in tumors of the breast and ovary in affected women. Because these are largely hormone regulated tissues, loss of BRCAI function may cause a perturbation of endocrine signaling pathways. Indeed, in a recent study by Fan et al. (1999), BRCAI was shown to inhibit ligand-dependent ERa transactivation activity in breast and prostate cancer cell lines. This finding suggests that one function of wt BRCAI in mammary epithelial cells is to inhibit estrogen-induced cell proliferation mediated by ERa signaling. Thus, a failure to oppose estrogen-stimulated cell growth may be another consequence of BRCAI loss that contributes to neoplastic transformation. Loss of BRCAI function, therefore, 125 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. most likely causes deficits in a number of critical cellular processes, the cumulative impact of which ultimately leads to tumor formation in affected individuals. Does BRCAI influence AR signaling? While germline mutations in BRCAI are associated with dramatically increased risk for breast and ovarian cancers, not all female carriers develop tumors in these tissues. This variable penetrance suggests that other genes are involved in modifying BRCAl- associated risk. As has already been discussed, the AR appears to be one such modifier (Rebbeck et al., 1999). Whether or not BRCAI plays a direct role in AR signaling remains to be seen, though its apparent participation in ERa signaling and its defined interactions with proteins involved in NR-mediated gene regulation (i.e., CBP/p300, the Pol II holoenzyme, and HD AC 1/2) suggest that a functional link may exist. Due to our long standing interest in the AR and its role in cancer development in the prostate, another hormone regulated tissue, we assessed the effects of BRCAI on ligand-dependent AR transactivation activity in prostate and breast cancer cell lines. Our findings indicate a direct role for BRCAI in AR-mediated target gene activation in both cell types and may offer profound insight into the complex etiology of cancer development in the breast and prostate. 126 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. MATERIALS AND METHODS Plasmids Mammalian expression vectors and reporter plasmids pCMV-hAR, pSG5.HA- GRIP1, pSG5-SRC-la, pcDNA3.1-AIBl, pcDNA-hAR(Q)n , pcDNA-hAR (NTD-LBD), pcDNA-hAR (DBD-LBD), pMMTV-CAT, and ARR3 TK-CAT were described in Chapter 3, Materials and Methods. ER expression vector pSG5-ERa (Green et al., 1988) and reporter plasmid ERE-coll60-CAT (Webb et al., 1995), that contains a single consensus ER response element (ERE) from the vitellogenin A2 gene upstream of the collagenase promoter (-60/+63), were described previously. Bacterial expression vectors encoding the GST fusion proteins GST-AR NTD (aa 1-555), GST-GRIP1 (aa 5-765), GST-GRIP1 (aa 563-1121), and GST-GRIP1 (aa 1121-1462) were also described previously (Ma et al., 1999). BRCAI expression vector pcDNA-BRCAl was constructed by inserting a pBSK- lhFL (Chen et al., 1999b)-derived Notl-Xhol full-length BRCAI fragment into the corresponding restriction endonuclease sites of pcDNA3.1/mycHisC(-) (Invitrogen, Carlsbad, CA). Mammalian vectors expressing overlapping fragments of BRCAI were constructed by inserting PCR amplified Kpnl-Xhol BRCAI fragments into the reciprocal restriction sites of a pcDNA3.1 (+) vector that contained the AR Kozak sequence [see Chapter 3, Materials and Methods regarding the construction o f pcDNA-hAR (DBD- LBD)]. The PCR primers used were the following: BRCAI (aa 1-404), F7 and R8; 127 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. BRCAI (aa 364-772), F8 and R9; BRCAI (aa 768-1142), F9 and RIO; BRCAI (aa 1085- 1529), F10 and R11; BRCAI (aa 1490-1863), FI 1 and R12 (Table 3.1). All forward (i.e., sense) PCR primers contained a 5’ Kpnl restriction site followed by the SV40 T antigen NLS. All reverse (i.e., anti-sense) PCR primers contained a 5’ Xhol restriction site followed by a hemaglutinin A (HA) tag. PCR fragments were amplified from pcDNA- BRCA1 plasmid DNA using the Elongase en2yme mixture (Life Technologies, Rockville, MD). Cell culture and transfections Cells obtained from the American Type Culture Collection (Manassas, VA) were maintained in RPMI (PC-3, DU-145, and HBL-100) or DMEM (MCF-7) medium that contained 10% FBS. Approximately 24 h prior to transfection, 106 (PC-3, DU-145, and HBL-100) or 5 x 105 (MCF-7) cells were seeded into each 60 mm dish. Cells were transfected in serum-free conditions with the Lipofectamine reagent (Life Technologies) according to the manufacturer’s protocol using 20 pi of reagent per dish. When appropriate, the total amount of DNA per dish was held constant by the addition of pcDNA3.1(+) vector. Following transfection, cells were grown for 24 h (DU-145, HBL- 100, and MCF-7) or 48 h (PC-3) in RPMI medium (without phenol red) that contained 5% charcoal/dextran-stripped FBS (Gemini Bio Products, Calabasas, CA) and, where indicated, DHT (1 or 10 nM) or 10 nM 17(3-estradiol (E2) for the last 24 h of growth. 128 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Whole-cell extract preparation, total cellular protein quantification, and CAT activity measurement were done as described in Chapter 3, Materials and Methods. GST pull-down assays GST pull-down assays were performed by John Park of the laboratory of Dr. Michael Press (University of Southern California, Los Angeles, CA) as described in Chapter 3, Materials and Methods. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 129 RESULTS BRCAI coactivates AR signaling through AR AF-1 To assess the role (s), if any, of BRCAI in AR signaling, prostatic carcinoma cells (i.e., PC-3 cells) were cotransfected with AR and BRCAI expression vectors, as well as with the ARR3 TK-CAT probasin reporter. With 2.5 pg of transfected pcDNA-BRCAl vector, a 2.5 fold DHT-dependent potentiation of AR transactivation activity was observed (Fig. 4.1 A). The response of AR signaling to BRCAI was apparently linear (Fig. 4.IB), though further experiments are needed to confirm this finding. In order to reveal which AR domain primarily is responsible for functional interactions with BRCAI, AF-1-containing AR (NTD-DBD) or AF-2-containing AR (DBD-LBD) was co-expressed in PC-3 cells with BRCAI. The constitutive transactivation activity o f AR (NTD-DBD) on the ARR3 TK-CAT reporter was enhanced nearly 3 fold by BRCAI (Fig. 4.2). This result indicates that BRCAI can potentiate AR signaling through interactions with AR AF-1. BRCAI did not, however, enhance AR (DBD-LBD) activity in the presence or absence of ligand, though no DHT-dependent AF-2 activity was measured in this experiment (i.e., in the absence o f co-expressed BRCAI). Subsequent cotransfection experiments with the pl60 coactivators, which are known to coactivate AR AF-2, revealed that AR (DBD-LBD) is non-functional on the ARRjTK promoter (data not shown). Thus, the failure o f BRCAI to enhance AR AF-2 activity may be promoter specific. To explore this possibility, another cotransfection 130 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. experiment was done using the MMTV-CAT reporter, which has been shown to support AR AF-2 activity (Fig. 3.9). In this experiment, BRCAI failed to enhance the activity of AR (DBD-LBD) (Fig. 4.5B). This result indicates that BRCAI does not functionally interact with isolated AR AF-2. BRCAI and the p l6 0 coactivators synergistically potentiate AR signaling in prostate- and breast-derived cell lines. Due to the pivotal role of the p i60 coactivators in NR signaling, the effect o f co expressed BRCAI on pl60-mediated coactivation of the AR was assessed. PC-3 cells were cotransfected with expression vectors for the AR, BRCAI, and the p i60 coactivators GRIP1, SRC-la, and AIB1. As expected, BRCAI and GRIP1 individually enhanced AR transactivation of the ARR3 TK-CAT reporter about 2 and 3 fold, respectively (Fig. 4.3). When co-expressed, however, AR transactivation activity was enhanced 12 fold. This combined BRCAI-GRIP 1 coactivation of AR signaling was synergistic since it was greater than the additive effects of BRCAI and GRIP1 measured independently. Similar results were obtained when either SRC-la or AIB1 were used, suggesting a generic BRCAl-pl60 functional interaction (Fig. 4.3). To rule out the possibility that the observed BRCAI effects on AR signaling were specific to PC-3 cells, an additional prostate carcinoma cell line (i.e., DU-145), an SV40- transformed breast epithelial cell line (i.e., HBL-100), and a breast cancer cell line (i.e., MCF-7), were used in cotransfection experiments. In DU-145 cells, as in PC-3 cells, 131 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. BRCAI and GRIP1 individually enhanced DHT-dependent AR transactivation o f the ARRjTK-CAT reporter (i.e., 2.5 and 5 fold, respectively) (Fig. 4.4). Likewise, when BRCAI and GRIP1 were co-expressed in this cell line, a 14 fold synergistic coactivation of AR signaling was observed. In HBL-100 cells, the effects of BRCAI were more dramatic. For example, BRCAI alone potentiated AR transactivation activity greater than 12 fold. In combination with GRIPl, moreover, BRCAI resulted in a nearly 45 fold enhancement of AR signaling. In MCF-7 cells, on the other hand, the effects of BRCAI generally were more modest. In this cell line, BRCAI and GRIPl individually potentiated AR activity less than 2 fold. Together, however, they did result in a 5 fold coactivation of AR signaling, consistent with observations made in the other cell lines. The relatively small BRCAI effects seen in MCF-7 cells may be due to high endogenous p i60 coactivator levels (Anzick et al., 1997). Together, the data presented in Figures 4.3 and 4.4 strongly support a functional role for BRCAI in pl60-mediated AR signaling. BRCAI enhances AR AF-2 activity in the presence o f p i 60 coactivator To further define BRCAl’s functionality in AR signaling, its ability to influence pl60-mediated coactivation of either AR (NTD-DBD) (i.e., AF-1) or AR (DBD-LBD) (i.e., AF-2) was assessed. As expected, both BRCAI and GRIPl individually enhanced the constitutive activity of AR (NTD-DBD) on the ARR3 TK-CAT reporter (Fig. 4.5A). Together, moreover, they resulted in nearly a 10 fold coactivation of AR (NTD-DBD) signaling. The response of AR (NTD-DBD) to BRCAI and GRIPl co-expression, 132 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. therefore, closely resembles that of liganded wt AR and demonstrates the functional importance of AR AF-1 in BRCA1-mediated coactivation of the AR. As mentioned above, BRCA1 failed to functionally interact with isolated AR AF-2 (Figs. 4.2 and 4.5B). BRCA1 did, however, potentiate GRIP 1-mediated coactivation of AR (DBD-LBD) on the MMTV-CAT reporter by 2 fold (Fig. 4.5B). This finding indicates that BRCA1 can functionally enhance DHT-dependent AR AF-2 activity through bound p i60 coactivator. Thus, it appears that the holo-AR LBD makes direct contacts with the p i60 coactivator which, in turn, recruits BRCA1. BRCAI physically interacts with the AR and GRIP 1 in GST pull-down assays To determine whether or not BRCAI makes physical contacts with the AR and/or GRIP1, GST pull-down experiments were performed by John Park of the laboratory of Dr. Michael Press (University of Southern California, Los Angeles, CA) in which in vitro translated and 3 5 S-labeled BRCAI was incubated with immobilized GST-AR (aa 1-555) or with various GST-fiised fragments of GRIP1 (i.e., aa 5-765, aa 563-1121, or aa 1122- 1462). In these experiments, BRCAI interacted with GST-AR (1-555) and with GST- GRIP1 (1122-1462) (Fig. 4.6B). This result indicates that in addition to functional interactions, BRCAI participates in physical interactions with the AR and with GRIP1. To identify which region(s) of BRCAI interacts with the AR NTD and with the C-terminus o f GRIP1, overlapping 3 sS-labeled fragments of BRCAI were incubated with GST-AR (1-555) or GST-GRIP1 (1122-1462). An N-terminal 404 aa fragment of 133 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. BRCAI interacted with both GST-fused proteins (Fig. 4.6C). An adjacent BRCAI fragment (i.e., aa 364-772), moreover, demonstrated weak binding to GST-GRIP1 (1122- 1462). In total, the in vitro interaction data presented in Figures 3.6B and 4.6 are highly suggestive of an AR/GRIP1/BRCA1 ternary complex mediated by reciprocal interactions among the AR NTD, the GRIP1 C-terminus, and an N-terminal region of BRCAI. Preliminary attempts to confirm the presence of such a complex in vivo by immunoprecipitation experiments have been unsuccessful (data not shown). This failure, however, may be due to low endogenous levels of BRCAI in the cell lines used for these experiments. BRCAI potentiates pl60-mediated ER signaling in prostate- and breast-derived cell lines To assess the role of BRCAI in ER signaling, PC-3 cells were transiently cotransfected with ERa and BRCAI expression vectors, as well as with the ERE-coll60- CAT reporter. BRCAI had no effect on E2-dependent ERa transactivation, even when 5 pg of pcDNA-BRCAl were used in the transfection (Fig. 4.7). Similar results were obtained in the HBL-100 and MCF-7 cell lines (Fig. 4.8). In the DU-145 cell line, however, a reproducible 2 fold repression of ERa transactivation was observed (Fig. 4.8). Thus, in general, BRCAI does not appear to strongly modulate E R a signaling, though it may exhibit some modest cell-specific effects. Because BRCAI has been shown to physically and functionally interact with the p i60 coactivator GRIP1, its ability to influence GRIP 1-mediated coactivation of ERa 134 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. signaling was assessed in the prostate- and breast-derived cell lines. Cells were transiently cotransfected with ERE-co 1160-CAT and expression vectors for ERa, BRCAI, and GRIP1. In PC-3, DU-145, and HBL-100 cells, BRCAI enhanced E2-dependent, GRIP 1-mediated ERa transactivation activity by about 2 fold (Fig. 4.8). Interestingly, the co-expression o f exogenous GRIP1 reversed the repressive effects o f BRCAI in DU- 145 cells. In MCF-7 cells, on the other hand, BRCAI had little or no effect on GRIP1- mediated ERa coactivation. It should be mentioned that E2-independent background stimulation o f the ERE-coll60-CAT reporter was very high in this cell line, limiting the measurable range o f GRIP1 coactivation (i.e., 1.5 fold). It may be, therefore, that co expression of GRIP1 and BRCAI leads to squelching of the reporter in MCF-7 cells. Nevertheless, the overall conclusion from these experiments is that BRCAI can markedly potentiate ER a signaling in the presence o f co-expressed p i60 coactivator in some cell lines. Exogenous BRCAI expression masks the AR poly-Q effect To assess the impact of BRCAI expression on the characterized AR poly-Q effect (see Chapter 3), PC-3 cells were transiently cotransfected with the ARR3 TK-CAT reporter and expression vectors for BRCAI, GRIP1, and the AR poly-Q variants. As expected from previous experiments (Fig. 3.13), about a 20% reduction in DHT- dependent AR transactivation activity was observed with increasing poly-Q size from 9- 42 glutamine residues (Fig. 4.9). This AR poly-Q effect was not apparent, however, in 135 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the presence o f transfected BRCAI. GRIP 1-mediated AR(Q)n transactivation, moreover, which was reduced 40-50% by increased poly-Q (Fig. 3.14), was not effected by AR poly-Q size in the presence of co-expressed BRCAI (Fig. 4.9). Thus, BRCAI may alleviate the inhibition imposed by increased A JR . poly-Q size on pl60-mediated coactivation o f the AR. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 136 FIGURE 4.1 LEGEND BRCAI potentiates AR transactivation activity. (A) PC-3 cells were transiently cotransfected with 2.0 pg ARR3 TK-CAT, 50 ng pCMV-hAR, and increasing amounts of pcDNA-BRCAl as indicated. Total transfected DNA was held constant by the addition of pcDNA3.1(+) vector when appropriate. Twenty-four hours after transfection, cells received fresh medium that contained 5% charcoal/dextran-stripped FBS and 1 nM DHT or ethanolic vehicle. Twenty-four hours later, whole-cell extracts were prepared and assayed for CAT activity as described in Materials and Methods. Data presented are the mean + SE of three independent dishes. Fold was measured relative to DFIT-dependent AR activity with no transfected BRCAI. (B) An alternate graphic representation of the data presented in A with the amount (pg) of transfected pcDNA-BRCAl depicted on a linear scale. 137 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 4.1 600 >» > 1 400 to I - < O 0) ■ J = 2 0 0 H B □ no horm one 1 nM DHT 0.1 0.5 pcDNA-BRCAl (jig) >» • w m m > O (0 I - < O 0) > iS o pcDNA-BRCAl (jig) 2.5 750 no horm one 1 nM DHT 500 250 0 1 2 138 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. fold FIGURE 4.2 LEGEND BRCAI works through AR AF-1. PC-3 cells were transiently cotransfected with 50 ng pCMV-hAR, 10 ng pcDNA-hAR (NTD-DBD), or 0.5 pg pcDNA-hAR (DBD-LBD), 2.0 pg ARRjTK-CAT, and 2.5 pg pcDNA-BRCAl as indicated. Total transfected DNA was held constant by the addition of pcDNA3.1(+) when appropriate. Transfection conditions were as described in the legend to Figure 4.1. Whole-cell extracts were prepared and assayed for CAT activity as described in Materials and Methods. Data presented are the mean + SE of three independent dishes. AR (NTD-DBD) is a constitutive activator of ARRjTK-CAT, and thus, potentiation of its activity by BRCAI was ligand-independent. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 139 FIGURE 4.2 500- £ 400- > 300- o CO < C O a > 200- £ ioo- | | no hormone PC-3 DHT r m arr3tk-cat AR(wt) AR (NTD-DBD) AR (DBD-LBD) BRCA1 + + + + + + + + + + + + + ................................... - + + + + - + + + - - + + - + + + + + + - + + 140 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 4.3 LEGEND Synergistic potentiation o f AR signaling by BRCAI and the p i60 coactivators. PC-3 cells were transiently cotransfected with 2.0 pg pSG5-GRIPl, pcDNA3.1-AJBl, or pSG5-SRC-la, 2.0 pg ARR3 TK-CAT, 25 ng pCMV-hAR, and 2.5 pg BRCAI as indicated. Total transfected DNA was held constant by the addition of pcDNA3.1(+) vector when appropriate. Transfection conditions were as described in the legend to Figure 4.1. Whole-cell extracts were prepared and assayed for CAT activity as described in Materials and Methods. Data presented are the mean + SE of three independent dishes. In each case, AR transactivation activity in the presence of both BRCAI and p i60 coactivator was greater than the additive effects of BRCAI and p i60 coactivator assayed separately. 141 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 4.3 50 0 >, 400H [> « 3 0 0 - i- < o | 200- 0 ) * " 1 0 0 ^ PC-3 DHT - + + + + + + + + ARR3 TK-CAT + + + + + + + + + AR + + + + + + + + + BRCA1 - + - + - + - + GRIP1 + + - - - - AIB1 — — — — — + + - — SRC-1 a — ^ 4 " 142 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 4.4 LEGEND Potentiation of AR signaling by BRCAI occurs in both prostate- and breast-derived cell lines. Prostate cell line DU-145 and breast cell lines HBL-100 and MCF-7 were transiently cotransfected with 2.0 pg ARR3 TK-CAT, 25 ng pCMV-hAR, 2.0 pg pSG5- GRIP1, and 2.5 pg pcDNA-BRCAl as indicated. Transfected DNA was held constant by the addition of pcDNA3.1(+) vector when appropriate. Transfection conditions were as described in the legend to Figure 4.1 except that 10 nM DHT was used. Whole-cell extracts were prepared and assayed for CAT activity as described in Materials and Methods. Data presented are the fold + SE of three independent dishes relative to DHT- dependent AR activity with no transfected BRCAI or GRIP1. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 143 hOS FIGURE 4.4 PIO* + + + + + + + + + ' + + + « + + + + « • , + + + + + ' + + + + i + + + « + + + + ' • ■ + + + + + ' + + + + » + + + i + + + + « « • “ i— & i? X < < S: < Q O X ^ £ ® S co X X < I u. o X X in Z > Q 144 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 4.5 LEGEND BRCAI potentiates GRIP 1-mediated coactivation o f AR AF-1 and AF-2. (A) PC-3 cells were transiently cotransfected with 2.0 pg ARR3 TK-CAT, 10 ng pcDNA-AR (NTD- DBD), 2.0 pg pSG5-GRIPl, and 2.5 pg pcDNA-BRCAl as indicated. Total transfected DNA was held constant by the addition of pcDNA3.1(+) vector when appropriate. Transfection conditions were as described in the legend to Figure 4.1 except that no hormone was added during the final 24 h of growth. Whole-cell extracts were prepared and assayed for CAT activity as described in Materials and Methods. Data presented are the mean +SE o f three independent dishes. (B) PC-3 cells were transiently cotransfected with 2.0 pg MMTV-CAT, 1.0 pg pcDNA-hAR (DBD-LBD), 2.0 pg pSG5-GRIPl, and 2.5 pg pcDNA-BRCAl as indicated. Transfection conditions were as described in the legend to Figure 4.1. Whole-cell extracts were prepared and assayed for CAT activity as described in Materials and Methods. Data presented are the mean +SE of three independent dishes. 145 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 4.5 + + + + + + + + + • + + « + + + + * • AifAjiae i v o aAjieiaj c c < A)iAj;oe i v o 3A i;e|aj 146 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. BRCA1 FIGURE 4.6 LEGEND BRCAI interacts with the AR NTD and the GRIP1 C-terminus. (A) Schematic diagrams of the structural/functional domains o f the AR, GRIP1, and BRCAI drawn to scale. Refer to Figures 3 and 5 for domain descriptions of the AR and GRIP1, respectively. Domains of BRCAI: RING, zinc-finger domain; NLS, nuclear localization signals; BRCT, BRCAI carboy terminus. Numbers represent relative amino acid positions. (B) Glutathione-Sepharose bound GST, GST-AR (1-555), GST-GRIP1 (5-765), GST-GRIP1 (563-1121), or GST-GRIP (1122-1462) was incubated with 3 5 S-labeled BRCAI transcribed and translated in vitro from pcDNA3.1 vector encoding full-length BRCAI. Bound BRCAI was eluted and analyzed by SDS-PAGE and autoradiography; shown for comparison is 10% of the total labeled protein incubated in each binding reaction (i.e., 10% input). (C) Glutathione-Sepharose bound GST, GST AR (1-555), or GST-GRIP1 (1122-1462) was incubated with 3 5 S-labeled, overlapping fragments o f BRCAI. Bound proteins were analyzed as above. 147 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 4.6 Q AR I ■ AF-1 P G I T AF-2 1 NTD bHLH/PAS OBD LBD NID 919 GRIP1 BRCAI TEE 563 765 CID/AD1 RING NLS H E 500 1863 GRIP1 BRCT BRCA1 148 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 4.7 LEGEND BRCAI does not potentiate or repress ER signaling in PC-3 cells. (A) PC-3 cells were transiently cotransfected with 1.0 pg ERE-coll60-CAT, 100 ng pSG5-ERa, and increasing amounts of pcDNA-BRCAl as indicated. Total transfected DNA was held constant by the addition of pcDNA3.1(+) vector when appropriate. Transfection conditions were as described in the legend to Figure 4.1 except that 10 nM E2 was used. Whole-cell extracts were prepared and assayed for CAT activity as described in Materials and Methods. Data shown are the mean + SE o f three independent dishes. (B) An alternate graphic representation of the data presented in A with the amount (pg) of transfected pcDNA-BRCAl depicted on a linear scale. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 149 FIGURE 4.7 300 >* > s o (0 0) □ no hormone ■ ■ 10 nM E2 200 - ■ 5 1 0 0 - (0 0) pcDNA-BRCAl (pg) B 3 0 0 no hormone 10 nM E2 2 0 0 - 100 - 1 2 4 5 pcDNA-BRCAl (pg) 150 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 4.8 LEGEND BRCAI potentiates pl60-mediated coactivation of ER signaling in prostate- and breast- derived cell lines. Cells were transiently cotransfected with 1.0 pg ERE-coll60-CAT, 25 ng pSG5-ERa, 2.0 pg pSG5-GRIPl, and 2.5 pg pcDNA-BRCAl as indicated. Total transfected DNA was held constant by the addition of pcDNA3.1(-f-) w'hen appropriate. Transfection conditions were as described in the legend to Figure 4.7. Whole-cell extracts were prepared and assayed for CAT activity as described in Materials and Methods. Data presented are fold + SE of three independent dishes relative to E2- dependent ER activity with no transfected GRIP1 or BRCAI. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 151 FIGURE 4.8 + + + + + + + + + • + + + « + + + + • • . + + + + + ' + + + + * + + + • + + + + * ■ . + + + + +' + + + + ■ + + + « + + + + ' « _ + + + + +' + + + + • + + + « + + + + > « o o o o o> CO C M l - U l < Q - < PIO* g 5 § s O 0 1 U i D C L U 152 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. P C -3 D U -145 H BL-100 MCF-7 FIGURE 4.9 LEGEND BRCAI masks the AR poly-Q effect. PC-3 cells were transiently transfected with 2.0 pg ARR3 TK-CAT, 25 ng of pcDNA-hAR(Q)n , 2.0 pg pSG5-GRIPl, and 2.0 pg pcDNA- BRCA1 as indicated. Total transfected DNA was held constant by the addition of pcDNA3.1(+) vector when appropriate. Transfection conditions were as described in the legend to Figure 4.1. Whole-cell extracts were prepared and assayed for CAT activity as described in Materials and Methods. CAT activities were normalized to AR(Q)n expression levels as determined by ligand binding assays (see Figure 3.11). Data presented are the mean + SE of three independent dishes. 153 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 4.9 >» o (S H < O 0) > iS 0) AR(Q)n BRCAI 2 5 0 - 150 - * ■ 6 0 -r 9 21 29 42 9 21 29 42 AR(Q)n BRCA1/GRIP1 9 21 29 42 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DISCUSSION BRCAI is an AR coactivator In this study, we investigated the role of BRCAI in AR signaling. In both breast and prostate cancer cell lines, BRCAI potentiated ligand-dependent AR transactivation of an androgen-responsive probasin reporter gene. This finding demonstrates a functional link between the AR and BRCAI and may partially explain why BRCAI mutations lead to the development o f breast tumors in some women. Given that androgen action is largely anti-proliferative in mammary epithelial cells (see Introduction), an assumption based on somewhat limited clinical and in vitro evidence, loss o f BRCAI function potentially would alleviate androgen-induced growth suppression by reducing AR- mediated signaling. Long-term exposure to reduced AR signaling (i.e., increased cellular proliferation), compounded by deficits in DNA repair and in cell-cycle control, would be expected to foster malignant transformation of breast epithelial cells, leading to tumor development. In prostate epithelial cells, on the other hand, loss of BRCAI function might be expected to protect against deleterious proliferation since AR signaling promotes growth of this cell type. Indeed, one could argue that BRCAI overexpression would contribute to prostate cancer risk by increasing AR ‘exposure’ over time, though it is difficult to reconcile the tumor suppressor functions of BRCAI (i.e., DNA repair, cell-cycle regulation) with such a proposal. Nevertheless, BRCAI mutations are infrequently found 155 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. in primary prostate tumors (Uchida et al., 1999), and though it remains a contentious issue, male carriers o f germline BRCAI mutations do not appear to be at increased risk for prostate cancer development (Langston et al., 1996; Hubert et al., 1999; Wilkens et al., 1999; Sinclair et al., 2000). The molecular basis fo r BRCAI function in AR coactivation In transfection experiments, BRCAI modestly enhanced the transactivation activities of wt AR and the AF-1-containing AR NTD (Figs. 4.1 and 4.2). Co-expression o f p i60 coactivator, moreover, markedly increased the 'BRCAI effect’ on the activities of both proteins (Figs. 4.3 and 4.5). This observation is clearly indicative of functional cooperativity between the pi 60 coactivators and BRCAI in mediating AR transactivation. It further may imply participation by BRCAI in the coordination and/or stabilization of the NR coactivator complex (see Chapter 3, Introduction). This notion is supported by the fact that BRCAI physically interacts with both the AR and GRIP1 in in vitro pull-down assays (Fig. 4.6). Specifically, BRCAI, through its RING finger- containing N-terminus, interacts with the AR NTD and with the C-terminus of GRIP1. Binding of BRCAI to the AR NTD indicates that, as with GRIP1 (Chapter 3), the AR NTD may directly recruit BRCAI to the target promoter. It is interesting that BRCAI interacts with the C-terminus of GRIP1 since this is the same region of GRIP1 that interacts with the AR NTD (Fig. 3.6). Taken together, these data are highly consistent 156 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. with the formation of an AR/GRIP1/BRCA1 ternary complex, one that possesses potent ligand-dependent transactivation activity on androgen responsive promoters. Because BRCAI apparently participates in a number of critical cellular processes and because no endogenous catalytic activity(s) has been yet associated with BRCAI, it has been proposed by Irminger-Finger et al. (1999) that “...BRCAI functions as an active vehicle for selected cellular proteins to converge on.” In this view, BRCAI serves as a dynamic scaffold-like binding protein around which complexes of various critical proteins interact depending on the cellular need. This model is attractive because it reconciles evidence from a plethora of studies implicating BRCAI function in diverse cellular activities. In terms of AR-mediated transcriptional regulation, BRCAI seems to serve as an efficient facilitator of activated transactivation. Based on our findings and those of others, it achieves this ‘transcriptional facilitation’ perhaps by linking the N- terminally associated NR coactivator complex to the C-terminally associated Pol II- containing PIC. Accordingly, BRCAI likely plays a critical role in modulating the effects of endocrine signals on cells by increasing the efficacy and accuracy of receptor- mediated transcriptional events. BRCAI-mediated inhibition o f ERa signaling? BRCAI previously was shown to repress ERa signaling (Fan et al., 1999). In that study, E2-dependent ERa transactivation of an ERE-containing luciferase reporter gene (i.e., ERE-TK-Luc) was completely inhibited by exogenously expressed BRCAI in both 157 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. breast and prostate cancer cell lines. Based on their findings, the authors logically concluded that loss o f BRCA1 function in mammary epithelial cells could promote carcinogenesis due to unopposed ERa-mediated signaling related to cellular proliferation. In light of our results with the AR, those of Fan et al. (1999) suggest that BRCA1 might discriminate between the two steroid hormone signaling pathways. In one instance, inhibiting signaling by ERa, and in another, activating signaling by the AR. To address this potentially important issue, a series of transfection experiments were performed to assess the influence o f BRCA1 on ERa signaling. In PC-3 cells, BRCA1 failed to demonstrate any effect on E2-dependent ERa-mediated transactivation, even when relatively large amounts of the BRCA1 expression vector were transfected (Fig. 4.7). Similar results were obtained in the HBL-100 and MCF-7 breast-derived cell lines (Fig. 4.8). In DU-145 cells, though, a roughly 50% BRCA1-mediated inhibition of ERa transactivation activity was observed, partially corroborating the findings of Fan et al. (1999). In obvious contrast to their findings, however, BRCA1 markedly potentiated ERa transactivation activity in the presence of co-expressed p i60 coactivator, even in DU-145 cells. This finding is consistent with a role for BRCA1 in facilitating NR- mediated transactivation, and casts considerable doubt on its role as a specific inhibitor of ERa signaling. That is not to say, however, that BRCA1 is incapable of participating in the repression o f transcription, even of that mediated by NRs. After all, BRCA1 has been shown to interact with components of the histone deacetylase complex (Yarden and Brody, 1999). 158 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. While our findings generally support a role for BRCA1 in both AR and ERa transactivation, some differences between BRCA l’s function in AR and ERa signaling are apparent. For example, BRCA1 enhanced ligand-dependent AR transactivation activity 2-3 fold in the absence of co-expressed p i 60 coactivator (Fig. 4.3). Presumably, this was due to direct recruitment of BRCA1 to the target promoter by the AR NTD (see Fig. 4.6). In contrast, BRCAl did not enhance ERa activity in the absence of co expressed p i60 coactivator [compare Fig. 4.8 (i.e., PC-3 cell data) to Fig. 4.3] and this may imply that no direct interaction occurs between BRCAl and ERa. This would not be too surprising considering the almost complete lack of homology between the AR and ERa NTDs (Tsai et al., 1999). BRCAl did potentiate pl60-mediated coactivation of both receptors, though its effect was considerably stronger in pl60/AR signaling (i.e., ~4 fold vs. ~2 fold with p 160/ERa). This result likely reflects differences in the stability of BRCAl interactions between the two receptors. With ERa, it appears that BRCAl is recruited exclusively through interactions with bound pi 60 coactivator. Regardless of the mechanism of recruitment, however, BRCAl appears to provide the same or similar enhancement function to both ERa and AR-mediated transactivation events. Reconciling the differences The data presented herein provide some compelling evidence that BRCAl participates directly in AR and ERa signaling. They further shed some light on the molecular basis of the obviously complex etiology of hereditary breast cancer. With 159 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. respect to the function of BRCAl in ERa signaling, our results clearly contradict the only other study that has been published on the subject (i.e., Fan et al., 1999). A close comparison of the two studies revealed that while similar reagents, cell lines, and techniques were used, Fan and colleagues transfected nearly 30-fold more ERa expression vector per equivalent experiment than did we. The ratio, moreover, of transfected ERa expression vector to reporter plasmid in their experiments was 1:1 compared to a ratio of 1:20 in ours. Lastly, Fan et al. (1999) utilized a 1 pM concentration of E2 to induce ERa responsiveness, nearly a 1000 fold excess considering the affinity of E2 to the receptor (i.e., Kd = <1 nM). While far from conclusive, these differences may suggest that the potent BRCAl-mediated inhibition of ERa signaling observed by Fan et al. (1999) was artifactual, possibly the result of squelching effects. Obviously, further studies need to be done to confirm or deny this assertion. Our findings imply that a reduction in BRCAl function would lead to decreased ERa signaling, and therefore, to an inhibition of estrogen-stimulated cellular proliferation. Along these lines, one could argue that BRCAl loss would be protective against breast cancer development. Despite this putative protective effect, however, concurrent accumulation of DNA damage, relaxed cell-cycle control, and diminished androgen-mediated growth inhibition due to reduced BRCAl function would still encourage malignancy. Tumor progression driven by these BRCAl-associated cellular insults could lead to a loss of estrogen regulation altogether. Indeed, breast tumors from women with BRCAl mutations are significantly less likely to be ER positive compared 160 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. to tumors from women without mutations (Karp et al., 1997). It also should be kept in mind that ER expression is generally associated with well-differentiated tumors that are likely to respond favorably to endocrine therapy (Isola et al., 1993). Thus, it is possible that reduced ERa signaling related to loss of BRCAl function could encourage dedifferentiation and the development of a more aggressive tumor phenotype. To end at the beginning As was discussed at the beginning of this chapter, the AR appears to be a modifier of breast cancer penetrance among BRCAl mutation carriers. Rebbeck et al. (1999) showed that carriers with at least one long AR CAG allele (i.e., > 28 CAG repeats) were diagnosed with breast cancer earlier than women with only short alleles. Interestingly, this ‘CAG effect’ was not observed among women without defined BRCAl mutations (Spurdle et al., 1999; Given et al., 2000). This implies that only in cells with diminished BRCAl function, do differences in AR transactivation activity, conferred by variations in poly-Q length, become manifest phenotypically. Our findings demonstrate that BRCAl markedly enhances AR signaling. It could be, therefore, that in cells with normal BRCAl levels (i.e., BRCAl*'*), AR activity is maintained sufficiently high (i.e., over a threshold level) so that poly-Q-mediated differences in AR transactivation activity are less relevant to cellular transformation. Indeed, our findings may indirectly support this hypothesis. BRCAl overexpression in PC-3 cells abolished the inhibitory effect of increasing poly-Q length on pl60-mediated coactivation o f the AR (compare Figs. 3.14 161 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. and 4.9). Thus, in women with germline BRCAl mutations, breast epithelial cells are under reduced androgen-mediated growth inhibition and tumors develop more rapidly in those women expressing less efficient ARs. Ultimately, these studies, in total, may suggest that germline BRCAl mutations predispose cells to a unique cancer etiology that involves AR-mediated pathways. Cells not harboring BRCAl mutations must identify and commit to pathways o f progression that may or may not involve the AR. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 162 CHAPTER 5 Identification and cloning o f androgen receptor interacting factors INTRODUCTION We and others have shown that Caucasian men who inherit short AR CAG alleles are at increased risk for the development of advanced prostate cancer (see Chapter 2). We and others further have shown that short AR CAG alleles encode receptors with increased transactivation potential relative to those encoded by long AR CAG alleles (see Chapter 3). These observations have led to the hypothesis that prostatic epithelial cells exposed to the increased activity of receptors with short poly-Q tracts (i.e., encoded by short CAG alleles) are susceptible to malignant transformation due to increased androgen-dependent cell division (Ross et al., 1998). To reveal the molecular basis for poly-Q-dependent modulation of AR transactivation activity, a two-tiered approach was undertaken. In one study, detailed in Chapter 3, AR interactions with the p i60 coactivators were explored. The take-home message from these experiments was that increasing poly-Q length inhibits pi 60- mediated coactivation of the AR (Irvine et al., 2000). Our interpretation of this finding, on a molecular level, was that longer poly-Q tracts exert allosteric stresses on the AR NTD tertiary structure that indirectly affect p i60 coactivator binding, which occurs 163 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. downstream o f the poly-Q tract. Perturbed AR NTD-pl60 interactions, therefore, may be one explanation for the ‘poly-Q effect’ on AR transactivation activity. In addition to assessing the impact of AR poly-Q variation on p i60 coactivator function, a parallel study, designed to identify novel factors that bind directly to the poly- Q tract, was initiated. The rationale for pursuing this approach stems from the observation that stretches of polyglutamine can form so-called polar zipper protein- protein interaction motifs (Stott et al., 1995). Thus, assuming that the AR poly-Q acts as an interface for protein binding, the yeast two-hybrid system was used to screen a human testis cDNA expression library for putative poly-Q interacting factors. The design of the poly-Q-containing AR NTD bait protein and the results of the two-hybrid screen are described herein. The yeast two-hybrid system The yeast two-hybrid system essentially was developed around a single biochemical phenomenon: certain transcriptional activators, like the yeast GAL4 protein, are modular in nature such that their DNA binding and transactivation capacities reside in separate domains which can function autonomously (Ma and Ptashne, 1987). When fused to two proteins that were known to interact, Fields and Song (1989) showed that these separate GAL4 domains could activate the transcription o f a GAL4-responsive reporter gene in yeast (i.e., S. cerevisiae). Thus, interactions between the ‘two GAL4 hybrid’ proteins brought the GAL4 AD to the target promoter and resulted in measurable 164 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. reporter gene activation. This novel two-hybrid system, therefore, provided a genetic means to assay protein-protein interactions in vivo. Chien et al. (1991) demonstrated the utility o f this system in identifying novel proteins that interact with a protein of interest (i.e., the yeast SIR4 protein). Using a GAL4 DBD-fiised SIR4 bait protein, they screened a yeast genomic library encoding GAL4 AD-fused proteins and identified SIR4- interacting protein (SFI1). The yeast-two hybrid system, therefore, is a powerful genetic tool that can be used to study interactions between known proteins or to identify new partners o f a protein o f interest. Despite often being characterized as high-risk ‘fishing expeditions’, yeast two- hybrid screens have been used quite successfully in recent years to dissect innumerable genetic pathways that govern cellular decision-making capabilities. Signal transduction, cell-cycle regulation, and apoptosis pathways, for example, all are dependent on protein- protein interactions to elicit appropriate cellular responses and all, to some extent, have been elucidated by yeast-two hybrid protocols (e.g., Reed, 1997). While many successes have been reported, it is prudent to remember that like all techniques, the yeast two- hybrid screen suffers from certain limitations. First and foremost, the bait protein must not activate reporter gene expression by itself. Also, the bait protein must fold properly to provide a ‘native’ target for putative interacting factors. Another limitation is that weak interactions (i.e., with equilibrium dissociation constants below 10-50 pM) are not detectable by most two-hybrid systems (Brent and Finley, 1997). Finally, some interactions may be identified that are presumably physiologically irrelevant (e.g., the two 165 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. proteins never occur in the same cellular compartment) and therefore, must be ignored. Overall, the yeast-two hybrid screen provides an effective means by which to identify and clone genes for proteins that interact with a protein o f interest, though caution should be observed during its design and throughout the interpretation of results. Proteins that interact with the AR Many proteins have been identified as AR interacting proteins (see summary table: http://www.mcgill.ca/androgendb). Most of these were identified in yeast-two hybrid screens based on their abilities to interact with NR LBDs in a ligand-dependent manner. The p i60 coactivator, GRIP1, for example, was isolated from a murine embryo cDNA library using the holo-GR LBD as bait (Hong et al., 1996). Like GREP1, several of these AR interacting proteins or ARIPs act as coactivators of AR signaling while others, act as corepressors. Relevant to the present study, a subset of ARIPs were isolated based on their interactions with the AR NTD. The 160 kDa AR-associated protein 160 (ARA160), identical to TATA modulatory factor (TMF), was identified in a far-Western blot screen of a human testis cDNA library using an AR NTD probe (Hsiao et al., 1999). This protein interacts with the AR NTD in a ligand-dependent fashion and enhances AR activity in transfected cells. Because ARA160 also enhances GR and PR activities, it is likely to be a general steroid receptor AF-1 coactivator. Another AR N-terminal interacting protein is the Ets transcription factor ERM. Members of the Ets family of 166 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. transcription factors bind cognate cis-elements (i.e., Ets binding sites) and cooperate with AP-1 to activate the expression of several genes encoding matrix metalloproteinases (MMPs) (Schneikert et al., 1996). Ligand-activated AR, through its NTD, interacts with DNA-bound ERM and represses MMP gene expression (Schneikert et al., 1996). Interestingly, ERM-mediated AR repression of MMP expression occurs in the absence of AR DNA binding. Perhaps, from our standpoint, the most intriguing AR NTD interacting protein to be characterized to date is Ras-related nuclear protein/AR-associated protein 24 (Ran/ARA24). Ran/ARA24 originally was characterized as a small nuclear GTPase with cellular roles in nuclear protein import, nuclear RNA export, cell cycle progression, mitotic regulation, DNA synthesis, and RNA synthesis and repair (reviewed by Rush et al., 1996). Hsiao et al. (1999) recently reported an additional role for Ran/ARA24 in AR signaling. Ran/ARA24 was isolated from a human brain cDNA library in a yeast two- hybrid screen using the AR NTD poly-Q region as bait. Functional characterization revealed that Ran/ARA24 coactivated the AR in a ligand-dependent manner in transfected PC-3 cells and that the coactivation efficiency of Ran/ARA24 decreased with increasing AR poly-Q length. While the data presented by Hsiao et al. (1999) are, for the most part, convincing, the authors failed to show that the loss o f Ran/ARA24-mediated coactivation (i.e., with increasing AR poly-Q size from Q25-Q49) was not due to diminished AR protein levels. In light of our findings that show dramatically reduced AR(Q)J 0 protein levels in 167 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. transfected PC-3 cells compared to AR(Q)2 9 (see Chapter 3), this is a real possibility. In addition to this potential problem, it is difficult to reconcile the diversity of functions attributed to Ran/ARA24. In terms of the AR, it may be that Ran/ARA24 participates directly in AR nuclear import, and therefore, may preferentially import receptors with shorter poly-Q tracts. Obviously, more studies are needed to clarify the exact nature of Ran/ARA24-AR interactions. The nature o f hypothesis-generating research In order to move forward in one’s understanding o f a particular scientific question, oftentimes it is necessary to perform experiments, the possible outcomes of which are difficult to predict. While complete failure is always a risk with such endeavors, the opportunity to strike upon a novel observation is seductively present. In order to further our understanding of AR function, we have embarked on a path of discovery that could lead to the identification and isolation o f important AR modulatory factors. We have not thrown caution to the wind, but rather, have taken care to pursue a line of study that is well-supported by experimental evidence. Nevertheless, naive errors in experimental design or limitations in detection imposed by the yeast two-hybrid system may prove to be formidable challenges to our effort. 168 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. MATERIALS AND METHODS Plasmids Yeast expression vectors pCLl (encoding full-length GAL4), pVA3 (encoding murine p53 aa 72-390), pTD l (encoding SV40 T antigen aa 84-708), and pLAM5’ (encoding human lamin C aa 66-230) are commercially available (Clontech, Palo Alto, CA). Yeast expression vectors encoding GAL4 DBD fused with AR NTD (aa 5-445), AR NTD (aa 5-347), and AR NTD (aa 5-156) were constructed by inserting EcoPl-BamHL PCR fragments into the corresponding restriction sites of pGBT9 (Clontech). The fragments were amplified from genomic DNA using the following PCR primers: AR NTD (aa 5-445), F12 and R13; AR NTD (aa 5-347), F12 and R14; and AR NTD (aa 5-156), F12 and R15 (Table 3.1). Mammalian expression vector pcDNA-ARIP29 was constructed in sequential cloning steps. First, an Nhel-Kpril PCR fragment containing the AR Kozak sequence (primers F2 and R3) was inserted into the corresponding sites o f pcDNA3.1(+). Second, a Kpnl-EcoPA PCR fragment (primers F13 and R16) was inserted into the restored Kpnl site and the downstream EcoRI site of the pcDNA3.1(+) multiple cloning site. Primer F13 encodes the SV40 T antigen NLS (i.e., 5 ’ -CCCAAGAAGAAGCGTAAG-3 ’) and primer R16 encodes a polyhistidine (His)6 epitope tag [i.e., antisense 5’-(ATG)6 -3’], both in frame with the ARIP29 ORF. Mammalian expression vectors pcDNA-MAGE-11 (ORF-1), -M AGE-ll (ORF-2), and -MAGE-11 (ORF-3) were constructed exactly as 169 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. pcDNA-ARIP29 using the following primers to amplify Kpnl-EcoRI PCR fragments: MAGE-11 (ORF-1), F14 and R17; MAGE-11 (ORF-2), F15 and R17; and MAGE-11 (ORF-3), F16 and R18. Primers FI4, F I5, and F I6 encode the SV40 T antigen NLS and primer R17 encodes the (His)6 epitope in frame with the ARIP5 ORFs. Titer determination and amplification o f the pACT2 human testis cDNA library A 1.0 ml aliquot of E. coli BNN132 culture transformed with the human testis MATCHMAKER cDNA library (Clontech; CAT# HL4035AH; LOT# 6090518) was thawed on ice. The cDNA library was made from mRNA extracted from normal, whole testes pooled from 11 Caucasians (ages 10-61) and was cloned into the XhoUEcoKL restriction sites of vector pACT2. The library contained approximately 3.2 x 106 independent clones with an average cDNA insert size of 1.4 kb (i.e., according to the Clontech certificate of analysis). To estimate the titer of the library, 1 pi of the thawed culture was diluted 106 in LB and aliquots of the dilution (i.e., 50 and 100 pi) were plated on separate LB plates that contained ampicillin (50 pg/ml). Following growth at 30° C for 18 h, colonies were counted and the titer [colony forming units (cfu)/ml] calculated as follows: (no. colonies/plating volume) x 103 x 103 x 103 = cfu/ml. The titer was determined to be approximately 2 x 109 cfu/ml [i.e., (86 colonies/50 pi) x 109 ]. To amplify the pACT2 cDNA library, approximately 50,000 cfu were plated onto each of 150, 15cm LB/amp plates. Due to the instability of the diluted library, plating was done in batches of 20 (i.e., 2 pi of stock library were diluted in 6 ml LB with brief 170 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. vortexing and 300 pi were spread on each plate). Plates were allowed to dry for 15 min, and then, were incubated at 30° C. After 18 h, 5 ml LB was transferred to each plate, which was then scraped systematically from top to bottom with a Costar (Cambridge, MA) cell scraper. Using a sterile pipette, the LB-bacterial mixture was transferred to a chilled 1 L Erlenmeyer flask on ice. The pooled LB-bacterial mixture from all 150 plates (i.e., nearly 1 L) was diluted to 2 L with LB and incubated at 30° C for 2 h with shaking at 250 rpm. Plasmid DNA was isolated from the expanded library culture in several sequential batches using the Qiagen MEGA Plasmid Isolation Kit (Qiagen, Valencia, CA). Plasmid DNA pellets were dissolved in sterile TE (pH 8.0), pooled, and quantified by spectrophotometric analysis (i.e., 3.5 mg total yield). Yeast two-hybrid library screening The library-scale cotransformation procedure was carried out exactly as described in the Yeast Protocols Handbook (Clontech; CAT# PT3024-1). To prepare competent yeast cells for transformation, 5, fresh 2-3 mm CG-1945 colonies, grown for several days on YPD/agar, were used to inoculate 150 ml YPD medium. The culture was incubated for 12-18 h at 30° C with shaking at 250 rpm. This stationary phase culture was then used to inoculate 1 L YPD medium to an appropriate density (OD6 0 0 = 0.2-0.3). The log phase culture was incubated as above for about 3 h until the target density was reached (i.e., OD6 0 0 = 0.4-0.6). Yeast cells were collected by centrifugation at 1000 x g for 5 min and the resulting pellet was washed lx in sterile ddH,0. The washed pellet was re- 171 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. suspended in 8 ml of freshly prepared sterile lx TE/LiAc solution [10 mM Tris-HCl, 1 mM EDTA, 100 mM LiAc, (pH 7.5)]. For the library-scale cotransformation, 8 ml competent CG-1945 cells, 1000 pg pGBT9-AR NTD (5-156), 500 pg pACT2 human testis cDNA library, and 20 mg denatured herring sperm carrier DNA (Life Technologies, Rockville, MD) thoroughly were mixed in a sterile 50 ml tube by vortexing. Next, 60 ml PEG/LiAc solution [40 % polyethylene glycol 4000 (avg. MW = 3,350; Sigma, St. Louis, MO), 10 mM Tris-HCl, 1 mM EDTA, 100 mM LiAc] were added with thorough vortexing to mix. Following a 30 min incubation at 30° C with shaking at 200 rpm, 7 ml sterile DMSO were added to the cotransformation reaction and mixed by gentle inversion. The reaction mixture was then subjected to heat shock at 42° C for 15 min with occasional swirling to mix. Cells were chilled on ice for 2 min, pelleted by centrifugation (5 min/1000 x g ), washed lx with sterile ddH2 0 , and re-suspended in 5 ml TE [10 mM Tris-HCl, 1 mM EDTA, (pH 7.5)]. Aliquots of 330 pi were spread onto each of 32, 15 cm plates containing leucine (Leu)-, tryptophan (Trp)-, and histidine (His)-deficient minimal medium. To estimate the cotransformation efficiency of the screen, 100 pi of a 1:1000 dilution of the transformation reaction was spread onto a 10 cm LeuVTrp" plate. Plates were allowed to dry for 15 min and then incubated at 30° C for 5-10 days. Colonies exhibiting robust His+ prototrophy (i.e., 4-5 mm colonies) were picked and streaked onto fresh LeuVTrp* plates and grown for an additional 5 days at 30° C. Cotransformation efficiency (cfu/pg library plasmid DNA) was calculated using the following equation: (cfu on 1:1000 172 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. dilution plate) (total suspension vol. of screen) - * • (pi plated) (dilution factor) (pg pACT2 library plasmid). Yeast plasmid isolation and recovery in E. coli To isolate pACT2 library plasmids from yeast transformants, a 10 mm2 area of each re-streaked yeast clone was transferred to a 1.5 ml eppendorf tube using a sterile loop. The yeast sample was thoroughly re-suspended in 30 pi sterile suspension buffer [10 mM Tris-HCl, 1 mM EDTA, 4.5 units/ml lyticase (Sigma)]. Following a 30 min incubation at 37° C, 170 pi sterile lysis buffer [2 % Triton X-100, 1 % SDS, 100 mM NaCl, 10 mM Tris-HCl (pH 8.0), I mM EDTA] were added to the yeast suspension. The 200 pi solution was then transferred to another 1.5 ml tube that contained ~200 pi acid- washed glass beads (425-600 pm; Sigma). Next, 200 pi phenol:chloroform:isoamyl alcohol (25/24/1) (Sigma) were added and the tube was vortexed at maximum rpm for 5 min. Following centrifugation at 14, 000 rpm for 10 min in a tabletop microfuge, the aqueous (upper) phase was transferred to a fresh 1.5 ml tube to which 8 pi 10 M ammonium acetate were added. DNA was ethanol-precipitated according to standard methods and the pellet was re-suspended in 20 pi sterile ddH2 0. Electrocompetent HB101 E. coli (Invitrogen, Carlsbad, CA), which harbor a leu B6 auxotrophic mutation that can be complemented by the pACT2-encoded LEU 2 gene, were transformed with yeast plasmid DNA (i.e., 1.5/20 pi) by electroporation using a Bio-Rad (Hercules, CA) Gene Pulser. Following a 1 h recovery incubation (i.e., at 37° 173 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. C/250 rpm) in 1 ml SOC medium, transformed E. Coli were pelleted by centrifugation (1000 x g for 5 min) and then, re-suspended in 1 ml M9/Leu-deficient minimal medium (see Sambrook et al., 1989). Cells were pelleted once again and re-suspended in 500 pi M9/Leu-deficient minimal medium, of which 150 pi were plated on an M9/Leu-deficient plate that contained ampicillin (50 pg/ml), thiamine-HCL (1 mM), and pro line (10 pg/ml). The plate was allowed to dry for 10 min and then incubated at 37° C for 48 h. Plasmid DNA, purified by standard methods from 3 independent HB101 colonies per plate, was double-digested with XhoUEcoRl to release the cDNA insert. Clones with unique Xhol/EcoRl digestion patterns were sequenced with the Thermo Sequenase Kit (Amersham-Pharmacia Biotech, Piscataway, NJ) using flanking 5’ and 3’ primers. Yeast two-hybrid /3-galactosidase assays Small-scale yeast transformation, cellular extract preparation, and measurement of (3-gal activity by the luminescent detection of a fluorescent substrate were described in Chapter 3 (Materials and Methods). Qualitative detection of (3-gal activity by the colony- lift filter assay (Clontech) was done as follows. A sterile 75 mm VWR 410 grade filter was placed over the surface of a yeast plate. The filter was gently rubbed with forceps to facilitate the even transfer of colonies, was carefully lifted off of the yeast plate, and was repeatedly frozen in liquid N, for 10 sec and thawed at room temperature for I min (i.e., 3 consecutive cycles). The filter was then placed colony-side up on another filter which had been pre-soaked in Z-buffer (see Chapter 3, Materials and Methods) that contained 5- 174 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. bromo-4-chloro-3-indolyl-P-D-galactopyranoside (X-gal) at 330 ng/ml. The filters were incubated at 30° C for several hours to allow for blue color development. Yeast protein extract preparation Yeast protein extracts were prepared using the urea/SDS method of Printen and Sprague (1994). Briefly, a 5 ml stationary phase culture of a particular CG-1945 transformant, selected for 12-18 h in Leu- or Trp-deficient minimal medium, was used to inoculate 50 ml of YPD medium. The culture was grown for ~6-8 h at 30° C with shaking at 250 rpm until the appropriate density was reached (i.e., OD6 0 0 = 0.4-0.6). The 50 ml log phase culture was divided in half and immediately poured into two, sterile 50 ml tubes filled halfway with ice. Yeast were pelleted by immediate centrifugation (i.e., at 1000 x g for 5 min at 4° C) and then, re-suspended in 25 ml ice-cold sterile ddH2 0. Following a second centrifugation, the supernatant was discarded and the yeast pellet was immediately snap-frozen in liquid N2 before being stored at -70° C. The frozen yeast pellet was solubilized in pre-warmed (i.e., 60° C) complete cracking buffer [8 M urea, 5 % SDS, 40 mM Tris-HCl (pH 6.8), 0.1 mM EDTA, 1 % anhydrous P-mercaptoethanol, 0.4 mg/ml bromphenol blue] that contained yeast protease inhibitors [pepstatin A, leupeptin, benzamidine, aprotinin, and phenylmethyl-sulfonyl fluoride (PMSF)] and then transferred to a 1.5 ml eppendorf tube that contained acid- washed glass beads. Following a 10 min incubation at 70° C, the tube was vortexed at maximum rpm for 1 min and then centrifuged at 14,000 rpm for 5 min at 4° C in a 175 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. microfiige. The supernatant was transferred to a pre-chilled 1.5 ml tube on ice while the pellet was boiled for 5 min, vortexed, centrifuged, and the second supernatant was then combined with the first. The total yeast protein extract was boiled for 3 min and then immediately frozen at -70° C. Western analysis Equivalent amounts of yeast protein extracts were subjected to western blot analysis essentially as described for mammalian protein extracts in Chapter 3 (Materials and Methods). Transferred yeast proteins were probed with mouse monoclonal anti- GAL4 DBD antibody RK5C1 (Santa Cruz Biotechnology, Santa Cruz, CA) or mouse monoclonal anti-GAL4 AD antibody C-10 (Santa Cruz Biotechnology) at 0.1 pg/ml. Horseradish peroxidase-conjugated anti-mouse IgG (Santa Cruz Biotechnology) was used as the secondary antibody at 0.16 pg/ml (1:2500 dilution). Northern analysis A 3’ ARIP29 cDNA fragment was used to probe the Human Multiple Tissue Northern (MTN) Blot II (Clontech; CAT# 7759-1). The 630 nt fragment was released from pACT2-ARIP29 plasmid DNA by Ncol/Xhol double digestion and was purified using the Qiaquick Gel Extraction Kit (Qiagen). Approximately 50 ng of the ARIP29 fragment were [3 2 P]labeled using the Ready-To-Go DNA-labeling Bead Kit (Amersham- 176 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Pharmacia Biotech). The labeled probe was purified using a Probe-Quant G-50 column (Amersham-Pharmacia Biotech) and was quantified by scintillation counting. The MTN blot, which contained approximately 2 pg of poly A" RNA per lane from eight different human tissues (spleen, thymus, prostate, testis, ovary, small intestine, colon, and blood leukocyte), was placed into a 50 ml polypropylene Falcon tube (Becton Dickinson & Co., Lincoln Park, NJ) that contained 7 ml ExpressHyb hybridization buffer (Clontech) and was equilibrated to 68° C in a rotating Robbins Scientific (Sunnyvale, CA) hybridization oven. The pre-hybridization solution was replaced with 7 ml ExpressHyb buffer (equilibrated to 68° C) that contained the labeled ARIP29 probe at a concentration of 2 x 106 cpm/ml (or ~4 ng/ml). The blot was incubated for 1 h at 68° C with continuous shaking and then was washed extensively with Solution I (2x SSC, 0.05 % SDS) at room temperature followed by Solution II (O.lx SSC, 0.1 % SDS) at 50° C. The blot was covered in plastic wrap and exposed to Kodak X-OMAT film for 24 h. Following the ARIP29 analysis, the MTN blot was stripped and probed with a human p- actin cDNA probe (Clontech) to control for mRNA loading variation. Cell culture and transfections The maintenance and transfection of mammalian cell lines was described in Chapter 3 (Materials and Methods). 177 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Sequence database searching and ORF analysis To identify unknown nucleotide and translated amino acid sequences, National Center for Biotechnology Information (NCBI) non-redundant DNA and protein databases (available at http://www.ncbi.nlm.nih.gov) were searched using the BLAST (Altschul et al., 1990) and Gapped BLAST (Altschul et al., 1997) search programs. The DNA databases that were searched include: GenBank; EMBL; DDBJ; PDB sequences; and GenBank EST. The protein databases that were searched include: GenBank CDS translations; PDB; SwissProt; SPupdate; and PIR. Open reading frame analyses were done using the Gene Jockey II (version 2.0) program (Biosoft, Cambridge, UK). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 178 RESULTS Cloning and characterization o f GAL4-AR NTD yeast two-hybrid ‘ bait’ proteins To identify proteins that interact with the AR NTD, particularly with the poly-Q region, various exon 1 fragments of the AR gene were cloned in-frame with the GAL4 DBD ORF in vector pGBT9. Because the primary limitation of the traditional yeast two- hybrid system is that the bait protein must not efficiently activate reporter gene expression by itself, and because the AR NTD possesses potent transcriptional activation potential, three different GAL4 DBD-AR NTD fusion proteins, containing different structural/functional features of the AR NTD (Fig. 5.1), were expressed in yeast reporter strain CG-1945 (Fig. 5.2; Table 5.1). This strain contains two integrated reporter genes under the control of GAL4 response units: HIS3, a nutrition selection marker encoding an enzyme required for histidine synthesis; and the bacterial lacZ gene, a color selection marker encoding (3-galactosidase. Both the GAL4 DBD-AR NTD (5-445) and -AR NTD (5-347) fusion proteins strongly activated the GAL4-responsive reporter genes, resulting in efficient His" prototrophy (due to HIS3 expression) and substantial P~gal activity (Fig. 5.2A; Table 5.1). The GAL4 DBD-AR NTD (5-156) fusion protein, on the other hand, resulted in no growth on His-deficient medium and no P-gal activity above background levels (Fig. 5.2A; Table 5.1). Western analysis with an anti-GAL4 DBD antibody confirmed that GAL4 DBD-AR NTD (5-156) expression was stable in the CG-1945 strain despite its negligible impact on reporter gene activation (Fig. 5.2B). Based on 179 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. these preliminary experiments, GAL4 DBD-AR NTD (5-156), encoding the poiy-Q and some flanking sequences, was identified as the most suitable bait protein for a two-hybrid screen. Yeast two-hybrid screen o f a human testis cDNA library Yeast strain CG-1945 was cotransformed with pGBT9-AR NTD (5-156) and a pACT2 human testis cDNA library and then, selected on plates containing Leu-, Trp-, and His-deficient minimal medium. Transformants that were His' prototrophic and exhibited robust growth (i.e., a colony size of 4-5 mm) on Leu'/Trp'/His' medium were picked, re-streaked on fresh LeuVTrpYHis' plates, and then, assayed for (3-gal activity. A representative (3-gal screen of several His' prototrophic clones is shown in Figure 5.3. Note that in this particular experiment, only 1/3 of the clones positive for HIS3 expression, were also positive for lacZ expression. In total, 4 independent screens of the library were conducted and an estimated 3.3 x 106 pACT2 cDNA clones were assessed for interactions with the GAL4-AR NTD (5-156) bait protein (see Table 5.2 for a summary of the yeast two-hybrid results). Of the His' prototrophic colonies, 159 were further assayed for P-gal activity and of these, only 32 were positive. These 32 ‘double positive’ clones contained pACT2 library plasmids encoding putative androgen receptor interacting proteins or ARIPs. The pACT2 library plasmids encoding putative ARIPs were isolated from yeast transformants and then re-introduced into the CG-1945 strain along with pGBT9-AR 180 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. NTD (5-156) to confirm two-hybrid interactions. Surprisingly, in only 11 out of 32 cases were the interactions verified (Table 5.2). To identify false positives [i.e., pACT2 plasmids encoding proteins that activate reporter gene expression by themselves (self activators) or that interact with GAL4 DBD instead of the bait protein] among the ‘confirmed’ clones, CG-1945 was transformed with each pACT2-ARIP plasmid alone and in combination with pGBT9 vector, which expresses GAL4 DBD. ARIPs 3 and 18 were self-activators whereas ARIPs 20 and 22 interacted with GAL4 DBD (Table 5.2). Thus, 7 candidate ARIPs (i.e., 5, 8, 25, 28, 29, 31, and 32) were identified in the screen. The pACT2 cDNA inserts encoding these proteins were partially sequenced and the obtained sequence information was used to search the NCBI-maintained non-redundant data bases (see Materials and Methods). Five of the 7 ARIPs (i.e., 5, 25, 28, 31, and 32) corresponded to the same gene located at Xq28 (see below). ARIP29 was identical to human ASB-3 variant except at its extreme N-terminus (see below). Finally, the cDNA encoding ARIP8 was identical to human CGR19 (i.e., cell growth regulatory gene 19; Madden et al., 1996) though it was out o f frame. Preliminary characterization o f ARIPs encoded at Xq28 ARIPs 5, 25, 28, 31, and 32 reproducibly interacted with AR NTD (5-156) in yeast two-hybrid assays (Fig. 5.4). These proteins were found to be encoded by a common gene located on the X chromosome at Xq28. Figure 5.5 depicts the presumed exonic structure of this Xq28 gene as inferred from sequence identity with an Xq28 181 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. cosmid (exons I and II) and with the MAGE-11 locus (exons III-IV) (De Plaen et al., 1994). The cDNAs encoding ARIPs 5 and 28 were identical and the ARIP5/28 reading frame was arbitrarily defined as ORF-1 (Fig. 5.5). The cDNAs encoding ARIPs 31 and 32 were truncated relative to ARIP5/28, but encoded the same ORF-I protein. Interestingly, the ARIP25 cDNA, while derived from the same Xq28 message, encoded a protein in ORF-2 (Fig. 5.5). All of the Xq28 cDNA clones contained the two known MAGE-11 gene exons (i.e., exons IV and V; Fig. 5.5) identified by De Plaen et al. (1994). Thus, it appears that they are derived from novel transcripts of the MAGE-11 gene that include sequences from 3 previously unknown 57 exons. Exons I and II correspond to Xq28 cosmid U21A12 (GenBank accession # U69569) nucleotides <38331-38406 and 39387-39488, respectively. Exon III corresponds to nucleotides 683-778 of the MAGE-11 gene (GenBank accession # U10686). All of the intervening introns (i.e., corresponding to the novel 5’ exons) are bordered by splice donor (i.e., GT) and splice acceptor (i.e., AG) signals (data not shown). In addition, ORF analysis revealed an in-frame ATG 110 codons upstream of the MAGE-11 start codon defined previously by De Plaen et al. (1994). This newly identified ATG is in exon II (i.e.. Fig. 5.5), is preceded by an in frame stop codon, and resides in a reasonably conserved Kozak sequence (i.e., T GAGGGATGG). Figure 5.6 depicts the putative amino acid sequences of the MAGE-11 protein and the Xq28 encoded ARIPs. Translation of MAGE-11 from the newly identified, upstream 182 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ATG predicts a protein with a calculated molecular mass o f 48, 000 M ,. compared to one o f 35, 540 M ,. beginning at MetII0. Interestingly, this ‘new’ MAGE-11 protein encodes an authentic NLS near its N-terminus. ARIPs 31 and 32 are truncated variants o f the ARIP5/28 protein, which comprises 104 amino acids and has a predicted molecular mass o f approximately 11, 440 M,. (Fig. 5.6). The putative ARIP25 protein is a 16 amino acid polypeptide with a predicted molecular mass of 1,760 To confirm that the pACT2- ARJP plasmids were indeed directing the stable expression of proteins of the appropriate predicted sizes, western analysis was done using an anti-GAL4 AD antibody (Fig. 5.7). Faint, but discernible, bands migrating at or near the 34 kDa protein marker were present in extracts from pACT2-ARIP5, 28, 31 and 32 yeast transformants. These immunopositive bands were detected at higher molecular weight than expected since the GAL4 AD-ARIP5 fusion protein has a predicted molecular mass of about 24, 000 M ,.. No immunopositive band was visible in extracts from pACT2-ARIP25 transformants (Fig. 5.7). The isolation of 4 unique clones of the same Xq28/MAGE-11 cDNA from the yeast-two hybrid library screen suggests a specific interaction between AR NTD (5-156) and the Xq28/M4G£ -/1 encoded proteins. While this is an attractive explanation o f the screening results, it is unlikely to be true since pACT2-ARIP25 encodes a different protein than the other Xq28IMAGE-11 clones. Therefore, the possibility exists that the observed interactions with AR NTD (5-156) are mediated by mRNA and not protein encoded by the Xq28IMAGE-11 clones. To directly test this hypothesis, MAGE-11 183 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. cDNA sequences were cloned into the mammalian expression vector pcDNA3.1(+) in all 3 reading frames. None o f the MAGE-11 constructs appreciably affected ligand- dependent AR signaling when transfected into PC-3 cells (Fig. 5.8A). The MAGE-11 (ORF-2) construct, however, modestly did enhance GRIP1 mediated coactivation o f the AR in a ligand dependent fashion (Fig. 5.8B). Similar results were obtained using the MAGE-11 (ORF-1) and (ORF-3) constructs (data not shown). Preliminary characterization o f ARIP29 ARIP29 interacted weakly but reproducibly with AR NTD (5-156) as well as with AR NTD (5-347) in yeast two-hybrid assays (Fig. 5.9). The ARIP29 cDNA contains a partial ORF (i.e., 1425 nt) with a defined 3’ end (data not shown). The ORF encodes a protein of 475 aa with a calculated molecular mass of 52, 000 Mr (Fig. 5.10A). A search of the non-redundant protein data bases (see Materials and Methods) revealed that the C- terminal 454 aa of the putative ARIP29 protein were identical to human ankyrin repeat- containing suppresser of cytokine signaling (SOCS) box protein 3 (i.e., hASB-3) (Hilton et al., 1998). ARIP29/hASB-3 contains 7 conserved ankyrin motifs N-terminal to the highly conserved SOCS box (Fig. 5.10A and B). It also contains 3 leucine-rich NR boxes. Stable ARIP29/hASB-3 protein expression was observed in yeast, as detected by an anti-GAL4 AD antibody (Fig. 5.7). To reveal the size(s) of the endogenous ARIP29/hASB-3 mRNA and to assess its expression in various human tissues, northern blot analysis was done using a 3 ’ ARIP29 184 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. cDNA probe (Fig. 5.11). A single radiographic band migrating at or near the 2.4 kb RNA marker was identified in most tissues analyzed, though its expression was relatively higher in the testis, ovary, and prostate (Fig. 5.11). Thus, ARIP29/hASB-3 mRNA expression occurs in tissues that are regulated, at least partially, by AR signaling. To assess the impact o f ARIP29 on ligand-dependent AR signaling, the AREP29 cDNA sequence was cloned into the mammalian expression vector pcDNA3.1(+) and transiently transfected into PC-3 cells. ARIP29 modestly enhanced the ligand-dependent transactivation activity o f AR(Q)9 but had negligible effects on AR(Q)4 2 activity (Fig. 5.12A). To determine whether ARIP29 stabilized AR protein expression in the presence or absence of ligand, PC-3 cells were cotransfected with AR and ARIP29 expression vectors. Western analysis on extracts from transfected cells revealed that ARIP29 expression increased DHT-dependent stabilization of AR protein levels (Fig. 5.12B). 185 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 5.1 LEGEND Schematic diagram indicating the amino acid boundaries of three AR NTD fragments used as bait proteins in the yeast two-hybrid system. Despite sharing a common N- terminus (i.e., Leiij), each fragment contains various structural/functional features of the AR NTD. The smallest AR NTD fragment comprises aa 5-156 and includes the polyglutamine stretch (Q) and the N-terminal region of TAU-1. The middle-sized fragment comprises aa 5-347 and encompasses most of the TAU-1 core domain. The largest fragment spans aa 5-445 and includes, in addition to Q and TAU-1, the polyproline repeat (P) and the N-terminal half of TAU-5. These AR NTD fragments were expressed as fusions with the GAL4 DBD using the pGBT9 yeast expression vector. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 186 FIGURE 5.1 1 MEVQjLGDGRV jT PR PPSK T Y R < 2Q Q < 2Q Q < 2Q Q Q G A F Q N L F Q SV R E V IQ N P G P R H PE A A SA A P P 5 1 G A S I iI iZ iI iQQQ QQQQQQQETS PRQQQQQQGE DGSPQAHRRG 1 0 1 P T G Y L V L D E E Q Q PSQPQSAXi EC H PE R G C V P E PG A A V A A SK G L PQ Q LPA PP 1 5 1 ^ E D D S ^ A P S T L S L L G P T F P G D S S C S A D L K D IL S E A S T M Q L LQQQQQEAVS 2 0 1 E G S S S G R A R E R S G A P T S S K D N Y IaG G T STX S DNAK ELCK A V SVSM GLGVEA 2 5 1 I 1E H I 1 S P G E Q I 1 RG D C K Y APLL G V P P A V R P T P CAPZiAECKGS LL D D SA G K ST 3 0 1 E D T A E Y S P F K GGYTKGDEGE S D G C S G S A A A G S S G T L E L P S T^SLYKjSGAL DYGSAWAAAA 3 5 1 D E A A A Y Q S R D Y Y N F P L A L A G P P P P P P P P H P H A R 1K L E N P L 4 0 1 A Q C R Y G D L A S LHGAGAAGPG S G S P S A A A S S SW HTX.FTAEE GQLYGjPCGGG ESDFTAPDVW 4 5 1 GGGGGGGGGG GGGGGGGGGG EAG AVAPYG Y TR PPQ G DAG Q 5 0 1 Y PG G M V SR V P Y P S P T C V K S E MGPWMDSYSG PYGDMRLE 5 3 8 NLS AF-1a AF-2 AF-1b | DBD NTD LBD 1 101 919 J2ZQ TAU-1 3 6 0 _ _ _ _ _ _ _ _ 5 2 8 TAU-5 AR N T D ( 5 - 1 5 6 ) 1 - AR N TD ( 5 - 3 4 7 ) 2 ■ AR N T D ( 5 - 4 4 5 ) 3 187 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 5.2 LEGEND Background activities and expression levels of AR NTD-GAL4 DBD proteins. (A) S. cerevisiae yeast reporter strain CG-1945 was transformed with pGBT9 vector encoding AR NTD (aa 5-156)-, AR NTD (aa 5-347)-, or AR NTD (aa 5-445)-GAL4 DBD as indicated. Yeast extracts were prepared from transformants selected in Trp-deficient medium and were assayed for (3-gal activity by chemiluminescent detection of a fluorescent substrate. Data presented are the mean + SE from three different yeast transformants and are representative o f at least two independent experiments. (B) Determination of AR NTD-GAL4 DBD expression by Western analysis. Yeast protein extracts were prepared from CG-1945 transformants selected in Trp-deficient medium. Equivalent amounts of extract were probed with anti-GAL4 DBD antibody and analyzed by autoradiography. MWM, molecular weight protein marker; C (control), protein extract prepared from an untransformed CG-1945 colony; GAL4 DBD, protein extract prepared from a CG-1945 colony transformed with empty pGBT9 vector. 188 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 5.2 A AR NTD 5 156 AR NTD 347 AR NTD 445 - » ■ ro co -k io o o o o o o o o o o © o relative [3-galactosidase activity B Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3000 FIGURE 5.3 LEGEND Representative screen to confirm interactions between AR NTD (5-156) and putative interacting proteins. S. cerevisiae yeast reporter strain CG-1945 was cotransformed with pACT2 vector encoding unknown proteins fused to GAL4 AD and/or pGBT9 vector encoding AR NTD (5-156). Yeast extracts were prepared from His+ prototrophic transformants selected in Trp- and/or Leu-deficient medium and were assayed for (3-gal activity by chemiluminescent detection of a fluorescent substrate. Data presented are the mean + - SE of three different yeast transformants and are representative of two independent experiments. Control plasmids: pCLl, YCp50-derived vector encoding full-length GAL4; pVA3, pAS2-l vector encoding murine p53 (aa 72-390); pTDl, pACT2 vector encoding SV40 large T-antigen (aa 84-708); pLAM5’, pAS2-l vector encoding human lamin C (aa 66-230). 190 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 5.3 PGBT9-AR (5-156) pACT2-testis cDNA library pCL1 pVA3/pTD1 pl_AM5'/pTD1 pGBT9-AR (5-156) unknown-1 unknown-2 unknown-3 unknown-4 unknown-5 unknown-6 unknown-7 unknown-8 unknown-9 unknown-10 unknown-11 unknown-12 1 1 - ■ r o r o r o GO o o U 1 o © o © o © © o relative p-galactosidase activity 191 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 5.4 LEGEND Interactions of AR NTD (5-156) with proteins encoded at the Xq28/M4G£-// locus. S. cerevisiae yeast reporter strain CG-1945 was cotransformed with pACT2 vector encoding ARIP 5, 25, 28, 31, or 32 fused to GAL4 AD and/or pGBT9 vector encoding AR NTD (5-156)-GAL4 DBD. Yeast extracts were prepared from transformants selected in Trp- and/or Leu-deficient medium and were assayed for p-gal activity by chemiluminescent detection of a fluorescent substrate. Data presented are the mean + SE of three different yeast transformants. See legend to Fig. 5.3 for description of controls. *Only a single pVA3/pTDl transformant was analyzed in this experiment (i.e., n = 1). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 192 FIGURE 5.4 IA *r o > r — • o o m + ■ + ■ + ■ + ■ + ■ + • + ■ + + + + + ■ “l“ “1 o o o o 0 o o o 01 to SIA ■ + I I + I + I A);Ai)oe asep;so)3e|e6-£| dAiie|ai 193 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ARIP32 FIGURE 5.5 LEGEND Presumed exonic structure of the Xq28/MAGE-11 clones. Library pACT2 plasmids encoding ARIPs 5, 25, 28, 31, and 32 were isolated and sequenced using both 5’ and 3’ primers. Sequence data revealed that the cDNAs encoding ARIPs 5 and 28 were identical and that the cDNAs encoding ARIPs 25, 31, and 32 were independent truncated clones of the ARIP5/28 cDNA. Each cDNA clone corresponded to five exons (i.e., I-V) embedded in the Xq28 cosmid U21A12 (GenBank accession # U69569) or the MAGE-11 locus GenBank accession # U10686). ARIPs 5/28, 31, and 32 are encoded in the arbitrarily defined ORF-1, while ARIP25 is encoded in ORF-2. MAGE-11 is encoded in ORF-2’ since it begins two codons downstream of the ARIP25/ORF-2 stop. The newly identified putative MAGE-11 start codon and the originally published MAGE-1 1 start codon are in bold font. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 5.5 [-► 5,28 [-► 31 [-► 25 lA G O STC TTTC T G A G G G G T T 3 T C T T G A G A G T G G C A G A G G G C A G C G G G T C C A G G C T C C A T G A G G A G G C A A G •-► 32 ORF-2 sto p J V tPutative MAGE-11 s ta rt codon (Irvine and Coetzee, 1999)] C C T T G G G A A T C tT G JjG G G A T G G A G A C T C A G T T C C G C A G A G G G G G T C T G G G G T G C A G C C C T G C C A G C A T C A A G A G G A A G A A G A A G A G G G A G G A C T C A G G A G A C TTTG G A CTC C A G G T G A G C A T T A TG TTC TC A G A G G A C G A C T T C C A G C C A A C A G A A A G A G C C C C A T A T G G T C C A C A A C T A C A G T G G T C C C A G G A T C T G C C A A G A G T C C A G ORF-1 s top G TTTTfrAGkG A A C A G G C C A A C C T G G A G G A C A G G A G T C C C A G G A G A A C C C A G A G G A T C A C T G G A G G A G A A C A A G T [MAGE-11 sta rt codon (De Plaen e t al., 1994)] G C T G T G G G G C C C C A T C A C C C A G A T A T T T C C C A C A G T T C G G C C T G C T G A C C T A A C C A G A G T C A TC A TG C C T C T T G A G C A A A G A A G T C A G C A C T G C A A G C C T G A G G A A G G C C T T C A G G C C C A A G A A G A A G A C C T G G G C C T G G T G G G T G C A C A G G C TC T C C A A G C T G A G G A G C A G G A G G C T G C CTTCTTCTCC T C T A C T C T G A A T G T G G G C A C T C T A G A G G A G TTG C C TG C TG C T G A G T C A C C A A G T C C TC C C C A G A G T C C T C A G G A A G A G T C CTTCTCTCCC A C T G C C A T G G A TG CCA TCTT T G G G A G C C T A T C T G A T G A G G G C T C T G G C A G C C A A G A A A A G G A G G G G C C A A G T A C C T C G C C T G A C C T G A T A G A C C C T G A G T CCTTTTCCCA A G A T A T A C T A C A T G A C A A G A TA A TTG A TTT G G TTC A TTTA TTG C TC C G C A A G T A T C G A G T C A A G G G G C T G A T C A C A A A G G C A G A A A T G C T G G G G A G T G T C A T C A A A A A T T A T G A G G A C T A C T T T C C T G A G A TA T T T A G G G A A G C C T C T G T A T G C A T G C A A CTG CTCTTTG G C A T T G A T G T G A A G G A A G T G G A C C C C A C T A G C C A C TC C TA TG TC C TTG TC A C C T C C C T C A A CCTCTCTTA T G A T G G C A T A C A G T G T A A T G A G C A G A G C A T G C C C A A G T C T G G C C T C C T G A T A A T A G T C C T G G G T G T A A T C T T C A T G G A G G G G A A C T G C A T C C C T G A A G A G G T T A T G T G G G A A G T C C T G A G C A T T A T G G G G G T G T A T G C T G G A A G G G A G C A CTTCCTCTTT G G G G A G C C C A A G A G G C T C C T T A C C C A A A A T T G G G T G C A G G A A A A G T A C C T G G T G T A C C G G C A G G T G C C C G G C A C T G A T C C TG C A TG C TA T G A G T TC C T G T G G G G T C C A A G G G C C C A C G C T G A G A C C A G C A A G A T G A A A G T T C T T G A G T A C A T A G C C A A T G C C A A T G G G A G G G A T C C C A C T TC TTA CCCA T ORF-21 stop C C C T G T A T G A A G A T G C T TT G A G A G A G G A G G G A G A G G G A G T C fT G flG C A T G A G A T G C A A C C A G G G C C A G C G G G C A G G G A A A T G G G C C A A T G C A T G C T T C A G G G C C A C A C C C A G C A G T T T C C C T G T C C T G T G T G A A A T C A G G C CCA TTC TTC C CTCTG TG TTT G A T G A G A G A A G TC A G TG TTC T C A G T A G T A G A A G G C A C A G T G A A T G G A A G G G A A C A C A T T G TA TA C TG C C T TTA G G TTTC T C TTC C A TC G G G T G A C T T G G A G A TTTG TTTT TGTTTCCCTT TG G TA A TTTT C A A A T A T TG T T C C T G T A A T A A A A G T T T TA G TTA G C TTC A A C A T C T A A G T G T A T G G A T G A T A C T G A C C A C A CA TG TTG TTT T G C T TA TC C A T TTC A A G TG C A A G T G T T T G C C A T T T T G T A A A A C A T TT T G G G A A A T C T T C C A TCTTG C TG T G A T T T G C A A T AG G TA TTTTC T T G G A G A A T G T A A G A A C T T A A C A A T A A A G C T G A A C T G G T G T T G T G A A A C A G A G A A A Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 5.6 LEGEND Putative amino acid sequences for MAGE-11 and Xq28-encoded ARIPs. Based on sequencing data and ORF analyses, predicted amino acid sequences were determined for MAGE-11 and the Xq28-encoded ARIPs. The newly identified MAGE-11 initiation codon (i.e., M,) is in bold font, the putative MAGE-11 NLS is underlined, and the original MAGE-11 initiation codon (i.e., now M 1 U ) is boxed. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 196 FIGURE 5.6 Putative MAGE-11 protein METQFRRGGL GCSPASIKRK KKREDSGDFG LQVSIMFSED DFQPTERAPY GPQLQWSQDL PRVQVFREQA NLEDRS PRRT QRITGGEQVL WGPITQIFPT VRPADLTRVI 0PLEQP.SQHC KPEEGLQAQE EDLGLVGAQA LQAEEQEAAF FSSTLNVGTL EELPAAESPS PPQSPQEESF SPTAMDAIFG SLSDEGSGSQ EKEGPSTSPD LIDPESFSQD ILHDKIIDLV HLLLRKYRVK GLITKAEMLG SVIKNYEDYF PEIFREASVC MQLLFGIDVK EVDPTSHSYV LVTSLNLSYD GIQCNEQSMP KSGLLIIVLG VIFMEGNCIP EEVMWEVLSI MGVYAGREHF LFGEPKRLLT QNWVQEKYLV YRQVPGTDPA CYEFLWGPRA HAETSKMKVL EYIANANGRD PTSYPSLYED ALREEGEGV Putative ARIP5/28 (ORF-1) SVFLRGVLRV AEGSGSRLHE EASLGNLRDG DSVPQRGSGV QPCQHQEEEE EGGLRRLWTP GEHYVLRGRL PANRKSPIWS TTTWPGSAK SPGF Putative ARIP32 (ORF-1) VFLRGVLRV AEGSGSRLHE EASLGNLRDG DSVPQRGSGV QPCQHQEEEE EGGLRRLWTP GEHYVLRGRL PANRKSPXWS TTTWPGSAK SPG Putative ARIP31 (ORF-1) VLRV AEGSGSRLHE EASLGNLRDG DSVPQRGSGV QPCQHQEEEE EGGLRRLWTP GEHYVLRGRL PANRKSPIWS TTTWPGSAK SPGF Putative ARIP25 (ORF-2) QRAAGPGSMR RQALGI 197 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 5.7 Expression of ARIP-GAL4 AD proteins determined by Western analysis. Yeast protein extracts were prepared from CG-1945 transformants selected in Leu-deficient medium. Equivalent amounts of extract were probed with anti-GAL4 AD antibody and analyzed by autoradiography. MWM, molecular weight protein marker; C (control), protein extract prepared from an untransformed CG-1945 colony. An immunopositive band corresponding to ARIP5 is indicated by the arrow. 198 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 5.8 LEGEND Assessment o f MAGE-11 mRNA function on AR transactivation in mammalian cells. (A) PC-3 cells were cotransfected with 2.0 jag ARR3 TK-CAT, 50 ng pCMV-hAR. and 2.0 pg pcDNA-M4GE-11 (ORF-1), -MAGE-11 (ORF-2), or -MAGE-11 (ORF-3) as indicated. The total amount of DNA per transfection was held constant by the addition of pcDNA3.1(+) vector when appropriate. Twenty-four hours after transfection, cells received fresh medium that contained 5% charcoal/dextran-stripped FBS and 10 nM DHT or ethanolic vehicle. Twenty four hours later, whole-cell extracts were prepared and assayed for CAT activity as described in Chapter 3 (Materials and Methods). Data presented are the mean + SE of three independent dishes. (B) PC-3 cells were cotransfected with 2.0 pg ARR3 TK-CAT, 50 ng pCMV-hAR, 2.0 pg pSG5-GRIPl, and 2.0 pg pcUNA-MAGE-11 (ORF-2) as indicated. Transfection conditions and data presentation as in A . 199 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 5.8 > Q > « 25- a > 1 I no hormone 10 nM DHT a r r 3tk-cat ARwt MAGE-11 (ORF-1) MAGE-11 (ORF-2) MAGE-11 (ORF-3) + + + + + + + + + + + + + + + + + + B 200 □ no hormone 10 nM DHT fi* 150 a r r 3t k-cat ARwt GRIP1 MAGE-11 (ORF-2) + + + + + + + + + + + + + + + + + + + + + + 200 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 5.9 LEGEND Interactions between ARIP29 and the AR NTD. S. cerevisiae yeast reporter strain CG- 1945 was cotransformed with pACT2 vector encoding ARTP29 fused to GAL4 AD and/or pGBT9 vector encoding AR NTD (5-156)- or AR NTD (5-347)-GAL4 DBD. Yeast extracts were prepared from transformants selected in Leu- and/or Trp-deficient medium and were assayed for P-gal activity by chemiluminescent detection of a fluorescent substrate. Data presented are the mean + SE o f three different yeast transformants. 201 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 5.9 ARIP29 ARIP29 ARIP29 relative p-galactosidase activity Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2000 FIGURE 5.10 LEGEND Putative ARIP29 amino acid sequence. (A) Based on cDNA sequence data and ORF analysis, the predicted amino acid sequence of ARIP29 was determined. This partial ARIP29 protein sequence comprises 475 amino acids and contains 7 conserved ankyrin- like motifs (underlined), 3 leucine-rich NR boxes (boxed), and a conserved C-terminal suppressor of cytokine signaling (SOCS) box (shaded). (B) Alignment of the 7 ARIP29 ankyrin-like motifs with a consensus sequence. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 203 FIGURE 5.10 Putative ARIP29 protein sequence 1 LGRSSGDAAS DQPFRCRSGS TADSSENYIK MKTFEGFCAL HLAASQGHWK 51 IVQILL.EAGA DPNATTLEET TPLFLAVENG QIDVLRLLLQ HGANVNGSHS LI 101 MCGWNSLHQA SFQDNAEIIK DLLRKGANKE CQDDFGITPL FVAAQYGKLE 151 SI.SXZ.XSSGA NVNCQALDKA TPLFIAAQEG HTKCVELLLS SGADPDLYCN _ _________________ 1 1 5 _ _ _ _ _ _ _ _ _ _ _ I 201 EDSVTQLPIHA AAQMGHTKIL DLLXPLTNRA CDTGLNKVSP VYSAVFGGHE 1 ® ------ .__________________________ I 251 ZK3.EXI.UtNG YSPDAQACLV FGFSSPVCMA FQKDCEFFGI VNILLKYGAQ IZ ________________________________________ 301 INELHLAYCL KYEKFSIFRY FLRKGCSLGP WNHIYEFVNH AIKAQAKYKE 351 WLPHLLVAGF DPLXLLCNSW IDSVSIDTLI FTLEFTNWKT LAPAVERMLS 401 ARASNAWILQ 451 B -G-TPLH-AA— GH---V— LL— GA— N----- motif ^ j consensus 1 -GFCALHLAASQGHWKIVQILLEAGADPN 2 - ETTPLFLAVENGQIDVLRLLLQHGAJNVN 3 -GWNSLHQASFQDNAEIIKLLLRKGAN 4 -GITPLFVAAQYGKLESLSILISSGANVN 5 -KATPLF IAAQEGHTKCVELLLSSGADPDLYCN 6 - W Q L.P IHAAAWMGHTKILDLLIPLTNRACDTGLN 7 -FSSPVCMAFQKEFFGIVNILLKYGAQIN DC 204 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 5.11 LEGEND ARIP29 mRNA expression in human tissues. A commercially obtained northern blot containing approximately 2 pg of poly A" RNA per lane from eight different human tissues was probed with a 630 nucleotide 3’ ARIP29 cDNA probe. As a control for RNA loading, the blot was stripped and re-probed with a human (3-actin cDNA probe. Relative positions of RNA molecular weight standards are indicated. 205 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 5.11 206 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 5.12 LEGEND Assessment o f ARIP29 function on AR transactivation and expression in mammalian cells. (A) PC-3 cells were cotransfected with 2.0 pg MMTV-CAT, 2.0 pg pcDNA- ARIP29, and 500 ng of pcDNA-hAR(Q)9 or -hAR(Q)4 2 as indicated. Total transfected DNA per dish was held constant by the addition of pcDNA3.1(+) vector when appropriate. Transfection conditions and data presentation as in Fig. 5.8A. (B) ARIP29- dependent stabilization of AR protein expression. PC-3 cells were transfected with 3.0 pg pCMV-hAR and/or 3.0 pg pcDNA-ARIP29. Twenty-four hours after transfection, cells received fresh medium that contained 5% charcoal/dextran-stripped FBS and 10 nM DHT or ethanolic vehicle. Twenty-four hours later, whole-cell extracts were prepared and normalized for total protein. Equivalent amounts of extract were probed with anti- AR antibody. Radiographic bands were analyzed by scanning densitometry. Relative AR stabilization due to DHT alone or to DHT and ARIP29 expression is indicated. MWM, molecular weight marker; C (control), extract prepared from PC-3 cells transfected with pcDNA3.1(+). 207 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. FIGURE 5.12 120 - C = J no hormone 10 nM DHT I* 90 H ‘ •S § < 6 0 ^ I to P 30 H n l r ^ i n j f n . MMTV-CAT + + + + + + + + AR(Q)g + + + + - AR(Q>42 - + + + + ARIP29 - + + - - + + B kDa 116 — 97 *•* 66 ✓ relative AR stabilization 1.00 1.47 I 1.00 2.44 + + + + + anti-hAR ARwt ~| ARIP29 DHT J Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. TABLE 5.1 Summary of preliminary yeast two-hybrid studies CG-1945 transformant growth on deficient media* P-galactosidase activity5 V L'/H' W W'/H' L /W L'/W'/H' filter lift assay chemiluminescence pCLl + + - - - - +++ +++ pVA3/pTDl + + + + + + ++ ++ pLAM51/pTDl + - + - + - - - pGBT9-AR (5-156) - - + - - - - - pGBT9-AR (5-347) - - +/- - - ++ ++ pGBT9-AR (5-445) - - + + - - +++ +++ * growth on plates deficient in: L, leucine; W, tryptophan; H, histidine. 5 relative p-galactosidase activities measured by the two different techniques: none; +, low; ++, substantial; +++, high. to o TABLE 5.2 Results o f yeast two-hybrid screen o f a human testis cDNA library with the AR NTD (5-156) bait protein No. independent library clones screened: -3 x 1 06. No. of histidine prototrophs picked and screened for P-galactosidase activity: 159. No. of histidine prototrophs positive for p-galactosidase activity: 32. false positives ARIP# confirmed' self-activator GAL4 interaction sequence identity 1 2 § 3 + + 4 - 5 + - Xq28 cosmid U21A12 6 § 7 - 8 + - human CGR-19 9 - 10 - 11 - 12 - 13 - 14 - 15 - 16 - 17 - 18 + + 19 ¥ 20 + + 21 - 22 + + 23 - 24 - 25 + - Xq28 cosmid U21A12 26 - 27 - 28 + - Xq28 cosmid U 21A 12 29 4- - human ASB-3 variant 30 - 31 + - Xq28 cosmid U21A12 32 + - Xq28 cosmid U21A12 into the yeast reporter strain CG-1945. 5 yeast cotransformed with the pACT2 plasmid and pGBT9-AR (5-156) failed to grow on His' plates. ¥ no pACT2 library plasmid was recovered from clone. 210 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DISCUSSION One o f the primary goals of this thesis research was to understand how poly-Q variation in the AR NTD modulates receptor transactivation activity. In Chapter 3, AR transactivation activity was shown to decrease with increasing poly-Q length, an effect due, at least in part, to inhibition of pl60-mediated coactivation. Presumably, increased poly-Q length negatively affects interactions at the downstream pl60-AR NTD interface. In addition to the p i 60 coactivators, other, as yet unidentified, proteins may interact with the AR NTD and differentially modulate receptor activity depending on poly-Q length. To identify such proteins, a human testis cDNA library was screened with an AR NTD- derived bait protein using the yeast two-hybrid system (Fields and Song, 1989; Chien et al., 1991). The bait protein used in the yeast two-hybrid screen included AR amino acids 5- 156. This rather small AR NTD fragment, comprising the poly-Q and some flanking sequences, was used by default since larger NTD fragments demonstrated significant background transactivation activity when expressed in the yeast reporter strain. Because of this, only those proteins that interact with the AR ‘poly-Q region’ potentially would have been identified in the screen. Other possibly relevant proteins, which interact elsewhere on the AR NTD, would have been missed due to this design limitation. 211 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Identification and confirmation ofputative AR interacting proteins Overall, 32 yeast cotransformants were identified, from a library screen of some 3.3 x 106 pACT2 cDNA clones, that were positive for both HIS3 and LacZ reporter gene expression. Surprisingly, only 11 of these 32 were confirmed following isolation o f the pACT2 plasmids (i.e., encoding the putative ARIPs) and re-introduction into the yeast reporter strain. Because S. cerevisiae strains can maintain several different selectable plasmids, one possible explanation for this result is that irrelevant pACT2 library plasmids were isolated from some of the ‘double positive’ yeast clones. This possibility seems unlikely, however, since a minimum of 3 independent plasmid preparations were evaluated for each yeast clone. In all but one case, the same pACT2 plasmid (i.e., as determined from restriction endonuclease digest patterns and direct sequencing) was repeatedly recovered from an individual yeast colony. In light of this, it is probable that spontaneous mutations occurred within some of the yeast clones that allowed for reporter gene activation and selection despite no relevant AR NTD-ARIP interactions. While some of the ‘confirmed’ ARIPs proved to be false positives, others specifically and reproducibly interacted with the AR NTD bait. In all, 7 putative ARIPs were identified in the yeast two-hybrid screen, 5 o f which corresponded to melanoma antigen gene (MAGE)-II encoded at Xq28 (Rogner et al., 1995). The isolation of multiple independent clones of the same cDNA, while far from convincing in and of itself, supports the presence of a biologically relevant interaction. The remaining 2 ARIPs (i.e., 8 and 29) were isolated only once during the screen, though they may be 212 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. encoded on rare transcripts, and therefore, not well-represented in the testis cDNA library. ARIPs encoded at the Xq28/MAGE-11 locus Sequencing revealed that 5 of the 7 putative ARIPs were encoded by cDNAs derived from novel transcripts o f the MAGE-11 gene. MAGE-11 is one member o f a family of highly conserved genes, clustered on the long arm of the X chromosome, that are expressed in tumors o f various histological types including, metastatic melanomas and carcinomas of the breast and bladder (De Plaen et al., 1994). Due to their limited expression in normal tissues (i.e., primarily in the testis and placenta), the MAGE genes and their protein products have attracted considerable interest in recent years due to their potential efficacy in tumor immunotherapy protocols (e.g., see Hoon et al., 1995). While the functions of the MAGE proteins remain unknown, MAGE-11 was recently shown to be a predominantly nuclear protein with an experimental molecular mass of about 48 kDa (Jurk et al., 1998). This finding was puzzling since the M AG E-ll cDNA cloned by De Plaen et al. (1994) predicts a protein of 35.5 kDa lacking a consensus NLS. Our isolation of novel MAGE-11 transcripts, which contain sequences from three previously unknown 5’ exons, has resolved this apparent conflict between the observed and predicted biochemical characteristics o f MAGE-11. These novel cDNAs encode a putative MAGE-11 protein with a molecular mass of 48 kDa that contains an authentic NLS at its N-terminus (Irvine and Coetzee, 1999). 213 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. MAGE-II mRNA function inA R transactivation? Despite the presence o f the newly defined, full-length MAGE-11 ORE in each of the Xq28 cDNA clones, MAGE-11 is not a putative ARIP. Indeed, none of the Xq28 pACT2 vectors expresses MAGE-11 as a fusion with GAL4 AD (see Figs. 5.5 and 5.6). Four of the 5 Xq28 vectors express length variants of the same protein (i.e., ARIP5/28) encoded in the arbitrarily defined ORF-1. The other Xq28 vector, on the other hand, encodes a short polypeptide (i.e., ARIP25) in ORF-2, which immediately precedes the MAGE-11 ORF (see Fig. 5.5). Thus, in the yeast two-hybrid screen, 5 independent pACT2 clones containing the MAGE-11 cDNA were isolated that express two different and apparently artifactual proteins (i.e., formed by the random ligation of cDNA into the pACT2 cloning site) as fusions with GAL4 AD, neither o f which is MAGE-11. While these Xq28 pACT2 clones do not express the same protein, they do, however, express the same mRNA molecule. It is possible, therefore, that it is the MAGE-11 transcript which interacts with the AR NTD. If this were the case, the MAGE- 11 mRNA itself would have to possess transactivation potential since the GAL4 AD is not present as a modular protein. Even though such a proposal seems perhaps unorthodox or even far-fetched, it is not without precedent. Indeed, Lanz et al. (1999) recently identified steroid receptor RNA activator (SRA) from a two-hybrid screen using the PR NTD (i.e., A isoform) as bait. This novel transcript was shown to selectively enhance AF-1-mediated transactivation of several steroid receptors, including the AR, 214 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. when over-expressed in mammalian cells (Lanz et al., 1999). SRA transcript, furthermore, was shown to be present in the p i60 coactivator complex in vivo (Lanz et al., 1999). Thus, mRNA transcripts can interact with NRs and participate in the regulation of receptor-mediated transcriptional activation. To assess whether or not MAGE-11 mRNA functions in AR-mediated transactivation, recombinant MAGE-11 was cloned into a mammalian expression vector in each reading frame. One would predict that all three MAGE-11 constructs would have similar effects on AR transactivation when transfected into mammalian cells if it is the hLAGE-11 transcript that possesses regulatory potential. Unfortunately, none of the MAGE-11 constructs had an appreciable effect on ligand-dependent AR signaling when transiently expressed in PC-3 cells. All of the constructs did, however, modestly enhance GRIP 1-mediated coactivation of the AR in a ligand-dependent fashion. Thus, preliminarily, it does appear that the MAGE-II transcript may play some, albeit limited, role in AR transactivation. It may be that MAGE-11 mRNA, like SRA, serves as a facilitator of NR signaling by allowing for the efficient interaction of the liganded and DNA-bound receptor with the pl60-containing coactivation complex. The failure of MAGE-11 mRNA to impact AR signaling by itself may be a reflection of limiting p i60 concentrations in this cell line, though this remains to be shown. Certainly, some important experiments must be done in order to confirm a role for the MAGE-11 transcript in AR transactivation. For example, transfection experiments should be done in the presence of the de novo protein synthesis inhibitor cycloheximide. 215 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Under such conditions, MAGE-11 transcript expressed from a transiently transfected vector should retain its ability to coactivate pl60-mediated AR transactivation. Also, immunoprecipitation assays should be used to confirm that MAGE-11 mRNA associates with the AR and/or the pl60-containing coactivator complex in vivo. Assuming that positive results were obtained from these experiments, one could begin to assess the secondary structure of the mRNA molecule and identify which regions or ‘domains’ are important for its putative function in AR signaling. Perhaps the most important question to be addressed when considering the putative role of a novel protein/RNA partner is whether or not its interaction with the target protein makes logical scientific sense. For instance, if the two proteins are localized in different cellular compartments or are not coincidentally expressed in the same tissue(s), than it is likely that observed interactions between them are artificial. In the present case, MAGE-11 mRNA and the ligand-activated AR are both localized to the nuclear compartment of transfected cells. Whether or not they are simultaneously expressed in the same cells or tissues in vivo (i.e., in the body) is another question entirely. Weak MAGE-11 mRNA expression has been detected by RT PCR in a number of malignant tissues and cell lines including HeLa cells (De Plaen et al., 1994; Jurk et al., 1998). Importantly, neither prostate tumors nor prostate cancer cell lines were evaluated for MAGE-11 mRNA expression in these studies. The expression of MAGE-11 in normal tissues was found to be limited to testis and piacenta, though prostate tissue was not evaluated (De Plaen et al., 1994). Thus, MAGE-11 mRNA expression may occur in 216 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. some normal and malignant tissues that are regulated by AR signaling though more studies are required to directly address this issue. In conclusion, preliminary assessment o f the MAGE-11 transcript indicates that it may play a role in ligand-dependent AR transactivation. Thus far, only quite modest effects on AJR activity have been observed for MAGE-11 mRNA and these were dependent on the expression of exogenous p i60 coactivator. Nevertheless, it seems unlikely that MAGE-11 mRNA, possessing transactivation potential, would randomly interact with the AR NTD in yeast and also turn out to have a reproducible effect on AR action in mammalian cells. It is difficult, however, to reconcile the evolutionary implications o f a gene whose RNA and protein products serve perhaps disparate functions in the cell. ARIP 29 ARIP 29, which weakly but reproducibly interacted with the AR NTD in yeast two-hybrid assays, appears to be a splice variant o f hASB-3. Except for its N-terminal 19 amino acids, ARIP29 is identical to hASB-3, which is a member of a newly identified family of proteins that contain a SOCS box (i.e., suppressor of cytokine signaling box) C- terminal to multiple ankyrin repeats (Hilton et al., 1998). While the function(s) of hASB- 3 is unknown, other SOCS box-containing proteins are involved in the negative regulation of cytokine signal transduction (Starr et al., 1997; Endo et al., 1997). Specifically, SOCS proteins, upregulated in response to cytokines, inhibit Janus 217 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. kinase/signal transducer and activator o f transcription (JAK/STAT)-dependent signaling pathways in the cytoplasm (Starr et al., 1997; Endo et al., 1997). It is important to note that SOCS proteins with defined roles in cytokine signaling inhibition possess an SH2 domain upstream o f the SOCS box, not ankyrin motifs like hASB-3 (Hilton et al., 1998). ARIP29/hASB-3 contains 7 conserved ankyrin-like repeats, which are found in a plethora o f nuclear and cytoplasmic proteins exhibiting much functional diversity (Bork, 1993). Multiple ankyrin repeats are believed to form tightly stacked a-helices, each with a hydrophobic patch (i.e., including conserved leucine residues) projecting outwards to facilitate protein-protein interactions (Bork et al., 1993). It has not escaped our attention that the core structure of the ankyrin motif is very reminiscent of the NR box (see Chapter 3), which also forms a helical structure with conserved leucine residues projecting outward (Torchia et al., 1997). Whether the ankyrin repeats o f ARIP29/hASB-3 mediate interactions with the AR NTD remains unknown at present. ARIP29/hASB-3 is expressed in the prostate and may modulate AR signaling Full-length ARIP29 is encoded in an approximately 2.4 kb transcript that is expressed in a number of human tissues including prostate, testis, and ovary. ARIP29 expression, therefore, occurs in some tissues that are subject to regulation by androgens. To assess whether or not ARIP29 plays a role in ligand-dependent AR signaling, it was transiendy expressed in PC-3 cells from a mammalian expression vector that encoded an 218 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. N-terminal NLS fused to ARIP29. Interestingly, ARIP29 modestly enhanced the transactivation activity of AR(Q)9 but had little effect on that of AR(Q)4 2 . This result may suggest that ARIP29 differentially modulates AR activity according to poly-Q size. One possible mechanism by which ARIP29 could increase AR transactivation activity is by increasing receptor stability, which in turn, would increase AR bio availability. Indeed, in transfected PC-3 cells, relative DHT-activated AR protein levels were increased in the presence of co-expressed ARIP29. Thus, the apparent effects of ARIP29, albeit slight, on ligand-dependent AR signaling may be due to its binding and stabilizing of the activated receptor. Obviously, these conclusions are based on preliminary findings that must be confirmed in future studies. One question of particular interest is whether ARIP29-mediated AR stabilization decreases with increasing AR poly-Q size. This partially might explain the poly-Q effect on AR transactivation activity. Do physiological interactions between ARIP29/hASB-3 and the AR make logical scientific sense? The AR is a well-characterized, dynamic regulatory factor that likely participates in multiple signal transduction pathways, in addition to the classical androgen signaling axis. It is expressed widely in human tissues wherein its normal life-cycle involves localization in both the cytoplasmic and nuclear compartments of the cell. ARIP29/hASB-3, on the contrary, is a novel, poorly characterized protein of unknown function. Based on its evolutionary relationship to other SOCS proteins, most likely it is a predominantly cytoplasmic protein involved primarily in the negative regulation of 219 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. cytokine signaling. That is not to say, however, that its involvement in cross-talk between the cytokine and androgen signaling pathways is precluded. Clearly, cells must integrate a myriad o f external signals in order to coordinate gene expression that ultimately results in the ‘appropriate’ cellular response. Thus, interactions of physiological relevance between the AR and ARIP29/hASB-3, two proteins that interact in yeast two-hybrid assays and demonstrate overlapping tissue expression patterns, may very well occur. Some final considerations In an attempt to identify novel proteins that interact with the AR NTD and modulate AR transactivation activity, a yeast two-hybrid screen was conducted of a testis cDNA library using an AR NTD-derived bait protein. This approach, while sound and potentially powerful in its scope, is largely ‘hypothesis-generating’ as opposed to ‘hypothesis driven’. Because of this, the data gleaned from such a study may be, at least initially, mostly descriptive in nature. In other words, the significance of a particular interaction immediately may not be appreciated, and therefore, experimental evidence is accumulated in an attempt to explain the union of two apparently disparate partners. The data presented in this chapter is a case in point. Following the identification of an RNA transcript and a novel protein that each interact with the AR NTD, a series of preliminary experiments were done to support or deny an underlying functional relevance to the interactions. While far from complete, the data presented with respect to the MAGE-11 220 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. transcript and ARIP29 are compelling in this regard and form a foundation upon which future studies can be built. 221 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. SUMMARY OF PRINCIPAL FINDINGS The purpose of this thesis investigation was four-fold. Firstly, it was to determine the role of AR gene allelic variation in prostate cancer risk. Secondly, it was to functionally characterize the effects of this allelic variation on AR transcriptional activation. Thirdly, it was to define the roles of various steroid receptor coactivators in modulating endocrine signaling in hormone regulated tissues. Lastly, it was to identify novel proteins that interact with the AR NTD. This work has resulted in several significant observations. 1. The distribution o f AR CAG size alleles correlates with risk of prostate cancer in African-American, Caucasian, and Asian populations. Short AR CAG alleles are most prevalent in high-risk African-Americans and least prevalent in low-risk Asians. 2. Short AR CAG alleles confer increased risk o f advanced prostate cancer in Caucasian men. 3. The AR and PSA loci synergistically interact to confer relatively high risk of prostate cancer in Caucasian men. 4. There is a highly significant inverse correlation between poly-Q length and AR transactivation activity in prostatic carcinoma cells. 222 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5. The p i60 coactivator, GRIP1, physically and functionally interacts with the AR NTD. This interaction is mediated by the C-terminus of GRIP1 and the TAU-5 region of the ARNTD. 6. Increasing AR poly-Q length inhibits p i60 coativator-mediated potentiation of AR signaling. 7. BRCA1 is a coativator of the AR in both prostate and breast cancer cell lines, working primarily through AR AF-1. 8. BRCA1 and the p i60 coactivators synergistically potentiate AR signaling in both prostate and breast cancer cell lines. 9. BRCA1 physically interacts with the AR NTD and with the GRIP1 C-terminus. These interactions are mediated by the N-terminal region of BRCA1 that contains a RING finger interaction domain. 10. BRCA1 enhances ER signaling in prostate and breast cancer cell lines, but only in the presence of co-expressed p i60 coactivator. 11. The poly-Q region of the AR NTD interacts with a novel protein, ARIP29/hASB-3, which is putatively involved in the negative regulation of cytokine signaling. In mammalian cells, ARIP29/hASB-3 may increase ligand-dependent AR expression. 223 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. REFERENCES Aamisalo, P., Palvino, J. J., and Janne, O. A. CREB-binding protein in androgen receptor-mediated signaling, Proc. Natl. Acad. Sci. USA. 95: 2122-2127, 1998. Alen, P., Claessens, F., Verhoeven, G., Rombauts, W., and Peeters, B. The androgen receptor amino-terminal domain plays a key role in p i60 coactivator-stimulated gene transcription, Mol. Cell. Biol. 19: 6085-6097, 1999. Altman, D. G., Lausen, B., Sauerbrei, W., and Schumacher, M. Dangers of using "optimal" cutpoints in the evaluation of prognostic factors, J. Natl. Cancer Inst. 86: 829-835, 1994. Altschul, S. F., Madden, T. L., Schaffer, A. A., Zhang, J., Zhang, Z., Miller, W., and Lipman, D. J. Gapped BLAST and PSI-BLAST: a new generation o f protein database search programs, Nucleic Acids Res. 25: 3389-3402, 1997. Altschul, S. F., Gish, W., Miller, W., Myers, E. W., and Lipman, D. J. Basic local alignment search tool, J. Mol. Biol. 215: 403-410, 1990. Anderson, S. F., Schlegel, B. P., Nakajima, T., Wolpin, E. S., and Parvin, J. D. BRCA1 protein is linked to the RNA polymerase II holoenzyme complex via RNA helicase A, Nat. Genet. 19: 254-256, 1998. Anzick, S. L., Kononen, J., Walker, R. L., Azorsa, D. O.,Tanner, M. M., Guan, X-Y., Sauter, G., Kallioniemi, O-P., Trent, J. M., and Meltzer, P. S. AIB1, a steroid receptor coactivator amplified in breast and ovarian cancer, Science. 277: 965- 968, 1997. Aprelikova, O. N., Fang, B. S., Meissner, E. G., Cotter, S., Campbell, M., Kuthiala, A., Bessho, M., Jensen, R. A., and Liu, E. T. BRCA1-associated growth arrest is RB- dependent, Proc. Natl. Acad. Sci. USA. 96: 11866-11871, 1999. Arbizu, T., Santamaria, J., Gomez, J. M., Quilez, A., and Serra, J. P. A. family with adult spinal and bulbar muscular atrophy X-linked inheritence and associated with testicular failure, J. Neurol. Sci. 59: 371-382, 1983. Aumuller, G. Morphologic and endocrine aspects o f prostatic function, Prostate. 4: 195- 214, 1983. 224 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Bartsch, G. and Rohr., H. P. Comparative light and electron microscopic study of the human, dog, and rat prostate, Urol. Int. 35: 91-104, 1980. Bell, J. I. Polygenic disease, Curr. Opin. Genet. Dev. 3: 466-469, 1993. Benlel, J. M., and Tilley, W. D. Androgen receptors in prostate cancer, J. Endocrinol. 151: l-l 1, 1996. Berg, J. M. DNA binding specificity of steroid receptors, Cell. 57: 1065-1068, 1989. Berrevoets, C. A., Doesburg, P., Steketee, K., Trapman, J., and Brinkmann, A. O. Functional interactions of the AF-2 activation domain core region of the human androgen receptor with amino-terminal domain and with the transcriptional coactivator TIF2 (transcriptional intermediary factor 2), Mol. Endocrinol. 12: 1172-1183, 1998. Berry, R., Schroeder, J. J., French, A. J., McDonnel, S. K., Peterson, B. J., Cunningham, J. M., Thibodeau, S. N., and Schaid, D. J. Evidence for a prostate cancer- susceptibility locus on chromosome 20, Am. J. Hum. Genet, in press, 2000. Berstein, L., and Ross, R. K. Cancer in Los Angeles: a portrait of incidence and mortality. Los Angeles: University of Southern California Press, 1991. Berthon, P., Valeri, A., Cohen-Akenine, A., Drelon, E., Paiss, T., Wohr, G., Latil, A., et al. Predisposing gene for early-onset prostate cancer, localized on chromosome lq42.2-43, Am. J. Hum. Genet. 62: 1416-1424, 1998. Bertwistle, D., and Ashworth, A. Functions of the BRCA1 and BRCA2 genes, Curr. Opin. Genet. Dev. 8: 14-20, 1998. Birrell, S. N., Bentel, J. M., Hickey, T. E., Ricciardelli, C., Weger, M. A., Horsfall, D. J., and Tilley, W. D. Androgens induce divergent proliferative responses in human breast cancer cell lines, J. Steroid Biochem. Molec. Biol. 52: 459-467, 1995. Bork, P. Hundreds o f ankyrin-like repeats in functionally diverse proteins: mobile modules that cross phyla horizontally, Proteins. 17: 363-374, 1993. Bradford, M. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding, Anal. Biochem. 72: 248- 254, 1976. 225 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Bratt, O., Borg, A., Kristoffersson, U., Lundgren, R., Zhang, Q-X., and Olsson, H. CAG repeat length in the androgen receptor gene is related to age at diagnosos of prostate cancer and response to endocrine therapy, but not to prostate cancer risk, Br. J. Cancer. 81: 672-676, 1999. Brent, R., and Finley, R. L. Understanding gene and allele function with two-hybrid methods, Annu. Rev. Genet. 31: 663-704, 1997. Breslow, N. E., and Day N. E. Statistical methods in cancer research Vol. 1. Lyon: IARC, 1980. Brown, C. J., Goss, S. J., Lubahn, D. B., Joseph, D. R., Wilson, E. M., French, F. S., and Willard, H. F. Androgen receptor locus on the human X chromosome: regional localization to X qll-12 and description of a DNA polymorphism, Am. J. Hum. Genet. 44: 264-269, 1989. Brzozowski, A. M., Pike, A. C. W., Dauter, Z., Hubbard, R. E., Bonn, T., Engstrom, O., Ohman, L., Greene, G. L., Gustafsson, J-A., and Carlquist, M. Molecular basis of agonism and antagonism in the oestrogen receptor, Nature. 389: 753-758, 1997. Butler, R., Leigh, P. N., McPhaul, M. J., and Gallo, J-M. Truncated forms of the androgen receptor are associated with polyglutamine expansion in X-linked spinal and bulbar muscular atrophy, Hum. Mol. Genet. 7: 121-127, 1998. Carter, B. S., Bova, G. S., Beaty, T. H., Steinberg, G. D., Childs, B., Isaacs, W. B., and Walsh, P. C. Hereditary prostate cancer: epidemiologic and clinical features, J. Urol. 150: 797-802, 1993. Carter, B. S., Beaty, T. H., Steinberg, G. D., Childs, B., and Walsh, P. C. Mendelian inheritance of familial prostate cancer, Proc. Natl. Acad. Sci. USA. 89: 3367- 3371, 1992. Chai, Y., Cui, J., Shao, N., Reddy, E. S. P., and Rao, V. N. The second BRCT domain of BRCA1 proteins interacts with p53 and stimulates transcription from the pliW A Fl/C IPl promoter, Oncogene. 18: 263-268, 1999. Chamberlain, N. L., Whitacre, D. C., and Miesfeld, R. L. Delineation of two distinct type I activation functions in the androgen receptor amino-terminal domain, J. Biol. Chem. 271: 26772-26778, 1996. 226 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chamberlain, N. L., Driver, E. D., and Miesfeld, R. L. The length and location o f CAG trinucleotide repeats in the androgen receptor N-terminal domain affect transactivation function, Nucleic Acids Res. 22: 3181-3186, 1994. Chan, J. M., Stampfer, M. J., Giovannucci, E., Gann, P. H., Ma, J., Wilkinson, P., Hennekens, C. H., and Poliak, M. Plasma insulin-like growth factor-I and prostate cancer risk: a prospective study, Science. 279: 563-566, 1998. Chang, C., Kokontis, J., and Liao, S. Structural analysis of complementary DNA and amino acid sequences o f human and rat androgen receptors, Proc. Natl. Acad. Sci. USA. 85: 7211-7215, 1988. Chapman, M. S., and Verma, I. M. Transcriptional activation by BRCA1, Nature. 382: 678-679, 1996. Chen, D., Ma, H., Hong, H., Koh, S. S., Huang, S-M., Schurter, B. T., Aswad, D. W., and Stallcup, M. R. Regulation of transcription by a protein methyltransferase, Science. 284: 2174-2177, 1999a. Chen, H., Lin, R. J., Schiltz, R. L., Chakravarti, D., Nash, A., Nagy, L., Privalsky, M. L., Nakatani, Y., and Evans, R. M. 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, 1997. Chen, J. J., Silver, D., Cantor, S., Livingston, D. M., and Scully, R. BRCA1, BRCA2, and Rad51 operate in a common DNA damage response pathway, Cancer Res. 59: 1752s-1756s, 1999b. Chien, C.-T., Bartel, P. L., Stemglanz, R., and Fields, S. The two-hybrid system: a method to identify and clone genes for proteins that interact with a protein of interest, Proc. Natl. Acad. Sci. USA. 88: 9578-9582, 1991. Choong, C. S., Kemppainen, J. A., and Wilson, E. M. Evolution of the primate androgen receptor: a structural basis for disease, J. Mol. Evol. 47: 334-342, 1998. Choong, C. S., Kemppainen, J. A., Zhou, Z., and Wilson, E. M. Reduced androgen receptor gene expression with the first exon CAG repeat expansion, Mol. Endocrinol. 10: 1527-1535, 1996. Coetzee, G. A., and Ross, R. K. Prostate cancer and the androgen receptor, J. Natl. Cancer Inst. 86: 872-873, 1994. 227 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Cohen, P. Serum insulin-like growth factor-I levels and prostate cancer risk: interpreting the evidence, J. Natl. Cancer Inst. 90: 876-879, 1998. Cohen, P., Peehl, D. M., and Rosenfeld, R. G. Biological effects of prostate specific antigen as an insuline-iike growth factor binding protein-3 protease, J. Endocrinol. 142: 407-415, 1994. Collingwooa, T. N., Umov, F. D., and Wolffe, A. P. Nuclear receptors: coactivators, corepressors and chromatin remodeling in the control of transcription, J. Mol. Endocrinol. 23: 255-275, 1999. Cook, P. J., Doll, R., and Fellingham, S. A. A mathmatical model for the age distribution of cancer in man, Int. J. Cancer. 4: 93-112, 1969. Comeiis, R. S., Neuhausen, S. L., Johansson, O., Aarson, A., Kelseil, D. et al. High allele loss rates at 17ql2-q21 in breast and ovarian tumors from BRCAl-linked families, Genes Chrom. Cancer. 13: 203-210, 1995. Correa-Cerro, L., Wohr, G., Haussler, J., Berthon, P., Drelon, E., Mangin, P., Fournier, G., Cussenot, O., Kraus, P., Just, W., Paiss, T., Cantu, J. M., and Vogel, W. (CAG)nCAA and GGN repeats in the human androgen receptor gene are not associated with protstate cancer in a French-German population., Eur. J. Hum. Genet. 7: 357-362, 1999. Craft, N., Shostak,Y., Carey, M., and Sawyers, C. L. A mechanism for hormone- independent prostate cancer through modulation o f androgen receptor signaling by the HER-2/neu tyrosine kinase, Nat. Med. 5: 280-285, 1999. Crook, T., Crossland, S., Crompton, M. R., Osin, P., and Gusterson, B. A. p53 mutations in BRCA1-associated familial breast cancer, Lancet. 350: 638-639, 1997. Culig, Z., Hobisch, A., Cronauer, M. V., Radmayr, C., Trapman, J., Hittmair, A., Bartsch, G., and Klocker, H. Androgen receptor activation in prostatic tumor cell lines by insuline-like growth factor-I, keratinocyte growth factor, and epidermal growth factor, Cancer Res. 54: 5474-5478, 1994. Cunha, G. R., Donjacour, A. A., Cooke, P. S., Mee, S., Bigsby, R. M., Higgins, S. J., and Sugimura, Y. The endocrinology and developmental biology of the prostate, Endocrine Rev. 8: 338-362, 1987. 228 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Danielian, P. S., White, R., Lees, J. A., and Parker, M. G. Identification of a conserved region required for hormone-dependent transactivation by steroid hormone receptors, EMBO J. 11: 1025-1033, 1992. De Plaen, E., Arden, K., Traversari, C., Gaforio, J. J., Szikora, J-P., De Smet, C., Brasseur, F., van der Bruggen, P., Lethe, B., Lurquin, C., Brasseur, R., Chomez, P., De Backer, O., Cavenee, W., and Boon, T. Structure, chromosomal localization, and expression of 12 genes of the MAGE family, Immunogenetics. 40: 360-369, 1994. Ding, X. F., Anderson, C. M., Ma, H., Hong, H., Uht, R. M., Kushner, P. J., and Stallcup, M. R. 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, 1998. Doesberg, P., Kuil, C. W., Berrevoets, C. A., Steketee, K., Faber, P. W., Mulder, E., Brinkmann, A. O., and Trapman J. Functional in vivo interaction between the amino-terminal, transactivation domain and the ligand binding domain of the androgen receptor, Biochemistry. 36: 1052-1064, 1997. Dowsing, A. T., Yong, E. L., Clark, M., McLachlan, R. I., de Kretser, D. M., and Trounson, A. O. Linkage between male infertility and trinucleotide repeat expansion in the androgen receptor gene, Lancet. 354: 640-643, 1999. Dupont, A., Gomez, J-L., Cusan, L., Koutsilieris, M., and Labrie, F. Response to flutamide withdrawal in advanced prostate cancer in progression under combination therapy, J. Urol. 150: 908-913, 1993. Edwards, A., Hammond, H. A., Jin, L., Caskey, T., and Chakraborty, R. Genetic variation at five trimeric and tetrameric tandem repeat loci in four human population groups, Gemomics. 12: 241-253, 1992. Edwards, S. M., Badzioch, M. D., Minter, R., Hamoudi, R., Collins, N., Ardem-Jones, A., Dowe, A., Osbome, S., Kelly, J., Shearer, R., Easton, D. F., Saunders, G. F., Deamaley, D. P., and Eeles, R. A. Androgen receptor polymorphisms: association with prostate cancer risk, relapse and overall survival, Int. J. Cancer (Pred. Oncol.). 84: 458-465, 1999. Eeles, R., and Kadouri, L. BRCAl/2 carriers and endocrine risk modifiers, Endocrine- Rel. Cancer. 6: 521-528, 1999. 229 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Eeles, R. A., Durocher, F., Edwards, S., Teare, D., Badzioch, M., Hamoudi, R., Gill, S., Biggs, P., Deamaley, D., Ardem-Jones, A., et al. Linkage analysis of chromosome lq markers in 136 prostate cancer families, Am. J. Hum. Genet. 62: 653-658, 1998. Ekman, P., Gronberg, H., Matsuyama, H., Kivineva, M., Bergerheim, U. S. R., and Li, C. Links between genetic and environmental factors and prostate cancer risk, Prostate. 39: 262-268, 1999. Endo, T. A., Masuhara, M., Yokouchi, M., Suzuki, R., Sakamoto, H., Mitsui, K., Matsumoto, A., Tanimura, S., Ohtsubo, M., Misawa, H., et al. A new protein containing an SH2 domain that inhibits JAK kinases, Nature. 387: 921-924, 1997. Fan, S., Wang, J-A., Yuan, R., Ma, Y., Meng, Q., Erdos, M. R., Pestell, R. G., Yuan, F., Aubom, K. J., Goldberg, I. D., and Rosen, E. M. BRCA1 inhibition of estrogen receptor signaling in transfected cells, Science. 284: 1354-1356, 1999. Feng, W., Ribeiro, R. C. J., Wagner, R. L., Nguyen, H., Apriletti, J. W., Fletterick, R. J., Baxter, J. D., Kushner, P. J., and West, B. L. Hormone-dependent coactivator binding to a hydrophobic cleft on nuclear receptors, Science. 280: 1747-1749, 1998. Fenton, M. A., Shuster, T. D., Fertig, A. M., Taplin, M-E., Kolvenbag, G., Bubley, G. J., and Balk, S. P. Functional characterization of mutant androgen receptors from androgen-independent prostate cancer, Clin. Cancer Res. 3: 1383-1388, 1997. Feunteun, J. Breast cancer and genetic instability: the molecules behind the scenes, Mol. Med. Today 263-267, 1998. Fields, S., and Song, O. A novel genetic system to detect protein-protein interactions, Nature. 340.^246-246, 1989. Fischbeck, K. H. Other X-linked Neuromuscular Diseases. In: L. P. Rowland, Wood, D. S., Schon, E. A., and DiMauro, S. (ed.) Molecular genetics in diseases of brain, nerve, and muscle. New York: Oxford University Press, 1989. Franks, L. M. Latent carcinoma of the prostate, J. Pathol. Bacteriol. 68: 603-616, 1954. Fronsdal, K., Engedal, N., Slagsvold, T., and Saatcioglu, F. CREB binding protein is a coactivator for the androgen receptor and mediates cross-talk with AP-1, J. Biol. Chem. 273: 31853-31859, 1998. 230 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Gao, T., Marcelli, M., and McPaul, M. J. Transcriptional activation and transient expression of the human androgen receptor, J. Steroid Biochem. Molec. Biol. 59: 9-20, 1996. Gast, A., Schneikert, J., and Cato, C. B. N-terminal sequences of the human androgen receptor in DNA binding and transrepressing functions, J. Steroid Biochem. Molec. Biol. 65: 117-123, 1998. Gerber, H.-P., Seipel, K., Georgiev, O., Hofferer, M., Hug, M., Rusconi, S., and Schaffner, W. Transcriptional activation modulated by homopolymeric glutamine and proline stretches, Science. 263: 808-811, 1994. Gibbs, M., Stanford, J. L., Jarvik, G. P., Janer, M., Badzioch, M., Peters, M. A., Goode, E. L., Kolb, S., Chakrabarti, L., Shook, M., Basom, R., Ostrander, E. A., and Hood, L. A genomic scan o f families with prostate cancer identifies multiple regions of interest, Am. J. Hum. Genet, in press, 2000. Gibbs, M., Chakrabarti, L., Stanford, J. L., Goode, E. L., Kolb, S., Schuster, E. F., Buckley, V. A., Shook, M., Hood, L., Jarvik, G. P., and Ostrander, E. A. Analysis o f chromosome lq42.2-43 in 152 families with high risk o f prostate cancer, Am. J. Hum. Genet. 64: 1087-1095, 1999. Gietz, D., St. Jean, A., Woods, R. A., and Schiestl, R. H. Improved method for high efficiency transformation of intact yeast cells, Nucleic Acids Res. 20: 1425, 1992. Giguere, V., Hollenberg, S. M., and Evans, R. M. Functional Domains of the human glucocorticoid receptor, Cell. 46: 645-652,1986. Giniger, E., Vamum, S. M., and Ptashne, M. Specific DNA binding o f GAL4, a positive regulatory protein of yeast, Cell. 40: 767-774, 1985. Giovannucci, E., Platz, E. A., Stampfer, M. J., Chan, A., Krithivas, K., Kawachi, I., Willett, W. C., and Kantoff, P. W. The CAG repeat within the androgen receptor gene and benign prostatic hyperplasia. Urology. 53: 121-125, 1999a. Giovannucci, E., Stampfer, M. J., Chan, A., Krithivas, K., Gann, P. H., Hennekens, C. H., and Kantoff, P. W. CAG repeat within the androgen receptor gene and incidence of surgery for benign prostatic hyperplasia in U. S. physicians, Prostate. 39: 130-134, 1999b. 231 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Giovannucci, E., Stampfer, M. J., Krithivas, K., Brown, M., Brufsky, A., Talcott, J., Hennekens, C. H., and Kantoff, P. W. The CAG repeat within the androgen receptor gene and its relationship to prostate cancer, Proc. Natl. Acad. Sci. USA. 94: 3320-3323, 1997. Given, H. F., Radboume, R., Oag, H., Merritt, S., Barclay, E., Hanby, A. M., Lamlum, H., McGrath, J., Curran, C., and Tomlinson, I. P. The androgen receptor exon 1 trinucleotide repeat does not act as a modifier of the age of presentation in breast cancer, Eur. J. Cancer. 36: 533-534, 2000. Gleason, D. F., Mellinger, G. T., et al. Prediction of prognosis for prostatic adenocarcinoma by combinded histological grading and clinical staging, J. Urol. Ill: 58-64, 1974. Goldenberg, I. S., Sedransk, N., Volk, H., Segaloff, A., Kelley, R. M., and Haines, C. R. Combined androgen and antimetabolite therapy of advanced female breast cancer, Cancer. 36: 308-310, 1975. Gowen, L. C., Avrutskaya, A. V., Latour, A. M., Koller, B. H., and Leadon, S. A. BRCA1 required for transcription-coupled repair of oxidative DNA damage, Science. 281: 1009-1012, 1998. Green, S., Kumar, V., Theulaz, I., Wahli, W., and Chambon, P. The N-terminal DNA- binding 'zinc finger' of the oestrogen and gluccocorticoid receptors determines target specificity, EMBO J. 7; 3037-3044, 1988a. Green, S., Issemann, I., and Sheer, E. A versatile in vivo and in vitro eukaryotic expression vector for protein engineering, Nucleic Acids Res. 16: 369, 1988b. Greenlee, R., Murray, T., Bolden, S., and Wingo, P. Cancer statistics, 2000, CA Cancer J. Clin. 50: 7-33, 2000. Gronberg, H., Damber, L., Damber, J-E., and Lennart, I. Segregation analysis of prostate cancer in Sweden: support for dominant inheritance, Am. J. Epidemiol. 146: 552- 557, 1997. Hackem, R., de la Pompa, J. L., Elia, A., Potter, J., and Mak, T. W. Partial rescue of Brcal^ ^ early embryonic lethality by p53 or p21 null mutation, Nat. Genet. 16: 298-302, 1997. 232 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Hackenberg, R., and Schulz, K-D. Androgen receptor mediated growth control o f breast cancer and endometrial cancer modulated by antiandrogen- and androgen-like steroids, J. Steroid Biochem. Molec. Biol. 56: 113-117, 1996. Haile, D. T., and Parvin, J. D. Activation of transcription in vitro by BRCA1 carboxyl- terminal domain, J. Biol. Chem. 274: 2113-2117, 1999. Hakimi, J. M., Schoenberg, M. P., Rondinelli, R. H-, Piantadosi, S., and Barrack, E. R. Androgen receptor variants with short glutamine of glycine repeats may identify unique subpopulations of men with prostate cancer, Clin. Cancer Res. 3: 1599- 1608, 1997. Hardy, D. O., Scher, H. I., Bodenreider, T., Sabbatini, P., Zhang, Z-F., Nanus, D. M., and Catterall, J. F. Androgen receptor CAG repeat lengths in prostate cancer: correlation with age of onset, J. Clin. Endocrinol. Met. 81: 4400-4405, 1996. He, B., Kemppainen, J. A., Voegel, J. J., Gronemeyer, H., and Wilson, E. M. 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, 1999. Heery, D. M., Kalkhoven, E., Hoare, S., and Parker, M. G. A signature motif in transcriptional co-activators mediates binding to nuclear receptors, Nature. 387: 733-736, 1997. Hill, J., Donald, K. A. I. G., and Griffiths, D. E. DMSO-enhanced whole cell yeast transformation, Nucleic Acids Res. 19: 5791, 1991. Hilton, D. J., Richardson, R. T., Alexander, W. S., Viney, E. M., Wilson, T. A., Sprigg, N. S., Starr, R., Nicholson, S. E., Metcalf, D., and Nicola, N. A. Twenty proteins containing a C-terminal SOCS box form five structural classes, Proc. Natl. Acad. Sci. USA. 95: 114-119, 1998. Hiiji, K. F., Tan S. J., and Elashoff, R. M. A quasi-exact test for comparing two binomial proportions, StatMed. 10: 1137-1153, 1991. Hong, H., Kohli, K., Garabedian, M. J., and Stallcup, M, R. GRIP1, a transcriptional coactivator for the AF-2 transactivation domain of steroid, thyroid, retinoid, and vitamin D receptors, Mol. Cell. Biol. 17: 2735-2744, 1997. 233 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Hong, H., Darimont, B. D., Ma, H., Yang, L., Yamamoto, K. R., and Stallcup, M. R. An additional region of coactivator GRIP1 required for interaction with the hormone- binding domains of a subset of nuclear receptors, J. Biol. Chem. 274: 3496-3502, 1999. Hong, H., Kohli, K., Trivedi, A., Johnson, D. L., and Stallcup, M. R. GRIP1, a novel mouse protein serves as a transcriptional coactivator in yeast for the hormone binding domains of steroid receptors, Proc. Natl. Acad. Sci. USA. 93: 4948-4952, 1996. Hoon, D. S. B., Yuzuki, D., Hayashida, M., and Morton, D. L. Melanoma patients immunized with melanoma cell vaccine induce antibody responses to recombinant MAGE-1 antigen, J. Immunol. 154: 730-737, 1995. Hsiao, P.-W., Lin, D-L., Nakao, R., and Chang, C. The linkage of Kennedy's neuron disease to ARA24, the first identified androgen receptor polyglutamine region- associated coactivator, J Biol. Chem. 274: 20229-20234, 1999. Hsieh, C.-L., Oakley-Girvan, I., Gallagher, R. P., Wu, A. H., Kolonel, L. N., Teh, C-Z., Halpem, J., West, D. W., Paffenbarger Jr., R. S., and Whittemore, A. S. Prostate cancer susceptibility locus on chromosome lq: a confirmatory study, J. Natl. Cancer Inst. 89: 1893-1894, 1997. Hubert, A., Peretz, T., Manor, O., Kaduri, L., Wienberg, N., Lerer, I., Sagi, M., and Abeliovich, D. The Jewish Ashkenazi founder mutations in the BRCA1/BRCA2 genes are not found at an increased frequency in Ashkenazi patients with prostate cancer, Am. J. Hum. Genet. 65: 921-924, 1999. Huggins, C., Stevens Jr., R. E., and Hodges, C. V. Studies on prostate cancer II. The effects of castration on advanced carcinoma o f the prostate gland, Arch. Surg. 43: 209-223, 1941. Humphrey, P. A., Frazier, H. A., Vollmer, R. T., and Paulson, D. F. Stratification of pathologic features in radical prostatectomy specimens that are predictive of elevated initial postoperative serum prostate-specific antigen levels, Cancer. 71: 1821-1827, 1993. Ikonen, T., Palvino, J. J., and Janne, O. A. Interaction between the amino- and carboxy- terminal regions of the rat androgen receptor modulates transcriptional activity and is influenced by nuclear receptor coactivators, J. Biol. Chem. 272: 29821- 29828,1997. 234 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Ingles, S. A., Ross, R. K., Yu, M. C., Irvine, R. A., La Pera, G., Haile, R. W., and Coetzee, G. A. Association o f prostate cancer risk with genetic polymorphisms in vitamin D receptor and androgen receptor, J Natl. Cancer Inst. 89: 166-170, 1997. Irminger-Finger, I., Siegel, B. D., and Leung, W-C. The functions of breast cancer susceptibility gene 1 (BRCA1) product and its associated proteins, Biol. Chem. 380: 117-128, 1999. Irvine, R. A., and Coetzee, G. A. Additional upstream coding sequences of MAGE-11, Immunogenetics. 49: 585, 1999. Irvine, R. A., Ma, H., Yu, M. C., Ross, R. K., Stallcup, M. R., and Coetzee, G. A. Inhibition of pl60-mediated coactivation with increasing androgen receptor polyglutamine length, Hum. Mol. Genet. 9: 267-274, 2000. Irvine, R. A., Yu, M. C., Ross, R. K., and Coetzee, G. A. The CAG and GGC microsatellites of the androgen receptor gene are in linkage disequalibrium in men with prostate cancer, Cancer Res. 55: 1937-1940, 1995. Isaacs, J. T. Prostatic cancer: an overview, Cancer Metastasis Rev. 12: 1-2, 1993. Isaacs, W. B., Bova, G. S., Morton, R. A., Bussemakers, M. J. G., Brooks, J. D., and Ewing, C. M. Molecular biology of prostate cancer, Seminars in Oncology. 21: 514-521, 1994. Isola, J. J. Immunohistochemical demonstration o f androgen receptor in breast cancer and its relationship to other prognostic factors, J. Pathol. 170: 31-35, 1993. Jacobsen, S. J., Katusic, S. K., Bergstralh E. J., Oesterling J. E., Ohrt, D., Klee, G. G., Chute, C. G., et al. Incidence of prostate cancer diagnosis in the eras before and after serum prostate-specific antigen testing, JAMA. 274: 1445-1449, 1995. Jenster, G., Trapman, J., and Brinkmann, A. O. Nuclear import of the human androgen receptor, Biochem. J. 293: 761-768, 1993. Jenster, G., van der Korput, H. A. G. M., Trapman, J., and Brinkmann, A. O. Identification o f two transcription activation units in the N-terminal domain of the human androgen receptor, J. Biol. Chem. 270: 7341-7346, 1995. Jenster, G., van der Korput, H. A. G. M., van Vroonhoven, C., van der Kwast, T. H., Trapman, J., and Brinkmann, A. O. Domains o f the human androgen receptor involved in steroid binding, transcriptional activation, and subcellular localization, Mol. Endocrinol. 5: 1396-1404, 1991. 235 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Jurk, M., Kremmer, E., Schwarz, U., Forster, R., and Winnacker, E-L. MAGE-11 protein is highly conserved in higher organisms and located predominantly in the nucleus, Int. J. Cancer. 75: 762-766, 1998. Karp, S. E., Tonin, P. N., Begin, L. R., Martinez, J. J., Zhang, J. C., Poliak, M. N., and Foulkes, W. D. Influence of BRCA1 mutations on nuclear grade and estrogen receptor status of breast carcinoma in Ashkenazi Jewish women, Cancer. 80: 435- 441, 1997. Kasper, S., Rennie, P. S., Bruchovsky, N., Lin, L., Cheng, H., Snoek, R., Dahlman- Wright, K., Gustafsson, J-A., Shiu, R. P. C., Sheppard, P. C., and Matusik, R. J. Selective activation o f the probasin androgen-responsive region by steroid hormones, J. Mol. Endocrinol. 22: 313-325, 1999. Kasper, S., Rennie, P. S., Bruchovsky, N., Sheppard, P. C., Cheng, H., Lin, L., Shiu, R. P. C., Snoek, R., and Matusik, R. J. Cooperative binding of androgen receptors to two DNA sequences is required for androgen induction of the probasin gene, J. Biol. Chem. 269: 31763-31769, 1994. Kawasaki, T., Ushiyama, T., Ueyama, H., Inoue, K., Mori, K., Ohkubo, I., and Hukuda, S. Polymorphic CAG repeats of the androgen receptor gene and rheumatoid arthritis, Ann Rheum. Dis. 58: 500-502, 1999. Kazemi-Esfaijani, P., Trifiro, M. A., and Pinsky, L. Evidence for a repressive function of the long polyglutamine tract in the human androgen receptor: possible pathogenetic relevance for the (CAG)n-expanded neuropathies, Hum. Mol. Genet. 4: 523-527, 1995. Kirkels, W. J. Natural history o f prostate cancer. In: Z. Petrovich, Baert, L., and Brady, L. W. (ed.) Carcinoma of the prostate: innovations in management, pp. 13-19. Berlin: Springer-Verlag, 1996. Knudsen, K. E., Arden, K. C., and Cavenee, W. K. Multiple G1 regulatory elements control the androgen-dependent proliferation of prostatic carcinoma cells, J. Biol. Chem. 273: 20213-20222, 1998. Koivisto, P., Kononen, J., Palmberg, C., Tammela, T., Hyytinen, E., Isola, J., Trapman, J., Cleutjens, Noordzij, A., Visakorpi, T., and Kallioniemi, O-P. Androgen receptor gene amplification: a possible molecular mechanism for androgen deprivation therapy failure in prostate cancer, Cancer Res. 57: 314-319, 1997. 236 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Kokontis, J. M., and Liao, S. Molecular action of androgen in the normal and neoplastic prostate, Vitamins and Hormones. 55; 219-307, 1999. Kosary, C. L., Ries, L. A. G., Miller, B. A., Hankey, B. F., Harras, A., and Edwards, B. K. SEER cancer statistics review, 1973-1992: tables and graphs. NIH publ 96- 2789. Bethesda: National Cancer Institute, 1995. Kozlowski, J. M., Ellis, W. J., and Grayhack, J. T. Advanced prostatic carcinoma: early v's late endocrine therapy. In: G. L. Andriole, and Catalona, W. J. (ed.) The urologic clinics of North America, pp. 15-24. Philadelphia: W. B. Saunders Co., 1991. La Spada, A. R., Roling, D. B., Harding A. E., Warner, C. L., Spiegel, R., Hausmanowa- Petrusewicz, I., Yee, W. C., and Fischbeck, K. H. Meiotic stability and genotype- phenotype correlation of the trinucleotide repeat in X-linked spinal and bulbar muscular atrophy, Nat. Genet. 2: 301-304, 1992. La Spada, A. R., Wilson, E. M., Lubahn, D. B., Harding, A. E., and Fischbeck, K. H. Androgen receptor gene mutations in X-linked spinal and bulbar muscular atrophy, Nature. 352: 77-79, 1991. Langley, E., Zhou, Z., and Wilson, E. M. Evidence for an antiparallel orientation of the ligand-activated human androgen receptor dimer, J. Biol. Chem. 270: 29983- 29990, 1995. Langston, A. A., Stanford, J. L., Wicklund, K. G., Thompson, J. D., Blazej, R. G., and Ostrander, E. A. Germ-line BRCA1 mutations in selected men with prostate cancer, Am. J. Hum. Genet. 58: 881-885, 1996. Lanz, R. B., McKenna, N. J., Onate, S. A., Albrecht, U., Wong, J., Tsai, S. Y., Tsai, M-J., and O'Malley, B. W. A steroid receptor coactivator, SRA, functions as an RNA and is present in an SRC-1 complex, Cell. 97: 17-27, 1999. Lavinsky, R., Jepsen, K., Heinzel, T., Torchia, J., Mullen, T-M., Schiff, R., Del-Rio, A. L., Ricote, M., Ngo, S., Gemsch, J., Hilsenbeck, S. G., Osborne, C. K., Glass, C. K., Rosenfeld, M. G., and Rose, D. W. Diverse signaling pathways modulate nuclear receptor recruitment o f N-Cor and SMRT complexes, Proc. Natl. Acad. Sci. USA. 95: 2920-2925, 1998. Lea, O. A., Kvinnsland, S., and Thorsen, T. Improved measurement of androgen receptors in human breast cancer, Cancer Res. 49: 7162-7167, 1989. 237 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Legius, E., Vanderschueren, E., Spiessens, C., D'Hooghe, T., and Matthijs, G. Association between CAG repeat number in the androgen receptor gene and male infertility in a Belgian study, Clin. Genet. 56: 166-167, 1999. Lehmann, E. L. Nonparametrics: statistical methods based on ranks. San Francisco: Holden-Day, 1975. Lindzey, J., Kumar, M. V., Grossman, M., Young, C , and Tindall, D. J. Molecular mechanisms o f androgen action, Vitamines and Hormones. 49: 383-432, 1994. Lobaccaro, J.-M., Lumbroso, S., Belon, C., Galtier-Dereure, F., Baringer, J., Lesimple, T., Heron, J-F., Pujol, H., and Sultan, C. Male breast cancer and the androgen receptor gene, Nat. Genet. 5: 109-110, 1993. Lubahn, D. B., Brown,T. R., Simental, J. A., Higgs, H. N., Migeon, C. J., Wilson, E. M., and French, F. S. Sequence of the intron/exon junctions of the coding region of the human androgen receptor gene and identification of a point mutation in a family with complete androgen insensitivity, Proc. Natl. Acad. Sci. USA. 86: 9534-9538, 1989. Lubahn, D. B., Joseph, D. R., Sullivan, P. M., Willard, H. F., French, F. S., and Wilson, E. M. Cloning of the human androgen receptor complementary DNA and localization to the X chromosome, Science. 240: 327-330, 1988. Luisi, B. F., Xu, W. X., Otwinowski, Z., Freedman, L. P., Yamamoto, K. R., and Sigler, P. B. Crystallographic analysis of the interaction of the gluccocorticoid receptor with DNA, Nature. 352: 497-505, 1991. Lumbroso, R., Beitel, L. K., Vasiliou, D. M., Trifiro, M, A., and Pinsky, L. Codon-usage variants in the polymorphic (GGN)n trinucleotide repeat of the human androgen receptor gene, Hum. Genet. 101: 43-46. 1997. Ma, H., Hong, H., Huang, S-M., Irvine, R. A., Webb, P., Kushner, P. J., Coetzee, G. A., and Stallcup, M. R. Multiple signal input and output domains of the 160- kilodalton nuclear receptor coactivator proteins, Mol. Cell. Biol. 19: 6164-6173, 1999. Ma, J., and Ptashne, M. Deletion analysis of GAL4 defines two transcriptional activating segments, Cell. 48: 847-853, 1987. MacLean, H. E., Wame, G. L., and Zajac, J. D. Localization of functional domains in the androgen receptor, J. Steroid Biochem. Molec. Biol. 62: 233-242, 1997. 238 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Madden, S. L., Galella, E. A., Riley, D., Bertelsen, A. H., and Beaudry, G. A. Induction of cell growth regulatory genes by p53, Cancer Res. 56: 5384-5390, 1996. Mahtre, A. N., Trifiro, M. A., Kaufinan, M., Kazemi-Esfaijani, P., Figlewicz, D., Rouleau, G., and Pinsky, L. Reduced transcriptional regulatory competence of the androgen receptor in X-linked spinal and bulbar muscular atrophy, Nat. Genet. 5: 184-188, 1993. Mangelsdorf, D. J., Thummel, C., Beato, M., Herrlich, P., Schutz, G., Umesono K., Blumberg, B., Kastner, P., Mark, M., Chambon, P., and Evans, R. E. The nuclear receptor superfamily: the second decade, Cell. 83: 835-839, 1995. McEwan, I. J., and Gustafsson, J-A. Interaction of the human androgen receptor transactivation function with the general transcription factor TFIIF, Proc. Natl. Acad. Sci. USA. 94: 8485-8490, 1997. Mclndoe, R. A., Stanford, J. L., Gibbs, M., Jarvik, G. P., Brandzel, S., Neal, C. L., Li, S., Gammack, J. T., Gay, A. A., Goode, E. L., Hood, L., and Ostrander, E. A. Linkage analysis of 49 high-risk families does not support a common familial prostate cancer-susceptibility gene at lq24-25. Am. J. Hum. Genet. 61: 347-353, 1997. Mclnemey, E. M., Tsai, M-J., O'Malley, B. W., and Katzenellenbogen, B. S. Analysis of estrogen receptor transcriptional enhancement by a nuclear receptor hormone receptor coactivator, Proc. Natl. Acad. Sci. USA. 93: 10069-10073, 1996. McKenna, N. J., Lanz, R. B., and O'Malley, B. W. Nuclear receptor coregulators: cellular and molecular biology, Endocrinol. Rev. 20: 321-344, 1999. McNeal, J. E. The prostate gland: morphology and pathobiology, Monogr. Urol. 4: 3-33, 1983. Miki, Y., Swensen, J., Shattuck-Eidens, D., Futreal, P. A., Harshman, K., Tavtigian, S., Liu, Q., et al. A strong candidate for the breast and ovarian cancer susceptibility gene BRCA1, Science. 266: 66-71, 1994. Mitsumori, K., Terai, A., Oka, H., Segawa, T., Ogura, K., Yoshida, O., and Ogawa, O. Androgen receptor CAG repeat length polymorphis in benign prostatic hyperplasia (BPH): correlation with adenoma growth, Prostate. 41: 253-257, 1999. 239 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Monroe, K. R., Yu, M. C., Kolonel, L. N., Coetzee, G. A., Wilkens, L. R., Ross, R. K., and Hendersen, B. E. Evidence of an X-linked or recessive genetic component to prostate cancer risk, Nat. Med. 1: 827-829, 1995. Moras, D., and Gronemeyer, H. The nuclear receptor ligand-binding domain: structure and function, Curr. Opin. Cell Biol. 10: 384-391, 1998. Moynahan, M. E., Chiu, J. W., Roller, B. H., and Jasin, M. Brcal controls homology- directed DNA repair, Mol. Cell. 4: 511-518, 1999. Muir, C. S., Waterhouse, J., Mack, T., Powell, J., Whelan, S., and Cook, P. J. Cancer statistics in five continents. IARC sci publ 88, Vol. 5. Lyon: IARC, 1987. Nagashima, T., Seko, K., Hirose, K., Mannen, T., Yoshimura, S., Arima, R., Nagashima, K., and Morimatsu, Y. Familial bulbo-spinal muscular atrophy associated with testicular atrophy and sensory neuropathy: autopsy case report of two brothers, J. Neurol. Sci. 87: 141-152, 1988. Narod, S. A., Dupont, A., Cusan, L., Diamond, P., Gomez, J-L., Suburu, R., and Labrie, F. The impact of family history on early detection of prostate cancer, Nat. Med. 1: 99-101, 1995. Nazareth, L. V., and Weigel. N. L. Activation of the human androgen receptor through a protein kinase A signaling pathway, J. Biol. Chem. 271: 19900-19907, 1996. Neuhausen, S. L., Famham, J. M., Kort, E., Tavtigian, S. V., Skolnick, M. H., and Cannon-Albright, L. A. Prostate cancer susceptibility locus HPC1 in Utah high- risk pedigrees, Hum. Mol. Genet. 8: 2437-2442, 1999. Ohnuki, Y., Mamell, M. M., Babcock, M. S., Lechner, J. F., and Kaighn M. E. Chromosomal analysis of human prostatic adenocarcinoma cell lines, Cancer Res. 40: 524-534, 1980. Onate, S. A., Boonyaratanakomkit, V., Spencer, T. E., Tsai, S. Y., Tsai, M-J., Edwards, D. P., and O'Mally, B. The steroid receptor coactivator-1 contains multiple receptor interacting and activation domains that cooperatively enhance the activation function 1 (AF1) and AF2 domains of steroid receptors, J. Biol. Chem. 273: 12101-12108, 1998. Ouchi, T., Monteiro, A. N. A., August, A., Aaronson, S. A., and Hanafusa, H. BRCA1 regulates p53-dependent gene expression, Proc. Natl. Acad. Sci. USA. 95: 2302- 2306, 1998. 240 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Pao, G. M., Janknecht, R., Ruffner, H., Hunter, T., and Verma, I. M. CBP/p300 interact with and function as transcriptional coactivators of BRCA1, Proc. Natl. Acad. Sci. USA. 97: 1020-1025, 2000. Poujol, N., Lobaccaro, J-M., Chiche, L., Lumbroso, S., and Sultan, C. Functional and structural analysis of R607Q and R608K androgen receptor substitutions associated with male breast cancer, Mol. Cell. Endocrinol. 130: 43-51, 1997. Poulin, R., Simard, J., Labrie, C., Petitclerc, L., Dumont, M., Lagace, L., and Labrie, F. Down-regulation of estrogen receptors by androgens in the ZR-75-1 human breast cancer cell line, Endocrinol. 125: 392-399, 1989. Printen, J. A., and Sprague, G. F. Protein-protein interactions in the yeast pheromone response pathway: Ste5p interacts with all members of the MAP kinase cascade, Genetics. 138: 609-619, 1994. Quigley, C. A., De Beilis, A., Marschke, K. B., El-A wady, M. K., Wilson, E. M., and French, F. S. Androgen receptor defects: historical, clinical, and molecular perspectives, Endocrine Rev. 16: 271-321, 1995. Rahman, N., and Stratton, M. R. The genetics of breast cancer susceptibility, Annu. Rev. Genet. 32: 95-121, 1998. Rao, A., and Cramer, S. D. Identification o f a polymorphism in the ARE 1 region of the PSA promoter, Proc. Am. Assoc. Cancer Res. 40: 65, 1999. Rebbeck, T. R., Kantoff, P. W., Krithivas, K., Neuhausen, S., Blackwood, M. A., Godwin, A. K., Daly, M. B., Narod, S. A., Garber, J. E., Lynch, H. T., Weber, B. L., and Brown, M. Modification o f BRCA1 -associated breast cancer risk by the polymorphic androgen receptor CAG repeat, Am. J. Hum. Genet. 64: 1371-1377, 1999. Reed, J. C. Double identity for proteins of the Bcl-2 family, Nature. 387: 773-776, 1997. Roeder, R. G. The role of general initiation factors in transcription by RNA polymerase II, Trends Biochem. Sci. 21: 327-335, 1996. Rogner, U. C., Wilke, K., Steck, E., Korn, B., and Poustka, A. The melanoma antigen gene (MAGE) family is clustered in the chromosomal band Xq28, Genomics. 29: 725-731, 1995. 241 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Ross, R. K., Pike, M. C., Coetzee, G. A., Reichardt, J. K. V., Yu, M. C., Feigelson, H., Stanczyk, F. Z., Kolonel, L. N., and Henderson, B. E. Androgen metabolism and prostate cancer: establishing a model of genetic susceptibility, Cancer Res. 58: 4497-4504, 1998. Ross, R. K., and Coetzee, G. A. The epidemiology and etiology of prostate cancer. In: Z. Petrovich, Baert, L., and Brady, L. W. (ed.) Carcinoma of the prostate: innovations in management, pp. 1-11. Berlin: Springer-Verlag, 1996. Roy, A. K., Lavrovsky, Y., Song, C. S., Chen, S., Jung, M. H., Velu, N. K., Bi, B. Y., and Chattel]ee, B. Regulation of androgen action, Vitamines and Hormones. 55: 309- 352,1999. Rubinsztein, D. C., Leggo, J., Coetzee, G. A., Irvine, R. A., Buckley, M., and Ferguson- Smith, M. A. Sequence variation and size ranges of CAG repeats in the Machado- Joseph disease, spinocerebellar ataxia type 1 and androgen receptor genes, Hum. Mol. Genet. 4: 1585-1590, 1995a. Rubinsztein, R. C.5 Amos, W., Leggo, J., Goodbum, S., Jain, S., Li, S-H., Margolis, R. L., Ross, C. A., and Ferguson-Smith, M. A. Microsatellite evolution-evidence for directionality and variation in rate between species, Nat. Genet. 337-343, 1995b. Ruizeveld de Winter, J. A., Janssen, P. J. A., Sleddens, H. M. E. B., Verleun-Mooijman, M. C. T., Trapman, J., Brinkmann, A. O., Santerse, A. B., Schroder, F. H., and van der Kwast, T. H. Androgen receptor status in localized and locally progressive hormone refractory human prostate cancer, Am. J. Pathol. 144: 735- 746, 1994. Rush, M. G., Drivas, G., and D'Eustachio, P. The small nuclear GTPase Ran: how much does it run?, BioEssays. 18: 103-112, 1996. Sadar, M. D. 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, 1999. Sadi, M. V., and Barrack, E. R. Image analysis of androgen receptor immunostaining in metastatic prostate cancer, Cancer. 71: 2574-2580, 1993. Saurin, A. J., Borden, K. L. B., Boddy, M. N., and Freemont, P. S. Does this have a familiar RING?, Trends Biochem. Sci. 21: 208-214, 1996. 242 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Schaid, D. J., McDonnell, S. K., Blute, M. L., and Thibodeau, S. N. Evidence for autosomal dominant inheritance of prostate cancer, Am. J. Hum. Genet. 62: 1425- 1438, 1998. Schiestl, R. H., and Gietz, R. D. High efficiency transformation of intact yeast cells using single stranded nucleic acids as carrier, Curr. Genet. 16: 339-346, 1989. Schneikert, J., Peterziel, H., Defossez, P-A., Klocker, H., Launoit, Y., and Cato, A. C. B. Androgen receptor-Ets protein interaction is a novel mechanism for steroid hormone-mediated down-modulation of matrix metalloproteinase expression, J. Biol. Chem. 271: 23907-23913, 1996. Schoenberg, M. P., Hakimi, J. M., Wang, S., Bova, G. S., Epstein, J. I., Fischbeck, K. H., Walsh, P. C., and Barrack, E. R. Microsatellite mutation (CAG24-18) in the androgen receptor gene in human prostate cancer, Biochem. Biophys. Res. Commun. 198: 74-80, 1994. Schroder, F. H., Hermanek, P., Denis, L., Fair, W. R., Gospodarowicz, M. K., and Macaluso, M. The TNM classification of prostate cancer, Prostate. 4 (Suppl): 129- 138, 1992. Scully, R., Chen, J., Ochs, R. L., Keegan, K., Hoekstra, M., Feunteun, J., and Livingston, D. M. Dynamic changes of BRCA1 subnuclear localization and phosphorylation state are initiated by DNA damage, Cell. 90: 425-435, 1997a. Scully, R., Anderson, S. F., Chao, D. M., Wei, W., Ye, L., Young, R. A., Livingston, D. M., and Parvin, J. D. BRCA1 is a component of the RNA polymerase II holoenzyme, Proc. Natl. Acad. Sci. USA. 94: 5605-5610, 1997b. Scully, R., Ganesan, S., Vlasakova, K., Chen, J., Socolovsky, M., and Livingston, D. M. Genetic analysis of BRCA1 function in a defined tumor cell line, Mol. Cell. 4: 1093-1099,1999. Shiau, A. K., Barstad, D., Loria, P. M., Cheng, L., Kushner, P. J., Agard, D. A., and Greene, G. L. The structural basis of estrogen receptor/coactivator recognition and the antagonism o f this interaction by tamoxifen, Cell. 95: 927-937, 1998. Sinclair, C. S., Berry, R., Schaid, D., Thibodeau, S. N., and Couch, F. J. BRCA1 and BRCA2 have a limited role in familial prostate cancer, Cancer Res. 60: 1371- 1375, 2000. 243 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Smith, J. R., Freije, D., Carpten, J. D., Gronberg, H., Xu, J., Isaacs, S. D., Brownstein, M. J., Bova, G. S., Guo, H., Bujnovszky, P., et al. Major susceptility locus for prosate cancer on chromosome I suggested by a genome-wide search, Science. 274: 13371-1374, 1996. Snoek, R., Bruchovsky, N., Kasper, S., Matusik, R. J., Gleave, M., Sato, N., Mawji, N. R., and Rennie, P. S. Differential transactivation by the androgen receptor in prostate cancer cells, Prostate. 26: 256-263, 1998. Sogani, P. C., Israel, A., Lieberman, P. H., Lesser, M. L., and Whitmore, W. F. Gleason grading o f prostate cancer: a predictor of survival, Urol. 25: 223-227, 1985. Spencer, T. E., Jenster, G., Burcin, M. M., Allis, C. D., Zhou, J., Mizzen, C. A., Mckenna, N. J., Onate, S. A., Tsai, S. Y., Tsai, M-J., and O'Malley, B. W. Steroid receptor coactivator-1 is a histone acetyltransferase, Nature. 289: 194-198, 1997. Spitz, M. R., Currier, R. D., Feuger, J. J., Babaian, R. J., and Newell, G. R. Familial patterns o f prostate cancer: a case-control analysis, J. Urol. 146: 1305-1307, 1991. Spurdle, A. B., Dite, G. S., Chen, X., Mayne, C. J., Southey, M. C., Batten, L. E., Chy, H., Trute, L., McCredie, M. R. E., Giles, G. G., Armes, J., Venter, D. J., Hopper, J. L., and Chenevix-Trench, G. Androgen receptor exon 1 CAG repeat length and breast cancer in women before age forty years, J. Natl. Cancer Inst. 91: 961-966, 1999. Stanford, J. L., Just, J.J., Gibbs, M., Wichlund, K. G., Neal, C. L., Blumenstein, B. A., and Ostrander, E. A. Polymorphic repeats in the androgen receptor gene: molecular markers for prostate cancer risk, Cancer Res. 57: 1194-1198, 1997. Starr, R., Willson, T. A., Viney, E. M., Murray, L. J. L., Rayner, J. R., Jenkins, B. J., Gonda, T. J., Alexander, W. S., Metcalf, D., Nicola, N. A., and Hilton, D. J. A family of cytokine-inducible inhibitors of signaling, Nature. 287: 917-921, 1997. Steinberg, G. D., Carter, B. S., Beaty, T. H., Childs, B., and Walsh, P. C. Family history and the risk o f prostate cancer, Prostate. 17: 337-347, 1990. Stenoien, D. L., Cummings, C. J., Adams, H. P., Mancini, M. G., Kavita, P., Demartino, G. N., Marcelli, M., Weigel, N. L., and Mancini, M. A. Polyglutamine-expanded androgen receptors form aggregates that sequester heat shock protein, proteosome components and SRC-1, and are suppressed by the HDJ-2 chaperone, Hum. Mol. Genet. 8: 731-741, 1999. 244 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Stott, K., Blackburn, J. M., Butler, P. J. G., and Perutz, M. Incorporation o f glutamine repeats makes protein oligomerize: implications for neurodegenerative diseases, Proc. Natl. Acad. Sci. USA. 92: 6509-6513, 1995. Szelei, J., Jimenez, J., Soto, A. M., Luizzi, M. F., and Sonnenschein, C. Androgen- induced inhibition o f proliferation in human breast cancer MCF7 cells transfected with androgen receptor, Endocrinol. 138: 1406-1412, 1997. Talmud, P., Tybjaerg-Hansen, A., Bhatnagar, D., Mbewu, A., Miller, J. P., Durrington, P., and Humphries, S. Rapid screening for specific mutations in patients with a clinical diagnosis o f familial hypercholesterolaemia, Atherosclerosis. 89: 137- 141, 1991. Talpin, M.-E., Bubley, G. J., Shuster, T. D., Frantz, M. E., Spooner, A. E., Ogata, G. K., Keer, H., N., and Balk, S. P. Mutation o f the androgen receptor gene in metastatic androgen-independent prostate cancer, N. Engl. J. Med. 332: 1393-1398, 1995. Tilley, W. D., Buchanan, G., Hickey, T. E., and Bentel, J. M. Mutations in the androgen receptor gene are associated with progression of human prostate cancer to androgen independence, Clin. Cancer Res. 2: 277-285, 1996. Tilley, W. D., Marcelli, M., Wilson, J. D., and McPhaul, M. J. Characterization and expression of a cDNA encoding the human androgen receptor, Proc. Natl. Acad. Sci. USA. 86: 327-331, 1989. Torchia, J., Glass, C., and Rosenfeld, M. G. Co-activators and co-repressors in the integration of transcriptional responses, Curr. Opin. Cell Biol. 10: 373-383, 1998. Torchia, J., Rose, D. W., Inostroza, J., Kamei, Y., Westin, S., Glass, C. K., and Rosenfeld, M. G. The transcriptional co-activator p/CIP binds CBP and mediates nuclear-receptor function, Nature. 387: 677-684, 1997. Touhimaa, P. Control of cell proliferation in male accessory sex glands. In: E. Spring- Mills, and Hafez, E. S. E. (ed.) Male accessory sex organs, pp. 131. New York: Elsevier/North Holland, 1980. Touhimaa, P., and Niemi M. Cell renewal and mitogenic activity of testosterone in male sex accessory glands. In: D. Brandes (ed.) Male accessory sex organs: structure and function in mammals, pp. 329. New York: Academic Press, 1974. Truss, M., and Beato, M. Molecular interactions of steroid receptors, Endocrine Rev. 14: 459, 1993. 245 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Tsai, M.-J., Clark, J. H., Schrader, W. T., and O'Malley, B. W. Mechanisms of action of hormones that act as transcription-regulatory factors. In: J. D. Wilson, Foster, D. W., Kronenberg, H. M., and Larsen, P. R. (ed.) Williams textbook of endocrinology, 9th edition, pp. 55-94. Philadelphia: W. B. Saunders Co., 1999. Tsai, Y. C., Simoneau, A. R., Spruck III, C. H., Nichols, P. W., Steven, K., Buckley, J. D., and Jones, P. A. Mosaicism in human epithelium: macroscopic monoclonal patches cover the urothelium, J. Urol. 153: 1697-1700, 1995. Tut, T. G., Ghadessy, F. J., Trifiro, M. A., Pinsky, L., and Yong, E. L. Long polyglutamine tracts in the androgen receptor are associated with reduced trans activation, impaired sperm production, and male infertility, J. Clin. Endocrinol. Metab. 82: 3777-3782,1997. Uchida, T., Wang, C., Sato, T., Gao, J., Takashima, R., Irie, A., Ohori, M. and Koshiba, K. BRCA1 gene mutation and loss of heterozygosity on chromosome 17q21 in primary prostate cancer, Int. J. Cancer (Pred. Oncol.). 84: 19-23, 1999. Van de Voorde, W. M. Pathology of prostatic carcinoma. In: Z. Petrovich, Baert, L., and Brady, L. W. (ed.) Carcinoma of the prostate: innovations in management, pp. 27-50. Berlin: Springer-Verlag, 1996. Webb, P., Nguyen, P., Shinsako, J., Anderson, C., Feng, W., Nguyen, M. P., Chen, D., Huang, S-M., Subramanian, S., McKinemey, E., Katzenellenbogen, B. S., Stallcup, M. R., and Kushner, P. J. Estrogen receptor activation function 1 works by binding pl60 coactivator proteins, Mol. Endocrinol. 12: 1605-1618, 1998. Webb, P., Lopez, G. N., Uht, R. M., and Kushner, P. J. Tamoxifen activation of the estrogen receptor/AP-1 pathway: potential origin for the cell-specific estrogen like effects of antiestrogens, Mol. Endo. 9: 443-456, 1995. Wiener, J. S., Teague, J. L., Roth, D. R., Gonzales Jr., E. T., and Lamb, D. J. Molecular biology and function of the androgen receptor in genital development, J. Urol. 157: 1377-1386, 1997. Wilkens, E. P., Freije, D., Xu, J., Nusskem, D. R., Suzuki, H., Isaacs, S. D., Wiley, K., Bujnovszky, P., Meyers, D. A., Walsh, P. C., and Isaacs, W. B. No evidence for a role of BRCA1 or BRCA2 mutations in Ashkenazi Jewish families with hereditary prostate cancer, Prostate. 39: 280-284, 1999. Wilson, J. D. The promiscuous receptor: prostate cancer comes of age, N. Engl. J. Med. 332: 1440-1441, 1995. 246 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Wooster, R., Mangion, J., Eeles, R., Smith, S., Dowsett, M., Averill, D., Barrett-Lee, P., Easton, D. F., Ponder, B. A. J., and Stratton, M. R. A germline mutation in the androgen receptor gene in two brothers with breast cancer and Reifenstein syndrome, Nat. Genet. 2: 132-134, 1992. Xu, J., Meyers, D., Freije, D., Isaacs, S., Wiley, K., Nusskem, D., Ewing, C., Wilkens, E., Bujnovszky, P., Bova, S., Walsh, P., et al. Evidence for a prostate cancer susceptibility locus on the X chromosome, Nat. Genet. 20: 175-179, 1998. Xu, L., Glass, C., and Rosenfeld, M. G. Coactivator and corepressor complexes in nuclear receptor function, Curr. Opin. Genet. Dev. 9: 140-147, 1999. Xue, W., Irvine, R. A., Ross, R. K., Coetzee, G. A., and Ingles, S. A. Susceptibility to prostate cancer: interaction between genotypes at the androgen receptor and prostate-specific antigen loci, Cancer Res. 60: 839-841, 2000. Yarden, R. I., and Brody, L. C. BRCA1 interacts with components of the histone deacetylase complex, Proc. Natl. Acad. Sci. USA. 96: 4983-4988, 1999. Yeh, S., Lin, H-K., Kang, H-Y., Thin, T. H„ Lin, M. F., and Chang, C. 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, 1999. Yoshida, K. I., Yano, M., Chiba, K., Honda, M., and Kitahara, S. CAG repeat length in the androgen receptor gene is enhanced in patients with idiopathic azoospermia, Urology. 54: 1078-1081, 1999. Yu, M. C., Skipper, P. S., Taghizadeh, K., Tannenbaum, S. R., Chan, K. K., Hendersen, B. E., and Ross, R. K. Acetylator phenotype, amino-hemoglobin adduct levels, and bladder cancer risk in White, Black, and Asian men in Los Angeles, California, J. Natl. Cancer Inst. 86: 712-716, 1994. Zhang, L., Leeflang, E. P., Yu, J., and Amheim, N. Studying human mutations by sperm typing: instability of CAG trinucleotide repeats in the human androgen receptor gene, Nat. Genet. 7; 531-535, 1994. Zhang, M., Magit, D., and Sager, R. Expression of maspin in prostate cells is regulated by a positive Ets element and a negative hormonal response element site recognized by androgen receptor, Proc. Natl. Acad. Sci. USA. 94: 5673-5678, 1997. 247 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Zhu, X., Daffada, A. A. I., Chan, C. M. W., and Dowsett, M. Identification of an exon 3 deletion splice variant androgen receptor mRNA in human breast cancer, Int. J. Cancer. 72: 574-580, 1997. 248 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

The role of tumor-associated and normal human brain endothelial cells in angiogenesis

PDF

Identification of oncogenes cooperating in murine mammary tumorigenesis and transgenic mouse models of breast cancer

PDF

Studies on the regulation of p53 and investigation of functional significance of interactions between p53 and its binding proteins

PDF

Transcriptional regulation mediated by androgen receptor at the prostate specific antigen locus

PDF

Wnt factors: Genes, functions and cooperative signaling in mammary cell transformation

PDF

Studies of hepatitis B virus and hepatitis C virus infection

PDF

Role of PKC-delta in deferoxamine-induced cell death in salivary epithelial Pa-4 cells

PDF

The effects of progesterone on cell cycle and differentiation status of human ovarian cancer cells

PDF

Immunogenicity of acute lymphoblastic leukemia: In vivo and in vitro studies of Bcr -Abl specific immune responses

PDF

The effect of interleukin-7 and C-Kit signaling throughout the ontogeny of naive CD4+ lymphocytes from hematopoietic stem cells to maturity

PDF

The effect of cyclooxygenase inhibitors on cell cycle regulation and proliferation of glioblastoma and lymphoma cell lines

PDF

Leukemia cell vaccine generation using lentiviral vectors

$Molecular and cellular characterization of nickel -induced C3H/10T½ Cl 8 cell transformation$

PDF

Molecular and cellular characterization of nickel -induced C3H/10T½ Cl 8 cell transformation

PDF

Mechanistic study of RAG -mediated initiation of V(D)J recombination

PDF

Rational design of DNA epitope vaccines

PDF

Regulation of RNA polymerase I transcription by CK2 and TAF1

PDF

Functional analysis of MSX2 and its role in skull patterning

PDF

RNA viruses and cell membrane: Hepatitis C virus and coronavirus

PDF

Molecular mechanisms of androgen independence in prostate cancer

PDF

The interactions of host cell proteins with mouse hepatitis virus RNA

Asset Metadata

Creator Irvine, Ryan Andrew (author)

Core Title The androgen receptor: Its role in the development and progression of cancers of the prostate and breast

Degree Doctor of Philosophy

Degree Program Molecular Microbiology and Immunology

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag biology, molecular,health sciences, immunology,health sciences, oncology,OAI-PMH Harvest

Language English

Contributor Digitized by ProQuest (provenance)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-78617

Unique identifier UC11328264

Identifier 3018090.pdf (filename),usctheses-c16-78617 (legacy record id)

Legacy Identifier 3018090.pdf

Dmrecord 78617

Document Type Dissertation

Rights Irvine, Ryan Andrew

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA