University of Southern California Dissertations and Theses

Functional analyses of androgen receptor structure pertaining to prostate cancer

(USC Thesis Other)

Functional analyses of androgen receptor structure pertaining to prostate cancer

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

FUNCTIONAL ANALYSES OF ANDROGEN RECEPTOR STRUCTURE
PERTAINING TO PROSTATE CANCER

by

Howard Chung-Hao Shen

A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(MOLECULAR EPIDEMIOLOGY)

May 2007

Copyright 2007 Howard Chung-Hao Shen
ii
DEDICATION

To Julianne Chun, my wife, for her unfaltering love and support
through all the ups and downs of my graduate studies.

To Wei-Chiang and Daisy Shen, my parents, for supporting me
in the many ways that only parents can.

To Sherlyn and Garrett Goo, my parents in-law, for welcoming me
into their lives and in doing so making mine better.

To Jerry Shen, my brother, for challenging me with the question:
“Why are you getting a Ph.D.?”

And only now, at its completion, do I have an answer:
“Because with it I hope to be more.”

iii
ACKNOWLEDGMENTS

I wish to express gratitude to Dr. Gerhard Coetzee, my graduate advisor and
laboratory head, whose trust in my work has granted me great autonomy throughout
the course of my doctoral research. His faith in my abilities to develop hypotheses,
design experiments, carry out procedures, and interpret results has been pivotal for
shaping my approach to science. It is through the confidence that he has shown in me
that I have been able to build my own self-confidence as a scientist.

Thank you to Dr. Sue Ingles, who has generously given her time and guidance from
day one of my entry into the Molecular Epidemiology Ph.D. program. Her sense of
reason and willingness to hear my opinions about my academic and research paths
has allowed me to recognize my strengths and develop them further.

Thank you to Dr. Michael Stallcup, for sharing his scientific insights and
collaborative spirit. My appreciation of his knowledge is surpassed only by that of
the openness with which he has welcomed me into his laboratory’s meetings week
after week, and though I cannot say from which I have benefited more, they both
equally reflect upon his generosity and stature.

Thank you to my fellow lab mates: Grant Buchanan, Li Jia, Jennifer Prescott,
Omar Khalid, Allison Walters, Michael Henderson, and Marcus Wantroba. They
have each, in their own way, contributed to the laboratory environment in which
I have been able to find success.

And thank you to Dr. Paul Love, Dr. Connie Sommers, and Dr. Alex Grinberg.
Their lasting gift to me is in being greater mentors than they will ever fully realize,
all the while making me think we were just having fun.
iv
TABLE OF CONTENTS

DEDICATION ....................................................................................................... ii

ACKNOWLEDGMENTS .................................................................................... iii

LIST OF FIGURES ............................................................................................. vii

ABSTRACT .......................................................................................................... ix

CHAPTER 1: Clinical overview of prostate cancer................................................. 1
Development and morphology of the prostate................................................... 1
Biology of prostate cancer................................................................................ 3
Epidemiology of prostate cancer in the United States ....................................... 5
Epidemiology of prostate cancer worldwide ..................................................... 7
Risk factors for prostate cancer......................................................................... 8
PSA screening................................................................................................ 11

CHAPTER 2: The role of the androgen receptor in prostate cancer....................... 13
The nuclear receptor superfamily ................................................................... 13
The AR gene and protein structure ................................................................. 15
The AR DNA-binding domain (DBD)............................................................ 19
The AR ligand-binding domain (LBD)........................................................... 25
The canonical androgen signaling axis ........................................................... 29
The dependence of prostate cancer on AR signaling ....................................... 35

CHAPTER 3: Signaling motifs of the AR transactivation domain......................... 39
INTRODUCTION ............................................................................................. 39
Localization of transactivation potential within the AR NTD.......................... 42
The polymorphic CAG and GGC microsatellite repeats.................................. 43
A highly conserved signature sequence .......................................................... 45
The role of phosphorylation in receptor function ............................................ 46
The role of sumoylation at negative regulatory motifs .................................... 48
MATERIALS AND METHODS........................................................................ 49
Site-directed mutagenesis and plasmid preparation......................................... 49
Cell culture and transfection........................................................................... 51
RESULTS.......................................................................................................... 52
Interference assays reveal the two main activation functions in AR NTD ....... 52
Poly-glycine GGC repeat length impacts transactivation ................................ 55
Substitution mutations reveal contributions of various signaling motifs.......... 55

v
DISCUSSION.................................................................................................... 62
Signaling motifs and trinucleotide expansions ................................................ 62
The flexibility of AR NTD signaling is apparent in transient transfection
systems .......................................................................................................... 64

CHAPTER 4: Involvement of p160 coactivators in AR N/C interaction ............... 67
INTRODUCTION ............................................................................................. 67
The AF-2 activation function of steroid receptor LBDs .................................. 71
MATERIALS AND METHODS........................................................................ 72
Preparation of plasmids .................................................................................. 72
Cell culture and transfection........................................................................... 77
Western blotting analysis................................................................................ 78
RESULTS.......................................................................................................... 79
AR transactivation proceeds in the absence of N/C interaction ....................... 79
GRIP1 can restore N/C interaction in the AR FQNAA mutant........................ 79
GRIP1 bridging of AR N/C interaction is dependent
on two coactivator motifs ............................................................................... 87
GRIP1 bridging of AR N/C interaction is not required for coactivation .......... 88
DISCUSSSION.................................................................................................. 92
Variety in the inter-domain interactions of nuclear receptors .......................... 92
GRIP1 acts as an AR coactivator through different mechanisms..................... 93
The paradox of AR FQNLF competition with GRIP1 for binding of AF-2...... 97
The mechanistic significance of AR N/C interaction....................................... 98

CHAPTER 5: Recruitment of the 26S proteasome by the AR Hinge domain ...... 100
INTRODUCTION ........................................................................................... 100
The nuclear localization signal ..................................................................... 100
The Hinge contains binding sites for AR cofactors ....................................... 101
AR is targeted for degradation via ubiquitination.......................................... 102
The role of the 26S proteasome in AR regulation ......................................... 103
The AR Hinge contains a putative PEST sequence ....................................... 110
MATERIALS AND METHODS...................................................................... 112
Site-directed mutagenesis and plasmid preparation....................................... 112
Cell culture and transfection......................................................................... 115
Chromatin Immunoprecipitation (ChIP) assay .............................................. 115
RESULTS........................................................................................................ 120
The AR Hinge inhibits N/C domain interactions........................................... 120
Inhibition of proteasome function results in loss of AR transactivation......... 126
Deletions of the AR Hinge radically alter receptor activity ........................... 129
Proteasome recruitment to the PSA gene is DHT responsive ........................ 133
DISCUSSSION................................................................................................ 136
Proteasome inhibition by MG132................................................................. 136
The Hinge as a split communication line to the proteasome.......................... 137

vi
RESEARCH SUMMARY ................................................................................. 142

REFERENCES .................................................................................................. 145
vii
LIST OF FIGURES

2.1 Schematic map of human X-chromosome…………………………. 18

2.2 Amino acid residues of the AR DBD……………………………… 24

2.3 5α-dihydrotestosterone (DHT)…………………………………….. 28

2.4 The canonical androgen signaling axis…………………………….. 32

2.5 Nuclear receptor-mediated transcription on chromatin……………. 34

3.1 Schematic domain maps of several Type I and Type II
hormone receptors…………………………………………………. 41

3.2 Interference assays reveal transactivation potential of AR NTD….. 54

3.3 Effect of alterations in poly-glycine GGC repeat size
on AR transactivation……………………………………………… 57

3.4 Effect of AR NTD signature sequence (ANTS) deletion
on receptor transactivation………………………………………… 59

3.5 Effect of amino acid substitutions at various AR NTD
signaling motifs on transactivation………………………………… 61

4.1 Domain map of GRIP1 constructs used in AR N/C
interaction assays…………………………………………………... 70

4.2 Overview of the mammalian two-hybrid assay system……………. 76

4.3 Disparity between the processes of AR N/C interaction
and receptor transactivation………………………………………... 82

4.4 Mammalian two-hybrid assay results from AR N/C interaction
scenarios…………………………………………………………… 84

4.5 Mammalian two-hybrid assay results indicating that relative
interaction levels between GRIP1 and AR NTD FQNLF versus
AR NTD FQNAA are identical……………………………………. 86

viii
4.6 Comparison of the transactivation ability of full-length
AR FQNLF versus AR FQNAA in the presence of GRIP1
constructs………………………………………………………….. 91

4.7 Hypothetical models of AR N/C interaction and GRIP1
bridging of AR domains as observed using the mammalian
two-hybrid assay system…………………………………………… 96

5.1 Schematic model of the 26S proteasome…………………………... 109

5.2 The AR Hinge domain……………………………………………... 114

5.3 Overview of steps involved in the ChIP assay……………………... 119

5.4 Mammalian two-hybrid data demonstrating impact of
AR Hinge on N/C domain interactions…………………………….. 123

5.5 Presence of the AR Hinge impacts AF-2 transactivation………….. 125

5.6 AR transactivation on various promoters is dependent on
proteasome function……………………………………………….. 128

5.7 Transactivation of AR Hinge-deletion constructs…………………. 132

5.8 ChIP data revealing occupancy of AR and proteasome on
the endogenous PSA gene locus…………………………………… 135

ix
ABSTRACT

Various peptide regions of the human androgen receptor (AR) were characterized for
their potential role in mediating the androgenic signaling response that is integrally
linked to prostate cancer development and resistance to treatment. A series of site-
directed mutant AR constructs was created that targeted putative signaling motifs
within the molecule, and was utilized in functional assays to determine how
deviations from the wild-type AR sequence at these sites impacted receptor activity.
It was shown that the AR amino-terminal transactivation domain (NTD) possesses
great compensatory ability, as independent disruption of several sites had little
impact on receptor function in prostate-cancer derived cell lines. Size modulations of
the glycine trinucleotide repeat in the NTD had a direct effect on receptor signaling,
suggestive of this region having important influence on NTD structure. It was
determined that a motif located in the proximal region of the NTD contributes
greatly to overall receptor function through mediating ligand-dependent interactions
between the AR NTD and ligand-binding domains (LBD). In addition, it was
revealed that one mechanism by which the p160 nuclear receptor coactivators
enhance AR function is through enhancement of this inter-domain communication.
In related studies it was shown that the AR Hinge domain imparts a pivotal
contribution to normal AR signaling through mediating communication between AR
and the 26S proteasome. Dual roles of the proteasome in AR signaling were
characterized based on observations that separate signaling motifs within the
x
AR Hinge control a balance between the apo-receptor and holo-receptor responses to
proteasome influence. These findings yield insight into how specific structural
components of the AR cooperatively function within the entirety of the actively
signaling receptor, and will contribute to new strategies for therapeutically targeting
the AR molecule in the context of the aberrant androgen signaling axis that is evident
in prostate cancer.
1
CHAPTER 1
Clinical overview of prostate cancer

Development and morphology of the prostate

The human prostate is a walnut-sized organ located immediately below the bladder
in men where it surrounds approximately 40 mm of the urethra exiting the internal
sphincter (Van de Voorde, 1996). Prostatic secretions are a major component of
semen, and they function both to neutralize the pH of the urethra and as an aid to
sperm motility via the action of proteolytic enzymes.
In the developing fetus the prostate begins as epithelial outgrowths (prostatic
buds) emerging from the endodermal urogenital sinus and extending into the
surrounding mesenchymal tissue (Cunha, Donjacour et al., 1987). These growths
undergo two distinct phases of growth and differentiation during which the complex
network of secretory ducts of the adult prostate are formed. The first phase of ductal
morphogenesis is prenatal, with the second phase taking place during puberty. Upon
reaching adulthood the prostate no longer grows as a consequence of very low
epithelial cell turnover (Senius, Pieltila et al., 1974).
Androgens are the male steroid hormones that drive prostate development.
This group includes molecules such as testosterone (T), dihydrotestosterone (DHT),
dehydroepiandrostenone (DHEA), and androstenedione. These hormones play key
signaling roles in a diversity of physiological systems including the immune system,
neuronal/brain development, maintenance of skeletal muscle, deposition of adipose
2
tissue, hair growth/loss, and more obviously in the development of the male
reproductive system and of secondary traits associated with the male phenotype.
Testosterone (T) is the major androgen in men, and in adults nearly 95% of the
circulating T is produced from the testis with the remaining 5% originating from the
adrenal glands (Coffey, 1992). With respect to early prostate development,
T produced by Leydig cells in the testis of the fetus at week eight of gestation begins
to promote morphogenesis of the epididymis, vas deferens, and seminal vesicles
from the Wolffian duct precursor (Wiener, Teague et al., 1997). The immature
prostate that arises during this prenatal development phase retains its rudimentary
form until the onset of puberty, during which the second phase of organ growth is
heralded by a surge in T production that results in marked enlargement and maturity
of the tissues (Wiener, Teague et al., 1997). Once mature, maintenance of the
prostate retains dependence on androgens for both morphology and function.
Removal of the androgen signal results in degeneration of the organ, as evidenced by
castrated men who experience involution of the prostate as a result of dramatic
epithelial and stromal cell apoptosis (Lindzey, Kumar et al., 1994).
The mature prostate is comprised of three roughly concentric glandular
regions extending outward from the urethral wall; these zones are designated as
central, transitional, and peripheral, and are distinct in their histology and function.
The zones share the common feature that their ducts and acini are lined with
secretory epithelium, below which is a layer of basal and endocrine-paracrine cells
(McNeal, 1988). The central zone comprises 25% of the prostate with its ductal
structure arising close to the ejaculatory duct openings, and is the zone that is
3
typically resistant to disease (McNeal, 1981). The transition zone is the exclusive site
of origin for benign prostatic hyperplasia (BPH), a common non-life-threatening
condition. Lastly, 70% of the prostate is composed of the peripheral zone; it is in this
zone where prostate cancer primarily occurs.

Biology of prostate cancer

Cancer of the prostate is a disease marked by deregulation of cellular proliferation
that leads to defects in ductal morphogenesis and aberrant enlargement of the organ
(Cunha, Donjacour et al., 1987). Abnormal, yet benign, congenital variations of
structures in the prostate are frequent, as ultrasound anatomical studies have
suggested that as high as 18% of men exhibit some improper ductal morphogenesis
in their otherwise healthy prostate (Villers, Terris et al., 1990). BPH is an example
of a common condition that often is confused with prostate cancer upon initial
discovery during examination. Other common conditions that lead to misdiagnosis
include post-inflammatory atrophy, basal cell hyperplasia, and duct-acinar dysplasia,
any of which may be confused with carcinoma in biopsy material (McNeal, 1988).
With regard to cases of true cancer, approximately 70% of prostate
carcinomas develop in the peripheral zone of the organ, with the remaining cases
typically involving disease of the transitional zone (Van de Voorde, 1996). Nearly
98% of prostate cancers are adenocarcinomas that exhibit identifying histological
features such as abolishment of the distinct basal cell layer, improper invasion of
4
stroma by malformed ductal epithelium, and nuclear anaplasia (Van de Voorde,
1996).
Aberrant proliferation of prostate epithelia, whether as a consequence of BPH
or of carcinoma, is marked by increases in the organ’s production of prostate specific
antigen (PSA), a normal component of prostatic secretions that functions as a
proteolytic enzyme in conditioning the seminal fluid. PSA can be detected
circulating in the blood and has thus become a convenient and widely accepted
biomarker for prostate cancer.
Upon diagnosis of prostate cancer, assessment is made according to
histological grade and clinical staging of the tumor. The grading system devised by
Gleason in the early 1970s is still used in common practice to this day (Gleason and
Mellinger, 1974). A Gleason grade from 1-5 is assigned to tissue to describe the
extent of glandular differentiation, with higher values indicating progressively less
differentiation and thus evidence of more aggressive disease. Typically two locations
on the same tissue slide are graded in this manner and the combined values yield a
Gleason score ranging from 2-10. The Gleason score of 7 is an important cutoff that
clinically translates to a far worsened prognosis. The popularity of this grading
system has led to Gleason scoring as the clinical gold standard for prognosis.

5
Epidemiology of prostate cancer in the United States

Carcinoma of the prostate represents a major health problem in the United States,
ranking as the most common solid tumor type reported in men. Data from the
National Cancer Institute SEER registry program has indicated that in the year 2003
men experienced nearly twice as many incident cases of prostate cancer as compared
to carcinomas of the lung. The compiled data from 2005 suggests that of the 700,000
incident cases of all types of cancer reported in men, nearly one-third, or
approximately 230,000 cases, were prostate cancer. These figures place the US rates
among the top worldwide and are particularly bolstered by African-Americans and
US Caucasians having the highest rates of any men globally. In fact, the American
Cancer Society (ACS) has reported that from 1998-2000, the likelihood in the US of
a man developing prostate cancer was 17%, or in other words, 1 in every 6 men
(ACS 2004). Overall, incidence rates in the US have increased steadily since the
early 1970s with a large spike peaking in 1992 that coincided with the
implementation of widespread PSA screening in 1988-89 (Crawford, 2003).
Following this peak, incidence rates decreased at a rate of 10.8% per year, eventually
returning to pre-PSA screening levels by the latter half of the 1990s (Hankey, Feuer
et al., 1999; Parkin, 2001; Quinn and Babb, 2002; Schaid, 2004). This well-
documented rise and fall has been attributed to the screening bias introduced by the
proportion of men receiving the PSA test for the first time (Legler, Feuer et al.,
1998). Unfortunately the subsequent return to the trends prior to the introduction of
PSA testing reveals incidence rates that are still steadily rising for high-risk
6
Americans. In part this may be attributable to the increased life expectancy of
American men. Another noteworthy observation is that the overall increasing
incidence rate is not reflected in the relatively low mortality rates that accompany
this disease. Carcinoma of the prostate accounts for only 10% of the cancer deaths in
US men, with mortality rates actually declining between 1993 and 1997 and
subsequently leveling off (Hankey, Feuer et al., 1999). This disparity between
incidence and mortality figures likely results from the typical profile of prostate
cancer having a long natural history of disease progression, which, when combined
with the effectiveness of PSA screening as an alert for cancer progression following
initial treatment, allows interventions such as androgen ablation therapies a greater
window of opportunity for prolonging patient survival. Alternatively, clinically non-
relevant prostate cancers may be over-diagnosed by PSA screening. Knowledge
gained from prospective randomized studies has led to the adoption of earlier
hormone ablation therapy as an accompaniment to surgical or radiation treatment,
which has resulted in better patient outcome compared against the latter treatments
alone (Messing, Manola et al., 1999).
Incidence and mortality rates vary among racial groups in the US (Haas and
Sakr, 1997; Parkin, 2001; Chu, Tarone et al., 2003; Crawford, 2003; Cunningham,
Ashton et al., 2003; Gronberg, 2003). According to ACS 2004 reports, age-adjusted
incidence rates from 1996-2000 were highest for African-American men (272.1 per
100,000) followed by Whites (164.3 per 100,000), then Hispanics (137.2 per
100,000), Asians/Pacific Islanders (100.0 per 100,000), and finally American
Indian/Native Alaskans (53.6 per 100,000). It is striking that whereas the
7
discrepancy between the highest US rates in African-Americans versus the lowest in
Asians is roughly four fold, mortality in the former group is nearly six fold higher
(ACS 2004).

Epidemiology of prostate cancer worldwide

Prostate cancer incidence rates are on the rise globally, with the disease ranking as
the third most common cancer diagnosed in men in the year 2000 (Hsing, Tsao et al.,
2000; Parkin, 2001; Quinn and Babb, 2002). Overall, international statistics reveal
that prostate cancer is the sixth most common form of any cancer (Gronberg, 2003;
Deutsch, Maggiorella et al., 2004). With the increase in the worldwide incidence rate
steady at 1.7% per year since the mid 1980s, there were approximately 1.5 million
prevalent cases of prostate cancer at the beginning of the current century (Parkin,
2001). However, the steady worldwide increase rate belies the fact that there is a
substantial range of incidence rates across geographic locations. High-risk countries
such as the US and Canada continue to report higher case numbers in comparison to
low-risk Asian countries (Hsing, Tsao et al., 2000). From 1988-1992, the US led the
world in incident cases with Canada, Sweden, Australia, and France rounding out the
top five. Even so, rates in the US were far ahead of the next nation in line; it has been
reported that the age-standardized incidence rates of US men prior to 1992 were
more than double that of Sweden and Australia, more than triple that of Europe, and
ten times those reported by Asian countries such as Japan, India, and China (Quinn
and Babb, 2002). In fact, China reported the lowest prostate cancer rates of any
8
nation during that time period (Crawford, 2003; Gronberg, 2003). However, it
remains unclear to what extent disparities in numbers reported among the countries
are due to true geographically associated risk factors as opposed to differences in
public health policy. The use and reliability of cancer registries varies from nation to
nation, as does the awareness of the general public and their access to health care
required for diagnosis and reporting. Indeed, insufficient prostate cancer screening
has been cited as a potential contributing cause of unreliable numbers for developing
countries (Hsing, Tsao et al., 2000).
There is great debate over the underlying causes for such wide discrepancies
in reported incidence rates worldwide. Variations in the influences of external risk
factors, genetic factors, and accessibility to health care, all related to cultural
diversity, have been theorized by some to be primary causes (Gronberg, 2003).
Changes in cultural practices brought on by modernization have been cited by others.
For example, it has been suggested that for Asian countries such as China and Japan,
while improvements in detection and diagnostic procedures may now be contributing
to higher incidence reporting, so might the increasing adoption of more western
habits such as diets that are higher in fat and meat (Parkin, 2001).

Risk factors for prostate cancer

As with most cancers a man’s age presents the greatest risk factor for developing
carcinoma of the prostate. Studies have suggested that as high as 10% of men in their
early 20s may already have cancer cells in their prostate, and this grouping increases
9
through adulthood to include nearly 50% of men by age 80 (Sakr, Haas et al., 1993).
Prostate cancer is rarely diagnosed in men younger than age 50, yet subsequently it is
the cancer with the fastest increasing risk associated with age (Haas and Sakr, 1997).
The general public may regard the saying “If a man lives long enough he’ll get
prostate cancer” as more adage than certainty, but such data reveal a solid foundation
for this claim. The disparity in incidence and mortality rates among the racial groups
in the US has led to several theories suggesting factors such as genetic susceptibility,
androgen metabolism, and ethnic dietary differences as having underlying
associations with disease (Shibata and Whittemore, 1997; Parkin, 2001). When
prostate cancer mortality rates are examined against dietary data across different
countries, a highly suggestive trend emerges that links increased dietary fat
consumption with elevated mortality (Carroll and Khor, 1975). Conversely it has
been suggested that dietary factors such as the consumption of tomatoes, broccoli,
and vitamin D may provide some protective effects, as does apparently having higher
levels of selenium (Haas and Sakr, 1997). As additional support to the role of diet as
a risk factor, studies have shown that people who immigrate develop incidence and
mortality rates approaching those of their new host nation (Shibata and Whittemore,
1997; Parkin, 2001; Crawford, 2003).
Other reported risk factors for prostate cancer include tobacco smoking
(Haas and Sakr, 1997; Hickey, Do et al., 2001; Roberts, Platz et al., 2003; Oefelein
and Resnick, 2004), alcohol (Dennis and Hayes, 2001; Platz, Leitzmann et al., 2004;
Schoonen, Salinas et al., 2005), sexual behavior (Rosenblatt, Wicklund et al., 2001;
10
Giles, Severi et al., 2003; Hsing and Chokkalingam, 2006), and benign prostatic
hyperplasia (Haas and Sakr, 1997; Guess, 2001; Hammarsten and Hogstedt, 2002).
Family history of the disease may play a role in predicting risk (Domitrz,
Jedrzejowska et al., 2001; Gelmann, 2002; Crawford, 2003; Schaid, 2004; Hsing and
Chokkalingam, 2006), and genetic variations in a few candidate genes have
received scrutiny; these include polymorphisms of the androgen receptor and PSA
genes (Irvine, Yu et al., 1995; Shibata and Whittemore, 1997; Gronberg, 2003). The
polymorphic CAG trinucleotide repeat region of the androgen receptor gene has been
strongly implicated in epidemiologic studies as a potential contributor to variations
in prostate cancer risk among racial groups (Irvine, Yu et al., 1995; Buchanan, Irvine
et al., 2001). In a modestly sized study conducted by our research group the
prevalence of short CAG alleles (< 22 repeats) was highest (75%) in African-
American males with the highest risk for prostate cancer, intermediate (62%) in
intermediate-risk non-Hispanic whites, and lowest (49%) in Asians at very low risk
for prostate cancer (Irvine, Yu et al., 1995). Whereas there is substantial data to
support the idea that trinucleotide repeat sizes in genes may play an important role in
explaining racial differences in prostate cancer risk, there remains no consensus as to
which genes could serve as clinically useful biomarkers. Currently some promising
genetic markers include PCA3, hypermethylation of select genes, and the recently
described phenomenon of gene translocations/fusions between the androgen-
responsive TMPRSS2 gene with neighboring chromosome 21 genes ERG and
ERTV1 belonging to the Ets family of transcription factors (Tomlins, Rhodes et al.,
2005). However, until the efficacy of using specific genes as markers is proven the
11
use of PSA screening remains the most widely accepted clinical biomarker for
prostate malignancy. As it stands currently, the ACS considers only family history,
race/ethnicity, and age as the three confirmed risk factors for prostate cancer.

PSA screening

The US Food and Drug Administration (FDA) approved the PSA test in 1986 as a
prognostic tool for monitoring disease progression in prostate cancer patients
(Hankey, Feuer et al., 1999; Schaid, 2004). At the time, the two main diagnostic
tools for prostate cancer consisted of either transrectal ultrasound (TRUS) or needle
biopsy, neither of which were highly efficacious (Potosky, Miller et al., 1995). In
1988-89 medical professionals began implementing the PSA test for use in
diagnostic screening.
The ACS recommends that annual screening for prostate cancer begin at age
50 for men who have an expected life span greater than ten years. Such screening
consists of receiving a digital rectal exam in conjunction with PSA testing (ACS
2004). The range of values for PSA monitoring in the blood typically extends from a
low reading of 2.5 ng/ml to significantly high levels of 20 ng/ml and greater.
Historically the value 4.0 ng/ml has been considered the cutoff above which a man
would undergo additional testing such as needle biopsy. Using this guideline, the
PSA test has a sensitivity ranging approximately 67.5-80% (Carroll, Coley et al.,
2001). In recent years there has been a push to lower this cutoff in the hopes of
increasing test sensitivity. Even though there have been suggestions that test results
12
in the lower ranges from 2.0 ng/ml upward may still be indicative of prostate cancer,
the reduction in test specificity resulting from lowering the cutoff to that level may
be unacceptable (Gann, Hennekens et al., 1995). Consequently it has been proposed
that the cutoff be set at 2.6 ng/ml as a balance between increasing test sensitivity and
minimizing loss of specificity (Punglia, D'Amico et al., 2003). The utility of using
PSA as a biomarker is far from perfect, as BPH and other non-life-threatening
conditions of the organ may also result in elevated PSA production. But as a
prognostic tool following identification of cancer, PSA monitoring is highly
effective. It has been proposed that pre-therapy PSA levels can serve as a good
prognostic tool for determining course of treatment, as low PSA levels may indicate
a cancer that is less likely to recur whereas higher PSA levels may warrant a more
aggressive treatment plan. It is also accepted that increases in PSA levels following
definitive prostate cancer treatment is a sure sign of disease progression, and the rate
at which the levels rise has a significant impact on patient prognosis. For example, a
PSA rise of 0.4 ng/ml within three years of surgery is accompanied by a significantly
higher mortality risk than such a rise taking place outside of that span (Hiramatsu,
Maehara et al., 1997).
13
CHAPTER 2
The role of the androgen receptor in prostate cancer

The nuclear receptor superfamily

Nuclear receptors (NR) are members of a large gene superfamily that have evolved
as transcription factors that regulate a broad range of fundamental metazoan life
processes. Type I NRs include receptors for androgens (AR), progestins (PR),
estrogens (ER), glucocorticoids (GR), and mineralocorticoids (MR). In their inactive
states these Type I receptors are typically sequestered by chaperone molecules such
as heat shock proteins that prevent the NR from influencing basal promoter
transcription. Upon ligand-induced activation these receptors form homodimers that
subsequently bind palindromic DNA response elements (Dennis and O'Malley,
2005). Type II NRs include receptors for thyroid hormone (TR), vitamin D (VDR),
and retinoids (RAR and RXR). These receptors bind to head-to-tail oriented DNA
response elements and do so as heterodimers with RXR. Type II receptors are found
constitutively bound to DNA in the absence of ligand, where they exert inhibition of
basal transcription through associations with corepressors. Upon ligand binding,
Type II receptors undergo an exchange of corepressors/coactivators that then favors
gene transcription. Lastly, a third category of NRs consists of the so-called orphan
receptors for which cognate ligands have yet to be identified.
Upon binding of DNA, the NRs exert control over gene expression by
modulating the recruitment of the general transcription factors (GTF) and
14
RNA polymerase II (RNA Pol II) to promoter regions. Interactions between NRs and
transcription factors such as coactivators and corepressors can either be direct or via
bridging through intermediate proteins. Ultimately the actions of the NRs affect the
assembly of TATA-binding protein (TBP) and TBP-associated factors (TAF) to form
the TFIID complex, which then organizes TFIIB, TFIIF, and RNA Pol II into the
pre-initiation complex (PIC).
Very early on in evolutionary history the primitive NRs most likely
functioned as monomers that could interact with the basic components of the
transcription machinery without the requirement of binding a cognate ligand. As
members of the NR superfamily evolved to make use of ligand signaling and
undergo homo- and heterodimerization as part of their function, the roles of these
receptors became more complex. This added complexity allowed for the evolution of
higher organisms by providing a highly sophisticated intersecting network of gene
regulation that ultimately influences every stage of life from development and
maturation to maintenance of physiological systems throughout adulthood (Owen
and Zelent, 2000).
In keeping with the common structural features of most NRs the two
absolutely fundamental domains of the Type I steroid receptors are the DNA-binding
domain (DBD) and C-terminal ligand-binding domain (LBD). The DBDs of different
steroid receptors are very similar in their amino acid sequences, a fact most likely
resulting from constraints placed on the structure and DNA binding function of this
peptide domain through evolution (Schwabe and Teichmann, 2004). In contrast,
LBDs among these receptors show greater diversity.
15
It is believed that the ER was the evolutionary precursor to all other steroid
receptors, as primitive orthologs of the gene have been identified in ancient
invertebrates such as mollusks (Thornton, Need et al., 2003). In contrast, none of the
other adrenal and sex steroid receptors have been found in any of the sequenced
genomes of invertebrates to date. Over time genomic amplification of this ancestral
ER probably gave rise to the other steroid receptors, and as the role of these
molecules became more complex in living organisms so did their distinguishing
structural features such as the expansion of their N-terminal transactivation domains
(NTD) (Owen and Zelent, 2000). It has been theorized that the adoption of ligand-
based regulation of the NRs and the subsequent diversification of the steroid receptor
line in particular may represent pivotal events in evolution that allowed for the
development of complex transcriptional networks as seen in higher vertebrates
(Baker, 2003).

The AR gene and protein structure

The human AR is a single-copy gene located on the q arm of the X-chromosome at
position Xq11.2-q12 (Brown, Goss et al., 1989). The genomic location of the AR
gene is highly conserved through mammals, and it has been proposed that this
suggests a potentially significant cooperation of the AR with other syntenic genes
during development (Spencer, Watson et al., 1991).
The AR gene is oriented with its 5’ end centromeric and consists of 8 exons
spanning approximately 80 kb of the chromosome (Chang, Kokontis et al., 1988;
16
Lubahn, Brown et al., 1989; Tilley, Marcelli et al., 1989)(Fig. 2.1). Exon 1
encompasses a large 5’ UTR (~1.1 kb) and codes for the entire NTD of the AR.
Exons 2 and 3 encode the DBD, and exons 4-8 the Hinge and LBD domains. Exon 8
additionally encompasses a large 3’ UTR (~7 kb). The exons together code for a
~2757 bp open reading frame within a 10.6 kb mRNA (Chang, Kokontis et al., 1988;
Kuiper, Faber et al., 1989; Lubahn, Brown et al., 1989; Tilley, Marcelli et al., 1989).
The AR DBD is the most highly conserved of the domains through vertebrate
evolution, followed by the LBD and Hinge (Thornton and Kelley, 1998). In stark
contrast, the AR NTD, which has expanded through evolution, displays conservation
only in two relatively short regions at amino acids 1-53 and at 360-429 (Choong,
Kemppainen et al., 1998).
An overview of the molecular biology of the AR is presented by an often
cited review by Dr. Edward Gelmann (Gelmann, 2002). The mature AR molecule is
a 919 amino acid (aa) protein that consists of the NTD from aa 1-538, the DBD from
aa 539-627, the Hinge from aa 628-658, and the LBD from aa 659-919. These
domains contain functional elements that contribute in a modular fashion to the
overall activity of the receptor (Jenster, van der Korput et al., 1992).
17
Figure 2.1 LEGEND

A) Schematic map of human X-chromosome showing the position of the
androgen receptor gene locus on the q-arm at Xq11.2 – Xq12. The gene is
oriented with a centromeric 5’ end and encompasses approximately 80 kb.
B) Map of the AR gene depicting the 8 exons and associated introns. Exon 1
contains an approximately 1.1 kb 5’ untranslated region (UTR) and encodes
the entire NTD. Exons 2 and 3 encode the separate Zn fingers of the DBD.
Exons 4-8 encode the Hinge and LBD. Exon 8 also contains an
approximately 7 kb 3’ UTR. TIS = Transcription initiation site.
C) Diagram of the 4 main functional domains of the mature AR molecule
with amino acid boundaries numerically designated.

18

19
The AR DNA-binding domain (DBD)

As the DBD is the region of highest conservation across not only the AR of different
species but also across the steroid receptors as a whole, its structure is perhaps the
best understood of the different domains (Fig 2.2). The human AR DBD is 79%
identical to that of PR, 76% identical to GR, and 56% identical to ER. Furthermore,
human AR DBD is 100% identical to that found in rat (Gelmann, 2002). The role of
the DBD is two-fold. First it targets and anchors the AR to appropriate binding loci
in the genome, and second it mediates dimerization of the receptor to facilitate
downstream signaling. One might presume that as transcription factors NRs are
solely recruited to locations in the genome where active assembly of basal
transcription machinery is taking place, i.e. the promoters of regulated genes. This is
proving not to be the case. Recent work has revealed that there are a substantial
number of ER binding sites throughout the human genome that are not restricted to
canonical upstream promoter regions of genes (Carroll, Liu et al., 2005).
The classic model of steroid NR binding to DNA involves the presence of
short nucleotide sequences being recognized by the DBD as direct anchor points for
the receptor (Yamamoto, 1985). These hormone response elements (HRE) are
composed of short hexameric nucleotide repeats that are found in pairs separated by
a short span of spacer nucleotides (Evans, 1988). The orientation of the two
hexameric half-sites can be either directly repeating (head-to-tail) as is the case for
sequences recognized by the Type II receptors, or as an inverted repeat (head-to-
head) as for the Type I steroid receptors. The AR, GR, MR, and PR together
20
recognize repeats that involve a consensus 5’TGTTCT3’ half-site. On the other hand,
ER binds to a consensus 5’TGACCT3’ half-site response element that is much more
reminiscent of those utilized by TR and RXR (Beato, 1989; Zilliacus, Wright et al.,
1995). The critical bases in the arrangement are the guanines (G) and cytosines (C)
located at positions 2 and 5 of each half-site; these represent the specific nucleotides
that are contacted by residues of the receptor DBD. The ability of the different
steroid receptors to discriminate among the HREs found in genes and only affect
appropriate targets is clearly crucial in vivo. It is particularly striking then that these
receptors for the most part target the same consensus HRE signal and utilize such
highly homologous DBDs. Evidence suggests that specificity is contributed by
variations in the non 2 and 5 conserved positions of the consensus HRE
(Schoenmakers, Verrijdt et al., 2000; Verrijdt, Schoenmakers et al., 2000). For
example, it has been shown that the presence of a thymine at nucleotide position -4
of the HRE in several androgen responsive genes determines AR, but not GR,
recognition. Mutations of this thymine to alanine allow these genes to be regulated
by GR as well (Verrijdt, Schoenmakers et al., 2000). These variations in HREs may
be recognized by residues in the C-terminal extension (CTE) of the DBD that differ
among the steroid receptors (Melvin, Roemer et al., 2002). Additional selectivity
may be a function of amino acids proximally located in the DBD itself; this is the
case with differing DBD residues between ER and GR resulting in different DNA
recognition (Zilliacus, Dahlman-Wright et al., 1991). Frequency and spatial
arrangement of tandem HREs in DNA may also aid in amplifying responses from
specific receptors. This is observed in the enhancer regions of some AR-target genes
21
such as PSA where the clustering of multiple AREs results in enhanced AR binding
and control (Grad, Dai et al., 1999).
In recent years non-classical models of AR function have developed that
suggest AR may be recruited to target genes via protein-protein interactions with
other transcription factors that themselves are already bound to DNA, thus bypassing
the need for AREs at the site. Alternatively the AR may function in a completely
non-genomic manner, imparting functional control over cellular processes through
interactions with other signaling molecules in the cytoplasm of the cell outside the
nucleus entirely (Baron, Manin et al., 2004; Bonaccorsi, Marchiani et al., 2006).
Whereas it is certain that the roles of these two non-classical mechanisms of AR
signaling will become better appreciated over time, currently the understanding is
that for most androgen-responsive genes that implement a high level of AR
regulation the receptor functions through direct DNA binding at AREs.
Interaction with cognate DNA by AR is accomplished by two zinc finger-like
motifs in the DBD, one coded by exon 2 and the other by exon 3 of the AR gene.
Each of these zinc fingers has 4 cysteines positioned to articulate a single Zn
2+
ion
per finger and in doing so project a short peptide loop that communicates with the
major groove of the DNA alpha helix (Hard, Kellenbach et al., 1990). Direct contact
with DNA is mediated by residues in a short stretch at the base of the first zinc finger
known as the proximal-box, or P-box (Nguyen, Steinberg et al., 2001). The P-box
residues directly bind the G and C nucleotides at positions 2 and 5 of the ARE half-
sites. The second zinc finger contains a short stretch of residues known as the
distal-box, or D-box, that is conserved in the DBDs of other steroid receptors and
22
serves as an interaction surface for dimerization (Dahlman-Wright, Wright et al.,
1991; Giwercman, Ivarsson et al., 2004). The full significance of AR
homodimerization remains to be elucidated, but the varying orientations of ARE
half-sites in different androgen-responsive promoters may induce dimers with
different AR orientations, thus encouraging alternate structures of the receptor
dependent on promoter context (Aumais, Lee et al., 1996).

23
Figure 2.2 LEGEND

A) Amino acid residues of the AR DBD arranged as two zinc fingers.
Outlined in black are the P-box involved in directly contacting DNA and
D-box involved in receptor dimerization. The C-terminal extension
responsible for ARE specificity and the proximal portion of the Hinge
domain are indicated by gray arrows. NLS = nuclear localization signal.
B) Consensus hormone response elements (HRE) recognized by the Type I
and Type II hormone receptors. Orientation of each half-site is indicated by
gray arrows.
C) Examples of endogenous androgen responsive elements (ARE).

24

25
The AR ligand-binding domain (LBD)

The three-dimensional structure of the AR ligand binding domain has been
determined through crystallography (Sack, Kish et al., 2001)(Fig 2.3). Whereas
significant differences exist in the primary amino acid sequence of the LBDs among
steroid receptors (in some cases demonstrating as little as 20% sequence homology),
the overall tertiary structures of the domains are remarkably similar (Tanenbaum,
Wang et al., 1998). The steroid receptor LBDs fold into 12 helices that form a
two-part structure, with one half being rather rigid in conformation and the other
being more flexible. It is the latter more flexible half that defines the ligand binding
pocket. In the AR this pocket is primarily formed by helices 3, 5 and 10 (Sack, Kish
et al., 2001). It is often noted that AR lacks a readily distinguishable helix 2, but to
preserve consistent nomenclature with the LBDs of other steroid receptors, the
remaining helices retain their homologous numbering. Interestingly the flexibility of
the helices surrounding the ligand binding pocket suggests that throughout evolution
amino acid substitutions within them were tolerated without major impact on overall
domain structure (Johnson, Wilson et al., 2000). This easing of constraints may have
allowed for the diversity of suitable ligands to develop that ultimately led to the
differentiation of the steroid receptor family (Schwabe and Teichmann, 2004).
Upon binding of ligand agonist, a conformational change is induced in the
LBD whereby helix 12, the most C-terminal of the helices, swings over to form a
clamp or lid that stabilizes the occupied ligand binding pocket (Nettles and Greene,
2005). In doing so a hydrophobic interaction surface is formed, known as
26
activation function 2 (AF-2), which serves as an important site for the recruitment of
cofactors such as members of the p160 family of NR coactivators.
As the DBD and LBD are considered the two fundamental domains shared by
all NRs, the above descriptions of their structure in the AR can be generalized to all
of the steroid receptor family members. This homology is not true for the NTD and
Hinge domains, both of which exhibit a high degree of variability. The features and
functions of these regions that distinguish the AR warrant far more detailed
characterization and thus serve as the basis for the projects pursued in this
Ph.D. thesis.

27
Figure 2.3 LEGEND

A) 5α-dihydrotestosterone (DHT).
B) The 12 α-helices that comprise the AR LBD. The hormone binding pocket is
formed by helices 3, 5, and 10. Identities of residues involved in contacting
bound DHT are circled. Position of activation function 2 (AF-2) is indicated.

28

29
The canonical androgen signaling axis

The mature AR is found in the cytoplasm of the cell associated with molecular
chaperones such as heat shock proteins and immunophilins that assist in proper
trafficking and maintenance of the receptor in its inactive state (Ratajczak, Ward et
al., 2003; Thomas, Dadgar et al., 2004)(Fig 2.4). Testosterone and other androgens
are sequestered in the blood by sex-hormone binding globulins (SHBG) that limit
their bioavailability to target tissues (Siiteri, 1979). Following diffusion into the cell,
T is metabolized through 5α-reductase activity into DHT, with the latter exhibiting a
nearly 10-fold greater bioactivity than the former as a ligand for AR (Toth and
Zakar, 1982). Upon binding of DHT to the ligand binding pocket of AR, an
exchange of co-chaperone proteins takes place that allows the transport of the
activated receptor to the nucleus (Heinlein and Chang, 2001; Cheung-Flynn,
Prapapanich et al., 2005). In the past it was thought that once in the nucleus the AR
becomes dissociated from chaperones to allow the receptor freedom to carry out its
regulatory roles. However, newer evidence paints a picture of the subnuclear AR
participating in a cyclic pattern of chaperone associations that helps regulate
transactivation on target genes in a revolving manner (Elbi, Walker et al.,
2004)(Fig 2.5). Ultimately, this cycling allows the AR to recognize and bind the
AREs found in target DNA loci of the genome, homodimerize, and recruit
complexes that mediate 1) ATP-dependent chromatin remodeling, such as the
SWI/SNF complex, 2) acetylation and methylation of histones, such as the
p160 coactivator complex, and 3) assembly of the basal transcription machinery,
30
such as by the TRAP/DRIP complex (Glass, Rose et al., 1997; Rosenfeld and Glass,
2001; Xu and Li, 2003). Likewise, a coactivator/corepressor exchange occurs that
further encourages receptor transactivation through such events as removal of factors
possessing histone deacetylation (HDAC) activities (Hobisch, Hoffmann et al., 2000;
Masiello, Cheng et al., 2002; Liao, Chen et al., 2003; Perissi, Aggarwal et al., 2004).
The resulting relaxation of chromatin and assembly of the RNA Pol II transcription
machinery ultimately leads to the expression of the target gene (Powell, Christiaens
et al., 2004).
31
Figure 2.4 LEGEND

The canonical androgen signaling axis.
1) In the cytoplasm, the nascent AR peptide associates with chaperone proteins such
as Hsc70/Hsp90 and immunophilins.
2) The mature AR adopts a highly structured LBD through chaperone-mediated
folding. At this point the NTD remains largely unformed.
3) DHT binds to the LBD resulting in conformational changes that expose AF-2 and
begin to collapse the NTD structure. An exchange of chaperones takes place that
facilitates transport of activated receptor.
4) AR is translocated to the nucleus through associations of chaperones with nuclear
pore complexes.
5) Subnuclear AR engages in cycling on/off target DNA, where binding of
coregulators further collapse the NTD into structures competent for regulating
transcription via recruitment of chromatin-remodeling factors and the basal
transcription machinery.

32

33
Figure 2.5 LEGEND

Nuclear receptor-mediated transcription on chromatin. The NR dimer binds to
cognate hormone response elements (HRE) and recruits the SWI/SNF complex for
chromatin remodeling, the GRIP1 p160 complex of histone acetylating and
methylating proteins, and the TRAP/DRIP complex that assembles the basal
transcription machinery. Gene expression is regulated through the cyclical exchange
of NR and cofactors at the locus.
34

35
The dependence of prostate cancer on AR signaling

Tissues of the prostate are dependent on androgens not only for growth and
development but for adult maintenance as well (Lindzey, Kumar et al., 1994). It has
been well established that prostate cancers regress in response to androgen
deprivation (Huggins C, 1941). The clinical impact of this is clearly shown by the
fact that over 80% of men with diagnosed prostate cancer respond to androgen
ablation therapies such as orchiectomy or treatment with LHRH agonists/antagonists
or antiandrogens such as hydroxyflutamide or bicalutamide (Miron, 1996). These
hormone treatment strategies are the only option for patients who have progressed to
metastatic disease. Even so, success is essentially palliative as inevitably the tumors
return in forms that are no longer responsive to hormone manipulations (Kozlowski,
Ellis et al., 1991). This transition of the disease from androgen dependence to
ablation resistance carries with it very poor patient prognosis, with the majority of
patients eventually succumbing to the disease (Denis and Griffiths, 2000). In general
the AR gene is normal and expressed in primary cancers, and continues to be
expressed following failure of androgen ablation treatment (Unni, Sun et al., 2004).
With the knowledge that the AR signaling axis remains intact following first-line
hormone deprivation therapy, current disease models regard progression to ablation
resistance not as a loss of AR signaling per se but as a condition whereby the AR
function has been altered to allow continued signaling in environments of castrate-
level androgens (Bentel and Tilley, 1996). Three studies exemplify the
AR-dependence of ablation-resistant prostate cancer cells. First, disruption of AR
36
expression by antibodies or ribozymes has been shown to inhibit proliferation of
ablation-resistant cells in the absence of androgens (Zegarra-Moro, Schmidt et al.,
2002). Second, increased AR expression was determined to be necessary and
sufficient to convert androgen-sensitive prostate cancer cells to an ablation-resistant
state (Chen, Welsbie et al., 2004). Third, a study showed that specific expression in
mouse prostate epithelial cells of an AR transgene harboring a gain-of-function
mutation led to prostate cancer development in 100% of study animals, revealing that
aberrant AR signaling is sufficient to cause cancer (Han, Buchanan et al., 2005).
The AR gene is amplified in 20-30% of the patients experiencing disease
recurrence after primary ablation therapy, perhaps as a cellular adaptation to
compensate for lowered androgen availability (Visakorpi, Hyytinen et al., 1995).
Studies have reported a two-fold increase in both AR and PSA in prostate tumor
samples compared to normal tissue (Koivisto and Helin, 1999; Linja, Savinainen et
al., 2001). Another study revealed that in the CWR-R1 and C4-2 cell models of
recurrent prostate cancer, the AR has higher steady state levels than normal, which
contribute to a hypersensitive cellular response to low DHT levels (Gregory, Johnson
et al., 2001). Thus it is thought that increased AR levels and activity may augment
the sensitivity of the androgen signaling axis as a contribution to disease progression
following ablation therapy.
Several research groups have characterized somatic mutations in the AR gene
that occur in prostate cancer. These mutations can often be identified in primary
tumor cells from pre-ablation tissue samples. But, in those situations they represent
only a small subpopulation of cells within the diseased organ. It is thought that the
37
selective growth pressures placed on the tumor from hormone ablation allow these
small groups of affected cells to expand, and thus the mutations are more readily
identified in tissues taken from tumors that have relapsed following failure of
ablation therapy (Taplin, Bubley et al., 1999). These mutations often allow the AR to
respond to anti-androgens as agonists or make use of other ligands inappropriately.
One of the best characterized is the T877A mutation found in LNCaP cells, a
prostate cancer cell line derived from a lymph node metastasis which, as a result of
this change, has an AR that is responsive to hydroxyflutamide as an agonist
(Wilding, Chen et al., 1989). Studies that have compiled these somatic alterations
seen in prostate cancers have shown that missense mutations of the AR co-locate to
specific regions of the receptor. This observation has served as a tool for identifying
sequences in AR domains that are potentially important for receptor signaling
(Buchanan, Greenberg et al., 2001).
Additionally, upregulation of other cellular factors or signaling pathways that
enhance AR function may be selected for in the depleted hormone environment of
post-ablation prostate tumors. It has been shown that the p160 family of NR
coactivators, which are potent enhancers of AR function, can be detected at elevated
levels in many prostate cancer tissues (Zhou, Yan et al., 2005). The HER2/Neu
signaling cascade has been implicated in driving AR function through activation of
MAPK pathways in prostate cancer cells (Yeh, Lin et al., 1999). This has been
shown to induce Akt-driven phosphorylation of AR in a manner that promotes cell
survival through decrease of AR-mediated apoptosis (Lin, Yeh et al., 2001). The
cytokine growth factors have also been implicated in enhancing AR signaling in
38
prostate carcinomas. As an example, Interleukin-6 (IL-6) has been demonstrated to
positively affect the AR signaling pathway in a ligand-dependent manner that may
contribute to continued cell growth in low hormone conditions (Hobisch, Eder et al.,
1998). However, it is clear that such effects are dependent on the relative levels of
such factors, as over-amplification of IL-6 signaling can also have inhibitory effects
on AR function in vitro (Jia, Kim et al., 2003). Recently our group has identified a
subclass of AR-responsive genes that demonstrate expression in the absence of
androgens that is facilitated by the AR-mediated loosening of chromatin structure
(Jia, Kim et al., 2003). Such genes may contribute to hyper-responsive AR effects in
ablation-resistant tumors.
It is now understood that the transition from androgen dependence to ablation
resistance experienced by the vast majority of prostate cancers following primary
hormone ablation therapy is not simply due to a loss of AR involvement as once
thought; the picture is far more complicated and involves appreciating several
mechanisms by which tumor cells can adapt and compensate under selective
pressures. These paths all converge on the AR as the central signaling hub in prostate
cells, and thus researchers now recognize that the functional AR signaling pathway
continues to exist, albeit perhaps in altered form, throughout the course of disease
progression in response to treatment. An understanding of the myriad mechanisms
by which the AR is able to achieve this functional flexibility in prostate cancer has at
its core the need to fully characterize the structure of the AR molecule itself.
In doing so we may gain insight into how the receptor communicates, compensates,
and ultimately controls the genes that influence disease.
39
CHAPTER 3
Signaling motifs of the AR transactivation domain

INTRODUCTION

The N-terminal transactivation domain (NTD) of the AR comprises over half of the
receptor sequence making it comparable in size to those of the GR, MR, and PR
among the Type I NRs. In contrast, the much shorter NTD of the ER is reminiscent
of those found in the Type II receptors such as TR and VDR, an observation that
supports the theory that as the steroid receptors evolved from a primordial ER, the
NTDs expanded and diversified in response to the broadening roles of steroid
signaling (Fig 3.1). In a simplified functional model of the AR domains, the LBD
acts as an on/off switch for the receptor and the DBD as an anchor that directs the
molecule’s regulatory activities to proper gene targets. The role of the NTD then, is
that of subsequently coordinating the recruitment of other proteins that will influence
transcriptional activity at targeted DNA, ultimately modulating gene expression.
The NTD possesses the action potential of the receptor, and in that sense the other
AR domains serve as regulators of when and where the NTD function is directed.
The NTD is the least conserved of the AR domains. Although greater than 90% of
AR DBD and 70% of AR LBD amino acid sequences have been conserved across
species during evolution, the human AR NTD shares less than 40% homology with
reptiles (Xenapus laevis), 30% with modern fish (Oncorhynus mykiss) and 15% with
ancient fish species (Astatatotilapia burtoni) (Dr. Grant Buchanan, personal
40
communication). This lack of sequence homology once again suggests that the
increasing complexity of physiological systems evolving in higher vertebrates led to
alterations of NTD structures to accommodate novel signaling interactions with other
cellular proteins.

41

Figure 3.1 LEGEND

Schematic domain maps of several Type I and Type II hormone receptors aligned to
contrast the variations in NTD size. Note size conservation of other domains and
homology of overall receptor structure.

42
Localization of transactivation potential within the AR NTD

The transactivation potential of the AR NTD is mainly localized to two overlapping
activation functions: AF-1 spanning aa 142-485, and AF-5 from aa 351-528. These
regions encompass a number of peptide motifs such as microsatellite repeats,
protein-protein interaction surfaces, and phosphorylation and sumoylation regulatory
sites (Jenster, van der Korput et al., 1995; Callewaert, Van Tilborgh et al., 2006).
The tertiary structures of these two broad regions have not been determined as it is
believed that prior to receptor activation, the entire NTD may exist as a large, fluid
peptide domain in contrast to the highly structured DBD and LBD. Such a state has
been previously determined for the NTD of GR (Dahlman-Wright, Baumann et al.,
1995). This globular nature establishes a requirement for cytoplasmic associations
with chaperone molecules such as Hsc70/Hsp90 and immunophilins to achieve
proper folding of the NTD into a mature conformation competent for receptor
function (Reid, Kelly et al., 2002). Upon activation of the receptor, associations with
NR-binding proteins (such as coactivators and corepressors) recruited to chromatin
collapse the NTD into structures that are optimal for transactivation signaling.
Evidence of this has been shown in the case of the transcription factor TFIIF, which
binds AR AF-1 in a manner that induces α-helical structure formation in the region
(McEwan and Gustafsson, 1997; Kumar, Betney et al., 2004). It is likely that this
evolutionary approach of the AR NTD remaining unstructured prior to activation
allows for a variety of NTD binding surfaces to arise in response to different
signaling contexts. One only has to consider that the AR is responsible for mediating
43
a wide range of androgenic cellular responses to appreciate the subtlety with which
the AR NTD must appropriately communicate with other molecules. Thus, creation
of interaction surfaces in the AR NTD is in large part allosterically dependent on the
AR-binding partners that are co-localized with the receptor. It is noteworthy that the
two activation functions appear to operate with some independence, as the primary
transactivation potential of the NTD in a full-length AR molecule is mediated by
AF-1 whereas the focus reverts to the AF-5 region in a truncated AR lacking the
LBD (Jenster, van der Korput et al., 1995). One implication of this is that the LBD
has an inhibitory influence on the NTD in the absence of ligand activation to prevent
inappropriate receptor activity, and there is evidence that this may involve
recruitment of a peptide stretch found just upstream of the DBD in AF-5 as a means
of inhibiting DNA binding (Liu, Wang et al., 2003). This is one of several examples
of specific peptide features within the AF-1 and AF-5 regions that control NTD
activity. These will now be discussed in greater detail.

The polymorphic CAG and GGC microsatellite repeats

There are two polymorphic trinucleotide repeats in the AR NTD. The first consists of
a stretch of ~21 CAG repeats coding for the amino acid glutamine (Q) spanning
aa 58-78, and the second is a stretch of GGN repeats that code for glycines (G)
spanning aa 449-472, the majority of these codons being GGC. In the general
population the size distribution of the AR CAG allele ranges from 6-39 repeats in
North American men but exhibits significant variation among racial-ethnic groups.
44
An estimated 65% of African-American males possess alleles shorter than 22 CAG
repeats compared to 53% of whites and 34% of Asian-Americans (Edwards, 1992).
The presence of shorter CAG and GGC repeats was observed in a high-risk group of
African-American men, corroborating previous findings of an allelic component to
prostate cancer predisposition (Irvine, Yu et al., 1995). In the laboratory, shorter
repeat size has been shown to result in increased activity of the receptor in vitro, with
complete deletion of the CAG repeat resulting in an AR that is several-fold more
active than wild-type (Chamberlain, Driver et al., 1994; Irvine, Ma et al., 2000;
Callewaert, Christiaens et al., 2003). Similar to trinucleotide repeats found in other
genes, it has been shown that the AR CAG microsatellite has expanded though the
course of evolution (Rubinsztein, Leggo et al., 1995; Choong, Kemppainen et al.,
1998). Over-expansion of CAG repeat size has detrimental effects on AR signaling.
Abnormal repeats >40 CAG are associated with diseases such as Huntington’s
disease and spinal and bulbar muscular atrophy (SBMA), commonly called
Kennedy’s disease (Mangiarini, Sathasivam et al., 1997; Domitrz, Jedrzejowska et
al., 2001; Greenland, Beilin et al., 2004). This finding has been attributed to the fact
that AR molecules with >40 CAG repeats show markedly decreased transactivation
activity in vitro even though ligand binding affinities of these molecules are not
affected (Chamberlain, Driver et al., 1994). In addition there is evidence that CAG
repeat size influences the ability of the AR molecule to engage in inter-domain
interactions that are thought to be crucial for normal receptor activity. The range of
16-29 CAG repeats that is optimal for this interaction is found in an estimated 90%
of the population (Buchanan, Yang et al., 2004). Parallels exist in the proposed
45
function of the GGC repeats, with modulations outside the 16-18 GGCs that
represent approximately 90% of the population resulting in decreased AR function
(Ding, Xu et al., 2005). Thus there may have been an evolutionary bias in the
expansion of the CAG and GGC repeats up to an optimal length to favor AR domain
interactions as a means of AR regulation.

A highly conserved signature sequence

Although there is little sequence homology among AR NTD from distantly related
species, an estimated 18-21% of the domain’s amino acid residues are engaged in
forming the secondary structural elements important for signaling (He, Fischer et al.,
1990). It is particularly striking then, that a fourteen amino acid AR NTD signature
sequence (ANTS) is highly conserved in the AR from all species for which coding
sequence is known (Buchanan, Greenberg et al., 2001). This sequence, which in
humans consists of
233
AKELCKAVSVSMGL
246
, is unique to the AR. Considering
the suggested functional importance of this sequence by its unusual conservation and
location within the AF-1 region, it seems likely that ANTS plays an important role in
regulating AR activity. There is evidence that ANTS serves as an interaction site for
factors that modulate receptor stability and may pose as a built-in inhibitory region
that attenuates receptor activity via the balance of stabilization versus degradation
(He, Bai et al., 2004; He, Gampe et al., 2004). Further understanding of this motif
and identification of additional AR cofactor proteins that signal through it are
needed.
46

The role of phosphorylation in receptor function

The newly synthesized AR quickly undergoes post-translational modification via
phosphorylation, with the resulting modified form of the receptor adopting a higher
molecular weight of ~110 kDa (Zhou, Kemppainen et al., 1995). The AR is
phosphorylated at about a dozen serine residues throughout the receptor, with the
majority of these sites located within the NTD. These include constitutively
phosphorylated positions such as serine 94, as well as those that undergo
phosphorylation associated with receptor activation, such as serine positions 16, 81,
256, 308, and 424 (Gioeli, Ficarro et al., 2002). There is quite a bit of evidence
suggesting phosphorylation is a normal means of regulating AR signaling both
directly and indirectly. Of particular interest are those pathways that potentially
activate the receptor without the binding of ligand that become amplified in
advanced hormone refractory tumors. These pathways involve extracellular stimuli
such as growth factors, cytokines, and HER-2/Neu acting through multi-tiered kinase
signaling cascades resulting in AR phosphorylation. Thus these different yet often
overlapping pathways have been extensively investigated for their regulatory roles in
the signaling of various nuclear receptors (Hobisch, Eder et al., 1998; Yeh, Lin et al.,
1999; Shen, Horwitz et al., 2001; Gianni, Bauer et al., 2002).
Activators of protein kinase A (PKA) have been shown to modulate
expression of the AR-mediated PSA gene independent of androgen in prostate cancer
derived cell lines (Blok, de Ruiter et al., 1998). The dose-dependent effects observed
47
in these experiments were conditional on the presence of AR, as evidenced by
inhibition of expression when the anti-androgen bicalutamide was introduced into
the cells to interfere with AR function (Nazareth and Weigel, 1996). It would be
inaccurate, though, to state that the functional consequences of AR phosphorylation
are necessarily stimulatory on target genes; the cytokine IL-6 has been shown to be
both an AR activator as well as a repressor in vivo (Hobisch, Eder et al., 1998; Jia,
Kim et al., 2003). In addition, serine 515, which is near the carboxyl-end of AF-5,
has been shown to be a phosphorylation target of the MAPK signal cascade (Wong,
Burghoorn et al., 2004). Interestingly, experiments conducted using forskolin, a
known stimulant of the MAPK pathway, have yielded conflicting results. LNCaP
prostate cancer cells treated with forskolin have been shown to exhibit enhanced AR
activity as indicated by increased expression of endogenous PSA (Jia, Kim et al.,
2003), whereas this same cell line has been used to demonstrate that forskolin
induces de-phosphorylation of the AR, which results in impaired ligand binding
affinity (Blok, de Ruiter et al., 1998). Currently lacking is an overall characterization
of how the different AR NTD serine positions are individually relevant.
This certainly contributes to discrepant interpretations of how phosphorylation
impacts receptor function. It is also quite likely that the overlapping nature of kinase
signaling cascades within the cell contributes to difficulties in the isolation of
specific responses. Further work is needed to elucidate the complex nature of these
kinase mechanisms that act upon the AR.

48
The role of sumoylation at negative regulatory motifs

The AR is also post-translationally modified by the addition of the 100-amino acid
protein SUMO-1 at two lysine residues in the NTD in a process referred to as
sumoylation (Poukka, Karvonen et al., 2000). These two NTD sites were originally
identified as negative regulatory motifs, and their sequences in human AR,
385
IKLE
388
and
519
VKSE
522
, both closely match the consensus for targeting by the
Ubc-type enzymes that serve as E3 SUMO ligases (Iniguez-Lluhi and Pearce, 2000).
This process is highly reminiscent of ubiquitination in that homologous E1, E2, and
E3 enzymes facilitate the activation and transfer of SUMO-1 to appropriate target
substrates (Wu and Mo, 2007). However, unlike ubiquitination that results primarily
in the irreversible flagging of proteins for degradation via the 26S proteasome
machinery, sumoylation is a reversible process that mediates several different effects
on targeted proteins. Such outcomes include activation, repression, intra-cellular
localization, as well as degradation (Muller, Hoege et al., 2001).
In the case of AR and other steroid receptors, proteins belonging to the
family of Protein Inhibitors of Activated STAT (PIAS) serve as E3 SUMO ligases
(Kotaja, Karvonen et al., 2002). Two members of this group, PIAS1 and PIASxα,
have been shown to repress ligand-dependent transactivation of AR concordant with
enhancement of receptor sumoylation (Nishida and Yasuda, 2002). Interestingly, the
activation of the AR by ligand binding appears to be a prerequisite for sumoylation, a
finding that implicates the covalent attachment of SUMO as a means of deactivating
AR molecules that are already engaged in transcriptional signaling. This repression
49
appears to be reversed by the actions of Zimp10, a recently characterized AR
coactivator that has been shown to conditionally bind to sumoylated AR (Sharma, Li
et al., 2003). Whereas the current understanding of AR sumoylation still leaves
much to be desired, it is clear that this mechanism represents another layer of
regulatory control targeting the AR NTD.

MATERIALS AND METHODS

Site-directed mutagenesis and plasmid preparation

Several AR expression plasmids were used as templates for mutagenesis. The
plasmids pcDNA3.1 AR and cmv-AR both express full-length AR under control of
CMV promoter. The pcDNA AR NTD/DBD plasmid expresses a truncated receptor
that codes for aa 1-647 and is constitutively active. The plasmid VP16 AR-NTD is
the intact transactivation domain aa 1-538 fused behind the activation domain of
VP16. This construct was used in mammalian two-hybrid interaction assays to assess
whether different AR NTD signaling motifs play a role in inter-domain interactions.
Site-directed mutagenesis was performed on these AR templates to introduce
mutations that disrupt regions believed to contain key signaling motifs. This was
accomplished using protocols published in the Stratagene Quick-Change
Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA). Briefly, DNA oligomers
of 40-60 bp were designed that consisted of intentional mismatched bases (resulting
in coding substitutions or deletions) flanked on either end by at least 15 bp with
50
100% homology to the target AR region. As oligos of high quality were needed for
this application, they were prepared and purchased from Operon Biotechnologies
(Huntsville, Alabama, USA). For each desired mutation, a pair of such oligos was
designed that annealed to the same identical span of target DNA but on opposite
strands of the plasmid template. PCR was then performed using high fidelity
Pfu Turbo DNA polymerase (Stratagene) and the resulting product was treated with
Dpn1 (a methylation-dependent restriction endonuclease) to cleave away the
template plasmid in the reaction tubes while leaving the mutation-containing PCR
product intact. The isolated plasmid was then transformed into XL-1 Blue
Supercompetant cells (Stratagene) and clones were selected on LB-Agar plates under
ampicillin selection.
We utilized several luciferase reporter plasmids for assaying activity of the
mutant AR constructs. The reporter PSA(540)-luc consists of firefly luciferase driven
by the androgen responsive 548 bp PSA promoter region (-541 to +7) preceded by
1450 bp of the upstream PSA enhancer region (-5322 to -3873) (Bristol-Myers
Squibb, Princeton, NJ). The reporter PB3-luc (ARR3-tk-luc/tk81-PB3) is a
mammalian expression vector that contains firefly luciferase linked to three copies of
the androgen responsive minimal rat probasin promoter (-244 to -96) ligated in
tandem to the thymidine kinase (tk) enhancer element provided by Dr. R.J. Matusik
from The Vanderbilt Prostate Cancer Center, Nashville, TN (Snoek, Bruchovsky et
al., 1998; Zhang, Thomas et al., 2000).
For mammalian one- and two-hybrid assays, the pGK1 reporter plasmid was
used that expresses firefly luciferase under control of a GAL4-responsive promoter.
51
As a control for driving this reporter, pm3vp16 plasmid was used. This expresses a
fusion protein of GAL4 DBD with VP16 AD that constitutively drives expression of
luciferase from pGK1.

Cell culture and transfection

PC3 cells (ATCC, Manassas, VA) were maintained in RPMI 1640 media
(USC/Norris Cancer Center Microchemical Core facility) that was supplemented
with 5% (v/v) heat-inactivated fetal bovine serum (FBS). COS7 cells (ATCC,
Manassas, VA) were maintained in DMEM hi-glucose media (USC/Norris) also
supplemented with 5% FBS. Cells were plated in 96-well plates (3-mm wells) at the
density of 1x10
4
cells/well and incubated for 24 hours in either RPMI 1640 or
DMEM. Subsequent transient transfection of plasmids was conducted using
LipofectAMINE 2000 reagent (Invitrogen, Carlsbad, CA) as per manufacturer’s
suggested protocol. Briefly, AR and appropriate reporter plasmids were pre-mixed in
Opti-MEM I reduced serum media (Invitrogen, Carlsbad, CA) and subsequently
introduced to LipofectAMINE 2000 reagent resulting in approximately 100ng/well
of total plasmid and 0.4 µl/well of transfection reagent to achieve a total well volume
of 50 µl media. DNA/LipofectAMINE mixes were incubated for 30 minutes at room
temperature to allow formation of micelle complexes. Cells were then exposed to the
transfection mix for 3 hours, after which media was changed to fresh RPMI 1640 or
DMEM for overnight incubation. The next day, media was replaced with either fresh
RPMI 1640 or DMEM containing 10 nM DHT or EtOH vehicle for subsequent
52
24 hour treatment. After incubation, cells were lysed with 1x Passive Lysis Buffer
(Promega, Madison, WI) and extracts assayed for luciferase activity using Promega
Luciferase Assay System measured on a MLX Microtiter Plate Luminometer
(Dynex, Chantilly, VA).

RESULTS

Interference assays reveal the two main activation functions in AR NTD

When pm3vp16 was co-transfected in COS7 cells with pGK1 reporter, strong
constitutive luciferase activity resulted. Co-expression of VP16 AR-NTD fragments
interfered with this constitutive reporter activity (Fig 3.2). The extent of interference
was dependent on the fragment of AR NTD involved. Neither co-expression of the
empty VP16 vector nor of VP16 AR NTD coding aa 487-557 (the C-terminal of AR
NTD) interfered with reporter activity. Expression of a peptide of AR NTD spanning
aa 351-426 showed greater impact, attenuating the reporter signal by approximately
50% compared to empty VP16 vector alone. However, even greater repression of
signaling was observed when peptides of the regions AR aa 141-356 and 404-501
were expressed. These two regions coincide with the designated AF-1 and AF-5
activation functions of the AR NTD, suggesting that this type of interference assay
may be useful in the future for identifying potential activation motifs within
transcription factor domains.

53
Figure 3.2 LEGEND

Interference assays reveal transactivation potential of AR NTD. COS7 cells were
transfected with the pm3vp16 construct that encodes a fusion protein of GAL4 DBD
with VP16 AD (GAL4-VP16) along with GAL4-responsive luciferase reporter using
methods described in the chapter text.
A) Co-expression of various AR NTD constructs differentially interferes with GAL4-
VP16 ability to drive luciferase expression, suggesting that the AR NTD sequesters
transcription factors away from the reporter. Interference profile across several
AR NTD fragments suggests AR amino acid stretches 141-356 and 404-501 harbor
strong transactivation potentials, coinciding with AR AF-1 and AF-5.
B) In an identical system, co-expression of full-length AR interferes with reporter
expression in a DHT-dependent manner, suggesting that only through ligand
activation does the AR transactivation potential compete for transcription factors
against GAL4-VP16. Expression of AR constructs having increasing CAG
trinucleotide repeats sizes results in little change in interference until the n=50
construct is used, which loses much of its ability to interfere with GAL4-VP16
activity. This is consistent with the reported loss of transactivation strength exhibited
by AR with >40 CAG.
54

55
Poly-glycine GGC repeat length impacts transactivation

Modulations of the poly-glycine trinucleotide repeat size in the AR NTD affected
receptor activity (Fig 3.3). Compared to wild-type AR having a repeat size of
17x GGC, constructs with 9x GGC and 8x GGC yielded progressively lower
transactivation levels. Finally, complete deletion of the GGC tract created a receptor
that could only achieve 20% of original transactivation.

Substitution mutations reveal contributions of various signaling motifs

The series of site-directed mutations designed in this study demonstrated that none of
the investigated signaling motifs were absolutely necessary for AR transactivation on
an ectopically expressed reporter. However, the data also suggest that achieving
optimal receptor activity involves their combined contributions. Deletion of the
ANTS signature sequence increased activity of full-length receptor by approximately
two fold, but this effect was not seen in the constitutively active, truncated AR
NTD/DBD molecule (Fig 3.4). The motif that appears to have the highest impact on
proper AR function is
23
FQNLF
27
, which is known to mediate domain interactions
between the AR NTD and LBD in an agonist-activated receptor. The mutation
23
FQNAA
27
lowered transactivation of the AR down to 30% of wild-type in both
COS7 and PC3 cells (Fig 3.5). Manipulation of the sumoylation sites at either NRM1
or NRM2 had no substantial impact on transactivation. Likewise, the S515A and
S515E substitutions made at the putative MAPK phosphorylation site did not have
pronounced effects on transactivation, although notably the exposure of the cells to
56
forskolin, an extracellular inducer of MAPK signaling cascades, seemed to have
repressive effects on transactivation across all constructs with greatest impact on the
S515A/E mutants.

57

Figure 3.3 LEGEND

Effect of alterations in poly-glycine GGC repeat size on AR transactivation. COS7
cells were transfected with AR constructs having GGC microsatellite repeat sizes of
n=17, 9, 8, and 0 along with PB3-luc reporter as per methods described in chapter
text. Cells were stimulated with 10 nM DHT for 24 hours.

58
Figure 3.4 LEGEND

Effect of AR NTD signature sequence (ANTS) deletion on receptor transactivation.
AR containing deletion of amino acids 233-246 was transfected into PC3 cells along
with PSA(540)-luc reporter and stimulated with 10 nM DHT for 24 hours.
A) Full length AR ΔANTS construct exhibits identical basal activity as wild-type
AR, but has two-fold higher DHT-stimulated response.
B) Truncated AR ΔANTS construct consisting of AR NTD/DBD without LBD
shows identical constitutive transactivation with wild-type on PSA(540), PB3, and
MMTV luciferase reporters. Together the data suggest that the deletion of ANTS
sequence only impacts AR signaling through the AF-1 that is responsive to ligand
binding in full-length receptor, but does not impact signaling through AF-5 that is
the dominant activation function in the truncated AR NTD/DBD protein.
59

60
Figure 3.5 LEGEND

Effect of amino acid substitutions at various AR NTD signaling motifs on
transactivation. PC3 cells were transfected with AR constructs, GRIP1 coactivator,
and PSA(540)-luc reporter as per methods described in chapter text. Cells were
treated with 10 nM DHT and/or 50 µM forskolin for 24 hours to induce MAPK
kinase activity. WT = wild-type AR. FQNAA = AR with FQNLF-to-FQNAA
mutation. NRM1* and NRM2* = AR with K-to-R substitution at the two
sumoylation sites. S515A and S515E = AR with substitution of S-to-A or S-to-E at
the S515 phosphorylation site.
61

62
DISCUSSION

Signaling motifs and trinucleotide expansions

As the physiological roles of androgens diversified in higher organisms, there was an
obligatory increase in the complexity of AR signaling (Thornton and Kelley, 1998;
Owen and Zelent, 2000; Schwabe and Teichmann, 2004). Characterization of the
roles of the unique AR NTD signaling motifs may directly address how the AR is
differentially regulated from the other steroid receptors.
The
23
FQNLF
27
motif greatly contributes to optimal transactivation of wild-
type AR. It is believed that the main mechanism by which this is accomplished is
through the FQNLF directly interacting with the AF-2 of the AR LBD, thereby
mediating inter-domain folding (He, Kemppainen et al., 2000). However, FxxLF
motifs are also used by other AR coregulators such as ARA70/RFG, ARA55/Hic-5,
and ARA54 to target the AR AF-2 (He, Minges et al., 2002). Newer evidence
suggests that the FQNLF of the AR itself may be the binding site of co-regulatory
molecules that alter AR transactivation through modulating receptor folding (Bai, He
et al., 2005). It is noteworthy that in our current study, alteration of the FQNLF to
FQNAA did not completely abolish transactivation, suggesting that AR NTD can
continue to recruit factors necessary for activity even in the absence of N/C
interaction. This is perhaps evidence that the direct N/C interaction was a more
recent evolutionary adaptation that further distinguished AR from other steroid
receptors.
63
The instability of trinucleotide repeats in genes has been implicated in the
development of several diseases including fragile site syndromes, myotonic
dystrophy and several neurodegenerative disorders (Bates and Lehrach, 1994;
Ashley and Warren, 1995; Hummerich and Lehrach, 1995). Of the neuronal diseases,
expansions of the AR poly-glutamine (CAG) microsatellite have been associated
with Huntington’s disease, spinal and bulbar muscular atrophy and spinocerebellar
ataxia type 1 (Bates and Lehrach, 1994). However, to date there has been no
conclusive evidence that the GGC repeat size is associated with disease. The variably
sized GGC trinucleotide repeat follows an invariable short sequence of six glycine
codons 3xGGT, 1xGGG, 2xGGT that are often counted in the entire GGN tract.
In 2004 an extensive meta-analysis summarized over eight separate studies that
prostate cancer cases had a mean difference of only 0.09 fewer GGN repeats than did
controls, and subsequently questioned the biological significance of this variation
(Zeegers, Kiemeney et al., 2004). Our own group first examined potential
associations between GGC repeat size and risk of prostate cancer in 1995 (Irvine, Yu
et al., 1995). This combined CAG and GGC microsatellite study involved genotypic
comparisons of AR from blood-extracted DNA of 68 prostate cancer cases with
those of 123 healthy controls. The study concluded that there was a protective effect
of having 16 GGC repeats compared to other sizes (relative risk of GGC
not 16=1.18). This effect was more pronounced when CAG<20 size was taken into
account (relative risk when CAG<20 and GGC not 16=2.10). In addition, there was
much higher indication of linkage disequilibrium between the two microsatellite
repeats in cases compared to control men. In the current study, we wished to
64
determine what effect modulations of the GGC tract alone would have on receptor
activity. The data suggest that AR activity is proportional to GGC repeat size but, as
the microsatellite sizes constructed for these assays were artificial and do not reflect
the normal range seen in humans, these in vitro observations cannot as yet be
extrapolated to the physiology of real disease.

The flexibility of AR NTD signaling is apparent in transient transfection systems

In this study we utilized the ectopic expression of various AR constructs to gauge the
impact of site-directed AR NTD mutations on receptor activity. We found that, with
only one exception, targeting single peptide motifs through amino acid substitutions
did not result in clear or dramatic effects on the ability of AR to drive androgen-
responsive reporters. This may be indicative of the highly flexible nature of AR NTD
signaling whereby having multiple protein-protein interaction surfaces within this
domain is compensatory and allows transactivation to proceed despite disruptions in
a specific signaling motif. It also seems likely that the evolution of such an array of
signaling sequences addressed separate and distinct regulatory mechanisms that are
not necessarily active simultaneously and may not be represented well in the cellular
conditions used in this study. Another caveat inherent to the use of transiently
transfected plasmid constructs is that the resulting signaling pathway may have an
artificiality that does not extrapolate well to normal physiology. Overexpression of
proteins from transfected plasmid may swamp the endogenous pool of AR
coregulators in the cell, resulting in signal output that resembles more of an on/off
65
response rather than a gradient. This effect masking may also be compounded by
experimentally manipulated ligand levels falling outside the normal physiological
range, which results in either under-activation of the available receptor pool or an
over-saturation of receptor signal that leaves little margin for fluctuation. To date
there is no reliable method for determining the intra-cellular levels of DHT in situ, as
this is influenced by a number of metabolic factors including hormone conversion
rates and transport. In cell culture then, concentrations of ligand are typically chosen
based on empirical evidence to yield favorable signal-to-noise ratios. The ligand
concentrations used for these experiments may require further titration in order to hit
a balance in receptor activation versus saturation that allows for observation of
potentially subtle variations in receptor behavior. Lastly, the choice of cellular
background for these types of experiments is of paramount concern. For the current
work, the choice of using PC3 prostate cancer cells was made in consideration of
1) that these cells do not possess an endogenous AR signaling response against
which a transfected AR construct would be competing, and 2) that a prostate cancer-
derived cell line would be an appropriate background in which to expect strong
signaling from the transfected receptors. In retrospect, whereas the first point
remains a valid concern for this type of experimentation, on the second point perhaps
we were a bit too optimistic with our expectation that the select NTD mutations
would impact AR function enough to discern clearly in a cellular environment
essentially selected for maximal androgen signaling. Indeed, with the exception of
the
23
FQNLF
27
motif, none of the substitution mutations had appreciable effect. It is
possible that within these PC3 cells there exists an environment that is simply too
66
favorable for AR signaling, perhaps as a result of the selective outgrowth of these
cancer cells having led to altered coactivator/corepressors levels that strongly
enhance the ability of AR to function.
67
CHAPTER 4
Involvement of p160 coactivators in AR N/C interaction

INTRODUCTION

Nuclear receptors regulate the transcription of target genes through modulating
chromatin structure and initiating the assembly of the RNA Pol II transcription
machinery. These tasks are accomplished through recruitment of coactivators and
corepressors in a highly choreographed exchange brought on by ligand binding.
The p160 family members of NR coactivators, so named for their approximate
molecular weight of 160 kDa, directly bind NRs and potently enhance receptor-
mediated transcription via secondary associations with factors possessing histone
acetylation and methylation activities. These activities result in the relaxation of
chromatin, thereby allowing access to DNA by other transcription factors. The p160
family, alternatively called the steroid receptor coactivator (SRC) family, consists of
three related members: SRC-1 (NcoA-1), SRC-2 (GRIP1, TIF2, NcoA-2), and
SRC-3 (p/CIP, RAC3, ACTR, AIB1, TRAM-1). These proteins were first
recognized for their NR coactivator functions in the mid 1990s (Onate, Tsai et al.,
1995; Hong, Kohli et al., 1996; Torchia, Rose et al., 1997). The p160s are general
coactivators of the steroid receptors in that each of the family members can enhance
the ligand-stimulated transcriptional regulation of multiple NRs (Xu and O'Malley,
2002). As a result there is apparent redundancy of function in vivo among the p160s,
68
but knock-out animal models have shown some differences in resulting phenotypes
(Xu and Li, 2003).
All three of the p160 coactivators show high structural homology (Fig 4.1).
In the p160 molecule, the primary interaction surface for NRs is localized
approximately halfway between the N and C-termini of the coactivator where three
LxxLL motifs (where L is leucine and x is an unspecified residue) together form
what is commonly termed the NR-box domain (Ma, Hong et al., 1999). These
LxxLL motifs physically interact with the hormone-bound LBDs of the steroid
receptors. The C-terminal of the p160 protein contains two distinct activation
domains, AD-1 and AD-2, which are involved in recruiting the secondary factors
that alter chromatin structure through modification of histone tails. The more
proximal AD-1 regulates recruitment of CBP/p300 histone acetyltransferase
activities, whereas the distal AD-2 recruits the methyltransferase activities of
CARM1 (coactivator-associated arginine methyltransferase 1) and PRMT1 (protein
arginine methyltransferase 1) (Chen, Ma et al., 1999; Chen, Huang et al., 2000;
Stallcup, Chen et al., 2000; Ma, Baumann et al., 2001; Strahl, Briggs et al., 2001;
Teyssier, Chen et al., 2002).

69
Figure 4.1 LEGEND

Domain map of GRIP1 constructs used in AR N/C interaction assays. Top diagram
depicts full-length GRIP1 WT coactivator amino acids 1-1462 and positioning of
functional domains that are highly homologous among the p160 family members.
NR-boxes that contain LxxLL motifs are indicated in Roman numerals.
AD-1 = activation domain 1, recruitment site for CBP/p300 histone acetylation
activities. AD-2 = activation domain 2, recruitment site for CARM1 and PRMT1
histone methylation activities. Middle diagram is of truncated GRIP1 (1-1121) that
lacks AD-2 and cannot bind AR NTD. Lower diagram is of the GRIP1 NR*
construct that has the second and third NR-boxes altered to LxxAA (asterisk) and
demonstrates impaired binding to AR AF-2. (Image adapted from Ma et al., 1999).
70

71
The AF-2 activation function of steroid receptor LBDs

The ligand binding domains of steroid receptors undergo conformational
rearrangement in response to agonist binding. This results in the formation of the
highly conserved protein-protein interaction surface known as AF-2 that is the
primary site for p160 coactivator recruitment (Henttu, Kalkhoven et al., 1997; Feng,
Ribeiro et al., 1998; Tanenbaum, Wang et al., 1998; Bevan, Hoare et al., 1999;
Warnmark, Treuter et al., 2003; Elhaji, Stoica et al., 2006). Early work demonstrated
that upon ligand activation, the AR NTD and LBD domains interact in what has been
referred to as an AR N/C interaction facilitated by coactivators (Wong, Zhou et al.,
1993; Ikonen, Palvimo et al., 1997). More recently it was shown that AF-2 of most
nuclear receptors is specifically occupied by the LxxLL-like motifs of coregulatory
proteins. Of these, the p160 family is the most widely characterized (Ma, Hong et al.,
1999). In contrast, AF-2 of the activated AR is predominantly occupied by the AR’s
own LxxLL-like peptide,
23
FQNLF
27
in humans, found in the AR NTD (He and
Wilson, 2002). This interaction is thought to mediate AR homodimerization and thus
be important for AR function in vivo (Li, Fu et al., 2006). A puzzling observation is
that the AR behaves quite differently than the other steroid receptors in that its AF-2
interacts with greater specificity and avidity toward this single
23
FQNLF
27
in its own
NTD than with the NR-boxes of p160 coactivators. This suggests competition
whereby N/C interaction and p160 binding may be mutually exclusive (He, Bowen et
al., 2001).
72
It has been shown through GST-pulldown assays that the p160 coactivator
GRIP1 not only interacts with the AR AF-2 region through its LxxLL NR-boxes, but
also with the AR NTD through the GRIP1 C-terminal domain (Ma, Hong et al.,
1999; Irvine, Ma et al., 2000). Evidence of physical protein-protein interactions
between GRIP1 and the separate AR NTD and LBD domains has led us to
hypothesize that GRIP1 may bridge the AR domains, thereby facilitating or
stabilizing AR inter-domain communication. We have proposed that the primary
consequence of this facilitated bridging is increased AR transactivation activity.
In the present study we assessed whether AR N/C interaction and
transactivation of target genes are mechanistically linked and how the p160
coactivator GRIP1 participates in the two processes.

MATERIALS AND METHODS

Preparation of plasmids

Mutations in AR expression plasmids were constructed utilizing the protocol
published in the Stratagene Quick-Change Site-Directed Mutagenesis Kit
(Stratagene, La Jolla, CA). Mutations in full-length receptor were constructed using
our laboratory’s 8.4 kb pcDNA AR plasmid that contains a 3050 bp AR cDNA insert
expressing human AR aa 1-919 under control of the CMV promoter. The
subsequently constructed plasmid pcDNA AR FQNAA codes for full-length AR
with the
23
FQNLF
27
to
23
FQNAA
27
mutation. In experiments assaying full-length
73
receptor activity, two androgen-responsive luciferase reporters were used. The
reporter plasmid PSA(540)-luc consists of firefly luciferase driven by the androgen
responsive 548 bp PSA promoter region (-541 to +7) preceded by 1450 bp of the
upstream PSA enhancer region (-5322 to -3873) (Bristol-Myers Squibb, Princeton,
NJ). PB3-luc (ARR3-tk-luc/tk81-PB3) is a mammalian expression vector that
contains firefly luciferase linked to three copies of the androgen responsive minimal
rat probasin promoter (-244 to -96) ligated in tandem to the thymidine kinase (tk)
enhancer element provided by Dr. R.J. Matusik from The Vanderbilt Prostate Cancer
Center, Nashville, TN (Snoek, Bruchovsky et al., 1998; Zhang, Thomas et al., 2000).
For mammalian two-hybrid assays examining AR N/C interaction, the AR
NTD and LBD domains were separately cloned into vectors derived from the
Clontech Mammalian Matchmaker Two-Hybrid Assay Kit (Clontech, Palo Alto, CA,
USA). The construct VP16-AR NTD codes for wild-type AR transactivation domain
amino aa 1-538 fused behind a portion of the VP16 activation domain (aa 411-455),
constitutively expressed by SV40 early promoter. The plasmid VP16-AR FQNAA
was subsequently derived from this and differs by containing the AR
23
FQNAA
27

mutation. The construct GAL4-AR LBD is the AR LBD aa 644-919 fused behind the
DNA binding domain of GAL4 (aa 1-147) also under control of SV40 promoter.
As a reporter for AR N/C interaction, the plasmid pGK1-luc was used that expresses
firefly luciferase under control of a GAL4-responsive promoter (Fig 4.2).
Three GRIP1 expression constructs were a gift from the laboratory of
Dr. Michael Stallcup (University of Southern California) and were used to determine
the effects of p160 coactivator involvement in both full-length AR transactivation as
74
well as AR N/C interaction assays. These consisted of the plasmids PSG5 GRIP1
coding for full-length GRIP1 aa 1-1462, PSG5 GRIP1 (1-1121) coding for a
truncated form of the protein lacking the C-terminal AD-2, and PSG5 GRIP1 NR*
coding for full-length protein that has the second and third LxxLL motifs in the
NR-box domain mutated to LxxAA.

75
Figure 4.2 LEGEND

Overview of the mammalian two-hybrid assay system.
1) Plasmids encoding fusion peptides of AR NTD with VP16 AD, and AR LBD with
GAL4 DBD were co-transfected along with a GAL4-responsive luciferase reporter
into COS7 cells as per methods described in the chapter text.
2) Cells were allowed to express the fusion peptides for 24 hours, after which DHT
stimulation was introduced for a subsequent 24 hours.
3) The GAL4 DBD binds to GAL4 responsive elements on the reporter, but the
AR LBD has little transactivation potential alone. Interactions between the DHT-
activated AR LBD and AR NTD result in VP16 AD recruiting transcription factors
to the reporter. The readout of luciferase activity thus serves as a proxy for the AR
peptide interactions.
76

77
Cell culture and transfection

COS7 cells (ATCC) were maintained in DMEM hi-glucose media (USC/Norris) that
was supplemented with 5% (v/v) heat-inactivated fetal bovine serum (FBS) and used
between passages 7-12. Cells were plated in 96-well plates (3-mm wells) at a density
of 2x10
4
cells/well and incubated for 24 hours in DMEM. Subsequent transient
transfection of plasmids was conducted using LipofectAMINE 2000 reagent
(Invitrogen, Carlsbad, CA) as per manufacturer’s suggested protocol. In preparing
DNA mixes for transfection, molar amounts of expression vectors were balanced
with their empty counterparts as controls, and total nanogram amounts of DNA were
balanced by addition of pCAT-basic, a eukaryotic promoter-less vector. Under these
conditions, experimental and control situations contained the identical concentrations
of total DNA per transfection (i.e., ng/transfection) and identical copy number of
promoter elements (i.e., molar amounts). After transfection, cells were grown in
Opti-MEM I reduced serum medium (Invitrogen, Carlsbad, CA) in the presence of
10 nM DHT (Sigma Chemical Co., St. Louis, MO) or EtOH vehicle for 24 hours in
the case of mammalian two-hybrid assay, or 36 hours in the case of full-length AR
transactivation assay. After incubation cells were lysed with 1x Passive Lysis Buffer
(Promega) and extracts assayed for luciferase activity using Promega Luciferase
Assay System measured on a MLX Microtiter Plate Luminometer (Dynex).

78
Western blotting analysis

COS7 cells were plated in 60 mm dishes at a density of 6x10
5
cells/dish and grown in
DMEM hi-glucose media for 24 hours prior to DNA transfection using
LipofectAMINE 2000 reagent. Treatment with DHT and subsequent incubation
times paralleled the 96-well plate protocols described above. Cells were then
harvested in 100 µl RIPA buffer (10 nM sodium phosphate, 2 mM EDTA,
150 mM NaCl, 50 mM NaF, 0.1% SDS, 1% IGEPAL CA-630,
1% sodium deoxycholate, 0.2 mM Na
3
VO
4
pH 7.2) that contained a cocktail of
mammalian protease inhibitors (Sigma-Aldrich, St Louis, MO). Equal volumes of
each extract were analyzed by SDS-PAGE. Proteins were transferred to Hybond-P
membrane (Amersham Pharmacia Biotech) and probed with anti-AR (N20) (Santa
Cruz Biotechnology, Santa Cruz, CA), anti-GAL4-DBD monoclonal antibody
(Clontech), anti-VP16-AD polyclonal antibody (Clontech) and anti-actin antibody
(Santa Cruz Biotechnology). HRP-conjugated anti-rabbit IgG (Santa Cruz
Biotechnology) and goat-anti-mouse HRP-conjugated antibody (BD Biosciences,
San Jose, CA) were used as secondary antibodies. Detection was performed using
Western Blotting Luminol Reagent (Santa Cruz Biotechnology) according to
manufacturer’s suggested protocol. Detection

was performed using the Enhanced
Chemiluminescence Western Blotting

System (Amersham Pharmacia Biotech)
according to the manufacturer’s

protocol. To demonstrate equivalent loading of
proteins in samples, Coomasie blue staining of gels was prepared. Chemiluminescent
79
images were analyzed using the Fluor-S Max MultiImager Quantification

System
(Bio-Rad, Hercules, CA).

RESULTS

AR transactivation proceeds in the absence of N/C interaction

Activity of the full-length AR receptor and its ability to engage in N/C domain
interactions are both dependent on DHT stimulation, but to different extents.
Concentrations of DHT necessary to yield half-maximal activity were about 1 nM
and 10 nM for transactivation activity and N/C interaction respectively, indicating a
disparity between the two processes. It is notable that at 1 nM DHT, full-length
receptor transactivation activity had already achieved half-maximum rates, whereas
N/C interaction was barely detectable over background. The data reveal that the
former process can proceed without the latter (Fig 4.3).

GRIP1 can restore N/C interaction in the AR FQNAA mutant

When both wild-type AR NTD and AR LBD domains were co-expressed in cells,
treatment with DHT resulted in strong interaction not demonstrated by each fusion
protein expressed with irrelevant binding partners (Fig 4.4). This interaction was
enhanced nearly ten fold when wild-type GRIP1 was co-expressed, suggesting that
GRIP1 may stabilize the AR N/C interaction by physically bridging the two AR
domains. Recruitment of GRIP1 to GAL4-AR LBD in the absence of an interacting
80
AR NTD resulted in comparatively negligible activity, demonstrating that increased
luciferase activity was not an artifact of inherent GRIP1 effects on reporter
transcription. When AR NTD containing the alteration
23
FQNAA
27
was co-expressed
with wild-type AR LBD, treatment with DHT alone resulted in an N/C interaction
that was only marginally above background level. This indicates that the wild-type
23
FQNLF
27
motif is indeed a key interaction surface for DHT-driven AR N/C
interaction. However, DHT-dependent N/C interaction involving the
23
FQNAA
27
mutant was dramatically restored when wild-type GRIP1 was co-expressed in the
assay, suggesting that the enhancing effect of GRIP1 on N/C interaction could still
operate in the presence of the altered FQNAA motif. The restored luciferase activity
could not be attributed to artifacts of non-AR directed GRIP1 function as once again,
GRIP1 recruitment to GAL4-AR LBD alone did not result in appreciable activity on
the same measurement scale. Therefore the recovery of
23
FQNAA
27
N/C interaction
by GRIP1 is evidence that the enhancement of AR N/C interaction is likely due to
binding surface(s) in the AR NTD, other than the FQNLF motif, that are physically
engaging the coactivator (Fig 4.5).

81
Figure 4.3 LEGEND

Disparity between the processes of AR N/C interaction and receptor transactivation.
COS7 cells were transfected with GAL4-AR LBD and VP16-AR NTD constructs
along with GAL4-luc reporter for mammalian two-hybrid assay, or CMV-AR and
PB3-luc reporter to assay for transactivation as described in the chapter text. Cells
were treated with varying concentrations of DHT for 24-36 hours depending on
assay. Maximum luciferase activities from each assay were normalized to 1.0 on the
y-axis and represented as relative light units (RLU). TA = receptor transactivation
response. N/C = mammalian two-hybrid reporter response.
82

83
Figure 4.4 LEGEND

Mammalian two-hybrid assay results from AR N/C interaction scenarios.
DHT-induced interaction of AR LBD with wild-type AR NTD FQNLF is enhanced
ten fold in the presence of GRIP1 WT, but much less so by GRIP1 (1-1121) or
GRIP1 NR* constructs. DHT-induced interaction of AR LBD with mutant AR NTD
FQNAA is negligible compared to wild-type response, but GRIP1 WT substantially
rescues impaired interaction. However, neither GRIP1 (1-1121) nor GRIP1 NR* can
rescue interaction comparably to GRIP1 WT. Control plasmids: GAL4 = empty
vector expressing GAL4 DBD alone. VP16 = empty vector expressing VP16 AD
alone.
84

85
Figure 4.5 LEGEND

A) Mammalian two-hybrid assay results indicating that relative interaction levels
between GRIP1 and AR NTD FQNLF versus AR NTD FQNAA are identical.
The data indicate that GRIP1 does not depend on the AR FQNLF for binding to
the AR NTD, and supports the model of GRIP1 rescuing the impaired AR
FQNAA N/C interaction ability through bridging at an alternate AR site.
Controls: GAL4-p53 = fusion peptide of GAL4 DBD with p53, VP16-T = fusion
peptide of VP16 AD with large-T antigen.
B) Western immunoblot analysis of the VP16-AR NTD FQNLF/FQNAA and
GAL4-AR LBD fusion peptides when co-expressed and stimulated by DHT.
Antibodies against the VP16 AD and GAL4 DBD domains were used as probes.
Antibody against actin used as protein loading control. Co-expression of GRIP1
constructs results in modest stabilizing effects on the AR NTD peptides but has
no apparent effect on the AR LBD.
86

87
GRIP1 bridging of AR N/C interaction is dependent on two coactivator motifs

We reasoned that GRIP1 mutants lacking the ability to bind either the AR NTD or
AR LBD would fail to properly bridge the domains and thus fail to enhance the
DHT-driven AR N/C interaction. It has been previously demonstrated that there are
physical interactions between two independent subdomains of GRIP1 with the
separate AR NTD and AR LBD (Ma, Hong et al., 1999; Irvine, Ma et al., 2000).
To address the implications of these findings in the present experiments, we utilized
two GRIP1 mutant constructs that exhibit impaired binding to either the AR NTD or
LBD. The first contains a truncation that eliminates the C-terminal 341 amino acids
[GRIP1 (1-1121)], thus losing the region surrounding AD-2

previously shown to be
necessary for GRIP1/AR NTD binding (Ma, Hong et al., 1999). The second has
LxxLL to LxxAA alterations in NR-boxes 2 and 3 (GRIP1 NR*), thereby
eliminating the ability of GRIP1 to bind AF-2 in the AR LBD (Ma, Hong et al.,
1999). When the two mutant GRIP1 constructs were introduced individually into
cells along with co-expressed wild-type AR NTD and LBD fragments,
DHT-dependent N/C interaction was unaffected by GRIP1 (1-1121) mutant but
marginally enhanced by GRIP1 (NR*) mutant. The latter observation suggests that
even in the absence of GRIP1/AR LBD binding, the recruitment of GRIP1 to the
transcription complex may still occur through GRIP1 associations with the AR NTD.
Furthermore, Western immunoblot analyses showed a small but consistent increase
in the levels of AR NTD fusion proteins in the presence of GRIP1 and
GRIP1 mutants, suggesting stabilization of AR NTD by the p160 coactivator. Within
88
the same assay conditions, results involving the AR FQNAA construct were
dramatically different. Whereas wild-type GRIP1 rescued the inhibited AR FQNAA
participation in N/C interaction as previously noted, the two GRIP1 mutants were
largely unable to do so. This indicates that both AR NTD- and LBD-interacting
domains of GRIP1 are necessary to achieve the putative bridging activity.
Once again the small induction of AR N/C interaction seen in response to the
co-expression of the GRIP1 NR* mutant is likely due to the activity of GRIP1 NR*
recruited to the AR NTD. It is clear that substantially lower AR N/C interaction
activity occurred between AR FQNAA and AR LBD in the presence of either GRIP1
mutant as compared to interactions involving wild-type AR FQNLF.

GRIP1 bridging of AR N/C interaction is not required for coactivation

We next examined the effects of GRIP1 on full-length wild-type and
23
FQNAA
27

mutant AR transactivation (Fig 4.6). Co-expressed wild-type GRIP1 resulted in
approximately a three fold increase in DHT-dependent receptor activity. When either
of the two mutant GRIP1 constructs were substituted, the receptor transactivation
activity was still enhanced over that of DHT stimulation alone, albeit modestly.
Compared to wild-type AR, the coactivation profile of the AR FQNAA mutant by
GRIP1 was similar. Importantly, substantial transactivation activity was observed by
this AR mutant (approximately 30% of AR FQNLF activity) even in the presence of
mutant GRIP1 molecules. This is in direct contrast to these same GRIP1 mutants
failing to rescue N/C interaction activity between AR FQNAA and AR LBD.
89
Thus, a disparity between levels of AR N/C interaction and AR transactivation
activity was evident, indicating not only that AR N/C interaction is not necessary for
transactivation but that the functional contributions of GRIP1 as a coactivator of AR
involve mechanisms beyond simply the bridging of the AR domains.

90
Figure 4.6 LEGEND

Comparison of the transactivation ability of full-length AR FQNLF versus AR
FQNAA in the presence of GRIP1 constructs. COS7 cells were transfected with AR,
GRIP1, and PSA(540)-luc reporter plasmids as per methods listed in chapter text.
The DHT-induced activity of AR FQNAA was approximately 30% that of AR
FQNLF. Transactivation of both constructs was substantially enhanced by GRIP1
WT and GRIP1 NR*, but to a much lesser extent by GRIP1 (1-1121).
Inset: Western immunoblot analysis of expression levels of both AR constructs.
91

92
DISCUSSSION

Variety in the inter-domain interactions of nuclear receptors

Other nuclear receptors have been shown to utilize the interplay between their AF-1
and AF-2 regions as a response to ligand binding (Metivier, Penot et al., 2001;
Bommer, Benecke et al., 2002). However, unlike AR these receptors are often
entirely dependent on adapter proteins such as the p160s or CBP/p300 to achieve this
folding, as exemplified by the estrogen receptor (both ER α and β) (Kobayashi,
Kitamoto et al., 2000). Furthermore the functional consequences of inter-domain
interactions across different receptors are inconsistent. In the progesterone receptor
(PR), domain interaction between the hinge/amino-terminal regions has been shown
to increase activity; for PPAR-γ, these types of interactions result in weakened ligand
binding affinity (Tetel, Jung et al., 1997; Shao, Rangwala et al., 1998; Wardell,
Kwok et al., 2005). Inter-domain interactions may be used by the mineralocorticoid
receptor (MR) to adjust the receptor’s transactivation response level to different
ligands (Rogerson and Fuller, 2003). Other steroid receptors such as glucocorticoid
receptor (GR) apparently do not engage in detectible N/C interactions at all
(Spanjaard and Chin, 1993). It is striking then, that the AR N/C interaction is distinct
in its ability to occur not only with the assistance of the p160 coactivators, but from
direct and high-affinity binding mediated by the receptor’s own NTD FQNLF motif
with the LBD AF-2.
93
GRIP1 acts as an AR coactivator through different mechanisms

In this study we demonstrated that the GRIP1 molecule acts as a physical bridge
between AR NTD and LBD in an agonist-activated receptor, thus aiding in the AR
inter-domain interactions that contribute greatly to receptor function. When
mutations were introduced into either of the GRIP1 regions that bind AR, this
bridging capacity was abolished (Fig 4.7). However, even in the absence of aiding
AR N/C interaction, GRIP1 still was able to enhance full-length receptor activity by
up to two fold. This demonstrated that mechanisms other than AR domain bridging
are components of GRIP1 coactivator function.
The family of p160 nuclear coactivators, consisting of SRC-1, GRIP1/TIF2,
and SRC-3/AIB1, recruit chromatin-modifying factors such as CBP/p300 and p/CAF
to sites of active transcription (Glass, Rose et al., 1997; Webb, Nguyen et al., 1998;
McKenna, Lanz et al., 1999). Indeed, our own data indicate that GRIP1 could boost
the ability of the lone GAL4-AR LBD peptide signaling on the GAL4-responsive
luciferase reporter. This is noteworthy because the AR AF-2 region contained within
the LBD is a very weak activation function and shows practically no transactivation
potential without coactivator assistance. It is also readily apparent that whereas
GRIP1 functions as a domain-bridging factor in some other receptors such as ER, in
others such as GR, the contribution of the coactivator most likely has nothing to do
with domain interactions at all (Spanjaard and Chin, 1993; Cheskis, McKenna et al.,
2003). This raises the question as to why the AR evolved to make use of GRIP1 as
an N/C domain bridging factor and whether this change may have contributed to the
94
AR AF-2 losing potency over evolutionary time. Evolutionary changes to the
AR NTD may have provided new structural mechanisms through which regulation of
the receptor could take place and thus this facilitated new roles for pre-existing AR
binding partners. As an example, such cofactor binding has been finely tuned over
time by modulations of the AR CAG trinucleotide repeat size (Alen, Claessens et al.,
1999; Irvine, Ma et al., 2000; Buchanan, Yang et al., 2004). The binding of GRIP1 to
the NTD of other steroid receptors often serves to activate an AF-1 that is otherwise
weak compared to AF-2 (Webb, Nguyen et al., 1998). However, in the case of AR
where the relative strengths of the AF-1 and AF-2 are reversed, perhaps there was
some other influence that selected for additional roles of GRIP1.

95
Figure 4.7 LEGEND

Hypothetical models of AR N/C interaction and GRIP1 bridging of AR domains as
observed using the mammalian two-hybrid assay system.
A-D) DHT-stimulated AR LBD recruits AR NTD FQNLF.
B) GRIP1 WT enhances interaction through domain bridging.
C) GRIP1 (1-1121) fails to bridge AR domains due to lack of AR NTD binding.
D) GRIP1 NR* fails to bridge AR domains due to lack of AR LBD binding, but still
enhances overall reporter signal via remaining binding to AR NTD.
E-H) DHT-stimulated AR LBD fails to recruit AR NTD FQNAA.
F) GRIP1 WT rescues impaired N/C interaction via bridging domain irrespective of
AR FQNAA mutation.
G-H) Neither GRIP1 (1-1121) nor GRIP1 NR* constructs can successfully bridge
due to each failing to bind either AR NTD or AR LBD domains.
96

97
The paradox of AR FQNLF competition with GRIP1 for binding of AF-2

It has been shown that the FQNLF motif in the AR NTD binds to AF-2 with greater
specificity and affinity than is demonstrated by the LxxLL NR-boxes of GRIP1,
suggesting AR N/C interaction inhibits p160 coactivator binding (He, Bowen et al.,
2001). This creates a counterintuitive model, since we know both processes are
integral to receptor activity. Recent work has suggested an elegant solution whereby
both binding events coexist but are temporally separated, and further suggests that
N/C interaction mediated by FQNLF binding may result in different molecular
conformations than that of GRIP1-mediated bridging (Bai, He et al., 2005).
Our group identified the melanoma antigen gene protein MAGE-11 as an AR
binding partner through screening of a human testis cDNA library, and subsequently
others have shown that the factor directly binds the AR NTD (Irvine and Coetzee,
1999; Bai, He et al., 2005). MAGE-11 binds to a 21-amino acid stretch of the AR
NTD that has the FQNLF framed at its center. Intriguingly the AF-2 of the AR LBD
binds to the middle 11-amino acids of the same sequence. This strongly suggests that
MAGE-11 binding at the site can completely block FQNLF from interacting with
AF-2, thereby leaving AF-2 available for interactions with other cofactors (Bai, He
et al., 2005). It is proposed that the FQNLF-mediated N/C interaction is an inhibitory
event that requires alleviation by coregulators such as MAGE-11 to allow for
p160-enhanced AR transactivation. In support of this, it has been shown that
MAGE-11 expression levels are higher in prostate cancer cell lines that also express
AR, such as LNCaP, C4-2, CWR-R1, and LAPC4, compared with those that do not,
98
such as PC3 and DU145. Also, MAGE-11 is expressed in androgen-regulated tissues
such as prostate and testis, whereas it is undetectable in liver and lung (Bai, He et al.,
2005).

The mechanistic significance of AR N/C interaction

Results from our current study suggest that approximately 70% of the transactivation
potential of wild-type AR on a transfected luciferase reporter is contributed by the
receptor undergoing N/C interaction. However, another research group has since
suggested that N/C interaction is an absolute requirement for AR regulation of
chromatin-integrated promoters (Li, Fu et al., 2006). Even more recently, the same
group has demonstrated that whereas a truncated AR lacking the LBD can
constitutively activate a luciferase reporter, this ability is lost on chromatin targets
(Li, Zhang et al., 2007). Those authors have shown that the presence of a
trans-expressed AR NTD can assist in the ability of a chromatin-bound AR LBD
fragment to drive transcription, suggesting that transactivation on chromatin requires
the interaction of both NTD and LBD domains in some form, even as separated
peptides.
The apparent prerequisite of N/C interaction for successful AR
transactivation on integrated reporters but not on transiently transfected ones
strongly suggests that a primary functional role of N/C interaction is to enhance the
recruitment of chromatin-modulating factors. It is plausible that even though
recruitment of the p160 coactivators can still take place at the AR AF-2 in the
99
absence of an AR NTD, the subsequent secondary recruitment of histone
acetylating/methylating factors may benefit from AR conformations achieved
through N/C interaction. In its absence, the stability of the complex formed by these
coactivators may be compromised. Another possibility is that primary recruitment of
other chromatin-modifying groups by AR, such as components of the SWI/SNF
complex, may be dependent on AR N/C for their binding signals. Lastly, the ability
of AR to shed corepressors, such as NCoR or SMRT that condense chromatin via
histone deacetylation, may be dependent on interference by N/C interaction. In any
of these scenarios, the absence of AR N/C interaction may have consequences on
transactivation that are much more apparent on chromatin-integrated reporters versus
ectopically introduced ones.

100
CHAPTER 5
Recruitment of the 26S proteasome by the AR Hinge domain

INTRODUCTION

Historically the Hinge domain has received less recognition as a functionally distinct
signaling region of the AR compared to the NTD, DBD, or LBD. In part this stems
from the tendency to group the Hinge with the C-terminal extension of the DBD or
with the downstream LBD in keeping with a three-domain AR model. However in
recent years, evidence points to the Hinge possessing several key functions that not
only distinguish it from other regions of the receptor but also indicate its absolute
necessity for normal AR signaling.

The nuclear localization signal

The Hinge spans amino acids 628-658 of the receptor. The nuclear localization
signal (NLS) of the AR is contained within amino acids 617-633 and consists of a
bipartite motif with short, basic arginine- and lysine-composed sequences from both
the DBD and Hinge separated by 10 intervening residues:
RKCYEAGMTLGARKLKK (Jenster, van der Korput et al., 1992; Zhou, Sar et al.,
1994). The specific mechanism by which this NLS targets the AR to the nucleus
remains unclear, but its ligand dependence has led to the hypothesis that functional
101
availability of this signal is conditioned upon conformational changes brought on by
receptor activation. (Zhou, Sar et al., 1994).

The Hinge contains binding sites for AR cofactors

There are several known AR cofactors that bind to the DBD-Hinge region. SNURF
(small nuclear RING finger protein) is a NR coactivator that is believed to traffic
hormone-activated AR into the nucleus and retard exiting of the receptor after ligand
dissociation (Poukka, Karvonen et al., 2000). Ubc9 is a homolog of the yeast
E2 class of ubiquitin conjugating enzymes and has been shown to function as an
E3 SUMO ligase towards AR, binding to the NLS portion of the Hinge and resulting
in sumoylation of the receptor (Poukka, Aarnisalo et al., 1999). Notably, disruption
of the NLS results in a loss of Ubc9 binding and subsequent attenuation of ligand-
dependent but not basal AR transactivation (Poukka, Aarnisalo et al., 1999).
The nuclear receptor corepressor SMRT (silencing mediator for retinoid and thyroid
hormone receptors) also recognizes the Hinge as an interaction site (Liao, Chen et
al., 2003). In contrast, NCoR, a highly related partner of SMRT, does not utilize the
Hinge directly for its interactions with AR (Wang, Lu et al., 2001). These
corepressors are thought to inhibit NR-regulated transcription via their recruitment of
histone deacetylation (HDAC) activities, and are found bound to unliganded steroid
receptors presumably as a mechanism for inhibiting improper receptor activity
(Jones and Shi, 2003).
102
AR is targeted for degradation via ubiquitination

The ubiquitin-proteasome pathway is an enzymatic cascade of events beginning with
a three step process whereby target substrates are flagged (Ciechanover and
Schwartz, 1998). In the first step, free ubiquitin is activated via the ATPase-
dependent functions of an ubiquitin-activating enzyme E1. In step two, the activated
ubiquitin is transferred to a ubiquitin-conjugating enzyme E2. The third step involves
the transfer of ubiquitin to an E3 ligase that specifically recognizes substrate proteins
and subsequently attaches the ubiquitin to target lysine residues. The ubiquitination
process is repeated, resulting in poly-ubiquitin chains that serve as recognition flags
for the cellular degradation machinery. This three-tiered system allows for a single
E1 enzyme to activate ubiquitin for several different E2 conjugating enzymes that in
turn pass the ubiquitin on to an even broader range of E3 ligases, ultimately
determining specificity. To date, there have been over 40 putative E2s and over
500 E3s identified in the human genome (Wong, Parlati et al., 2003). The known
E3 ubiquitin ligases fall into either the HECT (homologous to E6-associated protein
C-terminus) type or RING (really interesting new gene) type based on similarities of
their functional domains. In addition, a subclass of RING E3 ligases has been
characterized that involve a 75-amino acid domain known as the U-box that is
necessary for communication with E2 ubiquitin conjugating partners (Cyr, Hohfeld
et al., 2002).
There are currently two identified E3 ligases that target the AR for
ubiquitination. The first is Mdm2, which was originally characterized as an
103
E3 ubiquitin ligase for the tumor suppressor p53 (Honda, Tanaka et al., 1997). It has
since been shown that Mdm2 ubiquitinates AR as a crucial step in Akt
phosphorylation-dependent AR degradation (Lin, Wang et al., 2002). The second is
CHIP (carboxyl-terminus of Hsc70-interacting protein), a member of the U-box
containing RING-like E3 ubiquitin ligases (Connell, Ballinger et al., 2001; Jiang,
Ballinger et al., 2001; Murata, Minami et al., 2001). The involvement of CHIP in the
regulation of NRs first came to light when it was observed that CHIP alters the
steroid-mediated activity of GR through receptor degradation (Connell, Ballinger et
al., 2001; Jiang, Ballinger et al., 2001; Murata, Minami et al., 2001). CHIP has also
been implicated as an E3 ligase for ER (Reid, Hubner et al., 2003). It is now
recognized that CHIP interacts directly with the AR, for which it has ubiquitin ligase
activities that result in receptor turnover (He, Bai et al., 2004; Rees, Lee et al., 2006).
As with Mdm2 activity, there is evidence that this CHIP-AR communication is
associated with AR phosphorylation status (Rees, Lee et al., 2006).

The role of the 26S proteasome in AR regulation

The 26S proteasome is a large holoenzyme complex that functions as the major
cellular protease system responsible for ATP-dependent protein degradation
(Bochtler, Ditzel et al., 1999; Glickman and Ciechanover, 2002). The complex
consists of two main functional components, the 20S core and the 19S regulatory cap
(Fig 5.1-A). In eukaryotes, the 20S core consists of two copies each of fourteen
subunits that are grouped into α or β types. These subunits are arranged into four
104
rings made of seven units each, stacked to form a barrel with the β rings in the
middle and α rings on the ends (Baumeister, Walz et al., 1998). The 20S functions as
the proteolytic component of the proteasome, as the β subunits possess trypsin- and
chymotrypsin-like protease activities that degrade substrates within the barrel
chamber (Kopp, Hendil et al., 1997). The 20S core is flanked on either end by 19S
caps composed of eighteen distinct subunits each that form a complicated lid/base
structure (Gorbea, Taillandier et al., 1999). The 19S subunits utilize ATPase activity
to recognize proteins labeled for degradation and unfold them in preparation for
processing by the 20S core.
Historically the role of the proteasome has been considered that of a
“recycling bin” for processing unfolded or misfolded peptides labeled via the
ubiquitin E1-E2-E3 enzyme system. However, in recent years it has been shown that
different components of its machinery are involved in the regulation of gene
transcription (Swaffield, Melcher et al., 1995; Xu, Singer et al., 1995; Ferdous,
Gonzalez et al., 2001; Gonzalez, Delahodde et al., 2002). In the early 1990s,
researchers recognized that several of the 19S cap subunits have activities linked to
transcription, but their involvement was thought only to reflect cytosolic
proteasome-mediated turnover of transcription factors (Melcher and Johnston, 1995;
Chang, Gonzalez et al., 2001). However, through the adoption of chromatin
immunoprecipitation (ChIP) assays as a means of interrogating gene promoters, a
few of these 19S components have now been observed associated with active
transcription complexes on DNA (Fig 5.1-B). One of the first to be characterized in
this way was Rpt6/S8/Sug1, one of six highly-conserved ATPases of the 19S cap.
105
Rpt6 was originally identified as a modulator of the GAL4 C-terminal activation
domain (Swaffield, Bromberg et al., 1992). It has since been demonstrated that the
19S components Rpt6, Rpt4/Sug2, and Rpt5/TBP-1 are recruited to GAL1-10
promoters within 10 minutes of their activation (Gonzalez, Delahodde et al., 2002).
Notably, subunits of the 20S core proteasome have not been detected at these
promoters, suggesting that the 19S regulatory caps have a functional capacity even in
the absence of a complete 26S holoenzyme. Other 19S subunits that have been
observed recruited to active promoters include Rpt3/TBP-7 and Rpn2/S1 (Kinyamu,
Chen et al., 2005). The 19S has also been shown to modulate coactivator complexes
such as yeast SAGA (Spt-Ada-Gcn5-acetyltransferase), which, in addition to other
evidence, implicates direct proteasome involvement in the chromatin-modifying
steps of transcription by both non-steroid driven and steroid receptor-mediated
systems (Kinyamu and Archer, 2004; Lee, Ezhkova et al., 2005).
A particularly noteworthy observation is that the Rpt6/S8/Sug1 subunit of the
19S interacts with the AF-2 of several NRs such as vitamin D receptor (VDR),
retinoid X receptor (RXR), thyroid hormone receptor (TR), and ER (vom Baur,
Zechel et al., 1996; Masuyama and MacDonald, 1998; Parker, 1998; Gianni, Bauer
et al., 2002). This suggests that the 19S regulatory proteasome, or at least its
components, may be directly recruited by NRs in a ligand-dependent manner
concomitant with receptor stimulation.
It has been demonstrated that 26S proteasome involvement is essential for
AR, ER, and PR to achieve efficient transactivation on target genes. The
20S catalytic core has been observed engaged along the entire sequence of actively
106
transcribed ER-target genes, implicating the proteasome not only in initiation of
transcription but also with elongation (Zhang, Sun et al., 2006). This is different than
GAL4-driven systems where the 20S core is only seen at the 3’ end of the gene,
apparently waiting to complex with the 19S regulatory unit that can be found
associated with RNA Pol II throughout the gene (Gillette, Gonzalez et al., 2004). In
the case of PR, it has been shown that this proteasome-RNA Pol II association is
vital for the basal transcription complex to properly form on target promoters
(Dennis, Lonard et al., 2005). There is mounting evidence that the proteasome is
involved not only in the turnover of NRs, but with the cyclic turnover of coactivators
as well (Lonard, Nawaz et al., 2000; Reid, Hubner et al., 2003). This is the case with
GR signaling during which the contributions of coactivators such as CBP/p300 are
regulated by their proteasome-mediated degradation (Li, Su et al., 2002). Inhibition
of proteasome activity results in a marked increase in GR activity and alters the
receptor’s intracellular trafficking (Deroo, Rentsch et al., 2002).
In recent years it has become clear that the AR is dependent on proteasomal
involvement for proper transactivation. The proteasome influences AR function both
in the cytoplasm as well as in the nucleus. It has been suggested that the proteasome
may play a role in aiding the nuclear translocation of AR upon ligand activation,
perhaps by disrupting the interaction of AR with coregulators that maintain the
receptor in the cytoplasm (Lin, Altuwaijri et al., 2002). In the nucleus, an important
role of the proteasome is to control the cyclic loading and removal of AR at target
gene promoters; it has been observed that inhibition of proteasome function results in
increased AR occupancy on promoters, which suggests that the receptor becomes
107
immobilized on DNA. However, this AR build-up surprisingly results in attenuated
mRNA synthesis from the gene, supporting the hypothesis that proper receptor
cycling at the promoter is necessary for efficient transcription (Kang, Pirskanen et
al., 2002).

108
Figure 5.1 LEGEND

Schematic model of the 26S proteasome.
A) The 26S holoenzyme complex consists of the 20S catalytic core flanked by
19S regulatory caps on either end. Poly-ubiquinated target substrates are unfolded by
the 19S cap, which releases the ubiquitin. The unfolded protein is fed into the barrel-
like 20S core where proteolytic activity cleaves it into small peptides that are
released and subsequently recycled.
B) Expanded diagram of the 19S regulatory cap revealing the identified subunits
comprising the lid and base. The subunit Rpt6/S8/Sug1 has been shown to directly
bind to a number of transcription factors, several of which are listed. Of particular
interest are the nuclear receptors, which interact with Rpt6 through their AF-2
activation functions.
109

110
The AR Hinge contains a putative PEST sequence

The mechanisms behind AR communication with the 26S proteasome on target DNA
remain unclear. In the mid 1980s, researchers interested in the underlying factors
governing cellular protein turnover rates observed that many rapidly degraded
proteins contain short peptide regions enriched in the amino acids proline (P),
glutamate (E), serine (S), and threonine (T) (Rogers, Wells et al., 1986). Although it
is likely no consensus sequence exists, these enriched regions have been identified in
many regulatory proteins known to have short cellular half-lives such as Myc, Jun,
Fos, and p53 (Rechsteiner and Rogers, 1996). When these PEST regions are
mutated, the resulting proteins often exhibit dramatically altered stabilities
(Rechsteiner, 1990). It is now believed that the predominant pathway through which
PEST sequences signal protein degradation is the 26S proteasome (Rechsteiner and
Rogers, 1996).
Based on output from a software algorithm designed to identify putative
PEST sequences in candidate proteins, a likely PEST region has been identified in
the AR Hinge (Sheflin, Keegan et al., 2000). This 21-amino acid stretch in human
AR consists of
638
KLQEEGEASSTTSPTEETTQK
658
and is highly conserved across
mammalian species (Fig 5.2-A). It scores highly as a PEST sequence when
compared with other identified sequences from known PEST-regulated proteins.
To date it remains unclear just how these PEST sequences function in protein
degradation, but there is evidence that these signals may be targets for kinase
activities that result in phosphorylation-dependent conformational changes.
111
These changes recruit the activities of the ubiquitin-proteasome pathway either
through direct interactions with subunits of the 26S proteasome or through
recognition by E3 ubiquitin ligases (Rechsteiner and Rogers, 1996; Sheflin, Keegan
et al., 2000).
A recent study that set out to characterize the role of the AR Hinge concluded
that the Hinge plays an inhibitory role on AR activity by virtue of its PEST
sequence, which leads to receptor degradation and subsequent attenuation of
transactivation (Tanner, Claessens et al., 2004). In that study, partial Hinge deletions
resulted in a multifold increase in overall receptor activity on a transiently
transfected reporter. It is notable, though, that the deletions created by that group
involved a 19-amino acid stretch from aa 628-646 that constituted only the proximal
half of the documented PEST sequence and included deletion of the upstream
RKLKK signal believed to be part of the receptor’s NLS.
In our current study, we wished to address the functional role of the Hinge in
communications between the AR and 26S proteasome machinery. To accomplish
this, deletions of the Hinge were created with attention towards leaving the NLS
portion of the domain intact while altering regions spanning the entire documented
PEST sequence. We simultaneously modulated cell-wide 26S proteasome function
using the proteasome inhibitor drug MG132 to determine whether any observed
impact on AR transactivation could be related directly to 26S proteasome
involvement.

112
MATERIALS AND METHODS

Site-directed mutagenesis and plasmid preparation

Mutations in AR expression plasmids were constructed utilizing the protocol
published in the Stratagene Quick-Change Site-Directed Mutagenesis Kit
(Stratagene). The plasmid cmv-AR WT expresses wild-type AR aa 1-919 under
control of the constitutively active CMV promoter. Using this plasmid as a template,
several deletion constructs targeting the Hinge region were created. The construct
cmv-AR ΔPEST expresses AR with Hinge residues 638-658 deleted, removing the
entire documented PEST sequence. Two additional constructs were made as subsets
of this: cmv-AR Δ638-646 has the proximal half of the PEST sequence deleted and
cmv-AR Δ647-658 harbors a deletion of the distal half (Fig 5.2-B). To explore the
effect of the Hinge region on AR N/C interaction, the construct GAL4-AR
Hinge/LBD was created that expresses AR aa 628-919 (i.e., the entire Hinge region
and LBD) fused in frame to the GAL4 DNA-binding domain. This construct was
used in conjunction with the GAL4-AR LBD and VP16-AR NTD constructs
previously created for use in mammalian two-hybrid assays (Fig 5.2-C).

113
Figure 5.2 LEGEND

The AR Hinge domain.
A) Map of the AR domains showing coordinates of the Hinge boundaries
aa 628-658. The expanded detail of the Hinge amino acid sequence shows the
location of the NLS bordering the upstream DNA-binding domain C-terminal
extension (DBD CTE) as well as residues downstream of the Hinge leading into the
ligand-binding domain. The documented PEST sequence aa 638-658 are indicated.
B) Residues of the PEST sequence that were deleted to construct the AR ΔPEST,
Δ638-646, and Δ647-658 receptors used in transactivation assays.
C) Schematic maps of the GAL4 DBD fusion peptides with AR LBD (aa 644-919)
and Hinge/LBD (aa 628-919) regions for use in both N/C domain interaction assays
as well as mammalian one-hybrid AF-2 transactivation assays.

114

115
Cell culture and transfection

COS7 cells (ATCC) and HeLa cells with stably-integrated GAL4-luc reporter (gift of
Michael Stallcup’s laboratory, USC) were maintained in DMEM hi-glucose media
(USC/Norris) that was supplemented with 5% (v/v) heat-inactivated fetal bovine
serum (FBS). Cells were plated in 96-well plates (3-mm wells) at a density of
1x10
4
cells/well and incubated for 24 hours in DMEM. Subsequent transient
transfection of plasmids was conducted using LipofectAMINE 2000 reagent
(Invitrogen) as per manufacturer’s suggested protocol. After 3 hours transfection,
media was replaced with fresh DMEM and cells were allowed to rest overnight. The
next day, media was changed to DMEM supplemented with 5 µM MG132
proteasome inhibitor (Calbiochem, EMD Biosciences, Inc., La Jolla, CA), or DMSO
vehicle. Cells were treated for 2 hours with MG132. Media was then changed to
DMEM supplemented with 10 µM DHT or EtOH vehicle and cells were allowed to
incubate for 24 hours. After DHT treatment, cells were lysed with Passive Lysis
Buffer (Promega) and extracts assayed for luciferase activity using Promega
Luciferase Assay System measured on a MLX Microtiter Plate Luminometer
(Dynex).

Chromatin Immunoprecipitation (ChIP) assay

LNCaP cells (ATCC) were grown in RPMI 1640 media (USC/Norris) supplemented
with 5% FBS. Cells were seeded in 150 mm dishes at a density of 6x10
6
cells/dish in
116
30 ml RPMI 1640 supplemented with 5% charcoal dextran-stripped (CSS) FBS and
allowed to grow for 3 days. Media was changed to RPMI 1640 5% CSS containing
5 µM MG132 (Calbiochem) or DMSO vehicle and allowed to incubate for 2 hours.
Media was then changed to RPMI 1640 5% CSS containing 10 µM DHT or EtOH
vehicle and allowed to incubate for an additional 2 hours.
Media was aspirated off and cells were treated for 10 minutes at room
temperature with 20 ml 1% formaldehyde (Sigma, St. Louis, MO) to cross-link
protein/DNA complexes. Cells were washed with ice-cold PBS and harvested in ice-
cold PBS + 1x protease inhibitor cocktail (Sigma) into 1.5 ml microcentrifuge tubes.
Cells were pelleted and resuspended in SDS Lysis Buffer (1% SDS, 10 mM EDTA,
50 mM Tris-HCl pH 8.1, 2x protease inhibitor cocktail). Cell lysate was sonicated to
shear chromatin to fragments approximately 200 bp - 1 kb in size. From this
solubilized chromatin, 100 µl aliquots were dispensed per IP into 1.5 ml tubes.
To each sample was added 900 µl Dilution Buffer (0.01% SDS, 1.1% Triton X 100,
1.2 mM EDTA pH 8.1, 16.7 mM Tris-HCl pH 8.1, 167 mM NaCl, 1x protease
inhibitor cocktail). To each sample 40 µl Protein G Sepharose bead slurry
(Amersham Pharmacia) and 2 µg sheared salmon sperm DNA (Gibco) were added.
Sample tubes were put on 4ºC rotator for 1 hour to pre-clear. Beads were pelleted
and the supernatant transferred to a new 1.5 ml tube. Various antibodies were added
for IP: AR N-20 (Santa Cruz Biotechnology, Inc., Santa Cruz, CA #sc-816),
Normal Rabbit IgG (Santa Cruz #sc-2027), 19S proteasome S1 (Abcam Inc.,
Cambridge, MA #ab-2941). Samples were allowed to immunoprecipitate overnight
117
on a 4ºC rotator. The next morning 40 µl Protein G Sepharose slurry and
2 µg sheared salmon sperm DNA were added to each sample and tubes were allowed
to incubate for 1 hour on a 4ºC rotator. Beads were pelleted and supernatant carefully
discarded. Beads were then washed for 5 minutes on a 4ºC rotator with 1 ml each of
the following, followed each time by gentle pelleting of beads: Low Salt Buffer
(0.1% SDS, 1% Triton X 100, 2 mM EDTA, 20 mM Tris-HCl, 150 mM NaCl),
High Salt Buffer (0.1% SDS, 1% Triton X 100, 2 mM EDTA, 20 mM Tris-HCl,
500 mM NaCl), LiCl Buffer (0.25 M LiCl, 1% Igepal CA-630, 1% Deoxycholate,
1 mM EDTA, 10 mM Tris-HCl), two washes of 1x TE (10 mM Tris-HCl,
1 mM EDTA). After washes, beads were resuspended twice on a rotator at room
temperature in freshly prepared Elution Buffer (1% SDS, 0.1 M NaHCO
3
). To each
tube, 20 µl 5 M NaCl was added. Tubes were incubated at 65ºC overnight to reverse
cross-linking. The next morning, 10 µl 0.5 M EDTA, 20 µl 1 M Tris-HCl pH 6.5,
and 2 µl 10 mg/ml Proteinase K were added to each tube. Tubes were incubated at
45ºC for 1 hour. DNA was subsequently recovered by phenol/chloroform extraction
and ethanol precipitation using 2 µl glycogen (20 mg/ml) as carrier. Air-dried DNA
was resuspended in 100 µl water for use in PCR analysis. Real-time PCR was
performed using AmpliTaq Gold PCR Master Mix (Applied Biosystems,
Branchburg, NJ) run on an MJ Research OPTICON DNA Engine PCR machine (MJ
Research Inc., Waltham, MA). An overview of the steps involved in the ChIP assay
is depicted in Figure 5.3.
118
Figure 5.3 LEGEND

Overview of steps involved in the ChIP assay.
1) LNCap cells were treated with 5 µM MG132 proteasome inhibitor or DMSO
vehicle for 2 hours, followed by 10 nM DHT or EtOH vehicle treatment for 2 hours
to stimulate AR. Cells were then exposed to 1% formaldehyde to cross-link
protein/protein and protein/DNA complexes.
2) Cells were lysed and chromatin containing AR bound to target sites (i.e. AREs)
was isolated and purified.
3) Chromatin was sheared through sonication to fragment sizes approximately
200 bp - 1 kb.
4) Immunoprecipitation was performed using antibodies against AR and proteasome
conjugated to Protein G Sepharose beads. Beads were washed, and DNA eluted and
purified as described in chapter text.
5) Quantitative real-time PCR was performed on the purified ChIP DNA using
primers and probes designed to interrogate the PSA gene locus. Illustration depicts
PCR results showing DHT-dependent AR occupancy of target DNA.
119

120
RESULTS

The AR Hinge inhibits N/C domain interactions

Through mammalian two-hybrid assays comparing the DHT-induced interaction
between the VP16-AR NTD peptide and GAL4-AR LBD or GAL4-AR Hinge/LBD
peptides, we observed that the presence of the Hinge domain led to 95% inhibition of
the N/C interaction signal. Co-expression of the p160 coactivator GRIP1, which we
have previously shown enhances AR N/C interactions through domain bridging, had
little restorative effect on the inhibited interaction (Fig 5.4-A). This loss of signal
could not be explained as an artifact of protein degradation, as Western analysis
revealed that the GAL4-AR LBD and GAL4-AR Hinge/LBD fragments were
equally detectible using antibodies directed either at the AR LBD or the GAL4 DBD
(Fig 5.4-B).
When we assayed for the ability of the AR LBD AF-2 activation function to
transactivate with or without the Hinge present, mammalian one-hybrid assays
demonstrated that whereas the inclusion of the Hinge seemed to enhance the
DHT-induced activity of AF-2 by approximately three fold, the overall signal levels
still remained only marginally above background (Fig 5.5). This finding is consistent
with the AR possessing an intrinsically weak AF-2. The co-expression of GRIP1
enhanced the ability of DHT-induced GAL4-AR LBD to transactivate by
approximately two hundred fifty fold, reflecting the recruitment of GRIP1 to AF-2
and the subsequent contribution of the p160 coactivator to reporter transcription.
121
Again, however, the inclusion of the Hinge region was strongly inhibitory, as GRIP1
only enhanced the activity of the DHT-induced GAL4-AR Hinge/LBD fragment to
levels 5% of what was observed when GRIP1 aided the LBD without the Hinge
present. We next addressed whether these results were applicable to AR-target
promoters in a chromatin context. When the experiment was repeated using HeLa
cells that had stably-integrated GAL4-responsive luciferase reporter, the results were
similar. This demonstrated that the inhibitory effect of the AR Hinge on AF-2
transactivation was not limited to AR activity on transiently transfected promoters,
but is applicable to promoters that are complexed in chromatin as well.

122
Figure 5.4 LEGEND

Mammalian two-hybrid data demonstrating impact of AR Hinge on N/C domain
interactions.
A) COS7 cells were transfected as per protocols described in chapter text.
VP16-AR NTD = fusion peptide of VP16 activation domain with AR transactivation
domain aa 1-538. GAL4-AR LBD = fusion peptide of GAL4 DNA-binding domain
(DBD) with AR ligand-binding domain (LBD) aa 644-919.
GAL4-AR H/LBD = fusion of GAL4 DBD with AR Hinge and LBD aa 628-919.
Luciferase reporter activity is indicative of DHT-dependent interaction of AR NTD
with LBD, which is enhanced by co-expression of GRIP1 as described in Chapter 4.
Presence of the Hinge greatly inhibits N/C interaction, which largely cannot be
rescued by GRIP1 bridging of domains.
B) Western immunoblot data showing relative levels of the GAL4-AR LBD and
GAL4-AR H/LBD peptides as detected by antibodies against AR C-terminal and
GAL4 DBD. Expected size of GAL4-AR LBD = 46 kDa. Expected size of
GAL4-AR H/LBD = 48 kDa. Treatment with 5 µM MG132 proteasome inhibitor
stabilizes the GAL4-AR constructs.
123

124
Figure 5.5 LEGEND

Presence of the AR Hinge impacts AF-2 transactivation. COS7 cells were transfected
as described in chapter text. DHT-stimulation of GAL4-AR LBD results in little
plasmid reporter activity, consistent with AR possessing a weak AF-2.
Co-expression of GRIP1 substantially enhances DHT-driven AR AF-2 activity.
Inclusion of the AR Hinge domain has inhibitory effect on the ability of GRIP1 to
coactivate AR AF-2. In HeLa cells with stably-integrated GAL4-luc reporter, the
fusion protein GAL4-VP16AD that contains both the GAL4 DBD and VP16 AD
constitutively drives reporter expression as an experimental control. On this
stably-integrated reporter the presence of the AR Hinge inhibits AF-2 in a manner
similar to results from transiently transfected reporter.
125

126
Inhibition of proteasome function results in loss of AR transactivation

Treatment with MG132 proteasome inhibitor resulted in pronounced attenuation of
the DHT-induced activity of AR on several androgen-responsive luciferase reporters
(Fig 5.6). On MMTV promoter, activity of wild-type AR was decreased by
approximately 25% in the presence of the drug, coinciding with a marked increase in
basal activity. However, on the PSA and Probasin promoters, both of which
demonstrate higher specificity for AR than does MMTV, proteasome inhibition
attenuated AR transactivation >90% compared to when only DMSO vehicle was
present. This effect of MG132 was not a result of cellular toxicity, as within the same
experiment dexamethasone-induced GR signaling on MMTV promoter remained
unaffected by proteasome inhibition. Western immunoblot assays determined that
neither DHT nor MG132 treatment substantially affected AR steady-state expression
levels.

127
Figure 5.6

AR transactivation on various promoters is dependent on proteasome function. COS7
cells were transfected as described in chapter text. Transactivation of wild-type AR
was assessed on MMTV, PSA, and Probasin promoter-driven luciferase reporters.
All reporters exhibit a high degree of DHT dependence for AR transactivation.
Inhibition of proteasome by MG132 treatment results in attenuated AR activity.
Dexamethasone-driven GR transactivation on MMTV-driven luciferase reporter in
the same cells is unaffected by proteasome inhibition, demonstrating that MG132
effect on AR activity is not an artifact of global inhibition of transcription or drug
toxicity. Western immunoblot analysis reveals MG132 results in minimally
increased AR stabilization.
128

129
Deletions of the AR Hinge radically alter receptor activity

We next examined the impact of various Hinge deletions on overall receptor
transactivation (Fig 5.7). The AR ΔPEST construct, which has a deletion of the
entire documented PEST sequence aa 638-658, retains only 20% of wild-type
activity on the MMTV-luc reporter. Likewise, this loss in signaling is replicated in
the attenuated activity of either the AR Δ638-646 construct that has the proximal half
of the PEST sequence deleted, or the AR Δ647-658 construct that has the distal half
deleted. Therefore, the presence of both halves of the PEST sequence appears
necessary for the Hinge region to contribute normally to AR signaling. Western
analysis indicated no substantial variation in the stabilities of these proteins brought
on by the deletions.
Concomitant inhibition of cellular proteasome function by MG132 resulted in
highly variable activation profiles among the different AR Hinge constructs. In the
presence of MG132, DHT-induced wild-type AR transactivation on MMTV-luc once
again was lowered by approximately 25% compared to DMSO control. The AR
ΔPEST mutant that lacked the entire PEST sequence aa 638-658 exhibited a
remarkable response whereby proteasome inhibition alone led to a substantial
increase in basal (non DHT) signaling to levels nearly 50% greater than those
achieved by DHT stimulation when proteasome function was intact. In fact,
subsequent addition of DHT to stimulate this receptor resulted in no change to the
already elevated basal levels. When we examined the response to proteasome
130
inhibition of the AR Δ638-646 receptor, the striking increase in basal activity
observed of the AR ΔPEST construct was replicated. However, exposure of this
AR Δ638-646 receptor to DHT in the presence of MG132 successfully induced a
signal increase of at least two fold over its elevated basal activity, an observation that
contrasted with the complete loss of agonist induction exhibited by AR ΔPEST. With
regard to the AR Δ647-658 receptor, which lacked the distal half of the PEST
sequence, inhibition of the proteasome had no substantial impact on either the basal
or already-attenuated DHT-induced activity.

131
Figure 5.7 LEGEND

Transactivation of AR Hinge-deletion constructs. COS7 cells were transfected with
wild-type AR (AR WT), AR with deletion of the entire documented PEST sequence
(AR ΔPEST), AR with deletion of the proximal half of the PEST sequence
(AR Δ638-646), or AR with deletion of the distal half of the PEST sequence
(AR Δ647-658) along with MMTV-luc reporter. Each of the PEST deletions results
in a receptor that only achieves approximately 20% of wild-type response to DHT,
suggesting that the entire PEST sequence is required for normal transactivation.
Inhibition of proteasome function by MG132 results in a substantial elevation in
basal (non DHT) transactivation by the ΔPEST construct that is replicated by
deletion of the proximal half of the PEST alone (aa 638-646). The Δ638-646
receptor, but not the ΔPEST receptor, could be stimulated further by DHT. This
suggests that normal agonist-induced transactivation requires presence of the distal
half of the PEST for proper signaling. The receptor lacking only the distal half,
AR Δ647-658, shows heavy attenuation of signaling that is unresponsive to
modulations of proteasome function by MG132. Western immunoblot analysis
reveals little fluctuations in AR protein stability caused by Hinge deletions.
132

133
Proteasome recruitment to the PSA gene is DHT responsive

Next we wished to determine if the apparent communications between AR and
the 26S proteasome observed on transiently transfected luciferase reporters held
physiological relevance in the chromatin context of endogenous AR-target genes
(Fig 5.8-A). ChIP analyses demonstrated that after 2 hour DHT treatment the AR
could be detected occupying the -4 kb upstream AREIII enhancer region (A) and
-170 bp ARE I promoter region (C) of the PSA gene locus (coordinates relative to
transcription start site). DHT-induced AR occupancy of a -2 kb non-ARE containing
intermediate site (B) was only marginally above background (Fig 5.8-B). Similarly,
DHT-dependent occupancy of the proteasome was observed at locations (A) and (C),
suggesting recruitment of proteasome components to the PSA gene by ligand-
activated AR. Exposure to MG132 did not substantially affect DHT-induced AR
occupancy on the PSA gene at the interrogated sites. However, MG132 treatment
resulted in dramatically reduced detection of proteasome recruitment to the same
chromatin sites, suggestive of a drug-induced failure of communication between AR
and the 26S proteasome machinery. The loss of DHT-induced proteasome
recruitment to the PSA gene in response to MG132 treatment occurs concomitantly
with the loss of PSA gene transcription as determined by real-time RT-PCR analysis
(Fig 5.8-C). This data thus replicates in the context of an endogenous gene the
effects of proteasome inhibition on AR-mediated transcription that were observed in
the transient transfection reporter systems.
134
Figure 5.8 LEGEND

ChIP data revealing occupancy of AR and proteasome on the endogenous PSA gene
locus.
A) Map of the upstream regulatory region of the PSA gene showing locations of the
PCR primer/probe sets used to interrogate occupancy at the -4 kb enhancer
(ARE III), -170 bp promoter (ARE I), and intervening -2 kb non-AR binding control
region of DNA.
B) ChIP results indicating DHT-dependent occupancy of AR at the enhancer and
promoter regions of PSA. Inhibition of proteasome by MG132 does not dramatically
alter AR occupancy at these sites. The 19S proteasome occupies the PSA enhancer
and promoter in response to DHT treatment, implying its recruitment to the PSA
gene by AR. However, MG132 treatment results in a substantial loss of detectible
proteasome at the sites.
C) Real-time mRNA expression analysis demonstrates that AR-mediated expression
of the PSA gene is greatly attenuated in response to MG132 treatment. Together the
data suggest that loss of AR-mediated PSA expression is not a function of lost AR
binding, but rather of failed proteasome recruitment necessary for proper receptor
function.
135

136
DISCUSSSION

Proteasome inhibition by MG132

It is well accepted that cellular protein levels of AR are regulated by the ubiquitin-
proteasome pathway; the influence of the 26S proteasome on AR degradation can be
observed through the use of commercially available proteasome inhibitor drugs such
as MG132, lactacystin, and acetyl-ubiquitin (Fenteany and Schreiber, 1998).
Whereas the latter two are commonly used for in vitro biochemical assays, MG132 is
a short peptide drug that is readily internalized by cell membranes and thus is highly
effective for use in tissue culture experiments. Others have demonstrated through
immunoblot assays that upon treatment with MG132, cellular buildup of
poly-ubiquitinated AR moieties can be observed (Lin, Wang et al., 2002). This is
believed to occur as a result of MG132 disabling the β subunits of the 20S catalytic
core, resulting in 26S proteasome complexes effectively becoming choked with
partially-catalyzed substrate and thus incapacitated (Zhu, Wani et al., 2005).
Notably, our own Western blot data indicate that although MG132 treatment on
COS7 cells led to a minute increase in AR levels, the effects were not consistent with
substantial interference with AR degradation. We believe that this may be a
consequence of the drug treatment protocol. Comparisons of published NR studies
reveal that the concentration of MG132 used in cell culture for proteasome inhibition
typically ranges from 1 µM - 20 µM. Likewise, the exposure time of cells to the drug
ranges from a brief 30 minute treatment prior to ligand stimulation in some protocols
137
to as long as 24 hour concomitant exposure of drug with ligand in others. It is
apparent that cell lines exhibit varying sensitivities to MG132 and although there is
an understanding that cells can survive up to 24 hours with complete proteasome
inhibition, our pilot toxicity studies indicate that for COS7 cells a 2 hour treatment
with 5 µM MG132 prior to removal of drug and replacement with ligand is suitable
for pronounced effects on AR transactivation. It remains to be seen if modifying the
MG132 treatment protocol could result in more obvious effects on cellular AR
stability, but our current protocol is beneficial in that the lack of pronounced
variation in AR levels aids our ability to distinguish true transactivation effects from
those of protein stabilization.

The Hinge as a split communication line to the proteasome

In the current study we have demonstrated that the AR Hinge, and more specifically
aa 638-658 that constitute the recognized PEST sequence, influences AR
transactivation through recruitment of the 26S proteasome. We observed that an AR
construct lacking the entire PEST sequence only maintained 20% of wild-type
activity under normal cellular conditions, but surprisingly, when proteasome function
was inhibited by MG132, the basal activity of this receptor was enhanced nearly
five fold. We have dismissed the idea that this phenomenon is simply an effect of
inhibited proteasome-mediated turnover, as we observed that there was only a
minimal increase in AR protein levels. This strongly suggests that recognition of the
AR PEST sequence by the proteasome aids in maintaining low basal receptor
138
activity on promoters. Remarkably, DHT-stimulation of this AR ΔPEST receptor in
the presence of MG132 failed to drive reporter signal any higher than the already
elevated basal levels, indicating that the absence of proteasome function not only
increased basal activity, but simultaneously crippled agonist-induced transactivation.
These observations suggest a scenario in which the Hinge apparently has dual
communication lines with the proteasome; one mediates basal repression of the
receptor and another is necessary for proper ligand-stimulated activity. We further
scrutinized the PEST sequence for evidence of these two disparate signals. Among
the repertoire of chaperone proteins that complex with the inactive AR in the
cytoplasm, the immunophilin co-chaperone αSGT (small glutamine-rich
tetratricopeptide-containing protein) binds the AR through use of its TPR domain
that recognizes the residues
640
QEEGE
644
in the AR Hinge (Dr. Grant Buchanan,
personal communication). The immunophilins have been shown to assist in the
proper maturation and folding of nuclear receptors through direct associations with
Hsc70, Hop, and Hip, thus implicating αSGT as having an influence on how the heat
shock protein complex directs misfolded peptides into the ubiquitin-proteasome
pathway. In preliminary unpublished work, our group has shown that overexpression
of αSGT results in increased stabilization of AR in the cytoplasm, presumably by
sequestration to chaperone complexes that deter the AR from either 1) unnecessarily
entering into the ubiquitin-proteasome pathway for degradation or 2) improperly
translocating to the nucleus in an unliganded state. Knowledge of the functional
effects of αSGT leads us to believe that at least one of the underlying causes of
disrupted proteasome communication observed in our current study is deletion of
139
this αSGT recognition sequence
640
QEEGE
644
found in the proximal half of the
PEST. We believe that loss of this αSGT-AR association alone may not be sufficient
to drive improper AR signaling, as it is plausible that disruptions in the proper
chaperone regulation of AR may result in a compensatory elevation in ubiquitin-
proteasome degradation. However, in the absence of proteasome function brought on
by MG132, unliganded AR that is free from αSGT and proper chaperone influence
may be improperly transported to the nucleus en masse resulting in a marked
increase in basal transactivation. In support of this conclusion, the elevated basal
activity of the ΔPEST receptor can be replicated by the Δ638-646 deletion that
removes only the proximal half of the PEST containing the αSGT recognition site.
However, a notable difference between the ΔPEST and the Δ638-646 receptors is
that whereas in the presence of MG132 the former failed to respond to agonist
stimulation, the latter was clearly inducible by addition of DHT. This leads us to
believe that while the QEEGE motif in the PEST proximal half is one signal that
influences AR communication with the proteasome, there must be another in the
distal half necessary for the proteasome to mediate proper agonist-induced
transactivation. Upon closer examination of the residues comprising the distal half of
the PEST sequence, we became aware that this stretch contains the motif
650
SPTEETT
656
that has been shown through peptide array analysis to be an AR
binding site for CHIP, one of the E3 ubiquitin ligases for AR. It has been shown that
CHIP strongly binds to this peptide stretch dependent on phosphorylation of S650,
T655, and T656 (Rees, Lee et al., 2006). We propose that it is through this distal half
of the PEST that the AR is able to recruit, via CHIP E3 ubiquitin ligase activity, the
140
proteasome machinery required for agonist-induced receptor transactivation on target
promoters. As evidence to support this, the AR Δ647-658 receptor that lacks this
SPTEETT recognition site for CHIP has a greatly attenuated DHT-stimulated
activity that does not fluctuate in the presence or absence of MG132, consistent with
a receptor that has lost the ability to recruit proteasome for transactivation and thus
remains naïve to any subsequent modulations of cellular proteasome function.
Based on the complementary data obtained from the AR Δ638-646 and
AR Δ647-658 receptors that separately represent deletions of the proximal and distal
halves of the documented PEST sequence, we conclude that the role of the PEST
sequence in the AR Hinge is to control communications between AR and the
26S proteasome machinery. It is apparent that there are two disparate signaling
motifs present in the PEST, one in each half we examined. Whereas they both
influence recognition of AR by the ubiquitin-proteasome system, they differentially
recruit distinct roles of proteasome involvement in AR signaling. On one hand, the
QEEGE motif found in the proximal half of the PEST sequence mediates
associations between AR and chaperone complexes in the cytoplasm, thereby
stabilizing and protecting AR from proteasomal degradation. On the other, the
SPTEETT motif found in the distal half of the PEST sequence may play a crucial
role in the nucleus, where it recruits the proteasome for AR cycling on chromatin and
thus promotes efficient transactivation. Together these two signaling motifs within
the AR Hinge create a balanced and wonderfully choreographed model in which the
seemingly disparate faces of the 26S proteasome, as both an AR degradation
machine and as a promoter of AR activity, can be reconciled.
141
One noteworthy feature of the two-motif AR PEST model presented here is
that the QEEGE and SPTEETT sequences are separated in the Hinge by only 5
intervening residues. An intriguing implication from this is that not only may these
two motifs signal different pathways of proteasome involvement, but their proximity
may effectively create a situation in which recruitment of their cognate AR cofactors,
i.e., αSGT to the proximal PEST half and CHIP to the distal, are mutually exclusive.
In the current study we have not examined this, yet based on our functional
understanding of these AR cofactors it seems likely that their associations with the
receptor define an exchange of binding partners rather than a synergistic complex
involving both simultaneously. In recent years it has become recognized that
transcription factors do not simply discard associations with their chaperone
complexes upon translocation from cytoplasm to nucleus, but rather, undergo an
exchange of factors (such as Hsp90 and immunophilins) in the nucleus that aid in
their proper cycling to target chromatin (Pratt, Galigniana et al., 2004). It is likely
then, that the disparate stabilization and degradation signals that neighbor one
another in the AR Hinge play a role in this cofactor exchange undertaken by AR as it
is shuttled from the cytoplasm to the nucleus. Knowledge of the bipartite signaling
character of the AR PEST sequence brings us a step closer to understanding the
different paths through which AR communicates with the proteasome. It is
tantalizing to view the AR Hinge as a potential target for novel drugs aimed at
disrupting this communication with the goal of inhibiting the aberrant AR activity
experienced in diseases such as hormone-refractory prostate cancer.

142
RESEARCH SUMMARY

The projects comprising this Ph.D. thesis focused on examining various signaling
motifs within the AR through site-directed mutagenesis and functional assays that
assessed impact on receptor conformation, cofactor recruitment, and ultimately
transactivation.
It was shown that disruption of the
23
FQNLF
27
motif located in the proximal
region of the AR NTD resulted in a 70% loss of receptor activity. This effect
occurred concomitantly with a complete loss of AR ability to engage in the N/C
domain interactions thought to be integral to normal AR function. However, the
presence of GRIP1, a member of the p160 family of NR coactivators, restored the
N/C interaction. Further examination revealed that one of the mechanisms by which
GRIP1 enhances AR activity is through physically bridging the AR domains, thus
encouraging a conformation of the AR favoring transactivation. Intriguingly, the
region of the AR NTD involved in this GRIP1 bridging did not involve the
23
FQNLF
27
motif, a finding that distinguishes N/C interaction mediated by the
receptor’s own NTD sequence from that achieved by coactivator assistance. This
distinction advances our understanding of how modulations of the N/C interaction
may be achieved through the functional disruption of GRIP1 and, potentially, other
binding partners.
In a parallel study, it was revealed that the AR Hinge domain mediates
communications between the AR and the 26S proteasome. Specifically,
AR recruitment of the ubiquitin-proteasome pathway was shown to be regulated by a
143
putative PEST sequence present in receptor amino acids 638-658. Deletion of this
PEST resulted in an AR that achieved only 20% of wild-type agonist-induced
activity. It was shown that the two halves of this PEST direct different roles of the
proteasome to the AR. The proximal half of the PEST contains a
640
QEEGE
644
motif
that communicates with the ubiquitin-proteasome pathway through complexing AR
with cytoplasmic chaperone proteins that stabilize the receptor and repress basal
(i.e., non DHT-induced) activity. When this motif was disrupted, the receptor
exhibited a five fold elevation in basal transactivation on target promoters in
response to MG132 treatment. The distal half of the PEST contains an
650
SPTEETT
656
motif that recruits ubiquitin E3 ligase activities resulting in AR
degradation. An AR lacking this motif exhibited attenuated agonist-induced
transactivation similar to that of the other PEST deletions, yet was non-responsive to
further manipulations of proteasome function by MG132.
It is likely that the Hinge region plays an important role in the proteasome-
mediated cycling of AR on target promoters. It was observed that components of the
proteasome occupy the upstream regulatory region of the endogenous PSA gene in
response to DHT treatment, consistent with their recruitment by AR. In the presence
of MG132, the AR remained at the PSA locus, but proteasome occupancy was no
longer detectible. This result, when coupled with PSA mRNA expression data
showing loss of gene expression, led to the conclusion that attenuated AR
transactivation in response to MG132 was not a result of disrupted AR binding to
DNA but rather, was a consequence of failed proteasome recruitment. Thus, the
AR Hinge contains a bimodal communication signal between AR and the
144
26S proteasome. This signal consists of two motifs located side-by-side in the PEST
sequence of the receptor that balance AR stabilization by chaperones in the
cytoplasm against ubiquitination signals that promote AR cycling in the nucleus.
In conclusion, the collective work contained within this Ph.D. thesis aimed to
elucidate the functional significance of several peptide regions of the AR. There is
significant importance assigned to understanding how the structural components of
the molecule contribute to its overall signaling abilities. The AR is the central hub
through which androgen signaling takes place in the cell and thus, insight into
structural aspects of how this is achieved will advance the ongoing pursuit of novel
drugs that target the aberrant androgen signaling pathways underlying prostate
cancer progression.
145
REFERENCES

Alen, P., Claessens, F., et al. The androgen receptor amino-terminal domain plays a
key role in p160 coactivator-stimulated gene transcription. Mol Cell Biol
19(9):6085-97, 1999.

Ashley, C. T., Jr. and Warren, S. T. Trinucleotide repeat expansion and human
disease. Annu Rev Genet 29:703-28, 1995.

Aumais, J. P., Lee, H. S., et al. Function of directly repeated half-sites as response
elements for steroid hormone receptors. J Biol Chem 271(21):12568-77,
1996.

Bai, S., He, B., et al. Melanoma antigen gene protein MAGE-11 regulates androgen
receptor function by modulating the interdomain interaction. Mol Cell Biol
25(4):1238-57, 2005.

Baker, M. E. Evolution of adrenal and sex steroid action in vertebrates: a ligand-
based mechanism for complexity. Bioessays 25(4):396-400, 2003.

Baron, S., Manin, M., et al. Androgen receptor mediates non-genomic activation of
phosphatidylinositol 3-OH kinase in androgen-sensitive epithelial cells. J
Biol Chem 279(15):14579-86, 2004.

Bates, G. and Lehrach, H. Trinucleotide repeat expansions and human genetic
disease. Bioessays 16(4):277-84, 1994.

Baumeister, W., Walz, J., et al. The proteasome: paradigm of a self-
compartmentalizing protease. Cell 92(3):367-80, 1998.

Beato, M. Gene regulation by steroid hormones. Cell 56(3):335-44, 1989.

Bentel, J. M. and Tilley, W. D. Androgen receptors in prostate cancer. J Endocrinol
151(1):1-11, 1996.

Bevan, C. L., Hoare, S., et al. The AF1 and AF2 domains of the androgen receptor
interact with distinct regions of SRC1. Mol Cell Biol 19(12):8383-92, 1999.

Blok, L. J., de Ruiter, P. E., et al. Forskolin-induced dephosphorylation of the
androgen receptor impairs ligand binding. Biochemistry 37(11):3850-7,
1998.

146
Bochtler, M., Ditzel, L., et al. The proteasome. Annu Rev Biophys Biomol Struct
28:295-317, 1999.

Bommer, M., Benecke, A., et al. TIF2 mediates the synergy between RARalpha 1
activation functions AF-1 and AF-2. J Biol Chem 277(40):37961-6, 2002.

Bonaccorsi, L., Marchiani, S., et al. Non-genomic effects of the androgen receptor
and Vitamin D agonist are involved in suppressing invasive phenotype of
prostate cancer cells. Steroids 71(4):304-9, 2006.

Brown, C. J., Goss, S. J., et al. Androgen receptor locus on the human X
chromosome: regional localization to Xq11-12 and description of a DNA
polymorphism. Am J Hum Genet 44(2):264-9, 1989.

Buchanan, G., Greenberg, N. M., et al. Collocation of androgen receptor gene
mutations in prostate cancer. Clin Cancer Res 7(5):1273-81, 2001.

Buchanan, G., Irvine, R. A., et al. Contribution of the androgen receptor to prostate
cancer predisposition and progression. Cancer Metastasis Rev 20(3-4):207-
23, 2001.

Buchanan, G., Yang, M., et al. Structural and functional consequences of glutamine
tract variation in the androgen receptor. Hum Mol Genet 13(16):1677-92,
2004.

Callewaert, L., Christiaens, V., et al. Implications of a polyglutamine tract in the
function of the human androgen receptor. Biochem Biophys Res Commun
306(1):46-52, 2003.

Callewaert, L., Van Tilborgh, N., et al. Interplay between two hormone-independent
activation domains in the androgen receptor. Cancer Res 66(1):543-53, 2006.

Carroll, J. S., Liu, X. S., et al. Chromosome-wide mapping of estrogen receptor
binding reveals long-range regulation requiring the forkhead protein FoxA1.
Cell 122(1):33-43, 2005.

Carroll, K. K. and Khor, H. T. Dietary fat in relation to tumorigenesis. Prog Biochem
Pharmacol 10:308-53, 1975.

Carroll, P., Coley, C., et al. Prostate-specific antigen best practice policy--part I:
early detection and diagnosis of prostate cancer. Urology 57(2):217-24, 2001.

Chamberlain, N. L., Driver, E. D., et al. The length and location of CAG
trinucleotide repeats in the androgen receptor N-terminal domain affect
transactivation function. Nucleic Acids Res 22(15):3181-6, 1994.
147

Chang, C., Gonzalez, F., et al. The Gal4 activation domain binds Sug2 protein, a
proteasome component, in vivo and in vitro. J Biol Chem 276(33):30956-63,
2001.

Chang, C. S., Kokontis, J., et al. Molecular cloning of human and rat complementary
DNA encoding androgen receptors. Science 240(4850):324-6, 1988.

Chen, C. D., Welsbie, D. S., et al. Molecular determinants of resistance to
antiandrogen therapy. Nat Med 10(1):33-9, 2004.

Chen, D., Huang, S. M., et al. Synergistic, p160 coactivator-dependent enhancement
of estrogen receptor function by CARM1 and p300. J Biol Chem
275(52):40810-6, 2000.

Chen, D., Ma, H., et al. Regulation of transcription by a protein methyltransferase.
Science 284(5423):2174-7, 1999.

Cheskis, B. J., McKenna, N. J., et al. Hierarchical affinities and a bipartite
interaction model for estrogen receptor isoforms and full-length steroid
receptor coactivator (SRC/p160) family members. J Biol Chem
278(15):13271-7, 2003.

Choong, C. S., Kemppainen, J. A., et al. Evolution of the primate androgen receptor:
a structural basis for disease. J Mol Evol 47(3):334-42, 1998.

Chu, K. C., Tarone, R. E., et al. Trends in prostate cancer mortality among black
men and white men in the United States. Cancer 97(6):1507-16, 2003.

Ciechanover, A. and Schwartz, A. L. The ubiquitin-proteasome pathway: the
complexity and myriad functions of proteins death. Proc Natl Acad Sci U S A
95(6):2727-30, 1998.

Coffey, D. S. The molecular biology, endocrinology and physiology of the prostate
and seminal vesicles. Walsh, P. C., Retick, A. B., Stamey, T. A. and Vaughn,
J. E. D., Eds. Campbell's Urology, 1381-1428: W.B. Saunders Co., 1992.

Connell, P., Ballinger, C. A., et al. The co-chaperone CHIP regulates protein triage
decisions mediated by heat-shock proteins. Nat Cell Biol 3(1):93-6, 2001.

Crawford, E. D. Epidemiology of prostate cancer. Urology 62(6 Suppl 1):3-12, 2003.

Cunha, G. R., Donjacour, A. A., et al. The endocrinology and developmental biology
of the prostate. Endocr Rev 8(3):338-62, 1987.

148
Cunningham, G. R., Ashton, C. M., et al. Familial aggregation of prostate cancer in
African-Americans and white Americans. Prostate 56(4):256-62, 2003.

Cyr, D. M., Hohfeld, J., et al. Protein quality control: U-box-containing E3 ubiquitin
ligases join the fold. Trends Biochem Sci 27(7):368-75, 2002.

Dahlman-Wright, K., Baumann, H., et al. Structural characterization of a minimal
functional transactivation domain from the human glucocorticoid receptor.
Proc Natl Acad Sci U S A 92(5):1699-703, 1995.

Dahlman-Wright, K., Wright, A., et al. Interaction of the glucocorticoid receptor
DNA-binding domain with DNA as a dimer is mediated by a short segment
of five amino acids. J Biol Chem 266(5):3107-12, 1991.

Denis, L. J. and Griffiths, K. Endocrine treatment in prostate cancer. Semin Surg
Oncol 18(1):52-74, 2000.

Dennis, A. P., Lonard, D. M., et al. Inhibition of the 26S proteasome blocks
progesterone receptor-dependent transcription through failed recruitment of
RNA polymerase II. J Steroid Biochem Mol Biol 94(4):337-46, 2005.

Dennis, A. P. and O'Malley, B. W. Rush hour at the promoter: how the ubiquitin-
proteasome pathway polices the traffic flow of nuclear receptor-dependent
transcription. J Steroid Biochem Mol Biol 93(2-5):139-51, 2005.

Dennis, L. K. and Hayes, R. B. Alcohol and prostate cancer. Epidemiol Rev
23(1):110-4, 2001.

Deroo, B. J., Rentsch, C., et al. Proteasomal inhibition enhances glucocorticoid
receptor transactivation and alters its subnuclear trafficking. Mol Cell Biol
22(12):4113-23, 2002.

Deutsch, E., Maggiorella, L., et al. Environmental, genetic, and molecular features of
prostate cancer. Lancet Oncol 5(5):303-13, 2004.

Ding, D., Xu, L., et al. Effect of GGC (glycine) repeat length polymorphism in the
human androgen receptor on androgen action. Prostate 62(2):133-9, 2005.

Domitrz, I., Jedrzejowska, M., et al. [Kennedy's disease: expansion of the CAG
trinucleotide]. Neurol Neurochir Pol 35(1 Suppl):107-14, 2001.

Elhaji, Y. A., Stoica, I., et al. Impaired helix 12 dynamics due to proline 892
substitutions in the androgen receptor are associated with complete androgen
insensitivity. Hum Mol Genet 15(6):921-31, 2006.

149
Evans, R. M. The steroid and thyroid hormone receptor superfamily. Science
240(4854):889-95, 1988.

Feng, W., Ribeiro, R. C., et al. Hormone-dependent coactivator binding to a
hydrophobic cleft on nuclear receptors. Science 280(5370):1747-9, 1998.

Fenteany, G. and Schreiber, S. L. Lactacystin, proteasome function, and cell fate. J
Biol Chem 273(15):8545-8, 1998.

Ferdous, A., Gonzalez, F., et al. The 19S regulatory particle of the proteasome is
required for efficient transcription elongation by RNA polymerase II. Mol
Cell 7(5):981-91, 2001.

Gann, P. H., Hennekens, C. H., et al. A prospective evaluation of plasma prostate-
specific antigen for detection of prostatic cancer. Jama 273(4):289-94, 1995.

Gelmann, E. P. Molecular Biology of the Androgen Receptor. J Clin Oncol
20(13):3001-3015, 2002.

Gianni, M., Bauer, A., et al. Phosphorylation by p38MAPK and recruitment of SUG-
1 are required for RA-induced RAR gamma degradation and transactivation.
Embo J 21(14):3760-9, 2002.

Giles, G. G., Severi, G., et al. Sexual factors and prostate cancer. BJU Int 92(3):211-
6, 2003.

Gillette, T. G., Gonzalez, F., et al. Physical and functional association of RNA
polymerase II and the proteasome. Proc Natl Acad Sci U S A 101(16):5904-
9, 2004.

Gioeli, D., Ficarro, S. B., et al. Androgen receptor phosphorylation. Regulation and
identification of the phosphorylation sites. J Biol Chem 277(32):29304-14,
2002.

Giwercman, Y. L., Ivarsson, S. A., et al. A novel mutation in the D-box of the
androgen receptor gene (S597R) in two unrelated individuals Is associated
with both normal phenotype and severe PAIS. Horm Res 61(2):58-62, 2004.

Glass, C. K., Rose, D. W., et al. Nuclear receptor coactivators. Curr Opin Cell Biol
9(2):222-32, 1997.

Gleason, D. F. and Mellinger, G. T. Prediction of prognosis for prostatic
adenocarcinoma by combined histological grading and clinical staging. J
Urol 111(1):58-64, 1974.

150
Glickman, M. H. and Ciechanover, A. The ubiquitin-proteasome proteolytic
pathway: destruction for the sake of construction. Physiol Rev 82(2):373-428,
2002.

Gonzalez, F., Delahodde, A., et al. Recruitment of a 19S proteasome subcomplex to
an activated promoter. Science 296(5567):548-50, 2002.

Gorbea, C., Taillandier, D., et al. Assembly of the regulatory complex of the 26S
proteasome. Mol Biol Rep 26(1-2):15-9, 1999.

Greenland, K. J., Beilin, J., et al. Polymorphic CAG repeat length in the androgen
receptor gene and association with neurodegeneration in a heterozygous
female carrier of Kennedy's disease. J Neurol 251(1):35-41, 2004.

Gronberg, H. Prostate cancer epidemiology. Lancet 361(9360):859-64, 2003.

Guess, H. A. Benign prostatic hyperplasia and prostate cancer. Epidemiol Rev
23(1):152-8, 2001.

Haas, G. P. and Sakr, W. A. Epidemiology of prostate cancer. CA Cancer J Clin
47(5):273-87, 1997.

Hammarsten, J. and Hogstedt, B. Calculated fast-growing benign prostatic
hyperplasia--a risk factor for developing clinical prostate cancer. Scand J
Urol Nephrol 36(5):330-8, 2002.

Hankey, B. F., Feuer, E. J., et al. Cancer surveillance series: interpreting trends in
prostate cancer--part I: Evidence of the effects of screening in recent prostate
cancer incidence, mortality, and survival rates. J Natl Cancer Inst
91(12):1017-24, 1999.

Hard, T., Kellenbach, E., et al. Solution structure of the glucocorticoid receptor
DNA-binding domain. Science 249(4965):157-60, 1990.

He, B., Bai, S., et al. An androgen receptor NH2-terminal conserved motif interacts
with the COOH terminus of the Hsp70-interacting protein (CHIP). J Biol
Chem 279(29):30643-53, 2004.

He, B., Bowen, N. T., et al. Androgen-induced NH2- and COOH-terminal
Interaction Inhibits p160 Coactivator Recruitment by Activation Function 2.
J. Biol. Chem. 276(45):42293-42301, 2001.

He, B., Gampe, R. T., Jr., et al. Structural basis for androgen receptor interdomain
and coactivator interactions suggests a transition in nuclear receptor
activation function dominance. Mol Cell 16(3):425-38, 2004.
151

He, B., Kemppainen, J. A., et al. FXXLF and WXXLF sequences mediate the NH2-
terminal interaction with the ligand binding domain of the androgen receptor.
J Biol Chem 275(30):22986-94, 2000.

He, B., Minges, J. T., et al. The FXXLF motif mediates androgen receptor-specific
interactions with coregulators. J Biol Chem 277(12):10226-35, 2002.

He, B. and Wilson, E. M. The NH(2)-terminal and carboxyl-terminal interaction in
the human androgen receptor. Mol Genet Metab 75(4):293-8, 2002.

He, W. W., Fischer, L. M., et al. Molecular cloning of androgen receptors from
divergent species with a polymerase chain reaction technique: complete
cDNA sequence of the mouse androgen receptor and isolation of androgen
receptor cDNA probes from dog, guinea pig and clawed frog. Biochem
Biophys Res Commun 171(2):697-704, 1990.

Henttu, P. M., Kalkhoven, E., et al. AF-2 activity and recruitment of steroid receptor
coactivator 1 to the estrogen receptor depend on a lysine residue conserved in
nuclear receptors. Mol Cell Biol 17(4):1832-9, 1997.

Hickey, K., Do, K. A., et al. Smoking and prostate cancer. Epidemiol Rev 23(1):115-
25, 2001.

Hiramatsu, M., Maehara, I., et al. Aromatase in hyperplasia and carcinoma of the
human prostate. Prostate 31(2):118-24, 1997.

Hobisch, A., Eder, I. E., et al. Interleukin-6 regulates prostate-specific protein
expression in prostate carcinoma cells by activation of the androgen receptor.
Cancer Res 58(20):4640-5, 1998.

Hobisch, A., Hoffmann, J., et al. Antagonist/agonist balance of the nonsteroidal
antiandrogen bicalutamide (Casodex) in a new prostate cancer model. Urol
Int 65(2):73-9, 2000.

Honda, R., Tanaka, H., et al. Oncoprotein MDM2 is a ubiquitin ligase E3 for tumor
suppressor p53. FEBS Lett 420(1):25-7, 1997.

Hong, H., Kohli, K., et al. GRIP1, a novel mouse protein that serves as a
transcriptional coactivator in yeast for the hormone binding domains of
steroid receptors. Proc Natl Acad Sci U S A 93(10):4948-52, 1996.

Hsing, A. W. and Chokkalingam, A. P. Prostate cancer epidemiology. Front Biosci
11:1388-413, 2006.

152
Hsing, A. W., Tsao, L., et al. International trends and patterns of prostate cancer
incidence and mortality. Int J Cancer 85(1):60-7, 2000.

Huggins C, S. R., Hodges CV. Studies on prostate cancer II. The effect of castration
on advanced carcinoma of the prostate gland. Arch Surg 43:209-228, 1941.

Hummerich, H. and Lehrach, H. Trinucleotide repeat expansion and human disease.
Electrophoresis 16(9):1698-704, 1995.

Ikonen, T., Palvimo, J. J., et al. Interaction between the amino- and carboxyl-
terminal regions of the rat androgen receptor modulates transcriptional
activity and is influenced by nuclear receptor coactivators. J Biol Chem
272(47):29821-8, 1997.

Iniguez-Lluhi, J. A. and Pearce, D. A common motif within the negative regulatory
regions of multiple factors inhibits their transcriptional synergy. Mol Cell
Biol 20(16):6040-50, 2000.

Irvine, R. A. and Coetzee, G. A. Additional upstream coding sequences of MAGE-
11. Immunogenetics 49(6):585, 1999.

Irvine, R. A., Ma, H., et al. Inhibition of p160-mediated coactivation with increasing
androgen receptor polyglutamine length. Hum Mol Genet 9(2):267-74, 2000.

Irvine, R. A., Yu, M. C., et al. The CAG and GGC microsatellites of the androgen
receptor gene are in linkage disequilibrium in men with prostate cancer.
Cancer Res 55(9):1937-40, 1995.

Jenster, G., van der Korput, H. A., et al. Identification of two transcription activation
units in the N-terminal domain of the human androgen receptor. J Biol Chem
270(13):7341-6, 1995.

Jenster, G., van der Korput, J. A., et al. Functional domains of the human androgen
receptor. J Steroid Biochem Mol Biol 41(3-8):671-5, 1992.

Jia, L., Kim, J., et al. Androgen receptor activity at the prostate specific antigen
locus: steroidal and non-steroidal mechanisms. Mol Cancer Res 1(5):385-92,
2003.

Jiang, J., Ballinger, C. A., et al. CHIP is a U-box-dependent E3 ubiquitin ligase:
identification of Hsc70 as a target for ubiquitylation. J Biol Chem
276(46):42938-44, 2001.

153
Jones, P. L. and Shi, Y. B. N-CoR-HDAC corepressor complexes: roles in
transcriptional regulation by nuclear hormone receptors. Curr Top Microbiol
Immunol 274:237-68, 2003.

Kang, Z., Pirskanen, A., et al. Involvement of proteasome in the dynamic assembly
of the androgen receptor transcription complex. J Biol Chem 277(50):48366-
71, 2002.

Kinyamu, H. K. and Archer, T. K. Modifying chromatin to permit steroid hormone
receptor-dependent transcription. Biochim Biophys Acta 1677(1-3):30-45,
2004.

Kinyamu, H. K., Chen, J., et al. Linking the ubiquitin-proteasome pathway to
chromatin remodeling/modification by nuclear receptors. J Mol Endocrinol
34(2):281-97, 2005.

Kobayashi, Y., Kitamoto, T., et al. p300 mediates functional synergism between AF-
1 and AF-2 of estrogen receptor alpha and beta by interacting directly with
the N-terminal A/B domains. J Biol Chem 275(21):15645-51, 2000.

Kopp, F., Hendil, K. B., et al. Subunit arrangement in the human 20S proteasome.
Proc Natl Acad Sci U S A 94(7):2939-44, 1997.

Kotaja, N., Karvonen, U., et al. PIAS proteins modulate transcription factors by
functioning as SUMO-1 ligases. Mol Cell Biol 22(14):5222-34, 2002.

Kozlowski, J. M., Ellis, W. J., et al. Advanced prostatic carcinoma. Early versus late
endocrine therapy. Urol Clin North Am 18(1):15-24, 1991.

Kuiper, G. G., Faber, P. W., et al. Structural organization of the human androgen
receptor gene. J Mol Endocrinol 2(3):R1-4, 1989.

Kumar, R., Betney, R., et al. Induced alpha-helix structure in AF1 of the androgen
receptor upon binding transcription factor TFIIF. Biochemistry 43(11):3008-
13, 2004.

Lee, D., Ezhkova, E., et al. The proteasome regulatory particle alters the SAGA
coactivator to enhance its interactions with transcriptional activators. Cell
123(3):423-36, 2005.

Legler, J. M., Feuer, E. J., et al. The role of prostate-specific antigen (PSA) testing
patterns in the recent prostate cancer incidence decline in the United States.
Cancer Causes Control 9(5):519-27, 1998.

154
Li, J., Fu, J., et al. A role of the amino-terminal (N) and carboxyl-terminal (C)
interaction in binding of androgen receptor to chromatin. Mol Endocrinol
20(4):776-85, 2006.

Li, J., Zhang, D., et al. Structural and Functional Analysis of Androgen Receptor in
Chromatin. Mol Endocrinol, 2007.

Li, Q., Su, A., et al. Attenuation of glucocorticoid signaling through targeted
degradation of p300 via the 26S proteasome pathway. Mol Endocrinol
16(12):2819-27, 2002.

Liao, G., Chen, L. Y., et al. Regulation of androgen receptor activity by the nuclear
receptor corepressor SMRT. J Biol Chem 278(7):5052-61, 2003.

Lin, H. K., Altuwaijri, S., et al. Proteasome activity is required for androgen receptor
transcriptional activity via regulation of androgen receptor nuclear
translocation and interaction with coregulators in prostate cancer cells. J Biol
Chem 277(39):36570-6, 2002.

Lin, H. K., Wang, L., et al. Phosphorylation-dependent ubiquitylation and
degradation of androgen receptor by Akt require Mdm2 E3 ligase. Embo J
21(15):4037-48, 2002.

Lin, H. K., Yeh, S., et al. Akt suppresses androgen-induced apoptosis by
phosphorylating and inhibiting androgen receptor. Proc Natl Acad Sci U S A
98(13):7200-5, 2001.

Lindzey, J., Kumar, M. V., et al. Molecular mechanisms of androgen action. Vitam
Horm 49:383-432, 1994.

Liu, G. Z., Wang, H., et al. Identification of a highly conserved domain in the
androgen receptor that suppresses the DNA-binding domain-DNA
interactions. J Biol Chem 278(17):14956-60, 2003.

Lonard, D. M., Nawaz, Z., et al. The 26S proteasome is required for estrogen
receptor-alpha and coactivator turnover and for efficient estrogen receptor-
alpha transactivation. Mol Cell 5(6):939-48, 2000.

Lubahn, D. B., Brown, T. R., et al. 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 U S A 86(23):9534-8, 1989.

155
Ma, H., Baumann, C. T., et al. Hormone-dependent, CARM1-directed, arginine-
specific methylation of histone H3 on a steroid-regulated promoter. Curr Biol
11(24):1981-5, 2001.

Ma, H., Hong, H., et al. Multiple signal input and output domains of the 160-
kilodalton nuclear receptor coactivator proteins. Mol Cell Biol 19(9):6164-
73, 1999.

Mangiarini, L., Sathasivam, K., et al. Instability of highly expanded CAG repeats in
mice transgenic for the Huntington's disease mutation. Nat Genet 15(2):197-
200, 1997.

Masiello, D., Cheng, S., et al. Bicalutamide functions as an androgen receptor
antagonist by assembly of a transcriptionally inactive receptor. J Biol Chem
277(29):26321-6, 2002.

Masuyama, H. and MacDonald, P. N. Proteasome-mediated degradation of the
vitamin D receptor (VDR) and a putative role for SUG1 interaction with the
AF-2 domain of VDR. J Cell Biochem 71(3):429-40, 1998.

McEwan, I. J. and Gustafsson, J. Interaction of the human androgen receptor
transactivation function with the general transcription factor TFIIF. Proc Natl
Acad Sci U S A 94(16):8485-90, 1997.

McKenna, N. J., Lanz, R. B., et al. Nuclear receptor coregulators: cellular and
molecular biology. Endocr Rev 20(3):321-44, 1999.

McNeal, J. E. The zonal anatomy of the prostate. Prostate 2(1):35-49, 1981.

McNeal, J. E. Normal histology of the prostate. American Journal of Surgical
Pathology 12(8):619-33, 1988.

Melcher, K. and Johnston, S. A. GAL4 interacts with TATA-binding protein and
coactivators. Mol Cell Biol 15(5):2839-48, 1995.

Melvin, V. S., Roemer, S. C., et al. The C-terminal extension (CTE) of the nuclear
hormone receptor DNA binding domain determines interactions and
functional response to the HMGB-1/-2 co-regulatory proteins. J Biol Chem
277(28):25115-24, 2002.

Messing, E. M., Manola, J., et al. Immediate hormonal therapy compared with
observation after radical prostatectomy and pelvic lymphadenectomy in men
with node-positive prostate cancer. N Engl J Med 341(24):1781-8, 1999.

156
Metivier, R., Penot, G., et al. Synergism between ERalpha transactivation function 1
(AF-1) and AF-2 mediated by steroid receptor coactivator protein-1:
requirement for the AF-1 alpha-helical core and for a direct interaction
between the N- and C-terminal domains. Mol Endocrinol 15(11):1953-70,
2001.

Miron, L. [The hormonal and chemotherapy of prostatic cancer]. Rev Med Chir Soc
Med Nat Iasi 100(3-4):37-43, 1996.

Muller, S., Hoege, C., et al. SUMO, ubiquitin's mysterious cousin. Nat Rev Mol Cell
Biol 2(3):202-10, 2001.

Murata, S., Minami, Y., et al. CHIP is a chaperone-dependent E3 ligase that
ubiquitylates unfolded protein. EMBO Rep 2(12):1133-8, 2001.

Nazareth, L. V. and Weigel, N. L. Activation of the human androgen receptor
through a protein kinase A signaling pathway. J Biol Chem 271(33):19900-7,
1996.

Nettles, K. W. and Greene, G. L. Ligand control of coregulator recruitment to
nuclear receptors. Annu Rev Physiol 67:309-33, 2005.

Nguyen, D., Steinberg, S. V., et al. A G577R mutation in the human AR P box
results in selective decreases in DNA binding and in partial androgen
insensitivity syndrome. Mol Endocrinol 15(10):1790-802, 2001.

Nishida, T. and Yasuda, H. PIAS1 and PIASxalpha function as SUMO-E3 ligases
toward androgen receptor and repress androgen receptor-dependent
transcription. J Biol Chem 277(44):41311-7, 2002.

Oefelein, M. G. and Resnick, M. I. Association of tobacco use with hormone
refractory disease and survival of patients with prostate cancer. J Urol 171(6
Pt 1):2281-4, 2004.

Onate, S. A., Tsai, S. Y., et al. Sequence and characterization of a coactivator for the
steroid hormone receptor superfamily. Science 270(5240):1354-7, 1995.

Owen, G. I. and Zelent, A. Origins and evolutionary diversification of the nuclear
receptor superfamily. Cell Mol Life Sci 57(5):809-27, 2000.

Parker, M. G. Transcriptional activation by oestrogen receptors. Biochem Soc Symp
63:45-50, 1998.

Parkin, D. M. Global cancer statistics in the year 2000. Lancet Oncol 2(9):533-43,
2001.
157

Perissi, V., Aggarwal, A., et al. A corepressor/coactivator exchange complex
required for transcriptional activation by nuclear receptors and other
regulated transcription factors. Cell 116(4):511-26, 2004.

Platz, E. A., Leitzmann, M. F., et al. Alcohol intake, drinking patterns, and risk of
prostate cancer in a large prospective cohort study. Am J Epidemiol
159(5):444-53, 2004.

Potosky, A. L., Miller, B. A., et al. The role of increasing detection in the rising
incidence of prostate cancer. Jama 273(7):548-52, 1995.

Poukka, H., Aarnisalo, P., et al. Ubc9 interacts with the androgen receptor and
activates receptor-dependent transcription. J Biol Chem 274(27):19441-6,
1999.

Poukka, H., Karvonen, U., et al. Covalent modification of the androgen receptor by
small ubiquitin-like modifier 1 (SUMO-1). Proc Natl Acad Sci U S A
97(26):14145-50, 2000.

Poukka, H., Karvonen, U., et al. The RING finger protein SNURF modulates nuclear
trafficking of the androgen receptor. J Cell Sci 113 ( Pt 17):2991-3001, 2000.

Powell, S. M., Christiaens, V., et al. Mechanisms of androgen receptor signalling via
steroid receptor coactivator-1 in prostate. Endocr Relat Cancer 11(1):117-30,
2004.

Pratt, W. B., Galigniana, M. D., et al. Role of hsp90 and the hsp90-binding
immunophilins in signalling protein movement. Cell Signal 16(8):857-72,
2004.

Punglia, R. S., D'Amico, A. V., et al. Effect of verification bias on screening for
prostate cancer by measurement of prostate-specific antigen. N Engl J Med
349(4):335-42, 2003.

Quinn, M. and Babb, P. Patterns and trends in prostate cancer incidence, survival,
prevalence and mortality. Part I: international comparisons. BJU Int
90(2):162-73, 2002.

Rechsteiner, M. PEST sequences are signals for rapid intracellular proteolysis.
Semin Cell Biol 1(6):433-40, 1990.

Rechsteiner, M. and Rogers, S. W. PEST sequences and regulation by proteolysis.
Trends Biochem Sci 21(7):267-71, 1996.

158
Rees, I., Lee, S., et al. The E3 ubiquitin ligase CHIP binds the androgen receptor in a
phosphorylation-dependent manner. Biochim Biophys Acta 1764(6):1073-9,
2006.

Reid, G., Hubner, M. R., et al. Cyclic, proteasome-mediated turnover of unliganded
and liganded ERalpha on responsive promoters is an integral feature of
estrogen signaling. Mol Cell 11(3):695-707, 2003.

Reid, J., Kelly, S. M., et al. Conformational analysis of the androgen receptor amino-
terminal domain involved in transactivation. Influence of structure-stabilizing
solutes and protein-protein interactions. J Biol Chem 277(22):20079-86,
2002.

Roberts, W. W., Platz, E. A., et al. Association of cigarette smoking with
extraprostatic prostate cancer in young men. J Urol 169(2):512-6; discussion
516, 2003.

Rogers, S., Wells, R., et al. Amino acid sequences common to rapidly degraded
proteins: the PEST hypothesis. Science 234(4774):364-8, 1986.

Rogerson, F. M. and Fuller, P. J. Interdomain interactions in the mineralocorticoid
receptor. Mol Cell Endocrinol 200(1-2):45-55, 2003.

Rosenblatt, K. A., Wicklund, K. G., et al. Sexual factors and the risk of prostate
cancer. Am J Epidemiol 153(12):1152-8, 2001.

Rosenfeld, M. G. and Glass, C. K. Coregulator codes of transcriptional regulation by
nuclear receptors. J Biol Chem 276(40):36865-8, 2001.

Rubinsztein, D. C., Leggo, J., et al. 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(9):1585-90, 1995.

Sack, J. S., Kish, K. F., et al. Crystallographic structures of the ligand-binding
domains of the androgen receptor and its T877A mutant complexed with the
natural agonist dihydrotestosterone. Proc Natl Acad Sci U S A 98(9):4904-9,
2001.

Sakr, W. A., Haas, G. P., et al. The frequency of carcinoma and intraepithelial
neoplasia of the prostate in young male patients. J Urol 150(2 Pt 1):379-85,
1993.

Schaid, D. J. The complex genetic epidemiology of prostate cancer. Hum Mol Genet
13 Spec No 1:R103-21, 2004.

159
Schoenmakers, E., Verrijdt, G., et al. Differences in DNA binding characteristics of
the androgen and glucocorticoid receptors can determine hormone-specific
responses. J Biol Chem 275(16):12290-7, 2000.

Schoonen, W. M., Salinas, C. A., et al. Alcohol consumption and risk of prostate
cancer in middle-aged men. Int J Cancer 113(1):133-40, 2005.

Schwabe, J. W. and Teichmann, S. A. Nuclear receptors: the evolution of diversity.
Sci STKE 2004(217):pe4, 2004.

Senius, K. E., Pieltila, J., et al. A simple method for the clinical determination of the
mitotic activity of the human prostate in vitro. J Clin Pathol 27(11):880-2,
1974.

Shao, D., Rangwala, S. M., et al. Interdomain communication regulating ligand
binding by PPAR-gamma. Nature 396(6709):377-80, 1998.

Sharma, M., Li, X., et al. hZimp10 is an androgen receptor co-activator and forms a
complex with SUMO-1 at replication foci. Embo J 22(22):6101-14, 2003.

Sheflin, L., Keegan, B., et al. Inhibiting proteasomes in human HepG2 and LNCaP
cells increases endogenous androgen receptor levels. Biochem Biophys Res
Commun 276(1):144-50, 2000.

Shen, T., Horwitz, K. B., et al. Transcriptional hyperactivity of human progesterone
receptors is coupled to their ligand-dependent down-regulation by mitogen-
activated protein kinase-dependent phosphorylation of serine 294. Mol Cell
Biol 21(18):6122-31, 2001.

Shibata, A. and Whittemore, A. S. Genetic predisposition to prostate cancer: possible
explanations for ethnic differences in risk. Prostate 32(1):65-72, 1997.

Siiteri, P. K. Sex hormone production and action. Arthritis Rheum 22(11):1284-94,
1979.

Snoek, R., Bruchovsky, N., et al. Differential transactivation by the androgen
receptor in prostate cancer cells. Prostate 36(4):256-63, 1998.

Spanjaard, R. A. and Chin, W. W. Reconstitution of ligand-mediated glucocorticoid
receptor activity by trans-acting functional domains. Mol Endocrinol 7(1):12-
6, 1993.

Spencer, J. A., Watson, J. M., et al. The androgen receptor gene is located on a
highly conserved region of the X chromosomes of marsupial and monotreme
as well as eutherian mammals. J Hered 82(2):134-9, 1991.
160

Stallcup, M. R., Chen, D., et al. Co-operation between protein-acetylating and
protein-methylating co-activators in transcriptional activation. Biochem Soc
Trans 28(4):415-8, 2000.

Strahl, B. D., Briggs, S. D., et al. Methylation of histone H4 at arginine 3 occurs in
vivo and is mediated by the nuclear receptor coactivator PRMT1. Curr Biol
11(12):996-1000, 2001.

Swaffield, J. C., Bromberg, J. F., et al. Alterations in a yeast protein resembling HIV
Tat-binding protein relieve requirement for an acidic activation domain in
GAL4. Nature 360(6406):768, 1992.

Swaffield, J. C., Melcher, K., et al. A highly conserved ATPase protein as a mediator
between acidic activation domains and the TATA-binding protein. Nature
374(6517):88-91, 1995.

Tanenbaum, D. M., Wang, Y., et al. Crystallographic comparison of the estrogen and
progesterone receptor's ligand binding domains. Proc Natl Acad Sci U S A
95(11):5998-6003, 1998.

Tanner, T., Claessens, F., et al. The hinge region of the androgen receptor plays a
role in proteasome-mediated transcriptional activation. Ann N Y Acad Sci
1030:587-92, 2004.

Taplin, M. E., Bubley, G. J., et al. Selection for androgen receptor mutations in
prostate cancers treated with androgen antagonist. Cancer Res 59(11):2511-5,
1999.

Tetel, M. J., Jung, S., et al. Hinge and amino-terminal sequences contribute to
solution dimerization of human progesterone receptor. Mol Endocrinol
11(8):1114-28, 1997.

Teyssier, C., Chen, D., et al. Requirement for multiple domains of the protein
arginine methyltransferase CARM1 in its transcriptional coactivator function.
J Biol Chem 277(48):46066-72, 2002.

Thornton, J. W. and Kelley, D. B. Evolution of the androgen receptor: structure-
function implications. Bioessays 20(10):860-9, 1998.

Thornton, J. W., Need, E., et al. Resurrecting the ancestral steroid receptor: ancient
origin of estrogen signaling. Science 301(5640):1714-7, 2003.

161
Tilley, W. D., Marcelli, M., et al. Characterization and expression of a cDNA
encoding the human androgen receptor. Proc Natl Acad Sci U S A 86(1):327-
31, 1989.

Tomlins, S. A., Rhodes, D. R., et al. Recurrent fusion of TMPRSS2 and ETS
transcription factor genes in prostate cancer. Science 310(5748):644-8, 2005.

Torchia, J., Rose, D. W., et al. The transcriptional co-activator p/CIP binds CBP and
mediates nuclear-receptor function. Nature 387(6634):677-84, 1997.

Toth, M. and Zakar, T. Different binding of testosterone, 19-nortestosterone and
their 5 alpha-reduced derivatives to the androgen receptor of the rat seminal
vesicle: a step toward the understanding of the anabolic action of
nortesterone. Endokrinologie 80(2):163-72, 1982.

Unni, E., Sun, S., et al. Changes in androgen receptor nongenotropic signaling
correlate with transition of LNCaP cells to androgen independence. Cancer
Res 64(19):7156-68, 2004.

Van de Voorde, W. M. Pathology of prostatic carcinoma. Petrovich, Z., Baert, L. and
Brady, L. W., Eds. Carcinoma of the prostate: innovations in management,
27-50. Berlin: Springer-Verlag, 1996.

Verrijdt, G., Schoenmakers, E., et al. Change of specificity mutations in androgen-
selective enhancers. Evidence for a role of differential DNA binding by the
androgen receptor. J Biol Chem 275(16):12298-305, 2000.

Villers, A., Terris, M. K., et al. Ultrasound anatomy of the prostate: the normal gland
and anatomical variations. J Urol 143(4):732-8, 1990.

Visakorpi, T., Hyytinen, E., et al. In vivo amplification of the androgen receptor
gene and progression of human prostate cancer. Nat Genet 9(4):401-6, 1995.

vom Baur, E., Zechel, C., et al. Differential ligand-dependent interactions between
the AF-2 activating domain of nuclear receptors and the putative
transcriptional intermediary factors mSUG1 and TIF1. Embo J 15(1):110-24,
1996.

Wang, Q., Lu, J., et al. Ligand- and coactivator-mediated transactivation function
(AF2) of the androgen receptor ligand-binding domain is inhibited by the
cognate hinge region. J Biol Chem 276(10):7493-9, 2001.

Wardell, S. E., Kwok, S. C., et al. Regulation of the amino-terminal transcription
activation domain of progesterone receptor by a cofactor-induced protein
folding mechanism. Mol Cell Biol 25(20):8792-808, 2005.
162

Warnmark, A., Treuter, E., et al. Activation functions 1 and 2 of nuclear receptors:
molecular strategies for transcriptional activation. Mol Endocrinol
17(10):1901-9, 2003.

Webb, P., Nguyen, P., et al. Estrogen receptor activation function 1 works by
binding p160 coactivator proteins. Mol Endocrinol 12(10):1605-18, 1998.

Wiener, J. S., Teague, J. L., et al. Molecular biology and function of the androgen
receptor in genital development. J Urol 157(4):1377-86, 1997.

Wilding, G., Chen, M., et al. Aberrant response in vitro of hormone-responsive
prostate cancer cells to antiandrogens. Prostate 14(2):103-15, 1989.

Wong, B. R., Parlati, F., et al. Drug discovery in the ubiquitin regulatory pathway.
Drug Discov Today 8(16):746-54, 2003.

Wong, C. I., Zhou, Z. X., et al. Steroid requirement for androgen receptor
dimerization and DNA binding. Modulation by intramolecular interactions
between the NH2-terminal and steroid-binding domains. J Biol Chem
268(25):19004-12, 1993.

Wong, H. Y., Burghoorn, J. A., et al. Phosphorylation of androgen receptor isoforms.
Biochem J 383(Pt 2):267-76, 2004.

Wu, F. and Mo, Y. Y. Ubiquitin-like protein modifications in prostate and breast
cancer. Front Biosci 12:700-11, 2007.

Xu, J. and Li, Q. Review of the in vivo functions of the p160 steroid receptor
coactivator family. Mol Endocrinol 17(9):1681-92, 2003.

Xu, J. and O'Malley, B. W. Molecular mechanisms and cellular biology of the
steroid receptor coactivator (SRC) family in steroid receptor function. Rev
Endocr Metab Disord 3(3):185-92, 2002.

Xu, Q., Singer, R. A., et al. Sug1 modulates yeast transcription activation by Cdc68.
Mol Cell Biol 15(11):6025-35, 1995.

Yamamoto, K. R. Steroid receptor regulated transcription of specific genes and gene
networks. Annu Rev Genet 19:209-52, 1985.

Yeh, S., Lin, H. K., et al. From HER2/Neu signal cascade to androgen receptor and
its coactivators: a novel pathway by induction of androgen target genes
through MAP kinase in prostate cancer cells. Proc Natl Acad Sci U S A
96(10):5458-63, 1999.
163

Zeegers, M. P., Kiemeney, L. A., et al. How strong is the association between CAG
and GGN repeat length polymorphisms in the androgen receptor gene and
prostate cancer risk? Cancer Epidemiol Biomarkers Prev 13(11 Pt 1):1765-
71, 2004.

Zegarra-Moro, O. L., Schmidt, L. J., et al. Disruption of androgen receptor function
inhibits proliferation of androgen-refractory prostate cancer cells. Cancer Res
62(4):1008-13, 2002.

Zhang, H., Sun, L., et al. The catalytic subunit of the proteasome is engaged in the
entire process of estrogen receptor-regulated transcription. Embo J
25(18):4223-33, 2006.

Zhang, J., Thomas, T. Z., et al. A small composite probasin promoter confers high
levels of prostate-specific gene expression through regulation by androgens
and glucocorticoids in vitro and in vivo. Endocrinology 141(12):4698-710,
2000.

Zhou, H. J., Yan, J., et al. SRC-3 is required for prostate cancer cell proliferation and
survival. Cancer Res 65(17):7976-83, 2005.

Zhou, Z. X., Kemppainen, J. A., et al. Identification of three proline-directed
phosphorylation sites in the human androgen receptor. Mol Endocrinol
9(5):605-15, 1995.

Zhou, Z. X., Sar, M., et al. A ligand-dependent bipartite nuclear targeting signal in
the human androgen receptor. Requirement for the DNA-binding domain and
modulation by NH2-terminal and carboxyl-terminal sequences. J Biol Chem
269(18):13115-23, 1994.

Zhu, Q., Wani, G., et al. Deubiquitination by proteasome is coordinated with
substrate translocation for proteolysis in vivo. Exp Cell Res 307(2):436-51,
2005.

Zilliacus, J., Dahlman-Wright, K., et al. DNA binding specificity of mutant
glucocorticoid receptor DNA-binding domains. J Biol Chem 266(5):3101-6,
1991.

Zilliacus, J., Wright, A. P., et al. Structural determinants of DNA-binding specificity
by steroid receptors. Mol Endocrinol 9(4):389-400, 1995.

Abstract (if available)

Abstract Various peptide regions of the human androgen receptor (AR) were characterized for their potential role in mediating the androgenic signaling response that is integrally linked to prostate cancer development and resistance to treatment. A series of site-directed mutant AR constructs was created that targeted putative signaling motifs within the molecule, and was utilized in functional assays to determine how deviations from the wild-type AR sequence at these sites impacted receptor activity. It was shown that the AR amino-terminal transactivation domain (NTD) possesses great compensatory ability, as independent disruption of several sites had little impact on receptor function in prostate-cancer derived cell lines. Size modulations of the glycine trinucleotide repeat in the NTD had a direct effect on receptor signaling, suggestive of this region having important influence on NTD structure. It was determined that a motif located in the proximal region of the NTD contributes greatly to overall receptor function through mediating ligand-dependent interactions between the AR NTD and ligand-binding domains (LBD). In addition, it was revealed that one mechanism by which the p160 nuclear receptor coactivators enhance AR function is through enhancement of this inter-domain communication. In related studies it was shown that the AR Hinge domain imparts a pivotal contribution to normal AR signaling through mediating communication between AR and the 26S proteasome. Dual roles of the proteasome in AR signaling were characterized based on observations that separate signaling motifs within the AR Hinge control a balance between the apo-receptor and holo-receptor responses to proteasome influence.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Co-chaperone influence on androgen receptor signaling and identification of androgen receptor genes in prostate cancer

PDF

Modulation of Runx proteins by steroid hormone receptors

PDF

Identification of novel androgen receptor target genes in prostate cancer

PDF

Dietary and supplementary folate intake and prostate cancer risk

PDF

Monoamine oxidase A inhibitors and androgen receptor antagonists regulate mitochondrial function in prostate cancer cells

PDF

Homologous cell systems for the study of progression of androgen-dependent prostate cancer to castration-resistant prostate cancer

PDF

Genes and environment in prostate cancer risk and prognosis

PDF

Association of comorbidity with prostate cancer tumor characteristics in African American men

PDF

Exploration of the roles of cancer stem cells and survivin in the pathogenesis and progression of prostate cancer

PDF

Study of bone morphogenetic protein-2 and stromal cell derived factor-1 in prostate cancer

PDF

Dietary carcinogens and genetic variation in their metabolism: epidemiological studies on the risk of selected cancers

PDF

Chemical and biological studies of novel ligands of the human androgen receptor

PDF

Association between informed decision-making and mental health-related quality of life in long term prostate cancer survivors

PDF

Targeting molecular signals involved in the development of castration resistant prostate cancer

PDF

Identification and fine-mapping of genetic susceptibility loci for prostate cancer and statistical methodology for multiethnic fine-mapping

PDF

Prostate cancer disparities among Californian Latinos by country of origin: clinical characteristics, incidence, treatment received and survival

PDF

The mechanism of recruitment of Tip60 to ER target genes

PDF

A resistance exercise and protein supplementation intervention for prostate cancer survivors on androgen deprivation therapy

PDF

Functional characterization of a prostate cancer risk region

PDF

Development and validation of survey instrument designed for cervical cancer screening in Malawi, and other low resource settings

Asset Metadata

Creator Shen, Howard Chung-Hao (author)

Core Title Functional analyses of androgen receptor structure pertaining to prostate cancer

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Molecular Epidemiology

Publication Date 04/19/2007

Defense Date 03/05/2007

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

Tag androgen receptor,nuclear receptor,OAI-PMH Harvest,prostate cancer,steroid receptor

Language English

Advisor Coetzee, Gerhard A. (committee chair), Ingles, Sue A. (committee member), Stallcup, Michael R. (committee member)

Creator Email hcs@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m411

Unique identifier UC1482485

Identifier etd-Shen-20070419 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-477600 (legacy record id),usctheses-m411 (legacy record id)

Legacy Identifier etd-Shen-20070419.pdf

Dmrecord 477600

Document Type Dissertation

Rights Shen, Howard Chung-Hao

Type texts

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

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu