Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Infrequent androgen receptor mutations in primary prostate tumors from men residing in Singapore and Los Angeles
(USC Thesis Other)
Infrequent androgen receptor mutations in primary prostate tumors from men residing in Singapore and Los Angeles
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
INFREQUENT ANDROGEN RECEPTOR MUTATIONS IN PRIMARY
PROSTATE TUMORS FROM MEN RESIDING IN
SINGAPORE AND LOS ANGELES
Copyright 2005
by
Marcus Allan Wantroba
A Thesis Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(MOLECULAR EPIDEMIOLOGY)
August 2005
Marcus Allan Wantroba
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UMI Number: 1 4 3 0 4 0 8
INFORMATION TO USERS
The quality of this reproduction is dependent upon the quality of the copy
submitted. Broken or indistinct print, colored or poor quality illustrations and
photographs, print bleed-through, substandard margins, and improper
alignment can adversely affect reproduction.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if unauthorized
copyright material had to be removed, a note will indicate the deletion.
®
UMI
UMI Microform 1430408
Copyright 2006 by ProQuest Information and Learning Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
ProQuest Information and Learning Company
300 North Zeeb Road
P.O. Box 1346
Ann Arbor, Ml 48106-1346
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
DEDICATION
Charles O’Reilly and Paul Wantroba
Patriarchs
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
iii
ACKNOWLEDGEMENTS
Mom and dad, thank you for all your support, love, and patience and for instilling in
me the importance and value of honor and family. Mom, thanks for all the sacrifices
you’ve made with me in mind (30 years worth), for putting up with dad and I (our
philosophy, religion, and science debates), and for all the time you’ve spent
proofreading this. Dad, thank you for your guidance, wisdom, opinion and, most of
all, your tolerance.. .especially when I ignored the previous three. You have made
Grandpa Paul very proud. But take my word for it.. .Philosophy complicates the
simple; Science simplifies the complicated. Sarah and Anders, you’re the best sister
and brother-in-law a guy could have. Thank you Unky Paul and Auntie Jill for
everything. Thank you to all the Wrobels, Bodinofs, O’Reilly’s, Austgens, Dunns,
and Wickerts for being such a close and supportive family. Thank you to the Dennerts
for making me feel at home. Matty, Becks, CJ, Magoo, Drew, Ken, Joel, JJ, Michael,
Katie, Tony, Julie, Howie, Jen, Li, Omar, the state of California, the City of Los
Angeles, and Bukowski, thanks for making life fun.
Thanks to Drs. Coetzee (for the opportunity to work in your lab), Pinski, Yu, and
Bujarski for all your time, guidance, and knowledge.
Emily, my better half, thank you for being supportive and tolerant while I pursue my
ambitions and for calling me whenever a police chase is on. Love you Bug... and
Mizzax-shizzle. Goodnight Max 12/16/03
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
iv
CONTENTS
DEDICATION ii
ACKNOWLEDGEMENTS iii
LIST OF TABLES vi
LIST OF FIGURES vii
ABSTRACT viii
INTRODUCTION 1
THE PROSTATE 3
Basic Anatomy 3
History of Prostate Anatomy 4
Zonal Anatomy of the Prostate 5
Histology of the Prostate 8
EPIDEMIOLOGY OF PROSTATE CANCER 12
United States Epidemiology 12
World Epidemiology 14
PSA Screening 17
Potential Risk Factors 22
THE ANDROGEN RECEPTOR 23
Androgens 23
History of Androgen Receptor Research 24
The Androgen Receptor Gene and Transcription 26
Structure and Function of the Androgen Receptor 31
Androgen Receptor, Mutations, and Prostate Cancer 39
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
V
STUDY OVERVIEW 45
MATERIALS AND METHODS 48
Study Design 48
Study Subjects 51
Proteinase K Digestion 54
PCR Amplification 55
PCR Product Purification and Sequencing 60
Verification of Sequences 61
AR-CAG Repeat Analyses 62
RESULTS 64
Los Angeles Study 64
Singapore Study 69
DISCUSSION 75
Mutation Frequency 75
M886V Mutation 80
LA and Singapore AR-CAG Associations 83
BIBLIOGRAPHY 88
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
vi
LIST OF TABLES
TABLE 1. PCR Product Sizes
TABLE 2. First Round Reaction Primers
TABLE 3. Nested Round Reaction Primers
TABLE 4. Sequencing Results of LA Study
TABLE 5. LA Study Characteristics of Variables
TABLE 6. LA Study Distribution of CAG Repeats by Interval
TABLE 7. LA Study Results of Analyses
TABLE 8. Sequencing Results of Singapore Study
TABLE 9. Singapore Study Characteristics of Variables
TABLE 10. Singapore Study Distribution of CAG Repeats by Interval
TABLE 11. Singapore Study Case-Only Results of Analyses
TABLE 12. Singapore Study Case-Control Results of Analyses
55
57
59
65
66
67
68
70
72
73
74
74
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
v ii
LIST OF FIGURES
FIGURE 1. Structure of the Androgen Receptor Gene 27
FIGURE 2. Structure of the Androgen Receptor Coding Regions 32
FIGURE 3. Structure of Androgen Receptor mRNA and Protein 33
FIGURE 4. Frequencies and locations of AR mutations in disease 40
FIGURE 5. Location of Mutation Collocation Sites in the Androgen Receptor 49
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
v iii
ABSTRACT
We performed case-only studies of African-American, Latino, and Chinese men in
order to: (i) Assess the frequency of androgen receptor (AR) somatic mutations in
prostate tumors from who have not received androgen ablation therapy and: (ii)
Investigate the associations between AR-CAG repeat length and tumor characteristics.
Additionally, we performed a case-control analysis to estimate PCa risk associated
with PCa and AR-CAG repeat length in Chinese men. AR mutations occurred
extremely infrequently in primary tumors from men who have not undergone
androgen ablation therapy, which indicates limited etiological role. However,
significant associations were obtained between AR-CAG repeat lengths and Gleason
Grade, as well as age at diagnosis, in African-American and Latino men. No
significant association between AR-CAG repeat length and disease risk in Chinese
men was observed. We conclude that although germline AR-CAG variation is
involved in PCa progression, somatic mutations in the AR play virtually no role in
untreated PCa,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1
INTRODUCTION
The American Cancer Society estimates that over 230,000 new cases of prostate
cancer will be reported in 2004 in the United States with nearly 30,000 deaths
occurring (American Cancer Society (ACS)). Estimates for California account for
approximately 10 percent of each of these statistics. These projections infer that
prostate cancer is the most common type of cancer diagnosed among men in 2004,
accounting for 33 percent of all new cancer cases. Additionally, prostate cancer will
also be responsible for 10 percent of all cancer related deaths among men in 2004,
second only to lung cancer (ACS). For the period 1996 to 2000, the age-adjusted
mortality rate for prostate cancer in the United States was 32.9 per 100,000 people
(ACS).
Testosterone and DHT are two steroid hormones that are involved in the
regulation of several biological processes related to male growth and development.
The androgen receptor (AR) signaling pathway is responsible for regulating all these
processes. The AR, a class I nuclear receptor, has been implicated in the etiology of
prostate cancer and has been studied extensively. Of particular interest is the possible
involvement of somatic mutations in the AR in the development, progression, and
epidemiology of prostate cancer. While somatic mutations have been shown to exist
in tumors in men who have undergone androgen ablation therapy, which suggests an
androgen-depleted environment provides a selection pressure for certain mutations,
little is known about somatic mutations in untreated tumors.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2
We performed case-only and case-control studies on three separate populations
of men in order to: (i) Assess the frequency of somatic mutations in prostate cancer
tumors that have not been subjected to androgen ablation therapy and: (ii) to assess the
possible role of AR-CAG repeat polymorphism in disease predisposition and
progression. In addition to reporting our findings in these studies, this thesis will
provide basic background information on the molecular biology of the AR and the
epidemiology of prostate cancer.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3
THE PROSTATE
Basic Anatomy
In order to understand the complexities of cancer and other diseases of the prostate,
some basic information regarding this gland is essential. To begin, the prostate, along
with the seminal vesicles and bulbourethral glands, provides secretions that make up
the seminal fluid. The prostate has been described as resembling a chestnut in
appearance and being conical in shape (Van de Voorde 1996). The gland surrounds
the urethra with the base of the prostate situated against the bladder and the point
forming the prostate apex (Timms 1997; Netter 2001). The prostatic urethra, which
serves as an anatomical reference point, runs through the middle of the gland between
the prostate apex and bladder neck with an approximated 35 degree angle occurring in
the center (McNeal 1988a). Extending from this angle towards the prostate apex is
the verumontanum, which is the posterior wall of the prostatic urethra and is the point
where the prostatic utricle and the ejaculatory ducts join to the prostatic urethra
(McNeal 1972; McNeal 1981a; McNeal 1988a). There is a considerable protuberance
in the prostatic urethra in this region, specifically near the proximal end of the
verumontanum. In addition to the prostatic utricle and ejaculatory ducts, many
prostatic ducts with evenly distributed acini form junctions with the prostatic urethra
at this location (McNeal 1988a). The prostatic urethra is differentiated into two
segments at the proximal end point of the verumontanum (McNeal 1981a; McNeal
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4
1988a). The distal segment of the prostatic urethra includes the verumontanum and
the urethral section that extends to the prostate apex. The prostatic urethra between
the verumontanum and the bladder neck is referred to as the proximal segment. Since
the prostate is an anatomically diverse structure, this simplified description of prostate
anatomy, which is for the most part a summary of McNeal’s work, is intended to serve
as a reference and it does not represent the true complexity of the gland.
History o f Prostate Anatomy
Lowsley provided one of the earliest in-depth anatomical descriptions of the human
prostate in 1912 when he observed five distinct lobes (two lateral lobes, a posterior
lobe, a middle lobe, and an anterior lobe) in embryonic and fetal prostates with each
lobe designating a location of prostate ducts budding from the prostatic urethra
(Lowsley 1912). Lowsley concluded that these five lobes, which comprise the
prostate, develop from five separate groups of tubules that first appear in
approximately the twelfth week of embryo development (Lowsley 1912). Much
debate over the anatomy of the prostate has occurred since Lowsley’s description of
these prenatal prostatic lobes. Several authors, including Franks, Huggins, Hutch,
Tisell, and McNeal, challenged previous anatomic, histologic, and pathologic
descriptions of the prostate during the 1900’s, reflecting the complex nature of the
gland (McNeal 1980). Throughout the century, inconsistent terminology also
contributed to confusion regarding anatomical descriptions.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5
McNeal’s work in prostate anatomy, spanning nearly forty years, demonstrates
some of the most extensive study of the gland. McNeal points out that Franks
dismissed Lowsley’s concept of lobes in the adult prostate in 1954 (McNeal 1981b).
The theory of histologic homogeneity within the gland remained intact until 1968
when McNeal reported histologic heterogeneity in the glandular region (or in his
words “true prostate”) of the prostate that resulted in the identification of two distinct
zones which he designated the “central” and “peripheral” (McNeal 1968). In 1981,
McNeal described four unique anatomical regions that formed a more complete
picture of the zonal anatomy of the prostate: the peripheral zone, the central zone, the
preprostatic region, and the anterior fibromuscular stroma (McNeal 1981b). In a 1988
publication, McNeal refined his zonal anatomy of the prostate by detailing glandular
and nonglandular regions (McNeal 1988a). McNeal’s zonal anatomy represents the
culmination of decades of study on the prostate and currently serves as the best model
for prostate anatomy.
Zonal Anatomy o f the Prostate
There are four distinct types of nonglandular tissue in the prostate: the anterior
fibromuscular stroma, preprostatic sphincter, striated sphincter, and the prostatic
capsule (McNeal 1988a). The prostate is surrounded, except for areas near the
prostate apex and bladder neck, by a prostatic capsule that is composed of a layer of
smooth muscle encapsulated by an outer membrane comprised of collagen (McNeal
1988a). The majority of the prostate, except for the most distal and proximal regions,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6
is contained within the prostatic capsule. The anterior fibromuscular stroma is a
structure comprised of smooth muscle cells that spans the anterior prostate from the
bladder neck to the prostate apex (McNeal 1981b; McNeal 1988a). Although the
anterior fibromuscular stroma does not contribute to prostatic function, it does
influence the overall shape of the prostate. The proximal prostatic urethra is sheathed
in the preprostatic sphincter, which is comprised of smooth muscle (McNeal 1984;
McNeal 1988a). The anterior section of the distal prostatic urethra is covered by
skeletal muscle and is termed the (striated) prostatic sphincter (McNeal 1972). This
striated sphincter is an extension into the prostatic capsule of the external sphincter,
which is distal to the prostate apex (McNeal 1972; McNeal 1984). Nerves and
vascular tissues located within the prostate can also be considered nonglandular.
The glandular region of the prostate is posterior to the prostatic urethra and is
connected via a series of ducts to the urethra that are utilized in the secretion of the
prostatic fluid, which makes the prostate an exocrine gland. The glandular region
consists of three zones: the transition zone, peripheral zone, and central zone (McNeal
1988a).
The transition zone, which McNeal first classified as a component of the
preprostatic region in 1981 (McNeal 1981b), envelops the proximal urethra and
accounts for approximately five percent of the glandular region (McNeal 1981a;
McNeal 1981b; McNeal 1988a). The ducts of the transition zone originate at the
approximate juncture of the distal and proximal prostatic urethra (McNeal 1981b). In
1988, McNeal et al reported that approximately 20% of prostate cancers originate in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7
the transition zone (McNeal 1988b). In a 2003 study of transition zone carcinomas in
Western Australian men, Shannon et al reported findings that suggests a portion of
transition zone cancers, once thought the majority of which were clinically irrelevant,
are capable of progression and could result in poor prognoses (Shannon 2003). The
transition zone is the primary origin of benign prostatic hyperplasia (BPH), which can
be treated by a procedure known as a transurethral resection of the prostate (TURP)
(McNeal 1981a). BPH can also be found in another area McNeal classified as a
component of the preprostatic region in 1981, the periurethral gland region. This
small glandular area of the proximal prostatic urethra represents less than one percent
of the total glandular region of the prostate (McNeal 1981a).
The central zone, which accounts for approximately 25% of the glandular
region, surrounds the ejaculatory ducts and is enveloped by much of the peripheral
zone (McNeal 1981b). The ducts of the central zone originate from a concentrated
region of the verumontanum, branch out towards the most proximal portion of the
prostate, and are considerably larger than the ducts of the peripheral and transition
zones (McNeal 1988a). The base of the prostate is composed of the most developed
portion of the central zone. Nearly 10% of prostate cancers arise in the central zone,
which was once thought to be resistant to disease (McNeal 1981a; McNeal 1988b).
The peripheral zone, which surrounds the distal urethra posterior to the
transition zone, accounts for approximately 70% of the glandular region (McNeal
1981a). The ducts of the peripheral zone originate from the prostatic urethra in two
lines spanning the length of the distal segment (McNeal 1988a). Since the peripheral
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8
zone occupies such a large percentage of the glandular region of the prostate, it is
reasonable to hypothesize that more cancers would appear in this zone compared to
the other zones of the prostate. It is estimated that approximately 70% of prostate
cancers develop in this zone (McNeal 1988b). The peripheral zone is the site of origin
for most prostate carcinomas (McNeal 1981b) and prostatitis (inflammation or
infection of the prostate) (Van de Voorde 1996). Another condition that affects this
zone almost exclusively is prostatic intraepithelial neoplasia (PIN) (Haas 1997), which
is marked by cellular alterations that occur in the epithelial lining.
Histology o f the Prostate
The prostatic epithelium is spread over a layer of fibromuscular stromal cells (not to
be confused with the anterior fibromuscular stroma zone) and consists of three cell
types (Bonkhoff 1993b; Feldman 2001). The first cell type, columnar secretory cells,
lines the ductal lumen, is androgen dependent, and is responsible for secreting
prostatic fluid and proteins such as prostate specific antigen (PSA) (Feldman 2001;
Wang 2001b; Marker 2003). Schalken and van Leenders reported that these secretory
cells are well differentiated, have little ability for proliferation, and have a high
apoptotic index (Schalken 2003).
The second cell type, basal cells, forms a basal membrane between the
secretory cells and fibromuscular stroma, are undifferentiated, have an increased
ability to proliferate, and have a low apoptotic index (Wang 2001b; Schalken 2003).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9
Basal cells are known to express estrogen receptor (ER), progesterone receptor (PR),
and androgen receptor (AR). AR expression in basal cells, however, is considerably
weaker compared to AR expression in secretory cells (Bonkhoff 1993b). Although
prostatic basal cells express AR, they are generally considered to be androgen
independent. It has been proposed that a subpopulation of cells located in the basal
layer may be a form of stem cell, capable of differentiating and responding to
androgens (Bonkhoff 1993a; Wang 2001b; Schalken 2003). There is currently much
debate regarding basal cells, the basal layer, and prostatic cell differentiation and
regeneration. While basal cells are considered to be capable of differentiating into
secretory and neuroendocrine cells (Schalken 2003), possibly due to the effects of
androgens (Bonkoff 1993a), they are not required for this differentiation (Kurita
2004).
The third cell type found in prostate epithelia is the previously mentioned
neuroendocrine cell, which as they are named, possess characteristics of endocrine
cells and neurons. Neuroendocrine cells could possibly influence prostatic growth
regulation, differentiation, and secretion (Abrahamsson 1993; di Sant’Agnese 1996).
The distribution of these cells throughout the prostate has resulted in much research
regarding their specific functions. Although these cells are found throughout all areas
of the prostate, they have been found to exist in higher numbers in the major ducts
(Abrahamsson 1993) and are more abundant in the peripheral zone compared to the
transition zone (Van de Voorde 1996).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10
The fibromuscular stroma that underlies the epithelial layer is comprised of a
myriad of cell types, including smooth muscle cells, fibroblasts, endothelial cells,
dendritic cells, mast cells, and lymphocytes. Some of these cells may respond to
androgens or be responsible for growth factor production (Feldman 2001).
Farnsworth suggests that growth factors influencing interactions between stromal and
epithelial cells may be the basis for structural and functional differences that exist
between the extremes of the ducts of the glandular epithelium (Farnsworth 1999).
While AR in the stroma regulates prostatic development, growth, and regeneration
(Kurita 2004), neuroendocrine cells may be involved in stroma/epithelial interactions
that are related to prostate function (Abrahamsson 1993).
The relationships between the various cell types and their contributions to
prostatic growth, development, and maintenance are representative of the complexity
of the prostate itself. While a great deal is known about the signaling mechanisms of
specific cells and their involvement in prostate function, relatively new areas of
research in prostate cellular biology, such as investigations of prostatic stem cells, are
revealing that much is still unknown.
In summary, the prostate is a highly organized exocrine gland consisting of
three specific zones that comprise glandular and nonglandular regions. The three
zones of the prostate, as defined by McNeal, designate specific regions of prostatic
function and disease. Prostatic diseases including BPH and cancer can affect these
zones to different extents. Finally, the cellular components of the glandular region,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
11
specifically the secretory, basal, and neuroendocrine cells, are responsible for the
function, development, and maintenance of the gland via an intricate signaling
network.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
12
EPIDEMIOLOGY OF PROSTATE CANCER
United States Epidemiology
Like many cancers, the incidence of prostate cancer increases with age. Prostate
cancer is rarely diagnosed in men younger than age 50 and increases with age at a
faster pace than any other cancer (Haas 1997). While the incidence rate of prostate
cancer is increasing in men younger than 65, more than 70% of all prostate cancers are
found in men older than 65 (ACS 2004). For the years 1998 to 2000, the American
Cancer Society reports that the probability of developing prostate cancer for a man in
the United States over the course of his life was 17.15% (approximately one in every
six men) (ACS 2004).
An interesting pattern of prostate cancer incidence in the United States
emerged in the late 20th century. Many authors have noted that annual prostate cancer
incidence rates increased dramatically from 1987 due to PSA screening, peaked in
1992 (190.1 per 100,000 men (Crawford 2003)), decreased for three years
(approximately 10.8% per year (Hankey 1999)), and then returned to pre-PSA
screening levels (168.9 per 100,000 men (Crawford 2003)) between 1995 and 1999
(Hankey 1999; Parkin 2001; Quinn 2002; Crawford 2003; Schaid 2004). In 1998,
Legler et al reported that prostate cancer incidence rate trends observed between 1988
and 1994 in the United States were nearly identical to the trends for first-time PSA
tests for the same period (Legler 1998). The authors concluded that the increase, peak,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
13
and ensuing decline in incidence rates were related to the proportion of men receiving
PSA tests for the first time. The trend in annual mortality rate for prostate cancer in
the United States followed a pattern similar to the incidence rate by increasing rather
dramatically from the late 1980’s to a peak of 26.7 per 100,000 men in 1992 with a
subsequent decrease (Crawford 2003; ACS 2004).
It is well documented that incidence and mortality rates vary among ethnic and
racial groups in the United States (Haas 1997; Parkin 2001; Chu 2003; Crawford
2003; Cunningham 2003; Gronberg 2003; Ntais 2003). Age adjusted incidence rates
for prostate cancer during the period 1996 to 2000 in the United States were highest
among African-American men (272.1 per 100,000), followed by Whites (164.3 per
100.000), and Hispanics/Latinos 137.2 per 100,000) (ACS 2004). The lowest
incidence rates for the period were found in Asian Americans and Pacific Islanders
(100.0 per 100,000) and American Indians and Native Alaskans (53.6 per 100,000)
(ACS 2004). According to the American Cancer Society, the age-adjusted mortality
rates of prostate cancer during the period 1996 to 2000 for African-American men
(70.0 per 100,000) were more than double the mortality rates for White men (30.2 per
100.000) in the United States (ACS 2004). In 1995, Ndubuisi et al reported that in
addition to having higher incidence rates, African-American men were more likely to
be diagnosed with distant disease than White men in the District of Columbia, even
after adjustment for socio-economic status (SES), which resulted in worse prognoses
and a disparity between mortality rates. (Ndubuisi 1995). While American Indians
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
14
and Native Alaskans had higher incidence rates than Asian Americans and Pacific
Islanders, the mortality rates for American Indians and Native Alaskans (13.9 per
100,000) were lower than Asian Americans and Pacific Islanders (21.9 per 100,000)
(ACS 2004). The observed disparities in incidence and mortality rates for prostate
cancer across racial and ethnic groups in the United States and other countries have
resulted in theories such as genetic susceptibility, androgen metabolism, and ethnic
dietary differences being associated with the disease (Shibata 1997; Parkin 2001).
However, several authors have referred to studies showing that people who have
immigrated to another country, such as Japanese men to the United States, develop
incidence and mortality rates approaching those of the new nation of residence, which
conflict theories regarding genetic predisposition (Haas 1997; Shibata 1997; Parkin
2001; Crawford 2003; Gronberg 2003).
World Epidemiology
Numerous authors have cited the fact that prostate cancer is currently the sixth most
common form of cancer in the world (Parkin 2001; Quinn 2002; Gronberg 2003;
Deutsch 2004). As the third most common cancer diagnosed in men and with over
half a million estimated new cases being diagnosed in 2000, the incidence of prostate
cancer is increasing worldwide (Hsing 2000a; Parkin 2001; Quinn 2002; Gronberg
2003). At the turn of the century, the worldwide prevalence of prostate cancer was
estimated to be over 1.5 million and prostate cancer incidence has steadily increased
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
15
1.7% per year globally since the mid 1980’s (Parkin 2001). Although the increase in
worldwide prostate cancer incidence may be stable, incidence varies by geographic
region. Hsing et al reported in 2000 that the largest increases in incidence occurred in
high-risk countries such as the United States and Canada while Iow-risk countries
experienced smaller increases in incidence (Hsing 2000a). Aside from differences in
geographically related potential risk factors for prostate cancer, there are other
possible reasons for the disparities in incidence between countries. For example,
Hsing theorized that along with deficient cancer registries and poor diagnosis, low-risk
countries had insufficient prostate cancer screening compared to high-risk countries
and that may have an effect on differences in the magnitudes of increase in the
incidence rates (Hsing 2000a).
As previously stated, African-American men in the United States are
considered to have some of the highest incidence rates in the world while residents of
Tianjin, China have the lowest (Crawford 2003; Gronberg 2003). From 1988 to 1992,
the highest incidence rates of prostate cancer occurred in the United States, Canada,
Sweden, Australia, and France (Hsing 2000a). The incidence rates in the United
States, however, were considerably higher than any other country. In 2002 Quinn and
Babb reported that the age-standardized incidence rates of prostate cancer before 1992
in the United States were more than double the incidence rates found in Sweden and
Australia, more than triple the rates in Europe, and ten times the rates found in Asian
countries such Japan, India, and China (Quinn 2002). This diversity in incidence rates
among countries represents differences in prostate cancer risk in various geographic
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
16
locations. After 1992, the incidence rates in France, Canada, and Australia decreased
similarly to the trend in prostate cancer incidence in the United States, (Parkin 2001).
While the incidence may have decreased temporarily after the peak in 1992, an
overall increase in incidence is continuing in both high-risk and low-risk countries;
however, the reasons for these increases differ among countries representing each
level of risk (Hsing 2000a; Parkin 2001; Gronberg 2003). There is currently much
debate over what is actually influencing the increases in incidence rates that are being
observed in different geographic locations. Hsing’s previously mentioned theory is
just one example of many ideas that have been put forth in attempts to explain
differences in increasing incidence rates among countries. Gronberg postulates that
much of the variation in incidence that exists in different countries and ethnic groups
can be attributed to a mixture of the varying effects of differences in external risk
factors, genetic factors, and health care (Gronberg 2003). In low-risk countries, such
as China and Japan, Parkin suggests that while some of the increase in incidence may
be due to improved detection and diagnostic methods, increased consumption of fat
and meat may also be contributing to the rising incidence (Parkin 2001). It is also
entirely possible that epidemiologists in some countries may just be mistaken when it
comes to estimating prostate cancer incidence rates. Osegbe challenged the notion of
Nigeria being classified as a low-risk nation and determined the incidence rates of
prostate cancer in that country may have been egregiously underestimated, by nearly
13-fold, and may in fact be comparable to the rates found in African-American men in
the United States (Osegbe 1997).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
17
High-risk countries, such as the United States, also show significantly better
survival of prostate cancer than low-risk countries, but Parkin contends that this is
most likely the result of screening and detection of latent cancer (Parkin 2001).
Similar to what is occurring in the United States, scientists in other countries are
attempting to answer the questions regarding the relationship between PSA screening
and prostate cancer mortality rates. In 1999, Labrie et al reported the results from a
randomized controlled trial of prostate cancer screening involving PSA testing in
metropolitan Quebec, Canada (Labrie 1999a). The findings in this study, which
involved more than 46,000 men, revealed a 69% reduction in cause-specific death due
to prostate cancer in screened vs. unscreened men. It must be noted, however, that the
Labrie report generated much debate over the association between PSA screening and
mortality rate reduction, which highlights the need for carefully executed studies
regarding PSA screening and mortality (Alexander 1999; Boer 1999; Labrie 1999b;
Labrie 1999c; Labrie 2002).
PSA Screening
During the last two decades, the means of prostate cancer diagnosis has affected
incidence rates substantially. In 1995, Potosky et al concluded that the tremendous
increase in incidence of prostate cancer observed in the United States between 1986
and 1991 was due primarily to advanced detection of disease by PSA screening
(Potosky 1995). The United States Food and Drug Administration (FDA), in 1986,
originally approved the PSA test for the purpose of monitoring prostate cancer
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
18
progression in patients (Hankey 1999; Schaid 2004). However, in approximately
1988, medical professionals began using the test as a method for
screening for prostate cancer. Potosky et al reported that between 1986 and 1988
(before the PSA test became widely utilized for screening) the increasing incidence of
prostate cancer could be attributed to increasing use of two other methods of detection:
transrectal ultrasound (TRUS) and needle biopsy, which strengthened the argument
that improved diagnostic methods may affect incidence rates (Potosky 1995). After
six years of being used as a method for detecting prostate cancer, the FDA officially
approved the test for the purpose of screening in 1994 (Platz 2004a; Schaid 2004).
The American Cancer Society recommends that men over age 50 with an expected life
span of greater than 10 years be screened annually for prostate cancer with the PSA
test and the digital rectal exam (DRE) (ACS 2004). For high-risk individuals, such as
African-American men and men with a first degree relative diagnosed with prostate
cancer, the ACS recommends screening should begin at age 45.
While epidemiologists are attempting to determine the risks and benefits of
widespread screening, there is currently much research being done in the area of
improving the PSA test as well. A PSA serum concentration greater than 4.0 ng/ml
typically warrants a needle biopsy and is considered to be a potential indicator of
prostate cancer with a sensitivity of approximately 67.5 to 80% (Carroll 2001). Gann
et al reported in 1995 that lower PSA concentrations, in the range of 2.0 - 4.0 ng/ml,
could be indicative of prostate cancer but reasoned lowering the threshold to these
values could decrease specificity (Gann 1995). Conversely, Punglia et al advocate the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
19
lowering of the PSA threshold to 2.6 ng/ml because it would increase sensitivity with
only a minimal reduction in specificity (Punglia 2003a). Highlighting the debate over
and importance of acceptable sensitivity and specificity, Stamey et al point out that
Punglia’s suggested threshold of 2.6 ng/ml is not only indicative of prostate cancer,
but also BPH (Stamey 2004). The discussions regarding the PSA concentration
threshold represent only one side of the problem involving sensitivity and specificity.
Besides the PSA concentration threshold, Gann et al concluded that sensitivity and
specificity could also be affected by patient age and cancer aggressiveness (Gann
1995). Additionally, the PSA test is vulnerable to verification bias that can artificially
inflate the sensitivity and decrease the specificity. In a study of the effects of
verification bias on PSA screening, Punglia et al concluded that sensitivity and
specificity could be improved if verification bias is controlled for (Punglia 2003a;
Punglia 2003b). Other methods to improve the sensitivity and specificity of PSA
screening currently being explored are age-specific PSA, PSA velocity, and prostate
volume-adjusted PSA (Wilt 2003). Although improvements in PSA screening
techniques are valuable, understanding the risks (common false positives, needless
biopsies, harmful therapies, and excessive anxiety) and benefits of early detection may
at the present time be of more importance.
In 1999, Hankey et al provided evidence that showed PSA screening had an
effect on the incidence and mortality of prostate cancer for several years in the United
States (Hankey 1999). The authors suggest that an observed decrease in distant
disease incidence, the spectacular increase in incidence followed by a decrease, the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 0
decrease in age-specific mortality rates that represented a period effect, and lead-time
effects indicate incidence and mortality trends are directly related to PSA screening.
However, Hankey and colleagues also hypothesize that an observed decrease in age-
specific mortality rates could be the result of attribution bias (Feuer 1999) and
advances in therapy such as androgen blockade (Hankey 1999). In 2002, Chu et al
concluded that a disease stage incidence trend is related to the decrease in age-adjusted
mortality observed in the early 1990’s in African-American and White American men
(Chu 2003). The authors contend that ameliorated survival of those diagnosed with
distant disease does not contribute to the association between observed decreases in
distant disease incidence and mortality and therefore, PSA testing is responsible for
the decrease in mortality. Conversely, in a study of prostate cancer incidence and
mortality trends, Coldman et al concluded in 2003 that PSA screening had no effect on
the decreases in mortality that were observed in the 1990’s in British Columbia
(Coldman 2003). Wilt points out that prostate cancer mortality, compared to
incidence, was low before PSA testing and advanced treatments (Wilt 2003). In
addition, the low mortality may possibly be due to the slow progression of the disease
and the fact that frequent diagnosis occurs in the elderly, who likely die of causes
other than the cancer (Wilt 2003). Citing the standard value of 4.0 ng/ml as the
cutpoint for detection, the United States Preventive Services Task Force (USPSTF)
verified the ability of the PSA test to detect prostate cancer in an early stage but
determined the evidence for PSA testing was unconvincing in regards to reducing
mortality (USPSTF 2002a; USPSTF 2002b). Weinmann et al reported in 2004 that a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 1
matched case-control study of prostate cancer screening and mortality resulted in a
reduction in mortality that could be attributed to screening by DRE and PSA testing;
however, the net effect of each of the screening methods separately could not be
determined (Weinmann 2004). It should be noted, however, that the Weinmann study
had low statistical power and the results were questionable in general. The disparities
in these reports (Hankey 1999; USPSTF 2002b; Chu 2003; Coldman 2003; Wilt 2003;
Weinmann 2004) represent the perplexity of the relationship between PSA screening
and mortality rates and indicate the need for thorough and accurate analyses of the
risks and benefits of widespread screening.
The current lifetime risk of death due to prostate cancer is currently 3.4%
(including men who do not seek treatment), which suggests most individuals who are
diagnosed with prostate cancer do not die as a result of the disease (Wilt 2003). This
is an important fact to bear in mind when considering the risk/benefit analysis of PSA
screening and any potential impact on mortality rates. Sirovich et al reported that
more men aged 50 and older had been screened for prostate cancer than colorectal
cancer, which has been confirmed by randomized controlled trials (RCT) to be
beneficial in reducing mortality, even though the advantages of PSA screening are
vague (Sirovich 2003). The study found that by 2001, 75% of American men aged 50
and older had ever had a PSA screening test, compared to 63% for colorectal cancer,
and the probability of a man having a PSA test increased with age until age 80
(Sirovich 2003). The USPSTF reports that there are two RCTs currently ongoing that
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 2
will hopefully elucidate any potential advantages to prostate cancer screening;
however, it may be several years before results are reported (USPSTF 2002b).
Potential Risk Factors
Since prostate cancer is a multifactorial disease, the identification of potential risk
factors has been a daunting task that has frequently produced inconclusive results and
contradictory conclusions among studies. Risk factors for prostate cancer have been
reported to include tobacco smoking (Haas 1997; Hickey 2001; Roberts 2003;
Oefelein 2004), alcohol (Dennis 2001; Platz 2004b; Schoonen 2005), dietary factors
(Chan 2001a; Chan 2001b; Kolonel 2001; Parkin 2001; Crawford 2003; Gronberg
2003; Nelson 2003; Tseng 2004), sexual behavior (Haas 1997; Rosenblatt 2001; Giles
2003; James 2004; Leitzmann 2004), androgen levels (Gann 1996; Nelson 2003; Ntais
2003), BPH (Haas 1997; Guess 2001; Hammarsten 2002), prostatitis (Nelson 2003;
Platz 2004c; Roberts 2004), specific genetic factors (Shibata 1997; Parkin 2001;
Gronberg 2003; Nelson 2003; Ntais 2003; Deutsch 2004; Schaid 2004), family history
(Cunningham 2003; Gronberg 2003; Deutsch 2004; Schaid 2004; Negri 2005),
race/ethnicity (Shibata 1997; Gronberg 2003; Ntais 2003; Schaid 2004), and age (Haas
1997; Parkin 2001; Crawford 2003; Gronberg 2003; Schaid 2004). However, the
American Cancer Society considers only the latter three as being confirmed risk
factors (ACS 2004).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
23
THE ANDROGEN RECEPTOR
Androgens
Crystallized extracts from male sexual organs, theorized to have been purified
androgens, were being used in China as early as 200 BC for treating people who
appeared to have a deficiency in activity consistent with being male (Liao 2002). The
most abundant form of human androgen is testosterone, which is produced by
conversion of androstenedione primarily in the testes by Leydig cells and in lesser
amounts by peripheral tissue (Chung 2002; Lee 2003). In the prostate, the testosterone
metabolite 5a-dihydrotestosterone (DHT) is produced by 5a-reductase 1 and 2 (Eder
2001; Lee 2003). Whereas DHT functions as a more potent androgen than
testosterone, responsible for the development of prostate, bulbourethral glands, and
other reproductive structures, DHEA and androstenedione are weaker androgens than
testosterone (Chung 2002).
Circulating androgens, under strict regulatory control by the hypothalamus-
pituitary-testis axis (Chung 2002), exist primarily in two forms, testosterone and DHT.
Testosterone and DHT may be bound to sex hormone binding globulin (SHBG) and
albumin or they can exist in non-bound forms that are free to bind and activate the AR
(Lee 2003). Testosterone and DHT are the primary hormones responsible for male
development during embryogenesis (Eder 2001). In addition, these two hormones are
involved in the regulation of several other biological processes related to growth and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 4
development, including spermatogenesis (McPhaul 2002). The androgen receptor-
signaling pathway is responsible for regulating many of these processes.
History of Androgen Receptor Research
In the early half of the 20th century, the structures and identities of male hormones
were determined and purified androgens were first utilized in hospitals as medical
treatments (Liao 2002). Charles Huggins and Clarence V. Hodges discovered in 1941
that metastatic prostate cancer, which at the time was known to be associated with
increased activity of serum acid phosphatases, responds to androgen ablation therapy
via castration or estrogen injection (Huggins 2002). Liao et al demonstrated that
androgens play a vital role in the regulation of gene expression in the 1960’s by
affecting amino acid incorporation into protein (Liao 1962) and by controlling mRNA
levels (Liao 1965). In order to understand the mechanisms by which these events
occurred required further investigation into the androgen-signaling axis.
In the late 1960’s, the discovery of the AR was marked by the publications of
three separate studies. In July of 1968, Anderson and Liao concluded that a tissue-
specific receptor for DHT existed in prostatic nuclear chromatin (Anderson 1968). In
November of the same year, Bruchovsky and Wilson reported that radioactive
testosterone and DHT were bound to nuclear protein, suggesting the existence of a
potential receptor (Bruchovsky 1968). In 1969, Mainwaring verified the existence of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
25
a “steroid receptor” protein with androgen specificity and limited tissue distribution
(Mainwaring 1969). The findings in these studies not only resulted in the
identification of the AR; they verified DHT, as opposed to testosterone, as the primary
ligand for the receptor.
After the discovery of the AR, studies were undertaken to identify the
chromosomal location of the sequence coding for the protein. In 1970, Lyon and
Hawkes investigated testicular feminization (Tfm), which was theorized to be related
to androgen insensitivity, and determined that the Tfm gene was X-linked (Lyon
1970). In 1975, Meyer and colleagues studied genetic transmission of androgen
insensitivity and determined that a DHT receptor was X-linked and corresponded to
Lyon’s Tfm locus in mouse (Meyer 1975). In 1981, Migeon et al mapped the AR to
the centromeric region between Xpl 1 and Xql3 on the human X chromosome
(Migeon 1981). Based on the knowledge that androgen insensitivity syndromes
involve the AR, follow X-linked inheritance, and that no other steroid receptor genes
exist on the X chromosome, Lubahn and colleagues were able to narrow down the
location of the AR to an area between the centromere and Xql3 (Lubahn 1988).
Lubahn et al used a method that involved screening an X chromosome library with an
oligonucleotide probe, that was generated from homologous sequences of the DNA-
binding regions of other known human steroid receptors, to locate fragments for
cloning (Lubahn 1988). Along with Lubahn and colleagues, Chang et al and Trapman
et al also reported success with AR cloning in 1988 (Chang 1988a; Trapman 1988). In
1989, Brown et al refined the location of the AR gene to Xql 1-12 (Brown 1989). In
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 6
the same year, Kuiper and colleagues reported that the AR gene exists as a single copy
consisting of eight exons and seven introns, which extends across more than 90 kb of
chromosomal DNA (Kuiper 1989). The chromosomal location of the AR has now
been identified as Xql 1.2-12.
The Androgen Receptor Gene and Transcription
The AR, a class I nuclear receptor, is encoded by eight exons comprising 2.7 kb,
which are separated by seven introns that represent more than 79 kb (Heinlein 2002).
The exons range in size from 117 bases (exon three) to 1.6 kb (exon one) (Tilley 2003)
and the introns range in size from 700 bases (intron seven) to over 26 kb (intron one)
(Kuiper 1989; Heinlein 2002). Exons one and eight contain a 5’ untranslated region
(5’ UTR) and a 3’ untranslated region (3’ UTR), respectively (Figure 1). The
functional domains encoded by these eight exons will be discussed in the next section,
Structure and Function of the Androgen Receptor. The 5’ UTR is approximately 1.1
kb and begins with either one of two transcription start sites (TIS) at positions +1 and
+12 (Tilley 1990). The large 3’ UTR is approximately 6.8 kb and has two functional
sequences for polyadenylation (Faber 1991).
A large promoter region that contains many transcriptional regulatory elements
extends for thousands of bases upstream from the 5’ UTR (Heinlein 2002). The AR
promoter region does not contain TATA or CCAAT elements (Tilley 1990), which
characteristically indicates the existence of multiple transcription initiation sites (Lee
2003). In 1993, Faber et al identified a minimal promoter element that spans positions
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 7
-5 to +57 (Faber 1993) and in 1996, Takane and McPhaul identified a core promoter
bounded by positions -74 to +87 (Takane 1996). Analyses of the AR promoter have
involved a wide range of regions resulting in the identification of several sequences
that could potentially regulate transcription (Tilley 1990; Mizokami 1994a; Takane
1996; Chen 1997; Kinoshita 2000).
FIGURE 1. Structure of the androgen receptor gene
5’
US la u d II
3'
-600
-541 -500 -120 -60 -46 t>
i i i
I I i l l
NF1 CRE
Purflyr I Exon 1 2 3 4 5 6 7 8 |
5’ UTR 3' UTR
The human androgen receptor gene spans approximately 90 kb. Eight exons and
seven introns comprising more than 82.0 kb are Hanked by 5' and 3’ UTRs. The
ORF (shaded) represetted by exons 1-8 is encoded by 2.7 kb. A 1.1 kb 5' UTR is
located at the beginning of exon one. Transcription initiation sies (TIS) I and II are
located at positions+1 and +12, respectively. A large 3' UTR of approximately 7.0
kb is located at the end of exon eight. The promoter region spans several kb,
however only 600 bases are represented here. Sequences for transcription factor
recognition are located at positions -46 (Spl), -508 (CRE), and - 541 (HF1). The
homoputineihoniDpyrimidine (homo punfryr) stretch spans positions -120 to -60.
(Lubahn 1989; Tilley 1990,2003; Fabtr 1991,1993; Heirilein 2002)
A binding site for Spl transcription factor is located within a GC box at position -46
and a homopurine/homopyrimidine (homo pur/pyr) region is found between positions
-120 and -60 (Tilley 1990; Tilley 2003). The Splbinding site and homo pur/pyr
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 8
regions are conserved sequences in the AR promoters of human, rat, and mouse (Lee
2003). Sequences recognized by helix-loop-helix DNA binding proteins are located at
positions -161 to -156 and -19 to -14 (Takane 1996). Mizokami and colleagues
identified several sequences involved in transcriptional regulation, such as activating
protein-1 (AP-1), cyclic AMP response element (CRE), nuclear factor-1 (NF-1),
retinoic acid response element (RARE), and IL-6 (interleukin-6) response element,
throughout 2.3 kb of the promoter region (Mizokami 1994a). Including many of these
response elements, there are at least 22 potential transcription factor binding sites
known to exist within AR promoters in many species (Chen 1997). While there are
many potential binding sites for transcription factors, some appear to function in a
species-specific manner. Although NF-1 and nuclear factor kB (NF-kB) binding
have been shown to be involved in AR transcriptional suppression in rat, it is
unknown how these transcription factors function in human AR transcription
(Heinlein 2002).
Two distinct mechanisms exist for transcription initiation from TIS I and TIS
II, which are located at positions +1 and +12, respectively (Faber 1993). Faber et al
reported in 1993 that transcription from TIS I, which produces the major transcription
product, was dependent upon the minimal promoter sequence located between
positions -5 and +57 and transcription from TIS II was influenced by Spl binding at
the -46 position (Faber 1993). Spl binding, however, did not affect transcription
initiation from TIS I. Pugh and Tjian reported that Spl may recruit transcription
factor HD (TFIID) and TFIID binding protein leading to the formation of the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 9
transcription preinitiation complex (Pugh 1991). Chen et al determined that multiple
weak Spl binding interactions occur in the homo pur/pyr region upstream from
position -46 and postulated that this activity may represent a source of Spl for the
binding motif associated with TIS II transcription (Chen 1997). Based on deletion
analyses, it was theorized that the homo pur/pyr region might also play a role in
transcription initiated from TIS I, possibly via general transcription factor associations
or by acting as a transcription initiator in a similar fashion to a TATA box (Mizokami
1994a).
Emphasizing the prevalence of species-specific regulatory sequences located
within AR promoters, a CRE element has been defined in the mouse AR promoter
(Lindzey 1993), while one has not been identified in the rat AR promoter (Lee 2003).
In 1992 Blok et al reported that dibutyryl-cAMP (dbcAMP) affected transcriptional
regulation of the AR gene in a cell type-dependent manner, even though a CRE
element was not known to exist at the time (Blok 1992). In a series of experiments,
the authors demonstrated that dbcAMP stimulated AR expression in Sertoli cells,
however in LNCaP cells, there was no effect on expression. In 1994 Mizokami et al
observed that a functional CRE, with a two base mismatch compared to the putative
CRE sequence, exists at position -508 in the AR promoter (Mizokami 1994a).
Transcription may also be influenced by methylation occurring in a CpG island
of approximately 3 .0 kb that spans the AR promoter region, the 5’ UTR, and the
coding sequence of exon one (Kinoshita 2000). Epigenetic mechanisms involved in
transcriptional regulation of the AR may play important roles in prostate cancer. It has
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
30
been shown that the AR promoter of normal prostate epithelial tissue, as well as the
AR expressing prostate cancer cell lines LNCaP and PC-3, is not hypermethylated,
while non-AR expressing cell lines, such as Dul45, DuPro, and PPC-1, demonstrated
various degrees of promoter methylation (Jarrard 1998). The aforementioned studies
illustrate the complexity involved with AR transcription that results from regulatory
elements and regions, various transcription factors, and methylation of the AR
promoter. Additionally, these components represent only a fraction of what is known
about AR transcriptional regulation.
As previously stated, the human AR mRNA contains 5’ and 3’ UTRs.
Mizokami and Chang determined that the 5’ UTR is vital to AR mRNA translational
control (Mizokami 1994b). The infrequency of mutations in the 5’ UTR, including
somatic mutations in prostate cancer cases, indicate the sequence is highly conserved,
reflecting functional importance (Cabral 2004). In contrast to the 5’ UTR, the 3’ UTR
provides a source of variation at the mRNA level. Faber and colleagues reported the
existence of two mRNA species for the human AR with sizes of 11.0 and 8.5 kb,
suggesting the possibility of alternative splicing of the 6.8 kb 3’ UTR (Faber 1991).
Fischer et al reported that AR mRNA isoforms might be involved with hormone-
dependent cell proliferation and represent potential mechanisms of transcriptional
regulation of AR target genes (Fischer 1993).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
31
Structure and Function of the Androgen Receptor
In 1988, Chang et al reported the human AR has a molecular mass of approximately
94 kDa and is comprised of 918 amino acids (Chang 1988b). A year later, Tilley et al
found that the AR consisted of 917 amino acids, coding for a protein of 98,918 Da
(Tilley 1989). Brinkmann et al also reported in 1989 that the cDNA sequence of the
AR had an open reading frame of 2,751 bases, which encoded a protein comprised of
917 amino acids with a molecular mass in excess of 98 kDa (Brinkmann 1989). In the
same year, Lubahn et al identified the intron/exon splice sites in the human AR and
reported a coding region of 2,757 bases, which corresponded to an AR containing 919
amino acids (Lubahn 1989). In 1991, Kuiper et al determined that the AR is produced
as a 110 kDa protein that undergoes post-translational modifications, resulting in a 112
kDa protein (Kuiper 1991). Five years later, in 1996, Trapman and Brinkmann
referred to a 2,730 base sequence encoding a 910 amino acid AR. (Trapman 1996).
Variation in the size of the AR is due to two polymorphic trinucleotide repeat
regions in exon one that obviously affect the number of amino acids in the protein
(Trapman 1996). Variable length polyglutamine (CAGn ) and polyglycine (GGNn )
repeats in the first exon are responsible for variations in the total number of AR amino
acids. For example, the 917 amino acid AR defined by Tilley contained 20 glutamines
between residues 58 and 77 and 23 glycines between residues 448 and 470 (Tilley
1989). In contrast, the 919 amino acid AR defined by Lubahn contained 21
glutamines between residues 58 and 78 and 24 glycines between residues 449 and 472
(Lubahn 1989). Two shorter polyglutamine regions are found at amino acid locations
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 2
84-89 and 193-197 and polyproline and polyalanine repeats are marked by amino
acids 372-379 and 398-402, respectively (Faber 1989; Lubahn 1989). The Androgen
Receptor Gene Mutations Database (ARDB 2004) utilizes a 919 amino acid AR model
for assigning mutation locations (Gottlieb 2004). For this manuscript a sequence of
919 amino acids with variable polyglutamine and polyglycine tracts will be referenced
(Figure 2).
FIGURE 2. Structure of the androgen receptor coding regions
5’ UTR
Paly- CAG Fo^-GGN
Eight exons repre setting 11.0 kb are separtedby varible length introns. The
ORF (shaded) encodes a protein of 919 amino acids. The 5 ’ and 3 ’ UTRs are
located at 1he beginning of exon one and end of exon eight; respectively. The
poly-Gln(amino acids 58 -78) and pbly-Gly(amino acids 449 - 472) regons
are located in the ORF of exon one (Lubahn 1989; Heinlein 200 2; Tilley 2003).
Exon 1
3’ UTR
In addition to the existence of mRNA isoforms, two different AR protein
isoforms, A and B, are generated from a single mRNA through the utilization of two
translation initiation sites (Figure 3 A) (Hirata 2003). The AR-B isoform is the fully
intact wild type (WT) receptor, whereas the AR-A isoform contains a truncated N-
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
33
terminal end. Co-transfection experiments failed to detect any significant differences
in the functional activities of these isoforms; however cell- and promoter-specific
differences in functionality were observed (Gao 1998).
FIGURE 3. Structures of androgen receptor mRNA and protein
1100 1665 3857
+1 ATG ATG TGA
f*
|
UTR Exon 1 2 3 4 5 6 7 8 3’ UTR |
B
I
TRANSLATION
360-535
AF-5
110-360
AF-1 504-535
559-624
i-DBD—
625-669
•HMGE-t
FXXLF
23-27
-Fbly-Gln
58-78
>-Poly-GV
WXXLF +49-472
667-919
- L B D —
TGA
919
874-910
AF-2
1
433-437
QHF
670-673
A ARrcRNA. Two ATG sites gve nse to two ARisofbrms. AR-B (ATG
1100) is the larpr isofbrm, whereas the smaller AR-A (ATG 1665) isoform
lacks 188 N-terrrinal amino adds.
B. AR-B Protein The NTD is encoded by exon one. AF-1, AF-5, and
LIAF regions are tocated within the NTD. Coding fbrFXXLF and WXXLF
motifs are also found in the NTD. Regons of exons two and three encode
the DBD. Exons three and four both encode portions ofthe hinge regon
Coding for the LBDis comprised of exons four, five, six, seven and eight
The QPIF regon defines the boundary of 1he hinjp regonand the LBD. AF-
2 is located in the LBD. Amino add ranges coding for these regions are
listed above orbdowthe corresponding domain The coding regons ofthe
exons are as Mows;
Exon 1 residues 1-537 Earn 5 724-772
Exon 2 538-588 Exon 6 773-815
Exon 3 589-627 Exon7 816-867
Exon 4 628-723 Exon 8 868-919
Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission.
3 4
Once translated, the unliganded AR is associated with a multiprotein
chaperone complex that is comprised of Hsp40, Hsp70, Hsp90, Hop/p60, BAG-1 and
the immunophilins Cyp40 and FKBP51/52 (Heinlein 2002). The chaperone complex
provides the capability for the AR to bind ligand, which occurs in the cytoplasm and
results in the chaperone complex dissociating from the AR (Tilley 2003). After ligand
binding and chaperone complex dissociation, the AR is able to homodimerize possibly
via an interaction between the amino- and carboxyl-terminal domains (N/C
interaction) and translocate to the nucleus via the action of importin-a and -p (Black
2004). Once inside the nucleus, the AR can bind DNA and associate with
coregulators (Tilley 2003). Coactivators, such as SRC-1, GRIP-1/TIF2, and
p300/CBP enhance the transcriptional ability of the AR on target genes (Tilley 2003)
while corepressors, such as SRY, HDAC1, and NcoR, inhibit the transcriptional
activity of AR (Wang 2005). This simplified description of AR activation represents a
ligand-dependent pathway. However, ligand-independent activation of the AR can
also occur through such means as HER-2/neu and IL-6 via the MAPK, STAT3, and
Akt pathways (Culig 2003). The steps involved in the process of AR maturity from
translation to transactivation are dependent on specific functional domains and
subdomains encoded in the AR. Like other members of the nuclear receptor
superfamily, the AR contains an N-terminal domain (NTD), a DNA-binding domain
(DBD), and a ligand-binding domain (LBD) (Figure 3B). These three domains
perform vital functions that allow the AR to operate as a transcription factor in the
form of a ligand-dependent homodimer.
Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission.
35
Similar to the chicken PR, the NTD of the human AR is encoded entirely by
exon one (Faber 1989) and accounts for approximately 59% of the entire protein.
However, the NTD is the least conserved domain found in steroid receptors (Tilley
1989). In 1989, Faber et al reported that the AR-NTD does not appear to share any
apparent homology with the NTD of other steroid receptors, although there was
similarity in the amino acid content to the human PR-NTD (Faber 1989). In the same
year, Brinkmann et al theorized that a high acidic amino acid content of a region
between amino acids 100 and 325 of the AR-NTD might be related to transcriptional
regulation (Brinkmann 1989). In 1995, Jenster et al published the discovery of two
transcription activation regions in the AR-NTD encoded by amino acids 1 - 485 and
360 - 528 (Jenster 1995). Since then, these two “transcription activation units”,
referred to as TAU-1 and TAU-5 by Jenster, have been renamed activation functions
AF-1 (amino acids 110 - 360) and AF-5 (amino acids 360 - 535), respectively (Scher
2004). In the AR these activation functions overlap but are considered separate
entities with unique sequence characteristics (Jenster 1995). The activation functions
regulate AR activity on target genes in a cell- and promoter-specific manner
(Buchanan 2001a). A region, encoded by amino acids 502 - 535, is associated with
both ligand-dependent and ligand-independent transactivation (Buchanan 2001a) and
has been shown to interact with p i60 coactivators, CBP, and TFIIF (Irvine 2000).
Other features of the AR-NTD include two LXXLL-like motifs, the FXXLF
and WXXLF, which span amino acids 23-27 and 433-437, respectively (Figure 3B)
(Tilley 2003). The LXXLL-like motif 2 3 FQNLF2 7 has been found to be an integral
Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission.
3 6
component in DHT-induced N/C interaction (Shen 2005). However, N/C interaction
may enhance transactivation activity but it does not appear to be necessary for this
activity. Coregulators such as GRIP1 (Shen 2005), ARA70, and SRC-1 (Hsu 2005)
have been shown to influence the N/C interaction, possibly by acting as a bridging
molecule. In general, N/C interaction, transactivation activity, and coregulator effects
together represent a very complex interrelationship in the function of AR and the
details of which are beyond the scope of this manuscript.
Also included in the AR-NTD are the previously mentioned two polymorphic
trinucleotide microsatellites. The length of the C AG has been shown to have a direct
affect on transcriptional activity of the AR (Irvine 2000; Beilin 2000). In addition, the
length of the CAG repeat has been shown to be associated with PCa risk (Irvine 1995;
Ingles 1997; Stanford 1997; Hsing 2000b; Balic 2002), and Kennedy’s disease (Poletti
2004). Whereas PCa has typically been associated with shorter CAG repeats and
GGC repeats (Chang 2002), Kennedy’s disease is associated with expansions of the
CAG repeat (Hirawat 2003).
The AR-DBD is encoded by amino acids 559 - 624 (Chang 1988b; Buchanan
2001b) and shares significant homology with the DBD of the human glucocorticoid
receptor (GR) and PR (Brinkmann 1989). The DBD of steroid receptors is highly
conserved and is marked by the presence of eight cysteines, which form the basis for
two zinc fingers (Riegman 1991; Heinlein 2002). These two zinc fingers are involved
in receptor homodimerization and DNA recognition via binding to the major groove
(Tilley 2003). Jenster et al demonstrated that removal of the AR-DBD resulted in a
Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission.
3 7
receptor incapable of transactivation (Jenster 1991). The AR-DBD enables AR
homodimers to bind androgen response elements (AREs) of androgen-regulated genes.
AREs are typically hexameric half-sites containing either inverted or direct repeats
that are separated by three base pairs of spacer DNA (Shaffer 2004). Examples of
AREs include a specific sequence, AGAACAgcaAGTGCT, found in the PSA gene
promoter (Riegman 1991) and a consensus sequence, GGA/TACAnnnTGTTCT,
identified by Roche et al (Roche 1992). However, the actual mechanisms involved
with target gene expression are not simply a matter of the AR binding to an ARE. It
has recently been reported that the AR-NTD may be involved in DNA binding
selectivity (Brodie 2005). This suggests that in addition to intermolecular N/C
interactions that have been shown to enhance transactivation of target genes,
intramolecular mechanisms may also be involved in AR function.
Marking the boundary of the AR-DBD and AR-LBD is the hinge region,
which shows little homology with the corresponding domain o f other steroid receptors
(Tilley 2003), encoded by amino acids 625 - 669. The AR-DBD/hinge region
contains a nuclear localization signal (NLS) that has been mapped to an area
encompassed by amino acids 617 - 633 (Zhou 1994). In the absence of ligand, AR
mutants with this region deleted remained in the cytoplasm; however, the addition of
ligand (R1881) resulted in translocation to the nucleus (Jenster 1991). Although this
suggests that the putative NLS serves an important purpose in AR function, it also
indicates that there may be a secondary, ligand-dependent signal for nuclear
localization as well. The hinge has also been implicated in regulation of
Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission.
38
transactivation via post-translational modifications (Heinlein 2002) and by inhibiting
AF-2 activity in the LBD (Wang 2001a). In addition, some studies have shown that
the AR-DBD/hinge region functions as a protein-protein interaction site with AP-1
with both stimulatory (Bubulya 1996) and inhibitory effects (Wang 2001a) on
transactivation being reported. The approximate location where the hinge meets the
AR-LBD corresponds to amino acids 670 - 673 (Q, P, I, and F residues) and this
region has also been shown to be associated with AR transactivation (Buchanan
2001b).
The human AR-LBD is encoded by amino acids 667 - 919, (Chang 1988b;
Buchanan 2001b) and, like the AR-DBD, exhibits considerable homology with the
corresponding domain of the human GR and PR (Brinkmann 1989). In sequence
comparisons with M. musculus, R norvegicus, S. canaria, O. cuniculus andX. laevis
the human AR-LBD shared over 80% homology (Poujol 2000). The AR-LBD
contains the highly conserved “signature sequence” of steroid receptors, which spans
amino acids 703 - 732 (Buchanan 2001a). The ligand-dependent activation function,
AF-2, is located in helix 12 of the AR-LBD between amino acids 874 - 910 (Tilley
2003). Deletion studies have shown that truncation of the AR that results in the
removal of either all or part of the AF-2 prevents binding of the ligand metribolone
(R1881) (Jenster 1991). Additionally, the AF-2 serves as the primary binding site for
p i60 coactivators, which have an LXXLL-like motif (NR Box) that facilitates the
interaction (Buchanan 2001a; Shen 2005). These p i60 coactivators will then recruit
Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission.
3 9
secondary cofactors that will enhance receptor activity (Tilley 2003). In addition, the
AF-2 can form an N/C interaction with the 2 3 FQNLF2 7 of the AR-NTD (Shen 2005).
In summary the AR is capable of functioning as a ligand-dependent and
ligand-independent homodimer. Three functional domains, the NTD, DBD, and LBD,
which contain several sub-components, can influence nuclear localization,
homodimerization, coregulator recruitment, intramolecular signaling, and
transactivation of androgen regulated genes. The AR-DBD and AR-LBD share
considerable homology with the corresponding domains of other steroid receptors;
however, the AR-NTD is not highly conserved.
Androgen Receptor, Mutations, and Prostate Cancer
The first mutation in prostate cancer was found in the LNCaP cell line (T868A) and
was reported in 1990 by Veldscholte et al (Veldscholte 1990). As of 2004, there have
been 605 mutations identified in the AR (Gottlieb 2004), which makes the AR the
most commonly mutated transcription factor (Montgomery 2001). More than 500
different mutations associated with androgen insensitivity syndrome (AIS) have been
identified in the AR (Gottlieb 2004) and 72% have been reported to collocate to
regions that represent 11% of the coding sequence (Scher 2004). In stark contrast,
there have only been 85 mutations identified in prostate cancer (Gottlieb 2004) and
they collocate primarily to six specific regions that collectively represent 15% of the
AR coding sequence. It has been estimated that 86% of all LBD prostate cancer
Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission.
4 0
mutations collocate to three regions, amino acids 670 - 678, 701 - 730, and 874 - 910,
that are discrete from AIS mutations (Figure 4) (Scher 2004).
7 IGURE 4. Frequencies and locations of AR mutations in disease
AR-NTD (1-537)
I I i l l 1 1 I LLU I I U I _______
A.
hPCa
B.
hPCa
IE
S4-78
llL
I
2S6-268 502-S35
AR-LBD (667-919)
670-678 701-730 874-910
c.
AIS
1 m k L
700-712 730-784 827-870
A. Prcporticsul fveqa«Mcie of nutation; occunmg in the AR-NTD in prostata cancer Shaded
areas with the amino acid range helow indicate prostate cancel mitationcollocation sites
8. Proportional frequencies o f mutation* occurring in the AR-LBD in prostate cancer. Shaded
areas with the atnno acid range helovr indicate prostate cancer nmlationcolLocatioat sites.
C. ftojortional frequencies o f mutations occurring in the AR-LBD in AIS. Shaded areas with O re
amino acid range below indicate AS mntalum collocation sites.
Vertical bars cotraspond to the percentage o f total mtssewe reutahoro found irteach domain
Image adapted from Schoret al 2CD4
Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission.
41
Several studies have investigated mutations that occurred in post-androgen
ablation therapy tumors (Culig 1993; Taplin 1995, 2003; Haapala 2001; Hyytinen
2002). These studies reported frequencies of mutations in 7% (Culig 1993), 10%
(Taplin 2003), 33% (Hyytinen 2002), 36% (Haapala 2001), and 50% (Taplin 1995).
Studies of primary prostate tumors prior to androgen ablation therapy have suggested
the presence of mutations may be as high as 36% (Thompson 2003) to 44% (Tilley
1996). Other studies have suggested lower frequencies of 0% (Marcelli 2000), 2.5%
(Schoenberg 1994), and 3.8% (Newmark 1992). Most of these studies have been
small, with less than 40 subjects, while the Marcelli study consisted of 99 pretreated
tumor samples. With only one large study performed, the role AR mutations play in
primary prostate tumor progression prior to androgen ablation therapy is still
debatable.
Although prostate cancer mutations have been found in partial and complete
AIS (PAIS and CAIS) these occurrences have been very rare (Yong 2000; Mooney
2003). Typically, prostate cancer mutations are somatic, gain-of-function mutations,
whereas AIS mutations are germline, loss-of-iunction mutations (Gottlieb 2004).
However, there are a few examples of germline AR mutations associated with male
breast cancer (Wooster 1992; Lobaccaro 1993) and prostate cancer (Koivisto 2004).
Androgen receptor mutations can impact prostate cancer progression through
enhanced activation by adrenal androgens, non-androgenic steroids, and anti
androgens (Culig 2003). Typically, specific activation of the AR occurs with native
ligands; however, mutant AR may utilize alternative ligands, which results in
Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission.
4 2
promiscuous activation of the AR. In addition, coactivators such as ARA70 have also
been implicated in reduced ligand specificity (Heinlein 2004) and hypersensitivity of
the AR to adrenal androgens (Rahman 2004) in prostate cancer.
As stated above, the AR-C AG repeat has been found to affect AR
transcriptional activity as an inverse function of AR-CAG repeat length, which may
contribute to prostate cancer risk (Irvine 2000; Beilin 2000; Buchanan 2004). The
normal ranges of lengths for the polyglutamine tract have been reported to be 6 - 36
with an average polyglutamine length of 20 (Giovannucci 1997; Buchanan 2004).
Studies of racial-ethnic groups and AR-CAG repeat length have consistently found
shorter CAG repeats in African-Americans compared to other groups (Sartor 1999;
Bennett 2002). Whereas African-Americans have been shown to have the shortest
CAG lengths, relatively speaking, Whites and Latinos are slightly higher (Balic 2002),
and Asians have the longest CAG repeat tracts (Irvine 1995). Findings such as these
have fueled theories that the AR-CAG length may confer prostate cancer risk since
African-Americans and Asians have the highest and lowest risk, respectively (ACS
2004). However, over the years the literature has become increasingly murky
regarding the association between AR-CAG length and prostate cancer risk.
Some studies have shown an association between short CAG repeats and
increased risk for PCa (Irvine 1995; Ingles 1997; Stanford 1997; Hsing 2000b; Balic
2002), while other studies have reported no such association (Bratt 1999; Chen 2002).
In regard to prostate cancer grade, associations between short CAG and higher grade
have been reported (Giovannucci 1997) but findings that refute these observations
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
43
have also been published (Bennett 2002; Stanford 1997; Gsur 2002). In 2004 Zeegers
et al reported a significant association between shorter CAG repeat and increased risk
but they also found an absolute difference of less than one CAG between cases and
controls, which questions the biological significance of the shortened CAG on cancer
progression (Zeegers 2004). In the largest study to date, Freedman et al were unable
to find a statistically significant association between AR-CAG length and prostate
cancer risk (Freedman 2005). Studies that indicated significant associations exist
between short CAG repeats and increased risk for PCa facilitated experiments that
demonstrated increased transcriptional activity of AR with shorter CAG repeats and
generated theories that suggested a transcriptionally more active AR influences PCa
progression (Ding 2004; Wang 2004). Buchanan et al, however, studied the activity
of AR with various sized CAG repeats and a mutant AR with markedly increased
transcriptional activity that contained a polyglutamine tract that was interrupted by the
presence of two leucine residues (Buchanan 2004). The results of the studies suggest
that CAG repeats outside a critical range of 16 - 29 repeats would mediate disease
phenotype, rather than dichotomous values that represent short or long, since AR with
CAG repeats within the critical range can be considered functionally equal. As a result
of this finding, future epidemiologic studies of the association between CAG repeat
length and PCa risk need to address the biological aspects of AR function.
In summary, few prostate cancer mutations have been identified compared to
AIS mutations. In addition, PCa and AIS mutations collocate to distinct regions of the
AR. Mutant AR can affect the progression of PCa by promiscuous activation,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4 4
utilizing adrenal androgens, non-androgenic steroids, and anti-androgens. While PCa
mutants are typically regarded as somatic, gain-of-fimction mutations, AIS mutations
are usually germline, loss-of-fimction mutations. Shorter CAG length results in an AR
with increased activity; however the associations between CAG length and prostate
cancer risk and grade are inconclusive. Finally, some small studies have suggested
AR mutations may have a significant involvement in PCa progression prior to
androgen ablation therapy; however, one large study refutes these findings, which
indicates more studies are needed.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
45
STUDY OVERVIEW
The main objectives of this study were to (i) identify somatic mutations in the AR and
determine the frequency that they occurred in primary prostate tumors prior to
androgen ablation therapy in Chinese, African-Americans, and Latinos and (ii)
determine if the AR-CAG repeat polymorphism is associated with overall PCa risk in
Chinese as well as age of disease onset and/or disease Gleason Grade in Chinese
living in Singapore and African-Americans and Latinos living in Los Angeles.
It is well documented that AR mutations have been found in prostate cancers
occurring after hormone ablation therapy (Culig 1993; Taplin 1995; Tilley 1996;
Haapala 2001). However, information regarding AR mutation frequency in untreated
prostate cancers (PCa) is limited. In addition, early studies focused on screening
exons two through eight since the GC rich exon one, which comprises 58% of the AR,
was difficult to PCR amplify. In 1996 Tilley et al observed base changes in the AR
that lead to amino acid substitutions in 11 out of 25 primary tumors (44%) with 50%
of the mutations occurring in the NTD (Tilley 1996). This result suggested pre
existing mutations in the AR might be selected for in prostate tumors during the
progression to androgen independence instead of being acquired following androgen
ablative treatment. In 2000, Marcelli et al reported that 8/38 (21%) pre-androgen
ablation stage Di prostate cancers contained a total of 11 mutations (Marcelli 2000).
Citing the fact that there are few studies involving the identification of AR mutations
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4 6
in pre-treatment prostate cancers, Thompson et al looked at 14 primary tumors and
observed frequent missense mutations (29%) in high-grade PCa before androgen
ablation therapy (Thompson 2003). Based on findings such as these that suggest a
relatively large proportion of prostate tumors contain AR mutations, we attempted to
determine the precise frequency of somatic mutations in prostate tumors that were not
previously subjected to androgen ablation therapy in three distinct study populations.
One study, the Singapore Study, involved men of Chinese descent from the
Singapore Chinese Health Study cohort (Hankin 2001). Virtually all Singapore
Chinese originated from a defined geographic location in Southern China. Therefore,
this study population represents a genetically more homogeneous population in
contrast to studies involving American populations, which are subject to genetic
admixture. The genetic homogeneity that is offered by a genetically isolated
population provides an advantage over genetically heterogeneous populations in the
assessment of gene-disease associations (Lander 1994). Genetic homogeneity is not
relevant when researching somatic mutations and therefore does not have any bearing
on our case-only analyses. However, genetic homogeneity is desirable for
investigating germline genetics and will be advantageous in our case-control analyses.
Subjects in this study did not receive hormone ablation therapy. Therefore, any
verifiable somatic mutations identified in the course of our study would not be the
result of selection pressure due to androgen deprivation.
In our second study, the Los Angeles Study, two racial-ethnic groups (African-
American and Latino) from the Los Angeles area were represented. As with the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4 7
Singapore Study, the subjects for this study did not receive androgen ablation therapy,
negating the possibility for somatic mutations resulting from androgen deprivation
selection pressure.
In both of the above-mentioned studies, the AR-CAG repeat polymorphism was
also examined for associations with Gleason Grade, race, and age at diagnosis, as well
as in a case-control analysis in Chinese. While many studies have investigated the
AR-CAG repeat polymorphism (Ingles 1997; Giovannuci 1997; Hakimi 1997;
Stanford 1997; Bratt 1999; Chen 2002; Gsur 2002; Tsujimoto 2004), most of these
studies have been limited to Caucasian men and few have investigated study
populations similar to ours (Hsing 2000b; Xue 2000; Bennet 2002; Balic 2002).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4 8
MATERIALS AND METHODS
Study Design
Initially one of our objectives was to identify somatic mutations in the AR gene
and measure the frequency of these mutations in prostate tumors in different racial-
ethnic groups as a function of Gleason Grade. We hypothesized that gain-of-fimction
mutations in the AR gene are selected for in advanced tumors. Secondly, we wished
to investigate the associations between polyglutamine repeat number (C AG„) and
tumor grade in each of our two study populations. Thirdly, we wanted to examine the
potential association between CAG repeat length and prostate cancer predisposition in
Singapore Chinese men in a case-control analysis. In order to accomplish these goals,
we intended to PCR amplify and sequence the entire androgen receptor gene for each
subject in our two study populations. However, due to cost constraints, we decided to
modify our plan and only investigate six specific regions that were previously shown
to be sites of mutation collocation (Buchanan 2001a). Three of these collocation sites
are found in exon one (encoding the AR-NTD) and three are found in exons four, five
and eight, which encodes parts of the LBD. In the NTD, the three mutation
collocation sites targeted for PCR amplification corresponded to amino acids 58-78
(the polyglutamine repeat), 256 - 268 (a region with potential transactivation activity),
and 502 - 535 (an area of coactivator contact, the LIAF) (Figure 5 A.) (Buchanan
2001a; Scher 2004; WD Tilley personal communication with GA Coetzee).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4 9
FIGURE 5. Location of mutation collocation sites in the AR
A.
110
h-
AR-NTD (1-537)
AF1
3 6 0
I —
AF5 53S
HI
PolyQ
M3
58-78 256-268 502-535
B ,
AR-LBD (667-919)
: ■
i 1
670-678 701-730 874-910
A. Three c olio cation sites in the NTD are indicated b y shaded regons and the
corresponding amino acid residues are listed below. The amino acid residues
corresponding to AF-1, AF-5, and the poly^utamine regonare indicated above the
diagram
B. Three collocation sites in 1he LBD are represented by shaded re gons with the
corresponding amino acid residues listed below the diag'am
Together, the three regions of mutation collocation in the NTD represent 7% of the
total coding sequence of the AR and 12% of the NTD. Approximately 72% of all
NTD mutations collocate to these sites. Since the NTD is GC rich (a problem for PCR
amplification) we used Cloned Pfu Polymerase (Stratagene product #600159), which
is a high fidelity enzyme (error rate 1.3 x 10"6 ) (Cline 1996), for all PCR
amplifications. Pfu is unique among commercially available temperature resistant
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
50
polymerases in that it is stable up to 98°C allowing GC rich templates to melt
completely for efficient amplification. In the LBD, the three sites of mutation
collocation represent amino acids 670 - 678 (QPIF region), 701 - 730 (region of the
steroid receptor signature sequence), and 874 - 910 (AF-2) (Figure 5B). In 2004
Scher et al reported that 86% of all LBD mutations collocate to these three regions,
which accounts for 7% of the AR coding sequence and 29% of the LBD (Scher 2004).
By focusing on these six regions of mutation collocation in the NTD and LBD we
saved money at the cost of a minimal reduction in our ability to detect all possible PCa
mutations in the AR; we reasoned that this way we should be able to detect at least 70-
80% of mutations if they exist also in our study populations at the collocation sites as
defined above.
For both studies, Singapore and LA, we designed a case-only analysis for
comparison of normal adjacent tissue sequences to tumor tissue sequences, whereas
case-control analysis was utilized in addition in the Singapore Study for assessment of
the potential association between CAG repeat length and prostate cancer risk. We
were blinded to the racial-ethnic background (for the LA Study), Gleason Grade, and
age of diagnosis of each subject in the studies until completion of DNA sequencing.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
51
Study Subjects
Singapore Study
This study population involved men of Chinese descent from the Singapore Chinese
Health Study cohort, which was established between 1993 and 1998 (Hankin 2001).
This cohort was created from residents or citizens of Singapore who live in
government-built housing estates. Materials for the case-only analyses consisted of
123 microdissected paraffin embedded tissue slides, corresponding to 123 study
subjects. We were provided with information on these subjects that included age at
diagnosis, source hospital, Gleason grade, and tumor stage (TNM staging for only 37
subjects), to which we were blinded. We excluded 15 of these subjects (subjects scy5,
scy7, scy9, scylO, scyl4, scy20, scy24, scy26, scyl56, me82, mel28, nuhl62,
nuhl94, nuh203, and nuh239) from analyses since we were unable to identify either
normal adjacent or tumor tissue on the paraffin embedded tissue slides. For two
subjects (subjects scy75 and scy76) PCR amplification failed for both the normal
adjacent and tumor tissue for several mutation collocation sites, which resulted in the
exclusion of these subjects. These 17 exclusions left us with 106 subjects, equaling
212 total normal adjacent and tumor tissue samples for our case-only sequence
analyses. Of these 106 remaining subjects, seven (subjects scy23, scy31Bl, scy35Bl,
scyl38, scyl41, scyl42, and scyl53) were determined to be doubles after sequencing
and were subsequently excluded from statistical analyses. Of the remaining 99
subjects, 12 subjects (scy30, scy32B2, scy50, scy79, scy85, scy86, scyl27, scyl34,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
52
scyl46, scyl58, gmc47, and nuhl66) were of mixed race (Indonesian Chinese,
Malaysian Chinese, or Non-Singaporean Chinese), were diagnosed with squamous
metaplasis, or PCR amplification of the AR-I mutation collocation site failed, resulting
in our inability to sequence the polyglutamine repeat microsatellite and were excluded
from statistical analyses. After we finished with sequencing, we matched the Gleason
grade and age to each of the subjects involved in this study. The total number of
subjects used for statistical analyses was 87 and we were able to obtain mutation
collocation site sequence information for 106 subjects.
For the case-control analysis of this study, we were provided with information
on age when blood was drawn for AR gene sequencing and CAG repeat length for 85
unmatched Singapore Chinese controls from the Singapore Chinese Health Study
cohort (Hankin 2001). These 85 controls and our 87 cases were used to study the
association between CAG repeat length and prostate cancer risk after adjustment for
age.
Los Angeles (LA) Study
Subjects in this study were participants in the Multiethnic Cohort Study (MEC), which
consists of more than 200,000 men and women, enrolled between 1993 and 1996,
aged 45-75 residing in Hawaii and California (Kolonel 2000). The California Cancer
Registry and the Surveillance, Epidemiology, and End Results (SEER) cancer
registries in Hawaii and Los Angeles track incident cancer cases in the MEC
(Freedman 2005). Our sample set for sequencing consisted of 90 paraffin embedded
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
53
tissue slides, each slide corresponding to a study subject. We were able to obtain AR
mutation collocation site sequence data for all but one subject (subject 39) because
PCR amplification failed. This subject was therefore excluded from analyses. Upon
completion of sequencing, we were able to match the racial-ethnic background,
Gleason grade, and age to each of the subjects involved in this study. Of the 89
remaining subjects, we were provided with a second slide of paraffin embedded tissue
for each of three subjects (subjects 21, 58, and 62). These extra slides did not differ in
AR sequence or Gleason grade from the initial slide and were therefore excluded from
analyses. In addition, information regarding age at diagnosis was missing for one
subject (subject 48), who was excluded, resulting in a dataset consisting of 85 subjects
with sequence information for our case-only analyses. Our subjects were of either
African-American or Latino descent. However, ten subjects were of mixed racial-
ethnic background and were excluded from our analyses. One subject (subject 40)
was of African-American and “Other” background, two subjects (subjects 79 and 80)
were of African-American and White background, one subject (subject 19) was of
African-American, Latino, and White background, two subjects (subjects 20 and 43)
were of African-American, White, and “Other” background, and four subjects
(subjects 57, 58, 65, and 83) were of Latino and White background. Of these
remaining 75 subjects, one subject (subject 3) was missing Gleason Grade information
and was excluded from histologic grade analyses. Our final study sample set contained
75 subjects for statistical analyses and 89 subjects with mutation collocation site
sequence information.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
54
Proteinase K Digestion
For the Singapore Study, microdissected paraffin embedded tissue slides from each
subject were stained and analyzed by Dr. Louis Dubeau of the USC/Norris Cancer
Center. For the LA Study, microdissected paraffin embedded tissue slides from each
subject were previously stained and analyzed by the laboratory of Dr. Richard Cote of
the USC Pathology Department. Staining and analysis resulted in regions of tumor
and normal adjacent tissue on each slide being demarcated. We estimate that the
regions marked as tumor were 90% tumor. Next, a stained and analyzed slide was
used as a tracing guide for a corresponding unstained slide for marking regions of
normal adjacent and tumor tissue. A marked unstained slide was then used for DNA
extraction for normal adjacent and tumor tissue for each subject. Two 0.5 ml reaction
tubes for Proteinase K (Sigma Aldrich product #P2308) digestion were used for each
subject, one for normal adjacent tissue and one for tumor tissue. For the Singapore
study, each tissue sample was digested in a 50.0 pi reaction volume consisting of 2.5
pi of Proteinase K stock solution (20.0 mg/ml), 5.0 pi Proteinase K lOx reaction
buffer (100.OmM Tris-Cl, l.OmM EDTA, 5% Tween 20), and 42.5 pi sterile H2O . For
the LA Study, each tissue sample was digested in a 100.0 pi reaction volume
consisting of 5.0 pi of Proteinase K stock solution, 10.0 pi Proteinase K lOx reaction
buffer, and 85.0 pi sterile H2O . Approximately one microliter of the reaction was
applied to the demarcated area (either normal adjacent or tumor tissue) on a slide and a
sterile scalpel (Feather brand disposable #15 scalpels product #25-2975) was used to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
55
scrape the tissue and reaction liquid together. The resulting liquid-tissue mixture was
then deposited into the appropriately marked Proteinase K digestion tube, gently
mixed by flicking the bottom of the tube, and overlayed with one drop of mineral oil.
The process of scraping the microdissected slides was performed in a sterile, plasmid-
free hood with sterile scalpels to minimize the risk of contamination. The samples
were then placed into a thermal cycler (MJ Research, Inc. PTC-100) for the digestion.
Reaction conditions were 55C for five hours followed by 95C for ten minutes to
inactivate the Proteinase K.
PCR Amplification
For all samples of the two studies, a first round PCR and nested round PCR were
performed for all six mutation collocation targets. The region 701 - 730,
corresponding to target AR-V, contains the C-terminal end of exon four and the N-
terminal end of exon five. Therefore, in order to PCR amplify this region of mutation
collocation, we created one set of primers (for each first round and nested round
amplification) for the exon four component (target AR-V4) and one set of primers for
the exon five component (target AR-V5). As a result, there are seven sets of primers
each for the first round and nested round of PCR, which correspond to the six regions
of mutation collocation. Nested PCR amplification product sizes are listed in Table 1.
TABLET Nested PCR product sizes for seven mutation collocation site targets
Collocation Site Target AR-I AR-II AR-III AR-IV AR-V4 AR-V5 AR-VI
Size of Product (bp) 227 100 238 116 144 89 214
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
56
First Round PCR
All first round PCR reactions were prepared in a sterile, plasmid-free hood that is used
for PCR only. Designated first round PCR pipettors and barrier resistant pipet tips
were used to minimize the risk of contamination. All reaction solutions were
combined in a master reaction tube on ice. The master reaction was calculated to
include total number of samples, five negative controls, and one positive control.
Each sample reaction consisted of 0.5 pi Cloned Pfu Polymerase (Stratagene product
#600159), 2.5 pi lOx Reaction Buffer for Cloned Pfu Polymerase, 1.25 pi Dimethyl
Sulfoxide (Sigma Aldrich product #D2650), 0.5pl lOmM dNTPs (Sequencing Grade
dNTP Set from Amersham, product #27203501), 1.0 pi sense primer (10.0 pMols per
pi), 1.0 pi antisense primer (10.0 pMols per pi), and 16.25 pi sterile H2O for a final
volume of 23.0 pi.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
57
TABLE 2. First round reaction primers
For each mutation collocation site target region, the sense primer is listed above the
antisense primer. Primers for AR-I were designed by Howard C. Shen (GA Coetzee
Laboratory, University of Southern California). The reverse primer for AR-III and
forward primer for AR-IV were designed by Lubahn et al (Lubahn 1989). All other
primers were designed with the aid of Vector NTI software (InforMax-Invitrogen).
Target
Name First Round Primers________________ Exon Location________ X
5' CTTTCC AGAATCTGTTCCAGAG 3'
AR-I 5' CCT CAT CCA GGA CCA GGT AGC C 3'
EXON 1 55°C
5' GTG TGG AGG CGT TGG AGC AT 3'
AR-I I 5' GAA CCT TTG CAT TCG GCC AA 3'
EXON 1 48°C
5' CAT CCT GGC ACA GTG TCT TCA C 3'
AR-III 5' AG A ACA CAG AGT GAC TCT GCC 3'
EXON 1 62°C
5’ ACC AGC CCC ACT GAG GAG ACA A 3'
AR-IV 5' TGC AAA GGA GTC GGG CTG GT 3' EXON 4 62°C
5' AGG TGT AGT GTG TGC TGG AC 3'
AR-V4 5' CCA CTT CCC TTT TCC TTA CC 3' EXON 4 53°C
5' TAC CCA GAC TGA CCA CTG CC 3'
AR-V5 5' AAA CAC CAT GAG CCC CAT CC 3' EXON 5 61 °C
5' GAG GCC ACC TCC TTG TCA ACC CTG 3'
AR-VI 5' GGG GTG GGG AAA TAG GGT TT 3’ EXON 8 53°C
Each sample reaction, including positive and negative controls, was aliquoted from the
master reaction into individual wells of a 96-well PCR plate (USA Scientific product
#1402-9596) on ice. Next, 2.0 pi of each sample and positive control genomic DNA
were added to appropriate reaction wells. Nothing was added to the five negative
controls. The PCR plate was sealed using Microseal A Film (MJ Research, Inc.,
product #MSA-5001) and placed into a thermal cycler (MJ Research, Inc. PTC-100)
for the reaction. Reaction conditions were 98°C for two minutes for initial template
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
58
denaturing followed by 30 cycles of 98°C for one minute, X°C for one minute (X =
annealing temperature for target region primers, see Table 2), and 74°C for one
minute. A final extension at 74°C for four minutes followed the 30 cycles.
Nested Round PCR
Designated nested round PCR pipettors and barrier resistant pipet tips were used to
minimize the risk of contamination. All reaction solutions are combined in a master
reaction tube on ice. The master reaction was calculated to include total number of
samples, five negative controls, and one positive control. Each sample reaction
consisted of 0.5 pi Cloned Pfu Polymerase, 2.0 pi lOx Reaction Buffer for Cloned Pfu
Polymerase, 1.0 pi Dimethyl Sulfoxide, 0.5pl lOmM dNTPs, 1.0 pi sense primer (10.0
pMols per pi), 1.0 pi antisense primer (10.0 pMols per pi), and 14.0 pi sterile H2O for
a final volume of 20.0 pi.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
59
TABLE 3. Nested round reaction primers
For each mutation collocation site target region, the sense primer is listed above the
antisense primer. Primers for N-AR-I were designed by Howard C. Shen (GA
Coetzee Laboratory, University of Southern California). All other primers were
designed with the aid of Vector NTI software (InforMax-Invitrogen).
Target
Name Nested Reaction Primers
Amino
Acids
Collocation
Amino Acids X
5' GAG CGT GCG CGA AGT CAT CCA G 3'
Asn36 5 8 -7 8
N-AR-I 5' CTG TGG GGC CTC TAC GAT GGG C 3' to Gln96 Poly-Gin 60°C
5' GAG CAT CTG AGT CCA GGG GA 3'
Gln260
to
Pro280
256 - 268
AF-1 Region N-AR-I 1 5' CCA ATG GGG CAC AAG GAG TC 3'
48°C
5' GCG AGG CGG GAG CTG TAG C3'
Ala478
5 0 2 -5 3 5 to
N-AR-lll 5' CTG GGC CGA AAG GCG ACA TT 3' Met537 LIAF 59°C
5' CTG AGG AGA CAA CCC AGA AG 3'
Leu659
to
Gly683
6 7 0 -6 7 8
Hinge-LBD N-AR-IV 5' GTC GTG TCC AGC ACA CAC TA 3'
62°C
5' AGG TGT AGT GTG TGC TGG AC 3'
His689 701 - 730
Signature
Sequence N-AR-V4 5' CCA CTT CCC TTT TCC TTA CC 3'
to
Pro723 53°C
5' TGA CCA CTG CCT CTG CCT CT 3'
Phe725
to
Ile737
701 - 730
Signature
Sequence N-AR-V5 5' CCC CAT CCA GGA GTA CTG AA 3' 55°C
5' CCT TGT CAA CCC TGT TTT TCT CC 3'
Ile868
8 7 4 -9 1 0 to
N-AR-VI 5' TAG GGT TTC CAA TGC TTC ACT 3' End920 AF-2 62°C
Each sample reaction, including positive and negative controls, was aliquoted into
individual wells of a 96-well PCR plate on ice. 5.0 pi of each sample and
corresponding positive and negative controls from the first round PCR were then
added to the appropriate reaction wells. Next, the PCR plate was sealed using
Microseal A Film and placed into a thermal cycler for the reaction. Reaction
conditions were 98°C for two minutes for initial template denaturing followed by 30
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6 0
cycles of 98°C for one minute, X°C for one minute (X = annealing temperature for
target region primers, see Table 3), and 74°C for one minute. A final extension at
74°C for four minutes followed the 30 cycles.
PCR Product Purification
All nested PCR samples and controls were run on 2.0% agarose gel (PCR Plus
Agarose from EM Science, product #EM-2010). In situations where any of the
negative controls resulted in a positive PCR amplification the entire experiment was
discarded. Bands were removed with a sterile scalpel under UV light and PCR
products were purified using Qiaquick gel extraction columns (Qiaquick Gel
Extraction Kit from Qiagen, product #28706). Column eluate was desiccated in a
Savant Speed Vac SC 110 and DNA was resuspended in 20.0 pL of sterile H2O.
PCR Product Sequencing
Two microliters of sample and 10.0 pMols of primer were sent for sequencing
(Applied Biosystems ABI Prism 377 DNA Sequencer) in the Norris Cancer Center
Biochemical Core (NCCB) sequencing facility. We estimate our sensitivity to detect
mutations to be 25% based on the computer algorithm that was used by the NCCB to
determine base identity. Each sample was sequenced with both the sense and antisense
primers. Resulting data were aligned to the Tilley Androgen Receptor sequence
(Tilley 1989) using Sequence Navigator software.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6 1
Verification of Mutations
Potential mutations were defined as any change that deviated from the wild-type (WT)
AR sequence that was observed in both sense and antisense primed sequence
chromatograms, which would indicate the potential presence of about 25% mutant
AR. For the polyglutamine repeat, a somatic expansion (more CAG repeats in the
tumor tissue than normal adjacent tissue) or contraction (fewer CAG repeats in the
tumor tissue than normal adjacent tissue) of the CAG tract was considered a potential
mutation if the CAG repeat difference between the normal adjacent and tumor tissue
was three repeats or more. A sequence that resulted in a potential mutation was
selected for further verification analysis. Three, four, or five new first round PCR
amplifications and corresponding nested PCR amplifications were performed for each
sample that contained a potential mutation. The number of verification reactions
(three, four, or five) performed for a specific sample was determined by the available
amount of original template DNA for that specific sample. Verification procedures
were subject to the same methods for PCR amplification, PCR product purification,
and PCR product sequencing as previously reported.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6 2
AR-CAG Repeat Analyses
Each study, the LA and Singapore, was statistically analyzed as a single study
population without mixing of the data between studies. All statistical analyses were
performed in SAS (version 8e). For the case-only analyses, Fisher’s Exact and
Cochran-Mantel-Haenzel % 2 tests of association were used to examine the
relationships between normal adjacent tissue CAG repeat lengths and tumor Gleason
grade. Fisher’s Exact test was used for unadjusted analyses whereas the Cochran-
Mantel-Haenzel x2 test (1 df) was used in analyses that required adjustments for
covariates (age and race). For all analyses, two-sided p values are reported and an a
value of 0.05 was used to establish statistical significance. For the Singapore case-
only analysis a Wilcoxon signed-rank test used to test the null hypothesis that the
median difference between normal adjacent and tumor tissue CAG repeat lengths was
zero. For the case-control analysis of the Singapore study, logistic regression analyses
were performed to estimate odds ratios and the corresponding 95% confidence
intervals for prostate cancer risk in relation to CAG repeat length.
Tumor grade was examined as a categorical variable with high-grade tumors
defined by Gleason grades of seven and greater and low-grade tumors defined by
Gleason grades of six and less. Normal adjacent tissue CAG repeat length was
analyzed as a two-category variable with “short” CAG repeats defined by lengths less
than or equal to 22 and “long” CAG repeats defined by lengths 23 and greater, which
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
63
is based on previous literature (Hsing 2000b; Freedman 2005) Age was analyzed as a
two category variable with subjects classified as being less than or equal to 64 years
old or 65 and older.
For the LA Study three case-only analyses were performed. For the first two
analyses, African-Americans and Latinos were analyzed separately. For the third
analysis, data for African-Americans and Latinos were combined for a complete LA
Study analysis of the relationship between normal adjacent tissue CAG repeat length
and tumor grade with adjustments for race and age.
For the Singapore Study a case-only analysis was performed to examine the
relationships between normal adjacent tissue CAG repeat length and age and tumor
grade in Singapore Chinese men. Secondly, a case-control analysis was performed
that utilized data from our 87 Singapore Chinese cases and 85 Singapore Chinese
controls to investigate the relationship between prostate cancer risk and normal
adjacent tissue C AG repeat length.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6 4
RESULTS
LA Study Results
Mutation Frequency
We performed first round and subsequent nested round PCR amplifications on the
normal adjacent and tumor tissue samples for 90 subjects (180 total PCR samples),
which included three duplicate subjects, for each of the six mutation collocation sites.
Since our mutation collocation site target AR-V required two separate amplifications
(targets AR-V4 and AR-V5), we had seven PCR amplifications to perform for each of
our 180 samples. This resulted in a total of 1260 possible positive PCR
amplifications, of which we were able to obtain 1242 (98.6%) (Table 4). Sequence
analyses of duplicate and original samples, from the three subjects for whom
duplicates were provided, resulted in no sequence variation from WT for any of the six
mutation collocation site targets. For any given collocation site target approximately
50% of the chromatograms for either sense or antisense primed sequences resulted in
deviations from the WT sequence, as indicated by doublet peaks in the chromatogram.
However, the corresponding complimentary sequence chromatogram indicated the
WT sequence was indeed present for all but six samples. Therefore, six potential
mutations were observed in 87 subjects (6.9%) from the 1242 positive PCR
amplifications (< 0.5%) (Table 4). None of the six potential mutations identified in
1242 sequences during initial sequencing of all mutation collocation sites were
reproduced in subsequent verification PCR amplifications.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
65
TABLE 4. Sequencing results of LA study
Collocation Site PCR Performed Positive PCR % Positive Potential Mutations
AR-I 180 178 98.9 % 2
AR-II 180 178 98.9 % 0
AR-III 180 172 95.6 % 2
AR-IV 180 178 98.9 % 0
AR-V4 180 180 100.0 % 1
AR-V5 180 178 98.9 % 0
AR-VI 180 178 98.9 % 1
Total 1260 1242 98.6 % 6
AR-CAG Repeat Associations
Our sequencing analyses resulted in 97.3% concordance between normal adjacent and
tumor tissue CAG repeat lengths. Therefore, our statistical analyses only utilized
normal adjacent CAG length data. Table Five shows characteristics of age at
diagnosis, CAG repeat length, and tumor Gleason grade for the 75 subjects used in
statistical analyses for the LA Study. The mean ages for African-Americans, Latinos,
and African-Americans/Latinos combined were 62.2, 64.6, and 63.4, respectively.
The mean CAG lengths for African-Americans, Latinos, and African-
Americans/Latinos combined were 19.3, 22.6, and 21.1, respectively. The ranges of
CAG lengths were 13 -2 6 , 17-29, and 13-29, for African-Americans, Latinos, and
African-Americans/Latinos combined, respectively. When CAG repeats were
categorized into two-CAG repeat intervals, two peaks were observed for the African-
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6 6
Americans in the 17 - 18 CAG and 21-22 CAG repeats categories whereas one peak,
in the 21 - 22 CAG repeats category, was observed for Latinos (Table 6). Tumor
grades ranged from four to nine for the entire study population. Tumor grade
information was missing for one African-American subject.
TABLE 5. LA Study characteristics of age, CAG repeat length, and tumor grade
African-American Latino Combined
N = 36 N = 39 N = 75
N (% ) N (% ) N (% )
A ge at D xa
< 64 21 (58.3) 19 (48.7) 40 (53.3)
65 + 15 (41.7) 20 (51.3) 35 (46.7)
Mean + SD 62.2 + 5.7 64.6 + 4.4 63.4 + 5.2
Median 64 65 64
CAGn Length
< 2 2 29 (80.6) 20 (51.3) 49 (65.3)
23 + 7 (19.4) 19 (48.7) 26 (34.7)
Mean + SD 19.3 + 3.4 22.6 + 3.0 21.1 + 3.6
Median 18 22 21
Tumor Grade
Low Grade < 6 21 (60.0) 22 (56.4) 43(58.1)
High Grade 7 + 14 (40.0) 17 (43.9) 31 (41.9)
Unknown 1 0 1
a) Age at diagnosis
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
67
TABLE 6. LA Study distribution of CAG repeats by two-repeat interval
African-American Latino Combined
N = 36 N = 39 N = 75
CAG Repeat Length N (%) N (%) N (%)
< 16 5 (13.9) 0 5 (6.7)
17-18 14 (38.9) 4(10.3) 18 (24.0)
19-20 3 (8.3) 5 (12.8) 8 (10.7)
21-22 7 (19.4) 11(28.2) 18 (24.0)
23-24 3 (8.3) 8 (20.5) 11 (14.7)
25-26 4(11.1) 8 (20.5) 12 (16.0)
27 + 0 3 (7.7) 3 (4.0)
The results of our association analyses for CAG repeat length, age, race, and
tumor grade are presented in Table Seven. No significant findings can be reported for
African-Americans, however, statistically significant associations were observed
between short (fewer than 23) CAG repeats and diagnosis before age 65 (p = 0.001)
and low-grade tumors (p = 0.02) among Latinos. After adjustment for race, subjects
diagnosed with prostate cancer after age 65 were more likely than subjects diagnosed
before age 65 to have long (23 or more) CAG repeats (80.0% v 48.6%; p = 0.007).
After adjustment for age, Latinos were more likely than African-Americans to have
long CAG repeats (48.7% v 19.4%; p = 0.01). After adjustment for age and race,
subjects presenting with high-grade tumors were more likely than subjects presenting
with low-grade tumors to have long CAG repeats (51.6% v 23.3%; p = 0.02).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6 8
TABLE 7. LA Study results of analyses for associations between CAG repeat length
and age at diagnosis, tumor grade, and race
Af-Am (N = 36) Latino (N = 39) Combined (N = 75)
CAGn Length________CAGn Length________C AGn Length
Variable
< 2 2
N (row %)
23 +
N (row %)
< 2 2
N (row %)
23 +
N (row %)
< 2 2
N (row %)
23 +
N (row %)
A ge at D x
< 6 4
65 +
17 (81.0)
12 (80.0)
4 (19.0)
3 (20.0)
15 (79.0)
5 (25.0)
4(21.0)
15 (75.0)
32 (80.0)
17 (48.6)
8 (20.0)
18(51.4)
P
= 1.00 = 0.001 = 0.007“
Race
Af-Amd
Latino
29 (80.6)
20 (51.3)
7 (19.4)
19 (48.7)
P
= 0.01b
Grade0
Low Grade < 6
High Grade 7 +
18 (85.7)
10 (71.4)
3 (14.3)
4 (28.6)
15 (68.2)
5 (29.4)
7(31.8)
12 (70.6)
33 (76.7)
15 (48.4)
10 (23.3)
16 (51.6)
P
= 0.31° = 0.02° = 0.02°
a) Adjusted for race
b) Adjusted for age at diagnosis (< 64,65 +)
c) African-Americans, Latinos adjusted for age at diagnosis (< 64,65 +); Combined adjusted for
age at diagnosis (< 64,65 +) and race
d) Af-Am = African American
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6 9
Singapore Study Results
Mutation frequency
After numerous attempts, PCR amplification failed for target AR-III for a vast
majority of samples. Therefore, we could only analyze five mutation collocation sites
with six sets of target primers. We performed first round and subsequent nested round
PCR amplifications on the normal adjacent and tumor tissue samples for 106 subjects
(212 total PCR samples), which included seven duplicate subjects, for each of the five
mutation collocation sites. Since our mutation collocation site target AR-V required
two separate amplifications (targets AR-V4 and AR-V5), we had six PCR
amplifications to perform for each of our 212 samples. This resulted in a total of 1272
possible positive PCR amplifications, of which we were able to obtain 1207 (94.9%)
(Table 8). Sequence analyses of duplicate and original samples, from the seven
subjects for whom duplicates were provided, resulted in no sequence variation from
WT for any of the five mutation collocation site targets. Like the LA study results,
approximately 50% of the chromatograms for either sense or antisense primed
sequences resulted in deviations from the WT sequence, as indicated by doublet peaks
in the chromatogram,. However, the corresponding complimentary sequence
chromatogram indicated the WT sequence was present for all but 22 samples.
Therefore, we identified 22 potential mutations in 99 subjects (22.2%) from the 1207
positive PCR amplifications ( 1.8%) (Table 8).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
70
TABLE 8. Sequencing results of Singapore Study
Collocation Site PCR Performed Positive PCR % Positive Potential Mutations
AR-I 212 187 88.2 % 10
AR-II 212 210 99.1% 1
AR-IV 212 212 100.0 % 1
AR-V4 212 209 98.6 % 5
AR-V5 212 212 100.0 % 1
AR-VI 212 177 83.5 % 4
Total 1272 1207 94.9 % 22
Three of the 22 potential mutations identified during the initial sequencing of
were reproduced in subsequent verification PCR amplifications. Therefore, three
verified mutations were observed in 99 subjects (3.0%). Two subjects had somatic
contractions of at least three CAG repeats in the polyglutamine tract, which were
reproduced in subsequent verification PCR amplifications. In addition to the two
verified CAG contractions, we identified a Met to Val mutation at position 886 (ATG
to GTG) in the normal adjacent and tumor tissue of subject scyl49. In order to verify
the existence of this mutation, a new Proteinase K digestion was performed for normal
adjacent and tumor tissue samples. Four verification PCR amplifications were
performed for each sample (normal adjacent and tumor). All amplifications and
subsequent sequencing verified the existence of the M886V mutation in the normal
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
71
adjacent and tumor tissue samples. The number of CAG repeats identified in subject
scyl49 was 22 for both the normal adjacent and tumor tissue. Subject scyl49 was
diagnosed with prostate cancer, Gleason grade nine, at age 68.
AR-C AG Repeat Associations
The results of our sequencing analyses showed 70.9% concordance between normal
adjacent and tumor tissue CAG repeat lengths for our cases (p = 0.29). Our statistical
analyses only utilized normal adjacent polyglutamine length data for cases. Table
Nine shows characteristics of age and CAG repeat length for the 87 cases and 85
controls used in statistical analyses. The mean and median age at diagnosis for cases
is 71.2 and 72, respectively. For the controls, the mean and median ages at time of
specimen collection were 58.3 and 58, respectively. The mean and median CAG
repeat lengths for cases were 22.1 and 22, respectively. For controls the mean
polyglutamine length was 21.9 and the median was 22. The majority of cases (61.0%)
and controls (57.6%) had fewer than 23 CAG repeats. The ranges of polyglutamine
lengths were 6 -3 0 and 14 - 29 for cases and controls, respectively. Information on
CAG length was missing for five cases. When normal adjacent polyglutamine lengths
were categorized into two CAG repeat intervals, the distributions of cases and controls
were very similar (Table 10). The range for tumor grades for cases was 2 -1 0 and
information on tumor grade was missing for 13 cases.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
72
TABLE 9. Singapore Study characteristics of age, CAG repeat length, and tumor
grade
Cases Controls
N = 87 N = 85
N (%) N (%)
A ge8
< 71 40 (46.0) 77 (90.6)
72 + 47 (54.0) 8 (9.4)
Mean + SD 71.2 + 7.1 58.3 + 8.1
Median 72 58
CAGn Length
< 2 2 50 (61.0) 49 (57.6)
23 + 32 (39.0) 36 (42.4)
Unknown 5 0
Mean + SD 22.1 + 3.6 21.9 + 2.8
Median 22 22
Tumor Grade
Low Grade < 6 35 (47.3)
High Grade 7 + 39 (52.7)
Unknown 13
a) Age at diagnosis for cases; age at time of blood specimen collection for controls
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
73
TABLE 10. Singapore Study distribution of CAG repeats by two-repeat interval
Cases Controls
N = 87 N = 85
CAG Repeat Length N (%) N (%)
< 16 6(7.3) 5 (5.9)
17-18 1 (1.2) 3 (3.5)
19-20 10 (12.2) 12(14.1)
21-22 33 (40.2) 29 (34.1)
23-24 13 (15.8) 22 (25.9)
25-26 13(15.8) 12(14.1)
27 + 6(7.3) 2(2.4)
Unknown 5 0
Case-only findings regarding possible associations between CAG repeat
length, age, and tumor grade are shown in Table 11. No statistically significant
association between age at diagnosis and CAG repeat length was observed (p = 1.00).
After adjustment for age, no statistically significant association between CAG repeat
length and tumor grade was observed (p = 0 .68).
The results of our case-control analysis of possible associations between CAG
repeat length and prostate cancer risk are presented in Table 12. After adjustment for
age no statistically significant association between prostate cancer risk and CAG
repeat length was observed (OR, 1.18; 95% Cl, 0.58 - 2.39).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7 4
TABLE 11. Singapore Study case-only (N = 87) results of analyses for associations
between CAG repeat length and age at diagnosis and tumor grade
Variable
CAG <22
N (row %)
CAG 23 +
N (row %)
Age at Dx
<71 23 (60.5) 15 (39.5)
72 + 27(61.4) 17 (38.6)
P
= 1.00
Grade3
Low Grade < 6 22 (66.7) 1 1 (33.3)
High Grade 7 + 23 (62.2) 14 (37.8)
P
= 0.69
a) adjusted for age at diagnosis (<71, 72 +)
TABLE 12. Singapore Study case-control results of analyses for associations between
CAG repeat length and prostate cancer risk
Cases N = 87 Controls N = 85
N (%) N (%) OR (95 %CI)a
Normal CAG Length
< 2 2 50(61.0) 49 (57.6) 1.18(0.58, 2.39)
23 + 32 (39.0) 36 (42.4) 1.0
a) adjusted for age (< 71,72 +)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
75
DISCUSSION
Mutation Frequency
Many studies have examined AR mutations in post-treatment tumors after androgen-
ablation therapy failure (Culig 1993; Taplin 1995, 2003; Tilley 1996; Haapala 2001;
Hyytinen 2002). In one study, mutations found in two post-treatment high-grade
tumors were not identified in the corresponding pre-treated tumor samples (Hyytinen
2002). This brings into question the relevance of AR mutations in the etiology of
primary prostate cancer prior to androgen ablative treatments. However, few studies
have been undertaken to investigate the frequency of AR mutations in pre-treated
primary tumors (Newmark 1992; Schoenberg 1994; Marcelli 2000; Thompson 2003)
and they have typically involved less than 40 subjects.
We examined the frequency of occurrence of AR mutations in six regions of
mutation collocation in untreated primary tumors in two distinct study populations. In
our LA Study 6.9% of our 87 subjects of African-American or Latino descent had
potential mutations that occurred within the six regions of mutation collocation. We
were unable to verify their existence in subsequent PCR amplifications. In our
Singapore Study 22.2% of our 99 subjects contained potential mutations. Three of
these potential mutations (3.0%) were verified and included two somatic contractions
of the CAG repeat and one germline M886V mutation. In total, of 186 subjects, we
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
76
were only able to identify three verifiable mutations. We interpret the presence of
potential mutations not verifiable by subsequent PCR analyses as a result of either
founder effects by chance amplifications or the existence of mutation containing AR
molecules that occur at a frequency bordering on our level of detection sensitivity. Of
course the possibility remains that somatic mutations in the AR exist in regions
outside the collocation sites we have sequenced. We think this is unlikely because the
existence of collocation sites imply functional selection of mutations due to gain-of
function of the AR, giving tumor clones a growth advantage. Such mutations
therefore play a minimal role in PCa progression in the absence of androgen ablation
treatment.
Our findings contradict previously published literature that suggests AR
mutations may be frequent in primary prostate tumors prior to androgen ablation
therapy (Heinlein 2004). Tilley et al reported in 1996 that 44% (11/25) of stage C or
D primary prostate tumors contained mutations that resulted in an altered amino acid
(Tilley 1996). Similarly, Thompson et al found that 36% (5/14) of untreated primary
tumors and 14% (1/7) of bone, lymph node, and rectum metastases contained
mutations (Thompson 2003). Our findings are supported, however, by reports that
suggest AR mutations are infrequent in untreated tumors. Newmark et al observed
mutations in 3 .8% (1/26) of subjects with untreated, organ-confined stage B prostate
cancer (Newmark 1992). A somatic contraction of the CAG repeat (CAG24 ->
CAGig) was discovered in pretreated clinically localized disease in one subject out of
40 (2.5%) (Schoenberg 1994). Marcelli et al reported that no mutations were observed
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7 7
in any of the 99 subjects in the study who presented with stage B untreated prostate
cancer, which lead them to conclude that the AR probably plays an insignificant role
in early phases of the disease (Marcelli 2000).
Most of these previously mentioned studies utilized either denaturing gradient
gel electrophoresis (DGGE) (Newmark 1992) or single-stranded conformation
polymorphism (SSCP) analysis (Tilley 1996; Marcelli 2000; Thompson 2003) to
detect mutations before direct sequencing. Although SSCP has become established as
a primary method of mutation detection (Elo 1995; Tilley 1996; Hyytinen 2002;
Thompson 2003; Koivisto 2004; MacLean 2004), the materials associated with the
method, such as radioactive labels and autoradiography film, can be expensive. In
order to avoid the costs associated with SSCP analyses we decided to directly
sequence PCR products for mutation detection. It can be argued that the limitation of
our direct sequencing sensitivity may result in a failure to detect mutations. However,
the sensitivity of direct sequencing has been demonstrated to have the ability to detect
point mutations if they are present in 10% of genomic template DNA, which is
comparable to the 10-15% mutation detection sensitivity of SSCP analysis (Marcelli
2000). We are confident that we can detect mutations if they represent 10 - 25% of
our PCR template DNA sample. Thus, we conclude our observation of low AR
mutation frequency in pre-treated primary prostate tumors is accurate. In addition, if
we are indeed missing AR mutations that exist below our range of sensitivity of 10 -
25%, it can be argued that this small proportion of mutant AR plays an insignificant
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
78
role in primary tumors prior to androgen ablation therapy. This is not to say that these
mutations may, however, provide a selective growth advantage via altered androgen
responsiveness for cells after initiation of androgen ablative treatment (Tilley 1996).
Another potential limitation of our study is PCR sensitivity. It has been
suggested that failure to use tumor-enriched DNA may result in an inability to detect
mutations (Marcelli 2000). By using demarcated microdissected tissue samples,
which we estimate contained at least 90% tumor tissue, we maximize our ability to
detect somatic mutations in tumor by minimizing the amount of non-mutation
containing normal adjacent tissue DNA that can be unintentionally introduced into our
tumor DNA samples during extraction. However, microdissected paraffin embedded
tissue also has drawbacks that can affect PCR amplification. It has been shown that
the use of paraffin embedded tissue may result in poor quality template DNA that can
adversely affect DNA polymerase fidelity (Shiao 1997). In order to minimize the
possibility of sequencing artifacts introduced by this phenomenon we chose to use
Stratagene’s Pfu DNA polymerase. This polymerase has been shown to have a low
error rate, which is 1.3 x lO^/base (Cline 1996). This high fidelity combined with our
10 - 25% sequencing sensitivity makes the detection of mutations due to Pfu error
highly improbable. However, since we are using paraffin embedded tissue and we
can’t rule out the possibility of founder effects that by chance will result in the
amplification of PCR products that contain erroneously incorporated nucleotides, it is
a possibility that we could see some errors. Since we were only able to verify three of
our 28 potential mutations, it is likely that the 25 unverifiable potential mutations were
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7 9
erroneously amplified as a result of these phenomena. In addition, we believe our
Singapore Study paraffin embedded tissue samples may have produced DNA template
of poorer quality compared to our LA Study since they were prepared, packaged, and
shipped from China, which may explain the higher percentage of potential mutations
in the Singapore Study. However, another possibility is that these 25 potential
mutations really did exist in our template DNA in an amount bordering on our
sensitivity and we were just unable to reproduce these products in subsequent
verification reactions. If this were the case, then the low frequency of these mutations
in the tumor makes it unlikely that these mutations had any relevance to tumor
progression up to this stage. However, as stated above, these mutations may provide a
selective growth advantage after the initiation of androgen ablation therapy.
The problem of contamination is always an issue when performing PCR.
Since our lab regularly uses several different mutant AR plasmids, we went to great
lengths to prevent contamination in every step of our study. During the extraction
process, which was performed in a sterile PCR workstation dedicated for the purpose,
sterile scalpels and plasmid-free pipettors with barrier tips were used. During the PCR
process, first round reactions were performed in a sterile plasmid-free hood that is
only used for PCR purposes. Pipettors in the hood are only used for first round PCR
experiments and barrier tips are used to prevent contamination of this equipment. As
an additional precaution, UV light is constantly on in the hood and PCR workstation
when they are not being used. For nested round PCR, a separate set of nested round-
only pipettors was used. We are confident that our PCR procedures were free of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
80
plasmid contamination as a result of these precautions and the facts that: 1. Our use of
several water blanks for each PCR experiment served as an indicator for potential
contamination, which if it were present resulted in the disposal of the entire batch;
2. We routinely obtained different CAG repeat lengths during for different tumor
samples, which suggests a contaminating plasmid of fixed CAG length was not
consistently amplified; 3. Not one of our nearly 5,000 sense and antisense primed
sequences resulted in the detection of any of the numerous mutant plasmids used in
our laboratory; 4. Four of our seven mutation collocation site targets utilized intron-
based primers, which would not amplify AR plasmid DNA.
In summary, we looked at a total of 187 subjects from two different studies
that represented three different racial backgrounds. The fact that we were only able to
observe 3/186 (1.6%) mutations, a result comparable to the previously mentioned
findings of Marcelli, Newmark, and Schoenberg (Newmark 1992; Schoenberg 1994;
Marcelli 2000), in pretreated primary tumors represents a dearth of evidence for a
significant effect of AR mutations in the etiology of prostate cancer prior to androgen
ablation therapy.
M886V Mutation
Although the M886V mutation has been previously identified in AIS (Ghadessy
1999), our discovery of the M886V mutation represents a novel finding in the sense
that this particular mutation has not been identified in an individual with prostate
cancer. Ghadessy et al first identified this mutation in three unrelated, infertile males
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
81
in 1999 who had normal male karyotypes and were near-azoospermic. (Ghadessy
1999). External male genitalia and impaired spermatogenesis is indicative of MAIS
(Shkolny 1999), the phenotype presented by these three individuals. Ghadessy and
colleagues determined the M886V mutation is unlikely to exist in the general
population (p = 0.027) and that the association between the mutation and severe
oligospermia is statistically significant (p = 0.005) (Ghadessy 1999). Unfortunately,
we do not have any information on the phenotypic status of our subject in regard to
androgen insensitivity.
What makes our finding of the M886V mutation in a subject diagnosed with
prostate cancer interesting is the fact that mutations involved with prostate cancer are
usually not found in any of the three forms of AIS (Yong 2000; Mooney 2003).
Prostate cancer is typically associated with somatic, gain-of-function mutations in the
AR whereas AIS is associated with germline, loss-of-fimction mutations in the AR
(Gottlieb 2004). We believe our finding represents a germline mutation since we
observed this mutation in both normal adjacent and tumor tissue. In addition to our
finding, other germline mutations of the AR associated with cancer have also been
reported. The R726L mutation has been estimated to exist in 2% of Finnish prostate
cancer patients (Mononen 2000); however, it probably has little influence on prostate
cancer predisposition (Koivisto 2004). The germline mutations R607Q (Wooster
1992) and R608K (Lobaccaro 1993) were identified in males with breast cancer and
PAIS. To date, only six mutations (all in the LBD) have been identified that result in
an amino acid substitution that has been observed in both prostate cancer and AIS
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
82
(PAIS or CAIS) (ARDB 2004). Our identification of the M886V mutation in prostate
cancer represents the first time an AR mutation has been identified in both MAIS and
prostate cancer.
At the present time, there are only six publications (Ghadessy 1999; Yong
2000; Thompson 2001; Wang 2001a; Ghali 2003; Yong 2003) that discuss the M886V
mutation and only two of these (Ghadessy 1999; Thompson 2001) actively
investigated the functional activity of the mutation, which reveals little is known about
M886V. Ghadessy et al reported no difference in androgen-binding properties, a
reduced transactivation capacity (50 - 70% of WT receptor), and reduced DNA
binding activity (50% DNA binding activity of WT receptor) for M886V (Ghadessy
1999). In addition, it was found that the mutation impaired LBD-LBD (50% WT
activity), LBD-NTD (75% WT activity), and LBD-TIF2 (40% reduced coactivator
function) interactions, leading the authors to conclude the mutation does not affect
ligand binding but it does negatively affect receptor activity by impairing coactivator
function (Ghadessy 1999). Thompson et al, in a study comparing several mutations
associated with AIS and prostate cancer, reported findings consistent with Ghadessy’s
in that the M886V mutant had normal ligand binding activity and slightly decreased
transcriptional activity (Thompson 2001). However, in contrast to Ghadessy’s
observations, coexpressed GRIP1 resulted in more than a four-fold increase in
transcriptional activity for M886V and two other prostate cancer mutations, V715M
and R726L (the Finnish prostate cancer mutation), whereas the increase for four
mutants associated with CAIS or PAIS was only three fold. In addition, Thompson
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
83
et al did not observe a decrease in DNA binding activity or impairment of N/C
interaction of M886V. To summarize, Thompson et al found that M886V, V715M,
and R726L had DNA binding activity, N/C interaction capacity, and GRIP1 activation
comparable to WT receptor; but, M886V and R726L had reduced transcriptional
activity (Thompson 2001).
In conclusion, we identified an M886V germline mutation, a mutation
typically associated with MAIS, in an individual with high-grade prostate cancer.
Based on the report from Thompson et al (Thompson 2001), the M886V mutation
might play a role in the etiology of prostate cancer in a functional capacity similar to
other prostate cancer mutations, such as V715M and R726L. However, the relevance
of R726V to disease progression has been downplayed (Koivisto 2004) and it would
appear more likely that the M886V mutation does not play a significant role in the
development of prostate cancer either.
LA and Singapore AR-CAG Repeat Associations
In the LA Study, there was a 97.3% concordance between our normal adjacent
tissue and tumor tissue CAG repeat lengths, which is in agreement with a previously
reported value of 97.5% (Bennett 2002). Our Singapore Study resulted in 70.9%
concordance. This variation may be due to lower quality template DNA and/or
founder effects that by chance amplify an erroneous PCR product, as previously
mentioned. Our means, medians, and ranges of CAG repeats in African-Americans,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
84
Latinos, and Chinese are consistent with previously published observations (Irvine
1995; Sartor 1999; Hsing 2000b; Bennett 2002; Balic 2002;), which indicates our
method of direct sequencing was adequate for accurate determination of CAG length.
In the LA Study, we did not observe statistically significant associations
between number of CAG repeats and age at diagnosis or tumor grade in our African-
American subjects. However, in our Latino subjects, we did observe statistically
significant associations between short (< 22) CAG repeats and younger age (< 64) at
diagnosis (p = 0.001) and low-grade (< 6) tumors (p = 0.02). When African-
Americans and Latinos were combined for analyses, statistically significant
associations were observed for short CAG repeats and race after adjustment for age at
diagnosis (p = 0.01), younger age at diagnosis after adjustment for race (p = 0.007),
and low-grade tumors after adjustment for age at diagnosis and race (p = 0.02).
In our Singapore Study case-only analyses we did not observe any statistically
significant associations between CAG repeat length and age at diagnosis or tumor
grade. In the case-control analysis we found an insignificant 18% increased risk for
prostate cancer associated with short CAG repeat length (OR 1.18, 95% Cl 0.58 -
2.39). Although this finding is insignificant, it is consistent with other reports that
suggest an increased risk for prostate cancer associated with shorter CAG in Chinese
men. Hsing et al reported a statistically significant 65% increased risk for prostate
cancer in Chinese men who had CAG repeats shorter than 23 (Hsing 2000b).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
85
Our findings in the LA Study support previous literature regarding associations
between short CAG repeats and younger age at diagnosis (Bratt 1999) and race (Sartor
1999; Hsing 2000b; Bennett 2002; Balic 2002). However, contrary to our findings,
statistically significant associations between short CAG repeats and high-grade tumors
(Giovannucci 1997) have been observed, as well as a lack of significant association
between CAG length and grade (Stanford1997, Gsur 2002). To our knowledge we are
the first to observe a significant association between short CAG repeats and low-grade
tumors. It has been reported that a shorter polyglutamine tract will result in a
transcriptionally more active AR that can affect the progression of prostate cancer
(Ding 2004). Irvine et al demonstrated that increasing CAG repeat length had a
negative affect on pi 60 coactivation of the AR (Irvine 2000). Wang et al observed
that a polyglutamine tract containing nine CAG repeats demonstrated increased
transactivation ability due to enhanced coactivator recruitment resulting from an
altered ligand-induced conformation (Wang 2004). In addition, the shorter CAG
length also augmented prostate cancer cell proliferation. Based on findings such as
these it would be logical to theorize that shorter CAG lengths would be associated
with higher-grade tumors. Since we observed the short CAG/low tumor grade
association in our LA Study and no association between CAG length and tumor grade
in our Singapore Study, it can be hypothesized that we may be observing an influence
of PSA screening detection bias. PSA screening has been established as a method of
detection since 1988 in the United States (Platz 2004a), whereas mass PSA screening
in China is a recent development (Zhang 2004). In 2003 Chu et al reported that a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
86
decrease in prostate cancer mortality rates could be attributed to earlier detection by
PSA screening as evidenced by decreasing distant disease incidence and mortality
rates (Chu 2003). If transcriptionally hyperactive AR, owing to shorter CAG length,
results in earlier detection due to an increase in serum PSA, then it is plausible that we
would see an association between shorter CAG and lower-grade tumors. The fact that
we did not find this in the Singapore Study could be because PSA screening has only
recently been implemented in Singapore and the majority of detected tumors are
advanced (Zhang 2004). Giovannucci et al reported an association between high-
grade prostate cancer and short CAG repeat length; however, all the subjects in the
study had entered in 1982 and had been diagnosed by 1995 (Giovannucci 1997).
Higher-grade tumors may be over represented in the Giovannucci study since PSA
screening wasn’t initiated until 1988 and distant disease incidence didn’t begin
declining until approximately 1992 (Chu 2003). By comparison, our LA Study
consisted of subjects collected between 1993 and 1996, which is after the decline in
distance disease incidence began. Our Singapore Study subjects were collected
between 1993 and 1998 and contained a majority of high-grade tumors that can be
assumed were not PSA detected, which is consistent with recently published reports
(Hsing 2000b; Gao 2005). To our knowledge, our LA Study of the association
between CAG repeat length and tumor grade is one of only three studies that utilized
subjects collected after the observed decline in distant disease incidence began.
Stanford et al reported finding no association between CAG repeat length and grade in
subjects collected between 1993 and 1994 (Stanford 1997). Similarly, in 2002 Gsur et
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
87
al observed no association between CAG repeat length and tumor grade in subjects
collected between 1998 and 2001 (Gsur 2002). In order to verify an association
between CAG repeat length and tumor grade, future studies have to consider the
influence of PSA screening on their sample population.
In summary, we found statistically significant associations between shorter
CAG repeat length and younger age at diagnosis and lower-tumor grade in a mixed
racial population from Los Angeles. While these findings were significant in Latinos
when races were analyzed separately, we did not find significant associations in
African-Americans. Similarly, in a study of Singapore Chinese men we did not find
any statistically significant associations in case-only or case-control analyses. Based
on our observations from the LA and Singapore Studies of significant and
insignificant associations between short CAG repeat length and low-grade tumors,
respectively, we theorize detection bias in PSA screening may influence the
relationship between CAG repeat length and tumor grade.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
88
BIBLIOGRAPHY
Abrahamsson PA and di Sant’ Agnese PA. 1993. Neuroendocrine cells in the human
prostate gland. Journal of Andrology 14 (5): 307 - 309.
Alexander FE and Prescott RJ. 1999. Reply to Labrie et al. Prostate 40 (2): 135 -
136.
American Cancer Society (ACS). 2004. Facts and figures.
Anderson KM and Liao S. 1968. Selective retention of dihydrotestosterone by
prostatic nuclei. Nature 219 (151): 277 - 299.
ARDB. 2004. Androgen Receptor Database, http://www.androgendb.mcgill.ca.
Balic I, Graham ST, Troyer DA, Higgins BA, Pollock BH, Johnson-Pais TL,
Thompson IM, and Leach RJ. 2002. Androgen receptor length polymorphism
associated with prostate cancer risk in Hispanic men. Journal of Urology 168
(5): 2245 - 2248.
Beilin J, Ball EMA, Favaloro JM, and Zajac JD. 2000. Effect of the androgen
receptor CAG repeat polymorphism on transcriptional activity: specificity in
prostate and non-prostate cell lines. Journal of Molecular Endocrinology 25
(1): 85 - 96.
Bennett CL, Price DK, Kim S, Liu D, Jovanovic BD, Nathan D, Johnson ME,
Montgomery JS, Cude K, Brockbank JC, Sartor O, and Figg W. 2002. Racial
variation in the CAG repeat lengths within the androgen receptor gene among
prostate cancer patients of lower socioeconomic status. Journal of Clinical
Oncology 20 (17): 3599 - 3604.
Black BE and Paschal BM. 2004. Intranuclear organization and function of the
androgen receptor. Trends in Endocrinology and Metabolism 15 (9): 411 -
417.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8 9
Blok LJ, Themmen APN, Peters AHFM, Trapman J, Baarends WM, Hoogerbrugge
JW, and Grootegoed JA. 1992. Transcriptional regulation of androgen
receptor gene expression in Sertoli cells and other types. Molecular and
Cellular Endocrinology 88 (1-3): 153 - 164.
Boer R and Schroder FH. 1999. Quebec randomized controlled trial on prostate
cancer screening shows no evidence for mortality reduction. Prostate 40 (2):
130-131.
Bonkhoff H and Remberger K. 1993a. Widespread distribution of nuclear androgen
receptors in the basal cell layer of the normal and hyperplastic human prostate.
Virchows Archiv A Pathological Anatomy and Histopathology 422 (1): 35 -
38.
BonkhofFH, Stein U, and Remberger K. 1993b. Androgen receptor status in
endocrine-paracrine cell types of the normal, hyperplastic, and neoplastic
human prostate. Virchows Archiv A Pathological Anatomy and
Histopathology 423 (4): 291 - 294.
Bratt O, Borg A, Kristoffersson U, Lundgren R, Zhang QX, and Olsson H. 1999.
CAG repeat length in the androgen receptor gene is related to age at diagnosis
of prostate cancer and response to endocrine therapy, but not to prostate cancer
risk. British Journal of Cancer 81 (4): 672 - 676.
Brinkmann AO, Faber PW, van Rooij HCJ, Kuiper GGJM, Ris C, Klaassen P, van der
Korput JAGM, Voorhorst MM, van Laar JH, Mulder E, and Trapman J. 1989.
The human androgen receptor: domain structure, genomic organization, and
regulation of expression. Journal of Steroid Biochemistry 34 (1-6): 307 - 310.
Brodie J and McEwan IJ. 2005. Intra-domain communication between the N-terminal
DNA-binding domains of the androgen receptor: modulation of androgen
response element DNA binding. Journal of Molecular Endocrinology 34 (3):
603-615.
Brown CJ, Goss SJ, Lubahn DB, Joseph DR, Wilson EM, French FS, and Willard HF.
1989. Androgen receptor locus on the human X chromosome: regional
localization to Xql 1-12 and description of a DNA polymorphism. American
Journal of Human Genetics 44 (2): 264 - 269.
Bruchovsky N and Wilson JD. 1968. The intranuclear binding of testosterone and
5a-androstan-17P-ol-3-one by rat prostate. Journal o f Biological Chemistry
243 (22): 5953 - 5960.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
90
Bubulya A, Wise S, Shen XQ, Burmeister LA, and Shemshedini L. 1996. c-Jun
can mediate androgen receptor-induced transactivation. Journal of Biological
Chemistry 271 (40): 24583 - 24589.
Buchanan G, Greenberg NM, Scher HI, Harris JM, Marshall VR, and Tilley WD.
2001a. Collocation of androgen receptor gene mutations in prostate cancer.
Clinical Cancer Research 7 (5): 1273 - 1281.
Buchanan G, Yang M, Harris JM, Nahm HS, Han G, Moore N, Bentel JM, Matusik
RJ, Horsfall DJ, Marshall VR, Greenberg NM, and Tilley WD. 2001b.
Mutations at the boundary of the hinge and ligand binding domain of the
androgen receptor confer increased transactivation function. Molecular
Endocrinology 15 (1): 46 - 56.
Cabral DF, Santos A, Ribeiro ML, Mesquita JC, Carvalho-Salles AB, and Hackel C.
2004. Rarity of DNA sequence alterations in the promoter region of the
human androgen receptor gene. Brazilian Journal of Medical and Biological
Research 37 (12): 1789- 1794.
Carroll P, Coley C, McLeod D, Schellhammer P, Sweat G, Wasson J, Zietman A, and
Thompson I. 2001. Prostate-specific antigen best practice policy - part I:
early detection and diagnosis of prostate cancer. Urology 57 (2): 217 - 224.
Chan JM and Giovannucci EL. 2001a. Dairy products, calcium, and vitamin D and
risk of prostate cancer. Epidemiologic Reviews 23 (1): 87 - 92.
Chan JM and Giovannucci EL. 2001b. Vegetables, fruits, associated micronutrients,
and risk of prostate cancer. Epidemiologic Reviews 23 (1): 82 - 86.
Chang C, Kokontis J, and Liao S. 1988a. Molecular cloning of human and rat
complementary DNA encoding androgen receptor. Science 240 (4850): 324 -
326.
Chang C, Kokontis J, and Liao S. 1988b. Structural analysis of complementary DNA
and amino acid sequences of human and rat androgen receptors. Proceedings
of the National Academy of Sciences 85 (19): 7211 -7215.
Chang BL, Zheng SL, Hawkins GA, Isaacs SD, Wiley KE, Turner A, Carpten JD,
Bleecker ER, Walsh PC, Trent JM, Meyers DA, Isaacs WB, and Xu J. 2002.
Polymorphic GGC repeats in the androgen receptor gene are associated with
hereditary and sporadic prostate cancer risk. Human Genetics 110 (2): 122 -
129.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
91
Chen S, Supakar PC, Vellanoweth RL, Song CS, Chatteijee B, and Roy AK. 1997.
Functional role of a conformationally flexible homopurine/homopyrimidine
domain of the androgen receptor gene promoter interacting with Spl and a
pyrimidine single strand DNA-binding protein. Molecular Endocrinology 11
(1): 3 - 15.
Chen CP, Chem SR, Wang TY, Wang W, Wang KL, and Jeng CJ. 1999. Androgen
receptor gene mutations in 46,XY females with germ cell tumours. Human
Reproduction 14 (3): 664 - 670.
Chen C, Lamharzi N, Weiss NS, Etzioni R, Dightman DA, Barnett M, DiTommaso D,
and Goodman G. 2002. Androgen receptor polymorphisms and the incidence
of prostate cancer. Cancer Epidemiology, Biomarkers, and Prevention 11(10
pt 1): 1033 - 1040.
Chu KC, Tarone RE, and Freeman HP. 2003. Trends in prostate cancer mortality
among Black men and White men in the United States. Cancer 97 (6): 1507 -
1516.
Chung B and Hu M. 2002. Androgen biosynthesis and degradation. In: Androgens
and androgen receptor: mechanisms, functions, and clinical applications.
Chang C. 1 -1 5 . Norwell, Massachusetts: Kluwer Academic Publishers.
Cline J, Braman JC, Hogrefe HH. 1996. PCR fidelity of Pfu DNA polymerase and
other thermostable DNA polymerases. Nucleic Acids Research 24 (18): 3546
-3551.
Coldman AJ, Phillips N, and Pickles TA. 2003. Trends in prostate cancer incidence
and mortality: an analysis of mortality change by screening intensity.
Canadian Medical Association Journal 168 (1): 31 — 35.
Crawford ED. 2003. Epidemiology of prostate cancer. Urology 62 (Suppl 6 A):
3 -1 2 .
Culig Z, Hobisch A, Cronauer MV, Cato AC, Hittmair A, Radmayr C, Eberle J,
Bartsch G, and Klocker H. 1993. Mutant androgen receptor detected in an
advanced-stage prostatic carcinoma is activated by adrenal androgens and
progesterone. Molecular Endocrinology 7(12): 1541 - 1550.
Culig Z, Klocker H, Bartsch G, Steiner H, and Hobisch A. 2003. Androgen receptors
in prostate cancer. Journal of Urology 170 (4 pt 1): 1363 - 1369.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9 2
Cunningham GR, Ashton CM, Annegers JF, Souchek J, Klima M, and Miles B.
2003. Familial aggregation of prostate cancer in African-Americans and White
Americans. Prostate 56 (4): 256 - 262.
Dennis LK and Hayes RB. 2001. Alcohol and prostate cancer. Epidemiologic
Reviews 23 (1): 110-114.
Deutsch E, Maggiorella L, Eschwege P, Bourhis J, Soria JC, and Abdulkarim B.
2004. Environmental, genetic, and molecular features of prostate cancer.
Lancet Oncology 5 (5): 303 - 313.
Ding D, Xu L, Menon M, GPV Reddy, and Barrack ER. 2004. Effect of a short CAG
(glutamine) repeat on human androgen receptor function. The Prostate 58 (1).
23 - 32.
di Sant’ Agnese PA and Cockett ATK. 1996. Neuroendocrine differentiation in
prostatic malignancy. Cancer 78 (2): 357 - 361.
Eder IE, Culig Z, Putz T, Nessler-Menardi C, Bartsch G, and Klocker H. 2001.
Molecular biology of the androgen receptor: from molecular understanding to
the clinic. European Urology 40 (3): 241 - 251.
Faber PW, Kuiper GGJM, van Rooij HCJ, van der Korput JAGM, Brinkmann AO,
Trapman J. 1989. The N-terminal domain of the human androgen receptor is
encoded by one, large exon. Molecular and Cellular Endocrinology 61 (2):
257 - 262.
Faber PW, van Rooij HCJ, van der Korput HAGM, Baarends WM, Brinkmann AO,
Grootegoed JA, and Trapman J. 1991. Characterization of the human
androgen receptor transcription unit. Journal of Biological Chemistry 266
(17): 10743 - 10749.
Faber PW, van Rooij HCJ, Schipper HJ, Brinkmann AO, and Trapman J. 1993. Two
different, overlapping pathways of transcription initiation are active on the
TATA-less human androgen receptor promoter. Journal of Biological
Chemistry 268 (13): 9296 - 9301.
Farnsworth WE. 1999. Prostate stroma: physiology. Prostate 38 (1): 60 - 72.
Feldman BJ and Feldman D. 2001. The development of androgen-independent
prostate cancer. Nature Reviews Cancer 1 (1): 34 - 45.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
93
Feuer EJ, Merrill RM, and Hankey BF. 1999. Cancer surveillance series:
interpreting trends in prostate cancer - part II: cause of death misclassification
and the recent rise and fall in prostate cancer mortality. Journal of the National
Cancer Institute 91 (12): 1025 - 1032.
Freedman ML, Pearce CL, Penney KL, Hirschhom JN, Kolonel LN, Henderson BE,
Altshuler D. 2005. Systematic evaluation of genetic variation at the androgen
receptor locus and risk of prostate cancer in a multiethnic cohort study.
American Journal of Human Genetics 76 (1): 82 - 90.
Fischer L, Catz D, and Kelley D. 1993. An androgen receptor mRNA isoform
associated with hormone-induced cell proliferation. Proceedings of the
National Academy of Sciences 90 (17): 8254 - 8258.
Gann PH, Hennekens CH, and Stampfer MJ. 1995. A prospective evaluation of
plasma prostate-specific antigen for detection of prostatic cancer. Journal of
the American Medical Association 273 (4): 289 - 294.
Gann PH, Hennekens CH, Ma J, Longcope C, and Stampfer MJ. 1996. Prospective
study of sex hormone levels and risk of prostate cancer. Journal of the National
Cancer Institute 88 (16): 1118- 1126.
Gao HW, Li YL, Wu S, Wang YS, Zhang HF, Pan YZ, Zhang L, Tateno H, Sato I,
Kuwahara M, and Zhao XJ. 2005. Mass screening of prostate cancer in a
Chinese population: the relationship between pathological features of prostate
cancer and serum prostate specific antigen. Asian Journal of Andrology 7 (2):
159- 163.
Gao T and McPhaul MJ. 1998. Functional activities of the A and B forms of the
human androgen receptor in response to androgen receptor agonists and
antagonists. Molecular Endocrinology 12 (5): 654 - 663.
Ghadessy FJ, Lim J, Abdullah AAR, Panet-Raymond V, Choo CK, Lumbroso R, Tut
TG, Gottlieb B, Pinsky L, Trifiro MA, and Yong EL. 1999. Oligospermic
infertility associated with an androgen receptor mutation that disrupts
interdomain and coactivator (TIF2) interactions. Journal of Clinical
Investigation 103 (11): 1517 - 1525.
Giles GG, Severi G, English DR, McCredie MRE, Borland R, Boyle P, and Hopper
JL. 2003. Sexual factors and prostate cancer. BJU International 92 (3): 211 -
216.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
94
Giovannucci E, Stampfer MJ, Krithivas K, Brown M, Brufsky A, Talcott J,
Hennekens CH, and KantofFPW. 1997. The CAG repeat within the androgen
receptor gene and its relationship to prostate cancer. Proceedings of the
National Academy of Sciences 94 (7): 3320 - 3323.
Gottlieb B, Beitel LK, Wu JH, and Trifiro M. 2004. The androgen receptor gene
mutations database (ardb): 2004 update. Human Mutations 23 (6): 527 - 533.
GronbergH. 2003. Prostate cancer epidemiology. Lancet 361 (9360): 859-864.
Gsur A, Preyer M, Haidinger G, Zidek T, Madersbacher S, Schatzl G, Marberger M,
Vutue C, and Micksche M. 2002. Polymorphic CAG repeats in the androgen
receptor gene, prostate specific antigen polymorphism and prostate cancer risk.
Carcinogenesis 23 (10): 1647 - 1651.
Guess HA. 2001. Benign prostatic hyperplasia and prostate cancer. Epidemiologic
Reviews 23 (1): 152- 158.
Haapala K, Hyytinen ER, Roiha M, Laurila M, Rantala I, Helin HJ, and Koivisto PA.
2001. Androgen receptor alterations in prostate cancer relapsed during a
combined androgen blockade by orchiectomy and bicalutamide. Laboratory
Investigation 81(12): 1647 - 1651.
Haas GP and Sakr WA. 1997. Epidemiology of prostate cancer. CA: A Cancer
Journal for Clinicians 47 (5); 273 - 287.
Hakimi JM, Schoenberg MP, Rondinelli RH, Piantadosi S, and Barrack ER. 1997.
Androgen receptor variants with short glutamine or glycine repeats may
identify unique subpopulations of men with prostate cancer. Clinical Cancer
Research 3 (9): 1599 - 1608.
Hammarsten J and Hogstedt B. 2002. Calculated fast-growing benign prostatic
hyperplasia: a risk factor for developing clinical prostate cancer.
Scandinavian Journal o f Urology and Nephrology 36 (5): 330 - 338.
Hankey BF, Feuer EJ, Clegg LX, Hayes RB, Legler JM, Prorok PC, Ries LA, Merrill
RM, and Kaplan RS. 1999. 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. Journal of the National Cancer
Institute 91 (12): 1017 - 1024.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
95
Hankin JH, Stram DO, Arakawa K, Park S, Low SH, Lee HP, Yu MC. 2001.
Singapore Chinese health study: development, validation, and calibration of
the quantitative food frequency questionnaire. Nutrition and Cancer 39 (2):
187- 195.
Heinlein CA and Chang C. 2002. Structural and functional analysis of the androgen
receptor. In: Androgens and androgen receptor: mechanisms, ftinctions, and
clinical applications. Chang C. 17 - 55. Norwell, Massachusetts: Kluwer
Academic Publishers.
Heinlein CA and Chang C. 2004. Androgen receptors in prostate cancer. Endocrine
Reviews 25 (2): 276-308.
Hickey K, Do KA, and Green A. 2001. Smoking and prostate cancer. Epidemiologic
Reviews 23 (1): 115 - 125.
Hirata S, Shoda T, Kato J, and Hoshi K. 2003. Isoform/variant mRNAs for sex
steroid hormone receptors in humans. Trends in Endocrinology and
Metabolism 14 (3): 124 - 129.
Hirawat S, Budman DR, and Kreis W. 2003. The androgen receptor: structure,
mutations, and antiandrogens. Cancer Investigation 21 (3): 400 - 417.
Hsing AW, Tsao L, Devesa SS. 2000a. International trends and patterns of prostate
cancer incidence and mortality. International Journal of Cancer 85 (1): 60 -
67.
Hsing AW, Gao YT, Wu G, Wang X, Deng J, Chen YL, Sesterhenn IA, Mostofi FK,
Benichou J, and Chang C. 2000b. Polymorphic CAG and GGN repeat lengths
in the androgen receptor gene and prostate cancer risk: a population-based
case-control study in China. Cancer Research 60 (18): 5111-5116.
Hsu CL, Chen YL, Ting HJ, Lin WJ, Yang Z, Zhang Y, Wang L, Wu CT, Chang HC,
Yeh S, Pimplikar SW, and Chang C. 2005. Androgen receptor (AR) NH2-
and COOH-terminal interactions result in the differential influences on the
AR-mediated transactivation and cell growth. Molecular Endocrinology 19
(2): 350-361.
Huggins C and Hodges CV. 2002. Studies on prostatic cancer I. The effect of
castration, of estrogen, and of androgen injection on serum phosphatases in
metastatic carcinoma of the prostate. Journal of Urology 168 (1): 9 -12.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9 6
Hyytinen ER, Haapala K, Thompson J, Lappalainen I, Roiha M, Rantala I, Helin HJ,
Janne OA, Vihinen M, Palvimo JJ, and Koivisto PA. 2002. Patterns of
somatic androgen receptor gene mutations in patients with hormone-refractory
prostate cancer. Laboratory Investigation 82 (11): 1591 - 1598.
Ingles SA, Ross RK, Yu MC, Irvine RA, La Pera G, Haile RW, and Coetzee GA.
1997. Association of prostate cancer risk with genetic polymorphisms in the
vitamin D receptor and androgen receptor. Journal of the National Cancer
Institute 89 (2): 166 - 170.
Irvine RA, Yu MC, Ross RK, and Coetzee GA. 1995. The CAG and GGC
microsatellites of the androgen receptor gene are in linkage disequilibrium in
men with prostate cancer. Cancer Research 55 (9): 1937 - 1940.
Irvine RA, Ma H, Yu MC, Ross RK, Stallcup MR, and Coetzee GA. 2000. Inhibition
Of pl60-mediated coactivation with increasing androgen receptor
polyglutamine length. Human Molecular Genetics 9 (2): 267 - 274.
James WH. 2004. A hypothesis on the sexual behavior of men who are destined to
develop prostate cancer. International Journal of Epidemiology 33: 1 -2 .
Jarrard DF, Kinoshita H, Shi Y, Sandefiir C, Hoff D, Meisner LF, Chang C, Herman
JG, Isaacs WB, and Nassif N. 1998. Methylation of the androgen receptor
promoter CpG island is associated with loss of androgen receptor expression in
prostate cancer cells. Cancer Research 58 (23): 5310 - 5314.
Jenster G, van der Korput HAGM, van Vroonhoven C, van der Kwast TH, Trapman J,
and Brinkmann AO. 1991. Domains of the human androgen receptor involved
in steroid binding, transcriptional activation, and subcellular localization.
Molecular Endocrinology 5 (10): 1396- 1404.
Jenster G, van der Korput HAGM, Trapman J, and Brinkmann AO. 1995.
Identification of two transcription activation units in the N-terminal domain of
the human androgen receptor. Journal of Biological Chemistry 61 (4): 7341 -
7346.
Kinoshita H, Shi Y, Sandefiir C, Meisner LF, Chang C, Choon A, Reznikoff CR, Bova
GS, Friedl A, and Jarrard DF. 2000. Methylation of the androgen receptor
minimal promoter silences transcription in human prostate cancer. Cancer
Research 60 (13): 3623 - 3630.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9 7
Koivisto PA, Hyytinen ER, Matikainen M, Tammela TLJ, Ikonen T, and Schleutker J.
2004. Germline mutation analysis of the androgen receptor gene in Finnish
patients with prostate cancer. Journal of Urology 171 (1). 431 - 433.
Kolonel LN, Henderson BE, Hankin JH, Nomura AM, Wilkens LR, Pike MC,
Stram DO, Monroe KR, Earle ME, Nagamines FS. 2000. A multiethnic
cohort in Hawaii and Los Angeles: baseline characteristics. American Journal
of Epidemiology 151 (4): 346 - 357.
Kolonel LN. 2001. Fat, meat, and prostate cancer. Epidemiologic Reviews 23 (1): 72
-8 1 .
Kuiper GGJM, Faber PW, van Rooij HCJ, van der Korput JAGM, Ris-Stalpers C,
Klaassen P, Trapman J, and Brinkmann AO. 1989. Structural organization of
the human androgen receptor gene. Journal of Molecular Endocrinology 2 (3):
R 1 -R 4 .
Kuiper GGJM, de Ruiter PE, Grootegoed JA, and Brinkmann AO. 1991. Synthesis
and post-translational modification of the androgen receptor in LNCaP cells.
Molecular and Cellular Endocrinology 80 (1-3): 65 -7 3 .
Kurita T, Medina RT, Mills AA, and Cunha GR. 2004. Role of p63 and basal cells
in the prostate. Development 131 (20): 4955 - 4964.
Labrie F, Candas B, Dupont A, Cusan L, Gomez JL, Suburu RE, Diamond P,
Levesque J, and Belanger A. 1999a. Screening decreases prostate cancer
death: first analysis of the 1988 Quebec prospective randomized controlled
trial. Prostate 38 (2): 83 - 91.
Labrie F and Candas B. 1999b. Various statistical analyses indicate a marked
reduction of prostate cancer death in the Quebec trial. Prostate 40 (2): 132 -
134.
Labrie F and Candas B. 1999c. The Quebec study shows a 69% decrease in prostate
cancer death. Prostate 40 (2): 137.
Labrie F and Candas B. 2002. Interpreting the Quebec prostate cancer study.
Canadian Medical Association Journal 167 (4): 340.
Lander ES and Schork NJ. 1994. Genetic dissection of complex traits. Science 265
(5181): 2037-2047.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
98
Lee DK and Chang C. 2003. Endocrine mechanisms of disease: expression and
degradation of androgen receptor: mechanism and clinical implication.
Journal of Clinical Endocrinology and Metabolism 88 (9): 4043 - 4054.
Legler JM, Feuer EJ, Potosky AL, Merrill RM, and Kramer BS. 1998. The role of
prostate-specific antigen (PSA) testing patterns in the recent prostate cancer
incidence decline in the United States. Cancer Causes and Control 9 (5): 519 —
527.
Leitzmann MF, Platz EA, Stampfer MJ, Willett WC, and Giovannucci E. 2004.
Ejaculation frequency and subsequent risk of prostate cancer. Journal of the
American Medical Association 291 (13): 1578 - 1586.
Liao S and Williams-Ashman HG. 1962. An effect of testosterone on amino acid
incorporation by prostatic ribonucleoprotein particles. Proceedings of the
National Academy of Sciences 48: 1956 - 1964.
Liao S. 1965. Influence of testosterone on template activity of prostatic ribonucleic
acids. Journal of Biological Chemistry 240 (3): 1236 - 1243.
Liao S, Fukuchi J, and Kao YH. 2002. Molecular action of androgens. In: Hormone
therapy in breast and prostate cancer. Jordan VC and Furr BJA. 311-324.
Totowa, New Jersey: Humana Press, Inc.
Lindzey J, Grossmann M, Kumar MV, and Tindall DJ. 1993. Regulation of the 5’-
flanking region of the mouse androgen receptor gene by cAMP and androgen.
Molecular Endocrinology 7 (12): 1530- 1540.
Lobaccaro JM, Lumbroso S, Belon C, Galtier-Dereure F, Bringer J, Lesimple T,
Namer M, Cutuli BF, Pujol H, and Sultan C. 1993. Androgen receptor gene
mutation in male breast cancer. Human Molecular Genetics 2(11): 1799 -
1802.
Lowsley OS. 1912. The development of the human prostate gland with reference to
the development of other structures at the neck of the urinary bladder.
American Journal of Anatomy 13 (3): 299 - 349.
Lubahn DB, Joseph DR, Sullivan PM, Willard HF, French FS, and Wilson EM. 1988.
Cloning of human androgen receptor complementary DNA and localization to
the X chromosome. Science 240 (4850): 327 - 330.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
99
Lubahn DB, Brown TR, Simental JA, Higgs HN, Migeon CJ, Wilson EM, and French
FS. 1989. Sequence of the introns/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. Proceedings of the National
Academy of Sciences 86 (23): 9534 - 9538.
Lyon MF and Hawkes SG. 1970. X-linked gene for testicular feminization in the
mouse. Nature 227 (5264): 1217- 1219.
Mainwaring WIP. 1969. A soluble androgen receptor in the cytoplasm of rat
prostate. Journal of Endocrinology 45 (4): 531 - 541.
Marcelli M, Ittmann M, Mariani S, Sutherland R, Nigam R, Murthy L, Zhao Y,
DiConcini D, Puxeddu E, Esen A, Eastham J, Weigel NL, and Lamb D J.
2000. Androgen receptor mutations in prostate cancer. Cancer Research 60
(4): 944 - 949.
Marker PC, Donjacour AA, Dahiya R, and Cunha GR. 2003. Hormonal, cellular, and
molecular control of prostatic development. Developmental Biology 253 (2):
165 - 174.
McKenna NJ and O’Malley BW. 2002. Minireview: nuclear receptor coactivators -
and update. Endocrinology 143 (7): 2461 -2465.
McNealJE. 1968. Regional morphology and pathology of the prostate. American
Journal of Clinical Pathology 49 (3): 347 - 357.
McNeal JE. 1972. The prostate and prostatic urethra: a morphological synthesis.
Journal of Urology 107 (6): 1008 - 1016.
McNealJE. 1980. Anatomy of the prostate: an historical survey of divergent views.
The Prostate 1: 3 - 13.
McNealJE. 1981a. Normal and pathologic anatomy of prostate. Urology 17
(Suppl 3): 11 - 16.
McNealJE. 1981b. The zonal anatomy of the prostate. The Prostate 2: 35 - 49.
McNeal JE and Bostwick DG. 1984. Anatomy of the prostatic urethra. Journal of the
American Medical Association 251 (7): 890 - 891.
McNealJE. 1988a. Normal histology of the prostate. American Journal of
Surgical Pathology 12 (8): 619 - 633.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
100
McNeal JE, Redwine EA, Freiha FS, and Stamey TA. 1988b. Zonal distribution of
prostatic adenocarcinoma: correlation with histologic pattern and direction of
spread. American Journal o f Surgical Pathology 12(12): 897 - 906.
McPhaul MJ. 2002. Androgen receptor mutations and androgen insensitivity.
Molecular and Cellular Endocrinology 198 (1-2): 61 - 67.
Migeon BR, Brown TR, Axelman J, and Migeon CJ. 1981. Studies of the locus for
androgen receptor: localization on the human X chromosome and evidence for
homology with the tfm locus in the mouse. Proceedings of the National
Academy of Sciences 78 (10): 6339 - 6343.
Mizokami A, Yeh SY, and Chang C. 1994a. Identification of 3’, 5’ - cyclic
adenosine monophosphate response element and other cis-acting elements in
the human androgen receptor gene promoter. Molecular Endocrinology 8(1):
77 - 88.
Mizokami A and Chang C. 1994b. Induction of translation by the 5’-untranslated
region of human androgen receptor mRNA. Journal of Biological Chemistry
269 (41): 25655-25659.
Meyer WJ, Migeon BR, and Migeon CJ. 1975. Locus on human X chromosome for
dihydrotestosterone receptor and androgen insensitivity. Proceedings of the
National Academy of Sciences 72 (4): 1469 - 1472.
Mononen N, Syrjakoski K, Matikainen M, Tammela TL, Schleutker J, Kallioniemi OP
Trapman J, and Koivisto PA. 2000. Two percent of Finnish prostate cancer
patients have a germ-line mutation in the hormone-binding domain of the
androgen receptor gene. Cancer Research 60 (22): 6479 - 6481.
Montgomery JS, Price DK, and Figg WD. 2001. The androgen receptor gene and its
influence on the development and progression of prostate cancer. Journal of
Pathology 195 (2): 138 - 146.
Mooney SD, Klein TE, AltmanRB, TrifiroMA, and Gottlieb B. 2003. Afunctional
analysis of disease-associated mutations in the androgen receptor gene.
Nucleic Acids Research 31 (8): e42.
Ndubuisi SC, Kofie VY, Andoh JY, and Schwartz EM. 1995. Black-White
differences in the stage at presentation of prostate cancer in the District of
Columbia. Urology 46 (1): 71 - 77.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
101
Negri E, Pelucchi C, Talamini R, Montella M, Gallus S, Bosetti C, Franceschi S, and
La Vecchia C. 2005. Family history of cancer and the risk of prostate cancer
and benign prostatic hyperplasia. International Journal of Cancer 114 (4): 648
-652.
Nelson WG, De Marzo AM, and Isaacs WB. 2003. Mechanisms of disease:
prostate cancer. New England Journal of Medicine 349 (4): 366 - 381.
NetterFH. 2001. Atlas of human anatomy: second edition. Teterboro: ICON
Learning Systems.
Newmark JR, Hardy DO, Tonb DC, Carter BS, Epstein JI, Isaacs W, Brown T, and
Barrack ER. 1992. Androgen receptor gene mutations in human prostate
cancer. Proceedings of the National Academy of Sciences. 89 (14): 6319 -
6323.
Ntais C, Polycarpou A, and Tsatsoulis A. 2003. Molecular epidemiology of prostate
cancer: androgens and polymorphisms in androgen-related genes. European
Journal of Endocrinology 149 (6): 469 - 477.
Oefelein MG and Resnick MI. 2004. Association of tobacco use with hormone
refractory disease and survival of patients with prostate cancer. Journal of
Urology 171 (6 Pt. 1): 2281 - 2284.
OsegbeDN. 1997. Prostate cancer in Nigerians: facts and nonfacts. Journal of
Urology 157 (4): 1340 - 1343.
Parkin DM. 2001. Global cancer statistics in the year 2000. Lancet Oncology 2 (9):
533 - 543.
Platz EA, De Marzo AM, and Giovannucci E. 2004a. Prostate cancer association
studies: pitfalls and solutions to cancer misclassification in the PSA era.
Journal of Cellular Biochemistry 91 (3): 553 - 571.
Platz EA, Leitzmann MF, Rimm EB, Willett WC, and Giovannucci E. 2004b.
Alcohol intake, drinking patterns, and risk of prostate cancer in a large
prospective cohort study. American Journal of Epidemiology 159 (5): 444 -
453.
Platz EA and De Marzo AM. 2004c. Epidemiology of inflammation and prostate
cancer. Journal of Urology 171 (2 Pt. 2): S36 - S40.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
102
Poletti A. The polyglutamine tract of androgen receptor: from functions to
dysfunctions in motor neurons. 2004. Frontiers in Neuroendocrinology 25
(1): 1 -2 6 .
Potosky AL, Miller BA, Albertsen PC, and Kramer BS. 1995. The role of
increasing detection in the rising incidence of prostate cancer. Journal of the
American Medical Association 272 (7): 548 - 552.
Poujol N, Lobaccaro JM, Chiche L, Lumbroso S, and Sultan C. 1997. Functional and
structural analysis of R607Q and R608K androgen receptor substitutions
associated with male breast cancer. Molecular and Cellular Endocrinology 130
(1-2): 4 3 -5 1 .
Poujol N, Wurtz JM, Tahiri B, Lumbroso S, Nicolas JC, Moras D, and Sultan C.
2000. Specific recognition of androgens by their nuclear receptor. Journal of
Biological Chemistry 275 (31): 24022 - 24031.
Pugh BF and Tjian R. 1991. Transcription from a TATA-less promoter requires a
multisubunit TFIID complex. Genes and Development 5 (11): 1935 - 1945.
Punglia RS, D’Amico AV, Catalona WJ, Roehl KA, and Kuntz KM. 2003a. Effect of
verification bias on screening for prostate cancer by measurement of prostate-
specific antigen. New England Journal of Medicine 349 (4): 335 - 342.
Punglia RS, Catalona WJ, and Kuntz KM. 2003b. Verification bias in screening for
prostate cancer. New England Journal of Medicine 349 (17): 1672 - 1673.
Quinn M and Babb P. 2002. Patterns and trends in prostate cancer incidence,
survival, prevalence, and mortality. Part I: international comparisons. BJU
International 90 (2): 162- 173.
Rahman M, Miyamoto H, and Chang C. 2004. Androgen receptor coregulators in
prostate cancer: mechanisms and clinical implications. Clinical Cancer
Research 10 (7): 2208 - 2219.
Riegman PHJ, Vlietstra RJ, van der Korput JAGM, Brinkmann AO, and Trapman J.
1991. The promoter of the prostate-specific antigen gene contains a functional
androgen responsive element. Molecular Endocrinology 5 (12): 1921 - 1930.
Roberts WW, Platz EA, and Walsh PC. 2003. Association of cigarette smoking with
extraprostatic prostate cancer in young men. Journal of Urology 169 (2): 512 -
516.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
103
Roberts RO, Bergstrahl EJ, Bass SE, Lieber MM, and Jacobsen SJ. 2004. Prostatitis
as a risk factor for prostate cancer. Epidemiology 15 (1): 93 - 99.
Roche PJ, Hoare S A, and Parker MG. 1992. A consensus DNA-binding site for the
androgen receptor. Molecular Endocrinology 6 (12): 2229 - 2235.
Rosenblatt KA, Wicklund KG, and Stanford JL. 2001. Sexual factors and the risk of
prostate cancer. American Journal of Epidemiology 153 (12): 1152- 1158.
Sartor O, Zheng Q, and Eastham JA. 1999. Androgen receptor gene CAG repeat
length varies in a race-specific fashion in men without prostate cancer.
Urology 53 (2): 378-380.
Schaid DJ. 2004. The complex genetic epidemiology of prostate cancer. Human
Molecular Genetics 13 (Spec No. 1): R103 - R121
Schalken JA and Van Leenders G. 2003. Cellular and molecular biology of the
prostate: stem cell biology. Urology 62 (Suppl 5A): 11 - 20.
Scher HI, Buchanan G, Gerald W, Butler LM, and Tilley WD. 2004. Targeting the
androgen receptor: improving outcomes for castration-resistant prostate cancer.
Endocrine-Related Cancer 11 (3): 459-476.
Schoenberg MP, Hakimi JM, Wang S, Bova GS, Epstein JI, Fischbeck KH, Isaacs
WB, Walsh PC, and Barrack ER. 1994. Microsatellite mutation
(CAG24-> 18) in the androgen receptor gene in human prostate cancer.
Biochemical and Biophysical Research Communications 198 (1): 74 - 80.
Schoonen WM, Salinas CA, Kiemeney LA, and Stanford JL. 2005. Alcohol
consumption and risk of prostate cancer in middle-aged men. International
Journal of Cancer 113 (1): 133 - 140.
Shaffer PL, Jivan A, Dollins DE, Claessens F, and Gewirth DT. 2004. Structural
basis of androgen receptor binding to selective androgen response elements.
Proceedings of the National Academy of Sciences 101 (14): 4758-4763.
Shannon BA, McNeal JE, and Cohen RJ. 2003. Transition zone carcinoma of the
prostate gland: a common indolent tumour type that occasionally manifests
aggressive behavior. Pathology 35 (6): 467 - 471.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
104
Shen HC, Buchanan G, Butler LM, Prescott J, Henderson M, Tilley WD, and Coetzee
GA. 2005. GRIP1 mediates the interaction between the amino- and carboxyl-
termini o f the androgen receptor. Journal of Biological Chemistry 386 (1): 69
-7 4 .
Shiao YH, Buzard GS, Weghorst CM, and Rice JM. 1997. DNA template as a
source of artifact in the detection of p53 gene mutations using archived tissue.
Biotechniques 22 (4): 608 - 612.
Shibata A and Whittemore AS. 1997. Genetic predisposition to prostate cancer:
possible explanations for ethnic differences in risk. Prostate 32 (1): 65 - 72.
Shkolny DL, Beitel LK, Ginsberg J, Pekeles G, Arbour L, Pinsky L, and Trifiro
MA. 1999. Discordant measures of androgen-binding kinetics in two mutant
androgen receptors causing mild or partial androgen insensitivity, respectively.
Journal of Clinical Endocrinology and Metabolism 84 (2): 805 - 810.
Sirovich BE, Schwartz LM, and Woloshin S. 2003. Screening men for prostate and
colorectal cancer in the United States: does practice reflect the evidence?
Journal of the American Medical Association 289 (11): 1414 - 1420.
Stamey TA, Caldwell M, McNeal JE, Nolley R, Hemenez M, and Downs J. 2004.
The prostate specific antigen era in the United States is over for prostate
cancer: what happened in the last 20 years? Journal of Urology 172 (4 pt. 1):
1297- 1301.
Stanford JL, Just JJ, Gibbs M, Wicklund KG, Neal CL, Blumenstein BA, and
Ostrander EA. 1997. Polymorphic repeats in the androgen receptor gene:
molecular markers of prostate cancer risk. Cancer Research 57 (6): 1194 —
1198.
Takane KK and McPhaul MC. 1996. Functional analysis of the human androgen
receptor promoter. Molecular and Cellular Endocrinology 119(1): 8 3 -9 3 .
Taplin ME, Bubley GJ, Shuster TD, Frantz ME, Spooner AE, Ogata GK, Keer HN,
Balk SP. 1995. Mutation of the androgen-receptor gene in metastatic
androgen-independent prostate cancer. New England Journal of Medicine 332
(21): 1393 - 1398.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
105
Taplin ME, Rajeshkumar B, Halabi S, Werner CP, Woda BA, Picus J, Stadler W,
Hayes DF, Kantoff PW, Vogelzang NJ, and Small EJ. 2003. Androgen
receptor mutations in androgen-independent prostate cancer: cancer and
leukemia group b study 9663. Journal of Clinical Oncology 21 (14): 2673 -
2678.
Thompson J, Saatcioglu F, Janne OA, and Palvimo JJ. 2001. Disrupted amino- and
carboxyl terminal interactions of the androgen receptor are linked to androgen
insensitivity. Molecular Endocrinology 15 (6): 923 - 935.
Thompson J, Hyytinen ER, Haapala K, Rantala I, Helin HJ, Janne OA, Palvimo JJ,
and Koivisto PA. 2003. Androgen receptor mutations in high-grade prostate
cancer before hormonal therapy. Laboratory Investigations 83 (12): 1709 -
1713.
Tilley WD, Marcelli M, Wilson JD, and McPhaul MJ. 1989. Characterization and
expression of a cDNA encoding the human androgen receptor. Proceedings of
the National Academy of Sciences 86 (1): 327 - 331.
Tilley WD, Marcelli M, and McPhaul MJ. 1990. Expression of the human
androgen receptor gene utilizes a common promoter in diverse human tissues
and cell lines. Journal of Biological Chemistry 265 (23): 13776 - 13781.
Tilley WD, Buchanan G, Hickey TE, and Bentel JM. 1996. Mutations in the
androgen receptor gene are associated with progression of human prostate
cancer to androgen independence. Clinical Cancer Research 2 (2): 277 - 285.
Tilley WD, Buchanan G, and Coetzee GA. 2003. Androgen receptor signaling in
prostate cancer. In: Hormones, genes, and cancer. Henderson BE, Ponder B,
and Ross RK. 288 - 315. New York: Oxford University Press.
Timms BG. 1997. Anatomical perspectives of prostate development. In: Prostate,
basic and clinical aspects. NazRK. 2 9 -5 1 . Boca Raton: CRC Press LLC.
Trapman J, Klaassen P, Kuiper GGJM, van der Korput JAGM, Faber PW, van Rooij
HCJ, Geurts van Kessel A, Voorhorst MM, Mulder E, and Brinkman AO.
1988. Cloning, structure, and expression of a cDNA encoding the human
androgen receptor. Biochemical and Biophysical Research Communications
153 (1): 241-248.
Trapman J and Brinkmann AO. 1996. The androgen receptor in prostate cancer.
Pathology, Research and Practice 192 (7): 752 - 760.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
106
Tseng M, Breslow RA, DeVellis RF, and Ziegler RG. 2004. Dietary patterns and
prostate cancer risk in the national health and nutrition examination survey
epidemiological follow-up study cohort. Cancer Epidemiology, Biomarkers,
and Prevention 13 (1): 71 - 77.
Tsujimoto Y, Takakuwa T, Takayama H, Nishimura K, Okuyama A, Aozasa K, and
Nonomura N. 2004. In situ shortening of the CAG repeat length within the
androgen receptor gene in prostatic cancer and its possible precursors. The
Prostate 58 (3): 283 - 290.
United States Preventive Services Task Force (USPSTF). 2002a. Screening for
prostate cancer: recommendation and rationale. Annals of Internal Medicine
137 (11): 915 -916.
United States Preventive Services Task Force (USPSTF). 2002b. Screening for
prostate cancer: An update of the evidence for the U.S. Preventive Services
Task Force. Annals of Internal Medicine 137 (11): 917 — 929.
Van de Voorde WM. 1996. Pathology of prostatic carcinoma. In: Carcinoma of the
prostate. Petrovich Z, Baert L, and Brady LW. 27 - 50. Berlin Heidelberg:
Springer-Verlag.
Veldscholte J, Ris-Stalpers C, Kuiper GG, Jenster G, Berrevoets C, Claasen E, van
Rooij HC, Trapman J, Brinkmann AO, and Mulder E. 1990. A mutation in
the ligand binding domain of the androgen receptor of human LNCaP cells
affects steroid binding characteristics and response to anti-androgens.
Biochemical and Biophysical Research Communications 173 (2): 534 - 540.
Wang Q, Lu JH, and Yong EL. 2001a. Ligand- and coactivator-mediated
transactivation function (AF2) of the androgen receptor ligand-binding domain
is inhibited by the cognate hinge region. Journal of Biological Chemistry 276
(10): 7493 - 7499.
Wang Y, Hayward SH, Cao M, Thayer KA, Cunha GR. 2001b. Cell differentiation
lineage in the prostate. Differentiation 68 (4 - 5): 270 - 279.
Wang Q, Udayakumar TS, Vasaitis TS, Brodie AM, and Fondell JD. 2004.
Mechanistic relationship between androgen receptor polyglutamine tract
truncation and androgen-dependent transcriptional hyperactivity in prostate
cancer cells. Journal of Biological Chemistry 279 (17): 17319 - 17328.
Wang L, Hsu CL, and Chang C. 2005. Androgen receptor corepressors: an overview.
The Prostate 63 (2): 117 - 130.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
107
Weinmann S, Richert-Boe K, Glass AG, and Weiss NS. 2004. Prostate cancer
screening and mortality: a case-control study (United States). Cancer Causes
and Control 14 (2): 133 - 138.
Wilt TJ. 2003. Prostate cancer: epidemiology and screening. Reviews in Urology 5
(Suppl 6): S3 - S9.
Wooster R, Mangion J, Eeles R, Smith S, Dowsett M, Averill D, Barrett-Lee P, Easton
DF, Ponder BA, and Stratton MR. 1992. A germline mutation in the androgen
receptor gene in two brothers with breast cancer and Reifenstein syndrome.
Nature Genetics 2 (2): 132 - 134.
Xue W, Irvine RA, Yu MC, Ross RK, Coetzee GA, and Ingles SA. 2000.
Susceptibility to prostate cancer , interaction between genotypes at the
androgen receptor and prostate-specific antigen loci. Cancer Researcg 60 (4):
839-841.
Yong EL, Lim J, Qi W, Ong V, and Mifsud A. 2000. Molecular basis of
androgen receptor diseases. Annals of Medicine 32 (1): 15 - 22.
Yong EL, Loy CJ, and Sim KS. 2003. Androgen receptor gene and male infertility.
Human Reproduction Update 9 (1): 1-7.
Zeegers MP, Kiemeney LALM, Nieder AM, and Ostrer H. 2004. How strong is the
association between CAG and GGN repeat length polymorphisms in the
androgen receptor gene and prostate cancer risk? Cancer Epidemiology,
Biomarkers, and Prevention 13 (11): 1765 - 1771.
Zhang HF, Wang HL, Xu N, Li SW, Ji GY, Li XM, Pan YZ, Zhang L, Zhao XJ, and
Gao HW. 2004. Mass screening of 12,027 elderly men for prostate carcinoma
by measuring serum prostate specific antigen. Chinese Medical Journal 117
(1): 67-70.
Zhou ZX, Sar M, Simental JA, Lane MV, and Wilson EM. 1994. A ligand-
dependent bipartite nuclear targeting signal in the human androgen receptor.
Journal of Biological Chemistry 269 (18): 13115 - 13123.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
A case-control study of passive smoking and bladder cancer risk in Los Angeles
PDF
BRCA1 mutations and polymorphisms in African American women with a family history of breast cancer identified through high throughput sequencing
PDF
CYP17 polymorphism and risk for colorectal adenomas
PDF
Single-nucleotide polymorphisms in 17beta-hydroxysteroid dehydrogenase type III and prostate cancer risk
PDF
Dietary fiber intake and atherosclerosis progression: The Los Angeles Atherosclerosis Study
PDF
Cigarettes and alcohol in relation to colorectal cancer within the Singapore Chinese Health Study
PDF
Extent, prevalence and progression of coronary calcium in four ethnic groups
PDF
Colorectal cancer risks in Singapore Chinese: Polymorphisms in the insulin-like growth factor-1 and the vitamin D receptor
PDF
A pilot survey of medical abortion knowledge and practices among obstetrician/gynecologists and family practitioners in Los Angeles County
PDF
Association of vitamin D receptor gene polymorphisms with colorectal adenoma
PDF
Effect of hormone therapy on the progression of carotid-artery atherosclerosis in postmenopausal women with and without established coronary artery disease
PDF
Determinants of mammographic density in African-American, non-Hispanic white and Hispanic white women before and after the diagnosis with breast cancer
PDF
beta3-adrenergic receptor gene Trp64Arg polymorphism and obesity-related characteristics among African American women with breast cancer: An analysis of USC HEAL Study
PDF
Agreement between administrative and clinical data for term newborns with congenital malformations
PDF
Comparisons of metabolic factors among gestational diabetes mellitus probands, siblings and cousins
PDF
Development and evaluation of standardized stroke outcome measures in a population of stroke patients in rural China
PDF
A descriptive analysis of medication use by asthmatics in the Children's Health Study, 1993
PDF
Association between latchkey status and smoking behavior in middle school children
PDF
Descriptive epidemiology of thyroid cancer in Los Angeles County, 1972-1995
PDF
Evaluation of the accuracy and reliability of self-reported breast, cervical, and ovarian cancer incidence in a large population-based cohort of native California twins
Asset Metadata
Creator
Wantroba, Marcus Allan
(author)
Core Title
Infrequent androgen receptor mutations in primary prostate tumors from men residing in Singapore and Los Angeles
School
Graduate School
Degree
Master of Science
Degree Program
Molecular Epidemiology
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
biology, biostatistics,health sciences, public health,OAI-PMH Harvest
Language
English
Contributor
Digitized by ProQuest
(provenance)
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c16-46808
Unique identifier
UC11337880
Identifier
1430408.pdf (filename),usctheses-c16-46808 (legacy record id)
Legacy Identifier
1430408.pdf
Dmrecord
46808
Document Type
Thesis
Rights
Wantroba, Marcus Allan
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA
Tags
biology, biostatistics
health sciences, public health