University of Southern California Dissertations and Theses

Single-nucleotide polymorphisms in 17beta-hydroxysteroid dehydrogenase type III and prostate cancer risk

(USC Thesis Other)

Single-nucleotide polymorphisms in 17beta-hydroxysteroid dehydrogenase type III and prostate cancer risk

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content SINGLE NUCLEOTIDE POLYMORPHISMS
IN 17P-HYDROXYSTEROID DEHYDROGENASE TYPE III
AND PROSTATE CANCER RISK
by
Eugene Kim
A Thesis Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHRERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(MOLECULAR EPIDEMIOLOGY)
MAY 2003
Copyright 2003 Eugene Kim
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UMI Number: 1416561
UMI
UMI Microform 1416561
Copyright 2003 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.
UNIVERSITY OF SOUTHERN CALIFORNIA
THE GRADUATE SCHOOL
UNIVERSITY PARK
LOS ANGELES, CALIFORNIA 90089-1695
This thesis, written by
under the direction o f h &qr thesis committee, and
approved by all its members, has been presented to and
accepted by the Director o f Graduate and Professional
Programs, in partial fulfillment o f the requirements fo r the
degree o f c f C,C Z^ a L
Director
Date May 1 6 . 2 0 0 3
Thesis Committee
Chair
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Dedication
To the people who make life bearable...
my parents,
my sister,
my brother,
and my husband.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Acknowledgements
First of all, it is my greatest pleasure to acknowledge my advisor, Dr. Juergen
Reichardt, for his support. I would like to express my most sincere and deepest
gratitude to him not only for his scientific guidance but also for his understanding
and patience.
I also with to thank the committee members, Dr. Giske Ursin and Dr. Sue Ingles,
for their invaluable comments, helpful feedbacks, and motivating encouragements.
My colleagues at USC, Katia Margiotti and Dr. C. Leigh Pearce, deserve no less
acknowledgement for their support to complete this thesis.
Also, I thank Duk-won, Nancys (Ya-hsuan and Shih-chi), Jene, Cheryl, Hui-lee,
Mark, Troy, Cluadia, Lucio, Mark, Jione, Tim, Takako, and Terry for being good
fiiends during my studies at USC.
Finally, I would like to further thank my family for their emotional support and
understanding. I owe millions of thanks to my parents for consistently inspiring me
to keep seeking delight of learning. I thank my sister, Yoo-Jeong, and my brother,
Jung-FIwan, for being there for me when I need. Last but most, I would like to offer
my deepest appreciation to my husband, Hyun Woong Shin, for his love and
endurance.
iii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table of Contents
Dedication ii
Acknowledgements iii
List of Tables vi
List of Figures vii
Abstract viii
1 Introduction 1
1.1 Descriptive Epidemiology 2
1.1.1 Histopathology of Prostate Cancer 2
1.1.2 Incidence Pattern in the US: Age,
Racial/Ethnical Variation and Time Trend 3
1.2 Biological Rational of Prostate Cancer 6
1.2.1 Androgens 6
1.2.2 Genetic Component of Prostate Cancer 9
1.2.3 Type III 17(3-Hydroxysteroid Dehydrogenase 10
1.3 Main Hypothesis 12
2 Preliminary Data 15
3 Method & Materials 29
3.1 Multiethnic Cohort in Hawaii and Los Angeles (MEC) 29
3.2 Study Subjects 30
3.3 Experimental Method 31
3.3.1 Multiplex PCR 32
3.3.2 Multiplex Automated Primer Extension Analysis (MAPA) 34
3.3.3 Automated DNA Sequencing 36
iv
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3.3.4 Restriction Fragment Length Polymorphism (RFLP) 37
3.4 Statistical Analysis 38
3.4.1 Case-Cohort Design 38
3.4.2 Data Analysis 40
4 Results 43
4.1 Results from the Experiments 43
4.2 Results from Data Analysis 48
4.3 Results from Hardy-Weinberg Law 54
5 Conclusion 58
5.1 Association between SNPs in the HSD17B3 Gene and 5 8
Prostate Cancer Risk
5.1.1 Study Population and Subjects 59
5.1.2 Study Design and Power 60
5.1.3 Experimental Methods 63
5.1.4 Hardy-Weinberg Equilibrium 65
5.1.5 Inconsistent Results between Preliminary Data and this 66
study
5.2 Future Direction 67
5.3 Conclusion 69
References 72
V
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
List of Tables
Table Title Page
2.1 Association of the G289S SNP in the HSD17B3 gene with 26
Prostate Cancer
4.1 Demographic characteristics between cases and controls 49
4.2 Mean and Standard Deviation of Age, BMI, BMI at Age 21 50
4.3 Distribution of the V3II, II02F, and G289S Genotypes and ORs 51
for the association between genotypes and prostate cancer risk
4.4 ORs for the association between the V3II genotype and prostate 52
cancer risk by racial/ethnic group
4.5 ORs for the association between the G289S variants and prostate 53
cancer risk by racial/ethnic group
4.6 Distribution of SNPs by Ethnicity within the Controls 54
4.7 Observed vs. Expected Numbers of the G289S by Ethnicity 56
5.1 Expected case-control sets calculated from relative risk and 62
allele frequency (80% power and 0.05 significance; Quanto)
vi
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
List of Figures
Figure Title Page
1.1 Incidence Pattern of Prostate Cancer in Ethnic Variation 4
1.2 Time Trend of Prostate Cancer among African Americans and 5
Caucasians in the US from 1973 to 1995
1.3 Androgen Action 7
1.4 Biosynthetic Reduction of Androstenedione to Testosterone 1 1
1.5 Schematic Diagram of the Human HSD17B3 Gene with the 12
Position of the SNPs
4.1 Multiplex PCR 44
4.2 Results of MAPA 45
4.3 Sequencing Results 46
4.4 Results of RFLP with the Ban I Restriction Enzyme 47
vii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Abstract
It has been hypothesized that several genes involved in androgen metabolism
may be associated with prostate cancer risk. Among these, the HSD17B3 gene
encodes Type III 17(3 - hydroxy steroid dehydrogenase, which converts
androstenedione to testosterone primarily in the testis. In this study, we assessed the
association between three genotypes that result in amino acid substitutions at codon
31, 102 and 289 in the HSD17B gene and risk of prostate cancer among 549 cases
and 449 controls in a case-cohort study within the Los Angeles/Hawaii Multiethnic
Cohort Study (MEC). Restriction fragment length polymorphism (RFLP) and
multiplex automated primer extension analysis (MAPA, were used to screen
genotyping data. We did not find any statistically significant association between any
of the missense substitutions in the HSD17B3 gene and prostate cancer risk in all the
ethnic groups combined (White, Black, Latino, and Japanese in the US). The odds
ratio of having the VI genotype in V31I is 1.10 (95% Cl (confidence intervals)
=0.54-2.22) compared to the W genotype, and the odds ratio of having the GS+SS
genotype in G289S is 1.31 (95% 0=0.89-1.96), compared to the GG genotpye.
However, among Japanese-American men, compared to those with two G alleles in
G289S, those with one or more S allele had an OR of 1.97 (95% 0=1.05-3.71).
viii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
These results suggest in general the HSD17B3 variants are not good biomarkers of
susceptibility to prostate cancer. Our finding in the Japanese-American group
suggests that the effects of these variants in the HSD17B3 gene might be more
important in Asian population.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter 1: Introduction
Prostate cancer is the second most common cancer among men in the U.S
(American Cancer Society, Cancer Facts and Figures, 2002: http://www.cancer.org).
189,000 new cases and 30,200 deaths were expected in the U.S. during 2002 (Jemal
et al., 2002) and this would be approximately 30% of total cancer incidences and the
second leading cause of cancer death in men (American Cancer Society, Cancer
Facts and Figures, 2002: http://www.cancer.org). At present, men in the US have a
one in five lifetime risk of being diagnosed with prostate cancer, and approximately
20% among them will have a metastatic disease upon diagnosis (Cancer statistics
2001: http://www.cancer.orgV
Due to the high incidence and mortality and the associated enormous cost of
cancer detection and treatment, prostate cancer has become a heavy burden on public
health. Although intensified efforts have been made to understand the etiology of
prostate cancer in the past several decades, it remains unclear. However, some factors
such as age, race/ethnicity, family history, dietary fat and hormonal factors have been
implicated as potential risk factors (Bernstein & Ross, 1991; Whittemore et al., 1995;
Ross & Schottenfeld, 1996; Lesko et al., 1996; Peehl, 1999; Andersson et al., 1996;
Gann et al., 1995).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1.1 Descriptive Epidemiology
1.1.1 Histopathology of Prostate Cancer
The prostate is a muscular gland surrounding the urethra below the
bladder; it contributes to a secretion called the seminal fluid (McNeal, 1992).
In humans, the gland is divided into three zones (peripheral, transition, and
central), of which most carcinomas arise in the peripheral zone (Miller &
Torkko, 2001). Spread of prostate cancer occurs by direct local invasion and
through the blood stream and lymph. Several clinical staging and histologic
grading systems have been introduced to define how serious a tumor is, and
the most commonly used methods in the US are the tumor/node/metastasis
(TNM) system for staging and the Gleason system for grading (Fleming et al.,
1997; Spencer et al., 1998). The TNM system is based on observations of
presence, size, location and degree of invasion of tumor cells. It describes the
extent of the primary tumor (T stage), the absence or presence of spread to
nearby lymph nodes (N stage), and the absence or presence of distant spread,
or metastasis (M stage). The Gleason system evaluates how effectively
cancer cells are able to structure themselves into glands resembling those of
the normal prostate (Gleason, 1977). The Gleason pattern score consists of 2
to 10 (2 = best differentiated, 10 = poorest differentiation), and it has been
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
shown that a well-differentiated tumor is not less aggressively malignant than
a poorly differentiated one.
1.1.2 Incidence Pattern in the US: Age, Racial/Ethnical Variation and Time
Trend
Incidence of prostate cancer increases with age. In fact, the increase in
prostate cancer incidence by age is faster than any other major cancer
(Cancer statistics, 2001: http://www.cancer.org).
The most distinctive incidence pattern has been observed in
racial/ethnic variation (Figure 1.1): the highest at-risk population is
African-Americans, followed by non-Hispanic white and Hispanic white, and
Asian Americans with the lowest in the U.S (Bernstein & Ross, 1991). For
decades, African-Americans have had the highest prostate cancer risk in the
world (Parkin et al., 2001). Non-Hispanic whites in the US also have
relatively high rates in comparison to other populations around the world,
even though their age-adjusted incidence rate is much less than that of blacks
(225.0/100,000 compared with 149.2/100,000).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
White Hispanic
White Non-Hispanic
Black
Asian/Pacific Islander
Americ an Indian
0 S O too 150 200 250
Age-adjusted Incidence rate (per 100,000)
Figure 1.1 Incidence Pattern of Prostate Cancer in Ethnic Variation.
Age adjusted incidence rate by race/ethnicity in US from 1990 to 1995
(obtaind from SEER data: http://seer.cancer.gov).
Incidence rates for prostate cancer in the US have increased since the
early 1970s, followed by a sharp increase in 1991, and peaked in 1992 to
1993. Since then there has been a gradual decline as shown in Figure 1.2
(Ries et al., 2002). There was a sharp rise in incidence in the early1990s,
which might be attributed to the introduction of a new screening method
(PSA testing).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
o 300
Series 1
250
5 200
150
100
1993 1989 1985 1977 1981 1973
Year
Figure 1.2 Time Trend of Prostate Cancer among African Americans
(series 1) and Non-Hispanic Whites (series 2) in the US from 1973 to 1995
(obtained from SEER data: http://seer.cancer.gov).
1.2 Biological Rational of Prostate Cancer
1.2.1 Androgens
It has been hypothesized for several decades that prostate cancer is
androgen dependent (Huggins & Hodges, 1941). In early studies, Huggins et
al. demonstrated that growth of prostate cancer cells is inhibited by lowering
androgens through either castration or estrogen injection (Huggins & Hodges,
1941). On the other hand, increasing androgen levels by androgen injection
activates tumor cell growth (Huggins & Hodges, 1941).
5
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Androgens are male sex hormones, and cell proliferation in prostate
depends on androgenic metabolism. Figure 1.3 shows how androgens activate
cellular proliferation in the prostate. Testosterone in the circulation goes into
prostate cells and is converted to dehydrotestosterone (DHT), and binding of
DHT to the androgen receptor (AR) initiates to activate cellular growths in
the prostate. This androgen action within the prostate is determined by DHT
concentration as well as other factors including testosterone levels.
Many studies have supported that prostate cancer is related to high
levels of androgens (Noble, 1977; Henderson et al., 1982; Gann et al., 1995).
It was also observed that testosterone levels are high in African Americans,
intermediate in Caucasians and Latinos, and lower in Asians (Honda et al.,
1988; Ross et al., 1992), and this pattern is similar to the risk rates of the
prostate cancer among the ethnic groups.
Moreover, many molecular epidemiologic studies showed that single
nucleotide polymorphisms (SNPs) in the genes related to androgen metabolic
pathways might be associated with prostate cancer risk (Lunn et al., 1999;
Makridakis et al., 1999; Febbo et al., 1999), although the association is weak
and some results are not consistent (Kantoff et al., 1997; Wadelius et al., 1999;
Pearce et al., 2002). This weak association might be because the pattern of
inheritance in multifactorial diseases, like prostate cancer, does not follow
6
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Mendelian laws due to multiple genetic and environmental causalities (Risch,
2000).
Androstenedione
17b-HSD
i '
Testosteron
I
Androgen-Responsive Cell
Ligand
Binding
Dimerualion and
Phosphorylation^^.
ah k An )
y r " H G P ■ >
rts>
hsp
ahJ
e u o s
C o -a c fK a to l
n s c iu itn w n *
^ ? r n | ■
■ r X Z iJ
Androgen-Response
element
I
Target Gene Activation
f ~ l ~ T . ,
fPSA f Growth ( f Survival
Biological Responses
Figure 1.3 Androgen Action. Testosterone circulates in the blood,
where it is bound to albumin and sex-hormone-binding globulin (SHBG), and
only a small percentage of testosterone enters the prostate cells and activates
cellular proliferation. (Modified from Feldman & Feldman, 2001).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Thus, if increasing androgen levels are truly related to a high risk of
prostate cancer, genes regulating androgen levels of humans might be good
candidate genes, as one of the multiple causal factors for prostate cancer.
1.2.2 Genetic Component of Prostate Cancer
A family history of prostate cancer increases the risk of prostate
cancer in sporadic cases (Whittemore et al, 1995). It has been consistently
found in many epidemiologic studies that the frequency of prostate cancer
among the first degree relatives of prostate cancer patients is higher than the
frequency of prostate cancer either in the relatives of men without prostate
cancer or in the general male population (Whittemore et al., 1995; Cerhan et
al., 1999). This finding suggests that a series of genes likely contribute to
prostate cancer risk in subtle ways; lifelong exposure to the effects of these
genes increases prostate cancer risk.
The racial-ethnic variations, one of the strongest risk factors in
prostate cancer incidence, have suggested that genetic components play an
important role of the risk of the disease (Ross et al., 1999) as much as
environmental factors. Recently, it has become genetic components are
empathized on the etiology o f prostate cancer.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1.2.3 Type III 17B-Hydroxysteroid Dehydrogenase
Type III 17(3-hydroxysteroid dehydrogenase converts androstenedione
to testosterone (Figure 1.4), using NADPH as a cofactor (Luu-The et al.,
1995). Testosterone is the biologically active form of androgen and plays a
key role for development and growth of the internal male reproductive
strictures (Moghrabi et al., 1998). Complete deficiency of type III
17(3-hydroxysteroid dehydrogenase leads to male pseudohermaphroditism,
having the internal male reproductive stricture (epididymes, seminal vesicles
and vas deferens) and the female external genitalia (lack of male urethra,
prostate, penis and scrotum) (Geissler et al., 1994). Type III
17 ( 3 - hydroxysteroid dehydrogenase is cmcial to develop the prostate.
Inactive Active
O H
NADPH
Testosterone Andro stene dione
Figure 1.4 Biosynthetic Reduction of Androstenedione to Testosterone.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The enzyme is encoded by the HSD17B3 gene, which is located in
chromosome band 9q22, consists of 11 exons (Figure 1.5) and spans
approximately 60 kb of genomic DNA (Geissler et al., 1994). Although at
least 18 mutations and one polymorphism in this gene were reported
(Geissler et al., 1994; Anderson et al., 1996; Moghrabi et al., 1998), the
previous studies were mainly focused on research of male
pseudohermaphroditism. However, our preliminary data shows that cSNPs
(coding region Single Nucleotide Polymorphisms) in the HSD17B3 gene may
also be related to the etiology of prostate cancer (see Chapter 2 Preliminary
Data).
G289S
GGT—^ AGT
I102F
A TT-sU TT
m IV
it VII VII1 I - '
XI
Figure 1.5 Schematic Diagram of the Human HSD17B3 Gene with the
Position of the Single Nucleotide Polymorphisms (Reproduced with
permission from Katia Margiotti, affiliation).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1.3 Main Hypothesis
I propose that polymorphisms in the HSD17B3 gene may increase
prostate cancer susceptibility. That is, if there is a genetic variance on this gene,
the enzyme activity may increase, the level of testosterone may elevate, and then
the risk of prostate cancer may increase. Also, I hypothesized -as others have-
that some of the differences of prostate cancer risk among the ethnic groups
might be explained by variation over decades of life in androgen metabolism and
synthesis (Haiman, 2001). For this, I assessed the association between the
HSD17B3 genotypes and the risk of prostate cancer in a case-cohort study within
the Hawaii-Los Angeles Multi-ethnic Cohort Study of Diet and Cancer. This
study includes men from four different ethnic groups (Caucasian-American,
African-American, Latino-American, and Japanese-American). Three amino acid
substitutions in HSD17B3 gene coding regions (V31I: Valine at codon 31
replaced by Isoluecineon exon 1,1102F: Isoluecine at codon 102 replaced by
Phenylalanine on exon 4, and G289S: Glycine at codon 289 replaced by Serine
on exon 11) were screened in the study population to see if the frequency of these
polymorphisms is different between cases and controls. Based on our previous
study (done by Katia Margiotti), there are seven SNPs in the intronic region, one
SNP in the 5’ untranslated region, one silent polymorphisms in the coding resign,
and only these three amino acid substitutions in the HSD17B3 gene.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Polymorphisms causing amino acid changes affect enzymatic kinetics more
likely than SNPs in the intronic region and/or silent polymorphisms. Thus, we
focused on those amino acid substitutions in this study. The G289S was
previously reported (Moghrabi et al, 1998), and the V31I and the I102F were
found in our labs by Katia Margiotti.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter 2: Preliminary Data (The Prostate, 2002; 53:65-68)
Association of the G289S Single Nucleotide Polymorphism in the
HSD17B3 Gene
with Prostate Cancer in Italian Men
Katia Margiotti 2’4, Eugene Kim * ’3, C. Leigh Pearce 3, Enrico Spera 4 , Giuseppe
Novelli4 and Juergen K. V. Reichardt ’’2’3’5
1 Institute for Genetic Medicine
2 Department of Biochemistry and Molecular Biology
3 Department of Preventive Medicine
USC Keck School of Medicine
Los Angeles, CA; USA
4 Sezione di Genetica
Universita’ di Tor Vergata
Roma; Italy
5 Corresponding Author
Institute for Genetic Medicine
USC Keck School of Medicine
1 3
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
IGM 240
2250 Alcazar Street
Los Angeles, CA 90089-9075; USA
Phone: 323-442-1529
FAX: 323-442-2764
Email: reichard@hsc.usc.edu
Abstract
Background: Prostate cancer is a significant public health problem in this
country. Substantial data support a plausible role for androgens in the etiology of this
disease. The human HSD17B3 gene encodes the testicular (or type III)
17 3 - hydroxysteroid dehydrogenase enzyme which catalyzes testosterone
biosynthesis in men.
Methods: We have investigated the G289S (glycine at codon 289 replaced by
serine) polymorphism at the HSD17B3 locus as a candidate SNP (single nucleotide
polymorphism) for prostate cancer risk in constitutional DNA from 103 Italian
prostate cancer cases and 109 Italian disease-free centenarians to assess the role of
this SNP in susceptibility to prostate cancer.
Results: The G289S polymorphism confers a significant increase in risk for
prostate cancer (OR= 2.5; CI= 1.03-6.07) in our study population.
14
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Conclusions: Our data are consistent with a plausible role of the G289S SNP
in prostate cancer susceptibility. Therefore, the HSD17B3 gene may be a plausible
candidate gene for prostate cancer risk.
Introduction
It is estimated that in 2001, over 198,100 men will be diagnosed with prostate
cancer and 31,500 men will die from it in the United States (1). In fact, prostate
cancer is the most commonly diagnosed serious malignancy among men in this
country (1). Prostate cancer is rare before the age of 30 years, but the rate of increase
thereafter is exponential such that by the time a man is 90 years old, he has almost a
100 percent chance of having the disease (e.g. 2). There is a large variation in
prostate cancer incidence rates between racial/ethnic groups in the United States,
highest among African-Americans, intermediate among Caucasian-Americans and
Hispanic-Americans, and lowest among Asian-Americans (e.g. 3). These
epidemiologic data provide support for a significant genetic component to prostate
cancer risk, although environmental risk factors must also be involved. In addition,
molecular epidemiological and biochemical evidence suggests that variations on
androgen levels may play a role in the risk for prostate cancer (e.g. 4). Testosterone
plays an important role in normal and abnormal prostate development. Both
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
testosterone and dihydrotestosterone have been have proven to have a variable, but
generally correlation with prostate cancer risk (e.g. 6).
In men, testosterone is produced in large amounts by the testes (7). This
tissue is -in fact- the main source of the hormone in males (7). Testosterone is then
irreversibly metabolized to dihydrotestosterone, mainly in the prostate gland.
Dihydrotestosterone (or, much less efficiently, testosterone) is bound by the
androgen receptor (7). This complex then translocates to the cell nucleus where it
activates transcription of several genes with androgen-responsive elements in their
promoters (7). This process ultimately results in increased DNA synthesis.
Dihydrotestosterone is known to promote DNA synthesis and cell replication in the
prostate (7). Thus, increased prostatic androgens levels may increase the chance for
prostate cancer risk. Testosterone is synthesized from androstenedione and activates
the androgen receptor to initiate the male-like development of the Wolffian
duct-derived internal genitalia (epididymis, vas deferens, seminal vesicles, and
ejaculatory ducts) (8).
The 17|3-hydroxysteroid dehydrogenase type 3 enzyme which is encoded by
the human HSD17B3 gene is an integral membrane protein located in the
endoplasmic reticulum and uses NADPH as a cofactor (8). The human gene
encoding the testicular 17(3-hydroxysteroid dehydrogenase type III isozyme,
HSD17B3, maps to chromosome band 9q22 and is comprised of 11 exons spread out
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
over at least 60Kb of genomic DNA (8). Testicular 17(3 -hydroxysteroid
dehydrogenase activity provides most of this hormone in men and, therefore, is
crucial to normal male sexual development, a complex process that requires the
correct developmental interpretation of both genetic and hormonal signals. Mutations
in HSD17B3 cause a rare human disorder, male pseudohermaphroditism (8). Males
with this disorder are female at the birth but develop male musculature and other
secondary sex characteristics at puberty (8). These mutations however, do not appear
to be involved in prostate diseases in adults. Moghrabi et al. (9) reported a missense
substitution, G289S, resulting from a single nucleotide polymorphism (SNP) in exon
11 of the HSD17B3 gene. The polymorphic allele encodes a protein with a serine
rather than a glycine at position 289 (by replacing the normal GGT codon with
AGT).
We examined the hypothesis that the HSD17B3 locus is a candidate gene for
prostate cancer by genotyping the G289S polymorphism in a total of 103 Italian
prostate cancer cases and 109 Italian centenarian disease-free men. We report here a
statistically significant increased risk for prostate cancer (OR=2.5; CI= 1.03-6.07) in
this study that supports the hypothesis that the HSD17B3 gene may play a significant
role in prostate cancer risk.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Materials and Methods
Study Population
We screened 103 Italian prostate cancer cases and 109 Italian “centenarian”
disease-free men which we previously described in detail (10). Briefly, each patient
aged 53-85 years was accurately evaluated by histopathological, biochemical and
immunological analysis. All patients were considered sporadic since they did not
have an affected first degree relative. The “centenarians” were recruited from a group
of unrelated centenarians and other controls with an age distribution ranging from
66-134 years (10). No familial history of prostate cancer was documented on the
basis of medical records available. They were also considered free of any symptom
of prostate cancer by digital rectal examination (DRE) and prostate specific antigen
(PSA) level.
Genotyping
Genomic DNA was isolated from peripheral blood lymphocytes of 109
Italian unaffected centenarians and 103 Italian prostate cancer cases (10). We
screened for the G289S mutation by PCR amplification of exon 11 of the HSD17B3
gene (with the primers 5'- GATGAACTGAGGTACTTGTTATT and
3' - GAGGA A A AGGTTGT GCTGGACTCCT custom synthesized by Life
Technologies; Rockville, MD, USA), the polymerase chain reaction thermocycler
programs was: 30 cycles of 95 °C for 30 sec, 60 °C for 30 sec and 72 °C for 30 sec
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
followed by SNaPshot analysis (PE Biosystems; Foster City, CA, USA) primer
extension genotyping (11). We used 5 ( U .l of the SNaPshot™ Ready Reaction Mix (PE
Biosystems; Foster City, CA, USA) with approximately 0.15 pmoles of purified PCR
DNA template and 0.15 pmoles of the extension primer
(5'-AGCCTGATCCCGGCCTGGGCCTTCTACAGC) in a 10 |il final reaction
volume. This primer was extended in a TouchDown PCR machine (Hybaid;
Middlesex, UK) as follows: 10 sec at 96 °C, 5 sec at 50 °C, and 30 sec at 60 °C, for
25 cycles as reported (11). After primer extension, each reaction was treated with 1
unit of calf intestine alkaline phosphatase (CIAP) (Life Technologies; Rockville, MD,
USA) for 1 hour at 37 °C, followed by al5 min incubation at 72 °C, for CIAP
inactivation. All samples were run on an ABI PRISMr 377 automated DNA
sequencer (PE Biosystems; Foster City, CA, USA), following the recommendations
of the manufacturer, on a 5% Long Ranger™, 6M urea polyacrylamide gel (FMC
BioProducts; Rockland, ME, USA). After the run, samples were analyzed using
GeneScan™ 3.1 software (PE Biosystems; Foster City, CA, USA).
Statistical Analysis
Chi square tests were used to compare the genotypes among cases and
controls as reported previously (10). Odds ratios, 95% confidence intervals (Cl) and
p-values were determined and are presented as previously described (10).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Furthermore, all significance levels quoted are two-sided. Statistical analyses were
performed using the SAS program (Version 8.2; Cary, NC).
Results and Discussion
We have resequenced the exonic and adjacnt intronic regions of the
HSD17B3 gene and found the G289S missense substitution to be the most common
SNP. In addition we found two novel amino acid substitutions, one SNP in the 5’
UTR (untranslated region in exon 1), one silent substitution and nine intronic
substitutions (data not shown). We screened 103 Italian prostate cancer cases and 109
Italian “centenarian” disease-free men for the common G289S missense SNP in the
human testicular (or type III) 17P-hydroxysteroid dehydrogenase (HSD17B3) gene.
The distribution of G289S alleles in cancer subjects and unaffected “centenarians” is
shown in the table. The presence of at least one copy of the S allele (i.e. both hetero-
and homozygote men) was significantly higher in Italian prostate cancer than among
Italian centenarian normals (Table). Thus, a significantly increased risk for prostate
cancer was observed for individuals with at least one S allele (OR=2.5; CI=
1.03-6.07; Table). We assumed that there is no dose-response effect, so the
heterozygous (GS) and homozygous (SS) men were combined in the analysis to
achieve greater significance (Table). Finally, we note that common confounders such
as ethnicity and family history are probably not impacting our investigations reported
20
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
here, since all men are Caucasian, and in fact specifically of Italian ethnicity, and
they also have no family history of prostate cancer.
The G289S polymorphism confers a statistically significant increased risk for
prostate cancer in this study (Table). In short, we report here evidence for a plausible
contribution of the HSD17B3 locus to prostate cancer risk in Italy. Therefore, the
G289S SNP should be investigated in biochemical and kinetic detail to determine its
precise functional significance. In addition, molecular epidemiologic investigations
of this polymorphism should focus on racially and ethnically diverse populations to
confirm the risk associated with prostate cancer we report here (cf. Table). We also
note that previous investigations of the G289S SNP in polycystic ovary syndrome
failed to uncover any contribution to this particular disease (9). However, prostatic
phenotypes were apparently not investigated for the G289S polymorphism in that
study (9).
In summary, this is the first report investigating the testicular (or type HI)
17p-hydroxysteroid dehydrogenase (HSD17B3) locus as a candidate gene for
prostate cancer susceptibility. Our data suggest that the HSD17B3 gene may be a
plausible candidate gene for prostate cancer susceptibility. It, therefore, should be
systematically scanned to catalogue all SNPs, their functional significance should be
determined biochemically and ethnically diverse case/control (or case/cohort) studies
should be undertaken to establish the role of the HSD17B3 gene in prostate cancer
21
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
susceptibility. Finally, the HSD17B3 gene should be included in polygenic models of
prostate cancer etiology (12).
Table 2.1 Association of the G289S SNP in the HSD17B3 Gene With Prostate
Cancer.
Genotype Centenarian Cases OR 95% C l
Normals
GG (wt)
GS&SS
101
8 (8 + 0)
86
17(15 + 2) 2.5 1.03-6.07
The heterozygote (GS) and homozygote polymorphic (SS) genotypes were
combined in the analysis. GG (wt) is the normal genotype. OR is the odds ratio and
95 % Cl is the 95 % confidence interval.
Acknowledgements
This work was supported by grants from the NCI (CA83112 to JKVR) and the
Ministero della Sanita’ (to GN).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
References
1. Greenlee RT, Murray T, Bolden S (2001) Cancer Statistics, 2001, CA Cancer
J. Clin. 50, 7-33.
2. Schmid HP (1994) The Diagnostic and Therapeutic Window for Localized
Carcinoma of the Prostate, Ann. Urol. 28. 178-183.
3. Bernstein L, Ross RK (1991) Cancer in Los Angeles County: a Portrait of
Incidence and Mortality: University of Southern California Press, Los
Angeles, CA
4. Makridakis NM, Ross RK, Pike MC, Crocitto LE, Kolonel LN, Pearce CL,
Henderson BE, Reichardt JK. (1999) Association of Missense Substitution in
the SRD5A2 Gene with Prostate Cancer in African-American and Hispanic
Men in Los Angeles, USA, Lancet 354, 975-978.
5. Noble RL (1977) The Development of Prostate Adenocarcinoma in Nb Rats
Following Prolonged Sex Hormone Administration, Cancer Res. 37.
1929-1933.
6. Hsing AW. (2001) Hormones and Prostate Cancer: What's Next? Epidemiol.
Rev. 23, 42-58.
7. Cheng E, Lee C, Grayhack J (1993) Endocrinology of the Prostate, in: Lepor
H, Lawson RK,( eds.) Prostate Diseases, Saunders, Philadelphia, PA, 57-71.
8. Geissler WM, Davis DL, Wu L, Bradshaw KD, Patel S, Mendonca BB,
Elliston KO, Wilson JD, Russell DW, Andersson S (1994) Male
Pseudohermaphroditism Caused by Mutations of Testicular 17(3-
Hydroxysteroid Dehydrogenase 3, Nat. Genet. 7, 34-39
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9. Moghrabi N, Hughes IE, Dunaif A and Andersson S (1998) Deleterious
missense mutations and silent polymorphism in the human
17beta-hydroxysteroid dehydrogenase 3 gene (HSD17B3), J. Clin.
Endocrinol. Metab. 83, 2855-2860.
10. K Margiotti, F Sangiuolo, A De Luca, F Froio, CL Pearce, V Ricci-Barbini, F
Micali, M. Bonafe ,C Franceschi, B Dallapiccola, G Novelli, JKV Reichardt
(2000) Evidence for an Association Between the SRD5A2 (Type II Steroid
5a-Reductase) Locus and Prostate Cancer in Italian Patients, Dis. Markers16
147-150
11. Makridakis, NM and Reichardt, JKV (2001) MAPA: Multiplex Automated
Primer Extension Analysis: Simultaneous Genotyping of up to Four
Polymorphisms, Biotech. 3J_, 13 74-13 80
12. Ross RK, Pike MC, Coetzee GA, Reichardt JK, Yu MC, Feigelson H,
Stanczyk FZ, Kolonel LN, Henderson BE (1998) Androgen Metabolism and
Prostate Cancer: Establishing a Model of Genetic Susceptibility, Cancer Res.
58, 4497-4504.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter 3: Method & Materials
3.1 Multiethnic Cohort in Hawaii and Los Angeles (MEC)
This large cohort study, which began in 1993, consists of 215,251 men
and women in Hawaii and Southern California (Kolonel et al, 2000). Hawaii
and Southern California are ideal locations for collecting multiethnic samples
since there are unique histories of immigration in both areas. Selection of
subjects was based on drivers’ license files, which covered all socioeconomic
status, in both states. The voters’ registration files for Japanese-American
elderly women in Hawaii and the Health Care Financing Administration files
for African-Americans in California were also used. The ethnic identifier based
on last names was used in Los Angeles County for sampling in order to obtain
similar portions of the four ethnic groups of interest in the study population.
The ethnic distribution is 16.3% African-American, 22 % Latino, 22.9%
Caucasian, and 26.4% Japanese-American. Age was distributed from 45 to 75
at baseline. At the beginning of the study, all subjects in the study population
answered a self-administered, 26-page questionnaire, which includes
demographic data, dietary information, personal behaviors (smoking and
alcohol assumption), use of medications, personal history of medical conditions,
and family history of cancer. The cohort has been followed by the complex
25
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
tracking database in the Hawaii center. Based on this system, the follow-up
information about the cohort members including addresses, cancer incidences,
and death, has been updated in each location.
3.2 Study Subjects
The MEC is regularly linked to Los Angeles County cancer survey
program (SEER) Cancer Registry and State Cancer Registries in California and
Hawaii. Incident prostate cancer patients have been detected as cases since
1995.
A sub-cohort has been selected randomly among the cohort members,
who donated their blood samples to be used as controls. They were selected
before cases were identified. Blood samples have been also collected for cases.
All subjects provided written informed consent.
Totally 998 subjects were included (549 incident prostate cases and 449
controls). The racial/ethnic distribution is 213 (109/104) Caucasians, 239
(156/83) African Americans, 300 (158/142) Latin Americans, and 246 (126/120)
Japanese Americans (case/control). In this analysis, instead of age at entry into
the MEC, age at diagnosis for cases and age at blood collection for controls
w ere used, because cases and controls were included in the study at those time
points. Information about education, occupation, family history of prostate
26
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
cancer, medical history of enlarged prostate, medical history of prostate surgery,
medical history of vasectomy, body mass index (BMI), and BMI at age 21 were
also obtained. Severity of disease was included only for cases.
3.3 Experimental Methods
We genotyped three single nucleotide polymorphisms (SNPs) in three
different coding regions of the HSD17B3 gene: V31I (Valine at codon 31
replaced by Isoluecine) on exon 1,1102F (Isoluecine at codon 102 replaced by
Phenylalanine) on exon 4, and G289S (Glycine at codon 289 replaced by
Serine) on exon 11. G289S is a known polymorphism (Moghrabi et al., 1998),
and V31I and I102F are new polymorphisms found in our lab by Katia
Margiotti. Two different methods were used for obtaining genotype data:
multiplex automated primer extension analysis (MAPA) and Restriction
Fragment Length Polymorphism (RFLP).
To avoid observational bias during screening genotypes, I was blinded by
the disease status of samples whilst screening genotypes.
3.3.1 Multiplex PCR
Approximately 100 to 200 ng of genomic DNA, extracted from white
cell of blood samples, was used in a polymerase chain reaction (PCR) to
27
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
amplify exon 1 and exon 4 of the HSD17B3 gene at the same time. For this
multiplex PCR, the following two pairs of primers were used (synthesized by
Invitrogen);
Exon 1 5’ ACGGCCAGGGCTGAAACAGTCTGTT 3' (Forward)
5’ CATGAGATGGAACACTCCCT 3' (Reverse)
Exon 4 5’ TGGATCCCTGTTCATTAAAAAAACT 3' (Forward)
5’ GATGTATGACAACAAGCTTTGCATC 3' (Reverse)
The PCR was performed in a MJ PTC-200 Thermal Cyclers, using 25 pmol
of each primer, IX PCR buffer, 1.5 mM MgCt, and 1.75 unit/reaction of Taq
DNA polymerase and 50 pM of each deoxynucleotide triphosphate (all from
Invitrogen). The best condition for the amplification was an initial
denaturation for 3 min at 95°C, followed by 30 cycles of denaturation for 30
sec at 95°C, annealing for 30 sec at 58°C and extension for 30 sec at 72°C.
The additional extension was performed for 7 min at 72°C as the last step.
Multiplex PCR of Exon 1 and Exon 4 was detected on 1.8% agarose,
IX TAE (Tris-Acetate-EDTA buffer) gels as two bands with a 40bp
difference. The two bands were excised together and DNA was extracted
from the gel using the Gel Extraction Kit (Qiagen).
2 8
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3.3.2 Multiplex Automated Primer Extension Analysis (MAPA)
These purified samples, in which unincorporated PCR primers and
dNTPs were removed, were genotyped using the ABI PRISM® SNaPshot™
ddNTP primer extension kit (PE Biosystems) on an ABI prism 377 automated
Sequencer (Makridakis & Reichardt, 2001). 1.5 pi of the purified samples
were mixed with 0.15 pmol of the proper primer(s), designed to end one base
before the known point mutation(s), and 5pl of SNaPshot Ready Reaction
Premix containing dideoxynucleotide triphosphates (ddNTPs). In order to
detect variants of the V3II and the I102F in a single reaction, two primers
with different lengths were used (synthesized by Invitrogen):
V31I: 5' CGAAGTGCGTGAGATTCTCCAGATGT-G (wild type) or
A (mutation) 3’ (26 bases)
I102F: 5' ACAGGGAGGAGTGTGAAG-A (wild type) or T
(mutation) 3’ (18 bases)
The thermal cycling profile for the reaction was 25 cycles of denaturation for
10 sec at 95°C, annealing for 5 sec at 50°C, and extension for 30 sec at 60°C.
To remove unincorporated ddNTPs, calf intestinal phosphatase (CIP: New
England BioLabs) was added to the reaction tube and the samples were
incubated at 37°C for 1 hour followed by deactivation at 72°C for 15 min. For
loading on a gel, 3pl of the samples was mixed with 3pl of loading dye,
29
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
consisting of 5 unit deionized formamide and 1 unit of loading buffer
(Perkin-Elmer).
The samples were loaded on 5% Long Ranger™ polyacrylamide, IX
TBE (Tris-Borate-EDTA buffer) gel with 6M urea (BMA). For
polymerization, the gel should be prepared with TEMED
(Tetramethylethylenediamine) and ammonium persulfate two hour before
loading. The samples were incubated at 95°C for 5 min to denature two
strands of DNA and stored in ice to maintain the single stranded
conformation. 2pl of the samples were loaded on the gel and electrophoresed
for 40 min in an ABI Prism 377 automated sequencer. After the run, samples
were analyzed using GeneScan® 3.1 software (Applied Biosystems). The
genotypes were shown as colors of fluorescent peaks on the
electropherograms. 10% of the samples, which were randomly selected, were
regenotyped to check the consistency.
3.3.3. Automated DNA Sequencing
All heterozygous mutated samples were regenotyped by automated
DNA sequencing to assess the consistency of the data. PCR products (the
same primers were used in 2.3.2) were purified using an agarose gel
extraction kit (Qiagen). The purified products were analyzed y
30
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
electrophoresis on a 2% agarose gel to ensure the quality of the template.
Sequencing reactions were carried out in the total 20pl volume containing 3.2
pmol of primers (the same one in 2.3.2), 4pl of the PCR product, 4pl of ABI
Prism Terminator Big Dye Cycle Sequencing Ready Reaction Mix and 7pl of
ddKhO. The reaction was 35 cycles of denaturation for 10 sec at 95°C,
annealing for 5 sec at 50°C, and extension for 4 min at 60°C. The solution
was purified with Autoseq G-50 columns (Pharmacia Biotech). The solution
was then dried by placing inside a DNA Speed Vac for 20 minutes. The pellet
is then re-suspended with 5pl of Formamide and lpl of loading buffer
(Perkin-Elmer). 1.5pl of the suspension is loaded onto a 5% long ranger gel.
Electrophoresis is performed with the ABI Prism 377 DNA sequencer for 2
hours. Nucleotide sequences were analyzed with Sequence Navigator
software (Applied Biosystem).
3.3.4. Restriction Fragment Length Polymorphism (RFLP)
For genotyping of the G289S polymorphism (G to A replacement in
the first base at codon 289), the DNA was digested with a restriction enzyme
and the DNA fragments were separated according to molecular size using gel
electrophoresis. Exon 11, including this SNP, was amplified by the following
primers (synthesized by Invitrogen):
31
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Exon 11 5' GAGGAAAAAGGTTGTGCTGGACTCCT 3' (Forward)
5’ GATGAACTGAGGTACTTGTTATTCC 3' (Reverse)
The amplification condition was 30 cycles of 95°C for 30 sec, 59°C
for 30 sec, and 72°C for 30 sec.
The PCR products were incubated with 10 units of the Ban I
restriction enzyme (New England BioLabs) at 37°C for 16 hours and then ran
on the 1.8% agarose, IX TAE gels to determine the genotypes. The Ban I
recognizes g/gyrcc sequence (y is pyrimidine and r is purine). Thus, a sample
with G allele, having the restriction site, is cut into two 103 base pair
fragments and, in contrast, one with A allele, which the Ban I enzyme cannot
recognize, is a 206 base pair, uncleaved fragment. Since one allele has the
recognition site and the other does not, the heterozygous mutation shows two
bands, 103 and 206 base pairs.
To test the reproducibility of genotyping, 10% of the samples were
randomly selected and regenotyped.
3.4 Statistical Analysis
3.4.1 Case-Cohort Design
A case-cohort study and a nested case-control study are commonly
used as more economical methods to analyze cohort data. While controls are
32
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
recruited from cohort members by matching with cases in the nested
case-control study, controls in the case-cohort study are from a sub-cohort
who has been randomly selected in the cohort before identifying cases
(Prentice, 1986). In a case-cohort study, all cases are included regardless of
sub-cohort status, but controls are included only if they are in the sub-cohort
(Prentice, 1986). The subjects can be divided to three sub-groups: cases not
from the subcohort but from the MEC, cases from the subcohort, and the
subcohort controls. The cases, who are from the sub-cohort, should serve
both as a case and a control in a case-cohort analysis, and this makes data
analysis for a case-cohort study difficult (Barlow et al, 1999).
Generally, a pseudo-likelihood Cox regression model is used for a
case-cohort study (Prentice, 1986). This is a weighted Cox regression model,
and there are three different weighting schemes to handle the three
sub-groups differently (Prentice, 1986; Self & Prentice, 1988; Barlow, 1994).
Estimations of parameters in the relative risk function involve the comparison
of the covariates for cases to those for all other subjects who constitute a
“risk set.” The risk set is defined as a group of subjects at risk at each
timepoint that someone actually did fail. A case outside the sub-cohort is not
included in this risk set until just before failure. If the case is included in the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
earlier risk set, this leads to the bias because knowledge about the future case
status of that individual is used improperly (Barlow et al., 1999).
3.4.2 Data Analysis
Controls in this study were selected from the MEC subjects who
donated their blood samples, which were collected before the cases were
identified. This sampling method shows that it is a case-cohort study.
Even though this is a case-cohort study, we used an alternative way to
analyze data in this study, which is a case-control analysis considered as the
MEC standard. That is, if we assume that the MEC is one large population,
this case-cohort study is considered as a case-control study. That is, subjects
with newly diagnosed prostate cancer are incident cases regardless of
subcohort status and subcohort members except the subcohort cases are
controls. Because it is assumed that the cases are included in the study when
they have diagnosis while the controls are included when their blood samples
are collected, age variables are differently obtained between cases and
controls. Dr. Pearce compared the results from each analysis for a case-cohort
design and a case-control design within the MEC and showed they are very
similar (Pearce CL, Department of Preventive Medicine, USC; personal
34
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
communication). Thus, we used a standard case-control analysis, which is
much simpler that the case-cohort analysis, for this study.
Unconditional logistic regression was used to model the association
between risk of prostate cancer and SNPs in the HSD17B3 gene. All analyses
were adjusted for age at diagnosis for cases/at blood collection for controls
and for ethnicity. To assess if V31I, I102F, and G289S genotypes are
associated with prostate cancer risk, the adjusted odds ratios (OR)/ 95%
confidence interval (Cl) were obtained. Also, the age-adjusted ORs were
obtained within each ethnicity and compared to examine whether there are
the different risks among the four ethnic groups. Finally, the study population
was checked in accordance with Hardy-Weinberg equilibrium by Pearson’s
chi-square goodness-of-fit statistic. All presented Ps are two-sided.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter 4: Results
4.1 Results from the Experiments
Of the total of 998 subjects, genotype data for 17 subjects for V31I /
I102F and 136 subjects for G289S could not be obtained, either because the
DNA sample was depleted or because of PCR amplification failure. 98.72%
of cases and 97.77% of controls and 87.43% of cases and 85.08% of
subcohort controls were screened for V31I /I102F and G289S, respectively.
In this study, multiplex automated primer extension analysis (MAPA),
which is developed, based on a commercially available protocol (SNaPshot),
in our lab (Makridakis & Reichardt, 2001), and restriction fragment length
polymorphism (RFLP) were used to genotype DNA samples.
A miltiplex PCR produced two different parts of the HSD17B3 gene
(exon 1 and exon 4) for MAPA. It was shown on the agarose gel as two bands
with 40 bp differences as shown in figure 4.1.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
After gel extraction purification to reduce the amount of unused
primers and dNTPs, the multiplex PCR templates were used in MAPA. The
genotyping results of V31I and I102F were marked together as different
colors of peaks on the electropherograms (Figure 4.2).
Marker Blank Arnplicon containing exonl and exon 4
800 bp
350 bp
250 bp
200 bp
150 bp
50 bp
Figure 4.1 Mutliplex PCR. Exonl and exon 4 were amplified together in a single
reaction. The two bands are 40 base pair different.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 4.2 Results of MAPA.
a. Homozygous wild type both in V31I and I02F (a blue peak
and a green peak, respectively).
b. Heterozygous mutation only in V31I due to G to A
substitution (a blue peak and a green peak were shown
together).
c. Heterozygous mutation only in I102F due to A to T
substitution (a red peak and a green peak were shown together).
Since MAPA is relatively new, reliability of this method was
compared with automated sequencing (Figure 4.3). All heterozygous mutated
samples in exon 1 and exon 4 were regenotyped by sequencing to assess
consistency of the data: this resulted in 100% concordance.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
b
m . MC T T C A '
1 2 0
: a
% A A K \.C A T C
Figure 4.3 Sequencing Results.
a. Heterozygous mutation of V31I in exon 1: C (a blue peak) to
T (a red peak) substitution with the reverse primer
b. Heterozygous mutation of I102F in exon 2: A (a green peak)
to T (a red peak)
The genotypes of G289S were screened by RFLP, which is a
well-known and accurate method. The result was shown in figure 4.4.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 4.4 Results of RFLP with the Ban I Restriction Enzyme.
A. Undigested.
B. Homozygous mutation with one band (206 bp).
C. Heterozygous mutation with two bands
(103 bp and 206 bp).
D. Wild type with one band (103 bp).
E. Blank
To assess the reproducibility of the genotype data in each method,
10% of random samples were regenotyped; all genotypes matched the initial
genotypes.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4.2 Results from Data Analysis
Demographic characteristics between cases and controls are displayed
in Table 4.1 and 4.2. Cases were not significantly different from controls in
education, medical history of vasectomy, BMI, and BMI at age 21. However,
cases were more likely to have a family history of prostate cancer (p=0.04)
and a medical history of enlarged prostate (p=0.005) than controls. Cases
were statistically significantly older when they were diagnosed for prostate
cancer compared to controls at blood collection (p < 0.001).
Table 4.3 shows the results of the genotyping data and the odds ratios.
The II genotype in V31I and the FF genotype in I102F were not found in our
study. However, the SS genotype in G289S was found in the Japanese group
(thirteen subjects: four cases and nine controls). Since the frequency of the
SS genotype was rare and we assumed that there is no dose-response effect,
the GS genotype and the SS genotype were combined for the analysis.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 4.1 Demographic characteristics between cases and controls.
No. of
cases
(549)
(%)
No. of
subcohort
controls
(449)
(%)
Education
<10th 94 (17.34) 65 (14.64)
11 - 1 2 th 137 (25.28) 91 (20.50)
Some 152 (28.04) 126 (28.38)
college/V ocational 159 (29.34) 162 (36.49)
College grad or higher
Family history
No family history 420 (85.71) 378 (90.21)
1st degree relative 70 (14.29) 41 (9.79)
Enlarged prostate
Yes 114 (20.77) 63 (14.03)
No 435 (79.23) 386 (85.97)
Vasectomy
Yes 90 (16.39) 67 (14.92)
No 459 (83.61) 382 (85.08)
Severity of prostate
cancer
(only for cases)
Localized
Advanced
201
324
(38.28)
(61.72)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 4.2 Mean and Standard Deviation of Age, BMI, BMI at Age 21.
Variable N Mean (Standard Deviation)
Age
Case 547 66.87 (Max: 80, Min:
46)
(6.40)
Control 449 62.35 (Max: 80, Min:
46)
(8.32)
BMI
Case 545 26.43 (3.74)
Control 446 26.50 (3.94)
BMI at age 21
Case 530 22.26 (3.19)
Control 426 22.44 (2.87)
We did not find any significant association between variants of V3II
and G289S and prostate cancer risk in all the ethnic groups combined (Table
4.3), after controlling for age and ethnicity. The odds ratio of having the VI
genotype in V31I is 1.10 (95% Cl (confidence intervals) =0.54-2.22)
compared to the W genotype, and the odds ratio of having the GS+SS
genotype in G289S is 1.31 (95% 0=0.89-1.96), compared to the GG
genotype. These results were unchanged when we included only cases with
advanced disease (the VI vs. the W genotype in V3II: OR=1.10, 95%
0=0.51-2.39 & the GS+SS vs. the GG genotype in G289S: OR=1.14, 95%
0=0.78-1.67).
43
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 4.3 Distribution of the V31I, I102F, and G289S Genotypes and ORs
for the association between genotypes and prostate cancer risk.
SNPs Case Control Total
OR**
(95% Cl)
P
W 522 422 944 1.10
V31I 0.79
VI 20 17 37 (0.54-2. 25)
II 541 438 979
I102F***
IF 1 1 2
GG 394 327 721
1.31
G289S GS* 82 46 128 0.19
(0.89- 1.96)
SS* 4 9 13
*GS and SS were combined for the odds ratio.
** Adjusted by age at diagnosis for cases/blood collection for controls
and ethnic groups.
*** Not analyzed since the frequency of the variant is too low.
Table 4.4 and 4.5 show the results from each ethnic group. We found
a statistically significant association between the GS+SS genotype in G289S
and prostate cancer risk in Japanese men (OR=1.97, 95% CI=1.05-3.71).
However, there is no consistent result across the other ethnic groups: the
GS+SS genotype in G289S is a risk factor for Caucasian men while a
protective factor for African-American men, even though the findings were
not significant and the number of the variants was too small.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 4.4 ORs for the association between the V31I genotype and
prostate cancer risk by racial/ethnic group.
Genotype
W VI OR*
Ethnicity Case Control Case Control (95% Cl)
P
White 108 100 0 1
Black 148 80 7 0
Latino 154 137 1 2
0.27
(0.02 - 3.00)
0.28
Japanese 112 106 12 14
0.93
(0.38-2.29)
0.87
*Adjusted by age at diagnosis for cases/ blood collection for controls
Table 4.5 ORs for the association between the G289S variants and
prostate cancer risk by racial/ethnic group.
Genotype
GG GS+SS OR*
Ethnicity Case Control Case Control (95% Cl)
P
White 83 81 12 5
1.61
(0.52-4.96)
0.41
Black 120 56 14 8
0.75
(0.29- 1.92)
0.55
Latino 126 120 12 10
0.97
(0.39-2.41)
0.95
Japanese 65 70 48 32
1.97
(1.05-3.71)
0.04
*Adjusted by age at diagnosis for cases/ blood collection for controls
45
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Since the frequencies of the F allele in I102F are too low (only two
subjects with F allele are found), I102F was not further analyzed.
4.3 Results from Hardy-Weinberg Law
Only the subcohort controls were used to assess if the study
population is in Hardy-Weinberg Equilibrium (HWE). Table 4.5 showed the
distribution of the genotyping data by ethnicity within only controls.
Table 4.6 Distribution of SNPs by Ethnicity within the Controls.
SNPs White Black Latino Japanese Total
W 100 80 136 106 422
V31I 439
VI 1 0 2 14 17
IT 101 80 1387 120 438
I102F 439
IF 0 0 1 0 1
GG 81 56 120 70 327
G289S GS 5 8 10 23 46 382
SS 0 0 0 9 9
Expected numbers and Chi-square tests for HWE were obtained for
V31I and G289S:
For V31I,
pA = # o f V alleles / total # o f alleles = (2(422) + 17} / 2(439) *
0.9805
46
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
qA = # of I alleles / total # of alleles = {17 + 2(0)} / 2(439) = 0.0194
E (W ) = npA 2 = 439 (0.9806)2 = 422.13
E (VI) = n: 2- pA - qA = 439- 2- 0.9806- 0.0194= 16.70
E (II) = nqA 2 = 439 (0.0194)2 = 0.17
X 2 dM = {(422 - 422.13)2 / 423.13} + {(17 - 16.70)2 / 16.70} + {(0
0.17)2 / 0.17}= 0.1754 (p = 0.68)
Thus, the population is in Hardy-Weinberg Equilibrium.
For G289S,
pA = # of V alleles / total # of alleles = 2(327) + 46 / 2(382) = 0.9162
qA = # of I alleles / total # of alleles = 46 + 2(9) / 2(382) = 0.0838
E (W ) = npA 2 = 382 (0.9162)2 = 320.66
E (VI) = n 2- pA - qA = 382- 2- 0.9162- 0.0838 = 58.66
E (II) = n qA 2 = 430 (0.0838)2 = 3.02
X 2dfH = {(327 - 320.66)2 / 320.66} + {(46 - 58.66)2 / 58.66} + {(9
3.02)2 / 3.02} = 14.6989 (p = 0.0001)
Thus, the population is not in Hardy-Weinberg Equilibrium.
Since we detected that the population is not in HWE in G289S, we
compared the observed frequencies of G289S with expected numbers in each
ethnic group (Table 4.6).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 4.7 Observed vs. Expected Numbers of the G289S by Ethnicity.
G289S
White Black Latino Japanese
Obs Exp Obs Exp Obs Exp Obs Exp
GG 81 83.51 56 56.31 120 120.31 70 65.12
GS 5 4.84 8 7.44 10 9.51 23 32.76
SS 0 0.07 0 0.25 0 0.19 9 4.12
Japanese men had less of the GG and GS genotype but much more of
the SS genotype compared to the expected. Only the Japanese group is not in
HWE (x2df=i=9.25, p=0.002). This result might imply that there are
evolutionary changes in the Japanese-American group due to founder effect
and/or genetic drift. Also, in most cases, this implies that there is a
genotyping error, but in our case we mled out this possibility because only
one group among the four groups is not in HWE. If there is a genotyping
error, we should find other groups are not in HWE either. Another
explanation for this is a sampling error in control selection. Since study
subjects can choose multiple ethnicities in the questionnaire, it is possible
that some subjects who are not pure Japanese are included in the Japanese
group.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter 5: Conclusion
5.1 Association between SNPs in the HSD17B3 Gene and Prostate Cancer
Risk
The results presented in this study did not agree with our preliminary
data suggesting an association between variants in the HSD17B3 gene and
prostate cancer risk (Margiotti, 2002). Although the biological rationale for
our hypothesis - association between androgens and prostate cancer- is
plausible, we could not find an association of amino acid substitutions in the
HSD17B3 gene with prostate cancer in all the ethnic groups combined or in
and individual ethnic group, expect the Japanese group. There was a
significant increasing risk of prostate cancer with the GS+SS genotype in
G289S among Japanese-American men. However, a small, nonsignificant
protective association between the GS+SS genotype in G289S and prostate
cancer was found in the African-American and Latino group. While these
results indicate no strong or consistent pattern across the racial/ethnic groups,
these might suggest that effects (or degree of effects) of the variants are
different among the ethnic groups.
The following sections are issues to consider when we interpreted the
results in our study.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5.1.1 Study Population and Subjects
Like prostate cancer, the reported incidence rates of several common
cancers have striking differences among the four ethnic groups in the US
(Miller et al., 1996). It is not known yet which factors explain interethnic
differences in the incidence of prostate cancer. It is one of our strengths in our
study that the MEC has a great potential to perform interethnic comparisons
of exposure-disease relations since it includes a variety of ethnic groups
within a single study and uses a common data collection methodology in all
groups, (Kolonel et al., 2000). These interethnic comparisons will help us to
understand etiology of prostate cancer even though we did not find a clue in
our study.
The probability that we have some undetected cases in this study is
quite high, since prostate cancer is very common in older men and
approximately 65% of men in their sixties have evidence of prostatic
adenocarcinoma at autopsy (Sakr, 1994). These contaminated controls might
be a problem in our study. We cannot predict which direction the bias goes to,
because the frequencies of the variants among the contaminated controls are
unknown. However, if the variants are truly related to prostate cancer risk,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
the bias goes towards to the null, and this might explain partially the null
findings in our study.
5.1.2 Study Design and Power
A case-cohort design is economical for studying a large cohort
because covariate information needs to be assembled only on cases and
selected controls, which is only a fraction of those of the full cohort (Prentice,
1986; Wacholder et al., 1989; Langholz & Thomas, 1990). Since the controls
are randomly selected from the cohort, selection of this subcohort is
independent of disease status of themselves and characteristics of the cases
and, therefore, is relatively fast (Barlow et al., 1999; Wacholder, 1991). Also,
the subcohort is independant of the cases, so the subcohort can be used for
several outcomes.
Calculation of the power of a case-cohort design is relatively
complicated because the power depends on the size of the overlap in the sets
of subjects at risk at each pair of event times (Wacholder, 1991). For exact
power calculation in this study, therefore, we need to know all information
about the full cohort and this is beyond our study. However, we can calculate
an approximate estimation of the samples size. Table 5.1 shows required
case-control sets with each allele frequency and relative risk.
51
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 5.1 Expected case-control sets calculated from relative risk and
allele frequency (80% power and 0.05 significance; Quanto).
Allele
frequency 1.4
Relative Risk
1.5 2.0
0.05 1384 929 288
0.1 819 555 179
0.2 582 399 136
With 80% power and 0.05 level of significance, we need 582 cases
and 582 controls to detect an OR of 1.4 when allele frequency is 0.2.
However, the allele frequencies of the variants in our study are less than 0.1.
Thus, we need over 1500 samples to detect an OR of 1.4. This power
calculation shows that we do not have enough power to detect the true
association between SNPs in the HSD17B3 gene and prostate cancer risk,
when we assumed the effects caused by these variants are small.
It is hard to achieve enough power to detect the true association in our
study because allele frequencies of SNPs in the HSD17B3 were relatively
rare. However, allele frequencies in the Japanese group were different from
the other three ethnic groups: for V31I, the frequency of the I allele was less
52
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
than 0.002 except for Japanese men (0.05), and the frequency of the S allele
was approximately 0.05 in the rest of the groups while it is over 0.21 in the
Japanese groups. Furthermore, our on-going study, in which we screen the
same SNPs in Chinese men in China, also supports that the allele frequencies
of the variants in V31I and G289S are higher among Asians. The frequencies
are even little bit higher in Chinese men than Japanese-American men
(approximately 0.07 for the I allele in V3II and 0.25 for the S allele in
G289S). Thus, Asians are probably a better population to study the
association between SNPs in the HSD17B3 gene and prostate cancer risk,
compared to other ethnic groups.
5.1.3 Experimental Methods
In this study, two different methods were used for determining the
genotypes. While MAPA is relative new, RFLP has been a common method
for conducting genetic studies for the last decades. RFLP used to be laborious
because Southern blots with radio-labeled probes were used, but nowadays
this is not a problem because RFLPs can usually be typed by PCR (Strachan
& Read, 1999). MAPA is a modified technique of a commercially available
method in order to screen several nucleotide substitutions at the same time
(Makridakis & Reichardt, 2001). This relatively new method is not only less
53
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
laborious and less expensive but also much faster than sequencing. However,
while DNA sequencing and RFLP have been known for their accuracy (Eng
& Vijg, 1997; Persons & Heflich, 1997), the reliability of MAPA was not
completely established. Here, we assessed reliability and reproducibility of
MAPA by comparing the results with DNA sequencing and regenotyping.
The results clearly indicated that MAPA has reliability.
Since both techniques for screening genotype in our study are reliable
and reproducible, we ruled out the fact that the experimental errors are related
to our null findings.
5.1.4 Hardy-Weinberg Equilibrium
Historical demonstration proved that Mendelian inheritance does
allow variation to be maintained (Hardy, 1908; Weinberg, 1908). There are
several assumptions for the Hardy-Weinberg law: random mating with
respect to genotype, no selection, no mutation, no migration, discrete
generations (without inbreeding), and infinite population size (Hardy, 1908;
Weinberg, 1908). If there is violation to one of these assumptions, the
population is not in Hardy-Weinberg Equilibrium. Our study population is
not in Hardy-Weinberg Equilibrium, due to the Japanese-American group.
This is because an allele may increase or decrease in frequency simply
54
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
through chance due to genetic drift or founder effects and/or because there is
a sampling error in control selection. In either case, the result from the
Japanese group, therefore, is not reliable since we are not sure these allele
frequencies to remain constant over time (barring any specific evolutionary
forces acting upon this locus).
5.1.5 Inconsistent Results between Preliminary Data and this study
The results from our study, except one from the Japanese group, are
not consistent with the preliminary data from the Italian samples. The results
from these two groups, the Italian and Japanese groups, show that prostate
cancer risk might increase with the GS+SS genotype in G289S. However,
there are problems in these samples. First, since this study was not designed
by epidemiologists, it was not constructed properly for a epidemiologic study.
That is, the controls, “centenarian normals,” are not from the same base
population as the cases, who are hospital-based, in the Italian samples.
Second, the Japanese group is not in the HWE, so there might be
evolutionary process on this locus.
Due to these problems, the statistically significant association
between the GS+SS genotype in G289S and prostate cancer risk as well as
inconsistency between the two studies is questionable.
55
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
However, we can observe some consistency between Caucasians in
Italy and in the US; the GS+SS genotype increases risk of prostate cancer.
Also, a recent study shows that expression levels of the 17(3-HSD type III
gene are much higher in prostate cancer patients than in people without the
cancer. These imply that variants of this gene might be related to prostate
cancer risk (Koh et al, 2002). It may be worth to assess if these variants in the
HSD17B gene affects the enzymatic kinetics and susceptibility to prostate
cancer. However, it dose not seem that they are strong predictors of prostate
cancer.
5.2 Future Direction
It is not known if these amino acid substitutions affect enzyme activity.
Alterations of enzyme kinetics might lead to a different risk of prostate cancer
in individuals. For example, the A49T (alanine at codon 49 converted to
threonine) mutation in the SRD5A2 gene increases steroid 5-a reductase type
II and, therefore, may increases prostate cancer risk (Makridakis et al., 1999;
Jaffee et al., 2000; Makridakis, 2000). Thus, it is worth to assess the effect of
variants in these SNPs on enzyme function.
For this, in vitro expression analysis of the altered proteins will need to
be performed. The SNPs will be constructed in the HSD17B3 cDNA via
56
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
mutagenesis using standard techniques and transfected into mammalian cells.
Then enzyme activity will be tested with standard biochemical assay to assess
Michaelis-Menten constants. Measurements such as optimal pH, K m, Kj, and
V m ax will be made.
The M - length cDNA insert corresponding to the HSD17B3 gene can
be ligated within a vector to produce a recombinant plasmid. The accuracy and
efficiency of the ligation reaction will then be evaluated with restriction digest
analysis. In order to introduce the desired nucleotide substitutions, site-directed
mutagenesis can be performed utilizing the PCR fragment obtained from the
genomic DNA bearing the V31I and/or G289S substitutions, or site-directed
mutagenesis with overlap extension PCR with primers that contain die
nucleotide variants. The DNA sequences of the newly introduced Augments in
the recombinant plasmids will then be confirmed with direct sequencing.
Once the sequence is confirmed, the mutant plasmids will be
electroporated into mammalian cells with mock (blank), wild type (cDNA), and
mutant (mutant cDNA), and then grown. (3-galactosidase control plasmids will
also be co-electroporated to monitor transfection efficiency. Forty-eight hours
after transfection, 17|3-hydroxysteroid dehydorgenase activities will be
measured with the amount of conversion of [1 4 C]-labeled androstenedione to
testosterone with NADPH as a cofactor. Kinetic constants such as pH optima,
57
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Vm ax, Km , and K| will be measured for both wild type and mutant enzymes to
compare the difference.
5.3 Conclusion
We do not find a significant association between the amino acid
substitutions in the HSD17B3 gene and prostate cancer risk in all the ethnic
groups combined. However, there is a significant association between the
GS+SS genotype in G289S and the disease among Japanese-American men.
Also, there might be different effects of this amino acid substitution on
different ethnic groups. We do not know yet if the variants in the gene affect
the activity of human 17(3 - hydroxy steroid dehydrogenase type III. However, it
is possible that one polymorphism increases activity and the other decreases it,
for examples, A49T and V89L (Valine at codon 89 converted to Lysine) in the
SRD5A2 gene (Makridakis et al., 2000). If this occurs, the effects are canceled
out by each other. A future study on enzyme kinetics, therefore, might give us
some idea about understanding of the role of genetic variation in the risk of
prostate cancer.
Moreover, it has been reported that common polymorphisms of several
susceptibility genes, such as the AR, SRD5A2, and CYP17 genes, involved in
androgen metabolism are linked to prostate cancer (Reichardt et al., 1995;
58
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Hsing et al., 2000; Wadelius et al., 1999). Since prostate cancer is a
multifactorial disease caused by multiple environmental and genetic factors, it
is possible that the several mutations in the androgen metabolic genes,
including the variants in the HSD17B3 gene, alter androgsn levels together and,
therefore, increase/decrease prostate cancer risk. Future studies need to
examine these combined effects on prostate cancer.
In general, it seems that the HSD17B3 variants are not good biomarkers
of susceptibility to prostate cancer. However, our finding in the
Japanese-American group suggests that the effects of these variants in the
HSD17B3 gene might be important in Asian population. To our knowledge,
there has been no large population-based epidemiologic study for the
association between SNPs in the HSD17B3 gene and prostate cancer risk. Thus,
our study is the first step for assessing this association.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
References:
American Cancer Society, Cancer facts and Figures. 2001, Atlanta, GA
(www.cancer.org).
Anderson S, Geissler WM, Wu L, et al. Molecular genetics and pathophysiology of
17P-hydroxysteroid dehydrogenase 3 deficiency. J Clin Endocrinol Metab 1996; 81:
130-36.
Andersson SO, Wolk A, Bergstrom R et al. Energy, nutrient intake an prosate cancer
risk: a population based case-control study in Sweden. Int J Cancer 1996; 68:
716-22.
Barlow WE. Robust variance estimation for the case-cohort design. Biometrics 1994;
50: 1064-1072.
Barlow WE, Ichikawa L, Rosner D, Izumi S. Analysis of case-cohort designs. J Clin
Epidemiol 1999; 52(12): 1165-1172.
Bernstein L, Ross RK. Cancer in Los Angeles County: a Portrait of Incidence and
Mortality. University of Southern California Press, 1991.
Cancer Statistics, 2001. American Cancer Society website (www.cancer.org).
Cerhan JR, Parkr AS, Putnam SA, et al. Familial history and prostate cancer in a
population-based cohort of Iowa men. Cancer Epidemiol Biomarkers Prev 1999; 8:
53-60.
Eng C, Vijg J. Genetic testing: the problems and the promise. Nat Biotechnol 1997;
17:365-70.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Febbo PG, Kantoff PW, Platz EA, et al. The Y89L polymorphism in the
5a-reductase type 2 gene and risk of prostate cancer: a case-control analysis.
Cancer Res 1999; 59: 5878-81.
Feldman BJ, Feldman D. The development of androgen-independent prostate cancer.
Nat Rev Cancer 2001 Oct; 1(1): 34-45.
Prostate IN: Fleming ID, Cooper JS, Henson DE, et al, eds. AJCC cancer staging
Manual. 5th ed. Philadelphia, PA: Lippincott-Raven, 1997: 219-24.
Gann PH, Hennekens CH, Longcope C, et al. A Prospective Study of Plasma
Hormone Levels, Nonhormonal Factors, Development of Benign Prostatic
Hyperplasia. Prostate 1995; 26(1): 40-49.
Geissler WM, Davis DL, Wu L, et al. Male Pseudohermaphroditism Caused by
Mutations of Testicular 17(3 -Hydroxysteroid Dehydrogenase 3. Nature Gen 1994; 7:
34-39.
Gleason DF. Histologic grading and staging of prostate carcinoma. In: Tannenbaum
M, ed. Urologic pathology: the prostate. Philadelphia, PA: Lea and Febiger, 1997:
171-97.
Haiman CA, Stampfer MJ, Giovannucci E, et al. The Relationship between a
Polymorphism in CYP17 with Plasma Hormone Levels and Prostate Cancer, 2001.
Cancer Epidemiol Biomarkers Prev. 2001; 10: 743-748.
Hardy GH. Mendelian proportions in a mixed population. Science 1908; 28: 49-50.
Henderson BE, Ross RK, Pike MC, Casagrande JT. Endogenous Hormones as a
Major Factor in Human Cancer. Cancer Res 1982; 42: 3232-39.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Honda GD, Bernstein L, Ross RK, et al. Vasectomy, Cigarette Smoking, and Age at
First Sexual Intercourse as Risk Factors for Prostate Cancer in Middle-Aged Men. Br
J Cancer 1988; 57:326-331.
Hsing AW, Gao YT, Wu G, et al. Polymorphic CAG and GGN repeat lengths in the
androgen receptor gene and prostate cancer risk: a population-based case-control
study in China. Cancer Res 2000; 60: 5111 - 6.
Huggins C, Hodges CV. Studies on Prostatic Cancer. Cancer Res 1941; 1: 293-97.
Jaffe JM, Malkowicz SB, Walker AH, et al. Association of SRD5A2 genotype and
pathological characteristics of prostate tumors. Cancer Res 2000; 60: 1626-30.
Jemal A, Thomas A, Murray T, Thun M. Cancer Statistics, 2002. CA Cancer J Clin
2002;52:23-47.
Kantoff PW, Febbo PG, Giovannucci E, et al. A polymorphism of the 5a-reductase
gene and its association with prostate cancer: a case-control analysis. Cancer
Epidemiol Biomarkers Prev 1997; 6: 189-92.
Koh E, Noda T, Kanaya J, Namiki M. Differential expression of 17(3 -Hydroxysteroid
Dehydrogenase isozyme genes in prostate cancer and noncancer tissues. The Prostate
2002; 53: 154-9.
Kolonel LN, Henderson BE, Hankin JH, et al. A Multiethnic Cohort in Hawaii and
Los Angeles: Baseline Characteristics. Am J of Epidemiol 2000; 151: 346-357.
Langholz B, Thomas DC. Nested case-control and case-cohort methods of sampling
fiam a cohort: a critical comparison. Am J Epidemiol 1990; 131(1): 169-76.
Lesko SM, Rosenberg L, Shapiro S. Family history and prostate cancer risk. Am J
Epidemiol 1996; 144: 1041-7.
6 2
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Lunn RM, Bell DA, Mohler JL, Taylor JA. Prostate Cancer Risk and Polymorphism
in 17 Hydroxylase (CYP17) and Steroid Reductase (SRD5A2). Carcinogenesis 1999;
20: 1727-31.
Luu-The V, Zhang Y, Poirier D, Labrie F. Characteristics of Human Types 1, 2, and
3 17|3-Hydroxysteroid Dehydrogenase Activities: Oxidation/Reduction and
Inhibition. J Steroid Biochem Molec Biol 1995; 55(5/6): 581-587.
Makridakis N, Salle E, Reichardt JKV. Biochemical and pharmacogenetic
dissection of human steroid 5-a reductase type II. Pharmacogenetics 2000; 10:
407-13.
McNeal JE. Prostate. In: Sternberg SS, Ed. Histology for pathologists. New York,
NT: Raven Pres, 1992: 740-63.
Makridakis NM, Ross RK, Pike MC, et al. Association of Mis-sense Substitution in
SRD5A2 Gene with Prostate Cancer in African-American and Hispanic Men in Los
Angeles, USA. Lancet 1999; 354(9183): 975-8.
Makridakis N, Reichardt JKV. MAPA: Multiplex Automated Primer Extension
Analysis. Simultaneous Genotyping of up to Four Polymorphisms. Biotechniques
2001; 31:1374-80.
Margiotti K, Kim E, Pearce CL, et al. Association of the G289S Single Nucleotide
Polymorphism in the HSD17B3 Genewith Prostate Cancer in Italian Men. The
Prostate 2002; 53: 65-68.
Miller BA, Kolonel LN, Bernstein L, et al, eds. Racial/ethnic patterns of cancer in
the United States 1988-1992. Bethesda, MD: National Cancer Institute, 1996. (NIH
publication no. 96-4104).
Miller GJ, Torkko KC. Natural history of prostate cancer-epidemiologic
considerations. Epidemiol Rev 2001; 23(1): 14-18.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Moghrabi N, Hughes IA, Dunaif A, Andersson S. Deleterious Missense Mutations
and Silent Polymorphism in the Human 17 Beta-Hydroxysteroid Dehydrogenase 3
Gene. J Clin EndoMetabol 1998; 83: 2855-2860.
Nobel RL. The Development of Prostate Adenocarcinoma in Nb Rats Following
Prolonged Sex Hormone Administration Cancer Res 1977; 37; 1929-33.
Parkin DM, Bray F, Ferlay J, Pisani P. Estimating the world cancer burden:
Globocan 2000. Int J Cancer 2001; 94(2): 153-6.
Parsons BL, Heflich RH. Genotypic selction methods for the direct analysis of point
mutations. Mutat Res 1997; 387: 97-121.
Pearce CL, Makridakis NM, Ross RK, et al. Steroid 5-a Reductase type IIV89L
subsititution is not associated with risk of prostate cancer in a multiethnic population
study. Cancer Epidemiol Biomarkers Prev 2002; 11: 417-418.
Peehl DM. Vitamin D and Prostate Cancer Risk. Eur Urol 1999; 35: 392-394.
Potosky AL, Miller BA, Albertson PC, et al. The role of increasing detection in the
rising incidence of prostate cancer. JAMA 1995; 273: 548-552.
Prentice RL. A Case-Cohort Design for Epidemiologic Cohort Studies and Disease
Prevention Studies. Biometrika 1986; 73: 1-11.
Spencer JA, Chng WJ, Hudson E, et al. Prostate specific antigen level and Gleason
score in predicting the stage of newly diagnosed prostate cancer. Br J Radiol 1998:
71: 1130-5.
Reichardt JKV, Makridakis N, Henderson BE, et al. Genetic variability of the
human SRD5A2 gene: implications for prostate cancer risk. Cancer Res 1995; 55:
3973-5.
64
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Ries LAG, Kosaiy Cl, Hankey BF, et al. SEER cancer statistics review, 1973-1997,
Bethesda, MD: US Department of Health and Human Services, National Institute of
Health. National Cancer Institute, (http://seer.cancer.gov/csr/1973_1999/, 2002).
Risch N. Searching for Genetic Determinants in the New Millennium Nature 2000;
405: 847-856.
Ross RK, Bernstein L, Lobo RA, et al. 5a-Reductase Activity and Risk of Prostate
Cancer among Japanese and US White and Black Males. Lancet 1992; 339: 887-9.
Ross RK, Schottenfeld D. Prostate cancer. In: Schottenfeld D, Fraumeni JF Jr., eds.
Cancer epidemiology and prevention. 2nd ed. New York, NY: Oxford University
Press 1996: 1180-205.
Ross RK, Coetzee GA, Pearce CL, et al. Androgen Metabolism and Prostate Cancer:
Establishing a Model of Genetic Susceptibility. Eur Urol 1999; 35(5-6): 355-61.
Sakr WA, Grignon Dj, Grissman JD, et al. High grade prostatic intraepithelial
neoplasia (HGPIN) and prostatic adenocarcinoma between the ages of 20-69: an
autopsy study of 249 cases. In vivo 1994; 8: 439-43.
Self SG, Prentice RL. Asymptotic distribution theory and efficiency results for
case-cohort studies. Ann Stat 1988; 16: 61-81.
Strachan T, Read AP. Human molecular genetics 2 2n d edition. A John Wiley &
Sons, Inc., Publication 1999; 273.
Thompson RS, Barlow WE, Taplin SH et al. A population-based case-cohort
evaluation of the efficacy of mammographic screening for breast cancer. Am J
Epidemiol 1994; 140(10): 889-901.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
W acholder S, Gail MH, Pee D, Brookmeyer R. Alternative variance and
efficiency calculations for the case-cohort design. Biometrika 1989; 76(1):
117-23.
Wacholder S. Practical considerations in choosing between the case-cohort and
nested case-control designs. Epidemiology 1991; 2(2): 155-58.
Wadelius M, Andersson AO, Johansson JE, et al. Prostate cancer associated with
CYP17 genotype. Pharmacogenetics 1999; 9: 635-9.
Weinberg W. Uber den Nachweis der verebung beim menschen. Jahreshefte Des
Vereins Fur Vaterlandische Naturkunde in Wurttemberg 1908; 64: 368-82.
Whittemore AS, Wu AH, Kolonel LN, et al. Family history and prostate cancer risk
in black, white, and Asian men in the United Sates and Canada. Am J Epidemiol
1995; 141:732-40.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Insulin-like growth factor 1 genotype, phenotype and breast cancer risk, by racial/ethnic group

PDF

Colorectal cancer risks in Singapore Chinese: Polymorphisms in the insulin-like growth factor-1 and the vitamin D receptor

PDF

CYP17 polymorphism and risk for colorectal adenomas

PDF

Infrequent androgen receptor mutations in primary prostate tumors from men residing in Singapore and Los Angeles

PDF

Recreational physical activity and risk of breast cancer: The California Teachers Study

PDF

BRCA1 mutations and polymorphisms in African American women with a family history of breast cancer identified through high throughput sequencing

PDF

Polymorphisms in genes involved in steroid hormone metabolism and mammographic density changes in women randomized to menopausal estrogen and progesterone therapy

PDF

Cigarettes and alcohol in relation to colorectal cancer within the Singapore Chinese Health Study

PDF

A case-control study of passive smoking and bladder cancer risk in Los Angeles

PDF

Association between body mass and benign prostatic hyperplasia in Hispanics: Role of steroid 5-alpha reductase type 2 (SRD5A2) gene

PDF

Association of vitamin D receptor gene polymorphisms with colorectal adenoma

PDF

Validation of serum cotinine as a biomarker of environmental tobacco smoke exposure: Validation with self-report and association with subclinical atherosclerosis in non-smokers

PDF

The association between recreational physical activity and mammographic density

PDF

Progression of carotid intima-media thickness and plasma antioxidants: The Los Angeles Atherosclerosis Study

PDF

The influence of family structures on adolescent smoking among multicultural adolescents in Hawaii

PDF

Lymphedema: Impact on breast cancer survivors

PDF

Stability of infectious human immunodeficiency virus type 1 (HIV-1) in human body fluids

PDF

Selective laser trabeculoplasty for the treatment of glaucoma

PDF

Post-intensive care unit mechanical ventilation: Relationship of infections to outcomes of weaning from prolonged mechanical ventilation

PDF

beta3-adrenergic receptor gene Trp64Arg polymorphism and obesity-related characteristics among African American women with breast cancer: An analysis of USC HEAL Study

Asset Metadata

Creator Kim, Eugene (author)

Core Title Single-nucleotide polymorphisms in 17beta-hydroxysteroid dehydrogenase type III and prostate cancer risk

Degree Master of Science

Degree Program Molecular Epidemiology

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

Tag biology, genetics,biology, molecular,health sciences, oncology,health sciences, public health,OAI-PMH Harvest

Language English

Contributor Digitized by ProQuest (provenance)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-305229

Unique identifier UC11342021

Identifier 1416561.pdf (filename),usctheses-c16-305229 (legacy record id)

Legacy Identifier 1416561.pdf

Dmrecord 305229

Document Type Thesis

Rights Kim, Eugene

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