Subcellular localization of

zebrafish androgen receptor

in the presence of natural
ligands and antiandrogens

Molecular Biology and Environment Biology

Master (Fall 2007-Fall 2008)
JingYuan Feng
Supervisors: Ole Andersen and Lene Juel Rasmussen

Roskilde University
2
Abstract
In vertebrates, androgens are responsible for male primary and secondary sexual
characteristics and behavior. Androgens exert their functions by interacting with the
androgen receptor. This signaling pathway can be disrupted by exogenous chemicals
that can either mimic or antagonize the actions of the hormones. In this study, the
subcellular localization of androgen receptor from Zebrafish (Danio rerio) (zfAR)
was studied by expressing the yellow fluorescence protein (YFP)-zfAR fusion protein
in Hela and ZFL cell lines. Unliganded zfAR resides relatively even between
cytoplasm and nucleus, and it undergoes nuclear translocation in response to the
binding of ligands such as 11-ketotestorsterone (11-KT) and dihydrotestosterone
(DHT). Mutations created by site-directed mutagenesis in the predicted nuclear
localization sequence (KKCFEVGMTLGARKLRK) severely destroyed the 11-KT
induced nuclear translocation of zfAR. Furthermore, environmental antiandrogenic
chemicals bisphenol A, p, p’-DDE, vinclozolin and fenitrothion did not induce nuclear
translocation of zfAR alone, but significantly inhibited the 11-KT induced nuclear
import of zfAR at concentration 10-6M. Combinational effects of these four chemicals
were larger than effects observed with single chemical, but no cumulative effects
could be concluded in this study.

3
4
Acknowledgement
I would like to thank my supervisors Professor Ole Andersen and Professor Lene Juel
Rasmussen for the time and money they spend on me. I would also like to give my
special thanks to Ph.D Anne Jørgensen for providing valuable suggestions for both my
experiments and writing. I would like to thank Ph.D Claus Desler for teaching me
how to culture cell lines, and for the discussion we have had about my project. I
would also like to thank Ph.D Sascha Liberti, Post Doc. Anne Lützen, and Merete
Rasmussen for their help in the daily work in the laboratory. I would like to thank my
friend master student Henrik Jacobsen for his proofreading of the thesis and his help
in improving some of the graphs in the thesis. I would also like to thank all the people
from Lene Juel Rasmussen’s laboratory for providing a happy and comfortable
working environment.

5
6
Table of Contents

Abstract 3

Acknowledgement 5

Abbreviations and definitions 9

Introduction 11
1. Androgen and Androgen receptor 11
1.1 Androgen signaling 12
1.2 Nuclear import of androgen receptor 13
1.3 Androgen receptor subnuclear localization 15
1.4 Zebrafish androgen receptor and androgen signaling 15
1.4.1 Androgen signaling in fish 15
1.4.2 Characterization and structure of zebrafish androgen
receptor 16
2. Environmental Androgen Disrupting Chemicals (ADCs) 17
2.1 Action Mechanisms of ADCs 19
2.1.1 Bisphenol A (BPA) 20
2.1.2 P, p’-DDE 21
2.1.3 Vinclozolin 22
2.1.4 Fenitrothion 23
2.2 Mixture effects of antiandrogens 24
2.3 Zebrafish in endocrine disrupting chemicals evaluation 26

Materials and Methods 29

Construction of expression vector 29
Mutagenesis 29
Cell culture 31
Transfection and microscopy 32

7
 Chemicals 33
Cytotoxicity assessment 34
Studies of antiandrogenic chemicals 34

Results 37
Construction of expression vector pEYFP-zfAR 37
ZfAR subcellular localization 38
Bioinformatics search for Nuclear Localization Sequence (NLS) within
zfAR 42
Mutagenesis and the subcellular localization of zfAR mutants 43
Quantification of Hela and ZFL cells and cytotoxicity test with MTT
assay 48
Subcellular localization of zfAR in the presence of antiandrogens 49
Subcellular localization of zfAR in the presence of antiandrogen and
11-KT 51
Mixture effects 55

Discussion 57
Subcellular localization of zfAR upon exposure to 11-KT 57
Subcellular localization of zfAR upon exposure to ADCs 61

Conclusion 67

References 69

8
Abbreviations and definitions
11-KT: 11-ketotestosterone

ADC(s): Androgen Disrupting Chemical(s)

ARE: Androgen Response Element

BPA: Bisphenol A

DBD: DNA binding Domain

DHT: 5α-dihydrotestosterone

DMSO: dimethyl sulfoxide

EDC(s): Endocrine Disrupting Chemical(s)

FSH: Follicle Stimulating Hormone

hAR: human Androgen Receptor

LBD: Ligand Binding Domain

NLS: Nuclear Localization Sequence

NES: Nuclear Export Signal

SHBG: Sex Hormone Binding Globulin

T: Testosterone

TEF: Toxic Equivalency Factor

YFP: Yellow Fluorescence Protein

zfAR: zebrafish Androgen Receptor

9
10
Introduction
Despite extensive differences in sex determination and in sexual development and
function between different types of vertebrates (fish, amphibians, birds, reptiles and
mammals), some mechanisms are common, namely the involvement of androgens
and estrogens in development and function of male and female primary and
secondary sexual characteristics and behavior. The function of these steroids can be
disrupted by exogenous chemicals that can either mimic or antagonize their actions.
Therefore, environmental releases of industrially and agriculturally generated
contaminants represent considerable risk in human and wildlife health. A battery of
tests screening for endocrine disrupting potential among the existing and newly
generated chemicals was recommended by the Endocrine Disrupter Screening and
Testing Advisory Committee (EDSTAC) (1998) convened by the U.S. Environmental
Protection Agency (EPA), and by an OECD (Organization for Economic Co-operation
and Development) working group (EDTA-Endocrine Disrupter Testing and
Assessment Task Force).Understanding the signaling pathways of the steroids is the
first step towards a fair risk assessment of potential Endocrine Disrupting Chemicals
(EDCs).

1. Androgen and Androgen receptor

In vertebrates, the seminiferous tubules are lined with germ cells and Sertoli cells.
Follicle-stimulating hormone (FSH) released from anterior pituitary binds to its
receptors on Sertoli cells and consequently stimulates spermatogenesis in the germ
cells after sexual maturity. Sertoli cells support the development of the sperm and are
responsible for androgen receptor synthesis. Between the seminiferous tubules,
Leydig cells produce and secrete androgens, especially testosterone.

The androgens are responsible for triggering development of the primary male sexual
characteristics in the embryo, and the secondary male sexual characteristics at the
time of puberty. The androgens also contribute to general growth, i.e. the synthesis of
myofibrillar proteins in muscle, and male psychological and behavioral changes.
11
1.1 Androgen signaling

Steroids hormone are delivered to their target sites through endocrine, paracrine and
autocrine secretion. Circulating steroids rapidly reach equilibrium among different
body fluid compartments. About 98% of the lipophilic steroid hormones are
transported by sex hormone binding globulin (SHBG) and 2% of the steroid
hormones are circulating free and thus are biologically active (You 2004).

The basic mode of action for sex steroids is to interact with ligand-specific
intracellular nuclear receptors. The activated nuclear receptors bind to hormone
response elements upstream structural genes and alter their rates of transcription.
These actions of steroids are thus genomic, which are typically slow. Steroid
hormones also exert rapid cellular actions that, with or without the involvement of
their nuclear receptors, do not require alteration in gene expression and it is therefore
termed the non-genomic activity.

Both genomic and nongenomic pathways play important roles in androgen signaling.
Nongenomic action of androgens can be mediated through membrane or membrane
associated androgen receptor/binding proteins, changes in membrane flexibility,
changes in Ca2+, activation of second messenger pathway or a combination of these
mechanisms [reviewed by (Feldman and Feldman 2001; Foradori et al. 2008;
Heinlein and Chang 2002; Michels and Hoppe 2008)]. In the genomic pathway, free
testosterone enters cells, and is immediately metabolized to the more potent androgen
5α-dihydrotestosterone (DHT). DHT binds to the inactive nuclear androgen receptor.
In many, but not all, cells in tissue culture, inactivated androgen receptor is
predominantly cytoplasmic (Black and Paschal 2004). Upon ligand binding, androgen
receptor undergoes a series of events including homodimerization, phosphorylation,
and conformational change that allow the androgen receptor to enter the nucleus via
an intrinsic nuclear localization signal. Inside the nucleus, AR interact with a specific
sequence on DNA located upstream the promoter termed Androgen Response
Element (ARE) (Feldman and Feldman 2001; Hirawat et al. 2003; Roy et al. 2001)

12
(figure 1). This interaction facilitates the downstream gene transcription by recruiting
numbers of coregulator proteins such as CBP, p300, and other nuclear receptor
coactivateor pCAF, SRCI and SRC3. Ultimately RNA Pol II is recruited and bind to
the promoter of target genes [reviewed in (Culig et al. 2003; Lee and Chang 2003a)].

Figure 1. Genomic pathway of androgen signaling. Free testosterone enters cells, and is
immediately metabolized to the more potent androgen 5α-dihydrotestosterone (DHT). DHT
binds to the inactive nuclear androgen receptor. Upon ligand binding, androgen receptor
undergoes a series of events including homodimerization, phosphorylation, and
conformational change that allow the androgen receptor to enter the nucleus. AR interacts with
androgen response element and facilitates the downstream gene transcription. Picture
adapted from (Feldman and Feldman 2001).

1.2 Nuclear import of androgen receptor

The genomic pathway of androgen signaling is mediated through the nuclear

androgen receptor(s). Nuclear androgen receptors are members of the steroid receptor
superfamily of ligand-dependent transcription factors that generally consist of an

13
amino-terminal hypervariable transcriptional activation domain, a central highly
conserved DNA-binding domain (DBD), and a C-terminal ligand-binding domain
(LBD). DBD and LBD are connected by a hinge region (Hirawat et al. 2003; Roy et
al. 2001).

Nuclear translocation of AR is one of the critical steps, and this ligand-dependent

import of AR is regulated by several coregulators and an intrinsic nuclear localization
sequence (NLS). The first identified NLS motif was that of the SV40 large T-antigen
(PKKKRKV). A typical SV40-like NLS is a small sequence of seven or eight amino
acids containing a single cluster of basic residues. This type of NLS is simple and
usually sufficient in inducing nuclear transport (LaCasse and Lefebvre 1995).
Mutagenesis studies of human androgen receptor (hAR) revealed that it contains
another type of NLS, the bipartite type (Zhou et al. 1994), which is more complex and
more dependent on other functional domains in the protein compared to the first type
of NLS (Saporita et al. 2003) . For example, the phosphorylation of the specific site X
in hAR stabilizes the ligand/receptor homodimer, marking the ligand-receptor
complex for translocation to the nucleus (Edwards and Bartlett 2005; Lee and Chang
2003b; Zhu et al. 2001). A bipartite nuclear targeting sequence consists of two
clusters of basic amino acids separated by 10 amino acids, e.g.
RKCYEAGMTLGARKLKK in hAR (Zhou et al. 1994). The C-terminal basic
sequence is similar to the NLS of SV40 large T-antigen. However, both basic clusters
are essential for efficient nuclear import. In hAR, the NLS locates in the DNA binding
and hinge regions at amino acids 617-633, and is modulated by interactions between
the NH2- and carboxyl-terminal regions (Zhou et al. 1994).

In human cells, inactivated androgen receptors are bound by chaperone proteins such
as heat-shock protein 70 (Hsp 70) and Hsp 90 in cytoplasm. Upon ligand binding, the
androgen receptor undergoes conformational changes that expose the NLS. The
exposed NLS can be recognized by import machinery such as importin-α, which, with
importin-β, mediates docking interaction onto the nuclear-pore complex.
Translocation of the receptor complex through the nuclear membrane is inferred to

14
directly interact with the RanGTP (Paschal 2002; Roy et al. 2001).

1.3 Androgen receptor subnuclear localization

Localization studies of GFP chimeras of AR show that the subnuclear distribution of

the protein adopts a punctuate pattern which are commonly referred to as “nuclear
foci” (Black and Paschal 2004; Roy et al. 2001; Tomura et al. 2001). The
hyperspeckled distribution of agonist-bound GFP-AR is clearly correlated with both
decreased solubility and mobility (Marcelli et al. 2006). Fluorescence image and
three-dimensional reconstruction studies suggested that the agonist-bound AR is
localized between euchromatin and heterochromatin, which gives rise to nuclear foci
numbering between 250 and 400 (Tomura et al. 2001). The formation of foci is
dependent on agonist binding of AR. In contrast, antagonist bound AR is distributed
more uniformly throughout the nucleus (Karvonen et al. 2002). The underlying
composition of this subnuclear compartment is shown to be associated with the ability
of agonist bound AR to interact with other transcription co-activator, such as CBP and
p160 (Black and Paschal 2004; Karvonen et al. 2002; Marcelli et al. 2006). This is
also confirmed by the fact that flutamide-bound AR (flutamide is a pharmaceutical
antiandrogen) does not have transactivity for downstream genes, even though nuclear
translocation occurs but fails to form nuclear foci. Therefore, nuclear foci represent
transcriptionally active forms of the receptor while homogeneous receptor distribution
represents transcriptionally inactive forms (Kumar et al. 2006). The agonist binding
of AR has significant influence on normal function of AR as it specifically initiates a
complex cascade of events that include nuclear translocation, hyperspeckle formation,
and co-activator recruitment (Marcelli et al. 2006).

1.4 Zebrafish androgen receptor and androgen signaling

1.4.1 Androgen signaling in fish

Knowledge about androgen signaling in teleost fish is currently very limited. In fish,
11-ketotestosterone, testosterone and androstenedione are the main functional

15
androgens and it is believed that the action of androgens is primarily mediated
through the classic genomic pathway by binding to intracellular androgen receptors.
At the same time, the nongenomic signal pathway has also been described in specific
functions (fx oocyte maturation) in some fish and amphibian species (Braun and
Thomas 2004; White et al. 2005).

Until now androgen receptors from many fishes have been cloned, including
zebrafish, Atlantic croaker, mosqitofish, three-spine stickleback, fathead minnow etc.
Both membrane AR and nuclear AR have been discovered (Braun and Thomas 2004;
White et al. 2005). While in some fish species two types of AR have been described,
in other species, only one type of AR has been discovered. When two ARs have been
identified in the same specie, the ARs can differ from each other in several ways
including tissue expression, physiological functions, specificity in binding ligands and
mechanism of action.

1.4.2 Characterization and structure of zebrafish androgen receptor

Zebrafish androgen receptor gene (zfar) was firstly cloned and characterized by
Jorgensen et al (2007) and Hossain et al. (2008) independently. In both studies, only a
single ar locus was found. Blast results show that the single locus located at
chromosome 5, and is 118.24 kb long and contains 13 exons (Hossain et al. 2008).
The sequence analysis reveals that the zfAR has the three typical nuclear hormone
receptors functional domains: an amino-terminal hypervariable transcriptional
activation domain (TAD), a central highly conserved DNA-binding domain (DBD),
and a C-terminal ligand-binding domain (LBD). A putative leucine zipper and two
zinc fingers motifs located within the LBD contribute to the dimerization of receptors
and DNA binding upon ligand binding (Hossain et al. 2008; Jorgensen et al. 2007).

Tissue-specific expression analysis showed that the zfar was expressed ubiquitously
in all adult tissues, with sexually dimorphic expression in the gonad and muscle. The
ar transcript was maternally deposited into the embryo (Hossain et al. 2008).

KT was found to be the most potent agonist of zebrafish AR among KT, DHT, T

16
(testosterone), A (androstenedione) and E2 (estradiol) in the gene expression assay
(Hossain et al. 2008), even though DHT has the highest binding affinities in
competition with 3H-T among DHT, KT, T and A (Jorgensen et al. 2007).

Despite that zebrafish is an important model organism in developmental biology and

ecotoxicology, the mechanism of sex determination and gonadal sex differentiation
has not been elucidated. Several studies indicate the importance of androgen signaling
in zebrafish sexual development (Kinnberg et al. 2007; Orn et al. 2003), but
knowledge concerning the androgen molecular action is very limited. Thus it is
interesting to study the role of zfAR in androgen signaling in order to understand
more about the molecular mechanism of zebrafish sex differentiation. The
investigation of the subcellular localization of zfAR will provide the first hint of
androgen signaling in zebrafish. Based on the conservation of the androgen signaling
pathway among different species, our hypothesis is that the inactivated zfAR locates
in cytoplasm, and upon binding to its natural ligand zfAR undergoes nuclear
translocation via an intrinsic NLS. To test our hypothesis, subcellular localization of
zfAR was studied by tagging it to a yellow fluorescence protein (YFP), and the
function of a hypothetical NLS within zfAR was studied using site-directed
mutagenesis.

2. Environmental Androgen Disrupting Chemicals (ADCs)

An endocrine-disrupting chemical (EDC) is defined as “an exogenous substance or

mixture that alters function(s) of the endocrine system and consequently produces
adverse health effects in an intact organism, or its progeny or Populations” by the
World Health Organization (2002). EDCs include some organic chemicals, as well as
some metals used extensively in the past, especially in industry and agriculture. EDCs
are found in many of the every day products we are currently using including plastic
bottles, metal food cans, detergents, flame retardants, food, toys, cosmetics and
pesticides, and therefore possibly represent a human risk. Besides synthetic sources,
EDCs also have natural sources including human and animal hormones and phyto-

17
and mycoestrogens found in sewage, animal husbandry runoff, and intentionally or
accidentally as food and feed ingredients (Goksoyr 2006).

EDCs have diverse structures, but most of them and their metabolites share lipophilic
and persistent nature. They accumulate and biomagnify in different environmental
compartments, including aquatic biota. Many of these chemicals have been linked
with developmental, reproductive, neural, immune, and other problems in wildlife.

According to the phenomenological similarities caused by EDCs and the most

frequently studied hormone receptors, EDCs can be roughly divided into the
following groups: estrogenic chemicals (e.g. diethylstilbestrol), antiandrogens (e.g.
vinclozolin), thyroid-disrupting chemicals (e.g. PCDDs and PCBs) and AhR agonists
(e.g. the dioxin-like chemicals and some PCBs) (Kortenkamp 2007).

In the aquatic environment, chemicals with estrogenic activity have been the focus of
a great number of studies (Brevini et al. 2005; Goksoyr 2006; Sonnenschein and Soto
1998). Many chemicals that were in extensive industrial and agricultural use have
been banned or limited because of their endocrine disrupting potential, such as some
PCBs.

Since the early 1990’s, studies have revealed decreased semen quality in Danish men
and increased incidences of prostate cancer, hypospadia, and cryptorchidism
(Andersen et al. 2000). Hereafter, chemicals known to disrupt androgen signaling
pathways became a research hot spot, and this group of chemicals are therefore called
androgen disrupting chemicals (ADCs) (Gray et al. 1994; Gray et al. 1999b; Kelce et
al. 1995; Tamura et al. 2001; Vinggaard et al. 2006). In contrast to estrogens that
feminize males, ADCs interfere with natural androgen functions and thereby cause
malformation of male reproductive tissue or inhibition of development and function
of male characteristics. In fact, some chemicals that had previously been shown to be
estrogen agonists are also antiandrogens (Lee et al. 2003; Sohoni and Sumpter 1998).
In some studies using a rodent model, the endpoints to detect antiandrogenic effects
after in utero exposure of male rats include changes in the anogenital distance

18
(distance between penis and anus), retained nipples, and alteration in the weight of
sexual organs and accessory glands. Another assay used for screening
(anti)androgenic compound is the Hershberger assay where the weights of male
reproductive tract tissues are measured after in vivo exposure (Christiansen et al. 2008;
Gray et al. 1999a; Gray et al. 1994; Gray et al. 1999b; Hass et al. 2007; Metzdorff et
al. 2007; Rider et al. 2008). Classes of chemicals currently known to interfere with
the androgen signaling pathways include dicarboximide fungicides (e.g.vinclozolin),
organochlorine-based insecticides (e.g. p,p’-DDE), organophosphate insecticide (e.g.
fenitrothion), conozole fungicides (prochloraz), plasticizers (phthalates, BPA), and
urea-based herbicides (linuron) (Wilson et al. 2008).

Many of these androgen disrupting chemicals are found not only to antagonize
androgen functions in mammals, but also in other vertebrates, including fishes
(Arukwe 2001; Bayley et al. 2003; Kinnberg et al. 2007; Kiparissis et al. 2003; Leon
et al. 2007; Martinovic et al. 2008).

2.1 Action mechanisms of ADCs

Hormones are maintained through a dynamic balance between steroidogenesis and

steroid inactivation to ensure their physiological function. Although many steps of the
sensitive endocrine regulatory system can be influenced by different external stimuli,
most effects of endocrine disruptors observed and explained until now are attributed
to the function of the gonads, which control the development of sexual differentiation,
secondary sex characteristics, and functioning of sex organs (Goksoyr 2006). As
presented before, physiological functions of sexual hormones are largely mediated
through the activation of ligand-specific steroid hormone receptors. Therefore,
exogenous compounds can disrupt normal hormonal functioning by either interacting
with the receptors or altering ligands bioavailability through interfering with steroid
hormones biosynthesis, transportation, action in target organs and the coordinated
metabolism. Further, since endocrine homeostasis and hormone secretion require
complex regulations and coordination, interference of exogenous compounds with

19
biosynthesis and metabolism of one particular hormone could indirectly interfere with
other hormones via cross-talk of signal pathways or feedback regulation (You 2004).

Androgen disrupting chemicals also work through multiple mechanisms. One of the
major mechanisms for many of the antiandrogens that have been described to date is
to directly interact with the androgen receptor and thereby interfere with functions of
the natural androgen (Wilson et al. 2008). Since normal androgen signaling requires a
series of events, interference from antiandrogenic chemicals could happen during any
of the steps or a combination of several steps (Lee et al. 2003; Wilson et al. 2008).
The mode of action for antiandrogens have primarily been verified through in vitro
experiments such as receptor binding assays, subcellular localization studies of
androgen receptors, and reporter gene assays (Kelce et al. 1995; Lee et al. 2003; Soto
et al. 1998). Another group of antiandrogens, such as certain phthalates, does not
antagonize the androgen receptor directly, but instead inhibits fetal testis hormone
production including testosterone (Gray et al. 2006; Wilson et al. 2008). Some
antiandrogens, as mentioned before, are also estrogen receptor agonists (Sohoni and
Sumpter 1998). Thus antiandrogen is often used broadly in terms of counteracting the
effects of androgens in a functional sense, as opposed to referring only to AR
antagonists. However, this project mainly focuses on the narrow sense of
antiandrogens, even though some of the chemicals used here may work in multiple
modes of action. The structures of the four chosen antiandrogens are shown below:

2.1.1 Bisphenol A (BPA)

BPA is one of the industrial compounds that have attracted much attention because of
its widespread use. It is a monomer used in polycarbonate plastic and polystyrene
resins and as dental sealants. Harsh detergents, acidic or high temperature liquids

20
cause depolymerazation of the products resulting in leaching of BPA and its
derivatives into foods, infant formula, and saliva etc. BPA was first known to elicit
estrogenic activity by competing with estradiol for binding to the estrogen receptor
and increased MCF-7 cell proliferation (Krishnan et al. 1993). Since then great
attention has been raised concerning the influence of BPA on the environmental
estrogen pool. Numerous in vitro and in vivo studies have focused on the potential of
the compound to affect ER transactivation in human or yeast cells and to mimic the
physiological function of natural estrogen in mammals (Gaido et al. 1997; Gould et al.
1998; Kurosawa et al. 2002). Bisphenol A was verified as a weak estrogen receptor
agonist (Gaido et al. 1997). However, in vivo studies show different physiological
effects of bisphenol A compared to estradiol, and this leads to the conclusion that
bisphenol A interacts with the estrogen receptor alpha in a distinct manner from
estradiol (Gould et al. 1998). Further, in vitro studies found that Bisphenol A is also a
human androgen receptor antagonist, an aryl hydrocarbon receptor (AhR) antagonist
and an aromatase activity modulator (Bonefeld-Jorgensen et al. 2007; Lee et al. 2003;
Nativelle-Serpentini et al. 2003; Xu et al. 2005). Thus BPA has the potential to affect
several cellular pathways, including gene expression regulated by the steroid
receptors ER and AR, the conversion of testosterone into estrogen by aromatase, and
the function of AhR, involved in synthesis of steroids such as estrogens and
metabolism of steroids and xenobiotic compounds (Bonefeld-Jorgensen et al. 2007).

2.1.2. P, p’-DDE

The p, p’-DDE (dichloro- diphenyldichloroethylene), a metabolite of the pesticide

DDT, is one of the first identified antiandrogens (Kelce et al. 1995; Sohoni and
Sumpter 1998). It is a well-known environmental pollutant that possesses estrogenic
activity and accumulates in tissues because of slow metabolic degradation and
excretion. Kelce et al. (1995) provided compelling evidence for its antiandrogenic
effects from both in vitro and in vivo studies. In vitro, p, p’-DDE competitively binds
to the rat androgen receptor, and blocks androgen-induced human AR mediated
transcription activity. In developmental studies with the rat, p, p’-DDE exposed dams

21
gave birth to male pups that exhibited reduced anogenital distance and retained
thoracic nipples (Kelce et al. 1995). Also, the onset of pubertal maturation was
delayed in male rats treated with p, p’-DDE and the adult male rats had reduced
androgen-dependent seminal vesicle and ventral prostate weights (Kelce et al. 1995).
Increase in testosterone-repressed prostatic message 2 (TRPM-2) messenger RNA
levels and a decline in androgen-induced prostate binding protein subunit 3 (C3)
mRNA levels also support the proposed antiandrogen effects of p, p’-DDE (Kelce et
al. 1995).

2.1.3 Vinclozolin

Vinclozolin is a fungicide used on food crops (soft fruits and vegetables). It is

classified as a group C chemical (possible human carcinogen) by U.S. EPA, because
of its antiandrogenic activity and association with prostate cancer. It is one of the
environmental antiandrogens that has been studied the most until now, and no other
mechanism than this has been suggested. It was first reported to cause antiandrogenic
effects during male rat sexual differentiation (Gray et al. 1999a; Gray et al. 1994), and
later in vitro studies found that not only vinclozolin itself competitively antagonized
androgen binding, its two main degradation products, 2-[[(3,5-dichlorophenyl)-
carbamoyl]oxy] -2-methyl-3-butenoic acid (M1) and 3',5'-dichloro-2-hydroxy-2-
methylbut-3-enanilide (M2) are even more potent antagonists of rat and human
androgen receptor binding (Kelce et al. 1994). In addition, M2 inhibits
receptor-mediated transactivation in rats by blocking the binding of androgen-bound
AR to the ARE on DNA (Kelce et al. 1997). Addition of vinclozolin to rat serum at
37℃ resulted in more than 90% conversion to M1 and M2 within 2 hours. It is
therefore suggested that the demasculinizing effects of vinclozolin exposure in vivo
was mediated via the metabolites M1 and/or M2 (Kelce et al. 1994). Current in vivo
evidence is mostly based on rodent models, and even though comparative physiology
makes it very likely that humans respond to vinclozolin exposure like rodents, the
only study based on human occupational exposure to vinclozolin for 1 to 13 years
found no evidence of antiandrogenic or other reproductive effects (Zober et al. 1995).

22
In fathead minnow, the detection of the vinclozolin metabolites M1 and M2 is
consistent with studies in rat (Makynen et al. 2000). However, inconsistent results
have been reported regarding its antiandrogenic effects among fishes. Makynen et al.
(2000) found no significant effects on reproduction at concentration as high as 1200
µg/L in fathead minnow, while Martinovic et al. (2008) claimed to confirm
antiandrogenic effects of vinclozolin in the same species, with decreased fecundity
and expression of male secondary sexual characteristics in vinclozolin-exposed fish.
Vinclozolin has also been found to cause antiandrogenic effects in male guppy
(Bayley et al. 2003). Binding studies show that the binding affinity of vinclozolin to
AR depends on the type of AR. In fathead minnow, vinclozolin, M1 and M2 had a
low binding affinity to the putative androgen receptor in brain and gonadal extract
(Makynen et al. 2000), but a high affinity for androgen receptor from some tissues in
kelp bass and Atlantic croaker has been reported (Sperry and Thomas 1999). In a
study on global gene expression profiling of androgen disruption in medaka,
vinclozolin shared similar antiandrogenic expression profiles with flutamide, a
prototypic antiandrogen (Leon et al. 2008).

2.1.4 Fenitrothion

The organophosphothioate pesticide fenitrothion (F), [0,0-dimethyl-0-(4-nitro-m-tolyl)

phosphorothioate], is an organophosphate insecticide that has been widely used since
1959 to control insects in agriculture and for flies, mosquito, and cockroach control in
public health campaigns. As a result, the human health effects especially among
applicators, gardeners, and farmers, associated with exposure to fenitrothion are of
great concern. Tamura and colleagues (2001) recognized that fenitrothion shares
structural similarities with the pharmaceutical antiandrogen flutamide and the
environmental antiandrogenic herbicide linuron. Their preliminary work showed that
fenitrothion is an antiandrogen in rat and its antiandrogenic activity was proved by in
vitro assays (Tamura et al. 2001). Fenitrothion competitively antagonize
DHT-dependent hAR activation in a luciferase reporter assay in HepG2 cells, and it
significantly reduced weights of accessory sex glands in Hershberger assay (Tamura

23
et al. 2001). It is also found to be an antiandrogen in three-spined stickleback
(Katsiadaki et al. 2006; Sebire et al. 2008). It disrupted not only spiggin (androgen
specific expressed protein) production in the kidney in male three-spined stickleback,
but also androgen-dependent sexual behavior such as nest building and courtship
behavior (Sebire et al. 2008). However, other in vivo studies with rats on fenitrothion
have been inconclusive about its anti-androgenic activity (Sohoni et al. 2001; Sunami
et al. 2000). Fenitrothion is, as all the other organophosphate compounds, also an
acetylcholinesterase inhibitor, and therefore a neurotoxic compound.

2.2 Mixture effects of antiandrogens

Even though it is common sense that biological systems exposed to a mixture of

hundreds or thousands of chemicals, studies on mixture effects of EDCs is relatively
delayed compared to mounting reports on individual chemical effects. It is mainly
because mixture studies are considered to be challenging, both conceptually and
experimentally. In spite of the difficulties, studies on combination effects of endocrine
disrupting chemicals, especially androgen disrupting chemicals, have been conducted
(Birkhoj et al. 2004; Christiansen et al. 2008; Gray et al. 2006; Hass et al. 2007;
Hotchkiss et al. 2004; Jarfelt et al. 2005; Metzdorff et al. 2007; Nellemann et al. 2003;
Rider et al. 2008).

Mixture effects of chemicals are termed additive, antagonistic, or synergistic

indicating whether the individual chemicals in the mixture act together to produce
equal, diminishing or enhancing effects compared to the reference effects predicted
on the basis of information about all components in the mixture. It is therefore
important to choose the correct model to predict the anticipated reference effect.
Additivity is usually assumed in mixture studies, and two concepts are available to
describe the action of chemical mixtures depending on the presumed modes of action
of the mixture components: dose addition and independent action (also called
“response addition”). Dose addition applies to mixtures of chemicals that act through
similar mode of action, and therefore based on this assumption, one chemical in the

24
mixture can be replaced by a fraction of another chemical at equieffective
concentration without diminishing the overall combined effects. The toxic equivalency
factor (TEF) approach is a widely used application of dose addition. Independent
action is used for combinations of chemicals with distinct modes of action, and every
component in the mixture provokes effects independent of all the other agents.
Recently, to address the cumulative toxicity of complex mixtures containing
chemicals with both same and different mechanisms, an integrate addition concept
has been introduced, and it integrates dose addition and response addition into a
single model (Rider et al. 2008).

The choice of additivity model has significant implications in risk assessment (Rider
et al. 2008). It is obviously impossible to test all the combination of chemicals in the
environment, therefore prediction of mixture effects is important to assess the
potential risk before regulations are made. Often human exposure levels of single
chemicals are lower than the toxicity testing concentrations because of low resolving
power of the most of in vivo and many in vitro assays. When predicting combination
effects of mixtures with each component presenting at a dose level below or at the
dose that produce no effect, response addition does not predict adverse effect, while
dose addition does. Therefore, according to the dose addition concept we risk a higher
exposure to chemicals that can cause cumulative adverse effects even if each single
chemical does not if the regulations are made on chemical-chemical basis.

Mixture studies of antiandrogenic chemicals provide more evidence supporting the

dose addition concept than for response addition concept, regardless that the
components have the same or different target in their mode of action. Nellemann et al.
(2003) mixed vinclozolin and procymidone, both of which antagonize androgen
receptor, and observed additivity in vivo as well as in vitro by comparing the observed
effects with predicted effects by using an isobole method. Using the same method,
Birkhøj et al. (2004) observed additive antiandrogenic effects in vitro of five
commonly used pesticides acting dissimilarly. Metzdorff et al. (2007) studied the
dysgenesis, changes of genitals and perturbation of gene expression in male rats after

25
in utero exposure to a mixture of three androgen receptor antagonists each at
concentration where no effect were observed for individual compounds. The three
chemicals were vinclozolin, flutamide and procymidone, which share a common
antiandrogenic mechanism by antagonizing the androgen receptor, and the joint
effects were dose additive with all measured endpoints (Metzdorff et al. 2007). The
effects on sexual differentiation in rat were studied by Hass et al. (2007) with the
same combination of chemicals, and once again, the effects of the mixture of similar
acting antiandrogens were accurately predicted by the dose addition concept. In a
more complicated experimental setup, mixture effects of seven antiandrogens that
alter androgen signaling pathway via diverse mechanisms were studied in vivo (Rider
et al. 2008). By comparing the predicted responses generated by models of dose
addition, response addition or integrated addition, Rider et al. (2008) concluded that
response addition and integrated addition underestimated mixture effects, and that
chemicals that disrupt fetal tissue development during sexual differentiation act in a
cumulative, dose-additive manner irrespective of the specific cellular mechanism of
toxicity. Almost all the above studies claimed to get “something from nothing” results
concerning the joint antiandrogenic effects.

2.3 Zebrafish in endocrine disrupting chemical evaluation

Wildlife populations are likely to be exposed to contaminants such as insecticides,

pesticides and fungicides, because these chemicals usually end up in the aquatic
ecosystems by surface water flow over fields or through the waste water from sewage
plants.

A series of international test guidelines for detecting EDCs in human and various
animal species has been developed in numerous laboratories organized by the
Organization for Economic Co-operation and Development (OECD). Among these
medaka, fathead minnow, three-spined stickleback and zebrafish have been
recommended as test species in the Fish Sexual Development Test (FSDT) aiming at
detecting endocrine disrupting chemicals affecting fish. The FSDT measures the

26
change in the yolk protein (vitellogenin) concentration and sex ratio as the main
endpoints. The principle of the proposed test method is based on the existing
international guideline OECD TG 210 “Fish Early Life Stage Toxicity Test”, where
exposure starts with fertilized eggs and in the FSDT test continues until sexual
differentiation is finished.

However, compared to the abundant studies of antiandrogens in the other fish species,
investigations in zebrafish are relatively sparse. In the present project, the recently
characterized zebrafish androgen receptor was used to test antiandrogens by
investigating the subcellular localizations of the protein in the presence of four known
antiandrogens. As previously mentioned, antiandrogens can potentially target any step
in the androgen signaling pathway, however, this project mainly focuses on their
influence on the nuclear transportation of AR in response to an androgen. Previous
studies have shown varying results depending on the investigated antiandrogens.
Human AR bound vinclozolin and p, p’-DDE and translocated to the nucleus, but
failed to form nuclear foci, while BPA failed to induce the nuclear transportation of
hAR (Lee et al. 2003; Roy et al. 2001; Tomura et al. 2001). In addition, studies
investigating androgen receptor nuclear transportation with antiandrogens are
relatively rare compared to those with pharmaceutical antiandrogens such as casodex
and flutamide. This lack of knowledge and inconsistency in previously obtained
results has made it difficult to form a general hypothesis. However, the following
hypothesis is made: the presence of a single antiandrogen induces nuclear
translocation of zfAR, but fails to induce the formation of nuclear foci; while when
the antiandrogen is present together with 11-ketotestosterone, nuclear foci formation
is inhibited in a dose-response manner. Another aim of this project is to investigate
the mixture effects of the four antiandrogens in the presence 11-ketotestosterone. A
dose-addition response is expected because of previous compelling evidence.

27
28
Materials and methods
Construction of expression vector

Expression plasmid pEYFP-zfAR was constructed by Anne Jørgensen (unpublished).

To help the reader of this thesis, the construction of the expression plasmid is
described briefly. Total cDNA was prepared from zebrafish tissue (Jorgensen et al.
2007). To construct the expression vector that express YFP-zfAR fusion protein, full
length zfar was first amplified by PCR from the total cDNA preparation. At the same
time, BamHI sites were induced flanking the PCR fragment by primers listed in Table
1. PCR was performed according to the protocol for Expand Long Template PCR
system (Roche). Amplification conditions were: 94°C for 30 s (94°C for 15 s, 65°C
for 30 s, and 72°C for 4 min) for 40 cycles followed by a 7 min extension period at
72°C. PCR product was gel purified using the High Pure PCR Product Purification
Kit (Roche). The AR PCR product and the pEYFP-C1 vector were digested with
BamHI, according to manufacturer’s instructions (Fermentas). pEYFP-C1 vector was
de-phosphorylated using Shrimp Alkaline Phosphatase (SAP) according to
manufacturers instructions (USB Corporation). Inserts and pEYFP-C1 vector were
ethanol precipitated and the AR gene was ligated into pEYFP-C1 with T4 ligase
(Invitrogen) and were fully sequenced. The resulting fusion protein has YFP at the N
terminus, which allows convenient visualization by fluorescence microscopy.

Table 1 Oligonucleotide primers used for cloning the Danio rerio androgen receptor into
pEYFP-C1 expression vector.

Primer Amplicon Sense primer 5´-3´ Antisense primer 5´-3´

EYFP-AR 2607 nt ggatccatggaggttcgggtcggg ggatcctcatttgtggaacaggattgg

Restriction sites are underlined.

Mutagenesis

Five zfAR mutants (Table 2) were generated using pEYFP-zfAR expression vector as
template. Mutagenesis was performed using site-directed mutagenesis with pfuultra

29
high fidelity DNA polymerase (Stratagene). Primers (Table 2) were designed by The
QuikChange® Primer Design Program on Stratagene’s website
(http://www.stratagene.com/tradeshows/feature.aspx?fpId=118). Standard PCR
conditions were 1 cycle of 95oC 2 min; 30 cycles of 95oC 30 sec, 65oC 30 sec, 72oC 6
min; and ended with 72oC 7 min. Mutant pEYFP-zfARR582K583L584R585K586A was
generated with a slightly different PCR condition with 1 cycle of 95oC 2 min; 30
cycles of 95oC 30 sec, 70oC 30 sec, 72oC 6 min, and ended with 72oC 7 min. The
PCR products were digested with DpnI for 1 hour and then transformed into Top 10
competent cells by electroporation and selected on kanamycin containing LB plates.
All zfAR mutants were confirmed by fully sequencing.

Table 2 Plasmids and primers in site-directed mutagenesis

Plasmids Primers Primer Sequence

5'-aactgcccatcctgccgtctgggggcgtgttttga
Armut1_S
ggttggaatgac-3'
pEYFP-zfARK570G,K571A
5'-gtcattccaacctcaaaacacgcccccagacggca
Armut1_AS
ggatgggcagtt-3'

5'-ggttggaatgaccctgggagccattggacagatga
Armut2_S
aaggtcc -3'
pEYFP-zfARdel582-586
5'-ggacctttcatctgtccaatggctcccagggtcat
Armut2_AS
tccaacc -3'

5'-ttttgaggttggaatgaccctgggagccgccgcgg
Armut3_S
cggcggcgattggacagatgaaaggtccggatgag-3'
pEYFP-zfARR582K583L584R585K586A
5'-ctcatccggacctttcatctgtccaatcgccgccg
Armut3_AS
ccgcggcggctcccagggtcattccaacctcaaaa-3'

5'-ccctgggagcccgcatgctgatgatgattggacag
Armut4_S
atgaa-3'
pEYFP-zfAR K583R585K586M
5'-ttcatctgtccaatcatcatcagcatgcgggctcc
Armut4_AS
caggg-3'

pEYFP-zfARK570G,K571A,del582-586*

* Plasmid pEYFP-ARK570G,K571A,del582-586 is a double mutant and it was constructed through two

steps: plasmid pEYFP-ARK570G,K571A was first constructed using primers Armut1_S and
Armut1_AS, and then it was used as template and further mutation was induced using primer
Armut2_S and Armut2_AS.

30
Cell culture

ZfAR subcellular distribution was investigated in Hela cells and ZFL cells. Hela cells
were grown in DMEM supplemented with 5% FBS and 0.1% penicillin/ streptomycin
at 37oC, with 5% CO2 in 75 cm2 cell culture flasks. Cells were subcultured every 3-5
days. For this purpose, they were first rinsed with 5ml PBS (Gibco) and then detached
with 1ml of trypsin for 3-5 min at 37oC. Detachment was observed under a
microscopy. Trypsination was stopped by addition of 5ml of 5% FBS containing
medium. Cells were then split into two new 75 cm2 cell culture flasks.

Zebrafish liver cells ZFL (ATCC CRL- 2643) were grown in 50% Leibovitz’s L-15
medium with 2mM L-glutamine (Gibco 11415), 35% Dulbecco’s modified Eagle’s
medium with 4.5 g/l glucose and 4mM L-glutamine (Gibco 52100), 15% Ham’s F12
with 1mM L-glutamine (Gibco 21700), (all without sodium bicarbonate), and
supplemented with 0.15 g/l sodium bicarbonate, 15mM HEPES, 0.01 mg/ml insulin
(Sigma I-1882), 50ng/ml EGF (unfiltered, Sigma), and 5% heat-inactivated fetal
bovine serum. Growth media were freshly mixed every time before use. To prepare
the medium, DMEM and Hams’ F12 were first dissolved in 0.8-0.9x volume of
distilled water. The pH of the medium was adjusted to 7.4 with 5 M NaOH, and water
was added to achieve the final (1×) volume. Then the medium was filtered through
0.2-µm filters. Insulin stock solution (10 mg/ml) was prepared in acidified water (100
µl of glacial acetic acid in 10 ml of sterile water). Cells were grown at 28oC, with
100% air in 75 cm2 cell culture flasks. Cells were subcultured every 5-7 days. For this
purpose, they were first rinsed with 5ml PBS (Gibco) and then detached with 1ml of
trypsin for 5-8 min at room temperature. Detachment was observed under a
microscopy. Trypsination was stopped by addition of 5 ml of 10% FBS containing
medium. Cells were centrifuged for 10 min at 200 g. Cell pellets were resuspended in
serum-free medium and split into two new 75 cm2 cell culture flasks containing
serum-free medium. Cells were allowed to settle and attach to the bottom by
incubating at 28oC for 30 min, and then 5% FBS was added.

31
Transfection and microscopy

2x105 cells per well were seeded in a 24-wells plate 24h before transfection. At
90-95% confluence, the cells were transfected with 0.8µg DNA/well. Transient
transfections of Hela cells and ZFL cells with pEYFP-zfAR and its mutant derivatives
were performed with Lipofactamine 2000 (Life Technologies Inc., USA) according to
manufacture’s recommendation. For the purpose of observing nuclear translocation of
zfAR induced by 11-KT and the effects of antiandrogens, the natural ligand 11-KT
and/or test chemicals were added at the time of transfection, and fluorescence was
observed with a Leica DMIRB fluorescence microscope after 24 hours and/or 48
hours. The cells were classified into five categories according to a scoring system
widely used to examine subcellular localization of steroid receptors (Jenster et al.
1993; Poukka et al. 2000; Sackey et al. 1996): N, dominantly nuclear fluorescence;
N>C, nuclear fluorescence exceeding cytoplasmic fluorescence; N=C, equal nuclear
and cytoplasmic fluorescence; N<C, cytoplasmic fluorescence exceeding nuclear
fluorescence; C, exclusive cytoplasmic fluorescence (representative cells for each
category are shown in figure 2). In this project, the categories were assigned numbers
representing the percent of fluorescence in the nucleus: N, 95%; N>C, 75%; N=C,
50%; N<C, 25%; C, 0%, meaning if a cell has been rated as N, then this cell counts in
the calculation of nuclear translocation percentage (see below) as 95%, if as N>C, it
counts as 75% nuclear fluorescence, and so on. For each sample, 100-200 positive
cells were rated. The percentage of cells that fall in each category was noted as N%,
(N>C)%, etc. To get an overall view of how much protein was in the nucleus, the
following way was used to calculate the “nuclear translocation percentage”: nuclear
translocation percentage = N% x 95%+ (N>C)% x 75% + (N=C)% x 50% +(N<C)%
x 25% + C% x 0%.

32
Figure 2. According to the localization of fluorescence, cells were classified into five
categories according to a scoring system widely used to examine subcellular localization
of steroid receptors (Jenster et al. 1993; Poukka et al. 2000; Sackey et al. 1996): N,
dominantly nuclear fluorescence; N>C, nuclear fluorescence exceeding cytoplasmic
fluorescence; N=C, equal nuclear and cytoplasmic fluorescence; N<C, cytoplasmic
fluorescence exceeding nuclear fluorescence; C, exclusive cytoplasmic fluorescence
Representative pictures from experiments were shown.

Chemicals

The natural androgen 11-ketotestosterone (cat. No: K8250) and the test chemicals
bisphenol A (cat. No: 239658), p,p’-DDE (cat. No: PS696), vinclozolin (cat. No:
45705) and fenitrothion (cat. No: 442592) were all purchased from Sigma-Aldrich.
Since all test compounds are hydrophobic substance, each compound was dissolved
in DMSO so that all stock solutions were 1000 times concentrated compared to the
final test concentrations and the final concentration of DMSO in cell culture medium
does not exceed 0.1%. For all experiments that need mixture of chemicals, master
mixtures of chemicals were made so that each ingredient was 1000 times
concentrated compared to the test concentration.

33
Cytotoxicity assessment

MTT assay for quantifications of Hela cells and ZFL cells was established as follows:
1,250, 2,500, 5,000, 7,500, and 10.000 cells/well for Hela cells and for ZFL cells,
7,500, 10,000, 25,000, 75,000 and 100,000 cells/well were plated in 96 well plate in
200 µl phenol red free growth medium (or normal growth medium for ZFL cells) and
allowed to settle. Three replicates were setup for each cell density. At the same time,
wells containing only medium without cells were prepared and used later for
background correction. After 24 h incubation the medium was removed by gentle
aspiration. 100 µl phenol red free DMEM plus 10 µl 12 mM MTT stock solution were
premixed and immediately added into each well. After incubation with MTT for 4h at
37℃, 90µl medium was removed and 50µl of DMSO was added into each well to
dissolve formazan. The plate was incubated at 37 ℃ (or 28 ℃ in the case of ZFL cells)
for 15 min before absorbance at 540 nm was measured by a Bio-Tex Synergy
microplate reader. Absorbance in wells with cells was corrected by subtracting the
medium blank.

The procedures of MTT assays used for cytotoxicity tests were based on the above
protocol. 7,500 cells/well (or 50,000cells/well) of Hela cells (or ZFL cells) were
plated in 96 well plate. Also, wells with 0.1% DMSO were set up to test the solvent
cytotoxicity, and wells with only cells and medium were set up as control. Cells were
exposed to test chemicals for 24 hours before MTT was added. After corrected
absorbance in wells with cells by subtracting the medium blank, the toxicity of the
compound was calculated as absorbance with test compounds / absorbance without
compound *100%.

Studies of antiandrogenic chemicals

The subcellular localization of zebrafish androgen receptor in the presence of the

antiandrogenic chemicals bisphenol A, p-p’-DDE, vinclozolin, and fenitrothion was
studied by observing the localization of fluorescence in Hela cells transfected with
pEYFP-zfAR as described previously. The nuclear translocation percentages were
34
calculated and used for comparison between each sample. In all experiments, a
solvent control with 0.1% DMSO was always included and used as background
translocation. Each experiment was repeated at least once.

The nuclear translocation of zfAR in Hela cells was first studied in the presence of
each chemical at concentrations of 10-9M, 10-8M, 10-7M, 10-6M, and 10-5M without
11-ketotestosterone. Then the subcellular locations of zfAR were investigated in the
presence of both 6 nM 11-ketotestosterone and one anti-androgenic chemical. Tested
concentrations for each compound were 10-9M, 10-8M, 10-7M, 5x10-7M, 10-6M,
5x10-6M, and 10-5M. Meanwhile, one sample with only 6 nM 11-KT was setup as a
positive control. Dose-response curves were generated for each test chemical and
used for evaluating the potency of each chemical. Nuclear translocation percentages
for samples that had been treated with both 6 nM 11-KT and antiandrogen were
compared with the control (only 6nM 11-KT), and the lowest concentrations which
caused statistically significant difference (student’s t-test two-tailed, P=0.05) were
called the Lowest Observed Effect concentration (LOEC), and the concentration
below the LOEC are called No Observed Effect Concentration (NOEC).

To study the mixture effects of these four environmental chemicals, two master
mixtures with a ratio of 2:2:1:1 of BPA: p, p’-DDE: vinclozolin: fenitrothion were
made. The concentration of each ingredient in both mixtures was 1000 times higher
than the tested concentration, so that the final test concentration for mixture 1 (Mix1)
would be 5x10-7M BPA, 5x10-7M p,p’DDE, 10-7M vinclozolin, 10-7M fenitrothion,
and 6nM 11-KT, and for mixture 2 (Mix 2) would be 10-6M BPA, 10-6M p,p’DDE,
5x10-7M vinclozolin, and 5x10-7M fenitrothion and 6nM 11-KT. Meanwhile, one
sample with only 6 nM 11-KT was setup as a positive control. The inhibition of
nuclear translocation induced by 11-KT was calculated as: Inhibition percent =
(nuclear translocation percentage with 6 nM-KT – nuclear translocation percentage
with both 6nM 11-KT and the antiandrogenic chemical)/ nuclear translocation
percentage with 6nM 11-KT x 100%.

35
36
Results
Construction of Expression Vector pEYFP-zfAR

In order to study the subcellular location of zfAR, a fusion protein with yellow
fluorescence protein (YFP) and zfAR was constructed. This construction was fully
sequenced, and the resulting protein sequence of zfAR was aligned with the protein
sequence from NCBI (NP_001076592) by the AlignX function in Vector NTI
Advance 10 software (Invitrogen). Three amino acids were found to be different
(figure 3): P4R, G54D, and D682G (amino acids written in front of numbers are from
NP_001076592, numbers correspond to the position of the amino acid, amino acids
written after numbers are from pEYFP-zfAR).

Figure 3. Alignment analyses between protein sequence of zfAR in the fusion protein
expressing vector pEYFP-zfAR and the protein sequence from NCBI (NP_001076592) by
Vector NTI Advance 10 (Invitrogen). Differences between the two protein sequence are
circled.

In order to determine the degree of conservation of the three amino acids, a second
alignment was conducted with androgen receptors from Atlantic croaker, Japanese
medaka, Gold fish, Japanese eel and Rainbow trout. In figure 4, P4R and G54D from

37
zfAR seems to locate in a hypervariable region of the protein, while the aspartic acid
in D682G locates in a highly conserved region and is itself conserved among all
compared species (figure 4). This high conservation might indicate a functional
importance of this aspartic acid. Therefore, the G682 in pEYFP-zfAR was changed
into an aspartic acid (D) by site-directed mutagenesis. The resulting plasmid is used in
all the experiments in this project.

ZfAR Subcellular Localization

In order to study the subcellular location of zfAR with and without the presence of the
natural ligand 11-KT, the fusion protein YFP-zfAR was expressed in human cervical
cancer cell line Hela and zebrafish liver cell line ZFL.

First, subcellular localizations of YFP and YFP-zfAR with/without the presence of

androgen were studied in Hela cells. YFP showed even distribution between
cytoplasmic and nuclear compartments regardless of the addition of androgen (figure
5). In contrast, YFP-zfAR chimeras located mostly in cytoplasm in some cells or
evenly in cytoplasm and nuclei in other cells without the adding of ligand, and
exposure to both 10nM dihydrotestosterone (DHT) and 10nM 11-ketotestosterone
(11-KT) resulted in efficient transfer of YFP-zfAR from cytoplasm into nuclei (figure
5). Through this pilot experiment, cell specific distribution of YFP-zfAR was
observed, therefore a systematic approach was used in following experiments in
which at least 100 cells were analyzed. Further, because the androgen receptor
investigated in this project is from zebrafish, I would have preferred to analyze the
subcellular location of the protein in the zebrafish cell line ZFL. However, due to low
transfection efficiency in these cells, a systematic investigation was not performed,
while the high transfection efficiency of Hela cells permitted the analyses. 11-KT was
used as the test androgen because 11-KT is the major natural androgen in fish.

38
Figure 4. Alignment analyses of androgen receptors from zebrafish, Atlantic croaker,
Japanese medaka, Gold fish, Japanese eel and Rainbow trout by Vector NTI Advance
10 (Invitrogen). P4R, G54D, and D682G in zebrafish AR (amino acids written in front of
numbers are from NP_001076592, numbers stand for the position of the amino acid,
amino acids written after numbers are from pEYFP-zfAR.) are circled. D682G is
conserved among all compared species

39
Figure 5. Subcellular localization of YFP and YFP-zfAR in Hela cells in response to
androgens. A-B: Hela cells expressing YFP. A: no ligand added; B: 10 nM 11-KT added. C-D:
Hela cells expressing YFP-zfAR. C: no ligand added; D: 10nM dihydrotestosterone (DHT)
added; E: 10nM of 11-ketotestosterone (11-KT) added.

In the systematic analysis of translocation, the localization of the YFP-zfAR chimeras

in Hela cells were classified into five categories (see figure 2 in material and methods)
and an index called nuclear translocation percentage was introduced to indicate the
overall percentage of protein in nucleus (see material and methods). In the absence of
11-KT, zfAR distributed evenly between cytoplasm and nucleus (N=C) in almost 60%
of the observed Hela cells, and in around 20% of the cells more cytoplasmic than
nuclear (N<C) fluorescence was observed (figure 6). The overall nuclear translocation
percent was 47.04% (table 3). Exposure to 1nM 11-KT resulted in a transfer of
72.68% of zfAR from cytoplasm into nucleus (table 3). Around 65% of the cells
transfected with zfAR showed nuclear exceeding cytoplasm (N>C) or exclusively
nuclear (N) fluorescence, and around 33% of the cells showed an even distribution
between nucleus and cytoplasm (figure 6). As the concentration of 11-KT increased to
10nM, a more complete transfer of zfAR into nucleus was observed. 90% of the cells

40
showed predominantly nuclear (N) or nuclear exceeding cytoplasmic (N>C)
fluorescence (figure 6), which results in an overall nuclear translocation of zfAR to
90.03% (table 3).

Figure 6. Nuclear translocation of YFP-zfAR in response to 11-KT. Hela cells were transfected
with plasmid pEYFP-AR At the time of transfection, 1nM or 10nM of the natural ligand
11-ketotestosterone was added. At 24 hours after transfection, localization of fluorescence was
analyzed. For each sample, 100-200 cells were observed and placed in to five categories: N,
exclusively nuclear fluorescence; N>C, nuclear fluorescence exceeding cytoplasmic
fluorescence; N=C, equal nuclear and cytoplasmic fluorescence; N<C, cytoplasmic
fluorescence exceeding nuclear fluorescence; C, exclusively cytoplasmic fluorescence. Each
experiment was repeated twice. Representative pictures taken for samples without addition of
ligand, with 1 nM 11-ketotestosterone and with 10 nM 11-ketotestosterone are shown.

The higher nuclear translocation percentage induced by 10nM 11-KT compared to

1nM 11-KT indicated a concentration-dependent transfer of zfAR from cytoplasm
into nucleus. Therefore a dose-response study was performed to investigate the
relationship between the concentration of the androgen and the import of zfAR.
Furthermore, in order to establish whether the nuclear translocation is cell cycle
dependent, 48 hour samples were also analyzed in the study. A clear dose-response
fashion was observed for the 11-KT induced nuclear translocation of zfAR both at 24
and 48 hours after transfection (figure 7). The curves derived from the two time
points do not deviate, and therefore in all the following experiments localization of
zfAR was only analyzed at 24 hours.

41
Figure 7. Dose-response curves for zfAR nuclear translocation induced by
11-ketotestosterone at 24 hours and 48 hours after addition of ligand. The curves from the two
time points do not deviate from each other (Two-sample Kolmogorov-smirnov test, P>0.1).

In summary, zfAR was imported to nucleus from cytoplasm in response to addition of

11-KT and DHT. This androgen regulated distribution of zfAR in the cells showed a
dose-response relationship that did not change when measured at 24 hours and 48
hours.

Bioinformatics search for Nuclear Localization Sequence (NLS) within zfAR

The androgen-dependent nuclear translocation of zfAR led to a search for a nuclear

localization sequence (NLS) within zfAR. To determine the NLS, the amino acid
sequence of zfAR was aligned with the human androgen receptor (hAR) amino acid
sequence using ClustalW. The NLS sequence in hAR has previously been identified
(Zhou et al. 1994).

Nuclear localisation sequence

Homo sapiens AR (604-620) RKCYEAGMTLGARKLKK

Danio rerio AR (570-586) KKCFEVGMTLGARKLRK
:**:*.*********:*

*: identical residues between the ARs, : highly conserved residues, . semi-conserved amino
acids.

42
A bipartite type of NLS (KKCFEVGMTLGARKLRK ) ranging from 570 aa to 586
aa overlapping the DNA binding domain and the hinge region was predicted in zfAR.
The first two basic amino acid (Lys-Lys) and the last five basic amino acids
(Arg-Lys-Leu-Arg-Lys) separated by 10 amino acids make up the most characteristic
feature of a bipartite type nuclear localization sequence.

Mutagenesis and the subcellular localization of the zfAR mutants

To determine if the predicted NLS is sufficient for nuclear import of zfAR,

site-directed mutagenesis was conducted to substitute specific amino acids in the
sequence. Five mutants were made targeting the first two amino acids and the last five
amino acids of the predicted NLS (KK…RKLRK ) (figure 8). All five plasmids
containing the mutated YFP-zfAR gene were transfected into Hela cells and ZFL cells,
and the subcellular localization of the fusion proteins were observed with and without
11-KT.

Figure 8. Mutagenesis targeting of the bipartite NLS predicted in zebrafish androgen receptor
and the structure of relevant amino acids. Changed or deleted amino acids are underlined or
are shown as”-“. DBD: DNA Binding Domain. LBD: Ligand Binding Domain

43
In mutant 1 (zfARK570G, K571A) the first two basic lysines were substituted with
two small neutral amino acids (glycine and alanine) in order to determine the
significance of the charge of the two lysines. In the absence of 11-KT, 8.60% of the
mutated chimeras were located in nuclei and the representative distribution observed
in almost 80% of the cells was predominantly cytoplasmic (C) (figure 9 and table 3).
Exposure to 1nM and 10nM 11-KT resulted in increasing nuclear translocation
percentage to 28.69% and 48.35%, respectively (table 3). This indicates a
concentration-dependent partial nuclear transport of the mutated protein, and probably
a second nuclear import signal. Mutant 2 (zfARdel582-586) was designed to
investigate the significance of the second cluster of basic amino acids. Nuclear import
of mutant 2 was very low compared to that of the wild type protein (figure 9), even
though slight increases of nuclear translocation percentage to 25.58% and 27.98%
were observed when 1nM and 10nM 11-KT was added respectively (table 3). In
contrast to the first mutant whose nuclear translocation percentage increased almost
20% as the ligand concentration increased from 1nM to 10nM, the nuclear
translocation percentage for mutant 2 only increased 2%, indicating a disturbed ligand
binding function that probably resulted from the conformational change caused by the
deletion. To minimize the possible destruction of protein folding caused by the
deletion, mutant 3 (zfARR582K583L584R585K586A) was constructed, in which the
introduced five alanines differ from the original amino acids both in size and charge.
In addition, in order to eliminate the size difference, methionine was used in mutant 4
(zfARK583R585K586M). Compared to the first two mutants, less nuclear
translocation was observed for both mutant 3 and 4 when 1nM 11-KT was present
(figure 9 and table 3), and 12% and 13% increase in nuclear translocation percentage
were observed when the ligand concentration increased to 10nM for mutant 3 and 4
respectively. The last mutant 5 (zfARK570G, K571A, del582-586) is a double mutant
with changes in both clusters of basic amino acid in the predicted NLS. Nuclear
import of the mutant was severely disrupted (figure 9), but a nuclear translocation of
23.39% was still observed when 10nM 11-KT was added (table 3).

44
Another observation with the mutated zfAR is the formation of fluorescence foci
(figure 9), indicating receptor cluster, in the presence of 10nM 11-KT, especially in
cells transfected with mutant 1 (zfARK570G, K571A) and mutant 5 (zfARK570G,
K571A, del582-586). Receptor clusters were not observed in cells transfected with the
wild type zfAR regardless of the presence of 11-KT or in cell transfected with
mutated zfAR without the addition of 11-KT.

Table 3. Nuclear translocation of zfAR wild type and its mutants

ARwt Mutant 1 Mutant 2 Mutant 3 Mutant 4 Mutant 5

47.04% 8.60% 10.52% 7.18% 4.71% 6..90%

±8.25% ±4.64% ±0.70% ±1.69% ±0.05% ±1.89%

Without ligand

72.68% 28.69% 25.58% 11.32% 6..57% 8..97%

±4.35% ±7.50% ±5.90% ±4.34% ±1.54% ±2.03%

1nM 11-KT

90.03% 48.35% 27.98% 23.62% 19.98% 23.39%

±2.27% ±8.07% ±3.92% ±6.98% ±2.70% ±1.93%

10nM 11-KT

45
Figure 9. Mutagenesis of the bipartite NLS interfere with nuclear translocation of YFP-zfAR.
Hela cells were transfected with plasmid pEYFP-AR or the mutants: pEYFP-ARK570G,K571A,,
pEYFP-ARdel582-586,pEYFP-ARR582K583L584R585K586A,pEYFP-ARK583R585K586M, pEYFP-ARK570G,K571A,
del582-586. At the time of transfection, 1nM or 10nM of the natural ligand 11-ketotestosterone was
added. At 24 hours after transfection, localization of fluorescence was analyzed. For each
sample, 100-200 cells were observed and placed in to five categories: N, exclusively nuclear
fluorescence; N>C, nuclear fluorescence exceeding cytoplasmic fluorescence; N=C, equal
nuclear and cytoplasmic fluorescence; N<C, cytoplasmic fluorescence exceeding nuclear
fluorescence; C, exclusively cytoplasmic fluorescence. Each experiment was repeated twice.
Representative pictures taken for samples without adding ligand, with 1nM
11-ketotestosterone and 10nM 11-ketotestosterone are shown.

46
Preliminary investigations of the subcellular distribution of zfAR and its five mutants
were performed with ZFL cells. By observing individual ZFL cells instead of at least
100 cells in the case of Hela cells, a subjective impression was that the location of
zfAR in the zebrafish cell line is quite similar to that in Hela cells, and that the
mutations in NLS destroyed the nuclear translocation of zfAR (figure 10). Picture for
the double mutant 5 (zfARK570G, K571A, del582-586) is not shown due to low
picture quality.

Figure 10. Mutagenesis of the bipartite NLS interfere with nuclear translocation of YFP-zfAR in
ZFL cells. ZFL cells were transfected with plasmid pEYFP-AR or its mutant derivatives:
pEYFP-ARK570G,K571A,,pEYFP-ARdel582-586,pEYFP-ARR582K583L584R585K586A,pEYFP-ARK583R585K586M.
At the time of transfection, 1nM or 10nM of the natural ligand 11-ketotestosterone was added.

47
In summary, mutations in NLS, either in the first or the second cluster of the basic
amino acid, severely damaged the 11-KT dependent nuclear translocation of zfAR. A
higher nuclear translocation of YFP-zfARK570G, K571A was observed upon exposure to
10nM 11-KT compared to the other mutants.

Quantification of Hela cells and ZFL cells and cytotoxicity test using MTT assay

Before studying the effects of BPA, p, p’-DDE, vinclozolin and fenitrothion on the
subcellular localization of zfAR, it is necessary to investigate the possible toxicity
they may have in Hela cells and ZFL cells. For this purpose, the MTT assay was
chosen. The principle of this assay is based on 3-(4,5-dimethylthiazol-2-yl)-2,5-
diphenyltetrazolium bromide(MTT) reduction and it is widely used for quantitatively
assessing cell viability and proliferation. It is generally believed that MTT reduction
into formazan is dependent on cellular redox activity in living cells, and is therefore
an indicator for mitochondrial function. MTT itself is a pale yellow tetrazolium salt
which can be reduced by the mitochondrial dehydrogenase enzyme from viable cells
to a dark blue formazan crystal. Formazan is impermeable to cell membranes and thus
accumulate within healthy cells. Addition of a detergent results in solubilisation of the
cells and release of the crystals which are soluble. The number of surviving cells is
directly proportional to the level of the formazan product created. The color can then
be quantified using a colorimetric assay.

To establish this assay with Hela cells and ZFL cells, the first step was to study
influencing parameters, such as cell density. Since ZFL cells need to be grown at
28℃, MTT incubation was conducted at 28℃ instead of 37℃ as suggested by the
manufacturer. A linear relationship between cell number and absorption was
established for up to 10,000 cells/well (R=0.957) for Hela cells (figure 11a) and
100,000 cells/well (R=0.943) for ZFL cells (figure 11b). Since cell density of 7,500
cells/well for Hela cells and 50,000 cells/well for ZFL cells resulted in reasonably
strong signals, those seeding densities were used for further investigation of toxicity
of the four anti-androgens in Hela cells and ZFL cells.

48
Figure 11. Quantitation of Hela cells (a) and ZFL (b) cells using MTT assay. Cells were
incubated with MTT for 4h in a microplate. Absorbance measurements at 540 nm were made
using a microplate reader. Each data point represents the mean value of samples in triplicate.

No significant toxicity of the four antiandrogens was observed for Hela cells and ZFL
cells from concentration 10-8M to 10-5M (figure 12).

Figure 12. MTT assay for testing cytotoxicity in Hela cells (a) and ZFL cells (b) of BPA,
-8 -7 -6 -5
p,p’-DDE, vinclozolin and fenitrothion at concentration 10 M, 10 M, 10 M, and 10 M. MTT
assay were performed after cells had been exposed to chemical for 24 hours. Absorbance of
samples treated with chemicals was compared with that of untreated samples. Inhibitions of
absorbance were calculated. Each data point is average of 3 replicates.

Subcellular localization of zfAR in the presence of androgen disruptors

In order to establish whether the four tested antiandrogens, BPA, p, p’-DDE,

vinclozolin and fenitrothion alone induce nuclear translocation of zfAR, Hela cells
and ZFL cells were transiently transfected with pEYFP-zfAR, and the subcellular

49
fluorescence distributions were observed after treating cells with each antiandrogen
over a concentration range from 10-9 to 10-5M. The distribution of zfAR in both Hela
cells (figure 14 A-H) and ZFL cells (figure 14 a-c) did not deviate from the pattern
observed when no ligand was added, except that in Hela cells 10-9M BPA induced
nuclear translocation (almost 80%) (figure 13 and figure 14 A), but not at other higher
tested concentrations (figure 13 and figure 14 E). 10-5M vinclozolin also slightly
induced nuclear translocation of zfAR (figure 13 and figure 14 G), while 10-5M
fenitrothion seemed to slightly reduce AR translocation into the nucleus compared to
the control when no ligand was added (figure 13 and figure 14 H).

-9
Figure 13. Localization of zfAR in the presence of antiandrogens at concentration 10 M,
-8 -7 -6 -5
10 M, 10 M, 10 M, and 10 M in Hela cells. Concentration 0 represents the control sample
without adding any ligand. BPA: bisphenol A; V: vinclozolin; F fenitrothion. Each data point is
the average of 2 replicates.

50
Figure 14. Subcellular localization of YFP-zfAR in the presence of single antiandrogen in Hela
cells and ZFL cells. At the time of transfection, BPA, p,p’-DDE, vinclozolin and fenitrothion
-9 -5
were add into the media at concentration from 10 M to 10 M. Pictures were taken at 24 hours
-9
after transfection. A-D: Hela cells treated with 10 M of BPA, p,p’-DDE, vinclozolin and
-5
fenitrothion; E-H: Hela cells treated with 10 M of BPA, p,p’-DDE, vinclozolin and fenitrothion;
-5
a-c: ZFL cells treated with 10 M of BPA, p,p’-DDE, and vinclozolin. Picture for ZFL cells
-5
treated with 10 M fenitrothion is missing because of low picture quality.

Subcellular localization of zfAR in the presence of antiandrogens and 11-KT

Since the antiandrogens themselves did not result in nuclear import of the zfAR, it
was interesting to investigate if the co-addition of antiandrogen would disrupt the

51
nuclear translocation of zfAR induced by 11-KT. 6nM of 11-KT was mixed with each
individual antiandrogen over a concentration range from 10-9 to 10-5M, and added to
Hela cells transiently transfected with the YFP-zfAR fusion protein. The subcellular
fluorescence was then observed. While 6nM 11-KT alone induced around 85% of
zfAR nuclear translocation as observed previously, even distribution of fluorescence
between the nuclear and cytoplasmic compartment was observed in Hela cells treated
with 6 nM 11-KT and one of the four antiandrogens at 5 µM, indicating the nuclear
import of zfAR was destroyed by the adding of around 1000 fold of antiandrogen
(figure 15). This inhibition of nuclear transport of zfAR by the antiandrogen showed a
concentration-dependent manner (figure 16). In figure 16, the lowest concentration
that cause observable effects (Lowest Observed Effect Concentrations (LOEC)) for
bisphenol A and p, p’-DDE were 10-6M, with inhibition of nuclear translocation of
14.31% and 14.48%, respectively (table 4 and 5) Vinclozolin and fenitrothion showed
significant effect on nuclear translocation of zebrafish androgen receptor at
concentration of 5x10-7M, with inhibition of 19.64% and 12.85% respectively (table 4
and 5). By comparing the inhibition percentage, vinclozolin seems to be the most
potent antiandrogen among the four (figure 16). As the concentration increase, all four
chemicals reach their maximum effects at concentration of 5x10-6M.

52
Figure 15. 11-KT dependent nuclear translocation of zebrafish androgen receptor in Hela cells
-9 -6
was disrupted by addition of antiandrogen (a) 6nM 11-KT together with 10 M, 10 M and
-6 -9 -6 -6
5x10 M bisphenol A (BPA); (b) 6nM 11-KT together with 10 M, 10 M and 5x10 M p,p;-DDE;
-9 -7 -6 -9
(c) 6nM 11-KT plus 10 M, 5x10 M and 5x10 M vinclozolin (V); (d) 6nM 11-KT plus 10 M,
-7 -6
5x10 M and 5x10 M fenitrothion (F). Hela cells were transfected with plasmid pEYFP-AR.
-9 -8 -7 -7 -6 -6
Androgen disrupting chemicals (concentration at 10 M, 10 M, 10 M, 5x10 M, 10 M, 5x10
-5
M, and 10 M) and 6nM 11-KT were added at the same time as transfection. After 18-24 hours,
fluorescence was observed under a fluorescence microscope. Representative pictures from
-9
samples with androgen disrupting chemicals at the lowest concentration (10 M), at the lowest
concentrations that caused statistically significant difference (LOEC, P=0.05) and at the
-6
concentration that caused maximum effect (5x 10 M) are shown.

53
Figure 16. 6nM 11-KT induced zfAR nuclear translocation is inhibited by single antiandrogen
-9 -8 -7
in a concentration dependent manner (tested concentrations were 10 M, 10 M, 10 M,
-7 -6 -6 -5
5x10 M, 10 M, 5x10 M, and 10 M) Concentration 0 represents the control with only 6nM
11-KT. Nuclear translocation percentages for samples treated with antiandrogen were
compared with the control. Each data point is average of 2-3 replicates.

Mixture effects

To test if the prevailing dose addition theory in predicting combinational effects of

chemicals can be applied in our system, the nuclear translocation of zfAR in the
presence of mixtures of BPA, p, p’-DDE, vinclozolin and fenitrothion with mixing
ratio 2:2:1:1 was investigated in two concentrations together with 6nM 11-KT. The
ratio and the concentration of the mixtures were designed based on single chemical
effects from the previous experiments (figure 16). The first mixture (Mix 1) consists
of 5x10-7M BPA, 5x10-7M p, p’-DDE, 10-7M vinclozolin, and 10-7M fenitrothion,
corresponding to NOEC for the individual compound. In the presence of this mixture,
the nuclear translocation of zebrafish androgen receptor induced by 6nM 11-KT was
not significantly inhibited, even though a 7.23% decrease of nuclear translocation of
zfAR was observed. This inhibition percentage is quite similar to the case observed
with each single chemical. The second mixture (Mix 2) consists of 10-6M BPA, 10-6M
p, p’-DDE, 5x10-7M vinclozolin, and 5x10-7M fenitrothion, corresponding to LOEC
54
for the individual compound. A 22.38% inhibition of 11-KT induced nuclear
translocation was observed. Compared to the inhibition by each single chemical, the
mixture effect is larger but not enough to conclude cumulative effects from its single
component.

#
Table 4 . Inhibition of nuclear translocation by single chemicals and mixture 1

6nM BPA p,p’-DDE Vinclozolin Fenitrothion

-7 -7 -7
Mix 1*
11-KT 5x10 M 5x10 M 10 M 10-7M

Nuclear
87.54% 83.42% 81.29% 84.19% 84.20% 81.21%
Translocation
±5.79% ±1.83% ±0.47% ±2.57% ±2.69% ±2.68%
percent

Inhibition percent 0% 4.71% 7.14% 3.82% 3.82% 7.23%

* Mix1: mixture of 5x10-7M BPA, 5x10-7M p, p’-DDE, 10-7M vinclozolin, and 10-7M fenitrothion.

#
Table 5 . Inhibition of nuclear translocation by single chemicals and mixture 2

6nM BPA p, p’-DDE Vinclozolin Fenitrothion

Mix 2*
11-KT 10-6M 10-6M 5x10-7M 5x10-7M

Nuclear ++
87.54% 75.01% 74.86% 70.34% 76.28% 67.94%
Translocation
±5.79% ±5.66% ±5.10% ±5.21% ±4.22% ±1.85%
percent

Inhibition percent 0% 14.31% 14.48% 19.64% 12.85% 22.38%

-6 -6 -7 -7
* Mix2: mixture of 10 M BPA, 10 M p, p’-DDE, 5x10 M vinclozolin, and 5x10 M fenitrothion
++ : indicates statistically significance (P< 0,01, student’s t-test, two-tailed)

#
Results of inhibition caused by single chemical are average of 2-3 repeats, whereas results of

mixture effects are based on 2 repeats.

55
56
Discussion
Subcellular localization of zfAR upon exposure to 11-KT

The use of YFP tagged zfAR allows image analysis of living cells to determine the
subcellular localization of zfAR in response to ligand exposure. The results showed
that in Hela cells 40% to 50% of the unliganded zfAR population located in nuclei
while the rest located in cytoplasm. This pattern is different from that inferred from
the hAR model in which unliganded human androgen receptor resides largely in the
cytoplasm, complexed with heat shock proteins (Kumar et al. 2006). The presence of
unliganded zfAR in both nucleus and cytoplasm might be correct, but it may also be
the result of experimental variances such as overexpression of the protein in
transiently transfected cells. Contradictory conclusions have also been presented for
human AR, even though the consensus has been reached after applying sophisticated
techniques and refinement of the previous studies. Jenster et al (1993) studied the
subcellular localization of human AR by immunostaining within different cell lines
where hAR was transiently expressed. A cell line specific subcellular distribution of
the unliganded receptor was reported. hAR expressed in Hela cells was predominantly
nuclear, whereas mainly cytoplasmic staining was observed when it was expressed in
the monkey kidney cells COS-1 (Jenster et al. 1993). The author attributed this cell
line specific observation to the differences in membrane-translocation efficiency and
the components involved in trapping the AR within each cell line. Hereafter,
subcellular localization of unliganded human AR have been studied in monkey kidney
cells COS-1, COS-7, prostate cancer cells PC3 and Hela cells (Georget et al. 1997;
Marcelli et al. 2006; Saporita et al. 2003; Tomura et al. 2001). hAR distributed
exclusively in cytoplasm in the absence of agonists in COS-1, COS-7 and PC3 cells,
while Hela cells expressing GFP-hAR did not show exclusively cytoplasmic
distribution of the protein (Marcelli et al. 2006). Thus, the cell line specific
distribution of unliganded hAR might also apply to zfAR. However, a
ligand-regulated nuclear export signal (NES) in human androgen receptor was later

57
identified and it is reported to be dominant over the NLS in the DNA binding domain
(Saporita et al. 2003), meaning that in the absence of agonists, the NES is sufficient to
keep hAR cytoplasmic even with a coexisting intact NLS. This makes it more difficult
to explain the nuclear import of hAR in Hela cells when androgen is absent. In
addition, another possible explanation for the observed nuclear distribution of zfAR
could be that trace amounts of steroids in the serum used in cell culturing might be
able to induce zfAR translocation since the requiring concentration of agonist is
extremely low (as low as 1nM). However, since serum is used in culturing all the cell
lines mentioned above, this explanation seems invalid. But it would still be interesting
to investigate the localization of zfAR in Hela cells grown in serum-free medium
because zfAR may be more sensitive to agonists. Further, phenol red in the cell
culture medium is a weak estrogen (Berthois et al. 1986), and since cross reaction of
estrogens with androgen receptor has been reported (Marcelli et al. 2006; Roy et al.
2001), phenol red may also contribute to the high background of nuclear fluorescence.

Upon ligand exposure, zfAR was translocated into nucleus efficiently in a

dose-dependent manner. The observed nuclear transportation of zfAR in response to
11-KT agrees well with the previous observation in human androgen receptor
(Georget et al. 1997; Jenster et al. 1993; Kumar et al. 2006; Marcelli et al. 2006;
Saitoh et al. 2002; Tyagi et al. 2000; Whitaker et al. 2004; Zhou et al. 1994).
According to the studies with hAR, nuclear import of hAR occurs within 10-30
minutes, with concomitant formation of punctuated foci in the nuclei (Georget et al.
1997; Marcelli et al. 2006). However, due to the experimental setup in this project
where 11-KT was added at the time of transfection, a real time imaging analysis of the
nuclear translocation of zfAR was not conducted. In a pilot experiment, 11-KT was
added at 24 hours after transfection, and pictures were taken at time 0, 10 min, 20min,
30 min, 45min and 1hour. Because a single cell was followed in the experiment, the
cell culture was placed under the microscope at room temperature without supply of
5% CO2. After 1 hour, even though a slight redistribution of fluorescence did occur,
no clear sign of nuclear translocation of YFP-zfAR was observed (pictures not shown).

58
The suboptimal physiological condition, i.e. low temperature, low pH, could inhibit
the normally efficient nuclear import. Nucleus foci were not observed because of the
limited quality of the pictures taken. But due to the seemingly similar androgen
signaling mediated by zfAR and hAR, zfAR is speculated to locate in special
subnuclear compartments in association with genes being expression regulated by
zfAR. By applying sophisticated cell biological and molecular approaches, a highly
dynamic picture of how hAR and its partners mediate gene expression is emerging,
and this may be highly informative for the future study of androgen signaling in
zebrafish.

Site-directed mutagenesis of the putative NLS within zfAR revealed that the import of
zfAR is mediated through a bipartite sequence that span the DNA binding and the
hinge regions. The sequence KKCFEVGMTLGARKLRK, consists of two clusters of
basic amino acids (underlined) that facilitate zfAR nuclear import. Replacement or
deletion of the amino acids in either of the two regions severely damaged the nuclear
transfer of zfAR induced by 11-KT exposure. The main reason for the failure of
nuclear import of zfAR could be that the NLS recognition and/or interaction with the
importins were damaged. In addition, mutations in the NLS may also interfere with
other critical steps in the nuclear import process. Conformational changes of the
protein resulting from amino acid changes may influence the ligand recognition of the
LBD, phosphorylation of the receptor upon ligand binding, dimerization of the
receptor or a combination of those steps. However, some certain extent of nuclear
import of zfAR still remained, especially in mutant 1 whose Lys-Lys in NLS were
replaced by Gly-Ala, while Arg-Lys-Leu-Arg-Lys remains intact. This indicates that
even though the Lys-Lys play an important role in the nuclear import process, the
Arg-Lys-Leu-Arg-Lys sequence can function independently to a certain extent. The
Lys-Lys do not seem to induce nuclear import independently because the nuclear
translocation percentage observed in mutant 4 zfARK583R585K586M was not higher than
that observed in the double mutant 5 zfARK570G, K571A, del582-586. In fact, mutant 4
zfARK583R585K586M seems to have the lowest nuclear translocation, even though less

59
conformational change is expected as the changes in amino acids in mutant 4
zfARK583R585K586M were relatively few, both regarding the number of amino acids, and
the size of amino acids. In mutant 4, the wild type NLS KK…RKLRK was changed to
KK…RMLMM, and therefore only three amino acids were changed and the sizes of
M and R/K are similar. Thus the less nuclear translocation observed with mutant 4
leads to the conclusion that failure of nuclear import is rather because of the
interference in NLS recognition and/or interaction with the import machinery than
because of significant change in the protein conformation.

Further, around 20% of the zfAR mutated in the second basic region tend to locate in
the nucleus, and two explanations are proposed based the analogy to hAR and hGR
(human glucocorticoid receptor). The first explanation is that there might be another
NLS in the LBD of zfAR. A second ligand-dependent NLS has been identified in the
LBD of GR in addition to the NLS in the DBD and the hinge region (Picard and
Yamamoto 1987). Similarly, a ligand-dependent NLS also exists in the LBD of hAR,
which is capable of inducing nuclear import in the absence of the NLS in the DBD
and hinge region (Jenster et al. 1992; Poukka et al. 2000; Saporita et al. 2003).
Because the LBD of AR is a highly conserved domain, it is likely that a similar NLS
also exist in zfAR. The second explanation could be that a possible nuclear export
signal (NES) in the LBD of zfAR was disrupted due to the mutations. A
ligand-dependent NES was identified in the LBD of hAR and in the absence of
agonist, the active NES dominate over the NLS and keep the AR in cytoplasm,
whereas ligand exposure inactivate the NES and results in nuclear import of AR
mediated by NLS (Saporita et al. 2003). Mutations in the hinge area of AR may cause
conformational change of the protein thus interfering with the repression of NES by
the agonist and leading to leakage of AR into the nucleus. A putative NES in the LBD
of zfAR has been predicted by ClustalW (figure 16), and therefore it is plausible that
the conformational changes resulting from mutations in NLS indirectly leads to
nuclear import of zfAR. This theory is also consistent with the results where the
mutants (except zfARK570G, K571A) with more significant changes in amino acids had

60
higher nuclear translocation percentage in the presence of 11-KT.

Nuclear export sequence

++ + + + + +
Homo sapiens AR(731-792) LMVFAMGWRSFTNVNSRMLYFAPDLVFNEYRMHKSRMYSQCVRMKHLSQEFVLLQVTQEEFL
Danio rerio AR(694-755) MMVFALGWRSYKNANARMLYFAPDLVFNDRRMHVSSMYEHCVQMRHLSQEFGWLQITPQEFL
:****:****:.*.*:************: *** * **.:**:*:****** **:* :***

Homo sapiens AR(793-799) CMKALLL

Danio rerio AR (756-762) CMKAMLL
****:**

Figure 17. Androgen receptor sequence alignment of the ligand bindind domain of Homo
sapiens and Danio rerio. Nuclear export sequence identified in Homo sapiens is used. The
sequence homology and similarity of NES was determined using ClustalW analysis. * identical
residues between the ARs, : highly conserved residues, . semi-conserved amino acids, +
residues that interact with the ligand.

Another observation worth noticing is the formation of receptor clusters (represented

by the green dots in the pictures) in cells expressing zfAR mutants especially mutants
with the replacement of the Lys-Lys with Gly-Ala in the first basic domain of NLS
which overlaps the DBD. This is mainly seen when 11-KT is added, particularly when
10nM 11-KT is present. The formation of intracellular punctuated foci in response to
10nM 11-KT is not very likely to be a sign of degradation of the protein since no such
foci have been observed in the cells untreated with 11-KT and it is rarely seen in cells
treated with 1nM 11-KT. Similar receptor clusters formation in response to ligand
adding has also been reported in hAR with mutations in the DBD (Jenster et al. 1993;
Marcelli et al. 2006). No clear explanation was provided in any of the articles, but an
indication of alteration of the DBD function was suggested. A future experiment
intending to investigate whether the formation of receptor cluster is associated with
cell apoptosis would be interesting.

Subcellular localization of zfAR upon exposure to ADCs

Subcellular localizations of YFP-zfAR in the presence of four antiandrogens over a

range from 10-9M to 10-5M with/without co-treatment of 6nM 11-KT were analyzed.
None of the four compounds alone induced nuclear translocation of zfAR, except that

61
10-9M BPA caused around 80% of the zfAR translocation and 10-5M vinclozolin also
showed a slight induction of nuclear translocation of zfAR. Co-treatment of cells with
the antiandrogen and 6nM 11-KT resulted in inhibition of nuclear translocation of
zfAR in a dose-response manner.

BPA is believed to be an androgen antagonist that can affect multiple steps in the
activation and function of human AR, including inhibiting the binding of native
androgens to AR, AR nuclear localization, AR interaction with coregulators, and its
subsequent transactivation (Lee et al. 2003). At concentration from 10-8M to 10-5M,
BPA did not induce nuclear translocation of zfAR which is similar to result from the
human androgen receptor (Tomura et al. 2001). BPA was also reported to maximally
inhibit DHT binding to androgen receptor at concentration 50nM and the inhibition is
non-competitive (Lee et al. 2003). This may indicate another distinct interaction site
between BPA and the androgen receptor other than the ligand binding pocket. This
may be a potential explanation for the observed nuclear translocation of zfAR
induced by 10-9M BPA. Treatment of cells with BPA in the presence of 6nM 11-KT
revealed a failure of nuclear import of zfAR. Significant decrease in nuclear
translocation occurred at 1µM and a maximum inhibition effect was observed at 5µM.
Inconsistent results have been reported with human AR. Lee et al. (2003) observed
that the nuclear translocation of GFP-hAR fusion protein in the presence of
testosterone was affected by the addition of 10µM BPA, while Tomura et al. (2001)
showed that co-treatment of cells with 1µM of BPA and 10nM DHT preserved
DHT-induced fluorescence focus formations in the nucleus. Concentration difference
in these two tests probably contributes to the contradictory results, since a dramatic
inhibition of nuclear translocation of AR was seen as the concentration of BPA
increase from 1µM to 5µM in the present study.

In contrast to the observation in this project that vinclozolin and p, p’-DDE did not
induce nuclear translocation of zfAR, these two antiandrogens were found to
successfully promote nuclear import of human AR at concentration of 1µM (Roy et al.
2001; Tomura et al. 2001). Pure antiandrogens such as hydroxyflutamide (OHF) and

62
bicalutamide (CAS) are examples of antiandrogens that does not destroy nuclear
translocation of hAR (Marcelli et al. 2006; Poukka et al. 2000; Tomura et al. 2001;
Tyagi et al. 2000), however, they were usually much slower and less effective than the
natural ligands (Poukka et al. 2000). This group of nonhormonal ligands elicit their
antiandrogenic effects by inhibiting the subsequent steps that result in AR
transactivation (Marcelli et al. 2006; Poukka et al. 2000; Tomura et al. 2001; Tyagi et
al. 2000), and they were also found to disrupt the formation of nuclear foci induced
by DHT (Tomura et al. 2001). However, in our zebrafish model, the interaction of
zfAR with vinclozolin and p, p’-DDE failed to induce the first step, namely the
import of AR into the nuclear compartment, and it also inhibit the interaction between
11-KT and the androgen receptor. Even though vinclozolin in the concentration 10-5M
showed a slight induction of nuclear translocation of zfAR, the dramatic inhibition of
11-KT action indicate a different effect than what was seen in the hAR. This may
indicate that the two chemicals interact with zfAR and hAR in distinct manners.

Fenitrothion showed similar results as vinclozolin and p, p’-DDE. It is previously

demonstrated to competitively antagonize DHT-dependent hAR activation (Tamura et
al. 2001). However, no study has previously investigated the subcellular localization
of AR in the presence of fenitrothion. Thus, this is the first time that fenitrothion is
reported to affect the androgen signaling of zfAR at its nuclear translocation level.
Other examples of antiandrogens that have been shown to cause nuclear translocation
failure include agriculture fungicide procymidone and chlozolinate (Roy et al. 2001).
Binding of agonists to hAR is known to induce conformational changes of the AR
resulting in dissociation of AR with its chaperon protein and altered interactions
between the N and C-termini because of the repositioning of helix 12 in the LDB
(Roy et al. 2001). Thus the suggested mechanism for the mode of action for this
group of antiandrogens is that the conformational transition induced by these
chemicals is inadequate for dissociation of the receptor from other cytoplasmic
proteins and/or exposure of the NLS to promote its interaction with importins for its
nuclear import (Roy et al. 2001). Other steps critical for the nuclear transportation

63
may also be disrupted, such as the phosphorylation of the receptor upon ligand
binding and the dimerization of receptor.

Due to the nonparallel dose-response curve in inhibiting 11-KT induced nuclear

translocation of zfAR, potency of the four antiandrogens is hard to compare based on
the present results. However, according to the LOECs, vinclozolin is the most potent
antiandrogen among the four. This is in agreement with previous results. In a
screening of chemicals for anti-androgenic activity based on inhibition of luciferase
activity induced by R1881, vinclozolin was the most potent antiandrogenic compound
among the others including bisphenol A and p, p’-DDE (Roy et al. 2004). Gray et al.
(1999) also demonstrate that p, p’-DDE is a less potent AR ligand compared to
vinclozolin. In another study, the potency of fenitrothion to competitively inhibit
DHT-dependent human androgen receptor activation of the luciferase reporter was
found to be 8-fold higher than that of p, p’-DDE (Tamura et al. 2001).

The nonparallel dose-response curve created problems in the study of combinational

effects of the chemicals, where the toxic equivalency factor (TEF) approach is
generally used to derive the mixture ratio. However, in this study the TEF vary
depending on the effect level chosen for deriving their numerical values. Therefore
the mixture ratio of the four chemicals used to test the combination effects was
derived from the NOEC and LOEC of the individual compounds. Even though there
were compelling evidences that the combination effects of antiandrogens are dose
additive regardless of the mode of actions, this conclusion could not be verified in this
study. The theory for the cumulative effects from a mixture of antiandrogens that have
distinct mode of actions is that even though the chemicals are acting independently at
a molecular and cellular level, they all target a common pathway, the androgen
signaling pathway. Measurement of endpoints that are far downstream from the
biochemical mechanisms of the individual chemicals can not distinguish decreased
androgen signaling resulting from testosterone synthesis inhibition and that resulting
from antagonists binding to ARs (Rider et al. 2008). In this project, a single step in
the androgen signaling pathway is measured as endpoint, even though many events

64
have to proceed to allow this step to occur. However, nuclear translocation of zfAR
can still be seen as a common target since the single chemical experiments showed
that all four chemicals disrupted 11-KT induced nuclear transport regardless of which
step(s) they may affect.

The potential dose addition effects of the mixture with each of its component at their
NOEC were probably not large enough to be detected because of insufficient number
of test chemicals. Eight and seven chemicals were used in two of the well-known
mixture effects studies where “something from nothing” was reported (Rider et al.
2008; Silva et al. 2002). For the mixture in which each of the components was at the
concentration of its LOEC, a higher inhibition of nuclear translocation was observed
with the mixture compared to that with single chemical. However, it still does not
seem to be high enough to be concluded as dose additive. Limiting resolving power of
the present method may have eliminated the chance to make this conclusion. The
method used in this project is far from the ideal way for acquiring statistically
relevant data at the single cell level, since it is operator-dependent and laborious.
Therefore, it was not realistic to include single chemical controls in every repeat of
the mixture experiments. Combinational effects had to be compared with single
chemical effects that were observed in previous experiments. Because of the variance
resulting from the expression level of the protein, transfection efficiency, the high
background nuclear fluorescence in Hela cells, and the inevitable changing criteria of
the picture analyzer etc, changes of fluorescence partitioning between nucleus and
cytoplasm were sometimes not easy to reach a statistically significant level. High
throughput microscopy that can analyze thousands of cells from a 96-well plate was
used in a study of the quantifying effects of ligands on androgen receptor nuclear
translocation, intranuclear dynamics and solubility (Marcelli et al. 2006). This
approach seems to be an extremely powerful technology for single cell analysis.

For future research, it will be interesting to clearly describe the mechanism of action
of the four antiandrogens in inhibiting discrete step in androgen actions. Competitive
binding assays can be conducted to study the binding kinetics of these chemicals to

65
zfAR. Mutagenesis in the LBD of zfAR can reveal more details about the interaction
between ligands and zfAR. For vinclozolin and p, p’-DDE comparative studies
between human AR and zfAR could be interesting to determine why they do not
interfere with nuclear translocation of hAR but dose interfere with the zfAR. Further
analysis is needed to clearly describe the details in interaction between the antagonists
and the androgen receptor, which could be conducted by structure-function analysis.
A method with higher resolving power should be used to study the mixture effects
where highly quantitative data are required for accurate prediction.

66
Conclusion
By using a YFP-zfAR fusion protein, zfAR was found to undergo nuclear
translocation in response to DHT and 11-KT. Site-directed mutagenesis of a predicted
NLS within zfAR revealed that this nuclear import is mainly mediated through a
bipartite type of NLS (KKCFEVGMTLGARKLRK) ranging from the DBD to the
hinge region of zfAR. Both the first and the second clusters of basic amino acids
(underlined) in the NLS are important in recognizing and/or interacting with the
import machinery, but the second domain seems more essential than the first basic
domain.

By measuring the nuclear translocation of zebrafish androgen receptor as the major

endpoint, the four environmental chemicals, BPA, p, p’-DDE, vinclozolin and
fenitrothion, which have been previously demonstrated to be antiandrogens in
mammalian, were shown also to elicit androgen antagonistic effects in our in vitro
system with zebrafish. The action of 11-KT was inhibited by these four antiandrogens
in a dose-response manner. Interestingly, p, p’-DDE and vinclozolin, that do not
disrupt hAR nuclear translocation, seem to strongly inhibit this step in zfAR action.
Mixture effects of these four chemicals were larger than effects observed with the
single chemicals, but no cumulative effects can be concluded from this study.

67
68
References:
Andersen AG, Jensen TK, Carlsen E, Jorgensen N, Andersson AM, Krarup T, Keiding N, Skakkebaek
NE. 2000. High frequency of sub-optimal semen quality in an unselected population of young
men. Hum Reprod 15(2):366-72.
Arukwe A. 2001. Cellular and molecular responses to endocrine-modulators and the impact on fish
reproduction. Mar Pollut Bull 42(8):643-55.
Bayley M, Larsen PF, Baekgaard H, Baatrup E. 2003. The effects of vinclozolin, an anti-androgenic
fungicide, on male guppy secondary sex characters and reproductive success. Biol Reprod
69(6):1951-6.
Berthois Y, Katzenellenbogen JA, Katzenellenbogen BS. 1986. Phenol red in tissue culture media is a
weak estrogen: implications concerning the study of estrogen-responsive cells in culture. Proc
Natl Acad Sci U S A 83(8):2496-500.
Birkhoj M, Nellemann C, Jarfelt K, Jacobsen H, Andersen HR, Dalgaard M, Vinggaard AM. 2004. The
combined antiandrogenic effects of five commonly used pesticides. Toxicol Appl Pharmacol
201(1):10-20.
Black BE, Paschal BM. 2004. Intranuclear organization and function of the androgen receptor. Trends
Endocrinol Metab 15(9):411-7.
Bonefeld-Jorgensen EC, Long M, Hofmeister MV, Vinggaard AM. 2007. Endocrine-disrupting
potential of bisphenol A, bisphenol A dimethacrylate, 4-n-nonylphenol, and 4-n-octylphenol in
vitro: new data and a brief review. Environ Health Perspect 115 Suppl 1:69-76.
Braun AM, Thomas P. 2004. Biochemical characterization of a membrane androgen receptor in the
ovary of the atlantic croaker (Micropogonias undulatus). Biol Reprod 71(1):146-55.
Brevini TA, Zanetto SB, Cillo F. 2005. Effects of endocrine disruptors on developmental and
reproductive functions. Curr Drug Targets Immune Endocr Metabol Disord 5(1):1-10.
Christiansen S, Scholze M, Axelstad M, Boberg J, Kortenkamp A, Hass U. 2008. Combined exposure
to anti-androgens causes markedly increased frequencies of hypospadias in the rat. Int J
Androl 31(2):241-8.
Culig Z, Klocker H, Bartsch G, Steiner H, Hobisch A. 2003. Androgen receptors in prostate cancer. J
Urol 170(4 Pt 1):1363-9.
Edwards J, Bartlett JM. 2005. The androgen receptor and signal-transduction pathways in
hormone-refractory prostate cancer. Part 1: Modifications to the androgen receptor. BJU Int
95(9):1320-6.
Feldman BJ, Feldman D. 2001. The development of androgen-independent prostate cancer. Nat Rev
Cancer 1(1):34-45.
Foradori CD, Weiser MJ, Handa RJ. 2008. Non-genomic actions of androgens. Front Neuroendocrinol
29(2):169-81.
Gaido KW, Leonard LS, Lovell S, Gould JC, Babai D, Portier CJ, McDonnell DP. 1997. Evaluation of
chemicals with endocrine modulating activity in a yeast-based steroid hormone receptor gene
transcription assay. Toxicol Appl Pharmacol 143(1):205-12.
Georget V, Lobaccaro JM, Terouanne B, Mangeat P, Nicolas JC, Sultan C. 1997. Trafficking of the
androgen receptor in living cells with fused green fluorescent protein-androgen receptor. Mol
Cell Endocrinol 129(1):17-26.
Goksoyr A. 2006. Endocrine disruptors in the marine environment: mechanisms of toxicity and their

69
 influence on reproductive processes in fish. J Toxicol Environ Health A 69(1-2):175-84.
Gould JC, Leonard LS, Maness SC, Wagner BL, Conner K, Zacharewski T, Safe S, McDonnell DP,
Gaido KW. 1998. Bisphenol A interacts with the estrogen receptor alpha in a distinct manner
from estradiol. Mol Cell Endocrinol 142(1-2):203-14.
Gray LE, Jr., Ostby J, Monosson E, Kelce WR. 1999a. Environmental antiandrogens: low doses of the
fungicide vinclozolin alter sexual differentiation of the male rat. Toxicol Ind Health
15(1-2):48-64.
Gray LE, Jr., Ostby JS, Kelce WR. 1994. Developmental effects of an environmental antiandrogen: the
fungicide vinclozolin alters sex differentiation of the male rat. Toxicol Appl Pharmacol
129(1):46-52.
Gray LE, Jr., Wilson VS, Stoker T, Lambright C, Furr J, Noriega N, Howdeshell K, Ankley GT,
Guillette L. 2006. Adverse effects of environmental antiandrogens and androgens on
reproductive development in mammals. Int J Androl 29(1):96-104; discussion 105-8.
Gray LE, Jr., Wolf C, Lambright C, Mann P, Price M, Cooper RL, Ostby J. 1999b. Administration of
potentially antiandrogenic pesticides (procymidone, linuron, iprodione, chlozolinate, p,p'-DDE,
and ketoconazole) and toxic substances (dibutyl- and diethylhexyl phthalate, PCB 169, and
ethane dimethane sulphonate) during sexual differentiation produces diverse profiles of
reproductive malformations in the male rat. Toxicol Ind Health 15(1-2):94-118.
Hass U, Scholze M, Christiansen S, Dalgaard M, Vinggaard AM, Axelstad M, Metzdorff SB,
Kortenkamp A. 2007. Combined exposure to anti-androgens exacerbates disruption of sexual
differentiation in the rat. Environ Health Perspect 115 Suppl 1:122-8.
Heinlein CA, Chang C. 2002. The roles of androgen receptors and androgen-binding proteins in
nongenomic androgen actions. Mol Endocrinol 16(10):2181-7.
Hirawat S, Budman DR, Kreis W. 2003. The androgen receptor: structure, mutations, and
antiandrogens. Cancer Invest 21(3):400-17.
Hossain MS, Larsson A, Scherbak N, Olsson PE, Orban L. 2008. Zebrafish androgen receptor: isolation,
molecular, and biochemical characterization. Biol Reprod 78(2):361-9.
Hotchkiss AK, Parks-Saldutti LG, Ostby JS, Lambright C, Furr J, Vandenbergh JG, Gray LE, Jr. 2004.
A mixture of the "antiandrogens" linuron and butyl benzyl phthalate alters sexual
differentiation of the male rat in a cumulative fashion. Biol Reprod 71(6):1852-61.
Jarfelt K, Dalgaard M, Hass U, Borch J, Jacobsen H, Ladefoged O. 2005. Antiandrogenic effects in
male rats perinatally exposed to a mixture of di(2-ethylhexyl) phthalate and di(2-ethylhexyl)
adipate. Reprod Toxicol 19(4):505-15.
Jenster G, Trapman J, Brinkmann AO. 1993. Nuclear import of the human androgen receptor. Biochem
J 293 ( Pt 3):761-8.
Jenster G, van der Korput JA, Trapman J, Brinkmann AO. 1992. Functional domains of the human
androgen receptor. J Steroid Biochem Mol Biol 41(3-8):671-5.
Jorgensen A, Andersen O, Bjerregaard P, Rasmussen LJ. 2007. Identification and characterisation of an
androgen receptor from zebrafish Danio rerio. Comp Biochem Physiol C Toxicol Pharmacol
146(4):561-8.
Karvonen U, Janne OA, Palvimo JJ. 2002. Pure antiandrogens disrupt the recruitment of coactivator
GRIP1 to colocalize with androgen receptor in nuclei. FEBS Lett 523(1-3):43-7.
Katsiadaki I, Morris S, Squires C, Hurst MR, James JD, Scott AP. 2006. Use of the three-spined
stickleback (Gasterosteus aculeatus) as a sensitive in vivo test for detection of environmental

70
 antiandrogens. Environ Health Perspect 114 Suppl 1:115-21.
Kelce WR, Lambright CR, Gray LE, Jr., Roberts KP. 1997. Vinclozolin and p,p'-DDE alter
androgen-dependent gene expression: in vivo confirmation of an androgen receptor-mediated
mechanism. Toxicol Appl Pharmacol 142(1):192-200.
Kelce WR, Monosson E, Gamcsik MP, Laws SC, Gray LE, Jr. 1994. Environmental hormone
disruptors: evidence that vinclozolin developmental toxicity is mediated by antiandrogenic
metabolites. Toxicol Appl Pharmacol 126(2):276-85.
Kelce WR, Stone CR, Laws SC, Gray LE, Kemppainen JA, Wilson EM. 1995. Persistent DDT
metabolite p,p'-DDE is a potent androgen receptor antagonist. Nature 375(6532):581-5.
Kinnberg K, Holbech H, Petersen GI, Bjerregaard P. 2007. Effects of the fungicide prochloraz on the
sexual development of zebrafish (Danio rerio). Comp Biochem Physiol C Toxicol Pharmacol
145(2):165-70.
Kiparissis Y, Metcalfe TL, Balch GC, Metcalfe CD. 2003. Effects of the antiandrogens, vinclozolin and
cyproterone acetate on gonadal development in the Japanese medaka (Oryzias latipes). Aquat
Toxicol 63(4):391-403.
Kortenkamp A. 2007. Ten years of mixing cocktails: a review of combination effects of
endocrine-disrupting chemicals. Environ Health Perspect 115 Suppl 1:98-105.
Krishnan AV, Stathis P, Permuth SF, Tokes L, Feldman D. 1993. Bisphenol-A: an estrogenic substance
is released from polycarbonate flasks during autoclaving. Endocrinology 132(6):2279-86.
Kumar S, Saradhi M, Chaturvedi NK, Tyagi RK. 2006. Intracellular localization and nucleocytoplasmic
trafficking of steroid receptors: an overview. Mol Cell Endocrinol 246(1-2):147-56.
Kurosawa T, Hiroi H, Tsutsumi O, Ishikawa T, Osuga Y, Fujiwara T, Inoue S, Muramatsu M, Momoeda
M, Taketani Y. 2002. The activity of bisphenol A depends on both the estrogen receptor
subtype and the cell type. Endocr J 49(4):465-71.
LaCasse EC, Lefebvre YA. 1995. Nuclear localization signals overlap DNA- or RNA-binding domains
in nucleic acid-binding proteins. Nucleic Acids Res 23(10):1647-56.
Lee DK, Chang C. 2003a. Molecular communication between androgen receptor and general
transcription machinery. J Steroid Biochem Mol Biol 84(1):41-9.
Lee HJ, Chang C. 2003b. Recent advances in androgen receptor action. Cell Mol Life Sci
60(8):1613-22.
Lee HJ, Chattopadhyay S, Gong EY, Ahn RS, Lee K. 2003. Antiandrogenic effects of bisphenol A and
nonylphenol on the function of androgen receptor. Toxicol Sci 75(1):40-6.
Leon A, Teh SJ, Hall LC, Teh FC. 2007. Androgen disruption of early development in Qurt strain
medaka (Oryzias latipes). Aquat Toxicol 82(3):195-203.
Leon A, Wu PS, Hall LC, Johnson ML, Teh SJ. 2008. Global gene expression profiling of androgen
disruption in Qurt strain medaka. Environ Sci Technol 42(3):962-9.
Makynen EA, Kahl MD, Jensen KM, Tietge JE, Wells KL, Van Der Kraak G, Ankley GT. 2000. Effects
of the mammalian antiandrogen vinclozolin on development and reproduction of the fathead
minnow (Pimephales promelas). Aquat Toxicol 48(4):461-475.
Marcelli M, Stenoien DL, Szafran AT, Simeoni S, Agoulnik IU, Weigel NL, Moran T, Mikic I, Price JH,
Mancini MA. 2006. Quantifying effects of ligands on androgen receptor nuclear translocation,
intranuclear dynamics, and solubility. J Cell Biochem 98(4):770-88.
Martinovic D, Blake LS, Durhan EJ, Greene KJ, Kahl MD, Jensen KM, Makynen EA, Villeneuve DL,
Ankley GT. 2008. Reproductive toxicity of vinclozolin in the fathead minnow: confirming an

71
 anti-androgenic mode of action. Environ Toxicol Chem 27(2):478-88.
Metzdorff SB, Dalgaard M, Christiansen S, Axelstad M, Hass U, Kiersgaard MK, Scholze M,
Kortenkamp A, Vinggaard AM. 2007. Dysgenesis and histological changes of genitals and
perturbations of gene expression in male rats after in utero exposure to antiandrogen mixtures.
Toxicol Sci 98(1):87-98.
Michels G, Hoppe UC. 2008. Rapid actions of androgens. Front Neuroendocrinol 29(2):182-98.
Nativelle-Serpentini C, Richard S, Seralini GE, Sourdaine P. 2003. Aromatase activity modulation by
lindane and bisphenol-A in human placental JEG-3 and transfected kidney E293 cells. Toxicol
In Vitro 17(4):413-22.
Nellemann C, Dalgaard M, Lam HR, Vinggaard AM. 2003. The combined effects of vinclozolin and
procymidone do not deviate from expected additivity in vitro and in vivo. Toxicol Sci
71(2):251-62.
Orn S, Holbech H, Madsen TH, Norrgren L, Petersen GI. 2003. Gonad development and vitellogenin
production in zebrafish (Danio rerio) exposed to ethinylestradiol and methyltestosterone.
Aquat Toxicol 65(4):397-411.
Paschal BM. 2002. Translocation through the nuclear pore complex. Trends Biochem Sci 27(12):593-6.
Picard D, Yamamoto KR. 1987. Two signals mediate hormone-dependent nuclear localization of the
glucocorticoid receptor. Embo J 6(11):3333-40.
Poukka H, Karvonen U, Yoshikawa N, Tanaka H, Palvimo JJ, Janne OA. 2000. The RING finger
protein SNURF modulates nuclear trafficking of the androgen receptor. J Cell Sci 113 ( Pt
17):2991-3001.
Rider CV, Furr J, Wilson VS, Gray LE, Jr. 2008. A mixture of seven antiandrogens induces
reproductive malformations in rats. Int J Androl 31(2):249-62.
Roy AK, Tyagi RK, Song CS, Lavrovsky Y, Ahn SC, Oh TS, Chatterjee B. 2001. Androgen receptor:
structural domains and functional dynamics after ligand-receptor interaction. Ann N Y Acad
Sci 949:44-57.
Roy P, Salminen H, Koskimies P, Simola J, Smeds A, Saukko P, Huhtaniemi IT. 2004. Screening of
some anti-androgenic endocrine disruptors using a recombinant cell-based in vitro bioassay. J
Steroid Biochem Mol Biol 88(2):157-66.
Sackey FN, Hache RJ, Reich T, Kwast-Welfeld J, Lefebvre YA. 1996. Determinants of subcellular
distribution of the glucocorticoid receptor. Mol Endocrinol 10(10):1191-205.
Saitoh M, Takayanagi R, Goto K, Fukamizu A, Tomura A, Yanase T, Nawata H. 2002. The presence of
both the amino- and carboxyl-terminal domains in the AR is essential for the completion of a
transcriptionally active form with coactivators and intranuclear compartmentalization common
to the steroid hormone receptors: a three-dimensional imaging study. Mol Endocrinol
16(4):694-706.
Saporita AJ, Zhang Q, Navai N, Dincer Z, Hahn J, Cai X, Wang Z. 2003. Identification and
characterization of a ligand-regulated nuclear export signal in androgen receptor. J Biol Chem
278(43):41998-2005.
Sebire M, Scott AP, Tyler CR, Cresswell J, Hodgson DJ, Morris S, Sanders MB, Stebbing PD,
Katsiadaki I. 2008. The organophosphorous pesticide, fenitrothion, acts as an anti-androgen
and alters reproductive behavior of the male three-spined stickleback, Gasterosteus aculeatus.
Ecotoxicology.
Silva E, Rajapakse N, Kortenkamp A. 2002. Something from "nothing"--eight weak estrogenic

72
 chemicals combined at concentrations below NOECs produce significant mixture effects.
Environ Sci Technol 36(8):1751-6.
Sohoni P, Lefevre PA, Ashby J, Sumpter JP. 2001. Possible androgenic/anti-androgenic activity of the
insecticide fenitrothion. J Appl Toxicol 21(3):173-8.
Sohoni P, Sumpter JP. 1998. Several environmental oestrogens are also anti-androgens. J Endocrinol
158(3):327-39.
Sonnenschein C, Soto AM. 1998. An updated review of environmental estrogen and androgen mimics
and antagonists. J Steroid Biochem Mol Biol 65(1-6):143-50.
Soto AM, Michaelson CL, Prechtl NV, Weill BC, Sonnenschein C, Olea-Serrano F, Olea N. 1998.
Assays to measure estrogen and androgen agonists and antagonists. Adv Exp Med Biol
444:9-23; discussion 23-8.
Sperry TS, Thomas P. 1999. Identification of two nuclear androgen receptors in kelp bass (Paralabrax
clathratus) and their binding affinities for xenobiotics: comparison with Atlantic croaker
(Micropogonias undulatus) androgen receptors. Biol Reprod 61(4):1152-61.
Sunami O, Kunimatsu T, Yamada T, Yabushita S, Sukata T, Miyata K, Kamita Y, Okuno Y, Seki T,
Nakatsuka I and others. 2000. Evaluation of a 5-day Hershberger assay using young mature
male rats: methyltestosterone and p,p'-DDE, but not fenitrothion, exhibited androgenic or
antiandrogenic activity in vivo. J Toxicol Sci 25(5):403-15.
Tamura H, Maness SC, Reischmann K, Dorman DC, Gray LE, Gaido KW. 2001. Androgen receptor
antagonism by the organophosphate insecticide fenitrothion. Toxicol Sci 60(1):56-62.
Tomura A, Goto K, Morinaga H, Nomura M, Okabe T, Yanase T, Takayanagi R, Nawata H. 2001. The
subnuclear three-dimensional image analysis of androgen receptor fused to green fluorescence
protein. J Biol Chem 276(30):28395-401.
Tyagi RK, Lavrovsky Y, Ahn SC, Song CS, Chatterjee B, Roy AK. 2000. Dynamics of intracellular
movement and nucleocytoplasmic recycling of the ligand-activated androgen receptor in
living cells. Mol Endocrinol 14(8):1162-74.
Vinggaard AM, Hass U, Dalgaard M, Andersen HR, Bonefeld-Jorgensen E, Christiansen S, Laier P,
Poulsen ME. 2006. Prochloraz: an imidazole fungicide with multiple mechanisms of action.
Int J Androl 29(1):186-92.
Whitaker HC, Hanrahan S, Totty N, Gamble SC, Waxman J, Cato AC, Hurst HC, Bevan CL. 2004.
Androgen receptor is targeted to distinct subcellular compartments in response to different
therapeutic antiandrogens. Clin Cancer Res 10(21):7392-401.
White SN, Jamnongjit M, Gill A, Lutz LB, Hammes SR. 2005. Specific modulation of nongenomic
androgen signaling in the ovary. Steroids 70(5-7):352-60.
Wilson VS, Blystone CR, Hotchkiss AK, Rider CV, Gray LE, Jr. 2008. Diverse mechanisms of
anti-androgen action: impact on male rat reproductive tract development. Int J Androl
31(2):178-87.
Xu LC, Sun H, Chen JF, Bian Q, Qian J, Song L, Wang XR. 2005. Evaluation of androgen receptor
transcriptional activities of bisphenol A, octylphenol and nonylphenol in vitro. Toxicology
216(2-3):197-203.
You L. 2004. Steroid hormone biotransformation and xenobiotic induction of hepatic steroid
metabolizing enzymes. Chem Biol Interact 147(3):233-46.
Zhou ZX, Sar M, Simental JA, Lane MV, Wilson EM. 1994. A ligand-dependent bipartite nuclear
targeting signal in the human androgen receptor. Requirement for the DNA-binding domain

73
 and modulation by NH2-terminal and carboxyl-terminal sequences. J Biol Chem
269(18):13115-23.
Zhu Z, Becklin RR, Desiderio DM, Dalton JT. 2001. Identification of a novel phosphorylation site in
human androgen receptor by mass spectrometry. Biochem Biophys Res Commun
284(3):836-44.
Zober A, Hoffmann G, Ott MG, Will W, Germann C, van Ravenzwaay B. 1995. Study of morbidity of
personnel with potential exposure to vinclozolin. Occup Environ Med 52(4):233-41.

74
75