$3

I

 

r] v:

.. 9.
3m?!
IV

l _.
flaw... .u. 7"..an
mm. uh
a...“
r i.
:nkfl.

a. :
bun.

. .:.....,.........§
f3]! an}.
Ly. ..

v‘v‘. . I
a!!!
:7»

1:29.
.3 t

 

gr"

3..

 

Michigan's't'ate
University

This is to certify that the
dissertation entitled

APPLICATION OF TOXICOGENOMIC APPROACHES TO
STUDY CHEMICAL-INDUCED EFFECTS ON THE
HYPOTHALAMIC- PITUITARY- GONADAL (HPG) AXIS OF
THE JAPANESE MEDAKA (ORYZIAS. LA TIPES)

presented by

Xiaowei Zhang

has been accepted towards fulfillment
of the requirements for the

 

 

Zoology -
PhD degree in Environmental Toxicology
, /"'"7 g ’i .- M \
gép/é’l/‘A/ / A/ ’074—4/1

 

fl, Major Profeséot’s Si?
6‘ A /1 §/// .17ng ‘
/

Date

MSU is an affirmative-action. equal-opportunity employer

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5108 K IProi/Acc8-Pres/ClRC/Dateoue indd

APPLICATION OF TOXICOGENOMIC APPROACHES To STUDY
CHEMICAL-INDUCED EFFECTS ON THE HYPOTHALAMIC-
PITUITARY- GONADAL (HPG) AXIS OF THE JAPANESE MEDAKA
(0R YZIA S. LA TIPES)

By

Xiaowei Zhang

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Zoology

&
Environmental Toxicology

2008

ABSTRACT
APPLICATION OF TOXICOGENOMIC APPROACHES TO STUDY CHEMICAL-
INDUCED EFFECTS ON THE HYPOTHALAMIC- PITUITARY- GONADAL (HPG)
AXIS OF THE JAPANESE MEDAKA (ORYZIAS. LATIPES)
By
Xiaowei Zhang
System models utilizing genomic approaches can be powerful tools for

mechanistic toxicological research. This dissertation describes the development and
validation of a real time polymerase chain reaction (RT-PCR) array for studying
chemical-induced effects on gene expression of selected endocrine pathways along the
hypothalamic-pituitary-gonadal (HPG) axis of the small, oviparous fish, the Japanese
medaka (Oryzias latipes). The Japanese medaka HPG PCR array combines the
quantitative performance of SYBR® Green-based real-time PCR with the multiple gene
profiling capabilities of a microarray to examine expression profiles of 36 genes
associated with endocrine pathways in brain, liver and gonad. The performance of the
Japanese medaka HPG PCR array was evaluated by examining effects of five model
compounds, the synthetic estrogen, 17a-ethinylestradiol (EEZ), the anabolic androgen,
17B-trenbolone (TRB), the aromatase inhibitor, fadrozole (FAD), the imidozole-type
fungicides, prochloraz (PCZ) and ketoconazole (KTC) on the HPG axis of the Japanese
medaka. A pathway-based approach was implemented to analyze and visualize
concentration-dependent mRNA expression in the HPG axis of Japanese medaka. Four-
month-old medaka were exposed to different concentration of chemicals for 7 d in a static
renewal exposure system and the exposure concentrations were EEZ (5, 50, 500 ng/L), or

TRB (50, 500, 5000 ng/L) or PCZ (3, 30, 300 pg/L) or KTC (3, 30, 300 ug/L) or SOug

F AD/L. TRB, PCZ, KTC or FAD caused time- dependent reductions in fecundity by
Japanese medaka, but EE2 did not. The compensatory response to EEZ exposure
included the down-regulation of male brain GnRH R I and testicular CYP17. Despite
their different biochemical properties, TRB or FAD caused Similar responses in Japanese
medaka, such as lesser fecundity and down-regulation of V T G and CHG genes in the
liver of females. Compensatory responses to TRB in the female HPG axis included up-
regulation of brain GnRH R I] and ovary CYP19A. Exposure to FAD for 8 h resulted in
an 8-fold and 71-fold down-regulation of expression of estrogen receptor alpha (ER-a)
and CH6 L, respectively in female liver. TRB caused Similar down-regulation but the
effects were not observed until 32 h of exposure. These results support the hypothesis
that FAD reduces plasma E2 more quickly by inhibiting aromatase enzyme activity than
does TRB, which inhibits production of the E2 precursor testosterone. Exposure to KTC
or PCZ significantly down-regulated expression of ER—a and egg precursors in livers of
males and females. However, PCZ was more potent than KTC both in modulating
transcription and in causing lesser fecundity. Correlation analysis indicated that ER-a
plays a primary role in the transcription of V T G and CH0 genes in livers. The mRNA
level of the five egg precursors and ER-a in livers of females was log-log related to the
ecologically relevant endpoint, fecundity. Overall, the organ- gender- and concentration—
specific gene expression profiles derived by the Japanese medaka HPG axis RT—PCR
array provides a powerful tool to not only delineate chemical-induced modes of action,

but also to quantitatively evaluate chemical induced adverse effects on reproduction.

To My Parents:

Lanxin Zhang (gfiéfifl) and Xiaohuan Chen (WEBER)

To My Wife:

Jiali Miao (filififi)

ACKNOWLEDGEMENTS

First and foremost, I wish to thank my esteemed mentor, Professor John P. Giesy,
not only for inspiring and encouraging me to pursue a career in Environmental Science,
but also for offering me this challenging research opportunity. He was readily available
for me, as he so generously is for all of his students. More than anyone else, his
influence has continued to my development as an environmental scientist.

Second, my thanks go out to Dr. Markus Hecker, Dr. John Newsted, and Dr. Paul
Jones, whom are the Co-PIS of this project. Especially to Dr. Hecker, he had put
tremendous efforts in this research. I deeply appreciate the expertise and support from
my committee members, Professor Kaminski and Dr. Rob Halgren.

My research for this dissertation was made more efficient through the use of
several genomic research facilities. Thus I gladly express my gratitude to Dr. Annette
Thelen and Dr. Jeff Landgraf at the Research Technology Support Facility (RTSF) in
Michigan State University, especially for the training and kind assistance.

My family deserves a great deal of credit for my development and achievement. I
thank my parents for all their love and support. Finally, I offer my most genuine thanks

to my wife J iali Miao, whose unwavering compassion, faith and love carries me through

life.

TABLE OF CONTENTS

LIST OF TABLES ................................................................................. VII
LIST OF FIGURES .............................................................................. VIII
KEY TO ABBREVIATIONS ..................................................................... XI
CHAPTER I
INTRODUCTION

References .................................................................................... 8
CHAPTER 2

REAL TIME PCR ARRAY TO STUDY EFFECTS OF CHEMICALS ON THE
HYPOTHALAMIC-PITUITARY-GONADAL AXIS OF THE JAPANESE MEDAKA

Introduction ................................................................................. 13

Materials and Methods ..................................................................... 16

Results ....................................................................................... 29

Discussion ................................................................................... 39

References ................................................................................... 45
CHAPTER 3

TIME-DEPENDENT TRANSCRIPTIONAL PROFILES OF GENES OF THE
HYPOTHALAMIC-PITUITARY-GONADAL (HPG) AXIS IN MEDAKA (O.
LATIPES) EXPOSED TO F ADROZOLE AND I7B-TRENBOLONE

Introduction ................................................................................. 53

Materials and Methods ..................................................................... 56

Results ....................................................................................... 65

Discussion ................................................................................... 68

References ................................................................................... 74
CHAPTER 4

RESPONSES OF THE MEDAKA HPG AXIS PCR ARRAY AND REPRODUCTION
TO PROCHLORAZ AND KETOCONAZOLE

Introduction ................................................................................. 78
Materials and Methods ..................................................................... 81
Results ....................................................................................... 89
Discussion ................................................................................... 92
References ................................................................................... 96
CHAPTER 5
CONCLUSION
References ................................................................................. 108

vi

LIST OF TABLES

Table 2.1. Gene list ofthe medaka HPG PCR array system ................................... 21
Table 2.2. Primer sequences of selected medaka HPG axis genes ........................... 23

Table 2.3. Chemical induced effects on medaka gonadal-somatic index (GSI), hepatic-
somatic index (HIS) and brain-somatic index (BSI) ........................................... 25

Table 2.4. Changes in expression for HPG genes in EE2-exposed medaka fish. Genes
exhibited an over two-fold or significant change (p-value < 0.05) in expression between
control and exposed medaka are listed ........................................................... 28

Table 2.5. Changes in expression for HPG genes in TRB-exposed medaka fish. Genes
exhibited an over two-fold or Significant change (p-value < 0.05) in expression between
control and exposed medaka are listed ........................................................... 36

Table 2.6. Spearman rank correlation coefficients (numbers) and probabilities (*)
between hepatic expression levels of ER-a , Eli-,6. AR-a mRNA and other genes. . . . . ....37

Table 3.1. Transcriptional response profiles of HPG axis pathways in medaka fish
exposed to 50 pg FAD /L. Gene expression was expressed as the fold change comparing
to the corresponding solvent controls ............................................................ 61

Table 3.2. Transcriptional response profiles of HPG axis pathways in medaka fish
exposed to 2.0 pg TRB /L. Gene expression was expressed as the fold change comparing
to the corresponding solvent controls ............................................................ 62

Table 4.1 Transcriptional response profiles of HPG axis pathways in medaka fish
exposed to prochloraz (PCZ). Gene expression was expressed as the fold change
comparing to the corresponding vehicle controls ............................................... 84

Table 4.2 Transcriptional response profiles of HPG axis pathways in medaka fish
exposed to ketoconazole (KTC). Gene expression was expressed as the fold change
comparing to the corresponding vehicle controls ............................................... 85

Table 5.1 Effects ofdifferent chemicals on fecundity ofJapanese medaka in 7 d
exposure ............................................................................................. 104

vii

LIST OF FIGURES

Figure 2.1. Cumulative fecundity in medaka exposed to EE2 (A) or TRB (B) in a 7-d
test. Data represent the mean cumulative number of eggs per female collected from 3
replicate tanks, each containing 6 pairs of fish. The asterisks indicate a Significant
different (p-value < 0.05) from control group ................................................... 24

Figure 2.2. Volcano plots of chemically induced changes in gene expression pattern in
males and females. Data are from medaka exposed to 500 ng EE2/L or 5000 ng TRB/L.
Genes plotted farther from the either the x or y- axis have larger changes in gene
expression. Thresholds for fold-change (vertical lines, 2-fold) and significant difference
(horizontal line, p < 0.01) were used in this display ............................................ 26

Figure 2.3. Striped view of concentration dependent response profile in EE2 exposure
of male Japanese medaka. Gene expression data from medaka treated by 5, 50 and 500
ng EE2/L are shown as striped color sets on the selected endocrine pathways along the
medaka HPG axis. The legend listed in the upper right corner of the graph describes the
order of the three EE2 concentration and the eight colors designating different fold
thresholds. LH, lutinizing hormone; FSH, follicle-stimulating hormone; E2, 170-
estradiol; T, testosterone; HDL, high-density lipid; LDL, low-density lipid ............... 34

Figure 2.4. Striped view of concentration dependent response profile in TRB exposure
of female Japanese medaka. Gene expression data from medaka treated by 50, 500 and
5000 ng TRB/L are shown as striped color sets on the selected endocrine pathways along
the medaka HPG axis. The legend listed in the upper right corner of the graph describes
the order of the three TRB concentration and the eight colors designating different fold
thresholds ............................................................................................. 35

Figure 2.5. Correlation of brain expression level of brain CYPI 93 vs. ovary CYPI9A.
Spearman rank correlation coefficients for female -0.676 (p-value < 0.001) male 0.266
(p-value = 0.102) .................................................................................... 38

Figure 3.]. Cumulative fecundity in medaka exposed to 2.0 pg TRB/L or 50 pg
FAD/L in a 7-d test. Data represent the mean and standard deviation of cumulative
number of eggs per female collected from 2 replicate tanks, each containing 5 pairs of
fish. The asterisks indicate a Significant difference (p-value < 0.05) from vehicle control
group. ................................................................................................ 60

Figure 3.2. Striped view of Time dependent response profile in female Japanese
medaka exposed to 50 pg F AD/ L. The legend listed in the upper right corner of the
graph describes the order ofthe three sampling time points and the eight colors
designating different fold thresholds. LH, lutinizing hormone; FSH, follicle-stimulating
hormone; E2, l7B-estradiol; T, testosterone; HDL, high-density lipid; LDL, low-density
lipid ................................................................................................... 63

viii

Figure 3.3. Striped view of Time dependent response profile in female Japanese
medaka exposed to 2.0 pg TRB/L. The legend listed in the upper right corner of the
graph describes the order of the three sampling time points and the eight colors
designating different fold thresholds. LH, lutinizing hormone; FSH, follicle-stimulating
hormone; E2, l7B-estradiol; T, testosterone; HDL, high-density lipid; LDL, low-density
lipid ................................................................................................... 64

Figure 4.]. Cumulative fecundity of Japanese medaka exposed to PCZ or KTC for 7 d.
In the PCZ exposure, values are means (n=2) of the cumulative number of eggs,
expressed on a per female basis. In the KTC exposure, n = l. Asterisks indicate a
significant different from control group: *,p < 0.05 ........................................... 86

Figure 4.2. Striped view of concentration dependent response profile in PCZ exposure
of female Japanese medaka. Gene expression data from medaka treated by 3.0, 30 and
300 pg PCZ/L are shown as striped color sets on the selected endocrine pathways along
the medaka HPG axis. The legend listed in the upper right comer of the graph describes
the order of the three PCZ concentrations and the eight colors designating different fold
thresholds. LH, lutinizing hormone; FSH, follicle-stimulating hormone; E2, 170-
estradiol; T, testosterone; HDL, high-density lipid; LDL, low-density lipid ................ 87

Figure 4.3. Striped view of concentration dependent response profile in KTC

exposure of female Japanese medaka. Gene expression data from medaka treated by 3.0,
30 and 300 pg KTC/L are shown as striped color sets on the selected endocrine pathways
along the medaka HPG axis. The legend listed in the upper right corner of the graph
describes the order of the three KTC concentrations and the eight colors designating
different fold thresholds. LH, lutinizing hormone; FSH, follicle-stimulating hormone; E2,
l7B-estradiol; T, testosterone; HDL, high-density lipid; LDL, low-density lipid ........... 88

Figure 5.1. Heatmap of the concentration-dependent gene expression profiles in livers of
chemical exposed females. Gene tree was constructed by pearson correlation metric.
Chemical tree was constructed by ‘ToxClust” method, where the dissimilarity between
any two chemicals was calculated by the distance between the concentration—dependent
response curves in the exposure of both chemical ............................................. 105

Figure 5.2. Scree plot of the percentage of variance explained by each of the Principal
Components (PC) as a percentage of the total variance of the gene expression. The six
hepatic genes included in the analysis were ER-a, VT G I. VTG II, C HG L, ('HG H and
CH0 HM ............................................................................................ 106

Figure 5.3. Relationship between fecundity and gene expression in livers of females.
A: fecundity vs hepatic index, the broken line Shows the trend of data. B: Simple linear
regression of IoglO-transformed fecundity and hepatic index. The functions describing
the relationship are: Hepatic index = 0.236 *loglO (ER-(I)+ 0.326 *logl 0 (VTG I)+ 0.537
*Iog10(VTG II)+ 0.472 *loglO (CHG L) + 0.343 *loglO (CHG H) + 0.457 *IoglO
(CHG HM). The formula for the regression model was: loglO (fecundity) = 1.616 ——
0.4493 * logIO(-hepatic index) .................................................................. 107

ix

“Images in this dissertation are presented in color.”

ANOVA:

AR:

881:

CHG:

CHG H:

CHG HM:

CHG L:

Ct:

CYP:

CYP17:

DMSO:

E2:

EDCm

EE2:

ELISA:

ER:

FAD:

FSH:

GSI:

HDL:

HI:

HIS:

KEY TO ABBREVIATIONS

Analysis of variance

androgen receptors
brain-somatic index
choriogenin

choriogenin H

choriogenin Hminor
choriogenin L

cycle threshold

cytochrome P450

cytochrome P450 c170t hydroxylase. 17,20-lyase
dimethyl sulfoxide
l7B-estradiol

Endocrine disrupting chemicals
170i -ethinylestradiol
enzyme—linked immunosorbent assay
estrogen receptors

fadrozole

follicle-stimulating hormone
gonadal-somatic index
high-density lipid

hepatic index

hepatic-somatic index

xi

HPG:

lACUC:

KT:

KTC :

LDL:

LH:

MOA:

MSU:

OECD:

PCI:

PCR:

PC Z:

RACE:

RT-PC R:

ThR:

TRB:

VTG:

KEY TO ABBREVIATIONS (Cont’d)

hypothalamic-pituitary-gonadal

Institutional Animal Care and Use Committee
1 I-ketotesterone

ketoconazole

low-density lipid

lutinizing hormone

modes of action

Michigan State University

Organization of Economic Cooperation and Development
first principle component

polymerase chain reaction

prochloraz

rapid amplification of cDNA end

real-time polymerase chain reaction
testosterone

thyroid hormone receptors

1 7B-trenbolone

vitellogenin

xii

Pg:

pL:

min:

mg:

ml:

ng:

Pg:

yr:

UNITS OF MEASURE

gram
hour

liter

micro gram
microliter
minutes
milligram
milliliter
nanogram
picogram

year

xiii

Chapter I

Introduction

The issue of potential endocrine disruption by chemicals was highlighted by recent
legislation mandating that chemicals and formulations be screened for their potential to
modulate the endocrine system before they are manufactured or used in certain processes
(Safe Drinking Water Act Amendments of 1995 - Bill Number 8.1316; Food Quality
Protection Act of 1996 - Bill Number PL. 104-170). Currently knowledge of chemical-
induced endocrine disruption is largely limited to the pathways mediated through several
classical steroid hormone receptors, including estrogen receptors (ER), androgen
receptors (AR) and thyroid hormone receptors (T hR). Our recent studies have evaluated
the chemicals induced effect on steriodogenesis which alter the rates as well as absolute
and relative concentrations of hormones produced by steroidogenic cells by altering the
expression of steroidogenic enzymes (Gracia et al. 2006; Hecker et al. 2006; Hilscherova
et al. 2004; Zhang et al. 2005). Endocrine disrupting chemicals (EDCS) could alter
normal patterns of gene expression either by direct (steroid hormone receptor mediated
pathways) or compensatory effects (V illeneuve et al. 2007). Historically, studies on
endocrine disrupting induced chemicals have generally focused on several endpoints and
one tissue at one specific time in the development of an organism. However, considering
the complicated endocrine system in humans and wildlife, current chemical screening
tools are limited to either narrow molecular targets or restricted methods. What is needed
is a sensitive, flexible monitoring tool that allows for the screening of a multiple
molecular targets genes in multiple tissues simultaneously at any stage of development
and allow for mechanism based toxicity prediction and assessment.

A significant degree of conservation has been shown in the basic aspects of the

hypothalamic-pituitary-gonadal (HPG) axis among vertebrates (Ankley and Johnson

2005; Danger et al. 1990). Teleost fish, such as the Japanese medaka (Oryzias latipes),
zebrafish (Danio rerio) and fathead minnow (Pimephales promelas), have been
suggested to be appropriate models for testing EDCS in terms of both ecological impacts
and species extrapolation (Ankley and Villeneuve 2006; Villeneuve et al. 2007). The
medaka is a small, oviparous (egg-laying) freshwater fish native to Asia. Its physiology,
embryology and genetics have been extensively studied for more than 100 y (Wittbrodt et
al. 2002). The medaka represents an important test system for environmental research
and is widely used for testing endocrine disrupters in ecotoxicology (Pastva et al. 2001;
Villalobos et al. 2003). The medaka has several major advantages compared to the
zebrafish (Danio rerio), another commonly used model organism that are important for
the proposed study. First, to date nothing is known about the exact mechanism of sex
determination in zebrafish, an aspect that is highly relevant when testing EDC effects on
reproductive endocrinology. Conversely, the Japanese medaka has clearly defined sex
chromosomes, and sex determination system is the same as human: XX-XY (summarized
in Wittbrodt et al., 2002). Thus, it has been suggested that the Japanese medaka may
provide a valuable model to study the disruption of sex differentiation caused by
chemical in human. Secondly, medaka is hardier and less susceptible to disease than the
zebrafish (Wittbrodt et al., 2002). Another advantage of the Japanese medaka is its rapid
development and ease of breeding, producing eggs on a regular schedule under the
appropriate conditions of lighting and temperature. Also, there are closely related marine
and freshwater species of medaka. Because of these advantages the medaka has been
recognized by the international scientific community as an attractive complementary

model system to the zebrafish. Furthermore, a draft genome sequence of the freshwater

Japanese medaka has been assembled and over 20 thousands genes have been predicted
(Kasahara et al. 2007), which allow the application of toxicogenomics to evaluate
toxicity of different chemicals across the genomes of these fish and extrapolate the
observed toxicity in these models to humans and other Species.

Transcriptional profiling methods, like microarray and real- time PCR, are
powerful tools for examining chemical mechanisms or modes of action (MOA) and could
potentially be used to support aspects of regulatory decision making in ecotoxicology
(Ankley et a1 2007). Microarray technology can scan expression profiles of multiple
genes. The medaka has a relatively well-characterized genome. However, it lacks robust
annotation for many gene products. Therefore, because of absence of baseline
information of a large proportion of array spots, full interpretation of data collected by
medaka microarray is impossible. On the other hand, real-time polymerase chain
reaction (Real time -PCR) is a sensitive and reliable technique enabling quantitative
quantification of mRN A in biological samples. Real-time PCR methods have greater
precision for quantification of changes in gene expression than does the microarray. In
addition, the lesser expense of the real time PCR technique relative to that of the
microarray technique allows robust investigation on the studied chemical by examining
more concentrations, organs and replicates. Therefore, we have developed a medaka
HPG PCR array system that combines the quantitative performance of SYBR® Green-
based real-time PCR with the multiple gene profiling capabilities of a microarray to
examine chemical-induced gene expression profiles along the HPG axis. In the present
study, we have selected a suite of the functionally relevant genes associated with the

pathways of concern (HPG) based on literatures. All the genes investigated here either

have a cDNA sequence that is characterized in the NCBI database or have been
sequenced by our group using rapid amplification of cDNA end (RACE) techniques in
this study.

To evaluate the performance of the medaka HPG PCR array system and to
develop the associated data analysis and visualization tools, five chemicals were selected
as model chemicals to which medaka were exposed. The selected model chemicals
included estrogen 170I-ethinylestradiol (EE2), anabolic androgen 17B-trenbolone (TRB),
potent aromatase inhibitor fadrozole, fungicide prochloraz, and ketoconazole. These
chemicals have displayed different MOAS, but all have been shown to adversely affect
animal reproduction through the HPG axis. To elucidate the molecular mechanism
associated with the adverse effects by these chemicals, concentration /time —dependent
experiment design were applied.

I have included three closely related studies as chapters of my dissertation.
However, during my program of graduate, I have also conducted research in the field of
in vitro Endocrine Toxicology and Dioxin/PCB Toxicology. Thus, in addition to the
three chapters included in my dissertation, each of which has been submitted as a
manuscript to a journal, I have completed several other collateral studies and been
involved as a co-investigator in several additional studies in the general area of molecular
toxicology of endocrine disrupting chemicals that have resulted in publications in peer-
reviewed journals. In particular, I have been involved with the development of the
H295R steroidogenesis assay, which currently in the final phases of global validation by
the Organization of Economic Cooperation and Development (OECD). In addition to

completing the Ph.D. degree in the Toxicology track of the Environmental Toxicology

and Zoology, while at MSU I have simultaneously completed an MS in Statistics, for
which I concentrated on techniques appropriate for toxico-genomics. The manuscripts

that I have authored or co-authored while at MSU include the following.

In vitro Endocrine Toxicology

Zhang, X., Yu, R.M., Jones, P.D., Lam, G.K., Newsted, J .L., Gracia, T., Hecker M.,
Hilscherova K, Sanderson T, Wu R.S., Giesy JP. (2005). Quantitative RT-PCR

methods for evaluating toxicant-induced effects on steroidogenesis using the
H295R cell line. Environ Sci Technol. 39(8):2777-85.

Hilscherova, K., Jones, P.D., Gracia, T., Newsted, J.L., Zhang, X., Sanderson, J .T., Yu,
R.M.K., Wu, R.S.S., Giesy, J .P. (2004). Assessment of the effects of chemicals
on the expression of ten steroidogenic genes in the H295R cell line using real-
time PCR. Toxicol Sci. 81, 78-8.

Gracia, T., Hilscherova, K., Jones, P.D., Newsted, J .L., Higley, E.B., Zhang, X., Hecker,
M., Murphy, M.B., Yu, R.M., Lam, P.K., Wu, R.S., Giesy, J .P. (2007).
Modulation of steroidogenic gene expression and hormone production of H295R
cells by pharmaceuticals and other environmentally active compounds. T oxicol

Appl Pharmacol. 225(2), 142-153.

Gracia, T., Hilscherova, K., Jones, P.D., Newsted, J .L., Zhang, X., Hecker, M., Higley,
E.B., Sanderson, J .T., Yu, R.M.K., Wu, R.S.S., Giesy, J .P. (2006). The H295R
system for evaluation of endocrine-disrupting effects. Ecotoxicol Environ Saf.
65(3), 293-305.

Zhang, X., Jones, P.D., Newsted, J .L, Hecker, M., Gracia, T., Hilscherova, K., Yu,
R.M.K., Wu, R.S.S. Giesy, J.P. (2008). Comparison of Human H295R, JEG-3
and Rat R2C Cell lines for Determining Effects on Steroidogenesis, Toxicol Appl
Pharmacol. (Submitted)

Zhang, X., Newsted, J .L, Jones, P.D. Giesy, J .P. (2008). “ToxClust” --- a new data
mining method for the analysis of concentration and/or time-dependent biological

data. J Chem Inf Model (Submitted)

Dioxin/PC B Toxicology

Fung, C.N., Zheng, G.J., Zhang, X., Wong, H.L., Giesy, J.P., Fang Z., Lam. P.K.S.
(2005). Risk Posed by Concentrations of Trace Organic Contaminants in Coastal
Sediments in the Pearl River Delta, China. Mar Pollut Bull. 50, 1036-1049.

So, M.K., Zhang, X., Giesy, J .P., Wong, H.L., Zheng J.G., Lam. P.K.S. (2005).
Organochlorines, and dioxin-like compounds in green-lipped mussels Perna
viridis from Hong Kong mariculture zones. Mar Pollut Bull. 51, 677-87.

Xu, Y., Yu, R.M., Zhang, X., Murphy, M.B., Giesy, J .P., Lam, M.H., Lam, P.K.S., Wu,
R.S.S., Yu, H. (2006). Effects of PCBs and MeSOZ-PCBS on adrenocortical
steroidogenesis in H295R human adrenocortical carcinoma cells. Chemosphere

63(5), 772-734.

Zhang, X., Moore, J .N., Zwiernik, M.J., Hecker, M., Newsted, J.L., Jones, P.D., Bursian,
S.J., Giesy, J.P. (2007). Sequencing, and characterization of mixed function
monooxygenase genes CYP1A1 and CYP1A2 of Mink (Mustela vison) to facilitate
study on dioxin-like compounds. Toxicol App] Pharmacol. (Submitted)

Reproductive and Endocrine Toxicology of Fish

Zhan X., Park, J ., Tompsett, A.R., Hecker,M., Jones, P.D., Newsted, J .L., Giesy, J.P.
(2007). Development and validation of a medaka brain-gonadal-liver axis model
and a real time-PCR array method to facilitate the mechanistic classification of
endocrine-disrupting chemicals (EDCS). Aquat T oxicol. (Accepted)

Zhang, X., Park, J ., Tompsett, A.R., Hecker,M., Jones, P.D., Newsted, J .L., Giesy, J .P.
(2008). Time-dependent transcriptional profiles of genes of the hypothalamic-
pituitary-gonadal (HPG) axis in medaka (0. Iatipes) exposed to fadrozole and
17B-trenbolone. Environ Toxicol Chem. (Accepted)

Zhan X., Hecker,M., Jones, P.D., Newsted, J .L., Giesy, J .P. (2008). Responses of the

Medaka HPG axis PCR array and reproduction to prochloraz and ketoconazole.
Environ Sci Techno]. (Submitted)

REFERENCES

Ankley, G.T., Johnson, RD. (2005). Small Fish Models for Identifying and Assessing the
Effects of Endocrine-disrupting Chemicals. ILAR J 45(4), 469-483.

Ankley, G.T., Villeneuve, D.L. (2006). The fathead minnow in aquatic toxicology: past,
present and future. Aquat. Toxicol. 78(1), 91-102.

Danger, J .M., Tonon, M.C., Jenks, B.G., Saint-Pierre, S., Martel, J .C., Fasolo, A., Breton,
B., Quirion, R., Pelletier, G., Vaudry, H. (1990). Neuropeptide Y: localization in
the central nervous system and neuroendocrine functions. F undam Clin
Pharmacol 4(3), 307-340.

Gracia, T., Hilscherova, K., Jones, P.D., Newsted, J .L., Zhang, X., Hecker, M., Higley,
E.B., Sanderson, J.T., Yu, R.M.K., Wu, R.S.S., Giesy, J .P. (2006). The H295R

system for evaluation of endocrine-disrupting effects. Ecotoxicol Environ Saf
65(3), 293-305.

Hecker, M., Newsted, J .L., Murphy, M.B., Higley, E.B., Jones, P.D., Wu, K., Giesy, J .P.
(2006). Human adrenocarcinoma (H295R) cells for rapid in vitro determination of

effects on steroidogenesis: hormone production. T oxico] Appl Pharmacol 217(1),
1 14-124.

Hilscherova, K., Jones, P.D., Gracia, T., Newsted, J .L., Zhang, X., Sanderson, J .T., Yu,
R.M.K., Wu, R.S.S., Giesy, J .P. (2004). Assessment of the effects of chemicals on

the expression of ten steroidogenic genes in the H295R cell line using real-time
PCR. Toxicol Sci. 81(1), 78-89.

Kasahara, M., Naruse, K., Sasaki, S., Nakatani, Y., Qu, W., Ahsan, B., Yamada, T.,
Nagayasu, Y., Doi, K., Kasai, Y., Jindo, T., Kobayashi, D., Shimada, A., Toyoda,
A., Kuroki, Y., Fujiyama, A., Sasaki, T., Shimizu, A., Asakawa, S., Shimizu, N.,
Hashimoto, S., Yang, J., Lee, Y., Matsushima, K., Sugano, S., Sakaizumi, M.,
Narita, T., Ohishi, K., Haga, S., Ohta, F., Nomoto, H., Nogata, K., Morishita, T.,
Endo, T., Shin, I., Takeda, H., Morishita, S., Kohara, Y. (2007). The medaka draft
genome and insights into vertebrate genome evolution. Nature 447(7145), 714-
719.

Pastva, S.D., Villalobos, S.A., Kannan, K., Giesy, J .P. (2001). Morphological effects of
Bisphenol-A on the early life stages of medaka (Oryzias latipes). Chemosphere
45(4-5), 535-541.

Villalobos, S.A., Papoulias, D.M., Pastva, S.D., Blankenship, A.L., Meadows, J ., Tillitt,
D.E., Giesy, J .P. (2003). Toxicity of o,p'-DDE to medaka d-rR strain after a one-
time embryonic exposure by in ovo nanoinjection: an early through juvenile life
cycle assessment. Chemosphere 53(8), 819-826.

Villeneuve, D.L., Larkin, P., Knoebl, 1., Miracle, A.L., Kahl, M.D., Jensen, K.M.,
Makynen, E.A., Durhan, E.J., Carter, B.J., Denslow, N.D., Ankley, GT. (2007).
A graphical systems model to facilitate hypothesis-driven ecotoxicogenomics

research on the teleost brain-pituitary-gonadal axis. Environ Sci T echno]. 41(1),
321-330.

Wittbrodt, J., Shima, A., Schartl, M. (2002). Medaka--a model organism from the far East.
Nat Rev Genet. 3(1), 53-64.

Zhang, X., Yu, R.M., Jones, P.D., Lam, G.K., Newsted, J.L., Gracia, T., Hecker, M.,
Hilscherova, K., Sanderson, T., Wu, R.S., Giesy, J .P. (2005). Quantitative RT-

PCR methods for evaluating toxicant-induced effects on steroidogenesis using the
H295R cell line. Environ Sci Techno]. 39(8), 2777-2785.

Chapter 2

Real time PCR array to study effects of chemicals on the Hypothalamic-

Pituitary—Gonadal axis of the Japanese medaka

Xiaowei Zhang”, Markus Hecker2‘3‘5, June-Woo Park”, Amber R. Tompsettl‘z, John
Newsted2'4, Kei Nakayama", Paul D. J onesz’5 , Doris Aug, Richard Kongs, Rudolf S.S.

Wu8, John P. Giesy1‘2‘5’7‘8‘.

I Department of Zoology, Michigan State University, East Lansing, MI. USA
2 National Food Safety and Toxicology Center and Center for Integrative Toxicology,
Michigan State University, East Lansing, MI. USA
3 ENTRIX, Inc., Saskatoon, SK, Canada
4 ENTRIX, Inc., Okemos, MI, USA
5 Toxicology Centre, University of Saskatchewan, Saskatoon, SK, Canada
6 Center for Marine Environmental Studies (CMES), Ehime University
7 Dept. Biomedical Veterinary Sciences, University of Saskatchewan, Saskatoon, SK,
Canada

8 Dept. Biology & Chemistry, City University of Hong Kong, Hong Kong, SAR, China

10

ABSTRACT

System models utilizing genomic approaches can be powerful tools for mechanistic
toxicological research. This paper describes the development and validation of a PCR
array for studying chemical-induced effects on gene expression of selected endocrine
pathways along the hypothalamic-pituitary-gonadal (HPG) axis of the small, oviparous
fish, the Japanese medaka (Oryzias latipes). The Japanese medaka HPG PCR array
combines the quantitative performance of SYBR® Green-based real-time PCR with the
multiple gene profiling capabilities of a microarray to examine expression profiles of 36
genes associated with endocrine pathways in brain, liver and gonad. The performance of
the Japanese medaka HPG PCR array was evaluated by examining effects of two model
compounds, the synthetic estrogen, 17a-ethinylestradiol (EE2) and the anabolic androgen,
17B-trenbolone (TRB) on the HPG axis of the Japanese medaka. Four-month-old
medaka were exposed to three concentrations of EE2 (5, 50, 500 ng/L) or TRB (50, 500,
5000 ng/L) for 7 d in a static renewal exposure system. A pathway-based approach was
implemented to analyze and visualize concentration-dependent mRN A expression in the
HPG axis of Japanese medaka. The compensatory response to EE2 exposure included the
down-regulation of male brain GnRH R I and testicular C YPI 7. The down—regulation of
AR-a expression in brain of EE2-exposed males was associated with suppression of male
sexual behavior. Compensatory responses to TRB in the female HPG axis included up-
regulation of brain GnRH R 11 and ovary steroidogenic CYP19A. Overall, the results

suggested that the Japanese medaka HPG PCR array has potential not only as a screening

11

tool of potential endocrine disrupting chemicals but also in elucidating mechanisms of

action.

Keywords: 170r-ethinylestradiol, 17B-trenbolone, expression profile, steroidogenesis,

fecundity, endocrine, fish

12

INTRODUCTION
Knowledge of chemical-induced endocrine disruption is largely limited to the pathways
mediated through several steroid hormone receptors, including estrogen receptors GER)
androgen receptors (AR) and thyroid hormone receptors (ThR). More recently,
increasing efforts have been underway to evaluate alternate pathways of chemical
interaction with the endocrine system such as effects on steroidogenesis, which can alter
the rates as well as absolute and relative concentrations of hormones produced by an
organism by altering the expression of steroidogenic enzymes ( Ankley et a]. 2005;
Gracia et a]. 2006; Hecker eta]. 2006; Hilscherova eta]. 2004; Villeneuve eta]. 2007a;
Zhang eta]. 2005).

Endocrine-Disrupting Chemicals (EDCS) can alter normal patterns of gene
expression either by direct (steroid hormone receptor-mediated pathways) or
compensatory effects (V illeneuve et al. 2007b). Historically, studies of interactions of
EDCS with organisms have focused primarily on a few endpoints in one tissue at one
specific time in the development of an organism. However, considering the complicated
nature of endocrine systems in humans and wildlife, current chemical screening tools are
often limited to a few molecular targets. What is needed is a sensitive, flexible
monitoring tool that allows for the screening of a multiple molecular target genes in
multiple tissues simultaneously at any stage of development and allows for mechanism-
based toxicity prediction and assessment.

A significant degree of evolutionary conservation has been found to occur in the
basic aspects of the hypothalamic-pituitary-gonadal (HPG) axis among vertebrates

(Ankley and Johnson 2005). Teleost fish, such as the Japanese medaka (Oryzias latipes),

13

zebrafish (Danio rerio) and fathead minnow (Pimephales promelas), have been
suggested to be appropriate models for testing EDCS relative to both ecological relevance
and species extrapolation (Ankley and Villeneuve 2006). The Japanese medaka is a
small, oviparous, freshwater fish, native to Asia, for which extensive information on
physiology, embryology and genetics has been developed (Wittbrodt et a]. 2002; Pastava
eta]. 2001; Villalobos et a]. 2003). Recently a marine medaka model (Oryzias
melastigma) has also been developed for ecotoxicological study (Kong et a]. 2007).

Transcriptional profiling methods, like microarray and real-time (quantitative)
polymerase chain reaction (RT-PCR or Q-PCR), are powerful tools for examining
chemical mechanisms or modes of action (MOA) and could potentially be used to
support aspects of regulatory decision making in ecotoxicology (Ankley et a] 2007). In
the present study a Japanese medaka hypothalamic-pituitary-gonadal (HPG) PCR array
system was developed that combines the quantitative performance of SYBR® Green-
based RT-PCR with the multiple gene profiling capabilities of a microarray to examine
chemical-induced gene expression profiles along the HPG axis. Based on literature, a
suite of functionally relevant genes associated with the pathways of concern (HPG) were
selected. All the genes investigated here either have a cDNA sequence that has been
characterized in the NCBI database or have been sequenced using rapid amplification of
cDNA end (RACE) techniques in our laboratory.

To evaluate the performance of the Japanese medaka HPG PCR array and to
develop the associated data analysis and visualization tools, fish were exposed to two
model chemicals, the synthetic estrogen, 170r-ethinylestradiol (EE2) and the anabolic

androgen, l7B-trenbolone (TRB). The aims of the study were: 1) to select a suite of

14

functionally relevant genes to develop a Japanese medaka HPG model; 2) to clone and
sequence the selected genes that lack cDNA evidence; 3) to develop SYBR Green based
RT-PCR methods for the selected genes; 4) to conduct 7-d exposures with EE2 and TRB
to examine the gene expression patterns in brain, liver and gonad; 5) to test the
hypothesis that the chemicals can induce concentration-dependent, organ-specific gene
expression response patterns; and 6) to compare gene changes and effects at other

biologically relevant levels such as fecundity.

15

MATERIALS AND METHODS

Chemicals and regents
17 a-ethinyl estradiol (EE2), 17B-trenbolone (TRB) and dimethyl sulfoxide (DMSO)

were obtained from Sigma (St. Louis, MO).

Animals

Male and female wild-type 0. latipes were obtained from the aquatic culture unit at the
US Environmental Protection Agency Mid-Continent Ecology Division (Duluth, MN,
USA). The fish were cultured in flow-through tanks in conditions that facilitated
breeding (23-24°C; 16:8 light/dark cycle), and that were in accordance with protocols
approved by the Michigan State University Institutional Animal Care and Use Committee

(MSU-IACUC).

Total RNA isolation and reverse-transcription PCR

Total RNA was extracted from tissue samples using the Agilent Technologies Total RNA
Isolation Mini Kit (Agilent Technologies, Palo Alto, CA) according to the manufacturer’s
protocol. Purified RNA was stored at -80 °C until analysis. First-strand cDNA synthesis
was performed using Superscript III first-strand synthesis SuperMix and Oligo-dT
primers (Invitrogen, Carlsbad, CA). Briefly, a 0.5 to 2 pg aliquot of total RNA was
combined with 1 pL of 50 pM Oligo(dT)20, 1 pL of annealing buffer, and RNase-free
water to a final volume of 8 pL. Mixes were denatured at 65 °C for 5 min and then

quickly cooled on ice for 2 min. Reverse transcription was performed after adding 10 pL

16

2X first-stand reaction mix, and 2 pL SuperScript III/RNaseOUT enzyme mix. Reactions
were incubated at 50 °C for 50 min and, on completion, were inactivated at 85 °C for 5
min. To digest RNA, 1.25 pL RNase H (Invitrogen, Inc., Carlsbad, CA) was added
before incubation at 37 °C for 30 min. The cDNA synthesis reactions were stored at -20

°C until further analysis.

Gene selection and model development

A total of 36 genes representing key signaling pathways and functional processes within
the Japanese medaka HPG axis were selected for study based on the teleost “Graphical
Systems Model” previously proposed by (V illeneuve et al. 2007b) and Japanese medaka
species-specific literature (Table 2.1). In addition, reference genes, B-actin, 16S rRNA
and RPL-7, were selected as internal quantitative controls. The Japanese medaka HPG
transcriptional model was constructed and visualized using GenMAPP 2.1 (Salomonis et

a]. 2007).

Cloning and sequencing

Of the 36 selected HPG genes, 14 genes did not have cDNA sequences available in the
public NCBI Genbank database and cDNA cloning was conducted based on predicted
transcript sequences. Briefly, the corresponding homologous genes were identified from
the ensembl Japanese medaka genome (hfipJ/wwwensemblorgZngias latipes/index.
ht_r_n_l). Gene-specific primers were designed based on the predicted sequences for each of
the studied genes. 5’-RACE and 3’-RACE PCR reactions were performed using a BD

SMART RACE cDNA amplification kit (BD-Biosciences Clontech, Palo Alto CA)

17

according to the manufacturer’s protocol. Purified PCR products were cloned into a
plasmid vector or directly sequenced using the corresponding primers by an ABI 3730

Genetic Analyzer (Applied Biosystems, Foster City, CA).

Real-time PCR reaction

Real-time Q-RT-PCR was performed by using an ABI 7900 high throughput real time
PCR System in 384-well PCR plates (Applied Biosystems, Foster City, CA) (Table 2.2).
PCR reaction mixtures for one hundred reactions contained 500 pL of SYBR Green
master mix (Applied Biosystems, Foster City, CA), 2 pL of 10 pM sense/anti-sense gene-
specific primers, and 380 pL of nuclease-free distilled water (Invitrogen). A final
reaction volume of 10 pL was made up with 2 pL of diluted cDNA and 8 pL of PCR
reaction mixtures using a Biomek automation system (Beckman Coulter, Inc., Fullerton,
CA). The PCR reaction mix was denatured at 95 °C for 10 min before the first PCR
cycle. The thermal cycle profile was: denaturizing for 15 s at 95 °C; annealing for 30 s at
60 °C; and extension for 30 s at 72 °C. A total of 40 PCR cycles were used. PCR
efficiency, uniformity, and linear dynamic range of each Q-RT-PCR assay were assessed

by the construction of standard curves using DNA standards.

C hemica] exposure

Japanese medaka (14 wk old) were acclimated in 10-L tanks filled with 6 L of carbon-
filtered water for a period of 12 (1 prior to initiation of experiments. Fish were held at
24°C with a 16:8 light/dark cycle. Females were first separated from males by visual

morphological determination and 5 female and 5 male fish were put into each tank.

18

Half of the water in each tank (3 L) was replaced daily with fresh carbon-filtered water.
Overall mortality for all fish during the acclimation period was one. After the
acclimation period, fish were exposed to EE2 or TRB in a 7-d static renewal exposure
scenario. Each treatment was replicated in triplicate tanks and consisted of a vehicle
control (DMSO with a final concentration of 1:10000 v/v water), 5, 50, and 500 ng/L
EE2 and 50, 500, and 5000 ng/L TRB. Half of the water in each tank (3L) was replaced
daily with fresh carbon-filtered water dosed with the appropriate amount of chemicals.
Water quality parameters (temperature, pH, hardness, dissolved oxygen, ammonia-
nitrogen and nitrate-nitrogen) were measured daily. Eggs produced during the previous
day were counted and recorded before the replacement of water. No mortalities were
observed in any treatment during the exposure period. At the end of the 7-d exposure
period fish were euthanized in Tricaine S solution (Western Chemical, Femdale, WA,
USA), and total weight and snout-vent length were recorded for each fish. For gene
expression analysis, 4-6 males and females were randomly sampled from the three
replicate tanks of each treatment. Tissues from brain, liver, and gonads were collected
and preserved in RNAlater storage solution (Sigma, St Louis, M0) at -20°C until
analysis. The quantification of target gene expression was based on the comparative
cycle threshold (Ct) method with adjustment of PCR efficiency according to the
procedures described previously (Zhang eta]. 2005). To increase the reliability of
comparative Ct method —based gene expression quantification, the average Ct value of
multiple reference genes was used as reference Ct. ,B-actin, 16S rRNA and RPL-7 were

used as reference genes in brain and gonad tissues. Only 16S rRNA and RPL-7 were used

19

as reference genes in liver tissue because the hepatic expression of ,B-actin has been

shown to be responsive to estrogenic chemicals (Zhang and Hu 2007).

Data analysis

Gene expression was calculated as fold-change relative to the average expression in the
vehicle control. Statistical analyses were conducted using the R project language
(http://www.r-project.org/). Prior to conducting statistical comparisons, normality was
evaluated by Shapiro-Wilks test and if necessary, data were log-transformed to
approximate normality. Differences were evaluated by AN OVA followed by a pair-wise
t-test. Levels of statistical significance are p < 0.05 unless specified. To examine the
relationship between the expression level of steroid hormone receptors and other HPG
genes in different tissues, Spearman rank correlation analysis was conducted
independently for females and males. Because each receptor gene was compared with
multiple genes, the chance of false correlation increased. Therefore correlations with p <

0.01 were considered to be statistically significant.

20

 

.8382 Dm £85888me amazoweasov Eopoammxohiném QmEfim
*wmvom _ Dm mwmocoonEBm Acme—0:58 mmdmv own—088nm E FED
EVER. 28882205 0332343 223m 389:. aeosooso 5:0
*cmvofl 3m £85onng 83268 <00 ESE—wifioéxeam EDEE
omowwoOQ £85on8on £085 >833on 88m $8828on «Em.
smegma. Canaan 0:3 = 558:3; 2 Es
838m? Seneca 0:9A _ 55mo=§> N 8:
_N:.wo £85on0me 3‘: om: oEoEooSD meD
3.03 m>< £88828on m3 omvm oEoEoofiD MEAED
wcomwa £85828on <2 omvm 0:858pr VEAED
owwm3m< 285828on m: omvm oEoEooSD ME FED
mommwoma. 8582 8E8 E 25 8&8“: 5:8 Sm 8:6
whonmomas 8888 023$ = 09$ 8382 5:0 EN Ean
mnohmomxx 8388 839m H cab 8888 5ch E ExzD
Sachem? 8382 28on Son 8382 cowobmm MEN
v3me 838m: 28on «an? “Smooch cowobmm SEE
momchomaq. 8:58.. 388m 8:58.. Sweats/E 5-5x
mmmgom/E 28::on 0858: mammofirfiaobopacow ombéoEEm EeDm
ommgomxx 2888: 2858: mammo—oTfimobonmcow ogre—802 E~E=D\E
ommfivom/E 28883 = 0888: mEmeoHéEobocgow ombéoonD E Eszo
Ago—o8“? omcomme cowobmm A :EowotonD E DED
hcammomz 88am“: cowobmm 8&8: :EowotocD EE DED
mocowo 3838 cowobmm I EcowotonD E DED
mmfl _ > 8838 cowobmm mme 5853 NSE ESEV
manor? 8%.. £856 <m 8E oeoaooso 5&8
35.5mm fewofiu «Ea Z 28D 833555<

 

88m? Eta MDm Om: 8.82: on: me am: 280 ~.~ 03.;

21

88m 28 8 888:8 8:288 <ZQo .8th ...

 

82: 58
wowenm
2.88...
.228 _ 8
22.028
.2828
.8828
84% _ 8
8:38
18:58
88:88
.8328
8228
E28 _ 8

8% 8:880m
85w 8:880m
88w 888.8%
8888:
2888:
888822
2888888on
8888 294
2888:
88883
8888:
8888 282
8888 828m
8888 8588

EH 8882 .mEomonE
88m 80m
<58 me—

88:88 880 89m 2585

m 88 .2588

880 < 88 838%..

8880888 28898888098
8888 88882: 32.88 :wi
88838 .89 .2888: 9828824
> 2288882

880 288 8888; 82088.20
8888 88888. 3828 33
8888 2888: 888884
8888 8888: 882588 22:0m

m EEK
2.2.8 2.
ME

2 2.3.35
mm 2.5.2.62
2% 2.3.82
82.28
EEQE
2.5

x 8&8sz
2EEO
EEQE
EEE
EEME

2.280 .—.N ozah

22

Table 2.2. primer sguences of selected medaka HPG axis genes

 

 

Abbreviation Sense Antisense

C YP3A GAGATAGACGCCACCTTCC ACCTCCACAGTTGCCTI'G
Annexin max2 CTGATCGTGGCTCTGATGAC CTGCTGAGGTGTTCTGGAAG
CHG H TGGCAAGGCACTGGAGTATCAC CTGAGGCTTCGGCTGTGGATAG
CHG HM GGAGCCATI’ACCAGGGACAG AAGTTCCACACGCAAGATI’CC
CHG L TCCTGTCTCTGACTCTGAATGG GC‘I'I’GGCTCGTCCTCACC
cGnRH II TGTCTCGGCTGGTTCTAC GAGTCTAGCTCCCTCTI‘CC
mfGnRH GTGTCGCAGCTCTGTGTI’C AGTATI'TCAGTTCTCGCTI’CCC
anRH GATGATGGGCACAGGAAGAGtG GGGCACWGCATCTI’CAGGA
AR-a ACCTGGCTCACTTCGGACAC TCTGACGCCGTACTGCTCTG
ER-a GAGGAGGAGGAGGAGGAGGAG GTGTACGGTCGGCTCAACWC
ER-fl GCTGGAGGTGCTGATGATGG CGAAGCCCTGGACACAACTG
GnRH RI TCCGACGAGCCGCATCTG GATGAAGCCGACGACGATGAC
GnRH R11 GCAGCGGCACAGACATCATC GGACAGCACAATGACCACAGAC
GnRH RIII ACTI'CCAGAGGAGCCAGTTGAG GCCAGCCAAGAGTCGTTGTC

C YPI 1 B CTAGACGACGTGGCGAAAGACT ccrcrscrccrcrrccrrcrce
C YPI 9A CTCTTCCTGGGTGTI'CCTGTTG GCTGCTGTC‘ITGTGCCTCTG
CYP19B TCCTGATAACCCTGCTGTCTCG GTTGGTCTGCCTGATGCTGTTC
C YP 1 7 CGACCACCACCGTACTCAAATG TCTGGATAATGGATCAGGTAGGtG
VTG I ACTCTGCTGCTGTGGCTGTAG AAGGCGTGGGAGAGGAAAGTC
VTG II TCGCCGCAAGAGCAAGAC CTGGAGGAGCTGGAAGAACTG
StAR TGACAGGTI‘I'GAGAAAGAATG CAATGCGAGAACTI’AGAAGG
HMGR CTGCTGCTGGCTGTCAAG GCTGGCGGCTGC‘ITI‘ATG

C YP21 CAGCCAGCAATGTCATCAC TCAGCAGTGGGAAGGAATC
CYPI 1A GCTGCATCCAGAACATCTATCG GACAGCTTGTCCAACATCAGGA
33-HSD GGGCGGGACGAAACTCAG GGAGGCGGTGTGGAAGAC
FSHR TTCAGGCCACTGATGATGTTATCG CCTI'CGTGGGTTCCAGTGAGT
LHR GTCCTGGTCATCCTGCTCGTIAG AACCGGGAGATGGTCAG'ITI’GT
LDLR GTGCTACGAAGGCTACGAGATG AGGTCAATGCGGCGGATI'TC
GTHa GCAGAACGGAGGGATGAAGGAG ATTGGAGTAGGTGTCGGCTGTG
Ne urOpepY CTTCCACAGTCAAGTI’ACAAC TGATCTGCAAGGACGAATG
LHbeta GCCAGCCAGTCAAGCAGAAG TCCACCGTATGACAGCCAGAG
HDLR TCTGCCGAACTGTCACTGTC CCACCTGGTCGTCGATGATG
20,3-HSD CAAAGGCATCGGCCTGGCCATTGT GGCGGCGTTGWGATAAGGACATC
Activin BA GATGGTGGAAGCAGTGAAG 'I'I'CTI'GATGGCGTTGAGTAG
Activin BB GGCTAATCGGCTGGAATG CATGCGGTACTGGTTCAC
Inhibin A CGTTTCCCTTCCAGCCTTC AAGAGCG'ITGCGGATGAG

16S CGATCAACGGACCGAG‘I‘I’ACC AATAGCGGCTGCACCATTAGG
Beta actin GATGAAGCCCAGAGCAAGAGG CATCCCAGTTGGTAACAATACGG
RPL- 7 GTC GCC me 070 CAC AAA G AAC TI’C AAG ccr GCC AAc AAc

 

23

 

 

 

850885 1...
emcoomumg IT
85... mg Jr

85:00 I0!

 

 

 

 

 

 

8883 mm... “m

0

OL 09 09 017 08 OZ 0L

enema; ed 8669 peielnwnoov

 

 

 

 

 

 

 

80
m o v N o
_
0
L
0
Z
0
8
0
.7
O
C..
I 0
88888 1.:
429.8188 IT 9
8% 88 J... I 0
.968 .6.
I. /_
0

 

 

mhsmoaxm «mm n<

8:08 8828 :88 God V 8:85-26 888.86 88588

a 888:. 83:82“ 8? 82.8 8 888 c 888:8 some .888 8888 m :88 888:8 8888 88 88 .8 898::
9:83:80 80E 8: 88888 «ED 88 6-8 a 5 Amy my: 8 A<v ma 8 8898 8388 E 58:88 0283280

elewa; Jed $569 peremwnoov

3 288

24

Table 2.3. Chemical induced effects on medaka gonadal-somatic index
(GSI), hepatic-somatic index (HSI) and brain-somatic index (BSI) 8’ b’ °.

 

 

 

Female Male
Treatment BSI GSI HSI BSI GSI HSI
Control 1.61 10.85 4.18 1.64 1.28 2.14
5 ng EE2/L 1.67 9.40 6.40 1.70 1.68 3.80*
50 ng EE2/L 1.43 10.08 3.67 1.84 1.62 3.69
500 ng EE2/L 1.64 10.01 5.48 1.44 1.96 3.86*
50 ng TRB /L 1.00 14.51 5.70 1.46 1.43 2.27
500 ng TRB /L 1.32 15.72* 4.87 1.48 1.73 3.90*
5000 ng TRB /L 1.51 11.96 3.66 1.66 1.34 2.38
‘ n = 4 ~6,

b 881 = brain weight x 100/ body weight;
681 = gonad weight x 100/ body weight;
HS] = liver weight x 100 / body weight

° * p-value < O. 05 comparing control by pairwise Wilcox test.

25

Figure 2.2. Volcano plots of chemically induced changes in gene expression pattem in
males and females. Data are from medaka exposed to 500 ng EE2/L or 5000 ng TRB/L.
Genes plotted farther from the either the x or y- axis have larger changes in gene
expression. Thresholds for fold-change (vertical lines, 2—fold) and significant difference

(horizontal line, p < 0.01) were used in this display.

26

Smodmmcmco Eon

NONIVioimi
_______.

 

 

9m

 

 

 

 

 

\ f 56230 I
35me 53:51 _ O 7:195
.
95.1%,... . O 3595 T
—
N 035196 I
. O
_ 3:. FES
.
Adi—ton: 1.51... m: ooom un.
amodomcmco 28
m N F o N1 v:
_ _ w u _ _ e
O 0%? J. O
O _ 0 _ O
O 8 O. O I
_ .
_ _
_ _
IIIIIIIII .llllLIll’lIlllllll
" " “menafgo
_ _
,memflnio .0 u 1
. —
_ _
. _

 

 

 

§me5 Jam 9. com um

(0L60‘1)en12/\—d

(0L601lanleA-d

Amwodwmcmco Eon.

 

 

 

mi mi mi
_ _ _ b _
0
3:05
0
5:95
0
meco
0
QEImzo

 

 

 

3.3): Jxmm—P m: ooom “0

88459.20 yo...

3 2.0an
____________H

jzfmm
3:55 0 O

0‘ .E u
3.49.0 . 4 I :0
. 7.: 0F)
0 3:. OF)

 

 

Ens: .fimm as com n<

OL- * 1?-

EL—

(OLBO‘IlGWBA-d

(0 L50‘|)an;2/\_d

27

Table 2.4. Changes in expression for BHG genes in EE2-exposed medaka fish. Genes
exhibited an over two-fold or significant change (p-value < 0.05) in expression between
control and exposed medaka are listed. 8"

 

 

 

 

EE2/Male EE2/Female
Tissue Gene 5ng/L 50ng/L 500m/L 5ng/L 50ng/L 500ngLI:_
ER-a -1.14 -2.13 -1.72 1.93 2.19 1.62
ER-fl 1.09 -1.52 -2.1 2.26 2.08 1.83
AR-a -2.27* -2.29" -2.08* 1.26 1.51 1.54
NeuropepY 1.94* 1.19 -1.47 1.26 1.47 1.67
mfGnRH -1.23 1.12 -1.4 1.12 1.13 1.02
. anRH 1.15 1.29 -1.89 -1.29 1.39 -1.18

B'a'" cGnRH II -1 . 14 -1.57 1.05 1.09 1.24 1.95
GnRH RI -1.39 -3.98 -10* 3.06 2.78 2.31
GnRH RII -1.2 -1.45 -1.61 1.65 1.81 1.34
GnRH RIII 1.15 -1.74 -1.98 2.09 2.83 1.81
GTHa -1.52 -13.33 -2.62 1.37 -2.29 -8.26
LH beta 1.18 1.11 -1.02 1.6 -2.26 —3.58
CYP1QB 1.2 1.18 2.17"“ 1.16 1.11 1.92“"
ER-a -1.18 -1.15 -1.26 -1.35 -1.08 -1.69
ER-fi -1.13 1.06 -1 1.33 -1.14 1.47
AR-a -1.25 1.24 1.27 1.06 -1.02 -1.62
FSHR -1.02 1.1 -2.31 1.11 1.02 -1.27
LHR -1.04 1.28 -1.09 -1.78 1.14 -1.76
HDLR -1.86 1.03 -3.13 1.17 -1.23 -1.54
LDLR 1.46 1.07 -1.87 1.13 -1.09 1.05
HMGR -1.59 1.06 ~2.31 1.12 2.62 3.77
StAR -1.42 1.06 -1.47 -1.22 1.14 1.4

Gonad CYP11A 1.01 1.25 -2.36 -1.14 1.12 -2.72
CYP11B -1.52 1.01 -3.11 1.32 3.48 4.11
CYP17 -1.09 -1.22 -8* 1.42 1.25 -1.98
CYP19A -1.1 1.6 1.2 -1.31 -1.07 -5*
CYP21 -1.39 1.22 1.15 1.04 3.48 3.86
ZOB-HSD -1.28 1.66 -2.77 1.52 1.01 -1.06
3B-HSD -1.17 1.3 -2.15 -1.17 -1.17 -2.08
ActinBA -1.22 1.45 1.06 1.18 3.28 4.14
ActinBB -1.51 1.04 -1.01 1.06 -1.08 1.41
Inhibin A -1.29 1.1 -2.27 -1.55 -1.16 -2.08
ER—a -1.15 4.16“ 12.33”” -1.35 1.32 1.59
ER-fl -2.17 -1.57 -1.65 1.18 -1.17 1.63
AR-a 2.94 5.49 2.3 9.88 1.23 2.29
VTG l -1 .04 164*" 3205*“ 1.29 1.27 1.33

Liver VTG II 1 .78 325*” 9004*“ -1 .52 1.02 1.21
CHG H -2.19 2.65 67.8“" -1.28 1.27 2.28
CHG HM -1.74 10.8*** 111*“ -1.36 -1.13 -1.03
CHG L 2.83 11.5*** 1270*“ -1.47 -1.19 1.02
CYP3A -1.53 1.6* -3.13* -1.1 -1.18 -1.12
Annexin max2 1.22 -1.08 -1.76* 1.68 -1.22 -1.46

 

‘ animal replicate (n= 4-6).
b all P < 005, M P < 00], in" P < (1.001

28

RESULTS

1. Construction of medaka HPG pathways focused PCR array.

The 36 HPG genes selected in this study included peptide hormones, steroid receptors,
peptide hormone receptors, steroidogenic genes, selected estrogen responsive genes,
catabolic gene, lipid receptors, and yolk precursor genes (Table 2.1). For the 14 genes
that lacked transcript evidence, their cDNAs were cloned, sequenced, and submitted to
GenBank/DDBJ/EMBL. The 14 gene partial transcript sequences and their Accession

Numbers are HMGR (EU159456), CYP21 (EU1594512 ), CYP1 IA (EU159458), 3,6-HSD

(EU159459), FSHR (EU159460), LDLR (EU159461), GT Ha (EU047760), NeuropepY
(EU047761), LH-fl (EU047762), HDLR (EU159466), 20,8-HSD (EU159462), activin BA
(EU159463), activin BB (EU159464), and inhibin A (EU159465). In addition, three
house keeping genes including ,B-actin, 16S and RPL-7 were also measured as references
of gene expression calculation. Of the 39 genes targeted in this study, 16, 15, and 24
genes were measured in the brain, liver and gonad tissue, respectively.

The Japanese medaka HPG real-time PCR array system was designed and
optimized based on the SYBR Green detection method, which makes this PCR array
system flexible and widely applicable. The generation of single, gene-specific amplicons
of each primer set was confirmed by the real time PCR dissociation curve of each gene
and by agarose gel electrophoresis characterization. The RT-PCR results demonstrate the
high degree of plate-to-plate, run-to-run and replicate-to-replicate reproducibility inherent

in the technology.

29

2. Chemical-induced effects on fish fecundity and indices of body condition

Exposure to EE2 did not affect fecundity, but exposure to TRB significantly reduced
fecundity (Figure 2.1). There was no statistically significant effect on the accumulated
egg production for any Japanese medaka exposed to EE2. No statistically significant
differences in fecundity were observed between the control and 50 ng TRB/L for any
durations of exposure. However, after 3 d of exposure, egg production by fish exposed to
500 or 5000 ng TRB/L was significantly less than that of the unexposed fish.
Furthermore, the rate of egg production in the 5000 ng TRB/L exposure almost
completely inhibited egg production by the end of the study.

Exposure to either TRB or EE2 resulted in statistically significant increases of the
gonad-somatic index (GSI) and hepatic—somatic index (HSI) determined at the end of
study (Table 2.3). Exposure to 5 or 500 ng EE2/L resulted in significantly greater HSI
values in male Japanese medaka. No treatment-related effects of EE2 were observed for
females. The only statistically significant effects of TRB were greater HSI in males and

GSI in females exposed to 500 ng TRB/L.

3. Chemical-induced effects on HPG gene expression

To evaluate overall changes in gene expression patterns for all tissues in males and
females exposed to 500 ng EE2/L and 5000 ng TRB/L were first examined by ‘volcano
plot’ where Logz-transformed fold-changes in gene expression were plotted against t-test
p-values (Figure 2.2). Genes plotted farther from the central axis’s have greater fold-
changes and p-values. In male Japanese medaka exposed to 500 ng EE2/L (Figure 2.2A),

7 genes were significantly up-regulated in liver. The magnitudes of change for different

30

genes ranged from 3- to 9000-fold, and genes displaying such an increase in abundance
included VTG I, VTG II, choriogenin L (CHG L), choriogenin Hminor (CHG HM),
choriogenin H (CHG H) and ER-a. Only one gene, CYP3A, was down-regulated by EE2
in the liver. EE2 caused no statistically significant alteration in genes in the liver of
females, but gonadal CYP19A was significantly less (-5 fold) and brain CYP19B was
significantly greater (+1.92 fold) relative to that of the controls (Figure 2.2.B).

Exposure of TRB caused different gene expression profiles from that of EE2 in
both males and females (Figure 2.2.C, D). Expression of CHG L, CHG HM, VTG I and
VTG II in liver were significantly down-regulated by TRB in both males and females.
Furthermore, greater expression of gonadal aromatase (CYP19A) and less expression of
brain aromatase (CYP19B) was observed in females exposed to TRB, but not in males.
Other effects on gene expression of females included a greater expression of CHG L, and

lesser expression of both CYP3 and HMGR in liver.

3.] EE2- induced gene expression profile

Exposure to EE2 resulted in concentration-dependent alterations in gene expressions in
livers of Japanese medaka that were specific to males (Figure 2.3; Table 2.4).
Statistically significant up—regulation was observed for VTG I, VTG II, CHG L, and CHG
HM in males exposed to 50 or 5000 ng EE2/L. CHG H was less sensitive to EE2
exposure than for CHG HM and CHG L for which significant changes were only
observed in fish exposed to 500 ng EE2/L. Expression of ER-a was in liver was directly

proportional to EE2 concentration, and were up-regulated 4.5-fold and 12-fold by

31

exposure to 50 and 500 ng EE2/L, respectively. Expression of C YP3A was significantly
down-regulated in livers of males exposed to 500 ng EE2/L.

Exposure to EE2 for 7 d affected fewer genes and caused lesser fold changes of
gene expression in testes, ovaries and brains than in livers of male Japanese medaka.
Exposure to 500 ng EE2/L caused significant down-regulation of CYP1 7 and CYP19A
expression in testes and ovaries, respectively. In brains of males, exposure to EE2
resulted in significant down-regulation of expression of AR-a at all concentrations and
concentration —dependently decreased the expression of GnRH R] with a 10-fold change
at 500 ng EE2/L. Furthermore, EE2 exposure down regulated neuropep Y at 500 ng
EE2/L, and up-regulated the expression of CYP19B at 500 ng EE2/L. Expression of
CYP19B was significantly up-regulated (approximately 1.9- fold) in brains of females

exposed to 500 ng EE2/L.

3.2 TRB- induced gene expression profile

TRB differentially regulated gene expression along the HPG axis in Japanese medaka
(Figure 2.4; Table 2.5). In livers of females, exposed to 5000 ng TRB/L the expression
of VTG I, VTG II, CHG L, CHG H, or CHG HM was significantly down-regulated.
Conversely, the expression of C YP3A and annexin max2 in liver of females was up-
regulated in a concentration —dependent manner. In livers of males, TRB exposure
significantly down- regulated VTG I, VTGII, CHG H, and CHG HM expression when
they were exposed to 500 or 5000 ng TRB/L. A bimodal response was observed for
CHG L where a statistically significant up-regulation in expression was observed in

males exposed to 50 or 500 ng TRB/L and a statistically significant down-regulation of

32

expression when exposed to 5000 ng TRB/L. Exposure to 50 ng TRB/L did not affect
the expression of the selected genes in testis or ovaries. However TRB down- regulated
HDLR and up-regulated CYP19A in ovaries at 500 and 5000 ng TRB/L. In testis, TRB
exposure down regulated expression of HMGR, StAR and CYP1 I B. In brain, exposure to
TRB did not affect the gene expression in males, but up—regulated cGnRH II and down-

regulated CYP1 9 in females.

4. Correlations between expression of receptors and other genes
Expression of receptor genes and other genes in liver of both males and females were
inter-correlated. Liver ER-a did not correlate to either ER-fl or AR-a in either males or
females (Table 2.6). However, ER-a was significantly correlated with CHG HM, CHG H,
CHG L, VTG I and VTG II in both sexes. In male Japanese medaka, hepatic ER-fl was
significantly correlated with CYP3A.

Furthermore, the expression of CYP19B in brain was significantly and negatively
correlated with the ovary expression of CYP19A (Figure 2.5). But this relationship was

not significant in males.

33

Figure 2.3 Striped view of concentration dependent response profile in EE2 exposure
of male Japanese medaka. Gene expression data from medaka treated by 5, 50 and 500
ng EE2/L are shown as striped color sets on the selected endocrine pathways along the
medaka BHG axis. The legend listed in the upper right corner of the graph describes the
order of the three EE2 concentration and the eight colors designating different fold
thresholds. LH, lutinizing hormone; FSH, follicle-stimulating hormone; E2, 176-
estradiol; T, testosterone; HDL, high-density lipid; LDL, low-density lipid.

Medaka HPG axie Map

W

ColorSete

 

 

 

 

    
 

Hypothalamus Gonadotropin releasing hormones

“an“ A in an r p)

 

[EEC-—

Pituitary

 

 

 

 

 

 

 

   

 

gonadotrophe FSH L”
m Gonad Aciivln BA \
ER hem Acuvin BB \
Inhibin A

Progesterone —> 17-Hydroxyprogeeterone

l—q F

Preemnolom 11-deoxyeorlieol

 

Aldoetenedione

m“ Testosterone —-

V
.1

 

 

 

 

 

 

Cholesterol Cholesterol ‘7': WWW“) Betndlol -—
OWN-MM” WW“ 20betas 11-ketoteetoeterone -—> KT
I membrene membrene
E __,,,,,,,,__,
element
r’ "‘ ""‘ rm?”
cvp

 

 

 

 

<'

 

 

Figure 2.4

thresholds.

estradiol; T, testosterone; HDL, high-density lipid; LDL, low-density lipid.

Medaka HPG axis Map

Exmssion Dareset

Color Sets:

 

 

 

Hypothalamus

Gonadotropin releasing hormones

a In nup\

 

 

 

 

 

 

 

 

 

 

 

  
 

 

 

 

 

 

  

 

 

 

 

 

Striped view of concentration dependent response profile in TRB exposure
of female Japanese medaka. Gene expression data from medaka treated by 50, 500 and
5000 ng TRB/L are shown as striped color sets on the selected endocrine pathways along
the medaka HPG axis. The legend listed in the upper right corner of the graph describes
the order of the three TRB concentration and the eight colors designating different fold
LH, lutinizing hormone; FSH, follicle-stimulating hormone; E2,

 

“fl—aIll
Pituitary \ -i-
gonadotrophs FSH L"
an al , Gonad \_ i
5"” \‘MI. 'l'l'
@jjmpm mane! m...
Progesterone —-> 17-Hydroxyprogeeterone
HDL
tot 1m m ““7
-HMG-CoA
- T_eetosterone- —'> T
Cholesterol? Cholesterol 171.20b-P(MIS_) Estrada:
WWI WW MS 11-ketoteetoeterone ——> KT
membrane membrane
_, ”W
m“ We,
—» mag-ass
-

 

 

 

 

1713-

Table 2.5. Changes in expression for BHG genes in TRB-exposed medaka fish. Genes
exhibited an over two-fold or significant change (p-value < 0.05) in expression between
control and exgsed medaka are listed. “‘b

 

 

 

 

Male Female
Tissue Gene 50ng/L 500ng/L SOOOng/L 50ng/L SOOng/L 5000ng/L
ER-a 1.61 2.86 1.14 -1.77 —1.12 1.11
ER-[i 1.18 1.68 1.01 -2.5 -1.01 -1.11
AR-a -1.52 -1.12 -1.17 1.03 -1.12 -1.07
NeuropepY 1.11 1.66 1.1 1.21 1.97 -1.19
mfGnRH -1.04 -1.17 -1.17 1.07 1.36 -1.03
_ anRH 1.26 1.62 1.03 -1.06 1.31 -1.52
B’a'“ cGnRH // -1.46 -1.41 -154 2.21 " 2.19" 1.88"
GnRH RI -1.21 1.9 -1.27 -1.44 -1.04 1.27
GnRH Rll 1.06 1.08 1.05 -1.07 -1.03 -1.48
GnRH RM 1.34 2.69 1.02 -1.15 1.05 1.34
GTH alpha -4.07 -13.2 -2.53 -333 2.36 -1.03
LH beta -3.34 -3.23 -1.75 -100 2.05 -2.98
CYP19B -1.08 1.04 -1.43 —1.07 1.06 -2.14**
ER-a -1.36 -1.48 -1.15 -1.1 2.24 313*
ER-[i -1.09 -1.59 -1.08 -1.15 1.2 1.34
AR-a -1.13 -1.63 -1.31 -1.02 1.25 -1.02
FSHR -1.29 —1.57 1.27 1.09 1.38 2.26
LHR -1.04 -1.4 -1.22 1.01 1.88 1.79
HDLR -1.77 -3.73 -1.58 -1.4 -2.59* -2.98*
LDLR -1.27 -1.74 1.3 -1.05 —1.87 -1.18
HMGR -1.4 -2.79** -1.72 —1.26 3.73 3.06
StAR -1.34 -2.18** -2.06* -1.51 1.36 1.23
CYP11A 1.08 -1.43 -1.34 -1.16 1.72 1.17
Gonad CYPttB -1.61 -2.58* -1.56 -1.06 5.15 3.89
CYP17 -2.22 -2.49 -1.19 -1.15 1.23 -1.06
CYP19A 1.27 -1.17 -1.54 -1.19 4.14 5.82"
CYP21 1.05 -1.73 -1.63 -1.96 4.49 2.7
ZOB-HSD -1.13 -2.3 -1.42 -1.17 -1.34 -1.12
38-HSD -1.25 -1.69 -1.07 -1.32 2.1 3.07
ActinBA 1.17 ~1.1 -1.36 -1.32 4.73 4.4
ActinBB -1.29 -1.34 -1.31 -1 -1.52 -1.92
lnh/binA -2.02 -2.5 -1.19 -1.55 1.36 2.37
ER-a -1.75 -4.94** -3.26* 1.02 1.29 -2.51
ER-[t 1.58 -1.26 1.25 1.06 2.22 2.56
AR-a 2.25 -1.51 2.03 1.65 1.22 1.11
. VTG I -2.3 -3.9** -10.7** -1.62 -2.48 -12.5*
L'Ve' VTG II 1.28 -19.1*** -37.0*** -2.01 -27 —1 73*“
CHG H -2.13 -21.1*** —69.1*** -1.31 1.04 -14.1*
CHG HM -2.27 -4.75*** -21.5*** -2.38 -2.01 62.2“”
CHG L 314* 7.3* -3.81* -2.26 -1.29 -72.3***
CYP3A 1.66 -1.04 1.8 -1.58 1.92“ 303"“
Annexin max2 1.28 -1.03 1.23 -1.27 1.41 2.6”

 

" animal replicate in each treatment group (1124-6).

b * p < 0.05, ** p < 0.01. *** p < 0.001.

36

Table 2.6. Spearman rank correlation coefficients (numbers) and probabilities (*)

between hepatic expression levels of ER-a , ER-B, AR-a mRNA and other genes.

a.b,c

 

 

Female Male

ER-a ER-[I AR-a ER-a ER-li AR-a
ER-fl 0.08 -0014
AR-a 0.132 0.512 ** 0.12 0.156
CHG HM 0.798 *** -0.209 0.132 0.76 *** -0.011 0.213
CHG H 0.887 *** 0.098 0.152 0.747 *** 0.19 0.251
CHG L 0.834 *** -0.201 0.22 0.596 *** -0.043 0.212
VTG I 0.69 *** -0.207 0.094 0.817 *** 0.043 0.146
VTG I] 0.813 *** -0.207 0.141 0.848 *** -0.073 0.217

0.521

CYP3A 0.035 0.28] 0.05.5 -0.l46 *** 0.072
A nnexin max2 -0. 124 0.246 -0.007 -0. 17 0.315 -0. 144

 

“ Gene expressed level in each animal was calculated as fold change comparing to the average

expression level in control group.

b Analyses were conducted separately within female (11 2 31) and male (n = 40) groups

“ **p<0.01, *** p<0.001.

37

Figure 2.5

(p-value = 0.102).

CYP19A (fold)

Correlation of brain expression level of brain CYPIQB v.s. ovary CYP19A.
Spearman rank correlation coefficients for female -0.676 (p-value < 0.001) male 0.266

10

32

-32

40

Brain CYP198 v.s. Gonadai CYP19A

 

 

 

 

 

 

 

 

\-
I Female ‘. I I
Male I ‘ I
I
- IAIA
‘A
A .‘A .A A
u ‘
A ‘ IPIL a!
A .
A " "
A ' i
I‘,\'-
I I “
I r l I
-10 -32 1 32

CYP19B (fold)

38

DISCUSSION

Japanese medaka HPG axis RT -PCR array

The Japanese medaka HPG real-time PCR array developed for assessing chemical
induced effects on the endocrine pathways of brain, liver, and gonad was found to be
reliable for transcriptional profiling and the results can provide mechanistic knowledge of
chemical induced-effects in a systematic manner. Most of the 36 genes selected in the
Japanese medaka HPG real-time PCR array haven’t been previously examined in teleost
exposures to model chemical EE2 or TRB by the time of study. However, for the genes
previously studied in Japanese medaka and other fishes during similar exposures, their
reported changes in transcription are consistent with the results reported here in terms of
both magnitudes and directions of change (Lee et al. 2002, Martyniuk et al. 2007,
Miracle et al. 2006). For example, V T G I, one of yolk precursor genes, was up-regulated
by [64-fold and 3205-fold in male Japanese medaka from 7 d exposures to 50 and 500 ng
EE2 /L, respectively in this study. An about 700-fold up-regulation for the homologous
V TO 1' were reported in zebrafish exposed for 21 d to 10 ng EE2/L (Martyniuk et al.
2007). In fathead minnows, exposure to EE2 for 24 h also resulted in up-regulation of
VTG] (Miracle et al. 2006). In the present study, exposure to 50 ng EE2/L significantly
Lip-regulated expression of CHG HM and CHG L, but not CHG H, which is consistent
with the report that estrogenic chemicals dose-dependently up-regulated mRNA
expression of CHG subunits and that CHG L was more responsive than CHG H (Lee et al.
2002). Furthermore, up-regulation of ER-a in livers of fish exposed to EE2 similar to

that observed in our study has been previously reported for Japanese medaka (Yamaguchi

39

et al. 2005), zebrafish (Martyniuk et al. 2007) and fathead minnow (F ilby et al. 2007).
Finally, the statistically significant up-regulation (2-fold) of brain neuropeptide 1’ gene
expression in Japanese medaka exposed to 5 ng EE2/L observed in this study was
consistent with the 3-fold increase reported to occur in zebrafish exposed to 10 ng EE2/L
for 21 d (Martyniuk et al. 2007). These comparisons have not only verified the reliability
of the measurement by the Japanese medaka HPG axis PCR array, but also demonstrated
that the responses of the HPG axis pathway were similar among teleosts.

Quantification of changes in expression of genes in the HPG axis of Japanese
medaka by RT-PCR array provides unique information to develop hypothesises about the
transcriptional machinery. For example, expression of the V T G and CHG genes were
significantly correlated to expression of ER-a expression but not to either ER-fl or AR-a
in liver of both sexes exposed to EE2 or TRB (Table 2.6). This indicates that the VT G
and CHG genes were both primarily regulated by ER-a in the liver of Japanese medaka
exposed to the two model chemicals. Although TRB is considered to be an AR agonist
(Ankley et al. 2003), the hepatic ER mediated pathway is also responsive to TRB
exposure. Teleost fish such as Japanese medaka, have two distinct aromatase genes, the
gonadal (CYP19A) and brain (CYP19B) forms (Kuhl et al. 2005). The estradiol
synthesized by gonadal aromatase has critical impacts on reproductive and sexual
functioning, and brain aromatase activity can modulate neurogenic activity in the brain
(Callard et al. 1995). However, the relationship between the transcriptional regulation of
brain C YP19A and that of gonadal C YP19B has not been examined previously. The
negative correlation between the transcription of brain CYP19B and ovary CYP19A

indicates different physiological roles of brain and ovary aromatase activity.

40

17a-ethinylestradiol exposure

Exposure to EE2 had been reported to elevate VTG concentration and cause
feminization of male Japanese medaka and zebrafish (Om et al. 2006). Recent toxico-
genomic investigations have improved understanding of the molecular mechanisms of
EE2 effects in fish (Martyniuk et al. 2007; Moens et al. 2007; Santos et al. 2007).
However, these studies only focused on single tissue types, such as liver or gonad.
Transcriptional responses observed in the Japanese medaka HPG axis RT-PCR array
provide systematic information on EE2-induced mechanisms on gene expression. In this
study, the concentration dependent up-regulation of ER-a, V T G and CHG genes in livers
of males exposed to EE2 suggests that exposure to estrogen has resulted in increase of
endogenous estrogen concentration in Japanese medaka. The greater expression of V T G
and CHG genes in EE2 exposed males could produce extra yolk and envelop protein in
liver and explains the greater H81 in EE2 exposed male Japanese medaka (Table 2.3).
The HPG axis pathway in Japanese medaka displayed compensatory feedback
mechanism to EE2 exposure. In male brain, the expression GnRH R I and GTHa were
down-regulated. If less expression of these genes can be translated into lower protein
expression, it could lead to insufficient GnRH signaling and a fall in gonadotrophin
secretion, which could consequently down-regulate gonadal steroidgenesis. The
observed gonadal gene expression profile further confirms this hypothesis. EE2 exposure
down-regulated the testicular expression of CYP17, which is one of the key enzymes

involved in estrogen synthesis. In ovary, expression of CYP19A also displayed

41

significant down-regulation (5-fold), which is in favor of decreased estrogen synthesis to
compensate for redundant endogenous estrogen.

Expression responses of other HPG axis genes have also characterized the
mechanisms induced by EE2 exposure. For example, expression ofAR-a was down-
regulated in brains of males exposed to EE2. The agonist of AR, androgen is well known
to act on the brain to modify male sexual behavior and other brain functions. Therefore in
brains of EE2 exposed Japanese medaka, the lesser expression of A R-a may be associated
with some alterations in sexual behavior and olfactory preference for receptive females.
This hypothesis has been confirmed by the previous observation that exposure to another
form of estrogen, l7B-estradiol, suppressed sexual behavior (following, dancing, floating,
and crossing) in male Japanese medaka (Oryzias latipes) (Oshima et al. 2003). In another
example, the down-regulation of the expression of C YP3A in liver of males exposed to
500 EE2 ng/L is consistent with reduced production of P450 isoforms in E2 exposed
male fish (Kashiwada et al. 2007). In fish, CYP3A plays a major role in the metabolism
of endogenous compounds, including steroids (Miranda et al., 1989). The less expression
of C YP3A is related to the greater ER transcript that may down-regulate C YP3A in favor
of maintaining high endogenous hormone concentrations during reproduction

(Kashiwada et al. 2007).

/ 7/)’-trenbolone exposure
TRB is an active metabolite of trenbolone acetate which is used as a growth
promoter for farm animals. It has been reported to cause adverse effects on immune

responses and reproduction (Hotchkiss and Nelson 2007; Miller et al. 2007). In fish,

42

TRB has been reported to cause masculinization of female zebrafish and decrease V TG
production in both. zebrafish and Japanese medaka (Masanori et al. 2006; Orn et al. 2006).
TRB is an androgen receptor agonist and it has a greater affinity for the fish androgen
receptor than that of the endogenous ligand, testosterone (Ankley et al. 2003).
Nevertheless, TRB exposure has also been related to alterations of estrogen related
pathways in recent literature. Down-regulation of VT G mRNA (VII and Vt3) were
observed in female fathead minnows in a 21 d exposure to 50 and 500 ng TRB/L
(Miracle et al. 2006). In Japanese medaka, TRB exposure for 7 d down—regulated
expression of VTG-1, V T G-II, CHG H, CHG L and CHG HM in liver of both males and
females. The reduced expression of V TG and CHG genes in TRB exposed females could
produce less yolk and envelop protein and explains the reduced fecundity in TRB
exposed Japanese medaka (Figure 2.1). These results suggest a lower concentration of
endogenous estrogen (NB-estradiol) in liver of TRB exposed Japanese medaka because
hepatic VT G and CHG genes were primarily regulated by estrogen. The down-regulation
of ER—a in liver of TRB exposed Japanese medaka further supports this hypothesis.
Furthermore, TRB exposure reduced concentrations of plasma 17B-estradiol and
testosterone In female fathead minnows (Ankley et al. 2003).

The mechanism by which the androgen TRB inhibits the production of 170-
estradiol is still unclear. It has been hypothesized that decreased testosterone production
is likely to represent a compensatory response to exposure to exogenous androgen, TRB.
Consequently, less endogenous testosterone results in less l7B—estradiol because I7B-
estradiol is produced by conversion of testosterone by aromatase (Miracle et al. 2006).

The expression response of the Japanese medaka HPG axis supports this hypothesis. In

43

male brain, the down-regulation of the expression of GTHa and LH-/)’ could result in
decreased luteinizing hormone concentrations. The brain gonadotrophin LH plays a
major role in the regulation of gonadal steroidogenesis. In accordance with the decreased
number of LH transcripts in brain, the expression oftesticular steroidogenic genes
involved in the synthesis oftestosterone and estrogen, including HMGR, StAR and
CYPIIB, showed down-regulation in TRB exposed Japanese medaka.

TRB exposure induced gender-specific responses along the HPG axis in Japanese
medaka. Previous reports on the adverse effects by TRB have mainly focused on
masculinization offemales (Ankley et al. 2003; Miracle et al. 2006). Nevertheless,
males of Japanese medaka were more sensitive to TRB exposure than were females to the
reduction of expression of estrogen receptor, VT 0 and CHG genes in liver. Although the
physiological function of the expression of the V T G and CHG genes in males is largely
unknown, these genes have been shown to rapidly respond in male fish exposed to
estrogens and anti-estrogens in both field and laboratory studies (Hutchinson et al. 2006).
The higher susceptibility of estrogen-responsive pathways to fluctuations in the
concentration of endogenous estrogen in males than in females may be due to the lower
basal estrogen concentration and estrogen metabolizing capability in males. Conversely
females showed a greater compensatory response to TRB exposure to reduce endogenous
estrogen than males, which includes the up-regulation of brain GnRH R I] and ovarian
CYP19A. These gender-specific transcriptional response profiles in the HPG axis of TRB

exposed Japanese medaka provide a unique signature of TRB exposure.

44

CONCLUSIONS

The HPG -PCR array system developed in this study represents a sensitive,
reliable and flexible monitoring tool to research chemical—induced effects along the HPG
axis in Japanese medaka. The Japanese medaka HPG PCR array developed in the study
combines the quantitative performance of SYBR Green-based real-time PCR with the
multiple gene profiling capabilities of a microarray to examine chemical-induced gene
expression profiles along the HPG axis. The pathway-based approach implemented in
this study provides a valuable tool to analyze and visualize concentration-dependent
responses induced by chemical in the HPG axis of Japanese medaka.

Overall, this study demonstrated that profiling of HPG transcripts by RT-PCR
array methods can discriminate estrogenic and androgenic EDCs, and represents a useful
tool to systematically evaluate chemical induced molecular responses in multiple
pathways and multiple organs. Mechanistic information derived from the results can be

used in diagnostic and predictive assessments ofthe risk of EDCS.

Acknowledgements

This study was supported by a grant from the US EPA Strategic to achieve Results
(STAR) to J.P. Giesy, M. Hecker and PD. Jones (Project no. R-831846). The research
was also supported by a grant from the University Grants Committee of the Hong Kong
Special Administrative Region, China (Project No. AoE/P-04/04) to D. Au and J.P. Giesy

and a grant from the City University of Hong Kong (Project no. 70021 17).

45

REFERENCE

Ankley, G. T., Jensen, K. M., Makynen, E. A., Kahl, M. D., Korte, J. J., Hornung, M. W.,
Henry, T. R., Denny, J. S.. Leino, R. L., Wilson, V. S., Car‘don, M. C., Hartig, P.
C., Gray, L. E., 2003. Effects ofthe androgenic growth promoter l7-beta-
trenbolone on fecundity and reproductive endocrinology ofthe fathead minnow.
Environ. Toxicol. Chem. 22, 1350-1360.

Ankley, G. T., Johnson, R. D., 2005. Small Fish Models for Identifying and Assessing
the Effects of Endocrine-disrupting Chemicals. ILAR J 45, 469-483.

Ankley, G. T., Villeneuve, D. L., 2006. The fathead minnow in aquatic toxicology: past,
present and future. Aquat. Toxicol. 78, 91-102.

Ankley, G. T., Kalil, M. D., Jensen, K. M., Hornung, M. W., Korte, J. J., Makynen, E. A.,
Leino, R. L., 2002. Evaluation ofthe aromatase inhibitor fadrozole in a short-term
reproduction assay with the fathead minnow (Pimephales prornelas). Toxicol. Sci.
67, 121-130.

Callard, G. V., Kruger, A., Betka, M., 1995. The goldfish as a model for studying
neuroestrogen synthesis, localization, and action in the brain and visual system.
Environ. Health Perspect. 103, 51-57.

Filby, A. L., Thorpe, K. L., Maaek, G., Tyler, C. R., 2007. Gene expression profiles
revealing the mechanisms of anti-androgen- and estrogen-induced feminization in
fish. Aquat. Toxicol. 81, 219-231.

Gracia, T., Hilscherova, K., Jones, P. D., Newsted, J. L., Zhang, X., Hecker, M., Higley,

E. B., Sanderson. J. T., Yu, R. M. K., Wu, R. S. S., Giesy, J. P., 2006. The

46

H295R system for evaluation of endocrine-disrupting effects. Ecotoxicol. Environ.
Saf.65,293-305.

Hecker, M., Newsted, J. L., Murphy, M. B., Higley, E. B., Jones, P. D., Wu, R., Giesy, J.
P., 2006. Human adrenocarcinoma (H295R) cells for rapid in vitro determination
of effects on steroidogenesis: hormone production. Toxicol. Appl. Pharmacol 217,
114-124.

Hilscherova, K., Jones, P. D., Gracia, T., Newsted, J. L., Zhang, X. W., Sanderson, J. T.,
Yu, R. M. K., Wu, R. S. S., Giesy, J. P., 2004. Assessment ofthe effects of
chemicals on the expression of ten steroidogenic genes in the H295R cell line
using real-time PCR. Toxicol. Sci. 81, 78-89.

Hutchinson, T. H., Ankley, G. T., Segner, H., Tyler, C. R., 2006. Screening and testing
for endocrine disruption in fish-biomarkers as "signposts," not "traffic lights," in
risk assessment. Environ. Health Perspect. l 14, 106—114.

Kashiwada, S., Kameshiro, M., Tatsuta, H., Sugaya, Y., Kullman, S. W., Hinton, D. E.,
Goka, K., 2007. Estrogenic modulation of CYP3A38, CYP3A40, and CYP19 in
mature male medaka (Oryzias latipes). Compar. Biochem. Phys. Part C: Toxicol.
Phannacol.l45,370-378.

Kuhl, A. J., Manning, 8., Brouwer, M., 2005. Brain aromatase in Japanese medaka
(Oryzias latipes): Molecular characterization and role in xenoestrogen-induced
sex reversal. J Steroid Biochem. Mol Biol. 96, 67-77.

Lee, C., Na, J. G., Lee, K. C., Park, K., 2002. Choriogenin mRNA induction in male
medaka, Oryzias latipes as a biomarker of endocrine disruption. Aquat. Toxicol.

61,233-241.

47

Martyniuk, C. J., Gerrie, E. R., Popesku, J. T., Ekker, M., Trudeau, V. L., 2007.
Microarray analysis in the zebrafish (Danio rerio) liver and telencephalon after
exposure to low concentration of l7alpha-ethinylestradiol. Aquat. Toxicol. 84,
38-49.

Miracle, A., Ankley, G., Lattier, D., 2006. Expression oftwo vitellogenin genes (vgl and
vg3) in fathead minnow (Pimephales promelas) liver in response to exposure to
steroidal estrogens and androgens. Ecotoxicol. Environ. Saf. 63, 337-342.

Miranda, C.L., Wang, J.L., Henderson, M.C., Buhler, DR, 1989. Purification and
characterization of hepatic steroid hydroxylases from untreated rainbow trout.

Arch. Biochem. Biophys. 268, 227—238

Om, S., Yamani, S., Norrgren, L.,2006. Comparison of vitellogenin induction, sex ratio,
and gonad morphology between zebrafish and Japanese medaka after exposure to
17alpha-ethinylestradiol and l7beta-trenbolone. Arch. Environ. Contam. Toxicol.
51, 237-243.

Pastva, S. D., Villalobos, S. A., Kannan, K., Giesy, J. P., 2001. Morphological effects of
Bisphenol-A on the early life stages of medaka (Oryzias latipes). Chemosphere 45,
535-541.

Salornonis, N., Hanspers, K., Zambon, A.C., Vranizan, K., Lawlor, S.C., Dahlquist, K.D.,
Doniger, S.W., Stuart, J., Conklin, B.R., Pico. A.R., 2007. GenMAPP 2: new
features and resources for pathway analysis. BMC Bioinformatics, 8: 217.

Villalobos, S. A., Papoulias, D. M., Pastva, S. D., Blankenship, A. L., Meadows, J., Tillitt,

D. E., Giesy. J. P. (2003). Toxicity of o,p'-DDE to medaka d-rR strain after a one-

48

time embryonic exposure by in ovo nanoinjection: an early through juvenile life
cycle assessment. Chemosphere 53, 819-826.

Villeneuve, D. L., Ankley, G. T., Makynen, E. A., Blake, L. S., Greene, K. J., Higley, E.
B., Newsted, J. L., Giesy, J. P., Hecker, M., 2007. Comparison of fathead minnow
ovary explant and H295R cell-based steroidogenesis assays for identifying
endocrine-active chemicals. Ecotoxicol. Environ. Saf. 68, 20-32.

Villeneuve, D. L., Larkin, P., Knoebl, 1., Miracle, A. L., Kahl, M. D., Jensen, K. M.,
Makynen, E. A., Durhan, E. J., Carter, B. J., Denslow, N. D., Ankley, G. T., 2007.
A graphical systems model to facilitate hypothesis-driven ecotoxicogenomics
research on the teleost brain-pituitary-gonadal axis. Environ. Sci. Technol. 41,
321-330.

Wittbrodt, J., Shima, A., Schartl, M., 2002. Medaka--a model organism from the far East.
Nat. Rev. Genet. 3, 53-64.

Yamaguchi, A., Ishibashi, H., Kohra, S., Arizono, K., Torninaga, N., 2005. Short-term
effects of endocrine-disrupting chemicals on the expression of estrogen-
responsive genes in male medaka (Oryzias latipes). Aquat. Toxicol. 72, 239-249.

Zhang, X., Yu, R. M., Jones, P. D., Lam, G. K., Newsted, J. L., Gracia, T., Hecker, M.,
Hilscherova, K., Sanderson, T., Wu, R. S., Giesy, J. P., 2005. Quantitative RT-
PC R methods for evaluating toxicant-induced effects on steroidogenesis using the
H295R cell line. Environ. Sci. Technol. 39, 2777-2785.

Zhang, Z., Hu, J., 2007. Development and validation of endogenous reference genes for
expression profiling of medaka (Oryzias latipes) exposed to endocrine disrupting

chemicals by quantitative real-time RT-PCR. Toxicol. Sci. 95, 356-368.

49

Chapter 3

Time-dependent transcriptional profiles of genes of the hypothalarnic-pituitary-gonadal

(HPG) axis in medaka (O. latipes) exposed to fadrozole and l7B-trenbolone

Xiaowei Zhangu’“, Markus Ileckeri‘§‘“, June-Woo Park“, Amber R. Tompsettl‘i, Paul D.
Jonesi’”, John Newstedi‘", Doris Auii, Richard Kong”, Rudolf S.S. Wu“: and John P.

++
1-++

- , firm“
GresyT +

1' Department of Zoology, Michigan State University, East Lansing, MI. USA
3: National Food Safety and Toxicology Center and Center for Integrative Toxicology,
Michigan State University, East Lansing, MI. USA
§ ENTRIX, Inc., Saskatoon, SK, Canada
|| ENTRIX, Inc., Okemos, MI, USA
# Toxicology Centre, University of Saskatchewan, Saskatoon, SK, Canada
‘1‘1‘ Dept. Biomedical Veterinary Sciences, University of Saskatchewan, Saskatoon, SK,
Canada

:1 Dept. Biology & Chemistry, City University of Hong Kong, Hong Kong, SAR, China

50

ABSTRACT

Both the anabolic androgen l7B-trenbolone (TRB) and the aromatase inhibitor fadrozole
(FAD) can cause decreased plasma concentrations of estrogen (E2) and reduce fecundity
of fish. However, the underlying mechanisms and the molecular pathways involved are
largely unknown. This study was designed to assess time-dependent effects of FAD and
TRB on the transcriptional responses of the hypothalamic-pituitary-gonadal (HPG) axis
of Japanese medaka (Oryzias latipes). Fourteen wk-old Japanese medaka were exposed
to 50ug FAD/L or 2ug TRB/L in a 7-d static renewal and expression profile of 36 HPG
axis genes were measured by means of a medaka HPG real time PCR (RT-PCR) array
after 8 h, 32 h or 7 d of exposure. Exposure to TRB or FAD caused lesser fecundity of
Japanese medaka and down-regulated transcription of vitellogenin (V TG) and
choriogenin (CHG) genes in the liver of females. Exposure to FAD for 8 h resulted in an
8-fold and 71-fold down-regulation of expression of estrogen receptor alpha (ER-a) and
CHG L, respectively in female liver. l7B-trenbolone caused similar down-regulation of
these genes but the effects were not observed until 32 h of exposure. These results
support the hypothesis that FAD reduces plasma E2 more quickly by inhibiting aromatase
enzyme activity than does TRB, which inhibits production -of the E2 precursor
testosterone. Exposure to FAD and TRB resulted in rapid (after 8 h) down-regulation of
LHR and LDLR in the testis to compensate excessive androgen level. Overall, the
molecular responses observed in the present study differentiate the mechanisms of the

reduced fecundity by TRB and FAD.

51

Keyword- fecundity, activin, vitellogenin, HPG axis, Real time PCR array

52

INTRODUCTION

Over the last decade, molecular biomarkers have been successfully applied in the
screening and testing of endocrine disrupting chemicals in fish [1]. While molecular
biomarkers have potential to aid in the elucidation of causative modes of action (MOA)
and may in some cases allow for prediction of possible adverse effects of chemicals at
higher organizational levels, to date this potential has not been realized [2, 3]. It is still
difficult to relate changes in expression of single genes to population-level fitness. One
example ofa successful biomarker has been the yolk protein precursor vitellogenin (V TG)
which has been shown to rapidly respond in fish exposed to estrogens and anti-estrogens
in both filed and laboratory studies [1]. Recent studies have demonstrated that changes in
VTG caused by several model chemicals can be quantitatively translated into adverse
health effects at apical and population-level in fathead minnow (Pimephales promelas)
[4,5]. Measurement of the VTG in plasma using immunology-based detection methods
has been applied in various fish species [6]. Measurements of VTG mRNA transcripts
using real time PC R methods have been demonstrated to be advantageous over protein
determination due to the rapid induction of the gene (early warning) and less
susceptibility to cleavage [1]. Additional molecular biomarkers that have been shown to
serve as useful indicators for the interaction of EDCs with the teleost hypothalamic-
pituitary-gonadal (l-IPG) axis include nuclear hormone receptors [7], steroidogenic
enzymes (e. g. aromatase) [8], and gonadotropins (follicle-stimulating hormone and
luteinizing hormone) [9]. These studies have demonstrated that molecular responses,

specifically gene profiling, of one or two functional-relevant biomarkers, can facilitate a

53

better understanding of the mechanisms of estrogenic disruption in fish. However,
chemicals can cause similar responses of biomarkers at certain target organs by acting
through different mechanisms. For example, both the specific inhibitor of aromatase
fadrozole (FAD) and the androgen receptor agonist 17 B-trenbolone (TRB) inhibit hepatic
estrogen receptor (ER) and V T G transcripts, and consequently impair fecundity in
fathead minnow [10, l l] and Japanese medaka (Oryzias latipes) [12, 13]. To better
understand these types of interactions, more comprehensive and integrated molecular
approaches are needed to evaluate potential endocrine disrupting chemicals (EDCS) with
unknown mechanisms.

Recently, Villeneuve et al. summarized the current understanding of the
reproduction-related molecular pathways within the teleost HPG-axis [14]. Using the
Japanese medaka as a small fish model, we have developed and validated a reliable,
sensitive and flexible real time PCR (RT-PCR) array to systematically study chemical-
induced effects at multiple endocrine pathways in brain, gonad and liver [13]. 36 genes
has been selected from the four modules representing (1) gonadotropin synthesis and
release in the hypothalamus and pituitary, (2) cholesterol transport, (3) steroidogenesis in
the gonads, and (4) production of egg precursors in the liver [13]. Through systematic
monitoring of key genes along the HPG-axis in fish exposed to EDC of concern, it is
anticipated that a better understanding of chemical induced mechanisms of actions (MOA)
can be achieved that ultimately will aid in improving chemical risk assessment.

In the present study, we used FAD and TRB as model compounds to investigate
different MOAs resulting in the same net effect. Because of the direct inhibitory effect of

estrogen production by FAD, we hypothesized that FAD exposure would elicit responses

54

more rapidly on gonadal steroidogenesis —related genes and hepatic estrogen-responsive
genes than would TRB. Time—dependent transcriptional responses of key genes along
the HPG axis were examined after exposure to FAD and TRB using the medaka HPG

RT-PCR array [13].

55

MATERIALS AND METHODS

Compounds and reagents
Fadrozole was provided by Novartis, Inc. (Summit, NJ, USA). l7B-trenbolone

(TRB) and dimethyl sulfoxide (DMSO) was obtained from Sigma (St. Louis, MO).

Animals

Male and female wild-type Japanese medaka (Oryzias latipes) used in this study
originated from the aquatic culture unit at the US Environmental Protection Agency Mid-
Continent Ecology Division (Duluth, MN, USA). The fish were maintained in flow-
through tanks in conditions that facilitated breeding (23-24 °C; 16:8 light/dark cycle) in
accordance with protocols approved by the Michigan State University Institutional

Animal Care and Use Committee (MSU-IACUC).

Chemical exposure

Chemical exposure was carried out by a static renewable system as previously
described [13]. Briefly, studies were conducted in lO-L tanks filled with 6 L of carbon-
filtered water. Each day during the exposure half of the water in each tank (3 L) was
replaced with fresh carbon-filtered water containing the appropriate amount of chemical
or solvent. Each tank contained 5 male and 5 female 14 wk -old medaka as determined
by secondary sexual characteristics of the fins. Each treatment had 2 replicate tanks that
were sampled at each time point. After the acclimation period, medaka were exposed to

one of three treatments: vehicle control (DMSO with a final concentration of 1:10000 v/v

56

water), 50 ug FAD/ L or 2 ug TRB/L. Concentrations of the chemicals were selected
based on the results of previous studies [12, 13] so that the selected concentrations would
inhibit fecundity of medaka without causing mortality. Exposures started at midnight
(12:30 AM) of the first day. Eggs produced during the previous 24 h period were
counted and recorded before the replacement of water. No mortality of adult medaka was
observed in any treatment. The three sampling times were 8:30 AM ofthe fist day (8 h),
day 2 (32 h) and day 7 (152 h) of exposure. When sampling fish were euthanized in
Tricaine S solution (Western Chemical, Ferndale, WA, USA), and total weight and snout-
vent length were recorded for each fish. For gene expression analysis, 4-6 males and
females were randomly sampled from the two replicate tanks of each treatment. Tissues
from brain, liver, and gonads were collected and preserved in RNAlater storage solution

(Sigma, St Louis, MO) at -20°C until analysis.

Total RNA isolation and reverse transcription PCR

Total RNA was extracted from individually tissues and first strand cDNA was
separately made for further quantification of transcripts [13]. Total RNA was extracted
by use ofthe Agilent Total RNA Isolation Mini Kit (Agilent Technologies, Palo Alto, CA)
according to the manufacture’s protocol. Purified RNA was stored at -80 °C until
analysis. First-strand cDNA synthesis was performed using Superscript III first-strand
synthesis SuperMix and Oligo-dT primers (Invitrogen, Carlsbad, CA). Briefly, a 0.5 to 2
ug aliquot oftotal RNA was combined with 1 [1L of 50 uM of Oligo(dT)30. 1 [LL of
annealing buffer, and RNase-free water to a final volume of 8 [IL Mixes were denatured

at 65 0C for 5 min and then quickly cooled on ice for 2 min. Reverse transcription was

57

performed after adding 10 uL 2X first-stand reaction mix, and 2 uL SuperScript
lll/RNaseOUT enzyme mix. Reactions were incubated at 50 °C for 50 min and, on
completion, were inactivated at 85 °C for 5 min. RNA was digested by adding 1.25 uL
RNase H (Invitrogen, Carlsbad, CA) then incubated at 37 °C for 30 min. cDNA was

stored at —20 °C until further analysis.

Real time —PC R array measurement

Gene expression in brain, liver and gonad was quantified by use of the medaka
HPG axis PCR array described previously (Table 2.1) [13]. Briefly, real time,
quantitative polymerase chain reaction (Q-RT-PCR) was performed by using a 384-well
A81 7900 high throughput real time PCR System (Applied Biosystems, Foster City, CA).
PCR reaction mixtures sufficient for one hundred reactions contained 500 uL of SYBR
Green master mix (Applied Biosystems, Foster City, CA), 2 uL of 10 uM sense/anti-
sense gene-specific primers, and 380 uL of nuclease-free distilled water (Invitrogen). A
final reaction volume of 10 uL was made up with 2 uL of diluted cDNA and 8 [LL of PCR
reaction mixtures using a Biomek automation system (Beckman Coulter, Inc., Fullerton,
CA). Expression of target genes was quantified by use ofthe comparative cycle
threshold (Ct) method according to methods reported elsewhere [15]. The average Ct
value of the three reference genes (/i-actir‘2, RPL-7 and 16s) was used as reference for the

expression calculation oftarget genes

Statistical analysis

58

Statistical analyses were conducted using the R project language (http://www.r-
project.org/). Analysis of fecundity data was using analysis of variance (ANOVA) model,
in which the effects of time (day), chemical (TRB or FAD) and their interaction on the
daily recorded egg production were examined. Prior to conducting statistical
comparisons of gene expression, the assumption of normality of distributions of data was
evaluated by the Shapiro-Wilk’s test. If necessary, data was log-transformed to
approximate the normal probability function. Differences in magnitudes of expression
among genes were evaluated by use of ANOVA followed by pair—wise t-test. Differences

with p < 0.05 were considered to be statistically significant.

59

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

w s m m e m N F m s o m e m N F
r _ _ _ _ _ _ _ _ _ _ _ _ _ F
T O I 0
v v
.. -m m ..H H H -m m
m ..H HQ m
B e
-a w. - a. a
m m
r m w r m w
w w
0 a, 0 a.
Q<m Iml w mm... lei w
0.2.5 1 9:0 I
_ a O L¢f I M m _ a 0 ldi 1 n0... %
Eamonxo 2396mm. 9.3898 0.3.2.9.... 83-:

95% 35:8 22:? Eofi God v 02.9-3 oocobmtw Emoficwa a 8865 mxthma
2F .nmc mo when m win—5:8 some .353 28:8: N Eo¢ 628:8 295% Log wwwo Co 598:: 033383 Co cosmSoe cum—=85
was :38 05 E8052 San— .«moa “Yb a E AE<L w: om Lo 45”: w: ON 8 @3098 8162: E @6553 3:22:30 fin FSH,.—

6O

Table 3.1 Transcriptional response profiles of HPG axis pathways in medaka fish
exposed to 50 ug FAD /L. Gene expression was expressed as the fold change
comparing to the corresponding solvent controls 3.5

 

 

 

 

, Female Male

Tissue gene 8 h 32 h 7 d 8 h 32 h 7 d
ER-a 1.22 -1.12 1.15 1.13 -1.03 -1.14
ER-B 1.25 1.2 1.48 1 .13 2.27* -1.52
NeuropepY 1.15 -1.6 1.16 1.46 -1.06 -1.21
mfGnRH 1.61 1.13 1.15 1.43 -1.07 -1.08

Brain anRH 1.17 -1.08 -1.01 1.16 -1.05 -1.13
cGnRH II 1.65 -1.34 1.76 1.05 -1.34 1.25
GnRH RI 1.62 -1.04 -1.34 -1.16 1.18 -1.29
GnRH R/I 1.05 1.01 -1.04 1.13 1.11 -1.18
GnRH RIII 1.06 -1.19 -1.02 1.03 1.21 -1.08
GTHa -1.44 1.71 3.84*** -1.57 -1.3 -1.2
LH-B -1.44 -1.62 1.47 -2.33 -2.67 -3.26
CYP1QB -1.20 -3.11*** -2.90*** -1.50* 2.92*** -3.11***
ER-a 1.32 1.84 -1.18 1.05 256*“ -1.01
ER-B 1.52 -1.67 -1.76 -1.06 1.78* 1.17
AR-a 1.19 -1.07 -1.15 -1.15 1.49* 1.06
FSHR 1.19 1.77 1.97 1.02 223* -1.10
LHR -1.07 1.10 -1.36 -6.71* -3.33 -1.19
HDLR -1.21 -2.29*** -2.56* -1.06 182* -1.03

Gonad LDLR 1.12 1.11 -1.04 -2.09* 1.01 -1.44
HMGR 1.14 -3.55 -3.23 -5.15 1.90* -1.02
StAR 1.04 1.52 1.48 -1.04 2.17* 280*
CYP11A -1.15 1.19 -1.08 1.45 1.70 1.34
CYP1tB 1.50 -5.62* -3.28 -1.16 229* 1.92*
CYP17 1.22 1.19 1.45 2.03 3.13* 1.56
CYP19A 2.02 419* 3.99 1.80 -1.56 1.11
CYP21 1.06 -2.64* -2.84 -1.30 -2.37* 1.06
ZOB-HSD 1.09 -1.11 -1.11 1.11 1.05 -1.12
3B-HSD 1.21 2.26 1.25 1.27 2.49" 1.44
CYP3A -1.47 -13.28 -9.17** -1.09 -1.75 1.30
Inhibin A 1.78 1.15 1.16 1.11 1.86* 1.34
Activin BA 1.50 -5.54* -3.81 1.17 -2.59* 1.53
Activin BB 1.20 -1.61* -2.81* -1.21 1.33 1.01

Liver ER-a -8.46** -4.23* -8.47** 1.09 2.42*** -5.16***

ER-B 1.30 -1.74 1.11 1.12 1.09 1.20
AR-a -2.04 -1.27 -1.36 -1.08 -1.07 -1.16
VTG I 1.96 -2.54 -63.2* -5.23 5.68 -125**
VTG II -1.81 -20.3*** -355*** -2.47 14.8 -41.1***
CHG H 1.22 -1.31 -19.6** -2.56 5.57 -178***
CHG HM -2.19 -17.0*** -61.2*** -2.37 -2.37* -52.0***
CHG L -71.4*** -62.1*** -32.1*** -1.17 -1.34 -34.8***
Annexin max2 1.41 3.27" 1.68 -1.17 1.43 1.39

 

“ animal replicate (n= 4-6).

b It: p<005’ lint P< 001, air-ini- P<0-OOI.

61

Table 3.2. Transcriptional response profiles of HPG axis pathways in medaka fish
exposed to 2.0 ug TRB /L. Gene expression was expressed as the fold change comparing

 

 

 

to the corresponding solvent controls 3“”
Female Male

Tissue gene 8 h 32 h 7 d 8 h 32 h 7 d
ER-a -1.14 1.08 -1.12 1.01 -1.05 -1.13
ER-B -1.93* -1.11 1.63 1.14 1.71 1.07
NeuropepY -1.19 -1.18 1.03 1.84 -1.21 1.13
mfGnRH -1.8 1.08 -1.16 2.15" -1.07 1.45
anRH -1.27 -1.46 1.35 1.39 1.08 1.09
cGnRH II 1.1 1.37 1.56 -1.12 1.1 1.04

Brain GnRH RI -1.48 1.34 -2.12** 1.16 1.03 1.44

GnRH R]! -1.16 1.06 -1.07 1.04 1.13 -1.36**
GnRH R!” -1.22 1.04 -1.32* -1.00 1.42"“r -1.06
GTHa -1.28 -1.52 -2.85 -1.13 -1.66 -1.33
LH-B 1.01 -2.15 -3.68 1.03 -1.59 -2.36
CYP1QB 1.04 -1.75* -1.95** 1.17 -1.41 -1.16
ER-a 2.51* 1.01 1.20 -1.15 2.72“” -1.22
ER-B 1.53 -1.73 -1.75 -1.35 1.75* -1.11
AR-a 2.16 -2.31*** ~1.58 -1.31 1.56" -1.27
FSHR 369*“ 1.73 1.86 -1.16 222* -1.4
LHR 7.61 1.1 1.07 -13.6** -2.07 1.02
HDLR 1.42 -2.57*** -3.45* -1.14 2.37" -1.48
LDLR 2.21" -1.09 2.00 -2.53** 1.02 -1.17
HMGR -2.8 -3.59 -4.32 -14.9* 1.80* -1.34

Gonad StAR -1.16 -2.99* -1.09 -1.34 -1.51 -1.34
CYP11A -2.2 -3.17"’* -1.81 1.10 -1.09 -1.61*
CYP11B -3.04 -3.76 -3.67 -1.46 1.02 -1.47
CYP17 -1.58 -4.05*** -2.48 -1.02 1.62 -2.12*
CYP19A 297* 1.35 3.05 1.29 -1.39 -1.20
CYP21 -2.03 -2.51 -2.16 -2.12 -1.31 -1.22
20 B-HSD 1.13 -1.11 1.08 1.28 1.25 -1.31
3B-HSD 3.18" 1.24 -1.01 -1.46 1.89 -1.47
CYP3A -11.9 -10.7 -8.14* -1.12 5.09 -1.28
Inhibin A 3* —1.27 -1.14 -1.16 206* -1.68
Activin BA -3.17 -4.86 -4.49 -1.09 -2.04 1.10
Activin 88 -1.11 -2.17*** -3.56* -1.68 1.46" -1.08
ER-a 1.57 —8.08* -5.8* 274* 2.72"“Hr -2.19*
ER-B 2.09 -3.09 -1.71 -1.07 -1.00 2.09
AR-a 1.8 -3.25 -2.35 -1.06 -1.73 1.04

Liver VTG I 3.41" -4.45 -52.7* 6.91 1.92 -115***

VTG II 1.03 -29.8*** -145*** 13.9 1.19 -16.9***
CHG H 1.98 -5.97*** -73.7*** 1.66 1.43 -86.7**"’
CHG HM 1.23 -24.1*** -84.4*** 2.84 7.51“” -29.8***
CHG L -1.28 -119*** -93.0*** 1.76 -2.17 -25.0**'
Annexin max2 -1.04 1.73 1.47 1.09 1.16 1.66

 

" animal replicate (n= 4-6).

h * p<0.05, ** p<0.01, *** p<0.001.

62

Figure 3.2 Striped view of Time dependent response profile in female Japanese
medaka exposed to 50 ug FAD/L. The legend listed in the upper right comer of the
graph describes the order of the three sampling time points and the eight colors
designating different fold thresholds. LH, lutinizing hormone; F SH, follicle-stimulating
hormone; E2, l7B-estradiol; T, testosterone; HDL, high-density lipid; LDL, low-density

lipid.

  

Medaka HPG axis Map W.
Color

 

 

 

 

Hypothalamus Gonadotropin releasing hormones

—

 

 

 

 

 

 

 

 

 

 

 

    

|o£ruinr [lorinuliiflon'imhm
——
Pituitary \ i\ I
gonadotrophs FSH ”'1
E31113! Gonad 1 m _
.... :0
mm inhibin

Progesterone —-—-> 17-Hydroxyprogesterone
HDL

LDL [_

 

 

 

 

 

i Pregnenolone 1 1-deoxyeortlsol Aidostenedione
HMG-CoA ,~———— ________.__ i
-! ] Testosterone —- T
1 ma
Cholesterol—1—> Cholesterol 173. ZOb-PWISI Essa-die: 22
Outer mitochondrial: Inner mitochondrial 20betas 11-ketotestosterone _- KT
I membrane . membrane
.—> ___ Estrogen «m-
Fawn?“ 1E- -
Em obmn,\\ Liver
—> mm m
ran:
an

 

 

 

 

 

 

 

 

63

Figure 3.3 Striped view of Time dependent response profile in female Japanese
medaka exposed to 2.0 ug TRB/L. The legend listed in the upper right comer of the
graph describes the order of the three sampling time points and the eight colors
designating different fold thresholds. LH, lutinizing hormone; FSH, follicle-stimulating
hormone; E2, 17B-estradiol; T, testosterone; HDL, high-density lipid; LDL, low-density
lipid.

 

   
   
    
    
  
 

 

 

 

 

 

 

Medaka HPG axis Map Exemsien 2.3m
Colors“;
"1
Brain
._. Hypothalamus Gonadotropin releasing hormones
_ I cGnRHII I] mfGnRH ”anRH]
GnRH . m m 'nr

 

 

 

 

 

 

 

 

 

 

 

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

FSH L”
Gonad \
Inhibin
Progesterone 17-Hydroxyprogesterone
—> La 1-
i PMMMIOM 11-deoxycortisol Aldoetenedione
HMG-CoA [My—“TMMM”
-' l M.,... —’ T
[ Em
Cholesterol ——]—> Cholesterol 17a-M'Plu'sl ] Estradlol - 52,
Oldermfiochondflfll I “Wm 20th 11-ketomtoeterone —> KT
membrane I membrane
__ BMW”,
responsive Liver
element ---------- . ,
—.
<] J

 

64

RESULTS

Chemical induced effects on medaka fecundity

Both exposures to TRB and to FAD reduced the egg production of Japanese
medaka in a time-dependent manner. There were no differences among daily production
of eggs in the control group during the exposure. However, fewer eggs were produced
after exposure to 2 ug TRB /L or 50 ug FAD /L for 3 or more days (Figure 3.1).
Fecundity was time-dependent; and there was a statistically significant interaction
between the main classification variables of “chemical” and “time” (p < 0.001) for both

FAD and TRB.

Time course ofHPG gene expression profiles: FAD

Exposure to 50 ug FAD /L caused time-dependent changes in expression of some
of the genes studied. Exposure to FAD down-regulated expression of yolk precursor and
egg envelop precursor genes, including VT G I, VTG II, CHG H, CHG HM and CHG L in
liver of both sexes (Table 3.1, Figure 3.2). Expression of ER-a was the only steroid
receptor in liver of females that was significantly affected by FAD. Expression of both
ER-a and CHG L were significantly down-regulated in the liver of females exposed to
FAD relative to that of unexposed females after 8 h. While down-regulation of
expression of most genes in the liver of females, was like that of CHG H/HM, V T G
changed only slightly as a function of time while for some genes such as ER-a, there
were no changes in effects as a function of time. In the case of CHG L down-regulation

of gene expression was greatest after 8 h, with the magnitude of the effect becoming less

65

as a function of exposure duration. Annexin max2 was 33-fold greater than that ofthe
control after 32 h. Down-regulation of gene expression of genes in the liver occurred
earlier in females than in males. ER-a in liver of males was not affected after 8 h. The
effect of FAD on V TC 1 in liver of males was greater than that in females while the effect
on V TC 1! was less than that in females.

Changes in gene expression in the gonads of male and female medaka exposed to
FAD were generally less responsive than those observed in the liver. Exposure to FAD
for 8 h caused no statistically significant alteration in expression of any of the genes
analyzed in ovary. However, HDLR, CYP21, CYP1 IB and the activins were down-, and
CYP19A was up-regulated in ovaries after 32 h of exposure. After 7 d exposure to FAD,
ovarian CYP3A, HDLR and Activin BB were significantly down-regulated while CYP19A
was up-regulated 4-fold (p-value < 0.068). Effects of FAD on the testis were broader and
more evident than those observed in ovaries. LDLR and LHR were significantly down-
regulated in testis after 8 h. After 32 h CYP21 and Activin BA were significantly down-
regulated. Genes that were up-regulated after 32 h of exposure included the testicular
hormone receptors ER-a, ER-fl. A R-a, FSHR, steroidogenic genes, including HMGR,
StAR, ("Y/’17, CYP1/B, and 3/I-HSD, HDLR, and inhibin A. After 7 d StAR and CYP1/B
were the two genes for which significant up-regulation in expression occurred in the
testis.

Exposure to 50 ug FAD/L caused a time-dependent down-regulation in
expression of CYPI9B in brains of both males and females. The other gene for which
expression was significantly altered in the female brain was GTHa with a 3.84-fold up-

regulation after 7 d. ER-/)’ was the only other gene that was significantly affected in male

66

brain (2.3-fold up-regulation). This effect, however, occurred only after 32 h of exposure

to FAD.

Time course ofHPG gene expression profiles: T RB

TRB affected genes in the liver and gonad of both male and female medaka. I7B-
trenbolone down-regulated expression of genes coding for yolk precursor and egg
envelope in liver of both male and female medaka in a time-dependent fashion (Table 3.2,
Figure 3.3). However, down-regulation in TRB exposure only occurred after 32 h. After
8 h, VTG l in females and ER-a in males were slightly up-regulated in liver. In ovary,
significant up-regulation after 8h was observed for the receptors (ER-a, FSHR and LDLR),
some steroidogenic enzymes (3/i-HSD, CYP19A) and Inhibin alpha. In contrast, after 32
h down-regulation was observed for ovarian expression of AR, CYP1 IA, CYP17, HDLR,
StAR and Activin BB. After 7 d, C YP3A, HDLR and Activin BB were significantly down-
regulated in liver of TRB exposed females. In testis, LHR, LDLR and HMGR were
down-regulated by TRB after 8 h, while the up-regulated genes at 32 h included ERA,
HDLR and inhibin—(1..

Some genes were affected by TRB exposure in the brain of both male and female
medaka. WIS-trenbolone caused time-dependent down-regulation of CYP1 9B transcripts
in brain of females but not males. l7B-trenbolone also down-regulated female ER-fl at 8
h, and GnRH R I and GnRH R1]! at 7 d. GnRH R1] was down—regulated in male brain at 7
d, but mfGnRH and GnRH Rlll were down-regulated at 8 h and 32 h, respectively, though

to a less extent.

67

DISCUSSION

F adrozo/e exposure

FAD is a potent aromatase inhibitor and has been shown to suppress production
ofestrogen, l7B-estradiol (E2), in different in vitro and in vivo systems including the
human H295R cell line and gonad tissues of different fishes [8, 10, 16]. The decreased
E2 production caused by exposure to FAD is due to its inhibition of aromatase enzyme
activity, which catalyzes the conversion of C 19 androgens to C18 estrogens such as E2.
In a study with adult fathead minnow that were exposed to increasing concentrations of
FAD between 2 and 50 ug / L in a short-term (21 d) study, a concentration-dependent
reduction in fecundity was observed [10]. Similarly, exposure to 50 ug FAD /L
significantly reduced fecundity in Japanese medaka. In accordance with the reduced egg
production, time—dependent down-regulation of egg precursor genes including VTGs and
CHGs was observed in liver of FAD exposed females (Figure 3.2). Prior to the decrease
of V TG M] and C HG H/HM, FAD exposure first down- regulated the expression of ER-a
in liver of females after 8 h. These results not only confirm the primary role of ER-a in
the transcriptional regulation of egg precursor in Japanese medaka [13], but also suggest
that exposure to 50 ug FAD /L could decrease the endogenous E2 concentration as early
as within 8 h. Furthermore, males and females display different patterns of gene
expression in the exposure of FAD. Down-regulation of ER-a and egg precursor genes in
males were slower than for females to FAD-caused decrease of E2 concentration, which
is consistent to the key function of E2 in female reproduction. In FAD exposed females

expression of CHG L decreased earlier than VTGs. Conversely, CHG HM in liver of

68

male medaka responded more rapidly than the other V T G and CHG genes. These results
suggest that the regulatory mechanisms of VTGs and CHGs were different in male and
female medaka.

Ovary response of gene expression compensates the inhibitory effect of FAD on
E2 production in Japanese medaka. The up-regulation of CYP19A transcription is
apparently a compensatory response to cope with reduced E2 synthesis. Because FAD
inhibits the conversion of C 19 androgens to C18 E2, it may result in excess production of
androgen, which is confirmed by the down-regulation of ovarian HDLR and C YPI I B
transcripts in FAD exposed Japanese medaka at 32 h. CYP1 lB mRN A encodes for the
key enzyme involved in synthesis of 1 IB-hydro-testosterone, the direct precursor of the
active non-aromatizable teleost androgen ll-ketotesterone. HDLR is one of the key
transport proteins for cholesterol, precursor for all steroids. If the down-regulated
transcription of HDLR and C YPI I B in ovary by FAD lead to a correspondingly less
production of functional proteins, the activity of l l-ketotesterone would be expected to
be less.

Gonadal transcriptional responses also explain the other adverse effects observed
in FAD exposed fish. For example, transcription of CYP1 IB and the cholesterol
transferring protein, StAR were up-regulated at 7 d, which could potentially lead to the
increased production of l l-ketotesterone (KT) and/or testosterone (T). If the testicular
response to FAD is similar between the Japanese medaka and fathead minnow, this result
explains the observed increase of plasma T and KT concentrations and marked
accumulation of sperm in the testes of FAD exposed fathead minnow [10]. In female, the

down-regulation of ovarian activin BA and activin BB might be connected to the retarded

69

oocyte maturation observed in FAD-exposed Japanese medaka [12]. Activins are dimeric
proteins consisting of two inhibin 6 subunits, BA and BB. And the three forms of activins,
activin-A, -B and -A B, are produced by homo- and hetero-dimerization of the two inhibin
[3 subunits. In vertebrates, activins have been identified as important regulators of the
reproductive axis [17]. In fish, the activin system has also been indicated to be involved
in gonadotropin-regulated ovarian functions, such as oocyte maturation. Specifically, it
has been suggested that activin-A mediates gonadotropin-induced oocyte maturation in
zebra fish [17]. Our results support the hypothesis that the retarded oocyte maturation by
FAD exposure could be related to the inhibition of activin gene expression.

Brain response is also consistent to the inhibitory effect of E2 production by FAD.
F adrozole exposure reduced the expression of brain aromatase (C YPI 9B) transcript in
both males and females. C YPI9B mRNA reduction has also been observed in FAD
treated fathead minnow [8]. Different from CYP19A, teleost brain C YPI9B is regulated
by estrogen responsive element (ERE) [18]. Therefore, FAD exposure reduces the local
E2 concentration in brain by both inhibiting brain aromatase activity and less circulating

E2 level due to inhibited ovarian aromatase activity.

1 7-fl-trenbolone exposure

l7-beta—trenbolone is the primary metabolite of trenbolone acetate, which is used
to promote growth in cattle. 17B-trenbolone has been characterized as a potent androgen
in both in vitro and. in vivo studies with mammals and fish [1 1]. Although TRB has
different biochemical properties from FAD, it induced similar response in Japanese

medaka, such as less fecundity and down-regulation of egg precursor genes in liver.

70

These observations can be explained by decreased endogenous E2 production by TRB
exposure. Reduced plasma steroid (T and E2) and V TG concentrations have been
observed in females of TRB treated fathead minnow, [l 1]. It has been postulated that
exposure to exogenous androgen such as TRB leads to the compensatory response of
decreased endogenous androgen (T and KT) production and in turn, the decreased E2
production since E2 is converted from T by CYP19 aromatase in vertebrates [19].
Similar to FAD, TRB eventually elicits less E2 level and greater concentration of
‘functional’ androgen. In TRB exposed females, the less E2 not only resulted in the
down-regulation of brain aromatase (CYP19B) and ovarian activin BB, but also elicited
compensatory responses, including greater ovarian CYP19A and less C YP3A. While
testicular LHR, LDLR and HMGR were down- regulated at 8 h to compensate androgenic
TRB exposure by slowing down steroidogenesis.

However, TRB treated Japanese medaka also displayed gene expression patterns
different from what were observed in FAD exposure, which can be explained by the
different dynamic change of E2 level in FAD or TRB exposed fish (Figure 3.3). Because
of the direct inhibition of aromatase by FAD, we hypothesize that FAD -induced estrogen
reduction can be more instant than that of TRB, which indirectly repress estrogen
production by inhibited its precursor testosterone. This subtle difference between the two
chemicals can be observed at the time —dependent gene expression changes in Japanese
medaka. For example, TRB exposure decreased expression of ER-a, VT Gs and CHGs
transcripts in liver of females after 32h, instead of 8 h in case of FAD. Nevertheless,
significant increases of ER-a in males and V T G I in females were observed in TRB

treatment at 8 h. This leads to the hypothesis that TRB exposure might initially cause a

71

temporary increase in the availability of aromatizable androgen, which in turn leads a
slight increase of E2 production. The temporary E2 production by TRB exposure up-
regulated the expression of ovarian ER-a, FSHR, LHR and inhibin A at 8 h. C YPI9B
transcripts in brain are highly sensitive in response to estrogen exposure [20]. But
CYP19B did not respond to TRB exposure in brain of males, but was down-regulated by
FAD, which suggests that the tempered decrease of estrogen by TRB exposure might not
affect the endogenous E2 level in brain. On the other hand, TRB exposure might cause
androgen ‘surge’ more rapidly than that of FAD. Compensatory response to the exposure
of androgenic TRB could be found in the reduction of gonadal CYP17, the key enzyme
synthesizing androstenedione, the direct precursor of testosterone. Decreased plasma
concentrations of l l-ketotestosterone has been observed in TRB exposed male fathead
minnows after 21 d exposure [1 1]. However, such alteration could not be seen in the
FAD treatment.

Overall, the present study examined the molecular responses in the medaka HPG
axis to exposure of anabolic androgen TRB and aromatase inhibitor FAD. F adrozole
induced a more rapid reduction of hepatic estrogen-responsive pathways than did TRB,
while exposure to FAD and TRB resulted in rapid (after 8 h) down-regulation of LHR
and LDLR in the testis to compensate excessive androgen level. The time-dependent
molecular responses by FAD and TRB developed in the present study not only help to
elucidate the mechanism of the reduced fecundity, but also present unique signatures for
the two model chemicals. Overall the results from this study demonstrated the great
utility of systematic approach of HPG axis real time PCR array in the testing of potential

EDCS using Japanese medaka.

72

Acknowledgement- This study was supported by a grant from the US. EPA Strategic to
Achieve Results (STAR) to J .P. Giesy, M. Hecker and PD. Jones (Project no. R-83 1846).
Prof. Giesy was supported by an at large Chair Professorship at the Department of
Biology and Chemistry and Research Centre for Coastal Pollution and Conservation, City
University of Hong Kong and by a grant from the University Grants Committee of the
Hong Kong Special Administrative Region, China (Project no. AoE/P-O4/04) to D. Au

and J .P. Giesy and a grant from the City University of Hong Kong (Project no. 70021 17).

73

[101

REFERENCES

Hutchinson TH, Ankley GT, Segner H, Tyler CR. 2006. Screening and testing for
endocrine disruption in fish-biomarkers as ""signposts"" not ""traffic lights"" in risk
assessment. Environ Health Perspect 114: 106-114.

Ankley GT, Daston GP, Degitz SJ, Denslow ND, Hoke RA, Kennedy SW, Miracle
AL, Perkins EJ, Snape J, Tillitt DE, Tyler CR, Versteeg D. 2006. Toxicogenomics
in regulatory ecotoxicology. Environ Sci T echnol 40: 4055-4065."

Fielden MR, Kolaja KL. 2006. The state-of-the-art in predictive toxicogenomics.
Curr Opin Drug Discov Devel 9: 84-91.

Miller DH, Jensen KM, Villeneuve DL, Kahl MD, Makynen EA, Durhan EJ,
Ankley GT. 2007. Linkage of biochemical responses to population-level effects: a
case study with vitellogenin in the fathead minnow Pimephales promelas. Environ
Toxicol Chem 26: 521-527.

Thorpe KL, Benstead R, Hutchinson TH, Tyler CR. 2007. Associations between
altered vitellogenin concentrations and adverse health effects in fathead minnow
Pimephales promelas. Aquat Toxicol 85: 176-183.

Jones PD, Coen WD, Tremblay L, Giesy JP. 2000. Vitellogenin as a biomarker for
environmental estrogens. Wat Sci T echnol 42. l-l4.

Yamaguchi A, Ishibashi H, Kohra S, Arizono K, Tominaga N. 2005. Short-term
effects of endocrine-disrupting chemicals on the expression of estrogen-responsive
genes in male medaka Oryzias latipes. Aquat T oxicol 72: 239-249.

Villeneuve DL, Knoebl l, Kahl MD Jensen KM Hammermeister DE Greene KJ
Blake LS Ankley GT. 2006. Relationship between brain and ovary aromatase
activity and isoform-specific aromatase mRNA expression in the fathead minnow
Pimephales promelas. Aquat Toxicol 76: 353-368.

Villeneuve DL, Miracle AL, Jensen KM, Degitz SJ, Kahl MD, Korte JJ, Greene KJ,
Blake LS, Linnum AL, Ankley GT. 2007. Development of quantitative real-time
PCR assays for fathead minnow Pimephales promelas. gonadotropin beta subunit

mRNAs to support endocrine disruptor research. Comp Biochem Physiol C Toxicol
Pharmacol 145: 171-183. ~

Ankley GT, Kahl MD, Jensen KM, Homung MW, Korte JJ, Makynen EA, Leino
RL. 2002. Evaluation of the aromatase inhibitor fadrozole in a short-term
reproduction assay with the fathead minnow Pimephales promelas. T oxicol Sci 67:
121-130.

74

[11]

[12]

[13]

[14]

Ankley GT, Jensen KM, Makynen EA, Kahl MD, Korte JJ, Homung MW, Henry
TR, Denny JS, Leino RL, Wilson VS, Cardon MC, Hartig PC, Gray LE. 2003.
Effects of the androgenic growth promoter l7-beta-trenbolone on fecundity and

reproductive endocrinology of the fathead minnow. Environ T oxicol Chem 22:
1350-1360.

Tompsett AR, Park J, Zhang X, Jones PD. Newsted JL, Au DWT, Chen EXH, Yu
RMK, Wu RSS, Kong RYC, Giesy JP, Hecker M. 2007. Development and
validation of an in situ hybridization system to detect gene expression along the
HPG-axis in Japanese medaka Oryzias latipes. Aquat T oxicol (submitted).

Zhang X, Hecker M, Park J, Tompsett AR, Nakazawa K, Jones PD, Newsted JL,
Au DWT, Kong RYC, Wu RSS, Giesy JP. 2008. Development and validation of a
PCR array method to study chemical induced effects on Brain-Hepatic-Gonadal
axis of the Japanese medaka. Aquat T oxicol (Accepted).

Villeneuve DL, Larkin P, Knoebl 1, Miracle AL, Kahl MD, Jensen KM, Makynen
EA, Durhan EJ, Carter BJ, Denslow ND, Ankley GT. 2007. A graphical systems
model to facilitate hypothesis-driven ecotoxicogenomics research on the teleost
brain-pituitary-gonadal axis. Environ Sci Technol 41: 321-330.

[15] Zhang X, Yu RM, Jones PD, Lam GK, Newsted JL, Gracia T, Hecker M,

[16]

[17]

[181

[19]

Hilscherova K, Sanderson T, Wu RS, Giesy JP. 2005. Quantitative RT-PCR
methods for evaluating toxicant-induced effects on steroidogenesis using the
H295R cell line. Environ Sci Technol 39: 2777-2785

Hecker M, Newsted J L, Murphy MB, Higley EB, Jones PD, Wu R, Giesy JP. 2006.
Human adrenocarcinoma H295R. cells for rapid in vitro determination of effects on
steroidogenesis: hormone production. T oxicol Appl Pharmacol 217: 1 14-124

DiMuccio T, Mukai ST, Clelland E, Kohli G, Cuartero M, Wu T, Peng C. 2005.
Cloning of a second form of activin-betaA cDNA and regulation of activin-betaA
subunits and activin type II receptor mRN A expression by gonadotropin in the
zebrafish ovary. Gen Comp Endocrinol 143: 287-299.

Pellegrini E, Menuet A, Lethimonier C, Adrio F, Gueguen MM, Tascon C, Anglade
I, Pakdel F, Kah O. 2005. Relationships between aromatase and estrogen receptors
in the brain of teleost fish. Gen Comp Endocrinol 142: 60-66.

Miracle A, Ankley G, Lattier D. 2006. Expression of two vitellogenin genes vgl
and vg3. in fathead minnow Pimephales promelas. liver in response to exposure to
steroidal estrogens and androgens. Ecotoxicol Environ Saf 63: 337-342.

[20] Kuhl AJ, Manning S, Brouwer M. 2005 . Brain aromatase in Japanese medaka

(Oryzias latipes): Molecular characterization and role in xenoestrogen-induced sex
reversal. J Steroid Biochem Mol Biol 96: 67-77.

75

Chapter 4

Responses of the Medaka HPG axis PCR array and reproduction to prochloraz and

ketoconazole

Xiaowei Zhangl'”, Markus Hecker:‘§‘#, June-Woo Park”, Amber R. Tompsett”, Paul D.
Jones”, John Newsted‘t‘", Doris Au”, Richard Kong”, Rudolf S.S. Wu” and John P.
Giesyl‘ :‘g‘fi‘n.

T Department of Zoology, Michigan State University, East Lansing, MI. USA
I National Food Safety and Toxicology Center and Center for Integrative Toxicology,
Michigan State University, East Lansing, MI. USA
§ ENTRIX, Inc., Saskatoon, SK, Canada
|| ENTRIX, Inc., Okemos, MI, USA
# Toxicology Centre, University of Saskatchewan, Saskatoon, SK, Canada
H Dept. Biomedical Veterinary Sciences, University of Saskatchewan, Saskatoon, SK,
Canada

11 Dept. Biology & Chemistry, City University of Hong Kong, Hong Kong, SAR, China

76

ABSTRACT

Effects of two model imidozole-type fungicides, prochloraz (PCZ) and ketoconazole
(KTC), on the hypothalamic-pituitary-gonadal (HPG) axis of the Japanese medaka
(Oiszias latipe) were examined by use of real time PCR (RT-PCR) array. F ourteen-wk-
old Japanese medaka were exposed for 7 d to concentrations of PCZ or KTC from 3.0 to
300 pg /L. Exposure to KTC or PCZ caused significantly less fecundity of Japanese
medaka and down-regulated expression of estrogen receptor (ER)-a and egg precursors
in livers of males and females. These effects are consistent with inhibition of gonadal 17-
B estradiol (E2) and testosterone (T) production by both KTC and PCZ. However, PCZ
was more potent than KTC both in modulating transcription and causing lesser fecundity.
Exposure to 30 ug PCZ/L resulted in 50% less fecundity and significant down-regulation
of V T G II expression, but KTC did not cause such effects at this concentration. Exposure
to PCZ caused a compensatory up-regulation in CYP19A and CYP17 expression in the
ovary, while KTC did not. The ecologically relevant endpoint, fecundity was log-log

related to mRNA level of six genes in livers of females.

Keywords: imidozole, vitellogenin, fecundity, HPG axis RT-PCR array, potency.

77

INTRODUCTION

Concerns about possible links between natural and man-made substances on endocrine
and reproductive systems in humans and wildlife (1,2) have resulted in various
international agencies initiating projects to develop new guidelines for the screening and
testing of potential endocrine disrupting chemicals in vertebrates (3,4,5). These assays
include: 1) structure activity relationships and in vitro assays that could be used to
identify potential chemical candidates; 2) short-term in vivo assays to demonstrate
relevant activity in intact animal models; and 3) long-term assays that evaluate different
reproductive and developmental stages of animals (6). Small teleost fish, such as the
Japanese medaka (Oryzias latipes) and fathead minnow (Pimephales promelas), are
appropriate models for testing endocrine disrupting chemicals (EDCS), not only from the
perspective of existing ecological impacts, but also in terms of among-species
extrapolation. The Japanese medaka is a promising small fish model test organism
because individuals are small and can be easily reared and brought into reproductive
condition. Large numbers of fish can be cultured and tested in a small area. The
physiology, embryology, and genetics of the medaka are well known because they have
been extensively studied for more than 100 yr (7). The Japanese medaka has clearly
defined sex chromosomes and sex determination (7). In addition, all mRNA/cDNA
sequences used for this project, which were necessary to design appropriate RNA probes,
are available online in the NCBI database (www.ncbi.nlm.nih.gov). Finally, there is a

marine species of medaka (Oryzias melastigma) that is very similar to the freshwater

78

species, such that these two species simultaneously provides a test system that can be
applied to freshwater, marine, and brackish ecosystems (8).

System models utilizing genomic approaches can be useful tools for mechanistic
toxicological studies of EDCS. It has been suggested that mechanistic information
derived from changes in molecular or biochemical biomarkers can be used to aid in
extrapolation of effects among species and chemicals (9,10). However, it is important to
understand linkages between alteration at the molecular and biochemical levels and
ecological relevance of adverse effects at the individual and population levels that might
relate to fitness. We have previously developed an HPG-axis-based real time PCR (RT-
PCR) array system using the Japanese medaka (l 1). To evaluate the chemical-induced
effects on reproductive endocrinology, the HPG axis-based RT-PCR array systematically
examines the transcriptional expression of 36 functionally relevant genes in brain, gonad
and liver of Japanese medaka. In addition, reproductive performance, including fertility
and fecundity are observed during the exposure. This method not only furthers our
understanding of chemical modes of action along the HPG axis, but also provides
opportunity to examine the relationship between transcriptional responses of HPG axis
and fecundity in Japanese medaka.

Two imidozole fungicides ketoconazole (KTC) and prochloraz (PCZ) were
chosen as model chemicals to assess the utility ofthe HPG-axis-based real time PCR
array system and reproductive performance of the Japanese medaka. These commonly
used fungicides have been reported to affect reproduction and development in fish and
wildlife (12,13). lmidazole fungicides were designed to inhibit a cytochrome P450 (CYP)

enzyme involved in ergosterol synthesis of fungi (1.4). However, it has also been shown

79

that these fungicides can inhibit other C YP genes, including steroidogenic cytochrome
P450 cl7ahydroxylase, 17,20-lyase (C YPI 7) and aromatase (CYP19) in mammals and
fish (15,16). Recently, KTC and PCZ have been shown to reduce both 17fl-estradiol (E2)
and/or testosterone (T) production in vitro in H295R human adenocarcinoma cells and
fathead minnow ovary explants (17,18). Although these chemicals are structurally
similar, they have also displayed different modes of action (MOA) on other biological
pathways within the reproductive tract and the HPG axis. For example, PCZ, but not
KTC, has been identified as an androgen receptor (AR) antagonist in rat and fathead
minnow (19,20).

The objective of this study was to evaluate the effects of the two fungicides, KTC
and PCZ, during a short-term exposure, on the transcriptional profiles of the key
pathways within the medaka HPG axis. It was hypothesized that the observed different
transcriptional responses by the two chemicals are due to different modes of action on the
HPG axis of Japanese medaka. In the present study, the quantitative relationship between
hepatic transcriptional responses and egg reproduction of medaka was investigated during

7-d laboratory exposures.

80

MATERIALS AND METHODS

Animals and Exposure

Male and female wild-type 0. latipes were maintained in flow-through tanks in
conditions that facilitated breeding (23-24 °C; 16:8 light/dark cycle) using the protocol
previously described (1 1,21 ,22). Before exposure, l4-wk-old, Japanese medaka were
acclimated in 10-L tanks filled with 6 L of carbon-filtered water for a period of7 (1 prior
to initiation of experiments. Each tank contained 5 male and 5 female fish. One half of
the water in each tank (3 L) was replaced daily with fresh carbon-filtered water.
Ketoconazole (KTC) and prochloraz (PCZ) were obtained from Sigma-Aldrich (St. Louis,
MO) and dissolved in the least amount of dimethyl sulfoxide (DMSO) (Sigma-Aldrich,
St. Louis, MO) possible to produce a stock solution of known concentration. After the
acclimation period, fish were exposed to vehicle control (DMSO with a final
concentration of 1210000 v/v water), 3.0, 30 and 300 ug KTC/L, or 3.0, 30 and 300 ug
PCZ/L in a 7—d static renewal exposure scenario. Exposures started at 8:30 AM and one
half of the water in each tank (3 L) was replaced with fresh carbon-filtered water dosed
with the appropriate amount of chemicals at 8:30 AM each day during exposure. Eggs
produced during the previous 24 hr period were counted and recorded before the
replacement of water. No mortalities were observed at any treatment during the exposure
period. Fish were euthanized in Tricaine S solution (Western Chemical, Ferndale, WA,
USA), and total weight and snout-vent length were recorded for each fish. Samples of
brain, liver, and gonads were collected and preserved in RNA/ate]: storage solution

(Sigma, St Louis, M0) at -20 UC until analysis of gene expression. Masses of body, brain,

81

liver and gonad of each fish were recorded. Indices including, hepatic-somatic index

(HSI), gonadal-somatic index (GSI), and brain-somatic index (BSI) were calculated.

Real time —PCR array measurement

Processing of tissues followed the previously reported protocol (1 1). Briefly, total RNA
was individually extracted from tissues according to the manufacture’s protocol with
Agilent Total RNA Isolation Mini Kit (Agilent Technologies, Palo Alto, CA). First-
strand cDNA synthesis was performed using Superscript III first-strand synthesis
SuperMix and Oligo-dT primers (Invitrogen, Carlsbad, CA). The measurement of gene
expression in brain, liver and gonad tissues was conducted using the medaka HPG axis
PCR array system described previously (1 1). Briefly, real-time Q-RT-PCR was
performed by using a 384-well ABI 7900 high throughput real time PCR System
(Applied Biosystems, Foster City, CA). PCR reaction mixtures for one hundred reactions
contained 500 uL of SYBR Green master mix (Applied Biosystems, Foster City, CA), 2
uL of 10 uM sense/anti-sense gene-specific primers, and 380 uL of nuclease-free distilled
water (Invitrogen). A final reaction volume of 10 uL was made with 2 uL of diluted
cDNA and 8 uL of PCR reaction mixtures using a Biomek automation system (Beckman
Coulter, Inc., Fullerton, CA). Quantification of target gene expression was based
comparative cycle threshold (Ct) method with adjustment of PCR efficiency according to
a previous study (23). The average ct value of the three reference genes (beta-actin,

RPL-7 and 16s) was used as reference for the expression calculation of target genes

Statistical analyses

82

Statistical analyses were conducted using the R project language (http://www.r-
project.org/). Fecundity data was analyzed using analysis of variance (ANOVA), in
which the effects of time (day) and chemicals on the daily recorded egg production were
examined. Levels of expression of genes in tissues were expressed as the fold change
relative to the average value of the average of the vehicle control. Prior to conducting
statistical comparisons of gene expression value Bartlett Test were performed to check
homogeneity of variances of data. Normality of the distributions of data was evaluated
by Shapiro-Wilk’s test. If necessary, data was log-transformed to approximate normality.
Differences of relative gene expressions among treatments were evaluated by ANOVA

followed by pair-wise t-test. Differences with p < 0.01 were considered to be significant.

83

 

Table 4.1 Transcriptional response profiles of HPG axis pathways in medaka fish
exposed to prochloraz (PCZ). Gene expression was expressed as the fold change
comparing to the corrgsponding vehicle controls a‘b

 

 

 

 

Female Male
Tissue gene 3 pg /L 30 pg /L 300 pg /L 3 pg /L 30 pg /L 300 pg /1.
ER-a -1.06 1.79 -l.26 1.78 297* 1.06
ER-l)’ 1.15 1.05 1.15 -1.07 1.17 -l.17
AR-a 1.03 -1.22 -1.08 1.16 1.14 1.13
NPY -1.58 -1.2 -1.42 -1.06 1.03 -1.35
Brain cGnRH II 196* -1.35 1.35 1.22 -l.l 1.06
mfGnRH -1.19 1.85 -l.69 1.7 366* 1.71
anRH -1.8* -1.38 -l.26 1.32 1.46 -l.02
GnRHRl 1.07 1.63 —1.46 1.65 2.6 1.47
GnRH Rll -1.29 —1.49** -l.31 1.14 1.02 1.09
GnRHRl/l -1.17 -l.27 -1.53* 1.21 1.17 -l.07
CYPI9B -1.09 -1.05 —2.69*** -l.l9 -1.61** -2.76***
ER-a -1.26 -l.5 -l.09 1.25 -l.02 1.22
ER-[J’ -1.39 -1.68 -2.26* 1.25 1.06 1.35
AR-a -1.29 1.01 1.02 1.3 1.14 1.81**
FSHR -l.l9 1.1 1.39 2.12 1.43 1.34
LHR -l.74 -1.46 1.01 1.54 1.31 -1.11
HDLR -1.05 -1.02 -2.74* 1.59 1.33 1.12
LDLR 1.01 1.05 1.04 1.25 1.53* 1.28
HMGR -l.76 -2.39 -3.73* -l.l9 -1.5 -l.62*
Gonad StAR -1.36 -1.0 -1.04 1.53 -l.03 1.4
CYP1/A -1.04 1.34 1.33 1.62 1.46 266*
CYP1/B -1.74 -2.57 -4.0* 1.65 1.75 3.02**
CYP17 1.19 1.8* 3.46*** 1.51 1.58 3.48**
CYP19A -1.39 2.46** 2.38** 248* 2.76 1.84
CYP3A -1.02 -1.4 -l.86 -3.64 -2.83 -2.34
3/)’HSD -l.67 -l.01 1.42 1.41 1.29 2.39**
ltz/tibin A 1.07 1.11 1.28 1.53 -l.05 1.22
ActivinB.»1 -2.12 ~3.09* —8.33* 257* 4.41** 2.04
ActivinBB -1.09 -l.2 -l.03 1.66 1.53 1.09
ER-a -l.0 -1.27 -54.5*** 1.34 -1.09 -1.72
ER—fl 1.32 2.0 2.02 1.2 1.41 1.88**
AR-a 1.3 1.69 -1.78 1.01 -1.11 -l.33
VTG l 1.07 -2.07 ~58.7*** 366* 2.85 -4.37
Liver VTG 11 -1.37* -3.35*** -4000*** 2.72 1.12 -9.11*
CHGH 1.07 ~1.93 -113*** 3.49** 232* —5.8*
CHG HM -1.32 —2.62** -877*** 1.55 1.3 -10.5***
CHG L -1.01 -1.78 -1020*** 4.27 2.78 -4.87*
Annexin.ma.r2 1.6 1.8 221* 1.01 -l .32 1.02

 

“ animal replicate (n= 4-6).

“ *1)< 0.01, **p< 0.001.

 

84

Table 4.2 Transcriptional response profiles of HPG axis pathways in medaka fish
exposed to ketoconazole (KTC). Gene expression was expressed as the fold change
comparingto the corresponding vehicle controls “3

 

 

 

 

KTC / Female KTC [Male
Tissue Gene 3 pg /L 30pg /L 300 pg /L 3 pg /L 30 pg /L 300 pg /L
ER-a -1.1 -1.27 -1.07 1.52 1.34 1.83
ER-p’ -1.01 1.26 -1.11 1.16 -1.06 -1.01
AR—a -1.02 -1.09 -1.12 1.15 -1.13 -1.16
NPY 1.01 -1.08 —l.04 -2.03 -1.37 -1.05
cGnRH]! -1.21 1.23 -1.01 -1.46 1.01 1.07
Brain mfGnRH -l.52 -1.84 -1.33 2.44 1.33 2.98
anRI-l -1.44 -1.68 -1.39 -1.55 -1.08 -1.15
GnRHRI -1.03 —1.18 -1.59 1.97 1.39 2.54
GnRHRII -1.54* -l.3l -1.63** -1.01 -1.06 -1.25
GnRHRIII -1.48* -1.44 -1.67** 1.1 -1.15 -1.19
CYP19B 1.06 1.14 -1.39 -1.4 -1.33 -1.7*
ER-a 1.29 -l .61 -1.49 -1.27 —1.21 1.65
ER-fl -1.78 -1.59 -2.35** -l.05 -1.15 1.53**
AR-a -1.45 -1.32 -2.36** -l.12 1.2 1.86*
FSHR 1.31 -1.05 1.21 1.03 1.4 1.61
LHR -1.6 -2.25 -3.l8** 1.17 1.31 2.09**
HDLR -1.17 -1.83 -1.29 1.04 1.38 1.3
LDLR -1.06 1.22 -1.45 1.11 1.41 1.75“
HMGR -1.75 -2.89* -3.17* -1.86* 1.31 -1.31
Gonad StAR -1.13 -2.03* -3.14*** -l.78 1.46 1.51
CYPIIA -1.03 -1.61* -1.87** -1.29 -1.43 1.34
CYP1/B -2.38 -3.17 -7.3** -1.11 1.22 228*
CYP17 1.17 -1.2 1.25 -1.44 -1.61 1.56
CYP19A 1.18 -l.39 1.08 —1.07 1.91* 2.17*
CYP3A -2.96 -2.47 1.43 -1.15 -13.6 -18.4
3,8HSD -1.16 -1.63 -2.04* -1.18 -1.17 1.59
Inhibin A 1.03 -1.69 -1.55 1.21 1.14 1.56
ActivinBA -2.67 -3.01 -8.18** 1.07 1.83 282*
ActivinBB -1.31 -1.12 -2.01** 1.08 2.13 1.37
ER-a -l.77* -1.21 -3.24*** -l.5 -1.96 6.2***
ER-p’ -1.02 1.14 1.28 -1.04 -1.45 1.19
AR-a -1.04 1.3* -2.71*** -1.23 -2.12** -3.76***
Liver VTG I -1.38 -1.49 -54.0*** 28.1 3.31 -6.46
VTG II -1.46* -1.08 -79.6*** 2.97 1.13 -1.12
CHG H -l.6 -l.64 -35.6** 1.3 1.39 -22.1**
CHG HM -1.09 1.23 -83.0*** 1.39 -l.08 -l9.6***
CHGL 1.11 1.21 -77.8** 2.52 5.96 -1.38
Annexin max2 -1.17 1.48 12.1*** -1.55 -l .92 6.45***

 

"’ animal replicate (n= 4-6).

b *p<0.01, **p<0.001.

 

85

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

m o v N o w o v m o
_ _ _ _ _ W _ _ _ _ L W
i 0 % i 0 m
U U
- m m. I m .m.
.A. .. m
N N
i w .o i w .o
w. m.
.55. 958 it % So... 95% ix: nm
do; 95». IT I W 6 anon anon IT I W 6
5.55. mam 1w: 5 Smog mam Jr 5
6:80 not W. 6:28 LY %
w w
e m
a a

2:898 20chooouox ”m 9.3398 H.259.“ ”<

.36 v Q ..._ ”92on 65:8 Eoem “meagre :80:in a 8865
mV—mCBw< ._ H c .230me UFM 2: E 223 2mE£ com a no 83298 .mwwo mo 89:5: 3.53585 2: go ANNE
88E 2e. $3? .2585 NUm 2: 5 6 5 e8 Obv— 8 ND; 8 vomomxo 8358 03:33 go @6563 0252:50

2. 2:”:—

86

Figure 4.2.

Striped view of concentration dependent response profile in PCZ exposure

of female Japanese medaka. Gene expression data from medaka treated by 3.0, 30 and
300 pg PCZ/L are shown as striped color sets on the selected endocrine pathways along
the medaka HPG axis. The legend listed in the upper right comer of the graph describes
the order of the three PCZ concentrations and the eight colors designating different fold
thresholds. LH, lutinizing hormone; FSH, follicle-stimulating hormone; E2, 1713-
estradiol; T, testosterone; HDL, high-density lipid; LDL, low-density lipid.

Medaka HPG axis Map

W
CebrSeu:

 

 

 

 

 

Hypothalamus

 

Gonadotropin releasing hormones

 

 

 

 

 

 

Pituitary \

gonadotrophs FSH LH
Em Gonad m \
m \
AR alpha m

    

Progesterone ———-> —17-l-lydroxyprogesterone

 

 

 

 

 

 

 

W0_n°'°"0 11-deoxycortleol Aldoeunedlone
HMG-CoA '
j 1 Testosterone ——
1 % cw
Cholesterol Cholesterol 17‘: ”b'flmsl Estradlol
1 Outer mfichondrleli longer mltochondriel gomus 11.ntounomm ——l
Vltell
_. m .. Estrogen n-
ER be“ mponm "3 Liver
_. mm HEM

 

  
      

 

 

 

 

 

 

87

Figure 4.3. Striped view of concentration dependent response profile in KTC

exposure of female Japanese medaka. Gene expression data from medaka treated by 3.0,
30 and 300 pg KTC/L are shown as striped color sets on the selected endocrine pathways
along the medaka HPG axis. The legend listed in the upper right corner of the graph
describes the order of the three KTC concentrations and the eight colors designating
different fold thresholds. LH, lutinizing hormone; FSH, follicle-stimulating hormone; E2,
l7B—estradiol; T, testosterone; HDL, high-density lipid; LDL, low-density lipid.

Medaka HPG axle Map

 

 

 

HYPOthalamus Gonadotropin releasing hormones

 

 

Mm, Leeann. 1| mfgnBH "“3...ij
m
”In.

  

 

 

 

 

 

 

 

 

 

 

Pituitary

gonadotrophs FSH ”‘1
Gonad mm
Ell! m
[MW

Progesterone ———. 17-Hydroxyprogesterone

l—q

MMWW 11-deoxycortieol Aldoetenedlone

-cvp1 “““tmne —

Cholesterol —1—-> Cholesterol 17" 2°” W") 1 Eetndlol-

    

 

 

 

 

 

  
      

  

Outer mitochondrlel Inner mitochondrial gums 11.x.tomtoeterone —
I membrene membrane
VItello e
a m ism,»
ER beta "SW-‘1? —_"’"°9° Liver
element ——
_. mm °"°"°9

 

 

 

 

 

 

88

RESULTS

F ecundity. Statistically significant differences were observed on the cumulative egg
productions by medaka from different treatments (Figure 4.1). Both PCZ exposure and
KTC exposure induced concentration-dependent decrease of fish fecundity in a time-
dependent manner. There was no statistically significant difference between cumulative
egg production by fish exposed to 3.0 pg PCZ/L and that of unexposed fish. However.
exposure to 30 pg and 300 pg PCZ/L resulted in significantly less fecundities of 50% and
18%, respectively, relative to that of medaka exposed to the vehicle control. The
proportions of eggs produced, relative to the solvent control were 89%, 84.2%, and
20.3% when medaka were exposed to 3.0, 30, or 300 pg KTC/L, respectively. Only
exposure to 300 pg KTC/L resulted in statistically significant less fecundity. None of the
concentrations of either KTC or PCZ caused any statistically significant effects on any of

the body indices, HIS, GSI or BSI.

Transcriptional response to prochloraz. Sex- and organ-specific transcriptional
patterns were observed in the PCZ treated medaka fish. There was a statistically
significant, concentration-dependent down-regulation of hepatic genes including ER-a,

V TO I, VT G 11, choriogenin L (CHGL), choriogenin H (CHG H) and CHG Hminor (CHG
HM) in females exposed to PCZ (Table 4.1, Figure 4.2). Exposure of males to 300 pg
PCZ /L caused the greatest (54-fold) less ER-a expression, relative to that of the vehicle
control. Alternatively, responses of genes in the livers of males exposed to PCZ were
more variable and of lesser magnitude than that of the females. Slight up-regulation of

VTG I and CHG H transcript were observed in liver of males exposed to 3.0 pg PCZ/L.

89

However, exposure to 300 pg PCZ/L significantly down-regulated expression of V T G 11,
CHG H, and CHG HM in livers of males. In contrast, exposure of males to 300 pg
PCZ/L caused a statistically significant and concentration dependent up-regulation in
expression of ER-[i in livers.

Exposure to PCZ caused differences in transcriptional responses of genes in both
ovaries and testis. There were no statistically significant differences between levels of
transcription in ovaries from fish exposed to 3.0 pg PCZ /L and that of the vehicle control.
Concentration-dependent up-regulation of CYP1 7 and CYP19A were observed in ovaries
of females exposed to PCZ with the significant differences observed in females exposed
to 300 pg PCZ /L. Conversely, expression of ovarian activin BA was inversely
proportional to PCZ exposure concentration. PCZ exposure caused concentration-
dependent up-regulation expression of AR-a and steroidogenic CYP1 1B, CYP1 1A, CYP17
and 3,8 HSD in testes. Expression to 30 pg PCZ /L caused statistically significant up-
regulation of Activin BA in liver of males.

Exposure to PCZ caused only minor effects in the brains of medaka. PCZ caused
dose-dependent down-regulation of brain-type aromatase (CYP19B) in both male and
female after 7 d of exposure. GnRH RI] and GnRH RIII were also down-regulated in

brains of PCZ exposed females, but not in males.

Transcriptional response to ketoconazole. Exposure to KTC altered the transcriptional
expression of E2 responsive genes in livers of males and females. Consistent with the
lesser fecundity, exposure to KTC caused less mRNA of ER-a, VTG 1, VTG ll, CHG L,

CHG H and CHG HM in liver of females (Table 4.2). Choriogenin H and CHG HM

90

were decreased, while ER-a was increased by 62-fold in livers of males exposed to 300
pg KTC /L. KTC caused reduction of hepatic AR-a mRNA and an increase in annexin
max2 transcript in both males and females exposed to 300 pg KTC / L.

Changes in opposite direction were observed in testis and ovaries of KTC exposed
Japanese medaka. KTC exposure caused concentration-dependent down-regulation of
ovarian receptors ER-fl, AR-a, LHR, steroidogenic genes HMGR, StAR, CYP1 1A,

CYP1 1B, and activin subunits, activin BA and activin BB. Conversely, exposure to 300
pg KTC /L caused a statistically significantly up—regulated the ER-B, LHR, and LDLR
genes in testes. In addition, transcription of CYP19A (aromastase) and activin BA genes
were up-regulated in a dose-dependent manner, with significant alteration caused by
exposure to 300 pg KTC /L.

KTC caused down-regulation of gene expression in brains of both males and
females. Concentrations of GnRH R11 and GnRH RIII were both less in brains of females
exposed to 300 pg KTC /L. C YPI 9B was slightly down-regulated (-1.7-fold change; p =

0.022) in brains of males exposed to 300 pg KTC /L.

91

DISCUSSION

Prochloraz. PCZ is a fungicide that can inhibit other C YP enzymes including
steroidogenic CYP17 and CYP19 in mammals and fish (15,16). Exposure of rats to PCZ
resulted in less production of T in the testes, due to inhibition of CYP17 activity (15). In
fathead minnows, plasma concentrations of T and 1 l-ketotestosterone were less in PCZ
exposed males and plasma E2 was less in PCZ exposed females (19). While PCZ was a
more potent suppressor of E2 production than T production as observed in the H295R
and fathead minnow ovary explant assays (18). The inhibitory effects on production of
E2 and T elicited comprehensive transcriptional responses in gonads, livers, and brains of
Japanese medaka exposed to PCZ. Because the plasma sex steroid hormones are
primarily secreted by the gonad, the concentration-dependent up-regulation of gonadal

C YP19A and CYP17 in males and females can compensate for inhibition of gonadal E2
and T production by PCZ exposure. Nevertheless, the increased mRNA level of gonadal
C YP19A and CYP17 did not change the decrease of circulating concentration of E2. In
liver, expression of ER-a mRNA and VT G as well as CHG genes were down-regulated in
PCZ exposed females and males, which suggests the local E2 concentration was
decreased in livers. As a consequence, egg production of PCZ exposed medaka was
reduced in a time and concentration —dependent manner. In brain, PCZ exposure down-
regulated brain CYP19B in both males and females, which suggests that the local E2
concentration was also decreased by PCZ exposure because brain CYP19B is primarily
regulated by E2 through the estrogen responsive element (ERE) on its promoter sequence

(24). These results demonstrate that transcriptions of genes along the HPG axis react to

92

EDCs in an organized manner among organs, and systematic investigation of the HPG
axis can help elucidate chemical-induced MOAs

Changes in transcription of other gonadal genes are also consistent with the
effects on fish previously ascribed to PCZ. The activin system is involved in
gonadotropin-regulated ovarian functions, such as oocyte maturation thus, the
concentration-dependent decrease in activin BA transcript would be expected to result in
decreased fecundity. Activin-A, a homo-dimerization of two activin BA subunits, has
been suggested to mediate gonadotropin-induced oocyte maturation in zebra fish (25).
Although the transcriptional regulation mechanism of activin BA is unknown in medaka,
its reduced expression is consistent with lesser fecundity caused by PCZ. Conversely,
transcription of activin BA in the testes was increased by exposure of Japanese medaka to
30 pg PCZ/L for 7 d. In Japanese eels stimulation of activin B mRNA is accompanied by
spennatogonia proliferation in after gonadotropin treatment (26). While the relative
number of spennatogonia was increased when fathead minnows were exposed to
concentrations of 30 to 300 pg PCZ/L exposed (19). Those results suggest a role of
activins in the onset of spennatogenesis in Japanese medaka and activin transcripts can

be used as a biomarker of chemical induced -effects on reproduction in males.

Ketoconazole exposure Similar to PCZ, KTC can also reduce E2 and/ or T production
in mammals and fish (17,18), and exposure to KTC induced similar response profiles to
those of PCZ, including inhibition of hepatic VTG and CHG transcripts, the reduction of
fecundity. Although the underlying mechanism has not been identified, PCZ and KTC

both down-regulated the expression of GnRH R II/ III in brains of females. The

93

inhibition of GnRH R was possibly a part of negative feedback mechanism that would
reduce the responsiveness of the pituitary to GnRH stimulation and lead to less secretion
of gonadatrophins in females. This hypothesis was supported by the down-regulation of
steroidogenic pathways except CYP17 and CYP19A in ovaries of Japanese medaka
exposed to either PCZ or KTC.

Previous studies suggested that KTC is a more potent inhibitor of one or more
upstream steroidogenesis enzymes than it is an inhibitor of aromatase, while the
aromatase enzyme is more sensitive to inhibition by PCZ than the upstream targets (18).
These differences can be reflected by the different transcriptional responses observed in
Japanese medaka. For example, exposure to 30 pg PCZ /L caused a 50% less fecundity
and less expression of V TC [1 and CHG HM relative to controls, while the same
concentration of KTC did not cause such effects. In addition, expression of CYP19B was
down-regulated in brains of females exposed to 300 pg PCZ/L but not of KTC. These
results are consistent with the previous observation that PCZ was more than 10-fold more
potent than KTC on the reduction of E2 production by fathead minnow ovary explants
(18). However, KTC was a more potent suppressor of T production than E2 production.
The expression of A R-(t was significantly down-regulated in livers of KTC exposed
males and females, but was unaffected in the PCZ exposure.

In summary, the present study demonstrated that transcriptional profiling with the
Japanese medaka HPG axis RT-PCR array provides a systematic understanding of PCZ
or KTC induced effect on the HPG axis of Japanese medaka. The medaka HPG PCR
array system combines the quantitative performance of real-time PCR with the multiple

gene profiling capabilities ofa microarray to examine expression profiles of over 30

94

genes along the endocrine pathways in brain, liver and gonad. The organ- gender- and
concentration —specific gene expression profiles derived by the Japanese medaka HPG
axis RT-PCR array provides a powerful tool to not only delineate chemical-induced

modes of action, but also to quantitatively evaluate chemical induced adverse effects

Acknowledgements

This study was supported by a grant from the US EPA Strategic to Achieve Results
(STAR) Program awarded to J .P. Giesy, M. Hecker and PD. Jones (Project no. R-
831846). The research was also supported by a grant from the University Grants
Committee of the Hong Kong Special Administrative Region, China (Project no. AoE/P-
04/04) and a grant from the City University of Hong Kong (Project no. 7002234) to D.
Au and J .P. Giesy. Prof. Giesy was supported by an at large Chair Professorship at the
Department of Biology and Chemistry and Research Centre for Coastal Pollution and
Conservation, City University of Hong Kong and by an “Area of Excellence” Grant (AoE

P-04/04) from the Hong Kong University Grants Committee.

95

9

10

REFERENCES

Kavlock, R. J.; Daston, G. P.; DeRosa, C.; Fenner-Crisp, P.; Gray, L. E.; Kaattari,
S.; Lucier, G.; Luster, M.; Mac, M. J.; Maczka, C.; Miller, R.; Moore, J.; Rolland,
R.; Scott, G.; Sheehan, D. M.; Sinks, T.; Tilson, H. A. Research needs for the risk
assessment of health and environmental effects of endocrine disruptors: a report
of the US. EPA-sponsored workshop. Environ. Health Perspect. 1996, 104
(Suppl 4), 715-740.

Ankley, G.; Mihaich, E.; Stahl, R.; Tillitt, D.; Colbom, T.; McMaster, D.; Miller,
R.; Bantlem, J.; Campbell, P.; Denslow, N.; Dickerson, R.; Folmar, L.; Fry, M.;
Giesy, J.; Gray, L.E.; Guiney, P.; Hutchinson, T.; Kennedy, 8.; Kramer, V.,
LeBlanc, G.; Mayes, M.; Nimrod, A.; Patino, R.; Peterson, R.; Purdy, R.; Ringer,
R.; Thomas, P.; Touart, L.; Van Der Kraak, G.; Zacharewski, T. Overview of a
workshop on screening methods for detecting potential (anti-) .
estrogenic/androgenic chemicals in wildlife. Environ. T oxicol. Chem. 1998, 17,
6887.

OECD. Report of the First Meeting of the OECD Endocrine Disrupter Testing
and Assessment (EDTA) Working Group. 1998, OECD Paris.

OECD. The Fourth Meeting of the OECD Validation Management Group (VMG)
for the Screening and Testing of Endocrine Disrupters. 2003, OECD, Paris.

EDSTAC. Endocrine Disruptor Screening and Testing Advisory Committee Final
Report. 1998.

Kanno, J .; Onyon, L.; Haseman, J .; Fenner-Crisp, P.; Ashby, J .; Owens, W. The
OECD program to validate the rat uterotrophic bioassay to screen compounds for

in vivo estrogenic responses: phase 1. Environ. Health Perspect. 2001, 109(8),
785-794.

Wittbrodt, J.; Shima, A.; Schartl, M. Medaka--a model organism from the far East.
Nat. Rev Genet. 2002, 3(1), 53-64.

Kong, R.Y.; Giesy, J.P.; Wu, R.S.; Chen, E.X.; Chiang, M.W.; Lim, P.L.; Yuen,
B.B.; Yip, B.W.; Mok, H.O.; Au, D.W. Development ofa marine fish model for
studying in vivo molecular responses in ecotoxicology. Aquat. T oxicol. 2007,

86(2), 131-41.

Ankley, G.T.; Daston, G.P.; Degitz, S.J.; Denslow, N.D.; Hoke, R.A.; Kennedy,
S.W.; Miracle, A.L.; Perkins, E.J.; Snape, J.; Tillitt, D.E.; Tyler, C.R.; Versteeg,
D. Toxicogenomics in regulatory ecotoxicology. Environ. Sci. T echnol. 2006,
40(13), 4055-4065.

Villeneuve, D.L.; Larkin, P.; Knoebl, 1.; Miracle, A.L.; Kahl, M.D.; Jensen, K.M.;
Makynen, E.A.; Durhan, E.J.; Carter, B.J.; Denslow, N.D.; Ankley, OT A

96

ll

12

l3

14

15

16

17

graphical systems model to facilitate hypothesis-driven ecotoxicogenomics
research on the teleost brain-pituitary—gonadal axis. Environ. Sci. Techno]. 2007,
41(1), 321-330.

Zhang, X.; Hecker, M.; Park, J .; Tompsett, A.R.; Nakazawa, K.; Jones, P.D.;
Newsted, J.L.; Au, D.W.T.; Kong, R.Y.C.; Wu, R.S.S.; Giesy, J. P. Development
and validation of a PCR array method to study chemical induced effects on Brain-
Hepatic-Gonadal axis of the Japanese medaka. Aquat. T oxico]. 2008 (Accepted)

Gray, L.E.; Kelce, W.R.; Wiese, T.; Tyl, R.; Gaido, K.; Cook, J .; Klinefelter, G.;
Desaulniers, D.; Wilson, E.; Zacharewski, T.; Waller, C.; Foster, P.; Laskey, J .;
Reel, J.; Giesy, J.; Laws, S.; McLachlan, J.; Breslin, W.; Cooper, R.; Di Giulio, R.;
Johnson, R.; Purdy, R.; Mihaich, E.; Safe, 8.; Sonnenschein, C.; Welshons, W.;
Miller, R.; McMaster, S.; Colbom, T. Endocrine screening methods workshop
report: Detection of estrogenic and androgenic hormonal and antihorrnonal

activity for chemicals that act via receptor or steroidogenic enzyme mechanisms.
Reprod. T oxico]. 1997, 11(5), 719-750.

Raun Andersen, H.; Vinggaard, A.M.; Hoj Rasmussen, T.; Gjermandsen, I.M.;
Cecilie Bonefeld-Jorgensen, E. Effects of Currently Used Pesticides in Assays for
Estrogenicity, Androgenicity, and Aromatase Activity in Vitro. T oxico]. Appl.
Pharmacol. 2002, 179(1), 1-12.

Henry, M.J.; and Sisler, H.D. Effects of sterol biosynthesis-inhibiting (SBI)
fungicides on cytochrome P-450 oxygenations in fungi. Pestic. Biochem. Physio].
1984, 22(3), 262—275.

Blystone, C.R.; Lambright, C.S.; Howdeshell, K.L.; Furr, J.; Stemberg, R.M.;
Butterworth, B.C.; Durhan, E.J.; Makynen, E.A.; Ankley, G.T.; Wilson, V.S.;
LeBlanc, G.A.; Gray, L.E.Jr. Sensitivity of Fetal Rat Testicular Steroidogenesis to
Maternal Prochloraz Exposure and the Underlying Mechanism of Inhibition.

T oxico]. Sci. 2007, 97(2), 512-519.

Kan, P.B.; Hirst, M.A.; Feldman, D. Inhibition of steroidogenic cytochrome P-
450 enzymes in rat testis by ketoconazole and related imidazole anti-fungal drugs.

J. Steroid Biochem. 1985, 23(6), 1023-1029.

Hecker, M.; Newsted, J.L.; Murphy, M.B.; Higley, E.B.; Jones, P.D.; Wu, R.;
Giesy, J .P. Human adrenocarcinoma (H295R) cells for rapid in vitro
determination of effects on steroidogenesis: hormone production. T oxico]. Appl.
Pharmacol. 2006, 217(1), 114-124.

Villeneuve, D.L.; Ankley, G.T.; Makynen, E.A.; Blake, L.S.; Greene, K.J.; Higley,
E.B.; Newsted, J .L.; Giesy, J .P.; Hecker, M. Comparison of fathead minnow
ovary explant and H295R cell-based steroidogenesis assays for identifying
endocrine-active chemicals. Ecotoxicol. Environ. Saf 2007, 68(1), 20-32.

97

20

21

22

23

24

25

26

27

Ankley, G.T.; Jensen, K.M.; Durhan, E.J.; Makynen, E.A.; Butterworth, B.C.;
Kahl, M.D.; Villeneuve, D.L.; Linnum, A.; Gray, L.E.; Cardon, M.; Wilson, V.S.
Effects of two fungicides with multiple modes of action on reproductive
endocrine function in the fathead minnow (Pimephales promelas). T oxicol. Sci.

2005, 86(2), 300-308.

Earl Gray, L.; Wilson, V.S.; Stoker, T.; Lambright, C.; Furr, J.; Noriega, N.;
Howdeshell, K.; Ankley, G.T.; Guillette, L. Adverse effects of environmental

antiandrogens and androgens on reproductive development in mammals. Int. J.
Androl. 2006, 29(1), 96-104.

Tompsett, A.R.; Park, J.; Zhang, X.; Jones, P.D.; Newsted, J.L.; Au, D.W.T.;
Chen, E.X.H.; Yu, R.M.K.; Wu, R.S.S.; Kong, R.Y.C.; Giesy, J.P.; Hecker, M.
Development and validation of an in situ hybridization system to detect gene
expression along the HPG-axis in Japanese medaka, Oryzias latipes. Aquat.

T oxico]. 2008, (sumitted)

Zhang, X.; Park, J.; Hecker, M.; Tompsett, A.R.; Jones, P.D.; Newsted, J.L.; Au,
D.W.T.; Kong, R.Y.C.; Wu, R.; Giesy, J .P. Time-dependent transcriptional
profiles of hypothalamic-pimitary-gonadal (HPG) axis to fadrozole and l7beta-
trenbolone in medaka (O. latipes). Environ. T oxico]. Chem. 2008, (Accepted)

Zhang, X, Yu, R.M.; Jones, P.D.; Lam,G.K.; Newsted, J .L.; Gracia, T.; Hecker,
M.; Hilscherova, K.; Sanderson, T.; Wu, R.S.; Giesy, J .P. Quantitative RT-PCR

methods for evaluating toxicant-induced effects on steroidogenesis using the
H295R cell line. Environ. Sci. Techno]. 2005, 39: 2777-2785

Pellegrini, E.; Menuet, A.; Lethimonier, C.; Adrio, F.; Gueguen, M.M.; Tascon,
C.; Anglade, 1.; Pakdel, F .; Kah, 0. Relationships between aromatase and
estrogen receptors in the brain of teleost fish. Gen.Comp.Endocrinol. 2005,
142(1—2), 60—66.

DiMuccio, T.; Mukai, S.T.; Clelland, E.; Kohli, G.; Cuartero, M.; Wu, T.; Peng, C.

Cloning ofa second form of activin-betaA cDNA and regulation of activin-betaA
subunits and activin type II receptor mRNA expression by gonadotropin in the
zebrafish ovary. Gen. Comp. Endocrinol. 2005, 143(3), 287-299.

Nagahama, Y.; Miura, T.; Kobayashi, T. The onset of spennatogenesis in fish.
Ciba F oand. Symp. 1994,182, 255-267.

Miller, D.H.; Jensen, K.M.; Villeneuve, D.L.; Kahl, M.D.; Makynen, E.A.;
Durhan, E.J.; Ankley, G.T. Linkage of biochemical responses to population-level
effects: a case study with vitellogenin in the fathead minnow (Pimephales
promelas). Environ. Toxicol. Chem. 2007, 26(3), 521-527.

98

Chapter 5

Conclusion

99

The HPG -PCR array system developed in this study represents a sensitive,
reliable and flexible monitoring tool to research chemical-induced effects along the HPG
axis in Japanese medaka. Microarray and real-time PCR are currently the two most
popular transcriptional profiling techniques. While the Japanese medaka has a relatively
well-characterized genome, it lacks robust annotation for many gene products. Therefore,
because of absence of baseline information of a large proportion of array spots, full
interpretation of data collected by a Japanese medaka microarray is not yet possible.
Alternatively, real-time polymerase chain reaction (real-time PCR) is a sensitive and
reliable technique enabling reliable quantification of mRNA in biological samples. Real-
time PCR methods have greater precision for quantification of changes in gene
expression than do the microarray techniques. In addition, real time PCR techniques are
relatively less expensive than the microarray methodologies. Thus, the real-time PCR
techniques are more powerful for investigation of chemical effects because they allow
higher throughputs at lesser cost. The Japanese medaka HPG PCR array developed in
this study combines the quantitative performance of SYBR Green-based real-time PCR
with the multiple gene profiling capabilities of a microarray to examine chemical-induced
gene expression profiles along the HPG axis. Furthermore, the pathway-focused PCR
array for Japanese medaka is an open system, in which primer sets for new biomarker
genes are ready to be added when new information is available.

The results of the study demonstrated that profiling of HPG transcripts by RT-
PCR array method represents a powerful tool for mechanistic toxicological studies of
EDCS. It has been suggested that mechanistic information derived from changes in

molecular or biochemical biomarkers can be used to aid in extrapolation of effects among

100

species and chemicals (1,2). However, it is important to understand linkages between
alteration at the molecular and biochemical levels and ecological relevance of adverse
effects at the individual and population levels that might relate to fitness. Fecundity
represents the potential reproductive capacity of an organism or population. In the
present study, four out of five selected model chemicals significantly reduced fecundity
of Japanese medaka, in a concentration—dependent manner (Table 5.1) (3-5). A similar
pattern was identified in the gene expression of liver tissue of females(Figure 5.1). Four
of the five chemicals studied, including KTC, PCZ, TRB and FAD, induced
concentration dependent down-regulation of six hepatic genes, which consists of ER-a.
V T G I, V TC 11, CHG L, CHG H and CHG HM. Principal component analysis on the
transcript expression of the selected hepatic genes among chemical treatments confirmed
that expression of these genes was correlated, with the first principle component (PC 1)
explaining 96.3% of the variance among expression of the six genes (Figure 5.2). To
reduce the dimension of gene expression data and to simplify their relationship with fish
fecundity, we further developed a hepatic index (HI; Equation 1) for each treatment by

multiplying the fold change in gene expression by the PCI.

HI = 0.236 *logw (ER-a) + 0.326 *loglo (VTG l) + 0.537 *log10(VTG II) + 0.472 *logm

(CHG L) + 0.343 *loglo (CHG H) + 0.457 *loglo (CHG HM) (l)

The H1 is a sum of log-transforrned expression levels of the six hepatic genes with
similar weighting, which represents the overall expression level of this cluster of gene.

Plotting log-fecundity as a function of H1 revealed that fecundity was directly

101

proportional to H1 (Figure 5.3A). Furthermore, a log—log, relationship was developed for

the log—fecundity as a function log-HI (Equation 2) (Figure 5.3B).

log”) fecundity = 1.616 — 0.4493 * log”) H1 (2)

The coefficient of determination for this relationship (r2) was 0.864 and the analysis of
variance test indicated that the linear relationship was statistically significant (n =1 1, F:
56.5, p < 0.001)

A major challenge in the emerging field of ecotoxicogenomics is to define the
relationships between chemically induced changes in gene expression and alterations in
conventional toxicological parameters ( 1,6). The result, for the first time, quantitatively
linked the alteration of gene expression to ecologically-relevant endpoints such as
fecundity. All of the pre-selected hepatic genes are functionally relevant to fish fecundity.
Of the 6 selected hepatic genes, VT G 1 and V TG II are yolk precursor while CHG L, CHG
H and C HG HM are egg envelop precursors, which all are regulated by E2 through ER-a.
A similar linear relationship had also been found between the production of vitellogenin
(VTG), and the reproductive success of fish (7). However, the mRNA measurement by
the RT-PCR method offers many advantages over the V T G assay based on enzyme-linked
immunosorbent assay (ELISA). First, alterations in V T G mRNA would precede changes
of VTG protein, which makes the mRNA response a more rapid response. Second, the
inherent amplification of RT-PCR method makes the measure more sensitive than ELISA
V TC assay and needs only a small amount of tissue, which would reduce the number of

fish for screening of chemical. This is especially true when small fish species is used.

102

Finally since the hepatic index, integrates the responses of six genes, and thus
circumvents the variation of any single gene, it provides a more reliable prediction of
effects of chemicals on fecundity. Therefore, the hepatic index has a potential to be used
in the quantitative assessment of chemical induced effects on reproduction of Japanese
medaka in short-term exposure.

In summary, the present study demonstrated that transcriptional profiling with the
Japanese medaka HPG axis RT-PCR array provides a systematic understanding of PCZ
or KTC induced effect on the HPG axis of Japanese medaka. The medaka HPG PCR
array system combines the quantitative performance of real-time PCR with the multiple
gene profiling capabilities of a microarray to examine expression profiles of over 30
genes along the endocrine pathways in brain, liver and gonad. The organ- gender- and
concentration —specific gene expression profiles derived by the Japanese medaka HPG
axis RT-PCR array provides a powerful tool to not only delineate chemical-induced

modes of action, but also to quantitatively evaluate chemical induced adverse effects.

103

Table 5.1. Effects of different chemicals on fecundity of Japanese
medaka in 7 d exposure.

 

 

 

Chemical Conc. F ecundity (%)
5 ng/L 91.0%
EE2 50 ng/L 92.4%
500 ng/L 65.1%
50 ng/L 99.8%
TRB 500 ng/L 46.1% (*)
5000 ng/L 26.0% (*)
3 ug/L 95.5%
Prochloraz 30 ug/L 49.8% (*)
300 ug/L 18.0% (*)
3 ug/L 89.7%
Ketoconazole 30 ug/L 84.2%
300 ug/L 20.3% (*)
1 ug/L NA.
10 ug/L N.A.
Fadrozole 50 ug/L 20.4% (*)
100 ug/L N.A.
*, p < 0.05

104

Figure 5.1 Heatmap of the concentration-dependent gene expression profiles in livers of
chemical exposed females. Gene tree was constructed by pearson correlation metric.
Chemical tree was constructed by ‘ToxClust” method, where the dissimilarity between
any two chemicals was calculated by the distance between the concentration—dependent
response curves in the exposure of both chemical.

Fold Chang

 

CHGH_L

 

CHGL L
ERA L

AR L
VTG.|| L
VTG.| L
nnexinM2 L
ERB L

_i
E
E
I
(.2

5000 ng/L

TRB 500 ng/L
50 ng/L

300 uglL

PRO 30 uglL
3 uglL

30 ug/L

KTC 30 uglL
3 uglL

100 ug/L

FAD 10.0 uglL
1.0 uglL

500 nglL

EE2 50 nglL
5 nglL

 

105

Figure 5.3 Relationship between fecundity and gene expression in livers of females.
A: fecundity vs hepatic index, the broken line shows the trend of data. B: Simple linear
regression of logIO-transformed fecundity and hepatic index. The functions describing
the relationship are: Hepatic index = 0.236 *log10 (ER-(I)+ 0.326 *log10 (VTG I)+ 0.537
*log10 (VTG II) + 0.472 *loglO (CHG L) + 0.343 *logIO (CHG H) + 0.457 *log10
(CHG HM). The formula for the regression model was: log10 (fecundity) = 1.616 —

0.4493 * log10(-hepatic index).

107

Figure 5.2 Scree plot of the percentage of variance explained by each of the Principal
Components (PC) as a percentage of the total variance of the gene expression. The six

hepatic genes included in the analysis were ER-a, VT G l, VTG II, CHG L, C HG H and
CHG HM

 

Scree plot
g- 96.3%
0 _.
(I)
g
8~
C
5
d’ 35
o ;
N
O _ [——_—1

 

 

 

 

 

 

 

PC1 PCZ PC3 PC4 PCS PC6

106

 

xmvc.o:mam1

POI No: 00.. 7 m.. at
_ _ _ b _ _

 

:80 s. 38$ “ 22.3-”. a.

".

 

 

 

x2e... ozone: Eben m> 36.58... 39... “m

09 07 08 DZ

08

(%Wpunoed

$9.. 26ng

 

 

 

 

xwuc. 0:32.. m> 35:38... “a.

DZ

08

07

08

021

(%)Aiipunoed

108

REFERENCES

Ankley, G.T.; Daston, G.P.; Degitz, S.J.; Denslow, N.D.; Hoke, R.A.; Kennedy,
S.W.; Miracle, A.L.; Perkins, E.J.; Snape, J.; Tillitt, D.E.; Tyler, C.R.; Versteeg,
D. Toxicogenomics in regulatory ecotoxicology. Environ. Sci. Techno]. 2006,
40(13), 4055-4065.

Villeneuve, D.L.; Larkin, P.; Knoebl, 1.; Miracle, A.L.; Kahl, M.D.; Jensen, K.M.;
Makynen, E.A.; Durhan, E.J.; Carter, B.J.; Denslow, N.D.; Ankley, G.T. A
graphical systems model to facilitate hypothesis-driven ecotoxicogenomics
research on the teleost brain-pituitary-gonadal axis. Environ. Sci. Technol. 2007,

41(1),321-330.

Zhang, X., Park, J ., Tompsett, A.R., Hecker,M., Jones, P.D., Newsted, J .L., Giesy,
J .P. Development and validation of a medaka brain-gonadal-liver axis model and
a real time-PCR array method to facilitate the mechanistic classification of
endocrine-disrupting chemicals (EDCS). Aquat T oxicol. 2008. (Accepted)

Zhang, X., Park, J ., Tompsett, A.R., Hecker,M., Jones, P.D., Newsted, J .L., Giesy,
J .P. Time-dependent transcriptional profiles of genes of the hypothalamic-
pituitary-gonadal (HPG) axis in medaka (0. latipes) exposed to fadrozole and
l7B-trenbolone. Environ T oxico] Chem. 2008. (Accepted)

Zhang, X., Hecker,M., Jones, P.D., Newsted, J .L., Giesy, J .P. Responses of the
Medaka HPG axis PCR array and reproduction to prochloraz and ketoconazole.
Environ Sci T echno]. 2008. (Submitted)

Moggs, J. G. Molecular responses to xenoestrogens: mechanistic insights from
toxicogenomics. T oxico]. 2005, 213(3), 177-193.

Miller, D.H.; Jensen, K.M.; Villeneuve, D.L.; Kahl, M.D.; Makynen, E.A.;
Durhan, E.J.; Ankley, G.T. Linkage of biochemical responses to population-level
effects: a case study with vitellogenin in the fathead minnow (Pimephales
promelas). Environ. T oxicol. Chem. 2007, 26(3), 521-527.

109

lljllllljjlllljllljljll[lljl‘llllll