a
w
w

“fin ch

 

N

-'-¢§x0:)lfltfi

”n: u:

 

 

«Hm _LlBRARY
2 Michigan State
2 u )9) _ University

This is to certify that the
dissertation entitled

Development and validation of novel molecular techniques to
elucidate mechanisms of endocrine disruption

presented by
June—Woo Park

has been accepted towards fulfillment
of the requirements for the

Doctoral degree in Department of Zoology-
Environmental Toxicologv

 

 

 

 

Major Profesér’s Signature
(/ - {.1 - ()S’

 

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

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K IProj/Acc8Pres/ClRC/DateDue indd

DEVELOPMENT AND VALIDATION OF NOVEL MOLECULAR TECHNIQUES
TO ELUCIDATE MECHANISMS OF ENDOCRINE DISRUPTION
By

June-Woo Park

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
DEVELOPMENT AND VALIDATION OF NOVEL MOLECULAR TECHNIQUES
TO ELUCIDATE MECHANISMS OF ENDOCRINE DISRUPTION
By

June-Woo Park

Understanding the effects of chemicals on the endocrine systems of vertebrates requires
development of methods that enable the research of the specific molecular responses to
these compounds. The overall objective of my dissertation research was to develop and
validate novel sensitive and reliable histological and molecular techniques that can be
used to elucidate modes of chemical action on the endocrine systems of vertebrates.
During the first phase of my research, I Optimized a SYBR® Green I-based
quantitative reverse transcription polymerase chain reaction (Q RT PCR) technique to be
used as a sensitive means to research the effects of EDCs on aromatase gene expression
in testicular tissue of male Xenopus laevz's. This Optimization included a comparison of
different PCR quantification methods. The comparison revealed that the comparative CT
method was optimal for the quantification of gene expression in X. laevz's testis. The
optimized Q RT PCR method was then validated by examining induction of CYP19a
mRNA gene expression in ovary and testes after exposure to forskolin, a known
aromatase inducer. There was little aromatase enzyme activity or C YPI 9a gene
expression and the two parameters were not significantly correlated. The optimized Q
RT PCR methodology was successfully used to demonstrate that the herbicide atrazine

does not up-regulate C YP19a gene expression in gonads of male X. laevis.

In the second phase of my dissertation research, I optimized an in situ
hybridization methodology using fluorescent labeling (FISH) for use in whole mounts of
a small fish, the Japanese medaka (01321215 latipes). The developed FISH methods
allowed for the evaluation of gene expression profiles Simultaneously in multiple target
tissues in sections of Japanese medaka. The key issue that was addressed during the
optimization studies was reduction of auto-fluorescence of tissues and components of the
in Situ hybridization (ISH) procedure, which is one of the major limitations in the
application of FISH on tissue sections. This was done using a combination of chemical
treatment (sodium borohydride) and an advanced confocal microscopy system. The
optimized FISH system was validated in a test exposure with the aromatase inhibitor
fadrozole by revealing tissue Specific expression of the CYP19a gene. Moreover, the
combination of FISH with histological analysis provided insights into the molecular
changes at the cellular level, indicating that the observed changes were primarily due to a
change in cell composition rather than an increase in gene expression per cell. Using the
Optimized FISH methodology, I further examined short-term effects of 17(1-
ethinylestradiol (EE2) and l7B-trenbolone (TB) on changes of three key gene
(Vitellogenin II, androgen receptor, and CYP19a) expressions in male and female
Japanese medaka. Both chemicals affected fecundity and gonad histology of medaka.
Expression of the Vit II gene was gender and tissue specific in medaka and was induced
after exposure to EE2. The AR gene was observed in both ovary and liver, but TB
Significantly induced AR gene expression in ovary only. Expression of the aromatase
gene (CYP19a) was associated primarily with early stage oocytes and was up-regulated

by EE2 at lesser concentrations but down-regulated at greater concentrations.

eve—s “Jim W‘l e—e- Ara-Isa as; was acme "1‘4 as! new.
asst-:- quI ovat- oI—s— axis 22H. smear era-3- saw.

ACKNOWLEDGEMENT

First Of all, I am deeply grateful to my academic advisor Dr. John P. Giesy for his
guidance and encouragement which have allowed me to achieve academic goal, and my
committee members, Dr. Patrick Muzzall, Dr. Steven Bursian, Dr. Jack Harkema, and Dr.
Markus Hecker for their thoughtful and critical advices throughout my Ph.D. dissertation
project. Most of all, I would like to give my special thank to Dr. Markus Hecker for his
assistance and guidance for my Ph.D. project.

I would like to thank our Aquatic Toxicology lab colleagues, Hoon, Amber, Eric,
and Howard for their fish husbandry and technical help, and Dr. Melinda Frame for her
analytical help on fluorescence images.

I would also like to tell my Sincere gratitude to my wife, Jung-Eun Lee who has
always encouraged me during my time in Michigan, and I would like to share this joy
with my mother, my brother and sister and with my lovely son Jonathan Park. Finally, I
would like to give my special thank to the grace of God and believers in my church who
have prayed for my studies at MSU.

This study was supported by a grant from the US Environmental Protection
Agency Strategic to Achieve Results (STAR) to JP. Giesy, M. Hecker, and PD. Jones
(Project no. R-831846), an Area of Excellence 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 Hong Kong University Grants Council (Project no. 7002234)

to D. Au and JP. Giesy.

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................. ix
LIST OF FIGURES ............................................................................................................ x
ABBREVIATIONS ........................................................................................................ xiv
INTRODUCTION
Background ...................................................................................................................... l
Approaches to examine EDC modes of action in vertebrate species ............................... 3
Model species ................................................................................................................... 5
Study Objectives ............................................................................................................... 9
References ...................................................................................................................... 14
CHAPTER 1
Abstract .......................................................................................................................... 19
Introduction .................................................................................................................... 20
Materials and Methods ................................................................................................... 22
Animals ....................................................................................................................... 22
Isolation of total RNA and first-strand cDNA synthesis ............................................ 23
Real-time PCR using SYBR Green I .......................................................................... 24
Synthesis of plasmid DNA standards ......................................................................... 25
Quantification Of CYP19 mRNA expression .............................................................. 26
Confirmation of test system using positive controls .................................................. 28
CYP19 aromatase activity .......................................................................................... 29
Statistical analysis ...................................................................................................... 30
Results ............................................................................................................................ 30
RT PCR amplification efficiencies, linearity and reproducibility .............................. 30
Comparison of different quantification methods for CYP19 gene expression ........... 34
Comparison of CYP19 gene expression and aromatase activity ................................ 37
Gonadal C YPI 9 gene expression after exposure to forskolin .................................... 38
Discussions ..................................................................................................................... 39

vi

Development and optimization of Q-RT PCR system to quantify CYP19 gene

expression in male X. laevz's ....................................................................................... 39
Comparison of different gene quantification methods ............................................... 42
Comparison of C YPI9 gene expression with aromatase enzyme activity ................. 44
Implications for toxicological assessment of environmental pollutants ..................... 45
Acknowledgement .......................................................................................................... 47
References ...................................................................................................................... 48
CHAPTER 2
Abstract .......................................................................................................................... 54
Introduction .................................................................................................................... 55
Materials and Methods ................................................................................................... 59
Test chemical .............................................................................................................. 59
Culture of Japanese medaka ....................................................................................... 59
F adrozole exposure ..................................................................................................... 59
ISH procedure ............................................................................................................. 60
Q RT PCR procedure .................................................................................................. 67
Histology .................................................................................................................... 67
Statistics ...................................................................................................................... 68
Results ............................................................................................................................ 70
Reduction of autofluorescence ................................................................................... 70
Tissue and cell specificity Of C YPI 9a gene expression ............................................. 71
Fadrozole exposure ..................................................................................................... 74
Discussions ..................................................................................................................... 78
Optimization of FISH ................................................................................................. 78
Validation of FISH method ........................................................................................ 80
Utilization of FISH ..................................................................................................... 81
Comparison of FISH data to morphometric and histological results ......................... 83
Acknowledgement .......................................................................................................... 85
References ...................................................................................................................... 86
CHAPTER 3
Abstract .......................................................................................................................... 92
Introduction .................................................................................................................... 93

vii

Materials and Methods ................................................................................................... 97

Test chemicals ............................................................................................................ 97
Culture of Japanese medaka (0. latipes) .................................................................... 97
Chemical exposures .................................................................................................... 97
ISH procedure ............................................................................................................. 98
Histology .................................................................................................................. 104
Statistics .................................................................................................................... 105
Results .......................................................................................................................... 105
Weight, length, biological indices and fecundity ..................................................... 105
Histology of medaka exposed to EE2 or TB ............................................................ 108
Chemical-induced gene expression changes by ISH ................................................ I 11
Discussions ................................................................................................................... 1 l7
Optimization of in situ hybridization analysis .......................................................... 1 17

F ecundity, histology and gene expression Of medaka exposed to EE2 .................... I 18

F ecundity, histology and gene expression of medaka exposed to TB ...................... 123
Acknowledgement ........................................................................................................ 126
References .................................................................................................................... l 27
CONCLUSION ................................................................................................................ 133

viii

LIST OF TABLES

Table 1.1 Reproducibility and precision of standard curve method for CYP19 and
GAPDH plasmid DNA ...................................................................................................... 34

Table 1.2 Diverse gene expression quantification methods and aromatase enzyme
activities in individual male X. laevz's ................................................................................ 36

Table 1.3 Pearson correlation coefficient (r) and probabilities (p) between the different
parameters measured .......................................................................................................... 38

Table 2.1 Probes with primers, GenBank accession number, amplicon size, and cycling
condition for conventional PCR ......................................................................................... 69

Table 2.2 Targeted genes with primers, GenBank accession number, amplicon Size, and
cycling conditions for Q RT PCR analysis ........................................................................ 69

Table 3.1 Probes with primers, GenBank accession number, amplicon size, and cycling
condition for conventional PCR ....................................................................................... 100

Table 3.2 Body weight (g) and length (mm) of Japanese medaka used in this study ...... 106

LIST OF FIGURES

Figure 1.1 GA PDH plasmid DNA standard curve. (A) Amplification curves of six
dilutions of GA PDH plasmid DNA standard from 1 x 101 to 1 x 106 copies/ttL. (B) GAPDH
plasmid DNA standard curve plotting the log copies/ILL (x) of GAPDH plasmid DNA
against CT (y), the equation was calculated by linear regression analysis (r2=0.998 and
105.7% Of PC R efficiency). (C) Melting curve Of .PCR products. showing specificity of
the reaction. (D) 1.5% agarose gel electrOphoresis of the PCR products in the serially
diluted samples ................................................................................................................... 32

Figure 1.2 C YP19 plasmid DNA standard curve. (A) Amplification curves Of Six dilutions
Of CYP19 plasmid DNA standard from 1><10l to 1><106 copies/uL. (B) CYP19 plasmid
DNA standard curve plotting the log copies/ILL (x) Of CYP19 plasmid DNA against C T
(y), the equation was calculated by linear regression analysis (r2=0.994 and 96.7% of
PCR efficiency). (C) Melting curve OfPCR products, showing specificity of the reaction.
(D) 1.5% agarose gel electrophoresis of the PCR products in the serially diluted samples
............................................................................................................................................ 33

Figure 1.3 Comparisons among quantification methods for measuring CYP19 mRNA
expression used in the Q-RT PCR system. ER and 2AACT were calculated from standard
curve method and comparative CT method, respectively. C1“ ratio represents the ratio of

C T value of CYP19 to CT value of GAPDH. (A) Represents comparison of CYPI 9 gene
expression from standard curve method to that from comparative CT method. (B)
Represents comparison Of CYI’I 9 gene expression from standard curve method to that
from CT ratio of CYP19/GAPDH. (C) Represents comparison of C1’P19 gene expression
from C T ratio of CYP/ 9/(}APDH to that from comparative CT method ............................ 37

Figure 1.4 Fold-change (x-change, mean i SD.) of CYP19 mRNA in testicular (A) and
ovarian (B) explants oernopus laevis after exposure to 100 uM forskolin (100 uM) for
20h, using the standard curve method for quantification omeNA. SC = solvent control
(0.1% DMSO) .................................................................................................................... 39

Figure 2.1 Brief steps of mRNA in Situ hybridization with fluorescence labeled riboprobe
............................................................................................................................................ 66

Figure 2.2 Autofluorescence images of juvenile medaka ovary and emission spectra of
the sections obtained after excitation with a 488 laser (A, B, and C) of CLSM and

X

autofluorescence image of section after applying Linear spectral unmixing (D) of CLSM.
(A). Section not subjected to ISH procedure; (B). section in situ hybridized without probe
and without SB treatment; (C). Section in Situ hybridized. without probe and with SB
treatment; (D). Section in Situ hybridized without probe and SB treatment after
application of linear unmixing; (E). Autofluorescence intensities of the sections. PO =
previtellogenic oocytes. Scale bar = 100 um .................................................................... 71

Figure 2.3 Expression of C YPI 9a mRNA in the ovary of juvenile Japanese medaka after
hybridization of longitudinal whole mount sections with a fluorescence riboprobe.
Expression of CYP19a mRNA was detected in the ovary hybridized with antisense probe
(A); Very weak C YP19a detection was Observed in the oocytes hybridized with sense
probe (B); no signal in the ovary hybridized without probe (C). P0 = previtellogenic
oocytes. Scale bar = 100 um ............................................................................................. 72

Figure 2.4 Fold—changes of C YPI 9a mRNA gene expression by Q RT PCR analysis in
gonads of Japanese medaka exposed to fadrozole (l , 10, and 100 jig/L), using
comparative CT method for quantification of mRNA. ,6 actin served as the internal

control gene. All data are expressed as the median :1: the interquartile range. One-way
ANOVA was used to analyze dada by treatment groups for each tissue and sex separately,
followed by SNK test for multiple comparisons. Different letters indicate Significant
difference between treatment (p<0.05). ............................................................................. 73

Figure 2.5 Expression of CYP19a mRNA in the ovary of juvenile Japanese medaka using
the optimized in Situ hybridization. Strong Signal detection of C YPI 9a mRNA in the
early stage of oocytes was Observed, while low C YPI 9a gene expression was localized in
the outer layer of follicular cell layer of the vitellogenic and matured oocytes. P0 =
previtellogenic oocytes. Bar = 100 um ............................................................................. 73

Figure 2.6 Expression of C YPI 9a mRNA in the ovary of juvenile Japanese medaka using
the optimized FISH technique in the control (A) and 100 jig/L of fadrozole treatment
group (B) for 7 days. Fluorescence signal intensity of C YPI 9a expression in randomly
selected three early stage of oocytes in the ovary of Japanese medaka exposed to
fadrozole (C) and each bar represents means :1: SD. of 4 female fish. Significantly low
signal of negative control (sense probe, right panel) and highly expressed CYP19a
mRNA in the ovary exposed to 100 jig/L of fadrozole was Observed. Scale bar = 100 um
............................................................................................................................................ 76

Figure 2.7 Expression of CYP19a mRNA in the brain tissue of female Japanese medaka
in the control (A), and brain tissue (B) and liver tissue (C) of 100 jig/L of fadrozole

xi

treatment group by the optimized in Situ hybridization using fluorescence antisense
riboprobe. No fluorescence signal detection was Observed in brain and liver tissues. Bar =
100 um ............................................................................................................................... 77

Figure 2.8 Gonadal somatic index (GSI) and liver somatic index (LSI) Of male (A) and
female (B) Japanese medaka exposed to fadrozole for 7 days. Bars represent mean and
error bars are standard deviation ........................................................................................ 77

Figure 3.1 Brief steps of mRNA in Situ hybridization with fluorescence labeled riboprobe
.......................................................................................................................................... 102

Figure 3.2 Mean values (i SEM) of liver somatic index (LSI) and gonado-somatic index
(GSI) in Japanese medaka exposed to EE2 (A) or TB (B). Significant differences relative
to the control are indicated with an asterisk (p < 0.05, n = 4 ~ 6) ................................... 107

Figure 3.3 Cumulative numbers of viable fertilized eggs spawned by male and female
Japanese medaka exposed to EE2 (A) or TB (B). Each treatment consisted of triplicate
tanks, and each tank contained 6 pairs of medaka. Significant differences relative to the
control are indicated with an asterisk (p < 0.05) .............................................................. 107

Figure 3.4 H & E stained cross section of gonads of Japanese medaka. (A) testis of
control (lOOX, bar = 20 pm). SZ: spermatozoa, ST: spermatid, SC: spennatocyte, and
SG: spermatogonia. (B) control ovary (40X, bar = 50 um). PR: primary oocyte, PO:
previtellogenic oocyte, VO: vitellogenic oocyte, and MO: matured oocyte. (C) testis of
male exposed to 500 ng/L of EE2 (200x, (40 and 400X in the boxes) bar = 10 um).
Testis-ova observed in the form of perinuclear stage, degrading spermatozoa and greater
proportion of connective tissue were observed. (D) ovary of female exposed to 500 ng/L
of EE2 (40X, bar = 50 um). AO: atretic oocyte, and SST: somatic stromal tissue. There
were fewer mature oocytes, more atretic oocytes, and larger volume of somatic stromal
tissue. (E) testis of male exposed to 5,000 ng/L of TB ( IOOX, bar = 20 um). Accelerated
spermatozoa development and fewer spennatogonia were Observed. (F) ovary of female
exposed to 5,000 ng/L of TB (40X, bar = 50 um). Predominant matured oocytes and
fewer previtellogenic oocytes were observed. ................................................................ I 10

Figure 3.5 H & E stained cross section of liver of Japanese medaka. (A) control male
showing eosinophilia, (B) male liver Of fish exposed to 500 ng/L of EE2 and (C) female
liver of control Japanese medaka Showing intense staining with hematoxylin. (D)

xii

number of hematoxylin-stained stains (mean i SEM). Significant differences relative to
the control are indicated with an asterisk (p < 0.05). Bar = 50 pm ................................ l l 1

Figure 3.6 Expression of vitellogenin II mRNA in the gonads (A and B) and liver (C and
D) of Japanese medaka after hybridization of longitudinal whole mount sections with a
fluorescence riboprobes using Optimized ISH. Expression of Vit II mRNA was detected
in the testes (A, bar = 200 um) exposed to EE2 and control ovary (B, bar = 100 pm) of
fish with hybridization of antisense probe, especially strongly in the region of
Spennatogonia in testes and primary stage of oocytes in ovary, respectively. Very weak
detected fluorescence signal in the section hybridized with sense probe. Vit II expression
in the male liver (C, bar = 50 pm) of Japanese medaka exposed to EE2 (500 ng/L) was as
high as that in the section of female liver (D, bar = 50 um) hybridized with Vit II
antisense probe. Display channel was set to green for antisense probe labeled with Alexa
Fluor 488 and to red for autofluorescence ....................................................................... 1 13

Figure 3.7 Expression of CYP19a (A — D) and AR (E and F) in the ovary and liver of
female Japanese medaka using the FISH. Expression of CYP19a was very lowly
detected in the ovary hybridized with sense probe (A), while it was specifically detected
in the cytoplasm of primary oocytes of control ovary (B, bar = 100 um), ovary exposed to
EE2 (C), and ovary exposed to TB (D, bar = 100 um). AR mRNA expression was
detected, but lowly, in the ovary (E, bar = 100 um) and liver (F, bar = 50 um) exposed to
TB. Display channel was set to green for antisense probe labeled with Alexa F luor 488
and to red for autofluorescence ........................................................................................ 1 14

Figure 3.8 Fluorescence intensity of Vit II in testes (A), ovary (B), and liver (C) of
Japanese medaka exposed to EE2. Each bar represents mean i SEM. Significant
differences relative to the control are indicated with an asterisk (p < 0.05) .................... l 16

Figure 3.9 Fluorescence intensity of AR in ovary (A) and liver (B) of Japanese medaka
exposed to TB. Each bar represents mean i SEM. Significant differences relative to the
control are indicated with an asterisk (p < 0.05) .............................................................. 1 16

Figure 3.10 Fluorescence intensity of CYP19a in randomly selected three primary stage
of oocytes in the ovary of Japanese medaka exposed to EE2 (A) and TB (B), and each bar
represents mean i SEM. CYP19a mRNA expression was not significantly different
among treatment in both EE2 and TB exposure (p > 0.05) ............................................. 1 l7

xiii

AR (1
AR [3
CLSM
CT

CV
CYP19
DEPC
DIG
DMSO
DNA
DNAse
dNTP
E2
EDC
EE2
ER

ER
ERE
FISH
GAPDH
GSI

GtH

ABBREVIATIONS

Androgen Receptor (1

Androgen Receptor [3

Confocal Laser Scanning Microscopy
Threshold Cycle

Coefficient Of Vairation
Cytochrome P450 aromatase
Diethyl Pyrocarbonate
Digoxygenin

Dimethylsulfoxide
Deoxyribonucleic Acid
Deoxyribonulease
Deoxyribonucleotide Triphosphate
17-[3 Estradiol

Endocrine Disrupting Compounds
l7a-Ethinyl Estradiol

Estrogen Receptor

Expression Ratio

Estrogen Responsive Element
Fluorescent In Situ Hybridization
Glyceraldehyde-3-phosphate Dehydrogenase
Gonadal Somatic Index

Gonadotropin Hormone
xiv

H & E
HPG axis
IACUC
IHC
ISH
LSI
MO
mRNA
PCR
PO

PR

Q RT PCR
RNA
RNAse
rRNA
SB

SC

SD
SEM
SC

ST

SZ

TB

Hematoxylin and Eosin
Hypothalamus Pituitary Gonadal Axis
lnstituted Animal Care and Use Committee
ImmunO-Histochemistry

In Situ Hybridization

Liver Somatic Index

Matured Oocytes

messenger RNA

Polymerase Chain Reaction
Previtellogenic Oocytes

Primary Oocytes

Quantitative Reverse Transcription Polymerase Chain Reaction
Ribonucleic Acid

Ribonuclease

Ribosomal Ribonucleic Acid

Sodium Borohydride

Spermatocytes

Standard Deviation

Standard Error of the Mean
Spermatogonia

Spermatids

Spermatozoa

l 7B-Trenbolone

XV

US EPA United States of Environmental Protection Agency
Vit II Vitellogenin 11.

V0 Vitellogenic Oocytes

xvi

INTRODUCTION

1. Background

During the past 20 years a significant amount of research has been conducted to
characterize the potential of chemicals to interact with the endocrine system of vertebrate
species, so called endocrine disrupters (EDCS). While some of these chemicals occur
naturally in plants or fungi such as phytoestrogens, others are of anthropogenic origin and
are represented by a wide variety Of chemicals including fungicides, certain
polychlorinated dibenzo-p-dioxins and polychlorinated biphenyls, organotins, and
polycyclic aromatic hydrocarbons, plasticizers, and synthetic hormones (Norris, 2007).
EDCs are of increasing concern in context with both human and environmental risk
assessments due to their potential to adversely impact key functions of organisms such as
reproduction, energetics, growth, and development. The US EPA has officially defined
an EDC as an exogenous agent that interferes with the production, release, transport,
metabolism, binding, action, or elimination Of natural hormones in the body responsible
for the maintenance of homeostasis, reproduction, development, and/or behavior
(Kavlock et al., 1996). In EurOpe, an EDC is defined as an exogenous substance that
causes adverse health effects in an intact organism, or its progeny, consequent to changes
in endocrine function, while a potential endocrine disruptor is a substance that possesses
properties that might be expected to lead to endocrine disruption in an intact organism
(European Commission 1996). Of importance here is the concept that endocrine
disruptors encompass more than just environmental hormones, such as xenoestrogens and

-androgens, and include any agent that can adversely affect any aspect of the endocrine

system. Advances in biomarkers have allowed progress in assessing the association
between chemical exposures and potential adverse effects in wildlife. Generally, a
biomarker can be defined as a cellular, histological, biochemical, or molecular change
that can be measured in the product of an interaction between chemical and target
molecule, resulting in the prediction of a relationship between chemical exposure and its
effect in an organism. Molecular biomarkers using genomics and proteomics can be
useful tools for measurement of exposure to very small concentrations of chemicals in an
organism.

To date, most studies employing biomarkers have involved the expression Of one
to a few gene products that are known to be related to exposure to Specific chemical
classes. While the expression of specific genes has proven useful in assessing exposure
to specific chemicals, the limited number of gene products assessed does not make it
possible to distinguish patterns of gene expression that may be used to differentiate
exposure and effects of different chemicals or chemical classes. New techniques in
molecular biology make it possible to detect alterations in the expression of many or all
genes in an organism as a result Of exposure to chemicals or other environmental
stressors. Historically, studies have primarily focused on one tissue at one Specific time
in the development of an organism. Chemicals, to which humans and wildlife might be
exposed, can result in a disruption of the endocrine system by direct and indirect
mechanisms. For instance, some chemicals are direct acting agonists or antagonists while
others act indirectly by modulating signal transduction or cybernetic systems. For
example, the triazine herbicide atrazine does not bind to the estrogen receptor (ER), but

in vitro in a mammalian cell system, atrazine has been found to up-regulate the

expression of aromatase (C YP] 9), the enzyme that transforms testosterone to estradiol.
Although atrazine does not act like a typical estrogen via binding the ER, in mammalian
cell systems it is likely to result in an estrogenic effect by increasing endogenous
estradiol production (Sanderson et al., 2000). Thus, simple, targeted screening methods
such as receptor binding assays or even receptor mediated functional assays may not
identify these other types Of effects. Also, expression of different neurO-endocrine
systems and their specific components can vary greatly during development (Sanderson
et al., 2001). Some genes are only expressed in certain tissues, while others are
expressed in specific tissues at only certain times of development. Also, when using
laboratory small animal model species, the small amounts of individual tissues available
for study and the difficulty in excising them from small organisms has limited the
efficacy of these techniques to determine effects during critical windows of time during
ontogenesis.

Therefore, sensitive and flexible monitoring tools are needed that allow for the
screening of multiple genes in multiple tissues simultaneously at any stage Of

development, without the need to dissect out the small critical tissues.

2. Approaches to examine EDC modes of action in vertebrate species

The recent advent of genomic sciences opened new perspectives to scientists
researching modes of chemical action. There are four commonly used methods for
detection and quantification of transcription, which are northern blotting, RNAse
protection assays, quantitative reverse transcription polymerase chain reaction (Q RT

PCR) and in Situ hybridization (ISH) (Bustin, 2000). Briefly, northern blotting can

provide information about mRNA size, alternative splicing, and the integrity of mRNA
samples. The RNAse protection assay is useful for mapping transcript initiation and
termination sites and for discriminating between related mRNAS of similar Size (Bustin,
2000). While these methods have proven useful in developing gene expression profiles
that can be related to potential modes of action of EDCS, I focused on two methods, Q
RT PCR and ISH because they are the most sensitive and allow for the detection Of
spatial changes of gene expression, respectively. Also, northern blotting and RNAse
protection assays require relatively large amounts of mRNA, and thus, do not provide the
necessary sensitivity to detect low expression of genes such as CYP19a in gonads of male
frogs which was the aim of Chapter 1.

Quantitative (real-time) reverse transcriptase polymerase chain reaction (Q RT
PCR) is a sensitive and flexible technique that can detect small quantities of mRNA in
small amounts of tissue (Bustin, 2000 and 2002) and has been applied successfully as a
tool for the investigation of multiple fiinctionally related genes in endocrine disrupter
research (Rotchell and Ostrander, 2003). This technique, which amplifies the number of
copies of mRNA many times, can theoretically measure as little as a Single molecule of
the target mRNA (Bej et al., 1991). However, there are some disadvantages of using RT
PCR such as non-specific DNA amplification due to high sensitivity of Taq polymerase,
miS-incorporation Of nucleotides by lack of 3’-5’ exonulease capacity in T aq polymerase,
and difficulty in PCR amplification of long transcripts (> 5k bp) (Bej et al., 1991). Most
of all, based on empirical observation, it was difficult to extract large amount of RNA
from individual small tissue such as brain and testis of fish without pooling them, and, in

some cases extracted RNA was not purified and/or degraded. Enough input of non-

degraded RNA as template is critical for reverse transcription efficiency to synthesize
cDNA and in turn will affect on the efficiency of target gene quantification. Therefore,
additional methods such as in Situ hybridization that allow for detection and
quantification of intact mRNA directly in the tissue of interest without extraction or
further treatments such as reverse transcription and amplification are needed.

In Situ hybridization (ISH) involves the specific annealing of labeled probes to
complementary sequences of interest in fixed tissues, followed by Visualization of the
labeled probes. The major advantage of this method is that it provides a sensitive means
to localize and quantify the mRNA for specific genes in organs and tissues of interest in a
manner that is consistent with other methods that are used to detect lesions including
histopathology and immuno-histochemistry. ISH also provides a direct visualization of
the spatial location of specific sequences, which is crucial for elucidation of the
organization and function of genes. As a result, ISH has become an important tool in a
number of fields, such as the research of chromosomal arrangement, viral infection, and

analysis of gene function (Wilkinson, 1992; Jin and Lloyd, 1997).

3. Model species

There is increasing concern regarding the effects of EDCS on aquatic vertebrates
such as amphibians and fish because they are continuously exposed and in direct contact
with these compounds. Especially amphibians have been regarded as good indicator
Species for the exposure to environmental contaminants because they are subject to both
aqueous and terrestrial exposure routes. Recently, there was an increasing concern about

the potential of triazine herbicides to interact with the endocrine system of male frogs by

inducing aromatase which catalyzes the conversion of androgens to estrogens, resulting
in an increase of endogenous estrogen production and subsequently causing feminization
or demasculinization of males (Hayes et al., 2002). However, the investigation of this
proposed mode of action failed to date due to limitations in the sensitivity of the applied
tests systems, especially in males where the endpoints Of interest (gonadal aromatase)
were naturally very low, and typically below the method detection limits of the utilized
assays (Hecker et al., 2004). Therefore, to increase our ability to determine possible
changes in aromatase activity in the testis, a more sensitive test system such as an
optimized Q RT PCR method was needed by examining subtle effects on the aromatase
gene (CYP19a) expression.

In the first phase of my dissertation research, I developed a Q RT PCR method for
detection and quantification of CYP19a gene expression in the testes of male African
clawed frog (Xenopus laevis). This species has been used as one of the key models to
assess the effects of atrazine on the endocrine system of frogs (Hayes et al., 2002; Carr et
al., 2003; Coady et al., 2004; Hecker et al., 2004 and 2005a). The African clawed frog is
a carnivorous frog native in Africa, and inhabits warm and stagnant grassland ponds.
Tadpoles metamorphose within 2 or 3 mo of egg laying, and frogs reach sexual maturity
within one or two years, depending on breeding conditions such as temperature and
frequency of feeding. X. laevis is easily maintained in the laboratory, and has been
commonly used as model species for developmental studies (Wu and Gerhart, 1991).

In general, four principal actions for how EDCS can affect reproduction include
estrogenic, antiestrogenic, androgenic, and antiandrogenic effects that can lead to

feminization, neutralization of sexual differentiation, masculinization, and feminizing

effects, respectively (Kloas, 2002). Generally sexual differentiation is primarily
determined by on a sex-specific basis that leads to functional gonadal sex development
and that in turn is responsible for secondary sexual differentiation. Sexual differentiation
of frogs is regulated by the ratio of sex steroids, estrogens and androgens, in developing
eggs and embryos and can be shifted by supplements of estrogens or androgens by
changing the ratio. At the time Of hatching, expression of estrogen and androgen receptor
mRNAs are increased, indicating that early stage after hatching is sensitive for sexual
differentiation. For example, Xenopus frog larvae develop into 50% males (22) and 50%
females (ZW) under normal physiological conditions, however, exposure tO estrogen shift
sex ratio to feminization, while treatments of methyltestosterone and dihydrotestosterone
induced significant masculinization. Sexual differentiation also can be shifted by
regulation of ER and AR mediated cellular responses. Antiandrogens, such as
vinclozolin, caused feminization by suppressing the androgen receptor-mediated cellular
processes, leading indirectly to alter the ratio resulting in feminization. Antiestrogens,
such as tamoxifen, leaded to neutralization by blocking estrogen-induced developmental
processes in genetic females and males, resulting in a general depression of gonadal
development (Kloas, 2002). However there are still debates whether such general
characterization of modes of action is applicable to all amphibians. Only estrogen always
caused a clear feminization, while the effects of masculization by testosterone are
dependent on the Species. Also antiestrogenic and antiandrogenic chemicals resulted in
contradictory data (Rastogi and Chieffi, 1975).

In the second phase of my dissertation research, I developed and validated

techniques for the research of spatial changes in gene expression using a whole animal

approach, the Japanese medaka (OryZI'as latipes). The test system applied in these
studies utilized whole mount sections, an approach requiring a small model fish. 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 years
(Wittbrodt et al., 2002). The medaka represents an important test system for
environmental research and is widely used for testing endocrine disrupters in
ecotoxicology. Another advantage of the medaka is its rapid development and ease of
breeding, producing eggs on a regular schedule under the appropriate conditions of
lighting and temperature. Currently, several medaka genome projects are underway, and
over 90% of the medaka genome has already been sequenced so that DNA sequences are
already available for most of the genes of interest for this study. This strain exhibits
sexual dimorphism with males being orange-red, while females are white, and based on
their coloring the genetic sex can be determined. This was a great advantage for this
study because the genetic sex of each individual could be determined prior to the
initiation of the studies and at time of sampling. Furthermore, our research team had
already developed some of the basic techniques required for my studies, including the
sectioning and tissue fixing and preparation of ISH (Kong, et al., 2008; Tompsett et al.,
2008)

Many environmental chemicals exhibit estrogenic or androgenic activity.
Inappropriate induction of vitellogenin in juvenile or male fish has become one of the
most notable biological responses Of fish involved in EDC exposure, especially in
estrogenic compounds. The most dramatic increase of vitellogenin levels appear to be in

fish exposed to sewage discharges which contain sufficiently high concentration of

estrogenic compounds (Purdom et al., 1994). Vitellogenin induction was also exhibited
in fish exposed to municipal sewage discharges which might have natural and/or
synthetic estrogen (Folmar et al., 1996). Phytoestrols in pulp and paper mill effluents,
such as B—sitosterol also induced Vitellogenin levels in male and juvenile female fish
(Tremblay and Vad Der Kraak, I999). Vitellogenin induction in male can cause reduced
calcium concentration in scales and skeleton, enlarged livers, kidney damage and reduced
testicular growth (Herman and Kincaid, I988; Harries et al., 1997). Further male
genotypes become apparently normal female phenotypes as a result of feminization effect
during the sensitive part of gonadal development, such as development of testis-ova, and
oviduct (Wester and Canton, 1986; Gimeno et al., 1996). Thus, these estrogenic effects
will reduce male’s reproductive fitness, such as reduced fertilization success by
impairment in the sperm quality and abnomial reproductive behavior. Androgenic and
antiandrogenic effects on fish have not been well documented as the estrogenic and
antiestrogenic effects. Several studies have reported the masculinization of female
exposed to kraft mill effluents containing stigmastanol degradation product which have
androgenic properties. These masculinization features include development of
gonopodium in female and hermaphroditic conditions in male (Howell and Denton, 1989;

Bortone et al., 1989; Bortone and Cody, 1999).

4. Study Objectives
The overall objective of my dissertation research was to develop and validate
novel sensitive and reliable molecular techniques that can be used to elucidate the

endocrine modes of chemical action in vertebrates. The techniques developed allow for

the determination of mechanisms of toxic action of single chemicals or complex mixtures.
Specifically, using the developed and validated methods I examined potential effects of
chemicals at the level of gene expression by measuring the amount Of mRNA and
evaluated how changes in gene expression relate to reproductive functioned in the test

organism. The specific objectives of my research were:

0 To develop and Optimize Q RT PCR analysis to measure lowly expressed genes
such as CYP19 in testicular tissue of male X. laevis

0 To develop and optimize an ISH protocol using fluorophore-labeled probes to
detect specific mRNA sequences in whole animal sections of Japanese medaka (0.
latipes).

0 To validate the FISH methods developed in this study by examining CYP19a
mRNA in medaka after exposure to the specific aromatase inhibitor fadrozole, and
by comparing the gene expression data with that obtained during parallel Q RT
PCR analysis using the techniques developed during the above studies with X.
laevis

o~ To characterize the effects of l7u-ethinylestradiol (EE2) and 17B-trenbolone (TB)
on the tissue specific expression of CYP19a, vitellogenin ll (Vit II), and androgen
receptor a (AR) in whole sections of medaka using the Optimized FISH protocol.

0 Compare the changes in gene expression with histological, physiological, and/or
organismal responses to further our understanding of the molecular mechanisms of

the tested EDCS.

l0

Once the methods had been developed and validated, they could be adapted for
use with other genes and/or species of interest and used to efficiently and completely
screen for endocrine disruptor effects. Moreover, these methods have the potential to
Significantly improve molecular profiling approaches to better understand mechanisms of
action of individual compounds or complex mixtures, improving risk assessment of the
chemicals in general to both humans and wildlife. The testable hypotheses of my

research were:

0 Developed Q RT PCR analysis is sensitive to detect and quantify weakly
expressed gene such as CYP19a in testicular tissues of male Xlaevis.

o Developed Q RT PCR analysis can be applied as a tool for studying enzymes with
low activity.

0 ISH method is a proper tool to detect changes in target gene expression profiles
along the HPG-axis.

0 Gene expression profiles of key genes by means of ISH Show sex-and tissue-
specific patterns.

0 There are specific effects on changes of target gene expression profiles in an
organism after exposing to “model” compounds.

0 Specific changes in gene expression profiles by means of ISH are related with
histological relevant endpoints.
In Chapter 1, I optimized a Q RT PCR technique to be used as a sensitive mean to

research the effects of EDCS on aromatase gene expression in amphibians (Park et al.,

2006). Specifically, I developed an assay to measure mRNA for CYP19a (aromatase). I

ll

then applied the technique to test the hypothesis that the common pesticide, atrazine,
could up-regulate expression of gonadal aromatase in the African clawed frog (Xenopus
laevis). Due to low detection limit in enzymatic assays to measure aromatase enzyme
activity in testicular tissues, it had been difficult to test this hypothesis previously. Thus,
I optimized a SYBR“?9 Green I-based quantitative reverse transcription polymerase chain
reaction (Q RT PCR) method to quantify CYP19a mRNA in testicular tissue of male X.
laevis. This optimization included a comparison of different PCR quantification methods,
which revealed that of the methods tested, the absolute standard curve and the
comparative CT method were optimal for the quantification of gene expression in X.
laevis testis. Due to the labor intensity of the standard curve method, however, it was
decided to use the comparative CT method for future studies. The optimized Q RT PCR
method was then validated by examining induction of CYP19a mRNA gene expression in
ovary and testes after exposure to forskolin, a known aromatase inducer. Although both
aromatase activity, determined by the tritium release assay, and CYP19a mRNA were
detectable in testes of X. laevis, there was little aromatase enzyme activity, or CYP19a
gene expression and the two parameters were not Si gnificantly correlated. The Q RT
PCR methodology optimized in this phase was used to measure CYP19a gene expression
changes in gonads of male X. laevis exposed to the herbicide atrazine and it was
successfully used to demonstrate that the atrazine does not up-regulate CYP19a gene
expression in the tissues (Hecker et al., 2005a, and b).

In Chapters 2 and 3, I optimized an in situ hybridization methodology using a
fluorescent labeling (FISH) for use in whole mounts of a small fish, the Japanese medaka

(OryZI'as latipes) (Park et al., 2008a, b; Tompsett et al., 2008; Zhang et al., 2008a, b, c).

12

The FISH methods developed allowed for the evaluation of gene expression profiles
simultaneously in multiple target tissues in sections of Japanese medaka. The key issue
that was addressed during the Optimization studies was reduction of auto-fluorescence of
tissues and components of the ISH procedure, which is one of the major limitations in the
application of FISH on tissue sections. This was done using a combination of chemical
treatment (sodium borohydride) and an advanced confocal microscopy system. The
optimized FISH system was validated in a test exposure with the aromatase inhibitor
fadrozole by revealing tissue Specific expression of the CYP19a gene. F adrozole (100
ug/L) up-regulated CYP19a expression and this trend was comparable with that obtained
from Q RT PCR analysis. Moreover, the combination of FISH with histological analysis
provided insights into the molecular changes at the cellular level, indicating that the
Observed changes were primarily due to a change in cell composition rather than an
increase in gene expression per cell. Using the optimized FISH methodology, I further
examined short-term effects of 17a-ethinylestradiol (EE2) and 17B-trenbolone (TB) on
changes of three key genes (Vitellogenin II, androgen receptor, and C YP19a) expressions
in male and female Japanese medaka (Oryzias latipes) (Park et al., 2008b). Both
chemicals affected fecundity and gonad histology of medaka. Expression of the Vit II
gene was gender and tissue Specific in medaka and was induced after exposure to EE2.
The AR gene was observed in both ovary and liver, but TB significantly induced AR gene
expression in only the ovary. Expression of the aromatase (CYP19a) gene was primarily
associated with early stage oocyte and was up-regulated by EE2 at lesser concentrations

but down-regulated at greater concentrations.

References

Bej, AK., Mahbubani, MH., Atlas, RM., 1991. Amplification Of nucleic acids by
polymerase chain reaction (PCR) and other methods and their applications.
Critical Reviews in Biochemistry and Molecular Biology, 26: 301—334.

Bortone, SA., and Cody, RP. 1999. Morphological masculinization in poeciliid females
from a paper mill effluent receiving tributary of the St. johns River, Florida, USA.
Bulletin of the Environmental Contamination and Toxicology, 63: 150-156.

Bortone, SA., Davis, WB., and Bundrick, CM. 1989. Morphological and behavioral
characters in mosquito fish as potential bioindication of exposure to kraft mill

effluent. Bulletin of the Environmental Contamination and Toxicology, 43: 370-
377.

Bustin, SA. 2000. Absolute quantification of mRNA using real-time reverse transcription
polymerase chain reaction assays. Journal Of Molecular Endocrinology, 25: 169—
193.

Bustin, SA. 2002. Quantification of mRNA using real-time reverse transcription PCR
(RT-PCR): trends and problems. Journal of Molecular Endocrinology, 29: 23-39.

Carr, JA., Gentles, A., Smith, EE., Goleman, WL., Urquidi, LJ., Thuett, K., Kendall, RJ.,
Giesy, JP., Gross, TS., Solomon, KR., and Van Der Kraak, G. 2003. Response of
larval Xenopus laevis to atrazine: assessment of gonadal and laryngeal
morphology, Environmental Toxicology and Chemistry, 22: 396-405.

Coady, KK., Murphy, M., Villeneuve, DL., Hecker, M., Jones, PD., Carr, JA., Solomon,
K., Smith, EE., Van Der Kraak, G., Kendall, RJ., and Giesy, JP. 2004. Effects of
atrazine on metamorphosis, growth and gonadal development in the green frog

(Rana clamitans). Journal of Toxicology and Environmental Health, Part A, 67:
941-957.

European Commission. 1996. European workshop on the impact of endocrine disruptors
on human health and wildlife. Weybridge, 2-4 Dec. 1996. Report EUR 17594,
Environment and Climate Research Programme, DG XII, European Commission.

Gimeno, S., Gerritsen, A., Bowmer, T., and Komen, H. 1996. Femiziation of male carp.

l4

Nature, 384: 221-222.

Folmar, LC., Denslow, ND, Rao, V., Chow, M., Crain, DA., Enblom, J., Marcino, J., and
Guillette, LJ. 1996. Vitellogenin induction and reduced wrum testosterone
concentrations in feral male carp (Cyprinus carpio) captured near a major

metropolitan sewage treatment plant. Environmental Health Perspectives, 104:
1096-1 101.

Harries, JE., Sheahan, DA., Jobling, S., Matthiessen, P., Neall, P., Sumpter, JP., Tylor,
T., and Zaman, N. 1997. Estrogenic activity in five United Kingdom rivers
detected by measurement of vitellogenesis in caged male trout. Environmental
Toxicology and Chemistry, 16: 534-542.

Hayes, TB., Collins, A., Lee, M., Mendoza, M., Noriega, N., Stuart, AA., Vonk, A., 2002.
Hermaphroditic, demasculinized frogs after exposure to the herbicide atrazine at
low ecologically relevant doses. Proc. Natl. Acad. Sci. U. S. A. 99, 5476—5480.

Hecker, M., Giesy, JP., Jones, PD., Jooste, AM., Carr, JA., Solomon, KR., Smith, EE.,
Van Der Kraak, G., Kendall, RJ., du Preez, L. 2004. Plasma sex steroid
concentrations and gonadal aromatase activities in African clawed frogs
(Xenopus laevis) from South Africa. Environmental Toxicology and Chemistry,

23: 1996-2007.

Hecker, M., Kim, WJ., Park, JW., Murphy, MB., Villeneuve, DL., Coady, K., Jones,
PD., Solomon, KR., Van Der Kraak, G., Carr, JA., Smith, L. EE., du Preez, L.,
Kendall, RJ., and Giesy, J P. 2005a. Effects of estradiol and atrazine on plasma
sex steroid concentrations, gonadal aromatase activity and ultrastructure of the
testis in Xenopus laevis. Aquatic Toxicology, 72: 383-396.

Hecker, M., Park, JW., Murphy, MB., Jones, PD., Solomon, KR., Van Der Kraak, G.,
Carr, J A., Smith, EE., du Preez, L., Kendall, RJ., and Giesy, JP. 2005b. Effects of
atrazine on CYP19a gene expression and aromatase activity in testes and on sex
steroid concentrations in plasma of male African clawed frogs (Xenopus Laevis).
Toxicological Sciences, 86: 273-280.

Herman, RL., and Kincaid, HL. 1988. Pathological effects of orally administrated
estradiol to rainbow trout. Aquacluture, 72: 165-172.

l5

Howelll, WM., and Denton, TE. 1989. Gonopodial morphogenesis in female
mosquitofish, Gambusia aflinis affim's, mascluinized by exposure to degradation
products from plant sterols. Environmental Biology of Fishes, 24: 43-51.

Jin, L., and Lloyd, RV. 1997. In Situ hybridization: Methods and applications. Journal of
Clinical Laboratory Analysis, 11: 2-9.

Kavlock, R., DAston, GP., DeRosa, C., Fenner-Crisp, B, Gray, LE., Kaatari, S., Lucier,
G., Luster, L., Mac, MJ., Macza, C., Miller, R., Moore, J., Rolland, R., Scott, G.,
Sheehan, DM., Sinks, T., and Tilson, HA. 1996. Research needs for the risk
assessment of health and environmental effects of endocrine disruptors: A report
of the US. EPA-Sponsored workshop. Environmental Health Perspective, 104
(Suppl): 715-740.

Kong, RYC., Giesy, JP., Wu, RSS., Chen, EXH., Chiang, MWL., Lim, PL., Yuen, BBH.,
Yip, BWP., Mok, HOL., and Au, DWT. 2008. Development of a marine fish
model for studying in vivo molecular responses in ecotoxicology. Aquatic
Toxicology, 86:131-141.

Kloas, W. 2002. Amphibians as a model for the study of endocrine disruption.
International Review of Cytology, 216: 1-57

Norris, DO. 2007. An overview of chemical bioregulation in vertebrates. In: Norris, D.O.
(Ed), Vertebrate endocrinology, Elsevier Academic Press, Boston, MA, pp: 1-29

Park, JW., Hecker, M., Murphy, MB., Jones, PD., Solomon, KR., Van Der Kraak, G._,
Carr, JA., Smith, EE., du Preez, L., Kendall, RJ., and Giesy, JP. 2006.
Development and optimization of a Q-RT PCR method to quantify CYP19 mRNA
expression in testis of male adult Xenopus laevis: comparisons with aromatase
enzyme activity. Comparative Biochemistry and Physiology, Part B, 144: 18-28.

Park, JW., Tompsett, A., Newsted, JL., Jones, PD., Au, D., Kong, R., Wu, RSS., Giesy,
JP., and Hecker, M. 2008a. Fluorescence in Situ hybridization techniques (FISH)
to detect changes in CYP19a gene expression of Japanese medaka (OryZias
latipes). Toxicology and Applied Pharmacology, (Submitted).

Park, JW., Tompsett, A., Newsted, JL., Jones, PD., Au, D., Kong, R., Wu, RSS., Giesy,
JP., and Hecker, M. 2008b. Effects of ethinylestradiol and trenbolone on

l6

histology and gene expression of Japanese medaka (Oryzias latipes) using a
combination of fluorescence in Situ hybridization (FISH) and traditional histology.
Toxicological Sciences (Submitted).

Purdom, CE., Hardiman, PA., Bye, VJ ., Eno, NC., Tyler, CR., and Sumpter, JP. 1994.
Estrogenic effects of effluents from sewage treatment works. Chemistry and
Ecology, 8: 275-285.

Rastogi, PK., and Chieffi, G. 1975. The effects of antiestrogens and anitandrogens in
nonmammalian vertebrates. General and Comparative Endocrinology, 26: 79-91.

Rotchell, J M., and Ostrander, GK. 2003. Molecular markers of endocrine disruption in
aquatic organisms. Journal of Toxicology and Environmental Health, Part B:
Critical Review, 6: 453-496.

Sanderson, JT., Letcher, RJ., Heneweer, M., Giesy, JP., and Van den Berg, M. 2001.
Effects of Chloro-S—Triazine Herbicides and Metabolites on Aromatase (CYP19)
Activity in Various Human Cell Lines and on Vitellogenin Production in Male
Carp Hepatocytes. Environmental Health Perspectives, 109: 1027-103 1.

Sanderson. JT., Seinen, W., Giesy JP., and van den Berg, M. 2000. 2-chloro-S-Triazine
Herbicides Induce Aromatase (CYP-19) Activity in H295R Human
Adrenocortical Carcinoma Cells: A Novel Mechanism for Estrogenicity.
Toxicological Sciences 54:121-127.

Tompsett, AR., Park, JW., Zhang, X., Jones, PD., Newsted, JL., Au, DTW., Chen, EHX.,
Yu, RMK., Wu, RSS., Kong, RYC., Giesy, JP., and Hecker, M. 2008.
Development and validation of an in Situ hybridization system to detect gene
expression along the HPG-axis in Japanese medaka, 0ryzias latipes. Achieves of
Environmental Contamination and Toxicology, (Submitted).

Tremblay, L., and Van Der Kraak, G. 1999. Comparison between the effects of the
phytosterol b-sitosterol and pulp and paper mill efflurents on sexually immature
rainbow trout. Environmental Toxicology and Chemistry, 18: 329-336.

Wester, PW., and Canton, J H. 1986. Histopathological study of Oryzias latipes (Medaka)
after long-term B-hexachlorocyclohexane exposure. Aquatic Toxicology, 9: 21-
45.

17

Wilkinson, DG. 1992. The theory and practice of in Situ hybridization. In: Wilkinson,
D.G. (Ed), In Situ hybridization: A practical approach, Oxford University Press,
New York, NW, pp: 1—13.

Wittbrodt, J ., Shima, A., and Schartl, M. 2002. Medaka — a model organism from the Far
East. Natural Reviews Genetics. 3: 53-64.

Wu, M., and Gerhart, J. 1991. Raising Xenopus in the laboratory, In: Kay, BK., Peng, HB.
(Eds), Xenopus laevis: Practical uses in cell and molecular biology, Academic
Press, Inc., San Diego, CA, pp: 3-17.

Zhang, X., Hecker, M., Park, JW., Tompsett, AR., Newsted, J L., Nakayama, K., Jones,
PD., Au, D., Kong, R., Wu, RSS., and Giesy, JP. 2008a. Real time PCR array to
study effects of chemicals on the Hypothalamic—Pituitary-Gonadal axis of the
Japanese medaka. Aquatic Toxicology, (Submitted).

Zhang, X., Park, J W., Hecker, M., Tompsett, AR., Jones, PD., Newsted, JL., Au, D.,
Kong, R., Wu, RSS., and Giesy, JP. 2008b. Time-dependent transcriptional
profiles of hypothalamic-pituitary—gonadal (HPG) axis in medaka (0. latipes)
exposed to fadrozole and l7beta-trenbolone. Aquatic Toxicology, (Submitted).

Zhang, W., Park, J W., Tompsett, AR., Jones, PD., Newsted, JL., Au, D., Wu, RSS.,
Giesy, J P., and Hecker, M. 2008c. Responses of the Medaka HPG axis PCR array
and reproduction to prochloraz and ketoconazole. Environmental Sciences and
Technology, (Submitted).

CHAPTER 1

Development and Optimization of a Q-RT PCR method to quantify CYP19 mRNA
expression in testis of male adult Xenopus laevis: Comparisons with aromatase enzyme

activity

Abstract

Due to limitations of the currently used enzymatic assays, it is difficult to determine
aromatase activity in testicular tissue of amphibians. Quantitative reverse transcription
polymerase chain reaction (Q-RT PCR) is a sensitive and reliable technique to detect low
amounts of mRNA for specific genes. This study was designed to develop and optimize
a SYBR Green I-based Q-RT PCR method to quantify CYP19 mRNA in testicular tissue
from male Xenopus laevis. Four quantification methods for measuring C YP19 mRNA
expression were compared. The established test system proved to be highly sensitive
(detectable mRNA copies < 10), reproducible (interassay CV < 5.4%, intraassay CV <
0.9%), precise and Specific for the CYP19 gene. To confirm the validity of the applied
test system, an ex vivo testicular and ovarian explant study with a known inducer of
aromatase, forskolin, was conducted. Forskolin induced C YPI 9 gene expression in both
ovarian (3.7-fold) and testicular (2.6-fold) explants. Of the four quantification methods,
the absolute standard curve and the comparative CT method appear to be optimal as
indicated by their highly significant correlation (r2 = 0.998, p < 0.001). In conclusion, we
recommend the comparative CT method over the standard curve method because it is

more economical in terms of both cost and labor. Although both aromatase activity and

l9

CYP19 mRNA were clearly detectable in testes of X. laevis, both aromatase enzyme
activity and C YPI 9 gene expression were very low. Also, no significant relationships
were found between aromatase enzyme activity and gene expression. This is likely due
the fact that the aromatase enzyme may have been dormant at the developmental stage

the frogs were in during the experiment.

Introduction
The cytochrome P450 enzyme aromatase is the key enzyme that catalyzes the conversion
Of androgens to estrogens and represents the rate-limiting step in estrogen biosynthesis.
The protein that catalyzes the aromatization of steroid hormones is encoded by the
CYP19 gene (Thompson and Siiteri, 1974; Simpson et al., 1994). Estrogens, especially
estradiol-170 (E2), have been shown to play a key role in ovarian development,
reproductive function and sexual differentiation in various amphibian species (Miyashita
et al., 2000; Miyata and Kubo, 2000; Kuntz et al., 2003a; Kato et al., 2004). Thus,
disruption Of either activity or production of this enzyme is likely to result in altered
developmental or reproductive biology of organisms. Due to its key function in estrogen
biosynthesis and associated reproductive processes, aromatase has been considered as an
important endpoint to assess the exposure to compounds that may interact with
reproductive endocrinology in vivo and in vitro (Sanderson et al., 2002; Hayes et al.,
2002; Rotchell and Ostrander, 2003).

Recently, concern was raised about the potential Of triazine herbicides to interact
with the endocrine system of male frogs by inducing aromatase resulting in an increase of

endogenous estrogen production and subsequently causing feminization or

20

demasculinization Of males (Hayes et al., 2002). Although studies by Sanderson et a1.
(2002) and Roberge et a1. (2004) have found that high concentrations of triazine
herbicides can induce aromatase in mammalian cells in culture, to date there have been
no reports of this mechanism Of action being observed in vivo in amphibians. This may
be due to the fact that testicular aromatase enzyme activities are often low and are thus
difficult to detect because they are near the detection limits of the commonly used
enzymatic assays (Hecker et al., 2004). Therefore, to increase our ability to determine
possible changes in aromatase activity in the testis, a more sensitive test system is needed
that allows for detecting even subtle changes. One way to examine the potential for such
subtle effects on the expression of aromatase activity is by measuring the changes in the
expression of C YP19 mRNA. Quantitative (real-time) reverse transcriptase polymerase
chain reaction (Q-RT PCR) is a sensitive and flexible technique that can detect small
quantities of mRNA in small amounts of tissue (Bustin, 2000 and 2002). This technique,
which amplifies the number of copies of mRNA many times, can theoretically measure as
little as a Single molecule of the target mRNA (Linz et al., 1990; Bej et al., 1991).

There have been few studies analyzing C YPI9 gene profiles in the African
clawed frog (Xenopus laevis) or in amphibians in general (Miyashita et al., 2000;
Akatsuka et al., 2004; Kuntz et al., 2004). None of above studies, however, have focused
on adult males and, to our knowledge, Q-RT PCR methods using reliable quantification
methods have not yet been applied to quantify the gene expression levels of CYP19 in
testes of X. laevis. It is known that CYP19 is differentially expressed based on the sex or
life-stage in most vertebrate species (Miyashita et al., 2000; Liu et al., 2004; Sakata et al.,

2005; F orlano and Bass, 2004) and that one cannot Simply extrapolate between sexes,

21

especially with regard to effects Of chemical exposure. Therefore, the objective of this
study was to develop and Optimize a Q-RT PCR procedure to measure the expression
level of CYP19 in testicular tissue of male X. laevis. To facilitate accurate quantification,
a cDNA standard was produced that could be used for the determination of absolute copy
numbers of CYP19 mRNA in addition to the relative quantification determined by
comparison to the expression of housekeeping genes. Furthermore, comparison of

C YP19 gene expression in male with aromatase enzyme activities was conducted to

establish a link between expression and function of gonadal aromatase in male X. laevis.

Materials and Methods

Animals

Adult male X. laevis, 30—50g, were purchased from XenOpus Express (Plant City, FL,
USA). Each frog was treated with 0.06% NaCl upon their arrival at the laboratory to
reduce the risk of possible infections. Frogs were acclimated for several weeks at the
Michigan State University‘s Aquatic Toxicology Laboratory before the experiment was
initiated. During acclimation, animals were held in 600-L fiberglass tanks under flow-
through conditions. The photoperiod was 12:12-h light/dark. Frogs were fed Nasco frog
brittle (Nasco, Fort Atkinson, WI, USA) three times per week ad libitum. Fourteen frogs
were under static renewal conditions individually in 40 L aquarium for 36 days, with
50% of water renewal every 3 days. Feeding regimen, temperature, and photoperiod
during the exposures were consistent with acclimatization conditions. All procedures
used during all phases of this study were in accordance with protocols approved by the

Michigan State University Instituted Animal Care and Use Committee (IACUC).

22

Isolation of total RNA and first-strand cDNA synthesis

Frogs were sampled on exposure day 36 and were anesthetized by immersion in 250 mg/l
MS-222 (tricaine methanesulfonate). Total RNA was isolated from testes of 14 male X.
laevis using the SV Total RNA Isolation System (Promega, Madison, WI, USA)
following the manufacturer's specifications with minor modifications to maximize the
efficiency of total RNA isolation. Briefly, tissues were homogenized using a Kontes
pestle and lysed in microcentrifuge tubes with guanidine thiocyanate and B-
mercaptoethanol mixture. After centrifugation to remove precipitated proteins and
cellular debris, nucleic acids were precipitated with ethanol and bound to a glass fiber
membrane. All samples were treated with RN ase-free DNase I at room temperature for
15min to remove the chromosomal DNA. RNA integrity was checked by denaturing
agarose gel electrophoresis (not shown) and 260:280nm absorbance ratio (2.33 i 1.03)
using a DU530 UV/VIS spectrOphotometer (Beckman Coulter, Inc., CA, USA).
Concentrations of total RNA were determined using the RiboGreenTM RNA quantitation
reagent (Molecular Probes, Inc., OR, USA) in a TD700 laboratory fluorometer (Turner
BioSystemS, Sunnyvale, CA, USA). Purified RNA was stored at —80°C until further
analysis.

A sample containing 500ng Of total RNA was used to synthesize single-strand
cDNA in accordance with the manufacturer's directions (SuperScriptTM First-Strand
Synthesis System for RT PCR, Invitrogen, CA, USA). Briefly, prior to reverse
transcription, total RNA was treated with DNAse I to remove potential chromosomal

DNA. Then, 1.25uL of 12—18 Oligo(dT) (0.5ug/uL) and lOmM dNTP mix were added

23

to the total RNA, and incubated at 65°C for 5min. The reaction was stopped by chilling
the test solution on ice. Reaction mixture (10X RT buffer, 25mM MgClg, 0.1M DTT and
recombinant ribonuclease inhibitor) was added to the RNA/primer mixture and incubated
at 42°C for 2min. SuperScript II reverse transcriptase (1.25uL of SOU M-MLV) was
added and the reaction mixture was incubated at 42°C for 50min, followed by a second
incubation at 70°C for 15min. TO confirm complete removal Of possible genomic
contamination, a negative control (sample without reverse transcriptase) was run in
parallel in the Q-RT PCR system, which resulted in no amplification of the PCR product
(data not shown). To improve sensitivity Of the PCR to amplify the CYP19 mRNA from
cDNA, the RNA template from the cDNAzRN A hybrid molecule was removed by
digestion with Escherichia coli RNase H (2U/uL) after first-strand cDNA synthesis took

place.

Real-time PCR using S YBR Green I

To determine the accumulation of the PCR product, SYBR Green I dye was used as a
real-time reporter of the presence of double-stranded DNA. The expression level of

C YPI9 mRNA was normalized to an internal control gene, glyceraldehyde-3-phosphate
dehydrogenase (GAPDH). Both cDNA sequences were obtained from the public
GenBank database of NCBI. The X. laevis CYP19 gene primer [forward primer:
5'CGGTTCCATATCGTTACTTCC3’, reverse primer:
5'GCATCTTCCTCTCAATGTCTG3', amplicon length (bp): 140] was designed in our
laboratory based on consideration of GC content, length, secondary structure and melting

temperature of the primer using the program Beacon Designer 2 (PREMIER Biosoft lntl.,

24

Palo Alto, CA, USA). Sequences for the GAPDH gene primer [forward primer: 5'GCT
CCT CTC GCAAAG GTC AT3’, reverse primer: 5'GGG CCA TCC ACT GTC TTC
TG3', amplicon length (bp): 101] was Obtained from the published literature (Wiechmann
and Smith, 2001). Primer specificity was verified by a single distinct peak Obtained
during the melting curve analysis of the SYBR Green-based RT PCR system and by
DNA sequencing of the PCR amplicons separated by gel electrophoresis. Best results
were obtained at a dilution of the reverse-transcribed samples of 1/4 and 1/20 for C YP] 9
and GAPDH, respectively. All PCR reactions were performed in a SmanCycleI® II
(Cephid, Sunnyvale, CA, USA). PCR master mix was prepared on ice with 10X SYBR
Green 1 buffer containing 3uL of MgClz (25mM/1.5mL), 0.5uL Of dNTP mix with dUTP
(12.5mM/ 1 mL), proper primers (sense primer/antisense primer, 9.8 pM/uLz7.3 pM/uL
for C YPI 9 and sense primer/antisense primer, 9.3 pM/uL:11.3 pM/uL for GAPDH), 0.65
units of AmpliTaq GoldTM DNA polymerase (5U/pL) and 0.25 units of AmpErase
(lU/uL). SpL of diluted reverse-transcribed samples were added to 20uL Of the PCR
master mix. The PCR reaction mix was denatured at 95°C for 10min before the first PCR
cycle. The thermal cycle profile was: (1) denaturation for 158 at 95°C, (2) annealing for
305 at 60°C and (3) extension for 303 at 72°C. A total of 50 PCR cycles was used for

amplification due to the low C YPI 9 copy numbers in many of the samples.

Synthesis ofplasmid DNA standards
PCR products of C YPI 9 and GAPDH were separately ligated into the pGEM T vector
(Promega, Madison, WI, USA) following manufacturer's specifications. Sequence

validity of the cloned amplicons was confirmed by automatic DNA sequencing and

25

followed by a BLAST2 analysis (National Center for Biotechnology Information, NCBI
(xx-=\\-’\.v.ricbi.nlm.nih.gov), Bethesda, MD, USA) with their corresponding sequences in
GenBank. The concentrations of purified plasmids (C YPI 9 plasmid DNA and GAPDH
plasmid DNA) that spanned the target regions for forward and reverse primers were
measured by using TD700 laboratory fluorometer (Turner design, CA, USA) with
molecular probes' RibOGreenTM DNA quantitation reagent (Molecular Probes, Inc., OR,
USA). These measured plasmids were converted to copy numbers/uL according to the
formula below (Eq. (1)):

Number of DNA molecules per uL = (ng/uL ° 1.515+Nbp) ° 6.023 - 10H (1)
where Nbp = size of dsDNA (plasmid size plus DNA insert size) expressed as bp.

To evaluate PCR efficiency, uniformity and linear dynamic range of each Q-RT
PCR assay, standard curves for C YPI 9 and GAPDH were constructed using serial

dilution of PCR product-inserted plasmid DNA standards (1 X10'—1X10° copies/uL).

Quantification ofC Y P] 9 mRNA expression

There are two methods that are commonly used for the analysis of data obtained from the
RT PCR system. These include relative measurements, where the change in expression
of the mRNA of interest is compared to that of an internal housekeeping gene that is
assumed to be unaffected by the study treatment(s) (comparative CT method). This
method does not require any standards and is generally sufficient to demonstrate changes
in gene expression. The more accurate method is to develop standards of either mRNA

or the appropriate cDNA so that a standard curve can be developed to which the results of

26

the PCR from a sample can be compared (absolute standard curve method). To assure

the accuracy of measurements, both methods were applied and the results compared.

Comparative CI method

 

This method in which the expression of the C YPI9 target gene (cDNA made from
mRNA) was normalized to that of GAPDH in each RT PCR reaction (referred to as CT) is
the most commonly used method. Differences between median ACT of test group and
ACT of each sample were expressed as AACT. The fold difference (2°ACT) of gene
expression in C YP19 was calculated for each sample. While this method is accurate and
generally gives reliable results, the absolute quantification method, which relies on a

standard curve for each gene, is more accurate.

Absolute standard curve method
In addition to the comparative method, an absolute method, based on standard curves
developed for each transcript was generated from a dilution series of synthesized plasmid
cDNA standards and a linear regression model was applied to quantify the data (Eq. (2)).
Y = aX + b (2)
where Y = CT value, a = the Slope Of the standard curve, X = logarithm of the total copy
numbers and b = y-intercept.
The amount of mRNA present in the original RNA extract was determined using
the Q—RT-PC method. Data were expressed as the CT value, which is the cycle number
when a reaction reaches the threshold (level of detection of increasing fluorescence)

(Girault et al., 2002). Determination of transcript abundance (mean of CT value) of the

27

C YPI9 and the GAPDH genes were conducted in triplicate. The copy numbers of C YPI9
and GAPDH cDNA were calculated (Eq. (2)). TO compensate for variations in RNA
amount and RT efficiency, the copy number of C YPI 9 was normalized to that Of the
internal gene (GAPDH). GAPDH was selected as the internal control (housekeeping
gene) because it has been reported to be expressed at lesser levels than other
housekeeping genes, such as [i actin and 18S rRNA (Wiechmann and Smith, 2001).
GAPDH is a consistently expressed gene, making it suitable as an internal standard for Q
RT PCR assays (Raaijmakers et al., 2002). Expression ratio (ER) Of mRNA copy
numbers between C YPI9 and GAPDH in the same sample was also calculated (Eq. (3)).
ER = mRNA copy number of C YPI 9/mRNA copy number of GAPDH (3)

Quantitative (real-time) RT PCR efficiencies were calculated as follows (Eq. (4)).

Efficiency (%) = [[IOH/ a)1-1] - 100 (4)

where a is the Slope of the standard curve derived from Eq. (3).

C onfirmation oftest system using positive controls

To confirm the validity of the developed methods, testicular and ovarian tissues from
adult X. laevis were exposed to a model compound, forskolin (Sigma-Aldrich, St. Louis,
MO), that is known to induce C YPI 9 ovarian gene expression (Watanabe and Nakjin,
2004). Briefly, ovarian and testicular tissues were harvested and plated in Medium 199
(Hepes supplemented with 0.1mM IBMX and 1 ug/mL 25-hydroxycholesterol) in 24-well
plates (Corning, NY, USA) (testis: approx. 0.1 g/well, ovary: approx. 0.5 g/well). Prior
to transferring tissue from male frogs to plates, each testis was dissected into eight pieces

of equal size. Testicular fragments from all animals were then combined and four pieces

28

were randomly assigned to each well to minimize variation of C YPI 9 gene expression
due to inter-individual differences. Exposure concentrations were 0 and lOOuM forskolin
using DMSO as solvent carrier. A solvent control was run in the forskolin experiment to
test for possible effects of DMSO on CYP19 gene expression. Experiments were
conducted over a time period of 20h at 25°C. After exposure, CYP19 gene expression
was measured in tissue using the methods described above. Due to limitations in the
amount of tissue available, no measurements of aromatase activity could be conducted in

parallel.

C YPI 9 aromatase activity

Aromatase activity was measured following the protocol of Lephart and. Simpson (1991)
with minor modifications. Less than 0.5g of gonadal tissue was homogenized in 600uL
of ice-cold gonad buffer (50mM KPO4, ImM EDTA, lOmM glucose-6-phosphate, pH
7.4). The homogenate was incubated with 300nM 3H-androst-4-ene-3, l 7-dione
(25.9Ci/nmol; Lot NO. 3467-067; Cat. NO. NET—926; New England Nuclear, MA, USA),
0.5 IU/mL glucose-6-phosphate (Sigma Cat. # G6378) and ImM NADP (Sigma Cat. # N-
0505) at 37°C and 5% CO; for 90min. Tritiated water released from each sample was
extracted and activity determined by liquid scintillation counting. Aromatase activity
was expressed as pmol androstenedione converted/h/mg protein. The specificity of the
reaction for the substrate was determined by use of a competitive test with non-labeled
androstenedione and the use of the specific aromatase inhibitor fadrozole (Novartis
Pharma AG, Basel, CH). Addition of large amounts of androstenedione reduced tritiated

water formation to the concentrations found in the tissue blanks. Furthermore, addition

29

of fadrozole during the tritium-release assay reduced aromatase enzyme activity in a
dose-dependent manner with concentrations of SuM and greater resulting in complete
inhibition of enzyme activity to the levels measured in the blanks. This demonstrated
that the activity being measured was specific for aromatase. Protein concentrations were
determined using the Bradford assay (Bradford, 1976) with bovine serum albumin as the

protein standard (Sigma-Aldrich, St. Louis, MO, USA).

Statistical analysis

Statistical analyses in this study were conducted using SYSTAT 10 (SPSS Inc., Chicago,
IL, USA). Data sets were tested for normality using Kolmogorov—Smimov's one sample
test. The Pearson correlation analysis was used to evaluate the relationship between
CYP19 enzyme activity and C YPI 9 mRNA expression, and a linear regression model
was used to quantitatively determine relationships among gene quantification methods in
the Q-RT PCR system. The Student's t-test was used to examine differences in gene
expression between C YPI 9 and GAPDH. The criterion for significance in all statistical

tests was p < 0.05.

Results

RTPCR amplification efficiencies, linearity and reproducibility

Specificity of the PCR reaction, accuracy of mRNA quantification and sensitivity and
linearity of SYBR Green based Q-RT PCR for C YPI 9 and GAPDH in adult male X.
laevis were determined. Real-time PCR amplification curves for the two genes obtained

with the SmartCycler‘fi’ were very reproducible and indicated that primers were selective

30

and effective in producing the Specific PCR products (Figs. HA and 1.2A). The melting
curves (Figs. 1.1C and 1.2C) generated at the end Of the PCR reaction Show that all
amplicons of the C YPl9/GAPDH plasmid DNA standard had the same melting
temperature (81°C). This result indicates that no primer—dimers were formed during the
reactions (Figs. 1.1C and 1.2C). To further validate the specificity ofthe assay, gel
electrophoresis (1.5% agarose) was performed on the PCR products Obtained from
serially diluted plasmid DNA standards (Figs. MD and 1.2D). The results from the gel
electrophoreses demonstrate that the amplification was specific for the ~ 140bp and ~

10 lbp products of CYP19 and GAPDH, respectively.

The accuracy of mRNA quantification, and sensitivity and linearity of SYBR
Green-based Q-RT PCR were examined using a lO-fold serial dilution of each plasmid
DNA standard. Efficiencies during the exponential phase were 96.7% and 105.7% for
CYP19 and GAPDH, respectively. The relationship between threshold cycle (CT) and the
log copy number of plasmid DNA standard was linear with r2 > 0.99 for both genes,
indicating that the CT values changed proportionally with serial dilution of the samples.
The reproducibility of the techniques within and between assays was tested, using serial
dilutions of CYP19 and GAPDH plasmid cDNA standards. Intraassay variabilities were
assessed by evaluating the coefficient of variation (CV) for three replicates in each
dilution within one run (Table 1.1). Interassay variabilities were assessed by conducting
three different assays performed in triplicate of each dilution over a period of 3 days
(Table 1.1). Intraassay CVs of CT for both genes were very small (< 1.2%), indicating

that the assays were highly reproducible for determining expression Of both genes.

31

Although greater than intraassay CVs, interassay CT values were also small with CVS <

5.4% for both genes.

   

 

 

 

 

(A)
... (B)
GAPDH :..
y = - 3.2086x + 32.128
8 40 r2 = 0.998
C
0)
g 30 (I
:3 0°" ‘
= 1 0 s.
E: 20 0
,-’ --- o
‘ C
10610510410310210l 1o
10 20 30 40
(C) Cycles 0 2 4 6
Log copy number
C
8
3
8
3
E
Ill“ 10 3 Ill2 101

 

 

 

60 70 80 90
Degree, C

Figure 1.1 GAPDH plasmid DNA Standard curve. (A) Amplification curves OfSix
dilutions of GAPDH plasmid DNA standard from l>< 101 to l><106 copies/uL. (B) GAPDH
plasmid DNA standard curve plotting the log copies/ILL (x) of GAPDH plasmid DNA
against CT (y), the equation was calculated by linear regression analysis (r2=0.998 and
105.7% OfPCR efficiency). (C) Melting curve OfPCR products, showing specificity of
the reaction. (D) 1.5% agarose gel electrophoresis ofthe PCR products in the serially
diluted samples.

32

Fluorescence

Fluorescence

(A)

 

CYPl9
',/I l'
-.».'/..’ .44.-
10610510410310210

 

 

  

 

 

4o

 

10 20 30
(C) Cycles
CY P l 9
60 70 so 90
Degree, C

CT

 

 

(B)
40 y = - 3.4024x + 32.31
r2 = 0.994
30’ 0.,
.0,
'0~-.
20 - c _,
.0 ..
,0
10'
0 L l L
2 4 6
Log copy number

(D)

 

 

Figure 1.2 CYPI9 plasmid DNA standard curve. (A) Amplification curves of six dilutions
of C1’PI9 plasmid DNA standard from 1 x 10' to 1X 10‘) copies/uL. (B) C1’P19 plasmid
DNA standard curve plotting the log copies/UL (x) of CYPI9 plasmid DNA against CT
(y), the equation was calculated by linear regression analysis (r2=0.994 and 96.7% of
PCR efficiency). (C) Melting curve OfPCR products, Showing specificity ofthe reaction.
(D) 1.5% agarose gel electrophoresis ofthe PCR products in the serially diluted samples.

33

Table 1.1 Reproducibility and precision of standard curve method for C YPI9 and

 

 

 

 

 

GAPDH plasmid DNA.
lntra assay a Inter assay b
C 11)::O1g:::SI/1:E)DNA CT mean values C SD d CV e C;;::::” SD CV
1 x 106 12.37 0.04 0.31 12.57 0.23 1.79
1 x 105 15.18 0.06 0.36 15.41 0.32 2.08
1 x 104 18.56 0.16 0.84 18.62 0.15 0.83
1 x 1()3 21.53 0.11 0.50 21.84 1.16 5.33
1 x 102 25.18 0.11 0.44 25.65 0.72 2.80
1 x 1()1 29.60 0.03 0.11 28.77 0.83 2.90
CAPlzggiISSS/IESDNA CT mean values SD CV CL$::n SD CV
1 x 106 13.11 0.12 0.95 13.20 0.08 0.61
1x105 16.02 0.18 1.12 16.03 0.10 0.64
1 x 104 19.23 0.10 0.54 19.32 0.10 0.50
1 x 103 22.22 0.11. 0.49 22.60 0.47 2.06
1 x 102 25.52 0.04 0.15 25.81 0.35 1.34
1 x 10 29.28 0.05 0.16 29.73 1.02 3.43

 

a; intra assay was assessed by evaluating the coefficient Of variation (CV) for each dilution ofthe plasmid
using three replicates within run

b; inter assay was assessed by evaluating the coefficient of variation (CV) for each dilution ofthe plasmid
using three assays with three replicates over 3 different days

0; average of number Of cycles when fluorescence crossed threshold

(1; SD = standard deviation from the mean

e; CV = coefficient of variation (%)

Comparison of different quantification methods/Or C YPl 9 gene expression

Serial dilutions (1X10l to 1>< 106 copies/uL) of C YPI 9 and GAPDH plasmid DNA
standards were used to quantify gene amplification rates for the genes of interest. The
results demonstrated that the SYBR Green-based Q-RT PCR assay allowed for the

quantification of small amounts of CYPI9 mRNA (10 copies/reaction) in all 14 adult

34

male X. laevis. Initial copy numbers for both genes in all 14 samples were determined by
use of the standard curve method. GAPDH exhibited significantly greater abundances of
the transcript with a mean CT value Of 22.9 i 0.62 (mean :1: SD.) than CYPI9 with a
mean CT value Of 28.1 i 1.4 (p < 0.001). Mean copy numbers for all samples were 25.2
i 23.4 copies/uL and 802.61 i 350.9 copies/11L for CYPI9 and GAPDH, respectively.
Because of a number of factors such as varying amounts of mRNA in the
samples, differences in reverse transcription efficiency and potential presence Of PCR
reaction inhibitors can influence the gene amplification reaction, the use Of an internal
control is necessary to normalize the measurements. The simplest way to quantify
mRNA in RT PCR systems, the use of the CT value ratio (CT of target gene/CT of internal
gene), was also applied to quantify CYP19 gene expression (Table 1.2). The similar
efficiencies Observed for the two genes in this PCR assay allow for the use Of the
comparative CT method for quantifying C YP19 gene expression after normalization to
gene expression of the internal gene. The fold differences (2%") of C YPI 9 gene
expression of all 14 samples were calculated using the comparative CT method. The
average fold difference was not equal, but very close to 1.0. In addition to the
calculations above, C YPI 9 gene expression was measured using the standard curve
method, where the expression was determined as copy numbers obtained from C YP19
plasmid standard curve or as ER (Eq. (3)) normalized to the internal control (Table 1.2).
All four quantification methods were compared to each other using a linear
regression (r2) model to determine the compatibility of different quantification
approaches (Fig. 1.3). The comparative CT method and the standard curve method were

the most highly correlated in all comparisons (r2 = 0.997, p < 0.001). The relationship

35

between CT value ratio and comparative CT method or CT value ratio and standard curve
method was less strong, with r2 = 0.902 and r2 = 0.916, respectively. The coefficient for
the correlation between the uncorrected C YPI 9 copy number and the results from the
comparative CT method was the lowest overall (r2 = 0.608, p = 0.001).

Table 1.2 Diverse gene expression quantification methods and aromatase enzyme
activities in individual male X. laevis.

 

 

Replicate CT ratio a 2AACT b ER C err/9 copy (2:33:52? Elem-[y
, g proteln)
1 1.260 0.443 0.014 9.961 3.095
2 1.293 0.267 0.008 6.301 5.063
3 1.259 0.442 0.014 9.436 2.760
4 1.194 1.057 0.034 9.184 4.886
5 1.207 1.045 0.031 24.781 9.454
6 1.211 1.000 0.030 24.669 3.210
7 1.260 0.421 0.013 7.565 1.779
8 1.200 1.214 0.036 34.603 21.144
9 1.195 1.437 0.041 63.769 13.057
10 1.313 0.182 0.006 3.497 11.907
11 1.164 1.950 0.059 30.842 8.621
12 1.168 1.807 0.055 27.991 19.842
13 1.255 0.526 0.016 16.070 15.669
14 1.171 2.042 0.058 84.733 9.362

 

a; CT value of CYPI9 / CT value of GAPDH

b; comparative CT method

c; expression ratio calculated using standard curve method
(1: number of mRN A copies (Standard curve method)

36

(A) (B)

007- 007-
0.051' 0.06 ’ 0.0
0.051 a: 005’

1: 004L 111 004

  
 

0-03 ' y - 0.029311 + 0.0007 0'03 C
001 L 0.01

y - 0374311 + 0.4882
r2-0.916

O

 

 

 

 

1.16 1.2 1.26 1.3 1.36

2“ CT CT ratio

1.32 - . y--0.0712x+12954
13' . t2-0302

CT ratio
a

 

 

 

Figure 1.3 Comparisons among quantification methods for measuring Cl?! 9 mRNA
expression used in the Q-RT PCR system. ER and 2AM" were calculated from Standard
curve method and comparative CT method, respectively. CT ratio represents the ratio of
Cr value of CYPI9 to CT value of GAPDH. (A) Represents comparison of CYP19 gene
expression from standard curve method to that from comparative C-r method. (B)
Represents comparison Of (.‘1'1’19 gene expression from Standard curve method to that
from CT ratio of Cl’l’l 9/GAI’DH. (C) Represents comparison of C'Yl’l9 gene expression
from C '1‘ ratio of CYP19/GAPDH to that from comparative C '1- method.

Comparison o/‘CYPI 9 gene expression and aromatase activity
Aromatase activity was measurable in all frog testes analyzed with activities ranging
from 1.78 to 21.14 fmol/h/mg protein. Variability among individuals was relatively great

with a CV of 69%. This variability was Similar to those observed for changes in gene

expression: 63% for the standard curve method and 64% for the comparative C1 method.

37

However, when comparing aromatase enzyme activities with C YPI 9 gene expression
determined by either the comparative CT method or the Standard curve method in the
same frogs, no Significant correlations could be Observed (r = 0.404, p = 0.152; r = 0.399,

p = 0.158, respectively) (Table 1.3).

Table 1.3 Pearson correlation coefficients (r) and probabilities (p) between the different
parameters measured.

 

 

C1'Pl9 cepy ER ZAACT CT ratio Aggilil‘trlil‘t'cfe
CYPI 9 copy 1
ER 0.743 (0.002) 1
2°ACT 0.780 (0.001 ) 0.998 (<0.000) 1
CT .3110 -0.669 (0.009) -0957 (<0.000) 41.950 (<0.000) 1
mix?“ 0.339 (0.236) 0.399 (0.158) 0.404 (0.152) -0334 (0.243) 1

 

Bold numbers indicate significant correlation. Negative numbers indicated negative relationships.

Refer to Table 1.2 for explanations

Gonadal C Y Pl 9 gene expression after exposure to forskolin

Exposure of gonadal tissues to forskolin resulted in an increase of C YPI 9 gene
expression in both ovarian and testicular explants (Fig. 1.4). The greatest induction was
Observed in ovarian tissue with a 3.74-fold induction of C YPI 9 gene expression
compared to the solvent controls. In testicular tissue, C YPI 9 mRNA copy numbers were
increased 2.62-fold. The above results were achieved using the standard curve method.
However, similar patterns were observed when applying other quantification methods

such as C T ratio method (data not Shown).

38

 

 

 

(A)
U
m 3.2 ..
,9.
9)
E 2.4 ..
.2
2
c) 1.6 "
S
DD
(6
f, 0.8x
><
0 1 1 1
5 (B)
U
m 4
8
E 3 --
E
8
U 2
C
01)
‘° 1 ..
‘5.)
X
0 - 5 .‘ - 1
Blank SC lOOmM

Figure 1.4 Fold-change (x-change. mean i SD.) of(.'YPI9 mRNA in testicular (A) and
ovarian (B) explants oernopus laevis after exposure to 100 11M forskolin (100 1.1M) for
2011, using the Standard curve method for quantification of mRNA. SC :- solvent control
(0.1% DMSO).

Discussion

Development and optimization on-R T PCR system to quantify C YPI 9 gene expression
in male X. laevis

The conditions of SYBR Green-Q-RT PCR analysis for detecting C YPI 9 mRNA in testes
from male X. laevis were established and optimized. The two-step Q-RT PC R method

was selected over the one-step method due to its higher sensitivity, lesser risk of primer—

3‘?

dimer formation during PCR reaction and lesser risk of contamination with genomic
DNA (Vandesompele etal., 2002). It was possible to detect small quantities of C YPI9
mRNA (as few as 10 copies/reaction) in gonadal tissue (< 100mg) without prior cDNA
amplification or a nested PCR approach, which requires a secondary amplification of the
target gene using the PCR product from an initial gene amplification to improve
sensitivity and specificity.

SYBR Green was chosen for the detection Of amplicons during the PCR reaction
because it is relatively inexpensive while its sensitivity, reproducibility and dynamic
range were comparable to that of the fluorescent probe method (Lekanne Deprez et al.,
2002). The melting curve analyses revealed that the obtained signal for both C YPI 9 and
housekeeping gene were specific, and did not result in the amplification of unwanted
gene products. No primer—dimers were formed. The SYBR Green dye detection system
proved to be highly sensitive with a method detection limit of as few as 10 copies of the
target gene per reaction. The routine treatment of RNA samples with DNase I minimized
co-amplification Of pseudo-genes, which are genetically similar to the original gene but
are not expressed, or non-specific DNA which the primer may have found (Kreuzer et al.,
1999).

Quantitative analysis of gene expression is often achieved by normalization to
the amplification of housekeeping genes as internal controls. Ideally, the internal control
gene Should be expressed at a constant level among different cell populations and
individuals and Should be unaffected by experimental conditions (Thellin etal., 1999).
GAPDH is a gene that has these characteristics, which make it a usefill and effective

housekeeping gene to control for these types of variations (Wiechmann and Smith, 2001).

40

The use of the GAPDH as an internal control provides more accurate results since it not
only compensates for sample-to-sample variations but also circumvents technical
problems such as total RNA extraction efficiency and reverse transcription efficiency.
However, there are studies that suggested that, in some cases, GAPDH might not be
appropriate as an internal control for every RT PCR system. Some mammalian species
showed unstable gene expression of GAPDH during the cell cycle (Mansur et al., 1993)
and during different developmental stages (Calvo et al., 1997). A different study with
humans found that GAPDH mRN A transcription levels can also vary widely among
individuals (Bustin et al., 1999). In contrast, in our study, little variation in the
expression of GAPDH was Observed among individuals. This observation indicates that
GAPDH is a Suitable housekeeping gene for determining changes in CYP19 expression in
testes of X. laevis that are of similar developmental stage. However, this study was not
designed to address effects of different developmental stages on the expression of
GAPDH and, therefore, when conducting a developmental study the appropriateness of
the GAPDH as a housekeeping gene would need to be further validated.

In order to obtain accurate and reproducible results, the PCR reaction should
have efficiency as close to 100% as possible. At this efficiency, the template doubles
after each cycle. Efficiencies of the PCR reactions were very close to the desired
efficiency of 100% for both C YPI 9 and GAPDH, indicating that the increase in gene
expression is directly proportional to the number amplification cycles. Furthermore, the
small interassay variabilities among experiments conducted on 3 different days and the
low CVS for the calculated CT values for all experiments demonstrate the reproducibility

and precision of the established test system.

41

In conclusion, the Q-RT PCR method developed to quantify CYP19 gene
expression in male X. laevis in this study is sufficiently sensitive to allow the
measurement of Single digit copies of total RNA. This sensitive and precise assay is a
useful tool that allows for quantifying specific types of mRNA that are expressed at low
levels in certain tissues such as C YPI 9 in testes of male frogs and that allows for direct

comparison Of gene expression levels between samples.

Comparison of dijjerent gene quantification methods

In this study, four quantification methods were applied to quantify C YPI 9 gene
expression and were then compared to identify the optimal method for quantification. In
the first method, CYP19 mRNA copy numbers were calculated from the absolute
standard curve obtained by serial 10-fold dilutions of a cloned plasmid standard without
referring to the housekeeping gene. This method allowed estimation of the number of
copies of C YPI9 mRNA present in the unknown samples. However, estimates of copy
numbers of C YPI9 mRNA calculated from the linear equation derived from the absolute
standard curve method did not appear to give an accurate estimate of the actual
expression of C YPI 9 mRNA molecules present in the sample. The inaccuracy of the
estimate was indicated by the low correlation of these copy numbers with the
housekeeping gene-corrected copy numbers or the calculated ratio from the comparative
CT method. This correlation was improved once the copy numbers were normalized to
the internal control (expressed as ER). This demonstrates that the use of an internal
control such as GAPDH is critical to accurately quantify CYP19 gene expression profile

in male X. laevis when using Q-RT PCR. The Si gnificant correlations among all three

42

quantification methods using GAPDH as the housekeeping gene demonstrate the
applicability of all of these methods to quantify C YPI 9 gene expression in X. laevis testes.
However, compared to the very strong relationship between the standard curve and the
comparative CT method (r2 = 0.997, p < 0.001), the CT value ratio was less predictive for
the standard curve method (r2 = 0.916, p < 0.001) or the comparative C T method (r2 =
0.902, p < 0.001 ). Thus, we conclude that both the comparative CT method and the
standard curve method are optimal quantification methods to estimate low levels Of
C .YPI9 gene expression in testicular tissue of X. laevis. There have been few studies
using Q-RT PCR to measure aromatase mRNA expression in male African clawed frogs
(Miyashita et al., 2000; Kuntz et al., 2004). These studies reported gene expression of
C YP/ 9 without normalization (Miyashita et al., 2000), or simply by the ratio of C YP] 9/5/'
I (Kuntz et al., 2004) and, to date, and to the best of our knowledge, no study has been
conducted to measure C YPI 9 mRNA level in male X. laevis using more accurate RT
PCR quantification methods. The results from our study confirm that the simple ratio
between housekeeping and C YPI9 gene is not as accurate and sensitive as more
sophisticated methods such as the standard curve or comparative CT method.

The advantage of the comparative CT method over the absolute standard curve
method is that this method eliminates the need to construct a standard curve, which is a
time consuming and laborious process, allowing Simple quantification of the relative gene
expression of paired samples. Therefore, use of the economical and efficient comparative
CT method is recommended as the preferable method to quantify C YPI 9 gene expression

in testicular tissue of X. laevis.

43

Comparison of C YPl 9 gene expression with aromatase enzyme activity

While mRNA quantification of C YPI 9 provides important information on the regulation
of protein synthesis, it may not directly reflect aromatase enzyme activity due to
posttranscriptional control of enzyme activity. An earlier study reported that differences
in C YPI9 gene expression between males and females were not proportional to aromatase
enzyme activity in another amphibian Species, the newt, Pleurodeles waltl (Kuntz et al.,
2004). These authors hypothesized that this lack in correlation might be due to
differences in the posttranscriptional regulation of aromatase. Posttranscriptional factors
that can influence the net activity of the enzyme aromatase can be either due to
modifications of the mRNA that lead to differential translation within a tissue or can be
due to posttranslational modifications that alter the stability of functionality of the protein
(Balthazart, et al., 2001; Genissel, et al., 2001).

Even though there is evidence that estrogens, which are catalyzed by aromatase,
play a stimulatory role in germ cell development including spermatogonial division, germ
cell viability and differentiation, acrosome biogenesis and function of the Spermatozoa in
rodents (O'Donnell et al., 2001), at present little is known of aromatase expression and
the role of estrogens in the testis of amphibians. It appears that estrogens are involved in
multiple actions of male reproductive system of amphibians during certain developmental
stages (Fasano et al., 1989; Cobellis et al., 2002). The fact that aromatase enzyme
activities in our study were very low and not correlated with C YPI9 gene expression
indicates that this enzyme may have been dormant or at basal levels in non-active
breeding conditions of frogs applied in this research. However, the confirmation

experiments using an inducer of aromatase, forskolin, have demonstrated that CYP19

44

gene expression can be modulated (increased) both in ovarian and testicular tissue of
Xenopus, indicating that stimulation of the enzyme results in a specific response at the
gene expression level. This result indicates that the established Q-RT PCR system
represents a valid method to determine alterations in the expression of C I’Pl 9 in male

testis.

Implicationsfor toxicological assessment of environmental pollutants

Aromatase regulation and activity play a pivotal role in sexual development and in
communicating reproductive processes in vertebrates. While in ovarian tissues the
formation of estrogens from androgens via the enzyme aromatase is an essential process
for gonadal maturation in males both expression and activity of aromatase are low in the
testis during the maturation phase, which mainly depends on androgens. Accurate
transcriptional regulation of the genes encoding steroidogenic enzymes such as aromatase
is critical for the regulation of sex steroid homeostasis that is essential for ordinary sexual
development processes in animals (Yamada et al., 1995). Thus, improper and untimely
changes in C YP19 gene expression may affect reproductive success in animals (Trant et
al., 2001; Kuntz eta1., 2003b). Therefore, the quantitative analysis of C YPI 9 mRNA
expression can be an important marker for detection of developmental and reproductive
disruption by EDCS in animals of both sexes. In fact, a series of chemicals have been
reported to have the potential to directly or indirectly disturb steroidogenesis simply by
interfering with the regulation of C YPI 9 gene expression either in vivo or in vitro
(Connor et al., 1996; Sanderson et al., 2000; Miyata and Kubo, 2000; Kazeto et al., 2004).

In the recent controversy about possible effects of pesticides and/or other environmental

45

contaminants on reproduction and development in amphibian species, it was
hypothesized that abnormal sexual development such as compromised reproductive
functions and/or characteristics may be due to the induction of aromatase by these
chemicals causing a decrease of endogenous androgens in males (Hayes et al., 2002).
However, although a series of studies has been conducted to identify effects of the
exposure to triazine herbicides on aromatase activity or C YPl 9 gene expression in fish or
amphibians (Hayes et al., 2002; Kazeto et al., 2004; Lavado et al., 2004; Hecker et al.,
2004), it has proved to be difficult to establish a direct link between exposure to these
chemicals and changes in gonadal aromatase. As outlined previously, this is likely due to
the fact that aromatase enzyme activities are low in adult testicular tissue, often being
below or just above the method detection limits of enzymatic assays. The Q-RT PCR
technique established in this study represents a method that can help to overcome this
difficulty as it is capable of identifying very small amounts of C YP19 mRNA and has
been successfully used to determine gene expression in the testis of X. laevis.

In conclusion, the Q-RT PCR system established and optimized in this study
represents a highly sensitive, rapid and reliable method to detect and measure very small
quantities of CYPI9 mRNA in small amounts oftissue. Although CYPI9 mRNA
expression does not seem to directly reflect aromatase enzyme activity in testicular tissue,
the developed Q-RT PCR method is a powerful tool due to determine changes in the
regulation of protein synthesis of aromatase that will be helpful in researching general
regulatory mechanisms in the reproductive endocrinology of X. laevis. Furthermore, this
method can be used as a highly sensitive marker in toxicological studies to identify

effects of environmental contaminants at the pretranslational level of aromatase.

46

Currently, a parallel study is underway that uses this Q-RT PCR method to determine the

effects of atrazine on testicular aromatase in X. laevis.

Acknowledgement

We thank A. Hosmer for many helpful comments on experimental design. We also thank
C. Bens, R. Bruce and S. Williamson. This research was facilitated by the Atrazine
Endocrine Ecological Risk Assessment Panel, Ecorisk, Inc., Femdale, WA and sponsored

by Syngenta Crop Protection, Inc.

47

References

Akatsuka, N., Kobayashi, H., Watanabe, E., lino, T., Miyashita, K., Miyata, S., 2004.
Analysis of genes related to expression of aromatase and estradiol regulated

genes during sex differentiation in Xenopus embryos. Gen. Comp. Endocrinol.
136, 382—388.

Balthazart, J., Baillien, M., Ball, GE, 2001. Phosphorylation processes mediate rapid
changes of brain aromatase activity. J. Steroid Biochem. Mol. Biol. 79, 261—277.

Bej, A.K., Mahbubani, M.H., Atlas, RM., 1991. Amplification of nucleic acids by
polymerase chain reaction (PCR) and other methods and their applications. Crit.
Rev. Biochem. Mol. Biol. 26, 301—334.

Bradford, M., 1976. A rapid and sensitive method for quantitation of microgram
quantities of protein utilizing the principle of protein-dye binding. Anal.
Biochem. 72, 248—254.

Bustin, S.A., 2000. Absolute quantification of mRNA using real-time reverse
transcription polymerase chain reaction assays. J. Mol. Endocrinol. 25, 169—193.

Bustin, S.A., 2002. Quantification of mRNA using real-time reverse transcription PCR
(RT-PCR): trends and problems. J. Mol. Endocrinol. 29.

Bustin, S.A., Gyselman, V.G., Williams, N.S., Dorudi, S., 1999. Detection of
cytokeratins 19/20 and guanylyl cyclase C in peripheral blood of colorectal
cancer patients. Br. J. Cancer 79, 1813-1820.

Calvo, E.L., Boucher, C., Coulombe, Z., Morisset, J ., 1997. Pancreatic GAPDH gene
expression during ontogeny and acute pancreatitis induced by caerulein.
Biochem. Biophys. Res. Commun. 235, 636—640.

Cobellis, G., Meccariello, R., Fienga, G., Pierantoni, R., Fasano, S., 2002. Cytoplasmic
and nuclear Fos protein forms regulate resumption of spermatogenesis in the frog,
Rana esculenta. Endocrinology 143, 163—170.

48

Connor, K., Howell, J., Chen, 1., Liu, H., Berhane, K., Sciarretta, C., Safe, S.,
Zacharewski, T., 1996. Failure of chloro-S-triazine-derived compounds to induce
estrogen receptor-mediated responses in vivo and in vitro. F undam. Appl.
Toxicol. 30, 93—101.

Fasano, S., Minnucci, S., DiMatteo, L., D'Antonio, M., Pierantonio, R., 1989.
Intraatesticular feedback mechanisms in the regulation of steroid profiles in the
frog, Rana esculenta. Gen. Comp. Endocrinol. 75, 335—342.

Forlano, P.M., Bass, AH, 2004. Seasonal plasticity of brain aromatase mRNA
expression in Gila: divergence across sex and vocal phenotype. J. Neurobiol. 65,
37—49.

Genissel, C., Levallet, J ., Carreau, S., 2001. Regulation of cytochrome P450 aromatase
gene expression in adult rat Leydig cells: comparison with estradiol production. J.
Endocrinol. 168, 95—105.

Girault, 1., Lerebours, F ., Tozlu, S., Spyratos, F., Tubiana-Hulin, M., Lidereau, R., Bieche,
I., 2002. Real-time reverse transcription PCR assay of C YPI 9 expression:
application to a well-defined series of post-menopausal breast carcinomas. J.
Steroid Biochem. Mol. Biol. 82, 323—332.

Hayes, T.B., Collins, A., Lee, M., Mendoza, M., Noriega, N., Stuart, A.A., Vonk, A.,
2002. Hermaphroditic, demasculinized frogs after exposure to the herbicide
atrazine at low ecologically relevant doses. Proc. Natl. Acad. Sci. U. S. A. 99,
5476—5480.

Hecker, M., Giesy, J.P., Jones, P.D., Jooste, A.M., Carr, J.A., Solomon, K.R., Smith, E.E.,
Van der Kraak, G., Kendall, R.J., Du Preez, L., 2004. Plasma sex steroid
concentrations and gonadal aromatase activities in African clawed frogs
(Xenopus laevis) from South Africa. Environ. Toxicol. Chem. 23, 1996—2007.

Kato, T., Matsui, K., Takase, M., Kobayashi, M., Nakamura, M., 2004. Expression of
P450 aromatase protein in developing and in sex-reversed gonads of the XX/XY
type of the frog Rana rugosa. Gen. Comp. Endocrinol. 137, 227—236.

49

Kazeto, Y., Place, A.R., Trant, J .M., 2004. Effects of endocrine disrupting chemicals on
the expression of CYP19 genes in zebrafish (Danio rerio) juveniles. Aquat.
Toxicol. 69, 25—34.

Kreuzer, K.A., Lass, U., Landt, O., Nitsche, A., Laser, J ., Ellerbrok, H., Pauli, G., Huhn,
D., Schmidt, CA, 1999. Highly sensitive and specific fluorescence reverse
transcription—PCR assay for the pseudogene-free detection of beta-actin
transcripts as quantitative reference. Clin. Chem. 45, 297—300.

Kuntz, S., Chardard, D., Chesnel, A., Grillier-Vuissoz, 1., F lament, S., 2003a. Steroids,
aromatase and sex differentiation Of the newt Pleurodeles waltl. Cytogenet.
Genome Res. 101, 283—288.

Kuntz, S., Chesnel, A., Duterque—Coquillaud, M., Grillier-Vuissoz, I., Callier, M.,
Doumon, C., Flament, S., Chardard, D., 2003b. Differential expression of P450
aromatase during gonadal sex differentiation and sex reversal of the newt
Pleurodeles waltl. J. Steroid Biochem. MOI. Biol. 84, 89—100.

Kuntz, S., Chardard, D., Chesnel, A., Ducatez, M., Callier, M., F lament, S., 2004.
Expression of aromatase and steroidogenic factor 1 in the lung of the urodele
amphibian Pleurodeles waltl. Endocrinology 145, 3111—3114.

Lavado, R., Thibaut, R., Raldua, D., Martin, R., Porte, C., 2004. First evidence of
endocrine disruption in feral carp from the Ebro River. Toxicol. Appl. Pharrnacol.
196, 247—257.

Lekanne Deprez, R.H., Fijnvandraat, A.C., Ruijter, J .M., Moorman, AF, 2002.
Sensitivity and accuracy of quantitative real-time polymerase chain reaction
using SYBR green I depends on cDNA synthesis conditions. Anal. Biochem. 307,
63—69.

Lephart, E.D., Simpson, ER, 1991. Assay of aromatase-activity. Methods Enzymol. 206,
477—483.

Linz, U., Delling, U., Rubsamenwaigmann, H., 1990. Systematic studies on parameters
influencing the performance of the polymerase chain-reaction. J. Clin. Chem.
Clin. Biochem. 28, 5—13.

50

Liu, X., Liang, B., Zhang, S., 2004. Sequence and expression of cytochrome P450
aromatase and FTZ-F 1 genes in the protandrous black porgy (A canthopagrus
schlegeli). Gen. Comp. Endocrinol. 138, 247—254.

Mansur, N.R., Meyer-Siegler, K., Wurzer, J.C., Sirover, M.A., 1993. Cell cycle
regulation of the g1yceraldehyde-3-phosphate dehydrogenase/uracil DNA
glycosylase gene in normal human cells. Nucleic Acids Res. 21, 993—998.

Miyashita, K., Shimizu, N., Osanai, S., Miyata, S., 2000. Sequence analysis and
expression of the P450 aromatase and estrogen receptor genes in the Xenopus
ovary. J. Steroid Biochem. Mol. Biol. 75, 101—107.

Miyata, S., Kubo, T., 2000. In vitro effects of estradiol and aromatase inhibitor treatment
on sex differentiation in Xenopus laevis gonads. Gen. Comp. Endocrinol. 1 19,
105—1 10.

O'Donnell, L., Robertson, K.M., Jones, M.E., Simpson, ER, 2001. Estrogen and
spermatogenesis. Endocr. Rev. 22, 289—318.

Raaijmakers, M.H., van Emst, L., de Witte, T., Mensink, E., Rayrnakers, RA, 2002.
Quantitative assessment of gene expression in highly purified hematopoietic cells

using real-time reverse transcriptase polymerase chain reaction. Exp. Hematol.
30, 481—487.

Roberge, M., Hakk, H., Larsen, G., 2004. Atrazine is a competitive inhibitor of
phosphodiesterase but does not affect the estrogen receptor. Toxicol. Lett. 154,
61—68.

Rotchell, J .M., Ostrander, GK, 2003. Molecular markers of endocrine disruption in
aquatic organisms. J. Toxicol. Environ. Health, Part B Crit. Rev. 6, 453—495.

Sakata, N., Tamori, Y., Wakahara, M., 2005. P450 aromatase expression in the
temperature-sensitive sexual differentiation of salamander (Hynobius retardants)
gonads. Int. J. Dev. Biol. 49, 417—425.

51

Sanderson, J .T., Seinen, W., Giesy, J .P., van den Berg, M., 2000. 2-Chloro-Striazine
herbicides induce aromatase (CYP19) activity in H295R human adrenocortical
carcinoma cells: a novel mechanism for estrogenicity? Toxicol. Sci. 54, 121—127.

Sanderson, J .T., Boerrna, J ., Lansbergen, G.W.A., van den Berg, M., 2002. Induction and
inhibition of aromatase (CYP19) activity by various classes of pesticides in
H295R human adrenocortical carcinoma cells. Toxicol. Appl. Pharmacol. 182,

44—54.

Simpson, E.R., Mahendroo, M.S., Means, G.D., Kilgore, M.W., Hinshelwood, M.M.,
Grahamlorence, S., Amameh, B., Ito, Y.J., Fisher, C.R., Michael, M. D.,
Mendelson, C.R., Bulun, SE, 1994. Aromatase cytochrome-P450, the enzyme
responsible for estrogen biosynthesis. Endocr. Rev. 15, 342—355.

Thellin, O., Zorzi, W., Lakaye, B., De Borman, B., Coumans, B., Hennen, G., Grisar, T.,
Igout, A., Heinen, E., 1999. Housekeeping genes as internal standards: use and
limits. J. Biotechnol. 75, 291—295.

Thompson, E.A., Siiteri, PK., 1974. Involvement of human placental microsomal
cytochrome-P-450 in aromatization. J. Biol. Chem. 249, 5373—5378.

Trant, J .M., Gavasso, S., Ackers, J., Chung, B.C., Place, A.R., 2001. Developmental
expression of cytochrome P450 aromatase genes (CYP19a and CYPI9b) in
zebrafish fry (Danio rerio). J. Exp. Zool. 290, 475—483.

Vandesompele, J., De, P.A., Speleman, F., 2002. Elimination of primer—dimer artifacts
and genomic coamplification using a two-step SYBR green I realtime RT-PCR.
Anal. Biochem. 303, 95—98.

Watanabe, M., Nakjin, S., 2004. Forskolin up-regulates aromatase (CYP19) activity and
gene transcripts in the human adrenocortical carcinoma cell line H295R. J.
Endocrinol. 180, 125—133.

Wiechmann, A.F., Smith, A.R., 2001. Melatonin receptor RNA is expressed in

photoreceptors and displays a diurnal rhythm in Xenopus retina. Brain Res. Mol.
Brain Res. 91, 104—111.

52

Yamada, K., Harada, N., Honda, S., Takagi, Y., 1995. Regulation of placenta specific
expression of the aromatase cytochrome-P-450 gene-involvement of the
trophoblast-specific element-binding protein. J. Biol. Chem. 270, 25064—25069.

53

CHAPTER 2

Fluorescence in situ hybridization techniques (FISH) to detect changes in CYP19a gene

expression of Japanese medaka (Oryzias latipes)

Abstract

The aim of this study was to develop a sensitive in situ hybridization methodology using
fluorescence labeled riboprobes (FISH) that allows for the evaluation of gene expression
profiles simultaneously in multiple target tissues of whole fish sections of Japanese
medaka (Oryzias latipes). To date FISH methods have been limited in their application
due to auto-fluorescence of tissues, fixatives or other components of the hybridization
procedure. An optimized FISH method, based on confocal fluorescence microscopy was
developed to reduce the auto-fluorescence signal. Because of its tissue- and gender-
specific expression and relevance in studies of endocrine disruption, gonadal aromatase
(CYP19a) was used as a model gene. The in Situ hybridization (ISH) system was
validated in a test exposure with the aromatase inhibitor fadrozole. The optimized FISH
method revealed tissue specific expression of the CYP19a gene. Furthermore, the assay
could differentiate the abundance of CYP19a mRNA among cell types. Expression of
CYP19a was primarily associated with early stage oocytes, and expression gradually
decreased with increasing maturation. No expression of CYP19a mRNA was observed in
other tissues such as brain, liver, or testes. F adrozole (100 ug/L) caused up-regulation of
CYP19a expression, a trend that was confirmed by RT-PCR analysis. In a combination

approach with gonad. histology, it could be shown that the increase in CYP19a expression

54

as measured by RT-PCR on a whole tissue basis was due to a combination of both
increases in numbers of CYP19a-containing cells and an increase in the amount of

CYP19a mRNA present in the cells.

Introduction
In recent years, an increasing number of genomic and/or proteomic techniques have been
developed to identify mechanisms of toxic action in organisms exposed to environmental
contaminants. Most of the methods that were developed to meet these objectives such as
RT-PCR, Northern blotting, and RNAse protection assays rely on high yields of RNA
extracted from whole tissues. However, the limitations of these techniques are that they
often fail to detect gene expression of low-abundance mRNA in small tissues, or they do
not allow localizing changes within certain tissues or cell types. Some genes are only
expressed in certain tissues, while others are expressed in specific tissues at only certain
times of development (Sanderson et al., 2001). Especially when using small laboratory
animal model species, the limited amount of individual tissues available for study and the
difficulty in excising them from the organisms has limited the efficacy of these
techniques to determine effects during critical windows of time during ontogenesis.
Many of the efforts in endocrine disruptor research have focused on individual
endpoints such as receptor-mediated effects (Otsuka, 2002). However, such targeted
screening methods may not be sufficient when disruptions are induced through indirect
mechanisms. Some chemicals can act as direct agonists or antagonists to certain
receptors while others act indirectly by modulating signal transduction, or affecting gene

expression or substrate concentrations. For example, the triazine herbicide atrazine does

55

not bind to the estrogen receptor (ER), but in vitro in a mammalian cell system, atrazine
has been found to up-regulate the expression of aromatase (CYP19), the enzyme that
transforms testosterone to estradiol. Although atrazine does not act like a typical
estrogen via binding the ER, in mammalian cell systems it can, under some situations, at
relatively great concentrations result in estrogenic effects by increasing endogenous
estradiol production (Sanderson etal., 2000). As a result, it is not only important to
develop methodologies that allow for evaluation of chemical-induced effects in multiple
target tissues simultaneously, but also to determine subtle effects on multiple endpoints
simultaneously within these tissues.

Whole-animal in Situ hybridization (ISH) is a promising method for determining
spatial changes in gene expression (Tompsett et al., 2008; Zhang et al., 2008a and b).
This methodology allows determination of effects on expression of multiple genes in
multiple tissues simultaneously, and it can be used simultaneously with standard
histology (Peterson and McCrone, 1993; Lichter, 1997; Hrabovszky et al., 2004; J ezzini
et al., 2005; Ijiri et al., 2006). One of the major advantages of ISH is that it allows
detection of changes in expression of mRNA for specific genes in organs, tissues, and/or
cells of interest in a manner that is consistent with other methods that are used to detect
lesions, including histopathology and immuno-histochemistry (IHC) (Streit and Stern,
2001). The principle underlying ISH is the hybridization of specifically-labeled probes to
the complimentary mRNA sequences in tissue or cells. A number of different
visualization techniques can be applied to detect an ISH signal including radionucleotides,
enzyme linked systems (e. g. biotin, digoxygenin), and fluorophores. Each label type has

strengths and weaknesses depending on application. Radiolabelled probes have been

56

widely used to detect specific mRNA sequence in tissues or embryos since the detection
of mRNA in invertebrates and vertebrates was originally developed using the
radiolabelled probe systems (Simeone, 1999). ISH utilizing radiolabelled probes have
been found to be more sensitive and reliable than some other methods such as enzyme
linked or fluorophore-based systems. However, radioisotope-based techniques have a
number of disadvantages such as a relatively poor resolution, relatively long exposure
times for auto-radiographic visualization, and they are expensive and require special
certifications in many institutions and extra precautions in the laboratory (Braissant and.
Wahli, 1998; Simeone, 1999; Pemthaler and Amann, 2004; Tompsett et al., 2008). In
contrast, enzymatic detection systems such as digoxygenin are very sensitive but tend to
be variable. More recently, the application of fluorescent labeling techniques has
considerably improved ISH due to the advantage of using different fluorescent tags to
Simultaneously detect different gene sequences (Wilkinson, 1999). However, application
of fluorescent in situ hybridization (FISH) methods to detect Specific mRNA in tissue
sections has not been explored to the same extent as radioisotope or enzyme based
methods due to issues with sensitivity and/or auto-fluorescence of tissues (Dirks, 1990;
Wilkinson, 1999; Andreeff and Pinkel, 1999). Recent improvements in fluorescence
labeling techniques render FISH techniques an increasingly useful tool. To effectively
utilize FISH, however, a number of technical limitations needed to be overcome. Key
issues include probe penetration of sections, auto-fluorescence of tissues, non-Specific
binding of probe, type of target tissues and Species, and sample preparation (Wilkinson,
1999). Hence, there was need for the development and optimization of FISH methods to

overcome these issues.

57

The main objective of this study was to develop and optimize an ISH protocol that
uses fluorophore-labeled probes to detect specific mRNA sequences in whole animal
sections of Japanese medaka (Oryzias latipes). Specific goals of this study were: (1)
develop and optimize methods to design fluorescent riboprobes for use in ISH; (2)
develop and optimize methods to reduce auto- and background fluorescence in fish tissue
sections by using a combination of chemical treatment and advanced confocal
microscopy techniques; (3) validate the FISH methods developed in this study using Q
RT PCR; and (4) use the optimized FISH methods to examine changes in gonadal
CYP19a gene expression in Japanese medaka exposed to a competitive pharmaceutical
inhibitor of the aromatase enzyme, fadrozole. The physiology, embryology, and genetics
of the Japanese medaka have been extensively studied in the past, and more recently, this
species has been used as a model in endocrine disrupter research (Wittbrodt et al., 2002).
The Japanese medaka has clearly defined sex chromosomes and sex determination
(summarized in Wittbrodt et al., 2002). Cytochrome P450 aromatase, encoded by the
CYP19 gene, is the key enzyme in estrogen biosynthesis from androgens (Simpson et al.,
1994), and it has been extensively used as an endpoint to assess the exposure of
endocrine disrupting compounds (EDCS) due to its relation with reproductive processes
(Sanderson et al., 2000; Hayes et al., 2002; Rotchell and Ostrander, 2003; Hecker et al.,
2006). Fadrozole has been reported to affect CYP19a gene expression (Villeneuve et al.,
2006) but was also shown to result in series of other physiological effects in fish
including plasma estradiol concentrations, gonadal pathologies, and fecundity (Afonso et

al., 1999; Ankley et al-, 2002; Fenske and Segner, 2004).

58

Materials and Methods
Test chemical
The fadrozole (CG8016949A; MW: 259.74g) used in this research was provided by

Novartis Pharma AG (Basel, CH).

Culture ofJapanese medaka

Japanese medaka were obtained from the aquatic culture at the US EPA Mid-Continent
Ecology Division (Duluth, MN, USA). Medaka were held in flow through systems under
conditions facilitating breeding (23-24 0C, 16:8 light/dark). All procedures used during
all phases of this study were in accordance with protocols approved by the Michigan

State University Instituted Animal Care and Use Committee (IACUC).

F adrozole exposure

Prior to initiation of exposure experiments, 12-14 wk old medaka were placed into 10 L
tanks with 6 L of carbon filtered tap water and acclimated for 12 (1 under the same
conditions as in the subsequent exposure of fadrozole. One fish died during acclimation.
Each treatment group consisted of replicate tanks, and each tank contained 5 male and 5
female fish. After the acclimation period, fish were exposed to l, 10, or 100 11g fadrozole
/L or carbon filtered tap water as a control in a 7 (1 static renewal exposure. Every day
one half of the water in each tank (3 L) was replaced with fresh carbon filtered water
dosed with the appropriate amount of an aqueous fadrozole stock (5 mg/L). Fish were
fed Aquatox flake food (Aquatic Ecosystems, Apopka, FL, USA) ad libidum once daily

and held at 24 °C with a 16:8 light/dark cycle. Water quality parameters were measured

59

daily and values were within a normal range for water quality, as follows in all tanks:
temperature (24 °C), pH (7.89-8.13), ammonia nitrogen (<0.02-0.04 mg/L), nitrate
nitrogen (<0.02-0.3 mg/L), dissolved oxygen (4.3-6.9 mg/L), and hardness (370-480 mg
CaCO3/L).

After 7 d of exposure medaka were euthanized in Tricaine S (50 mg/mL)
(Western Chemical, Femdale, WA, USA). Weight and snout length were recorded. Fish
were separated into two groups, one group was for ISH and consisted Of 2 fish per sex per
tank, and a second group that was to be used for Q RT PCR procedures and included
three fish per sex and treatment group. Fish from the ISH group were fixed for ISH and
histological investigations as described below. For the Q RT PCR group, the brain, liver,
and gonads were dissected from the fish and weighted individually. The liver somatic
index (LSI) and gonadal somatic index (GSI) were calculated as follows (Equations 1 and
2):

LSI = (liver weight/body weight)* 100 (1)

GSI = (gonad weight/body weight)* 100 (2)

The organs were then placed into cryovials and stored in liquid nitrogen for later RNA

extraction.

ISH procedure

Prgmration of sections

For ISH, fish were processed using methods adapted from Kong et al. (2008). Briefly,
fish were gross dissected to remove fins, tail, skull roof, otoliths, and opercula. The body

cavity was opened to improve the penetration of fixative (80% Histochoice MB (EMS,

6O

Hatfiled, PA, USA), 2% paraforrnaldehyde, and 0.05% glutaraldehyde) for better internal
organ fixation. Fish were then immersed in individual vials containing the fixative, and
allowed to fix over night at room temperature. After approximately 22 h, fish samples
were removed from the fixative, and were washed with 70% methanol and dehydrated
through a graded methanol series (80%, 95%, and 100%), and then cleared in chloroform
at 4 °C. Fixed and cleared samples were infiltrated with melted Paraplast Plus paraffin
(McCormick Scientific, St. Louis, MO, USA) at 60 °C. The paraffin was allowed to
harden overnight at room temperature, and paraffin blocks were stored under RNase free
conditions at 4 °C until sectioning.

Fish samples were sectioned on a rotary AO-820 microtome (American Optical,
Buffalo, NY, USA) that had been cleaned and decontaminated with absolute ethanol and
RNase—Zap (Sigma-Aldrich, St. Louis, MO, USA). Serial sections were cut at 711m. The
sections were floated out onto a 40 °C water bath, and placed on Superfrost Plus slides
(Erie Scientific, Portsmouth, NH, USA), followed by drying at 40 °C overnight. Slides

were stored in RNase-free containers at room temperature until used for ISH.

Fluorescence labeled riboprobe synthesis

 

In this study, RNA probes (riboprobes) were used for the detection of target mRNA. All
sequences used to design RNA probes were obtained from the NCBI database

(wwwncbi.nlrn.nih.gov). To synthesize riboprobes, reverse-transcribed first-strand

 

cDNA was used as a template in a conventional PCR with appropriate primers to amplify
PCR products (Table 2.1). The probes were designed to be approximately 500 bp long
using Beacon Designer 2 software (PREMIER Biosoft Int., Palo Alto, CA, USA). Probe

61

length was chosen based on a review of Wilkinson (1999) that reported that either too
Short or too long probes may give weaker signals possibly due to either low specificity to
target transcript or low penetration efficiency into tissue, respectively. The sequence of
the riboprobe to detect CYP19a mRNA was compared with all sequences of known genes
in Japanese medaka using Blast2 analysis (NCBI, Bethesda, MD, USA), and no sequence
homogeneity was found except for the target gene.

The PCR products were cloned into a pGEM T-Easy vector (Promega, Madison,
MI, USA) following manufacturer’s direction SO that it was flanked by two different
RNA polymerase initiation sites (T7 and SP6). Sequence validity of cloned amplicons
was confirmed by automatic DNA sequencing and followed by a BLAST2 analysis with
their corresponding sequences in GenBank. In order to synthesize sense and antisense
probes for CYP19a, cloned plasmids were digested with SalI and NcoI (Invitrogen,
Carlsbad, CA, USA), respectively. Complete digestion was confirmed with
electrophoresis on agarose gel (data not Shown).

Sense and antisense riboprobes for CYP19a mRNA were synthesized using in
vitro transcription. Briefly, the sense riboprobes were transcribed with T7 polymerase
with their respective plasmids, while the antisense probes were transcribed with SP6
polymerase using manufacturer’s direction (Roche, Indianapolis, IN, USA). Sizes of
synthesized riboprobes were confirmed by MOPS-formaldehyde gel electrophoresis (data
not shown), followed by purification using lithium chloride precipitation (Ambion, Foster
City, CA, USA).

Synthesized riboprobes were labeled with ULYSIS Nucleic Acid Labeling Kits

(Alexa Fluor 488, Molecular Probes, Eugene, OR, USA). After labeling, the riboprobes

62

were purified with a gel filtration based spin column, Micro Bio-Spin 30 Columns in
RNase-Free Tris (Bio-Rad, Hercules, CA, USA) to remove excess, unincorporated
fluorescent dyes. The quality of fluorescence labeled riboprobes was confirmed by
MOPS—formaldehyde gel electrophoresis without ethidium bromide (data not shown).
The quantity of the riboprobes was measured using a spectrophotometer (260 nm).
.Riboprobes were separated into aliquots and stored at -80 °C. Probes were used within

few days after synthesis to minimize degradation of RNA degradation.

Fluorescent in situ hybridization

Slides on which whole histological sections of medaka had been placed were incubated at
60 °C for 1 h to allow paraffin to melt and to fuse the sections to the slide (Fig. 2.1). The
slides were then de-paraffinized and rehydrated as follows; washing with (a) xylene 3
times for 3 minutes, (b) 100% ethanol 2 times for 3 min, (c) PBS for 5 minutes, and (e)
diethyl pyrocarbonate (DEPC)-treated water for 1 min. In order to reduce auto-
fluorescence signal originating either from the tissues or from the fixative, slides were
then treated 3 times for 20 min with 10 mg/mL sodium borohydride (SB) (Sigma-Aldrich,
Saint-Louis, MO, USA) in PBS (Billinton and Knight, 2001). Slides were then rinsed
with PBS and DEPC treated water and the sections were perrneableized for 20 min with
0.2N HCl. To increase specificity of probes to target mRNA, slides were treated with 0.1
U/mL of RNase-free DNase I (Roche, Indianapolis, IN, USA), followed by inactivation
of DNase I with DNase stop solution (10mM Tis-HCl, 150mM NaCl and 20mM EDTA).
Sections were acetylated in triethanolamine-HCI buffer plus 0.25% acetic anhydride (2

times) (Sigma—Aldrich, Saint-Louis, MO, USA), and then pre-hybridized with

63

hybridization buffer (2X SSC, 50% deionized formamide, 1X Denhardt’s solution, 0.4
ug/mL of sonicated salmon sperm DNA, 1% SDS, 20% dextran sulfate, and DEPC-
treated water to make 31.25 mL) without probes for l h in a humid box at 43°C to reduce
non-Specific binding of probes. Probes were denatured at 90 °C for 1.0 min and chilled
on ice. Sections were hybridized with the riboprobe (2 ng/ 11L) at 43 °C for 16 h. The
negative control consisted of sections that were hybridized with equal amounts of sense
probe under the same ISH conditions as described above. Following hybridization, the
slides were washed in 3 SSC gradient (4X at room temperature (5 min), 2X at 37 °C (10
min), 2X at RT (5 min), and 0.2X at room temperature (5 min)). Following washing,
slides were air dried, mounted with fluoromount G (EMS, Hatfield, PA, USA), and left in
a dark chamber until the mounting medium was completely dry. Thereafter, slides were
kept in the slide box to prevent photo-bleaching and photo-activation of fluorescent
molecules until analysis by confocal microscopy. The optimum concentration of SB to
quench auto- and background fluorescence signals was determined during preliminary
studies with SB at 10 or 20 mg/mL, copper (II) sulfate at 10 or 100 mM, and a

combination of SB and copper (II) sulfate.

Confocal laser scanning microscopy (CLSM) image analysis

Spectral distributions of fluorescence were determined with a Zeiss LSM 510 Meta
confocal system (Carl Zeiss, Jena, Germany). Images were collected with a Zeiss EC
Plan NEOFLUAR 10X (Carl Zeiss, Jena, Germany). Using the 458-nm line of an argon

laser for excitation, l6 spectrally resolved images were recorded at 10.7 um intervals

64

from 475.5 um — 646.7 um. Sections of ovary were imaged with a pinhole aperture
corresponding to an optical thickness of < 2.0 pm. ,

Several endogenous and/or fixative-induced fluorophores have broad band
emission spectra that make separation from emission Spectra related to riboprobe
fluorophores particularly difficult. In this study, the effort focused on the separation of
the individual spectral components associated with auto-fluorescence, background, and
Alexa F luor 488 dye. Auto-fluorescence and background spectral components were
obtained from ovary sections (early stage oocytes) and other tissues, respectively, on
Slides hybridized without probe to account for these variables in the overall fluorescence
associated with these sections. The specific spectrum of Alexa Fluor 488 dye was
obtained directly from the dye reagent. Once the number of significant and independent
sources of the specific spectral components was defined by means of CLSM, the linear
spectral unmixing was applied to separate the individual components, to remove
autofluorescence Signal in the recorded images (Fig. 2.2) and to isolate the specific
fluorescence signal of the Alexa Fluor 488 dye. With each set of ISH experiments, a
section with no probe added was included to control for alterations in the
autofluorescence spectral shape by photo-bleaching that can result from consecutive laser
scans.

Each set of ISH slides had at least one ISH section with sense probe that served
as a negative control. Images of the ovary were collected at 10X magnification. For
quantification, 3 different portions were randomly selected in the ovary of each section
hybridized with CYP19a antisense probe and then 3 early Stage of oocytes in each portion

were randomly selected (less than 100 pm in diameter). Based on morphological criteria

65

developed by Iwamatsu et al. (1988), those oocytes are classified into the previtellogenic
stage. Then we applied the above described spectral unmixing method to quantify the
intensities of the true Alexa F luor signal in these oocytes. Due to the detection of low
quantities of Alexa F luor dye in the section hybridized with sense probe, we normalized

the CYP19a antisense signal to the CYP19a sense Signal in each set ofISH.

Deparaffinization and rehydration

3

Removal of autofluorescence

5

Sections permeabilization

«I

DNase treatment

I

Acetylation

l
Prehybridization

a
Hybridization

5

Posthybridization wash and mounting

8

Confocal microscope analysis

Figure 2.1. Brief steps of mRNA in Situ hybridization with fluorescence labeled
riboprobe.

66

Q RT PCR procedure
Total RNA isolation, cDNA synthesis, and RT PCR were conducted as described by Park
et a1. (2006). To confirm complete removal of possible genomic contamination, a
negative control (sample without reverse transcriptase) was run in parallel with each PCR
experiment, and which resulted in no amplification Of the PCR product (data not shown).
TO improve sensitivity of PCR amplification, the cDNAzRNA hybrid molecules were
removed by digestion with E. coli RNase H after first-strand cDNA synthesis. The
expression level of CYPI9 mRNA was normalized to an internal control gene, ,[1’ actin.
To determine the accumulation of the PCR product, SYBR Green I dye
(bioMérieux, Marcy l'Fttnle. France) was used as a real-time reporter of the presence of
double-stranded DNA. All cDNA sequences were obtained from the public GenBank
database Of NCBI. The primers were obtained using Beacon Designer 2 (Premier Biosoft
International, Palo Alto, CA, USA). Primer sequences and conditions used for Q RT
PCR analysis are given (Table 2.2). Primer specificity was verified by a single distinct
peak obtained during the melting curve analysis of the SYBR Green I based real time
PCR system and by DNA sequencing of the PCR amplicons separated by gel

electrophoresis (data not shown).

Histology

Histological changes in medaka ovaries were evaluated using H & E Stained sections
(Tompsett et al., 2008). Briefly, Slides were de-paraffinized in xylene and rehydrated
through a descending ethanol series (100, 95, and 70%). Slides were then stained in

Harris’ hematoxylin (EMS, Hatfield, PA, USA) for 3 min, processed through acid alcohol,

67

ammonia, and ethanol washes, and then stained in 1% Eosin Y (EMS, Hatfield, PA,
USA) in 80% ethanol for l min. Slides were then dehydrated through an ethanol series
(70, 95, and 100%) and cleared in xylene. Slides were preserved under glass cover Slips
using Entellan mounting medium (EMS, Hatfield, PA, USA) and allowed to dry. Images
of the gonad on each slide were recorded using a Camedia C-3040 ZOOM digital camera
(Olympus, Center Valley, PA, USA) attached to an Olympus BX41 microscope (Optical

Analysis Corporation, Nashua, NH, USA).

Statistics

Statistical analyses in this study were conducted using SAS (SAS Institute Inc. Cary, NC,
USA). Prior to analysis, data sets were tested for normality using the Shapiro Wilk’s test.
One way ANOVA test was applied to test for differences of CYP19a gene expression
across all treatment groups in both Q RT PCR and ISH, followed by the Student-
Newman-Keul’s test for multiple comparisons. The 2-tailed Spearman rank correlation
analysis was used to evaluate the relationship between CYP19a gene expression levels by
means of Q RT PCR analysis and ISH analysis. The criterion for significance in all

statistical tests was p<0.05.

68

 

UHUHOOHOO<HOHHOHUOH<DF

 

 

 

 

- . . SB: :3
CS? omEc mtmo ow xcwEm \O<D<UUUD.E.OUUHHOO<O . m
. 1 . UPU<U<UUH<OPPOUUHUFUU
\ \ a 3
ommm owow 2&0 we woomwQ .\O<OHHH<FUUFOH<HHUOUUUU Eoib
EDEE Dov $.25: Uov AdeE Uov
55:85 m::wo::< :ocfigwcoo 6N6 .SnEs: A m- m .om:omz:m 8:03 £2:an mac—Sm wooowam
523:5. :o_mmooo< . . . \ , : . .
:oEEoo wEUAU . .
Ewen—ago.

MU; PM 0 :8 m:o:6:oo wczoxo Em .ONB :oOszm £383: :ommmooom diam—:00 .3255 :23 85w BBmSH Nd 03m...

 

UFOH<O<DDHHOHUUO<O<<O

 

, \_ o .
00R: onm mag co: wcomwo .O<UHO<O©PUHDDH<<HHDHUU ixib
AdeE Uov EOEE Dov AdeE Dov
:ommfizxm w::mo::< :o_§._2m:oQ “Em 59:2:

 

- .om:omz:m 3:8 wcoE: me o:
. :oO_EE< :o_mmuoo< hm .m . \ v . m a A—
:02950 w::o>U . .

 

do; .m:o::o>:8 :8 :oE::oo @298 “Em 6N6 :oozaEe .5965: 5688.“ xswmfio 5253 5:5 monoi _.m Bank

69

Results

Reduction of auto/luorescence

Initial experiments revealed strong auto-fluorescence in non-treated sections (Figs. 2.2A-
B). Of the treatments tested, only sodium borohydride (SB) and SB + copper (II) sulfate
significantly decreased autofluorescence, but there were no significant difference in the
Signal intensity between the two treatments (data not shown). Therefore, SB was used in
all further experiments. However, complete removal of autofluorescence to levels
Observed in the negative controls was not possible (Figs.‘2.2C and 2.213). When SB-
treated sections were subjected to linear spectral unmixing of CLSM the background
fluorescence signal previously observed in medaka ovaries was reduced or eliminated.
(Fig. 2.2D) relative to traditional excitation/emission filter system laser microscopy (Fig.
2.2C). Thus, for further analysis it was decided to treat all samples with 10mg/mL SB to
reduce and/or minimize the signal of autofluorescence, and then to subject each image to

the above described spectral unmixing method.

70

 

 

 

 

Emission wavelength (nm) Intensrty
A B C
513 1356.1 869.3 733.2
524 1035.6 979.6 769.9
534 1131.0 1087.1 818.7

 

Figure 2.2 Autofluorescence images ofjuvenile medaka ovary and emission Spectra of
the sections obtained after excitation with a 488 laser (A, B, and C) ofCLSM and
autofluorescence image of section after applying Linear Spectral unmixing (D) ofCLSM.
(A). Section not subjected to ISH procedure; (B). section in situ hybridized without probe
and without SB treatment; (C). Section in situ hybridized without probe and with SB
treatment; (D). Section in situ hybridized without probe and SB treatment after
application oflinear unmixing; (E). Autofluorescence intensities ofthe sections. P0 =
previtellogenic oocytes. Scale bar = 100 um.

Tissue and cell specificity ofCYPI 9a gene expression
Longitudinally sectioned whole medaka hybridized with CYP19a antisense probe

exhibited a fluorescence signal that was Specific to ovary (Fig. 2.3A). The specificity of

71

hybridization was demonstrated with negative controls using sense probe, which gave a
very weak fluorescence signal (Fig. 2.3B). Sections hybridized in the absence of probe
showed no signal (Fig. 2.3C). The greatest fluorescence was found in thecal cells.
granulosa cells, and premature early Stage pro-vitellogenic oocytes. CYP19a expression
in vitellogenic oocytes was much less and matured oocytes cells exhibited little
fluorescence (Fig. 2.3). If present, CYP19a expression was found only in the outer layers
of matured oocytes (Fig. 2.5). Furthermore. fluorescence Specific for positive ('1’1’I9a
staining was observed only in ovary and no CYP19a mRNA could was detected by 1811
in other tissues such as the brain or liver (Fig. 2.7). When measured by RT-PCR,
CYP19a gene expression was greatest in ovary, while both brain and liver tissues

expressed little CYP19a mRNA (data not Shown).

 

Figure 2.3 Expression of CYP19a mRNA in the ovary of juvenile Japanese medaka after
hybridization of longitudinal whole mount sections with a fluorescence riboprobe.
Expression of CYP19a mRNA was detected in the ovary hybridized with antisense probe
(A); Very weak CYP19a detection was observed in the oocytes hybridized with sense
probe (B); no Signal in the ovary hybridized without probe (C). P0 = previtellogenic
oocytes. Scale bar = 100 um.

72

Gonads
16 . 13 Female

I Male

 

Fold change of CYP19a gene expression
normalized to B actln
(X)

 

 

0- l

CTR 1 10 100
Fadrozole Treatment (pg/L)

Figure 2.4 Fold-changes of CYP19a mRNA gene expression by Q RT PCR analysis in
gonads of Japanese medaka exposed to fadrozole (1, 10, and 100 ug/L), using
comparative CT method for quantification of mRNA. [1' actin served as the internal

control gene. All data are expressed as the median :t the interquartile range. One—way
ANOVA was used to analyze dada by treatment groups for each tissue and sex separately,
followed by SNK test for multiple comparisons. Different letters indicate significant
difference between treatment (p<0.05).

‘5

Follicular cell layer

Poifi

 

Figure 2.5 Expression of CYP19a mRNA in the ovary ofjuvenile Japanese medaka using
the Optimized in situ hybridization. Strong Signal detection of CYP19a mRNA in the
early stage of oocytes was observed, while low CYP19a gene expression was localized in
the outer layer of follicular cell layer of the vitellogenic and matured oocytes. P0 =
previtellogenic oocytes. Bar = 100 um.

73

F adrozole exposure

Gene expression

 

Expression of CYP19a in ovaries of medaka exposed to fadrozole, as measured by both
RT PCR and ISH, was greatest in medaka exposed to 100 11g fadrozole /L. RT PCR
analyses showed that there was a significant, dose-dependent up-regulation of CYP19a
gene expression in ovary of fadrozole-exposed females (Fig. 2.4). Exposure to 100 11g
fadrozole/L caused a l4-fold up-regulation of CYP19a expression in ovaries, relative to
that of controls. While there was also greater CYP19a expression measured by ISH, due
to the relative great variation and the semi-quantitative nature of the analyses these
differences were not statistically significant. Furthermore, the 1.4 fold change observed
by use of ISH was less than that demonstrated by use of RT-PCR (Fig. 2.6). While the
up—regulation of CYP19a expression caused by 100 11g fadrozole /L showed the same
increasing trend than that determined by RT-PCR, ISH was unable to demonstrate effects
for lesser concentrations of fadrozole. Although trends in CYP19a expression were
similar between quantification between ISH and Q RT PCR, no statistically significant

correlation was observed (r2 = 0.015, n = 16, p = 0.957).

F adrozole exposure - Mortality, morpho-metric and histological endpoints

No mortality was observed in any of the exposure groups or the controls during the
course of the experiment. Fadrozole caused no statistically significant changes in liver
somatic (LSI) and gonadal somatic (GSI) indices of Japanese medaka (p > 0.05) (Fig.

2.8). However, in a parallel study exposure to 100 pg/L fadrozole caused histological

74

effects. While female medaka fish from the control, 1.0, and 10 ug fadrozole /L
treatments had oocytes in all stages of development, oocytes of females exposed to 100
pg fadrozole /L were predominantly in later Vitellogenic stages, similar in size, and
lacked a distinct yolk globule (stage VII and VIII of development as classified by
Iwamatsu et al. (1988)). None of the ovaries in fish from this treatment group had stage

IX or mature (with a pink-staining yolk globule) oocytes.

75

 

250 r

(C)
33,“ 200 -
C1
(1)
.13
T, 150 —
I:
.29
g 100 —
I:
0)
8
8 50 <
O
2
Ln
0 4

 

 

CTR 1 10 100
F adrozole treatment (pg/L)

Figure 2.6 Expression of CYP19a mRNA in the ovary of juvenile Japanese medaka using
the optimized FISH technique in the control (A) and 100 ug/L of fadrozole treatment
group (B) for 7 days. Fluorescence signal intensity of CYP19a expression in randomly
selected three early stage of oocytes in the ovary of Japanese medaka exposed to
fadrozole (C) and each bar represents means 1 SD. of 4 female fish. Significantly low
Signal of negative control (sense probe, right panel) and highly expressed C YP/ 9a

mRNA in the ovary exposed to 100 ug/L of fadrozole was observed. Scale bar = 100 um.

76

 

Figure 2.7 Expression of CYP19a mRNA in the brain tissue of female Japanese medaka
in the control (A), and brain tissue (B) and liver tissue (C) of 100 ug/L of fadrozole
treatment group by the optimized in situ hybridization using fluorescence antisense
riboprobe. No fluorescence signal detection was observed in brain and liver tissues. Bar :

100 um.

   

 

 

 

 

A B
'9' L81 -e’»- 1.81
4.00 ~ 4 GSI ”-00 ' —. cs1
3.00 9.00
2.00 ‘ 6.00
1.00 - 3.00
0 ‘ V l l 0 l' I I I
CTR l 10 100 CTR 1 10 100
F adrozole treatment (pg/L) Fadrozole treatment (pg/L)

Figure 2.8 Gonadal somatic index (GSI) and liver somatic index (LSI) of male (A) and
female (B) Japanese medaka exposed to fadrozole for 7 days. Bars represent mean and
error bars are standard deviation.

77

Discussion

Optimization of FISH

Optimization of fluorescence ISH protocols, as described in this study, resulted in the
development of a test system that allows for the identification of gene expression in
paraffin embedded whole mount sections of small fish with relatively great sensitivity
and spatial specificity. To date, the application of fluorescence microscopy techniques in
ISH has been limited by the relatively great auto-fluorescence of many tissue types or
components (e. g. fixative) used in the assay (SzOllOsi et al., 1995). Typically, the
emission spectrum of auto-fluorescence is very broad compared to the spectra of
exogenous source such as a fluorescently labeled probe, complicating the use of
fluorescent probes further because auto-fluorescence cannot be avoided by simply
choosing dyes with excitation and emission spectra out of the range of the spectra of
autofluorescent molecules. Sodium borohydride (SB) is a known blocker of aldehyde
groups that are generated after reactions of amines and protein molecules with aldehyde
fixatives, and eventually these aldehyde groups combine covalently to any amino groups,
resulting in aldehyde-induced fluorescence (Beisker et al., 1987). However, the capacity
of SB to reduce auto-fluorescence seems to be species and/or tissue specific. For
example, attempts to reduce the auto-fluorescence in the aldehyde fixed-human bone
marrow paraffin sections failed, and even resulted in an increase of fluorescence
(Baschong etal., 2001). Other studies have reported that SB quenched the fixative
induced auto-fluorescence in brain tissue of mammals (Tagliaferro et al., 1997; Clancy
and Cauller, 1998). While in our study treatment with SB significantly reduced auto-

fluorescence, the specific signal of Alexa Fluor 488 dye was still obscured by the

78

relatively strong autofluorescence Signal when using traditional fluorescent microscopy
techniques, which indicates that aldehyde-induced fluorescence was only one contributor
to the Observed auto-fluorescence. Furthermore, the small difference in the strength of
the auto-fluorescence signal of sections that did and did not undergo ISH implies that the
components of the ISH procedures used in this study did not significantly contribute to
auto-fluorescence. The true dye signal could be detected after complete removal auto-
fluorescence signal by use of linear spectral unmixing with CLSM to isolate the Alexa

F luor 488 dye signal from the remaining auto-fluorescence. The linear spectral unmixing
technique is able to identify and separate specific spectral components in fluorescence
images (Chorvat et al., 2005). The application of this procedure allowed isolation of the
dye-Specific signal that allowed clearly discerning between background and gene probe
signal. The significantly greater fluorescence intensity in sections hybridized with
antisense probe compared to those hybridized with the sense probe indicated the
specificity of the developed ISH protocol. However, the optimization techniques applied
did not completely reduce the fluorescence signals in samples processed with sense probe.
Possible reasons for this could be that probes were not completely removed from the
tissue during post-hybridization washing steps, or that some Of the tissue
autofluorescence signal corresponded to the specific spectrum of Alexa Fluor 488 dye.
Regardless of these artifacts, the FISH system developed in this study allowed for spatial
determination of Specific gene signals in medaka with high resolution and sensitivity,
which represents a significant improvement compared to most of the previous FISH

approaches.

79

Validation of FISH method

The Q RT PCR analysis revealed differential expression of CYP19a mRNA in brain and
ovary of juvenile female medaka (data not Shown). Average CT values for ovary and
brain were 21 and 35, indication a 2 x 1014 difference in abundance of CYP19a transcript
between these tissues. Significant greater induction of CYP19a mRNA in ovary
compared to brain is in accordance with findings of other studies with teleost fish
(Callard et al., 2001; Kishida and Callard, 2001 (zebrafish); Villeneuve et al., 2006
(fathead); Liu et al., 2007 (catfish)). The results of the ISH revealed a clear signal for
CYP19a mRNA only in ovary. However, while no expression of CYP19a in brain could
be observed with ISH, expression was detectable at very low levels by use of Q RT PCR.
This result indicates that ISH on whole animal sections is relatively less sensitive than
use of Q-RT-PCR with excised tissue. This difference in sensitivity is likely due to the
fact that CYP19a was amplified during the PCR process while the ISH signal is
proportional to the absolute amount of mRNA present in the tissue. Furthermore, the RT
PCR method applied here utilized mRNA extracted from an entire gonad while the ISH
procedure is limited to the visualization of a small section of a tissue. Now that the utility
of the ISH method has been demonstrated for whole tissue mounts, future work could
increase tissue-Specific sensitivity by use of in situ PCR.

While qualitative determinations of up-regulation of C YP19a expression by
fadrozole could be detected both by the ISH on whole fish sections and RT-PCR with
excised tissues the correlation between the magnitude of change measured by the two
methods was poor (r2 = 0.015). This result is in accordance with a parallel study that

investigated the effects of fadrozole on CYP19a gene expression in the same fish using

80

radionucleotide based ISH (Tompsett et al., 2008). The increase in ovarian CYP19a
expression after exposure to fadrozole compares well to findings Of Villeneuve et al.
(2006) in juvenile female fathead minnow. It has been hypothesized that this increase in
gene expression, which is opposite to that of enzyme activity, is most likely due to a
compensatory response through increased GtH from pituitary in response to decreased
levels of E2 (Kime, 1998; Tompsett et al., 2008). However, these findings indicate that
the developed fluorescent ISH method is indicative of changes at the gene expression
level that were previously reported in the same experiment using different analytical (RT-

PCR) or detection (radionucleotide ISH) systems.

Utilization of FISH

The high resolution and integration into classical histological analysis suggests that FISH
is a useful technique for use in studies aiming to elucidate mechanisms of effects of
chemicals. However, while FISH allows the localization of mRNA in a number of
tissues simultaneously with comparison to histological responses, FISH is less sensitive
and more variable than Q RT PCR. Thus, currently the results of FISH in whole animals
sections are semi-quantitative. This is in accordance with a parallel study using
radionuceotide based ISH, and which also reported relatively great variation between
replicate slides and difficulties to quantify the ISH Signal (Tompsett et al., 2008).
Utilization of ISH for localization and expression of mRNA in tissues or whole
organisms are not novel, but the improvement made by reduction in auto-fluorescence
has made it more useful. Specifically, the expression of the CYP19a gene has been

studied in various teleost fish species previously (Goto-Kazeto et al., 2004; Kobayashi et

81

al., 2004; Dong and Willett, 2008; Wang and Orban, 2007; Tompsett et al., 2008).
However, these studies have employed non-fluorescent visualization systems such as
enzyme or radionucleotide labeled probes. While such systems represent useful
approaches to identify the spatial and tissue specific expression of genes, they are
associated with a series of issues that limit their applicability. Enzyme based detection
systems such as digoxygenin and biotin require immunochemical steps that are time and
sometimes cost intensive. Furthermore, while more sensitive than fluorescence based
systems, enzymatic or immunological amplification steps for probe visualization
typically Show a poor correlation between signal intensity and the abundance of mRNA
(Day etal., 2007). One of the limitations of the use of radionucleotide detection methods
is the limited resolution and sensitivity by which tissue specific changes in gene
expression can be determined (Day et al., 2007; Tompsett et al.,2008). While there are
alternative detection methods such as silver-grain based techniques that are highly
sensitive and provide good resolution (Hrabovzky et al., 2004), these techniques are labor

intensive and very expensive.

The FISH approach applied in this study allowed specific, high resolution
detection of CYP19a in whole mount female medaka sections. CYP19a was specific to
the ooplasma of early stage oocytes that were less than 100 pm in diameter and possessed
high expression of CYP19a, and the follicle cell layer of previtellogenic and vitellogenic
stage of oocytes, albeit less expressed, which is consistent with the results of other studies
with teleost fishes, such as killifish (Dong and Willett, 2008), zebrafish (Goto-Kazeto et
al., 2004; Wang and Orban, 2007), and Atlantic croaker (Nunez and Applebaum, 2006).

Overall, CYP19a mRNA gene expression gradually lessened with maturation of oocytes:

82

early stage of oocytes >> pre- and vitellogenic oocytes > mature oocytes. Expression of
CYP19a has been reported to be primarily localized in the ooplasm of primary growth
staged oocytes of developing killifish (F undulus heteroclitus) and gradually declined
from stage I oocytes (previtellogenic follicle) to non-detectable levels at stage IV oocytes
(matured follicles) (Dong and Willett, 2008). Expression of CYP19a mRNA was also
Observed in the follicular layer of developing oocytes in the medaka (Suzuki et al., 2004).
In Atlantic croaker, CYP19a mRNA transcripts were most expressed in previtellogenic
ovary (Nunez and Applebaum, 2006). In zebrafish, a greater abundance of CYP19a
mRNA was observed in the mid-vitellogenic oocytes than in the primary stage of oocytes,
which were embedded in connective tissue (Goto-Kazeto et al., 2004). This may be
indicative of differences in synthesis of E2 and associated vitellogenesis among species.
Overall, these findings indicate the validity of the FISH system used in our study as it
reproduced the findings of other studies with teleost fish species. The utilization of the
FISH could be further confirmed by the tissue specificity of CYP19a hybridization that

was limited to the ovary.

Comparison of FISH data to morphometric and histological results

No significant effects on morphometric parameters, such as organ or body sizes, were
observed to be caused by exposure to fadrozole, which is likely due to the relatively short
duration of exposure. The histological analysis indicated that an effect on development
of the ovary in response to fadrozole exposure occurred. This supports the hypothesis
that aromatase is important in sexual development and gonadal maturation in medaka.

This conclusion is supported by the results of studies that have shown that inhibition of

83

aromatase by specific inhibitors caused a decrease in plasma E2 and vitellogenin levels,
and ultimately inhibited oocyte growth in teleost fishes (Ankley et al., 2002; Suzuki et al.,
2004; Sun et al., 2007).

One of the major advantages of FISH is that it allows the localization of a
specific gene at the tissue and/or cellular level. This tissue or cell specific response
provides a clue for understanding relationships between molecular and histological
processes. While both molecular approaches utilized in this study revealed an increase in
CYP19a gene expression at the greatest dose tested, the magnitude of the effect was very
different, and based on this results the ISH data would have not allowed predicting the
great increase measured by RT-PCR. By combining the ISH with standard histology,
however, it was possible to demonstrate that there was a change in the number of earlier
stage oocytes, the types of cells that are characterized by increased CYP19a expression.
Given the relatively small change in CYP19a gene expression measured by ISH when
compared to RT-PCR, therefore, it can be concluded that the change in tissue
composition - increase in the number Of cells that have increased CYP19a expression — is
likely to be the main factor resulting in the increase in gonadal aromatase expression in
response to fadrozole exposure.

In summary, the optimized FISH method developed in this study allowed to
detect CYP19a mRNA expression in the ovary of medaka. The method not only can
provide useful information relative to temporal and spatial changes in gene expression,
but also can aid in explaining molecular changes at the level of histological observation

with the ultimate goal of being able to link histological changes to potential pathologies.

84

Furthermore, FISH has the potential to be further Optimized using multiple riboprobes

with different emission spectrums simultaneously with the same section.

Acknowledgement

This study was supported by a grant from the US. EPA Strategic to Achieve Results
(STAR) to JP. Giesy, M. Hecker and PD. Jones (Project no. R-831846). 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-04/04) to D. Au and JP. Giesy

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

85

References

Afonso, L.O.B., Iwama, G.K., Smity, J ., Donaldson, E.M., 1999. Effects of the aromatase
inhibitor fadrozole on plasma sex steroid secretion and ovulation rate in female

coho salmon, Oncorhvnchus kisutch, close to final maturation. Gen. Comp.
Endocrinol. 113, 221-229.

Andreeff, M., Pinkel, D., 1999. Introduction to fluorescence in situ hybridization:
principles and clinical applications. John Wiley & Sons, New York, p 455

Ankley, G.T., Kahl, M.D., Jensen, K.M., Homung, M.W., Korte, J.J., Makynen, E.A.,
Leino, RL., 2002. Evaluation of the aromatase inhibitor fadrozole in a Short-term

reproduction assay with the fathead minnow (Pimephales promelas). Toxicol. Sci.
67, 121-130.

Baschong, W., Suetterlin, R., Laeng RH, 2001. Control of autofluorescence of archival
fonnaldehyde-fixed, paraffin-embedded tissue in confocal laser scanning
microscopy (CLSM). J. Histochem. Cytochem. 49, 1565-1671.

Beisker, W., Dlobeare, F., Gray, J.W., 1987. An improved immunocytochemical
procedure for hi gh-sensitivity detection of incorporated bromodeoxyuridine.
Cytometry 8, 235-239.

Billinton, N., Knight, A.W., 2001. Seeing the wood through the trees: A review of
techniques for distinguishing green fluorescent proteins from endogenous
autofluorescence. Anal. Biochem. 291, 175-197.

Braissant O., Wahli W., 1998. A simplified in Situ hybridization protocol using non-
radioactively labeled probes to detect abundant and rare mRNAs on tissue
sections. Biochemica 1, 10-16.

Callard, G.V., Tchoudakova, A.V., Kishida, M., Wood, E., 2001. Differential tissue
distribution, developmental programming, estrogen regulation and promoter

characteristics of CYP19 genes in teleost fish. J. Steroid Biochem. Mol. Biol. 79,
305-314.

86

Chorvat, D.Jr., Kirchnerova, J ., Cagalinec, M., Smolka, J ., Mateasik, A., Chorvatova, A.,
2005. Spectral unmixing of flavin autofluorescence components in cardiac
myocytes. Biophys. J. 89, L55—L57.

Clancy, B., Cauller, LJ., 1998. Reduction of background autofluorescence in brain
sections following immersion in sodium borohydrate. J. Neurosci. Methods. 83,
97-102.

Day, D.J., Mrkusich, E.M., Miller, JH., 2007. Comparative quantitation of mRNA
expression in the central nervous system using fluorescence in Situ hybridization.
Methods Mol. Biol. 353,125-142.

Dirks, R.W., 1990. Simultaneous detection of different mRNA sequences coding for
neuropeptide hormones by in situ hybridization using FITC and biotin labeled
oligonucleotides. J. Histochem. Cytochem. 38, 467-473.

Dong, W., Willet, KL, 2008. Local expression of CYP19A1 and CYP19A2 in
developing and adult killifish (F undulus heteroclitus), Gen. Comp. Endocrinol.
155, 307-317.

Fenske, M., Segner, H., 2004. Aromatase modulation alters gonadal differentiation in
developing zebrafish (Danio rerio). Aquat. Toxicol. 67, 105-126.

Goto-Kazeto, R., Kight, K.E., Zohar, Y., Place, A.R., Trant, J .M., 2004. Localization and
expression of aromatase mRNA in adult zebrafish. Gen. Comp. Endocrinol. 139,
72-84.

Hayes, T.B., Collins, A., Lee, M., Mendoza, M., Noriega, N., Stuart A.A., Vonk, A.,
2002. Hermaphroditic, demasculinized frogs after exposure to the herbicide

atrazine at low ecologically relevant doses. Proc. Natl. Acad. Sci. U. S. A. 99,
5476-5480.

Hecker, M., Sanderson, J., Karbe, L., 2006. Supression of aromatase activity in
populations of bream (Abramis brama) from the river Elbe, Germany.
Chemosphere 66, 542-552.

87

Hrabovszky, E., Kallo, I., Steinhauser, A., Merchenthaler, I., Coen, C., Peterson, 8.,
Liposits, Z., 2004. Estrogen receptor-beta in oxytocin and vasopressin neurons of
the rat and human hypothalamus: Immunocytochemical and in situ hybridization
studies. J. Comp. Neurol. 473, 315-333.

Ijiri, S., Takei, N., Kazeto, Y., Todo, T., Adachi, S., Yamauchi, K., 2006. Changes in
localization of cytochrome P450 cholesterol side-chain cleavage (P450300) in

Japanese eel testis and ovary during gonadal development. Gen. Comp.
Endocrinol. 145, 75-83.

Iwamatsu, T., Ohta, T., Oshima, E., Sakai, N., 1988. Oogenesis in the medaka Oryzias
latipes-stages of oocytes development. 2001. Sci. 5, 353-373.

Jezzini, S., Bodnarova, M., Moroz, L., 2005. Two-color in situ hybridization in the CNS
oprlysia californica. J. Neurosci. Methods. 149, 15-25.

Kime, DE, 1998. Introduction to fish reproduction. In: Kime, D.E. (Ed.), Endocrine
disruption in fish. Kluwer Academic Publishers, Massachusetts, p. 81-101.

Kishida, M., Callard, G.V., 2001. Distinct cytochrome P450 aromatase isoforms in
zebrafish (Danio rerio) brain and ovary are differentially programmed and
estrogen regulated during early development. Endocrinology 142, 740-750.

Kobayashi, Y., Kobayashi, T., Nakamura, M., Sunobe, T., Morrey, C.E., Suzuki, N.,
Nagahama, Y., 2004. Characterization of two types of cytochrome P450
aromatase in the serial-sex changing gobiid fish, T rimma okinawae. Zool. Sci. 21,
417-425.

Kong, R., Giesy, J.P., Wu, R., Chen, E., Chiang, M., Lim, R, Yuen, B., Yip, B., Mok, H.,
Au, D. 2008. Development of a marine fish model for studying in vivo molecular
responses in ecotoxicology. Aquat. Toxicol. 86, 131-141.

Lichter, P., 1997. Multicolor fishing: What's the catch? Trends Genet. 13, 475-480.

Liu, Z., Wu, F., Jiao, B., Zhang, X., Hu, C., Huang, B., Zhou, L., Huang, X., Wang, Z.,
Zhang, Y., Nagahama, Y., Cheng, C.H.K., Wang D., 2007. Molecular cloning of
doublesex and mab-3-related transcription factor 1, forkhead transcription factor

88

gene 2, and two types of cytochrome P450 aromatase in southern catfish and their
possible roles in sex differentiation. J. Endocrinol. 194, 223-341.

Nunez, B.S., Applebaum, S.L., 2006. Tissue- and sex-specific regulation of CYP19al
expression in the Atlantic croaker (Micropogonias undulates). Gen. Comp.
Endocrinol. 149, 205-216.

Otsuka, F., 2002. Gene expression assay for hazard assessment of chemicals. Ind. Health.
40, 113-120.

Park, J.W., Hecker, M., Murphy M.B., Jones P.D., Solomon, K.R., Van Der Kraak, G.,
Carr, JA., Smith, E.E., du Preez, L., Kendall, RJ., Giesy, J.P., 2006.,
Development and optimization of a Q-RT-PCR method to quantify CYP19a
mRNA expression in testis of male adult Xenopus laevis: Comparison with

aromatase enzyme activity. Comp. Biochem. Physiol., Part B: Biochem. Mol.
Biol. 144, 18-28.

Pemthaler, A., Amann R., 2004. Simultaneous fluorescence in situ hybridization of
mRNA and rRNA in environmental bacteria. Appl. Environ. Microbiol. 70, 5426-
5433.

Peterson, 8., McCrone, S., 1993. Use of dual-label in situ hybridization histochemistry to
determine the receptor complement of specific neurons. In: Valentino, K.L.,
Eberwine, J.J., Barchas, J .D. (Eds), In situ hybridization applications to
neurobiology. Oxford University Press, New York, pp. 78.

Rotchell, J .M., Ostrander, GK, 2003. Molecular markers of endocrine disruption in
aquatic organisms. J. Toxicol. Environ. Health, Part B. 6, 453-495.

Sanderson, J .T., Letcher, R.J., Heneweer, M., Giesy, J.P., van den Berg, M., 2001.
Effects of chloro-S-triazine herbicides and metabolites on aromatase (CYP19)
activity in various human cell lines and on vitellogenin production in male carp
hepatocytes. Environ. Health Perspect. 109, 1027-1031.

Sanderson, J .T., Seinen, W., Giesy, J .P., van den Berg, M., 2000. 2-chloro-S-triazine
herbicides induce aromatase (CYP-l9) activity in H295R human adrenocortical
carcinoma cells: A novel mechanism for estrogenicity. Toxicol. Sci. 54, 121-127.

89

Simeone, A., 1999. Detection of mRNA in tissue sections with radiolabelled riboprobe.
In: Wilkinson, D.G. (ed.), In situ hybridization: A practical approach (2nd Ed).
IRL Press, Oxford, pp. 69-84.

Simpson, E.R., Mahendroo, M.S., Means, G.D., Kilgore, M.W., Hinshelwood, M.M.,
Grahamlorence, S., Amameh, B., Ito, Y.J., Fisher, C.R., Michael, M.D.,
Mendelson, C.R., Bulun, SE, 1994. Aromatase cytochrome-P450, the enzyme
responsible for estrogen biosynthesis. Endocr. Rev. 15, 342-355.

Streit, A., Stem, C., 2001. Combined whole-mount in-situ hybridization and
immnunohistochemistry in avian embryos. Methods 23, 339-344.

Sun, L., Zha, J ., Spear, P.A., Wang, Z., 2007. Toxicity of the aromatase inhibitor
letrozole to Japanese medaka (Oryzias latipes) eggs, larvae and breeding adults.
Comp. Biochem. Physiol., Part C: Toxicol. Pharmacol. 145, 533-541.

Suzuki, A., Tanaka, M., Shibata, N., 2004. Expression of aromatase mRN A and effects of
aromatase inhibitor during ovarian development in the medaka, Oryzias latipes. J.
Exp. Zool. 301A, 266-273.

SzOllOSi, J ., Lockett, S.J., Balazs, M., Waldman, RM., 1995. Autofluorescence correction
for fluorescence in situ hybridization. Cytometry 20, 356-361.

Tagliaferro, P., Tandler, C.J., Ramos, A.J., Pecci Saavedra, J ., Brusco, A., 1997.
Immunofluorescence and glutaraldehyde fixation. A new procedure based on the
Schiff-quenching method. J. Neurosci. Methods. 77, 191—197.

Tompsett, A.R., Park, J.W., Zhang, X., Jone, P.D., Newsted, J .L., Au, D.W.T., Chen,
E.X.H., Yu, R., Wu, R.S.S., Kong, R.Y.C., Giesy, J.P., Hecker, M., 2008.
Development and validation of an in Situ hybridization system to detect gene
expression alogn the HPG-axis in Japanese medaka, Oryzias latipes. Aquat.
Toxicol. (Submitted).

Villeneuve, D.L., Knoebl, I., Kahl, M., Jensen, K., Hammermeister, D., Greene, K.,
Blake, L., Ankely, G., 2006. Relationship between brain and ovary aromatase
activity and isofonn-specific aromatase mRNA expression in the fathead minnow
(Pimephales promelas). Aquat. Toxicol. 76, 353-368.

90

Wang, X.G., Orban, L., 2007. Anti-mullerian hormone and l l beta-hydroxylase Show
reciprocal expression to that of aromatase in the transforming gonad of zebrafish
males. Dev. Dyn. 236, 1329-1338.

Wilkinson, DG., 1999. The theory and practice of in situ hybridization. In: Wilkinson,
D.G. (ed.), In situ hybridization: A practical approach (2nd Ed). IRL Press, Oxford,
pp. 1-21.

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

Zhang, X., Hecker, M., Park, J .W., Tompsett, A.R., Newsted, J .L., Nakayama, K., Jones,
P.D., Au, D., Kong, R., Wu, R.S.S., Giesy, J.P., 2008a. Real time PCR array to
study effects of chemicals on the Hypothalamic-Pituitary-Gonadal axis of the
Japanese medaka. Aquat. Toxicol. (Submitted).

Zhang, X., Park, J.W., Hecker, M., Tompsett, A.R., Jones, P.D., Newsted, J.L., Au, D.,
Kong, R., Wu, R.S.S., Giesy, J.P., 2008b. Time-dependent transcriptional profiles
of hypothalamic-pituitary-gonadal (HPG) axis in medaka (0. latipes) exposed to
fadrozole and l7beta-trenbolone. Aquat. Toxicol. (Submitted).

91

CHAPTER 3

Effects of ethinylestradiol and trenbolone on histology and gene expression of Japanese
medaka (Oryzias latipes) using a combination Of fluorescence in situ hybridization

(FISH) and traditional histology

Abstract

Short—term effects of Wot-ethinylestradiol (EE2) and 17B-trenbolone (TB) on the
hypothalamus pituitary gonadal (HPG) axis of male and female Japanese medaka
(Oryzias latipes) were determined by use of a combination of traditional histology (H&E
Staining), and fluorescence in situ hybridization (FISH). Four-month-old medaka were
exposed to EE2 at 5, 50, or 500 ng/L or to TB at 50, 500, or 5,000 ng/L for 7 d in static
renewal exposure system and effects on expression of vitellogenin II (Vit II), androgen
receptor (AR), and gonadal aromatase (CYP19a) were determined. Exposure to either
500 ng EE2/L or 5000 ng TB/L resulted in significantly less fecundity relative to the
controls as well as histological changes in ovary and testis. Greater intensity of staining
of hepatocytes with hematoxylin was observed in males exposed to EE2 and was
correlated with expression of the Vit II gene. This result suggests that greater staining of
cells, which is related to greater amounts of genetic material, suggesting that the degree
of staining with hematoxylin in hepatocytes can be used to examine the mechanisms of
action for EE2 in fish. FISH used in conjunction with whole fish sections not only
allowed for simultaneous determination of gene expression in multiple target tissues, but

also revealed tissue-and/or gender-Specific differences in the abundance of target mRNAs.

92

Expression of the Vit II gene was typically detected in liver and ovary of unexposed
females, but little expression was observed in liver and testis of unexposed males.
However, in male fish exposed to EE2, expression of the Vit II gene was up-regulated in
both testis and liver relative to that of controls. There was little expression of the AR
gene observed in ovary or liver of control females. Exposure of females to TB
significantly up-regulated AR gene expression in ovary but did not alter the AR gene
expression in liver. In ovary, expression of aromatase (CYP19a) was primarily
associated with early stage oocytes. The synthetic estrogen (EE2) caused up-regulation
of expression of CYP19a at concentrations less than 50 ng EE2/L, but exposure to 500 ng
EE2/L caused down—regulation. FISH was able to localize changes in gene expression at
the cellular/tissue level and provided useful spatial information on changes in these genes

in a whole fish model.

Introduction

To identify the molecular mechanisms of toxic action of a chemical, it is necessary to
detect and quantify the expression of mRNAs that encode for proteins involved in key
processes of the pathway of interest. Most of the techniques utilized have been based on
reverse transcription-polymerase chain reaction (RT-PCR), Northern blotting, or RNase
protection assays. These methods rely on either high yields of RNA extraction from bulk
tissues or only focus on a few genes in one tissue at one specific time in the development
of an organism. However, these methods are limited when used with small animal model
species such as Japanese medaka (Oryzias latipes) or zebrafish (Danio rerio) that are

commonly used in the assessment of effects and risks resulting from the exposure to EDC

93

chemicals. The small amounts of individual tissues available for study and the difficulty
in excising these tissues from these small organisms has limited the efficacy of these
techniques to determine effects during critical windows of time during ontogenesis or
Simultaneously in multiple tissues. Therefore, innovative, sensitive, and flexible
techniques are needed that allow for the identification of multiple molecular target genes
after chemical exposure in multiple tissues simultaneously at any stage of ontogeny.

One promising tool is fluorescence in situ hybridization (FISH) (Park et al.,
2008; Tompsett et al., 2008; Zhang et al., 2008). This technique allows for the direct
visualization of specific mRNA sequences in tissues, individual cells, and/or cell
structures. FISH has been shown to have the potential to be a powerful tool by
combining molecular biology with histology to evaluate gene expression associated with
specific cell types in a tissue (Wilkinson, 1999; Hayat, 2004; Park et al., 2008). Principle
underlying FISH is that molecular probes of complimentary mRNA sequences attached
to a fluorescent molecule can be hybridized to mRNA in tissues or cells of fixed tissue.
Measuring spatial and temporal changes in gene expression as a consequence of chemical
exposure can provide information relative to modes of action within tissues as regulation
of genes among cell types and tissues. Localization of Specific genes at the tissue or
cellular level can help to further our understanding of gene expression patterns in context
with pathologies as determined by parallel histology.

In a previous study, a FISH protocol was optimized to detect spatial expression
of mRNA in whole mount sections of Japanese medaka and validated (Park et al., 2008).
Here I report the results of a study in which the optimized FISH method was used to

evaluate effects of two model EDCS, the synthetic estrogen, ethinylestradiol (EE2) and

94

the synthetic androgen Nil-trenbolone (TB). These two compounds have known and
fairly Specific modes of endocrine action. EE2 is an analogue to the endogenous estrogen
(”B-estradiol), is a strong estrogen receptor (ER) agonist (Lee et al., 2002), and
represents one of the most potent xenoestrogens known to be present in the aquatic
environment (Rotchell and Ostrander, 2003). The US. Geological survey found EE2 in
several surface waters of the United States at concentrations as great as 0.831 11g EE2/L
(Kolpin et al., 2002). Trenbolone is the product of the hydrolysis of trenbolone acetate,
which is a synthetic androgen that is a mammalian androgen receptor (AR) agonist and is
used as a growth promoter for cattle in the USA (Blankvoort et al., 2001; Wilson et al.,
2002; Durhan et al., 2006). TB has the potential to adversely affect aquatic organisms
due to its relative long half-life in water and soil (Schiffer et al., 2001), and its potential
to cause disorders in reproductive endocrine functions in fish including the
masculinization of females (Davis et al., 2000; Ankley et al., 2003; Oms et al., 2006;
Seki et al., 2006).

The Japanese medaka is a common model species in the field of endocrine
disrupter research (Wittbrodt et al., 2002), and was chosen in this study because of the
information available on its physiology, embryology and genetics and the fact that it is
small and easy to rear and induce to reproduce under laboratory conditions.

The model genes selected for this study, Vitellogenin II (Vit II), androgen
receptor (AR) and gonadal aromatase (CYP19a) are considered important biomarkers for
the exposure to EDCS, specifically androgens and estrogens, and thus were thought to be
useful as examples of the capabilities of FISH in whole fish histological sections.

Vitellogenin is under strict control of estrogen, and is found at only very low

95

concentrations in male fish under normal physiological circumstances. It can serve as a
marker for both exposures of males to estrogens and exposure of females to anti-
estrogens or other maturation inhibiting compounds (Kime et al., 1998). The key enzyme
responsible for the conversion of androgens to estrogens is aromatase. The products of
this reaction, specifically l7B-estradiol (E2), are critical in ovarian development,
reproductive function, and sexual differentiation, so that disruption of either activity or
production of the enzyme could alter developmental or reproductive processes of an
organism. Due to its key function in estrogen synthesis, aromatase has been considered
an important biomarker to assess the exposure of EDCS (Rotchell and Ostrander, 2003).
AR is a nuclear receptor that is activated by binding of androgens. It functions as a
transcription factor to regulate androgen specific gene expression. Thus, binding of AR
with xenoandrogens could interfere with processes such as normal male or female
gonadal development.

The objective of these studies was to use FISH to characterize the effects of EE2
and TB on the tissue specific expression of CYP19a, vitellogenin II, and androgen
receptor a in whole mount sections of medaka. Furthermore, the changes in gene
expression were compared to a series of histological, physiological, and/or organismal
responses to further our understanding of the molecular mechanism of action of EE2 and

TB.

96

Materials and Methods

Test chemicals

Chemicals used in this research were EE2 (Ha-ethinylestradiol; l701-Ethynyl- 1 ,3,5(10)-
estratriene-3, l 7B-diol l9-Nor-l ,3,5(1 0),l7a-pregnatrien-20-yne-3,l 7-diol; Si gma-Aldrich,
St. Louis, MO, USA) and l713-trenbolone. TB (l713-Hydroxyestra-4,9,l l-trien-3-one;
Sigma-Aldrich, St. Louis, MO, USA). Dimethyl sulfoxide (DMSO) was used as a carrier

solvent to deliver EE2 and TB treatments in water at a final concentration of 0.01 % (v/v).

Culture ofJapanese medaka (0. latipes)

Japanese medaka were obtained from the aquatic culture at the US EPA Mid-Continent
Ecology Division (Duluth, MN, USA). The medaka were held in flow through systems
under conditions to facilitate breeding (23-24 °C, 16:8 light/dark). Medaka were fed
Aquatox flake food (Aquatic Ecosystems, Apopka, FL, USA) ad Iibitum once daily, and
brine shrimp (Artemia) twice daily. All procedures used during all phases of this study
were in accordance with protocols approved by the Michigan State University Instituted

Animal Care and Use Committee (IACUC).

Chemical exposures

Prior to initiation of exposures experiments, 12- to l4-wk old medaka were placed into

IO-L tanks with 6-L of carbon filtered tap water and acclimated for 12 d under the same
conditions as in the subsequent exposures. Each treatment group consisted of triplicate
tanks, and each tank contained 5 male and 5 female medaka. After the acclimation

period, medaka were exposed to 5.0, 50, or 500 ng EE2/L or 50, 500, or 5,000 ng TB/L.

97

Carbon filtered tap water containing equal amount of DMSO than the chemical treatment
groups served as a control. Every day during the exposure phase of the study, one half of
the water in each tank (3 L) was replaced with fresh carbon filtered water dosed with the
appropriate amount of EE2 (5 mg/L in DMSO) or TB stock (50 mg/L in DMSO). Eggs
produced during the past 24h were counted before the replacement of water. Medaka
were fed Aquatox® (Aquatic Ecosystems, Apopka, FL, USA) flake food and brine
shrimp once daily, and the tanks were kept at 24 °C and 16:8 light/dark. Water quality
parameters (temperature, pH, hardness, dissolved oxygen, ammonia nitrogen, and nitrate
nitrogen) were measured daily and values were within a normal range for water quality.
After exposure for 7 d, medaka were euthanized in Tricaine S (50 mg/mL) (Western
Chemical, Femdale, WA, USA), weighted and snout length was measured. Medaka were
separated into two groups; one group was for used in FISH analysis as described in this
study and the second group was used in a different study utilizing Q RT PCR analysis
that is described elsewhere (Zhang et al., 2008). Medaka designated for analysis by use
of FISH, were fixed as described below. To determine the effects of chemicals on
hepatic and gonadal growth, liver and gonads were weighted and the liver somatic index

(LSI) and gonadal somatic index (GSI) were derived (Equations 1 and 2).

LS1 = (liver weight/body weight)* 100 (l)
GSI = (gonad weight/body weight)*100 (2)
[SH procedure

1. Preparation of sections

98

Sections were prepared for FISH by use of methods adapted from Park et al. (2008), and
Kong et al. (2008). In brief, fish were dissected to remove fins, tail, Skull roof, otoliths,
and opercula. The body cavity was Opened to improve the penetration of fixative (80%
Histochoice MB (EMS, Hatfiled, PA, USA), 2% paraformaldehyde, and 0.05%
glutaraldehyde) for better internal organ fixation. Medaka were then immersed in the
fixative, and allowed to fix over night at room temperature. Fixed medaka were washed
with methanol and dehydrated through a graded methanol series, and then cleared in
chloroform at 4 °C. Fixed and cleared samples were infiltrated with melted Paraplast
Plus paraffin (McCormick Scientific, St. Louis, MO, USA) and the resulting paraffin
blocks were stored under RNAse free conditions at 4 °C until sectioning.

Medaka were sectioned on a rotary AO-820 microtome (American Optical,
Buffalo, NY, USA) under RNAse free conditions using absolute ethanol and RNase-Zap
(Sigma—Aldrich, St. Louis, MO, USA). Serial sections were cut at 7 pm and placed on
Superfrost Plus slides (Erie Scientific, Portsmouth, NH, USA). Slides were stored in

RNase-free containers at room temperature until used for ISH.

2. Fluorescence labeled riboprobe synthesis

All procedures to synthesize fluorescence labeled riboprobes were adapted from the study
of Park et al. (2008) with minor modifications. To synthesize the riboprobes, reverse-
transcribed first-strand cDNA was used as a template in a conventional PCR with
corresponding primers to amplify PCR products of CYP19a, vitellogenin II ( Vit II), and
androgen receptor-01 (AR) (Table 3.1). Probes for FISH were designed using the Beacon

Designer 2 (PREMIER Biosoft Int., Palo Alto, CA, USA) to have lengths Of

99

 

U<U<<OUOFH<UOHU<HPU<DFH

 

. . .. - - 5 . 5:0wo 0::

00R: o»? m Siva or: Emile? UhUH<UUFH<UO<UH<UUH<U<U : . : .>
- 1 UH<UU<U<<OUO<<<HUUH<UU d

.\ 6

oo. mm 0?? 80:00 CV 3839:: 5H0<OU<<O<<UO<OUOHO -8508: :0wo:::<
UHOH<O<OOHHOHUUO<O<<O
a
09mm onm Diva :34 woomwm C<UHO<OOPUHODH<<HHOHUU o _ m>U
30m .05: Dov 30m .05: Do: A000 .058
:o_m:0:xm w::w0:3< U6: w5:3:m:0Q 05m 8:53: A.m-.m 8805580808 .0 :05 : 3 0:35

 

85:50 M556 8023:3304 8880030..

 

.80: 33550530 :8 30:50:00 @508 :5: .05m 50:95:: 8:53: 358000.: 0:58:00 805:: :53 8:05 _.m 2:3: :1

.:8:0:5 .:o 0:0w Ema: 0:: :8

5080 :58 83 3:0:0ono: 00:038.: 0: 3:: 355005 08:83 5 8:0w :305 :6 808338 2:: :53 30:39:00 83 $485

:0w:m: w5::oa8::8 80:0: 0: 0:288: :08 :0 00:03.38 0: :1 5038038: .038: 0:5 3030:0880 35:30:08: 32 :o 5:885:

:0w:m: 0: 3:21:88 32 8:50 0: 03: 3:80: 5:56 8:83 03m 85 8:83 3: COO: 8:: 0:05: w:o_ 00: :o E: 9: 30::

m8: to:m 00: 8:50 8:: 30:58: :5: 833: 585053 :0 3030: a :o :08: 88:0 83 ::w:0_ 0:83 .3: cow 30853859:

100

The method used to clone the PCR product into pGEM T-Easy vector (Promega,
Madison, MI, USA) has been described previously (Park et al., 2008). In order to
synthesize the sense probes, their corresponding cloned plasmids were digested with Sail
(Invitrogen, Carlsbad, CA, USA) for CYP19a and Vit II gene, and Spel (New England
Biolabs, Ipswich, MA, USA) for AR gene. For antisense probes, cloned plasmids were
digested with Ncol (Invitrogen, Carlsbad, CA, USA) for CYP19a and Vit II, and SacII
(New England Biolabs, Ipswich, MA, USA) for AR gene. I confirmed the digestion with
electrophoresis on agarose gel by observing single band with size of plasmid plus
inserted PCR product (data not shown). Sense and antisense riboprobes were synthesized
using in vitro transcription and labeled with fluorescence dye (Alexa Flour 488,

Molecular Probes, Eugene, OR, USA), as described in the method of Park et al. (2008).

3. Fluorescent in situ hybridization

A brief summary of the steps in the ISH is provided (Fig 3.1). The procedures of ISH
and washing steps were in accordance with the methods depicted in Park et al. (2008)
with minor modification to improve probe specificity to the mRNA sequence of interest.
Microtome sections of medaka were hybridized with the riboprobe (1.5 ng/uL for AR and
Vit II, and 2 ng/pL for CYP19a) at 43 °C for 17 h. To evaluate the specificity of binding
of antisense probe there were sections that received equal amount of sense probe.
Expression of the Vit II gene was measured in male and female medaka exposed to EE2
and AR gene expression was measured in female medaka exposed to TB. CYP19a gene

expression was measured in female medaka exposed to both EE2 and TB.

101

Deparaffinization and rehydration

l

Removal of autofluorescence

1

Sections permeabilization

I

DNase treatment

J

Acetylation

l
Prehybridization

J
Hybridization

l

Posthybridization wash and mounting

I.

Confocal microscope analysis

Figure 3.1 Sequence of steps of mRNA in situ hybridization with fluorescence labeled
riboprobe.

4. Confocal laser scanning microscopy (CLSM) image analysis

Distribution of the fluorescent probes were identified and quantified by use of confocal
fluorescence microscopy. Expression of genes was detected and quantified by use of the
LSM 510 Meta system (Carl Zeiss, 07740 Jena, Germany) as described by Park et al.
(2008). To account for background auto-fluorescence due to tissues and/or components
of the hybridization procedure, individual spectral components associated with auto-
fluorescence, background, and Alexa Flore 488 dye were separated using the confocal
system. Auto-fluorescence and background spectral components were obtained from the

section of each tissue (ovary, testes, and liver) on the slide hybridized without probe. The

102

specific spectrum of Alexa Flour 488 dye was obtained directly from dye reagent. Once
defined, the number of significant and independent sources of the specific spectral
components using CLSM was then subjected to linear spectral unmixing to separate the
individual components and to remove auto-fluorescence signal in the recorded images to
obtain the specific fluorescence signal of Alexa Fluor 488 dye. To avoid the alteration of
the auto-fluorescence spectral shape by photo-bleaching that can result from consecutive
laser scans, each set of ISH experiments had at least one ISH section with sense probe for
antisense probe specificity and one ISH section without probe for the separation of
spectral components.

Images of the tissues were collected at 10X magnification. Expression of the
CYP19a gene was quantified by randomly selecting 3 different areas of ovary sections
hybridized with CYP19a antisense probe and then quantifying the fluorescence in 3 early
stage of oocytes (less than 100 pm in diameter) randomly selected in each area. Based on
morphological criteria developed by Iwamatsu et al. (1988), those oocytes were classified
as being in the previtellogenic stage (less than 200 pm) of oogenesis. Previtellogenic
oocytes were classified as being primary oocytes (less than 100 pm) or pre-vitellognic
oocytes (100 - 200 pm) according to their size. Fluorescence intensity was divided by
the total area of the collected oocytes to compensate variation caused by the size of
oocytes. Gene expression in testis and liver, was measured in a randomly selected area of
tissue (approx. 1/4 of entire testis) and two areas of tissue (each approx. 1/10 of entire
liver), respectively. To measure gene expression in the ovary, the same method as
described for the quantification of CYP19a gene expression was followed. The

fluorescence of testis and liver, was collected from the whole tissue in the image, and

103

then divided by the area across which the signal was measured. Also, due to the
detection of small amounts of Alexa Flour dye specific signal in the sections hybridized
with sense probe, the antisense signal was normalized to sense signal in each ISH

experiment.

Histology

Histological changes in medaka ovaries were evaluated using H & E stained sections
(Tompsett et al., 2008). Briefly, slides were de-paraffinized in xylene and rehydrated
through a descending ethanol series (100, 95, and 70%). Slides were then stained in
Harris’ hematoxylin (EMS, Hatfield, PA, USA) for 3 min, processed through acid alcohol,
ammonia, and ethanol washes, and then stained in 1% Eosin Y (EMS, Hatfield, PA,
USA) in 80% ethanol for l min. Slides were then dehydrated through an ethanol series
(70, 95, and 100%) and cleared in xylene. Slides were preserved under glass cover slips
using .Entellan mounting medium (EMS, Hatfield, PA, USA) and allowed to dry. Images
of the gonad and liver on each slide were recorded using a Nikon Eclipse TE3OO
microscope with image software (SPOT, Diagnostic Instrument, Inc). The intensity of
staining of hepatocytes with hematoxylin, a measure of the amount of genetic material
present in the tissue, was measured with image analysis software (Image J 1.38X,
National Institute of Health, USA). Briefly, digitized images of livers were segmented to
obtain the purple color on the image by setting the Hue histogram to 212 - 255, which
represents the genetic materials in the cell stained with Hematoxylin. Purple stains

greater than 400 pixels in size were enumerated.

104

Statistics

Statistical analyses were conducted using SAS (SAS Institute Inc. Cary, NC, USA). Data
sets were tested for normality using the Shapiro-Wilk’s test, and were log-transformed if
necessary to achieve normality. Statistical differences between treated groups were
determined using one way analysis of variance (ANOVA), followed by the Student-
Newman-Keuls test for multiple comparisons. For comparison of means of two groups,
the Student’s t test was applied. The relationship between the degree of staining of
hepatocytes with hematoxylin and expression of Vit II gene as measured by FISH, was
investigated by use of non—parametric 2-tailed Spearman rank correlation. The criterion

for significance in all statistical tests was p<0.05.

Results
Weight, length, biological indices and fecundity
The wet weight and length of medaka exposed to either EE2 or TB were not significantly
different from those of control medaka (Table 3.2). In female medaka, EE2 did not
significantly affect LSI or GSI, while in male medaka exposure to all concentrations of
EE2 resulted in statistically greater LSIs compared to the controls (Fig 3.2A). TB caused
a statistically significant effect on LSI at concentrations greater of equal to 50 ng TB/L in
females while the concentration required to cause a statistically significant effect on GSI
was 500 ng. For males, the only statistically significant effect was observed for LSI after
exposure to 500 ng TB/L (Fig 3.2B).

Production of eggs was approximately 48% less in medaka exposed to 5000 ng

EE2/L than in the controls. However, this difference was not statistically significant (Fig

105

.God A 3: m3. ::0 mm ::0: E 0::0E:00:: w:0E0 :0: 30 ::w:0_ ::0 £903 30 00:88:: ::00_.::w:m 0Z ...
.::w:0_ 50:: 0: :0:00:: 0:: E :0:E=: 0:: .Emm a :02: ...
.OmEQ ::3 :0:00:: 003 3:0:w _0:::00 ::0>_0m ...

 

20.000000:

Aowdaowomv Ammoamwom: Ammdfiwmhmv 3::anva Godanwémv 33:05.8:

 

0 0E0
868on vodfiwmd vodaomd modflmmd modfimmd modi md modfi md _ 3
2:603on defimodmv $2.330.an $0085.08 GYOHNQVNV Anvdamwfimv 2.0600363 0.02
.odahmd Nodfiofio 8.3”de modfiumd Sdfinmd modammd Sdfiomd
Good com ow cam ow m

 

0:::0U ::0>_0m
3:03 05 3:80 05 _

 

2:30 0::: E :00: 0:0:08 000:030a 30 ASE: ::w:0_ ::0 Am: ::m:03 >:0m Nam 030.0

4%: w: 000m 0: 0500308 0:: 30 >0: ::000m 0:: :080 :00000 30:29:00 :m0E_0 :0_:0=:0:3 wwm

.Ammgm wr: 30.500300: JR: w: ooom :0 cow 0: :0m03x0 0:0:0E E 000— $3.0 ::0 o: 003 :0::0::0:3 ww0 0>::0_:E:U .A<m.m

lOé

 

 

 

I GSI o LSI 16 I GSI o LSI
6 male female
4 12 . T r r
8 . l
2 l ‘i l
0 0 i '
CTR 5 50 500 CTR 5 50 0
EE2 exposure (ng/L) EE2 exposure (ng/L)
B
6 male - GSI a LSI 20 female ' GSI D L31
4 I 16
12
2 ii 8
il {:13 4
o . l 0
500 5000 500 5000
TB exposure (ng/L) TB exposure (ng/ L)

Figure 3.2 Mean values (:t SEM) of liver somatic index (LSI) and gonado-somatic index
(GSI) in Japanese medaka exposed to EE2 (A) or TB (B). Significant differences relative
to the control are indicated with an asterisk (p < 0.05, n = 4 ~ 6).

 

 

 

 

 

U, A B
O0
30 1500 __.~ control 1500 - -,_ control
‘H _ EE2 5 n L l
0 1200 + g/ 1200 q 4: TB 50 ngL I
53 "U ——.— EE2 50 ng/L ——0- TB 500 ng/L ,
E 8 900 + EE2 500 ng/L ..~' 900 ~u~ TB 5000 ng/L ,.;:_11..}
c m " 1', I
2 g 600 600 - If... r.
‘0' C; ---...~»~.°‘ I"
g 300 300 v/"’:, ‘43”: ;._—n~———:s—-—n— II A
:3
Q 0 () . .

0 l 2 3 4 5 6 7 0 l 2 3 4 5 6 7

Exposure (Day) Exposure (Day)

Figure 3.3 Cumulative numbers of fertilized eggs spawned by male and female Japanese
medaka exposed to EE2 (A) or TB (B). Each treatment consisted of triplicate tanks, and
each tank contained 6 pairs of medaka. Significant differences relative to the control are
indicated with an asterisk (p < 0.05)

107

Histology ofmedaka exposed to EEZ or TB

Both testes and ovaries of control medaka exhibited complete gonadal differentiation and
maturation characteristic for this species (Figs 3.4A and 3.48). Testicular tissue was well
developed and all types of germ cells were present with Spermatogonia (SG) and
sperrnatocytes (SC) located in the periphery and Spermatids (ST) and spermatozoa (SZ)
in the center of the testis. All stages of oocyte maturation were present in the ovary,
including primary oocytes (PR), previtellogenic oocytes (PO), vitellogenic oocytes (V0),
mature oocytes (MO) that had a clearly expressed yolk globule, and visible corpora lutea.

Exposure to EE2 caused changes in both the ovary and testes. All types of germ
cells were present in testes of males exposed to 500 ng EE2/L, but spermatozoa were
starting to degrade and there was a greater proportion of connective tissue present.
Testicular ovarian follicles (perinucleus stage) were observed in the testes of only one
male medaka exposed to 500 ng EE2/L (Fig 3.4C). Exposure of females to EE2 resulted
in alterations of the composition of cells with different maturation stages in the ovary.
There were fewer mature oocytes, more atretic oocytes, and larger volume of somatic
stromal tissue (Fig 3.4D).

Exposure to TB resulted in profound effects on the ovary, but few effects on
testes. The only effects of TB on the histology of testes were observed in medaka
exposed to 5,000 ng TB/L. While all types of germ cells were present in the testes of
medaka exposed to this concentration, there was an accelerated development of
spermatozoa such that fewer Spermatogonia were observed compared to the controls (Fig
3.4E). Exposure of females to 5,000 ng TB/L resulted in the predominant cell type being

mature oocytes and there were fewer previtellogenic oocytes present compared to

108

controls (Fig 3.4F). Most of the mature oocytes exhibited signs of disruption of yolk
accumulation and increased yolk vacuolization.

EE2 caused changes in the histology of livers of both males and females. Liver
tissue of control males was mainly stained with eosin which stains extracellular or
intracellular proteins in the cytoplasm of hepatocytes. After exposure to 500 ng EE2/ L,
there was a shift in the staining of hepatocytes in livers of males towards predominantly
hematoxylin, which is an indicator of the presence of nucleic acids (Fig 3.5). The
number of stains with hematoxylin in livers of males exposed to 500 ng EE2/L was
significantly greater than in livers of unexposed males. There was no statistically
significant difference in numbers of hematoxylin stains between livers of male exposed to
EE2 and that of control females (Fig 3.5D). Expression of Vit [1 mRNA in liver of males,
as measured by FISH, was directly proportional to the number of hematoxylin-stained

stains (r2 = 0.821, n = 23, p < 0.05).

lO9

 

 

 

 

 

 

 

 

Figure 3.4 H & E stained cross section of gonads of Japanese medaka. (A) testis of
control (lOOX, bar = 20 um). SZ: spermatozoa, ST: spermatid, SC: sperrnatocyte, and
SG: spermatogonia. (B) control ovary (40X, bar = 50 um). PR: primary oocyte, PO:
previtellogenic oocyte, VO: vitellogenic oocyte, and MO: matured oocyte. (C) testis of
male exposed to 500 ng/L of EE2 (200x, (40 and 400K in the boxes) bar = 10 pm).
Testis-ova observed in the form of perinuclear stage, degrading spermatozoa and greater
proportion of connective tissue were observed. (D) ovary of female exposed to 500 ngL
of EE2 (40X, bar = 50 pm). AO: atretic oocyte, and SST: somatic stromal tissue. There
were fewer mature oocytes, more atretic oocytes, and larger volume of somatic stromal
tissue. (E) testis of male exposed to 5,000 ng/L of TB (lOOX, bar = 20 pm). Accelerated
spermatozoa development and fewer Spermatogonia were observed. (F) ovary of female
exposed to 5,000 ng/L of TB (40X, bar = 50 pm). Predominant matured oocytes and
fewer previtellogenic oocytes were observed.

110

 

2000 -

 

1600 .
U)
*6
% 1200 .
Q:
0
5 800
.D .
E
z!
z
400 -
0 .

 

CTR male E500 male CTR female
Figure 3.5 H & E stained cross section of liver of Japanese medaka. (A) control male
showing eosinophilia, (B) male liver of fish exposed to 500 ng/L of EE2 and (C) female
liver of control Japanese medaka showing intense staining with hematoxylin. (D)
number of hematoxylin-stained stains (mean i SEM). Significant differences relative to
the control are indicated with an asterisk (p < 0.05). Bar = 50 um.
Chemical-induced gene expression changes by ISH
l. Vitellogenin [1 mRNA expression in male and female medaka exposed to EE2
Expression of Vit II mRNA in livers and gonads of both male and female medaka was
detected by use of optimized FISH (Fig 3.6). Exposure to EE2 caused up-regulation of
Vit [1 mRNA in tissues of male and female medaka, with the exception of ovary. In

sections of ovary hybridized with Vit II antisense probe fluorescence intensity signal was

111

specific to previtellogenic oocytes, rather than matured oocytes. Sections hybridized
with sense probe were not fluorescent (Fig 3.6B, left panel). Vit 1] mRNA expression in
ovary of control fish was not significantly different from that in ovary of medaka exposed
to 500 ng EE2/L (Fig 3.88). Although there was no significant difference in location or
specific intensity of fluorescence in sections of testes between EE2-exposed and control
medaka, the average fluorescence intensities of the exposed fish were approximately 2-
fold greater than the controls (Fig 3.8A). Vit [1 mRNA expression was detectable in
female livers, but there was no difference between livers of females exposed to EE2 and
the controls (Fig 3.8C). The fluorescence intensity in livers of control males was not
different from that of sections hybridized with sense probe. Fluorescence from Vit II
antisense probe was present primarily in regions of the testes where Spermatogonia were
developing and rarely in the region of matured spermatozoa. Intensities of fluorescence
were in the following order: Spermatogonia >> Spermatocytes and
spermatid>>spermatozoa) (Fig 3.6A). In livers of males, expression of the Vit II gene
was significantly greater in exposed fish, relative to that of controls (Figs 3.6C and 3.8C).
Expression of the Vit II gene was greatest in livers of males exposed to 500 ng EE2/L and

was comparable to that observed in livers of females (Figs 3.6D and 3.8C).

112

 

Figure 3.6 Expression of vitellogenin II mRNA in the gonads (A and B) and liver (C and
D) of Japanese medaka after hybridization of longitudinal whole mount sections with a
fluorescence riboprobes using optimized ISH. Expression of Vit II mRNA was detected
in the testes (A, bar = 200 pm) exposed to EE2 and control ovary (B, bar = 100 pm) of
fish with hybridization of antisense probe, especially strongly in the region of
sperrnatogonia in testes and primary stage of oocytes in ovary, respectively. Very weak
detected fluorescence signal in the section hybridized with sense probe. Vit II expression
in the male liver (C, bar = 50 um) of Japanese medaka exposed to EE2 (500 ng/L) was as
high as that in the section of female liver (D, bar = 50 pm) hybridized with Vit II
antisense probe. Display channel was set to green for antisense probe labeled with Alexa
F luor 488 and to red for autofluorescence.

2. AR mRNA expression changes in female medaka exposed to TB

While AR mRNA could be detected in both ovary and liver of female medaka by use of
FISH antisense probe (Figs 3.7E and 3.7F), there was no fluorescence in tissues
hybridized with sense probe (data not shown). In the ovary, TB caused dose-dependent
greater expression of AR gene expression. No significant changes in AR gene expression

in liver tissue were observed among concentrations of TB (Figs 3.9A and 3.9B). Overall,

AR mRNA expression in tissues of both control and exposed female medaka was low

113

except for AR mRNA gene expression in ovaries of medaka exposed to 5000 ng TB/L

(Fig 3.9A).

C

c
';

 

Figure 3.7 Expression of CYP19a (A — D) and AR (E and F) in the ovary and liver of
female Japanese medaka using the FISH. Expression of CYP19a was very lowly
detected in the ovary hybridized with sense probe (A), while it was specifically detected
in the cytoplasm of primary oocytes of control ovary (B, bar = 100 um), ovary exposed to
EE2 (C), and ovary exposed to TB (D, bar = 100 um). AR mRNA expression was
detected, but lowly. in the ovary (E, bar : 100 pm) and liver (F, bar I 50 um) exposed to
TB. Display channel was set to green for antisense probe labeled with Alexa Fluor 488
and to red for autofluorescence.

114

3. CYP19a mRNA expression changes in female medaka exposed to EE2 or TB
Hybridization of longitudinal sections whole medaka with CYP19a antisense probe
showed that this gene was predominantly expressed in the ovary (Fig 3.7B). The
specificity of hybridization was demonstrated by ISH using sense probe, and which
resulted in a very weak fluorescence signal (Fig 3.7A). The greatest fluorescence
intensity was associated with premature early stages of oocytes and previtellogenic
oocytes. Expression of CYP19a was less in vitellogenic oocytes, and matured oocytes
cells expressed little or no CYP19a (Figs 3.7B - D). Exposure to concentrations of EE2
as great as 50 ng/L caused a significant and dose-dependent up-regulation of CYP19a
expression in ovary. At the greatest dose, however, this trend was reversed and CYP19a
expression was less than that in ovaries of control females.

Exposure of females to TB resulted in a dose-dependent up-regulation of CYP19a
gene expression in the ovary that was similar to the pattern observed after exposure to
EE2. The effect was dose-dependent to 500 ng TB/L, but less at 5000 ng TB/L. While
these differences were observable, they were not statistically significant due to the

relatively great variation that was observed among replicates (Figs 3.10A and 3.10B).

115

g A 5.0E-05 " 4013-03
3% A 5% ' B
*5 3 405-05 2 3
._ 0 -— 0 3.05-03
G o :: a)
.2 ta 3-05-05 .2 ‘0
§% 20505 g? 205-03 .
5.5.» ’ ' 5.99
V) In
” 0 1.05-05 ° 0 "05‘03
< 8 < E
E2 § 005+00 5 § 005+00
11)
E E CTR 5 so 500 E g CTR 500
T: 2 I _-
E; L: _ EE2 Exposure (mg/L) 5 L: EE2 Exposure (ng/L)
q) A
> NE **
= 1.6E-04 :-
E 3 C **
r: 8
0 125-04
‘5. ED 8.0E-05 * *
5 '72
D
‘2‘: § 405-05
% a
= g 005+00
2; 5 CTR-M 5-M SO-M 500-M CTR-F 500-5

EE2 Exposure (ng/L)

Figure 3.8 Fluorescence intensity of Vit II in testes (A), ovary (B), and liver (C) of
Japanese medaka exposed to EE2. Each bar represents mean i SEM. Significant
differences relative to the control are indicated with an asterisk Q) < 0.05).

 

 

: a?
N _ a 30505
E5 1.25 03 A * ,5 3 B
0 3 .S 0
.~ 0 a
5 5 3-05-04 '5 g 205-05 '
2 Eu 0%?
CLUE 0) a)
5 0 405-04 <3 3 1.0E-05
0
< c: o
v E a
E 2% 5 a
E 5- 005+00- is E 005+oo '
fi 5 CTR so 500 5000 L“ CTR 50 500 5000
TB exposure (ng/L) TB exposure (ng/L)

Figure 3.9 Fluorescence intensity of AR in ovary (A) and liver (B) of Japanese medaka
exposed to TB. Each bar represents mean i SEM. Significant differences relative to the
control are indicated with an asterisk (p < 0.05).

116

 

 

 

Fluorescence signal/area (umz)

 

Fluorescence signal/area (umz)
AR mRNA expression in ovary

AR mRNA expression in ovary

CTR 5 50 500 CTR 50 500 5000
EE2 exposure (ng/L) TB exposure (ng/L)

Figure 3.10 Fluorescence intensity of CYP19a in randomly selected three primary stage
of oocytes in the ovary of Japanese medaka exposed to EE2 (A) and TB (B), and each bar
represents mean i SEM. CYP19a mRNA expression was not significantly different
among treatment in both EE2 and TB exposure (p > 0.05).

Discussion

Optimization of in situ hybridization analysis

Using the optimized FISH method described previously (Park et al., 2008), it was
possible to identify and to quantify expression of the CYP19a, Vit II, and AR genes in
paraffin embedded whole mount sections of Japanese medaka. The optimized FISH
method utilized spectral linear unmixing of the fluorescence signals and sodium
borohydride treatment that significantly reduced background fluorescence. However,
even with these adaptations, there was a weak signal detectable in sections hybridized
with sense probe. Similar background fluorescence using FISH has been observed
previously and has been hypothesized to be either due to incomplete removal of non-
hybridized probes during post-hybridization, or to auto-fluorescence of the tissue and/or

ISH components at the same wave-length than the specific probe signal (Alexa Fluor

117

488; Park et al., 2008). The variation in fluorescence signal limited statistical power.
However, significantly greater fluorescence intensity was observed in sections hybridized
with antisense probe compared to those hybridized with the sense probe, which indicates
the specificity of the developed ISH protocols. This result is similar to those of other
studies using fluorescent or radionuclide-based ISH where relatively great variation
between replicate slides was also observed such that it was difficult to quantify gene
expression by use of the ISH (Park et al., 2008; Tompsett et al., 2008). Thus, in future
studies using FISH, the number of replicates would need to be optimized to demonstrate
statistical significance at a specified power. While this might limit the utility of FISH as
a screening tool with small numbers of replicates, the high resolution and integration into
classical histological analysis renders studied FISH technique a powerful tool for

mechanistic research.

F ecundity, histology, and gene expression of medaka exposed to E52

Exposure to EE2 resulted in fewer eggs produced by Japanese medaka, a result that is
consistent with the results of a study by Scholz and Gutzeit (2000) where exposure of
Japanese medaka 10 or 100 ng EE2/L resulted in reduced egg production and a lesser
GSI than in "unexposed medaka. Production of fewer fertile eggs by medaka exposed to
E2 may be due to impairment of the reproductive system in females or deficient sperm
and/or suppression of sexual behavior in males (Kang et al., 2002). In this study, based
on histological examination, testes from medaka exposed to 500 ng EE2/L were normal,
except for some degeneration of spermatozoa. However, histological examination of

ovaries of female medaka exposed to EE2 revealed fewer oocytes and atrophy of the

118

ovary, which indicates that the reduced fecundity observed at this stage was caused by
impaired oocyte development. However, exposure duration was only 7 d, and it cannot
be excluded that the degenerative changes observed in the testis would have been
progressed over time, and thus, also impacting fecundity due to effects in the males as
well as those observed for the females.

The FISH method used in our study allowed the detection of a specific Vit I]
antisense signal in both the liver and gonads of Japanese medaka sections. In contrast,
Vit II was rarely detected in the testis or liver of control male medaka. This is in
accordance with findings of Wang et al. (2005) and Islinger et al. (2003) who did not
detect vitellogenin in testes or liver of control male zebrafish using northern blotting and
RT PCR analysis, respectively. In this study, exposure to EE2 caused greater expression
of Vit II in both liver and testes of male medaka. In the testis, fluorescence specific to Vit
II was localized in Spermatogonia which are located in the peripheral region of testis. It
has been previously reported that Japanese medaka exposed to exogenous estrogen
accumulated vitellogenin protein, measured by immuno-histochemistry, in the cytoplasm
of sperrnatocytes in the seminiferous tubule, but not in sperrnatogonia (Kobayashi et al.,
2005). The fact that in this study fluorescence of the Vit II mRNA probe was rarely
detected in the sperrnatocytes of testis after treatment of male medaka with EE2 (Fig
3.6A) implies that synthesis and accumulation of vitellogenin in the testis can occur at
different locations.

lip-regulation of Vit II in testis of males exposed to EE2 is in accordance with
findings in another oviparous fish, the zebrafish (Islinger et al., 2003; Wang et al., 2005).

Similarly, xenoestrogens up-regulated vitellogenin mRNA expression in testes of sea

H9

bream (Diplodus sargus) (Pinto et al., 2006) and resulted in greater concentrations of
vitellogenin in testis of rainbow trout (Skillman et al., 2006). Vitellogenin transcription
is activated through binding of an estrogen to the ER. Therefore, a greater number of
ERs in the testis might explain the up-regulation of vitellogenin mRNA expression.
Although the presence of ERs in testis, either at the gene or protein level, was not
quantified in the current study, significant up-regulation of transcript or protein levels of
estrogen receptors after estrogen treatment have been reported in other studies with
medaka (Contractor et al., 2004), goldfish (Carassius auratus) (Nelson et al., 2007), sea
bream (Pinto et al., 2006), and rainbow trout (Skillman et al., 2006). Based on results
obtained using ISH, exposure of tilapia (Oreochromis aureus) to 17B-estradiol (E2)
induced synthesis of vitellogenin mRNA in the testes (Ding et al., 1993). This indicates
the presence of functional ER in the testes. Induction of Vit II mRNAs in testes of E2-
treated male medaka was more evident than in ovary of E2-treated female fish. This
indicates that males were more sensitive to the effects of EE2 than were females.

The ovary of teleost fish has been regarded as the destination of vitellogenin
produced in the liver, while it had been believed that liver is the primary place for
synthesis of vitellogenin (Wahli, 1988). As a consequence, to date most efforts have
focused on the deposition and accumulation of vitellogenin in oocytes, but information
on the synthesis or gene expression of vitellogenin in the gonads is rare. Based on the
optimized FISH method, Vit II mRNA expression was detected in the cytoplasm of
previtellogenic oocytes in both control fish and EE2 exposed fish, and expression of the
Vit II gene ovary of EE2-treated medaka was greater than in control fish. This is

consistent with the findings of studies with spotted ray (Torpedo mamorata) that revealed

I20

active synthesis of vitellogenin in granulosa cells associated with both previtellogenic
and vitellogenic oocytes. This demonstrates that these cells play a role in vitellogenesis
in the ovary (Pisco et al., 2004). Vitellogenin mRNA in ovary of zebrafish, measured by
use of ISH, was reported to be slightly greater in ovary after exposure to 17B-estradiol
(Wang et al., 2005). However, these authors stated that vitellogenin mRNA was
expressed in adipose tissues of ovary, not in oocytes. In the study on which I report here,
fluorescence of the Vit 1] mRNA probe was localized in the cytoplasm of previtellogenic
oocytes. The relevance of vitellogenesis in ovaries of fish is unclear but it may be an
additional source of vitellogenin when hepatic vitellogenesis is disrupted. Another
possibility is that ovarian vitellogenesis could have a supporting function in context with
maturation of the previtellogenic oocyte.

The observation that Vit [1 mRNA expression was not observed in liver of
control males, but once exposed to EE2, Vit II mRNA expression was significantly up-
regulated, is consistent with previous studies, in which exogenous estrogen induced either
vitellogenin mRNA expression or protein level in male liver of teleosts (Zhang et al.,
2002; Islinger et al., 2003; Van der Ven et al., 2003; Scholz et al., 2004; Kobayashi et al.,
2005; Wang et al., 2005; Skillman et al., 2006; Miracle et al., 2006). As observed for the
gonads, there was a gender-specific difference in the sensitivity of response to EE2
exposure as measured by Vit II expression in the liver, with males being more sensitive
than females. The level of Vit II expression in livers of males exposed to 500 ng EE2/L
was similar or greater than that measured in livers of control and exposed females, which
is in accordance with findings in zebrafish (Wang et al., 2005). Although Vit [1 mRNA

expression was detected throughout the entire liver section, there appeared to be a

12]

tendency to slightly greater gene abundance in the outer layer of the liver. This result
suggests that the surface regions are primarily involved with vitellogenesis, which has
also been reported in zebrafish (Wang et al., 2005).

In a previous study, it has been proposed that H & E staining can also be used to
detect changes in mRNA content in cells or tissues (Van der Ven et al., 2003). The
significantly greater number of stains with hematoxylin in liver of males exposed to EE2
is an indication that there was more genetic material such as mRNA in the hepatocytes,
and which was assumed to be related to the greater Vit [1 mRNA production observed
with FISH. The greater number of cells staining with hematoxylin in liver observed this
study was consistent with similar results in zebrafish (Van der Ven et al., 2003). The
intensity of staining with hematoxylin is a function of the greater amount of genetic
material. The greater staining observed in liver in this study as well as similar results
obtained by ISH suggests that the greater incidence of hematoxylin—stained cells in the
liver is useful as a screening method for estrogen stimulation of male fish (Van der Ven
et al., 2003). This hypothesis was confirmed in the present study by the strong
correlation between the two methods (r2 = 0.821, p < 0.05).

As observed in a previous study (Park et al., 2008), CYP19a mRNA expression
as measured by FISH was most prominent in the cytoplasm of early stage of oocytes,
which is consistent with findings in other teleost species, such as killifish (F undulus
heteroclitus) (Dong and Willett, 2008), zebrafish (Goto-Kazeto et al., 2004; Wang and
Orban, 2007), and Atlantic croaker (Micropogom'as undulates) (Nunez and Applebaum,
2006). Although not statistically significant, expression of CYP19a mRNA was directly

proportional to EE2 concentration, except for the greatest concentration where it was less

122

than the controls. The mechanism of EE2 interaction with CYP19a gene expression,
induction at low doses and reduction at high dose, is unknown. There have been studies
that have found that EE2 at relatively low concentrations increased expression of CYP19a
in ovary. In fathead minnows, exposure to 10 ng EE2/L caused up-regulation of CYP19a
gene expression (F ilby et al., 2007), and exposure to 100 ng EE2/L resulted in up-
regulation of CYP19a gene expression in Japanese medaka (Scholz and Gutzeit, 2000).
Other studies have found that exposure to relatively great concentrations of estrogens
resulted in down-regulation of CYP19a gene expression in adult zebrafish (5 ug EE2/L;
Ortiz-Zarragoitia et al., 2006), juvenile zebrafish (~ 30 pg EE2/L; Kazeto et al., 2004),
and embryo zebrafish (~ 270 u g E2/L; Kishida and Callard, 2001). It has been suggested
that the down-regulation of CYP19a mRNA was not controlled by transcription due to
the lack of the ERE on its 5’-flanking region, but that this inhibiting effect was more

likely due to a direct effect of EE2 on gametogenesis (Kazeto et al., 2004).

F ecundity, histology, and gene expression of medaka exposed to TB

Effects of TB on reproductive and related functions were more pronounced than those
observed for EE2, with the two greatest concentrations completely inhibiting egg
production shortly after initiation of the exposure experiments. Similar results have also
been observed in other teleost species such as channel catfish (Ictalurus punctatus)
(Davis et al., 2000) and fathead minnow (Ankley et al., 2003). Histological observations
of the gonads revealed that the lesser fecundity was caused by impairment of gonadal
development in both males and females. Males and females exhibited an increase in the

proportion of spermatozoa and stimulated spermatogenesis, and impairments of

l23

vitellogenesis and oocyte development, respectively. Disruption of vitellogenin
accumulation in the ovary was most likely due to TB causing lesser plasma vitellogenin
concentrations, as has been demonstrated for fathead minnows (Ankley et al., 2003).
Other studies have also found similar histological changes in the testis as were observed
in our study (Ankley et al., 2003; Orn et al., 2006).
The optimized FISH method revealed that AR mRNA expression was induced in

a dose-dependent manner in the ovary but no significant changes were observed in the
liver. This observation is in agreement with those of a previous study in which
expression of AR mRNA in liver, measured by means of Q RT PCR, was not altered
(Zhang et al., 2008). While in our study, a dose-dependent increase in AR mRNA
abundance was observed in the ovary, no comparable effects were observed when AR
was measured using RT-PCR (Zhang et al., 2008). The reasons for this difference are not
clear. Exposure of mosquitofish (Gambusia affinis aflinis) to TB resulted in a rapid
increase in the expression of the AR-a and AR-fl in the anal fin of females (Sone et al.,
2005). That result confirmed that TB has the potential to up—regulate AR gene expression.

The results of the FISH revealed tissue specific detection of CYP19a mRNA that
was specific to the cytoplasm of early stage of oocytes. No expression of C YP1 9a was
observed in brain or liver, which is consistent with the results of a previous study (Park et
al., 2008). This indicates that the ovary is the predominant tissue of estrogen synthesis by
gonadal aromatase. C YP19a mRNA expression in the ovary of fish exposed to TB was
not significantly different from the control fish, but there appeared to be a trend towards
increased abundanceCYP19a mRNA in fish from all TB treatment groups when

compared to the controls. This trend was confirmed by a parallel study utilizing Q RT

124

PCR analysis, which also showed a clear dose-dependent increase of CYP19a gene
expression in fish from the same experiment (Zhang et al., 2008). It has been shown in a
different study that exposure to androgens such as TB can up- and down-regulate CYPl9
gene expression depending on the tissues fish (Filby et al., 2007). Thus, due to its tissue-
and/or species-specificity, the mode of action of TB on CYP19a gene expression still
remains unclear. However, induction of CYP19a mRN A expression in ovary exposed to
TB in this study might be a compensatory response to the decreased levels of plasma
estrogen caused by this chemical (Ankley et al., 2003; Jensen et al., 2006). A comparable
increase in ovarian CYP19a expression has also been observed after the exposure of
medaka to an aromatase inhibitor, fadrozole, a result which suggests a similar
compensatory mechanism in response to decreased endogenous estrogen (Park et al.,
2008; Tompsett et al., 2008).

In summary, the optimized FISH method used in this study was able to detect and
quantify changes in Vit 11, AR, and CYP19a mRNA transcripts in tissues of Japanese
medaka exposed to two common endocrine disruptors, the xenoestrogen, EE2 and the
androgen, TB. The primary strength of the FISH approach was its ability to localize
changes in the expression of target genes at the cellular and/0r tissue level that provided
useful information on spatial changes in gene expression that can be used to further our
understanding of molecular changes at the histological level. However, due to the semi-
quantitative nature of this method and the variability observed in the measured signal,
additional work is needed to optimize these methods. Specifically, additional effort is
needed relative to further reduce background fluorescence that in some instances made it

difficult to evaluate gene expression at the cellular level for low expressed genes.

125

Additional efforts should focus on developing more precise approaches to quantifying
fluorescence signals with tissues or cell types within tissues such that some of the
variability observed in this study is reduced. Regardless of these rather minor
uncertainties, the use of FISH methodology represents a valuable tool to further our
understanding of the molecular mechanisms of action of chemicals and to aid in linking

effects at the molecular level to pathologies.

Acknowledgement

This study was supported by a grant from the US Environmental Protection Agency
Strategic to Achieve Results (STAR) to J .P. Giesy, M. Hecker, and PD. Jones (Project
no. R-831846), an Area of Excellence 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 Hong Kong University Grants Council (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.

126

References

Ankley G.T., Jensen, K.M., Makynen E.A., Kahl, M.D., Korte, J.J., Homung, M.W.,
Henry T.R., Denny, J.S., Leino, R.L., Wilson, V.S., Cardon, M.C., Hartig PC,
and Gray LE. 2003. Effects of the androgenic growth promoter I7-B-trenbolone
on fecundity and reproductive endocrinology of the fathead minnow.
Environmental Toxicology and Chemistry, 22: 1350-1360.

Blankvoort, BM., de Groene, BM., van Meeteren-Kreikamp, AP., Witkamp, RF.,
Rodenburg, RJ., and Aarts, J M. 2001. Development of an androgen reporter gene
assay (AR-LUX) utilizing a human cell line with an endogenously regulated
androgen receptor. Analytical Biochemistry, 298: 93-102.

Contractor, RG., Foran, CM., Li, S., and Willett, KL. 2004. Evidence of gender- and
tissue-specific promoter methylation and the potential for ethinylestradiol-induced
changes in Japanese medaka(01yziaslatipes) estrogen receptor and aromatase
genes. Journal of Toxicology and Environmental Health, Part A, 67: 1-22.

Davis, KB., Morrison, J ., and Galvez, J I. 2000. Reproductive characteristics of adult
channel catfish treated with trenbolone acetate during the phenocritical period of
sex differentiation. Aquaculture, 189: 351-360.

Ding, J L., Ng, WK., Lim, EH., and Lam, TJ. 1993. In situ hybridization shows the tissue
distribution of vitellogenin gene expression in Oreochromis aureus. Cytobios, 73:
197-208.

Dong, W., and Willet, KL. 2008. Local expression of CYP19A1 and CYP19A2 in
developing and adult killifish (F undulus heteroclitus), General and Comparative
Endocrinology, 155: 307-317.

Durhan, EJ., Lamb‘right, CS., Makynen, EA., Lazorchak, J ., Hartig, PC., Wilson, VS.,
Gray, LE., and Ankley, GT. 2006. Identification of metabolites of trenbolone
acetate in androgenic runoff from a beef feedlot. Environmental Health
Perspective, 114: 65-68.

Filby, AL., Thorpe, KL., Maack, G., and Tyler, CR. 2007. Gene expression profiles
revealing the mechanism of anti-androgen and estrogen-induced feminization in
fish. Aquatic Toxicology, 81: 219-231.

127

Goto-Kazeto, R., Kight, KE., Zohar, Y., Place, AR., and Trant, JM. 2004. Localization
and expression of aromatase mRN A in adult zebrafish. General and Comparative
Endocrinology, 139: 72-84.

Hayat, M.A., 2004. Comparision of immunochemistry, in situ hybridization, fluorescence
in situ hybridization, and chromogenic in situ hybridization, p. 3-1 1. In Hayat,
M.A. (ed.), Handbook of immunochemistry and in situ hybridization of human
carcinomas (V01. 1). Elsevier Academic Press, Burlington, MA.

Islinger, M., Willimski, D., Volkl, A., and Braunbeck, T. 2003. Effects of 17a-
ethinylestradiol on the expression of three estrogen-responsive genes and cellular

ultrastructure of liver and testes in male zebrafish. Aquatic Toxicology, 62: 85-
103.

Iwamatsu, T., Ohta, T., Oshima, E., and Sakai, N. 1988. Oogenesis in the medaka
Oryzias latipes-stages of oocytes development. Zoological Sciences, 5: 353-373.

Jensen, KM., Makynen, EA., Kahl, MD., and Ankley, GT. 2006. Effects of the feedlot
contaminant 17a-trenbolone on reproductive endocrinology of the Fathead
minnow. Environmental Science and Technology, 40: 3112-3117.

Kang, IJ., Yokota, H., Oshima, Y., Tsuruda, Y., Yamaguchi, T., Maeda, M., Imada, N.,
Tadokoro, H., and Honjo, T. 2002. Effect of 17B-estradiol on the reproduction of
Japanese medaka (Oryzias latipes). Chemosphere, 47: 71-80.

Kazeto, Y., Place, AR., and Trant, JM. 2004. Effects of endocrine disrupting chemicals

on the expression of CYP19 genes in zebrafish (Danio rerio) juveniles. Aquatic
Toxicology, 69: 25-34.

Kime, DE. 1998. Introduction to fish reproduction, p. 81-101. In Endocrine disruption in
fish. Kluwer Academic Publishers, Massachusetts, MA.

Kishida, M., and Callard, GV. 2001. Distinct cytochrome P450 aromatase isoforms in
zebrafish (Danio rerio) brain and ovary are differentially programmed and
estrogen regulated during early development. Endocrinology, 142: 740-750.

Kobayashi, K., Tamotsu, S., Yasuda, K., and Oishi, T. 2005. Vitellogenin-
128

immunohistochemistry in the liver and the testis of the medaka, Oryzias latipes,
exposed to 17B-estradiol and p-nonylphenol. Zoological Science, 22: 453-461.

Kolpin, DW., Furlong, ET., Meyer, MT., Thurman, BM., Zaugg, SD., Barber, LB., and
Buxton, HT. 2002. Pharmaceuticals, hormones, and other organic wastewater
contaminants in US. streams, 1999-2000: A national reconnaissance.
Environmental Science and Technology, 36: 1202-1211.

Kong, RYC., Giesy, JP., Wu, RSS., Chen, EXH., Chiang, MWL., Lim, PL., Yuen, BBH.,
Yip, BWP., Mok, HOL., and Au, DWT. 2008. Development of a marine fish
model for studying in vivo molecular responses in ecotoxicology. Aquatic
Toxicology, 86: 131-141.

Lee, C., Jeon, S.H., Na, J .G., Choi, Y.J. and Park, K. 2002. Sensitivities of mRNA
expression of vitellogenin, chriogenin, and estrogen receptor by estrogenic
chemicals in medaka, Oryzias latipes. Journal of Health Science, 48:441-445.

Miracle, A., Ankley, G., and 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. Ecotoxicology and Environmental Safety,
63: 337-342.

Nelson, ER., Wiehler, WB., Cole, WC., and Habibi, HR. 2007. Homologous regulation
of estrogen receptor subtypes in goldfish (Carassius auratus). Molecular
reproduction and development, 74: 1105-1112.

Nunez, BS., and Applebaum, SL. 2006. Tissue- and sex-specific regulation of CYP19a1
expression in the Atlantic croaker (Micropogonias undulates). General and
Comparative Endocrinology, 149: 205-216.

Om, S., Yamani, S., and Norrgren, L. 2006. Comparison of vitellogenin induction, sex
ratio, and gonad morphology between zebrafish and Japanese medaka after
exposure to Wet-ethinylestradiol and l7B-trenbolone. Archives of Environmental
Contamination and Toxicology, 51: 237-243.

Ortiz-Zarragoitia, M., Trant, J M., and Caaraville, MP. 2006. Effects of dibutylphthalate
and ethynylestradiol on liver peroxisomes, reproduction, and development of
zebrafish (Danio rerio). Environmental Toxicology and Chemistry, 25: 2394-

129

2404.

Park, JW., Hecker, M., Zhang, X., Tompsett, AR., Newsted, J L., Jones, PD., Au, DWT.,
Kong, RYC., Wu, RSS., and Giesy, JP. 2008. Fluorescence in situ hybridization
techniques (FISH) to detect changes in CYP19a gene expression of Japanese
medaka (Oryzias latipes). Toxicology and Applied Pharmacology, (Submitted).

Pinto, PIS., Singh, PB., Condeca, JB., Teodosio, HR., Power, DM., and Canario, AVM.
2006. ICI 182,780 has agonistic effects and synergizes with estradiol-17Bin fish
liver, but not in testes. Reproductive Biology and Endocrinology, 27: 4:67.

Pisco, M., Valiante, S., Romano, M., Ricchiare, L., Liguoro, A., Laforgia, V., Limatola,
E., and Andereuccetti, P. 2004. Ovarian follicle cells in T orepedo marmorata
synthesize vitellogenin. Molecular reproduction and development, 67: 424-429.

Rotchell, J .M. and Ostrander, GK. 2003. Molecular markers of endocrine disruption in
aquatic organisms. Journal of Toxicology and Environmental Health Part B:
Critical Reviews, 6: 453-495.

Schiffer, B., Daxenberger, A., Meyer, K., and Meyer, HHD. 2001. The fate of trenbolone
acetate and melengestrol acetate after application as growth promoters in cattle:
Environmental studies. Environmental Health Perspectives, 109: 1145-1151.

Scholz, S., and Gutzeit, H0. 2000. 17-a-ethinylestradiol affects reproduction, sexual
differentiation and aromatase gene expression of the medaka (Oryzias latipes).
Aquatic Toxicology, 50: 363-373.

Scholz, S., Kordes, C., Hamann, J ., and Gutzeit, H0. 2004. Induction of vitellogenin in
vivo and in vitro in the model teleost medaka (Oiyzias latipes): comparison of
gene expression and protein levels. Marine Environmental Research, 57: 23 5-244.

Seki, M., Fujishima, S., Nozaka, T., Maeda, M., and Kobayashi, K. 2006. Comparison of
response to l7B-estradiol and l7B-trenbolone among three small fish species.
Environmental Toxicology and Chemistry, 25: 2742-2752.

Skillman, AD., Nagler, JJ., Hook, SE., Small, JA., and Schultz, IR. 2006. Dynamics of
Non-ethynylestradiol exposure in rainbow trout (Oncorhynchus mykiss):

130

absorption, tissue distribution, and hepatic gene expression pattern.
Environmental Toxicology and Chemistry, 25: 2997-3005.

Sone, K., Hinago, M, Itamoto, M, Katsu, Y., Watanabe, H., Urushitani, H., Tooi, O.,
Guillette Jr., LJ., and Iguchi, T. 2005. Effects of an androgenic growth promoter
l7B-trenbolone on masculinization of mosquitofish (Gambusia afj'inis affinis).
General and Comparative Endocrinology, 143: 151-160.

Tompsett, AR., Park, JW., Zhang, X., Jones, PD., Newsted, JL., Au, DWT., Chen, EXH.,
Wu, RSS., Kong, RYC., Giesy, JP., and Hecker, M. 2008. Development of
validation of an in situ hybridization system to detect gene expression along the
HPG-axis in Japanese medaka, Oryzias latipes. Achieves of Environmental
Contamination and Toxicology, (Submitted).

Van der Ven, LTM., Holbech, H., Fenske, M., Van den Brandhof, EJ., Gielis-Proper, FK.,
and Wester, PW. 2003. Vitellogenin expression in zebrafish Danio rerio:
evaluation by histochemistry, immunohistochemistry, and in situ mRNA
hybridization. Aquatic Toxicology, 65: 1-1 1.

Wahli, W. 1988. Evolution and expression of vitellogenin genes. Trends in Genetics, 4:
227-232.

Wang, H., Tan, JTT., Emelyanov, A., Korzh, V., and Gong, Z. 2005. Hepatic and
extrahepatic expression of vitellogenin genes in the zebrafish, Danio rerio. Gene,
356: 91-100.

Wang, XG., and Orban, L. 2007. Anti-mullerian hormone and 11 beta-hydroxylase show
reciprocal expression to that of aromatase in the transforming gonad of zebrafish
males. Developmental Dynamics, 236: 1329-1338.

Wilkinson, DG., 1999. The theory and practice of in situ hybridization, p. 1-21. In
Wilkinson, D.G. (ed.), In situ hybridization: A practical approach (2nd Ed). IRL
Press, Oxford, NY.

Wilson, VS., Lambright, C., Ostby, J ., and Gray, LE. 2002. In vitro and in vivo effects of
17B-trenbolone: a feedlot effluent contaminant. Toxicological Sciences, 70: 202-
211.

131

Wittbrodt, T., Shima, A., and Schartl, M., 2002. Medaka-a model organism from the Far
East. Nature Reviews Genetics, 3: 53-64.

Zhang, L., Khan, IA, and Foran, CM. 2002. Characterization of the estrogenic response
in genistein in Japanese medaka (Oryzias latipes). Comparative Biochemistry and
Physiology, Part C 132: 203-211.

Zhang, X., Hecker, M., Park, JW., Tompsett, AR., Newsted, J L., Nakayama, K., Jones,
PD., Au, DWT., Kong, RYC., Wu, RSS., and Giesy, JP. 2008. Real time PCR
array to study effects of chemicals on the Hypothalamic-Pituitary-Gonadal axis of
the Japanese medaka. Aquatic Toxicology, (Submitted).

132

CONCLUSION

The aim of my dissertation research was to develop and validate sensitive and reliable
molecular techniques that can be used to elucidate the endocrine modes of chemical
action in vertebrates. The focus of my research was on wildlife species such as
amphibians and fish. The developed and validated methods were utilized for the
integrated examination of chemical effects at the molecular, histological, and organismal
level by measuring the amount of mRN A, assessing tissue morphology, and by
evaluating how changes in gene expression and tissue morphology relate to effects on
reproductive functions in the test organism. The techniques developed allowed for the
determination of mechanisms of toxic action of single chemicals.

In phase I, the Q RT PCR method developed to quantify CYP19a gene expression
in testicular tissues of male X. laevis was sufficiently sensitive, rapid, and reproducible to
allow the measurement of single digit copies of total RNA. This sensitive and precise
assay could be a useful tool that allows for quantifying specific types of mRNA that are
expressed at low levels in certain tissues such as CYP19a in testes of male frogs and that
allows for direct comparison of gene expression levels between samples. In fact, the Q
RT PCR technique developed in this study has been successfully applied in a parallel
studies to demonstrate that atrazine does not interact with aromatase gene expression in
either male or female X. laevis (Hecker et al., 2005a in introduction). Furthermore, the Q
RT PCR method developed forX. laevis during these studies was further optimized to aid
as a validation tool in the subsequent FISH studies of spatial changes in gene expression

using medaka as a whole animal model. The methods were used to test the hypothesis

133

that atrazine, a commonly used herbicide, was up-regulating CYP19a activity in the testes
of male frogs and thus causing feminization that could contribute to population declines.
It was found that atrazine did not affect expression of CYP19a in the gonads of frogs.
This information was critical to a decision of the Science Advisory Panel of the US EPA
when they made a decision to support registration of this important agricultural chemical.
Thus, my published work had impacts on the science and on national policy decisions.
The optimized FISH method developed was sufficiently sensitive to detect and
quantify changes of gene expression in tissues of Japanese medaka after exposure to
certain types of EDCS. Also, FISH allowed for the determination of tissue-and/or
gender-specific differences in gene expression as well as how different tissues interact in
response to chemical exposure. This method not only provided useful information
relative to spatial changes in gene expression at the cellular and/or tissue level, but also
could help in the understanding of molecular changes at the level of histological
observation with the ultimate goal of being able to link histological changes to potential
pathologies. However, it should be mentioned here that there are still some limitations
regarding the use of FISH for routine chemical characterization. These limitations are
primarily related to issues regarding auto-fluorescence of tissues or components of the
ISH procedure, the semi-quantitative nature due to the variability observed in the
measured signal among sections for some genes, and the labor intensity of this method.
While complete removal of auto-fluorescence still poses a challenge, once removed, it
might allow detecting and quantifying gene expression at the cellular level with a
resolution and sensitivity that would even allow the quantification of very lowly

expressed genes. Furthermore, FISH has the potential to be further optimized using

134

multiple riboprobes with different emission spectra simultaneously with the same section
allowing for the simultaneous detection of multiple genes. The potential to such
improvements in combination with the here demonstrated capacity of FISH to detect
modes of chemical action and to relate these to pathologies renders this technique a

powerful future tool for chemical risk assessment.

135

   

11111111111111.11111111.111