 

. . .
42... . .
5 ..
I P a

:>:..m.‘.....=..:sa..~
t. ,
r.
. 32.5.
5:, in...»

4.5.,
I.

.. , ‘ .
63...?“mhmww V n . ..

.35“ rahww‘ 1

ﬁgs... um:

«45..

hr:

m5
1,.
a

#2::-

. .3. .
3mm
. .3. .
.i.
I... L 1944..
‘11 It.hu
.. fet- ,.t£ﬁ}lz.l.[
112“.) I unruly.
“:1.
41.3.
sun:
5) 4.
. v... . . u.
a . ﬁtcﬁawiubﬁl
I.
:3... .
ﬁndinnhnuf
. u. 4.:
9.0.1944“

.mmﬁ

 

 

 

 

7'21. E313

J
2063

Wu x’A/

J

IVEI‘SI

LIBRARY
higan State

1c
Un

E

This is to certify that the
dissertation entitled

AN INVESTIGATION INTO THE MECHANISM OF ACTION
OF ATRAZINE AND ITS EFFECTS ON DEVELOPING RANA
CLAMITANS ANDXENOPUS LAEVIS

presented by
Katherine Kemler Coady

has been accepted towards fulﬁllment
of the requirements for the

Ph.D. degree in Zoology

Major Professor's ture
8/ 3/05 {75

Date

MSU is an Aﬁinnative Action/Equal Opportunity Institution

 

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

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8/01 clelRC/DatoDuo.p65-p.15

AN INVESTIGATION INTO THE MECHANISM OF ACTION OF ATRAZINE AND
ITS EFFECTS ON DEVELOPING RANA CLAMITANS AND XENOPUS LAEVIS
By

Katherine Kemler Coady

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Department of Zoology

2003

ABSTRACT
AN INVESTIGATION INTO THE MECHANISM OF ACTION AND EFFECTS OF
ATRAZINE ON DEVELOPING AMPHIBIANS
By

Katherine Kemler Coady

Concern for the amphibian class has increased in the last two decades. A high incidence
of deformed ﬁo gs among certain populations and a grong awareness of amphibian
population declines across the globe have contributed to the fear that amphibians are
facing environmental stressors that may be deleterious to their future survival. Pesticide
exposure, whether from agricultural runoff or atmospheric deposition, is a suspected
environmental stressor thought to be adversely affecting amphibian communities.
Atrazine, one of the most widely used herbicides in the United States and Canada, has
been reported to cause abnormal reproductive development in amphibian populations via
disruption of the endocrine system. This laboratory investigation was undertaken to
elucidate the mechanism of action of atrazine in amphibians and to examine the effects of
atrazine on developing frogs. Two frog species, the native green frog (Rana clamitans)
and the non-native African-clawed frog (Xenopus laevis), were chronically exposed to
waterborne test solutions, including 0, 0.1, 1.0, 10, or 25 ug/L atrazine, 0.005% ethanol
(EtOH), or 0.1 mg/L estradiol or dihydrotestosterone carried in 0.005% EtOH. Test
solutions were renewed by 50% every 72 hrs. Survival, metamorphosis, growth, sex
ratios, as well as gonad and larynx development were examined as endpoints in this

study. In addition, sex steroid concentrations, and P450 aromatase activity were

examined in X. laevis ﬁogs. Aromatase, an important enzyme in reproductive
development, is responsible for converting androgens into estrogens. Atrazine exposure
at environmentally relevant concentrations did not signiﬁcantly affect frog mortality in
either species. Time to metamorphosis as well as length and weight of frogs at
metamorphic completion did vary between treatments in some cases, but these
parameters were not consistently affected across atrazine concentrations. Sex ratios were
not altered in the green frog or the African-clawed frog in response to atrazine exposure.
However, when green frog tadpoles were exposed to DHT, over 97% of the
metamorphosing frogs were classiﬁed as males. Neither laryngeal nor gonadal
development appeared to be affected by atrazine in the exposed frogs. As expected
DHT-exposed frogs had signiﬁcantly larger larynx muscle areas compared to frogs in
other treatments. Sex steroid levels, including WIS-estradiol and testosterone, were not
affected in response to atrazine exposure. Likewise, aromatase activity in the brains and
gonads of atrazine-exposed frogs was not different from controls. However, aromatase
activity was signiﬁcantly altered in E2 exposed frogs. In conclusion, chronic atrazine
exposure at environmentally relevant concentrations did not affect amphibian survival,
growth, or reproductive development. Based on these ﬁndings, it appears that low-level
atrazine exposure does not disrupt the amphibian endocrine system via interference with
the steroidogenic enzyme, aromatase. Future research using a greater range of atrazine
concentrations and focusing on the possible interactions of atrazine with the amphibian
endocrine system, including the hypothalamus-pituitary axis, is necessary to discount

atrazine as an endocrine disruptor in amphibians.

Cepyright by
KATHERINE KEMLER COADY
2003

ACKNOWLEDGMENTS
I would like to thank my graduate advisory committee including Dr. John Giesy, Dr.
Steve Bursian, Dr. Heather Eisthen, and Dr. Don Hall.

Special thanks to all the aquatic toxicology lab personnel at Michigan State University,
including Margaret Murphy, Dr. Markus Hecker, Dr. Paul Jones, Randy Szwast, John
Longcore, Eric Higley, Amber Tompsett, Carolyn Sageman, and Sandy Mazzoni.
Histological work was done at Texas Tech University and at Michigan State University’s
Clinical Center. Many hours were spent sectioning and staining frog gonad samples, so

thanks to Rick, Amy, Josilyn, and Cathy for the great histology samples.

TABLE OF CONTENTS

LIST OF TABLES ................................................................................... ix
LIST OF FIGURES ................................................................................. xi
INTRODUCTION .................................................................................. 1
CHAPTER ONE .................................................................................... 7
Abstract ....................................................................................... 8
Study Objectives ............................................................................ 8
Test Materials ............................................................................... 9
Methods ...................................................................................... 9
Exposure methods .................................................................. 9
Water quality and exposure veriﬁcation ....................................... 1 1
Methods at exposure termination ................................................ 12
Statistical methods ............................................ g .................... 14
Results ....................................................................................... 15
Atrazine exposure verification ................................................... 15
Water quality ....................................................................... l7
Mortality ........................................................................... 17
Metamorphosis ..................................................................... 19
Sex ratios ........................................................................... 23
Gonadal anomalies ............................................................... 24
Discussion .................................................................................. 28
Conclusion ................................................................................. 33

vi

CHAPTER TWO .................................................................................. 35

Abstract ..................................................................................... 36
Study Objectives ......................................... . .................................. 36
Test Materials .............................................................................. 37
Methods ..................................................................................... 37
Exposure methods ................................................................. 37
Analytical methods ................................................................ 39
Statistical methods ................................................................ 42
Results ....................................................................................... 43
Atrazine exposure verification ................................................... 43

Water quality ...................................................................... 45
Mortality ........................................................................... 45
Metamorphosis .................................................................... 46

Gross gonadal anomalies ....................................................... 47
Histological gonadal anomalies ................................................ 5 1
Sex ratios ........................................................................... 53
Laryngeal dilator muscle area ................................................... 54
Discussion .................................................................................. 56
Conclusion ................................................................................. 61
CHAPTER THREE ............................................................................... 62
Abstract ..................................................................................... 63
Study Objectives ........................................................................... 64
Methods ..................................................................................... 64

vii

Test materials and exposure methods .......................................... 64

Sample collection methods ....................................................... 64

Blood collection and sex steroid measurements .............................. 65

Tissue collection and aromatase activity measurements ..................... 66

Statistical methods ................................................................ 67

Results ....................................................................................... 68
Atrazine concentrations .......................................................... 68

Aromatase activity ................................................................ 69

Hormones ........................................................................... 74

Discussion .................................................................................. 79
Conclusion .......................................................................................... 81
REFERENCES ..................................................................................... 82

viii

LIST OF TABLES

CHAPTER ONE

Table 1: Nominal and measured concentrations of atrazine (jig/L) in exposure water as
measured by ELISA at Michigan State University (MSU) and Syngenta Crop Protection
Inc ..................................................................................................... 16

Table 2: Average percent (%) mortality for R. clamitans exposed to atrazine or sex
steroids over 506 d .................................................................................. 19

Table 3: Number of R. clamitans exposed to atrazine or sex steroids over 506 d and
surviving to initiation and completion of metamorphosis ..................................... 20

Table 4: Mean ages, lengths, and weights at metamorphosis for R. clamitans exposed to
positive controls and various concentrations of atrazine ....................................... 23

Table 5: Percent male and female R. clamitans exposed to positive controls and various
concentrations of atrazine ......................................................................... 24

Table 6: Percent gross gonadal anomalies in R. clamitans exposed to positive controls
and various concentrations of atrazine ........................................................... 26

Table 7: Percent histological gonadal anomalies in R. clamitans exposed to positive
controls and various concentrations of atrazine ................................................ 27
CHAPTER TWO

Table 1: Nominal and measured concentrations of atrazine in test solutions as measured
by Enzyme-Linked Immunosorbent Assay at Michigan State University (MSU) and

Syngenta Crop Protection Inc ..................................................................... 44

Table 2: Percent gross gonadal anomalies in Nieuwkoop-Faber stage 66 X. laevis
exposed to EtOH, DHT, E2, or various concentrations of atrazine ........................... 50

Table 3: Percent gross gonadal anomalies in X. laevis exposed to EtOH, DHT, E2, or
various concentrations of atrazine until 2-3 mo post-metamorphosis ........................ 50

Table 4: Percent gonadal anomalies at the histological level in male X. laevis exposed to

EtOH, DHT, E2, or various concentrations of atrazine until 2-3 mo post-
metamorphosis ...................................................................................... 52

ix

Table 5: Laryngeal dilator muscle cross-sectional areas for male and female X. laevis
exposed to EtOH, DHT, EtOH, or various concentrations of atrazine ...................... 55

CHAPTER THREE

Table 1: Gonadal aromatase activity (fmol/h/mg protein) in male and female juvenile X.
laevis exposed to positive controls and various concentrations of atrazine .................. 70

Table 2: Brain aromatase activity (pmol/h/mg protein) in male and female juvenile X.
laevis exposed to positive controls and various concentrations of atrazine .................. 73

Table 3: Plasma estradiol concentrations (ng/mL) in juvenile X. laevis exposed to
positive controls and various concentrations of atrazine ....................................... 76

Table 4: Plasma testosterone concentrations (ng/mL) in juvenile X. laevis exposed to
positive controls and various concentrations of atrazine ....................................... 78

LIST OF FIGURES

CHAPTER ONE
Figure 1: Cumulative mortality of R. clamitans for each treatment group .................. 18

Figure 2: Cumulative number of R. clamitans frogs completing metamorphosis for each

treatment group ...................................................................................... 20
Figure 3: Mean ages of R. clamitans initiating and completing metamorphosis ........... 21
Figure 4: Mean snout-vent length of R. clamitans at metamorphic completion ............ 22

Figure 5: Normal and abnormal gonad morphology in post-metamorphic

Rana clamitans ....................................................................................... 26
Figure 6: Histological gonadal anomalies in post-metamorphic Rana clamitans ......... 27
CHAPTER TWO

Figure 1: Normal and abnormal gonadal morphology of juvenile male and female X.
laevis 2-3-months post-metamorphosis ...................................... j .................... 49

Figure 2: Cross section of rudimentary herrnaphroditic gonads of post-metamorphic X.
laevis .................................................................................................. 53

Figure 3: Laryngeal dilator muscle areas in Nieuwkoop-Faber stage 66 X. laevis exposed
to EtOH, DHT, E2, and various concentrations of atrazine ................................... 56

CHAPTER THREE

Figure l: Gonadal aromatase activity in male and female X. laevis exposed to positive
controls and various concentrations of atrazine ................................................. 71

Figure 2: Brain aromatase activity in male and female X. laevis exposed to positive
controls and various concentrations of atrazine ................................................. 74

Figure 3: Estradiol concentrations in juvenile male X. laevis exposed to positive controls
and various concentrations of atrazine ........................................................... 77

xi

INTRODUCTION

In recent decades, there has been a growing concern among scientists and
conservationists for the status of amphibian populations worldwide. Diverse members of
the amphibian class, including Costa Rica’s golden toad (Bufo periglenes) and Australia’s
unique gastric brooding frog (Rheobatrachus silus), have been completely eliminated in
recent years, with scientists left trying to discover the cause of their disappearance
(Pounds and Crump, 1994; Laurance etal., 1996). In addition to amphibian extinctions, a
greater number of limb deformities among frog populations have also been noted in the
United States and Canada (Ouellet et al., 1997; Sower et al., 2000). By some accounts,
the increasing reports of amphibian declines and deformities indicate that amphibian
populations are now expiring at an unprecedented rate. This has fueled increased
research into the causes of amphibian declines.

The mechanisms underlying amphibian declines are complex, and several causal
factors are probably involved depending on region and habitat (Blaustein and Kiesecker,
2002). These causal factors include climate changes (Pounds and Crump, 1994),
ultraviolet radiation (Ankley et al., 1998; Blaustein et al., 2003), acid rain (Rowe et al.,
1992), introduction of exotic species (Kats and Ferrer, 2003), parasitism (Lefcort and
Blaustein, 1995), infectious diseases (Daszak et al., 2003), habitat destruction, and
environmental contaminants (Blaustein et al., 2003).

Environmental contaminants, especially pesticides and industrial byproducts, are
ubiquitous in the environment due to atmospheric transport. Amphibians, due to their
permeable skins and life history (incorporating both aquatic and terrestrial habitats), are

considered especially at risk of exposure to environmental contamination. Pesticide use,

in particular, has been a suspected contributor to amphibian declines. This is especially
true in cases where there is a close association between wetland habitat and agricultural
areas. Low-lying wetlands can act as a sink for various agricultural toxicants, and
amphibians, due to their dependence on aquatic systems, can be exposed to various
toxicants via agricultural runoff, drift, or aerial deposition. Most pesticides are applied in
the spring when frogs and other amphibians are congregating near water bodies in
preparation for reproductive activities. Pesticide exposure may affect larval amphibians
to a greater extent than adults, since, in most cases, the larvae are restricted to water
bodies that may be contaminated. In addition, early life stages are often more sensitive to
pollutants, due to the complex and sensitive developmental processes taking place
(McKim, 1977).

Pesticide exposure has been shown to alter growth and metamorphosis in
developing anurans. Juvenile ranids exposed to triphenyltin fungicides showed decreased
growth rates and increased time to metamorphosis (F ioramonti et al., 1997). Atrazine
exposure at concentrations greater than 40 ug/L caused a decrease in the length and
weight of Xenopus laevis and Hyla versicolor metamorphs as compared to controls
(Diana et al., 2000; Sullivan and Spence, 2003).

Pesticide exposure has also been linked to immunosuppresion in certain frog
populations. A reduction in antibody production was observed in northern leopard frogs
(Rana pipiens), when the frogs were injected with sub-lethal doses of various pesticides
(Gilbertson et al., 2003). Studies have suggested that pesticide exposure alters various
immune responses in amphibians, which consequently affects their ability to combat

parasitic infections (Kiesecker, 2002; Christin et al., 2003).

Some pesticides target the nervous system in frogs, causing an array of adverse
effects. For example, exposure to the insecticide endosulfan affected the neuromuscular
development in ranid and bufonid tadpoles, causing paralysis (Berrill et al., 1998).
Organophosphate insecticides, used in California’s Sierra Nevada, may be contributing to
amphibian declines in that area by interfering with cholinesterase activity in the nervous
system. Depression of cholinesterase activities has been shown to reduce tadpole
activity, increase uncoordinated swimming behavior, and lead to greater tadpole mortality
(Sparling et al., 2001). Triphenyltin exposure also appears to affect the nervous system
of developing frogs. Triphenyltin-exposed tadpoles exhibited decreased swimming and
feeding activity compared to tadpoles in the control treatment (Semlitsch et al., 1995).

Some pesticides may have the ability to interact with and consequently disrupt the
amphibian endocrine system, thereby altering hormone-dependent processes, such as
development and reproduction. For example, methoxychlor has been shown to inhibit
progesterone-induced maturation of X. laevis oocytes (Pickford and Morris, 1999).
Dichlorodiphenyl trichloroethane (DDT), a now banned insecticide, is suspected of
disrupting corticosterone activities in exposed amphibians (Hayes et al., 1997). During a
single year of a multi-year study, the prevalence of intersex individuals among cricket
frog populations (Acris crepitans) was increased in those areas with greater pesticide
exposure (Reeder et al., 1998). In addition, deformed frogs collected in New Hampshire
were found to contain lower levels of androgens and gonadotropin releasing hormones.
Both the deformities and the disrupted hormone levels in these frogs were attributed to

pesticide exposure (Sower et al., 2000).

Atrazine (6-chloro-4-ethylamino-6-isopropyl-amino-s-triazine) is a suspected
endocrine-disrupting chemical (Cooper et al., 2000) and is also the most widely used
herbicide in the Midwestern United States and Canada (Solomon et al., 1996). The
herbicide was ﬁrst registered for commercial use in the United States in 1959 and has
been applied mainly on corn, sorghum, and sugarcane crops to control annual broadleaf
and grass weeds. The water solubility of atrazine is 33 mg/L at 22°C, thus the herbicide
can partition into the water and reach aquatic systems via runoff from agricultural ﬁelds
(Solomon et al., 1996). The half-life of atrazine can be as little as 2 d to as long as 800 (1,
depending on pH and other environmental factors (Solomon et al., 1996). Environmental
concentrations of atrazine rarely exceed 20 ug/L in North American water bodies;
however, runoff events periodically contaminate nearby aquatic systems with greater
atrazine concentrations (Solomon et al., 1996).

While atrazine is not acutely toxic at environmentally relevant concentrations
(Morgan et al., 1996), chronic exposure has been linked to endocrine disruption in
amphibians (Hayes et al., 2002; Hayes et al., 2003). Field surveys of frog populations
have suggested a trend between atrazine-exposure and altered reproductive development
(Reeder et al., 1998; Hayes et al., 2003; Hecker et al., 2003). However, a causal
relationship between atrazine-exposure and reproductive responses has not been
established in these cases. There have been several studies investigating the effects of
atrazine exposure on laboratory populations of African clawed frogs (Xenopus laevis)
(Hayes et al., 2002; Carr et al., 2003). One series of laboratory exposures reported that
atrazine increased the incidence of gonadal anomalies, decreased laryngeal muscle size,

and altered sex steroid concentrations in male X. laevis (Hayes et al., 2002). However, in

a similar exposure study, there were no observable changes in laryngeal muscle size and
fewer observed incidences of gonadal anomalies in frogs exposed to atrazine (Carr et al.,
2003). Other studies reported that exposure to 21 jig/L atrazine for 48 hrs during
development resulted in decreased testicular volume in male X. laevis tadpoles and
decreased numbers of primary germ cells in females (Tavera-Mendoza et al., 2002a;
Tavera-Mendoza et al., 2002b). Thus, some authors have suggested that atrazine may be
affecting gonadal development and consequently reproduction in some sensitive
amphibian populations via endocrine disruption.

The proposed mechanism of action by which atrazine disrupts the amphibian
endocrine system is via induction of P450 aromatase (aromatase) (Hayes et al., 2002).
The aromatase enzyme is a product of the CYP 19 gene family, so named because it
converts androgens to estrogens by removing the methyl group at carbon 19 (C19) and
aromatizes the steroid A ring (Melo and Rarnsdell, 2001). Atrazine, at relatively high
concentrations, was reported to induce aromatase activity in a mammalian cell line
(Sanderson et al., 2000). Thus, it has been proposed that atrazine increases the activity of
aromatase not only in vitro, but also in vivo, reducing plasma concentrations of
testosterone (T) and increasing plasma concentrations of estradiol (E2), resulting in
demasculinization or feminization of male frogs (Hayes et al., 2002; Hayes et al., 2003).
Aromatase activity and plasma sex steroid concentrations were investigated in Xenopus
laevis collected from reference areas and areas of atrazine use in South Africa (Hecker et
al., 2003). In this study, there were no signiﬁcant differences in gonadal aromatase
activity between frogs collected from reference areas and areas of atrazine use. However,

female X laevis collected from areas of atrazine use had signiﬁcantly lower plasma

concentrations of both T and E2 in comparison to female frogs collected from reference
locations (Hecker et al., 2003).

While the effects of atrazine on frogs are equivocal, they suggest that atrazine
may be affecting steroidogenesis and/or gonad development in amphibian populations via
endocrine disruption. Thus, the following research was conducted to elucidate the
mechanism of action of atrazine in amphibians and to investigate the effects of
environmentally relevant concentrations of atrazine on developing frogs.

Two anuran species, one North American native and one non-native, were
selected as test animals for laboratory exposures involving atrazine. The green frog
(Rana clamitans) was selected as a representative amphibian of North American
wetlands. Green frogs commonly occur in agro-ecosystems of the Midwestern United
States, where atrazine is frequently used in corn crop production (Gillilland et al., 2001).
These frogs deposit eggs in quiet waters and their larval stage is adapted to life in a pond.
Green frog tadpoles may not undergo metamorphosis for a year or two depending on
environmental conditions. Following metamorphosis, the adult green frog is terrestrial,
but remains closely associated with ponds and wetlands (Conant and Collins, 1998).

African clawed frogs (Xenopus laevis) are native to South Africa (Duellrnan and
Trueb, 1994). Unlike native North American frogs, X. laevis will breed year round under
laboratory conditions. X. laevis is purely aquatic throughout its lifecycle, and is an easy
amphibian to care for in the laboratory. Thus, these frogs are commonly used in various
scientiﬁc investigations. Since there is widespread use of African-clawed frogs in
toxicology research, the selection of this test species allows for more precise comparisons

between this and other toxicology studies.

CHAPTER ONE
EFFECTS OF ATRAZINE ON METAMORPHOSIS, GROWTH, AND GONADAL

DEVELOPMENT IN THE GREEN FROG (RANA CLAMI T ANS)

Abstract
Embryos of the green frog (Rana clamitans) were collected from the ﬁeld and

exposed to one of six waterborne treatments for 273 (1 (mid July 2001 to mid-April 2002).
The treatments were 0, 10, or 25 jig/L atrazine, 0.005% ethanol (EtOH), or 0.1 mg/L
estradiol or dihydrotestosterone carried in 0.005% EtOH. The exposures began
approximately 5 d post-hatch. Treatments were applied in a static renewal system with a
50% test solution replacement approximately every 3 (1. Following the exposure period,
tadpoles were reared in freshwater until metamorphosis or until study termination (at day
506). Time to initiate and complete metamorphosis, stage-speciﬁc mortality, length and
weight at metamorphosis, and gross morphology and histology of the gonads were
examined. At environmentally relevant concentrations, atrazine did not consistently
affect grth or metamorphosis. Compared to controls, the length of the larval period

was greater in tadpoles exposed to 10 ug/L atrazine. However, the length of the larval

period was not different between tadpoles in the control and 25 jig/L atrazine treatments.
Neither gross gonadal morphology nor histopathology of the gonads was altered in
response to atrazine exposure in post-metamorphic frogs. This study provides evidence
that environmentally relevant concentrations of atrazine do not adversely affect the

growth or reproductive development of Rana clamitans.

Study Objectives
The overall hypothesis investigated in this study was that waterborne

concentrations of atrazine could affect gonadal development of the green frog (Rana

clamitans) when exposed during critical phases of development. The objectives of this

study were: (1) to develop and validate methods of husbandry and exposure for
conducting laboratory studies with R. clamitans and (2) to determine the response of
larval R. clamitans to atrazine by assessing metamorphosis and reproductive indices
when animals are exposed during larval development. Indices evaluated included the
number of frogs initiating and completing metamorphosis, time to metamorphosis, fresh
post-mortem body weight and snout-vent length at metamorphic completion, and the

incidence of gross and histological gonadal anomalies.

Test Materials
Atrazine (CAS number 1912-24-9; purity 97.1%) was obtained from Syngenta

Crop Protection Inc. (Greensboro, NC, USA). Estradiol (E2) (CAS number 50-28-2,
purity 98%) and dihydrotestosterone (DHT) (CAS number 521-18-6) were purchased
from Sigma Chemical Co. (St. Louis, MO, USA). Ethanol (EtOH) (CAS ntunber 64-17-
5, purity 200 proof, 100% USP grade) was purchased from AAPER alcohol (Shelbyville,
KY). All chemicals were dissolved in UV-treated laboratory freshwater. EtOH was used
as a carrier solvent to deliver DHT and E2 treatments in the freshwater, thus the 0.005%

EtOH treatment group served as a solvent control.

Methods
Exposure methods

R. clamitans eggs were collected as a single mass of fertilized eggs from a rural
pond (42°41 ’20”N, 084°16’59”W) in Michigan on July 10, 2001. This pond has no

history of pesticide contamination and no predatory ﬁsh populations, but does have a

stable population of R. clamitans. Eggs were brought back to the laboratory and batched
in 600-mL beakers containing 500 mL of aerated pond water. Pond water was gradually
replaced with laboratory freshwater. Hatching began on July 13, 2001. On July 18, 2001
tadpoles had reached Gosner stage 25 (Gosner, 1960) and the exposure period began.
Approximately 30 free-swimming tadpoles were randomly assigned to 10-L exposure
tanks containing 4 L of test solution. Each exposure treatment was replicated in 9 tanks
for a total of 54 exposure tanks in the study. Tadpoles were housed in a windowless
environmental chamber with a light exposure regime of 12 hrs of light followed by 12 hrs
of darkness.

The test solutions used in the exposure were untreated control (UV -treated
freshwater only), 0.005% EtOH, 0.1 mg/L E2, 0.1 mg/L DHT, as well as 10 or 25 ug/L
atrazine. Test materials were applied in a static renewal exposure regime and were
renewed by 50% every 72 hours with minor exceptions. During test solution renewals, 2
L of medium were removed by siphoning from each tank. Each treatment had its own
siphon tube assigned by color code to avoid cross-contamination. New batches of the
various test solutions were made up in large glass tanks and delivered by siphon to 20-L
polypropylene carboys. The carboys were then used to deliver the new test solutions to
the exposure tanks.

On September 22, 2001, after 67 d of exposure, tadpoles were transferred from
10-L glass tanks into 20-L tanks containing 16 L of test solution. The transfer was
necessary due to tadpole growth and increasing levels of nitrite and ammonia nitrogen in
the smaller exposure tanks. Aﬁer a total of 273 d of static renewal exposure, tadpoles

were transferred to a continuous ﬂow-through system of clean freshwater. Flow-through

10

tanks were made by boring two 4 cm diameter holes on opposing sides of lO—L glass
tanks, and covering these holes with 1 mm mesh material to allow water ﬂow in and out
of the tanks. Flow-through tanks were housed in large acrylic tanks containing 80 L of
continuously renewing freshwater. From that time, tadpoles were reared in laboratory
freshwater until metamorphosis occurred. At the initiation of metamorphosis, marked by
forelimb emergence (Gosner stage 42), tadpoles were removed from the ﬂow-through
system and were housed as individuals or in small groups in 10-L glass tanks containing
approximately 500 mL of freshwater. Tanks holding metamorphosing frogs were tilted at
an angle such that both a wet and dry surface was available for the developing frogs. In
total, the study lasted for 506 d. Tadpoles were exposed for 273 days and then reared in
clean freshwater for the remaining 233 d of the study. After 506 d, tadpoles that had not
reached initiation of metamorphosis were sacriﬁced in tricaine methanesulfonate (MS-

222) and stored at -20°C.

Water quality and exposure veriﬁcation

Prior to test solution renewals, water quality parameters such as pH, total
ammonia nitrogen, nitrite nitrogen, hardness, and dissolved oxygen were measured in
select replicate tanks. Water quality parameters were measured using Larnotte test kits
for ammonia nitrogen, nitrite nitrogen, and hardness (Aquatic Ecosystems, Apopka, FL,
USA). Dissolved oxygen was measured by an YSI model 57 oxygen meter (Y SI Inc.
Marion, MA, USA), and pH was measured using an Orion research pH meter, model

710A (Therrno Orion Inc., Beverly, MA, USA).

11

Water samples were collected from all treatments following each test solution
renewal in order to measure atrazine concentrations in the exposure water.
Concentrations of atrazine were measured by enzyme-linked immunosorbent assay
(ELISA) at Michigan State University (MSU) and also by ELISA at Syngenta Crop
Protection Inc. (Greensboro, NC, USA). At MSU, concentrations of atrazine in dilute
stock solutions and replicate tanks were measured using the Envirogard® Triazine 96-
well plate kit (Strategic Diagnostics Newark, DE, USA; product # 7211000). The
method detection limit of the Envirogard® Triazine 96-well plate kit was 0.025 rig/L
atrazine. A subset of the water samples analyzed at MSU was also analyzed at Syngenta
Crop Protection Inc. At Syngenta, concentrations of atrazine were measured with the
Beacon Analytical triazine plate kit (Beacon Analytical Systems, Portland, ME, USA).
The method detection limit of the Beacon kit was 0.05 ug/L atrazine. When the
concentrations of atrazine, as measured by ELISA, deviated from expected
concentrations, water samples were compiled and analyzed by gas chromatography/mass
selective detection (GC-MS) at Syngenta Crop Protection Inc. to verify the results of the
ELISA. For GC-MS analyses, water samples were buffered to a pH of 10, and analytes
were extracted by partitioning into ethyl acetate. Samples were then evaporated to
dryness and reconstituted in acetone. GC-MS analyses were performed in the selected

ion monitoring mode (Y okley and Cheung, 2000).
Methods at exposure termination

Throughout the study, mortalities and number of frogs initiating and completing

metamorphosis were recorded daily. At metamorphic completion, marked by complete

12

tail resorption (Gosner stage 46), ﬁ'ogs were euthanized by immersion in MS-222,
weighed, and measured for snout-vent length. Then the frogs were preserved in Bouin’s
ﬁxative for 48 hrs followed by 70% ethanol for long-terrn storage. When frogs were
transferred into ethanol, code numbers were assigned so that the frog’s treatment origin
would be unknown in subsequent analyses. All frogs completing metamorphosis and
some frogs near metamorphic completion (Gosner stages 44 - 46) were examined for
gross gonadal morphology. Frogs were dissected under an Olympus SZ4O
stereomicroscope and gonads were examined for sex classiﬁcation and gross gonadal
anomalies. Digital photographs were taken of each specimen. The incidence of gross
gonadal anomalies was enumerated and the nature of the anomaly described.

A random subset of the frogs examined for gross morphology of the gonads was
also examined for histology of the gonads. Gonads were removed from the ﬁxed frogs,
embedded in parafﬁn and serially cross-sectioned at 5 pm intervals. Every fourth cut
section of the gonad was mounted onto a slide, such that every 20 um was represented in
the analysis. Slides were stained with eosin and hematoxylin. All slides of sectioned
ovaries and testes were observed under a compound microscope without knowledge of
sample origin. When they occurred, histological anomalies were described and
enumerated.

Historically, a number of terms have been used to describe both normal and
abnormal gonadal development in vertebrates, including frogs (Dumont, 1972). More
recently, other terms have been used to describe the phenomenon in ﬁsh (Sumpter et al.,
1996; Gimeno, 1998) and frogs (Carr et al., 2003). In this study, juvenile frogs

containing both testicular and ovarian tissue are referred to as rudimentary

l3

herrnaphrodites. Rudimentary hermaphrodites function as one sex, but also contain
gonadal tissue of the opposite sex (Van Tienhoven, 1983). Rudimentary
hermaphroditism can be characterized by two different morphologies. The intersex
morphology occurs when testicular and ovarian tissues are separated rostral/caudal or
left/right in the same individual. The mixed sex morphology occurs when both testicular
and ovarian tissue are mixed together in the same gonad. When observed at the
histological level, the mixed sex morphology is characterized by the occurrence of
testicular oocytes. Testicular oocytes were characterized as such if the oocytes within the
testes contain an intact nucleus, nucleoli within the nucleus, and a surrounding squamus

epithelial layer.

Statistical methods

Fresh batches of treatment solutions (before being added to exposure tanks) and
replicate exposure tanks were sampled and analyzed for concentrations of atrazine.
These concentrations were incorporated into the total average atrazine concentration
calculated for that treatment. Measured concentrations of atrazine from replicate tank
samples were averaged across tanks for a single date. This average concentration was
then used as a single data point when calculating treatment averages. For any given
treatment, concentrations in the ﬂesh batches of treatment solutions were indicative of
atrazine exposure for the date that they were used. Non-parametric 95% conﬁdence
intervals were calculated for mean measured concentrations of atrazine by bootstrapping

the data array.

14

Kolmogorov-Smimov’s One Sample test with Lillifor’s transformation was used
to assess whether observations were normally distributed. When data were normally
distributed, analysis of variance (AN OVA) followed by Fisher’s least signiﬁcant
difference (LSD) was used to detect signiﬁcant differences between treatment groups.
When observations for a parameter were not normally distributed, the non-parametric
Kruskal-Wallis test was used to detect differences in the treatment groups. The non-
parametric Mann-Whitney U test was used to test for signiﬁcant differences between
treatment groups. The Chi-square test was used to detect signiﬁcant deviations in
expected sex ratios in each replicate tank, and Pearson’s Chi-square was used to test for
signiﬁcant differences in the incidences of gross and histological gonadal anomalies. The
criteria for signiﬁcance that was used in all statistical tests was p<0.05. The probability
of a type H error (l-B), or power, in statistical tests was set at 0.8, and the detectable
differences among treatment groups were calculated based on sample size and the
average variation of the parameter under investigation.

Data sets were analyzed in two separate groups such that the treatment groups
were compared to the appropriate control treatment in the experimental design. The
atrazine data set included all atrazine treatments and the untreated control. The positive

control data set included the DHT, E2 and EtOH solvent control treatments.

Results
Atrazine exposure verification
When measured by ELISA at MSU, concentrations of atrazine were similar to the

nominal concentrations. Similar results were found when water samples were analyzed

15

by immunoassay at Syngenta Crop Protection Inc. (Table 1). Small concentrations of
atrazine were detected in control exposure water (mean i standard deviation = 0.10 i
0.16 jig/L atrazine). To verify the presence of atrazine in control treatment tanks and to
cross check the results of the ELISA methods, select water samples from each treatment
(having the greatest concentrations of atrazine as measured by ELISA) were compiled
and measured by GC-MS at Syngenta Crop Protection Inc. The ranges of GC-MS-
measured atrazine concentrations in the control, 10 and 25 jig/L treatment samples were
4.0x10‘1- 4.3 x10], 18-29, and 33-50 jig/L atrazine, respectively. These values represent
the greatest concentrations of atrazine present during the exposure period for each

treatment.

Table 1: Nominal and measured concentrations of atrazine (pg/L) in exposure water as
measured by ELISA at Michigan State University (MSU) and Syngenta Crop Protection

Inc. (95% conﬁdence intervals in parentheses)

 

 

Treatment Nominal (Syngenta) (MSU)
concentrations Average measured Average measured
concentration concentration
a Control 0 0.14 (0.07-0.23) 0.10 (0.06 - 0.17)
10 ug/L atrazine 10 15.91 (12.03 —19.90) 11.76 (10.21 -l3.65)
25 ug/L atrazine 25 27.95 (24.92-31.24) 28.23 (25.14 - 31.60)

 

' For calculations, a proxy value of either 0.0125 or 0.025 pg/L atrazine was assigned to values below the

ELISA method detection limits at MSU or Syngenta, respectively.

16

Water quality

Water quality was measured in both the static exposure tanks and the ﬂow-
through freshwater tanks. The average temperature (1 standard deviation) in the static
exposure tanks was 21.3 i 1.4 °C. The average nitrite nitrogen and total ammonia
nitrogen concentrations (i standard deviation) in static exposure tanks were 0.22 i 0.44
and 0.04 i 0.05 mg/L, respectively. Static tanks had an average pH of 7.87, an average
hardness concentration (i standard deviation) of 426 :t 45 mg/L C3CO3, and an average
dissolved oxygen content (i standard deviation) of 8.0 i 0.7 mg/L. The average
temperature (i standard deviation) in the ﬂow-through tanks was 24.8 i 0.5 °C. Both
nitrites and total ammonia nitrogen had average concentrations of 0.02 mg/L in the ﬂow-
through tanks. The average dissolved oxygen concentration (i standard deviation) in the
ﬂow- through tanks was 6.1 i 1.7 mg/L, and the average hardness concentration (i

standard deviation) was 424 i 26 mg/L CaCO3.

Mortality

Mortality was greatest during the ﬁrst month of the exposure period, and tended
to decrease as the tadpoles aged. Mass mortality events occurred early in the study
within a time span of 24 hrs. Due to tank-speciﬁc mortality, several experimental
replicates were lost. Mortality decreased aﬁer transferring the tadpoles from 4 L to 16 L
in the static exposure regime. Mortality also decreased aﬁer transferring tadpoles from
the 16 L static tanks into the ﬂow-through water system after 273 d (Figure 1). Average
total mortality over 506 d was 76.5%. There were no signiﬁcant differences in mortality

among atrazine treatments and the untreated controls (ANOVA, p=0.34). Given the

17

sample sizes and variation in mortality in this experiment, a distinction of 19% mortality

was differentiated between treatments with 80% power (1 43). Mortality did differ among

the positive control treatments and the EtOH solvent control (ANOVA, p=0.01). The

DHT treatment had fewer deaths than the E2 treatment (LSD, p=0.003), but mortality in

this group was not different from the EtOH control treatment. There were also no

signiﬁcant differences in mortality between the E2 treatment and the EtOH controls.

Average mortality is listed for each treatment group (Table 2).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

i

'u i ‘

g 150i

5 mETOI-l

I

8 —E2

--——10uolLdrazine
50

1 f —250dl.atrazln:] i
l 1 J l
: 0 1 i A I 1
l 1 51 101 151 201 251 301 351 401 451 501
l
l 3WD!!!

 

‘ Figure 1: Cumulative mortality of R. clamitans for each treatment group

Arrows indicate the days in which tadpoles were transferred to new exposure vessels.

18

Table 2: Average percent (%) mortality for R. clamitans in different treatment groups

 

 

over the 506 (1 study period

Treatment Initial N Average % Mortality

Untreated controls 285 79.2

EtOH 280 74.8

DHT 291 62.7

E2 282 * 85.7

10 jig/L atrazine 292 73.1

25 jig/L atrazine 292 83.3

 

EtOH = 0.005% ethanol, DHT = 0.1 mg/L dihydrotestosterone, E2 = 0.1 mg/L estradiol

* indicates a signiﬁcant difference from the DHT treatment at p < 0.05

Metamorphosis

The ﬁrst incidence of initiation of metamorphosis was observed on exposure day
99. The ﬁrst incidence of completion of metamorphosis was observed on exposure day
112. As of exposure day 143, 10 tadpoles had completed metamorphosis. Between
exposure day 143 (December 7, 2001) and 285 (April 28, 2002), no tadpoles initiated
metamorphosis. An increased rate in development occurred after exposure day 285 (post
April 28, 2002) (Figure 2). Total numbers of green frogs surviving to initiation and

completion of metamorphosis are listed (Table 3).

l9

Table 3: Nrunber of R. clamitans exposed to different treatment groups over 506 d and

surviving to initiation and completion of metamorphosis

 

 

Treatment Initial N # Frogs Initiating 1 # Frogs Completing
Metamorphosis Metamorphosis

Untreated controls 285 58 (20.4%) 44 (15.4%)
EtOH 280 69 (24.6%) 47 (16.8%)
DHT 291 104 (35.7%) 75 (25.8%)
E2 282 40 (14.2%) 33 (11.7%)
10 ug/L atrazine 292 77 (26.4%) 64 (21.9%)
25 jig/L atrazine 292 48 (16.4%) 37 (12.7%)

 

 

I
l

EtOH = 0.005% ethanol, DHT = 0.1 mg/L dihydrotestosterone, E2 = 0.1 mg/L estradiol

 

—Control

 

a: 80-]
.E .
£312 1
E 8 601
8‘9 '
t o 404:
21%
3
E 20
3
o
0
200 250

 

Figure 2: Cumulative number of R. clamitans ﬁogs completing metamorphosis for each

treatment group

Age at initiation and completion of metamorphosis was different among the

 

#MHIMWW“ ET 0 H
10 ug/L atrazine

 

 

 

DHT

25 ug/L atrazine

 

 

d'wif-
all“
[1:
5.15
300 350 400 450 500 550
Day

 

 

__J

atrazine treatment groups and the untreated controls (Kruskal-Wallis, p=0.021 [initiation]

p = 0.010 [completion]). Pair-wise comparisons revealed that frogs in the 10 ug/L

20

atrazine treatment initiated and completed metamorphosis at a signiﬁcantly older age as
compared to both untreated control frogs and frogs exposed to 25 ug/L atrazine (Figure
3).

Age at initiation of metamorphosis was also different among the positive control
treatment groups and the EtOH controls (Kruskal-Wallis, p=0.029). Pair-wise
comparisons revealed that frogs in the E2 treatment initiated metamorphosis at a younger
age as compared to both EtOH control frogs and frogs exposed to DHT (Mann-Whitney
U, p=0.049 and 0.013, respectively; Figure 3).

if DAgenat initiation

j I Age at completion

01
O

N
001
CO

 

  

Mean age (d)
N 0) (.0
O
O

  

 

   

TIT _.i .14 i 1‘1_%

‘ Control 10 uglL 25 uglL ETOH DHT E2
atrazine atrazine

l __ _

 

Figure 3: Mean ages of R. clamitans initiating and completing metamorphosis (error bars
represent the standard error of the mean)

" Signiﬁes a signiﬁcant differences from untreated control and 25 ug/L atrazine treatments at p< 0.05

" Signiﬁes signiﬁcant differences from the EtOH and DHT treatments at p<0.05

Snout-vent length at metamorphic completion was different among the atrazine

treatment groups and the untreated controls (ANOVA, p=0.03). Fisher’s LSD test

21

revealed that frogs in the 10 jig/L atrazine treatment were shorter than frogs in the 25

ug/L atrazine treatment (p=0.009) but were not shorter than the controls (p=0.15). There
were no differences in snout-vent length among frogs in the positive control treatments
and the EtOH controls (ANOVA, p= 0.15) (Figure 4). At 80% power, a difference in

length of 0.22 cm was detectable among these treatments.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

j ' "2.701 '5 F A C F l
E ' E l
1 .9. 2.60 5 I
l 5 I : l
a, 1
.3 2.50 < I 1 l i
*- : T
5 2.40 -< I : l.
a l :
3 2.30 ~ 5
C 1
”’ i
2.20 *- . . .1 . . .
Control 10 ug/L 25 ug/L ETOH DHT £2
atrazine atrazine

 

 

Figure 4: Mean snout-vent length of R. clamitans at metamorphic completion (error bars
represent the standard error of the mean)

* Signiﬁes a signiﬁcant differences from the 25 ug/L atrazine treatment at p< 0.05

There were no signiﬁcant differences in body weight at metamorphic completion
among the atrazine treated frogs and the untreated controls (ANOVA, p=0.07). Likewise,
there were no differences in weight at metamorphic completion among frogs in the
positive control treatments and the EtOH controls (ANOVA, p=0.47). The detectable

weight difference among all treatment groups in this experiment was found to be 0.36 g

22

at 80% power. Average ages, lengths, and weights at metamorphosis are listed for each

treatment (Table 4).

Table 4: Mean ages, lengths, and weights at metamorphosis for R. clamitans exposed to

positive controls and various concentrations of atrazine (Std. error in parentheses)

 

 

Treatment Mean age at Mean age at Mean Mean snout-vent

initiation (d) a completion ((1) b weight (g) length (cm)

Untreated Control 328 (8.5) 337 (10.0) 1.76 (0.05) 2.52 (0.08)
EtOH 350 (7.4) 360 (9.5) 1.56 (0.04) 2.47 (0.06)

DHT 350 (7.1) 369 (7.9) 1.50 (0.03) 2.42 (0.05)

E2 ** 330 (4.8) 342 (4.6) 1.64 (0.09) 2.57 (0.06)

10 ug/L atrazine * 362 (6.5) * 377 (7.3) 1.48 (0.03) * 2.39 (0.04)
25 ug/L atrazine 335 (7.7) 342 (7.9) 1.64 (0.05) 2.54 (0.07)

 

EtOH = 0.005% ethanol, DHT = 0.1 mg/L dihydrotestosterone, E2 = 0.1 mg/L estradiol
' Initiation of metamorphosis was marked by the ﬁrst emergence of forelimbs.

b Completion of metamorphosis was deﬁned as complete tail resorption.

* Indicates a signiﬁcant difference from the untreated control at p < 0.05.

" Indicates a signiﬁcant difference from the EtOH control at p < 0.05.

Sex ratios

Sex was determined during gross examination for gonadal anomalies. The sex
ratios of frogs from untreated control and atrazine treatments were not signiﬁcantly
different from an expected 50:50 male to female ratio (Chi-square). Likewise, sex ratios
in the EtOH and E2-exposed tanks were not signiﬁcantly different from an expected
50:50 male to female ratio (Chi-square). However, all tanks in the DHT treatment had

sex ratios skewed in favor of more male frogs (Chi-square). A total of 97.7% of the

23

examined frogs in the DHT treatment were males. Sex ratios for each treatment are listed

(Table 5).

Table 5: Percent male and female R. clamitans exposed to positive controls and various

concentrations of atrazine 3‘

 

 

Treatment N Males Females
Untreated Control 51 43.1 56.9
EtOH b 57 47.4 50.9
DHT 86 97.7 2.30
E2 38 36.8 63.2
10 jig/L atrazine 72 40.3 59.7
25 ug/L atrazine 44 40.9 59.1

 

EtOH = 0.005% ethanol, DHT = 0.1 mg/L dihydrotestosterone, E2 = 0.1 mg/L estradiol

' All frogs surviving to Gosner stage 44 were examined for sex determination

b As the gonads from one frog in the EtOH treatment were not located, the percentages do not total 100%

Gonadal anomalies

Several types of gross gonadal anomalies were observed: discontinuous gonad,
multiple testes, and size irregularity. Discontinuous gonad was characterized by
abnormal segmentation of the gonadal tissue. Frogs with multiple testes were
characterized by multiple testicular-like gonads on one side of the frog. Size irregularity
was characterized either by large size discrepancies between gonad pairs or unusually
large or small gonads (Figure 5).

The incidence of gross gonadal anomalies ranged from 0.0 — 5.9% across all

treatment groups (Table 6). A large size incongruity between gonad pairs was the most

24

commonly observed gross gonadal anomaly (Table 6). No hermaphroditic gonads
(testicular and ovarian tissue in one individual) were identiﬁed during gross gonadal
investigations. A discontinuous gonad was detected only in- the 25 jig/L atrazine
treatment (Table 6). Gonad or gonad pairs could not be located in ﬁve ﬁ'ogs. It is
unclear whether this was due to dissection errors or if these constituted an anomaly. Two
frogs in both the E2 and DHT treatments were missing one gonad upon inspection. In
addition, the gonads in one frog in the EtOH control treatment could not be located.
Using Pearson’s Chi-square test, no differences in the incidence of gross gonadal
anomalies were found among all treatment groups (Chi-square, p= 0.404 [discontinuous
gonads], p=0.527 [multiple testes], p=0.456 [size incongruity]).

When gonads were examined at the histological level, gonadal anomalies, such as
rudimentary hermaphroditism, were observed in the green frogs. Two different
morphologies of rudimentary hermaphroditism were observed: intersex gonads and
testicular oocytes (Figure 6).

When examined histologically, the incidence of gonadal anomalies, across all
treatment groups, ranged from 0.0 to 12.9%. Most of the gonadal anomalies occurred in
male frogs (Table 7). However, one frog in the E2 treatment group, identiﬁed as a
female during the gross gonadal inspections, had testicular oocytes. Testicular oocytes
were the most common histological anomaly. This anomaly occurred in untreated
control, EtOH, and E2-exposed frogs, but not in DHT, 10, or 25 jig/L atrazine-exposed
frogs (Table 7). Using Pearson’s Chi-square test, no differences in the incidence of
histological gonadal anomalies were found among all treatment groups (Chi-square, p=

0.299 [testicular oocytes], p=0.673 [intersex]).

25

Table 6: Percent gross gonadal anomalies in R. clamitans exposed to positive controls

and various concentrations of atrazine ‘

 

 

Treatment N Size incongruity Discontinuous Multiple Missing
between gonads gonad testes gonad

Control 51 5.9 0.0 2.0 0.0
EtOH 57 0.0 0.0 0.0 1.8
DHT 86 3.5 0.0 0.0 2.3
E2 38 5.3 0.0 0.0 5.3
10 jig/L atrazine 73 1.4 0.0 1.4 0.0
25 jig/L atrazine 44 2.3 2.3 0.0 0.0

 

EtOH = 0.005% ethanol, DHT = 0.1 mg/L dihydrotestosterone, E2 = 0.1 mg/L estradiol

'All frogs surviving to Gosner stage 44 were examined for gonadal anomalies

 

      

Figure 5: Normal and abnormal gonad morphology in post-metamorphic Rana clamitans
a) Normal testes, b) Normal ovaries, c) Gonad with size irregularity from the
dihydrotestosterone treatment, (1) Discontinuous ovary from the 25 jig/L atrazine

treatment, e) Multiple testes from the 10 jig/L atrazine treatment (Bar represents 1 mm)

26

          

.558” '1 - . '
amorphic Rana clamitans a) Gonad

\L— - ~ ,4.

Figure 6: Histological gonadal anomalies in post-met
with testicular oocytes from the untreated control treatment b) Intersex gonad from the 25

jig/L atrazine treatment (Bar represents 200 um)

Table 7: Percent histological gonadal anomalies in R. clamitans exposed to positive

controls and various concentrations of atrazine

 

 

Treatment N Rudimentary N Rudimentary N Rudimentary
Group males hermaphroditism females hermaphroditism total hermaphroditism
W Testicular Intersex Testicular Intersex
oocytes oocytes oocytes

 

Control 17 12 5.9 25 0.0 0.0 42 4.8 2.4
EtOH 20 5.0 0.0 24 0.0 0.0 44 2.3 0.0

DHT 22 0.0 4.5 2 0.0 0.0 24 0.0 4.2

E2 12 25 0.0 19 5.3 0.0 31 13 0.0

10 ug/L 20 0.0 0.0 25 0.0 0.0 45 0.0 0.0
atrazine

25 ug/L 16 0.0 6.3 20 0.0 0.0 36 0.0 2.8

atrazine

 

EtOH = 0.005% ethanol, DHT = 0.1 mg/L dihydrotestosterone, E2 = 0.1 mg/L estradiol

27

Discussion
Since there were no standard protocols available before our research was initiated,

methods for rearing R. clamitans in the laboratory needed to be developed. The observed
level of mortality may have been natural for this species or due to overcrowding in static
tanks. At times, water quality, such as relatively great concentrations of nitrites, may also
have contributed to tadpole mortality in the static tanks (Marco et al., 1999).

In addition to its possible effects on mortality in this study, overcrowding also is
known to decrease rates of grth and development in amphibians (Gillespie, 2002;

Tej edo and Reques, 1992). In this study, when ﬁ'ogs were transferred to larger aquaria or
removed from static tanks and placed in a ﬂow-through system, mortality decreased and
the rate of development increased (Figure 1). The increased rate of development for
tadpoles in the continuous ﬂow-through system was most likely due to the combined
inﬂuences of an increase in temperature, better water quality, and the onset of spring.
However, seasonal cues appeared to have the most important role in the regulation of
metamorphic development. Even though green frog tadpoles were kept under constant
temperature and lighting conditions in the laboratory, they still responded to seasonal
cues. This was evident in that no tadpole initiated metamorphosis in the winter during
the natural hibernation period for R. clamitans. However, metamorphosis greatly
increased in the spring when the natural hibernation period was over.

In this study, the use of Rana clamitans, a species with a long larval period,
simulates a worst-case scenario of atrazine exposure in an aquatic ecosystem. However,
the long spans of time from hatching to completion of metamorphosis makes R. clamitans
a difﬁcult species for routine developmental toxicology tests. Age at metamorphosis was

variable. A small set of tadpoles underwent metamorphosis the fall of the year they were

28

collected (2001), however the majority of tadpoles underwent metamorphosis spanning
the period from late April until November of the following year (2002). This result is
consistent with what has been observed in wild populationsof R. clamitans. In the ﬁeld,
most R. clamitans eggs laid near the end of the summer (after July 10‘“) did not hatch
until the following year in the spring. Egg masses that were laid before June 25th tended
to metarnorphose by early or mid fall of the same year (Martof, 1956). The R. clamitans
eggs used in this study were collected near the end of the spawning period. Individuals
that initiated metamorphosis during the ﬁrst season after their collection were generally
from tanks that had experienced high mortality rates during the ﬁrst month of the
exposure. Therefore, the few frogs that underwent metamorphosis in the ﬁrst year
experienced lesser tadpole densities during the pre-metamorphic phase of development.
It is possible that the reduced crowding early in development allowed for increased
growth, which enabled an earlier onset of metamorphosis (Tej edo and Reques, 1992).

To limit mortality and maximize development in future laboratory exposures with
R. clamitans, it is recommended that tadpoles be reared at lower densities in static tanks
(<1 tadpole/ L) or in a continuous ﬂow-through system. In addition, although they are
not as closely associated with aquatic systems throughout their life span, other species,
such as the wood frog (Rana sylvatica) or the northern leopard frog (Rana pipiens) have
shorter developmental periods and are possible alternative ranids that could also be used
in developmental toxicology research in aquatic systems. They were not selected as
experimental animals for this study because their populations are not as strongly
established in Michigan.

The results are discussed in terms of nominal exposure concentrations. However,

29

the actual average concentrations of atrazine were slightly different. The nominal
concentrations of 10 and 25 ug atrazine/L were actually approximately 12 and 28 pg
atrazine/L, respectively. The average concentration of atrazine measured in the control
exposure was 0.1 ug atrazine/L. Thus, the actual measured concentrations of atrazine
were slightly greater than the nominal concentrations. Average concentrations of atrazine
measured by Syngenta were greater than the concentrations measured at MSU. This was
due in part to the sub-selection of water samples that deviated from the expected nominal
atrazine concentrations for further analysis by Syngenta. It is unknown why small
concentrations of atrazine were measured in the control tanks. It is possible that other
compounds were able to cross-react with the triazine ELISA resulting in false positive
measures of atrazine in the control water. However, it seems most likely that small
amounts of atrazine were periodically present in the laboratory water supply. The small
concentrations of atrazine in the control water did not affect the conclusions of the study,
since there were no statistically signiﬁcant, dose-dependent effects of atrazine observed
on any of the parameters investigated.

Concentrations of atrazine in the aquatic environment rarely exceed 20 jig/L
(Solomon et al., 1996). Atrazine is generally applied in the spring and reaches aquatic
systems in pulses due to agricultural runoff events. The continuous exposure of 10 and
20 jig/L atrazine to green frog tadpoles for 273 days overestimates the expected duration
of atrazine exposure in the environment. Thus, the exposure regime in this laboratory
study models a worst-case scenario of atrazine exposure in natural frog populations.

Continuous exposure to 10 or 25 jig/L atrazine did not affect mortality of R.

clamitans. This result is consistent with other amphibian toxicology research involving

30

atrazine (Morgan etal., 1996; Battaglin and Fairchild, 2002). Atrazine causes amphibian
mortality at much greater concentrations (LCSO = 410 jig/L atrazine) than those used in
this study (Battaglin and Fairchild, 2002).

Length and age at metamorphic completion were signiﬁcantly different among
treatment groups. However, atrazine exposure did not affect these parameters in a
concentration-dependent manner. Tadpoles exposed to 10 jig/L atrazine had lower
mortality (73.1%) in comparison to the control (79.2%) and 25 [lg/L atrazine (83.3%)
treatments (Table 2). Thus, there were more tadpoles per tank in the 10 jig/L atrazine
treatment, and these frogs were older and smaller when they underwent metamorphosis
as compared to frogs in the control and 25 jig/L atrazine treatments. Tadpoles exposed to
25 jig/L atrazine had similar mortality as the controls. Consequently, frogs exposed to 25
jig/L atrazine did not differ from controls in age, weight, or length at completion of
metamorphosis. Therefore, the difference in time to complete metamorphosis between
treatment groups appeared to be a result of density-dependent tank effects on relative
growth rates rather than atrazine treatment. Frogs exposed to E2 had greater mortality
and metamorphosed sooner than frogs in the EtOH controls. Since frogs exposed to 10
ug/L atrazine metamorphosed at an older age as compared to controls, and frogs in the
E2 treatment metamorphosed at a younger age as compared to controls, this suggests that
any effects on the maturation process caused by atrazine are not the result of an E2-
mediated mechanism. In an exposure study with larval gray tree frogs (Hyla versicolor),
atrazine did not affect size and age at metamorphosis of tadpoles exposed to atrazine
concentrations < 200 ug/L (Diana et al., 2000). In our study, chronic exposure to 10 and

25 jig/L atrazine did not consistently affect age, length, or weight of R. clamitans at

31

completion of metamorphosis. However, an exposure to more than two concentrations of
atrazine would be necessary to fully elucidate a possible dose-response relationship
between atrazine exposure and tadpole development.

Even though tadpoles were not exposed to atrazine for their entire developmental
history, sex ratio data indicated that the exposure period of 273 d was a critical time for
green frog gonad development. Since R. clamitans exposed to DHT were 97.7% male, it
can be assumed that the exposure period and duration were sufﬁcient to allow effects of
exogenous substances on gonadal development. Unlike African clawed frogs (X. laevis),
R. clamitans are not fenrinized when exposed to exogenous estradiol, but instead are
masculinized when exposed to exogenous androgens, such as DHT (F oote and Witschi,
1939). This difference in responses to hormones demonstrates that endocrine regulation
of maturation and sexual development is different among frog species (Wallace et al.,
1999). Therefore, hormones and hormone mimics could induce different responses in
different frog species.

The fact that exposure to atrazine did not affect sex ratios in R. clamitans is
similar to the results reported for X. laevis exposed to atrazine, where the sex ratio was
not altered from an expected 50:50 ratio (Carr et al., 2003). While it has been
hypothesized that atrazine can feminize male frogs through up-regulation of aromatase,
the enzyme that converts testosterone to estradiol (Hayes et al., 2002; Hayes et al., 2003),
sex reversal was not observed in R. clamitans.

It has been reported that atrazine exposure causes hermaphroditism in some
species of frogs (Hayes et. a1, 2003). The fact that the incidence of gross and histological

gonadal anomalies was not different in control and atrazine-treated R. clamitans suggests

32

that this is not the case for all species exposed to similar concentrations of atrazine. It is
possible that different species or strains of amphibians have different types and
magnitudes of response to atrazine exposure. In addition, some amphibians are known to
have hermaphroditic characteristics during phases of their natural reproductive
development. In particular, testicular oocytes have been reported to be a natural
developmental occurrence in some strains of ranid species, including, but not limited to,
Rana temporaria (Witschi, 1929), Rana catesbeiana (I-Isu and Liang, 1970), and Rana
curtipes (Gramapurohit et al., 2000). In addition, bufonid toads, whether genetic males
or genetic females, develop ovaries, which become compressed into the Bidder’s organs
in front of the deﬁnitive gonad at the time of metamorphosis (Wallace et al., 1999).
Therefore, based on the results of this study and the reported incidences of natural
background levels of hermaphroditism in the scientiﬁc literature, gonadal anomalies in R.

clamitans do not appear to be related speciﬁcally to the presence or absence of atrazine.

Conclusion

In summary, chronic exposure to environmental concentrations of atrazine did not
affect mortality, growth, sex ratios, or incidence of gonadal anomalies in developing R.
clamitans. While another study has linked atrazine exposure to the occurrence of
hermaphroditism in leopard ﬁ'ogs (Rana pipiens) (Hayes et al., 2003), we did not observe
a similar phenomenon in laboratory exposed Rana clamitans. In our study,
hermaphroditism in Rana clamitans occurred irrespective of exposure to atrazine. The
difference in the results of the two studies could be attributed to the difference in frog

species examined. Gonadal development may differ between Rana pipiens and Rana

33

clamitans. In addition, part of the Hayes et a1. 2003 study relied on data from ﬁeld
collected frogs. An array of factors, apart from atrazine, could inﬂuence the development
of hermaphroditism in ﬁeld-collected frogs. In a controlled laboratory exposure, we
observed that exposure to atrazine at environmentally relevant concentrations did not

affect mortality, growth, or reproductive development in R. clamitans.

34

CHAPTER TWO
EFFECTS OF ATRAZINE ON METAMORPHOSIS, GROWTH, AND LARYNGEAL
AND GONADAL DEVELOPMENT IN THE AFRICAN-CLAWED FROG (HNOPUS

LAEVIS)

35

Abstract
Eight replicates of 30 African clawed ﬁ'ogs (Xenopus laevis) were exposed to one

of eight waterborne treatments including: 0, 0.1, 1.0, 10, or 25 jig/L atrazine, 0.005%
ethanol (EtOH), or 0.1 mg/L estradiol (E2) or dihydrotestosterone (DHT) containing
0.005% EtOH. Frogs were exposed from 72 h post-hatch until 2 to 3 mo post-
metamorphosis via a 3-d static renewal exposure regimen. Multiple endpoints were
measured in the developing frogs, including time to complete metamorphosis, mortality,
length and weight at metamorphosis, gross morphology of the gonads, and histology of
the larynx and gonads. Male frogs had larger laryngeal dilator muscles than females,
whereas male and female frogs exposed to DHT had larger laryngeal dilator muscle areas
as compared to frogs in all other treatments. Frogs exposed to E2 had greater incidences
of testicular oocytes. Atrazine at concentrations between 0.1 and 25 jig/L did not
signiﬁcantly affect any of the parameters measured in this study. These results suggest
that environmentally relevant concentrations of atrazine do not affect growth,

metamorphosis, or gonadal or laryngeal development of X. laevis.

Study Objectives
The present study was undertaken to evaluate the effects of chronic atrazine

exposure on growth, development, and reproductive morphology of developing X. laevis.
The objectives of this study were to determine the effects of atrazine on growth,
metamorphosis, and reproductive indices of larval X. laevis exposed from 72 hr after
hatching until completion of metamorphosis. Indices evaluated at the completion of

metamorphosis included: number of ﬁogs initiating and completing metamorphosis, time

36

to metamorphosis, body weight, snout-vent length, gonad development, and laryngeal
dilator muscle size. Gonads were examined at the gross and histological levels for
structural anomalies, such as rudimentary hermaphroditism (deﬁned as the presence of

both ovarian and testicular tissue in the same individual).

Test Materials
Atrazine (CAS number 1912-24-9; purity 97.1%) was obtained from Syngenta

Crop Protection Inc. (Greensboro, NC, USA). Estradiol (E2) (CAS number 50-28-2,
purity 98%) and dihydrotestosterone (DHT) (CAS number 521-18-6) were purchased
from Sigma Chemical Co. (St. Louis, MO, USA). Ethanol (EtOH) (CAS number 64-17-
5, purity 200 proof, 100% USP grade) was purchased from AAPER alcohol (Shelbyville,
KY). All chemicals were dissolved in the Frog Embryo Teratogenesis Assay- Xenopus
(F ETAX) test medium comprised of laboratory reverse osmosis water containing salts in
the following concentrations: 0.625 g/L NaCl; 0.030 g/L KCl; 0.015 g/L CaClz; 0.096
g/L NaHC03; 0.060 g/L CaSO4*2H20; and 0.075 g/L MgSO4 (ASTM, 1991). EtOH was
used as a carrier solvent to deliver DHT and E2 treatments in FETAX; thus, the 0.005%

EtOH treatment group served as a solvent control for these hormone exposure groups.

Methods
Exposure methods

Sexually mature X. laevis obtained from Xenopus Express® (Plant City, FL) were
induced to breed in the laboratory by injecting human chorionic gonadotropin into the

dorsal lymph sac. Twelve hours later, fertilized eggs were collected and de-jellied in a

37

2% solution of L-cysteine in FETAX medirun (ASTM, 1991). Viable embryos were
sorted into lots of 100 and left in FETAX medium overnight at 22° C. On the next day,
larvae were randomly assigned with the use of random number tables to glass Petri dishes
(100 mm diameter) containing 20 mL of FETAX (35 embryos per dish). The exposure
period began after the larvae were 72 hr old, at which time all viable larvae were
transferred from glass petri dishes containing FETAX solution and randomly assigned to
10L glass tanks containing 4 L of the appropriate test solution. At the start of the
exposure, approximately 30 X. laevis larvae were placed in each replicate tank. Each
treatment group was replicated eight times, for a total of 64 replicate tanks. Tadpoles
were housed in a windowless environmental chamber with a light exposure regime of 12
hrs of light followed by 12 hrs of darkness.

Test materials were applied in a static renewal exposure regime. Test solutions
were renewed by 50% replacement every 72 hr. The test solution renewal process began
with the removal of 2L of the test solution by siphoning followed by replacement with 2L
of the appropriate fresh test solutions. Each treatment had its own siphon tube designated
by color code to avoid cross-contamination. At the time of test solution renewals, 15 mL
water samples were taken from the stock solutions and the atrazine-treated tanks to test
for actual atrazine concentrations.

Throughout the study, mortalities were monitored and recorded on a daily basis
for each replicate tank. In addition, the number of tadpoles initiating and completing
metamorphosis was recorded daily for each replicate tank. The Nieuwkoop-Faber (NF)
staging method was used to determine when tadpoles initiated and completed

metamorphosis (Nieuwkoop and Faber, 1967). At the initiation of forelimb emergence

38

\\
.
.JC.‘
I / )

(NF stage 58), glass dividers surrounded by a 1 mm mesh material were constructed and
placed into the 10L aquariums, such that tadpoles at different stages of metamorphosis
could be kept physically separate and be easily enumerated. At the time of metamorphic
completion (NF stage 66), time to initiate and complete metamorphosis, snout-vent
length, and weight data were recorded for each metamorphosing frog. A subset of the
frogs was euthanized at NF stage 66 for assessment of gross morphology of the gonads
and for histological analyses of gonad and larynx tissue. Frogs were euthanized by
immersion in MS-222.

The remaining NF stage 66 X. laevis were reared in 4L of test solution until
approximately 1 mo post-metamorphosis, at which time they were transferred to 40L
aquaria containing 10 L of the test solution to increase space and maximize growth.
These frogs were exposed to the various test solutions until 2-3 mo post-metamorphosis.

In total, the exposure period was 185 d (from Dec 21, 2001 — June 24, 2002).

Analytical methods

Concentrations of atrazine in the exposure solution were measured by Enzyme-
Linked Irnmunosorbent Assay (ELISA) at Michigan State University (MSU) and
Syngenta Crop Protection Inc. (Greensboro, NC, USA). At MSU concentrations of
atrazine in fresh renewal batches and in replicate tanks were measured using the
Envirogard® Triazine 96-well plate kit (Strategic Diagnostics Newark, DE; product #
7211000). The method detection limit of the Envirogard® Triazine 96-well plate kit was
0.025 jig/L atrazine. A subset of the treatment solution samples analyzed at MSU was

also analyzed at Syngenta Crop Protection Inc. At Syngenta, concentrations of atrazine

39

were measured with the Beacon Analytical triazine plate kit (Beacon Analytical Systems,
Portland, ME, USA). The method detection limit of the Beacon kit was 0.05 ug/L
atrazine. When the concentrations of atrazine, as measured by ELISA, deviated from
expected concentrations, treatment solution samples were analyzed by gas
chromatography/mass selective detection (GC-MS) at Syngenta Crop Protection Inc. to
verify the results of the ELISA. For GC-MS analyses, treatment solution samples were
buffered to a pH of 10, and analytes were extracted by partitioning into ethyl acetate.
Samples were then evaporated to dryness and reconstituted in acetone. GC-MS analyses
were performed in the selected ion monitoring mode (Yokley and Cheung, 2000).

Aquarium water quality parameters were measured weekly using Larnotte test kits
for total ammonia nitrogen, nitrite nitrogen, and hardness (Aquatic Ecosystems, Apopka,
FL, USA). Dissolved oxygen was measured by a YSI model 57 oxygen meter (Y SI Inc.
Marion, MA, USA), and pH was measured using an Orion research pH meter -model
710A (Therrno Orion Inc., Beverly, MA, USA).

Gross morphology of the gonads was examined in a subset of NF stage 66 X.
laevis and in a subset of frogs that were allowed to develop to 2-3 mo post
metamorphosis. Frogs examined for gross gonadal morphology were euthanized in MS-
222 and preserved in Bouin’s fixative for 48 hr and then transferred to a solution of 70%
ethanol for storage. Frogs were dissected under an Olympus SZ40 stereomicroscope and
gonads were evaluated for sex classiﬁcation and gross gonadal anomalies. Digital
photographs were taken of each specimen. The incidence of gross gonadal anomalies

was enumerated and types of anomalies described.

40

A subset of the frogs examined for gross morphology of the gonads was also
processed for gonad histology. A total of 25 NF stage 66 frogs found to have anomalous
gonadal morphology were sectioned and stained at Texas Tech University (Lubbock, TX,
USA). Gonads were embedded in parafﬁn wax and serially cross-sectioned at 7 urn
intervals. All sections were mounted and stained with eosin and hematoxylin by standard
methods. In addition, a total of 50 frogs per treatment (25 males and 25 females) were
examined for gonad histology 2 -3 mo post-metamorphosis. These ﬁogs were sectioned
and stained at Michigan State University (East Lansing, MI, USA). Gonads were
removed from the frogs, embedded in parafﬁn and serially sectioned at 5 pm intervals.
Every fourth 5 pm section of the gonad was mounted onto a slide, such that every 20 um
was represented in the analysis. Slides were stained with eosin and hematoxylin. All
slides of sectioned ovaries and testes were evaluated under a compound microscope with
the analyst having no knowledge of sample treatment group. When they occurred,
histological anomalies were noted.

NF stage 66 frogs were processed for laryngeal histology at Texas Tech
University as described by Carr et al. (2003). The laryngeal regions were removed from
preserved X. laevis and imbedded in parafﬁn. Specimens were coronally sectioned at 8
pm and stained with hematoxylin and eosin. Every twentieth section of the laryngeal
dilator muscle was digitally photographed and the cross-sectional areas of the laryngeal
dilator muscle was measured using “Simple PCI” software (Compix Inc. Imaging
Systems, Cranberry Township, PA, USA) (Carr et al., 2003). Aﬁer measuring a series of
laryngeal muscle cross-sectional areas for each frog, the largest measured area for the

right and left dilator muscle was selected and used in statistical analyses.

41

Statistical methods

Fresh batches of treatment solutions (before being added to exposure tanks) and
replicate exposure tanks were sampled and analyzed for concentrations of atrazine.
These concentrations were incorporated into the total average atrazine concentration
calculated for that treatment. Measured concentrations of atrazine ﬁ'om replicate tank
samples were averaged across tanks for a single date. This average concentration was
then used as a single data point when calculating treatment averages. For any given
treatment, concentrations in the fresh batches of treatment solutions were indicative of
atrazine exposure for the date that they were used. Non-parametric 95% conﬁdence
intervals were calculated for mean measured concentrations of atrazine by bootstrapping
the data array.

Mortality was calculated based only on the observed number of tadpoles/frogs
that died in each replicate tank. Percent mortality for each treatment tank was calculated
using the Kaplan-Meier method, since frogs were removed from the study at different
times during the exposure (Lee, 1992). This method estimates mortality rates based on
changing numbers of individuals at risk for mortality throughout the study period.

Kolmogorov-Smirnov’s One Sample test with Lillifor’s transformation was used
to assess whether or not data sets were normally distributed. When data were normally
distributed, analysis of variance (AN OVA) and Fisher’s least signiﬁcant difference
(LSD) post hoc test were used to detect signiﬁcant differences between treatment and
control groups. When data were not normally distributed, the non-parametric Kruskal-
Wallis test was used to detect differences among treatment groups. If the Kruskal-Wallis

test showed signiﬁcant differences among the treatment groups, the Mann-Whitney U test

42

was used to evaluate differences between treatment groups. The Chi-square test was used
to detect signiﬁcant deviations in expected sex ratios in each replicate tank. Pearson’s
chi-square was used to detect differences in incidences of gross and histological gonadal
anomalies among treatment groups. The criteria used for signiﬁcance in all statistical
tests was p<0.05. The probability of a type 11 error (l-B), or power, in all statistical tests
was set at 0.8, and detectable differences among treatment groups were calculated based
on sample size and the average variation of the parameter under investigation.

Data sets were analyzed such that treatment groups were compared to the
appropriate control treatment in the experimental design. Atrazine treatments were
compared to the untreated controls. E2 and DHT treatments were compared to the EtOH

solvent controls.

Results
Atrazine exposure verification

Atrazine concentrations measured by ELISA in the control, 0.1, 1.0, 10, and 25
ug/L atrazine treatments were greater than expected (Table 1). Similar results were
found when a subset of the treatment solution samples was analyzed by immunoassay at
Syngenta Crop Protection Inc. (Table 1). To verify the concentrations of atrazine as
measured in the immunoassays, several treatment solution samples were compiled and
atrazine concentrations were measured by GC-MS at Syngenta Crop Protection Inc.
Concentrations of atrazine ranging from <0.05 to 0.27 jig/L were detected in the

untreated control solution when measured by GC-MS. Two composite treatment solution

43

samples from the 25 jig/L atrazine treatment were analyzed by GC-MS. The measured

concentrations of atrazine for these composite samples were 33 and 35 jig/L.

Table 1: Nominal and measured concentrations of atrazine in test solutions as measured
by Enzyme-Linked Irnmunosorbent Assay at Michigan State University (MSU) and

Syngenta Crop Protection Inc. a

 

 

Treatment Nominal MSU Syngenta
concentration Average measured Average measured
of atrazine concentrations of atrazine concentrations of atrazine
(Hg/L) (rig/L)b (lug/L)"
Control 0 0.16 (0.11 — 0.20) 0.11(0.07-0.15)
EtOH 0 0.01 (0.01-0.01) Not measured
DHT 0 0.02 (0.01-0.02) Not measured
E2 0 0.04 (0.03-0.04) Not measured
0.1 jig/L 0.1 0.22 (0.17-0.26) 0.23 (0.16 — 0.31)
atrazine
1.0 rig/L 1.0 1.0 (9.3x10'l—1.1) 1.4 (1.2-1.7)
atrazine
10 jig/L 10 16.4 (14.1-18.9) 11.4 (9.72-13.3)
atrazine
25 jig/L 25 28.9 (24.2-35.1) 25.1 (22.3-28.1)
atrazine

 

EtOH = 0.005 % ethanol; DHT = 0.1 mg/L dihydrotestosterone; E2 = 0.1 mg/L estradiol

3 For these calculations, a proxy value of either 0.0125 or 0.025 jig/L atrazine was assigned to values less

than the ELISA method detection limits of either 0.025 or 0.05 jig/L atrazine, respectively.

b 95% conﬁdence intervals in parentheses

44

Water quality

Water temperatures over the 185 (1 exposure ranged from 17 to 23 °C. The mean
water temperature (i standard deviation) during the 185 d exposure was 20 i 1.1 °C. The
mean nitrite nitrogen and total ammonia nitrogen concentrations (:1: standard deviation) in
exposure water were 0.47 i 0.78 and 0.23 :l: 0.41 mg/L, respectively. The mean
dissolved oxygen concentration (i standard deviation) was 7.4 i 0.8 mg/L. The mean
hardness of the exposure water (1 standard deviation) was 138 i 15 mg/L CaCO3, and the

mean pH (i standard deviation) was 7.6 i 0.3.

Mortality

The number of surviving frogs at risk for mortality changed during the exposure
period due to either euthanization of some tadpoles failing to initiate metamorphosis or
scheduled subsampling of NF stage 66 frogs. In addition, ﬁve frogs eScaped from the
tanks. Therefore, the number of surviving frogs was taken into account when calculating
percent mortality for each replicate tank. Mortality in each replicate tank was averaged
to calculate overall percent mortality for each treatment group. Average percent
mortality in the control treatment was 11.3%. Average percent mortality in the 0.1, 1.0,
10, and 25 jig/L atrazine treatments was 20.4, 11.5, 18.7 and 20.3%, respectively.
Average percent mortality in the EtOH, DHT, and E2 treatments was 14.4, 21.3 and
8.3%, respectively. The total average observed mortality across all treatment groups was
16.1%. There were no statistical differences in observed mortality among treatments

(n=8 tanks/ treatment) (Kruskal-Wallis, p= 0.298). Given the sample sizes and variation

45

in mortality in this experiment, a difference of 21.7 % mortality was detected between

treatments with 80% power (l-B).

Metamorphosis

At completion of metamorphosis (NF stage 66) age, snout-vent length, and weight
were determined for each frog. Due to tank effects, the experimental units in these
analyses were the mean tank values for each measured parameter (n=8/treatment). In the
entire study, a total of 17 surviving tadpoles did not initiate metamorphosis. Therefore,
98.9% of surviving frogs initiated metamorphosis during the study period. There were no
statistically signiﬁcant differences in age at completion of metamorphosis among
treatment groups when treatments were compared to the appropriate controls (AN OVA
atrazine and control treatments, p= 0.986; ANOVA E2, DHT and EtOH treatments, p=
0.703). Given the sample sizes and variation in age at metamorphosis in this experiment,
an age difference of 5.25 d was discemable between treatments with 80% power (1-0).
The average age at metamorphic completion across all treatments was 72.8 d with a
standard deviation (SD) of 8.19 d.

There were no statistically signiﬁcant differences in snout-vent length at
completion of metamorphosis when treatment groups were compared to the appropriate
control treatment (ANOVA atrazine and control treatments, p= 0.066; AN OVA E2, DHT
and EtOH treatments, p=0.512). In this experiment, a 0.13 cm difference in length was
discemable between treatments at 80% power. A pair-wise comparison of snout-vent
lengths between EtOH and untreated control frogs showed that EtOH frogs were longer

(mean 1: SD = 1.89 i 0.15 cm) than untreated control frogs (mean i SD = 1.75 i: 0.20

46

cm; t test, p= 0.032). The mean snout-vent length at completion of metamorphosis across
all treatment groups was 1.85 cm (SD = 0.17 cm).

There were no signiﬁcant differences in weight at completion of metamorphosis
when treatment groups were compared to the appropriate control treatment (AN OVA
atrazine and control treatments, p=0.220; ANOVA E2, DHT and EtOH treatments, p=
0.311). Given the sample sizes and variation in weight in this experiment, a difference of
0.19 g was detected between treatments with 80% power. A pair-wise comparison found
that EtOH -treated ﬁ'ogs were signiﬁcantly heavier (mean :1: SD = 0.85 i 0.23 g) than the
untreated controls (mean i SD = 0.70 i 0.24 g; t test, p= 0.046). The total average

weight at metamorphic completion across all treatment groups was 0.78 g (SD =0.22 g).

Gross gonadal anomalies

Three types of gross gonadal anomalies were observed: discontinuous gonad,
rudimentary hermaphroditism, and size irregularity. Discontinuous gonad was
characterized by abnormal segmentation of the gonadal tissue. Rudimentary
hermaphroditism was characterized by the appearance of immature testicular and ovarian
tissue in a single individual (Van Tienhoven, 1983). Two types of rudimentary
hermaphroditism were observed in juvenile X. laevis. The gonads of some individuals
contained masses of ovarian and testicular tissue separated left/right or rostral/caudal.
This type of rudimentary hennaproditism was termed intersex. In other hermaphroditic
individuals, there was co-occurrence of both ovarian and testicular tissue in a single

gonad, and this was termed mixed sex. Size irregularity was characterized either by large

47

size discrepancies between gonad pairs or unusually large or small gonads (Figure 1).
Other gonadal anomalies, regarding shape and pigmentation, were noted as well.

The most commonly occurring gross gonadal anomaly in the stage 66 frogs was
discontinuous gonad (Table 2). There were no statistically signiﬁcant differences in the
incidence of gross gonadal anomalies among treatments for NF stage 66 frogs (Chi-
square). Gross gonadal anomalies observed in individuals 2-3 mo post-metamorphosis
were similar to those observed in stage 66 frogs, with the most commonly observed gross
gonadal anomaly being discontinuous gonad (Table 3). No statistically signiﬁcant
differences in the incidence of gross gonadal anomalies were observed among treatments

when frogs were 2—3 mo post-metamorphosis (Chi-square).

48

 

Figure 1: Normal and abnormal gonadal morphology of juvenile male and female X.
laevis 2-3-months post metamorphosis. a) Normal ovaries b) Normal testes c) Mixed sex
gonad from a 10 ug/L atrazine-exposed frog d) Discontinuous testes from a 0.1 ug/L
atrazine-exposed frog e) Intersex gonad in a frog from an estradiol-exposed frog f) Gonad

size irregularity in a frog from the 10 [lg/L atrazine treatment

49

Table 2: Percent gross gonadal anomalies in Nieuwkoop-Faber stage 66 X. laevis

exposed to EtOH, DHT, E2, or various concentrations of atrazine.

 

 

 

 

Treatment Group N % % Rudimentary Hermaphrodites % Size % Other
Discontinuous % Mixed Sex % Intersex Irregularities Anomalies
Gonads

Control 45 2.2 0.0 0.0 2.2 2.2
EtOH 45 0.0 0.0 0.0 0.0 2.2
DHT 42 4.8 2.4 0.0 2.4 4.8
E2 46 6.5 4.4 2.2 0.0 2.2
0.1 ug/L atrazine 40 5.0 0.0 0.0 0.0 7.5
1.0 jig/L atrazine 46 2.2 0.0 0.0 2.2 0.0
10 jig/L atrazine 43 7.0 0.0 0.0 0.0 4.7
25 ug/L atrazine 39 5.1 2.6 0.0 0.0 0.0

 

EtOH = 0.005 % ethanol; DHT = 0.1 mg/L dihydrotestosterone; E2 = 0.1 rug/L estradiol

Table 3: Percent gross gonadal anomalies in X. laevis exposed to EtOH, DHT, E2, or

various concentrations of atrazine until 2-3 mo post-metamorphosis.

 

 

Treatment Group N % Discontinuous % Rudimentary % Size Irregularities
Gonads Hermaphrodites (mixed sex)

Control 75 1.4 0.0 2.7
EtOH 75 2.7 0.0 1.3
DHT 72 1.4 0.0 0.0
E2 77 2.6 0.0 0.0
0.1 jig/L atrazine 71 4.2 0.0 0.0
1.0 jig/L atrazine 79 1.3 0.0 2.5
10 pg/L atrazine 73 4.1 2.7 0.0
25 jig/L atrazine 67 3.0 0.0 0.0

 

EtOH = 0.005 % ethanol; DHT = 0.1 mg/L dihydrotestosterone; E2 = 0.1 mg/L estradiol

50

Histological gonadal anomalies

The most common gonadal anomaly among NF stage 66 frogs at the histological
level of observation was rudimentary hermaphroditism (Van Tienhoven, 1983). At the
histological level, rudimentary hermaphroditism was characterized either by
rostral/caudal or left/right separation of testicular and ovarian tissue (intersex) or by the
occurrence of testicular oocytes. Testicular oocytes were coded as such if the oocytes
had an intact nucleus, nucleoli within the nucleus, and a surrounding squamus epithelial
layer. Rudimentary hermaphroditism was observed in four NF stage 66 individuals, all
of them in the E2 treatment group. In addition, one control frog appeared to have
unidentiﬁed tissue surrounding the testes that appeared abnormal.

When the gonads from frogs exposed to the treatments for 2-3 mo post-
metamorphosis were examined at the histological level, a greater percentage of
rudimentary herrnaphrodites, including intersex gonads (separated ovarian and testicular
tissue) and testicular oocytes, were observed (Table 4). Testicular oocytes occurred in
male frogs from all treatments, while intersex gonads were observed in male frogs from
the control and EtOH treatments, as well as the 0.1 and 1.0 ug/L atrazine treatments
(Table 4). In general, when testicular oocytes were observed, only one or two oocytes
were present in the entire testis (Figure 2). However, several gonads in the E2 treatment
and one gonad in thelO jig/L atrazine treatment had greater numbers of oocytes mixed in
with testicular tissue (Figure 2). Few histological anomalies were noted during gonadal
examinations from frogs classiﬁed as females during gross gonadal inspections. Most of
the anomalies noted in female frogs were small or underdeveloped ovaries containing

relatively few or no eggs. Two ﬁ'ogs (8.0%) in the E2 treatment classiﬁed as females

51

from gross gonadal inspections had testicular tissue containing oocytes. These were most
likely genetic males developing female gonads in response to E2 exposure (Witchi,
1967).

There were no signiﬁcant differences in the occurrence of any histological
gonadal anomaly observed among control and atrazine treated frogs (Chi-square).
However, during histological examinations, a signiﬁcant difference in the occurrence of
testicular oocytes was observed between DHT and EtOH treated frogs (Chi-square,

=0.039). DHT-treated frogs had signiﬁcantly fewer gonads with testicular oocytes than

did frogs exposed to EtOH (Table 4).

Table 4: Percent gonadal anomalies at the histological level in male X. laevis exposed to

EtOH, DHT, E2, or various concentrations of atrazine until 2-3 mo post-metamorphosis.

 

 

 

Treatment Group N % Rudimentary Hermaphrodites % Other Anomalies
% Testicular Oocytes % Intersex gonads

Control 25 8.0 16.0 0.0
EtOH 25 20.0 4.0 0.0
DHT 25 "' 4.0 0.0 0.0
E2 25 32.0 0.0 0.0
0.1 jig/L atrazine 25 12.0 4.0 0.0
1.0 jig/L atrazine 25 8.0 4.0 0.0
10 jig/L atrazine 25 12.0 0.0 0.0
25 jig/L atrazine 25 8.0 0.0 0.0

 

EtOH = 0.005 % ethanol; DHT = 0.1 mg/L dihydrotestosterone; E2 = 0.1 mg/L estradiol

"‘ Indicates a signiﬁcant difference from the EtOH control at p< 0.05.

52

:; testicular
. . .o'ocyte

   

 

'k‘m‘M '1

t , “4,11.-

Figure 2: Cross section of rudimentary hermaphroditic gonads of 2-3 mo post-
metamorphic X. laevis. a) Intersex gonad from an ethanol-exposed frog b) Testicular
oocytes from an estradiol-exposed frog c) Testicular oocytes from a control frog (1)
Testicular oocytes from a 10 ug/L atrazine exposed frog (Bar in each picture represents

200 um)

Sex ratios
There were no consistent deviations from the expected 50:50 sex ratio (Chi-square).
Several tanks did have a signiﬁcantly different sex composition from the expected 50:50

ratio, but the ratio was not consistently skewed in favor of one sex over the other, and

53

varied from tank to tank. One EtOH-exposed tank, one DHT-exposed tank, and two 0.1
ug/L atrazine-exposed tanks had skewed sex ratios in favor of male frogs. One EtOH -
exposed tank and one E2-exposed tank had skewed sex ratios in favor of female ﬁ'ogs.
There were no signiﬁcant differences in % females or % males among atrazine
treated tanks and untreated control tanks (ANOVA, p= 0.108 and 0.137, respectively).
Likewise, there were no signiﬁcant differences in % females or % males among the DHT,
E2, and EtOH treatments (ANOVA, p= 0.111 and 0.232, respectively). In this
experiment, a sex ratio difference of 18% was discemable among treatments with 80%

power.

Laryngeal dilator muscle area

Overall, male frogs had laryngeal muscle dilator cross-sectional areas that were
signiﬁcantly greater than those of females (Mann-Whitney U, p= 0.0001). In most
treatment groups, laryngeal dilator muscles in males were larger than females (Figure 3).
Due to the statistically signiﬁcant differences in muscle area between the two sexes, the
data was stratiﬁed by sex for subsequent statistical analyses.

There were no statistically signiﬁcant differences in laryngeal muscle area among
male frogs in the atrazine treatments and the untreated controls (Kruskal-Wallis, p=
0.476; Table 5). Given the sample sizes and variation in male laryngeal dilator muscle
area in this experiment, a difference of 0.08 mm2 was detected between treatments with
80% power. A signiﬁcant difference in male laryngeal muscle area was detected among

the E2, DHT, and EtOH treatments (Kruskal-Wallis, p=0.008). Pair-wise comparisons

54

revealed that male frogs exposed to DHT had greater laryngeal muscle area as compared
to males in all other treatment groups (Figure 3).

There were no signiﬁcant differences in laryngeal muscle area among female
frogs in the atrazine treatments and the untreated controls (Kruskal-Wallis, p= 0.181;
Table 5). Given the sample sizes and variation in female laryngeal dilator muscle area in
this experiment, a difference of 0.12 mm2 was detected between treatments with 80%
power. Female frogs exposed to DHT had signiﬁcantly greater laryngeal muscle area
(Kruskal-Wallis, p=0.0001) than did females in all other treatment groups (Figure 3).
The laryngeal muscle area of female frogs exposed to DHT was not signiﬁcantly
different from the muscle area of male frogs exposed to DHT (Mann-Whitney U, p=

0.815).

Table 5: Laryngeal dilator muscle cross-sectional areas for male and female X. laevis

exposed to EtOH, DHT, EtOH, or various concentrations of atrazine

 

 

Treatment N Mean Left Standard N Mean Left Standard
Males Laryngeal Muscle error Females Laryngeal error
Area in Males Muscle Area in
(mmz) Fermles (mmz)

Control 13 0.214 0.009 10 0.153 0.01
EtOH 14 0.192 0.020 5 0.185 0.028
DHT l4 "‘ 0.304 0.022 10 "' 0.346 0.068

E2 7 0.223 0.023 10 0.147 0.008
0.1 ug/L atrazine 14 0.212 0.024 9 0.211 0.025
1.0 jig/L atrazine 14 0.228 0.011 9 0.166 0.012
10 ug/L atrazine 14 0.217 0.013 8 0.147 0.013
25 ug/L atrazine 13 0.204 0.013 9 0.152 0.007

 

EtOH = 0.005 % ethanol; DHT = 0.1 mg/L dihydrotestosterone; E2 = 0.1 mg/L estradiol

* Indicates a signiﬁcant difference from all other treatment groups at p < 0.05.

55

 

      

 

 

 

.305 ,. ' ‘m ' TTT T FTC—11 'TT—
E DFemales

{73’ 0.4 - ‘W’ ‘

I: 0.3 l

, 8

l; 0.2

lg 01 ~- I

1% '

1 0.1ug/L ETOH our 52

Z atrazine

 

Figure 3: Laryngeal dilator muscle areas in Nieuwkoop-Faber stage 66 X. laevis exposed
to EtOH, DHT, E2, and various concentrations of atrazine (Bars represent mean :1:

standard error of the mean)

EtOH = 0.005% ethanol; DHT = 0.1 mg/L dihydrotestosterone; E2 = 0.1 ug/mL estradiol

"' indicates a signiﬁcant difference from all other treatments, p<0.05

Discussion
Results are discussed in terms of the nominal exposure concentrations. However,

the actual concentrations of atrazine measured by immunoassay at Michigan State
University were slightly greater than the nominal concentrations. Concentrations of
atrazine measured by immunoassay at Syngenta Crop Protection Inc. were similar to
concentrations measured by ELISA at Michigan State University. On occasion, the
ELISAs indicated that there was a small amount of atrazine in the control tanks. Atrazine
was detected in control treatment solution at an average value ranging from 0.11 to 0.16
jig/L atrazine. When a subset of control treatment solution samples was analyzed by GC-
MS, atrazine was detected at similar concentrations. The presence of atrazine in control

tanks may be explained by aerial transport and deposition of droplets of treatment

56

solutions produced by bubbling air stones from nearby tanks. It is also possible that
small amounts of atrazine were periodically present in laboratory RO water used in the
FETAX solution. Even though small concentrations of atrazine were detected in
untreated control tanks, this does not affect the conclusions of the study, since there were
no statistically signiﬁcant, dose-dependent effects of atrazine observed on any of the
parameters investigated. The fact that there were no measurable concentrations of
atrazine in the ethanol control tanks allows it to be used as a no-atrazine control
treatment. When this was done, there was still no dose-response relationship for any of
the measurement endpoints. This observation argues against any atrazine related effects.

The observation that atrazine, under the conditions and at the doses used in this
experiment, did not have signiﬁcant effects on mortality of X. laevis is consistent with
other amphibian toxicity assays (Morgan et al., 1996; Battaglin and F airchild, 2002).
Atrazine causes mortality in amphibians at much greater concentrations (LCso 2 410 jig/L
atrazine) than those used in this study (Battaglin and Fairchild, 2002).

The observation that chronic exposures to atrazine at concentrations between 0.1
and 25 jig/L did not affect age, length, nor weight at completion of metamorphosis is
consistent with the results of other studies that have also observed no effect of atrazine on
these parameters at similar concentrations (Allran and Karasov, 2000; Diana et al., 2000).
In an exposure study with larval gray tree frogs (Hyla versicolor), atrazine did not affect
tadpole growth at environmentally relevant concentrations (Diana et al., 2000).

Frogs exposed to EtOH were longer and heavier than those in untreated control
tanks. This may have resulted from increased availability of food in ethanol containing

tanks. Ethanol in exposure tanks appeared to aggregate food particles in the water

57

column. This was observed by the noticeably greater murkiness in EtOH -containing
tanks.

In this study, exposure to environmentally relevant concentrations of atrazine did
not affect the incidence of either gross or histological gonadal anomalies in post-
metamorphic X. laevis. However, other researchers have linked atrazine exposure with
the demasculinization of gonads in male X. laevis (Hayes et al., 2002; Hayes et al., 2003).
The results from this study indicate that gonadal anomalies in X. laevis exposed to
atrazine until they were 2-3 mo post-metamorphosis are not related to atrazine
concentration. The lack of a signiﬁcant concentration-response relationship and the
presence of testicular oocytes and intersex gonads in EtOH controls (containing no
measurable concentrations of atrazine) suggest that these types of deformities may not be
related to atrazine exposure, but rather are a component of normal ontological
development. Hennaphroditism occurs normally in other anurans (W itchi, 1929; Hsu
and Liang, 1970; Grarnapurohit et al., 2000) and therefore may also be a normal
occurrence in some strains of developing X. laevis. Further research into the histological
morphology of X. laevis gonads is necessary to determine the background levels of
rudimentary hermaphroditism in juvenile frogs under both laboratory and ﬁeld
conditions.

The results of this study are in contrast to the studies of Hayes et al. (2002), who
also studied the effects of atrazine on growth, metamorphosis, and gonadal and laryngeal
development in X. laevis. It is not possible to directly compare the results obtained in our
study with those of Hayes et a1. (2002) due to differences in the age and strain of the

examined frogs. In the study by Hayes et al. (2002), X. laevis were exposed to atrazine

58

until metamorphic completion (NF stage 66), but no X. laevis were exposed to atrazine 2-
3 mo post-metamorphosis and then examined for gonadal anomalies. In addition,
different strains of X. laevis were used in the two studies, and gonadal development and
metamorphosis are not necessarily synchronized in different frog strains (Gramapurohit
et al., 2000). As a result, the differences between our study and that of Hayes et al.
(2002) could be explained by differences in the degree of gonadal development in the
examined frogs.

The observation that exposure to atrazine did not affect sex ratios is consistent
with the results of other studies with X. laevis (Carr et al., 2003). While it was expected
that E2 exposed X. laevis would have skewed sex ratios in favor of increased numbers of
females, this phenomenon was not observed in this study. Some researchers have
reported 90-100% females when X. laevis were exposed to similar concentrations of E2
(Miyata et al., 1999; Miyata and Kubo, 2000; Hayes et al., 2002). However, exposure to
E2 under a similar regime as used in the present study produced a proportion of female
frogs that was similar to the present study (Carr et al., 2003). The balanced sex ratio in
most E2-exposed tanks in this study may have resulted from rapid metabolism and
breakdown of the E2. Since E2 was only applied every third day during the exposures, it
is possible that the hormone was metabolized too rapidly by the frogs and therefore was
below the threshold concentration required to cause altered sexual development. If E2
was administered daily or in a ﬂow-through system, the results may have been different.
E2 exposure was used as a positive control for possible estrogen receptor mediated

effects, but was not used to model the fate of waterborne atrazine treatments.

59

Laryngeal development in ﬁ'ogs is a sexually dimorphic process, and the
formation of a larynx capable of male calling behavior is androgen-dependent. Under
normal conditions the laryngeal dilator muscle of male X. laevis is larger than that of
females (Fischer et al., 1995; Tobias et al., 1991; Tobias et al., 1993). It has been
hypothesized that atrazine could decrease plasma concentrations of testosterone (T) in X.
laevis by up-regulating the expression of aromatase, the enzyme that converts T to E2
(Hayes et al., 2002). This decrease in available androgen for conversion into DHT could
then result in a decrease in laryngeal dilator muscle volume (Hayes et al., 2002). Thus,
dilator muscle size was selected as an indicator of disruption of plasma hormone balance,
especially of androgens during development. Theoretically, this measurement endpoint
can serve as an integrating measure of androgen-dependent processes that would respond
to subtle changes in androgen status during critical periods of development when it is not
possible to measure plasma hormone concentrations. Such a measure should be able to
detect changes that occur in small windows during development as well as changes in
androgen status in tissues that might not be observed in the plasma.

The observation that exposure to atrazine had no signiﬁcant effect on the size of
the laryngeal dilator muscle size in either males or females is consistent with a similar
study that observed no effect on laryngeal dilator muscle size (Carr et al., 2003). Unlike
the ﬁndings of this study, a decreased larynx muscle size in atrazine exposed X. laevis
has been previously reported (Hayes et al., 2002). However, in that study, there was no
concentration-dependent response and the relationship was only elucidated when the
proportions of individuals greater than the controls was investigated, instead of

investigating the central tendencies or variances in more standard statistical assessments.

60

As expected, male X. laevis in this study had greater larynx muscle size than females.
Because the DHT treatment resulted in statistically greater laryngeal muscle sizes for
both male and female X. laevis, it can be concluded that the frogs were responsive to
changes in androgen status. Thus, the response to DHT was as would be expected.
Therefore, if atrazine was acting as an androgen agonist or was increasing androgen

levels, we would have expected to detect a change in the laryngeal muscle size.

Conclusion

In summary, chronic exposure to environmental concentrations of atrazine did not
affect mortality, growth, sex ratios, laryngeal dilator muscle area, or incidence of gonadal
anomalies in developing X. laevis. Alternative ﬁndings resulting from other studies could
be due, in part, to differences in the exposure regime, X. laevis age or strain, larynx
muscle area measurement techniques, diet, or exposure conditions. However, based on
the ﬁndings in these studies, it appears that exposure to atrazine at environmentally

relevant concentrations does not disrupt development in juvenile X. laevis.

61

CHAPTER THREE
EFFECTS OF ATRAZINE ON AROMATASE ACTIVITY AND PLASMA SEX

STEROID LEVELS IN AFRICAN-CLAWED FROGS (XENOPUS LAEVIS)

62

Abstract

Eight replicates of thirty African clawed frogs (Xenopus laevis) per replicate were
exposed to one of eight waterborne treatments including: 0, 0.1, 1.0, 10, or 25 ug/L
atrazine, 0.005% ethanol (EtOH), or 0.1 mg/L 17B-estradiol (E2) or dihydrotestosterone
(DHT) carried in 0.005% EtOH. Frogs were exposed ﬁ‘om 72 hr post-hatch until 2 to 3
mo post-metamorphosis. Exposures were via a 3-d static renewal exposure regimen.
Aromatase activity in brain and gonadal tissue and circulating concentrations of both E2
and testosterone (T) were examined in juvenile frogs at the end of the exposure. Gonadal
aromatase activity was statistically greater in female frogs as compared to male frogs.
There were no atrazine-related effects on aromatase activity in the brains or gonads in
either male or female X. laevis. However, exposure to E2 decreased gonadal aromatase
activity in female frogs and increased brain aromatase activity in male frogs. Female X.
laevis had greater plasma E2 concentrations as compared to male frogs. Male ﬁogs
exposed to 1.0 ug/L atrazine had statistically lower concentrations of E2 in their plasma
as compared to controls; however, this effect was not consistent across atrazine doses. In
addition, male frogs exposed to E2 had statistically greater concentrations of E2 in their
plasma as compared to controls. There were neither sex nor treatment-related effects on
plasma T concentrations due to atrazine exposure. We conclude that at environmentally
relevant concentrations, atrazine does not affect aromatase activity or plasma

concentrations of E2 or T in juvenile X. laevis.

63

Study Objectives

The objectives of this study were to determine the effects of atrazine on 1) brain
and gonadal aromatase activity and 2) concentrations of circulating T and E2 in control
and atrazine-exposed X. laevis reared under laboratory conditions.
Methods
Test materials and exposure methods

X. laevis were exposed to one of eight treatment solutions from 72 hrs post-hatch
until 2-3 mo post-metamorphosis. Treatment solutions were dissolved in the Frog
Embryo Tertatogenesis Assay-Xenopus (FETAX) medium (ASTM, 1991). The treatment

groups were 0 (Control), 0.1, 1.0, 10, and 25 ug/L atrazine, 0.005% ethanol (EtOH), 0.1

mg/L dihydrotestosterone (DHT), or 0.1 mg/L 17B-estradiol (E2). DHT and E2 were
delivered to FETAX medium in EtOH, so the EtOH treatment group functioned as a
solvent control for the DHT and E2 treatments. Each treatment was replicated in eight
tanks, for a total of 64 tanks in the entire study. Test materials were applied in a static
renewal exposure regime. Test solutions were renewed by 50% replacement every 72 hr.
The light cycle during the exposure was 12 hrs of light followed by 12 hrs of darkness.
Throughout the study, treatment solution samples were collected to test for actual
atrazine concentrations. In total, the exposure period was 185 d (from Dec 21, 2001 —

June 24, 2002).

Sample collection methods
At the termination of the exposure period, a subset of the ﬁogs was selected for

blood and tissue collection, with the use of random number tables. Sixty-four individuals

64

(one from each replicate tank) were randomly selected, and necropsies were conducted
between June 11 and June 24, 2002. To control for circadian cycles in circulating

hormone levels, individual frogs were sampled within a 3 hr time window from 8:30 to
11:30 am. Blood, as well as gonad and brain tissue, were collected to test for treatment

effects on sex steroid hormone concentrations and aromatase activity, respectively.

Blood collection and sex steroid measurements

Frogs were anesthetized by immersion in tricaine methanesulfonate (MS-222).
As soon as frogs were anesthetized, blood was collected by cardiac puncture. Heparin
coated syringes equipped with 25G 5/ 8 needles were used in the blood collection. Blood
samples were kept on ice for several hours and then centriﬁrged at 11,000g to separate the
plasma. Plasma samples were stored at —80 °C until extraction for sex steroids. Plasma

samples (50 uL) were extracted twice with 2.5 mL diethyl ether. Ether fractions were

nitrogen evaporated to dryness and the hormone extract was re-constituted in 250 pL
phosgel buffer (1.4375 g NaZHPO4, 0.32 g NaHzPO4, 0.25 g gelatin, and 0.025 g
thimersol in 250 mL nanopure water) for use in the enzyme-linked immunosorbent assay
(ELISA). Ten pL of 0.0002 pCi/ttL 3H-labeled testosterone [1,2,6,7-3H(N)] (Perkin
Elmer Life Sciences, Boston, MA, USA) was dosed into each plasma sample before
extraction. Following the extraction procedure, the radioactivity in a fraction of the ﬁnal
extract was quantiﬁed in a liquid scintillation counter to test for recoveries.
Concentrations of T and E2 in plasma extracts were measured by competitive ELISA as
described by Cuisset et al. (1994) with modiﬁcations (Hecker, 2002). In the competitive

ELISA, steroids of the sample extract compete with acetylchloinesterase-labeled steroid

65

for the binding site on the polyclonal rabbit anti-steroid antibody. Antiserum to T was
obtained from Dr. D.E. Kime (Shefﬁeld, UK). Cross reactivities of the T antiserum are
described in Nash et al. (2000). The antiserum to E2 (Cayman Chemical product #
482250, Ann Arbor, MI) was reported to cross-react with estradiol-3-glucoronide (17%),
estrone (4%), estriol (0.57%), T (0.1%) and Son-dihydrotestosterone (0.1%). For all other
steroids, cross-reactivities were reported to be less than 0.1%. The steroid ELISAs were

performed using 96-well COSTAR high binding plates (Corning Inc, product # 9018).

Tissue collection and aromatase activity measurements

Following blood collection and euthanasia, both gonads were removed and ﬂash
frozen in liquid nitrogen. Frog carcasses were then stored at -80°C until frozen solid. At
that point, the entire brain mass was removed from the frog carcass by slicing open the
cranium of the frozen frogs using a razor blade. Brain tissue was then. ﬂash frozen in
liquid nitrogen.

The tritium labeled water release assay was used to measure aromatase activity in
brain and gonad tissue (Lephart and Simpson, 1991). Gonad tissue was removed from
liquid nitrogen storage, and homogenized in 600-900 11L of buffer containing 50 mM
KPO4, 1 mM EDTA, and lOmM glucose 6-phosphate. Brain tissue was removed from
liquid nitrogen storage and homogenized in a similar buffer containing 10 mM KPO4, 100
mM KCl, IOmM dithiothreitol, 1 mM EDTA, and 10 mM glucose 6-phosphate. A 500
uL aliquot of the tissue homogenate was incubated with 395 uL 10-3H-androstenedione
(Perkin Elmer Life Sciences, Boston, MA, USA), 100 uL of 10 mM NADP, and 5 uL

glucose 6-phosphate dehydrogenase (100 IU/ml) for 2 hrs at 37 °C. After incubation,

66

samples were centrifuged for 2 min at 11,000g, and the supernatant was split into two
200 uL fractions. These ﬁactions were extracted with 500 11L chloroform by vortex
mixing for 1 min. Samples were then centrifuged for 2 min. at 11,000g. Following
centrifugation, 100 uL of the supernatant was vortexed for 1 min with 100 uL of a
dextran-coated charcoal suspension (5% charcoal and 0.5% dextran) to remove remaining
3H-labeled androstenedione substrate. Samples were then centrifuged for 15 min at

1 1,000g. A 125 uL-aliquot of the supernatant was added to 4 mL of scintillation cocktail
and quantiﬁed in a liquid scintillation counter. Protein content was quantiﬁed for each
tissue homogenate (Bradford, 1976). Aromatase activity for gonad tissue was reported as
fmol/h/mg protein, and aromatase activity for brain tissue was reported as pmol/h/mg

protein.

Statistical methods

Kolmogorov-Smirnov’s One Sample test with Lillifor’s transformation was used
to assess whether or not data sets were normally distributed. When data were normally
distributed, analysis of variance (AN OVA) and Fisher’s least signiﬁcant difference
(LSD) post hoc test were used to detect signiﬁcant differences between treatment and
control groups. When data were not normally distributed, the non-parametric Kruskal-
Wallis test was used to detect differences among treatment groups. If the Kruskal-Wallis
test showed signiﬁcant differences among the treatment groups, the Mann-Whitney U test
was used to decipher differences between treatment groups. The criteria used for
signiﬁcance in all statistical tests was p<0.05. The probability of a type II error (l-B), or

power, in statistical tests was set at 0.8, and detectable differences among treatment

67

groups were calculated based on sample size and the average variation of the parameter
under investigation.

Data sets were analyzed as two separate groups such that treatment groups were
compared to the appropriate control treatment. The atrazine treatments were compared
against the untreated control treatment. The DHT and E2 treatments were compared

against the EtOH solvent control.

Results
Atrazine concentrations

The concentrations of atrazine in the exposure tanks were measured by ELISA at
Michigan State University and Syngenta Crop Protection Inc. In short, measured
concentrations of atrazine were close to nominal concentrations, however, small amounts
of atrazine were detected in the control treatment solution from time to time. At
Michigan State University an average measured concentration of 0. l6 :1: 0.17 ug/L
atrazine was detected in the control solution. When measured by GC-MS, atrazine was
detected in control solutions at concentrations ranging from <0.05 to 0.27 jig/L. Average
measured concentrations of atrazine in the nominal 0.1, 1.0, 10, and 25 jig/L atrazine
treatments were 0.22 i 0.13, 1.0 i 0.30, 16.4 i 7.23, and 28.9 i 16.7 jig/L atrazine,
respectively. Average measured concentrations of atrazine in the EtOH, DHT, and E2
treatments were 0.01 i 0.00, 0.02 i 0.01, and 0.04 i 0.03, respectively. At Syngenta
Crop Protection Inc. average measured concentrations of atrazine in the nominal 0, 0.1,
1.0, 10, and 25 ug/L atrazine treatments were 0.11 i 0.08, 0.23 i 0.13, 1.44 i 0.42, 11.4

i 2.91, and 25.1 i 4.77 ug/L atrazine, respectively.

68

Aromatase activity

Aromatase activities in the gonads of juvenile female X. laevis were signiﬁcantly
greater than in males (Mann-Whitney U, p = 0.0001; Figure 1). The mean (i SD)
ovarian aromatase activity was 420 i 486 fmol/h/mg protein, while the mean (i SD)
testicular aromatase activity was 10.8 i 52.8 fmol/h/mg protein. There was no detectable
aromatase activity in 77% of the male frogs examined.

There were no statistically signiﬁcant differences in testicular aromatase activity
among male frogs exposed to atrazine and untreated control frogs (Kruskal-Wallis, p =
0.07 5). Neither were there statistically signiﬁcant differences in testicular aromatase
activity among DHT, E2 and EtOH -exposed frogs (Kruskal-Wallis, p = 0.382; Figure 1;
Table 1). Given the sample size and variability in testicular aromatase activity, a
difference of 95.5 ﬁnol/h/mg protein was discemable among treatments with 80% power.

There were no statistically signiﬁcant differences in ovarian aromatase activity
among atrazine treatments and untreated controls (Kruskal-Wallis, p = 0.821). Given the
sample size and variability in ovarian aromatase activity, a difference of 766 frnol/h/mg
protein was discemable among treatments with 80% power. Females exposed to E2 had

signiﬁcantly less ovarian aromatase activity than did females exposed to EtOH (Mann-

Whitney U, p = 0.0003; Figure 1; Table 1).

69

Table l: Gonadal aromatase activity (ﬁnol/h/mg protein) in male and female juvenile X.

laevis exposed to positive controls and various concentrations of atrazine ‘

 

 

Treatment N Males Standard N , Females Standard

Error Error
Mean Median Mean Median

Control 19 8.92 0.0 4.19 16 532 272 162
EtOH 21 5.55 0.0 2.67 17 614 403 127
DHT 21 7.22 0.0 3.21 15 202 111 52.3
E2 25 13.1 0.0 4.19 15 * 106 " 37.9 44.7
0.1 ug/L atrazine 24 8.66 0.0 4.67 10 331 232 117
1.0 ug/L atrazine 23 35.8 0.0 28.8 15 490 209 156
10 ug/L atrazine 18 1.92 0.0 1.92 18 507 332 116
25 ug/L atrazine 21 0.94 0.0 0.94 16 496 354 113

 

 

a All frogs with aromatase activity less than the method detection limit (MDL) were assigned a value of 0.0

for statistical analyses and summaries.

* Indicates a signiﬁcant difference ﬁ'om the EtOH control at p < 0.05.

70

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

fé—ﬁﬂw——TT” _-i¢ #7,—- VﬂAﬂ-ﬂ lMales

E A 800.00 1 F le

ﬁg ‘ . U ema SJ

a, § 600.00 -

g a.

U! -

Eg 400.00

2 £ 1.

< '6 200.00 a ﬂ

8 .E

5 V 000 — e e j-
i 0 . ' f I I DHT E2 I

Control 0.1ug/L 1.0 ug/L 1O ug/L 25 ugIL ETOH
atrazine atrazine atrazine atrazine ‘

l

 

Figure l: Gonadal aromatase activity in male and female X. laevis exposed to positive
controls and various concentrations of atrazine. Bars represent mean :t standard error of

the mean. * indicates a signiﬁcant difference from ethanol control, p<0.05.

EtOH = 0.005% ethanol; DHT = 0.1 mg/L dihydrotestosterone; E2 = 0.1 mg/L estradiol

Brain aromatase activity also was statistically different between male and female
X. laevis (Mann-Whitney U, p = 0.024; Figure 2; Table 2). Overall, there was greater
aromatase activity in the brains of female frogs as compared to male frogs. However,
this was not a consistent trend in all treatment groups. Male frogs in the EtOH and E2
treatment group had greater aromatase activity than the female frogs (Table 2). The
mean aromatase activity in the brains of male and female frogs was 0.73 :l: 0.64 and 0.90
i 0.69 pmol/h/mg, respectively.

There were no statistically signiﬁcant differences in brain aromatase activity
among male atrazine-exposed X. laevis and untreated controls (Kruskal-Wallis, p =
0.410). Given the sample size and variability in male brain aromatase activity, a

difference of 0.83 pmol/h/mg protein was discemable among treatments with 80% power.

71

Signiﬁcant differences in brain aromatase activities of male X. laevis were observed
among the DHT, E2, EtOH treatments (Kruskal-Wallis, p = 0.024). The E2-exposed X.
laevis had signiﬁcantly greater brain aromatase activity as compared to the DHT-exposed
X. laevis (Mann-Whitney U, p = 0.012; Figure 2; Table 2).

Among the female X. laevis, there were no signiﬁcant differences in brain
aromatase activity among atrazine-treated X. laevis and the untreated controls (Kruskal
Wallis, p = 0.885). Likewise, in female X. laevis there were no signiﬁcant differences in
brain aromatase activities among the DHT, E2, and EtOH treatments (Kruskal-Wallis, p
= 0.597; Figure 2; Table 2). Given the sample size and variability in female brain
aromatase activity, a difference of 1.1 pmol/h/mg protein was discemable among

treatments with 80% power.

72

Table 2: Brain aromatase activity (pmol/h/mg protein) in male and female juvenile X.

laevis exposed to positive controls and various concentrations of atrazine ’

 

 

Treatment N Males Standard N Females Standard
Error Error
Mean Median Mean Median

Control 18 0.71 0.54 0.12 15 1.05 0.81 0.22

EtOH 21 0.93 0.89 0.17 15 0.81 0.70 0.12

DHT 19 0.54 0.47 0.09 15 0.63 0.46 0.11

E2 24 * 1.07 * 0.91 0.16 15 0.86 0.48 0.20

0.1 ug/L 24 0.58 0.55 0.09 10 0.93 0.85 0.22
atrazine

1.0 ug/L 22 0.71 0.47 0.15 16 0.95 0.86 0.17
atrazine

10 ug/L atrazine 18 0.43 0.27 0.10 18 0.98 0.90 0.16

25 ug/L atrazine 21 0.80 0.81 0.16 16 0.94 0.59 0.23

 

 

a All frogs with aromatase activity less than the MDL were assigned a value of 0.0 for statistical analyses

and summaries.

* Indicates a signiﬁcant difference from the DHT treatment at p < 0.05.

73

I
l
l
l
'
T
l
l
l
l
l

 

I Males
121 Females

 

 

 

 

 

   

 

 

(pmollhlmg protein)

 

 

 

 

 

 

 

 

Brain Aromatase Activity:

 

. Control 0.1ug/L 1.0 ug/L 10 ug/L 25 ug/L ETOH DHT 52 l
J atrazine atrazine atrazine atrazine |
[W _

Figure 2: Brain aromatase activity in male and female X. laevis exposed to positive

 

controls and various concentrations of atrazine. Bars represent mean i standard error of

the mean. * indicates a signiﬁcant difference from estradiol treatment, p<0.05.

EtOH = 0.005% ethanol; DHT = 0.1mg/L dihydrotestosterone; E2 = 0.1mg/L estradiol

Hormones

Concentrations of T and E2 were measurable in the plasma of both male and
female X. laevis by ELISA. However, E2 concentrations in both male and female
juvenile frogs were sometimes less than the method limit of detection (7.8 pg/mL).
Juvenile female X. laevis had greater plasma E2 concentrations than juvenile male frogs
(Kruskal-Wallis, p = 0.020). The mean concentrations of E2 were 27.0 :t 126 and 40.8 i
227 ng/mL for males and females, respectively.

There were statistically signiﬁcant differences in plasma E2 concentrations
among male frogs exposed to control and atrazine treatments (Kruskal-Wallis, p = 0.003;
Table 3). Male frogs exposed to 1.0 ug/L atrazine had signiﬁcantly lower E2
concentrations than untreated control, 0.1 pg/L and 25 ug/L atrazine exposed male frogs

(Mann-Whitney U, p = 0.001 , 0.015, and 0.0001, respectively; Figure 3). However,

74

concentrations of E2 in X. laevis exposed to 0.1, 10 and 25 ug/L atrazine were not
different from untreated controls. Signiﬁcant differences in plasma E2 concentrations
were also observed among males exposed to the DHT, E2, and EtOH treatments
(Kruskal-Wallis, p = 0.025). Plasma concentrations of E2 in male X. laevis exposed to
E2 were signiﬁcantly greater than concentrations of E2 in the EtOH -exposed frogs
(Mann-Whitney U, p = 0.008; Figure 3).

Among the female X. laevis, there were no signiﬁcant differences in E2
concentrations among any of the atrazine treatments and the untreated controls (Kruskal-
Wallis, p = 0.079). Likewise, there were no signiﬁcant differences in E2 concentrations
among the DHT, E2, and EtOH treatments (Kruskal-Wallis, p = 0.086). Given the
sample size and large variability in female estradiol levels, a difference of 495 ng/mL

was discemable among treatments with 80% power.

75

Table 3: Plasma estradiol concentrations (ng/mL) in juvenile X. laevis exposed to

positive controls and various concentrations of atrazine '

 

 

Treatment N Standard N Females Standard
Error Error
Mean Median Mean Median

Control 20 35.2 3.47 15.4 15 97.5 1.88 60.4
EtOH 21 6.03 0.30 5.08 18 0.66 0.23 0.22
DHT 21 18.7 1.47 5.96 16 149 0.35 147

E2 25 ** 68.6 ** 3.92 57.8 15 22.2 2.44 8.61

0.1 ug/L atrazine 26 12.9 0.80 4.64 12 42.5 0.17 26.3
1.0 ug/L atrazine 24 * 0.29 * 0.24 0.03 16 5.14 0.27 2.74
10 ug/L atrazine 20 58.0 0.53 39.6 17 8.98 0.18 8.52
25 ug/L atrazine 23 18.2 3.52 5.32 17 12.4 0.32 6.12

 

 

a All estradiol concentrations less than the method detection limit (0.0078 ng/mL) were set to a value of

0.005 ng/mL for statistical analyses and sununaries. “ Indicates a signiﬁcant difference from the control

treatment at p < 0.05. " Indicates a signiﬁcant difference from the EtOH control at p < 0.05.

76

 

 

 

 

 

 

Estradiol (ug/mL)
00
o

 

 

 

 

 

 

 

1 b ._ 9

I I l

‘ Control 0.1ug/L 1.0 ug/L 10ug/L 25 ug/L ETOH DHT E2
‘ atrazine atrazine atrazine atrazine

 

 

 

Figure 3: Estradiol concentrations in juvenile male X. laevis exposed to positive controls

and various concentrations of atrazine. Bars represent mean 3: standard error of the mean.

a, b signify tests for signiﬁcant differences at p< 0.05 among the atrazine data set; c, d signify tests for
signiﬁcant differences at p<0.05 among the positive controls; EtOH = 0.005% ethanol; DHT = 0.1 mg/L

dihydrotestosterone; E2 = 0.1 mg/L estradiol

There were no signiﬁcant differences in plasma T concentrations between
juvenile male and female X. laevis (Kruskal-Wallis, p = 0.170). There were no
statistically signiﬁcant differences in T concentrations among male frogs in the control
and atrazine treatments (Kruskal-Wallis, p = 0.270). There also were no statistically
signiﬁcant differences in concentrations of T among the E2 and DHT treatments and the
EtOH controls (Kruskal-Wallis, p = 0.187; Table 4). Given the sample size and
variability in male testosterone levels, a difference of 37 ng/mL was discemable among
treatments with 80% power.

There were no statistically signiﬁcant differences in concentrations of T among

atrazine treatments and the untreated controls for female X. laevis (Kruskal-Wallis, p =

77

0.179). Likewise, there were no statistically signiﬁcant differences in plasma
concentrations of T among females in the DHT, E2, and EtOH treatments (Kruskal-
Wallis, p = 0.363; Table 4). Given the sample size and variability in female testosterone
levels, a difference of 83 ng/mL was discemable among treatments with 80% power.
Concentrations of E2 and T were positively correlated (R2 = 0.81). Individual
frogs with greater concentrations of T also tended to have greater concentrations of E2.
However, neither concentrations of T nor E2 were correlated with either gonad or brain

aromatase activity (R2 < 0.005 in all cases).

Table 4: Plasma testosterone concentrations (ng/mL) in juvenile X. laevis exposed to

positive controls and various concentrations of atrazine “

 

 

Treatment N Male Standard N Female Standard

Error Error
Mean Median Mean Median

Control 20 17.2 1.39 8.17 15 24.6 1.39 12.9
EtOH 21 5.60 0.74 3.57 18 1.11 0.61 0.30
DHT 21 3.30 1.46 0.99 16 25.7 1.04 24.3
E2 25 11.7 1.56 7.74 15 6.61 0.70 4.16
0.1 ug/L atrazine 26 3.57 0.77 1.15 12 24.2 0.94 17.3
1.0 ug/L atrazine 24 1.10 0.67 0.18 16 1.00 0.75 0.23
10 ug/L atrazine 20 16.3 1.48 11.0 17 2.79 0.82 1.82
25 ug/L atrazine 23 2.49 1.24 0.57 17 1.69 0.71 0.58

 

 

a All testosterone concentrations less than the method detection limit (0.0078 ng/mL) were set to a value of

0.005 ng/mL for statistical analyses and summaries.

78

Discussion
During the 185 (1 exposure, small concentrations of atrazine were detected in

untreated control tanks. This may have resulted from volatilization and deposition of
atrazine ﬁom exposure tanks into control tanks, or atrazine may have been periodically
present at low levels in the laboratory water supply. However, this does not affect the
conclusions of the study, since there were no statistically signiﬁcant, concentration-
dependent effects of atrazine observed on any of the parameters investigated.

Aromatase is a critical enzyme for ovarian differentiation in ﬁsh and amphibians
(Miyata et al., 1999; Baroiller and D’Cotta, 2001). Therefore, as expected, female X.
laevis had greater aromatase activity in their gonads compared to male frogs. As in
teleost ﬁshes, there was lower aromatase activity (per unit protein) in the gonadal tissue
of X. laevis as compared to the brain tissue (Melo and Ramsdell, 2001; Pasmanik and
Callard, 1988).

Overall, female X. laevis had greater brain aromatase activity than male frogs, but
this result was not consistent in all treatment groups. In ﬁsh, brain aromatase activity
differs spatially in the male and female brain (Melo and Ramsdell, 2001). In addition,
brain aromatase activity was reported to vary seasonally in both frogs and ﬁsh (Pasmanik
and Callard, 1988; Guerriero et al., 2000) and was quantitatively different in male and
female goldﬁsh (Carassius auratus) only during a single month of the year (Guerriero et
al., 2000). In the present study, it is possible that the sex difference in brain aromatase
activity among X. laevis is also a transient observation.

Aromatase activity was measurably different in frog gonads and brains exposed to
E2 and DHT. Exposure to E2 resulted in lesser gonadal aromatase activity in female X.

laevis relative to controls. This phenomenon was not apparent in the male ﬁ'ogs. Most

79

male individuals had gonadal aromatase activity that was less than the method detection
limit, resulting in poor resolution for detecting signiﬁcant decreases in aromatase activity
among treatment groups. In the case of female X. laevis, exogenous application of E2
may have been functioning in a negative feedback system with the aromatase enzyme,
thus lowering the enzyme’s activity (Yue et al., 2001). In male frogs, brain aromatase
activity was elevated in E2-exposed frogs as compared to DHT-exposed frogs. This
result is similar to other studies that report a positive feedback mechanism between E2
administration and male brain aromatase activity (Melo and Ramsdell, 2001).

There were no atrazine related effects on either gonad or brain aromatase activity
in either male or female X. laevis. These results are consistent with another study in
which aromatase activity did not differ between X. laevis collected from reference and
atrazine use areas (Hecker et al., 2003). Aromatase activity was not correlated with
circulating concentrations of either T or E2 in X. laevis. This result does not support the
hypothesis that atrazine feminizes male frogs by increasing estrogen or demasculinizes
frogs by decreasing testosterone levels via upregulation of aromatase, the enzyme that
transforms T into E2 (Hayes et al., 2002; Hayes et al., 2003).

Atrazine exposure did not affect circulating levels of T in either male or female
X. laevis. In addition, there were no differences in circulating concentrations of E2 in
female frogs exposed to atrazine. In males, however, a signiﬁcant difference in plasma
concentrations of E2 was detected among frogs in the 1.0 ug/L atrazine treatment as
compared to controls. Yet, this result was not consistent across atrazine concentrations,
leading to the conclusion that atrazine was not responsible for depressed levels of E2 in

this case. Since circulating concentrations of both E2 and T levels were variable in

80

juvenile X. laevis, and because sex steroids are not at their maximum level in juvenile
frogs (Tobias et al., 1998; Kang et al., 1995), alterations in T and E2 concentrations may
be more difﬁcult to detect in juvenile ﬁogs. In adult X. laevis, the plasma concentrations
of both E2 and T were depressed in female frogs collected from atrazine use areas as
compared to reference areas (Hecker et al., 2003). However, this decrease in hormone
levels could not be causally linked with atrazine exposure, since multiple pesticides and
other unknown environmental factors also could have resulted in lowered hormone levels
in these ﬁ'ogs (Hecker et al., 2003). Thus, it cannot be concluded that atrazine alone is
capable of causing depressed sex steroid concentrations in X. laevis. For a clearer picture
of how exogenous compounds may disrupt sex steroid levels in amphibians, increased
research into the normal background titers of circulating E2 and T and an increased

understanding of normal sex steroid cycling is necessary in juvenile and adult frogs.

Conclusion

At measured concentrations ranging from 0.01 to 28.9 113/L, atrazine did not
affect aromatase activity or circulating concentrations of T or E2 in developing X. laevis.
This conclusion does not support the hypothesis that environmentally relevant
concentrations of atrazine disrupt hormone levels in frogs via upregulation of the enzyme
aromatase (Hayes et al., 2002; Hayes et al., 2003). While atrazine may cause an effect on
aromatase activity in vitro in mammalian cell lines (Sanderson etal., 2000), the same
may not be true in vivo in frogs. Therefore, it appears that atrazine, at environmentally
relevant concentrations, does not affect reproductive development in X. laevis frogs via

upregulation of the aromatase enzyme.

81

REFERENCES

10.

11.

. Ankley GT, Tietge J E, Defoe DL, Jensen KM, Holcombe GW, Durhan EJ, and

Diamond SA. 1998. Effects of ultraviolet light and methoprene on survival and
development of Rana pipiens. Environ. T oxicoI. Chem. 17(12): 2530-2542.

Allran J W and Karasov WH. 2000. Effects of atrazine and nitrate on northern
leopard (Rana pipiens) larvae exposed in the laboratory from posthatch through
metamorphosis. Environ. T oxicol. Chem. 19(11): 2850-2855.

American Society for Testing and Materials (ASTM). 1991. Standard guide for
conducting the frog embryo teratogenesis assay-Xenopus (F etax). ASTM E 1439-
91. Philadelphia, PA.

Baroiller J F and D’Cotta H. 2001. Environment and sex determination in farmed

ﬁsh. Comparative Biochemistry and Physiology C-Toxicology & Pharmacology
130: 399- 409.

Battaglin W and F airchild J. 2002. Potential toxicity of pesticides measured in

Midwestern streams to aquatic organisms. Water Science and Technology 45: 95-
103.

Berrill M, Coulson D, McGillivray L, and Pauli B. 1998. Toxicity of endosulfan
to aquatic stages of anuran amphibians. Environ. T oxicol. Chem. 17(9): 1738-
1744.

Blaustein AR and Kiesecker JM. 2002. Complexity in conservation: lessons from
the global decline of amphibian populations. Ecology Letters 5(4): 597-608.

Blaustein AR, Romansic JM, Kiesecker J M, and Hatch AC. 2003. Ultraviolet
radiation, toxic chemicals and amphibian population declines. Diversity and
Distributions 9: 123-140.

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

Carr J A, 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 growth, metamorphosis, and
gonadal and laryngeal morphology. Environ. T oxicol. Chem. 22: 396-405.

Christin MS, Gendron AD, Brousseau P, Menard L, Marcogliese DJ, Cyr D, Ruby
S, and Foumier M. 2003. Effects of agricultural pesticides on the immune system

82

12.

13.

14.

15.

16.

17.

l8.

19.

20.

21.

22.

23.

of Rana pipens and on its resistance to parasitic infection. Environ. T oxicol.
Chem. 22(5): 1127-1133.

Conant R and Collins J T. 1998. Peterson Field Guides: Reptiles and Amphibians
Eastern/Central American. Houghton Mifﬂin Co. New York, New York.

Cooper RL, Stoker TE, Tyrey L, Goldman J M, and McElroy WK. 2000. Atrazine
disrupts the hypothalamic control of pituitary-ovarian ﬁmction. Toxicological
Sciences 53: 297-307.

Cuisset B, Pradelles P, Kime DE, Kuhn ER, Barbin P, Davail S, and LeMenn F.
1994. F. Comp. Biochem. Physiol. 108:229—241.

Dazak P, Cunningham AA, and Hyatt AD. 2003. Infectious disease and
amphibian population declines. Diversity and Distributions 9: 141-150.

Diana SG, Resetarits WJ, Schaeffer DJ, Beckmen KB, and Beasley VR. 2000.
Effects of atrazine on amphibian growth and survival in artiﬁcial aquatic
communities. Environ. Toxicol. Chem. 19: 2961-2967.

Duellman WE and Trueb L. 1994. Biology of Amphibians. The John Hopkins
University Press, Baltimore, Maryland.

Dumont JN. 1972. Oogenesis in Xenopus laevis (Daudin). I. Stages of oocyte
development in laboratory maintained animals. J Morphol. 136: 153-7 9.

Fioramonti E, Semlitsch RD, Reyer H, and F ent K. 1997. Effects of triphenyltin
and pH on the growth and development of Rana lessonae and Rana esculenta
tadpoles. Environ. T oxicol. Chem. 16(9): 1940-1947.

Fischer LM, Catz D, and Kelley DB. 1995. Androgen-directed development of

the X. laevis larynx: control of androgen receptor expression and tissue
differentiation. Developmental Biology 170: 1 1 5-126.

Foote CL and Witschi E. 1939. Effect of sex hormones on the gonads of frog
larvae (Rana clamitans): Sex inversion in females; stability in males. The
Anatomical Record 75: 75-83.

Gilbertson M, Haffner GD, Drouillard KG, Albert A, and Dixon B. 2003.
Immunosuppression in the northern leopard frog (Rana pipiens) induced by
pesticide exposure. Environ. T oxicol. Chem. 22 (1): 101-110.

Gillespie GR. 2002. Impacts of sediment loads, tadpole density, and food type on

the growth and development of tadpoles of the spotted tree ﬁog Litoria spenceri:
an in-stream experiment. Biological Conservation 106: 141 -150.

83

 

 

’75.}. w mm. rlﬂmlllrl

 

 

24. Gillilland CD, Summer CL, Gillilland MG, Kannan K. Villeneuve DL, Coady
KK, Muzzall P, Mehne C, Giesy JP. 2001. Organochlorine insecticides,
polychlorinated biphenyls, and metals in water, sediment, and green frogs from
southwestern Michigan. Chemosphere 44: 327-339.

25. Gimeno S. 1998. Demasculinization of sexually mature male common carp,
Cyprinus carpio, exposed to 4-tert-pentylphenol during spermatogenesis. Aquatic
T oxicol. 43: 93-109.

26. Gosner KL. 1960. A simpliﬁed table for staging anuran embryos and larvae with
notes on identiﬁcation. Herpetologica 16: 183-190.

27. Gramapurohit NP, Shanbhag BA, and Saidapur SK. 2000. Pattern of gonadal sex
differentiation, development, and onset of steroidogenesis in the hog, Rana
curtipes. General and Comparative Endocrinology 119: 256-264.

28. Guerriero G, Roselli CE, Paolucci M, Botte V, and Ciarcia G. 2000. Estrogen
receptors and aromatase activity in the hypothalamus of the female ﬂog, Rana

esculenta. Fluctuations throughout the reproductive cycle. Brain Research 880:
92-101.

29. Hayes TB, Wu TH, and Gill TN. 1997. DDT-like effects as a result of
corticosterone treatment in an anuran amphibian: is DDT a corticoid mimic or a
stressor? Environ. T oxicol. Chem. 16(9): 1948-1953.

30. Hayes TB, Collins A, Lee M, Mendoza M, Noriega N, Stuart AA, and Vonk A.
2002. Hermaphroditic, demasculinized frogs after exposure to the herbicide
atrazine at low ecologically relevant doses. Proceeding of the National Academy
of Sciences of the United State of America 99: 5476-5480.

31. Hayes TB, Haston K, Tsui M, Hoang A, Haeffele C, and Vonk A. 2003. Atrazine-
induced hermaphroditism at 0.1 ppb in American leopard frogs (Rana pipens):
laboratory and ﬁeld evidence. Environmental Health Perspectives 111: 568-575.

32. Hecker M. 2002. Natural variability of endocrine functions and their modulation
by anthropogenic inﬂuences: investigations of the bream (Abramis brama[L.])
along the Elbe River, and in a reference site. Ph.D. thesis at the University of
Hamburg, Germany.

33. Hecker M, Giesy JP, Jones PD, Alarik MJ, du Preez L. 2003. Plasma sex steroid
concentrations and gonadal aromatase activities in African clawed frogs (Xenopus
laevis) from the com-growing region of South Africa. Submitted.

34. Hsu C and Liang H. 1970. Sex races of Rana catesbeiana in Taiwan.
Herpetologica 26: 214-221.

84

35. Kang L, Marin M, and Kelley D. 1995. Androgen biosynthesis and secretion in
developing Xenopus laevis. Gen Comp Endocrinol 100: 293-307.

36. Kats LB and Ferrer RP. 2003. Alien predators and amphibian declines: review of

two decades of science and the transition to conservation. Diversity and
Distributions 9: 99-1 10.

37. Kiesecker JM. 2002. Synergism between trematode infection and pesticide
exposure: A link to amphibian limb deformities in nature? PNAS available at
www.pnas.org/cgi/doi/10.1073/pnas.152098899.

38. Laurance WF, McDonald KR, and Speare R. 1996. Epidemic disease and the
catastrophic decline of Australian rain forest frogs. Conservation Biology 10: 406-
413.

39. Lee ET. 1992. Statistical Methods for Survival Data Analysis. 2nd ed. New York:
John Wiley & Sons.

40. Lefcort H and Blaustein AR. 1995. Disease, predator avoidance, and vulnerability
to predation in tadpoles. Oikos 74: 469-474.

41. Lephart ED and Simpson ER. 1991. Assay of aromatase activity. Methods in
Enzymology 206: 477-483.

42. Marco A, Quilchano C, and Blaustein AR. 1999. Sensitivity to nitrate and nitrite

in pond-breeding amphibians from the Paciﬁc Northwest, USA. Environ. Toxicol.
Chem. 18(12): 2836-2839.

43. Martof B. 1956. Growth and development of the green frog, Rana clamitans,
under natural conditions. The American Midland Naturalist 55: 101 -1 17.

44. McKim JM. 1997 . Evaluation of tests with early life stages of ﬁsh for predicting
long-term toxicity. J .F ish.Res.Board Can. 34: 1148-1154.

45. Melo AC and Ramsdell J S. 2001. Sexual dimorphism of brain aromatase activity
in medaka: induction of a female phenotype by estradiol. Environ. Health
Perspec. 109 (3): 257-264.

46. Miyata S, Koike S, and Kubo T. 1999. Hormonal reversal and the genetic control
of sex differentiation in Xenopus. Zoological Science 16:335-340.

47. Miyata S and Kubo T. 2000. In vitro effects of estradiol and aromatase inhibitor

treatment on sex differentiation in Xenopus laevis. Gen Comp Endocrinol 119:
105-1 1 0.

85

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

58.

Morgan MK, Scheuerman PR, Bishop CS, and Pyles RA. 1996. Teratogenic
potential of atrazine and 2,4-D using FETAX. Journal of Toxicology and
Environmental Health 48: 151-168.

Nash JP, Davail-Cuisset B, Bhattacharyya S, Suter HC, LeMenn F, and Kime DE.
2000. An enzyme linked immunosorbent assay (ELISA) for testosterone,
estradiol, and 17,20 beta-dihydroxy-4-pregenen-3-one using acetylcholinesterase
as tracer: application to measurement of diel patterns in rainbow trout
(Oncorhynchus mykiss). Fish Physiol. Biochem. 22:355-363.

Nieuwkoop PD and Faber J. 1967. Normal Table of Xenopus laevis (Daudin). A
systematical and chronological survey of the development from fertilized egg till
the end of metamorphosis. 2"d edition. Amsterdam, North-Holland Pub. Co.

Ouellet M, Bonin J, Rodrigue J, DesGranges J, and Lair S. 1997. Hindlimb
deformities (ectromelia, ectrodactyly) in free-living anurans from agricultural
habitats. Journal of Wildlife Diseases 33: 95-104.

Pasmanik M and Callard GV. 1988. Changes in brain aromatase and Sa-reductase
activities correlate signiﬁcantly with seasonal reproductive cycles in goldﬁsh
(Carassius auratus). Endocrinology 122: 1349-1356.

Pickford DB and Morris ID. 1999. Effects of endocrine-disrupting contaminants
on amphibian oogenesis: methoxychlor inhibits progesterone-induced maturation
of Xenopus laevis oocytes in vitro. Environ. Health Perspec. 107(4): 285-291.

Pounds J A and Crump ML. 1994. Amphibian declines and climate disturbance:
the case of the golden toad and the harlequin frog. Conservation Biology 8: 72-85.

Reeder AL, Foley GL, Hansen LG, Wikoff B, Faeh S, Eisold J, Wheeler MB,
Warner R, Murphey JE, and Beasley VR. 1998. Forms and prevalence of
intersexuality and effects of environmental contaminants on sexuality in cricket
frogs (Acris crepitans). Environmental Health Perspectives 106: 261-265.

Rowe CL, Sadinski WJ, and Dunson WA. 1992. Effects of acute and chronic
acidiﬁcation on three larval amphibians that breed in temporary ponds. Arch.
Environ. Contam. T oxicol. 23: 339-350.

Sanderson J T, Seinen W, Giesy JP, and van den Berg M. 2000. 2-chloro-s-triazine
herbicides induce aromatase (CYP19) activity in H295R human adrenocortical
carcinoma cells: a novel mechanism for estrogenicity? T oxicol. Sci. 54:121-127.

Semlitsch RD, Foglia M, and Mueller A. 1995. Short-term exposure to

triphenyltin affects the swimming and feeding behavior of tadpoles. Environ.
T oxicol. Chem. 14(8): 1419-1423.

86

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

Solomon KR, Baker DB, Richards RP, Dixon DR, Klaine SJ, LaPoint TW,
Kendall RJ, Weisskopf RJ, Giddings JM, Giesy JP, Hall LW, and Williams WM.
1996. Ecological risk assessment of atrazine in North American surface waters.
Environ. T oxicol. Chem. 15: 31-74.

Sower SA, Reed KL, and Babbitt KJ. 2000. Limb malformations and abnormal
sex hormone concentrations in frogs. Environ Health Perspec 108: 1085-1090.

Sparling DW, Fellers GM, and McConnell LL. 2001. Pesticide and amphibian
population declines in California, USA. Environ. T oxicol. Chem. 20: 1591-1595.

Sullivan KB and Spence KM. 2003. Effects of sublethal concentrations of
atrazine and nitrate on metamorphosis of the African clawed frog. Environ.
T oxicol. Chem. 22(3): 627-635.

Sumpter JP, Jobling S, and Tyler CR. 1996. Oestrogenic Substances in the
Aquatic Environment and Their Potential Impact on Animals, Particulary Fish. In:
E. W. Taylor (Ed.). New York: Cambridge University Press, 283p.

Tavera-Mendoza L, Ruby S, Brousseau P, Foumier M, Cyr D, and Marcogliese
D. 2002a. Response of the amphibian tadpole (Xenopus laevis) to atrazine during
sexual differentiation of the testis. Environ. T oxicol. Chem. 21: 527-531.

Tavera-Mendoza L, Ruby S, Brousseau P, Foumier M, Cyr D, and Marcogliese
D. 2002b. Response of the amphibian tadpole (Xenopus laevis) to atrazine during
sexual differentiation of the ovary. Environ. T oxicol. Chem. 21: 1264-1267.

Tej edo M and Reques R. 1992. Effects of egg size and density on metamorphic
traits in tadpoles of the natterj ack toad (Bufo calamita). Journal of Herpetology
26: 146-152.

Tobias ML, Marin ML, and Kelley DB. 1991. Temporal constraints on androgen

directed laryngeal masculinization in Xenopus laevis. Developmental Biology
147: 260-270.

Tobias ML, Marin ML, and Kelley DB. 1993. The roles of sex, innervation, and
androgen in laryngeal muscle of Xenopus laevis. The Journal of Neuroscience

13(1): 324-333.

Tobias ML, Tomasson J, and Kelley DB. 1998. Attaining and maintaining strong
vocal synapses in female Xenopus laevis. J Neurobiol 37(3): 441-448.

Van Tienhoven, A., 1983. Reproductive Physiology of Vertebrates. 2nd edition,
Cornell University Press, Ithaca, NY.491 pp.

87

71.

72.

73.

74.

75.

Wallace H, Badawy GMI, and Wallace BMN. 1999. Amphibian sex
determination and sex reversal. Cellular and Molecular Life Sciences 55: 901-
909.

Witschi E. 1929. Studies on sex differentiation and sex determination in
amphibians. The Journal of Experimental Zoology 54: 157-223.

Witchi E. 1967. Biochemistry of sex differentiation in vertebrate embryos. In The

Biochemistry of Animal Development: Biochemical Control and Adaptations in
Development, Vol 2. Academic Press, New York, NY, USA, pp 193-225.

Yokley RA and Cheung MW. 2000. Analytical method for the determination of
atrazine and its dealkylated chlorotriazine metabolites in water using gas
chromatography/mass selective detection J Agric. Food Chem. 48: 4500-4507.

Yue W, Berstein LA, Wang JP, Clark GM, Hamilton CJ, Demers LM, and Santen

RJ. 2001. The potential role of estrogen in aromatase regulation in the breast.
Journal of Steroid Biochemistry and Molecular Biology. 79: 157-164.

88

’lalllllllllllllllllllll'llllllllllllll
3 1293 02470 0415