Effects of p,p'-DDE on Reproduction and Biomarkers of Endocrine Disruption in Fathead Minnows (Pimephales promelas)

ray_e ( .pdf )

EFFECTS OF p,p'-DDE ON REPRODUCTION AND BIOMARKERS OF
ENDOCRINE DISRUPTION IN FATHEAD MINNOWS (Pimephalespromelas)















By

ELIZABETH JORDAN RAY


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA


2006

































Copyright 2006

by

Elizabeth Jordan Ray
































This work is dedicated to the loving and cheerful spirit of my grandpa, Owen.















ACKNOWLEDGMENTS

The completion of this project would not have been possible without input of

several people. First, I thank my advisor, Dr. David Barber, for his direction and

availability, without which I would not have had the opportunity to learn and do the

variety of skills required for the successful completion of this project. I also thank my

committee members, Dr. Nancy Denslow and Dr. Madan Oli, for all their input and

advice. My family and friends were an essential source of support throughout the course

of these years. There are also several lab members and co-workers at the Center for

Environmental and Human Toxicology and the Aquatic Toxicology Facility that I thank:

Kathy Childress and Kevin Kroll for their advice on fish care and handling, Greg

Robbins for assisting with fish care, Kathleen Jensen at the US EPA in Duluth for helping

with RIA validation, Scott Wasdo and Nancy Szabo for their chemical advice, Joe Griffitt

for help with gene analysis, and fellow lab members Alex McNally and Roxana Weil for

all the little things.
















TABLE OF CONTENTS

page

A C K N O W L E D G M E N T S ................................................................................................. iv

L IST O F TA B LE S ............................ ................... ...... .............. .. vii

LIST OF FIGURES ............. .. ..... ...... ........ ....... .......................... viii

INTRODUCTION .......................... ........ .. ... .... ........ ...............

Endocrine D isruption in Teleosts ........................................ ........................... 2
Biom arkers of Endocrine Disruption................................ ......................... ........ 7
p,p'-D D E in the Environm ent......... ................................................. ............... 12
p,p'-DDE: An Endocrine Disruptor............................................................... 14
Linking Biomarkers of Endocrine Disruption to Fish Populations ............................17

M A TERIA L S A N D M ETH O D S............................................................ .....................21

G en e ral M eth o d s ................................................................................................... 2 1
Fish H holding C onditions........................................................... ............... 21
R productive M measures ............................................... ............................ 22
Plasma and Tissue Collection............... ........ .................................... 22
Determination ofp,p'-DDE Content....................................... ............... 23
Determ nation of Plasm a 17p-Estradiol ................................... .................24
Vitellogenin mRNA Quantification ....................................... ............... 25
E x p erim mental Set-U p ......................................................................... ................... 2 6
Pilot E xperim ent ...................... ...... ....... ........ .. ...... .. .... .... ............ .. .... 26
p,p'-DDE Dose-Response Experiment I: Accumulation Rate and
Reproductive Output of Adults.......................................................27
Survival, Development, and Reproductive Output of Offspring.........................28
p,p'-DDE Dose-Response Experiment II: Collection of Biological Materials ...29
Statistical A analyses ........................................................ ...... .. .... ...... .. 29

R E S U L T S ................................................................................ 3 1

Pilot E xperim ent ................ .. .. ................................ ............ .. .......... ............ 31
p,p'-DDE Dose-Response Experiments I & II: Adult 17p-Estradiol, GSI, and
R production .................................................. ...........32
Effects of in ovo Exposure on Survival, Development, and Reproduction ..............34
Identification of B iom arkers............................................................ .....................36









D ISC U SSIO N ................................................................................................ ....... 47

Effects of Adult Exposure to p,p'-DDE on Reproduction, 17p-Estradiol, and GSI...47
Effects of in ovop,p'-DDE Exposure on Survival, Development, and
Reproduction ............. ..... ......... .... ...............52
Biom arkers .................................. ........................ ...... ..... ........ 55
Conclusions and Future D irections....................................... .......................... 56

A P P E N D IX .................................................................................................................. 5 8

HORMONE DETERMINATION BY RIA .............................................................58

R E F E R E N C E S ........................................ ............................................ ................. .. 6 0

B IO G R A PH IC A L SK E TCH ..................................................................... ..................68
















LIST OF TABLES

Tables p

1-1. Classes of endocrine disrupting compounds and examples.............................. 20

2-1. Primer sequences used for quantitative real-time PCR. .........................................30

3-1. Survival probabilities for offspring spawned from adults fed p,p'-DDE.................37
















LIST OF FIGURES


Figures p

1-1. Schematic of basic signals within the hypothalamus-pituitary-gonad (HPG) axis. ....3

1-2. Active horm ones found in fish. .............................................................................5

1-3. DDT and selected metabolites ........................ ................... ............... 13

3-1. The number of eggs produced per female before and after exposure to either
p,p'-DDE of flutamide as compared to control................................... ............... 37

3-2. Comparison of plasma 17p-estradiol levels in males and females within and
am ong each treatm ent group.. ............................. ............................................... 38

3-3. Meanp,p'-DDE muscle tissue concentration (Ltgp,p'-DDE / g wet weight
m uscle tissue). .........................................................................38

3-4. Comparison of plasma 17p-estradiol levels in males and females from each
treatm ent group .......................................................................39

3-5. Mean GSI values of adults treated with p,p'-DDE-contaminated feed..................39

3-6. Cumulative number of eggs produced per female in adults of each treatment
g ro u p ...................... .. .. ......... .. .. ................................................... 4 0

3-7. Egg fertilization and hatch success...................................... ......................... 40

3-8. M maternal transfer ofp,p'-DDE to eggs ....................................... ............... 41

3-9. Mean length (mm) and weight (g) + S.E. of offspring at four months after hatch...41

3-10. The age when male characteristics or first reproduction was observed in
offspring from each treatment group.............................. ...............42

3-11. Sex ratio of offspring in each treatment group................................................. 42

3-12. GSI values of offspring in each treatment group........... ......... ............... 43

3-13. Cumulative egg production of offspring through nine months of age. ................43









3-14. Values of mean female plasma estradiol, mean female GSI, and the percent of
eggs fertilized are given for each treatment group............................................44

3-15. Values of mean plasma estradiol and the percent of eggs per female are given
for each treatm ent group ................................................ .............................. 45

3-16. Values of mean adult female plasma estradiol and mean adult GSI for fish
exposed as adults and for fish exposed in ovo are given for each treatment
group ................. ..................................... ...........................46















Abstract of Thesis Presented to the Graduate School of the University of Florida
in Partial Fulfillment of the Requirements for the
Degree of Master of Science

EFFECTS OF p,p'-DDE ON REPRODUCTION AND BIOMARKERS OF
ENDOCRINE DISRUPTION IN FATHEAD MINNOWS (Pimephalespromelas)

By

Elizabeth Jordan Ray

August 2006

Chair: David Barber
Major Department: Interdisciplinary Ecology

The phenomenon of endocrine disruption, which includes impaired reproduction

and survival, has been widely researched over the past decade. However, few studies

have linked biomarkers of endocrine disruption to population-level outcomes. The most

stable metabolite of the organochlorine pesticide DDT is p,p'-DDE, a common

environmental contaminant. The objective of this experiment was to investigate the

effects ofp,p'-DDE on biomarkers of endocrine disruption as they related to survival and

reproduction of fathead minnows (Pimephalespromelas). Fish were exposed to 1.63,

11.48, 104.25, or 900 [tgp,p'-DDE / g feed or 1208 tg flutamide / g feed in three

separate experiments. Reproductive output, 17p-estradiol, and gonadosomatic index

(GSI) were measured in adults exposed to p,p'-DDE via feed. Survival, development,

GSI and reproductive output were measured to nine months of age in eggs from those

adults to determine second-generation effects ofp,p'-DDE exposure. Reproductive

output was impaired in fish exposed to 104.25 [tgp,p'-DDE / g feed as well as their









offspring, but not in fish exposed to flutamide. 17p-estradiol as measured by enzyme-

linked immunosorbent assay (ELISA) was elevated in males exposed to flutamide and

1.63, 11.48, and 104.25 tpgp,p'-DDE / g feed. p,p'-DDE and flutamide do not act by the

same mechanism at high doses (900 tpgp,p'-DDE and 1208 ptg flutamide / g feed).

Neither GSI nor 17p-estradiol levels were correlated top,p'-DDE concentration in feed

or fish muscle tissue. The percent of eggs fertilized, GSI and 17p-estradiol measured

from adults exposed to p,p'-DDE through feed were directly related to each other and

inversely related to GSI of offspring. Male GSI of offspring of adults exposed to 104.25

pg / g feed was significantly higher than any other group. GSI in adult males was

inversely related to the number of eggs spawned per female. These data have important

implications for the effects of in ovo exposure to endocrine disrupting compounds.

Further, data such as these can be used to model the population growth rates and relate

them to biomarkers, such as GSI, of exposure to endocrine disrupting compound.















INTRODUCTION

Scientific and popular literature have reported on the effects of endocrine

disrupting compounds (EDCs) in fish for decades. Endocrine disrupting compounds are

those that affect the normal functioning of that system, typically resulting in adverse

effects on an organism, its progeny, and/or a population. Endocrine disruption has

caused widespread public concern regarding the sustainability of fish populations and

sparked copious scientific research in recent years. Laboratory and field studies have

shown reproductive dysfunction in animals and humans exposed to a variety of EDCs

(Cook et al. 2003; Gray et al. 2001; LeBlanc et al. 1997; Noaksson et al. 2003). Specific

groups of chemicals have been investigated for endocrine-disrupting effects, including

pesticides, therapeutic hormones administered to humans and mammals, and byproducts

of industrial processes. More specific types of these groups of compounds include

organochlorine and organophosphate pesticides, components of oral contraceptives,

plasticizers, fire retardants, and jet-fuel residues (Table 1-1).

The reproductive cycle of fish is controlled by the endocrine system. The

endocrine system is regulated by the hypothalamus, pituitary, and gonad, collectively

known as the HPG axis. By definition, EDCs alter the normal functioning of the HPG

axis. Disruptions of the HPG axis are most conspicuously manifested in gonads, which

unlike the hypothalamus and pituitary, undergo visible stages of reproductive

development in fathead minnows. Because the gonad is responsible for reproductive

output, it is the primary link between an altered endocrine system and demonstrated









reproductive dysfunction. Thus, gonadal function is an easily measured and relevant

marker of exposure to an EDC. Production of the female-specific egg yolk precursor

protein, vitellogenin, is also an important measure of endocrine disruption in males.

Measures of gonad health and vitellogenin are commonly used biomarkers of endocrine

disruption. Relationships between biomarkers of endocrine disruption measured in an

individual fish and population-level outcomes are currently poorly understood.

Endocrine Disruption in Teleosts

The teleostean endocrine system involves several organs and myriad molecular

signaling pathways. To understand endocrine disruption, one must first understand the

normal endocrine system. Environmental cues, such as photoperiod, temperature and

presence of other fish, trigger a cascade of signals that prepare teleosts for reproduction.

In teleosts, steroid hormones are typically considered the ultimate molecular factors

influencing reproductive development, maturity, and release of gametes.

The steroidogenic process is controlled by the HPG axis, beginning in the

hypothalamus when environmental cues stimulate the release of gonadotropin-releasing

hormone (GnRH; Fig. 1-1). Gonadotropin-releasing hormone begins a cascade of signals

to stimulate reproductive preparedness. The release of GnRH stimulates the pituitary to

secrete two types of gonadotropins, GtH I and GtH II that act on steroidogenic tissues

(Arcand-Hoy and Benson 2001; Hu et al. 2001). GtH I and GtH II are the teleostean

analogs of the mammalian follicle-stimulating hormone (FSH) and luteinizing hormone

(LH), respectively (Arcand-Hoy and Benson 2001). The steroidogenic process is

triggered when gonadotropins (GtH I and GtH II) reach the gonads and attach to

hormone-responsive cellular receptors in the cell membrane (Young et al. 2005). While

most research has focused on GtH I and GtH II., other hormones and factors may also










play a role in steroidogenesis in the gonad, but their potency and mechanism of action are

not well-described and hence, they are not addressed here (Van der Kraak et al. 1998).


Environmental Signals Hypothalamus
Day length Gonadotropin-releasing hormone
Temperature (GnRH)

inhibit

Pituitary
GTH I & GTH II



Gonad
Liver Gonad
Vitellogenin 17-Estradiol .
Testosterone
S11-Ketotestosterone

Egg





Figure 1-1. Schematic of basic signals within the hypothalamus-pituitary-gonad (HPG)
axis.

Each gonadotropin plays a specific role in the process of reproduction, which

includes oocyte development and ovulation. GtH I is primarily responsible for oocyte

development in females and spermiation in males. In females, GtH I binds to specific

receptors on the follicle, stimulating testosterone production and its subsequent

aromatization to estradiol. Estradiol then binds to the estrogen receptor in endocrine

active tissues, signaling a cascade of events that contribute to oocyte development,

including production of vitellogenin in the liver (Ding 2005). As oocytes and sperm

develop, the expression of GtH II increases relative to GtH I (Van der Kraak et al. 1998).

GtH II is primarily responsible for stimulating the production of progesterone, also

known as maturation-inducing hormone (MIH), which is believed to be responsible for









the final oocyte maturation and ovulation. Several hormones shown to have MIH effects

include progestens, cortisol, and deoxycorticosterone (Nagahama et al. 1994). The

outcomes produced by each gonadotropin, such as oocyte development and ovulation, are

a result of the amount of type of steroid produced.

Production of specific steroids is ultimately triggered by an increase in cAMP,

which occurs when gonadotropins attach to hormone-responsive receptors in cell

membranes of steroidogenic tissues (Hu et al. 2001). cAMP then binds to response

elements in the promoter regions of genes coding for factors involved in steroidogenesis,

thereby increasing the expression of those factors. Cholesterol is the initial substrate for

all steroid hormones, including estrogens, progesterones, and androgens. The first and

rate-limiting step of this process begins with production of steroidogenic acute regulatory

protein (StAR), which facilitates the transfer of cholesterol from the cytoplasm into

mitochondria (Young et al. 2005). In mitochondria P450 side-chain cleavage enzyme

(P450scc) catalyses the conversion of cholesterol into pregnenolone. Pregnenalone can

then be converted to a number of steroid hormones through several different pathways.

Progesterones are formed from pregnenolone in reactions catalyzed by P450cl7, 33-

hydroxysteroid dehydrogenase (30-HSD), and 20p-hydorysteroid dehydrogenase (203-

HSD). Androstenedione and androstenediol, which are formed from pregnenolone and

progesterones, are the substrates converted to testosterone. Testosterone can then be

converted to 17p-estradiol or 11-ketotestosterone by P450 aromatase or P45011P and

110-HSD, respectively (Thibaut and Porte 2004; Young et al. 2005). In fish,









17p-estradiol is the active estrogen, but there is uncertainty regarding the roles of

testosterone or 11-ketotestosterone in androgenic activity (Kime 1995). Though the

process of steroidogenesis occurs only in specific tissues, they may have effects on other

systems.

OH
OH OH O




HO O
17-B-Estradiol Testosterone 11-keto-Testosterone

Figure 1-2. Active hormones found in fish.

Once steroids are produced, they can remain in steroidogenic tissue or travel to

other organs where they undergo metabolism, cause feedback inhibition, or regulate a

variety of responses including, vitellogenesis, development of secondary sex

characteristics, or reproduction itself (Arcand-Hoy and Benson 2001). In a negative

feedback loop, steroids target the hypothalamus and pituitary and inhibit further signaling

of steroid production via GnRH or gonadotropins, respectively (Arcand-Hoy and Benson

2001; Young et al. 2005). Steroids are metabolized primarily in the liver, but can also be

metabolized in other tissues (Sonderfan et al. 1989). Phase I enzymes metabolize steroid

hormones by hydroxylation and dehydrogenation and Phase II enzymes, such as UDP-

glucuronosyltransferase and sulfotransferase, are responsible for steroid metabolism by

conjugation (Matsui et al. 1974; Waxman 1988).

When steroids reach target organs they bind to hormone receptors. There are

several steroid hormone receptors that regulate expression of a suite of target genes via

genomic interaction by binding to promoters in those target genes. Recent evidence









suggests that effects elicited by hormones may also arise via a non-genomic pathway

(Loomis and Thomas 2000). Most hormonal responses, however, are believed to be

mediated through receptors and genomic interaction, which is a common target for EDCs

(Filby and Tyler 2005). There are at least three forms of estrogen receptors in fish (ERa,

ERP, ERy) to which estradiol binds (Filby and Tyler 2005). In female teleosts, binding of

estradiol to ERs stimulates transcription and production of vitellogenin, an egg yolk

protein precursor important in egg development (Young et al. 2005). There are two

likely forms of androgen receptor in teleosts, which regulate androgen-controlled genes

(ARa and ARP; Wilson et al. 2004). As hormones bind to these receptors, they regulate

expression of target genes through genomic interaction.

Endocrine disruption can occur when any part of the complex system of signals

within the HPG axis is altered. While disruption may occur at numerous targets, many

studies show EDCs act by interacting with hormone receptors (Kelce et al. 1995; Wilson

et al. 2004). A significant deviation of plasma hormone levels from normal is often an

indicator of endocrine disruption. Additionally, because males do not produce eggs, they

do not produce significant amounts of vitellogenin. Thus, an induction of vitellogenin in

male fish signifies endocrine disruption and, more specifically, exposure to estrogenic

compounds (Denslow et al. 1997; Nash et al. 2004). Estrogen and androgen receptors are

also likely targets of EDCs, where EDCs bind the receptor (Chedrese and Feyles 2001;

Kelce et al. 1995; Young et al. 2005). However, studies suggest that receptors' binding

affinities may differ among species, which has important implications for species to

species extrapolations (Wilson et al. 2004).









Endocrine disrupting compounds may also affect offspring during oocyte

development or embryonic development. Adverse effects to offspring may occur if

estradiol and/or progesterone are altered during the reproductive cycle because of their

critical influence on egg yolk protein and timing of release of oocytes (Nimrod and

Benson 1997). Additionally, embryonic exposure of fish to EDCs may influence normal

sexual development and growth (Arcand-Hoy and Benson 2001). Fathead minnows

undergo a series of physiological changes as they develop into reproductive male or

female adults. Those changes, which include phenotypic sex determination and

development of secondary sex characteristics, are largely controlled by levels of steroid

hormones (von Hofsten and Olsson 2005). An alteration in those steroid hormones may

lead to developmental dysfunction.

There is also mounting evidence that exposure of an adult to an EDC can affect its

offspring through epigenetic mechanisms (Anway et al. 2005; Collas 1998). Recently,

studies have been published on the effects of exposure to EDCs during egg development,

or in ovo. A multiple-generation study of fish exposed to enthynylestradiol found

reproductive dysfunction in offspring exposed in ovo (Nash et al. 2004). That study

showed no change in fertilities of adults exposed, but a reduction in fertility of the second

generation, even after depuration. The connection between abnormal steroid hormone

and vitellogenin levels in adults and altered reproductive output of their offspring is not

well understood. Thus, this study focused on biomarkers of p,p'-DDE exposure in adults

as they relate to effects in reproductive output of those adults and their offspring.

Biomarkers of Endocrine Disruption

Biomarkers are defined as biological responses that deviate from normal as a result

of exposure to a given stimulus (Mayer et al. 1992). These responses can be measured at









different levels of biological organization and include changes in gene expression,

hormone concentrations, reproductive output, and from a broader perspective,

populations (Korte et al. 2000; Sepulveda et al. 2002). There are two groups of

biomarkers: those of exposure and those of effect. Biomarkers of exposure are those that

simply indicate that an organism has been exposed to an EDC at some level. Biomarkers

of effect are those that indicate a degree of exposure with an EDC sufficient to result in

an impact on a higher level of biological organization. The distinction between

biomarkers of effect and biomarkers of exposure is often dependent on the endpoint of

interest. Much effort has been placed towards developing biomarkers of exposure and

effect of EDCs in fish by correlating exposure to contaminants with altered gene

expressions, hormone concentrations, and reproductive output (Ankley et al. 2001; Foran

et al. 2002; Giesy et al. 2000). Connections among these factors tend to vary, but some

are conserved across species and compounds.

Population- and organism-level biomarkers of exposure to EDCs include skewed

sex ratio, gonadosomatic index (GSI), and age at first reproduction. A skewed sex ratio

is often used as an indicator of a wild population exposed to EDCs. Physiological

biomarkers such as GSI, inappropriate presence of intersex gonads or secondary sex

characteristics, and age at sexual maturity require some measure of individual fish

(Ankley et al. 2001; Monnosson et al. 1997; Sepulveda et al. 2002). Reduced GSI has

been found in fish exposed to estrogenic and anti-androgenic contaminants in the

laboratory and in fish inhabiting contaminated sites (Ankley et al. 2001; Panter et al.

1998; Sepulveda et al. 2002). GSI, however, can be unaffected by exposure to

compounds that exhibit other endocrine disrupting properties (Bayley et al. 2002). Few









studies of fathead minnows have observed intersex gonads, which is the presence of

characteristics of both ovaries and testes in the gonad (Mills and Chichester 2005). Age

at first maturity is not a common biomarker for EDC exposure, which is probably due to

the time and difficulty involved in making such a measurement. Nonetheless, studies on

population dynamics have shown that time of first reproduction can play an important

role in population growth rate (Levin et al. 1996).

Molecular and genetic biomarkers that indicate exposure to contaminants include

DNA adducts, steroid hormones, vitellogenin induction in males, tumors, and thinned

eggshells (Denslow et al. 1997). Genetic biomarkers for reproductive dysfunction in

fathead minnows include abnormal mRNA levels for ERa, AR, vitellogenin, the

steroidogenic enzyme cytochrome P450 17a-hydroxylase, 17,20,lyase (P450c17), and

P450scc or aromatase (Denslow et al. 1997; Halm et al. 2003; Rolland et al. 1997;

Wilson et al. 2004). While mRNA expression is a useful biomarker that also elucidates

possible mechanisms of EDC action, abnormal steroid hormone and vitellogenin protein

levels are more common biomarkers of EDC exposure. In fish, the primary focus of

molecular biomarkers of exposure and effect has been on plasma vitellogenin, 170-

estradiol, testosterone, and 11-ketotestosterone (Giesy et al. 2000; Mills and Chichester

2005; Sepulveda et al. 2002).

Steroid hormone levels as well as the ratio of estrogens to androgens are also

common biomarkers of exposure and effect. Appropriate levels of steroid hormones are

used as a biomarker because they can be indicative of adverse affects on normal

reproduction. Identification of abnormal hormone levels first requires knowledge of

normal hormone levels during the reproductive cycle of each sex of a species of fish.









Jensen et al. (2001) described the basic reproductive biology of the fathead minnow

(Pimephalespromelas). Average plasma 17pestradiol and testosterone concentrations of

females were 5.97 1.12 and 3.08 0.34 ng / ml and of males were 0.40 + 0.13 and 9.11

0.92 ng / ml, respectively. Investigations in several species of fish, including fathead

minnows and largemouth bass, exposed to EDCs relate altered steroid hormone levels

with other reproductive endpoints such as GSI and egg output. For example, Giesy et al.

(2000) found a significant positive correlation between plasma estradiol and the number

of eggs produced per female fathead minnow exposed to 4-nonylphenol. However,

Makynen et al. (2000) found a reduction in female GSI, but no change in plasma steroid

hormone levels in fathead minnows exposed to vinclozolin. Additionally, largemouth

bass inhabiting contaminated lakes in Florida had both hormonal and reproductive

abnormalities (Guillette et al. 1994; Sepulveda et al. 2002). Steroid hormones can be

sensitive biomarkers of exposure to EDCs, regardless of their connection to other

reproductive endpoints.

Changes in steroid hormone levels can further affect production of vitellogenin,

which is initiated when estradiol bind to an ER. Because vitellogenin production is

activated by the ERs, its induction in males is often used as a biomarker of exposure to

estrogenic compounds (Denslow et al. 1999; Korte et al. 2000). Expression of

vitellogenin is sensitive to estrogenic action, but the response in mRNA levels is not as

persistent as the protein itself. This is because male fish do not have a mechanism for

clearing vitellogenin from the body (Korte et al. 2000). Although vitellogenin expression

in males is indicative of exposure to estrogenic compounds, it is not a consistently

reliable marker of reproductive dysfunction. A study in fathead minnows exposed to the









estrogenic compound 4-nonylphenol, found that correlations between vitellogenin and

estradiol differed when the study was repeated by the same investigator (Giesy et al.

2000). Additionally, plasma estradiol, but not vitellogenin was related to egg production

(Giesy et al. 2000). Further, studies show vitellogenin expression levels in females are

not necessarily correlated to hatching success of eggs from adults exposed to an

estrogenic compound during development (Cheek et al. 2001). Thus, transcript level of

vitellogenin is a good biomarker of recent exposure and vitellogenin protein level is a

better biomarker of exposure history to estrogenic compounds, but neither is necessarily

indicative of altered reproductive capacity.

The consequences of reproductive malfunctions caused by EDCs can be

detrimental to fish populations, which people depend upon for food and recreation (Cook

et al. 2003). Thus, it is important to understand how fish will respond to endocrine

disruptors to maintain healthy fish populations, especially in restoration (Bayley et al.

2002) sites. The present study measured a suite of biomarkers in fathead minnows

exposed to p,p'-DDE with the goal of providing a more comprehensive understanding of

the connections among them in the context of population level outcomes.

The fathead minnow is a small member of the minnow family, Cyprinidae, that is

easily raised in the laboratory and commonly used in toxicity assays. This species

reaches sexual maturity at 4-5 months and lives up to 4 years in the wild, where it feeds

primarily on invertebrates. Spawning activity can be induced by environmental

conditions such as photoperiod and water temperature. In a study of the basic

reproductive biology of fathead minnows, females spawned approximately 85 eggs every

four days (Jensen et al. 2001). Steroid hormones and GSI varies at each point in the









spawning cycle of females, but not males (Jensen et al. 2001). Vitellogenin is

occasionally detected in male fish not exposed to EDCs and is at a relatively constant

level throughout the spawning cycle in females (Jensen et al. 2001). The fathead minnow

is a good model for endocrine disruption studies because it is easy to work with in a

laboratory setting and there is considerable information on its basic reproductive biology.

p,p'-DDE in the Environment

Organochlorine pesticides (OCPs) are banned in most industrialized countries, but

their use continues in developing countries because they are relatively inexpensive,

effective, and easily manufactured. OCPs include compounds commonly known as

DDT, toxaphene, chlordane, methoxychlor, vinclozolin, and dieldrin. Because of

massive volumes used, atmospheric processes, and their persistence, they remain

common environmental contaminants (Kalantzi et al. 2001; Matsumura 1985).

Specifically, dichlorodiphenyltrichlorethane (DDT) is an organochlorine insecticide and

persistent environmental pollutant which is banned in the United States. DDT was

widely used as an insecticide in the US from the time it was discovered by Peter Miller

in 1939 until it was banned in 1972 (Carr and Chambers 2001). Detectable levels of

DDTs have been found in biological, geological, and atmospheric samples since the

1960s.

The primary metabolites of DDT are dichlorodiphenyl-dichloroethane (DDD),

dichlorodiphenyl-dichloroethylene (DDE), and dichlorodiphenylchloroethane (DDMU)

(Fig. 1-3). Isomers of DDT and its metabolites are collectively referred to as DDTs.

Soils are a major sink for DDTs (Grau and Peterle 1979), but they are flushed into

aquatic systems during flood events (Miglioranza et al. 2003). In soils, two isomers of

DDT, o,p-DDT and p,p-DDT, are dechlorinated to o,p-DDD and p,p-DDD by anaerobic









microorganisms (Huang et al. 2001). Finally, DDD isomers are degraded to isomers of

DDE. Most ingestion of DDTs in humans is believed to be ofp,p'-DDE itself, which is

absorbed in the gastro-intestinal tract and stored indefinitely in fat tissues (Moffat 1986).

A survey of people not occupationally exposed to DDT found levels as high as 17 [tg / g

fat (Moffat 1986). This study focuses onp,p'-DDE, the most stable and often most

abundant metabolite of DDT (Huang et al. 2001; Spies and Thomas 1997).


CI CI CI CI



CI Cl CI Cl Ci CI

p,p-DDT p,p-DDE p,p-DDD


Figure 1-3. DDT and selected metabolites.

Although DDT is banned in the US, its metabolites remain common contaminants

in locations of spills, sites of heavy and/or continued use domestically and abroad, and in

areas of atmospheric deposition. The ubiquitous and liberal use of DDT before it was

banned led to heavy contamination of many sites around the USA. In 1965 DDT was

applied at the rate of 4 lb / acre in southern Arizona (Matsumura 1985). Soils in Florida

wetland restoration sites that were formerly farmland contained up to 4,200 [tgp,p'-DDE

/ kg soil in the early 1990s, while fish tissues from sites had greater than 190 tg DDTs /

kg (Marburger et al. 2002). Typically, concentrations are directly related to the extent to

which DDT was used in a particular region (Kalantzi et al. 2001). Natural atmospheric

processes, however, can transport DDTs from areas of use to more pristine environments.

(Catalan et al. 2004) foundp,p'-DDE in the ng / g range in fish muscle tissue from a high

mountain lake of the Pyrenees (2240 m above sea level), where atmospheric deposition









was the sole source of OCPs. Microorganisms that degrade DDT intop,p'-DDE were

found in soils where the parent compound was never applied (Miglioranza et al. 2003).

The concentration of total OCPs in those soils was 656 ng / g dry weight. Though DDT

has been banned in many places, its continued use and physical properties are potentially

problematic in several regions of the world.

Studies analyzing DDTs from different levels of biological organization

demonstrate the abundance ofp,p'-DDE in the environment, as well as its propensity to

bioaccumulate. A study of DDTs and other chlorinated compounds found p,p'-DDE was

the most prevalent and abundant contaminant in fish tissues from Latvian freshwaters

(Valters et al. 1999). p,p'-DDE concentrations were found as high as 20 [tg / g in fish

ovaries (Marburger et al. 2002) and 5.8 tg / g in alligator eggs (Guillette et al. 1994)

from heavily contaminated sites in Florida. Additionally, kelp bass in coastal waters of

California had average liver concentrations of 3.43 tg / g DDTs, of which greater than

97% was p,p'-DDE (Spies and Thomas 1997). p,p'-DDE (log Kow = 6.5) is a highly

lipophilic compound prone to bioaccumulation. Fish muscle tissue concentrations of

p,p'-DDE can be as much greater than concentrations found in water. For example, in a

site where lake water contained 7.4 pgp,p'-DDE / L (parts per trillion), invertebrates

averaged 40.06 ngp,p'-DDE / g (parts per million), and brown trout (Salmo trutta) had

57.23 ngp,p'-DDE / g (Catalan et al. 2004). In California sea otters (Enhydra lutris)

DDTs were as high as 5,900 ng / g in liver and 4,600 ng / g in kidney of, while prey

concentrations ranged from 0.08 to 12.9 ng / g (Kannan et al. 2004).

p,p'-DDE: An Endocrine Disruptor

Initial studies on DDTs in wildlife focused on egg shell-thinning effects in birds,

especially raptors. The fist study of DDT on fish and wildlife was conducted in 1946 by









Cottom and Higgins. p,p'-DDE was shown to cause reproductive malfunctions in avian

species as far back as the 1960s (Heath et al. 1969). Additional observations and

investigations suggested DDTs contribute to endocrine disruption in fish (Macek 1968).

In past studies, the reproductive capacity of fishes was adversely affected by p,p'-

dichlorodiphenyldichloro-ethylene (DDE), the most persistent metabolite of DDT

(Bayley et al. 2002; Mills et al. 2001). There is significant evidence suggestingp,p'-

DDE adversely affects fish reproduction and populations. Surveys in Lake Michigan

showed a likely connection between a skewed sex ratio in bloater (Coregonus hoyi)

populations and p,p'-DDE concentration. The percent of female bloaters in Lake

Michigan returned to normal as fishp,p'-DDE concentrations decreased from

approximately 3.5 tg / g in 1969 to 0.75 tg / g in the early 1980s (Monnosson et al.

1997).

Several authors suggest that OCPs affect hormone homeostasis through steroid

synthetic and metabolic pathways (Hornung et al. 2004; Spies and Thomas 1997; Thibaut

and Porte 2004). Spies and Thomas (1997) found plasma estradiol levels in fish

decreased with concentration of DDTs. On a more mechanistic level, p,p'-DDE inhibited

steroid synthesis in mammalian ovary cells at 10 atM (Chedrese and Feyles 2001), but

enhanced steroid synthesis in fish testicular cells at 100 PM (Thibaut and Porte 2004).

Reproductive abnormalities were observed in male guppies exposed top,p'-DDE during

sexual differentiation, a time susceptible to the effects of endocrine disruption (Bayley et

al. 2002).

In mammals, p,p'-DDE affects transcription of androgen-controlled genes by

binding to the androgen receptor. In vitro studies in mammalian cells found 200 nM









p,p'-DDE inhibited half the androgenic transcriptional activity induced by a testosterone

(Kelce et al. 1995). Those same in vitro studies also suggestp,p'-DDE binds the AR,

allowing it to enter the nucleus, but preventing the AR from inducing androgen-

dependent genes. Kelce et al. (1995) foundp,p'-DDE acts as an androgen inhibitor with

potency similar to that of hydroxyflutamide (200 nM = IC50) in mammalian cell lines. In

fathead minnows, p,p'-DDE had a binding affinity for AR similar to dihydrotestosterone,

at concentrations of 20 and 22 nM, respectively (Wilson et al. 2004). However, activity

of the bound AR was not measured in that study. Consequently, there is no confirmation

that p,p'-DDE bound to the AR actually inhibits transcription of androgen-dependent

genes in fathead minnows.

Environmental concentrations ofp,p'-DDE have been found at 80 times the

concentrations that cause these anti-androgenic effects in vitro (Guillette et al. 1995;

Kelce et al. 1995; Monnosson et al. 1997). Other in vitro studies suggest thatp,p'-DDE

may increase granulosa cell growth by stimulating progesterone synthesis, but not

progesterone synthesis stimulated by 17p-estradiol (Chedrese and Feyles 2001; Crellin et

al. 1999). It is, therefore, likely thatp,p'-DDE does not disrupt hormone homeostasis by

interfering at the ER, but with another component of the steroidogenic pathway. p,p'-

DDE has also been shown to increase granulosa cell growth similar to, but less potently

than, 17p-estradiol in mammalian cells (Chedrese and Feyles 2001). From that same

study, Chedrese and Feyles (2001) found thatp,p'-DDE decreased progesterone, a

hormone required for normal ovulation. Although those studies were conducted in

mammalian cells, they indicate thatp,p'-DDE may not be acting at the estrogen-receptor.

Extrapolating the anti-androgenic or estrogenic activity found in those mammalian









studies to fish may be invalid. Several in vitro studies suggest there may be differences

in binding of contaminants to ARs among mammals and teleosts, between species of

teleosts, and between tissues of a single species (Bayley et al. 2002; Makynen et al. 2000;

Wells and Van Der Kraak 2000).

Given the uncertainty in the link between AR binding in vitro and in vivo activity,

Bayley et al. (2002) concluded that sex characteristics and reproduction themselves were

the best measure of reproductive dysfunction. Male guppies (Poecilia reticulata)

exposed tol0 [tg / g p,p'-DDE during sexual development had a sex ratio skewed toward

females, increased time to male development, and altered secondary sex characteristics

and sperm count (Bayley et al. 2002). A captive population of trout exposed to 10, 40, or

80 [tg / gp,p'-DDE in ovo did not have altered sex ratio or reproductive dysfunction upon

reaching sexual maturity (Carlson et al. 2000). Several other reproductive endpoints,

including gonodosomatic index (GSI), egg production, and fertilization success, were

also not affected in trout exposed top,p'-DDE in ovo (Carlson et al. 2000). Increased

mortality was, however, observed in progeny spawned from males treated withp,p'-DDE

in ovo and uncontaminated females. As those studies show, the effects ofp,p'-DDE on

fish can vary greatly among species and can depend on the life-stage at the time of

exposure.

Linking Biomarkers of Endocrine Disruption to Fish Populations

Reproductive effects of EDCs have been observed extensively at biochemical and

physiological levels of biological organization in fishes. Linking biomarkers measured in

individual fish to population-level outcomes has only recently been attempted (Grist et al.

2003) and links between biomarkers and real population-level effects have yet to be

confirmed (Mills and Chichester 2005; Segner 2005). Most studies investigating the









effects of EDCs on population parameters and demographic changes have been on

invertebrates because they have shorter generation times and are easier to raise in the

laboratory than fishes (Barata et al. 2002; Mauri et al. 2003; Raimondo and McKenney

2005). Much of the concern and research of EDCs, however, is related to fish because

humans depend on them for nutrition and income. Consequently, there is a disparity

between the knowledge of population-level outcomes and the amount of data on

biomarkers endocrine disruption in fish.

A limited number of recent studies have attempted to link adult exposure of fishes

to EDCs with changes in population growth rate. A study by Miller and Ankley (2004)

computed the effects of a synthetic androgen on density-dependent population growth

rates of fathead minnows. Grist et al. (2003) investigated the contributions of

demographic parameters to changes in population growth rate of fathead minnows

exposed to ethynylestradiol during development. They found fertilities contributed more

than survival probabilities to the highly significant correlation between ethynylestradiol

concentration and population growth rate. Those studies do not, however, account for

effects of in ovo exposure on the survival probabilities and fertilities.

Few investigations on multi-generational effects of contaminant have been made.

Demographic parameters of offspring exposed to EDCs maternally may be of great

importance. Nash et al. (2004) exposed two generations of fathead minnows to

environmentally relevant concentrations of the potent estrogen, ethynylestradiol. That

study showed no change in fertilities of exposed adults, but a reduction in fertility of the

second generation, even after depuration. As Nash et al. (2004) concluded, those

findings carry major implications for population-level impacts of long-term exposure of









fish to EDCs. Therefore, it is important to investigate second-generation effects of EDC

exposure.

Ankley et al. (2001) developed a protocol for measuring reproductive effects of

sub-chronic exposure to EDCs of fathead minnows, a species commonly used in such

assays. That and other studies call for a deeper knowledge of the connection between

contaminant tissue burdens, reproductive dysfunction, and population effects (Chedrese

and Feyles 2001; Foster et al. 2001; Gray et al. 2002; Orlando et al. 1999). Employing an

exposure methodology similar to that described by Ankley et al. (2001), I investigated the

effects of parental exposure top,p'-DDE on a suite of biomarker and population

parameters of fathead minnows. Because EDCs act on the reproductive system by

definition, I expected fertilities to change more than survival among treatment groups.

Feed concentrations ofp,p'-DDE were chosen to represent environmentally

relevantp,p'-DDE body burdens (Marburger et al. 2002; Muller 2003). In this case, the

stimulus for altered biological responses is exposure to various concentrations ofp,p'-

DDE through the diet. This study had three overall goals: to assess the effects ofp,p'-

DDE on reproduction and endocrinology of fathead minnows, to assess the effect of in

ovo exposure top,p'-DDE on survival and development, and to link those effects with

biomarkers of exposure. To this end, I conducted a pilot experiment in which fish were

exposed to high levels ofp,p'-DDE and the anti-androgen flutamide. Then, I conducted a

dose-response experiment, from which the effects of in ovo exposure were assessed.

Finally, the dose-response experiment was repeated to obtain additional biological

materials to use for measuring biomarkers.










Table 1-1. Classes of endocrine disrupting compounds and examples.
Compound class Type Examples
organochlorine pesticides (OCPs) DDT, methoxychlor, toxaphene

Pesticides organophosphate pesticides (OPs) TEPP, chlorpyrifos, malathion
extracts of( hC /i, \',,iil il
pyrethrins flowers
flowers
polycyclic aromatic hydrocarbons (PAHs) benzo(A)pyrene, aflatoxin
Industrial
bypducts plasticizers di(n-butyl)pthalate
byproducts
polychlorinated biphenyls (PCBs) Aroclor 1254
Hormonal ethynylestradiol
Hormonal p birth control pills, hormone therapy ethynylestradiol
therapeutics
Natural hormone estradiol, testosterone
(Cook et al. 2003; Macek 1968; Miller and Ankley 2004; Mills et al. 2001; Nash et al. 2004; Thompson et
al. 2004; Valters et al. 1999)















MATERIALS AND METHODS

Three in vivo experiments were conducted on fathead minnows (Pimephales

promelas). Each experiment was designed to optimize collection of the endpoint of

interest. The focus of the initial pilot experiment was to assess ifp,p'-DDE caused

reproductive dysfunction similar to the anti-androgen flutamide. Upon observing

reproductive dysfunction in fish administered a high level ofp,p'-DDE, a dose-response

experiment to determine a no observed effect level (NOEL) was conducted. During that

second experiment fish were fed one of three environmentally plausible p,p'-DDE

concentrations. That experiment focused on the accumulation rate ofp,p'-DDE in adults

and survival, development, and reproduction of offspring spawned from those adults. In

a third experiment, p,p'-DDE was administered to fish similar to the second experiment

to obtain additional biological materials for hormone and gene analyses. All experiments

were conducted at the University of Florida Aquatic Toxicology Facility under the same

general environmental conditions in accordance with IACUC protocols. Differences in

experimental conditions are described below. Methods for determination ofp,p'-DDE

content, plasma 17p-estradiol levels, and mRNA expression were identical across

experiments unless otherwise noted.

General Methods

Fish Holding Conditions

Fish were housed in flow-through tanks supplied with dechlorinated water and kept

on a 16 hours light: 8 hours dark schedule. Fish were exposed top,p'-DDE through









contaminated Silvercup Trout Chow (Zeigler Brothers, Inc.). Food was stored at 4 OC.

Water temperature was measured daily, while dissolved oxygen (DO), pH, hardness

measured as CaCO3, and total ammonia were monitored weekly. Water quality

parameters were as follows for flow-through tanks: DO was 8.5-8.9 mg/L; pH was 8.7 +

1; hardness was 40 2 mg CaCO3/L; and total unionized ammonia was always less than

0.5 mg/L in flow-through tanks.

Reproductive Measures

Reproductive output was measured as the number of eggs produced, percentage of

eggs fertilized, and percentage of fertilized eggs that hatched. One spawning substrate

(3-inch sections of 3-inch diameter polyvinyl chloride pipe) per male fish was kept in

each tank at all times. Spawning substrates were checked for eggs each afternoon. If

eggs were present, the spawning substrate was removed, eggs counted, and placed in an

aerated 2 L glass beaker filled with approximately 1.75 L dechlorinated water and 25 ml

blackwater extract (Aquatic Ecosystems, Inc.). Blackwater extract was used to prevent

fungal infection on eggs, which was determined the best method for preventing fungal

infection. The number of eggs fertilized was determined by counting eggs that developed

eyes 2-3 days post spawn (dps). Water within the beaker was changed when fertilization

was determined. Digital overhead photographs were used to count the number of eggs

that hatched 5-8 dps, depending on when eggs were no longer present on the substrate.

Plasma and Tissue Collection

At the end of each exposure period adult fish were anesthetized with MS-222

(100mg/L buffered with 200 mg NaHCO3/L), killed by decapitation, bled, and tissue

collected (Ankley et al. 2001). Blood was collected from the caudal sinus in heparinized

micro-hematocrit capillary tubes (Fisher Scientific Company), centrifuged at 1,500 X g









for 10 minutes, and plasma was removed and frozen at -80 OC. Gonads were excised,

weighed for determination of the gonadosomatic index (GSI = [gonad weight / body

weight] x 100), and flash frozen in liquid nitrogen. Liver and brain were removed, flash

frozen, and stored at -80 C. Carcasses were eviscerated and stored at -20 OC until

analyzed forp,p'-DDE content.

Determination ofp,p'-DDE Content

Fish were analyzed for muscle tissue or egg p,p'-DDE concentration (wet weight)

by gas chromatography / mass spectrometry by the method described in (Glesleichter et

al. 2005) and modified as follows. One gram or 2.5 g were sectioned from the

eviscerated carcass posterior of the opercle of each female and male fish, respectively.

The fish tissue was homogenized by a Tekman Tissumizer (Tekman Company) with 3 [tg

dl0-phenanthrene as an internal standard (Protocol Analytical, LLC), 2.5 times tissue

weight ofNa2SO4 (A.C.S. Grade, Fisher Scientific Company), and 7 ml n-hexanes

(A.C.S. Gade, Fisher Scientific Company), vortexed, and centrifuged for 15 minutes at

approximately 100 X g. The supernatant was decanted and the homogenate was

extracted twice more with 3 ml n-hexanes. Extracts were combined to yield a total

extracted volume of 13 ml for each tissue sample. The extract was dried under a stream

of nitrogen at 35 C. The dried extract was reconstituted in 3 ml acetonitrile (Optima

Grade, Fisher Scientific) and eluted through a pre-conditioned SPE-C18 cartridge

(Agilent Technologies), which was repeated once. The cartridge was rinsed with 1 ml

acetonitrile. The eluate was then passed through an SPE-NH2 cartridge (Varian, Inc.)

and the glass tube containing the eluate was rinsed with 1 ml acetonitrile, which was also

placed over the SPE-NH2 cartridge. The final eluate was dried under a stream of









nitrogen at 35 C and reconstituted in 1 ml 3 [g dl0-pyrene / ml cyclohexane (Ultra

Scientific).

A Shimadzu 17A gas chromatograph (Shimadzu Scientific Instruments) with HP-

5MS column dimensions of 29 m x 0.25 mm coupled with a Shimadzu QP-5000 were

used for analyte separation and detection. One microliter of reconstituted extract was

injected into a splitless inlet at 275 C. Analytes were separated using the following

program: initial oven temperature was 100 C held for 2.5 min. Temperature was ramped

to 190 OC at 15 C / min, then to 250 C at 5 OC/min and finally to 2900C at 200C/min,

which was held for 5 min. Initial carrier flow was 1.4 ml/min. This was reduced to 1

ml/min at 2.5 min for the remainder of the program. Interface temperature was

maintained at 2800C. Mass spectrometer was operated in selected ion monitoring (SIM)

mode and m/z 246 and 317 were used forp,p'-DDE and m/z 188 was collected for dl0-

phenanthrene. Quantitation was performed using the ratio of area of m/z 246 to m/z 188

for each p,p'-DDE concentration.

Determination of Plasma 17p-Estradiol

Hormone levels were determined by enzyme immunoassay (EIA) after plasma was

extracted with organic solvent. Briefly, 180 dl EIA buffer was added to 10 [l plasma to

increase aqueous volume. The plasma and buffer mixture was then extracted twice with

0.75 ml ethyl ether (Pesticide Grade, Fisher Scientific Company). The ethyl ether extract

was evaporated under a gentle stream of nitrogen in a water bath at 30 OC. The extract

was reconstituted in 200 l EIA buffer, mixed by vortex, and placed on an orbital shaker

at 4 C overnight. This method of extraction and reconstitution was validated by

counting a known amount of extracted and reconstituted 3H-estradiol stock solution and

comparing it with a known amount of unadulterated stock solution. The efficiency of the









extraction and resuspension was greater than 90 percent. Estradiol standards were

prepared by bringing 10ld of each 15 ng/ml, 10 ng/ml, 5 ng/ml, 2.5 ng/ml, 1.25 ng/ml,

0.625 ng/ml, 0.312 ng/ml, 0.156 ng/ml, 0.078 ng/ml, and 0.039 ng/ml to a total volume of

200 [l in EIA buffer, mixed by vortex, and placed on an orbital shaker at 4 OC overnight.

Standards and reconstituted sample were mixed by vortex immediately prior to 173-

estradiol determination. Each standard and sample was measured for estradiol in

duplicate according to protocol for estradiol EIA kit (Cayman Chemical Company). The

quantifiable range for this assay was from 0.156 to 15 ng estradiol / ml plasma. An effort

to analyze estradiol in small volumes of plasma by radioimmunoassay was made, but

acceptable validation was not achieved (see Appendix).

Vitellogenin mRNA Quantification

RNA was isolated from liver using Trizol (Invitrogen), following the

manufacturer's instructions and reconstituted in RNA Secure (Ambion). RNA quality

was verified on ethidium bromide-stained 1.5% formaldehyde-agarose electrophoresis

gels. RNA was considered to have acceptable purity when the A260 nm/A280 nm ratios

were greater than 1.8, as determined on a spectrophotometer (NanoDrop Technologies).

cDNA was made by reverse transcription-polymerase chain reaction (PCR) using random

decamer primers and 2 [tg DNA-free RNA according to manufacturer's instructions for

the RETRO-Script Kit (Ambion). The quality of cDNA was verified by gel

electrophoresis of products from the following PCR program using 18S primers: the

reaction was held at 94 OC for 2 min, then 35 cycles of 30 sec at each 94 C, 55 C, and

72 C, and a final extension of 5min at 72 C. Real time-PCR was conducted using pairs

of oligonucleotide primer sequences for vitellogenin (Table 1), using 18S as the

housekeeping gene (Filby and Tyler 2005).









Experimental Set-Up

Pilot Experiment

The goal of the pilot experiment was to determine ifp,p'-DDE affected

reproductive output and 17p-estradiol levels of fathead minnows similar to the anti-

androgen flutamide. To meet this objective, adult fathead minnows (7-10 months old)

were obtained from a local fish breeder (Fish Soup, Newberry, FL. One male and two

females were housed in 5-gallon flow-through tanks. There were six replicate tanks

within each treatment group, totaling 18 tanks for the entire experiment. Mean water

temperature was 22 + 2 C. Treatment groups included a control group (vehicle only

feed), a positive control group (1208 [tg flutamide / g feed), and ap,p'-DDE group (900

tg p,p'-DDE / g feed). Contaminant concentrations were measured by the Analytical

Toxicology Core Laboratory at the University of Florida. Fish feed was coated with

menhaden oil and acetone containing the appropriate chemical concentrations, mixed,

and placed under a fume hood overnight. p,p'-DDE (2,2-bis(4-chlorophenyl)-l,1-

dichloroethylene, 99.4% purity) was obtained from Aldrich Chemical Company. Fathead

minnows housed at the facility consumed menhaden oil-coated feed without

discrimination.

Adult survival and reproductive output were measured for 21 days prior to

chemical exposure. During that pre-exposure period fish were fed approximately 0.5 g

feed twice and live brine shrimp, Artemia (Great Lake Artemia) once daily. Once egg

production was established during the pre-exposure period, the 26-day exposure period

immediately followed the pre-exposure period. During the exposure period fish were fed

~ 2 gp,p'-DDE -contaminated feed twice daily and no brine shrimp. For this experiment,

the total number of eggs per female was used as the sole measure of reproductive output.









At the conclusion of the exposure period fish were sacrificed and plasma 17p-estradiol

levels were determined as described above.

p,p'-DDE Dose-Response Experiment I: Accumulation Rate and Reproductive
Output of Adults

Adult fathead minnows (5-7 months old) were obtained from Aquatic Biosystems,

Inc. (Fort Collins, CO). Each treatment group consisted of four males and eight females

housed in 30-gallon flow-through tanks. Water temperature for experiment two was 25 +

1 C in flow-through tanks and 25 2 C in egg-housing beakers. p,p'-DDE was

dissolved in acetone, sprayed onto fish food, mixed for several minutes, and placed under

a fume hood to allow the acetone to evaporate overnight to obtain nominal concentrations

of 2.5 utg/g, 25 utg/g, and 250 utg/g. Actual feed concentrations ofp,p'-DDE were 0.03 [tg

/ g, 1.63 tg / g, 11.48 [tg / g, and 104.25 [tg / g, which will be referred to as control, low,

medium, and high groups, respectively. Each tank was fed approximately 2 g of food at

both 10 AM and 6 PM daily for 29 days. Fathead minnows housed at the facility

consumed vehicle only feed without discrimination.

Reproductive output measured daily throughout the exposure period, as described

above. Upon determining sex at sacrifice, it was evident that each treatment group did

not have the same number of females throughout the experiment. This is due to the fact

that some males did not display secondary sex characteristics and therefore were

mistaken for females when allocated to each tank at the outset of the experiment.

Consequently, the number of eggs produced in each treatment group was calculated on a

per female basis.

One male and two females from each treatment group were killed and tissues

collected as described above on days 7, 14, 21, and 29 of exposure top,p'-DDE -









contaminated feed. Body burden was determined by gas chromatography / mass

spectrometry as described above. Reproductive output was monitored as described above

during the 28-day exposure period. Eggs collected during the 28-day exposure period

were grouped into four clutches based on the day they were spawned: clutch one was

spawned from 2 d to 9 d; clutch two was spawned from 10 d to 16 d; clutch three was

spawned from 17 d to 23 d; clutch four was spawned from 24 d to 29 d. Clutch three and

four were used to study the effects of in ovo p,p'-DDE exposure, as described in the

following section.

Survival, Development, and Reproductive Output of Offspring

Offspring studies were conducted on eggs spawned from thep,p'-DDE dose-

response response experiment of control adults or adults exposed to low, medium, and

high doses ofp,p'-DDE. Offspring for each treatment group were pooled based on day

spawned into a 2.5-gallon flow-through tank. Survival of offspring was measured weekly

for three weeks post-hatch by counting fish in digital overhead photographs taken of each

beaker. During those three weeks offspring were kept in 2 L flow-through tanks and fed

live Artemia, hatched in artificial sea water. After four weeks, 100 offspring (juveniles)

from clutch four were taken from each treatment group and placed in 5-gallon flow-

through tanks with spawning substrates. Offspring from the low treatment group were

taken from clutch three because adults did not spawn during the final six days of the

experiment. All further studies on offspring were conducted on these 100 fish. Survival

of these fish was measured monthly. Secondary sex characteristics and first spawn where

monitored approximately every other day after four months post-spawn. The offspring

were transferred to 30 gal flow-through tanks and provided spawning substrates at 3.5

months of age. Reproductive output of offspring was measured, as described above, for









one week each month after the group initiated spawning. Offspring length and weight

were measured approximately four months post-spawn and killed for tissue collection

and measurement at approximately nine months of age. These fish were not analyzed for

p,p'-DDE content, estradiol or gene expression.

p,p'-DDE Dose-Response Experiment II: Collection of Biological Materials

Fathead minnows were bred in-house from fish received from Fish Soup

(Newberry, FL). At the time of use the fish were 12-15 months old. A fresh batch of

p,p'-DDE -contaminated feed was prepared to achieve concentrations similar to those in

the firstp,p'-DDE dose-response experiment. Actual p,p'-DDE concentrations for this

batch of feed were 0.03 [tg / g, 1.38 [tg / g, 15.58 [tg / g, and 104.73 [tg / g, which was fed

to the control, low, medium, and high groups, respectively. Reproductive output was not

measured during this experiment. Plasma and tissues from this experiment were used to

determine 17p-estradiol levels, p,p'-DDE tissue concentration, and vitellogenin mRNA

expression in fish with body burdens similar to those achieved in the firstp,p'-DDE dose-

response experiment.

Statistical Analyses

Analysis of variance (ANOVA) followed by post-hoc comparisons using Tukey's

HSD were used to test for significant differences among treatment groups. Data were

log-transformed to meet the assumption of normality as necessary. Two-tailed Pearson's

correlations of measures taken in the same fish were computed. Values are reported as

the mean + standard error of the mean (S.E.). Significance was set at p<0.05. All

statistical analyses were conducted in SPSS 13.0 for Windows (SPSS, Chicago, IL,

USA).









Two-tailed Pearson's correlation ofp,p'-DDE concentration (control, low, medium,

or high) to biomarkers in fish exposed through feed (i.e. GSI and plasma) estradiol and

fish exposed in ovo (i.e. GSI) and population-level effects (i.e. survival and measures of

reproduction) to identify biomarkers of multi-generational effects. Because there were no

replicates, these correlations are not robust, and therefore only simple relationships were

further investigated and displayed.

Table 2-1. Primer sequences used for quantitative real-time PCR.
Gene Forward Primer Reverse Primer
ll 5'-GCT GCT GCT CCA TTT CAA AAG- 5'-GTG AGA GTG CAC CTC AAC GC-
Vitellogenin 3' 3'















RESULTS

Pilot Experiment

Egg output was measured in treatment and control groups during a pre-exposure

and exposure period. Fish in each tank demonstrated a capability for spawning during

the pre-exposure period. Fish in thep,p'-DDE group ceased spawning activity after just

three days of exposure to the p,p'-DDE, while fish in the control and flutamide groups

continued to spawn during the exposure period (Fig. 3-1). During the exposure period

the p,p'-DDE group spawned less than 16 eggs per female, while the control group

spawned 195 eggs per female and the flutamide group spawned 240 eggs per female.

Plasma 17p-estradiol (estradiol) was measured in control fish and fish treated with

p,p'-DDE or flutamide to elucidate whether the two compounds act by the same

mechanism in the fathead minnow. Mean plasma estradiol in males from the flutamide

group was 4.0 + 0.56 ng / ml, which was significantly greater than 2.02 0.35 ng / ml in

the control group and 1.88 0.27 ng / ml in thep,p'-DDE group (ANOVA, p<0.01; Fig.

3-2). There were no significant differences in estradiol levels in females among

treatment groups (ANOVA, p=0.29). The mean estradiol level in females in thep,p'-

DDE group (4.82 2.65 ng / ml) was, however, approximately half of the mean levels

measured in both the control (8.96 1.72 ng / ml) and flutamide (10.50 2.75 ng / ml)

groups. Plasma estradiol levels were significantly different between males and females

of the control group (p=0.01), but not the p,p'-DDE or flutamide groups (ANOVA,

p=0.63 and p=0.12, respectively). Although plasma estradiol levels were much greater in









females (10.51 2.75 ng / ml) than males (4.01 0.58 ng / ml) in the flutamide group,

the difference between the two sexes was not statistically significant (ANOVA, p=0.12).

p,p'-DDE Dose-Response Experiments I & II: Adult 17p-Estradiol, GSI, and
Reproduction

p,p'-DDE accumulation rates in fish tissue and reproductive output of adults were

measured from fish in thep,p'-DDE dose-response experiment I. Biomarkers including

adult GSI, plasma 17p-estradiol, and vitellogenin gene expression reported here were

measured from fish from p,p'-DDE dose-response experiment II. Survival of adult fish

was not affected in any of the treatment groups.

Concentration ofp,p'-DDE in adult muscle tissues was quantified in three fish at

seven-day intervals during the exposure period. p,p'-DDE tissue concentrations appeared

to approach, but not at reach steady state by the end of the experiment (29 days; Fig. 3-3

A and Fig. 3-3 B). Fish taken at the end of the exposure achievedp,p'-DDE muscle

tissue concentrations (wet weight) of 0.38 0.05 tg /g, 3.11 + 0.57 [tg / g, and 21.14 +

2.72 [tg / g in the low, medium, and high dose groups, respectively. The concentration of

p,p'-DDE in control fish, 0.12 0.10 atg / g (n=6), was used as the concentration on day

one for each group.

Plasma from these fish was used to determine 17p-estradiol levels. Plasma

estradiol levels, as measured by EIA, differed among treatment groups. Females in the

high group had mean plasma estradiol level of 6.70 2.53 ng/ml (n=5), which was lower

than the control, low, and medium groups with average levels of 9.38 1.09 ng/ml (n=5),

10.82 2.04 (n=5), and 10.48 2.13 (n=5), respectively. There were no significant

differences in female estradiol levels among treatment groups (Fig. 3-4; ANOVA,

p=0.51). Plasma estradiol in females was not significantly correlated top,p'-DDE









muscle tissue concentration or GSI (R2 =-0.34, p=0.35 and R2 =-0.49, p=0.60 ,

respectively). Males in the low, medium and highp,p'-DDE groups had mean plasma

estradiol levels of 7.72 2.85 ng/ml (n=5), 3.85 0.38 ng/ml (n=4), 5.58 2.63 ng/ml

(n=4), whereas 0.37 0.12 ng/ml (n=6) was the mean in the control group. Males in all

p,p'-DDE groups had significantly higher plasma estradiol levels than males in the

control group (Fig. 3-4; ANOVA, p<0.01). Plasma estradiol in males was not

significantly correlated top,p'-DDE muscle tissue concentration or GSI (R2 =-0.26,

p=0.82 and R2 =-0.19, p=0.68, respectively). Plasma estradiol was significantly different

between males and females in the control (ANOVA, p<0.01) and medium groups only

(ANOVA, p=0.03).

While there were no significant differences of either sex among treatment groups,

the high groups tended to have noticeably lower GSI than control (Fig. 3-5). GSI in

males was 1.34 0.23 (n=6) in the control group, 1.16 0.26 (n=6) in the low group,

1.15 0.29 (n=6) in the medium group, and 0.97 0.13 (n=4) in the high group

(ANOVA, p=0.81; Fig 3-5 A). In females GSI was 7.96 0.73 (n=6) in the control

group, 9.32 1.53 (n=6) in the low group, 8.36 1.27 (n=6) in the medium group, and

5.67 0.98 (n=7) in the high group (ANOVA, p=0.16; Fig 3-5 B). GSI was not

significantly correlated top,p'-DDE muscle tissue concentrations in males or females

(p=0.56 and p=0.13, respectively).

No adverse effect ofp,p'-DDE treatment was observed in reproductive output in

fish from the low and mediump,p'-DDE groups. Fish in the high group, however,

spawned 77 percent fewer eggs than the control group at the conclusion of the experiment

(Fig. 3-6). The percent of eggs successfully fertilized was at least 90 percent in the









control, low and medium groups, but only 65 percent in the high group (Fig. 3-7 A).

Similar to fertilization success, the percent of fertilized eggs that hatched did not show a

dose-related response (Fig. 3-7 B). Percent hatch in the low group was 68 percent, but in

all other groups percent hatch exceeded 93 percent.

Frozen tissue samples of liver were used to compare expression of vitellogenin in

treated fish to control fish. Results for this analysis were not definitive due to

amplification of multiple products by primers for the housekeeping gene, 18S. Because

multiple products were amplified, 18S could not be used to normalize and quantify the

relative expression of the gene of interest, vitellogenin. To troubleshoot this problem, I

used a new stock of the 18S primer, which also amplified multiple products. I tried

amplifying other housekeeping genes, including L8 and P-actin, but had no success.

Effects of in ovo Exposure on Survival, Development, and Reproduction

Concentration ofp,p'-DDE was measured in eggs spawned from adults p,p'-DDE

dose-response experiment II. p,p'-DDE in eggs spawned from the control group was

below the limit of detection. The concentration ofp,p'-DDE in the eggs from the

medium and high groups was approximately 30 percent of the concentration found in

adult females in the medium and high p,p'-DDE groups (Fig. 3-8). Eggs from the

medium group contained 1.40 [tgp,p'-DDE /g (n=l) and adults contained 3.11 0.37 [tg

p,p'-DDE /g muscle tissue (n=3). In the high group, eggs contained 7.60 [tgp,p'-DDE /g

(n=l) and adult muscle tissue had 21.14 2.72 (n=3). Eggs spawned from the low group

were not available forp,p'-DDE analysis.

All measurements of offspring are from eggs spawned in the p,p'-DDE

accumulation and reproductive output of adults experiment. Weekly survival was

measured for the first three weeks of life to assess early life-stage mortality related to









p,p'-DDE consumption of adults. There was considerable variability among treatment

groups in weekly offspring survival during the first three weeks after hatch (Table 3-1).

Offspring in the low and control group had identical average weekly survival

probabilities during the first three weeks after hatching. The medium offspring showed

the highest survival probability of 0.87 and the high group the next highest of 0.79. After

one month of life, there were only minor differences in monthly survival probabilities

among treatment groups, which were greater than or equal to 0.97.

The effects of in ovo exposure ofp,p'-DDE on FHM development were measured

as body size at four months of age, appearance of secondary sex characteristics, and age

at first reproduction. Length and weight were measured when fish reached

approximately four months of age to assess differences in growth among treatment

groups. Fish in the low group weighed significantly more than all other groups

(ANOVA, p<0.01; Fig. 3-9). The medium group displayed secondary sex characteristics

15 days before offspring in the control and low groups (Fig. 3-10), while the high group

displayed secondary sex characteristics 22 days after the control and low groups (Fig. 3-

10). Age at first reproduction in the high group was 33 days greater than in the control,

low, and medium groups. The low group showed sexual development identical to control

offspring.

Fish exposed in ovo were killed at nine months of age for determination of sex

and gonad weight. Sex ratios differed considerably among treatment groups, but showed

no dose-dependent trend (Fig. 3-11). The control and high group ratios were close to

50:50, but the low group was skewed toward females and the medium toward males.

Average male GSI from offspring in the high group was significantly greater than all









other groups (ANOVA, p<0.01; Fig. 3-12 A). Females GSI did not differ significantly

among treatment groups (ANOVA, p=0.07; Fig. 3-12 B).

Reproductive output of fish exposed in ovo did not show the same pattern among

treatment groups as either sex ratio or GSI. The total number of eggs produced in each

group over nine months from greatest to least was from the low, medium, control, and

high groups (Fig. 3-13 A). Cumulative egg production was also calculated on a per

female basis, where sex ratio was used to estimate the number of females in each

treatment group during each collection period. On a per female basis, the number of eggs

per female from greatest to least was from the medium, low, control, and high groups

(Fig. 3-13 A). By both measurements of egg production, the high group produced

considerably less eggs than all other treatment groups.

Identification of Biomarkers

Correlations of among average measure of biomarkers and reproductive effects

were conducted using each treatment group (n=4). Biomarkers and reproductive effects

that were related were graphed for visual representation. Among adults, average female

estradiol and GSI were directly related to the percent of eggs fertilized and average male

estradiol was related to the number of eggs per female (Fig. 3-14 and Fig. 3-15).

Average female estradiol and GSI were inversely related to GSI in both sexes of their

offspring (Fig. 3-16).










Table 3-1. Survival probabilities for offspring spawned from adults fedp,p'-DDE during
days 23-29. Survival probabilities were measured weekly for the first three
weeks and monthly from one month through eight months post-hatch. Sample
sizes are in parentheses next to each survival probability for 1-3 weeks post-
hatch. After one month, fish populations were normalized to 100 offspring in
each treatment group.
Average Survival Probabilities by Treatment Group


Control Low Medium High

Mean weekly survival 0.69 0.69 0.87 0.79
for three weeks post (n=254) (n=288) (n=645) (n=152)
hatch
Mean monthly 0.99 0.99 0.97 0.98
survival for 1-8
months post hatch


Control
- 900 ppm DDE
1208 ppm flutamide


-25 -15 -5 5 15 25
Time (days)

Figure 3-1. The number of eggs produced per female before and after exposure to either
p,p'-DDE of flutamide as compared to control. Eggs per female during the
pre-exposure period is shown to the left of the y-axis and eggs per female after
fish were administered contaminated feed is shown to the right of the y-axis.
There were 10 females in the control group, 7 in the p,p'-DDE group, and 12
in the flutamide group.












14 Male
I I Female

a 12 -
1
05
10 -
S*

8-



6-


Sa


2
S10 5 12


Control DDE Flutamide

Treatment Group

Figure 3-2. Comparison of plasma 173-estradiol levels in males and females within and
among each treatment group. Values represent mean S.E. Significant
differences among males among treatment groups are indicated by different
letters (p<0.05). Significant differences between males and females within a
treatment group are indicated by (p<0.05).


05-


04
.3
"04-
t-



0
O
CD
0
0 02-
0)


.U
03

Q
Q


0 7 14

Study Day


21 28


-*- Low


uu -' i --- i --- i --- i ----
0 7 14 21 28

Study Day


Figure 3-3. Mean p,p'-DDE muscle tissue concentration (tg p,p'-DDE / g wet weight
muscle tissue). A)p,p'-DDE measured in three fish from each treatment
group every seven days during exposure. B) Expanded view of meanp,p'-
DDE tissue concentrations measured in three fish from the low dose group
every seven days during exposure. The concentration of three control fish
represents the starting p,p'-DDE body burden for all treatment groups. Values
are the mean S.E.


30-


. 25-

0-


0 15-
0





0
(D 10-

- 5

Q


I I I I I












14


12
E
0)
10
0
-o

- 8
a,
(D

S 6
E
cj



r 2


0
0


Control Low Medium High

Treatment Group


Figure 3-4. Comparison of plasma 17j-estradiol levels in males and females from each
treatment group. Values represent mean + S.E. Significant differences
between males and females within a treatment group are indicated by *
(p<0.05). Different letters indicate statistically significant differences of the
same sex among treatment groups (p<0.05). The number males in each
treatment group were n=6 (control), n=5 (low), n=4 (medium), and n=4
(high). The number females in each treatment group were n=6 (control), n='
(low), n=5 (medium), and n=5 (high).
A Males


1.6 -

1.4-

L 1.2-
(U)
S1.0-

0 0.8-
r
0.6-

0.4 -

0.2 -


m control
i low
medium
- high


uJ



CU

* I
4-


0.0 'I1- I. I
Control Low Medium High

Treatment Group


B Females


_T_


Control


Low Medium High
Treatment Group


Figure 3-5. Mean GSI values of adults treated withp,p'-DDE-contaminated feed. A)
Male GSI. B) Female GSI. Error bars represent one standard error. Sample
size is six for the control, low, medium groups and five for the high group.











500



400
(D

E
a
uL 300

1-
0)
C.
()
' 200-

O
UJ

4-
- 100-
E
z


-- Control
-v- Low
-- Medium
-o- High


Exposure Day


Figure 3-6. Cumulative number of eggs produced per female in adults of each treatment
group The number of females in each treatment group were n=7 (control), n=4
(low), n=7 (medium), and n=8 (high).


- Control
- Low
- Medium
High


120
A
"o
N 100
1:
L 80-
0)
0) 60-
4-
S40


S20
0.
G-----


0 -1-----


Figure 3-7. Egg fertilization and hatch success. A) The percent of eggs fertilized in
each treatment group. B) The percent of fertilized eggs that hatched in each
treatment group. The number of eggs in each treatment group were n=307
(control), n=457 (low), n=645 (medium), and n=234 (high).











25



20












-
S15-
0)


o 10-



5-


Adult Body Burden
II Egg Burden














1"


Control


Low


Medium


High


Treatment Group

Figure 3-8. Maternal transfer ofp,p'-DDE to eggs.


S mean length
50 mean weight

4cd cd

40 T n


Control Low Medium High

Treatment Group


Figure 3-9. Mean length (mm) and weight (g) + S.E. of offspring at four months after
hatch (n=50 for each treatment group). Error bars represent one standard
error. Different letters above each bar indicate a significant difference
between treatment groups (p<0.05).


1.8

1.6

1.4

1.2
0)
1.0
-c
0)
0.8 oa

0.6

0.4

0.2

0.0















160 -


140 -


120 -


100 -


Male Characteristics
I First Reproduction


Control


Low


Medium


High


Treatment Group
Figure 3-10. The age when male characteristics or first reproduction was observed in
offspring from each treatment group.


Control Low Medium High

Treatment Group
Figure 3-11. Sex ratio of offspring in each treatment group (n=50 for each group).













A Males


Control
I low
I medium
I high


12-


10-

8-

+1
0" 6-
0

a, 4-

2-
2-


a a a

FT -


Control


- 0-


Low Medium High

Treatment Group


B Females


Control Low Medium

Treatment Group


Figure 3-12. GSI values of offspring in each treatment group. A) Mean male GSI. B)
female GSI. Error bars represent one standard error. Significantly different
values (p<0.05) are indicated by different letters above each bar.


25, B


E2o


) 15
0)

S0

EControl Low Medium High



Control Low Medium High


Control Low Medium High


Figure 3-13. Cumulative egg production of offspring through nine months of age. A)
Cumulative number of eggs produced from offspring. B) Cumulative number
of eggs produced per female from offspring.


LU
(5
+1
(1) 10-


a,
5-


High


c(
U)
) 500





0)
(D 400

300,
3
E
3 200
O
100

0












female GSI
female estradiol (ng / ml)
% eggs fertilized


.000 -
.00 00 00


- 100


"o
-90 N


C)
()
0)
-80 LU




- 70




- 60


Medium


Treatment Group


Figure 3-14. Values of mean female plasma estradiol, mean female GSI, and the percent
of eggs fertilized are given for each treatment group to show relationships
among the variables. Estradiol and GSI values are on the lefty-axis and the
percent of eggs fertilized is on the righty-axis.


Control


------a-----















- adult male estradiol (ng / ml)
- -v #eggs / female


0-



Control


K
K
K


Medium


- 60


- 50



- 40 D
oo

- 30 ,

()

- 20 LU



- 10


-30
-0


High


Treatment Group
Figure 3-15. Values of mean plasma estradiol and the percent of eggs per female are
given for each treatment group to show relationships among the variables.
Estradiol values are on the left y-axis and the number of eggs per female is on
the right y-axis.
















A----a-


female GSI
female estradiol (ng / ml)
% eggs fertilized


-, -, -- -00


-/s--


Low


Medium


- 100




-o


w
U)
0)
-0
- 80 U.1




- 70




S60


High


Treatment Group
Figure 3-16. Values of mean adult female plasma estradiol and mean adult GSI for fish
exposed as adults and for fish exposed in ovo are given for each treatment
group to show relationships among the variables. Estradiol and GSI values
are on the lefty-axis.


Control


I I I I















DISCUSSION

Effects of Adult Exposure to p,p'-DDE on Reproduction, 17p-Estradiol, and GSI

One of the primary goals of the dose-response experiments was to achieve a

relatively constant body burden which resembled that of wild fish from previous studies.

Fish in the low and medium p,p'-DDE groups achieved a muscle tissue concentration of

approximately 30 percent and the high group achieved approximately 20 percent of the

feed concentration ofp,p'-DDE. Largemouth bass accumulated greater than 30 percent

ofp,p'-DDE in the diet after 30 days of exposure to 3.6 [tg / g feed (Muller 2003), which

is similar to the accumulation rate of fathead minnows in this study. Thus, the doses in

the low and medium groups of the dose response experiments are environmentally

relevant, but the doses in the high group and the pilot study are probably out of the range

of environmental relevance.

In terms of population level parameters, survival probabilities were identical in

adults in the dose response experiments because there were no mortalities at any dose.

However, the number of eggs laid per female was considerably reduced in the highp,p'-

DDE group. This suggests that the no observed effect level (NOEL) under these

experimental conditions is between 10 [tg / g and 100 [tg / g in the feed. Egg production

in the high group did not differ from other groups until about day 14 ofp,p'-DDE

exposure. It was between days 7 and 14 thatp,p'-DDE muscle tissue concentration in the

high group increased rapidly from 7.26 0.46 to 8.54 6.82 [tg / g. Because there was

no discernable difference in eggs per female before day 14 in the high group and









reproduction was not affected in the medium group at any point during the study, it is

likely that the NOEL is between the day 29 p,p'-DDE muscle concentration in the

medium group (3.0 + 0.57 tg / g) and the day 14p,p'-DDE muscle tissue concentration

in the high group (8.54 6.82 tg / g).

The percent of eggs that were fertilized and that hatched were measured to isolate

which sex might be responsible for any reproductive dysfunction observed. The percent

of eggs fertilized in the high group was much lower than any other group. This type of

response have previously been observed in guppies exposed to 100 [tgp,p'-DDE / g feed,

which had a marked decrease in sperm count (Bayley et al. 2002). Alternatively,

fertilization success of male trout injected with p,p'-DDE was not altered. The percent of

fertilized eggs that hatched in the low group was also much lower than any other group,

which indicates a problem with egg rather than sperm quality. Along with data from

other studies, my results suggest there may be a problem with fertilization in the high

group, in which males had significantly elevated estradiol. Without replication, however,

these results are inconclusive.

The pilot experiment demonstrated thatp,p'-DDE impairs reproduction and does

not have the same effect as the mammalian anti-androgen, flutamide, at similar doses in

feed. Fathead minnows treated with p,p'-DDE stopped spawning after just three days of

exposure, while those treated with flutamide spawned at a relatively constant rate

throughout the pre-exposure and exposure periods. In other studies, waterborne

flutamide reduced fecundity of fathead minnows (Jensen et al. 2004). The route of

exposure is a likely factor that contributed to the difference in effect of flutamide and

p,p'-DDE on reproductive output: fish in this study were administered p,p'-DDE and









flutamide via feed and fish other studies were exposed to waterborne flutamide. The

control group seemed to have a lower spawning rate during the final ten days of the

experiment, but the rate of spawning inp,p'-DDE treated fish was zero throughout the

final 25 days of exposure. Thus, p,p'-DDE caused reproductive impairment because

exposed fish completely stopped spawning before the spawning rate of control fish

declined.

Plasma 17p-estradiol (estradiol) levels were measured in fathead minnows from the

pilot experiment to investigate ifp,p'-DDE acted by a similar mechanism as the

antiandrogen flutamide. The high dose ofp,p'-DDE in the pilot experiment (900 pg p,p'-

DDE / g feed) decreased estradiol in females, though not significantly, but flutamide

significantly increased estradiol in males, the same result of a study where fathead

minnows were exposed to waterborne flutamide (Jensen et al. 2004). p,p'-DDE did not

affect estradiol in male fish in the pilot experiment, which had estradiol values almost

identical to the control group. Despite the fact that males treated with flutamide had

plasma estradiol levels twice that of control, no effect on reproductive output was

observed. In other studies, reproductive out was adversely affected in fish treated with

waterborne flutamide (Jensen et al. 2004).

The first experiment proved that p,p'-DDE can cause reproductive dysfunction in

fathead minnows, albeit at a relatively high dose. In that experiment estradiol levels were

also considerably reduced in females treated withp,p'-DDE. At the lower and

environmentally relevant levels ofp,p'-DDE in the dose response experiment, female

estradiol levels were unaffected and male plasma estradiol levels increased as a result of

p,p'-DDE treatment. Similar to the result in males of the pilot study, trout injected with









p,p'-DDE (30 [tg / g) did not have increased vitellogenin, which is a response of

exposure to estrogens (Donohoe and Curtis 1996). Thus, p,p'-DDE was not acting as an

estrogen in that study. The dose in that study on trout was much greater than the

concentration found in the ovary in fathead minnow administered 104 [tgp,p'-DDE / g

feed in this study. The lack of an estrogenic response to a high dose ofp,p'-DDE in trout

coincides with no statistically significant changes in estradiol in the pilot experiment,

where fish also received a very high dose (900 [tg / g feed). A compensatory mechanism

may explain the difference in responses observed at very high versus environmentally

relevant doses.

Hormone receptors are common targets of endocrine disruptors (Young et al.

2005). In other studies, p,p'-DDE and flutamide had similar effects on mammalian

hormone receptor binding (Kelce et al. 1995) and on secondary sex characteristics in

guppies, a viviparous fish (Bayley et al. 2002). However, in studies on oviparous fish

cells, including other cyprinids, flutamide did not bind the AR, butp,p'-DDE did (Wells

and Van Der Kraak 2000). Additionally, when investigating AR binding in brain,

ovarian, and testicular tissues from goldfish (Curassius auratus) and rainbow trout

(Oncorhynchus mykiss), Wells and Van der Kraak (2000) found thatp,p'-DDE bound

only to AR from goldfish testes. The difference in AR binding affinities among species,

in conjunction with different changes in estradiol, suggests that p,p'-DDE and flutamide

do not act solely by similar anti-androgenic mechanisms in fatheads minnows. In this

experiment, estradiol levels in fish treated with p,p'-DDE or flutamide were not similarly

altered, which contradicts findings of these two compounds in some other species. The









data here corroborate evidence that an EDC's mode of action may differ greatly among

species.

Steroid synthetic and metabolic pathways are another possible target of the

endocrine disruption caused byp,p'-DDE in this study. Although a very high dose of

p,p'-DDE did not cause the same affect on reproduction or estradiol as flutamide, fish

responded to lower doses ofp,p'-DDE in a manner consistent with flutamide in the pilot

study: estradiol was elevated in males treated with flutamide and in males from all p,p'-

DDE groups in the dose response study. Thibaut and Porte (2004) found thatp,p'-DDE

(200 tM) increased synthesis of maturation inducing hormones (MIH), specifically

pregnenolone, in carp ovarian cells by increasing 200-HSD activity. An overall increase

in pregnenolone is likely to result in higher levels of hormones, as was seen inp,p'-DDE

-treated fish in the dose response experiment, because it is the substrate from which sex

steroid hormones are made. Largemouth bass exposed top,p'-DDE under laboratory

conditions (Muller, unpublished data) and in kelp bass (Paralabrax clathratus) from sites

contaminated with DDTs had decreased steroid hormone levels (Spies and Thomas

1997), which is the opposite response of fathead minnows in this study. Again, this

suggests species specific responses to EDCs.

The fact that there was no change in estradiol in females at environmentally

relevant doses, but there was a change in males indicates that there is a sex specific target

ofp,p'-DDE or that the target ofp,p'-DDE is more sensitive in males. Two potential

targets ofp,p'-DDE that might cause increased estradiol in males are P450aromatase and

Phase II enzymes (e.g. UDP-glucuronosyltransferase and sulfotranserfase). If

P450aromatase activity is stimulated, more testosterone would be converted to estradiol.









Alternatively, if Phase II enzymes were inhibited, estradiol would not be metabolized and

subsequently excreted (Thibaut and Porte 2004). An increase in estradiol production

would result inp,p'-DDE binding to factors that inhibit the release of GtH I, which is a

stimulant of estradiol and testosterone production (Arcand-Hoy and Benson 2001).

Plasma GtH I was decreased in fish from a site contaminated with DDTs and PCBs

(Spies and Thomas 1997), indicating that chlorinated compounds such as p,p'-DDE, may

be disrupting the endocrine system in the hypothalamus and pituitary, not just the steroid

synthetic or metabolic pathways in the gonad. Speculating the target enzyme or factor in

the HPG axis that increases estradiol in males is risky without those additional data on

other hormones and enzyme activity.

Gonadosomatic index is often used as an indicator of endocrine status in fish

(Jensen et al. 2001; Mills et al. 2001; Noaksson et al. 2003). Gonadosomatic index did

not differ significantly among treatment groups nor correlate with p,p'-DDE content or

estradiol levels. However, both males and females in the high group of the dose response

experiment showed a trend toward decreased GSI. Mills et al. (2001) found that none of

the measures of endocrine status including GSI, hepatosomatic index, estradiol,

vitellogenin, and gonadal development changed in juvenile summer flounder

(Paii i/ h/hy dentatus) exposed top,p'-DDE intravenously. Bayley et al. (2002) found

similar results in guppies, where GSI was the only androgen controlled characteristic not

affected byp,p'-DDE. Similarly, in this study GSI was not an indicator of adult exposure

to p,p'-DDE.

Effects of in ovop,p'-DDE Exposure on Survival, Development, and Reproduction

The concentrationp,p'-DDE in fish was not changing significantly during the

collection of eggs used for survival and development studies. A relatively stable level of









p,p'-DDE in adults was achieved by allowing adult fish to accumulate p,p'-DDE for at

least 17 days before collecting eggs. Egg concentration ofp,p'-DDE in the medium

group (1.40 tg / g) was comparable to total DDTs found in largemouth bass ovaries in

site contaminated with p,p'-DDE in Florida (2.82 + 0.49 tg / g; (Marburger et al. 2002).

Survival, development, and reproduction were monitored for nine months in eggs

spawned from adults exposed to p,p'-DDE. Eighty-nine percent of eggs in the control

group were successfully fertilized and 93% of those fertilized eggs hatched. The hatch

rates in this study are comparable to that of control fish in other studies, where

hatchability is greater than 90 percent (Ankley et al. 2001). These rates indicate that fry

rearing conditions were amenable to successful development.

Survival probabilities are important parameters for projecting population growth

rates (Caswell 2001). In this study, early survival of fathead minnows exposed top,p'-

DDE in ovo was not directly dependent on dose. Fish in the low group had early survival

probabilities identical to the control group. Because estradiol was not affected in

females, one would not expect a change in offspring survival caused by altered egg

quality, which is largely controlled by estrogen during oocyte development (Arcand-Hoy

and Benson 2001). However, Cheek et al. (2000) found dose-dependent survival of eggs

exposed to an estrogenic compound, which was not a result observed in the present study.

It is imperative to understand differences in sexual maturity among groups because

age at first reproduction can be an important determinant of population growth rate

(Levin et al. 1996). Fish in the low group had identical survival and sexual development

as the control group, though they were larger at four months of age. In Cheek et al.

(2000), fish exposed to the estrogen o,p'-DDT during development were also









significantly larger than control fish. However, density may have been an issue in that

study because there were fewer fish in the DDT group than the control group, which was

not normalized among treatment groups during fry rearing. The time to first reproduction

and appearance of secondary sex characteristics was delayed in fish exposed to in the

high dose (104 [tg / g) ofp,p'-DDE in ovo. Kelce et al. (1995) showed pubertal rats

exposed top,p'-DDE had increased body weight and delayed puberty. This study on

p,p'-DDE suggests that precociousness can result from low dose exposure and delayed

maturity from high dose exposure.

Not only did fish in the high group have delayed sexual maturity, but male fish

exposed in ovo also had abnormally high GSI when measured at nine months of age. To

my knowledge, there are no reports of male GSI values as high as those of females, as

observed in this study. This is especially notable given that males with drastically

increased GSI were only exposed top,p'-DDE in ovo and were allowed to develop in an

uncontaminated environment. Rats exposed to a dose of 10-20 tg p,p'-DDE / g as

developing fetuses, which may be comparable to the high dose of this study, also had

male reproductive abnormalities (Gray et al. 2001). This result implies thatp,p'-DDE

may cause early life stage changes of important factors, including epigenetic factors and

hormone homeostasis, that play a role in gonadal development. The high GSI values

measured in the high group may be the result of intersex gonads, though no male gonads

appeared to have follicles. If males were receiving estrogenic signals during

development, they may develop intersex gonads and therefore, histological analyses on

these fish should be conducted.









Reproductive output of fish exposed in ovo varied among treatment groups. When

measured as total egg output, the low group produced more eggs over nine months than

any other group. Measured as eggs per female, however, the medium group produced

more eggs than all other groups. The disparity in these two measures can be attributed to

a skewed sex ratio toward females in the low group and males in the medium group. The

high group spawned considerably fewer eggs throughout the experiment and fewer eggs

per female, a finding that cannot be attributed to a skewed sex ratio, which was close to

50:50 in the high group.

Biomarkers

A correlation matrix of average values of biomarkers and reproductive output was

produced to identify potential relationships. Biomarkers and reproductive measures that

were significantly correlated were then graphed for visual representation of the

relationship because no statistical tests could be conducted on parameters for which there

was no replication for each treatment group (i.e. number of eggs per female, percent of

eggs fertilized, etc.).

Female GSI and estradiol of adults exposed top,p'-DDE via feed were related to

one another and the percent of eggs fertilized. There was no significant relationship

between female GSI and estradiol when correlated using individual data, which is

probably due to the fact that GSI remains relatively constant but estradiol changes

throughout the spawning cycle of fathead minnows (Jensen et al. 2001). Additionally,

the percent of eggs fertilized was also directly related to adult female GSI and estradiol.

This is interesting because few studies investigate fertilization success, which was related

to GSI and estradiol in this study. GSI and estradiol are typically related to egg output,

which was not the case in this study (Jensen et al. 2001).









Adult female GSI and estradiol were also inversely related to GSI of their offspring

(i.e. eggs only exposed in ovo). Mean female GSI of adults in the high group was lower

than in all other groups, but GSI of their offspring were much higher than other groups.

Differences in GSI of both sexes in adults and their offspring were less obvious in the

low and medium groups, but the inverse relationship held.

Interestingly, adult male estradiol was also inversely related to the number of eggs

per female in adults exposed top,p'-DDE through feed. This may be due to effects of

estradiol on male courtship behavior. In guppies, Bayley et al (2002) found thatp,p'-

DDE did not affect male courtship behavior, but it did affect male secondary sex

characteristics and sperm count.

Conclusions and Future Directions

The US Food and Drug Administration's action level for issuing human

consumption advisories in fish is 5 tg p,p'-DDE / g muscle (Kennish and Ruppel 1996),

a level that may cause reproductive dysfunction in adults according to this study. NOEL

for reproductive dysfunction in adults and their offspring is observed at muscle tissue

concentrations between 3.0 + 0.57 [tg / g and 8.54 6.82 [tg / g. Muscle concentrations

as low as 0.38 0.05 [tgp,p'-DDE / g corresponded to marked changes in estradiol. In

males, increased plasma estradiol was an indicator of exposure, but did not increase with

dose. Therefore, the level of increase in estradiol was not indicative of the p,p'-DDE

concentration to which fish were exposed.

In this experiment there was no replication of each treatment level. However, the

total sample size of fish studied was similar to that of many EDC assays in fathead

minnows (Grist et al. 2003). Effects on offspring are not as conclusive as those on adults.

Effects observed in fish exposed in ovo would be more robust with replication because









differences in low and medium groups were subtle (e.g. female GSI) or without variation

(e.g. egg production).

This study has preliminarily linked biomarkers, such as adult GSI and offspring

GSI, to population parameters such as viable eggs, which can be used to develop matrix

population models. Modeling populations of treated fish would better define actual

population outcomes that may result fromp,p'-DDE treatment. The population level data

collected in this study, including survival probabilities and fertilities, should be used to

project population growth rates under each treatment and/or combinations of treatments.














APPENDIX
HORMONE DETERMINATION BY RIA

Plasma hormone concentrations of 17p-estradiol and testosterone were determined

by radioimmunoassay, as previously described (Jensen et al. 2001; Kagawa et al. 1981).

The assays were optimized for the antibodies and 3H-hormones used. The assay is

described as follows: plasma was diluted 1:5 for females and 1:4 for males with 0.1 M

PBS (pH 7.6). One-hundred fifty microliters 0.1 M PBS were added to 6 [l of diluted

plasma. The diluted plasma was extracted twice with 1.5 ml ethyl ether, the extract was

evaporated, and reconstituted in 120 [l assay buffer (0.01 M PBS with 1% BSA, pH 7.4).

One hundred microliters of reconstituted extract or 17p-estradiol standard, 100 [l

estradiol antibody (Fitzgerald Industries International, Inc.; catalogue number 20-ER06;

1:16,000 final dilution), and 100 al of 0.05 tCi/ml 3H-17p-estradiol (Amersham

Biosciences; catalogue number 125-250UCI) were added to a microcentrifuge tube and

incubated for two hours at 25 C, then placed in an ice-water bath for 15 minutes. Four-

hundred microliters of Dextran-coated charcoal solution (1.5 g activated charcoal, 0.15 g

Dextran (Sigma-Aldrich; catalogue numbers C-3345 and D-4751, respectively), and 300

ml of 0.1 M PBS, pH 7.6) were added to each tube and the tubes were returned to the ice-

water bath for another 15 minutes. The tubes were centrifuged at 3,000 rpm for 30

minutes at 4 C, 0.5 ml of the supernatant was placed into a scintillation vial to which 4.5

ml of scintillation fluid were added (Fisher Scientific, catalogue numbers SX18-4).

Tritium was counted for two minutes for each sample on a scintillation counter (Beckman

Coulter LS 6000 IC).






59


The estradiol RIA method described above could not be validated. Values

calculated were artificially high. After several troubleshooting experiments as well as

sending samples to another lab, I determined the antibody binds to multiple components

in the extract. This may be remedied by using ether with higher purity. However, I

believe the root of the problem lies with the specificity of the antibody.















REFERENCES


Ankley GT, Jensen KM, Kahl MD, Korte JJ, Makynen EA. 2001. Description and
evaluation of a short-term reproduction test with the fathead minnow (Pimephales
promelas). Environ Toxicol Chem 20(6):1276-1290.

Anway MD, Cupp AS, Uzumcu M, Skinner MK. 2005. Epigenetic transgenerational
actions of endocrine disruptors and male fertility. Science 308(5727):1466-1469.

Arcand-Hoy LD, Benson WH. 2001. Toxic responses of the reproductive system. In:
Taget Organ Toxicity in Marine and Freshwater Teleosts (Schlenk D, Baatrup E,
eds). New York:Taylor and Francis, 175-202.

Barata C, Baird D, Soares A. 2002. Demographic responses of a tropical cladoceran to
cadmium: Effects of food supply and density. Ecol Appl 12(2):552-564.

Bayley M, Junge M, Baatrup E. 2002. Exposure of juvenile guppies to three
antiandrogens causes demasculinization and a reduced sperm count in adult
males. Aquat Toxicol 56(4):227-239.

Carlson DB, Curtis LR, Williams DE. 2000. Salmonid sexual development is not
consistently altered by embryonic exposure to endocrine-active chemicals.
Environ Health Perspect 108(3):249-255.

Carr RL, Chambers JE. 2001. Toxic responses of the nervous system. In: Target Organ
Toxicity Marine and Freshwater Teleosts (Schlenk D, Benson W, eds). New
York:Taylor and Francis, 26-95.

Caswell H. 2001. Matrix Population Models. Sunderland:Sinauer Associates, Inc.

Catalan J, Ventura M, Vives I, Grimalt JO. 2004. The roles of food and water in the
bioaccumulation of organochlorine compounds in high mountain lake fish.
Environ Sci Technol 38(16):4269-4275.

Chedrese PJ, Feyles F. 2001. The diverse mechanism of action of
dichlorodiphenyldichloroethylene (DDE) and methoxychlor in ovarian cells in
vitro. Reprod Toxicol 15(6):693-698.

Cheek A, Brouwer T, Carroll S, Manning S, McLachlan J, Brouwer M. 2001.
Experimental evaluation of vitellogenin as a predictive biomarker for
reproductive disruption. Environ Health Perspect 109(7):681-690.










Collas P. 1998. Modulation of plasmid DNA methylation and expression in zebrafish
embryos. Nucleic Acids Res 26(19):4454-4461.

Cook PM, Robbins JA, Endicott DD, Lodge KB, Guiney PD, Walker MK, et al. 2003.
Effects of aryl hydrocarbon receptor-mediated early life stage toxicity on lake
trout populations in Lake Ontario during the 20th century. Environ Sci Technol
37(17):3864-3877.

Crellin NK, Rodway MR, Swan CL, Gillio-Meina C, Chedrese PJ. 1999.
Dichlorodiphenyldichloroethylene potentiates the effect of protein kinase A
pathway activators on progesterone synthesis in cultured porcine granulosa cells.
Biol Reprod 61(4):1099-1103.

Denslow ND, Chow M, Chow MM, Bonomelli LCF, Heppell SA, Sullivan CV. 1997.
Development of biomarkers for environmental contaminants affecting fish. In:
Chemically Induced Alterations in Functional Development and Reproduction of
Fishes (Rolland RM, Gilbertson M, Peterson RE, eds). Pensacola, FL:SETAC
Press, 73-86.

Denslow ND, Chow MC, Kroll KJ, Green L. 1999. Vitellogenin as a biomarker for
estrogen and estrogen mimics. Ecotoxicology 8:385-398.

Ding LD. 2005. Vitellogenesis and Vitellogenin Uptake into Oocytes. In: Hormones and
their Receptors in Fish Reproduction (Melamed P, Sherwood N, eds).
Hackensack, NJ:World Scientific Publishing Co. Pte. Ltd., 254-276.

Donohoe R, Curtis LR. 1996. Estrogenic activity of chlordecone, o,p'-DDT, o,p'-DDE in
juvenile rainbow trout: induction of vitellogenesis and interaction with hepatic
estrogen binding sites. Aquatic Toxicol 36:31-52.

Filby AL, Tyler CR. 2005. Molecular characterization of estrogen receptors 1, 2a, and 2b
and their tissue and ontogenic expression profiles in fathead minnow (Pimephales
promelas). Biol Reprod 73(4):648-662.

Foran C, Peterson B, Benson W. 2002. Influence of parental and developmental cadmium
exposure on endocrine and reproductive function in Japanese medaka (Oryzias
latipes). Comp Biochem Phys C 133(3):345-354.

Foster EP, Fitzpatrick MS, Feist GW, Schreck CB, Yates J. 2001. Gonad organochlorine
concentrations and plasma steroid levels in White sturgeon (Acipenser
transmontanus) from the Columbia River, USA. Bull Environ Contam Toxicol
67(2):239-245.









Glesleichter J, Manire CA, Szabo NJ, Cortes E, Carlson J, Lombardi-Carlson L. 2005.
Organochlorine concentrations in bonnethead sharks (Sphyrna tiburo) from Four
Florida Estuaries. Arch Environ Contam Toxicol 48(4):474-483.

Giesy J, Pierens S, Snyder E, Miles-Richardson S, Kramer V, Snyder S, et al. 2000.
Effects of 4-nonylphenol on fecundity and biomarkers of estrogenicity in fathead
minnows (Pimephales promelas). Environ Toxicol Chem 19(5):1368-1377.

Grau BL, Peterle TJ. 1979. Compartmentalization of 14C-DDT in an experimental old-
field plot. Bull Environ Contam Toxicol 22(6):846-853.

Gray LE, Jr., Ostby J, Wilson V, Lambright C, Bobseine K, Hartig P, et al. 2002.
Xenoendocrine disrupters-tiered screening and testing: filling key data gaps.
Toxicology 181-182:371-382.

Gray LE, Ostby J, Furr J, Wolf CJ, Lambright C, Parks L, et al. 2001. Effects of
environmental antiandrogens on reproductive development in experimental
animals. Human Reproduction Update 7(3):248-264.

Grist E, Wells N, Whitehouse P, Brighty G, Crane M. 2003. Estimating the effects of 17
alpha-ethinylestradiol on populations of the fathead minnow Pimephales
promelas: Are conventional toxicological endpoints adequate? Environ Sci
Technol 37(8):1609-1616.

Guillette LJ, Jr., Gross TS, Gross DA, Rooney AA, Percival HF. 1995. Gonadal
steroidogenesis in vitro from juvenile alligators obtained from contaminated or
control lakes. Environ Health Perspect 103 Suppl 4:31-36.

Guillette LJ, Jr., Gross TS, Masson GR, Matter JM, Percival HF, Woodward AR. 1994.
Developmental abnormalities of the gonad and abnormal sex hormone
concentrations in juvenile alligators from contaminated and control lakes in
Florida. Environ Health Perspect 102(8):680-688.

Halm S, Kwon JY, Rand-Weaver M, Sumpter JP, Pounds N, Hutchinson TH, et al. 2003.
Cloning and gene expression of P450 17alpha-hydroxylase,17,20-lyase cDNA in
the gonads and brain of the fathead minnow Pimephales promelas. Gen Comp
Endocrinol 130(3):256-266.

Heath R, Spann J, Kreitzer J. 1969. Marked DDE impairment of mallard reproduction in
controlled studies. Nature 224(5214):47-&.

Hornung MW, Jensen KM, Korte JJ, Kahl MD, Durhan EJ, Denny JS, et al. 2004.
Mechanistic basis for estrogenic effects in fathead minnow (Pimephales
promelas) following exposure to the androgen 17alpha-methyltestosterone:
conversion of 17alpha-methyltestosterone to 17alpha-methylestradiol. Aquatic
Toxicol 66(1):15-23.










Hu M, Chiang E, Tong S, Lai W, Hsu N, Wang L, et al. 2001. Regulation of
steroidogenesis in transgenic mice and zebrafish. Molecular and Cellular
Endocrinol 171(1-2):9-14.

Huang HJ, Liu SM, Kuo CE. 2001. Anaerobic biodegradation of DDT residues (DDT,
DDD, and DDE) in estuarine sediment. J Environ Sci Health B 36(3):273-288.

Jensen KM, Kahl MD, Makynen EA, Korte JJ, Leino RL, Butterworth BC, et al. 2004.
Characterization of responses to the antiandrogen flutamide in a short-term
reproduction assay with the fathead minnow. Aquat Toxicol 70(2):99-110.

Jensen KM, Korte JJ, Kahl MD, Pasha MS, Ankley GT. 2001. Aspects of basic
reproductive biology and endocrinology in the fathead minnow (Pimephales
promelas). Comp Biochem Physiol C 128(1):127-141.

Kagawa H, Takano K, Nagahama Y. 1981. Correlation of plasma estradiol-17 beta and
progesterone levels with ultrastructure and histochemistry of ovarian follicles in
the white-spotted char, Salvelinus leucomaenis. Cell Tissue Res 218(2):315-329.

Kalantzi OI, Alcock RE, Johnston PA, Santillo D, Stringer RL, Thomas GO, et al. 2001.
The global distribution of PCBs and organochlorine pesticides in butter. Environ
Sci Technol 35(6):1013-1018.

Kannan K, Kajiwara N, Watanabe M, Nakata H, Thomas N, Stephenson M, et al. 2004.
Profiles of polychlorinated biphenyl congeners, organochlorine pesticides, and
butyltins in southern sea otters and their prey. Environ Toxicol Chem 23(1):49-
56.

Kelce WR, Stone CR, Laws SC, Gray LE, Kemppainen JA, Wilson EM. 1995. Persistent
DDT metabolite p,p'-DDE is a potent androgen receptor antagonist. Nature
375(6532):581-585.

Kennish MJ, Ruppel BE. 1996. DDT contamination in selected estuarine and coastal
marine finfish and shellfish of New Jersey. Arch Environ Contam Toxicol
31(2):256-262.

Kime D. 1995. The effect of reproduction on fish. Rev Fish Biol Fisher 5(1):52-95.

Korte J, Kahl M, Jensen K, Pasha M, Parks L, LeBlanc G, et al. 2000. Fathead minnow
vitellogenin: Complementary DNA sequence and messenger RNA and protein
expression after 17 beta-estradiol treatment. Environ Toxicol Chem 19(4):972-
981.

LeBlanc GA, Bain LJ, Wilson VS. 1997. Pesticides: multiple mechanisms of
demasculinization. Mol Cell Endocrinol 126(1):1-5.










Levin L, Caswell H, Bridges T, DiBacco C, Cabrera D, Plaia G. 1996. Demographic
responses of estuarine polychaetes to pollutants: Life table response experiments.
Ecol Appl 6(4):1295-1313.

Loomis AK, Thomas P. 2000. Effects of estrogens and xenoestrogens on androgen
production by Atlantic croaker testes in vitro: evidence for a nongenomic action
mediated by an estrogen membrane receptor. Biol Reprod 62(4):995-1004.

Macek K. 1968. Reproduction in brook trout (Salvelinusfontinalis) fed sublethal
concentrations of DDT. J Fish Res Board Can 25(9):1787-&.

Makynen EA, Kahl MD, Jensen KM, Tietge JE, Wells KL, Van Der Kraak G, et al. 2000.
Effects of the mammalian antiandrogen vinclozolin on development and
reproduction of the fathead minnow (Pimephalespromelas). Aquatic Toxicol
48(4):461-475.

Marburger J, Johnson W, Gross T, Douglas D, Di J. 2002. Residual organochlorine
pesticides in soils and fish from wetland restoration areas in central Florida, USA.
Wetlands 22(4):705-711.

Matsui M, Kinuyama Y, Hakozaki M. 1974. Biliary metabolites of testosterone and
testosterone glucosiduronate in the rat. Steroids 24(4):557-573.

Matsumura F. 1985. Toxicology of Insecticides. New York:Plenum Press.

Mauri M, Baraldi E, Simonini R. 2003. Effects of zinc exposure on the polychaete
Dinophilus gyrociliatus: a life-table response experiment. Aquatic Toxicol
65(1):93-100.

Mayer FL, Versteeg DJ, McKee MJ, Folmar LC, Graney RL, McCume DC, et al. 1992.
Physiological and nonspecific biomarkers. In: Biomarkers: Biochemical,
Physiological, and Histological Markers of Anothropogenic Stress (Huggett RL,
Kimble RA, Mehrle PM, Bergman HL, eds). Boca Raton:Lewis, 5-86.

Miglioranza KS, Aizpun de Moreno JE, Moreno VJ. 2003. Dynamics of organochlorine
pesticides in soils from a southeastern region of Argentina. Environ Toxicol
Chem 22(4):712-717.

Miller DH, Ankley GT. 2004. Modeling impacts on populations: fathead minnow
(Pimephalespromelas) exposure to the endocrine disruptor 17beta-trenbolone as
a case study. Ecotoxicol Environ Saf 59(1):1-9.

Mills L, Chichester C. 2005. Review of evidence: Are endocrine-disrupting chemicals in
the aquatic environment impacting fish populations? Sci Total Environ 343(1-
3):1-34.









Mills LJ, Gutjahr-Gobell RE, Haebler RA, Horowitz DJ, Jayaraman S, Pruell RJ, et al.
2001. Effects of estrogenic (o,p'-DDT; octylphenol) and anti-androgenic (p,p'-
DDE) chemicals on indicators of endocrine status in juvenile male summer
flounder (Paralichthys dentatus). Aquat Toxicol 52(2):157-176.

Moffat AC, ed. 1986. Clarke's isolation and identification of durgs in pharmaceuticals,
body fluids, and post-mortem material. London:The Pharmeceutical Press.

Monnosson E, Kelce WR, Mac M, Gray LE. 1997. Environmental antiandrogens:
potential effects on fish reproduction and development. In: Chemically Induced
Alterations in Functional Development and Reproduction of Fishes (Rolland RM,
Gilbertson M, Peterson RE, eds). Pensacola, FL:SETAC Press, 53-60.

Muller JK. 2003. An evaluation of dosing methods and effects ofp,p'-DDE and dieldrin
in Florida largemouth bass (Micropterus salmiodesfloridanus). Gainesville,
Fla.:University of Florida. Available: http://purl.fcla.edu/fcla/etd/UFE0001107.

Nagahama Y, Yoshikuni M, Yamashita M, Tanaka M. 1994. Molecular endocrinology of
fish. In: Fish Physiology (Sherwood N, Hew CL, eds). San Diego:Academic
Press.

Nash JP, Kime DE, Van der Ven LT, Wester PW, Brion F, Maack G, et al. 2004. Long-
term exposure to environmental concentrations of the pharmaceutical
ethynylestradiol causes reproductive failure in fish. Environ Health Perspect
112(17):1725-1733.

Nimrod AC, Benson WH. 1997. Assessment of estrogenic activity in fish. In: Chemically
Induced Alterations in Functional Development and Reproduction of Fishes
(Rolland RM, Gilbertson M, Peterson RE, eds). Pensacola, FL:SETAC Press.

Noaksson E, Linderoth M, Bosveld AT, Balk L. 2003. Altered steroid metabolism in
several teleost species exposed to endocrine disrupting substances in refuse dump
leachate. Gen Comp Endocrinol 134(3):273-284.

Orlando EF, Denslow ND, Folmar LC, Guillette LJ, Jr. 1999. A comparison of the
reproductive physiology of largemouth bass, Micropterus salmoides, collected
from the Escambia and Blackwater Rivers in Florida. Environ Health Perspect
107(3):199-204.

Panter GH, Thompson RS, Sumpter JP. 1998. Adverse reproductive effects in male
fathead minnows exposed to environmentally relevant concentrations of the
natural oestrogens, oestradiol, and oestrone. Aquatic Toxicol 42:243-253.

Raimondo S, McKenney CL, Jr. 2005. Projected population-level effects of thiobencarb
exposure on the mysid, Americamysis bahia, and extinction probability in a
concentration-decay exposure system. Environ Toxicol Chem 24(3):564-572.









Rolland RM, Gilbertson M, Peterson RE, eds. 1997. Chemically Induced Alterations in
Functional Develoment and Reproduction of Fishes. Racine: SETAC Press.

Segner H. 2005. Developmental, reproductive, and demographic alterations in aquatic
wildlife: Establishing causality between exposure to endocrine-active compounds
(EACs) and Effects. Acta Hydrochimica Hydrobiologia 33(1): 17-26.

Sepulveda M, Johnson W, Higman J, Denslow N, Schoeb T, Gross T. 2002. An
evaluation of biomarkers of reproductive function and potential contaminant
effects in Florida largemouth bass (Micropterus salmoidesfloridanus) sampled
from the St. Johns River. Sci Total Environ 289(1-3):133-144.

Sonderfan AJ, Arlotto MP, Parkinson A. 1989. Identification of the cytochrome P-450
isozymes responsible for testosterone oxidation in rat lung, kidney, and testis:
evidence that cytochrome P-450a (P450IIA1) is the physiologically important
testosterone 7 alpha-hydroxylase in rat testis. Endocrinology 125(2):857-866.

Spies RB, Thomas P. 1997. Reproductive and endocrine status of female kelp bass from a
contaminated site in the Southern California Bight and estrogen receptor binding
of DDTs. In: Chemically Induced Alterations in Functional Development and
Reproduction of Fishes (Rolland RM, Gilbertson M, Peterson RE, eds).
Pensacola, FL:SETAC Press, 113-133.

Thibaut R, Porte C. 2004. Effects of endocrine disrupters on sex steroid synthesis and
metabolism pathways in fish. J Steroid Biochem Mol Biol 92(5):485-494.

Thompson CJ, Ross SM, Gaido KW. 2004. Di(n-butyl) phthalate impairs cholesterol
transport and steroidogenesis in the fetal rat testis through a rapid and reversible
mechanism. Endocrinol 145(3):1227-1237.

Valters K, Olsson A, Asplund L, Bergman A. 1999. Polychlorinated biphenyls and some
pesticides in perch (Percafluviatilis) from inland waters of Latvia. Chemosphere
38(9):2053-2064.

Van der Kraak G, Chang JP, Janz DM. 1998. Reproduction. In: The Physiology of Fishes
(Evans DH, ed). Boca Raton, FL:CRC Press, LLC, 465-490.

von Hofsten J, Olsson PE. 2005. Zebrafish sex determination and differentiation:
involvement ofFTZ-F1 genes. Reprod Biol Endocrinol 3:63.

Waxman DJ. 1988. Interactions of hepatic cytochromes P-450 with steroid hormones.
Regioselectivity and stereospecificity of steroid metabolism and hormonal
regulation of rat P-450 enzyme expression. Biochem Pharmacol 37(1):71-84.






67


Wells K, Van Der Kraak G. 2000. Differential binding of endogenous steroids and
chemicals to androgen receptors in rainbow trout and goldfish. Environ Toxicol
Chem 19(8):2059-2065.

Wilson VS, Cardon MC, Thornton J, Korte JJ, Ankley GT, Welch J, et al. 2004. Cloning
and in vitro expression and characterization of the androgen receptor and isolation
of estrogen receptor alpha from the fathead minnow (Pimephalespromelas).
Environ Sci Technol 38(23):6314-6321.

Young G, Kusakabe M, Nakamura I, Lokman PM, Goetz FW. 2005. Gonadal
Steroidogenesis in Teleost Fish. In: Hormones and Their Receptors in Fish
Reproduction (Melamed P, Sherwood N, eds). Hackensack, NJ:World Scientific
Publishing Co. Pte. Ltd., 155-223.















BIOGRAPHICAL SKETCH

Liza J. Ray was born to a farmer and swim instructor in Southern California where

she lived until attending the University of Michigan. While at Michigan, she studied and

loved the wetland environments of that region. After being cold for four years, she was

given the opportunity to enter graduate school in the Interdisciplinary Ecology program at

the University of Florida School of Natural Resources and Environment. She is finally

starting to overcome her fear of alligators (Alligator mississippiensis).

PAGE 1

EFFECTS OF p,p Â’-DDE ON REPRODUCTION AND BIOMARKERS OF ENDOCRINE DISRUPTION IN FATHEAD MINNOWS ( Pimephales promelas ) By ELIZABETH JORDAN RAY A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2006

PAGE 2

Copyright 2006 by Elizabeth Jordan Ray

PAGE 3

This work is dedicated to the loving and cheerful spirit of my grandpa, Owen.

PAGE 4

iv ACKNOWLEDGMENTS The completion of this project would not have been possibl e without input of several people. First, I thank my adviso r, Dr. David Barber, for his direction and availability, without which I would not ha ve had the opportunity to learn and do the variety of skills required for the successful completion of this project. I also thank my committee members, Dr. Nancy Denslow and Dr. Madan Oli, for all their input and advice. My family and friends were an e ssential source of support throughout the course of these years. There are also several la b members and co-workers at the Center for Environmental and Human Toxicology and the Aquatic Toxicology Facility that I thank: Kathy Childress and Kevin Kroll for thei r advice on fish care and handling, Greg Robbins for assisting with fish care, Kathleen Jensen at the US EPA in Duluth for helping with RIA validation, Scott Wasdo and Nancy S zabo for their chemical advice, Joe Griffitt for help with gene analysis, and fellow la b members Alex McNally and Roxana Weil for all the little things.

PAGE 5

v TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES............................................................................................................vii LIST OF FIGURES.........................................................................................................viii INTRODUCTION...............................................................................................................1 Endocrine Disruption in Teleosts.................................................................................2 Biomarkers of Endocrine Disruption............................................................................7 p,p Â’-DDE in the Environment.....................................................................................12 p,p Â’-DDE: An Endocrine Disruptor............................................................................14 Linking Biomarkers of Endocrine Disruption to Fish Populations............................17 MATERIALS AND METHODS.......................................................................................21 General Methods.........................................................................................................21 Fish Holding Conditions......................................................................................21 Reproductive Measures.......................................................................................22 Plasma and Tissue Collection..............................................................................22 Determination of p,p Â’-DDE Content...................................................................23 Determination of Plasma 17 -Estradiol..............................................................24 Vitellogenin mRNA Quantification....................................................................25 Experimental Set-Up..................................................................................................26 Pilot Experiment..................................................................................................26 p,p Â’-DDE Dose-Response Experiment I: Accumulation Rate and Reproductive Output of Adults........................................................................27 Survival, Development, and Reproductive Output of Offspring.........................28 p,p Â’-DDE Dose-Response Experiment II: Co llection of Biologi cal Materials...29 Statistical Analyses.....................................................................................................29 RESULTS........................................................................................................................ ..31 Pilot Experiment.........................................................................................................31 p,p Â’-DDE Dose-Response Experi ments I & II: Adult 17 -Estradiol, GSI, and Reproduction..........................................................................................................32 Effects of in ovo Exposure on Survival, Development, and Reproduction................34 Identification of Biomarkers.......................................................................................36

PAGE 6

vi DISCUSSION....................................................................................................................4 7 Effects of Adult Exposure to p,p Â’-DDE on Reproduction, 17 -Estradiol, and GSI...47 Effects of in ovo p,p Â’-DDE Exposure on Survival, Development, and Reproduction..........................................................................................................52 Biomarkers..................................................................................................................55 Conclusions and Future Directions.............................................................................56 APPENDIX....................................................................................................................... .58 HORMONE DETERMINATION BY RIA................................................................58 REFERENCES..................................................................................................................60 BIOGRAPHICAL SKETCH.............................................................................................68

PAGE 7

vii LIST OF TABLES Tables page 1-1. Classes of endocrine disr upting compounds and examples.......................................20 2-1. Primer sequences used for quantitative real-time PCR.............................................30 3-1. Survival probabilities for o ffspring spawned from adults fed p,p Â’-DDE..................37

PAGE 8

viii LIST OF FIGURES Figures page 1-1. Schematic of basic signals within th e hypothalamus-pituitary-gonad (HPG) axis.....3 1-2. Active hormones found in fish...................................................................................5 1-3. DDT and selected metabolites..................................................................................13 3-1. The number of eggs produced per fe male before and after exposure to either p,p Â’-DDE of flutamide as compared to control........................................................37 3-2. Comparison of plasma 17 -estradiol levels in males and females within and among each treatment group....................................................................................38 3-3. Mean p,p Â’-DDE muscle tissue concentration ( g p,p Â’-DDE / g wet weight muscle tissue)...........................................................................................................38 3-4. Comparison of plasma 17 -estradiol levels in males and females from each treatment group........................................................................................................39 3-5. Mean GSI values of adults treated with p,p Â’-DDE-contaminated feed....................39 3-6. Cumulative number of eggs produced per female in adults of each treatment group.........................................................................................................................4 0 3-7. Egg fertilization and hatch success...........................................................................40 3-8. Maternal transfer of p,p Â’-DDE to eggs.....................................................................41 3-9. Mean length (mm) and we ight (g) S.E. of offspring at four months after hatch...41 3-10. The age when male characteris tics or first reproduction was observed in offspring from each treatment group........................................................................42 3-11. Sex ratio of offspring in each treatment group.......................................................42 3-12. GSI values of offspr ing in each treatment group....................................................43 3-13. Cumulative egg pr oduction of offspring through nine months of age...................43

PAGE 9

ix 3-14. Values of mean female plasma estrad iol, mean female GSI, and the percent of eggs fertilized are given for each treatment group...................................................44 3-15. Values of mean plasma estradiol and the percent of eggs per female are given for each treatment group..........................................................................................45 3-16. Values of mean adult female plasma estradiol and mean adult GSI for fish exposed as adults and for fish exposed in ovo are given for each treatment group.........................................................................................................................4 6

PAGE 10

Abstract of Thesis Presented to the Gra duate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science EFFECTS OF p,p Â’-DDE ON REPRODUCTION AND BIOMARKERS OF ENDOCRINE DISRUPTION IN FATHEAD MINNOWS ( Pimephales promelas ) By Elizabeth Jordan Ray August 2006 Chair: David Barber Major Department: Interdisciplinary Ecology The phenomenon of endocrine disruption, which includes impaired reproduction and survival, has been widely researched ove r the past decade. However, few studies have linked biomarkers of endocrine disrupt ion to population-level outcomes. The most stable metabolite of the orga nochlorine pesticide DDT is p,p Â’-DDE, a common environmental contaminant. The objective of this experiment was to investigate the effects of p,p Â’-DDE on biomarkers of endocrine disrupt ion as they related to survival and reproduction of fathead minnows ( Pimephales promelas ). Fish were exposed to 1.63, 11.48, 104.25, or 900 g p,p Â’-DDE / g feed or 1208 g flutamide / g feed in three separate experiments. Reproductive output, 17 -estradiol, and gonadosomatic index (GSI) were measured in adults exposed to p,p Â’-DDE via feed. Survival, development, GSI and reproductive output were measured to nine months of age in eggs from those adults to determine second-generation effects of p,p Â’-DDE exposure. Reproductive output was impaired in fish exposed to 104.25 g p,p Â’-DDE / g feed as well as their

PAGE 11

offspring, but not in fish exposed to flutamide. 17 -estradiol as measured by enzymelinked immunosorbent assay ( ELISA) was elevated in males exposed to flutamide and 1.63, 11.48, and 104.25 g p,p Â’-DDE / g feed. p,p Â’-DDE and flutamide do not act by the same mechanism at high doses (900 g p,p Â’-DDE and 1208 g flutamide / g feed). Neither GSI nor 17 -estradiol levels we re correlated to p,p Â’-DDE concentration in feed or fish muscle tissue. The percen t of eggs fertilized, GSI and 17 -estradiol measured from adults exposed to p,p Â’-DDE through feed were directly related to each other and inversely related to GSI of offspring. Male GSI of offspring of a dults exposed to 104.25 g / g feed was significantly higher than any other group. GSI in adult males was inversely related to the number of eggs spaw ned per female. These data have important implications for the effects of in ovo exposure to endocrine disrupting compounds. Further, data such as these can be used to model the population growth rates and relate them to biomarkers, such as GSI, of exposure to endocrine disrupting compound.

PAGE 12

1 INTRODUCTION Scientific and popular liter ature have reported on th e effects of endocrine disrupting compounds (EDCs) in fish for d ecades. Endocrine disrupting compounds are those that affect the normal functioning of th at system, typically resulting in adverse effects on an organism, its progeny, and/or a population. Endocrine disruption has caused widespread public concern regarding th e sustainability of fi sh populations and sparked copious scientific research in recen t years. Laboratory and field studies have shown reproductive dysfunction in animals and humans exposed to a variety of EDCs (Cook et al. 2003; Gray et al. 2001; LeBlanc et al. 1997; Noak sson et al. 2003). Specific groups of chemicals have been investigated for endocrine-disrupti ng effects, including pesticides, therapeutic hormones administered to humans and mamm als, and byproducts of industrial processes. Mo re specific types of thes e groups of compounds include organochlorine and organophosphate pesticides components of oral contraceptives, plasticizers, fire retardants, and jet-fuel residues (Table 1-1). The reproductive cycle of fish is cont rolled by the endocrine system. The endocrine system is regulated by the hypotha lamus, pituitary, and gonad, collectively known as the HPG axis. By definition, EDCs alter the normal functioning of the HPG axis. Disruptions of the HPG axis are most conspicuously manifested in gonads, which unlike the hypothalamus and pituitary, undergo visible stages of reproductive development in fathead minnows. Because the gonad is responsible for reproductive output, it is the primary link between an al tered endocrine system and demonstrated

PAGE 13

2 reproductive dysfunction. Thus, gonadal functio n is an easily measured and relevant marker of exposure to an EDC. Producti on of the female-specific egg yolk precursor protein, vitellogenin, is also an important measure of endoc rine disruption in males. Measures of gonad health and vitellogenin ar e commonly used biomarkers of endocrine disruption. Relationships between biomarkers of endocrine disruption measured in an individual fish and population-level ou tcomes are currently poorly understood. Endocrine Disruption in Teleosts The teleostean endocrine system involve s several organs and myriad molecular signaling pathways. To understand endocrine disruption, one must first understand the normal endocrine system. Environmental cu es, such as photoperiod, temperature and presence of other fish, trigge r a cascade of signals that pr epare teleosts for reproduction. In teleosts, steroid hormones are typically considered the ultimate molecular factors influencing reproductive development, ma turity, and release of gametes. The steroidogenic process is controll ed by the HPG axis, beginning in the hypothalamus when environmental cues stimul ate the release of gonadotropin-releasing hormone (GnRH; Fig. 1-1). Gonadotropin-rel easing hormone begins a cascade of signals to stimulate reproductive preparedness. The re lease of GnRH stimulat es the pituitary to secrete two types of gonadotropins, GtH I a nd GtH II that act on st eroidogenic tissues (Arcand-Hoy and Benson 2001; Hu et al. 2001). GtH I and GtH II are the teleostean analogs of the mammalian follicle-stimulating hormone (FSH) and luteinizing hormone (LH), respectively (Arcand-Hoy and Bens on 2001). The steroidogenic process is triggered when gonadotropins (GtH I and Gt H II) reach the gonads and attach to hormone-responsive cellular receptors in th e cell membrane (Young et al. 2005). While most research has focused on GtH I and Gt H II., other hormones and factors may also

PAGE 14

3 play a role in steroidogenesis in the gonad, but their potency and mechanism of action are not well-described and hence, they are not a ddressed here (Van der Kraak et al. 1998). Gonad17 -Estradiol Testosterone 11-Ketotestosterone Egg Hypothalamus PituitaryEnvironmental SignalsGonadotropin-releasing hormone (GnRH) GTH I & GTH II Liver Vitellogenin Day length Temperature Presence of other fish inhibit Figure 1-1. Schematic of basic signals w ithin the hypothalamus-pituitary-gonad (HPG) axis. Each gonadotropin plays a specific role in the process of reproduction, which includes oocyte development and ovulation. GtH I is primarily responsible for oocyte development in females and spermiation in males. In females, GtH I binds to specific receptors on the follicle, stimulating te stosterone production and its subsequent aromatization to estradiol. Estradiol then binds to the estrogen receptor in endocrine active tissues, signaling a cascade of events that contribute to oocyte development, including production of vitell ogenin in the liver (Ding 2005) As oocytes and sperm develop, the expression of GtH II increases relative to GtH I (V an der Kraak et al. 1998). GtH II is primarily responsible for stimu lating the production of progesterone, also known as maturation-inducing hormone (MIH), wh ich is believed to be responsible for

PAGE 15

4 the final oocyte maturation and ovulation. Se veral hormones shown to have MIH effects include progestens, cortisol, and deoxycor ticosterone (Nagahama et al. 1994). The outcomes produced by each gonadotropin, such as oocyte development and ovulation, are a result of the amount of t ype of steroid produced. Production of specific steroids is ultima tely triggered by an increase in cAMP, which occurs when gonadotropins attach to hormone-responsive receptors in cell membranes of steroidogenic tissues (Hu et al. 2001). cAMP then binds to response elements in the promoter regions of genes coding for factors involve d in steroidogenesis, thereby increasing the expression of those factor s. Cholesterol is the initial substrate for all steroid hormones, including estrogens, pr ogesterones, and androgens. The first and rate-limiting step of this process begins w ith production of steroi dogenic acute regulatory protein (StAR), which facilitates the transfer of cholesterol from the cytoplasm into mitochondria (Young et al. 2005). In mitoc hondria P450 side-chain cleavage enzyme (P450scc) catalyses the conversi on of cholesterol into preg nenolone. Pregnenalone can then be converted to a numb er of steroid hormones through several different pathways. Progesterones are formed from pregnenol one in reactions catalyzed by P450c17, 3 hydroxysteroid dehydrogenase (3 -HSD), and 20 -hydorysteroid dehydrogenase (20 HSD). Androstenedione and androstenediol which are formed from pregnenolone and progesterones, are the substrates converted to testosterone. Testosterone can then be converted to 17 -estradiol or 11-ketotestoste rone by P450 aromatase or P45011 and 11 -HSD, respectively (Thibaut and Port e 2004; Young et al. 2005). In fish,

PAGE 16

5 17 -estradiol is the active estrogen, but there is uncertainty regarding the roles of testosterone or 11-ketotest osterone in androgenic ac tivity (Kime 1995). Though the process of steroidogenesis occurs only in spec ific tissues, they may have effects on other systems. Figure 1-2. Active hormones found in fish. Once steroids are produced, they can remain in steroidogenic tissue or travel to other organs where they undergo metabolism, cause feedback inhibition, or regulate a variety of responses including, vitelloge nesis, development of secondary sex characteristics, or reprod uction itself (Arcand-Hoy and Benson 2001). In a negative feedback loop, steroids target the hypothalamu s and pituitary and inhi bit further signaling of steroid production via GnRH or gonadotropins, respectiv ely (Arcand-Hoy and Benson 2001; Young et al. 2005). Steroids are metaboliz ed primarily in the liver, but can also be metabolized in other tissues (Sonderfan et al 1989). Phase I enzymes metabolize steroid hormones by hydroxylation and dehydrogenation and Phase II enzymes, such as UDPglucuronosyltransferase and su lfotransferase, are responsib le for steroid metabolism by conjugation (Matsui et al. 1974; Waxman 1988). When steroids reach target organs they bind to hormone receptors. There are several steroid hormone receptors that regulate expression of a suite of target genes via genomic interaction by binding to promoters in those target genes. Recent evidence OH O Testosterone OH O H 17--Estradiol OH O O 11-keto-Testosterone

PAGE 17

6 suggests that effects elicite d by hormones may also aris e via a non-genomic pathway (Loomis and Thomas 2000). Most hormonal responses, however, are believed to be mediated through receptors and genomic interaction, which is a common target for EDCs (Filby and Tyler 2005). There are at least thr ee forms of estrogen receptors in fish (ER ER ER ) to which estradiol binds (Filby and Tyle r 2005). In female teleosts, binding of estradiol to ERs stimulates transcription and production of vitellogenin, an egg yolk protein precursor important in egg deve lopment (Young et al. 2005). There are two likely forms of androgen receptor in teleosts which regulate androgen-controlled genes (AR and AR ; Wilson et al. 2004). As hormones bi nd to these receptors they regulate expression of target genes through genomic interaction. Endocrine disruption can occur when any pa rt of the complex system of signals within the HPG axis is altered. While di sruption may occur at numerous targets, many studies show EDCs act by interacting with hormone receptors (Kelce et al. 1995; Wilson et al. 2004). A significant de viation of plasma hormone levels from normal is often an indicator of endocrine disruption. Additiona lly, because males do not produce eggs, they do not produce significant amounts of vitellogenin. Thus, an induction of vitellogenin in male fish signifies endocrine disruption a nd, more specifically, exposure to estrogenic compounds (Denslow et al. 1997; Nash et al. 2004). Estrogen and androgen receptors are also likely targets of EDCs, where EDCs bind the receptor (Chedrese and Feyles 2001; Kelce et al. 1995; Young et al. 2005). However, studies sugg est that receptorsÂ’ binding affinities may differ among species, which has important implications for species to species extrapolations (Wilson et al. 2004).

PAGE 18

7 Endocrine disrupting compounds may al so affect offspr ing during oocyte development or embryonic development. A dverse effects to offspring may occur if estradiol and/or proge sterone are altered during the repr oductive cycle because of their critical influence on egg yol k protein and timing of rel ease of oocytes (Nimrod and Benson 1997). Additionally, embryonic exposure of fish to EDCs may influence normal sexual development and growth (Arca nd-Hoy and Benson 2001). Fathead minnows undergo a series of physiological changes as they develop into reproductive male or female adults. Those changes, which include phenotypic sex determination and development of secondary sex characteristics, are largely controlled by levels of steroid hormones (von Hofsten and Olsson 2005). An alteration in those steroid hormones may lead to developmental dysfunction. There is also mounting evidence that exposur e of an adult to an EDC can affect its offspring through epigenetic mechanisms (Anway et al. 2005; Collas 1998). Recently, studies have been published on the effects of exposure to EDCs during egg development, or in ovo A multiple-generation study of fi sh exposed to enthynylestradiol found reproductive dysfunction in offspring exposed in ovo (Nash et al. 2004). That study showed no change in fertilities of adults expos ed, but a reduction in fertility of the second generation, even after depuration. The c onnection between abnormal steroid hormone and vitellogenin levels in adults and altered reproductive output of their offspring is not well understood. Thus, this study focused on biomarkers of p,p Â’-DDE exposure in adults as they relate to effects in reproductive output of those adults and their offspring. Biomarkers of Endocrine Disruption Biomarkers are defined as biological res ponses that deviate from normal as a result of exposure to a given stimulus (Mayer et al 1992). These responses can be measured at

PAGE 19

8 different levels of biologi cal organization and include changes in gene expression, hormone concentrations, reproductive output and from a broader perspective, populations (Korte et al. 2000; Sepulveda et al. 2002). There are two groups of biomarkers: those of exposure and those of eff ect. Biomarkers of exposure are those that simply indicate that an organism has been e xposed to an EDC at some level. Biomarkers of effect are those that indicate a degree of exposure with an EDC sufficient to result in an impact on a higher level of biological organization. The distinction between biomarkers of effect and biomarkers of e xposure is often dependent on the endpoint of interest. Much effort has been placed towards developi ng biomarkers of exposure and effect of EDCs in fish by correlating expos ure to contaminants with altered gene expressions, hormone concentrations, and repr oductive output (Ankley et al. 2001; Foran et al. 2002; Giesy et al. 2000). Connections among these fact ors tend to vary, but some are conserved across species and compounds. Populationand organism-lev el biomarkers of exposure to EDCs include skewed sex ratio, gonadosomatic index (GSI), and age at first reproduction. A skewed sex ratio is often used as an indicator of a wild population exposed to EDCs. Physiological biomarkers such as GSI, inappropriate pres ence of intersex gonads or secondary sex characteristics, and age at sexual maturity require some measure of individual fish (Ankley et al. 2001; Monnosson et al. 1997; Sepulveda et al. 2002). Reduced GSI has been found in fish exposed to estrogenic and anti-androgenic contaminants in the laboratory and in fish inhabiting contaminated sites (Ankley et al 2001; Panter et al. 1998; Sepulveda et al. 2002). GSI, how ever, can be unaffected by exposure to compounds that exhibit other endocrine disr upting properties (Bayley et al. 2002). Few

PAGE 20

9 studies of fathead minnows have observed in tersex gonads, which is the presence of characteristics of both ovaries and testes in the gonad (Mills and Chichester 2005). Age at first maturity is not a common biomarker for EDC exposure, which is probably due to the time and difficulty involved in making such a measurement. Nonetheless, studies on population dynamics have shown that time of first reproduction can play an important role in population growth rate (Levin et al. 1996). Molecular and genetic biomarkers that in dicate exposure to contaminants include DNA adducts, steroid hormones, vitellogenin in duction in males, tumors, and thinned eggshells (Denslow et al. 1997). Genetic biomarkers for reproductive dysfunction in fathead minnows include abnormal mRNA levels for ER AR, vitellogenin, the steroidogenic enzyme cytochrome P450 17 –hydroxylase,17,20,lyase (P450c17), and P450scc or aromatase (Denslow et al. 1997; Halm et al. 2003; Rolland et al. 1997; Wilson et al. 2004). While mRNA expression is a useful biomarker th at also elucidates possible mechanisms of EDC action, abnormal steroid hormone and vitellogenin protein levels are more common biomarkers of EDC exposure. In fish, the primary focus of molecular biomarkers of exposure and e ffect has been on plasma vitellogenin, 17 estradiol, testosterone, and 11ketotestosterone (Giesy et al 2000; Mills and Chichester 2005; Sepulveda et al. 2002). Steroid hormone levels as well as the ratio of estroge ns to androgens are also common biomarkers of exposure and effect. Appropriate levels of steroid hormones are used as a biomarker because they can be indicative of adverse affects on normal reproduction. Identification of abnormal hor mone levels first requires knowledge of normal hormone levels during the reproductive cycle of each sex of a species of fish.

PAGE 21

10 Jensen et al. (2001) descri bed the basic reproductive bi ology of the fathead minnow ( Pimephales promelas ). Average plasma 17 estradiol and testosterone concentrations of females were 5.97 1.12 and 3.08 0.34 ng / ml and of males were 0.40 0.13 and 9.11 0.92 ng / ml, respectively. Investigations in several species of fish, including fathead minnows and largemouth bass, exposed to EDCs relate altered ster oid hormone levels with other reproductive endpoints such as GSI and egg output. For example, Giesy et al. (2000) found a significant positive correlation between plasma estr adiol and the number of eggs produced per female fathead mi nnow exposed to 4-nonylphenol. However, Makynen et al. (2000) found a reduction in female GSI, but no change in plasma steroid hormone levels in fathead minnows exposed to vinclozolin. Additionally, largemouth bass inhabiting contaminated lakes in Florida had both hormonal and reproductive abnormalities (Guillette et al 1994; Sepulveda et al. 2002). Steroid hormones can be sensitive biomarkers of exposure to EDCs regardless of their connection to other reproductive endpoints. Changes in steroid hormone levels can further affect production of vitellogenin, which is initiated when estradiol bind to an ER. Because vitellogenin production is activated by the ERs, its inducti on in males is often used as a biomarker of exposure to estrogenic compounds (Denslow et al. 1999; Korte et al. 2000). Expression of vitellogenin is sensitive to estrogenic action, but the response in mRNA levels is not as persistent as the protein itself. This is because male fish do not have a mechanism for clearing vitellogenin from the body (Korte et al. 2000). Although v itellogenin expression in males is indicative of exposure to es trogenic compounds, it is not a consistently reliable marker of reproductive dysfunction. A study in fathead minnows exposed to the

PAGE 22

11 estrogenic compound 4-nonylphenol, found that correlations between vitellogenin and estradiol differed when the study was repeated by the same investigator (Giesy et al. 2000). Additionally, plasma estradiol, but not vitellogenin was related to egg production (Giesy et al. 2000). Further, studies show vitellogenin expr ession levels in females are not necessarily correlated to hatching succe ss of eggs from adults exposed to an estrogenic compound during development (Cheek et al. 2001). Thus, transcript level of vitellogenin is a good biomarker of recent exposure and vitellogenin protein level is a better biomarker of exposure history to estr ogenic compounds, but neither is necessarily indicative of altere d reproductive capacity. The consequences of reproductive malfunctions caused by EDCs can be detrimental to fish populations, which pe ople depend upon for food and recreation (Cook et al. 2003). Thus, it is important to understand how fish will respond to endocrine disruptors to maintain healthy fish populations, especially in restoration (B ayley et al. 2002) sites. The present study measured a suite of biomarkers in fathead minnows exposed to p,p Â’-DDE with the goal of providing a mo re comprehensive understanding of the connections among them in the context of population level outcomes. The fathead minnow is a small member of the minnow family, Cyprinidae, that is easily raised in the laborator y and commonly used in toxic ity assays. This species reaches sexual maturity at 4-5 months and liv es up to 4 years in the wild, where it feeds primarily on invertebrates. Spawning activity can be induced by environmental conditions such as photoperiod and water temperature. In a study of the basic reproductive biology of fathead minnows, fema les spawned approximately 85 eggs every four days (Jensen et al. 2001) Steroid hormones and GSI va ries at each point in the

PAGE 23

12 spawning cycle of females, but not males (Jensen et al. 2001). Vitellogenin is occasionally detected in male fish not exposed to EDCs and is at a relatively constant level throughout the spawning cy cle in females (Jensen et al 2001). The fathead minnow is a good model for endocrine disruption studie s because it is easy to work with in a laboratory setting and there is considerable information on its basic reproductive biology. p,p Â’-DDE in the Environment Organochlorine pesticides (OCPs) are banne d in most industriali zed countries, but their use continues in developing countries because they are relatively inexpensive, effective, and easily manufactured. OCPs include compounds commonly known as DDT, toxaphene, chlordane, methoxychlor, vi nclozolin, and dieldrin. Because of massive volumes used, atmospheric processe s, and their persistence, they remain common environmental contaminants (K alantzi et al. 2001; Matsumura 1985). Specifically, dichlorodiphenyltrichlorethane (DDT) is an organochlorine insecticide and persistent environmental pollutant which is banned in the United States. DDT was widely used as an insecticide in the US fr om the time it was discovered by Peter Mller in 1939 until it was banned in 1972 (Carr and Ch ambers 2001). Detectable levels of DDTs have been found in biological, geologi cal, and atmospheric samples since the 1960s. The primary metabolites of DDT are di chlorodiphenyl-dichloroethane (DDD), dichlorodiphenyl-dichloroet hylene (DDE), and dichlorodi phenylchloroethane (DDMU) (Fig. 1-3). Isomers of DDT and its metabolite s are collectively referred to as DDTs. Soils are a major sink for DDTs (Grau and Peterle 1979), but they are flushed into aquatic systems during flood events (Miglioranza et al. 2003). In soils, two isomers of DDT, o,p -DDT and p,p -DDT, are dechlorinated to o,p -DDD and p,p -DDD by anaerobic

PAGE 24

13 microorganisms (Huang et al. 2001). Finall y, DDD isomers are degraded to isomers of DDE. Most ingestion of DDTs in humans is believed to be of p,p Â’-DDE itself, which is absorbed in the gastro-intestina l tract and stored indefinitely in fat tissues (Moffat 1986). A survey of people not occupationally expos ed to DDT found levels as high as 17 g / g fat (Moffat 1986). This study focuses on p,p Â’-DDE, the most stable and often most abundant metabolite of DDT (Huang et al. 2001; Spies and Thomas 1997). Figure 1-3. DDT and selected metabolites. Although DDT is banned in the US, its metabolites remain common contaminants in locations of spills, sites of heavy and/or continued use domestically and abroad, and in areas of atmospheric deposition. The ubiquito us and liberal use of DDT before it was banned led to heavy contamination of ma ny sites around the USA. In 1965 DDT was applied at the rate of 4 lb / acre in southe rn Arizona (Matsumura 1985). Soils in Florida wetland restoration sites that were formerly farmland contained up to 4,200 g p,p Â’-DDE / kg soil in the early 1990s, while fish tissues from sites had greater than 190 g DDTs / kg (Marburger et al. 2002). Typi cally, concentrations are directly related to the extent to which DDT was used in a particular region (K alantzi et al. 2001). Natural atmospheric processes, however, can transport DDTs from areas of use to more pristine environments. (Catalan et al. 2004) found p,p Â’-DDE in the ng / g range in fish muscle tissue from a high mountain lake of the Pyrenees (2240 m above sea level), where atmospheric deposition Cl Cl Cl Cl Cl Cl Cl Cl p,p-DDE p,p-DDD CCl3 Cl Cl p,p-DDT

PAGE 25

14 was the sole source of OCPs. Micr oorganisms that degrade DDT into p,p Â’-DDE were found in soils where the parent compound was never applied (Miglioranza et al. 2003). The concentration of total OCPs in thos e soils was 656 ng / g dry weight. Though DDT has been banned in many places, its continue d use and physical prope rties are potentially problematic in several regions of the world. Studies analyzing DDTs from differen t levels of biological organization demonstrate the abundance of p,p Â’-DDE in the environment, as well as its propensity to bioaccumulate. A study of DDTs a nd other chlorinated compounds found p,p Â’-DDE was the most prevalent and abundant contaminant in fish tissues from Latvian freshwaters (Valters et al. 1999). p,p Â’-DDE concentrations were found as high as 20 g / g in fish ovaries (Marburger et al. 2002) and 5.8 g / g in alligator eggs (Guillette et al. 1994) from heavily contaminated sites in Florida. Additionally, kelp bass in coastal waters of California had average live r concentrations of 3.43 g / g DDTs, of which greater than 97% was p,p Â’-DDE (Spies and Thomas 1997). p,p Â’-DDE (log Kow = 6.5) is a highly lipophilic compound prone to bioa ccumulation. Fish muscle tissue concentrations of p,p Â’-DDE can be as much greater than concentrations found in water. For example, in a site where lake water contained 7.4 pg p,p Â’-DDE / L (parts per trillion), invertebrates averaged 40.06 ng p,p Â’-DDE / g (parts per million), and brown trout ( Salmo trutta ) had 57.23 ng p,p Â’-DDE / g (Catalan et al. 2004). In California sea otters ( Enhydra lutris ) DDTs were as high as 5,900 ng / g in liver and 4,600 ng / g in kidney of, while prey concentrations ranged from 0.08 to 12.9 ng / g (Kannan et al. 2004). p,p Â’-DDE: An Endocrine Disruptor Initial studies on DDTs in wildlife focused on egg shell-thinning effects in birds, especially raptors. The fist study of DDT on fish and wildlife was conducted in 1946 by

PAGE 26

15 Cottom and Higgins. p,p Â’-DDE was shown to cause repr oductive malfunctions in avian species as far back as the 1960s (Heath et al. 1969). Additional observations and investigations suggested DDTs contribute to en docrine disruption in fish (Macek 1968). In past studies, the reproductive capacity of fishes was adversely affected by p,p Â’dichlorodiphenyldichloro-ethylene (DDE), the most persistent metabolite of DDT (Bayley et al. 2002; Mills et al. 2001). There is significan t evidence suggesting p,p Â’DDE adversely affects fish reproduction and populations. Surveys in Lake Michigan showed a likely connection between a skewed sex ratio in bloater ( Coregonus hoyi ) populations and p,p Â’-DDE concentration. The percent of female bloaters in Lake Michigan returned to normal as fish p,p Â’-DDE concentrations decreased from approximately 3.5 g / g in 1969 to 0.75 g / g in the early 1980s (Monnosson et al. 1997). Several authors suggest that OCPs aff ect hormone homeostasis through steroid synthetic and metabolic pathways (Hornung et al. 2004; Spies and Thomas 1997; Thibaut and Porte 2004). Spies and Thomas (1997) found plasma estradiol levels in fish decreased with concentration of DDTs. On a more mechanistic level, p,p Â’-DDE inhibited steroid synthesis in mammalian ovary cells at 10 M (Chedrese and Feyles 2001), but enhanced steroid synthesis in fish testicular cells at 100 M (Thibaut and Porte 2004). Reproductive abnormalities were observed in male guppies exposed to p,p Â’-DDE during sexual differentiation, a time susceptible to the effects of endocrine disruption (Bayley et al. 2002). In mammals, p,p Â’-DDE affects transcription of androgen-controlled genes by binding to the androgen receptor. In vitro studies in mammalian cells found 200 nM

PAGE 27

16 p,p Â’-DDE inhibited half the androgenic transcri ptional activity induced by a testosterone (Kelce et al. 1995). Those same in vitro studies also suggest p,p Â’-DDE binds the AR, allowing it to enter the nucleus, but pr eventing the AR from inducing androgendependent genes. Ke lce et al. (1995) found p,p Â’-DDE acts as an androgen inhibitor with potency similar to that of hydroxyflutamide (200 nM = IC50) in mammalian cell lines. In fathead minnows, p,p Â’-DDE had a binding affinity for AR similar to dihydrotestosterone, at concentrations of 20 and 22 nM, respectiv ely (Wilson et al. 2004). However, activity of the bound AR was not measured in that st udy. Consequently, there is no confirmation that p,p Â’-DDE bound to the AR actually inhibits transcription of androgen-dependent genes in fathead minnows. Environmental concentrations of p,p Â’-DDE have been found at 80 times the concentrations that cause these anti-androgenic effects in vitro (Guillette et al. 1995; Kelce et al. 1995; Monno sson et al. 1997). Other in vitro studies suggest that p,p Â’-DDE may increase granulosa cell growth by stim ulating progesterone synthesis, but not progesterone synthesis stimulated by 17 -estradiol (Chedrese and Feyles 2001; Crellin et al. 1999). It is, therefore, likely that p,p Â’-DDE does not disrupt hormone homeostasis by interfering at the ER, but with another component of the steroidogenic pathway. p,p Â’DDE has also been shown to increase granul osa cell growth similar to, but less potently than, 17 -estradiol in mammalian cells (Chedres e and Feyles 2001). From that same study, Chedrese and Feyles (2001) found that p,p Â’-DDE decreased progesterone, a hormone required for normal ovulation. A lthough those studies were conducted in mammalian cells, they indicate that p,p Â’-DDE may not be acting at the estrogen-receptor. Extrapolating the anti-androgenic or es trogenic activity found in those mammalian

PAGE 28

17 studies to fish may be invalid. Several in vitro studies suggest ther e may be differences in binding of contaminants to ARs among ma mmals and teleosts, between species of teleosts, and between tissues of a single spec ies (Bayley et al. 2002; Makynen et al. 2000; Wells and Van Der Kraak 2000). Given the uncertainty in the link between AR binding in vitro and in vivo activity, Bayley et al. (2002) concluded that sex char acteristics and reproduction themselves were the best measure of reproduc tive dysfunction. Male guppies ( Poecilia reticulata ) exposed to10 g / g p,p Â’-DDE during sexual development had a sex ratio skewed toward females, increased time to male development, and altered secondary sex characteristics and sperm count (Bayley et al. 2002). A cap tive population of trout exposed to 10, 40, or 80 g / g p,p Â’-DDE in ovo did not have altered sex rati o or reproductive dysfunction upon reaching sexual maturity (Carlson et al. 2000). Several other reproductive endpoints, including gonodosomatic index (GSI), egg production, and fertilization success, were also not affected in trout exposed to p,p Â’-DDE in ovo (Carlson et al. 2000). Increased mortality was, however, observed in pr ogeny spawned from males treated with p,p Â’-DDE in ovo and uncontaminated females. As those studies show, the effects of p,p Â’-DDE on fish can vary greatly among species and can depend on the life-stage at the time of exposure. Linking Biomarkers of Endocrine Disruption to Fish Populations Reproductive effects of EDCs have been observed extensively at biochemical and physiological levels of biologica l organization in fishes. Link ing biomarkers measured in individual fish to population-le vel outcomes has only recently been attempted (Grist et al. 2003) and links between biomarkers and real population-level effects have yet to be confirmed (Mills and Chichester 2005; Segne r 2005). Most studies investigating the

PAGE 29

18 effects of EDCs on population parameters and demographic changes have been on invertebrates because they have shorter gene ration times and are easier to raise in the laboratory than fishes (Barata et al. 2002; Mauri et al. 2003; Raimondo and McKenney 2005). Much of the concern and research of EDCs, however, is related to fish because humans depend on them for nutrition and inco me. Consequently, th ere is a disparity between the knowledge of population-leve l outcomes and the amount of data on biomarkers endocrine disruption in fish. A limited number of recent studies have at tempted to link adult exposure of fishes to EDCs with changes in popul ation growth rate. A study by Miller and Ankley (2004) computed the effects of a synthetic andr ogen on density-depende nt population growth rates of fathead minnows. Grist et al. (2003) investigated the contributions of demographic parameters to changes in population growth rate of fathead minnows exposed to ethynylestradiol during development. They found fertilit ies contributed more than survival probabilities to the highly significant corr elation between ethynylestradiol concentration and population growth rate. Those studies do not, however, account for effects of in ovo exposure on the survival probabilities and fertilities. Few investigations on multi-generational effects of contaminant have been made. Demographic parameters of offspring exposed to EDCs maternally may be of great importance. Nash et al. (2004) exposed two generations of fathead minnows to environmentally relevant concentrations of the potent estrogen, ethynylestradiol. That study showed no change in fertil ities of exposed adults, but a reduction in fertility of the second generation, even after depuration. As Nash et al. (2004) concluded, those findings carry major implications for populati on-level impacts of long-term exposure of

PAGE 30

19 fish to EDCs. Therefore, it is important to investigate second-generation effects of EDC exposure. Ankley et al. (2001) developed a protoc ol for measuring reproductive effects of sub-chronic exposure to EDCs of fathead minnows, a species commonly used in such assays. That and other studies call for a deeper knowledge of th e connection between contaminant tissue burdens, reproductive dys function, and populati on effects (Chedrese and Feyles 2001; Foster et al 2001; Gray et al. 2002; Orland o et al. 1999). Employing an exposure methodology similar to that described by Ankley et al. (2001) I investigated the effects of parental exposure to p,p Â’-DDE on a suite of bi omarker and population parameters of fathead minnows. Because EDCs act on the reproductive system by definition, I expected fertilit ies to change more than survival among treatment groups. Feed concentrations of p,p Â’-DDE were chosen to represent environmentally relevant p,p Â’-DDE body burdens (Marburger et al. 2002; Muller 2003). In this case, the stimulus for altered biological responses is exposure to various concentrations of p,p Â’DDE through the diet. This study had three overall goals: to a ssess the effects of p,p Â’DDE on reproduction and endocrinology of fath ead minnows, to assess the effect of in ovo exposure to p,p Â’-DDE on survival and development, and to link those effects with biomarkers of exposure. To this end, I conducted a pilot experiment in which fish were exposed to high levels of p,p Â’-DDE and the anti-androgen flutamide. Then, I conducted a dose-response experiment, fr om which the effects of in ovo exposure were assessed. Finally, the dose-response experiment was repeated to obtain additional biological materials to use for measuring biomarkers.

PAGE 31

20 Table 1-1. Classes of endocrine disrupting compounds and examples. Compound class Type Examples organochlorine pesticides (OCPs) DDT, methoxychlor, toxaphene organophosphate pesticides (OPs ) TEPP, chlorpyrifos, malathion Pesticides p yrethrins extracts of Chrysanthemum flowers p olycyclic aromatic hydrocarbons (PAHs)benzo(A)pyrene, aflatoxin p lasticizers di( n -butyl)pthalate Industrial byproducts p olychlorinated biphenyls (PCBs) Aroclor 1254 Hormonal therapeutics birth control pills, hormone therapy ethynylestradiol Natural hormone estradiol, testosterone (Cook et al. 2003; Macek 1968; Miller and Ankley 2004; Mills et al. 2001; Nash et al. 2004; Thompson et al. 2004; Valters et al. 1999)

PAGE 32

21 MATERIALS AND METHODS Three in vivo experiments were conducted on fathead minnows ( Pimephales promelas ). Each experiment was designed to optimize collection of the endpoint of interest. The focus of the initial pilot experiment was to assess if p,p Â’-DDE caused reproductive dysfunction similar to the anti-androgen flutamide. Upon observing reproductive dysfunction in fish administered a high level of p,p Â’-DDE, a dose-response experiment to determine a no observed effect level (NOEL) was conducted. During that second experiment fish were fed one of three environmentally plausible p,p Â’-DDE concentrations. That experiment focused on the accumulation rate of p,p Â’-DDE in adults and survival, development, and reproduction of offspring spawned from those adults. In a third experiment, p,p Â’-DDE was administered to fish similar to the second experiment to obtain additional biological materials for hormone and gene analyses. All experiments were conducted at the University of Florid a Aquatic Toxicology Facility under the same general environmental conditions in accordance with IACUC protocols. Differences in experimental conditions ar e described below. Methods for determination of p,p Â’-DDE content, plasma 17 -estradiol levels, and mRNA e xpression were identical across experiments unless othe rwise noted. General Methods Fish Holding Conditions Fish were housed in flow-through tanks s upplied with dechlorinated water and kept on a 16 hours light: 8 hours dark schedule. Fish were exposed to p,p Â’-DDE through

PAGE 33

22 contaminated Silvercup Trout Chow (Zeigler Br others, Inc.). Food was stored at 4 C. Water temperature was measured daily, wh ile dissolved oxygen (DO), pH, hardness measured as CaCO3, and total ammonia were monitored weekly. Water quality parameters were as follows for flow-t hrough tanks: DO was 8.5-8.9 mg/L; pH was 8.7 1; hardness was 40 2 mg CaCO3/L; and total unionized ammoni a was always less than 0.5 mg/L in flow-through tanks. Reproductive Measures Reproductive output was measured as the num ber of eggs produced, percentage of eggs fertilized, and percentage of fertilized eggs that ha tched. One spawning substrate (3-inch sections of 3-inch di ameter polyvinyl chloride pipe) per male fish was kept in each tank at all times. Spawning substrates were checked for eggs each afternoon. If eggs were present, the spawning substrate was removed, eggs counted, and placed in an aerated 2 L glass beaker fille d with approximately 1.75 L dechlorinated water and 25 ml blackwater extract (Aquatic Ecosystems, Inc.). Blackwater extract was used to prevent fungal infection on eggs, which was determin ed the best method for preventing fungal infection. The number of eggs fertilized wa s determined by counting eggs that developed eyes 2-3 days post spawn (dps). Water within the beaker was changed when fertilization was determined. Digital overhead photographs were used to count the number of eggs that hatched 5-8 dps, depending on when eggs were no longer present on the substrate. Plasma and Tissue Collection At the end of each exposure period adult fish were anesthetized with MS-222 (100mg/L buffered with 200 mg NaHCO3/L), killed by decapitation, bled, and tissue collected (Ankley et al. 2001). Blood was collected from the caudal sinus in heparinized micro-hematocrit capillary tubes (Fisher Scientific Company), centrifuged at 1,500 X g

PAGE 34

23 for 10 minutes, and plasma was removed and fr ozen at -80 C. Gonads were excised, weighed for determination of the gonadosomatic index (GSI = [gonad weight / body weight] 100), and flash frozen in liquid nitr ogen. Liver and brain were removed, flash frozen, and stored at -80 C. Carcasses were eviscerated and stored at -20 C until analyzed for p,p Â’-DDE content. Determination of p,p Â’-DDE Content Fish were analyzed for muscle tissue or egg p,p Â’-DDE concentration (wet weight) by gas chromatography / mass spectrometry by th e method described in (Glesleichter et al. 2005) and modified as follows. One gram or 2.5 g were sectioned from the eviscerated carcass posterior of the opercle of each female and male fish, respectively. The fish tissue was homogenized by a Tekman Tissumizer (Tekman Company) with 3 g d10-phenanthrene as an internal standard (Protocol Analytical, LLC), 2.5 times tissue weight of Na2SO4 (A.C.S. Grade, Fisher Scientific Company), and 7 ml n-hexanes (A.C.S. Gade, Fisher Scientific Company) vortexed, and centrifuged for 15 minutes at approximately 100 X g. The supernatant was decanted and the homogenate was extracted twice more with 3 ml n-hexanes. Extracts were combined to yield a total extracted volume of 13 ml for each tissue sa mple. The extract was dried under a stream of nitrogen at 35 C. The dried extract wa s reconstituted in 3 ml acetonitrile (Optima Grade, Fisher Scientific) and eluted through a pre-conditioned SPE-C18 cartridge (Agilent Technologies), which was repeated on ce. The cartridge was rinsed with 1 ml acetonitrile. The eluate was then passed through an SPE-NH2 cartridge (Varian, Inc.) and the glass tube containing the eluate was rinsed with 1 ml acetonitrile, which was also placed over the SPE-NH2 cartridge. The fi nal eluate was dried under a stream of

PAGE 35

24 nitrogen at 35 C and reconstituted in 1 ml 3 g d10-pyrene / ml cyclohexane (Ultra Scientific). A Shimadzu 17A gas chromatograph (Shima dzu Scientific Instruments) with HP5MS column dimensions of 29 m 0.25 mm coupled with a Shimadzu QP-5000 were used for analyte separation and detection. On e microliter of recons tituted extract was injected into a splitless inlet at 275 C. Analytes were separated using the following program: initial oven temperature was 100 C held for 2.5 min. Temperature was ramped to 190 C at 15 C / min, then to 250 C at 5 C/min and finally to 290C at 20C/min, which was held for 5 min. Initial carrier fl ow was 1.4 ml/min. This was reduced to 1 ml/min at 2.5 min for the remainder of the program. Interface temperature was maintained at 280C. Mass spectrometer was operated in selected ion monitoring (SIM) mode and m/z 246 and 317 were used for p,p Â’-DDE and m/z 188 was collected for d10phenanthrene. Quantitation was performed us ing the ratio of area of m/z 246 to m/z 188 for each p,p Â’-DDE concentration. Determination of Plasma 17 -Estradiol Hormone levels were determined by enzyme immunoassay (EIA) after plasma was extracted with organic solvent. Briefly, 180 l EIA buffer was added to 10 l plasma to increase aqueous volume. The plasma and bu ffer mixture was then extracted twice with 0.75 ml ethyl ether (Pesti cide Grade, Fisher Scientific Co mpany). The ethyl ether extract was evaporated under a gentle stream of nitr ogen in a water bath at 30 C. The extract was reconstituted in 200 l EIA buffer, mixed by vortex, a nd placed on an orbital shaker at 4 C overnight. This method of extr action and reconstitution was validated by counting a known amount of extracted and reconstituted 3H-estradiol stock solution and comparing it with a known amount of unadultera ted stock solution. The efficiency of the

PAGE 36

25 extraction and resuspension was greater than 90 percent. Estradiol standards were prepared by bringing 10 l of each 15 ng/ml, 10 ng/ml, 5 ng/ml, 2.5 ng/ml, 1.25 ng/ml, 0.625 ng/ml, 0.312 ng/ml, 0.156 ng/ml, 0.078 ng/ml, and 0.039 ng/ml to a total volume of 200 l in EIA buffer, mixed by vortex, and placed on an orbital shaker at 4 C overnight. Standards and reconstituted sample were mixed by vortex immediately prior to 17 estradiol determination. Each standard a nd sample was measured for estradiol in duplicate according to protocol for estradio l EIA kit (Cayman Chemical Company). The quantifiable range for this assay was from 0.156 to 15 ng estradiol / ml plasma. An effort to analyze estradiol in small volumes of plasma by radioimmunoassay was made, but acceptable validation was not achieved (see Appendix). Vitellogenin mRNA Quantification RNA was isolated from liver usi ng Trizol (Invitrogen), following the manufacturerÂ’s instructions and reconstitu ted in RNA Secure (Ambion). RNA quality was verified on ethidium bromide-staine d 1.5% formaldehyde-agarose electrophoresis gels. RNA was considered to have acceptable purity when the A260 nm/A280 nm ratios were greater than 1.8, as determined on a spectrophotometer (NanoDrop Technologies). cDNA was made by reverse transcription-poly merase chain reaction (PCR) using random decamer primers and 2 g DNA-free RNA according to manufacturerÂ’s instructions for the RETRO-Script Kit (Ambion). The quality of cDNA was verified by gel electrophoresis of products from the follo wing PCR program using 18S primers: the reaction was held at 94 C for 2 min, then 35 cycles of 30 sec at each 94 C, 55 C, and 72 C, and a final extension of 5min at 72 C Real time-PCR was conducted using pairs of oligonucleotide primer sequences for v itellogenin (Table 1), using 18S as the housekeeping gene (Filby and Tyler 2005).

PAGE 37

26 Experimental Set-Up Pilot Experiment The goal of the pilot experiment was to determine if p,p Â’-DDE affected reproductive output and 17 -estradiol levels of fathead minnows similar to the antiandrogen flutamide. To meet this objectiv e, adult fathead minnows (7-10 months old) were obtained from a local fish breeder (F ish Soup, Newberry, FL. One male and two females were housed in 5-gallon flow-through tanks. There were six replicate tanks within each treatment group, totaling 18 tanks for the entire experiment. Mean water temperature was 22 2 C. Treatment gr oups included a control group (vehicle only feed), a positive control group (1208 g flutamide / g feed), and a p,p Â’-DDE group (900 g p,p Â’-DDE / g feed). Contaminant concentrat ions were measured by the Analytical Toxicology Core Laboratory at the University of Florida. Fish feed was coated with menhaden oil and acetone containing the a ppropriate chemical concentrations, mixed, and placed under a fume hood overnight. p,p Â’-DDE (2,2-bis(4-chlorophenyl)-1,1dichloroethylene, 99.4% purity) was obtained from Aldrich Ch emical Company. Fathead minnows housed at the facility consum ed menhaden oil-coated feed without discrimination. Adult survival and repro ductive output were measur ed for 21 days prior to chemical exposure. During that pre-exposur e period fish were fed approximately 0.5 g feed twice and live brine shrimp, Artemia (G reat Lake Artemia) once daily. Once egg production was established during the pre-e xposure period, the 26-day exposure period immediately followed the pre-exposure period. During the exposure period fish were fed ~ 2 g p,p Â’-DDE -contaminated feed twice daily and no brine shrimp. For this experiment, the total number of eggs per female was used as the sole measure of reproductive output.

PAGE 38

27 At the conclusion of the exposure period fish were sacrificed and plasma 17 -estradiol levels were determined as described above. p,p Â’-DDE Dose-Response Experiment I: Accumulation Rate and Reproductive Output of Adults Adult fathead minnows (5-7 months old) we re obtained from Aquatic Biosystems, Inc. (Fort Collins, CO). Each treatment gr oup consisted of four males and eight females housed in 30-gallon flow-through tanks. Wate r temperature for experiment two was 25 1 C in flow-through tanks and 25 2 C in egg-housing beakers. p,p Â’-DDE was dissolved in acetone, sprayed onto fish food, mixed for several minutes, and placed under a fume hood to allow the acetone to evaporate overnight to obtain nominal concentrations of 2.5 g/g, 25 g/g, and 250 g/g. Actual feed concentrations of p,p Â’-DDE were 0.03 g / g, 1.63 g / g, 11.48 g / g, and 104.25 g / g, which will be referre d to as control, low, medium, and high groups, respectively. Each tank was fed approximately 2 g of food at both 10 AM and 6 PM daily for 29 days. Fathead minnows housed at the facility consumed vehicle only feed without discrimination. Reproductive output measured daily thr oughout the exposure period, as described above. Upon determining sex at sacrifice, it was evident that each treatment group did not have the same number of females throughout the experiment. This is due to the fact that some males did not display secondary sex characteristics and therefore were mistaken for females when allocated to each tank at the outset of the experiment. Consequently, the number of eggs produced in each treatment group was calculated on a per female basis. One male and two females from each tr eatment group were killed and tissues collected as described above on da ys 7, 14, 21, and 29 of exposure to p,p Â’-DDE -

PAGE 39

28 contaminated feed. Body burden was determined by gas chromatography / mass spectrometry as described above. Reproductive output was monitored as described above during the 28-day exposure period. Eggs co llected during the 28day exposure period were grouped into four clutches based on th e day they were spawned: clutch one was spawned from 2 d to 9 d; clutch two was sp awned from 10 d to 16 d; clutch three was spawned from 17 d to 23 d; clutch four was sp awned from 24 d to 29 d. Clutch three and four were used to study the effects of in ovo p,p Â’-DDE exposure, as described in the following section. Survival, Development, and Repro ductive Output of Offspring Offspring studies were conducte d on eggs spawned from the p,p Â’-DDE doseresponse response experiment of control adults or adults exposed to low, medium, and high doses of p,p Â’-DDE. Offspring for each treatm ent group were pooled based on day spawned into a 2.5-gallon flow-through tank. Su rvival of offspring was measured weekly for three weeks post-hatch by counting fish in digital overhead photographs taken of each beaker. During those three weeks offspring we re kept in 2 L flow-through tanks and fed live Artemia, hatched in artifi cial sea water. After four weeks, 100 offspring (juveniles) from clutch four were taken from each tr eatment group and placed in 5-gallon flowthrough tanks with spawning substrates. O ffspring from the low treatment group were taken from clutch three because adults did not spawn during the final six days of the experiment. All further studies on offspring were conducted on these 100 fish. Survival of these fish was measured monthly. Seconda ry sex characteristics and first spawn where monitored approximately every other day afte r four months post-spawn. The offspring were transferred to 30 gal flow-through ta nks and provided spawning substrates at 3.5 months of age. Reproductive output of offs pring was measured, as described above, for

PAGE 40

29 one week each month after the group initia ted spawning. Offspri ng length and weight were measured approximately four months post-spawn and killed for tissue collection and measurement at approximately nine months of age. These fish were not analyzed for p,p Â’-DDE content, estradiol or gene expression. p,p Â’-DDE Dose-Response Experiment II: Collection of Biological Materials Fathead minnows were bred in-house from fish received from Fish Soup (Newberry, FL). At the time of use the fish were 12-15 months old. A fresh batch of p,p Â’-DDE -contaminated feed was prepared to ac hieve concentrations similar to those in the first p,p Â’-DDE dose-response experiment. Actual p,p Â’-DDE concentrations for this batch of feed were 0.03 g / g, 1.38 g / g, 15.58 g / g, and 104.73 g / g, which was fed to the control, low, medium, and high gr oups, respectively. Reproductive output was not measured during this experiment. Plasma and tissues from this experiment were used to determine 17 -estradiol levels, p,p Â’-DDE tissue concentration, and vitellogenin mRNA expression in fish with body burdens similar to those achieved in the first p,p Â’-DDE doseresponse experiment. Statistical Analyses Analysis of variance (ANOVA) followed by post-hoc comparisons using TukeyÂ’s HSD were used to test for significant diffe rences among treatment groups. Data were log-transformed to meet the assumption of nor mality as necessary. Two-tailed PearsonÂ’s correlations of measures taken in the same fi sh were computed. Values are reported as the mean standard error of the mean (S.E.). Significance was set at p<0.05. All statistical analyses were conducted in SPSS 13.0 for Windows (SPSS, Chicago, IL, USA).

PAGE 41

30 Two-tailed PearsonÂ’s correlation of p,p Â’-DDE concentration (c ontrol, low, medium, or high) to biomarkers in fish exposed th rough feed (i.e. GSI and plasma) estradiol and fish exposed in ovo (i.e. GSI) and population-level effect s (i.e. survival and measures of reproduction) to identify biomarkers of multi-ge nerational effects. Because there were no replicates, these correlations are not robust, and therefore onl y simple relationships were further investigated and displayed. Table 2-1. Primer sequences used for quantitative real-time PCR. Gene Forward Primer Reverse Primer Vitellogenin 5Â’-GCT GCT GCT CCA TTT CAA AAG3Â’ 5Â’-GTG AGA GTG CAC CTC AAC GC3Â’

PAGE 42

31 RESULTS Pilot Experiment Egg output was measured in treatment and control groups during a pre-exposure and exposure period. Fish in each tank dem onstrated a capability for spawning during the pre-exposure period. Fish in the p,p Â’-DDE group ceased spawning activity after just three days of exposure to the p,p Â’-DDE, while fish in the control and flutamide groups continued to spawn during the exposure peri od (Fig. 3-1). During the exposure period the p,p Â’-DDE group spawned less than 16 eggs per female, while the control group spawned 195 eggs per female and the flutamide group spawned 240 eggs per female. Plasma 17 -estradiol (estradiol) was measured in control fish and fish treated with p,p Â’-DDE or flutamide to elucidate whet her the two compounds act by the same mechanism in the fathead minnow. Mean plas ma estradiol in male s from the flutamide group was 4.0 0.56 ng / ml, which was significantly greater than 2.02 0.35 ng / ml in the control group and 1.88 0.27 ng / ml in the p,p Â’-DDE group (ANOVA, p<0.01; Fig. 3-2). There were no significant differen ces in estradiol levels in females among treatment groups (ANOVA, p=0.29). The mean estradiol level in females in the p,p Â’DDE group (4.82 2.65 ng / ml) was, however, a pproximately half of the mean levels measured in both the control (8.96 1.72 ng / ml) and flutamide (10.50 2.75 ng / ml) groups. Plasma estradiol levels were signi ficantly different between males and females of the control group (p=0.01), but not the p,p Â’-DDE or flutamide groups (ANOVA, p=0.63 and p=0.12, respectively). Although plasma estradiol levels were much greater in

PAGE 43

32 females (10.51 2.75 ng / ml) than males (4.01 0.58 ng / ml) in the flutamide group, the difference between the two sexes was not statistically significant (ANOVA, p=0.12). p,p Â’-DDE Dose-Response Experiments I & II: Adult 17 -Estradiol, GSI, and Reproduction p,p Â’-DDE accumulation rates in fish tissue a nd reproductive output of adults were measured from fish in the p,p Â’-DDE dose-response experiment I. Biomarkers including adult GSI, plasma 17 -estradiol, and vitellogenin gene expression reported here were measured from fish from p,p Â’-DDE dose-response experiment II. Survival of adult fish was not affected in any of the treatment groups. Concentration of p,p Â’-DDE in adult muscle tissues was quantified in three fish at seven-day intervals during the exposure period. p,p Â’-DDE tissue concentrations appeared to approach, but not at reach steady state by the end of the experime nt (29 days; Fig. 3-3 A and Fig. 3-3 B). Fish taken at the end of the exposure achieved p,p Â’-DDE muscle tissue concentrations (w et weight) of 0.38 0.05 g /g, 3.11 0.57 g / g, and 21.14 2.72 g / g in the low, medium, and high dose gr oups, respectively. The concentration of p,p Â’-DDE in control fish, 0.12 0.10 g / g (n=6), was used as the concentration on day one for each group. Plasma from these fish was used to determine 17 -estradiol levels. Plasma estradiol levels, as measured by EIA, diffe red among treatment groups. Females in the high group had mean plasma estradiol level of 6.70 2.53 ng/ml (n=5), which was lower than the control, low, and medium groups w ith average levels of 9.38 1.09 ng/ml (n=5), 10.82 2.04 (n=5), and 10.48 2.13 (n=5), respectively. There were no significant differences in female estradiol levels among treatment groups (Fig. 3-4; ANOVA, p=0.51). Plasma estradiol in female s was not significantly correlated to p,p Â’-DDE

PAGE 44

33 muscle tissue concentration or GSI (R2 =-0.34, p=0.35 and R2 =-0.49, p=0.60 respectively). Males in the low, medium and high p,p Â’-DDE groups had mean plasma estradiol levels of 7.72 2.85 ng/ml (n=5 ), 3.85 0.38 ng/ml (n=4), 5.58 2.63 ng/ml (n=4), whereas 0.37 0.12 ng/ml (n=6) was the mean in the control group. Males in all p,p Â’-DDE groups had significantly higher plasma estradiol levels than males in the control group (Fig. 3-4; ANOVA, p<0.01). Plasma estradiol in males was not significantly correlated to p,p Â’-DDE muscle tissue concentration or GSI (R2 =-0.26, p=0.82 and R2 =-0.19, p=0.68, respectively). Plasma es tradiol was significantly different between males and females in the contro l (ANOVA, p<0.01) and medium groups only (ANOVA, p=0.03). While there were no significant differen ces of either sex among treatment groups, the high groups tended to have noticeably lo wer GSI than control (Fig. 3-5). GSI in males was 1.34 0.23 (n=6) in the control group, 1.16 0.26 (n=6) in the low group, 1.15 0.29 (n=6) in the medium group, and 0.97 0.13 (n=4) in the high group (ANOVA, p=0.81; Fig 3-5 A). In female s GSI was 7.96 0.73 (n=6) in the control group, 9.32 1.53 (n=6) in the low group, 8.36 1.27 (n=6) in the medium group, and 5.67 0.98 (n=7) in the high group (ANOVA, p=0.16; Fig 3-5 B). GSI was not significantly correlated to p,p Â’-DDE muscle tissue concentrations in males or females (p=0.56 and p=0.13, respectively). No adverse effect of p,p Â’-DDE treatment was observed in reproductive output in fish from the low and medium p,p Â’-DDE groups. Fish in the high group, however, spawned 77 percent fewer eggs than the control group at the conclusion of the experiment (Fig. 3-6). The percent of e ggs successfully fertilized wa s at least 90 percent in the

PAGE 45

34 control, low and medium groups, but only 65 percent in the high group (Fig. 3-7 A). Similar to fertilization success, the percent of fertilized eggs that hatched did not show a dose-related response (Fig. 3-7 B). Percent hatch in the low group was 68 percent, but in all other groups percent hatch exceeded 93 percent. Frozen tissue samples of liver were used to compare expression of vitellogenin in treated fish to control fish. Results for this analysis were not definitive due to amplification of multiple products by prim ers for the housekeeping gene, 18S. Because multiple products were amplified, 18S could not be used to normalize and quantify the relative expression of the gene of interest, vitellogenin. To troubleshoot this problem, I used a new stock of the 18S primer, which also amplified multiple products. I tried amplifying other housekeeping genes, including L8 and -actin, but had no success. Effects of in ovo Exposure on Survival, Development, and Reproduction Concentration of p,p Â’-DDE was measured in eggs spawned from adults p,p Â’-DDE dose-response experiment II. p,p Â’-DDE in eggs spawned from the control group was below the limit of detecti on. The concentration of p,p Â’-DDE in the eggs from the medium and high groups was approximately 30 percent of the concentration found in adult females in the medium and high p,p Â’-DDE groups (Fig. 3-8). Eggs from the medium group contained 1.40 g p,p Â’-DDE /g (n=1) and adults contained 3.11 0.37 g p,p Â’-DDE /g muscle tissue (n=3). In the high group, eggs contained 7.60 g p,p Â’-DDE /g (n=1) and adult muscle tissue had 21.14 2.72 (n=3). Eggs spawned from the low group were not available for p,p Â’-DDE analysis. All measurements of offspring are from eggs spawned in the p,p Â’-DDE accumulation and reproductive output of adults experiment. Weekly survival was measured for the first three weeks of life to assess early life-stage mortality related to

PAGE 46

35 p,p Â’-DDE consumption of adults. There was considerable variab ility among treatment groups in weekly offspring survival during the first three weeks after hatch (Table 3-1). Offspring in the low and control group had identical average weekly survival probabilities during the first th ree weeks after hatching. Th e medium offspring showed the highest survival probability of 0.87 and the high group the next highest of 0.79. After one month of life, there were only minor di fferences in monthly survival probabilities among treatment groups, which were grea ter than or equal to 0.97. The effects of in ovo exposure of p,p Â’-DDE on FHM development were measured as body size at four months of age, appearan ce of secondary sex characteristics, and age at first reproduction. Length and weight were measured when fish reached approximately four months of age to assess differences in growth among treatment groups. Fish in the low group weighed si gnificantly more than all other groups (ANOVA, p<0.01; Fig. 3-9). Th e medium group displayed secondary sex characteristics 15 days before offspring in the control a nd low groups (Fig. 3-10), while the high group displayed secondary sex characteristics 22 da ys after the control and low groups (Fig. 310). Age at first reproduction in the high group was 33 days gr eater than in the control, low, and medium groups. The low group showed sexual development identical to control offspring. Fish exposed in ovo were killed at nine months of age for determination of sex and gonad weight. Sex ratios differed consid erably among treatment groups, but showed no dose-dependent trend (Fig. 3-11). The c ontrol and high group ratios were close to 50:50, but the low group was skewed toward fe males and the medium toward males. Average male GSI from offspring in the hi gh group was significantly greater than all

PAGE 47

36 other groups (ANOVA, p<0.01; Fi g. 3-12 A). Females GSI di d not differ significantly among treatment groups (ANOVA, p=0.07; Fig. 3-12 B). Reproductive output of fish exposed in ovo did not show the same pattern among treatment groups as either sex ratio or GSI. The total number of eggs produced in each group over nine months from greatest to least was from the low, medium, control, and high groups (Fig. 3-13 A). Cumulative egg production was also calculated on a per female basis, where sex ratio was used to estimate the number of females in each treatment group during each collection period. On a per female basis, the number of eggs per female from greatest to least was from the medium, low, control, and high groups (Fig. 3-13 A). By both measurements of egg production, the high group produced considerably less eggs than all other treatment groups. Identification of Biomarkers Correlations of among average measure of biomarkers and reproductive effects were conducted using each treatment group (n=4 ). Biomarkers and reproductive effects that were related were graphed for visual representation. Among adults, average female estradiol and GSI were directly related to the percent of eggs fertilized and average male estradiol was related to the number of e ggs per female (Fig. 3-14 and Fig. 3-15). Average female estradiol and GSI were invers ely related to GSI in both sexes of their offspring (Fig. 3-16).

PAGE 48

37 Table 3-1. Survival probabilities fo r offspring spawned from adults fed p,p Â’-DDE during days 23-29. Survival probabilities were measured weekly for the first three weeks and monthly from one month thr ough eight months post-hatch. Sample sizes are in parentheses next to each survival probability for 1-3 weeks posthatch. After one month, fish populations were normalized to 100 offspring in each treatment group. Average Survival Probabilities by Treatment Group Control Low Medium High Mean weekly survival for three weeks post hatch 0.69 (n=254) 0.69 (n=288) 0.87 (n=645) 0.79 (n=152) Mean monthly survival for 1-8 months post hatch 0.99 0.99 0.97 0.98 0 50 100 150 200 250 300 350 400 -25-15-551525Time (days)# eggs / # females Control 900 ppm DDE 1208 ppm flutamide Figure 3-1. The number of eggs produced per fe male before and after exposure to either p,p Â’-DDE of flutamide as compared to control. Eggs per female during the pre-exposure period is show n to the left of the y -axis and eggs per female after fish were administered contaminated feed is shown to the right of the y -axis. There were 10 females in the control group, 7 in the p,p Â’-DDE group, and 12 in the flutamide group.

PAGE 49

38 Treatment Group ControlDDEFlutamideMean Plasma 17-Estradiol (ng/ml) 0 2 4 6 8 10 12 14 Male Female a a b*6 10 7 5 6 12 Figure 3-2. Comparison of plasma 17 -estradiol levels in males and females within and among each treatment group. Values re present mean S.E. Significant differences among males among treatment groups are indicated by different letters (p<0.05). Significant differences between males and females within a treatment group are indicated by (p<0.05). Study Day 07142128DDE Tissue Concentration (g/g) 0 5 10 15 20 25 30 Low Medium High Study Day 07142128DDE Tissue Concentration (g/g) 0.0 0.1 0.2 0.3 0.4 0.5 Low A B Study Day 07142128DDE Tissue Concentration (g/g) 0 5 10 15 20 25 30 Low Medium High Study Day 07142128DDE Tissue Concentration (g/g) 0.0 0.1 0.2 0.3 0.4 0.5 Low Study Day 07142128DDE Tissue Concentration (g/g) 0 5 10 15 20 25 30 Low Medium High Study Day 07142128DDE Tissue Concentration (g/g) 0.0 0.1 0.2 0.3 0.4 0.5 Low A B Figure 3-3. Mean p,p Â’-DDE muscle tissue concentration ( g p,p Â’-DDE / g wet weight muscle tissue). A) p,p Â’-DDE measured in three fish from each treatment group every seven days during exposure. B) Expanded view of mean p,p Â’DDE tissue concentrations measured in three fish from the low dose group every seven days during exposure. The concentration of three control fish represents the starting p,p Â’-DDE body burden for all treatment groups. Values are the mean S.E.

PAGE 50

39 Treatment Group ControlLowMediumHighMean Plasma 17-estradiol (ng/ml) 0 2 4 6 8 10 12 14 Male Female a*b b b* Figure 3-4. Comparison of plasma 17 -estradiol levels in male s and females from each treatment group. Values represent m ean S.E. Significant differences between males and females within a treatment group are indicated by (p<0.05). Different letters indicate sta tistically significant differences of the same sex among treatment groups (p<0.05). The number males in each treatment group were n=6 (control), n=5 (low), n=4 (medium), and n=4 (high). The number females in each treatment group were n=6 (control), n=5 (low), n=5 (medium), and n=5 (high). Figure 3-6. Mean GSI values of male adults treated with DDE-contaminated feed. Error bars represent one standard error. Sample si ze is six for the control, low, medium groups and five for the high group. Figure 3-7. Mean GSI values of female adu lts treated with DDE-contaminated feed. Error bars represent one standard error. Samp le size is six for the control, low, medium groups and seven for the high group. Figure 3-5. Mean GSI values of adults treated with p,p Â’-DDE-contaminated feed. A) Male GSI. B) Female GSI. Error bars represent one standard error. Sample size is six for the control, low, medi um groups and five for the high group. Treatment Group ControlLowMediumHighMean GSI +/S.E. 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 control low medium high Treatment Group ControlLowMediumHigh Mean GSI +/S.E. 0 2 4 6 8 10 12 A Males B Females Treatment Group ControlLowMediumHighMean GSI +/S.E. 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 control low medium high Treatment Group ControlLowMediumHigh Mean GSI +/S.E. 0 2 4 6 8 10 12 A Males B Females

PAGE 51

40 Exposure Day 051015202530Number of Eggs per Female 0 100 200 300 400 500 Control Low Medium High Figure 3-6. Cumulative number of eggs produced per female in adults of each treatment group The number of females in each tr eatment group were n=7 (control), n=4 (low), n=7 (medium), and n=8 (high). Percent of Eggs Fertilized 0 20 40 60 80 100 120 Control Low Medium High Percent of Fertilized Hatched 0 20 40 60 80 100 A B Percent of Eggs Fertilized 0 20 40 60 80 100 120 Control Low Medium High Percent of Fertilized Hatched 0 20 40 60 80 100 A B Figure 3-7. Egg fertilization and hatch success. A) The percent of eggs fertilized in each treatment group. B) The percent of fertilized eggs that hatched in each treatment group. The number of eggs in each treatment group were n=307 (control), n=457 (low), n=645 (medium), and n=234 (high).

PAGE 52

41 Treatment Group ControlLowMediumHighDDE (g /g) 0 5 10 15 20 25 Adult Body Burden Egg Burden Figure 3-8. Matern al transfer of p,p Â’-DDE to eggs. Treatment Group ControlLowMediumHighLength (mm) 0 10 20 30 40 50 Weight (g) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 mean length mean weight a b a a cd c d cd Figure 3-9. Mean length (mm) and weight (g) S.E. of offspring at four months after hatch (n=50 for each treatment group). Error bars represent one standard error. Different letters above each bar indicate a significant difference between treatment groups (p<0.05).

PAGE 53

42 Treatment Group ControlLowMediumHighAge of Offspring (days) 80 100 120 140 160 180 Male Characteristics First Reproduction Figure 3-10. The age when male characteri stics or first reproduction was observed in offspring from each treatment group. Treatment Group ControlLowMediumHigh% female or % male 0 20 40 60 80 % female % male Figure 3-11. Sex ratio of offspring in each treatment group (n=50 for each group).

PAGE 54

43 Treatment Group ControlLowMediumHighMean GSI S.E. 0 2 4 6 8 10 12 control low medium high Treatment Group ControlLowMediumHighMean GSI S.E. 0 5 10 15 20 A MalesB Females a aa b Treatment Group ControlLowMediumHighMean GSI S.E. 0 2 4 6 8 10 12 control low medium high Treatment Group ControlLowMediumHighMean GSI S.E. 0 5 10 15 20 A MalesB Females a aa b Figure 3-12. GSI values of offspring in each treatment group. A) Mean male GSI. B) female GSI. Error bars represent one st andard error. Signi ficantly different values (p<0.05) are indicated by di fferent letters above each bar. ControlLowMediumHighcumulative eggs 0 100 200 300 400 500 600 700 ControlLowMediumHighcumulative eggs / female 0 5 10 15 20 25 A B ControlLowMediumHighcumulative eggs 0 100 200 300 400 500 600 700 ControlLowMediumHighcumulative eggs / female 0 5 10 15 20 25 A B Figure 3-13. Cumulative egg production of offs pring through nine months of age. A) Cumulative number of eggs produced from offspring. B) Cumulative number of eggs produced per female from offspring.

PAGE 55

44 Treatment Group ControlLowMediumHighFemale 17-estradiol and GSI 0 2 4 6 8 10 12 % Eggs fertilized 60 70 80 90 100 110 female GSI female estradiol (ng / ml) % eggs fertilized Figure 3-14. Values of mean female plasma estradiol, mean female GSI, and the percent of eggs fertilized are given for each treatment group to show relationships among the variables. Estradiol and GSI values are on the left y -axis and the percent of eggs fertilized is on the right y -axis.

PAGE 56

45 Treatment Group ControlLowMediumHighMale 17-estradiol 0 2 4 6 8 Eggs per female 0 10 20 30 40 50 60 adult male estradiol (ng / ml) #eggs / female Figure 3-15. Values of mean plasma estradio l and the percent of eggs per female are given for each treatment group to show relationships among the variables. Estradiol values are on the left y-axis and the number of eggs per female is on the right y -axis.

PAGE 57

46 Treatment Group ControlLowMediumHighFemale 17-estradiol and GSI 0 2 4 6 8 10 12 % Eggs fertilized 60 70 80 90 100 110 female GSI female estradiol (ng / ml) % eggs fertilized Figure 3-16. Values of mean adult female pl asma estradiol and mean adult GSI for fish exposed as adults and for fish exposed in ovo are given for each treatment group to show relationships among the va riables. Estradio l and GSI values are on the left y -axis.

PAGE 58

47 DISCUSSION Effects of Adult Exposure to p,p Â’-DDE on Reproduction, 17 -Estradiol, and GSI One of the primary goals of the doseresponse experiments was to achieve a relatively constant body burden which resembled that of wild fish from previous studies. Fish in the low and medium p,p Â’-DDE groups achieved a muscle tissue concentration of approximately 30 percent and the high group achieved approximately 20 percent of the feed concentration of p,p Â’-DDE. Largemouth bass accumulated greater than 30 percent of p,p Â’-DDE in the diet after 30 days of exposure to 3.6 g / g feed (Muller 2003), which is similar to the accumulation rate of fathead minnows in this study. Thus, the doses in the low and medium groups of the dose response experiments are environmentally relevant, but the doses in the high group and the pilot study are probably out of the range of environmental relevance. In terms of population level parameters, su rvival probabilities were identical in adults in the dose response experiments because there were no mortalities at any dose. However, the number of eggs laid per fema le was considerably reduced in the high p,p Â’DDE group. This suggests that the no obs erved effect level (NOEL) under these experimental conditions is between 10 g / g and 100 g / g in the feed. Egg production in the high group did not differ from other groups until about day 14 of p,p Â’-DDE exposure. It was between days 7 and 14 that p,p Â’-DDE muscle tissue concentration in the high group increased rapidly from 7.26 0.46 to 8.54 6.82 g / g. Because there was no discernable difference in eggs per female before day 14 in the high group and

PAGE 59

48 reproduction was not affected in the medium group at any point during the study, it is likely that the NOEL is between the day 29 p,p Â’-DDE muscle concentration in the medium group (3.0 0.57 g / g) and the day 14 p,p Â’-DDE muscle tissue concentration in the high group (8.54 6.82 g / g). The percent of eggs that were fertilized and that hatche d were measured to isolate which sex might be responsible for any re productive dysfunction observed. The percent of eggs fertilized in the high group was much lower than any other group. This type of response have previously been observed in guppies exposed to 100 g p,p Â’-DDE / g feed, which had a marked decrease in sperm count (Bayley et al. 2002). Alternatively, fertilization success of male trout injected with p,p Â’-DDE was not altered. The percent of fertilized eggs that hatched in the low gr oup was also much lower than any other group, which indicates a problem with egg rather than sperm quality. Along with data from other studies, my results suggest there may be a problem with fertilization in the high group, in which males had significantly elevated estradiol. Without replication, however, these results are inconclusive. The pilot experiment demonstrated that p,p Â’-DDE impairs reproduction and does not have the same effect as the mammalian anti-androgen, flutamide, at similar doses in feed. Fathead minnows treated with p,p Â’-DDE stopped spawning afte r just three days of exposure, while those treated with flutamid e spawned at a relatively constant rate throughout the pre-exposure and exposure pe riods. In other studies, waterborne flutamide reduced fecundity of fathead mi nnows (Jensen et al. 2004). The route of exposure is a likely factor that contributed to the difference in effect of flutamide and p,p Â’-DDE on reproductive output: fish in this study were administered p,p Â’-DDE and

PAGE 60

49 flutamide via feed and fish other studies we re exposed to waterborne flutamide. The control group seemed to have a lower spaw ning rate during the final ten days of the experiment, but the rate of spawning in p,p Â’-DDE treated fish was zero throughout the final 25 days of exposure. Thus, p,p Â’-DDE caused reproductive impairment because exposed fish completely stopped spawning be fore the spawning rate of control fish declined. Plasma 17 -estradiol (estradiol) levels were measured in fathead minnows from the pilot experiment to investigate if p,p Â’-DDE acted by a similar mechanism as the antiandrogen flutamide. The high dose of p,p Â’-DDE in the pilot experiment (900 g p,p Â’DDE / g feed) decreased estrad iol in females, though not significantly, but flutamide significantly increased estradiol in males, the same result of a study where fathead minnows were exposed to waterborne flutamide (Jensen et al. 2004). p,p Â’-DDE did not affect estradiol in male fish in the pilot experiment, which had estradiol values almost identical to the control group. Despite the f act that males treate d with flutamide had plasma estradiol levels twice that of c ontrol, no effect on re productive output was observed. In other studies, reproductive out wa s adversely affected in fish treated with waterborne flutamide (Jensen et al. 2004). The first experiment proved that p,p Â’-DDE can cause reproductive dysfunction in fathead minnows, albeit at a re latively high dose. In that experiment estradiol levels were also considerably reduced in females treated with p,p Â’-DDE. At the lower and environmentally relevant levels of p,p Â’-DDE in the dose response experiment, female estradiol levels were unaffected and male plasma estradiol levels increased as a result of p,p Â’-DDE treatment. Similar to the result in males of the pilot study, trout injected with

PAGE 61

50 p,p Â’-DDE (30 g / g) did not have increased vi tellogenin, which is a response of exposure to estrogens (Donohoe and Curtis 1996). Thus, p,p Â’-DDE was not acting as an estrogen in that study. The dose in that study on trout was much greater than the concentration found in the ovary in fathead minnow administered 104 g p,p Â’-DDE / g feed in this study. The lack of an estrogenic response to a high dose of p,p Â’-DDE in trout coincides with no statistically significant chan ges in estradiol in the pilot experiment, where fish also received a very high dose (900 g / g feed). A compensatory mechanism may explain the difference in responses obser ved at very high versus environmentally relevant doses. Hormone receptors are common targets of endocrine disruptors (Young et al. 2005). In other studies, p,p Â’-DDE and flutamide had similar effects on mammalian hormone receptor binding (Kelce et al. 1995) and on secondary sex characteristics in guppies, a viviparous fish (Bayley et al. 2002) However, in studies on oviparous fish cells, including other cyprinids, fl utamide did not bind the AR, but p,p Â’-DDE did (Wells and Van Der Kraak 2000). Additionally, wh en investigating AR binding in brain, ovarian, and testicular tissues from goldfish ( Curassius auratus ) and rainbow trout ( Oncorhynchus mykiss ), Wells and Van der Kraak (2000) found that p,p Â’-DDE bound only to AR from goldfish testes. The differe nce in AR binding affinities among species, in conjunction with different cha nges in estradiol, suggests that p,p Â’-DDE and flutamide do not act solely by similar anti-androgenic mechanisms in fatheads minnows. In this experiment, estradiol levels in fish treated with p,p Â’-DDE or flutamide were not similarly altered, which contradicts findings of these two compounds in some other species. The

PAGE 62

51 data here corroborate evidence that an EDCÂ’s mode of acti on may differ greatly among species. Steroid synthetic and metabolic pathways are another possible target of the endocrine disruption caused by p,p Â’-DDE in this study. Although a very high dose of p,p Â’-DDE did not cause the same affect on repr oduction or estradiol as flutamide, fish responded to lower doses of p,p Â’-DDE in a manner consistent with flutamide in the pilot study: estradiol was elevated in males treated with flutamide and in males from all p,p Â’DDE groups in the dose response study. Thibaut and Porte (2004) found that p,p Â’-DDE (200 M) increased synthesis of maturation inducing hormones (MIH ), specifically pregnenolone, in carp ovar ian cells by increasing 20 -HSD activity. An overall increase in pregnenolone is likely to result in higher levels of hormones, as was seen in p,p Â’-DDE -treated fish in the dose response experiment because it is the s ubstrate from which sex steroid hormones are made. Largemouth bass exposed to p,p Â’-DDE under laboratory conditions (Muller, unpublished data) and in kelp bass ( Paralabrax clathratus ) from sites contaminated with DDTs had decreased ster oid hormone levels (Spies and Thomas 1997), which is the opposite re sponse of fathead minnows in this study. Again, this suggests species specific responses to EDCs. The fact that there was no change in estradiol in females at environmentally relevant doses, but there was a change in males indicates that there is a sex specific target of p,p Â’-DDE or that the target of p,p Â’-DDE is more sensitive in males. Two potential targets of p,p Â’-DDE that might cause increased estr adiol in males are P450aromatase and Phase II enzymes (e.g. UDP-glucuronosyltran sferase and sulfotranserfase). If P450aromatase activity is stimulated, more test osterone would be converted to estradiol.

PAGE 63

52 Alternatively, if Phase II enzymes were inhibi ted, estradiol would not be metabolized and subsequently excreted (Thibaut and Port e 2004). An increase in estradiol production would result in p,p Â’-DDE binding to factors th at inhibit the release of GtH I, which is a stimulant of estradiol and testosterone production (A rcand-Hoy and Benson 2001). Plasma GtH I was decreased in fish from a site contaminated with DDTs and PCBs (Spies and Thomas 1997), indicating th at chlorinated compounds such as p,p Â’-DDE, may be disrupting the endocrine system in the hypot halamus and pituitary, not just the steroid synthetic or metabolic pathways in the gonad. Speculating the target enzyme or factor in the HPG axis that increases estradiol in males is risky without those additional data on other hormones and enzyme activity. Gonadosomatic index is often used as an indicator of endocrine status in fish (Jensen et al. 2001; Mills et al. 2001; Noakss on et al. 2003). Gona dosomatic index did not differ significantly among treatme nt groups nor correlate with p,p Â’-DDE content or estradiol levels. However, both males and fe males in the high group of the dose response experiment showed a trend toward decreased GS I. Mills et al. (2001) found that none of the measures of endocrine status includi ng GSI, hepatosomatic index, estradiol, vitellogenin, and gonadal development changed in juvenile summer flounder ( Paralichthys dentatus ) exposed to p,p Â’-DDE intravenously. Ba yley et al. (2002) found similar results in guppies, where GSI was the only androgen controlled characteristic not affected by p,p Â’-DDE. Similarly, in this study GSI was not an indicator of adult exposure to p,p Â’-DDE. Effects of in ovo p,p Â’-DDE Exposure on Survival, Development, and Reproduction The concentration p,p Â’-DDE in fish was not changi ng significantly during the collection of eggs used for survival and develo pment studies. A relatively stable level of

PAGE 64

53 p,p Â’-DDE in adults was achieved by allowing adult fish to accumulate p,p Â’-DDE for at least 17 days before collecting eggs. Egg concentration of p,p Â’-DDE in the medium group (1.40 g / g) was comparable to total DDT s found in largemouth bass ovaries in site contaminated with p,p Â’-DDE in Florida (2.82 0.49 g / g; (Marburger et al. 2002). Survival, development, and reproduction we re monitored for nine months in eggs spawned from adults exposed to p,p Â’-DDE. Eighty-nine percent of eggs in the control group were successfully fertilized and 93% of those fertilized eggs hatched. The hatch rates in this study are compar able to that of control fish in other studies, where hatchability is greate r than 90 percent (Ankley et al. 2001) These rates indicate that fry rearing conditions were amenable to successful development. Survival probabilities are important para meters for projecting population growth rates (Caswell 2001). In this study, early survival of fathead minnows exposed to p,p Â’DDE in ovo was not directly dependent on dose. Fi sh in the low group had early survival probabilities identical to the control group. Because estradiol was not affected in females, one would not expect a change in offspring survival caused by altered egg quality, which is largely controlled by es trogen during oocyte development (Arcand-Hoy and Benson 2001). However, Cheek et al. ( 2000) found dose-dependent survival of eggs exposed to an estrogenic compound, which was not a result observed in the present study. It is imperative to understand differences in sexual maturity among groups because age at first reproduction can be an important determinant of population growth rate (Levin et al. 1996). Fish in the low group ha d identical survival and sexual development as the control group, though they were larger at four months of age. In Cheek et al. (2000), fish exposed to the estrogen o,p Â’-DDT during development were also

PAGE 65

54 significantly larger than contro l fish. However, density may have been an issue in that study because there were fewer fish in the DDT group than the control group, which was not normalized among treatment groups during fr y rearing. The time to first reproduction and appearance of secondary sex characteristics was delayed in fish exposed to in the high dose (104 g / g) of p,p Â’-DDE in ovo Kelce et al. (1995) showed pubertal rats exposed to p,p Â’-DDE had increased body weight a nd delayed puberty. This study on p,p Â’-DDE suggests that precociousness can resu lt from low dose exposure and delayed maturity from high dose exposure. Not only did fish in the high group have delayed sexual maturity, but male fish exposed in ovo also had abnormally high GSI when measured at nine months of age. To my knowledge, there are no reports of male GSI values as high as those of females, as observed in this study. This is especially notable given that ma les with drastically increased GSI were only exposed to p,p Â’-DDE in ovo and were allowed to develop in an uncontaminated environment. Rats exposed to a dose of 10-20 g p,p Â’-DDE / g as developing fetuses, which may be comparable to the high dose of this study, also had male reproductive abnormalities (Gray et al. 2001). This result implies that p,p Â’-DDE may cause early life stage changes of important factors, including ep igenetic factors and hormone homeostasis, that play a role in gonadal development. The high GSI values measured in the high group may be the result of intersex gonads, though no male gonads appeared to have follicles. If males were receiving estrogenic signals during development, they may develop intersex gonads and therefore, hist ological analyses on these fish should be conducted.

PAGE 66

55 Reproductive output of fish exposed in ovo varied among treatment groups. When measured as total egg output, the low group pr oduced more eggs over nine months than any other group. Measured as eggs per female, however, the medium group produced more eggs than all other groups. The disparity in these two measures can be attributed to a skewed sex ratio toward females in the lo w group and males in the medium group. The high group spawned considerably fewer eggs throughout the experiment and fewer eggs per female, a finding that cannot be attributed to a skewed sex ratio, which was close to 50:50 in the high group. Biomarkers A correlation matrix of average values of biomarkers and reproductive output was produced to identify potential relationships. Biomarkers and reproduc tive measures that were significantly correlated were then graphed for visual representation of the relationship because no statistical tests could be conducted on parameters for which there was no replication for each treatment group (i.e number of eggs per female, percent of eggs fertilized, etc.). Female GSI and estradiol of adults exposed to p,p Â’-DDE via feed were related to one another and the percent of eggs fertilized. There wa s no significant relationship between female GSI and estradiol when co rrelated using individual data, which is probably due to the fact that GSI remains relatively constant but estradiol changes throughout the spawning cycle of fathead mi nnows (Jensen et al. 2001). Additionally, the percent of eggs fertilized was also directly related to adult female GSI and estradiol. This is interesting because few studies inves tigate fertilization success, which was related to GSI and estradiol in this study. GSI and estradiol are typically related to egg output, which was not the case in this study (Jensen et al. 2001).

PAGE 67

56 Adult female GSI and estradiol were also in versely related to GSI of their offspring (i.e. eggs only exposed in ovo ). Mean female GSI of adults in the high group was lower than in all other groups, but GSI of their offs pring were much higher than other groups. Differences in GSI of both sexes in adults and their offspring were less obvious in the low and medium groups, but the inverse relationship held. Interestingly, adult male estradiol was also inversely related to the number of eggs per female in adults exposed to p,p Â’-DDE through feed. This may be due to effects of estradiol on male courtship behavior. In guppies, Bayley et al (2002) found that p,p Â’DDE did not affect male courtship behavior but it did affect male secondary sex characteristics and sperm count. Conclusions and Future Directions The US Food and Drug Administrati onÂ’s action level for issuing human consumption advisories in fish is 5 g p,p Â’-DDE / g muscle (Kennish and Ruppel 1996), a level that may cause reproductive dysfuncti on in adults according to this study. NOEL for reproductive dysfunction in adults and thei r offspring is observed at muscle tissue concentrations between 3.0 0.57 g / g and 8.54 6.82 g / g. Muscle concentrations as low as 0.38 0.05 g p,p Â’-DDE / g corresponded to marked changes in estradiol. In males, increased plasma estradiol was an indi cator of exposure, but did not increase with dose. Therefore, the level of increase in estradiol was not indicative of the p,p Â’-DDE concentration to which fish were exposed. In this experiment there was no replicati on of each treatment level. However, the total sample size of fish studied was simila r to that of many EDC assays in fathead minnows (Grist et al. 2003). E ffects on offspring are not as c onclusive as those on adults. Effects observed in fish exposed in ovo would be more robust with replication because

PAGE 68

57 differences in low and medium groups were s ubtle (e.g. female GSI) or without variation (e.g. egg production). This study has preliminarily linked biomar kers, such as adult GSI and offspring GSI, to population parameters such as viable eggs, which can be used to develop matrix population models. Modeling populations of tr eated fish would better define actual population outcomes that may result from p,p Â’-DDE treatment. The population level data collected in this study, including survival probabilities and fertil ities, should be used to project population growth rates under each treatm ent and/or combinations of treatments.

PAGE 69

58 APPENDIX HORMONE DETERMINATION BY RIA Plasma hormone concentrations of 17 -estradiol and testoste rone were determined by radioimmunoassay, as previously described (J ensen et al. 2001; Kagawa et al. 1981). The assays were optimized for the antibodies and 3H-hormones used. The assay is described as follows: plasma was diluted 1: 5 for females and 1:4 for males with 0.1 M PBS (pH 7.6). One-hundred fifty micr oliters 0.1 M PBS were added to 6 l of diluted plasma. The diluted plasma was extracted tw ice with 1.5 ml ethyl ether, the extract was evaporated, and reconstituted in 120 l assay buffer (0.01 M PBS with 1% BSA, pH 7.4). One hundred microliters of r econstituted extract or 17 -estradiol standard, 100 l estradiol antibody (Fitzg erald Industries International, Inc.; catalogue number 20-ER06; 1:16,000 final dilution), and 100 l of 0.05 Ci/ml 3H-17 -estradiol (Amersham Biosciences; catalogue number 125-250UCI) were added to a microcentrifuge tube and incubated for two hours at 25 C, then placed in an ice-water bath for 15 minutes. Fourhundred microliters of Dextran-coated charco al solution (1.5 g activ ated charcoal, 0.15 g Dextran (Sigma-Aldrich; catalogue number s C-3345 and D-4751, respectively), and 300 ml of 0.1 M PBS, pH 7.6) were added to each tu be and the tubes were returned to the icewater bath for another 15 minutes. Th e tubes were centrifuged at 3,000 rpm for 30 minutes at 4 C, 0.5 ml of the supernatant was placed into a scintillation vial to which 4.5 ml of scintillation fluid were added (Fis her Scientific, catalogue numbers SX18-4). Tritium was counted for two minutes for each sample on a scintillation counter (Beckman Coulter LS 6000 IC).

PAGE 70

59 The estradiol RIA method described above could not be validated. Values calculated were artificially hi gh. After several troubleshooting experiments as well as sending samples to another lab, I determined the antibody binds to multiple components in the extract. This may be remedied by us ing ether with higher purity. However, I believe the root of the problem lies with the specificity of the antibody.

PAGE 71

60 REFERENCES Ankley GT, Jensen KM, Kahl MD, Kort e JJ, Makynen EA. 2001. Description and evaluation of a short-term reproducti on test with the fathead minnow ( Pimephales promelas ). Environ Toxicol Chem 20(6):1276-1290. Anway MD, Cupp AS, Uzumcu M, Skinner MK. 2005. Epigenetic transgenerational actions of endocrine disruptors and ma le fertility. Science 308(5727):1466-1469. Arcand-Hoy LD, Benson WH. 2001. Toxic resp onses of the reproductive system. In: Taget Organ Toxicity in Ma rine and Freshwater Teleos ts (Schlenk D, Baatrup E, eds). New York:Taylor and Francis, 175-202. Barata C, Baird D, Soares A. 2002. Demogra phic responses of a tropical cladoceran to cadmium: Effects of food supply an d density. Ecol Appl 12(2):552-564. Bayley M, Junge M, Baatrup E. 2002. E xposure of juvenile guppies to three antiandrogens causes demasculinization and a reduced sperm count in adult males. Aquat Toxicol 56(4):227-239. Carlson DB, Curtis LR, Williams DE. 2000. Salmonid sexual development is not consistently altered by embryonic expos ure to endocrine-active chemicals. Environ Health Pers pect 108(3):249-255. Carr RL, Chambers JE. 2001. Toxic responses of the nervous system. In: Target Organ Toxicity Marine and Freshwater Tele osts (Schlenk D, Benson W, eds). New York:Taylor and Francis, 26-95. Caswell H. 2001. Matrix Population Models Sunderland:Sinauer Associates, Inc. Catalan J, Ventura M, Vives I, Grimalt JO 2004. The roles of food and water in the bioaccumulation of organochlorine com pounds in high mountain lake fish. Environ Sci Technol 38(16):4269-4275. Chedrese PJ, Feyles F. 2001. The diverse mechanism of action of dichlorodiphenyldich loroethylene (DDE) and methoxyc hlor in ovarian cells in vitro. Reprod Toxicol 15(6):693-698. Cheek A, Brouwer T, Carroll S, Manning S, McLachlan J, Brouwer M. 2001. Experimental evaluation of vitelloge nin as a predictive biomarker for reproductive disruption. Environ He alth Perspect 109(7):681-690.

PAGE 72

61 Collas P. 1998. Modulation of plasmid DNA me thylation and expression in zebrafish embryos. Nucleic Acids Res 26(19):4454-4461. Cook PM, Robbins JA, Endicott DD, Lodge KB Guiney PD, Walker MK, et al. 2003. Effects of aryl hydrocarbon receptor-mediat ed early life stage toxicity on lake trout populations in Lake Ontario duri ng the 20th century. Environ Sci Technol 37(17):3864-3877. Crellin NK, Rodway MR, Swan CL, Gillio-Meina C, Chedrese PJ. 1999. Dichlorodiphenyldichloroethylene potentia tes the effect of protein kinase A pathway activators on progesterone synthesi s in cultured porcine granulosa cells. Biol Reprod 61(4):1099-1103. Denslow ND, Chow M, Chow MM, Bono melli LCF, Heppell SA, Sullivan CV. 1997. Development of biomarkers for environm ental contaminants affecting fish. In: Chemically Induced Alterations in F unctional Development and Reproduction of Fishes (Rolland RM, Gilbertson M, Pete rson RE, eds). Pensacola, FL:SETAC Press, 73-86. Denslow ND, Chow MC, Kroll KJ, Green L. 1999. Vitellogenin as a biomarker for estrogen and estrogen mimi cs. Ecotoxicology 8:385-398. Ding LD. 2005. Vitellogenesis and Vitellogeni n Uptake into Oocytes. In: Hormones and their Receptors in Fish Reproduction (Melamed P, Sherwood N, eds). Hackensack, NJ:World Scientific Publishing Co. Pte. Ltd., 254-276. Donohoe R, Curtis LR. 1996. Es trogenic activity of chlord econe, o,p'-DDT, o,p'-DDE in juvenile rainbow tr out: induction of vite llogenesis and interaction with hepatic estrogen binding sites. Aquatic Toxicol 36:31-52. Filby AL, Tyler CR. 2005. Molecular characteri zation of estrogen receptors 1, 2a, and 2b and their tissue and ontogenic expression profiles in fathead minnow (Pimephales promelas). Biol Reprod 73(4):648-662. Foran C, Peterson B, Benson W. 2002. Influen ce of parental and developmental cadmium exposure on endocrine and reproductive f unction in Japanese medaka (Oryzias latipes). Comp Biochem Phys C 133(3):345-354. Foster EP, Fitzpatrick MS, Feist GW, Schr eck CB, Yates J. 2001. Gonad organochlorine concentrations and plasma steroid le vels in White sturgeon (Acipenser transmontanus) from the Columbia Rive r, USA. Bull Environ Contam Toxicol 67(2):239-245.

PAGE 73

62 Glesleichter J, Manire CA, Szabo NJ, Cort es E, Carlson J, Lombardi-Carlson L. 2005. Organochlorine concentrati ons in bonnethead sharks (Sphyrna tiburo) from Four Florida Estuaries. Arch Envi ron Contam Toxicol 48(4):474-483. Giesy J, Pierens S, Snyder E, Miles-Rich ardson S, Kramer V, Snyder S, et al. 2000. Effects of 4-nonylphenol on fecundity and bi omarkers of estrogenicity in fathead minnows (Pimephales promelas). Environ Toxicol Chem 19(5):1368-1377. Grau BL, Peterle TJ. 1979. Compartmentalizat ion of 14C-DDT in an experimental oldfield plot. Bull Environ C ontam Toxicol 22(6):846-853. Gray LE, Jr., Ostby J, Wilson V, Lambright C, Bobseine K, Hartig P, et al. 2002. Xenoendocrine disrupters-tiered screening and testing: filling key data gaps. Toxicology 181-182:371-382. Gray LE, Ostby J, Furr J, Wolf CJ, Lamb right C, Parks L, et al. 2001. Effects of environmental antiandrogens on reproductive development in experimental animals. Human Reproduction Update 7(3):248-264. Grist E, Wells N, Whitehouse P, Brighty G, Crane M. 2003. Estimating the effects of 17 alpha-ethinylestradiol on populat ions of the fathead minnow Pimephales promelas : Are conventional toxicological endpoints adequate? Environ Sci Technol 37(8):1609-1616. Guillette LJ, Jr., Gross TS, Gross DA, Rooney AA, Percival HF. 1995. Gonadal steroidogenesis in vitro from juvenile alli gators obtained from contaminated or control lakes. Environ Hea lth Perspect 103 Suppl 4:31-36. Guillette LJ, Jr., Gross TS, Masson GR, Matte r JM, Percival HF, Woodward AR. 1994. Developmental abnormalities of the gonad and abnormal sex hormone concentrations in juvenile alligators from contaminated and control lakes in Florida. Environ Health Perspect 102(8):680-688. Halm S, Kwon JY, Rand-Weaver M, Sumpter JP, Pounds N, Hutchinson TH, et al. 2003. Cloning and gene expression of P450 17a lpha-hydroxylase,17,20-lyase cDNA in the gonads and brain of the fathead minnow Pimephales promelas. Gen Comp Endocrinol 130(3):256-266. Heath R, Spann J, Kreitzer J. 1969. Marked DDE impairment of mallard reproduction in controlled studies. Nature 224(5214):47-&. Hornung MW, Jensen KM, Korte JJ, Kahl MD, Durhan EJ, Denny JS, et al. 2004. Mechanistic basis for estrogenic eff ects in fathead minnow (Pimephales promelas) following exposure to the androgen 17alpha-methyltestosterone: conversion of 17alpha-methyltestosterone to 17alpha-methylestradiol. Aquatic Toxicol 66(1):15-23.

PAGE 74

63 Hu M, Chiang E, Tong S, Lai W, Hsu N, Wang L, et al. 2001. Regulation of steroidogenesis in transgenic mice a nd zebrafish. Molecular and Cellular Endocrinol 171(1-2):9-14. Huang HJ, Liu SM, Kuo CE. 2001. Anaerobic biodegradation of DDT residues (DDT, DDD, and DDE) in estuarine sediment. J Environ Sci Health B 36(3):273-288. Jensen KM, Kahl MD, Makynen EA, Korte JJ, Leino RL, Butterworth BC, et al. 2004. Characterization of responses to the an tiandrogen flutamide in a short-term reproduction assay with the fathead minnow. Aquat Toxi col 70(2):99-110. Jensen KM, Korte JJ, Kahl MD, Pasha MS, Ankley GT. 2001. Aspects of basic reproductive biology and endocrinology in the fathead minnow ( Pimephales promelas ). Comp Biochem Physiol C 128(1):127-141. Kagawa H, Takano K, Nagahama Y. 1981. Corre lation of plasma estradiol-17 beta and progesterone levels with ultr astructure and histochemistry of ovarian follicles in the white-spotted char, Salvelinus le ucomaenis. Cell Tissue Res 218(2):315-329. Kalantzi OI, Alcock RE, Johnston PA, San tillo D, Stringer RL, Thomas GO, et al. 2001. The global distribution of PCBs and organoc hlorine pesticides in butter. Environ Sci Technol 35(6):1013-1018. Kannan K, Kajiwara N, Watanabe M, Nakata H, Thomas N, Stephenson M, et al. 2004. Profiles of polychlorinated biphenyl c ongeners, organochlorine pesticides, and butyltins in southern sea otters and their prey. Environ Toxicol Chem 23(1):4956. Kelce WR, Stone CR, Laws SC, Gray LE, Kemppainen JA, Wilson EM. 1995. Persistent DDT metabolite p,p'-DDE is a potent androgen receptor antagonist. Nature 375(6532):581-585. Kennish MJ, Ruppel BE. 1996. DDT contaminati on in selected estuarine and coastal marine finfish and shellfish of New Jersey. Arch Environ Contam Toxicol 31(2):256-262. Kime D. 1995. The effect of reproduction on fish. Rev Fish Biol Fisher 5(1):52-95. Korte J, Kahl M, Jensen K, Pasha M, Pa rks L, LeBlanc G, et al. 2000. Fathead minnow vitellogenin: Complementary DNA seque nce and messenger RNA and protein expression after 17 beta-est radiol treatment. Envir on Toxicol Chem 19(4):972981. LeBlanc GA, Bain LJ, Wilson VS. 1997. Pesticides: multiple mechanisms of demasculinization. Mol Ce ll Endocrinol 126(1):1-5.

PAGE 75

64 Levin L, Caswell H, Bridges T, DiBacco C, Cabrera D, Plaia G. 1996. Demographic responses of estuarine polychaetes to pollu tants: Life table re sponse experiments. Ecol Appl 6(4):1295-1313. Loomis AK, Thomas P. 2000. Effects of estrogens and xenoestrogens on androgen production by Atlantic croaker testes in vitro: evidence for a nongenomic action mediated by an estrogen membrane receptor. Biol Reprod 62(4):995-1004. Macek K. 1968. Reproduction in brook trout ( Salvelinus fontinalis ) fed sublethal concentrations of DDT. J Fish Res Board Can 25(9):1787-&. Makynen EA, Kahl MD, Jensen KM, Tietge JE Wells KL, Van Der Kraak G, et al. 2000. Effects of the mammalian antiandroge n vinclozolin on development and reproduction of the fathead minnow ( Pimephales promelas ). Aquatic Toxicol 48(4):461-475. Marburger J, Johnson W, Gross T, Douglas D, Di J. 2002. Residual organochlorine pesticides in soils and fish from wetland re storation areas in ce ntral Florida, USA. Wetlands 22(4):705-711. Matsui M, Kinuyama Y, Hakozaki M. 1974. B iliary metabolites of testosterone and testosterone glucosiduronate in the rat. Steroids 24(4):557-573. Matsumura F. 1985. Toxicology of Insecticides. New York:Plenum Press. Mauri M, Baraldi E, Simonini R. 2003. Eff ects of zinc exposure on the polychaete Dinophilus gyrociliatus: a life-table response experiment. Aquatic Toxicol 65(1):93-100. Mayer FL, Versteeg DJ, McKee MJ, Folmar LC, Graney RL, McCume DC, et al. 1992. Physiological and nonspecific biomar kers. In: Biomarkers: Biochemical, Physiological, and Histological Mark ers of Anothropogenic Stress (Huggett RL, Kimble RA, Mehrle PM, Bergman HL, eds). Boca Raton:Lewis, 5-86. Miglioranza KS, Aizpun de Moreno JE, Mo reno VJ. 2003. Dynamics of organochlorine pesticides in soils from a southeastern region of Argentina. Environ Toxicol Chem 22(4):712-717. Miller DH, Ankley GT. 2004. Modeling impacts on populations: fathead minnow ( Pimephales promelas ) exposure to the endocrine di sruptor 17beta-trenbolone as a case study. Ecotoxicol Environ Saf 59(1):1-9. Mills L, Chichester C. 2005. Review of evid ence: Are endocrine-dis rupting chemicals in the aquatic environment impacting fish populations? Sci To tal Environ 343(13):1-34.

PAGE 76

65 Mills LJ, Gutjahr-Gobell RE, Haebler RA, Horowitz DJ, Jayaraman S, Pruell RJ, et al. 2001. Effects of estrogenic (o,p'-DDT; oc tylphenol) and anti-androgenic (p,p'DDE) chemicals on indicators of endocrin e status in juvenile male summer flounder (Paralichthys dentatus). Aquat Toxico l 52(2):157-176. Moffat AC, ed. 1986. Clarke's isolation and id entification of durgs in pharmeceuticals, body fluids, and post-mortem material London:The Pharmeceutical Press. Monnosson E, Kelce WR, Mac M, Gray LE. 1997. Environmental antiandrogens: potential effects on fish reproduction and development. In: Chemically Induced Alterations in Functional Development and Reproduction of Fishes (Rolland RM, Gilbertson M, Peterson RE, eds). Pensacola, FL:SETAC Press, 53-60. Muller JK. 2003. An evaluation of dosing methods and effects of p,p '-DDE and dieldrin in Florida largemouth bass ( Micropterus salmiodes floridanus ). Gainesville, Fla.:University of Fl orida. Available: http://purl.fcla.edu/fcla/etd/UFE0001107 Nagahama Y, Yoshikuni M, Yamashita M, Tanaka M. 1994. Molecu lar endocrinology of fish. In: Fish Physiology (Sherwood N, Hew CL, eds). San Diego:Academic Press. Nash JP, Kime DE, Van der Ven LT, Wester PW, Brion F, Maack G, et al. 2004. Longterm exposure to environmental concentrations of the pharmaceutical ethynylestradiol causes reproductive failure in fish. Environ Health Perspect 112(17):1725-1733. Nimrod AC, Benson WH. 1997. Assessment of estr ogenic activity in fish. In: Chemically Induced Alterations in Functional De velopment and Reproduction of Fishes (Rolland RM, Gilbertson M, Peterson RE, eds). Pensacola, FL:SETAC Press. Noaksson E, Linderoth M, Bosveld AT, Ba lk L. 2003. Altered steroid metabolism in several teleost species exposed to endocr ine disrupting substances in refuse dump leachate. Gen Comp Endocrinol 134(3):273-284. Orlando EF, Denslow ND, Folmar LC, Guille tte LJ, Jr. 1999. A comparison of the reproductive physiology of largemouth bass, Micropterus salmoides collected from the Escambia and Blackwater Rivers in Florida. Environ Health Perspect 107(3):199-204. Panter GH, Thompson RS, Sumpter JP. 1998. Adverse reproductive effects in male fathead minnows exposed to environmenta lly relevant concentrations of the natural oestrogens, oestradiol, and oestrone. Aquatic Toxicol 42:243-253. Raimondo S, McKenney CL, Jr. 2005. Projected population-level effects of thiobencarb exposure on the mysid, Americamysis ba hia, and extinction probability in a concentration-decay exposure system. Environ Toxicol Chem 24(3):564-572.

PAGE 77

66 Rolland RM, Gilbertson M, Peterson RE, eds. 1997. Chemically Induced Alterations in Functional Develoment and Reproducti on of Fishes. Racine:SETAC Press. Segner H. 2005. Developmental, reproductive, and demographic alterations in aquatic wildlife: Establishing causality between exposure to endocri ne-active compounds (EACs) and Effects. Acta Hydr ochimica Hydrobiologia 33(1):17-26. Sepulveda M, Johnson W, Higman J, De nslow N, Schoeb T, Gross T. 2002. An evaluation of biomarkers of reproductive function and potential contaminant effects in Florida largemouth bass ( Micropterus salmoides floridanus ) sampled from the St. Johns River. Sci Total Environ 289(1-3):133-144. Sonderfan AJ, Arlotto MP, Parkinson A. 1989. Identification of the cytochrome P-450 isozymes responsible for testosterone oxidation in rat lung, kidney, and testis: evidence that cytochrome P-450a (P450IIA 1) is the physiologically important testosterone 7 alpha-hydr oxylase in rat testis. En docrinology 125(2):857-866. Spies RB, Thomas P. 1997. Reproductive and endoc rine status of female kelp bass from a contaminated site in the Southern Calif ornia Bight and estrogen receptor binding of DDTs. In: Chemically Induced Alte rations in Functional Development and Reproduction of Fishes (Rolland RM, Gilbertson M, Peterson RE, eds). Pensacola, FL:SETAC Press, 113-133. Thibaut R, Porte C. 2004. Effects of endocri ne disrupters on sex st eroid synthesis and metabolism pathways in fish. J Ster oid Biochem Mol Bi ol 92(5):485-494. Thompson CJ, Ross SM, Gaido KW. 2004. Di(nbutyl) phthalate impairs cholesterol transport and steroidogenesis in the fetal rat testis through a ra pid and reversible mechanism. Endocrinol 145(3):1227-1237. Valters K, Olsson A, Asplund L, Bergman A. 1999. Polychlorinated biphenyls and some pesticides in perch ( Perca fluviatilis ) from inland waters of Latvia. Chemosphere 38(9):2053-2064. Van der Kraak G, Chang JP, Janz DM. 1998. Reproduction. In: The Physiology of Fishes (Evans DH, ed). Boca Raton, FL:CRC Press, LLC, 465-490. von Hofsten J, Olsson PE. 2005. Zebrafish se x determination and differentiation: involvement of FTZ-F1 genes. Reprod Biol Endocrinol 3:63. Waxman DJ. 1988. Interactions of hepatic cy tochromes P-450 with steroid hormones. Regioselectivity and stereospecificity of steroid metabolism and hormonal regulation of rat P-450 enzyme expr ession. Biochem Pharmacol 37(1):71-84.

PAGE 78

67 Wells K, Van Der Kraak G. 2000. Differen tial binding of endogenous steroids and chemicals to androgen receptors in rai nbow trout and goldfish. Environ Toxicol Chem 19(8):2059-2065. Wilson VS, Cardon MC, Thornton J, Korte JJ, Ankley GT, Welch J, et al. 2004. Cloning and in vitro expression and characterizati on of the androgen rece ptor and isolation of estrogen receptor alpha from the fathead minnow ( Pimephales promelas ). Environ Sci Technol 38(23):6314-6321. Young G, Kusakabe M, Nakamura I, Lokman PM, Goetz FW. 2005. Gonadal Steroidogenesis in Teleost Fish. In: Hormones and Their Receptors in Fish Reproduction (Melamed P, Sherwood N, eds). Hackensack, NJ:World Scientific Publishing Co. Pte. Ltd., 155-223.

PAGE 79

68 BIOGRAPHICAL SKETCH Liza J. Ray was born to a farmer and swim instructor in Southern California where she lived until attending the Univ ersity of Michigan. While at Michigan, she studied and loved the wetland environments of that region. After being cold for four years, she was given the opportunity to enter graduate schoo l in the Interdisciplin ary Ecology program at the University of Florida School of Natural Resources and Environment. She is finally starting to overcome her fear of alligators ( Alligator mississippiensis ).