TOXICOLOGICAL SCIENCES 120(S1), S93–S108 (2011)

doi:10.1093/toxsci/kfq329

Advance Access publication October 25, 2010

Endocrine Disruption: Historical Perspectives and Its Impact on theFuture of Toxicology Testing

Mary Sue Marty,1 Edward W. Carney, and Justin Craig Rowlands

Toxicology and Environmental Research and Consulting, The Dow Chemical Company, Midland, Michigan

1To whom correspondence should be addressed at Toxicology and Environmental Research and Consulting, The Dow Chemical Company, 1803 Building,

Midland, MI 48674. Fax: (989) 638-9863. E-mail: mmarty@dow.com.

Received August 13, 2010; accepted October 15, 2010

The endocrine system is one of the body’s major

homeostatic control systems whose aim is to maintain normal

functions and development in the face of a constantly changing

environment. Working in tandem with the nervous system,

which mainly is responsible for rapid and immediate responses,

the approximately 30 different glands comprising the endocrine

system tend to act in a slower and more sustained manner to

regulate processes as diverse as the female reproductive cycle,

bone growth, cell proliferation, and psychosocial behavior.

Multiple endocrine glands also work in concert with one

another to form complex feedback loops, which tightly regulate

critical physiological processes.

Like all homeostatic control systems, the capacity to

maintain physiological parameters within normal bounds is

finite, and when this capacity is exceeded by chemical

exposures, drugs, or environmental stressors, adverse

consequences can ensue. In the mid-1990’s, concerns about

the potential for environmental chemicals, drugs, and other

stressors to alter endocrine physiology rapidly mounted and

received a great deal of attention within the toxicology

community as well as in the public media. These circumstances

spurred a flurry of research on mechanisms of toxicity, new

questions about how to deal with cumulative exposures to

endocrine-active compounds (EACs) from man-made, dietary,

and endogenous sources, and the creation of major chemical

screening and testing programs focused on endocrine-mediated

modes of toxic action.

In this review, we will trace the path of knowledge regarding

the effects of exogenous agents on the endocrine system, the

events which gave rise to the explosion of interest in this topic

in the mid-1990’s, and how it led to a fundamentally different

paradigm in toxicity testing as well as a great deal of basic

research on mechanisms of endocrine-mediated toxicity. We

first cover the history of testing for endocrine-mediated

toxicity, followed by coverage of key historical events in

basic science, and then bring the two together to discuss

implications of the science on testing both now and for the

future.

EARLY HISTORY OF TOXICITY TESTING FOR

ENDOCRINE-MEDIATED EFFECTS

A casual observer might easily be led to believe that the

ability of exogenous substances to interfere with the endocrine

system was unknown until relatively recently, but in fact, this is

far from true. Even the ancients were familiar with the actions

of various herbal preparations to modulate what we now know

to be hormonally regulated processes. For centuries, farmers

have observed reproductive problems in female sheep and

cows grazing on pastures rich in certain clover species, which

later were found to contain estrogenic compounds such as

coumestrol (Adams, 1995). As early as the 1930’s, the ability

of both natural and synthetic chemicals to interact with

endogenous hormone receptors was already well established

(Schueler, 1946; Sluczewski and Roth, 1948; Walker and

Janney, 1930).

Starting around the 1940’s, steroidal compounds began to be

used in the livestock industry to modulate reproductive cycles

and to enhance rate and efficiency of body weight gain, while

compounds such as ethinyl estradiol, mestranol, norethynodrel,

and megestrol acetate were being developed as contraceptive

agents and to treat specific disease conditions in humans

(Newburgh, 1975). In parallel with these developments,

toxicologists began to investigate these compounds for po-

tential adverse effects.

Diethylstilbestrol (DES), a ‘‘synthetic estrogen’’ with similar

potency to 17b-estradiol, was first prescribed in 1938 to

� The Author 2010. Published by Oxford University Press on behalf of the Society of Toxicology. All rights reserved.For permissions, please email: journals.permissions@oup.com

at University of N

airobi on Novem

ber 16, 2013http://toxsci.oxfordjournals.org/

Dow

nloaded from

prevent miscarriage and premature births. This drug was

particularly significant as the daughters of women who took

DES were found to have an increased incidence of clear-cell

carcinoma, a rare vaginal cancer. As a result, in 1971, the Food

and Drug Administration (FDA) advised physicians to stop

prescribing DES to pregnant women. Further research has

revealed an increased risk for breast cancer in DES mothers

and also possible links with reproductive tract abnormalities in

DES sons and daughters (Giusti et al., 1995). It has been

estimated that 5–10 million Americans received DES during

pregnancy or were exposed to the drug in utero.

Other episodes of endocrine-mediated toxicity also occurred

via unintentional exposure to hormonally active compounds,

such as polychlorinated biphenyls (PCBs) and polychlorinated

dibenzofurans (PCDFs). Some members of this large family of

structurally related congeners can act on the estrogen and

arylhydrocarbon (Ah) receptor systems and/or modulate

thyroid hormone function. In Japan (1968) and Taiwan

(1979), a few thousand women were exposed to PCBs and

their pyrolysis products (PCDFs) following consumption of

contaminated rice oil. The offspring of these mothers exposed

to PCBs and PCDFs tended to be smaller at birth and

exhibited delays in neurological development (Aoki, 2001).

Subsequent studies on PCB exposure at lower levels have

alleged other adverse effects, although these data remain

somewhat controversial.

In the environment, the presence of natural hormones in

wastewater treatment outfalls in the United States was noted as

early as 1965 (Stumm-Zollinger and Fair, 1965), and this work

expanded in 1970 to include synthetic estrogens used as birth

control agents (Tabak and Bunch, 1970). By the mid-1970’s,

scientists began to detect other pharmaceutical agents near

wastewater outlets in the United States (Garrison and Pope,

1975; Hignite and Azarnoff, 1977). Concerns about dichlor-

odiphenyltrichloroethane (DDT) in the environment were

highly publicized in Rachel Carson’s book, Silent Spring,

although a specific link to endocrine-mediated toxicity was yet

to come. Despite the aforementioned events in both wildlife

and humans, endocrine-mediated toxicity did not garner any

special attention above and beyond other mechanisms of

toxicity known at the time. Instead, it tended to be regarded as

just one of many potential mechanisms that could lead to

certain adverse outcomes, such as reproductive toxicity,

developmental toxicity, or cancer.

THE BIRTH OF ‘‘ENDOCRINE DISRUPTORS’’

This mindset changed rather abruptly in the mid-1990’s

following a series of publications suggesting that pesticides and

other man-made chemicals were threatening the reproductive

capability and intelligence of future generations of humans and

wildlife (Colborn et al., 1996, 1993; Hunter and Kelsey, 1993).

Of particular impact was a book entitled Our Stolen Future, Are

We Threatening Our Fertility, Intelligence and Survival?—A Scientific Detective Story written by Colborn et al. (1996)

with a forward penned by then Vice President Al Gore. These

and other authors proposed that many EACs elicited effects at

doses far lower than toxicities caused by other modes of

action (MOAs) and thus required special regulation. Also

around the same time were a number of more specific reports

alleging associations between endocrine-disrupting chemicals

and declining sperm counts (Carlsen et al., 1992; Sharpe and

Skakkebaek, 1993), increases in breast cancer (Davis et al.,1993), cryptorchism and hypospadias (Barlow et al., 1999),

and developmental abnormalities in wildlife species (Sumpter,

1995). The level of concern spiked even higher following a

1996 report claiming > 1000-fold level of synergy among

a mixture of four weakly estrogenic agents as assessed in a

yeast estrogen receptor (ER) transcriptional activation assay

(Arnold et al., 1996). Ironically, the latter results could not be

reproduced by others (Ashby et al., 1997; Ramamoorthy

et al., 1997) and a National Institutes of Health inquiry

revealed intentional falsification of the results leading to

retraction of the original report, which had been published in

Science (McLachlan, 1997).

Despite uncertainties about the scientific validity of these

reports and the true extent of the problem, both the scientific

community and public reacted as if this were an approaching

tsunami. Media coverage of what would previously have been

considered an esoteric science topic became mainstream

headline news. Prominent government and scientific organ-

izations such as the World Health Organization, International

Union of Pure and Applied Chemistry and the U.S. National

Research Council reacted by gathering endocrinologists,

toxicologists, and other public health professionals from

around the world to capture their knowledge of the state-of-

the-science and to recommend research to address the issue.

These efforts are documented in a series of comprehensive

review papers (IPCS, 2002; Lintelmann et al., 2003; NRC,

1999), which also helped spur a plethora of research on

endocrine-mediated toxicity. The IPCS (2002) report was

particularly significant in that it described a structured weight-

of-evidence approach to guide the application of new research

for addressing causal relationships between exposure to

EACs and adverse health outcomes. In fact, application of

weight-of-evidence approaches has failed to provide con-

vincing support for many of the early claims that initially

sparked the issue, such as alleged links between EACs and

declining sperm counts (Fisch, 2008) and breast cancer

(summarized at: http://epi.grants.cancer.gov/LIBCSP/ ). Re-

ported effects in aquatic species have pointed more toward

natural and synthetic hormones (e.g., birth control pills) as

causative agents, more so than industrial chemicals as

originally suggested (Jobling et al. 2006). Nonetheless, the

term endocrine disruptor has been adopted in the vernacular

of toxicologists, public health professionals, and even the

general public and appears here to stay.

S94 MARTY, CARNEY, AND ROWLANDS

at University of N

airobi on Novem

ber 16, 2013http://toxsci.oxfordjournals.org/

Dow

nloaded from

Somewhat unwittingly, the new emphasis on endocrine

disruption also marked the beginning of a fundamental change

away from the traditional ‘‘outside-in’’ approach to toxicology,

in which adverse outcomes are identified first followed by

mechanistic research to understand these effects. With

endocrine disruption came a reversed approach in which

identification of mechanism came first followed by research to

define possible consequences of activating these mechanisms.

The implications of this change are quite significant and are

discussed later in this review.

GENESIS OF ENDOCRINE SCREENING AND TESTING

At around the same time that these alarming reports were

emerging, the U.S. Safe Drinking Water Act was being updated

while the Food Quality Protection Act was being created to

enhance protection of children and other vulnerable subpopula-

tions. This temporal convergence led to a 1996 congressional

mandate requiring the Environmental Protection Agency (EPA) to

implement a screening and testing program to detect endocrine

disruptors. Despite the nascent state of the science at the time,

Congress gave EPA just 2 years to establish the new program. To

assist EPA, an advisory committee known as the Endocrine

Disruptor Screening and Testing Advisory Committee (EDSTAC)

was formed to evaluate protocols for approximately 50 different

endocrine assays and select a subset to comprise a ‘‘Tier I’’

screening battery for biological activity involving the estrogen,

androgen, and thyroid (EAT) hormone systems.

Although limiting screening to these three hormone systems

might seem arbitrary, these three appeared to have the broadest

range of potential effects as well as the greatest number of

ligands with which they interact (Blair et al., 2000).

Compounds that can bind steroid hormone receptors include

industrial chemicals such as bisphenol A (BPA), certain

alkylphenol ethoxylates, phthalates, parabens, benzophenones,

and environmental contaminants like DDT, methoxychlor,

some PCBs, dioxins and furans. There also are many

pharmaceuticals purposely designed to have hormonal activity,

such as DES, contraceptive agents, and the selective ER-

modulators (SERMs) used in the treatment of diseases such as

osteoporosis, as well as numerous plant-derived compounds

such as genistein, daidzein, naringenin, genistin, daidzin,

glycitein, and puerarin (Boue et al., 2003) and resveratrol, a

constituent of red wine (Klinge et al., 2003). Finally, exposure

to synthetic and plant-derived estrogens occurs against a

background of endogenous estrogens, such as estradiol-17b(E2), estrone (E1), and estriol (E3), the levels of which in

blood and tissues can vary dramatically depending on gender,

age, and physiological state (e.g., pregnancy, pre-, or post-

menopause).

Beyond need to cover so many different chemistries and

sources, the challenge laid before EDSTAC was further

complicated by the desire to screen for both receptor agonists

and antagonists, as well as compounds which act indirectly.

Indirect mechanisms are plentiful and include the inhibition of

steroid biosynthesis or modulation of the hypothalamic-

pituitary-gonadal (HPG) axis. An even more complex example

begins with activation of the Ah receptor (AhR), which can

ultimately oppose estrogen signaling.

Ironically, some compounds bearing the very properties

that are the targets of screening can actually be beneficial. Studies

of endocrine-disrupting AhR ligands have led to the development

of selective AhR modulators, such as ring-substituted

diindolylmethanes (DIMs), for chemotherapy of breast,

endometrial, pancreatic, and prostate cancer (Safe et al.,2008). Substituted furans and indoles have been licensed by a

pharmaceutical company as antitumor development compounds

(Safe, personal communication). Some of the DIMs have been

further modified into anticarcinogenic AhR-inactive analogs,

which activate orphan nuclear receptors that induce genes

associated with growth inhibition and cancer cell death.

EPA’S ENDOCRINE DISRUPTOR SCREENING PROGRAM

TIER I ASSAYS

In 1998, EDSTAC released its report outlining a screening

and testing program to assess the potential of compounds to

interact with the endocrine system. EDSTAC recommended

that a tiered approach be adopted to examine endocrine activity

with a focus on the potential for compounds to interact with the

EAT hormone systems. The Tier I assays were designed to

screen for potential interaction with these endocrine systems

and designed to be an efficient screening process to screen

a large number of chemicals for endocrine activity. In Tier I,

EDSTAC opted to prioritize assays with greater sensitivity (but

possibly lower specificity) to avoid false-negative results.

EDSTAC described the importance of using a ‘‘weight-of-

evidence’’ approach to evaluate Tier I screening results. Tier II

tests, such as the multigeneration reproductive toxicity study,

were to be used to characterize dose-response relationships,

whether endocrine-mediated adverse effects for EAT occur, as

well as providing data for risk assessment. During this time

period, EPA also took on an extensive revision of the

multigeneration and some other Tier II tests in order to

enhance their sensitivity to endocrine disruptors.

The Tier I assays included in vitro, in vivo mammalian, and

in vivo nonmammalian screening assays. EDSTAC noted that

EACs may alter endocrine function by affecting the availability

of a hormone at a target tissue (e.g., synthesis, transport,

metabolism, HPG/hypothalamic-pituitary-thyroid [HPT] axis

function, etc.) or by affecting the cellular response to a hormone

(e.g., receptor binding). The original Tier I battery proposed by

EDSTAC (EDSTAC, 1998) included the following: (1) ER

binding/transactivation assay, (2) androgen receptor (AR)

binding/transactivation assay, (3) steroidogenesis assay with

minced testis, (4) rodent 3-day uterotrophic assay (sc route),

ENDOCRINE-MEDIATED TOXICITY S95

at University of N

airobi on Novem

ber 16, 2013http://toxsci.oxfordjournals.org/

Dow

nloaded from

(5) rodent 20-day pubertal female assay with thyroid, (6) rodent

5- to 7-day Hershberger assay, (7) frog metamorphosis assay,

and (8) fish gonadal recrudescence assay. In addition, four

alternate assays were identified: (1) placental aromatase assay,

(2) modified rodent 3-day uterotrophic (ip), (3) rodent 15-day

intact adult male assay with thyroid, and (4) rodent 20-day

thyroid/pubertal male assay. These alternative assays, if

validated, could be used to replace other Tier I assays, namely,

either the intact adult male assay or the pubertal male assay

would replace the pubertal female and Hershberger assays.

Per the Food Quality Protection Act (1996), the endocrine

screening program was required to use ‘‘appropriate validated

test systems’’; consequently, the EPA launched the endocrine

disruptor screening program (EDSP) validation program.

Based on results of methods development and validation work,

some of the proposed Tier I in vitro assays were adopted,

whereas others have undergone some modifications. The final

list of EPA Tier I EDSP screening assays appears in Table 1.

Both the ER binding and ER transactivation assays have been

adopted into Tier I. The ER binding assay measures the ability

of a radiolabeled ligand ([3H]-17b-estradiol) to bind to rat

uterine cytosolic ER in the presence of increasing test material

concentrations. A compound is positive for ER binding if it

effectively competes with [3H]-estradiol binding in a concentration-

related manner, and an inhibitory concentration that decreases

radioligand binding by 50% (IC50) is obtained. The ER

competitive binding assay cannot distinguish agonists from

antagonists. Test guidelines also were issued for the Stably

Transfected ER Transcriptional Activation assay using HeLa-

9903 cells (Organization for Economic Co-operation and

Development [OECD] TG 455; OPPTS 890.1300), although

detection of agonists is all that is required. Although having

limited metabolic capability, the ER transactivation assay

reportedly has good concordance with in vitro ER binding and

in vivo uterotrophic responses for active parent compounds

(Takeyoshi, 2006); however, the test guideline criteria for

a positive response (i.e., if a compound achieves a PC10, which

is 10% of the positive control response at 1nM 17b-estradiol)

may result in numerous false-positive results and jeopardize the

predictiveness of this assay.

AR binding using rat prostate cytosolic preparations also is

included in Tier I (OPPTS 890.1150), although an AR

transactivation assay for use in Tier I screening is still being

developed. The AR binding assay shares many of the same

advantages and limitations as the ER binding assay.

To detect compounds affecting steroidogenesis, the minced

testis assay was originally proposed. This assay was designed

to detect enzyme inhibition in the steroid biosynthesis pathway

by measuring decreased testosterone production; however, the

assay was removed from consideration because of high false-

positive and false-negative responses and difficulty assessing

Leydig cell cytotoxicity in the mixed population of cells

present in minced testis preparations (Powlin et al., 1998). The

steroidogenesis assay was replaced with a cell-based system

that uses a human cell line, the H295R cells (Hecker et al.,2006, 2008). This assay is designed to measure production of

testosterone and 17b-estradiol by the H295R cells in the

presence of varying concentrations of test compound to

identify alterations in steroid biosynthesis. Because of the

limited metabolic capacity of these cells, the assay primarily

examines the ability of the parent material to interfere with

steroidogenesis. A positive result occurs if there is a significant

difference in steroid production in the presence of less than

20% cytotoxicity.

Aside from steroidogenesis alterations, a separate in vitroassay was developed specifically to identify compounds

affecting aromatase (CYP19), the enzyme that converts

androgens to estrogens. The aromatase assay examines effects

of the parent compound on recombinant human aromatase

activity by measuring the production of heavy water (3H2O)

during the conversion of 3H-androstenedione to estrone.

Compounds that competitively inhibit aromatase activity will

decrease estrone production and consequently 3H2O levels in

a dose-related manner.

Interlaboratory validation has been conducted with these

Tier I in vitro assays to verify each assay’s ability to detect

known EACs operating via relevant MOAs, although the size

of the validation programs (i.e., number of test compounds,

number of laboratories, etc.) has differed by assay. One issue

with the in vitro EDSP assays is that the assays have limited

ability to metabolize substances; thus, if a metabolite is the

EAC, it may not be detected by these assays. Options to

improve metabolic capacity are under consideration. Although

in vitro assays provide useful information on potential MOA

and can help clarify in vivo assay results, previous studies have

suggested that in vivo data should carry more weight than

in vitro data when determining the need for Tier II testing (e.g.,

Charles et al., 2005).

The uterotrophic assay went through an extensive validation

program, which was led by the OECD Endocrine Disruptor

TABLE 1

U.S. EPA EDSP Tier I Assays

In vitro assays

ER binding—rat uterine cytosol

ER—(hERa) transcriptional activation—human cell line (HeLa-9903)

AR binding—rat prostate cytosol

Steroidogenesis—human cell line (H295R)

Aromatase—human recombinant microsomes

In vivo assays

Uterotrophic (rat)

Hershberger (rat)

Pubertal female (rat)

Pubertal male (rat)

Amphibian metamorphosis (frog)

Fish short-term reproduction

Note.http://www.epa.gov/scipoly/oscpendo/pubs/assayvalidation/

tier1battery.htm#assays.

S96 MARTY, CARNEY, AND ROWLANDS

at University of N

airobi on Novem

ber 16, 2013http://toxsci.oxfordjournals.org/

Dow

nloaded from

Testing and Assessment (EDTA) Task Force and the

Validation Management Group (VMG)—Mammalian. This

validation process, led by Japanese representatives on the

EDTA, included a detailed background document prepared by

the OECD (2003) and multiphase intra- and interlaboratory

validation studies involving approximately 20 laboratories to

show transferability, reproducibility, and reliability. This

validation program tested a potent estrogen (ethinyl estradiol),

weak estrogen agonists in both a coded (blinded) and dose-

response manner to evaluate sensitivity, and negative

compounds to verify specificity. This work culminated in

a series of publications (Ashby et al., 2003; Kanno et al.,2001, 2003a,b; Owens and Koeter, 2003; Owens et al., 2003)

and an OECD test guideline in 2007. The U.S. EPA test

guideline was released in 2009.

The assay uses female rats (or mice) with nonfunctional

HPG axis to yield sensitive and consistent increases in uterine

weight in response to stimulation by estrogenic materials.

Ultimately, the test guideline outlines the application of the

uterotrophic assay for the detection of compounds with

estrogenic activity; use of this assay to identify antiestrogens

is supported by the test guideline, but not required. Overall, the

uterotrophic assay appears to perform well for the identification

of estrogenic compounds, although the ovariectomized model

has been recognized as more specific for identification of

estrogenic activity than the immature model. The immature

model, which has an intact HPG axis, can detect aromatizable

androgens and is more sensitive to body weight–mediated

changes in uterine weights. The alternate uterotrophic assay,

which used ip dosing, was not included in the validation

because of potential localized effects of the test compounds on

uterine tissue.

The Hershberger assay also underwent an extensive multi-

laboratory validation process, led by U.S. EPA through the

OECD EDTA VMG—mammalian group (Owens et al., 2006,

2007). The Hershberger assay uses castrated adult male rats

with a nonfunctional HPG axis. Although originally proposed

with a 5- to 7-day exposure period, rats are treated with the

test compound for 10 days by oral gavage or sc dosing in

the adopted version of the assay. Increases in accessory sex

tissue (AST) weights (i.e., glans penis, ventral prostate,

seminal vesicles with fluid, Cowper’s glands, and levator

ani-bulbocavernosus muscles) indicate potential androgenicity.

To evaluate antiandrogenic potential, rats are given the

test compound (oral gavage or sc) concurrent with 0.2 or

0.4 mg/kg/day testosterone propionate sc. If the test compound

inhibits the testosterone-induced increase in AST weights, then

the compound has potential antiandrogenic activity or is

a potential 5a-reductase inhibitor. Two or more organ weights

must be significantly altered to yield a positive result. Overall,

the Hershberger assay appears to perform well for the

identification of potential androgens, antiandrogens, and

5a-reductase inhibitors. During the Hershberger validation

studies, a weanling (non-castrated) animal model was included;

however, the weanling model did not detect weak antiandrogens

as consistently as the castrated adult model (Freyberger and

Schladt, 2009) and therefore was not included in either the EPA

or the OECD test guideline.

Initially, the 20-day female pubertal assay was recommen-

ded, and the 20-day male pubertal assay was an alternate EDSP

screen. Prior to initiation of the interlaboratory validation work,

the exposure period for the pubertal male assay was increased

from 20 to 30 days, and ultimately, both assays were adopted

into the Tier I EDSP battery. One of the strengths of the male

and female pubertal assays is that these assays look for changes

in young animals during a dynamic period of endocrine

activity. The pubertal assays are the most apical of the Tier I

in vivo mammalian assays, but these assays are multimodal,

capable of detecting EACs that operate through a variety of

MOAs (i.e., strong and weak (anti)estrogenic and (anti)andro-

genic compounds, steroid biosynthesis inhibitors, aromatase

inhibitors, 5a-reductase inhibitors, thyroid-active agents, and

compounds affecting the HPG or HPT axis). Some of these

MOAs are not detected by other Tier I in vitro or mammalian

assays; thus, it may be difficult to dismiss a positive result in

the pubertal assays even if there is some suspicion that the

primary MOA is not endocrine mediated.

One limitation of the pubertal assays is that it may be

difficult to determine the MOA for potential EACs. Some

endpoints assessed in these assays (puberty onset, estrous

cycle, and organ weights) can be altered by chemicals that

operate via an endocrine MOA or by nonspecific/systemic

toxicity. During the validation program, some compounds were

not identified for their primary MOA. For example, phenobar-

bital, which is a weak thyroid-active compound, did not alter

thyroid weights, thyroid histopathology, or serum T4 or thyroid

stimulating hormone levels in the male pubertal assay during

the validation program but rather showed signs of antiandro-

genicity, which included delayed preputial separation and

decreased reproductive and AST weights (U.S. EPA, 2007).

There remains some question regarding the specificity of the

pubertal assays. The Interagency Coordinating Committee on

the Validation of Alternative Methods defines specificity as the

proportion of inactive substances that are correctly identified.

During assay validation, 2-chloronitrobenzene (2-CNB) was

selected as a negative control chemical for the male and female

pubertal assays; however, 2-CNB caused thyroid effects,

delayed vaginal opening, and altered adrenal and uterine

weights in females, and delayed preputial separation, decreased

serum testosterone levels, decreased androgen-dependent organ

weights, and had testis histopathological effects in males. The

EPA selected hydroxyatrazine as an alternate negative control

compound and recently reported that it was negative in both the

male and the female pubertal assays (Stoker and Zorrilla,

2010). Only a subset of these data has been published. Once

available, the full data set is likely to provide additional

information on setting maximum tolerated dose levels and

interpreting pubertal assay results (i.e., provide perspective for

ENDOCRINE-MEDIATED TOXICITY S97

at University of N

airobi on Novem

ber 16, 2013http://toxsci.oxfordjournals.org/

Dow

nloaded from

hydroxyatrazine results reported in Laws et al., 2003). With

few negative compounds included in the pubertal assay

validation, it is unclear whether the male and female pubertal

assays are prone to nonspecific effects, which will produce

a high frequency of false-positive results.

One alternative assay proposed as part of the EDSP was the

15-day intact adult male rat assay. Approximately 30 EACs,

across a variety of MOAs, have been evaluated using the

15-day intact male assay including estrogen agonists/antagonists,

androgen agonists/antagonists, steroid biosynthesis inhibitors,

thyroid-active compounds, and prolactin modulators. The

validation program also examined assay specificity by evaluating

the impact of body weight effects on assay endpoints (O’Connor

et al., 2000) and including a negative control compound, allyl

alcohol (O’Connor et al., 2008). The intact male assay was

specifically designed as an MOA screening study, whereby

researchers could compare the ‘‘fingerprint’’ of the organ weight,

histopathology, and hormonal endpoints for an unknown

compound with data generated for a series of positive controls

to help characterize the MOA (O’Connor et al., 2002). There was

some concern regarding the use of numerous serum hormone

measurements in this assay as serum hormone levels are

inherently variable. However, data developed during the

validation process have shown that serum hormone measure-

ments are reliable with the group sizes employed (n ¼ 15 per

dose). Of greater concern for the U.S. EPA was the inability of

the intact male assay to reliably detect weak antiandrogens

(O’Connor et al., 1999), which would require that the assay be

paired with an alternative assay (e.g., Hershberger assay) for Tier

I screening. Because of this weakness, the 15-day intact male

assay was not included in the EPA EDSP.

In addition to the in vitro and in vivo mammalian assays, two

nonmammalian assays also are included in the Tier I EDSP:

the amphibian metamorphosis assay and the 21-day fish

reproduction screen. The amphibian metamorphosis assay

(OPPTS 890.1100) provides an environmental screen to detect

compounds that have the potential to interfere with the HPT

axis. Xenopus laevis tadpoles are exposed for 21 days starting

at Nieuwkoop and Faber stage 51 (Nieuwkoop and Faber,

1994) and examined for mortality, developmental stage, hind

limb length, snout-vent length, wet weight, and thyroid

histopathology. Because the rate of metamorphic development

is controlled by thyroid hormone in X. laevis, developmental

stage (which is determined by the inspection of various

morphological changes associated with metamorphosis) and

hind limb length (which is normalized by snout-vent length) are

used to determine if development is accelerated, asynchronous,

delayed, or unaffected by the test compound after either 7 or

21 days of exposure. Many of the endpoints assessed in this

assay are apical and can be altered by MOAs that are not

specific for thyroid effects. Some endpoints (e.g., snout-vent

length and wet weight) provide information on whether the

test compound alters growth and/or causes systemic toxicity.

If any changes in developmental stage are observed

(i.e., advanced, delayed, or asynchronous development), thyroid

histopathology is an important endpoint to confirm a specific

thyroid effect. To facilitate interpretation, laboratories are

encouraged to include positive control compounds to verify

assay performance and build historical data. Thus, the

amphibian metamorphosis data should be considered in

a weight-of-evidence approach to determine potential thyroid

activity.

Originally, the fish gonadal recrudescence assay was

proposed as an EDSP Tier I screening assay. In this assay,

fish were examined for compound effects on maturation from

a regressed state (recrudescence) in response to an increased

photoperiod and temperature regime. Endpoints included

secondary sex characteristics, ovary and testis weight increases,

gonadosomatic index, gamete maturation, and induction of

vitellogenin. However, feasibility studies indicated that there

was high variability in the gonadal recrudescence response in

continuous spawning species like the fathead minnow (Touart,

2005); thus, the recrudescence assay was not included in the

EDSP. Instead, the U.S. EPA has opted for the 21-day fish

reproduction screen (OPPTS 890.1350) in the Tier I battery as

opposed to the OECD 21-day fish assay (OECD TG 230). This

screen uses a minimum of 96 reproductively mature adult

fathead minnows (alternate models allowed per OECD)

exposed continuously for 21 days to a minimum of three test

concentrations and a control. The EPA 21-day fish reproduc-

tion screen includes additional parameters not measured in the

OECD 21-day fish screen, including fecundity, fertility,

gonadal histopathology, and sex steroid concentrations. In

addition, adult survival, secondary sex characteristics (such as

coloration patterns and tubercle scores), reproductive behavior,

vitellogenin, and gonadal-somatic index (GSI) are evaluated.

Secondary sex characteristics, vitellogenin, and plasma sex

steroids are more predictive of an endocrine MOA, although

the high variability seen in plasma sex steroids makes them less

robust and a lower priority than vitellogenin when blood

plasma quantities are limited. In contrast, fecundity is an

indicator of general reproductive condition and not endocrine

specific. Thus, evaluations in the fish reproduction screen are

a combination of endocrine-specific endpoints and apical

endpoints that can be altered by non-endocrine MOAs. With

these apical endpoints and the variability in some of the

observations, weight of the evidence will be employed in

determining the potential of the compound to interact with the

endocrine system of fathead minnow.

Both environmental screens must manage issues related to

the physicochemical properties of various test materials, the

need to develop a solvent-free test system (if possible),

development of analytical methods for dose verification, and

the need to generate range-finding data to assist in dose

selection. Appropriate dose setting is critical to differentiate

systemic toxicity from genuine endocrine-mediated effects.

In 2009, the U.S. EPA released its first list of priority

compounds for endocrine screening. The initial priority list

S98 MARTY, CARNEY, AND ROWLANDS

at University of N

airobi on Novem

ber 16, 2013http://toxsci.oxfordjournals.org/

Dow

nloaded from

consisted of 67 chemicals, primarily pesticide-active ingre-

dients (58) and some pesticide inerts (9). For pesticide-active

ingredients, substantial hazard data are available because of

registration requirements. As part of the EPA EDSP program,

companies were allowed to submit other scientifically relevant

information (OSRI), describing available toxicity data for

a compound and known health risks with an emphasis on

endocrine-related effects. OSRI information for many of the

priority compounds is currently under review by the EPA;

however, it is not clear what information will be viewed as

‘‘functionally equivalent’’ to the Tier I assays. If functionally

equivalent data are available for a compound, companies may

be allowed to omit one or more of the Tier I EDSP assays

based on these existing data.

The goal of the Tier I EDSP is to identify substances that have

the potential to interact with the estrogen, androgen, or thyroid

hormone system for further testing in Tier II. Given that there are

apical endpoints in Tier I assays that may be altered by nonspecific

effects and there is some redundancy to detect endocrine MOAs

across different Tier I assays, it is beneficial to use a weight-of-

evidence approach to determine whether there is sufficient data to

warrant Tier II testing (Fig. 1). Previous toxicity data generated

with standardized test guidelines and/or validated methods should

be included in this weight-of-evidence analysis.

One example of applying weight of evidence can be

illustrated with methoxychlor, an estrogenic compound

included in validation studies for several of the Tier I EDSP

screening assays. It is recognized that methoxychlor is

metabolized in vivo to an estrogenic metabolite, 2,2-bis

(p-hydroxyphenyl)-1,1,1-trichloroethane (HPTE). Thus, me-

thoxychlor was not included as part of the estrogen binding

assay validation work; however, other binding assays using

competitive binding to uterine cytosol ERs were negative for

methoxychlor (e.g., Shelby et al., 1996). Methoxychlor was

included in the ER transactivation assay validation, wherein its

estimated EC50 was 1 3 10�5M. The authors concluded that

methoxychlor exhibited a response curve that was consistent

with other weak estrogens. Methoxychlor was positive in some

AR binding assays (Charles et al., 2005), but it was negative in

the aromatase assay. In vivo, methoxychlor was positive in the

uterotrophic assay (Kanno et al., 2003a,b), although it was

more readily detected with oral exposure than sc injection and

sc injection is the preferred route for administration in the EPA

uterotrophic test guideline. Methoxychlor was negative in the

Hershberger assay at doses that caused a significant increase in

liver weights (Charles et al., 2005) and positive in both the

male and the female pubertal assays. The estrogenic properties

of methoxychlor also were identified in the fish reproduc-

tion assay. There were no data on methoxychlor’s effects in

the steroidogenesis or amphibian metamorphosis assays. Based

on a weight-of-evidence approach, methoxychlor would be

identified as an estrogenic compound with possibly a mixed

mode of action that affects androgen-dependent endpoints.

Except where indicated by other citations, results of these

validation studies can be located online (http://www.epa.gov/

scipoly/oscpendo/).

Based on a weight-of-evidence approach, compounds

deemed positive in Tier I will undergo Tier II testing to

characterize hazards and develop data for risk assessment. The

specific tests that will be included in Tier II have not been

definitively identified. Possible tests under consideration for

inclusion in Tier II include the two-generation mammalian (rat)

reproductive toxicity study, the two-generation amphibian

(frog) reproductive toxicity study, the two-generation avian

reproductive toxicity study, fish life cycle toxicity study, and

the invertebrate (mysid) life cycle toxicity study.

LOW-DOSE ISSUE AND MIXTURES

Concern over endocrine disruptors has been further

heightened because of reports that such compounds can elicit

unique low-dose effects and exhibit nonmonotonic dose-

response relationships that may not be detected in regulatory

toxicity studies, which are typically conducted at higher dose

levels. The low-dose hypothesis has been a contentious issue

FIG. 1. This diagram illustrates some of the redundancy in Tier I EDSP

assays. In a weight-of-evidence approach, a pattern of effects across related

assays is easily interpreted (e.g., positive ER binding, ER transactivation, and

uterotrophic assays coupled with estrogenic responses in the pubertal and fish

short-term reproduction assays). However, in many cases, the pattern will not

be as easily discerned. If results of in vitro and in vivo assays disagree, weight

of evidence requires careful consideration of the in vivo results and previous

toxicity data when determining whether the compound is a potential EAC and

warrants Tier II testing. Aside from the expected MOA relationships depicted

in this figure, other, more complex MOA relationships are not shown (e.g., the

ability to detect agents with mixed MOAs like methoxychlor or HPG axis

compounds like atrazine).

ENDOCRINE-MEDIATED TOXICITY S99

at University of N

airobi on Novem

ber 16, 2013http://toxsci.oxfordjournals.org/

Dow

nloaded from

among some toxicologists. The National Toxicology Program

(NTP), with support from the U.S. EPA, organized a panel to

evaluate the scientific evidence on the low-dose effects of

endocrine disruptors (Melnick et al., 2002), defining low-dose

effects as ‘‘biologic changes that occur in the range of human

exposures or at doses lower than those typically used in the

standard testing paradigm of the U.S. EPA for evaluating

reproductive and developmental toxicity.’’ The panel

concluded that (1) there was evidence for low-dose effects in

laboratory animals, although in some cases, these effects have

not been replicated; (2) the shape of the dose-response curve

depended on the endpoint and dosing regimen used; and (3) the

current reproductive/developmental toxicity testing paradigm

should be reviewed and possibly revised with respect to dose

selection, animal models used, ages evaluated, and endpoints

measured. Some reasons identified for the inability to

reproduce low-dose effects across studies included intrauterine

position, strain and substrain differences in response, dietary

levels of phytoestrogens or differences in caloric intake,

differences in housing conditions (caging, bedding, etc.), and

seasonal variation. The statistics subgroup (Haseman et al.,2001) also identified a number of important factors that should

be considered in the design, analysis, and interpretation of

experimental studies to help resolve low-dose issues, including

study sensitivity (power), reproducibility, control for litter

effect, quality control in data collection and management,

procedures for data analysis, interpretation of the significance

of biological changes, and data selectivity. Ultimately, the U.S.

EPA concluded, ‘‘Until there is an improved scientific

understanding of the low-dose hypothesis, EPA believes that

it would be premature to require routine testing of substances

for low-dose effects in the Endocrine Disruptor Screening

Program’’ (U.S. EPA, 2002).

Following the NTP low-dose report, a number of key

studies by Naciff et al. (2002, 2009, 2010) have examined

changes in expression of up to 38,500 genes in various cell

lines or uterine tissues exposed to very low doses of EACs,

including BPA at concentrations down to 1nM. These studies

collectively provide support for monotonic dose-responses for

these compounds and showed negligible changes in gene

expression at very low-dose levels. Furthermore, several

comprehensive weight-of-evidence reviews for BPA have

concluded that nonmonotonic dose-response curves and low-

dose effects from BPA are not well supported by available

data (Goodman et al., 2009; Kamrin, 2004, 2007). The Center

for the Evaluation of Risks to Human Reproduction (Chapin

et al., 2008) cited many of the issues related to using low-dose

study data when evaluating BPA. Despite these reports, the

Endocrine Society issued a scientific statement on endocrine-

disrupting chemicals that included support of the low-dose

hypothesis (Diamanti-Kandarakis et al., 2009); thus, the

debate continues.

The low-dose controversy has evolved to consider issues

related to exposures to mixtures of EACs at low, environmentally

relevant dose levels. Although risk assessment approaches for

cumulative and aggregate risk have been in place for many

years (reviewed in Lipscomb et al., 2010), interest in

cumulative exposures to multiple EACs has received increased

attention in recent years (Carney et al., 2010; Hotchkiss et al.,2008; Kortenkamp, 2008). Mixtures have been reported to

have additive effects (e.g., Gennings et al., 2004), less than

additive effects (e.g., Charles et al., 2007), greater than additive

effects (e.g., Laetz et al., 2009), or even counteracting effects

(e.g., Tanida et al., 2009). Overall, data indicate that the

interactions of chemicals in mixtures are dose dependent.

Using mixture data in the context of environmental risk

assessment is even more complex when one considers natural

exposures to EACs in food and pharmaceuticals (Gaido et al.,1998; Safe et al., 1997). Exposures to environmental chemicals

typically involve low doses of each chemical within a mixture

and, given the safety factors used in risk assessment, exposures

generally below the threshold for toxicity. Thus, mixtures

research requires a focus on the behavior of mixtures at

environmentally relevant exposures because results obtained at

high-dose levels may not apply in real-world scenarios. This

concept also was espoused by the Expert Working Group on

Mixtures Toxicology, which was convened by the Society of

Toxicology and the Society for Environmental Toxicology and

Chemistry (Teuschler et al., 2002).

HISTORICAL DEVELOPMENTS IN THE BASIC SCIENCE

OF ENDOCRINE SIGNALING

When the history of endocrinology is written, it is likely that

the 20th century will be neatly divided into two epochs with the

first half being the era of biochemistry when the hormones were

discovered and the emphasis was on proteins and the second half

of the 20th century, the era of genomics when the genes for the

hormone and nuclear receptors (NRs) were identified. The word

hormone was first coined in 1905, but it was not until 1926 that

the first hormones were characterized in detail when Kendeall

and Reichstein isolated and determined the structures of

cortisone and thyroxine (The Nobel Assembly at Karolinska

Institutet, 1950). Soon thereafter, in 1929, Adolf Butenandt and

Edward Adelbert Doisy independently isolated and determined

the structure of estrogen (Tata, 2005). These discoveries along

with other experimental observations allowed Clark to propose

his model for receptor-mediated processes in his publication of

the theory of occupancy. The receptor theory has been refined

since the 1930’s along with the vocabulary used to classify

chemicals that bind to NRs or ligands, which are now classified

into agonist, antagonist, partial agonist, and inverse agonist

based on the biological effects elicited through NRs (Table 2).

Potential EAC’s may produce their effects through inter-

actions with NRs. Under normal physiological conditions, the

intracellular NRs are bound by their endogenous signaling

molecules, or receptor ligands, such as the steroid hormones,

S100 MARTY, CARNEY, AND ROWLANDS

at University of N

airobi on Novem

ber 16, 2013http://toxsci.oxfordjournals.org/

Dow

nloaded from

thyroid hormones, retinoids, vitamin D, and other as yet

identified endogenous ligands. These receptor ligands either

diffuse directly across the plasma membrane of cells or are

sometimes produced in the same cells and bind to the NR. The

commonly understood mechanism of action for NRs involves

the initial ligand binding that results in receptor activation,

binding at specific DNA response elements in nuclear

chromatin, followed by recruitment of a variety of cofactors

that modify the chromosomal organization and interaction with

the transcription complex, thereby altering the rates of

transcription of genes. So, binding of a ligand to an NR is

only the first step in a complex cascade of events leading to

changes in gene expression and subsequent cell signaling. With

the completion of the human genome, it is now clear that there

are some 48 NRs expressed in humans, of which only half have

identified ligands such as steroid hormone receptors that

constitute only a minority of this large superfamily of receptors

(Gronemeyer et al., 2004).

Like the NR superfamily, coregulators are themselves a large

and mixed family of proteins and possess a varied array of

enzymatic activities necessary for chromosomal maintenance

and gene regulation (Roeder, 2005). They can be categorized

based upon their wide-ranging functional properties including

acetyltransferases, ubiquitin ligases, ATP-coupled chromatin

remodeling complexes, protein methylases, cell cycle regu-

lators, RNA helicases, and bridging proteins that facilitate

direct contact with components of the basal transcription

machinery (McKenna and O’Malley, 2003). There are

currently 285 coregulators identified and many are expressed

differentially by cell type, tissue type, gender, developmental

stage, and species (Lonard et al., 2007; Smith and O’Malley,

2004). Moreover, coregulators can be differentially recruited to

the same receptor by ligand-induced conformational changes in

the NR. Thus, with 48 NRs and 285 plus coregulators, the

complexity of these endocrine signaling pathways is apparent.

It is important to note that all these concepts of agonists,

antagonists, partial agonists/antagonists, or inverse agonists are

conditional and specific to a given gene, in a specific cell and

tissue under specific conditions. For example, a given ligand

may be an agonist for a gene in a cell and at the same time be

an antagonist for a different gene in the same cell; or a ligand

can be an agonist for a gene in one cell type and an antagonist

for the same gene in a different cell type or even the same cell

type but under different conditions. Potency is a conditional

measure that depends on parameters of affinity and efficacy, as

well as tissue receptor numbers and ligand bioavailability.

Thus, the intrinsic efficacy of an EAC, whether agonist,

antagonist, partial agonist/antagonist, or inverse agonist for

a given NR, is dependent upon both chemical properties of the

EAC and the biological environment of the NR (Jordan, 2003).

These differences in the properties of ligands are often lost in

the discussion of EACs. An EAC’s potency is dependent on its

affinity for an NR, and some putative EACs have ER binding

affinities that are 10,000–100,000 lower than ER binding

affinities of potent, natural, or contraceptive estrogens

(Gutendorf and Westendorf, 2001). For example, the affinity

of BPA for ERs is approximately 10,000-fold weaker than that

of estradiol (Kuiper et al., 1998). What this means in practical

terms is that a cell would need to be exposed to 10,000

molecules of BPA to produce the same effect through the ER

that one molecule of estradiol would produce through that same

receptor. Furthermore, the amount of BPA that one would need

to consume, breath, or absorb through the skin would be even

greater because the absorption, distribution metabolism, and

elimination of the BPA would limit the amount of BPA

available to the cells. Even if an EAC was highly potent in one

cell type, it does not mean it will have the same efficacy in

another cell type or the same cell type under different

conditions as efficacy is independent of ligand affinity and

instead reflects a ligand’s ability to induce a stable coactivator

binding site on the NR (discussed further below). Thus, two

ligands can have similar potencies for inducing a response

through an NR, but the efficacy for these two ligands inducing

the response can vary greatly and this can have a profound

impact on the ability of these ligands to produce the same

apical effects through the NR. The lower potencies and

different efficacies of EACs relative to the natural hormones

need to be understood and considered in the discussions of the

potential impact of EACs on human health and the environment.

It took another three decades after the elucidation of the

structure of estrogen before the ER was isolated in 1961 by

Elwood Jensen (Tata, 2005) and another two decades passed

before the new tools of molecular biology allowed Pierre

Chambon’s group to clone the gene for the human ER (Green

et al., 1986). The 1980’s ushered in an era of rapid discovery of

the receptor genes for most of the major hormones, which was

followed in the 1990’s by the cloning of novel nuclear orphan

receptors for which there are no known ligands. It was also

during this period, in 1996, that Jan-Ake Gustaffson’s group

cloned a second ER, now called ERb (Kuiper et al., 1996),

requiring the observations of the previous decades of estrogen

TABLE 2

Properties of Chemicals that Affect NRs

Ligand Chemical that binds reversibly to an NR

Affinity A ligand’s capacity to bind to an NR

Agonists Ligand that produce an NR conformation

that is transcriptionally active

Antagonists Ligands that prevent agonist binding or induce

an NR conformation resulting in an action

distinct from that produced by an agonist

Partial agonists Ligands that possess weak agonist activity

Inverse agonists Ligands that induce an NR conformation that

results in transcriptional repression

Efficacy A measure of the capacity of the ligand to

activate, prevent activation, or inactivate an NR

Potency A measure of the concentration or amount of

ligand needed to produce an effect through an NR

ENDOCRINE-MEDIATED TOXICITY S101

at University of N

airobi on Novem

ber 16, 2013http://toxsci.oxfordjournals.org/

Dow

nloaded from

studies to be viewed through a new lens. No doubt there will be

additional chemicals identified as ligands for some of the 48

NRs identified in humans, thereby necessitating an under-

standing of their potential for behaving as EACs.

ESTROGENS: FACTORS AFFECTING AFFINITY, POTENCY,

AND EFFICACY

One of the best understood examples of an endocrine

signaling system that brings all these concepts together is that

of estrogens. A half-century has passed since the initial reports

of a cognate ER, and it is now known that ERa and ERbexist, as well as less well-understood additional subtypes. The

distribution of these two receptors is variable, with tissues such

as bone having predominantly the b isoform, uterus containing

mainly a, but many tissues expressing both a and b.

Furthermore, a and b often regulate opposing functions

providing a feedback loop for finite control of estrogen

signaling. The biological effects elicited by specific ER ligands

are highly cell and tissue specific. One of the best known

examples of this phenomenon is the drug tamoxifen, which has

been used for treatment of early stage breast cancer and as

a chemoprevention agent for women at high risk for this

disease (Jordan, 2003). Tamoxifen is an ER antagonist in breast

cancer but an ER agonist in the bone and uterus (Fong et al.,2007). Another example is the phytoestrogen resveratrol,

which shows more activity in cells that uniquely express

ERb or that express higher levels of ERb than ERa and also

appears to exhibit estrogen antagonist activity for ERa in

certain promoter contexts (Bowers et al., 2000). In addition, only

a small percentage of ER ligands are pure estrogen agonists or

antagonists. Most phytoestrogens and xenoestrogens are partial

agonists, meaning that they are unable to induce a maximal

response matching that of the endogenous ligand, even at very

high concentrations (Zhu, 2005). The latter compounds tend to

act as agonists under some conditions but act as antagonists

when their concentration is very high relative to more potent

ligands. The research from numerous laboratories over the last

five decades has revealed the explanations for many of these

observations.

Differences in the tissue-specific receptor agonist and

antagonist activities have long been recognized for the steroid

hormone receptors and form the basis for the development of

drugs for treatment of hormone-dependent diseases (Ai et al.,2009; Nettles and Greene, 2005). Tamoxifen’s tissue-specific

activity is now thought to be because of recruitment of different

coregulators to the ER by tamoxifen in breast tissue versus

uterine and bone tissue (Gronemeyer et al., 2004). Tamoxifen

was the first of what are now called selective ER modulators or

SERMs where the cellular response to SERMs is determined

by cell type– and promoter-specific differences in coregulator

recruitment by the ER. Thus, coregulators extend the actions of

the same hormone receptor by producing cell type–specific

positive or negative stimuli from the same hormone ligand or

different receptor ligands depending on the cell-specific

coregulator availability and/or ligand-specific induced receptor

conformations with differential coregulator affinities (Fig. 2).

Selective receptor modulators (SRMs) have been observed for

other NRs including peroxisome proliferator–activated recep-

tors, ARs, progesterone receptors, as well as the AhR

(Gronemeyer et al., 2004; Zhang et al., 2008). Thus, SRM

appears to be a common feature of ligand-activated transcrip-

tion factors that explains the observed selectivity that different

ligands confer to NRs on modulation of gene expression and

the resultant tissue changes both within and across tissue and

cell types (Ai et al., 2009; Zhang et al., 2008). No doubt, EACs

will share similar characteristics with SRMs.

All the discoveries discussed above and more have been

harnessed to develop in vitro and in vivo assays to identify

chemicals with endocrine-modulating activities. The very

FIG. 2. The actions of nuclear receptors are ligand dependent and cell

context dependent. Nuclear receptors (R) form a different tertiary structure

depending on the ligand that activates them (EAC-1 or EAC-2). This in turn

causes the receptors to recruit different coregulator complexes, e.g., coactivator

complexes or corepressor complexes that lead to gene induction or repression,

respectively. In Cell A, the two different ligands (EAC-1 and EAC-2) bind to

the same receptor but induce two different tertiary structures, which in turn

cause the receptor to recruit a coactivator complex in the case of EAC-1 bound

receptor or a corepressor complex in the case of EAC-2 bound receptor.

Whereas in Cell B, the opposite occurs with EAC-1 producing a corepressor

complex recruitment through the receptor, but EAC-2 produces a coactivator

complex recruitment through the receptor. Thus, an individual EAC can act as

an agonist, antagonist, partial agonist/antagonist, or inverse agonist depending

on the coregulators present in the specific cell. Although not shown, the gene

architecture influences nuclear receptor-coregulator recruitment as well.

Coactivator (coactivator complex); corepressor (corepressor complex); TBP,

A, B, F, pol II (RNA polymerase II holoenzyme: RNA polymerase II and

general transcription factors); TATA (TATA box promoter site); transcription

(RNA transcript).

S102 MARTY, CARNEY, AND ROWLANDS

at University of N

airobi on Novem

ber 16, 2013http://toxsci.oxfordjournals.org/

Dow

nloaded from

nature of an NR being a trans-acting transcriptional regulator

allows for their exogenous expression in numerous cell types

and cell lines to study the effects of potential ligands as

agonists, antagonists, etc. These studies have led to the

adoption of some assays into the EDSP battery of tests and

many others that are used in the ToxCast panel of assays. The

cloning of these receptors has also facilitated the development

of NR knockout (KO) mouse models that provided important

confirmatory evidence for the physiological role of hormones

in development, reproduction, and normal physiology as well

as some unexpected observations.

With the advent of the KO models, it was possible to

develop transgenic receptor knock in mouse models that

express human NRs in place of their endogenous mouse

receptor. The development of the ER KO mouse (ERKO

mouse) by the National Institute of Environmental Health

Sciences group led by Ken Korach demonstrated that both

sexes of these mutant animals are infertile and show a variety

of phenotypic changes, some of which are associated with the

gonads, mammary glands, reproductive tracts, and skeletal

tissues (Lubahn et al., 1993). The ERKO mice and other

hormone receptor KO mouse models can be used to confirm

the endocrine effects of putative EACs and demonstrate

important differences in the effects produced by endocrine

chemicals compared with endogenous hormones. By knocking

in the human gene for an NR into the appropriate receptor KO

mouse, new transgenic mouse models are made that allow

comparison of putative EACs acting through the human

receptor or mouse receptor. The observations from such

studies are important to consider in the weight of evidence

for predicting the effects of EACs to humans. Future mouse

models are being developed that express the human proteins for

multiple NRs and accessory proteins, and these will allow

further refinement of the effects of EACs on human pathways.

In addition, the technology is now available to make KO rats

and other species, and this will be followed by the ability to

humanize these models as well, thereby furthering our ability to

understand the human relevance for EACs identified in wild-

type laboratory models and in vitro assays.

FUTURE DIRECTIONS IN ENDOCRINE DISRUPTOR

SCREENING AND TESTING

Endocrinology has come far since the term hormones were

first introduced in 1905. What started as a simple concept has

evolved into a very complex array of hormones, receptors, and

cofactors. Because a chemical ligand for an NR can have

multiple intrinsic efficacies (i.e., induce several NR responses)

depending upon the cellular and tissue environment of the NR,

relying upon in vitro cell and biochemical assays to define EACs

has important implications. With suspect EACs possessing the

potential for multiple induced NR responses, the validity for

transferring inferences about the efficacy of the EAC-induced

effect obtained for one response to another response is

questionable. Moreover, the problem of confounding is even

more problematic when considering that measured parameters of

in vitro high-throughput screening systems are by nature dif-

ferent from those found for receptor systems in vivo. Therefore,

the results from these in vitro screening systems may seldom

accurately reflect the intrinsic efficacies of NR ligands in vivoand challenge the validity for extrapolating observations to

humans. Even if a single in vitro system was reflective of

a human response under certain conditions, the fact that the

intrinsic efficacy of an EAC is dependent upon the cellular

context of the NR means that multiple systems and or conditions

would need to be employed to ensure that all the possible

conditions in vitro reflect similar conditions in humans in vivo.

Although this review only touched briefly on some of the

key developments to date concerning endocrine-mediated

toxicity and safety assessment, it should be readily apparent

that the issue has raised more questions and challenges than

have been answered. Most likely, we are only at the very

beginning of a long journey toward improved understanding of

the basic mechanisms of endocrine-mediated toxicity and the

extent to which relevant exposures to hormonally active agents

impact human and environmental health. As we forge ahead,

this active and often controversial topic has already accelerated

the science of toxicology and risk assessment and has

redirected it into some fundamentally different directions.

One such direction is down. This is not meant in a negative

sense but refers to the examination of much lower, environ-

mentally relevant dose levels than have been examined in the

traditional high-dose toxicology paradigm. This new direction

was spurred by research, albeit controversial, that hormonally

active agents might not adhere to the monotonic dose-response

behavior, which has been a fundamental principle of toxicology

since its inception as a field of science. However, further prog-

ress in this area is possible thanks to modern technology, such

as toxicogenomics, which allows one to examine for precursor

changes on the cellular and molecular level (Naciff et al., 2009)

and thus help to better understand the fundamental biology

underlying dose-response for environmentally relevant expo-

sures. In addition, the increased utilization of human, rather than

animal, cells holds promise for enhancing human relevance

even further.

Another research direction points upward, namely, toward

the investigation of higher numbers of chemicals and other

stressors and their potential cumulative effects. For years,

toxicologists have tended to avoid studying cumulative effects

because it was considered an intractable problem. This was

especially so for hormonally active agents because of the vast

array of ligands and multiplicity of sources, which includes not

only man-made chemicals but also compounds in plants,

endogenous hormones, and even nonchemical stressors.

However, because of new high-throughput tools, refined

statistical methods, and economical experimental designs for

mixture studies (Gennings, 1996), questions about combined

ENDOCRINE-MEDIATED TOXICITY S103

at University of N

airobi on Novem

ber 16, 2013http://toxsci.oxfordjournals.org/

Dow

nloaded from

exposures to hormonally active agents can be addressed in

a scientifically rigorous and efficient manner.

This movement toward evaluating more chemicals is also

being combined with a fundamental change toward mechanism-

based testing (inside out) versus the traditional adverse effects-

based system (outside in) as described in the landmark report

by the U.S. NRC Committee on Toxicity Testing and 1365

Assessment of Environmental Agents, Board of Environmental

Studies and Toxicology, Institute for Laboratory Animal

Research (2007) report entitled ‘‘Toxicity Testing in the 21st

Century: A Vision and a Strategy.’’ Within this report,

endocrine screening is highlighted as an area of toxicological

concern that could benefit from the use of molecular-based

screening methods. Progress has been made toward implementing

this approach. EPA’s ToxCast program contains numerous

endocrine-related screening assays, including ER and AR

receptor binding assays, ER, AR, and thyroid receptor reporter

gene assays, an assay for thyrotropin-releasing hormone

receptor, and an aromatase enzymatic activity assay (Houck,

2009). These assays are conducted in high-throughput mode,

so results are generated quickly. However, there are concerns

about relying on these in vitro and alternative assays for

regulatory decisions in determining potential endocrine activity

(e.g., Bus and Becker, 2009) because of the lack of

toxicokinetic considerations (absorption, distribution, metabolism,

and excretion), the imperfect concordance of ToxCast assays

with in vivo data, and achieving agreement on a mechanism-

based definition of adverse. In the short run, these data may

contribute to the weight of evidence for some test compounds

in determining potential endocrine activity. Furthermore, if the

vision of ‘‘Toxicology in the 21st Century’’ is realized, such an

approach could prove to be far more efficient and cost effective

than the existing EDSP assays, some of which are resource

intensive and use large numbers of animals to screen for

a limited number of MOAs.

Finally, future research in this area needs to look outward,

meaning that information on specific components of the

endocrine system (e.g., specific cell culture responses) needs

to be interpreted in the context of whole-body physiology,

preferably that of humans. In the case of environmental

species effects, integrative frameworks such as the ‘‘adverse

outcomes pathway’’ concept have recently been developed to

address links between events at the molecular level all the

way up through adverse outcomes at the population level

(Ankley et al., 2010). Because the endocrine system

integrates responses across distant cells, tissues, and organs,

a reductionist approach that solely focuses on isolated

components of an endocrine circuit can generate incomplete

answers and may even guide public health professionals in the

wrong direction. In this same vein, basic principles of

toxicology and in particular, toxicokinetics, should not be

forgotten and need to be part of any research program if it is

to be meaningful for risk assessment. For example, recent

research on the toxicokinetics of BPA and genistein has

highlighted the tremendous impact of hepatic first pass

metabolism in converting these compounds to their

biologically inactive glucuronide forms (Teeguarden and

Barton, 2004; Teeguarden et al., 2005). Therefore, route of

exposure can have a dramatic effect on the percentage of free,

biologically active compound and consequently the effects

produced. The key point is that as we delve deeper into the

details of individual mechanisms, it is essential that these new

findings be interpreted in a holistic, physiological framework.

Fortunately, new tools in exposure modeling, physiologically

based pharmacokinetic modeling, and systems biology are

increasingly available to assist with experimental design, data

integration, and data interpretation to achieve this goal.

CONCLUSION

Concerns about EACs have come to the forefront of

toxicology over the past 15 years. The precise impact of

EACs is difficult to discern as EACs can be beneficial or

adverse, producing different effects in different tissues at

different life stages. To complicate this issue further, exposure

to low-level environmental EACs occurs in an environment in

which natural EACs (e.g., phytoestrogens) and/or exposures

to pharmaceuticals contribute to the background hormonal

mileau. The focus on endocrine disruption represents a new

paradigm in toxicology, wherein the MOA, rather than

toxicological outcome, is the focus of chemical screening.

Development of the EDSP has been more complex than

originally envisioned, and the difficulty in identifying assays,

efficiently conducting those assays, and data interpretation

have become apparent. As endocrine issues have increased in

scope, increased research has led to a better understanding of

receptor biology. To take full advantage of this new

knowledge and efficiently meet the needs to screen com-

pounds for endocrine activity, the EDSP should be iterative

and allow for reevaluation of earlier assumptions based on the

state of the science. Clearly, the past 15 years have

demonstrated that legislative mandates for specific types of

toxicity testing that precede scientific understanding are

difficult to implement and limit the latitude to adapt to

evolving science. Even so, the significant resources dedicated

to developing and validating the panel of EDSP assays have

led to an improved understanding of the challenges involved

with implementing and interpreting EDSP assays, and these

lessons will be beneficial to eventual implementation of the

EPA’s ToxCast assays and human primary cell assays as

proposed in the National Research Council report Toxicity

Testing in the 21st Century.

ACKNOWLEDGMENTS

The authors would like to thank Drs James Bus, Matt

Lebaron, and Katie Coady for their thoughtful reviews of the

manuscript and Dr David Geter for contributing Figure 1.

S104 MARTY, CARNEY, AND ROWLANDS

at University of N

airobi on Novem

ber 16, 2013http://toxsci.oxfordjournals.org/

Dow

nloaded from

REFERENCES

Adams, N. R. (1995). Detection of the effects of phytoestrogens on sheep and

cattle. J. Anim. Sci. 73, 1509–1515.

Ai, N., Krasowski, M. D., Welsh, W. J., and Ekins, S. (2009). Understanding

nuclear receptors using computational methods. Drug Discov. Today

14, 486–494. Available at: http://www.ncbi.nlm.nih.gov/pubmed/19429508.

Ankley, G. T., Bennett, R. S., Erickson, R. J., Hoff, D. J., Hornung, M. W.,

Johnson, R. D., Mount, D. R., Nichols, J. W., Russom, C. L.,

Schmieder, P. K., et al. (2010). Adverse outcome pathways: a conceptual

framework to support ecotoxicology research and risk assessment. Environ.

Toxicol. Chem. 29, 730–741.

Aoki, Y. (2001). Polychlorinated biphenyls, polychlorinated dibenzo-p-dioxins,

and polychlorinated dibenzofurans as endocrine disrupters—what we have

learned from Yusho disease. Environ. Res. 86, 2–11.

Arnold, S. F., Klotz, D. M., Collins, B. M., Vonier, P. M., Guillette, L. J., Jr.,

and McLachlan, J. A. (1996). Synergistic activation of estrogen receptor with

combinations of environmental chemicals. Science 272, 1489–1492.

Ashby, J., Lefevre, P. A., Odum, J., Harris, C. A., Routledge, E. J., and

Sumpter, J. P. (1997). Synergy between synthetic oestrogens? Nature

385, 494.

Ashby, J., Owens, W., Odum, J., and Tinwell, H. (2003). The intact immature

rodent uterotrophic bioassay: possible effects on assay sensitivity of

vomeronasal signals from male rodents and strain differences. Environ.

Health Perspect. 111, 1568–1570.

Barlow, S., Kavlock, R. J., Moore, J. A., Schantz, S. L., Sheehan, D. M.,

Shuey, D. L., and Lary, J. M. (1999). Teratology Society Public Affairs

Committee position paper: developmental toxicity of endocrine disruptors to

humans. Teratology 60, 365–375.

Blair, R. M., Fang, H., Branham, W. S., Hass, B. S., Dial, S. L., Moland, C. L.,

Tong, W., Shi, L., Perkins, R., and Sheehan, D. M. (2000). The estrogen

receptor relative binding affinities of 188 natural and xenochemicals:

structural diversity of ligands. Toxicol. Sci. 54, 138–153.

Boue, S. M., Wiese, T. E., Nehls, S., Burow, M. E., Elliott, S., Carter-

Wientjes, C. H., Shih, B. Y., McLachlan, J. A., and Cleveland, T. E. (2003).

Evaluation of the estrogenic effects of legume extracts containing

phytoestrogens. J. Agric. Food Chem. 51, 2193–2199.

Bowers, J. L., Tyulmenkov, V. V., Jernigan, S. C., and Klinge, C. M. (2000).

Resveratrol acts as a mixed agonist/antagonist for estrogen receptors alpha

and beta. Endocrinology 141, 3657–3667.

Bus, J. S., and Becker, R. A. (2009). Toxicity testing in the 21st century: a view

from the chemical industry. Toxicol. Sci. 112, 297–302.

Carlsen, E., Giwercman, A., Keiding, N., and Skakkebaek, N. E. (1992).

Evidence for decreasing quality of semen during past 50 years. BMJ

305, 609–613.

Carney, E., Woodburn, K. B., and Rowland, S. J. (2010). Endocrine active

chemicals. In Principles and Practice of Mixtures Toxicology

(M. M. Mumtaz, Ed.), pp. 421–442. WILEY-VCH Verlag GmbH & Co.,

Weinheim, Germany.

Chapin, R. E., Adams, J., Boekelheide, K., Gray, L. E., Jr., Hayward, S. W.,

Lees, P. S., McIntyre, B. S., Portier, K. M., Schnorr, T. M., Selevan, S. G.,

et al. (2008). NTP-CERHR expert panel report on the reproductive and

developmental toxicity of bisphenol A. Birth Defects Res. B 83, 157–395.

Charles, G. D., Gennings, C., Tornesi, B., Kan, H. L., Zacharewski, T. R.,

Gollapudi, B. B., and Carney, E. W. (2007). Analysis of the interaction of

phytoestrogens and synthetic chemicals: an in vitro/in vivo comparison.

Toxicol. Appl. Pharmacol. 218, 280–288.

Charles, G. D., Kan, H. L., Schisler, M. R., Gollapudi, B. B., and Marty, M. S.

(2005). A comparison of in vitro and in vivo EDSTAC test battery results

for detecting antiandrogenic activity. Toxicol. Appl. Pharmacol. 202,

108–120.

Colborn, T., Dumanoski, D., and Myers, J. P. (1996). In Our Stolen Future. Are

We Threatening Our Fertility, Intelligence and Survival?—A Scientific

Detective Story. Penguin Books USA, New York.

Colborn, T., vom Saal, F. S., and Soto, A. M. (1993). Developmental effects of

endocrine-disrupting chemicals in wildlife and humans. Environ. Health

Perspect. 101, 378–384.

Davis, D. L., Bradlow, H. L., Wolff, M., Woodruff, T., Hoel, D. G., and Anton-

Culver, H. (1993). Medical hypothesis: xenoestrogens as preventable causes

of breast cancer. Environ. Health Perspect. 101, 372–377.

Diamanti-Kandarakis, E., Bourguignon, J.-P., Giudice, L. C., Hauser, R.,

Prins, G. S., Soto, A. M., Zoeller, R. T., and Gore, A. C. (2009). Endocrine-

disrupting chemicals: an Endocrine Society scientific statement. Endocr.

Rev. 30, 293–342.

Endocrine Disruptor Screening and Testing Advisory Committee (EDSTAC).

(1998). Endocrine Disruptor Screening and Testing Advisory Committee

(EDSTAC) Final Report, August 1998. Available at: http://www.epa.gov/

endo/pubs/edspoverview/finalrpt.htm. Accessed August 13, 2010.

Fisch, H. (2008). Declining worldwide sperm counts: disproving a myth. Urol.

Clin. North Am 35, 137–146.

Fong, C. J., Burgoon, L. D., Williams, K. J., Forgacs, A. L., and

Zacharewski, T. R. (2007). Comparative temporal and dose-dependent

morphological and transcriptional uterine effects elicited by tamoxifen

and ethynylestradiol in immature, ovariectomized mice. BMC Genomics

8, 151.

Food Quality Protection Act (FQPA). (1996). Public Law 104-170, 110 STAT.

1489, Available at: http://www.epa.gov/scipoly/oscpendo/pubs/fqpa.pdf

Accessed August 13, 2010.

Freyberger, A., and Schladt, L. (2009). Evaluation of the rodent Hershberger

bioassay on intact juvenile males—testing of coded chemicals and

supplementary biochemical investigations. Toxicology 262, 114–120.

Gaido, K., Dohme, L., Wang, F., Chen, I., Blankvoort, B., Ramamoorthy, K.,

and Safe, S. (1998). Comparative estrogenic activity of wine extracts and

organochlorine pesticide residues in food. Environ. Health Perspect.

106(Suppl. 6), 1347–1351.

Garrison, A. W., and Pope, J. D. (1975). In GC/MS Analysis of Organic

Compounds in Domestic Wastewaters. Chemical Congress of the North

American Continent. Ann Arbor Science, Ann Arbor, Michigan.

Gennings, C. (1996). Economical designs for detecting and characterizing

departure from additivity in mixtures of many chemicals. Food Chem.

Toxicol. 34, 1053–1058.

Gennings, C., Carter, W. H., Jr., Carney, E. W., Charles, G. D.,

Gollapudi, B. B., and Carchman, R. A. (2004). A novel flexible approach

for evaluating fixed ratio mixtures of full and partial agonists. Toxicol. Sci.

80, 134–150.

Giusti, R. M., Iwamoto, K., and Hatch, E. E. (1995). Diethylstilbestrol

revisited: a review of the long-term health effects. Ann. Intern. Med.

122, 778–788.

Goodman, J. E., Witorsch, R. J., McConnell, E. E., Sipes, I. G., Slayton, T. M.,

Yu, C. J., Franz, A. M., and Rhombert, L. R. (2009). Weight-of-evidence

evaluation of reproductive and developmental effects of low doses of

bisphenol A. Crit. Rev. Toxicol. 39, 1–75.

Green, S., Walter, P., Greene, G., Krust, A., Goffin, C., Jensen, E., Scrace, G.,

Waterfield, M., and Chambon, P. (1986). Cloning of the human oestrogen

receptor cDNA. J. Steroid Biochem. 24, 77–83.

Gronemeyer, H., Gustafsson, J. A., and Laudet, V. (2004). Principles for

modulation of the nuclear receptor superfamily. Nat. Rev. 3, 950–964.

Gutendorf, B., and Westendorf, J. (2001). Comparison of an array of in vitro

assays for the assessment of the estrogenic potential of natural and synthetic

estrogens, phytoestrogens and xenoestrogens. Toxicology 166, 79–89.

ENDOCRINE-MEDIATED TOXICITY S105

at University of N

airobi on Novem

ber 16, 2013http://toxsci.oxfordjournals.org/

Dow

nloaded from

Haseman, J. K., Bailer, A. J., Kodell, R. L., Morris, R., and Portier, K. (2001).

Statistical issues in the analysis of low-dose endocrine disruptor data.

Toxicol. Sci. 61, 201–210.

Hecker, M., Giesy, J. P., Jones, P. D., Higley, E. B., Newsted, J. L., and

Mehrle, P. (2006). In Influence of Cell Passage and Freeze/Thaw Events on

Basal Production of 17b-Estradiol and Testosterone by H295R Cells.

Interim Draft Report, Contract Code # GS-10F-0041L. ENTRIX, Inc.

September, 2006. Available at: www.epa.gov/endo/pubs/appendix_i-h295r_

culture_protocol.pdf. Accessed August 13, 2010.

Hecker, M., Giesy, J. P., and Timm, G. (2008). Multi-laboratory Validation of

the H295R Steroidogenesis Assay to Identify Modulators of Testosterone and

Estradiol Production. Report, Contract Code # GS-10F-0041L. ENTRIX,

Inc. February, 2008. Available at: www.epa.gov/endo/pubs/h295r_validation_

study_interim_report.pdf. Accessed August 13, 2010.

Hignite, C., and Azarnoff, D. L. (1977). Drugs and drug metabolites as

environmental contaminants: chlorophenoxyisobutyrate and salicylic acid in

sewage water effluent. Life Sci. 20, 337–341.

Hotchkiss, A. K., Rider, C. V., Blystone, C. R., Wilson, V. S., Hartig, P. C.,

Ankley, G. T., Foster, P. M., Gray, C. L., and Gray, L. E. (2008). Fifteen

years after ‘‘Wingspread’’—environmental endocrine disrupters and human

and wildlife health: where we are today and where we need to go. Toxicol.

Sci. 105, 235–259.

Houck, K. (2009). In What Can ToxCast� Tell Us about Endocrine Disruption?

ISRTP Workshop: The Endocrine Disruptor Screening Program—What Can

Screening Results Tell Us About Potential Adverse Endocrine Effects.

Bethesda, MD. Available at: http://www.isrtp.org/index.htm. Accessed

August 13, 2010.

Hunter, D. J., and Kelsey, K. T. (1993). Pesticide residues and breast cancer:

the harvest of a silent spring? J. Natl. Cancer Inst. 85, 598–599.

International Programme on Chemical Safety (IPCS). (2002). Global

assessment of the state-of-the-science of endocrine disruptors. Damstra T.,

Barlow, S., Bergman, A., Kavlock, R. and Van der Kraak, G., Eds.,

WHO publication no. WHO/PCS/EDC/02.2. World Health Organization,

Geneva.

Jobling, S., Williams, R., Johnson, A., Taylor, A., Gross-Sorokin, M.,

Nolan, M., Tyler, C. R., van Aerle, R., Santos, E., et al. (2006). Predicted

exposures to steroid estrogens in U.K. rivers correlate with widespread

sexual disruption in wild fish populations. Environ. Health Perspect.

114(Suppl. 1), 32–39.

Jordan, V. C. (2003). Antiestrogens and selective estrogen receptor modulators

as multifunctional medicines. 1. Receptor interactions. J. Med. Chem.

46, 883–908.

Kamrin, M. A. (2004). Bisphenol A: a scientific evaluation. MedGenMed.

6(3), 7.

Kamrin, M. A. (2007). The ‘‘low dose’’ hypothesis: validity and implications

for human risk. Int. J. Toxicol. 26, 13–23.

Kanno, J., Onyon, L., Haseman, J., Fenner-Crisp, P., Ashby, J., and Owens, W.

(2001). The OECD program to validate the rat uterotrophic bioassay to

screen compounds for in vivo estrogenic responses: phase 1. Environ. Health

Perspect. 109, 785–794.

Kanno, J., Onyon, L., Peddada, S., Ashby, J., Jacob, E., and Owens, W.

(2003a). The OECD program to validate the rat uterotrophic bioassay. Phase

2: coded single-dose studies. Environ. Health Perspect. 111, 1550–1558.

Kanno, J., Onyon, L., Peddada, S., Ashby, J., Jacob, E., and Owens, W.

(2003b). The OECD program to validate the rat uterotrophic bioassay. Phase

2: dose-response studies. Environ. Health Perspect. 111, 1530–1549.

Klinge, C. M., Risinger, K. E., Watts, M. B., Beck, V., Eder, R., and

Jungbauer, A. (2003). Estrogenic activity in white and red wine extracts.

J. Agric. Food Chem. 51, 1850–1857.

Kortenkamp, A. (2008). Low dose mixture effects of endocrine disrupters:

implications for risk assessment and epidemiology. Int. J. Androl.

31, 233–240.

Kuiper, G. G., Enmark, E., Pelto-Huikko, M., Nilsson, S., and Gustafsson, J. A.

(1996). Cloning of a novel receptor expressed in rat prostate and ovary. Proc.

Natl. Acad. Sci. U.S.A. 93, 5925–5930.

Kuiper, G. G., Lemmen, J. G., Carlsson, B., Corton, J. C., Safe, S. H., van der

Saag, P. T., van der Burg, B., and Gustafsson, J. A. (1998). Interaction of

estrogenic chemicals and phytoestrogens with estrogen receptor beta.

Endocrinology 139, 4252–4263.

Laetz, C. A., Baldwin, D. H., Collier, T. K., Herbert, V., Stark, J. D., and

Scholz, N. L. (2009). The synergistic toxicity of pesticide mixtures:

implications for risk assessment and the conservation of endangered pacific

salmon. Environ. Health Perspect. 117, 348–353.

Laws, S. C., Ferrell, J. M., Stoker, T. E., and Cooper, R. L. (2003). Pubertal

development in female Wistar rats following exposure to propazine, and

atrazine biotransformation by-products, diamino-s-chlorotriazine and

hydroxyatrazine. Toxicol. Sci. 76, 190–200.

Lintelmann, J., Katayama, A., Kurihara, N., Shore, L., and Wenzel, A. (2003).

Endocrine disruptors in the environment. Pure Appl. Chem. 75, 631–681.

Lipscomb, J. C., Lambert, J. C., and Teuschler, L. (2010). Chemical mixtures

and cumulative risk assessment. In Principles and Practice of Mixtures

Toxicology (M. M. Mumtaz, Ed.). WILEY-VCH Verlag GmbH & Co.,

Weinheim, Germany.

Lonard, D. M., Lanz, R. B., and O’Malley, B. W. (2007). Nuclear receptor

coregulators and human disease. Endocr. Rev. 28, 575–587.

Lubahn, D. B., Moyer, J. S., Golding, T. S., Couse, J. F., Korach, K. S., and

Smithies, O. (1993). Alteration of reproductive function but not prenatal

sexual development after insertional disruption of the mouse estrogen

receptor gene. Proc. Natl. Acad. Sci. U.S.A. 90, 11162–11166.

Martin, R.-P., Wright, A. N., Miyamoto, J., and Wermuth, C. G. (1998).

Natural and anthropogenic environmental oestrogens: the scientific basis for

risk assessment. Pure Appl. Chem. 70, iv.

McKenna, N. M., and O’Malley, B. W. (2003). Nuclear receptor signaling:

concepts and models. NURSA: The Nuclear Receptor Signaling Atlas.

Available at: http://www.nursa.org/flash/gene/nuclearreceptor/start.html.

Accessed July 15, 2010.

McLachlan, J. A. (1997). Synergistic effect of environmental estrogens: report

withdrawn. Science 277, 462–463.

Melnick, R., Lucier, G., Wolfe, M., Hall, R., Stancel, G., Prins, G., Gallo, M.,

Reuhl, K., Ho, S.-M., Brown, T., et al. (2002). Summary of the National

Toxicology Program’s report of the endocrine disruptors low-dose peer

review. Environ. Health Perspect. 110, 427–431.

Naciff, J. M., Hess, K. A., Overmann, G. J., Torontali, S. M., Carr, G. J.,

Tiesman, J. P., Foertsch, L. M., Richardson, B. D., Martinez, J. E., and

Daston, G. P. (2005). Gene expression changes induced in the testis by

transplacental exposure to high and low doses of 17{alpha}-ethynyl

estradiol, genistein, or bisphenol A. Toxicol. Sci. 86, 396–416.

Naciff, J. M., Jump, M. L., Torontali, S. M., Carr, G. J., Tiesman, J. P.,

Overmann, G. J., and Daston, G. P. (2002). Gene expression profile induced

by 17alpha-ethynyl estradiol, bisphenol A, and genistein in the developing

female reproductive system of the rat. Toxicol. Sci. 68, 184–199.

Naciff, J. M., Khambatta, Z. S., Reichling, T. D., Carr, G. J., Tiesman, J. P.,

Singleton, D. W., Khan, S. A., and Daston, G. P. (2010). The genomic

response of Ishikawa cells to bisphenol A exposure is dose- and time-

dependent. Toxicology 270, 137–149.

Naciff, J. M., Khambatta, Z. S., Thomason, R. G., Carr, G. J., Tiesman, J. P.,

Singleton, D. W., Khan, S. A., and Daston, G. P. (2009). The genomic

response of a human uterine endometrial adenocarcinoma cell line to

17alpha-ethynyl estradiol. Toxicol. Sci. 107, 40–55.

S106 MARTY, CARNEY, AND ROWLANDS

at University of N

airobi on Novem

ber 16, 2013http://toxsci.oxfordjournals.org/

Dow

nloaded from

National Research Council (NRC) (1999). Hormonally Active Agents in the

Environment. National Academy Press, Washington, DC.

National Research Council (NRC) Committee on Toxicity Testing and

Assessment of Environmental Agents, Board of Environmental Studies

and Toxicology, Institute for Laboratory Animal Research. (2007). Toxicity

Testing in the 21st Century. A Vision and a Strategy. National Academies

Press, Washington, DC.

Nettles, K. W., and Greene, G. L. (2005). Ligand control of coregulator

recruitment to nuclear receptors. Annu. Rev. Physiol. 67, 309–333.

Newburgh, R. W. (1975). Toxicology of the reproductive system. In

Toxicology The Basic Science of Poisons. (L. J. Casarett and J. Doull,

Eds.), pp. 261–274. Macmillan Publishing Co., Inc, New York, NY.

Nieuwkoop, P. D., and Faber, J. (1994). In Normal Table of Xenopus laevis.

Garland Publishing Inc., New York, NY.

O’Connor, J. C., Cook, J. C., Marty, M. S., Davis, L. G., Kaplan, A. M., and

Carney, E. W. (2002). Evaluation of Tier I screening approaches for detecting

endocrine-active compounds (EACs). Crit. Rev. Toxicol. 32, 521–549.

O’Connor, J. C., Davis, L. G., Frame, S. R., and Cook, J. C. (2000). Evaluation

of a Tier I screening battery for detecting endocrine-active compounds

(EACs) using the positive controls testosterone, coumestrol, progesterone,

and RU486. Toxicol. Sci. 69, 79–91.

O’Connor, J. C., Frame, S. R., Davis, L. G., and Cook, J. C. (1999). Detection

of the environmental antiandrogen p,p#-DDE in CD and Long-Evans rats

using a Tier I screening battery and a Hershberger assay. Toxicol. Sci.

51, 44–53.

O’Connor, J. C., Marty, M. S., Becker, R. A., Snajdr, S., and Kaplan, A. M.

(2008). Results of the negative control chemical allyl alcohol in the 15-day

intact adult male rat screening assay for endocrine activity. Birth Defects Res.

B Dev. Reprod. Toxicol. 83, 117–122.

Organization for Economic Co-operation and Development (OECD). (2003).

Detailed Background Review of the Uterotrophic Bioassay, OECD Series on

Testing and Assessment, No. 38, OECD Publishing.

Owens, W., Ashby, J., Odum, J., and Onyon, L. (2003). The OECD program to

validate the rat uterotrophic bioassay. Phase 2: dietary phytoestrogen

analyses. Environ. Health Perspect. 111, 1559–1567.

Owens, W., Gray, L. E., Zeiger, E., Walker, M., Yamasaki, K., Ashby, J., and

Jacob, E. (2007). The OECD program to validate the rat Hershberger

bioassay to screen compounds for in vivo androgen and antiandrogen

responses. Phase 2 dose-response studies. Environ. Health Perspect. 115,

671–678.

Owens, W., and Koeter, H. B. (2003). The OECD program to validate the rat

uterotrophic bioassay: an overview. Environ. Health Perspect. 111,

1527–1529.

Owens, W., Zeiger, E., Walker, M., Ashby, J., Onyon, L., and Gray, L. E., Jr.

(2006). The OECD program to validate the rat Hershberger bioassay to

screen compounds for in vivo androgen and antiandrogen responses. Phase 1:

use of a potent agonist and a potent antagonist to test the standardized

protocol. Environ. Health Perspect. 11, 1259–1265.

Powlin, S. S., Cook, J. C., Novak, S., and O’Connor, J. C. (1998). Ex vivo and

in vitro testis and ovary explants: utility for identifying steroid biosynthesis

inhibitors and comparison to a Tier I screening battery. Toxicol. Sci.

46, 61–74.

Ramamoorthy, K., Wang, F., Chen, I. C., Norris, J. D., McDonnell, D. P.,

Leonard, L. S., Gaido, K. W., Bocchinfuso, W. P., Korach, K. S., and

Safe, S. (1997). Estrogenic activity of a dieldrin/toxaphene mixture

in the mouse uterus, MCF-7 human breast cancer cells, and yeast-

based estrogen receptor assays: no apparent synergism. Endocrinology

138, 1520–1527.

Roeder, R. G. (2005). Transcriptional regulation and the role of diverse

coactivators in animal cells. FEBS Lett. 579, 909–915.

Safe, S., Connor, K., Ramamoorthy, K., Gaido, K., and Maness, S. (1997).

Human exposure to endocrine-active chemicals: hazard assessment prob-

lems. Regul. Toxicol. Pharmacol. 26, 52–58.

Safe, S., Papineni, S., and Chintharlapalli, S. (2008). Cancer chemotherapy

with indole-3-carbinol, bis(3’-indolyl)methane and synthetic analogs.

Cancer Lett. 269, 326–338.

Schueler, F. W. (1946). Sex hormonal action and chemical constitution. Science

103, 221–223.

Sharpe, R. M., and Skakkebaek, N. E. (1993). Are oestrogens involved in

falling sperm counts and disorders of the male reproductive tract? Lancet

341, 1392–1395.

Shelby, M. D., Newbold, R. R., Tully, D. B., Chae, K., and Davis, V. L.

(1996). Assessing environmental chemicals for estrogenicity using a combi-

nation of in vitro and in vivo assays. Environ. Health Perspect.

104, 1296–1300.

Sluczewski, A., and Roth, P. (1948). Effects of androgenic and estrogenic

compounds on the experimental metamorphoses of amphibians. Gynecol.

Obstet. 47, 164–176.

Smith, C. L., and O’Malley, B. W. (2004). Coregulator function: a key to

understanding tissue specificity of selective receptor modulators. Endocr.

Rev. 25, 45–71.

Stoker, T. E., and Zorrilla, L. M. (2010). The effects of endocrine disrupting

chemicals on pubertal development in the rat: use of the EDSP pubertal

assays as a screen. In Endocrine Toxicology, 3rd ed. (J. C. Eldridge and

J. T. Stevens, Eds.), pp. 27–81. Informa Healthcare, New York, NY.

Stumm-Zollinger, E., and Fair, G. M. (1965). Biodegradation of steroid

hormones. J. Water Pollut. Control Fed. 37, 1506–1510.

Sumpter, J. P. (1995). Feminized responses in fish to environmental estrogens.

Toxicol. Lett. 82–83, 737–742.

Tabak, H. H., and Bunch, R. L. (1970). Steroid hormones as water pollutants.

I. Metabolism of natural and synthetic ovulation-inhibiting hormones by

microorganisms of activated sludge and primary settled sewage. Dev. Ind.

Microbiol. 11, 367–376.

Takeyoshi, M. (2006). In Draft Report of Pre-validation and Inter-laboratory

Validation for Stably Transfected Transcriptional Activation (TA) Assay to

Detect Estrogenic Activity—The Human Estrogen Receptor Alpha Mediated

Reporter Gene Assay Using hER-HeLa-9903 Cell Line. Chemicals

Evaluation and Research Institute, Japan. Available at: http://www.oecd.

org/document/62/0,3343,en_2649_34377_2348606_1_1_1_1,00.htm

Accessed August 13, 2010.

Tanida, T., Warita, K., Ishihara, K., Fukui, S., Mitsuhashi, T., Sugawara, T.,

Tabuchi, Y., Nanmori, T., Qi, W.-M., Inamoto, T., et al. (2009). Fetal and

neonatal exposure to three typical environmental chemicals with different

mechanisms of action: mixed exposure to phenol, phthalate, and dioxin

cancels the effects of sole exposure on mouse midbrain dopaminergic nuclei.

Toxicol. Lett. 189, 40–47.

Tata, J. R. (2005). One hundred years of hormones. EMBO Rep. 6, 490–496.

Teeguarden, J. G., and Barton, H. A. (2004). Computational modeling of

serum-binding proteins and clearance in extrapolations across life stages and

species for endocrine active compounds. Risk Anal. 24, 751–770.

Teeguarden, J. G., Waechter, J. M., Jr.., Clewell, H. J., III, Covington, T. R.,

and Barton, H. A. (2005). Evaluation of oral and intravenous route

pharmacokinetics, plasma protein binding, and uterine tissue dose metrics of

bisphenol A: a physiologically based pharmacokinetic approach. Toxicol.

Sci. 85, 823–838.

Teuschler, L., Klaunig, J., Carney, E., Chambers, J., Conolly, R.,

Gennings, C., Giesy, J., Hertzberg, R., Klaassen, C., Kodell, R., et al.

(2002). Support of science-based decisions concerning the evaluation of the

toxicology of mixtures: a new beginning. Regul. Toxicol. Pharmacol. 36,

34–39.

ENDOCRINE-MEDIATED TOXICITY S107

at University of N

airobi on Novem

ber 16, 2013http://toxsci.oxfordjournals.org/

Dow

nloaded from

The Nobel Assembly at Karolinska Institutet. (1950). The Nobel Prize in

physiology or medicine 1950. (Edward C. Kendall, Tadeus Reichstein, and

Philip S. Hench). Available at: http://nobelprize.org/nobel_prizes/medicine/

laureates/1950/. Accessed July 15, 2010.

Touart, L. (2005). In The Story of the Fish Screening Assay in EPA’s Endocrine

Disruptor Screening Program. Available at: http://www.epa.gov/endo/pubs/

edmvac/fish_screening_story.htm. Accessed April 4, 2002.

U.S. Environmental Protection Agency (U.S. EPA). (2002). EPA Statement

Regarding Endocrine Disruptor Low-Dose Hypothesis. U.S. Environmental

Protection Agency, Washington, DC. Available at: http://www.epa.gov/

endo/pubs/edmvs/lowdosepolicy.pdf. Dated March 26, 2005. Accessed

August 13, 2010.

U.S. Environmental Protection Agency (U.S. EPA). (2007). Integrated

Summary Report for Validation of a Test Method for Assessment of PubertalDevelopment and Thyroid Function in Juvenile Male Rats as a Potential

Screen in the Endocrine Disruptor Screening Program Tier-1 Battery. U.S.

Environmental Protection Agency, Office of Science Coordination and

Policy and Office of Research and Development, Washington, DC. Available

at: http://www.epa.gov/endo/pubs/male_pubertal_isr.pdf Accessed August

13, 2010.

Walker, B. S., and Janney, J. C. (1930). Estrogenic substances. II. Analysis of

plant sources. Endocrinology 14, 389–392.

Zhang, S., Rowlands, C., and Safe, S. (2008). Ligand-dependent interactions of

the Ah receptor with coactivators in a mammalian two-hybrid assay. Toxicol.

Appl. Pharmacol. 227, 196– 206. Available at: http://www.ncbi.nlm.nih.gov/

pubmed/18048071.

Zhu, B. T. (2005). Mechanistic explanation for the unique pharmacologic

properties of receptor partial agonists. Biomed. Pharmacother. 59,

76–89.

S108 MARTY, CARNEY, AND ROWLANDS

at University of N

airobi on Novem

ber 16, 2013http://toxsci.oxfordjournals.org/

Dow

nloaded from