Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
State of the Science Evaluation:
Nonmonotonic Dose Responses as They Apply
to Estrogen, Androgen, and Thyroid Pathways
and EPA Testing and Assessment Procedures
U.S. Environmental Protection Agency
Jointly developed by:
Office of Research and Development
Office of Science Policy
National Health and Environmental Effects Research Laboratory
National Center for Environmental Assessment
National Center for Computational Toxicology
Office of Chemical Safety and Pollution Prevention
Office of Pesticide Programs
Office of Pollution Prevention and Toxics
Office of Science Coordination and Policy
June 2013
1 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Table of Contents
Foreward 9
Executive Summary 10
1. Introduction 20
1.1 Background 21
1.2 Scope of the Current Evaluation 27
1.3 Central Scientific Questions 28
2. General Biological Concepts and Statistical Considerations for NMDR 28
2.1 Pharmacokinetic Processes 28
2.2 Modes of Action and Adverse Outcome Pathways 29
2.3 Determinants of Nonmonotonic Dose-Response 31
2.4 The Hypothalamic-pituitary-gonadal and Hypothalamic-pituitary-thyroid Axes 34
2.5 Statistical Issues and Experimental Design 40
3. NMDRs from in vitro Studies 42
4. NMDRs from in vivo Studies 55
4.1 Aquatic Models 55
4.1.1 Literature Search and Analysis, Aquatic Species 56
4.1.2 NMDRs in the Estrogen Hormone System in Aquatic Species 62
4.1.2.1 Estrogen Receptor Agonists 62
4.1.2.2 Selective Estrogen Receptor Modulator (Tamoxifen) 64
4.1.3 NMDRs in the Androgen Hormone Pathway in Aquatic Species 64
4.1.3.1 Androgens 64
4.1.3.2 Androgen Receptor Antagonists 69
4.1.3.3 Steroid Synthesis Inhibitors 70
4.1.3.4 Role of Compensatory Processes in NMDR - Examples from Fish Time-Course Studies
with Three EDCs 73
4.1.4 NMDRs in the Thyroid Hormone Pathway in Aquatic Species 79
4.1.4.1 Fish 79
4.1.4.2 Amphibians 82
4.2 Mammalian Models 83
4.2.1 Literature Search and Selection Strategy for E and A Pathways 83
4.2.1.1 Specific Considerations in Reviewing the Literature on E and A 85
4.2.2 Estrogen Hormone System 92
4.2.2.1 Ethinyl Estradiol (EE2) [A.l.a] 94
4.2.2.2 17(B Estradiol (E2) [A.l.b] 96
4.2.2.3 Diethylstilbestrol (DES) [A.l.c] 96
4.2.2.4 Genistein [A.l.d] 97
4.2.2.5 Bisphenol A [A.l.k] 99
4.2.2.6 Selective Estrogen Receptor Modulators (SERMs) [A.2] 101
4.2.2.7 Tamoxifen [A.2.b] 101
4.2.2.8 Aromatase Inhibitors (Al: block androgen to estrogen synthesis) [A.3] 102
4.2.2.9 Exemestane [A.3.b] 102
4.2.3 Androgen Hormone System [B] 102
4.2.3.1 Vinclozolin [B.l.b] 106
4.2.3.2 Procymidone [B.l.c] 106
4.2.3.3 Phthalates [B.2] 107
4.2.3.4 Semicarbazide [B.4] 115
4.2.3.5 Multiple Molecular Initiating Events [B.5] 115
2 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
4.2.3.6 Androgen Receptor Agonists [B.6] 116
4.2.3.7 Selective Androgen Receptor Modulators (SARMS) [B.7] 118
4.2.4 Thyroid 119
4.2.4.1 Environmental Contaminants and Thyroid Disruption 120
4.2.4.2 Literature Search and Analysis 121
4.2.4.3 Review of Three Major Targets for Nonmonotonic Thyroid Pathway Disruption 122
4.2.4.4 Results of Literature Analysis 129
4.2.4.5 Conclusions for Thyroid Studies 133
4.3 Human Studies and Epidemiology 136
4.3.1 Context for Human Studies in this Review 136
4.3.2 Interpreting Epidemiological Evidence 136
4.3.2.1 Multiple Exposures (Combined or in Sequence) 137
4.3.2.2 Exposure Assessment 137
4.3.3 Moving Forward - Converging Evidence 138
4.3.4 Summary 139
5. Conclusions 139
5.1 General Conclusions 139
5.1.1 Overall Conclusions: Estrogen, Androgen, and Thyroid 142
5.2 Central Scientific Questions and Answers 144
5.2.1 Do nonmonotonic dose responses (NMDRs) exist for chemicals and if so under what
conditions do they occur? 144
5.2.2 Do NMDRs capture adverse effects that are not captured using our current chemical
testing strategies (i.e., false negatives)? 145
5.2.3 Do NMDRs provide key information that would alter EPA's current weight of evidence
conclusions and risk assessment determinations, either qualitatively or quantitatively?
146
5.3 Summary 147
6. Bibliography 148
7. Appendices 177
3 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
List of Figures
Figure 1.1: Examples of Monotonic and Nonmonotonic Dose Response Curves 22
Figure 2.1: Key event dose-response framework organizes available information on the
multiple kinetic and dynamic events that occur between an initial dose and the effect of
concern 29
Figure 2.2: Perturbation by a xenobiotic of normal biological function may be overcome by
adaptive responses (homeostasis), while higher doses of the xenobiotic overwhelm the
adaptive capability, driving the system away from its normal biological functions to frank
toxicity 31
Figure 2.3: (A) Modeling of dose-response relationships for DNA-adduct levels as a function of
dose of an exogenous xenobiotic 33
Figure 2.4: Hypothalamic-pituitary-gonadal (HPG) axis 38
Figure 2.5: Regulation of thyroid hormone production, release, transport, metabolism and
action 39
Figure 2.6: Schematic diagram of the activation of a cytoplasmic steroid hormone receptor. ..40
Figure 3.1: Comparisons of in vitro points of inflection and in vivo or environmental exposure
concentrations using T47D KBIuc cell luciferase activity induction 51
Figure 4.1: Effects of a 21-d water-borne exposure to 17b-trenbolone on plasma
concentrations of (a) testosterone (T), (b) 17|3-estradiol (E2) and vitellogenin in female fathead
minnows 66
Figure 4.2: Effects of a 28-d dietary exposure to methyltestosterone on sex ratio in Nile tilapia.
68
Figure 4.3: Time-course (8-d exposure and 8-d recovery) effects of fadrozole on female fathead
minnow 76
Figure 4.4: Time-course (8-d exposure and 8-d recovery) effects of prochloraz on female
fathead minnow 77
Figure 4.5: Time-course (8-d exposure and 8-d recovery) effects of trenbolone on female
fathead minnow 78
Figure 4.6: Glans Penis Weight after exposure to a mixture of procymidone and dibutyl
phthalate 86
Figure 4.7: Comparison of results for preputial separation in male rats after exposure to EE2. 87
Figure 4.8: Comparison of results for dorsolateral prostate weight in male rats after exposure
to EE2 88
Figure 4.9: Representation of possible background dietary phthalate concentration 90
Figure 4.10: Histograms illustrating effects on vaginal opening and age at first estrus in DEHP
exposed rats 91
Figure 4.11: An adverse outcome pathway model for the effects of TDCs 121
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
List of Tables
Table 2.1: Endpoints with nonmonotonic dose-response and causative influences that are
dominant over different dose ranges* 32
Table 3.1: in vitro NMDRs as evaluated by the Danish Centre on Endocrine Disruptors DCED
(2013) 45
Table 4.1: Select studies from literature review displaying some evidence of non-monotonicity.
59
Table 4.2: NMDR From Studies Evaluating the Mammalian Estrogen Hormone System 93
Table 4.3: NMDR From Studies Evaluating Androgen Hormone System 104
Table 4.4: From a thyroid literature review of more than 1153 references, 1831 mammalian or
in vitro chemical-studies were evaluated for the presence of NMDR 134
5 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Contributors to the Document:
Contributions included formulating the Science Questions, structuring the approach to
evaluation, formulating conclusions, and drafting the main document and appendices.
Gerald Ankley
Jason Aungst
Don R. Bergfelt
Rory Conolly
Ralph Cooper
Kevin Crofton
Vicki Dellarco
Suzanne Fitzpatrick
Mary Gilbert
L. Earl Gray
Keith Houck
Abigail Jacobs
Kristi Jacobs
Ronald Lorentzen
Mary Manibusan
Eva McLanahan
Mark Miller
Sunni Mumford
Rita Schoeny
Jennifer Seed
Joe Tietge
Dan Villeneuve
Vickie Wilson
Douglas C. Wolf
EPA/ORD
FDA/CFSAN
EPA/OCSPP
EPA/ORD
EPA/ORD
EPA/ORD
EPA/OCSPP
FDA/CFSAN
EPA/ORD
EPA/ORD
EPA/ORD
FDA/CDER
FDA/CFSAN
FDA/CFSAN
EPA/OCSPP
EPA/ORD
EPA/ORD
NIH/NICHD
EPA/ORD
EPA/OCSPP
EPA/ORD
EPA/ORD
EPA/ORD
EPA/ORD
Reviewers
Internal EPA Review
Many of the above Contributors reviewed and commented on several interim drafts of
the document. Additional reviewers from USEPA are listed below.
Daniel Axelrad EPA/OA
Tina Bahadori EPA/ORD
Steve Bradbury EPA/OCSPP
Laurel Celeste EPA/OGC
6 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Michael Firestone EPA/OA
Ronald Hines EPA/ORD
Robert Kavlock EPA/ORD
Glenn Paulson EPA/OA
Kathleen Raffaele EPA/OSWER
Santhiny Ramasamy EPA/OW
External Peer Review
External peer review will be conducted by a panel of the National Research Council,
National Academies of Sciences U.S.A.
Communications and Outreach
Monica Linnenbrink EPA/ORD
Literature Search and Management
Ryan Jones EPA/ORD
Danielle Moore EPA/ORD
Diane LeBlond EPA/ORD
Technical Editing
William Wooge EPA/OCSPP
Other Acknowledgments
We wish to express our thanks to Stacy Katz (EPA/ORD) and Gail Robarge (EPA/ORD) for their
assistance in the preparation of this document. Sandcastle Technology Associates Ltd. and ICF
International served as contractors to USEPA in preparation of portions of this document. We
are grateful to other colleagues for their continual input through scientific dialogue.
7 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
At a Glance
EPA's State of the Science Evaluation: Nonmonotonic Dose
Responses as They Apply to Estrogen, Androgen, and Thyroid
Pathways and EPA Testing and Assessment Procedures.
Why we did this review
The risk assessment process is
dependent upon the selection of
doses and the characterization of
dose-response relationships. There
could be implications for selection
of doses in toxicology studies and
incorporation of results into risk
assessments.
Background
Nonmonotonic dose responses are
measured biologic effects with dose
response curves that contain a
point of inflection where the slope
of the curve changes sign at one or
more points within the tested
range. They are of specific concern
in the context of chemical testing
and risk assessment because they
do not follow the typically expected
linear or threshold dose response,
wherein increasing dose is
associated with increasing
frequency or severity of effect.
For further information contact
USEPA, Office of Research and
Development/Office of Science
Policy at
comments.nmdr(3epa.gov
The full report is at:
http://epa.gov/ncct/edr/nas-
review.html
What We Did
Developed a state of the science document providing a judgment on
the degree to which nonmonotonic dose-responses are evidenced in
the scientific literature and to evaluate the extent to which they may
impact USEPA's chemical testing and risk assessment.
What We Found
Nonmonotonic dose responses (NMDR) after exposure to xenobiotics
do occur in biological systems but are generally not common. Where
NMDRs were observed, biological endpoints closest to the molecular
initiating event were more likely to identify a point of inflection
(change of direction in slope) then those effects further downstream,
including the apical adverse outcomes. The goal of chemical testing is
to identify the potential for hazard after exposure to the xenobiotic of
concern, not to identify and describe 100% of all the possible
biological effects. As such, the current testing approaches perform
this function successfully and, based on the current evaluation, are
highly unlikely to mischaracterize a chemical that has the potential to
adversely perturb the endocrine system due to an NMDR.
What We Concluded
• NMDRs do occur in estrogen, androgen, and thyroid systems in
ecological and mammalian studies.
• NMDRs are not unexpected in vitro particularly when evaluating
high dose levels and/or lower-order biological endpoints in estrogen
androgen or thyroid systems.
• NMDRs are not commonly identified in estrogen, androgen, or
thyroid systems in vivo and are rarely seen in apical endpoints after
low-dose and/or long-term exposure.
• The nature of a dose response will vary over time, and
nonmonotonicity due to compensation may be observed.
• NMDRs observed in endocrine endpoints may be biologically
relevant and should be evaluated in context with the totality of the
available scientific data, including epidemiologic and human studies.
• There is currently no reproducible evidence that the early key
events involved in the expression of NMDRs that are identified at low
dose are predictive of adverse outcomes that may be seen in humans
or wildlife populations for estrogen, androgen or thyroid endpoints
• Therefore, current testing strategies are unlikely to mischaracterize,
as a consequence of NMDR, a chemical that has the potential for
adverse perturbations of the estrogen, androgen or thyroid
pathways.
8 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Fo reward
Within the community of environmental and public health scientists, the discourse around the
shape of the dose response curve has been energized in recent years especially as applied to
chemicals that induce adverse health outcomes through interaction with endocrine systems.
This topic and its relevance to the effectiveness of our chemical testing and evaluation
strategies is of great interest to EPA; thus, in 2012 EPA's Office of Chemical Safety and Pollution
Prevention (OCSPP) requested that the Office of Research and Development (ORD) conduct a
rapid and high priority expert review of the state of the science on the specific topic of lack of
monotonicity in dose-response for the endocrine disruptor mode of action. The particular
interest of the EPA review was effects mediated by alterations in the estrogen, androgen and
thyroid hormone systems. Several international agencies, including the Joint Research Centre
of the European Commission and the European Food Safety Authority, recently conducted
comparable reviews, each focusing on the areas most relevant to their mission.
In response to the request from OSCPP, senior scientists in ORD, together with other Agency
colleagues and federal partners conducted a review of more than 2000 scientific documents
over a six month period. ORD's approach to this rapid review was to capitalize on its extensive
in-house expertise in estrogen, androgen, and thyroid hormone systems. To this end and to
meet the timeline of this review, these expert groups initially worked independently drawing on
their intimate knowledge of the most relevant literature in their areas of expertise, before
gathering for the overall evaluation that led to the conclusions and responses to science
questions. Our conclusions are, of course, bounded by the content of the available scientific
literature, which was generally not developed with the express purpose of rigorously evaluating
the nature of dose-response relationships for endocrine-related molecular initiating events. It
should also be noted that the experimental animal literature is almost exclusively oriented
toward single chemical exposures, whereas the real world environmental situation in which
humans and wildlife are exposed is typically a complex mix of chemicals and other stressors,
making comparisons among these disciplines challenging.
Finally, we at EPA are committed to further evolving our Path Forward objectives that reflects
excellence in science, and is grounded in pragmatic, timely, relevant, and mission driven
objectives. This exercise is a seminal case study in our progress as it reflects the agility and
responsiveness of EPA science and scientific staff in the ability to address a complex topic of
high import for EPA and the nation. The awareness of the importance of the issue motivated
the request for this NRC review. On behalf of the Agency, let me thank you in advance for your
efforts to evaluate our document and provide guidance on how we might improve it. We look
forward to your engagement - acknowledging that your task is no less daunting and time-
sensitive than ours was.
Robert Kavlock
Deputy Assistant Administrator for Science
Office of Research and Development
US Environmental Protection Agency
9 | Page
-------
Draft Deliberative
Executive Summary
State of the Science: Nonmonotonic Dose Responses V7
The U.S. Environmental Protection Agency (USEPA) chartered a group of scientists to evaluate
the impact of potential nonmonotonic dose response relationships (NMDRs), as they apply to
estrogen-, androgen-, and thyroid-based modes of action and the Agency's testing and
assessment procedures. The goals were to develop a state of the science document providing a
judgment on the degree to which NMDR are identifiable in the scientific literature, and to
evaluate the extent to which their existence may impact USEPA's chemical testing and risk
assessment programs.
NMDRs are measured biologic effects with dose response curves that contain a point of
inflection where the slope of the curve changes sign at one or more points within the tested
range. NMDRs are of specific concern in the context of chemical testing and risk assessment
because they do not follow expected dose response curves, wherein increasing dose is
associated with increasing frequency or severity of effect. Although, the USEPA is concerned
about the full domain of potential NMDR impacts, the current analysis focuses solely on effects
caused by endocrine disrupting chemicals (EDC) that act via perturbation of the estrogen,
androgen, or thyroid hormone systems. For the purposes of this document, an EDC is defined
as an exogenous substance or mixture that alters function(s) of the endocrine system and
consequently causes adverse health effects in an intact organism, or its progeny, or
(sub)populations WHO (2002). Consistent with existing USEPA and National Toxicology
Program (NTP) activities and assessments, the current state-of-science document considered an
effect to be adverse only if it displayed a change in morphology, physiology, growth,
development, reproduction, or life span of a cell or organism, system, or (sub)population that
results in an impairment of functional capacity, an impairment of the capacity to compensate
for additional stress, or an increase in susceptibility to other influences. Where appropriate,
this document also distinguished adverse effects from adaptive responses, which has been
defined as a process whereby a cell or organism responds to a xenobiotic but without
impairment of function.
The USEPA has a well-documented and long standing interest in the shape of dose response
curves, particularly at the lowest end of the range of exposures resulting in observed effects.
USEPA has published guidance for dose response assessment, the step in risk assessment that
estimates the likelihood of adverse effects at environmental levels of exposure. The USEPA is
cognizant of the fact that the risk assessment process, therefore, is highly dependent upon the
selection of dose levels in toxicology studies, the choice of measured endpoints and the ensuing
characterization of dose-response relationships. The Agency has long recognized the
importance of understanding the determinants that influence the biological response to
exposures over a wide range of dose levels, including those that are environmentally relevant.
10 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
The USEPA also recognizes that if NMDRs are associated with low dose levels (i.e., the range of
typical human exposures or doses lower than those typically used in standard testing protocols)
of exposure to EDCs, there could be implications for the selection of dose levels in standard
toxicology studies and incorporation of results from such studies into risk assessments.
In a June 15, 2012 memorandum, the USEPA's Office of Chemical Safety and Pollution
Prevention's (OCSPP) Assistant Administrator requested the USEPA Office of Research and
Development (ORD) "evaluate the impact of potential nonmonotonic dose response
relationships, particularly as they apply to endocrine-based modes of action, on EPA testing and
assessment procedures." In response, ORD convened a working group to develop this NMDR
State of the Science Assessment. The working group included scientists, managers and
technical experts from USEPA/ORD and OCSPP, the Food and Drug Administration (FDA), the
National Institute of Environmental Health Sciences (NIEHS), and the National Institute of Child
Health and Human Development (NICHD).
Although NMDRs have been reported in a wide range of biological systems, the current state of
the science document is limited to an overview of biological and statistical concepts pertinent
to NMDRs for adverse health effects of chemicals acting on the estrogen, androgen, or thyroid
hormone systems. This assessment builds on existing scientific data, reports, and reviews but is
not intended to cover all endocrine systems that may be disrupted by environmental
exposures, to assess test methodologies for detecting EDCs, or to address specific EPA testing
methodologies, risk assessments or risk management decisions. Rather, it evaluates the peer-
reviewed scientific literature in which associations between chemical exposures and adverse
outcomes have been demonstrated or hypothesized to occur via perturbation of estrogen,
androgen, or thyroid hormone systems. The roles of pharmacokinetics, key molecular events,
and modes of action as they impact the dose response curves for early developmental and
reproductive endpoints and statistical analyses, are considered. We used an expert-driven
approach as the most expedient way to accomplish our goal of a state of the science evaluation
that cut across several disciplines.
We identified three central scientific questions to be addressed in the current state of the
science assessment. The central scientific questions, and summarized responses, are these.
1. Do nonmonotonic dose responses (NMDRs) exist for chemicals and if so under
what conditions do they occur?
Yes, we concluded that exposures to chemicals can result in NMDRs for specific
endpoints. NMDRs arise from complex relationships between the dose of toxicant at its
target site and the effect of interest WHO (2012). NMDRs are biologically plausible and
11 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
can arise when the biological system that is activated in response to toxicant exposure
consists of at least two activities that can act in opposition to each other WHO (2012;
Conolly and Lutz (2004). We determined that NMDRs are more frequently identified in
these types of studies: in vitro studies, high-dose range studies, and short-term studies.
Assays that provide data at a lower level of biological organization (such as proteomics
or transcriptomics) are more likely to identify NMDRs than studies that provide data on
apical adverse events further downstream from the molecular initiating event.
Reproducibility of NMDRs is important in establishing plausibility of a response and its
potential applicability as part of the hazard characterization. Factors that influence
reproducibility include:
¦ Study design - dose selection, sample size, organism strain, diet, housing
environment, statistical methods;
¦ Robustness of physiology - physiologic compensation producing changes in slope;
and
¦ Competing processes- induction of metabolism, repair, or independent
mechanisms.
2. Do NMDRs capture adverse effects that are not captured using our current
chemical testing strategies (i.e., false negatives)?
There are certainly adverse biological changes that may occur in a nonmonotonic
manner that would not be captured using current testing strategies. No testing strategy
is able to assess all potential adverse effects, for all biological systems, in all tissues, for
all species, in all developmental time points.
As the work group progressed with its review and discussions of the science, it became
clear that our second question needed to be further defined so that it could be more
accurately and fully answered in the context of the science as evaluated. Thus, question
2 was further expanded for clarification.
¦ Are there adverse effects with NMDRs that are not being identified using the current
chemical testing strategies?
¦ Are there NMDRs for adverse effects below the no observed adverse effect levels
(NOAELS) or benchmark doses (B rived from the current testing strategies?
¦ Do EPA chemical testing strategies detect relevant adverse effects for chemicals
which prodi specific endpoints?
12 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Chemicals that operate through endocrine modes of action (MoA) have multiple targets
across organs, tissues, and cellular systems in various species, and across all life stages.
It is not possible or feasible for chemical testing to measure or analyze all possible
endpoints for all chemical MoA in all tissues. The objective of USEPA's chemical testing
strategy is not to identify all possible adverse effects, but rather, to identify sensitive
endpoints relevant to human or ecological health, providing confidence that adverse
effects are not being induced at dose levels below what was determined to be a NOAEL.
For estrogen, androgen or thyroid MoA that provide adequate information to make an
assessment, our evaluation shows that there is not sufficient evidence of NMDRs for
adverse effects below the NOAELS or BMD derived from the current testing strategies.
For some MoA, however, the scientific database remains too limited to conclude this
with certainty.
While there are biological changes that may occur in a nonmonotonic manner in the low
dose region, our review indicates that reproducible NMDRs for adverse effects occur in
the high dose region of the dose response curve. Thus, the current testing approaches
do not fail to identify or establish appropriate NOAELS in the low rose range of
exposure, even if not all effects for every chemical are identified. The extensive
evaluation conducted in the present review as well as almost two decades of experience
with screening assays for hazard identification indicate that these assays do not fail to
detect chemicals with endocrine activity for the estrogen or androgen hormone
systems. Dose response assessment is not an issue for screening assays. NMDRs would
be problematic only if a chemical with estrogen, androgen, or thyroid activity produced
an effect in vivo at a dose below those used in screening, and the chemical had no effect
on estrogen, androgen, or thyroid related endpoints at the higher screening dosage
levels. Although, such NMDRs have been hypothesized they have not been
demonstrated reproducibly, and none were found in the present evaluation.
Our assessment of the adequacy of the current testing assays concludes that a number
of standardized short- and long-term assays are sensitive in detecting chemicals that
interfere with the estrogen, androgen and thyroid signaling pathways. Specifically, the
EDSP screening battery can detect disruption of these pathways using combined in vitro
and in vivo assays in mammalian and aquatic models. Standard multigenerational test
guidelines have measures that are sensitive to disruption of the estrogen and androgen
signaling pathways. While these studies are considered the current standard for
assessing the potential of a chemical to be a reproductive toxicant and for use in setting
NOAELS, they are not without limitations. USEPA testing strategies are reviewed
13 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
periodically to assure they are incorporating the most sensitive and biologically relevant
endpoints.
Further, if an objective of testing is to define the shape of the dose response curve and
thereby identifying potential NMDRs, then the three treated groups and a control group
used in current guideline studies may not be sufficient for this purpose. Modifications
that could lead to a more clearly defined dose response characterization and increase
the statistical power to detect low dose effects may be appropriate in specific instances.
3. Do NMDRs prowicle key information that would alter EPA's current weight of
evidence conclusions and risk assessment determinations, either qualitatively
01* quantitatively?
Data from studies in which NMDRs are identified may be biologically relevant and as
such should be evaluated in context with the totality of the available scientific data in
weight of evidence (WoE) conclusions and risk assessment determinations. These data
should be considered and analyzed, as all data are, and factored into the WoE based on
standard criteria including, but not limited to, conduct of the studies, representation of
biological processes that are relevant to the evaluation, biological plausibility, and
reproducibility. NMDRs can have impact on both qualitative and quantitative risk
assessments, but cannot be considered in isolation from other data for the chemical and
biological response being considered.
In summary, nonmonotonic dose responses after exposure to xenobiotics do occur in biological
systems but are generally not common. Where NMDRs were observed, biological endpoints
closest to the molecular initiating event were more likely to identify a point of inflection; those
effects further downstream, including the apical adverse outcomes, are most commonly
monotonic in their dose-response. The goal of chemical testing is to identify the potential for
hazard after exposure to the xenobiotic of concern, not to identify and describe 100% of all the
possible biological effects. As such, the current testing approaches perform this function
successfully and, based on the current evaluation, are highly unlikely to mischaracterize a
chemical that has the potential to adversely perturb the endocrine system due to an NMDR.
14 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
¦ si idusions:
l.
NMDRs do occur in estrogen, androgen, and thyroid systems as evidenced
in ecological and mammalian studies.
2.
NMDRs are not unexpected in vitro particularly when evaluating high dose
levels and/or lower-order biological endpoints in estrogen, androgen, or
thyroid systems.
3.
NMDRs are not commonly identified in estrogen, androgen, or thyroid
systems in vivo and are rarely seen in apical endpoints after low-dose
and/or long-term exposure.
4.
The nature of a dose response will vary over time, and nonmonotonicity
due to compensation may be observed.
5.
NMDRs observed in endocrine endpoints may be biologically relevant and
should be evaluated in context with the totality of the available scientific
data, including epidemiologic and human studies.
6.
There is currently no reproducible evidence that the early key events
involved in the expression of NMDRs that are identified at low dose are
predictive of adverse outcomes that may be seen in humans or wildlife
populations for estrogen, androgen or thyroid endpoints.
7.
Therefore, current testing strategies are unlikely to mischaracterize, as a
consequence of NMDR, a chemical that has the potential for adverse
perturbations of the estrogen, androgen or thyroid pathways.
15 | Page
-------
Draft Deliberative
Acronyms
State of the Science: Nonmonotonic Dose Responses V7
33-HSD 33-Hydroxysteroid Dehydrogenase
AGD Anogenital Distance
ANOVA Analysis of Variance
AOP Adverse Outcome Pathway
AR Androgen Receptor
ARE Androgen Response Element
BDNF Brain-Derived Neurotrophic Factor
BBP Butyl Benzyl Phthalate
BMD Bench Mark Dose
BPA Bisphenol A
BW Body Weight
cAMP Cyclic adenosine monophosphate
CCB 1-Chloro 4-(Chloromethyl) Benzene
ChAT Choline acetyltransferase
c-Kit Protein on surface of many cell types
CI04 Perchlorate
CNS Central Nervous System
CYP Cytochrome P450s
DBP Dibutyl Phthalate
DCHP Dicyclohexyl Phthalate
DEHP Di-2-ethylhexyl Phthalate
DES Diesthylstilberstrol
df Degrees of Freedom
DHP Dihexyl Phthalate
DHT Dihydrotestosterone
DI1/DI2/DI3 Deiodinase 1, 2, or 3
DIBP Diisobutyl Phthalate
DINP Diisononyl Phthalate
DPeP Dipentyl Phthalate
E Estrogen
E2 Estradiol
EAS Endocrine Active Substances
ED Endocrine Disruptor
EDC Endocrine Disrupting Chemical
EDSP Endocrine Disruptor Screening Program
EDSTAC Endocrine Disruptor Screening and Testing Advisory Committee
EE2 Ethynyl Estradiol
16 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
EFSA
European Food Safety Agency
ER
Estrogen Receptor
ERR
Estrogen Related Receptor
ETU
Ethylene thiourea
F1/F2/F3
1st, 2nd, or 3rd, Filial Generation
FSH
Follicle Stimulating Hormone
GD
Gestation Day
GLP
Good Laboratory Practices
GnRH
Gonadotropin Releasing Hormone
GSH
Glutathione
GSI
Gonadal Somatic Index
HCB
Hexachlorobenzene
HPG
Hypothalamic Pituitary Gonadal
HPT
Hypothalamic Pituitary Thyroidal
HRE
Hormone response element
Insl3
Insulin-like Factor 3
IPCS
International Programme on Chemical Safety
IRIS
Integrated Risk Information System
Kd
Equilibrium dissociation constant
KEDRF
Key Event Dose Response Framework
LABC
Levator Ani Bulbo-Cavernosus
LE
Long Evans
LH
Luteinizing Hormone
LOEL
Lowest Observed Effect Level
MAP kinase
Mitogen-activated protein kinase
MBP
Myelin Basic Protein
MCT
Monocarboxylate Transporter
MEHP
Methylethyl Hexyl Phthalate
MIE
Molecular Initiating Event
MMI
Methimazole
Mo A
Mode of Action
MT
17a-methyltestosterone
NCTR
National Center for Toxicological Research
NIEHS
National Institute for Environmental Health Sciences
NIS
Sodium-Iodine Symporter
NMDR
Nonmonotonic Dose Response
NMDRC
Nonmonotonic Dose Response Curve
NOAEL
No Observed Adverse Effect Level
NOEL
No Observed Effect Level
17 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
NRC National Research Council
NTP National Toxicology Program
OATP Organic Anion Transporting Polypeptide
OECD Organization for Economic Cooperation and Development
OHF Hydroxyflutamide
P Parental Generation
PBPK Physiologically Based Pharmacokinetic
PCB Polychlorinated Biphenyl
PCR Polymerase Chain Reaction
PHAH Polyhalogenated Aromatic Hydrocarbon
PND Postnatal Day
PNECs Predicted no effect concentrations
PPS Preputial Separation
PTI l-methyl-3-propylimidazole -2-thione
PTU Propylthiouracil
RIA Radioimmunoassay
SARM Selective Androgen Receptor Modulator
sc Subcutaneous
SD Sprague Dawley
SE Standard Error
SEM Semicarbazide
SERM Selective Estrogen Receptor Modulator
SR-B1 Scavenger Receptor B-l
StAR Steroid Acute Regulatory Protein
T Testosterone
T3 Triiodothyronine
T4 Thyroxine
TB Trenbolone
TBG Thyroid Binding Globulin
TCDD Tetrachlorodibenzodioxin
TDC Thyroid Disrupting Chemicals
TH Thyroid Hormone
TPO Tyroperoxidase
TR Thyroid Receptor
TRH Thyroid Releasing Hormone
TSH Thyroid Stimulating Hormone
TTR Transthyretin
UGTs/UDGPTs Uridine Diphophogluronyltranferases
VO Vaginal Opening
18 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
VTG Vitellogenin
WHO World Health Organization
WoE Weight of Evidence
19 | Page
-------
Draft Deliberative
1. Introduction
State of the Science: Nonmonotonic Dose Responses V7
In June of 2012, USEPA chartered a group of scientists, managers, and technical experts to
evaluate the existence and impact of nonmonotonic dose response (NMDR) relationships, as
they apply to endocrine-based modes of action, within current USEPA testing and assessment
procedures. The group was charged to
develop a state of the science document
providing a weight of evidence
judgment on the degree to which
NMDRs are identified in the scientific
literature associated with evaluating the
effects of endocrine active agents and
the degree to which their existence may
impact USEPA's chemical testing and
risk assessment.
Although USEPA is concerned about the full domain of potential NMDR impacts, the current
analysis focuses solely on endocrine disrupting chemicals (EDCs) impacting the estrogen,
androgen, or thyroid hormone systems. It is not intended to be a comprehensive treatise on
any specific chemical, mode of action
(MoA), or toxicological effect. The current
analysis focuses primarily on effects
observed in the developing organism or on
reproductive function. Moreover, it does
not address cumulative effects of chemical
exposures, except very broadly in the
context of MoA and observed dose-effect
discussions.
This document relied upon publicly available, published literature. The authors of this
document have commented in some places on statistical tests or interpretations of data in that
literature. This state of the science evaluation did not include formal statistical or
mathematical reanalysis of the published data.
A structured, consistent and formal systematic review of the literature, with defined criteria for
study selection, may have been desirable. This has been recommended recently by both the
National Research Council and our Science Advisory Board for exploring individual issues or
single chemicals. Given the established timelines and availability of resources for this project,
we used our expert-driven approach as the most expedient way to accomplish our goal of a
Non Monotonic Dose Responses (NMDRs) -
measured biological effects with dose
response curves that contain a point of
inflection where the slope of the curve
changes sign at one or more points within
the tested range.
Endocrine Disrupting Chemical (EDC)~ an
exogenous substance or mixture that
alters function(s) of the endocrine system
and consequently causes adverse health
effects in an intact organism, or its
progeny, or (sub)populations. (WHO 2002)
20 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
state of the science evaluation that cut across several disciplines. Guidelines for systematic
review are evolving, and we look forward to applying new methodologies to complex questions
such as those posed in this review.
In Section 5, Conclusions, the EPA work group has characterized its overall confidence in the
extent and circumstances under which NMDRs are observed. Throughout the document the
EPA workgroup has noted knowledge gaps, uncertainties, quality of data, and some scientific
issues in the interpretation of dose response data. By doing so, we have conveyed the
limitations of the current evaluation.
The data analyses and conclusions of this state of science assessment represent the authors'
current thinking on NMDR. This document does not establish or necessarily reflect USEPA
policy or test guidelines or guidance. Its purpose is to provide scientific information to aid
USEPA senior leadership, program managers, and risk assessors in their effort to develop
policies grounded, at a most fundamental level, in sound science.
For other general information about this assessment or other questions, the reader is referred
to the USEPA, Office of Research and Development/Office of Science Policy at
comments. nmdr@epa.gov.
1.1 Background
The USEPA has a well-documented and long standing interest in the shape of dose response
curves, particularly at the lowest end of the range of exposures resulting in observed effects.
USEPA has published guidance for dose response assessment (e.g., U.S. EPA (2005a)). the step
in risk assessment that estimates the likelihood of adverse effects at environmentally relevant
levels of exposure. The dose response steps of the risk assessment process, therefore, are
highly dependent upon the selection of dose levels in toxicology or exposure brackets in human
studies and the ensuing characterization of dose-response relationships. The Agency has
recognized the importance of understanding the determinants that influence the biological
response to exposures over a wide range of dose levels, particularly those that are
environmentally relevant.
Risk assessors extrapolate from effects observed in experimental systems to environmental
exposures, in some instances relying on assumptions of low dose linearity or of an exposure
associated with no adverse effect (e.g., threshold). With few exceptions (such as essential
elements), it is expected that lower exposures to biologically active chemicals will result in
decreased measures of response. Responses that consistently increase or decrease, across a
range of exposures, are said to exhibit monotonic dose responses.
21 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
This expectation has been subject to challenge e.g., Welshons et al. (1999; Nagel et ai. (1997;
vom Saal et al. (1997). According to these reports, NMDR were demonstrated when laboratory
animals were exposed to "low dose" levels of putative EDCs. The question was raised whether
these chemicals, by virtue of their ability to interfere with the endocrine system, were unique in
their dose response relationships. Early on, other scientists were unable to replicate these
findings Ashby et ai. (1999; Cagen et ai. (1999a), which created controversy and stirred debate
as to the validity of the original reports of NMDRs.
NMDR are measured biological effects with dose response curves that contain a point of
inflection where the slope of the curve changes sign at one or more points within the tested
range. As described in Conolly and Lutz (2004), NMDRs are most commonly reported with
either a decrease in the level of response observed in controls at low dose(s) followed by an
increase at higher doses (called a U-shape or J-shape); or vice versa (an "inverted U" or p-
shape) WHO (2012).
Monotonic Curve Monotonic Curve Non-monotonic Curve Non-monotonic Curve
Figure 1.1: Examples of Monotonic and Nonmonotonic Dose Response Curves.
(Reproduced from Fagin (2012), Nature 490: 462-465)
The USEPA recognized that if NMDRs were associated with low-dose levels of exposure to EDC,
there could be implications for the selection of dose levels in standard toxicology studies and
the incorporation of results from such studies into risk assessments.
The USEPA convened the Endocrine Disruptor Screening and Testing Advisory Committee
(EDSTAC), an expert panel to provide advice on the development of a screening program for
endocrine disrupters. The panel addressed the "low dose hypothesis," as it was referred to at
that time, in its 1998 report to EPA EDSTAC (1998). The EDSTAC report stated that if low-dose
phenomena are reproducible, generalizable, and related to adverse effects, the implications for
regulatory toxicity testing and risk assessment are profound. Evaluation of responses from
exposure to EDCs at doses below traditional toxicology studies or at environmental levels is
22 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
important but different from
determining the shape of the dose-
response curve across many doses,
including very low exposure levels.
While an NMDR can coincide with low
exposures, it can also be identified at any
point along the dose-response
continuum which is a separate issue
from determining the lowest dose that can result in an effect. The report further advised that
more research to address the nature of the dose-response curves for exogenous EDC was
necessary to inform toxicology study designs, the use of data in risk assessments, and to resolve
the then underlying uncertainties and controversy.
Responding to the EDSTAC conclusions is complicated by the fact that no single definition of
"low dose" is consistently used by the scientific community. Low dose is a relative term that is
evaluated in a chemical- and testing-specific context for environmental and biological
relevance. Several different definitions have been proposed and used e.g., Welshons et al.
(2006; Melnick et al. (2002; Brucker-Davis et al. (2001).
For Example:
¦ "Any biological changes occurring in the range of typical human exposures" NTP (2001)
¦ "Doses lower than those typically used in standard testing protocols, i.e., doses below
those tested in traditional toxicology assessments" NTP (2001)
¦ "A dose below the lowest dose at which a biological change (or damage) for a specific
chemical has been measured in the past, any dose below the lowest observed effect
level (LOEL) or lowest observed adverse effect level (LOAEL)" Welshons et al. (2006)
The current NMDR state of the science assessment uses the National Toxicology Program (NTP)
definitions: a biological change occurring in the range of typical human exposures or at doses
lower than those typically used in standard testing protocols NTP (2001).
Likewise there are multiple definitions of adverse effect, which are used by risk assessors. In
2002, EPA defined an adverse effect as "a biochemical change, functional impairment, or
pathologic lesion that affects the performance of the whole organism, or reduces an organism's
ability to respond to an additional environmental challenge" USEPA (2002). This NMDR state of
the science document considers the concepts of mode of action, toxicity pathways and adverse
outcome pathways (see section 2.2); in particular we consider the adaptive capability of
organisms or populations that encounter environmental stress. To that extent we include the
systems biology oriented description of adversity by Keller et al. (2012).
Low Dose Effect- a biological change
occurring in the range of typical human
exposures or at doses lower than those
typically used in standard testing
protocols NTP (2001).
23 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
In 2000, an NTP panel concluded that the traditional multigeneration reproduction testing
paradigm used by EPA has not revealed major reproductive or developmental effects in
laboratory animals exposed to endocrine active agents at doses approaching the no observed
adverse effect levels (NOAELs) set by the standardized testing protocols NTP (2001). The peer
review panel evaluated data from major, selected studies that supported the presence or
absence of low-dose effects in laboratory animals and that would be relevant for human health
assessments. In particular, the NTP panel was interested in evaluating the scientific
underpinnings of dose-response relationships for reproductive toxicants, including
nonmonotonicity within the low dose range.
The NTP panel also recommended that EPA periodically review testing paradigms used for
assessments of reproductive and developmental toxicity to see if changes are needed regarding
dose selection, animal model selection, age when animals are evaluated, and the endpoints
being measured following exposure to endocrine active agents. In 2002, the USEPA issued a
statement that additional research was needed to support a better understanding of the low-
dose hypothesis USEPA (2002). The statement noted that an improved understanding of the
mechanisms of action by which endocrine-active agents exert their effects would determine
whether existing testing protocols needed modification
(http://www.epa.gov/endo/pubs/edmvs/lowdosepolicy.pdf ). Further, the USEPA indicated
that it would be monitoring the outputs of ongoing research for applicability to its Endocrine
Disruptor Screening Program (EDSP). The statement left open the possibility that USEPA could
require low-dose testing on a case-by-case basis if relevant information on specific chemicals
Adverse Effect - a measured endpoint that displays a
change in morphology, physiology, growth, development,
reproduction, or life span of a cell or organism, system,
or population that results in an impairment of functional
capacity, an impairment of the capacity to compensate
for additional stress, or an increase in susceptibility to
other influences Keller et al. (2012).
Boettcher et al. (2011: Hirabayashi and Inoue (2011: EFSA (2010: Diamanti-Kandarakis et al.
(2009: Kortenkamp (2008: Sekizawa (2008: Kamrin (2007: Kortenkamp (2007: Scholze and
Kortenkamp (2007: Welshons et al. (2006: Owens and Chaney (2005: Haseman et al. (2001).
became available.
There have been many
other workshop reports,
reviews, statements, and
other publications on the
issue of low-dose, NMDR,
and endocrine disruptors
e.g., Birnbaum (2013:
DCED (2013: UNEP (2013:
Birnbaum (2012:
Vandenberg et al. (2012:
Zoeller et al. (2012:
24 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Congress also showed an interest in this topic. In 2009, the Endocrine Disruption Prevention
Act (HR 4190 and S 5210) was introduced with the following statement: "Congress finds
that traditional toxicology and risk assessment, which evaluate one chemical at a time, and
only at high concentrations, have failed to sufficiently address the effects of low doses of
chemicals...". At a 2010 hearing on EDCs in drinking water, Representative Jim Moran testified
that even "infinitesimally low levels of exposure" to EDCs could cause adverse effects, quoting a
2009 statement from the Endocrine Society. In its response to Representative Moran, the
USEPA committed to convening a workshop to bring together expert scientists to share their
findings, characterize the state of the science, identify data gaps and determine the best way to
address the issue.
This internal EPA 2011 workshop was convened for the following purposes: 1) to re-examine
the state of the science with an eye toward determining whether current study designs and/or
risk assessment approaches should be modified; 2) to determine whether there is a need for an
international workshop, and if so, what its goals would be, which organizations could be
potential co-organizers, and how the path forward should be mapped; and 3) to develop a
statement summarizing the USEPA's perspective on NMDRCs and endocrine disruptors. This
state of the science evaluation is an outgrowth of that workshop.
Contemporary reviews that deal in total or in part with NMDR have been prepared for different
purposes. Vandenberg et al. (2012)reviewed over 800 papers with the objective of evaluating
the literature for the existence of NMDR in the context of biological mechanisms for these
responses. Based on their description of application of a weight of evidence (WoE) approach,
they conclude that "when nonmonotonic dose-response curves occur, the effects of low doses
cannot be predicted by the effects observed at high doses. Thus, fundamental changes in
chemical testing and safety determination are needed to protect human health." Zoeller et a I.
(2012) note their objective is to use principles of endocrinology to make recommendations for
testing programs designed to identify EDC. The adequacy or relevance of testing and screening
programs for EDC has also been addressed by the Danish Centre on Endocrine Disrupters DCED
(2013). which concluded the following: "The current information requirements in REACH are
not designed for the identification of endocrine disrupters, although certain endpoints and
assays may give some indication of endocrine disrupting effects." By contrast to other reviews,
this conclusion was not based solely on the observation of NMDR.
25 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
The European Commission Joint Research Centre (JRC 2013) states that the purpose of the
European Union Endocrine Disrupters Expert Advisory Group is: "to provide detailed reflections
on scientific issues relevant to endocrine disrupting substances, not specific to any regulatory
framework, including advice/orientation on scientific criteria for the identification of endocrine
disrupting substances." They agreed upon elements for identification of EDCs, but they were
not able to provide a full evaluation of the adequacy of the currently available test methods.
The JRC report relies in part on Kortenkamp et al. (2011).
EFSA (2013) reviewed existing information
related to the testing and assessment of
endocrine active substances (EASs) and
EDCs. They set three criteria that define
an EDC: "the presence of i) an adverse
effect in an intact organism or a
(sub)population; ii) an endocrine activity;
and iii) a plausible causal relationship
between the two". EFSA noted the lack of consensus in the scientific community on the
reproducibility and relevance to risk assessment of NMDR; they stated that the quality and
robustness of data for studies reporting NMDR should be assessed, as it is for any other studies.
Finally, the WHO UNEP report UNEP (2013) was not focused on NMDR, but rather on the state
of the science for links between EDC exposure and human disease, with a recommendation for
reducing exposures to such compounds. However, section 1.2.4 of UNEP (2013) describes
some instances and rationales for NMDR in vitro. Most of the above contemporary reviews
suggest the need for additional review of dose response relationships for potential EDC.
In response to recommendations for continued review of nonmonotonicity, the USEPA
convened a working group to generate this state of the science assessment of NMDRs for
chemicals acting via the estrogen, androgen, or thyroid hormone systems, in support of
USEPA's continued review of testing and risk assessment processes. Congruent with other
contemporary reviews (e.g., DCED (2013)) we based our approach to evaluating the literature
on concepts common to WoE.
The principles and criteria for weighing and integrating different lines of evidence articulated in
existing EPA documents U.S (2005. 2002. 1998. 1996. 1991) are applicable to evaluating data on
endocrine active chemicals. USEPA refers to the WoE approach as "...a collective evaluation of
all pertinent information so that the full impact of biological plausibility and coherence is
adequately considered." U.S (1999). In its recommendations to EPA, the Endocrine Disruptor
Screening and Testing Advisory Committee (EDSTAC) referred to the WoE approach as "...a
process by which trained professionals judge the strengths and weaknesses of a collection of
Weight of Evidence (WoE) - an integrative
and interpretive process routinely used by
EPA and other risk assessors to evaluate
health and ecological toxicity in a manner
that takes into account all relevant
scientific and technical information.
26 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
information to render an overall conclusion that may not be evident from consideration of the
individual data" EDSTAC (1998).
The USEPA uses a WoE process to make determinations on the hazard and risk from exposure
to a chemical stressor; for an example see the WoE document on the EDSP website
(www.epa.gov/endo). All WoE processes involve a number of common steps: assembling the
relevant data; evaluating that data for quality and relevance; and an integration of the different
lines of evidence to support conclusions concerning a property of the substance. WoE
considers both negative and positive studies, but it is not a simple tallying of the number of
plusses and minuses. The significant issues, strengths, and limitations of the data and the
uncertainties that deserve serious consideration are presented, and the major points of
interpretation highlighted.
In our evaluation we did not assign categories of likelihood for NMDR, as is done in many types
of hazard identification (see for example descriptions of WoE categories for human cancer in
U.S. EPA (2005a) and the WoE categories in DCED (2013)). Nor was there an attempt to design
exclusion/inclusion criteria for studies uncovered in all literature searches; for reasons of
resource limitation, this was done primarily for the description of the data on the thyroid
pathway. For sections of our review, judgments of likelihood were based on relevant examples
rather than a formal WoE. All discussions, however, considered instances wherein NMDR were
demonstrated or were not substantiated; the description of the latter studies can be found in
appendices to this document.
1.2 Scope of the Current Evaluation
NMDRs may be observed for specific endpoints in any biological system, but the scope of this
state of the science document is limited to chemicals affecting the estrogen, androgen and
thyroid pathways. Our review is narrowly focused on discussion of existence of, biological
rationales for, and relevance of NMDRs to risk assessment. When identified, endpoints that
may be affected by multiple modes of action or alternative pathways were considered in the
analysis for compounds with primary modes of action via the estrogen, androgen, or thyroid
systems. In these cases, alternative pathways were not specifically investigated. This state of
the science document also does not consider cumulative stressors.
This evaluation builds on existing scientific data reviews and documents and is not intended to
cover all endocrine systems that may be disrupted by environmental exposures; to assess test
methodologies for detecting EDCs; or to address specific EPA testing methodologies, risk
assessments or risk management decisions. Our document does not attempt to establish a link
between endocrine active compounds and human disease, nor does it make a statement on the
27 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
likelihood of EDC to have effects at low or environmental exposure levels. This state of the
science document does not deal generally with demonstrations of effects at low dose, but
rather with documentation and explication of NMDR as defined in section 1.2.
1.3 iral Scientific Questions
Prior to the formation of the working group an NMDR Steering Committee identified three
central scientific questions to be addressed in the current state of the science assessment. This
state of the science evaluation addresses these three central scientific questions for the
estrogen, androgen and thyroid pathways.
1. Do nonmonotonic dose responses (NMDR) exist for chemicals and if so under what
conditions do they occur?
2. Do NMDRs result in adverse effects that are not captured using our current chemical
testing strategies (i.e., false negatives)?
3. Do NMDRs provide key information that would alter EPA's current weight of evidence
conclusions and risk assessment determinations, either qualitatively or quantitatively?
The rest of the document deals in depth with the evaluation of information to answer the
above questions.
2. General Biological Concepts and Statistical Considerations for NMDR
The concepts of Mode of Action (MoA) (described in Boobis et al. (2008; U.S (2005; Sonich-
Mullin et al. (2001) guided the evaluation of the data related to NMDR in the present report.
MoA is defined as a set of key events starting from interaction of a xenobiotic with a cellular
receptor and proceeding to the apical effect. The roles of pharmacokinetics in determining the
shapes of dose response curves are important in the interpretation of NMDRs. The concepts
presented in this section are generally applicable to mammals including humans and to
nonmammalian species (ecological toxicology) unless otherwise noted. Although exposure is
not a focus of this evaluation, it should be noted that the concept of "low dose" implies some
knowledge of actual or expected exposures.
2.1 Pharmacokinetic Processes
Many pharmacokinetic processes are saturable, such as energetically mediated transport across
membranes, bioactivation, and detoxification. The rates of these kinds of processes are initially
linear with concentration but become sublinear as saturation is approached. Pharmacokinetics
can, thus, impart nonlinearities to the dose-response curve. Characterization of concentrations
28 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
or closes at which such nonlinearities occur, relative to the closes used in experimental work
and environmental exposures, reduces uncertainty in dose-response characterization and in
any associated risk assessments. Physiologically based pharmacokinetic (PBPK) models are
important tools for these characterizations. Appropriately developed and validated PBPK
models are useful for predicting pharmacokinetic behaviors for situations that lack
experimental data, is often the case with interspecies and high-to-low dose extrapolations
Mclanahan et al. (2012).
Initial Dose (intake or exposure)
r
Biological interaction or process (absorption)
Modification of
absorption
Interaction or process {e.g. transport/distribution/excretion)
Interaction or process (e.g., metabolism)
Modification of large I tissue
exposure
4
Interaction or process (in target tissue)
Homeosratic compensation,
adaptation, repair
I
ULTIMATE EFFECT OF CONCERN
Homeosiatic compensation,
adaptation, repai r
Figure 2.1: Key event dose-response framework organizes available information on the
multiple kinetic and dynamic events that occur between an initial dose and the effect of
concern. Events are indicated generically here; but, for a given pathway, many specific kinetic
and dynamic events may occur. (Fig. 2 in Julien et al. (2009))
2.2 Modes of Action and Adverse Outcome Pathways
There has been a longstanding view that MoA analysis improves dose response evaluation and
extrapolation to environmental and human exposures of interest Boobis et al. (2008; Boobis et
al. (2006; U.S (2005). To ensure rigor and transparency in the evaluation of MoA, a framework
was developed in conjunction with the International Programme for Chemical Safety (IPCS)
Boobis et al. (2008; Boobis et al. (2006) and by the U.S (2005). This MoA framework provides a
weight of evidence approach based on considerations for causality as originally articulated by
29 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Sir Austin Bradford Hill for epidemiologic studies Hill (1965). These include considerations of
dose response and temporal concordance; consistency, specificity, biological plausibility; and
coherence. The Hill considerations provide a framework for consistency and transparency in
evaluation of data and are used internationally including when evaluating endocrine disrupting
chemicals WHO (2002). While the MoA framework was originally developed for human health
effects, it has also been applied to ecological effects. In this context, the term adverse outcome
pathway (AOP) has been used, but it is conceptually similar to the MoA Ankley et al. (2010).
Another conceptually similar framework, the Key Event Dose-Response Framework (KEDRF)
Julien et al. (2009) has the advantage of including a more explicit consideration of
pharmacokinetics and dose (Fig. 2.1). For the purposes of this document, MoA, AOP, and
KEDRF are used interchangeably.
Interaction of the active form of the xenobiotic with its tissue and cell-specific target site
constitutes the initial key event, or molecular initiating event, in the MoA. This leads to a
sequence of changes that progress through time and levels of biological organization, leading
finally to an endpoint of concern. Key events are defined as measurable and quantifiable rate
limiting biological steps that lead to the development of an adverse consequence at the organ,
organism, and/or population level. The set of key events is considered necessary for the
development of the adverse effect, but none are sufficient by themselves to cause the adverse
effect and, generally, are not considered, individually, as adverse effects.
Resilience, or adaptation, is an important consideration in evaluating a MoA or AOP. When the
magnitude of the perturbation associated with the initial key event is small, cells and tissues
exhibit resilience, consisting of adaptive responses that maintain homeostasis Andersen et al.
(2005). Propagation of the initial effect along the sequence of key events to the final, apical
event occurs when the adaptive capacity is overwhelmed by higher doses (Fig. 2.2). The tight
regulation of steroid hormone levels is a good example of this kind of homeostatic control.
30 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Exposure-
Tissue dose
J
Biological interaction
Perturbation
mortality
Figure 2.2: Perturbation by a xenobiotic of normal biological function may be overcome by
adaptive responses (homeostasis), while higher doses of the xenobiotic overwhelm the
adaptive capability, driving the system away from its normal biological functions to frank
toxicity. (Fig. 5 in Andersen etal. (2005)).
2.3 Determinants of Nonmonotonic Dose-Response
Conolly and Lutz (2004) noted that "The first interaction of a toxic agent with its primary
biological target molecule follows the law of mass action, which results in a monotonic dose
response." Thus, nonmonotonic responses must arise from more complex relationships
between the dose of the toxicant at its target site and the effect of interest. In studies using
multiple hormone sensitive cell lines, NMDR can be produced from integration of two or more
monotonic responses. One may see NMDR related to receptor down-regulation where the
receptor number is inversely proportional to the amount or concentration of the ligand. NMDR
may be identified when very high concentrations of the compound are toxic to the cells which
would result in cell degeneration or killing and a damping of the response at high dose. NMDR
may be identified due to differences in receptor affinity where multiple receptors bind a ligand
resulting in complex dose response effects WHO (2012).
Conolly and Lutz (2004) developed computational models to illustrate several mechanisms that
can generate NMDR:
1. The level of cyclic AMP as a function of the activation of adenosine A1 and A2
receptors by phenylisopropyladenosine.
31 | Page
-------
2.
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Androgen-mediated gene expression in which combined exposure to native
androgen and a synthetic analog interact competitively at the androgen receptor
to form a series of homo- and heterodimers with differing abilities for promotion
of gene expression.
DNA damage leading to induction of repair wherein the induced repair capacity
also repairs DNA damage from background processes.
The rate of mutation from DNA damage when the damage activates checkpoints
in the cell cycle, with the longer-duration checkpoints providing additional time
for DNA repair before the replicative synthesis of DNA that fixes the mutation.
3.
4.
Table 2.1: Endpoints with nonmonotonic dose-response and causative influences that are
dominant over different dose ranges*.
Endpoint with
nonmonotonic
dose-response
Toxicant or ligand
Dominant influence 1**
Dominant influence 2**
cAMP
Phenylisopropyladenosine
Adenosine A1 receptor,
down regulation of cAMP
(low dose)
Adenosine A2 receptor, up
regulation of cAMP (high
dose)
Androgen-
mediated
gene expression
hydroxyflutamide
(+ 10-7 M
dihydrotestosterone)
homodimers that promote
gene expression
(low and high dose)
heterodimers that inhibit
gene expression
(mid dose)
DNA damage
Xenobiotic that directly
damages DNA
DNA adducts increase with
increasing dose of
xenobiotic
(low and high dose)
induction of DNA repair
reduces adduct level
(mid dose)
Mutation
Xenobiotic that directly
replicative DNA synthesis
cell cycle checkpoint delays
damages DNA
(but no induction of repair)
converts DNA adducts into
mutations
(low and high dose)
replicative DNA synthesis
(mid dose)
*Adapted from data presented in Conolly and Lutz (2004).
**ln this context, "low dose" refers to the low dose end of the nonmonotonic dose-response
curve, and "high dose" to the high dose end. "Mid dose" falls in between low and high dose.
It is possible, in these four examples (Table 2.1), to distinguish two classes of mechanisms that
give rise to NMDR. Two of the examples involve adaptive responses of the exposed tissue -
32 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
induction of DNA repair and activation of ceii cycle checkpoints. The other two examples,
cAMP and androgen-mediated gene expression, do not involve adaptation but, rather, simply
reflect constitutive biology. Since induction of an enzymatic repair process requires some
minimum amount of time, the appearance in this case of the NMDR has a temporal aspect in
that nonmonotonicity will not be seen if the interval between exposure and measurement of
the relevant endpoint is too brief. In general however, for risk assessment, we are concerned
with both short- and long-term exposures, so the dependence of some nonmonotonic
responses on this kind of adaptation does not make them in any sense irrelevant.
Dose 01 xenobtotic
Dose of >enot>ict»c
Dose ol xenobiotic
Dcse of xenobioUc
Figure 2.3: (A) Modeling of dose-response relationships for DNA-adduct levels as a function
of dose of an exogenous xenobiotic. Numbers indicate increasing efficacy of saturable
induction of DNA repair, as illustrated in B. (C) Decrease of background DNA adducts due to the
induced repair. (D) Total adducts (background plus exogenous), obtained by superimposition of
A and C.
The four examples, though diverse, are all characterized by the presence of more than one
influence on the shape of the dose-response curve, with each influence dominant over a
different range of doses (Fig. 2.3). Conolly and Lutz (2004) used "parameter sweeps" to
characterize how these influences combine to generate nonmonotonicity. In these parameter
33 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
sweeps value of a key parameter was varied to produce a corresponding set of dose-response
curves. For the case of nonmonotonicity due to induction of DNA repair, a sweep on the
parameter determining the efficacy or degree of induction of DNA repair was conducted (Fig.
2.3). When the efficacy of induction was minimal, the dose-response curve was monotonic,
while maximal efficacy generated a nonmonotonic curve (Fig. 2.3, panel D). Interestingly,
intermediate efficacy leads to a dose-response curve where, at low dose, the increase in the
adduct burden due to the xenobiotic is closely balanced by a decrease in adducts due to the
induction of repair capacity, resulting in a threshold-like curve. It is important to note here that
the threshold-like behavior is not due to a single action of the xenobiotic but rather to the
addition of two opposing influences that happen to balance each other over a range of doses.
Similar results were obtained by Conolly and Lutz (2004) for the other three cases that they
examined; sweeping on a key parameter led from monotonic dose-response, through an
intermediate, threshold-like regimen to a clearly nonmonotonic response. These results
suggest that the conditions under which nonmonotonicity arises may be only subtly different
from those generating monotonic responses, possibly involving no more than a quantitative
difference in a single component of the system under study.
A general principle can be deduced from the forgoing examples. NMDR can arise when the
biological system that is activated in response to chemical exposure consists of at least two
activities that can act in opposition to each other. If we consider the nonlinear nature of
signaling pathway biology e.g., Bhalla et al. (2002; Hoffmann et al. (2002), the ensemble of
inducible cellular stress response pathways Simmons et al. (2009) and, in the case of the
endocrine system, the tight homeostatic regulation of hormone levels, it seems quite possible
that under some conditions and for some types of endpoints, NMDR relationships would be
expected. The key concern is not so much about the existence of NMDR but, rather, whether
or not nonmonotonicity occurs over dose-ranges that are relevant to humans and to nonhuman
target species and for endpoints of regulatory concern.
2.4 tothalamic-Pituitary-Gonadal and Hypothalamic-Pituitary-Thyroid Axes
The preceding discussion of determinants of the shapes of dose-response curves, whether
monotonic or nonmonotonic, is relevant to EDCs but up to this point has been generic rather
than focused specifically on endocrine endpoints. This section provides an overview of the
biology of the hypothalamic-pituitary-gonadal (HPG) and hypothalamic-pituitary-thyroid (HPT)
axes with the goal of providing necessary background for the evaluations in Sections 2, 3, and 4
of in vitro and in vivo data on NMDR.
The HPG and HPT axes function as integrated systems to produce and tightly regulate blood and
tissue levels of estrogen (E), testosterone (T), and the thyroid hormones (TH) (Figs. 2.4 and 2.5).
34 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
E and T bind to their respective high affinity cytoplasmic receptors, usually referred to as
estrogen receptor (ER) and androgen receptor (AR). TH, specifically T3, binds to nuclear thyroid
receptors (TR).
Estradiol 173 (E2) is a potent natural steroid produced primarily by gonadal tissues in most
vertebrates. The "normal" physiological level of E2 varies greatly among different species,
genders, life stages and reproductive status. Levels required for normal reproductive function
at one stage of life can produce adverse effects at another stage of life. In adulthood, both
higher and lower than "normal" E2 levels also can have adverse effects. The estrogens estrone
and estriol also are normally found in vivo and are only slightly less potent than E2. These
hormones affect many, if not all, tissues in the body in a tissue-specific manner via one of two
nuclear receptors, ERa or ER|3, or less well characterized cell membrane receptors. Tissue-
specific responses arise from differences in receptor levels, levels of coactivators and
corepressors, E2 metabolism, receptor stability, different target gene estrogen response
elements, gene silencing, and other factors.
The two highly related estrogen receptors, ERa and ERp have different tissue distributions and
activate different sets of genes. Although ERa and ERp bind many ligands with similar affinity, a
number of chemical structures have been shown to have selectivity towards one of the pair.
Ligand binding to the ER or AR receptor induces a conformational change resulting in
translocation to the nucleus for receptors located in the cytoplasm and receptor dimerization.
Dimerization is required to form a complex capable of binding to specific DNA sequences,
termed response elements, in the promoters of their target genes. The ligand-bound receptor
homodimer has a 3-dimensional shape that varies from ligand to ligand, resulting in unique
coregulator recruitment profiles and forming "interactomes" that moderate the endocrine
activity of the transcriptional complex, thus potentially altering patterns of target gene
expression O'Malley et al. (2012; Huang et al. (2010). The mRNA produced is translated into
protein that then directs physiologic responses to maintain homeostasis, regulate growth, or
control other specific processes. O'Malley et al. (2012) reported that interactomes associated
with estrogen receptor binding to its natural ligand consist commonly of approximately 10
protein partners that combine into one functional unit. Human cells contain about 11,500
unique gene products that code for proteins that directly or indirectly regulate nuclear receptor
function as components of interactomes. Much less is known about the signaling mechanisms
following activation of the membrane-associated estrogen receptors, but these pathways do
not involve direct transcriptional regulation. Tissue-specific responses arise from differences in
levels of receptors, coactivators and corepressors, receptor stability, gene silencing and other
factors. In addition, EDC-specific variations in the makeup of interactomes can induce gene
expression patterns that differ from those induced by E2 itself.
35 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
The androgen signaling pathway shares many molecular and cellular traits with the estrogen
signaling pathway. There is only one AR in mammals, and as the gene resides on the X
chromosome, males will have only one copy Gao et al. (2005). In some species of fish there are
two AR (ARa and ArP) due to whole genome duplication that occurred at a point in
evolutionary history Ogino et al. (2009). There are two physiologically active androgens in
mammals; testosterone is the major regulatory steroid in many androgen-dependent tissues,
while other tissues rely upon the conversion of testosterone to dihydrotestosterone (DHT) by
5a reductase Gao et al. (2005). There also are two active androgens in fish, testosterone and
11-ketotestosterone; however, the physiological relationship between the two in terms of
function is uncertain. In the androgen hormone system, as with the estrogen hormone system,
tissue-specific coregulators elicit tissue-specific responses.
Thyroid hormone is produced by the thyroid gland (Fig. 2.5) and is essential for normal
development and growth. While TH production is tightly controls by the HPT axis, there are a
large number of targets for thyroid disrupting chemicals that are outside this axis and also
control circulating and tissue levels of THs (see also section 4.2.3.1). Importantly, there is a lack
of detailed information on the temporal and dose-dependence of the multi-tissue feedback-
control systems that regulate TH homeostasis.
The developing brain is one of the most vulnerable organs to thyroid hormone (TH) excess or
insufficiency. THs regulate neuron proliferation and migration, synaptogenesis, synaptic
plasticity and myelination in the developing brain (Williams, 2008). A key feature of TH action
in brain is the temporal sequence of events it supports, a feature that increases the complexity
of determining the impact of xenobiotic-induced alterations in thyroid function Howdeshell
(2002).
In the circulation, TH is tightly bound to serum transport proteins for delivery to the many TH-
dependent organs. The neuroendocrine control of thyroidal TH synthesis and secretion is very
sensitive to negative feedback exerted by circulating TH and involves various molecular
mechanisms such as specific expression patterns of TR genes, TH transporters and deiodinases
Chiamolera and Wondisford (2009; Fekete and Lechan (2007; Fliers et al. (2006). T4 is a
prohormone that is converted into the active form, T3, by type 2 5'-deiodinase, but may also
have non genomic effects via membrane-bound receptors specific for T4. Nongenomic effects
of TH have been reported for blood vessels, heart, and brain Davis et al. (2008) and may
influence cell proliferation and cancer.
Endocrine activity in the context of this document refers to biological effects resulting from
these hormones circulating through the blood to target tissues where they bind to their
receptors to implement their respective activities (Fig. 2.6). An EDC may bind to ER or AR or TR
36 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
as a full or partial agonist, mimicking wholly or in part the ability of the endogenous hormone
to activate the receptor. Alternatively, an EDC may be an antagonist, binding to the receptor
but not activating it and thereby blocking the activity of the endogenous hormone. Differences
in chemical structure between EDCs and their corresponding endogenous hormones can lead to
differences in the conformational change of the receptor that occurs upon ligand binding.
These conformational differences can lead to differences in the signaling downstream from the
receptor, resulting in different patterns of gene expression (Fig. 2.6).
Patterns of gene expression can also be changed when EDC alter normal levels of endogenous
hormones. This MoA involves the activation of nuclear receptors (e.g., CAR, PXR) by the parent
EDC or metabolite. Activation of these nuclear receptors results in an induction of Phase I, II,
and III hepatic proteins, increasing hormone clearance and thereby lowering circulating levels
Crofton and Zoeller (2005; Hill et al. (1998; Capen (1997). For example, methyl-tertiary butyl
ether and wholly vaporized gasoline induce metabolism in the liver and other tissues of
enzymes that catabolize endogenous estrogen Moser et al. (1998; Moser et al. (1997; Moser et
al. (1996a; Moser et al. (1996b; Standeven and Goldsworthy (1994; Standeven et al. (1994b;
Standeven et al. (1994a). Phase I CYP inducers include pesticides and pharmaceuticals Lake
(2009; Martignoni et al. (2006) Phase II enzyme induction includes the enzyme family that
metabolize thyroid hormone, UDP-glucuronosyltransferase enzymes, cytosolic
sulfotransferases, and GSH S-transferase enzymes, as well as some Phase III cellular
transporters Omiecinski et al. (2011; Martignoni et al. (2006).
Steroid hormone receptor activity can also be modified through allosteric interactions Kumar
and Mcewan (2012). In this case receptor activity is modified not by classical "lock and key"
ligand binding at the active site but by binding elsewhere on the receptor molecule that alters
its three dimensional structure and, thereby, it's signaling behavior. For example, Joseph et al.
(2009) describe "second site" ligands identified in pharmacological studies that antagonize the
activity of AR. While allosteric interactions are a potential class of key events for environmental
endocrine active xenobiotics, there do not currently appear to be documented cases of this
mode of action.
37 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Stress,
Nutrition,
Exercise,
Seasonal cues
Tostostorono
A Progesterone receptors
• Estrogen receptors
¦ Androgen receptors
Y Inhibin receptors
Lepliri receptors
Progesterone
Figure 2.4: Hypothalamic-pituitary-gonadal (HPG) axis. Blood and tissue levels of sex steroids
are tightly regulated in this multi-tissue, feedback-controlled system. The hypothalamus
produces gonadotropin-releasing hormone (GnRH); the pituitary produces luteinizing hormone
(LH) and follicle stimulating hormone (FSH), and the gonads (testis, ovary) produce testosterone
and estrogen. (Fig. 2.4 from www.Health-7.com)
38 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Hypothalamus
6 Excretion
In Bile
5 Metabolism
in Liver
UDPGT, EROD,
PROD, Deiodinases
1 TH Regulation
s Negative feedback
via TRU, TSH, T3, T4
2 TH Synthesis
TPO, I, NIS, TSHr
4 Peripheral
Deiodinases
Serum Binding Proteins
TBG, TTR, Albumin
TH Action
7 Local Transport Proteins mcts.oatp
8 Local Deiodinases 02,03
9 Thyroid Receptors trp, tru
» Structural/Functional Impact
Adapted from Boas el al. Eur J Endo, 2006
* q Gene Transcription, Neurogenesis, Migration, Synaptic impairments, learning
Synaptogenesis, Myelination ' * deficits, hearing loss, visual defects
Figure 2.5: Regulation of thyroid hormone production, release, transport, metabolism and
action. Control of thyroid hormone action in tissues is regulated by a combination of the HPT
axis and other extrathyroid tissue processes. The HPT feedback process regulates synthesis and
release via the hypothalamic production of thyrotropin-releasing hormone (TRH); the pituitary
production of Thyroid Stimulating Hormone (TSH) and the Thyroid gland production and release
of thyroid hormones (T4 and T3). In most mammals the majority of T3 is produced at the target
tissue from deiodination of T4 and impacts T3 and T4 levels in both tissues and blood. The liver
controls catabolism of THs and can thus indirectly regulate TH levels in the blood. Finally, local
tissue regulation of transport proteins and receptors can regulate tissue level transcription.
(Fig.2.5 from Gilbert et al. (2012))
39 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Cytoplasm
Figure 2.6: Schematic diagram of the activation of a cytoplasmic steroid hormone receptor.
After hormone binding (A) the HSP complex dissociates from the receptor (B), the hormone
receptor complex translocates to the nucleus (C), dimerizes (D) and binds to a hormone
responsive element (HRE) in the promoter region of a specific gene (E), After binding to the
HRE different coregulators of transcription are recruited (F), which are responsible for
transcriptional activation (Fig. 2.6 from Riedmaier et al. (2009)).
2.5 Statistical Issues and Experimental Design
When determining the appropriate analytical approach, the underlying assumptions of the
statistical tests, which often include normality, monotonicity, and balanced groups, need to be
considered. In cases where these assumptions are violated, tests that have been developed to
avoid these common assumptions should be used. New methods have been developed in
many areas to overcome these challenges in order to improve estimation of nonmonotonic
dose responses Bretz and Hothorn (2003; Takizawa et al. (2000; Chen (1999). in particular,
Gaylor et at. (2004) described methods for the analysis of J-shaped dose-response curves for
both continuous and quantal endpoints. J-shaped curves are NMDR in which the curve initially
has a negative slope downward from a nonzero control level, which then inflects upward to a
positive slope at higher doses. Gaylor et al. (2004) suggest experimental designs to identify
40 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
nonmonotonicity and include information regarding numbers of doses and group sizes. In
general, they show that much larger sample sizes are needed to detect J-shaped effects at low
doses, particularly for quantal data by comparison to continuous data. This work highlights that
principles of statistical design of experiments need to be utilized when determining
experimental doses and sample sizes to maximize the power to detect J-shaped curves. Since
Gaylor et al. (2004) was restricted to J-shaped curves, its relevance to NMDR generally may be
limited.
Multiple testing problems can arise and potentially confound identification of NMDR. Several
methods have been developed to correct for multiple comparisons in order to control the
familywise error rate. Familywise error rate is the term used in the multiple comparisons
literature to mean that the overall alpha level is kept to 0.05 (or another pre-specified level) for
all hypothesis testing performed. These methods range from the most conservative Bonferroni
approach where the type I error rate is set to be alpha/n, where n is the number of tests
performed, to more advanced computerized resampling methods, including bootstrapping and
Monte Carlo simulations. As pointed out by Bretz and Hothorn (2003), a trend test is not useful
if it does not "control the probability of incorrectly declaring a dose to be effective, when in fact
it is not effective."
Limitations in experimental design, in particular the number of animals in each dose group, can
lead to inaccurate interpretation of the data, including indicating nonmonotonicity when the
true response is monotonic. Though small sample sizes may be unavoidable for logistical
reasons, they may contribute to considerable random variation. Lutz et al. (2005)
demonstrated that NMDRs can arise due to random variation in a quantal endpoint when the
true underlying dose-response has a positive, linear slope.
Use of larger group sizes or replication of entire studies may not be practical for large-scale
studies, such as carcinogen bioassays or multigeneration reproductive toxicology studies.
However, when results are potentially significant for our understanding of toxicological
mechanisms and their attendant regulatory implications, experiments should be evaluated for
statistical power and replicated when possible. Jasny et al. (2011). in a special issue of Science
on Data Replication and Reproducibility, noted that "Replication - the confirmation of results
and conclusions from one study obtained independently in another - is considered the scientific
gold standard." For large scale studies, where study replication or modification of study design
is not practical, it is desirable to use ancillary studies of pharmacokinetics and the MoA to gain
insights into expected dose-response behaviors.
The number of dose groups in current guideline studies is considered adequate to inform most
risk-based decisions, but they may not be sufficient to fully describe the shape of an NMDR. It
41 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
is generally true that the shape of any dose response curve can be more clearly defined by
using more dose groups and by adjusting group sizes using statistical power calculations OECD
(2011; Blystone et al. (2010; Hotchkiss et al. (2008).
3. NMDRs from in vitro Studies
The occurrence of NMDR in vitro has been well documented in the scientific literature with
examples presented in recent review articles (e.g., Vandenberg et al. (2012); DCED (2013)). The
current state of the science document does not include an additional exhaustive review of the
in vitro literature, but rather discusses general concepts for nonmonotonicity in vitro,
incorporates observations from recent review articles, and cites examples from primary
literature, where applicable.
NMDRs have been reported for a variety of in vitro assays targeting endocrine receptors, and
incidence of their observation is likely to increase with the recent impetus to develop and
implement high-throughput toxicological screening technologies. To understand the
implications of in vitro NMDR findings, it is useful to begin with a mechanistic understanding of
common assays and potential causes of NMDRs. For the estrogen and androgen receptors, the
great majority of the reported biological activity of xenobiotic chemicals (especially E and A
signaling) is through binding to the ligand-binding site of the receptor, resulting in a modulation
of the receptor activity manifested as agonistic, antagonistic or selective receptor modulation.
The original in vitro assays targeting endocrine receptors measured direct binding of
radiolabeled E and A ligands to preparations of crude, partially purified or highly purified
receptors. Alternatively, binding of non-labeled ligands could be detected indirectly through
competitive displacement of radiolabeled ligand. With such binding assays, NMDRs have not
been reported except as artifacts. Binding of the ligands to nuclear receptors follows the law of
mass action in which a ligand binds reversibly to the receptor and the fractional occupancy of
the receptor is a function of the ligand concentration and the Kd, the equilibrium dissociation
constant for the ligand/receptor pair. Thus, as measured in a radioligand binding assay, the
response is a monotonic, competitive binding curve. When behavior of the assay deviates from
this with a nonmonotonic response, it is characterized by decreased binding at high
concentrations. Such effects are explained either by insolubility of the compound at higher
concentrations or by assay interference artifacts, such as aggregation or denaturation, common
at higher concentrations.
Binding of the receptor ligand to its binding site is the initial step in receptor activation. This
alone, however, does not provide information about whether the compound has functional
agonist or antagonist behavior. Functional cellular reporter gene assays, called transactivation
assays, have traditionally been used to determine this behavior. These assays measure ligand-
42 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
bound receptor binding to a hormone-response element on DNA, recruitment of coregulator
proteins, and changes in the expression of measurable reporter gene products. Major
variations in transactivational assay approaches include: full-length receptor vs. ligand-binding
domain; stable vs. transient transfection of the reporter gene constructs and receptor;
endogenous vs. transfected coregulator proteins; and the specific reporter gene product.
Finally, these assays can be run in agonist mode to measure the ability of the test compound to
activate or increase receptor activity, or in antagonist mode performed in the presence of a
single concentration of a receptor agonist, measuring the ability of the test compound to inhibit
or reduce agonist-induced activity.
NMDRs can occur in transactivational assays due to cytotoxicity. Dose response effects caused
by a chemical are often impacted by cell death at high levels. Since many compounds have
multiple molecular targets that are affected at different concentrations, compounds that bind
and activate steroid receptors at one concentration can interact with other cellular proteins at
higher concentrations, leading to cytotoxicity. In a reporter gene assay, this is manifested as
increasing reporter gene activity with increasing compound concentration up to a point at
which there is significant cytotoxicity and concurrent loss of reporter gene activity. An example
is the phytoestrogen genistein. Genistein binds and activates the estrogen receptor with a Kd of
20 nM but also affects a number of other important targets such as protein kinases and
topoisomerase at higher concentrations, leading to cytotoxicity and cell death. Measuring a
transactivation assay for genistein with the ER will result in a NMDR with loss of activity at
higher concentrations.
Use of transactivation assays serve as surrogate assays for endogenous target gene expression
and corresponding encoded protein levels, the primary biological activity regulated by nuclear
receptors. Measurement of gene or protein expression in other types of in vitro systems suffers
the same proclivity to NMDR effects as do transactivation assays. An illustrative example of this
is provided by Schmieder et al. (2004) who examined induction of the egg yolk protein
precursor vitellogenin (VTG) in rainbow trout liver slices exposed to several model ER agonists.
They observed a significantly depressed induction of VTG at high versus low doses with a
number of putative estrogens (o,p'-DDT, monohydroxy-methoxychlor, p-nonylphenol), a
response that appeared to be attributable to overt toxicity of the chemicals to the liver cells. In
that paper Schmieder et al. (2004) noted the importance of evaluating cell viability in order to
accurately interpret responses to EDCs in this type of in vitro system.
In a recent review, Vandenberg et al. (2012) identified 80 examples from the literature, which
provide evidence of nonmonotonicity in vitro. The NMDRs in these studies were from a wide
range of chemical classes including natural and xenobiotic hormones, phytoestrogens, plastics,
surfactants, metals, PAHs, PCBs, PBDEs, pesticides, and dioxin. The nonmonotonic effects were
43 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
predominantly seen in lower-order biological endpoints closer to the molecular initiating event.
Included in these endpoints were measures of protein or hormone levels, gene expression, and
cell number. Subsequent analysis of these 80 NMDRs by The Danish Centre on Endocrine
Disruptors DCED (2013) revealed that cytotoxicity was the most commonly observed
mechanism for causing NMDRs in vitro. Nearly half (45%) of the in vitro studies had NMDRs
that were likely caused by cytotoxicity. Further, the authors established that an additional 22%
were "false NMDRs" caused by study design (i.e., mixtures of chemicals, inappropriate
statistical analyses for establishing nonmonotonicity). The remaining third of the studies either
showed evidence of NMDRs (16%) or provide evidence that may or may not be due to NMDRs
(17%). These study evaluations are provided in Table 3.1.
44 | Page
-------
Draft Deliberative - Do NOT Cite, Quote, or Distribute
State of the Science: Nonmonotonic Dose Responses V7
Table 3.1: in vitro NMDRs as evaluated by the Danish Centre on Endocrine Disruptors DCED (2013).
C1ienBic.il' by
chrmi>nt
Nonmonotonic effect
C ell type
Mrfs,
Evaluation
Natural karmonet
17R-Estradiol
Cell number
MCF? breast cancer cells
155. -16
Cvtctox
Dopamine uptake
Fetal hypothalamic cells {primary)
"7 i
Maybe
pERK levels, prolactin release
GH3®6/F10 pituitary cells
41. "IS. "IS
NMDR
R-HexDsamiffidase release
HMC-1 mast cells
::c
CVtctox?
Cell number
Vascular smooth muscle cells
~2l
Fahe
Production ofL-PGDS, a steep-promo'iiie substance
U251 glioma cells
C'vntox
5a-Dih.ydrotestosteraiie
Cell number
LNGaP-FGC prostate cancer cells
490
Cvtctax
Cell number, kinase activity
Vascular smooth muscle cells
721
Falie
Sa-Androsteiieiioiie
Cell number
INCiP-FG€ prostate ciacer cells
499
Cytotox
Coiticosterone
Milochrondrial oxidation, calcium flux
Cortical, Demons (primary)
723
Cvtctox
Itraik.
Markers of tpoptosis (in absence of glucose)
Pancreatic R-cells (primary)
"724
Cytotox
Pi.'ze'.te: ."Tie
Cell number
LNCaP-FGC prottate cancer cells
49 o
Cytotox
Piolactici
Testosterone release
Adult rat testicular cells (primary)
TV^
Cytotox
liC'G
Testosterone release
Adult rat testicular cells (primary)
71'i
CjKto
t3
Rate of protein pkesphotytotiaii
Cerebral coites cells (primary, synaptosomes)
726
MMDS.
LPL naRNA expression
White ! i at primary)
"T 1 -¦
False
GH
1GF-I expression
Kepatccytes l.ptmiaiy cntats &om alva sea bream i
~2S
false
P1! ,ti-ni"nf i'ic! .'-I,'i 'jat's
BEt>
Cell number
MCF" h.ei?t caacet eels
716
Cvtctax
Pre!aeftr. release
GH3 3d F". 0 pituitarycells
41
NMDP
Ethinyl eitradiol
CXCLI2 secreuoc
MCF"" tieait caace; cells, T47D breast cancer cells
729
Maybe
R1SS". i synthetic acdiceea)
Cell number
LMaP-FCrC celb
499
C YtCtCCv
Treabcloce
Induction of uuctemxfei
PJL-Wl fi-.hhvet eels
730
NMDR
Plastics
BPA
Cell number
MCF7 breast cancer ceils
135. 716
Crtecox
Dopamine effiux
PC 12 rat tumor cells
40
XNIDR
pERKlevels, innacellr.bi Cj cliiiisei, prolactin release GH3.'Bfe'F10 pituitary cells
4!. 7ig
XNIDR
Cell number
LNCif prostate cancer cells
"31
Fai:e
DEHP
Number of colonies
Bcfiiinrta'n cvli nodB siifiiitS bacteria
' J _
Maybe
Di-is-octyl pbtfaalate
Number of coknies
E. coM and B, suMlis bacteria
•— i
> J —
Maybe
45 | Page
-------
Draft Deliberative - Do NOT Cite, Quote, or Distribute
State of the Science: Nonmonotonic Dose Responses V7
Chemicals by
chemical class
Nonmonotonic effect
Cell type
Ms,
Evaluation
Detm'gmtU, sx'/m-fniffs
Octyipkeaol
Cell number
MCF7 breast cancer cells
716
C vts/tox
Dopamine uptake
Fetal hypoiiaiainic cells (primary)
717
M.ivbe
nERK levels
GH3/B6/F1Q pituitary cells
"IS
nmdr
HCG-stimulated testosterone levels
Leydig cells (primary)
753
XMDR
fcopylpfajol
pERK levels
GB3.«B®FU pituitary cells
~ES
NMDR
Nonylpheaal
pEBK levels, prolactin release
GH3.®&T10 pitaitaiy cells
41. ""IS
N51DR
R-Hesosaminidage release
HMC-1 mast cells
720
Cvtctox
Ceil number
MCF7 bresst cancer cells
135
Cvte:ox
PAH
Phenamfarene
.-ill-trsni letiici; acid activitv
P19 embryonic carcinoma cells
754. 735
Mavte
Renz(a)acridiiie
AII-tijib leticcsc acid activity
Pitembryonic carcinoma cells
754
Maybe
Naphthalene
jiC G-vtjuul.tted te 4c -..teione
Pseci of k liifi'vli testes
"36
C'vtc'ox
B-iiapliiioflavoae
aC G- s fcnir.hted teoto -Aetone
Pwres of eoldfrti testes
736
C\tv"!OX
Retene
IiC G-?t;a:r.l,ited tesfoi-leione
?^si:cl solifuli'estei
"36
Cvtefox
flamy metah
Lead
E-jnoeer.. testosterone, and Cortisol lewis
Fo;fvitellcgeaic -cllules (isolated fiom cafthlii
737
Cvteto.i
Cadmium
Expsei5ioo cf jMiogeaews semes
Kuaun -fadcmfTrnl enictliel-.j". celk.
73 S
Maybe
o«rrf natural anriaxtdrnm
Geuiitem
Cell n-.mibei
Caco-2BBe colon idenocarcinoma cells
75 a
Cvtctox
CXCLil secretion, cell number
74~D bieast can;ei cells
-? 2 a
Maybe
Cell number, ceE im-aston, MMP-S activity
PC3p:ostjte ;?.r„e« cells
74.fl
Cvtctox
2-+-
pJNK levels, Ca fins
Or.'- B6 F10 pituitary cells
^1?
XMDR
Coumesterol
ftol»etia release, pERK levels.
GtS'R&FiCi pituitary ceils
IP
NMDR
Psidezia
Prolactin release, pERK levels
GH3,'B#F10 pituitary cells
719
XMDR
Cell number
MCF7 breast cancer cells
155
C'ytCtOX
Cell cumber
LoVo coloo cancer cells
741
Mavbe
Resveritol
Exprccion of aagiogaiesis genes
Human umbilical vein endothelial cells
742
NMDR
Trans-resverafcrai
2-+-
pEEK levels C « flux
GH3.'B&'F10 pituitary cells
719
NMDR
Artelastochromene
Cell iivuiifc.et
MCF7 breast cancer cells
743
M.vbe
Caipelastoiiiran
C eil numbei
MCF7 breast cancer cells
743
Maybe
Biochanin A
Induction cf esfcoeer.-5er..;j;-Te ger.es
MCF7 breait cancer cells
744.
Maybe
Licoflavone C
Bielmtioiici esfcosen-jen-abve aeaes
Yeast trio-assay
745
Maybe
Qoercetin
Aiomatjse activity
H295R adrenocortical carcinoma cells
746
Cytotox
Cell ir.imtei
SCC-25 oral squamous cardnoma cells
747
Maybe
46 | Page
-------
Draft Deliberative - Do NOT Cite, Quote, or Distribute
State of the Science: Nonmonotonic Dose Responses V7
Clwiaieals. by
chemical class
N011.iB0110t11.11lc effect
Cell type
Rrfs,
Evaluinon
TCBD
Cell number. Bene expression
MliSYi breast cells
748
Maybe
FC±
PC B-~4
Cell viability, GoRH peptide levels
&T1-7 hypothalamic celb
749
Cvtctox
PCB-1 IS
Cell viability, GnRH peptide levels
GT1-7 hypothalamic cells
749
C'ytotox
Arc-dor 1242 ,K3 uuxrure)
P-Hexosamtrndase release
HMC-1 0M5T celli
720
C'vtccox
PC'P rmxrose
Hpcp^its of cumulus ceis
Occjle-cjinuhn complete? :pnnui/, front pigs)
750
Falw
H« bickl^'
Glvpho spin te-hei 'tucids (Roua4-
¦Upi Cell death. aromatase jcfcvitv. EF.p activity
r>oGl ..
751
Cytotox
Atraziae
Cell Blunter
IEC--5 ifJesEir,sl celb
752
False
Insecticides
Eadosal&n
Cell number
IEC-'S celh
752
False
p -Hexosamimdase release
HMC-1 mast cell-;
720
Cvtetox
ATPase activity of P-glycoprotein
CEO cell exriacn
~53
Maybe
Diaziaoa
Cett cumber
IEC-6ii;fe5tu:al celh
~5 2
False
DieMrin
p -Heios«mmid«se release
HMC-1 aa-;t celli
< 20
CvtctOA
DDT
Cell number
MCF~ bieast :a:ice; cell;
144
Not evaluated-
DDE
P-Hexosaminidase release
HMC-1 aiast cell-:
720
C vtojoj;
Prolactin release
&H_! 3? F10 pituitary cells
41
NMDR
3-Metliytolfoiiyl-DD E
CmtisotndddoitasK release, sterwiegemc genes
E2P?R adieno cortical cmamom* cells
754
Cyt>:tox
Fungicides
Hexacblarobenzene
Trmsciiptioatl activity is the presence of DHT
PC testate taccer cell-;
755
Cvtcnx
Pfocliloraz
Aldosterone, progesterone, and cartkosterane levels;
expression of steroidogenic genes
E2P5R adrenocortical cell-.
756
Cytofox?
Ketocecazole
Aldosterone secretion
H2P5R adrenocortical celh
757
False
Fungicide nn.xntie'.
Aldosterone secretion
H295R adrenocortical ceil-.
757
False
PBDE
PBDE-49
Activation of ryanodine receptor 1
HEK29? ceil (membrane-; i
758
C'ytetox
PBDE-99
Espies-ion cf GAF43
Cerebral cortex cells (primary)
759
Cyterox
47 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
In addition to cytotoxicity, other effects can result in NMDRs in vitro. Competition between
ligands, ligand-induced down-regulation of receptors Alarid (1999), squelching due to limited
availability of cofactors Bocquel et al. (1989), shutting-off of endocrine responses at high levels
Sonnenschein et al. (1989), 1989), and mixed dimerization of ligands or receptors Chang et al.
(2006; Wong et al. (1995) may lead to NMDRs. In general these mechanisms of
nonmonotonicity have biological underpinnings that explain their occurrence.
In addition, NMDRs may be seen as an artifact of a specific in vitro assay and may not be
repeatable in other transactivation systems. Increasing chemical concentrations may interfere
with the reporter gene products by mechanisms such as chemical quenching of fluorescent
signal or inhibition of a reporter gene product enzymatic activity. Sometimes the pharmacology
of the assay system has not been optimized appropriately resulting in NMDR at high
concentrations due to transcriptional squelching from rate-limiting steps in the signaling
pathway. As an example of this, in transiently transfected assays utilizing recombinant
receptors, a large excess of receptors relative to the more limiting endogenous coregulators
could result in the generation of "unproductive transcription complexes" that lack sufficient
coregulators and/or accessory proteins on the response elements to permit transcription. This
occurs only at high concentrations of ligand and associated high receptor occupancy, whereas
at low concentrations the relatively few active ligand-bound receptors have sufficient
availability of coregulators and accessory proteins to signal appropriately.
An NMDR has been reported in vitro for the AR antagonist hydroxyflutamide (OHF, active
metabolite of flutamide) via mixed agonist and antagonist activity. OHF binds AR and inhibits 4-
dihydrotestosterone (DHT)-induced gene expression with montonic effects in the nanomolar to
micromolar concentration range Wilson et al. (2002). At higher concentrations, however, OHF
shows a loss of inhibition of DHT-induced activity beginning at around 10 micromolar, an
activity attributable to agonist activity at this concentration as confirmed by activity in the
absence of DHT Wong et al. (1995; Kelce et al. (1994). This mixed antagonist:agonist activity
appears as a NMDR Wilson et al. (2002). The biological significance of these high
concentration-effects remains unclear. The in vivo dose response data for flutamide provides
no evidence that the androgenic activity seen in vitro is expressed in vivo Miyata et al. (2002;
Mcintyre et al. (2001). In addition, several other AR antagonists also are androgenic in vitro,
and for some of these chemicals, the concentration that produces AR antagonism of androgen
in the test system media also induces AR-dependent gene expression when androgen is not
present in the media. Most of these, however, appear to act as antiandrogens in vivo. A
complete mechanistic understanding of these effects is lacking. Of note is the report that
hydroxyflutamide can activate the MAP kinase pathway which would lead to ligand-
independent activation of AR Lee et al. (2002). This would be consistent with
polypharmacology of the compound occurring at concentrations higher than required for
48 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
receptor binding. An alternative proposal is an effect occurring only when receptor dimer
subunits bind to hydroxyflutamide or vinclozolin and require high chemical concentrations
Wong et al. (1995).
Estrogen may bind to additional receptors beyond the steroid nuclear receptors. The 7-
transmembrane receptor GPR30 was found to bind estrogen with a Kd of approximately 5 nM,
10-fold higher concentration than the Kd for the ER itself Revankar et al. (2005; Thomas et al.
(2005). This has also been noted for Bisphenol A (BPA), however, several other laboratories
were unable to demonstrate estradiol binding to GPR30 or estradiol-activated signal
transduction in GPR30-expressing cells. Binding to the GPR30 receptor has been reported to
induce production of cAMP and intracellular calcium mobilization. Sensitive assay technologies
exist for these endpoints and have been used to characterize GPR30 cellular activity. Some of
these studies have resulted in apparent NMDR at relatively low concentrations. These studies
are difficult to evaluate, however. The responses are often of very low magnitude, and,
although statistically significant, have little biological validation. The GPR30 receptor was
originally an orphan GPCR receptor discovered through DNA sequence analysis and only later
found to bind estrogen. Further, it remains to be determined if estrogen is a physiological
ligand for GPR30 regulating important in vivo functions. This receptor does not mediate
estrogenic responses in reproductive organs in mice Otto et al. (2009). It is critical to consider
the validity of the linkage of the in vitro activity to in vivo adverse effects for the evaluation of
potential endocrine disrupting activity. Further, data gaps exist regarding the potential
organizational and/or activational effects of GPR30 receptor binding and the significance of
NMDRs.
Other components of the complex nuclear receptor signaling pathways could be manifested as
NMDR in cellular assays. To this point the in vitro discussion has largely focused on binding to
the ligand binding domain of the receptor, which results in conformational change that creates
the coactivator binding site and permitting initiation of a functional transcriptional complex.
However the amino-terminal region of the receptor is also transcriptionally active and not
regulated directly by ligand-binding.
Other signaling pathways such as the MAP kinase pathway or tethering by the aromatic
hydrocarbon receptor or other transcription factors such as AP-1 and STAT5, can modulate ER
activity through this region and have been reported to generate NMDR. Again, although effects
on receptor activity can be demonstrated in vitro, there is no evidence that this translates to an
in vivo effect. As an example, Bulayeva and Watson (2004) show results for a group of reported
xenoestrogens and estrogen on the MAP kinase signaling pathway. Effects on phosphorylation
of ERK were shown at subpicomolar concentrations followed by a decline in activity, and then
an increase in activity for some of the chemicals at higher concentrations. However, no positive
49 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
control for the MAP kinase pathway, such as epidermal growth factor, was used. Hence the
marginal 20-40% maximal increase in signal over basal conditions in this in vitro assay system
cannot be put into the context of effects in vivo for the endocrine or any other physiological
system.
The point of inflection for in vitro nonmonotonic curves is typically seen at high doses above the
dose range known to cause overt effects in vivo or that are measured in environmental samples
(Figure 3.1). Therefore, a key question is how relevant is the in vitro concentration at the
nonmonotonic point of inflection when compared to in vivo tissue levels or environmental
exposures? A comparison of the in vitro concentration-response curve for E2-induced
luciferase activity is presented in Figure 3.1a. Superimposed on the graph are the doses at
which in vivo reproductive effects for apical endpoints have been seen in mammals Biegel et al.
(1998b; Biegel et al. (1998a) and fish Hirai et al. (2006). In this example the reproductive effects
occur at exposures more than three orders of magnitude lower than the in vitro high dose point
of inflection at which the curve turns downward. Similar differences between in vitro points of
inflection and in vivo or environmental exposure concentrations exist for many chemicals,
including EE2, BPA, Trenbolone, and Testosterone (Figures 3.1b-e). In our assessment of the
literature, no examples were identified wherein the point of inflection in vitro was near or
below the concentration producing adverse effects in animals.
In summary, NMDRs are sometimes seen with in vitro assays measuring estrogen and androgen
receptor signaling pathways. However, these are generally resulting from assay artifacts,
cytotoxicity, or improperly controlled experimental conditions, in particular lack of appropriate
positive controls. None of the reported in vitro effects relating to NMDRs appear to have
significant evidence linking them to in vivo adverse outcomes. The majority of the evidence
linking in vitro assays to in vivo effects is readily explained by the well described pharmacology
of the steroid hormone receptors. Even the complex phenomena of selective receptor
modulators follow law of mass action behavior and can be translated from in vitro systems to in
vivo effects as exemplified by the selective estrogen receptor modulator (SERM) raloxifene,
which shows in vitro agonistic activity for bone-derived cell types but antagonist activity for
uterine and mammary cells, and tamoxifen, which shows antagonist activity towards mammary
cells but agonist for uterine Shane and Brown (2002). In vitro systems do provide effective
means for measuring potential in vivo activity for important endocrine targets. While one can
identify specific cases where NMDRs could pose technical challenges for testing protocols, e.g.,
testing for estrogenic activity only at a single very high concentration where assay interference
(cytotoxicity or artifact) resulted in a false negative interpretation, these situations would be
adequately addressed by testing a broad concentration-response that covered a range from
picomolar to hundreds of micromolar.
50 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Figure 3.1: Comparisons of in vitro points of inflection and in vivo or environmental exposure
concentrations using T47D KBIuc cell luciferase activity induction.
E2 T47D KBIuc
10i
o
¦ 1MB
o
D
"D
C
2
O
5-
0-
r-r
Serum E2 in
female rats fed 2.5
ppm E2 reducing
F1 female
fecundity
Welter cone
inducing
100% ova/testis in
m edaka
> > JV
rrmq rm
<5 S.
N«>
E2 ppt concentration
A. Comparison of the in vitro concentration-response curve for estradiol (E2) with
concentrations found in the serum of acyclic female rats and water concentrations inducing
ova-testes in 100% of the male fish. In vitro data from Wilson (personal communication),
rat data from Biegel et al, (1998b); Biegel et al. (1998a) and medaka data from Hirai et ai,
(2006).
51 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
EE2 T47D KBLUC
CM
LU
O
4-
o
+-»
C
y N
K>» K?
Concentration ppt
»
I
^ i TTHTmi^~niwy i img i iwpTH^TTi^
k$ fcS ^ ^ ^
Serum EE2 level in
rats treated with a
very high oral dose of
EE2 - 1m g/kg
Cone in whole lake
causing fish
population crash
B, Comparison of the in vitro concentration-response curve for ethinyl estradiol (EE2) with
concentrations found in the serum of rats treated with a very high dose of EE2 and water
concentrations causing a population crash and near extinction of fathead minnows in a
seven year whole lake study. In vitro data from Wilson (personal communication), rat data
from Twaddle et al. (2003) and fathead minnow data from Kidd et al. (2007).
52 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
BISPHENOLA: NMDRC
T47D KBLUC ERTA ASSAY
25-i[ Cmli77rrr*BP^TTduTlinT^eon^ai^HKUS-ToTjig/rgTv7Trarj
204
15-
10-
5-
0
I I I llliq I I I lllll| I III lll|
Water cone inducing m ale fish vitellogenin
i —i —i i hi
c?>N c?>N <^N
N*
Concentration (ppm)
C, Comparison of the in vitro concentration-response curve (Wilson, personal communication)
for free bisphenol A (BPA) with concentrations found in the serum of neonatal rats Doerge
et al. (2010c; Doerge et al. (2010a) and neonatal and adult monkeys and mice Sieli et al.
(2011; Doerge et al. (2010b) treated orally with a high doses of BPA (compared to human
exposure levels) and water concentrations of BPA causing estrogenic effects in fathead
minnow Ankley et al. (2010).
53 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
T renbolone
NMDRC
Cytotoxicity
K >$> & ^ ^ ^
a v> N ^ ^ ^ <5
Concentration (ppb)
TB cone in tissue associated with infertility in the
fathead minnow (Ankley et al, 2003). BCF about 10x
TB cone in amniotic fluid in female fetus resulting in
reproductive tract malformations (Hotchkiss et al., 2010)
D. Comparison of the in vitro concentration-response curve for trenbolone (TB) with
concentrations found in the amniotic fluid of rats treated with a dose of TB that induces
reproductive tract malformations in female rat offspring Hotchkiss et al. (2010) and tissue
concentrations in adversely affected adult fathead minnows exposed to TB in the water, In
vitro data from Wilson (personal communication), rat data from Hotchkiss et al, (2010) and
fathead minnow data from Ankley et al. (2010).
54 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Testosterone
ppt
serum T in pregnant rats at a dose inducing reproductive
tract malformations in 100% of the female offspring
E. Comparison of the in vitro concentration-response curve for testosterone with
concentrations found in the serum of pregnant rats on gestational day 19, treated with a
dose of testosterone that induces reproductive tract malformations in female rat offspring
Wolf etal. (2002).
4. NMDRs from in vivo Studies
4.1 Aquatic Models
The occurrence of NMDRs has been well documented in the field of ecotoxicology e.g., Kefford
et al. (2008). And, as observed for some types of human health-oriented tests with EDCs, there
are examples of NMDRs generated with endocrine-active chemicals in ecologically-relevant,
non-mammalian species. In general, however, the occurrence of NMDRs in tests contributing
to ecological risk assessments have not been as significant an issue as they are for human
health assessments. There are multiple reasons for this. First, and probably most importantly,
selection of concentrations for ecotoxicology testing typically includes actual or predicted
chemical concentrations observed in the field. Thus, environmentally relevant dosing may be
part of protocol development so that there is little need to extrapolate from high- to low-dose
55 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
effects; the latter are measured directly. Second, at least in some instances, there is less of a
challenge in ecotoxicology relative to extrapolation of dose-response relationships across
species. That is, it often is possible to directly evaluate the species of concern, or closely-
related taxa, which is not the case for human health assessments. Finally, the occurrence of
NMDRs in ecological risk assessments historically has been of lesser concern than for human
health because a comparatively greater degree of uncertainty has generally been acceptable for
the former. Tests that have been conducted with non-mammalian species lend insights as to
the nature, frequency, and mechanistic underpinnings of NMDRs.
4.1.1 Literatu xh and Analysis, Aquatic Species
We undertook a relatively focused literature search and associated analysis from which were
identified illustrative examples of dose-responses relationships (including NMDRs) for different
classes of HPG-active chemicals. These included those that interact with the estrogen and
androgen receptors as well as inhibitors of enzymes involved in steroid synthesis, such as
aromatase. For the present analysis we sought to identify studies encompassing full life-cycles
or, when not available, longer-term experiments during portions of the life-cycle expected to be
sensitive to endocrine-active chemicals. For example, in the present effort short-term lethality
studies were not assessed. In addition, where possible we focused on studies that examined
apical endpoints clearly related to endocrine function (e.g., aspects of early development or
active reproduction). Ideal study would have at least four treatment groups, however, in some
instances studies with fewer doses were included for evaluation in order to include additional
chemicals and pathways in the analysis.
To facilitate an efficient literature survey, we focused on a subset of 28 model chemicals (Table
4.1) that are known to interact with HPG pathways of concern for which, at least some data
exist for non-mammalian species in the peer-reviewed literature. By far the best studied HPG-
active chemicals in fish are ER agonists, xenoestrogens. The two most commonly tested
compounds are 17a-ethynylestradiol (EE2) and BPA, both of which were included in our
analysis. We also were able to identify a longer-term, multi-dose fish study with the SERM,
tamoxifen. Chemicals selected as model ARagonists for the analysis were the synthetic steroids
17|3-trenbolone (TB) and 17a-methyltestosterone (MT). Flutamide and vinclozolin were
evaluated as prototype AR antagonists. Finally, fish studies with fadrozole, trilostane,
prochloraz and ketoconazole are included as illustrative examples of the nature of dose-
response curves with known inhibitors of sex steroid synthesis.
Table 4.1: Model Chemicals for Aquatic Models Assessment.
ESTROGEN SIGNALING PATHWAY ANDROGEN SIGNALING PATHWAY
56 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Estrogens
Ethinyl Estradiol
Estradiol
Diethylstibestrol
Genistein
Zearalenone and Zeranol
Bisphenol A
Methoxychlor
Chlordecone
Octylphenol
Nonylphenol
Selective Estrogen Receptor Modulators
Raloxifene
Tamoxifen
Lasofoxifene
Aromatase Inhibitors
Fenarimol
Letrozole
Fadrozole
Anastrazole
Exemestane
To obtain representative literature for our analysis, a multi-phased literature search was
conducted. Initially, a personal ProCite database maintained by one of the authors of the
document (GTA) was searched for studies with non-mammalian species using the 11 chemicals
identified in the previous paragraph. This database has been maintained for more than 20
years, and contains about 10,000 entries on different topics in environmental toxicology. To
generate additional coverage, searches focused on effects of the 11 target chemicals,
predominantly in fish, were conducted using PubMed, and the ECOTOX database U.S (2007).
The ECOTOX database is comprised of greater than 40,000 papers concerning the toxicity of
chemicals to, primarily, aquatic species. Prior to inclusion in the ECOTOX database, all papers
and their data are reviewed as to their quality using a set of well-defined criteria U.S (2007).
The examples and the associated analyses described in Section 4 relative to HPG-active
chemicals were restricted largely to studies with fish for practical considerations (i.e., available
data), and because there is well-documented structural and functional conservation of most
characteristics of the HPG axis across vertebrate classes (Norris 2006). Although there are
examples of NMDRs for putative EDCs in invertebrate species e.g.. BPA in snails; Oehlmann et
al. (2006). there is substantial uncertainty as to whether chemicals known or suspected to
impact HPG function in vertebrates operate via similar pathways in invertebrates. We also
restricted the scope of our review and analysis to those chemicals with adequately described
mode(s) of action as they pertain to the HPG axis. For example, due to uncertainties associated
with the data from atrazine exposed amphibians and fish, we did not include those in our
analysis Tillitt et al. (2010; Haves et al. (2003).
57 | Page
Androgen Receptor Antagonist
Flutamide
Vinclozolin
5-alpha reductase inhibitors
Finasteride
Semicarbazide
Inhibitors of P450 enzymes in steroidogenesis
Prochloraz
Androgen Receptor Agonists
Testosterone
Trenbolone
Selective Androgen Receptor Agonists
Ostarine
Andarine
Phthalates
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Additional studies were reviewed to assess the occurrence of NMDRs associated with the
thyroid hormone system offish and amphibian species. Numerous chemicals have been
investigated for thyroid effects in fish and amphibians. Among these are chemicals of
environmental concern with relatively little information regarding thyroid activity, as well as
recognized thyroid disrupting chemicals with relatively well established mechanisms of action.
Among the established mechanisms of action operative in thyroid disruption, this review
focused on those chemicals which have been shown to inhibit the sodium-iodide symporter
(NIS) or the thyroid peroxidase (TPO) enzyme as the molecular initiating event. Both NIS
transport and TPO activity represent critical steps in TH synthesis, and their inhibition can result
in a hypothyroid state. Perchlorate is a well known competitive inhibitor of NIS, which
concentrates iodide in the thyroid follicular cell. Methimazole and propylthiouracil (PTU) are
well known inhibitors of TPO, the enzyme that covalently binds iodide to tyrosine residues of
thyrogobulin and links two iodinated tyrosines to form iodinated thyronines, primarily as T4, in
the thyroid follicular lumen.
Published papers were identified through electronic searches (PubMed and ECOTOX) using
both chemical names such as, PTU, ethylene thiourea, propylthiourea, methimazole,
perchlorate, and protein names such as, sodium iodide symporter, NIS, thyroperoxidase, and
TPO. Searches were taxonomically limited to fish and amphibian species. References in
acquired papers were also reviewed for relevant publications.
The endpoints used to determine thyroid disrupting effects typically include: circulating T3
and/or T4 concentrations, thyroid gland histology, and selected expression of genes associated
with thyroid hormone homeostasis and action. Some studies included developmental
endpoints. Nearly all studies evaluated the effect of the test chemical on survival and growth,
measurements of weight and/or length, as a general indicator of toxicity. Studies that did not
report thyroid-specific endpoints were excluded, as were studies with single concentrations of
the test chemical. Statistical tests used in the studies were accepted as reported.
58 | Page
-------
Draft Deliberative - Do NOT Cite, Quote, or Distribute
State of the Science: Nonmonotonic Dose Responses V7
Table 4.1: Select studies from literature review displaying some evidence of non-monotonicity.
All studies are described in detail in the main text (pages 59-69).
Chemical
Species
(Life Stage)
Concentration
or Dose
Route
Test Type
(Duration)
References
Ethinyl estradiol
Fathead
minnow
(all)
0.32, 0.96, 3.5, 9.6,
23 ng/L
Water
Full Life Cycle
(150 days)
Parrott and Blunt (2005)
Bisphenol A
Zebrafish
(adult)
0.01, 0.1, 1, 10, 100
Hg/L
Water
Short Term
(4 days)
Villeneuve etal. (2012)
Bisphenol A
Fathead
minnow
(adult)
0.01, 0.1, 1, 10, 100
Hg/L
Water
Short Term
(4 days)
Villeneuve etal. (2012)
Bisphenol A
Fathead
minnow
(adult)
1, 16, 160, 640, 1280
Hg/L
Water
Partial Life Cycle
(164 d)
Sohoni et al. (2001)
Tamoxifen
Fathead
minnow
(all)
0.01, 0.08, 0.18,
0.56, 0.64, 1.8, 5.12,
5.6, or 18 ng/L
Water
Partial (42 d) or Full
Life Cycle (284 days)
Williams et al. (2007)
Trenbolone
Fathead
minnow
(adult)
0.005, 0.05, 0.5, 5.0,
50 ng/L
Water
Partial Life Cycle (21
days)
Anklev et al. (2003)
Methyl
testosterone
Nile tilapia
(juvenile)
3.75, 7.5, 15, 30, 60,
120, 240, 480, 600,
1200 mg/kg
Diet
Partial Life Cycle (28
days)
Phelps and Okoko (2011)
Methyl
testosterone
Zebrafish
(juvenile)
0.026, 0.05, 0.1,
0.26, 0.5, 1.0 ng/L
Water
Partial Life Cycle (40
days)
Orn etal. (2003)
59 | Page
-------
Draft Deliberative - Do NOT Cite, Quote, or Distribute
State of the Science: Nonmonotonic Dose Responses V7
Chemical
Species
(Life Stage)
Concentration
or Dose
Route
Test Type
(Duration)
References
Vinclozolin
Goldfish
(adult)
100, 400, 800 ng/L
Water
Partial Life Cycle (30
days)
Hatef et al. (2012)
Fadrozole
Fathead
minnow
(adult)
1.85, 5.55, 16.7, 50
Hg/L
Water
Short Term
(7 days)
Villeneuve etal. (2006)
Prochloraz
Zebrafish
(juvenile)
32, 38, 75, 100, 150,
300, 320, 600 ng/L
Water
Partial Life Cycle (60
days)
Holbech et al. (2012)
Ketoconazole
Fathead
minnow
(adult)
6, 25, 100, 400 ng/L
Water
Partial Life Cycle (21
days)
Anklev et al. (2007)
Ammonium perchlorate
Zebrafish
(adult)
18, 677 mg/L
(as perchlorate)
Water
Partial Life Cycle (56
days)
Patino et al. (2003)
Ammonium Perchlorate
Fathead
minnow
(embryo)
0.85, 8.47, 84.7
mg/L
(as perchlorate)
Water
Developmental
(28 days)
Crane et al. (2005)
6-Propylthiouracil
Zebrafish
(larvae)
2.5, 10, 25, 50 mg/L
Water
Partial Life Cycle (35
days)
Schmidt and Braunbeck (2011)
Methimazole
Fathead
minnow
(all)
32, 100, 320 ng/L
Water
Partial Life Cycle (84
days)
Crane et al. (2006)
Amitrole
Chinese rare
minnow
(juvenile)
1, 10, 100, 1000,
10,000 ng/L
Water
Partial Life Cycle
(28 days)
Li etal. (2009)
Magnesium perchlorate
Chinese rare
5, 50 ng/L
Water
Partial Life Cycle
Li etal. (2011)
60 | Page
-------
Draft Deliberative - Do NOT Cite, Quote, or Distribute
State of the Science: Nonmonotonic Dose Responses V7
Chemical
Species
(Life Stage)
Concentration
or Dose
Route
Test Type
(Duration)
References
minnow
(larvae and
adult)
(as perchlorate)
(21 days)
Ethylenethiourea
X. laevis
(larvae)
1.0, 2.5, 10, 25, 50
mg/L
Water
Developmental (90
days)
Opitz et al. (2006)
6-Propylthiouracil
X. tropicalis
(larvae)
2, 5, 10, 20, 75 mg/L
Water
Developmental
(14 days)
Carlsson and Norrgren (2007)
61 | Page
-------
Draft Deliberative
4.1.2
State of the Science: Nonmonotonic Dose Responses V7
NMDRs in the Estrogen Hormone System in Aquatic Species
4.1.2.1 Estrogen Receptor Agonists
A variety of estrogens have been studied in vivo in fish, with endpoints ranging from molecular
and biochemical responses (e.g., induction of the egg yolk protein precursor VTG in males) to
apical developmental and reproductive outcomes more typically used in risk assessment. Many
of these studies have focused on environmentally-relevant concentrations of the chemicals,
sometimes for relatively extended periods of time (i.e., multiple generations). For example,
Kidd et al. (2007) conducted a field study in which a whole lake was treated with EE2 for three
years (during the summer months), and impacts on the entire system (including fish
populations) were assessed.
Probably the most studied xenoestrogen in fish is EE2, which has been associated with
feminization of males exposed to municipal waste water treatment plants Purdom et al. (1994)
Routledge et al. (1998). A number of full and partial life-cycle tests have been conducted with
EE2 in a variety of fish species, including fathead minnows, zebrafish and medaka, all well-
characterized small fish models. There have been several independent analyses of these
studies in an attempt to derive robust predicted no-effect water concentrations (PNECs) for
reproductive effects of EE2, as a basis for generating protective water quality criteria for the
estrogen Caldwell et al. (2012; Caldwell et al. (2008; U.S (2008). These study reviews involved
critical evaluation of both the experimental designs employed (e.g., use of multiple EE2
concentrations, including those relevant to the environment) and quality of the data generated
(e.g., analytical confirmation of EE2 concentrations in the test water, validity of methods used
to measure endpoints, and nature of concentration-response relationships). Based on data
from approximately 10 of the highest quality EE2 chronic fish studies, a predicted no effect
concentrations (PNEC) for adverse apical effects of 0.1 ng EE2/L has been proposed, which is a
concentration lower than that reported in effluents, but above that typically found in surface
waters Caldwell et al. (2012). NMDRs were identified in these long-term studies, and one of
them reported nonmonotonic data at an EE2 test concentration below 1 ng/L. Parrott and
Blunt (2005) conducted a full life-cycle (150 day) test with the fathead minnow using EE2 water
concentrations of 0, 0.32, 0.96, 3.5, 9.6 and 23 ng/L. Concentrations above 3.5 ng/L resulted in
substantial reproductive effects, with essentially no viable eggs produced. Fecundity
(production of eggs) exhibited an NMDR as it was significantly greater than control values in the
0.32 and 0.96 treatments, but fertility also was reduced in these treatments; these off-setting
observations served to mitigate overall EE2 impacts in terms of total number of viable embryos
(i.e., number of fertile eggs). One complication relative to interpretation of the Parrott and
Blunt (2005) study was that EE2 test concentrations were not confirmed analytically in the 0.32
and 0.96 ng/L treatment groups. Overall, considered in isolation, it would be difficult to derive
62 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
a robust PNEC based on the shape of the low-end dose-response data in the Parrott and Blunt
(2005) study, however, when combined with other available chronic fish data, Caldwell et al.
(2012; Caldwell et al. (2010) felt that the derived value of 0.1 ng EE2/L would be protective of
effects on reproduction.
BPA acts both as an ER agonist and AR antagonist in fish Ekman et al. (2012), but most of the
work conducted concerning endocrine-related effects of BPA in fish has focused on the
chemical's estrogenic properties. Like EE2, BPA is commonly detected in a wide variety of
environmental matrices for a review see Staples et al. (1998). BPA is much less potent than EE2
relative to effects in fish but, nonetheless, estrogenic responses (e.g., induction of VTG) have
been observed at the low |ag/L water concentrations of BPA occasionally encountered in the
environment Kang et al. (2007). Villeneuve et al. (2012) reported a relatively unusual dose-
response relationship for BPA in studies with the fathead minnow and zebrafish. In those
experiments, reproductively-active male and female fish were exposed for 4 d to 0.01, 0.1, 1, 10
or 100 |ag BPA/L water, following which ovarian gene expression (determined using
microarrays) and plasma VTG concentrations were assessed. In both zebrafish and fathead
minnows, NMDRs were observed for expression of ovarian genes. Specifically, the greatest
number of differentially-expressed genes, by far, was observed in the 10 |ag/L treatment group.
In the fathead minnow, the second largest number of differentially-expressed genes after the 4-
d exposure was seen in fish exposed to 0.01 |ag BPA/L. In contrast to the nonmonotonic effects
of BPA on gene expression, there was a concentration-dependent induction of VTG in BPA-
exposed males of both species, with the threshold for response occurring in the 10 |ag/L
treatment group. Plasma VTG also was induced in a monotonic fashion in fathead minnow
females; there was no indication of increased VTG in zebrafish females.
There also have been longer-term experiments concerning the effects of BPA on endocrine-
mediated responses and reproductive success in the fathead minnow Mihaich et al. (2012;
Sohoni et al. (2001). Both studies exposed reproductively-active adult fish to BPA for 164 d at
water concentrations of 1,16, 160, 640, and 1280 |ag/L by Sohoni et al and at 1, 16, 64, 160,
and 640 |j,g/L by Mihaich et al. (2012). Reductions in egg production were observed only at
1280 |-ig/L; however, changes in the relative gonad weight (gonadal-somatic index; GSI) and
histology also occurred at lower test concentrations. Plasma VTG in both sexes was induced at
BPA concentrations of > 64 |ag/L. Most endpoints in the two studies exhibited monotonic
concentration-response relationships. One clear example of an NMDR, however, was the GSI of
females in the Sohoni et al. (2001) study. After 43 d of exposure GSI was slightly, but
significantly greater than controls in after treatment with 1,160 and 640 |ag BPA/L, but was not
affected in animals exposed to 16 or 1280 |ag BPA/L. The cause of this is uncertain, but it is
noteworthy that by conclusion of the assay (164 d) there was a significant monotonic (negative)
relationship between BPA exposure concentration and female GSI. Another example of a
63 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
possible NMDR from the Sohoni et al. (2001) study involved the relative distribution of
testicular male sex cell types across the various treatment groups at 164 d, however, Mihaich et
al. (2012) questioned the functional significance of this observation since successful egg
production and fertility occurred at all but the very highest BPA concentration.
4.1.1 sctive Estrogen Receptor Modulator (Tamoxifen)
One fish study was identified in which the long-term toxicological effects of the SERM
tamoxifen were examined Williams et al. (2007). Partial (42-d) and full-life (284 d) life-cycle
tests were conducted with fathead minnows exposed to tamoxifen concentrations in the water
ranging from 0.01-18 |ag/L (0.01, 0.08, 0.18, 0.56, 0.64, 1.8, 5.12, 5.6, or 18 ug/L, depending on
stage in the lifecycle). Very limited monitoring data (from the UK) suggested that tamoxifen
could occur in surface water at a concentration as high as 0.2 |ag/L. In the 42-d study,
conducted with sexually mature fish, tamoxifen decreased plasma VTG concentrations in
females to about the same degree at test concentrations of 0.56 |ag tamoxifen/L and higher,
which is consistent with antagonism of the estrogen receptor. However, in the full life-cycle
test, tamoxifen did not affect VTG concentrations in F0 or F1 females at test concentrations up
to, and including 5.12 |ag/L. Interestingly, the SERM induced VTG in a dose-dependent manner
in F0 males exposed to 0.64 and 5.12 |ag tamoxifen/L, as well as in F1 males in the 5.16 |ag/L
treatment group. The drug did not affect any endpoints related to reproduction (egg
production, hatching success) at concentrations < 5.12 |ag/L. There were some effects of
tamoxifen on larval (Fl) size (weight, length) in the higher dose groups (0.08, 0.64, 5.12 |ag/L),
but this effect was transitory in that it was observed 28 d post-hatch, but not at 112 d. Larval
size at 112 d was smaller in fish exposed to 0.01 |ag tamoxifen/L than in controls or the other
treatment groups. It is uncertain, however, whether this seeming nonmonotonic growth
response should be considered biologically significant Williams et al. (2007). and/or whether it
is related to estrogen receptor-mediated responses to tamoxifen in the fish.
4.1.3 NMDRs in the Androgen Hormone Pathway in Aquatic Species
4.1.3.1 Androgens
Compared to chemicals that interact with estrogen receptors (primarily agonists), less is known
concerning the toxicological effects of AR agonists in long-term fish tests. Although fewer in
number, we identified several full and partial life-cycle tests with fish exposed to the synthetic
steroids Trenbolone and methyl-testosterone, including some examples of NMDRs.
Ankley et al. (2003) describe a study in which effects of the steroidal androgen TB, a growth-
enhancer used for livestock production, was evaluated at multiple biological levels of
64 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
organization in reproductively-active fathead minnows, resulting in NMDRs for some
biochemical endpoints but monotonicity of apical endpoints. The fish were exposed to water-
borne TB concentrations spanning 5 log units (0.005, 0.05, 0.5, 5.0, or 50 |ag/L) for 21 days.
Based on studies at beef feedlots, the lower two test concentrations would be considered
environmentally-relevant Durhan et al. (2006). Consistent with the anabolic nature of TB, body
weight of exposed female fathead minnows was increased in a concentration-dependent
manner. The androgen also caused morphological masculinization of the females, resulting in a
concentration-dependent induction of male-like dorsal nuptial tubercles. Exposure to TB had
relatively rapid and profound effects on cumulative egg production in the fish, with monotonic
response. In contrast to these apical responses, several key biochemical measures of HPG
function monitored in TB-exposed female fathead minnows exhibited a U-shaped dose-
response relationship, with decreased responses occurring at the higher test concentrations.
There was some evidence of this relative to plasma VTG concentrations, with much more
pronounced nonmonotonic relationships for plasma 17|3-estradiol (E2) and, especially,
testosterone (T) concentrations (Fig. 4.1a-c). The fact that these three endpoints are
functionally related to one another lends veracity to the nature of the U-shaped dose-response
curve. Specifically, T is the metabolic precursor to E2 (via a reaction catalyzed by cytochrome
P450 aromatase [CYP19]), and E2 activation of hepatic estrogen receptors is responsible for
VTG production in fish. The basis of the U-shaped dose-response relationship(s) is uncertain,
but the response is suggestive of some sort of compensatory mechanism in the fish. However,
further exploration of this possibility would require measurements of basic endocrine function
and signaling (e.g., temporal changes in gonadotrophins) not included in the original study.
65 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
6.00
4.50
~ 3.00
£
O)
1.50
0.00
6.00
4.50
g 3.00
0.00
30
24
I 18
D)
E
.¥ 12
c
<1)
D)
_o
6
>
0
Figure 4.1: Effects of a 21-d water-borne exposure to 17b-trenbolone on plasma
concentrations of (a) testosterone (T), (b) 17(J-estradiol (E2) and vitellogenin in female
fathead minnows.
Anklev et al. (2003)
Following the work by Anklev et al. (2003) several other researchers have also used TB as a
model androgen in fish studies. For example, Seki et al. (2006) described a study in which adult
male and female fathead minnow, Japanese medaka and zebrafish were exposed to TB for 21-d
at water concentrations of 0.05, 0.5 and 5 |-ig/L. Consistent with expectations, the androgen
masculinized female fathead minnows and medaka, and depressed (female) plasma VTG
concentrations in all three species. Absolute sensitivity relative to impacts of TB on these
endpoints was species-specific, but all responses identified were monotonic. Hemmer et al.
(2008) also conducted a 21-d study with the sheepshead minnow, an estuarine species,
Control 0.005 0.05 0.5 5.0
Trenbolone (ug/l)
66 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
exposed to 0.005, 0.05 and 5 |ag TB/L water. There were no significant effects noted at the
lower two TB concentrations, but 5 |ag/L significantly reduced egg production. Decreased egg
production corresponded with reduced plasma VTG concentrations in female sheepshead
minnows although, due to among individual variability, this response was not statistically
significant. Cripe et al. (2010) assessed the effects of TB on sheepshead minnow reproduction
in a three-generation (42 week) experiment that started with reproductively mature adults (F0)
and concluded with an assessment of reproductive success of F2 animals. Target water
concentrations were 0.01, 0.04, 0.2,1.0 and 5.0 |ag TB/L. Some of the somatic measurements
exhibited nonmonotonic relationships. For example, F1 and F2 fish from the two lowest TB
treatments were longer than the controls by 90 or 28 dph, respectively, and the GSI of F0
females was larger in fish from the 1.0 than 5.0 treatment group than in untreated animals.
However, these types of variations did not seem to be systematic. Apical endpoints directly
related to reproduction were monotonic when different from control in the study by Cripe et al.
(2010). Although there were differences in the absolute sensitivity of the fish across the three
generations (i.e., in terms of lowest observed effect concentrations of TB), effects on
cumulative fecundity, fertility, embryo hatch, and percentages of abnormal embryos and
infertile eggs all exhibited monotonic dose-response relationships. Cripe et al. (2010) also
noted depressions in plasma VTG concentrations in F0 females from the highest two TB
treatment groups.
Effects of TB also have been evaluated in sexual development assays with zebrafish. Holbech et
al. (2006) exposed newly-fertilized zebrafish eggs to target concentrations of 0.005, 0.05, 0.5 or
5 |ag TB/L for 59 d, following which they evaluated phenotypic (gonadal) sex of the fish. In all
but the lowest treatment group 100% of the zebrafish were scored as males at the end of the
test—there was no indication of non-monotonicity in the responses. Morthorst etal. (2010)
conducted an experiment and published a set of results qualitatively similar to that of Holbech
etal. (2006). except that they monitored sex of the fish (held in clean water) for up to 170 d
post-exposure. Effects of the TB on sexual differentiation were irreversible in that timeframe.
Steroidal chemicals, both estrogens and androgens, have been used for many years in fish
culture to produce mono-sex populations with desirable attributes such as enhanced growth.
Although a substantial amount of work has been done in this area, the types of studies
conducted to support aquacultural practices have limited relevance to assessing potential
toxicological properties and risks of endocrine-active chemicals for a number of reasons: (1)
exposure often is dietary, rather than via the water, leading to uncertainties as to actual dose
(supporting, for example, extrapolation of results to field settings); (2) concentrations of
chemicals added to the diet typically are quite large—much higher than could possibly occur in
the environment; and (3) many of the steroids used for aquaculture have not been implicated
as environmental contaminants.
67 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
A recent study by Phelps and Okoko (2011) reflects all of these characteristics, but nonetheless
is quite interesting in terms of producing a relatively novel NMDR. In that study Nile tilapia fry
were fed diets containing 3.75, 7.5,15, 30, 60, 120, 240, 480, 600, or 1200 mg/kg of
methyltestosterone (MT, a very commonly used steroid in aquaculture) for 28d, following
which the animals were sexed by examination of the gonads. A robust inverted U-shaped dose-
response was observed, wherein intermediate doses produced completely male populations,
while in the control and high-dose groups fish exhibited a normal sex ratio of around 50:50
(Figure 4.2). Past studies with MT have shown that it can be converted to methylestradiol in
fish (via CYP19), thereby producing seemingly "paradoxical" estrogenic and androgenic
responses simultaneously Hornung et al. (2004; Ankley et al. (2001). However, Phelps and
Okoko (2011) did not think that this adequately explained the unusual dose-response they
observed. Further work would be required to determine the mechanistic basis of this NMDR.
MT mg/kg diet
Figure 4.2: Effects of a 28-d dietary exposure to methyltestosterone on sex ratio in Nile
tilapia.
Phelps and Okoko f20111
Other more toxicologic studies with MT have not typically produced NMDRs. For example,
Pawlowski etal. (2004) exposed adult fathead minnows to water concentrations of 0.1, 1, 5 and
50 |ag MT/L for 21 d during gonadal recrudescence (moving from a "winter-type" non-spawning
condition to a "summer-type" condition, through temperature and photoperiod manipulation),
68 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
following which a number of endpoints related to HPG function were measured. Consistent
with previous studies (e.g., Ankley et al. (2001)), MT exhibited dual endocrine activities,
masculinizing females (production of male secondary sex characteristics) and feminizing males
(inducing VTG). Exposure to MT also inhibited egg production in fish subsequently moved to
clean water for 21 d. Effects of MT on secondary sex characteristics, VTG, fecundity and fertility
were all monotonic. Kang et al. (2008) exposed adult Japanese medaka to 0.0225, 0.0468,
0.0881, 0.188 and 0.380 |ag MT/L of water for 21 d. Fecundity and fertility of the fish, and VTG
concentrations in females were depressed in a concentration-dependent manner at
concentrations > 0.0468 |ag/L. The relative gonad weight of females was increased to a
relatively similar degree in the four higher treatment groups. Seki et a I. (2004) conducted a
two-generation experiment, also with medaka, at nominal water concentrations of 0.00031,
0.00098, 0.00313, 0.01 and 0.032 |ag/L. The test was initiated with newly-fertilized eggs,
continued through reproduction by the F0, and concluded after 60 d post-hatch of the F1
generation. The highest test concentration strongly skewed the F0 gonadal sex toward males
by 61 d post-hatch. The lower test concentrations did not cause additional marked effects
during the F0 generation, but in F1 animals MT induced hepatic VTG in males (0.01 |ag/L group),
and reduced hepatic VTG in females, to about the same extent, in all the remaining lower
treatment groups (i.e., 0.00031-0.01 |-ig/L).
One toxicologically-relevant study with MT did exhibit a NMDR. Orn et al. (2003) exposed
juvenile (20 days post-hatch) zebrafish to MT water concentrations ranging from 0.026, 0.05,
0.1, 0.26, 0.5, or 1 |ag/L for 40 d, following which gonad morphology and whole body VTG
concentrations were assessed. Complete sex reversal occurred in all fish exposed to MT, with
no ovaries observed in animals from any treatment group. Whole-body VTG concentrations in
the fish exhibited a "U-shaped" dose-response relationship, with smallest values occurring in
the intermediate treatment groups. This response could plausibly be linked to the conversion
of methyl testosterone to methylestradiol, and subsequent activation of the estrogen receptor
at the higher exposure concentrations.
4.1.3.2 Androgen Receptor Antagonists
No fish full life-cycle tests with the AR antagonists flutamide or vinclozolin were identified.
However, endocrine-related effects of the two chemicals have been examined in multiple
partial life-cycle tests with fish. Bayley et al. (2002) fed a diet amended with 0.01 or 1 mg
flutamide/kg to guppies from birth to adulthood (ca. 26 wk) and evaluated several endpoints.
They found that flutamide caused a skewing of sex ratio toward females (60-70% of the total
population) at both doses. Sexual maturation of the males was delayed in a dose-dependent
manner, and size of males at maturation was significantly reduced in the high flutamide
treatment. Flutamide also caused a decrease in gonopodium length (a male secondary sex
69 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
characteristic) in both treatment groups. In a reproduction study with flutamide, Jensen et al.
(2004) exposed adult fathead minnows via water to 50 or 500 |ag flutamide/L for 21-d. Egg
production was significantly reduced in the high treatment group, and the flutamide caused
slight alterations in plasma sex steroid and/or VTG concentrations in both sexes, but all of these
changes were monotonic.
Bayley et al. (2002) also examined the effects of the pesticide vinclozolin on development of
guppies fed a diet with 0.1 or 10 mg/kg for 26 wk starting at birth. Responses observed were
sex ratios biased toward females, delayed maturation and growth in males, and decreased
gonopodium length. Vinclozolin also caused a dose-dependent decrease in sperm count in the
guppies. In a follow-up study, Bayley et al. (2003) reported that male guppies fed a diet
containing 0.1,1 or 10 mg vinclozolin/kg for 6 wk exhibited concentration dependent decreases
in sperm count and number of first-clutch juveniles sired. Martinovic etal. (2008) assessed the
effects of vinclozolin on reproductively-active fathead minnows exposed via the water to 60,
255 or 450 |ag/L. They found a dose-dependent decrease in egg production in the fish, as well
as a depression in normal secondary sex characteristics in males. Other vinclozolin-induced
changes in plasma VTG concentrations, sex steroid production and/or expression of transcripts
for the androgen receptor occurred in male or female fish, but none of these alterations
exhibited NMDRs. Hatef etal. (2012) exposed goldfish to 100, 400 or 800 |ag vinclozolin/L for 1
month. They noted a significant decrease in sperm quantity and quality in males from the high
treatment group. Plasma concentrations of 11-ketoteststerone (a fish-specific androgen) in the
males exhibited an NMDR, with significantly elevated levels of the steroid in the 100 |ag/L
treatment group and depressed levels in the 800 |ag/L fish. The increase in 11-ketoteststerone
at the lower vinclozolin concentration observed by Hatef et al. (2012) might reflect a
compensatory response in the fish resulting from reduced androgen signaling. The authors
suggested that this increase in 11-ketoteststerone could be responsible for a lack of adverse
effects on sperm quantity and quality in the 100 |ag/L treatment group.
4.1.3.3 Steroid Synthesis Inhibitors
A number of model chemical inhibitors of sex steroid synthesis have been tested in short-term
or partial life-cycle assays with fish. For example, several fish studies have been conducted with
the drug fadrozole, a relatively specific inhibitor of CYP19 A, which originally was developed to
treat estrogen-dependent breast cancers in humans. In a short-term study, Villeneuve et al.
(2006) exposed sexually-mature female fathead minnows for 7 d to 1.85, 5.55, 16.7 or 50 |ag
fadrozole/ L water. They found that fadrozole reduced aromatase activity in both the brain
(CYP19B) and ovaries (CYP19A) of the fish. Effects in the brain followed a monotonic dose-
response relationship, whereas inhibition of aromatase activity by fadrozole in the gonad
exhibited an inverted U-shaped dose response. Fadrozole also produced a concentration-
70 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
dependent increase in expression of CYP19A in the ovaries, leading Villeneuve etal. (2006) to
speculate that the basis of the NMDR was due to interplay between the increased CYP19A
transcripts (a compensatory response) and direct inhibition of the enzyme. By contrast, Ankley
etal. (2002) saw no evidence of NMDRs in a longer-term, partial life-cycle study with fadrozole
in adult fathead minnows, conducted at water concentrations of 2, 10 and 50 |ag/L. Over the
course of the 21 d reproduction test, the drug caused concentration-dependent reductions in
egg production in the fish, as well as decreased plasma E2 and VTG concentrations in females.
In a study focused on sexual development, Kwon et al. (2000) exposed Nile tilapia genetic
females to fadrozole via the diet for 30 d starting 7 d post-hatch, and assessed effects on
phenotypic sex ratios. They found a dose-dependent increase in the incidence of males at
doses ranging from 0 to 200 mg fadrozole/kg food; at dietary concentrations from 200 to 500
mg/kg, the percentage of males remained constant as about 92-96%.
Fish partial life-cycle studies also have been conducted with prochloraz, trilostane and
ketoconazole, all known inhibitors of sex steroid synthesis in mammals. Ankley et al. (2005)
described a 21-d fathead minnow reproduction test with prochloraz, an imidazole designed to
inhibit fungal CYP14a-demethylase (CYP51), which catalyzes a key step in ergosterol
biosynthesis supporting cell wall formation. Prochloraz, like many other fungicides designed to
inhibit fungal CYP51s, also inhibits a variety of vertebrate CYPs involved in steroid production,
including CYP19 and CYP17 (hydoxylase/lyase). Fathead minnows exposed to 30, 100 and 300
lag prochloraz/L water exhibited a concentration-dependent decrease in egg production over
the course of 21 d, a response coincident with monotonic decreases in plasma E2 and VTG in
the female fish Ankley et al. (2005). Similarly, Zhang et al. (2008) reported a dose-dependent
decrease in hepatic expression of VTG transcripts and egg production in adult Japanese medaka
exposed for 7 d to 3, 30 or 300 |ag prochloraz/L.
The effects of prochloraz also have been assessed in studies focused on sexual differentiation
and development in fish. Kinnberg et al. (2007) exposed zebrafish to 16, 64 or 202 |ag
prochloraz/L water for 60 d starting at 24 h post-hatch. The authors reported a significant shift
in sex ratio in fish from the 202 |ag/L treatment, with a slight bias toward males. Similar to
observations made by Ankley et al. (2005) and Zhang et al. (2008). prochloraz decreased
(whole-body) concentrations of VTG in female zebrafish. Concentrations of VTG in males were
depressed in the 202 |ag/L group, but they were slightly elevated in the two lower treatment
groups. Although this endpoint displayed an NMDR, the biological significance of the response
is uncertain as VTG levels in all males from the study were very low (several orders of
magnitude below females), and VTG has no known functional role in male fish. Holbech et al.
(2012) describe an effort focused on standardization of a test protocol for endocrine-active
chemicals, in which the effects of prochloraz on sexual development were evaluated in
zebrafish and the fathead minnow, in a ring test conducted by five labs. Animals were treated
71 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
with prochloraz from days 0 through 60 post-hatch to concentrations of 32, 38, 75,100,150,
300, 320, or 600 |ag/L water. In both species prochloraz caused a significant skewing of the sex
ratio towards males, and a depression of VTG concentrations in females. With one exception,
all of the effects data were monotonic. One of the labs testing zebrafish did not observe
statistically significant masculinization at the highest prochloraz concentration tested but did
see significant masculinization at the next two lower test concentrations. However, two other
labs involved in the ring test reported significant masculinization of zebrafish at concentrations
comparable to the high treatment group in the first lab Holbech et al. (2012).
Trilostane is a human pharmaceutical designed to inhibit 3|3-hydroxysteroid dehydrogenase
(3|3HSD), which catalyzes the conversion of pregnenolone to progesterone. Villeneuve et al.
(2008) evaluated the effects of trilostane on endocrine function in several studies with the
fathead minnow. Initial studies showed that trilostane effectively inhibited the in vitro
production of E2 by ovary explants in a dose-dependent manner, while a 21-d in vivo exposure
at 60, 300 and 1500 |ag trilostane/L indicated a concentration-dependent depression of plasma
VTG concentrations in females, and a significant decrease in egg production in the high
treatment group.
Ankley et al. (2007) described a series of fathead minnow studies with ketoconazole, a
fungicide capable of inhibiting multiple steriodogenic CYPs in mammals, most notably CYP17
and CYP11A (CYP cholesterol side-chain cleavage). In an initial 7-d range-finding experiment^
ketoconazole depressed plasma VTG concentrations in female fish, a response consistent with
inhibition of steroid synthesis. In a subsequent 21 d reproduction study, fish were exposed to
6, 25, 100 and 400 |ag ketoconazole/L water; ex vivo production of T by gonad explants from
both sexes was decreased in the three highest treatment groups. However, this did not
translate into decreased plasma concentrations of T or E2 in either sex, or VTG in females. The
unexpected lack of impact of ketoconazole on these in vivo functional measures of HPG status
appeared to be due to compensatory responses in the fish, manifested as increased GSI, altered
gonad histopathology (e.g., interstitial cell proliferation in males), and up-regulation of genes
coding for steriodogenic proteins, including CYP17 and CYP11A. An NMDR was observed
relative to effects of ketoconazole on production of eggs, which was significantly depressed in
the 25 and 400, but not 100 |ag/L treatment groups. It is uncertain whether the compensatory
molecular and histological responses were the cause of the nonmonotonic effect of
ketoconazole on egg production. Zhang et al. (2008) conducted a 7 d reproduction test with
3.0, 30, and 300 ug/L ketoconazole in the Japanese medaka. They reported a significant
decrease in egg production at the highest concentration tested (300 |ag/L), with no evidence of
a nonmonotonic relationship at two lower test concentrations (3 and 30 |-ig/L).
72 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
4.1.3.4 Hole of Compensatory Processes in NMDR - Examples fr h
Time-Course Studies with Three EDCs
Some of the fish EDC studies summarized in the previous section are suggestive of the
occurrence of NMDRs due to compensatory mechanisms. To explore this hypothesis more
directly, data from a series of intensive time-course studies with fathead minnows exposed to
different classes of EDCs Ankley et al. (2009a) were reassessed. In the experiments described
below, the animals were exposed to fadrozole, prochloraz or TB (Villeneuve Ekman et al. (2011;
Ankley et al. (2009b; Villeneuve et al. (2009). Although the three chemicals operate via
different molecular initiating events, past studies have shown that they all effectively depress T
and/or E2 synthesis and plasma concentrations of VTG in female fathead minnows, ultimately
resulting in depressed egg production Ankley et al. (2010).
The time-course studies with fadrozole, prochloraz and TB utilized the same basic experimental
design Ankley et al. (2009a). Sexually mature fathead minnows (of both sexes) were continually
exposed to two different, analytically confirmed, concentrations of the chemicals via water in a
continual-flow system. The two test concentrations used for each chemical were indexed to
data from 21-d reproduction studies Ankley et al. (2003; Ankley et al. (2002), such that the high
concentration was associated with substantial reproductive impacts (e.g., total cessation of egg
production), while the low concentration was expected to have slight (or no) effects on egg
production. Fish were sampled 1, 2, 4 and 8 d after initiation of the chemical exposure, and 1,
2, 4 and 8 d after ceasing exposure (i.e., animals were held in constantly-renewed control
water). A variety of measurements indicative of HPG function were made, including plasma VTG
concentrations, ex vivo gonadal production and plasma concentrations of T and E2, and gonadal
expression (determined via real-time quantitative polymerase chain-reaction [PCR]) of several
genes coding for proteins involved in endocrine signaling and steroid production. Specifically,
NMDRs were defined as those wherein one (or both) of the treatment groups differed
significantly from the control at a given sampling time, but the relationship of the two
treatment groups to one another did not exhibit a monotonic rank order. Studies with more
than two test concentrations per time point would have, of course, been preferable for the
analysis, but the data from these studies nonetheless illustrate the role of compensatory
processes in producing unanticipated dose-response relationships.
Consistent with inhibition of CYP19A, exposure of fish to fadrozole depressed ex vivo ovarian E2
production and plasma concentrations of E2 and VTG in female fathead minnows Villeneuve et
al. (2009). Reductions in E2 were significant, and concentration-dependent, on days 1 and 2 of
the exposure in both the 3 and 30 |ag/L treatment groups. While plasma E2 remained
depressed in the high treatment throughout the course of the 8 d exposure, in the 3 |ag
fadrozole/L treatment, E2 concentrations showed significant recovery during the exposure
73 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
phase of the test to levels comparable to controls by 8 d. Upon cessation of the fadrozole
exposure, plasma E2 concentrations in the 30 |ag/L treatment group returned (within 2 d) to
control levels, but in the low fadrozole treatment there appeared to be an "overshoot" in
plasma E2 concentration relative to controls on day 1 post-exposure, such that there was an
NMDR. The behavior of plasma E2 concentrations, especially in the low-dose treatment was
indicative of compensation in the female fish. A component of this compensatory response
appeared to be related to up-regulation of genes coding for several key proteins involved in
steroidogenesis, including CYP19A, CYP11A and steroid acute regulatory protein. There were
instances over the course of the study in which expression of one or more of these genes did
not vary in a monotonic fashion. For example, on day 1 of the recovery phase of the test,
expression of CYP19A and CYP11A transcripts were higher in fish from the 3 |ag/L treatment
than in controls, while levels of the two gene products in fish from the 30 |ag/L treatment group
were intermediate to those of the control and 3 |ag/L treatment groups.
Prochloraz appears to depress steroid synthesis in fish via inhibition both of CYP19A and CYP17
Ankley et al. (2005). In a time-course study, ovarian ex vivo E2 production and plasma E2
concentrations in females both were decreased by exposure to prochloraz Ankley et al.
(2009b). After 1 d of exposure, the two test concentrations of the pesticide (30 and 300 |ag/L)
had depressed plasma E2 concentrations in a dose-dependent manner. Plasma E2
concentrations remained depressed in the high treatment during the 8 d exposure phase of the
test, but in the 30 |ag/L group, E2 recovered to levels comparable to or exceeding the controls
by day 4 of the exposure. Ex vivo E2 production by ovarian explants displayed a similar pattern
relative to plasma responses observed on day 4, in that E2 synthesis in the low dose group was
significantly greater than controls, while production in females exposed to 300 |ag prochloraz/L
was lower than control values. Hence, the plasma and ex vivo E2 data from the day 4 females
both exhibit nonmonotonic relationships due to the compensatory overshoot-type response in
the 30 |ag/L treatment. Interestingly, during the exposure phase of the test, plasma
concentrations of VTG were depressed in the high- but not low-dose prochloraz treatment. Up-
regulation of several ovarian genes coding for proteins involved in steroid production, including
CYP19A, CYP11A and CYP17, were observed in prochloraz-exposed animals Ankley et al.
(2009b).
Past experiments have shown that TB decreases plasma T and E2 concentrations in female
fathead minnows, presumably through feedback inhibition of steroidogenesis (Fig. 4.1a-c;
Ankley et al. (2003)). In the time course study, exposure of females to both test concentrations
of TB (0.05, 0.5 |-ig/L) depressed ex vivo E2 production and plasma E2 concentrations within 1-2
d Ekman et al. (2011). Plasma E2 levels in fish from the low-dose group exhibited what was
thought to be a compensatory response, rebounding during the exposure phase of the test to
concentrations comparable to (or even perhaps exceeding) the controls, while in the high dose
74 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
group plasma E2 concentrations in the female fish returned to control levels only after
cessation of TB exposure. Plasma VTG concentrations were slightly decreased in the 0.05 |ag
TB/L group during the exposure, returning to control levels within 1 d of depuration; VTG
concentrations in the 0.5 |ag/L females were strongly depressed during the chemical exposure,
but also recovered to control levels by the latter part of the recovery phase of the experiment.
Relative expression of several ovarian genes involved in HPG function changed over the course
of the TB exposure. Expression of transcripts for CYP11A in fish from the low dose group were
lower, higher and lower than corresponding control values on exposure days 2, 4, and 8,
respectively. On each of these occasions, the dose-response relationship between the control
and treatment groups was nonmonotonic. There were also several occurrences of NMDR
between treatment level and expression of gene products for VTG and the AR in the ovaries of
TB-exposed females.
The fathead minnow studies with fadrozole, prochloraz and TB are a subset of a larger group of
time-course experiments with different endocrine-active chemicals, several of which show
comparable results to those described herein Ankley et al. (2009a). For example, similar
response patterns, including the occurrence of NMDRs for endpoints reflecting gene expression
and sex steroid status, have been observed with ketoconazole (an inhibitor of multiple
steriodogenic CYPs other than, or in addition to, CYP19A) and trilostane, a specific inhibitor of
33-hydrosysteroid dehydrogenase, which converts pregnenolone to progesterone in the
steriodogenic cascade (Ankley et al. 2011; 2012). Although these different time-course studies
were not designed specifically to generate or examine the basis of NMDRs, they nonetheless
are illustrative of the important role of compensation in generating NMDR. Although there
were variations among the time-course studies with fadrozole, prochloraz and TB, several
common patterns were observed. For example, plasma E2 concentrations (Figs 3b; 4b; 5b)
exhibited a consistent depression early (1-2 d) at both the low and high test concentrations.
Subsequent to this, but still during the chemical exposure, plasma E2 in the low-dose groups
(i.e., at water concentrations associated with minimal effects on egg production in a 21-d
reproduction test) exhibited compensatory behavior, returning to (and occasionally exceeding)
control levels of E2. In the high dose groups (i.e., at concentrations associated with substantial
impacts on egg production over 21 d), plasma E2 concentrations tended not to return to levels
comparable to controls until after the chemical exposure had ceased. It was during these
periods of active compensation and early recovery when NMDRs for endpoints related to
steroid production and/or gene expression were most commonly observed. This pattern
suggests, perhaps, that NMDRs may be more prevalent in shorter-term assays, especially during
periods of system disequilibrium. Further, these data clearly highlight the time-dependent
nature of dose-response relationships.
75 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Figure 4.3: Time-course (8-d exposure and 8-d recovery) effects of fadrozole on female
fathead minnow.
(a) ovarian synthesis of 17|3-estradiol (E2); plasma concentrations of (b) E2 and (c) vitellogenin
(Vtg); and ovarian expression of mRNA for (d) cytochrome P450 (CYP) 19A, (e) CYP11A, and (f)
steroid acute regulatory protein (StAR). Differences from control are indicated by "#" and
from Villeneuve et al. (2009).
76 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Day Day
Figure 4.4: Time-course (8-d exposure and 8-d recovery) effects of prochloraz on female
fathead minnow.
(a) ovarian synthesis of 17|3-estradiol (E2); plasma concentrations of (b) E2 and (c) vitellogenin
(Vtg); and ovarian expression of mRNA for (d) cytochrome P450 (CYP) 19A, (e) CYP11A, and (f)
CYP17. Differences from control are indicated by asterisks from Ankley et al. (2009b).
77 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Day Day
Figure 4.5: Time-course (8-d exposure and 8-d recovery) effects of trenbolone on female
fathead minnow.
(a) ovarian synthesis of 17|3-estradiol (E2); plasma concentrations of (b) E2 and (c) vitellogenin
(Vtg); and ovarian expression of mRNA for (d) cytochrome P450 (CYP) 19A, (e) androgen
receptor (AR), and (f) Vtg receptor. Differences from control are indicated by asterisks (from
Ekman et al. (2011)).
78 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
The time-course experiments also provide insights as to the types of endpoints that might be
expected to exhibit NMDRs. For example, in the fadrozole, prochloraz and TB time-course
studies there were several instances where changes in gene expression were nonmonotonic.
There also were examples of either plasma steroid concentrations or ex vivo steroid production
exhibiting nonmonotonic relationships with dose. Based on this we hypothesize that responses
at molecular and biochemical levels, such as those involved in feedback regulation of
homeostasis (including compensation), may more commonly exhibit nonmonotonic
characteristics than more integrated, downstream endpoints. Notably, this is consistent with
the results of other studies described above, such as the 21-d fathead minnow reproduction
test with TB, where NMDRs were observed for a number of biochemical endpoints in females,
but not apical responses such as changes in weight, induction of male secondary sex
characteristics or egg production Fig, la-c; Ankley et al. (2003). Villeneuve etal. (2012) and
Osachoff et al. (2012) further highlight the variability that can be observed in terms of the
nature of dose-response relationships for changes in gene expression. As described above,
Villeneuve etal. (2012) noted NMDRs for the effects of BPA on global gene expression
(measured via microarray) in the ovaries of zebrafish and fathead minnows. Osachoff et al.
(2012) conducted time-course microarray experiments with Chinook salmon exposed to
different concentrations of sewage effluent which, depending upon the gene examined,
produced a variety of monotonic and nonmonotonic relationships that varied over time in livers
of the animals.
Finally, and most significantly, although NMDRs were observed in the short-term time course
studies with fadrozole, prochloraz and TB for gene expression and steroid endpoints, longer-
term (21-d) studies with these same three chemicals, in the same dose range used for the time-
course work, did not produce NMDRs for apical endpoints such as changes in secondary sex
characteristics, growth, fertility or fecundity Ankley et al. (2005; Ankley et al. (2003; Ankley et
al. (2002).
4.1.4 NMDRs roid Hormone Pathway in Aquatic Species
4.1.4.1 Fish
There are no existing test guidelines routinely used to assess the effects of chemicals on thyroid
function in fish. Therefore, the literature reviewed here is comprised of studies that include
various chemicals, species, protocols, and endpoints. The most commonly studied species was
zebrafish, and fewer studies used stickleback, fathead minnow, Chinese rare minnow, eastern
mosquitofish, coho salmon, and sea lamprey. A total of 38 studies were evaluated. Five were
eliminated due to use of a single test concentration. Fourteen studies were eliminated based
on the use of chemicals with insufficiently defined MoA. Two studies were eliminated due to
79 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
lack of thyroid-specific endpoints. Of the remaining 17 studies, 14 included the NIS inhibitor,
perchlorate, and 6 included a TPO inhibitor.
There were 14 studies in fish using perchlorate anion, which included more than 1 exposure
concentration and included thyroid specific endpoints. These include studies with embryos,
larvae, juveniles, subadults, and adults. The number of exposure concentrations ranged from 2
to 7 and the durations of the different studies ranged from 3 days to 16 weeks. Thyroid
histology was the most commonly used endpoint, followed by plasma and whole body
measurements of T4 and/or T3.
Thyroid histological changes included follicular cell hypertrophy and hyperplasia, changes in
follicular size and colloid characteristics, and overall thyroid gland size and sensitive endpoints
of thyroid hormone disruption commonly incorporated into toxicology studies Grim et al.
(2009); metrics used to report these endpoints are often variable. In most studies examining
this endpoint in fish, thyroid histology was affected monotonically.
Apparent NMDRs for histological changes were generally confounded by overt toxicity of the
test compound, poor study design, or failure to normalize changes in gland size to body weight.
For example, Patino et al. (2003), showed increased follicular cell hyperplasia and colloid
depletion at 18 ppm perchorate as compared to the controls, whereas 677 ppm perchlorate
had no effect. General toxicity and lethality was observed in the 677 ppm treatment, and
samples were taken at different times at the two dose levels (4 weeks earlier in high dose vs the
low dose group). Similarly, Crane et al. (2005) evaluated the effects of ammonium perchlorate
on fathead minnows and reported an apparent NMDR with follicular cell height where exposure
to 1 and 10 ppm were significantly larger than the controls, but exposure to 100 ppm resulted
in significantly reduced follicular cell heights compared to the 10 ppm exposure. However, this
study also observed significant reductions in growth at 10 and 100 ppm of about 40 and 60 %
respectively, indicating that the two higher exposure concentrations were at toxic
concentrations.
Finally, Schmidt and Braunbeck (2011) evaluated the effects of 2.5,10, 25, and 50 mg PTU/L
water on zebrafish and reported that histological effects on thyroid follicles and follicular cells
occurred in a concentration dependent manner, as did the reduction in whole body T4
concentrations. By contrast, morphometric analysis of the pituitary suggested an NMDR in
total pituitary area and adenohypohyseal area, which was highest at an intermediate PTU
treatment. However, the normalized area of the adenohypophysis, using the ratio of the
adenohypophysis: neurohypophysis, increased in a monotonic manner. Indexing of
morphometric data is commonly used to avoid scaling errors associated with differences in
body size, so it remains unclear if this observation is meaningful based on the un-indexed data.
80 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Circulating concentrations of T4 and T3, are also used as indicators of thyroid toxicity in aquatic
species but they are often impractical measurements due to the relatively small size of the test
organisms and thus limited blood volume. Some studies have evaluated thyroid hormones
using whole body extracts. However, whole body T4 measurements are largely inconclusive
and highly variable for technical reasons Crane et al. (2006; Park et al. (2006). Crane et al.
(2006) used both whole body and circulating T4 measurements to evaluate the effects of
methimazole on fathead minnows, depending on the size of the organism at the time of
sampling. Circulating thyroid hormone levels at 84 days were not statistically analyzed in the
females due to sample size problems, but there were no significant differences observed in the
circulating T4 and T3 measurements in the males. However, an apparent NMDR was observed
in the 28-day measurements of whole body T4 concentrations, where T4 was significantly
reduced at 32 and 100 ug/L but not at 320 ug/L methimazole. At 56 days, whole body T4 was
increased at 320 ug/L only. Whole body T3 was reduced at 320 ug/L at 28 days, but a slight
decrease was observed only at the 100 ug/L at 56 days. The authors suggest that the NMDRs in
whole body thyroid hormone measurements could be due to either compensatory mechanisms
or from altered thyroid hormone metabolism. No data are presented in support of either
suggestion. The inconsistent direction of change and the previously noted potential problems
with whole body thyroid hormone measurements contribute to uncertainty about these
observations. Furthermore, studies in zebrafish with PTU did observe monotonic reductions in
circulating T4 and T3 van der Ven et al. (2006) and whole body T4 concentrations Schmidt and
Braunbeck (2011).
Finally, two studies in fish evaluated thyroid hormone-dependent gene expression in response
to two thyroid disruptors, ammonium perchlorate and amitrole (Li, Zha et al., 2011)Li et al.
(2009). For the most part, gene expression exhibited montonic concentration-responses, with
one time point exhibiting an NMDR. Li, Zha et al, 2011 evaluated the effects of 5 and 50 ppb
perchlorate on the expression of genes for the iodothyronine deiodinase enzymes (Dll, DI2,
and DI3) and sodium iodide symporter (NIS) in the brain and liver of adult Chinese rare
minnows. They also measured DI2 and NIS in larvae. Expression of these genes was analyzed
following 7, 14, and 21 days of exposure. Circulating T4 and T3 were measured in adult plasma
at the end of the 21 day exposure. Changes in DI2 and NIS expression in the larvae were
monotonic within each sampling event, though the direction of change was inconsistent among
the different sampling times. Effects on hepatic expression of Dll, DI2, DI3, and NIS were
monotonic in males within each sampling event, though direction of change was again
inconsistent among the sampling times. DI3 and NIS expression in female livers were
monotonic at 7 and 21 days, but nonmonotonic at 14 days. Effects on brain expression of DI2,
DI3, and NIS were monotonic in females, with DI3 showing strong and consistent reductions at
most time points. The same three genes in males exhibited both monotonic and nonmonotonic
81 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
responses of varying magnitude and direction, depending upon the sample time. The gene
expression response patterns in this study are complicated, and the observation of apparent
nonmonotonic responses was dependent on when the samples were taken and were not
consistent throughout the duration of the study.
Li et al. (2009) evaluated the effects of 1 to 10,000 ng/L amitrole using juvenile Chinese rare
minnows exposed for 28 days. Expression of Dll, DI2, transthyretin (TTR), and thyroid hormone
receptor alpha (TRa) were measured in liver and brain. The only gene showing a possible
NMDR was that of hepatic DI2, which was significantly elevated in the low and intermediate
concentrations, but was above or at control levels in the highest concentrations. However,
hepatocyte degeneration was also observed at the high concentration, indicating that
cytotoxicity was occurring.
4.1.4.2 Amphibians
Numerous studies have been conducted with amphibians regarding thyroid disruption, as
amphibian metamorphosis is a well described thyroid hormone-dependent process. Many
studies have used single exposure concentrations and, therefore, were not considered in this
review. Recently, however, several multiple concentrations studies have been published.
These were largely executed in the development of a test guideline for regulatory purposes
OECD (2009; U.S. EPA (2009). Some studies are now appearing in the literature based on the
relatively new test guidelines. The most commonly studied species is Xenopus laevis.
The amphibian data reviewed here showed little evidence of nonmonotonic behavior in
developmental endpoints, thyroid histology, thyroid hormones, and gene expression. Potential
nonmonotonic responses were observed in histological endpoints of two studies. Opitz et al.
(2006) evaluated the developmental, histological, and molecular effects of ethylene thiourea
(ETU) exposure on X. laevis larvae using 1.0, 2.5, 10, 25, and 50 mg ETU/L for up to 90 days.
Metamorphic development and most of the morphological and histological endpoints changed
monotonically with the exception of follicular cell height, which was significantly decreased at
10 mg/L, but increased at 25 and 50 mg/L. This apparent NMDR was not interpreted further in
the Opitz et al. (2006) discussion. In any case, the prevalence of the remainder of the
histological observations, including follicular cell hypertrophy, increased monotonically with
ETU concentration. Indeed, five of the seven histological endpoints were affected at 10 mg/L.
Thus, despite an internal contradiction in the data, the overall histological profile indicates a
monotonic response.
Carlsson and Norrgren (2007) evaluated the effects of 2, 5,10, 20, and 75 mg/L 6-PTU on
development and thyroid gland histology of X. tropicalis larvae in a 14 day exposure. The
82 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
developmental and histological effects were observed to be monotonic, with the exception of
follicular lumen area, which increased through 10 mg/L but began to decrease at 20 mg/L. This
apparent NMDR is attributable to collapse of the follicle associated with the depletion of colloid
in the higher dose; an observation made in other studies of TH synthesis inhibitors. Reinforcing
this explanation is the fact that the follicular lumen areas at the 75 mg PTU /L concentration
were excluded from analysis due to severe follicular lumen collapse, which prevented accurate
measurements.
NMDR in fish and amphibians exposed to TPO and NIS inhibitors were infrequent. Where they
were observed, they were identified to be the result of toxicity at high doses, associated with
temporal sampling which may not have adequately represented the toxicodynamics of the
response (including compensation), resulting from highly variable and inconsistent
measurements, and/or normal observations associated with specific phenomena. In all cases
the apparent nonmonotonic responses were identified in isolated, single studies.
4.2 Mammalian Models
4.2.1 Literatu xh and Selection Strategy for E and A PathwafS
A large database of journal articles and other reports was examined resulting in the review of
numerous dose response curves for different EDC mechanisms of toxicity leading to disruption
of the estrogen orandrogen hormone system. Each article was evaluated to determine if an
NMDR occurred. The focus of this review was on in vivo EDC effects with NMDRs identifiable
from a broad range of doses, including the low dose range. The evaluation determined if the
observed NMDRs were consistent among similar studies on the same chemical, and identify if
the observed NMDR were robust and reproducible. The evaluation, when possible, attempted
to identify whether or not NMDR were clearly adverse effects or were upstream endpoint
causally linked to an adverse effect. Where possible there was an attempt to identify
uncertainties in the data base on EDCs with respect to NMDRs.
The general approach taken was to review the in vivo literature for well characterized low dose
effects reported for drugs, pesticides and toxic substances that disrupt the estrogen or
androgen signaling pathways. If these chemicals induced NMDR effects in the low dose range,
then it is possible that other chemicals displaying the same mechanisms of toxicity also would
induce NMDR.
A multipronged strategy was utilized to obtain references for this evaluation. A list was
produced that included chemicals that directly disrupt E or A signaling pathways at key points,
including AR or ER nuclear receptor binding or by altering steroid hormone synthesis. Articles
83 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
and documents evaluated included peer reviewed literature, US and international regulatory
agency risk assessments, as well as peer review committee and agency review articles and
summaries. A second approach evaluated key articles that were cited in the above documents
or guideline reports. Appendices A and B contain the review and evaluation of each individual
study.
The preference for inclusion was for studies with six or more dose groups by oral exposure to
the test chemical because these provide more robust information on the shape of the dose
response curve than do studies with fewer dosage levels. However, excluding studies with four
or five dose levels would have severely restricted the numbers of studies available for review
and would have also excluded many of the studies cited by others as displaying NMDRs
(Vandenberg et a!., 2012). Therefore, we also included in the evaluation in vivo studies of EDCs,
which encompassed a broad dose-range and have at least four or more groups (a control and
three treated groups).
In addition to the standard phenotypic endpoints, some genomic studies that utilized a
subcutaneous dosing regimen are included since they examined the effects of hormonally
active chemicals over a very broad dose response range with a comprehensive assessment of
EDC-induced altered gene expression in reproductive tissues. Studies using subcutaneous
dosing also are included when there were few, if any, oral studies with a specific estrogen or
androgen MoA.
The literature reviewed for the A pathway contains a very robust data base for some of the
chemicals. Some of the identified studies provided unique insights on how often one can
identify NMDRs, the conditions under which they occur, the potential relevance of NMDRs in
risk assessment and the general shape of the dose response curve. In addition to laboratory
animal studies, several dose-response studies have evaluated the beneficial and adverse effects
of androgens and selective androgen receptor modulators (SARMS) in both young and aged
men.
The studies were evaluated based on the robustness of the data available in the publication and
relative to other publications reporting on the same chemical.
More detail on study selection is provided on a chemical-by-chemical basis in Appendices A and
B. In total, more than 70 studies were reviewed, which had six or more dose groups and over
200 with four or more dose groups.
84 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
4.2.1.1 Specific Considerations in iewiewiing the Literature on E and A
Evaluation of the data was done to determine the robustness of the NMDR. An NMDR was
considered robust if it was reproducible and biologically plausible. In some cases, even though
the NMDR was not reproduced in additional studies because they had not been done, it was
still considered well supported base on its biological plausibility. There were a number of
situations in evaluating the studies that would suggest that the NMDR was an incidental finding,
was not biologically plausible, or while it might have been reported by the study authors was
not considered to be supported by the data by the authors of the current document. These
cases included the following: identification of the frequency of one effect declining as a more
severe one develops; a failure to correct for multiple comparisons or other inappropriate
statistical analyses; reporting of NMDR at dosage levels below background levels of the
xenobioticl. In some cases the study authors interpreted the data as an NMDR based on the
shape of the curve but in the absence of statistically significant differences from control. And
finally some NMDR were identified that were associated with overt toxicity or adverse effects
at lower doses. In this evaluation of E and A pathways we paid particular attention to the issues
in evaluating potential endocrine and other systemic effects in the overall WoE for the
observation of NMDR.
Several robust examples of NMDR were identified for adverse effects, but these were not
common. Exposure to androgenic chemicals has been shown to produce robust NMDRs in
utero and during adult life in female and male rats, respectively. For example, Robaire et a I.
(1979) found that administering increasing doses of testosterone to adult male rats resulted in
a biphasic response in testis weight and sperm production. This response also has been
demonstrated repeatedly in young adult male rats Dykman (1981: Robaire et al. (1979: Ewing et
al. (1977: Walsh and Swerdloff (1973). rabbits Ewing et al. (1973). and rhesus monkeys Ewing et
al. (1976). Similar NMDR effects on the testis and sperm production have not been reported for
human males; however, administration of high doses is precluded by the onset of adverse
effects, especially in aged men. Administration of testosterone to mature intact male rats,
rabbits, and rhesus monkeys causes a reduction in LH, followed by declines in testis androgen
levels, sperm production and testis weight without causing increases in serum testosterone or
androgen- dependent organ weights. However, as testosterone dosage levels are increased
above the nadir of the NMDR, testis weight and sperm production levels are partially restored
due to increasing levels of intratesticular testosterone from the serum (Figure 2, from Robaire
et al. (1979). redrawn below, Appendix B.6.b.l).
These are cases in which an effect displays an NMDR that peaks in the mid-dose range of a
study and then reverses direction at a higher dosage level as one effect is replaced by another
more biologically important effect. In the example below (Figure 4.6), glans penis weight
85 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
reaches a nadir in the mixture group exposed in utero to 50 % of the highest dose of dibutyl
phthalate (DBP) plus procymidone, whereas males exposed to the highest dose of the mixture
display organ weights greater than, not less than, control values (Appendix B.l.c.6). This results
from the fact that in the mid-dose group the glans was reduced in size but was not malformed;
by contrast almost all of the tissues from the high dose group were too malformed to weigh.
These kinds of NMDRs are not uncommon with chemicals that induce malformations or
histopathological alterations.
PROCYMIDONE PLUS DBP STUDY
GLANS PENIS WEIGHT - NMDRC
% TOP DOSE: MIXTURE GROUPS
Figure 4.6: Glans Penis Weight after exposure to a mixture of procymidone and dibutyl
phthalate.
Glans penis weight reaches a nadir in the mixture group exposed in utero to 50 % of the highest
dose of DBP plus procymidone whereas males exposed to the highest dose of the mixture
display organ weights greater than the control Hotchkiss et al. (2010).
There were occasions wherein an NMDR was identified by the study authors based on what
may be considered inappropriate statistical tests. This includes, for example, failure to account
for multiple comparisons by adjusting the p-value required for statistical significance in studies
with large numbers of endpoints and dosage levels. The NTP/NCTR five generation studies with
four dose levels of the estrogenic chemicals genistein (Appendix A.l.d.l) and EE2 (Appendix
A.l.a.l) included hundreds of endpoints (examination of both sexes, frequent measures of
growth and food consumption, multiple organ weights and histopathology, landmarks of
puberty, fertility, estrous cyclicity, etc). More dose levels of each chemical were included in the
one-generation dose range finding studies. Even if these chemicals had no endocrine or other
toxicity at the dosage levels used, one would reasonably expect to detect falsely as positive
some treatment related effects if one inappropriately compared each dose group to the control
86 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
value for every effect with a t-test, or an analysis of variance (ANOVA) accepting a p-value of
0.05 as a significant difference. In some of these cases the NMDR was not consistently
identified across experiments. In the NCTR/NTP EE2 dose range finding study NTP (2010), two
endpoints among all those measured in F1 males displayed NMDRs; age at preputial separation
(PPS) (Figure 4.6) and dorsolateral prostate weight (Figure 4.8). However when examined in
the F1 and F2 generations of the five generation study EE2 did not induce NMDRs in the same
dose range. Further increasing the lack of support for a true NMDR is lack of concordance with
other effects in the study and lack of consistency with other studies. Accelerated onset of PPS
at 25 ppb EE2 suggests that EE2 is androgenic or that it induces an earlier onset of increased
androgen levels. However, no other androgen-dependent tissue was "androgenized" at this
dose and these effects have not been seen in any other EE2 study or with other estrogenic
chemicals.
LU
O
QL
LU
0.
(lose R.iiMjfi y inilin¦! Study
MMDRfoi .nje aPieputi.il Separation
in F1 Mole SD R-ite
46
¦14
« 42
4)
< 40
38
liiii.il
V ft O ft ft ft
L JY cJO* <»¦
' V
EE2 ppb
DOSE RANGE FINDING STUDY
PERCENT DISPLAYING PPS
BY 50 DAYS OFAGE
100
80
GO
40
20
EE2ppb
Dose Romje Fimliixj Study
NMDRfoi «Kje
CL
Q.
(5
4>
44
42
4 0"
OlO I
rr*I
1
10
100
TTT¥1
1000
EE2 ppb
AGE AT PPS
FIVE GENERATION EE2 STUDY
FT
nun. i in
ill I I I 1 -> O I I I
.V ~ ~ ^ o 2 10 50
ft V ft ft ft ft 4
n* f)* ffy' <§v
EE2 ppb
Figure 4.7: Comparison of results for preputial separation in male rats after exposure to EE2.
An NMDR was identified for PPS in the dose range finding study, that dose range did not
accelerate PPS in similarly exposed males from the F1 and F2 multigenerational study. In the
87 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
multigenerational study, several endpoints (out of the hundreds measured over 4 generations)
displayed NMDRs; that is they were increased or decreased in a statistically significant fashion
versus control and an intermediate dose but not at a higher dosage level NTP (2010; Delclos et
al. (2009).
Dorsolateral Prostate Weight
DOSE RANGE FINDING STUDY
Dorsolateral Prostate Weight of
F1 and F2 males from the 5 Generation Study
EE2 ppb
p* 300-
^ 250-
O) 200-
"<5
ft
¦ Five Gen study F1 day 50
~ Five Gen study F1 day 90
¦ Five Gen study F2 day 50
~ Five Gen study F2 day 90
OOOO C5 C5 C5 C5 O O O O
EE2ppb
Figure 4.8: Comparison of results for dorsolateral prostate weight in male rats after exposure
to EE2.
An NMDR was identified for dorsolateral prostate weight in the dose range finding study but
not in similarly exposed males from the F1 and F2 multigenerational study NTP (2010; Delclos
et al. (2009).
Another finding that raised uncertainty in the identification of a NMDR was statistical analyses
in studies using multiple pups from the same litter or repeated measures on the same endpoint
on the same pup at different ages. This error can inflate the error degrees of freedom and
underestimate the error mean square, resulting in inflated significant F and t-values. An
example would be of a small study wherein3 pregnant rats per dose group are treated with
DEHP at four dose levels (control and three treated groups. If F1 serum testosterone is
recorded for 3 pups/litter repeatedly (5 observations/Fl male offspring over a lifetime) an
ANOVA analysis that fails to account for litter effects and repeated measures would have an F
value for a main effect of DEHP on serum testosterone levels based upon 176 degrees of
freedom in error; this is by contrast to 11 degrees of freedom in error given corrections got
litter effects and repeated measures.
An additional confounding variable occurs for certain xenobiotics that happen to be ubiquitous
in the environment, and thus routinely contaminate feed and bedding in rodent studies.
Several of the chemicals discussed in this current evaluation are ubiquitous in the environment
and are found in a majority of environmental and human urine samples. Some are present in
88 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
human diets and, not surprisingly, they also are found in animal diets; this can be at
concentrations orders of magnitude above those used by investigators in what is suggested to
be a low dose study. Some study design and analysis and interpretation of the data have not
taken those background levels into consideration. For example, several publications have
reported that phthalates are found in animal's diets at low ppm concentrations and are
reported at even higher concentrations in the litter used for cage bedding. One study reported
that 100% of the animal diets and beddings contained DBP, BBP and DEHP and another
reported that these exposures produced measurable phthalate metabolite levels in the tissues
(Kondo et al. (2010) (Appendix B.2.a). In the case of the phthalates, the validity of the results
from studies reporting NMDRs at dose levels below the reported levels in diets and beddings
would likely be of questionable value.
In figure 4.9, the administered levels of DEHP fed to pregnant mice in this particular study falls
well within the range of phthalate concentrations identified in a survey of control animal diets
(Appendix B.2.e.2). Failure to control for this dietary confounder raises uncertainty about the
actual exposure levels in the study. Thus, there would be uncertainty about the NMDR
conclusion as the F-value for a treatment effect was not significant, and the reported effect is a
small increase in anogenital distance (AGD), whereas the phthalate syndrome phenotype is
characterized by shortened AGD in newborn male rats. Do et al., (2012) (Appendix B.2.e.2)
orally administered DEHP from GD 9 to 18 at 0, 0.5, 1.0, 5, 500, 50,000, and 500,000 M-g/kg/d
and examined maternal and fetal (1M position males only) hormones on GD 18. In figure 4.6,
the range of the estimated daily phthalate intake from the control animal's diets is shown in red
(estimate from Kondo etal., 2010, Appendix B.2.a). The fraction of total phthalate exposure
contributed by the diet and unaccounted for in the study could be considered substantial and
raises questions regarding an NMDR in such a study.
Animal diets may also contain heavy metals, mycotoxins, phytoestrogens, dioxins and PCBs and
other contaminants at varying levels
(www.fda.gov/AnimalVeterinary/Products/AnimalFoodFeeds/Contaminants/default.htm).
These additional contaminants may interact with the test material.
89 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Possible Dietary Phthalate levels
DEHP micrograms per kg per day
Figure 4.9: Representation of possible background dietary phthalate concentration.
Possible dietary phthalate levels (red box) based on calculation from Kondo et al (2010) related
to doses as provided to mice in Do et al. (2012).
In some cases NMDR have been reported when no treated group is statistically significantly
different from control, and observation of NMDR appears to be based on inspection of the
data. For example, Vandenberg et al. (2012) reported that DEHP data from Grande et al. (2006)
displayed NMDRs for the ages at vaginal opening (VO) and first estrus in female rats (Appendix
B.2.a.3). In contrast, Grande et al. (2006) reported "A significant delay in the age at vaginal
opening (approximately 2 days) at 15 mg DEHP/kg bw/day and above, as well as a trend for a
delay in the age at first estrus at 135 and 405 mg DEHP/kg bw/day"; the study authors did not
report NMDRs (see figure 4.10).
90 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
AGE AT VAGINAL OPENING
* INDICATES STATISTICALLY SIGNIFICANT
EFFECTS NOTED BY AUTHOR
o
>
•4-I
CO
0)
O)
<
40-|
A A A A A ,.<5 ,.<5 ,.<5
ft* ft* ft* ft* ft* \*" V ,*>•
Dose DEHP mg/kg/d
(/)
3
s_
+->
LU
*•>
(/)
CO
0)
O)
<
AGE AT FIRST ESTRUS
NO STATISTICALLY SIGNIFICANT
EFFECTS NOTED BY AUTHOR
A A A A A A A -Si
ft' ft* ft* ft* ft* N* <3* ,{5* ^3*^*^3*
Dose DEHP mg/kg/d
Figure 4.10: Histograms illustrating effects on vaginal opening and age at first estrus in DEHP
exposed rats.
Grande et al. (2006). The histogram of age at first estrus has the appearance of indicating an
NMDR although none of the doses are statistically different from control. Yellow box indicates
calculated background phthalate exposures from feed based on Kondo et al (2010).
Identification of a NMDR after high dose exposure associated with overt toxicity or adverse
effects that occur at doses below the NMDR introduces uncertainty. Chemicals that induce
overt toxicity also may indirectly disrupt the endocrine system and secondarily induce
reproductive and endocrine alterations. For example, Vandenberg etal. (2012) listed
semicarbazide (SEM) as an EDC that disrupted male rat pubertal development in an NMDR
fashion (Appendix B.4.C). SEM acts as an osteolathyrogen, and induces osteochondral and
vascular lesions in young rats due to impaired cross-linking reactions of collagen and elastin
(Appendix B.4.b). In addition, teratogenic effects such as induction of cleft palate and aortic
aneurysms also have been reported (Appendix B.4.a). However, the literature provides little
support for this chemical as an EDC; moreover the identified NMDR is observed at a dose level
that produces a severe reduction in weight gain during dosing and is higher than doses of SEM
that produces lesions of the joints and vascular system.
In summary, in this evaluation of E and A pathways we paid particular attention to the above
issues in evaluating potential endocrine and other systemic effects in the overall WOE for the
observation of NMDR.
91 | Page
-------
Draft Deliberative
4.2.2
State of the Science: Nonmonotonic Dose Responses V7
Estrogen Hormone System
In addition to natural steroidal estrogens, many chemicals including plant and fungal estrogens,
pharmaceuticals in the environment, toxic substances and pesticides also bind the estrogen
receptors of mammals. Other than some of the estrogenic pharmaceuticals the xenoestrogens
bind ERs with affinities orders of magnitude lower than naturally occurring 17p estradiol (E2).
For this reason relatively high exposure levels are required to produce estrogenic effects for
many xenoestrogens. For example, oral administration of E2 and ethinyl estradiol (EE2) induce
increases in uterine weight at low M-g/kg dose levels whereas bisphenol A (BPA), bisphenol AF
and methoxychlor require doses four to five orders of magnitude higher to induce a
uterotrophic response of similar magnitude. In some cases the endocrine activity may be
displayed only at doses equal to or above the induction of some other systemic toxicity; for
example, with the tremorogenic pesticide chlordecone (Kepone), estrogenicity is observed only
at doses equal to or above those inducing some other form of systemic toxicity. The interaction
of a chemical with ER or induction of an estrogen-dependent response does not enable one to
predict with certainty what other estrogenic effects may occur, the dose that will produce an
effect, or the shape of the dose response curve in each tissue. Tissue-specific responses arise
from differences in these factors: receptor levels, levels of coactivators and corepressors, E2
metabolism, receptor stability, different target gene estrogen response elements, gene
silencing, and other factors. The studies that identified NMDR are described below with
additional evaluations of a broader collection of chemicals and studies presented in Appendix
A.
92 | Page
-------
Draft Deliberative - Do NOT Cite, Quote, or Distribute
State of the Science: Nonmonotonic Dose Responses V7
Table 4.2: NMDR From Studies Evaluating the Mammalian Estrogen Hormone System.
Chemical
Species
Group Size
Doses Given
Dosing
route
Reference(s)
Appendix
location
Ethinyl estradiol
rat
5
0.1, 1.5, 5, 25, 100 or 200 ppb
feed
NTP (2007)
A.l.a.l
Ethinyl estradiol
rat
?
2, 10, or 50 ppb
feed
Delclos et al. (2009)
A.l.a.l
Ethinyl estradiol
rat
19-43 dams
0.05, 0.15, 0.5, 1.5, 5, 15, or 50 |J.g/kg
gavage
Howdeshell etal. (2008)
A.l.a.4
Ethinyl estradiol
mouse
10-12 dams
0.002, 0.02, 0.2, 2, 20, 200 jag/kg
gavage
Thayer et al. (2001)
A.l.a.5
Ethinyl estradiol
rat
10, 5 per
block
0.01, 0.05, 0.2 mg/kg
gavage
Andrews et al. (2002)
A.l.a.9
17(3 Estradiol
mouse
10
0.005, 0.05, 0.5, 2.5, 5, 10, or 50 ppm
feed
Tvl etal. (2008a)
A.l.b.2
17(3 Estradiol
mouse
25
0.001, 0.005, 0.05, 0.15, or 0.5 ppm
feed
Tvl etal. (2008c)
A.l.b.3
Diethylstil bestrol
mouse
?
0.002, 0.02, 0.2, 2, 20, and 200 ng/g
gavage
vom Saal etal. (1997)
A.l.c.l
Diethylstil bestrol
mouse
10
1, 5, 10, or 15)J,g/kg
gavage
Nagao et al. (2012)
A.I.e.2
Diethylstil bestrol
rat
10
0.2, 1.0, or 5.0 )J.g/kg
gavage
Kim et al. (2002b)
A.I.e.6
Diethylstil bestrol
rat
10
10, 20, or 40 J-tg/kg
gavage
Shin et al. (2009)
A.I.e.7
Genistein
rat
10 dams
5, 25, 100, 250, 625, or 1250 ppm
feed
Delclos et al. (2001)
A.l.d.l
Genistein
rat
10
5, 25, 100, 250, 625, or 1250 ppm
feed
NTP Tox Report (2007, 08)
A.l.d.l
Genistein
rat
9 dams
5, 50, 500, 1000 ppm
feed
Akingbemi et al. (2007)
A.l.d.6
Genistein
rat
~17
5, 100, 500 ppm
feed
Dalu etal. (2002)
A.l.d.2
Nonylphenol
rat
25
2, 10, or 50 mg/kg
gavage
Nagao et al. (2001)
A.1.H.3
Bisphenol A
rat
28
0.01, 0.1, 1.0, 10 ppm
water
Cagen et al. (1999b)
A.l.k.2
Bisphenol A
rat
20 dams
0.015, 0.3, 4.5, 75, 750, or 7500 ppm
feed
Tvl etal. (2002)
A.l.k.3
Bisphenol A
mouse
28
0.018, 0.18, 1.8, 30, 300, or 3500
ppm
feed
Tvl etal. (2008b)
A.l.k.4
Bisphenol A
rat
25
0.2, 2, 20, or 200 ng/kg
gavage
Ema et al. (2001)
A.l.k.5
Bisphenol A
rat
10
0.33, 3.3, or 33 ppm
feed
Kobavashi etal. (2012)
A.l.k.7
Bisphenol A
rat
24
0.15, 1.5, 75, 750, or 2250 ppm
feed
Stump et al. (2010)
A.l.k.8
Tamoxifen
rat
10
10, 50 or 200 ng/kg
gavage
Kim et al. (2002b)
A.2.b.l
Tamoxifen
rat
10
5, 30, or 200 ng/kg
gavage
Kennel et al. (2003)
A.2.b.2
Exemestane
rat
8 or 30
2, 4, 5, 10, 20, 40, 50, 90, 100, 125,
200, 250, 500, 810, 1000 mg/kg
gavage
Beltrame etal. (2001)
A.3.b.l
93 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
4.2.2.1 Ethinyl Estradiol (EE2) [A.l.a]1
There are several robust, well designed, comprehensive, low to high dose, multigenerational
and transgenerational EE2 studies conducted with rats. In addition, there are a number of
shorter-term, mechanistic studies that examined the effects of EE2 over a broad dose range
that included upstream endpoints. The literature on the estrogen EE2 provides a rich data base
to address questions about dose response, sensitivity of endpoints and critical life stages for
induction of adverse effects of estrogenic chemicals. The NTP executed a series of studies with
EE2 and other well characterized estrogenic chemicals, specifically to examine the shape of the
dose response curve and to identify endpoints sensitive to estrogens for potential inclusion in
multigenerational studies of chemicals displaying estrogenicity in screening assays.
The NTP executed a series of studies with EE2 and other well characterized estrogenic
chemicals specifically to examine the shape of the dose response curve and to identify
endpoints sensitive to estrogens for potential inclusion in multigenerational studies of
chemicals displaying estrogenicity in screening assays Delclos et al. (2009; Latendresse et al.
(2009) [A.l.a.l]. These studies included measurement of fertility, fecundity, maternal and litter
measures, repeated observations of growth and food consumption throughout the life cycle,
AGDat birth, neonatal developmental landmarks, pubertal landmarks, estrous cyclicity, organ
weights, histopathology, ovarian oocyte counts and sperm counts over 5 generations.
The first study was a dose range finding none generation study with 7 dose levels (0, 0.1,1.5, 5,
25, 100 or 200 ppb) in the feed. Exposure began on GD 7 and continued through gestation,
lactation and directly to F1 in the diet until necropsy at 50 days of age. Dietary exposure of the
dams continued through lactation. Pups from five litters, culled to eight per litter with an equal
sex distribution on postnatal day (PND) 2 were maintained on the same dosed feed as their
mother after weaning until sacrifice at PND 50. Maternal body weight gain was reduced during
pregnancy by about 50% in the 200 ppb dose group and by 30% at 100 ppb and mean pup
weight at birth also was reduced in these two dose groups. The percentage of F1 males
displaying PPS at 50 days of age was dramatically reduced from 85% in controls to 20% in the
200 ppb dose group. The age at PPS displayed an NMDR, being accelerated by about 1.6 days
at 5 and 25 ppb. In F1 males, terminal body weight, ventral prostate weight, testis sperm count
and testes weights were reduced at 200 ppb, whereas the dorsolateral prostate weight was
identified as a NMDR with a significant increase only at 5 ppm. Histopathological alterations
were noted in reproductive and nonreproductive organs at 100 and 200 ppb and mammary
gland ductal hyperplasia was present at doses of 25 ppb and higher dose levels, all monotonic
responses Latendresse et al. (2009) [A.l.a.l]. In F1 females, VO was accelerated in the 25, 100,
1 The bracketed alpha-numeric identifier corresponds with the specific location in the appendix.
94 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
and 200 ppb groups and body and ovarian weights were reduced at 200 ppb and ovarian,
uterine and vaginal tissues displayed histological abnormalities at 200 ppb; however no NMDRs
were identified.
The data from the dose range finding study was used to select dosage levels for a 5 generation
study (n>30 mated pairs per group) wherein 0, 2, 10, 50 ppb was fed in the diet. In the
multigenerational study, vaginal opening was accelerated at 50 ppb in the Fl, F2 and F3
generations, but not in the F4 generation (unexposed) Delclos et al. (2009) [A.l.a.l]). Ethinyl
estradiol accelerated the attainment of puberty in females provided 50 ppb under continuous
exposure conditions (Fi and F2) and in the F3 where dosing was terminated at weaning.
Perturbation of the estrous cycle (prolonged cycles, aberrant cycles, time in estrus) in young
females after VO and prior to mating was observed in the F^ and F2 generations. In males,
statistically significant inductions of male mammary gland hyperplasia (Fg through F3
generations) and mild mineralization of renal tubules (F1 and generations) were observed.
The majority of these effects in male and female rats were observed at 50 ppb.
Sporadic NMDRs that did not repeat in similarly exposed groups of rats from different
generations were identified in this very large and comprehensive study of a model estrogenic
chemical. Although PPS displayed an NMDR in the dose range finding study, that dose range
did not accelerate PPS in similarly exposed males from the Fl and F2 multigenerational study.
In the multigenerational study, several endpoints (out of the hundreds measured over 4
generations) displayed NMDRs being increased or decreased significantly versus control and
an intermediate dose but not at a higher dosage level. None of these effects were consistent
across similarly exposed generations. Although dorsolateral prostate weight displayed an
NMDR in the dose range finding study, that dose range did not increase the weight of this
tissue in similarly exposed males from the Fl and F2 multigenerational study. In addition, no
histopathological lesions were detected in this tissue.
Howdeshell et al. (2008)[A.l.a.41 studied the effects of gestational and lactational oral gavage
exposure to 0.05, 0.15, 0.5, 1.5, 5, 15, or 50 0g EE2/kg on the postnatal development and
reproductive function of the male rats. The numbers of live pups at weaning was increased at
0.15 and 1.5 M-g/kg/d but not at higher dosage levels. Increased survival is not typically
considered an adverse effect.
Thayer etal. (2001)[A.l.a.51 treated pregnant mice with 0.002, 0.02, 0.2, 2, 20, or 200 0g
EE2/kg by gavage and found body weight at necropsy and AR per prostate in male offspring,
exhibited NMDRs following exposure.
Administration of 0.01, 0.05, 0.2 mg EE2 /kg by gavage to adult female and male rats (OECD
Test Guideline no. 407) produced adverse effects at relatively high dosage levels as compared
95 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
to dosage levels producing effects in developing rats. Thyroid stimulating hormone (TSH) levels
displayed an NMDR in the first block, being statistically significantly increased at 50 |ag/kg/d,
but not at a higher dose. However, this effect was not seen in the second block Andrews et al.
(2002) [A. 1.a.9].
4.2.; radiol (E2) [A.l.b]
Mice were exposed to 0.005, 0.05, 0.5, 2.5, 5, 10, or 50 ppm E2 in feed during a 2-week pre-
breeding period, mating, gestation and lactation in a one generation study Tyl et al.
[2008a)[A.l.b.2]. Of the 100 or so measurements (with multiple contrasts for each versus
control values) only seminal vesicle weight in the P0 generation displayed an apparent NMDR,
being increased versus the control group at 0.005 and decreased at 5 ppm and above. In a
companion two generation study, mice were fed 0.001, 0.005, 0.05, 0.15, or 0.5 ppm E2 during
an 8 week pre-breeding period, mating, gestation and lactation Tyl et al. (2008c)[A.l.b.31. The
NMDR on P0 seminal vesicle weight in the one generation study at 0.005 ppm was not a
statistically significant effect in the latter study. In the latter study, thyroid weight was the only
endpoint among more than 60 to display an NMDR, being decreased only at 0.005 ppm;
however, this effect was not seen in the former study. In summary, 2 of 160 endpoints
displayed apparent NMDRs in these two E2 studies with mice; and neither effect was
replicated.
4.2.2.3 Diethylstilbestrol (DES) [A.l.cJ
DES is a recognized developmental toxicant in humans. It was synthesized in 1938 for
treatment of menopausal symptoms in women. Studies that followed shortly thereafter
identified that in utero administration of DES produced uterine malformations in female rat
offspring and that long-term treatment was carcinogenic. DES is still an environmental
estrogen of concern, as it is used today in aquaculture in some countries, in order to induce sex
reversals and control growth in some fish species.
vom Saal etal. (1997)[A.l.c.l) treated pregnant mice with 0.002, 0.02, 0.2, 2, 20, and 200 ng
DES/g/d by gavage and reported increased prostate weights versus controls at 0.02 and 2
ng/g/d with a statistically significant decline at 200 ng/g/d in mice following prenatal exposures.
In similar studies by Ashby etal. (1999)[A.l.c.l1 and Cagen etal. (1999a)[A.l.k.21 neither
identified the prostate weight change in mice exposed to 0.2 ng DES/g, which was the only dose
of DES provided in these two studies as it was used as a concurrent positive control.
Nagao et al. (2012)[A.l.c.21 examined placental function and gene expression after oral
treatment with 1, 5, 10, or 15 |ag DES/kg by gavage to pregnant mice on GD 4 to 8. All effects
96 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
displayed monotonic dose responses except expression of ERa mRNA in the placentas of male
(but not female) embryos; ERa mRNA was expressed only in male embryos of dams exposed to
DES at 5 M-g/kg/d and in no other dose group. ER|3, ERR|3 and ERRX mRNA were not statistically
significantly affected at any dosage level.
Kim et al. (2002b)[A.l.c.61 exposed 21 to 41 day old female rats to 0.2, 1.0, or 5.0 |ag DES/kg by
oral gavage and NMDRs were identified for liver and kidney weights which were decreased only
at 1 |ag/kg/d. Other reported effects demonstrated monotonic dose responses.
Shin etal. (2009)[A.l.c.71 studied the effects from oral gavage administration of 10, 20, or 40 |ag
DES/kg on pubertal development in male rats, exposed from PND 33 to 53. Among the
endpoints measured, liver weight showed an NMDR, increasing at 10 and 20 but declining at 40
M-g/kg/d concurrent with a 20% reduction in terminal body weight. In other studies of perinatal
exposure of rats or mice to DES Ashby et al. (1999)[A.l.c.l1; Cagen et al. (1999a, b)[A.l.k.2];
Odum et al. (2002)[A.l.c.51; Nagao and Yoshimura (2009)[A.l.c.21; Ohta et al. (2012)[A.l.c.41),
no effects displayed NMDRs.
4.2.2.4 Geniste .d]
Genistein is one of several phytoestrogens found at high levels in many animal and human
diets. Thus, a series of studies were conducted by the NTP. Rats were administered 5, 25,100,
250, 625, or 1250 ppm genistein in the diet, which spanned the levels found in human diets. Of
the >60 individual dose response curves generated in the dose range finding study one showed
an apparent NMDR; that was for the effects of genistein on the righting reflex in F1 female (but
not male) pups Delclos et al. (2001)[A.l.d.ll). Three different dose levels, 50, 100, and 500
ppm, were administered in the multigeneration study NTP (2008)[A.l.d.l1). This study resulted
in more than 200 individual dose response curves of which a few were NMDRs. The age at VO
was accelerated in an NMDR fashion at 5 ppm in females from the F3 generation, but not in any
other generation. The percent time spent in different stages of the estrous cycle also displayed
NMDRs in some cases; this effect was not consistent from generation to generation. Female
adrenal weights were reduced in only the 5 ppm group of the F2 generation (but not F0, F1 or
F3 generations); pituitary gland weight was increased only in the 100 ppm group of the F0
generation (but not Fl, F2 or F3 generations); spleen weight was increased in the 5 ppm group
of the Fl generation (but not F0, F2 or F3); and thymus weight was decreased in the 100 ppm
group of the F3 generation (but not F0, Fl or F2). None of these NMDRs were consistently
expressed across generations, sexes, or doses. The NMDR of genistein reported by Akingbemi
et al. (2007 (A.l.d.6]) on Fl male AGD, body weights at weaning and testis weights, were not
replicated in this study in neonatal males from any of the filial generations that were exposed
to 5, 50, 500, 1000 ppm genistein in utero and during lactation.
97 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Three NTP studies NTP (2007)[A.l.d.l1) administered genistein in the diet of rats from the time
of conception, throughout pregnancy (0, 0.5, 9, or 45 mg/kg), and through lactation (0.7,15, or
75 mg/kg). One study continued exposures for 20 weeks post birth (0.4, 8, or 44 mg/kg to
females and 0.4, 7, or 37 mg/kg to males) and another one continued exposures for 2 years
(0.3, 5, or 29 mg/kg to females and 0.2, 4, or 20 mg/kg to males). There were more than 200
dose responses that were evaluated; 40 tissues were examined histologically and many tissues
were weighed in the 3 studies. Effects exhibiting NMDRs were observed, however, the same
response was not seen in both sexes or in more than one of the three cohorts. The NMDR
identified were for these effects: reduced incidence of mammary gland fibroadenomas in the 5
ppm group in one of 3 cohorts; reduced incidence of pituitary adenomas in male rats in the 5
ppm group in one of 3 cohorts; increased incidence of preputial gland squamous cell carcinoma
in male rats in one of 3 cohorts at 100 ppm; reduction in benign neoplasms in all organs in the
100 ppm group in one of 3 cohorts; reduced incidence of uterine stromal polyps in 1 of 3
cohorts in the 5 ppm group; variations in body weight; increased brain weight in females in 1 of
3 cohorts in the 100 ppm group; increased pituitary weight in females in 1 of 3 cohorts in the
100 ppm group; and increased spleen weight in females in 1 of 3 cohorts in the 5 ppm group.
Some NMDR were observed in effects that might not be considered adverse, such as reduced
tumor incidence.
An NMDR was identified from a two-generation reproduction study in rats given 5, 100, 500
ppm genistein in feed (Dalu et al. (2002)[A.l.d.21). This was for ERa in the dorsolateral prostate
at 100 ppm (F1 but not F2) in the cohort exposed through their mothers and then directly to
genistein (G/G) but not in the cohort switched to a control diet at the time of weaning (G/C).
An NMDR was also identified for ER|3 protein levels in the ventral prostate of the F1 but not the
F2 at 100 ppm in the G/C but not G/G cohorts. The NMDR for serum testosterone, which
Akingbemi et al. (2007)[A.l.d.6D reported as increased at 5 ppm but not at higher dosage levels
in rats given 5, 50, 500, 1000 ppm genistein in the feed was not replicated in the Dalu et al.
[2002)[A.l.d.2] study.
The study by Akingbemi et al. (2007)[A.l.d.6D reported a number of findings that were not
seen in other studies of genistein at similar exposure levels. Akingbemi et al. (2007) did not
administer genistein, but rather they fed pregnant LE rats casein-based diets containing whole
soybean as sources of protein with isoflavone concentrations of 0, 5, 50, 500, or 1000 ppm from
GD 12 to PND21. They measured serum concentrations of free and conjugated insoflavones in
male rats and dams at 21 d postpartum. NMDRs were identified for increased body weights
and longer AGD in male rats at postnatal day (PND) 5 (at 5 and 50 ppm but not higher). In
addition, serum testosterone levels and in vitro Leydig cell testosterone production were
increased only at 5 ppm in prepubertal males, and serum luteinizing hormone (LH) was
98 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
increased only at 50 ppm in adult male offspring (fed the same diet as the dam until PND 90).
They noted that testis weight was reduced at all dose levels, which has not been reported other
studies of genistein.
4.2.2.5 Bisphenol A [A.l.kJ
The selection criteria for studies reviewed herein on BPA used the approach described in the
scientific reviews by governmental and regulatory agencies. The studies included for discussion
in this document were selected from those identified by these groups as being "adequate for
evaluation" and of "high utility" in the respective evaluations Chapin et al. (2008). The current
state of the science evaluation does not revisit conclusions on hazard characterization for BPA.
Rather we review the shapes of the dose response curves only from adequate, high utility
studies that used oral dosing, with a broad range of dosage levels (at least 3 treated levels and a
negative control). In most perinatal and adult studies in rats and mice exposed to BPA (Kim et
al. (2002a)[a.l.k.ll1; Tinwell et al. (2002)[A.l.k.l1; Ashby et al. (2003)[A.l.k.l31; Chitra et al.
(2003b); Chitra et al. (2003a)[A.l.k.l41; Yamasaki et al. (2003)[A.l.k.l21; Howdeshell et al.
(2008)[A.l.k.91; Kobayashi et al. (2010)[A.l.k.61; Ryan et al. (2010a); Ryan et al.
(2010b)[A.l.k.l01) hundreds of endpoints that were measured but did not display NMDR. The
specific examples of NMDR are summarized below.
Cagen et al. (1999b)[A.l.k.21 conducted a study in rats given 0.01, 0.1, 1.0, 10 ppm in water to
examine the effects of prenatal and lactational BPA exposure on reproductive development of
rats. Of all of the endpoints examined only male sex ratio was affected in an NMDR manner,
being increased in the 0.1 ppm group (56.7% males versus 48.4% in control) but not at any
other dose level. However, the same effect was not seen in other stud ies of BPA in rats in this
dose range.
Tyl et al. (2002)[A.l.k.31 performed a multigenerational study of BPA in Sprague Dawley rats
given 0.015, 0.3, 4.5, 75, 750, or 7500 ppm BPA in the feed. NMDRs were identified for several
measures across the study. There were no reproducible NMDRs across generations in the
current study on any endpoint. Of the 52 dose response comparisons for body and organ
weights in this set of studies, relatively few had identifiable NMDR. Male rats had NMDR for
liver, kidney, testes, and seminal vesicle with coagulating gland weights seemingly randomly
across the generations. Female rats had NMDR for liver, ovary, and uterus weight across the
generations as well. There did not appear to be any consistency across gender; for example,
the liver weight NMDR inflection point was in the 0.3 ppm group in the F2 males whereas it was
the 0.015 ppm group in the F2 females. Of the 11 reproductive parameters evaluated in male
and female rats across the 4 generations, two NMDR were identified: the number of
implantation sites and the total pups per liter both in the 0.3 ppm group of the F3 generation.
99 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
NMDR were identified for only 3 of the 17 reproductive development parameters reported in
this paper: the number of live pups/liter in the F3 generation exposed to 0.3 ppm; AGD in
females across the 0.3-170 ppm groups in the F2 generation; and the age of PPS in males across
the 0.3-75 ppm groups in the F2 generation. Given the large number of statistical comparisons,
it could be suggested that the number of NMDRs identified in the study does not exceed what
could occur by chance variation.
BPA was administered in the diet of mice at 0.018, 0.18, 1.8, 30, 300, or 3500 ppm in a two
generation reproductive toxicity study Tyl et al. (2008b)[A.l.k.41). The NMDRs identified for
seminal vesicle weight, 21 day survival, and thymus weight were not consistent across the
generations or cohorts.
Ema etal. (2001)[A.l.k.51 conducted a multigenerational reproductive toxicity study
administering 0.2, 2, 20, or 200 jag BPA/kg via gavage to rats. Numerous measurements were
reported across the 3 generations evaluated in this study. Several NMDR were identified. Of
the 50 separate measurements reported for AGD in males and females (different ages across
two generations) occasional NMDR responses were identified; however, NMDR seen in the F1
were not replicated in the F2. NMDR for negative geotaxis was identified in F2 males (but not
F2 females or F1 males or females); NMDR related to accelerated mid-air righting reflex was
identified for F1 males (but not F1 females or F2 males or females). Of the approximately 60
organ weights taken at necropsy of F1 and F2 weanlings, two NMDR were identified in males
but not females. Of the approximately 24 sperm measures taken in F1 and F0 males, an NMDR
was identified for one; that is, a decrease in abnormal sperm from 1.6% in controls to 0.6% in a
mid-dose group in F1 but not F0 adult males. In addition, an NMDR was identified in F1 adult
males of a reduction in testis weight. However, there were no consistent NMDR across
generations or genders.
Kobayashi etal. (2012)[A.l.k.71 treated pregnant SD rats with 0.33, 3.3, or 33 ppm BPA in the
diet from GD6 to PND21. F1 pups were not exposed to BPA directly after weaning and were
necropsied at 5 weeks or 3 months of age. An NMDR was noted for increased body weights of
F1 males, but not females, at 0.33 ppm at 12 and 13 weeks of age (2 of 11 measurements of
body weight) and in F1 females at 6 and 11 weeks of age at 33 and 0.33 ppm, respectively.
Stump etal. (2010)[A.l.k.81 reported on the potential of BPA to induce functional and/or
morphological effects to the nervous system of F1 offspring from dietary exposure of 0.15,1.5,
75, 750, or 2250 ppm during gestation and lactation. In that study NMDR were identified for
some endpoints. Body weights were measured more than 15 times in P0 males and females,
and NMDRs were identified at 5 preweaning ages in the 75 ppm group in males but not in P0
females. These effects did not persist after weaning. Weight gain was recorded for both sexes
100 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
for four time periods. In one of the eight measures of F1 weight gain during lactation male, but
not female, weight gain was increased at 0.15 and 75 ppm. There is an indication of an NMDR
in the Biel Water Maze as a low-dose effect in males, but not female F1 rats. The
interpretability of some of the results in this study have been questioned. A Euopean Food
Safety Assessment Panel concluded that "the study by Stump et al. (2010) cannot be used for
the assessment of the effects of BPA on learning and memory due to methodological
limitations. A number of studies addressing other neurobehavioural endpoints (e.g., learning
and memory behaviour, anxiety-related behaviour and gender-specific behaviour) were
considered invalid or inadequate for risk assessment purposes by the Panel. The Panel does
not consider the currently available data as convincing evidence of neurobehavioural toxicity of
BPA" EFSA (2010)[A.l.k.81.
4.2.1 sctive Estrogen Receptor Modulators (SERMs) [A.2]
SERMS are synthetic drugs used as an option for the treatment of breast cancer, and
osteoporosis in women, and for other conditions arising from adverse effects of estrogens or
their lack. In this regard, SERMs have been developed and studied in vitro using cell lines from
different estrogen-responsive tissues. They have also been studied in rat models
(orchidectomized males and female rats) for their ability to have selective, beneficial effects on
tissues like bone, without inducing potentially adverse effects on other tissues (for example,
breast and prostate in females and males, respectively). These chemicals can act as estrogens
in some tissues and as antiestrogens in others. The first SERM approved for pharmaceutical use
was tamoxifen, but due to its stimulatory effect on the endometrium, which in some patients
was determined to be a negative response, it is currently not indicated for osteoporosis.
4.2.1 loxifen [A.,
Kim et al. (2002b)[A.2.b.l1) performed the USEPA guideline EDSP pubertal female assay with
several EDCs including tamoxifen given by gavage at 10, 50 or 200 |J-g/kg. NMDRs were
identified for thyroid gland weight, thymus weight, estradiol, TSH, and T3. These effects were
related to pituitary-thyroid function (thyroid weight, serum TSH, and serum T3>and ovarian
estradiol production. Kennel et al. (2003)[A.2.b.21 administered 5, 30, or 200 (j,g tamoxifen/kg
by oral gavage for 28 days to young adult male and female rats in an OECD 407 assay. Liver
weight displayed an NMDR in one block, being increased at 5 |ag/kg/day, but this was not
repeated in the other block. In addition, uterine weight (with cervix) displayed an NMDR, being
heavier after 5 |ag/kg/day and lighter at 30 and 200 |ag/kg/day (all groups differed statistically
significantly from control). However, a one generation reproduction study in rats treated with
0.12, 0.6, or 3 |j,g tamoxifen/kg by gavage did not demonstrate any NMDR for any endpoints
measured Yamasaki et al. (2005)[A.2.bl.31.
101 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
4.2.2.8 Aromatase Inhibitors (Al: block androgen to estrogen synthesis)
[A.3]
Aromatase is a cytochrome P450 enzyme that converts androgens to estrogens by aromatizing
the A ring of the steroid molecule. It is found in the ovary, brain and other tissues in the body.
As a result of the structural similarity of this enzyme to other cytochrome P450 enzymes in the
steroid synthesis pathway (among other pathways), aromatase inhibitors also can affect the
synthesis of other hormones and disrupt multiple signaling pathways. In general, aromatase
inhibitors have been developed as drugs for treatment of breast cancer and ovarian cancer in
postmenopausal women.
Few multigenerational animal studies with continuous exposure were found for chemicals that
act as aromatase inibitors. The aromatase inhibitors all appear to induce dystocia and to delay
delivery due to inhibition of estrogen synthesis at this critical stage of pregnancy. In order to
avoid the induction of dystocia, multigenerational studies terminate dosing several days before
the end of pregnancy in the rat and reinitiate dosing at birth of the pups.
4.2.2.9 Exemestane [A.3.b]
Beltrame etal. (2001)[A.3.b.l1) describe a large number of parameters evaluated in
developmental toxicity as well as male and female reproductive toxicity studies of exemestane
with reported doses of 2, 4, 5, 10, 20, 40, 50, 90, 100, 125, 200, 250, 500, 810, 1000 mg/kg by
gavage. The male/female sex ratio displayed an NMDR with an increased percentage of males
in two mid-dose groups in one cohort in one of the two studies but not the other three cohorts.
When Fls were mated to produce the F2, a sex ratio also displayed an NMDR with a reduced
sex ratio in a midrange dose group. A few effects with NMDRs were identified in the Beltrame
et al. (2001) study at doses above those that produced adverse effects. These received less
weight in the evaluation of NMDR as a common finding (section 4.2 literature search and
evaluation).
4.2.3 Androgen Hormone Systt
The androgen signaling pathway shares many molecular and cellular traits with the estrogen
signaling pathway. However, there are also some notable differences. Unlike the E pathway
which has several ligand activated nuclear receptors there is only one wildtype androgen
receptor (AR) in mammals. In addition, there are two physiologically active androgens. While
testosterone is the major regulatory steroid in many androgen-dependent tissues, other tissues
rely upon the conversion of testosterone to dihydrotestosterone (DHT) by the enzyme 5a
102 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
reductase. In the A pathway, as with the E pathway, there are tissue-specific cofactors that
imbue tissue-specific responses and multiple forms of the androgen response elements (AREs)
on different genes. AREs fall into two general classes, one of which is specific for AR and others
that also are activated by the progesterone receptor and the glucocorticoid receptor.
103 | Page
-------
Draft Deliberative - Do NOT Cite, Quote, or Distribute
State of the Science: Nonmonotonic Dose Responses V7
Table 4.3: NMDR From Studies Evaluating Androgen Hormone System.
Chemical
Species
Group Size
Doses Given
Dosing
route
Reference(s)
Appendi
X
location
Vinclozolin
rat
10 dams, 25 pups
3, 12, or 200 mg/kg
gavage
Hellwig et al. (2000)
B.l.b.2
Vinclozolin
rat
20
40, 200, 1000 ppm
feed
Matsuura etal. (2005)
B.l.b.3
Vinclozolin
rat
25 dams
4, 20, 100 mg/kg
feed
Schneider et al. (2011)
B.l.b.7
Procymidone
rat
8 or 16 dams
5, 10, 25, 50, 100, or 150 mg/kg
gavage
Metzdorff et al. (2007)
B.l.c.4
DEHP
rat
17 pairs (PO)
1.5 (control), 10, 30, 100, 300, 1000, 7500,
and 10,000 ppm
feed
Blvstone etal. (2010)
B.2.a.l
DEHP
rat
11-16 dams
0.015, 0.045, 0.135, 0.405, 1.215, 5, 15,
45, 135, and 405 mg/kg
gavage
Grande et al. (2006)
B.2.a.3
DEHP
rat
11-16 dams
0.015, 0.045, 0.135, 0.405, 1.215, 5, 15,
45, 135, and 405 mg/kg
gavage
Grande et al. (2007)
B.2.a.4
DEHP
rat
20
0.015, 0.045, 0.135, 0.405, 1.215, 5, 15,
45, 135, and 405 mg/kg
gavage
Andradeefa/. (2006a)
B.2.a.5
DEHP
rat
10-12
0.015, 0.045, 0.135, 0.405, 1.215, 5, 15,
45, 135, and 405 mg/kg
gavage
Andrade etal. (2006b)
B.2.a.6
DEHP
rat
6-15 dams
3, 10, 30, 100, 300, 600, or 900 mg/kg
gavage
Christiansen etal. (2010)
B.2.a.8
DEHP
rat
10-12 dams
20, 100, or 500 mg/kg
gavage
Dalsenter et al. (2006)
B.2.a.7
DEHP
rat
40
10, 500 or 750 mg/kg
gavage
Geetal. (2007)
B.2.C.3
DEHP
rat
1, 10, 100, 200 mg/kg
gavage
Akingbemi etal. (2001)
B.2.C.1
DEHP
rat
50
5, 50, 500, or 5000 ppm
feed
Poon et al. (1997)
B.2.C.5
DEHP
rat
4
10, 100, 500 mg/kg
gavage
Vo etal. (2009)
B.2.d.2
DEHP
mouse
10 dams
0.05, 5, or 500 mg/kg
feed
Pocar et al. (2012)
B.2.e.l
DEHP
mouse
9-21 dams
0.5, 1.0, 5, 500, 50,000, or 500,000 Hg/kg
gavage
Do etal. (2012)
B.2.e.2
DBP
rat
20 dams
50, 250, or 500 mg/kg
gavage
Zhang et al. (2004a)
B.2.b.4
DBP
rat
6-8 dams
20, 200, 2000, or 10,000 ppm
feed
Lee et al. (2004)
B.2.b.5
DBP
rat
?
4, 20, 100 or 500 mg/kg
gavage
Mahood et al. (2007)
B.2.b.6
DBP
rat
7 dams
0.1, 1.0, 10, 30, 50, 100, or 500 mg/kg
gavage
Lehmann etal. (2004)
B.2.b.7
DBP
rat
20
0.1, 1.0, 10, 100, or 500 mg/kg
gavage
Bao et al. (2011)
B.2.d.l
Semicarbazide
rat
10
250, 500, 1000 ppm
feed
Takahashi etal. (2009)
B.4.b
104 | Page
-------
Draft Deliberative - Do NOT Cite, Quote, or Distribute
State of the Science: Nonmonotonic Dose Responses V7
Chemical
Species
Group Size
Doses Given
Dosing
route
Reference(s)
Appendi
X
location
Semicarbazide
rat
20 dams
40, 75, or 140 mg/kg
gavage
Maranghi etal. (2010)
B.4.C
Semicarbazide
rat
20 dams
40, 75, or 140 mg/kg
gavage
Maranghi etal. (2009)
B.4.C
Prochloraz
rat
8 dams
3.9, 7.8, 15.6, 31.3, 62.5, or 125 mg/kg
gavage
Blystone et al. (2007a)
B.5.a.2
Testosterone
priopionate
rat
4 dams
0.1, 0.5, 1, 2, 5, or 10 mg
sc
Wolf et al. (2002)
B.6.b.2
Testosterone
enanthate
Human
male
10-14
25, 50, 125, 300, and 600 mg/week
im
Grav et al. (2005)
B.6.b.l2
SARM
JNJ28330835
rat
10
2.5, 1, 30 mg/kg
gavage
Allan etal. (2007)
B.7.f
105 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
4.2.3.1 Vinclozc ..b]
Most perinatal studies conducted in rats or mice with exposure to vinclozolin (Ostby et al.
(1999)[B.l.b.l1; Flynn et al. (2001)[B.l.b.61; Owens and Chaney (2005)[B.l.b.81; Hass et al.
(2007)[B.l.b.41; Metzdorff et al. (2007)[B.l.b.4, B.l.c.4]; Christiansen et al. (2009)[B.l.b.51) did
not identify effects that displayed NMDRs. Those that did are presented below.
Hellwig etal. (2000)[B.l.b.21) administered vinclozolin by oral gavage of 1, 3, 6, 12, or 200
mg/kg to Long Evans (LE) and 3, 12, or 200 mg/kg to Wistar rats from day 14 of pregnancy until
day 3 of lactation, and they monitored the reproductive development of all the male offspring
from every litter through adulthood. In the Wistar strain, terminal body weight (at 3 mg/kg/d),
testes weight (at 12 mg/kg/d), epididymal weight (at 3 and 12 mg/kg/d) and coagulating gland
weight (at 3 mg/kg/d) displayed slight, but statistically significant, NMDRs. However, none of
these NMDRs were reproduced in the LE strain. Additionally, it is not clear whether the
statistical analysis of the organ weight data was adjusted for litter effects; failure to make such
an adjustment can affect statistical significance, as the error degrees of freedom (df) is
increased, and the standard error (SE) is decreased. NMDRs were identified for body and organ
in one rat strain weights but not the other study, nor were they seen in the study by Ostby et al.
[1999)[B.l.b.l]) in LE rats given 3.125, 6.25, 12.5, 25, 50, or 100 mg vinclozolin/kg by gavage.
Matsuura etal. (2005)[B.l.b.31 examined the effects of 40, 200,1000 ppm vinclozolin in the
diet on male and female rats of vinclozolin administration over two generations. There were
more than 200 dose response curves, each with 3 group comparisons with the control for a
total of 600 statistical comparisons. An NMDR was identified in liver weights for males but not
females at 40 ppm, and in diestrus uterine weight for the P0 generation only. These were not
seen in subsequent generations, or both sexes in the same generation (for liver weight).
In a one generation study by Schneider et al. (2011)[B.l.b.71). vinclozolin was administered in
the diet for nominal dosing of 4, 20, or 100 mg/kg for two (P0 females) to four (P0 males) weeks
prior to mating through PND 21. After weaning, F1 animals were maintained on the diet
through PND 70. The effects observed that displayed NMDR were these.
¦ Percent pup survival from PND 4 to 21 was reduced by 11%; however, litter sizes at PND
4 and 21 were not reduced
¦ Two of six measured antibody titers were reduced and displayed NMDRs
4.2.3.2 Procymido .c]
106 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Procymidone is a dicarboximide fungicide like vinclozolin and displays AR antagonism in vitro
and in vivo. There is a reasonably comprehensive published data set on procymidone including
in vitro AR binding and gene expression assays, short-term in vivo assays with some genomic
data, and multigenerational studies. The dose levels used in these studies range from low
mg/kg/d to hundreds of mg/kg/d. Most of the studies used in regulatory agency risk
assessments, including several with relatively low procymidone exposure levels are summarized
in the assessment documents; however, the dose response data are not available for public
review. An NMDR was noted for left and right testis weights in male rat offspring exposed
during gestation to 5,10, 25, 50, 100, or 150 mg procymidone/kg by gavage with weights
reportedly increased at 10 mg/kg/d but reduced at 100 and 150 mg/kg/d; the weights differed
from the control value by only 1.3 and 0.9 mg, with standard errors of about 1.2 mg Metzdorff
et al. (2007)[B.l.c.41). In other perinatal studies rats were exposed to procymidone at 5, 10, 25,
50, 100, or 150 mg/kg by gavage Hass et al. (2007)[B.l.c.51), 3, 10, 30, and 100 mg/kg by gavage
Owens et al. (2007)[B.l.c.l1). None of the other endpoints that were measured displayed
effects with NMDR.
4.2.3.3 Phthalates [B.2]
Some phthalate esters administered to pregnant rats in utero cause male reproductive tract
abnormalities, fetal loss, abortions and skeletal malformations. The reproductive effects in the
fetal male rats arise from abnormal testicular androgen and insulin-like factor 3 (insl3) hormone
synthesis. By contrast, these chemicals apparently induce pregnancy loss by reducing maternal
ovarian progesterone synthesis. In young pubertal males, these same phthalates disrupt Sertoli
cell function, hormone synthesis and induce testicular atrophy in a wide range of mammalian
species.
The list of phthalate esters that induce these effects continues to grow as more research is
performed on this chemical class: DEHP, DBP, butyl benzyl phthalate (BBP), diisobutyl phthalate
(DIBP), dipentyl phthalate (DPeP), dicyclohexyl phthalate (DCHP), dihexyl phthalate (DHP),
diisononyl phthalate (DINP); and methylethyl hexyl phthalate (MEHP). Of these DPeP is the
most potent, and DINP is only weakly positive for induction of the above effects. Studies also
demonstrate that some phthalates do not produce reproductive effects at any dose. When
mixture studies are conducted with phthalates they act in a dose-additive manner Howdeshell
et al. (2008) where the phthalate mixture effect is a function of the dose and relative potency
of each chemical.
There are numerous robust multigenerational and pubertal studies of phthalates, providing a
relatively comprehensive data base for examination of the shape of the dose response curves
107 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
over a broad range of closes. A few examples of apparent low-dose NMDRs have been reported
in the literature, none of which were replicated across studies.
Low-dose phthalate studies that failed to monitor background phthalate levels in control diets
and bedding are difficult to interpret. Some authors have reported effects at dosage levels that
were likely below the background phthalate exposure levels reported in control animals. For
example, the control dietary values determined by measurement of background contamination
levels reported by Blystone etal. (2010) is 1.5 ppm (equivalent to about 0.12 mg/kg/d); this
value exceeds the levels of DEHP administered in several low-dose studies (Andrade et al.
[2006c)[B.2.a.5], Andrade et al. (2006b)fB.2.a.5. B.2.a.6], Andrade et al. (2006c)[B.2.a.51;
Grande et al. (2006)[B.2.a.31; Grande et al. (2007)[B.2.a.3. B.2.a.4] and Do et al.
(2012)[B.2.e.21). Similarly, Kondo etal. (2010)[B.21) found DBP and DEHP in each of 12
untreated rodent diets tested (0.14 to 1.41 ppm) as well as all of 13 bedding materials (0.02 to
7.6 ppm) examined. The data, when further analyzed to produce a daily intake results in the
background level consumed by rodents that can be as much as 1.4 |ag/g diet consumed. If a
pregnant rat eats about 20 g food/d then the resulting daily consumption can be as high as 84.7
l-ig/kg body weight/d. Pregnant mice eat about 5 g per day resulting in as much as 213.6 M-g/kg
body weight/d Kondo et al. (2010). Phthalates also were found in cereals, fat, oil, and other
products in levels up to 10 ppm by Wormuth et al. (2006) and up to 58 ppm by Jarosova et a I.
(2009). Therefore, it is likely that tissues from control animals contain phthalate and phthalate
metabolites Jarosova et al. (2009). The existence of phthalates as background contamination
confounds the interpretation of studies that failed to control for background phthalate
exposure and attempt to expose test animals to levels of phthalates falling below that
background level.
Since several of the phthalate esters disrupt development in rats via a common mechanism of
toxicity the following descriptions have been organized by species (rat then mouse),
developmental period (in utero and lactation then pubertal) and then by chemical (DEHP then
DBP).
4.2.3.3.1 Effects of Phthalates Esters in Rats: In ui d
lactational studies: DEI a]
Grande etal. (2006)[B.2.a.31. Grande etal. (2007)[B.2.a.3. B.2.a.4]) and Andrade et al. (2006c);
Andrade et al. (2006a)rB.2.a.51) studied the effects of 0.015, 0.045, 0.135, 0.405, 1.215, 5, 15,
45, 135, and 405 mg DEHP/kg by gavage on Wistar rat offspring that were exposed by oral
gavage from GD 6 to PND 2. The low concentrations in these studies are in the reported range
for background feed levels Kondo et al. (2010). Body weights of females necropsied at PND 1 (2
per litter) displayed an NMDR, being increased at 0.045, 1.125 and 5 mg/kg/d, and liver weights
108 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
were increased in the two highest dosage groups Grande et al. (2006)[B.2.a.31). The effects on
body and liver weights seen on PND 1 were not present in females necropsied at weaning on
PND 22. Several effects were reported as having NMDRs in the review by Vandenberg et al.
(2012); however, these conclusions are not consistent with the study authors' analysis Grande
et al. (2006)[B.2.a.31). Grande etal. (2006)[B.2.a.31) reported that all maternal and offspring
lactational indices, litter sizes and body weights were unaffected by DEHP treatment at any
dose level. Body weights of females necropsied at PND 1 (2 per litter) displayed an NMDR,
being increased at 0.045, 1.125 and 5 mg/kg/d and liver weights were increased in the two
highest dosage groups. The effects on body and liver weights seen on PND 1 were not present
in females necropsied at weaning on PND 22. AGD and nipple retention in F1 females on PND
22 also were not affected by DEHP treatment. They did detect a dose related delay in VO in 15-
405 mg/kg/d dose groups, all being delayed by about 2 days, but the age at first estrus was not
affected. Several effects including "vaginal opening, and first estrous" were reported as NMDRs
by Vandenberg et al. (2012)(Table 7), however, these conclusions are not consistent with
Grande et al. (2006) analyses of the same data.
Maternal kidney weight displayed an NMDR, being increased in dams treated with DEHP at
0.045 mg/kg/d Grande et al. (2007)[B.2.a.3. B.2.a.4]). Similar to F1 females Grande et al.
(2006)[B.2.a.31 and Andrade et al. (2006c)[B.2.a.51 reported that body weights of males
necropsied at PND 1 (2 per litter) displayed an NMDR, being increased at 0.045, 1.125 and 5
mg/kg/d. Liver weights were also increased in the two highest dosage groups. Also similar to
their female siblings, the effects of DEHP on body and liver weights at PND 1 were not present
in males necropsied at weaning on PND 22.
In Andrade et al. (2006c)[B.2.a.51 an NMDR was identified for AGD at PND 22, being increased
in the 0.015 mg/kg/d DEHP group versus controls and decreased at 405 mg DEHP/kg/d. An
NMDR for testis weight was identified, and it was statistically significantly increased at 5,15, 45
and 135 mg/kg/day on PND 22. This observation qualitatively differed from exposure to higher
doses in Andrade et al. (2006c)[B.2.a.51. These effects also differ from testis weight data from
other studies in prepubertal male rats: studies wherein animals were exposed to DEHP at 3, 10,
30, 100, 300, 600, or 900 mg/kg Christiansen et al. (2010)[B.2.a.81) and to 10, 100, 300, or 900
mg;kg Noriega et al. (2009)[B.2.c.41) by gavage. The lack of an NMDR on the age of PPS is
consistent with the observations of Noriega etal. (2009)[B.2.c.41. Noriega et al. (2009) did not
replicate the NMDR identified at 1 and 10 mg DEHP/kg/d which was reported in Ge et a I.
(2007)[B.2.a.5 and B.2.C.3]) who treated rats with 10, 500, or 750 mg/kg by gavage. Andrade et
al. (2006a)[B.2.a.51 measured a number of endpoints, some of which displayed NMDRs at doses
consistent with reported background dietary DEHP levels Kondo et al. (2010).
109 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Dalsenter et al. (2006)[B.2.a.71 evaluated the effects of DEHP on rats treated with 20, 100, or
500 mg/kg by gavage for reproductive function and sexual behavior of F1 male rat offspring rats
after in utero and lactational oral gavage treatment. Weaning weight displayed an NMDR being
increased only at 100 mg/kg/d, the intermediate dose administered.
Christiansen etal. (2010)[B.2.a.81 administered 3, 10, 30,100, 300, 600, and 900 mg
DEHP/kg/day from GD 7 to PND16 by oral gavage in two blocks. NMDRs were seen but were
generally not reproducible from block to block. The NMDR that were identified in this study
were in F1 males (but not females) as the weaning weight was increased only after treatment
with 100 mg/kg/d in block 1 but not block 2. Post-implantation loss was increased only after
treatment with 10 mg/kg/d, in block 2 but not block 1. The diagnosis of "Mild" genital
dysgenesis (on a score of 0 to 3, mild=l) was increased in 16 day old males in all dose groups
except 30 mg/kg/d. The biological significance of this change is uncertain since malformation of
the external genitalia has not been described for this dose in adult F1 animals. NMDRs were
identified for the gene expression of PCB C3 and ODC in block 2 but not 1. NMDR were also
identified for some organ weights across the study. Right but not left testis weight was reduced
in block 1 (but not 2) after treatment with 100 mg/kg/d; this was not statistically different from
control after treatment with 300 mg/kg/d. Levator ani bulbo-cavernosus (LABC) weight was
decreased in block 1 in all treated groups except 600 mg/kg/d. In block 2, 10 and 30 mg/kg/d
slightly reduced LABC weight, but 100 mg/kg/d did not. Adrenal weight was reduced in block 1
after treatment with 10, 100 and 900 mg/kg/d but not in the other dose groups; the effects at
10 and 100 mg/kg/d were not replicated in block 2. Many of these NMDR were no longer
apparent when the two blocks were pooled and analyzed together.
4.2.3.3.2 Effects of Phthalates Esters in Rats: In ui d
lactational studies: DBP [B.2.b]
Several perinatal studies with DBP exposure in rats treated with 250, 500, or 750 mg/kg
Mvlchreest et al. (1998UB.2.b.31. 100, 250, 500 mg.kg Mvlchreest et al. (1999UB.2.b.21..5. 5, 50,
or 100 mg/kg Mvlchreest et al. (2000)[B.2.b.l1. and 10, 50, or 500 mg/kg (Wyde et al.
(2005)[B.2.b.81) by gavage did not identify any effects with NMDRs. Those studies that did are
presented below.
Zhang et al. (2004b)[B.2.b.41) administered 50, 250, and 500 mg/kg DBP by oral gavage from GD
1 to PND 21, and the male offspring were examined through adulthood. Of the 20 endpoints
measured (with 60 potential statistical comparisons with the control group), prostate weight
had an NMDR, being reduced only at 250 mg/kg/d, a dose at which several other endpoints also
were statistically significantly affected in a monotonic manner.
110 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
In a study by Lee etal. (2004)[B.2.b.51, pregnant rats were treated with 20, 200, 2000 or 10,000
ppm DBP in the diet from GD15 to PND21. Several NMDRs were identified in this study. The
number of live pups was reduced and body weights were increased in the low dose group only
at PND2. The age at puberty in males occurred 1.3 days earlier at 200 ppm than controls but
not at higher doses. Pituitary weight in males at week 11 was increased at 20, 200 and 2000
ppm, but not in females or in males at other ages. Prostate weight was increased at week 11 in
the 200 ppm group but not at 20 weeks. An NMDR was also identified for the severity of
histological alterations in male mammary gland of the low and middle dose groups at 11 and 20
weeks. No mammary gland effects were present at weaning or in females.
In a study by Mahood et al. (2007)[B.2.b.61 pregnant Wistar rats were gavaged daily with 4, 20,
100 or 500 mg DBP/kg from GD 13.5 to either GD20.5 (fetal samples) or GD21.5 (postnatal
tissue), and the male fetuses and F1 adult male offspring were examined. All of the effects
displayed monotonic responses. The frequency of mild Leydig cell clusters in GD 21.5 testes
showed an NMDR, however, the severity of the testis alterations progressed from medium at
100 mg/kg/d to large at 500 mg/kg/d.
Lehmann etal. (2004)[B.2.b.71 administered 0.1,1.0,10, 50, 100, or 500 mg DBP/kg by oral
gavage to pregnant rats on GD 12 to 19. Fetal testes were isolated on GD19, and changes in
gene and protein expression and testicular testosterone concentration were measured. The
protein expression and testosterone concentrations displayed monotonic responses, and these
effects were statistically significant at 50 mg/kg/day and higher. Of the mRNAs evaluated, SR-
Bl, 33-HSD and c-Kit displayed NMDRs, being significantly reduced at the lowest dosage levels
(0.1 and 1 mg/kg/d) but not at 10 mg/kg/d. Attempts to replicate the low dose NMDR effects
of DBP on SR-B1 and 3|3-HSD were unsuccessful, whereas some of the reductions reported at
the higher dosage levels were replicated (LE Gray personal communication with author).
4.2.3.3.3 Effects of Phthalates Esters in Rats: Peripubertal
exposure effects of DEHP on male rat reproductive
development [B.2.c]
Phthalate treatment during peripubertal development causes testicular lesions in several
mammalian species including the rat, ferret, guinea pig, and some strains of mice and hamsters;
the sensitivity of the species varies considerably. Some effects have also been reported for
male and female marmosets. The constellation of effects of peripubertal treatment in the
young males includes testicular atrophy, seminiferous tubule hypospermatogenesis, reduced
sperm counts, infertility, delayed puberty, reduced androgen-dependent tissue weights and
transient alterations in serum androgen levels.
Ill | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Results from multigenerational studies demonstrate that some of the effects of phthalate
treatment, such as delayed puberty, are consistently seen only when the treatment continues
after weaning through the peripubertal period, rather than with in utero and lactational
exposure alone. Multigenerational studies with pubertal dosing of 250, 500, or 1000 mg
DBP/kg or 200, 400, 100 mg/kg, have not reported NMDRs for PPS Wolf et al. (1999).
An examination of the literature on the effects of DEHP on pubertal development indicates that
two of the studies have reported NMDRs on pubertal measures in young male rats treated with
10, 500, or 750 mg/kg Ge et al. (2007)[B.2.c.31; Vo et al. (2009)[B.2.d.21) for the same
endpoints. These NMDRs are not consistent with other data from the same laboratory
(Akingbemi et al. (2001)[B.2.c.l1; Akingbemi et al. (2004)[B.2.c.21) or with other laboratories
(Noriega et al. (2009)[B.2.c.41; Andrade et al. (2006c)[B.2.a.51; Wolf et al. (1999)[B.5.b.41.
Ge et al. (2007) did studies in the same laboratory as those by Akingbemi et al. (2001)[B.2.c.l1
and Akingbemi et al. (2004)[B.2.c.21. In their study, Ge et al. (2007)[B.2.c.31 dosed rats with
DEHP by oral gavage at 10, 100, or 200 mg/kg Akingbemi et al. (2001)[B.2.c.l1.10 or 100 mg/kg
Akingbemi et al. (2004)[B.2.c.21), and 10, 500 or 750 mg/kg Ge et al. (2007)[B.2.c.3 ]; this was
done by gavage for 28 days from PND 21 to 49. This study reports an NMDR for the age at
puberty in treated males, with a 1.8 day acceleration at 10 mg/kg/d; no statistically significant
effect at 500 mg/kg/d; and a 6.9 day delay at 750 mg/kg/d. The authors also reported that
body and seminal vesicle weights and serum T levels were increased at 10 mg/kg/d. It appears
that the authors randomly assigned rats to treatment groups rather than controlling for
weaning body weight as recommended in the EPA EDSP test guideline. This may introduce a
confounder in the pubertal male assay, and failure to control for this confounder could explain
the acceleration in PPS and other effects attributed to this 10 mg/kg/d dosing. In the
Akingbemi etal. (2001)[B.2.c.l1 study, this dose did not cause an increase in body weight in
similarly exposed males, whereas LH was increased, an effect not seen by Ge et a I.
(2007)[B.2.c.31. There is difficulty in interpreting whether serum hormone changes are adverse
in that the two studies from the same laboratory report serum T and LH levels that differ by
two fold from the control group across the studies; this is greater than the reported DEHP
effects at 10 mg/kg/d. In addition, Noriega etal. (2009)[B.2.c.41) did not see accelerated
puberty or increased serum T levels in males exposed to DEHP at 10 mg/kg/d, whereas
statistically significant delays in puberty were seen at 300 mg/kg/d and above. Similar studies
with DBP given at 0.1, 1.0, 10, 100 or 500 mg/kg by gavage (Bao et al. (2011)[B.2.d.l1) a
phthalate with the same mechanism of toxicity and mode of action as DEHP, did not identify
any of the hormonal changes reported by Ge etal. (2007)[B.2.c.31.
When Ge etal. (2007)[B.2.c.31) dosed rats with DEHP for 14 days from PND 21 to 34 none of the
low dose effects from exposure on PND 21-49 were noted at 10 mg/kg/d. The authors also
112 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
reported that MEHP in vitro produced an NMDR on Leydig cell T production, increased above
baseline at 10~4and 10"3 M and then decreased at 10~2 M.
While effects on PPS and body and reproductive organ weights with an NMDR were observed in
the Ge etal. (2007)[B.2.c.31, they were not seen by Noriega et al. (2009)[B.2.c.41 in either of
two rat strains studied, and some of the other effects reported in Ge et al. are not consistent
with findings from other publications.
4.2.3.3.4 Pubertal effects of DEHP in male rat [B.2.c]
Results from a study wherein rats were dosed with DEHP from PND 23 until necropsy at 56 or
98 days of age reported no NMDR Noriega et al. (2009)[B.2.c.41. However, Poon et al.
(1997)[B.2.c.51 performed a 90 day study with 5, 50, 500, or 5000 ppm DEHP in the diet in
young male and female SD rats; this study conformed to OECD Guidelines under GLP. Of the
endpoints measured and effects observed, NMDR were identified for two of six serum
biochemistry indices in female, but not male, rats.
4.2.3.3.5 Pubertal effects of DBP in male rat [B.2.d]
In Bao et al. (2011)[B.2.d.l1) five week old male SD rats were administered DBP orally for 30
days, and reproductive organ weights, testicular histopathology and serum hormonal levels
were measured at necropsy. In addition, proteomic analysis was performed on testes to
determine if any proteins were affected by DBP treatment; however, these analyses were not
conducted in the 100 and 500 mg/kg/d groups. Several nonmonotonic effects were identified.
¦ Serum E2 was increased at 0.1 and 500 mg/kg/d but not in other dose groups.
¦ Serum LH was increased in all groups, but the increase was not statistically significant at
1.0 mg/kg/d.
¦ Of the twenty proteins from the testis reported to be affected in the low dose groups
(10 mg/kg/d and below) the dose response data are only shown for 4 of these, one of
which (vimentin) displayed an NMDR (n=3 pergroup).
In addition, peripubertal DBP-treatment did not induce a NMDR on serum testosterone at low
doses, in contrast to some of the aforementioned reports where DEHP was administered during
puberty.
Vo etal. (2009)[B.2.d.21 administered 10, 100, and 500 mg/kg DEHP by gavage to rats from PND
21 to 35. Some authors have identified NMDR for seminal vesicle weight, epididymal weight,
and testicular expression of steroidogenesis genes as measured by rtPRC Vandenberg et al.
(2012). In reviewing the Vo et al. (2009) study we authors of the current evaluation note that
113 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
the data suggest no alteration of the three steroidogenesis genes measured, StAR, CYPllal and
HSD3|31,by DEHP treatment at any dose level (Fig 4, a,b,c in Vo et al. (2009)[B.2.d.21). However
Vo et a I. (2009) [B.2.d.2] do identify an additional set of 2 genes evaluated in the testis,
whichdo express NMDRs. Vo et al. (2009) considered these to be markers of exposure. In
summary NMDR were identified for two reproductive organ weights and for mRNA expression
levels for two testis genes. The small sample size (n=4/group) and lack of concordance of the
organ weight effects reported here with several other studies raises uncertainty about the
biological significance of these results.
4.2.3.3.8 Effects of Phthalates Esters in Mice: In litem and
lactational exposure effects of DEHP on male rat
reproductive development [B.2.e]
NMDRs have been described at low dose levels for some endpoints after in utero DEHP
exposure in mice. However, there is uncertainty in interpreting some of these results as they
report administered DEHP at levels below the published background levels in rodent diets of
this ubiquitous contaminant. In addition, the effects are small and are not necessarily adverse.
At present there is debate in the scientific community about the ability of phthalates, such as
DEHP, to demasculinize fetal mice as is seen in the fetal rat. The link between the changes seen
in the fetal mouse to postnatal reproductive outcomes is not established.
DEHP was administered to mice at 0.05, 5, and 500 mg/kg in the diet during pregnancy and
lactation, and the male and female offspring were examined and necropsied at PNDs 21 and 42
Pocar et al. (2012)[B.2.e.l1. DEHP induced an NMDR response on testis weight, being reduced
only in the low dose group. The analysis of these data does not appear to be adjusted for litter
effects, which can inflate the statistical significance of the reported effects. In vitro fertilization
studies identified NMDR associated with reductions in the cleavage and blastocyst rates in the
low dose group. It is not reported how many times the in vitro assays were replicated as the
analyses appear to be based upon a single pool of oocytes and fertilized ova for each dose
group. However, the expression of genes related to steroidogenesis were altered in the ovary,
testis, and pituitary of the offspring in a dose related manner.
Do etal. (2012)[B.2.e.21) administered 0.5, 1.0, 5, 500, 50,000, or 500,000 fig/kg DEHP by
gavage from GD 9 to 18 orally to mice and examined maternal and fetal (1M males only)
hormones on GD 18. Although the study was conducted in three blocks (incomplete block
design) only the pooled data are presented in the paper. Data for fetal males from intrauterine
positions other than 1M is not presented. The effects are small, and none of the effects were
statistically significantly altered by DEHP-treatment (all F values from the ANOVA are non-
significant; >0.05). The study authors, however, report several of these effects as NMDR. The
biological significance of the reported effects (small increases in serum T), is not clear as T
114 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
changes dramatically in the dam and fetal male during this stage of pregnancy. There is
uncertainty as to comparison of the administered doses of DEHP to reported background levels
in the animal's diet and bedding. Some of the administered doses are below the levels of
phthalates reported in animal diets and bedding Kondo et al. (2010).
4.2.3.4 Semicarbazide [B.4]
Semicarbazide is a metabolite of the banned antibiotic nitrofurazone and a breakdown product
of azodicarbonamide, which is used as a blowing agent in plastic gaskets EFSA (2005).
Semicarbazide is known to inhibit enzymes, such as lysyl oxidase, semicarbazide-sensitive
amine oxidase and glutamic acid decarboxylase. Semicarbazide acts as osteolathyrogen, and
induces osteochondral and vascular lesions in young rats due to impaired cross-linking
reactions of collagen and elastin Takahashi et al. (2009)[B.4.b1). In addition, teratogenic effects,
such as induction of cleft palate and aortic aneurysms, have also been reported. It has been
described as an example of an EDC that displayed an NMDR for the age at puberty in male rats.
Semicarbazide was administered at 250, 500, and 1000 ppm in the diet to 6 week old male and
female rats for 90 days Takahashi et al. (2009)[B.4.b1. The study included a large number of
endpoints, a few of which displayed NMDRs at doses above the NOAEL; namely one of the >40
hematology measures and food consumption in male rats.
Semicarbazide was administered at 40, 75, and 140 mg/kg to 23 day old male and female rats
by oral gavage for 28 days Maranghi et al. (2010; Maranghi et al. (2009)[B.4.c1. Puberty in
males was accelerated at 40 and 75 mg/kg/d but delayed at 150 mg/kg/d. Given all the lesions
seen with semicarbazide in the lowest dose level, taken together with what is known about the
mechanism of toxicity for the bone and vascular lesions, we authors of this state of the science
evaluation do not consider semicarbazide an EDC. The NMDR effect on puberty at high dosage
levels is of uncertain biological releavance.
4.2.3.5 Multiple Molecular Initiating Events [B.5]
It has been shown that a number of environmental chemicals disrupt the androgenic signaling
pathway in an antiandrogenic manner impacting multiple toxicity pathways. Pesticides such as
linuron and prochloraz act as AR antagonists and inhibitors of fetal T synthesis. In contrast to
the phthalates (discussed above), prochloraz and linuron reduce fetal T production without
inhibiting mRNA expression of steroidogenic enzymes or the insl3 hormone. Although the
chemicals that disrupt androgen signaling in the fetal male rat produce some malformations in
common, the specific profiles of effects are chemical specific.
115 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
4.2.3.5.1 Prochloraz [B.5.a]
In the first of a series of experiments (Blystone et al. (2007a)[B.5.a.l1. Blystone et al.
(2007b)[B.5.a.21), SD weanling male rats were dosed by gavage with prochloraz at 31.3, 62.5, or
125 mg/kg/day of from 23 to 42 or 51 days of age. There was a statistically significant delay in
PPS at 125 mg/kg/day PCZ and several of the androgen-dependent organ weights were
decreased significantly. At both ages, serum testosterone levels and ex vivo testosterone
release from the testis were statistically significantly decreased whereas serum progesterone
and 17a-hydroxyprogesterone levels were statistically significantly increased at dose levels
below those that affected PPS or reproductive organ weights. The hormone results suggested
that PCZ was inhibiting CYP17 activity. None of the effects displayed an NMDR; however, the
dose response curves varied statistically significantly among the groups necropsied at 42 and 51
days of age, as did the LOEL for organ weight reductions being 31.3 mg/kg at 42 days and 62.5
at 51 days of age. In the second pubertal study rats were dosed by gavage with prochloraz at 0,
3.9, 7.8,15.6, 31.3, or 62.5 mg/kg/d. Serum testosterone levels and ex vivo testosterone
production were significantly reduced at 15.6 mg/kg/d whereas ex vivo androstenedione
production was significantly reduced at 7.8 mg/kg/d with 3.9 mg/kg/d being a NOEL. Two
related endpoints showed NMDRs in this study as body weight at necropsy and glans penis
weights were increased at 3.9 mg/kg/d. In order to determine if prochloraz displayed AR
antagonism in vivo independent of its effects on testosterone synthesis, castrated immature
male rats were dosed with androgen and 0,15.6, 31.3, 62.5, or 125 mg/kg/day PCZ for 10-11
days (Hershberger assay). In this assay, androgen-sensitive seminal vesicle and LABC weights
were significantly decreased at 125 mg/kg/d. Liver weight was increased at 31.3 mg/kg/d and
above and serum LH was decreased at 125 mg/kg/d. These effects were all dose related. In
this study, glans penis weight displayed an NMDR being reduced only at 15.6 mg/kg/d, a
response in the opposite direction from the NMRD seen in the pubertal study. Effects
described in other studies from prenatal or pubertal exposures to prochloraz displayed
monotonicity, and no other effects with NMDRs were seen Noriega et al. (2005)[B.5.a.41;
Christiansen et al. (2009)[B.5.a.31.
4.2.3.6 Androgen Receptor Agonists [B.6]
There appears to be no publically available rodent multigenerational study with administered
androgen agonist in the diet or by oral gavage over a broad dose range. Most data sets have
been developed for submission for pharmaceutical product approval. The dietary and oral
gavage studies on the synthetic androgen trenbolone have been reviewed, and detailed
summaries are available on the internet. These do not, however, allow one to examine the
shapes of the dose response curves.
116 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
This class of chemicals provides some examples of NMDRs for clearly adverse effects by several
androgens including testosterone, Trenbolone, and methyltestosterone in mammals and fish.
The NMDRs identified with androgens in the mammalian studies administered the chemical
either by subcutaneous injection or through the use of implants. The relevance of these routes
of administration to human health can be debated. They were not considered to be useful for
the present evaluation based on the literature search and evaluation criteria (section 4.2
above). When effects were compared in the Hershberger Assay between injection versus oral
administration for androgens, the spectrum of effects are similar, but the oral route was
approximately 80 fold less potent Wilson et al. (2002)[B.6.a.l1). One study wherein an NMDR
was identified administered testosterone sc during pregnancy and evaluated the male and
female offspring Wolf et al. (2002)[b.6.b.21. In these studies, adverse effects were identified
after exposure at doses below which the NMDR were identified; thus, the characterization of
the NMDR would not affect determination of a NOAEL in the studies.
In addition to rodents, the effects of androgens have been studied extensively for decades in
many vertebrate species, including humans. In this section of the review, a number of
experimental studies on the dose-related effects of androgens in humans are presented. It is
clear from these studies that few if any NMDRs are identifiable associated with testosterone
administration to humans.
4.2.3.8. aolone [B.C
Trenbolone binds the AR and is a potent androgen agonist both in vitro and in vivo. Extensive
dose response studies with trenbolone have been conducted in a number of mammalian
species including rats, domestic animals, and nonhuman primates. Some of these studies are
summarized in an FDA document (NADA-138-612 Finaplix at
http://www.fda.gov/AnimalVeterinary/Products/ApprovedAnimalDrugProducts/FOIADrugSum
maries/ucmlll214.htm). Summaries from a multigeneration and supplementary hormonal
study conducted in rats (James, Huntington Research Centre), a preliminary oral toxicity study
in young male and female Cynomolgus monkeys (Sortwell, Huntington Research Centre), and a
hormonal study in female Rhesus Macaques (Hess, Oregon Regional Primate Research Center)
report effects on reproductive or hormonal endpoints following exposure to trenbolone
acetate; however, they do not indicate NMDRs. The actual data are not available for review.
In other studies, prenatal trenbolone exposure to 0.1, 0.5,1.0, or 2.0 mg/kg Hotchkiss et al.
[2007)[B.6.a.l] by gavage and 12.5, 25, 50, 100, or 200 |ag IP (Wilson etal. (2002)rB.6.a.21)
resulted in androgenic effects on developmental and reproductive endpoints that were all dose
related except that pup weight was reduced in the high dose group at birth but not at 13 days
of age (Wilson et al. (2002)[B.6.a.21). Findings from Hershberger Assays showed monotonic
117 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
dose related changes from oral administration of 0.3,1.5, 8, and 40 mg/kg trenbolone by
gavage Owens et al. (2007)[B.6.b.41); however, sc administration produced an NMDR at high
dose on body weight gain over the ten-day dosing period.
4.2.3.6.2 Testosterone [B.6.b]
Investigators have demonstrated pronounced effects with NMDRs from sc or implants of
testosterone on the testis. It has been repeatedly demonstrated in young adult male rats
Robaire et al. (1979; Ewing et al. (1977; Walsh and Swerdloff (1973), rabbits Ewing et al. (1973),
and rhesus monkeys Ewing et al. (1976) that increasing doses of testosterone can result in a
reduction in LH. This is followed by declines in testis androgen levels, sperm production and
testis weight without causing increases in serum testosterone of androgen-dependent organ
weights. However, as testosterone dosage levels are increased above the nadir of the NMDR,
testis weight and sperm production levels are partially restored due to increasing levels of
intratesticular testosterone from the serum.
In addition, Wolf etal. (2002)[B.6.b.21 administered 0.1, 0.5, 1, 2, 5, or 10 mg testosterone
priopionate sc on GD 14-18. Although most effects displayed monotonic, dose-related changes,
three effects displayed pronounced NMDRs; these were for uterine weight, incidence of uterine
hydrometrocolpos, and survival of female offspring after puberty. Other prenatal Hotchkiss et
al. (2007)[B.6.b.31 or Hershberger studies Owens et al. (2006)[B.6.b.41 exposing rats to
testosterone sc, did not produce effects that displayed NMDRs.
In a series of studies, the dose-related effects of testosterone were described for a number of
physiological processes in healthy, eugonadal men, 18-35 years of age. No effects with NMDRs
were identified (Bhasin et al. (2001)[B.6.b.51; Singh and Nocerino (2002)[B.6.b.71; Sinha-Hikim
et al. (2002)[B.6.b.61; Storer et al. (2003)[B.6.b.81; Coviello et al. (2005)[B.6.b.l01; Bhasin et al.
(2005)[B.6.b.ll1). In a study to determine if sexual function in older men (libido, sexual activity,
and erectile function) was affected by graded doses of testosterone (Gray et al.
(2005)[B.6.b.l21). some effects did display NMDRs; these included overall sexual function score
and waking erection frequency.
4.2J ictiwe Androgen Receptor Modulators (SARM:
Over the last ten years SARMS have been developed for oral treatment of androgen-responsive
tissues, which provide the beneficial effect with minimal side effects. None of the currently
available SARMS are completely selective for the desirable anabolic effects on muscle and/or
bone without producing undesirable androgenic side effects in sensitive tissues such as the
prostate gland. SARMs have been developed that are orally active without causing liver
118 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
damage; in contrast to testosterone, more potent and selective SARMs have been developed
with enhanced tissue-selectivity. In elderly men with osteopenia or osteoporosis, it is desirable
to have a SARM targeting bone and muscle tissue but with lesser or no activity on the prostate
or testes. A SARM for women would ideally stimulate bone retention or libido and other sexual
function, without negative side effects such as masculinization, increased LDL/HDL ratios, or
liver dysfunction.
Since the typical target populations of SARMS are aging men and women, there are few
multigenerational test guideline-type studies. The animal models in use in the pharmaceutical
industry to evaluate the SARMS are often the adult castrated male or female rat. In addition, as
is the case for SERMS, in vitro assays have been useful to screen chemicals for tissue specific
effects using cell lines derived from the different steroid-responsive tissues; this is because
some of the selective responses of different tissues to SARMS arise from differences in tissue
specific coactivators or corepressors. In addition, differential metabolism of the different
steroidal androgens can also significantly alter the toxicity profile of the chemical (e.g.,
aromatizable or not or activated or inactivated by 5a reductase). The studies that were
available for review presented typical monotonic dose responses to exposure and are described
in Appendix B.7.
4.2.^ fold
Thyroid hormones are responsible for regulating day-to-day function of many biological
systems in mammals and other vertebrates, and they are essential regulators of early
development. As briefly described in Section 2, thyroid hormone levels and thyroid hormone
action are regulated by multi-tissue feedback-control systems, including the classic HPT
neuroendocrine steroid hormone feedback system that controls TH production and release. TH
affects gene transcription through nuclear receptors encoded by two genes, TRa or TR|3 and TH-
mediated gene transcription is particularly important for brain development Williams (2008). A
key feature of TH action in brain is the temporal sequence of events it supports, a feature that
increases the complexity of determining the impact of xenobiotic-induced alterations in thyroid
function. Transport of TH in the blood is controlled by serum binding proteins, thyroid binding
globulin (TBG) and transthyretin (TTR) have the highest affinity forTHs. Specific transporter
proteins (monocarboxylate transporter, MCT and organic anion-transporting polypeptide,
OATP) located on blood vessels, astrocytes, and neurons actively take up T3 and T4 (Williams,
2008; Bernal, 2011). Tissue concentrations of THs are also controlled by a number of cellular
mechanisms, most notably the expression and activity of deiodinase enzymes which provide an
additional means of control over TH-dependent gene regulation Bernal (2011; Williams and
Bassett (2011). In addition to the direct molecular action of TH binding to target receptors,
non-genomic actions of TH have also been identified Davis et al. (2008). T4 and its presumed
119 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
iodothyronamine metabolites, have rapid physiological action both in vitro and in vivo
indicating that TH may also act at non-nuclear sites.
4.2.4.1 Environmental Contaminants at oid Disruption
Environmental contaminants that cause adverse outcomes by acting as thyroid disrupting
chemicals (TDCs) (Fig 2.7) have been shown to affect one or more pathways in the thyroid
hormone system, which can then lead to an adverse outcome Murk et al. (2013; Crofton (2008;
Kohrle (2008; Brucker-Davis (1998; Capen (1997). Current knowledge indicates that the
majority of thyroid disrupting effects are mediated by molecular intiating events (MIEs) at sites
other than direct interference with thyroid receptor (TR) binding in the target tissue (Figure
4.11). This includes both thyroidal and extrathyroidal sites of action Capen (1997) that extend
well beyond the HPT axis. Several compounds are known to inhibit thyroperoxidase (TPO), a
critical enzyme for the synthesis of T3 and T4 in the thyroid gland, as the molecular initiating
event. There are also several anions (e.g., bromate, chlorate, nitrate, perchlorate, thiocyanate,)
that compete with iodine transport by the sodium-iodine symporter (NIS) Eskandari (1997;
Wolff (1964) Dohan et al. (2003). The NIS is essential for concentration of sufficient amounts of
iodine in the thyroid gland for adequate TH production. Competition for transport between
these environmental contaminants and iodine reduces iodine uptake into the thyroid gland,
resulting in a decrease of available iodine for production of TH, decreased TH production and
secretion, and a compensatory increase in circulating TSH. A variety of extrathyroidal
mechanisms also affect TH levels by: altering binding to hormone transport proteins, which may
change tissue availibity of TH, increased hepatic clearance of TH or TH catabolites, which will
lower cirrulating levels of TH, or inhibition of deiodination within peripheral tissues, which will
decrease tissue levels of T3 and decrease transcriptional processes Murk et al. (2013; Crofton
(2008; Capen (1997). The primary downstream consequences of most of these these effects
are to alter TH-directed transcription via changes in circulating or tissue concentrations of THs
(Figure 4.11).
120 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Figure 4.11: An adverse outcome pathway model for the effects of TDCs. Abbreviations: MIE,
molecular initiating event; NIS, sodium iodine symporter: T3, triiodothyronine; T4, thyroxine;
TH, thyroid hormone; TSH, thyroid stimulating hormone; TTR, transthyretin; UGT, uridine
diphosphate glucuronyltransferase (modified from Miller et al. (2009): Crofton and Zoeller
(2005).
4.2.4.2 Literature Search and Analysis
To evaluate the possibility that nonmonotonic dose-response relationships exist for many
thyroid active chemicals, two approaches were used:
1) An in depth review was performed for a chemical representative of one of three widely
accepted MIEs for which an extensive literature was available. The MIEs included
thyroperoxidase (TPO) inhibition; inhibition of the sodium iodide symporter (NIS); and
nuclear receptor activation induced upregulation of TH metabolism. These MIEs were
represented by propylthiouracil (PTU), perchlorate (CI04~), and polyhalogenated
aromatic hydrocarbons (PHAHs), respectively. These specific chemicals were chosen
because each is characterized by a rich database that encompasses studies from many
laboratories, contain a diversity of endpoints, and a provide data from a broad range of
doses and dosing scenarios.
121 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
2) The second approach adopted to evaluate the incidence and impact of NDMR for the
thyroid system was an extensive review of the published literature for evidence of
NMDRs for thyroid endpoints. The specifics of the literature search can be found in
Appendix C. Briefly, evaluation of 1153 papers ceded a total of 814 relevant references
and 2060 chemical-studies, of which 229 were non-mammalian in vivo chemical-studies
and 1831 were mammalian in vivo or in vitro chemical-studies. Chemical-studies were
defined as independent determinations of a dose or concentration-response for a
chemical (see Appendix C for details). The bulk of which were mammalian based
studies.
The publications containing the 1831 mammalian and in vitro chemical-studies were then
evaluated for the presence of NMDR effects at low doses that would impact regulatory action
(e.g, lowest effective dose for a chemical). The evaluation was conducted according to criteria
outlined in the 'Decision Tree' in Figure 1 of Appendix C. Briefly, the criteria for inclusion were:
Filter 1: Minimum of 3 dose levels plus a control evaluated
Filter 2: Evidence of a statistically significant NMDR on any thyroid endpoint
Filter 3: Absence of observations at lower dose levels in the study that would have been
used to determine the LOEL/LOAEL
Filter 4: a) Absence of other published reports on this chemical where effects were
observed at low levels.
b) Absence of other published reports for effects on other endpoints that would
have been used to determine the LOEL/NOEL below the dose identified as an
NMDR.
c) Absence of study quality concerns or statistical power issues that weakened
confidence in the NMDR observation.
4.2.4.3 iewiew of Three Major Targets for Nonmonotonic Thyroid
Pathway Disruption
4.2.4.3.1 TPO Inhibition to Reduces TH Synthesis-
Propylthioruacil Case Study
Several chemicals, including methimazole and PTU are known to inhibit TPO, an enzyme in the
thyroid gland critical for the synthesis of T3 and T4 Gardner et al. (1986: Cooper et al. (1983). In
peripheral tissues, PTU also inhibits deiodinases that convert T4 to its active form, T3 Cooper et
al. (1983). PTU has been used extensively as an experimental tool to study the impact of TH
disruption on physiological systems. As such, we have focused on PTU for a thorough review as
a representative classic thyroid disruptor because of the breadth of its database in addition to
the wide array of outcome measures that have been quantitatively examined in a dose-
response manner.
122 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Given the essential critical role of TH in early development and the greater susceptibility of the
developing fetus and neonate to TH insufficiencies, PTU has been utilized in a multitude of
studies as a model compound to alter thyroid status during a critical period of brain
development. Although the majority of these are restricted to a single high dose of PTU that
induced a severe state of hypothyroidism, a number of papers (~20) were identified that
evaluated a range of doses. Reviewed here are studies that evaluated the impact of graded
levels of thyroid disruption induced by PTU on serum markers and a variety of outcome
measures in response to adult, neonatal, or fetal exposure to PTU.
A number of early reports administering multiple concentrations of PTU to adult male or female
laboratory rats in the drinking water Cooper et al. (1983; Mannisto et al. (1979), food O'Connor
et al. (2002; Hood et al. (1999; O'Connor et al. (1999), or by oral gavage Cho et al. (2003)
focused on the impact of PTU on serum TH profiles. Studies varied not only in route, but also in
durations of exposures (days to weeks) and dose ranges, but in all cases three to five PTU
concentrations were used. For all of these reports, monotonic dose-responses were most
commonly observed for serum T3, T4 and TSH. Longer duration exposures led to more robust
declines in serum T3 and T4, and reductions were evident at lower doses with increasing time
of exposure. The lowest dose tested in these studies was the equivalent of 0.025 mg/kg/day
delivered in the food and for which a 20% decline in serum T4 was observed that was further
reduced with increasing dose O'Connor et al. (1999). Slightly higher doses were accompanied
by declines in serum T3 and increases in serum TSH. Higher doses for the longer exposure
periods were often accompanied by body weight deficits and increases in liver and thyroid
weights.
O'Connor et al. (2002) administered PTU in food (0, 0.025, 0.25, 1.0 and 10 mg/kg/day) and
reported that thyroid weight and thyroid histopathology appeared to be more sensitive than
serum hormones in short duration exposures (15 vs 30 days), with significant deviations from
control values evident in these measures in the lowest dose group tested (0.1 mg/kg/day). All
measures impacted by PTU, however, followed a monotonic dose-response pattern. In
ovariectomized female rats administered PTU for only 5 days, TH changes were restricted to the
highest dose group (10 mg/kg/day). Body weight, liver weight and reproductive hormones
were not affected with this shorter dosing regimen.
Hood et al. (1999) examined two TPO inhibitors, PTU and methimazole (MMI), with extensive 7-
point dose groups in adult male rats. PTU or MMI were administered in food at concentrations
of 0,1, 3, 10, 30, 100 and 300 ppm (equivalent to 0.06-18 mg/kg/day) for 21 days. Monotonic
dose-dependent reductions in total and free T4, and increases in TSH and thyroid weight were
observed and were first evident at a dose of 0.8 mg/kg/day; observations consistent with the
123 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
findings of O'Connor et al. (2002) described above. However, an NMDR was seen for a related
TPO-inhibitor, methimazole, for which a slight, but significant, increase in total T4 and free T3
were seen at the lowest dose tested, followed by monotonic declines as the dose increased.
In a developmental dosing paradigm, Blake and Henning (1985) reported on serum TH and
growth in rat pups exposed to 1, 5, 10, 50 and lOOppm PTU through the drinking water of the
dam beginning at birth of the pups and continuing throughout lactation. Serum T4 measured at
weaning was reduced at the lowest dose (~0.09 mg/kg/day), and body weight deficits were
evident at 5ppm (0.4 mg/kg/day) and above. Goldey etal. (1995) also assessed hormones,
growth, motor activity and auditory function in offspring of dams exposed to PTU (0,1, 5, 25
ppm in drinking water) beginning in late gestation (GD17) and continuing to weaning. No
effects on body weight were seen at doses less than 25ppm. Serum hormones were reduced in
a dose-dependent manner at the 5 and 25 ppm. Motor activity, acoustic startle and auditory
threshold were altered at 5ppm and above.
Using a similar dosing paradigm, Sawin et al. (1998) reported comparable effects on serum TH
and growth parameters resulting from PTU exposure. Brain weights were reduced in adult
male offspring at the two highest doses. Brain weight in female offspring, however, was
increased at the 5ppm dose with no change at the higher dose levels, suggestive of an NMDR.
The weight difference was small and may not be biologically relevant, in the light of the
following observations: control brain weights were more variable than the other groups;
decreases rather than increases in brain weight are expected from TH insufficiency; and no
effects were seen in brain/body weight ratios. Choline acetyltransferase (ChAT), an enzyme
necessary for synthesis of the neurotransmitter, acetylcholine, was reduced in the prefrontal
cortex of the adult brain at the 15ppm dose only, also suggestive of an NMDR, whereas
reductions were seen in the hippocampus at all dose levels. No changes were seen in either
brain region using another cholinergic marker, hemicholinium binding, or in ChAT activity
assessed at PND25. PTU-induced reductions in ChAT at earlier ages followed a monotonic dose
response pattern and were significantly reduced at the two highest doses.
A lower dose range of PTU was examined by Axelstad et al. (2008) using oral gavage
administration of PTU (0, 0.8,1.6, and 2.4 mg/kg/day) to pregnant rat dams beginning in early
(GD7 as compared to GD17 in studies described above) and terminating prior to weaning on
PND17. Dose-dependent reductions in serum TH were seen in dams on GD15 and in pups on
PND16, and the reduction was evident in neonates at the lowest dose assessed. A series of
apical measures of neurotoxicity including motor activity, spatial learning, and auditory function
were negatively impacted at one or both of the highest dose groups assessed with no evidence
of non-monotonicity.
124 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Neurological effects were reported at slightly lower closes achieved through drinking water
exposure to pregnant dams from GD6 to weaning of the pups on postnatal day (PND) 21 Sharlin
et al. (2010). Abnormalities in white matter composition of oligodendrocytes and astrocytes in
the corpus callosum and anterior commissure were detected at 2 and 3ppm PTU (0.18 and 0.27
mg/kg/day equivalents) and were linearly related to a monotonic decrease in serum T4 Sharlin
et al. (2008). Expression of DI2 and transporter protein MCT8 were increased in the
hippocampus in a dose-dependent manner and taken as evidence of induction of compensatory
responses to protect against thyroid hormone insufficiency Sharlin et al. (2010). However,
there was no evidence of nonmonotonicity in any of these brain-derived molecular and
structural endpoints.
Two apical neurotoxicity endpoints, synaptic physiology in the hippocampus and a test of
associative learning (trace fear conditioning) were evaluated in adult euthyroid offspring
following developmental PTU exposure Gilbert (2011). Serum T4 in pups declined in a
monotonic manner with significant decreases in all dose groups. Declines in dam serum T4
were restricted to the two highest dose groups. Excitatory and inhibitory transmission in the
hippocampus was dose-dependently reduced at the two highest doses. Synaptic plasticity, was
impaired at all dose levels, but to a greater extent at the intermediate dose level. Trace fear
conditioning was also altered, with apparent "improved learning" evident at low doses and no
change at the highest dose tested. The authors hypothesized that "improved learning" at the
lower doses may reflect increased anxiety in developmentally hypothyroid animals that served
to augment "freezing" behavior, the dependent measure in fear learning. The reversal of this
impairment at the higher doses may reflect a learning impairment that counteracts the
increased tendency to freeze. Observations of learning impairments at these levels of TH
insufficiency induced by PTU are in agreement with other reports of cognitive deficits at the
higher dose levels Axelstad et al. (2008; Gilbert and Sui (2006).
Lasley and Gilbert (2011) reported a monotonic effect on serum T4 accompanied by a
nonmonotonic pattern in expression of brain-derived neurotrophic factor (BDNF) in the
hippocampus and cortex of adult offspring of 1, 2, 3, and 10 ppm PTU drinking water exposed
dams. Decreases in BDNF protein were observed at low doses (1 and 2ppm, ~0.09 and 0.18
mg/kg/day in pregnant rat), a return to control at 3ppm (~0.27mg/kg/day), and exceeding
control levels in females at the highest dose level of lOppm (~0.9 mg/kg/day). No such change
was observed in another TH-responsive brain region (the cerebellum), nor were alterations
observed in any of these three regions in the brain of the neonate. Neurotrophins are central
to many aspects of CNS development Lewin (1996). but elevated levels of BDNF and other
neurotrophins also accompany brain trauma and injury where they promote neuroprotection,
recovery, or repair Pi Fausto et al. (2007; Chen et al. (2005). The authors suggest the
125 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
neuroprotective induction of BDNF serves to mask the effects of TH insufficiency evident at
lower doses and results in the observed biphasic dose-response pattern.
4.2.4.3.2 Sodium-Iodide Symporter (MIS) - Perchlorate Case
Study
Iodine is an essential element forTH synthesis. Several anions (e.g., perchlorate, thiocyanate,
nitrate, bromate) compete with iodine transport by the sodium-iodine symporter (NIS) Dohan
et al. (2003; Eskandari (1997; Wolff (1964). NIS ion transport is essential for concentration of
sufficient amounts of iodine in the thyroid gland to secure adequate TH production and
competition for transport iodine transport by environmental contaminants reduces iodine
uptake into the thyroid gland, resulting in decreased available iodine for production of TH.
When gland-level iodine deficiency is of sufficient magnitude to result in a decrease in TH
production and secretion, a compensatory increase in circulating TSH may occur with
subsequent enlargement of the thyroid gland. A large number of studies have been conducted
to evaluate the effects of these molecules in laboratory rodents, humans, and other species.
The best studied compound in the group of NIS inhibitors is perchlorate (CI04~). Perchlorate is a
ubiquitous environmental contaminant, with measurable levels found in ground and surface
waters, and has both natural and manmade sources Dasgupta et al. (2006). Perchlorate
exposure to humans can be through water or food Murray et al. (2008; Kirk et al. (2005). In
2005, a committee of the National Research Council (NRC) reviewed the state of the science on
perchlorate NRC (2005) to inform the USEPA's Integrated Risk Information System (IRIS)
Assessment for perchlorate U.S. EPA (2005b). The NRC Committee suggested a point-of-
departure for quantitative risk assessment of 0.007 mg/kg-day based on inhibition of NIS
thyroidal iodide uptake. More recently. Vandenberg et al. (2012) reviewed the perchlorate
literature and, while they found weak evidence of low-dose effects of perchloratej, they-did not
report evidence of NMDR associated with perchlorate exposure. Determination of the
perchlorate dose that may induce an adverse effect, how much of a decrease in TH is a concern,
and in what population, are all issues still surrounding the toxicity of perchlorate.
Perchlorate represents an environmental thyroid disruptor for which a very large experimental
database exists. This database was extensively reviewed by EPA U.S. EPA (2005b) and the
National Research Council NRC (2005) as part of regulatory determination of perchlorate
human health hazard risks. Overall, the shapes of the dose-response curves associated with the
early events following inhibition of uptake of iodine at the NIS are typically monotonic in nature
that exhibits a threshold that has been used as the point of departure to characterize the
hazard (e.g., U.S. EPA (2005b); Yu et al. (2002)). In our NMDR review, only a limited number of
perchlorate studies reporting NMDR were identified (summarized in Appendix C).
126 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Mannisto et al. (1979) administered 0, 10, 50, 100, and 500 mg/L potassium perchlorate in
drinking water for 5 days to male rats and found a statistically significant decrease in serum T3
at 50 and 100 mg/L and a return to near control values at 500mg/L. However, a decrease in
serum T4 was seen at all doses greater than or equal to 50 mg/L.
Thuett et al. (2002a); Thuett et al. (2002b) administered low doses (0, InM, luM, ImM;
approximately 0.01, 0.1 and 1.5 |ag/kg-day) of ammonium perchlorate in drinking water to male
and female mice during mating and throughout gestation and lactation. The pups were
euthanized at PND21 and measurements taken. A statistically significant increase in plasma T4
was observed in the low and mid dose groups, and no change was observed at the high dose in
PND21 mice. Thyroid histopathology analysis found a decrease in active follicles at the low and
high dose, and the middle dose did not result in a statistical difference from control. The
changes noted in this paper are inconsistent with each other and what is known about the
thyroid axis, as well as what is seen in the majority of the thyroid literature for perchlorate.
One other NMDR was reported for serum hormones in a two-generation rat study completed
by York et al. (2001). In this study slight increases in serum T4 or T3 were observed in low dose
groups (0.3 and 3.0 mg/kg-day with declines as the dose was increased York et al. (2001).
However, the lowest effect associated with an NMDR in this study was not the sole
determinant of the study wide NOEL. York et al. (2004) also reported an NMDR in the F1 male
generation of rats born to dams treated with 0, 0.1,1.0, 3.0, or 10.0 mg/kg-day perchlorate
from gestation day 0 to lactation day 10. A statistically significant decrease in follicular lumen
area was reported at the low (0.1 mg/kg-day), second intermediate (3.0 mg/kg-day), and high
(10 mg/kg-day) dose in F1 male pups; however, no change in follicular lumen area was found at
the 1st intermediate dose (1.0 mg/kg-day). Additionally, the standard deviation for the 1.0
mg/kg-day dose that did not show a decrease in follicular lumen was very large, almost 3-times
the SDs for the other dose groups, and data from this study was previously subjected to re-
analysis by outside experts and a significant main effect of treatment on lumen size for all doses
at PND5 was found U.S. EPA (2005b).
Gilbert and Sui (2008) evaluated the dose-response for perchlorate in a developmental study.
They assessed serum hormone levels as well as a apical functional endpoints (excitatory and
inhibitory synaptic transmission, synaptic plasticity). Monotonic dose-response patterns were
observed for the functional measures, but an NMDR was seen for serum TSH in pups on PND14,
with an increase at the intermediate dose but no change at the highest dose. This pattern was
not seen in pups at PND21 or dams, nor was it accompanied by any change in serum hormones
in pups at any age. Dam T4 was dose-dependently decreased at all dose levels and TSH
increases were limited to the high dose group.
127 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Overall, considering the large number of studies in which extensive dose-response relationships
for perchlorate are been described, the report of NMDR curves is rare. In cases in which NMDR
were reported, these were often isolated instances of low magnitude, occurred at doses above
those that were defined as the critical determinants of the study or overall chemical NOEL.
Furthermore, several findings were not consistent with what is known about the thyroid axis
biology, and in the case of York et al. (2004), a re-review of study results U.S. EPA (2005b)
concluded that the purported NMDR histopathology was in fact monotonic in nature.
4.2.4.3.3 Activation of Nuclear Receptors/ Altered Hepatic
Metabol slyhalogenated Aromatic Hydrocarbons
Case Study
A number of xenobiotics decrease circulating TH as a consequence of their induction of
metabolizing enzymes in the liver. This particular MoA involves the activation of nuclear
receptors in the liver (e.g., CAR, PXR) when the the compound or metabolite acts as a ligand.
Activation of these nuclear receptors results in an induction of a series of Phase II and III hepatic
proteins increasing TH clearance and resulting in lower circulating levels of serum TH Crofton
and Zoeller (2005; Hill et al. (1998; Capen (1997). Phase II induction includes the enzyme family
that metabolize and eliminate thyroid hormone, UDP-glucuronosyltransferase (UGT) enzymes,
cytosolic sulfotransferases, and GSH S-transferase enzymes, as well as some Phase III cellular
transporters Martignoni et al. (2006); Omiecinski et al. (2011). The increased metabolism and
excretion of T4 by UGT in the liver, results in decreased circulating T4, a required key event
that leads to an adverse outcome Crofton and Zoeller (2005; Hill et al. (1998). As with the other
thyroid disruptors, the key events and adverse outcomes associated with this mode of action
tend to follow a monotonic dose-response.
The class of chemicals known as polyhalogenated aromatic hydrocarbons (PHAHs), include a
number of groups and mixtures of chemicals that have been shown to disrupt thyroid hormone
homeostasis by interacting with nuclear receptors and stimulating induction of enzymes
Crofton and Zoeller (2005). In studies performed in rats that were exposed to individual or
mixtures of PHAHs (e.g., PCBs, dioxins, furans) there was evidence for NMDR in some endpoints
evaluated, including serum T3 and T4. Li et al. (2001; Desaulniers et al. (1997; Gray et al. (1993;
van Raaii et al. (1993). In a study by Zoeller et al. (2000) pregnant rats were treated with PCBs
throughout gestation and lactation and a number of endpoints were examined in offspring
including serum TH and myelin basic protein (MBP) mRNA in the brain. The expression levels of
MBP exhibited nonmonotonicity, but all other endpoints, followed a typical monotonic dose
response with the lowest effect level provided by the changes in T4.
128 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
There were additional studies that used an extensive number or range of doses. NMDR were
not found in an in vitro study that employed 7 concentrations of Aroclor 1254 in COS-7 cells
transfected with thyroid receptor beta (TR(3) and a concentration range from 1 pM to 100 uM
Bogazzi et al. (2003). An in vivo rat study of 18 PHAHs (including two dioxins, four
dibenzofurans, and twelve PCBs) evaluated 7-9 dose groups and dose ranges designed to cover
human exposures; no evidence of NMDR was found Crofton et al. (2005). In two additional
reports, there are some instances of NMDR at lower concentrations of exposure to one PCB
congener, PCB77. Desaulniers et al. (1997) reported increased serum T4 at 0.06, 0.6 and 6
ug/kg/day, followed by a decrease at the highest dose (60 ug/kg-day).
4.2.4.4 Results of Literature Analysis
As noted in section 4.2.3.2, four filters were used to select studies from the literature review for
more in depth analyses of NMDR results (further described in Appendix C). From the total 1153
references, 814 were deemd relevant. These 814 references yieled 2060 chemcial- studies, of
which there were 1005 mammalian and in vitro chemical-studies with 3 or more dose groups
and a control (Filter 1). Of these, only 46 NMDR chemical-studies from 42 different papers
were identified, representing 38 chemicals (Table 4.4 and Appendix C). Two important
conclusions can be made from this analysis. First, the number of NMDR reports is small
compared to the entirety of the reviewed papers (only 46 out of 1005 chemical studies).
Second, is that a large portion of the extant literature is comprised of reports with too few dose
groups to make a determination of the existence of NMDR (Appendix C.5). Almost half of the
papers (45%) used only 1 or 2 dose groups; only 6% of the in vivo chemical-studies used 5 or
more dose groups. A synopsis of each of these 46 chemical-studies (from 42 papers) and
justification for the filtering applied in each instances is provided in Appendix C. Of these 46
chemical-studies, 28 were eliminated based on the Filter 3 (effect is not the sole driver of the
study-wide LOAEL/NOAEL) yielding a total of 18 chemical-studies papers with evidence of a
nonmonotonic low dose effects on some aspect of thyroid function. Based on the criteria for
Filter 4, only 8 of these 18 chemical-studies remained. This amounts to a small percentage of
total number of chemical-studies that contained mammalian thyroid-related endpoints and 3 or
more dose groups (8 of 1005, or 0.8%) and even smaller proportion of the total number of
chemical-studies from all of published literature that was reviewed (8 of 2060, or 0.38%).
When observed, the most common nonmonotonic effect was an increase in serum T4 orT3 at
low doses followed by return to control levels (i.e., chloromethyl benzene, tamoxifen) or
declines at higher doses (tert-butyl methyl-phenol, methimazole, PCB110, PCB77). The MIEs
identified for the few chemical that produced NMDR were TPO inhibition (MMI) and nuclear
receptor activated up-regulation of hepatic metabolism (chloromethyl benzene, PCB77,
PCB110, thioxpyr). The vast majority of reports for chemicals that act via these MIEs failed to
129 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
report NMDRs for any TH related endpoints. The MIEs for the chemicals described in the three
papers were not readily discernible (propylparaben and tertbutylmethylphenol), although
tamoxifen is a well known estrogen receptor antagonist. The nature of the NMDRs of the suite
of chemicals found in this review did not readily lend themselves to categorization based on the
proposed MIE. A full description of the findings on these papers is included in Appendix C, and
each are briefly summarized below.
CCB (l-chloro-4-(chloromethyl)benzene, 0,10, 50, 250 mg/kg-day) was administered to male
and female SD rats via oral gavage daily for 28 days, beginning at 8 weeks of age (Yamasaki et
al., 2012). A slight (~20%) but statistically significant increase in serum T4 was observed at the
middle dose in female rats, with no change in male rats. No other changes in serum T3, TSH or
thyroid histopathology were noted in either sex after exposure to chloromethyl benzene. The
absence of decreases in serum T4 at high CCB concentrations is inconsistent with a proposed
hepatic mode of action for this compound. Interestingly, the authors concluded that no
endocrine-mediated effects, including thyroid dysfunction were observed in any groups of rats
treated with l-chloro-4-(chloromethyl) benzene Yamasaki et al. (2008). In this same
publication, a statistically significant monotonic dose-dependent increase in serum T4 was
reported in response for a related benzene derivative (1,3-diethyl benzene). No other
published studies on any effects, endocrine or otherwise, of this chemical could be found.
This same group of investigators Yamasaki et al. (2008), using an identical study design,
reported on the thyroid effects of oral exposure to 4,4'-butylidenebis-(2-tert-butyl-5-
methylphenol (0, 5, 25, 125 mg/kg-day). A statistically significant increase in serum T4 (15%)
was observed at the low dose, no change at the middle dose, and a decline in serum T4 (17%)
at the highest dose tested. Unlike the effects on serum T4 reported a bove for CCB exposure,
this alteration in serum hormone was detected in males but not females. Increases in TSH were
detected in both sexes at the high dose. Thyroid histopathology was performed only in the high
dose animals, and no abnormalities were noted. The authors compared these effects to PTU,
but provided no explanation of the increases in T4 at the lowest dose. No other published
reports on endocrine effects of this compound were found.
Hood et al. (1999) administered MMI or PTU (0, 1, 3, 10, 30, 100, 300 ppm) to adult male rats in
the diet for 21 days. Both MMI and PTU inhibit the synthesis enzyme TPO in the thyroid gland.
Statistically significant increases in serum total T4 (~30%, RIA) and free T3 (~25%) were
observed at the 3ppm dose level (~0.18mg/kg/day) followed by monotonic dose-dependent
declines at higher doses. A similar pattern, although less robust and below the level of
statistical significance, was also observed for PTU. Important to note in this study was the wide
dose range employed (1 - 300 mg/kg/day) and used 6 dose groups. This type of design is rare
130 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
for published in vivo studies and makes it easier to discern possible low-dose NMDR effects. No
additional studies were located for this dose of MMI or lower.
A similar profile for serum T4 was seen after a 13-week dietary exposure of young female rats
to PCB77 (3,3',4,4'-tetrachlorobiphenyl, 0,10, 100,1000, 10,0000 ppb) Desaulniers et al.
(1997). Statistically significant increases in serum T4 were observed at 10, 100, lOOOppb (20-
45%) with declines (50%) at the highest dose tested. The declines in serum T4 at the highest
dose were accompanied by slight increases in serum TSH and up-regulation of UDP-GTs in the
liver. These food concentrations correspond to dose of 0.06, 0.6 and 6 ug/kg/day and represent
some of the lowest doses for which an effect on thyroid or any other in vivo endpoint has been
reported. A related chemical, PCB28 (2,4,4'-trichlorobiphenyl) assessed in the same study did
not alter any of the thyroid endpoints assessed.
PCB110 (0, 8, 32, 48, 96 mg/kg/day) was administered alone or in a mixture contaminated with
PCB126, ip, in corn oil vehicle, to juvenile female SD rats on PND21 and 22 Li et al. (1998). As
seen with PCB77 Desaulniers et al. (1997), a statistically significant increase in serum T4 was
observed at the lowest dose, followed by a monotonic decrease in T4 at higher dose levels.
Hepatic microsomal enzymes were not induced at lower dose of PCB110, which may have
permitted mobilized T4 to reach higher serum levels. This conclusion is supported by induction
of metabolic pathways at all dose levels of the PCB110 plus PCB126 mixture where a monotonic
dose-response pattern was observed for serum T4.
Contrasting with previous studies, three additional chemicals with NMDR were identified,
which exhibited decreases in serum T4 (propylparaben) orT3 (thiazopyr, tamoxifen) at lower
doses with return to control levels or reductions at higher doses. Vo et al. (2010) reported on a
number of thyroid and reproductive endpoints for a series of para ben antimicrobial agents.
Propylparaben was administered to young male and female rats by oral gavage (0, 62.5, 250,
1000 mg/kg/day) for 20 days. A statistically significant decrease in serum T4 (42% based on a
commercial ELIZA kit) at 250 mg/kg of propylparaben was observed, but not at higher or lower
doses, and there were no effects on any other parameter assessed (thyroid weight, liver
weight, or female reproductive endpoints). The only NMDR thyroid effect reported for
propylparaben was restricted to this nonmontonic decrease in serum T4, while the other five
paraben compounds assessed resulted in effects in a number of thyroid and other reproductive
endpoints, but none of these effects was nonmonotonic. No other studies of propylparaben
were found.
An NMDR for serum T3 was reported for the pre-emergent herbicide thiazopyr Hotz et al.
(1997). Thiazopyr was administered (0,10, 30, 100,300, 1000, 3000 ppm) to adult male SD rats
for 56 days. Significant reductions in serum T3 (~20%, RIA) were detected at the mid-dose of
131 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
30ppm with increases in T3 at the highest dose tested. Effects on serum to T4 and TSH were
limited to reductions and increases, respectively, at the highest dose of 3000ppm. However,
within this study, the high dose increases in serum T3 were not consistently observed, while
increases in TSH, liver and thyroid weights were replicated at 3000 ppm. As such, a reduction in
serum T3 at the 30ppm dose level constitutes the nonmonotonic effect of concern. UGTs were
increased in liver but only at the two highest doses. A decrease in serum T3 at the mid- dose
level in the absence of effect on T4 is not consistent with an upregulation of hepatic
metabolism. No other thiazopyr studies were found.
The estrogen receptor antagonist, tamoxifen was administered (0, 10, 50, 200 M-g/kg) by oral
gavage to immature female SD rats for 20 days and thyroid glands and blood collected 24 hours
after the last dose Kim et al. (2002b). Thyroid gland weights were increased at the middle two
doses but were not significantly different from controls at the high dose. Both serum TSH and
T3 were also increased at these intermediate doses, but no change was detected in serum T4 at
any dose level. Increases in TSH are consistent with increases in thyroid weight, but not with
increases in serum T3. Another report in the literature using a very similar dosing regimen did
not replicate these observations Kennel et al. (2003). In this report male rats were unaffected
at any dose of tamoxifen (5, 30 or 200 |ag/kg/day), and females exhibited a decrease in serum
T4 with no changes in other serum markers or in thyroid gland weight.
In summary, a very limited number of publications were identified with NMDR characteristics at
the lower end of tested dose ranges. The nature of NMDR was not generally consistent with
categorization of the chemicals based on the proposed MIE. The most common nonmonotonic
effect identified was an increase in serum T4 orT3 at low doses followed by return to control
levels (i.e., chloromethyl benzene, tamoxifen) or declines at higher doses (tert-butyl methyl-
phenol, methimazole, PCB110, PCB77). Although there are no well-documented mechanistic
explanations for the small increases in serum hormones found, there are some plausible, and
testable, hypotheses that could explain the results. The small increases in THs could result from
the activation of homeostatic mechanisms to maintain euthyroid serum hormone levels,
including increased sensitivity to TSH during initial response, modulation by deiodinases,
recovery of T3 metabolites, and enterohepatic circulation (e.g., Brabant et al. (1992). Overall,
the magnitude of increase in serum TH was typically small, occurred infrequently, and may
represent the transitory adjustments of a dynamic system in flux. The biological significance of
these changes is open to debate.
132 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
4.2.4.5 Conclusions for Thyroid Studies
Both the case-study approach and literature review of TDCs identified a very small percentage
of studies with evidence of NMDR for thyroid disruption at the low end of the dose-response
curve. In approximately half of the cases where nomonotonicity was detected, it was not for an
endpoint that defined the critical effect or represented the sole determinant that would drive
the NOAEL or LOAEL for risk assessment. As discussed above, only eight instances were
identified where the NMDR profile was at the lowest dose levels reported in the study. The
reported effects were most often on serum T3 or T4 and in a direction inconsistent with the
proposed MoA identified for those chemicals. Furthermore, the NMDR was often inconsistent
with the known downstream adverse effect of primary concern, or the effect seen for this same
endpoint at higher dose levels. The declines in serum TH observed at low doses in some studies
were inconsistent with other thyroid related measures or with other reports in the literature.
Replication of many of these effects is recommended before consideration for use in regulatory
decisions. Overall, a more complete understanding of the temporal dynamics of compensatory
processes in the mutli-tissue regulation of thyroid hormones and thyroid hormone action would
aid interpretation of the biological significance of many of the NMDRs reported here.
133 | Page
-------
Draft Deliberative - Do NOT Cite, Quote, or Distribute
State of the Science: Nonmonotonic Dose Responses V7
Table 4.4: From a thyroid literature review of more than 1153 references, 1831 mammalian or in vitro chemical-studies were evaluated
for the presence of NMDR. A filtering protocol developed to help evaluate NMDR data was applied to these studies. Approximately one-
half of these studies were eliminated as they had fewer than 3 dose groups and a control. This table summarizes the 42 papers that report
NMDR curves for thyroid related endpoints. Detailed information on each of these studies is available in Appendix C. The four filtering
criteria used to determine NMDR are described in the text and the Appendix, and only 8 chemical studies from 8 separate papers passed all
filtering criteria (Filter Applied=Discuss). Each of these appears in the Appendix and is also discussed further in the text (Section 4.2.3.4).
Chemical
Species
Group Size
Doses Given
Study
Type
Filter
Applied
Reference(s)
17a-methyltestosterone
Rat
9-10
5, 20, 80 mg/kg-day
Gavage
3
Okazaki et al. (2002)
17|5-estradiol (E2)
Mouse
16-25
0.001, 0.005, 0.05, 0.15, 0.5ppm
Feed
3
Tyl et al. (2008c)
l-chloro-4-(chloromethyl)benzene
Rat
5
10, 50, 250 mg/kg-day
Gavage
Discuss
Yamasaki et al. (2012)
l-methyl-3-propylimidazole -2-
thione (PTI)
Rat
15
5,10, 25, 75 mg/kg-day
Gavage
3
Biegel et al. (1995)
2,3,7,8-tetrachlorodibenzo-p-dioxin
(TCDD)
Rat
12
6.25, 12.5, 25, 50, 100 ng/kg
Gavage
4a
Potter et al. (1986)
4,4'-butylidenebis(2-tert-butyl-5-
methylphenol)
Rat
10
5, 25,125 mg/kg-day
Gavage
Discuss
Yamasaki et al. (2008)
5-ethylidene-2-norbornene
Rat
10
52, 148, 359 ppm
Vapor
3
Ballantvne et al. (1997)
Amiodarone
In vitro
3#
0.5, 1, 5, 10, 15 nM
In vitro
4c
Freitas et al. (2011)
Chlorpyrifos
Rat
12 pups
1,10,100 mg/kg
Gavage
3
Jeong et al. (2006)
d-d-T80-prallethrin
Rat
24
120, 600, 3000 ppm
Feed
3
Seki et al. (1987)
Dibromoacetonitrile
Rat
10
0.1, 1, 10, 100 ppm
Water
3
Poon et al. (2003)
Diethylstilbestrol (DES)
Rat
10
10, 20, 40 ng/kg-day
Gavage
3
Shin et al. (2009)
Ethylenethiourea (ETU)
Rat
5
5, 25, 125, 250, 500 ppm
Feed
4c
Graham et al. (1975; Graham
et al. (1973)*
Furan
Rat
12
0.03, 0.12, 0.5, 2.0, 8.0 mg/kg
Gavage
3
Gill et al. (2010)
Hexachlorobenzene (HCB)
Rat
3-5
3, 10, 30, 100, 300, 1000 ppm
Gavage
3
van Raaii et al. (1993)
Imidazole
Rat
8-10
20, 60, 200 ppm
Water
3
Comer et al. (1985)
Methimazole (MMI)
Rat
4-5
3, 10, 30, 100, 300, 1000 ppm
Feed
Discuss
Hood et al. (1999)
Methoxychlor
Rat
9-10
20,100, 500 mg/kg
Gavage
3
Okazaki et al. (2001)
Parabens
Vo et al. (2010)
Propylparaben
Rat
10
62.5, 250, 1000 mg/kg
Gavage
Discuss
Isobutylparaben
Rat
10
62.5, 250, 1000 mg/kg
Gavage
3
134 | Page
-------
Draft Deliberative - Do NOT Cite, Quote, or Distribute
State of the Science: Nonmonotonic Dose Responses V7
Chemical
Species
Group Size
Doses Given
Study
Type
Filter
Applied
Reference(s)
Isopropylparaben
Rat
10
62.5, 250, 1000 mg/kg
Gavage
3
Butylparaben
Rat
10
62.5, 250, 1000 mg/kg
Gavage
3
Pentabromodiphenyl ether (DE-71)
Mouse
6
0.8, 4.0, 20, 100, 500 mg/kg
Gavage
4a
Fowles et al. (1994)
Pentachlorodiphenyl ethers
Rosiak et al. (1997)
PCDE 35
Rat
6 dams
25, 50,100 mg/kg-day
Gavage
4c
PCDE37
Rat
6 dams
50, 75,100 mg/kg-day
Gavage
4c
Perchlorate (CI04)
Mouse
10 dams
1 nM, ImM, ImM
Water
4c
Thuett et al. (2002a); Thuett
et al. (2002b)*
Perchlorate (CI04)
Rat
30 dams
0.3, 3.0, or 30.0 mg/kg
Water
3
York et al. (2001)
Perchlorate (CI04)
Rat
25 dams
0.1,1.0, 30,10.0 mg/kg-day
Water
4c
York et al. (2004)
Perchlorate (CI04)
Rat
5-6
10, 50, 100, 500 mg/L
Water
3
Mannisto et al. (1979)
Perchlorate (CI04 )
Rat
16-27 litters
30, 300, 1000 ppm
Water
3
Gilbert and Sui (2008)
PCB Aroclor 1254
Rat
15
5, 50, 500 ppm
Feed
4b
Collins and Capen (1980)
PCB Aroclor 1254
Rat
6 dams
1, 4, 8 mg/kg
Feed
3
Zoeller et al. (2000)
PCB110
Rat
4-11
8, 32, 48, 96 mg/kg
i.p.
Discuss
Li et al. (1998)
PCB149
Rat
5-9
8, 32, 96 mg/kg/day
i.p.
4c
Li et al. (2001)
PCB77
Rat
10
10, 100, 1000, 10000 ppb
Feed
Discuss
Desaulniers et al. (1997)
Potassium Bromate (KBr03)
Rat
50
0.02, 0.1, 0.2, 0.4 g/L
Water
3
Wolf et al. (1998)
Propazine (DACT)
Rat
15
16.7, 33.8, 67.5,135 mg/kg/day
Gavage
3
Laws et al. (2003)
Propylthiouracil (PTU)
Rat
10-14
1, 2, 3 ppm
Water
3
Gilbert et al. (2012)
Propylthiouracil (PTU)
Rat
21 dams
1, 2, 3,10 ppm
Water
3
Laslev and Gilbert (2011)
Propylthiouracil (PTU)
Rat
5
5, 15, 25 ppm
Water
3
Sawin et al. (1998)
Propylthiouracil (PTU)
Rat
12
2.5, 5, 10, 25 ppm
Water
4a
Gordon et al. (2000)
Saisentong
Rat
10
5,10,15 mg/kg-day
Gavage
4c
Zhang et al. (2010)
Tamoxifen
Rat
10
10, 50, 200 ng/kg
Gavage
Discuss
Kim et al. (2002b)
Thiazole-Zn
Rat
10
40,100, 200 mg/kg-day
Gavage
3
Yang et al. (2013)
Thiazopyr
Rat
20
10, 30, 100,300, 1000, 3000 ppm
Feed
Discuss
Hotz et al. (1997)
Triclosan
Rat
8-15
3, 30,100, 200, 300 mg/kg-day
Gavage
3
Zorrilla et al. (2009)
135 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
4.3 Human Studies and Epidemiology
4.3.1 Context for Human Studies in this iewiew
A major focus of this document (as seen in sections 4.1 and 4.2) has been on in vivo animal
data, with an emphasis on what can be learned about frequency of NMDR in reproducible
toxicology studies. Given the impact and relevance to current test rules, a particular effort has
been made to review those toxicology reports that conform to test guidelines and Good
Laboratory Practices. However, concerns about the adequacy of these tests and testing
strategies have largely been prompted by findings in the epidemiologic literature and in human
studies that show health effects of EDCs at low doses and environmentally relevant
concentrations.
Resources for this EPA review did not support a systematic literature search and review of
human studies. However, a number of recent publications and reports from international
health organizations provided relevant insights into the state of the science and were helpful in
informing the Agency's understanding of the literature. These included two separate reports in
2013 from the EFSA Scientific Committee and the European Joint Research Centre (JRC) that
provided guidance on identification, evaluation, and hazard assessments of endocrine active
and endocrine disrupting compounds EFSA (2013; JRC (2013), as well as the World Health
Organization (WHO) report, State of the Science of Endocrine Disrupting Chemicals (2012). This
WHO report is an update of the scientific knowledge related to concerns about potential
adverse health effects of endocrine disrupting chemicals on humans and wildlife and does not
focus on NMDR perse WHO (2012). In addition the Vandenberg et at. (2012) review was
helpful in identifying topics for discussion of some of the considerations listed below.
4.3.2 Interpreting Epidemiological Evidence
There are increasing numbers of reports in the epidemiological literature showing possible
health effects associated with EDCs at low doses Vandenberg et al. (2012; WHO (2012). In
these settings, low dose is usually meant to describe doses at environmentally relevant levels.
As such, epidemiological studies provide a context for evaluating and understanding
associations between low dose chemical exposures and human health, and together with other
human studies they have provided much of the basis for further exploration into evaluating
health effects at low doses. Epidemiological studies have made important contributions to
understanding the impact of various exposures on multiple health endpoints and are a critical
link between toxicological testing and evaluating adverse effects in humans.
136 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Historically, WoE approaches have been used to incorporate multiple data streams from
different study types to inform risk and hazard assessments. But currently, there is no
consensus guidance on how epidemiological or human studies might be used to shape or alter
the design of a toxicity testing study. This is primarily due to profound differences between
epidemiological and toxicological studies that both enrich and complicate the state of the
science. Some of these issues are highlighted below.
4.3.2.1 Multiple Exposure ibined or in Sequence)
Standard toxicological tests are designed to assess the impact of individual chemicals for
hypothesized health impacts using controlled experiments. Epidemiological studies, on the
other hand, attempt to find causal associations between a health outcome and an exposure.
However, in observational and non-interventional studies, those exposures typically occur
concurrent exposure to other chemical and non-chemical stressors that may or may not modify
the observed effect. In addition, the observed population may have experienced prior
exposures to the same or different sets of stressors that could also affect the nature and
magnitude of the observed effect to current exposures. Because most epidemiological studies
are not designed to include control populations, it is difficult to provide a definitive attribution
of causation to the specific exposure of interest EFSA (2013; Vandenberg et al. (2012).
Moreover, many exposures are highly correlated, introducing methodological issues; such as
collinearity, high dimensionality, and potential synergistic or inhibitory effects. This further
exacerbates the problems with treating environmental exposures observed in epidemiological
studies as single entities JRC (2013). Several advances have been made in this area and
methods have been designed to identify subsets of mixtures Gennings et al. (2010). to assess
mixtures that are also affected by a limit of detection Herring (2010). to accommodate joint
analysis of high-dimensional biomarker data Zhang et al. (2012). and to model interactions
Yeatts et al. (2010; Moser et al. (2005; Charles et al. (2002).
4.3.2.2 Exposure Assessment
Another consideration in human studies is that when they are measured, exposures are
evaluated in environmental or biological media. While these measurements are more
realistically representative of environmental exposures, in comparing these human studies to
animal toxicology studies, there are typically two kinds of challenges. First it is difficult to
extrapolate from the measured exposures to internal exposures or dose delivered to the target
organ, which are the exposures of interest in typical toxicology tests. This is particularly
important for EDCs, as they produce tissue-specific effects WHO (2012). Second, there may be
137 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
important discrepancies in how the data are measured and reported, including the way in
which measurement error is handled, this may affect how the study results are interpreted.
For example, biomarkers are increasingly used for biological assessment of exposure. When
exposure levels are low, measurement sensitivity becomes increasingly important, and often
results in a large number of measurements taken that are reported to be below the limit of
detection. Although measurements are produced both above and below this cut point, in many
cases all points below the limit of detection are reported as "not detected." The way in which
data below the limit of detection are handled in epidemiological studies is not always clearly
specified; this may introduce bias into the exposure assessment and has implications for the
understanding and identification of both low dose effects and NMDRs (Schisterman
Schisterman et al. (2006; Lubin et al. (2004; Richardson and Ciampi (2003). Approaches have
been developed for more accurate and efficient estimation when dealing with data subject to a
limit of detection Herring (2010; Schisterman and Little (2010; Vexler et al. (2008; Perkins et al.
(2007).
4.3.3 Mowing Forwa nverging Evidence
While methodological issues may create difficulties for interpreting their findings in the context
of toxicological data, population and epidemiological studies are essential to identifying
adverse effects of EDCs in human populations EFSA (2013; JRC (2013). Observational studies
allow for the study of various exposures that otherwise would be unethical to deliver in a
randomized fashion to humans. Several criteria or principles have been proposed to guide
causal inference from epidemiological studies Rothman et al. (2008; Hill (1965). These
principles include important considerations such as biologic plausibility and temporality. In
particular, the ability to see consistent results across studies of different designs and
populations is critical for inferring causality. New large longitudinal epidemiologic studies are
being undertaken to further evaluate relationships between EDCs and relevant health
endpoints Buck Louis et al. (2011b; Buck Louis et al. (2011a). Moreover, these studies are
increasingly using validated biomarkers of exposure and are allowing for the study of chemical
mixtures to further our understanding in this area.
The Endocrine Society has noted strong evidence for adverse reproductive outcomes after
exposure to EDCs for both males and females Zoeller et al. (2012; Diamanti-Kandarakis et al.
(2009). Others have also recognized that EDCs have been shown to affect a wide array of
fecundity and fertility endpoints Mendola and Buck Louis (2010; Buck Louis et al. (2006; Toft et
al. (2004). Alhough effects have been observed for several intermediate endpoints, there have
been limited data published on human exposure and fecundity endpoints, including gynecologic
disorders. However, emerging evidence shows that EDCs are associated with a higher risk of
138 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
endometriosis, including these examples: select phthalate concentrations Buck Louis et al.
(2013); urinary concentrations of benzophenone-type UV filters Kunisue et al. (2012); persistent
organochlorine pollutants Buck Louis et al. (2012a; Cooney et al. (2010); and perfluorinated
chemicals Louis et al. (2012). In addition, persistent environmental chemicals Buck Louis et al.
(2013), metals Buck Louis et al. (2012b), and PBDEs Harley et al. (2010) have been observed to
be associated with reduced fecundity in humans, and PCBs with pregnancy loss Pollack et al.
(2011). These studies overcome several important limitations from previous studies, including
prospective assessment of reproductive outcomes, reporting biomarkers of exposure, and
assessment of chemical mixtures.
4.3.4 Summary
As noted previously, this assessment did not include a systematic evaluation of the open
scientific literature or data for epidemiologic evidence for NMDR of EDCs in humans. The
authors of this Agency review, however, recognize that the data from new and emerging
epidemiology and human studies suggest a narrative that may be different from that which can
be constructed from data from toxicology tests. Clearly there is a need, within the community
of environmental health scientists, to examine and reconcile these differences, and to develop
guidance for synthesizing the evidence from and strengthening the designs of future human
studies and toxicity tests EFSA (2013; WHO (2012). Ideally, epidemiologic studies of high
quality and adequate statistical power need to be integrated with the in vitro and in vivo
laboratory aquatic and mammalian animal toxicological studies to fit into the overall regulatory
database. This is particularly important in the investigation of common MoA for chemicals
across different levels of biological organization. This approach will help identify relevant
adverse outcome pathways for the putative agent(s) and contribute to discerning the
plausibility and coherence of the associations in the findings. Although epidemiologic data are
limited, they have direct relevance to our understanding of associations between low dose
chemical exposures and human health. Consideration of the measurement process, including
limit of detection and chemical mixtures, can improve our ability to obtain unbiased estimates
of the associations between EDCs and human health outcomes to facilitate their use in risk
assessment EFSA (2013: JRC(2013: WHO (2012).
5. Conclusions
5.1 General Conclusions
The literature on NMDRs has been extensively and critically evaluated for estrogen, androgen
and thyroid hormone pathways. Based on these evaluations, the WoE supports the conclusion
that NMDRs can occur and, in many instances, are explainable in terms of basic biological
139 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
processes. NMDR are more likely to be identified after short-term exposures than in longer-
term or chronic studies. One reason for this observation is that in short term studies higher
doses may be administered without causing overt toxicity than is possible in longer term or
chronic studies. The longer term study may achieve an adverse effect at a lower exposure level
as the duration of exposure is increased; thus, administration of the higher dose producing an
NMDR in a short-term study is precluded. Another possible reason that NMDRs may be more
commonly observed in short-term studies is that the animals are exposed during one, or part of
one, stage of development, whereas long-term multigenerational studies include exposures
during the full life cycle of all of the reproductive events sensitive to estrogen pathway
disruption. Therefore, any effects observed in the latter would be a result of the cumulative
exposure over multiple life stages. Furthermore the shape of the dose response curve can
change over time in studies with repeated samples taken at multiple time points as
compensatory mechanisms alter pharmacokinetics and toxicokinetics in a tissue-specific
manner.
NMDR occur in vivo but are not common and tend to be seen at high doses. Evaluation of
NMDR across multiple studies for the same chemical often shows a lack of reproducibility. This
lack of reproducibility may be due to variations in sample size, experimental design issues,
inappropriate statistical analysis, or lack of a true NMDR. When NMDRs are present and
reproducible in in vivo studies, they are typically associated with apical adverse effects after a
high dose exposure.
The mode of action/adverse outcome pathway concepts should be used to inform the
evaluation of data to determine if an NMDR is described and is biologically plausible.
Evaluating the key events in the pathway in the context of dose response enables sound
conclusions to be drawn as to the influence of an NMDR on characterization of the adverse
outcome and its relevance for risk characterization. Minimum conditions are necessary for an
NMDR to be present. NMDR can occur when at least two opposing influences shape the curve,
with the effect of each influence dominant over a different range of doses (WHO, 2012). For
example, disruption of estrogen- and androgen-mediated endpoints can occur when
xenobiotics act as ligands for ER and AR or change the activities of enzymes involved in steroid
hormone biosynthesis or metabolic clearance. When NMDRs are present they tend to arise
from the summation or interaction of multiple processes happening to different degrees at
different points along the dose-response spectrum. Potency of the agent as well as its ability to
induce repair processes can also impact the description of dose-response and likelihood of an
NMDR.
The relevance of an NMDR to human or ecological health should be evaluated for the range of
doses or exposures over which the nonmonotonic range occurs. When some or all of the doses
140 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
or exposures associated with nonmonotonicity significantly exceed those known or expected to
occur at environmental levels of exposure, then the nonmonotonic relationship may not be
relevant to the evaluation of health effects.
Nonmonotonicity can also arise by chance. Statistical power and methods, including addressing
multiple comparisons, should be carefully evaluated. In addition, important studies should be
repeated whenever feasible, as independent replication of NMDR is considered to provide the
most scientific support. When replication is not possible, ancillary studies that provide data on
pharmacokinetics and the MoA can provide insight into expected dose-response behaviors.
One biological process that theoretically can give rise to NMDR involves compensatory (or
adaptive) responses. Organisms have evolved a wide range of mechanisms to maintain
homeostatic conditions conducive to normal function(s) when impacted by stressors. The
vertebrate endocrine system has been comparatively well-studied in terms of homeostasis
Nichols et al. (2011). Various feedback mechanisms operate at different biological levels of
organization to maintain a dynamic homeostasis supporting normal reproduction and
development. Given the very active nature of feedback signaling in biological systems,
maintenance of homeostatic conditions through compensatory responses could be the basis for
nonmonotonic features of dose-response relationships in some types of studies.
Fish exposed to exogenous chemicals exhibited several instances wherein changes in gene
expression were nonmonotonic. There also were examples of either plasma steroid
concentrations or ex vivo steroid production exhibiting nonmonotonic relationships with dose.
However, in time-course studies, there were no examples of NMDR for endpoints closer in
biological level of organization to the apical endpoints examined. These data suggest that
responses at molecular and biochemical levels, such as those involved in feedback regulation of
homeostasis (including compensation), may more commonly exhibit nonmonotonic
characteristics than more integrated, downstream, apical, endpoints.
This review did not include a systematic evaluation of the epidemiologic literature for NMDRs in
human estrogen, androgen, or thyroid systems. However, data from epidemiology and human
studies may provide a perspective not captured by traditional toxicologic testing. Thus, there is
a need within the community of environmental health scientists to develop guidance for
synthesizing evidence from various study types and to strengthen the design of future human
studies and next generation toxicity tests EFSA (2013; WHO (2012). Ideally, epidemiologic
studies of high quality and adequate statistical power should be integrated with the in vitro and
in vivo aquatic and mammalian testing data for common MoA for chemicals across different
levels of biological organization.
141 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Conclusions:
1. NMDRs do occur in estrogen, androgen, and thyroid systems as evidenced in
ecological and mammalian studies.
2. NMDRs are not unexpected in vitro particularly when evaluating high dose
levels and/or lower-order biological endpoints in estrogen androgen or
thyroid systems.
3. NMDRs are not commonly identified in estrogen, androgen, or thyroid
systems in vivo and are rarely seen in apical endpoints after low-dose and/or
long-term exposure.
4. The nature of a dose response will vary over time, and nonmonotonicity due
to compensation may be observed.
5. NMDRs observed in endocrine endpoints may be biologically relevant and
should be evaluated in context with the totality of the available scientific data,
including epidemiologic and human studies.
6. There is currently no reproducible evidence that the early key events involved
in the expression of NMDRs that are identified at low dose are predictive of
adverse outcomes that may be seen in humans or wildlife populations for
estrogen, androgen or thyroid endpoints.
7. Therefore, current testing strategies are unlikely to mischaracterize, as a
consequence of NMDR, a chemical that has the potential for adverse
perturbations of the estrogen, androgen or thyroid pathways.
5.1.1 Overall Conclusions: Estrogen, Androgen, and Thyroid
Effects with NMDRs were frequently observed in studies evaluating estrogenic or androgenic
activity of chemicals in vitro. However, they tend to be indentified when high concentrations
are used in the medium; these concentrations projected to in vivo scenarios would be too high
to be environmentally relevant or attainable. Therefore, while the data from in vitro studies
provide valuable insights into the endocrine activity and potential mechanisms of action of
chemicals, the relevance of the dose response curves from these studies to understanding dose
response relationships in vivo is highly questionable.
When NMDRs were seen in vivo for estrogen or androgen signaling pathway effects, the
observations were often from large multigenerational studies wherein numerous
measurements were taken and a great many comparisons were made across control and
exposed groups of animals over several generations. Given the large number of measurements
and comparisons made in these types of studies, one would expect to see some NMDRs by
142 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
chance. To determine biological relevance one would generally weight more heavily results
from experiments that have been replicated or explained mechanistically.
In some cases, NMDR were observed in non-multigeneration studies, but they were not
replicated in other studies using similar exposure scenarios. This was often attributed to small
sample sizes, questionable experimental designs (e.g., route of exposure not relevant to
environmental scenarios), and/or inappropriate statistical analyses (e.g., not accounting for
litter effects). Reproducible effects following an NMDR were seen for several in vivo studies for
adverse, downstream phenotypic effects. In these cases, the effects were usually at high and
not low doses.
In general, NMDR for the estrogen and androgen signaling pathways, while identified for some
apical endpoints, are more likely with genomic and hormonal measures than with phenotypic
effects and endpoints such as fertility, histopathology or malformations. These types of
measures tend to be more variable and less reproducible, which could explain why similar
findings often were not observed in studies with comparable exposures.
For a specific endocrine mechanism of toxicity, the shapes of the dose response curves will be
tissue- and cell-type specific due to the complexity involved in the tissue-specific molecular
events modulating the different responses to a chemical that disrupts estrogen or androgen
signaling. In the case for estrogens, responses are frequently tissue specific due to tissue
specific ratio of ERa and ER|3, which may antagonize one another when activated. Likewise,
there may be different levels of the two estrogen receptors, corepressors and coactivators that
regulate effects on gene expression, metabolism, distribution and elimination. In addition, the
set of compounds for which NMDR for the estrogen hormone system were identified were
most frequently synthetic hormones that were designed to have endocrine activity. Some of
the reasons for tissue specificity for estrogens are also applicable for the androgen signaling
pathway. Different androgen sensitive tissues have different corepressors and coactivators
that regulate the effects of the chemical's ability to alter gene expression and metabolism,
distribution and elimination of the chemical can vary from tissue to tissue.
Studies conducted in mammalian models in vivo rarely resulted in estrogen or thyroid signaling
pathway effects that were NMDR. Studies conducted in mammalian models in vivo resulted in
androgen signaling pathway effects that were nonmonotonic more often than reported for the
estrogen or thyroid signaling pathways. Furthermore, when NMDRs occurred in these hormone
systems they often were not reproducible within or across studies.
In the review of the mammalian, fish, and amphibian literature for xenobiotics that have been
shown to perturb the thyroid hormone system, evidence of NMDR for thyroid disruption was
143 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
found in a small number of instances. In fish and amphibians, the appearance of NMDR was
often confounded by toxicity induced at high doses, inadequate temporal sampling, or from
variable and inconsistent measurements. In mammals, when nomonotonicity was detected, it
was often not the critical effect or sole determinant that would determine the NOAEL or LOAEL.
In the instances wherein the NMDR was at the lowest dose levels, the reported effects were on
serum T3 or T4 and were inconsistent with the proposed MoA identified for those chemicals. In
addition, the declines in serum TH observed at low doses in some studies were inconsistent
with other thyroid related measures, and suggest that there are observations that require
replication before being given further consideration. Overall, a more complete understanding
of the temporal dynamics of compensatory processes in thyroid hormone regulation and
activity would assist interpretation of the biological relevance of NMDR when they are
identified.
5.2 iral Scientific Questions and Answers
We identified three central scientific questions to be addressed in the current state of the
science assessment. The central scientific questions, and summarized responses are these.
5.2.1 Do nonmonotonic dose responses (NMDRs) exist for chemicals and if so
under what conditions do they occur?
Yes, we concluded that exposures to chemicals can result in NMDRs for specific endpoints.
NMDRs arise from complex relationships between the dose of toxicant at its target site and the
effect of interest WHO (2012). NMDRs are biologically plausible and can arise when the
biological system that is activated in response to toxicant exposure consists of at least two
activities that can act in opposition to each other Conolly and Lutz (2004). WHO (2012). We
determined that NMDRs are more frequently identified in in vitro studies, high dose range
studies, or short-term studies. Assays that provide data at a lower level of biological
organization (such as proteomics or transcriptomics) are more likely to identify NMDRs than
studies that provide data on apical adverse events further downstream from the molecular
initiating event. Reproducibility of NMDRs is important in establishing plausibility of a response
and its potential applicability as part of the hazard characterization. Factors that influence
reproducibility include:
¦ Study design - dose selection, sample size, organism strain, diet, housing environment,
statistical methods;
¦ Robustness of physiology - physiologic compensation producing changes in slope; and
¦ Competing processes- induction of metabolism, repair, or independent mechanisms.
144 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
5.2.2 Do NMDRs capture adverse effects that are not captured using our
current chemical testing strategies (i.e., false negatives)?
There are certainly adverse biological changes that may occur in a nonmonotonic manner that
would not be captured using current testing strategies. No testing strategy is able to assess all
potential adverse effects, for all biological systems, in all tissues, for all species, in all
developmental time points.
As the work group progressed with its review and discussions of the science, it became clear
that our second question needed to be further defined so that it could be more accurately and
fully answered in the context of the science as evaluated. Thus, question 2 was further
expanded for clarification.
¦ Are there adver cts with NMDRs that are not being identified using the current
chemical testing strategies?
¦ Are there NMDRs for adverse effects below the no observed adverse effect levels
(NOAELS) or benchmark doses (B :rived from the current testing strategies?
¦ Do EPA chemical testing strategies detect relevant adverse effects for chemicals which
produce NMDR for specific end points?
Chemicals that operate through endocrine modes of action (MoA) have multiple targets across
organs, tissues, and cellular systems in various species, and across all life stages. It is not
possible or feasible for chemical testing to measure or analyze all possible endpoints for all
chemical MoA in all tissues. The objective of USEPA's chemical testing strategy is not to identify
all possible adverse effects, but rather, to identify sensitive endpoints relevant to human or
ecological health, providing confidence that adverse effects are not being induced at dose
levels below what was determined to be a NOAEL.
For estrogen, androgen or thyroid MoA that provide adequate information to make an
assessment, our evaluation shows that there is not sufficient evidence of NMDRs for adverse
effects below the NOAELS or BMD derived from the current testing strategies. For some MoA,
however, the scientific database remains too limited to conclude this with certainty.
While there are biological changes that may occur in a nonmonotonic manner in the low dose
region, our review indicates that reproducible NMDRs for adverse effects occur in the high dose
region and not the low dose region of the dose response curve. Thus, the current testing
approaches do not fail to identify or establish appropriate NOAELS in the low rose range of
exposure, even if not all effects for every chemical are identified. The extensive evaluation
conducted in the present review, as well as almost two decades of experience with screening
145 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
assays for hazard identification, indicate that these assays do not fail to detect chemicals with
endocrine activity for the estrogen or androgen hormone systems. Dose response assessment
is not an issue for screening assays. NMDRs would be problematic only if a chemical with
estrogen, androgen, or thyroid activity produced an effect in vivo at a dose below those used in
screening, and the chemical had no effect on estrogen, androgen, or thyroid related endpoints
at the higher screening dosage levels. Although such NMDRs have been hypothesized, they
have not been demonstrated reproducibly, and none were found in the present evaluation.
Our assessment of the adequacy of the current testing assays concludes that a number of
standardized short- and long-term assays are sensitive in detecting chemicals that interfere
with the estrogen, androgen and thyroid signaling pathways. Specifically, the EDSP screening
battery can detect disruption of these pathways using combined in vitro and in vivo assays in
mammalian and aquatic models. Standard multigenerational test guidelines have measures
that are sensitive to disruption of the estrogen and androgen signaling pathways. While these
studies are considered the current standard for assessing the potential of a chemical to be a
reproductive toxicant and for use in setting NOAELS, they are not without limitations. USEPA
testing strategies are reviewed periodically to assure they are incorporating the most sensitive
and biologically relevant endpoints.
Further, if an objective of testing is to define the shape of the dose response curve and thereby
identifying potential NMDRs, then the three treated groups and a control group used in current
guideline studies may not be sufficient for this purpose. Modifications that could lead to a
more clearly defined dose response characterization and increase the statistical power to
detect low dose effects may be appropriate in specific instances.
5.2.3 Do NMDRs prowicle key information that would alter EPA's current
weight of evidence conclusions and risk assessment determinations,
either qualitatively or quantitatively?
Data from studies in which NMDRs are identified may be biologically relevant and as such
should be evaluated in context with the totality of the available scientific data in weight of
evidence (WoE) conclusions and risk assessment determinations. These data should be
considered and analyzed, as all data are, and factored into the WoE based on standard criteria
including, but not limited to, conduct of the studies, representation of biological processes that
are relevant to the evaluation, biological plausibility, and reproducibility. NMDRs can have
impact on both qualitative and quantitative risk assessments, but cannot be considered in
isolation from other data for the chemical and biological response being considered.
146 | Page
-------
Draft Deliberative
5.3 Summary
State of the Science: Nonmonotonic Dose Responses V7
NMDR after exposure to xenobiotics occur in estrogen, androgen, and thyroid hormone
systems but are generally not common. Where NMDRs were observed, they were often at high
dose. Biological endpoints closest to the molecular initiating event were more likely to identify
a nonmonotonic point of inflection than those effects further downstream, including the apical
adverse outcomes. The goal of chemical testing is to identify the potential for hazard after
exposure to the xenobiotic of concern, not to identify and describe 100% of all the possible
biological effects. As such, the current testing approaches perform this function successfully
and, based on the current evaluation, are unlikely to mischaracterize a chemical that has the
potential to adversely perturb the endocrine system due to an NMDR.
147 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
6, Bibliographf
Linked Citations
Akingbemi. BT; Braden, TP; Kemppainen, BW; Hancock. KD; SherrilI. JD; Cook. SJ; He. X; Supko,
^(2007). Exposure to phytoestrogens in the perinatal period affects androgen
secretion by testicular Leydig cells in the adult rat. Endocrinology 148: 4475-4488.
http://dx.doi.org/10.1210/en.2007-Q327.
Akingbemi. BT; Ge. R; Klinefelter. GR; Zirkin, BR; Hardy, MP. (2004). Phthalate-induced Leydig
cell hyperplasia is associated with multiple endocrine disturbances. PNAS 101: 775-780.
http://dx.doi.org/10.1073/pnas.03Q5977101.
Akingbemi. BT; Youker, RT; Sottas. CM; Ge. R; Katz, E; Klinefelter. GR; Zirkin. BR; Hardy, MP.
(2001). Modulation of rat Leydig cell steroidogenic function by di(2-
ethylhexyl)phthalate. Biol Reprod 65: 1252-1259.
http://dx.doi.Org/10.1095/biolreprod65.4.1252.
Alarid, ET. (1999). Proteasome-Mediated Proteolysis of Estrogen Receptor: A Novel Component
in Autologous Down-Regulation. Mol Endocrinol 13:1522-1534.
http://dx.doi.Org/10.1210/me.13.9.1522.
Allan, GF; Tannenbaum, P; Sbriscia, T; Linton, 0; Lai, MT; Haynes-Johnson, D; Bhattacharjee, S;
Zhang, X; Sui, Z; Lundeen, SG. (2007). A selective androgen receptor modulator with
minimal prostate hypertrophic activity enhances lean body mass in male rats and
stimulates sexual behavior in female rats. Endocrine 32: 41-51.
http://dx.doi.org/10.1007/sl2020-007-90Q5-2.
Andersen, ME: Dennison, JE; Thomas, RS; Conolly, RB. (2005). New directions in incidence-dose
modeling [Review]. Trends Biotechnol 23: 122-127.
http://dx.doi.Org/10.1016/i.tibtech.2005.01.007.
Andrade, AJ; Grande, SW; Talsness, CE; Gericke, C; Grote, K; Golombiewski, A: Sterner-Kock, A:
Chahoud, I. (2006a). A dose response study following in utero and lactational exposure
to di-(2-ethylhexyl) phthalate (DEHP): Reproductive effects on adult male offspring rats.
Toxicology 228: 85-97. http://dx.doi.Org/10.1016/i.tox.2006.08.020.
Andrade, AJ: Grande, SW: Talsness, CE: Grote, K; Chahoud, I. (2006b). A dose-response study
following in utero and lactational exposure to di-(2-ethylhexyl)-phthalate (DEHP): Non-
monotonic dose-response and low dose effects on rat brain aromatase activity.
Toxicology 227: 185-192. http://dx.doi.Org/10.1016/i.tox.2006.07.022.
Andrade, AJ: Grande, SW: Talsness, CE: Grote, K; Golombiewski, A: Sterner-Kock, A: Chahoud, I.
(2006c). A dose-response study following in utero and lactational exposure to di-(2-
ethylhexyl) phthalate (DEHP): Effects on androgenic status, developmental landmarks
and testicular histology in male offspring rats. Toxicology 225: 64-74.
http://dx.doi.Org/10.1016/i.tox.2006.05.007.
Andrews, P; Freyberger, A: Hartmann, E; Eiben, R; Loof, I: Schmidt, U; Temerowski, M; Folkerts,
A: Stahl, B; Kayser, M. (2002). Sensitive detection of the endocrine effects of the
estrogen analogue ethinylestradiol using a modified enhanced subacute rat study
148 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
protocol (OECD Test Guideline no. 407). Arch Toxicol 76: 194-202.
http://dx.doi.org/10.1007/s00204-002-Q337-7.
Ankley, GT; Bencic, DC; Breen, MS; Collette, TW; Conolly, RB; Denslow, ND; Edwards. SW;
Ekman, PR; Garcia-Reyero, N; Jensen. KM; Lazorchak, JM; Martinovic. D; Miller. DH;
Perkins. EJ; Orlando. EF; Villeneuve, PL; Wang. RL; Watanabe, KH. (2009a). Endocrine
disrupting chemicals in fish: developing exposure indicators and predictive models of
effects based on mechanism of action. Aquat Toxicol 92:168-178.
http://dx.doi.Org/10.1016/i.aquatox.2009.01.013.
Ankley, GT; Bencic. PC; Cavallin, JE; Jensen. KM; Kahl, MP; Makynen, EA; Martinovic. P;
Mueller. NP; Wehmas. LC; Villeneuve. PL. (2009b). Pynamic nature of alterations in the
endocrine system of fathead minnows exposed to the fungicide prochloraz. Toxicol Sci
112: 344-353. http://dx.doi.org/10.1093/toxsci/kfp227.
Ankley, GT; Bennett. RS; Erickson, RJ; Hoff, PJ; Hornung, MW; Johnson. RP; Mount. PR; Nichols.
JW; Russom. CL; Schmieden PK; Serrrano, JA; Tietge, JE; Villeneuve. PL. (2010). Adverse
outcome pathways: a conceptual framework to support ecotoxicology research and risk
assessment [Review]. Environ Toxicol Chem 29: 730-741.
http://dx.doi.org/10.10Q2/etc.34.
Ankley, GT; Jensen, KM; Purhan, EJ; Makynen, EA; Butterworth, BC; Kahl, MP; Villeneuve, PL;
Linnum, A; Gray, LE; Cardon, M; Wilson, VS. (2005). Effects of two fungicides with
multiple modes of action on reproductive endocrine function in the fathead minnow
(Pimephales promelas). Toxicol Sci 86: 300-308. http://dx.doi.org/10.1093/toxsci/kfi202.
Ankley, GT; Jensen, KM; Kahl, MP; Korte, JJ; Makynen, EA. (2001). Pescription and evaluation of
a short-term reproduction test with the fathead minnow (Pimephales promelas).
Environ Toxicol Chem 20: 1276-1290.
Ankley, GT: Jensen, KM: Kahl, MP: Makynen, EA: Blake, LS; Greene, KJ; Johnson, RP: Villeneuve,
PL (2007). Ketoconazole in the fathead minnow (Pimephales promelas): reproductive
toxicity and biological compensation. Environ Toxicol Chem 26: 1214-1223.
Ankley, GT: Jensen, KM: Makynen, EA: Kahl, MP: Korte, JJ: Hornung, MW: Henry, TR; Penny, JS;
Leino, RL: Wilson, VS; Cardon, MC; Hartig, PC: Gray, LE. (2003). Effects of the androgenic
growth promoter 17-beta-trenbolone on fecundity and reproductive endocrinology of
the fathead minnow. Environ Toxicol Chem 22: 1350-1360.
Ankley, GT: Kahl, MP: Jensen, KM: Hornung, MW: Korte, JJ: Makynen, EA: Leino, RL. (2002).
Evaluation of the aromatase inhibitor fadrozole in a short-term reproduction assay with
the fathead minnow (Pimephales promelas). Toxicol Sci 67: 121-130.
Ashby, J: Tinwell. H; Haseman, J. (1999). Lack of effects for low dose levels of bisphenol A and
diethylstilbestrol on the prostate gland of CF1 mice exposed in utero. Regul Toxicol
Pharmacol 30:156-166. http://dx.doi.org/10.10Q6/rtph.1999.1317.
Ashby, J: Tinwell, H; Lefevre, PA: Joiner, R; Haseman, J. (2003). The effect on sperm production
in adult Sprague-Pawley rats exposed by gavage to bisphenol A between postnatal days
91-97. Toxicol Sci 74:129-138. http://dx.doi.org/10.1093/toxsci/kfg093.
Axelstad, M; Hansen, PR: Boberg, J: Bonnichsen, M; Nellemann. C; Lund, SP; Hougaard, KS;
Hass, U. (2008). Pevelopmental neurotoxicity of Propylthiouracil (PTU) in rats:
Relationship between transient hypothyroxinemia during development and long-lasting
behavioural and functional changes. Toxicol Appl Pharmacol 232: 1-13.
http://dx.doi.Org/10.1016/i.taap.2008.05.020.
149 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Ballantyne, B; Norris, JC; Dodd, DE; Klonne, PR; Losco, PE; Neptun, DA; Price. SC; Grasso, P.
(1997). Short-term and subchronic repeated exposure studies with 5-ethylidene-2-
norbornene vapor in the rat. J Appl Toxicol 17:197-210.
http://dx.doi.org/10.1002/(SICI )1099-1263( 199707)17:4<197::AI D-J AT429>3.0.CQ:2-L.
Bao, AM: Man. XM; Guo. XJ; Dong. HB; Wang. FQ; Sun. H; Wang. YB; Zhou. ZM; Sha. JH. (2011).
Effects of di-n-butyl phthalate on male rat reproduction following pubertal exposure.
Asian J Androl 13: 702-709. http://dx.doi.org/10.1038/aia.2Qll.76.
Bayley, M;Junge, M; Baatrup, E. (2002). Exposure of juvenile guppies to three antiandrogens
causes demasculinization and a reduced sperm count in adult males. Aquat Toxicol 56:
227-239.
Bayley, M; Larsen, PF; Baekgaard, H; Baatrup. E. (2003). The effects of vinclozolin, an anti-
androgenic fungicide, on male guppy secondary sex characters and reproductive
success. Biol Reprod 69: 1951-1956. http://dx.doi.org/10.1095/biolreprod.103.017780.
Beltrame. D; di Salle. E; Giavini, E; Gunnarsson, K; Brughera, M. (2001). Reproductive toxicity of
exemestane, an antitumoral aromatase inactivator, in rats and rabbits. Reprod Toxicol
15: 195-213. http://dx.doi.org/10.1016/S0890-6238(01)00120-4.
Bernal, J. (2011). Thyroid hormone transport in developing brain [Review]. Curr Opin Endocrinol
Diabetes Obes 18: 295-299. http://dx.doi.org/10.1097/MED.0b013e32834a78b3.
Bhalla, US; Ram. PT; Iyengar, R. (2002). MAP kinase phosphatase as a locus of flexibility in a
mitogen-activated protein kinase signaling network. Science 297: 1018-1023.
http://dx.doi.org/10.1126/science.1068873.
Bhasin, S; Woodhouse, L; Casaburi, R; Singh, AB; Bhasin, D; Berman, IN; Chen, X; Yarasheski, KE;
Magliano, L; Dzekov, C; Dzekov, J; Bross, R; Phillips, J; Sinha-Hikim, I; Shen, R; Storer, TW.
(2001). Testosterone dose-response relationships in healthy young men. Am J Physiol
Endocrinol Metab 281: E1172-E1181.
Bhasin. S; Woodhouse. L; Casaburi. R; Singh. AB; Mac. RP; Lee. M; Yarasheski. KE; Sinha-Hikim. I;
Dzekov. C; Dzekov. J; Magliano. L; Storer. TW. (2005). Older men are as responsive as
young men to the anabolic effects of graded doses of testosterone on the skeletal
muscle. J Clin Endocrinol Metab 90: 678-688. http://dx.doi.org/10.1210/ic.2004-1184.
Biegel. LB; Cook. JC; Hurtt. ME; O'Connor. JC. (1998a). Effects of 17 beta-estradiol on serum
hormone concentrations and estrous cycle in female Crl:CD BR rats: Effects on parental
and first generation rats. Toxicol Sci 44: 143-154.
http://dx.doi.org/10.1006/toxs.1998.2469.
Biegel. LB; Cook. JC; O'Connor. JC; Aschiero. M; Arduengo. AJ; Slone. TW. (1995). Subchronic
toxicity study in rats with l-methyl-3-propylimidazole-2-thione (PTI): effects on the
thyroid. Fundam Appl Toxicol 27: 185-194.
Biegel. LB; Flaws. JA; Hirshfield. AN; O'Connor. JC; Elliott. GS; Ladies. GS; Silbergeld. EK; Van Pelt.
CS; Hurtt. ME; Cook. JC; Frame. SR. (1998b). 90-day feeding and one-generation
reproduction study in Crl:CD BR rats with 17 beta-estradiol. Toxicol Sci 44: 116-142.
http://dx.doi.org/10.1006/toxs.1998.2468.
Birnbaum. LS. (2012). Environmental chemicals: evaluating low-dose effects [Editorial]. Environ
Health Perspect 120: A143-A144. http://dx.doi.org/10.1289/ehp.1205179.
Birnbaum. LS. (2013). State of the science of endocrine disruptors. Environ Health Perspect 121:
A107. http://dx.doi.org/10.1289/ehp.1306695.
150 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Blake. HH; Henning, SJ. (1985). Effect of propylthiouracil dose on serum thyroxine, growth, and
weaning in young rats. Am J Physiol 248: R524-R530.
Blystone, C; Kissling, G; Bishop. J; Chapin, R; Wolfe. G; Foster. P. (2010). Determination of the
di-(2-ethylhexyl) phthalate NOAEL for reproductive development in the rat: importance
of the retention of extra animals to adulthood. Toxicol Sci 116: 640-646.
http://dx.doi.org/10.1093/toxsci/kfql47.
Blystone, CR; Furr, J; Lambright, CS; Howdeshell, KL; Ryan, BC; Wilson. VS; Leblanc, GA; Gray, LE.
(2007a). Prochloraz inhibits testosterone production at dosages below those that affect
androgen-dependent organ weights or the onset of puberty in the male Sprague Dawley
rat. Toxicol Sci 97: 65-74. http://dx.doi.org/10.1093/toxsci/kfm004.
Blystone, CR; Lambright. CS; Howdeshell. KL; Furr. J; Sternberg. RM; Butterworth, BC; Durhan,
EJ; Makynen, EA; Ankley, GT; Wilson. VS; Leblanc. GA; Gray, LE. (2007b). Sensitivity of
fetal rat testicular steroidogenesis to maternal prochloraz exposure and the underlying
mechanism of inhibition. Toxicol Sci 97: 512-519.
http://dx.doi.org/10.1093/toxsci/kfm055.
Bocquel, MT; Kumar, V; Strieker, C; Chambon, P; Gronemeyer, H. (1989). The contribution of
the N- and C-terminal regions of steroid receptors to activation of transcription is both
receptor and cell-specific. Nucleic Acids Res 17: 2581-2595.
Boettcher, M; Kosmehl, T; Braunbeck, T. (2011). Low-dose effects and biphasic effect profiles: is
trenbolone a genotoxicant? Mutat Res 723: 152-157.
http://dx.doi.Org/10.1016/i.mrgentox.2011.04.012.
Bogazzi, F; Raggi, F; Ultimieri, F; Russo, D; Campomori, A; Mckinney, JD; Pinchera, A; Bartalena,
L; Martino, E. (2003). Effects of a mixture of polychlorinated biphenyls (Aroclor 1254) on
the transcriptional activity of thyroid hormone receptor. J Endocrinol Invest 26: 972-978.
Boobis, AR; Cohen, SM; Dellarco. V; McGregor, D; Meek, ME: Vickers, C; Willcocks, D; Farland,
W. (2006). IPCS framework for analyzing the relevance of a cancer mode of action for
humans. Crit Rev Toxicol 36: 781-792. http://dx.doi.org/10.1080/1040844060Q977677.
Boobis, AR: Doe, JE; Heinrich-Hirsch. B; Meek, ME: Munn, S; Ruchirawat, M; Schlatter, J: Seed, J:
Vickers, C. (2008). IPCS framework for analyzing the relevance of a noncancer mode of
action for humans [Review]. Crit Rev Toxicol 38: 87-96.
http://dx.doi.ore/10.1080/10408440701749421.
Brabant, G; Bergmann, P; Kirsch, CM: Kohrle. J: Hesch, RD; von Zur Muhlen. A. (1992). Early
adaptation of thyrotropin and thyroglobulin secretion to experimentally decreased
iodine supply in man. Metabolism 41:1093-1096. http://dx.doi.org/10.1016/0Q26-
0495(92)90291-H.
Bretz, F; Hothorn, LA. (2003). Statistical analysis of monotone or non-monotone dose-response
data from in vitro toxicological assays. Altern Lab Anim 31 Suppl 1: 81-96.
Brucker-Davis. F. (1998). Effects of environmental synthetic chemicals on thyroid function
[Review]. Thyroid 8: 827-856.
Brucker-Davis, F; Thayer, K; Colburn, T. (2001). Significant effects of mild endogenous hormonal
changes in humans: considerations for low-dose testing [Review]. Environ Health
Perspect 1: 21-26.
Buck Louis, GM; Chen, Z; Peterson, CM: Hediger, ML: Croughan, MS: Sundaram, R; Stanford, JB;
Varner, MW; Fujimoto, VY; Giudice, LC; Trumble, An: Parsons, PJ; Kannan, K. (2012a).
151 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Persistent Lipophilic Environmental Chemicals and Endometriosis: The ENDO Study.
Environ Health Perspect 120: 811-816. http://dx.doi.org/10.1289/ehp.1104432.
Buck Louis. GM; Hediger, ML; Peterson. CM; Croughan, M; Sundaram, R; Stanford. J; Chen. Z;
Fujimoto. VY; Varner, MW; Trumble, A; Giudice, LC; Group. ESW. (2011a). Incidence of
endometriosis by study population and diagnostic method: the ENDO study. Fertil Steril
96: 360-365. http://dx.doi.Org/10.1016/i.fertnstert.2011.05.087.
Buck Louis. GM; Lynch, CD; Cooney, MA. (2006). Environmental influences on female fecundity
and fertility [Review]. Seminars in Reproductive Medicine 24:147-155.
http://dx.doi.org/10.1055/s-2006-944421.
Buck Louis. GM; Schisterman, EF; Sweeney, AM; Wilcosky, TC; Gore-Langton, RE; Lynch, CD;
Boyd Barr, D; Schrader, SM; Kim, S; Chen, Z; Sundaram, R. (2011b). Designing
prospective cohort studies for assessing reproductive and developmental toxicity during
sensitive windows of human reproduction and development-the LIFE Study. Paediatric
and Perinatal Epidemiology (Online) 25: 413-424. http://dx.doi.org/10.1111/j. 1365-
3016.2011.01205.x.
Buck Louis, GM; Sundaram, R; Schisterman, EF; Sweeney, AM; Lynch, CD; Gore-Langton, RE;
Chen, Z; Kim, S; Caldwell, KL; Barr, DB. (2012b). Heavy metals and couple fecundity, the
LIFE Study. Chemosphere 87:1201-1207.
http://dx.doi.Org/10.1016/i.chemosphere.2012.01.017.
Buck Louis, GM; Sundaram, R; Schisterman, EF; Sweeney, AM; Lynch, CD; Gore-Langton, RE;
Maisog, J; Kim, S; Chen, Z; Barr, DB. (2013). Persistent Environmental Pollutants and
Couple Fecundity: The LIFE Study. Environ Health Perspect 121: 231-236.
http://dx.doi.org/10.1289/ehp.12Q5301.
Bulayeva, NN; Watson, CS. (2004). Xenoestrogen-induced ERK-1 and ERK-2 activation via
multiple membrane-initiated signaling pathways. Environ Health Perspect 112: 1481-
1487.
Cagen, SZ; Waechter, JM; Dimond, SS; Breslin, WJ; Butala, JH; Jekat, FW; Joiner, RL; Shiotsuka,
RN; Veenstra, GE; Harris, LR. (1999a). Normal reproductive organ development in CF-1
mice following prenatal exposure to bisphenol A. Toxicol Sci 50: 36-44.
http://dx.doi.Org/10.1093/toxsci/50.l.36.
Cagen, SZ: Waechter, JM: Dimond, SS: Breslin, WJ: Butala, JH: Jekat, FW: Joiner, RL: Shiotsuka,
RN: Veenstra, GE: Harris, LR. (1999b). Normal reproductive organ development in Wistar
rats exposed to bisphenol A in the drinking water. Regul Toxicol Pharmacol 30: 130-139.
http://dx.doi.org/10.1006/rtph.1999.1340.
Caldwell, DJ; Mastrocco, F; Anderson, PD; Lange. R; Sumpter, JP. (2012). Predicted-no-effect
concentrations for the steroid estrogens estrone, 17|3-estradiol, estriol, and 17a-
ethinylestradiol. Environ Toxicol Chem 31: 1396-1406.
http://dx.doi.org/10.10Q2/etc.1825.
Caldwell, DJ: Mastrocco, F; Hutchinson, TH; Lange, R; Heijerick, D; Janssen, C; Anderson, PD:
Sumpter, JP. (2008). Derivation of an aquatic predicted no-effect concentration for the
synthetic hormone, 17 alpha-ethinyl estradiol. Environ Sci Technol 42: 7046-7054.
Caldwell, DJ: Mastrocco, F; Nowak, E; Johnston, J: Yekel, H; Pfeiffer, D; Hoyt, M; Duplessie, BM;
Anderson, PP. (2010). An assessment of potential exposure and risk from estrogens in
drinking water. Environ Health Perspect 118: 338-344.
http://dx.doi.org/10.1289/ehp.0900654.
152 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Capen, CC. (1997). Mechanistic data and risk assessment of selected toxic end points of the
thyroid gland [Review]. Toxicol Pathol 25: 39-48.
http://dx.doi.org/10.1177/0192623397025001Q9.
Carlsson, G; Norrgren, L. (2007). The impact of the goitrogen 6-propylthiouracil (PTU) on West-
African clawed frog (Xenopus tropicalis) exposed during metamorphosis. Aquat Toxicol
82: 55-62. http://dx.doi.Org/10.1016/i.aquatox.2007.01.005.
Chang. EC; Frasor, J; Komm, B; Katzenellenbogen, BS. (2006). Impact of estrogen receptor beta
on gene networks regulated by estrogen receptor alpha in breast cancer cells.
Endocrinology 147: 4831-4842. http://dx.doi.org/10.1210/en.2006-Q563.
Chapin, RE; Adams. J; Boekelheide, K; Gray, LE. Jr.; Hayward, SW; Lees. PS; Mclntyre, BS; Portier,
KM; Schnorr, TM; Selevan, SG; Vandenbergh, JG; Woskie, SR. (2008). NTP-CERHR expert
panel report on the reproductive and developmental toxicity of bisphenol A. Birth
Defects Res B Dev Reprod Toxicol 83: 157-395. http://dx.doi.org/10.1002/bdrb.20147.
Charles. GD; Gennings, C; Zacharewski, TR; Gollapudi, BB; Carney, EW. (2002). An approach for
assessing estrogen receptor-mediated interactions in mixtures of three chemicals: a
pilot study. Toxicol Sci 68: 349-360.
Chen. Q; Long. Y; Yuan. X; Zou, L; Sun. J; Chen. S; Perez-Polo, JR; Yang. K. (2005). Protective
effects of bone marrow stromal cell transplantation in injured rodent brain: synthesis of
neurotrophic factors. J Neurosci Res 80: 611-619. http://dx.doi.org/10.1002/inr.20494.
Chen. Yl. (1999). Rank-based tests for dose finding in nonmonotonic dose-response settings.
Biometrics 55:1258-1262.
Chiamolera, Ml; Wondisford, FE. (2009). Minireview: Thyrotropin-releasing hormone and the
thyroid hormone feedback mechanism [Review]. Endocrinology 150:1091-1096.
http://dx.doi.org/10.1210/en.2008-1795.
Chitra. KC; Latchoumycandane. C; Mathur. PP. (2003a). Induction of oxidative stress by
bisphenol A in the epididymal sperm of rats. Toxicology 185: 119-127.
http://dx.doi.org/10.1016/S0300-483X(02)00597-8.
Chitra. KC: Rao. KR; Mathur. PP. (2003b). Effect of bisphenol A and co-administration of
bisphenol A and vitamin C on epididymis of adult rats: a histological and biochemical
study. Asian J Androl 5: 203-208.
Cho. SD; Kim. JH; Kim. DY; Lee. YS; Kang. KS. (2003). Pre-validation study for OECD enhanced
test guideline 407 protocol by gavage for 4 weeks using propylthiouracil and tamoxifen.
Toxicol Lett 144: 195-204.
Christiansen. S; Boberg. J: Axelstad. M; Dalgaard. M; Vinggaard. A: Metzdorff. S; Hass. U. (2010).
Low-dose perinatal exposure to di(2-ethylhexyl) phthalate induces anti-androgenic
effects in male rats. Reprod Toxicol 30: 313-321.
http://dx.doi.Org/10.1016/i.reprotox.2010.04.005.
Christiansen. S; Scholze. M; Dalgaard. M; Vinggaard. AM: Axelstad. M; Kortenkamp. A: Hass. U.
(2009). Synergistic disruption of external male sex organ development by a mixture of
four antiandrogens. Environ Health Perspect 117: 1839-1846.
http://dx.doi.org/10.1289/ehp.0900689.
Collins. WT. Jr; Capen. CC. (1980). Ultrastructural and functional alterations of the rat thyroid
gland produced by polychlorinated biphenyls compared with iodide excess and
deficiency, and thyrotropin and thyroxine administration. Virchows Arch B Cell Pathol
Incl Mol Pathol 33: 213-231. http://dx.doi.org/10.1007/BF02899183.
153 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Comer. CP; Chengelis, CP; Levin. S; Kotsonis, FN. (1985). Changes in thyroidal function and liver
UDPglucuronosyltransferase activity in rats following administration of a novel imidazole
(SC-37211). Toxicol Appl Pharmacol 80: 427-436.
Conolly, RB; Lutz, WK. (2004). Nonmonotonic dose-response relationships: Mechanistic basis,
kinetic modeling, and implications for risk assessment. Toxicol Sci 77:151-157.
http://dx.doi.org/10.1093/toxsci/kfh007.
Cooney, MA; Buck Louis. GM; Hediger, ML; Vexler, A; Kostyniak, PJ. (2010). Organochlorine
pesticides and endometriosis. Reprod Toxicol 30: 365-369.
http://dx.doi.Org/10.1016/i.reprotox.2010.05.011.
Cooper. PS; Kieffer, JD; Halpern, R; Saxe, V; Mover. H; Maloof, F; Ridgway, EC. (1983).
Propylthiouracil (PTU) pharmacology in the rat. II. Effects of PTU on thyroid function.
Endocrinology 113: 921-928. http://dx.doi.org/10.1210/endo-113-3-921.
Coviello, AD; Matsumoto, AM; Bremner, WJ; Herbst, KL; Amory, JK; Anawalt, BP; Sutton. PR;
Wright. WW; Brown. TR; Yan, X; Zirkin, BR; Jarow, JP. (2005). Low-dose human chorionic
gonadotropin maintains intratesticular testosterone in normal men with testosterone-
induced gonadotropin suppression. J Clin Endocrinol Metab 90: 2595-2602.
http://dx.doi.org/10.1210/ic.2004-08Q2.
Crane. HM; Pickford, PB; Hutchinson. TH; Brown. JA. (2005). Effects of ammonium perchlorate
on thyroid function in developing fathead minnows, Pimephales promelas. Environ
Health Perspect 113: 396-401. http://dx.doi.org/10.1289/ehp.7333.
Crane. HM; Pickford. PB; Hutchinson. TH; Brown. JA. (2006). The effects of methimazole on
development of the fathead minnow, Pimephales promelas, from embryo to adult.
Toxicol Sci 93: 278-285. http://dx.doi.org/10.1093/toxsci/kfl063.
Cripe, GM; Hemmer, BL; Raimondo, S; Goodman. LR; Kulaw, PH. (2010). Exposure of three
generations of the estuarine sheepshead minnow (Cyprinodon variegatus) to the
androgen, 17beta-trenbolone: effects on survival, development, and reproduction.
Environ Toxicol Chem 29: 2079-2087. http://dx.doi.org/10.10Q2/etc.261.
Crofton. KM. (2008). Thyroid disrupting chemicals: mechanisms and mixtures [Review]. Int J
Androl 31: 209-223. http://dx.doi.Org/10.llll/i.1365-2605.2007.00857.x.
Crofton. KM; Craft. ES; Hedge. JM; Gennings. C; Simmons. JE; Carchman. RA; Carter. WH. Jr;
Pevito. MJ. (2005). Thyroid-hormone-disrupting chemicals: Evidence for dose-
dependent additivity or synergism. Environ Health Perspect 113:1549-1554.
Crofton. KM; Zoeller. RT. (2005). Mode of action: neurotoxicity induced by thyroid hormone
disruption during development-hearing loss resulting from exposure to PHAHs
[Review], Crit Rev Toxicol 35: 757-769. http://dx.doi.org/10.1080/104084405910073Q4.
Palsenter. PR; Santana. GM; Grande. SW; Andrade. AJ; Araujo. SL. (2006). Phthalate affect the
reproductive function and sexual behavior of male Wistar rats. Hum Exp Toxicol 25: 297-
303. http://dx.doi.org/10.1191/0960327105ht624oa.
Palu. A; Blaydes. BS; Bryant. CW; Latendresse. JR; Weis. CC; Pelclos. KB. (2002). Estrogen
receptor expression in the prostate of rats treated with dietary genistein. J Chromatogr
B Analyt Technol Biomed Life Sci 777: 249-260. http://dx.doi.org/10.1016/S157Q-
0232(02)00346-X.
Pasgupta. PK; Pyke. JV; Kirk. AB; Jackson. WA. (2006). Perchlorate in the United States. Analysis
of relative source contributions to the food chain. Environ Sci Technol 40: 6608-6614.
http://dx.doi.org/10.1021/es061321z.
154 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Davis. PJ; Leonard, JL; Davis. FB. (2008). Mechanisms of nongenomic actions of thyroid
hormone [Review]. Front Neuroendocrinol 29: 211-218.
http://dx.doi.Org/10.1016/i.yfrne.2007.09.003.
DCED (Danish Centre on Endocrine Disrupters). (2013). Input for the REACH-review in 2013 on
endocrine disrupters. Denmark: DTU Food, National Food Institute.
Delclos, KB; Bucci, TJ; Lomax. LG; Latendresse, JR; Warbritton, A; Weis, CC; Newbold, RR.
(2001). Effects of dietary genistein exposure during development on male and female
CD (Sprague-Dawley) rats. Reprod Toxicol 15: 647-663.
http://dx.doi.org/10.1016/S0890-6238(01)00177-0.
Delclos. KB; Weis. CC; Bucci. TJ; Olson. G; Mellick, P; Sadovova, N; Latendresse. JR; Thorn. B;
Newbold. RR. (2009). Overlapping but distinct effects of genistein and ethinyl estradiol
(EE(2)) in female Sprague-Dawley rats in multigenerational reproductive and chronic
toxicity studies. Reprod Toxicol 27: 117-132.
http://dx.doi.Org/10.1016/i.reprotox.2008.12.005.
Desaulniers, D; Poon, R; Phan, W; Leingartner, K; Foster. WG; Chu, I. (1997). Reproductive and
thyroid hormone levels in rats following 90-day dietary exposure to PCB 28 (2,4,4'-
trichlorobiphenyl) or PCB 77 (3,3'4,4'-tetrachlorobiphenyl). Toxicol Ind Health 13: 627-
638.
Pi Fausto, V; Fiore, M; Tirassa, P; Lambiase, A; Aloe. L. (2007). Eye drop NGF administration
promotes the recovery of chemically injured cholinergic neurons of adult mouse
forebrain. Eur J Neurosci 26: 2473-2480. http://dx.doi.org/10.1111/j. 1460-
9568.2007.05883.x.
Diamanti-Kandarakis, E; Bourguignon, JP; Giudice, LC; Hauser, R; Prins, GS; Soto. AM; Zoeller,
RT; Gore. AC. (2009). Endocrine-disrupting chemicals: an Endocrine Society scientific
statement [Review]. Endocr Rev 30: 293-342. http://dx.doi.org/10.1210/er.2009-00Q2.
Do. RP; Stahlhut. RW; Ponzi. D; Vom Saal. FS; Taylor. JA. (2012). Non-monotonic dose effects of
in utero exposure to di(2-ethylhexyl) phthalate (DEHP) on testicular and serum
testosterone and anogenital distance in male mouse fetuses. Reprod Toxicol 34: 614-
621. http://dx.doi.Org/10.1016/i.reprotox.2012.09.006.
Doerge. PR: Twaddle. NC; Vanlandingham. M; Fisher. JW. (2010a). Pharmacokinetics of
bisphenol A in neonatal and adult Sprague-Pawley rats. Toxicol Appl Pharmacol 247:
158-165. http://dx.doi.Org/10.1016/i.taap.2010.06.008.
Poerge. PR: Twaddle. NC: Woodling. KA; Fisher. JW. (2010b). Pharmacokinetics of bisphenol A
in neonatal and adult rhesus monkeys. Toxicol Appl Pharmacol 248: 1-11.
http://dx.doi.Org/10.1016/i.taap.2010.07.009.
Poerge. PR: Vanlandingham. M; Twaddle. NC: Pelclos. KB. (2010c). Lactational transfer of
bisphenol A in Sprague-Pawley rats. Toxicol Lett 199: 372-376.
http://dx.doi.Org/10.1016/i.toxlet.2010.09.022.
Pohan. 0: Pe La Vieja. A: Paroder. V; Ridel. C; Artani. M; Reed. M; Ginter. CS; Carrasco. N.
(2003). The sodium/iodide symporter (NIS): characterization, regulation, and medical
significance. Endocr Rev 24: 48-77. http://dx.doi.org/10.1210/er.2001-0Q29.
Purhan. EJ; Lambright. CS: Makynen. EA; Lazorchak. J: Hartig. PC: Wilson. VS; Gray. LE; Ankley.
GT. (2006). Identification of metabolites of trenbolone acetate in androgenic runoff
from a beef feedlot. Environ Health Perspect 114 Suppl 1: 65-68.
155 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Dykman, DP. (1981). Temporal Effects of Testosterone-Estradiol Polydimethylsiloxane
Subdermal Implants on Pituitary, Leydig Cell, and Germinal Epithelium Function and
Daily Serum Testosterone Rhythm in Male Rats. Biol Reprod 25: 235-243.
http://dx.doi.Org/10.1095/biolreprod25.2.235.
EDSTAC (Endocrine Disruptor Screening and Testing and Advisory Committee). (1998).
Endocrine Disruptor Screening and Testing Advisory Committee, Final Report, Volume I-
II. http://www.epa.gov/scipoly/oscpendo/pubs/edspoverview/finalrpt.htm.
EFSA (European Food Safety Authority). (2005). Opinion of the Scientific Panel on food
additives, flavourings, processing aids and materials in contact with food (AFC) related
to Semicarbazide in food. EFSA J 6: 219. http://dx.doi.Org/10.2903/j.efsa.2005.219.
EFSA (European Food Safety Authority). (2010). Scientific opinion on bisphenol A: Evaluation of
a study investigating its neurodevelopmental toxicity, review of recent scientific
literature on its toxicity and advice on the Danish risk assessment of Bisphenol A. EFSA J
8:1829-1821-1829-1826. http://dx.doi.Org/10.2903/i.efsa.2010.1829.
EFSA (European Food Safety Authority). (2013). Scientific Opinion on the hazard assessment of
endocrine disruptors: Scientific criteria for identification of endocrine disruptors and
appropriateness of existing test methods for assessing effects mediated by these
substances on human health and the environment.
Ekman, PR; Hartig, PC; Cardon, M; Skelton, DM; Teng, Q; Durhan, EJ; Jensen. KM; Kahl, MP;
Villeneuve, PL; Gray, LE; Collette, TW; Ankley, GT. (2012). Metabolite profiling and a
transcriptional activation assay provide direct evidence of androgen receptor
antagonism by bisphenol A in fish. Environ Sci Technol 46: 9673-9680.
http://dx.doi.org/10.1021/es3Q14634.
Ekman. PR; Villeneuve. PL; Teng. Q; Ralston-Hooper, KJ; Martinovic-Weigelt, P; Kahl. MP;
Jensen. KM: Purhan. EJ: Makynen. EA; Ankley. GT: Collette. TW. (2011). Use of gene
expression, biochemical and metabolite profiles to enhance exposure and effects
assessment of the model androgen 17|3-trenbolone in fish. Environ Toxicol Chem 30:
319-329. http://dx.doi.org/10.1002/etc.406.
Ema. M; Fujii. S; Furukawa. M; Kiguchi. M; Ikka. T; Harazono. A. (2001). Rat two-generation
reproductive toxicity study of bisphenol A. Reprod Toxicol 15: 505-523.
http://dx.doi.org/10.1016/S0890-6238(01)00160-5.
Eskandari. S. (1997). Thyroid Na+/I-Symporter. MECHANISM, STOICHIOMETRY, ANP
SPECIFICITY. J Biol Chem 272: 27230-27238.
http://dx.doi.org/10.1074/ibc.272.43.2723Q.
Ewing. LL; Pesjardins. C; Irby. PC: Robaire. B. (1977). Synergistic interaction of testosterone and
oestradiol inhibits permatogenesis in rats. Nature 269: 409-411.
http://dx.doi.org/10.1038/269409a0.
Ewing. LL: Schanbacher. B; Pesjardins. C; Chaffee. V. (1976). The effect of subdermal
testosterone filled polydimethylsiloxane implants on spermatogenesis in rhesus
monkeys (Macaca mulata). Contraception 13: 583-596.
Ewing. LL: Stratton. LG: Pesjardins. C. (1973). EFFECT OF TESTOSTERONE
POLYPIMETHYLSILOXANE IMPLANTS UPON SPERM PROPUCTION, LIBIPO ANP
ACCESSORY SEX ORGAN FUNCTION IN RABBITS. Reproduction 35: 245-253.
http://dx.doi.Org/10.1530/irf.0.0350245.
156 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Fagin, D. (2012). Toxicology: The learning curve. Nature 490: 462-465.
http://dx.doi.org/10.1038/49Q462a.
Fekete, C; Lechan, RM. (2007). Negative feedback regulation of hypophysiotropic thyrotropin-
releasing hormone (TRH) synthesizing neurons: role of neuronal afferents and type 2
deiodinase [Review]. Front Neuroendocrinol 28: 97-114.
http://dx.doi.Org/10.1016/i.yfrne.2007.04.002.
Fliers. E; Unmehopa, UA; Alkemade, A. (2006). Functional neuroanatomy of thyroid hormone
feedback in the human hypothalamus and pituitary gland [Review]. Mol Cell Endocrinol
251: 1-8. http://dx.doi.Org/10.1016/i.mce.2006.03.042.
Flynn, KM; Delclos, KB; Newbold, RR; Ferguson. SA. (2001). Behavioral responses of rats
exposed to long-term dietary vinclozolin. J Agric Food Chem 49: 1658-1665.
http://dx.doi.org/10.1021/if00Q8893.
Fowles, JR; Fairbrother, A; Baecher-Steppan, L; Kerkvliet, Nl. (1994). Immunologic and
endocrine effects of the flame-retardant pentabromodiphenyl ether (DE-71) in C57BL/6J
mice. Toxicology 86: 49-61. http://dx.doi.org/10.1016/0300-483X(94)90052-3.
Freitas, J; Cano, P; Craig-Veit, C; Goodson, ML; Furlow, JD; Murk. AJ. (2011). Detection of
thyroid hormone receptor disruptors by a novel stable in vitro reporter gene assay.
Toxicol In Vitro 25: 257-266. http://dx.doi.Org/10.1016/i.tiv.2010.08.013.
Gao, W; Bohl, CE; Dalton, JT. (2005). Chemistry and structural biology of androgen receptor
[Review]. Chem Rev 105: 3352-3370. http://dx.doi.org/10.1021/cr020456u.
Gardner. DF; Cruikshank, DP; Hays, PM; Cooper. PS. (1986). Pharmacology of propylthiouracil
(PTU) in pregnant hyperthyroid women: correlation of maternal PTU concentrations
with cord serum thyroid function tests. J Clin Endocrinol Metab 62: 217-220.
http://dx.doi.org/10.1210/icem-62-l-217.
Gaylor. DW; Lutz. WK; Conolly. RB. (2004). Statistical analysis of nonmonotonic dose response
relationships: Research design and analysis of nasal cell proliferation in rats exposed to
formaldehyde. Toxicol Sci 77:158-164. http://dx.doi.org/10.1093/toxsci/kfh008.
Ge. RS; Chen. GR; Dong. Q; Akingbemi. B; Sottas. CM: Santos. M; Sealfon. SC; Bernard. DJ;
Hardy. MP. (2007). Biphasic effects of postnatal exposure to diethylhexylphthalate on
the timing of puberty in male rats. J Androl 28: 513-520.
http://dx.doi.org/10.2164/iandrol.106.0019Q9.
Gennings. C; Sabo. R; Carney. E. (2010). Identifying subsets of complex mixtures most
associated with complex diseases: polychlorinated biphenyls and endometriosis as a
case study. Epidemiology 21 Suppl 4: S77-S84.
http://dx.doi.org/10.1097/EDE.0b013e3181ce946c.
Gilbert. ME. (2011). Impact of low-level thyroid hormone disruption induced by propylthiouracil
on brain development and function. Toxicol Sci 124: 432-445.
http://dx.doi.org/10.1093/toxsci/kfr244.
Gilbert. ME: Rovet. J: Chen. Z; Koibuchi. N. (2012). Developmental thyroid hormone disruption:
prevalence, environmental contaminants and neurodevelopmental consequences.
Neurotoxicology 33: 842-852. http://dx.doi.Org/10.1016/i.neuro.2011.ll.005.
Gilbert. ME: Sui. L. (2006). Dose-dependent reductions in spatial learning and synaptic function
in the dentate gyrus of adult rats following developmental thyroid hormone
insufficiency. Brain Res 1069: 10-22. http://dx.doi.Org/10.1016/i.brainres.2005.10.049.
157 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Gilbert. ME; Sui, L. (2008). Developmental exposure to perchlorate alters synaptic transmission
in hippocampus of the adult rat. Environ Health Perspect 116: 752-760.
http://dx.doi.org/10.1289/ehp.11089.
Gill. S; Bondy, G; Lefebvre, D; Becalski, A; Kavanagh, M; Hou, Y; Turcotte, A; Barker. M; Weld. M;
Vavasour. E; Cooke. G. (2010). Subchronic oral toxicity study of furan in Fischer-344 rats.
Toxicol Pathol 38: 619-630. http://dx.doi.org/10.1177/019262331Q368978.
Gordon. CJ; Becker. P; Padnos, B. (2000). Comparison of heat and cold stress to assess
thermoregulatory dysfunction in hypothyroid rats. Am J Physiol Regul IntegrComp
Physiol 279: R2066-R2071.
Graham. SL; Davis. KJ; Hansen. WH; Graham. CH. (1975). Effects of prolonged ethylene thiourea
ingestion on the thyroid of the rat. Food Cosmet Toxicol 13: 493-499.
http://dx.doi.org/10.1016/0015-6264(75)90001-2.
Graham. SL; Hansen. WH; Davis. KJ; Perry, CH. (1973). Effects of one-year administration of
ethylenethiourea upon the thyroid of the rat. J Agric Food Chem 21: 324-329.
http://dx.doi.org/10.1021/if60187a036.
Grande. SW; Andrade, AJ; Talsness, CE; Grote, K; Chahoud, I. (2006). A dose-response study
following in utero and lactational exposure to di(2-ethylhexyl)phthalate: effects on
female rat reproductive development. Toxicol Sci 91: 247-254.
http://dx.doi.org/10.1093/toxsci/kfil28.
Grande. SW; Andrade. AJ; Talsness. CE; Grote. K; Golombiewski, A; Sterner-Kock, A; Chahoud. I.
(2007). A dose-response study following in utero and lactational exposure to di-(2-
ethylhexyl) phthalate (DEHP): reproductive effects on adult female offspring rats.
Toxicology 229: 114-122. http://dx.doi.Org/10.1016/i.tox.2006.10.005.
Gray, LE. Jr.; Ostby, J; Marshall. R; Andrews. J. (1993). Reproductive and thyroid effects of low-
level polychlorinated biphenyl (Aroclor 1254) exposure. Fundam Appl Toxicol 20: 288-
294. http://dx.doi.org/10.1006/faat.1993.1038.
Gray. PB; Singh. AB; Woodhouse. LJ; Storer. TW; Casaburi. R; Dzekov. J; Dzekov. C; Sinha-Hikim.
I; Bhasin. S. (2005). Dose-dependent effects of testosterone on sexual function, mood,
and visuospatial cognition in older men. J Clin Endocrinol Metab 90: 3838-3846.
http://dx.doi.org/10.1210/ic.2005-Q247.
Grim. KC; Wolfe. M; Braunbeck. T; Iguchi. T; Ohta. Y; Tooi. 0; Touart. L; Wolf. DC; Tietge. J.
(2009). Thyroid histopathology assessments for the amphibian metamorphosis assay to
detect thyroid-active substances. Toxicol Pathol 37: 415-424.
http://dx.doi.org/10.1177/0192623309335063.
Harley. KG; Marks. AR; Chevrier. J; Bradman. A; Sjodin. A; Eskenazi. B. (2010). PBDE
concentrations in women's serum and fecundability. Environ Health Perspect 118: 699-
704. http://dx.doi.org/10.1289/ehp.0901450.
Haseman. JK; Bailer. AJ; Kodell. RL; Morris. R; Portier. K. (2001). Statistical issues in the analysis
of low-dose endocrine disruptor data. Toxicol Sci 61: 201-210.
Hass. U; Scholze. M; Christiansen. S; Dalgaard. M; Vinggaard. AM; Axelstad. M; Metzdorff. SB;
Kortenkamp. A. (2007). Combined exposure to anti-androgens exacerbates disruption of
sexual differentiation in the rat. Environ Health Perspect 115:122-128.
http://dx.doi.org/10.1289/ehp.9360.
158 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Hatef, A; Alavi, SM; Milia, S; Krist'an, J; Golshan, M; Fontaine. P; Linhart, O. (2012). Anti-
androgen vinclozolin impairs sperm quality and steroidogenesis in goldfish. Aquat
Toxicol 122-123: 181-187. http://dx.doi.Org/10.1016/i.aquatox.2012.06.009.
Hayes, T; Haston, K; Tsui. M; Hoang, A; Haeffle, C; Vonk, A. (2003). Atrazine-induced
hermaphroditism at 01 ppb in American leopard frogs (Rana pipiens): Laboratory and
field evidence. Environ Health Perspect 111: 568-575.
Hellwig, J; van Ravenzwaay, B; Mayer, M; Gembardt, C. (2000). Pre- and postnatal oral toxicity
of vinclozolin in Wistar and Long-Evans rats. Regul Toxicol Pharmacol 32: 42-50.
http://dx.doi.org/10.1006/rtph.2000.140Q.
Hemmer, MJ; Cripe, GM; Hemmer, BL; Goodman, LR; Salinas, KA; Fournie, JW; Walker, CC.
(2008). Comparison of estrogen-responsive plasma protein biomarkers and reproductive
endpoints in sheepshead minnows exposed to 17beta-trenbolone. Aquat Toxicol 88:
128-136. http://dx.doi.Org/10.1016/i.aquatox.2008.04.001.
Herring, AH. (2010). Nonparametric bayes shrinkage for assessing exposures to mixtures subject
to limits of detection. Epidemiology 21 Suppl 4: S71-S76.
http://dx.doi.org/10.1097/EDE.0b013e3181cf0058.
Hill, AB. (1965). The environment and disease: Association or causation? Proc R Soc Med 58:
295-300.
Hill, RN; Crisp, TM; Hurley, PM; Rosenthal, SL; Singh, DV. (1998). Risk assessment of thyroid
follicular cell tumors. Environ Health Perspect 106: 447-457.
Hirabayashi, Y; Inoue, T. (2011). The low-dose issue and stochastic responses to endocrine
disruptors. J Appl Toxicol 31: 84-88. http://dx.doi.org/10.10Q2/iat.1571.
Hirai, IN; Nanba, A; Koshio, M; Kondo, T; Morita, M; Tatarazako, IN. (2006). Feminization of
Japanese medaka (Oryzias latipes) exposed to 17beta-estradiol: formation of testis-ova
and sex-transformation during early-ontogeny. Aquat Toxicol 77: 78-86.
http://dx.doi.Org/10.1016/i.aquatox.2005.ll.001.
Hoffmann, A: Levchenko, A: Scott, ML: Baltimore, D. (2002). The IkappaB-NF-kappaB signaling
module: temporal control and selective gene activation. Science 298: 1241-1245.
http://dx.doi.org/10.1126/science.1071914.
Holbech, H; Kinnberg, K; Petersen, Gl; Jackson, P; Hylland, K; Norrgren, L; Bjerregaard, P. (2006).
Detection of endocrine disrupters: evaluation of a Fish Sexual Development Test (FSDT).
Comp Biochem Physiol C Toxicol Pharmacol 144: 57-66.
http://dx.doi.Org/10.1016/i.cbpc.2006.05.006.
Holbech, H; Kinnberg, KL; Brande-Lavridsen. IN; Bjerregaard, P; Petersen, Gl: Norrgren, L; Orn, S;
Braunbeck, T; Baumann, L; Bomke, C; Dorgerloh, M; Bruns, E; Ruehl-Fehlert. C; Green,
JW; Springer, TA; Gourmelon, A. (2012). Comparison of zebrafish (Danio rerio) and
fathead minnow (Pimephales promelas) as test species in the Fish Sexual Development
Test (FSDT). Comp Biochem Physiol C Toxicol Pharmacol 155: 407-415.
http://dx.doi.Org/10.1016/i.cbpc.2011.ll.002.
Hood, A; Liu, YP; Gattone, VH; Klaassen, CD. (1999). Sensitivity of thyroid gland growth to
thyroid stimulating hormone (TSH) in rats treated with antithyroid drugs. Toxicol Sci 49:
263-271.
Hornung, MW; Jensen, KM; Korte, JJ; Kahl, MP; Durhan, EJ; Denny, JS; Henry, TR; Ankley, GT.
(2004). Mechanistic basis for estrogenic effects in fathead minnow (Pimephales
159 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
promelas) following exposure to the androgen 17alpha-methyltestosterone: conversion
of 17alpha-methyltestosterone to 17alpha-methylestradiol. Aquat Toxicol 66:15-23.
Hotchkiss, AK; Furr, J; Makynen, EA; Ankley, GT; Gray, LE. Jr. (2007). In utero exposure to the
environmental androgen trenbolone masculinizes female Sprague-Dawley rats. Toxicol
Lett 174: 31-41. http://dx.doi.Org/10.1016/i.toxlet.2007.08.008.
Hotchkiss. AK; Rider. CV; Blystone, CR; Wilson. VS; Hartig, PC; Ankley, GT; Foster. PM; Gray, CL;
Gray, LE. (2008). Fifteen years after "Wingspread"~environmental
endocrine disrupters and human and wildlife health: where we are today and where we
need to go [Review]. Toxicol Sci 105: 235-259. http://dx.doi.org/10.1093/toxsci/kfn030.
Hotchkiss. AK; Rider. CV; Furr. J; Howdeshell, KL; Blystone, CR; Wilson. VS; Gray, LE. (2010). In
utero exposure to an AR antagonist plus an inhibitor of fetal testosterone synthesis
induces cumulative effects on F1 male rats. Reprod Toxicol 30: 261-270.
http://dx.doi.Org/10.1016/i.reprotox.2010.06.001.
Hotz, KJ; Wilson, AG; Thake, DC; Roloff, MV; Capen, CC; Kronenberg, JM; Brewster, DW. (1997).
Mechanism of thiazopyr-induced effects on thyroid hormone homeostasis in male
Sprague-Dawley rats. Toxicol Appl Pharmacol 142: 133-142.
http://dx.doi.org/10.1006/taap.1996.8032.
Howdeshell, KL. (2002). A model of the development of the brain as a construct of the thyroid
system [Review]. Environ Health Perspect 110 Suppl 3: 337-348.
Howdeshell, KL; Furr, J; Lambright, CR; Wilson, VS; Ryan, BC; Gray, LE. (2008). Gestational and
lactational exposure to ethinyl estradiol, but not bisphenol A, decreases androgen-
dependent reproductive organ weights and epididymal sperm abundance in the male
long evans hooded rat. Toxicol Sci 102: 371-382.
http://dx.doi.org/10.1093/toxsci/kfm306.
Huang, P; Chandra, V; Rastinejad, F. (2010). Structural overview of the nuclear receptor
superfamily: insights into physiology and therapeutics [Review]. Annu Rev Physiol 72:
247-272. http://dx.doi.org/10.1146/annurev-physiol-021909-135917.
Jarosova, A: Harazim, J: Suchy, P; Kratka, L; Stancova, V. (2009). The distribution and
accumulation of phthalates in the organs and tissues of chicks after the administration
of feedstuffs with different phthalate concentrations. Veterinarni Medicina 54: 427-434.
Jasny, BR: Chin, G; Chong, L; Vignieri, S. (2011). Data replication & reproducibility. Again,
and again, and again .... Introduction. Science 334: 1225.
http://dx.doi.org/10.1126/science.334.606Q.1225.
Jensen, KM: Kahl, MP: Makynen, EA: Korte, JJ; Leino, RL; Butterworth, BC: Ankley, GT. (2004).
Characterization of responses to the antiandrogen flutamide in a short-term
reproduction assay with the fathead minnow. Aquat Toxicol 70: 99-110.
http://dx.doi.Org/10.1016/i.aquatox.2004.06.012.
Jeong, SH; Kim, BY: Kang, HG; Ku, HO: Cho, JH. (2006). Effect of chlorpyrifos-methyl on steroid
and thyroid hormones in rat F0- and Fl-generations. Toxicology 220:189-202.
http://dx.doi.Org/10.1016/i.tox.2006.01.005.
Joseph, JD; Wittmann, BM; Dwyer, MA: Cui, H; Dye, DA: Mcdonnell. DP: Norris, JD. (2009).
Inhibition of prostate cancer cell growth by second-site androgen receptor antagonists.
PNAS 106:12178-12183. http://dx.doi.org/10.1073/pnas.09001851Q6.
JRC (Joint Research Centre). (2013). Key scientific issues relevant to the identification and
characterisation of endocrine disrupting substances.
160 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Julien. E; Boobis, AR; Olin, SS; llsi Research Foundation Threshold Working. G. (2009). The Key
Events Dose-Response Framework: a cross-disciplinary mode-of-action based approach
to examining dose-response and thresholds. Crit Rev Food Sci Nutr49: 682-689.
http://dx.doi.org/10.1080/1040839090311Q692.
Kamrin, MA. (2007). The "low dose" hypothesis: validity and implications for human
risk [Review], Int J Toxicol 26: 13-23. http://dx.doi.org/10.1080/109158106Q1117968.
Kang, IJ; Yokota, H; Oshima. Y; Tsuruda, Y; Shimasaki, Y; Honjo, T. (2008). The effects of
methyltestosterone on the sexual development and reproduction of adult medaka
(Oryzias latipes). Aquat Toxicol 87: 37-46.
http://dx.doi.Org/10.1016/i.aquatox.2008.01.010.
Kang. JH; Asai, D; Aasi, D; Katayama, Y. (2007). Bisphenol A in the aquatic environment and its
endocrine-disruptive effects on aquatic organisms [Review]. Crit Rev Toxicol 37: 607-
625. http://dx.doi.org/10.1080/104084407014931Q3.
Kefford, BJ; Zalizniak, L; Warne, MS; Nugegoda, D. (2008). Is the integration of hormesis and
essentiality into ecotoxicology now opening Pandora's Box? Environ Pollut 151: 516-523.
http://dx.doi.Org/10.1016/i.envpol.2007.04.019.
Kelce, WR; Monosson, E; Gamcsik, MP; Laws. SC; Gray, LE. (1994). Environmental hormone
disruptors: evidence that vinclozolin developmental toxicity is mediated by
antiandrogenic metabolites. Toxicol Appl Pharmacol 126: 276-285.
http://dx.doi.org/10.10Q6/taap.1994.1117.
Keller. DA; Juberg, PR; Catlin, IN; Farland, WH; Hess. FG; Wolf. DC; Doerrer, NG. (2012).
Identification and characterization of adverse effects in 21st century toxicology. Toxicol
Sci 126: 291-297. http://dx.doi.org/10.1093/toxsci/kfr350.
Kennel. P; Pallen, C; Barale-Thomas, E; Espuna, G; Bars. R. (2003). Tamoxifen: 28-day oral
toxicity study in the rat based on the Enhanced OECD Test Guideline 407 to detect
endocrine effects. Arch Toxicol 77: 487-499. http://dx.doi.org/10.1007/sQ0204-003-
0476-5.
Kidd. KA; Blanchfield. PJ; Mills. KH; Palace. VP: Evans. RE: Lazorchak. JM; Flick. RW. (2007).
Collapse of a fish population after exposure to a synthetic estrogen. PNAS 104: 8897-
8901. http://dx.doi.org/10.1073/pnas.06095681Q4.
Kim. HS: Han. SY: Kim. TS: Kwack. SJ: Lee. RD: Kim. IY: Seok. JH: Lee. BM: Yoo. SD: Park. KL.
(2002a). No androgenic/anti-androgenic effects of bisphenol-A in Hershberger assay
using immature castrated rats. Toxicol Lett 135: 111-123.
Kim. HS: Shin. JH: Moon. HJ: Kim. TS: Kang. IH: Seok. JH: Kim. IY: Park. KL: Han. SY. (2002b).
Evaluation of the 20-day pubertal female assay in Sprague-Dawley rats treated with DES,
tamoxifen, testosterone, and flutamide. Toxicol Sci 67: 52-62.
http://dx.doi.Org/10.1093/toxsci/67.l.52.
Kinnberg. K; Holbech. H; Petersen. Gl; Bjerregaard. P. (2007). Effects of the fungicide prochloraz
on the sexual development of zebrafish (Danio rerio). Comp Biochem Physiol C Toxicol
Pharmacol 145: 165-170. http://dx.doi.Org/10.1016/i.cbpc.2006.ll.002.
Kirk. AB; Martinelango. PK; Tian. K; Dutta. A: Smith. EE: Dasgupta. PK. (2005). Perchlorate and
iodide in dairy and breast milk. Environ Sci Technol 39: 2011-2017.
http://dx.doi.org/10.1021/esQ48118t.
Kobayashi. K; Kubota. H; Ohtani. K; Hojo. R; Miyagawa. M. (2012). Lack of effects for dietary
exposure of bisphenol A during in utero and lactational periods on reproductive
161 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
development in rat offspring. J Toxicol Sci 37: 565-573.
http://dx.doi.org/10.2131/jts.37.565.
Kobayashi, K; Ohtani, K; Kubota, H; Miyagawa, M. (2010). Dietary exposure to low doses of
bisphenol A: effects on reproduction and development in two generations of C57BL/6J
mice. Congenit Anom 50:159-170. http://dx.doi.org/10.1111/j.l741-
4520.2010.00279.x.
Kohrle, J. (2008). Environment and endocrinology: the case of thyroidology [Review]. Ann
Endocrinol 69:116-122. http://dx.doi.Org/10.1016/i.ando.2008.02.008.
Kondo, F; Okumura, M; Oka. H; Nakazawa, H; Izumi, S; Makino, T. (2010). Determination of
phthalates in diet and bedding for experimental animals using gas chromatography-
mass spectrometry. Bull Environ Contam Toxicol 84: 212-216.
http://dx.doi.org/10.1007/sQ0128-009-9919-x.
Kortenkamp, A. (2007). Introduction: endocrine disruptors-exposure assessment, novel end
points, and low-dose and mixture effects. Environ Health Perspect 115 Suppl 1: 7.
http://dx.doi.org/10.1289/ehp.10517.
Kortenkamp. A. (2008). Low dose mixture effects of endocrine disrupters: implications for risk
assessment and epidemiology [Review]. Int J Androl 31: 233-240.
http://dx.doi.Org/10.llll/i.1365-2605.2007.00862.x.
Kortenkamp. A; Martin. O; Faust. M; Evans. R; McKinlay, R; Orton, F; Rosivatz, E. (2011). State of
the art assessment of endocrine disrupters. Final report. (Project Contract Number
070307/2009/550687/SER/D3). Kortenkamp, A; Martin, O; Faust, M; Evans, R; McKinlay,
R; Orton, F; Rosivatz, E.
Kumar. R; Mcewan, IJ. (2012). Allosteric modulators of steroid hormone receptors: structural
dynamics and gene regulation [Review]. Endocr Rev 33: 271-299.
http://dx.doi.org/10.1210/er.2011-1033.
Kunisue. T; Chen. Z; Buck Louis. GM; Sundaram. R; Hediger. ML: Sun. L; Kannan. K. (2012).
Urinary concentrations of benzophenone-type UV filters in U.S. women and their
association with endometriosis. Environ Sci Technol 46: 4624-4632.
http://dx.doi.org/10.1021/es2Q4415a.
Kwon. JY; Haghpanah. V; Kogson-Hurtado. LM; Mcandrew. BJ; Penman. DJ. (2000).
Masculinization of genetic female nile tilapia (Oreochromis niloticus) by dietary
administration of an aromatase inhibitor during sexual differentiation. J Exp Zool 287:
46-53. http://dx.doi.org/10.1002/1097-010X(20000615)287:1<46::AID-JEZ6>3.0.CQ:2-X.
Lake. BG. (2009). Species differences in the hepatic effects of inducers of CYP2B and CYP4A
subfamily forms: relationship to rodent liver tumour formation [Review]. Xenobiotica
39: 582-596. http://dx.doi.org/10.1080/00498250903Q98184.
Lasley. SM; Gilbert. ME. (2011). Developmental thyroid hormone insufficiency reduces
expression of brain-derived neurotrophic factor (BDNF) in adults but not in neonates.
Neurotoxicol Teratol 33: 464-472. http://dx.doi.Org/10.1016/i.ntt.2011.04.001.
Latendresse. JR; Bucci. TJ; Olson. G; Mellick. P; Weis. CC; Thorn. B; Newbold. RR; Delclos. KB.
(2009). Genistein and ethinyl estradiol dietary exposure in multigenerational and
chronic studies induce similar proliferative lesions in mammary gland of male Sprague-
Dawley rats. Reprod Toxicol 28: 342-353.
http://dx.doi.Org/10.1016/i.reprotox.2009.04.006.
162 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Laws. SC; Ferrell, JM; Stoker. TE; Cooper. RL. (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.
http://dx.doi.org/10.1093/toxsci/kfg223.
Lee. KY; Shibutani, M; Takagi, H; Kato, IN; Takigami, S; Uneyama, C; Hirose, M. (2004). Diverse
developmental toxicity of di-n-butyl phthalate in both sexes of rat offspring after
maternal exposure during the period from late gestation through lactation. Toxicology
203: 221-238. http://dx.doi.Org/10.1016/i.tox.2004.06.013.
Lee. YF; Lin. WJ; Huang. J; Messing. EM; Chan. FL; Wilding. G; Chang. C. (2002). Activation of
mitogen-activated protein kinase pathway by the antiandrogen hydroxyflutamide in
androgen receptor-negative prostate cancer cells. Cancer Res 62: 6039-6044.
Lehmann, KP; Phillips. S; Sar, M; Foster. PM; Gaido, KW. (2004). Dose-dependent alterations in
gene expression and testosterone synthesis in the fetal testes of male rats exposed to di
(n-butyl) phthalate. Toxicol Sci 81: 60-68. http://dx.doi.org/10.1093/toxsci/kfhl69.
Lewin, GR. (1996). Physiology of the Neurotrophins. Annu Rev Neurosci 19: 289-317.
http://dx.doi.Org/10.1146/annurev.neuro.19.l.289.
Li. MH; Hsu. PC; Guo, YL. (2001). Hepatic enzyme induction and acute endocrine effects of
2,2',3,3',4,6'-hexachlorobiphenyl and 2,2',3,4',5',6-hexachlorobiphenyl in prepubertal
female rats. Arch Environ Contam Toxicol 41: 381-385.
http://dx.doi.org/10.1007/sQ02440010263.
Li. MH; Rhine. C; Hansen. LG. (1998). Hepatic enzyme induction and acute endocrine effects of
2,3,3',4',6-pentachlorobiphenyl in prepubertal female rats. Arch Environ Contam Toxicol
35: 97-103. http://dx.doi.org/10.1007/sQ02449900355.
Li. W; Zha, J; Spear. PA; Li. Z; Yang. L; Wang. Z. (2009). Changes of thyroid hormone levels and
related gene expression in Chinese rare minnow (Gobiocypris rarus) during 3-amino-
1,2,4-triazole exposure and recovery. Aquat Toxicol 92: 50-57.
http://dx.doi.Org/10.1016/i.aquatox.2009.01.006.
Li. W; Zhu. L; Zha. J: Wang. Z. (2011). Effects of decabromodiphenyl ether (BDE-209) on mRNA
transcription of thyroid hormone pathway and spermatogenesis associated genes in
Chinese rare minnow (Gobiocypris rarus). Environ Toxicol.
http://dx.doi.org/10.1002/tox.20767.
Louis. GM; Peterson. CM: Chen. Z; Hediger. ML: Croughan. MS: Sundaram. R; Stanford. JB;
Fujimoto. VY; Varner. MW; Giudice. LC; Kennedy. A: Sun. L; Wu. Q; Kannan. K. (2012).
Perfluorochemicals and endometriosis: the ENDO study. Epidemiology 23: 799-805.
http://dx.doi.org/10.1097/EDE.0b013e31826cc0cf.
Lubin. JH; Colt. JS; Camann. D; Davis. S; Cerhan. JR; Severson. RK; Bernstein. L; Hartge. P. (2004).
Epidemiologic evaluation of measurement data in the presence of detection limits.
Environ Health Perspect 112: 1691-1696.
Lutz. WK; Gaylor. DW; Conolly. RB; Lutz. RW. (2005). Nonlinearity and thresholds in dose-
response relationships for carcinogenicity due to sampling variation, logarithmic dose
scaling, or small differences in individual susceptibility [Review]. Toxicol Appl Pharmacol
207: S565-S569. http://dx.doi.Org/10.1016/i.taap.2005.01.038.
Mahood. IK: Scott. HM; Brown. R; Hallmark. IN; Walker. M; Sharpe. RM. (2007). In utero
exposure to di(n-butyl) phthalate and testicular dysgenesis: comparison of fetal and
163 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
adult end points and their dose sensitivity. Environ Health Perspect 115 Suppl 1: 55-61.
http://dx.doi.org/10.1289/ehp.9366.
Mannisto, PT; Ranta, T; Leppaluoto, J. (1979). Effects of methylmercaptoimidazole (MMI),
propylthiouracil (PTU), potassium perchlorate (KCI04) and potassium iodide (Kl) on the
serum concentrations of thyrotrophin (TSH) and thyroid hormones in the rat. Acta
Endocrinol 91: 271-281. http://dx.doi.Org/10.1530/acta.0.0910271.
Maranghi, F; Tassinari, R; Lagatta, V; Moracci, G; Macri, C; Eusepi, A; Pi Virgilio. A; Scattoni, ML;
Calamandrei, G. (2009). Effects of the food contaminant semicarbazide following oral
administration in juvenile Sprague-Dawley rats. Food Chem Toxicol 47: 472-479.
http://dx.doi.Org/10.1016/i.fct.2008.12.003.
Maranghi. F; Tassinari. R; Marcoccia, D; Altieri, I; Catone, T; De Angelis, G; Testai, E;
Mastrangelo, S; Evandri, MG; Bolle, P; Lorenzetti, S. (2010). The food contaminant
semicarbazide acts as an endocrine disrupter: Evidence from an integrated in vivo/in
vitro approach. Chem Biol Interact 183: 40-48.
http://dx.doi.Org/10.1016/i.cbi.2009.09.016.
Martignoni, M; Groothuis, GM; de Kanter, R. (2006). Species differences between mouse, rat,
dog, monkey and human CYP-mediated drug metabolism, inhibition and induction
[Review]. Expert Opin Drug Metab Toxicol 2: 875-894.
http://dx.doi.Org/10.1517/17425255.2.6.875.
Martinovic, D; Blake. LS; Durhan, EJ; Greene. KJ; Kahl, MP; Jensen. KM; Makynen, EA;
Villeneuve, PL; Ankley, GT. (2008). Reproductive toxicity of vinclozolin in the fathead
minnow: confirming an anti-androgenic mode of action. Environ Toxicol Chem 27: 478-
488. http://dx.doi.Org/10.1897/07-206R.l.
Matsuura, I; Saitoh. T; Ashina, M; Wako, Y; Iwata, H; Toyota, IN; Ishizuka, Y; Namiki, M; Hoshino,
IN; Tsuchitani. M; Ikeda. Y. (2005). Evaluation of a two-generation reproduction toxicity
study adding endpoints to detect endocrine disrupting activity using vinclozolin. J
Toxicol Sci 30: 163-188. http://dx.doi.org/10.2131/its.30.S163.
Mcintyre. BS; Barlow. NJ; Foster. PMP. (2001). Androgen-mediated development in male rat
offspring exposed to flutamide in utero: Permanence and correlation of early postnatal
changes in anogenital distance and nipple retention with malformations in androgen-
dependent tissues. Toxicol Sci 62: 236-249. http://dx.doi.Org/10.1093/toxsci/62.2.236.
Mclanahan. EP; El-Masri. HA; Sweeney. LM; Kopylev. LY; Clewell. HJ; Wambaugh. JF; Schlosser.
PM. (2012). Physiologically based pharmacokinetic model use in risk assessment-why
being published is not enough. Toxicol Sci 126: 5-15.
http://dx.doi.org/10.1093/toxsci/kfr295.
Melnick. R; Lucier. G; Wolfe. M; Hall. R; Stancel. G; Prins. G; Gallo. M; Reuhl. K; Ho. SM; Brown.
T; Moore. J; Leakey. J; Haseman. J; Kohn. M. (2002). Summary of the National Toxicology
Program's report of the endocrine disruptors low-dose peer review [Review]. Environ
Health Perspect 110: 427-431.
Mendola. P; Buck Louis. GM. (2010). Environmental contaminants, female reproductive health
and fertility. In TJ Woodruff; S Janssen; L Guillette; L Giudice (Eds.), (pp. 161-172). New
York: Cambridge University Press.
Metzdorff. SB; Palgaard. M; Christiansen. S; Axelstad. M; Hass. U; Kiersgaard. MK; Scholze. M;
Kortenkamp. A; Vinggaard. AM. (2007). Pysgenesis and histological changes of genitals
164 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
and perturbations of gene expression in male rats after in utero exposure to
antiandrogen mixtures. Toxicol Sci 98: 87-98. http://dx.doi.org/10.1093/toxsci/kfm079.
Mihaich, E; Rhodes. J; Wolf. J; van der Hoeven, IN; Dietrich. D; Hall. AT; Caspers, IN; Ortego, L;
Staples. C; Dimond. S; Hentges, S. (2012). Adult fathead minnow, Pimephales promelas,
partial life-cycle reproductive and gonadal histopathology study with bisphenol A.
Environ Toxicol Chem 31: 2525-2535. http://dx.doi.org/10.10Q2/etc.1976.
Miller. MP; Crofton, KM; Rice. DC; Zoeller, RT. (2009). Thyroid-disrupting chemicals:
interpreting upstream biomarkers of adverse outcomes [Review]. Environ Health
Perspect 117: 1033-1041. http://dx.doi.org/10.1289/ehp.0800247.
Miyata, K; Yabushita, S; Sukata, T; Sano, M; Yoshino, H; Nakanishi, T; Okuno, Y; Matsuo, M.
(2002). Effects of perinatal exposure to flutamide on sex hormones and androgen-
dependent organs in F1 male rats. J Toxicol Sci 27: 19-33.
http://dx.doi.org/10.2131/jts.27.19.
Morthorst, JE; Holbech, H; Bjerregaard, P. (2010). Trenbolone causes irreversible
masculinization of zebrafish at environmentally relevant concentrations. Aquat Toxicol
98: 336-343. http://dx.doi.Org/10.1016/i.aquatox.2010.03.008.
Moser, GJ; Wolf. DC; Harden. R; Standeven, AM; Mills. J; Jirtle, RL; Goldsworthy, TL. (1996a).
Cell proliferation and regulation of negative growth factors in mouse liver foci.
Carcinogenesis 17:1835-1840. http://dx.doi.Org/10.1093/carcin/17.9.1835.
Moser. GJ; Wolf. DC; Sar, M; Gaido, KW; Janszen, D; Goldsworthy, TL. (1998). Methyl tertiary
butyl ether-induced endocrine alterations in mice are not mediated through the
estrogen receptor. Toxicol Sci 41: 77-87. http://dx.doi.org/10.1006/toxs.1997.2366.
Moser. GJ; Wolf. DC; Wong. BA; Goldsworthy, TL. (1997). Loss of tumor-promoting activity of
unleaded gasoline in N-nitrosodiethylamine-initiated ovariectomized B6C3F1 mouse
liver. Carcinogenesis 18: 1075-1083. http://dx.doi.Org/10.1093/carcin/18.5.1075.
Moser, GJ; Wong, BA; Wolf, DC; Moss, OR; Goldsworthy, TL. (1996b). Comparative short-term
effects of methyl tertiary butyl ether and unleaded gasoline vapor in female B6C3F1
mice. Toxicol Sci 31: 173-183. http://dx.doi.org/10.1006/faat.1996.0Q89.
Moser, VC; Casey, M; Hamm, A; Carter, WH, Jr.; Simmons, JE; Gennings, C. (2005).
Neurotoxicological and statistical analyses of a mixture of five organophosphorus
pesticides using a ray design. Toxicol Sci 86: 101-115.
http://dx.doi.org/10.1093/toxsci/kfil63.
Murk, AJ; Rijntjes. E; Blaauboer, BJ; Clewell. R; Crofton, KM; Dingemans, MM; Furlow, JD;
Kavlock, R; Kohrle. J; Opitz, R; Traas, T; Visser, TJ; Xia, M; Gutleb, AC. (2013).
Mechanism-based testing strategy using in vitro approaches for identification of thyroid
hormone disrupting chemicals. Toxicol In Vitro 27: 1320-1346.
http://dx.doi.Org/10.1016/i.tiv.2013.02.012.
Murray, CW; Egan, SK; Kim, H; Beru, N; Bolger, PM. (2008). US Food and Drug Administration's
Total Diet Study: dietary intake of perchlorate and iodine. J Expo Sci Environ Epidemiol
18: 571-580. http://dx.doi.org/10.1038/si.ies.7500648.
Mylchreest, E; Cattley, RC; Foster, PM. (1998). Male reproductive tract malformations in rats
following gestational and lactational exposure to Di(n-butyl) phthalate: an
antiandrogenic mechanism? Toxicol Sci 43: 47-60.
http://dx.doi.org/10.1006/toxs.1998.2436.
165 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Mylchreest, E; Sar, M; Cattley, RC; Foster. PMD. (1999). Disruption of androgen-regulated male
reproductive development by di(n-butyl) phthalate during late gestation in rats is
different from flutamide. Toxicol Appl Pharmacol 156: 81-95.
http://dx.doi.org/10.10Q6/taap.1999.8643.
Mylchreest, E; Wallace. DG; Cattley, RC; Foster. PM. (2000). Dose-dependent alterations in
androgen-regulated male reproductive development in rats exposed to Di(n-butyl)
phthalate during late gestation. Toxicol Sci 55: 143-151.
http://dx.doi.Org/10.1093/toxsci/55.l.143.
Nagao, T; Kagawa, IN; Saito, Y; Komada, M. (2012). Developmental effects of oral exposure to
diethylstilbestrol on mouse placenta. J Appl Toxicol. http://dx.doi.org/10.10Q2/iat.2766.
Nagao, T; Wada, K; Marumo, H; Yoshimura, S; Ono, H. (2001). Reproductive effects of
nonylphenol in rats after gavage administration: a two-generation study. Reprod Toxicol
15: 293-315. http://dx.doi.org/10.1016/S0890-6238(01)00123-X.
Nagao, T; Yoshimura, S. (2009). Early embryonic losses in mice induced by diethylstilbestrol.
Congenit Anom 49: 269-273. http://dx.doi.Org/10.llll/i.1741-4520.2009.00247.x.
Nagel, SC; vom Saal, FS; Thayer, KA; Dhar, MG; Boechler, M; Welshons, WV. (1997). Relative
binding affinity-serum modified access (RBA-SMA) assay predicts the relative in vivo
bioactivity of the xenoestrogens bisphenol A and octylphenol. Environ Health Perspect
105: 70-76.
Nichols, JW; Breen, M; Denver, RJ; Distefano, JJ; Edwards, JS; Hoke, RA; Volz, DC; Zhang, X.
(2011). Predicting chemical impacts on vertebrate endocrine systems. Environ Toxicol
Chem 30: 39-51. http://dx.doi.org/10.10Q2/etc.376.
Noriega, NC; Howdeshell, KL; Furr, J; Lambright, CR; Wilson, VS; Gray, LE, Jr. (2009). Pubertal
administration of DEHP delays puberty, suppresses testosterone production, and
inhibits reproductive tract development in male Sprague-Dawley and Long-Evans rats.
Toxicol Sci 111: 163-178. http://dx.doi.org/10.1093/toxsci/kfpl29.
Noriega, NC: Ostby, J: Lambright, C; Wilson, VS: Gray, LE. (2005). Late gestational exposure to
the fungicide prochloraz delays the onset of parturition and causes reproductive
malformations in male but not female rat offspring. Biol Reprod 72: 1324-1335.
http://dx.doi.org/10.1095/biolreprod.104.031385.
NRC (National Research Council). (2005). Health implications of perchlorate ingestion.
Washington, DC: National Academies Press, http://www.nap.edu/catalog/11202.html.
NTP (National Toxicology Program). (2001). National Toxicology Program's report of the
endocrine disruptors low-dose peer review [NTP]. Research Triangle Park, NC.
http://ntp.niehs.nih.gov/index.cfm?obiectid=06F5CE98-E82F-8182-7FA81C02D3690D47.
NTP (National Toxicology Program). (2007). NTP technical report on the toxicology and
carcinogenesis study of genistein (Cas No. 446-72-0) in Sprague-Dawley rats (feed study)
[NTP] (pp. 1-240). (NTP TR 545; NIH Pub. No. 08-4430). Research Triangle Park, NC.
http://ntp.niehs.nih.gov/ntp/htdocs/LT rptsZtr545.pdf.
NTP (National Toxicology Program). (2008). NTP technical report on the multigenerational
reproductive study of genistein (Cas No. 446-72-0) in Sprague-Dawley rats (feed study)
[NTP] (pp. 1-266). (NTP TR 539; NIH Pub. No. 08-4477). Research Triangle Park, NC.
http://ntp.niehs.nih.gov/ntp/htdocs/LT rptsZtr539.pdf.
NTP (National Toxicology Program). (2010). Multigenerational reproductive toxicology study of
ethinyl estradiol (CAS No. 57-63-6) in Sprague-Dawley rats. 1-312.
166 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
O'Connor. JC; Frame. S. R.; Davis. LG; Cook. JC. (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. JC; Frame. S. R.; Ladies, GS. (2002). Evaluation of a 15-day screening assay using
intact male rats for identifying steroid biosynthesis inhibitors and thyroid modulators.
Toxicol Sci 69: 79-91.
O'Malley. BW; Malovannaya, A; Qin, J. (2012). Minireview: nuclear receptor and coregulator
proteomics-2012 and beyond [Review]. 26:1646-1650.
http://dx.doi.org/10.1210/me.2012-1114.
Odum. J; Lefevre, PA; Tinwell, H; Van Miller. JP; Joiner. RL; Chapin, RE; Wallis, NT; Ashby, J.
(2002). Comparison of the developmental and reproductive toxicity of diethylstilbestrol
administered to rats in utero, lactationally, preweaning, or postweaning. Toxicol Sci 68:
147-163. http://dx.doi.Org/10.1093/toxsci/68.l.147.
OECD (Organisation for Economic Co-operation and Development). (2009). Test No. 231:
Amphibian Metamorphosis Assay. In OECD Guidelines for the Testing of Chemicals,
Section 2: Effects on Biotic Systems. Paris, http://dx.doi.org/10.1787/9789264076242-
en.
OECD (Organisation for Economic Co-operation and Development). (2011). Test No. 443:
Extended One-Generation Reproductive Toxicity Study. In OECD Guidelines for the
Testing of Chemicals, Section 4: Health Effects.
http://dx.doi.org/10.1787/9789264122550-en.
Oehlmann. J; Schulte-Oehlmann, U; Bachmann, J; Oetken, M; Lutz, I; Kloas, W; Ternes, TA.
(2006). Bisphenol A induces superfeminization in the ramshorn snail Marisa
cornuarietis(Gastropoda: Prosobranchia) at environmentally relevant concentrations.
Environ Health Perspect 114 Suppl 1: 127-133.
Ogino. Y; Katoh. H; Kuraku. S; Yamada. G. (2009). Evolutionary history and functional
characterization of androgen receptor genes in jawed vertebrates. Endocrinology 150:
5415-5427. http://dx.doi.org/10.1210/en.2009-Q523.
Ohta. R; Ohmukai. H; Marumo. H; Shindo. T; Nagata. T; Ono. H. (2012). Delayed reproductive
dysfunction in female rats induced by early life exposure to low-dose diethylstilbestrol.
Reprod Toxicol 34: 323-330. http://dx.doi.Org/10.1016/i.reprotox.2012.04.014.
Okazaki. K; Imazawa. T; Nakamura. H; Furukawa. F; Nishikawa. A: Hirose. M. (2002). A repeated
28-day oral dose toxicity study of 17alpha-methyltestosterone in rats, based on the
'enhanced OECD Test Guideline 407' for screening the endocrine-disrupting chemicals.
Arch Toxicol 75: 635-642.
Okazaki. K; Okazaki. S; Nishimura. S; Nakamura. H; Kitamura. Y; Hatayama. K; Nakamura. A:
Tsuda. T; Katsumata. T; Nishikawa. A: Hirose. M. (2001). A repeated 28-day oral dose
toxicity study of methoxychlor in rats, based on the 'enhanced OECD test guideline 407'
for screening endocrine-disrupting chemicals. Arch Toxicol 75: 513-521.
Omiecinski. CJ; Vanden Heuvel. JP: Perdew. GH; Peters. JM. (2011). Xenobiotic metabolism,
disposition, and regulation by receptors: from biochemical phenomenon to predictors of
major toxicities [Review]. Toxicol Sci 120 Suppl 1: S49-S75.
http://dx.doi.org/10.1093/toxsci/kfq338.
Opitz. R; Hartmann. S; Blank. T; Braunbeck. T; Lutz. I: Kloas. W. (2006). Evaluation of histological
and molecular endpoints for enhanced detection of thyroid system disruption in
167 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Xenopus laevis tadpoles. Toxicol Sci 90: 337-348.
http://dx.doi.org/10.1093/toxsci/kfi083.
Orn, S; Holbech, H; Madsen, TH; Norrgren, L; Petersen. Gl. (2003). Gonad development and
vitellogenin production in zebrafish (Danio rerio) exposed to ethinylestradiol and
methyltestosterone. Aquat Toxicol 65: 397-411.
Osachoff, HL; Van Aggelen, GC; Mommsen, TP; Kennedy, CJ. (2012). Concentration-response
relationships and temporal patterns in hepatic gene expression of Chinook salmon
(Oncorhynchus tshawytscha) exposed to sewage. Comparative Biochemistry and
Physiology - Part D: Genomics and Proteomics 8: 32-44.
http://dx.doi.Org/10.1016/i.cbd.2012.10.002.
Ostby, J; Monosson, E; Kelce, WR; Gray, LE. Jr. (1999). Environmental antiandrogens: Low doses
of the fungicide vinclozolin alter sexual differentiation of the male rat. Toxicol Ind Health
15: 48-64. http://dx.doi.org/10.1177/0748233799015001Q6.
Otto. C; Fuchs, I; Kauselmann, G; Kern. H; Zevnik, B; Andreasen, P; Schwarz, G; Altmann, H;
Klewer, M; Schoor, M; Vonk, R; Fritzemeier, KH. (2009). GPR30 does not mediate
estrogenic responses in reproductive organs in mice. Biol Reprod 80: 34-41.
http://dx.doi.org/10.1095/biolreprod.108.071175.
Owens. JW; Chaney, JG. (2005). Weighing the results of differing 'low dose' studies of the
mouse prostate by Nagel, Cagen, and Ashby: quantification of experimental power and
statistical results. Regul Toxicol Pharmacol 43: 194-202.
http://dx.doi.Org/10.1016/i.yrtph.2005.07.001.
Owens. W; Gray, LE. Jr.; Zeiger, E; Walker. M; Yamasaki, K; Ashby, J; 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. http://dx.doi.org/10.1289/ehp.9666.
Owens. W; Zeiger. E; Walker. M; Ashby. J: Onyon. L; Gray. LE. 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 114: 1259-1265.
http://dx.doi.org/10.1289/ehp.8751.
Park. JW: Rinchard. J: Liu. F; Anderson. TA; Kendall. RJ; Theodorakis. CW. (2006). The thyroid
endocrine disruptor perchlorate affects reproduction, growth, and survival of
mosquitofish. Ecotoxicol Environ Saf 63: 343-352.
http://dx.doi.Org/10.1016/i.ecoenv.2005.04.002.
Parrott. JL; Blunt. BR. (2005). Life-cycle exposure of fathead minnows (Pimephales promelas) to
an ethinylestradiol concentration below 1 ng/L reduces egg fertilization success and
demasculinizes males. Environ Toxicol 20: 131-141.
http://dx.doi.org/10.1002/tox.20Q87.
Patino. R; Wainscott. MR: Cruz-Li. El: Balakrishnan. S; Mcmurry. C; Blazer. VS; Anderson. TA.
(2003). Effects of ammonium perchlorate on the reproductive performance and thyroid
follicle histology of zebrafish. Environ Toxicol Chem Chem.: press.
Pawlowski. S; Sauer. A: Shears. JA; Tyler. CR; Braunbeck. T. (2004). Androgenic and estrogenic
effects of the synthetic androgen 17alpha-methyltestosterone on sexual development
and reproductive performance in the fathead minnow (Pimephales promelas)
168 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
determined using the gonadal recrudescence assay. Aquat Toxicol 68: 277-291.
http://dx.doi.Org/10.1016/i.aquatox.2004.03.018.
Perkins. NJ; Schisterman, EF; Vexler, A. (2007). Receiver operating characteristic curve inference
from a sample with a limit of detection. Am J Epidemiol 165: 325-333.
http://dx.doi.org/10.1093/aie/kwk011.
Phelps. RP; Okoko, M. (2011). A non-paradoxical dose response to 17a-methyltestosterone by
Nile tilapia Oreochromis niloticus (L.): Effects on the sex ratio, growth and gonadal
development. Aquaculture Research 42: 549-558. http://dx.doi.org/10.1111/j. 1365-
2109.2010.02650.x.
Pocar, P; Fiandanese, N; Secchi, C; Berrini, A; Fischer. B; Schmidt. JS; Schaedlich, K; Borromeo, V.
(2012). Exposure to di(2-ethyl-hexyl) phthalate (DEHP) in utero and during lactation
causes long-term pituitary-gonadal axis disruption in male and female mouse offspring.
Endocrinology 153: 937-948. http://dx.doi.org/10.1210/en.2011-1450.
Pollack. AZ; Buck Louis. GM; Lynch, CD; Kostyniak, PJ. (2011). Persistent Organochlorine
Exposure and Pregnancy Loss: A Prospective Cohort Study. 2: 683-691.
http://dx.doi.org/10.4236/iep.2011.26Q79.
Poon, R; Chu, I; Lebel, G; Yagminas, A; Valli, VE. (2003). Effects of dibromoacetonitrile on rats
following 13-week drinking water exposure. Food Chem Toxicol 41: 1051-1061.
Poon. R; Lecavalier, P; Mueller. R; Valli. VE; Procter. BG; Chu. I. (1997). Subchronic oral toxicity
of di-n-octyl phthalate and di(2-ethylhexyl) phthalate in the rat. Food Chem Toxicol 35:
225-239. http://dx.doi.org/10.1016/S0278-6915(96)00064-6.
Potter. CL; Moore. RW; Inhorn, SL; Hagen, TC; Peterson. RE. (1986). Thyroid status and
thermogenesis in rats treated with 2,3,7,8-tetrachlorodibenzo-p-dioxin. Toxicol Appl
Pharmacol 84: 45-55. http://dx.doi.org/10.1016/0041-008X(86)90415-l.
Purdom. CE; Hardiman. PA: Bye. VVJ: Eno. NC; Tyler. CR; Sumpter. JP. (1994). Estrogenic Effects
of Effluents from Sewage Treatment Works. Chem Ecol 8: 275-285.
http://dx.doi.org/10.1080/02757549408Q38554.
Revankar. CM: Cimino. DF; Sklar. LA: Arterburn. JB; Prossnitz. ER. (2005). A transmembrane
intracellular estrogen receptor mediates rapid cell signaling. Science 307: 1625-1630.
http://dx.doi.org/10.1126/science.1106943.
Richardson. DB; Ciampi. A. (2003). Effects of exposure measurement error when an exposure
variable is constrained by a lower limit. Am J Epidemiol 157: 355-363.
Riedmaier. I: Becker. C; Pfaffl. MW; Meyer. HH. (2009). The use of omic technologies for
biomarker development to trace functions of anabolic agents [Review]. J Chromatogr A
1216: 8192-8199. http://dx.doi.Org/10.1016/i.chroma.2009.01.094.
Robaire. B; Ewing. LL; Irby. DC: Desjardins. C. (1979). Interactions of testosterone and estradiol-
17 beta on the reproductive tract of the male rat. Biol Reprod 21: 455-463.
http://dx.doi.Org/10.1095/biolreprod21.2.455.
Rosiak. KL; Seo. BW; Chu. I: Francis. BM. (1997). Effects of maternal exposure to chlorinated
diphenyl ethers on thyroid hormone concentrations in maternal and juvenile rats. J
Environ Sci Health B 32: 377-393. http://dx.doi.org/10.1080/03601239709373Q93.
Rothman. KJ; Greenland. S; Poole. C; Lash. TL. (2008). Causation and causal inference. In KJ
Rothman; S Greenland; TL Lash (Eds.), (3rd edition ed., pp. 5-31). Philadelphia:
Lippincott Williams and Wilkins.
169 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Routledge, EJ; Sheahan, D; Desbrow, C; Brighty, GC; Waldock, M; Sumpter, JP. (1998).
Identification of Estrogenic Chemicals in STW Effluent. 2. In Vivo Responses in Trout and
Roach. Environ Sci Technol 32:1559-1565. http://dx.doi.org/10.1021/es970796a.
Ryan, BC; Hotchkiss, AK; Crofton, KM; Gray, LE. (2010a). In utero and lactational exposure to
bisphenol A, in contrast to ethinyl estradiol, does not alter sexually dimorphic behavior,
puberty, fertility, and anatomy of female LE rats. Toxicol Sci 114: 133-148.
http://dx.doi.org/10.1093/toxsci/kfp266.
Ryan, KK; Haller, AM; Sorrell, JE; Woods. SC; Jandacek, RJ; Seeley, RJ. (2010b). Perinatal
exposure to bisphenol-a and the development of metabolic syndrome in CD-I mice.
Endocrinology 151: 2603-2612. http://dx.doi.org/10.1210/en.2009-1218.
Sawin, S; Brodish, P; Carter. CS; Stanton. ME; Lau, C. (1998). Development of cholinergic
neurons in rat brain regions: Dose-dependent effects of propylthiouracil-induced
hypothyroidism. Neurotoxicol Teratol 20: 627-635. http://dx.doi.org/10.1016/SQ892-
0362(98)00020-8.
Schisterman, EF; Little. RJ. (2010). Opening the black box of biomarker measurement error
[Editorial]. Epidemiology 21 Suppl 4: S1-S3.
http://dx.doi.org/10.1097/EDE.0b013e3181dda514.
Schisterman. EF; Vexler, A; Whitcomb, BW; Liu. A. (2006). The limitations due to exposure
detection limits for regression models. Am J Epidemiol 163: 374-383.
http://dx.doi.org/10.1093/aie/kwi039.
Schmidt. F; Braunbeck, T. (2011). Alterations along the hypothalamic-pituitary-thyroid axis of
the zebrafish (Danio rerio) after exposure to propylthiouracil. J Thyroid Res 2011:
376243. http://dx.doi.org/10.4Q61/2011/376243.
Schmieden PK; Tapper. MA; Denny, JS; Kolanczyk, RC; Sheedy, BR; Henry, TR; Veith, GD. (2004).
Use of trout liver slices to enhance mechanistic interpretation of estrogen receptor
binding for cost-effective prioritization of chemicals within large inventories. Environ Sci
Technol 38: 6333-6342.
Schneider, S; Kaufmann, W; Strauss, V; van Ravenzwaay, B. (2011). Vinclozolin: A feasibility and
sensitivity study of the ILSI-HESI Fl-extended one-generation rat reproduction protocol.
Regul Toxicol Pharmacol 59: 91-100. http://dx.doi.Org/10.1016/i.yrtph.2010.09.010.
Scholze, M; Kortenkamp, A. (2007). Statistical power considerations show the endocrine
disruptor low-dose issue in a new light. Environ Health Perspect 115 Suppl 1: 84-90.
http://dx.doi.org/10.1289/ehp.9364.
Seki, M; Fujishima, S; Nozaka, T; Maeda, M; Kobayashi. K. (2006). Comparison of response to 17
beta-estradiol and 17 beta-trenbolone among three small fish species. Environ Toxicol
Chem 25: 2742-2752.
Seki, M; Yokota, H; Matsubara, H; Maeda, M; Tadokoro, H; Kobayashi, K. (2004). Fish full life-
cycle testing for androgen methyltestosterone on medaka (Oryzias latipes). Environ
Toxicol Chem 23: 774-781.
Seki, T; Ito, S; Adachi, H; Yoshioka, K; Hosokawa, S; Yoshitake, A: Miyamoto, J: Kurata, Y. (1987).
[One-year chronic dietary toxicity study of d.d-T80-prallethrin in rats]. J Toxicol Sci 12:
397-428.
Sekizawa, J. (2008). Low-dose effects of bisphenol A: a serious threat to human health [Review].
J Toxicol Sci 33: 389-403.
170 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Shang, Y; Brown. M. (2002). Molecular determinants for the tissue specificity of SERMs. Science
295: 2465-2468. http://dx.doi.org/10.1126/science.1068537.
Sharlin, PS; Gilbert. ME; Taylor, MA; Ferguson. DC; Zoeller, RT. (2010). The nature of the
compensatory response to low thyroid hormone in the developing brain. J
Neuroendocrinol 22: 153-165. http://dx.doi.Org/10.llll/i.1365-2826.2009.01947.x.
Sharlin. PS; Tighe, D; Gilbert. ME; Zoeller. RT. (2008). The balance between oligodendrocyte and
astrocyte production in major white matter tracts is linearly related to serum total
thyroxine. Endocrinology 149: 2527-2536. http://dx.doi.org/10.1210/en.2007-1431.
Shin. JH; Kim. TS; Kang, IH; Kang, TS; Moon. HJ; Han. SY. (2009). Effects of postnatal
administration of diethylstilbestrol on puberty and thyroid function in male rats. J
Reprod Pev 55: 461-466. http://dx.doi.org/10.1262/jrd.20169.
Sieli, PT; Jasarevic, E; Warzak, PA; Mao. J; Ellersieck, MR; Liao, C; Kannan, K; Collet. SH; Toutain,
PL; Vom Saal, FS; Rosenfeld, CS. (2011). Comparison of serum bisphenol A
concentrations in mice exposed to bisphenol A through the diet versus oral bolus
exposure. Environ Health Perspect 119: 1260-1265.
http://dx.doi.org/10.1289/ehp.10Q3385.
Simmons. SO; Fan. CY; Ramabhadran, R. (2009). Cellular stress response pathway system as a
sentinel ensemble in toxicological screening [Review]. Toxicol Sci 111: 202-225.
http://dx.doi.org/10.1093/toxsci/kfpl40.
Singh. A; Nocerino, J. (2002). Robust estimation of mean and variance using environmental data
sets with below detection limit observations. Chemometr Intell Lab Syst 60: 69-86.
http://dx.doi.org/10.1016/S0169-7439(01)00186-1.
Sinha-Hikim, I; Artaza, J; Woodhouse, L; Gonzalez-Cadavid, IN; Singh. AB; Lee. Ml; Storer, TW;
Casaburi, R; Shen, R; Bhasin, S. (2002). Testosterone-induced increase in muscle size in
healthy young men is associated with muscle fiber hypertrophy. Am J Physiol Endocrinol
Metab 283: E154-E164. http://dx.doi.org/10.1152/aipendo.00502.2Q01.
Sohoni. P; Tyler. CR; Hurd. K; Caunter. J: Hetheridge. M; Williams. T; Woods. C; Evans. M; Toy. R;
Gargas. M; Sumpter. JP. (2001). Reproductive effects of long-term exposure to
Bisphenol A in the fathead minnow (Pimephales promelas). Environ Sci Technol 35:
2917-2925.
Sonich-Mullin. C; Fielder. R; Wiltse. J: Baetcke. K; Pempsey. J: Fenner-Crisp. P; Grant. P; Hartley.
M; Knaap. A: Kroese. P; Mangelsdorf. I: Meek. E; Rice. JM; Younes. M; Safety. IPoC.
(2001). IPCS conceptual framework for evaluating a mode of action for chemical
carcinogenesis. Regul Toxicol Pharmacol 34: 146-152.
http://dx.doi.org/10.1006/rtph.2Q01.1493.
Sonnenschein. C; Olea. IN; Pasanen. ME: Soto. AM. (1989). Negative controls of cell
proliferation: human prostate cancer cells and androgens. Cancer Res 49: 3474-3481.
Standeven. AM; Blazer. PG. Ill; Goldsworthy. TL. (1994a). Investigation of antiestrogenic
properties of unleaded gasoline in female mice. Toxicol Appl Pharmacol 127: 233-240.
http://dx.doi.org/10.10Q6/taap.1994.1157.
Standeven. AM; Goldsworthy. TL. (1994). Identification of hepatic mitogenic and cytochrome P-
450-inducing fractions of unleaded gasoline in B6C3F1 mice. J Toxicol Environ Health 43:
213-224. http://dx.doi.org/10.1080/15287399409531916.
171 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Standeven, AM; Wolf. DC; Goldsworthy, TL. (1994b). Interactive effects of unleaded gasoline
and estrogen on liver tumor promotion in female B6C3F1 mice. Cancer Res 54: 1198-
1204.
Staples. CA; Porn, PB; Klecka, GM; O'Block, ST; Harris. LR. (1998). A review of the environmental
fate, effects, and exposures of bisphenol A [Review]. Chemosphere 36: 2149-2173.
Storer, TW; Magliano, L; Woodhouse, L; Lee. ML; Dzekov, C; Dzekov, J; Casaburi, R; Bhasin, S.
(2003). Testosterone dose-dependently increases maximal voluntary strength and leg
power, but does not affect fatigability or specific tension. J Clin Endocrinol Metab 88:
1478-1485. http://dx.doi.org/10.1210/ic.2002-Q21231.
Stump. DG; Beck. MJ; Radovsky, A; Garman. RH; Freshwater. LL; Sheets. LP; Marty, MS;
Waechter, JM; Dimond. SS; Van Miller. JP; Shiotsuka, RN; Beyer, D; Chappelle, AH;
Hentges, SG. (2010). Developmental neurotoxicity study of dietary bisphenol A in
Sprague-Dawley rats. Toxicol Sci 115: 167-182. http://dx.doi.org/10.1093/toxsci/kfq025.
Takahashi, M; Yoshida, M; Inoue, K; Morikawa, T; Nishikawa, A. (2009). A ninety-day toxicity
study of semicarbazide hydrochloride in Wistar Hannover GALAS rats. Food Chem
Toxicol 47: 2490-2498. http://dx.doi.Org/10.1016/i.fct.2009.07.008.
Takizawa, T; Igarashi, T; Imamizo, H; Ikeda, T; Omichi, K; Kadota, T; Kawata, M; Sakaki, H; Terai,
H; Tsukamoto, 0; Totsuka, K; Handa, J; Hirata, A; Mizuma, H; Murakami, Y; Yamada, M;
Yokouchi, H. (2000). A Study on the Consistency between Flagging by Statistical Tests
and Biological Evaluation. 34: 501-509.
http://dx.doi.org/10.1177/00928615000340Q219.
Thayer, KA; Ruhlen, RL; Howdeshell, KL; Buchanan, PL; Cooke, PS; Preziosi, D; Welshons, WV;
Haseman, J; vom Saal, FS. (2001). Altered prostate growth and daily sperm production in
male mice exposed prenatally to subclinical doses of 17alpha-ethinyl oestradiol. Hum
Reprod 16: 988-996. http://dx.doi.Org/10.1093/humrep/16.5.988.
Thomas, M; Yu, ZG; Padgar, N; Varambally, S; Yu, JJ; Chinnaiyan, AM; Lieberman, AP. (2005).
The unfolded protein response modulates toxicity of the expanded glutamine androgen
receptor. J Biol Chem 280: 21264-21271. http://dx.doi.org/10.1074/ibc.M50014420Q.
Thuett, KA; Roots, EH; Mitchell, LP; Gentles, BA; Anderson, T; Kendall, RJ; Smith, EE. (2002a).
Effects of in utero and lactational ammonium perchlorate exposure on thyroid gland
histology and thyroid and sex hormones in developing deer mice (peromyscus
maniculatus) through postnatal day 21. J Toxicol Environ Health A 65: 2119-2130.
http://dx.doi.org/10.1080/00984100290Q71513.
Thuett, KA; Roots, EH; Mitchell, LP; Gentles, BA; Anderson, TA; Smith, EE. (2002b). In utero and
lactational exposure to ammonium perchlorate in drinking water: Effects on developing
deer mice at postnatal day 21. J Toxicol Environ Health A 65: 1061-1076.
http://dx.doi.org/10.1080/15287390276Q125228.
Tillitt, PE; Papoulias, PM; Whyte, JJ; Richter, CA. (2010). Atrazine reduces reproduction in
fathead minnow (Pimephales promelas). Aquat Toxicol 99:149-159.
http://dx.doi.Org/10.1016/i.aquatox.2010.04.011.
Tinwell. H; Haseman, J; Lefevre, PA; Wallis, N; Ashby, J. (2002). Normal sexual development of
two strains of rat exposed in utero to low doses of bisphenol A. Toxicol Sci 68: 339-348.
http://dx.doi.Org/10.1093/toxsci/68.2.339.
172 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Toft. G; Hagmar, L; Giwercman, A; Bonde, JP. (2004). Epidemiological evidence on reproductive
effects of persistent organochlorines in humans [Review]. Reprod Toxicol 19: 5-26.
http://dx.doi.Org/10.1016/i.reprotox.2004.05.006.
Twaddle. NC; Churchwell, Ml; Newbold, RR; Delclos, KB; Doerge, DR. (2003). Determination
using liquid-chromatography-electrospray tandem mass spectroscopy of ethinylestradiol
serum pharmacokinetics in adult Sprague-Dawley rats. J Chromatogr B Analyt Technol
Biomed Life Sci 793: 309-315. http://dx.doi.org/10.1016/S1570-0232(03)00331-3.
Tyl, RW; Myers, CB; Marr, MC; Castillo. NP; Veselica, MM; Joiner. RL; Dimond. SS; Van Miller. JP;
Stropp, GD; Waechter, JM; Hentges, SG. (2008a). One-generation reproductive toxicity
study of dietary 17beta-estradiol (E2; CAS No. 50-28-2) in CD-I (Swiss) mice. Reprod
Toxicol 25: 144-160. http://dx.doi.Org/10.1016/i.reprotox.2007.ll.012.
Tyl, RW; Myers, CB; Marr. MC; Sloan. CS; Castillo. NP; Veselica. MM; Seely, JC; Dimond. SS; Van
Miller. JP; Shiotsuka, RN; Beyer, D; Hentges. SG; Waechter. JM. (2008b). Two-generation
reproductive toxicity study of dietary bisphenol A in CD-I (Swiss) mice. Toxicol Sci 104:
362-384. http://dx.doi.org/10.1093/toxsci/kfn084.
Tyl, RW; Myers, CB; Marr, MC; Sloan, CS; Castillo, NP; Veselica, MM; Seely, JC; Dimond, SS; Van
Miller, JP; Shiotsuka, RS; Stropp, GD; Waechter, JM; Hentges, SG. (2008c). Two-
generation reproductive toxicity evaluation of dietary 17beta-estradiol (E2; CAS No. 50-
28-2) in CD-I (Swiss) mice. Toxicol Sci 102: 392-412.
http://dx.doi.org/10.1093/toxsci/kfn002.
Tyl, RW; Myers, CB; Marr, MC; Thomas, BF; Keimowitz, AR; Brine, PR; Veselica, MM; Fail, PA;
Chang, TY; Seely, JC; Joiner, RL; Butala, JH; Dimond, SS; Cagen, SZ; Shiotsuka, RN; Stropp,
GD; Waechter, JM. (2002). Three-generation reproductive toxicity study of dietary
bisphenol A in CD Sprague-Dawley rats. Toxicol Sci 68: 121-146.
http://dx.doi.Org/10.1093/toxsci/68.l.121.
U.S. EPA. (1991). Guidelines For Developmental Toxicity Risk Assessment. (600FR91001).
http://nepis.epa .gov/exe/ZyPU RL.cgi?Dockey=2000CA0L.txt.
U.S. EPA. (1996). Strategic Plan for the Office of Research and Development. (600R96059).
http://nepis.epa .gov/exe/ZyPU RL.cgi?Dockey=30005FU N.txt.
U.S. EPA. (1998). Guidelines for ecological risk assessment [EPA Report]. (EPA/630/R-95/002F).
Washington, DC. http://www.epa.gov/raf/publications/guidelines-ecological-risk-
assessment.htm.
U.S. EPA. (1999). Guidelines for carcinogen risk assessment [review draft] [EPA Report]. (NCEA-
F-0644). Washington, DC. http://www.epa.gov/raf/publications/pdfs/CANCER GLS.PDF.
U.S. EPA. (2002). Guidelines for ensuring and maximizing the quality, objectivity, utility, and
integrity, of information disseminated by the Environmental Protection Agency [EPA
Report], (EPA/260/R-02/008). Washington, DC.
http://www.epa.gov/quality/informationguidelines/documents/EPA InfoQualityGuideli
nes.pdf.
U.S. EPA. (2005). Guidelines for carcinogen risk assessment [EPA Report]. (EPA/630/P-03/001F).
Washington, DC. http://www.epa.gov/cancerguidelines/.
U.S. EPA. (2007). ECOTOX user guide: ECOTOXicology database system. Version 4.0. Available
online at http://cfpub.epa.gov/ecotox/ecotox home.cfm (accessed
173 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
U.S. EPA. (2008). White paper: Aquatic life criteria for contaminants of emerging concern.
Washington, DC.
http://water.epa.gov/scitech/swguiclance/standards/criteria/aqlife/cec.cfm.
U.S. EPA (U.S. Environmental Protection Agency). (2005a). Guidelines for carcinogen risk
assessment [EPA Report]. (EPA/630/P-03/001F). Washington, DC.
http://www.epa.gov/cancerguidelines/.
U.S. EPA (U.S. Environmental Protection Agency). (2005b). Health Implications of Perchlorate
Ingestion. Committee to Assess the Health Implications of Perchlorate Ingestion,
National Research Council, http://www.nap.edu/catalog.php7record id=11202.
U.S. EPA (U.S. Environmental Protection Agency). (2009). Endocrine disruptor screening
program test guidelines OPPTS 890.1350 : fish short-term reproduction assay [EPA
Report]. (EPA 740-C-09-007). Washington, DC: USEPA, Office of Chemical Safety and
Pollution Prevention. http://www.regulations.gov/#!documentPetail;P=EPA-HQ-OPPT-
2009-0576-0007.
UNEP (United Nations Environment Programme). (2013). State of the Science of Endocrine
Disrupting Chemicals - 2012. Available online at
http://www.unep.org/hazardoussubstances/UNEPsWork/StateoftheScience/tabid/1059
13/Default.aspx (accessed
USEPA. (2002). A review of the reference dose and reference concentration processes [EPA
Report], (EPA/630/P-02/002F). Washington, DC.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=51717.
van der Ven, LT; van den Brandhof, EJ; Vos, JH; Power. DM; Wester. PW. (2006). Effects of the
antithyroid agent propylthiouracil in a partial life cycle assay with zebrafish. Environ Sci
Technol 40: 74-81. http://dx.doi.org/10.1021/es050972c.
van Raaij. JA; Frijters. CM: van den Berg. KJ. (1993). Hexachlorobenzene-induced
hypothyroidism. Involvement of different mechanisms by parent compound and
metabolite. Biochem Pharmacol 46: 1385-1391.
Vandenberg. LN; Colborn. T; Haves. TB; Heindel. JJ; Jacobs. PR: Lee. DH; Shioda. T; Soto. AM:
vom Saal. FS; Welshons. WV; Zoeller. RT; Myers. JP. (2012). Hormones and endocrine-
disrupting chemicals: low-dose effects and nonmonotonic dose responses [Review].
Endocr Rev 33: 378-455. http://dx.doi.org/10.1210/er.2011-105Q.
Vexler. A: Liu. A: Eliseeva. E; Schisterman. EF. (2008). Maximum likelihood ratio tests for
comparing the discriminatory ability of biomarkers subject to limit of detection.
Biometrics 64: 895-903. http://dx.doi.Org/10.llll/i.1541-0420.2007.00941.x.
Villeneuve. PL: Blake. LS; Brodin. JD; Cavallin. JE; Durhan. EJ: Jensen. KM: Kahl. MP: Makynen.
EA; Martinovic. P; Mueller. NP; Ankley. GT. (2008). Effects of a 3beta-hydroxysteroid
dehydrogenase inhibitor, trilostane, on the fathead minnow reproductive axis. Toxicol
Sci 104: 113-123. http://dx.doi.org/10.1093/toxsci/kfn073.
Villeneuve. PL: Garcia-Reyero. N; Escalon. BL; Jensen. KM: Cavallin. JE: Makynen. EA: Purhan.
EJ: Kahl. MP: Thomas. LM; Perkins. EJ: Ankley. GT. (2012). Ecotoxicogenomics to support
ecological risk assessment: a case study with bisphenol A in fish. Environ Sci Technol 46:
51-59. http://dx.doi.org/10.1021/es201150a.
Villeneuve. PL: Knoebl. I: Kahl. MP: Jensen. KM: Hammermeister. PE; Greene. KJ: Blake. LS:
Ankley. GT. (2006). Relationship between brain and ovary aromatase activity and
174 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
isoform-specific aromatase mRNA expression in the fathead minnow (Pimephales
promelas). Aquat Toxicol 76: 353-368. http://dx.doi.Org/10.1016/i.aquatox.2005.10.016.
Villeneuve. PL; Mueller. ND; Martinovic, D; Makynen, EA; Kahl, MP; Jensen. KM; Durhan, EJ;
Cavallin, JE; Bencic, P; Ankley, GT. (2009). Pirect effects, compensation, and recovery in
female fathead minnows exposed to a model aromatase inhibitor. Environ Health
Perspect 117: 624-631. http://dx.doi.org/10.1289/ ehp.11891.
Vo. TT; Yoo, YM; Choi. KC; Jeung, EB. (2010). Potential estrogenic effect(s) of para bens at the
prepubertal stage of a postnatal female rat model. Reprod Toxicol 29: 306-316.
http://dx.doi.Org/10.1016/i.reprotox.2010.01.013.
Vo. TTB; Jung. EM: Pang. VH; Yoo. YM: Choi. KC: Yu. FH; Jeung. EB. (2009). Pi-(2 ethylhexyl)
phthalate and flutamide alter gene expression in the testis of immature male rats.
Reprod Biol Endocrinol 7: 104. http://dx.doi.org/10.1186/1477-7827-7-104.
vom Saal, FS; Timms. BG; Montano, MM; Palanza, P; Thayer, KA; Nagel, SC; Phar, MP; Ganjam,
VK; Parmigiani, S; Welshons, WV. (1997). Prostate enlargement in mice due to fetal
exposure to low doses of estradiol or diethylstilbestrol and opposite effects at high
doses. PNAS 94: 2056-2061. http://dx.doi.Org/10.1073/pnas.94.5.2056.
Walsh. PC; Swerdloff, RS. (1973). Biphasic effect of testosterone on spermatogenesis in the rat.
11: 190-193.
Welshons. WV; Nagel. SC; Thayer, KA; Judy, BM; Vom Saal. FS. (1999). Low-dose bioactivity of
xenoestrogens in animals: fetal exposure to low doses of methoxychlor and other
xenoestrogens increases adult prostate size in mice. Toxicol Ind Health 15: 12-25.
http://dx.doi.org/10.1191/074823399678846745.
Welshons, WV; Nagel, SC; vom Saal, FS. (2006). Large effects from small exposures. III.
Endocrine mechanisms mediating effects of bisphenol A at levels of human exposure
[Review]. Endocrinology 147: S56-S69. http://dx.doi.org/10.1210/en.2005-1159.
WHO (World Health Organization). (2002). Global assessment of the state-of-the-science of
endocrine disruptors. International Programme on Chemical Safety.
http://www.who.int/ipcs/publications/new issues/endocrine disruptors/en/.
WHO (World Health Organization). (2012). Science of Endocrine Pisrupting Chemicals.
Williams. GR. (2008). Neurodevelopmental and neurophysiological actions of thyroid hormone
[Review]. J Neuroendocrinol 20: 784-794. http://dx.doi.org/10.1111/i. 1365-
2826.2008.01733.x.
Williams. GR: Bassett. JH. (2011). Peiodinases: the balance of thyroid hormone: local control of
thyroid hormone action: role of type 2 deiodinase [Review]. J Endocrinol 209: 261-272.
http://dx.doi.org/10.1530/JQE-10-0448.
Williams. TP: Caunter. JE: Lillicrap. AP; Hutchinson. TH; Gillings. EG: Puffell. S. (2007).
Evaluation of the reproductive effects of tamoxifen citrate in partial and full life-cycle
studies using fathead minnows (Pimephales promelas). Environ Toxicol Chem 26: 695-
707.
Wilson. VS; Lambright. C; Ostby. J: Gray. LE. (2002). In vitro and in vivo effects of 17beta-
trenbolone: a feedlot effluent contaminant. Toxicol Sci 70: 202-211.
http://dx.doi.Org/10.1093/toxsci/70.2.202.
Wolf. CJ; Hotchkiss. A: Ostby. JS; LeBlanc. GA; Gray. LE. Jr. (2002). Effects of prenatal
testosterone propionate on the sexual development of male and female rats: A dose-
response study. Toxicol Sci 65: 71-86.
175 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Wolf. CJ; Ostby, JS; Gray, LE. Jr. (1999). Gestational exposure to 2,3,7,8-tetrachlorodibenzo-p-
dioxin (TCDD) severely alters reproductive function of female hamster offspring. Toxicol
Sci 51: 259-264. http://dx.doi.Org/10.1093/toxsci/51.2.259.
Wolf. DC; Crosby, LM; George. MH; Kilburn, S. R.; Moore. TM; Miller. RT; Deangelo, AB. (1998).
Time- and dose-dependent development of potassium bromate-induced tumors in male
Fischer 344 rats. Toxicol Pathol 26: 724-729.
Wolff. J. (1964). TRANSPORT OF IODIDE AND OTHER ANIONS IN THE THYROID GLAND [Review],
Physiol Rev 44: 45-90.
Wong. C; Kelce, WR; Sar, M; Wilson. EM. (1995). Androgen receptor antagonist versus agonist
activities of the fungicide vinclozolin relative to hydroxyflutamide. J Biol Chem 270:
19998-20003.
Wormuth, M; Scheringer, M; Vollenweider, M; Hungerbuhler, K. (2006). What are the sources
of exposure to eight frequently used phthalic acid esters in Europeans? Risk Anal 26:
803-824. http://dx.doi.Org/10.llll/i.1539-6924.2006.00770.
Wyde, ME; Kirwan, SE; Zhang. F; Laughter. A; Hoffman. HB; Bartolucci-Page, E; Gaido, KW; Yan,
B; You. L. (2005). Di-n-butyl phthalate activates constitutive androstane receptor and
pregnane X receptor and enhances the expression of steroid-metabolizing enzymes in
the liver of rat fetuses. Toxicol Sci 86: 281-290. http://dx.doi.org/10.1093/toxsci/kfi204.
Yamasaki, K; Ishii, S; Kikuno, T; Minobe, Y. (2012). Endocrine-mediated effects of two benzene
related compounds, l-chloro-4-(chloromethyl)benzene and 1,3-diethyl benzene, based
on subacute oral toxicity studies using rats. Food Chem Toxicol 50: 2635-2642.
http://dx.doi.Org/10.1016/i.fct.2012.05.035.
Yamasaki. K; Miyata, K; Shiraishi, K; Muroi, T; Higashihara, N; Oshima. H; Minobe. Y. (2008).
Uterotrophic assay, Hershberger assay, and subacute oral toxicity study of 4,4'-
butylidenebis(2-tert-butyl-5-methylphenol) and 3-(dibutylamino)phenol, based on the
OECD draft protocols. Arch Toxicol 82: 301-311. http://dx.doi.org/10.1007/sQ0204-007-
0250-1.
Yamasaki. K; Noda. S; Muroi. T; Mitoma. H; Takakura. S; Sakamoto. S. (2005). Effects of in utero
and lactational exposure to tamoxifen in SD rats. Toxicol Lett 156: 289-296.
http://dx.doi.Org/10.1016/i.toxlet.2004.12.001.
Yamasaki. K; Takeyoshi. M; Sawaki. M; Imatanaka. N; Shinoda. K; Takatsuki. M. (2003).
Immature rat uterotrophic assay of 18 chemicals and Hershberger assay of 30 chemicals.
Toxicology 183: 93-115.
Yang. H; Zhang. W; Kong. Q; Liu. H; Sun. R; Lin. B; Zhang. H; Xi. Z. (2013). Effects of pubertal
exposure to thiazole-Zn on thyroid function and development in female rats. Food Chem
Toxicol 53: 100-104. http://dx.doi.Org/10.1016/i.fct.2012.ll.003.
Yeatts. SD: Gennings. C; Wagner. ED: Simmons. JE; Plewa. MJ. (2010). Detecting Departure
From Additivity Along a Fixed-Ratio Mixture Ray With a Piecewise Model for Dose and
Interaction Thresholds. Journal of Agricultural, Biological, and Environmental Statistics
15: 510-522. http://dx.doi.org/10.1007/sl3253-010-003Q-x.
York. RG; Barnett. J: Brown. WR: Garman. RH; Mattie. PR: Dodd. D. (2004). A rat
neurodevelopmental evaluation of offspring, including evaluation of adult and neonatal
thyroid, from mothers treated with ammonium perchlorate in drinking water. Int J
Toxicol 23: 191-214. http://dx.doi.org/10.1080/1091581049Q475835.
176 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
York. RG; Brown. WR; Girard, MF; Dollarhide, JS. (2001). Two-generation reproduction study of
ammonium perchlorate in drinking water in rats evaluates thyroid toxicity. Int J Toxicol
20: 183-197. http://dx.doi.org/10.1080/109158101750408Q19.
Yu. KO; Narayanan, L; Mattie, PR; Godfrey, RJ; Todd. PIN; Sterner. TR; Mahle, DA; Lumpkin. MH;
Fisher. JW. (2002). The pharmacokinetics of perchlorate and its effect on the
hypothalamus-pituitary-thyroid axis in the male rat. Toxicol Appl Pharmacol 182:148-
159.
Zhang, B; Chen, Z; Albert, PS. (2012). Latent class models for joint analysis of disease prevalence
and high-dimensional semicontinuous biomarker data. Biostatistics 13: 74-88.
http://dx.doi.org/10.1093/biostatistics/kxr024.
Zhang, L; Wang, J; Zhu, GN. (2010). Pubertal exposure to saisentong: Effects on thyroid and
hepatic enzyme activity in juvenile female rats. Exp Toxicol Pathol 62:127-132.
http://dx.doi.Org/10.1016/i.etp.2009.03.001.
Zhang, X; Hecker, M; Tompsett, AR; Park, JW; Jones, PD; Newsted, J; Au, D; Kong, R; Wu, RS;
Giesy, JP. (2008). Responses of the medaka HPG axis PCR array and reproduction to
prochloraz and ketoconazole. Environ Sci Technol 42: 6762-6769.
Zhang, Y; Chen, B; Ding, X; Jiang, X. (2004a). [Reproductive and developmental toxicity of F1
male rats treated with DBP in utero and during lactation]. Wei Sheng Yan Jiu 33: 9-14.
Zhang, Y; Jiang, X; Chen, B. (2004b). Reproductive and developmental toxicity in F1 Sprague-
Dawley male rats exposed to di-n-butyl phthalate in utero and during lactation and
determination of its NOAEL. Reprod Toxicol 18: 669-676.
http://dx.doi.Org/10.1016/i.reprotox.2004.04.009.
Zoeller, RT; Brown, TR; Doan, LL; Gore, AC; Skakkebaek, NE; Soto, AM; Woodruff, TJ; Vom Saal,
FS. (2012). Endocrine-disrupting chemicals and public health protection: a statement of
principles from the endocrine society. Endocrinology 153: 4097-4110.
http://dx.doi.org/10.1210/en.2Q12-1422.
Zoeller, RT: Dowling, AL; Vas, AA. (2000). Developmental exposure to polychlorinated biphenyls
exerts thyroid hormone-like effects on the expression of RC3/neurogranin and myelin
basic protein messenger ribonucleic acids in the developing rat brain. Endocrinology
141: 181-189. http://dx.doi.Org/10.1210/en.141.l.181.
Zorrilla. LM; Gibson, EK; Jeffay, SC; Crofton, KM: Setzer, WR: Cooper, RL; Stoker, TE. (2009). The
effects of triclosan on puberty and thyroid hormones in male Wistar rats. Toxicol Sci
107: 56-64. http://dx.doi.org/10.1093/toxsci/kfn225.
Citations Not Linked
Li, W; Zha, J; Yang, L; Li, Z; Wang, Z. (2011). Regulation of iodothyronine deiodinases and
sodium iodide symporter mRNA expression by perchlorate in larvae and adult Chinese rare
minnow (Gobiocypris rarus). Marine Pollution Bulletin 63:350-355.
Norris, D.O. 2006. Vertebrate Endocrinology. Academic Press, San Diego, CA.
7. Appendices
177 | Page
-------
Draft Deliberative
State of the Science: Nonmonotonic Dose Responses V7
Appendix A: Mammalian Studies Describing the Effects of Chemicals That Disrupt the
Estrogen Signaling Pathways
Appendix B: Mammalian Studies Describing the Effects of Chemicals That Disrupt the
Androgen Signaling Pathways
Appendix C: Mammalian Studies Describing the Effects of Chemicals That Disrupt the
Thyroid Signaling Pathways
Appendix D: Aquatic Ecotoxicology Studies
178 | Page
------- |