ArticlePDF Available

Development of a putative adverse outcome pathway network for male rat reproductive tract abnormalities with specific considerations for the androgen sensitive window of development

Authors:

Abstract and Figures

Structured approaches like the adverse outcome pathway (AOP) framework offer great potential for depicting complex toxicological processes in a manner that can facilitate informed integration of mechanistic information in regulatory decisions. While this concept provides a structure for organizing evidence and facilitates consistency in evidence integration; the process, inputs, and manner in which AOPs and AOP networks are developed is still evolving. Following the OECD guiding principles of AOP development, we propose three AOPs for male reproductive tract abnormalities and derive a putative AOP network. The AOPs were developed using a fundamental understanding of the developmental biology of the organs of interest, paying close attention to the gestational timing of key events (KEs) to very specifically inform the domain of life stage applicability for the key event relationships (KERs). Chemical stressor data primarily from studies on low molecular weight phthalates (LMWPs) served to ‘bound’ the pathways of focus in this dynamic period of development and were integrated with the developmental biology data through an iterative process to define KEs and conclude on the extent of evidence in support of the KERs. The AOPs developed describe the linkage between 1) a decrease in Insl3 gene expression and cryptorchidism, 2) the sustained expression of Coup-tfII and hypospadias and 3) the sustained expression of Coup-tfII and altered Wolffian duct development/ epididymal agenesis. A putative AOP network linking AOP2 and AOP3 through decreased steroidogenic biosynthetic protein expression and converging of all AOPS at the population level impaired fertility adverse outcome is proposed. The network depiction specifies and displays the KEs aligned with their occurrence in gestational time. The pathways and network described herein are intended to catalyze collaborative initiatives for expansion into a larger network to enable effective data collection and inform alternative approaches for identifying stressors impacting this sensitive period of male reproductive tract development.
Content may be subject to copyright.
Development of a putative adverse outcome pathway network for male rat
reproductive tract abnormalities with specic considerations for the
androgen sensitive window of development
Christine M. Palermo
a,
, Jennifer E. Foreman
b
, Daniele S. Wikoff
c
, Isabel Lea
c
a
ExxonMobil Biomedical Sciences Inc., 1545 US Route 22 East, Annandale, NJ, United States
b
ExxonMobil Chemical Company, 22777 Springwoods Village Parkway, Spring, TX 77389, United States
c
ToxStrategies, Inc., 31 College Place, Suite B118, Asheville, NC 28801, United States
ARTICLE INFO
Keywords:
Male programming window
Phthalate
Adverse outcome pathway
Adverse outcome pathway network
ABSTRACT
Structured approaches like the adverse outcome pathway (AOP) framework offer great potential for depicting
complex toxicological processes in a manner that can facilitate informed integration of mechanistic informa-
tion in regulatory decisions. While this concept provides a structure for organizing evidence and facilitates con-
sistency in evidence integration; the process, inputs, and manner in which AOPs and AOP networks are
developed is still evolving. Following the OECD guiding principles of AOP development, we propose three
AOPs for male reproductive tract abnormalities and derive a putative AOP network. The AOPs were developed
using a fundamental understanding of the developmental biology of the organs of interest, paying close atten-
tion to the gestational timing of key events (KEs) to very specically inform the domain of life stage applica-
bility for the key event relationships (KERs). Chemical stressor data primarily from studies on low molecular
weight phthalates (LMWPs) served to boundthe pathways of focus in this dynamic period of development and
were integrated with the developmental biology data through an iterative process to dene KEs and conclude
on the extent of evidence in support of the KERs. The AOPs developed describe the linkage between 1) a
decrease in Insl3 gene expression and cryptorchidism, 2) the sustained expression of CouptfII and hypospadias
and 3) the sustained expression of CouptfII and altered Wolfan duct development/ epididymal agenesis. A
putative AOP network linking AOP2 and AOP3 through decreased steroidogenic biosynthetic protein expres-
sion and converging of all AOPS at the population level impaired fertility adverse outcome is proposed. The
network depiction species and displays the KEs aligned with their occurrence in gestational time. The path-
ways and network described herein are intended to catalyze collaborative initiatives for expansion into a larger
network to enable effective data collection and inform alternative approaches for identifying stressors impact-
ing this sensitive period of male reproductive tract development.
1. Introduction
Structured frameworks like the adverse outcome pathway (AOP)
framework and associated AOP networks offer great potential for
depicting complex toxicological processes in a manner which can facil-
itate informed integration of mechanistic information in chemical
safety assessments (OECD, 2017, 2018; Villeneuve et al., 2014a). Con-
ceptually, the AOP framework organizes data into a series of biological
events spanning multiple levels of biological organization, providing a
hypothesized path between two anchors: a molecular initiating event
(MIE) and an adverse outcome (AO) (Ankley et al., 2010; Villeneuve
et al., 2014b; OECD, 2018). The pathway is then characterized using
key events (KEs) and key event relationships (KERs) (Villeneuve
et al., 2014a; Villeneuve et al., 2014b; Knapen et al., 2015). AOPs
are often constructed initially as unidirectional linear pathways from
MIE through AO. They are recognized to be oversimplications of
https://doi.org/10.1016/j.crtox.2021.07.002
Received 6 March 2021; Revised 13 July 2021; Accepted 14 July 2021
Available online xxxx
2666-027X/© 2021 ExxonMobil Biomedical Sciences Inc. Published by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Abbreviations: AGD, Anogenital distance; AO, Adverse Outcome; AOP, Adverse Outcome Pathway; DBP, Dibutyl phthalate; DEHP, Di(2ethylhexyl)phthalate; DPP, Dipentyl
phthalate; DHT, 5αdihydrotestosterone; E, Embryonic day (ED1=GD1 gestational day 1); GD, Gestational day (GD1=ED1 embryonic day 1); KE, Key event; KER, Key event
relationship; LMWP, low molecular weight phthalate straight chain length of the esteried alcohols between 3 and 6 carbon atoms; MPW, male programming window.
Corresponding author at: 1545 US Route 22 East, Annandale, NJ 08801, United States.
E-mail address: Christine.palermo@exxonmobil.com (C.M. Palermo).
Current Research in Toxicology 2 (2021) 254271
Contents lists available at ScienceDirect
Current Research in Toxicology
journal homepage: www.elsevier.com/locate/crtox
the complexity of the biological systems which they represent (Knapen
et al., 2018) but aim to depict how an explicit perturbation by any
stressor may lead to an adverse effect (i.e. they are chemically agnos-
tic). Because AOPs, as well as the components of each AOP, cannot be
considered in isolation, the interactions among pathways, stressors,
taxa, and life stages are critical to AOP development and also highlight
the importance of AOP networks. Networks are simply an assembly of
two or more AOPs that share one or more KEs and allow for better
accommodation of more realistic biological scenarios (OECD, 2018;
Knapen et al., 2018). They can depict all of the possible AOs resulting
from a given MIE, or all of the different MIEs through which a given
AO can be caused. Networks acknowledge that even a single stressor
may induce toxicity via multiple mechanisms and/or interact with
multiple targets, the outcome of which may be fully dependent on
the occurrence of the perturbation and sequence of events relative to
windows of biological exposure. In this way, networks can provide a
pragmatic representation of the biological complexity and highlight
all of the possible perturbation points through which stressors may
evoke AOs in a sensitive life stage.
Though AOPs capture the toxicodynamic portion of a toxicity path-
way and are therefore chemically agnostic, focusing on data from one
stressor classcan be a useful means for dening or narrowing the
scope of an AOP/AOP network development effort. The term rat
phthalate syndrome(Foster, 2006) was coined to encompass a group
of effects observed in male rats from exposures to low molecular
weight phthalates (LMWP) (C3C6 backbone) during a critical devel-
opmental window of male sexual differentiation. These effects include
reproductive abnormalities characterized by malformations of the epi-
didymis, vas deferens, seminal vesicles, prostate, external genitalia
(hypospadias), cryptorchidism and testicular injury together with per-
manent changes (feminization) in the retention of nipples/areolae
(sexually dimorphic structures in rodents) and demasculinization of
the growth of the perineum resulting in a reduced anogenital distance
(AGD) (Foster, 2006). Despite decades of research on rat phthalate
syndrome, the understanding of relationships between conditions
and identication of early initiators remains obscure. For LMWPs, it
remains unclear if the multitude of effects are different manifestations
of a shared mechanism, have interrelated but diverging mechanistic
underpinnings, or represent outcomes of distinct biological pathways
that share a common window of developmental susceptibility and/or
common chemical trigger. Clarifying what is known and not know
about the connectivity between biological events and life stage appli-
cability of events during development will facilitate transferability of
the knowledge gleaned from these widely studied substances into
new applications such as informing chemical test strategies and alter-
native methods development.
To facilitate a more generalized and long term utility of the biolog-
ical response pathways embodied in rat phthalate syndrome, the
objective herein was to dene AOPs and a putative network associated
with adverse male reproductive tract outcomes informed by develop-
mental biology and the life stage specicity of KEs. We took a targeted
approach that started with attaining a fundamental biological under-
standing of the pathways that could possibly lead to hypospadias,
cryptorchidism and epididymal agenesis. As these outcomes could
occur via a multitude of pathways, the next step involved evaluation
of the empirical evidence from primarily (but not exclusively) LMWP
data to further inform the causal linkages between KEs. The three
AOPs were subsequently combined in a simple putative AOP network
that aligns with the temporality of the underlying biological processes
specic to the androgen sensitive window of biological development.
Rening life stage applicability was a key focus of this effort and rec-
ommendations for improving information capture for AOPs relevant to
reproductive development were made. The AOP conceptual paradigm
recognizes the need for collaboration, interdisciplinary interaction and
presents itself as a livingconstruct where AOPs and AOP networks
are modied, rened, and expanded as understanding and knowledge
advances (Villeneuve et al., 2014a). The network described herein is
meant to serve as a starting point for network expansion efforts to
include additional MIEs and AOs relevant to this developmental win-
dow. Simplifying the complex developmental biology into a pragmatic
AOP network is an important rst step towards identication of the
critical pathways and critical few network nodules important for
advancing alternative methods development for this susceptible life
stage.
2. Approach
2.1. AOP development
The focus of this effort was to develop AOPs in a manner compliant
with the principles and guidance set out by the OECD (OECD, 2018).
Systematic review guidance for AOP development was not available
at the onset of the project, and is, at the time of publication, still being
drafted. Systematic methods were considered and initially employed;
however, given the broad scope of the assessment, it was determined
that fully implementing such methods was not feasible. Rather, a tar-
geted approach which built on the existing knowledge base was used
to further elucidate the putative AOPs herein.
To dene KEs and KERs in male rat reproductive tract abnormality
pathways, two comprehensive literature searches were conducted in
PubMed. Both search strategies captured rodent studies in primary
and secondary (i.e. review) articles with no date exclusion applied.
Citation mining was performed when relevant review articles were
identied. The rst search captured mechanistic data related to devel-
opment of the reproductive tract in the rodent male programming win-
dow (MPW) and used key terms (e.g. development, hypospadias,
cryptorchidism, epididymal malformations) for the target organs (epi-
didymis, Wolfan duct, penis, testis). Articles of relevance were those
providing mechanistic insight to the normal and abnormal develop-
ment of the target organs of interest during gestation. The second
search captured studies evaluating the effects of LMWP exposure dur-
ing the MPW and used search terms such as phthalate and male repro-
ductive tract or hypospadias, or cryptorchidism, or androgen, or
testosterone, or epididymal agenesis, or testicular dysgenesis; or
phthalate syndrome. Articles were selected when the exposure
spanned the MPW, reported effects on the target organs, and/or pro-
vided mechanistic insight into phthalate mediated developmental tox-
icity. In some cases, studies reporting effects on the target organs from
exposures outside of the MPW were also captured to inform on the life
stage specicity of the KERs. Subsequent targeted investigative
searches were performed (in both the biology and toxicology elds)
to inform critical decision points (as described more below).
To establish AOPs, the molecular and cellular events that may lead
to cryptorchidism, hypospadias and epididymal malformations were
identied from the development biology literature and categorized
by level of biological organization and gestational timing. The timing
of in utero events are referred to as both embryonic day (E) and gesta-
tional day (GD) in the literature and these designations were used
interchangeably in this manuscript, as to retain the designations uti-
lized in the relied upon references. Based on the fundamental under-
standing of biology, more than one MIE could be responsible for
mediating the AOs. Therefore we narrowed our focus to dening the
KERs applicable to rat phthalate syndromeby initially considering
empirical evidence from the LMWP literature. Using an iterative pro-
cess of integrating data, KEs from the developmental biology literature
were ruled inwhen evidence from the stressor literature supported it;
or KEs were ruled outwhen counter evidence was present. When evi-
dence for a KE in relation to the entire AOP was minimal or con-
founded, targeted investigative literature searches were performed
and supporting information was assembled from either the develop-
C.M. Palermo et al. Current Research in Toxicology 2 (2021) 254271
255
mental biology literature and/or nonLMWP chemical stressors shown
to induce the AOs of interest.
2.2. Evidence integration and weight of evidence evaluation
Condence in each AOP was established using the weight of evi-
dence (WoE) considerations per the recommendations in the OECD
AOP Developers Handbook (OECD, 2018). The rst type of evidence
assembled as part of the AOP evaluation was biological plausibility.
Biological plausibility refers to the extent the relationship between
an upstream and downstream KE is supported by an understanding
of normal biological function (versus a response to a specic stressor).
Weight of evidence in support of biological plausibility for the KERs
was assessed consistent with the OECD handbook (OECD, 2018).
High: KER is well understood based on extensive previous docu-
mentation; established mechanistic basis and broad acceptance.
Moderate: KER is plausible based on analogy to accepted biological
relationships but biological relationship/scientic understanding not
completely established.
Low: empirical support for an association exists between KEs but
structural or functional relationship between KEs is not understood.
The second type of evidence assembled as part of the AOP evalua-
tion was the extent of empirical evidence for the KERs. The level of
empirical concordance for the KERs was assessed by determining
whether upstream KEs occurred at lower doses and earlier time points
than downstream KEs and/or the frequency and incidence of an
upstream KE was greater than the downstream KE at the same dose.
A quantitative characterization of the KERs was out of scope of this
effort. Rather focus was given to the qualitative evidence supporting
dose, temporal and incidence concordance of KERs (taking care to
identify where evidence was limited), and particularly to the concor-
dance across multiple KEs in support of the AOP. Weight of empirical
evidence for the KERs and AOP was assessed consistent with the OECD
handbook (OECD, 2018)asdened below:
High: there is dependent change in events following exposure to a
wide range of stressors (extensive evidence for dose, temporal and
incidence concordance) and few data gaps or conicting data. In this
case, dose, temporal and incidence concordance had to be demon-
strated in a welldesigned study or when combining data across multi-
ple studies for at least three LMWPs. The number of studies reporting a
dose response effect of both observations for at least three LMWPs was
also considered.
Moderate: dependent change in events is demonstrated following
exposure to a small number of specic stressors and some evidence
inconsistent with the expected pattern that can be explained by factors
such as experimental design. In this case, dose temporal and incidence
concordance demonstrated in a welldesigned study for at least one
LMWP. The number of studies reporting a dose response effect of both
observations for two or more LMWPs was also considered.
Low: dependent change in events following exposure to a specic
stressor is identied in limited or no studies (i.e. endpoints never mea-
sured in the same study or not at all), and/or lacking evidence of tem-
poral or doseresponse concordance, or identication of signicant
inconsistences in empirical dataset that do not align with the expected
pattern for the hypothesized AOP.
Indirect (or other) observational evidence was also relied upon to
inform the extent of empirical evidence for KERs. This type of evidence
considered how the pattern of observations within the collective and
broad data set (e.g. across animal strains, stressors) supported or
refuted a given KER. Indirect evidence was considered in addition to
the empirical concordance determinants above for an overall conclu-
sion on weight of empirical evidence.
The biological plausibility and empirical evidence ratings were
assessed together to conclude on the overall support for the AOP. Bio-
logical plausibility ranks highest among the different types of evidence
used to support the AOPs (OECD, 2017, 2018) as it requires an under-
standing of the structural and functional relationships in normal biol-
ogy and therefore weighted higher in the overall assessment. Empirical
evidence primarily reected dose response/incidence and temporal
relationship data from studies with LMWP and therefore served to
increase condence in the AOP. The available data did not allow for
assessment of the condence in supporting data for essentiality of
KEs within the AOP. Experimental approaches for establishing essen-
tiality of KEs, such as genetic knockout animal models or genetic over-
expression models, underpin the basic developmental biology
understanding and, as such, were part of the biological plausibility
evaluation.
2.3. AOP network development
To establish an AOP network that respects the gestational timing of
KE/KERs, each AOP was evaluated for conuence of KEs, considering
both the nature of the KE as well as its temporal occurrence in devel-
opment. AOPs were linked through putative shared KEs as the basis to
dene an AOP network.
The numbering for the MIEs, KEs, and AOs described herein refer to
the AOP network shown in Fig. 4.Supplementary Material provide
detailed descriptions of empirical evidence and biological plausibility
for KEs/KERs to facilitate independent assessment of the interpreta-
tions drawn in this analysis.
3. Results
3.1. AOPs
Male reproductive development is highly dynamic and complex. As
such, many perturbations of the system could possibly lead, in a causal
manner, to the AOs of interest. To narrow the scope, attention was
given to molecular and cellular perturbations during the male pro-
gramming window (MPW) of development (GD15.518.5) (Sharpe,
2020). LMWP literature largely shows the importance of GD
15.518.5 as the window in which the initiating event(s) that leads
to LMWP mediated AOs occurs (Howdeshell et al., 2017;
Kortenkamp, 2020); however, the timing of the other KEs in the out-
come pathways has been less rigorously dened. The time between
KEs in male reproductive development could be milliseconds, days
or in some instances months before the AO is realized. The alignment
of the KEs and KERs in the three AOPs described below aims to respect
gestational timing by carefully considering the window of chemical
exposure and life stage specicity of when observational measure-
ments were taken in the studies relied upon to build the AOP. In this
way, for example, testosterone levels measured in adulthood would
not be considered to provide much value in informing on hormonal
changes during gestation.
3.1.1. AOP1: INSL3 decrease to cryptorchidism
Cryptorchidism is an absence of one or both testes from the scro-
tum and reects an alteration in the normal process of testicular des-
cent. Testicular descent is characterized as a 2stage process, with
each phase regulated by different hormones and anatomical mecha-
nisms. The rst phase (transabdominal phase) beings around GD15.5
in rodents (Hutson et al., 1997, 2013, 2016; Klonisch et al., 2004)
and is characterized by regression of the cranial suspensory ligament
(CSL) and swelling of the gubernaculum to position the testes at the
bottom of the abdomen. In this phase, CSL regression is initiated by
testosterone (Amann and Veeramachanemi, 2007; Van der Schoot
and Emmen, 1996), and the swelling reaction in the gubernaculum
is stimulated primarily by insulinlike hormone 3 (INSL3) with some
studies implicating secondary augmentation by androgens (Adham
et al., 2002; Bogatcheva et al., 2003; Gorlov et al., 2002; Nef et al.
1999; Overbeek et al. 2001; Tomiyama et al., 2003; Zimmerman
C.M. Palermo et al. Current Research in Toxicology 2 (2021) 254271
256
et al., 1999). The second phase of testicular descent (inguinoscrotal
process) is largely controlled by androgens and occurs at approxi-
mately postnatal week 3 and involves movement of the testes from
the bottom of the abdomen to the base of the scrotum (Klonisch
et al 2004, Zimmerman et al., 1999).
In rats, gestational exposure during the MPW to dibutyl phthalate
(DBP) has been associated with increased incidence of cryptorchidism
(Welsh et al., 2008; van den Driesche et al., 2017; van den Driesche
et al., 2020). van den Driesche and colleagues (2017) exemplied
the window of susceptibility for initiating this AO showing exposure
to DBP during GD 15.518.5 only resulted in cryptorchidism, whereas
exposure to DBP during GD 19.521.5 did not (Fig. 1). This narrow
window of susceptibility implies the hormonal and anatomical mech-
anisms of relevance in deriving a pathway initiated by LMWPs and
leading to cryptorchidism are restricted to the transabdominal phase
of testicular descent (Hutson et al., 1997; Klonisch et al., 2004;
Amann and Veeramachaneni, 2007; Nation et al., 2011). Evaluation
of the events occurring during this phase of development as well as
observational data on LMWPs suggest a pathway with two KEs initi-
ated by an unknown MIE leading to one AO: 1) decreased Insl3 gene
expression; 2) agenesis of the gubernaculum ligaments; 3) cryp-
torchidism (AO) (Fig. 1).
The rst known event in this pathway was decreased Insl3 gene
expression in Leydig cells of the testis (KE3) (Fig. 1). Developmental
biology studies support an essential role of this testicular factor in nor-
mal gubernacular outgrowth and testicular descent (Nef and Parada,
1999; Zimmermann et al., 1999). INSL3 interacts with relaxinfamily
peptide receptor 2 (RXFP2) to stimulate gubernacular growth through
an unclear mechanism (Bogatcheva et al., 2003). Developmental
expression of Insl3 transcripts coincides with the onset of testicular
descent with the transcript rst detected as early as GD13.5 in mice
(Feng et al., 2009). INSL3 protein is rst detectable in rats at
GD15.5, expression peaks at GD17.519 before returning to low levels
at GD 21.5 (McKinnell et al., 2005). Very little is known about the reg-
ulation of Insl3 in development hindering the ability to hypothesize
the upstream MIE for AOP1. One study implicated COUPTFII as being
an important regulator of Insl3 in Leydig cells (MendozaVillarroel
et al., 2014). In this study, a signicant decrease in Insl3 mRNA levels
and an absence of testosterone were apparent in mice with prepubertal
depletion of COUPTFII. LMWPs, however, are associated with
increased expression of COUPTFII (van den Driesche et al., 2012)
casting doubt on the role of COUPTFII as the MIE for this pathway.
Two other transcription factors, SF1 and DAX1, have also been
shown to affect mouse Insl3 expression in vitro.SF1 enhanced Insl3
promoter activity and increased transcription whereas, DAX1 inhib-
ited promoter activity and decreased transcription of Insl3 (Adham
et al., 2004). In the rat testis, Insl3 has a relatively short promoter
region characterized by three putative binding sites for SF1 transcrip-
tion factors (Ivell and Bathgate, 2002; Sadeghian et al., 2005). It is
interesting to speculate that displacement of SF1 from the Insl3 pro-
moter via COUPTFII might create a common thread with AOP2; how-
ever data and mechanistic understanding are currently too limited to
support this as the MIE for AOP1.
The level of condence in the KERs in AOP1 based on biological
plausibility is summarized in Table 1. The condence in the KERs
downstream of the MIE is high, based on well documented mechanistic
understanding in the developmental biology literature that strongly
support a role for INSL3 and altered gubernacular development during
the transabdominal phase of testicular descent (Table 1). There is also
extensive literature describing and evaluating the ability of LMWPs to
induce cryptorchidism and the general consensus in this literature con-
cludes a role for INSL3 in mediating the effects (Foster, 2006; Wilson
et al., 2008; Howdeshell et al., 2015; Gray et al., 2016). The develop-
mental biology literature supports well the proposed KERs and a high
degree of condence in AOP1.
The extent of empirical support for the KERs in AOP1 is summa-
rized in Table 1 and range from moderate to high. Though a dose
responsive effect following exposure to a small number of LMWPs is
supported for KE3, KE6 and AO1 (KE increased in incidence/severity
with increasing dose of phthalate, See Supplemental Table 1), empiri-
cal evidence to support concordance of KERs downstream from the
MIE ranged from low to high. One study with DPP, shows strong dose,
temporal and incidence concordance between Insl3 expression, guber-
naculum effects and testicular descent (Table 2) (Grey et al., 2016),
supporting high condence in the AOP. As all KEs were very rarely
measured in a single study, concordance of KEs for a wider range of
LMWPs was assessed by compiling observations from multiple studies
(Supplemental Table 2). The low frequency of observations for KE6 in
this data set was a notable limitation in the condence assessment of
KER concordance. Variation in experimental design, measurement
methods, and measurement time points also challenged the calibration
of these data. Overall, concordance between KEs was moderately sup-
ported when combining data across multiple studies for DBP and
DEHP, and not well supported for BBP (likely due to limited data).
Inguinoscrotal descent
Gonad to testis
Unknown
(MIE)
Agenesis of
gubernacular
ligaments
(KE6)
Cryptorchidism
(AO1)
Decreased
Insl3 gene
expression
(KE3)
E15.5 E16.5 E17.5 E18.5
Transabdominal phase of testicular descent
64% Cryptorchidism
PND19
E13.5
Testicular
differentiation
begins
0% Cryptorchidism
E19.5 E20.5 PND6
INSL3 peak
E17
750 mg/kg DBP
Fig. 1. INSL3 reduction to cryptorchidism pathway (AOP1). Leydig cells (in the testes) are the cellular target for KE3, the gubernaculum is the target organ for
KE6, and the testes the target organ for the AO, cryptorchidism (AO1). KEs outlined in blue are proposed to occur during the MPW. Events are aligned to the
developmental timeline such that events sharing the same vertical plane occur in the same developmental window. The horizontal planes depict: 1) biological
process occurring, 2) embryonic day (E), 3) AOP, and 4) subset of data that supports the pathway being specic to the indicated developmental window. This data
demonstrates exposure to DBP during the MPW (lled red bar), but not later in gestation (open red bar) induces cryptorchidism (van den Driesche et al. 2017).
Note: KEs are numbered according to the network (Fig. 4). (For interpretation of the references to colour in this gure legend, the reader is referred to the web
version of this article.)
C.M. Palermo et al. Current Research in Toxicology 2 (2021) 254271
257
Indirect (or other) empirical evidence from the broad biological data
set depicts a pattern of observations consistent with the KERs and
includes 1) Wistar rats exposed to LMWPs had a higher incidence of
gubernacular lesions and lower Insl3 mRNA levels compared to Spra-
gue Dawley rats with a higher level of Insl3 and no gubernacular
lesions; 2) LMWPexposed rat phenotype was identical to that
observed in Insl3 KO mice which displayed freely moving testes that
were not attached to the abdominal wall by the CSL or gubernaculum;
3) undescended testes and gubernacular malformations cooccurred in
adult male SpragueDawley rats exposed in utero to a mixture of
phthalates (Supplemental Table 3). In totality, these evidence support
moderate condence in the AOP based on consideration of empirical
concordance and indirect observational evidence.
Considering the major decision points and uncertainties in this
pathway: in the transabdominal phase of testicular descent, the posi-
tion of the testis is a function of both the swelling of the gubernaculum
and the regression of the CSL (Klonisch et al., 2004, Nation et al.,
2011). Early biology studies implicate a role for androgens in CSL
regression and propose inappropriate outgrowth of the CSL could lead
to cryptorchidism (van der Schoot and Emmen, 1996; Emmen et al.,
1998). Because LMWPs impact androgens during GD 15.518.5, it is
also important to consider an alternative AOP where reduced andro-
gen and impaired CSL regression could lead to cryptorchidism. Analy-
sis of the available literature show comparatively little evidence to
support a role for the CSL (Supplemental Table 4). The LMWP litera-
ture does not substantiate well, a link between stunting of the CSL
and transabdominal testicular descent failure although there are only
a few studies that capture CSL observations (Gray et al., 2000; Gray
et al., 2009); moreover, the developmental biology literature does
not support well impaired CSL regression as a primary cause of cryp-
torchidism during the transabdominal phase (van der Schoot and
Emmen, 1996; Emmen et al., 1998; Adham and Agoulnik, 2004;
Kassim et al., 2010; Mamoulakis et al., 2015). It is also important to
consider an alternative AOP where reduced androgen could work in
concert with INSL3 or alone to cause gubernacular agenesis. Particu-
larly as the severity of LMWP mediated cryptorchidism has been cor-
related with greater androgen impact in some studies (Howdeshell
et al., 2015; van den Driesche et al., 2017). However the pattern of
observations in the collective LMWP data set are inconsistent with a
causal link between testosterone (Supplemental Table 4). Most nota-
bly, experiments in different rat strains (Supplemental Table 3) show
a strainspecic sensitivity to LMWP mediated cryptorchidism impli-
cating INSL3 rather than testosterone in testicular descent (Wilson
et al., 2007). LMWP studies in SpragueDawley rats show a greater
reduction in testosterone levels is associated with lower rates of cryp-
torchidism as compared to studies in Wistar rats demonstrating a
greater impact on INSL3 and a higher incidence of cryptorchidism
and gubernacular lesions (Mylchreest et al., 1998; Mylchreest et al.,
1999; Mylchreest and Foster, 2000; Mylchreest et al., 2000;
McKinnell et al., 2005; Wilson et al., 2007). The developmental biol-
ogy literature also supports a change in testosterone is not necessary
to alter gubernacular development. Developmental biology evidence
does show testosterone plays an important role in the postnatal
Table 1
Assessment of biological plausibility and empirical evidence in support of the KERs in AOP1.
KER
(Adjacency of KEs)
WoE Conclusion
WoE Rationale
Biological Plausibility WoE
KE3 leads to AO1 (Indirect)
High
Established developmental biological knowledge based on molecular biology studies (e.g. Insl3 knockout studies) support INSL3 is necessary
in promoting normal testicular descent (Overbeek et al., 2001; Gorlov et al., 2002; Bogatcheva et al., 2003; Tomiyama et al., 2003; Rugarli
and Langer, 2012; Imaizumi et al., 2015).
KE3 leads to KE 6
(Direct)
High
Established developmental biological knowledge based on molecular biology studies (e.g. Insl3 knockout and mutation studies) support
INSL3 disruption impairs gubernacular development (Nef and Parada, 1999; Zimmermann et al., 1999; Overbeek et al., 2001; Adham et al.,
2002; Gorlov et al., 2002; Bogatcheva et al., 2003; Tomiyama et al., 2003).
KE6 leads to AO1 (Direct)
High
Established developmental biological knowledge based on targeted molecular biology studies and observational associations support
abnormal gubernacular development impedes testes descent (Ivell and Hartung, 2003; Adham and Agoulnik, 2004; Klonisch et al., 2004;
Amann and Veeramachaneni, 2007; Hutson et al., 2013; Hutson et al., 2016; Hadziselimovic, 2017).
Established biological knowledge that the higher testicular temperature environment resulting from the physical location of cryptorchid
testes (away from the scrotum) causes germ cell depletion and infertility (Setchell, 1998).
Empirical Evidence WoE
KE3 lead to AO1 (Indirect)
High
Dose, temporal and incidence concordance well supported in a single study on DPP (Table 2), and for a small number of LMWPs when
combining data across multiple studies (Supplemental Table 2: BBP data limited, DBP and DEHP consistent with KER). Pattern of
observations within a broader biological dataset supports the KER (Supplemental Table 3). A dependent change in both events is observed in
the same study for a number of LMWPs (Supplemental Table 1).
KE3 leads to KE6 (Direct)
Moderate
Dose, temporal and incidence concordance well supported in a single study on DPP (Table 2), support for concordance is challenged across a
small number of LMWPs when combining data across multiple studies (Supplemental Table 2: BBP data limited, DBP consistent and DEHP
inconsistent but data limited). Pattern of observations within the broader dataset supports the KER (Supplemental Table 3). A dependent
change in both events is observed in a same study for a number of LMWPs (Supplemental Table 1). Inconsistent data possibly implicating
failed CSL regression (and not gubernacular agenesis) in the LMWP literature (Supplemental Table 4) is comparatively minimal to the
evidence supporting involvement of the gubernaculum.
KE6 leads to AO1 (Direct)
Moderate
Dose, temporal and incidence concordance well supported in a single study on DPP (Table 2). Support for concordance is challenged across a
small number of LMWPs when combining data across multiple studies (Supplemental Table 2: BBP and DEHP consistent with the KER; DBP
inconsistent but data limited). A dependent change in both events is observed in the same study for a number of LMWPs (Supplemental
Table 1). Inconsistent data possibly implicating failed CSL regression (see above).
KE3 = decreased Insl3 expression; KE6 = agenesis of the gubernaculum; AO1 = cryptorchidism; KER = key event relationships; WoE = weight of evidence.
Table 2
Dose, temporal and incidence concordance of KEs in AOP1. Observations as
reported by Gray et al. (2016) on dipentyl phthalate (DPP) support concordance
of KEs.A dash () indicates no effect; a plus (+) indicates effect observed;
number of pluses indicates increased severity/incidence. Severity/incidence of
effect weighting are provided as part of Supplemental Table 2. An AOP
supported by empirical data shows early events occurring at lower doses and
with higher severity than later events. GD = gestational day; mo. = months.
Temporal Concordance
Dose
Cecnadr
ocno
Dose
mg/kg/d
KE 3
(GD18)
KE 6
(6-7 mo.)
(6-7 mo.)
0
-
-
11
-
-
33
+
-
100
++
++
300
+++
+++
C.M. Palermo et al. Current Research in Toxicology 2 (2021) 254271
258
inguinoscrotal phase of testis descent (Nation et al., 2009; Hutson
et al., 2013; Hutson et al., 2016; Hadziselimovic, 2017) as is described
in AOP288 (https://aopwiki.org/); however, the life stage specicity
of AOP288 to postnatal development excludes the relevance of these
KEs to cryptorchidism initiated during the MPW. Based on limited evi-
dence, a role for the CSL was ruled out.
Considering the pathway as a whole, the extent of evidence is con-
cluded as high in support of AOP1 upon consideration of high biolog-
ical plausibility, moderate empirical evidence, and uncertainties.
Condence is high in support of the life stage relevance of this AOP
during fetal life and more specically the gestational timing of the
MIE, KE1 and KE6 to the MPW of fetal development (van den
Driesche et al., 2017). Despite the unknown MIE, condence in this
AOP for use in regulatory application is considered high based on
the condence in the downstream KERs, however clarity on the MIE
may improve advancement of alternative approaches capable of
screening or identifying developmental stressors.
3.1.2. AOP2: sustained expression of coup-TFII to hypospadias
Development of the anogenital tract is a complex process which is
temporally and spatially dependent (Supplemental Fig. 1). Signaling
needs to occur in a complex continuum to have appropriate differenti-
ation of the anorectal tract and genitourinary tract, as well as appropri-
ate growth of external genitals. In the early stages coordinated
signaling across time is hormonally independent, and needs to occur
in various disparate structures simultaneously for normal development
to occur. Later, signaling becomes hormonally responsive in order to
sexually differentiate the genital tract into the male or female form.
Disruption of this process can lead to various malformations of differ-
ing severity depending upon what signals are disrupted and at what
time point. The precursor to the external genital is the genital tubercle
(GT). Growth of the GT consists of two phases. The rst phase consists
of the initial outgrowth and patterning of the GT, which occurs in both
males and females, and results in the bipotential GT (Miyagawa et al.,
2009; Petiot et al., 2005). This phase occurs from roughly GD10.5
when the genital swellings rst appear to GD15/15.5 which is the start
of sexual differentiation of the GT (Ipulan et al., 2014; Murishma et al.,
2015). The end of Phase I coincides with the start of the male program-
ing window (MPW) (Petiot et al., 2005; Welsch et al., 2008; van den
Driesche et al., 2012; Miyagawa et al., 2009). During the rst phase
growth and patterning occurs the same in males and females, with
no, as yet, discernable differences. The second phase is hormonally
mediated and entails either continued growth and differentiation of
the penis, or the arrest of outgrowth and differentiation into the cli-
toris (Petiot et al., 2005; Suzuki et al., 2014).
Of interest to this assessment were processes specic to develop-
ment of the external male genitalia and the disruptions leading to
hypospadias. Penile malformations elicited by prenatal exposure to
LMWPs in rats are typied by the development of hypospadias arising
from a proximal shift in the urethral meatus (Supplemental Table 1;
(Foster, 2006). Using a rat model, gestational exposure to DBP during
the MPW was sufcient to cause hypospadias, with gestational expo-
sures outside this gestational window unable to elicit this abnormality
(van den Driesche et al., 2017). A similar effect pattern is observed
after treatment with androgen receptor (AR) agonist utamide
(Foster and Harris, 2005) and 5αreductase inhibitor nasteride
(Clark et al., 1993) where treatment at GD17/GD18 and GD1617,
respectively, induced hypospadias but treatment earlier or later did
not. Even though each stressor operates via a distinct molecular initi-
ating event, exposure of each only induces hypospadias during a short
developmental time window in rats, which supports the developmen-
tal occurrences during this window as being critical androgen medi-
ated events in normal penile development.
This narrow window means the anatomical mechanisms of rele-
vance include growth of the GT as well as the formation of the internal
urethra and proper positioning of the urethral meatus at the tip of the
penis (Georgas et al., 2015). Evaluation of molecular events occurring
during urethral development together with observational data on
LMWPs suggest a pathway with four KEs initiated by an unknown
MIE leading to one AO: 1) sustained COUPTFII expression in Leydig
cells; 2) decreased steroidogenic biosynthetic protein expression
(StAR, CYP11A1, CYP17); 3) decreased 5αdihydrotestosterone
(DHT); 4) inhibition of urethral tube closure (decrease in Mafb expres-
sion) and 5) hypospadias (Fig. 2).
Fig. 2. Sustained expression of COUP-TFII to hypospadias pathway (AOP2). Leydig cells are the cellular target for KE1 & KE2, the genital tubercle is the target
organ for KE5 & KE7, and the penis is the target organ for the AO, hypospadias (AO2). KEs outlined in blue are proposed to occur during the MPW. Events are
aligned to the developmental timeline such that events sharing the same vertical plane occur in the same developmental window. The horizontal planes depict: 1)
biological process occurring, 2) embryonic day (E), 3) AOP, and 4) subset of data that supports the pathway being specic to the indicated developmental window.
These data demonstrate a single 50 mg/kg dose (Foster and Harris 2005) or repeated 100 mg/kg dosing (Welsh et al., 2008) of the AR antagonist utamide needs
to occur within the MPW to induce hypospadias. Repeat dosing with 750 mg/kg/d DBP (van den Driesche et al., 2017) leads to a high incidence of hypospadias
when delivered during the MPW. Exposure during the late gestational window (open bar) result in a low incidence (in the range of background) of hypospadias
(4.3%). Note: KEs are numbered according to the network (Fig. 4). (For interpretation of the references to colour in this gure legend, the reader is referred to the
web version of this article.)
C.M. Palermo et al. Current Research in Toxicology 2 (2021) 254271
259
The rst proposed event in this pathway is sustained expression of
COUPTFII in Leydig cells of the testis (KE1). Greater than 85% of Ley-
dig cells are COUPTFII positive at E15.5, the beginning of the male
programing window and roughly the time when testosterone produc-
tion begins to increase. Following E15.5 there is a decrease in Coup
TFII positive Leydig cells with only ~10% of the cells remaining posi-
tive by E21.5 which is consistent with the increasing levels of andro-
gen during this developmental window (van den Driesche et al
2012). The highest levels of COUPTFII positive cells precede steroid
synthesis and increases in steroid levels are dose dependently associ-
ated with a decline in COUPTFII expression, supporting the potential
for a causal relationship between the decline in COUPTFII and
steroidogenesis. COUPTFII has been shown to act as a transcriptional
repressor of gene expression, possibly through competition for occu-
pancy of binding sites. The ability of COUPTFII to suppress SF1 acti-
vation of KE2 (Shibata et al., 2003) genes is central to the proposed
AOP. Plummer et al. (2007) demonstrated impacts on the expression
of steroidogenesis genes by DBP exposure only during the time period
when COUPTFII is decreasing during normal development (GD 17.5
and GD 19.5) with no effect on GD15.5 before COUPTFII expression
normally declines in Leydig cells. COUPTF II is an orphan nuclear
receptor which interacts with the corepressors NCoR, SMRT and
RIP13, to silence transcription by active repression and transrepression
(Qiu et al., 1994; Bailey et al., 1997). LMWPs have been shown to acti-
vate some members of the nuclear receptor superfamily, (i.e. PPARs)
(Bility et al., 2004). The MIE may be specic ligand activation of
COUPTFII, or potential disruption of required cofactors. Further
research is required to identify precise MIE for COUPTFII sustained
expression.
One of the important features of this pathway is that androgen
effects are mediated by DHT rather than testosterone and this is sup-
ported by an abundance of literature demonstrating a role for DHT
in urethral tube development (Table 3). The site of androgen synthesis
changes during development with androgen precursors rst synthe-
sized in fetal Leydig cells on GD 14.5 in rodents when cholesterol
was transported to the mitochondria by steroidogenic acute regulatory
protein (StAR). In fetal Leydig cells, lack of a key conversion enzyme
(HSD17β3) means conversion of cholesterol to testosterone occurs out-
side the Leydig cells (Wen et al., 2016). Conversion of testosterone to
the more potent DHT occurs at the site of DHT action (i.e. DHT target
tissues). This information supports that testosterone is not present in
the biological compartment (i.e. Leydig cells) specied for KE1 and
KE2 during this life stage. As testosterone is not the active androgen
responsible for the differentiation of the genital tubercle (target organ)
for this AOP, it was not included as a KE in the AOP for hypospadias.
The proposed AOP aims to depict the critical few events necessary to
progress from the unknown MIE to AO. Assessment of testosterone
could be used as an indirect measurement of KE2 or KE5, as it is one
of a few intermediary steps between these KEs. Anogenital distance
(AGD), however, is considered a more reliable indicator of DHT levels
during the male programming window. Using anogenital distance
(AGD) as a biomarker gives an assessment of DHT level throughout
the whole MPW window (McIntyre et al., 2001). It is important to
emphasize that while AGD is an important indicator of the DHT envi-
ronment during development (Dean et al., 2012; Dean and Sharpe,
2013; Suzuki et al., 2015) a change in AGD is not a KE in this AOP
because a reduction in AGD is not essential for hypospadias to occur.
It is the change in DHT (which the AGD measurement reects) which
is the causal event. This has been demonstrated in knock out studies in
mice where AGD effects were disassociated from hypospadias (Suzuki
et al., 2014). As such AGD is considered a useful biomarker but not a
KE in the progression of this AOP.
Appropriate DHT signaling is necessary for proper penile develop-
ment but not sufcient as additional factors need to occur to ensure the
proper closure of the urethral tube, which, if disrupted, lead to
hypospadias, suggesting alternative androgen independent AOPs for
hypospadias exist as well. Evidence of the lack of sufciency is pro-
vided by data from Mafb knockout (KO) mice as these animals demon-
strate a failure in urethral tube formation and closure despite having
similar levels of DHT in the GTs as their wild type counter parts
(Suzuki et al., 2014). Mafb KO also show no discernable difference
from wild type animals in AGD measures, which further supports this
measure as a reliable marker of DHT levels and not a causal factor in
the AOP (Suzuki et al., 2014). One confounding factor reported by
Foster and Harris (2005) showed a single exposure to utamide on
GD 16 or GD 19 impacted AGD, but did not lead to hypospadias. These
gestational days coincides with the period before proximal fusion of
the urethra begins, and after urethra fusion is complete which is con-
sistent with the proposed AO but suggests AGD may not be a perfect
indicator as it is inuenced outside of the window of effect. We suggest
the addition of KE7, inhibition of urethral tube closure, and propose
Mafb gene expression as an indicator of disruption, to further capture
the AO progression. Unlike some of the other genes which are critical
for normal GT growth, deletion of Mafb does not affect development of
the GT prior to the critical androgen dependent programming window
(Suzuki et al., 2014). Additional work suggest that Fgf10 and FgfR2,
also appear to be critical for proper urethral tube closure, though these
genes appear to be related to both growth and fusion events (Petiot
et al., 2005, Leung et al., 2016). Temporal expression patterns of Mafb
are consistent with developmental timing of urethral tube closure as
gene expression is evident by GD16.5, the start of proximal fusion
(Suzuki et al., 2014) and the KO animals provide direct evidence that
MAFB has a key role in urethral tube closure. There is also strong bio-
logical plausibility for a KER between DHT and Mafb expression. Data
from Androgen receptor (Ar) KO mice, provide evidence of a causal
relationship between DHT and Mafb expression, as Ar KO male GTs
have down regulated Mafb signaling as compared to wild type animals
(Suzuki et al., 2014). Addition of a KE specic for urethral tube closure
may also facilitate network expansion as it differentiates the proposed
AOP based on androgen dependent mesenchymal differentiation to
form the urethra and what would be a separate AOP for androgen
dependent organ size control, which would branch after DHT.
(Suzuki et al., 2014). As no experimental evidence provided dose
and temporal concordance data for Mafb expression, further work
would be needed to explore this potential KE for AOP2 and the utility
of Mafb as a marker.
The level of condence in the KERs of AOP2 based on considera-
tion of biological plausibility is summarized in Table 3 and range from
low to high. In general, the developmental biology supports well
the latter half of the AOP, with reduced insight provided into the
early/upstream KEs. The biological and mechanistic understanding
of the steroidogenesis pathway is well documented and the importance
of DHT in urethral tube development broadly demonstrated and
accepted. However, the ability of the transcriptional repressor
COUPTFII to suppress steroidogenesis genes has not been widely
established. Additionally, the mechanistic underpinnings of DHTs
effects on the urethral tube and necessity of another key molecular
event in the urethral tube downstream of DHT is not understood. Con-
sidering the pathway as a whole, the biological plausibility is moder-
ate based on the wellestablished link between androgen and
urethral tube development; but limited understanding of upstream reg-
ulation of steroidogenesis genes.
The extent of empirical support for the KERS in AOP2 is summa-
rized in Table 3. Though a dose responsive effect following exposure
to LMWPs is supported for most KE, empirical evidence to support con-
cordance of KERs ranged from low to moderate. A study by van den
Driesche et al (2012) demonstrates well dose and incidence concor-
dance of the early KEs (Table 4), however measurement of these
events on the same gestational day (e21.5) prevents assessment of tem-
poral concordance and did not measure KE5 and KE7. A study by Gray
et al. (2016), demonstrates dose, incidence and temporal concordance
between KE2, KE5 and AO2 (Supplemental Table 5), but does not mea-
C.M. Palermo et al. Current Research in Toxicology 2 (2021) 254271
260
sure KE1 or KE7. Compiling observations from multiple studies on DBP
with various experimental designs provides moderate support for con-
cordance between KERs (Supplemental Table 6), with KE7 continuing
to be a key observational gap in the LMWP literature. KERs are sup-
ported by indirect observational evidence showing 1) overlapping
DNA promotor binding sites for SF1, a transcription factor important
for steroidogenic gene transcription, and COUPTFII on StAR,
Cyp11A1 and Cyp17, and 2) colocalization of AR expression and Mafb
in cell population of GT critical for urethral tube closure (Supplemen-
tal Table 7). Considering the pathway as a whole, the extent of empir-
ical evidence in support of the sequence of events laid out in AOP2 is
concluded as moderate upon consideration of empirical concordance
and indirect observational evidence.
The KEs in this AOP are linked to developmental events which are
thought to occur on specic gestation days, and the proposed AOP is
disruption of the occurrence of events, not regression of said event.
However some of the data suggest this may be incorrect as the AO
occurs in animals treated with a stressor after completion of the devel-
opmental event. Specically the developmental literature evaluated
suggests urethral tube closure is complete by E17.5, however in one
study (Foster and Harris 2005) hypospadias are observed after a single
treatment on GD18 and in a second study (van den Driesche et al.,
2017) hypospadias occur after treatment begins on GD19.5 (Fig. 2).
These data could suggest an alternative to KE5 or could be explained
by discrepancies in the data. In van den Driesche et al., (2017) only
a single animal was observed with mild glandular hypospadias arising
Table 3
Assessment of biological plausibility and empirical evidence in support of the KERs in AOP2.
KER
(Adjacency of KEs)
WoE Conclusion
WoE Rationale
Biological Plausibility WoE
KE 1 leads to KE 2
(Direct)
Moderate
Established mechanistic basis that SF-1 regulates StAR, Cyp11a1, and Cyp17 genes. Coup-TFII is a known transcriptional repressor and
steroidogenesis genes have SF-1 and COUP-TFII overlapping binding sites in the promoter. Limited biological and mechanistic evidence that
COUP-TFII displaces SF-1 from binding to target genes leading to down regulation of genes critical for steroidogenesis (van den Driesche et al.,
2012). Some evidence based on dose response studies with COUP-TF1 which shares a conserved binding domain with COUP-TFII (Shibata et al.,
2003).
KE2 leads to KE5
(Direct)
High
Established biological knowledge of the steroidogenic process and that enzymatic disruption impacts male hormones (DHT) levels (Hasegawa
et al., 2000;Hu et. al., 2002;Barsoum and Yao 2006;Miller and Walter 2007;Turcu and Auchus 2015).
KE5 leads to AO1(Indirect)
High
Established biological knowledge across a range of chemicals (utamide, nasteride) and clinical outcomes that disruption of DHT (reduced
production or interference with receptor binding) causes hypospadias (Kim et al., 2002; Turcu and Auchus, 2015). Established mechanistic basis
that the importance of testosterone in external genital development is exclusively as a precursor for DHT; testosterone does not play a role. (Hib
and Ponzio, 1995; Tian and Russell, 1997).
KE5 leads to KE7
(Direct)
Moderate
Strong but limited biological evidence that MAFB (serving as an indicator of urethral tube closure) is androgen dependent (sexually dimorphic
expression, co-localization with AR receptor expression, and AR dependent expression) (Suzuki et al., 2014).
KE7 leads to AO2(Direct)
High
Established biological knowledge based on Mafb-/- mice with hypospadias phenotype (Suzuki et al., 2014). Limited mechanistic basis:
transcription factor MAFB expressed in the mesenchymal cell population of the genital tubercle adjacent to the urethral plate (cell population
critical for urethral tube formation and closure) (Nishida et al., 2008; Miyagawa et al., 2009; Suzuki et al., 2014).
KE2 leads to AO2
(Indirect)
Moderate
Male Cyp11a1 KO mice have external female genitalia which demonstrates a direct link with the development of male external genitalia and a
male like urethra (Hu et al., 2002).
Empirical Evidence WoE
KE1 leads to KE2*
(Direct)
Moderate
Dose and incidence concordance well supported in a single study on DBP (Table 4). Dose, incidence concordance challenged when combining data
across multiple studies for DBP (Supplemental Table 6). Pattern of observations within the broader dataset supports a relationship between KEs
(Supplemental Table 7).
KE2* leads to KE5**
(Direct)
Moderate
Dose and incidence concordance well supported in a single study on DPP (Supplemental Table 5). Dose and incidence concordance supported
when combining data across multiple studies for DBP (Supplemental Table 6). A dependent change in both events is observed in the same study
for a number of LMWPs (Supplemental Table 1)
KE5** leads to AO2
(Indirect)
Moderate
Dose and incidence concordance well supported in a single study on DPP (Supplemental Table 5). Data lacking to support concordance when
combining data across multiple studies for DBP (Supplemental Table 6). Association between severity of DHT impact (as measured by severity of
AGD reduction) and hypospadias incidence has been demonstrated (Supplemental Table 7). A dependent change in both events is observed in the
same study for a number of LMWPs (Supplemental Table 1).
KE5
**
lead to KE7
(Direct)
Low
Data on markers of urethral tube disruption (KE7) were not found. Pattern of observations within the broader dataset supports the KER
(Supplemental Table 7)
KE7 leads to AO2
(Direct)
Low
Data on markers of urethral tube disruption (KE7) were not found in. Pattern of observations within the broader dataset supports the KER
(Supplemental Table 7).
*The WoE assessment focused on data measuring steroidogenesis gene transcription/protein expression. **The WoE assessment relied on AGD as a surrogate
measure for DHT during the MPW. Testosterone was not used as a surrogate for KE2 or KE5 in this WoE assessment. KE1 = sustained Coup-TFII expression;
KE2 = decreased steroidogenic biosynthetic protein expression (StAR, CYP11A1, CYP17); KE5 = decreased DHT; KE7 = inhibition of urethral tube closure;
AO2 = hypospadias; KER = key event relationships; WoE = weight of evidence.
Table 4
Dose-response and temporal concordance of KE in AOP2. Observations as
reported in van den Driesche et al. (2012) on DBP support concordance of early
KEs with the AO. A dash (-) indicates no effect; a plus (+) indicates effect
observed; number of pluses indicates increased severity/incidence; a blank cell
indicates event not measured. Severity/incidence of effect weighting are
provided as part of Supplemental Table 5. An AOP supported by empirical data
shows early events occurring at lower doses and with higher severity than later
events.
Temporal Concordance
Dose
ecnadrocnoC
Dose
mg/kg/d
KE 1
(E21.5)
KE 2
(E21.5)
KE 5
KE 7
AO
(adults)
0
-
20
+
100
++
500
+++
++
+
C.M. Palermo et al. Current Research in Toxicology 2 (2021) 254271
261
from late window exposure (750 mg/kg DBP), which could be due to a
nonchemically mediated background incidence effect, or from a fetus
with a delayed gestational stage (resulting in exposure during the win-
dow of sensitivity). A mild glandular hypospadias would be consistent
with disruption at the very end of urethral tube development. The inci-
dence level of hypospadias in Foster and Harris (2005) of 20% sug-
gests a true chemically mediated event and not background
incidence, however the discrepancy could be due to how the AOP
was constructed, rather than an error in the data. The developmental
landmarks described in Fig. 2 were identied from mouse literature
(Petiot et al., 2005; Georgas et al., 2015; Ipulan et al., 2016) and des-
ignated by E whereas the chemical challenge data were developed in
rats and designated by GD in the studies (Foster and Harris 2005;
Welsh et al., 2008). The Foster and Harris (2005) data could indicate
there is not a direct alignment between mouse and rat development
and closure of the urethral tube in rat occurs later, sometime between
GD18 and GD19.
A difference in developmental timing, and thus window of suscep-
tibility, is supported by data in mice that show when mice are treated
with utamide at E17.5 and E18.5 hypospadias do not occur (Zheng
et al., 2015), which is consistent with the developmental information
in mice indicating urethral closure is complete by E17.5. The MPW,
which is dened as the window of susceptibility for this AOP, has been
less well dened in mice. Some references indicate sexually dimorphic
differences beginning at GD15 (Seifert et al., 2008) and other indicate
the MPW is set at GD 15.5, the same as rats (Cohn et al., 2011;
Miyagawa et al., 2009). Some of the primarily literature in mice
demonstrates that utamide treatment between GD15.516.5 and
GD16.517.5 causes demasculinization of the external genitalia,
whereas treatment between GD13.514.5 and GD14.515.5 has no
effect on male genital development (Miyagawa et al., 2009). Consis-
tent with the male data, females in this study treated with testosterone
propionate (TP) during the window of sensitivity (E15.5E16.5)
develop masculinized genitalia (Miyagawa et al., 2009). These data
are consistent with a MPW in mice similar to that observed in rats.
However the data generated by Zheng et al. (2015) suggests the
MPW in mice may be different. Treatment of male mice with utamide
did not have any impacts when it occurred at E17.5E18.5 whereas
treatment on E12.5E13.5, E13.5E14.5, E14.5E15.5, or
E15.5E16.5 resulted in varying degrees of hypospadias (Zheng
et al., 2015). Though treatment from E15.5E16.5 is consistent, the
earlier treatments lie outside the window of susceptibility. The data
from Zheng et al. (2015) suggest the AOP as dened here should be
taxonomically limited to rats until the discrepancies in the mouse data
are resolved. In addition species differences in phthalate literature (re-
viewed in Johnson et al., 2012) suggest the MIE may be taxonomically
restrained.
Considering the pathway as a whole, the extent of evidence is con-
cluded as moderate in support of AOP2 upon consideration of moder-
ate biological plausibility, moderate empirical evidence, and
uncertainties. Condence is high in support of the life stage relevance
of this AOP during fetal life and more specically the gestational tim-
ing of the KE upstream of the AO to the MPW of fetal development
(Foster and Harris, 2005; Welsh et al., 2008; van den Driesche et al.,
2017). The reduced condence in the early KERs should be considered
in applications of this AOP. Clarity on the MIE and increased con-
dence in early KERs, would improve utility in regulatory applications
and inform the advancement of alternative approaches capable of
screening or identifying developmental stressors.
3.1.3. AOP3: sustained COUP-TFII to epididymal agenesis and altered
sperm maturation
The Wolfan ducts (WD) are the progenitors of the epididymis, vas
deferens and seminal vesicles. There are three developmental pro-
cesses that are considered to be important during the development
of the WD: (1) mesonephros formation, (2) stabilization (~GD
15.518.5) and differentiation (~GD19birth) of the WD, and (3) post-
natal differentiation (Murashima et al., 2015). The WD develops in
both the male and female embryo as part of the rst process. However,
during the second process, regression of the duct occurs in females
whereas androgen mediated stabilization (E 1518) and retention of
the ducts occurs in males (Murashima et al., 2015;Shaw and
Renfree, 2014;Welsh et al., 2007). Following stabilization in males,
tubular elongation, and morphological differentiation occur to form
the epididymis, vas deferens, and seminal vesicle. Ductal elongation
and coiling of the epididymis continue from ~E18 until after birth,
with regions of the epididymis becoming morphologically distinct
after birth (Murashima et al., 2015).
Epididymal malformations are one of the most prevalent AOs
observed following exposure to LMWPs in the MPW (Mylchreest
et al., 1998; Barlow and Foster, 2003; Foster, 2006). The malforma-
tions have been most commonly assessed in adult animals that were
exposed as fetuses in utero and are broadly characterized as epididy-
mal agenesisdescribing reduced number of ducts, malformed, absent,
or partial epididymides; in the fetus (~GD21.5) the outcome is
described as decreased coiling (Mylchreest and Foster, 2000;
Mylchreest et al., 2002; Barlow and Foster, 2003). Targeted studies
with LMWPs and the AR antagonist, utamide, support exposure from
GD 15.517.5 as necessary and sufcient to induce epididymal agene-
sis (Foster and Harris, 2005; Welsh et al., 2007; Wilson et al., 2007).
This timing overlaps with the stabilization phase (~GD 15.518.5) of
Wolfan duct (WD) development (the progenitor of the epididymis),
a process known to be facilitated by androgens (Hannema and
Hughes, 2007; Welsh et al., 2007; Swain, 2017). Combining what is
known about the mechanisms of WD stabilization and observational
data on LMWPs (Supplemental Table 1), AOP3 can be described by
the following events: 1) sustained expression of COUPTFII, 2)
decreased steroidogenic biosynthetic protein expression (StAR,
CYP11A1, CYP17), 3) reduced testosterone levels at the site of action
in the WD, 4) altered WD development/epididymal agenesis, 5)
altered sperm maturation (Fig. 3).
Unlike AOP2 where DHT acts as the active local androgen, testos-
terone is thought to act directly on the WD to mediate stabilization
(Shaw and Renfree, 2014; Murashima et al., 2015). Due to the gesta-
tional timing of early KEs in the MPW, we consider the target organ
of testosterone action to be the WD (and not the epididymis). As exper-
imental observations tend to be measured late in gestation or postna-
tally (i.e. after WD differentiation), the AO is characterized in the
epididymis. Some studies have indicated that the epididymal changes
induced during gestation increase in incidence/severity (Fig. 3) and/or
may not manifest as clear malformations until adulthood (Barlow and
Foster, 2003). Therefore, while the life stage applicability of the AOP
is fetal life, the optimal timing to measure the AO may be in later life
stages.
The level of condence in the KERs of AOP3 based on biological
plausibility is summarized in Table 5 and range from low to high.
While the basic understanding of WD development is sufcient to for-
mulate a pathway, the fundamental understanding of the mechanisms
regulating WD development is limited. A role of testosterone in the sta-
bilization phase of WD development has been established, however
the upstream regulating events and role for other androgens are
knowledge gaps. Considering the pathway as a whole, the biological
plausibility for AOP3 is concluded to be moderate.
The extent of empirical support for the KERs in AOP3 is summa-
rized in Table 5 and range from low to high. Though a dose responsive
effect following exposure to a small number of LMWPs is supported for
KE1, KE2, KE4, AO3 (KE increased in incidence/severity with increas-
ing dose of LMWP, See Supplemental Table 1), empirical evidence to
support concordance of KERs was limited for most KERs. In particular,
all of the KEs in AOP3 were never measured together in a single study.
For example, data from Gray et al. (2016) supports well dose, temporal
and incidence concordance amongst KE2, KE3 and AO3 following
C.M. Palermo et al. Current Research in Toxicology 2 (2021) 254271
262
exposure to DPP (Table 6), however does not measure KE1 or AO4.
Evaluating temporal concordance for early KEs was also hindered
based on when data were collected. For example, van den Driesche
et al (2012) (Supplemental Table 10) supports dose and incidence con-
cordance amongst KE1, KE2 and KE3, however all measurements were
collected on the same day, and downstream KEs (AO3 and AO4) were
not assessed. As all KEs were not measured in a single study, concor-
dance of KEs was assessed by compiling observations from multiple
studies on DBP. Variation in experimental design and measurement
time points challenged this analysis, which ultimately does not support
well a dose and temporal concordance between KERs (Supplemental
Table 9). Condence in the KER between testosterone and epididymal
agenesis is increased upon consideration of indirect observational evi-
dence (Supplemental Table 11) showing 1) SD rats exposed to LMWPs
have a higher incidence of epididymal agenesis and a greater decrease
in testosterone compared to Wistar rats comparably exposed which
show a lesser effect on testosterone levels and lower incidence of epi-
didymal agenesis; 2) the epididymal lesion described following LMWP
exposure during the MPW is the same characterization of the lesion
following exposure in the same gestational window to the AR antago-
nist utamide. The limited number of studies evaluating sperm matu-
ration parameters (versus spermatogenesis) was a notable data gap
within the collective database. In totality, the limited availability of
evidence to evaluate concordance of KERs supports a conclusion of
moderate empirical evidence for the sequence of events laid out in
AOP3.
Considering the major decision points and uncertainties in this
pathway: the initial events in AOP3 are shared with AOP2 with an
unknown MIE occurring between GD 15.5 and GD 18.5, leading to sus-
tained expression of CouptfII in fetal Leydig cells. Increased COUP
TFII levels may promote increased steroidogenesis from fetal Leydig
cells and/or play a direct role in WD regression independent from or
enhanced by reduced androgen synthesis. CouptfII knockout models
show COUPTFII to be a suppressor of the mesenchymeepithelium
crosstalk responsible for WD regression (Zhao et al., 2017) which sug-
gests sustained COUPTFII alone could destabilize the WD leading to
abnormal development. As the temporal window of WD stabilization
overlaps with the MPW, it has been proposed that androgens may
antagonize COUPTFII in WD mesenchymal cells to support stabiliza-
tion potentially through androgen effects on growth factor signaling
(Hannema and Hughes, 2007; Swain, 2017; Zhao et al., 2017). As
mentioned earlier, COUPTFII is an orphan nuclear receptor and
LMWPs have been shown to activate some members of the nuclear
receptor superfamily, (i.e. PPARs) (Bility et al., 2004). Therefore plau-
sible alternatives to the AOP shown in Fig. 3 include one where sus-
tained COUPTFII in the WD is necessary and sufcient to cause
epididymal agenesis or where sustained COUPTFII in the WD is fur-
ther enhanced by the reduction of testosterone in the testes. Difculty
separating and experimentally manipulating androgen signaling and
COUPTFII in target tissues challenges the ability to untangle these
interdependent processes. Whether or not initial KEs occur in the
WD, LMWP data suggest androgen plays a role in either directly
mediating or modulating the events of this pathway (Supplemental
Table 12). Currently the biological plausibility of testicular testos-
terone impacting WD stabilization rather than COUPTFII alone is
better supported when considering the developmental biology
literature and chemical stressor data together.
The specicity of gestational timing for the KE/KERs in AOP3 is
another point of uncertainty in this analysis. As described above,
experimental evidence supports the gestational timing of the early
KEs to be during the MPW (Foster and Harris, 2005; Welsh et al.,
2007; Wilson et al., 2007). However, the development and differenti-
ation of the WD depends on continuous androgen stimulation that
extends beyond the MPW and into postnatal life (Shaw and Renfree,
2014). We consider this long duration of androgen dependence nota-
ble to the proposed AOP for two reasons. First, many of the KERs
dened in AOP3 likely have relevance beyond the restricted gesta-
tional timing proposed here. As shown in Fig. 3, a single 50 mg/kg
exposure to the AR antagonist utamide on GD19 (outside the
MPW) adversely affects epididymis size. Likewise, the LMWP litera-
ture supports that the AOs of AOP3 can arise when exposure occurs
both inside and outside the MPW (Kay et al., 2014; Czubacka et al.,
2021). Therefore, KERs downstream from the MIE are plausibily rele-
vant later in development and in other life stages. Second, reduced
Stabilization of the Wolffian Ducts Clear differentiation of epididymis
Unknown
(MIE)
Altered
sperm
maturation
(AO4)
E15.5 E16.5 E17.5 E18.5E13.5 E19.5 E20.5E13
Leydig Cell
differentiation
begins
Sustained
expression
of
COUPTFII
(KE1)
Sustained ↓
Testosterone
(KE4)
Epididymal
agenesis
(AO3)
50 mg/kg
Flutamide
0% ↓Epididymis size
20% Epididymis missing
E14.5
Stabilization of
Wolffian Ducts
↓CYP11A1
(KE2b)
↓StAR
(KE2a)
CYP17
(KE2c)
20% ↓Epididymis size
10% Epididymis missing
20% ↓Epididymis size
0% Epididymis missing
20% ↓Epididymis size
0% Epididymis missing
100 mg/kg Flutamide ↓ WD Differentiation (less coiling) %Missing s egments E21.5 0%,
PND17 0%, PND42 40%, PND70 50%
No effect on coiling, no missing segments
E21.5, PND17, PND42, PND70
%Missing Segments of WD/Epididymis E21.5 11%, PND17 13%, PND42 63%, PND70, 83%
Testosterone secretion begins
Fig. 3. Sustained COUP-TFII to epididymal agenesis and altered sperm maturation (AOP3). Leydig cells are the cellular target for KE1, the epididymis is the target
organ leading to altered sperm maturation as the AOs. KEs outlined in blue are proposed to occur during the MPW. KEs are aligned to the developmental timeline
such that events sharing the same vertical plane occur in the same developmental window. The horizontal planes depict: 1) biological process occurring, 2)
embryonic day (E), 3) AOP, and 4) subset of data that supports the pathway being specic to the indicated developmental window. These data demonstrate a
single dose of the AR antagonist utamide causes different effects on the epididymis (when measured at PND95-105) depending on when in gestation the exposure
occurs (red circles indicates timing of single exposure) (Foster and Harris, 2005); extending the duration of utamide exposure to include the late gestational
window (red lled bars) leads to an earlier onset of observable epididymal lesions (E21.5) and higher incidence later in life (PND17, PND42 & PND70) than
utamide exposure during the MPW only; and treatment during the late gestational window (open bar) has no effects on the epididymis (E21.5, PND17, PND42, &
PND70). Note: KEs are numbered according to the network (Fig. 4). (For interpretation of the references to colour in this gure legend, the reader is referred to the
web version of this article.)
C.M. Palermo et al. Current Research in Toxicology 2 (2021) 254271
263
testosterone may need to be sustained for days (or longer) to impact
the epididymis. As shown in Fig. 3, exposure to the AR antagonist u-
tamide from GD15 until birth results in an earlier onset of observable
epididymal impacts and higher incidence in adulthood than exposures
restricted to the MPW only. It has been shown that in utero exposure to
DBP at high doses during the MPW only produces a sustained reduc-
tion in testosterone levels that persists into adulthood (van den
Driesche et al., 2017). Therefore, it is possible a sustained reduction
in testosterone that extends over a period of days or longer is necessary
to elicit AO3. The renement in the KE description of AO3 may also
inform the life stage restriction of this AOP. As shown in Fig. 3, the nat-
ure of the epididymal effects may vary depending on timing of expo-
sure, likely due to the differentiation status of the WD at the time of
the reduced testosterone perturbation. As mentioned earlier, studies
tend to collect multiple KE measurements simultaneously, capturing
a snapshot in time typically after the MPW. This challenges the con-
dence in the gestational timing of KEs/KERs specied here. More
detail on the nature, function and cellular state of the MIE would also
inform life stage specicity. At this time however, available data sup-
port that later KERs in this AOP are plausibly relevant to other life
stages; and the duration of reduced testosterone sufcient for mani-
festing the AO is unclear with initiation of reduced testosterone levels
plausibly restricted to the MPW (Supplemental Table 12).
Table 5
Assessment of biological plausibility and empirical evidence in support of the KERs in AOP3.
KER
(Adjacency of KEs)
WoE Conclusion WoE Rationale
Biological Plausibility
WoE
KE1 leads to KE 2
(Direct)
Moderate See Table 3
KE2 leads to KE4
(Direct)
High
Established biological knowledge of the steroidogenic process and that disruption of steroidogenic enzymes impacts male hormone (testosterone)
levels (Barsoum and Yao, 2006; Hasegawa et al., 2000; Hu et al., 2002; Miller and Walter, 2007; Turcu and Auchus, 2015).
KE1 leads to AO3
(Indirect)
Low
Temporal and tissue-specic COUP-TFII ablation claried its role in developmental processes (Qin et al., 2008; Lin et al., 2011). Data assessing the
essentiality of COUP-TFII in WD, while limited, implicates COUP-TFII as an active promoter of WD elimination (Zhao et al., 2017). This evidence does
not support this KER, however, as it implicates COUP-TFII in a direct role in the WD itself rather than in the testes.
KE2 leads to AO3
(Indirect)
Moderate
While plausible that impacting the key rate limiting process in testosterone synthesis (i.e. StAR delivery of cholesterol, to the inner mitochondrial
membrane) will have dramatic impact on testosterone levels and subsequent impacts on epididymal development, data to support this are not
available. Male mice lacking StAR have normal epididymis and some capacity for steroidogenesis (Caron et al., 1997). Male Cyp11a1 null mice have a
smaller epididymis than their wild type counter parts. Histological sections of the Cyp11a1 null epididymis revealed reduced tubule size, smaller
columnar epithelium, and absence of microvilli (Hu et al., 2002). Male Cyp17a1 KO mice lack Wolfan derivatives and are infertile (Aherrahrou et al.,
2020). Data from these KO animals support that a cumulative effect on expression of these genes would plausibly impact epididymal formation.
KE4 leads to AO3
(Direct)
High
Androgen broadly proposed in the developmental biology literature as a primary factor in WD stabilization with a direct role for testicular testosterone
accepted (Higgins et al., 1989; Hannema and Hughes, 2007; Shaw and Renfree, 2014; Murashima et al., 2015; Swain, 2017). Targeted studies with
androgen receptor antagonist, utamide, during the period of WD stabilization (GD 15.517.5) leads to epididymal agenesis (Foster and Harris, 2005;
Welsh et al., 2007).
AO3 leads to AO4
(Direct)
High
Well accepted role of the epididymis in supporting sperm maturation. Fertility is dependent on sperm maturation in the epididymis (Joseph et al.,
2009).
Empirical Evidence WoE
KE1 leads to KE2*
(Direct)
Moderate See Table 3
KE2* leads to KE4
(Direct)
Moderate
Dose and incidence concordance well supported in a single study on DPP (Table 6) and when data were combined across multiple studies for DBP
(Supplemental Table 9). Change in both events is observed in the same study for a number of LMWPs (Supplemental Table 1).
KE4 leads to AO3
(Indirect)
High
Dose, temporal and incidence concordance supported in a single study on DPP (Table 6). Dose and incidence concordance supported for DBP when
data were combined across multiple studies (Supplemental Table 9). Pattern of observations within the broader dataset supports a relationship
between KEs (Supplemental Table 11). Change in both events is observed in the same study for a number of LMWPs (Supplemental Table 1)
AO3 leads to AO4
(Direct)
Low
Data on sperm maturation parameters limited. Change in both events is observed in the same study for DBP (Supplemental Table 1).
*This assessment focused on data measuring steroidogenesis gene transcription/protein expression. Testosterone was not used as a surrogate for KE2 or KE5 in this
WoE assessment. KE1 = sustained expression of COUP-TFII; KE2 = decreased steroidogenic biosynthetic protein expression (StAR, Cyp11a, Cyp17); KE4 = re-
duced testicular testosterone levels; AO3 = altered Wolfan duct (WD) development/epididymal agenesis; AO4 altered sperm maturation.
BP = biological plausibility; KER = key event relationships; DS KE = downstream key event; US KE = upstream key event; ND = not determined; WoE = weight
of evidence.
Table 6
Dose, temporal and incidence concordance of KEs in AOP. Observations as
reported in Gray et al. (2016) on DPP, supports concordance of a subset of KEs.
A dash () indicates no effect; a plus (+) indicates effect observed; number of
pluses indicates increased severity/incidence; a blank cell indicates event not
measured. Note for KE2, the severity of effect at a given dose varied among the
steroidogenesis genes. The severity indices here reects an average impact
between StAR, CYP17 and CYP11. Severity/incidence of effect weighting are
provided as part of Supplemental Table 9. While sperm parameters were
reported, parameters of impact on sperm maturation were not reported. An AOP
supported by empirical concordance shows early events occurring at lower doses
and with higher severity than later events. GD = gestational day;
mo. = months.
Temporal Concordance
Dose
ecnadrocnoC
Dose
mg/kg/d
KE 1
KE2
(GD18)
KE 4
(GD 18)
AO3
(6-7 mo.)
AO4
(6-7 mo.)
0
-
-
-
11
+
+
-
33
+
++
-
100
++
+++
+
300
+++
+++
+++
C.M. Palermo et al. Current Research in Toxicology 2 (2021) 254271
264
Considering the pathway as a whole, the extent of evidence is con-
cluded as moderate in support of AOP3 upon consideration of moder-
ate biological plausibility, moderate empirical evidence, and
uncertainties. Condence is high in support of the life stage relevance
of the entire AOP during fetal life based on the restricted gestational
timing of the MIE to the MPW (Foster and Harris, 2005; Welsh
et al., 2007; Wilson et al., 2007). However, the condence is low with
respect to the duration of testosterone reduction needed to manifest
the AOs. Condence in this AOP for use in regulatory application is
considered high if the uncertainties described are respected. Clarity
on the MIE may improve advancement of alternative approaches cap-
able of screening or identifying developmental stressors.
3.2. Putative AOP network for male reproductive outcomes
The three AOPs described in this work were assembled into a puta-
tive AOP network that aims to more realistically represent the biology
occurring during this narrow window of gestational development.
Identication of shared KEs amongst the AOPs considered both the
sameness of the biological or physiological state and sameness of the
gestational timing of when this same KE occurred in each pathway.
As a general description, the network portrays primarily a divergent
motif (Knapen et al., 2018) with the androgen independent pathway,
AOP1, possibly sharing an unknown MIE with the androgen dependent
pathways, AOP2 and AOP3, with all 3 pathways sharing a common
population level AO (i.e. impaired fertility). A lack of a convergence
point at testosterone clearly depicts a variable role of testosterone in
all three networked AOPs. The condence in the network is low due
to the unidentied MIEs. These AOPs are considered to be open and
under development and should be iteratively rened and elaborated
via incorporation of new/existing relevant data.
Attention to gestational timing of KEs and KERs was a key focus
throughout this effort and care was taken to align the KEs/KERs with
the gestational timing of their causal importance in the network depic-
tion. In doing so, the life stage applicability of this network is dened
in a rened manner. The evidence on LMWPs clearly supports the tim-
ing of the stressor perturbation as needing to occur during the MPW to
initiate all three AOPs. This means the function or structure of the MIE
(s) may likely be restricted to this narrow window of development.
While the unknown identity of the MIE(s) makes this impossible to
conclude, it is clear that the life stage applicability of at least one early
KE/KER in this network is restricted to the MPW. Unlike AOP1/2,
KERs downstream from the MIE in AOP3 are plausibly relevant to
other life stages. For example the KER of reduced testosterone (KE4)
leading to epididymal agenesis (AO3) is likely relevant in postnatal life
and the KER of reduced testosterone (KE4) leading to altered sperm
maturation relevant into adulthood. The life stage applicability of an
AOP is restricted by the most restrictive KER. A simple illustration of
this concept is that cryptorchidism cannot be induced in adulthood.
While some KERs in an AOP leading to cryptorchidism may be func-
tionally relevant during adulthood, an AOP describing cryptorchidism
is restricted to development.
Considering the major uncertainties in this network, the unidenti-
ed MIE is a clear data gap and limits condence in the network. A
common MIE is reasonably plausible based on indirect empirical evi-
dence that exposure during a discrete window of development (the
MPW) is necessary and sufcient for a common chemical initiator
(LMWPs) to induce all three AOs. However, it is equally plausible that
separate initiating events could be responsible. In either case it is likely
the MIE(s) occurs in fetal Leydig cells. With further delineation of
these pathways and developmental events, it could be determined that
AOP1 belongs in the network, or that AOPs 2 and 3 only converge on
KE1 and KE2 before diverging again. While the assumption of a com-
mon MIE is plausible, an undened MIE means the inclusion of AOP1
in this network is uncertain. Potential MIEs were explored as part of
this effort and some were discussed already in earlier sections.
Reduced activation of PPARαhas been proposed as responsible for
reduced steroidogenesis (AOP#18 https://aopwiki.org)(Nepelska
et al., 2017), however the condence in this MIE is questioned based
on studies describing the lack of PPARαresponsive gene expression in
the fetal testis following LMWP exposure, inability of PPARαantago-
nists to reduce fetal testis steroidogenic gene expression or testos-
terone production (Hannas et al., 2011) as well as other empirical
evidence (as discussed in (Arzuaga et al., 2019; Gray et al., 2021;
Clewell et al., 2020). Further evaluation of the molecular biology liter-
ature plausibly implicates the nuclear receptor NR4A1 as a possible
MIE based on its demonstrated ability to regulate INSL3 (Robert
et al., 2006) and testosterone biosynthesis genes (HSD3B2, rat
Cyp17a1, mouse Hsd3b1, and mouse StAR promoters) (Zhang and
Mellon, 1997; Hong et al., 2004; Martin and Tremblay, 2005). More
recently, phospholipase A2 (cPLA2) enzyme inhibition has been pro-
posed as a possible MIE responsible for reduced steroidogenesis
(Clewell et al., 2020). In this proposal, inhibition of cPLA2 leads to
inhibition of arachidonic acid (AA) release from intracellular stores
resulting in reduced steroidogenesis gene transcription. The plausibil-
ity of this is supported in the literature based on role of AA in mediat-
ing StAR gene transcription. The biological plausibility of cPLA2 or AA
leading to reduced INSL3 was not addressed in this proposal with over-
all condence in this MIE concluded as low by Clewell et al. (2020).
The uncertainty in the MIE denotes the putative nature of this pro-
posed network. The limitation this uncertainty poses to the utility of
this network will depend on the specic network application.
A lack of empirical data in this instance to support the convergence
point at the AO (impaired fertility) at the population level is also a
point of uncertainty worth a brief mention. Clearly the testes, epi-
didymis, and penis all serve a role in male fertility, supporting a bio-
logically plausible connection point. While LMWPs have been
associated with effects on fertility indices e.g. sperm, fertility index
(Kay et al., 2014; Yost et al., 2019), careful evaluation of these data
to inform the KER (i.e. initiated in the MPW, relevance of indices to
the target organs) was out of scope of this effort. The uncertainty in
the population level convergence point adds to the putative nature
of this proposed network.
Expansion of the putative network (Fig. 4) was explored via connec-
tivity to information available in the AOPWiki (https://aopwiki.org).
The AOPWiki was searched for KE titles of similar nature to those
herein. Hits were eliminated if the information in the title, KER descrip-
tion and/or life stage applicability domain clearly indicated a mismatch
(see Supplemental Table 13 for search terms and results). The remain-
ing KEs and associated KERs and AOPs were evaluated for relevance to
this network. Three KEs were identied as similar in title and descrip-
tion to those herein: KE #1613, Decrease, dihydrotestosterone levels
(KE#5, herein); KE#1616 cryptorchidism (AO1, herein); and
KE#406 impaired fertility (AO5). Taking KE#1613 (decrease, DHT
levels) rst, only 1 of the 4 associated KERs (KER # 1935: Decrease,
DHT levels leads to decrease AR activation) was concluded as relevant
as it would rene further the mechanistic events occurring between
KE5 and KE7 in AOP2. Regarding KE#1616 (cryptorchidism), the KE
is relevant as it describes the same AO captured in AOP1, however
the associated KER (#1938 impaired inguinoscrotal phase leads to mal-
formation, cryptorchidism) is irrelevant as it refers to the second phase
of testis descent which is outside of the narrowly dened gestational
window herein. This AOP however would be relevant to a network
expansion effort aimed at depicting the sequential nature of AOPs
across developmental stages leading to cryptorchidism.
4. Discussion
The objective of this review was to develop AOPs as a means to
evaluate the biological plausibility and empirical evidence supporting
(or refuting) the linkage and gestational timing of biological events
C.M. Palermo et al. Current Research in Toxicology 2 (2021) 254271
265
involved in mediating male reproductive tract abnormalities during
fetal development. As it is a large effort to describe all of the possible
perturbations through which stressors may lead to male reproductive
tract abnormalities, we focused our efforts on characterizing the path-
ways through which LMWPs perturb biology during the androgen sen-
sitive window of development. As it is also a large effort to describe
the pathways responsible for all of the male reproductive tract effects
associated with LMWP exposure, we focused on the pathways leading
to cryptorchidism, hypospadias and epididymal agenesis. The AOPs
described range in condence from moderate to high and operate
through androgen independent (AOP1, Fig. 1) and androgen
dependent (AOPs 2 and 3; Figs. 2 and 3, respectively) mechanisms.
With regard to the impact of androgens, both testosterone and DHT
play an important role. To more realistically represent the complexity
of the biology and possible stressor perturbation points during this
dynamic period of development, a putative network was established.
As testosterone was not a common KE across pathways, the putative
network does not converge in the center in a bowtie manner at testos-
terone as has been previously illustrated in the literature (Howdeshell
et al., 2017; Clewell et al., 2020). Instead we propose a putative net-
work linking AOP2 and AOP3 through decreased steroidogenic biosyn-
thetic protein expression with convergence of all AOPS at the
population level impaired fertility AO. While extensive data were
relied upon to underpin the proposed AOPs, the condence in each
of the AOPs varies and gaps remain, including the MIEs for all AOPs.
Through development of KER descriptions and WoE evaluation in a
manner consistent with the OECD AOP framework (OECD, 2018), this
work provides an ample contribution towards a holistic understanding
of what is known and not known about the possible perturbation
points and AOs relevant to this narrow window of gestational
development.
A key challenge in generating this AOP network depicting pro-
cesses that are dynamic, highly interconnected and occurring in a very
short window of developmental time was avoiding inference and
assumption due to study design and information limitations. Detailed
time course studies restricted to the 3 day window of relevance (i.e.
MPW) were limited making it difcult to assign cause and effect and
untangle early androgen dependent pathways from those initiated
later in gestation. Available studies tended to involve high dose levels
to ensure induction of reproductive abnormalities, hindering dose con-
cordance assessments. The ability to assess cause and effect was also
limited by the timing of observations. Studies often assessed endpoints
simultaneously rather than sequentially over time; and observations
were often made after the gestational time points of interest. For exam-
ple, measurement of testosterone outside of the MPW, and in particu-
lar in postnatal or adult animals, was considered of limited utility for
informing the prenatal biology occurring in the MPW, especially if the
stressor exposure was not limited to the MPW. Reliance on develop-
mental biology literature was useful for clarifying, to some extent,
where extrapolation or inference of the stressor observational data
was reasonable versus unfounded. Targeted studies using lower
potency substances targeting multiple MIEs with careful dose spacing
and observational time points may prove useful to further separate
intertwined processes. Identication, standardization and utilization
of biomarkers of in utero effect are also seemingly useful in efforts
to improve condence and clarity in the network. For example AGD
Fig. 4. Putative network to male reproductive AOs initiated in the MPW. The gure should be read from left to right and is organized in developmental time from
top to bottom. Everything sharing the same vertical plane is thought to occur in that developmental window. The horizontal planes depict: 1) linear pathway
network, 2) biological process and 3) gestational timing of events (E = embryonic day). KEs outlined in blue are those impacted during the MPW. The solid arrows
indicate KER identied as part of the AOP and the dashed arrows indicate a mechanistic link where the event was not identied as a keypart of the AOP.
Hypothesized AOPs are underscored by a colored pathway- green for AOP1, blue for AOP2 and purple for AOP3. The dashed green between the MIEs indicate they
may be unique, as depicted, or shared. For the network it was determined that enough evidence existed to assume the unknown MIE for AOPs 2 and 3 is shared as
the two AOPs share the same early KEs and it is unlikely that two distinct MIEs would converge on the same early KE. AOP1 could not be linked to the other AOPs
with any known early KEs therefore, until otherwise demonstrated, the MIE is depicted as a unique event. (For interpretation of the references to colour in this
gure legend, the reader is referred to the web version of this article.)
C.M. Palermo et al. Current Research in Toxicology 2 (2021) 254271
266
measured postpartum is a recognized biomarker for a change in DHT
levels in utero (Sharpe, 2020), and as such serves as an inlife measure-
ment that allows for additional observations at later time points in
postnatal development. Unique experimental designs are likely
needed to improve condence in the proposed KERs and proposed net-
work convergence/divergence points.
Dening the taxonomic applicability domain for these AOPs was
out of scope of this effort. However, it should be emphasized that vari-
ations in mechanisms (molecular and anatomical) and gestational tim-
ing are expected to pose varying degrees of taxonomic restriction to
these AOPs. In this effort, evidence underpinning biological plausibil-
ity (i.e. the developmental biology literature) was primarily from mice
while the empirical evidence (i.e. the stressor data) was primarily from
rats. As there are differences in androgen responses following in utero
LMWP exposure between mice and rats (Johnson et al., 2012) and reg-
ulation of testosterone in fetal development is different in rodents than
humans (Sharpe 2020), taxonomic relevance is particularly important
to clarify for regulatory utility of these AOPs. For example, it is known
that there is a fundamental difference in the regulation of testosterone
production in the human fetal testis compared to rats (Scott et al.,
2009; Sharpe, 2020). The disparity in steroidogenesis regulation
between these two species is consistent with ndings in fetal human
testes where exposure to LMWPs, both in in vitro culture and as testis
xenotransplants in immune compromised mice, results in no suppres-
sion of testosterone production (Lambrot et al., 2009; Mitchell et al.,
2010; Heger et al., 2012). This suggests a potential impact to the tax-
onomic relevance of AOP2 and AOP3 in particular, as they are andro-
gen dependent. To advance development of alternative methods and
predictive models of human relevance, understanding of the taxo-
nomic variations is important to consider and incorporate into net-
work expansion efforts.
Previous publications have proposed other designs for male repro-
ductive AOP networks based on the effects of LMWPs. These include
an interrelated, bowtienetwork arising from both androgen
dependent and androgenindependent pathways (Arzuaga et al.,
2019; Kortenkamp, 2020) with some merging clearly at the cellular
testosterone response (Howdeshell et al., 2017; Clewell et al., 2020).
The purpose for which these networks were developed varies as does
the breadth of biological pathways they have integrated. However a
notable commonality between them was the use of broadly dened
network nodes (or KEs), where several biological, cellular or target
organ states were collapsed or groupedinto a single KE. For exam-
ple, some presented INSL3 and androgen in a single networking node
(Howdeshell et al., 2015; Howdeshell et al., 2017; Arzuaga et al.,
2019), some grouped various hormones into a generalized androgen
response(Arzuaga et al., 2019; Kortenkamp, 2020), and most com-
monly the AOs were merged or grouped in a manner depicting all
reproductive malformations as equivalent or able to substitute for
another (Howdeshell et al., 2015; Howdeshell et al., 2017; Arzuaga
et al., 2019; Kortenkamp, 2020). The grouping of KEs and manner in
which KEs were described in these published networks (Howdeshell
et al., 2015; Howdeshell et al., 2017; Arzuaga et al., 2019; Clewell
et al., 2020; Kortenkamp, 2020) challenged our ability to build upon
these efforts and hindered our ability to identify and address
inconsistencies.
A KE is meant to dene a single measurable event within a specic
biological level of organization, with a KER meant to dene a causal
and predictive relationship between KEs (OECD, 2018). As discussed
in Villeneuve et al. (2018a), the extent of biological resolution to
include in an AOP is determined by the developer, and an AOP can
be reduced to a minimal few KEs if the weight of correlative evidence
is sufcient. If a KE encompasses several changes this may prove prob-
lematic to the base utility of the AOP. For example, it is easily foreseen
how grouping of KEs will make quantication of KERs more difcult
and obscure the biological understanding needed for development of
alternative testing approaches for these outcomes. In the network pre-
sented in Fig. 4, the independence of cryptorchidism from testosterone
is well supported, as well as the reliance of hypospadias on DHT rather
than testosterone. Rening the KEs to specify androgens could provide
regulators an additional level of understanding when considering com-
pounds with less wellstudied effects, For example, a stressor could
theoretically produce offspring with normal testosterone levels but
also show cryptorchidism (due to impacts on INSL3) or hypospadias
(due to inhibitory effect on the enzymatic conversion of testosterone
to DHT). Arguably, the network depiction in Fig. 4 may obscure con-
nectivity of the biology as testosterone is not represented as the mech-
anistic precursor to DHT in AOP2. As steroidogenesis is a multistep
process, the level of biological resolution in the KE descriptions and
KE selections will impact network points. Just as the grouping of
KEs in the previously published networks (Howdeshell et al., 2015;
Howdeshell et al., 2017; Arzuaga et al., 2019; Clewell et al., 2020;
Kortenkamp, 2020) challenged our ability to build upon these efforts,
our choices of KE descriptors could hinder linkage to other AOPs. More
clarity on the extent of biological resolution needed for KEs will likely
come as more AOPs relevant to male reproductive development are
established.
An important area of focus during the construction of the AOPs/
AOP network herein, was on gestational temporality and aligning
the KE/KERs with gestational timing. Publications depicting AOPs/
AOP networks associated with perturbation of androgen mediated
reproductive development have, for the most part, acknowledged the
importance of GD15.518.5 as necessary and sufcient for LMWPs
to initiate the AOPs (i.e. trigger the MIE) (Howdeshell et al., 2017;
Arzuaga et al., 2019; Clewell et al., 2020; Kortenkamp, 2020). How-
ever they are not so careful to dene the window of development of
the other KEs, or depict the separation in developmental time among
networked AOPs. For example, Howdeshell et al (2017) utilizes an
AOP network to depict the multiple mechanisms and MIEs by which
LMWPs in combination with other stressors can interfere with andro-
gen singling to disrupt male reproductive development to help inform
an approach for cumulative risk assessment. The network depiction
appears to link together the transabdominal phase of testicular descent
(inuenced by INSL3 and relevant during gestation in rats), and the
inguinoscrotal phase of testicular descent (inuenced by androgen
and relevant during postnatal development in rats) in a manner por-
traying these pathways cooccur in biological time. However these
two phases of testicular descent occur weeks apart in development.
If cumulative risk assessment is concerned with how chemicals behave
when they are present together, it may be important to distinguish
pathways in a network that do not cooccur in biological time as this
informs what is plausible with respect to the trajectory of outcomes.
In addition, the inguinoscrotal phase of testicular descent (the
androgen dependent phase) occurs in utero in humans and postnatally
in rodents (Picut et al., 2017). This species difference in life stage
applicability of the AOPs may restrict the taxonomic validity of the
network or result in different network connections based on other
species differences during this period of development. During
development, life stage considerations may be the most restrictive
boundary of how broadly data may be extrapolated, including
potential human relevance. Life stage specicity is also important for
informing when data become evidence in a chemical hazard
assessment (e.g., testosterone levels measured in adulthood provides
little value in informing on the ability of a chemical to inuence
testosterone levels during gestation), for guiding experimental design
(e.g. duration of exposure or optimal observation time point), and
may be key to reconciling inconsistencies in a chemicals ability to
elicit effects across a group of studies with a wide range of
experimental designs.
Current OECD AOP guidance (OECD, 2018) identies life stage rel-
evance as an important consideration in bounding the biological
domain of applicability of an AOP. In the AOPwiki, structured terms
can be selected to identify life stage relevance with options to clarify
C.M. Palermo et al. Current Research in Toxicology 2 (2021) 254271
267
further in a free text eld. These structured terms are fairly broad (e.g.
developmental, fetal, juvenile) compared to the renement to gesta-
tional days/weeks depicted herein. To facilitate broader use of AOPs,
incorporation of more rened life stage information beyond in utero
or fetallife stage designation (e.g. gestational timing) should be
encouraged. Incorporation of life stage information to include gesta-
tional timing in the KE details will facilitate generation of AOP net-
works that depict the toxicodynamics during a narrow development
stage. This would allow for sequential network depictions of critical
molecular and morphological progressions that more appropriately
capture the complexity of developmental biology. Unfortunately, there
is variation in the literature for the different fetal developmental stages
in a given species and standard milestones across taxa that can be read-
ily applied do not exist (Picut et al., 2017). However, this may be
important to tackle in order to facilitate more informed linkages in
the AOPwiki between shared KEs and reuse of information by AOP
wiki users.
The OECD has recognized the importance of standardized organiza-
tion of information as critical to support condence in use of AOPs and
guidance for AOP development is available from many sources
(Villeneuve et al., 2014a; Fay et al., 2017; OECD, 2017; Knapen
et al., 2018; OECD, 2018). However, the manner in which AOPs are
developed and the pathway specicity with which they are character-
ized is still evolving and this experience suggests further clarity of the
OECD AOP development standards is warranted. First, and as dis-
cussed above, variation in KE naming conventions, limited KE descrip-
tions, and level of mechanistic abstraction when selecting KEs
continues to challenge the development of sharable KEs. Deviations
from the current OECD KE naming conventions may be needed to
ensure appropriate level of biological renement to improve connec-
tivity among AOPs (Leist et al., 2017; Villeneuve et al., 2018b). As
well, guidance on how explicitly to characterize an unknown MIE
would improve utility of these knowledge gaps. For example, inclusion
of maximal information on biological context and speculative molecu-
lar targets would avoid propagation of fragmented information and
could facilitate appropriate networking particularly if a plausible
MIE is entered into the AOPWiki (https://aopwiki.org) at a future
date. Second, standardizing approaches to data collection may be a
benecial advancement of the AOP development process. In this effort,
potential AOPs were rst characterized based on developmental biol-
ogy knowledge, followed by multiple rounds of targeted investigative
literature searches (in both the biology and toxicology elds) to inform
biological plausibility and empirical evidence to support KERs. While
stepwise and meticulous, clear documentation of the key questions
that drove the literature searches and clarity of literature relied upon
(versus excluded) would have enhanced transparency in WoE under-
pinning these AOPs. Expanding the AOP framework to include stan-
dards for evidence identication has been suggested as a means to
improve condence (De Vries et al., 2021; Kleinstreuer et al., 2016;
Leist et al., 2017); however the standards would have to be exible
to accommodate the iterative nature of the AOP development process
and evergreen status of AOPs. Finally, inclusion of tforpurpose cri-
teria may alleviate the evidence evaluation burden on AOP practition-
ers. It is currently the case that sufciency of an AOP for a given
purpose is user determined, even for AOPs that have gone through
the OECD AOP review program (OECD, 2014, 2020). Indeed the OECD
endorsement of AOPs refers to condence in the scientic review pro-
cess that the AOP has undergone rather than endorsement that the
AOP is sufciently complete for use in regulatory application (OECD,
2020). This means the AOP practitioner must commit to understand-
ing the developers scope, decision points and evidence basis, to inde-
pendently conclude if the AOP is t for their intended use. In this effort
succinctly capturing the basis of WoE while achieving adequate trans-
parency was a big challenge. Reviewing the lengthy evidence con-
tained herein poses a heavy burden on any user of these AOPs.
Establishing readiness for use criteria(e.g. as proposed by Coady
et al. (2019)) is likely important if the full potential of AOPs is to be
realized.
The breadth of data and interdisciplinary knowledge exist to
advance next generation toxicity test methods and the AOP framework
represents a conceptual construct to facilitate and foster collaborative
efforts to this end. The AOP network herein provides a foundation to
advance additional efforts. Efforts to clarify how changes at the molec-
ular level may relate to the AOs of concern by mining highthroughput,
highcontent databases could inform the MIE or lead to more rapid
development of additional networked AOPs. The systems approach
to identifying an array of potential molecular targets involved in male
reproductive development by Leung et al. (2016) provides insights to
this end; as does a recent publication by Gray et al. (2021) evaluating
genomic responses in fetal testes exposed to an array of chemicals.
Identifying the MIE would rapidly advance identication of a targeted
in vitro assay for inclusion in an integrated testing approach for male
reproductive development. Future efforts could also involve advancing
quantication of KERs through computational analysis of legacy dose
response data to generate mathematical prediction models. A clear
next step towards collaborative utility of this network is entry of these
AOPs into the AOPWiki.
CRediT authorship contribution statement
Christine M. Palermo: Conceptualization, Formal analysis, Writ-
ing review & editing. Jennifer E. Foreman: Conceptualization, For-
mal analysis, Writing review & editing, Visualization. Daniele S.
Wikoff: Writing original draft, Writing review & editing. Isabel
Lea: Writing original draft, Writing review & editing.
Declaration of Competing Interest
The authors declare the following nancial interests/personal rela-
tionships which may be considered as potential competing interests:
Christine Palermo and Jennifer Foreman are employees of a producer
of high molecular weight phthalates. Daniel Wikoff and Isabel Lea are
employees of ToxStrategies, a private consulting rm that provides ser-
vices on toxicology and risk assessment issues to private and public
organizations..
Acknowledgements
The authors gratefully acknowledge the efforts and work of a sig-
nicant number of contributing authors and collaborators who have
participated in the development, application evolution and advance-
ment of the adverse outcome pathway concept including the OECD
Extended Advisory Group on Molecular Screening and Toxicoge-
nomics and contributors to the OECDSponsored AOP knowledgebase.
Special thanks to Katy Goyak at ExxonMobil Biomedical Sciences for
her insights during conceptualization and manuscript review.
This work did not receive any specic grant from funding agencies
in the public, commercial, or notforprot sectors. ExxonMobil
Biomedical Sciences provided nancial support to ToxStrategies for
preparation of this manuscript. No authors received personal fees.
Appendix A. Supplementary data
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.crtox.2021.07.002.
References
Adham, I.M., Agoulnik, A.I., 2004. Insulin-like 3 signalling in testicular descent. Int. J.
Androl. 27 (5), 257265. https://doi.org/10.1111/j.1365-2605.2004.00481.x.
C.M. Palermo et al. Current Research in Toxicology 2 (2021) 254271
268
Adham, I.M., Steding, G., Thamm, T., Büllesbach, E.E., Schwabe, C., Paprotta, I., Engel,
W., 2002. The overexpression of the insl3 in female mice causes descent of the
ovaries. Mol Endocrinol (Baltimore, Md.) 16 (2), 244252.
Aherrahrou, R., Kulle, A.E., Alenina, N., Werner, R., Vens-Cappell, S., Bader, M.,
Schunkert, H., Erdmann, J., Aherrahrou, Z., 2020. CYP17A1 decient XY mice
display susceptibility to atherosclerosis, altered lipidomic prole and atypical sex
development. Sci Rep. 10, 8792. https://doi.org/10.1038/s41598-020-65601-0.
Amann, R.P., Veeramachaneni, D.N.R., 2007. Cryptorchidism in common eutherian
mammals. Reproduction (Cambridge, England). 133, 541561.
Ankley, G.T., Bennett, R.S., Erickson, R.J., Hoff, D.J., Hornung, M.W., Johnson, R.D.,
Mount, D.R., Nichols, J.W., Russom, C.L., Schmieder, P.K., Serrrano, J.A., Tietge, J.
E., Villeneuve, D.L., 2010. Adverse outcome pathways: a conceptual framework to
support ecotoxicology research and risk assessment. Environ. Toxicol. Chem. 29 (3),
730741. https://doi.org/10.1002/etc.v29:310.1002/etc.34.
Arzuaga, X., Walker, T., Yost, E.E., Radke, E.G., Hotchkiss, A.K., 2019. Use of the
Adverse Outcome Pathway (AOP) framework to evaluate species concordance and
human relevance of Dibutyl phthalate (DBP)-induced male reproductive toxicity.
Reprod. Toxicol. 96, 445458. https://doi.org/10.1016/j.reprotox.2019.06.009.
Bailey, P.J., Dowhan, D.H., Franke, K., Burke, L.J., Downes, M., Muscat, G.E., 1997.
Transcriptional repression by COUP-TF II is dependent on the C-terminal domain
and involves the N-CoR variant, RIP13delta1. J Steroid Biochem Mol Biol. 63,
165174. https://doi.org/10.1016/s0960-0760(97)00079-4.
Barlow, N.J., Foster, P.M.D., 2003. Pathogenesis of male reproductive tract lesions from
gestation through adulthood following in utero exposure to Di(n-butyl) phthalate.
Toxicol. Pathol. 31 (4), 397410.
Barsoum, I., Yao, H.-C., 2006. The road to maleness: from testis to Wolfan duct. Trends
Endocrinol Metab. 17 (6), 223228. https://doi.org/10.1016/j.tem.2006.06.009.
Bility, M.T., Thompson, J.T., McKee, R.H., David, R.M., Butala, J.H., Vanden Heuvel, J.
P., Peters, J.M., 2004. Activation of mouse and human peroxisome proliferator-
activated receptors (PPARs) by phthalate monoesters. Toxicol Sci. 82, 170182.
https://doi.org/10.1093/toxsci/kfh253.
Bogatcheva, N.V., Truong, A., Feng, S., Engel, W., Adham, I.M., Agoulnik, A.I., 2003.
GREAT/LGR8 is the only receptor for insulin-like 3 peptide. Mol. Endocrinol.
(Baltimore Md.) 17 (12), 26392646.
Caron, K.M., Soo, S.-C., Wetsel, W.C., Stocco, D.M., Clark, B.J., Parker, K.L., 1997.
Targeted disruption of the mouse gene encoding steroidogenic acute regulatory
protein provides insights into congenital lipoid adrenal hyperplasia. PNAS 94 (21),
1154011545.
Clewell, R.A., Leonard, J.A., Nicolas, C.I., Campbell, J.L., Yoon, M., Efremenko, A.Y.,
McMullen, P.D., Andersen, M.E., Clewell, H.J., Phillips, K.A., Tan, Y.-M., 2020.
Application of a combined aggregate exposure pathway and adverse outcome
pathway (AEP-AOP) approach to inform a cumulative risk assessment: A case study
with phthalates. Toxicol In Vitro 66, 104855. https://doi.org/10.1016/
j.tiv.2020.104855.
Clark, R.L., Anderson, C.A., Prahalada, S., Robertson, R.T., Lochry, E.A., Leonard, Y.M.,
Stevens, J.L., Hoberman, A.M., 1993. Critical developmental periods for effects on
male rat genitalia induced by nasteride, a 5α-reductase inhibitor. Toxicol. Appl.
Pharmacol. 119 (1), 3440.
Coady, K., Browne, P., Embry, M., Hill, T., Leinala, E., Steeger, T., Maślankiewicz, L.,
Hutchinson, T., 2019. When are adverse outcome pathways and associated assays
t for purposefor regulatory decision-making and management of chemicals?
Integr Environ Assess Manag. 15 (4), 633647. https://doi.org/10.1002/
ieam.4153.
Cohn, M.J., 2011. Development of the external genitalia: conserved and divergent
mechanisms of appendage patterning. Dev. Dyn. 240 (5), 11081115.
Czubacka, E., Czerczak, S., Kupczewska-Dobecka, M., 2021. The overview of current
evidence on the reproductive toxicity of dibutyl phthalate. Int. J. Occup. Med.
Environ. Health 34 (1), 1537.
Dean, A., Sharpe, R.M., 2013. Clinical review: anogenital distance or digit length ratio as
measures of fetal androgen exposure: relationship to male reproductive
development and its disorders. J. Clin. Endocrinol. Metab. 98, 22302238.
https://doi.org/10.1210/jc.2012-4057.
Dean, A., Smith, L.B., Macpherson, S., Sharpe, R.M., 2012. The effect of
dihydrotestosterone exposure during or prior to the masculinization programming
window on reproductive development in male and female rats. Int. J. Androl. 35,
330339. https://doi.org/10.1111/j.1365-2605.2011.01236.x.
De Vries, R.B.M., Angrish, M., Browne, P., Brozek, J., Rooney, A.A., Wikoff, D.S.,
Whaley, P., Edwards, S.W., et al, 2021. Applying evidence-based methods to the
development and use of adverse outcome pathways. ALTEX 38, 336347. https://
doi.org/10.14573/altex.2101211.
Emmen, J.M., McLuskey, A., Grootegoed, J.A., Brinkmann, A.O., 1998. Androgen action
during male sex differentiation includes suppression of cranial suspensory ligament
development. Hum Reprod. 13 (5), 12721280. https://doi.org/10.1093/humrep/
13.5.1272.
Fay, K.A., Villeneuve, D.L., LaLone, C.A., Song, Y., Tollefsen, K.E., Ankley, G.T., 2017.
Practical approaches to adverse outcome pathway development and weight-of-
evidence evaluation as illustrated by ecotoxicological case studies. Environ. Toxicol.
Chem. 36 (6), 14291449. https://doi.org/10.1002/etc.3770.
Feng, S., Ferlin, A., Truong, A., Bathgate, R., Wade, J. D., Corbett, S., et al., 2009. INSL3/
RXFP2 signaling in testicular descent. Ann N Y Acad Sci. 1160, 197-204.10.1111/
j.1749-6632.2009.03841.x
Foster, P.M.D., 2006. Disruption of reproductive development in male rat offspring
following in utero exposure to phthalate esters. Int J Androl. 29 (1), 140147.
Foster, P.M.D., Harris, M.W., 2005. Changes in androgen-mediated reproductive
development in male rat offspring following exposure to a single oral dose of
utamide at different gestational ages. Toxicol. Sci. 85, 10241032. https://doi.
org/10.1093/toxsci/k159 %J Toxicological Sciences.
Georgas, K. M., Armstrong, J., Keast, J. R., Larkins, C. E., McHugh, K. M., Southard-
Smith, E. M., et al., 2015. An illustrated anatomical ontology of the developing
mouse lower urogenital tract. Development (Cambridge, England). 142, 1893-
1908.10.1242/dev.117903
Gorlov, I.P., Kamat, A., Bogatcheva, N.V., Jones, E., Lamb, D.J., Truong, A., Bishop, C.E.,
McElreavey, K., Agoulnik, A.I., 2002. Mutations of the GREAT gene cause
cryptorchidism. Hum. Mol. Genet. 11, 23092318.
Gray Jr, L. E., Furr, J., Tatum-Gibbs, K. R., Lambright, C., Sampson, H., Hannas, B. R.,
Wilson, V. S., Hotchkiss, A., Foster, P. M. D., 2016. Establishing the Biological
Relevanceof Dipentyl Phthalate Reductions in Fetal Rat Testosterone Production
and Plasma and Testis Testosterone Levels. Toxicological Sciences. 149, 178-
191.10.1093/toxsci/kfv224.
Gray L.E, Lambright C.S., Conley J.M., Evans N, Furr J.R., Hannas B.R., Wilson V.S.,
Sampson H., Foster P.M.D., 2021. Genomic and Hormonal Biomarkers of Phthalate-
Induced Male Rat Reproductive Developmental Toxicity Part II: A Targeted RT-
qPCR Array Approach that Denes a Unique Adverse Outcome Pathway. Toxicol Sci.
Epub ahead of print. doi: 10.1093/toxsci/kfab053.
Gray Jr., L.E., Barlow, N.J., Howdeshell, K.L., Ostby, J.S., Furr, J.R., Gray, C.L., 2009.
Transgenerational effects of Di (2-ethylhexyl) phthalate in the male CRL:CD(SD) rat:
added value of assessing multiple offspring per litter. Toxicol Sci. 110, 411425.
https://doi.org/10.1093/toxsci/kfp109.
Gray Jr., L.E., Ostby, J., Furr, J., Price, M., Veeramachaneni, D.N., Parks, L., 2000.
Perinatal exposure to the phthalates DEHP, BBP, and DINP, but not DEP, DMP, or
DOTP, alters sexual differentiation of the male rat. Toxicol Sci. 58, 350365.
https://doi.org/10.1093/toxsci/58.2.350.
Hadziselimovic, F., 2017. On the descent of the epididymo-testicular unit,
cryptorchidism, and prevention of infertility. Basic Clin Androl. 27, 21. https://
doi.org/10.1186/s12610-017-0065-8.
Hannas, B.R., Lambright, C.S., Furr, J., Howdeshell, K.L., Wilson, V.S., Gray Jr., L.E.,
2011. Dose-response assessment of fetal testosterone production and gene
expression levels in rat testes following in utero exposure to diethylhexyl
phthalate, diisobutyl phthalate, diisoheptyl phthalate, and diisononyl phthalate.
Toxicol Sci. 123, 206216. https://doi.org/10.1093/toxsci/kfr146.
Hannema, S.E., Hughes, I.A., 2007. Regulation of Wolfan duct development. Horm.
Res. 67 (3), 142151.
Hasegawa, T., Zhao, L., Caron, K.M., Majdic, G., Suzuki, T., Shizawa, S., Sasano, H.,
Parker, K.L., 2000. Developmental roles of the steroidogenic acute regulatory
protein (StAR) as revealed by StAR knockout mice. Mol. Endocrinol. (Baltimore
Md.) 14 (9), 14621471.
Heger, N.E., Hall, S.J., Sandrof, M.A., McDonnell, E.V., Hensley, J.B., McDowell, E.N.,
Martin, K.A., Gaido, K.W., Johnson, K.J., Boekelheide, K., 2012. Human fetal testis
xenografts are resistant to phthalate-induced endocrine disruption. Environ Health
Perspect. 120 (8), 11371143. https://doi.org/10.1289/ehp.1104711.
Hib, J., Ponzio, R., 1995. The abnormal development of male sex organs in the rat using
a pure antiandrogen and a 5 alpha-reductase inhibitor during gestation. Acta
Physiol. Pharmacol. Ther. Latinoam. 45, 2733.
Higgins, S.J., Young, P., Cunha, G.R., 1989. Induction of functional cytodifferentiation
in the epithelium of tissue recombinants. II. Instructive induction of Wolfan duct
epithelia by neonatal seminal vesicle mesenchyme. Development 106, 235250.
Hong, C.Y., Park, J.H., Ahn, R.S., Im, S.Y., Choi, H.-S., Soh, J., Mellon, S.H., Lee, K.,
2004. Molecular mechanism of suppression of testicular steroidogenesis by
proinammatory cytokine tumor necrosis factor alpha. Mol. Cell Biol. 24 (7),
25932604. https://doi.org/10.1128/MCB.24.7.2593-2604.2004.
Howdeshell, K.L., Hotchkiss, A.K., Gray, L.E., 2017. Cumulative effects of
antiandrogenic chemical mixtures and their relevance to human health risk
assessment. Int. J. Hyg. Environ. Health 220 (2), 179188. https://doi.org/
10.1016/j.ijheh.2016.11.007.
Howdeshell, K.L., Rider, C.V., Wilson, V.S., Furr, J.R., Lambright, C.R., Gray, L.E., 2015.
Dose addition models based on biologically relevant reductions in fetal testosterone
accurately predict postnatal reproductive tract alterations by a phthalate mixture in
rats. Toxicol. Sci. 148 (2), 488502. https://doi.org/10.1093/toxsci/kfv196.
Hu, M.-C., Hsu, N.-C., El Hadj, N.B., Pai, C.-I., Chu, H.-P., Wang, C.-K.-L., Chung, B.-C.,
2002. Steroid deciency syndromes in mice with targeted disruption of Cyp11a1.
Mol. Endocrinol. 16, 19431950. https://doi.org/10.1210/me.2002-0055 %J
Molecular Endocrinology.
Hutson, J.M., Hasthorpe, S., Heyns, C.F., 1997. Anatomical and functional aspects of
testicular descent and cryptorchidism. Endocr. Rev. 18, 259280.
Hutson, J., Li, R., Vikraman, J., Loebenstein, M., 2016. What animal models of testicular
descent and germ cell maturation tell us about the mechanism in humans. Eur. J.
Pediatric Surgery 26 (05), 390398.
Hutson, J.M., Southwell, B.R., Li, R., Lie, G., Ismail, K., Harisis, G., Chen, N., 2013. The
regulation of testicular descent and the effects of cryptorchidism. Endocr. Rev. 34,
725752. https://doi.org/10.1210/er.2012-1089.
Imaizumi, N., Kwang Lee, K., Zhang, C., Boelsterli, U.A., 2015. Mechanisms of cell death
pathway activation following drug-induced inhibition of mitochondrial complex I.
Redox Biol. 4, 279288.
Ipulan, L.A., Suzuki, K., Matsushita, S., Suzuki, H., Okazawa, M., Jacinto, S., Hirai, S.-I.,
Yamada, G., 2014. Development of the external genitalia and their sexual dimorphic
regulation in mice. Sexual Development 8 (5), 297310.
Ipulan, L.A., Raga, D., Suzuki, K., Murashima, A., Matsumaru, D., Cunha, G., Yamada, G.,
2016. Investigation of sexual dimorphisms through mouse models and hormone/
hormone-disruptor treatments. Differentiation 91 (4-5), 7889. https://doi.org/
10.1016/j.diff.2015.11.001.
C.M. Palermo et al. Current Research in Toxicology 2 (2021) 254271
269
Ivell, R., Bathgate, R.A.D., 2002. Reproductive biology of the relaxin-like factor (RLF/
INSL3). Biol. Reprod. 67, 699705.
Ivell, R., Hartung, S., 2003. The molecular basis of cryptorchidism. Mol. Hum. Reprod.
9, 175181.
Joseph, A., Yao, H., Hinton, B.T., 2009. Development and morphogenesis of the
Wolfan/epididymal duct, more twists and turns. Dev. Biol. 325 (1), 614. https://
doi.org/10.1016/j.ydbio.2008.10.012.
Johnson, K.J., Heger, N.E., Boekelheide, K., 2012. Of mice and men (and rats):
phthalate-induced fetal testis endocrine disruption is species-dependent. Toxicol
Sci. 129, 235248. https://doi.org/10.1093/toxsci/kfs206.
Kassim, N.M., Russell, D.A., Payne, A.P., 2010. Does the cranial suspensory ligament
have a role in cryptorchidism? Cells Tissues Organs. 191 (4), 307315. https://doi.
org/10.1159/000260062.
Kay, V. R., Bloom, M. S., Foster, W. G., 2014. Reproductive and developmental effects of
phthalate diesters in males. Crit. Rev. Toxicol. 44, 467498.10.3109/
10408444.2013.875983.
Kim, K.S., Liu, W., Cunha, G.R., Russell, D.W., Huang, H., Shapiro, E., Baskin, L.S., 2002.
Expression of the androgen receptor and 5 alpha-reductase type 2 in the developing
human fetal penis and urethra. Cell Tissue Res. 307, 145153. https://doi.org/
10.1007/s004410100464.
Kleinstreuer, N.C., Sullivan, K., Allen, D., Edwards, S., Mendrick, D.L., Embry, M.,
Matheson, J., Rowlands, J.C., Munn, S., Maull, E., Casey, W., 2016. Adverse
outcome pathways: From research to regulation scientic workshop report. Regul.
Toxicol. Pharmacol. 76, 3950. https://doi.org/10.1016/j.yrtph.2016.01.007.
Klonisch, T., Fowler, P.A., Hombach-Klonisch, S., 2004. Molecular and genetic
regulation of testis descent and external genitalia development. Dev. Biol. 270
(1), 118. https://doi.org/10.1016/j.ydbio.2004.02.018.
Knapen, D., Angrish, M.M., Fortin, M.C., Katsiadaki, I., Leonard, M., Margiotta-Casaluci,
L., Munn, S., O'Brien, J.M., Pollesch, N., Smith, L.C., Zhang, X., Villeneuve, D.L.,
2018. Adverse outcome pathway networks I: Development and applications.
Environ. Toxicol. Chem. 37 (6), 17231733. https://doi.org/10.1002/etc.4125.
Knapen, D., Vergauwen, L., Villeneuve, D.L., Ankley, G.T., 2015. The potential of AOP
networks for reproductive and developmental toxicity assay development. Reprod.
Toxicol. 56, 5255. https://doi.org/10.1016/j.reprotox.2015.04.003.
Kortenkamp, A., 2020. Which chemicals should be grouped together for mixture risk
assessments of male reproductive disorders? Mol. Cell Endocrinol. 499, 110581.
https://doi.org/10.1016/j.mce:2019.110581.
Lambrot, R., Muczynski, V., Lécureuil, C., Angenard, G., Cofgny, H., Pairault, C.,
Moison, D., Frydman, R., Habert, R., Rouiller-Fabre, V., 2009. Phthalates impair
germ cell development in the human fetal testis in vitro without change in
testosterone production. Environ. Health Perspect. 117 (1), 3237. https://doi.org/
10.1289/ehp.11146.
Leist, M., Ghallab, A., Graepel, R., Marchan, R., Hassan, R., Bennekou, S.H., Limonciel,
A., Vinken, M., Schildknecht, S., Waldmann, T., Danen, E., van Ravenzwaay, B.,
Kamp, H., Gardner, I., Godoy, P., Bois, F.Y., Braeuning, A., Reif, R., Oesch, F.,
Drasdo, D., Höhme, S., Schwarz, M., Hartung, T., Braunbeck, T., Beltman, J.,
Vrieling, H., Sanz, F., Forsby, A., Gadaleta, D., Fisher, C., Kelm, J., Fluri, D., Ecker,
G., Zdrazil, B., Terron, A., Jennings, P., van der Burg, B., Dooley, S., Meijer, A.H.,
Willighagen, E., Martens, M., Evelo, C., Mombelli, E., Taboureau, O., Mantovani, A.,
Hardy, B., Koch, B., Escher, S., van Thriel, C., Cadenas, C., Kroese, D., van de Water,
B., Hengstler, J.G., 2017. Adverse outcome pathways: opportunities, limitations and
open questions. Arch. Toxicol. 91 (11), 34773505. https://doi.org/10.1007/
s00204-017-2045-3.
Leung, M.C., Phuong, J., Baker, N.C., Sipes, N.S., Klinefelter, G.R., Martin, M.T.,
McLaurin, K.W., Setzer, R.W., Darney, S.P., Judson, R.S., Knudsen, T.B., 2016.
Systems toxicology of male reproductive development: proling 774 chemicals for
molecular targets and adverse outcomes. Environ. Health Perspect. 124 (7),
10501061. https://doi.org/10.1289/ehp.1510385.
Lin, F.-J., Qin, J., Tang, K., Tsai, S.Y., Tsai, M.-J., 2011. Coup d'Etat: an orphan takes
control. Endocr. Rev. 32, 404421. https://doi.org/10.1210/er.2010-0021.
Mamoulakis, C., Antypas, S., Sofras, F., Takenaka, A., Sokitis, N., 2015. Testicular
descent. Hormones (Athens) 14.
Martin, L.J., Tremblay, J.J., 2005. The human 3beta-hydroxysteroid dehydrogenase/
Delta5-Delta4 isomerase type 2 promoter is a novel target for the immediate early
orphan nuclear receptor Nur77 in steroidogenic cells. Endocrinology 146, 861869.
https://doi.org/10.1210/en.2004-0859.
McIntyre, B.S., Barlow, N.J., Foster, P.M., 2001. Androgen-mediated development in
male rat offspring exposed to utamide in utero: permanence and correlation of
early postnatal changes in anogenital distance and nipple retention with
malformations in androgen-dependent tissues. Toxicol. Sci. 62, 236249. https://
doi.org/10.1093/toxsci/62.2.236.
McKinnell, C., Sharpe, R. M., Mahood, K., Hallmark, N., Scott, H., Ivell, R., et al., 2005.
Expression of insulin-like factor 3 protein in the rat testis during fetal and postnatal
development and in relation to cryptorchidism induced by in utero exposure to di
(n-Butyl) phthalate. Endocrinology. 146, 4536-4544
Mendoza-Villarroel, R.E., Robert, N.M., Martin, L.J., Brousseau, C., Tremblay, J.J., 2014.
The nuclear receptor NR2F2 activates star expression and steroidogenesis in mouse
MA-10 and MLTC-1 Leydig cells. Biol. Reprod. 91, 26. 10.1095/
biolreprod.113.115790.
Miller, W.L., Walter, L., 2007. Steroidogenic acute regulatory protein (StAR), a novel
mitochondrial cholesterol transporter. BBA 1771 (6), 663676.
Mitchell, R.T., Saunders, P.T.K., Childs, A.J., Cassidy-Kojima, C., Anderson, R.A.,
Wallace, W.H.B., Kelnar, C.J.H., Sharpe, R.M., 2010. Xenografting of human fetal
testis tissue: a new approach to study fetal testis development and germ cell
differentiation. Hum. Reprod. 25 (10), 24052414. https://doi.org/10.1093/
humrep/deq183.
Miyagawa, S., Satoh, Y., Haraguchi, R., Suzuki, K., Iguchi, T., Taketo, M.M., et al, 2009.
Genetic interactions of the androgen and Wnt/beta-catenin pathways for the
masculinization of external genitalia. Mol. Endocrinol. 23, 871880. https://doi.
org/10.1210/me.2008-0478.
Murashima, A., Xu, B., Hinton, B.T., 2015. Understanding normal and abnormal
development of the Wolfan/epididymal duct by using transgenic mice. Asian J.
Androl. 17, 749755. https://doi.org/10.4103/1008-682X.155540.
Mylchreest, E., Cattley, R.C., Foster, P.M., 1998. Male reproductive tract malformations
in rats following gestational and lactational exposure to Di(n-butyl) phthalate: an
antiandrogenic mechanism? Toxicol Sci. 43, 4760. https://doi.org/
10.1006/toxs.1998.2436.
Mylchreest, E., Foster, P.M., 2000. DBP exerts its antiandrogenic activity by indirectly
interfering with androgen signaling pathways. Toxicol. Appl. Pharmacol. 168,
174175. https://doi.org/10.1006/taap.2000.9031.
Mylchreest, E., Sar, M., Cattley, R.C., Foster, P.M.D., 1999. Disruption of androgen-
regulated male reproductive development by di(n-butyl) phthalate during late
gestation in rats is different from utamide. Toxicol. Appl. Pharmacol. 156 (2),
8195. https://doi.org/10.1006/taap.1999.8643.
Mylchreest, E., Sar, M., Wallace, D.G., Foster, P.M.D., 2002. Fetal testosterone
insufciency and abnormal proliferation of Leydig cells and gonocytes in rats
exposed to di(n-butyl) phthalate. Reprod. Toxicol. 16 (1), 1928. https://doi.org/
10.1016/S0890-6238(01)00201-5.
Mylchreest, E., Wallace, D.G., Cattley, R.C., Foster, P.M., 2000. Dose-dependent
alterations in androgen-regulated male reproductive development in rats exposed
to Di(n-butyl) phthalate during late gestation. Toxicol Sci. 55, 143151. https://doi.
org/10.1093/toxsci/55.1.143.
Nation, T.R., Balic, A., Southwell, B.R., Newgreen, D.F., Hutson, J.M., 2009. The
hormonal control of testicular descent. Pediatr. Endocrinol. Rev. 7, 2231.
Nation, T.R., Buraundi, S., Balic, A., Farmer, P.J., Newgreen, D., Southwell, B.R.,
Hutson, J.M., 2011. The effect of utamide on expression of androgen and estrogen
receptors in the gubernaculum and surrounding structures during testicular descent.
J. Pediatr. Surg. 46 (12), 23582362. https://doi.org/10.1016/j.
jpedsurg.2011.09.026.
Nef, S., Parada, L.F., 1999. Cryptorchidism in mice mutant for Insl3. Nat. Genet. 22 (3),
295299.
Nepelska, M., Odum, J., Munn, S., 2017. Adverse Outcome Pathway: Peroxisome
Proliferator-Activated Receptor αActivation and Reproductive Toxicity
Development and Application in Assessment of Endocrine Disruptors/
Reproductive Toxicants Applied In Vitro Toxicology. 3, 234-249.http://doi.org/
10.1089/aivt.2017.0004
Nishida, H., Miyagawa, S., Matsumaru, D., Wada, Y., Satoh, Y., Ogino, Y., Fukuda, S.,
Iguchi, T., Yamada, G., 2008. Gene expression analyses on embryonic external
genitalia: identication of regulatory genes possibly involved in masculinization
processes. Congenital Anomalies. 48 (2), 6367. https://doi.org/10.1111/j.1741-
4520.2008.00180.x.
OECD, 2017. Revised Guidance Document on Developing and Assessing Adverse
Outcome Pathways. Series on Testing & Assessment. No. 184. Series on Testing &
Assessment. No. 184.
OECD, 2018. Users' Handbook supplement to the Guidance Document for developing
and assessing Adverse Outcome Pathways. doi:https://doi.org/10.1787/
5jlv1m9d1g32-en. last accessed May 18, 2021.
OECD, 2020. Draft Guidance Document for the scientic review of Adverse Outcome
Pathways. https://www.oecd.org/chemicalsafety/testing/
Draft_GD_AOP_scientic_review_27_July.pdf. last accessed May 18, 2021.
Overbeek, P. A., Gorlov, I. P., Sutherland, R. W., Houston, J. B., Harrison, W. R.,
Boettger-Tong, H. L., Bishop, C. E., Agoulnik, A. I., 2001. A transgenic insertion
causing cryptorchidism in mice. Genesis (New York, N.Y.: 2000). 30, 26-35.
Petiot, A., Perriton, C.L., Dickson, C., Cohn, M.J., 2005. Development of the mammalian
urethra is controlled by Fgfr2-IIIb. Development 132, 24412450. https://doi.org/
10.1242/dev.01778.
Picut, C.A., Ziejewski, M.K., Stanislaus, D., 2017. Comparative Aspects of pre- and
postnatal development of the male reproductive system. Birth Defects Res. 110 (3),
190227. https://doi.org/10.1002/bdr2.1133.
Plummer, S., Sharpe, R.M., Hallmark, N., Mahood, I.K., Elcombe, C., 2007. Time-
dependent and compartment-specic effects of in utero exposure to Di(n-butyl)
phthalate on gene/protein expression in the fetal rat testis as revealed by
transcription proling and laser capture microdissection. Toxicol. Sci. 97 (2),
520532.
Qin, J., Tsai, M.-J., Tsai, S.Y., Mueller, U., 2008. Essential roles of COUP-TFII in Leydig
cell differentiation and male fertility. PLoS ONE 3 (9), e3285. https://doi.org/
10.1371/journal.pone.0003285.
Qiu, Y., Tsai, S.Y., Tsai, M.-J., 1994. COUP-TF an orphan member of the steroid/thyroid
hormone receptor superfamily. Trends Endocrinol Metab. 5 (6), 234239. https://
doi.org/10.1016/1043-2760(94)P3081-H.
Robert, N.M., Martin, L.J., Tremblay, J.J., 2006. The orphan nuclear receptor NR4A1
regulates insulin-like 3 gene transcription in leydig cells1. Biol. Reprod. 74,
322330. https://doi.org/10.1095/biolreprod.105.044560 %J Biology of
Reproduction.
Rugarli, E.I., Langer, T., 2012. Mitochondrial quality control: a matter of life and death
for neurons. EMBO J. 31, 13361349. https://doi.org/10.1038/emboj.2012.38.
Sadeghian, H., Anand-Ivell, R., Balvers, M., Relan, V., Ivell, R., 2005. Constitutive
regulation of the Insl3 gene in rat Leydig cells. Mol. Cell. Endocrinol. 241 (1-2),
1020.
Scott, H.M., Mason, J.I., Sharpe, R.M., 2009. Steroidogenesis in the fetal testis and its
susceptibility to disruption by exogenous compounds. Endocr Rev. 30, 883925.
https://doi.org/10.1210/er.2009-0016.
C.M. Palermo et al. Current Research in Toxicology 2 (2021) 254271
270