Use of the Adverse Outcome Pathway (AOP) framework to evaluate species concordance and human relevance of Dibutyl Phthalate (DBP)-induced male reproductive toxicity

Abstract

Dibutyl phthalate (DBP) is a phthalate ester used as a plasticizer, and solvent. Studies using rats consistently report that DBP exposure disrupts normal development of the male reproductive system in part via inhibition of androgen synthesis. However, studies using xenograft models report that in human fetal testis DBP exposure is unlikely to impair testosterone synthesis. These results question the validity of the rat model for assessment of male reproductive effects caused by DBP. The adverse outcome pathway (AOP) framework was used to evaluate the available evidence for DBP-induced toxicity to the male reproductive system. Three relevant biological elements were identified: 1) fetal rats are more sensitive than other rodents and human fetal xenografts to DBP-induced anti-androgenic effects, 2) DBP- induced androgen-independent adverse outcomes are conserved among different mammalian models and human fetal testis xenografts, and 3) DBP-induced anti-androgenic effects are conserved in different mammalian species when exposure occurs during postnatal life stages.

The views expressed are those of the authors and do not necessarily represent the views or policies of the US EPA.

Introduction

Dibutyl phthalate (DBP) is used as a plasticizer in cellulose plastics, solvent of dyes, a softener in adhesives, lacquers, varnishes and printing inks, and it also has wide usage in cosmetics [1–3]. DBP also had limited use in polyvinyl chloride plastics during the 1970s and 1980s, but it is no longer used for this application [4]. DBP enters the environment during its use and production but can also be released from materials containing DBP as it is not covalently bound to plastic polymers [1, 2]. The largest source of DBP exposure to humans is food [2, 5, 6]. Inhalation and dermal exposure to DBP can also occur, but exposure through those routes is considered minimal [1, 2]. Within the human population, children (ages 0 to 11 years) have the highest exposure to DBP [1, 7] Phthalates are metabolized by gut lipases to monoester metabolites, which are considered the bioactive form of phthalate compounds [1, 2]. Experimental studies have shown that DBP metabolites such as monobutyl phthalate (MBP), can reach the developing fetus [1, 2, 8, 9], and phthalate metabolites have been measured in amniotic fluid in both animals and humans [10].

Previous reviews/evaluations of the available epidemiological, toxicological, and mechanistic evidence conclude that DBP exposure can result in adverse responses in the male and female¹ reproductive systems [5, 8–11]. Epidemiological studies suggest that DBP exposure is associated with altered reproductive functions in adult men and androgen disruption in the developing human male fetus [11]

Most of the available toxicological evidence on DBP-induced effects in the male reproductive system comes from experimental studies using rats, an animal model considered informative of male reproductive toxicity [12, 13]. Experimental studies have shown that exposure to DBP, and other similar phthalates during pre- and postnatal life stages can affect the male reproductive system [10, 14, 15]. Gestational exposure effects include agenesis and malformations of the male reproductive system (e.g. cryptorchidism and hypospadias), reduced androgen production/levels, alterations in markers of fetal androgen production (e.g. reduced anogenital distance [AGD], nipple retention, and decreased reproductive organ weight), delayed preputial separation, and reduced fertility [2, 8–10, 16]. Studies using adult animals have also shown that male reproductive effects such as decreased testis weight, sperm production and tubular atrophy appear to be conserved in the mouse, rat, and guinea pig, but not in hamsters [1, 2, 17]. While the responses to phthalate exposures in the male reproductive system appear similar to those elicited by known androgen receptor antagonists, phthalates do not appear to interact with the rodent or human androgen receptor [18, 19]. Evidence from gestational and/or early postnatal exposure studies suggests that young animals are more sensitive to phthalate-induced testicular injury than adults [12, 14, 20–22].

Studies using fetal testis explants or xenografts² into rat or mouse hosts were designed to evaluate the human fetal testis response to phthalates, characterize the mechanisms by which these and other types of compounds can affect the development of the male reproductive system, and identify potential species-specific differences [23, 24]. These studies report decreased responsiveness for testosterone-dependent male reproductive outcomes during the masculinization programming window³ [5, 22, 24, 25]. Although concerns in the experimental design used for human ex vivo and xenograft studies have been identified (see discussion section), overall their findings present a contrast to the associations reported in several epidemiological studies and have called into question the validity of the rat model for assessment of human male reproductive effects associated with gestational phthalate exposure [23, 25].

Evaluation of the mode(s) of action (MOA) for chemical-induced adverse effects can help identify experimental models most relevant to human health, and inform the susceptibility of the types of effects that may occur during specific life stages [9, 23, 26, 27]. A qualitative analysis of the mechanistic evidence was performed to evaluate the human relevance of effects reported in animal models. Mechanisms/pathways that are conserved across species and experimental models are considered relevant to human health [26, 28, 29]. Thus, the analysis presented here was focused on a qualitative comparison of the available evidence from in vivo, ex vivo, xenograft, and cell culture models to determine species concordance of key events⁴ and adverse responses activated with DBP exposure.

Mechanism of action for DBP-induced male reproductive effects.

The mechanisms by which DBP exposure leads to alterations in the development and function of the male reproductive system has been evaluated by authoritative organizations and various research groups⁵. Analyses of the mode(s) of action for DBP-induced male reproductive effects has also relied primarily on evidence obtained from rat experiments (E.g. [8, 12]). A brief description of the established pathways by which DBP and other similar phthalates cause adverse effects in the male reproductive system is presented below. This narrative relies on previous analyses/reviews of the mechanisms by which DBP and other phthalates cause adverse effects in the male reproductive system.

Effects resulting from exposure during gestation

Gestational development of the male reproductive system is mediated by hormones produced in Leydig cells [LC] during the masculinization programming window [30]. Testosterone regulates development of both internal and external genitalia while insulin-like growth factor 3 (INSL3) regulates the initial descent of the testis to the scrotum [13, 31]. Exposure to DBP, and other phthalates that target the male reproductive system, interferes with Leydig cell functions and results in decreased expression of steroidogenic enzymes and testosterone production, and reduced INSL3 levels (see supplementary Figure 1) [8, 12, 32]. This in turn leads to an inhibition of androgen-dependent development of the male reproductive system resulting in agenesis and malformations such as hypospadias and cryptorchidism. These findings are supportive of epidemiological studies which report an association between DBP exposure and decreased AGD [11], a morphological marker considered indicative of androgen levels during the masculinization programming window [10, 33].

Gestational phthalate exposure has also been shown to affect Sertoli cell development, functions, and interactions with germ cells. These effects may result in altered spermatogenesis in postnatal mature life stages (see supplementary Figure 1) [34, 35]. These androgen-independent effects in the seminiferous cord are not associated with changes in Leydig cell function, and testosterone and INSL3 synthesis [24]. Since Sertoli cells are required for germ cell proliferation and development, phthalate-induced decreases in Sertoli cell numbers leads to lower germ cell numbers [15]. DBP-induced effects in fetal Sertoli cells include decreased cell proliferation and differentiation, abnormal cytoskeletal arrangement, cytoplasmic alterations, and disrupted functions and contact with germ cells [22, 36, 37]. Phthalate-induced alterations in Sertoli cell function/development have been hypothesized to alter male fertility by promoting germ cell effects such as increased formation of multinucleated gonocytes (MNGs) and germ cell death [38].

Overall, phthalate-induced effects from gestational exposure results in altered or incomplete development of the male reproductive system. DBP-induced outcomes reported in experimental studies using rats include decreases in reproductive organ weights, disrupted androgen production, reduced AGD, increased incidence of malformations such as hypospadias and cryptorchidism, agenesis of the prostate, epididymis and vas deferens; retention of thoracic areolae or nipples, degeneration of the seminiferous epithelium; interstitial cell hyperplasia of the testis, reduced Sertoli and germ cell numbers, and infertility [8, 32, 37, 39].

Effects resulting from exposures during postnatal life stages

Exposure to phthalate esters such as DBP or diethylhexyl phthalate (DEHP) during early postnatal and sexually mature life stages has been shown to affect normal homeostasis of the male reproductive system and alter reproductive organ development and function. Experimental studies also report that peripubertal individuals are more susceptible to DBP and other diortho phthalate ester-induced effects in the male reproductive system [34] [12, 14, 21, 22, 40].

The hypothalamic-pituitary-gonadal axis and associated hormones play an important role in the development and function of the male reproductive system (see supplementary Figure 2). The hypothalamus initiates production of gonadotropin-releasing hormone (GnRH), which in turn stimulates the anterior pituitary to produce luteinizing hormone (LH) and follicle-stimulating hormone (FSH) [41]. LH activates Leydig cells to produce testosterone, which directs growth and maturation of the male reproductive system ([31, 42]). FSH and testosterone activate Sertoli cells, which interact with germ cells to promote spermatogenesis. Testosterone produced by LCs and inhibin produced by Sertoli cells also initiate a negative feedback mechanism which regulates GnRH, LH, and FSH production in the hypothalamus and anterior pituitary [43]. During sexually mature life stages the hormones produced in the pituitary-gonadal axis continue to regulate normal Sertoli and Leydig cell functions and spermatogenesis [42].

Phthalate-induced effects observed in pubertal or sexually mature animals include decreased reproductive organ weights, degeneration of seminiferous tubules, delayed puberty, germ cell apoptosis, and reduced fertility [14, 34, 40, 44]. Evidence from cell culture studies suggests that Leydig cell and Sertoli cell development and functions are directly affected by phthalate esters such as DBP [12, 40, 45]. Downstream effects of disruptions in LC functions and decreased testosterone production/levels include reduced weight/size of the testis, urethra and accessory reproductive organs [34]. Downstream effects of Sertoli cell disruption include reduced physical and metabolic support of germ cells, resulting in germ cell damage and reduced steroidogenesis [34, 40]. Phthalate-induced disruption in SC functions is also evidenced by altered production/levels of inhibin and Anti-Müllerian hormone (AMH)⁶. These findings from experimental studies are consistent with epidemiological studies which report a robust association between DBP exposure and altered semen parameters in men [11]. Finally, since testosterone plays an important role in normal Sertoli cell functions [15, 24], phthalate-induced inhibition of testosterone production in Leydig cells can indirectly target Sertoli cells.

Methods

Literature search and screening

The objective of the literature search and screening was to compare DBP-induced mechanisms that are well established in the rat with other mammalian experimental models, including studies using human fetal testis xenografts. The literature search and screening are described in detail in the supplementary materials. Briefly, experimental and mechanistic studies were identified from four online databases (PubMed, Web of Science, Toxline, and Toxic Substances Control Act Test Submissions 2.0). Non-date-limited searches with no language restrictions were applied. The initial search was conducted as early as March 2012 and was followed by literature search updates every 6−12 months through September 2018. Literature search results were screened and housed in the U.S. EPA’s Health and Environmental Research Online (HERO) database. Database searches were supplemented by manual searches of citations from regulatory documents. The Population, Exposure, Comparators, and Outcome (PECO) criteria used to screen the literature are also available in the supplementary materials.

Ex vivo studies that expose human fetal testis tissue cultures to DEHP or its metabolite mono (ethylhexyl) phthalate (MEHP) were also included in this analysis as DBP and DEHP target the male reproductive system via the same proposed mode of action [2, 10, 12, 16]. Ex vivo studies using DEHP or MEHP were identified from literature reviews on phthalate-induced male reproductive effects in various experimental models including human fetal testis xenografts and ex vivo tissue cultures [22–25].

Publications that met PECO criteria went through a full text review to select studies for data extraction and development of a literature inventory. As recommended by U.S. EPA’s Framework for Assessing Health Risk of Environmental Exposures to Children and the World Health Organization International Programme on Chemical Safety, the available mechanistic and toxicological studies and endpoints that inform the mode of action for DBP-induced male reproductive effects were evaluated according to the life stage of exposure [26, 27, 46, 47]. This approach was applied to account for biological differences, including endocrine, cellular and molecular, and functional aspects that are specific to each life stage. Examples include differences in Leydig cell populations in the developing fetus versus adults [48], differences in cellular targets (e.g. Sertoli cells have been hypothesized to be primary cellular target of phthalate-induced toxicity during postnatal stages [21, 40]), and the role of the pituitary gonadal axis in the development and function of the male reproductive system during gestation, puberty, and sexually mature life stages [13, 30]. Studies that did not report the age or life stage of the test model were not considered for information extraction, inventory development, and further review. Gestation studies that did not expose animals during critical organogenesis stages in the development of the male reproductive system (starting at approximately gestation day (GD) thirteen in rodents and gestational week 6 in the marmoset [13, 49, 50]) were not considered for further analysis. Studies that exposed animals during both pre- and post-natal life stages were not considered for analysis as it is not possible to ascertain whether effects reported can be attributed to chemical-induced alterations during gestation, peri-puberty or sexually mature life sages.

Literature inventory

Studies that met the criteria described above after full-text review were incorporated into a database/inventory used to facilitate the evaluation on individual key events and concordance analysis of effects conserved across species. The inventory was developed in Microsoft Access and exported to Microsoft Excel, and captures the following information from the studies: the U.S. EPA HERO database identification number for each reference, list of authors, publication year, and publication title; test compound, exposure route and duration; the sex, species and strain, and age/life stage at exposure and endpoint evaluation of the test organisms; the sample sizes for exposure group and endpoint testing; a succinct description of the exposure outcomes; and the key events informed by each endpoint reported in the study (see supplementary mechanistic effects database). Three distinctions were applied to the available studies when considering the age/life stage of exposure: gestation, peri-puberty, and sexually mature. Whenever a study provided the age, but not the life stage of the test organisms estimates reported in the U.S. EPA Review of the Reference Dose and Reference Concentration Processes⁷ [51] were used to obtain an approximation of the life stage equivalent to the reported age of exposure.

Application of the MOA-AOP framework.

The available mechanistic and toxicological evidence was organized and evaluated in concordance with the framework and levels of biological organization used for Mode of Action (MOA) analysis for non-cancer effects and development of Adverse Outcome Pathways (AOP)⁸ [27, 29, 47, 52]. The available evidence on DBP-induced effects in the male reproductive system was organized according to the following levels of biological organization: molecular target(s), cellular response(s), tissue/organ response(s), and organism response(s). Molecular initiating events, key events, and adverse outcomes identified in previous mechanistic and MOA evaluations were incorporated into the framework (see [8, 53]. Outcomes considered indicative to each key event are summarized in Tables 1 and 2. As recommended in U.S. EPA 2005, the analysis described here was focused on the concordance of key events and adverse responses across species to obtain clarification on issues raised by human xenograft studies and the relevance of animal studies to human health [54].

Table 1:

Evidence interpretation on key events in MOA for phthalate-induced male reproductive effects after gestational exposures.

Key event	Endpoints considered informative of key events
Molecular level interactions
Leydig cells	Molecular level interactions between DBP metabolites and potential receptors involved in cell signal transduction in Leydig cells.
Sertoli cells and germ cells	Molecular level interactions between DBP metabolites and potential receptors involved in cell signal transduction in Sertoli cells and/or germ cells.
Cellular effects
Leydig cells	Histological effects: Leydig cell aggregation and/or altered tissue distribution, ↓ Leydig cell number, ↓ Leydig cell endoplasmic reticulum, alterations in cell cytoplasm [38, 44].
	Functional changes: ↓ steroidogenic enzyme mRNA or protein levels and/or activity, ↓ testicular or serum testosterone levels, ↓ testicular testosterone production, ↓ INSL3 mRNA and/or protein levels [8, 10, 12, 38].
	Phenotypical changes indicative of reduced LC function: ↓ AGD, ↑ incidence of nipple retention [10, 34, 39].
Sertoli cells and germ cells	Germ cell effects: ↑ number of multinucleated germ cells (MNGs), ↓ germ cell numbers, ↑ germ cell death [8, 22, 38].
	Sertoli cell effects: Altered Sertoli cell development and interactions with germ cells, ↓ Sertoli cell numbers, changes in mRNA/protein levels of testicular genes indicative of Sertoli cell function during gestation16 [22, 24, 38, 39].
Organ effects
Urethra	Incomplete/abnormal development: ↓ organ weight/size, delayed preputial separation [16, 22].
Accessory reproductive glands	Incomplete/abnormal development: ↓ organ weight for accessory reproductive organ weights (e.g. epididymis, seminal vesicle, prostate, etc.), malformed epididymis, prostate, and/or seminal vesicle [15, 16, 22].
Testis	Incomplete/abnormal development: ↓ organ weight, ↓ gubernaculum development. [1, 38]
	Histological effects: Changes in seminiferous cord diameter (or seminiferous tubule dilation), incomplete or malformed seminiferous tubules, ↑ incidence of Sertoli cell only tubules, ↓ epidydimal-ductular cross sections, interstitial cell hyperplasia, [34, 38, 39].
Organism effects
Loss of reproductive functions assessed in adult animals	↓spermatogenesis, ↓ fertility, sperm abnormalities, ↓sperm counts [10, 34, 38].
Malformations: urethra	Increased incidence and/or severity of hypospadias [10, 14].
Malformations: testis	Increased incidence and/or severity cryptorchidism, increased incidence of other malformations (e.g. fluid filled testis) ([10, 38].

Table 2:

Evidence interpretation on key events in MOA for phthalate-induced male reproductive effects after exposures during postnatal life stages.

Key event	Endpoints considered informative of key events
Molecular level interactions
Leydig cells	Molecular level interactions between DBP metabolites and potential receptors involved in cell signal transduction in Leydig cells.
Sertoli cells and germ cells	Molecular level interactions between DBP metabolites and potential receptors involved in cell signal transduction in Sertoli cells and/or germ cells.
Cellular effects
Leydig cells	Histological effects (e.g. hyperplasia, altered cell number) [22, 44].
	Functional changes: ↓ steroidogenic enzyme mRNA or protein levels and/or activity, ↓ testicular or serum testosterone levels/production [3, 14, 24].
	Phenotypical changes indicative of altered LC function: ↓AGD in animals exposed during early postnatal life stages [22].
Sertoli cells and germ cells	Germ cell effects: ↓ germ cell numbers, ↑ germ cell death, altered markers of germ cell maturation and spermatogenesis17 [3, 22, 38].
	Sertoli cell effects: Altered Sertoli cell development and germ cell attachment/interactions, cytoskeletal alterations, changes in mRNA/protein levels of testicular genes indicative of Sertoli cell development and functions18 [22, 34, 40, 126]
Organ effects
Pituitary gonadal axis	Alterations in organ weights and production/levels of FSH hormone, LH [34, 44, 150].
Accessory reproductive organs	Incomplete development, ↓ organ weight/size [3, 14, 44].
Testis	Incomplete development, reduced organ size: ↓testis weight [34, 44].
	Histological effects: seminiferous tubule alterations, ↑ incidence of germinal epithelium atrophy or Sertoli cell only tubules [1, 3, 22, 44].
Organism effects
Loss of reproductive functions	↓ fertility19, ↓ spermatogenesis or markers of spermatogenesis, ↓ sperm count and sperm motility, ↑ sperm abnormalities [22, 44, 126].

Results

Summary of male reproductive effects from exposures during gestation (Figure 1, Table 3).

Figure 1:

MOA pathway corresponding to DBP-induced male reproductive effects after gestational exposure.

Table 3.

Concordance analysis of key events in DBP MOA for gestational effects.

Key event	Evidence from animals	Evidence from human fetal testis xenografts or ex vivo tissue cultures	References
Molecular interactions
Leydig cells	Molecular target for DBP/MBP has not been identified.	Molecular target for DBP/MBP has not been identified.
Sertoli cells or germ cells	Molecular target for DBP/MBP has not been identified.	Molecular target for DBP/MBP has not been identified.
Cellular effects
Leydig cell effects	Rat studies consistently report that DBP/MBP exposure alters LC development and function. One study in rabbits reports supportive evidence. Most mouse studies report no exposure related effects. One study in marmosets reports no exposure related effects.	Evidence obtained from human xenograft and explant studies suggest lower sensitivity to DBP-induced LC effects.	Rat (see reviews cited above and supplementary database); rabbit [67]; mouse [65]; [66]; [63]; [64]; [73]); marmoset [68]; human xenograft or explant [59–63]
Sertoli cell and/or germ cell effects	Rats, rabbits, and mice studies report GC and SC effects. E.g. ↓ SC numbers, altered SC cytoskeleton and SC-GC interactions, impaired GC development, and ↑ GC apoptosis	Evidence obtained from human xenograft and explant studies suggest exposure resulted in GC and SC effects.	Rat (see reviews cited above and supplementary database); rabbit [67, 151]; mice [56, 63–65, 73, 151]; marmoset [68]; human xenograft [56, 60, 62, 63, 66]
Organ effects
Urethra	Rat studies reporting altered development resulting in malformations. One study using marmosets and one study in rabbits reported no exposure related effects	No available evidence.	Rat (see reviews cited above and supplementary database); marmoset [68]; rabbit [67]
Accessory reproductive organs	Rat studies consistently report altered and delayed development of accessory male reproductive organs. One study in rabbits reports supportive evidence. Mouse and marmoset studies and human xenograft studies report no exposure related effects.	Yes. Human xenograft studies evaluated accessory reproductive organ weights in rodent hosts suggest lower sensitivity to DBP-induced LC effects.	Rat (see reviews cited above and supplementary database); rabbit [67]; marmoset [68]; human xenograft [37]; [61]
Testis	Rat studies consistently report altered or delayed testis development. One study in rabbits reports supportive evidence. Majority of mouse studies report no exposure related effects. One study in marmoset reports no exposure related effects.	No available evidence.	Rat (see reviews cited above and supplementary database); rabbit [67]; mice [65, 73, 77, 122]; marmoset [68]
Organism effects
Reproductive functions	Rat studies report decreased sperm production, increased sperm abnormalities and reduced fertility. One study using rabbits reports alterations in sperm parameters. One study in marmosets reports no exposure related effects.	No available evidence.	Rat (see reviews cited above and supplementary database); rabbit [67]; marmoset [68]

Molecular interactions: Leydig cells and seminiferous cord (Sertoli cells and germ cells).

Many experimental studies have attempted to identify the molecular target(s) for DBP in Leydig cells, or in the seminiferous cords (i.e. germ cell and Sertoli cells) [24]. To date, a molecular target for DBP or its bioactive metabolite has not been identified. Activation of the peroxisome proliferator activated receptors (PPAR) has been proposed as a potential target for DBP [34, 55], and human fetal testis tissue cultures exposed to MEHP for three days display increased PPARγ mRNA levels [56]. However, the fetal rat testis does not express PPAR-responsive genes after exposure to phthalate esters, and gestational treatment with a PPARα or PPARγ−specific ligands failed to reproduce effects consistent with phthalate-induced responses in the male reproductive system [24, 57, 58]. The available evidence from experimental studies suggests that PPARs are not involved in phthalate-induced male reproductive effects.

Cellular effects: Leydig cell development/functions

Evaluation of the available studies using different mammalian experimental models or human fetal testis xenografts or explants suggest that Leydig cell responsiveness to DBP, MBP, or MEHP varies according to species (see Table 3). Unlike effects observed in gestational exposure studies using rats, ex vivo and xenograft studies using human fetal testis report no significant effect on Leydig cell functions. Studies from two independent groups using human fetal testis explants report that MEHP exposure did not affect basal or human chorionic gonadotropin, 22R-hydroxy-cholesterol, or LH – stimulated testosterone synthesis [59, 60]. Furthermore, MEHP exposure did not affect mRNA levels of INSL3, the cholesterol transporter steroidogenic acute regulatory protein (StAR) or the steroidogenic enzymes P450scc, or P450c17 [60]. Human fetal testis xenograft studies from two independent groups report that host animal exposure to DBP or MBP did not affect hCG-stimulated serum testosterone levels [61, 62]. Additionally, DBP had no significant effect on basal mRNA levels of the steroidogenic enzymes: CYP11A1, CYP17A1, or Scarb1 in human fetal testis xenografts [63]. Nevertheless, Heger et al. 2012 reported that DBP exposure lead to a ~80% decrease in INSL3 mRNA levels, while Mitchell et al. 2012 reported a ~30% reduction in testosterone, but these effects were not statistically significant [22].

Various in vivo studies using other mammalian models also report different results from those observed in the gestational rat studies. In vivo and ex vivo studies using various strains of mice report no significant change (or in some cases an increase) in basal testosterone concentrations and mRNA levels of steroidogenic enzymes, and no changes in AGD [24, 63–66]. However, a study using mouse fetal testis explants reported MEHP exposure lead to reduced responsiveness to LH-stimulated testosterone production [64]. The observations made in rat studies are supported by an experiment using Dutch-Belted rabbits which reported that gestational exposure to DBP decreased basal serum testosterone, but had no significant effect on gonadotropin releasing hormone (GnRH)-induced testosterone production [67]. A study on marmosets reported no significant effects on serum testosterone after gestational exposure to DBP [68].

Cellular effects: Seminiferous cord (Sertoli cells and germ cells)

An evaluation of the available evidence on effects in the seminiferous cord suggests that responses to DBP or MBP are conserved across species (see Table 3). Cellular and histopathological changes observed in the seminiferous cord include changes in the development, functions, and interactions between germ cells and Sertoli cells. In rats, these effects have been shown to result in reduced numbers of germ cells and sperm production [34, 38].

Germ cell effects such as increased cell death and formation of MNGs occur independently of DBP-induced alterations in testosterone production [12, 24], and appear to be conserved across species including rats, mice, and human fetal testis xenografts and explants. One in vivo study using marmosets reported that exposure to MBP during gestation had no statistically significant effects on germ cell or Sertoli cell numbers, but undifferentiated germ cell clusters were noted in 33% of the exposed animals [68]. Formation of MNGs, an effect that is observed in rats after exposure during late gestation [8, 69, 70], has also been observed in studies where human fetal testis xenografts were treated with DBP [62, 63, 66], and in several mouse strains exposed to DBP or MEHP [63, 66]. In vivo studies using rats have also observed that gestational exposures to DBP results in a decrease in the total number of germ cells [20, 69, 71]. These findings are supportive of the results from a study using human fetal testis xenografts which reported that exposure to DBP for 21 days caused a decrease in the total number of germ cells and the number of undifferentiated germ cells [66]. However, a human fetal testis xenograft study performed by a separate group reported that DBP exposure for 24 to 72 hours had no significant effect on germ cell numbers [63]. Studies which expose human, mice, and rat fetal testis explants, or in vivo mice to MEHP report decreased germ cell numbers and increased germ cell apoptosis [56, 60, 64, 72].

DBP-induced responses in Sertoli cells also appear to be conserved across species including rats, mice and human fetal testis explants and xenografts. In rats these effects are characterized by a reduction in number of Sertoli cells and alterations in their cytoskeleton, disrupted interactions between Sertoli cells and germ cells, and a decrease in AMH production [22]. In mice, exposure to MEHP or DBP was associated with decreased Sertoli cell numbers and reduced AMH protein levels [64, 73]. Studies using in vivo mice or human ex vivo tissue cultures report no DBP or MEHP-induced alterations in Sertoli cell numbers, but AMH levels were decreased in both species [60, 64, 72]. Furthermore, a study using human fetal testis xenografts reports that DBP exposure was associated with decreased mRNA levels of the follicle-stimulating hormone receptor (FSHR)⁹ [62].

Organ effects: Development of the testis, urethra, and accessory reproductive organs.

The analyses of the available studies that use different mammalian in vivo experimental models and human fetal testis xenografts or explants suggest that DBP (or MEHP)-induced responses on androgen-dependent development of the male reproductive system varies across species (see Table 3). In rats, and in one study using rabbits, gestational exposure to DBP was shown to inhibit testosterone-dependent development of the testis, urethra and accessory male reproductive organs [8, 67]. Evaluation of the available mouse studies that expose animals to DBP during gestation suggest that responses may be strain specific: Qin et al. 2017 and Gaido et al. 2007 respectively observed that C57BL/6 mice displayed decreased testicular weight [73] and increased seminiferous cord diameter [65], while NTP 1995 reported no significant changes in testis weights in B6C3F1 mice [65]. These discrepancies in responses may have also been due to differences in experimental design as the Qin et al. 2017 and Gaido et al. 2007 studies exposed animals during gestation only whereas the NTP 1995 study exposed animals during gestation and postnatal life stages. A study in marmosets [68] reported that exposure to MBP had no significant effects in the weights of the testis or accessory reproductive organs. Two studies using human fetal testis xenografts grafted into castrated mice reported no significant DBP or MBP-induced effects in the seminal vesicle weight of the host animal [61, 62], whereas in castrated animals grafted with rat fetal testis DBP exposure resulted in decreased seminal vesicle weight [61].

Organism effects: Malformations (hypospadias and cryptorchidism), and loss of fertility.

Analysis of the available evidence from experimental studies suggests that DBP-induced male reproductive effects at the organism level are also species-specific for some outcomes (see Table 3). In rats, gestational exposure to DBP is associated with increased incidence of cryptorchidism and hypospadias and reduced reproductive functions such as decreased sperm production and increased sperm abnormalities [8, 9]. These observations are supported by a study in rabbits which reported decreased sperm concentration and ejaculate volume, and increased abnormal sperm after gestational exposure to DBP [67]. However, a gestational study using marmosets reported that MBP exposure did not have a significant effect on the incidence of cryptorchidism or hypospadias, or on spermatogenesis in adult animals [68]. Studies using other animal models have not evaluated outcomes indicative of changes in fertility.

Summary of male reproductive effects from exposures during postnatal life stages (Figure 2, Table 4).

Figure 2:

MOA pathway corresponding to DBP-induced male reproductive effects after postnatal exposure.

Table 4.

Concordance analysis of key events in DBP MOA for early postnatal effects

Key event	Evidence in animal in vivo, xenograft and cell culture models	References
Molecular interactions
Leydig cells	A molecular target for DBP/MBP has not been identified.
Sertoli cells or germ cells.	A molecular target for DBP/MBP has not been identified.
Cellular effects
Leydig cell effects.	In vivo and cell culture rat studies consistently report that DBP/MBP exposure alters LC development and functions. Two studies in non-human primates (rhesus monkeys and marmosets) report evidence of altered LC development and functions. Most of in vivo studies using mice report effects that are consistent with observations made in rats. All cell culture studies using immortalized mouse cells or dog primary LCs report altered LC functions.	Rat (see reviews cited above and supplementary database); marmosets [59]; rhesus monkeys [121]; mice in vivo [77, 94, 95, 125]; mice cell culture [45] [97–100, 152]; dog cell culture [101]
Sertoli cell and/or germ cell effects.	Rat in vivo and cell culture studies report alterations in SC cytoskeleton, functions, SC-GC interactions, and GC maturation. One in vivo study using marmosets, one xenograft study using rhesus monkeys, and all available mouse in vivo and cell culture studies report evidence of SC and GC effects.	Rat (see reviews cited above and supplementary database); mice [17, 77, 83, 92, 94, 95]; marmosets [68]; rhesus monkeys [121]
Organ effects
Pituitary gonadal axis	Majority rat studies report alterations in LH & FSH levels. One study using rabbits reports increased hypothalamic GnRH levels while three mouse studies report no exposure related effects.	Rat (see reviews cited above and supplementary database); rabbits [67]; mice [77, 94, 95]
Accessory reproductive organs.	Rat studies report alterations in accessory reproductive organs weights and histopathology. One study in rabbits, and one xenograft study using non-human primates report supportive evidence for effects in seminal vesicle and prostate. Majority of mouse studies report no exposure related effects.	Rat (see reviews cited above and supplementary database); rhesus monkeys [121]; mice [77, 94, 122–124]
Testis.	Rat studies and most of the mouse studies report decreased weight and histopathological effects. One study in guinea pigs reported decreased weight. Syrian hamsters and one rabbits were not responsive. However, histopathological observations were observed in rabbits.	Rat (see reviews cited above and supplementary database); Guinea pigs and Syrian hamsters [17]; mice [17, 77, 83, 94, 123, 125]; rabbits [67].
Organism effects
Reproductive functions.	Rat studies report adverse effects on sperm parameters. One study in Guinea pigs and one study using rabbits report alterations in sperm parameters. Majority of mouse studies report alterations in sperm parameters and fertility.	Rat (see reviews cited above and supplementary database); Guinea pigs [17]; mice [77, 94, 122–125, 127]; rabbits [67].

Molecular interactions: Leydig cells and seminiferous cord (Sertoli cells and germ cells).

As described above, experimental studies have attempted to identify a molecular target for DBP in the testis. Two studies evaluated the potential role of DBP exposure during post-natal life stages on PPARα and γ functions. Zhang et al. (2014) exposed primary Sertoli cells obtained from peripuberal Sprague-Dawley rats to 100 μM DBP or MBP and reported that cells treated with DBP displayed increased protein levels and DNA binding of PPARα [74]. These results should be interpreted with caution as Zhang et al. 2014 observed PPARα-related effects in DBP-treated cells, but MBP exposed cells were not responsive [74]. Furthermore, Sertoli cells are not normally exposed to DBP as it is metabolized in the intestines and liver and absorbed almost entirely as the monoester metabolites [1, 2]. In peripubertal rats DBP exposure lead to a mild increase in PPARα responsive genes and only one of two assayed PPARγ responsive genes was affected by DBP exposure [75]. Overall these findings suggest that PPAR activation may play a role in DBP-induced effects in the testis after postnatal exposure, however additional studies would help determine whether these nuclear receptors play a critical role in phthalate-induced male reproductive toxicity.

Cellular effects: Leydig cell development/functions

Evaluation of the available evidence from experimental studies on postnatal exposures to DBP suggests that Leydig cell effects are conserved across species (see Table 4). Most of the available studies using rats report that oral exposure to DBP during postnatal stages results in altered Leydig cell functions. Peripubertal rats exposed to DBP had reduced testicular and serum testosterone levels [76–80], altered expression of steroidogenic enzymes [78, 81], and increased expression of AT1, a cellular stress marker expressed in Leydig cells [82]. However, two studies using Wistar rats reported no changes in serum testosterone in adult animals (PND90) after exposure during lactation [66], or increased testicular testosterone in peripubertal animals exposed for seven days [83]; and three studies using peripubertal Sprague-Dawley rats report no effects in testicular testosterone after exposure for seven [84] or twelve [85] days, or eight weeks [86]. The study by Alam et al. 2010 [84] also observed that acute exposure to DBP (3, 6, or 24 hours) decreased the testicular mRNA levels of steroidogenic enzymes and testosterone concentration. In sexually mature rats DBP exposure resulted in decreased testosterone levels, but this response appeared to vary according to the duration of the experiment. In Wistar and Sprague-Dawley rats DBP treatment for 14 or 15 days lead to decreased serum testosterone levels [87–89], but in Sprague-Dawley rats shorter exposures (3, 6, or 7 days) resulted in increased or no significant effects on serum testosterone [79, 90]. In-vivo findings on DBP-induced alterations in Leydig cell development/functions are supported by in vitro studies using primary testicular cells obtained from pre-weaning, peripubertal and sexually mature Sprague-Dawley rats, which report that treatment with DBP or MBP decreased androgen production [91, 92] and mRNA levels of genes associated with steroidogenesis as well as lipid and cholesterol metabolism [91, 93]. The discrepancies observed in rat studies using young animals represent an uncertainty in the available evidence and do not appear to be due to strain differences or other experimental design features. Although a majority of the available rat studies appear to suggest that DBP exposure alters Leydig cell functions, additional experiments may be necessary to identify variables which may contribute to the different responses described above.

The majority of in vivo and cell culture studies using mice provide supportive evidence that DBP- induced Leydig cell effects are conserved across species. An in vivo study using four-day-old C57BL/6J mice reported decreased AGD and serum testosterone after exposure to DBP for 14 days [94]. A follow up study by the same group also reported that although early life stage exposure did not alter serum testosterone levels in adult animals, treated animals had increased numbers of Leydig cells and reduced testicular mRNA levels of the steroidogenic enzyme CYP11A1 [95]. However, a subchronic toxicity study using B6C3F1 mice reported that exposure to DBP for 13 weeks had no significant effect on serum testosterone levels [77]. These differences in response could have been due to experimental design features such as the age of the experimental animals used: NTP, 1995 initiated exposure to DBP at post natal day 42 (PND42) while the Moody et al. 2013 began treatment on PND4. Cell culture studies using immortalized Leydig cell lines obtained from various mouse strains including C57BL/6J and BALB/c mice provide further evidence that DBP exposure can disrupt normal Leydig cell functions. In MA-10 and MLTC-1 cells generated from C57BL/6J mice, exposure to MBP or DBP resulted in reduced steroid hormone production and decreased mRNA levels of steroidogenic enzymes and INSL3 protein [45, 96–98]. These findings are supported by a study using TM3 Leydig cells generated from BALB/c mice where exposure to DBP was associated with an inhibition of LH-stimulated steroidogenesis [99]. Interestingly, two studies using MLTC-1 cells report that treatment with low concentrations of MBP (between 0.1 and 1 μM) caused a significant increase in steroidogenesis and in mRNA levels of steroidogenic enzymes and INSL3 protein [98, 100].

In addition to evidence obtained from experimental studies using rats or mice, in vivo studies using Dutch-Belted rabbits and marmosets also report that exposure to DBP during peripubertal life stages was associated with decreased serum testosterone levels [59, 67]. However, when sexually mature rabbits were exposed at the same level, there were no significant changes in serum testosterone [67]. These observations are consistent with previous analyses which conclude that earlier life stages are more sensitive to DBP-induced effects in the male reproductive system [12, 34, 40]. In neonatal marmosets, DBP-induced decrease in serum testosterone was observed after short term exposure for 5 hours [59]. When animals were exposed for 14 days there was an increase in Leydig cell volume, but no changes in serum testosterone. According to the study authors the increase in Leydig cell volume may have been due to elevated LH levels resulting from the initial decrease in steroidogenesis caused by MBP exposure [59].

Finally, studies using primary testicular cells or xenograft tissue transplants provide further supportive evidence suggesting that exposure to DBP during early postnatal life stages alters Leydig cell functions. A xenograft study using testicular tissue obtained from sexually immature rhesus macaques, reported that exposure to DBP for 14 or 28 weeks decreased mRNA levels of INSL3 and steroidogenic enzymes [59], and in primary testicular cells isolated from sexually mature dogs, treatment with DBP resulted in decreased INSL3 levels and testosterone production [101]. Overall, the available evidence suggests that DBP exposure during, early postnatal, peripubertal, and sexually mature life stages can affect Leydig cell development and functions on several mammalian species.

Cellular effects: Seminiferous cord development and functions

Evaluation of the available experimental evidence suggests that DBP-induced effects in seminiferous tubules are conserved across various species (including rats, mice, and rhesus macaques) (see Table 4). As described below, most of the available experimental studies report that DBP targets Sertoli and germ cell development and functions.

Evidence from in vivo studies using rats exposed at different ages and life stages consistently report that DBP alters normal cellular development and functions in the seminiferous cords. In peripubertal Wistar and Sprague-Dawley rats, exposure to DBP lead to increases in germ cell death, reduced number of spermatogenic cells [75, 79, 84, 102–104], as well as decreased testicular Zn levels¹⁰ [77, 83], and alterations in gene mRNAs associated with germ cell development and spermatogenesis [78, 81]. These studies also report that DBP exposure caused alterations in mRNA levels of genes indicative of Sertoli cell functions [75], increased apoptosis and histopathological malformations of Sertoli cells such as alterations in cytoskeleton and increased vacuolation¹¹ [79, 84–86, 102–104], and reduced the number of Sertoli cells [76, 78]. Most of the available studies report that exposure to DBP also affected seminiferous cord structure and functions in sexually mature rats, but three studies report no significant effects after exposure [79, 88, 105]. In Wistar and Sprague-Dawley rats treatment with DBP lead to decreased testicular Zn levels [106–108], reduced the number of germ cells [89, 106–113], and altered markers of Sertoli cell function and spermatogenesis [87, 107–110, 112, 114, 115].

In vitro studies using primary cells isolated from rat testis provide supportive evidence of observations made in in vivo studies. In primary Sertoli cells or Sertoli-germ cell co-cultures isolated from peripubertal Sprague-Dawley rats, exposure to MBP or DBP caused increased germ cell detachment, reduced viability, and increased expression of genes associated with apoptosis [116, 117]. In primary Sertoli cells, exposure to DBP or MBP lead to reduced cellular membrane integrity and cell viability [85, 117–119], increased permeability of Sertoli cell tight junctions and decreased mRNA and protein levels of tight junction proteins [85, 120], alterations in inhibin B mRNA and protein levels [85], and disruption of vimentin filaments and collapse of the cellular cytoskeleton [74, 103].

Most in vivo and cell culture studies using other mammalian models provide evidence that is consistent with observations made in rats both in vivo and in vitro. In peripubertal mice of various strains DBP exposure reduced germ cell numbers [17, 94], decreased Sertoli cell proliferation, and altered expression of AMH and other markers of Sertoli cell development and function [94]. Similar observations were made in rhesus macaques testis xenografts isolated from six-month-old animals in which exposure to DBP lead to a decrease in Sertoli cell numbers [121]. However, unlike the response observed in mice, DBP did not affect AMH or other markers of Sertoli cell maturation in the macaque xenografts, but the same study also reported decreased numbers of spermatocytes and spermatogonia when host animals were exposed to DBP [121]. Testicular Zn levels were decreased in peripubertal mice exposed to DBP for 7 days [83]. However, one study using B6C3F1 mice reported increased testicular Zn levels after DBP exposure for 13 weeks [77]. The observations from in vivo and xenograft studies on DBP-induced effects in germ cells are supported by cell culture studies using mouse GC-2 cells, a spermatocyte cell line isolated from BALB/c mice. Treatment with DBP reduced GC-2 cell viability, increased endoplasmic reticulum (ER) damage, and activation of ER-mediated autophagy [92]. One study using four-day-old marmosets exposed to DBP for fourteen days reported no significant changes in germ cell numbers [68], but studies that used rats or mice and exposed animals to similar DBP doses and durations (ten to twenty one days) report significant effects on Sertoli cells and germ cells [79, 85, 88, 94, 105, 107]. This suggests that marmosets may be less responsive to DBP-induced germ cell effects. However, McKinnell et al. 2009 is the only study that has evaluated outcomes informative of seminiferous tubule effects in marmosets, and additional experiments using different exposure durations are needed to fully characterize DBP-induced responses in this species [68]. Overall, the available evidence suggests that exposure to DBP during peripubertal and sexually mature life stages affects seminiferous tubule development and cellular functions in at least three different mammalian species.

Organ effects: Hypothalamic pituitary gonadal axis

Few in vivo studies have evaluated the potential effects of DBP exposure in the pituitary gonadal axis. The available evidence suggests that DBP-induced alterations in various pituitary hormones may be species-specific. However, the available evidence is variable and the direction of DBP-induced changes does not appear consistent from one study to another (see Table 4).

Several studies using Sprague-Dawley rats have reported DBP-induced alterations on serum LH levels. However, the magnitude and direction of these hormonal responses varies across studies. In both peripubertal and sexually mature animals, exposure to DBP for 12, 15, or 30 days resulted in decreased or increased serum LH levels [76, 79, 85, 88]. Studies in which animals were exposed for shorter periods of time (ranging from 3 hours to 6 days) report no effects or minor (non-statistically significant) decreases in serum LH [84, 90]. The contrasting observations amongst the studies cited above may have been due to variety in experimental design such as exposure durations and/or differences in responsiveness in young versus sexually mature animals.

In rats, DBP exposure resulted in altered serum FSH levels, but the direction of the induced changes appeared to be due to biological differences amongst strains and/or life stages. In peripubertal Sprague-Dawley rats, exposure to DBP increased serum FSH levels [79, 85] while in sexually mature animals, treatment with DBP lead to a decrease in serum FSH in Wistar rats [87, 89], but no effect in Sprague Dawley rats [79].

Studies using rabbits and mice suggest that these two species appear to be insensitive to DBP-induced effects in the hypothalamus and pituitary. Two studies using neonatal C57BL/6J mice reported that DBP exposure for 3, 10 or 17 days had no significant effect on serum LH [95] or FHS [94, 95], while a 13-week exposure study using B5C3F1 mice reported no response on pituitary histopathology [77]. In peripubertal and sexually mature Dutch-Belted rabbits, exposure to DBP for 8 to 12 weeks did not alter gonadotropin releasing hormone serum levels [67].

Organ effects: Accessory reproductive organs

In rats, most of the available studies report that exposure to DBP during peripubertal or sexually mature life stages results in alterations in the development and function of the accessory male reproductive organs. Postnatal exposure to phthalates, including DBP, has been shown to cause effects such as decreased epididymis, prostate, and seminal vesicle weight; delayed preputial separation; and structural degeneration of the epididymis and vas deferens (see Table 4, and supplementary mechanistic effects database) [14, 44].

Most of the available studies using other mammalian models report that exposure to DBP also results in altered accessory reproductive organ weight. However, mice appear to be less responsive. One study using mice reported decreased preputial gland weight [95] in exposed animals, but the same group observed no treatment related effects in seminal vesicle or prostate weights. Additional studies by other research groups note that when mice were treated with DBP during peripubertal life stages no effects were observed in prostate, epididymis, and seminal vesicle weight [122, 123]. In DBP-treated Dutch-Belted rabbits, accessory reproductive organs were decreased in peripubertal animals while sexually mature animals were not responsive [67]. These observations in rabbits are supported by a xenograft study which reported that mice implanted with testis tissue from six-month-old Rhesus macaques and exposed to DBP had decreased seminal vesicle weights [121].

Organ effects: Testis weight

In rats, postnatal exposure to phthalates has been shown to cause adverse testicular effects [34, 44]. Most available rat studies report that Sprague Dawley, Wistar, and F344 rats exposed to DBP had decreased testis weight and increased incidence of adverse histopathological outcomes such as seminiferous tubule degeneration, and elevated numbers of apoptotic cells; see supplementary mechanistic effects database (see Table 4, and supplementary mechanistic effects database).

Studies using other mammalian models do not present consistent evidence on DBP-induced alterations in testis weight (see Table 4). In peripubertal Syrian hamsters¹² and marmosets, exposure to DBP had no significant effect on testis weight or histopathology [17, 59, 68]. Experimental studies using peripubertal mice report an exposure-related decrease in testis weight [17, 83, 94, 124] and increased seminiferous tubule atrophy [17, 125]. However, several studies by other groups report no effects on testis weight and histopathology, and in some cases increased testis weight [77, 122, 123, 125]. Studies that report no significant effect on testis weight generally use sexually mature animals, and exposed animals for longer periods of time (ranging from 14 to 133 days) than studies that observed decreased testis weights (ranging from 7 to 56 days). In peripubertal Guinea pigs and in peripubertal and sexually mature New Zealand white rabbits, DBP exposure was associated with histopathological alterations in the seminiferous tubule [17, 67]. These findings are supported by a study using Rhesus macaque xenografts obtained from six-moth-old animals, which reported that DBP exposure increased the incidence of Sertoli Cell Only tubules¹³, and decreased grafted testicular tissue weight [121].

Organism effects: Loss of reproductive function

Evaluation of the available experimental evidence suggests that DBP-induced loss of reproductive functions is conserved across various species. As described below, most of the available experimental studies report that DBP exposure causes alterations in sperm parameters and fertility measures (see Table 4).

In rats, postnatal exposure to DBP has been consistently shown to disrupt/alter markers of male reproductive fertility [22, 44, 126]. Effects observed after peripubertal and sexually mature life stages include disruption in spermatogenesis, decreased sperm numbers and concentration, reduced sperm motility, and increased sperm abnormalities (see supplementary mechanistic effects database).

Most of the studies that use other mammalian species (including mice, guinea pigs, and rabbits) report outcomes that are consistent with observations made in rat experiments. Studies using C57BL/6J or Pzh:Sfis mice report that exposure to DBP during peripubertal life stages lead to delayed spermatogenesis, increased sperm abnormalities, and decreased sperm count and motility [94, 122, 124, 125]. In sexually mature Pzh:Sfis mice, DBP exposure was associated with decreased fertility and percent pregnancies [127]. However, evaluation of the available studies also suggests that susceptibility to DBP-induced effects on male fertility may also vary according to strain. In CD-1 and B6C3F1 mice, DBP exposure did not affect sperm motility, concentration, or incidence of abnormalities [77, 123]. The variability in responses observed amongst mouse strains could be due to toxicokinetic differences, but additional studies are needed. DBP-exposed guinea pigs during peripubertal life stages had decreased numbers of spermatocytes, spermatids, and spermatogonia [17], whereas an in vivo study using peripubertal and sexually mature rabbits reported increased sperm abnormalities, but no effects on other sperm parameters or mating behavior [67].

Discussion and conclusions

Comparisons across species: gestational exposure related effects

Based on the evaluation of the evidence from gestational DBP exposure studies rats appear to be more sensitive to DBP-induced anti androgenic effects than other mammalian models including human xenografts and ex vivo fetal testis tissue cultures. In vivo and xenograft studies using fetal mice as a model organism report various phthalate-induced male reproductive outcome which are like those observed in rats, but mice did not appear to be responsive to DBP exposure levels shown to cause changes in testosterone synthesis and downstream effects in fetal rats (see Table 3). In general, androgen-independent outcomes such as seminiferous tubule and germ cell effects have been observed in both rats and mice. In vivo and ex vivo studies using several strains of mice exposed to DEHP or dipentyl phthalate (DPP) during gestation provide mixed results: while several studies reporting effects consistent with those seen in the fetal rat [128–133], others observe no changes in androgen production or levels [64, 134, 135]. Additional studies are needed to determine whether these diverse responses to phthalate-induced anti-androgenic effects in the mouse model are due to strain differences or other experimental factors.

As discussed above, several studies have evaluated androgen-dependent and -independent effects of phthalate exposure on human fetal testis explants as well as in vivo xenografts (see Supplementary Figures 3 and 4). In both models, exposure to DBP or its bioactive metabolite MBP did not affect testicular testosterone production. Previous reviews and analyses of the human fetal testis xenograft and explant studies conclude that humans appear to be refractory to DBP or MEHP-induced inhibition of androgen synthesis [24, 25]. However, several issues have been identified in these studies and should be considered in a hazard and/or mode of action evaluation:

• In experiments that use human fetal testis ex vivo cultures, the potential variation in androgen synthesis by the individual tissue explants could affect the statistical power of the study [33, 39].

• The results obtained from the serum testosterone assay used in Mitchell et al. (2012) appeared highly variable. This has been attributed to the inherent biological variability of the samples, the small sample size (n = 3 to 13 individuals) used in the study, and possible technical difficulties [22, 33, 39].

• Although the fetal rat xenografts were responsive, a comparison with rat studies suggest that the methods used to measure testosterone levels in the Mitchell et al. 2012 xenograft study may not be as sensitive as those used for in vivo experiments [22].

• The age of the human fetal testis tissue used in the xenograft studies does not match the period of peak testosterone production during the masculinization programming window¹⁴. This has raised concerns about the responsiveness of this test model [33, 39].

However, human fetal testis explants and xenografts do appear to be sensitive to known disruptors of testicular androgen synthesis. Hallmark et al. 2007 used ketoconazole, an established inhibitor of testosterone synthesis¹⁵, as a positive control and reported that human fetal testis explants exposed to this agent displayed decreased basal and hCG-stimulated testosterone production. Furthermore, the xenograft study by Spade et al. 2014 also treated human fetal testis xenografts to abiraterone, a selective and irreversible CYP17A1 inhibitor [136], and they observed decreased serum testosterone in the exposed host animals. These findings suggest that the ex vivo and xenograft models are responsive to known disruptors of androgen synthesis and raise confidence in the strength and biological plausibility of effects reported in these studies.

Evaluation of effects that occur independently of phthalate-induced disruption in androgen synthesis (i.e. germ cell and Sertoli cell outcomes) suggests that adverse responses observed in the seminiferous tubule after gestational exposure are conserved across mammalian models including human fetal testis xenografts and ex vivo tissue incubations. In vivo studies using rats, rabbits, and mice report that exposure to DBP resulted in similar adverse responses including increased incidence of MNGs and germ cell apoptosis, decreases in germ cell and Sertoli cell numbers, and alterations of Sertoli cell cytoskeleton and Sertoli-germ cell interactions (see Table 3). An in vivo gestational exposure study using marmosets, however, reported no adverse germ cell or Sertoli cell effects. Studies using human fetal testis xenografts or ex vivo tissue cultures report androgen-independent outcomes that are consistent with effects observed in rats, rabbits and mice. Overall, the available evidence suggests that androgen-independent adverse responses are conserved across species, including human fetal testis xenografts and ex vivo tissue cultures. Experimental models such as the fetal rat or mouse are therefore considered informative of potential human responses in the seminiferous cord after gestational exposures to phthalates such as DBP.

Comparisons across species: postnatal exposure related effects

Evaluation of the evidence from DBP studies that expose animals during postnatal life stages suggests that androgen–dependent outcomes are conserved across species. Most of the available studies in which peripubertal and sexually mature rats were exposed to DBP report LC-related effects such as decreased testicular and serum testosterone levels, reduced Leydig cell numbers and mRNA levels of genes associated with steroid synthesis (see Table 4). Similar outcomes were reported in in vivo studies using rabbits, mice, and marmosets where exposures to DBP also lead to reduced testicular and serum testosterone levels, AGD, and expression of steroidogenic enzymes. These findings are supported by cell culture experiments using dog-, mice-, or rat- derived primary or immortalized cells, which report that exposure to DBP or MBP lead to reduced LC viability, altered expression of steroidogenesis-related genes, and decreased basal or stimulated testosterone production (see Table 4). Furthermore, a xenograft study using rhesus monkeys reported decreased mRNA levels for INSL3 and genes associated with steroidogenesis and reduced accessory reproductive organ weights in host animals. As noted above several studies using rats or mice reported contradicting evidence on DBP-induced effects on Leydig cell functions. In some cases (e.g. exposures in sexually mature rats) these discrepancies appear to be due to experimental design differences. Although the majority of the available studies suggest that DBP exposure disrupts Leydig cell functions these discrepancies represent a source of uncertainty and additional studies are needed to provide clarification.

Androgen-independent adverse outcomes also appeared to be conserved across species. In rats and mice exposure to DBP was associated with altered Sertoli cell development and functions, reduced germ cell numbers and disruptions in spermatogenesis. These findings are supported by mouse and rat cell culture studies that report increased Sertoli and germ cell damage after DBP treatment. A study in marmosets observed alterations in germ cell number, but these effects were not statistically significant. However, a xenograft study using young (six-month-old) rhesus monkeys reported that DBP exposure lead to disrupted spermatogenesis, and reduced Sertoli cell numbers and proliferation. Finally, in vivo studies that evaluate reproductive functions report reduced male fertility in DBP-exposed rats, mice, guinea pigs and rabbits.

Several studies have also evaluated the potential role of DBP exposure on LH and FSH production and levels, but the available results are highly variable. This represents a data gap in the proposed MOA. Future studies need to address the potential impact of DBP exposure on hypothalamic and pituitary functions.

Conclusions

Analysis of the mechanistic and toxicological evidence on DBP-induced male reproductive effects suggests that:

• Fetal rats are more sensitive than other mammalian vertebrates (including human fetal testis xenografts and ex vivo tissue cultures) to DBP-induced disruption in LC functions and effects that occur downstream of decreases in testosterone levels.

• However, adverse responses in fetal rat testis which are not regulated by testosterone (i.e. seminiferous cord effects) are conserved across species, including human fetal testis xenografts and ex vivo tissue cultures.

• Both androgen-dependent and -independent outcomes are conserved across most mammalian species when exposures occur during postnatal life stages.

Use of the AOP framework served as a practical tool for the evaluation of mechanistic and toxicological evidence generated from a variety of evidence sources including whole animal, cell culture, and human ex vivo and xenograft experimental models. In the analysis presented here, application of the AOP framework and inclusion of life stage considerations facilitated a transparent analysis for consistency and concordance. Development of a database/inventory also simplified the navigation and analysis of a large and complex number of studies.