Impact of endocrine disruptors on neurons expressing GnRH or kisspeptin and pituitary gonadotropins

Abstract

Reproduction is a complex process that is controlled centrally via a network of hypothalamic neurons to modulate the pulsatile release of gonadotropin-releasing hormone (GnRH) and subsequently pituitary gonadotropins. The gonadotropins, luteinizing hormone and follicle stimulating hormone, drive gametogenesis and hormone production from the gonads. The hypothalamic-pituitary exchange is controlled by gonadal steroids through negative and positive feedback mechanisms via steroid receptors. Due to the expression of these receptors, GnRH neurons, the hypothalamic neurons that control them, and pituitary gonadotropes are sensitive to exogenous compounds that interact with steroid and nuclear receptors or alter hormone production and metabolism. The compounds, called endocrine disrupting compounds (EDCs), are ubiquitous and persistent in human environments and could bioaccumulate in the body. EDCs include plasticizers (like bisphenol A), dioxin, polychlorinated biphenyls (PCBs), organochlorine pesticides, flame retardants, and perfluorinated akyl substances (PFAS). Numerous studies have reported that perinatal, juvenile, or adult exposure to these EDCs, primarily in rats, disrupt the hypothalamic control of pituitary gonadotropin production leading to disruption of gonadal steroid production and estrous cyclicity. The purpose of this review is to evaluate these studies primarily focusing on GnRH and kisspeptin neurons and anterior pituitary gonadotropins and to discuss the need for deeper investigations into the hypothalamic-pituitary-gonadal axis.

Introduction

Reproduction is governed by the hypothalamic-pituitary-gonadal (HPG) axis, as we have previously reviewed (Acevedo-Rodriguez et al., 2018). The execution of reproduction across most vertebrate species is dependent on hypothalamic neuropeptide/neurotransmitter production and release under the control of peripheral hormonal signals to regulate gametogenesis and mating. Gonadotropin-releasing hormone (GnRH) is the primary hypothalamic neuropeptide secreted in pulses that controls the release of pituitary gonadotropins, luteinizing hormone (LH) and follicle stimulating hormone (FSH), from the anterior pituitary gonadotropes (Acevedo-Rodriguez et al., 2018). The gonadotropins, in turn, stimulate the gonads to produce steroids i.e., testosterone (T), estrogens (primarily 17β-estradiol, E2), and progesterone (P4), and initiate gamete production. The gonadal steroids act as positive (only E2) and negative regulators of GnRH and gonadotropins by modulating neurons, including kisspeptin neurons, upstream of the HPG axis (Acevedo-Rodriguez et al., 2018), although neuro-progesterone produced by hypothalamic astrocytes has been implicated in the LH surge (Sinchak et al., 2020). Although many of the key pathways involved in the regulation of GnRH release are known, three neuronal populations, GnRH neurons and two kisspeptin neuron populations, act as an interconnected network to control the secretion of GnRH in pulses to temporally control the release of gonadotropins and the onset of puberty and the LH surge. Kisspeptin neurons, found in the anteroventral periventricular (AVPV) nucleus and the arcuate nucleus (ARC), have potent stimulatory actions on GnRH secretion and pulsatility and are also responsive to metabolic and physiological stressors to suppress reproduction (Acevedo-Rodriguez et al., 2018). AVPV kisspeptin neurons in rodents control the proestrous LH surge in females and contribute to the basal secretion of GnRH. ARC Kisspeptin neurons co-express Neurokinin B and Dynorphin (called KNDy neurons) and are the GnRH pulse generator during most of the estrous cycle producing the tonic release of GnRH and LH. All of these neurons and the pituitary gonadotropes express at least one subtype of estrogen receptors (ER) and/or androgen receptors (AR). As such, each level of the HPG axis is tightly regulated and modulated by hormonal and central signals to influence reproductive status (See Figure 1 for illustration).

Figure 1 – Modulation of GnRH and kisspeptin neurons by steroids controls gonadotropin production in the rodent.

(1) the gonadal steroids, 17β-estradiol (E2), progesterone (P4), and testosterone (T), control the HPG axis through negative feedback onto hypothalamic neurons and pituitary gonadotropes. During negative feedback when progesterone levels are also elevated, E2 exerts a negative influence on GnRH (blue) pulsatility. Testosterone exerts negative feedback onto kisspeptin neurons and gonadotropes. GnRH is released from the nerve terminals in the median eminence into the primary porta plexus to activate the production of luteinizing hormone (LH) and follicle stimulating hormone (FSH). (2) GnRH pulsatility is controlled primarily through arcuate KNDy (purple) neurons. Kisspeptin is released onto GnRH terminals in the median eminence. The release of kisspeptin is controlled, in part, by the positive influence of Neurokinin B (NKB) via Tac3 receptors expressed on neighboring KNDy neurons and by the negative influence of Dynorphin release onto neighboring KNDy neurons or unidentified arcuate neurons that suppress kisspeptin release from KNDy neurons. (3) During positive feedback, elevated E2 plasma concentrations augment AVPV Kiss1 (red) neuronal activity and kisspeptin gene expression to produce the surge of GnRH and subsequently LH. Green circles (+) denote positive feedback or neuronal activation. Red circles (−) denote negative feedback or neuronal suppression.

These hypothalamic neurons and pituitary cells are sensitive to endocrine influences during fetal, neonatal, and juvenile development and are controlled by the same hormone signals in the adult (Tena-Sempere, 2010; García-Galiano et al., 2012). Thus, they are important targets for anthropogenic compounds known to interact with steroid (ER, AR) and nuclear receptors (aryl hydrocarbon receptor (AhR)) expressed by hypothalamic neurons and pituitary gonadotropes. These compounds, called endocrine disrupting compounds (EDCs), exert their disruptive effects through multiple mechanisms as reviewed earlier (Waye and Trudeau, 2011; La Merrill et al., 2019), which include: 1) acting as agonists and antagonists of hormone receptors, 2) altering hormone receptor expression and hormone signaling, and 3) disrupting hormone production, distribution, transport, and clearance. Exposure to common EDCs occurs throughout the lifespan of humans and wild-life as many EDCs are environmentally persistent and accumulate in lipid-rich tissues in the body (León-Olea et al., 2014; La Merrill et al., 2019).

One of the most well-known and well-studied EDCs is bisphenol A (BPA). Previous reviews have examined the evidence that this common EDC is a potential threat to reproduction through its interactions with GnRH, kisspeptin, and the pituitary (Tena-Sempere, 2010; Huo et al., 2015; Santoro et al., 2019; Pivonello et al., 2020). Therefore, we will not be reviewing BPA but will focus on other well-known or emerging, but understudied, EDCs – dioxin and PCBs, organochlorine pesticides (DDT and methoxychlor (MXC)), perfluorinated alkyl substances (PFAS), and flame retardants. While many studies examining the influence of EDCs on the HPG axis have focused on the regulation of steroidogenic enzymes in the gonads and the production of steroid hormones (Waye and Trudeau, 2011; La Merrill et al., 2019), our review will focus solely on influences of these EDCs on hypothalamic GnRH and kisspeptin expression, GnRH pulsatility, and pituitary gonadotropin production from the pituitary and summarize their impact on the HPG axis. We will focus our discussion on rats and mice, exposed perinatally or as juveniles and adults and briefly review exposure to hypothalamic or pituitary cell cultures for a comprehensive, yet not exhaustive, assessment of their influence on the HPG axis. We will also attempt to relate these findings to human physiology and to highlight areas that demand further investigation.

Endocrine Disrupting Compounds

Humans are chronically exposed to dioxins, namely 2,3,7,8-tetrachlorodibenzodioxin (TCDD) and polychlorinated biphenyls (PCBs), which are a class of ~200 related compounds. Only co-planar PCBs act through similar mechanisms as dioxin, which is a potent agonist for the aryl hydrocarbon receptor (AhR) (Pessah et al., 2010; Larsson et al., 2015). Dioxins and PCBs are highly persistent in the environment and are known to bioaccumulate in animal and human adipose tissue leading to biomagnification in predators and human food sources (Schecter et al., 2006; McIntyre and Beauchamp, 2007). Dioxins can occur naturally through combustion whereas anthropogenic sources occur through industrial manufacturing of paper, pesticides, herbicides, and electronics, and through waste incineration leading to exposures greater than the tolerable weekly intake of 2 pg toxic equivalents (TEQ)/kg bw/week (Gilpin et al., 2003; Fernandes and Falandysz, 2021). While PCBs have been banned in the USA since 1979 and globally since 2001 (White and Birnbaum, 2009), human exposure persists in the 1–300 pg/g lipid range due to the leaching of PCBs from old construction materials and hazardous waste sites (Consonni et al., 2012; Koh et al., 2015) with a no-observed-adverse-affect level (NOAEL) of 6–9 mg/kg/d (Graceli et al., 2020). Although their main mechanism of toxicity is through AhR activation, expressed in the hypothalamus and pituitary gonadotropes (Cao et al., 2011), many PCBs and dioxins can interact with steroid receptors or modulate their expression to exert EDC-like effects (Dickerson and Gore, 2007; Akinola et al., 2020). Thus, the HPG axis is a target for these compounds as has been previously reviewed (Dickerson and Gore, 2007). Our review focuses specifically on the kisspeptin-GnRH neural circuit and the pituitary and will discuss TCDD and PCBs.

Another class of chemicals that are classified as EDCs are organochlorine pesticides. Widely used since the middle of the 20^th century, this class of pesticides includes DDT (1,1,1-trichloro-2,2-bis[4-chlorophenyl]ethane) and methoxychlor [1,1,1-trichloro-2,2-bis(p-methoxyphenyl)ethane, MXC]. The metabolites of DDT and MXC are well-known EDCs (Tiemann, 2008; Gore et al., 2011). Organochlorine pesticides and their metabolites possess estrogenic properties; resulting in adverse effects on the reproductive system in animal models (Cummings, 1997; Martyniuk et al., 2020). In particular, DDT and its primary metabolites (DDE and DDD) are ubiquitous, persistent in the environment, bioaccumulate in lipid-rich organs, and are detected in human samples (plasma 100–300 ng/g lipid) across the globe (Turusov et al., 2002; Rodríguez-Alcalá et al., 2015) with a NOAEL of 0.05 mg/kg/d and total daily intake (TDI) of 0.5 ng/kg/d (Graceli et al., 2020). Although DDT has not been in wide use in the USA and Canada since the early 1970s, exposures persist due to global transportation (Wu et al., 2020). MXC was developed as a replacement insecticide for DTT, but ultimately became obsolete and is no longer used. MXC has little endocrine activity and a low affinity for ERs. However, its primary metabolite HPTE (2,2,-bis-(p-hydroxyphenyl)-1,1,1-trichloroethane) has a high affinity for ERs and thus has become a model estrogenic EDC (Cummings, 1997). The influence of DTT and MXC on female fertility, ovarian function, and implantation in mammals was previously reviewed (Tiemann, 2008). Thus, we are focusing on the hypothalamic and pituitary impacts of these EDCs on the reproductive axis in rodents and cell cultures.

PFAS are synthetic chemicals that are insoluble in water and solvents, while also repelling water and oils. Thus, PFAS are used for both household and industrial applications from shampoos to cookware and insecticides to fire-retardant foams. Ingestion via food, water, or from food packaging and preparation is the main route of exposure in human populations (Domingo and Nadal, 2017). Common PFAS, perfluorooctanoic acid (PFOA) and perfluorooctane sulfonate (PFOS), are detected in human samples (blood, urine) from infants, children, and pregnant and nursing mothers at 1–25 ng/ml and are environmentally persistent, ubiquitous, and bioaccumulate in humans and wildlife (Ballesteros et al., 2017; Liu et al., 2018). PFAS have disruptive properties to the hypothalamic-pituitary-thyroid axis that play an important neurodevelopmental role in utero and neonatal neuromaturation through to adolescence and adulthood (Coperchini et al., 2017). In regards to reproduction, data from maternal PFOA exposure studies in rodents suggest that PFOA disrupts lactational efficiency in the dams leading to offspring mortality, delayed mammary development, and altered mammary gene expression (Patisaul et al., 2018). Furthermore, peripubertal PFOA exposure in female mice disrupts estrous cyclicity and genes involved in ovarian steroidogenesis in a strain-dependent manner (Zhao et al., 2012). It is only within the past decade that more attention has been focused on the effects of PFAS on the HPG axis in mammals.

Flame retardants are compounds added or adhered to furniture, clothing, toys, electronics, plastics, and other household and workplace furnishings to reduce combustibility to meet government fire safety regulations. Flame retardants are often volatile and are not chemically bound to the fabric, plastic, or materials. Flame retardants are typically found in house dust, food, and water and in the air of offices, airplanes, and other workplace environments (Xue et al., 2007; Hoffman et al., 2017; Wong and Durrani, 2017; Gravel et al., 2020; Percy et al., 2020). There are three main groups: polybrominated diphenyl ethers (PBDEs), novel brominated compounds, and organophosphate flame retardants (OPFRs). PBDEs have been largely phased out of use since the mid-2000s due to their bioaccumulation in humans and wildlife, and their neurological and endocrine toxicity (Costa and Giordano, 2007; van der Veen and de Boer, 2012; Hu et al., 2014; Abbasi et al., 2015). PBDEs have been replaced by novel brominated compounds and OPFRs (van der Veen and de Boer, 2012), which are not as persistent and bioaccumulate less than PBDEs (Yang et al., 2019). PBDEs were detected in human samples (placental and breast milk), 5–10 years after the phase-out, at concentrations ranging from 1–60 ng/g lipid (Adgent et al., 2014; Leonetti et al., 2016; Eskenazi et al., 2017) and standard safety levels set at 0.1–7 μg/kg/d (Graceli et al., 2020). Body burdens of OPFRs and their metabolites are increasing in human populations, with some metabolites found at levels 15 times higher in 2014–15 than in 2002–03 (Hoffman et al., 2017, 2018; Blum et al., 2019). For example, OPFRs are found in household dust at median concentrations of ~1–6 μg/g dust (triphenyl phosphate (TPP), tricresyl phosphate (TCP), and tris(1,3-dichloro-2-propyl)phosphate (TDCPP)), which are inadvertently ingested by a pregnant mother or a growing infant (Carignan et al., 2013; Meeker et al., 2013; Fan et al., 2014). Consequently, OPFR metabolites are detectable in humans during the perinatal period (Hoffman et al., 2015) in urine, breast milk, and blood samples ranging in concentrations from 1–100 ng/mL and 1–100 ng/g lipids (Butt et al 2014; Dishaw et al 2014; Hoffman et al., 2017, 2014; Kim et al., 2014; Sundkvist et al 2010). While OPFR exposures are associated with increased risk for high body mass index (BMI) in children (Boyle et al., 2019) and adverse neurological outcomes (Castorina et al., 2017; Doherty et al., 2019), little is known about their influence on the hypothalamus or pituitary. Summary Tables (Tables 1–4) are provided for each group of EDCs.

Table 1.

Summary of effects of dioxin on GnRH, kisspeptin, and gonadotropins

Source	Exposure	Dose	Model	GnRH and pulse	Kisspeptin	Gonadotropins
Kakeyama et al. (2008)	perinatal	200 or 800 ng/kg	Rat
Mutoh et al. (2006)	perinatal	0.01, 0.1, 1 μg/kg	Rat	NC		F, M: ↓LHβ, serum LH
Shi et al. (2007)	perinatal	1–200 ng/kg/week	Rat			F: ↑LH 1 h post-GnRH
Takeda et al. (2014)	perinatal	1 μg/kg/2 ml vehicle	Rat	M: ↓		M: ↓LH
Clements et al. (2009)	perinatal	5 μg/kg	Rat	M: ↓
Takeda et al. (2011)	perinatal	1 or 10 nM	Rat: PC			↓Lhβ, αGsu
Li et al. (1997)	juvenile	10 μg/kg	Rat			F: ↑LH and FSH
Gao et al. (2000)	juvenile	32 μg/kg	Rat			F: ↓LH and FSH
Petroff et al. (2003)	juvenile	8 or 32 μg/kg	Rat			F: ↑FSH and LH
Bookstaff et al. (1990)	adult	100 μg/kg	Rat	M:↓Gnrhr		M: ↓LH
Cao et al. (2011)	adult	0.5 and 10 μg/kg	rat			F: ↑Lhβ
Yin et al. (2012)	adult	100 ng/kg/d	Mouse	M:↓		M: ↓testicular LH, FSH
Trewin et al. (2007)	adult	3.2 nM	Rat explants	NC		NC
Huang et al. (2011)	transgeneration	0 and 2 ng/L	Mouse	NC		NC
Petroff et al. (2003)	cell	1–100 nM	GT1–7	NC
Solak et al. (2013)	cell	10 nM	Rat: GnV-3	↑↓Gnrh
Mueller & Heger (2014)	cell	10 mM	GripTite™ 293 MSR cells transfected with the Kiss1 promoter	↓Gnrh	↓Kiss1

NC, measured but no change; PC, pituitary cells; ↑, increase in expression of gene or protein; ↓, decrease in gene or protein; ↑↓, mixed effects; F, female; M, male.

Table 4.

Summary of effects of PFAS on GnRH, kisspeptin, and gonadotropins

EDC	Source	Exposure	Dose, mg/kg	Model	GnRH and Pulse	Kisspeptin	Gonadotropins
PFOA	Du et al. (2019)	perinatal (PND: 1–5)	0.1, 1,10	Rat		F: ↓AVPV; ↓ARC Kiss1, Kiss1r	F: ↑LH serum in high dose
PFOA	Du et al. (2019)	Juvenile (PND: 26–30)	0.1, 1, 10	Rat		F: ↑AVPV low dose; ↓AVPV high dose	F: ↑LH in high dose
PFOS	Lopez-Doval et al. (2014)	young adult	0.5, 1.0, 3.0, 6.0	Rat	M: ↓at 3 low doses		M: ↑LH, FSH at 2 low doses M: ↓LH at 0.5 and 2.0 kg M: ↑FSH all doses
PFOS	Lopez-Doval et al. (2016)	adult	1.0, 3.0, 6.0	Rat	M: NC: Gnrh		M: NC - Gnrhr M: ↓GnRH receptor protein
PFOS	Feng et al. (2015)	adult	0.1	Mouse	F: ↑in diestrus F: ↓in proestrus	F: ↓AVPV Kiss+ neurons F: ↓AVPV	F: ↓LH, FSH proestrus F:↑ LH diestrus
PFAS	Zhang et al. (2020)	adult	0.5, 2, 5	Mouse	↓at 2 and 5 mg/kg		↓LH at 2 and 5 mg/kg

NC: measured but no change. ↑, increase in expression of gene or protein. ↓, decrease in gene or protein. F = female. M = male.

Perinatal exposures

The perinatal period, which encompasses gestation and lactation periods and fetal, neonatal, and pre-weaning juvenile developmental windows, is especially sensitive to the perturbations from EDC exposure. The primary animal model used in perinatal EDC exposures is the rat, although studies using mice are now common. These animal models are used to assess the influence of EDCs on maternal programming and early juvenile development, which correlates to similarly sensitive windows in human development. In regards to dioxins, the prenatal (fetal) stage is highly sensitive. For example, a single dose of TCDD (0.800 μg/kg) at GD15 in Long-Evans hooded rats advanced the onset of puberty (vaginal opening) and first estrus day while increasing ovarian weights (Kakeyama et al., 2008). A single dose of TCDD (1 μg/kg) at GD15 did not alter GnRH expression but reduced fetal (GD20) pituitary expression of LHβ and serum LH concentrations. The authors provide evidence that the decrease in serum T is due to a reduction in steroidogenic enzymes e.g., steroidogenic acute regulatory protein (StAR), due to the decrease in LH production by rescuing StAR expression in the fetal testes from TCDD exposure via injection of equine chorionic gonadotropin (eCG) into the fetus at GD15 (Mutoh et al., 2006). In a study of hypothalamic explants from maternally-exposed Sprague-Dawley male offspring (PND35), a single dose of TCDD (5 μg/kg, GD15) suppressed KCl-evoked release of GnRH to 40% of controls (Clements et al., 2009). Suppression of GnRH was also observed at PND60 and 90 indicating that TCDD has long-lasting effects beyond puberty on GnRH release. Not only was GnRH release altered by TCDD exposure, but hypothalamic GnRH content and GnRH somal profiles (smooth, lightly thorny (spines), heavily thorny) were altered in the lateral POA with more smooth-type and fewer thorny-type GnRH neurons. Smoother cells suggest fewer presynaptic inputs and potentially less excitatory inputs onto GnRH neurons. These data suggest that single exposure to TCDD during fetal development can have long-lasting effects on reproductive parameters of the offspring, albeit at doses higher than typical human exposure.

Perinatal TCDD exposure can also induce long-lasting disturbances to reproduction. TDCC exposure (50 or 200 ng/kg/week by gavage) every week for four weeks across the perinatal period (GD14, GD21, PND7, PND14) delayed the onset of puberty and suppressed E2 production throughout the cycle (Shi et al., 2007). In ovariectomized females from this experiment, TCDD exposure at 200 ng/kg/week potentiated the induction of LH by GnRH 1 h post-injection with no change in FSH production. These doses were designed to be within the range of human exposure, whether across the reproductive lifespan or in high-risk populations occupational exposures. In another perinatal study supporting the long-term effects of TCDD exposure on adult GnRH activity, pregnant Wistar dams were orally dosed with TCDD (1 μg/kg per 2 ml of corn oil) at GD15 followed by sample collection at GD18, 19, 20, and 21 and PND0, 2, 4, 7, 14, 28, 56, and 70 (Takeda et al., 2014). TCDD exposure reduced hypothalamic GnRH expression in juvenile and young adult male offspring and reduced GnRH protein content in the hypothalamus, cerebrum, and cerebellum at PND70. However, GnRH mRNA was reduced in the hypothalamus across all postnatal ages, but only in the cerebrum or cerebellum at PND70. TCDD exposure also attenuated the fetal and neonatal expression of pituitary LH. Furthermore, adult (PND70) male offspring exposed to TCDD during gestation exhibited reduced mounting and intromissions when administered GnRH. Although testosterone was not measured, these effects on sex behavior may be due to steroidogenesis during development as steroidogenic acute regulatory protein (StAR) and 17α-hydroxylase (Cyp17) expression was reduced in fetuses and neonates suggesting that the organizational effects of TCDD endocrine disruption in the brain underlie the deficits in sex behavior. In a transgenerational mouse study, TCDD was administered via drinking water at 2 ng/L (2 ppt) to C57 BL/6J females starting at weaning (PND21). Dams were later mated and their offspring were studied over the next two generations (Huang et al., 2011). TCDD exposure reduced E2 levels in young adult females in the P0 and P1 generations and this was associated with a higher incidence of miscarriage and still-birth with no change in P4. These parameters are an indirect measurement of HPG activity/reproductive toxicity suggesting disruption of the HPG axis by TCDD at concentrations not exceeding exposure limits to humans.

PCBs are another EDC that can interact with both AhRs and steroid receptors as either parent compound or their metabolites. Only a few studies illustrating the potential disruption to GnRH, kisspeptin, and gonadotropin production due to perinatal PCB exposure exist. In two similar studies from the same laboratory, exposure of pregnant Sprague-Dawley dams to an estrogenic positive control (estradiol benzoate (EB), 50 μg/kg), Aroclor 1221 (A1221, 1 mg/kg), or a mixture of nonplanar PCBs with the highest body burden in humans (PCB 138, 153, 180, at 1 mg/kg) on GD16 and 18 increased anogenital distance (AGD) in male pups (PND1) and reduced AGD in female pups (Dickerson et al., 2011a). A1221 increased the number of apoptotic nuclei in the female AVPV, but not the medial preoptic area (MPOA). When POA gene expression from PND1 pups was examined by low-density PCR arrays, Kiss1 and Gnrh expression were not affected by the EDCs but kisspeptin receptor was reduced in males, yet increased in female pups (Dickerson et al., 2011a). In the second study using the same EDCs and experimental design but focusing on pups, juveniles, and young adult offspring, exposure to A1221, the PCB mixture, or EB increased anogenital distance in male pups from PND1 to 21, while only A1221 increased body weights in juvenile and adult female offspring (Dickerson et al., 2011b). All three treatments induced early onset of puberty in female offspring but delayed puberty (preputial separation) in male pups. These exposures reduced the number of ERα-expressing AVPV neurons (potentially kisspeptin) and the volume of the AVPV in adult female offspring. A1221 and PCBs also reduced AVPV kisspeptin expression and the number of GnRH neurons expressing c-fos, an early immediate gene indicating neuronal activation, on the afternoon of proestrus. The authors also reported main effects of PCBs on serum T and main effects of A1221 and PCBs on serum P4 in males (Dickerson et al., 2011b). In a multigenerational perinatal study, pregnant Sprague-Dawley Rats were dosed with A1221 (0.1, 1, or 10 mg/kg) on GD16 and 18 (Steinberg et al., 2008). In the F1 generation, A1221 treatment did not alter serum P4 and E2 but increased serum LH concentrations. In the F2 generation, all doses of A1221 reduced P4 during proestrus, and the high dose reduced P4 during diestrus. The high dose reduced F2 uterine weights during estrus and LH during proestrus. These three studies strongly suggest that disruption of GnRH and kisspeptin or its receptor by PCBs in the hypothalamus may lead to long-term disruption of the HPG axis that can extend through generations of female rats.

Epidemiological studies have demonstrated associations between DDT or its metabolites (DDE, DDD, etc.) and preterm and undersized neonates with a range of DDE from 3–178 ng/ml (Longnecker et al., 2001). While there are numerous studies on DDT and reproduction in rodent models, there are only a few that measure the activity of the HPG axis after perinatal DTT exposure. In a maternal exposure study using Wistar rat dams, DDE exposure (via food, 125 ppm, from GD1- PND21) did not alter the onset of puberty, estrous cyclicity, or serum levels of FSH and LH in young female offspring (Makita, 2008). In a postnatal exposure model, young female rats were exposed to the o,p´-DDT isomer or its metabolite, p,p´-DDE, during postnatal development (PND6 onward), followed by measurements of in vitro GnRH secretion and the onset of puberty and gonadotropin secretion (Rasier et al., 2007). In hypothalamic explants from juvenile females (PND15), o,p´-DDT and p,p´-DDT (10⁻⁶ to 10⁻⁴ M) reduced GnRH interpulse intervals (increased pulsatility) after 3–4 h in a concentration- and time-dependent manner with no effect on amplitude. In an in vivo experiment, female rats received daily subcutaneous injections of o,p´-DDT for 5, 10, or 35 days (10 or 100 mg/kg/d for PND6–10 and 10 mg/kg/d for PND6–15 and PND6–40). In PND15 female rats, serum LH concentrations were reduced, potentially through a negative feedback effect directly on gonadotropes during the pre-pubertal period, as the authors suggest. However, GnRH pulsatility was increased (lower interpulse intervals) after 10 and 100 mg/kg o,p´-DDT, which correlated with early onset of puberty (vaginal opening) and first estrus. In vitro exposure of hypothalamic explants collected from PND15 female pups to o,p´-DDT and p,p´-DDT (10⁻⁵-10⁻⁴ M) potentiated glutamate-evoked GnRH secretion by increasing the amplitude of pulses (Rasier et al., 2008). In vivo exposure to o,p´-DDT (10 mg/kg/d, PND6–10) accelerated pulsatile GnRH secretion and induced early onset of puberty in juvenile female rats. Antagonists to both ERα and AhR blocked the effects of o,p´-DDT on GnRH pulsatility. These studies are unique in that they analyze the effect of short-term exposure to a high dose estrogenic EDC on GnRH pulsatility, on both amplitude and frequency. These data also indicate transient exposure to doses of DDT higher than body burdens in pregnant women (Longnecker et al., 2001) alters GnRH pulsatility leading to early sexual maturation.

While MXC is no longer in use and human exposures are minimal, this EDC, and its strongly estrogenic metabolite HPTE (Hewitt and Korach, 2011), are useful models for reproductive disruption by estrogenic EDCs. Perinatal exposures to MXC in rodents disrupt reproductive parameters (puberty) and hypothalamic gene expression and pituitary gonadotropin production. In Fischer rats, aged female offspring exposed to daily injections of MXC (20 μg/kg or 100 mg/kg) from GD19 to PND7 through the dams exhibited reduced estrous cyclicity by 14 months of age compared to DMSO control, and the high dose of MXC induced complete anestrus by 17 months while all DMSO controls were still cycling (Gore et al., 2011). These effects on cyclicity were associated with lower serum E2 and elevated Esr1 (ERα) and Gnrhr expression in the POA, with no effect on POA Kiss1. In a similar perinatal exposure study of Fischer rats, female adult offspring from dams exposed to MXC (75 mg/kg/d from GD11 to PND7) exhibited early onset of puberty (vaginal opening), but no effect on Kiss1 expression in the arcuate nucleus (Roepke et al., 2016). In another maternal exposure study, MXC (1200 ppm) was introduced via a special diet from GD15 to PND10 in an unidentified rat strain. The MXC-containing diet reduced the percentage of LH⁺, FSH⁺, and prolactin⁺ cells in male juvenile offspring (PND21) and reduced LH⁺ cells in juvenile females (PND21). In the young adult offspring, MXC exposure increased the percentage of FSH⁺ and prolactin⁺ cells only in female offspring to 134% of control (Masutomi et al., 2004). In another perinatal exposure model using Wistar-Imamichi rats, MXC exposure via diet (24, 240, 1200 ppm, GD15 - PND10) decreased serum FSH and LH, but not T, after maturation in male offspring (Suzuki et al., 2004). In female offspring, the lordosis quotient was reduced by all doses of MXC which was associated with reduced proestrous LH surge concentrations in serum. Collectively, these data suggest that high doses of MXC during the perinatal period disrupt the HPG axis in both female and male rats.

While at doses higher than recently reported PFAS concentrations in human samples (Ballesteros et al., 2017; Liu et al., 2018), PFAS exposure appears to specifically target the steroid-sensitive kisspeptin neurons in juveniles that lead to changes in sexual maturation. In rats, the response to PFAS is dependent on the duration and dose of exposure, the developmental window of exposure, and the age and sex of the animal. In a recent study using neonatal and juvenile female Sprague-Dawley rats, subcutaneous injection of PFOA (0.1, 1, and 10 mg/kg) or PFOS (0.1, 1, and 10 mg/kg) during PND1–5 suppressed Kiss1 and Kiss1r mRNA expression in the AVPV and the ARC in the postpubertal adult rats compared to controls (Du et al., 2019). Conversely, in the juvenile exposure on PND 26–30 to a low dose of PFOA (0.1 mg/kg) increased AVPV Kiss1 expression whereas a high dose (10 mg/kg) decreased AVPV Kiss1. These effects on kisspeptin expression were associated with elevated kisspeptin fiber intensities at the low doses and suppressed kisspeptin fiber intensities at the high doses in the AVPV and ARC. Physiologically, these central effects correlated with elevated LH serum concentrations in neonates and an advancement of puberty onset (vaginal opening) and first estrus at 10 mg/kg for both neonatal and juvenile exposures.

There are few studies that examine the effects of flame retardants on GnRH, kisspeptin, and pituitary production of gonadotropins in rat perinatal or adult exposures were reviewed. In one study on the maternal exposure (GD10–18) of rats to PBDE 99 (1 and 10 mg/kg/d), more exposed adult female offspring exhibited irregular estrous cycles than controls (Faass et al., 2013). While GnRH, kisspeptin, or gonadotropins were not assessed, PBDE 99 exposure increased ERα expression in the MPOA and the ventromedial hypothalamus (VMH) while suppressing VMH ERβ and progesterone receptor (PR). As these two regions are important for sex behavior, these results indicate that suggesting that PBDEs could disrupt the hormonal control of sexual motivation and lordosis. In male rats, VMH PR mRNA induction in response to E2 was reduced by 1 mg PBDE 99, but increased by 10 mg (Faass et al., 2013). In a recent mouse study from our lab, maternal exposure to a mixture of OPFRs (TPP, TCP, TDCPP; 1 mg/kg/d from GD7 to PND14) induced transcriptional changes to several reproductive hypothalamic genes in neonatal (PND0) and juvenile (PND14) WT C57 pups (Adams et al., 2020). In female juvenile pups, the expression of Pdyn, Tac2, and Esr1 was elevated by maternal exposure to OPFRs. These changes are associated with an increase in body weight at PND14 and a decrease in anogenital distance in male pups at PND7, indicating an estrogenic effect of OPFR exposure or inhibition of the neonatal testosterone surge. Collectively, these few studies suggest that developmental exposures to PBDEs or OPFRs can disrupt hypothalamic expression of neuropeptides and steroid receptors. These findings warrant further investigation with particular attention on the mechanism of toxicity and the subsequent effects in pre- and post-pubertal rodents.

Adult Exposures

Exposures to EDCs are continuous and occur across the lifespan of an individual. However, most adult studies of EDC on the HPG axis in rodents do not use continuous lifelong exposures but focus on single dose or sub-chronic exposures to doses higher than human exposures. For example, exposures to TCDD at high doses influence hypothalamic and pituitary reproductive endpoints in young and adult rats. In a 1990 study, a single high dose of TCDD (100 μg/kg, oral) given to adult Sprague-Dawley male rats reduced plasma T, prevented the compensatory increase in plasma LH, and blocked the increase in GnRH receptors in the pituitary (Bookstaff et al., 1990). However, in castrated male rats, TCDD did not affect LH or GnRH plasma concentration or pituitary GnRH receptor expression suggesting that TCDD potentiates the negative feedback of androgens on the hypothalamus and pituitary. In immature (PND22–23), female Sprague–Dawley rats, TCDD injection can alter gonadotropin production in a dose- and time-dependent manner. A single dose of TCDD (10 μg/kg) by gastric intubation elicited a distinct increase in both LH and FSH serum concentration about 24 hours after dosing that was not observed in the controls (Li et al., 1997). FSH production was more sensitive to TCDD as FSH was induced at 0.3, 1, 3, 10, and 30 μg/kg while LH was induced at 3, 10, and 30 μg/kg. Collectively, these data suggest that a single high dose TCDD is estrogenic in potentiating the release of gonadotropins in female and by reducing androgen production in male rats. In another study using immature female rats, TCDD (32 μg/kg) injected alone reduced LH and FSH production 58 h after an injection of equine chorion gonadotropin (eCG) (which induces folliculogenesis) and significantly reduced the number of ova collected in the oviduct compared to control. However, TCCD injection did not alter the induction of LH or FSH when administered with GnRH suggesting that the disruption to ovulation is a direct effect on the ovary (Gao et al., 2000). In a third study using immature female Sprague-Dawley rats, TCDD (8 or 32 μg/kg orally) increased FSH and LH serum levels prematurely after eCG injection (Petroff et al., 2003). Furthermore, TCDD (32 μg/kg) increased serum LH and FSH compared to control when estradiol cypionate was also administered. In adult ovariectomized female Sprague-Dawley rats, TCDD at 0.5 and 10 μg/kg increased pituitary Lhb expression, which was suppressed by E2 (Cao et al., 2011). In an adult subchronic exposure model using an unidentified mouse model, oral TCDD exposure at a low dose (100 ng/kg/d for 7 weeks) reduced hypothalamic GnRH protein expression and testicular LH and FSH content in males (Yin et al., 2012). The disruptive effects of subchronic TCDD exposure were partially ameliorated by vitamin E supplementation by reducing oxidative stress and expression of AhR target genes (Hamm et al., 2000; Xu et al., 2008). These data indicate that TCDD may have age- and sex-dependent effects on GnRH actions in the pituitary and on gonadotropin production, as well as disruption to gonadal function through direct and indirect pathways. Thus, TCDD has the capability of disrupting the mouse HPG axis both at single high-dose exposure or subchronic low-dose exposures, although more investigations are needed.

In experiments focusing on adult PCB, DDT, and MXC exposures, both gonadotropin and steroid production are especially sensitive. In adult male Sprague Dawley rats, PCB 126 injections (6.25, 25, 100, or 400 μg/kg/d for 2 days) reduced serum LH but increased pituitary LH and FSH, indicating that PCB 126 suppresses the release of LH, driven by GnRH, but not to the same degree as synthesis (Desaulniers et al., 1999). In another acute PCB exposure study in young adult male Sprague Dawley rats, weekly injections of PCB126 (0.2 mg/kg) or PCB114 (20 mg/kg) for 2, 5, or 8 weeks altered serum T and LH (Han et al., 2010). In the 2-week PCB 126 and PCB 114 groups, serum T was suppressed and in the 5-week PCB126 group, serum T was also lower than controls, due to a reduction in the expression of steroidogenic enzymes in the testes. Consequently, serum LH in the 5-week PCB114 group was elevated and serum FSH in the 2- and 5-week PCB126 groups were lower compared to controls. Potentially many of the effects observed in PCB exposure are exerted in the gonads leading to suppression of the negative feedback signal (T) that in turn increases LH production. These effects may also be dependent on the sex or the strain of the rat. For example, in female Wistar rats subcutaneously injected with either PCB 180 or A1221 (10 mg/kg for 3 days) serum FSH was suppressed but not serum LH (Uslu et al., 2013). Exposure to DDT (50 and 100 mg/kg, 10 days) in adult male rats severely diminished serum T in a dose dependent manner (Ben Rhouma et al., 2001). Subsequently, FSH levels were elevated at both doses and LH levels at 100 mg/kg due to the lack of negative feedback from testosterone. In an adult exposure paradigm in Sprague-Dawley rats, MXC (subcutaneous injection, 25 mg/kg/d, 30 days) decreased plasma LH and T as compared to control with no effect on FSH (Lafuente et al., 2008). In ERα-overexpressing mice (tet-op-tTA/tet-op-luciferase mice/tet-op-ERα), MXC (i.p. injection, 64 mg/kg/d, 20 days) increased time spent in estrus in both littermate controls and ERα-overexpressing mice, and overall showed reduced serum FSH levels, with no effect on E2 although greater atresia of ovarian follicles was observed in the ERα-overexpressing mice (Tomic et al., 2006). This last study is one of the few studies that investigated a hormone receptor-mediated mechanism of HPG axis disruption in a rodent model. Collectively, these adult exposure studies to EDCs illustrate the importance of examining the entire HPG axis to determine the site of endocrine disruption as disruption to steroidogenesis in the gonads produces effects at the level of hypothalamus and/or pituitary hormone production.

An emerging threat to reproductive health is the pervasive EDC class – PFAS. In male Sprague-Dawley rats, subchronic oral exposures to PFOS (0.5, 1.0, 3.0. and 6.0 mg/kg/d, 28 days) administered during young adulthood (PND60), reduced hypothalamic GnRH expression at the three lowest doses, increased pituitary LH and FSH expression at 0.5 and 1.0 mg/kg, and decreased circulating LH levels in 0.5 and 3.0 mg/kg/d groups (López-Doval et al., 2014). However, all animals exposed to PFOS exhibited an increase in serum FSH. In a follow-up study from the same group, adult PFOS exposure increased serotonin concentrations in the anterior and mediobasal hypothalamus and the median eminence, suppressed hypothalamic neuropeptide Y expression, and increased expression of genes involved in nitric oxide signaling and metabolism (López-Doval et al., 2015). Interestingly, noradrenaline concentrations increased in the anterior hypothalamus and the median eminence, where the neurotransmitter likely stimulates GnRH and kisspeptin neuronal activity (Kalil et al., 2016; Terasawa, 2019). Since these three signals (noradrenaline, serotonin, nitric oxide) stimulate GnRH neurons and have a role in modulating GnRH release, the modulation of these signals by PFOS exposure suggests that adult exposure to these compounds may contribute to HPG axis dysregulation and infertility. In a final study from the same lab, PFOS exposure (1.0, 3.0, and 6.0 mg/kg/d, 28 days, oral gavage) in adult male rats did not modify pituitary GnRH receptor gene expression but suppressed receptor protein expression at all doses (López-Doval et al., 2016). Thus, high doses of PFOS administered to adults disrupts the expression of hypothalamic neuropeptide expression, neurotransmitter release, and pituitary gonadotropin production. This disruption may negatively influence estrous cyclicity. Indeed, injection of PFOS at 1 and 10 mg/kg/d for 14 days suppressed estrous cyclicity in adult female rats inducing irregular cycles or persistent estrus in 20–30% of the females. Females exposed to 10 mg/kg also exhibited elevated norepinephrine in the paraventricular hypothalamus but not the MPOA which correlated with elevated plasma corticosterone, suggesting that disruption of the HPA axis underlies the effects on estrous cyclicity (Austin et al., 2003).

In mice, studies on PFAS exposure and the hypothalamus have focused solely on female mice. In a chronic (6 month) PFOS exposure model via drinking water (0.1 mg/kg/d) in adult female ICR mice, POA GnRH mRNA expression was reduced by PFOS during proestrous but elevated during diestrus, which correlated with similar changes in hypothalamic GnRH protein content (Feng et al., 2015). Chronic PFOS exposure also modified LH and FSH dependent upon the stage of the estrous cycle. LH and FSH were reduced during proestrus, but only LH was elevated by PFOS during diestrus. These data correlated with a suppression in the number of AVPV Kiss⁺ neurons and AVPV Kiss1 expression. PFOS also reduced E2 serum concentrations suggesting that PFOS disrupts the E2-induced activation of AVPV kisspeptin neurons necessary for the generation of the LH surge. Indeed, PFOS extended the diestrous stage by one day and impaired the LH surge during proestrus, although not the sensitivity of LH to the positive feedback of E2 or kisspeptin injection. In a follow-up study from the same laboratory, subchronic exposure to PFOS (7–30 days, 10 mg/kg, oral) in adult female ICR mice reduced serum concentrations of LH and GnRH after 7 and 14 days of exposure and nearly tripled the length of diestrus, producing irregular cyclicity (Wang et al., 2018). PFOS exposure also reduced the number of Kiss1+ AVPV neurons and Kiss1 protein and gene expression during proestrus without altering ARC Kiss1. Furthermore, PFOS suppressed the E2-induced increase in AVPV Kiss1 mRNA and protein expression and the E2-induced hypothalamic GnRH content and the surge in LH. In a recent study from the same laboratory, another PFAS, PFOA, altered the HPG axis and hypothalamic kisspeptin expression in adult (12 weeks) ICR female mice (Zhang et al., 2020). PFOA, administered by gavage (0.5, 2, or 5 mg/kg/d for 28 days), reduced GnRH and LH serum concentrations at 2 and 5 mg/kg/d. PFOA suppressed the endogenous LH surge and the E2-induced LH surge potentially by attenuating the expression of AVPV Kiss1 mRNA through elevated liver FGF21 production (Zhang et al., 2020). FGF21 was recently demonstrated to suppress actions of vasopressin from the suprachiasmatic nucleus in activating AVPV kisspeptin during the proestrous surge of LH (Owen et al., 2013). Collectively, these data suggest that exposure to high doses of PFAS compounds influences the hypothalamic neurons that control the HPG (and the HPA axis), at the level of gene and protein expression, leading to dysregulation of reproduction. Determining the effects of environmentally relevant doses of these PFAS/PFOS on these genes and the HPG axis in a rodent model should be investigated further.

The influence of adult exposures to flame retardants (PBDEs or OPFRs) is very limited in rodents. Our recent publication on exposure of adult mice to two concentrations of PBDE-47 or a mixture of OPFRs (TPP, TCP, TDCPP) at 1 or 10 mg/kg/d for 28 days reported that Kiss1 expression in the arcuate was increased 2-fold in both sexes (intact males and ovariectomized females) (Krumm et al., 2018). These transcriptional effects were not observed in ERα knockout mice indicating that these flame retardants may act through estrogen signaling to influence gene expression. However, the change in Kiss1 expression did not lead to disruption of estrous cyclicity in intact females (Vail et al., 2020).

The previous sections discuss the impact of perinatal, juvenile, and adult exposures to EDCs on the HPG axis in rodents. In general, most EDC studies, including those above, use acute or subchronic exposures at concentrations higher than human or wildlife exposures. However, humans and wildlife are not exposed solely during these separate developmental windows but across the entire lifespan, from pre-conception to death. Thus, short-term exposure experiments do not accurately recapitulate the anthropogenic environment. We are not suggesting that these short-term experiments are without value as they are necessary to establish mechanisms and determine the most sensitive windows of exposure. Rather, that continuous exposure studies to lower environmentally relevant doses, although time-consuming and costly, would more closely reflect real-world exposure and better address the applicability of chronic EDC exposure to fertility issues and human health outcomes.

In vitro exposures

Using immortalized hypothalamic cell lines or hypothalamic explants provides advantages in studying the mechanisms behind reproductive endocrine disruption. There are only a few studies that have characterized TCDD effects on cells that express GnRH, kisspeptin, and pituitary gonadotropins. In immortalized GnRH cells (GT1–7), TCDD did not alter GnRH content or cell number due to the lack of AhR expression in these cells (Petroff et al., 2003). In explants of the hypothalamus and pituitary from adult Sprague-Dawley female rats, TCDD exposure (3.2 nM, 6 h) did not alter baseline GnRH production or GnRH pulsatility nor did TCDD alter baseline gonadotropin production or average peak amplitude (Trewin et al., 2007). In cultured pituitary cells from male and female Wistar rats at GD20, incubation with TCDD (1 or 10 nM, 1–24 h) attenuated the GnRH-induced expression of Lhb and αGsu potentially via interfering with PKA/PKC signaling (Takeda et al., 2011). In rat hypothalamic GnV-3 cells, TCDD (10 nM) exposure dysregulated Gnrh expression via period circadian protein homolog 1 (Per1) through the actions of AhR (Solak et al., 2013). Finally, in human embryonic kidney cells (GripTite™ 293 MSR) transfected with the Kiss1 promoter constructs and in rat hypothalamic primary cell culture, TCDD (10 mM) suppressed Kiss1 and Gnrh through xenobiotic response element activation (Mueller and Heger, 2014). These data indicate that TCDD suppresses both the neuropeptides (GnRH, kisspeptin) and gonadotropins that control the reproductive axis through the activity of AhR and its downstream effectors.

There are few studies examining the effects of PCBs, DDT, or MXC on the hypothalamic-pituitary circuit in cell models. In a study examining GnRH-stimulated gonadotropin production from primary pituitary cultures from Sprague-Dawley female rats, Aroclor 1242 (0.1, 1, 10, and 100 ppm) alone did not increase LH or FSH after 24 h preincubation except at the 100-ppm dose (Jansen et al., 1993). However, GnRH-stimulated LH and FSH secretion was augmented after 24 and 48 h incubation with A1242 (10 ppm). A1242 (50 ppm) decreased total LH in the presence of GnRH and total FSH only in the absence of GnRH. A1242 (10 ppm) also increased GnRH-stimulated LH release without E2-coincubation. Finally, in a recent study using mouse pituitary gonadotrope cells (LβT2 cells), p,p´-DDT and MXC stimulated mRNA expression of Fshβ and Lhβ in a dose-dependent manner; and both EDCs at 10⁻⁷ M induced FSH and LH protein production (Zhou et al., 2014). Collectively, these cell culture studies illustrate how estrogenic EDCs can alter genes of the HPG axis leading to disruption of puberty, estrous cyclicity, and reproduction, albeit at concentrations higher than typical human exposures.

Conclusions

Collectively, these studies indicate that EDC exposures impact the expression of GnRH and kisspeptin, influence the pulsatile release of GnRH, and subsequently alter gonadotropin production from the pituitary. These influences can occur from single exposures during perinatal, juvenile growth, and adulthood at low concentrations and lead to disruption of steroidogenesis, estrous cyclicity, and the onset of puberty (Bourguignon et al., 2013). Presumably, these EDC impacts are due to activation or influence of ERs, ARs, or in the case of dioxins and PCBs, AhRs, during the sensitive developmental (late gestation), prepubertal, and young adult periods. Future studies should characterize the role of these individual receptors and target neurons upstream of GnRH neurons such as AVPV kisspeptin and arcuate KNDy neurons. Another potential hypothalamic neuron involved in the control of GnRH is RFamide-Related Peptide-3 (RFRP-3) neurons, also known as gonadotropin-inhibiting hormone (GnIH) neurons, which act as a brake on both AVPV kisspeptin and GnRH neurons (Acevedo-Rodriguez et al., 2018). These neurons are also understudied in regards to exposure to endocrine disruptors. Certainly, further investigation is warranted in rodent models using environmentally relevant doses and exposure models.

More research is needed in rodent models and cell lines at relevant concentrations and throughout the lifespan with greater attention to mechanism and the neural control of the HPG axis (Figure 2 summarizes the effects of each class of EDCs on the various hypothalamic region and pituitary). Therefore, we are advocating for investigators to look beyond gene and protein expression in the hypothalamus and the pituitary. Researchers should incorporate classic neural techniques like whole-cell patch clamp electrophysiology to assess the influence of EDC on GnRH, kisspeptin, and RFRP-3 neuronal activity along with assessments of the innervation of kisspeptin and RFRP-3 inputs on GnRH neurons. We have recently characterized the influence of adult exposures to OPFRs on arcuate melanocortin neurons and found that these neurons are excited by OPFR exposure in female mice (Vail and Roepke, 2020), which suggests that other steroid-sensitive hypothalamic neurons may also be a target for adult exposure to EDCs. Another avenue of investigation is assessing LH pulsatility in wild-type and transgenic mouse models as LH pulsatility can be reliably measured using an ultra-high sensitive LH assay from small amounts (~6 μl) of blood (Steyn et al., 2011). These novel studies, when employed together, would assess the influence of EDCs at the level of the synapse and the circuit and the downstream control of LH, all of which are important for assessing the mechanisms of EDC toxicity and its influence on reproduction across species.

Figure 2 – Summary of tissue-specific actions of EDCs on reproductive hormones and neuropeptides.

GnRH neurons in the POA are targeted by PCBs, specifically A1221, DDT (and its metabolites), PBDEs, and TCDD. POA GnRH neurons express AhR and ERβ (ERbeta). Kisspeptin neurons in the AVPV are targeted by PCBs, specifically A1221, and PFAS/PFOA and express ERα (ERalpha). KNDy neurons in the arcuate nucleus are targeted by OPFRs, PBDEs, PFAS/PFOA, and TCDD and express AhR and ERα (ERalpha). Pituitary gonadotropes are targeted by A1221, DDT (and its metabolites), MXC, PCBs, PFAS/PFOA, and TCDD. Gonadotropes express AhR and ERα (ERalpha). In some studies, the effects of the EDCs on LH or FSH are due to disruption of steroidogenesis in the gonads and reflect a secondary effect of changes to negative feedback.

Our review has several limitations in regard to relevancy to human health. First, as we have highlighted throughout the review, many of these studies are using concentrations of EDCs higher than what is reported in the home and in occupational or natural environments. Thus, the direct translational relevance is not strong, in that regard. However, due to the continuous nature of human exposures to concentrations for a wide range of EDCs and for an expanding mixture of chemicals, such short-term or acute studies are necessary to demonstrate potential molecular mechanisms – hormone receptors, steroidogenic enzymes, epigenetic modifications of reproductive genes in the HPG axis, etc. Furthermore, few studies examine the effects of a mixture of EDCs from many classes of chemicals on the HPG axis. This is, in part, due to the limitations made by review panels and funding sources and their focus on the mechanisms instead of the necessary descriptive studies on the influences of multiple EDC exposures. Second, we focused on rats and mice as the majority of studies in this field are focused on mammalian reproduction. However, there is an increasing number of studies that use domesticated mammals, fish, and invertebrate animal models (Bellingham et al., 2016; Cuvillier-Hot and Lenoir, 2020; You and Song, 2021), which may be key in determining receptor-mediated mechanisms. As fish are more sensitive to these compounds than rodents, the impact of these EDCs on fish reproduction is a major cause of concern for aquatic species, including commercially relevant fish stocks. These findings can then be compared and integrated into studies in rodents and in epidemiological studies in humans. Finally, the response to EDCs is influenced by multiple factors – sex, age, and diet. Most of these studies and EDC studies focused on young (<6 months) animals, regardless of exposure window, and did not include interactions with diets. Maternal obesity or high-fat diet feeding can disrupt reproduction (Yao et al., 2021) and diet-induced obesity in adults also disrupts reproduction (Negrón and Radovick, 2020). The interactions of EDC exposures and obesity on reproduction are an understudied area of reproductive toxicology. In summary, these data in rodents and cell lines are highly suggestive that chronic exposures to EDCs in humans may have significant implications for public health and regulatory policy across the globe. Chronic exposures to EDCs across generations will influence the fertility of future generations and lead to an increase in reproductive diseases.

Table 2.

Summary of effects of PCBs on GnRH, kisspeptin, and gonadotropins

EDC	Source	Exposure	Dose	Model	GnRH and pulse	Kisspeptin	Gonadotropins
A1221; PCB138,153, 180	Dickerson et al. (2011a)	Perinatal GD16–18	1mg/kg	Rat	F:↓ # of c-fos+ GnRH cells	F:↓# of ERα+ AVPV neurons (kisspeptin?); ↓AVPV
A1221; PCB138,153, 180	Dickerson et al. (2011b)	perinatal PND1	1mg/kg	Rat	NC	NC: Kiss1; M:↓Kiss1r; F:↑Kiss1r
A1221	Steinberg et al. (2008)	multigenerational	0.1–10 mg/kg	Rat			F: ↑serum LH in F1; ↓proestrus LH in F2
A1242	Jansen et al. (1993)	adult - primary pituitary cultures	0.1–100 ppm	Rat			↑LH & FSH after GnRH-stimulation; ↓LH; ↓FSH w/ GnRH
PCB153	Desaulniers et al. (1999)	young adult	25 mg/kg/d	Rat			↓serum LH; No effect on pituitary LH
PCB126 PCB114	Han et al. (2010))	young adult	0.2 mg/kg PCB126; 20 mg/kg of PCB114	Rat			M:↓ serum FSH PCB126; ↑serum LH PCB114
PCB126	Desaulniers et al. (1999)	adult	6.25–400 mg/kg/d	Rat			M:↓ serum LH; ↑pituitary LH & FSH
PCB180A1221	Uslu et al. (2013)	adult	10 mg/kg	Rat			F: ↓ serum FSH

NC, measured but no change; ↑, increase in expression of gene or protein; ↓, decrease in gene or protein; F, female; M, male.

Table 3.

Summary of effects of organochlorine pesticides on GnRH, kisspeptin, and gonadotropins

EDC	Source	Exposure	Dose	Model	GnRH and pulse	Kisspeptin	Gonadotropin
DDE	Makita (2008)	perinatal	125 ppm	rat			F: NC
MXC	Roepke et al. (2016)	perinatal		F rat		NC ARC
MXC	Masutomi et al. (2004)	perinatal	1200 ppm	Rat			M: ↓% of LH+, FSH+ cells; F: ↓LH+ cells, ↑% of FSH+ cells
MXC	Suzuki et al. (2004)	perinatal	24, 240, 1200 ppm	M rat			M: ↓serum FSH, LH; F: ↓proestrous LH surge
o,p´-DDT, p,p´-DDE	Rasier et al. (2007)	juvenile	10, 100 mg/kg/d	F rat	↓GnRH IPI; ↑pulsatility in PND15		↓Serum LH concentrations
o,p´-DDT p,p´-DDT	Rasier et al. (2008)	juvenile	10−5-10−4 M	F rat	↑Glut-evoked GnRH, pulse amp, pulsatile GnRH
MXC	Dickerson et al. (2011b)	juvenile	20 μg/kg, 100 mg/kg	F rat		NC AVPV
DDT	Ben Rhouma et al. (2001)	adult	50mg, 100 mg/kg	Rat			M: ↑FSH, LH
MXC	Lafuente et al. (2008)	adult	25 mg/kg/d	Rat			↓plasma LH; NC FSH
MXC	Tomic et al. (2006)	adult	64 mg/kg/d	Mouse			↓serum FSH levels
p,p´-DDT MXC	Zhou et al. (2014)	cell	10−7 M	LβT2 cells			↑Stimulation Fshβ, Lhβ, FSH, LH protein

NC, measured but no change; ↑, increase in expression of gene or protein; ↓, decrease in gene or protein; F, female; M, male.