Epigenetic Mechanisms in the Actions of Endocrine-disrupting Chemicals: Gonadal Effects and Role in Female Reproduction

Authors


Author’s address (for correspondence): M Uzumcu, Department of Animal Sciences, School of Environmental and Biological Sciences, Rutgers, The State University of New Jersey, 84 Lipman Drive, New Brunswick, NJ 08901-8525, USA. E-mail: uzumcu@aesop.rutgers.edu

Contents

There is a heightened interest and concern among scientists, clinicians and regulatory agencies as well as the general public, regarding the effects of environmental endocrine-disrupting chemicals (EDCs). In this review, we identify the main epigenetic mechanisms and describe key ovarian processes that are vulnerable to the epigenetic actions of EDCs. We also provide an overview of the human epidemiological evidence documenting the detrimental effects of several common environmental EDCs on female reproduction. We then focus on experimental evidence demonstrating the epigenetic effects of these EDCs in the ovary and female reproductive system, with an emphasis on methoxychlor, an organochlorine pesticide. We conclude the review by describing several critical issues in studying epigenetic effects of EDCs in the ovary, including transgenerational epigenetic effects.

Introduction

Endocrine-disrupting chemicals are synthetic or natural compounds in the environment that can interfere with endocrine functions. The EDCs in the environment can be categorized as pesticides, plasticizers, industrial side products, pharmaceuticals, flame-retardants, phyto-oestrogens or heavy metals such as cadmium (Diamanti-Kandarakis et al. 2009; Iavicoli et al. 2009; Kortenkamp 2011).

Developmental Exposures to EDCs and the Barker Hypothesis

Exposure to EDCs during development is a bigger concern than exposure during adulthood. The health consequences of developmental exposure are mostly permanent or long-lasting, while the individuals who are exposed during adulthood usually regain their normal physiological functions after the withdrawal of chemicals. The long-lasting effects of EDCs can be viewed from the perspective of developmental origins of health and diseases (DoHAD) concept, often called the Barker hypothesis. This concept was first developed based on the observation that in utero exposure to poor nutrition can lead to various adult disorders, including metabolic and cardiovascular diseases (Barker 1995). The same concept applies to EDC exposure during the critical window(s) of development, which can alter the programming of the female reproductive organs, including the ovary and lead to infertility. Abnormal developmental programming occurs when the adverse environmental conditions alter the epigenetic mechanisms that modify gene expression patterns (Heijmans et al. 2009).

Epigenetic Mechanisms

Epigenetics is the study of mitotically and meiotically heritable changes in the gene function (i.e. gene expression) without changing the DNA sequence. Three common epigenetic mechanisms are DNA methylation, post-translational modifications of histone proteins (histone modifications) and non-coding RNA (ncRNA) (Jirtle and Skinner 2007), and they collectively make up the major components of the epigenome. Exposure to EDCs can alter these three epigenetic mechanisms in the ovary and various female reproductive organs, leading to gene expression changes (Zama and Uzumcu 2009; Bredfeldt et al. 2010; Luense et al. 2011).

DNA methylation

Methylation of DNA occurs at cytosine residues in CpG dinucleotides and plays a role in genomic imprinting, suppression of retrotransposons and X-chromosome inactivation, as well as gene expression regulation (Ehrlich 2003). CpGs are underrepresented in the genome, except in segments of DNA called CpG islands (CGIs), where the CG content is markedly increased and is associated with the regulatory sequences of the genes (Turner 2009). Methylation of DNA can interfere with transcription factor binding, leading to a reduction in gene expression. In the absence of DNA methylation, transcription factors can bind to the regulatory sequences leading to upregulation of gene expression (Yagi and Koshland 1981; Ehrlich 2003). The DNA methylation alterations in CpGs outside CGIs are now also considered to be important in gene regulation (Suzuki and Bird 2008), necessitating more genome-wide methylation studies.

Histone modifications

Histones are the protein components of the nucleosome, through which genomic DNA is packaged in eukaryotes. Post-translational modifications of histones at lysine, arginine, serine and threonine amino acids play significant roles in both structural and functional states of the chromatin (Hirose et al. 1985; Turner 2009). These modifications include acetylation of lysine, methylation of lysine and arginine and phosphorylation of serine and threonine. Whereas acetylation of lysine residues is associated with the relaxation of chromatin, allowing access to transcription factors and active transcription, deacetylation leads to compaction of the chromatin, preventing access of transcription factors and silencing the loci. Similarly, methylation or phosphorylation can be associated with gene activation or gene silencing, depending on the position of the amino acid modified [see (Sharma et al. 2005; Yoo and Jones 2006; Turner 2009) for more details].

Non-coding RNA

The importance of ncRNAs in biological processes has been recently recognized (Chang et al. 2006). Non-coding RNAs are transcripts without a clear open reading frame; therefore, they do not code proteins, but regulate the expression of other genes in cis and trans manners. The ncRNAs vary in size from microRNAs, or miRNAs [15–21 nucleotides (nt)], and small RNAs (100–200 nt) to large RNAs (>200 nt) and are involved in vital functions such as X-chromosome inactivation, genomic imprinting, transposon and virus silencing, and developmental patterning and differentiation [reviewed in (Costa 2008; Brosnan and Voinnet 2009)].

The above-mentioned three epigenetic mechanisms are known to influence each other’s roles in cellular development, differentiation and function (Li 2002; Costa 2008; Namihira et al. 2008). During establishment of the epigenetic marks, histone modifications lead to DNA methylation, whereas in maintenance or inheritance of the marks, DNA methylation plays a primary role. In addition, ncRNAs, histone modifications and DNA methylation work cooperatively in various processes, including X-chromosome inactivation (Avner and Heard 2001) and genomic imprinting (Bartolomei 2009; Brosnan and Voinnet 2009).

Major Epigenetic Modifications in the Mammalian Life Cycle

The epigenome, especially DNA methylation patterns, is stably maintained in somatic cells. However, major epigenetic programming occurs during two stages of development: (i) in the pre-implantation embryo and (ii) in primordial germ cells (PGCs) in the bipotential gonad (Reik et al. 2001). In the pre-implantation embryo, a genome-wide epigenetic reprogramming occurs through which the epigenetic marks, except on imprinted genes, are eliminated (Santos and Dean 2004), the zygote gains its totipotency, and embryonic cell lineage-specific methylation occurs, including in the PGCs. After the PGCs migrate from the extragonadal sites to the genital ridge, the DNA methylation patterns of the imprinted genes in the PGCs are also erased in both sexes (Fig. 1) (Hajkova et al. 2002) and re-established in a sex-specific manner. In males, the remethylation starts on embryonic day (E) 14 and is mostly completed by E16–17 (Ueda et al. 2000). In contrast, the female germ cell remethylation is initiated during post-natal day (PND) 1 through PND5 and continues throughout oocyte growth until the pre-antral follicle stage in mice (Obata and Kono 2002).

Figure 1.

 The stages of gonadal development in females (see the text for details). One of the processes that are particularly vulnerable to the epigenetic effects of endocrine-disrupting chemicals is epigenetic reprogramming (i.e. DNA demethylation and remethylation) in primordial germ cells (PGCs). During the genome-wide epigenetic programming in the pre-implantation embryo, the DNA methylation patterns in both male and female genomes are eliminated (not shown), except in imprinted genes (colour solid lines) whose methylation patterns are maintained, including in PGCs until their migration to the gonad. After the migration, the patterns are erased (dotted lines) and re-established (colour solid lines) in a sex-specific manner. The timeline shown is primarily based on the events in mice. Adapted from (Zama and Uzumcu 2010)

The two stages that are decribed above are likely to be the most vulnerable to the epigenetic effects of EDCs. If EDCs interfere with DNA methylation during the epigenetic reprogramming in the pre-implantation embryo, all three embryonic germ layers can be affected, and therefore all tissues of the organism (Dolinoy et al. 2007). If EDCs alter epigenetic reprogramming (i.e. imprinted genes) of the germline, subsequent generations can be affected (Stouder and Paoloni-Giacobino 2011). Although the effects that occur during the erasure can influence both sexes, the effect during remethylation can be sex-specific.

Ovarian Development and Function

In addition to effects that can occur during the two major epigenetic modifications described above, EDCs can have ovary-specific effects if the exposure takes place during the critical developmental processes in the ovary. In addition, the cyclic processes that occur in adult ovary throughout the female’s reproductive life which are closely regulated by local and endocrine factors, including estradiol (E2), progesterone (P4) and androgens, can be affected by EDCs that display oestrogenic, anti-oestrogenic and anti-androgenic actions (Zama and Uzumcu 2010). The specific ovarian processes are discussed using a timeline that applies to rodent species (Fig. 1). We focused on processes that are likely to be affected by EDCs, rather than describing the normal ovarian development and function in detail, which can be found in recent reviews (Edson et al. 2009; Gougeon 2010).

Oocyte nest breakdown and primordial follicle formation

Oocyte nest breakdown that leads to the formation of primordial follicles is mediated by oocyte apoptosis starting at E16.5 in mice (Pepling and Spradling 2001). By PND3–4, the primordial follicles are formed. Each primordial follicle consists of an oocyte at the centre surrounded by a single layer of squamous pre-granulosa cells (GCs). In vitro and in vivo experiments have shown that the steroids P4 and E2 inhibit primordial follicle formation by inhibiting oocyte nest breakdown, leading to multi-ovular follicles (MOF; presence of more than one oocyte within a follicle) (Kezele and Skinner 2003; Chen et al. 2007).

Primordial-to-primary follicle transition

The transition from primordial to primary follicles, which is also known as ‘initial recruitment’ (McGee and Hsueh 2000), ensures a steady supply of follicles for further growth and ovulation. Once the primordial follicle is recruited, the oocyte undergoes growth and maturation to gain competence for fertilization, and the pre-GCs differentiate into GCs by growing in size and becoming cuboidal and they proliferate. Most of the growing follicles (99.9%) are destined to become atretic, and the rest will ovulate after completing the following follicular stages: primary; secondary; pre-, early-, mid- and late antral; and pre-ovulatory (see (Armenti et al. 2008) for a detailed description of follicular stages).

Both E2 and P4 also have inhibitory roles in the initial recruitment, and in the ovaries that are exposed to E2 and P4in vitro, the transition is inhibited (Kezele and Skinner 2003). In addition, when newborn rats are injected with estradiol benzoate, their ovaries have more primordial follicles and fewer growing follicles (Ikeda et al. 2001). In contrast, cytochrome P450 aromatase knockout (ArKO) mice, which lack endogenous E2, have fewer primordial and primary follicles (Britt et al. 2004). These results show that whereas E2 is needed for early events in folliculogenesis, excessive E2 inhibits these events. In contrast, androgens stimulate the primordial-to-primary follicle transition but inhibit follicle growth beyond the primary stage (Yang and Fortune 2006; Yang et al. 2009).

Pre-antral and early-antral follicular growth

Follicular growth during primary and secondary stages is regulated by local factors. Starting with the secondary follicle stage, follicles begin to gain receptors for gonadotropins, follicle-stimulating hormone (FSH) and luteinizing hormone (LH) (Sokka and Huhtaniemi 1990). Around this time, follicles also gain a second somatic cell type, thecal cells. The follicular progression becomes dependent on gonadotropins during the pre-antral stage (Kumar et al. 1997). Granulosa and thecal cells cooperate in the production of E2 under the regulation of FSH and LH, respectively (Magoffin 2005). Increasing E2 concentrations cause negative feedback at the hypothalamus and pituitary levels, thereby reducing the secretion of gonadotropins.

Oestradiol plays a stimulatory role in later stages of folliculogenesis. This role is the clearest in transgenic mouse strains, in which ERβ is globally deleted via homologous recombination (Carpenter and Korach 2006) or ERα is deleted in thecal cells (Lee et al. 2009), the primary cell type that expresses ERα. Both mouse strains show inhibition in follicular maturation and ovulation. In contrast, while androgens are stimulatory during the initial stages of folliculogenesis, excess androgen is inhibitory and its receptor is actively down-regulated at the later stages of folliculogenesis (Tetsuka et al. 1995).

Follicular selection, maturation and ovulation

To complete the full course of folliculogenesis and ovulation, one follicle within the recruited cohort is selected in mono-ovulators, whereas multiple follicles are selected in polyovulators. The selection process is under complex regulation, which is primarily determined by the elevated gonadotropin responsiveness of the selected follicle(s) despite decreasing gonadotropin concentrations. A whole host of local growth factors, such as insulin-like growth factors (IGFs), plays significant roles throughout folliculogenesis including the final maturation (reviewed in (Edson et al. 2009; Sirotkin 2011). At the pre-ovulatory stage, elevated E2 exerts positive feedback at the hypothalamus, leading to the pre-ovulatory LH surge. During this stage, GCs also gain responsiveness to LH, which initiates numerous gene expression alterations that prepare the follicle for ovulation (Richards and Pangas 2010). The oocyte is released during the ovulation along with surrounding cumulus cells – a special subpopulation of GCs that are immediately adjacent to oocyte. The remaining cells of the follicle form the corpus luteum (CL), which is the source of the P4 that prepares the reproductive tract for the pregnancy. The formation, function and demise of the CL (in the case of continuous reproductive cycle) and the maintenance of CL (in case of pregnancy) are closely regulated events (Stocco et al. 2007), which are affected by EDCs.

Although the ovary plays a central role in female reproduction, its function is closely regulated by the hypothalamus and pituitary. The EDCs can act at the levels of the hypothalamus and pituitary as well as the ovary. However, in this review, we focus on the direct effects of select EDCs in the ovary.

Epidemiological Evidence Linking EDCs to Female Reproductive Disorders

Epidemiological studies strongly suggest that EDCs affect both male and female reproduction. One of the earliest reports related to the decline in sperm parameters, such as reduced sperm count and seminal volume in men during the previous 50 years, pointed out the harmful effects of EDCs (Carlsen et al. 1992). This time period closely corresponds to the time of the widespread use of synthetic EDCs. Although this 1992 report is controversial (Fisch et al. 1996; Younglai et al. 1998), there is agreement regarding a geographical region-dependent decline in sperm parameters (Swan et al. 2003), which supports an environmental aetiology. For example, several studies have shown that there are clear differences between male fertility parameters of men in Denmark and Finland. These studies show that semen quality parameters are lower while the incidence of testicular cancer, hypospadias and cryptorchidism is higher in men living in Denmark, including the Denmark-born offspring of immigrants, as compared to men from Finland (Boisen et al. 2004; Jorgensen et al. 2006; Myrup et al. 2008). The human data have been supported by findings from laboratory animals (Gray et al. 2001) and wildlife species (Guillette and Gunderson 2001). The pathologies in males are collectively named as testicular dysgenesis syndrome (TDS), which has been extensively reviewed (Toppari et al. 2010).

Females also have EDC-originated fertility problems, leading to ovarian dysgenesis syndrome (ODS) (Buck Louis et al. 2011). Although it is not as clearly defined or well accepted as TDS, the use of the term ODS is gaining support (Fowler et al. 2012). The impaired fecundity rate in the United States increased from 11% to 15% between 1982 and 2002 (Guzick and Swan 2006). Although various other confounding factors such as lifestyle changes can contribute to this decline, the role of EDCs is also strongly considered because of supporting studies. The incidence of female reproductive disorders, such as ovarian cancer, is on the rise in a region-dependent manner (reviewed in (Buck Louis et al. 2011)). Furthermore, numerous studies suggest a strong link between exposure to EDCs either during the adult life or during development and increased female reproductive disorders.

Diethylstilbestrol

The most convincing human evidence that oestrogenic EDC exposure during development can permanently affect female reproduction comes from reports of the effects of diethylstilbestrol (DES), a non-steroidal synthetic oestrogen. From the 1940s to the 1970s, DES was prescribed at doses of 5–150 mg/day to prevent miscarriages (Smith and Gabbe 1999). Numerous abnormalities in the reproductive, cardiovascular and immune systems have since been reported in both male and female offspring of women treated with DES, and similar effects have been demonstrated in animal models [see (Newbold 2004) for specific examples]. Some of these effects, such as irregular cylicity or ovarian cancer, are being observed in the granddaughters of DES-treated women as well (Blatt et al. 2003; Titus-Ernstoff et al. 2006). In the light of this multi-generational aspect, whether epigenetic mechanisms are involved is a significant question (Newbold et al. 1998).

Bisphenol A

Bisphenol A (BPA) is a plasticizer used in the manufacture of polycarbonate plastics and epoxy resins, and its total worldwide production exceeds 6 million tons per year. Exposure to BPA may occur through various materials that humans use daily, such as food and drink cans, baby bottles and carbonless paper [reviewed in (Zama and Uzumcu 2010)]. As a result, 95% of adults who were tested had detectable levels of BPA in their urine (Calafat et al. 2008). Urine BPA levels of women undergoing infertility treatment are negatively correlated with the number and quality of eggs retrieved, and with serum E2 levels (Mok-Lin et al. 2010; Fujimoto et al. 2011).

Genistein

Genistein is a well-studied isoflavonoid phyto-oestrogen derived from soy. Phyto-oestrogens were originally shown to have endocrine-disrupting potential in domestic animal species: newborn lambs born to ewes fed clover, a rich source for phyto-oestrogens, had reproductive abnormalities [reviewed in (Jefferson 2010)]. Soy products are popular because they are lactose-free substitutes for dairy products and are considered to be protective against breast cancer risks (Peeters et al. 2003). However, a major cause for concern is that babies who are fed soy formula consume an average of 6–9 mg/kg body weight, which is many-fold higher than the levels approved by the FDA for adults for cardio-protective purposes (Setchell et al. 1998; Johns et al. 2003). It is alarming to know that early life exposure to soy formula is associated with a greater risk of uterine fibroids in adulthood (D’Aloisio et al. 2010).

Phthalates

Phthalates are used as plasticizers in polyvinyl chlorides, personal care products, some flooring, car products, medical devices, and insect repellent. Although phthalates have been mostly studied for their effects on male reproduction, there is accumulating evidence that they adversely affect female reproduction as well. Di-(2-ethylhexyl) phthalate and its metabolite mono-(2-ethylhexyl) phthalate (MEHP) are the most commonly studied phthalates (reviewed in (Craig et al. 2011)). There is increasing evidence that exposure to phthalates may affect reproductive health in humans (Jurewicz and Hanke 2011). The levels of phthalates, especially MEHP, in the urine of women undergoing infertility treatments correlate with significantly higher risk of implantation failure (Ehrlich et al. 2010).

Methoxychlor

Methoxychlor (MXC) is a well-studied organochlorine pesticide used as a replacement for dichlorodiphenyltrichloroethane (DDT). It is an oestrogenic compound that demonstrates low-affinity binding to oestrogen receptors (Cummings 1997). The major MXC metabolites, 2,2-bis-(p-hydroxyphenyl)-1,1,1-trichloroethane (HPTE) and mono-OH MXC, can function as oestrogenic, anti-oestrogenic or anti-androgenic compounds (Gaido et al. 2000), and therefore, it is used as a model compound (Uzumcu and Zachow 2007). Epidemiological studies have shown that there is a strong association between developmental exposure to organochlorine pesticides and subsequent female fertility problems (reviewed in (Zama and Uzumcu 2010)). In addition, in a study that examined pesticide-use data and the breast cancer incidence rates in Hispanic women in California, the amount of MXC used was positively associated with increased breast cancer incidence (Mills and Yang 2006).

Experimental Evidence from In Vivo Studies Regarding the Effects of EDCs in the Ovary

Characteristic sets of defects in the ovary are the hallmark of EDC exposure in rodent models. For example, mice injected as little as a single dose of 10 μg/kg DES on E15 and examined at 7 months of age had numerous atretic follicles and no corpora lutea (an indicator of ovulation) (Wordinger and Highman 1984). Other studies in rodents of varying doses of DES administered either in utero (E9–16) or neonatally (PND1–5) demonstrated similar defects by adulthood (Haney et al. 1984; Tenenbaum and Forsberg 1985). Oestrous cyclicity was completely disrupted, and high levels of testosterone were found. A vacuolated interstitial tissue with lipid droplet inclusions and haemorrhagic cysts was also observed. A recent study proposes that while increased lipid droplet is caused by impaired steroidogenesis because of suppressed LH levels, the haemorrhagic follicles are the results of direct effects of DES in the ovary (Kakuta et al. 2012). Furthermore, there was a dose-dependent reduction in the number of litters as well as the number of oocytes ovulated after stimulation with exogenous gonadotropins (McLachlan et al. 1982) with such oocytes when used in IVF showing lower levels of fertilizability, suggesting reduced oocyte quality (Wordinger and Derrenbacker 1989; Iguchi et al. 1991). Most striking of all, an excessive number of MOFs is found in adult ovaries; such a finding is considered to be an indicator of reduced reproductive lifespan (Iguchi and Takasugi 1986; Iguchi et al. 1990). The above-mentioned phenotype is mostly recapitulated by exposure to other EDCs, including BPA (Howdeshell et al. 1999; Markey et al. 2005) and genistein (Jefferson et al. 2005; Chen et al. 2007). Phthalates also can decrease E2 production, prolong the oestrous cycle and cause anovulation (Davis et al. 1994; Lovekamp and Davis 2001); however, effects of phthalates on the ovarian follicular composition are not known.

Such oestrogenic actions of these EDCs are mediated via the ER signalling pathway (Kuiper et al. 1998; Drummond and Fuller 2010; Patisaul and Adewale 2009). Recent studies have shown that MOFs induced by neonatal exposure to 3 μg/kg DES are mediated by ERβ and not by ERα (Kirigaya et al. 2009). Diethylstilbestrol exposure was shown to reduce oocyte apoptosis (potentially suppressing oocyte nest breakdown) via ERβ signalling mechanisms. Furthermore, it was hypothesized that alterations in the germ cell-to-somatic cell ratio may affect the invasion of pre-GCs and basement membrane remodelling during primordial follicle formation (Kim et al. 2009). In contrast to ERβ signalling mechanisms involved in mediating ovarian effects, Couse and colleagues reported that ERα is essential for the mediation of DES effects in the uterus: αERKO female mice exhibited a complete resistance to the effects of DES, while βERKO mice did not (Couse and Korach 2004). Interestingly, in utero treatment of βERKO females with low doses of BPA did not enhance the oocyte meiotic defects usually caused by BPA, suggesting that the compound could act via the ERβ signalling pathway (Susiarjo et al. 2007). Furthermore, microarray-based gene expression analysis in the foetal ovary (E12–14.5) after low-dose BPA exposure revealed that mitotic cell cycle genes are significantly down-regulated, potentially leading to accelerated meiotic entry of the oogonia and maybe eventually leading to a reduced size oocyte pool after birth (Lawson et al. 2011). Another recent study in rats showed that BPA reduces the pool of primordial follicles by stimulating initial recruitment after a PND1-7 exposure (Rodriguez et al. 2010). Interestingly, Arase and co-workers reported an increase in oestrogen production potentially through increase in Cyp19a1 and Cyp11a1 expression in the mouse urogenital sinus, specifically in the mesenchymal component after BPA exposure between E9 and E15 (Arase et al. 2011). Thus, BPA has a double-edged effect on mitosis and meiosis as well as pre- and post-natal steroidogenesis.

Epigenetic Mechanisms Associated with the Female Reproductive System

Although there are now well-documented studies of the physiological and morphological effects of EDCs in the ovary, the early research providing evidence of an epigenetic component of EDC actions was predominantly on the uterus. For example, it is well known that DES caused T-shaped uteri and clear cell adenocarcinoma of the uterus, cervix and vagina in women whose mothers were exposed to DES during pregnancy (Herbst et al. 1971). In the numerous animal studies validating these human reports, developmentally DES-treated mice manifest malformations of the uterus, squamous metaplasia of the luminar and glandular epithelium, endometrial hyperplasia and leiomyomas, and oviductal proliferative lesions (McLachlan et al. 1980; Kitajewski and Sassoon 2000). Ovariectomized animals when supplemented with E2 are able to respond by a transient increase in gene expression and concomitant uterine proliferation and growth. When such a stimulus was removed, the uterus returned to its unstimulated state. However, when DES or E2 is administered during neonatal development, expression of immediate early genes such as c-fos, c-jun and c-myc as well as lactoferrin and EGF is upregulated even into adulthood (Nelson et al. 1994; Falck and Forsberg 1996). This was associated with hypomethylation of the promoter region of the lactoferrin gene in the adult uterus (Li et al. 1997). However, if animals were exposed for the same interval during adulthood, no such methylation or expression defects were observed, indicating the importance of developmental exposure. Subsequently, it was also found that exon 4 of the c-fos gene was extensively hypomethylated, while the promoter region and intron 1 were unaffected, thereby potentially allowing for the upregulation of c-fos expression (Li et al. 2003). Furthermore, recent work by Block and co-workers has shown that DES disrupts the regionalization of expression of Hoxa-9 and Hoxa-10 and the homoeotic anterior transformations associated with hypermethylation in the promoter and intron 1 regions of Hoxa-10 gene (Block et al. 2000). Additionally, ERα induction is necessary for activation of oestrogen-responsive gene expression in the uterus, including that of the lactoferrin and c-fos genes (Nelson et al. 1994; Falck and Forsberg 1996). As many of the above-mentioned genes are downstream of ER signalling, which is involved in direct actions of EDCs, it is imperative to thoroughly examine the potential role of epigenetic mechanisms in the regulation of ER expression after EDC exposure. In addition, PI3K/Akt signalling downstream of membrane-associated ER signalling caused reduction in trimethylation of the histone H3K27 in response to E2 and DES exposures. More interestingly, activation of this non-genomic signalling caused reprogramming of the uterine gene expression profile (Bredfeldt et al. 2010). Another example is that of the investigation by Tang et al. (2008), wherein it was demonstrated that neonatal DES/genistein exposure reduced DNA methylation and increased gene expression of nucleosomal binding protein 1 in adult uteri. These studies highlighted the age-dependent aspect of epigenetic reprogramming and also its interaction with steroid hormones.

Epigenetic Effects of MXC in the Ovary

Early studies that exposed females to MXC for 6–10 weeks during both in utero and early post-natal development periods show that exposed females have acceleration of the vaginal opening (sign of puberty), acceleration of the onset of the first oestrus, irregular cycles with persistent vaginal oestrus, reduced pregnancy rate and litter size despite apparent mating, and early reproductive senescence (Gray et al. 1989; Chapin et al. 1997; You et al. 2002). Serum E2 and P4 levels were altered with increased FSH levels (Chapin et al. 1997). The effects on the ovary were dramatic, with both folliculogenesis and ovulation being inhibited.

In our more recent study, female rats were treated for 12 days during foetal and neonatal development (E19–PND7) with 100 mg/kg/day MXC, higher than the dose normally encountered by humans or animals on a daily basis, but similar to doses used in the above studies. The exposed females displayed similar abnormalities in reproductive parameters and in ovarian morphology by adulthood despite the fact that the treatment period was much shorter (Armenti et al. 2008). A close examination of follicle composition showed that developmental MXC treatment did not affect the total number of follicles or follicles at primary and secondary stages in adult females. However, the number of pre-antral and early-antral follicles was increased and the number of CL was reduced, with numerous cystic follicles. Immunohistochemical staining and quantification of expression patterns of important regulators of ovarian functions revealed that while LHR, CYP11A1 and CYP19A1 levels were reduced, levels of anti-Mullerian hormone and androgen receptor were increased, and levels of StAR and ERα were unchanged (Armenti et al. 2008). Especially noteworthy was that ERβ level was unchanged in primary and secondary follicles, yet decreased dramatically in peri-antral stage follicles, which are very responsive to gonadotropins (unpublished observations). These data suggest that hormone-responsive follicles are most affected by EDC exposure.

Epigenetic analyses using bisulphite-sequencing PCR and methylation-specific PCR showed that MXC caused hypermethylation in multiple CpGs in two CGIs in ERβ promoter sequences (Zama and Uzumcu 2009). Further analysis has shown that the DNA methylation levels in the promoter regions of these genes were unchanged in neonatal ovaries (PND7) immediately after the exposure (A. M. Zama and M. Uzumcu, unpublished data). These data demonstrate the age dependence and hormone responsiveness of the epigenetic changes, which has also been shown in other tissues (e.g. uterus) with other compounds (e.g. DES, genistein) (Tang et al. 2008). The global DNA methylation analysis using methylation-sensitive arbitrarily primed PCR (AP-PCR) showed that there were multiple loci that were hypermethylated in MXC-treated ovaries (Zama and Uzumcu 2009). The majority of candidates were those encoding transcription factors or ribosomal proteins. Our most recent studies have involved genome-wide methylation analyses. Since assays such as AP-PCR depend on the availability of restriction enzyme sites, they may not capture all possible methylation events. Furthermore, because DNA methylation is dynamic and reversible (Bhutani et al. 2011), we hypothesized that genome-wide methylation analyses at both ages would be highly informative. Therefore, we examined the ovaries (i) immediately after exposure, at PND7, and (ii) after animals reached adulthood, at PND60. The methylated DNA immunoprecipitation protocol (MeDIP), which employs an antibody against the 5-methyl cytosine moiety, was conducted followed by hybridization to microarrays that included probes spanning the promoter regions as well as the CGIs of all RefSeq genes in the rat genome. This was in an effort to examine not only individual gene methylation patterns but also potential gene networks that were exhibiting altered methylation patterns in response to EDC exposure, immediately after exposure as well as in adulthood, long after the exposure ends. We have now found that multiple loci belonging to critical ovarian signalling pathways (e.g. PTEN signalling, IGF signalling) were hypermethylated and their gene expression was suppressed. These pathways are essential for early folliculogenesis as well as follicular maturation and ovulation.

Conclusions and Future Studies

Several aspects should be considered when the effects of EDCs in the ovary are evaluated. The ovary is the major source of the hormones that are mimicked by EDCs. The same hormones that the ovary produces also regulate the ovarian functions. Thus, the effects of EDCs in the ovary can be twofold. In addition, the ovary of a reproductive-aged adult consists of follicles in various developmental stages as well as corpora lutea. Therefore, both developing and adult ovaries require involvement of numerous factors that are spatio-temporally regulated. It is likely that these factors are regulated by various epigenetic mechanisms, including DNA methylation (Vanselow et al. 2010), histone modifications (LaVoie 2005) and ncRNAs (Carletti et al. 2010; Torley et al. 2011; da Silveira et al. 2012).

One of the most pressing issues in the field of epigenetics and EDCs is the transgenerational (TG) epigenetic effects. All epigenetic actions in the ovary have the potential to affect the subsequent generation(s) via the female germ cells. The effects that are transmitted through epigenetic mechanisms to an unexposed generation are described as epigenetic TG inheritance (Youngson and Whitelaw 2008). There have been reports in males that EDCs can cause TG epigenetic effects in experimental animals (Anway et al. 2005; Guerrero-Bosagna et al. 2010). Since then, there have been additional reports concerning the effects of EDCs in males (Salian et al. 2009; Stouder and Paoloni-Giacobino 2011). Although the described effects are also transmitted by the female germline, the impacts are less obvious. The potential reasons for the less prominent effects are not known but can be related to the timing of the treatment and the nature of the chemicals. Importantly, there are some reports on humans regarding TG effects of the environment (for examples, see Section Diethylstilbestrol and (Titus-Ernstoff et al. 2006).

Additional key questions or challenges related to studying the epigenetic effects of EDCs in the ovary include the following:

  • 1 Follicular stage–specific effects in the ovary: Tissues in the body are composed of different cell types, differentiations of which are believed to be primarily due to epigenetic modifications specific to a given cell type. Therefore, the epigenetic analysis in a whole tissue consisting of various cell types is considered to be problematic. An additional complexity in the post-natal ovary is the presence of the follicles at different stages of their development. When epigenetic analysis in the ovary is being considered, not only should different cell types be examined individually but also these cell types should be separated based on the follicle stage from which they are obtained.
  • 2 EDC mixture effects: Most studies examining the effects of EDCs in the ovary as well as in other organs are conducted with single chemicals. However, individuals are normally exposed to mixture of these chemicals in daily life. Therefore, studies that are examining the effects of mixtures with various approaches have been intensifying. These studies in other organs have shown that when chemicals are combined, their impact increases synergistically – much more dramatically than additive effects of individual compounds (Jacobsen et al. 2010; Rider et al. 2010). The mixture studies further show the complexity of the potential harmful effects of EDCs as mixtures (Kortenkamp et al. 2007).
  • 3 Epigenome-wide effects: The genome is relatively more resistant to the effects of environmental factors that individuals are likely to encounter daily, such as EDCs. In contrast, the epigenome can be vulnerable to such effects. In addition, as compared to the primary DNA sequence, the epigenome is much complex and dynamic. Furthermore, the importance of epigenome in the DoHAD is widely accepted. Therefore, a comprehensive understanding of the effects of the EDCs requires studying not only the methylome but also whole epigenome (including histone modifications and ncRNAs, especially miRNA). Such a broader and more precise understanding of epigenetic effects of EDCs will enable the development of effective prevention and therapy.

Acknowledgements

The authors thank Dr. Kathy Manger for her assistance in the preparation of this manuscript. The research results from our laboratory described in this review are supported in part by NIEHS grants ES013854 and ES017059 and the NIEHS-sponsored UMDNJ Center for Environmental Exposures and Disease grant NIEHS ES005022. Dr. Elif Oruc, visiting scholar from Cukurova University, Turkey, is supported in part by TUBITAK-BIDEB International Postdoctoral Research Scholarship Program.

Conflicts of interest

None of the authors have any conflicts of interest to declare.

Ancillary