There is growing interest in the possible health threat posed by endocrine-disrupting chemicals (EDCs), which are substances in our environment, food, and consumer products that interfere with hormone biosynthesis, metabolism, or action resulting in a deviation from normal homeostatic control or reproduction. In this first Scientific Statement of The Endocrine Society, we present the evidence that endocrine disruptors have effects on male and female reproduction, breast development and cancer, prostate cancer, neuroendocrinology, thyroid, metabolism and obesity, and cardiovascular endocrinology. Results from animal models, human clinical observations, and epidemiological studies converge to implicate EDCs as a significant concern to public health. The mechanisms of EDCs involve divergent pathways including (but not limited to) estrogenic, antiandrogenic, thyroid, peroxisome proliferator-activated receptor γ, retinoid, and actions through other nuclear receptors; steroidogenic enzymes; neurotransmitter receptors and systems; and many other pathways that are highly conserved in wildlife and humans, and which can be modeled in laboratory in vitro and in vivo models. Furthermore, EDCs represent a broad class of molecules such as organochlorinated pesticides and industrial chemicals, plastics and plasticizers, fuels, and many other chemicals that are present in the environment or are in widespread use. We make a number of recommendations to increase understanding of effects of EDCs, including enhancing increased basic and clinical research, invoking the precautionary principle, and advocating involvement of individual and scientific society stakeholders in communicating and implementing changes in public policy and awareness.
I. General Introduction to Endocrine Disruption
An endocrine-disrupting compound was defined by the U.S. Environmental Protection Agency (EPA) as “an exogenous agent that interferes with synthesis, secretion, transport, metabolism, binding action, or elimination of natural blood-borne hormones that are present in the body and are responsible for homeostasis, reproduction, and developmental process.” Our understanding of the mechanisms by which endocrine disruptors exert their effect has grown. Endocrine-disrupting chemicals (EDCs) were originally thought to exert actions primarily through nuclear hormone receptors, including estrogen receptors (ERs), androgen receptors (ARs), progesterone receptors, thyroid receptors (TRs), and retinoid receptors, among others. Today, basic scientific research shows that the mechanisms are much broader than originally recognized. Thus, endocrine disruptors act via nuclear receptors, nonnuclear steroid hormone receptors (e.g., membrane ERs), nonsteroid receptors (e.g., neurotransmitter receptors such as the serotonin receptor, dopamine receptor, norepinephrine receptor), orphan receptors [e.g., aryl hydrocarbon receptor (AhR)—an orphan receptor], enzymatic pathways involved in steroid biosynthesis and/or metabolism, and numerous other mechanisms that converge upon endocrine and reproductive systems. Thus, from a physiological perspective, an endocrine-disrupting substance is a compound, either natural or synthetic, which, through environmental or inappropriate developmental exposures, alters the hormonal and homeostatic systems that enable the organism to communicate with and respond to its environment.
The group of molecules identified as endocrine disruptors is highly heterogeneous and includes synthetic chemicals used as industrial solvents/lubricants and their byproducts [polychlorinated biphenyls (PCBs), polybrominated biphenyls (PBBs), dioxins], plastics [bisphenol A (BPA)], plasticizers (phthalates), pesticides [methoxychlor, chlorpyrifos, dichlorodiphenyltrichloroethane (DDT)], fungicides (vinclozolin), and pharmaceutical agents [diethylstilbestrol (DES)].
Natural chemicals found in human and animal food (e.g., phytoestrogens, including genistein and coumestrol) can also act as endocrine disruptors. These substances, whereas generally thought to have relatively low binding affinity to ERs, are widely consumed and are components of infant formula (1,2). A recent study reported that urinary concentrations of the phytoestrogens genistein and daidzein were about 500-fold higher in infants fed soy formula compared with those fed cow’s milk formula (3). Therefore, the potential for endocrine disruption by phytoestrogens needs to be considered.
A challenge to the field of endocrine disruption is that these substances are diverse and may not appear to share any structural similarity other than usually being small molecular mass (<1000 Daltons) compounds. Thus, it is difficult to predict whether a compound may or may not exert endocrine-disrupting actions. Nevertheless, in very broad terms, EDCs such as dioxins, PCBs, PBBs, and pesticides often contain halogen group substitutions by chlorine and bromine. They often have a phenolic moiety that is thought to mimic natural steroid hormones and enable EDCs to interact with steroid hormone receptors as analogs or antagonists. Even heavy metals and metalloids may have estrogenic activity, suggesting that these compounds are EDCs as well as more generalized toxicants. Several classes of EDCs act as antiandrogens and as thyroid hormone receptor agonists or antagonists, and more recently, androgenic EDCs have been identified.
The sources of exposure to EDCs are diverse and vary widely around the world. The situation is constantly evolving because some EDCs were banned decades ago and others more recently, with significant differences between countries. In this respect, migrating people provide a model to study cessation and/or onset of exposure depending on contamination of the original and new milieus. There are also several historical examples of toxic spills or contamination from PCBs and dioxins that show a direct causal relationship between a chemical and the manifestation of an endocrine or reproductive dysfunction in humans and wildlife. However, these types of single exposures are not representative of more common widespread persistent exposure to a broad mix of indoor and outdoor chemicals and contaminants. Industrialized areas are typically characterized by contamination from a wide range of industrial chemicals that may leach into soil and groundwater. These complex mixtures enter the food chain and accumulate in animals higher up the food chain such as humans, American bald eagles, polar bears, and other predatory animals. Exposure occurs through drinking contaminated water, breathing contaminated air, ingesting food, or contacting contaminated soil. People who work with pesticides, fungicides, and industrial chemicals are at particularly high risk for exposure and thus for developing a reproductive or endocrine abnormality.
Some EDCs were designed to have long half-lives; this was beneficial for their industrial use, but it has turned out to be quite detrimental to wildlife and humans. Because these substances do not decay easily, they may not be metabolized, or they may be metabolized or broken down into more toxic compounds than the parent molecule; even substances that were banned decades ago remain in high levels in the environment, and they can be detected as part of the body burden of virtually every tested individual animal or human (4,5). In fact, some endocrine disruptors are detectable in so-called “pristine” environments at remote distances from the site they were produced, used, or released due to water and air currents and via migratory animals that spend part of their life in a contaminated area, to become incorporated into the food chain in an otherwise uncontaminated region. Others, such as BPA, may not be as persistent [although recent evidence (e.g., Ref. 6) suggests longer half-lives) but are so widespread in their use that there is prevalent human exposure.
A. Important issues in endocrine disruption
A number of issues have proven to be key to a full understanding of mechanisms of action and consequences of exposure to EDCs. These have been reviewed previously in detail (7), and several of them are listed here in brief.
1. Age at exposure
Exposure of an adult to an EDC may have very different consequences from exposure to a developing fetus or infant. In fact, the field of endocrine disruption has embraced the terminology “the fetal basis of adult disease” (8) to describe observations that the environment of a developing organism, which includes the maternal environment (eutherian mammals), the egg (other vertebrates), and the external environment, interacts with the individual’s genes to determine the propensity of that individual to develop a disease or dysfunction later in life. In this Scientific Statement, we extend this concept beyond the fetal period to the early postnatal developmental period when organs continue to undergo substantial development. Thus, we will henceforward use the terminology “the developmental basis of adult disease.”
2. Latency from exposure
The developmental basis of adult disease also has implicit in its name the concept that there is a lag between the time of exposure and the manifestation of a disorder. In other words, consequences of developmental exposure may not be immediately apparent early in life but may be manifested in adulthood or during aging.
3. Importance of mixtures
If individuals and populations are exposed to an EDC, it is likely that other environmental pollutants are involved because contamination of environments is rarely due to a single compound. Furthermore, effects of different classes of EDCs may be additive or even synergistic (9).
4. Nontraditional dose-response dynamics
There are several properties of EDCs that have caused controversy. First, even infinitesimally low levels of exposure—indeed, any level of exposure at all—may cause endocrine or reproductive abnormalities, particularly if exposure occurs during a critical developmental window (10). Surprisingly, low doses may even exert more potent effects than higher doses. Second, EDCs may exert nontraditional dose-response curves, such as inverted-U or U-shaped curves (11). Both of these concepts have been known for hormone and neurotransmitter actions, but only in the past decade have they begun to be appreciated for EDCs.
5. Transgenerational, epigenetic effects
EDCs may affect not only the exposed individual but also the children and subsequent generations. Recent evidence suggests that the mechanism of transmission may in some cases involve the germline (12) and may be nongenomic. That is, effects may be transmitted not due to mutation of the DNA sequence, but rather through modifications to factors that regulate gene expression such as DNA methylation and histone acetylation.
B. The role of endocrinologists in discerning effects of EDCs
The field of endocrine disruption has particular pertinence to endocrinologists. In general, persistent endocrine disruptors have low water solubility and extremely high lipid solubility, leading to their bioaccumulation in adipose tissue. The properties of these substances are particularly well suited for study by endocrinologists because they so often activate or antagonize hormone receptors. There is no endocrine system that is immune to these substances, because of the shared properties of the chemicals and the similarities of the receptors (13) and enzymes involved in the synthesis, release, and degradation of hormones (Fig 1). Therefore, the role of this Scientific Statement is to provide perspectives on representative outcomes of exposures to endocrine disruptors and evidence for their effects in wildlife, laboratory animals, and humans.
Model of the endocrine systems targeted by endocrine-disrupting chemicals as discussed in this article. This figure demonstrates that all hormone-sensitive physiological systems are vulnerable to EDCs, including brain and hypothalamic neuroendocrine systems;...
II. Overview of Endocrine Disruption and Reproductive Health from a Clinical Perspective
A. Clinical aspects of endocrine disruption in humans
For a clinician taking care of an individual patient, there are numerous challenges in ascertaining EDC involvement in a particular disorder. Each person has unique exposure to a variety of both known and unknown EDCs. Individual differences in metabolism and body composition will create considerable variability in the half-life and persistence of EDCs, as well as their degradation in body fluids and tissues. Susceptibility to EDCs may vary according to genetic polymorphisms. In addition, human disorders are more likely the result of chronic exposure to low amounts of mixtures of EDCs. The latency between exposure to EDCs and occurrence of clinical disorders creates further challenges when one attempts to establish a relationship at the level of a given individual.
Epidemiological studies at the level of populations in a country or a region are crucial to alert researchers about geographical or secular trends in prevalence of disorders pointing to possible environmental factors. Registries with data on particular diseases or cell/organ donors may provide valuable contributions. For instance, the observation of adverse trends in male reproductive health together with declining sperm count in Denmark and other countries has led to the hypothesis of environmental contaminants being harmful to reproduction (14). Unfortunately, it is virtually impossible to make direct links between such epidemiological observations and exposure to given chemicals. Regional differences in certain reproductive disorders (infertility, cancer) that may be tied to contamination by compounds used locally such as in agriculture, industrial accident, or product misuse/abuse in subpopulations can also be informative (14,15). Finally, a comparison of disorders before and after migration to a new environment may reveal exposure and/or susceptibility to exposure to EDCs (16).
As already mentioned, a critical concern is the potential lag between exposure to EDCs and the manifestation of a clinical disorder. In humans, this period may be years or decades. In the case of reproduction, infertility cannot be assessed until the exposed individual has attained a certain age, again resulting in a lag between early exposure and manifestation of a dysfunction. Delayed or early puberty cannot be assessed until this event actually takes place, although timing of puberty could involve programming many years earlier during fetal life. Interestingly, an increased likelihood of early puberty was observed in subjects born with intrauterine growth retardation (IUGR) (17,18), suggesting a link between developmental programming and reproductive maturation. As discussed below, development of vaginal adenocarcinoma in women exposed fetally to DES (19) and the association of carcinoma in situ in the fetal testis with the development of testicular cancer in adulthood (14,20) are examples of links between the fetal environment and the occurrence of adult disease.
The timing of exposure is key to human disease because there are critical developmental periods during which there may be increased susceptibility to environmental endocrine disruptors. In those cases in which disruption is directed toward programming of a function, e.g., reproductive health, this may interfere with early life organization, followed by a latent period, after which the function becomes activated and the dysfunction can become obvious. For reproductive function in both humans and animals, fetal life is most vulnerable because there are rapid structural and functional events. The roles of sex steroids in sexual differentiation and thyroid hormones in brain development are of paramount importance at that time. Early postnatal life is also a time when maturation is still rapid (e.g., the central nervous system undergoes significant development at this time, including the hypothalamus which controls reproduction; see Section VII). The organization of the neuroendocrine control of reproduction is not completed at birth and remains sensitive to the interaction of steroids or EDCs neonatally such as has been shown for the control of ovulation in rodents. Breast or formula feeding could be of particular significance due to the capacity of human milk to concentrate EDCs in the former and the potential high intake of phytoestrogens in soy milk and/or plasticizers in formula-containing cans in the latter. It is apparent that the developmental basis of adult disease is an important concept for understanding endocrine disruption of reproductive function in humans.
B. Clinical dimorphism of EDCs on male and female reproduction
A spectrum of disorders throughout life, some of which are sexually dimorphic, can be related to endocrine disruption (Table 1). Male sexual differentiation is androgen-dependent (and potentially estrogen-dependent), whereas female differentiation occurs largely independently of estrogens and androgens. Therefore, it is expected that different disorders are seen in males and females as a result of EDC effects that overall mimic estrogens and/or antagonize androgens.
Disorders of the human reproductive system possibly involving EDCs in their pathogenesis: A sexually dimorphic life cycle perspective
In the male (Table 2), cryptorchidism, hypospadias, oligospermia, and testicular cancer have been proposed to be linked as the testicular dysgenesis syndrome (TDS) arising from disturbed prenatal testicular development (14,21). Such links are important because they could mean that several disorders occur at different periods throughout life in a single individual as a result of exposure to a given EDC (or mixture) at a particular period. The epidemiological data relating TDS with environmental disruptors are indirect, and we still lack direct evidence of EDC involvement in the pathogenesis of TDS in humans (see Section V). In the rodent, however, a TDS-like condition can be observed after fetal exposure to phthalates (20), and the reduced anogenital distance observed in the rat (22) was observed in a recent epidemiological study on human male newborns (23). Several studies have shown a strong association of low birth weight with hypospadias and cryptorchidism, suggesting that they have a common determinant (15).
Effects of some specific EDCs on the male reproductive system
Other pathologies in males are linked to EDC exposure. Prostate hyperplasia has been described after exposure to BPA (24). In adolescence, boys born with IUGR have small testes and elevated serum FSH, together with low inhibin B levels (18) that could be related to some of the TDS disorders. Divergent data have been reported on effects of EDCs on pubertal timing in the male (25).
In the female (Table 3), premature thelarche has been reported in girls exposed to phthalates (26), although these data need to be replicated. Sexual precocity presumably of peripheral origin initially and secondarily central could be related to exposure to the insecticide DDT in girls migrating for international adoption (17). A neuroendocrine mechanism is suggested by experiments in a rodent model (27) (see Section VII). An association of premature pubarche and ovulatory disorders with EDCs is suggested indirectly by links with IUGR at birth and metabolic syndrome in adulthood (18).
Effects of some specific EDCs on the female reproductive system
In the adult female, the first evidence of endocrine disruption was provided almost 40 yr ago through observations of uncommon vaginal adenocarcinoma in daughters born 15–22 yr earlier to women treated with the potent synthetic estrogen DES during pregnancy (19). Subsequently, DES effects and mechanisms have been substantiated in animal models (28). Thus, robust clinical observations together with experimental data support the causal role of DES in female reproductive disorders. However, the link between disorders such as premature pubarche and EDCs is so far indirect and weak, based on epidemiological association with both IUGR and ovulatory disorders. The implications of EDCs have been proposed in other disorders of the female reproductive system, including disorders of ovulation and lactation, benign breast disease, breast cancer, endometriosis, and uterine fibroids (29,30,31,32).
C. Experimental and clinical evidence of EDCs and potential mechanisms
In Tables 2 and 3, some experimental and clinical observations of disturbed reproductive systems are listed for selected EDCs. The evidence from human epidemiological studies is partial and indirect (see Section V). Mechanistic studies are ethically and practically very limited in humans and have to rely on data obtained using animal experiments (in vivo and in vitro models), although these models can have limitations. Clinical and experimental studies correlate DES effects quite convincingly in both sexes. In the male, rodent studies using phthalates and, to a lesser extent, PCBs model TDS entirely or partly. In the female, some rodent studies are consistent with DDT/dichlorodiphenyldichloroethylene (DDE) involvement in sexual precocity.
The following considerations emphasize some of the concepts emerging from the available data.
There may be transgenerational effects of EDCs due to overt mutation or to more subtle modifications of gene expression independent of mutation (i.e., epigenetic effects). Epigenetic effects of EDCs include context-dependent transmission (e.g., the causal factor persists across generations; Ref. 33) or germline-dependent mechanisms (i.e., the germline itself is affected; Refs. 12, 34, and 35). An example of germline transmission of an epigenetically modified trait is shown in a rat model for the fungicide vinclozolin and is manifested by a higher likelihood of metabolic disorders, tumors, and reproductive dysfunctions in the next four generations (12,34,35,36,37,38). In the case of DES, there are both human and experimental observations indicating heritability (19,28,39).
2. Diversity and complexity of mechanisms
EDCs often act via more than one mechanism. Some EDCs have mixed steroidal properties: for example, a single EDC may be both estrogenic and antiandrogenic. EDCs may be broken down or metabolized to generate subproducts with different properties. For instance, the estrogen agonist DDT is metabolized into the androgen antagonist DDE (27). The balance between estrogenic and androgenic properties of EDCs can be biologically significant because reproduction of both sexes involves an interplay of androgens and estrogens. In humans, early breast development occurs in girls with a highly active variant of CYP3A4, a cytochrome p450 enzyme involved in inactivating testosterone (40), and premature thelarche occurs with antiandrogenic phthalates (25). Similar androgen-estrogen interactions have been reported in DES-treated rats in which reduced androgen secretion or action sensitized the animals to the estrogenic effects of DES (41). Moreover, many organs are targeted by sex steroids and are thereby vulnerable to endocrine disruption, including the hypothalamic-pituitary-gonadal system, breast, uterus, cervix, vagina, brain, and nonreproductive tissues such as bone, muscle, and skin (Fig. 1). In the case of humans, a peripheral effect in the reproductive system (e.g., breast development) can result from direct EDC effects (peripheral puberty) and/or endogenous estrogen increase through premature neuroendocrine maturation (central puberty) (17,27), but these may be difficult to distinguish. For instance, EDC effects can involve altered ERα expression in hypothalamus (42) and epididymis (43) or uterus (44). Along with the direct influence of EDCs on estrogen or androgen actions, they can affect endogenous steroid production through negative and positive feedback, effects that may differ depending on developmental stage. Also, there are multiple levels of interactions with steroid action (receptor or promoters), synthesis (e.g., aromatase stimulation by atrazine), and metabolism [e.g., sulfotransferase (45)]. Finally, there are coexisting mechanisms not directly mediated at the hypothalamic-pituitary-gonadal (HPG) system. For instance, reproductive dysfunction can result from thyroid disruption (46) or nonspecific interference of reduced energy intake (47).
3. Limits of translational models
The in vivo animal models may be difficult to extrapolate to humans for several reasons, including species differences in ontogeny of reproductive system and functions, differences in metabolism of sex steroids, difficulty in estimating exposure to mixtures, and variable body burdens. As already mentioned, exposure to EDCs is complex. For example, mixtures are likely to be the usual form of exposure to EDCs, but they are difficult to approximate in experimental models. Moreover, the effects may not be additive; nevertheless, a combination of low doses of substances that individually are inactive may result in a biological perturbation (48). Despite these limitations, considering the substantial conservation of endocrine and reproductive processes across species, it is certainly reasonable to use animal models for understanding human processes, as long as these potential differences are taken into account.
III. Clinical and Translational Impacts of EDCs on Female Reproduction
A. Introduction to female reproductive development and function
Development and function of the female reproductive tract depends on coordinated biological processes that, if altered by endogenous or exogenous factors during critical periods of development or during different life stage, could have significantly adverse effects on women’s health and reproductive function and outcomes. For example, the full complement of cell types in the human ovary depends on successful germ cell migration from the yolk sac during the first trimester and differentiation into oocytes with associated somatic cells to form the functional unit of the primordial follicle by the second to third trimesters of gestation. Factors that interfere with germ cell migration or follicle formation can result in abnormal functioning of this tissue with significant reproductive consequences. Also, the oocyte is arrested in the diplotene stage of late prophase until meitic divisions occur beginning at puberty (meiosis I) and after fertilization (meiosis II), and abnormalities in these processes can have a profound impact on reproductive outcomes, such as aneuploidy, premature ovarian failure (POF), and miscarriage. In addition, whereas Mullerian tract formation begins at 8 wk gestation with fusion of the Mullerian ducts and subsequent differentiation into the uterus (endometrium, myometrium), cervix, and upper vagina, uterine differentiation with regard to formation of luminal epithelium, glandular epithelium, and stromal components is mostly a postnatal event, with functionality of response to steroid hormones beginning at puberty. Interference with these processes can predispose women to infertility, ectopic gestation, poor pregnancy outcomes, and other reproductive disorders that may be programmed during development (e.g., endometriosis, uterine fibroids). Thus, abnormal development or alterations at other times in the life cycle can alter anatomy and functionality of the female reproductive tract and thus can alter the reproductive potential of affected individuals and their offspring.
Most female reproductive disorders are well described with regard to clinical presentation, histological evaluation of involved tissues where applicable, and diagnostic classification. However, whereas few are polygenic inherited traits and some are due to infections, the pathogenesis of the vast majority of female reproductive disorders is not well understood. This has hindered a preventive strategy to their development and/or exacerbation, and in some cases limited the development of effective therapies for symptoms and associated morbidities.
A key question arises as to whether EDCs contribute to the development of female reproductive disorders, particularly those occurring during a critical window of susceptibility: in utero, neonatally, in childhood, during puberty, and during adulthood. There are increasing data from wildlife studies and laboratory studies with rodents, ungulates, and nonhuman primates that support a role of EDCs in the pathogenesis of several female reproductive disorders, including polycystic ovarian syndrome, aneuploidy, POF, reproductive tract anomalies, uterine fibroids, endometriosis, and ectopic gestation (for reviews, see Refs. 29 and 49,50,51,52,53,54; also see Table 4). Many of the mechanisms are understood and, moreover, are conserved between animals and humans. Herein, we describe some of the clinical implications of these associations.
Female reproductive disorders and their possible relationships to EDCs: Some experimental and human data
B. Polycystic ovarian syndrome (PCOS)
PCOS is a heterogeneous syndrome characterized by persistent anovulation, oligo- or amenorrhea, and hyperandrogenism in the absence of thyroid, pituitary, and/or adrenal disease (55,56,57). At the level of the ovary, there is recruitment and growth of follicles to the small antral stage, without selection of a dominant, preovulatory follicle, leading to accumulation of multiple, small, antral follicles (58). Hyperfunctioning of the theca and relative hypofunctioning of the granulosa cells accompany the acyclicity of the syndrome. Many, but not all women with PCOS have relatively high circulating levels of LH, compared with FSH, believed to be due to insensitivity to steroid hormone feedback. However, this does not fully account for the observed increase in thecal androgen production or the relative quiescence and sometimes frank FSH resistance of the granulosa cells. This complex disorder likely has its origins both within and outside the hypothalamic-pituitary-ovarian axis, and metabolic, neuroendocrine, and other endocrine regulators likely contribute to its manifestation. Obesity and insulin resistance occur in about 50% of women with PCOS, and obese women have a 12% risk of having PCOS (59). PCOS has multiple physiological processes (e.g., neuroendocrine functioning and feedback mechanisms, ovarian steroidogenesis, insulin resistance, and obesity) that are regulated by hormonal and metabolic parameters. Hence, endocrine disruption by environmental chemicals may indeed contribute to the pathogenesis of PCOS.
In sheep and rhesus monkeys, prenatal exposure to high levels of testosterone results in fetal programming of PCOS traits (60). Specifically, high levels of testosterone exposure at gestational d 40–60 and 100–115 result in rhesus monkey females who, in adulthood, have anovulatory infertility, hypersecretion of LH, elevated circulating levels of testosterone, neuroendocrine feedback defects, central adiposity and compensatory insulin resistance, and polycystic ovaries with ovarian hyperandrogenism and follicular arrest in adulthood (60,61). In the sheep model, a similar PCOS phenotype, along with IUGR and compensatory catch-up growth after birth, derives from prenatal exposure to exogenous testosterone (60,62). In rhesus monkey and sheep, unlike rodents, follicular differentiation is completed during fetal life. Thus, it is plausible that in utero exposure of human female fetuses to androgen-like EDCs could result in PCOS in adulthood, along with associated metabolic disorders. Very recent evidence for androgenic properties of personal-care products such as triclocarban (63) add to the possibility of environmental androgens, although a connection to PCOS has not yet been drawn.
There are numerous candidate genes associated with predisposition to developing PCOS in women (57,64), and how and if these interact with prenatal androgen-like factors to promote the PCOS phenotype in women has not been determined. Nonetheless, PCOS is a debilitating disorder in women, occurring in 6.6% of the reproductive-age population (65,66,67); it is a leading cause of subfertility and is associated with increased lifetime risks for cardiovascular disease and type II diabetes (55). In addition to these clinical impacts on patients, the cost to the health care system for PCOS diagnosis and treatment is substantial, totaling in 2004 about $4.4 billion in the United States alone (68). These facts underscore the need to understand potential EDC contributions to the development of PCOS in an effort to minimize such exposures and maximize prevention.
Other pathways may be involved in endocrine disruption of PCOS. Women with PCOS have higher levels of the EDC BPA (69), and increased testosterone in these women is consistent with decreased clearance of BPA (70). Although adult exposures do not necessarily imply earlier exposures in life, especially with EDCs of relatively short half-lives, there are data demonstrating nearly 5-fold higher levels of BPA in amniotic fluid compared with other body fluids, suggesting significant prenatal exposure (71). Although a cause and effect of BPA and PCOS have not been demonstrated definitively, the biological plausibility is interesting and worthy of further consideration.
C. Premature ovarian failure, decreased ovarian reserve, aneuploidy, granulosa steroidogenesis
POF (cessation of proper ovarian function before the age of 40) occurs in about 1% of reproductive-age women (72). Although in some cases the causation is known, for the vast majority of women with POF this is not the case, and there are stages of susceptibility during organogenesis and adult exposures that could contribute to POF.
Because the total ovarian follicle complement is established before birth in humans (73), anything that interferes with this, resulting in a decreased ovarian follicle resting pool, can result in POF. For example, disruption of germ cell migration from the genital ridge into the developing gonad results in ovarian dysgenesis. The resting pool undergoes a baseline level of apoptosis, and TNF-α, Fas ligand, and androgens stimulate this in the resting pool, as well as in the growing pool (74). Also, once a cohort of follicles is recruited during a given cycle in women, survival factors (FSH, estradiol, and growth factors, e.g., IGFs) are important for escape from apoptosis of the dominant follicle. Recent data in the mouse show that selective activation of the K-ras pathway in the oocyte results in rapid follicular development and depletion (75). Interestingly, adult and in utero exposures of mice to BPA have resulted in damage to oocytes (76,77). Specifically, adult exposures result in abnormalities in alignment of chromosomes on the meiotic spindle and aneuploidy, which, while not leading to ovarian senescence, does lead to aneuploid gametes and offspring (76). However, BPA given to pregnant dams during midgestation affects the developing ovary with resulting abnormalities in meiotic prophase, including synaptic defects, and mature animals exposed in utero have an increase in aneuploid oocytes and embryos (77). Such alterations also lead to cell cycle arrest and oocyte death, thus depleting the complement of normal oocytes (77). Currently, there are no data on in utero or adult exposure to BPA and aneuploidy in humans, but the possibility that there are parallels is compelling.
Interestingly, mice exposed in utero to DES, between d 9–16 gestation, have a dose-dependent decrease in reproductive capacity, including decreased numbers of litters and litter size and decreased numbers of oocytes (30%) ovulated in response to gonadotropin stimulation with all oocytes degenerating in the DES-exposed group, as well as numerous reproductive tract anatomic abnormalities (78). In women with in utero exposure to DES, Hatch et al. (79) reported an earlier age of menopause between the 43–55 yr olds, and the average age of menopause was 52.2 yr in unexposed women and 51.5 yr in exposed women. The effect of DES increased with cumulative doses and was highest in a cohort of highest in utero exposure during the 1950s (79). These observations are consistent with a smaller follicle pool and fewer oocytes ovulated, as in DES-exposed mice after ovulation induction (78).
Of interest are human data that demonstrate unequivocally that adult exposure in women to cigarette smoke results in decreased fecundity, decreased success rates in in vitro fertilization (IVF), decreased ovarian reserve (higher basal cycle d 3 FSH and stimulated parameters), earlier menopause by 1–4 yr, and an increased miscarriage rate (80,81). The mechanism appears to be mediated by the AhR-mediated apoptosis of oocytes, with accelerated loss of ovarian follicles. Interestingly, exposure of rats to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) in utero and through the end of reproductive life results in a dose-dependent onset of premature reproductive senescence, likely due to direct effects on ovarian function (82).
Thus, whereas POF may occur in a relatively small percentage of the population, there are several alarming signals that should not be ignored. For example, the age group with the fastest growing rate of involuntary subfertility is 15- to 24-yr-old women (83). Also, the known effects of environmental contaminants on oocyte survival, aneuploidy, decreased ovarian reserve, and infertility described above underscore how much at risk the population may be for reproductive compromise.
With regard to ovarian granulosa steroidogenesis, several EDCs have effects on this process (84). For example, TCDD (10 ppm) decreases FSH-stimulated LH receptor mRNA expression and half-life in cultured granulosa (85). DDE increases vascular endothelial growth factor and IGF-I expression in luteinized granulosa from IVF patients, suggesting a contribution to impaired steroidogenesis and perhaps infertility (86). Recently, Kwintkiewicz and Giudice (87,88) have shown, in preliminary studies, that BPA decreases proliferation and FSH-induced aromatase expression via activation of peroxisome proliferator-activated receptor γ (PPAR-γ) and increases IGF-I and IGF receptor type I in human granulosa-like tumor cells and luteinized human granulosa from IVF subjects. These data suggest that EDCs may have local effects on ovarian function in adult women.
D. Reproductive tract anomalies
Disruption of female reproductive tract development by the EDC DES is well documented (89). A characteristic T-shaped uterus, abnormal oviductal anatomy and function, and abnormal cervical anatomy are characteristic of this in utero exposure, observed in adulthood (90), as well as in female fetuses and neonates exposed in utero to DES (91). Some of these effects are believed to occur through ERα (92) and abnormal regulation of Hox genes (93,94). Clinically, an increased risk of ectopic pregnancy, preterm delivery, miscarriage, and infertility all point to the devastating effect an endocrine disruptor may have on female fertility and reproductive health (89). It is certainly plausible that other EDCs with similar actions as DES could result in some cases of unexplained infertility, ectopic pregnancies, miscarriages, and premature deliveries. Although another major health consequence of DES exposure in utero was development of rare vaginal cancer in DES daughters, this may be an extreme response to the dosage of DES or specific to pathways activated by DES itself. Other EDCs may not result in these effects, although they may contribute to the fertility and pregnancy compromises cited above. Of utmost importance clinically is the awareness of DES exposure (and perhaps other EDC exposures) and appropriate physical exam, possible colposcopy of the vagina/cervix, cervical and vaginal cytology annually, and careful monitoring for fertility potential and during pregnancy for ectopic gestation and preterm delivery (89,95).
E. Uterine leiomyomas
Uterine leiomyomas (fibroids) are benign smooth muscle tumors of the myometrium that can cause morbidity for women, including menorrhagia, abdominal pain, pelvic prolapse, and infertility and miscarriage (96). They are the most common tumor of the reproductive tract in women and comprise the leading cause for hysterectomy and the second leading cause of inpatient surgery in the United States, with health care costs exceeding $2 billion in 2004 (97). The prevalence rate of uterine leiomyomas is approximately 25–50%, with a preponderance occurring in African-American women (97). The greatest risk factor in adult women is prolonged exposure to unopposed estrogen. Whether in utero exposure to DES increases a woman’s lifetime risk of developing uterine fibroids is controversial, as the method to detect fibroids in two different studies influenced the outcome (98,99). Specifically, in a study of 1731 women exposed to DES and 848 matched unexposed controls, no association was found (P = 0.68) when histological confirmation after myomectomy or hysterectomy was used to document uterine fibroids (98). In contrast, when ultrasound was used to determine the presence of fibroids in DES-exposed vs. DES-unexposed women, a significant relationship was found (odds ratio, 2.4; 95% confidence interval, 1.1–5.4) in DES-exposed women and uterine fibroids (99). However, there are strong animal data to support development of uterine fibroids in adulthood after in utero exposure to EDCs, especially DES (for reviews, see Refs. 49, 50, and 52). Newbold et al. (100) reported that CD-1 mice develop uterine leiomyomas if exposed in utero or neonatally to DES, whereas unexposed mice do not. Furthermore, the Eker rat, which has a germ-line mutation in the rat homolog of the tuberus sclerosis complex 2 tumor suppressor gene, spontaneously develops uterine leiomyomas (101). The number, size, and growth rate of the fibroids increase significantly when the rat is exposed to DES on postnatal d 3–5 and 10–12, but not 17–19 (102), an effect that can be diminished with prior oophorectomy (102). These data overall strongly suggest developmental programming and gene-environment interactions for the increased risk of uterine lyomyomas in this rat model (103). In addition to mice, the Eker rat, and some dogs, the Baltic gray seal that has high organochlorine body burden also develops uterine leiomyomas (104). As with most environmental causes of abnormalities in the reproductive tract (and other tissues and organs), direct cause and effect relationships are difficult to establish. However, as in many of the other abnormalities in this Scientific Statement, the likelihood of such a relationship is plausible.
Endometriosis is an estrogen-dependent gynecological disorder associated with pelvic pain and infertility. It occurs in 6–10% of women and up to 50% of women with pelvic pain and infertility. In 2002, the total health care costs estimated in the United States for diagnosis and treatment of endometriosis totaled approximately $22 billion (105). There are suggestive animal data of adult exposure to EDCs and development of or exacerbation of existing disease, and there is evidence that in utero exposure in humans to DES results in an increased relative risk = 1.9 (95% confidence interval, 1.2–2.8) (106). Most striking are the observations of rhesus monkeys administered different doses of TCDD and their subsequent development of endometriosis (107,108). Although this study had low sample size and confounding variables that brought into question the relationship between endometriosis and TCDD (49,52,109), another study revealed that adult exposure of cynomolgus monkey to TCDD promotes growth and survival of endometriosis implants (110), indicating that this EDC is involved in the progression, if not pathogenesis, of this disorder. Similar data were obtained in rodent models of endometriosis in which human endometrium is transplanted into mouse and rat peritoneum, and the established lesions grew larger when animals were exposed to TCDD in utero and as adults (111,112), underscoring the estrogen (and EDC) dependence of this disorder.
There are also correlative findings of phthalate levels in plasma and endometriosis. For example, Cobellis et al. (113) found high plasma concentrations of di-(2-ethylhexyl)-phthalate in women with endometriosis, and an association of phthalate esters with endometriosis was found among Indian women (114). Thus, the evidence is accumulating of correlations between EDCs in the circulation of women with endometriosis, although a cause-and-effect relationship has yet to be established, which is not uncommon in reproductive environmental toxicity.
Endometriosis is believed to be due to retrograde menstruation and transplantation of endometrial fragments and cells into the peritoneal cavity. Because nearly all women have retrograde menstruation but relatively few have endometriosis, the disorder is also believed to involve a dysfunctional immune response, i.e., activated macrophages in the peritoneal cavity with robust secretion of inflammatory cytokines but without clearance of disease. An interesting model of early-life immune insult and developmental immunotoxicity suggests that in utero exposures to specific insults may reprogram the immune system, resulting in disorders such as chronic fatigue syndrome, cancer, and autoimmune disorders. Whether this has any relevance to the development or progression of endometriosis in adult women has not been explored but warrants further evaluation. Interestingly, TCDD and a therapy for endometriosis, danazol, both have effects on the adult immune system, although effects on the developing immune system are not known.
Although the infertility associated with endometriosis for the most part can be treated with advanced reproductive technologies, less success has been achieved with treatment of endometriosis-related pain. Because the pathogenesis of the associated pain is not known with certainty, therapies are empiric and include agents directed to minimize inflammation (nonsteroidal antiinflammatory drugs, danazol), progestins and androgens (to oppose estrogen actions), GnRH analogs (to inhibit gonadotropin secretion and thus ovarian estradiol production), and aromatase inhibitors (to inhibit estradiol synthesis by the ovary and endometriotic lesions), as well as surgical ablation or excision of the disease, when possible. Most of these therapies are effective in up to 50–60% of affected women, with either intolerable side effects (e.g., profound hypoestrogenism) or recurrence of pain (e.g., after surgery) (115). Thus, prevention is key to this disorder, as is understanding the pathogenesis so that therapies for pain can be devised appropriately and administered.
IV. Endocrine Disruptors, Mammary Gland Development, and Breast Cancer
It has been hypothesized that the significant increase of the incidence of breast cancer in the industrialized world observed during the last 50 yr may be due to exposure to hormonally active chemicals, particularly xenoestrogens (116). A similar increase in the incidence of testicular cancer and malformations of the male genital tract and decreased quantity and quality of human sperm have been observed during the same half century, again suggesting a link to the introduction of these chemicals into the environment (117) (see Sections II and V).
A. Windows of vulnerability to carcinogenic agents and “natural” risk factors
The standard risk factors for developing breast cancer include age at menarche, first pregnancy, menopause, lactation, and parity. All of these factors are related to lifetime exposures to ovarian hormones. It is also known that there are developmental periods of enhanced vulnerability (see Section I). For example, sensitivity to radiation is highest during puberty. Additionally, pregnancy increases the risk of breast cancer in the short term (118) and decreases it in the long term (119). More recently, epidemiological studies have revealed that the intrauterine environment may also influence the risk to develop breast cancer later in life. Studies comparing human dizygotic twins and single births revealed that the propensity to breast cancer is enhanced in female twins, and this outcome was attributed to excess estrogen exposure in dizygotic twins during gestation (120).
B. Theories of carcinogenesis
A majority of researchers support the idea that cancer is due to the accumulation of mutations in a cell [the somatic mutation theory (121)]. In contrast, supporters of the theory of developmental origins of adult disease are proposing that changes in the epigenome play a central role in carcinogenesis (see Section VI).
Both the genetic and epigenetic theories of carcinogenesis imply that cancer originates in a cell that has undergone genetic and/or epigenetic changes, which ultimately results in dysregulated cell proliferation (122). Alternatively, the tissue organization field theory postulates that carcinogenesis represents a problem of tissue organization, comparable to organogenesis gone awry, and that proliferation is the default state of all cells (123,124,125). According to this theory, carcinogens, as well as teratogens, would disrupt the normal dynamic interaction of neighboring cells and tissues during early development and throughout adulthood (126).
During postnatal life, the mammary gland undergoes massive architectural changes, comparable to those usually associated with organogenesis. These changes occur in response to alterations in endogenous hormone levels such as those associated with puberty and pregnancy and can be induced experimentally by endocrine manipulation. Many studies of endocrine disruptors have illustrated that developmental exposure to these exogenous hormone mimics can alter normal patterns of tissue organization and hence disrupt stromal-epithelial interactions (127,128). These changes may disturb important regulatory mechanisms and enhance the potential for neoplastic lesions.
C. Susceptibility of the breast during puberty and adulthood
Several epidemiological studies explored the link between exposure to endocrine disruptors and breast cancer incidence. In general, these are case-control studies that usually measure exposure to a single chemical at the time of breast cancer diagnosis. This type of study has produced inconsistent results. Prospective studies that measured exposure several years before cancer diagnosis revealed a positive link between breast cancer and chemical exposure to toxaphene (129) and DDT (130). In particular, a study linked DDT with an increased risk of breast cancer when the exposure was measured before 14 yr of age. This study used samples taken before the banning of DDT for agricultural use and hence represents higher exposures than those measured today. Humans, however, are exposed to a plethora of hormonally active chemicals with different metabolic profiles. Moreover, individuals living in the same area may be exposed to a different mixture of chemicals due to different diets and to migration history. These facts imply that a single chemical cannot be construed as a marker of total exposure. Not surprisingly, one case-control study reported a significant correlation between total xenoestrogen exposure and breast cancer (131).
How xenoestrogen exposure during the period of sexual maturity may result in mammary gland carcinogenesis remains unsolved; this is not surprising because the mechanisms underlying hormonal carcinogenesis are still unknown. One possibility, compatible with all the cancer hypotheses briefly discussed above, is that xenoestrogens may extend the length of the period of ductal growth and alveologenesis during the menstrual cycle. This period is also characterized by proliferative activity in the glandular epithelium. For example, ductal cell proliferation in the breast is maximal from the late follicular phase and throughout the luteal phase, i.e., when endogenous estrogen levels are high (132). The ubiquitous presence of xenoestrogens in foods, their persistence, and their lack of binding to the plasma carrier protein SHBG (127) may result in relatively constant levels in blood. These xenoestrogens would act additively with ovarian estrogens and thus advance by a few days the period of ductal growth. Hence, a small and maintained increase of estrogenic activity during the period of low ovarian output could be sufficient to “promote” carcinogenesis by increasing the number of cells that undergo proliferation menstrual cycle after menstrual cycle, an explanation consistent with the somatic mutation theory. An alternative explanation, consistent with the tissue organization field theory, is that estrogens acting as morphogens would enhance tissue remodeling through stroma epithelium interactions and increase the likelihood of producing alterations of tissue architecture. This notion is supported by data showing that recombination of normal mammary epithelial cells with stroma exposed to carcinogenic agents results in the development of epithelial neoplasias (133) and that conversely, recombination of mammary carcinoma cells with stroma from multiparous animals (which are refractory to carcinogens) results in the normalization of the neoplastic phenotype (126).
D. Susceptibility of the mammary gland during the perinatal period
Direct evidence of prenatal estrogen exposure and breast cancer risk is being gathered from the cohort of women born to mothers treated with DES during pregnancy and is discussed above (see Sections II and III). These women are now reaching the age at which breast cancer becomes more prevalent. In the cohort of these women who are aged 40 yr and older, there is a 2.5-fold increase in the incidence of breast cancer compared with unexposed women of the same age (134,135), suggesting that indeed, prenatal exposure to synthetic estrogens may play an important role in the development of breast neoplasms. Consistent with this, experiments in rats showed that prenatal exposure to DES resulted in increased mammary cancer incidence during adulthood (136,137). These experiments illustrated that rats exposed prenatally to DES and challenged with the chemical carcinogen dimethylbenzanthracene (DMBA) at puberty had a significantly greater incidence of palpable mammary tumors at 10 months of age than animals exposed prenatally to vehicle. In addition, the tumor latency period was shorter in the DES-exposed compared with the vehicle-exposed group (130). Both the epidemiological and experimental data are consistent with the hypothesis that excessive estrogen exposure during development may increase the risk of developing breast cancer.
In utero exposure to tamoxifen, an estrogen antagonist and partial agonist, has also been shown to increase the incidence of mammary tumors when the exposed offspring are challenged with DMBA at puberty. Eighteen weeks after the challenge, 95% of the tamoxifen-exposed animals developed tumors, compared with 50% of the vehicle-treated rats (138). However, in the above-mentioned studies, both DES and tamoxifen were administered at high pharmacological doses to reflect the medical use of these agents, whereas the effects of twinning mentioned above represent a physiological range of endogenous hormone levels to which developing fetuses are exposed.
E. Perinatal exposure to environmentally relevant levels of endocrine disruptors
There is a third type of exposure that needs to be addressed: the inadvertent and continuous exposure of fetuses to environmentally active chemicals, such as dioxins and BPA (Table 4).
Depending on the context (time of exposure, organ, presence or absence of estrogens) dioxins have either estrogenic or antiestrogenic effects. Despite cross-talk between the aryl hydrocarbon and ERs (139), the mechanisms underlying these opposite effects have yet to be elucidated. Rats exposed prenatally (gestational d 15) to TCDD and challenged with the chemical carcinogen DMBA at 50 d of age showed increased tumor incidence, increased number of tumors per animal, and shorter latency period than rats exposed prenatally to vehicle and to DMBA at 50 d of age. These TCDD-exposed animals had increased numbers of terminal end buds at puberty (140). Because these structures are believed to be the site where mammary cancer arises, these results were interpreted as evidence that TCDD increased the propensity to cancer by altering mammary gland morphogenesis. Interestingly, Fenton (31) showed that prenatal exposure to TCDD results in impaired development of terminal end buds that remain in the gland for prolonged periods, whereas in the normal animals terminal end buds are transient structures that regress when ductal development is completed.
2. BPA, a ubiquitous xenoestrogen
The ubiquitous use of BPA provides great potential for exposure of both the developing fetus, indirectly through maternal exposure, and the neonate, directly through ingestion of tinned food, infant formula, or maternal milk (11). Indeed, BPA has been measured in maternal and fetal plasma and placental tissue at birth in humans (141). A recently published study conducted by the Centers for Disease Control, the first using a reference human population, showed that 92.6% of over 2500 Americans had BPA in their urine (142). Measured urine concentrations were significantly higher in children and adolescents compared with adults. BPA has also been measured in the milk of lactating mothers. These data indicate that the developing human fetus and neonate are readily exposed to this chemical.
Postnatal development of Leydig cells involves transformation through three stages: progenitor, immature, and adult Leydig cells. The process of differentiation is accompanied by a progressive increase in the capacity of Leydig cells to produce testosterone (T). T promotes the male phenotype in the prepubertal period and maintains sexual function in adulthood; therefore, disruption of T biosynthesis in Leydig cells can adversely affect male fertility. The present study was designed to evaluate the ability of a xenoestrogen, methoxychlor (the methoxylated isomer of DDT [1,1,1-trichloro-2,2-bis(p-chlorophenyl)ethane]), to alter Leydig cell steroidogenic function. Purified progenitor, immature, and adult Leydig cells were obtained from, respectively, 21-, 35-, and 90-day-old Sprague-Dawley rats treated with graded concentrations of the biologically active metabolite of methoxychlor, 2,2-bis(p-hydroxyphenyl)-1,1,1-trichloroethane (HPTE), and assessed for T production. HPTE caused a dose-dependent inhibition of basal and LH-stimulated T production by Leydig cells. Compared to the control value, reduced T production by progenitor and immature Leydig cells was apparent after 10 h of HPTE treatment in culture; the equivalent time for adult Leydig cells was 18 h. The reversibility of HPTE-induced inhibition was evaluated by incubating Leydig cells for 3, 6, 10, 14, or 18 h and measuring T production after allowing time for recovery. After treatment with HPTE for 3 h, T production by immature and adult Leydig cells for the 18-h posttreatment period was similar to the control value, but that of progenitor Leydig cells was significantly lower. The onset of HPTE action and the reversibility of its effect showed that Leydig cells are more sensitive to this compound during pubertal differentiation than in adulthood. T production was comparable when control and HPTE-treated immature Leydig cells were incubated with pregnenolone, progesterone, and androstenedione, but HPTE-treated Leydig cells produced significantly reduced amounts of T when incubations were conducted with 22R-hydroxycholesterol (P < 0.01). This finding suggested that HPTE-induced inhibition of T production is related to a decrease in the activity of cytochrome P450 cholesterol side-chain cleavage enzyme (P450scc) and cholesterol utilization. The reduced steady-state mRNA level for P450scc in HPTE-treated Leydig cells was demonstrated by reverse transcription-polymerase chain reaction and densitometry. In conclusion, this study showed that HPTE causes a direct inhibition of T biosynthesis by Leydig cells at all stages of development. This effect suggests that reduced T production could be a contributory factor in male infertility associated with methoxychlor and, possibly, other DDT-related compounds.
It has been hypothesized that a global decline in the semen quality in men is associated with environmental factors, notably pesticides with estrogenic activity [1, 2]. Studies in laboratory animals and wildlife have also demonstrated a variety of reproductive anomalies in males exposed to estrogenic pesticides and related compounds . The natural estrogen, estradiol-17β (E), is known to be involved in gonadotropin regulation of Leydig cell steroidogenesis , fluid absorption in the male reproductive tract , and maintenance of male bone density . Estrogen receptors (ERs) are expressed in fetal, neonatal, and adult tissues, including the hypothalamus, pituitary, testis, and the excurrent duct system, suggesting that estrogen may support multiple activities in male reproduction [7–11]. The primary male hormone, testosterone (T), is produced almost exclusively by Leydig cells in the testis. Leydig cell differentiation is sensitive to inhibition by estrogen, evidenced by E-induced suppression of Leydig cell regeneration in rats treated with a cytotoxin, ethane dimethylsulfonate .
Environmental toxicants associated with infertility in wildlife and laboratory species include organophosphate and polychlorinated pesticides, as well as industrial chemicals. While DDT [1,1,1-trichloro-2,2-bis(p-chlorophenyl)ethane]) itself was banned in the United States in 1972, methoxychlor remains in agricultural use because of its diminished persistence in the atmosphere, rapid excretion from the body, and lower toxicity . However, this compound was associated with a decrease in the weight of the testis, seminal vesicles, and prostate in the adult rat  and the testis in rabbits . When administered to prepubertal rats, methoxychlor reduced testicular steroidogenesis, interstitial fluid T content, and spermatogenesis, suggesting a direct action on Leydig cells . It is now known that endocrine disrupters cause demasculinization in rodents through altered androgen biosynthesis . The estrogenic activities of methoxychlor were attributed to demethylated phenolic derivatives that result from its biotransformation in the liver. Furthermore, these metabolites, especially HPTE [2,2-bis(p-hydroxyphenyl)-1,1,1-trichloroethane], are known to bind the ER to a greater degree than methoxychlor itself [18, 19].
In the rat, Leydig cell development begins with proliferation and commitment of mesenchymal cells to the Leydig cell lineage by Day 14 postpartum; these are designated progenitor Leydig cells. Progenitor Leydig cells become immature Leydig cells between Days 21 and 35. Finally, immature Leydig cells differentiate into adult Leydig cells by Day 56 . Since these Leydig cell stages differ in their capacities for T production, T levels provide an index for assessing Leydig cell differentiation. In the prepubertal period, T promotes the development of male secondary sex characteristics and hormonal imprinting of the liver, prostate, and hypothalamus. In the adult, T supports spermatogenesis, sperm maturation, and sexual function . Therefore, disruption of T biosynthesis has implications for the masculinization process and fertility. This study was designed to characterize the effects of HPTE on male reproductive function through evaluation of T production capacity in differentiating Leydig cells.
Materials and Methods
Sodium bicarbonate (NaHCO3), Hepes, trypsin inhibitor, heparin, EDTA, BSA, bovine lipoprotein, Dulbecco's modified Eagle's medium nutrient mixture:Ham's F-12 (DMEM/F-12; 1:1 mixture without phenol red), E, diethylstilbestrol (DES), 5-pregnen-3β-ol-20-one (pregnenolone), dibutyryl cAMP (dbcAMP: N6,2′-O-dibutyryl adenosine 3′,5′-cyclic monophosphate), 22R-hydroxycholesterol (22R-CHO), albumin, Percoll, etiocholan-3β-ol-17-one, nicotinamide adenine dinucleotide (NAD), nitro blue tetrazolium, and gentamycin were purchased from Sigma Chemical Company (St. Louis, MO). Dulbecco's PBS, Medium 199, and 10-strength Hanks' Balanced Salt Solution were obtained from Life Technologies (New York, NY); and collagenase, dispase, and DNase from Boehringer Mannheim GmbH (Mannheim, Germany). The antiestrogen ICI 182,780 [7α-[9-(4,4,5,5,-pentafluoropentylsulphinyl)estra-1,3,5(10)-triene-3,17β-diol]] was a gift from Zeneca Pharmaceuticals (Cheshire, UK). Ovine LH was generously provided by the National Hormone and Pituitary Program (National Institute of Diabetes and Digestive and Kidney Diseases, Rockville, MD). HPTE was synthesized by Dr. W.R. Kelce (Monsanto Company, St. Louis, MO).
Male Sprague-Dawley rats were purchased from Charles River Laboratories (Wilmington, MA). Progenitor, immature, and adult Leydig cells were isolated from rats that were, respectively, 21, 35, and 90 days of age on the day of isolation; these isolations used, respectively, 50 male pups, 18 pubertal, and 10 sexually mature rats. The animals were killed by CO2 asphyxiation, according to a protocol approved by the Institutional Animal Care and Use Committee of Rockefeller University (Protocol 91200-R2). Leydig cells were isolated from the testes by collagenase digestion, centrifugal elutriation, and Percoll density gradient centrifugation according to the method of Klinefelter et al. . Cell yields were estimated with a hemocytometer, and purity was assessed by histochemical staining for 3β-hydroxysteroid dehydrogenase (3β-HSD) using 0.4 mM etiocholan-3β-ol-17-one as the enzyme substrate . Progenitor Leydig cells were 90% enriched for cells that stain weakly for the marker enzyme 3β-HSD, and immature and adult Leydig cell fractions were typically 95–97% enriched for cells that stain intensely.
Culture of Leydig Cells
The culture medium consisted of DMEM/F-12 buffered with 14 mM NaHCO3, containing 0.1% BSA and 0.5 mg/ml bovine lipoprotein . Progenitor and immature Leydig cells were cultured at a density of 0.5 to 1.0 × 106 cells per well in 6-well plates, and adult Leydig cells at 0.2 to 0.3 × 106 cells in 24-well plates (Corning-Costar Company, New York, NY), in an atmosphere containing 5% O2 and 5% CO2 and a temperature of 34°C. Leydig cells were allowed 2 h to attach in fresh medium alone that were thereafter replaced by test solutions (fresh media containing test compounds). Except for 22R-CHO, which was solubilized in dimethylsulfoxide (DMSO), test compounds were dissolved in 95% ethanol and kept as 10 mM stock solutions at −20°C. When required, stock solutions were serially diluted such that 1 μl in 1 ml of the culture media gave the desired concentration of test compound. Equivalent amounts (1 μl/ml) of the solvent were added to the media for the control cultures. This level (0.1%) of ethanol or DMSO in the media did not significantly alter T production by Leydig cells. Culture of Leydig cells with low amounts of LH maintains or increases the number of LH receptors in Leydig cells, but incubation with high doses for prolonged periods results in a decrease in the number of LH receptors and Leydig cell desensitization . To avoid desensitization, incubation of Leydig cells for longer than 6 h was done using a low concentration of ovine LH (10 ng/ml), while incubations of shorter duration were conducted with a high dose (100 ng/ml). T concentrations were determined by RIA as previously described, with an interassay variation of 7.8% . T production values were normalized to ng T/106 cells.
Dose Dependency of HPTE Effect
To establish that HPTE has an effect on T production, progenitor, immature, and adult Leydig cells were cultured for 18 h in the absence or presence of 0, 0.01, 0.05, 0.1, 0.2, 0.5, 1, 2, 5, and 10 μM HPTE. After treatment, cells were harvested following a 5-min incubation in a solution of 0.05% collagenase and 0.05% dispase in Medium-199 buffered with 8.45 mM NaHCO3 and 8.8 mM Hepes, containing 0.1% BSA and 0.0025% trypsin inhibitor, at pH 7.1–7.2. To facilitate the process of detachment, culture plates were subsequently placed on an orbital shaker at 70 cycles/min for 10 min; Leydig cells were removed thereafter. The viability of harvested cells was assessed by the trypan blue exclusion test. Aliquots of harvested cells (0.25–0.5 × 106) were subsequently incubated in fresh media for 3 h at 34°C. The incubations were conducted in the absence or presence of a maximally stimulating (100 ng/ml) dose of ovine LH . At the end of incubation, the tubes were centrifuged at 500 × g and stored at −20°C until RIA.
Onset and Reversibility of HPTE Effect
Dose-response data indicated that 1.11 ± 0.02 μM HPTE caused half-maximal inhibition (IC50) of T production by Leydig cells at all three stages of differentiation. Therefore, subsequent studies were conducted using HPTE at a concentration of 1 μM; cells remained attached throughout the incubation period. To ensure that this dose was not directly cytotoxic, a cell fluorescence assay (Promega, Madison, WI) was performed after incubation of Leydig cells in media containing 10 ng/ml LH for 18 h with and without 1 μM HPTE. After treatment, a premixed dye solution containing a tetrazolium salt was added to the cultures, and incubation was continued for another 4 h. Cell viability corresponded to cellular conversion of the tetrazolium salt into a colored formazan product and was quantified using an ELISA plate reader (Dynex Technologies, Chantilly, VA). The results were in agreement with those of the trypan blue exclusion test.
The time course for HPTE inhibition was analyzed in order to determine whether HPTE action results from an acute effect that occurs in a short period of time, e.g., less than 1 h, or is delayed as would be expected with changes in gene expression, e.g., requiring several hours to manifest. Progenitor, immature, and adult Leydig cells were cultured in media containing 10 ng/ml LH with and without 1 μM HPTE for 3, 6, 10, 14, or 18 h. At the end of each time point, incubation was stopped, and spent media were collected for assay of T concentrations by RIA.
The reversibility of HPTE effect, as indicated by the ability of Leydig cells to restore T production capacity after cessation of treatment, was evaluated by incubating progenitor, immature, and adult Leydig cells in media containing 10 ng/ml LH with and without 1 μM HPTE for 3, 6, 10, 14, or 18 h. At the end of treatment, spent media were replaced by fresh media containing 10 ng/ml LH, and the incubation was continued for another 18 h regardless of the initial treatment period. At the end of culture, media were collected and stored at −20°C until RIA.
Antagonism of HPTE Effect by Coincubation with Antiestrogen
ICI 182,780 is a pure antiestrogen that binds the two isoforms of the ER . This compound was used to determine whether HPTE inhibition of T production could be prevented by blockade of the Leydig cell ER. Immature Leydig cells were used for this determination and in subsequent studies because their sensitivity to HPTE action was intermediate between that of progenitor and that of adult Leydig cells. Leydig cells were incubated for 18 h with and without 1 μM HPTE, ICI 182,780 (5 μM) plus HPTE, or ICI 182,780 in medium containing 10 ng/ml LH. The 5 μM dose of ICI 182,780 was identified from pilot experiments as the concentration that prevented inhibition of T production by 1 μM HPTE in vitro. To test for the inhibitory effect of E and DES on Leydig cell steroidogenic function, immature Leydig cells were incubated for 18 h with graded doses (ranging from 10−11 to 10−6 M) of the two compounds. After the spent media were discarded and cells gently rinsed, incubation was continued in fresh media with and without 100 ng/ml LH for a 6-h period. Basal and LH-stimulated T production was measured in aliquots of the media collected at the end of this posttreatment period.
Evaluation of the T Biosynthetic Pathway
Immature Leydig cells were first incubated in culture media for 18 h with and without 1 μM HPTE. At the end of treatment, cells were gently rinsed without detachment, and incubation was continued for another 6 h in fresh media containing 100 ng/ml LH or 100 μM dbcAMP. These concentrations are known to maximally stimulate T production by Leydig cells in vitro [26, 27]. Previously exposed Leydig cells (HPTE, 18 h) were also incubated for 6 h in the presence of T precursors: 50 μM 22R-CHO or 20 μM of pregnenolone, progesterone, and androstenedione. These steroid substrates are known to diffuse readily into Leydig cells and are used at high doses to determine the capacity of their respective enzymes : cytochrome P450 cholesterol side-chain cleavage (P450scc; 22R-CHO), 3β-HSD (pregnenolone), cytochrome P450 17α-hydroxylase/17-20 lyase (progesterone), and 17β-hydroxysteroid dehydrogenase (androstenedione). Pilot studies confirmed that the concentrations of steroids used were substrate saturating for these enzymes in immature Leydig cells.
Of the four substrates tested, only 22R-CHO failed to reverse HPTE inhibition, indicating that the decline in T production by Leydig cells is most likely due to reduced capacity for cholesterol side-chain cleavage rather than its transport. Steady-state mRNA levels for P450scc and steroidogenic acute regulatory protein (StAR), responsible for transporting cholesterol to the inner mitochondrial membrane location of P450scc , were measured after immature Leydig cells were incubated for 12 h in media containing 10 ng/ml LH with and without 1 μM HPTE. Total RNA was isolated by a single-step method after cells were lysed in culture plates without detachment using phenol and guanidinium thiocyanate (Trireagent; Molecular Research Center, Cincinnati, OH) in accordance with the manufacturer's instructions. First-strand cDNA synthesis from 400 ng of total RNA was done using avian myeloblastosis virus reverse transcriptase, random primers, and deoxyNTPs at 37°C for 75 min, and the reaction was terminated by heating at 95°C for 5 min. Target cDNA was coamplified with RPS16 as the internal control in an aliquot of the synthesized product. Primers for the target cDNAs were synthesized on an oligonucleotide synthesizer (Gene Assembler Special; LKB, Rockville, MD) using their published sequences. The sequences were 5′-AGGTGTAGCTCAGGACTTCA-3′ (forward) and 5′-AGGAGGCTATAAAGGACACC-3′ (reverse) for P450scc , 5′-TTGGGCATACTCAACAACCA-3′ (forward) and 5′-ATGACACCGCTTTGCTCAG-3′ (reverse) for StAR , and 5′-AAGTCTTCGGACGCAAGAAA-3′ (forward), 5′-TTGCCCAGAAGCAGAACAG-3′ (reverse) for RPS16 . The expected product sizes were 399, 389, and 148 base pairs (bp) for P450scc, StAR, and RPS16, respectively. PCR was initiated by the addition of Taq DNA polymerase and continued for 35 cycles of 1 min each at 94°C, 57°C, and 72°C. Preliminary studies showed that the cDNAs of interest were amplified linearly between 15 and 35 cycles of PCR. PCR products were separated on 3% agarose, visualized by ethidium bromide staining, and quantified on a densitometer (Kodak Scientific Imaging Systems, New York, NY) using 0.65 μg of 100-bp DNA ladder as standard. The mass of PCR products for P450scc and StAR was normalized to RPS16. These experiments were conducted three times, and RT-PCR was replicated at least three times on each occasion.
Data are presented as mean ± SEM and represent the average of three separate experiments for each determination (total of six determinations; Figs. 1–5Table 1). The dose-response data (Fig. 1) were analyzed by a four-parameter logistic curve fitting using the SAAM II optimizer and IC50 values derived therefrom . Other data were examined by one-way ANOVA with multiple comparisons performed by the Duncan multiple range test to identify differences between groups.
T production (ng/106 cells) by control and HPTE-treated (1 μM, 18 h) immature Leydig cells that were incubated with LH, dbcAMP, or steroid substrates.a
|Basal||24 ± 2||8 ± 1*|
|LH||105 ± 5||30 ± 4*|
|dbcAMP||122 ± 12||41 ± 8*|
|22R-CHO||218 ± 7||93 ± 6*|
|Pregnenolone||1595 ± 58||1564 ± 75|
|Progesterone||1371 ± 87||1326 ± 104|
|Androstenedione||1825 ± 64||1791 ± 80|
|Basal||24 ± 2||8 ± 1*|
|LH||105 ± 5||30 ± 4*|
|dbcAMP||122 ± 12||41 ± 8*|
|22R-CHO||218 ± 7||93 ± 6*|
|Pregnenolone||1595 ± 58||1564 ± 75|
|Progesterone||1371 ± 87||1326 ± 104|
|Androstenedione||1825 ± 64||1791 ± 80|