Impact on Male Rat Reproductive Tract Development Vickie S

international journal of andrology ISSN 0105-6263

REVIEW ARTICLE Diverse mechanisms of anti-androgen action: impact on male rat reproductive tract development Vickie S. Wilson,* Chad R. Blystone,* Andrew K. Hotchkiss,* Cynthia V. Rider* and L. Earl Gray Jr*

*US Environmental Protection Agency, Ofﬁce of Research and Development, National Health and Environmental Effects Laboratory, Reproductive Toxicology Division, North Carolina, and North Carolina State University, Raleigh, NC, USA

Summary

Keywords: Scientists have identified environmental chemicals that display anti-androgenic androgen receptor, anti-androgen, linuron, activity via multiple mechanisms of action. Early studies focused on pesticides phthalate, prochloraz, reproductive acting as androgen receptor (AR) antagonists but it soon became apparent that development, vinclozolin was not the only endocrine mode by which compounds affected the androgen signalling pathway. Classes of chemicals currently known to interfere with the Correspondence: Vickie S. Wilson, US Environmental Protection androgen signalling pathway include dicarboximide fungicides (e.g. vinclozolin), Agency, ORD, National Health and organochlorine-based insecticides (e.g. p,p¢-DDT and -DDE), conazole fungi- Environmental Effects Laboratory, cides (e.g. prochloraz), plasticizers (phthalates) and urea-based herbicides (linu- Reproductive Toxicology Division, MD-72, ron). Phthalate esters (PEs) and vinclozolin appear to act primarily via a single 2525 E. Highway 54, Research Triangle Park, mechanism of action, while others such as linuron and prochloraz, appear to NC 27711, USA. display dual mechanisms of action. Exposure to PEs decreases mRNA expres- E-mail: [email protected] sion of key steroidogenic enzymes and also the peptide hormone insulin-like Received 30 October 2007; revised 13 peptide 3 (insl3) from the foetal Leydig cells. Hence, both androgen- and inls3- December 2007; accepted 18 December 2007 dependent tissues are affected. Vinclozolin and procymidone act solely through binding to the AR as antagonists thus blocking the action of androgen at the doi:10.1111/j.1365-2605.2007.00861.x cellular level but do not affect foetal testosterone synthesis or insl3 gene expression. The compounds linuron and prochloraz are AR antagonists but also inhibit foetal testosterone synthesis, although unlike the PEs, mRNA expression of steroidogenic enzymes and insl3 are not affected. All the above chemicals disrupt androgen signalling in the foetal male rat and produce some malformations in common, but the precise profiles of effects in the offspring are pathognomonic for each mode of action. For example, the ‘phthalate syndrome’ vs. the ‘vinclozolin syndrome’ each displays a profile of effects which is clearly different. In summary, as more and more molecular studies with anti- androgenic compounds are conducted, the number of mechanisms by which compounds can affect the androgen signalling pathway is likely to increase. Furthermore, the effects of mixtures of these compounds are just beginning to be explored.

pathologies (Jegou et al., 2001). As only a small percent- Introduction age of these lesions can be linked directly to known Concern has risen over the apparent increase in male genetic defects, developmental exposure to man-made reproductive health problems and the potential role of chemicals has been implicated in the increases in these endocrine disrupting chemicals (EDCs) in the aetiology reproductive malformations. of these conditions (Skakkebaek, 2002a). Declining sperm Compounds that disrupt endocrine signalling pathways counts and increased incidences of hypospadias, cryptor- can act via several mechanisms including binding to hor- chidism and testis cancer have all been reported. The mone receptors, modifying the production or metabolism worldwide doubling in the rate of testicular cancer over of endogenous hormones, or modifying the number of the last 40 years is one of the more robust of these hormone receptors. For example, androgen agonists bind

Journal compilation ª 2008 Blackwell Publishing Ltd • International Journal of Andrology 31, 178–187 178 No claim to original US government works V. S. Wilson et al. Diverse mechanisms of anti-androgen action to androgen receptors and mimic natural hormones signalling pathway through slightly different mechanisms, which may initiate a cascade of events resulting in inap- there are both similarities and differences in androgen- propriate expression of hormone-dependent genes. Con- dependent tissue effects and the profile of malformations versely, androgen receptor (AR) antagonists compete with which are produced. In the course of this summary, endogenous androgen for the receptor and prevent down- mechanism-based groups of anti-androgens was discussed stream activation of androgen-dependent gene expression. with focus on the pathognomic effects of each class of While early research focused on the ability of environ- anti-androgen. The terms ‘low’ dose and ‘high’ dose will mental compounds to impact androgen binding to its generally be used in the discussion. The specific doses receptor, it is now clear that it is not the only mechanism which fit each of these categories are dependent on the through which compounds can interfere with the andro- potency and mechanism of action of each compound dis- gen signalling pathway. In addition, some compounds cussed and will clearly be different for each individual appear to act through more than one mechanism, each of compound. Information on the actual dose can be found which having the potential impact on the androgen sig- in the reference literature. Generally, low dose refers to nalling pathway at different points. a dose near or just above the no observed adverse effect Mammalian gonadal development and sex differentia- level for the individual compound. Additionally, we cover tion occur during a relatively narrow time window. The the differential effects between two strains of rats exposed production of testosterone and other hormones by the to the same anti-androgen. Lastly, we discuss conclusions foetal testis during this period are critical factors for and future directions for this work including the study of the proper development of the male reproductive tract complex mixtures of anti-androgens with diverse mecha- (Schardein, 1993). In utero exposure to chemicals that nisms of action. disrupt androgen signalling during this window can result in alterations that may include reduced anogenital dis- Androgen receptor antagonists tance (AGD), increase in female-like nipple retention, epididymal agenesis, reduction in sex accessory gland Several environmental chemicals adversely affect male weights, cryptorchidism, hypospadias and reduced fertil- development by interfering with androgen receptor signal- ity. Generally, in utero exposure to anti-androgenic chem- ling during the critical periods (in utero or peri-pubertal) icals affects male offspring, while having little effect on of sex differentiation and maturation. Depending on the female offspring. Androgenic chemicals, conversely, have timing of developmental exposure, these anti-androgens been shown to induce malformations in exposed female affect androgen-sensitive organs and processes within the offspring and have little to no effect on males. male rat leading to altered neuroendocrine development The suite of effects following in utero exposure to and behaviour, reduced reproductive organ weights and endocrine disruptors can be indicative of the mechanism malformations of external genitalia and reproductive of action of the chemical exposure. In addition to the organs. Several anti-androgenic chemicals have been iden- effects mentioned above from androgenic and anti-andro- tified as androgen receptor (AR) antagonists including: genic exposure, environmental oestrogens have been the fungicides vinclozolin (Gray et al., 1994; Kelce & Wil- found to cause infertility or shorten the reproductive life- son, 1997), procymidone (Hosokawa et al., 1993; Ostby span of in utero exposed females. Compounds such as et al., 1999), prochloraz (Vinggaard et al., 2002; Noriega 2,3,7,8-tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD) elicit et al., 2005), the DDT metabolite p,p¢-DDE (Kelce et al., effects in the female reproductive tract including malfor- 1995), the herbicide linuron (Gray et al., 1999a; Lam- mations of the external genitalia, reduced ovarian weight bright et al., 2000) and polybrominated diphenyl ethers and can also decrease their reproductive lifespan. In fact, flame retardants (Stoker et al., 2005). a relatively low single oral dose of 1–2 lg ⁄ kg 2,3,7,8- In utero exposure to chemicals that disrupt AR signal- TCDD given to the dam during gestation can produce ling leads to a host of typical effects in male rat offspring. frank reproductive malformations in both male and External biomarkers of prenatal androgen disruption female offspring. Lastly, foetal germ cell toxicants, such as include the anogenital distance (AGD) and the juvenile busulfan, can cause infertility in females, reduce sperm nipple ⁄ areolae number. The AGD is defined as the dis- production in males and elicit histological lesions in both tance between the genital papilla and the anus; male ovaries and testes without affecting gross morphology of rodents have AGDs that are approximately twice the the reproductive tract in either sex. length as those of females (Vandenbergh & Huggett, The focus of this selective review was to summarize 1995; Gray et al., 1999a,b). Areolae (areolas) are dark our laboratory’s research on the effects of in utero areas surrounding the nipple bud and their presence as exposure to anti-androgens on the male rat reproductive measured at postnatal day 2–3 is indicative of adult nip- tract. While all these compounds impact the androgen ples. Adult female rats typically have 12 nipples whereas

Journal compilation ª 2008 Blackwell Publishing Ltd • International Journal of Andrology 31, 178–187 No claim to original US government works 179 Diverse mechanisms of anti-androgen action V. S. Wilson et al. males have none. Both of these biomarkers vary with agenesis, undescended testes and formation of a vaginal prenatal exposure to androgens or anti-androgens in pouch. The association between high dose exposures and females and males respectively (Gray et al., 1999a,b; resulting malformations is clear. Barlow et al., 2004; Hotchkiss et al., 2004, 2007b). Reduction of anogenital distance and ⁄ or retention of Testosterone synthesis inhibitors nipples in male rats are indicative of prenatal exposure to anti-androgens. Although there is no innate adversity Several phthalates and pesticides cause malformations in associated with shortened AGD and retained areolae, male rat offspring after in utero exposure, and do so in these relatively low dose effects persist through adulthood part or in whole, by inhibiting testosterone synthesis. The (Hotchkiss et al., 2004) and are predictive of malform- mechanism of inhibition varies among these environmen- ations in the adult male reproductive tract. tal toxicants. Data from foetal ex vivo testis incubations Perinatal exposure to androgens is critical for the mas- and testis mRNA experiments show that the phthalates culinization of behavioural phenotype in the males in diethylhexyl phthalate (DEHP), dibutyl phthalate (DBP) many species. However, in some species of mammals, the and benzylbutyl phthalate (BBP) decrease foetal testis tes- conversion of testosterone into oestrogen by aromatase is tosterone production and reduce the expression of steroi- critical for differentiation of male behaviour. In rats, most dogenic genes after in utero exposure (Parks et al., 2000; masculinization of the brain is not by testosterone per se, Barlow & Foster, 2003; Wilson et al., 2004b; Borch et al., but rather is either not influenced by steroid hormones at 2006; Howdeshell et al., 2007). Examples of the mRNA of all, or is masculinized through the conversion of testos- steroidogenic genes reduced by in utero phthalate treat- terone to oestrogen and subsequent interaction with the ment includes cyp17 (17a-hydroxylase ⁄ C17,20-lyase), oestrogen receptor. Although oestrogen is critical for mas- cyp11a (side-chain cleavage) and StAR (steroidogenic culinization of some behaviours in the rat, others, such as acute regulatory protein). juvenile play behaviour, are affected specifically by andro- In addition to inhibiting testosterone synthesis, the gens during the perinatal period. Social play is organized phthalates DEHP, BBP and DBP also reduce expression by the perinatal interaction of androgens with the AR of insulin-like peptide 3 (insl3), an important hormone (Meaney et al., 1983; Meaney & McEwen, 1986; Pellis & secreted by the Leydig cell necessary for development of Pellis, 1997). As such, juvenile males play more than the gubernacular ligament. This results in unsuccessful females. Neonatal exposure to AR antagonists results in decent of the testis (Wilson et al., 2004a). Reduction of a reduction of male play behaviour (Meaney et al., 1983; insl3 by phthalates is not produced by the AR antago- Hotchkiss et al., 2003). Another central nervous system nists, vinclozolin and procymidone, or the mixed anti- (CNS) target influenced by the AR signalling pathway is androgens linuron and prochloraz. Phthalate-induced the development of the spinal neurons that innervate the reduction of insl3 is consistent with the higher incidence male sex organs. In females, these neurons undergo onto- of cryptorchidism (unsuccessful descent of the testis) in genetic death during the perinatal period. In males, acti- male rats exposed in utero to phthalates compared with vation of the AR signalling pathway by androgens rescues pesticides that are AR antagonists, or pesticides that are these neurons from death (reviewed in Breedlove et al., weak AR antagonists which do not induce cryptorchidism 1999). Future research will examine whether perinatal (Gray et al., 2006). The combination of unique effects exposure to AR antagonists reduces the numbers of spinal associated with phthalates has been labelled the ‘phthalate neurons and the associated behaviours. Although limited syndrome’ (Foster, 2006). The adverse outcomes on male examples of androgen organization of behaviour exist in reproductive development because of phthalate exposure, the rat, it appears to be more common in the non-human which affect Wolfian duct and prostate differentiation primates (Cooke et al., 1998), as developmental exposure along with inducing hypospadias and cryptorchidism, to androgens in non-human primates enhances the male- have striking parallels with the reported human testicular typical expression of mounting behaviour in adults (Pom- dysgenesis syndrome (Sharpe, 2001; Skakkebaek, 2002b). erantz et al., 1985; Thornton & Goy, 1986). Recently, an informal review of the literature indicated Finally, disruption of the AR signalling pathway can that there were notable strain differences after in utero also lead to alterations of the reproductive tissues. For DEHP exposure between Sprague Dawley (SD) and example, at higher levels of anti-androgens, androgen-sen- Wistar male rats. The incidences of epididymal and sitive organs including ventral prostate, seminal vesicles, gubernacular lesions in response to DEHP exposure were epididymides and testes are subject to weight reductions found to vary depending on rat strain: Wistar rats have (Gray et al., 2006). At the highest levels of exposure a higher incidence of gubernacular lesions and lower a suite of typical malformations of the male reproductive incidence of epididymal lesions while SD rats display the tract occur. These include hypospadias, epididymal reverse as a result of DEHP treatment (Wilson et al.,

Journal compilation ª 2008 Blackwell Publishing Ltd • International Journal of Andrology 31, 178–187 180 No claim to original US government works V. S. Wilson et al. Diverse mechanisms of anti-androgen action

2007). To elucidate the mechanism behind these antagonism and reduced testosterone production differences, the effects of DEHP on foetal testosterone displayed by these chemicals may have a cumulative effect production and insl3 expression were investigated in each on the male offspring, but the relative contribution of strain. Although DEHP decreased foetal testis testosterone each mechanism needs further clarification. production and insl3 expression in each strain, the ratio Linuron is a urea-based herbicide that binds to the AR of the reduction in insl3 mRNA to testosterone produc- and competitively inhibits AR-DNA binding and gene tion indicated that testosterone production is relatively activation. Both ex vivo and in vitro treatment with linu- more affected in the SD rat strain than in the Wistar ron significantly reduce testosterone production from the strain while insl3 expression was more affected in the foetal testis without significantly affecting progesterone Wistar than the SD strain. The identification of the differ- production (Hotchkiss et al., 2004; Wilson et al., 2004b). ences in lesions and foetal hormones between the strains This reduction in foetal testosterone results in a profile could potentially be used as a tool to aid in understand- that more closely resembles phthalates than a pure AR ing the mode of action for the phthalates as reproductive antagonist such as vinclozolin (Hotchkiss et al., 2007a). toxicants as well as prove to be a model for genetic Briefly, males exposed in utero to linuron have modest differences in human populations. reductions in neonatal anogenital distance, and reduced While PE reduced foetal testosterone production by ventral prostate weight, seminal vesicle and epididymal decreasing mRNA expression of steroidogenic enzymes, weights as well as malformations of the epididymides and the fungicide prochloraz does not reduce foetal testoster- testis. Although this profile of effects is more commonly one production through the same mechanism. In utero associated with foetal androgen synthesis inhibitors such exposure to the pesticide prochloraz increases foetal pro- as the phthalates, certain aspects of the linuron profile are gesterone and decreases testosterone production through dissimilar to phthalates (Gray et al., 2006). For example, what appears to be direct inhibition of the activity of one insl3 gene expression is not affected by this chemical or more of the enzymes in the steroidogenic pathway (Wilson et al., 2004a) and, subsequently, in utero exposed without affecting the mRNA expression of those enzymes males do not display gubernacular ligament agenesis. In (Wilson et al., 2004b; Blystone et al., 2007a). Further- addition, expression of genes in the steroidogenic path- more, microsomal CYP17 hydroxylase activity was signifi- way that are targeted by phthalate treatment are not cantly inhibited by prochloraz (Ki = 865 nm or 326 ppb) reduced by maternal linuron treatment (V. S. Wilson, (Blystone et al., 2007a). In that study, prochloraz concen- personal communication). trations in amniotic fluid ranged from 78 to 1512 ppb Prochloraz is a fungicide that affects male rat develop- (207–4014 nm) and testosterone production was reduced ment. Adult male rats which were exposed in utero to when prochloraz reached approximately 500 ppb, which prochloraz have increased incidence of genital malforma- compares favourably with the experimentally determined tions and reduced reproductive organ weights (Noriega

Ki for CYP17 hydroxylase activity of 326 ppb. In utero et al., 2005; Vinggaard et al., 2005). Furthermore, peri- exposure to the phthalates or prochloraz, however, results pubertal exposure to prochloraz delays male pubertal in reduced testosterone levels during this critical develop- development (Blystone et al., 2007a). Evidence suggests mental window of sex differentiation of androgen-depen- that prochloraz inhibits CYP17 based upon the pattern of dent tissues. increased progestins and decreased androgens seen in ex vivo and in vivo experiments in the foetal and pubertal male rat (Wilson et al., 2004b; Blystone et al., 2007a,b). Mixed mechanisms Prochloraz inhibits the hydroxylase activity of CYP17 In addition to the AR antagonists and androgen synthesis in vitro at concentrations that are consistent with levels inhibitors described above, several chemicals have been measured in the rat amniotic fluid (Blystone et al., identified that appear to operate simultaneously by more 2007b). The data from in vitro AR antagonism and than one mechanism. These chemicals are herein referred inhibition of steroidogenesis studies suggest that the to as having a ‘mixed’ mechanism of action. Two such dominant mechanism of prochloraz is inhibition of tes- chemicals are the herbicide, linuron and the fungicide, tosterone production, not AR antagonism (Noriega et al., prochloraz. Both of these chemicals are AR antagonists 2005; Blystone et al., 2007b). However, as both mecha- which, similarly to vinclozolin and procymidone, directly nisms are affecting the AR signalling pathway, the compete with testosterone and dihydrotestosterone for contribution of the AR antagonism mechanism to the the AR (Lambright et al., 2000; Noriega et al., 2005). anti-androgen effects is likely magnified by the effect of Both linuron and prochloraz inhibit foetal testosterone decreased circulating testosterone. The dissimilarities of production (Hotchkiss et al., 2004; Wilson et al., 2004b; male malformations between linuron and prochloraz are Vinggaard et al., 2005; Blystone et al., 2007a,b). The AR interesting. At high doses, linuron induces epididymal

Journal compilation ª 2008 Blackwell Publishing Ltd • International Journal of Andrology 31, 178–187 No claim to original US government works 181 Diverse mechanisms of anti-androgen action V. S. Wilson et al.

agenesis and little hypospadias, while prochloraz induces estimated maximally effective dose (ED100) for inducing hypospadias with no epididymal agenesis (Gray et al., malformations (vinclozolin 15 mg ⁄ kg ⁄ day, procymidone 2006). It remains unclear why these chemicals, with 15 mg ⁄ kg ⁄ day, prochloraz 35 mg ⁄ kg ⁄ day, linuron reported similar mixed anti-androgen effects, would have 20 mg ⁄ kg ⁄ day and BBP, DBP and DEHP at dissimilar proﬁles of male malformations. 150 mg ⁄ kg ⁄ day). The mixture was administered at the

ED100 and 75, 50 and 25% of the ED100 level. Using this information, it was estimated that the ED treatment Mixtures of anti-androgens 100 was equivalent to vinclozolin 100 mg ⁄ kg ⁄ day and the pre- Although risk assessments are typically conducted on dicted reduction in AGD from each treatment using the a chemical-by-chemical basis, the 1996 Food Quality parameters was calculated from the vinclozolin linear Protection Act Law requires the USEPA to consider regression. Results of this study demonstrated that the cumulative risk from chemicals that act via a common mixture affected all androgen-dependent endpoints in mode ⁄ mechanism. Consequently, we are conducting stud- a dose-additive manner (Gray et al., 2006; Rider et al., ies with mixtures of anti-androgenic compounds in order 2008). Neonatal reductions in AGD in male rats can be to provide a framework for assessing the cumulative permanent and are highly correlated with an increase in in utero effects of ‘anti-androgenic’ compounds (Gray reproductive tract malformations; with more severe mal- et al., 2001, 2004a,b). In the initial studies, mixtures of formations being seen in the males with the shortest two compounds were evaluated both singly and in a AGD (McIntyre et al., 2002a,b; Hotchkiss et al., 2004). It binary mixture. Compounds were administered to SD rats is likely that cumulative toxicity will be seen on other from gestation day 14–18 at dosage levels equivalent to reproductive endpoints later in life. Additional results about one half of the effective dose which causes a 50% from this seven compound mixture are presented within incidence (ED50) of hypospadias and ⁄ or epididymal agen- the same issue of this journal (Rider et al., 2008). Results esis. The chemical pairs include: (i) two AR antagonists indicate that mixtures of chemicals that alter the andro- (vinclozolin plus procymidone, each at 50 mg ⁄ kg ⁄ day, no gen signalling pathway via diverse mechanisms do disrupt common active metabolite), (ii) two phthalate esters with male reproductive development in a cumulative manner. a common active metabolite (DBP and BBP, each at It appears that compounds which disrupt a common 500 mg ⁄ kg ⁄ day), (iii) two phthalate esters with different tissue during the period of sex differentiation, act in a active metabolites (DEHP and DBP, each at dose-additive manner irrespective of their speciﬁc cellular 500 mg ⁄ kg ⁄ day) (Howdeshell et al., 2007), (iv) a phtha- target. late ester plus an AR antagonist (DBP; 500 mg ⁄ kg ⁄ day) plus procymidone (50 mg ⁄ kg ⁄ day) and (v) linuron Conclusion (75 mg ⁄ kg ⁄ day) plus BBP (500 mg ⁄ kg ⁄ day) (Hotchkiss et al., 2004). At these doses, each chemical individually Anti-androgenic chemicals can target the androgen signal- was predicted to induce few, if any, reproductive tract ling pathway at various levels. The effects of AR antago- malformations; however, by mixing any two chemicals nists (vinclozolin and procymidone), testosterone together, they would induce reproductive tract malforma- synthesis inhibitors (phthalates) and mixed mechanism tions and affect androgen-dependent tissue weight in chemicals (linuron and prochloraz) have been discussed about 50% of the males. Results to date indicate that all herein. As summarized in Table 1, our research reveals combinations produced cumulative effects on the andro- that environmental chemicals can alter the androgen sig- gen-dependent tissues. As expected, only the phthalate nalling pathway via several distinct modes of action. ester combinations caused agenesis of the insl3-dependent Knowledge of the modes of action of these compounds gubernacular ligaments. Additional results demonstrate also allows us to make some predictions how individual that the phthalates need not have a common active tissues will be affected when anti-androgens are com- metabolite to produce cumulative dose-additive adverse bined. EDCs that alter differentiation of the same repro- effects (Howdeshell et al., 2007). ductive tissues during sexual differentiation produce Our mixture studies have now been extended to evalu- cumulative, apparently dose-additive, effects when com- ate a more complex mixture consisting of seven anti- bined even though the relative potency factors among androgens. The potency of each chemical relative to vinc- chemicals varies from tissue to tissue. lozolin was initially predicted by comparing ED50s from Given that severe alterations of sexual differentiation linear regression models of the dose response data for can be produced in rat laboratory studies, the question each chemical (Gray et al., 2006; Rider et al., 2008). In arises of what would be expected in exposed humans given the high dose group, termed hereafter as the ED100, each that humans are exposed to mixtures of compounds in chemical in the mixture was administered at 1 ⁄ 7th of its their environment. For many EDCs, we might expect to

Journal compilation ª 2008 Blackwell Publishing Ltd • International Journal of Andrology 31, 178–187 182 No claim to original US government works V. S. Wilson et al. Diverse mechanisms of anti-androgen action

Table 1 Anti-androgenic chemicals that impact the androgen signalling pathway can affect male reproductive development via several different mechanisms of action resulting in slightly differing proﬁles of effects

fl Testosterone production Binds Compound AR fl Insl3 fl mRNA flactivity w ⁄ no flmRNA ‘Low dose’ prominent malformations

Vinclozolin · 0 0 0 Retained nipples; Hypospadias; Agenesis of ventral prostate

Procymidone · 0 0 0 Similar to Vinclozolin

Linuron · 00 · Epididymal and testis abnormalities; No gubernacular agenesis

Prochloraz · 00 · Similar to Vinclozolin

Dibutyl phthalate 0 ·· – All three phthalates produce Benzybutyl phthalate 0 ·· – epididymal and testis abnormalities; Diethylhexyl phthalate 0 ·· – Gubernacular agenesis

fl: decrease, ·: a known mechanism, 0: does not act through this mechanism, flactivity w ⁄ no flmRNA: an · here indicates enzyme activity is decreased but expression levels of mRNA for the enzyme are not affected.

find adverse reproductive effects in only the most suscepti- to determine if the mechanism of action of some of these ble humans such as those exposed to levels of anti-andro- compounds is conserved between the test species and the gens during development if the range of the exposure species of concern (i.e. humans) and to determine if levels approaches those producing adverse effects in the exposure occurs during the critical period of reproductive developing rats. It cannot be ruled out that less affected development at levels which might be expected to be individuals might display permanent alterations in the problematic. absence of overt physical malformations. Observation that phthalate exposures were associated with reduced neonatal Disclaimer AGD was extended from studies with rats to humans (Swan et al., 2005). Swan et al. (2005) examined AGI The research described in this article has been reviewed (weight adjusted AGD) and other genital measurements in by the National Health and Environmental Effects relation to prenatal phthalate exposure in 134 boys, 2– Research Laboratory, ORD, US Environmental Protection 36 months of age, and found that urinary concentrations Agency, and approved for publication. Approval does not of four phthalate metabolites were inversely related to signify that the contents necessarily reflect the views and AGI. Their data support the hypothesis that prenatal phth- policies of the agency nor does the mention of trade alate exposure at environmental levels can adversely affect names or commercial products constitute endorsement or male reproductive development in humans in a manner recommendation for use. similar to that seen in rodent studies. In conclusion, while all the anti-androgenic compounds Acknowledgements discussed herein impact the androgen signalling pathway through slightly different mechanisms, there are both sim- The authors would like to thank Drs Kembra Howdeshell ilarities and differences in androgen-dependent tissue and Dalma Martinovic for their helpful comments and effects and the profile of malformations which is pro- constructive review of earlier versions of this manuscript. duced. As more and more molecular-based studies with anti-androgenic compounds are conducted, the mecha- References nisms by which compounds can impact the androgen signalling pathway are likely to expand. In addition, given Barlow, N. J. & Foster, P. M. (2003) Pathogenesis of male that humans are most often exposed to mixtures of low reproductive tract lesions from gestation through adulthood levels of compounds simultaneously, the next major fron- following in utero exposure to di(n-butyl) phthalate. Toxico- tier is the continued exploration of the combined effects logic Pathology 31, 397–410. of exposure to multiple anti-androgenic chemicals. Preli- Barlow, N. J., McIntyre, B. S. & Foster, P. M. (2004) Male minary animal studies have demonstrated that these reproductive tract lesions at 6, 12, and 18 months of age chemicals do not act independently, but contribute to following in utero exposure to di(n-butyl) phthalate. a cumulative toxicity. Additional research is also needed Toxicologic Pathology 32, 79–90.

Journal compilation ª 2008 Blackwell Publishing Ltd • International Journal of Andrology 31, 178–187 No claim to original US government works 183 Diverse mechanisms of anti-androgen action V. S. Wilson et al.

Blystone, C. R., Furr, J., Lambright, C. S., Howdeshell, K. L., (2006) Adverse effects of environmental antiandrogens and Ryan, B. C., Wilson, V. S., Leblanc, G. A. & Gray, L. E. Jr androgens on reproductive development in mammals. (2007a) Prochloraz inhibits testosterone production at International Journal of Andrology 29, 96–104; discussion dosages below those that affect androgen-dependent organ 105–108. weights or the onset of puberty in the male sprague dawley Hosokawa, S., Murakami, M., Ineyama, M., Yamada, T., Koy- rat. Toxicological Sciences 97, 65–74. ama, Y., Okuno, Y., Yoshitake, A., Yamada, H. & Miyamoto, Blystone, C. R., Lambright, C. S., Howdeshell, K. L., Furr, J., J. (1993) Effects of procymidone on reproductive organs Sternberg, R. M., Butterworth, B. C. et al. (2007b) Sensitiv- and serum gonadotropins in male rats. Journal of Toxicologi- ity of fetal rat testicular steroidogenesis to maternal prochlo- cal Sciences 18, 111–124. raz exposure and the underlying mechanism of inhibition. Hotchkiss, A. K., Ostby, J. S., Vandenbergh, J. G. & Gray, L. E. Toxicological Sciences 97, 512–519. Jr (2003) An environmental antiandrogen, vinclozolin, alters Borch, J., Metzdorff, S. B., Vinggaard, A. M., Brokken, L. & the organization of play behavior. Physiology and Behavior Dalgaard, M. (2006) Mechanisms underlying the anti-andro- 79, 151–156. genic effects of diethylhexyl phthalate in fetal rat testis. Toxi- Hotchkiss, A. K., Parks-Saldutti, L. G., Ostby, J. S., Lambright, cology 223, 144–155. C., Furr, J., Vandenbergh, J. G. & Gray, L. E. Jr (2004) A Breedlove, S. M., Cooke, B. M. & Jordan, C. L. (1999) The mixture of the ‘‘Antiandrogens’’ Linuron and butyl benzyl orthodox view of brain sexual differentiation. Brain, phthalate alters sexual differentiation of the male rat in a Behavior and Evolution 54, 8–14. cumulative fashion. Biology of Reproduction 71, 1852–1861. Cooke, B., Hegstrom, C. D., Villeneuve, L. S. & Breedlove, S. Hotchkiss, A. K., Furr, J., Makynen, E. A., Ankley, G. T. & Gray, M. (1998) Sexual differentiation of the vertebrate brain: L. E. Jr (2007a) In utero exposure to the environmental principles and mechanisms. Frontiers in Neuroendocrinology androgen trenbolone masculinizes female Sprague-Dawley 19, 323–362. rats. Toxicology Letters 174, 31–41. Foster, P. M. (2006) Disruption of reproductive development Hotchkiss, A. K., Lambright, C. S., Ostby, J. S., Parks-Saldutti, in male rat offspring following in utero exposure to phtha- L., Vandenbergh, J. G. & Gray, L. E. Jr (2007b) Prenatal late esters. International Journal of Andrology 29, 140–147; testosterone exposure permanently masculinizes anogenital discussion 181–145. distance, nipple development, and reproductive tract Gray, L. E. Jr, Ostby, J. S. & Kelce, W. R. (1994) Developmental morphology in female Sprague-Dawley rats. Toxicological effects of an environmental antiandrogen: the fungicide Sciences 96, 335–345. vinclozolin alters sex differentiation of the male rat. Howdeshell, K. L., Furr, J., Lambright, C. R., Rider, C. V., Toxicology and Applied Pharmacology 129, 46–52. Wilson, V. S. & Gray, L. E. Jr (2007) Cumulative effects of Gray, L. E. Jr, Wolf, C. J., Lambright, C., Mann, P., Cooper, dibutyl phthalate and diethylhexyl phthalate on male rat R. L. & Ostby, J. (1999a) Administration of potentially an- reproductive tract development: altered fetal steroid tiandrogenic pesticides (procymidone, linuron, iprodione, hormones and genes. Toxicological Sciences 99, 190–202. chlozolinate, p,p¢-DDE, and ketoconazole) and toxic sub- Jegou, B., Soto, A., Sundlof, S., Stephany, R., Meyer, H. & stances (dibutyl- and diethylhexyl phthalate, pcb 169 and Leffers, H. (2001) Existing guidlines for the use of meat ethane dimethane sulphonate) during sexual differentiation hormones and other food additives in Europe and USA. produces diverse proﬁles of reproductive malformations in APMIS Supplementum 103, S551–556. the male rat. Toxicology and Industrial Health 15, 94–118. Kelce, W. R. & Wilson, E. M. (1997) Environmental antian- Gray, L. E. Jr, Ostby, J., Monosson, E. & Kelce, W. R. (1999b) drogens: developmental effects, molecular mechanisms, and Environmental antiandrogens: low doses of the fungicide clinical implications. Journal of Molecular Medicine 75, 198– vinclozolin alter sexual differentiation of the male rat. 207. Toxicology and Industrial Health 15, 48–64. Kelce, W. R., Stone, C. R., Laws, S. C., Gray, L. E. Jr, Gray, L. E. Jr, Ostby, J., Furr, J., Wolf, C. J., Lambright, C., Kemppainen, J. A. & Wilson, E. M. (1995) Persistent ddt Parks, L. G. et al. (2001) Effects of environmental antian- metabolite p,p’-DDE is a potent androgen receptor antago- drogens on reproductive development in experimental nist. Nature 375, 581–585. animals. Human Reproduction Update 7, 248–264. Lambright, C., Ostby, J., Bobseine, K., Wilson, V., Hotchkiss, Gray, L. E. Jr, Ostby, J., Furr, J., Lambright, C., Hotchkiss, A. & A. K., Mann, P. C. & Gray, L. E. Jr (2000) Cellular and Wilson, V. S. (2004a) Cumulative effects of endocrine molecular mechanisms of action of linuron: an antiandro- disrupters (edcs): synergy or additivity? The Toxicologist 78, 74. genic herbicide that produces reproductive malformations in Gray, L. E. Jr, Ostby, J., Furr, J., Wolf, C., Lambright, C., male rats. Toxicological Sciences 56, 389–399. Wilson, V. & Noriega, N. (2004b) Toxicant-induced hypo- McIntyre, B. S., Barlow, N. J. & Foster, P. M. (2002a) Male spadias in the male rat. Advances in Experimental Medicine rats exposed to linuron in utero exhibit permanent changes and Biology 545, 217–241. in anogenital distance, nipple retention, and epididymal Gray, L. E. Jr, Wilson, V. S., Stoker, T., Lambright, C., Furr, J., malformations that result in subsequent testicular atrophy. Noriega, N., Howdeshell, K., Ankley, G. T. & Guillette, L. Toxicological Sciences, 65, 62–70.

Journal compilation ª 2008 Blackwell Publishing Ltd • International Journal of Andrology 31, 178–187 184 No claim to original US government works V. S. Wilson et al. Diverse mechanisms of anti-androgen action

McIntyre, B. S., Barlow, N. J., Sar, M., Wallace, D. G. & Decrease in anogenital distance among male infants with Foster, P. M. (2002b) Effects of in utero linuron exposure prenatal phthalate exposure. Environmental Health on rat wolffian duct development. Reproductive Toxicology Perspectives 113, 1056–1061. 16, 131–139. Thornton, J. & Goy, R. W. (1986) Female-typical sexual Meaney, M. J. & McEwen, B. S. (1986) Testosterone implants behavior of rhesus and defeminization by androgens given into the amygdala during the neonatal period masculinize the prenatally. Hormones and Behavior 20, 129–147. social play of juvenile female rats. Brain Research 398, 324–328. Vandenbergh, J. G. & Huggett, C. L. (1995) The anogenital Meaney, M. J., Stewart, J., Poulin, P. & McEwen, B. S. (1983) distance index, a predictor of the intrauterine position Sexual differentiation of social play in rat pups is mediated effects on reproduction in female house mice. Laboratory by the neonatal androgen-receptor system. Neuroendocrinol- Animal Science 45, 567–573. ogy 37, 85–90. Vinggaard, A. M., Nellemann, C., Dalgaard, M., Jorgensen, E. Noriega, N. C., Ostby, J., Lambright, C., Wilson, V. S. & Gray, B. & Andersen, H. R. (2002) Antiandrogenic effects in vitro L. E. Jr (2005) Late gestational exposure to the fungicide and in vivo of the fungicide prochloraz. Toxicological Sciences prochloraz delays the onset of parturition and causes repro- 69, 344–353. ductive malformations in male but not female rat offspring. Vinggaard, A. M., Christtiensen, S., Laier, P., Poulsen, M. E., Biology of Reproduction 72, 1324–1335. Breinholt, V., Jarfelt, K., Jacobsen, H., Dalgaard, M., Ostby, J., Kelce, W. R., Lambright, C., Wolf, C. J., Mann, P. & Nellemann, C. & Hasss, U. (2005) Perinatal exposure to the Gray, L. E. Jr (1999) The fungicide procymidone alters fungicide prochloraz feminizes the male rat offspring. sexual differentiation in the male rat by acting as an andro- Toxicological Sciences 85, 886–897. gen-receptor antagonist in vivo and in vitro. Toxicology and Wilson, V. S., Cardon, M. C., Thornton, J., Korte, J. J., Ankley, Industrial Health 15, 80–93. G. T., Welch, J., Gray, L. E. Jr & Hartig, P. C. (2004a) Clon- Parks, L. G., Ostby, J. S., Lambright, C. R., Abbott, B. D., ing and in vitro expression and characterization of the Klinefelter, G. R., Barlow, N. J. & Gray, L. E. Jr (2000) The androgen receptor and isolation of estrogen receptor alpha plasticizer diethylhexyl phthalate induces malformations by from the fathead minnow (pimephales promelas). Environ- decreasing fetal testosterone synthesis during sexual mental Science & Technology 38, 6314–6321. differentiation in the male rat. Toxicological Sciences 58, Wilson, V. S., Lambright, C., Furr, J., Ostby, J., Wood, C., 339–349. Held, G. & Gray, L. E. Jr (2004b) Phthalate ester-induced Pellis, S. M. & Pellis, V. C. (1997) The prejuvenile onset of gubernacular lesions are associated with reduced insl3 gene play fighting in laboratory rats (rattus norvegicus). expression in the fetal rat testis. Toxicology Letters 146, Developmental Psychobiology 31, 193–205. 207–215. Pomerantz, S. M., Roy, M. M., Thornton, J. E. & Goy, R. W. Wilson, V. S., Howdeshell, K. L., Lambright, C. S., Furr, J. & (1985) Expression of adult female patterns of sexual Earl Gray, L. Jr (2007) Differential expression of the phtha- behavior by male, female, and pseudohermaphroditic female late syndrome in male sprague-dawley and wistar rats after rhesus monkeys. Biology of Reproduction 33, 878–889. in utero dehp exposure. Toxicology Letters 170, 177–184. Rider, C. V., Furr, J., Wilson, V. S. & Grauy, L. E. Jr (2008) A mixture of seven antiandrogens induces reproductive malfor- Panel discussion mations in rats. International Journal of Andrology, doi: 10.1111/j.1365-2605.2007.00859.x A. Kortenkamp Schardein, J. L. (1993) Chemically Induced Birth Defects, pp You have given a specific definition of antiandrogenicity 281–358. Dekker, New York. and described a clear pattern of effects caused by andro- Sharpe, R. M. (2001) Hormones and testis development and gen receptor (AR) antagonism. Not all compounds which the possible adverse effects of environmental chemicals. are AR antagonists in vitro produce in vivo effects such as Toxicology Letters 120, 221–232. some of the azole fungicides which are antiandrogens but Skakkebaek, N. E. (2002a) Carcinoma in situ of the testis: have no effect on genital tract development. Can this lack Frequency and relationship to invasive germ cell tumours in of in vitro and in vivo correlation be explained by toxico- infertile men. Histopathology 41, 2. kinetics effects? Skakkebaek, N. E. (2002b) Endocrine disrupters and testicular dysgenesis syndrome. Hormone Research 57(Suppl. 2), 43. Stoker, T. E., Cooper, R. L., Lambright, C. S., Wilson, V. S., V. Wilson Furr, J. & Gray, L. E. (2005) In vivo and in vitro anti-androgenic effects of de-71, a commercial polybrominated diphe- Our laboratory performs in vitro assays and results are nyl ether (pbde) mixture. Toxicology and Applied not always predictive of in vivo effects. I am hesitant to Pharmacology 207, 78–88. speculate about the reasons for this discrepancy. Some Swan, S., Main, K., Liu, F., Stewart, S., Kruse, R., Calafat, A. AR antagonist compounds have mixed actions and it is et al. (2005) Study for future families research team. not always clear how the different actions interact.

Journal compilation ª 2008 Blackwell Publishing Ltd • International Journal of Andrology 31, 178–187 No claim to original US government works 185 Diverse mechanisms of anti-androgen action V. S. Wilson et al.

Toxicol., in press 2008) describing the effect of 400 chemi- A. Soto cals tested in the androgen receptor (AR) gene assay. This Some antiandrogens bind to the oestrogen receptor (ER) enabled the construction of a computer model for predict- including some phthalates such as butylbenzyl-, dibutyl- ing the activity of different molecules. Many compounds and the metabolites of methoxychoro- phthalate. What is have been assessed by this model and 7% of all substances the effect of fetal exposure to these compounds with oes- assessed are predicted to have antiandrogenic effects. trogenic and antiandrogenic activity: do you see an oes- trogenic effect in the fetuses or masculinisation of the R. Sharpe brain? Following bisphenol A (BPA) exposure, females act like males. Normally we view strain differences in laboratory animals as a hindrance and an annoying complication which con- tributes to variability in various responses. However, the V. Wilson differences in the effects of DEHP in Wistar and Sprague- We have not seen any abnormalities in sex development Dawley rats are perhaps an opportunity to study this as a or behavioural effects on female pups exposed in utero. model for geographic differences in human conditions. We look specifically for sex differentiation and androgenic For example, the differences in the disorders of the testi- effects. cular dysgenesis syndrome (TDS) between Denmark and Finland may reflect similar genetic differences in suscept- ibility to environmental factors. E. Gray Phthalates have no effect when administered to pubertal J. Spearow female rats and have no effect on oestrogen-dependent traits in the female. Not much work has been done on Wistar rats and Sprague-Dawley rats have both been the differences in female sexual behaviour. Phthalates did selected for large litter size which may have an endocrine not reduce mounting behaviour in male rats, but there is basis and is likely to have an effect on steroidogenic func- little evidence of masculinisation of female rats. tion. Have you considered looking at a strain of Wistar rats with smaller litter sizes as a comparison. V. Wilson V. Wilson We have no further information on these differences, but this discrepancy is a potential model for examining the The strains of rats which we use at present are satisfactory mechanism of INSL-3 activation and testosterone bio- for our mechanistic studies. synthesis. R. Ivell P. Foster Is the strain difference between Sprague-Dawley and Wis- We have heard at this Workshop that not all phthalates tar rats merely a quantitative effect, or are there differ- have adverse effects in rodents. Studies in the last century ences in the entire Leydig cell function in the 2 strains demonstrated a structure ⁄ activity relationship of phtha- using discriminating pathways for testosterone synthesis lates administered to pubertal rats, and the specific com- and INSL-3 transcription? pounds which were active in these rats were also effective in producing developmental abnormalities in fetal rats, V. Wilson affecting particularly the fetal testis. It has been found that the active phthalate must have a specific configura- We have no further information on these differences, but tion: it must be ortho and have a specific chain length this discrepancy is a potential model for examining the usually C4–C8, irrespective of the number of branching mechanism of INSL-3 activation and testosterone activa- side chains. tion.

C. Nellemann H. Patisaul A paper will shortly be published (Vinggaard et al. Screen- Phthalates have different effects in rats and mice. Have ing of 397 chemicals and development of a QSAR model any studies been performed on other non rodent mam- for androgen receptor antagonism of chemicals. Chem. Res. mals such as goats or sheep or non human primates?

Journal compilation ª 2008 Blackwell Publishing Ltd • International Journal of Andrology 31, 178–187 186 No claim to original US government works V. S. Wilson et al. Diverse mechanisms of anti-androgen action

However, it suggests that there are reproductive anoma- R. Sharpe lies produced in non human primates which might be We have studied marmosets that have been exposed to analogous to the effects seen in rats. high doses of monobutyl phthalate (MBP). When a single large oral dose is administered to neonatal male marmo- G. Bittner sets during the period of ‘‘mini puberty’’, there is a sig- niﬁcant reduction in testosterone levels in blood but of There are phthalates in plastics, cosmetics and other con- lesser magnitudes than seen in fetal rats after DBP expo- sumer products which have been shown not to bind to sure. We have ongoing studies in which MBP has been oestrogen receptor (ER) or androgen receptor (AR), but administered to pregnant marmosets during an 8 week in common usage may be exposed to UV light, heat or window: at birth, MBP exposed males show evidence of a other factors which might alter the chemistry. There is a delay in germ cell differentiation compared with controls, possibility of splitting the long chains which prevent AR which is also found in DBP ⁄ MBP exposed rats. This is and ER binding, or hydroxyl groups might be added to preliminary data and the possible long term effects give a product with hormone activity. in adults will not be assessable for another 15 months.

Journal compilation ª 2008 Blackwell Publishing Ltd • International Journal of Andrology 31, 178–187 No claim to original US government works 187