Endocrinology. 2006 Oct;147(10):4871-82.
Neonatal exposure to the phytoestrogen genistein alters mammary gland growth and developmental programming of hormone receptor levels.
Padilla-Banks E, Jefferson WN, Newbold RR.
Developmental Endocrinology and Endocrine Disruptor Section, Laboratory of Molecular Toxicology, National Institute of Environmental Health Sciences, National Institutes of Health, Research Triangle Park, North Carolina, USA.
[ Free full text ]

Developmental effects of genistein (Gen) on the mammary gland were investigated using outbred female CD-1 mice treated neonatally on d 1-5 by sc injections at doses of 0.5, 5, or 50 mg/kg.d. Examination of mammary gland whole mounts (no. 4) before puberty (4 wk) revealed no morphological differences in development after Gen treatment. However, mice treated with Gen-50 had stunted development characterized by less branching at 5 wk and decreased numbers of terminal end buds at 5 and 6 wk. Conversely, at 6 wk, Gen-0.5-treated mice exhibited advanced development with increased ductal elongation compared with controls. Measurements of hormone receptor levels showed increased levels of progesterone receptor protein and estrogen receptor-beta mRNA in Gen-0.5-treated mice compared with controls; ERalpha expression was decreased after all doses of Gen treatment. Lactation ability, measured by pup weight gain and survival, was not affected after neonatal Gen-0.5 and Gen-5. Mice treated with Gen-50 did not deliver live pups; therefore, lactation ability could not be determined. Evaluation of mammary glands in aged mice (9 months) showed no differences between Gen-0.5-treated mice and controls but mice treated with Gen-5 and Gen-50 exhibited altered morphology including reduced lobular alveolar development, dilated ducts, and focal areas of "beaded" ducts lined with hyperplastic ductal epithelium. In summary, neonatal Gen exposure altered mammary gland growth and development as well as hormone receptor levels at all doses examined; higher doses of Gen led to permanent long-lasting morphological changes.

From the full text article:

EXPOSURE TO HIGH levels of estrogens during perinatal life has been suggested to be a risk factor for developing breast cancer later in life (1). For example, female twins developing in an elevated intrauterine estrogen environment have an increased incidence of breast cancer; this supports the idea that the fetal hormonal environment plays a role in predisposing women to breast cancer later in life (2, 3, 4). Also, increased levels of circulating estrogens usually found in older pregnant women have been associated with an increased breast cancer risk in their daughters (5, 6). Conversely, preeclampsia characterized by restricted estrogen production in the placenta and high maternal blood pressure correlates with a decreased incidence of breast cancer in daughters (4). Furthermore, epidemiological studies following a cohort of women exposed prenatally to the synthetic estrogen diethylstilbestrol (DES) have found an increased incidence of breast cancer (18%) as they age (7, 8). These studies provide evidence that early estrogen exposure is a risk factor for the subsequent development of breast cancer. These human data are supported by experimental animal data that demonstrate the proliferative and carcinogenic potential of estrogens in its target tissues including the mammary gland (9, 10, 11, 12).

Rodent models of developmental exposure to DES have been useful in reproducing and predicting abnormalities and tumors in humans prenatally exposed to DES (10, 11, 13, 14), and have shown that the mammary gland is extremely sensitive to DES (13, 15, 16). Prenatal or neonatal treatment of rats with DES increased the incidence of mammary lesions (hyperplastic alveolar nodules, dysplasia, and neoplasia) and decreased tumor latency (15). Neonatal treatment of mice with DES increased sensitivity to hormones and carcinogens later in life (17). Similar studies using BALB/c mice treated prenatally with DES showed that palpable mammary tumors developed in the presence of mouse mammary tumor virus (18), and in CD-1 mice mammary tumors developed even in the absence of this virus (19).

Over the last 25 yr, epidemiological studies have suggested the incidence of breast cancer in the United States is on the rise (20, 21, 22). Because only 50% of breast cancer can be explained by known risk factors such as age, nationality, family, and reproductive history, there is increasing interest in examining the role of developmental exposures to environmental toxicants and estrogen-mimicking chemicals in the etiology of this disease. Reports in animal models have suggested that exposure during development to environmental chemicals with estrogenic activity increases the risk of mammary cancer (23) as well as causes early puberty (24) and abnormal estrous cyclicity (25). Also, exposure to the estrogenic pesticide dieldrin has been shown to increase the risk of breast cancer in exposed women and decrease their survival rate in a dose-dependent manner (26, 27). Furthermore, in utero exposure of mice to low, environmentally relevant doses of bisphenol A caused abnormal histoarchitecture and altered DNA synthesis in mammary tissues, two changes that have been associated with carcinogenesis in both rodents and humans (28). Taken together, these data provide strong evidence that exposure to environmental chemicals with hormone activity during critical stages of development is associated with an increase risk of breast cancer later in life.

The isoflavone genistein (Gen) is a naturally occurring phytoestrogen found in soy products. Gen, generally in the glycosylated form in the diet, accounts for more than 65% of the isoflavones found in soy. Although Gen-rich foods and supplements have been proposed for use by menopausal women because of potential beneficial effects, adverse effects can occur if exposure occurs during development, particularly if exposure occurs during critical periods of differentiation of reproductive and breast tissues (29, 30, 31). Therefore, it is of concern that infants are exposed to Gen through soy-based formulas (32) and newly marketed soy-enriched products designed to appeal to children with the intention of providing them a healthy diet. Infants consuming soy-based formula are exposed to 6–11 mg/kg·d of isoflavones (4–7 mg/kg·d of total Gen) resulting in circulating levels of approximately 1–5 µM total Gen (33). In contrast, adults consuming a moderate to large amount of soy in the diet are exposed to approximately 1 mg/kg·d of total Gen resulting in circulating levels of approximately 0.5 µM total Gen (32). Although the form of Gen predominantly found in dietary sources is the glycosylated form, genistin, recent studies have shown that genistin is quickly hydrolyzed to the unconjugated form genistein in the gut and readily absorbed by infants because glucuronidated Gen metabolites and other metabolites were found in their urine (34). Although the typical route of exposure for humans is through the diet, a previous pharmacokinetic study from our laboratory using sc injections of Gen at a dose of 50 mg/kg·d produced serum circulating levels of 6.8 µM (in female pups), which is similar to human infants drinking soy-based infant formulas (1–5 µM). Furthermore, the percentage of circulating levels of the aglycone form of Gen (unconjugated, estrogenically active form) is much higher in neonates compared with adults (35, 36), making the developmental period especially sensitive to perturbation by estrogenic compounds like Gen. Thus, infants are exposed to much higher levels of Gen than adults, particularly because soy-based infant formulas can be their sole source of nutrition for many months. Unfortunately, there have been few epidemiology studies examining long-term effects of soy infant formula and soy-enriched diets for children.

Numerous experimental studies have examined developmental effects of Gen in estrogen target tissues (37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47). Several studies from our laboratory using outbred CD-1 mice have shown that neonatal exposure to Gen (0.5–50 mg/kg·d) by sc injection on d 1–5 caused deleterious effects on the female reproductive tract including altered ovarian differentiation, subfertility/infertility, and development of uterine cancer later in life (39, 45, 46, 47). Another study using rats exposed to Gen (40 mg/kg·d) neonatally by oral gavage showed similar adverse effects on the female reproductive system (48). Changes in sexually dimorphic areas of the brain, brain function, and altered reproductive behavior (44, 49, 50) have also been described. In humans, an epidemiological study showed an association with vegetarian diets during pregnancy and reproductive tract malformation (hypospadias) in the male offspring (51), and other studies indicate soy products have hormonal effects in adult women (52, 53, 54, 55, 56, 57).

The National Toxicology Program has evaluated the effects of dietary Gen in Sprague Dawley rats for multigenerational and chronic effects on reproductive tract tissues and other estrogen target tissues (16). Histopathological examination of the mammary gland of female pups from rat dams fed 250-1250 ppm Gen (20–100 mg/kg·d) showed ductal/alveolar hyperplasia. In male pups, the most sensitive target tissue for endocrine disruptor effects was the mammary gland, which showed significant ductal/alveolar hyperplasia and hypertrophy at 25 ppm of Gen (2 mg/kg·d). Chronic effects of Gen in the National Toxicology Program study suggested that Gen treatment decreased the incidence of benign mammary fibroadenomas but increased the incidence of mammary adenocarcinoma (Delclos, K. B., personal communication). A recent study showed similar effects on the male mammary gland after dietary Gen treatment (300 and 800 ppm) during pregnancy and lactation; furthermore, Gen increased the sensitivity of the mammary gland to the pesticide methoxychlor, a proestrogen (58). These data suggest that mammary gland development is particularly susceptible to disruption by estrogenic compounds if exposure occurs during critical stages of differentiation.

Other reports described the long-term effects on the mammary gland after developmental exposure to Gen (59, 60, 61, 62, 63). The timing of exposure, route, and dose of Gen appeared to be important factors in the overall effect on the mammary gland. For example, one study exposing rats to Gen (1–10 mg/kg·d) through the diet from gestation d 0 through postnatal d 21 (weaning) showed resistance to dimethylbenz[a]anthracene (DMBA)-induced mammary tumors later in life (59). There was also reduced terminal end bud (TEB) development as well as immature terminal ductal structures that remained until adulthood. In another study from the same laboratory, exposure to high levels of Gen (500 mg/kg·d) in the diet on d 16, 18, and 20 reduced mammary tumors after DMBA treatment in adulthood (62, 63). These studies show that Gen exposure during development of the mammary gland caused altered morphogenesis and abnormal mammary gland responses later in life to environmental insults. In addition, Hilakivi-Clarke’s laboratory reported similar findings when rats treated with Gen (1 mg/kg·d) by sc injection on d 7–20 showed reduced DMBA-induced mammary gland tumors as well as reduced terminal duct formation (61). However, in another study from Hilakivi-Clarke’s laboratory, rats treated prenatally with Gen (0.1, 0.5, or 1.5 mg/kg·d) on gestation d 15–20 had increased susceptibility of mammary gland tumors induced by DMBA (60). These data taken together showed that the timing of Gen exposure influences the development of the mammary gland and its subsequent response to carcinogens later in life such that prenatal exposure to Gen increased sensitivity to DMBA, a known mammary carcinogen.

[...]

This study shows that neonatal exposure to the phytoestrogen Gen alters mammary gland morphogenesis long after the time of exposure, despite the lack of obvious effects before puberty. This suggests developmental programming of the mammary gland was altered such that at later time points the tissue responds differently to hormonal cues than would be expected. This was evidenced by altered hormone receptor levels as well as advanced mammary gland morphogenesis during puberty after neonatal exposure to the lower doses of Gen treatment (0.5 and 5 mg/kg) and delayed mammary gland morphogenesis in the highest dose of Gen (50 mg/kg). Although the development of the mammary gland was altered during the time of puberty in the Gen-0.5 and Gen-5 treatment groups, the ultimate function of the mammary gland appeared to be intact because mice from these treatment groups at 2 months of age were able to lactate and maintain pup growth. This would be expected because the morphological changes observed in the low dose of Gen treatment would have most likely led to expansion of the mammary gland through the fat pad at a faster rate but not necessarily affecting the function of the mammary gland. However, at a later time point (9 months), altered mammary gland morphology was apparent in the Gen-5 and Gen-50 treatment groups with less branching and less alveolar development compared with controls. In addition, some of the mice in the Gen-5 and Gen-50 groups exhibited the presence of dilated beaded ducts lined with hyperplastic ductal epithelium, which were not seen in controls at any age examined. The presence of these structures has been previously reported after prenatal exposure to the xenoestrogen zearalenone, supporting the idea that developmental exposure to an estrogenic compound caused this effect (72). Whether the lack of tertiary differentiation later in life is because of a direct effect on the mammary gland or to a lack of cyclicity and associated low levels of progesterone, is not known.

The differential effects on the mammary gland at low vs. high doses is not unique to this compound or this tissue. Several studies have shown an effect at a low dose and an opposite effect at a higher dose. For example, mice treated with the same low dose of Gen used in this study (Gen-0.5) have increased ovulation rates compared with control mice, whereas mice treated with the high dose (Gen-50) have decreased rates (46). Another study from our laboratory showed enhanced uterine responsiveness after neonatal exposure to low doses of DES and dampened uterine responsiveness after high doses (64). Others have also shown nonlinear effects on reproductive tissues such as the prostate after developmental exposure to endocrine-disrupting compounds including DES, bisphenol A, methoxychlor, arochlor, and ethynylestradiol (65, 66, 67, 68). Although the exact mechanisms controlling these observations remain unknown, there is clear evidence that biphasic effects occur. One possible explanation is that there are two separate mechanisms involved with one effect being mediated elsewhere in the endocrine system, perhaps the hypothalamus or pituitary, and the other effect being mediated in the tissue itself. One study supports this idea in that developmental exposure to low doses of Gen led to hyperresponsiveness of the pituitary, whereas higher doses of Gen led to lack of responsiveness (73). Additional study on differential effects at low vs. high doses is needed to determine the mechanism(s) responsible for each effect.

Mammary gland development has been studied for decades, but the exact mechanisms that control growth and development remain to be elucidated. The mammary gland is a rudimentary structure at birth in both rodents and humans (74). Extension and ramification of the mammary gland through the fat pad begins before puberty and continues throughout puberty. Additional differentiation of the mammary gland occurs in adulthood, responding to hormonal cues during the estrous cycle and pregnancy. Hormones such as estradiol and progesterone are necessary for proper mammary gland development and differentiation. Studies using transgenic mouse models of hormone receptors provide clues to the contributions of each receptor type on the development and subsequent differentiation of the mammary gland (75, 76).

Progesterone is important in mammary gland differentiation, and there are two forms of the PR, PR-A and PR-B; each one mediates differential effects in the mammary gland as well as other reproductive tissues (77). Transgenic mice overexpressing PR-A and PR-B showed defects in postpubertal differentiation of the mammary gland with increased ductal branching and hyperplasia in the PR-A-overexpressing mice (78) and reduced ductal branching in PR-B-overexpressing mice (79). Neither of these models showed differences in prepubertal development of the mammary gland, which suggest that PR and progesterone signaling is not the primary mechanism involved in early prepubertal development (77). In addition, transgenic mice lacking PR-A had normal mammary gland development (80) and mice lacking PR-B had reduced ductal branching associated with pregnancy (81). The fact that overexpression of PR-B or lack of PR-B caused decreased ductal branching suggests that the level of PR-B is carefully regulated and that differences in expression level can cause alterations in mammary gland morphogenesis.

Estrogens have also been implicated in mammary gland development and differentiation. There are two transgenic mouse models lacking either one or the other of the two receptor subtypes responsible for estrogen signaling. Mice lacking ER (ERKO) maintain a rudimentary duct even as adults (82). The lack of mammary development in these mice was found to result from a nonfunctioning pituitary with insufficient production of prolactin and subsequent reduction of progesterone in those mice (82). In fact, estrogen and progesterone treatment along with a transplanted functioning pituitary recovered mammary gland development in ERKO mice. Interestingly, the formation of TEBs and ductal elongation occurred in ERKO mice in the presence of estrogen and in absence of prolactin, suggesting that early mammary gland morphogenesis is estrogen dependent but not mediated by ER (82). Mice lacking ERß (ßERKO) have a less dramatic effect on the mammary gland, but ERß may play a role in some aspects of mammary gland development (75). These mice have less ductal elongation as evidenced by incomplete penetration of the fat pad as well as less alveolar development, although specific studies during the time of development were not researched (75). In contrast, another laboratory showed that mammary gland development in ßERKO mice developed independently of ERß and that they have the ability to lactate and appear relatively normal during adulthood (82). However, ßERKO mice have much smaller litter sizes and may not need the full capacity of the mammary gland to sustain the litter. Early developmental studies on the mammary gland of the ßERKO mice were also not conducted to determine what role ERß might play in early mammary gland development.

The findings from the current study show differences in receptor expression levels after neonatal treatment with Gen long after exposure occurred. The most striking difference was the reduced levels of ER mRNA after Gen treatment at both 5 and 6 wk of age. This is in agreement with previous studies that have also shown similar reductions in ER expression after developmental exposure to Gen (61). The lower levels of ER expression in the mammary gland after the highest dose of Gen are accompanied by reduced TEBs and reduced branching. This reduced or delayed expansion through the fat pad is reminiscent of the ERKO mice, which lack TEBs and do not expand into the fat pad because of a nonfunctioning pituitary in these mice. In addition, these studies have shown that ductal elongation during prepuberty is dependent on estrogen, although its effects are not mediated by ER (82). Therefore, the direct consequence of lower ER expression in the mammary gland after developmental exposure to Gen remains unclear.

Another interesting finding is the increased expression of ERß mRNA in the lower dose of Gen treatment. However, this increase in ERß mRNA was not accompanied by increased ERß protein expression in the mammary ductal epithelium as determined by immunohistochemistry. The most likely explanation is that the immunohistochemical staining shows only that the ductal epithelial cells are positive for ERß and that the cells across treatments have similar intensities but not how many cells are actually staining positive throughout the entire mammary gland. The increase in mRNA may be more of a reflection of the higher percentage of mammary ductal epithelial cells present in the Gen-0.5-treated samples as opposed to increased ERß mRNA in individual cells. Because there is increased ductal elongation in the low dose, one would expect there to be more mammary gland ductal epithelium relative to the number of mammary stromal cells and fat cells. The increased proportion of mammary ductal epithelial cells could explain the increased levels of ERß mRNA. Therefore, the increased ERß mRNA in the mammary gland after the low dose of Gen treatment supports the finding of increased mammary gland structure in this treatment group during this period of development.

Another possible mechanism by which Gen could alter the development of the mammary gland is by disrupting the hypothalamic-pituitary-gonadal (HPG) axis. This would impact the secretory pathway of pituitary and ovarian hormones that are necessary for mammary gland morphogenesis and differentiation. This has been suggested previously by Faber and Hughes (73) showing that rats exposed neonatally to low doses of Gen (0.01 mg/kg) produced higher levels of LH in response to GnRH. Data from our laboratory support this idea because mice treated with lower doses of Gen exhibited enhanced ovulation rates after exogenous stimulation with gonadotropins, suggesting hyperresponsiveness of the pituitary (46). Faber and Hughes (73) also showed that rats exposed to higher doses of Gen during neonatal life were associated with decreased pituitary responsiveness. Because alterations in HPG signaling have been observed in similarly treated animals, the pituitary hormones involved in mammary gland morphogenesis may be disrupted as well. Additional studies in our laboratory are underway to further characterize the responsiveness of the hypothalamus and pituitary after developmental exposure to Gen.

The effects observed herein as a consequence of exposure to Gen is not unique because developmental exposure to other environmental chemicals also altered mammary gland development and differentiation. For example, exposure to dioxin prenatally inhibits mammary gland morphogenesis (83). Exposure to synthetic estrogenic or antiestrogenic chemicals such as DES, tamoxifen, or ICI during development also alter mammary gland development (12, 84). There are also several xenoestrogens that have been shown to alter mammary gland development including bisphenol A, zearalenone, and resveratrol (28, 72). In addition, developmental exposure to the growth factor TGF has been shown to advance mammary gland growth during puberty (12), similar to our low-dose Gen-treated mice. Furthermore, IGF-I and GH are also known to play a role in mammary gland growth and differentiation and are thought to work in concert with estradiol to complete mammary gland morphogenesis (85). Future studies will be needed to determine possible effects of neonatal exposure to genistein on growth factors and growth factor signaling in these mice.

In summary, developmental exposure to Gen at environmentally relevant doses alters murine mammary gland morphogenesis during puberty, despite the lack of obvious effects before puberty. These data suggest that the brief exposure to estrogenic substances during development alters programming such that subsequent exposure to hormones results in altered growth and/or differentiation. This idea is supported by the fact that hormone receptor levels in the mammary gland are altered after neonatal Gen treatment; however, altered endocrine signaling from the HPG axis may also be involved in altered mammary gland development. There are also long-term effects on the mammary gland, including ductal epithelial hyperplasia in the higher doses of Gen treatment. This effect appears to coincide with lack of estrous cyclicity or persistent estrus in these mice, but a direct effect on the mammary gland epithelial cells during the time of treatment might also explain this effect. Whether these effects are observed in humans exposed developmentally to these compounds remains to be determined.

Categories: 2006, Soy, Phytoestrogens, Sexual development, Endocrine, Estrogen, Progesterone, Pituitary, Cancer, Hormone disruptors, Nutrition and diet