Perinatally Administered Bisphenol A as a Potential Mammary Gland Carcinogen in Rats

Background: Environmental exposure to bisphenol A (BPA) affects mammary gland development in rodents and primates. Prenatal exposure to environmentally relevant doses of BPA increased the number of intraductal hyperplasias and ductal carcinomas in situ by 50 days of age in Wistar-Furth rats. Objective: We aimed to determine whether BPA exposure of dams during gestation only or throughout lactation affects the incidence of mammary gland neoplasia in female offspring. Methods: We treated pregnant Sprague-Dawley rats with BPA at 0, 0.25, 2.5, 25, or 250 μg BPA/kg BW/day from gestational day (GD) 9 to birth and from GD9 to postnatal day (PND) 21. Mammary glands from BPA-exposed offspring were examined at four time points for preneoplastic and neoplastic lesions. To assess circulating BPA levels, we exposed pregnant rats to vehicle or 250 μg BPA/kg BW/day during gestation only or during gestation/lactation and analyzed sera from dams, fetuses, and nursing pups for total and unconjugated BPA. Results: Total and unconjugated BPA were detected in sera from 100% of dams and fetuses and 33% of pups exposed to 250 μg BPA/kg BW/day. Unconjugated BPA levels in exposed dams and fetuses (gestational) and in exposed dams and pups (gestational/lactational) were within levels found in humans. Preneoplastic lesions developed in BPA-exposed female offspring across all doses as early as PND50. Unexpectedly, mammary gland adenocarcinomas developed in BPA-exposed offspring by PND90. Conclusions: Our findings suggest that developmental exposure to environmentally relevant levels of BPA during gestation and lactation induces mammary gland neoplasms in the absence of any additional carcinogenic treatment. Thus, BPA may act as a complete mammary gland carcinogen. Citation: Acevedo N, Davis B, Schaeberle CM, Sonnenschein C, Soto AM. 2013. Perinatally administered bisphenol A acts as a mammary gland carcinogen in rats. Environ Health Perspect 121:1040–1046; http://dx.doi.org/10.1289/ehp.1306734


Introduction
Cumulative exposure to ovarian steroids during a woman's lifetime represents the most well-defined risk factor for the develop ment of breast cancer. Epidemiological studies have suggested that increased estrogen levels in the fetal environment are associated with an increased risk of breast cancer during adult life (Braun et al. 1995;Ekbom et al. 1992;Potischman and Troisi 1999). The synthetic estrogen diethylstilbestrol (DES), prescribed from the 1940s to the 1970s to prevent miscarriage, is recognized as a semi nal example of a human transplacental carcinogen for the multitude of adverse effects manifested in adult offspring. The effects of DES include increased risk of vaginal clear cell carcinoma, reproduc tive tract malformations, poor pregnancy outcomes, and compromised immune systems (Herbst et al. 1971;Hoover et al. 2011), as well as increased risk of developing breast cancer after 40 years of age (Hoover et al. 2011;Troisi et al. 2007). Another synthetic estrogen, bisphenol A (BPA), is currently one of the highest volume chemicals produced world wide, with a global production capacity of 11.5 billion pounds in 2008 (Burridge 2008;vom Saal et al. 2007). BPA is used in the production of poly carbonate plastics, epoxy resins, dental sealants and composites, and thermal receipt paper. Incomplete polymerization of BPA leads to leaching of the chemical and subsequent human exposure, as evidenced by the detection of BPA in human urine, serum, maternal and fetal plasma, amniotic fluid, placenta, and adipose tissue (Calafat et al. 2005;Fernandez et al. 2007;Ikezuki et al. 2002;Schönfelder et al. 2002;Vandenberg et al. 2007a;Zalko et al. 2011). Although oral exposure through ingestion of food and beverages was considered the main route in humans, recent studies have indicated that humans are also exposed to BPA through inhalation as well as via absorption through the skin and mucosal membranes of the mouth, and that these routes are not negligible ). The U.S. Environmental Protection Agency (EPA) has calculated an oral reference dose for BPA of 50 μg BPA/kg body weight (BW)/ day based on a lowest observed adverse effect level of 50 mg/kg BW/day (U.S. EPA 1993; Welshons et al. 2003). A review of more than two dozen biomonitoring studies that used analytical chemistry methods to measure BPA in healthy adults reported the detection of mean unconjugated BPA levels in the range of 1 ng/mL in blood (Vandenberg et al. 2010). The distinction between unconjugated and conjugated BPA is especially important in blood because the unconjugated form is considered the active form and has estrogenic activity (Thomas and Dong 2006;Watson et al. 2005). Recent pharmaco kinetic analy ses in non human primates suggested that daily oral exposure to 400 μg BPA/kg BW/day is sufficient to produce serum concentra tions of unconjugated BPA in the range measured in humans (Taylor et al. 2011) and that prenatal exposure to this dose altered the developing mammary glands of female rhesus monkeys (Tharp et al. 2012).
Several studies have reported that fetal exposure to low doses of BPA altered the development of the rodent mammary gland, which manifested from the time of exposure and was exacerbated at puberty and beyond (Markey et al. 2001;Munoz de Toro et al. 2005;Vandenberg et al. 2007bVandenberg et al. , 2008. Perinatal exposure to BPA increased estrogen and progesterone sensitivity in the mouse mammary gland (Ayyanan et al. 2011;Wadia et al. 2007). Together, these results suggest that peri natal exposure to BPA may increase the propensity to breast carcino genesis. In a study supporting this assessment, rats exposed prenatally to 2.5 μg BPA/kg BW/day showed a significant increase in the number of mammary gland intra ductal hyper plasias at postnatal day (PND) 50 and PND90 compared with controls; exposure to 250 or 1,000 μg BPA/kg BW/day resulted in the develop ment of ductal carcinomas in situ (DCIS) (Murray Background: Environmental exposure to bisphenol A (BPA) affects mammary gland development in rodents and primates. Prenatal exposure to environmentally relevant doses of BPA increased the number of intraductal hyperplasias and ductal carcinomas in situ by 50 days of age in Wistar-Furth rats. oBjective: We aimed to determine whether BPA exposure of dams during gestation only or throughout lactation affects the incidence of mammary gland neoplasia in female offspring. Methods: We treated pregnant Sprague-Dawley rats with BPA at 0, 0.25, 2.5, 25, or 250 μg BPA/kg BW/day from gestational day (GD) 9 to birth and from GD9 to postnatal day (PND) 21. Mammary glands from BPA-exposed offspring were examined at four time points for pre neoplastic and neo plastic lesions. To assess circulating BPA levels, we exposed pregnant rats to vehicle or 250 μg BPA/kg BW/day during gestation only or during gestation/lactation and analyzed sera from dams, fetuses, and nursing pups for total and unconjugated BPA. results: Total and unconjugated BPA were detected in sera from 100% of dams and fetuses and 33% of pups exposed to 250 μg BPA/kg BW/day. Unconjugated BPA levels in exposed dams and fetuses (gestational) and in exposed dams and pups (gestational/lactational) were within levels found in humans. Preneoplastic lesions developed in BPA-exposed female offspring across all doses as early as PND50. Unexpectedly, mammary gland adenocarcinomas developed in BPA-exposed offspring by PND90. conclusions: Our findings suggest that developmental exposure to environmentally relevant levels of BPA during gestation and lactation induces mammary gland neoplasms in the absence of any additional carcinogenic treatment. Thus, BPA may act as a complete mammary gland carcinogen.  Durando et al. (2007) reported that rats exposed prenatally to 25 μg BPA/kg BW/day displayed a higher number of ductal hyper plasias associated with desmo plasia in adult hood. These authors also found that nitroso methyl urea adminis tered at 50 days of age at doses that fail to induce tumors in control animals elicited the develop ment of mammary carcinomas in females peri natally exposed to BPA. In rats or mice exposed perinatally to BPA, administration of the chemical carcinogen dimethyl benz anthracene (DMBA) during adulthood resulted in increased tumor incidence and decreased tumor latency compared with animals exposed to DMBA alone (Jenkins et al. 2009;Weber Lozada and Keri 2011); BPA also shifted the window of susceptibility to DMBA (Betancourt et al. 2010). In addition, lactational exposure to BPA between PND2 and PND20 increased tumor incidence in conjunction with a DMBA challenge (Jenkins et al. 2009).
As a follow-up to our previous work in Wistar-Furth rats ), we examined the effect of duration of BPA exposure over a wide range of concentrations on the induction of preneoplastic lesions and DCIS in Sprague-Dawley rats, a strain used extensively for toxicology and carcino genesis studies by the National Toxicology Program. To relate this evidence to human biomonitoring data, we measured the internal levels of BPA in serum of these rats. Unexpectedly, at scheduled sacrifice times, we observed large mammary carcinomas (> 1 cm 2 in diameter) occurring at internal doses relevant to human exposure, suggesting that BPA may act as a complete carcinogen.

Materials and Methods
Animals. Sexually mature virgin female Sprague-Dawley rats (8-10 weeks of age; Taconic, Germantown, NY) were maintained in temperature-and light-controlled (14 hr/10 hr light/dark) conditions in the Tufts University School of Medicine Division of Laboratory Animal Medicine. Experimental procedures were approved by the Tufts University-Tufts Medical Center Animal Research Committee, and all animals were treated humanely and with regard for alleviation of pain in accordance with the Guide for the Care and Use of Laboratory Animals (Institute of Laboratory Animal Resources, National Research Council 2011). Cages, water bottles, and bedding tested negligible for estrogenicity by the E-SCREEN assay (Soto et al. 1992). Food (Harlan Teklad 2018 Rodent Diet, Harlan Teklad, Indianapolis, IN) was supplied ad libitum. Estrogenicity of the feed was measured at 8-15 fmol of estrogen equivalents per gram, a negligible amount (Soto et al. 1992). Female rats were mated with Sprague-Dawley males. The morning on which sperm was observed in vaginal smears was designated gestational day (GD) 1.
Fetal and neonatal exposure to BPA. To explore its role on neoplastic develop ment, we adminis tered BPA sub cutaneously via Alzet osmotic pumps (Durect Corp., Cupertino, CA), with the dose calculated based on the weight of the dam at day 7 of pregnancy. Dams (n = 9-12/dose/exposure period) were implanted with pumps on day 9 of pregnancy to adminis ter vehicle (50% dimethyl sulfoxide; Sigma Chemical Co., St. Louis, MO) or 0.25, 2.5, 25, or 250 μg BPA/kg BW/day. For convenience, these doses are subsequently referred to as BPA0.25, BPA2.5, BPA25, and BPA250, respectively.
We examined two different exposure periods. For animals exposed only through gestation, dams were implanted with pumps (catalog no. 2002; Durect Corp.) designed to deliver continuously up to 14 days (see Supplemental Material, Figure S1A). Pumps were implanted after 24 hr equilibration, according to the manufacturer's specifications. For animals exposed through gestation and lactation, dams were implanted with pumps (catalog no. 2004; Durect Corp) designed to deliver continuously up to 28 days (see Supplemental Material, Figure S1B). These pumps were implanted after 48 hr of equilibration. In both groups, animals delivered normally and litters were culled to 10 individuals on PND2. We distributed all female offspring (n = 9-12/dose/ age at sacrifice per exposure group) so that each litter was represented only once. We harvested mammary gland tissue at PND50, PND90, PND140, and PND200. The fourth left inguinal mammary gland was fixed and processed for paraffin embedding, and the contra lateral gland was whole mounted and stained with carmine as previously described .
Histological and whole mount analysis. Three 5-μm sections separated by 50 μm were used to assess the presence of pre neoplastic and neoplastic lesions in mammary glands of PND50 females, with five animals sampled per dose per exposure group. We visualized histological sections with an Axioskop 2 Plus microscope and captured images with an AxioCAm HR color digital camera and Axiovision software, version 4.5 (all from Carl Zeiss, München-Hallbergmoos, Germany). We assessed the incidence of total ductal hyperplasia as previously described ). Briefly, the leading edge and terminal end buds (TEBs) were localized; a 4-mm 2 box was drawn starting 400 μm from the most proximal TEB; and all of the ducts within this area were counted. Usual intra ductal hyperplasia (UDH) was charac terized by an increase in mono morphic ductular epithelial cell layers (≥ 3 cells thick) typified by cuboidal or columnar pseudo stratified epithelial cells that maintained a perpendicu lar orientation around the basement membrane. Preneoplastic lesions (atypical ductal hyper plasia; ADH) were charac terized by an increase in mono morphic to mildly pleo morphic ductular epithelial cell layers, typified by flattened cuboidal epithelial cells that more intensely exhibited eosinophilic cyto plasmic staining and/or slightly enlarged hyper chromatic or vesicular nuclei. Neoplastic lesions (DCIS) were distinguished from ADH by increased layers of dis organized, pleo morphic ductular epithelial cells bridging and occluding the lumen but main taining an intact basement membrane (Davis and Fenton 2013).
Quantification of circulating levels of BPA. We conducted a pilot experiment to determine the levels of detectability of BPA in the serum of pregnant dams. Based on dam weight at GD7, pregnant rats (n = 6/ dose) were implanted with pumps on GD9 to deliver vehicle, BPA25, or BPA250, and blood was collected by cardiac puncture 72 hr later. Samples were frozen and shipped to the Centers for Disease Control and Prevention, where serum samples were analyzed for total and unconjugated BPA as described by Ye et al. (2008). Briefly, serum was either treated with β-glucuronidase/sulfatase to estimate the concentration of total BPA (conjugated plus unconjugated), or processed without enzymatic treatment to estimate the concentration of unconjugated BPA. Then, serum concentrations were quantified using on-line solid phase extraction coupled to high performance liquid chromatography-isotope dilution tandem mass spectrometry. The limit of detection (LOD), the lowest amount of an analyte that can be detected with a defined probability, was 0.3 ng/mL. The limit of quantification (LOQ), calculated as three times the LOD, defines the point at which data attain statistical significance.
On the basis of these results, we conducted a second experi ment using either vehicle or volume 121 | number 9 | September 2013 • Environmental Health Perspectives BPA250 and divided the pregnant animals into two groups. One group (corresponding to the gestational-only exposure group) was designed to measure fetal exposure near the end of gestation to ensure the collection of a sufficient volume of fetal serum. Dams were weighed at GD17 to calculate the BPA dose to be adminis tered, and animals were implanted at GD18 with pumps (catalog no. 2002). On GD21 animals were killed, serum was collected from each dam and from fetuses; fetal sera from each litter were pooled (see Supplemental Material, Figure S1C).
The second group (corresponding to the gestational/lactational exposure group) was designed to measure BPA levels in dams and pups at PND10. Dams were weighed at GD7 and implanted with pumps (catalog no. 2004) on GD9. Animals delivered normally and litters were culled to 10 on PND2. Dams and pups were killed on PND10 and the serum collected (see Supplemental Material, Figure S1D); serum from the pups was pooled by litter. Serum samples were analyzed for total and unconjugated serum BPA at the Centers for Disease Control and Prevention as described above.
Statistical analyses. All calculated parameters and statistical significance were determined using SPSS statistical software (IBM SPSS Statistics 20; IBM, Chicago, IL). For the statistical calculations involving total and unconjugated BPA concentrations, we used the instrument-generated values, even if they were below the LOD, to run Student's t-tests. All results are presented as mean ± SD. Overall differences in preneoplastic lesions were analyzed by analysis of variance (ANOVA), and differences in tumor incidence were analyzed via the chi-square test. For all statistical tests, results were considered significant at p < 0.05.

BPA in serum of dams and their offspring following continuous exposure to BPA.
In the serum of dams, fetuses, and pups, BPA was present mainly in its conjugated form. Total and unconjugated BPA serum concentrations were detectable in 100% of pregnant dams and fetuses exposed gestationally to BPA250. The mean serum concentra tion of total BPA was signifi cantly higher in exposed fetuses compared with their vehicle controls, and the difference between mean unconjugated BPA levels between exposed fetuses and their controls approached significance (p = 0.051; Table 1). Unconjugated BPA was significantly higher in exposed dams compared with controls. The average total BPA in the serum of BPAexposed fetuses was four-times greater than in the serum of the dams (p = 0.004). In the group exposed gestationally/lactationally, total and unconjugated serum BPA were detectable in 100% of BPA-treated lactating dams, but total BPA was detectable in only 33% of the BPA-exposed pups (Table 2). Serum concentrations of total and unconjugated BPA were significantly higher in exposed lactating dams than in controls (Table 2). Total serum BPA measured in exposed pups was significantly higher than in controls; however, all values were below the LOQ. The mean total BPA concentration in the exposed pups (0.38 ng/mL) was significantly lower than in their respective dams (16.50 ng/mL; p = 0.03). Because BPA is ubiquitously present in the environment, extensive measures were taken throughout the process, from blood sampling to the final assay, to reduce contamination. Detection of predominantly Abbreviations: NA, not applicable; T:U, total BPA:unconjugated BPA. Values are expressed as mean ± SD; n = 4-6 animals/ treatment group. a Number of animals with BPA measurements at or above the LOD (0.3 ng/mL; the lowest amount of BPA that can be detected with a defined probability). b All concentrations were at or below the LOQ (0.9 ng/mL; the lowest amount of BPA that can be quantitatively determined with suitable precision and accuracy). c One or more concentrations were at or below the LOQ (0.9 ng/mL). *p < 0.05 compared with vehicle control by Student's t-test. Pups 2/6 a 0.38 ± 0.26* 0.2-0.8 b 0/6 < LOD 0.0-0.1 b NA Abbreviations: NA, not applicable; T:U, total BPA:unconjugated BPA. Values are expressed as mean ± SD; n = 6 animals/ treatment group. a Number of animals with BPA measurements at or above the LOD (0.3 ng/mL; the lowest amount of BPA that can be detected with a defined probability). b All concentrations were at or below the LOQ (0.9 ng/mL; the lowest amount of BPA that can be quantitatively determined with suitable precision and accuracy). c One or more concentrations were at or below the LOQ (0.9 ng/mL). *p < 0.05 compared with vehicle control by Student's t-test. Figure 1. Representative photomicrographs of preneoplastic and neoplastic lesions observed in PND50 mammary glands from BPA-exposed rats. (A) UDH (usual intra ductal hyperplasia) from rat after gestational/lactational exposure to BPA25. (B,C) ADH (atypical ductal hyperplasia) from rats after gestational/lactational exposure to BPA25 and BPA2.5, respectively. (D) DCIS from rat after gestational/lactational exposure to BPA25. Bar = 50 μm.

B C D
conjugated versus unconjugated BPA in the serum of dams, fetuses, and pups suggests that external contamination, if it occurred, was not systematic or extensive.

Preneoplastic and neoplastic lesions developed in perinatally exposed mammary glands.
We analyzed histological sections at PND50 for incidence of preneoplastic and neoplastic lesions (Figure 1). The incidence of UDH was not significantly different between the BPA and vehicle controls for either exposure period (data not shown). However, histological assessment of glands for ADH or DCIS showed that the incidence of ADH among glands from females exposed to all doses of BPA during both exposure periods ranged from 0 to 60% compared with no incidence (0%) in the vehicle controls (Table 3). We diagnosed DCIS in a gland from one female rat exposed gestationally/lactationally to BPA25 ( Figure 1D).
We observed mammary gland lesions in whole mounts by PND90 (Figure 2). Four animals exhibited proliferative lesions in their mammary glands at PND90 and PND140 following either gestational-only or gestational/ lactational exposure (Table 4): one lesion, diagnosed as a benign micro fibro adenoma, in a mammary gland from a vehicle-exposed animal at PND140, and three lesions that were diagnosed as lobular alveolar hyperplasia in three different females, one each exposed to BPA0.25, BPA25, or BPA250.
Malignant tumors developed following perinatal exposure to BPA. We detected tumors at PND90, PND140, and PND200 in animals exposed to BPA across all doses and exposure times. A total of six mammary gland tumors were observed in females exposed perinatally to BPA at doses ranging from BPA0.25 to BPA250 (n = 230; Table 4). Five tumors were diagnosed histopathologically as adenocarcinomas and one was diagnosed as a benign fibro adenoma (Figure 3). No malignant tumors were detected in any vehicletreated control animals (n = 65).

Discussion
Developmental exposure to BPA has been shown to increase the propensity to mammary gland carcinogenesis in rodents (Betancourt et al. 2010;Durando et al. 2007;Jenkins et al. 2009;Weber Lozada and Keri 2011). In the present study, we observed that peri natal exposure to human-rele vant internal doses of BPA-in the absence of additional exposure to chemical carcinogens-was associated with the induction of malignant mammary gland tumors and other lesions in adult female rats. To correlate dose and effect, we assessed the internal BPA dose in exposed dams and their offspring. In the BPA250 gestational exposure group, we detected significantly higher levels of unconjugated BPA in both dams and fetuses than in controls; these high levels were within the range of the levels detected in human serum (Table 1). This validates the use of osmotic pumps as an effective route of BPA adminis tration to the fetus.
Remarkably, at GD21 following gestational exposure to BPA, the average total BPA measured in fetuses was four times greater than that measured in dams. Although UDPglucuronosyltransferase-2B1-the major liver Table 3. Preneoplastic and neoplastic lesions observed at PND50 in histological sections of mammary glands of BPA-exposed females.   . Ratios indicate the number of animals with the lesion or tumor relative to the total number of animals per dose per exposure period for PND90, PND140, and PND200 combined. Differences in tumor incidence following treatment were compared with vehicle by chi-square test and found not to be significant. a Benign. b Malignant. c Tumor.
volume 121 | number 9 | September 2013 • Environmental Health Perspectives enzyme responsible for conjugation (inactivation) of BPA via glucu ronidation-shows little to no activity in the fetal rat liver (Yokota et al. 1999), the concentrations of glucuronidated BPA in the placenta and fetus have been shown to be higher than in maternal blood following administration of BPA to dams (Takahashi and Oishi 2000;Zalko et al. 2003). Total and unconjugated BPA concentrations measured in lactating dams at PND10 following gestational/lactational BPA exposure were higher than in their nursing pups (Table 2). In fact, unconjugated BPA was undetectable in all the pups examined, whereas total BPA was detected in 33% of the BPA-exposed pups. These results suggest that when a dam is continuously exposed to a constant dose of BPA during gestation and lactation, the neonate is exposed to lower levels of BPA than the fetus, probably because of the lack of transfer in milk. Doerge et al. (2010) reported comparable results, with serum concentrations of unconjugated BPA undetectable and total BPA approximately 300 times lower in suckling rat pups than in their dams; these authors attributed the low plasma levels observed in pups to low BPA intake from the mother's milk. Applying this relationship to our internal dose study, one may conclude that the pups from the BPA250 dams were lactationally exposed to 0.8 μg total BPA/kg BW/day, below the current U.S. EPA reference dose (U.S. EPA 1993). The increase in BPA levels in the dams at PND10 and the significant decrease in BPA in the pups compared with fetuses could also be attributed to maternal grooming practices, as suggested for other chemicals (White et al. 2007).
Oral administration of tritiated BPA has been shown to result in a linear relationship between the adminis tered dose and unconjugated serum BPA concentrations in both rodents and primates (Taylor et al. 2011). In the present study, the mean unconjugated serum BPA concentration in dams exposed either gestationally or gestationally/ lactationally to BPA250 was 1.25 ng/mL, even though the mean total BPA concentrations were almost three times higher in dams exposed gestationally/lactationally than only gestationally. Assuming linearity of circulating levels with dose, the concentrations of unconjugated BPA following exposure to BPA0.25, BPA 2.5, and BPA25 could be estimated around 0.00125, 0.0125, and 0.125 ng/mL, respectively, levels below the LOD of our method (0.3 ng/mL). However, at these undetectable levels, we have clearly seen effects in the mammary gland, both in the present data set and in our previous study .
Earlier studies provided evidence that perinatal exposure to low doses of BPA resulted in altered morpho genesis of the rodent mammary gland that first manifests during the exposure period. As the gland under goes further changes under the influence of ovarian and pituitary hormones, morphological alterations become more pronounced after the onset of puberty and throughout adulthood (Munoz de Toro et al. 2005;Vandenberg et al. 2007b). In the present study, nonneoplastic lesions diagnosed as lobular alveolar hyperplasia were observed in whole mounts of mammary glands at PND90 and PND140. These lesions have been associated with the adminis tration of xeno biotics that act as estrogen receptor (ER) agonists (Biegel et al. 1998) or dopaminergic receptor antagonists that cause prolactinemia in female rats (Lotz and Krause 1978;Lucas et al. 2007). A previous study showed that a single injection of the carcinogen N-nitroso methyl urea to post pubertal virgin female rats resulted in secretion of α-lactalbumin from lobular alveolar structures in the PND200 mammary gland (Murray et al. 2009). Therefore, the increased secretory material in the ducts and/or alveoli in the lesions observed in the present study may be due to xenobiotic-induced prolactinemia.
In this study, we examined whether exposure to environmentally relevant doses of BPA influenced the development of pre neoplastic and neo plastic lesions in Sprague-Dawley rats. Unlike our previous study in Wistar-Furth rats ), we observed no difference in the incidence of UDH in Sprague-Dawley females exposed to BPA compared with vehicle controls. This could be attributed to strain differences in the overall histoarchitecture of the mammary gland in the peripubertal phase (Fenton S, personal communication). Alternatively, the time course of development and regression of these lesions may be different in these strains. In the present study, both ADH and DCIS were identified only in glands of females exposed to BPA, at doses as low as BPA0.25.
To our knowledge, the induction of malignant tumor formation following develop mental exposure to environmentally relevant levels of BPA has not been previously reported. As an unexpected outcome of the present study, adenocarcinomas were identified in females as early as PND90 and at doses as low as BPA0.25. It is important to note that no vehicle-exposed control animals developed malignant tumors throughout the duration of the study. Historical data on the natural occurrence of neoplastic lesions in control female Sprague-Dawley rats from large carcino genicity trials established that there is evidence that spontaneous malignant tumors do not occur before PND210, and in fact most occurred after PND350 (Ikezaki et al. 2011;National Toxicology Program 2007Son and Gopinath 2004). Although the incidence of tumor development was not statistically significant, the highly adverse nature of carcinomas validates the biological importance of reporting this outcome, as this information warrants further studies. , and 200 (C,F) in gestationally and gestationally/lactationally exposed females. (A,B,D,E) Papillary adenocarcinoma from rats exposed gestationally to BPA250 (A) or BPA2.5 (D), or gestationallly/lactationally to BPA2.5 (B) or BPA25 (E). (C) Ulcerative adenocarcinoma from rat exposed gestationally to BPA0.25. (F) Fibroadenoma from rat exposed gestationally/lactationally to BPA2.5. Bar = 500 μm. Insets: Magnification at 10× base image to show detail of tumor morphology.  How BPA contributes to the initiation and progression of neoplasia is still unknown; however, perinatal exposure to low doses of BPA has been shown to affect the hypothalamic-pituitary-ovarian axis via a) altered develop ment of the hypothalamic nuclei essential for cyclic gonadotropin release (Rubin et al. 2006), b) disrupted estrous cyclicity of exposed individuals (Rubin et al. 2001), and c) increased sensitivity of the adult mammary gland to ovarian hormones (Ayyanan et al. 2011;Munoz de Toro et al. 2005). It is likely that BPA, like DES, may induce carcino genesis by acting as an estrogen. During fetal life, ERs α and β are present only in stromal cells; epithelial ER expression begins at the end of gestation (Vandenberg et al. 2007b). Wadia et al. (2013) reported that low-dose BPA exposure (250 ng/kg BW/ day) during mouse fetal development altered the composition and organization of the extra cellular matrix and accelerated maturation of the presumptive fat pad, an event necessary for ductal invasion and branching. This suggests that BPA acts directly on the stroma, which may in turn alter the development of the epithelium, as evidenced by increased ductal area and delayed lumen formation during the period of exposure (Vandenberg et al. 2007b). These data are compatible with the tissue organization field theory of carcinogenesis that posits that carcinogens alter the reciprocal inter actions between stroma and epithelium (Soto and Sonnenschein 2011), as shown by tissue recombination studies (Maffini et al. 2004).
According to the U.S. EPA, a carcinogen is a "chemical or physical agent capable of causing cancer" (U.S. EPA 2005), a definition that does not specify the mechanism(s) by which the cancer is induced; it only identifies the consequence of an insult. Thus, by this definition, and based on the data from the present study, BPA may act as a complete mammary gland carcinogen.

correction
The title and two summary statements were revised from the Advance Publication version of the manuscript to clarify the conclusions of the study.